Eur J Epidemiol (2009) 24:1–105 DOI 10.1007/s10654-009-9380-5
IEA-EEF European Congress of Epidemiology 2009 Epidemiology for Clinical Medicine and Public Health International Epidemiological Association was founded in 1954 by John Pemberton of Great Britain and Harold N Willard of the United States with the advice and help of the late Robert Cruickshank. The aims of the Association are to facilitate communication among those engaged in research and teaching in epidemiology throughout the world, and to engage in the development and use of epidemiological methods in all fields of health including social, community and preventive medicine and health services administration. The European Epidemiology Federation (EEF) is an official group within the IEA. It originated as the IEA European Group as ratified at the Council Meeting held in Japan in August 1996. All members of the IEA in Europe are automatically members of the EEF. Actions proposed or already taken by the European Epidemiology Federation include: • • •
Annual international IEA EEF meetings in Europe (equivalent to IEA Regional Meetings held elsewhere). Advocacy for epidemiological interests when issues of relevance for our research options are debated. Fostering closer relationships with WHO-EURO in Copenhagen to improve epidemiological expertise in all countries of Europe— and particularly those in Eastern Europe. • Improvement in training in epidemiology and public health in Eastern Europe. • Fostering closer associations between IEA and other European Epidemiological groups. • Organization of short-term training courses. IEA-EEF European Congress of Epidemiology 2009—EUROEPI2009 ‘‘Epidemiology for Clinical Medicine and Public Health’’ is the following annual IEA European Scientific Meeting. It refers to former European Meetings held—since IEA European Group ratification—in Bordeaux in 1998, Kaunas in 2000, Oxford in 2001, Toledo in 2003, Porto in 2004, Utrecht in 2006 and Cork in 2007. When World Congress of IEA takes place, European Meeting is not being organized. EUROEPI2009 is held in Warsaw, capital of Poland in the Old Library at Warsaw University—the largest university in Poland founded in 1816. The conference gathers about 250 general epidemiologists, clinical epidemiologists, clinicians, public health specialists, biostatisticians and representatives of other disciplines working for research both in clinical medicine and in public health from over 40 countries. The program covers all modern epidemiological disciplines, like cardioepidemiology, cancer epidemiology, perinatal and reproductive health epidemiology, infectious diseases epidemiology, genetic epidemiology, pharmacoepidemiology, environmental epidemiology, occupational epidemiology, social epidemiology and others. Thank you for your interest in coming to Warsaw, despite of the economic crisis and influenza pandemic which afflicted our planet this year. We wish you fruitful scientific discussions and nice stay in Warsaw! Susana Sans IEA Councillor for Europe Andrzej Zielin´ski Chairman of the Scientific Committee Katarzyna Szamotulska Chairman of the local Organizing Committee
123
2
IEA-EEF European Congress of Epidemiology 2009
Greetings from the Past President of the International Epidemiological Association Jorn Olsen, University of California, School of Public Health, Department of Epidemiology, Los Angeles, USA
Jorn Olsen is Professor and Chair of the Department of Epidemiology, School of Public Health, UCLA. He came to UCLA from Denmark in 2006 where he was head of the Danish Epidemiology Science Centre and was Principal Investigator of the Danish National Birth Cohort, a study among 100,000 pregnant women and their children. The aim of this study is to examine early determinants of health in a cohort that will be followed from conception to death. Most of Dr. Olsen s research has been in reproductive health but studies have also been done on cancer and other chronic diseases. Dr. Olsen has published more than 400 papers in peer-reviewed journals and his publications also include textbooks on epidemiology. He has been a member of the Medical Research Council in Denmark and serves as consultant for research councils in several countries. He serves on advisory committees for several research centers and on editorial boards for scientific journals and is Associate Editor of Human Reproduction. Past president of the International Epidemiological Association 2005–2008.
IEA-EEF European Congress of Epidemiology 2009
Key-note sessions Inauguration lecture Epidemic of Diabetes: Challenges and opportunities for epidemiologists Andrzej Krolewski, Harvard Medical School, Joslin Diabetes Center, Section on Genetics and Epidemiology, Boston, USA Dr. Andrzej Krolewski received his medical education at the Warsaw Medical University, Warsaw, Poland. After coming to the U.S. and completing fellowships in epidemiology at Harvard School of Public Health and molecular human genetics at Massachusetts Institute of Technology, he established a laboratory of Genetics and Epidemiology at the Joslin Diabetes Center. Capitalizing on his access to the 16,000 patient population of the Joslin Clinic, he developed research on the epidemiology and genetics of diabetes and its late complications. He has published over 200 papers in this area. His current research focuses on three topics: 1. Genetics and epidemiology of diabetes 2. Epidemiology and genetics of diabetic kidney complications 3. Application of genetic, genomic and proteomic methods to epidemiological studies For the last 20 years Dr. Krolewski’s laboratory has been a center for training of many post-doctoral fellows who use epidemiologic and genetic methods to study the etiology of diabetes and its complications. He pioneered this field by teaching courses on this topic for 15 years at the Harvard School of Public Health. Currently Dr. Krolewski is Senior Investigator and Head of laboratory on Genetic and Epidemiology at the Joslin Diabetes Center and a professor of medicine at Harvard Medical School, Boston, MA, USA.
Plenary lecture Cancer Survival in five continents (CONCORD study) Michel Coleman, London School of Hygiene and Tropical Medicine, London, UK Michel Coleman has been Professor of Epidemiology and Vital Statistics at the London School of Hygiene & Tropical Medicine since 1995. He was Deputy Chief Medical Statistician at the Office for National Statistics from 1995 to 2004 and Head of the Cancer and Public Health Unit at the School from 1998 to 2003. He has previously worked for the World Health Organisation at the WHO’s International Agency for Research on Cancer in Lyon (1987–1991), and was Medical Director of the Thames Cancer Registry in London (1991–1995). The interests of Dr. Coleman are trends in cancer incidence, mortality and survival, and the application of these tools to the public health control of cancer. He has published over 200 articles on these topics, including ‘‘Cancer survival in five continents: a worldwide population-based study (CONCORD)’’, the first worldwide analysis of cancer survival, with standard quality-control procedures and identical analytic methods for all datasets. CONCORD provides survival estimates for 1.9 million adults (aged 15– 99 years) diagnosed with a first, primary, invasive cancer of the breast (women), colon, rectum, or prostate during 1990–1994 and followed up to 1999, using individual tumour records from 101 population-based cancer registries in 31 countries on five continents.
3 Plenary lecture The 2009 influenza pandemic—unexpected beginning, uncertain future Johan Giesecke, Chief Scientist, European Centre for Disease Prevention and Control, Stockholm, Sweden Johan Giesecke is Chief Scientist and Head of the Scientific Advice Unit at the newly established Stockholm-based EU Agency ECDC (European Centre for Disease Prevention and Control). He is also Adjunct Professor of Infectious Disease Epidemiology at the Karolinska Institute Medical University in Stockholm. From a background as infectious disease clinician, he trained epidemiology at the London School of Hygiene and Tropical Medicine, and before joining ECDC was State Epidemiologist for Sweden for 10 years. He has been active in the efforts to harmonise infectious disease surveillance and control within the EU, and during a 1-year sabbatical to WHO Geneva in 1999/2000 he led the group working on the revision of the International Health Regulations. Research interests include: epidemic modelling, HIV/STIs, and very late sequelae of acute infections. He has published some 150 scientific papers, has written a textbook on infectious disease epidemiology and co-edited another. Plenary lecture Life course epidemiology: concepts and examples from the British 1946 birth cohort study and other life course cohorts Diana Kuh, MRC Unit for Lifelong Health and Ageing, MRC National Survey for Health and Development, London, UK Professor Diana Kuh is the Director of the MRC Unit for Lifelong Health and Ageing and of the MRC National Survey of Health and Development, the first British birth cohort study that has followed 5362 men and women since their birth in a week in March 1946. Diana is also the Principal Investigator for ‘Healthy Ageing across the Life Course (HALCyon)’, a collaborative research programme that brings together investigators on nine UK cohort studies to study lifetime influences on healthy ageing. Since 1987, Diana has used data from the MRC National Survey of Health and Development to study how biological, psychological and social factors at different stages of life, independently, cumulatively or interactively affect adult health and function and premature mortality. Her research interests are in adult physical capability and musculoskeletal function and their change with age, women’s health, cardiovascular function and wellbeing. Diana is internationally recognised for the advancement of the field of life course epidemiology where she is the co-editor of the acknowledged key texts: ‘‘A life course approach to chronic disease epidemiology’’ (2nd edition Oxford University Press 2004, with Professor Yoav Ben-Shlomo) and ‘‘A life course approach to women’s health’’ (Oxford University Press 2002, with Dr Rebecca Hardy). Diana has published over 100 articles in a wide range of peerreviewed journals. She was made Professor of Life Course Epidemiology at University College London in 2003.
Precongress courses, workshops and satellite meetings Generalized Linear Models using R Andrzej Gałecki, University of Michigan, Institute of Gerontology, Ann Arbor, USA Generalized linear models provide a unified framework for many commonly used analytical techniques, such as classical linear models,
123
4
IEA-EEF European Congress of Epidemiology 2009
logistic regression, Poisson regression, and Cox’s proportionalhazards for survival data Andrzej Gałecki is a Research Professor in the Department of Internal Medicine, and Institute of Gerontology at the University of Michigan Medical School, and holds Research Scientist position in the Department of Biostatistics at the University of Michigan School of Public Health. His background is in applied mathematics, medicine and epidemiology. His research interests lie in the development and application of statistical methods for analyzing correlated and overdispersed data in studies of elderly. Since his graduation from Warsaw University of Technology and Warsaw Medical University he has published over 60 papers in peer reviewed journals and has developed a few advanced SAS Macros.
Mixed Models Jos Twisk, VU University Medical Center, Department of Epidemiology and Biostatistics, Amsterdam, The Netherlands In this workshop, the basic concepts of mixed models will be explained with a special emphasis on the interpretation of the results of mixed model analyses. The following topics will be discussed: • • • • • • •
Basic principles of mixed model analysis Background of mixed model analysis Example of a mixed model with a two level structure Example of a mixed model with a three level structure What do we gain by using mixed models The use of mixed model analysis in longitudinal studies Software
Furthermore, at the end of the workshop there will be a short ‘quiz’ to evaluate whether the attendants of the workshop understood the theory and interpretation of mixed model analysis. Jos Twisk studied human movement science in Amsterdam, and after his graduation in 1990 he joined the research team of the Amsterdam Growth and Health Longitudinal Study (AGHLS). In 1995, he finished his PhD-thesis, which was related to this longitudinal study and moved to the department of Epidemiology (the EMGO-Institute). After his PhD, he supervised several projects within the AGHSL and participated as a teacher and coordinator in several postdoctoral courses given at the EMGO-Institute. In this period, he specialised himself in the methodological field of longitudinal data analysis and multilevel analysis and wrote two textbooks about it (both published by Cambridge University Press). In 2005, he became a professor on applied biostatistics and head of the department of methodology and applied biostatistics at the Institute of Health Sciences. He is now also head of the expertise centre of applied longitudinal data analysis. His main activities are statistical and methodological consultancies, teaching and applied methodological research.
Clinical epidemiology Diederick E. Grobbee, UMC Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands In the current era of evidence-based medicine, clinical epidemiology is increasingly being recognized as an important tool in the critical appraisal of available evidence and the design of new studies. Epidemiologic principles and methods provide and invaluable tool for applied clinical research. In the seminar the main aspects of clinical epidemiologic study design will be addressed, including the principles of diagnostic and prognostic research. In an interactive session, examples will be discussed and published literature will be analyzed.
Diederick Grobbee is Professor of Clinical Epidemiology at the University Medical Center Utrecht and chair of the Julius Center for Health Sciences and Primary Care. He obtained an MD at the Utrecht University and a PhD in Epidemiology at Erasmus University, Rotterdam. After a residency in Internal Medicine, he was appointed Associate Professor and subsequently full Professor of Clinical Epidemiology at Erasmus University. He was a Visiting Associate Professor at Harvard School of Public Health, Boston and is a Visiting Professor of Epidemiology and Medicine at Sydney University. He is a member of the Royal Dutch Academy of Arts and Sciences, board member of the Netherlands Heart Foundation and board member of the Academic Alliance for Clinical Trials. He is a past president of the Netherlands Epidemiological Association. At Utrecht University he directs the MSc and PhD programmes ‘‘Epidemiology’’. Diederick Grobbee has been on the editorial board of several journals and (inter)national scientific, search and site-visit committees. He is editor of the European Journal of Epidemiology and editor-in-chief of the European Journal for Cardiovascular Prevention and Rehabilitation. Diederick Grobbee has been a (principal) investigator in many large scale epidemiologic studies and randomized trials of interventions for the prevention and treatment of cardiovascular diseases. His experience covers the full range of epidemiologic study designs including trials, cohorts and case-control studies. Topics include hypertension, diabetes, risk factors in children and youngsters, coronary heart disease, women’s health, cardiovascular ageing, heart failure and non-invasive vascular imaging. In addition to research on drugs he examined the cardiovascular effects of various nutrients and dietary factors in cohort studies and trials. Apart from his interest in cardiovascular disease he has worked on the theoretical principles and methods of diagnostic and prognostic research and together with Arno W. Hoes published the textbook ‘‘Clinical Epidemiology’’(Boston, Jones&Bartlett, 2009). He supervised over 100 PhD fellows and published more than 800 scientific papers, chapters and books. Number of international citations (Google Scolar) 37860. Hirsch: h = 98. Methods of evaluation of cancer incidence and treatment effectiveness using cancer registries (Moderators: Magdalena Bielska—Lasota, Tapio Luostarinen) 1. Cancer predictions—Tadeusz Dyba, Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, Helsinki, Finland Dr. Tadeusz Dyba graduated from Cracow Economical University specialiazing in statistics and econometrics. He worked as a statistician at Cracow Cancer Registry, Karolinska Institute in Stockholm and, after defending his doctoral thesis in statistics at Helsinki University as a scientist at Finnish Cancer Registry. Dr. Dyba has published 40 scientific publications in cancer epidemiology and biostatistics and is specialized in cancer incidence prediction. 2. Visualizing spatial-temporal variation of disease incidence— Eero Pukkala, Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, Helsinki, Finland Eero Pukkala is Director of Statistics and epidemiologist at the Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, Helsinki. He is Professor of Public Health and Epidemiology responsible for the EU international postgraduate education network ‘‘Speading of Excellence in Cancer Control using Population-Based Registries and Biobanks’’, member of the Scientific Committee on Emerging and Newly Identified Health Risks of the European Commission, university lecturer and visiting professor in the University of Tampere. He has been an initiator, coordinator, participant of steering committee or national principal investigator in tens of international scientific studies and study programmes. He has published more than 500 peer-reviewed epidemiological publications, including studies on cancer and
IEA-EEF European Congress of Epidemiology 2009 other health outcomes related to physical and social environment, occupational hazards, and life habits, works on biological risk factors of cancer; familial clustering of cancer; evaluations of the effects of screenings and other interventions, survival studies, cancer predictions, cancer map presentations for numerous countries and methodological publications on register and biobank data quality, data protection or privacy issue (H—index exceeded 50 in Summer 2008). 3. Applications of cancer registry data in cancer control—Michel Coleman, London School of Hygiene and Tropical Medicine, London, UK (see under Plenary lecture above)
Precongress satellite meeting of the Reproductive Outcome and Migration (ROAM) international collaboration Reproductive Outcomes And Migration (ROAM): An international research collaboration was established in August 2005 and currently consists of 33 researchers from 13 countries. All collaboration members are self-selected researchers interested in migration and reproductive health. Although all aspects of migrant women’s reproductive health are of interest to ROAM, the collaboration has begun its work by focusing on perinatal health. Some of the work of ROAM can be seen in upcoming or previously published articles (Small R, Gagnon AJ, Gissler M, Zeitlin J, Bennis M, Glazier R, Haelterman E, Martens G, Urquia M, Vangen S. Somali women and their pregnancy outcomes post-migration: data from six receiving countries. British Journal of Obstetrics and Gynaecology. Dec 2008, 115(13): 1630–1640; Gissler M, Alexander S, Macfarlane A, Small R, Stray-Pedersen B, Zeitlin J, Zimbeck M, Gagnon A for the ROAM collaboration. Stillbirths and infant deaths among migrants in industrialised countries. Acta Obstetricia et Gynecologica Scandinavica 2009, 28: 134–148; Gagnon AJ, Zimbeck M, Zeitlin J and the ROAM collaboration. Migration to western industrialized countries and perinatal health: A systematic review. Social Science and Medicine. Accepted April 2009; Urquia M, Glazier R, Blondel B, Zeitlin J, Gissler M, McFarlane A, Ng E, Heaman M, Stray-Pedersen B, Gagnon AJ. International Migration and Adverse Birth Outcomes: Role of Ethnicity, Region of Origin and Destination. Journal of Epidemiology and Community Health. Accepted May 2009).
5 abstracts not evaluated by the reviewer(s), decisions were made by the Chairman of the Scientific Committee. There were 40 abstracts evaluated this way. This way, 24 abstracts have been rejected, 124 accepted for oral presentation and 127 accepted for poster presentation.
Table 1 Abstracts submitted and accepted by country Country
Submitted
Accepted Total
Poster
Oral
Algeria
1
1
1
0
Australia
2
2
1
1
Belarus
5
3
3
0
Belgium
5
5
1
4
Brazil
11
11
5
6
Canada
5
5
2
3
China
1
1
1
0
Colombia
1
1
1
0
Croatia
1
1
0
1
Cyprus
1
1
0
1
Czech Republic
6
6
3
3
Denmark
2
2
0
2
Egypt
1
1
0
1
Finland
6
6
3
3
France
6
5
3
2
15
14
5
9
Germany Hungary
4
4
3
1
India
3
2
2
0
4
4
0
4
21
17
16
1
Ireland
2
2
1
1
Israel
2
2
1
1
Italy
24
21
10
11
Kuwait
1
1
1
0
Latvia
1
1
1
0
Lithuania
3
3
2
1
There were submitted 275 abstracts from 43 countries (Table 1) by the19th April 2009. Each abstract was evaluated anonymously by 2 independent reviewers—members of the Scientific Committee. The following criteria of evaluation were applied:
Morocco
2
2
1
1
Netherlands
3
3
0
3
Nigeria
6
2
2
0
Norway
10
9
1
8
•
For differences between two reviewers less or equal 2, the average of two evaluations was taken and abstract
Philippines
4
2
2
0
Poland
37
34
21
13
was rejected, if mean score was less or equal 4.0 was accepted for poster presentation, if mean score was more or equal 4.5 and less or equal 6.5 was accepted for oral presentation, if mean score was more or equal 7.0.
Portugal
16
16
4
12
Romania
2
2
2
0
Russia
2
2
1
1
Serbia
3
3
3
0
South Korea
6
6
2
4
There were 170 abstracts evaluated this way.
Spain
Abstract submission and peer review evaluation
•
For differences in scores between two reviewers more than 2 and agreement in final grouping (reject, poster, oral) the final grouping was taken There were 65 abstracts evaluated this way.
•
For differences in scores between two reviewers more than 2 and lack of agreement in final grouping (reject, poster, oral) and for
Indonesia Iran
18
18
9
9
Sweden
2
2
0
2
Taiwan
1
1
0
1
Turkey
4
4
3
1
18
18
7
11
United Kingdom USA
7
5
3
2
Total
275
251
127
124
123
6
IEA-EEF European Congress of Epidemiology 2009
After sending abstract notification on 30th May 2009, 4 oral and 10 poster presentations have been withdrawn by the authors, 3 authors of oral presentations and 15 authors of poster presentations have not registered and 2 authors of oral presentations asked for change into poster presentations. Finally, there are 115 oral presentations and 104 poster presentations published in the supplement and grouped into the following thematic sessions, taking into account authors suggestions (Table 2):
Table 2 Presentations by topic Topic
Total n
Oral Poster presentations presentations %
n
%
n
young scientists from middle-income countries from the European Region (as defined by the World Bank). The scholarship covers the registration fee. Applicants should be epidemiologist in practice, teaching or research and be below the age of 40 years old unless there are exceptional circumstances. Their abstract had to be accepted for oral or poster presentation at the meeting. Applications for both kinds of awards had to be sent by 30th April 2009. Selection of candidates was done by the Chairman and Vicechairman of the Local Organising Committee and the Chairman of the Scientific Committee on 29th May 2009. There were 10 applications from 5 countries for the IEA Bursary and 15 applications from 6 countries for the IEA-EEF Scholarship. The IEA Bursary has been awarded to 3 persons and the IEA-EEF Scholarship—to 13 persons. Congratulations!
%
Cancer epidemiology
23
10.5
14
12.2
9
8.7
Cardioepidemiology
25
11.4
12
10.4
13
12.5
4
1.8
4
3.5
Perinatal epidemiology
22
10.0
12
10.4
10
9.6
Children and adolescent health epidemiology Lifecourse epidemiology
21
9.6
7
6.1
14
13.5
8
3.7
6
5.2
2
1.9
Infectious diseases epidemiology
25
11.4
11
9.6
14
13.5
Epidemiology of elderly
11
5.0
4
3.5
7
6.7
Genetic epidemiology
4
1.8
4
3.5
0
0.0
Clinical epidemiology
11
5.0
9
7.8
2
1.9
Pharmacoepidemiology
4
1.8
4
3.5
0
0.0
Health services research
13
5.9
4
3.5
9
8.7
Environmental epidemiology
7
3.2
4
3.5
3
2.9
Occupational epidemiology
13
5.9
7
6.1
6
5.8
Geert van der Heijden National representative of the Netherlands Epidemiological Society
Biostatistics
3
1.4
3
2.6
0
0.0
Methods
2
0.9
0
0.0
2
1.9
Hans-Werner Hense National representative of the German Association of Epidemiology
23
10.5
10
8.7
13
12.5
219 100.0 115
100.0
104
100.0
Diabetes mellitus epidemiology
Social epidemiology Total
Scientific Programme Committee
0.0 Andrzej Zielin´ski Vice President of the Polish Epidemiological Association—Chairman Susana Sans Mene´ndez European Councilor, International Epidemiological Association— European Epidemiology Federation—Vice Chairman Henrique Barros National representative Association
of
the
Portuguese
Epidemiological
Beatrice Blondel National representative of the Association of French Speaking Epidemiologists Diederick E. Grobbee UMC Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht
Tapio Luostarinen National representative of the Finnish Association of Epidemiology Jorn Olsen Past President of the International Epidemiological Association
Awards Two types of awards have been available to qualified young scientists willing to participate in the EUROEPI 2009 Congress: • •
IEA Bursaries IEA-EEF Scholarships
IEA BURSARIES are awarded by the International Epidemiological Association to young scientists from low-income countries (as defined by the World Bank). The bursary reimburses for the registration fee and travel/accommodation expenses. Additionally, the IEA membership will be given to the non-member applicant and the cost of a 3-year membership will be covered. Applicants should be epidemiologist in practice, teaching or research and be below the age of 40 years old unless there are exceptional circumstances. Their abstract had to be accepted for oral or poster presentation at the meeting. IEA-EEF SCHOLARSHIPS are awarded by the International Epidemiological Association - European Epidemiology Federation to
Andrzej Paja˛k Jagiellonian University Medical College, Institute of Public Health, Cracow Nereo Segnan National representative of the Italian Association of Epidemiology Jos Twisk VU University Medical Center, Department of Epidemiology and Biostatistics, Amsterdam Robert West Past national representative of the United Kingdom Society for Social Medicine Mirosław Wysocki Vice President of the Polish Epidemiological Association Jan Zejda President of the Polish Epidemiological Association
IEA-EEF European Congress of Epidemiology 2009
7
Local Organising Committee Katarzyna Szamotulska National Research Institute of Mother and Child, Warsaw—Chairman Ewa Mierzejewska National Research Institute of Mother and Child, Warsaw—Vice Chairman Magdalena Rosin´ska National Institute of Public Health—National Institute of Hygiene, Warsaw Izabela Sakowska National Institute of Public Health—National Institute of Hygiene, Warsaw
Grzegorz Bro_zek Medical University of Silesia, Katowice Barbara Kołłataj Medical University of Lublin, Lublin Iwona Ługowska National Research Institute of Mother and Child, Warsaw Longin Niemczyk Medical University of Warsaw, Warsaw Kinga Polan´ska Medical University of Ło´dz´, Ło´dz´
Paweł Steffanow National Institute of Public Health—National Institute of Hygiene, Warsaw
Krystyna Szafraniec Institute of Public Health—Jagiellonian University Medical College, Cracow
Alicja Krzy_zaniak Poznan University of Medical Sciences, Poznan
Dorota Szostak-Wegierek Medical University of Warsaw, Warsaw
123
8
IEA-EEF European Congress of Epidemiology 2009
EUROEPI 2009 Congress Epidemiology for Clinical Medicine and Public Health
Thursday, 27 August 2009 9.00–10.00
Inauguration lecture Epidemic of Diabetes: Challenges and opportunities for epidemiologists
Annual Scientific Meeting of the International Epidemiological Association—European Epidemiological Federation, Warsaw, 26th–29th August 2009
Andrzej Krolewski, Harvard Medical School, Joslin Diabetes Center, Section on Genetics and Epidemiology, Boston, USA
Programme 10.00–11.00
Early morning parallel sessions (see page 10)
Wednesday, 26 August 2009
Session 1—Lifecourse epidemiology I
9.00–13.00
Precongress morning course
Session 2—Biostatistics
Generalized Linear Models using R
Session 3—Clinical epidemiology I
Andrzej Gałecki, University of Michigan, Institute of Gerontology, Ann Arbor, USA
Session 4—Occupational epidemiology I 11.00–11.30
Coffee break
Precongress parallel afternoon course
11.30–12.50
Late morning parallel sessions (see page 11)
14.00–18.00
14.00–18.00
14.00–17.00
Mixed Models
Session 1—Cancer epidemiology I
Jos Twisk, VU University Medical Center, Department of Epidemiology and Biostatistics, Amsterdam, The Netherlands Precongress parallel afternoon course
Session 2—Perinatal epidemiology I Session 3—Cardioepidemiology I Session 4—Infectious diseases epidemiology I
Clinical epidemiology
12.50–14.30
Lunch break
Diederick E. Grobbee, UMC Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands
14.30–15.30
Plenary lecture Cancer Survival in five continents (CONCORD study) Michel Coleman, London School of Hygiene and Tropical Medicine, London, UK
Pre-congress parallel afternoon workshop Methods of evaluation of cancer incidence and treatment effectiveness using cancer registries (Moderators: Magdalena Bielska—Lasota, Tapio Luostarinen)
15.30–16.30
Cancer predictions—Tadeusz Dyba, Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research
Epidemiology of elderly Health services research I
Visualizing spatial-temporal variation of disease incidence—Eero Pukkala, Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research
Methods Perinatal epidemiology I Children and adolescent health epidemiology I
Applications of cancer registry data in cancer control—Michel Coleman, London School of Hygiene and Tropical Medicine, London, UK 9.00–18.00
Precongress parallel meeting of the Reproductive Outcome and Migration (ROAM) international collaboration
16.00–18.00
Registration
18.00–20.00
Greetings from the Past President of the International Epidemiological Association* and Welcome reception *Jorn Olsen, University of California, School of Public Health, Department of Epidemiology, Los Angeles, USA
123
Poster session (see page 18) ? coffee break Cancer epidemiology I Cardioepidemiology I
Infectious diseases epidemiology I Social epidemiology I Lifecourse epidemiology 16.30–17.50
Afternoon parallel sessions Session 1—Diabetes mellitus epidemiology Session 2—Children and adolescent health epidemiology I Session 3—Pharmacoepidemiology Session 4—Social epidemiology I
18.00–20.00
IEA-EEF Board meeting
18.00–21.00
Warsaw City Tour
IEA-EEF European Congress of Epidemiology 2009
9
Friday, 28 August 2009
Saturday, 29 August 2009
9.00–10.00
9.00–10.00
Plenary lecture The 2009 influenza pandemic—unexpected beginning, uncertain future Johan Giesecke, Chief Scientist, European Centre for Disease Prevention and Control, Stockholm, Sweden
10.00–11.00
European epidemiology in practice Representatives of national societies 10.00–11.00
Session 2—Social epidemiology III Session 3—Clinical epidemiology III
Session 1—Lifecourse epidemiology II
Session 4—Infectious diseses epidemiology III
Session 2—Children epidemiology II Session 4—Social epidemiology II 11.00–11.30
Coffee break
11.30–12.50
Late morning parallel sessions (see page 14)
11.00–11.30
Coffee break
11.30–12.50
Late morning parallel sessions (see page 17) Session 1—Cancer epidemiology III Session 2—Perinatal epidemiology III Session 3—Cardioepidemiology III
Session 1—Cancer epidemiology II 12.50–13.50
Session 4—Environmental epidemiology Closing ceremony
13.50–14.30
Lunch
15.00–20.00
More Highlight of Warsaw Tour
Session 2—Perinatal epidemiology II Session 3—Cardioepidemiology II Session 4—Infectious diseses epidemiology II 12.50–14.30
Lunch break
14.30–15.30
Plenary lecture
Early morning parallel sessions (see page 16) Session 1—Cancer epidemiology III
Early morning parallel sessions (see page 13)
Session 3—Clinical epidemiology II
Panel discussion
Life course epidemiology: concepts and examples from the British 1946 birth cohort study and other life course cohorts Diana Kuh, MRC Unit for Lifelong Health and Ageing, MRC National Survey for Health and Development, London, UK 15.30–16.30
Poster session (see page 18) ? coffee break Cancer epidemiology II Cardioepidemiology II Occupational epidemiology Environmental epidemiology Health services research II Perinatal epidemiology II Children and adolescent health epidemiology II Infectious diseases epidemiology II Clinical epidemiology Social epidemiology II
16.30–17.50
Afternoon parallel sessions (see page 15) Session 1—Genetic epidemiology Session 2—Epidemiology of elderly Session 3—Health services research Session 4—Occupational epidemiology II
18.00–18.30
Bus transfer to Łazienki Park
18.30–19.00
Visiting Łazienki Park
19.00
Gala Dinner at the Belvedere Restaurant
123
10
IEA-EEF European Congress of Epidemiology 2009
Sessions Detailed Programme Thursday, 27 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Lifecourse epidemiology I
Early Morning Session 3—Clinical epidemiology I
Moderator: Susana Sans Menendez
Moderator: Diederick E. Grobbee
IS POSTNATAL GROWTH ASSOCIATED WITH BLOOD PRESSURE IN CHILDHOOD? N. M. Davies, K. Tilling, F. Windmeijer, G. D. Smith, Y. Ben-Shlomo, M. S. Kramer, R. M. Martin (213)
DRUG USE IN POPULATION SCREENING: A PHARMACOEPIDEMIOLOGICAL ASPECTS J. Atthobari, S. T. Visser, R. Gansevoort, P. E. de Jong, L. T. W. de Jong-van den Berg (129)
LOW BIRTH WEIGHT AND OTHER VASCULAR-RELATED OUTCOMES OF PREGNANCY AND SOCIOECONOMIC FACTORS IN MOTHERS OF A PORTUGUESE BIRTH COHORT E. Alves, A. Azevedo, H. Barros (296)
DEVELOPMENT OF A MELANOMA RISK SCORING TOOL AND VALIDATION ACROSS DIFFERENT ETHNIC POPULATIONS C. Fortes, S. Mastroeni, L. Bakos, G. Antonelli, Livia, M. A. Pilla, M. Allotto, R. Bonamigo, P. Pasquini, C. F. Melchi (36)
INFANT NUTRITION IS NOT ASSOCIATED WITH INSULIN RESISTANCE MEASURES IN EARLY ADULTHOOD: THE BARRY-CAERPHILLY GROWTH COHORT STUDY D. M. Williams, R. M. Martin, G. D. Smith, D. P. Davies, K. George, M. M. Alberti, Y. Ben-Shlomo, A. McCarthy (262)
THE UNRECOGNIZED PREVALENCE OF CHRONIC KIDNEY DISEASE AMONG FAMILY MEMBERS OF END STAGE RENAL DISEASE PATIENTS S. S. Abd El-Hafeez, Z. M. Gad, Hala S El-Wakil, M. H. Hashsish (264)
Early Morning Session 2—Biostatistics
Early Morning Session 4—Occupational epidemiology I
Moderator: Jos Twisk
Moderator: Jan Zejda
MULTIPLE INDICATORS OF POLY DRUG USE USING STRUCTURAL EQUATION MODELING (SEM) V. Siciliano, S. Molinaro (160)
WOMEN WITH ANAEMIA IN PREGNANCY GET HEAVIER BABIES IN A HEAVILY POLLUTED TOWN IN THE RUSSIAN ARCTIC: A REGISTRY-BASED STUDY ON 23,950 SINGLETON BIRTHS, 1973–2001 E. Chumak, A. M. Grjibovski (123)
DIETARY PATTERNS IN THE POPULATION OF PORTO USING MIXED CONTINUOUS AND BINARY VARIABLES R. Gaio, J. P. da Costa, A. C. Santos, E. Ramos, C. Lopes (250) SPATIAL ANALYSIS OF DRUG-RELATED HOSPITAL ADMISSIONS: AN AUTO-GAUSSIAN MODEL TO ESTIMATE THE HOSPITALIZATION RATES IN ITALY. E. Colasante, S. Molinaro, F. Mariani (145)
123
HYPERTENSION IN WORKING POPULATION IN CROATIAPREVALENCE AND CONTROL RATE M.Gomzi, B. Winterhalter-Zvonar (76) CARCINOGENIC RISK IN BLACK-COAL MINERS WITHOUT PNEUMOCONIOSIS IN THE CZECH REPUBLIC - LONGITUDINAL STUDY (1992 - 2006) H. Tomaskova, Z. Jirak, A. Splichalova, P. Urban, V. Zavadilova (131)
IEA-EEF European Congress of Epidemiology 2009
11
Thursday, 27 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology I Moderator: Tapio Luostarinen BIOMARKERS EXPRESSION IN INTERVAL AND SCREENDETECTED BREAST CANCERS IN A BREAST CANCER SCREENING PROGRAMME IN BARCELONA L. Domingo, S. Servitja, J. Corominas, J. Martı´nez, F. Ferrer, F. Macia`, J. Albanell, X. Castells, M. Sala (175) HORMONE REPLACEMENT THERAPY USE AND INCIDENCE OF CENTRAL NERVOUS SYSTEM TUMOURS; THE MILLION WOMEN STUDY V. Benson, K. Pirie, J. Green, G. Reeves, V. Beral (97) RISK OF COLORECTAL CANCER AFTER RESECTION OF ADENOMAS V. Cottet, V. Jooste V, A. M. Bouvier, J. Faivre, C. Bonithon-Kopp (280) RECURRENT SORES BY ILL-FITTING DENTURES AND INTRA-ORAL SQUAMOUS CELL CARCINOMA IN SMOKERS J. L. F. Antunes, G. F. Vaccarezza, P. Michaluart-Ju´nior (3)
Late Morning Session 2—Perinatal epidemiology I Moderator: Beatrice Blondel HIGH PERINATAL MORTALITY RATE AMONG IMMIGRANTS IN BRUSSELS J. Racape´, M. De Spiegelaere, S. Alexander, M. Dramaix, P. Buekens, E. Haelterman (24) DURATION OF RESIDENCE AND ADVERSE BIRTH OUTCOMES AMONG IMMIGRANTS TO URBAN ONTARIO M. L. Urquia, J. W. Frank, R. H. Glazier, R. Moineddin (194) MATERNAL COUNTRY OF BIRTH VARIATIONS IN CAESAREAN SECTION: ANOTHER CAUSE FOR CONCERN? R. Small (78) CHILDBEARING HEALTH AND SERVICE NEEDS OF MIGRANTS AND REFUGEES TO MONTREAL, QUEBEC (CANADA) A. Gagnon, O. Wahoush, G. Dougherty, J. F. Saucier, C. L. Dennis, E. Stanger, B. Palmer, L. Merry, D. E. Stewart (275)
Late Morning Session 3—Cardioepidemiology I Moderator: Andrzej Paja˛k TWELVE YEAR TRENDS IN CARDIOVASCULAR RISK FACTORS IN PATIENTS WITH CORONARY HEART DISEASE. RESULTS FROM THE EUROASPIRE I, II AND III STUDIES IN ¨ NSTER THE REGION OF MU C. Prugger, J. Heidrich, J. Wellmann, R. Dittrich, R. Telgmann, S. M. Brand-Herrmann, U. Keil (186) PREVALENCE OF THE METABOLIC SYNDROME AND ITS COMPONENTS IN NORTHWEST RUSSIA: THE ARKHANGELSK STUDY O. Sidorenkov, O. Nilssen, T. Brenn, S. Martiushov, A. M. Grjibovski (70) LOWER SERUM ADIPONECTIN LEVELS INDEPENDENTLY PREDICT TYPE 2 DIABETES IN A KOREAN RURAL POPULATION J. K. Park, S. B. Koh (238) VALIDATION OF REFLECTANCE COLORIMETRIC METHOD FOR FASTING GLUCOSE AND TOTAL CHOLESTEROL DOS˜ O PRETO, SP, BRAZIL, 2006 AGE. RIBEIRA V. de Castro, S. A. de Moraes, I. C. M. de Freitas (234)
Late Morning Session 4—Infectious diseases epidemiology I Moderator: Andrzej Zielinski NEWLY DIAGNOSED HIV INFECTIONS IN DRUG USERS IN PORTUGAL: FREQUENCY AND CHARACTERISTICS OF FIRST-TIME AND FOLLOW-UP VISITORS OF DRUG TREATMENT CENTRES R. Lucas, M. J. Santos, L. Prasad, H. Barros (268) HEROIN USERS IN BELGIUM, 2001–2005: PATTERNS OF DRUGS USE, RISK BEHAVIOURS, INFECTIOUS DISEASES AND USE OF HEALTH SERVICES N. Deprez, M. Roelands (98) HOMOSEXUALLY ACQUIRED SYPHILIS IN NORWAY, 1999– 2007: AN ONGOING BATTLE I. Jakopanec, A. Grjibovski, P. Aavitsland (291) RISK FACTORS FOR EXTENSIVELY DRUG-RESISTANT TUBERCULOSIS: A SYSTEMATIC REVIEW M. Tavares, B. Lima, H. Barros (293)
123
12
IEA-EEF European Congress of Epidemiology 2009
Thursday, 27 August 2009 Parallel afternoon oral sessions: 16.30–17.50 Afternoon Session 1—Diabetes mellitus epidemiology
Afternoon Session 3—Pharmacoepidemiology
Moderator: Hans-Werner Hense
CONCOMITANT USE OF GASTROPROTECTIVE AGENTS (GPAS) AMONG NSAID/COX-2 SELECTIVE INHIBITOR USERS: A HOSPITAL REGISTER-BASED STUDY J. Atthobari, E. Kristin, I. Dwiprahasto (128)
INTERACTION BETWEEN GAMMA-GLUTAMYTRANSFERASE AND OBESITY ON RISK FOR PREVALENCE OF TYPE 2 DIABETES S. B. Koh, J. K. Park (239) OBESITY PREVALENCE AND AWARENESS AMONG ADULT POPULATION AND NHS OPERATORS IN THE LAZIO REGION, ITALY: RESULTS FROM THE PASSI SURVEILLANCE M. O Trinito, A. Capon, S. Iacovacci, A. Lancia (69) THE EPIDEMIOLOGY OF METABOLIC SYNDROME IN THE HUNGARIAN ADULT POPULATION: A REPRESENTATIVE SURVEY E. Szigethy, G. Sze´les, Z. Voko´, A. Nagy, A. Horva´th, T. Hidve´gi, ´ da´ny (278) G. Jermendy, G. Paragh, G. Blasko´, R. A DIABETES MELLITUS PREVALENCE AND CORRELATES IN A ˜ O PRETO, SP, BRAZILIAN POPULATION LIVING IN RIBEIRA BRAZIL, 2006. S. A. de Moraes, I. C. M. de Freitas (236)
Afternoon Session 2—Children and adolescent health epidemiology I Moderator: Henrique Barros PATHWAYS FROM INTERPARENTAL VIOLENCE TO VICTIMIZATION BY PEERS DURING ADOLESCENCE: THE ROLE OF SELF-ESTEEM AND PARENTING STYLE Ch. Roustit, B. Chaix, P. Chauvin (287) ASSESSMENT OF REPRODUCIBILITY FOR A SELF-DELIVERED ESPAD (THE EUROPEAN SCHOOL SURVEY PROJECT ON ALCOHOL AND OTHER DRUGS) QUESTIONNAIRE ABOUT USE OF ILLICIT DRUGS USING THE TEST-RETEST METHOD O. Curzio, S. Molinaro (204) OVERWEIGHT AND OBESITY IN ADOLESCENTS IN NORTHWEST RUSSIA: DOES THE PREVALENCE DEPEND ON WHETHER RUSSIAN OR INTERNATIONAL CRITERIA ARE APPLIED? S. L. Khasnutdinova, A. M. Grjibovski (125) UNDERWEIGHT, OVERWEIGHT AND OBESITY TRENDS IN WARSAW ADOLESCENTS DURING THE 35 YEARS PERIOD, 1971–2006 J. Charzewska, E. Chabros, B. Wajszczyk, Z. Chwojnowska, E. Kraszewska (279)
123
Moderator: Geert van der Heijden
PATTERNS OF CO-PRESCRIBING OF ANGIOTENSIN CONVERTING ENZYME INHIBITOR AND ANGIOTENSIN RECEPTOR BLOCKERS IN IRELAND W. A. H. Wan Md Adnan, N. L. Zaharan,K. Bennett, C. A. Wall, J. Feely (246) GENDER AND AGE DIFFERENCES IN THE ADHERENCE TO EVIDENCE-BASED DRUG THERAPY AFTER ACUTE MYOCARDIAL INFARCTION: A POPULATION BASED COHORT STUDY IN ROME, ITALY U. Kirchmayer, N. Agabiti, V. Belleudi, M. Davoli, D. Fusco, M. Stafoggia, M. Arca`, A. P. Barone, C. A. Perucci (96) DETERMINANTS OF ANTIDEPRESSANT AND TRANQUILLIZER USE IN THE GENERAL POPULATION J. Van der Heyden, L. Gisle, E. Hesse, S. Demarest, J. Tafforeau (162)
Afternoon Session 4—Social epidemiology I Moderator: Mirosław Wysocki SOCIOECONOMIC STATUS AS DETERMINANT FOR PARTICIPATION IN MAMMOGRAPHY SCREENING: ASSESSING THE DIFFERENCE BETWEEN USING WOMEN’S OWN VERSUS THEIR PARTNER’S M. Kjelle´n, M. von Euler-Chelpin (141) SOCIOECONOMIC DIFFERENCES IN SELF-MANAGEMENT OF CORONARY HEART DISEASE I. Forde, T. Chandola, M. Kivimaki (152) PREVALENCE OF OBSTRUCTIVE PULMONARY DISORDERS AND SYMPTOMS IN ADULT POPULATION OF CHORZOW— PRELIMINARY RESULTS OF GALEN SCREENING SURVEY G. Brozek, J. Zejda, M. Farnik, M. Kowalska (248) DIFFERENCES IN LIFE EXPECTANCY BETWEEN EAST AND WEST OF EUROPE M. Manczuk and the HEM project collborative group (219)
IEA-EEF European Congress of Epidemiology 2009
13
Friday, 28 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Lifecourse epidemiology II
Early Morning Session 3—Clinical epidemiology II
Moderator: Robert West
Moderator: Hans-Werner Hense
EFFECT OF PARENTAL SOCIAL CLASS ON ALCOHOL RELATED MORTALITY IN ADULTHOOD AND OLD AGE Flaminia Chiesa, Ilona Koupil (176)
CALIBRATION OF THE NEW YORK HEART ASSOCIATION FUNCTIONAL CLASSIFICATION SYSTEM BETWEEN OPERATORS M. Severo, R. Gaio, P. Lourenc¸o, M. Alvelos, P. Bettencourt, A. Azevedo (263)
LIFECOURSE FAMILY INCOME INFLUENCES ON ADULT ORAL HEALTH: THE 1982 PELOTAS BIRTH COHORT M. A. Peres, K. G. Peres, W. Murray Thomson, J. M. Broadbent, D. P. Gigante, B. L. Horta (157) EARLY CHILDHOOD SOCIOECONOMIC CORRELATES OF QUALITY OF LIFE IN ADOLESCENCE J. Mazur, A. Dzielska, K. Mikiel-Kostyra, E. Smolen (271)
SENSIBILITY AND RELATIVE INDEX OF SPECIFICITY OF THE INFLUENZA SYMPTOMS T. Vega, J. E. Lozano, R. O. de Lejarazu, M. Gutie´rrez, J. Bermejo, J. M. Eiros, J. Castrodeza (231) DETERMINANT FACTORS CONTRIBUTING TO THE DEATH OF ADULT INDONESIAN HAJJ PILGRIMS IN MECCA IN 1427 H ISLAMIC YEAR M. Pane, N. Kodim (180)
Early Morning Session 2—Children and adolescent health epidemiology II
Early Morning Session 4—Social epidemiology II
Moderator: Henrique Barros
MORTALITY AND CAUSES OF DEATH AMONG DRUGGED DRIVERS K. Karjalainen, T. Lintonen, A. Impinen, P. Lillsunde, P. Ma¨kela¨, O. Rahkonen, A. Ostamo (143)
IS PET OWNERSHIP IN EARLY CHILDHOOD ASSOCIATED WITH ASTHMA OR ALLERGIES IN SCHOOL AGE? METAANALYSES WITH ORIGINAL DATA FROM 11 EUROPEAN BIRTH COHORTS—A GA2LEN INITIATIVE T. Keil, S. Roll, K. H. Carlsen, P. Mowinckel, A. H. Wijga, B. Brunekreef, M. Torrent, G. Roberts, S. H. Arshad, I. Kull, M. Wickman, U. Kra¨mer, A. von Berg, E. Eller, E. A. Høst, S. Halken, C. Kuehni, B. Spycher, J. Sunyer, C. M. Chen, J. Heinrich, A. Arsanoj, C. Puig, O. Herbarth, K. Van Steen, S. N. Willich, U. Wahn, S. Lau, K. C. Lødrup Carlsen for the working group ‘Birth Cohorts’ of GALEN (Global Allergy and Asthma European Network) (179)
Moderator: Mirosław Wysocki
MENTAL HEALTH PROBLEMS AND CANNABIS USE IN THE GENERAL POPULATION, BELGIUM, 2004 M. Roelands (259) PREVALENCE OF ALCOHOL DEPENDENCE AND ASSOCIATED FACTORS. A BRAZILIAN POPULATION-BASED STUDY. S. A. de Moraes, I. C. M. de Freitas (237)
RESIDENTIAL EXPOSURE TO MOTOR VEHICLE EMISSIONS AND THE RISK OF WHEEZING AMONG 7–8 YEAR-OLD SCHOOLCHILDREN IN NICOSIA, CYPRUS N. Middleton, N. Nicolaou, S. Pipis, M. Zeniou, S. Kleanthous, P. Demokritou, P. Koutrakis, P. Yiallouros (114) LIFE PREVALENCE OF CHRONIC UPPER RESPIRATORY TRACT DISEASES AND ASTHMA AMONG CHILDREN RESIDING IN RURAL AREA NEAR A REGIONAL INDUSTRIAL PARK: CROSS-SECTIONAL STUDY I. Karakis, E. Kordysh, T. Lahav, A. Bolotin,Y. Glazer, H. Vardi, I. Belmaker, B. Sarov (40)
123
14
IEA-EEF European Congress of Epidemiology 2009
Friday, 28 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology II
Late Morning Session 3—Cardioepidemiology II
Moderator: Geert van der Heijden
Moderator: Andrzej Paja˛k
CUMULATIVE FALSE POSITIVE RISK AND ITS ASSOCIATED FACTORS IN A DATABASE FROM 10 POPULATION BASED BREAST CANCER SCREENING PROGRAMMES IN SPAIN
PASSIVE SMOKING AND SMOKING CESSATION AMONG PATIENTS WITH CORONARY HEART DISEASE ACROSS EUROPE. RESULTS FROM THE EUROASPIRE III STUDY J. Heidrich, U. Keil, J. Wellmann, D. De Bacquer, Z Reiner, C. Jennings, Z. Fras, C. Prugger (172)
R. Roma´n, M. Sala, R. Zubizarreta, D. Salas, N. Ascunce, F. Belvis, X. Castells, and the RAFP research group (86) RAW CABBAGE/SAUERKRAUT CONSUMPTION DURING ADOLESCENCE AND BREAST CANCER RISK IN POLAND D. R. Pathak, J. Charzewska, B. Zemla, Anna Kozaczka, El_zbieta Nowakowska, Dariusz Godlewski, Piotr Zurawski (282) SMOKING, AN INDEPENDENT RISK FACTOR FOR INVASIVE CERVICAL CANCER. A NORDIC JOINT STUDY T. Luostarinen, A. Simen Kapeu, Matti Lehtinen for the CCPRB-EU Network (81) HELICOBACTER PYLORI INFECTION AND GASTRIC CARDIA CANCER: SYSTEMATIC REVIEW AND META-ANALYSIS M. C. Pinto, B. Peleteiro, N. Lunet (265)
Late Morning Session 2—Perinatal epidemiology II Moderator: Beatrice Blondel PREVALENCE OF CONGENITAL HEART DEFECTS IN NEWBORNS IN GERMANY: RESULTS OF THE PAN STUDY H. W. Hense, A.Lindinger, G. Schwedler (166) LENGTH OF STAY ON NEONATAL UNITS OF VERY PRETERM BABIES IN A UK POPULATION B. Manktelow, E. Draper, D. Field (72) INFANT MORTALITY BY GESTATIONAL AGE GROUPS: A POPULATION BASED STUDY IN AN ITALIAN REGION S. Farchi, D. Di Lallo, F. Franco, A. Polo (230) ANTENATAL AND PERINATAL FACTORS IN CHILDHOOD TYPE 1 DIABETES: A CASE-CONTROL STUDY IN THE MIDDLE EAST A. Memon, S. Polack, M. A. Al-Khawari, M. Qabazard, A. Al-Adsani, M. Abdul-Rasul, R. R. Gulati, I. Siddique, N. A. Abdella, A. A. Shaltout, A. Suresh (241)
123
WHICH FACTORS INFLUENCE PHYSICAL ACTIVITY IN PERSONS WITH CARDIOVASCULAR DISEASE OR RISK FACTORS? A COMPARISON OF PATTERNS, SETTINGS AND ATTITUDES BY LEVEL OF CARDIOVASCULAR RISK C. Huy, K. Steindorf, D. Litaker, A. Thiel, C. Diehm (154) GENDER DIFFERENCES IN THE ASSOCIATIONS OF SOCIOECONOMIC POSITION AND EDUCATION WITH CORONARY HEART DISEASE INCIDENCE H. Arik, B. Adler, R. Sinnreich, N. Goldberger, J. D. Kark (272) DIET AND RURAL-URBAN MIGRATION IN INDIA: THE INDIAN MIGRATION STUDY L. Bowen, S. Ebrahim, B. De Stavola (206)
Late Morning Session 4—Infectious diseases epidemiology II Moderator: Paulo Moreira PHYSICIANS’ PRESCRIBING PRACTICES AND ANTIBIOTIC RESISTANCE IN GERMANY E. Velasco, I. Noll, A. Barger, T. Eckmanns (174) SEROEPIDEMIOLOGICAL DESCRIPTION OF CHLAMYDIAL INFECTION IN SAINT PETERSBURG, RUSSIA L. B. Kulyashova, L. A. Berezina, A. B. Zhebrun (139) RISK FACTORS AND IMPACT OF SURGICAL SITE INFECTION IN INDONESIANS HOSPITAL J. Atthobari, D. V. P. Sianipar, Suwardiman, E. M. C. Menur, I. Dwiprahasto (127) COULD STETHOSCOPES BE A SOURCE OF INFECTION? S. Rehman, H. Razzaq, O. Anwar, J. Macfie (74)
IEA-EEF European Congress of Epidemiology 2009
15
Friday, 28 August 2009 Parallel afternoon oral sessions: 16.30–17.50 Afternoon Session 1—Genetic epidemiology
Afternoon Session 3—Health services research
Moderator: Andrzej Kro´lewski
Moderator: Magdalena Bielska-Lasota
SUSCEPTIBILITY GENES AND AGE-RELATED MACULAR DEGENERATION H. W. Hense, A. Farwick, J. Wellmann, M. Stoll, D. Pauleikhoff (137)
DIAGNOSIS AND REFERRAL OF RHEUMATOID ARTHRITIS BY PRIMARY CARE PHYSICIAN: RESULTS OF A PILOT STUDY ON THE TOWN OF PISA, ITALY
ASSOCIATION BETWEEN CYTOKINE GENE POLYMORPHISMS AND GASTRIC PRECANCEROUS LESIONS B. Peleteiro, C. Carrilho, L. David, J. C. Machado, C. La Vecchia, Nuno Lunet (247) POLYMORPHISMS IN ANTIOXIDANT DEFENSE GENES AND SUSCEPTIBILITY TO HEPATOCELLULAR CARCINOMA IN A MOROCCAN POPULATION S. Ezzikouri, A. E. El feydi, R. Afifi, L. El kihal, M. Benazzouz, M. Hassar, P. Pineau, S. Benjelloun (100) USING GENERALIZED WEIGHTS OF NEURAL NETWORKS TO IDENTIFY GENE-GENE INTERACTIONS F. Guenther, N. Wawro, K. Bammann (184)
A. Della Rossa, R. Neri, R. Talarico, M. Doveri, A. Consensi, S. Salvadori, V. Lorenzoni, G. Turchetti, S. Bellelli, M. Cazzato, L. Bazzichi, P. Monicelli, S. Moscardini, S. Bombardieri (224) IDENTIFICATION OF VARIABLES OF ADVERSE EVOLUTION AT THE HOSPITAL ER IN PATIENTS WITH COPD EXACERBATION J. M. Quintana, S. Garcı´a, C. Esteban, A. Bilbao, J. L. Lobo, N. Gonza´lez, I. Lafuente, F. Aizpuru (87) PALLIATIVE AND SUPPORT HEALTH CARE AT HOME IN SPAIN. A HEALTH SENTINEL NETWORK STUDY T. Vega, J. E. Lozano, C. Quin˜ones, Y. Anes, C. Gomez, S. Fern´ lamo, on behalf of the a´ndez, M. Perucha, M. Gil, M. Ramos, R. A RECENT group (225) THE ACCURACY OF CAUSE-OF-DEATH CODING IN THE NETHERLANDS: A STUDY INTO RELIABILITY P. Harteloh, K. de Bruin, J. Kardaun (80)
Afternoon Session 2—Epidemiology of elderly
Afternoon Session 4—Occupational epidemiology II
Moderator: Robert West
Moderator: Wojciech Hanke
MAJOR LIFE EVENTS AND RISK OF PARKINSON’S DISEASE N. H. Rod, J. Hansen, E. Schernhammer, B. Ritz (66)
SENSITIVITY AND SPECIFICITY OF WORK ABILITY INDEX FOR DETERMINATION OF SICKNESS ABSENCE AND DISABILITY. S. M. Alavinia, M. Omidvar, A. Burdorf (89)
DISPARITY IN LOWER EXTREMITY FUNCTION BY AREA AND EDUCATIONAL LEVEL K. Yoon, S. Cho (256) SURVIVAL OF ELDERLY FEMALE BREAST CANCER PATIENTS IN THE WESTERN PART OF THE NETHERLANDS E. Bastiaannet, J. Portielje, L. G. M. van der Geest, A. J. M. de Craen, G. J. Liefers, C. J. H. van de Velde, R. G. J. Westendorp (138) PATTERNS OF MORTALITY AMONG OLDER CRACOW CITIZEN (22-YEARS OBSERVATION CRACOW STUDY) A. Pac, B. Tobiasz-Adamczyk, P. Brzyski (277)
COMPARISON OF CANCER RISK IN WORKERS WITH SILICOSIS AND UNEXPOSED POPULATION IN THE CZECH REPUBLIC A. Splichalova, H. Tomaskova, Z. Jirak, H. Lehocka, P. Urban, J. Holub, D. Pelclova, E. Hrncir, A. Eichlerova, M. Nakladalova, M. Bartnicka, A. Hejlek, L. Adamkova, D. Cermakova (130) WORK-RELATED RESPIRATORY SYMPTOMS IN NORWEGIAN SALMON INDUSTRY WORKER POPULATION O. Shiryaeva, L. Aasmoe, B. Bang (134) OCCUPATIONAL FATALITIES IN THE ARKHANGELSK REGION, 1996–2007 Zh. L. Varakina, A. M. Vyazmin, A. L. Sannikov, A. M. Grjibovski (284)
123
16
IEA-EEF European Congress of Epidemiology 2009
Saturday, 29 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Cancer epidemiology III
Early Morning Session 3—Clinical epidemiology III
Moderator: Magdalena Bielska-Lasota
Moderator: Hans-Werner Hense
PNEUMOCONIOSIS AND CARCINOGENIC RISK IN THE CZECH BLACK-COAL MINES H. Tomaskova, Z. Jirak, A. Splichalova, P. Urban, J. Holub, R. Gromnica, Z. Hajdukova, I. Landecka, V. Machartova, E. Korolova (133)
VALIDITY OF A SIMPLE SEVERITY SCALE USED AT THE ER TO PREDICT ADVERSE EVENTS ON PATIENTS WITH COPD EXACERBATION J. M. Quintana, S. Garcı´a, C. Esteban, A. Bilbao, J. A. Blasco, M. Bare, S. Vidal, E. Perea, J. L. Lobo (90)
TRENDS IN THE MORTALITY RATES OF MALIGNANT LUNG CANCER IN FEMALE INHABITANTS OF UPPER SILESIA,AN INDUSTRIAL AREA IN POLAND. M. Juszko-Piekut, Z. Kolosza, B. F. P. Zemla, A. Mozdzierz, J. Stojko (197)
RECURRENCE AND MORTALITY IN SECONDARY ATRIAL FIBRILLATION: A POPULATION-BASED STUDY R. Greenlee, P. Chyou, P. Sharma, P. Smith, J. Hayes, R. Mareedu, M. Ortiz, H. Vidaillet (192)
CIVIL STATUS, EDUCATION AND INCOME IN RELATION TO STAGE AND SURVIVAL OF RECTAL CANCER L. Olsson, F. Granstro¨m (276)
Early Morning Session 2—Social epidemiology III Moderator: David Batty HOW SOCIOECONOMIC STATUS INFLUENCES ROAD TRAFFIC INJURIES AND HOME INJURIES IN ROME L. Camilloni, F. Chini, S. Farchi, P. Giorgi Rossi, P. Borgia (217) FAMILY WEALTH MODIFIES THE RELATIONSHIP BETWEEN SELF-RATED HEALTH AND CHRONIC DISEASES STATUS IN EARLY ADOLESCENCE J. Mazur (191) RELATIONSHIP BETWEEN LOW HEALTH-RELATED QUALITY OF LIFE AND UNEMPLOYMENT P. Zagozdzon, J. Ejsmont (146)
INTER-OBSERVER AGREEMENT IN ABSTRACTION OF DATA FROM CLINICAL RECORDS E. Alves, V. Morais, S. Correia, A. Azevedo (260)
Early Morning Session 4—Infectious diseases epidemiology III Moderator: Andrzej Zielinski HELICOBACTER PYLORI INFECTION AND SOCIOECONOMIC STATUS IN A HIGH PREVALENCE DEVELOPED COUNTRY: PREVALENCE, INCIDENCE AND SEROREVERSION J. Bastos, B. Peleteiro, N. Lunet, H. Pinto, A. Marinho, T. Guimara˜es, C. La Vecchia, Henrique Barros (242) COVERAGE OF HIV SCREENING BEFORE THE FIRST CONTACT OF DRUG USERS WITH THE PUBLIC DRUG TREATMENT NETWORK IN PORTUGAL R. Lucas, M. J. Santos, L. Prasad, H. Barros (269) MORTALITY RATE AND OUTCOME FACTORS IN MIXED CRYOGLOBULINEMIA: THE IMPACT OF HCV A. Della Rossa, A. Tavoni, A. D’Ascanio, E. Catarsi, F. Marchi, W. Bencivelli, S. Salvadori, P. Migliorini, S. Bombardieri (226)
123
IEA-EEF European Congress of Epidemiology 2009
17
Saturday, 29 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology IIIA
Late Morning Session 3—Cardioepidemiology III
Moderator: Magdalena Bielska-Lasota
Moderator: Susana Sans Menendez
EFFECT OF SMOKING AND ALCOHOL PRICE INDEXING ON LUNG AND BREAST CANCER INCIDENCE IN DENMARK I. Soerjomataram, E. de Vries, G. Engholm, G. Paludan-Mu¨ller, H. Bronnum-Hansen, H. Storm, J. Barendregt (156)
KNOWLEDGE OF BLOOD PRESSURE LEVELS AND TARGETS IN CORONARY PATIENTS ACROSS EUROPE: RESULTS FROM THE EUROASPIRE III STUDY C. Prugger, D. De Bacquer, G. De Backer, G. Ambrosio, Z. Reiner, D. Gaita, D. Wood, K. Kotseva, U. Keil, J. Heidrich (185)
TREATMENT DELAY OF BREAST CANCER PATIENTS RECEIVING CURATIVE TREATMENT IN PODLASKIE VOIVODESHIP IN 2001–2002 POPULATION STUDY D. Mas´lach, M. Krzy_zak-Mas´lach, A. Szpak, M. Juczewska, K. Maksimowicz, U. Łapin´ska (249) CANCER INCIDENCE PREDICTION FOR THE YEARS 2010 AND 2015 S. Gozdz, R. Mezyk, T. Dyba (142)
HYPOTHESIS-ORIENTED FOOD PATTERNS AND INCIDENCE OF HYPERTENSION IN THE SUN PROSPECTIVE COHORT E. Toledo, F. Carmona-Torre, A. Alonso, J. Nun˜ez-Co´rdoba, F. Guille´n-Grima, M. A. Martı´nez-Gonza´lez (223) CALCIUM IS AN INDEPENDENT RISK FACTOR FOR HYPERTENSION; A POPULATION-BASED COHORT STUDY J. W. Kim, H. Kim, K. Ko, H. J. Koo, J. K. Park, C. Shin, S. S. Kim (182) EFFECT OF DEPRESSIVE SYMPTOMS ON THE INCIDENCE OF HYPERTENSION IN ADULTS: A POPULATION-BASED STUDY IN PORTO, PORTUGAL M. Pereira, D. Costa, A. Azevedo (243)
Late Morning Session 2—Perinatal epidemiology III Moderator: Beatrice Blondel PHYSICAL ACTIVITY AND FERTILITY IN WOMEN—THE NORTH-TRØNDELAG HEALTH STUDY S.L. Gudmundsdottir, W.D. Flanders, L.B. Augestad (39)
Late Morning Session 4—Environmental epidemiology Moderator: Jan Zejda THE EFFECTS OF AMBIENT TEMPERATURE ON THE RISK OF MYOCARDIAL INFARCTION IN LONDON, UK K. Bhaskaran (178)
RISK FACTORS FOR PRETERM BIRTH, LOW BIRTH WEIGHT AND SMALL FOR GESTATIONAL AGE BIRTHS M. Heaman, D. Kingston, B. Chalmers, R. Sauve, L. Lee, D. Young, S. Taylor-Clapp for the Maternity Experiences Study Group of the Canadian Perinatal Surveillance System (103)
FINDING CADMIUM THRESHOLD CONCENTRATION RELATED TO BREAST CANCER R. Kregzdyte, L. Strumylaite, D. Baranauskiene, O. Abdrkhmanov (106)
PREVALENCE AND SOCIO-DEMOGRAPHIC DETERMINANTS OF ANAEMIA IN PREGNANCY IN NORTHWEST RUSSIA: A REGISTRY-BASED STUDY ON 23,950 BIRTHS, 1973–2001 E. Chumak, A. M. Grjibovski (124)
SEARCHING FOR BIAS: RELIABILITY OF SELF-REPORTED HOUSEHOLD PESTICIDES USE AND ORS CORRECTION C. Fortes, S. Mastroeni, P. Boffetta, V. Salvatori, N. Melo, S. Bolli, P. Pasquini (35)
IMPACT OF EDUCATIONAL CAMPAIGN ON SOCIO-DEMOGRAPHIC FACTORS AFFECTING WOMEN’S KNOWLEDGE OF FOLIC ACID AND PRECONCEPTIONAL FOLIC ACID SUPPLEMENTATION IN REGARD TO NEURAL TUBE DEFECTS’ PREVENTION E. Mierzejewska, K. Szamotulska (283)
ENVIRONMENTAL EXPOSURE TO BIOCONTAMINANTS IN KINDERGARTENS IN POLAND A. Buczynska, M. Cyprowski, P. Soroka, I. Szadkowska-Stanczyk (300)
123
18
IEA-EEF European Congress of Epidemiology 2009
Poster sessions Thursday, 27 August 2009, 15.30–16.30
Friday, 28 August 2009, 15.30–16.30
Cancer epidemiology I
Cancer epidemiology II
Perinatal epidemiology I
Perinatal epidemiology II
Poster-53
Poster-190
Poster-4
Poster-92
Poster-136
Poster-88
Poster-68
Poster-150
Poster-211
Poster-101
Poster-212
Poster-228
Poster-221
Poster-102
Poster-240
Poster-151
Cardioepidemiology I
Children and adolescent health epidemiology I
Poster-183
Poster-83
Poster-187 Poster-193
Poster-105 Poster-120
Poster-201 Poster-30 Poster-58 Epidemiology of elderly
Poster-281
Poster-229 Poster-288
Cardioepidemiology II
Children and adolescent health epidemiology II
Poster-64
Poster-32
Poster-205
Poster-67
Poster-210
Poster-255
Poster-82
Poster-216
Poster-189
Poster-91
Poster-55
Poster-274
Poster-177
Poster-147
Poster-290
Poster-94
Poster-245
Poster-153
Infectious diseases epidemiology I
Poster-41
Poster-292
Poster-214
Poster-297
Poster-301
Poster-99
Poster-232
Poster-44
Poster-33
Poster-38
Poster-258
Poster-47
Poster-95
Poster-56
Poster-289
Poster-62
Poster-135
Poster-163
Poster-168
Poster-140
Poster-164
Poster-173
Poster-169
Poster-188 Poster-298
Health services research I Poster-158
Social epidemiology I Poster-198
Poster-155
Poster-270
Poster-253
Poster-161
Poster-42
Poster-251 Poster-220
Methods
Poster-200 Environmental epidemiology Poster-52 Poster-61 Poster-286 Health services research II Poster-85
Infectious diseases epidemiology II
Poster-181 Clinical epidemiology Poster-84 Poster-199 Social epidemiology II Poster-252
Poster-208
Poster-222
Poster-60
Poster-235
Poster-227
Poster-257
Poster-126
Poster-113
Lifecourse epidemiology
Poster-148
Poster-295
Poster-285
Poster-115
123
Occupational epidemiology
Poster-203
Poster-215 Poster-218
IEA-EEF European Congress of Epidemiology 2009
Oral Presentations Thursday, 27 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Lifecourse epidemiology I 213 IS POSTNATAL GROWTH ASSOCIATED WITH BLOOD PRESSURE IN CHILDHOOD? N. M Davies1,2, K. Tilling1, F. Windmeijer2, G. D. Smith1, Y. Ben-Shlomo1, M. S. Kramer3, R. M. Martin 1MRC Centre (CAiTE) Department of Social Medicine, University of Bristol, BRISTOL, United Kingdom, 2Centre for Market and Public Organisation, Department of Economics, University of Bristol, BRISTOL, United Kingdom, 3Department of Pediatrics and Department of Epidemiology, Biostatistics and Occupational Health, McGill University Faculty of Medicine, MONTREAL, Canada Background: It is hypothesised that there are critical periods in foetal life and infancy in which an individual’s growth rate determines later health outcomes. Objectives: To investigate whether variations in growth patterns in early life are associated with blood pressure at 6.5 years by modelling individual growth trajectories between birth and 5 years of age. Methods: We fitted a linear spline random-effects model with 2 knots thus dividing follow-up into 3 time periods, each with its own trajectory. The spline models were used to estimate 4 random-effects coefficients: birthweight; ‘early infant weight velocity’ (birth—3 mo); ‘late infant weight velocity’ (3 months–1 years) and ‘childhood weight velocity’ (1–5 years). Together, the coefficients are a summary of each child’s growth curve from birth to 5 years. The outcome was blood pressure at 6.5 years. OLS was used to estimate associations of each coefficient with blood pressure, adjusted for hospital and baseline confounders. The data were from 10,494 children from Belarus who were born in one of 31 hospitals that participated in a cluster-randomised trial of a breastfeeding promotion intervention. Each child had up to 13 measures of height and weight from birth to age 6.5 years. Results: Birthweight and weight velocity at all 3 timeperiods were positively associated with blood pressure. The change in systolic blood pressure per z-score increase in growth was 0.27 (95% confidence interval 0.07–0.47) for birthweight; 0.47 (0.26–0.68) for ‘early infant weight velocity’; 0.70 (0.50–0.90) for ‘late infant weight velocity’ and 0.96 (0.76–1.16) for ‘childhood weight velocity’. There was no evidence of differences by gender (interaction P-value[ 0.05). Conclusions: Children’s growth trajectories between birth and 5 years were positively associated with blood pressure at 6.5 years. Associations increased in magnitude with age.
296 LOW BIRTH WEIGHT AND OTHER VASCULAR-RELATED OUTCOMES OF PREGNANCY AND SOCIOECONOMIC FACTORS IN MOTHERS OF A PORTUGUESE BIRTH COHORT E. Alves1,2, A. Azevedo1,2, H. Barros1,2 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health - University of Porto, PORTO, Portugal Background: Socioeconomic position is a consistent determinant of cardiovascular health, but its effect on vascular-related outcomes of
19 pregnancy is not established. Objectives: To assess the association between low birth weight and other vascular-related outcomes of pregnancy and indicators of mother’s socioeconomic status. Methods: Within a population-based cohort study, 8125 pregnant women were recruited in all public hospitals of Porto in 2005–2006. Vascularrelated outcomes of pregnancy (combined endpoint including gestational hypertension, eclampsia/pre-eclampsia, gestational diabetes and low birth weight newborn) were considered according to clinical records. Mother’s and her parents’ educational levels, income, employment status, marital status and owning one’s house were used as socioeconomic factors. After exclusion of 132 multiple pregnancies and those with missing values, 977 affected by any of these outcomes were considered cases and the remaining 5849 taken as controls. Odds ratios (OR) and 95% confidence intervals (CI) were estimated using logistic regression. Results: Overall, 14.3% (95%CI 13.5–15.2) of women had a vascular-related outcome of pregnancy. The crude OR (95%CI) were 0.78 (0.68–0.90) for education [9 years; 0.88 (0.77– 1.02) for parents’ education C4 years; 0.81 (0.61–1.08) for monthly income C1500€ and 0.72 (0.58–1.02) for women that refused to report income, in comparison with \500€; 0.87 (0.75–1.02) for unemployment; 0.73 (0.53–1.01) for single mothers; 0.93 (0.80–1.07) for owning one’s house. When adjusting for age and body mass index, the prevalence of the endpoint was inversely associated with education [9 years (OR = 0.76; 95%CI 0.66–0.88) and income C 1500€ (OR = 0.67; 95%CI 0.48–0.93). When adjusting for each other, age and body mass index, the effect of education remained unchanged (OR = 0.79; 95%CI 0.66–0.93), while that of income was explained by education (OR = 0.78; 95%CI 0.55–1.10). Conclusions: This study suggests that low educational levels are the most important socio-economic determinant of low birth weight and other vascularrelated outcomes of pregnancy in this population.
262 INFANT NUTRITION IS NOT ASSOCIATED WITH INSULIN RESISTANCE MEASURES IN EARLY ADULTHOOD: THE BARRYCAERPHILLY GROWTH COHORT STUDY D. M. Williams1, R. M. Martin1, G. D. Smith1, D. P. Davies2, K. George3, M. M. Alberti3, Y. Ben-Shlomo1, A. McCarthy4 1 Department of Social Medicine, University of Bristol, BRISTOL, United Kingdom, 2Department of Child Health, University of Wales, College of Medicine, CARDIFF, United Kingdom, 3Department of Endocrinology and Metabolism, St Mary’s Hospital, Imperial College, LONDON, United Kingdom, 4Child Health Epidemiology Division, Health Research Board, DUBLIN, Ireland Background: Previous studies have suggested that infant nutrition may be involved in the metabolic programming of long-term susceptibility to insulin resistance. Aims: To test the hypothesis that increasing dried formula milk intake during infancy is inversely associated with insulin sensitivity and secretion measures in early adulthood (23–27 years). Methods: We conducted a prospective observational follow-up of 679 young adults, aged 23–27 years, who had originally been enrolled in the Barry Caerphilly Growth study at birth between 1972–1974. The primary outcomes of interest were insulin sensitivity (ISI0) and insulin secretion (CIR30) measures in young adulthood according to quartiles of dried milk consumption recorded by nurses during regular home visits in early infancy. Results: There was little evidence of any association between increasing quartiles of dried milk consumption during early infancy (\3 months) and fasting glucose, fasting insulin, insulin sensitivity or insulin secretion in adulthood. There was evidence of a positive association of dried milk consumption at
123
20 3 months of age with waist circumference in adulthood, corresponding to a 1.56 cm (95% CI: 0.64, 2.54 cm) increase in waist circumference per quartile of formula milk consumption after adjustment for sex, intervention group and age at follow-up. This association showed negligible attenuation after further adjustments for birth weight z score and social class in childhood (1.54 cm; 95% CI: 0.59, 2.49 cm). Conclusions: Our findings suggest no influence of early feeding habits on the long-term risk of developing insulin resistance, despite an observed effect on central adiposity in young adulthood. It is possible that the reported protective benefits of breast-feeding on insulin resistance and diabetes incidence in adulthood emerge later in adult life or they may have previously been exaggerated due to publication bias.
Early Morning Session 2—Biostatistics 160 MULTIPLE INDICATORS OF POLY DRUG USE USING STRUCTURAL EQUATION MODELING (SEM) V. Siciliano, S. Molinaro, National Research Council, Institute of Clinical Physiology, PISA, Italy Background: An important aspect to recognize in researching drug prevalence is poly drug use. Most drug users do not limit themselves to a single substance and typically use two or more drugs; in particular those who use illicit drugs (i.e. cannabis, cocaine or other hard drugs) also tend to use alcohol or cigarette. Objectives: The aim of this study is concentrated in the detailed relationship of poly drug use (expressed as underlying latent construct) to personal attitudes and/or interpersonal relationships, among adolescents. Methods: Data from ESPAD-Italia2005 database were provided by the National Research Council and a structural equation model was fitted with AMOS 16. A representative sub-sample of 4000 students 18 years old was randomly selected. Confirmatory-factor analysis considers 4 latent endogenous variables (licit drug use; cannabis and licit drug use; sedative and licit drug use; illicit drug use) as the underlying factors that explain the correlations between the observed multiple indicators of illicit and licit drug use. Results: Concerning licit drug use, high correlation were found with ‘‘absence of parental monitoring’’. For males it’s also related to ‘‘absence from school’’ (b = 0.12; P \ 0.01) and for females to ‘‘high family economic-social status’’ (b = 0.18; P \ 0.001). Cannabis and licit drug use was positively associated with ‘‘having many friend who use illicit drug’’ (b = 0.33; P \ 0.001), ‘‘satisfied with friends’’ and ‘‘go out most evenings’’ (b = 0.25; P \ 0.01). Sedative and licit drug use was negatively associated with ‘‘satisfied with interpersonal relationships’’ (b = -0.25; P \ 0.001) and positively associated with ‘‘illicit drug use’’. Concerning illicit drug use, correlations with ‘‘high family economic-social status’’ for males and with ‘‘go out most evenings’’ for females were found. Conclusions: This study has allowed the SEM have been used successfully and have great potential to be used in helping researchers to understand the complex relationships among various factors, playing an important role in the identification of possible interventions for drug use prevention.
250 DIETARY PATTERNS IN THE POPULATION OF PORTO USING MIXED CONTINUOUS AND BINARY VARIABLES R. Gaio1, J. P. da Costa2, A. C. Santos3, E. Ramos3, C. Lopes3 1 Department of Pure Mathematics, University of Porto Science School, PORTO, Portugal, 2Department of Applied Mathematics,
123
IEA-EEF European Congress of Epidemiology 2009 University of Porto Science School, PORTO, Portugal, 3Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal Background: Mixture models can be applied for the construction of dietary patterns, and have several advantages over the traditional cluster or principal component analysis. However, estimation difficulties may arise in the presence of very skewed or jumpy distributions. Objectives: To identify dietary patterns for the population of Porto using mixture models. Methods: As part of an ongoing Portuguese cohort study, participants were selected by random digit dialing. Information was collected by trained interviewers, using a structured questionnaire. Diet information was collected using an 86-item semi-quantitative food frequency questionnaire, for a total of 1488 women and 922 men. Separate, but parallel, analyses were carried out according to sex. Among 14 selected food groups, 5 were dichotomized due to either a high frequency ([45%) of food nonconsumption or a graphical evidence for an ordinal nature. Continuous variables were square-root transformed. For each food group, outliers were identified and eliminated from the study. Dietary patterns were constructed using conditional finite mixture models: the a priori probability was made conditional on age (3-level) and total energy intake, while the densities within each cluster multivariate normal distributions with diagonal variance-covariance matrices) were conditioned by calories only. Cluster parity was determined by the smallest Bayesian information criterion. Results: Preferred models for men and women were the 3 and 4-cluster solutions, respectively. Correspondent cluster profiles were ‘‘Healthy’’, ‘‘Meat/Alcohol’’ and ‘‘Sweets/Fastfood’’ for men, ‘‘Healthy’’, ‘‘Meat/Alcohol’’, ‘‘Sweets/ Fastfood’’ and ‘‘Moderate Consumption’’, for women. Average posterior probabilities within clusters ranged 85%-93%, and 76%-86%, respectively. Conclusions: The present methodology allows the identification of consistent dietary patterns for the Portuguese population. Advantages of this methodology arise from the fact that it is probabilistic and scale-invariant, takes into account the different scale types of the food groups, and allows for direct covariate(s) adjustment.
145 SPATIAL ANALYSIS OF DRUG-RELATED HOSPITAL ADMISSIONS: AN AUTO-GAUSSIAN MODEL TO ESTIMATE THE HOSPITALIZATION RATES IN ITALY E. Colasante, S. Molinaro, F. Mariani, National Council of Research, Institute of Clinical Physiology, ROME, Italy Background: The aim of this study is to evaluate, even if partially, how much the drug use phenomenon impacts on the Italian National Heatlh System throughout the estimation at local level (Local Health Unit) of the hospitalization rate caused by substance use and abuse such as opiates, barbiturates-sedatives-hypnotics, cocaine and cannabis, and keeping in mind the phenomenon distribution in the space and so the fact that what happens in a specific area depends on what is happening in the neighbourhoods close to it (spatial autocorrelation). Methods: Data from hospital discharge database were provided by the Ministry of Health and an auto-Gaussian model was fitted. The spatial trend can be a function of other explanatory variables or can simply be modeled as a function of spatial location. Both models were fitted and compared using the number of subjects kept in charge by Drug Addiction Services and the number of beds held by hospitals as covariates. Results: Concerning opiates use related hospitalizations, results show areas where the phenomenon was less prominent in 2001 (Lombardy, part of Liguria, Umbria, part of Latium, Campania,
IEA-EEF European Congress of Epidemiology 2009 Apulia and Sicily). In the following years, the hospitalization rates increased in some areas, such as the north of Apulia, part of Campania and Latium. A dependence of the opiates related hospitalization rates on the rate of subjects kept in charge by the Drug Addiction Services is highlighted. Concerning barbiturates-sedatives-hypnotics consumption, the best model is the one without covariates and estimated hospitalization rates are lower then 3 per thousand. The model with only the covariate ‘‘rate of subjects kept in charge by Drug Addiction Services’’ has been used both for cocaine and cannabis. In these two cases, more than a half of the Local Health Units report hospitalization rates lower than 0.5 per thousand. Conclusions: This study has allowed for the development of an indirect indicator for the phenomenon of drug use and it constitutes an effort in answering specific needs for the planning of health policies related to a field that, given the specificity of the phenomenon, is often difficult to detect and quantify by means of more common data analysis techniques. Moreover, it is important to highlight that, being this study a first attempt in applying this statistical methodology to data regarding drug addiction, it needs to be further improved.
Early Morning Session 3—Clinical epidemiology I 129 DRUG USE IN POPULATION SCREENING: A PHARMACOEPIDEMIOLOGICAL ASPECTS J. Atthobari1, S. T. Visser2, R. Gansevoort3, P. E. de Jong3, L. T. W. de Jong-van den Berg2 1Department of Pharmacology and Toxicology, Faculty of Medicine, Gadjah Mada University, YOGYAKARTA, Indonesia, 2Groningen Research Institute of Pharmacy, Groningen University, GRONINGEN, The Netherlands, 3 Department of Internal Medicine, University Medical Centre Groningen (UMCG), GRONINGEN, The Netherlands Background: The population based screening has been used as a basis for studies of drug use. Access to drug information from pharmacy database and link it to the health screening data may contribute several aspect of pharmacoepidemiologic research. Objective: This study aimed to know whether the population based screening (as predictor variable) could influence drug prescription and to assess the potential effects of a specific drug use (hormone contraceptives (HC) and statins) on population based screening (as outcome variable). Methods: We used the clinical data and 4.2 years follow-up data of the 8592 subjects with high risk for cardiovascular disease (enriched) who were participated in screening of the PREVEND (Prevention of REnal and Vascular ENd-stage Disease) cohort study. This clinical data were linked to the electronic pharmacy data from Inter-Action Data Base (IADB). Results: Compared with the non-screening group, screening-related drugs (antihypertensive, antilipidaemic, antidiabetic and antithrombotic) were prescribed more frequently for subjects in the screening group (HR = 1.20; 95% CI 1.13–1.29), whereas screening-unrelated drugs were used a comparable frequencies (HR = 1.04; 95% CI 0.99– 1.04) after 4.2 years of follow-up. On long term use, HC associated with worsening blood pressure, albuminuria and renal function. On the other hand, the cessation of HC resulted in an improvement of those outcomes. The statins use induced a rise in albuminuria (RRadj = 1.49; 95%CI 1.07–2.08), especially in higher dose RRadj = 1.71; 95%CI 1.11–2.62), and for longer duration RRadj = 1.70; 95%CI 1.07–2.82). Conclusions: A population based screening could be as a determinant and outcome factors of drug use. Use of large computerised pharmacy databases and record linkage has become increasingly important in pharmacoepidemiology research.
21
36 DEVELOPMENT OF A MELANOMA RISK SCORING TOOL AND VALIDATION ACROSS DIFFERENT ETHNIC POPULATIONS C. Fortes1, S. Mastroeni1, L. Bakos2, G. Antonelli1, Livia, M. Pilla1, M. Allotto1, R. Bonamigo3, P. Pasquini1, C. F. Melchi4 1Clinical Epidemiology Unit, IDI-IRCCS, ROME, Italy, 2 Department of Dermatology, Hospital de Clı´nicas de Porto Alegre, Federal University of Rio Grande do Sul, PORTO ALEGRE, Brazil, 3 Department of Dermatology, Fundac¸a˜o Faculdade Federal de Cieˆncias Me´dicas, PORTO ALEGRE, Brazil, 4VIII Dermatology Unit, IDI-IRCCS, ROME, Italy Background: Melanoma is an increasingly common malignancy of melanocytes, with incidence rates steadily increasing over the past several decades. Even if malignant melanoma incidence is still increasing, it is a relatively rare malignancy and therefore a screening campaign aimed to the entire population may show little return. Nevertheless, if individuals at high risk could be identified, surveillance could be targeted at this group. Objective: The objective of this study was to develop and validate an individual risk score for cutaneous melanoma across different ethic populations. Methods: With the summary coefficients of the risk factors for cutaneous melanoma, derived from a meta-analysis, a melanoma risk score (MRS) was developed. Common nevi, skin and hair and colour, freckles and sunburns in childhood were the variables included in the final predictive model. Two case-controls conducted in different countries (Italy, Brazil) were used to evaluate the predictive ability of the melanoma risk model The Receiver Operator Characteristic (ROC) analysis were used to evaluate the accuracy of the risk model. The test for equality was used to compare the Areas under the ROC curve (AUCs) of different risk models. Results:. The Area under the ROC curve (AUC) of the candidate model (MRS) was 0.79 (95% CI: 0.75– 0.83). The same model applied in the Brazilian population presented an AUC of 0.79% (95% CI.0.75–0.83). At the cut-off level of three and more, 89% of the cases were classified as ‘‘at risk for melanoma’’ in the Italian study and 80% in the Brazilian population. Conclusion: The MRS is a simple tool that identify subjects for preventive measures and may be used with reasonable confidence in different populations. MRS may help family doctors in referring patients to dermatological clinics and thus enhance early diagnosis.
264 THE UNRECOGNIZED PREVALENCE OF CHRONIC KIDNEY DISEASE AMONG FAMILY MEMBERS OF END STAGE RENAL DISEASE PATIENTS S. S. Abd El-Hafeez1, Z. M. Gad1, Hala S El-Wakil2, M. H. Hashsish1 1 High Institute of Public Health, University of Alexandria, ALEXANDRIA, Egypt, 2Faculty of Medicine, University of Alexandria, ALEXANDRIA, Egypt Background: The health burden of the renal disease is increasing world wide. The global challenge over the next decade is to shift the emphasis away from treatment of end stage renal disease (ESRD) to the early detection and prevention of chronic kidney disease (CKD). Several studies have reported that relatives of ESRD patients are at high risk for CKD, but little is known about the prevalence of CKD and associated epidemiological characteristics among this group. Objectives: To determine the prevalence of CKD among first and second degree adult relatives of ESRD patients, and to study the epidemiological characteristics of the CKD cases found among the
123
22 relatives of ESRD. Methods: A cross sectional study was conducted by voluntary screening of 400 subjects of the first and second degree adult (18 years or older) relatives of ESRD patients. Relatives were recruited from the different seven districts in Alexandria. The relatives were subjected to questionnaire interview (Data about age, gender, social class, personal and family histories of diabetes mellitus, hypertension, CKD were collected). Measurement of blood pressure, weight and height, and laboratory investigations including spot urine sample for detection of proteinuria by dipstick and serum creatinine determination to estimate GFR by MDRD equation. Results: The prevalence of CKD among the studied relatives was 57% and the proteinuria (?1 or more) was found among 21%. After logistic regression analysis, it was found that old age, hypertension, and previous history of kidney problems as stones are the risk factors for CKD among the studied relatives. Conclusion: CKD is quite prevalent among first and second degree relatives of ESRD patients. This high risk group might benefit from interventions to improve the detection and management of risk factors for progressive CKD for combating the current epidemic of ESRD especially in our developing country.
Early Morning Session 4—Occupational epidemiology I 123 WOMEN WITH ANAEMIA IN PREGNANCY GET HEAVIER BABIES IN A HEAVILY POLLUTED TOWN IN THE RUSSIAN ARCTIC: A REGISTRYBASED STUDY ON 23,950 SINGLETON BIRTHS, 1973–2001 E. Chumak1, A. M. Grjibovski1,2 1International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia, 2Norwegian Institute of Public Health, OSLO, Norway Background: previous studies have suggested that both too low and too high haemoglobin levels in pregnant women are associated with adverse pregnancy outcomes including foetal growth restriction. The prevalence of maternal anaemia in Russia reached almost 90% during economic crisis of the 1990 s was. Objectives: to study the effect of anaemia in pregnancy on foetal growth in a Russian arctic town. Methods: The study was performed in a heavily polluted by nickel and copper industry arctic town of Monchegorsk (65560 N; 32540 E). Data on all singleton births from 1973 through 2001 with known data on maternal haemoglobin levels during pregnancy were obtained from the Kola Birth Registry (n = 23,950). Maternal anaemia was defined as haemoglobin level below 120 g/l. Multiple linear regression was used to estimate individual effects of maternal anaemia on birthweight with adjustment for potential confounders. Additional adjustment for gestational age was made to separate the effects of anaemia on foetal growth from its effect on preterm delivery. Crude and adjusted birthweight differences with 95% confidence intervals (CI) were calculated. Results: The prevalence of anaemia in pregnancy in Monchegorsk increased from 47% in 1973 to 89% in 2001. Babies born to mothers with anaemia in pregnancy were heavier by 81 gram (95% CI: 67 - 95) than babies born to mothers with no anaemia. Adjustment for maternal occupation, age, marital status, parity, smoking, alcohol consumption, gestational age and year of delivery reduced these differences to 47 gram (95% CI: 35–59). Conclusions: the findings suggest that anaemia in pregnancy defined as haemoglobin level below 120 g/l is associated with slightly more favourable foetal growth in a heavily polluted town in the Russian Arctic. Potential mechanisms behind the observed association will be discussed.
123
IEA-EEF European Congress of Epidemiology 2009
76 HYPERTENSION IN WORKING POPULATION IN CROATIA- PREVALENCE AND CONTROL RATE M. Gomzi1, B. Winterhalter-Zvonar2 1Institute for Medical Research and Occupational Health, ZAGREB, Croatia, 2Health Centre, ZAGREB, Croatia Background: Hypertension, the important cardiovascular risk, was increasing in Croatian working population, while rates of hypertension control continued to be low. Objectives: The aim of the present study was to investigate the prevalence of arterial hypertension in the continental Croatian working population and to estimate the awareness and control of hypertension in relation to age, gender and certain aspects of occupation. Methods: The study comprised two cross-sectional surveys on arterial hypertension in working population. First survey included 26 643 workers from manufacturing, administrative and service sector in continental Croatia (mean age 36.4 years, range 18–56 years, 57% male). In second survey authors sampled 1123 workers from the first survey undergoing routine medical check-up, which involved a structured questionnaire and physical examination. Socioeconomic status of the study subjects was estimated by assessing their education level and type of occupation. Descriptive techniques and multivariate logistic regression were used. Results: In the first survey high blood pressure (C 140/90 mm Hg) had been diagnosed by occupational health service in 12,9% of workers. Women were hypertensive more often than men (13,8%:12,2%). Hypertension prevalence increased across the gradient from service sector to office workers to manufacturing: 9, 15 and 19 percent, respectively. Routine check-up in the second survey showed that 31.8% (33.7% of males and 29,3% of females) had high blood pressure. The rates of awareness, treatment and control tended to increase with age, and decrease with education level. Conclusions: The results implied that the blood pressure in the Croatian working population was not adequately controlled, particularly in younger workers, in workers with lower education level and in certain types of occupation. The evidence of occupational factors contributing to workers’ cardiovascular disease may have public health benefits—the introduction of health measures to enhance the knowledge about potential risks and rates of treatment.
131 CARCINOGENIC RISK IN BLACK-COAL MINERS WITHOUT PNEUMOCONIOSIS IN THE CZECH REPUBLIC—LONGITUDINAL STUDY (1992–2006) H. Tomaskova1,2, Z. Jirak2, A. Splichalova1, P. Urban3, V. Zavadilova2 1Institute of Public Health Ostrava, OSTRAVA, Czech Republic, 2University of Ostrava, Faculty of Health Studies, OSTRAVA, Czech Republic 3National Institute of Public Health, PRAHA, Czech Republic Background: In black coal mines, the main risk factor is dust containing crystalline silica that was classified by IARC as a human carcinogen in 1997. Objectives: In this longitudinal epidemiological study, carcinogenic risk was analysed in the cohort of 6,705 blackcoal ex-miners without pneumoconiosis. Methods: The sample included miners without pneumoconiosis who were working in the mining profession for at least 8 years. The data on individual and occupational history were combined with the data from the National Cancer Register and the National Population Register. The relative
IEA-EEF European Congress of Epidemiology 2009 risk of cancer (Cancer of lung, stomach, colon, kidney, bladder) was calculated as SIR (Standardized Incidence Ratio) over the period 1992–2006 between ex-miners and general Czech population. Cox proportional hazards model was used for analysis of relationship between lung cancer and smoking habits and duration of professional exposure controlled for age. Stata software was used for the data analysis. Results: In the cohort, 464 newly registered cases of cancer and 681 deaths were found in the period 1992–2006. The main causes of death were cardiovascular diseases (32%) and cancer (30%). SIR was not statistically significant higher for any specific cancer (Cancer
23 of lung N = 93, SIR = 0.87, stomach N = 20, SIR = 1.15, colon N = 35, SIR = 0.87, kidney N = 24, SIR = 0.66, bladder N = 19, SIR = 0.72). Data on smoking habits were available from 67% of workers of those 80% were smokers or ex-smokers. Average duration of exposure was 22.9 years (SD = 5.9). Only relative risk of lung cancer for smokers, controlled for age, was significant higher (HR = 3.22, 95% IC 1.66 - 6.25). Conclusions: The risk of specific cancers in miners without pneumoconiosis was comparable to the risk in general population. The main risk factor for lung cancer was smoking.
123
24
Thursday, 27 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology I 175 BIOMARKERS EXPRESSION IN INTERVAL AND SCREEN-DETECTED BREAST CANCERS IN A BREAST CANCER SCREENING PROGRAMME IN BARCELONA L. Domingo1, S. Servitja2, J. Corominas3, J. Martı´nez4, F. Ferrer4, F. Macia`1, J. Albanell2, X. Castells1, M. Sala1 1Evaluation and Clinical Epidemiology Service, IMIM-Hospital del Mar, BARCELONA, Spain, 2Oncology Service, IMIM-Hospital del Mar, BARCELONA, Spain, 3 Anatomical Pathology Service, IMIMHospital del Mar, BARCELONA, Spain, 4Image Diagnostic Service, IMIM-Hospital del Mar, BARCELONA, Spain Background: Interval cancers (IC) are breast cancers diagnosed after a negative screening mammography and before the following one. It was previously reported this group of cancers presents worse prognostic features than screen-detected cancers (SDC). This difference could be attributed to the biological characteristics of the tumours. Objectives: To compare biomarkers expression and prevalence of different phenotypes between IC and SDC and to evaluate their relation with tumour characteristics at diagnosis and clinical evolution at 2-year follow-up. Methods: Case-case study of women enrolled in the breast cancer screening program of Barcelona City, which presented IC or SDC between 1995–2008, matched on age and year of diagnosis. In tumour samples, p53, oestrogen receptor (ER), progesterone receptor (PR), and HER2 were examined with immunohistochemical methods. A bivariate analysis was performed to compare clinical and biological characteristics among both groups and logistic regression analysis to asses the adjusted risk of interval cancer. Results: We have studied 98 IC and 98 SDC. Family history of breast cancer (20.4%IC, 11.2%SDC; P = 0.078) and higher proportion of dense breasts (52.6%IC, 34.1%SDC; P = 0.017) were more frequently observed in women with IC. Tumours in the IC group presented lower expression of either ER and HER2. Triple-negative (TN) phenotype (ER-, PR-, HER2-) was more frequent in IC than in SDC group (20.2%, 6.4%, respectively = 0.006). At diagnosis, IC were larger (P \ 0.001) presenting more affected lymph nodes than SDC (43.7%, 25.0%, respectively = 0.009). TN phenotype is shown as the main independent variable for IC (OR = 3.60;95%CI = 1.12–11.52). IC presented more loco-regional relapses and distant metastasis than SDC (17.5%, 7.1%, respectively = 0.027). Conclusions: IC present more aggressive features than SDC. TN phenotype and high breast density are the strongest factors associated with IC. The study of IC could provide useful information to improve early detection. However, a complete IC classification is needed to fully understand the role of these biomarkers.
97 HORMONE REPLACEMENT THERAPY USE AND INCIDENCE OF CENTRAL NERVOUS SYSTEM TUMOURS; THE MILLION WOMEN STUDY V. Benson, K. Pirie, J. Green, G. Reeves, V. Beral for the Million Women Study Collaborators Cancer Epidemiology Unit, University of Oxford, OXFORD, United Kingdom
123
IEA-EEF European Congress of Epidemiology 2009 Background: There are approximately 48,000 people diagnosed with a tumour of the central nervous system (CNS) annually in Europe. They have poor prognosis with 39,000 deaths per year, yet little is known about risk factors of the tumours, although some previous studies have suggested that hormone replacement therapy (HRT) use might increase the risk. Objectives: To investigate the role of HRT and the risk of CNS tumours, by histological subtype. Methods: Over 1.1 million postmenopausal women recruited in 1996–2001 in the UK were followed for incident tumour registration. All CNS tumours and each of the tumour types (glioma, meningioma, and acoustic neuroma) were separate end points in a Cox proportional hazards model for investigation of various measures of HRT use. Preliminary Results: During an average of 5 years of follow-up, a total of 1,132 postmenopausal women with HRT information were diagnosed with a tumour of the CNS; 497 were gliomas, 270 meningiomas, and 112 acoustic neuromas. Preliminary findings show that current users of HRT are slightly more likely to develop a CNS tumour when compared to never users (RR = 1.15, 95% CI = 1.00–1.32). Past users of HRT were not at a significantly increased risk of developing a CNS tumour. Findings by type of HRT use and by the histological subtype of the tumours will be presented. Conclusion: Current users of HRT are at a slightly increased risk of developing CNS tumours.
280 RISK OF COLORECTAL CANCER AFTER RESECTION OF ADENOMAS V. Cottet1,2, V. Jooste V1,2, A. M. Bouvier1,2, J. Faivre1,2, C. Bonithon-Kopp1,3 1INSERM U866, DIJON, France, 2Registre Bourguignon des cancers digestifs, DIJON, France, 3INSERM CIE 1, CHU de Dijon, DIJON, France Background: Colonoscopic removal of adenomas is recognized as the most efficient way to prevent colorectal cancer, although the actual benefit in routine clinical practice is yet poorly assessed. Objectives: The aim of this population-based study was to estimate the risk of colorectal cancer after adenoma removal in routine practice, in comparison with the general population. Methods: This study relies on the only population-based Registry which collects since 1976 all cases of polyps and colorectal cancer diagnosed in a well-defined administrative area in France. All patients residing in Coˆte-d’Or area (506,755 inhabitants), who were diagnosed with a first colorectal adenoma or carcinoma in situ between January 1990 and December 1999 were included and followed-up until the end-point date (December 2003). The observed numbers of cases of colorectal cancer among the study population were compared with the expected numbers calculated on the basis of age- and gender- specific rates in the general population using data from the Burgundy Registry of Digestive Cancers. The ratio of observed to expected cases was reported as a standardized incidence ratios (SIR). Results: A total of 5,782 adenoma patients were included and 88 colorectal cancers were detected during a follow-up of 39,730 persons-years. The expected number of colorectal cancers in this population was 69. In comparison with the general population, the risk of colorectal cancer after firstadenoma removal was significantly increased (SIR = 1.28 [Confidence Intervals 95%: 1.02–1.57]), especially for advanced adenomas (SIR = 2.11 [CI 95%: 1.61–2.73]). Removal of non-advanced lesions was associated with a slightly decrease in the incidence of colorectal cancer (SIR = 0.71 [CI 95%: 0.47–1.01]). Conclusion: In actual conditions of clinical practice, the risk of developing colorectal cancer after adenoma removal remained greater than in the general population. Both the high individual carcinogenetic risk and shortcomings in follow-up probably contribute to these findings.
IEA-EEF European Congress of Epidemiology 2009
3 RECURRENT SORES BY ILL-FITTING DENTURES AND INTRA-ORAL SQUAMOUS CELL CARCINOMA IN SMOKERS J. L. F. Antunes1, G. F. Vaccarezza1, P. Michaluart-Ju´nior2 1 University of Sao Paulo, School of Dentistry, SAO PAULO, Brazil 2 University of Sao Paulo, School of Medicine, SAO PAULO, Brazil Background: Mouth cancer is the most incident neoplasm of the head and neck worldwide. Differences in incidence are well documented to associate with gender, age, tobacco smoking, alcohol consumption and dietary habits. However, the assessment of specific conditions of dental status has not produced conclusive results. Objectives: To examine whether denture use and recurrent sores caused by ill-fitting dentures are associated with mouth cancer in individuals exposed to tobacco. Methods: We conducted a hospital-based case-control study comprising 124 patients (current smokers and those who quitted after the diagnosis) with intra-oral squamous cell carcinoma attended at the head and neck surgery center of a large public hospital in Sao Paulo, Brazil (2006–2007). Controls were recruited from outpatient-units of the same hospital. Gender and age were controlled by individual pairing. The cumulative exposure to tobacco and alcohol was assessed using the methodology standardized by the International Agency for Research on Cancer. Dietary habits were assessed using a food frequency questionnaire specifically devised for a previous study, and considered items usually consumed in Brazil. Conditional logistic regression analysis assessed the effect of denture use and recurrent oral sores by ill-fitting dentures, as adjusted by covariates on the lifetime exposure to alcohol and tobacco, socioeconomic standings and dietary patterns. Results: The use of dentures did not associate with mouth cancer (Adjusted Odds Ratio 1.40, 95% confidence interval 0.51–3.87, P = 0.513). However, the report of recurrent sores caused by ill-fitting dentures showed significant association with the disease (Adj. OR 4.58, 95%CI 1.52–13.76, P = 0.007). Conclusions: The association between recurrent oral sores caused by ill-fitting dentures and mouth cancer in smokers reinforces the hypothesis that the chronic physical irritation of oral mucosa contributes to the topical carcinogenic effect of tobacco, which must be taken into careful consideration in the planning of dental services for adults and the elderly.
Late Morning Session 2—Perinatal epidemiology I 24 HIGH PERINATAL MORTALITY RATE AMONG IMMIGRANTS IN BRUSSELS J. Racape´1, M. De Spiegelaere2, S. Alexander1, M. Dramaix1, P. Buekens3, E. Haelterman1 1School of Public Health,Universite´ Libre de Bruxelles, BRUXELLES, Belgium, 2Health and Social Observatory of Brussels, BRUXELLES, Belgium, 3School of Public Health and Tropical Medicine, Tulane University, NEW ORLEANS, United States of America Background: A number of studies refers to the ‘‘Epidemiological paradox’’: despite their low socio economic status, immigrants have good pregnancy outcomes. However, perinatal mortality rates vary between ethnic groups, with minorities repeatedly having higher mortality rates. Objectives: Describe and measure inequalities in perinatal mortality according to maternal nationality and socio-economic status in Brussels. Methods: This study is a population-based cohort study using the data from linked birth and death certificates from the Belgian civil registration system. Data are related to all babies born during the 9-year period 1998–2006 and whose mother was living in Brussels, irrespective of the place of delivery (137 974 newborns). Perinatal and post-perinatal mortality were analysed
25 according to the nationality and the socio-demographic characteristics of the mother at birth. We used logistic regression to estimate the odds ratios (ORs) for the association between mortality and nationality. Results: Women of sub-Saharan Africa experience a 50% excess in perinatal mortality that mainly reflects a high rate of preterm deliveries, low birth weight and a low socio-economic level. Paradoxically, despite their favourable pregnancy outcomes, Maghrebians and Turkish experience a strong excess (50–70%) of perinatal mortality. Differences in age, parity distributions and multiple birth play no role, and the excess does not reflect low socio-economic level. This excess of perinatal mortality contrasts with the absence of excess of infant mortality. Conclusion: In Brussels, patterns of inequalities in perinatal mortality vary according to nationality and perinatal mortality is increased in particular ethnic groups independently of the socioeconomic status. We suggest that suboptimal care and low access to care may contribute to differences in perinatal mortality among nationalities.
194 DURATION OF RESIDENCE AND ADVERSE BIRTH OUTCOMES AMONG IMMIGRANTS TO URBAN ONTARIO M. L. Urquia1,2, J. W. Frank3, R. H. Glazier1,2,4,5, R. Moineddin4,5 1 Dalla Lana School of Public Health, University of Toronto, TORONTO, Canada, 2Centre for Research on Inner City Health, St. Michael’s Hospital, TORONTO, Canada, 3Scottish Collaboration for Public Health Research and Policy, EDINBURGH, United Kingdom, 4Institute for Clinical Evaluative Sciences, TORONTO, Canada, 5Department of Family and Community Medicine, University of Toronto, TORONTO, Canada Background: Migrant women contribute to more than one fifth of all live births in North America and several European countries. The ‘convergence hypothesis’ suggests that birth outcomes of immigrants would approach the level observed in the host population with increasing time in the new environment. Objectives: To examine the relation between duration of residence and adverse birth outcomes among immigrants to urban Ontario. Methods: We linked a Canadian database of migrants who obtained their permanent residence through 1985–2000 with mother-infant records (2002–2007) of all acute Ontario hospitals. We examined the associations between length of residence (as continuous and in 5-year groups) and singleton low birthweight (LBW) (\2,500 grams), preterm birth (PTB) (\37 completed weeks), and small for gestational age (SGA) (\10th percentile of the most recent sex-specific Canadian reference), among immigrants (N = 83,233). Multilevel logistic models accounted for the clustering of births into maternal countries of birth. Covariate selection was based on directed acyclic graphs. Comparisons with non-immigrants (N = 314,237) were also made. Results: After adjustment, duration of residence was linearly associated with increases in LBW and PTB [5-year adjusted odds ratio (95% confidence intervals): 1.08 (1.03–1.13), 1.14 (1.10–1.19) but not in SGA [0.99 (0.96–1.02)]. Recent immigrants (\5 years) had lower risk of PTB (4.7%) than non-immigrants (6.2%) but those with C15 years of stay were at higher risk (7.4%). These differences persisted after controlling for potential confounders (maternal age, parity, maternal education, marital status, immigrant class, official language knowledge, country of birth) and cohort of arrival. Conclusions: Our findings do not support the convergence hypothesis. The deterioration of birth outcomes of immigrants was driven by a shortening of gestational age with time spent in Canada. Hypothesized pathways involve maternal pre-pregnancy weight, health behaviors, and work conditions. Immigrants’ length of residence should be adopted as an indicator for surveillance.
123
26
78 MATERNAL COUNTRY OF BIRTH VARIATIONS IN CAESAREAN SECTION: ANOTHER CAUSE FOR CONCERN? R. Small, Mother & Child Health Research, La Trobe University, MELBOURNE, Australia Background: Increasing rates of caesarean section are the topic of much general debate, but caesarean section in immigrant women compared with their receiving country counterparts has received relatively little attention to date. Objectives: To describe and compare caesarean section rates in Victoria, Australia for the period 1999–2007 among women born in Australia and those born overseas in ‘nonEnglish speaking’ (NES) countries. Methods: Routinely-collected data from the Victorian Perinatal Data Collection on confinements and births to all Victorian women have been analysed for the period 1999– 2003, (to be updated to include data to 2007 for presentation). For 1999–2003, Australian-born women numbered 235,275 and there were 15 NES-countries with more than 1000 women giving birth (n = 37,310). Results: Similar overall proportions of Australian-born (25.7%) and immigrant women (24.7%) gave birth by caesarean section. However, the proportion of caesarean sections across the 15 maternal NES-countries varied widely, from 15.0% for women born in Cambodia to 37.6% for women born in Italy. Women from seven NES-countries had higher rates and eight had lower rates of caesarean section than their Australian-born counterparts. Stratification by parity and by patient admission status (public/private) had little effect on these maternal country of birth variations. Likewise, analyses restricted to ‘standard primiparas’, (ie to women having first births with no known complicating factors at the time of labour, and therefore considered to be at low obstetric risk for caesarean section), did little to reduce the observed variations. Conclusions: Explanations of these variations in caesarean section by maternal country of birth require further investigation and do not appear to be readily explained by obstetric complications or differences in maternal characteristics. The potential role of communication difficulties, migration experiences and knowledge of maternity care practices may all be relevant, but individual level data on these factors are rarely collected routinely.
275 CHILDBEARING HEALTH AND SERVICE NEEDS OF MIGRANTS AND REFUGEES TO MONTREAL, QUEBEC (CANADA) A. Gagnon1,2, O. Wahoush3, G. Dougherty1,2, J. F. Saucier4,5, C. L. Dennis6, E. Stanger7, B. Palmer8, L. Merry1, D. E. Stewart9 1 McGill University, MONTREAL, Canada, 2McGill University Health Centre, MONTREAL, Canada, 3McMaster University, HAMILTON, Canada, 4Universite´ de Montreal, MONTREAL, Canada, 5Ste. Justine’s Hospital, MONTREAL, Canada, 6University of Toronto, TORONTO, Canada, 7Language Services, Cross-cultural Health & Diversity, Vancouver Coastal Health, VANCOUVER, Canada, 8BC Women’s Hospital and Health Centre, VANCOUVER, Canada, 9University Health Network, TORONTO, Canada Introduction: Childbearing refugee and asylum-seeking women in industrialized countries may have harmful health outcomes and unmet health and social needs. The forced nature of their migration, separation from their families, lack of knowledge of host country language(s) and for some, a precarious immigration status and limited access to healthcare services, increases their vulnerability. Research Questions: (1) Do refugee or asylum-seeking women and their infants, experience more or different harmful childbearing health outcomes than non-refugee immigrant or receiving country-born
123
IEA-EEF European Congress of Epidemiology 2009 women? (2) Are harmful postpartum health outcomes un-addressed by the health care system associated with immigration status? Methods: Multi-site prospective (birth to 4 months) cohort study. Refugee, asylum-seeking, non-refugee immigrant, and Canadian-born women were recruited from hospital postpartum units. Health data were collected from medical records; general health and Background information were obtained through questionnaires. Research nurses collected data on maternal and infant health, services used, and migration history during home visits at 7–10 days and 4-months postbirth. These data were then classified by a nurse expert (blinded to research questions and immigration status) as providing evidence for the existence of a professional concern and whether it had been ‘un-addressed’ or ‘addressed’ by the healthcare system (based on professional practice guidelines). Results: All migrant groups experienced more professional concerns compared to Canadian-born women including: greater postpartum depression risk at 4 months, lack of social support, skipping meals due to lack of resources, and/or not knowing what to do in an emergency. These concerns were addressed less often among all migrant groups, with refugee women having the highest mean rate of un-addressed concerns. Relevance: Knowledge of the extent of need of childbearing women in all migrant groups as well as the response of the healthcare system to those needs will inform both immigration and health policy makers as well as providers of services.
Late Morning Session 3—Cardioepidemiology I 186 TWELVE YEAR TRENDS IN CARDIOVASCULAR RISK FACTORS IN PATIENTS WITH CORONARY HEART DISEASE. RESULTS FROM THE EUROASPIRE I, II AND III STUDIES ¨ NSTER IN THE REGION OF MU C. Prugger, J. Heidrich, J. Wellmann, R. Dittrich, R. Telgmann, ¨ NSTER, S. M. Brand-Herrmann, U. Keil, University of Mu¨nster, MU Germany Background: EUROASPIRE is a series of cross-sectional studies to evaluate secondary prevention of coronary heart disease (CHD) across Europe. The region of Mu¨nster participated in all three EUROASPIRE surveys (1994–1995, 1999–2000 and 2006–2007). Objectives: To investigate time trends of cardiovascular risk factors in coronary patients from the region of Mu¨nster over a period of 12 years. Methods: Patients were interviewed and examined at least 6 months after hospitalisation for CHD (CABG, PTCA, myocardial infarction, ischaemia) using standardised methods. We applied logistic regression models to investigate time trends in cardiovascular risk factors between surveys. Results: In the Mu¨nster region, a total of 392, 402 and 457 patients participated in EUROASPIRE I, II and III. The proportion of smokers (17% in EUROASPIRE I and II, 18% in EUROASPIRE III, P = 0.891) remained virtually unchanged, whereas the proportion of patients with raised blood pressure decreased slightly (61%, 69% and 55%, P = 0.001). The frequency of hypercholesterolaemia decreased considerably from 94% to 84% and 48% (P \ 0.001). A pronounced increase was observed in the proportion of patients with diabetes (14% in EUROASPIRE I and II, 23% in EUROASPIRE III, P \ 0.001) and obesity (23%, 31% and 43% in EUROASPIRE I, II and III, P \ 0.001). The proportion of patients reporting use of antihypertensive (from 80% to 89% and 93%, P \ 0.001) and lipid lowering (from 35% to 67% and 87% P \ 0.001) agents increased markedly between surveys. Conclusions: Over a period of 12 years, adverse changes in risk profiles of coronary patients were observed. The increasing trends in diabetes and obesity are particularly alarming. Despite rising use of cardioprotective medication, the majority of patients fail to reach treatment targets for
IEA-EEF European Congress of Epidemiology 2009 blood pressure and almost half of patients show elevated cholesterol levels. The observed trends emphasise the dire necessity to strengthen prevention strategies promoting lifestyle and behaviour change in coronary patients.
70 PREVALENCE OF THE METABOLIC SYNDROME AND ITS COMPONENTS IN NORTHWEST RUSSIA: THE ARKHANGELSK STUDY O. Sidorenkov1, O. Nilssen1, T. Brenn1, S. Martiushov2, A. M. Grjibovski3,4 1Institute of Community Medicine, University of Tromsø, TROMSØ, Norway, 2Department of Internal Medicine-II, Northern State Medical University, ARKHANGELSK, Russia, 3 Norwegian Institute of Public Health, OSLO, Norway, 4International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia Background: The metabolic syndrome (MetS) is a cluster of risk factors associated with cardiovascular (CVD) morbidity and mortality. Russia has one of the highest CVD mortality rates in the world. However, the prevalence of MetS in Russia remains unknown. Objectives: To estimate the prevalence of MetS and its components in an urban Russian setting. Methods: Altogether, 3705 Russian adults aged 18–90 years were enrolled in a cross-sectional study in Arkhangelsk (Northwest Russia). All subjects completed a questionnaire and underwent a physical examination. Blood tests were taken and analyzed in Tromsø, Norway. MetS was defined according to the modified National Education Cholesterol Education Program Adult Treatment Panel III (NCEP), American Heart Association / National Heart, Lung and Blood Institute (AHA/NHLBI) and International Diabetes Federation (IDF) criteria. The prevalence data were standardized using Segi world standard population. Results: The age-standardized prevalence rates of MetS among women were 19.8% (95% CI: 18.1–21.5), 20.6% (95% CI: 18.9–22.3) and 23.1% (95% CI: 21.3–24.9) by NCEP, AHA/NHLBI and IDF criteria, respectively. The corresponding rates for men were 11.5% (95% CI: 10.1–12.9), 13.7% (95% CI: 12.2–15.2) and 11.0% (95% CI: 9.7–12.4). Among subjects with MetS, central obesity was more common among women, while elevated triglycerides and blood glucose were more common among men. Conclusions: While the prevalence of MetS among Russian women is comparable to the data from Europe and the US, the prevalence of MetS among Russian men is considerably lower than among their European and North-American counterparts. Our results suggest that MetS is unlikely to be a major contributor to the CVD deaths in Russian men. Further studies on the associations between different criteria of MetS and CVD are needed for a better understanding of the mechanisms leading to exceptionally high rates of CVD mortality in Russia.
238 LOWER SERUM ADIPONECTIN LEVELS INDEPENDENTLY PREDICT TYPE 2 DIABETES IN A KOREAN RURAL POPULATION J. K. Park, S. B. Koh, Department of Preventive Medicine, and Institute of Genomic Cohort, Yonsei University Wonju College of Medicine, WONJU CITY, South Korea Background: Adiponectin, a fat-derived protein, has been proposed to play an important role in development of insulin resistance and atherosclerosis. Decreased adiponectin were shown in obesity, metabolic syndrome, type 2 diabetes, and cardiovascular disease. Objective: The
27 aim of this study was to assess relationship between adiponectin and the status of glucose tolerance and to investigate whether decreased adiponectin predicted type 2 diabetes in Korean population. Method: We enrolled 3508 subjects (mean age 56.3 ± 8.0 years; BMI 24.7 ± 3.2 kg/m2; 41.0% male) from the Koran Genomic Rural Cohort (KGRC). Clinical and biochemical metabolic parameters were measured. Serum adiponectin levels were measured by RIA (LINCO Research, Inc. USA). We defined normal glucose tolerance (NGT), impaired fasting glucose (IFG), impaired glucose tolerance (IGT), and type 2 diabetes using ADA criteria. Results: The prevalence of NGT, IFG or IGT and type 2 diabetes were 58.2%, 28.9%, and 13.1%. The level of adiponectin in female were significantly higher than those in male (12.1 ± 5.5 vs. 8.6 ± 4.5 lg/mL, P \ 0.001) and adiponectin levels in NGT, IFG or IGT, and type 2 diabetes showed a significant inverse linear pattern in each sex group (male 9.2 ± 4.5 vs. 8.2 ± 4.3 vs. 7.2 ± 4.1; female 12.8 ± 5.4 vs. 11.5 ± 5.5 vs. 9.7 ± 4.7 lg/mL, P \ 0.001). Adiponectin in each sex were significantly correlated with age, waist circumference (WC), triglycerides (TG), HDL, HOMA-IR, and FBS (correlation coefficient: male 0.18, -0.29, -0.24, 0.35, -0.13, and -0.11; female 0.13, -0.27, -0.23, 0.28, -0.17, and -0.14). The odds ratios (ORs) for type 2 diabetes were significantly decreased with increasing quartiles of adiponectin in each sex. After adjustment for age, WC, HOMA-IR, TG, and HDL cholesterol, ORs with increasing quartiles remained significant [male 1 vs. 0.57 (0.36– 0.77) vs. 0.49 (0.31–0.78) vs. 0.48 (0.29–0.79); female 1 vs. 0.75 (0.51–10.9) vs. 0.61 (0.40–0.92) vs. 0.40 (0.24–0.66)]. Conclusions: Adiponectin showed a strong inverse correlation with the status of glucose tolerance and lower adiponectin level was an independent predictor of type 2 diabetes in KGRC population.
234 VALIDATION OF REFLECTANCE COLORIMETRIC METHOD FOR FASTING GLUCOSE AND TOTAL CHOLESTEROL DOSAGE. ˜ O PRETO, SP, BRAZIL, 2006 RIBEIRA V. de Castro, S. A. de Moraes, I. C. M. de Freitas, University of Sa˜o ˜ O PRETO, Brazil Paulo, RIBEIRA Background: The benefits for health and quality of life offered by the control and reduction of glucose and cholesterol levels demand the use of simplified techniques to obtain these indicators in populationbased epidemiological studies. Objective: To assess the accuracy of the automated fasting glucose and total cholesterol dosage method used in a cross-sectional population-based epidemiological study to detect the prevalence of overweight and diabetes mellitus in the population aged 30 years or older living in Ribeira˜o Preto, SP, Brazil, in 2006. Methods: To verify glucose and total cholesterol levels, capillary blood and venous plasma samples were obtained simultaneously from 150 participants after 12 hours of fasting. Venous plasma dosage was measured in a reference laboratory (goldstandard) using enzymatic methods, and capillary blood samples were analyzed with portable equipment by means of the reflectance colorimetric method. The global accuracy of the automated method was estimated using ROC curves. Results: The areas under the ROC curves (AUC) were 0.886 (CI95%: 0.739–1.000), and 0.942 (CI95%: 0.905–0.979) for automated fasting glucose measures, and automated total cholesterol measures, respectively. Conclusions: The reflectance colorimetric method represents a highly accurate alternative to obtain fasting glucose and total cholesterol levels. The use of a portable, compact and easily operable equipment is one of the advantages of this alternative method to carry out population-based epidemiological research.
123
28
Late Morning Session 4—Infectious diseases epidemiology I 268 NEWLY DIAGNOSED HIV INFECTIONS IN DRUG USERS IN PORTUGAL: FREQUENCY AND CHARACTERISTICS OF FIRST-TIME AND FOLLOW-UP VISITORS OF DRUG TREATMENT CENTRES R. Lucas1,2, M. J. Santos3, L. Prasad1,2, H. Barros1,2,3 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health of the University of Porto, PORTO, Portugal, 3National Coordination for HIV/AIDS, LISBOA, Portugal Background: HIV risk and equity in access to testing and treatment are not uniform within drug user (DU) populations. Quantification and characterisation of new HIV cases may indentify target groups for prevention. Objectives: To quantify new HIV cases in first-time and followup visitors to drug treatment centres (DTC) in Portugal and to identify risk factors for the infection. Methods: From June to December 2007, 6930 visitors to 45 DTC in Portugal were systematically approached and invited to participate in a counselling, testing and referral programme. Participants were enrolled after informed consent. Rapid testing followed by confirmation was used to identify HIV-positive clients, who were referred to hospitals according to a pre-established protocol. Further data collection was conducted using a standardised questionnaire including sociodemographic and drug-use related characteristics. Associations between HIV test result and sociodemographic variables were estimated using odds ratios (OR) and 95% confidence intervals (95%CI). Results: Prior HIV prevalence was 12.1% in follow-up and 4.4% in first time visitors. We identified 89 new HIV cases. Among 862 first time visitors, HIV prevalence was 0.38% (95%CI:0.04–1.40) in non-IDU and 4.87% (95%CI:2.86–7.68) in IDU. In 6068 follow-up clients it was 0.33% (95%CI:0.14–0.66) in non-IDU and 1.69% (95%CI:1.30–2.16) among IDU. Testing positive was associated with female sex in IDU (OR = 2.08, 95%CI: 1.32–3.29 in follow-up; OR = 1.93, 95%CI: 0.68–5.47 in first visits) but especially in follow-up non-IDU (OR = 4.09, 95%CI: 1.43–11.71). In first visits the highest HIV frequency was observed in those under 25 years-old (OR = 4.07, 95%CI: 0.35–46.94 in IDU) and in those born outside Portugal (OR = 3.22, 95%CI: 0.87–11.96 in IDU; OR = 7.45, 95%CI:0.66– 84.40 in non-IDU). Educational level was inversely associated with HIV in follow-up clients (OR = 0.43, 95%CI: 0.24–0.78 in IDU; OR = 0.21, 95%CI: 0.03–1.59 in non-IDU). Conclusion: Rapid HIV testing improved access to early diagnosis, treatment and support of first time DTC visitors, especially in women, migrants and young people.
98 HEROIN USERS IN BELGIUM, 2001–2005: PATTERNS OF DRUGS USE, RISK BEHAVIOURS, INFECTIOUS DISEASES AND USE OF HEALTH SERVICES N. Deprez, M. Roelands, Scientific Institute of Public Health, BRUSSELS, Belgium Background: Injecting heroin use is a risk factor for ill health. Information about this group is lacking. Objectives: This study investigates the population of heroin users in Belgium. It aims to describe its socio-demographic characteristics, patterns of drug use, risk behaviour, related harmful consequences and use of health care facilities. Methods: A longitudinal study was set up and participants (N = 1034) were selected through a snowball sampling method.
123
IEA-EEF European Congress of Epidemiology 2009 Results: In the first phase, for 2 in 3 heroin users, social allowances were the main source of income. Median period between first use of non-injected heroin and first use of injected heroin was 1 year. Median period between first use of non-injected heroin and first request for treatment was 3 years. Median period between first use of non-injected heroin and first substitution treatment was 5 years. 76% never used a condom with an IDU partner. Prostitution was significantly more prevalent among women compared to men, 26% versus 3%. Almost half of heroin users reported to have experienced at least once an overdose. Related risk factors for having an overdose were imprisonment and intravenous drug use. 10% of the heroin users had visited an emergency room, 11% were admitted in hospital for at least one night. 56% of the respondents visited an outpatient clinic and 37% had contact with street workers. In the follow-up, significant decreases were observed in injecting use of heroin, cocaine and speedballs. Significant increases were found in substitution substances and hypnotics use. Some respondents quitted their IDU. Incarceration is found to be a risk factor for having a new overdose. Conclusions: Some characteristics of lifestyle and behavior in heroin users could be described. Some significant changes could be reported for people who were willing to participate in the follow-up study.
291 HOMOSEXUALLY ACQUIRED SYPHILIS IN NORWAY, 1999–2007: AN ONGOING BATTLE I. Jakopanec, A. Grjibovski, P. Aavitsland, Department of Infectious Disease Epidemiology, Norwegian Institute of Public Health, Nydalen, OSLO, Norway Background: Following reports of syphilis outbreaks among men who have sex with men (MSM) in many countries in the late nineties, the first increase in Norway was noted in 1999 with 40 cases reported to the Norwegian Surveillance System for Communicable Diseases (MSIS), compared to 2 in 1998. Primary, secondary or early latent stages of syphilis are mandatory anonymously notifiable in Norway. Objectives: In order to provide evidence for preventive measures we described specific trends in epidemiology of homosexually acquired syphilis in Norway. Methods: We analysed cases of homosexually acquired syphilis reported to MSIS in 1999–2007 by year, demographic characteristics, stage of disease, co-infection with other sexually transmitted infections (STI), contact partners and place of infection. Results: From 1999 to 2007, 304 homosexually acquired syphilis cases were reported. After a peak of 43 cases in 2002, two peaks in 2006 and 2007 followed with 56 and 54 cases. Syphilis was diagnosed in primary stage in 111 cases (36%), in secondary stage in 135 cases (44%) and early latent stage in 58 cases (19%). The median age was 37 years. Casual contacts accounted for 220 (72%) infections, while 51 (17%) were infected by a steady partner. Among the 234 (77%) cases, infected in Norway, 214 (90%) cases acquired their infection in Oslo city. Of 92 (30%), co-infected with another sexually transmitted infection, 68 (22%) were HIV positive. HIV co-infection increased from 15% in 1999 to 33% in 2007. Conclusions: Since 1999, an epidemic of syphilis has affected MSM in Norway with Oslo as the epicentre. One third of the cases are HIV positive. Only a third of the cases are diagnosed in the primary stage, indicating missed opportunities for earlier treatment, hence with prolonged infectivity. Practice of safer sex is needed to stop the epidemic.
293 RISK FACTORS FOR EXTENSIVELY DRUG-RESISTANT TUBERCULOSIS: A SYSTEMATIC REVIEW M. Tavares1,2, B. Lima1,2, H. Barros3,4, 1Department of Infectious Diseases, University of Porto Medical School, PORTO, Portugal,
IEA-EEF European Congress of Epidemiology 2009 2
Hospital de S. Joa˜o, PORTO, Portugal, 3Department of Epidemiology, University of Porto Medical School, PORTO, Portugal, 4 University of Porto Instuitute of Public Health, PORTO, Portugal Background: Extensively drug-resistant tuberculosis (XDR-TB) is emerging as a global public health problem. The current definition of XDR-TB (October 2006, WHO) implies resistance to isoniazid and rifampicin, to a fluroquinolone and to a second-line injectable agent. Its treatment is more expensive and difficult than multi drug resistant tuberculosis (MDR-TB) and the outcomes much severe. Thus, the identification of risk factors for XDR-TB is of paramount importance to design effective TB control strategies. Objective: Systematic review of published English language articles on risk factors for XDR-TB. Methods: Pubmed and Cochrane Library were searched using the terms ‘‘tuberculosis’’, ‘‘XDR TB’’ and ‘‘extensively drugresistant tuberculosis’’ and review articles references manually screened. We identified 192 articles, 151 were excluded by the
29 abstract. The remaining 41 articles were retrieved for full text detailed evaluation by two authors and 12 relevant articles were selected for final review. The criteria for inclusion was the 2006 XDR-TB case definition. Results: Some risk factors were consistently present, mainly previous TB treatment and length of treatment. Other conditions often associated were imprisonment, immigration, alcohol and HIV co-infection. Pre-XDR-TB points to an increased risk of XDR-TB but needs further assessment. The described quantitative risk for many of these factors is probably underestimated due to several limitations of individual studies namely: concerns with prompt diagnosis, laboratory quality-control for the testing of all firstand second-line drugs resistance, data collection on patient characteristics, and a general absence of standard procedures. Conclusion: The information regarding determinants of XDR-TB is scarce. However, regardless of the country surveyed, special emphasis should be given to minimize the risks of TB re-treatment to prevent the emergence of highly resistant TB.
123
30
Thursday, 27 August 2009 Parallel afternoon oral sessions: 16.30–17.50 Afternoon Session 1—Diabetes mellitus epidemiology 239 INTERACTION BETWEEN GAMMAGLUTAMYTRANSFERASE AND OBESITY ON RISK FOR PREVALENCE OF TYPE 2 DIABETES S. B. Koh, J. K. Park, Department of Preventive Medicine, and Institute of Genomic Cohort, Yonsei University Wonju College of Medicine, WONJU CITY, South Korea Background: Serum c-glutamytransferase (GGT) and obesity have been regarded as risk factors for type 2 diabetes. Some observational studies have shown that the association of obesity with the risk of type 2 diabetes according to serum GGT is either weakened or strengthened. However, the evidence of the interaction between GGT and obesity on the risk of diabetes is insufficient, and the association between visceral adiposity and diabetes according to serum GGT has not been examined. Objective: We examined the association between various parameters of obesity and the risk of prevalence of diabetes in accordance with serum GGT and analyzed the interaction between visceral adiposity and GGT on the risk of prevalence of diabetes in cross-sectional study. Methods: There were 2,948 participants, aged 40 to 70 years from Korean Rural Genomic Cohort (KRGC). The participants received at least one medical examination and completed a questionnaire during 2005–2006. We examined the risk of prevalence of diabetes according to quartiles of waist circumference (WC), visceral fat area (VFA) and serum GGT, respectively. We determined the combined effect of variables and the interaction between WC or VFA and GGT on the risk of diabetes. Results: Adjusted odds rations (ORs) for prevalence of diabetes showed an increased response to successive quartiles of WC, VFA, and GGT. Area under the curves for WC, VFA and GGT associated with prevalence of diabetes were similar, respectively. Adjusted Ors for the strata defined by pairs of variables were increased (ORs = 2.01, 2.68, 3.88 for GGT and VFA; ORs = 2.24, 2.82, 4.17 for GGT and WC). The risk of WC or VFA on prevalence of diabetes was not associated within the lower quartiles of GGT. However, risk of prevalence of diabetes was increased according to quartiles of GGT in all categories of WC or VFA. The interaction between WC or VFA and GGT in the risk of prevalence of diabetes was not statistically significant. Conclusion: Our findings revealed a combined effect of WC or VFA and serum GGT on the risk of prevalence of diabetes. These results indicate that serum GGT may be a more influential factor than abdominal obesity or visceral fat on the risk of prevalence of diabetes.
69 OBESITY PREVALENCE AND AWARENESS AMONG ADULT POPULATION AND NHS OPERATORS IN THE LAZIO REGION, ITALY: RESULTS FROM THE PASSI SURVEILLANCE M. O Trinito1, A. Capon2, S. Iacovacci3, A. Lancia1 1Prevention Department, Roma C Local Health Unit, ROME, Italy, 2Agency for Public Health—Lazio Region, ROME, Italy, 3Prevention Department, Latina Local Health Unit, LATINA, Italy
123
IEA-EEF European Congress of Epidemiology 2009 Background: Obesity and overweight prevalence have been growing over the last years through Europe. Effective prevention strategies are urgently needed to reduce consequences in terms of morbidity, mortality and public health costs. Awareness among general population and health operators should be adequately monitored. The Passi surveillance on behavioural and preventable risk factors is a useful source of information. Methods: Phone interviews from the first year of the Passi surveillance were analyzed. Target population was aged 18 to 69, subjects were selected through age and sex stratified random sampling. Interview items included: reported height and weight, weight self-perception, physical activity, advices received by NHS operators on diet and exercise. Factors associated to obesity and advice on diet were explored trough two multivariate logistic models. Results: Totally 2920 people were interviewed from 31/08/2007 to 31/08/2008. 31.1% were overweight (BMI 25–30) and 10.7% obese. Overall overweight and obesity prevalence increased with age both among men (18–34 years: 32.3% 60–69 years: 63.6%) and women (18–34 years: 13.9% 60–69 years: 57.2%). 27% of overweight subjects and 36% of obese were on diet. 44.8% of them were advised by their GP to exercise regularly and 63.2% to follow a diet. Multivariate logistic regression showed a significantly increased risk for obesity in subjects with lower education (junior high school or lower: (RR = 2.44)) and in those reporting economical problems (RR = 1.54). Among overweight and obese subjects GP advise on diet was significantly associated to higher education (RR = 1.53), obesity (RR = 4.42), female sex (RR = 2.69), previous myocardial infraction (RR = 1.85). Conclusions: The results confirm a high prevalence of overweight and obesity. Public awareness regarding the need of a correct diet and exercise is still low and not strongly supported by NHS operators. Ongoing PASSI surveillance aims at raising awareness among NHS staff and focusing interventions among unprivileged groups.
278 THE EPIDEMIOLOGY OF METABOLIC SYNDROME IN THE HUNGARIAN ADULT POPULATION: A REPRESENTATIVE SURVEY E. Szigethy1, G. Sze´les1, Z. Voko´1, A. Nagy1, A. Horva´th2, ´ da´ny1 T. Hidve´gi3, G. Jermendy4, G. Paragh5, G. Blasko´6, R. A 1 Department of Biostatistics and Epidemiology, Faculty of Public Health, University of Debrecen, DEBRECEN, Hungary, 2SanofiAventis Private Co Ltd., BUDAPEST, Hungary, 3Petz Alada´r County } Hungary, 4Bajcsy-Zsilinszky Hospital, Teaching Hospital, GYOR, BUDAPEST, Hungary, 51st Department of Internal Medicine, University of Debrecen, DEBRECEN, Hungary, 6Department of Pharmaceutical Management, University of Debrecen, DEBRECEN, Hungary Background: Hungary has high cardiovascular disease mortality. Recent studies reveal a high prevalence of many cardiovascular risk factors, including obesity, diabetes and hypertension. Objectives: The main objective of our study was to estimate the prevalence of metabolic syndrome (MS) and its treatment practices in a representative sample of the Hungarian population. Methods: Within the framework of the General Practitioners’ Morbidity Sentinel Stations Programme we selected a random sample of 2000 inhabitants aged 20–69 years, registered at participating general practitioners (GPs) in 2006. Physical examinations, blood sampling and the gathering of data related to treatment practices were performed by the GPs. Information on environmental and lifestyle factors was collected by means of a standardized self-administered questionnaire. Results:
IEA-EEF European Congress of Epidemiology 2009 The overall response rate was 91%. The age-adjusted prevalence of MS, according to the ATPIII criteria, was 26% in males aged 20–69 years, and 24.1% in females (P = 0.35). The IDF consensus gives higher figures, of 36.1% in men and 30.1% in women (P \ 0.01). In contrast, the proportion of patients known to have MS prior to the study was 7.8% in men and 6.1% in women. The prevalence of impaired glucose regulation (according to IDF criteria) was 17.1% (n = 309), of central obesity 72.4% (n = 1,301), of hypertension 58.7% (n = 1,059), of hypertrigliceridemia 34.9% (n = 630), and of reduced HDL cholesterol 29.7% (n = 536). Noteworthy, the prevalence of impaired glucose regulation was high (30%) in the age group 55–69 years and the prevalence of central obesity exceeded 40% in the youngest age group (20–24 years). There were no statistically significant regional differences in the prevalence figures. Conclusions: The high prevalence of metabolic syndrome in Hungary is a serious public health problem, as a major determinant of the high cardiovascular disease burden.
236 DIABETES MELLITUS PREVALENCE AND CORRELATES IN A BRAZILIAN ˜ O PRETO, SP, POPULATION LIVING IN RIBEIRA BRAZIL, 2006 S. A. de Moraes, I. C. M. de Freitas University of Sa˜o Paulo, ˜ O PRETO, Brazil RIBEIRA Background: The demographic-epidemiological transition and the current panorama of nutritional transition demand prevalence studies and the identification of factors associated with chronic outcomes in the Brazilian population. Objectives: To identify the prevalence of diabetes mellitus (DM) and its correlates in the adult population of Ribeira˜o Preto, SP, Brazil. Methods: Cross-sectional populationbased epidemiological study using three-stage cluster sampling, being census sector the primary sampling unit. The variability introduced in the third sampling fraction was corrected by attributing equal weights to the number of eligible units in each domicile, resulting in a weighted sample of 2197 participants aged 30 years or older, living in the urban area of Ribeira˜o Preto-SP, Brazil, in 2006. DM diagnosis was based on previous medical history/or meeting WHO cut-offs after OGTT. DM prevalence was estimated by points and 95% confidence intervals, considering the design effect. To investigate associated factors, crude and adjusted prevalence ratios were estimated by points and 95% confidence intervals, using Poisson regression. Results: The prevalence of diabetes mellitus was 15.02% (CI 95%: 12.68–17.36%). Among participants diagnosed as diabetic, 15% ignored their condition before the research. After adjusting for potential confounding, DM correlates remaining in the final model were represented by: ‘‘age’’ (PR = 2.62; CI95%: 1.31– 5.24) for participants aged 60 years and older; ‘‘DM familial antecedents’’ (PR = 1.74; CI 95%:1.28–2.37); ‘‘waist hip ratio’’ (PR = 1.70; CI95%:1.04–2.76) and ‘‘waist height ratio’’ (PR = 2.39; CI 95%:1.30–4.40), for participants classified in the third tercile; ‘‘5 or more medicines taken during the last 15 days’’ (PR = 2.61; CI 95%:1.57–4.34), and ‘‘attendance in outpatient centers’’ (PR = 2.13; CI 95%:1.21–3.73). Trend tests for correlates classified in more than two categories were statistically significant (P \ 0.05). Conclusion: The results showed high prevalence of diabetes mellitus in the study population as well as identified associated factors liable to public intervention policies.
31
Afternoon Session 2—Children and adolescent health epidemiology I 287 PATHWAYS FROM INTERPARENTAL VIOLENCE TO VICTIMIZATION BY PEERS DURING ADOLESCENCE: THE ROLE OF SELF-ESTEEM AND PARENTING STYLE Ch Roustit, B. Chaix, P. Chauvin Inserm, U707, Research Group on the Social Determinants of Health and Health Care, PARIS, France, UPMC Univ Paris 06, UMR S 707, PARIS, France Background: Violence exposure in childhood is associated with a higher risk of psychosocial maladjustment. Understanding how to break cycles of violence in which certain adolescents are trapped in their family, school or residential environment is one of the avenue for promoting adolescent health. Objectives: Our aim was to investigate the factors mediating the link between exposure to interparental violence in childhood and having experienced violence by peers in adolescence, intermediate factors at which specific interventions could be targeted. Methods: Study: Social and Health Survey of Children and Adolescents of Quebec (Canada) a cross-sectional survey of a representative sample of 2346 youths aged 13–16 years and 1983 parents. Outcome variables: having experienced violence at school for all youths and exposure to intimate partner violence for girls only. Main independent variable: lifetime witnessing of interparental violence. The potential mediating factors included adolescent self-esteem, family functioning (parental emotional support, authoritative parental control and parental psychological distress), and social support. Statistical analyses: logistic regression models. Results: Of the 2346 adolescents, 35% reported having experienced violence at school. Of the 619 girls who reported an intimate relationship, 38% reported victimization by their partner. Adolescents exposed to interparental violence were more likely to be victimized by their peers at school (OR = 1.68;95%CI = [1.35– 2.09]). Moreover, girls exposed to interparental violence were more likely to have experienced intimate partner violence (OR = 1.76;95%CI = [1.18–2.61]). Low self-esteem and a high level of negative authoritative parental control were found to be mediating factors of the association between interparental violence and victimization by peers. Conclusions: This study gives arguments for a cycle of violence across generations. Psychosocial interventions for breaking this cycle require focusing on the internal functioning of the family in order to improve the likelihood of a supportive parental relationship, to improve the adolescent’s self-esteem, and to change the model through which adolescents interact with their social environment.
204 ASSESSMENT OF REPRODUCIBILITY FOR A SELF-DELIVERED ESPAD (THE EUROPEAN SCHOOL SURVEY PROJECT ON ALCOHOL AND OTHER DRUGS) QUESTIONNAIRE ABOUT USE OF ILLICIT DRUGS USING THE TEST-RETEST METHOD O. Curzio, S. Molinaro Institute of Clinical Physiology, Italian National Council for Research, Department of Epidemiology, ROME, Italy Background: Considering the prevalence rate and the consequences of drug use, developing a reproducible, valid, flexible and cheap screening test is of crucial interest. The test-retest reproducibility is an important
123
32 indicator of temporal stability of a measurement. Objectives: To assess the reproducibility of the ESPAD questionnaire among students aged 15-19 with a gap of three-weeks between administrations. Methods: A test-retest methodology was used for a subgroup of ESPADItalia2004 sample. To verify the test-retest concordance for the variables relating to lifetime use of alcohol, cigarettes and illicit drugs the original ordinal variables as well as the same dicotomically recodified were used. Data analysis was made using Kappa and weighted Kappa for the answers given at the first and at the second administration. To evaluate the influence of gender and age on the concordance the method proposed by Lipsitz was used. Results: The questions about drug use were examined in ordinal form and show a good test-retest concordance and an excellent concordance for the answers relating to the use of cigarettes, alcohol, cannabis. Apart from the case of tranquillizers and sedatives, concordance increases when answers about drug use were examinated in dicotomic form. As to the effect of age adjusted for gender, 15-year-old subjects showed a significantly lower concordance, if compared with the 19-year-old ones. When the two covariates resulted significant in the model, the focus was put on specific values of k for the covariate levels. Conclusions: The ESPAD questionnaire is an instrument with a good reproducibility to quantify consumption of licit and illicit drugs. The results showed the effect of the gender and age covariates on the concordance of answers and suggested the importance of examining the concordance relating to the covariate levels.
125 OVERWEIGHT AND OBESITY IN ADOLESCENTS IN NORTHWEST RUSSIA: DOES THE PREVALENCE DEPEND ON WHETHER RUSSIAN OR INTERNATIONAL CRITERIA ARE APPLIED? S. L. Khasnutdinova1, A. M. Grjibovski2 1International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia, 2Norwegian Institute of Public Health, OSLO, Norway Background: Overweight and obesity is an increasing public health problem worldwide. The latest studies suggest an increase in obesity among Russian adults in the 2000 s. It is unknown if this increase occurred in adolescents, especially in the areas outside large cities. Furthermore, most of the Russian studies apply Russian criteria for defining overweight and obesity making it difficult to make meaningful comparisons with international data. Aim: To assess the prevalence of overweight and obesity among adolescents in a rural district in Northwest Russia by using international and Russian criteria. Methods: All 9–11th grade students (mean age 15.6 years) from all schools in Velsk district were asked to participate. The sample included 428 boys and 604 girls (response rate 87%). Overweight ? obersity and obesity were defined according to CDC2000, IOTF, WHO-2007 and Russian criteria. Differences between the estimates by Russian criteria and each of the international criteria were studied by McNemar tests. Results: The estimated prevalence of overweight ? obesity among boys was 9.7% (95%CI: 7.2–12.8), 10.8% (95%CI: 8.2–14.0), 12.4% (95%CI: 9.6–15.7) and 7.6% (95%CI: 5.5–10.5) by CDC-2000, IOTF, WHO-2007 and Russian criteria, respectively. The corresponding estimates for girls were 7.9% (95%CI 6.0–10.3), 7.7% (95%CI: 5.9–10.1), 8.9% (95%CI: 6.9–11.4) and 9.2% (95%CI: 7.1–11.7). The prevalence of obesity among boys was 2.9% (95%CI: 1.7–4.9), 2.5% (95%CI: 1.4–4.4), 6.1% (95%CI: 4.2–8.7) and 1.8% (95%CI: 0.9–3.5) by CDC-2000, IOTF, WHO-2007 and Russian criteria, respectively. The corresponding estimates for girls were 2.4% (95%CI 1.5–3.9), 1.6% (95%CI: 0.9–2.9), 3.7% (95%CI: 2.5–5.5) and 2.7% (95%CI:
123
IEA-EEF European Congress of Epidemiology 2009 1.7–4.3). Russian criteria significantly underestimated the prevalence of overweight ? obesity compared to all international references (all P \ 0.001) in boys, but not in girls. Conclusions: the prevalence of overweight and obesity in rural Northwest Russia is lower than in Europe. Russian criteria underestimate the prevalence of overweight and obesity in boys, but not in girls.
279 UNDERWEIGHT, OVERWEIGHT AND OBESITY TRENDS IN WARSAW ADOLESCENTS DURING THE 35 YEARS PERIOD, 1971–2006 J. Charzewska1, E. Chabros1, B. Wajszczyk1, Z. Chwojnowska1, E. Kraszewska2 1Department of Nutritional Epidemiology and Dietary Allowances, National Food and Nutrition Institute, WARSAW, Poland, 2Medical Center for Post Graduated Education, WARSAW, Poland Background: The shifts in diet and physical activity patterns in countries passing through economic transition, have potential to influence prevalence of nutritional disorders resulting in excess of obese as well as in underweighted children and youth. Objective: To determine time trends (1971–2006) in underweight, overweight or obesity in adolescents and to identify independent predictors of those disorders. Method: Data come from five successive surveys of randomly selected samples of adolescents (aged 11–15 years old), from Warsaw region. In total 10.941 pupils have been examined, with response rate varying from 55 to 87% depending on year. Classification of underweight, overweight and obese was done using Body Mass Index (BMI = kg/m2) specific for age and gender group of children (Cole, et al., 2000, 2007). Results: For underweight, tests for trend were not significant, however underweight was more frequent in girls (13.8% as average) than in boys (8.7%). In both genders, frequency of overweight pupils increased significantly; in boys from 8.9% in 1971 to 18.4% in 2006, and in girls from 8.2 to 12.1%. Similarly, tests for trend in obesity were also significant, and frequency of obese girls changed from 0.3 to 3.4% and in boys from 1.0 to 2.8%. Multivariate logistic regression models identified the most sensitive age group for development of analysed disorders to be age 13 for boys and age 12 for girls. Other independent predictors which influence underweight or overweight and obesity in adolescents were social stratas, year of study and nutritional factors. Conclusion: There was a significant increase in overweight and obesity during the 35 year period with greatest increase observed after year 2000, implicating obesogenic influences of some social situations.
Afternoon Session 3—Pharmacoepidemiology 128 CONCOMITANT USE OF GASTROPROTECTIVE AGENTS (GPAS) AMONG NSAID/ COX-2 SELECTIVE INHIBITOR USERS: A HOSPITAL REGISTER-BASED STUDY J. Atthobari, E. Kristin, I. Dwiprahasto, Department of Pharmacology & Toxicology, Faculty of Medicine, Gadjah Mada University, YOGYAKARTA, Indonesia Background: NSAIDs and COX2 have been associated with an increased risk of GI complications, in particular when risk factors are present. It is recommended that gastroprotective agents (GPAs) (i.e. misoprostol, proton-pump inhibitors, H2 receptor antagonists or antacids) be taken concomitantly to prevent NSAID-induced GI com-
IEA-EEF European Congress of Epidemiology 2009 plications. However, there are concerns that the rate of concomitant use of GPAs in NSAID/Cox2 users is too low. Objective: To investigate the prevalence of concomitant GPA among users of NSAIDs/COX2 and determine the factors associated with concomitant use of GPA and NSAIDs/COX2. Methods: We analyzed data on age, sex and dispensed drugs from an Indonesian private hospital prescribed drugs register on 2004–2006 (n = 139,942 prescriptions) and located 31,350 NSAID/ COX2 prescriptions for 9,455 patients. Logistic regression analysis was used for analyzing the association between the use of different NSAIDs/ COX2 and GPA, and between individual characteristics and use of GPAs. Results: Co-prescribing of GPAs were used by 5,550 (17%) of NSAID/COX-2 users. Celecoxib, meloxicam, ketoprofen, rofecoxib, glucosaminoglycan and diclofenac were most used concomitantly with GPAs. Celecoxib and meloxicam were most strongly associated with GPAs, OR 95%CI 10.09; 8.65–11.77 and 9.65; 8.08-11.54, respectively. Concomitant use of GPAs varied among individual NSAIDs. Use of combination NSAIDs (ORadj 1.16; 1.07-1.26), females (ORadj 2.38; 2.24–2.54), age [ 45 years old (ORadj 1.30; 1.21–1.40), and concomitant use of corticosteroid (ORadj 1.28; 1.14–1.45) were significantly associated with concomitant prescribing of GPAs during NSAIDs therapy. Conclusions: The rate of concomitant prescribing GPAs in NSAID users is low. Furthermore COX2 were used with GPAs more often than traditional NSAIDs. Use of combination NSAIDs, female, age [ 45, use of corticosteroid are associated with concomitant use of GPAs in NSAID users. Feedback to prescribers should be given to improve prescribing practices.
246 PATTERNS OF CO-PRESCRIBING OF ANGIOTENSIN CONVERTING ENZYME INHIBITOR AND ANGIOTENSIN RECEPTOR BLOCKERS IN IRELAND W. A. H. Wan Md Adnan1, N. L. Zaharan2, K. Bennett2, C. A. Wall1, J. Feely2 1Department of Renal Medicine, Adelaide & Meath Hospital, TALLAGHT, Ireland, 2Department of Pharmacology & Therapeutics, St James’s Hospital, DUBLIN, Ireland Background: Dual ACE inhibitors (ACE-I) and angiotensin receptor blockers (ARB) therapy may be prescribed for patients with heart failure, hypertension and proteinuria. The CALM and COOPERATE trials found better outcome with dual therapy compared to single therapy. Subsequent trials (VALIANT and ONTARGET) failed to demonstrate superiority and found increased adverse events. Aim: To examine the prescribing trends of dual ACE-I and ARB therapy in Ireland following CALM (December 2000), COOPERATE (December 2002), VALIANT (November 2003) and ONTARGET (April 2008) trials. Method: The Irish HSE-Primary Care Reimbursement Services database was used to identify patients C 16 years old co-prescribed ACE-Is and ARBs between January 2000 and January 2009 (n = 266,554). Rate of prescribing over 1000 general medical scheme (means tested for those \ 70 years and free for all C 70 years) population were calculated for each month. Odds ratio (OR) with 95% confidence interval (CI) were calculated to examine prescribing according to patient characteristics. Linear trend test was used to examine prescribing trends. The effects of the major trials on prescribing were examined using segmented regression analysis for 12-months pre- and post-trial (9-months for ONTARGET trial).
33 Results: There was a significant linear trend in overall ACE-I and ARB dual therapy prescribing over the 9-year study period (P \ 0.001). Rate of prescribing in January 2000 and January 2009 was 0.16 and 5.72, respectively. Females were less likely to receive dual therapy (OR 0.80[0.80,0.81]). Those aged 45-64 were more likely (OR 1.11[1.08,1.13]) while those aged C 65 were less likely (OR 0.93[0.91,0.95]) to receive dual therapy compared to those aged 16–44. Significant increases in prescribing were observed after CALM (P = 0.03) and VALIANT (P = 0.007) trials, but no significant changes were observed after COOPERATE (P = 0.87) and ONTARGET (P = 0.798) trials. Conclusion: Trend in dual ACE-I and ARB therapy prescribing do not appear to be affected by results from major trials.
96 GENDER AND AGE DIFFERENCES IN THE ADHERENCE TO EVIDENCE-BASED DRUG THERAPY AFTER ACUTE MYOCARDIAL INFARCTION: A POPULATION BASED COHORT STUDY IN ROME, ITALY U. Kirchmayer, N. Agabiti, V. Belleudi, M. Davoli, D. Fusco, M. Stafoggia, M. Arca`, A. P. Barone, C. A. Perucci Department of Epidemiology, Local Health Authority Roma E, ROME, Italy Background: Clinical guidelines recommend that patients with acute myocardial infarction (AMI) should be chronically treated with platelet aggregation inhibitors, beta blocking agents, agents acting on the renin-angiotensin system, and HMG CoA reductase inhibitors. Evidences of gender and age related differences in treatment adherence have been described, but population based European studies are limited. Objectives: To analyse age and gender differences in prevalence of evidence-based drug therapy of AMI patients. Methods: A cohort of 5518 AMI patients (35? years), resident in Rome was selected from the Hospital Information System. Drugs claimed after hospital discharge were retrieved from the regional drug dispense registry. Drug utilisation during 6 months follow-up after discharge was defined as cumulative number of prescriptions claimed, and corresponding mean number of therapy days covered, calculated through Defined Daily Doses. Chronic therapy was defined as two or more prescriptions. The association between polytherapy (at least one drug of each of the four categories) and individual characteristics was analysed through logistic regression. Results: The proportion of patients with chronic use was 74.1% for platelet aggregation inhibitors, 63.7% for agents acting on the reninangiotensin system, 63.4% for HMG CoA reductase inhibitors, and 50.4% for beta blocking agents. Among users, the available therapy days covered the entire follow-up period, except for beta blocking agents (102.7 mean therapy days available for 182 days of followup). Analysing poly-therapy, significant gender and age differences were observed: 51.3% of males and 37.7% of females received poly-therapy (age adjusted ORfemales = 0.76, 95%CI: 0.67–0.85). The probability of receiving poly–therapy decreased significantly with increasing age (reference: 35–64 years), adjusting for gender: OR 65–74 years = 0.83, 95%CI: 0.72–0.95; OR 75–84 years = 0.56, 95%CI: 0.49–0.64; OR 85? years = 0.20, 95%CI: 0.16–0.26. Conclusions: In Rome, adherence to evidence-based drug therapy after AMI varies between ATC groups. Use of poly-therapy is suboptimal and a significantly lower adherence is observed among females and older patients.
123
34
162 DETERMINANTS OF ANTIDEPRESSANT AND TRANQUILLIZER USE IN THE GENERAL POPULATION J. Van der Heyden, L. Gisle, E. Hesse, S. Demarest, J. Tafforeau Scientific Institute of Public Health, BRUSSELS, Belgium Introduction: Antidepressants and tranquillizers are the most commonly used psychotropic medicines. In this contribution we identify determinants of the use of these drugs in the general population after adjusting for the typical indications for these medicines: depression and anxiety/sleeping disorder respectively. More specifically we assess if determinants differ by type of medicine. Methods: Data were used from the Belgian Health Interview Survey conducted in 2004 in which information was available for 11220 respondents of 15 years and older on the medicines used in the 24 hours preceding the interview. Anxiety, depression and sleeping disorder were measured with the SCL-90R scale. The probability of tranquillizer and antidepressant use was studied by fitting a multinomial logistic regression model with 4 categories: no use, isolated use of antidepressants/ tranquillizers and combined use. Results: The use of antidepressants was significantly associated with depressive disorder, but not with anxiety disorder or sleeping disorder. Similarly, the use of tranquillizers was significantly associated with anxiety disorder and sleeping disorder, but not with depressive disorder. Age, gender, self-rated health status and chronic physical disease were significantly associated with both isolated and combined use of antidepressants and tranquillizers. No variation in use was observed in function of education and income. Reporting unsatisfactory social contacts was associated with antidepressant use (relative risk ratio 1.74; 95% CI 1.17–2.59), but not with tranquillizer use (relative risk ratio 1.15; 95% CI 0.83–1.59). Conclusions: At population level a higher use of tranquillizers and antidepressants is observed among persons with mental health conditions that are typical indications for these medicines. Factors contributing to higher use of both antidepressants and tranquillizers are being a women, older age, moderate to bad selfrated health and presence of chronic physical disease. Persons with unsatisfactory social contacts use more often an antidepressant, but do not present a higher consumption of tranquillizers.
Afternoon Session 4—Social epidemiology I 141 SOCIOECONOMIC STATUS AS DETERMINANT FOR PARTICIPATION IN MAMMOGRAPHY SCREENING: ASSESSING THE DIFFERENCE BETWEEN USING WOMEN’S OWN VERSUS THEIR PARTNER’S M. Kjelle´n, M. von Euler-Chelpin Department of Public Health, University of Copenhagen, COPENHAGEN, Denmark Background: Earlier research has shown that participation in mammography screening tends to vary across socioeconomic levels. However, several studies have suggested that measures of socioeconomic status (SES) on individual-, household- and neighborhood level, respectively, all are independently relevant for a comprehensive
123
IEA-EEF European Congress of Epidemiology 2009 analysis of health related outcomes. It is therefore to be expected that the impact of SES on participation in mammography screening varies depending on the choice of SES-indicator. Objective: We assessed the difference between using the woman’s own SES and using that of her household or spouse/partner as determinant of participation in mammography screening. Methods: Participation data from two mammography screening programs in Denmark, covering 73.415 women, were linked to a national SES classification system which provides data for each citizen as well as for their partner and household. We calculated the odds ratio (OR) of non-participation in mammography screening across SES levels using the woman’s own, the household’s, and her partner’s SES, respectively. Results: The results when basing SES on the woman’s own SES and the results when basing it on the spouse’s/partner’s SES differed significantly (P \ 0.001). When using the woman’s own SES, the odds ratio of non-participation showed a clear U-shape across SES levels, while when using the partner’s SES, the distribution of risk across SES levels was more flat. Conclusions: To what extent socioeconomic status was a determinant for screening participation strongly depended on whether the woman’s own SES or that of her partner was used in the analysis. The understanding of these patterns is important, for example, in order to target behavior change interventions correctly.
152 SOCIOECONOMIC DIFFERENCES IN SELF-MANAGEMENT OF CORONARY HEART DISEASE I. Forde, T. Chandola, M. Kivimaki University College London, LONDON, United Kingdom Introduction: Availability of simvastatin ‘over the counter’ (OTC) in the UK for those with moderate risk of cardiovascular disease (CVD) was associated with a sharp fall in prescribing. Previous studies had demonstrated equitable use of statins. We aimed to investigate the equity of statin use after deregulation and explore variations in knowledge of personal CVD risk. Methods: Analysis of rates of statin use (prescribed and OTC) and recall of personal CVDrisk among high (n = 2,451) and moderate (n = 1,180) CVD-risk participants in the Whitehall II survey, by ethnicity and socioeconomic position (SEP). Results: Mean age was 62.7 years; 91% participants were Caucasian; 47% of highest SEP. Amongst those with high-CVD risk, only 54% reported using prescribed statin and 37% were aware of being at increased risk. South Asian were more likely to report using prescribed statin than White (O.R. 1.73, P \ 0.001), there was no variation by SEP. Middle (O.R. 0.54, P = 0.05) and lowest (O.R. 0.11, P = 0.04) SEP participants were less likely to use OTC statin than highest SEP participants; South Asian participants were less likely to recall being at increased CVD risk than White participants (O.R. 0.65, P = 0.02); and mid- (O.R. 0.74, P = 0.001) and lowest (O.R. 0.52, P \ 0.001) SEP participants were less likely to recall than highest SEP. Conclusion: We find that prescribed statin use remains equitable after adjustment for need despite recent changes in availability. Fears of statin deregulation worsening health inequities appear to have been unfounded. Few high-risk participants were aware of their risk, with those of South Asian origin and lower SEP significantly less likely to be aware.
IEA-EEF European Congress of Epidemiology 2009 This may imply that health professionals communicate poorly with those groups most at risk of CVD. This is likely to impact on their ability to participate in self-management and may partly explain their worse outcomes.
248 PREVALENCE OF OBSTRUCTIVE PULMONARY DISORDERS AND SYMPTOMS IN ADULT POPULATION OF CHORZOW— PRELIMINARY RESULTS OF GALEN SCREENING SURVEY G. Brozek1, J. Zejda1, M. Farnik2, M. Kowalska1 1Department of Epidemiology, Faculty of Medicine, Medical University of Silesia, KATOWICE, Poland 2Department of Pneumonology, Faculty of Medicine, Medical University of Silesia, KATOWICE, Poland Background: The study was a part of multicenter GA2LEN Screening Survey. Present abstract shows the results from Chorzow (Poland). Objectives: The aim of the study was to assess the prevalence of respiratory symptoms and diseases in population living in large city located in Silesia voivodship. Methods: A cross-sectional questionnaire survey was performed in 6000 randomly selected inhabitans of Chorzow aged 15–75 years. After three contact attempts response rate was 45.7%. Out of 2740 participants 1263 (46.1%) were men and 1477 (53.9%) were women. The mean age was 43.9 ± 17.0 years, similar for man and women (P = 0.9). Results: Questionnairederived prevalence of physician diagnosed asthma was 6.6%, of COPD was 2.8% of subjects, asthmatic/spastic/obstructive bronchitis was diagnosed in 10.3%. Chest wheezing (last 12 months) was found in 24.9%, waking up with a feeling of tightness in the chest in 26.1%, waking up by an attack of shortness of breath in 15.8%, waking up by an attack of cough in 35%, and chronic phlegm production from the chest in 24.8% of subjects. There were no statistically significant differences between man and woman (P [ 0.05). The prevalence of mentioned diseases and symptoms increased with age and the differences between people younger and older than 50 years were statistically significant for all analyzed disorders (P \ 0.05).Smoking for at least 1 year declared by 53.1% responders: 64.0% men and 45.4 women (P = 0.00001). An interesting finding was that among smokers with at least one symptom and no diagnosis of asthma only
35 3.3% had COPD and 11.7% had asthmatic/spastic/obstructive bronchitis diagnosed ever by physician. Conclusions: The results suggest underdiagnosis of COPD in Chorzow.
219 DIFFERENCES IN LIFE EXPECTANCY BETWEEN EAST AND WEST OF EUROPE M. Manczuk and the HEM project collborative group Cancer Center and Institute of Oncology, Cancer Epidemiology and Prevention Division, WARSAW, Poland Objective: To measure the difference in life expectancy (LE) between eastern and western part of European Union and to determine how much of this difference is attributable to different causes of death and risk factors. Methods: Mortality and population data were obtained from the WHO statistics database. Data on tobacco and alcoholrelated deaths were gathered within HEM—Closing the Gap project. Standard methodology of LE difference decomposition was applied. Analyses were conducted separately by sex and age. Results were calculated in two groups of countries: EU10 (central and eastern European EU members) and the EU15 (western countries). Results: LE difference between EU10 and EU15 averaged in men 6.8 years. Mortality of young adults (20–44) contributed in 16% to the gap, middle-aged adults (45–64) in 45% and population after the age of 65–30%. Cardiovascular disease contributed in 40% to the gap, cancer in 16% and injuries in 23%. Tobacco contributed in 46% to the gap and alcohol in 25%. In women the LE difference averaged 4.7 years. Mortality of young adults contributed in 6% to the gap, middle-aged adults in 26% and at the age 65?—56%. CVD contributed in 47% to the gap, cancer in 24% and injuries in 11%. Tobacco contributed in 20% to the gap and alcohol in 6%. Conclusion: In 2002, the difference in male life expectancy at birth between the EU10 and the EU15 was 7 years, largely occurring in those aged 20? years, and largely due to CVD and injuries, with tobacco and alcohol being appreciable causes. In 2002, the difference in female life expectancy at birth between the EU10 and EU15 averaged almost 5 years, largely occurring in those aged 45? years, and largely due to CVD, followed by cancers and injuries, with tobacco and, to a lesser extent, alcohol being appreciable causes.
123
36
Friday, 28 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Lifecourse epidemiology II 176 EFFECT OF PARENTAL SOCIAL CLASS ON ALCOHOL RELATED MORTALITY IN ADULTHOOD AND OLD AGE Flaminia Chiesa, Ilona Koupil, Centre for Health Equity Studies, STOCKHOLM, Sweden Background: In a life course perspective, heavy drinking is considered to be one of the important trails that lead to deprived health and serious diseases, premature death and health inequalities. Research suggests that there are large differences between socioeconomic groups in alcohol related morbidity and mortality, in particular that alcohol-related mortality is substantially higher in disadvantaged socioeconomic groups. The persisting independent effect of early life social environment on alcohol related morbidity and mortality, is, however, less well established and it is not known to what extent it is independent of adult social characteristics. Objectives: The aim of this analysis was to study the effect of parental social class on alcohol related mortality, and to investigate whether and how individual social trajectories can mediate this potential association. Methods: 6314 men and 5854 women from the Uppsala Birth Cohort Multigenerational Study (UBCoS) who were born in Uppsala, Sweden 1915–1929. Data on parental social characteristics from archived obstetric records were linked with data from the national Censuses and from Swedish Registers of deaths and Migration. Subjects were followed up for mortality from 1960 until 2002. Univariable and multivariable Cox proportional hazard models were used to analyse differences in survival time. Results: Lower social class both in childhood and in adult life were associated with an increased risk of death from alcohol-related causes whereas being married and having higher household disposable income as adult indicated a protective effect. In multivariate analyses, taking the adult social characteristics into consideration, the increased risk on alcohol-related mortality in men with parental manual social class observed in the univariate analyses (HR: 1.7, 95% C.I.: 1.2–2.6), was moderately diminished. Conclusions: Our results indicate that the influence of parental social class on alcohol-related deaths is gender-specific and is partly mediated via adult socioeconomic circumstances.
157 LIFECOURSE FAMILY INCOME INFLUENCES ON ADULT ORAL HEALTH: THE 1982 PELOTAS BIRTH COHORT M. A. Peres1, K. G. Peres1, W. Murray Thomson2, J. M. Broadbent2, D. P. Gigante3, B. L. Horta3 1Department of Public Health, Federal ´ POLIS, Brazil, University of Santa Catarina, FLORIANO 2 Department of Oral Sciences, Faculty of Dentistry, University of Otago, DUNEDIN, New Zealand, 3Department of Social Medicine, Federal University of Pelotas, PELOTAS, Brazil Background: In spite of a plethora of evidence showing that socioeconomic status is strongly associated with oral health the dynamics of how socioeconomic status over time affects adults0 oral health remains unclear. Objectives: This study aimed to test how three models of lifecourse socioeconomic status—social origin,
123
IEA-EEF European Congress of Epidemiology 2009 accumulation, and change—predict a relevant oral health outcome in young adulthood. Methods: This study is nested in a population-based birth cohort initiated in 1982 in Pelotas, in the south of Brazil. The 5,914 live-born infants were investigated during the life course providing socioeconomic (family income, mothers schooling), demographic, and health-related variables. Oral health studies were carried out when the cohort participants were 15 (n = 888) and 24 years-old (n = 720). We used the number of unsound teeth by age 24 as the main outcome. Family income at birth, and at ages 15 and 23 were used to compose four lifecourse income trajectories groups: never poor, downwardly mobile, upwardly mobile and always poor. Potential mediators and confounders (such as skin colour, smoking status, pattern of dental attendance, and dental calculus) were also collected. Poisson regression with robust variance provided unadjusted and adjusted Prevalence Ratios (PR). All analysis was stratified by mother’s schooling. Results: The sample comprised 332 (46.1%) who were never poor, 131 (18.2%) downwardly mobile, 94 (13.1%) upwardly mobile, and 163 (22.6%) who were always poor. Adjusted analysis showed that, the higher the number of poverty episodes during life, the higher the proportion of unsound teeth in both high and low schooling groups. Poverty at birth was associated (PR 2.0 95% CI 1.1, 3.5) with unsound teeth at adulthood only among those from mothers with high schooling. Conclusions: These findings confirm the hypothesised association between social origin and the cumulative risk of unsound teeth at young adulthood.
271 EARLY CHILDHOOD SOCIOECONOMIC CORRELATES OF QUALITY OF LIFE IN ADOLESCENCE J. Mazur, A. Dzielska, K. Mikiel-Kostyra, E. Smolen Institute of Mother and Child, WARSAW, Poland Background: Infancy is considered as critical period during which poor socioeconomic (SES) conditions can have lasting effect on the further health chances. Objective: To examine early childhood SES correlates of health-related quality of life (HRQL) in Polish 13-year-old adolescents. Methods: A secondary analysis of data from three–wave longitudinal study on the determinants and benefits of breastfeeding was performed. In the first wave, all infants born in Poland between January 1st–10th, 1995 participated and data from maternity wards were analyzed. After 3 years, 20% of mothers were sampled and they reported on living conditions (LC) in the first year of child life (0—adequate to child needs; 1—not adequate). In the next postal survey, conducted at the beginning of 2008, 605 families participated. HRQL was assessed by single dimension KIDSCREEN10 index transformed into 0–100 score. Two SES indicators were employed: FAS (Family Affluence Scale) reported by children and perceived family wealth (PFW) reported by parents. Mean KIDSCREEN scores were compared by LC, FAS and PFW levels using ANOVA. In multivariate linear regression model, FAS and PFW were recoded into dummy variables. Results: In total sample, mean KIDSCREEN index was equal to 73.8 (SD = 12.53), with no difference between boys and girls. In one-factor ANOVA, eta-squared was equal to 0.007 for PFW (P = 0.123); 0.014 for FAS (P = 0.022) and 0.017 in the model including LC (P = 0.002). In stepwise linear regression, firstly LC and then low FAS were entered, accounting for 1.5% and 1.0% of change in R2, respectively. Conclusions: Independent of current family wealth, adolescents were more likely to report lower HRQL if family had serious financial problems in their early childhood. Adverse living conditions may accumulate over the life course and reduce quality of life.
IEA-EEF European Congress of Epidemiology 2009
Early Morning Session 2—Children and adolescent health epidemiology II 179 IS PET OWNERSHIP IN EARLY CHILDHOOD ASSOCIATED WITH ASTHMA OR ALLERGIES IN SCHOOL AGE? META-ANALYSES WITH ORIGINAL DATA FROM 11 EUROPEAN BIRTH COHORTS—A GALEN INITIATIVE T. Keil1, S. Roll1, K. H. Carlsen2, P. Mowinckel3, A. H. Wijga4, B. Brunekreef5, M. Torrent6, G. Roberts7, S. H. Arshad7, I. Kull8, M. Wickman9, U. Kra¨mer10, A. von Berg11, E. Eller12 E, A. Høst12, S. Halken12, C. Kuehni13, B. Spycher13, J. Sunyer14,15, C. M. Chen16, J. Heinrich16, A. Arsanoj9, C. Puig15, O. Herbarth17, K. Van Steen18, S. N. Willich1, U. Wahn1, S. Lau1, K. C. Lødrup Carlsen19 for the working group ‘Birth Cohorts’ of GA2LEN (Global Allergy and Asthma European Network) 1Charite´, BERLIN, Germany, 2 University of Oslo, OSLO, Norway, 3Ullevl University Hospital, OSLO, Norway), 4National Institute for Public Health and the Environment, BILTHOVEN, The Netherlands, 5University Medical Center, UTRECHT, The Netherlands, 6Ib-salut Menorca Health Area, MINORCA, Spain, 7University of Southampton, SOUTHAMPTON, United Kingdom, 8Stockholm County Council, STOCKHOLM, Sweden, 9Karolinska, STOCKHOLM, Sweden, 10University of Duesseldorf, DUESSELDORF, Germany, 11Marien-Hospital Wesel, WESEL, Germany, 12Odense University Hospital, ODENSE, Denmark, 13University of Bern, BERN, Switzerland, 14CREAL, BARCELONA, Spain, 15IMIM, BARCELONA, Spain, 16Helmholtz Zentrum Mu¨nchen, NEUHERBERG, Germany, 17University of Leipzig, LEIPZIG, Germany, 18University of Ghent, GHENT, Belgium, 19Ulleva˚l University Hospital, OSLO, Norway Background and Objective: There is uncertainty about the association of pet ownership and allergies in childhood. We aimed to examine if pet ownership in the first 2 years of life is a risk or a protective factor for asthma and allergies at the age of 6–10 years. Methods: Using the raw data from 11 European birth cohorts that recruited 26,521 children in total, we defined ‘‘clean’’ categories of pet ownership during the first 2 years of life for cat(s), dog(s), bird(s), and rodent(s). Asthma at 6–10 years was based on doctor-diagnosis, parent-reported wheezing and asthma medication, if at least 2 out of these 3 factors were present. Allergic rhinitis at 6–10 years was assessed by parentreported symptoms plus a positive serum test (IgE) to aero-allergen(s). We estimated adjusted odds ratios for each study using logistic regression analyses and, for the combined estimates, by random-effect meta-analyses. Results: Furry pet or bird ownership ranged from 20 to 60%. Asthma prevalence was between 7% and 12% in most cohorts. There were no significant associations between any type of pet ownership in early childhood and asthma at 6–10 years in adjusted estimates in the individual cohorts or in the meta-analyses (ref-category was ‘‘no pets’’): cat only (adjusted OR 1.02, 95%CI 0.83–1.25), dog (0.88, 0.68–1.14), bird (1.12, 0.81–1.55), rodent (1.08, 0.74– 1.56). The prevalence of allergic rhinitis in early school age was on average 6.8%. There were also no statistically significant associations between any type of pet ownership and allergic rhinitis at 6–10 years: cat only (adjusted OR 1.02, 95%CI 0.80–1.30), dog (0.77, 0.55–1.07), bird (1.28, 0.92–1.78), rodent (0.83, 0.48–1.45). Conclusions: The results of this unique analysis of combined European birth cohort data suggested that pet ownership in early childhood is neither a risk nor a protective factor for asthma or allergic rhinitis in early school age.
37
114 RESIDENTIAL EXPOSURE TO MOTOR VEHICLE EMISSIONS AND THE RISK OF WHEEZING AMONG 7–8 YEAR-OLD SCHOOLCHILDREN IN NICOSIA, CYPRUS N. Middleton1,2, N. Nicolaou3, S. Pipis2, M. Zeniou2, S. Kleanthous4, P. Demokritou5, P. Koutrakis5, P. Yiallouros2 1Department of Nursing, Cyprus University of Technology, NICOSIA, Cyprus, 2 Cyprus International Institute for the Environment and Public Health in association with Harvard School Public Health, NICOSIA, Cyprus, 3 Respiratory Research Group, School of Translational Medicine, Wythenshawe Hospital, MANCHESTER, United Kingdom, 4Air Quality Section, Department of Labour Inspection, NICOSIA, Cyprus, 5Department of Environmental Health, Harvard School of Public Health, BOSTON, United States of America Background: First-time investigation into the frequency of asthmatic symptoms among Cypriot children showed increased prevalence in the capital compared to all other areas. Exposure to traffic pollutants have been repeatedly shown to be associated with a range of cardiorespiratory effects. Differences in asthma prevalence on an island the size of Cyprus may reflect such environmental factors. Objective: To investigate the association between residential exposure to traffic pollution and the risk of asthmatic symptoms within the city of Nicosia. Methods: The addresses of participating children in Nicosia were geo-coded (N = 1,735) and levels of exposure were assessed using both (a) distance-based and (b) emissions-based indicators e.g. estimated levels of traffic pollutants at all roads within a pre-defined distance around the residence. Odds ratios of self-reported symptoms in relation to exposure were estimated in logistic regression models adjusting for person-based risk factors and co-morbidity as well as intra-school clustering. Results: Up to 2-fold differences in symptoms were observed; however, increased risk seemed to concentrate at distances less than 50 m from a main road and/or only among those experiencing the highest levels of cumulative exposure. Associations persisted after adjusting for other risk factors. While the direction of effect was apparent at longer distances, differences were not statistically significant. Adjusted odds ratios for current wheezing were 2.33 (95% CI 1.27, 4.30) amongst the quartile of participants exposed to the highest levels of particulate matter (PM) at all roads 50 m of their residence and 2.14 (95% CI 1.05, 4.35) for levels of nitrogen oxides (NOx). Conclusions: Children with the highest vehicular emissions burden around their residence seem to be at higher risk of reporting symptoms. However, due to the small number of children residing at close proximity to main roads and with no evidence of effect at longer distances, this alone would not explain the generally higher prevalence observed in the urban areas in Nicosia.
40 LIFE PREVALENCE OF CHRONIC UPPER RESPIRATORY TRACT DISEASES AND ASTHMA AMONG CHILDREN RESIDING IN RURAL AREA NEAR A REGIONAL INDUSTRIAL PARK: CROSS-SECTIONAL STUDY I. Karakis1,2, E. Kordysh2, T. Lahav2, A. Bolotin2,Y. Glazer2, H. Vardi2, I. Belmaker3, B. Sarov2 1School of Public Health, Haifa University, HAIFA, Israel; 2Department of Epidemiology, Ben-Gurion University of the Negev, BEERSHEBA, Israel,
123
38 3
Southern Regional Health Department of the Israeli Ministry of Health, BEERSHEBA, Israel Background: Our study was initiated by the Israel Ministry of Health. Objectives: The objective of this study was to investigate the link between the regional industrial park (IP) nearness and life prevalence (LP) of upper respiratory tract chronic diseases (URTCD) and asthma in children 0–14 years in Negev’s agricultural localities, Southern Israel. Methods: The cross-sectional study was conducted in seven localities during year 2002 (550 children). We used the three indirect exposure indicators: 1) the distance (less/ more than 20 km) from the IP, 2) the presence (yes/no) of the dominant wind direction from the IP toward a locality, 3) the mother’s odour complaints (yes/no) related to the IP. Medical diagnoses were collected from the clinic records. The following parts were included in the questionnaire: 1) demography, 2) the child’s birth and breast feeding history, 3) the parental respiratory health, smoking and occupational exposure, 4) odour related to the IP emissions. For statistical analysis multivariate logistic regressions with adjusted odds ratio (OR) within 95% confidence interval were used. Results: The multivariate analysis shows that the increased LP of URTCD in proximal localities is significantly associated with odour complaints (OR = 3.76, [1.16, 12.23]). The following factors were related to increase in URTCD—LP: 1) father’s lower education (by distance: OR = 2.62, [1.23, 5.57]; by wind direction: OR = 4.07, [1.65, 10.03]); 2) in-vitro fertilization (by distance: OR = 3.03, [1.17, 7.87]; by wind direction: OR = 4.34, [1.48, 12.72]). In the proximal localities, the increase in asthma LP is associated with: 1) wind direction (OR = 1.95, [1.01, 3.76]); 2) the child’s male gender (OR = 2.95, [1.48, 5.87]); 3) the child mother’s acute infectious diseases during the pregnancy (OR = 4.84, [1.33, 17.63]). Conclusions: The increased LP of children chronic respiratory morbidity is associated with the indirect measurements of exposure to the IP emissions. These results were similar to our previous reported findings among other subgroups.
Early Morning Session 3—Clinical epidemiology II 263 CALIBRATION OF THE NEW YORK HEART ASSOCIATION FUNCTIONAL CLASSIFICATION SYSTEM BETWEEN OPERATORS M. Severo1,2, R. Gaio3, P. Lourenc¸o4, M. Alvelos4, P. Bettencourt4, A. Azevedo1,2,4 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health of the University of Porto, PORTO, Portugal, 3 Department of Pure Mathematics, University of Porto Science School, PORTO, Portugal, 4Heart Failure Clinic, Department of Internal Medicine, Hospital S. Joa˜o, PORTO, Portugal Background: Studies have shown that the New York Heart Association (NYHA) functional classification is valid but not reliable. Objective: To calibrate the NYHA classification system between operators, aiming to increase its reliability. Methods: Within the first follow-up of a cohort, representative of the non-institutionalized adult population of Porto, Portugal, at baseline, 1136 individuals aged C45 years were assessed by 7 physicians experienced in the management of heart failure and an echocardiogram was performed. Subjects reporting breathlessness (n = 273) were presented to a 4-item questionnaire on functional capacity to characterize the severity of symptoms. Each subject was assessed by one physician, who administered the questionnaire and classified the subject’s functional capacity using the NYHA classification. The calibration of
123
IEA-EEF European Congress of Epidemiology 2009 the NYHA classifications was carried out by the concurrent method, using the 1-dimensional 2-PL graded response model (GRM), with the 4-items as anchor items and the 7 obtained NYHA classifications as target items. Item and ability parameters were estimated for all items simultaneously, assuming missing in the target items as random. The estimated ability (standard normal distribution) was used to predict the NYHA classification globally, for all operators, and separately, for each operator, by classification trees. Final agreement rates and kappa coefficients evaluated the prediction quality of the ability. Results: The inspection of the thresholds between NYHAcategories I-II and II-III, for each operator, demonstrates discrepancies between operators in terms of ability. Estimates of the first and the second thresholds for each operator ranged from -1.43 to 0.42, and from 0.94 to 2.90, respectively. The global and operator-specific agreement between the ability and the NYHA classification was 72.5% (kappa = 0.57) and 94.0% (kappa = 0.90), respectively. Conclusion: Thresholds between operators to classify patients in NYHA class were very discrepant. Concurrent calibration through GRM models can be used to calibrate a large number of operators on the same scale.
231 SENSIBILITY AND RELATIVE INDEX OF SPECIFICITY OF THE INFLUENZA SYMPTOMS T. Vega1, J. E. Lozano1, R. O. de Lejarazu2, M. Gutie´rrez1, J. Bermejo2, J. M. Eiros2, J. Castrodeza1 1Consejerı´a de Sanidad de Castilla y Leo´n, VALLADOLID, Spain, 2Hospital Clı´nico Universitario de Valladolid, VALLADOLID, Spain Background: The established networks on flu surveillance in the majority of the European countries have not normalized criteria of notification. This makes difficult to perform comparisons between countries with different health systems. Objectives: This work is committed to analyse the sensibility and the specificity, through a relative index, of symptoms and signs which characterize ILI. Methods: Data came from the Health Sentinel Network of Castilla y Leo´n (Spain), seasons 1996/1997 to 2007/2008. ILI symptoms were independently studied: fever, sudden onset, cough, sore throat, dyspnea, respiratory symptoms, shivers, asthenia, myalgias, headache, epidemic period and contact with a flu patient. Sensibility is calculated as the percentage of cases with the symptom among all culture positives cases. Relative Index of Specificity (RISp) has been calculated as the ratio between cases without the symptom and cases with the symptom, weighted by the 1/(1 - sensibility). Results: Swabs were taken in 1,016 cases of ILI. In 548 patients influenza was confirmed by culture. The most sensible constitutional symptom was fever 96.3%, followed by asthenia 93.4%. All local symptoms showed similar sensibility, around 83–84%. Other epidemiological variables such as the epidemic period and contact with a flu patient had a lower sensibility, 76.5% and 29.2% respectively. Contact with a flu patient was the most specific variable with RISp of 3.04, and headache the second one (RISp of 2.22). Local symptoms such as cough and sore throat have also a high RISp. The variable with the lower index was sudden onset (0.97). Conclusions: ILI symptoms have a high sensibility and, probably, any combination of them give us a definition with a high sensibility, an excellent positive predictive value in epidemic period and negative predictive value in non epidemic period. Constitutional symptoms are the most sensible, mainly fever and asthenia, which must bee considered as the basic symptoms of the optimal definition. In order to provide specificity, constitutional and local symptoms should be combined with epidemiological criteria and local symptoms.
IEA-EEF European Congress of Epidemiology 2009
180 DETERMINANT FACTORS CONTRIBUTING TO THE DEATH OF ADULT INDONESIAN HAJJ PILGRIMS IN MECCA IN 1427 H ISLAMIC YEAR M. Pane1, N. Kodim2 1Directorate General of Diseases Control and Environmental Health, Ministry of Health, JAKARTA, Indonesia, 2 Departement Epidemiology, Faculty of Public Health, University of Indonesia, JAKARTA, Indonesia Every year there are 3–3.5 Million Moslem come to Arab Saudi for Hajj, Indonesia has the biggest population in Hajj. The mortality rate of Indonesian hajj pilgrims is higher than Indonesian population mortality and other countries. For the last 10 years, the mortality rate of Indonesian hajj pilgrims fluctuated ranging from 2 to 3.8 per 1,000 per hajj season. Objective: The objective is to know determinant factors for the death of Indonesian hajj pilgrims aged more than 40 years in Mecca in 1427 H hajj season. This research conducted to the 149.537 ordinary hajj pilgrims. Design of the study was cross-sectional ecological studies. Compositional factors that contribute to the death of ordinary Indonesian hajj pilgrims for the age more than 40 years are age, sex, educational level, and length of stay in Mecca before Armina and preexisting diseases. The men of the aged more than 40 years have 1.78 times higher risk of death (95% CI: 1.54–2.16) compared to the women. The older the pilgrims the higher risk dying. Pilgrims who have higher education have 1.59 times risk of death (95% CI: 0.93–1.67), while those who had lower education have almost 2 times higher risk of death (95% CI: 1.08–1.95) compared to those who have higher education. Pilgrims who have lung diseases have 2.6 times risk of death (95% CI: 1.54–3.74), pilgrims who have metabolic diseases have 2.5 times risk of death (95% CI: 1.75–3.29) and Pilgrims who have cardio-cerebro vascular diseases have 1.3 times risk of death (95% CI: 1.09–1.63) compared to healthy pilgrims. Thin pilgrims have 2 times risk of death (95% CI: 1.59–2.41) compared to those who have normal weight.Pilgrims who arrived earlier in the second turn have 1.4 times higher risk of death (95% CI: 1.01–1.99) and those who arrived later in the second turn have 1.88 times higher risk of death (95% CI: 1.34–2.65) compared to those who arrive in the first week. The highest contribution based on individual level is the pilgrims of 60 to 69 years old, 36.35%. The second highest contribution is those of 79 years old, 30%, men contribute 27.6% and low education contribute 29%. While IMT and acclimatization in the third to eight days before arriving at Armina, each contributes to 11% and 12%, preexisting diseases contribute less than 6% to the death rate. Adjusted real per capita expenditure, population without access to health facilities, adult, literacy, population with health problems and population self-medicating which contribute to the death after being controlled by other variables.
Early Morning Session 4—Social epidemiology II 143 MORTALITY AND CAUSES OF DEATH AMONG DRUGGED DRIVERS K. Karjalainen1, T. Lintonen2, A. Impinen1, P. Lillsunde1, P. Ma¨kela¨1, O. Rahkonen4, A. Ostamo1,3 1National Institute for Health and Welfare, HELSINKI, Finland, 2Police College of Finland, TAMPERE, Finland, 3Tampere School of Public Health, TAMPERE, Finland, 4University of Helsinki, Department of Public Health, HELSINKI, Finland Background: People suspected of driving under the influence of drugs (DUID) can be studied as drugged drivers, but also as a sample of drug users. The mortality of drug users is often studied among cohorts using iv-drugs and admitted for detoxification or treatment programs. Thus, studying DUID suspects gives
39 complementary information about the mortality of drug users. Objectives: This study analysed mortality rates and causes of death among drugged drivers in Finland and compared them with the general Finnish population during 1993–2006. Methods: The register data consisted of 5,832 DUID suspects apprehended by the police. A reference group of 74,809 of individuals was drawn from the general Finnish population. Deaths were traced from National Death Register. The survival and the differences in mortality hazard were estimated by using Kaplan-Meier plots and Cox regression models. Results: Both male and female DUID suspects had almost ten times higher hazard of death in comparison with the population. Among male DUID suspects cause-specific hazards were highest when the cause of death was poisoning/overdose, violence or suicide. One fourth of DUID suspects were under the influence of drugs or alcohol at the time of death, whereas under one tenth of the population was intoxicated while dying. Male DUID suspects had twice as high hazard of death as females. Cases with a finding for one drug only were likely to survive longer than cases with poly-drug findings. DUID suspects who had a finding for medicinal drugs (especially benzodiazepines) had a higher hazard of death than DUID suspects with a finding for illicit drugs (especially amphetamines). Conclusions: DUID suspects apprehended by the police had an increased risk for premature death, in all observed causes of death. Findings for sedatives/tranquillizers indicated excessive mortality over findings for stimulants. Preventive actions should be aimed especially at benzodiazepine users.
259 MENTAL HEALTH PROBLEMS AND CANNABIS USE IN THE GENERAL POPULATION, BELGIUM, 2004 M. Roelands Scientific Institute of Public Health, BRUSSELS, Belgium Background: In Belgium, treatment services reported that mental disorders were common among illicit drug users demanding treatment. However, it was unknown whether this was a general phenomenon. Objectives: The study aimed to describe the distribution of mental disorders and cannabis use in a general population sample. Research questions were: what is the proportion of persons with mental disorders in cannabis users, and what is the odds ratio of having a mental disorder in users versus non-users? Methods: The data of the general population survey in 2004 in Belgium was analysed (N = 13,831). Data were collected with validated instruments (e.g. GHQ) and analysed with SPSS. Illicit drug use was measured as last month cannabis use (n = 203). The analysed sample was weighted. Results: The proportion of persons with a double diagnosis (who had a mental disorder and also used cannabis) in the general population was estimated to be 0.5% (depressive disorder), 0.4% (anxiety disorder) and 0.6% (depressive or anxiety disorder), respectively. 0.7% of the population had a psychological impairment and used cannabis. In this representative sample of the population, 17.6% of cannabis users were found to have also a depressive disorder, 13.5% an anxiety disorder and 20.9% of cannabis users had a depressive or anxiety disorder. 24.1% of cannabis users had a psychological impairment. The odds of having a mental disorder was higher in users compared to non-users; this was found regarding a depressive disorder (OR = 3.07), an anxiety disorder (OR = 2.68), a depressive or anxiety disorder (OR = 2.96) and psychological impairment (OR = 2.33). Analysis with a similar measure (reported depression) confirmed the findings. Conclusions: Although cannabis users have increased odds to have also a mental disorder, the proportion of persons with a mental disorder is limited in a populationbased sample of cannabis users.
123
40
237 PREVALENCE OF ALCOHOL DEPENDENCE AND ASSOCIATED FACTORS. A BRAZILIAN POPULATION-BASED STUDY S. A. de Moraes, I. C. M. de Freitas University of Sa˜o Paulo, ˜ O PRETO, Brazil RIBEIRA Background: Alcohol abuse entails serious consequences for people’s quality of life and health, contributing to a large number of preventable deaths. Objectives: To estimate the prevalence of alcohol consumption and the severity of this habit, and to identify correlates to alcohol dependence. Methods: Cross-sectional population-based epidemiological study using three-stage cluster sampling. The variability introduced in the third sampling fraction was corrected by attributing equal weights to the number of eligible units in each domicile, resulting in a weighted sample of 2197 participants aged 30 years or older, living in the urban area of Ribeira˜o Preto-SP, Brazil, in 2006. Prevalence estimates were based on Alcohol Use Disorder Identification Test (AUDIT). Participants were classified in four dependence zones. For the association study, crude and adjusted
123
IEA-EEF European Congress of Epidemiology 2009 prevalence ratios (PR) were calculated and the dependent variable was classified in two levels, using scores C 8 as the cut-off point. The design effect was considered in all estimates. Linear trend tests were applied to the variables with more than two categories. Results: The global prevalence of alcohol dependence was 20.2% (CI95%: 17.4– 22.9), with 43.5% (CI95%: 38.3–48.6) for men and 10.2% (7.6–12.7) for women. Among the youngest participants, the prevalence was 29% (22.9–35.1), and for those aged 60 years the prevalence was 10.5% (5.4–15.5). Of those classified in dependence zone 4, 7.6% were men and 1.2% women. In the final model, the following variables were associated with alcohol dependence: male gender (PR = 3.23; CI95%: 2.42–4.30), age ‘‘30–39’’ (PR = 2.29; CI95%:1.42– 3.68), education ‘‘4–7 years’’ (PR = 1.34; CI95%:1.04–1.72), marital status ‘‘alone’’ (PR = 1.38; CI95%:1.13–1.69), ‘‘smoker’’ (PR = 2.66; CI95%:1.87–3.79) and ‘‘Catholic’’ religion (PR = 2.84; CI95%: 1.76–4.59). Smoking showed dose–response effect related to alcohol abuse (P = 0.000). Conclusion: The prevalence of alcohol consumption was high, mainly among men. The results may contribute to the adoption of health promotion and prevention strategies, considering the consequences from alcohol abuse and dependence.
IEA-EEF European Congress of Epidemiology 2009
Friday, 28 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology II 86 CUMULATIVE FALSE POSITIVE RISK AND ITS ASSOCIATED FACTORS IN A DATABASE FROM 10 POPULATION BASED BREAST CANCER SCREENING PROGRAMMES IN SPAIN R. Roma´n1, M. Sala1, R. Zubizarreta2, D. Salas3, N. Ascunce4, F. Belvis1, X. Castells1, and the RAFP research group 1Evaluation and Clinical Epidemiology Service, IMIM-Hospital del Mar, BARCELONA, Spain, 2Servizo de Programas Poboacionais de Cribado, Direccio´n Xeral de Sau´de Pu´blica, SANTIAGO DE COMPOSTELA, Spain, 3Oficina del Plan de Ca´ncer, Conselleria de Sanitat, VALENCIA, Spain, 4Seccio´n de Deteccio´n Precoz, Instituto de Salud Pu´blica, PAMPLONA, Spain Background: False Positive (FP) recalls are and adverse effect of breast cancer screening programmes (BCSP). The effect of protocol differences among screening programmes in the FP risk has not been studied. Objectives: To estimate the risk of a false positive and its association with the protocol and women characteristics, in a cohort of women from 10 population-based BCSP in Spain. Methods: Retrospective Cohort study of women aged 45–69 years invited to participate in any of 10 screening programmes. Multilevel Discrete Time hazard models were performed to estimate the FP risk in any of the women screening participations and to estimate the effect of protocol and women characteristics on the FP risk for both invasive and non invasive procedures. The FP risk was projected up to 10 screening rounds (the period a women is invited to participate in BCSP). Results: We analyzed 4.778.613 mammographic screenings from 1.584.419 women. The FP risk of non-invasive procedures was 8.04% (95%CI = 8.00–8.08) for first screening and 3.49% (95%CI = 3.47–3.51) for successive screenings. The FP risk of invasive procedures was 0.984% (95%CI = 0.969–1.00) for first screening and 0.246% (95%CI = 0.241–0.252) for successive. The cumulative FP risk after 10 screenings was 25.31% (95%CI = 24.90– 25.73) for women who entered the screening programme at 50–51 years of age and 39.15% (95%CI = 38.75–9.56) for women who entered at 45–46 years. The factors associated with a higher FP risk of non–invasive procedures were: double reading of mammography (OR = 1.97; 95%CI = 1.91–2.04), one view image (OR = 1.16; 95%CI = 1.14–1.17), initial recommendation of an early recall (OR = 1.47; 95%CI = 1.44–1.50), perimenopausal status (OR = 1.29; 95%CI = 1.27–1.31), use of hormone replacement therapy (OR = 1.05; 95%CI = 1.03–1.07), previous invasive procedures (OR = 1.48; 95%CI = 1.45–1.51), familial history of breast cancer (OR = 1.18; 95%CI = 1.16–1.21) and age 44–49 years (OR = 1.33; 95%CI = 1.29–1.37). Conclusions: Over one third of women could have a false positive recall over 10 biennial screens. Different protocol and women characteristics influence the risk of a FP. Women should be informed of these risks.
282 RAW CABBAGE/SAUERKRAUT CONSUMPTION DURING ADOLESCENCE AND BREAST CANCER RISK IN POLAND D. R. Pathak1, J. Charzewska2, B. Zemla3, Anna Kozaczka4, El_zbieta Nowakowska5, Dariusz Godlewski6, Piotr Zurawski7 1 Departments of Epidemiology and Family Medicine, Michigan State
41 University, E. LANSING, United States of America, 2Department of Nutritional Epidemiology and Dietary Allowances, National Food and Nutrition Institute, WARSAW, Poland, 3Department of Oncological Epidemiology, Center of Oncology, Maria SklodowskaCurie Memorial Institute, GLIWICE, Poland, 4Department of Internal Diseases and Oncological Chemotherapy, Silesian Medical Academy, KATOWICE, Poland, 5Center of Oncology, POZNAN, Poland, 6 OPEN, POZNAN, Poland, 7Center of Oncology, BIALYSTOK, Poland Background: Epidemiological studies are inconsistent about high cruciferous vegetables consumption and reduced breast cancer (BC) risk despite numerous animal and in vitro studies showing anti-carcinogenic effects of specific compounds found in cruciferous vegetables. Potential explanations for these inconsistencies include: lack of variability in consumption in studied populations, lack of accounting for bioavailability of these compounds and focus mainly on adult consumption. Objectives: To determine the effect of high cabbage/sauerkraut consumption during adolescence on breast cancer risk. Methods: The Polish Women’s Health Study (PWHS) is a population based case-control study of BC among Polish-born immigrants to US (Chicago and Detroit) and native women in Poland (Gliwice, Katowice, Poznan and Bialystok). The in-home interview included demographic information, standard risk factor questionnaire and 143-item food frequency questionnaire with one section focusing on cabbage/sauerkraut foods. The two time periods of dietary data were: 1985–1989, i.e., immediately preceding introduction of market economy to Poland (representing adult consumption) and when participants were 12–13 years old, representing adolescence. Analyses are based on 825 BC cases and 880 controls who participated in the study. Results: Using multivariate conditional logistic regression we observed a non-significant decrease in BC risk for total cabbage/ sauerkraut consumption during adolescence. However, when analysis were restricted to raw cabbage/sauerkraut foods, thus accounting for bioavailability of the protective compounds, a significant 15% reduction in BC risk was observed for each increase in 1 serving/ week (trend odds ratio: OR = 0.85 with 95% confidence interval, 95% CI: 0.74–0.97). Conclusions: These initial results from the Polish component of the PWHS show a significant decrease in breast cancer risk, approximately 15%, for each increase in one serving/ week of raw sauerkraut/cabbage consumption in adolescence. This decrease in risk is observed after adjustment for adult sauerkraut/ cabbage consumption, reproductive risk factors and other lifestyle factors relevant to breast cancer risk.
81 SMOKING, AN INDEPENDENT RISK FACTOR FOR INVASIVE CERVICAL CANCER. A NORDIC JOINT STUDY T. Luostarinen1, A. Simen Kapeu2, Matti Lehtinen3 for the CCPRBEU Network 1Finnish Cancer Registry, HELSINKI, Finland 2National Public Health Institute, OULU, Finland 3University of Tampere, TAMPERE, Finland Background: Cervical infection with high-risk types of human papillomavirus (HPV) is considered necessary but not a sufficient cause of invasive cervical cancer (ICC). Exposure to tobacco smoke is of interest as a cofactor in cervical carcinogenesis. Objectives: The aim was to assess whether the association between smoking and cervical cancer is due to an independent role of smoking in cervical cancer or due to residual confounding by oncogenic HPV types. Methods: A nested case-control study within four Nordic serum banks containing serum samples from approximately 1,000,000 women was performed. Linkage to population-based cancer registries resulted to 588 cervical cancer cases diagnosed
123
42 between 1975–2002 of which 445 were squamous cell carcinomas (SCC). Cases were diagnosed on average 9.6 years after the serum sampling. Five randomly selected controls were matched to serum bank, age at serum sampling, and storage time. Tobacco smoke exposure was measured by serum cotinine. Women with cotinine level in category, 20–100 ng/mL and C100 ng/mL, were considered as light and heavy smokers, correspondingly. IgG antibodies specific for HPV types 16 and 18, Chlamydia trachomatis and herpes simplex virus type 2 were determined. Odds ratios were estimated by conditional and unconditional logistic regression. Results: Cotinine levels corresponding to both light and heavy smoking were associated with increased risk of ICC and SCC with and without adjustment for antibodies to HPV16/18. The odds ratios (OR) for ICC and SCC, 2.1 (95% CI: 1.4–3.2) and 2.7 (95% CI: 1.7–4.3), related to heavy smoking were significantly increased among HPV16/18-seropositive women. A dose-response relation between cotinine level and the risk of ICC and SCC was found both in HPV16/18-seropositive and negative women. Cotinine level was not associated with an excess risk of adenocarcinoma. Conclusions: Smoking is an independent risk factor for ICC/SCC in women infected with oncogenic HPVs.
265 HELICOBACTER PYLORI INFECTION AND GASTRIC CARDIA CANCER: SYSTEMATIC REVIEW AND META-ANALYSIS M. C. Pinto1, B. Peleteiro1,2, N. Lunet1,2 1Department of Hygiene and Epidemiology, Medical Faculty of the University of Porto, PORTO, Portugal, 2Institute of Public Health - University of Porto (ISPUP), PORTO, Portugal Background: Helicobacter pylori infection is the most important risk factor for non––cardia gastric cancer, but no association is recognised between infection and cardia cancer. However, two aetiologically distinct types of cardia gastric cancer were recently proposed, one associated with H. pylori-induced atrophic gastritis and the other associated with non-atrophic gastric mucosa. Objectives: We aimed to quantify the association between H. pylori infection and gastric cardia cancer through meta-analysis. Methods: We systematically reviewed published articles on the association between H. pylori infection and gastric cardia cancer. Cohort and case-control studies were identified in published meta-analyses on the association between H. pylori and gastric cancer (up to 2003) and in PubMed (from 2003 to January 2009). From each article we extracted crude and adjusted (whenever available) estimates of the association between infection and cardia and non-cardia (whenever available) cancer. Summary relative risk (RR) estimates and 95% confidence intervals (95%CI) were computed using random effects models. Heterogeneity was quantified through the I2 statistic. Stratified analyses were conducted for studies from low- and highrisk settings (gastric cancer age-standardised (world) incidence rate [20.0 per 100,000—GLOBOCAN 2002). Results: For cardia cancer, the crude RR estimate was 1.00 (95%CI: 0.81–1.24; I2 = 62.3%; 38 studies), and the adjusted RR estimate 1.05 (95%CI: 0.82–1.33; I2 = 56.7%; 26 studies). The adjusted RR estimate was 1.60 in high-risk settings (95%CI: 1.33–1.93; I2 = 0%; 12 studies) and 0.76 in low-risk settings (95%CI: 0.59–0.98; I2 = 30.9%; 14 studies). For non-cardia cancer, the adjusted RR estimate was 1.91 (95%CI: 1.23–3.00; I2 = 87.9%; 11 studies) in high-risk settings and 3.63 (95%CI: 2.66–4.95; I2 = 67.4%; 18 studies) in low-risk settings. Conclusions: In high-risk settings the association between H. pylori infection and gastric cancer is similar for cardia and non-cardia cancers.
123
IEA-EEF European Congress of Epidemiology 2009
Late Morning Session 2—Perinatal epidemiology II 166 PREVALENCE OF CONGENITAL HEART DEFECTS IN NEWBORNS IN GERMANY: RESULTS OF THE PAN STUDY H. W. Hense1, A. Lindinger2, G. Schwedler3 1Clinical Epidemiology Section, Institute of Epidemiology and Social Medicine, University of Muenster, MUENSTER; Germany 2Clinic of Pediatric Cardiology, University Hospital, HOMBURG/SAAR, Germany, 3Epidemiology Unit, Competence Network for Congenital Heart Defects, Germany Objective: To investigate the current prevalence of congenital heart defects (CHD) in newborns in Germany and to assess its relation to demographic and gestational factors. Design/Setting: PAN (‘‘Pra¨valenz angeborener Herzfehler bei Neugeborenen in Deutschland’’) is a nationwide study that includes prospectively all live births with cardiovascular malformations in Germany. Data were provided by a total of 260 participating institutions throughout Germany. CHD diagnoses were coded according to the standard set by the Association for Pediatric Cardiology. We here present the report on the first year of complete registration. Results: A total of 7,245 infants with CHD born between 1st July 2006 and 30th June 2007 were entered into the PAN database. Based on the denominator population of 673,287 live births in Germany during this period, the total CHD prevalence was 1.08%. The most common lesion was ventricular septal defect (all types) at a proportion of 48.9%, followed by atrial septal defect (17.0%), valvular pulmonary stenoses (6.1%), persistent arterial duct (4.3%) and aortic coarctation (3.6%). The most common cyanotic lesions were tetralogy of Fallot (2.5%) and D-transposition of the great arteries (2.2%). A single ventricle (all types) was identified in 2.8%, half of them being a hypoplastic left heart syndrome. Female gender was more common among mild cardiovascular defects (57.3%) while there was a striking predominance of male infants among those with severe lesions (58.4%). Preterm delivery and a birth weight below 2,500 g were about three times more common in infants with CHD than the national average of live births. The rate of CHD was doubled in multiple births. More than 80% of the diagnoses were made within three months after birth. Conclusions: The first registration year of the nationwide PAN study has revealed a prevalence of CHD in Germany of 1.08 per 100 live births. Haemodynamically insignificant cardiovascular lesions with minor degrees accounted for about two thirds of all defects. The occurrence of CHD was associated with prematurity, low birth weight and multiple births.
72 LENGTH OF STAY ON NEONATAL UNITS OF VERY PRETERM BABIES IN A UK POPULATION B. Manktelow, E. Draper, D. Field Department of Health Sciences, University of Leicester, LEICESTER, United Kingdom Background: The length of stay on a neonatal unit of very preterm babies is of interest to parents and neonatal teams. In addition, the anticipated move to a daily tariff for such services in England has made length of stay in such high cost low volume services of much interest to Commissioners. Objectives: The aim of this paper was to produce estimates of the length of stay on neonatal units of babies born at less than 33 weeks gestational age for a geographically defined UK population. These variations were then used to compare the observed lengths of stay between individual units in the study area. Methods: The data were obtained from The Neonatal Survey, an on-going study of neonatal care activity in the East Midlands and
IEA-EEF European Congress of Epidemiology 2009 Yorkshire regions of England. Estimated 10th, 50th and 90th percentiles for length of stay were estimated using least-absolute value regression by gestational age (completed weeks), birthweight (categorised into 250 g bands) and the need for respiratory care within the first 12 hours of life. All functions and their interactions were modelled using fractional polynomials. Results: Data for 4702 babies were included in the analysis, 1215 (26%) of whom required respiratory support within the first 12 hours of life. Simple to use charts of estimated length of stay were produced with estimated 10th, 50th and 90th percentiles for babies by gestational age, birthweight and requirement for early respiratory support. Although the model estimated the observed percentiles well across clinically relevant groups (e.g. gestational age, sex, year of birth), the prediction was much less good for individual units. Conclusion: It seemed highly likely that the lack of prediction when assessing the charts against individual units was the result of local policies and services governing the timing of discharge home.
230 INFANT MORTALITY BY GESTATIONAL AGE GROUPS: A POPULATION BASED STUDY IN AN ITALIAN REGION S. Farchi, D. Di Lallo, F. Franco, A. Polo, Public Health Agency Lazio Region, ROME, Italy Background: In Italy official statistics don’t produce infant mortality rates (IMR) by gestational age (GA) due to the lack of record linkage between mortality and birth records. Objectives: The aim of this population-based study was to estimate IMR by GA obtained from a record linkage procedure with attention on late preterm or at term births. Methods: Using data from the Lazio birth database2001–2005 linked to regional mortality registry2001-2006, we estimated IMR by GA. Four survival regression models were applied to estimate the hazard risk (HR) of neonatal (0–28 days) and post neonatal deaths ([ 28 days) for late preterm and at term infants. Results: The cohort was composed by 255,254 live births, 1112 died at 1 year (IMR = 4.4%). The IMR was: 524.5% at 22–27 weeks, 98.5 at 28–31, 11.2 at 32–36, 1.53 at 37-41 and 0.87 at 42 or more. The proportion of postneonatal deaths for the five GA groups was respectively: 9.1%, 13.1%, 23.5%, 44.9% and 33.3% (P \ 0.001). The results of the four models showed that neonatal morbidity factors were strongly associated to both neonatal and postneonatal deaths; migrant status (defined by mother’s country of birth) was predictor of 32–36 weeks neonatal mortality and 37–41 weeks post-neonatal mortality; younger mothers (\20 years) were at higher risk of postneonatal mortality in the 37–41 weeks and caesarean section was associated to neonatal and postneonatal mortality in the 37–41 weeks. Conclusion: Our study produce original data on IMR in a Italian region and it highlights a relevant higher risk of mortality of late preterm compared to at term births. It shows that clinical conditions at birth are important predictors even of late mortality. Non clinical factors, such as migrant status and caesarean section are strongly associated to mortality.
241 ANTENATAL AND PERINATAL FACTORS IN CHILDHOOD TYPE 1 DIABETES: A CASE-CONTROL STUDY IN THE MIDDLE EAST A. Memon1, S. Polack2, M. A. Al-Khawari3, M. Qabazard3, A. Al-Adsani3, M. Abdul-Rasul3, R. R. Gulati3, I. Siddique4,
43 N. A. Abdella4, A. A. Shaltout4, A. Suresh4 1Brighton and Sussex Medical School, BRIGHTON, United Kingdom, 2London School of Hygiene and Tropical Medicine, LONDON, United Kingdom, 3 Ministry of Health, KUWAIT, Kuwait, 4Faculty of Medicine, Kuwait University, JABRIYA, Kuwait Background: There has been a five-fold increase in the incidence of childhood Type 1 diabetes in Kuwait over the past two decades (from 4.0/100,000 in 1980–1981 to 21/100,000 in1992–1997). Little is known about the epidemiology of Type 1 diabetes in the Middle East, where the incidence appears to be increasing. Objective: To identify environmental factors in the antenatal/perinatal period that may be associated with the subsequent onset of Type1 diabetes during childhood. Design: Population based case–control study. Setting: Kuwait Childhood Diabetes Registry and the national network of diabetes and primary care clinics in Kuwait (population 2.8 million). Methods: 376 case–control pairs (aged 0–19 years) individually matched on age, gender, nationality, and district of residence. Information on factors in the antenatal/perinatal period was obtained through a personal interview with the mother of the subject, and the data were recorded in a structured questionnaire. Results: Univariate conditional logistic regression analysis showed that maternal age C 25 years (OR = 1.8, 95% CI: 1.3–2.6), paternal age C 30 years (2.1, 1.5–2.9), gestational diabetes (1.8, 1.1–3.2), birth order C 3 (2.3, 1.4–3.7), having [ 5 siblings (1.9, 1.1–2.1), diabetes in sibling(s) (8.3, 2.9–23.2), history of autoimmune disease in mother (1.7, 1.1–2.5), and history of congenital malformation in the family (3.9, 2.1–7.1) were associated with a significantly increased risk of Type 1 diabetes; whereas, maternal use of coffee, analgesics and vitamins/food supplements, and vaccination during the pregnancy were significantly protective. There was no association with parental consanguinity, parity, gestational age, caesarean delivery, or birth weight. In the multivariate analysis, birth order C4 and maternal use of vitamins/food supplements and vaccination during the pregnancy remained independently associated with type 1 diabetes. Conclusions: It appears that a host of environmental factors in the antenatal/perinatal period influence the risk of childhood Type 1 diabetes. Overall, the study presents a profile of increased risk with increasing age of parents, birth order, and number of siblings. Factors that affect maternal immune system during pregnancy may be protective, and require further study.
Late Morning Session 3—Cardioepidemiology II 172 PASSIVE SMOKING AND SMOKING CESSATION AMONG PATIENTS WITH CORONARY HEART DISEASE ACROSS EUROPE. RESULTS FROM THE EUROASPIRE III STUDY J. Heidrich1, U. Keil1, J. Wellmann1, D. De Bacquer2, Z Reiner3, C. Jennings4, Z. Fras5, C. Prugger1 1University of Muenster, MUNSTER, Germany, 2University of Ghent, GHENT, Belgium 3 University Hospital Center, ZAGREB, Croatia, 4Imperial College London, LONDON, United Kingdom, 5University Medical Center, LJUBLJANA, Slovenia Background: Passive smoking (environmental tobacco smoke, ETS) is a cardiovascular risk factor. Patients with coronary heart disease (CHD) are at particular risk from ETS since their risk of recurrent coronary events is high. Objectives: To describe the occurrence of passive smoking among patients with CHD and to investigate the association between passive smoking and smoking cessation. Methods: The EUROASPIRE III study is a cross-sectional study conducted in 2006–2007 among 8966 coronary patients aged up to 80 years from
123
44
IEA-EEF European Congress of Epidemiology 2009
22 European regions. Patients were interviewed on average 15 months after hospitalization for CHD using standardised methods. Smoking history, passive smoking and carbon monoxide (CO) in breath were assessed at interview. Results: Overall, 16% of patients were active smokers, 17% reported ETS exposure at home and 12% at workplace. Exposure was lower among non-smokers (11% and 9%, respectively) than in active smokers (44% and 29%, respectively). Home ETS exposure in non-smokers differed between regions (range 4% to 31%), women and men (15% vs. 10%, P \ 0.001), lower and higher education (12% vs. 10%, P = 0.007), and decreased with age (16% age \ 50 years vs. 8% age [ 70 years, P \ 0.001). Nonsmokers with ETS exposure showed higher mean CO levels in breath than those without ETS exposure (2.6 vs. 1.7 ppm P \ 0.001). Among patients who were smoking before hospitalization, the cessation rate at interview was substantially lower in those with home ETS exposure than in those without (28% vs.61%; adjusted odds ratio 0.27, 95% confidence interval 0.22–0.34). This finding applied also to ETS exposure at workplace (55% vs. 35%, OR 0.58; 95% CI 0.46– 0.74). Conclusions: A noteworthy proportion of CHD patients is exposed to passive smoking. A reduction of ETS exposure is important for secondary prevention. Additional passive smoking in active smokers is a major barrier for smoking cessation and needs to be addressed in quitting strategies.
and those with CVD or risk factors varied to a limited extent regarding the type, frequency and settings for PA. However, CVD risk stratified models demonstrated different patterns of associations with regular PA. Socioeconomic status, for example, was important for respondents with CVD-related risk factors, while attitudes towards PA were important for apparently healthy respondents and those with CVD. Conclusions: Our results point to the potential value of tailoring interventions to promote PA by accounting for the needs of differing clinical groups and influences that arise from these individuals’ socioeconomic, health, behavioural and psychosocial characteristics. Further studies are needed to investigate the impact of attitudes on PA in differing clinical groups to optimize health promotion strategies.
154 WHICH FACTORS INFLUENCE PHYSICAL ACTIVITY IN PERSONS WITH CARDIOVASCULAR DISEASE OR RISK FACTORS? A COMPARISON OF PATTERNS, SETTINGS AND ATTITUDES BY LEVEL OF CARDIOVASCULAR RISK
Background: An evolution in the association of socioeconomic position (SEP) with coronary heart disease (CHD), as a function of stages of economic development, has been described in Western countries. Elsewhere, less work has been done on this topic, and very little in women. Objectives: The aim of this study was to determine the existence, strength and direction of the association of SEP and education with CHD incidence in Israelis of both sexes. Methods: The populationbased Jerusalem LRC parent cohort which comprised 6921 Jewish adults initially examined in 1976–1980 was followed for 21–25 years. The analysis reported here is restricted to 2802 spouse pairs [mean ages (SD): men 49.1 (6.7), women 44.3 (5.5) years] who experienced 1296 documented coronary events during follow-up, as determined by record linkage and patient chart review. A 6-point SEP scale for men was based on their occupation and for women on their spouses’ occupation. Hazard ratios were estimated from age- and multivariable-adjusted time dependent Cox regression models that controlled for the conventional CHD risk factors. Results: In women we found strong graded ageadjusted associations of SEP and education with CHD incidence [HR = 5.4 (95% CI 3.2–9.2) for lowest vs highest household SEP, and HR = 3.2 (95% CI 2.2–4.5) for 0–4 vs 13? years of woman’s secular education], that were not mediated by conventional risk factors. The association with education was entirely explained by SEP. In women the population-attributable fractions for SEP and education (30% and 34%, respectively) far exceeded those due to any single coronary risk factor. In men the associations were significantly weaker. Conclusions: Low SEP and low education were the major determinants of CHD incidence among these Israeli women, but not in men. Low SEP Israeli women who were middle-aged in the 1970 s seem to have been exquisitely vulnerable to CHD.
C. Huy1, K. Steindorf2, D. Litaker1,3, A. Thiel4, C. Diehm5 1 Mannheim Institute of Public Health, Social and Preventive Medicine, University Medicine Mannheim, Heidelberg University, MANNHEIM, Germany, 2Unit of Environmental Epidemiology, German Cancer Research Center, HEIDELBERG, Germany, 3 Departments of Medicine, Epidemiology and Biostatistics, Case Comprehensive Cancer Center, Case Western Reserve University, CLEVELAND, United States of America, 4Department of Sports Science, University of Tuebingen, TUEBINGEN, Germany, 5 Department of Internal Medicine, Clinical Center and Academic Hospital of the University of Heidelberg, KARLSBADLANGENSTEINBACH, Germany Background: Little is known about physical activity (PA) patterns among adults at differing levels of cardiovascular risk. This information may inform the development of targeted interventions to address the public health burden of cardiovascular disease (CVD). Objective: The objective of this study was to describe and compare PA patterns and correlates between groups with different level of cardiovascular risk: those with and without a known CVD diagnosis and those with CVD-related risk factors only. Methods: In 2006, we conducted computer-assisted telephone interviews of 2,002 men and women aged 50 to 70 years, who resided in southern Germany. We assessed sports and leisure PA preferences, attitudes and the settings in which these activities took place. The groups were compared by using Chi square and non-parametric tests. Stratified multivariable regression models identified between-group differences in the association of a range of factors associated with regular PA (C1 h/week per 6-month period). Results: 677 (33.8%) of respondents had no medically diagnosed CVD or risk factor; 1,020 (51.0%) reported at least one cardiovascular risk factor but no disease; and 305 (15.2%) reported medically diagnosed CVD. Apparently healthy respondents
123
272 GENDER DIFFERENCES IN THE ASSOCIATIONS OF SOCIOECONOMIC POSITION AND EDUCATION WITH CORONARY HEART DISEASE INCIDENCE H. Arik1, B. Adler2, R. Sinnreich2, N. Goldberger2, J. D. Kark2 1 Dokuz Eylul University, Faculty of Medicine, Department of Public Health, IZMIR, Turkey, 2Hebrew University, Hadassah School of Public Health, JERUSALEM, Israel
206 DIET AND RURAL-URBAN MIGRATION IN INDIA: THE INDIAN MIGRATION STUDY L. Bowen, S. Ebrahim, B. De Stavola London School of Hygiene and Tropical Medicine, LONDON, United Kingdom Background: There is an emerging nutrition transition in India. Migrants from rural to urban areas offer an insight into the changes in
IEA-EEF European Congress of Epidemiology 2009 lifestyle factors that are happening. Objectives: To assess the impact of rural-urban migration on diet in India. Methods: Data used were from the Indian Migration Study- a cross-sectional sibling-pair study based in factories in North (Lucknow), Central (Nagpur) and South (Hyderabad, Bangalore) India. The places of origin of people working in factories in the four cities were identified. Migrants of rural origin and their rural dwelling siblings, and those of urban origin were recruited into the study. Dietary intake was measured by interviewer administered food frequency questionnaire and average daily intake of nutrients and food groups was derived. Results: A total of 6509 subjects (42% women, mean age 41) were included. Median daily energy intake was higher in migrant (2782 kcal) and urban (2938 kcal) subjects than in rural (2525 kcal) subjects, as was macronutrient intake (all P \ 0.001). Percentage of calories from fat and protein were higher in migrant and urban than rural subjects, and percentage of calories from carbohydrates was higher in rural than migrant and urban subjects (all P \ 0.01). Migrant and urban subjects reported a more energy-dense diet, a higher number of total foods and western foods eaten throughout the year, and a higher quantity and diversity of fruit and vegetables than rural subjects (all P \ 0.001). There wasn’t a clear pattern for foods of animal origin, with results varying by geographical region. Conclusions: Rural-urban migration is associated with important dietary changes, some of which may contribute to an increased risk of non-communicable diseases. Dietary differences are modifiable, and understanding the ways in which dietary behaviour changes with migration could help inform health promotion campaigns.
Late Morning Session 4—Infectious diseases epidemiology II 174 PHYSICIANS’ PRESCRIBING PRACTICES AND ANTIBIOTIC RESISTANCE IN GERMANY E. Velasco1, W. Espelage2, I. Noll1, A. Barger3, T. Eckmanns1 1 Robert Koch Institute, BERLIN, Germany, 2Federal Institute for Risk Assessment, BERLIN, Germany, 3Federal Ministry for Health, BERLIN, Germany Background and Objectives: Growing populations of older and immunocompromised patients, changes in epidemiology and unchecked use contribute to more consumption of antibiotics. Medical doctors (MDs) often have an important role alongside contributing factors. We conducted a national survey of MDs on their behaviours and expectations for intervention. We aimed to assess MD behaviours with and influences on antibiotic prescribing and resistance and the potential for related prevention interventions. Methods: We surveyed a representative sample of 10,610 MDs from stationary and ambulatory settings (respectively: 36%, 0% internists; 0%, 54% general practitioners; 32%, 11% surgery; 3%, 4% ear/nose/throat; 10%, 10% paediatrics; 5%, 3% urology; 9%, 13% gynaecology; 2%, 4% dermatology; 3%, \1% other) in 15 federal states. We developed study questions to capture baseline information and MD practice with antibiotics, and influences that affect MD behaviour in practice. Other questions solicited opinions about interventions to improve practice. Mailed questionnaires were distributed to participants via state medical associations. Results: Among survey respondents (n = 3,492; response rate = 33%), 66% reported that they prescribe antibiotics daily, and 90% indicated they prescribe at least weekly. Of all surveyed MDs, 60% reported that they think their own prescribing practice influences antibiotic resistance in their region. Of all MDs, 83% found it ‘‘important’’ to continually improve use of antibiotics through industry independent experts providing consultation, audits and feedback. Of all MDs, 96%
45 found it ‘‘important’’ to have provision of regional coverage of antibiotic resistance with appropriate feedback for practicing MDs, and 82% found it ‘‘important’’ to have regulations for prescribing antibiotics with appropriate feedback. Conclusion: Many MDs prescribe antibiotics often and believe their practice may have an affect on antibiotic resistance. They show interest in interventions, such as industryindependent consulting, which serve as an integral part of relevant prevention strategies for antibiotic resistance in Germany.
139 SEROEPIDEMIOLOGICAL DESCRIPTION OF CHLAMYDIAL INFECTION IN SAINT PETERSBURG, RUSSIA L. B. Kulyashova, L. A. Berezina, A. B. Zhebrun St. Petersburg Pasteur Institute, SAINT PETERSBURG, Russia Chlamydia trachomatis causes various reproductive organs diseases.There have been only fragmentary data of urogenital chlamydia infection within the territory of Russia. The objective of this work is to study the prevalence of infection associated with Chlamydia trachomatis among the citizens of St.Petersburg. In 2008 our institute conducted screening of 1678 individuals including 768 children and adolescents aged from 0 to 18 years old and 910 adults aged 19– 78 years old. The diagnostic groups included patients with chronic pelvic inflammatory disease, with malignant neoplasms of urogenital tract as well as patients without clinical manifestations of Chlamydia infection. Methods: IgG screening for Chlamydia trachomatis antibodies was performed using ELISA method with test-system produced by DRG and Medac. Results: The results of the screening showed that 73 (9.5%) screened children and adolescents possess IgG to Chlamydia trachomatis within the reference meanings. The analysis of different groups performance during the screening relative to age allowed to establish the following: with children aged 0–1 years. antibodies were discovered in 9.3% of the screenings; with children aged 5–12 years. this figure comprised 3.4%; with the adolescents group (13–18 years) it was marked by considerably biger number of seropositive patients comprising 17.3%. In the group of adult patients without clinical manifestations of Chlamydia infection the established infection rate was 11.6%, whereas among those with diagnosed infertility antibodies to Chlamydia trachomatis were discovered in 33% of the screenings. In the case of patients with malignant neoplasms of urogenital tract the figure of 30.6% was achieved. Conclusions: On the basis of the conducted screening high level of Chlamydia trachomatis infection among new-born children and sharp increase of the infection among adolescents were established. The findings argue in favour of an urgent need to develop anti-infection measures in order to prevent Chlamydia trachomatis infection among population.
127 RISK FACTORS AND IMPACT OF SURGICAL SITE INFECTION IN INDONESIANS HOSPITAL J. Atthobari1, D. V. P. Sianipar2, Suwardiman2, E. M. C. Menur2, I. Dwiprahasto1 1Department of Pharmacology, Faculty of Medicine, Gadjah Mada University, YOGYAKARTA, Indonesia, 2Department of Public Health, Faculty of Medicine, Gadjah Mada University, YOGYAKARTA, Indonesia Background: Surgical Site Infection (SSI) is a major problem in healthcare services particularly in developing countries. Previous
123
46 studies showed that incidence of SSI in Indonesian hospitals are varied between 3 and 18%. Nevertheless, the risk factors, impact on length of stay (LOS) and medical costs on SSI in Indonesian hospital is unknown. Objective: To know the risk factors of SSI and its impact on LOS and medical costs in Indonesian Hospitals. Methods: We conducted observational study using the hospitals database at 2002–2007. The subjects were all patients underwent obstetric surgery intervention and appendectomy in three district hospitals in West-Kalimantan, South-Sumatera and North-Sulawesi. SSI was defined according to the criteria of the Centers for Disease Control and Prevention (CDC). The data were analyzed using logistic regression analysis. Results: Among 860 subjects were included in the study, we identified 229 (26.7%) get SSI (annual incidence rate 5.4 cases/100 operations). The risk factors related to SSI were duration of surgery [45 minutes (OR = 2.0; 95%CI 1.1–3.6), non-elective surgery (OR = 5.1; 95%CI 2.4–10.6), postoperative LOS [7 days (OR = 3,2; 95%CI 1.5–6.5), preoperative leukocytes [10000/lL (OR = 2.2; 95%CI 1.3–3.7). Other risk factors such as age, preoperative LOS, antibiotic use, preoperative hemoglobin level and comorbidity were not statistically significant. The average LOS was higher among patients having SSI (5.8 days) for compare to 3.1 days for those without SSI (P \ 0.001). The estimate of incremental costs for patients with SSI was IDR 1.1 million (US$120). The number of postoperative visits in outpatient clinic for those with SSI was 2.3 times higher than those without SSI with incremental costs IDR 200,000 (US$22). Conclusion: The incidence of SSI in Indonesian hospitals is relatively high. Duration of surgery, non-elective surgery, postoperative LOS, leukocytes level are associated with the high incidence of SSI. The SSI is associated with a greater extra LOS and medical costs.
74 COULD STETHOSCOPES BE A SOURCE OF INFECTION? S. Rehman, H. Razzaq, O. Anwar, J. Macfie Combined Gastroenterology Research Department, Scarborough General Hospital, SCARBOROUGH, United Kingdom
123
IEA-EEF European Congress of Epidemiology 2009 Background: Nosocomial infections cause serious morbidity and mortality in the hospitalised patients. Many types of equipment including stethoscopes have been implicated in their transmission. Objectives: This study aim at evaluating, the stethoscopes carry potential pathogens including staphylococcus aurous and gram negative organisms. Healthcare professionals play significant role in preventing nosocomial infections, using effective decontaminant such as isopropylene alcohol. Methods: One-hundred and forty-three stethoscopes belonging to, senior physicians, residents, medical students’ physiotherapists and nursing staff at Surgery, Medicine, Paediatric wards, ITU, and CCU were sampled. Participants were asked of their habits of cleansing stethoscopes and agents used. Samples were taken one week apart for two weeks, with prior warning in the second stage of study. Microbial culture and antibiotic sensitivity testing were carried out. Results: Almost all (89.24%) of the samples were positive, with coagulase negative staphylococcus aurous was the commonest organism isolated. Two (2.15%) of the samples were positive for methicillin sensitive staphylococcus aurous (MSSA) and none of them were positive for methicillin resistant staphylococcus aurous (MRSA). No growth went up (from 1% to 7.5%) in the second stage of study, whereas the number of colony forming units (CFU) fell significantly from (0–50, mean = 20) to (0–30, mean = 8). Only 4.3% (8/143) of the health care professionals admit to clean their stethoscopes after every patient compared to 35.1% (66/143) who did so weekly, however 12.2% (23/143) of health care professionals never cleaned their stethoscopes. Alcohol wipe was the most frequently (39%) used cleansing agent. Conclusion: Stethoscopes carry potentially dangerous pathogens. MRSA is rare, but isolation of MSSA and gram negative organisms from stethoscopes poses real threat to critically ill patients on the wards and intensive care. Health professionals play equivocally significant role and campaigning does lead to increase frequency of cleansing.
IEA-EEF European Congress of Epidemiology 2009
Friday, 28 August 2009 Parallel afternoon oral sessions: 16.30–17.50 Afternoon Session 1—Genetic epidemiology 137 SUSCEPTIBILITY GENES AND AGE-RELATED MACULAR DEGENERATION H. W. Hense1, A. Farwick1, J. Wellmann1, M. Stoll1, D. Pauleikhoff2 1 ¨ NSTER, Germany, 2St.-FranziskusUniversity Mu¨nster, MU ¨ NSTER, Germany Hospital, MU Background: Little is known about the role of genetic variants in the early stages of age-related macular degeneration (AMD). Objective: We aimed to investigate how genetic variations within five well-defined genes relate to AMD severity. Methods: We analysed SNPs in the genes for complement factor H (CFH), age-related maculopathy susceptibility (ARMS2), HtrA serine peptidase 1 (HtrA1), complement factor B (CFB), and complement component 2 (C2)in 913 patients from the prospective Muenster aging and retina study (MARS). Severity scoring at baseline was based on the bilateral Rotterdam classification of fundus photographs, progression was based on changes in single eyes after a median of 2.6 years. Results: As compared to 180 subjects free of AMD at baseline, patients with very early AMD showed a significantly increased minor allele frequency (MAF) for CFH-rs1061170. With increasing severity of AMD, SNPs in CFH-rs1061170,as well as ARMS2-rs10490924, became consistently more common (P \ 0.001). HtrA1-rs11200638 was less clearly associated with increased AMD severity, C2-rs9332739 and CFB-rs641153 were unrelated. Progression to early AMD after 2.6 years was associated with the CFH-Gene variant (OR 1.9 [1.3–3.0] heterozygous; 2.9 [1.7–5.1] homozygous). However, CFH-polymorphisms were unrelated to progression to late stages as opposed to the ARMS2-gene (=R = 1.6 [0.9–2.7] heterozygous; OR = 3.55 [1.9–6.6] homozygous). Adjustment for baseline AMD stage in the contraletral eye attenuated, but did not remove the observed relations. Conclusions: Our findings indicate that genetic factors exert a major impact on onset and course of this chronic disease in the elderly: while the CFH-gene is relevant for the onset of AMD, both, CFH and ARMS2, and more weakly, the HtrA1gene, appear to account for the advancement of AMD.
247 ASSOCIATION BETWEEN CYTOKINE GENE POLYMORPHISMS AND GASTRIC PRECANCEROUS LESIONS B. Peleteiro1,2, C. Carrilho3,4, L. David5,6, J. C. Machado5, C. La Vecchia7,8, N. Lunet1,2 1Department of Hygiene and Epidemiology, Medical Faculty, University of Porto, PORTO, Portugal, 2Institute of Public Health, University of Porto, PORTO, Portugal, 3Department of Pathology, Medical Faculty, Eduardo Mondlane University, MAPUTO, Mozambique, 4Department of Anatomical Pathology, Maputo Central Hospital, MAPUTO, Mozambique, 5Institute of Molecular Pathology and Immunology, University of Porto, PORTO, Portugal, 6Department of Pathology, Medical Faculty, University of Porto, PORTO, Portugal, 7Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’, MILAN, Italy, 8Istituto di Statistica Medica e Biometria, Universita` di Milano, MILAN, Italy Background: Genetic variations within the IL1 and TNFA gene clusters have been associated with the increased risk of developing gastric cancer, but little is known about its effect on gastric precancerous lesions. Objectives: The aim of this study was to quantify the association between IL1B, IL1RN and TNFA gene polymorphisms
47 and the occurrence of gastric precancerous lesions, using original data from Portugal and Mozambique and conducting a systematic review and meta-analysis. Methods: Mozambican dyspeptic patients and Portuguese volunteer shipyard workers underwent an upper digestive endoscopy, and four biopsy specimens were collected from each subject. The same pathologists observed both samples and polymorphisms genotyping was performed. Published studies on the association between these polymorphisms and the development of gastric precancerous lesions (chronic atrophic gastritis and/or intestinal metaplasia) were identified by systematic review, and estimates were combined with fixed effects meta-analysis models. Results: All Mozambican participants presenting intestinal metaplasia were IL1RN*LL or IL1B-511 T. Within the Portuguese sample, the IL1RN (IL1RN*22 vs. LL: OR = 2.83, 95%CI: 1.15–6.96) and the IL1B-511 polymorphisms (IL1B-511 T vs. CC: OR = 1.86, 95%CI: 1.03–3.36) were associated with an increased risk of intestinal metaplasia. Fourteen studies were obtained through systematic review and included in the meta-analysis. Regarding IL1RN (9 studies), an increased risk of gastric precancerous lesions was observed for individuals having the L2 (OR = 1.32, 95%CI: 1.08–1.62, I2 = 36.3%) or the 22 (OR = 2.74, 95%CI: 1.89–3.96, I2 = 44.0%) compared with the LL genotype. No association was found for any of the IL1B511 (11 studies: CT vs. CC, OR = 0.98, 95%CI: 0.82–1.17, I2 = 59.3%; TT vs. CC, OR = 1.00, 95%CI: 0.79–1.25, I2 = 71.0%) or TNFA-308 genotypes (6 studies: GA vs. GG, OR = 0.90, 95%CI: 0.63–1.28, I2 = 13.3%; AA vs. GG, OR = 0.95, 95%CI: 0.38–2.38, I2 = 0.0%). Conclusions: An association was observed for IL1RN and gastric precancerous lesions, suggesting a role for this polymorphism in the early stages of gastric carcinogenesis.
100 POLYMORPHISMS IN ANTIOXIDANT DEFENSE GENES AND SUSCEPTIBILITY TO HEPATOCELLULAR CARCINOMA IN A MOROCCAN POPULATION S. Ezzikouri1, A. E. El Feydi2, R. Afifi2, L. El Kihal2, M. Benazzouz2, M. Hassar2, P. Pineau3, S. Benjelloun1 1Laboratoire des He´patites Virales, Institut Pasteur du Maroc, CASABLANCA, Maroc, 2Service de Me´decine C, CHU Ibn-Sina, RABAT, Maroc, 3Unite´ d’Organisation Nucle´aire et Oncoge´ne`se, INSERM U579, Institut Pasteur, PARIS, France Background: It is generally believed that the initiation of liver cancer is a consequence of cumulative genetic damage leading to genetic alterations and provoking uncontrolled cellular proliferation and/or aberrant programmed cell death, or apoptosis. Reactive oxygen species have been related to the etiology of cancer as they are known to be mitogenic and therefore capable of tumor promotion. Objectives: The aim of this study was to assess the role of common variation in 3 polymorphic genes (MnSOD Ala-9Val, GPX1 Pro198Leu and CAT -262 C [ T) coding for antioxidant defense enzymes in modulating individual susceptibility to hepatocellular carcinoma (HCC) using a case–control study (cases = 98 and controls = 222). Both cases and controls were recruited in Western-Central Morocco (Rabat and Casablanca) in a population of mixed berberic and arabic ethnicity. Methods: PCR-RFLP and sequencing methods were used to determine the genotype. Results: Odds ratios (OR) and 95% confidence intervals (95% CI) were estimated by logistic regression. Overall, there were no associations between genotypes GPX1 and HCC risk (OR = 1.16; 95% CI, 0.56–2.42; P = 0.685). The MnSOD Ala/Ala and CAT TT genotypes were more frequent in HCC group than in control group (P = 0.001 and P = 0.039, respectively). Further analyses stratified by gender or HCV infection revealed that men and HCV infected patients carrying CAT TT genotype had a higher risk to
123
48 develop HCC when compared with controls (OR = 15.94; 95% CI, 3.48–72.92; P \ 0.000001 and 12.01; 95% CI, 0.64–223.63, P = 0.023, respectively). When measuring the number of risk alleles of MnSOD, CAT, and GPX1 hypothetically related to a reduced protection against ROS, there was a significant relationship between hepatocellular carcinoma and carriage of risk alleles. Combined MnSOD Ala/Ala and GPx1 Leu/Leu had a synergistic effect on HCC risk, with an OR of 3.84 (P = 0.029). Furthermore an even more pronounced risk was observed when we combined MnSOD Ala/Ala and CAT TT (OR = 13.60, P = 0.023). Conclusions: It appears that variants in MnSOD, CAT, or GPX1 have an influence on HCC risk in this cohort. Furthermore, it is possible that cumulative defects in protection from oxidative stress may result in increased risk of the liver cancer in Moroccan population.
184 USING GENERALIZED WEIGHTS OF NEURAL NETWORKS TO IDENTIFY GENE-GENE INTERACTIONS F. Guenther, N. Wawro, K. Bammann Bremen Institute for Prevention Research and Social Medicine, University of Bremen, BREMEN, Germany Background: The investigation of complex diseases plays an important role in genetic epidemiology. Besides main effects, particularly the interplay of potential risk factors such as genetic causes and environmental factors is of interest. A major problem is the discrepancy between statistical and biological interaction, as there is no overall definition of interaction or epistasis for all disciplines. We study neural networks as a flexible statistical tool to model any functional relationship between covariates and response variables without the need for a priori specified structure. Objectives and Methods: Our aim is to resolve the discrepancy between what statistical tools are able to capture and what biologists have in mind. We apply neural networks to simulated data representing different types of biological gene––gene interaction and independence models in the situation of a case–control study with equal numbers of cases and controls as well as in the situation of a cohort study. In a first step, the generalized weights of the neural networks are studied to see how main effects can be quantified. As a second step, we focus on modeling interaction effects. The generalized weights of the neural networks represent the contribution of a covariate to the log odds, given all other covariates. In case of no epistasis, the generalized weights are expected to yield equal risk estimates for a gene, irrespective of the value of a second gene. Results and Conclusions: We present results of a simulation study for the case–control and the cohort design. We discuss the ability of the generalized weights to capture main effects and their use to resolve the discrepancy between the terms of biological and statistical interaction.
Afternoon Session 2—Epidemiology of elderly 66 MAJOR LIFE EVENTS AND RISK OF PARKINSON’S DISEASE N. H. Rod1, J. Hansen2, E. Schernhammer3, B. Ritz4 1Department of Social Medicine, Institute of Public Health, University of Copenhagen, COPENHAGEN, Denmark, 2Institute of Cancer Epidemiology, Danish Cancer Society, COPENHAGEN, Denmark, 3 Department of Epidemiology, Harvard School of Public Health, BOSTON, United States of America, 4Department of Epidemiology,
123
IEA-EEF European Congress of Epidemiology 2009 School of Public Health, UCLA, LOS ANGELES, United States of America Background: Major life events such as divorce, death of a spouse or a child, or long-term unemployment are stressful to most people and animal models have suggested a link between stress and onset of Parkinsonian symptoms. Objectives: In a large case–control study based on nationwide registries, we aim to address whether major life events are risk factors for Parkinson’s disease. Methods: Between 1986–2006, 13,695 Parkinson’s patients were identified in the Danish National Hospital Register. Each case was matched to five population controls. Information on major life events was ascertained from national registries. Results: Among men, number of life events was associated with risk of Parkinson’s disease in an inverse dose-response manner (P \ 0.0001). Compared to no events, three or more events were associated with a 42% lower risk of Parkinson’s disease (OR = 0.58; 95% CI: 0.34–0.99). Life events were not associated with Parkinson’s disease in women. In contrast, a higher risk of Parkinson’s disease was observed among women who had never been married (1.16; 1.04–1.29) and among men (1.47; 1.18–1.82) and women (1.30; 1.05–1.61) who have never been employees. Conclusions: The lower risk of Parkinson’s disease among men who had experienced life events was unexpected but might suggest a general ‘risk avoidance behavior’ in Parkinson’s patients.
256 DISPARITY IN LOWER EXTREMITY FUNCTION BY AREA AND EDUCATIONAL LEVEL K. Yoon, S. Cho, Graduate School of Public Health, Seoul National University, SEOUL, South Korea Background: Lower extremity function is known to be associated with subsequent disability, hospitalization, and mortality among older persons. The Short Physical Performance Battery (SPPB) is an objective assessment tool for evaluating lower extremity functioning. Objectives: We investigated the association of demographic and behavioral characteristics with lower extremity function in Korean elderly persons. Methods: This study included 217 men and 286 women aged 47–95 (mean 63.5) years in Korea, as a part of KLoSA Biomarker Pilot study. Lower extremity function was measured by SPPB tests. The battery included balance tests with various foot position such as side-by-side, semi-tandem, and tandem for thirty seconds, gait speed for 4 m, and timed rise from chair repeated five times. Each test scores from 0 to 4 with performance time. The complete summary score was 12 points. A score below 10 was regarded as poor performance. Portable bioimpedence device(Omran HBF359) was used for measuring skeletal muscle percentage of body weight. Age, sex, residential area(urban vs. rural), education, and behavioral factors were included in the analyses. Education was dichotomized as low (\10 years) and high (C10 years). Logistic regression was used for data analysis. Results: In univariate analyses, SPPB score was lower for older age, women, low education, rural residence, and higher muscle percentage. Occupation and behavioral factors (smoking, alcohol, exercise) were not associated. In multiple logistic regression adjusting for age and sex, poor performance was significantly associated with rural residence (odds ratio [OR] 2.0, P = 0.01) and lower education (OR 2.1, P = 0.05). After adjusting for these factors, skeletal muscle percentage lost significance for its association with SPPB score. Conclusion: SPPB performance was significantly poorer for elderly with low education and rural residence. Low extremity function may form an intermediate link between low socioeconomic position and higer disability and mortaliy, suggesting a potential point of intervention.
IEA-EEF European Congress of Epidemiology 2009
138 SURVIVAL OF ELDERLY FEMALE BREAST CANCER PATIENTS IN THE WESTERN PART OF THE NETHERLANDS E. Bastiaannet1, J. Portielje2, L. G. M. van der Geest3, A. J. M. de Craen1, G. J. Liefers1, C. J. H. van de Velde1, R. G. J. Westendorp1 1Leiden University Medical Centre, LEIDEN, The Netherlands, 2Haga Hospital, THE HAGUE, The Netherlands, 3 Comprehensive Cancer Centre West, LEIDEN, The Netherlands Background: A large proportion of breast cancer patients are diagnosed above 65 years and incidence in the elderly is expected to increase in the coming years. Specific clinical guidelines for elderly are not available due to lack of data, uncertainties concerning toxicity and competing death risks. Objectives: Aim of this study was to assess time trends in Relative Survival (RS) and Standardized Mortality Ratios (SMR) of (elderly) breast cancer patients in the western part of the Netherlands 1990–2005. Methods: Female breast cancer patients were selected from the regional cancer registry (Comprehensive Cancer Centre West). RS (Hakulinen method) was calculated according to year and age; Relative Excess Risks (RER) for year were calculated and corrected in multivariate model (Poisson distribution) for stage, histology, grade and treatment. SMR were calculated as ratio of observed and expected deaths based on corresponding general population (national life tables). Results: Overall, 18965 patients were included. Contrary to younger patients, RS did not increase over time for elderly; corrected for potential confounders RS did decrease for 86–90 (RER 1.10; P = 0.04) and [90 (RER 1.23; P = 0.04) years. Stage at diagnosis and treatment changed over time for patients 65–75; however stage did not change for patients C75 years, while treatment did change for patients 81–85 and 86–90 years. Patients aged 86–90 years received less surgery and more hormonal treatment (monotherapy). SMR decreased almost to 1.0 and differences between stages became smaller as age increased. Conclusions: Survival for the elderly did not improve in the western part of the Netherlands. Standardized mortality ratios according to age and stage are close to one suggesting that excess mortality due to breast cancer is low and gain of earlier diagnosis is not as large. Recently the FOCUS study started in the Netherlands to develop evidence-based clinical guidelines for elderly patients with breast cancer.
277 PATTERNS OF MORTALITY AMONG OLDER CRACOW CITIZEN (22-YEARS OBSERVATION CRACOW STUDY) A. Pac1, B. Tobiasz-Adamczyk2, P. Brzyski2 1Chair of Epidemiology and Preventive Medicine, Jagiellonian University Medical College, CRACOW, Poland 2Department of Medical Sociology, Chair of Epidemiology and Preventive Medicine, Jagiellonian University Medical College, CRACOW, Poland Objectives: The aim of the study was to assess mortality patterns of all causes death in relation to four types of predictors: sociodemographic, everyday activities, health status and health-related attitudes. Methods: A base-line study was performed over 1986/87 in a randomised sample of 2605 individuals (34.6% males and 65.7% females), aged 65 years and over, residents of Krakow city centre. Face to face interviews covered questions on demographic data, information on self-reported chronic diseases, current and previous lifestyle, occupational activity, marital status, living conditions, everyday activities and leisure time activities, self-rated health status, and attitudes towards positive health behaviors. The vital status of all individuals under study was ascertained by
49 monitoring city vital records and general mortality was analyzed. Statistical analysis was performed using Cox proportional hazard model. Results: During the 22 years of observation 81.1% of study population died. The highest impact on mortality had age and education level for males and age for females. Analysis of health status showed that diabetes (HR = 1.40; 95%CI: 1.07–1.84), asthma (HR = 1.30; 95%CI: 1.04–1.63), and heart attack (HR = 1.28; 95%CI: 1.04–1.58) in males as well as diabetes (HR = 1.76; 95%CI: 1.45–2.12), heart attack (HR = 1.27; 95%CI: 1.05–1.53) and CHD (HR = 1.13; 95%CI: 1.02–1.27) in females were factors increasing general mortality. We have also found that for both, men and women, doing everyday household activities (HR = 0.78— males; HR = 0.65—females) and working in the garden were protective factors (HR = 0.75—males; HR = 0.78—females). Conclusions: The higher level of education seems to be protective effect only in males. The higher mortality risk was assessed for people who suffered from diabetes, coronary heart disease, asthma or heart attack. People who are more involved in household activities are suppose to live longer.
Afternoon Session 3—Health services research 224 DIAGNOSIS AND REFERRAL OF RHEUMATOID ARTHRITIS BY PRIMARY CARE PHYSICIAN: RESULTS OF A PILOT STUDY ON THE TOWN OF PISA, ITALY A. Della Rossa1, R. Neri1, R. Talarico1, M. Doveri1, A. Consensi1, S. Salvadori2, V. Lorenzoni2, G. Turchetti3, S. Bellelli3, M. Cazzato1, L. Bazzichi1, P. Monicelli4, S. Moscardini4, S. Bombardieri1 1 Rheumatology Unit, Department of Internal Medicine, University of Pisa, PISA, Italy, 2CNR, Institute of Clinical Physiology, PISA, Italy, 3 Scuola Superiore S. Anna, PISA, Italy, 4General practitioner, PISA, Italy Background: Rheumatoid arthritis (RA) is a chronic inflammatory disease affecting diarthroidal joints. In Italy prevalence rates vary from 3.3/1000 to 4.6/1000 and very few studies on the pattern of diagnosis and referral of RA by primary care physicians has been performed. Objective: The aim of the study was to assess the prevalence of RA in Tuscany, Italy, to evaluate the reliability of the prevalence estimated by primary care physicians, and the economic impact of the disease. This multi-centre study involved the provinces of Pisa, Siena and Florence in Tuscany Region in the years 2006– 2007. Final results regarding the area of Pisa are presented. Methods: The RA subjects [18 years followed by each GP constituted the population studied. A stratified random sample of 19 general practitioner (GP), corresponding to a stratified random cluster sample of 26709 subjects (around 10% of target population), was drawn from the Tuscany registry of GPs: stratification variables were length of service, use of personal computer, geographical area. GPs was asked to fill out a questionnaire regarding RA patients and to send them to the tertiary Rheumatologic centre to confirm/discard the diagnosis, to collect the clinical and epidemiological data and data for the costs estimation. Results: The estimated prevalence of RA was 5.1/1000 (C.I 95% 4.1/1000–6.1/1000). The mean annual cost per patient with RA was estimated in 5530 Euro with high variability dependent on the degree of disability. More than 90% of the cost per patient is due to the medical and non medical direct components of costs. Conclusion: The prevalence of RA in Tuscany is comparable with similar prevalence studies in Italy. The annual cost per patient with RA is variable and strictly dependent on the level of disability. More than 90% of the overall cost is due to the direct burden of costs.
123
50
87 IDENTIFICATION OF VARIABLES OF ADVERSE EVOLUTION AT THE HOSPITAL ER IN PATIENTS WITH COPD EXACERBATION J. M. Quintana1, S. Garcı´a1, C. Esteban2, A. Bilbao3, J. L. Lobo4, N. Gonza´lez1, I. Lafuente1, F. Aizpuru5 1Unidad de Investigacio´n, Hospital Galdakao-Usansolo- CIBERESP, GALDAKAO, Spain, 2 Servicio de Respiratorio, Hospital Galdakao-Usansolo, GALDAKAO, Spain, 3Fundacio´n Vasca de Innovacio´n e Investigacio´n Sanitarias (BIOEF)–CIBERESP, SONDIKA, Spain, 4 Servicio de Respiratorio, Hospital de Txagorritxu, VITORIA, Spain, 5 Unidad de Investigacio´n, Hospital de Txagorritxu- CIBERESP, VITORIA, Spain Objectives: To identify variables routinely gathered at the hospital Emergency Room (ER) that may help to predict adverse evolution during the two first weeks in patients attended at the ER by a COPD exacerbation. Methods: Pilot data from a prospective cohort study of patients with COPD exacerbation from 2 hospitals, starting at the ER with follow-up during the hospital admission or discharge at home up to two weeks after presentation at the ER. At the ER possible clinical (respiratory frequency, frequency cardiac, presence of dyspnoea, instability at admission, oxygen saturation, age) and laboratory parameters (gasometry) that may predict adverse evolution were gathered at the time the decision of admission or discharge was made. Main outcome was composed by mortality, major complications, admission at ICU, or need for mechanical ventilation during the admission or the first week after discharge. Univariate analysis were made to determine what ER variables were related to the outcome of interest. Results: Data was available from 254 patients, of whom 48 (18.9%) presented adverse evolution. The univariate analysis allows to identify altered pH (P B 0.0001) hypercapnia (P = 0.013), and instability when arriving at urgencies (P = 0.016) (includes presence of accessory muscles use, cardiac arrest or hemodynamic instability) as predicting variables of adverse evolution. If we included to the previous sample those discharge from the ER, to altered pH (P B 0.0001) and hypercapnia (P = 0.0001) was added as predicting variable altered respiratory frequency (P = 0.03) while dyspnoea at the moment of the decision making at ER was in the limit of the statistical significancy (P = 0.06). Conclusions: This preliminary analysis allows identifying some possible predicting variables of adverse evolution in patients with COPD exacerbation. It could help physicians at the ER in the determining the severity of a patient. These results have to be corroborated with a greater sample and validated in other samples.
225 PALLIATIVE AND SUPPORT HEALTH CARE AT HOME IN SPAIN. A HEALTH SENTINEL NETWORK STUDY T. Vega1, J. E. Lozano1, C. Quin˜ones2, Y. Anes3, C. Gomez3, ´ lamo1, S. Ferna´ndez1, M. Perucha2, M. Gil1, M. Ramos3, R. A on behalf of the RECENT group 1Consejerı´a de Sanidad de Castilla y Leo´n, VALLADOLID, Spain, 2Consejerı´a de Salud de La Rioja, ˜ O, Spain 3Consejerı´a de Sanidad y Dependencia de LOGRON Extremadura, ME´RIDA, Spain Background: Palliative and support health care are, nowadays, a major clinic and ethical challenge in developed countries and a real burden for the sustainability of good-quality medical and social attention. Increasing life expectancy, social changes in family role
123
IEA-EEF European Congress of Epidemiology 2009 and the accessibility to new technologies and therapies have led to this public health problem. Objective: The objective of this study is to describe the magnitude and types of palliative and support care at home provided by the primary health care system in Spain. Methods: Five Spanish Health Sentinel Networks, with 218 GPs, covering a population of 282,216 people recorded during three months in 2008 every visit to patients in terminal phase or with a very advanced chronic disease to provide them palliative o support care at home. Results: Global frequency (prevalence) of patients receiving palliative or support care at home was estimated in 4.2 per 1,000 people aged 15 and more. This rate is significantly higher in women than in men (5.5 vs. 2.9 per 1,000). By age groups, the prevalence increases significantly from 75 to 84 with 16.1 per 1,000 and 85 and more with 49.7 per 1,000 people. The majority of the cases were patients with severe functional incapacity derived from a chronic disease but not in end of life care phase (87.5%). Terminal disease patients with a life expectancy less than 6 months represented only the 9.8%; a 2.7% share these two conditions. Conclusions: Palliative care is not an exclusive characteristic of patients with quickly progressive diseases such as cancer or neurodegenerative diseases. Most of them suffer from chronic conditions which in elderly people represent a very important public health problem. Women have twice more risk to arrive to this stage, and complementary analysis should be done to study the patterns and determinants to establish preventive and control policies.
80 THE ACCURACY OF CAUSE-OF-DEATH CODING IN THE NETHERLANDS: A STUDY INTO RELIABILITY P. Harteloh, K. de Bruin, J. Kardaun Statistics Netherlands, THE HAGUE, The Netherlands Background: Cause-of-death statistics are an important source of information for epidemiological research or policy decisions. The reliability or consistency of these statistics is important for interpreting trends in time or differences between populations. Attribution of observed differences to determinants of health should not be biased by variations in the coding or selection of the underlying cause of death. Objectives: We measured the reliability of cause-of-death statistics in order to improve internal consistency and external interpretations. Methods: A sample of 8,215 death certificates from the month of May 2005 was recoded (manually) by the four coders of Statistics Netherlands. Reliability was measured by calculating agreement between (intercoder) or within (intracoder) coders and by calculating the chance of reproducing the original code for the underlying cause of death. Results: The agreement between coders in coding and selecting the underlying cause-of-death was 78% (intercoder variation: 22%). The (mean) intracoder agreement was 89% (intracoder variation: 11%). The chance of reproducing the original code was 87%. Agreement among coders was associated with the specificity of the ICD-10 code (chapter, three digits, four digits), the nature of the disease (ICD-10 chapter), the age of the deceased and the number of diseases reported on the death certificate. For important causes of death such as cancers and acute myocardial infarction the reliability of cause-of-death statistics appeared to be high ([90%). For chronic diseases such as diabetes and renal insufficiency the reliability was low (\60%). Conclusions: When interpreting cause-of-death statistics reliability should be taken into account. The consistency of mortality statistics could be improved by (automatic application of) coding or selection rules, the interpretation by analysis based not only on underlying causes of death, but on comorbidity as well (multiple cause coding).
IEA-EEF European Congress of Epidemiology 2009
Afternoon Session 4—Occupational epidemiology II 89 SENSITIVITY AND SPECIFICITY OF WORK ABILITY INDEX FOR DETERMINATION OF SICKNESS ABSENCE AND DISABILITY S. M. Alavinia1, M. Omidvar1, A. Burdorf2 1North Khorasan University of Medical Sciences, Deputy of Education Research Administration, BOJNORD, Iran, 2Department of Public Health, Erasmus MC, University Medical Center Rotterdam, ROTTERDAM, The Netherlands Background: In order to prevent workers from quitting the workforce due to (work-related) disability, the concept of measuring work ability has been developed as a valuable tool to tailor interventions at individual level. For assessing workers’ work ability. Finnish researchers have constructed the so-called work ability index (WAI), that combines subjective experiences of one’s ability to cope with physical and mental requirements at work (performance at work) with information on diseases and consequent functional limitations, sick leave, and mental resources at work (enjoying daily tasks, active and vital life, optimism about the future). Objective: To determine the predictive value of WAI for sickness absence and disability and also determine its sensitivity and specificity for these two important events in occupational health. Method: Sickness absence data were collected from then 5,667 male Dutch workers in the construction industry in the Netherlands, who participated in the voluntary periodic medical examination in 2005 and had complete data and registered sickness absence. Poisson regression analysis was used to describe the effects of covariates on the sickness absence. A Cox regression analysis was performed to study the relation between work ability and disability. A Receiving Operator Characteristic (ROC) analysis was done to evaluate the area under the curve and sensitivity and specificity for both sickness absence longer than 3 months and disability. Results: The incidence of a sickness absence period was 0.31 per person year. ROC showed a low sensitivity (6%) and a low positive predictive value (9%) of a poor work ability for predicting sick leave lasting at least 3 months. The ROC curve had an area under the curve of 0.65 (P \ 0.001). Conclusion: Although the WAI is a good predictor for work related disability, it should be considered that especially the low sensitivity does not make the WAI a suitable test for identifying workers at risk for a longer duration of sickness absence.
130 COMPARISON OF CANCER RISK IN WORKERS WITH SILICOSIS AND UNEXPOSED POPULATION IN THE CZECH REPUBLIC A. Splichalova1,2, H. Tomaskova1,2, Z. Jirak2, H. Lehocka1,2, P. Urban3, J. Holub4, D. Pelclova5, E. Hrncir6, A. Eichlerova7, M. Nakladalova8, M. Bartnicka9, A. Hejlek10, L. Adamkova11, D. Cermakova12 1Institute of Public Health, OSTRAVA, Czech Republic, 2University of Ostrava, Faculty of Health Studies, OSTRAVA, Czech Republic, 3National Institute of Public Health, PRAGUE, Czech Republic, 4Institute of Health Information and Statistics of the Czech Republic, PRAGUE, Czech Republic, 5 Department of Occupational Medicine, General University Hospital, PRAGUE, Czech Republic, 6Department of Occupational Medicine, University Hospital Kralovske Vinohrady, PRAGUE, Czech Republic, 7Department of Occupational Medicine, Regional Hospital, PARDUBICE, Czech Republic, 8Department of Occupational Medicine, University Hospital, OLOMOUC, Czech Republic, 9 Department of Occupational Medicine, Hospital Podlesi, TRINEC, Czech Republic, 10Department of Occupational Medicine, Hospital
51 Ceske Budejovice, CESKE BUDEJOVICE, Czech Republic, 11 Occupational medicine physician, ZLIN, Czech Republic, 12 Department of Occupational Medicine, Regional Hospital, LIBEREC, Czech Republic Background: In 1997 the International Agency for Research on Cancer classified dust containing crystalline silica to group 1, as a human carcinogen. Many studies identified increased lung cancer risk in workers exposed to crystalline silica dust. Objectives: The aim of this study was to compare cancer risk between workers with diagnosed silicosis and unexposed general population in the Czech Republic. Methods: The study sample included 797 workers (men only) exposed to crystalline silica dust who were registered with silicosis in the National Register of Occupational Diseases in 1992–2001. These workers came from metallurgic, ore-mining, stone-cutting, ceramic and porcelain industry. The data on individual and occupational history of workers were linked with the data from the National Cancer Register and the National Population Register. Risks for cancer (lung, stomach, colon, kidney and bladder) were calculated as a standardized incidence ratios (SIRs) with 95% confidence interval (CI) based on the data on cancer incidence from 1992 to 2006. Results: Average age of study sample in time of diagnosed silicosis was 60.4, average exposure was 24.2 years. In period 1992–2006, 158 workers (19.8%) sicken of cancer. Significantly higher cancer risk was confirmed only for lung cancer (N = 61, SIR = 2.81, 95% CI: 2.2-3.8). The highest lung cancer risk was identified in ore-miners and breakers SIR = 3.86 (95% CI: 2.5–6.0), metal workers SIR = 3.79 (95% CI: 2.2–6.5) and metallurgist-founders SIR = 2.77 (95% CI: 1.6–4.8). Positive association was confirmed between smoking and incidence of lung cancer in workers with silicosis (P \ 0.01). No statistically significant differences were found for cancer incidence of stomach, colon, kidney and urinary bladder. Conclusions: The results demonstrate significantly higher lung cancer risk in workers having silicosis comparing with the Czech general population.
134 WORK-RELATED RESPIRATORY SYMPTOMS IN NORWEGIAN SALMON INDUSTRY WORKER POPULATION O. Shiryaeva1,2, L. Aasmoe1,3, B. Bang1,3 1Department of Occupational and Environmental Medicine, University Hospital North Norway, TROMSØ, Norway, 2Institute of Community Medicine, University of Tromsø, TROMSØ, Norway, 3Institute of Pharmacology, University of Tromsø, TROMSØ, Norway Background: Different types of environmental exposures in workplaces may cause airway symptoms. Reported symptoms relate to both upper and lower airways and include chest tightness, dyspnoea, cough and symptoms from nose. The salmon industry is of high economic importance in Norway, strongly contributing to employment opportunities. Processing of fish creates respirable particles carrying allergens, endotoxins, proteins. Inhalation of bioaerosol components is a suspected risk factor of respiratory symptoms and allergy. Objectives: Present study was set up to determine the prevalence of work-related airway symptoms, asthma and allergy; to explore the relationship between exposure to different biologic agents in salmon industry work places and respiratory effects; to compare the health data between exposed and unexposed study population. Methods: A cross-sectional survey was conducted on 251 participants, 98 employees of five salmon factories and 153 controls. A self-administered BMRC questionnaire with additional questions on work-related symptoms was used. Pulmonary function and nasal patency tests were performed. Exhaled NO, total and specific IgE to salmon and inhalant allergens were measured. Aerosolized constituents were collected in breathing zones by means of personal portable air pumps. The concentration of airborne particles
123
52 were quantified and correlated with health outcomes. Results: Prevalence of doctor-diagnosed asthma was 5.1%, allergy 8.1% and atopy 19.4% in exposed group. Prevalence of work-related airway symptoms was 26.1% for production workers and 9.1% for controls. Spirometric airflow limitation was found to be significantly higher in exposed group (16.2% versus 4.1% in controls). None from unexposed group has IgE to salmon. Conclusion: Our study demonstrates new insight into the work—related health outcomes in seafood industry and suggests that exposure to aerosolized allergens, endotoxins and proteins affects respiratory organs. The mechanisms involved can be immunological and non-immunological in nature. This knowledge gives an opportunity to prevent airway diseases and allergy that result from environmental exposure.
284 OCCUPATIONAL FATALITIES IN THE ARKHANGELSK REGION, 1996–2007 Zh L. Varakina1,2, A. M. Vyazmin2, A. L. Sannikov2, A. M. Grjibovski1,3 1International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia, 2Department of Public Health, Health Care and Social Work, Northern State Medical University, ARKHANGELSK, Russia, 3Norwegian Institute of Public Health, OSLO, Norway Background: Occupational fatalities are among the indicators of occupational safety. More than one third of a million of fatal occupational accidents occur every year worldwide. External causes of
123
IEA-EEF European Congress of Epidemiology 2009 death are the second main killers in Russia with occupational fatalities being important contributors. Northwest region is among the areas with the highest burden of occupational accidents in Russia. However, little is known about the actual incidence of occupational fatalities since no systematic reporting system has been established. Objectives: To describe occurrence of occupational fatalities in the Arkhangelsk region in 1996–2007. Methods: Data on all fatal occupational accidents reported in the Arkhangelsk region in 1996–2007 were obtained from the State Labor Inspection. The incidence of occupational fatal accidents was calculated by age, gender and occupations per 100 000 employed. Data on the number of employed were collected at the regional federal state statistics service. Results: Altogether, there were 734 occupational fatalities during the study period, 690 (94%) of them were men. The incidence of fatal occupational accidents decreased from 18.6 (95%CI: 14.1–23.1) in 1996 to 11.7 (95%CI: 7.9–15.4) in 2007 among men and from 1.3 (95%CI: 0.03–2.5) in 1996 to 0.3 (95%CI: 0.2–0.9) in 2006 among women. No fatal accidents occurred in women in 2007. Thirty eight percent of all fatal accidents occurred in age-group 40–49 years. Most occupational fatalities occurred in agriculture, hunting and forestry (33.5%), 22.8%—in manufacturing and mining, 12.3%—in transport and communication. The incidence of occupational fatalities in agriculture, hunting and forestry decreased from 43.9 (95%CI 28.7–59.1) in 1996 to 24.1 (95%CI 10.4–40.9) in 2007. Conclusions: The incidence of occupational fatalities in the region considerably decreased during the study period. However, it is still several times higher than in many European countries, particularly among men employed in agriculture, hunting and forestry.
IEA-EEF European Congress of Epidemiology 2009
Saturday, 29 August 2009 Parallel early morning oral sessions: 10.00–11.00 Early Morning Session 1—Cancer epidemiology III 133 PNEUMOCONIOSIS AND CARCINOGENIC RISK IN THE CZECH BLACK-COAL MINES H. Tomaskova1,2, Z. Jirak1, A. Splichalova1,2, P. Urban3, J. Holub4, R. Gromnica5, Z. Hajdukova6, I. Landecka7, V. Machartova8, E. Korolova9 1University of Ostrava, Medico-Social Faculty, OSTRAVA, Czech Republic, 2Institute of Public Health in Ostrava, OSTRAVA, Czech Republic, 3National Institute of Public Health, PRAHA, Czech Republic, 4Institute of Health Information and Statistics of the Czech Republic, PRAQUE, Czech Republic, 5Miner’s Hospital, OSTRAVA, Czech Republic, 6University Hospital with Policlinic, OSTRAVA, Czech Republic, 7Miner’s Hospital, KARVINA, Czech Republic, 8University Hospital, PLZEN, Czech Republic, 9University Hospital, HRADEC KRALOVE, Czech Republic Background: The possibility of the increased risk of lung cancer in black-coal miners cannot be explicitly, according to the present knowledge, either confirmed or eliminated. Objectives: The goal of this study was to assess carcinogenic risk between black-coal miners with pneumoconiosis and general population in the Czech Republic. Methods: The sample consisted of 2.504 black-coal miners with pneumoconiosis. These workers were registered in the National Register of Occupational Diseases from 1992 to 2001. The data on individual and occupational history were linked with the data from the National Cancer Register and the National Population Register. The relative risk of cancer (Cancer of lung, stomach, colon, kidney, bladder) was calculated as SIR (Standardized Incidence Ratio) with 95% CI over the period 1992 to 2006 between miners and general Czech population. Stata and OpenEpi software were used for the data analysis. Results: Coal workers’ pneumoconiosis (PN) was diagnosed in average age of 48.8 years (SD = 12.5). Average time of exposure was 21.1 years (SD = 7.9). In the sample 278 newly registered cases of cancer and 493 deaths were found in the period 1992–2006. Risk of lung cancer was statistically significant (N = 91, SIR = 2.25, 95% CI 1.82–2.75). Risk of lung cancer depended on the degree of the seriousness of PN (SIR = 1.05–5.49). The miners with lung cancer had significantly longer occupation exposure then miners without lung cancer. Data on smoking habits were available for 73% of workers. Of those 65.6% were smokers or ex-smokers. The Risk of other specific cancers was not statistically significantly higher (Cancer of stomach N = 9, SIR = 1.10; colon N = 18, SIR = 0.93; kidney N = 13, SIR = 0.97; bladder N = 15, SIR = 1.18). Conclusions: The risk of lung cancer in miners with pneumoconiosis was significantly higher than the risk in the general population.
197 TRENDS IN THE MORTALITY RATES OF MALIGNANT LUNG CANCER IN FEMALE INHABITANTS OF UPPER SILESIA, AN INDUSTRIAL AREA IN POLAND M. Juszko-Piekut1, Z. Kolosza2, B. F. P. Zemla2, A. Mozdzierz1, J. Stojko1 1Medical University of Silesia, Department of Hygiene, Bioanalysis and Environmental Studies, School of Pharmacy, KATOWICE, Poland, 2 Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Gliwice Branch, Department of Cancer Epidemiology, GLIWICE, Poland
53 The results of regression analysis of the years 1970–1993 show that lung cancer mortality rates in female inhabitants of the Upper Silesia increased annually by 3.6%. Aim: The aim of our study was to analyze the trends of lung cancer mortality rates in female inhabitants of Upper Silesia, an industrial area in Poland, over the last three decades. Material and methods: Age-specific and age-standardized (overall and truncated) rates were calculated by direct method using the world standard population. The joinpoint regression analysis was used to quantify changes in lung cancer mortality rates from 1975 to 2005. Results: The increasing trend of mortality rates of lung cancer in women, which can be presently observed in Poland, was also recorded in the Upper Silesia. The trend direction changed 3 times, namely in the years 1975–1991 the mortality rates increased by 4% per year, in 1991–1994 the trends stabilized, whereas in 1994–2005 a further annual increase in mortality rates by 4% was observed. Corresponding unfavorable trends are observed in 5-year age groups between ages 45 and 84 years. Only in young women (age 40–44) since 1988 till the end of the analyzed period, the trend stabilized. The cohort analysis confirmed the obtained results. Since the beginning of the previous century, the mortality rates of lung cancer in women have increased. The first signs of minimizing the risk are observed only in adults born after 1945. Conclusion: In the Upper Silesia, and in Poland, there are unfavorable trends of increased mortality rates of lung cancer in women, which might be related to tobacco smoking. However, the mortality trends in young female inhabitants of the Upper Silesia are less optimistic than the trends for young women in Poland. Nevertheless, the recorded stabilization indicates that lung cancer mortality rates will decrease in future.
276 CIVIL STATUS, EDUCATION AND INCOME IN RELATION TO STAGE AND SURVIVAL OF RECTAL CANCER L. Olsson1, F. Granstro¨m2 1Department of Molecular Medicine and Surgery, Karolinska Institutet, STOCKHOLM, Sweden, 2Centre for Clinical Research, So¨rmland County Council/Uppsala University, UPPSALA, Sweden Background: Stage distribution of cancer is associated with socioeconomic inequality but the relation to overall survival of rectal cancer is unclear. Better treatment stresses the importance of the issue. Objectives: Analyse overall survival of rectal cancer in relation to stage and socioeconomic variables. Methods: All patients in the Swedish Rectal Cancer Registry 1995–2005 (16 713) were included. Individual information on civil status and income was available for almost all, as well as education for patients \75 years. Survival (censored December 2008) was adjusted for age, sex and stage in a Cox regression model. Results:Rectal cancer stage I was most frequent in married (21.9%), university educated (24.8%) and Q4 income patients (23.1%). Overall 5-year survival was 46%; 51 vs 42% in married–unmarried and 59 vs 40% in income Q4-Q1 patients. For patients \ 75 years, unmarried, divorced and widowed men had an increased risk of dying (HR 1.3 (95% CI 1.21–1.47), HR 1.2 (95% CI 1.12–1.36) and HR 1.5 (95% CI 1.28–1.70)) compared to married men whereas in females only widows had an increased risk (HR 1.3 (95% CI 1.13–1.46)). Patients of income Q1, Q2, and Q3 had an increased risk of dying (HR 1.44 (95% CI 1.32–1.58), HR 1.36 (95% CI 1.25– 1.48) and HR 1.23 (95% CI 1.15–1.30)) compared to Q4 patients. Education had no independent impact on survival. For patients [ 75, the result was similar. 5-year survival increased from 45.2 to 47.4% during the second part of the study, but HR for socioeconomic variables were similar during both sub-periods. Conclusion: In particular low income was associated with poorer overall survival of rectal
123
54 cancer in a country with universal health care. To what extent differences in treatment contribute to this inequality need to be explored.
Early Morning Session 2—Social epidemiology III 217 HOW SOCIOECONOMIC STATUS INFLUENCES ROAD TRAFFIC INJURIES AND HOME INJURIES IN ROME L. Camilloni, F. Chini, S. Farchi, P. Giorgi Rossi, P. Borgia Agency for Public Health Lazio Region, ROME, Italy Background: Road Traffic Injuries (RTI) and Home Injuries (HI) are a relevant public health problem, especially among people living in deprived areas. Objective: To explore the relationship between morbidity, hospitalization, mortality from RTI and HI and socioeconomic status (SES) of the area of residence. Methods: Source: RTI and HI surveillance based on the integration between the Emergency Information System, the Hospital Information System and the Mortality Registry of Lazio region. The SES index (5 levels:1 ‘most privileged areas’; 5 ‘most deprived areas’) was derived using census data. For each injured subject the SES index of its census tract of residence was obtained. Population: Emergency Department (ED) admissions (year 2005) of residents in Rome, Italy. Analyses: For each level of SES, we computed rates for ED visits, hospitalisation and mortality, adjusted by sex and age. IRR and 95% confidence intervals have been estimated using Poisson Regression. A focus was made for HI occurred in children aged 0-14 and for RTI in adult aged 15–34. Results: Out of 127,129 ED visits, 9576 (7.5%) were followed by a hospitalization and 594 (0.5%) died within 30 days. SES was estimated for 109,993 cases, 87% of the total. The rates of RTI and HI visits were higher among the most deprived level of SES (IRR 1.27 CI95%:1.24–1.30; IRR 1.33 CI95%:1.29–1.37 respectively) compared to the most privileged ones; similar result was found for hospitalization (IRR 1.19 CI95%:1.08– 1.32; IRR 1.11 CI95%:1.01–1.22). A strong relation was found between RTI mortality rates and poor level of SES, while there was no association between HI mortality and SES. These findings were more marked for 0–4 year old HI injured and in adult 15–34 year old involved in RTI. Conclusion: RTI and HI incidence were associated to sociodemographic factors. This finding has implications for targeting injury prevention interventions and resources.
191 FAMILY WEALTH MODIFIES THE RELATIONSHIP BETWEEN SELF-RATED HEALTH AND CHRONIC DISEASES STATUS IN EARLY ADOLESCENCE J. Mazur Institute of Mother and Child, WARSAW, Poland Background: Self-rated health (SRH) is a commonly used single-item measure, which is strongly associated with physical and psychosocial aspects of health. Only few studies have investigated the relationship between SRH and other health indicators reported by adolescents living in various socioeconomic conditions. Objective: To investigate an interaction between chronic diseases and family wealth as determinants of SRH. Methods: A standard international questionnaire was administered to 5489 students aged 10.5–16.5 years, participating in the Health Behaviour in School-aged Children (HBSC) study conducted in Poland in 2006. To assess chronic disease status (CDS) students were asked whether they suffer from any long-standing condition diagnosed by a doctor. Family wealth was measured by the three-level Family Affluence Scale (FAS). A series of multivariate
123
IEA-EEF European Congress of Epidemiology 2009 logistic regression models adjusted for age and gender was estimated using SRH as dependent variable (very good/good health vs. fair/ poor). Validity of each model was assessed by the Hosmer-Lemeshow Chi-square analysis. Results: In the total sample, 14.2% of students reported fair or poor health. In the group with chronic diseases the corresponding number was 38.9% and 11.0% in the group without such conditions. P value for Hosmer–Lemeshow test was equal to 0.784 in the model including terms for CDS and interaction between FAS and CDS, by comparison to 0.367 in the model including only main effects for CDS and FAS, and 0.050 when both FAS, CDS and FASxCDS were considered. The latter model had the poorest calibration. Stratified analysis showed stronger association of CDS with SRH in poor families (OR = 5.16; 95% CI: 3.76–7.08) by comparison to wealthy families (OR = 3.97; 95% CI: 2.69–5.84). Conclusion: In early adolescence, socio-economic factors may modify the relationship between SRH and more objective health assessment and the impact of demographic factors.
146 RELATIONSHIP BETWEEN LOW HEALTH-RELATED QUALITY OF LIFE AND UNEMPLOYMENT P. Zagozdzon, J. Ejsmont Department of Hygiene and Epidemiology, Medical University of Gdansk, GDANSK, Poland Background: It is well known that unemployment can cause health problems to the exposed individual. Multiple epidemiological reports showed that inequity in health status is partly related to unemployment. The purpose of this population-based study was to describe health related quality of life among unemployed people compared to those who are employed, and to analyse variables related to quality of life for the respective groups. Methods: The sample consisted of 51 unemployed individuals and 51 sex, age and neighborhood-matched working individuals as a reference group. Unemployed individuals were randomly selected from unemployed population living in Gdansk City and Gdansk County who were registered as unemployed in Employment Office from January 1st 2007 to June 30th 2007. They all received a questionnaire about income, personal characteristics, civil status, educational level and self-reported quality of life based on SF36 form. Univariate and multivariate analysis was carried out by means of a logistic regression model. Results: The unemployed reported lower levels of quality of life compared to the reference group. In multivariate analysis unemployment showed and odds ratio 3.12 95% confidence interval (CI) 1.03–9.39 for low physical health related quality of life as compared to employment. The adjusted odds ratio for a poorer mental health quality of life associated with unemployment was 4.01 95% CI 1.26–12.76. Conclusion: There was worse health related quality of life among sample of unemployed individuals comparing to matched working people. More research is needed to understand the specific role of unemployment among different socioeconomic determinants of health related quality of life in Poland.
Early Morning Session 3—Clinical epidemiology III 90 VALIDITY OF A SIMPLE SEVERITY SCALE USED AT THE ER TO PREDICT ADVERSE EVENTS ON PATIENTS WITH COPD EXACERBATION J. M. Quintana1, S. Garcı´a1, C. Esteban2, A. Bilbao3, J. A. Blasco4, M. Bare5, S. Vidal6, E. Perea7, J. L. Lobo8 1Unidad de Investigacio´n, Hospital Galdakao-Usansolo- CIBERESP, GALDAKAO, Spain,
IEA-EEF European Congress of Epidemiology 2009 2
Servicio de Respiratorio, Hospital Galdakao-Usansolo, GALDAKAO, Spain, 3Fundacio´n Vasca de Innovacio´n e Investigacio´n Sanitarias (BIOEF) –CIBERESP, SONDIKA, Spain, 4 Agencia Lain Entralgo, MADRID, Spain, 5Unidad de Investigacio´n, Corporacio Parc Tauli, SABADELL, Spain, 6 Unidad de Investigacio´n, Hospital de Valme, SEVILLA, Spain, 7Unidad de Investigacio´n, Hospital Costa del Sol- CIBERESP, MARBELLA, Spain, 8Servicio de Respiratorio, Hospital de Txagorritxu, VITORIA, Spain Objectives: To validate a short severity scale for patients attended at the Emergency Room (ER) by a COPD exacerbation. Methods: Our short severity scale for patients attended at the ER by a COPD exacerbation was developed by our research team which included pneumologist and physicians working at the ER. Data was collected from a pilot study from a prospective cohort study of patients with COPD exacerbation from 10 hospitals. Collection of data started at the ER with follow-up during the hospital admission or discharge at home up to two weeks after presentation at the ER. Our short severity scale included the following variables: respiratory frequency, presence of hemodynamic instability at admission, PO2 or oxygen saturation, pH, presence of hypercapnia and consciousness level, measured through the Glasgow scale. Main outcome was composed by mortality, major complications, admission at ICU, or need for mechanical ventilation during the admission or the first week after discharge. We performed univariate analysis to study the correlation of our severity scale (three categories: mild, moderate and severe COPD exacerbation) with the outcome of interest. Results: Data was available from 951 patients, of whom 101 (10.6%) presented adverse evolution. The univariate analysis showed that only 34 (6.50%) patients classified as mild COPD exacerbation presented any adverse event, while in the moderate class were 25 (9.40%) and 42 (25.93) in the severe group (P B 0.0001). Compared with the mild group, the moderate had an Odds Ratio (OR) of 1.49 (95% Confidence Limits95% CL-0.87 to 2.56) while the severe group had an OR of 5.03 (95% CL: 3.07 to 8.25). The c parameter was of 0.66. Conclusion: This preliminary analysis shows an acceptable correlation of our severity scale with the presence of any adverse event in a short time. Nevertheless, its discriminative ability is low.
192 RECURRENCE AND MORTALITY IN SECONDARY ATRIAL FIBRILLATION: A POPULATION-BASED STUDY R. Greenlee1, P. Chyou1, P. Sharma2, P. Smith2, J. Hayes2, R. Mareedu3, M. Ortiz4, H. Vidaillet1,2 1Marshfield Clinic Research Foundation, MARSHFIELD, United States of America, 2Marshfield Clinic, MARSHFIELD, United States of America, 3Medical College of Wisconsin, MILWAUKEE, United States of America, 4Clinical Hospital, University of Chile, SANTIAGO, Chile Background: ACC/AHA/ESC guidelines for atrial fibrillation (AF) suggest care decisions for secondary AF (SAF), following reversible antecedents, should be considered separately from idiopathic (or primary) AF (PAF), a known predictor of stroke and mortality. The clinical course of SAF amongst community patients is not well described. Objectives: We evaluated the clinical course of SAF and PAF within the Marshfield Epidemiologic Study Area (MESA), a resource for population-based clinical epidemiology in Wisconsin, USA. Methods: MESA is a defined region of 60,000 where a single health system captures nearly all medical care. We verified all newly diagnosed AF, 1991–1995, with ECG review. Incident AF within
55 30 days of an acute predisposing event was considered SAF. AF without reversible antecedent was PAF. We assessed first recurrence and progression to chronic AF through 5 years post-diagnosis, and deaths through May 2007. Analysis included direct estimation of outcome rates, Kaplan–Meier survival curves, and Cox proportional hazards regression adjusting for age and gender. Results: There were 270 PAF and 231 SAF cases. Of SAF, 45% were diagnosed following acute myocardial infarction or cardiac surgery; other antecedents included non-cardiac surgery, pneumonia, sepsis, trauma, and seizures. Recurrence-free survival curves were similar after two months. Overall, rate of first recurrence was nominally but not significantly higher in SAF (n = 88) than PAF (n = 91) (27 vs. 19 per 100 personyears, adjusted hazard ratio (HR) 1.2, 95% confidence interval (CI) 0.89, 1.62). Chronic AF developed in SAF (11%) less often than in PAF (20%). Adjusted mortality was identical (HR 1.01, 95% CI 0.82, 1.24). Conclusions: In 5-year follow-up, recurrence was as common in SAF as in PAF, although SAF progressed to chronic AF less often. Long-term mortality was identical. AF perceived clinically as secondary to reversible causes may instead reflect emergence of underlying AF susceptibility, requiring corresponding adjustment in clinical decision-making.
260 INTER-OBSERVER AGREEMENT IN ABSTRACTION OF DATA FROM CLINICAL RECORDS E. Alves1,2, V. Morais1,2, S. Correia1,2, A. Azevedo1,2 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health - University of Porto (ISPUP), PORTO, Portugal Background: Clinical records are an important source of information for epidemiologic studies. Objectives: To evaluate the inter-observer variability in data abstraction from clinical records, by two reviewers, and to assess the agreement between data recovered from clinical records or obtained through structured questionnaires, applied by trained interviewers for research purposes, to mothers of a birth cohort. Methods: As part of a population-based cohort study of consecutive newborns, 8125 pregnant women were recruited in public hospitals of Porto, in 2005–2006. Among 3771 with missing data on birth weight, mother’s anthropometrics, hypertensive disorders of pregnancy or family history of diabetes, we randomly selected 80 per hospital (n = 400). Data from clinical records were abstracted by 2 blinded reviewers. Discordant data were re-reviewed and resolved by consensus. After correction of discordant data, we assessed agreement with data collected by questionnaire. Results: Mean (standard deviation (SD)) birth weight was 3142 (565) g as recorded by reviewer 1 and 3140 (566) g by reviewer 2, P = 0.19. Mother’s height was 162.0 (5.7) cm and 162.1 (5.6) cm, P = 0.36, and usual prepregnancy weight 62.5 (10.1) Kg and 62.4 (10.0) Kg, P = 0.28, respectively. The kappa coefficient (95% confidence interval (CI)) for the agreement on hypertensive disorders of pregnancy was 0.90 (0.80–1.00). When assessing the agreement between data collected by questionnaire and clinical records, mean (SD) birth weight was 3168 (521) g and 3167 (534) g, P = 0.95, respectively. Mother’s height was 161.4 (5.5) cm and 162.0 (5.4) cm, P \ 0.01, and usual prepregnancy weight 61.0 (9.9) Kg and 62.0 (10.6) Kg, P = 0.01, respectively. Kappa (95%CI) for agreement on hypertensive disorders of pregnancy was 0.60 (0.39–0.82) and on family history of diabetes 0.53 (0.42–0.65). Conclusions: Overall, the inter-observer variability was low. The importance of implementing standardized methods to abstract data from clinical records cannot be overemphasized to improve the validity of data.
123
56
Early Morning Session 4—Infectious diseases epidemiology III 242 HELICOBACTER PYLORI INFECTION AND SOCIOECONOMIC STATUS IN A HIGH PREVALENCE DEVELOPED COUNTRY: PREVALENCE, INCIDENCE AND SEROREVERSION J. Bastos1,2, B. Peleteiro1,2, N. Lunet1,2, H. Pinto 1, 2, A. Marinho3, T. Guimara˜es2, 4, C. La Vecchia5,6, Henrique Barros 1,2 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health of the University of Porto, PORTO, Portugal, 3Department of Immunology, University of Porto Medical School, PORTO, Portugal, 4Department of Biochemistry, University of Porto Medical School, PORTO, Portugal, 5 Laboratory of Epidemiology, Istituto di Richerche Farmacologiche ‘‘Mario Negri’’, MILAN, Italy, 6Istituto di Statistica Medica e Biometria, Universita` degli Studi di Milano, MILAN, Italy Background: The estimation of the prevalence of H. pylori infection in different age groups, and according to socio-demographic characteristics, contributes to predict the future burden of H. pylori related diseases and to a better understanding of the gastric cancer dynamics. Objectives: This study aims to quantify the prevalence and the incidence of H. pylori infection and to identify, and quantify the major correlates of infection in a Portuguese urban population. Methods: A representative sample of the non-institutionalised adult inhabitants of Porto, Portugal, was recruited as part of a health and nutrition survey. A venous blood sample was available for 2067 participants. Gender-, age- and education- adjusted prevalence ratios (RR) and corresponding 95% confidence intervals (95%CI) were computed using binomial or the modified Poisson regression, as appropriate. Poisson regression was used to compute incidence rate ratios (RR). Results: The overall prevalence of H. pylori infection was 84.2% (95%CI: 82.4–86.1%), ranging from 73.9% in subjects under 31 years to 89.6% in those between 61 and 70 years. There was an inverse relation between education and prevalence of H. pylori infection (C10 vs. B4 years: RR = 0.88, 95%CI: 0.84– 0.91). The prevalence of infection increased across age groups in more educated subjects (P for trend \ 0.001), but not among the less educated. The incidence rate was 3.20/100 person-years (95%CI: 2.00–5.40). There was an inverse relation between education and incidence of H. pylori infection ([9 vs. B9 years: RR = 0.24, 95%CI: 0.07–0.81). The seroreversion rate was 0.95/ 100 person-years (95%CI: 0.37–3.26). Conclusions: The prevalence of infection in Portugal remains among the highest in Europe, particularly among less educated and lower social class individuals. This suggests that stomach cancer incidence and mortality, in the near future, is likely to remain high in Portugal as compared to other western European countries.
269 COVERAGE OF HIV SCREENING BEFORE THE FIRST CONTACT OF DRUG USERS WITH THE PUBLIC DRUG TREATMENT NETWORK IN PORTUGAL R. Lucas1,2, M. J. Santos3, L. Prasad1,2, H. Barros1,2,3 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health of the University of Porto, PORTO, Portugal, 3National Coordination for HIV/AIDS, LISBOA, Portugal
123
IEA-EEF European Congress of Epidemiology 2009 Background: Quantifying the coverage of HIV testing before the first contact with the drug treatment network and the potential importance of these structures in early HIV diagnosis is relevant in settings with concentrated HIV epidemics. Objectives: To estimate the frequency of previous HIV testing among first time clients of drug treatment centres and the association between testing and sociodemographic and drug-use related characteristics. Methods: We conducted a survey at the Portuguese public network of drug treatment centres. From June to December 2007, 902 first time visitors to 45 centres in the country were systematically approached and invited to participate in a voluntary counselling and testing programme. Participants were enrolled after informed consent. A standardised questionnaire was applied by trained professionals. Associations between previous HIV testing and sociodemographic and drug-use related characteristics were estimated using odds ratios (OR) and 95% confidence intervals (95%CI) adjusted for sex, age and life injection drug use (IDU). Results: Mean (SD) age was 32.7 (8.3) years, 138 (15.7%) were women and 387 (42.9%) were IDU. Overall, 188 (28.8%) clients had had previous HIV testing of which 21.3% were positive. In those with negative result, 60.1% had conducted the test in the preceding year. Previous testing was more frequent among men (OR = 1.78, 95%CI: 1.08–2.95) and in older DU (reference \25 years-old; OR([44) = 2.98, 95%CI: 1.26–7.06). Participants who were positive for hepatitis C virus were less likely to have a previous HIV test result (OR = 0.29, 95%CI: 0.08–1.00). There were no differences in the frequency of a previous result according to IDU (OR = 1.02, 95%CI: 0.68–1.52) or, among IDU, according to duration of injection (per year OR = 1.03, 95%CI: 0.98–1.09) or with ever sharing injection material (OR = 1.02, 95%CI:0.42–2.50). Conclusion: Coverage of HIV testing in drug users before contact with specialised structures was low. Higher risk did not translate into more frequent testing.
226 MORTALITY RATE AND OUTCOME FACTORS IN MIXED CRYOGLOBULINEMIA: THE IMPACT OF HCV A. Della Rossa1, A. Tavoni2, A. D’Ascanio1, E. Catarsi2, F. Marchi2, W. Bencivelli2, S. Salvadori3, P. Migliorini2, S. Bombardieri1 1 Rheumatology Unit, Department of Internal Medicine, University of Pisa, PISA, Italy, 2Immunoallergology Unit, Department of Internal Medicine, University of Pisa, PISA, Italy, 3CNR, Institute of Clinical Physiology, PISA, Italy Background: Mixed cryoglobulinemia (MC) is a chronic small vessel vasculitis clinical syndrome with features like purpura, weakness and arthralgias. Some studies suggest that chronic HCV infection ‘‘per se’’ has not impact on survival, however HCV-negative MC patients show distinct clinical pictures and higher mortality rates as compared to HCV-positive subjects. Objective: The purpose of the study was to retrospectively evaluate mortality and outcome factors in patients affected by HCV related MC. Methods: In 70 patients affected by HCV related MC whose disease onset occurred after 1990 Clinical and serological features were collected and different organs involvement (skin, joints, peripheral vascular/neuropathy, sicca syndrome, liver, kidneys, abdominal vasculitis, B-cell lymphoma) was defined according to standardized criteria. Cryocrit, rheumatoid factor and complement factors were evaluated. Cumulative survival rates were computed by the Kaplan-Meier method and the difference between survival curves by the logrank statistics. Association between mortality and clinical variables and serologic parameters was evaluated by the Cox proportional hazard regression model. Results:
IEA-EEF European Congress of Epidemiology 2009 At the end of follow-up (2008) 46 (74.2%) patients were alive, 16 (25.8%) were deceased after 93 ± 56 months from disease onset and 8 were lost to follow-up. Causes of death were MC related or possibly MC related in 13 patients (81.2%). Patients with renal involvement, abdominal vasculitis and cryocrit [5% showed a significantly worse survival. The cox model supported the prognostic findings obtained. For abdominal vasculitis hazard ratio was 15.9 (CI95%: 2–125), for
57 renal involvement 23.8 (CI95%: 1.1–500) and for cryocrit 1.1 (CI95%: 0.74–167). Cumulative survival of MC patients 10 years after diagnosis was significantly lower then the general Italian population. Conclusion: The study confirms the increased risk of death for renal involvement and intestinal vasculitis found in other studies both by means of cumulative survival rates and Cox model which points them as strong predictors of mortality.
123
58
Saturday, 29 August 2009 Parallel late morning oral sessions: 11.30–12.50 Late Morning Session 1—Cancer epidemiology IIIA 156 EFFECT OF SMOKING AND ALCOHOL PRICE INDEXING ON LUNG AND BREAST CANCER INCIDENCE IN DENMARK I. Soerjomataram1, E. de Vries1, G. Engholm2, G. Paludan-Mu¨ller2, H. Bronnum-Hansen3, H. Storm2, J. Barendregt4 1Department of Public Health, Erasmus MC, ROTTERDAM, The Netherlands, 2 Danish Cancer Society, COPENHAGEN, Denmark, 3National Institute of Public Health, University of Southern Denmark, COPENHAGEN, Denmark, 4The University of Queensland, School of Population Health, HERSTON, Australia Background: Lung and breast cancers present a large burden in population health. Smoking and alcohol consumption are related to increased incident of both cancers. Increasing cigarette and alcoholic beverage price can reduce consumption, and may finally reduce the incident of these cancers. Objectives: We aimed to demonstrate the potential impact of smoking and alcohol intervention program i.e. price indexing on future burden of lung and breast cancer in Denmark. Method: We used Prevent v.3.01 to assess the changes in cancer incidence as a result of risk factor changes. Prevent is a computer simulation model that estimates the effect of changes in risk factor prevalence on disease occurrence and/or mortality accounting for trends and interventions. Incidence of lung and breast cancer in 2050 was predicted under two scenarios: total elimination of smoking and alcohol intake, and decreasing risk factors prevalence due to 10% annual increase in cigarette and alcohol beverage price. Danish data from the household survey, cancer registration and Eurostat were used. Results: A 60% lower lung cancer incidence can be expected if smoking were totally eliminated. Annual 10% price increase may reduce lung cancer incidence by 8 and 6 cases per 100,000 personyears, in men and women in 2050 (593 prevented new cases). Intervention to reduce alcohol consumption to the recommended level may lower breast cancer cases to 5,609 as compared to the expected 6,048 new cases in 2050. Annual 10% price increase in alcoholic beverages achieved similar reduction by 2050 to a prevention program that reduced intake to the recommended level. Conclusions: The future burden of lung and breast cancer could be markedly reduced by population intervention on smoking and alcohol consumption. Public health model such as Prevent illustrates the benefit of interventions and may serve as guidance in political decision-making.
249 TREATMENT DELAY OF BREAST CANCER PATIENTS RECEIVING CURATIVE TREATMENT IN PODLASKIE VOIVODESHIP IN 2001–2002 POPULATION STUDY D. Mas´lach1, M. Krzy_zak-Mas´lach1, A. Szpak1, M. Juczewska2, K. Maksimowicz2, U. Łapin´ska3 1Medical University of Bialystok, Department of Public Health, BIALYSTOK, Poland, 2Bialystok Oncology Center Voivodeship Cancer Registry, BIALYSTOK, Poland, 3Office of the Marshal of the Podlaskie Voivodeship, Department of Health, BIALYSTOK, Poland Background: Treatment delay is an important problem in current oncology. Knowledge about its impact on curability should be used to plan the treatment policy in the region. This can contribute to better
123
IEA-EEF European Congress of Epidemiology 2009 results of treatment. Results of population studies concerning treatment delay ought to be applied to organize the treatment in compliance with GMP (Good Medical Practice). Objectives: Evaluation of treatment delay from diagnosis to first treatment among breast cancer women diagnosed in Podlaskie Voivodeship in 2001–2002. Methods: Among of 698 primary breast cancer cases registered in CR (Voivodeship Cancer Registry) a cohort of 579 was established. Data collected from CR and medical records from hospitals in the region were combined. The earliest day of diagnosis was considered for calculation. The number of days from diagnosis to the first treatment was calculated and compared with recommendations of The Royal College of Physicians and The Royal College of Radiologists. The proportion of applied oncological treatment (surgical, chemotherapy, radiotherapy, hormonotherapy) was also calculated. Results: The following treatments were applied as the first ones: surgery (58.0%), chemotherapy (32.1%), hormonotherapy (8.5%) and radiotherapy (1.4%). The average waiting time was 42 days (median was 14 days). 34.5% patients waited less then 7 days, 35.8% from 8 to 28 days and 29.7% longer. In analysed cohort the recommendations for GMP concerning chemotherapy were applied to 34.4% of patients. The maximum acceptable time was applied to 30.1%. However, it should be pointed out that 35.5% of patients did not comply with this criteria. Conclusions: High proportion of patients whose curative treatment has been substantially delayed is one of reasons for poor results of 5-year survival in Poland. The study of 5-year survival rate of this cohort is in progress.
142 CANCER INCIDENCE PREDICTION FOR THE YEARS 2010 AND 2015 S. Gozdz1, T. Dyba2, R. Mezyk1 1Holycross Cancer Centre, KIELCE, Poland, 2Finnish Cancer Registry, HELSINKI, Finland Introduction: The Holycross Cancer Registry is one of the 17 Regional Registries of Malignant Tumors in Poland. It has been functioning since 1986 and is one of the departments of The Holycross Cancer Centre. The population covered by the Registry was 1297413 persons in 2002, including 663742 women and 633671 men. There are about 32 000 patients registered at present with active oncological diseases and 5295 new cases of malignant tumors were registered in 2002: 2376 for women and 2919 for men. Aim of the study: The aim of the study was to predict a new number of cancer cases in the Holycross Region in years 2010 and 2015 for the most common cancer sites in both sexes. Methods: The calculations were based on the proposed earlier prediction method (Dyba T., Hakulinen T., Pa¨iva¨rinta L.) making use of the Poisson assumption for the observed number of cases and calculations were performed using ready-made Stata macros. The base of prediction 1993–2002 was used to calculate age-specific and overall results both for number of cases and age-adjusted rates with accompanying 95% prediction intervals. The future predicted population numbers for the years 2010 and 2015 were used as standard populations, respectively, in the calculations. Results: The future predicted numbers of incidence and 95% prediction intervals for the years 2010 and 2015 were calculated. Prediction was made for most frequent tumors separately for women and for men. The most frequently cancers were: (1) for women: Breast, Skin, Cervix Uteri, Corpus Uteri, Bronchus&Lung, Colon, Stomach and Rectum; (2) for men: Bronchus&Lung, Prostate, Bladder, Skin, Stomach, Colon, Rectum, Larynx and Kidney. Conclusions: In comparison to the year 2002 one can predict further increase of incidence in women for Bronchus & Lung (2010—62%, 2015—88%), Breast (48%, 76%), Rectum (22%, 48%) both in 2010 as in 2015. Decrease of incidence is predicted for Stomach (24%, 30%), Corpus Uteri (21%, 5%) and Cervix Uteri (7%, 12%). For men the further increase of incidence for Bladder (74%, 127%), Kidney (74%, 121%), Prostate (56%, 113%), Colon (57%, 96%), Rectum (53%, 95%), Bronchus & Lung (27%, 52%), Larynx (26%, 35%) is predicted.
IEA-EEF European Congress of Epidemiology 2009
Late Morning Session 2—Perinatal epidemiology III 39 PHYSICAL ACTIVITY AND FERTILITY IN WOMEN—THE NORTH-TRØNDELAG HEALTH STUDY S.L Gudmundsdottir1, W.D. Flanders2, L.B. Augestad1 1Human Movement Science Programme, Faculty of Social Sciences and Technology Management, Norwegian University of Science and Technology, TRONDHEIM, Norway, 2Departments of Epidemiology and Biostatistics, Emory University, ATLANTA, United States of America Background: Reproductive dysfunction is reported to have a higher prevalence in athletes than non-athletes (Otis et al., 1997; Russell et al., 1984), with clinical consequences that may include infertility Objective: The purpose of this study was to evaluate the association between physical activity (PA) and fertility and parity in healthy women. Methods: A population based health survey (HUNT 1) was conducted during 1984–1986 in Nord-Trøndelag county, Norway and a follow up during 1995–1997 (HUNT 2). This study included 3,897 women, \45 years of age in HUNT2. PA was assessed with baseline questionnaire and fertility and parity by a questionnaire at follow up. Results: Increased frequency, duration and intensity of PA was associated with increased subfertility and frequency and duration of PA with voluntary childlessness. After adjusting for age, parity, smoking and marital status, women active on most days were 3.2 times more likely to have fertility problems than inactive women. Exercising to exhaustion was associated with 2.3 times the odds of fertility problems compared to low intensity. Women with the highest intensity of PA had the lowest frequency of continuing nulliparity and highest frequency of having 3 or more children during follow up (P \ 0.05). A sensitivity analyses including BMI as a confounder did not alter the results. Increased risk of infertility was only found for the small group of women reporting highest category of intensity and frequency of PA. We did not find any associations between lower levels of activity and fertility or parity. Conclusion: We focused on overall occurrence of infertility in the population but did not confirm it with biological testing. Awareness of possible risks of infertility should be endorsed among none-athletic women carrying out vigorous exercise.
103 RISK FACTORS FOR PRETERM BIRTH, LOW BIRTH WEIGHT AND SMALL FOR GESTATIONAL AGE BIRTHS M. Heaman1, D. Kingston2, B. Chalmers3, R. Sauve4, L. Lee5, D. Young6, S. Taylor-Clapp7 for the Maternity Experiences Study Group of the Canadian Perinatal Surveillance System 1University of Manitoba; WINNIPEG, Canada, 2McMaster University, HAMILTON, Canada, 3University of Ottawa, OTTAWA, Canada, 4 University of Calgary, CALGARY, Canada, 5British Columbia Perinatal Health Program, VANCOUVER, Canada, 6Dalhousie University, HALIFAX, Canada, 7Public Health Agency of Canada, OTTAWA, Canada Background: The rate of preterm birth (PTB) has steadily increased in Canada, while the rate of small for gestational age (SGA) birth has decreased. Objectives: The purpose of this study was to identify and compare risk factors for PTB, low birth weight (LBW), and SGA births using a national sample of Canadian women. Methods: An analysis of data collected by the Maternity Experiences Survey (MES) of the Canadian Perinatal Surveillance System was conducted. The MES used a randomly selected sample of recent mothers drawn
59 from the 2006 Canadian Census; participants completed a computer assisted telephone interview conducted by Statistics Canada (n = 6,421). Mothers were [15 years of age, had given birth to a singleton infant, and were living with their infant at the time of the interview (between 9 and 14 months postpartum). Backward stepwise multivariate logistic regression models were constructed for each outcome, entering risk factors for which there was evidence of a significant association from the univariate analyses. Results: The rate of PTB was 6.1%, LBW 4.7%, and SGA 7.9%. Risk factors for all three outcomes included smoking during pregnancy, low weight gain (\ 20 pounds), and nulliparity. Additional risk factors for PTB and LBW were maternal age [ 35 and antenatal hospitalization (indicative of a pregnancy complication), while risk factors for both PTB and SGA were short stature and history of previous PTB, and a risk factor for both LBW and SGA was being underweight prior to pregnancy. Reporting life as ‘‘very stressful’’ in the past year was a risk factor specific to PTB, while single marital status was specific to SGA. Conclusions: Modifiable risk factors were identified for the three outcomes, although there were some differences among PTB, LBW and SGA. Preventive measures such as smoking cessation and ensuring adequate weight gain during pregnancy may improve pregnancy outcomes among Canadian women.
124 PREVALENCE AND SOCIO-DEMOGRAPHIC DETERMINANTS OF ANAEMIA IN PREGNANCY IN NORTHWEST RUSSIA: A REGISTRY-BASED STUDY ON 23,950 BIRTHS, 1973–2001 E. Chumak1, A. M. Grjibovski1,2 1International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia, 2Norwegian Institute of Public Health, OSLO, Norway Introduction: Maternal anaemia is known to be associated with poor pregnancy outcomes. Previous studies from Northwest Russia show that the crisis of the 1990 s was accompanied by increase in preterm births, decrease in birthweight and increase in social variations in pregnancy outcomes. These changes may be at least partly attributed to increased prevalence of anaemia in Russian women, although the evidence is scarce. Aim: to assess the prevalence of anaemia in pregnancy and its determinants among Russian women during the Soviet time and transition using the first Russian population-based birth registry. Methods: All singleton births in Monchegorsk, Northwest Russia, 1973–2001 were included (n = 23,950). Multiple logistic regression was used to estimate independent effects of socio-demographic factors on the odds of anaemia in pregnancy. Crude and adjusted odds ratios (OR) with 95% confidence intervals (CI) were calculated. Results: The prevalence of maternal anaemia increased from 47% in 1973 to 89% in 2001. In crude analysis, women aged 25–29 and 30–34 years had higher odds for anaemia than women aged 20–24 years (OR = 1.07, 95%CI: 1.01–1.13 and OR = 1.10, 95%CI: 1.01–1.19, respectively). Women working as clerical staff were less likely (OR = 0.85; 95%CI: 0.79–0.91) while unemployed women were more likely (OR = 1.24, 95%CI: 1.05–1.47) to have anaemia in pregnancy than professionals. Unmarried women had higher odds for anaemia than their married counterparts (OR = 1.26, 95%CI: 1.16– 1.35). Women who drank alcohol in pregnancy were less likely to be anaemic (OR = 0.62, 95%CI: 0.42–0.91). No variations in the prevalence of anaemia by maternal socio-demographic characteristics remained significant after adjustment for age, occupation, marital status, parity, smoking, alcohol consumption, and year of delivery. Conclusions: The prevalence of maternal anaemia in Monchegorsk doubled during transition. Women from all social strata were equally affected by anaemia, which contradicts previous findings on pregnancy outcomes. The reasons for this contradiction will be discussed.
123
60
283 IMPACT OF EDUCATIONAL CAMPAIGN ON SOCIO-DEMOGRAPHIC FACTORS AFFECTING WOMEN’S KNOWLEDGE OF FOLIC ACID AND PRECONCEPTIONAL FOLIC ACID SUPPLEMENTATION IN REGARD TO NEURAL TUBE DEFECTS’ PREVENTION E. Mierzejewska, K. Szamotulska National Research Institute of Mother and Child, WARSAW, Poland Background: National Primary Prevention Program of Neural Tube Defects (NTD) based on an educational campaign was developed in Poland in 1997–2007 to disseminate the knowledge, that folic acid (FA), when supplemented during periconceptional period, is crucial for prevention of NTD’s in offspring. Objectives: To investigate whether campaign influenced socio-demographic determinants (educational level, economic status, place of residence, number of children, belief about the need of pregnancy planning) of women’s knowledge of FA and preconceptional FA supplementation. Methods: Cross-sectional studies on knowledge and behaviours toward FA among women were conducted in 2001 and 2007 years at primary health centres in 31 subregions. Standard questionnaires were administered to 314 women (2001) and 272 women (2007), aged 18–35, pregnant at the time of or during 2 years before the interview. Multiple logistic regression analysis was applied. Results: The proportion of women knowing the benefits of FA for foetus and the right time for supplementation increased from 21% in 2001 to 44.1% in 2007. The proportion of women taking FA already before the pregnancy increased from 11.5% to 28.7% respectively. The lack of proper knowledge of FA was significantly related to educational level both in 2001 and 2007 (OR = 10.3; 95%CI: 4.5–23.9 and OR = 6.3; 95%CI: 2.7–14.5 respectively for women in low vs high education). The lack of FA usage before pregnancy was also significantly related to educational level in both studies (OR = 5.4; 95%CI: 1.4–20.1 and OR = 4.0; 95%CI: 1.7– 9.5) respectively for women in low vs high education), but for the 2007 study the lack of preconceptional supplementation became significantly related to economic status (OR = 5.0; 95%CI: 1.4–17.9) for women with bad vs good economic status), place of residence (OR = 2.1; 95%CI: 1.1–4.1 for women living in the countryside vs towns) and belief that pregnancy should be planned (respectively OR = 1.9; 95%CI:1.0–3.5 for women rather convinced and OR = 5.4; 95%CI: 1.7–17.1 for women not convinced vs women absolutely convinced about need of pregnancy planning). Conclusions: Due to the campaign the impact of educational level on differences in knowledge and behaviours concerning FA among women has decreased, but also limitations on preconceptional use of FA independent of the lack of knowledge as socio-economic status and women’s beliefs were revealed.
Late Morning Session 3—Cardioepidemiology III 185 KNOWLEDGE OF BLOOD PRESSURE LEVELS AND TARGETS IN CORONARY PATIENTS ACROSS EUROPE: RESULTS FROM THE EUROASPIRE III STUDY C. Prugger1, D. De Bacquer2, G. De Backer2, G. Ambrosio3, Z. Reiner4, D. Gaita5, D. Wood6, K. Kotseva6, U. Keil1, J. Heidrich1 1 ¨ NSTER, Germany 2Ghent University, Mu¨nster University, MU GHENT, Belgium, 3Venice City Hospital, VENICE, Italy 4University Hospital Centre, ZAGREB, Croatia, 5University of Medicine
123
IEA-EEF European Congress of Epidemiology 2009 & Pharmacy, TIMISOARA, Romania, 6Imperial College London, LONDON, United Kingdom Background: Blood pressure control is a key issue in coronary patients. Patients’ knowledge of their own blood pressure level and therapeutic goals are important for compliance and treatment success. Objectives: To investigate knowledge of blood pressure levels and targets in a large sample of coronary patients across Europe. Methods: EUROASPIRE III is a cross-sectional study conducted in 2006–2007 among coronary patients aged\80 years from 22 European countries. Standardized interviews and examinations were performed on average 15 months after hospitalisation for coronary heart disease. Overall, 7686 patients using antihypertensive (AH) medication were included in the analyses. Knowledge of target blood pressure was classified according to current guidelines (cutpoint 140/90 mmHg; 130/ 80 mmHg in diabetic patients). We used logistic regression analysis to investigate factors associated with knowledge of target blood pressure. Results: Mean age was 62 (SD 9.5) years and 26% of the study population were female. 73% reported they were prescribed AH medication specifically for hypertension treatment. In total, 81% were aware of their blood pressure level and 58% provided correct knowledge of target blood pressure. Increasing age and lower education were inversely related to knowledge of target blood pressure (52% at age 70–79 years vs. 60% at age \50 years, OR 0.75, 95% CI 0.62–0.92; 73% in those with higher education vs. 46% in those with primary education, OR 2.21, CI 1.85–2.64). Diabetes and obesity were associated with poorer target knowledge (47% in diabetics vs. 62% in non-diabetics, OR 0.48, CI 0.43–0.54; 55% in obese vs. 59% in non-obese, OR 0.80, CI 0.71–0.89). Patients who reported lifestyle advice by a health professional provided correct target knowledge more frequently than those not advised. Conclusions: Knowledge of treatment targets for blood pressure is inappropriate in this high-risk population of coronary patients. More efforts are necessary to improve communication of treatment goals.
223 HYPOTHESIS-ORIENTED FOOD PATTERNS AND INCIDENCE OF HYPERTENSION IN THE SUN PROSPECTIVE COHORT E. Toledo1,2, F. Carmona-Torre2, M. Bes-Rastrollo2, J. Nun˜ezCo´rdoba2, F. Guille´n-Grima2, M. A. Martı´nez-Gonza´lez2 1 Department of Preventive Medicine and Quality Management, Hospital Virgen del Camino, PAMPLONA, Spain, 2Department of Preventive Medicine and Public Health, University of Navarra, PAMPLONA, Spain Background: Among the different risk factors for hypertension—the individual risk factor that causes most deaths worldwide- diet has received increasing interest. Several food patterns (FP) have been postulated as potentially able to prevent hypertension. However, scarce evidence exists in Mediterranean countries about FP and the risk of hypertension. Objectives: To study the association between adherence to several a priori-defined healthy FP and the risk of hypertension. Methods: This study was performed within the frame of the SUN [Seguimiento Universidad de Navarra (University of Navarra follow-up)] project, a prospective, multipurpose, dynamic cohort study. We followed-up 10,800 men and women (all of them university graduates) initially free of hypertension during a variable period of time (range 2–6 years, median: 4.6 years). During followup, 640 new cases of self-reported physician-diagnosed hypertension were observed. Baseline diet was assessed using a validated 136item food frequency questionnaire. Validated information about
IEA-EEF European Congress of Epidemiology 2009 non-dietary potential confounders was also gathered. We assessed the association between eight different hypothesis-oriented FP (Diet Quality Index-International, the Recommended Food-Score, the Quantitative Index for Dietary diversity, both in kcal and in g, the Healthy Eating Index, the Alternate Healthy Eating Index, the Dietary Guidelines for Americans Adherence Index and the Score of Attitudes Toward a Healthy Diet) and incident hypertension using multivariable Cox models. Results: In multivariable analyses, a higher adherence to the DASH diet was significantly associated with a lower risk for developing hypertension [P for trend 0.02]. No association with incident hypertension was observed for any of the other seven FP assessed. Conclusions: Our results support a longterm protection of the DASH diet against the incidence of hypertension, but we found no evidence of a similar inverse association with hypertension for any other a priori-defined healthy food pattern.
182 CALCIUM IS AN INDEPENDENT RISK FACTOR FOR HYPERTENSION; A POPULATION-BASED COHORT STUDY J. W. Kim1, H. Kim1, K. Ko1, H. J. Koo1, J. K. Park1, C. Shin2, S. S. Kim1 1Division of Epidemiology and Health Index, Center for Genome Science, Korea Center for Disease Control and Prevention, SEOUL, South Korea 2Institute of Human Genomic Study, College of Medicine, Korea University, SEOUL, South Korea Background: Recently, an association between primary hyperparathyroidism and hypertension was reported, but the correlation between plasma calcium level and hypertension is unclear. Objectives: To investigate correlations between plasma calcium along with various conventional risk factors and hypertension in a healthy population. Methods: Data from Ansan and Ansung cohorts from the Korean Genome and Epidemiology Study (KoGES) was used, which is an ongoing population-based prospective study of 7,417 Korean adults aged 40–69 years old that began in 2001 with 2-year follow up intervals until the tertiary survey in 2008. Receiver operation characteristics curve was used to determine the cutoff values of albumin-corrected calcium, urinary calcium/creatinine ratio, and logtransformed homeostasis model assessment (HOMA) index. Hazard ratios (HR) and 95% confidence intervals (CI) of corrected calcium and other conventional risk factors for hypertension were estimated using Cox’s proportional-hazards model. Results: On univariate analysis, plasma renin activity (PRA [ 0.65 ng/mL/h), corrected calcium ([9.5 mg/dL), abdominal obesity (male [ 90 cm, female [ 80 cm), triglyceride ([150 mg/dL), old age, log-transformed HOMA index ([0.67), body mass index ([23 kg/m2), area of residence (urban vs. rural), fasting glucose ([100 mg/dL), presence of prehypertension, microalbuminuria, and azotemia (creatinine clearance \ 60 mL/min/1.73m2) showed significant correlation with incidence of hypertension. The associations between alcohol consumption, smoking, urinary calcium/creatinine ratio, C-reactive protein, and exercise with hypertension were insignificant. On multivariate analysis, higher corrected calcium level (HR 1.33, 95% CI 1.02–1.73), presence of prehypertension (HR 2.23, 95% CI 1.977–2.515), triglyceride (HR 1.20, 95% CI 1.07–1.35), and presence of microalbuminuria (HR 1.572, 95% CI 1.26–1.96) showed positive correlation, whereas PRA [ 0.65 ng/mL/h (HR 0.69, 95% CI 0.60–0.79) and urban area of residence (HR 0.646, 95% CI 0.57–0.73) showed negative correlation with incidence of hypertension. Conclusions: This is the first investigation showing a correlation between plasma calcium and hypertension. Subclinical primary hyperparathyroidism may have played a role.
61
243 EFFECT OF DEPRESSIVE SYMPTOMS ON THE INCIDENCE OF HYPERTENSION IN ADULTS: A POPULATION-BASED STUDY IN PORTO, PORTUGAL M. Pereira, D. Costa, A. Azevedo Department of Hygiene and Epidemiology, University of Porto Medical School; Cardiovascular R&D Unit; Institute of Public Health of the University of Porto (ISPUP), PORTO, Portugal Background: Hypertension has been linked to several humor disorders, including depression, but the association between hypertension incidence and depressive symptoms has not been adequately examined in longitudinal studies. Thus, uncertainty remains regarding the temporal sequence of reported associations. Objectives: To determine if depressive symptoms independently predict hypertension incidence. Methods: We followed up 444 individuals (59.5% women), aged C18 years, randomly selected from the urban population of Porto, Portugal, stratified by age below/above 40 years, initially free of hypertension. The recruitment occurred from 1999 to 2003. Hypertension was defined as blood pressure C 140/90 mmHg (two measurements on a single occasion) or the use of prescribed antihypertensive medication. Depressive symptoms were assessed with the Beck Depression Inventory (BDI), and Poisson regression was used to analyze crude and adjusted incidence rate ratios (RR), excluding 25 participants under antidepressant medication in the previous year. We considered the possible confounding effect of age, sex, education, body mass index, alcohol use, and cigarette smoking. All analyses were weighted to account for the sampling procedure. Results: During a median (interquartile range) follow-up of 4.0 (2.5–7.1) years, generating 1854.5 person-years at risk, 93 individuals developed hypertension, with an overall incidence rate (95% confidence interval (CI)) of 44.0 (35.4–55.3) per 1000 person-years. The median (interquartile range) of BDI scores was 5 (2–10). A 5-point increase in the BDI score was not significantly associated with hypertension incidence in crude (RR, 0.94; 95% CI, 0.78–1.13) and adjusted analysis (RR, 0.86; 95% CI, 0.68–1.08). The use of antidepressant medication was also not significantly associated with hypertension incidence (RR, 0.99; 95% CI, 0.65–1.53). Conclusions: Depressive symptoms did not predict hypertension incidence in this sample. Very few participants had BDI scores corresponding to moderate to severe depressive symptoms. Previous studies have generated divergent results. Our estimate supports those reporting a null or inverse association.
Late Morning Session 4—Environmental epidemiology 178 THE EFFECTS OF AMBIENT TEMPERATURE ON THE RISK OF MYOCARDIAL INFARCTION IN LONDON, UK K. Bhaskaran, London School of Hygiene And Tropical Medicine, LONDON, United Kingdom Background: The effects of weather parameters, in particular ambient temperature, on overall and cardiovascular mortality are well documented. However, effects on specific cardiovascular outcomes are less clear. Objectives: We aimed to assess the shortterm effects of ambient temperature on the incidence of myocardial infarction (MI). Methods: Using MI incidence data for Greater London from the Myocardial Ischaemia National Audit Project
123
62 (MINAP), and meteorological data from 11 local UK Met Office monitoring stations, we performed a daily time series analysis over the period 2003–2006, using a Poisson model to investigate the effect of daily mean temperature on the daily number of MIs, allowing for effects delayed by up to 28 days. We adjusted for season and long-term trend, ozone and particulate pollution, relative humidity, day of week, holiday, and circulating infection levels. We used Akaike’s Information Criterion to select between a linear temperature model and threshold models for separate cold and heat effects. Results: A median of 18 MIs per day (IQR 15 to 21) were recorded. The mean daily temperature was 11.7C (standard deviation 5.8, range -1 to 27). There was a strong relationship between ambient temperature and daily MI incidence in all models (P \ .001). The best model emerging from our model building process suggested a detrimental effect of cold at temperatures \ 19C: we estimated a 3.6% (95% CI 1.8–5.4) total increase in MI risk over 28 days following a 1C drop in temperature, with the effect mainly occurring in the first 7 days. We found no temperature effect above the 19C threshold, though power to detect heat effects may have been limited with only 159 days (11%) having mean temperatures over 19C. Conclusions: Drops in ambient temperature appear to increase the short-term risk of MI; this effect may partly explain the increases in overall mortality associated with cold weather.
106 FINDING CADMIUM THRESHOLD CONCENTRATION RELATED TO BREAST CANCER R. Kregzdyte1,2, L. Strumylaite1, D. Baranauskiene1, O. Abdrkhmanov1, 1Institute for Biomedical Research, Kaunas University of Medicine, KAUNAS, Lithuania 2Department of Preventive Medicine, Kaunas University of Medicine, KAUNAS, Lithuania Background: Breast cancer is the most common cancer in women worldwide and in Lithuania as well. Many studies are performed in order to find etiopathogenesis of breast cancer. A hospital based case––control study is carried out by researchers of Kaunas University of Medicine. One of its aims is to assess a relationship between cadmium exposure and the disease. The first findings show that cadmium concentration in breast tumor and normal tissue differs statistically significantly. Objective: The objective of the followed analysis was to determine cadmium concentration in breast tissue over which the cancer is more likely. Material and Methods: The concentration of cadmium was determined in breast tissue samples of 57 breast cancer and 50 benign tumor patients. Tumor and normal tissue close to tumor samples were taken for the analysis. The metal concentration was determined by atomic absorption spectrometry (PerkinElmer, Zeeman 3030). Receiver operating characteristic (ROC) curves were used to detect a cut-off value. Results: The area under the curve for cadmium concentration in tumor was 0.684 (95% confidence interval 0.579–0.788; P = 0.001), that indicated cadmium to be a significant predictor of the disease status. The cut-off point for optimal test was chosen to optimize the rate of true positives and minimizing the rate of false positives. The minimum sum of false positives and false negatives squared was at the concentration of 28 ng/g. Calculated sensitivity was 0.72 and the specificity was 0.66 at this point. Conclusions: Revised estimate of cadmium threshold concentration made on enlarged sample size is valuable predictor of breast cancer and is useful in the further studies.
123
IEA-EEF European Congress of Epidemiology 2009
35 SEARCHING FOR BIAS: RELIABILITY OF SELF-REPORTED HOUSEHOLD PESTICIDES USE AND ORS CORRECTION C. Fortes1, S. Mastroeni1, P. Boffetta2, V. Salvatori1, N. Melo1, S. Bolli3, P. Pasquini1 1Clinical Epidemiology Unit, IDI-IRCCS, ROME, Italy, 2 International Agency for Research on Cancer, LYON, France, 3 National Institute of Health, ISS, ROME, Italy Background: Household pesticide exposure has been associated with cancer risk in both adults and children. We investigated the reliability of reported lifetime household pesticide exposure through repeated administration of a standardized questionnaire. Methods: A questionnaire including detailed questions on lifetime frequency and duration of pesticides use in non-occupational circumstances was administered on two occasions to 163 cutaneous melanoma cases and 113 controls. We investigated the agreement between the two measurements taken on average 12 months apart and studied the association between differences in the two measurements and a set of explanatory variables. According to the results of the reliability analysis we also corrected OR estimates from the main study. Results: Agreement for ever/never use of pesticides outdoors and indoors was 73.6% (Cohen’s Kappa = 0.42) and 69.6% (Cohen’s Kappa = 0.28) respectively. Agreement for duration and frequency of use of pesticides outdoors was 89.5% (Cohen’s Kappa = 0.48) and 92.0% (Cohen’s Kappa = 0.40) respectively while for duration and frequency of use of pesticides indoors agreement was 75.4% (Cohen’s Kappa = 0.32) and 77.4% (Cohen’s Kappa = 0.28) respectively. Agreement was higher for duration (97.4%; Cohen’s Kappa = 0.72) and use of pesticides on domestic animals (86.4%; Cohen’s Kappa = 0.68). Agreement was not associated with case/ control status. The corrected OR showed a moderate increase with a reinforcement of the effect of pesticides. Conclusion: Overall, there was a good reproducibility in self-reported exposure to pesticides and was comparable to that found in studies of other environmental exposures.
300 ENVIRONMENTAL EXPOSURE TO BIOCONTAMINANTS IN KINDERGARTENS IN POLAND A. Buczynska, M. Cyprowski, P. Soroka, I. Szadkowska-Stanczyk, Nofer Institute of Occupational Medicine, LODZ, Poland Background: A number of published studies demonstrate that children are exposed to biocontaminants occurring in various concentrations not only at home but also in kindergartens. Objectives: The purpose of the study was to measure indoor allergens from mite (Der p 1), cat (Fel d 1), dog (Can f 1), cockroach (Bla g 2) and (1 ? 3)-b-D-glucans (molds indicator) in three municipal kindergartens in Lodz, Poland. Methods: Levels of biocontaminants were investigated in settled dust samples collected from floors in play rooms (n = 90) and from cloths of children (n = 34). Allergens were determined with ELISA method and glucans content was determined by using the LAL assay. Results: Children attending to the kindergartens were exposed to allergens which concentrations did not exceed the ‘‘threshold’’ levels. The highest allergen load per unit of settled dust was found for Can f 1 (0.97 ± 4.49) and Fel d 1 (0.30 ± 4.43). Der p 1 and Bla g 2 did not exceed the mean level of 0.02 lg/g dust. The average concentration of (1 ? 3)-b-D-glucans was 0.62 ± 2.31 lg/g of settle dust. The significant differences between the objects were found for Der p 1 (P = 0.014) and
IEA-EEF European Congress of Epidemiology 2009 (1 ? 3)-b-glucans (P \ 0.0001). The seasonal variability of biocontaminants levels was confirmed for dog allergen Can f 1 (winter: 1.29 ± 2.14 lg/g dust, summer: 0.80 ± 7.87 lg/g dust; P = 0.029) and for (1 ? 3)-b-glucans (summer: 0.76 ± 2.39 lg/g dust, winter: 0.54 ± 1.48 lg/g dust; P = 0.015). We confirmed pet’s allergens (Can f1, Fel d 1) transportation by pet-keeping children to the
63 kindergartens environment. The mean concentrations of Can f 1 and Fel d 1 in dust from pet’s owners cloths were 3.63 ± 1.47 lg Fel d 1 /g dust and 59.24 ± 5.39 lg Can f 1 /g dust. Conclusions: Recommendations for biocontaminants avoidance should include a proper cleaning system, ventilation and reduction of soft toys in the kindergartens faciliti.
123
64
Poster Presentations Thursday, 27 August 2009, 15.30–16.30 Cancer epidemiology I 53 DIET AND THE RISK OF LUNG CANCER IN MEN INHABITANTS OF THE UPPER SILESIA, POLAND M. Juszko-Piekut1, Z. Kolosza2; B. F. P. Zemla2, A. Mozdzierz1, J. Stojko1 Medical University of Silesia, School of Pharmacy, Department of Hygiene, Bioanalysis and Environmental Studies, KATOWICE, Poland, 2Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Gliwice Branch, Department of Cancer Epidemiology, GLIWICE, Poland This study is a further investigation of diet influence on the lung cancer incidence in male inhabitants of the Upper Silesia, an industrial region of Poland. Aim: The aim of the study was to estimate the influence of diet types on the lung cancer risk, including squamous cell and small cell carcinomas, in men in the Upper Silesia. Material and methods: In a case-control study, the studied population comprised 237 men with squamous cell carcinoma and 122 men with small cell carcinoma, and 649 healthy controls. We investigated high-vegetable diet (very big, big and medium vegetables), low-vegetable diet (small amount of vegetables, small and medium amount of meat), meat diet (frequently consumed big amount of meat), high calorie diet (fatty and very fatty diet), high calorie meat diet (fatty, big amount of fried meat,). Odd ratios (ORs) and 95% confidence intervals (CIs) were estimated by logistic regression adjusted for age and smoking. Results: Each studied diet, compared to high-vegetable diet, is a lung cancer risk factor in male. Low-vegetable diet causes over 5-fold increase in the lung cancer risk, poses threat to non-smokers and increases the risk of squamous cell carcinoma. Meat diet increases the risk almost 3 times, but the increase is significant only for smokers and squamous cell carcinoma. We observed an increasing trend of the risk with growing number of calories in food and it was especially strong in nonsmokers. The risk tends to increase in the following order: meat diet, low-vegetable diet, high calorie diet and high calorie meat diet. Conclusion: Diet low in vegetables, as well as meat and high calorie diets are significant risk factors for developing lung cancer, and its leading histological types in men. The association between diet type and lung cancer risk is more pronounced for squamous cell carcinoma.
136 THE CLINICAL AND EPIDEMIOLOGICAL STUDY OF 673 SALIVARY GLANDS TUMORS—16 YEARS ANALYSIS FROM THE HOLYCROSS REGION IN POLAND M. Zio´łkowska1, S. Bien´1, R. Me˛z_ yk2 1Department of Otolaryngology, Head and Neck Surgery, Holycross Cancer Centre, KIELCE, Poland, 2Department. of Information Technology, Holycross Cancer Centre, KIELCE, Poland The 673 cases of salivary glands tumors, diagnosed and treated in District Hospital and Holy Cross Cancer Center, Kielce—Poland from 1993 to 2008 (16 years) was taken into analysis. The collected material was coming from relatively small region (district) in Poland, with population of 1280000 inhabitants. Along the time of observation, the significant increases of morbidity, both in benign, and malignant group of tumors was observed. The distribution of
123
IEA-EEF European Congress of Epidemiology 2009 localization of tumors reflected the data usually quoted in literature (parotid gland—78.9%, submandibular gland—10.3%, and sublingual and small salivary glands—10.8%), but the ratio of malignant tumors in the analyzed material 148/608 (23.6%) was unexpectedly high. The significant difference in mean age between malignant group (62.5 years), and benign group of tumors (48.1 years) was recorded. The authors postulate necessity of multicenter studies concerning the epidemiology of salivary gland tumors—the malignant, as well as the benign, and multicenter collection of tumors of a relatively rare pathology type. The reason to include a group of benign tumors into a routine epidemiological survey is a tendency of malignant transformation of some benign pathological types of salivary gland tumors.
211 FALSE POSITIVE VS. DETECTION RATE ON BREAST CANCER SCREENING IN SPAIN ACCORDING TO AGE OF STARTING SCREENINGS (45–49 VS 50–54 YEARS OLD) D. Salas1, J. Iba´n˜ez1, J. Velarde2, R. Roma´n2, M. Sala2, N. Ascunce3, R. Zubizarreta4, X. Castells2, grupo investigacio´n RAFP 1DGSP Conselleria Sanitat, VALENCIA, Spain, 2SEEC IMIM H del Mar Barcelona CIBER Epidemiologı´a SP, BARCELONA, Spain, 3SDP Instituto Salud Pu´blica Navarra, PAMPLONA, Spain, 4SPPC Conselleria Sanidade Galicia, SANTIAGO DE COMPOSTELA, Spain Background: 10 Breast cancer screening programmes participate in a study to estimate the aggregate probability for presenting false positive (FP) in mamographic screenings. Objective: To compare false positives and detection rate according to age of starting screenings (cohorts: 45–49/50–54 years old). Method: 4,778,613 mamographic screenings were perfomed between 1990 and 2006. The descriptive analysis compares the false positive rate of additional valuation (FPRAV), invasive tests (FPR-IT), detection rate (DR), as well as the predictive value for AV (PV-AV) and IT (PV-IT) of the two age cohorts. Initial (IS) or sucessive (SS) screenings were differentiated. The chi2 values the dependence ratio. Results: On the IS, the FPR-AV is 11.06% for the youngest group and 7.38% for the 50–54 years’, with FPR-IT of 1.09% and 0.97%. The DR is 3.2 and 3.7 per thousand women on each group with a PV-AV of 2.63 and 4.43, and a PV-IT of 20.72 and 25.44 on both cohorts. On the CS, the FPR-AV is 7.58% and 4.72% being the FPR-IT 0.41% and 0.31% with a DR of 2.4 per thousand cases on both cohorts. The PV-AV is 2.91 and 4.61, and PV-IT is resp. 34.33 and 41.08. All differences are statistically significant with a P \ 0.000 except for DR on SS. Conclusion: The screening main adverse effect is the FPR-IT, with little difference between both groups. The expected benefit depends on the DR and is quite similar for both groups, being of reduced significance on the SS. The biggest difference seen between groups was on the FPR-NIT. The aggregate probability analisys will allow to check whether these differences on the FPR rise or decrease on time.
221 FROM LOW SECRETION OF MELATONIN TO CANCER—POTENTIAL MOLECULAR MECHANISMS A. Bukowska, J. Marjan´ska, B. Pepłon´ska Department of Occupational and Environmental Epidemiology, Nofer Institute of Occupational Medicine, LODZ, Poland Background: Melatonin is a pineal gland hormone synthesised and secreted predominantly in the night. Epidemiological studies suggest
IEA-EEF European Congress of Epidemiology 2009 that work at night and circadian cycle disruption might be associated with increased risk of breast cancer and it was suggested that this might be partially explained by decreased melatonin production. Objectives: To review published data on the potential role of melatonin in inhibition and growth of cancer. Methods: The available articles published by December 2008 included in MEDLINE database were searched. Search terms included: melatonin, pineal gland, circadian cycle and cancer. Results: A lot of research have been performed on the role of melatonin since its discovery in 1958, including its antineoplastic effect. In vitro and animal experimental studies have shown that melatonin interacts with nuclear and membrane receptors that can lead to inhibition of tumour growth. Melatonin was shown to be an effective antioxidant, scavenging reactive oxygen species and protecting cells from oxidative damages. It was also reported that melatonin has ability to modulate the immune system, by regulating secretion of interleukin 4 (IL4) and interleukin 2 (IL2) and by increasing activity of NK cells and other cells of immune system. Some studies indicates that melatonin might also increase p53 and p21 proteins levels leading to inhibition of cell cycle and division of cancer cells. Reduction of tumour growth after melatonin administration was also reported. In humans, lower levels of melatonin have been observed in night shift workers when compared to day workers. Some epidemiological studies reported higher levels of estradiol and other hormones that was suggested to be a result of inhibited melatonin production. Conclusions: Although experimental studies provide clear evidence on the antineoplastic role of melatonin, still the data in humans are limited. More epidemiological research is warranted on biological effects in humans associated with low melatonin levels.
240 NIGHT SHIFT WORK AND RISK OF CANCER—A LITERATURE REVIEW J. Marjan´ska, A. Bukowska, B. Pepłon´ska Department of Occupational and Environmental Epidemiology, Nofer Institute of Occupational Medicine, LODZ, Poland Background: About 15–20% of the occupationally active people in Europe and the USA work at shift work that involves night work. Such work organization is the most prevalent in the health care, industrial manufacturing, mining, transport, communication, leisure and hospitality sectors. Shift-work that involves circadian disruption has been recently classified by the International Agency for Research on Cancer to Group 2 A—probably carcinogenic to humans based on definite evidence in animal studies and limited in human. Objectives: The objectives of this work was to systematically review epidemiological studies addressing an association between night-shift work and cancer risk. Methods: The available data from MEDLINE database were reviewed. Search terms included: shift work, night shift work and risk of cancers, melatonin and cancer. Results: The majority of the epidemiological studies performed up to date focused on association between shift work and breast cancer risk. Of the reviewed studies on breast cancer, two were register-linked case–control studies, two population based case–control studies, one population based register-linked case–control study, two prospective cohort studies and one population based register-linked cohort study. In only one out of eight studies no association was found while in others moderate increase (up to OR = 2.2; 95%CI 1.1–4.5 in nurses in Norway with [30 years of night shift work) were found. The increased cancer risk was reported in nurses, radio-telephone operators, graveyard shift workers and flight attendants. Few studies reported association between risk of colorectal, prostate, thyroid cancer and non-Hodgkin lymphoma. Conclusion: The epidemiology review suggest an association between work at night and cancer risk. A number of
65 limitations of studies including imprecise exposure assessment performed up date preclude definite conclusions. Further epidemiological studies on the populations of workers employed at night shift work are warranted.
Cardioepidemiology I 183 ASSOCIATION OF BODY MASS INDEX AND CARDIOVASCULAR EVENTS IN PATIENTS AT RISK FOR OR ESTABLISHED CORONARY HEART DISEASE J. Mu¨ller-Nordhorn1, H. Englert2, K. Wegscheider3, F. Sonntag4, H. Vo¨ller5, W. Meyer-Sabellek6, E. Windler7, H. A. Katus8, S. N. Willich6 1Berlin School of Public Health, Charite´ University Medical Center, BERLIN, Germany, 2University of Applied ¨ NSTER, Germany, 3Department of Medical Biometry Sciences, MU and Epidemiology, University of Hamburg, HAMBURG, Germany, 4 Cardiology Practice, HENSTEDT-ULZBURG, Germany, 5 ¨ DERSDORF, Rehabilitation Center for Cardiovascular Diseases, RU Germany, 6Institute for Social Medicine, Epidemiology and Health Economics, Charite´ University Medical Center, BERLIN, Germany, 7 Department of Internal Medicine, University of Hamburg, HAMBURG, Germany, 8Department of Cardiology, University of Heidelberg, HEIDELBERG, Germany Objective: Body mass index has been shown to be associated with risk of cardiovascular events in patients with established coronary artery disease. The objective of the present study was to assess the association between body mass index and cardiovascular events in both high-risk patients as well as in patients with coronary heart disease. Methods: In the ORBITAL (Open-label primary care study: Rosuvastatin-Based compliance Initiatives linked To Achievement of LDL goals) Study, patients with hypercholesterolemia were included if they had an indication for statin therapy according to the Joint European Guidelines. Patients were enrolled consecutively in 1961 primary care practices in Germany. Cardiovascular events were defined as myocardial infarction, stroke, or cardiac interventions during the 12-month follow-up. Results: A total of 7640 patients were included, with 5.2% having a cardiovascular event during follow-up. After adjustment for lipid-lowering, hypertensive and antidiabetic medication, risk of event was significantly and inversely associated with body mass index (relative risk [RR] 0.96 per unit increase, 95% confidence interval [CI] 0.93–0.99). Others variables associated with an increased risk were male gender, hypertension, decreased physical and mental Short Form (SF)-12 summary scores, and established coronary heart disease. There was no significant interaction between risk stratum and body mass index. Conclusion: The risk for cardiovascular events is inversely related to body mass index in both highrisk patients and patients with established coronary heart disease. This effect is independent from any medication received.
187 PHYSICAL ACTIVITY AND CORONARY HEART DISEASE: INTERVENTION AT THE COMMUNITY LEVEL G. Saatli1, G. Ergor1, B. Unal1, O. Aslan2 1Dokuz Eylul University, Department of Public Health, IZMIR, Turkey, 2Dokuz Eylul University, Department of Cardiology, IZMIR, Turkey Background: It is known that Physical Activity (PA) reduces Coronary Heart Disease (CHD) and risk factors. CHD risk for individuals are being estimated using Framingham Equation in Izmir, Balcova
123
66 district in an ongoing project named Balcova’nin Kalbi (BAK). Objectives: This study was aimed to reduce CHD risk and risk factors by an intervention in PA at community-based level. Methods: Participants were individuals whose CHD risk was assessed as intermediate level in the BAK project. People who met the criteria were randomized according to the streets on which the participants were living as an intervention (n = 409) or control (n = 393) groups. Results: For the first PA programme, people were invited by telephone for informing about PA programme. Fifty-two people participated in the programme. PA programme was planned twice per week for six weeks with a physical educator. Forty-one people completed this programme and their measurements for blood presure, height/weight, glucose, trigliserid, HDL, LDL, total cholesterol were made. In the evaluation of these parameters by using paired t test, there was significant difference in glucose, trigliserid, HDL and CHD risk score. The second PA programme invited participants by door to door visits of interwievers. Forty-nine people started the second PA programme and upon completion, their measurements were also taken. In the meantime, control group was also invited for measuring same parametres and 90 people has attended this evaluation. The analysis process of comparing intervention and control group measurments is ongoing. Conclusions: Encouraging the community to participate and continue the PA programme were the major difficulties in this intervention. Regardless of the obstacles this was the first intervention in Turkey which aimed to reduce CHD risk factors by a PA programme at the community level. The preliminary results show that there is a significant difference in some risk factors of CHD after a six weeks PA programme.
193 RAISING THE EFFECTIVENESS OF THE NATIONWIDE CVD PREVENTION PROGRAM IN POLAND. THE POLKARD PROJECT A. Pajak1, G. Broda2, W. Drygas3, Z. Gasior4, M. Janion5, E. Kawalec1, K. Szafraniec1, A. Szpak6, B. Wizner1, R. Wolfshaut-Wolak1, T. Zdrojewski7 for the POLKARD study group 1 Jagiellonian University Collegium Medicum, KRAKOW, Poland, 2 Institute of Cardiology, WARSAW, Poland, 3Medical University in Lodz, LODZ, Poland, 4Silesian Medical University, KATOWICE, Poland, 5Voivodhip Hospital in Kielce, KIELCE, Poland, 6Medical Academy in Bialystok, BIALYSTOK, Poland, 7Medical University in Gdansk, GDANSK, Poland Cardiovascular Disease (CVD) is a major public health burden in Poland. CVD mortality has been declining since 1991 but it is still high and causes a life expectancy gap between men and women. Results from large population surveys indicated that control of CVD risk factors is poor. National Primary Prevention Program (NPPP) aims active identification of high risk persons in general practices and providing them advising given by general practitioners. One set of the goals of the POLKARD Project (supported by Polish Ministry of Health) for the years 2007–2008 was: 1) to develop a system to control the effectiveness of the NPPP, and 2) to develop an implementation project for structured intervention. Studied group and study design: in all 16 Polish voivodships (provinces) 2–5 clinics were recruited to the program. Half participated in the NPPP (active clinics) and half did-not (non-active clinics). Random sample of registered persons at age 35–55 years was selected in each clinic (5700 persons in total). For each patient observations based on the review of medical documentation were collected using a standard questionnaire. Patients were invited for examination including interview, anthropometric measurements and assessment of blood pressure and blood lipids and glucose. Results: In the medical documentation, information on risk factors was missing in significant proportion of
123
IEA-EEF European Congress of Epidemiology 2009 patients. Lifestyle advising was more frequent in the active clinics. At examination, differences between active and non-active clinics in the proportion of persons reaching the goals for treatment were not significant. Lessons: (1) results of the study suggest that national prevention program is less effective than intervention in the research trials, (2) there is a need to implement more intensive, structured intervention program which would involve not only doctors but other medical occupational groups.
201 ASSOCIATION BETWEEN URBAN LIFE YEARS AND CARDIO-METABOLIC RISK: THE INDIAN MIGRATION STUDY S. Kinra1, L. Andersen1, Y. Ben-Shlomo2, L. Bowen1, D. Prabhakaran3, L. Ramakrishnan4, K. S. Reddy5, G. Davey Smith2, S. Ebrahim1 1London School of Hygiene & Tropical Medicine, LONDON, United Kingdom, 2University of Bristol, Department of Social Medicine, BRISTOL, United Kingdom, 3Centre for Chronic Disease Control, NEW DELHI, India, 4All India Institute of Medical Sciences, NEW DELHI, India, 5Public Health Foundation of India, NEW DELHI, India Background: Urban living is associated with an increase in cardiometabolic risks, but the speed at which these risks are accrued over time and their patterning by personal and socio-demographic circumstances is not known. Objectives: To investigate the hypotheses that: (1) cardio-metabolic risks increase linearly, and at the same speed, with time spent in an urban environment, and (2) the trajectories of these risk accruals do not vary substantially by age, marital status, household structure, occupation and socio-economic position of the participants. Methods: A cross-sectional sib-pair design was used. Migrant workers and their spouses from four factory sites in north, central and south India, together with their non-migrant rural siblings, were surveyed. Risk factor prevalence was examined by years of urban living estimated from lifetime residential history. Results: Data on 4,223 participants (39% women; mean age 41 years) were available. Body fat increased rapidly in the first decade, while blood pressure and insulin rose linearly and gradually over four decades of urban living. In men, body fat increased by 4.5% in the first decade and 0.4% per decade thereafter; systolic blood pressure rose by 1 mm Hg and fasting insulin by 1.1 mU/L in each of the four decades. Similar but attenuated patterns were noted in women. Men in manual occupations showed particularly stronger rises in cardiometabolic risk. Conclusion: Body fat increases rapidly on first moving to an urban environment, and translates into gradual but ongoing rises in other cardio-metabolic risks. Public health interventions should focus on control of obesity in newer migrants to urban areas, particularly addressing men engaged in manual occupations.
30 KNOWLEDGE, OPPINIONS, AND BMI, IN A SAMPLE OF ROMANIAN ADULTS C. A. Zugravu1, C. Rada2, G. Nitu1 1University of Medicine and Pharmacy ‘‘Carol Davila’’, BUCHAREST, Romania, 2 The Institute of Anthropology ‘‘Francisc I. Rainer’’, BUCHAREST, Romania Background: Romania, as other European countries, is confronted with an explosion of overweight. Objectives: Our study searches some links between BMI, as indicator of the nutritional status, and lifestyle opinions and knowledge, using a descriptive cross sectional design.
IEA-EEF European Congress of Epidemiology 2009 Methods: A randomly selected sample of Romanians (n = 512), dwelling in southern part of the country responded at a 48 items questionnaire and the results were analyzed by SPSS program, having as independent variable the BMI. Results: The results showed a strong correlation with the personal level of education (P \ 0.01), obese having a lower level than the rest. The self perception body image is essentially a result of BMI (P \ 0.01), being correct for normally- and over-weighted, and incorrect for obese (generally describing themselves as overweighed) and under-weighted (thinking to have the ideal BMI (69.6%), probably because of the contemporary beauty trends). Obesity is strongly correlated with a low level of satisfaction regarding health and with quantifying it as ‘‘acceptable’’, but without linking health problems with weight surplus. Planned changes in personal diet are strongly linked with BMI (P \ 0.01) and have different reasons: normally- and under-weighted try to make it more healthy, overweighed, to loose weight, and obese, to improve their state of health. BMI is also strongly connected with willingness of spending more on healthy, natural food (P = 0.002) and on ecological ways of home waste disposal (recycling, P \ 0.01), with knowledge about different dietary fats(omega 3, saturated, trans, CLA, etc), about health claims on food labels, organic foods and GMF (all: P \ 0.001), and with the general interest in ecological problems (global warming) (P = 0.004). For all topics, obese have the lowest level of knowledge and interest. Conclusions: Obesity, apart of being a medical problem, is linked with different behavioral patterns showing frequently a low level of general education. Improving knowledge might be a useful tool in dealing with nutritional problems.
58 EMPOWERMENT AND OBESITY AMONG EGYPTIAN WOMEN A. Refaat College of Health Sciences, Walden University, MINNEAPOLIS, United States of America Background: Obesity is a worldwide health problem. Objective: the current study aims to identify the influence of women empowerment level on obesity status. Methods: the study used EDHS 2005 data to identify women empowerment status in three areas: 1- education & awareness; 2- work & decision making; and 3- protection from violence. Areas were computed from different variables and examined against obesity using significance tests and ANOVA for the mean BMI. Level of women empowerment was computed from the three areas to be tested as determinant of obesity. Results: Mean BMI of the studied women was 29.5. Obesity (BMI C 30) reached 42% among Egyptian women, with statistically significant highest Mean BMI among older age (32.8), urban (31.0), widows (31.2) and richest ones (31.4). Composite variables of education & awareness showed higher BMI among women educated up to primary level (30.2) and those who has color TV or satellite dish (30.1) with absence of obesity from those with education and awareness. Composite variables of work and decision making showed higher BMI among women working in clerical positions (31.5); and paid in cash (30.6) only, meanwhile there was no obesity levels among those working and sharing in decision making. On the other hand, highest BMI was found among women exposed to multiple types of physical abuse (34.0) or emotional domestic violence (30.8) with absence of obesity among those protected from violence. Most of the studied women (73%) were moderately empowered with 17% only who could be considered highly empowered. There were no obesity levels among different levels of empowered women. Conclusion: the present study showed absence of obesity levels from empowered women with their different composing areas. However, it showed that Egyptian women tend to be overweight and there are groups that need change in life style to avoid obesity.
67
Epidemiology of elderly 41 STUDY OF INSOMNIA SYNDROM IN KAHRIZAK NURSING HOME F. Mousavi1, B. Golestan2 1Islamic Azad University Tehran Medical Branch, TEHRAN, Iran, 2Tehran Medical University, TEHRAN, Iran Objectives: 1. Describe the prevalence of insomnia symptoms and syndrome as defined by the DSM-IV-R. 2. Identify the relation of insomnia with other characteristics of long term care residents. Material, Method: This is a cross-sectional study that was carried out on a group of elderly people (n = 1200), living in Kahrizak nursing home. Finally after data accusation from aforementioned number of subjects, due to some healthcare file corruption, a sum of 772 subjects became eligible for statistical analysis. Individuals were eligible if they were 65 years or older and living in public nursing home. The data collected came from 2 sources: 1. Structured interview with 2 nurses who knew the resident, 2. Systematic review of the healthcare files. The questionnaire was composed of 5 main parts: 1. Background questions, 2. Medical history, 3. Translated Voyer questionnaire, 4. Translated MOSES questionnaire, 5. Insomnia symptoms (DSM-IVR). Results: Subjects’ age ranged from 65 to 107 years with the mean of 76.7 years, 248 (32.1%) were male and 524 (67.9%) were female. Subjects’ residence times ranged from 1 week to 299 month with the mean of 47.2 month.302 (39.1%) of the subjects were involved with insomnia syndrome. Moreover 432 (56%) had delay in sleeping, 357 (46.2%) had difficulty maintaining sleep, 362 (46.9%) complained from waking up during the first hours in the morning and 313 (40.5%) reported fatigue and sleepiness during the day. Conclusion: Results of this study showed a significant increase of insomnia cases than the report of United States of America (39.1% vs. 6.2%). Insomnia related factors are nearly the same as American study. Insomnia predictive factors were: Aging of old subjects, Dependence decrease, Cognition status decrease.
214 IN-HOSPITAL AND AT THREE MONTHS MORTALITY AMONG ELDERLY WITH HIP FRACTURE: A STUDY IN CENTRAL ITALY E. Castronuovo, P. Pezzotti, D. Di Lallo Agency For Public Health of Lazio Region, ROME, Italy Background: Among elderly, hip fractures represent one of the most important causes of morbidity and mortality. Mortality is often described during hospital stay while few studies evaluated this outcome after discharge. Objectives: Aim of the study was to evaluate elderly mortality associated to hip fracture during hospital stay and from discharge to 3 months and to estimate the effect of selected clinical and hospital characteristic on early (in-hospital) and after discharge mortality. Methods: We used regional hospital discharge database to identify hip fracture cases between January and September 2006, aged 65 years and more, residents in Lazio Region. A deterministic record-linkage procedure between the cohort and regional death registry was applied for mortality follow-up. The Kaplan-Meier method was used to evaluate survival curves. Cox regression analysis was used to compute crude and adjusted relative risks for in-hospital and after discharge (3 months) mortality. As predictors we considered age, gender, social condition, education degree, Charlson Comorbidity Index, treatment (conservative or surgery), hospital volume (admission per year). Results: 4.795 patients enrolled, the 79% percent were females; the median age was 82; 10% had two or more comorbidities. Fifty percent had a conservative treatment, 12% had a delay surgery
123
68 less 3 days and 73% more than two days. In- hospital mortality was 5.2%; mortality at 30, 60 and 90 days from discharge was respectively 5%, 8%, 9%. The in hospital and at 3-months after discharge mortality models showed an increased risk associated to: increasing age and number of comorbidities, male gender and conservative treatment. Conclusion: Our study shows a important mortality also after hospital discharge. The result of multivariate analysis indicates that the effect of strong predictors on early mortality, such as conservative treatment, still remain after hospital discharge suggesting the need of appropriate follow-up of these frailty subject by outpatients services.
232 ORAL HEALTH IN NON-INSTITUTIONALIZED ELDERLY WITH DIFFERENT FRAILTY LEVELS
IEA-EEF European Congress of Epidemiology 2009 population aged 60 years and over, was followed prospectively from 2001 through 2003. Sleep duration was self-reported in 2001, and HRQL was measured using the SF-36 questionnaire in 2001 and 2003. Analyses were adjusted for the main confounders. Results: In comparison with women who slept 7 hours, those with extreme sleep durations (B5 or C10 hours) reported worse scores on the SF-36 physical and mental scales in 2001. Among men, sleeping B5 hours was associated with a worse score in the role-physical scale in 2001. The magnitude of most of these associations was comparable with the reduction in HRQL associated with aging 10 years. Sleep duration in 2001 failed to predict changes in HRQL between 2001 and 2003. Conclusions: Extreme sleep durations are a marker of worse HRQL in the elderly.
L. B. Rihs1, M. L. R. de Sousa1, M. J. D. Diogo2, M. H. Guariento2, F. A. Cintra2, A. L. Ne´ri1 1Piracicaba Dental School, State University of Campinas, PIRACICABA, Brazil, 2Medical School, State University of Campinas, CAMPINAS, Brasil
289 GEOGRAPHICAL DIFFERENCES IN THE TREND OF HIP FRACTURES INCIDENCE IN PORTUGAL
The oral health interferes directly in the maintenance of the general health, and consequently in the quality of life. Thus, this study evaluated the oral health conditions in a population of frail elderly. Methods and materials: The oral exams followed the recommendations from the World Health Organization. The non-institutionalized sample was regularly attended at a geriatrics clinic. The elderly were divided into two groups: prefrail and frail (Fried et al., 2001). The Mann Whitney test was used in the analysis of data. Results: The sample was composed of 65 elderly with 60 years of age or older, with a mean age of 77.5 years (SD = 8.1). Among them, 66.2% (n = 43) were edentulous; and the mean of present teeth was 3.5 (SD = 7.8) and the DMFT of 28.7 (SD = 5.2), and most of this index was composed of the missing component (93.5%) and the DF-R index (average of decayed and/or filled roots) was 0.31 (SD = 0.95). Among dentate individuals, the DMFT was 24.4 and the missing component 20.3 (on average). When a comparison was made between the dental variables, dentate individuals showed better oral conditions. In relation to the frailty classification, 49.2% (n = 32) were prefrail and 50.8% (n = 33) were frail. When the dental variables were compared with the frailty condition in the population, none of the conditions was different between the two groups. A total of 83.7% of individuals who used prosthesis needed to replace it. Conclusions: Thus, it could be conclude that there are gaps in the care for this group and to reverse this reality, it is necessary to target specific and comprehensive dental programs aimed to promote health and to control the caries disease.
C. Oliveira1, F. Pina1,2 1Instituto de Engenharia Biome´dica, PORTO, Portugal, 2Faculdade de Medicina da Universidade do Porto, PORTO, Portugal
258 SLEEP DURATION AND HEALTH-RELATED QUALITY OF LIFE AMONG OLDER ADULTS: A POPULATION-BASED COHORT IN SPAIN R. Faubel, E. Lo´pez-Garcı´a, P. Guallar-Castillo´n, T. Balboa, J. L. Gutie´rrez-Fisac, J. R. Banegas, F. Rodrı´guez-Artalejo Department of Preventive Medicine and Public Health, School of Medicine, Universidad Auto´noma de Madrid—CIBER in Epidemiology and Public Health, MADRID, Spain Background: The few studies that have addressed the association between sleep duration and health-related quality of life (HRQL) were cross-sectional and small-sized, targeted young and middle-aged persons, and did not adjust for the main confounders. Objectives: To examine the cross-sectional and longitudinal relationship between habitual sleep duration and HRQL in older adults. Methods: A cohort of 3834 persons, representative of the non-institutionalized Spanish
123
Background: Hip fractures are the most severe consequence of osteoporosis, and related with high mortality, disability, and social costs. Objectives: This study analyzed the incidents rates and the percentage of variations of hip fractures, between 2000 and 2004, by administrative divisions (NUT III) of Portugal. Methods: All patients aged 50? years from the National Hospital Inpatients Data Registers, with diagnosis of hip femur (code 820.x from ICD9.CM) caused by low impact were included. It was made an adjustment to the trend by linear regression for all NUT III and calculated the percentage of variations for the study period. The age standardized incidence rates were also calculated. Results: 40,441 hip fractures were recorded being 9,099 (22.5%) in men and 31,342 (77.5%) in women with a mean age (± SD) of 78 ± 10.0 and 81 ± 8.6, respectively for men and women. The overall incidence rates of hip fractures was 122 (95% CI: 120–125) per 100,000 person-years in men and 341 (95% CI: 337–345) in women and the women/men incidence ratio was 2.8. The total increase in these five years was 17.2% in men and 18.7% in women however the trends differed markedly by region. It was observed in some regions a decreasing in tendency. The increase was higher in inland regions compared to costal regions. Differences of variation in incidents rates, per NUT III, for men and women were found. Some regions with higher incidents rates showed a tendency of stabilization or decrease of incidence while regions with lower incidence rates showed an increased trend. Conclusions: The results showed, in some regions in Portugal, a tendency already reported in other international studies, that is, a stabilization or decrease in the incidence of hip fractures. Although not yet explained, this can be due to prevention actions, changes in life style, or simply a cohort effect.
168 SELF-DEFINING BEING A VICTIM OF ABUSE IN RELATION TO THE HEALTH-RELATED QUALITY OF LIFE OF OLDER PEOPLE. A CRACOW STUDY B. Tobiasz-Adamczyk, M. Brzyska, P. Brzyski, M. S. Kopacz Department of Medical Sociology, Chair of Epidemiology and Preventive Medicine, Jagiellonian University Medical College, CRACOW, Poland
IEA-EEF European Congress of Epidemiology 2009 Labeling theory defines older age as secondary deviance, which, in consequence, increases the risk of different forms of abuse, predominantly from one’s family environment . Material and methods: This study examined the relationship between how being a victim of different types of abuse was defined by 290 older women and how this definition influenced their health-related quality of life (HRQoL). Data was initially collected in the homes of 456 older people using face to face interviews concerning their perception of different forms of abuse, as experienced by older acquaintances of the respondent, and their own self-definition of being a victim, based on experiences limited to older age. HRQoL was measured using the SF-36 test. Particular dimensions of HRQoL were defined as low provided their score was below the median. Risk of lower levels of HRQoL was done with logistic regression modeling using SPSS 15 for Windows. Results: The risk of restricted functional status due to one’s emotional state was 2.85 times greater in women who feel neglected by their family (95%CI = (1.29; 6.28)). The risk of restricted functional status due to pain is 3.29 times greater in women who experienced physical abuse in older age (95%CI = (1.12; 9.7)). The risk of rating one’s psychological health as low was 2.7 times greater in women who self-defined themselves as victims of psychological abuse perpetrated by their family in older age (95%CI = (1.30; 5.66)). Conclusions: The self-definition by older women of being a victim of different forms of abuse in older age significantly decreasing their HRQoL.
169 CHANGES IN THE EMOTIONAL STATUS OF OLDER PEOPLE AS A CONSEQUENCE OF FEELING NEGLECTED. A CRACOW STUDY M. Brzyska, B. Tobiasz-Adamczyk, P. Brzyski, M. S. Kopacz Department of Medical Sociology, Chair of Epidemiology and Preventive Medicine, Jagiellonian University Medical College, CRACOW, Poland The relationship between feeling neglected by family members, as selfreported by older people, and emotional status is very important for evaluating general health-related quality of life (HRQoL) in older age. This study examined this relationship as a predictor of general HRQoL. Materials and Methods: The study was conducted in a simple random sample of 456 older people, including 290 (63.7%) women. Face-toface interviews were performed among older citizens of Cracow, focusing on feelings of different forms of neglect, in relation to characteristics of their social network and family environment. HRQoL was measured using the SF-36 test. Particular dimensions of HRQoL were defined as low provided their score was below the median. Risk of lower levels of HRQoL was done with logistic regression modeling using SPSS 15 for Windows. Results: Almost 75% of those sampled were aged B75 years, 42.4% were currently married, and 38.1% widowed. Approximately half of the sample interacted socially with 1–3 individuals weekly (i.e., not co-resident) and *10% were socially isolated. The risk of restricted functional status due to one’s emotional state is 2 times greater in women (95%CI = (1.28; 3.12)), though lower among those who interacted socially with more than 4 individuals weekly (Exp(B) = 0.60; 95%CI = (0.39; 0.94)). The risk of rating one’s psychological health as low was 1.66 times higher in individuals who reporting that older people in Poland are not treated well (95%CI = (1.08; 2.56)). This risk is 1.88 times higher in women (95%CI = (1.23; 2.87)), and approximately 1.7 times higher in individuals reporting a secondary or lower level of education (95%CI = (1.01; 2.86)). Conclusions: Data confirm the significant relationship between feelings of neglect and emotional status as a predictor of general HRQoL.
69
Health services research I 158 THE QUALITY OF COLLABORATION AMONG MEDICAL RESEARCH CENTERS, UNIVERSITIES, HEALTH EXECUTIVES AND THE COMMUNITY IN IRAN S. Asefzadeh, Qazvin University of Medical Sciences, Health Management & Economics, QAZVIN, Iran A health research system is defined as a system for planning, coordinating, monitoring and health research aimed at effective and equitable national health development. This study was conducted to describe the quality of mutual collaboration among medical research centers, university, executive organization and the community. In this situation analysis qualitative methods were applied. In-depth interviews were held with the heads of 20 research centers (RCs), 6 senior executive health administrators at national level, 10 vice chancellors at university and 3 directors of private medical research firms. In addition to the interviews, records and evidence were reviewed. The findings reveal that 86% of the RCs’ annual budget is secured by public sources. The collaboration among the RCs, university and organization is informal rather that formal. The cooperation between RCs and university in policymaking, priority setting and knowledge production and utilization is stronger than that with other organs. The RCs have weak relationships with the community. About 20% of them have no relation with the private sector. The collaboration of RCs with the executive organizations and managers is also weak, especially in knowledge utilization and priority setting. It is concluded that the policies of medical research centers should be reoriented towards the basic health problem of the country facing executive organizations as well as the community. The collaboration between RCs and university, the private sector and NGOs should also be reinforced through networking and improvement of health information systems.
155 LIFESTYLE COUNSELLING IN PRIMARY HEALTH CARE—A SYSTEMATIC REVIEW ON ATTITUDES, KNOWLEDGE AND PRACTICE OF PHYSICIANS REGARDING THE PREVENTION OF CARDIOVASCULAR DISEASE C. Huy1, C. Diehm2, S. Schneider1 1 Mannheim Institute of Public Health, Social and Preventive Medicine, University Medicine Mannheim, Heidelberg University, MANNHEIM, Germany, 1 Department of Internal Medicine, Clinical Center and Academic Hospital of the University of Heidelberg, KARLSBADLANGENSTEINBACH, Germany Background: Modifiable risk factors like overweight, unhealthy diet, physical inactivity and smoking account for about 50% of cardiovascular disease (CVD), the world’s leading cause of death. Preventive measures including a healthy lifestyle can effectively reduce the risk of CVD. Primary care physicians (PCPs) play an important role in helping patients change their lifestyle. However, little is known to date about the knowledge, attitudes and practice of PCPs regarding lifestyle counselling. Objectives: This study provides a comprehensive overview on this topic from the physician’s perspective. Our objectives were to analyse practice patterns and to determine causes for possible deficits in CVD prevention: lack of knowledge, lack of motivation or other barriers. Methods: A search for literature published until 2008 was carried out in Medline,
123
70 BIOSIS Previews, PsycINFO and the CSA database. 57 out of 436 identified articles met the following inclusion criteria: (a) PCPs; (b) CVD prevention or health promotion; (c) surveys without intervention; (d) reports on attitudes, knowledge and/or practice of physicians. Results: Results from 36,202 PCPs were included in this review. Studies were mainly conducted in the U.S. (n = 25) and Europe (n = 25), and achieved response rates of 6–100%. There were hardly any studies from developing countries. PCPs’ knowledge and practice of lifestyle counselling varied widely across the studies. Overall, PCPs had strong beliefs and interests in health promotion. However, multiple barriers existed that prevented physicians from lifestyle counselling, e.g. patients’ compliance, lack of physician’s confidence, financial or organizational obstacles, lack of time or training. Conclusions: Although guidelines on CVD prevention are available worldwide, there is considerable variation in the knowledge and practice of PCPs. Broader dissemination of guidelines, education of physicians and patients, and better organization of services may improve the implementation of these guidelines. Further studies are needed to explore this issue in developing countries.
253 HIV/AIDS-RELATED KNOWLEDGE, PROFESSIONAL RISK AND PRACTICE OF HEALTH CARE PERSONNEL IN THE CITY OF NIS, SERBIA B. Kocic, B. Petrovic University of Nis, School of Medicine, Institute of Public Health, Department of Epidemiology, NIS, Serbia Background: Health care workers are key players in the prevention and management of HIV infection. Deficiencies of health care workers’ knowledge with regard to HIV may influence practice and reduce compliance with infection control recommendations. Objectives: To assess HIV/AIDS-related knowledge, professional risk and practice of health care workers in Nis. Methods: A cross-sectional study of 222 health care personnel was performed. The subjects were selected through two-stage stratified sampling (according to the number and occupation of the personnel in the health care strata). The data were collected by an anonymous questionnaire. MantelHaenszel chi-square testing was applied. Results: Medical personnel had deficient knowledge about the modes of HIV transmission. The majority of true answers were found relating to HIV transmission by blood, sperm/vaginal secretions, and sweat. The majority of false answers were found relating to HIV transmission by saliva/tears, urine, and feces. The group of doctors/dentists didn’t have generally better knowledge than the other group of nurses and laboratory technicians. Within the last year, 58% had skin contact with patient’s blood, 50% reported needlestick injuries, 80% mentioned cuts from sharp instruments and 34% reported contact of eye and other mucosa with patient’s blood. Skin contact with patient’s blood (P = 0.031), needlestick injuries (P = 0.009) and the cuts from sharps (P = 0.003) occurred most frequently in the group of nurses and laboratory technicians. One third of health personnel with accident asked for medical treatment. Almost 90% perceived high professional risk of acquiring HIV infection. Nearly two thirds did not use appropriate barriers during medical procedures with patients. If the HIV status of a patient was known, nurses and laboratory technicians had used additional personal protection more frequently than doctors and dentists (P = 0.001). Conclusions: Well-coordinated continuing education of HIV/AIDS for all categories of medical staff is recommended.
123
IEA-EEF European Congress of Epidemiology 2009
42 ASSESSMENT OF MEDICAL INTERNS’ ATTITUDES TOWARDS THEIR FUTURE CAREER F. Mousavi1, B. Golestan2, A. F. Nejad3 1Department of Community Medicine, Medical School, Islamic Azad University, Tehran Medical branch, TEHRAN, Iran, 2Department of Epidemiology & Biostatistics, School of Public Health, Tehran University of Medical Sciences, TEHRAN, Iran, 3Department of Medical Records, School of Allied Medicine, Tehran University of Medical Sciences, TEHRAN, Iran Objective: The objective of this study was to look at the factors affecting medical students’ attitudes towards their future career. Materials and Method: In a descriptive, cross-sectional research, 250 medical interns (94 males and 156 females) at the Islamic Azad University in Tehran were included in a study using an attitude questionnaire and demographic features. Their ages ranged from 22 to34 years with a mean of 25.98 years. They comprised all the medical interns studying medicine in the academic year 1384–1385 (2005–2006) in Qods clinic and Amir Al-Momenin, Javadiyeh, Bou-Ali, Javaheri, Torfeh, Kashani, Lavasani, and Naft hospitals. The data were analyzed by independent t-test, ANOVA and multiple regression. Results: The results showed that the total average attitude score was 45.36 (SD, 9.37) with a range of 19 to 78. The comparison of attitude scores in total and in each domain in terms of the background characteristics indicated that total attitude score differed with respect to the mother’s educational level, while attitude towards medical career varied in terms of sex, being engaged in a non-medical job and willingness to practice medicine after graduation. Attitude towards economic and working hours did not seem to be affected by the background variables. Conclusion: Negative attitudes towards their economic status, the future of medical practice, working hours, and obligatory services in deprived areas are among the factors which diminished the medical students’ motivation to practice medicine in the future.
Methods 148 MAPPING DRUG USERS AMONG VENETO REGION IN 2006 V. Lorenzoni, S. Salvatori, S. Molinaro, M. Urcioli Scalese, R. Potente, F. Mariani National Council of Research, Institute of Clinical Physiology, PISA, Italy Background: In epidemiological field knowing the spatial distribution of a phenomenon is interesting to understand its pattern in the area of interest but also to verify if there are factors related to the phenomenon under investigation. Analyzing drug addictions in these terms can help us to investigate the presence factors of ‘‘exposure’’ and can be useful to planning and/or reorganize Drug Addiction Services. Objectives: Our aim is to analyze spatial distribution of drug users distinguishing from opiate, stimulants and cannabis users living in Veneto region and in treatment to private and public Drug Addiction Services during 2006, in order to see if there are differences from different drugs users and to verify association with ecological covariates. Methods: Anonimous data were providede by private and public Drug Addiction Services of Veneto region. Age and sex adjusted prevalence were calculated for each drug and for each municipality. Estimated prevalences were also obtained using a hierarchical bayesian model that allow us to account for different sources of variability: clustering, heterogenity and ecological covariates (in this case altitude and population density). Trought this model we also calculated a posterior probability which allow us to compare estimated prevalence for each municipality with a reference value (region mean value). Results: For opiates, municipalities in the
IEA-EEF European Congress of Epidemiology 2009 nearby of Venezia and Caorle and other spread all over the region showed prevalence greater than the reference value. With respect to stimulants users some municipalities around Venezia and in the nearby of the greater cities had prevalence significantly greater than the mean. Finally, for cannabis users, a prevalence significantly greater than the mean was shown from the municipalities all along the coast and in the neighbourhood of Verona. A negative significantly association was found between altitude and stimulants users prevalence (IC 95%: – 0.003 to –0.00004). Conclusions: Opiates, stimulants and cannabis users have a different spatial distribution. This is a simple analisys that would be enriched including other possible sources of variability (eg. services avaibility, socio-economics index) but also using data from several years to analyse both spatial and time trend of the phenomenon.
285 GEOREFERENCING DATA IN MICRO-AREAS: STILL A CHALLENGE FOR THE USE OF GIS IN HEALTH STUDIES A. Olhero1, M. de Fa´tima de Pina1,2 1Instituto de Engenharia Biome´dica, PORTO, Portugal, 2Servic¸o de Higiene e Epidemiologia, Faculdade de Medicina do Porto, PORTO, Portugal In the last two decades the use of GIS has spread and become popular among health professionals. The availability of digital maps makes the georeferencing of data in administrative areas an easy task, although, when it comes to micro-areas, and the need to deal with addresses, the problems persist along the decades. The objective of this work is to describe the georeferencing process of participants from two cohorts (EpiPorto and Epiteen) in Porto, Portugal. EpiPorto is a cohort of adults, selected by random dialing, leaving in Porto, and EpiTeen is a cohort of teenagers, borne in 1990 and studying in Porto schools. We used digital map with streets centerlines, and initial and final numbering of each segment. Address were standardized before the georeferencing process, which had three phases: automatic (GIS locating the addresses), semi-automatic (operator interacting with GIS), and manual. EpiPorto had 2421 registers: 0.2% were lost because of incorrect or incomplete addresses; 70.5% were georeferenced automatically and 13% semi-automatically. The remaining 16.3% were manually georeferenced. EpiTeen had 1597 registers: 1.6% were lost; 31.9% georeferenced automatically, and 17.4% semiautomatically. The majority (49.1%) were georeferenced manually. During manual process, 216 and 300 participants, respectively from EpiPorto and Epiteen were contacted by phone, in order to correct the addresses. Time consuming was much higher with the Epiteen georeferencing because the effectiveness of using the GIS tools in the automatic process was very low. The quality of the addresses supplied by adolescents was considerably worse than the addresses from the group of adults. This can be due the fact that teenagers nowadays are not familiar with their own addresses, by not using conventional mail, rather e-mails or mobile messages. Despite GIS evolution and availability of digital maps, it is still a slow and difficult process to georeference addresses, just like 20 years ago.
Perinatal epidemiology I 190 ASSOCIATIONS BETWEEN HABITUAL PERICONCEPTIONAL NUTRITIONAL INTAKE AND ADVERSE BIRTH OUTCOMES AMONG LOW SOCIO-ECONOMIC STATUS ETHNIC MINORITY WOMEN IN ISRAEL K. Abu-Saad1, I. Belmaker2,3, D. Fraser1 1S. Daniel Abraham International Center for Health and Nutrition, Department
71 of Epidemiology and Health Services Evaluation, Faculty of Health Sciences, Ben-Gurion University of the Negev, BEER-SHEVA, Israel, 2Ministry of Health, JERUSALEM, Israel, 3Department of Epidemiology and Health Services Evaluation, Faculty of Health Sciences, Ben-Gurion University of the Negev, BEER-SHEVA, Israel Background: Preterm birth (PTB), low birth weight (LBW) and small size-for-gestational-age (SGA) are adverse birth outcomes (ABOs) that may have life-long consequences. Inadequate perinatal and periconceptional maternal dietary intake has been related to adverse pregnancy outcomes, particularly in low income populations. Objectives: To explore the associations between the habitual, periconceptional nutritional intake of pregnant Bedouin Arab women (the lowest socio-economic status (SES) group in Israel) and adverse birth outcomes. Methods: Periconceptional diet was assessed prospectively in non-primipara Bedouin women with singleton pregnancies, using a short targeted food frequency questionnaire (TFFQ). The TFFQ was designed to measure the habitual pre-pregnancy intake of protein, iron, zinc, calcium, folate, and omega-3 fatty acids. Individual nutrient measures were combined to create a habitual multi-nutrient intake scale (H-MNIS) for ranking women by the quality of their overall habitual diet. Data on maternal anthropometrics, obstetric history, index pregnancy, SES indicators, and birth outcomes were abstracted from the medical records. Results: The TFFQ was administered to a sample of 404 participants, whose mean ± SD gravidity and parity were 4.2 ± 2.8 and 3.7 ± 2.5, respectively. Birth outcomes for the index pregnancy were 9% each for PTB, LBW, and SGA; and a total of 18% had one or more ABO. In multivariate logistic regression analysis controlling for possible confounders, women in the lowest H-MNIS quartile had a 7.1-times higher risk (95%CI: 2.3–22.0) of having any ABO than women in the highest quartile (P for trend = 0.001), and the RR for PTB alone was similar. Women in the lowest quartile of complete protein intake had a 7.0–times higher risk (95%CI: 1.7–28.1) of LBW than women in the highest quartile (P for trend = 0.006). Conclusions: In this low-income population, a better quality periconceptional diet was associated with reduced risk for ABOs. Interventions to improve the habitual diet quality of Bedouin women of childbearing age may reduce ABO rates.
88 LIFE WITH A NEW BABY: HOW DO IMMIGRANT AND AUSTRALIAN-BORN WOMEN’S EXPERIENCES COMPARE? M. Bandyopadhyay1, R. Small1, L. Watson1, S. Brown2 1Mother & Child Health Research, Faculty of Health Sciences, La Trobe University, Melbourne, Australia, 2Healthy Mothers Healthy Families Research Group, Murdoch Childrens Research Institute, PARKVILLE, Australia Background: Very little is known about immigrant mothers’ experiences of life with a new baby, with the exception of studies on maternal depression. Objectives: To compare the post-childbirth experiences of Australian-born and immigrant mothers from nonEnglish speaking countries, including: physical and emotional health; practical and emotional support; views about partners, friendships, and health services. Methods: Data from a postal survey of recent mothers at six months postpartum in Victoria, Australia, were compared for aspects of life with a new baby for two groups of immigrant women: those born overseas in non-English speaking countries who reported speaking English very well (n = 460); and those born overseas in non-English speaking countries who reported speaking English less than very well (n = 184), and Australian-born women (n = 9796). Results: Compared with Australian-born women, both immigrant groupings were equally confident in caring
123
72 for their baby and in talking to health care providers about health problems postpartum. They were more likely to be breastfeeding at six months postpartum than Australian-born women (61% women proficient in English & 63% women less proficient in English vs 52%; OR = 1.41, 95% CI 1.2–1.7 & OR = 1.65, 95% CI 1.2–2.3). No differences were found in anxiety or relationship problems with partners. However, immigrant mothers less proficient in English had a higher prevalence of depression (28.8% vs 15%, OR = 2.40, 95% CI 1.9–3.0) and were more likely to report wanting more practical (65.2% vs 55.4%) and emotional (65.2% vs 44.1%) support. They were also more likely to have no ‘time out’ from baby care (47% vs 28%, OR = 2.28, 95% CI 1.7–3.1) and to report feeling lonely and isolated (39% vs 17%, OR = 3.08, 95% CI 2.3–4.2) after birth. Conclusions: Counter to common stereotypes, not all the findings indicated increased vulnerability among immigrant women. Greater effort needs to be made to improve the responsiveness of health and other support services to meet individual needs of women less proficient in English.
101 HEALTH CARE PROVIDERS’ PERCEPTIONS OF BARRIERS AND FACILITATORS OF PRENATAL CARE AMONG INNER-CITY WOMEN IN WINNIPEG, CANADA M. Heaman1, W. Sword2, M. Moffat3, L. Elliott1, M. Helewa1, P. Gregory3, C. Cook3, L. Tjaden3, H. Morris1 1University of Manitoba, WINNIPEG, Canada, 2McMaster University, HAMILTON, Canada, 3Winnipeg Regional Health Authority, WINNIPEG, Canada Background: Previous research, primarily from the United States, indicates that inner-city women are less likely to receive adequate prenatal care. However, little research has been conducted on this issue in a Canadian context. Objectives: The purpose of this study was to explore health care providers’ perceptions of barriers and facilitators of prenatal care use among inner-city women, and to obtain suggestions regarding models for delivery of prenatal care. Methods: In this qualitative component of a mixed methods study of factors associated with inadequate prenatal care among inner-city women in Winnipeg, Canada, a purposive sample of 24 health care providers took part in individual in-depth interviews. Participants included obstetricians, family practice physicians, midwives, nurse practitioners and nurses. Interviews were transcribed and content analysis was undertaken to identify themes. Results: Barriers grouped into themes of program/service characteristics (e.g., inconvenient location, lengthy office wait, scheduling problems), caregiver qualities (too busy, judgmental), health care system issues (lack of providers), and personal factors (woman’s lack of understanding of importance of prenatal care, transportation difficulties, childcare problems, social issues such as substance abuse). Facilitators included tangible rewards, geographic proximity, individualized care, interpreter services, and respectful caregivers. Suggestions for improving prenatal care grouped into themes of improved accessibility and convenience (e.g., provision of transportation, more flexible hours), program characteristics (a multidisciplinary ‘‘one stop shop’’ approach, group prenatal care, specific prenatal services for special needs groups such as adolescents and substance abusing women), and health care system improvements (establish community based clinics, expand midwifery services, increase public awareness of prenatal care). Conclusions: Several barriers to prenatal care exist. These results can be used to develop more effective models of prenatal care for innercity women.
123
IEA-EEF European Congress of Epidemiology 2009
102 INNER-CITY WOMEN’S PERCEPTIONS OF BARRIERS, FACILITATORS AND MOTIVATORS RELATED TO USE OF PRENATAL CARE IN WINNIPEG, CANADA M. Heaman1, W. Sword2, M. Moffat3, L. Elliott1, M. Helewa1, P. Gregory3, C. Cook3, L. Tjaden3, H. Morris1 1University of Manitoba, WINNIPEG, Canada, 2McMaster University, HAMILTON, Canada, 3Winnipeg Regional Health Authority, WINNIPEG, Canada Background: Previous research, primarily from the United States, indicates that inner-city women are less likely to receive adequate prenatal care. However, little research has been conducted on this issue in a Canadian context. Objectives: The purpose of this study was to explore the perceptions of inner-city women regarding barriers, facilitators, and motivators related to prenatal care use. Methods: In this qualitative component of a mixed methods study of factors associated with inadequate prenatal care among inner-city women in Winnipeg, Canada, a purposive sample of 26 women took part in individual in-depth interviews during pregnancy or the early postpartum period. Participants varied in age, parity, marital status, ethnicity, and use of prenatal care. Interviews were transcribed and content analysis was conducted to identify themes. Results: Barriers grouped into themes of program/service characteristics (e.g., geographic distance, lengthy office wait, scheduling difficulties), caregiver qualities (too busy, not sharing information), health care system issues (lack of consistent caregiver, lack of providers), and personal barriers (limited awareness of prenatal care services, lack of motivation, lack of social support, no perceived need for prenatal care, travel difficulties, childcare issues, financial problems). Facilitators grouped into themes of accessibility and convenience (geographic proximity, minimal wait times), social support (encouragement to attend prenatal care), positive caregiver qualities, transportation assistance, provision of child care, and tangible rewards. Women reported being motivated to seek prenatal care by their desire for a healthy mother and baby, an opportunity for social interaction, personal values, gaining knowledge and skills, and positive past experiences. Conclusions: Inner-city women’s utilization of prenatal care may improve by overcoming barriers to care, facilitating their attendance, and capitalizing on factors that motivate them to attend.
151 SEXUAL BEHAVIOUR OF THE YOUNG WOMEN LIVING IN BELARUS E. Tishchenko, M. Surmach Grodno State Medical University, GRODNO, Belarus Demographic crisis and sharp change of social stereotypes were the reason of many researchers’ interest to the sexual behavior of youth. This work purpose was to study Belarusian young women sexual behavior, to establish its safety from the medical point of view. The most risky period for reproductive health is bound with intimate life beginning. For its characteristic data of 404 women of 15–24 years old those pregnant for the first time or just after their first resultive pregnancy, as well as 267 women of the same age interrupting the first pregnancy are analyzed. Sample was formed according to official data of Belarus youth territorial distribution. Profound studying of young women sex behavior is spent among last year students of one of medical high schools (100 persons, 22–27 years). It is taped, that more than a half of the first sexual contacts of the young Byelorussians having experience of pregnancy, were unsafe from the medical point of view. Partly the reason was insufficient availability of the sexological help (only a quarter of the young women interrupting
IEA-EEF European Congress of Epidemiology 2009 pregnancy, and more than a third of the women expecting a child consider, that can receive sexologist’s consultation in due time), low medical-hygienic literacy (only half of the city women, and less than a half of the villages women are correctly informed in the field of sexuality). Knowledge in the sphere of physiology of sexuality is necessary for harmony of sex relations. At the same time, data of medical high school students interrogation show, that the knowledge itself is not a guarantee of safe sexual behavior. General approval of behavioral stereotypes, negative from the medical and social point of view, have led to unsafe sex behavior of the young women.
Children and adolescent health epidemiology I 83 ASSOCIATION OF FAMILY, PEERS AND THE MEDIA WITH YOUTH’S SEXUAL BEHAVIOUR M. Ruiz-Canela1, S. Carlos2, C. Lo´pez2, E. Toledo2,3, A. Osorio4, M. Calatrava2, J. de Irala2 1Department of Biomedical Humanities, School of Medicine, University of Navarra, PAMPLONA, Spain, 2 Department of Preventive Medicine and Public Health, University of Navarra, PAMPLONA, Spain, 3Department of Preventive Medicine and Quality Management, Hospital Virgen del Camino, PAMPLONA, Spain, 4Department of Education, University of Navarra, PAMPLONA, Spain Background: campaigns to prevent premature sex are very important from a public health perspective. However, the success of these campaigns is highly dependent on how environment is influencing sexual behaviour among adolescents. Objectives: to examine the association between messages received from family, friends and mass media, and premature sex among adolescents. Methods: a cross-sectional study of a random sample of 3,259 students (from 13 to 19 years old) from El Salvador. Participants were obtained through multi-stage sampling of clusters of schools. An anonymous selfadministered questionnaire related to relationships, love and sexuality was distributed at schools. Multiple logistic regression analysis was used to identify independent factors associated with having sexual relationships. Results: 638 out of 2515 valid respondents (25.3%) referred having had sexual relationships already. After adjusting for age, sex, economic level, religiosity, non-sexual risk behaviours (such as smoking cigarettes, using drugs or alcohol and attending discos), and time spent on school or cultural activities, the following risk factors for experiencing sexual intercourse were identified: discussing sexual relationships (OR = 1.15, IC95% [0.98–1.35]) or contraceptives and the prevention of STI (OR = 1.30, IC95% [1.02–1.66]) with parents; friends as frequent source of information on love and sexuality (OR = 1.27, IC95% [1.02–1.60]); perceived liberal attitudes on parents (OR = 1.80, IC95% [1.18–2.75]) or friends (OR = 1.62, IC95% [1.25–2.12]; and supporting the belief that HIV risk equals zero whenever condoms are used in sexual intercourses (OR = 1.48, IC95% [1.02–2.16]). On the contrary, the following protective factors were identified: receiving sexual abstinence messages from friends (OR = 0.66, IC95% [0.50–0.87]) or brothers (OR = 0.64, IC95% [0.50–0.81]). Conclusions: Messages received from family, friends and mass media are probably important external factors influencing sexual behaviours of adolescents. The influence of these factors should be taken into account when implementing public health strategies to prevent premature sex.
105 INJURIES AMONG YOUNG PEOPLE: INTEREST OF POPULATION SURVEYS C. Senterre1,2, M. Dramaix1, A. Leveˆque2 1De´partement de Biostatistique, Ecole de Sante´ Publique de l’Universite´ Libre de
73 Bruxelles, BRUSSELS, Belgium, 2Unite´ de recherches Epide´miologie, Traumatismes et Maladies chroniques, Ecole de Sante´ Publique de l’Universite´ Libre de Bruxelles, BRUSSELS, Belgium Background: In the ‘‘Child Injury Prevention report’’ (2008), WHO declared that ‘‘rare are the countries which collected good descriptive data’’. WHO recommends to ‘‘reinforce the quality and the quantity of the available data for prevention’’. In Belgium, no injuries monitoring system exists. Some informations are however available in ‘‘Health Surveys by Interview’’ which are carried out every 4 years, at the national level, since 1997. Objective: The main objective is to document how the health studies can provide interesting epidemiologic informations about a specific problem among young. Methods: Data are collected using three elements: a household questionnaire, a ‘‘face-to-face’’ questionnaire managed by an investigator and an individual auto-administered questionnaire (only for the ‘‘C15 years’’). Injuries are one of the topics investigated. In 1997, this module was in the individual questionnaire then we don’t have data concerning the less than 15 years. Since 2001, this topic is included in the ‘‘face-to-face’’ questionnaire. Data concerning the ‘‘\ 20 years’’ are available for 2001 and 2004. Results: 10.6% of the 2609 ‘‘\20 years’’ questioned in 2001 had at least one injury implying a doctor visit (12 months recall). This proportion was 9.4% in 2004 (n = 2284). The 2004 data shows that 27.7% are school injuries and 29.7% are home injuries. The stratification by age group shows that the injury proportion increases with age and also that injury locations change. Conclusion: Information from population surveys requires to be supplemented. The link with hospital databases, for example, would make it possible to have details on the circumstances, the gravity and the evolution of injuries. These links, impossible currently from an individual point of view, are possible from a population point of view. Their realization requires a framing of dialogue which is poor developed in Belgium.
120 PREVALENCE OF SMOKING AMONG 14–17 YEAR OLD SCHOOLCHILDREN FROM THE DISADVANTAGED AREAS IN ARKHANGELSK, NORTHWEST RUSSIA O. A. Kharkova1, A. G. Soloviev2, A. M. Grjibovski1,3 1International School of Public Health, Northern State Medical University, ARKHANGELSK, Russia, 2Northern State Medical University, ARKHANGELSK, Russia, 3Norwegian Institute of Public Health, OSLO, Norway Background: Smoking is one of the major public health problems worldwide. The prevalence of adult smoking in Russia is very high: more than two thirds of Russian men and about 25% of women are daily smokers. Less is known about adolescent smoking. Objectives: to estimate the prevalence of adolescent smoking in a disadvantaged area in Northwest Russia. Methods: Altogether, 188 schoolchildren from all three schools in two most disadvantaged districts in Arkhangelsk aged 14–17 years filled in an anonymous questionnaire about their smoking habits and knowledge on smoking and its effects on health (response rate 82%). Results: Altogether, 52.5% of boys and 47.4% of girls ever smoked. 48% of boys and 38% of girls smoke daily. Among smokers, 11.0% of boys and 4.0% of girls smoke more than 20 cigarettes a day. On average, boys smoke 6–10 cigarettes a day while girls smoke 4–8 cigarettes a day. The main reported reasons for smoking initiation were ‘‘curiosity’’, ‘‘coping with stress’’, ‘‘facilitation of conversations’’ and that ‘‘smoking makes look older’’. The least popular answer was that smoking was prestigious. Among nonsmokers, the most popular reasons for not smoking were: ‘‘smoking is dangerous to health’’, ‘‘do not want to be like others’’, ‘‘do not like how
123
74 it smells/tastes’’, ‘‘my parents do not allow me to smoke’’ and ‘‘no money to buy cigarettes’’. Lung cancer, lung diseases, different cancers, bronchitis, dental diseases, and cardiovascular diseases were mentioned as direct effects of smoking. Decrease of lifespan, infertility, birth defects, memory problems were mentioned by only few responders. Conclusions: The prevalence of adolescent smoking in the study area is very high in both genders. The observed prevalence of smoking in girls is substantially higher than smoking in adult women in Russia in general indicating a considerable problem for women’s health in the future.
210 ARE THE CHILDREN A RELIABLE SOURCE OF INFORMATION ABOUT THEIR HEALTHRELATED QUALITY OF LIFE? E. Cichocka-Jarosz1, B. Tobiasz-Adamczyk2, P. Brzyski2, G. Lis1, U. Jedynak1, J. J. Pietrzyk1, K. Profus3, J. Lange4, M. Kulus4, E. S´wiebocka5, M. Kaczmarski5, T. Małaczyn´ska6, O. S´widzicka6, G. Bre˛borowicz7, Z. Kycler7 1Jagiellonian University Medical College, Chair of Pediatrics, Department of Pediatrics, PolishAmerican Children’s Hospital, CRACOW, Poland, 2Jagiellonian University Medical College, Chair of Epidemiology and Preventive Medicine, Department of Medical Sociology, CRACOW, Poland, 3 Out-patient Department of Developmental and Clinical Psychology, University Children’s Hospital of Cracow, CRACOW, Poland, 4 Department of Children Pulmonology and Allergy, Medical University of Warsaw, WARSAW, Poland, 5Białystok Department of Pediatrics, Children’s Gastroenterology and Allergology, Medical Univerisity, BIALYSTOK, Poland, 6Regional Center of Allergology, Pulmonology and Mucoviscidosis, GDANSK, Poland, 7Chair of Pediatrics Department of Pneumonology, Allergology and Clinical Immunology, Poznan´ University of Medical Sciences, POZNAN, Poland Background: Venom (wasp and honey-bee) allergy might be life threatening event with fatal result, which may influence children’s as well as their parents quality of life (QoL). There are some doubts regarding accuracy of QoL data acquired from children. Thus, proxyresponders (usually parents) are used as a source of information regarding children’s QoL. Aim: The aim of this study was the estimation of concordance between QoL data obtained from children themselves and their parents. Methods: Seventy children (mean age 10.8, SD 2.2) from five clinical centers in Poland were examined with a structured questionnaire regarding health related QoL (HRQoL): general health perception, Oude-Elberink scale, expectations of outcome, and impact of specific venom immunotherapy (VIT). Children’s parents were examined with similar questions referring to their children’s quality of life. The strength of agreement between these two measurements was calculated using Spearman rank correlation coefficient. Results: QoL measurements completed by parents on behalf of their children vs completed by children themselves revealed low to moderate agreement. The highest concordance was achieved for questions concerning living areas where insects are present and anxiety after being stung (0.52). Lower agreement was achieved for feeling of restriction due to allergy during summer and how the child’s health was (self-)rated (0.46). The lowest agreement was achieved for questions concerning the influence of VIT on anxiety against being stung (0.13), falling in panic at insects presence and discomfort when looking for insects while playing outdoor (0.21). QoL scales achieved the highest agreement for summary scales (0.49). Conclusions: A child may serve as a reliable source of information concerning their own HRQoL. This suggests that healthcare personnel may more often communicate directly with the child, without parents support.
123
IEA-EEF European Congress of Epidemiology 2009
216 EATING ATTITUDES AND RELATED FACTORS: A SURVEY CARRIED OUT ON 45.000 ITALIAN STUDENTS F. Denoth, S. Molinaro, V. Siciliano Italian National Council for Research, Institute of Clinical Physiology, PISA, Italy Background: The increasing number of adolescents that develop an eating disorder points out the necessity to conduce specific surveys on large population to define and steer prevention strategies. Objectives: To describe, using the Eating Attitude Test short version, EAT 26, the prevalence of some eating risk behaviors. Moreover, this study wants to investigate the possible related factors or individual attitudes that can act as the base to develop an eating disorder. The EAT 26 is part of the ESPAD questionnaire used in the homonymic survey conduced in more then 500 Italians high schools. Methods: Logistic regression analysis was used to examine the possible related factors to some problematic eating behaviors. The analysis was performed on the total score of the EAT 26 test and on its subscale. Results: Girls obtained highest scores 6.5% compared to the 0.9% of the boys. Some important related characteristics are to have had sexual behavior of which you have regretted (OR 3.8, P \ 0.05), to have participated in games in which you spend money (OR 3.8, P \ 0.05), to use tranquilizers without medical prescription (OR 2.3, P \ 0.05), but also often do sport, have problems with parents and teachers or have been involved in fights. Protective factors against the risk of developing an eating disorder seem to be to perceive a medium-high economical status (OR 0.5, P \ 0.05), to perceive an high risk in using illegal psychotropic substances (OR 0.5, P \ 0.05) and to be satisfied of the relationship with parents (OR 0.4, P \ 0.05). Conclusions: Although the EAT test is not a complete instrument by itself to make a specific diagnosis of an eating disorder, it is a good screening test to estimate the possible risk factors that can lead to improper eating behaviors.
55 MULTIPLE MEASURES OF CHILDHOOD SOCIOECONOMIC CIRCUMSTANCES AND MORTALITY: EVIDENCE FROM THE BRITISH WHITEHALL II AND FRENCH GAZEL STUDIES S. Stringhini1, A. Dugravot1, M. Kivimaki2, M. Shipley2, M. Zins1, M. Goldberg1, J. E. Ferrie2, A. Singh-Manoux1,2,3 1INSERM U687, Hoˆpital Paul Brousse, VILLEJUIF, France, 2Department of Epidemiology and Public Health, University College London, LONDON, United Kingdom, 3Centre de Ge´rontologie, Hoˆpital Ste Pe´rine, AP-HP, PARIS, France Background: Father’s occupational position, education and height have been used to examine the effects of adverse childhood socioeconomic position on health. These measures are used interchangeably and it remains unknown if they predict mortality equally well. Methods: Pooled data from the Whitehall II and GAZEL occupational cohorts related to 18393 men and 7060 women at mean age of 44 or 45 years at the time childhood socioeconomic position was measured. They were followed up for all-cause and cause-specific mortality; mean follow-up 19.2 years for all-cause and 17.8 years for causespecific mortality. Results: During follow-up, 1487 participants died, the most common causes being cancer (n = 654) and cardiovascular disease (n = 290). The three measures of childhood socioeconomic position were differently related to mortality. Education was associated with overall, cancer, cardiovascular, and non-cancer-noncardiovascular mortality [age-, sex- and cohort-adjusted hazard ratios varied between 1.30 (95% CI 0.96 to 1.84) and 1.60 (95% CI 1.26 to 2.04)]. There was evidence of a U-shaped association between height
IEA-EEF European Congress of Epidemiology 2009 and all-cause, cancer and cardiovascular deaths. These associations were robust to adjustment for the other childhood socioeconomic indicators. Participants with manual parental occupational status had in general a slightly higher risk of mortality than those with nonmanual parental occupational status, except for cardiovascular deaths for which there was no association. Conclusions: This study shows that the association between childhood socioeconomic position and mortality depends on the specific socioeconomic indicator used and the cause of death being examined. Short height can be a misleading indicator of childhood socioeconomic adversity as taller people do not have a health advantage for all mortality outcomes.
147 USE OF SHORT SCALES TO ASSES CANNABIS RELATED PROBLEMS IN YOUNG POPULATION S. Molinaro, L. Bastiani, V. Siciliano, F. Mariani Institute of Clinical Physiology, National Research Council, PISA, Italy Background: For several years, surveys carried out on young people in general population in Italy have clearly shown the increasing usage of the most common drugs, including cannabis. The estimate of cannabis users in Italy showed 2.800.000 people aged from 15 to 64 years old and 350.000 students from Secondary School. The only ‘‘contact’’ with the cannabis (at least once in last month), it is not symptomatic of problematic use, then, it is necessary to provide a psychoactive cannabis abuse screening tools. This study aims to present a short screening test for cannabis abuse among adolescents in school population survey, the CAST, designed at French Monitoring Centre for Drug and Drug Addiction (OFDT). Methods: A sub-sample of students, that used cannabis was selected from original database. The ESPAD-Italia questionnaire include a short section with CAST items; each one screens for different aspects of harmful cannabis use by assessing the frequency of some events throughout an individual’s life. A optimal threshold for the CAST score was calculated using ROC curve related to a variable that indicate daily cannabis use (20 or more times in last month). Results: The mean score of CAST is different between no-daily users and daily users: from 1.1 ± 1.4 to 3.4 ± 1.5. Using ROC curve is calculated an optimal threshold equal 3. This cut-off identifies correctly 82% of students (sensibility:75%; specificity: 83%). Conclusions: In our analysis we found a best threshold 3 using daily prevalence use of cannabis and the CAST seems to be an efficient tool in order to screen the problem among adolescents. We scheduled to compare the CAST with SDS and M-CIDI, to validate the optimal thresholds of the CAST for the high risk of abuse defined. Thus, this research contributes to the definition of common European screening tool for cannabis use disorders.
Infectious diseases epidemiology I 292 THE INFLUENCE OF MMR VACCINATION ON DEVELOPMENT OF INFANTS FOLLOWED OVER THE FIRST THREE YEARS OF LIFE D. Mrozek-Budzyn1, A. Kieltyka1, E. Flak1, E. Mroz1, A. Skarupa1, R. Majewska1, W. Jedrychowski1, F. Pererra2 1Chair of Epidemiology and Preventive Medicine, Jagiellonian University Collegium Medicum, CRACOW, Poland, 2Columbia Center for Children’s Environmental Health, Mailman School of Public Health, Columbia University, NEW YORK, United States of America Background: A great deal of speculation exists concerning the possible associations of certain vaccines, particularly the measles-
75 mumps-rubella (MMR) vaccine, with autism. It was suggested that gastrointestinal and developmental symptoms constituted a syndrome, which may be triggered by this vaccine. Objectives: The objective of this analysis was to determine the relationship between vaccination to measles (MMR and single measles vaccination) and cognitive and psychomotor development during the first three years of life. Methods: All children included in this study were a sample followed in a collaborative study with Columbia University in New York on the susceptibility of the fetus and child to environmental factors. The cohort recruited prenatally in Krakow (Poland) consist of 504 children. The Bayley Scales of Infant Development second edition (BSID-II) were administered to infants in 12th, 24th and 36th month of life. The data on measles vaccination was available for 316 children. The multivariate logistic and linear regression was used to examine whether MMR vaccination preceded the poor BSID-II scores. Results: Children vaccinated with MMR to 24 month of life had significantly higher mental BSID-II scores in 36th month then vaccinated with single measles vaccine (103.7 ± 10.3 vs 97.9 ± 11.2). Scores ‘‘Mildly Delayed’’ or ‘‘Significantly Delayed’’ reached respectively 3.5% and 20.7% subjects. The relative risk of ‘‘delayed’’ was 7.2 times higher in ‘‘single measles vaccine’’ than in ‘‘MMR’’ group (95PU%: 2.41–21.7). After standardization (gender, maternal education, the birth order, lead level in cord blood, ETS during pregnancy) OR was equal 4.25, however lost its significance (95PU%: 0.91–20.0). Psychomotor scale did not show any relation with measles vaccination. Conclusions: Our study demonstrated that the MMR vaccine did not inhibit infants’ development. On the contrary, children vaccinated with ‘‘single measles vaccine’’ had tendency for higher risk of delayed mental scores.
297 CHLAMYDIA PNEUMONIAE INFECTION IN CHILDREN WITH TYPE 1 DIABETES MELLITUS B. Banecka1, B. Bienias´2, L. Szewczyk1 1Department of Pediatric Endocrinology and Neurology, Medical University in Lublin, LUBLIN, Poland, 2Department of Pediatric Nephrology, Medical University in Lublin, LUBLIN, Poland Background: With the blood from the lungs, contained in circulating monocytes, the tissue macrophages, Chlamydia pneumoniae (CP) can be passed on to the other organs. When the immune system is unable to eliminate Chlamydia, the macrophages occur to stimulate the release of pro-inflammatory cytokines, which makes the inflammation become a self-fuelling chronic process. The serological confirmation of CP infection is the presence of specific antibodies in blood serum. Material and methods: The study included 47 children with diabetes mellitus type1(DM1). The control group was represented by 37 children suspected of anatomical deviations in the urinary system. The diabetic patients were divided into two subgroups:23 children with recent onset DM1 and 24 children with long-lasting DM1. Serum IgM and IgG antibodies against CP were determined by immuno-assay. The statistic analysis was based on the Mann–Whitney test. Results: In the group of children with recent onset diabetes the level of IgM-antibodies was significantly higher than in the control group (Me = 6.3 vs. 4.4; P = 0.03) while the difference between the level of IgG remained insignificant. In the group of children with recent onset diabetes the IgM level was significantly higher than the level of IgG (Me = 6.3 vs. 3.1; P = 0.002)—In the group with long-lasting diabetes the IgG level was higher than in the control group and the difference was considered close to significance (Me = 4.35 vs. 2.9; P = 0.07) while the IgM level in the two groups did not differ significantly. Conclusions: The significantly higher, compared to the control group, level of IgMantibodies against CP in the group with recent onset DM1 as well as the
123
76 significantly higher level of IgM than that of IgG found in the group with recent onset DM1 may indicate a correlation between the CP infection and developing diabetes mellitus type 1.
IEA-EEF European Congress of Epidemiology 2009 statistically significant (P \ 0.005). Dot blot assay is considered a very discriminatory system for detecting Leishmania infection in fielded study and screening test. The Simplicity, reproducibility, high sensitivity and high specificity of the assay show determining the prevalence of sandfly infection.
44 EPIDEMIOLOGICAL SURVEY OF FOODBORNE BOTULISM IN IRAN DURING 2004–2008 H. R. Tavakoli1, M. Zeynali2 1Nutrition & Food Hygiene Department, School of Health and Health Research Center, Baqyatallah University of Medical Science, TEHRAN, Iran, 2Disease Management Center, Ministry of Health, Treatment and Medical Education, TEHRAN, Iran Background: Food-born botulism poisoning is one of the most hazardous food intoxication due to consumption of contaminated food such as meat, canned vegetable, traditional sea foods (smoked salted fish, fermented fish, spawn) and in rare cases dairy products. In Iran, for instance, and bunches of other countries, it is considered a health problem with variable prevalence rate. Materials and methods: Studying epidemiological aspect of food-borne botulism in Iran during a 5-year period is the goal of this survey. In this cross-sectional study, the data of the botulism cases in Iran during the year 2004–2008 have collected and analyzed epidemiologically on account of food items which produce it, processing type, affluence distribution in rural & urban vicinities and widespread types. Results: Overall, 341 suspicious cases of botulism have been recorded in Iran during this 5-year period. The most and the least cases pertained to the years 2008 (28.28%) and 2004 (12.6%) respectively. Among the provinces, Gilan lead the cases with 90 instances while the least amount related to West Azarbayjan with just 4 occasions (1.17%). According to the survey, Spawn and Salin fish (31.08%) are the most botulism producing factors. Indeed, the prevalence rate in rustic vicinities (56.59%) was punctually more than the cities(39.29%). Providing & consumption of canned foods in home was one of the most predominant causes of food poisoning (53.07%). Moreover, this study unraveled that A and E types with the rate of (18–47%) and (13.7–41.3%) respectively are the most common causes of engendering botulism food poisoning in Iran. Discussion and conclusion: It has been found from this survey that this kind of food poisoning had a escalation rate during the years 2004 till 2008 probably due to paying more attention to record and report the occurred cases in recent years in comparison to what was used to be. Some measures like general health education, not using non sanitary and traditional food processing, heating sufficient while consumption, not using non pasteurized dairy product and regular supervising by health inspector can prevent claimed issue.
47 COMPARISON DOT BLOTS AND CULTURE METHOD TO DETECTION LEISHMANIA ANTIGEN WITHIN NATURALLY INFECTED SANDFLIES IN ABARDEJ, IRAN H. Nekouie, N. Hassan Pasteur Institute of Iran, TEHRAN, Iran Leishmaniasis is a worldwide infectious disease. Leishmania. spp is found in the tropical and subtropical areas of Asia, Africa, and South America. cuntaneous Leishmaniasis is a major health problem in many rural in several provinces of Iran . Different methods were used to identify Leishmania antigen from natural foci of zoonotic cutaneous Leishmaniasis and vectors. We used Dot blot immunoassay for detection leishmanial antigen in the natural infection phlebotomus with comparison culture method. Of specimens 22.2% were positive in culture media, 37.5% were positive by Dot immunoblot. We can detect 10 ul of leishmania antigen in by Dot blot. This difference was
123
62 NOTIFIED IMPORTED INFECTIOUS DISEASES IN THE CZECH REPUBLIC J. Dlhy1,2; C. Benes2 1Regional Public Health Institute of the Kralovehradecky region, HRADEC KRALOVE, Czech Republic, 2 National Institute of Public Health, PRAGUE, Czech Republic Objectives: Based on the analysis of epidemiological characteristics, a trend specification for imported infectious diseases in the Czech Republic the objective of which is to draw attention to the epidemiologic meaning of travelling abroad and migration was made. Material and methods: In cooperation with the National reference centre for analysis of epidemiological data in the National Institute of Public Health, the Epidat information system databases from the whole Czech Republic as created from 1993 to 2007 were analysed in the software environment of 6.04d Epi Info version. Results: In the course of 1993–2007, 13 829 infectious diseases were imported into the Czech Republic from 172 destinations of the world. In individual years, the reported numbers of imports fluctuated within the range from 206 to 1 714 per a year. The highest rate (56% of all imports) fall to Czech tourists, 36% of infectious diseases were imported by foreigners. The most frequent countries of import provenance are Slovakia (1348 cases), Vietnam (1268 cases), and India (862 cases). According to the diagnosis of imported cases as stated, the highest rate was found out for salmonellosis (22%), campylobacteriosis (11%) and shigellosis (8%). Conclusions: In the Czech Republic, the Epidat system is used for the purposes of communicable diseases reporting and recording. The Epidat database analysis results aimed at the recorded cases of imported infectious diseases represent an important starting point for estimation of the public health risks connected with travelling abroad.
140 THE PREVALENCE OF H. PYLORI INFECTIONS IN ST. PETERSBURG AT PRESENT A. Svarval, R. Ferman, A. Zhebrun, O. Balabash Pasteur Institute, SAINT PETERSBURG, Russia Helictobacter pylori is the aetiological agent of active chronic gastritis and is involved in the development of peptic ulceration and possibly gastric carcinoma. Methods: In our laboratory we have screened 198 patients aged from 20 to 60 and older who were subjectively apparently healthy. ELISA-method was used for the serological screening. Availability of Cag A H. pylori antibodies was determined. Results: The conducted screening showed that 55.1% out of all screened patients were seropositive to Helictobacter pylori. The percentage of seropositive patients in different age groups was the following: in diagnostic group aged 20–29 it was 48.27%, among patients aged 30–39 it was 53.49%, whereas among patients aged 40–49 the percentage of seropositive individuals comprised 55.74%. For the diagnostic group of over 60 y.o. the percentage was lower and comprised 33.33%. Conclusions: Serological studies have revealed high prevalence of H.pylori infections. Taken this fact into consideration we suppose that the development of specific dispositional preventive measures is probably necessary, i.e. prevention of diseases among the population already infected
IEA-EEF European Congress of Epidemiology 2009 with H. pylori. This problem is worth further consideration. Suggestions with regard to introduction of national programs of mass H. pylori albation are being discussed and require further case studies. Specific vaccine treatment may be effective. However, it requires designing vaccines that are likely to have a better protective effect that natural H. pylori infection.
188 OVERVIEW ON EUROPEAN PASSENGER TRANSPORT STATISTICS/FLOWS: ANALYSIS OF STATISTICAL DATA FOR THE REACT PROJECT (RESPONSE TO EMERGING INFECTIOUS DISEASE: ASSESSMENT AND DEVELOPMENT OF CORE CAPACITIES AND TOOLS) M. Poorbiazar, O. Mohr, E. Velasco, T. Eckmanns, G. Krause, G. Poggensee Robert Koch Institute, Department of Infectious Disease Epidemiology, BERLIN, Germany Background: Increased mobility of persons and goods require coordinated preparedness and response to infectious diseases across Europe. REACT is an EC-funded multi-agency project to develop strategies and risk assessment tools for international contact tracing after exposure to infectious diseases in public means of transport. Currently most public health efforts concerning public passenger transport are focused on air travel while the importance of other means of transport is fairly unknown. Objectives: Identification of the importance of different means of transport according to passenger transport data for EU27 member states. Methods: European passenger transport data for 1995 to 2006 were obtained from Eurostat. Analyses were completed for EU27 according to passenger transport performance (TP) as the number of passengers times travelled distance, in billion passenger-kilometres (Bpkm). TP follows the ‘‘territoriality principle’’, excluding distances travelled trough transit countries or extraterritorial sea. Results: Total passenger TP in EU27 member states increased from 5.377 in 1995 to 6.333Bpkm in 2006. The biggest part was generated by passenger cars (3.854 to 4.601Bpkm) and stayed stable at *73% over the whole period. The second biggest part was air transport (335 to 547Bpkm) with a slightly rising proportion (6.3-8.6%) over the years, followed by busses/coaches (501 to 522Bpkm, 9.5–8.3%) and railway (348 to 384Bpkm, 6.6–6.1%). Conclusion: In passenger transport the share of passenger air transport has continuously increased over the last years. Due to the ‘‘territoriality principle’’, its actual performance is likely to be even higher. Limitations were as follows: absence of data on transport distance, total passengers carried and unstandardized data collection for bus transport. However, data demonstrate that busses/coaches and railway are also very important within European public passenger transport. Therefore it is necessary to assess the risk of transmission of infectious diseases not only in air transport but also in busses/coaches and trains.
77 access to healthcare and prevention methods. In a review of the literature, these factors have been cross-referenced and analyzed across geographical locations to determine possible similarities in the type, level and extent of these interactions. In particular, the balance among these influences may be unstable in regions where rice cultivation has been established relatively recently. Through this analysis, there are implications for varied rice cultivating techniques, in particular the System of Rice Intesification (SRI), a method that avoids constant flooding of rice fields in favor of intermittent irrigation. It is relevant to explore its potential relationship to relevant vector abundance and malaria transmission, as the SRI approach is being increasingly explored in farming communities throughout the world, in response to local economics and water usage. The findings are extrapolated to the country of India, where rice is grown throughout, in various climates and environments. It suggests that there are locations that may serve to further investigate this unique vector biology, using an epidemiological study design.
Social epidemiology I 198 SPATIO-TEMPORAL PATTERNS OF SELECTED SPECIFIC MORTALITY IN THE NORTHERN MORAVIAN REGION I. Tomasek, H. Tomaskova, P. Polaufova, H. Slachtova, A. Splichalova Institute of Public Health in Ostrava, OSTRAVA, Czech Republic Background: The mortality is generally decreasing in the Czech Republic. The Northern Moravian region has one of the highest mortality rates of all Czech regions. Objective: The aim of this study was to assess whether a mortality trend in 22 small geographical Northern Moravian units (ORP) over a period 2003–2006 follows a national trend and to identify the units with the highest mortality for future interventions. Methods: Posterior expected relative risk for total and selected specific mortality in ORPs was analysed. Bayesian space time model was used with random effects in both area specific intercept and linear time trend. Results: Total mortality risk in men was above one, with increasing time trend in 7 ORPs. Also in women the risk was above one with slightly increasing trend in all ORPs. Total cancer mortality risk in both genders was slightly below one with increasing trend. Lung cancer risk in men was constantly higher in north-west ORPs with decreasing trend in remaining ORPs. In women, lung cancer risk was low but the trend was increasing in all ORPs. Cardiovascular mortality risk trend in men has been increasing over time up to risk above one and decreasing in women. Respiratory mortality risk was above one with decreasing trend in both genders. Similarly, gastrointestinal mortality risk was above one in all ORPs in both genders with increasing trend in a part of ORPs. Breast cancer mortality risk was slightly below one in all ORPs but with increasing trend. Conclusion: This analysis showed that the mortality trend for selected diseases did not follow the same pattern of decreasing national trend of mortality in some ORPs in the Northern Moravian region. In some diseases, significant differences in mortality risk have been found among ORPs in the region.
298 METHODS OF RICE CULTIVATION AND MALARIA TRANSMISSION: A REVIEW J. K. Dhaliwal, J. S. Rai, Sikh Human Rights Group, LONDON, United Kingdom The correlation between traditional rice cultivation and increased mosquito abundance has been recognized, but its relation to malaria transmission is more complex, as it depends on such factors as vector species and breeding habits, disease endemicity, socioeconomics,
270 EPIDEMIOLOGY OF BURNS ADMITTED TO BURN UNITS IN ZANJAN, IRAN S. Mazloomzadeh1, M. Hejazi2, M. H. Moghimi3 1Faculty of Medicine, Zanjan University of Medical Sciences, ZANJAN, Iran, 2 Zanjan University of Medical Sciences, ZANJAN, Iran, 3Moosavi Hospital, Zanjan University of Medical Sciences, ZANJAN, Iran
123
78 Background: Burns are still one of the most common injuries and a main cause of disability in developing countries. Patients with such injuries usually need to undergo several years of rehabilitation, and psychological control. No information was available on the current epidemiological features of burns in Zanjan, northwest of Iran. Objectives: The purpose of this study was to determine the epidemiological and outcome characteristics of burn injuries in Zanjan. Methods: This prospective study was carried out on patients admitted to the burn units of shafiyeh and Moosavi Hospitals from October 2007 to September 2008. Information on patient demographics, etiology of burn, extent and anatomical areas burned, length of hospital stay, and outcome were collected. The age- and sex-specific incidence rates of burns per 100,000 person-years were calculated using the population data from the last national census in Iran 2007. Results: A total of 257 burn-injured patients were admitted to burn units. The annual incidence rates of burn injury admissions for men and women were 34.7 and 18.9 per 100,000 person-years, respectively. The incidence rates were 22.3 per 100,000 of urban population and 34.6 per 100,000 of rural population. The highest rate of admissions occurred in children younger than 10 years old for both sexes. This was observed in total, urban and rural populations. Flame was the most common cause of burn (45.9%). The majority of burns (77.4%) occurred at home. The mean total body surface area (TBSA) burned was 13.0%. The highest proportion of burns was observed in winter (32.3%). The mean length of hospital stay for burn patients was 10.2 days (range: \1 to 110). The mortality rate of burns was 0.41 per 100,000 person-years. Conclusions: The results of this study provide the necessary information to develop proper burn prevention programs to reduce the frequency of burns in this area.
161 SURVEY OF FAMILY RURAL AREA OF GOLSTAN PROVINCE FOR EARTHQUAKE PREPAREDNESS E. Ahmadnezhad1, A. Ardalan1, K. H. Naeni1, F. Fayyaz-Jahani2 Tehran University of Medical Sciences, TEHRAN, Iran, 2Zahedan University of Medical Sciences, ZAHEDAN, Iran During last decade, an average of 4,000 people were killed and 55,000 affected annually by natural disasters in I.R. Iran. Although the rural areas of the country with 34 million of Iranian population are highly vulnerable to natural disasters, no disaster risk management structure has been defined in the village level. While there is growing evidence that most top-down disaster risk management and responses programs fail to address specific local needs of vulnerable communities, ignore the potential of local resources and capacities. Golestan is a northern province of Iran and its rural areas are highly vulnerable to different types of natural hazard, mostly flash flood, earthquake and wild fire. Village Disaster Taskforce program has been a successful community-based model that was piloted in 2006 in a part of Golestan province focusing on flash floods. This project aims to reduce the risk of natural disasters in rural area of Golestan province through empowerment of the community by extending the VDT program to other part of the province with all-hazard approach. This is a cross sectional survey has been done in Golestan area by 210 families. It has 3 parts includes: Stakeholder analysis, Risk assessments and evaluation of family’s preparedness. We used a questionnaire for data collection. 146 families asked the questionnaires. 64 of them participated in risk assessment phase, then they didn’t participate in preparedness phase. Results of preparedness evaluation show that any family has no supplement or no practice for earthquake preparedness. Knowledge about them, are also very low. The existence of a ‘‘culture of coping with crisis’’ and ‘‘culture of disaster reduction’’ is
123
IEA-EEF European Congress of Epidemiology 2009 necessary. Risk assessment process would be involved participation of people and incorporating of vulnerability and capacity.
251 THE ASSOCIATION BETWEEN SOCIOECONOMIC FACTORS AND DENTAL OUTCOMES ALONG THE LIFE COURSE AND THE PREVALENCE OF ORAL IMPACTS OF DAILY PERFORMANCE IN BRAZILIAN 12 YEAR OLDS K. G. Peres1, M. A. Peres1, A. M. B. Menezes2, C. L. P. Arau´jo2, P. C. Hallal2 1Federal University of Santa Catarina, ´ POLIS, Brazil, 2Federal University of Pelotas, FLORIANO PELOTAS, Brazil Background: The way how individuals perceive and evaluate their health should be included in population health surveys. A variety of socio dental indicators have been developed and used to overcome the normative assessment bring contributions from psychology, sociology, economics, operational research and biostatistics. Objectives: We aimed to estimate the prevalence of oral health impacts on daily performance (OIDP) among 12 year-old Brazilian adolescents (n = 359) and its association with life course socioeconomic, clinical outcomes and dental services utilization. Methods: This study is nested in a population-based birth cohort in Southern Brazil. Exploratory variables were collected at birth, and at 6 and 12 year of age. The OIDP was analyzed as an ordinal outcome (OIDP = 0; 1; and 2 or above). Multivariable ordinal regression was performed. Results: The prevalence of OIDP = 1 was 30.1% (CI95%25.2;35.0) and OIDP C 2 was 28.0% (CI 95% 23.2; 32.8). The most common daily activity affected was eating (44.8%), follow by cleaning the mouth and smiling (15.6%, and 15.0%, respectively). The risk score was significantly associated with having incisal crowding (OR = 1.6, [CI 95% 1.0; 2.6]), untreated dental caries (OR = 1.9 [CI 95% 1.2; 2.9]), high proportion of teeth with gingival bleeding (OR = 2.0 [CI 95% 1.2; 3.3]), and dental pain (OR = 3.6 [CI 95% 1.8; 7.0]) at aged 12, after adjusting by early life factors. Conclusion: Our findings highlight the importance of the later dental outcomes on the adolescent’s quality of life independently of the social aspects and early oral health conditions.
220 PREVALENCE AND SOCIAL GRADIENT OF MUSCULOSKELETAL DISEASES IN CATALONIA. THE POPULATION HEALTH SURVEY OF CATALONIA (ESCA 2006) M. Larrosa, X. Surı´s, M. J. Pueyo, J. Auleda, P. Brugulat Master Plan on Musculoskeletal Diseases, Health Plan Unit1, Health Department of Catalonia, BARCELONA, Spain Introduction: The Health Survey of Catalonia (ESCA) is a population survey held by the Department of Health of Catalonia (over 7 million inhabitants) in 2006. Objective: To know the situation of musculoskeletal diseases in a population survey (ESCA) and to analyze its distribution according to sex, age, social class and educational level. Material and methods: Cross-sectional descriptive study. A total of 15,926 adults (aged 15 or over) were surveyed in a multi-stage stratified sampling, representative of sex and group of age of the country. Collected variables were sociodemographic characteristics, healthrelated behaviors, self-perceived health and quality of life, selfreported chronic health problems and use of health services. The musculoskeletal problems were grouped in four categories: chronic
IEA-EEF European Congress of Epidemiology 2009 joint disease (osteoarthritis, arthritis, and rheumatism); chronic low back pain (dorsal or lumbar); chronic upper back pain, and osteoporosis. Results: Chronic joint disease, low back pain (LBP), and upper back pain (UBP) are the main chronic problems of health. Chronic joint disease was reported by 22.7% of the general population; LBP by 29.6%, UBP by 27.4% and osteoporosis by 5.6%. The prevalence was higher in women: 30.3% versus 14.9%, 34.8% versus 24.3%, 36% versus 18.6% and 9.7% versus 1.5% respectively. After adjusting by age, being female increases risk of having chronic joint disease, LBP, UBP and osteoporosis (OR 2.6, 1.66, 2.45, 7.18 respectively). Social gradient for the four categories was observed with higher prevalence in lower social class and in people with low educational level. The BMI were significantly higher in people with chronic joint disease (27.12 versus 24.75) and LBP (26.1 versus 24.94). Conclusions: Chronic joint disease, LBP, and UBP are the main chronic problems of health reported by the population in Catalonia. Being female increases the probability of declaring these problems, especially osteoporosis. Social gradient was observed in the four categories.
208 RELATION BETWEEN SOCIO-ECONOMIC STATUS(SES) AND PULMONARY FUNCTIONS IN RESIDENTS OF KRAKOW AT AGE 45–70 YEARS R. C. Nagaraj, K. Szafraniec, A. Paja˛k Department of Epidemiology and Population Studies, Institute of Public Health, Faculty of Health Sciences, Jagiellonian University Medical College, CRACOW, Poland Low Socio-Economic Status (SES) is a well recognized determinant of poor health. The relation between SES and pulmonary functions was not extensively studied in the East European countries. Goal: To assess the relation between pulmonary function (FVC and FEV1) and characteristics of SES i.e. educational level, marital status, current employment, childhood socioeconomic circumstances and current material deprivation. Studied group and methods: The study sample consisted 3833 men and women at age 45–70 years, randomly selected from the population register of Krakow for the Polish part of the HAPIEE (Health, Alcohol, Psychosocial Factors in Eastern Europe) Project. Data on SES were collected by structured questionnaire. FEV1 and FVC were assessed by the Micro-Medical Microplus spirometer and the SPIDA 4 software (recording the curves, in addition to numerical results). Associations between SES and lung functions were estimated by multivariate linear regression. Results: In women, lower FVC and FEV1 were significantly associated with low own education, low education of parents, and current material deprivation, after adjusting for age, smoking and height. In men, after adjustment, current material deprivation and unmarried status were related to lower FEV1 but not FVC. In men, there was no relation between own education and pulmonary function but father’s education was inversely associated with FEV1. Current employment status was not related to FVC and FEV1 in either sex. Conclusions: Our results confirm that low SES is related to poor pulmonary function but the association differed between genders and varied by indicator of SES.
235 SITTING TIME CORRELATES IN A BRAZILIAN POPULATION—A CROSS SECTIONAL STUDY C. S. Suzuki, S. A. de Moraes, I. C. M. de Freitas University of Sa˜o ˜ O PAULO, Brazil Paulo, SA Background: Sedentary behavior, mainly sitting time, has been considered an important risk factor for cardiovascular diseases, obesity
79 and diabetes mellitus. Objectives: Estimate the prevalence of sitting time and identify correlates of this sedentary behavior. Methods: Cross-sectional population-based epidemiological study using threestage cluster sampling, being census sector the primary sampling unit. The variability introduced in the third sampling fraction was corrected by attributing equal weights to the number of eligible units in each domicile, resulting in a weighted sample of 2197 participants aged 30 years or older, living in the urban area of Ribeira˜o Preto-SP, Brazil, in 2006. To identify the predictors of sitting time, multiple linear regression models were built. In the univariate analysis, P \ 0.25 criterion (Wald Test) was used to select candidates for the multivariate model. The design effect was considered when calculating all estimates. Results: The reported mean time spent sitting/day was 280.94 minutes, and it was significantly higher (P \ 0.002) in men (306.18 minutes) than in women (270.25 minutes). The highest prevalence of sitting time was identified in the fourth quartile ([ 360 minutes), among participants aged 80 years and older: (63.65%; CI 95%: 37.91–83.39). In the multivariate analysis, the following variables were kept in the final model: gender (b = -37.661; CI 95%:-62.338–12.984); years of education (b = 5.569; CI 95%: 3.416–7.722); occupational time (b = -6.867; CI 95%:-10.738– 2.996); familial antecedents of obesity (b = 18.775; CI 95%: 6.997–30.553); total MET-min-week-1 (b = -0.008; CI 95%: -0.014–0.002),and smoke (b = 27.146; CI 95%:1.616–52.677). Conclusions: Considering that sedentary behavior, including sitting time, has been appointed as one of the correlates of chronic-degenerative diseases, this study reveals its importance by identifying the main predictors of sitting time, supporting the planning related to rational public health promotion and prevention policies.
Lifecourse epidemiology 295 FOETAL ORIGINS OF DEPRESSION? A SYSTEMATIC REVIEW OF BIRTH WEIGHT AND LIFETIME DEPRESSION W. Wojcik, W. Lee, M. Hotopf Dept of Psychological Medicine, Institute of Psychiatry, LONDON, United Kingdom Introduction: Models of illness which are informed by early developmental factors and behavioural exposures accumulating over the life course have been gaining currency over recent years, thanks to evidence from epidemiology, genetic and epigenetic research. Meanwhile, established evidence of an association between depression and cardiovascular disease is frustrated by negative findings in several trials of treating depression and cardiovascular outcomes. Could an upstream association of both depression and cardiovascular disease with developmental adversity in early life go some way to explaining the association between them? The ‘Barker hypothesis’ of fetal programming influencing a trajectory towards chronic ill health has yielded positive findings for diabetes, hypertension and heart disease, using birth weight as a proxy for foetal adversity. Evidence for a similar associations in depressive illness is less clear and to date there has been no systematic review. Objective: To review the evidence for an association between birth weight and depression in later life. Method: We conducted a systematic review. The search strategy included search strings in Medline, PsychInfo and Embase, a citation search from key papers and a hand search, yielding in excess of one thousand papers. Inclusion criteria were kept broad to aid an overall summary of an heterogeneous field, before applying a checklist to describe and rate each paper, each of which was rated by two independent investigators. Results and Conclusions: We discuss both the methodological observations yielded from this work, the overall findings, and their relevance to the broader field of psychological medicine.
123
80
115 FACTORS ASSOCIATED WITH PARTICIPATION AND ATTRITION IN THE FRENCH GAZEL COHORT M. Zins, J. F. Chastang, M. Coeuret-Pellicer, A. Leclerc, S. Bonenfant, A. Gue´guen, A. Ozguler, M. Goldberg INSERM Unite´ 687, VILLEJUIF, France Background: The GAZEL Cohort was launched in 1989 among employees aged 35-50 years of the French national utility company (E´lectricite´ de France-Gaz de France: EDF-GDF). Inclusion was through mailed self-questionnaires, and 20,625 subjects (15 011 men and 5,614 women) accepted to participate. The follow-up of the cohort includes a yearly mailed questionnaire and an invitation to attend a health screening centre (HSC) for a full medical examination. The participation rate was 44.5% at inception; each year about 75% complete their self-questionnaire; 44.7% attended a HCS. Objectives: To analyse the sociodemographic, occupational, lifestyle and health
123
IEA-EEF European Congress of Epidemiology 2009 factors associated with selection effects at inclusion, during the annual follow-up and the health examination respectively, and to quantify their role. Methods: at inception, the volunteers were compared to non-participants according to variables systematically collected in the personnel and medical databases of the company (absenteeism, mortality, occupational exposures) using logistic regression models; the probability of completing the annual selfquestionnaires was analysed through mixed models; logistic regression models were used to study the participation to the medical examination. Results: It was not always the same variables that play a role at the inclusion, follow-up and medical examination stages, and the magnitude of the selection effects varies according to each specific stage. Conclusions: This study took advantage of the extended follow-up of the cohort, and of the availability of various sources of data both for respondents and non-participants. It was possible to differentiate the specific role of numerous factors associated with participation at inception and during the follow-up of the cohort, thus allowing for a better understanding of the potential selection biases.
IEA-EEF European Congress of Epidemiology 2009
Friday, 28 August 2009, 15.30–16.30 Cancer epidemiology II 4 REFERENCE VALUES FOR THE ASSESSMENT OF QUALITY OF LIFE FOR PATIENTS WITH ORAL AND OROPHARYNGEAL CANCER J. L. F. Antunes1, F. P. de Andrade1, M. G. H. Biazevic1, M. B. de Carvalho2 1School of Dentistry, University of Sao Paulo, SAO PAULO, Brazil, 2Head and Neck Surgery Service, Heliopolis Hospital, SAO PAULO, Brazil Background: The self report of quality of life (QoL), for patients with head and neck cancer is documented by questionnaires assessing the impact of the disease and its treatment on several physiological, psychological and social domains. Objectives: To assess reference values for quality of life domains comprised by the University of Washington Quality of Life Questionnaire (UW-QoL, version 4) for persons without cancer, and to compare their answers with results obtained for patients with oral and oropharyngeal cancer previously to the primary surgery of tumor resection. Methods: A total of 141 patients attended at outpatients units of a large hospital in Sao Paulo, Brazil (2007–2008), were interviewed to assess reference values for the UW-QoL questionnaire, paired by sex and age with 47 patients hospitalized for surgery of oral and oropharyngeal squamous cell carcinoma in the same hospital. This questionnaire was already translated and validated for use in the Brazilian context. Overall and domain specific (pain, appearance, activity, recreation, swallowing, chewing, speech, shoulders, taste, saliva, humor and anxiety) ratings for QoL vary from 0 to 100, with higher figures indicating improved functional status. Results: The overall rating of QoL for patients without cancer was 91.1. This rating was significantly (P \ 0.01) higher than the corresponding figure for patients affected by cancer (80.6). Patients without cancer ranked significantly higher ratings for pain, appearance, swallowing, speech, taste and anxiety than their counterparts affected by the disease. Conclusions: Near 9% of the reduction in QoL ratings may be due to other causes associated with the aging process. This proportion is reduced, which reinforces the hypothesis that the questionnaire is specific for the assessment of impairments associated with head and neck cancer. However, the reduction of QoL reported by patients with oral and oropharyngeal cancer should not be entirely attributed to the disease and its treatment.
68 REDUCING BRAIN TUMOUR PATIENTS HOSPITALIZATION TROUGH A HOME CARE PROGRAM—ROME, ITALY A. Capon1, A. Pace2, A. Pompili2, A. Mastromattei1, V. Mantini3 Agency for Public Health - Lazio Region, ROME, Italy, Department of Neuroscience, Regina Elena National Cancer Institute, ROME, Italy, Direction for health planning, Lazio Region, ROMA, Italy Background: Despite aggressive antitumour treatment, the prognosis for brain tumour (BT) patients remains gloomy. Moreover existing data suggest that in the last stages of disease too many patients do not receive adequate palliative care. Methods: A comprehensive program of neuro-oncological home care for BT patients discharged by the National Cancer Institute of Rome was started in year 2000. The program includes periodic neurological surveys, neuro-rehabilitation at home, psychological support for patients and family and nurse assistance increasing up to a level of ‘‘hospice at home’’ in the late stages of the disease. The end-of-life quality of life for BT patients was
81 evaluated comparing hospitalization rates in the month previous to death among BT patients enrolled in the program (UNIT A) and BT patients discharged from a control unit (B). Date of death and hospital admissions were ascertained through record linkage with mortality register and hospital discharges archives. Results: From July 2002 to August 2008 141 deaths occurred among patients discharged by the 2 units (72 A, 71 B). Mean age at death in the 2 groups didn’t differ significantly (61.4 (a) vs. 63.6 (b)).Crude monthly hospitalization rate in unit A (8.3 9 100) was significantly lower than unit B (28.2 9 100) while Poisson regression age and sex-adjusted Incidence Rate Ratio for unit B compared to unit A was equal to 0.29 (IC 95% 0.12–0.74 P \ 0.009). Averages hospitalization costs in the month previous to death were significantly higher in the unit B (24 €) compared to unit A (517 €). Results didn’t change substantially when performing analysis over a two months period before death. Conclusions: Palliative programmes and home-care models of assistance may represent an alternative to in-hospital care for the management of patients dying from a brain tumour and may improve the quality of end-of-life care.
212 CANCER INCIDENCE, MORTALITY AND SURVIVAL IN POLAND U. Wojciechowska The Maria Skłodowska-Curie Memorial Cancer Center and Institute of Oncology, Department of Epidemiology and Cancer Prevention, WARSAW, Poland Background: Cancer is the second most frequent cause of death in Polish population. During 40 years of observation the absolute numbers of new cancer cases and cancer deaths are systematically growing. Objectives: The analysis presents changes in incidence and mortality from cancers in Poland in years 1965–2006 and estimation of 5-year relative survival rates for patients diagnosed in 1999–2002. Methods: Trends of incidence and mortality were based on data coming from 2 sources: cancer cases come from National Cancer Registry and deaths from the Central Statistical Office. For the analysis of time trends there were used rates standardized to the world population (ASW). The analysis of 5-year relative survival rates was done with the Hakulinen method (SURV3). Results: Increase in incidence and deaths in Poland is caused mainly by the aging of population. Percentage of people after the age of 60 years increased from 10 to 17%. The most frequently registered cancers in men in 2006 were lung, prostate, colorectal cancers, and in women: breast, colorectal and lung cancers. In majority of cancer sites, both incidence and mortality, there are observed increasing trends of standardized rates (ASW), only in a few localizations (lung and stomach in men and stomach and uterus in women) declining trends can be noticed. 5-year survival rates for overall cancer in people diagnosed in years 1999–2002 differs substantially by sex and amounts 31.4% in men and 50.3% in women. Conclusions: Poland is a country with high risk of cancer, in which multidirectional prevention actions can lead to lowering the cancer risk. Population screenings of breast, cervix and colorectal introduced in 2006, and conducted for many years health education programs based on the European Code against Cancer are actions, which will allow on substantial reduction of cancer mortality and through early diagnosis improve 5-year survival rates.
281 CANCER INCIDENCE IN BELARUS IN 1970–2005 N. Novikava1, I. Veyalkin2 1International Sakharov Environmental University, MINSK, Bielarus, 2 Republican Scientific and Practical Centre of Oncology and Medical Radiology, MINSK, Bielarus
123
82 Objective: To examine main trends in cancer incidence in republic over the last 36 years. Subjects and methods: Basing on the data from National Cancer Registry age-specific incidence rates were investigated and age-standardized incidence rates (ASRs, World Standard) and cumulative risks were calculated. Smoothing of rates was made with method of moving averages. Tendency equations were plotted using linear regression method. Results: 892,899 cases were registered during the years 1970 to 2005. The average annual incidence for all cancers was 250.2 per 100.00 (crude rate); 187.4 per 100,000 (ASR), of which the ASR for male was 224.85, and 158.25 per 100,000 for female. Mean cumulative risk was 37.63% (95% confidence interval (CI) 35.44–39.82) for male and 22.97% (95% CI 21.63–24.3) for female. The most common leading sites in rank were lung in male (mean ASR 53.6 per 100,000) and breast in female (mean ASR 27.09 per 100,000), stomach (45.31 in male and 20.53 in female), skin (20.51 in male and 17.83 in female), prostate (12.2) and cervix (12.08). Stomach cancer incidence declined significantly; breast, skin, thyroid, prostate and kidney (in male) cancer incidence increased significantly (particularly from the end of 80 s). Statistically significant (P \ 0.001) positive slopes were revealed (4.49 with standard error SE = 0.15 in male and 2.75 with SE = 0.099 in female for ASRs). The incidence was higher for urban population. Conclusions: Substantial changes in rates of various cancers have taken place. Monitoring of cancer incidence through population-based cancer registration did help to show the patterns of main cancers and time trends.
Cardioepidemiology II 64 RANDOMIZED, CONTROLLED, DOUBLE BLIND, CLINICAL TRIAL TO MEASURE EFFICACY AND SAFETY OF FLAVANOID-RICH COCOA IN TREATMENT OF PREHYPERTENSION AND TYPE I HYPERTENSION IN PANAMA V. Bayard1, J. Motta1, F. Chamorro1, R. Czako´2, N. Hollenberg3 1 Gorgas Memorial Institute, PANAMA, Panama, 2Johns Hopkins University, BALTIMORE, United States of America 3Harvard University, BOSTON, United States of America Background: Several clinical trials had evaluated the effect on hypertension of flavonoid–rich dark chocolate. The majority was on European men and with crossover design. Objective: To measure the impact of flavanoid-rich cocoa on the level of arterial tension in prehypertension to mild hypertension men of three ethnic group and women in Panama. Design: A randomized, controlled, double blind, parallel-group trial enrolling 195 males (136 latinos, 58 afro-antillean and 69 kunas) and 69 females; with a mean of 45 years old and untreated upper-range prehypertension or type I hypertension. Participants were randomly assigned in blocks by ethnic group to receive for 16 weeks either a cocoa mixture elaborated by MARS containing high level (B) of procyanidins (451.1 mg CP) and a control product with low level (A) of procyanidins (18.2 mg CP). Intention to treat analysis was applied. Results: From baseline to 16 weeks, the mean and 95% CI of change in blood pressure by type of cocoa and group were: Systolic blood pressure men group A = -6.07 (-8.04, -4.10) and in men group B = -5.13 (-7.40, -3.85). Systolic blood pressure women group A = -5.04 (-8.56, -1.52) and in women group B = -7.73 (-11.86, -3.59). Diastolic blood pressure men group A = - 3.34 (-5.03, -1.65) and in men group B = -0.74 (-2.62, 1.14). Diastolic blood pressure women group A = -4.70 (-7.54, -1.85) and women group B = -2.76 (-6.39, 0.882). This differences by high or low flavonol-rich cocoa were not significant, neither the effect on blood pressure by ethnic group. There were only minor
123
IEA-EEF European Congress of Epidemiology 2009 collateral effects like cephalea and hyperactivity in a small number of participants. Conclusion: Flavonol rich and low cocoa reduced the blood pressure and the effect was similar to other clinical trials among European people, without difference by ethnic group or sex.
205 PERFORMANCE OF ANTHROPOMETRIC MEASUREMENTS AND PERCENTAGE BODY FAT IN PREDICTING ELEVATED SERUM LIPIDS R. Meseri1, B. Unal1, R. Cehreli2, R. Ucku1 1Dokuz Eylul University, Faculty of Medicine, Department Public Health, IZMIR, Turkey, 2 Dokuz Eylul University, Institute of Oncology, IZMIR, Turkey Introduction: Obesity is defined as the excessive fat accumulation in the body. Obesity may influence serum lipids and may lead to coronary heart diseases and cancer. Objectives: To evaluate the performance of anthropometric measurements and percentage body fat (PBF) in predicting of elevated serum lipids and cancer. Methods: In this cross-sectional study, all the residents of an urban neighborhood (Teleferik–Balcova, Izmir) who were 30 years or above, not taking any lipid lowering medication were invited to participate in the study. In total 246 men and 637 women interviewed and their weight, height, waist/ hip circumference were measured and anthropometric indices (Body mass index; BMI, waist circumference; WC, waist hip ratio; WHR, waist height ratio; WHtR) were calculated. PBF was determined using a multi-frequency bioelectrical analysis (BIA). Fasting blood samples were collected for blood lipid analyses. Total cholesterol (TC), LDL-C, HDL-C and triglycerides were assessed and categorized using suggested cut-off points. ROC curves of anthropometric indices and PBF for each elevated serum lipid were drawn and the area under the curve was used to assess the performance of the anthropometric measurements. Results: In men, PBF and anthropometric measurements were not successful in predicting elevated TC, LDL-C and triglycerides (P [ 0.05 for all) where BMI, WC, WHtR had ability to predict decreased HDL-C levels significantly (P = 0.006, P = 0.033 and P = 0.042). In women PBF and anthropometric measurements were much more successful in predicting elevated serum lipids especially for triglycerides (area under the curve ranged between 65% and 68%; P \ 0.001). Conclusion: In men, except HDL-C, anthropometric measurements and PBF were not successful in predicting elevated serum lipids. In women, all the anthropometric measurements and PBF successfully predicted elevated serum lipids. In general, WHtR found to be an easy and practical measurement for prediction of elevated serum levels which can be used in routine services.
255 MENOPAUSE INCREASES OBESITY AND SERUM LIPIDS IN WOMEN R. Meseri1, B. Unal1, R. Cehreli2, R. Ucku1 1Dokuz Eylul University, Faculty of Medicine, Department Public Health, IZMIR, Turkey, 2 Dokuz Eylul University, Institute of Oncology, IZMIR, Turkey Introduction: Menopause has a detrimental effect on cardiovascular function and metabolism. Many studies report that obesity and elevated serum lipids increase in postmenopausal women. In Turkey there are not many studies conducted in the community, evaluating affect of menapouse in obesity and serum lipids. Objective: To determine the effect of menopause on obesity and serum lipids. Methods: In this cross sectional study, women participants of Heart of Balcova in Teleferik District were included (n = 700). Obesity and serum lipids were considered as the dependent variables where menopausal status was the independent variable. Obesity was
IEA-EEF European Congress of Epidemiology 2009 evaluated by anthropometric measurements (Body mass index; BMI, waist circumference; WC, waist hip ratio; WHR, waist height ratio; WHtR) and percentage body fat (PBF) by multi-frequency bioelectrical impedance analysis (BIA). Fasting blood samples were collected for blood lipid analyses. Serum lipids (Total cholesterol-TC, LDL-C, HDL-C and triglycerides) and anthropometric measurements were assessed and categorized using suggested cut-off points. Chisquare and t-test were used to test the significance (P \ 0.05) and ROC curves were drawn in order to evaluate the performance of anthropometric measurements and PBF in predicting elevated serum lipids. Results: Among the participants 389 (55.6%) were postmenopausal. Obesity is significantly higher in postmenapousal women (for all anthropometric measurements and PBF P \ 0.01). Mean TC, LDL-C, HDL-C and triglyceride was significantly higher among postmenapousal women (P \ 0.01 for all). Both in premenapousal and postmenapousal women none of the anthropometric measurements and PBF were able to predict elevated TC and LDL-C. In premenapousal women all the antropometric measurements and PBF significantly predict decreased HDL-C and elevated triglyceride. WHtR found to be best indicator in premenapousal women. Conclusion: After menapouse, obesity increases among women which may explain the elevated serum lipids. WHtR can be used in routine service to evalute obesity.
189 EPIDEMIOLOGY OF DIABETIC RETINOPATHY IN EUROPE: META-ANALYSIS BASED ON A SYSTEMATIC LITERATURE REVIEW E. Prokofyeva, R. Wilke, E. Zrenner Institute for Ophthalmic Research, University of Tuebingen, TUEBINGEN, Germany Background: Diabetic retinopathy (DR) is one of the most threatening complications of diabetes. Objectives: To study the epidemiology of diabetic retinopathy in Europe through a systematic literature review. Methods: A literature search was performed using Pub-MED with key words: diabetic retinopathy (DR), prevalence, incidence, populationbased, cross-sectional studies, and epidemiology. Inclusion criteria for the studies were: (a) performed on a healthy population or populations with diabetes mellitus of Caucasian origin at an age between 50 and 75, (b) diagnosed by ophthalmological examination with colour fundus photographs, (c) sample size [ 200, (d) published between 1990 and 2008. Results: 27 studies met the inclusion criteria: one WHO study; three European; two multicentre European; two multicentre within Spain and Germany; one German long-term follow-up study; seven population-based studies from Bulgaria, Germany, Italy, UK, Spain, and Rotterdam; six cross-sectional studies from Germany, France, and Spain; three European studies and two literature reviews. 8% of the global population and 3% of European adults are effected by DR. Prevalence of DR in the population over 60 was the highest in France (16.60%), followed by Germany (10.6%). Incidence of DR was the highest in the UK (43.3%), followed by Switzerland (42.3%), Poland (31.8%), and Germany (29.9%). Men in the range of 60–75 had a higher incidence of DR (29%) than women (16.51%). The mild form of DR (8.50%) was the most prevalent in Germany followed by non-proliferative (1.70%) and proliferative (0.60%). In the UK non-proliferative DR (48%) was more prevalent than proliferative (4%). The same trend was observed in Spain: 38.90% and 5.80%, respectively. Conclusions: France had the highest prevalence of DR in Europe. Incidence of DR was higher in men than in women and was similar across Europe, with the highest level in the UK. Non-proliferative DR was the most frequent in a majority of European countries.
83
274 THE PREVALENCE OF ASYMPTOMATIC PERIPHERAL VASCULAR DISEASE AND ASSOCIATED RISK FACTORS IN THE DORSET POPULATION Tanya Lesiw Bournemouth University, School of Health and Social Care, BOURNEMOUTH, United Kingdom Background: Peripheral vascular disease (PVD) is a result of atherosclerosis in the arteries of the lower limbs. This disease is highly disabling and is a strong marker of risk for major cardiovascular events. Prevalence rates for this disease vary greatly due to the large number of unidentified asymptomatic sufferers in the community. Furthermore it is believed that prevalence rates will be on the increase due to lifestyle choices currently being adopted by many individuals, which have proven favourable to the development of the atherosclerosis including smoking, poor diet and lack of physical exercise. Objectives: The present research aims to determine the true prevalence of asymptomatic peripheral vascular disease in the Dorset population, England. Secondary to the above risk factors of disease will be explored, through the use of a health status questionnaire. Methodology: This research has taken the form of an open-cross sectional observational study, on an age stratified population of males and females aged 40-89 drawn from the south of England. Participants who have no former diagnosis or presenting symptoms of PVD are invited to take part. The methodology employed to screen participants for PVD is a non-invasive Doppler ultrasound. This procedure is inexpensive, risk free and is 95% sensitive and 99% specific in detecting PVD regardless of symptoms. Results: To date a total of 391 participants have been non-invasively assessed for PVD using the Doppler ultrasound. In line with current diagnostic standards an ABPI of less than 0.9 is used to define disease presence. The prevalence of PVD in the sample examined so far is 26% with rates increasing with age. Conclusions: As recruitment and data collection is still ongoing in order to generate a true representation of the local population, both prevalence rates and the key themes identified so far will be discussed.
290 MONITORING CONTROL OF TYPE 2 DIABETES IN PRIMARY CARE IN HUNGARY A. Nagy, E. Szigethy, Z. Voko´, G. Sze´les, A. Keszei, T. Jenei, ´ da´ny Faculty of Public Health, Medical & Health Sciences R. A Centre, University of Debrecen, DEBRECEN, Hungary Background: General practice-based registration systems are used all over the world to monitor morbidity and the performance of primary health care. The General Practitioners’ Morbidity Sentinel Stations Program (GPMSSP) was launched by the School of Public Health, University of Debrecen, Hungary and the National Public Health and Medical Officer Service in May 1998 for these purposes. Objectives: The objective of our study was to investigate the control of diabetes in the primary health care services in Hungary. Methods: The study was performed in the framework of the GPMSSP. The study population consisted of 1466 randomly selected patients older than 35 years with type 2 diabetes. The data collection included data on treatment (specialists involved in care, medication, non-medical treatments), and on complications. The participating GPs performed a physical examination (height, weight, blood pressure, examination of the feet), and drew blood samples. HbA1c level was measured at baseline, at 3 and 6 month. Participants filled in a questionnaire about their socioeconomic status and health behaviour. Results: The control of diabetes was poor, 41% of the study population had a HbA1c level higher than or equal to 7.5%. 22% of the patients had a BMI larger than 35 kg/m2. The prevalence of complications was high already at young age. More
123
84 than 20% of the particpants had coronary artery diasease, and more than 10% had retinopathia at age 35–54 years. Conclusions: General practice based monitoring systems can provide reliable information on the occurrence of major non-communicable diseases and can provide framework for targeted epidemiological and health services research. Our study provided quantitative information on the inefficient control of diabetes in Hungary, which can be used for the development of health services.
245 PRESCRIBING OF ANTIDIABETES THERAPIES IN IRELAND N. L. Zaharan, K. Bennett, J. Feely Department of Pharmacology and Therapeutics, Trinity Centre of Health Sciences, St James’ Hospital, DUBLIN, Ireland Background: Diabetes mellitus (DM) prevalence is on the increase worldwide. Hypoglycaemic agents are prescribed to achieve better glycaemic controls in both type 1 (T1DM) and type 2 DM (T2DM) patients. Aims: (1) To determine the prevalence of T1DM and T2DM in adult population in Ireland (2) To examine the hypoglycaemic agents utilization trends over a 6-year period. Methods: The HSEPrimary Care Reimbursement Services prescribing database was used to identify patients aged C16 years prescribed hypoglycaemic agents from 1st January 2003 to 31st December 2008 (n = 550,318). T1DM patients were those prescribed insulin only while T2DM were those prescribed oral hypoglycaemic agents either alone or combined with insulin. Prevalence was determined using projected national population. Linear trend test was used to examine prescribing trends. Results: Overall, the prevalence of DM in adults increased from 2003 to 2008 (P = 0.004). The prevalence of T1DM was 0.56% in 2008 (P = 0.06). A significant increase in T2DM prevalence was observed (1.67% in 2003; 2.63% in 2008), (P = 0.002). Metformin and sulphonylureas remained the agents of choice for T2DM patients. Increased rate of metformin and decreased rate of sulphonylureas prescribing was observed (P \ 0.001). Prescribing for glitazones decreased after 2005 (P = 0.04). There was also an increase in the prescribing of fixed oral combination therapy (P = 0.02) and insulin (P \ 0.001). Insulin analogues were rapidly replacing human insulin as choice of therapy for T1DM patients (P \ 0.001). Increase in the prescribing of fast-acting and long-acting insulin was observed with reduced prescribing of intermediate-acting and combination insulin. The insulin analogues of choice were insulin aspart followed by insulin glargine. Conclusion: The prevalence of diabetes in Ireland is on the increase. The prescribing for T2DM patients is in line with guidelines with a move towards metformin and increasing use of insulin. Although expensive, insulin analogues are gaining popularity as first choice agents in T1DM patients.
Occupational epidemiology 301 ANALYSIS OF OCCUPATIONAL EXPOSURE TO HARMFUL FACTORS (BIOLOGICAL AGENTS AND IRRITANT GASES) AMONG SWINE CONFINEMENT WORKERS I. Szadkowska-Stanczyk, M. Sowiak, K. Brodka, A. Buczynska, M. Cyprowski, A. Kozajda, Nofer Institute of Occupational Medicine, LODZ, Poland Background: Swine production is characterized by organic dust exposures of high biological activity. It contains the microbial
123
IEA-EEF European Congress of Epidemiology 2009 components that can contribute to the pulmonary diseases of workers. Objectives: Comprehensive exposure assessment to bioaerosols (bacteria and molds), organic dust and irritant gases as well as cross-shift decline in lung function among polish pig farmers was studied. Materials and methods: Occupational exposure assessment was carried out in 30 swine farms in Poland. The questionnaire study and spirometry before and after the shift were done among 90 swine farms workers and 94 non-exposed controls. Organic dust was collected using personal aerosol samplers according to Polish norm. Inhalable and respirable fractions were analyzed. Ammonia and hydrogen sulphide were measured using Draeger pipes. Portable spirometer (Spirolab II) was used for pulmonary function measurements. Crossshift changes in FVC, FEV1 and FEV1/FVC parameters were analyzed. Percentages of cross-shift declines over 5% were compared between exposed and control group. Association between exposure parameters and lung function was correlated. Results: Swine farm workers were exposed to high concentrations of bacteria exceeding recommended hygienic standards (1 9 105 jtk/m3) ten times. However relatively high concentrations of fungi were below these recommendations (5 9 104 jtk/m3). There were found 17 different species of bacteria and 33 of fungi. The worst conditions were found in the barns without the mechanical ventilation as well as where the head of pigs was high (exceeding 500 animals) and when the bedding was used. The results of the study revealed significant cross-shift declines in certain lung function parameters among swine farmers in comparison with the control group. Conclusions: The results of the study revealed significant cross-shift declines in certain lung function parameters among swine farmers. Protection of workers’ respiratory system should be applied.
33 LUNG CANCER MORTALITY IN ARSENICEXPOSED WORKERS FROM A US CADMIUM RECOVERY PLANT T. Sorahan, University of Birmingham, BIRMINGHAM, United Kingdom Background: There is evidence that arsenic is a late-stage human lung carcinogen. Objectives: To investigate lung cancer risks in a cohort of cadmium recovery workers in relation to period from ceasing exposure to arsenic. Methods: The mortality experience (1940–2001) of a cohort of 625 male workers from a US cadmium recovery plant was compared with expectations based on US national mortality rates. Results: There was a statistically significant (P \ 0.05) negative trend in lung cancer standardised mortality ratios (SMRs) in relation to period from ceasing arsenic exposure. Conclusion: The findings are consistent with the hypothesis that arsenic is a late-stage human carcinogen.
95 DOES PERSONAL PROTECTIVE EQUIPMENT PROPERLY PROTECT THE WORKERS?- STUDY AMONG WOMEN WORKING IN VEGETABLEGROWING GREENHOUSE IN POLAND J. Jurewicz, W. Hanke, W. Sobala, D. Ligocka Nofer Institute of Occupational Medicine, LODZ, Poland Purpose: The objectives of the study was to evaluate the protection function of the personal protective equipment and assess the level of the azoxystrobin sprayed for a group of women working in a vegetablegrowing greenhouse. Methods: Exposure samples were collected on two days during the spring: first entry on the day after spraying of
IEA-EEF European Congress of Epidemiology 2009 azoxystrobin and second entry six day later. Dermal exposure was measured by using patches on the outside of clothing and sampling gloves underneath regular working gloves. Measurements of pesticide concentrations in gloves and patches were performed with liquid chromatography and mass spectrometry technique (LC-MS/MS). Results: This study has shown that workers in a Polish greenhouse are indeed exposed to pesticides at re-entry into the greenhouse after spraying pesticides on previous days. Dermal exposure levels to azoxystrobin have been detected on hands, shoulders and chest. Concentrations found on gloves biological protection workers six days after spraying were 10 times lower comparing to the concentrations found after restricted entry intervals expired. The opposite was found for women tending and harvesting vegetables. Conclusions: Women working in a vegetable-growing greenhouse and not directly engaged in the process of spraying have a measurable dermal exposure to pesticides . The personal protective equipment does not properly protect the workers and can be even the sources of the secondary exposure. More efficient personal protective equipment for proper protection of women working in vegetable greenhouses is needed.
135 THE FRENCH NATIONAL PROGRAM FOR POST-OCCUPATIONAL SURVEILLANCE OF SUBJECTS EXPOSED TO ASBESTOS M. Carton1, P. Rolland2, J. Home`re2, M. Nachtigal1, S. Bonnaud1, A. Serrano1, S. Audignon3, P. Brochard3, M. Goldberg1,2, M. Zins1, E. Imbernon2 1Inserm U687, Equipe Risques post-professionnelsCohortes (RPP-C), VILLEJUIF, France, 2Institut de veille sanitaire, De´partement sante´ travail, SAINT-MAURICE, France, 3Universite´ de Bordeaux 2, Isped, Laboratoire sante´ travail environnement, BORDEAUX, France Background and aims: The French national Institutes for Public Health Surveillance (InVS) and for Medical Research (Inserm) have established a surveillance program for retired subjects exposed to asbestos during their working life to identify exposed workers and to propose them a medical surveillance, to describe past exposures and their long-term health effects, and to assess the program in terms of benefits for health and compensation. Methods: Since 2005, two prospective cohorts of retired workers have been set up among former male salaried workers (‘‘Spirale’’ cohort) and self-employed craftsmen (‘‘Espri’’ cohort). In 2008, both cohorts cover one third of the population. Each year, a questionnaire is mailed to new retired workers (50,000 salaried workers and 4,500 craftsmen) to detect past occupational asbestos exposure. The exposure assessment is made by medical and industrial hygiene experts. The medical surveillance is proposed to subjects according to specific exposure criteria. The national medical databases are used for follow-up. Results: During the pilot stage (2005–2008), the participation rate was 24% for salaried workers (without reminder) and 67% for craftsmen (with one reminder). Heavy exposures were found for construction, shipbuilding and fabrication of metal products. About one half of retired craftsmen has been exposed during their working life, versus one quarter of the salaried workers. Among craftsmen who had a chest CT scan, about one quarter showed asbestos-related abnormalities or pathologies. Discussion and conclusions: The first results highlighted the expectation of retired workers for a medical surveillance. Its expansion to the whole territory will include around 250,000 male salaried workers retiring each year and 17,000 self-employed craftsmen. The assessment of the impact of the program on benefits for health and compensation during the cohort follow-up will provide guidance for public policy about surveillance of workers exposed to carcinogenic agents.
85
164 A CROSS-SECTIONAL STUDY IN FARMERS WORKING IN FLORICULTURE P. Liu, J. He, J. Qi, Z. Li, H. Gu, Q. Chen, J. Xing School of Public Health, Kunming Medical University, KUNMING, China Background: Flower cultivation is an economic mainstay in the countryside near Kunming, China. Approximately 70% of the flower-growing land consists of family farms. The flowers are cultivated in the greenhouse and mixed pesticides are widely used. In order to evaluate occupational pesticides exposure and its health effects on farmers working in the floriculture, we had this investigation. Objectives: 211 farmers (male 118 and female 93) working with floriculture were chosen as the exposure group. The control group was another 46 farmers (male 16 and female 30) living in the same village but had not worked in the floriculture at least 5 years. Methods: A cross-sectional study was held. A questionnaire was used to evaluate pesticide exposure and the knowledge about pesticides and self-reported symptoms in farmers. A physical examination was carried out for outcomes of health effects. Results: 74.4% farmers used pesticides more than 40 days in a year. More than half farmers mixed over five kinds of pesticide together. 80% farmers agreed that protection was needed when using pesticides, but few of them had protection and someone mixed pesticides even by naked hands. 70.6% farmers reported some toxic symptoms. After adjusted confounding factors such as sex, age, smoking and alcohol drinking, the numbers of mixed pesticides was associated with the toxic symptoms (r = 0.160, P \ 0.05). The physical examination showed that indexes of many systems between the two groups had no significant differences, but the heart rate of exposure group was significantly lower than that of the control group (P \ 0.01). Meanwhile, the ECG abnormality was much higher in the exposure group than that of the control group (P \ 0.01, OR = 2.39, 95% CI: 0.809–10.747). Conclusion: The farmers working in the floriculture had high risk of pesticide exposure and their cardiovascular system had affected.
200 OCCUPATIONAL DIFFERENCES IN SUBJECTIVE HEALTH AND HEALTH BEHAVIOUR AMONG LITHUANIAN ADULT POPULATION V. Kriaucioniene, J. Petkeviciene, J. Klumbiene Institute for Biomedical Research, Kaunas University of Medicine, KAUNAS, Lithuania Background: Information on occupational differences in life style is important since some occupational groups may be particularly vulnerable to the adverse health effects associated with health behaviour. Objectives: to evaluate occupational differences in subjective health and health behaviour among different occupational groups in Lithuania. Methods: In 1996–2006 six cross-sectional surveys have been carried out within the international Finbalt Health Monitor project. For every survey random sample of 3000 Lithuanians aged 20–64 was taken from the National Population Register. The response rates varied from 59.2% to 74.4%. The study material was collected by mailed questionnaires which included questions about occupation, subjective health and health behaviour. Results: Majority of respondents (40.5%) were officer or services workers, industrial workers 22%, agricultural workers 9.4% and 9% were unemployed. Officer or service employees evaluated their health as good or very good more often then industrial and agricultural workers or unemployed people
123
86 (respectively 52.2% versus 40.82%, 38.0% and 24.71%). The proportion of daily smoking was lower among men working in office or service (32.2%) compared with agricultural (53.6%), industrial (56.8%) workers or unemployed people (53.1%). No occupational differences were found in prevalence of daily smoking among women. Nutrition habits of officer or services workers were more in accordance with nutritional recommendation: they drank less high-fat milk, consumed less meat and meat products, more often used vegetable oil for cooking and more often consumed fresh vegetables and fruits. However officer or services workers were less physical active then others. Proportion of women drinking strong alcohol at least 2–3 times per month and proportion of men drinking bear at least once per week were the highest among officer or services workers. Conclusions: The obtained data on occupational differences in selfreported health and health behaviour should be taken in to account when planning interventions designed to eliminate these disparities.
Environmental epidemiology 52 THE ASSESSMENT OF PAH AND BAP CONTENT IN THE ATMOSPHERIC AIR A. Moz´dzierz, M. Juszko-Piekut, J. Stojko, M. Wyszyn´ska Medical University of Silesia, School of Pharmacy, Department of Hygiene, Bioanalysis and Environmental Studies, KATOWICE, Poland Background: The Upper Silesia Industrial Area, especially in its central part, is both the most industrialized and ecologically degraded area of Poland. Among all monitored air pollutants, the highest values, in comparison to the allowable norm, were recorded for benzo(a)pyrene, which poses one of the biggest health threats for the inhabitants of the area. Objectives:Thus we decided to evaluate the pollution level of atmospheric air by polycyclic aromatic hydrocarbon and benzo(a)pyrene in the Upper Silesia over the years 2000–2006. Methods:In the study, we used the statistical output data obtained from particular measurements taken in the Provincial Sanitary and Epidemiological Station in Katowice. The measurements were taken for the administrative units of the Upper Silesia. We calculated average concentration values for the summer and winter seasons as well as an average annual concentration for the following nine PAHs present in the particulate matter (PM10). PAHs were determined with the use of liquid chromatography. The measurement values of benzo(a)pyrene were compared to the allowable norm of an average annual concentration (1 ng/m3). Results: In the studied period, i.e. 2000–2006, average annual concentrations of benzo(a)pyrene were from 6.6 ng/m3 to 28.4 ng/m3. The annual allowable norm was exceeded in all measurement stations. It should be emphasized that there are huge differences between an average concentration of BaP recorded in the summer and winter (heating) season. The concentration values in the summer season are significantly lower. The median concentrations for the sum of PAHs were from 47.9 ng/m3 to 257.2 ng/m3. Conclusion: Average annual concentration of BaP over the years 2000–2006 in the Upper Silesia Industrial Area exceeded the allowable norm of 1 ng/m3 many times, whereas the sum of PAHs values were alarmingly high. Furthermore, seasonal variability in the values of BaP concentration and the sum of PAHs was observed.
61 URBAN AIR POLLUTION AND EMERGENCY VISITS IN TEHRAN HEART CENTER S. Saadat1, S. Sadeghian2, R. Hamidian3, M. A. Najafi4 1Sina Trauma Research Center, Tehran University of Medical Sciences, TEHRAN,
123
IEA-EEF European Congress of Epidemiology 2009 Iran, 2Heart Center Hospital, Tehran University of Medical Sciences, TEHRAN, Iran, 3Clinical and Experimental Unit, Tehran Heart Center hospital, Tehran University of Medical Sciences, TEHRAN, Iran, 4Pollution Modeling Division, Air Quality Control Company, Tehran Municipality, TEHRAN, Iran Background: Tehran—the capital of Iran—with a population of approximately 7.9 million, only 30% of the days in a year profits healthy air quality. Objectives: The aim of this study was to evaluate the effect of air pollution on the emergency visit, hospitalization and in-hospital mortality of acute coronary syndrome (ACS) patients referred to Tehran Heart Center (THC). Methods: Patients admitted to the emergency ward of THC between December 2006 and October 2007 with confirmed cardiovascular diseases were included in this cross sectional study. Meteorological data were obtained from ‘‘Air Quality Control Corporation Geophysics station’’ which is located about 1000 meters away from THC. A principal component analysis was utilized to extract a linear combination of air pollutants which represented eigenvalue [1. Then, a regression analysis was utilized to determine the air pollution exposure level and the frequency of patients admitted with unstable angina, myocardial infarction and arrhythmia. The principal component linear combination was considered as the exposure level for the regression analysis. Results: During the study period 3525 patients with confirmed diagnosis of ACS were admitted in THC. The principal component score was higher in admission date of patients expired during hospitalization compared to that of patients discharged alive from the hospital (P \ 0.05). After adjustment for the effect of age, sex, smoking, hypertension, hyperlipidemia and diabetes a statistically significant relationship between principal component score at the admission date of patients and the frequency unstable angina (P \ 0.05) and arrhythmia (P \ 0.05) was detected. Conclusion: Occurrence of acute air pollution is associated with increased emergency visit and hospitalization of patients referred with unstable angina and arrhythmia. It is also associated with in-hospital mortality of patients admitted with ACS.
286 PARTICULATE MATTER POLLUTION AND DAILY HOSPITAL ADMISSIONS IN KRAKOW, POLAND K. Szafraniec Jagiellonian University Medical College, Institute of Public Health, Department of Epidemiology and Population Studies, CRACOW, Poland Background: Despite the undertaken measures and the improvement achieved during last several years, Krakow is still considered as one of the most polluted cities in Poland. A question arises whether current levels of air pollution have measurable health impacts? Objectives: To explore the relationship between ambient particulate matter and daily admissions to hospitals in Krakow. Material and Methods: Exposure—daily concentrations of particulate matter (PM10) estimated on the basis of measurements provided by the monitoring network. Measurements taken on the same day of admission and cumulative average over 3 days were considered in the analysis. Health outcome—daily admissions for respiratory and cardiovascular diseases were examined in the all ages, and then in 0–14, 15–64 and greater than 65 years age groups. The study period covered 2001 to 2004. Generalized Additive Models technique with Poisson regression and non-parametric smoothing for time-related confounding factors was applied for statistical analysis. Results: Over the study period an average of 26 residents per day were admitted to Krakow hospitals with a respiratory condition, 0–14 year old children made up 38% of that. An average of 60 persons per day were admitted for cardiovascular conditions, the 60% were over the age of 65. The results of the analysis show significant positive associations between
IEA-EEF European Congress of Epidemiology 2009 PM10 and admissions for cardiovascular and respiratory disease in the all-ages and 65? years groups. Associations between PM10 levels and admissions for asthma were not statistically significant. A 10 lg/m3 increase in mean PM10 was associated with a 0.74% increase in the risk of admission for cardiovascular and with a 1.19% increase in the risk of admission for chronic airways obstruction. Cumulative exposures of 3 days generated larger pooled effect estimates than that of a single-day. Conclusions: The results suggest associations between PM10 and cardio-respiratory hospital admissions, specially in elderly.
Health services research II 85 VARIABILITY IN PRIORITISATION OF PATIENTS WAITING TO BE OPERATED ON CATARACTS S. Garcı´a1, A. Bilbao2, N. Gonza´lez1, I. Lafuente1, U. Aguirre1, C. Las Hayas1, K. Vrotsou1, I. Arostegui3, J. M. Quintana1 1Unidad de Investigacio´n, Hospital Galdakao-Usansolo, USANSOLO, Spain, 2 BIOEF: Fundacio´n Vasca de Innovacio´n e Investigaciones Sanitarias, SONDIKA, Spain, 3Departamento de Matema´tica Aplicada e Investigacio´n Operativa, Universidad del Pais Vasco, LEIOA, Spain Objectives: Since there are not standardized priority criteria, we think that patients with similar clinical and social characteristics could be prioritized differently. Our main goal was to explore the existence of variability in the cataract extraction prioritisation by means of the comparison of the scores in a priority scale which was developed previously by our group and the time in waiting list among participant centers. Methodology: We included 5214 patients older than 18 who were waiting to be intervened on cataracts in 12 hospitals in Spain. Sociodemographical variables were collected (age, gender, education, social support), variables components of priority scale (simple cataract or cataract with other pathologies, pre-intervention visual acuity (VA), contralateral and expected VA, laterality, visual function, appropriateness, social dependence). We also collected technical complexity and time in waiting list. Results: Statistical significant differences were encountered among centers in the time in waiting list (P \ 0.0001) and in the percentage of interventions classified as ‘‘high priority’’ by our priority scale (P \ 0.0001). Diffrerences in waiting times persisted even after adjust by pathology, age, social dependence, laterality and number of inappropriate interventions performed in each center. Differences in percentages og ‘‘high priority interventions’’ dessapeared after adjust by clinical and social variables. Conclusions: There is variability in time spent in waiting list by patients who have to be operated on cataracts. The use of an standardized priority tool could help to decreased it.
222 SELF-PERCEIVED HEALTH, RESTRICTION OF ACTIVITIES AND USE OF HEALTH SERVICES IN PEOPLE WITH MUSCULOSKELETAL DISEASES. THE POPULATION HEALTH SURVEY OF CATALONIA (ESCA 2006) X. Surı´s, M. Larrosa, M. J. Pueyo, J. Auleda, A. Mompart, P. Brugulat Master Plan on Musculoskeletal Diseases, Health Plan Unit1, Health Department of Catalonia, BARCELONA, Spain Introduction: The Health Survey of Catalonia (ESCA) is a population survey held by the Department of Health of Catalonia (over 7 million
87 people) in 2006. Objective: To know the self-perceived health, restriction of activities and the use of health services of people who report musculoskeletal (MS) problems in ESCA. Methods: Crosssectional descriptive study. A total of 15,926 adults were surveyed in a multi-stage stratified sampling, representative of sex and age in Catalonia. Collected variables were sociodemographic characteristics, self-perceived health and quality of life, self-reported chronic health problems and use of health services. MS problems were grouped in four categories: chronic joint disease (JD) (osteoarthritis, arthritis, and rheumatism); chronic low back pain (LBP); chronic upper back pain (UBP), and osteoporosis (OP). Self-perceived health was measured qualitatively and quantitatively with the lower values indicating worse health. Questions were made about the restriction of daily activities and about visits to any health professional in the last year and in the last 15 days. Results: Self-perceived health in those who reported any category of MS problems was significantly worse than the rest of surveyed population, after fitting by age and sex. Health status average was 55 vs 75.8 (P \ 0.001) in chronic JD; 61 vs 75.3 (P \ 0.001) in LBP; 60.8 vs 74.9 (P \ 0.001) in UBP, and 49.3 vs 72.3 (P \ 0.001) in OP. Restriction of activities in the last 15 days, daily activities in the last year and declaration of disability were significantly higher in people with the mentioned problems than in the rest of the population after adjusting by age and sex. People with any of the four problems reported significantly more consultations to the general practitioner, specialists, nurses, physiotherapists and osteopaths. Conclusions: MS problems are associated with worse self-perceived health, more restriction of activities in a short- and long- term and more use of health services.
227 CAUSES AND TYPES OF PALLIATIVE CARE IN SPAIN. A HEALTH SENTINEL NETWORK STUDY ´ . G. de Caso1, M. Margolles2, M. Miralles3, E. Arrieta1, J. A ´ lamo1, O. Zurriaga3, A. Lo´pez3, J. E. Lozano1, P. de la Iglesia1, R. A M. Gil1, T. Vega1 on behalf of the RECENT group 1Consejerı´a de Sanidad de Castilla y Leo´n, VALLADOLID, Spain, 2Consejerı´a de Salud y Servicios Sanitarios, Principado de Asturias, OVIEDO, Spain, 3Conselleria de Sanitat, Comunitat Valenciana, VALENCIA, Spain Background: Palliative care has been defined as the active and comprehensive care of patients with progressive and advanced disease by the treatment of the pain and other symptoms and the provision of psychological, social and spiritual support. The final aim is to obtain the major quality of life for the patient and his family. Objective: The objective of this study is to describe the causes (diseases) and the symptoms and other health problems treated at the end of the life or in patients with a very advanced chronic disease in Spain. Methods: During three months, in 2008, 218 GPs from five Spanish Health Sentinel Networks registered 1,192 home patient visits to provide palliative or support health care. Diseases and symptoms for which care was provided were recorded. Results: Most terminal patients (life expectancy less than 6 months) had cancer (82.9%), circulatory system diseases (35.9%) and injury, poisoning and other external causes (19.6%). Patients with chronic diseases which cause severe functional incapacity but not at the end of life phase suffered from circulatory system diseases (65.6%), nervous system diseases (32.6%) and musculoskeletal system and connective tissue diseases (32.2%). Pain treatment was the main symptom in terminal patients (76.1%) and urinary incontinence in chronic patients (56.8%). Other frequent symptoms that needed palliative care were mental disorders (depression, anxiety) in 41.9% and sleep disorders in 40.7%. Anorexia and asthenia were
123
88 more frequent in terminal patients than in chronic ones with percentages of 43.6% vs. 10.4% and 50.4% vs. 16.6% respectively. Conclusions: Palliative care is necessary in terminal patients as well as in patients with chronic conditions. Although cancer is the main cause in the formers and neurology degenerative diseases in the lasts, both patient types share causes (cardiovascular, metabolic and respiratory diseases) and symptoms (mental and sleep disorder, urinary incontinency etc). Data suggest that the distinction between these patients has not much more sense from the treatment and support care point of view.
126 THE USE OF NATIONAL HEALTH INSURANCE DATA TO DETERMINE HOSPITAL UTILIZATION PROPORTION FOR THE CALCULATION OF POPULATION AT RISK IN HOSPITAL-BASED INCIDENCE STUDIES T. Janevic1, E. Gutterman1, D. Mills1, D. Isaacman2, E. Ujhelyi3 1 Via Research, PRINCETON, United States of America, 2Wyeth Pharmaceuticals, COLLEGEVILLE, United States of America, 3 United Szt. Istvan and Szt. Laszlo Hospital, BUDAPEST, Hungary Background: A common challenge to hospital-based studies of disease incidence is determining the proportion of the population in the surveillance area which utilizes the study hospitals. Household surveys to determine hospital utilization can be inefficient and non-representative of true utilization patterns. Objective: We sought to determine the proportion of the population within the surveillance area, which utilizes five children’s hospitals in two counties in Hungary for an incidence study of invasive pneumococcal disease in children aged 28 days to 5 years. Methods: Data were obtained from the Hungarian National Health Insurance Fund for 2007 for children aged 28 days to 5 years visiting hospitals as inpatients or outpatients in the two counties of Budapest and Pest County. Visits were selected by ICD-10 code for diagnoses of conditions consistent with an underlying etiology of invasive pneumococcal disease (pneumonia, meningitis, bacteremia, unspecified viral infection, unexplained fever, septicemia, peritonitis, and osteomyelitis). A case was considered incident if there was no previous inpatient or outpatient visits by the child for same disease within 30 days. Hospitals were categorized as Study Hospitals and Non-Study Hospitals. The percent of visits to Study Hospitals was calculated by disease, county, and subregions within each county. Results: 56.8% of potential IPD-related visits were to Study Hospitals. Utilization percents varied by county, with 61.7% visiting a Study Hospital in Budapest and 48.9% visiting a Study Hospital in Pest County. Differences were greater by subregion, with utilization percents ranging from 4% to 93%. Utilization percents did not vary substantially by inpatient or outpatient status. Conclusions: Utilization of study hospitals varied greatly within the surveillance area. In countries with a national hospitalization database, proportional hospital utilization can be calculated and used to adjust denominators yielding more valid incidence rates within a surveillance area.
203 STAKEHOLDER ANALYSIS IN COMMUNITYBASED DISASTER RISK MANAGEMENT IMPROVEMENT PROJECT IN RURAL AREAS OF GOLESTAN PROVINCE, IRAN IN 2008–2009 E. Ahmadnezhad1, S. Hatami2, F. Fayyaz-Jahani3, N. Sepehrvand2, A. Ardalan1, K. H. Naieni1, A. M. Zanganeh4, A. Keshtkar4,
123
IEA-EEF European Congress of Epidemiology 2009 A. Nazari4 1Tehran University of Medical Sciences, TEHRAN, Iran, Urmia University of Medical Sciences, URMA, Iran, 3Zahedan University of Medical Sciences, ZAHEDAN, Iran, 4Golestan University of Medical Sciences, GOLESTAN, Iran
2
Stakeholder is an individual, community, group or organization which influences on outcomes of the projects or is influenced from. Stakeholder analysis is a technique to identify and assess the importance level of Key individuals, groups and organizations which could affect the successful implementation of the projects. A stakeholder analysis was performed in order to more precisely implementation of Community-Based Disaster Risk Management improvement program in the rural areas of Golestan province. We used the suggested method of WHO in analyzing. First, the main stakeholders identified by brain storming, and then by using personal interviews and snowball method, the other effective individuals and organizations were listed. Using Matrix and Map drawing method, finally the assessment of the impact and potential strategies for obtaining support or reducing obstacles of organizations and groups were listed and analyzed. Generally two organizations and four groups were identified as stakeholders of this study. Both two organizations took the A score, two groups took B score, and the other two took C score in the assessment of their impact. The prior successful experience of similar projects in this province and giving the right of choosing study villages to organizational stakeholders were some potential strategies for obtaining support. Involving native expert colleagues and appropriate history of corporation among two organizations played a positive role in implementing project. Involving native people of rural areas as stakeholder group also played a positive role in execution and continuing the project. Stakeholder analysis is necessary before performing health system projects. Stakeholder analysis in this projected resulted in more accurate determination of study villages, and approach to all hazards of the region including earthquake, flood and fire. The intervention plan was finalized after performing stakeholder analysis.
Perinatal epidemiology II 92 MULTICENTER, POLISH NATIONAL MOTHER AND CHILD COHORT STUDY K. Polanska, W. Hanke Nofer Institute of Occupational Medicine, LODZ, Poland Background: Exposures during prenatal and early postnatal period have implications for children’s health and may also have implications for morbidity and mortality occurring later in life. Prospective cohort study design allow for the identification of potential exposures that may influence pregnancy outcome and children’s health, verification of such exposures by biomarkers measurement and notification of any changes in exposure level. The aim: The study concentrate on the identification and evaluation of the effects of prenatal environmental exposure on pregnancy outcome and children’s health. Methods: Polish Mother and Child cohort study is the prospective population-cohort study conduct in 8 different regions of Poland. Whole cohort of 1300 mothers-child pairs will be established within 3 years period (2007– 2010). The recruitment and all scheduled visits are conducted in maternity unit or clinics in participated in the study districts. We include into the study women between 8–12 weeks of single pregnancy, not assisted with reproductive technology, and not expected to be finished as spontaneous abortion. All women with the serious chronic diseases specified in study protocol are excluded from the study. All women who agreed to participate in the study are interviewed by obstetrician and/or midwife using detail questionnaire. During that visit saliva and blood sample is collected. The second visit scheduled between 20 and 24 weeks of pregnancy include all elements from the
IEA-EEF European Congress of Epidemiology 2009 firs visit and additionally collection of urine sample. Third examination between 30–34 weeks of pregnancy contains: third questionnaire and collection of saliva, urine, blood and hair samples. At the time of delivery blood sample from the mother and cord blood is sampled. After delivery pregnancy outcome is notified and about 2 weeks postpartum breast milk is collected. Results: Up today the cohort comprise of 660 recruited pregnant women. All planned biological samples was collected from 269 women and from 60 women breast milk was collected.
150 CHALLENGES IN BELARUSIAN YOUTH REPRODUCTIVE POTENTIAL M. Surmach Grodno State Medical University, GRODNO, Belarus The official statistics testifies about negative medical-demographic tendencies in Belarus. While the infant mortality rate reached the value of the developed countries of Europe, the abortion’s level still remains high (45 on 100 live-births). Chronic pathology rate grows. Syphilis incidence rate level, despite depression, exceeds that for 1991 in 2 times. Transfer rates of HIV as a result of heterosexual contacts increase. The purpose of work was to analyze a condition of reproductive potential of young Belarusian women. The analysis of sociological interrogation materials got from 1500 women of 15–24 years old as well as medical documentation of 2022 pregnant women and women in childbirth of the same age was carried out. Territorial distribution of sample was planned according to distribution of Belarusian youth. Results show, that only 5.8% of young pregnant women and women in childbirth have no pathology. More than a half have a pathology of several organs and the systems, every fifth—diseases of genitals. About third the first pregnancies at young women terminate in artificial abortion. Almost 22% of the interrogated women’s first-borns were born as a result of pregnancy estimated by the women as untimely and undesirable. Even in case of realization of the quantity of children planned by youth as ideal number of children in a family (2.20 ± 0.02 the child for one woman), birth rate level can provide only simple reproduction. Despite appreciable successes of public health services activity in the sphere of motherhood and childhood protection, data testify about critical condition of young women living in Belarus reproductive potential. It is obvious that public health services, while dealing with challenges in Belarusian youth reproductive health protection, should change priorities from oriented towards simply medical result to the socially proved dimensions.
228 MATERNAL MORBIDITY AND MODE OF DELIVERY: A COHORT STUDY IN ITALY S. Farchi, A. Polo, F. Franco, D. Di Lallo Public Health Agency, Lazio region, ROME, Italy Background: Studies on maternal outcomes and mode of delivery report contradictory results. Objectives: Aim of this population-based study is to estimate the risk of maternal complications and deaths associated with mode of delivery in a population of low-risk pregnancies. Methods: Data on 324,883 at term singleton infants born in 2001-2007 were retrieved from birth and hospital discharge database of Lazio region, central Italy. Breech presentation births and chronic and pregnancy-related diseases in the mother and fetus were excluded. The remaining 273,789 births were analysed. Maternal outcomes
89 included: haemorrhage, Rupture of uterus, anaesthetic complications, Obstetric shock, Major puerperal infection, hysterectomy and Transfusion of blood. The association between mode of delivery (vaginal, vaginal operative, elective caesarean section and in labour caesarean section) and maternal outcomes was estimated by Odds Ratios (ORs) adjusted by maternal age, education, parity and country of birth. Results: The 62.8% of births were vaginal deliveries, the 2.4% were born by operative vaginal delivery, the 20.9% by elective caesarean section and the 13.8% by in labour caesarean section. Incidence of haemorrhage was 4.1/1000 deliveries, blood transfusion 1.8 and incidence of hysterectomy was 0.3. With the exception of risk of haemorrhage, lower in caesarean deliveries compared to vaginal (OR = 0.29; 95% CI: 0.20–0.42 for elective caesarean and OR = 0.40; 95% CI: 0.30–0.54 for in labour cesarean), women undergoing caesarean delivery had increased risk of severe maternal morbidity compared with women undergoing spontaneous vaginal delivery (hysterectomy: OR = 2.90; 95% CI: 1.20–7.00 for elective caesarean and OR = 3.93; 95% CI: 1.75–8.81 for in labour caesarean and obstetric shock: OR = 7.14; 95% CI: 2.99–17.06 for elective caesarean and OR = 3.19; 95% CI: 1.04–9.75 for in labour caesarean). Conclusions: Cesarean delivery, even an elective one, carries a significantly higher risk of severe maternal complications. Our results suggest the necessity of a depth evaluation of risks and benefits of the choice of mode of delivery.
229 IS THERE A DIFFERENTIAL MORTALITY OF OUTBORN STATUS ACCORDING TO LEVEL OF CARE OF HOSPITAL BIRTH AMONG VERY PRETERM BABIES? D. Di Lallo, F. Franco, S. Farchi., P. Pezzotti, A. Polo Public Health Agency, Lazio region, ROME, Italy Background: There are evidences that outborn very preterm births have higher risk of mortality compared to inborn, but few studies investigated on the association between outborn status by perinatal level of hospital of birth and mortality. Objective: The aim of this study is to estimate mortality of very preterm babies (less then 32 weeks of gestation or \1500 gr) admitted to NICUs by outborn status. Methods: Babies less than 32 weeks of gestation or \1500gr. admitted in NICUs in the Lazio region in 2007–2008 were considered. Data were extracted by the NICU Lazio network database. The information on outborn status was classified as inborn (births in an hospital with a NICU and never transferred), outborn from level I–II, and outborn from NICU. To estimate the 90 day mortality risk by outborn status, cox regression was applied adjusting for perinatal factors (gender, congenital malformations, apgar5 min, antenatal steroids and gestational age). To take into account their low probability to be transferred, babies deaths within the first day (n = 64) were excluded from the regression. Results:The study population was 1220 babies. The 27% were at 22–27 weeks of gestation, 59% were at 28–31 and 14% were at 32 weeks or more and \1500 gr.. According to outborn status, the 70.5% were inborn, 14.2% were outborn from level I–II and 15.3% were outborn from NICU. The in-hospital mortality rate was 16.8% ranging from 43.5% in 22–27 weeks babies to 5.1% in the 32 weeks or more and \1,500 g babies. Compared to inborn infants, babies transferred from level I–II and those transferred from NICU had two-fold risk of 90 days mortality (HRI-II level = 2.01; 95% CI: 1.29–3.31), (HRNICU = 2.46; 95% CI: 1.63–3.71). Conclusions: In our region we observed that about 30% of very preterm babies are transferred and that the risk of mortality associated to outborn status is independent from the hospital level of birth.
123
90
288 MATERNAL DIET THROUGHOUT PREGNANCY AND FOETAL GROWTH TRAJECTORIES E. Pinto1,2, M. Severo1,2, A. Cunha3, B. De Stavola4, I. dos Santos Silva1,4, H. Barros1,2 1Department of Hygiene and Epidemiology, University of Porto Medical School, PORTO, Portugal, 2Institute of Public Health, University of Porto, PORTO, Portugal, 3Ju´lio Dinis Maternity Hospital, PORTO, Portugal, 4 Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, LONDON, United Kingdom Background: Only weak associations between maternal diet during pregnancy and birth size have been reported in well-nourished populations, but the effect of maternal diet on foetal growth trajectories has not been previously examined. Objective: To investigate the effect of maternal diet on foetal growth trajectories. Methods: 218 singleton pregnant women from ‘‘Gerac¸a˜o XXI’’, a Portuguese birth cohort, were followed throughout pregnancy and their diet ascertained by repeated administrations of a validated semi-quantitative FFQ. Foetal weight was estimated from ultrasonographic foetal measures taken after the 17th gestational week (median number of scans/ woman: 2; range 2–6) using the Hadlock’s equation. Mixed-effect models were fitted, assuming gestational age as time variable, taking foetal growth as the outcome, nutrient intakes as the exposures, and maternal age, maternal smoking and foetal gender as confounders. The bottom and top fifths of the intake distributions were compared with the intermediate ones (= reference). Results: Median (25–75% percentiles) daily energy intake, birth weight and gestational age were, respectively, 2432 kcal (2056–2828), 3212 g (2860–3480) and 39 weeks (38–40). Total energy, protein and carbohydrate intakes were not associated with any measure of foetal growth. The effect of fat intake was modified by gestational age, with foetus of women in the top intake fifths of total fat (P = 0.06), saturated fat (P \ 0.01), and trans—(P \ 0.01) and omega-6 fatty acids (P \ 0.01) having a similar weight as those of women in the other fifths up to gestational weeks 20–25, but an increasing lower weight after that (*8.5 grams less weekly). Similar effect modifications were observed for abdominal circumference, and less strongly, for head circumference and femur length. Conclusions: Maternal total energy, protein and carbohydrate intakes had no effect on foetal growth. Highest intakes of total fat, and certain of its components, were associated with a slow down in foetal growth in the last half of pregnancy.
Children and adolescent health epidemiology II 32 NUTRITIONAL KNOWLEDGE, HEALTHY BEHAVIOR AND SELF EVALUATION OF BODY WEIGHT IN A SAMPLE OF ROMANIAN ADOLESCENTS C. A. Zugravu1, C. Rada2, G. Nitu1 1University of Medicine and Pharmacy ‘‘Carol Davila’’, BUCHAREST, Romania, 2 The Institute of Anthropology ‘‘Francisc I. Rainer’’, BUCHAREST, Romania Background: Adolescence is an age when healthy lifestyle knowledge has great chances to be established as an ally for the entire life. Programs focused on teenagers have to take in consideration the particularities of their behavior, in order to become efficient. Objectives: We investigated the nutritional/lifestyle knowledge and the self evaluation of the BMI, in a randomly selected sample of 1201
123
IEA-EEF European Congress of Epidemiology 2009 highschool teens from 10 Romanian towns, all learning in mathprofile schools. The sample is quite homogenous regarding the socioeconomical background (middle class). Methods: The descriptive cross sectional study was based on a self-completed questionnaire with 26 items and on the evaluation of weight and height of each participant, at the schools‘ nurseries. The results were analyzed by SPSS program. Results: Nutritional knowledge regarding each food group is generally correct. The suitable level of daily physical activity is underestimated, over a half (65.8%)of the teens having a sedentary life, mostly due to the overloaded school schedule. A percent of 17.3 smoke. The main sources of lifestyle information are Internet and TV programs(65.5%). Regarding BMI, 69% of the sample are between the 5th and the 85th percentile (WHO percentiles for children and adolescents), 27.6%, under the 5th percentile and 3.4%,between the 85th and the 95th percentile. Almost a half has an incorrect image of the body weight (think being overweighed) and wishes to get thinner. The body image discrepancy is significantly associated with underweight (P \ 0.01), sex (female, P \ 0.01), has no connection with age, smoking or overexercising and is connected with an insufficient food intake (eating less in order to lose weight : P \ 0.01). Conclusions: Even if well educated adolescents have a good nutritional knowledge, a special educational effort has to be paid in order to teach them to evaluate correctly their BMI so that the probability of eating disorders onset can be avoided.
67 QUALITY OF LIFE PROTOCOLS IN EARLY ASTHMA DIAGNOSIS IN CHILDREN M. Farnik1, W. Pierzchała1, G. Bro_zek2, J. E. Zejda2 1Department of Pneumonology, Silesian University of Medicine, KATOWICE, Poland, 2Department of Epidemiology, Silesian University of Medicine, KATOWICE, Poland Introduction: Asthma is considered an important problem in children and there is still need for early intervention strategies. The purpose of this study was to find out the impact of asthma on the health related quality of life (HRQL) of newly recognized pediatric asthma patients and their parents. Methods: Patients were identified by a general respiratory survey. HRQL was assessed by a disease-specific questionnaire for children, the Pediatric Asthma Quality of Life Questionnaire (PAQLQ) and general survey for parents, the Quality of Life in Child’ Chronic Disease Questionnaire (QLCCDQ). Results: 29 newly diagnosed asthma children aged 7–11 (mean age 8.65 SD = 1.39) were identified from 1822 based on random population sample. QLCCDQ—emotional domain showed a correlation to patient symptoms of wheezing (P \ 0.05) and dyspnea (0 \ 0.05). QLCCDQ symptom perception domain correlated with wheezing (P \ 0.05), while cough was the only symptom that correlated with PAQLQ—emotion domain (P = 0.01). Baseline spirometry showed correlation with HRQL—FEV1 %FVC and PAQLQ symptom domain (r = 0.42). More correlations were obtained during the exercise test: FEV1% FVC correlated with QLCCDQ emotional domain (r = 0.7); MEF25-75% correlated with family roles (r = 0.81) and social functioning (r = 0.81). Activity limitation on the PAQLQ correlated with the MEF 25–75% (r = 0.44) and PEF (r = 0.55). Conclusions: The most impaired HRQL domains differ in parents and children. In children activity limitation was the most impaired; while emotional functioning and symptom perception domains were most impaired in parents. Parents showed more correlations between the patient’s clinical status and the HRQL. The impact of disease on everyday child’ and parental functioning can provide valuable data for early management.
IEA-EEF European Congress of Epidemiology 2009
82 VALIDATION OF A CHILDHOOD ASTHMA SCREENING QUESTIONNAIRE J. E. Zejda1, A. Wypych-S´lusarska2, G. Bro_zek1, M. Kowalska1 1 Department of Epidemiology, Faculty of Medicine, Medical University of Silesia, KATOWICE, Poland, 2Department of Epidemiology, Faculty of Public Health, Medical University of Silesia, KATOWICE, Poland Background: Epidemiological evidence on the prevalence of childhood asthma justifies implementation of effective population-based screening programs. Standard questionnaires, however, have been developed to address research oriented goals. There is a need for short and duly validated screening questionnaires. Objectives: The goal of the study was to design and validate a short questionnaire, used in childhood asthma screening programs. Methods: A short (10 questions) questionnaire was designed according to the results of a crosssectional study on respiratory health in 6420 children aged 6– 16 years. All selected questions met the best sensitivity criterion (questionnaire-derived diagnosis of asthma as a ‘‘gold standard’’). The developed questionnaire was validated in a group of 180 children (direct clinical diagnosis of asthma by physician as a ‘‘gold standard’’). The procedure involved assessment of repeatability and reliability (sensitivity, specificity, Youden’s index), according to three scenarios: Scenario 1: category ‘‘asthma ? probable asthma’’ versus category ‘‘no asthma’’; Scenario 2: category ‘‘asthma’’ versus category ‘‘no asthma’’ ? ‘‘probable asthma’’; Scenario 3: category ‘‘asthma’’ versus category ‘‘no asthma’’. Results: Repeatability of the screening questions reached 80.6–91.9% level (kappa values: 0.14– 0.70). Scenario 2 provided the best reliability (chest wheeze ever: sensitivity = 91.7, specificity: 74.4, Youden’s index: 0.66, followed by attacks of dyspnea and history of obstructive bronchitis). Assessment agreement of screening questionnaire and direct medical diagnosis revealed a good agreement (77.3%). Conclusions: The designed and validated short screening questionnaire on childhood asthma has a good repeatability and reliability, thus it could be applied in childhood asthma population-based screening programs.
91 PRENATAL AND CHILDHOOD EXPOSURE TO PESTICIDES AND NEUROBEHAVIORAL DEVELOPMENT: REVIEW OF EPIDEMIOLOGICAL STUDIES J. Jurewicz, W. Hanke Nofer Institute of Occupational Medicine, Department of Environmental Epidemiology, LODZ, Poland Objectives: Conventional pesticides comprise a diverse group of substances intended to destroy, repel or control organisms identified as pests. Compared to studies of lead, mercury and PCBs few epidemiological studies have assessed the developmental neurotoxicity of pesticides. Materials and Methods: Epidemiological studies focused on the neurobehavioural development of children exposed to pesticides were identified by a search of the PUBMED, MEDLINE, EBSCO, AGRICOLA and TOXNET literature bases. Results: The results from the presented studies suggest that children exposure to pesticides may impaired development. Exposure to organophosporus pesticides (OP) in children can caused difficulties with tasks involving short-term memory, incrased reaction time, mental development and pervasive development problems. In neonates increased number of abnormal reflexes and mental and emotional symptoms in adolesecents. The results of the studies investigated the association between exposure to organochlorine pesticides and neurodevelopmental problems display inconsistent results. Whereas some studies found
91 reduction in mental and psychomotor function the other studies did not confirmed that. Conclusion: Information derived from epidemiological studies so far indicate the need to increase awareness among people and children exposed to pesticides about the association between use of pesticides and neurodevelopmental impairment. We should simply apply the principle of prudence, just in case.
177 PHYSICAL ACTIVITY AND ALCOHOL USE IN ADOLESCENCE: WHICH MEDIATORS EXPLAIN THE INDIRECT CAUSAL RELATION? M. Scatigna1, K. Panopoulou1, I. Carosi1, R. Gigante1, G. Sementilli1, F. Vigna-Taglianti2, S. Vadrucci2, L. Fabiani1 1University of L’Aquila, L’AQUILA, Italy, 2Piedmont Centre for Drug Addiction Epidemiology, GRUGLIASCO, Italy Regular practice of Physical Activity (PA) certainly provides physical, mental and social benefits in young people. However evidences on association between PA or sport and unhealthy behaviours in adolescents (tobacco, alcohol and drugs use, doping, eating disorders, violence) aren’t conclusive. Our aim is to describe the relation between different level of PA involvement and alcohol use/misuse in adolescents, analysing the mediation effects due to individual and environmental intermediate variables. A cross-sectional study has been performed in 2006 involving 3,488 Italian students of Abruzzo Region (48.5% males) 14.5 years aged on average. Physical adolescents’ profile has been referred to Organized PA (OPA) and each subject classified as ‘‘No-active’’, ‘‘Low-active’’ or ‘‘High-active’’. The outcome’s measures collected on alcohol are: actually daily drinking, actually at least weekly drinking and almost one time to get drunk in the last month. Moreover, the level of alcohol consumption expressed in a numeric scale (0–2 score). The indirect causal relation between OPA and alcohol consumption levels was studied by mean of mediation analysis. Crossing data on PA habits and alcohol consumption it appears a possible opposite influence depending by level of involvement, with some differences between genders. In males, only daily drink is significantly related to OPA-involvement, in protective way. In females High-level of OPA appears significantly expositive to higher alcohol consumption and misuse, while Low-level seems protective especially vs drunkenness. On the whole, intention to drink within 1 year, social norms on peers’ drunkenness occurrence, refusal skills, academic achievement orientation are worst in OPA High–active adolescents and they are significant mediators of alcoholic use more in males than in females. On the contrary self esteem is better and it’s a suppressor for both genders. The value of this work stays in individuating intermediates variables that could mediate preventive interventions in school and sport settings.
94 ENVIRONMENTAL TOBACCO SMOKE EXPOSURE AND PSYCHOMOTOR CHILD DEVELOPMENT K. Polan´ska, W. Hanke, W. Sobala, D. Ligocka Nofer Institute of Occupational Medicine, LODZ, Poland Background: In Poland about 30% of children is exposed to tobacco smoke coonhounds during prenatal period and more than 50% in postnatal one. Such exposure has serious health consequences including negative effect on child neurodevelopment. The aim of the study was to assess the effect of environmental tobacco smoke (ETS) exposure on psychomotor child development. Material and methods:
123
92 The study population consisted of 63 children with well assessed prenatal exposure to environmental tobacco smoke (three times analysis of cotinine level in saliva of pregnant women). Assessment of child ETS exposure within 1 year after birth was based on questionnaire conducted with mothers. The Bayley Scale for Infant and Toddler Development (BESID-III) was used for the evaluation of child neurodevelopment. Results: Multivariative analysis (including gender, birth order of the child and parental educational status) indicated the statistically significant association between prenatal exposure to ETS and cognitive child development (b = -4.0; P = 0.04). ETS exposure has also negative impact on motor (b = -2.7; P = 0.2) and language (b = -3.4; P = 0.08) abilities of the child although the results were not statistically significant. Conclusions: Maternal smoking was found to be related to a decrease child neurodevelopment although it impossible to separate the prenatal from postnatal exposure. All effort should be taken to eliminate the child ETS exposure.
153 EVALUATING SURVEILLANCE OF INVASIVE PNEUMOCOCCAL DISEASE AMONG CHILDREN IN PROSPECTIVE HOSPITAL-BASED STUDIES IN THREE CENTRAL EASTERN EUROPEAN COUNTRIES E. M. Gutterman1, T. Janevic1, R. M. Alexander2, B. Czech-Szczapa3 1 Via Research, PRINCETON, United States of America, 2 Wyeth Pharmaceuticals, COLLEGEVILLE, United States of America, 3Department of Preventive Medicine, Faculty of Health Sciences, Poznan University of Medical Sciences, POZNAN, Poland Background: Prospective surveillance within a hospital-based system can be used to estimate disease incidence. Quantification of enrollment effort is rarely described, although failure to identify eligible patients and other non-enrollment may lead to underestimated incidence. Objective: To determine proportional enrollment among a representative sample of eligible subjects in a prospective surveillance study. Methods: A strategy to evaluate proportional enrollment among eligible children was developed during a hospital-based, active surveillance study of children with invasive pneumococcal disease (IPD) conducted in eight hospitals within three countries in Central Eastern Europe. For randomly selected days, hospitals prepared computerized lists of children presenting to in- and outpatient departments, including: child’s initials, birthdate, residence, assigned ICD-10 diagnoses, time, and hospitalization decision. Diagnoses associated with most clinically eligible children and, therefore, included in the review were: septicemia (A41), pneumonia (J18), bronchitis (J20), bronchiolitis (J21) and fever unspecified (R50.9). Children meeting review criteria were identified and cross-referenced to children included in study log. For children not in study log, chart review by hospital staff eliminated children who visited hospital for follow-up or lacked other clinical criteria (e.g. bronchitis without fever [ 39F). Among total eligible children by hospital, outcomes were percent enrolled, missed, refused consent or other. Results: Among eligible children at the 8 hospitals, the numbers and percents enrolled were: 18/19 (94.7%), 30/33 (90.9%), 6/7 (85.7%), 11/14 (78.6%), 16/23 (69.6%), 47/75 (62.7%), 36/60 (60.0%) and 65/130 (50.0%). In the hospital that enrolled 50% of eligible children, enrollment among hospitalized children was 76.8%, while negligible enrollment occurred among children treated as outpatients. Conclusions: Although evaluation results were limited to selected ICD-10 diagnoses, the proportional enrollment results ranging from excellent to problematic, can be used to quantify extent of underestimation.
123
IEA-EEF European Congress of Epidemiology 2009 Proportional enrollment by weekend/weekday, diagnosis and hospitalization decision further clarified gaps and framed strategies to upgrade surveillance.
Infectious diseases epidemiology II 99 HIV DIFFERENTIALS IN INDIA: A CONCERN A. Verma, Regional Medical Research Centre for Tribals, JABALPUR, India Background: Out of 5 million HIV cases in South Asia, India contributes 2.4 million cases. Overall, around 0.3% of India’s population is living with HIV. Mere 0.1% increase in HIV prevalence will increase the estimated number of people living with HIV by over half a million. Further people living with HIV in India come from incredibly diverse backgrounds, cultures and lifestyles. It is difficult to combat HIV/AIDS; and emerged as a threat to achieve the targets of Millennium Development Goals (MDG). Objectives: Estimate the prevalence of HIV according to their socio-demographic status. Methods: National Family Health Survey-phase 3 (NFHS-3) was conducted during 2005–2006 through out the country. 109,041 households were studied, which covers 99 percent of India’s population living in all 29 states. The data is analyzed according to their socio-demographic characteristics, and some of them observed to be statistically significant. Results: Around 1.707 million adults are infected from HIV. The HIV prevalence rate is higher in urban areas. HIV prevalence rates are higher for men in every age group except age 15–19, having no education. HIV prevalence is highest for Buddhist/Neo-Buddhists, Christians. The highest HIV prevalence are for widowed/ divorced. Higher percent of uncircumcised men are HIV infected. Differentials by caste/ tribe are quiet small. Despite tribe endogamy, HIV is found among the tribal population. Knowledge of HIV/AIDS prevention methods differs markedly between women and men age 15–49. Few adults know each of the three ABC methods of prevention and comprehensive knowledge of HIV/AIDS. Less than half of the women age 15–49 and almost two thirds of men know that HIV can be transmitted from a mother to her baby, A few women and men know about the use of drugs. Conclusions: To attain the MDG goals, the IEC activities are required for prevention of HIV.
38 TIME SERIES CROSS-CORRELATION ANALYSIS OF HIV SEROPOSITIVITY AND PULMONARY TUBERCULOSIS AMONG MIGRANTS ENTERING KUWAIT S. Akhtar1, H. GH. H. Mohammad2 1Department of Community Medicine and Behavioural Sciences, Faculty of Medicine, Kuwait University, SAFAT, Kuwait, 2Ports and Borders Health Division, Ministry of Health, RUMAITHIYA, Kuwait Background: There is a paucity of published data on burden and pattern of dual infection with M. tuberculosis and HIV among migrants from South Asia, South-east Asia and sub-Saharan Africa entering Middle-East particularly Kuwait. Objectives: to assess the overall prevalence of HIV infection and pulmonary tuberculosis and evaluate ecological relationship between them. Methods: Time series cross-correlation analysis was used to determine the ecological timelagged relationship between the monthly proportions (per 100,000) of HIV seropositive and pulmonary tuberculosis cases among
IEA-EEF European Congress of Epidemiology 2009 migrant workers entered Kuwait from January 1, 1997 to December 31, 2006. Results: During the study period, overall prevalence (per 100,000) of HIV seropositivity and pulmonary tuberculosis among the migrants was 21 (494/2328582) (95% CI: 19–23), and 198 (4608/2328582) (95% CI: 192–204) respectively. Estimated crosscorrelation function revealed a significant positive correlation (0.292 ± 0.093) at lag-3 representing a positive relationship between the proportions of HIV seropositive (per 100,000) migrants tested 3 months earlier and the proportion of pulmonary tuberculosis (per 100,000) cases detected among migrants in a given month. Thus, the peak in proportion of pulmonary tuberculosis cases preceded the peak in proportion of HIV seropositive migrants indicating a direct time-lagged association between HIV seropositivity and prevalence of pulmonary tuberculosis among migrants. Conclusions: HIV infection seemed to have played a significant role in the reactivation of latent M. tuberculosis infection in this migrant population. While currently less evident, in near future however, tuberculosis and HIV/AIDS control programmes in the countries of origin of migrants may face a crucial challenge. Knowledge of serious consequences of association between HIV infection and pulmonary tuberculosis allows the promotion of public heath education to reduce the exposure to these infections. Future studies may focus on evaluating the impact of public health education program on this dual burden of HIV infection and tuberculosis in migrants
56 PREVALENCE OF HIV AND HEPATITIS C VIRUS INFECTIONS AND RELATED BEHAVIORAL DETERMINANTS AMONG INJECTING DRUG USERS REFERRED TO DROP-IN CENTERS IN IRAN A. Mirahmadizadeh, R. Majdzadeh, K. Mohammad, M. H. Forouzanfar Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, TEHRAN, Iran Background: Drop-in centers (DICs) are established to reduce harms of high-risk behavior in high-risk groups especially in injecting drug users (IDUs). This study conducted to determine high-risk behaviors, Hepatitis C virus (HCV), and human immunodeficiency virus (HIV) prevalence in IDUs seeking harm reduction services in Iran. Method: 15 centers of 48 centers covered by universities of medical sciences were selected with systematic random sampling. Information about demography, high-risk behaviors, and HIV and HCV infections of IDUs were obtained by interview. Results: Among 1531 subjects that were analyzed 96.1% were male, 47.8% were single, and 35.1% were married. The mean age of participants was 33 ± 9.0 years and the most prevalent age group was 26–35 years (45%). The median of duration of drug abuse and injection were 12 (range of 1 to 43) and 5 (range of 0 to 37) years respectively. 20.5% (95%CI: 17.94, 23.22) of participants were HIV positive test results, while prevalence of HCV was 43.4% (95%CI: 40.17, 46.62). The median number of injections were 21 per week for all injections and zero for injection with shared needles and syringes. 47.4% of the IDUs had sexual contact with someone other than their spouse. The proportion of unprotected homo and heterosexual contacts among IDUs were 19.4% and 37.4% respectively. Conclusion: IDUs have high rate of unsafe sexual contact and injection relating behaviors. High prevalence of HIV and HCV infection among this group implies high rate of transmission and exposure to risk of serious diseases. The study showed necessity of establishing and developing
93 harm reduction support to the majority of IDUs, to reduce transmission and burden of HIV and hepatitis C in Iran.
163 EFFECT OF HEPATITIS B IMMUNIZATION AMONG 1–10 YEARS OLD CHILDREN IN ZAHEDAN—IRAN M. Hashemi Shahri1, F. Fayyaz-Jahani1, S. Keyvanlo1, E. Ahmadnezhad2 1Zahedan University of Medical Sciences, ZAHEDAN, Iran, 2Tehran University of Medical Sciences, TEHRAN, Iran Introduction: Hepatitis B is still a major Health problem in both developing and developed countries. It can become chronic and it may result in Hepatocellular carcinoma or cirrhosis. The incidence of becoming chronic is more in younger adults. The hepatitis B vaccine is essential in preventing HBV infection and lowering its chronicity, however, the immune response influenced by various factors. Thus, in the present study we have evaluated the immunological response to hepatitis B vaccination among 1–10 years old children in Zahedan. Method: We enrolled 315 randomly selected 1–10 years old vaccinated children in this cross-sectional descriptive study. Serum samples were collected via venupuncture and tested for HBsAg, HBsAb, HBcAb via ELISA method. Result: The protection percentage of HBsAb was 88.5%, 85.7% and 40.9% during the 1st, 6th and 10th year after vaccination, respectively. Seroprevalence of HBsAg ? and HBcAb in all groups was 0.95 percent. The vaccine response was 0.95% compared with 5% of Healthy blood donors. Conclusion: According to these findings, universal Hepatitis B vaccination program at birth provides adequate protection against Hepatitis B virus infection at least for 10 years in both sexes, comparing the percentage of HBs Ag positive (0.95 in vaccinated children) with the results of healthy blood donors’ shows efficacy of vaccination. It is concluded that booster vaccination is not recommended, but further follow up studies among vaccinated adults could be considered.
173 PREOPERATIVE IMMUNISATION FOR HBV AT POLISH HOSPITALS AS A POSSIBLE PUBLIC HEALTH TOOL TO LIMIT THE SPREAD OF THE EPIDEMIC: A CROSS-SECTIONAL STUDY M. Ganczak1, Z. Szych2, M. Korzen´3 1Department of Hygiene, Epidemiology, and Public Health, Faculty of Health Sciences, Pomeranian Medical University, SZCZECIN, Poland, 2Department of Computer Science and Education Quality Research, Faculty of Medicine, Pomeranian Medical University, SZCZECIN, Poland, 3 Faculty of Computer Science and Information Systems, Westpomeranian University of Technology, SZCZECIN, Poland Objectives: to determine the immunisation coverage for HBV among surgical/gynaecologic patients and the prevalence of HBsAg to detect seropositive patients immunised for HBV without previous HBsAg screening. Methods: anonymous serosurvey of consecutive adult patients presenting to 16 randomly selected hospitals in West Pomerania, Poland, between February 2008/January 2009. Patients’ characteris-tics obtained by a questionnaire, serum samples assayed for HBsAg by ELISA test. Results: response rate was 87%. The
123
94 median age: 49 years; 1118 (68%) of 1652 study participants were females; 68% were hospitalised on surgical wards. Immunisation was reported by 60.4% patients (95%CI:58.0–62.7%) of these 81.7% had received the complete course of three doses; 7.6% (95%CI:6.4– 8.9%) did not remember if immunised, 32% (95%CI:29.8–34.3%) were not immunised. Of those immunised, 65.2% were immunised at the request of referring surgeons, the rest otherwise. Immunisation coverage was higher (P \ 0.0001) in those having elective surgery than in emergencies (813/1195; 68.0% vs 185/457; 40.1%). The multivariable regression model revealed that patients’ characteristics such as: age \40 years (OR 1.68; P \ 0.0006), a low educational status (OR 1.7; P \ 0.002), not having a history of any surgery in the past (OR 2.1, P \ 0.0001), tattoo application (OR 2.66; P \ 0.0001), as well as a hospital characteristics such as the emergency procedure (OR 2.14, P \ 0.0001) and gynaecologic ward (OR 1.78; P \ 0.002) were each associated with a greater odds of not being immunised. HBsAg prevalence was 0.6% (10/1652; 95%CI:0.3–1.1%). One in ten HBsAg-positive patients had been immunized for HBV without previous HBsAg screening. Conclusions: In addition to other preventive methods, the preoperative immunisation policy might serve as an effective public health tool to limit the spread of the epidemic. Young, low educated patients should be the group of a special interest. Some HBsAg-positive patients might be immunised for HBV without previous HBsAg screening, so clear recommendations for pre-immunisation testing should be established.
181 EPIDEMIOLOGICAL FEATURES OF HIV INFECTION IN SERBIA A. Grgurevic1, D. Simic2, Z. Gledovic1, Institute of Epidemiology, School of Medicine, University of Belgrade, BELGRADE, Serbia, Institute of Public Health of Serbia, BELGRADE, Serbia Background: HIV/AIDS represents one of the most destructive epidemic in the world population. It is estimated that 33 million people worldwide are currently living with HIV/AIDS. Annually about 2.7 million people become newly infected with HIV and 2 million people die due to AIDS related causes. Objective: The aim of the study was the analysis of epidemiological characteristics of HIV/AIDS in Serbia. Methods: The source of morbidity and mortality data for the period 1984–2008 were the annual reports of Institute of Public Health of Serbia. Population denominator data were obtained according to the official estimates of the Statistical Office of the Republic of Serbia, done on the basis of national censuses from 1981, 1991 and 2002 with extrapolation. In the data analysis incidence rates, mortality rates as well as incidence and mortality trends were calculated. Results: During the period 1984–2008 there were 2317 HIV infected persons, 1436 AIDS cases and 945 deaths caused by HIV related causes in Serbia. HIV incidence rates slightly increased in the period 1984–2008 (y = 0.723 ? 0.027x, P = 0.036) and the highest incidence rate (1.55/100,000) was in 2008. The trend of AIDS incidence rates showed also significantly increasing tendency during the observed period (y = 0.378 ? 0.026x, P = 0.008) with the highest rate in 1998 (1.44/100,000). AIDS mortality rates had reached a peak in 1996 (0.93/100,000) and began to decline in 2002. Analysis of AIDS mortality rates revealed a significant quadratic trend in the observed period (y = -0.226 ? 0.142x - 0.005 9 2, P \ 0.001). Conclusion: Although Serbia belongs to countries with low HIV/ AIDS prevalence, situation could be defined as unfavorable with tendency of worsening because of bed social and economic conditions, as well as, the presence of risk behavior.
123
IEA-EEF European Congress of Epidemiology 2009
Clinical epidemiology 84 DEVELOPMENT OF EXPLICIT APPROPRIATENESS CRITERIA FOR ADMISSION IN PATIENTS WITH ACUTE COPD EXACERBATION J. M. Quintana, S. Garcia, A. Escobar, C. Esteban, U. Aguirre, N. Gonzalez, and the IRYSS-Cataract Appropriateness Group Hospital Galdakao-Usansolo, USANSOLO, Spain Objectives: To develop explicit appropriateness criteria for admission of patients who come to emergency department with acute exacerbation of COPD and evaluate the reliability of the 2 panel of experts that help to develop them. Methods: Explicit appropriateness criteria were created through a variation of RAND appropriateness method. 820 scenarios were developed, from the combination of the following variables: age, presence of diabetes mellitus, presence of cardiopathy, prevision of patient treatment fulfillment, response to the previous treatments, basal COPD severity, COPD acute exacerbation severity, number of previous hospital admission in the last year, and need of domiciliary chronic oxygen therapy. Two panel of experts0 scores were analyzed:main panel, who was in a face to face meeting for a second round and the second panel whose second round was performed by mail. For the generation of results, CART analysis was used (Classification and Regression Trees). Results: We included results of the decision trees originated by the CART analysis. The main panel score mean was 6.61 in the first round and 6.49 in the second one (standard deviation (SD) ranged from 0.47 to 2.28 in the first round and from 0.39 to 0.9 in the second round). In the second panel the scores means were respectively 6.47 to 6.60 with a (SD) ranged from 0.44 to 1.74 in the first round and from 0.53 to 1.57 in the second one. The degree in agreement between panels about appropriateness scores was of 90.6% in the group of the appropriate category and 81.5% in the inappropriate category (kappa = 0.79). Conclusions: We have been able to synthesize the results in decision trees for an easier use between the clinicians. The panel who was meet in a face to face meeting for the second round showed less variability in the scores and more trend to regression to the mean than the second panel.
199 PREVALENCE OF MYOPIA AND HYPEROPIA IN SOME SELECTED EYE CARE CENTERS IN IMO STATE OF NIGERIA J. O. C. Anyanwu, I. D. Ezirim, C. W. Agbakwuru, T. U. Emegwamuo Optometry Department, Imo State University, OWERRI, Nigeria Background: This research work was done to determine the prevalence of myopia and hyperopia in some selected eye centers in Owerri Imo State Nigeria. Aims:To determine the prevalence of myopia and hyperopia among people in Owerri Imo State. Material and methods: The target population for this study includes 300 subjects of both male and female between the ages 3 and 70 years with corrected refractive error and -0.25cyl randomly selected from private and government eye clinics in Owerri Imo state. Simple percentage is used as method of data analysis to establish the prevalence of myopia and hyperopia according to age group, sex and refractive status. Results: Percentage distribution of prevalence data of 300 samples of myopia and hyperopia collected in some private and government eye clinics in Owerri Imo State shows that prevalence rate of hyperopia and myopia is high with a percentage rate of 61% and 39%. Between the age of
IEA-EEF European Congress of Epidemiology 2009 25–29 years, hyperopia occur more with prevalent rate of 13.3% myopia at this same age groups is more with prevalent rate of 6.7%, while between the age of 60–64 the prevalent rate is the lowest with hyperopia 1%, myopia1% too. It is observed that hyperopia occur more in female with prevalent rate of 38.7% of total population and 22.3% of male while myopia is found to be 22.3% in female and 16.7% in male. Conclusions: i. With the result of my research work, I therefore conclude that hyperopia is more prevalent than myopia in Owerri Imo State. ii. Myopia and hyperopia occur more in young adult of 25–29 eye group. iii. The gender difference shows that hyperopia is more prevalent in female than male while myopia is also more prevalent in female than in male iv. This information provide data for eye care planners in Owerri Imo State Nigeria and shows need for visual health education to reach the inhabitants of this area especially young adults.
Social epidemiology II 252 PREVALENCE OF INTIMATE PARTNER VIOLENCE AND ASSOCIATED FACTORS: A POPULATION-BASED STUDY IN SOUTHERN BRAZIL A. J. Anacleto1, K. Njaine1, G. Z. Longo1, A. F. Boing1, K. G. Peres2 1 Post graduate Program in Public Health, Universidade do Planalto Catarinense, LAGES, Brazil 2Post graduate Program in Public Health, ´ POLIS, Brazil Universidade Federal de Santa Catarina, FLORIANO Background: Intimate partner violence is considered a public health problem, playing an important role in individual physical health, and their self-esteem. In addition, women affected by intimate violence are more prone to a wide range of health problem including chronic pain, physical disability, drug and alcohol abuse, and depression. Objectives: The aims of this study were to estimate the prevalence of intimate partner violence and its social associated factors. Methods: A cross-sectional population-based study was carried out. A multi stage clustered sample of 1,042 women aged 20–59 years living in the urban area of Lages, SC, a medium sized city in Southern Brazil, was domiciliary interviewed. A questionnaire covering socioeconomic, and demographic variables was applied including the Conflict Tactics Scales—Form R to investigate verbal aggression, minor physical violence, and severe physical violence. Pearson chi-squared and linear trend test were used to test the associations. Results: The prevalence of verbal aggression, minor, and severe physical abuse within couples were 79.0%, 14.9%, and 9.3%, respectively. Couples with less than 30 years of age, with a per capita income lower than a half monthly Brazilian minimum wage, and living in places with more than two individuals by one bedroom, were more likely to report all of type of violence measured when compared with older couples, those with higher incomes, and those living in less crowding places, respectively. Conclusion: Socioeconomic disadvantaged are important aspects to take into account in order to reach effective strategies against the intimate partner violence.
60 EFFECTIVENESS OF SLIDES-BASED AND PAMPHLETS-BASED EDUCATION ON KNOWLEDGE AND ATTITUDE OF CONSCRIPTS ABOUT SUBSTANCE ABUSE, SHIRAZ-IRAN, 2008 A. A. Banihashemi1, A. Mirahmadizadeh2, A. Hemmati2, R. Parsapour3 1Social Security Assurance of Fars Province,
95 SHIRAZ, Iran, 2Shiraz University of Medical Sciences, SHIRAZ, Iran, 3Kermanshah University of Medical Sciences, KERMANSHAH, Iran Background: Attending of numerous high-risk young people as soldiers in military garrisons, makes these centers as one of the places in which any substance abuse prevention program could be effective. This study conducted to show the effectiveness of slides-based and pamphlets-based education on knowledge and attitude of conscripts about substance abuse. Method and material: This study conducted in 14 selected military garrisons in Fars province. At the first stage, we trained a group of focal points, as peers to educate own center’s conscripts. At the second stage, the came back to theirs centers to execute the prevention program. Educational materials included a series of PowerPoint slides and a series of pamphlets. To evaluate the effectiveness of the program, we compared the results of pre and posttests, using paired t-test. Findings: This study showed that the mean age of participants was 24.41 ± 3.77 years, the mean of participants’ education was 11.22 ± 3.5 official years, and prevention program significantly increased the scores of knowledge and attitude, (P \ 0.001). Conclusion: The finding of this study, like other studies, showed that prevention programs effectively promote the levels of both knowledge and attitude about substance abuse. Although prevention programs are effective, they are very difficult to implement in military places because security concerns and some other limitations. We recommend high-level coordination with directors and commanders of military centers, in advance. In addition to traditional drugs, information about alcohol, cigarette smoking, hookah and new substances must add to educational package.
257 PROBLEMATIC USE OF PSYCHOACTIVE SUNSTANCES IN PATIENTS UNDER TREATMENT AT PUBLIC ADDICTION CENTRES IN AN ITALIAN PROVINCE AND RELATED PREDICTORS M. Karakachoff1, M. Gori1, R. Lovaste2, M. Scalese1, C. Doveri1, L. Molteni2, R. Calla`2, S. Molinaro1 1Italian National Council for Research, Institute of Clinical Physiology, Epidemiology and Public Health Research, PISA, Italy, 2Public Addiction Treatment Centres of Trent, TRENT, Italy Background: Systematic review of scientific literature reveals the proven effectiveness of some treatments undertaken by drug abusers. That effectiveness can be explained by one or more of the following factors: treatment retention and reduction of drug use, reduction of health risk behaviours, reduction of overdoses and reduction of related infectious diseases, prostitution and sexual partners. Objectives:This study aims to assess the intermediate and final outcomes of drug abuser treatments analysing them in a broader view of personalized treatment programs, according to the procedures used in the Public Addiction Treatment Centres (SerTs) of Trento, and in relation to characterization of the patient and the motivational phase he is undergoing. Methods:Databases used referred to patients under treatment at Public Addiction Treatment Centres (SerTs) of Trento, between years 2006 and 2007. For each patient the following data was considered: age, sex, clinical situation, environment and toxicological and socio-economic characteristics. Patients were divided into two categories of drug users, according to the patient’s motivation to change his condition regarding drug addiction, which determines different types of treatment programs and different indicators of the results of the assessment. According to this breakdown and to the different indicators, logistic regression models have been adapted for the analysis of the outcomes
123
96 and their causes. Results: Results show that some therapeutic conditions are associated with the success of the programs. In particular, the motivation to change has proved to be the most important cause of the overall success of treatment. Other factors are related to: early entry, the continued retention in treatment and older age. There were no socio-economic or clinical conditions that may be considered to be statistically significant in determining the outcome of treatment programs. Conclusions: Although these are preliminary results, it seems interesting to note the possibility of assessing the outcomes of treatment programs according to the inclination and motivation to change of patients and also based on intermediate and global indicators which are representative of treatment program effectiveness.
113 HIGH MORTALITY AMONG PEOPLE SUSPECTED OF DRUNKEN DRIVING. AN 18-YEAR REGISTER-BASED FOLLOW-UP A. Impinen1, O. Rahkonen2, K. Karjalainen1, P. Lillsunde1, T. Lintonen3,4, P. Ma¨kela¨1, A. Ostamo1,4 1National Institute for Health and Welfare, HELSINKI, Finland, 2University of Helsinki, Department of Public Health, HELSINKI, Finland, 3Police College of Finland, TAMPERE, Finland, 4Tampere School of Public Health, TAMPERE, Finland Background: Alcohol drinking and driving under the influence (DUI) of alcohol are associated with numerous harmful health and social effects. DUI offenders are more likely to be involved with fatal road traffic accidents but there is little evidence of other excess mortality in DUI offender population. Objectives: The aim of this study was to examine overall and cause-specific mortality of DUI suspects compared to reference population with no history of DUI and to recognize risk factors of premature death. Methods: The data used were a register of all suspected drunken drivers with between April 1988 and December 2006 (N = 112,394) and their deaths (N = 14,636). All drivers with drugpositive samples were excluded. DUI suspects were compared to reference population with no DUI offence previous to case. Survival analysis methods were used to study the risk factors and to compute overall and cause-specific hazard ratios. Results: Alcohol causes, diseases of circulatory system and accidents were the most common causes of death amongst DUI suspects. DUI was linked with higher mortality in every observed group. Especially risk of death by alcoholrelated or external cause was high compared to reference population. Among women the DUI apprehension increased risk of death more than among men. Within the group of DUI suspects the risk of death was affected by age, sex, marital status, education, recidivism as well as time and observed blood alcohol level of the apprehension. Half of the DUI cases and every fifth of the references had alcohol as contributing factor to death. Conclusions: Drunken driving is a severe indicator of elevated risk of death. As premature deaths and preceding health harms are costly to society these deaths should be prevented more efficiently. Several factors increasing the risk of death could be used to recognize the high risk groups for whom to target interventions to.
215 ANALYSIS OF THE LATENCY PERIOD BETWEEN THE FIRST COCAINE USE AND THE FIRST REQUEST FOR TREATMENT: A PILOT STUDY IN LIGURIA REGION M. Gori, E. Colasante, I. Schizzi, S. Salvini, V. Lorenzoni, S. Molinaro National Council of Research, Institute of Clinical Physiology, PISA, Italy
123
IEA-EEF European Congress of Epidemiology 2009 Background: The aim of this study is to estimate the distribution of the period between the time of first use of cocaine and the time of first contact with a Drug Addiction Service and to evaluate the effects of some covariates (such as gender and age at first use of cocaine) on this time lag, defined as ‘‘latency period’’. Methods: Survival analysis was used to estimate the distribution of the latency period. In particular, Kaplan–Meier method was applied to estimate the survival functions and Log-Rank and Wilcoxon tests were emploied to compare them. Moreover, a Weibull regression model was fitted to assess the relationship of explanatory variables to the probability to seek for a treatment for the first time. These methods were applied to data provided by Liguria region and collected within the project ‘‘Epidemiological Surveillance of the Drug Addiction’’ during the years 2005–2007. Results: The analysis of the variable ‘‘age at first cocaine use’’ points out average latency times approximately around 9 years among those with onset of cocaine use before 20 years of age, 6 years among those with onset after 21 years of age. Latency period vary also according to the gender: 4 years among females and 8 years among males. Both Log-Rank and Wilcoxon test show that those differences are to be considered statistically significant (P-value = 0.00). According to the univariate analysis, results of the Weibull regression model show that women and those starting cocaine use after 21 years of age have more probability to claim earlier (respectively 80% and 60% related to the reference groups) the first treatment related to cocaine use. Conclusions: The distribution estimate of the time lag between the first use of cocaine and the first treatment demand constitutes a good epidemiological indicator of the attractiveness and rapidity of first treatment activation on behalf of the public health service, addressed to subjects with substance related problems. This pilot study points out that the distribution of the latency period is influenced both by gender and age at first use of cocaine; therefore women and those who began taking cocaine in the later years, are more likely to claim earlier the first treatment.
218 POPULATION SURVEY ABOUT GAMBLING PARTICIPATION AND GAMBLING-RELATED PROBLEMS AMONG ADULT RESIDENTS IN CATALONIA, SPAIN A. G. Iba´n˜ez1, R. Volberg2, M. J. Pueyo3, S. Melchiorre1, L. Baranda4 1 Mataro´ Hospital, BARCELONA, Spain, 2Gemini Research, Ltd., NORTHAMPTON, United States of America, 3Health Department of Catalonia, BARCELONA, Spain, 4TRS enterprise, BARCELONA, Spain Background: Although several forms of legal gambling, including casinos, lotteries and gaming machines, are well-established in Catalonia, very little is known about how Catalonians gamble, and what groups in the population experience problems related to their gambling. Objectives: The goals were: to assess the habits of gambling of Catalonian population; to estimate the prevalence of risk, problem and pathological gambling in the adult population; and to describe their characteristics. Methods: Cross-sectional descriptive study. During 2007, a multi-stage sampling representative of sex and group of age of the population of Catalonia (over 7 million inhabitants) was done. 3,000 adults were interviewed by telephone with a questionnaire used in American epidemiological studies and adapted and validated to our country. Problem with gambling was assessed using the NODS, a DSM-IV-based instrument designed for use in population studies, and the Canadian Problem Gambling Index. Variables about socio-demographic characteristics, reasons for gambling, mental health, self-perceived health and substance abuse were
IEA-EEF European Congress of Epidemiology 2009 also collected. A pilot study was done and quality in all the process was assured. Results: Prevalence of gambling was 52.2% in the last year and 90aˆ€TM2% any time in life. Lottery and bet on sports results were the favourite game. The life prevalence of risk gamblers was 2%, problem gamblers was 0.5% and pathologic gamblers 0.2%. Last
97 year prevalence was 1, 0.3 and 0.1%, respectively. Problem and pathological gamblers were more frequently male, young (between 18 and 34), began before 18 years, mainly with gaming machines, had more mental health problems, had worst self-perceived health and showed higher substance abuse.
123
.
IEA-EEF European Congress of Epidemiology 2009
99
Index: Authors and Page Numbers
A Aasmoe, L. Aavitsland, P. Abd El-Hafeez, S.S. Abdella, N.A. Abdrkhmanov, O. Abdul-Rasul, M. Abu-Saad, K. Adamkova, L. ´ da´ny, R. A Adler, B. Afifi, R. Agabiti, N. Agbakwuru, C.W. Aguirre, U. Ahmadnezhad, E. Aizpuru, F. Akhtar, S. Al-Adsani, A. ´ lamo, R. A Alavinia, S.M. Albanell, J. Alberti, M.M. Alexander, R.M. Alexander, S. Al-Khawari, M.A. Allotto, M. Alvelos, M. Alves, E. Ambrosio, G. Anacleto, A.J. Andersen, L. Anes, Y. Antonelli, G. Antunes, J.L.F. Anwar, O. Anyanwu, J.O.C. Arau´jo, C.L.P. Arca´, M. Ardalan, A. Arik, H. Arostegui, I. Arrieta, E. Arsanoj, A. Arshad, S.H. Ascunce, N. Asefzadeh, S. Aslan, O. Atthobari, J. Audignon, S. Augestad, L.B. Auleda, J. Azevedo, A.
51 28 21 43 62 43 71 51 30, 44 47 33 94 87, 78, 50 92 43 50, 51 24 19 92 25 43 21 38 19, 60 95 66 50 21 25, 46 94 78 33 78, 44 87 87 37 37 41, 69 65 21, 85 59 78, 19,
B Bakos, L. Balabash, O. Balboa, T. Bammann, K. Bandyopadhyay, M. Banecka, B. Banegas, J.R.
21 76 68 48 71 75 68
83
94 88, 93
87
55
81
88
64
32, 45
87 38, 55, 61
Bang, B. Banihashemi, A.A. Baranauskiene, D. Baranda, L. Bare, M. Barendregt, J. Barger, A. Barone, A.P. Barros, H. Bartnicka, M. Bastiaannet, E. Bastiani, L. Bastos, J. Bayard, V. Bazzichi, L. Bellelli, S. Belleudi, V. Belmaker, I. Belvis, F. Benazzouz, M. Bencivelli, W. Benes, C. Benjelloun, S. Bennett, K. Ben-Shlomo, Y. Benson, V. Beral, V. Berezina, L.A. Bermejo, J. Bes-Rastrollo, M. Bettencourt, P. Bhaskaran, K. Biazevic, M.G.H. Bien´, S. Bienias´, B. Bilbao, A. Blasco, J.A. Blasko´, G. Boffetta, P. Boing, A.F. Bolli, S. Bolotin, A. Bombardieri, S. Bonamigo, R. Bonenfant, S. Bonithon-Kopp, C. Bonnaud, S. Borgia, P. Bouvier, A.M. Bowen, L. Brand-Herrmann, S.M. Bre˛borowicz, G. Brenn, T. Broadbent, J.M. Brochard, P. Broda, G. Brodka, K. Bronnum-Hansen, H. Brown, S. Broz˙ek, G. Brozek, G. Brugulat, P.
51 95 62 96 54 58 45 33 19, 51 49 75 56 82 49 49 33 37, 41 47 56 76 47 33, 19, 24 24 45 38 60 38 61 81 64 75 50, 54 30 62 95 62 37 49, 21 80 24 85 54 24 44, 26 74 27 36 85 66 84 58 71 90, 35 78,
28, 56, 90
71
84 66
54, 87
56
66
91 87
123
100
IEA-EEF European Congress of Epidemiology 2009
Brunekreef, B. Brzyska, M. Brzyski, P. Buczynska, A. Buekens, P. Bukowska, A. Burdorf, A.
37 68, 49, 62, 25 64, 51
C Calatrava, M. Calla`, R. Camilloni, L. Capon, A. Carlos, S. Carlsen, K.H. Carmona-Torre, F. Carosi, I. Carrilho, C. Carton, M. Castells, X. Castrodeza, J. Castronuovo, E. Catarsi, E. Cazzato, M. Cehreli, R. Cermakova, D. Chabros, E. Chaix, B. Chalmers, B. Chamorro, F. Chandola, T. Charzewska, J. Chastang, J.F. Chauvin, P. Chen, C.M. Chen, Q. Chiesa, F. Chini, F. Cho, S. Chumak, E. Chwojnowska, Z. Chyou, P. Cichocka-Jarosz, E. Cintra, F.A. Coeuret-Pellicer, M. Colasante, E. Consensi, A. Cook, C. Corominas, J. Correia, S. Costa, D. Cottet, V. Cunha, A. Curzio, O. Cyprowski, M. Czako´, R. Czech-Szczapa, B.
73 95 54 30, 73 37 60 91 47 85 24, 38 67 56 49 82 51 32 31 59 82 34 32, 80 31 37 85 36 54 48 22, 32 55 74 68 80 20 49 72 24 55 61 24 90 31 62, 82 92
D da Costa, J.P. D’Ascanio, A. Davey Smith, G. David, L. Davies, D.P.
20 56 66 47 19
123
69 68, 69, 74 84 65
81
41, 64
41
59
84
Davies, N.M. Davoli, M. de Andrade, F.P. De Backer, G. De Bacquer, D. de Bruin, K. de Carvalho, M.B. ´ .G. de Caso, J.A de Castro, V. de Craen, A.J.M. de Fa´tima de Pina, M. de Freitas, I.C.M. de Irala, J. de Jong, P.E. de Jong-van den Berg, L.T.W. de la Iglesia, P. de Lejarazu, R.O. de Moraes, S.A. de Sousa, M.L.R. De Spiegelaere, M. De Stavola, B. de Vries, E. Della Rossa, A. Demarest, S. Demokritou, P. Dennis, C.L. Denoth, F. Deprez, N. Dhaliwal, J.K. Di Lallo, D. Diehm, C. Diogo, M.J.D. Dittrich, R. Dlhy, J. Domingo, L. dos Santos Silva, I. Dougherty, G. Doveri, C. Doveri, M. Dramaix, M. Draper, E. Drygas, W. Dugravot, A. Dwiprahasto, I. Dyba, T. Dzielska, A.
19 33 81 60 43, 50 81 87 27 49 71 27, 73 21 21 87 38 27, 68 25 44, 58 49, 34 37 26 74 28 77 43, 44, 68 26 76 24 90 26 95 49 25, 42 66 74 32, 58 36
E Ebrahim, S. Eckmanns, T. Eichlerova, A. Eiros, J.M. Ejsmont, J. El Feydi, A.E. El Kihal, L. Eller, E. Elliott, L. El-Wakil, H.S. Emegwamuo, T.U. Engholm, G. Englert, H. Ergor, G. Escobar, A. Espelage, W.
44, 66 45, 77 51 38 54 47 47 37 72 21 94 58 65 65 94 45
60
31, 40, 79
31, 40, 79
90 56
67, 89 69
73
45
IEA-EEF European Congress of Epidemiology 2009 Esteban, C. Ezirim, I.D. Ezzikouri, S.
50, 54, 94 94 47
F Fabiani, L. Faivre, J. Farchi, S. Farnik, M. Farwick, A. Faubel, R. Fayyaz-Jahani, F. Feely, J. Ferman, R. Ferna´ndez, S. Ferrer, F. Ferrie, J.E. Field, D. Flak, E. Flanders, W.D. Forde, I. Forouzanfar, M.H. Fortes, C. Franco, F. Frank, J.W. Fras, Z. Fraser, D. Fusco, D.
91 24 43, 35, 47 68 78, 33, 76 50 24 74 42 75 59 34 93 21, 43, 25 43 71 33
G Gad, Z.M. Gagnon, A. Gaio, R. Gaita, D. Ganczak, M. Gansevoort, R. Garcia, S. Garcı´a, S. Gasior, Z. George, K. Gigante, D.P. Gigante, R. Gil, M. Giorgi Rossi, P. Gisle, L. Glazer, Y. Glazier, R.H. Gledovic, Z. Godlewski, D. Goldberg, M. Goldberger, N. Golestan, B. Gomez, C. Gomzi, M. Gonzalez, N. Gonza´lez, N. Gori, M. Gozdz, S. Granstro¨m, F. Green, J. Greenlee, R. Gregory, P. Grgurevic, A. Grjibovski, A.
21 26 20, 60 93 21 94 50, 66 19 36 91 50, 54 34 37 25 94 41 74, 44 67, 50 22 94 50, 95, 58 53 24 55 72 94 28
54, 89 90
88, 93 84
62 89
38
54, 87
87
80, 85 70
87 96
101 Grjibovski, A.M. Gromnica, R. Gu, H. Guallar-Castillo´n, P. Guariento, M.H. Gudmundsdottir, S.L. Gue´guen, A. Guenther, F. Guille´n-Grima, F. Guimara˜es, T. Gulati, R.R. Gutie´rrez, M. Gutie´rrez-Fisac, J.L. Gutterman, E. Gutterman, E.M.
22, 27, 32, 52, 59, 73 53 85 68 68 59 80 48 60 56 43 38 68 88 92
H Haelterman, E. Hajdukova, Z. Halken, S. Hallal, P.C. Hamidian, R. Hanke, W. Hansen, J. Harteloh, P. Hashemi Shahri, M. Hashsish, M.H. Hassan, N. Hassar, M. Hatami, S. Hayes, J. He, J. Heaman, M. Heidrich, J. Heinrich, J. Hejazi, M. Hejlek, A. Helewa, M. Hemmati, A. Hense, H.W. Herbarth, O. Hesse, E. Hidve´gi, T. Hollenberg, N. Holub, J. Home`re, J. Horta, B.L. Horva´th, A. Høst, E.A. Hotopf, M. Hrncir, E. Huy, C.
25 53 37 78 86 84, 48 50 93 21 76 47 88 55 85 59, 26, 37 77 51 72 95 42, 37 34 30 82 51, 85 36 30 37 79 51 44,
I Iacovacci, S. Iba´n˜ez, A.G. Iba´n˜ez, J. Imbernon, E. Impinen, A. Isaacman, D.
30 96 64 85 39, 96 88
J Jakopanec, I. Janevic, T.
28 88, 92
88, 91
72 43, 60
47
53
69
123
102
IEA-EEF European Congress of Epidemiology 2009
Janion, M. Jedrychowski, W. Jedynak, U. Jenei, T. Jennings, C. Jermendy, G. Jirak, Z. Jooste V, V. Juczewska, M. Jurewicz, J. Juszko-Piekut, M.
66 75 74 83 43 30 22, 51, 53 24 58 84, 91 53, 64, 86
K Kaczmarski, M. Karakachoff, M. Karakis, I. Kardaun, J. Karjalainen, K. Kark, J.D. Katus, H.A. Kawalec, E. Keil, T. Keil, U. Keshtkar, A. Keszei, A. Keyvanlo, S. Kharkova, O.A. Khasnutdinova, S.L. Kieltyka, A. Kim, H. Kim, J.W. Kim, S.S. Kingston, D. Kinra, S. Kirchmayer, U. Kivimaki, M. Kjelle´n M. Kleanthous, S. Klumbiene, J. Ko, K. Kocic, B. Kodim, N. Koh, S.B. Kolosza, Z. Koo, H.J. Kopacz, M.S. Kordysh, E. Korolova, E. Korzen´, M. Kotseva, K. Koupil, I. Koutrakis, P. Kowalska, M. Kozaczka, A. Kozajda, A. Kramer, M.S. Kra¨mer, U. Kraszewska, E. Krause, G. Kregzdyte, R. Kriaucioniene, V. Kristin, E. Krzy_zak-Mas´lach, M. Kuehni, C. Kull, I.
74 95 37 50 39, 44 65 66 37 26, 88 83 93 73 32 75 61 61 61 59 66 33 34, 34 37 85 61 70 39 27, 53, 61 68, 37 53 93 60 36 37 35, 41 84 19 37 32 77 62 85 32 58 37 37
123
96
43, 60
74
30 64 69
91
Kulus, M. Kulyashova, L.B. Kycler, Z.
74 45 74
L La Vecchia, C. Lafuente, I. Lahav, T. Lancia, A. Landecka, I. Lange, J. Łapin´ska, U. Larrosa, M. Las Hayas, C. Lau, S. Leclerc, A. Lee, L. Lee, W. Lehocka, H. Lehtinen, M. Lesiw, T. Leveˆque, A. Li, Z. Liefers, G.J. Ligocka, D. Lillsunde, P. Lima, B. Lindinger, A. Lintonen, T. Lis, G. Litaker, D. Liu, P. Livia Lobo, J.L. Lødrup Carlsen, K.C. Longo, G.Z. Lopes, C. Lo´pez, A. Lo´pez, C. Lo´pez-Garcı´a, E. Lorenzoni, V. Lourenc¸o, P. Lovaste, R. Lozano, J.E. Lucas, R. Lunet, N. Luostarinen, T.
47, 50, 37 30 53 74 58 78, 87 37 80 59 79 51 41 83 73 85 49 84, 39, 28 42 39, 74 44 85 21 50, 37 95 20 87 73 68 49, 38 95 38, 28, 42, 41
M Machado, J.C. Machartova, V. Macia`, F. Majdzadeh, R. Majewska, R. Ma¨kela¨, P. Maksimowicz, K. Małaczyn´ska, T. Manczuk, M. Manktelow, B. Mantini, V. Marchi, F. Mareedu, R. Margolles, M. Mariani, F. Marinho, A.
47 53 24 93 75 39, 96 58 74 35 42 81 56 55 87 20, 70, 75 56
56 87
87
91 96
96
54
70, 96
50, 87 56 47, 56
IEA-EEF European Congress of Epidemiology 2009 Marjan´ska, J. Martin, R.M. Martı´nez, J. Mart´inez-Gonza´lez, M.A. Martiushov, S. Mas´lach, D. Mastroeni, S. Mastromattei, A. Mazloomzadeh, S. Mazur, J. McCarthy, A. Melchi, C.F. Melchiorre, S. Melo, N. Memon, A. Menezes, A.M.B. Menur, E.M.C. Merry, L. Meseri, R. Meyer-Sabellek, W. Mezyk, R. Michaluart-Ju´nior, P. Middleton, N. Mierzejewska, E. Migliorini, P. Mikiel-Kostyra, K. Mills, D. Mirahmadizadeh, A. Miralles, M. Moffat, M. Moghimi, M.H. Mohammad, H.G.H.H. Mohammad, K. Mohr, O. Moineddin, R. Molinaro, S. Molteni, L. Mompart, A. Monicelli, P. Morais, V. Morris, H. Moscardini, S. Motta, J. Mousavi, F. Mowinckel, P. Mozdzierz, A. Mroz, E. Mrozek-Budzyn, D. Mu´ller-Nordhorn, J. Murray Thomson, W.
64, 19 24 60 27 58 21, 81 77 36, 19 21 96 62 43 78 45 26 82 65 58, 25 37 60 56 36 88 93, 87 72 77 92 93 77 25 20, 95 87 49 55 72 49 82 67, 37 53, 75 75 65 36
N Nachtigal, M. Naeni, K.H. Nagaraj, R.C. Nagy, A. Naieni, K.H. Najafi, M.A. Nakladalova, M. Nazari, A. Nejad, A.F. Nekouie, H. Ne´ri, A.L. Neri, R. Nicolaou, N.
85 78 79 30, 83 88 86 51 88 70 76 68 49 37
65
62
54
64
95
31, 75, 70, 74, 95, 96
70 64, 86
103 Nilssen, O. Nitu, G. Njaine, K. Noll, I. Novikava, N. Nowakowska, E. Nunez-Co´ rdoba, J.
27 66, 90 95 45 81 41 60
O Olhero, A. Oliveira, C. Olsson, L. Omidvar, M. Ortiz, M. Osorio, A. Ostamo, A. Ozguler, A.
71 68 53 51 55 73 39, 96 80
P Pac, A. Pace, A. Pajak, A. Palmer, B. Paludan-Mu´ ller, G. Pane, M. Panopoulou, K. Paragh, G. Park, J.K. Parsapour, R. Pasquini, P. Pathak, D.R. Pauleikhoff, D. Pelclova, D. Peleteiro, B. Peplon´ska, B. Perea, E. Pereira, M. Pererra, F. Peres, K.G. Peres, M.A. Perucci, C.A Perucha, M. Petkeviciene, J. Petrovic, B. Pezzotti, P. Pierzchala, W. Pietrzyk, J.J. Pilla, M. Pina, F. Pineau, P. Pinto, E. Pinto, H. Pinto, M.C. Pipis, S. Pirie, K. Poggensee, G. Polack, S. Polanska, K. Polaufova, P. Polo, A. Pompili, A. Poorbiazar, M. Portielje, J. Potente, R. Prabhakaran, D.
49 81 66, 26 58 39 91 30 27, 95 21, 41 47 51 42, 64, 54 61 75 36, 36, 33 50 85 70 67, 90 74 21 68 47 90 56 42 37 24 77 43 88, 77 43, 81 77 49 70 66
79
30, 61 62
47, 56 65
78, 95 78
89
91 89
123
104
IEA-EEF European Congress of Epidemiology 2009
Prasad, L. Profus, K. Prokofyeva, E. Prugger, C. Pueyo, M.J. Puig, C.
28, 56 74 83 26, 43, 60 78, 87, 96 37
Q Qabazard, M. Qi, J. Quin˜ones, C. Quintana, J.M.
43 85 50 50, 54, 87, 94
R Racape´, J. Rada, C. Rahkonen, O. Rai, J.S. Ramakrishnan, L. Ramos, E. Ramos, M. Razzaq, H. Reddy, K.S. Reeves, G. Refaat, A. Rehman, S. Reiner, Z. Rihs, L.B. Ritz, B. Roberts, G. Rod, N.H. Rodrı´guez-Artalejo, F. Roelands, M. Roll, S. Rolland, P. Roma´n, R. Roustit, Ch. Ruiz-Canela, M.
25 66, 39, 77 66 20 50 46 66 24 67 46 43, 68 48 37 48 68 28, 37 85 41, 31 73
S Saadat, S. Saatli, G. Sadeghian, S. Sala, M. Salas, D. Salvadori, S. Salvatori, S. Salvatori, V. Salvini, S. Sannikov, A.L. Santos, A.C. Santos, M.J. Sarov, B. Saucier, J.F. Sauve, R. Scalese, M. Scatigna, M. Schernhammer, E. Schizzi, I. Schneider, S. Schwedler, G. Sementilli, G. Senterre, C. Sepehrvand, N. Serrano, A.
86 65 86 24, 41, 49, 70 62 96 52 20 28, 37 26 59 95 91 48 96 69 42 91 73 88 85
123
90 96
60
39
64
41, 64 64 56
56
Servitja, S. Severo, M. Shaltout, A.A. Sharma, P. Shin, C. Shipley, M. Shiryaeva, O. Sianipar, D.V.P. Siciliano, V. Siddique, I. Sidorenkov, O. Simen Kapeu, A. Simic, D. Singh-Manoux, A. Sinnreich, R. Skarupa, A. Slachtova, H. Small, R. Smith, G.D. Smith, P. Smolen, E. Sobala, W. Soerjomataram, I. Soloviev, A.G. Sonntag, F. Sorahan, T. Soroka, P. Sowiak, M. Splichalova, A. Spycher, B. Stafoggia, M. Stanger, E. Steindorf, K. Stewart, D.E. Stojko, J. Stoll, M. Storm, H. Stringhini, S. Strumylaite, L. Sunyer, J. Suresh, A. Surı´s, X. Surmach, M. Suwardiman Suzuki, C.S. Svarval, A. S´widzicka, O. S´wiebocka, E. Sword, W. Szadkowska-Stanczyk, I. Szafraniec, K. Szamotulska, K. Sze´les, G. Szewczyk, L. Szigethy, E. Szpak, A. Szych, Z.
24 38, 43 55 61 74 51 45 20, 43 27 41 94 74 44 75 77 26, 19 55 36 84, 58 73 65 84 62 84 22, 37 33 26 44 26 53, 47 58 74 62 37 43 78, 72, 45 79 76 74 74 72 62, 66, 60 30, 75 30, 58, 93
T Tafforeau, J. Talarico, R. Tavakoli, H.R. Tavares, M. Tavoni, A. Taylor-Clapp, S.
34 49 76 28 56 59
90
74, 75
71
91
51, 53, 77
64, 86
87 89
84 79, 86 83 83 66
IEA-EEF European Congress of Epidemiology 2009 Telgmann, R. Thiel, A. Tilling, K. Tishchenko, E. Tjaden, L. Tobiasz-Adamczyk, B. Toledo, E. Tomasek, I. Tomaskova, H. Torrent, M. Trinito, M.O. Turchetti, G.
26 44 19 72 72 49, 68, 69, 74 60, 73 77 22, 51, 53, 77 37 30 49
U Ucku, R. Ujhelyi, E. Unal, B. Urban, P. Urcioli Scalese, M. Urquia, M.L.
82 88 65, 82 22, 51, 53 70 25
V Vaccarezza, G.F. Vadrucci, S. van de Velde, C.J.H. van der Geest, L.G.M. Van der Heyden, J. Van Steen, K. Varakina, Zh.L. Vardi, H. Vega, T. Velarde, J. Velasco, E. Verma, A. Veyalkin, I. Vidaillet, H. Vidal, S. Vigna-Taglianti, F. Visser, S.T. Voko´, Z. Volberg, R. Vo¨ller, H. von Berg, A. von Euler-Chelpin, M. Vrotsou, K. Vyazmin, A.M.
25 91 49 49 34 37 52 37 38, 50, 87 64 45, 77 92 81 55 54 91 21 30, 83 96 65 37 34 87 52
W Wahn, U. Wahoush, O. Wajszczyk, B. Wall, C.A.
37 26 32 33
105 Wan Md Adnan, W.A.H. Watson, L. Wawro, N. Wegscheider, K. Wellmann, J. Westendorp, R.G.J. Wickman, M. Wijga, A.H. Wilke, R. Williams, D.M. Willich, S.N. Windler, E. Windmeijer, F. Winterhalter-Zvonar, B. Wizner, B. Wojciechowska, U. Wojcik, W. Wolfshaut-Wolak, R. Wood, D. Wypych-S´lusarska, A. Wyszyn´ska, M.
33 71 48 65 26, 43, 47 49 37 37 83 19 37, 65 65 19 22 66 81 79 66 60 91 86
X Xing, J.
85
Y Yiallouros, P. Yoon, K. Young, D.
37 48 59
Z Zagozdzon, P. Zaharan, N.L. Zanganeh, A.M. Zavadilova, V. Zdrojewski, T. Zejda, J. Zejda, J.E. Zemla, B. Zemla, B.F.P. Zeniou, M. Zeynali, M. Zhebrun, A. Zhebrun, A.B. Zins, M. Zio´łkowska, M. Zrenner, E. Zubizarreta, R. Zugravu, C.A. Zurawski, P. Zurriaga, O.
54 33, 88 22 66 35 90, 41 53, 37 76 76 45 74, 64 83 41, 66, 41 87
84
91 64
80, 85
64 90
123