HEALTHCAREANALYSIS VOL.3:15-26 (1995)

Psychology, Health Promotion and Aesthemiology Paul Bennett -t, Simon Murphy':, Douglas Carrollw and lan Ground82 1-University of Wales, Cardiff, ~:University of the West of England, wGlasgow Caledonian University, Scotland, I] Applied Philosophy Trust, Newcastle upon Tyne, UK

Paper One: Social Cognition Models as a Framework for Health Promotion: Necessary, but not Sufficient Paul Bennett, Simon Murphy and Douglas Carroll

Abstract

Much of health promotion is premised on the notion that health-related behaviours are under individual control, and strongly influenced by intra-psychic factors, including knowledge and attitudes. The emphasis placed on such factors has led to a neglect of the social and material context in which the individual is situated. This paper describes a number of psychological theories which have influenced health promotion, and suggests ways in which a wider set of psychological theories and methods, which take into account social and material factors, may more usefully inform health promotion initiatives.

Paul Bennett, Gwent Psychologyand Consultation Liaison PsychiatryServices,12ParkSquare,NewportNP9 4EL,Wales and Schoolof Psychology,Universityof Wales,SimonMurphy, Facultyof Healthand CommunityStudies,Universityof the Westof England,St. Matthias,Fishponds,BristolBS162JP, Douglas Carroll, Schoolof Psychology,GlasgowCaledonian University, Scotland, and Ian Ground, Applied Philosophy Trust, 194 Heaton Park Road, Heaton,Newcastleupon Tyne NE6 5AP, UK. CCC1065-3058/95/010015-12 9 1995 by John Wiley & Sons, Ltd.

Introduction

The majority of health promotion initiatives are premised upon the notion that health status is, to a large extent, consequent upon behaviour under individual controU According to this assumption, a primary task for those involved in health promotion is to identify the processes involved in determining behavioural choice and to develop programmes which take these into account when attempting to bring about appropriate behavioural change. These processes are frequently considered to lie within the individual, and are thought to involve attitudes, beliefs, personal competencies, and so on: that is, they are thought to be psychological in nature. The centrality accorded these psychological processes has had a powerful effect on health promotion, and many of the most innovative programmes in this field have made them the central target of their interventions.2 The use of psychological theory can be traced to the 'health education movement', which was one of the origins of health promotion. At its simplest, health education was premised on the notion that if people were given appropriate information from the 'right' source this would lead directly to attitudinal change: and this would, in turn, result in behavioural change. These latter assumptions have now been strongly challenged by empirical studies: 3 information does not necessarily lead to attitudinal change, and there is only a modest relationship between attitude and behavioural change. 4 The move in recent years from concentration on

16

health education programmes to the use of apparently more widely focused health promotion programmes reflects recognition of the shortcomings of the 'attitudinal change' approach. Despite this, there has often not been a significant change of focus. 'Community' health promotion programmes still frequently target the individual within the community as the determinant of his or her own behaviour rather than attempt to modify the structure of the community itself. 5 The psychological theories adopted by the 'health promotion movement' have been rationalistic and concerned with intra-psychic determinants of behaviour. Below we briefly describe some of these 'social-cognitive' theories. The list is not exhaustive, but nevertheless includes theories which have informed many internationally recognised health promotion programmes both in Europe 6 and the USA. 7 We go on to consider how psychological models and methods which take more account of the wider social context of behaviour may enhance health promotion initiatives while still making use of psychological theory as a framework.

The 'Social-Cognitive' Approach A number of psychological theories are subsumed under the broad heading 'social-cognitive theory'. While there is some degree of overlap between them, each emphasises differing aspects of cognitive processes. Thus they may be seen as competing theories, and they are not without their critics. 8 However, their value to health promotion is that together they form a framework within which to plan potential interventions. Central to all such approaches is the notion that knowledge alone is frequently insufficient to bring about attitudinal or behavioural change; other psychological processes need to be brought into play.

The Theoryof Reasoned Action The closest descendent of traditional attitudinal theory to be widely adopted by health promoters has been the theory of reasoned action. 9 This theory suggests that the closest predictor of behaviour is one's intention to engage in it. This

R BENNETT, S. MURPHYAND 13.CARROLL

'intention' is held to be derived from an interaction between the individual's attitudes towards the behaviour in question and his or her view of the prevailing social norms (that is, the individual's beliefs concerning the attitudes of 'important others' towards the behaviour). With this theory in mind, some health promotion initiatives have attempted to change 'population attitudes' and perceived norms. In North Karelia, for example, attempts were made through the media to alter television viewers' perceptions of dietary norms to encourage a reduction in the use of saturated fats. 1~

Social Learning Theory Social learning theory 11 makes use of a collection of theories and ideas. The two most pertinent to health promotion are the notions of 'vicarious learning' and 'self-efficacy'. 'Vicarious learning' is the process of developing skills, attitudes and knowledge through observation of others. In the initiation of smoking, for example, most new smokers learn inhalation techniques from observation of others and may only persevere after their first, frequently unpleasant, experience of inhalation because respected peers or parents demonstrate that smoking can be a pleasurable and 'cool' activity. Health promotion programmes based on social learning theory attempt to provide models of health promoting behaviours and, where necessary, behavioural change in such a manner that the observer comes to believe that this is a desirable behaviour, and that he or she is capable of making any necessary behavioural changes. The UK television series Psst ... the really useful guide to alcohol 12 for example, was meant to provide a model of appropriate drinking by showing a number of 'stars' and 'ordinary people' describing their drinking habits and explaining w h y they drank moderately. The series was also intended to promote behavioural change by means of the example set by one of the presenters who cut down his alcohol consumption as the series progressed. 'Self-efficacy', which is somewhat analogous to self-confidence, is thought to be central to the process of change: observers are unlikely to attempt change unless they believe they are likely to be successful. Accordingly, a primary goal of

SOCIALCOGNITION MODELS

interventions using 'vicarious learning' is to maintain or increase observers' confidence in their ability to achieve change. One way this can be achieved is through observation of 'coping' models; that is, through the observation of people similar to the observer who do not achieve effortless change, but who make successful or at least reasonably successful attempts at change. Such a model suggests to the observer that he or she too has the resources to change ('if they can do it, so can I') and demonstrates a variety of methods by which to do so. 13 The media output of both the North Karelia 1~and Stanford 7 coronary heart disease prevention projects provided such modelling through, for example, the televising of volunteer smoking cessation groups. Appropriate modelling of behaviour may also be achieved through the entertainment media. De Foe and Breed, 14 for example, worked with media professionals to incorporate appropriate models of moderate alcohol consumption in several US prime time television series.

Health Belief Theory The 'health belief model '14 was specifically developed in the attempt to explain and predict behaviour in 'health contexts'. The theory behind the model assumes that the likelihood of individuals engaging in a particular health-related behaviour (be it promoting or damaging) is a function of their perceptions of the relationship between that behaviour and disease, their perceived susceptibility to the disease, the seriousness of the disease, and the costs and benefits involved in engaging in a particular course of action. A final influence on the uptake of any behaviour is the presence of cues to action; that is, reminders to engage in certain behaviours. According to this theory, any health promotion programme needs to identify a clear link between behaviour and disease, to highlight the severity of the disease and to make it relatively easy to engage in behaviour likely to lead to a reduction in risk for that disease. In following such a policy, Heartbeat Wales 16A7 encouraged food labelling and increased access to 'healthy foods' in major retail outlets, developed the notion of 'healthy restaurants', and encouraged the wider advertising and distribution of low alcohol beer. This

17

attempt to influence the context of behaviours clearly acknowledges the importance of the environment in influencing behaviour, although still assumes ultimate personal control.

Problems with the 'Social-Cognitive' Approach 'Social-cognitive' theories have been of considerable importance and value in moving health promotion towards the status of 'applied science'. They provide a description of at least some of the processes involved in health-related decisionmaking, and accordingly, suggest intervention points for health promoters. However, the rigidity of the models forces researchers to explore clear hypotheses at the cost of discovering alternative explanations of behavioural choice. For example, the 'health belief theory' presupposes that the decision about whether or not to engage in health-related behaviours is a function of the implications of such behaviours for (usually) long-term health. But this may not always be the case. Both Axelson and Brinberg 18and Hayes and Ross, 19 for example, found the strongest predictors of dietary choice and dieting to be associated with the immediate effects of eating, particularly comfort and ability to perform physical exercise. In addition, the parsimony of 'social-cognitive' theories means they lose some of the richness, complexity and, conversely, the mindlessness, associated with much routine health behaviour. Thus they have three serious shortcomings: Routine or non-cognitively mediated health behaviours may be neglected or inappropriately addressed, a~ Much health behaviour is repetitive and habitual. Choices of food type, frequency of exercise, or smoking are frequently based on habit rather than immediate consideration. In a qualitative study, for example, Bremble, Bennett and Morgan 21 found that the food choice of the majority of respondents of middle-age or over was determined by habit, and they rarely considered alternatives. In addition, many respondents could not recall the original reasons for their dietary choice. Moreover, in a larger survey of 13000 people aa we found one 'socialcognition' model, the multi-dimensional

18

R BENNETT, S. MURPHYAND D CARROLL

health locus of control, 23 to account for between only 2-3% of dietary choices. 'Socialcognition' theories may be better at predicting one-off behaviours, such as attending health screening, than more repetitive behaviours which are frequently determined by habit and contextual factors, and engaged in at a routine, non-cognitive level. ii. The individual is frequently seen as the unit of analysis or intervention at the expense of factors in the wider social environment. This represents an unnatural division between behavioural and structuralist accounts of health behaviour, as such behaviour is situated within a social context, be it community, organisation or family. iii. An emphasis is placed on understanding and addressing individual risk factors in isolation, most frequently smoking, diet, exercise and alcohol consumption. Although this may be appropriate to a traditional medical/epidemiological understanding of health, it fails to address more holistic interpretations.24,25

An Emerging Psychological Approach The above, and other shortcomings, suggest that the predominance accorded by health promotion to 'social-cognition' theories can no longer be justified. A wider set of psychological theories and methods are more appropriate to health promotion. In particular: i.

the use of qualitative research to identify determinants of health-related behaviours ii. making programmes sensitive to cultural and sub-cultural determinants of health-related behaviours iii. understanding how health promotion initiatives are received by target populations iv. increasing environmental manipulations and V.

exploring possible psychological mechanisms mediating between relative adverse economic and material circumstances, and ill-health. Of course, these notions are already common

currency in several other disciplines (such as medical sociology and social anthropology) but are only slowly being adopted by psychologists involved in health promotion. 26,27 We acknowledge this debt, but argue that a wider adoption of these approaches may benefit the development of psychology as a whole, and its usefulness in informing health promotion initiatives in particular.

A Qualitative Approach Qualitative research methods may highlight reasons why people engage in many health-related behaviours, beyond those usually identified by rationalistic models. Both Graham 28and Jacobs o n , 29 for example, found that working class women smokers make a rational choice to use smoking as a means of controlling the stress which results from having to cope with adverse social and material circumstances--in the full knowledge of its long term health-damaging effects. As a result of the stresses involved in managing a family, frequently under adverse material circumstances, many such women reported that the time and effort required to quit were not available or not afforded a high priority. Accordingly, health promotion initiatives here may have to address much wider issues such as housing, child care and leisure facilities before effective attempts can be made to help such people try to stop smoking. Ethnographic methods have also been used to identify lay beliefs about the nature, cause and origins of AIDS in order to inform the development of materials for health educators. 3~ These materials have used people's knowledge and emotional reactions to AIDS and HIV infection as a starting point for education. By taking their perceptions of the causes of ill health and incorporating them in a participatory educational approach, a health initiative was adapted to accommodate and exploit existing beliefs in relevant social contexts. Qualitative research may also serve to identify inappropriate assumptions made by health promoters. Such methods, for example, have revealed the apparently simple notion of not engaging in unprotected sex without 'knowing your partner' advocated in a British AIDS campaign of

SOCIALCOGNITION MODELS

the late 1980s to be confusing and irrelevant to many of the young people at whom the intervention was targeted. It was not clear what 'knowing your partner' actually meant, nor how this may be used to determine condom use. The length of time considered necessary to 'get to know' a sexual partner was taken by many of the young people interviewed as simply learning their name. 31 More careful use of qualitative methods prior to the campaign may have prevented this communication failure. In calling for an increased use of qualitative research methods we are not suggesting that these may be used to replace the quantitative approach; instead, we would argue, the two approaches may combine to produce a fuller understanding of the processes involved in decisions about health-related behaviours.

Cultural and Social Determinants As has already been suggested, decisions regarding health behaviour are frequently based on factors far removed from the end-point of health. They may be strongly embedded within cultural or sub-cultural structures. Backett, 32 for example, in examining health behaviours in families, found a complex system of collective negotiations resulting in marked differences in male and female health behaviour. Dorn 33 reported how youthful drinking behaviour is frequently culturally determined. Forms of drink buying, particularly 'round buying', are seen as part of the group's collective response to the material conditions in which it is situated. This behaviour serves to demonstrate the independence and equality of individuals, while enabling reciprocal public exchange. Thus this culture serves to insulate young people from traditional health promotion programmes which are of little relevance to them as individuals and are likely to be strongly countered by the demands of their social group. Nevertheless, once cultural determinants of behaviour are identified health promoters may work within these boundaries. Leathar, 34 for example, interviewed young Glaswegian drinkers to identify determinants and factors controlling their drinking. As a result, a media-based intervention which focused on encouraging them to measure their drinking in 'units' and to keep to

19

the 'recommended' drinking limits was abandoned and was replaced by one which focused on the social costs of drinking excessively (losing control in front of friends, running short of money, and so on).

Responses to Health Promotion An increased use of process evaluation and phased or episodic research is needed to increase our understanding of how health promotion initiatives impact on their target population. The dominant concern of health promotion has been the measurement of intervention outcome effects, usually in terms of behaviour or risk-factor change, with the assumption that these have resulted directly from the intended intervention. While such outcomes have obvious utility, the concentration solely on behavioural or bio-medical outcomes means that the processes of reception, interpretation and response to any intervention by the target population cannot be understood. This emphasis on a traditional 'top down' view of behaviour change may usefully be replaced by one which continually identifies how cultural or sub-cultural processes influence the impact of any intervention, and which modifies initiatives accordingly.

Environmental manipulations The World Health Organisation is presently affording high priority to large scale participatory projects which involve communities in developing healthy environments. 35 These initiatives have obvious roots within the public health movement and seem far removed from psychological theory. Yet an exclusive focus on the wider social environment, and the neglect of individuals' relationships to it, may result in shortcomings of a similar magnitude to those which focus exclusively on intra-psychic processes. Indeed, it is possible that many such interventions may be guided by psychological research and theory. The health belief model makes a clear case for increasing the availability and decreasing the costs of health-promoting behaviours. Accordingly, increased availability (whether geographic, economic, or temporal) of stress buffers such as

20

leisure facilities or creches for working mothers in leisure time, may impact on the health of many disadvantaged social groups. A somewhat simpler strategy is suggested by Hunt and Martin 2~ who suggested a method by which routine behaviours could be changed through increasing the salience of these repetitive behaviours, such that they become more conscious and subject to active consideration. Such changes may be brought about quite simply, for example by moving foodstuffs to slightly different places on the supermarket shelves.

Material and Social Circumstances That the social circumstances in which the individual is situated may impact profoundly on their health is unarguable. More contentious are the explanations for such inequalities in health. Carroll, Bennett and Davey Smith 36,37 have argued that such differentials do not result solely from behavioural differences between various economic groups, although these, of course, play their part. 38 However, there is growing evidence that poor material circumstances may overwhelm the impact of behaviours typically considered to be damaging to health. For example, in a 7-year prospective study of coronary heart disease (CHD) in Danish men reported by Hein, Suadicani and Gyntelberg 39 a number of behavioural and biological risk factors contributed to risk for CHD differentially according to social classification. No significant variation in CHD incidence attributable to smoking was found in the lower socioeconomic groups, whereas, in contrast, smokers in the higher social groups evidenced a five-fold increase in risk for CHD in comparison to those who did not smoke. The process by which social inequality may confer increased risk for disease remains to be fully determined. There are clear differences in health behaviours and environmental insults according to social group, 4~ and these need to be taken into account in any intervention. However, two further psychological factors may be of relevance: differences in social support and differences in levels of stress, both of which are unequally distributed among different social groups, to the disadvantage of those in the lower socio-economic groups. 38,42,43 A number of

R BENNETT, S. MURPHYAND D~CARROLL

studies have identified a link between poor social relations and increased coronary mortality 44 and all-cause mortality. 45,46Analysis of one particular social tie, marriage, reveals a consistent picture of lower mortality and morbidity rates for married relative to single, divorced and widowed individuals. 41"47 Studies of the frequency of daily hassles, major life events, and stress at work reveal a small but significant relationship between the frequency and extent of such stressors and mortality and various indices of morbidity. 48 When the presence of stress is confounded by a lack of control over the cause of the stress, the outcome may be particularly damaging. Alfredsson, Spetz and Theorell, 49 for example, examined the prevalence of CHD across all occupational groups in Sweden. Occupations were classified in terms of high and low demand, and high and low control. CHD was most prevalent among workers in high demand occupations (typically blue collar workers) who had the least control over what they did at work, and when they did it. Should this speculation be substantiated by future research it would suggest a number of interventions and intervention points in future health promotion planning. Some interventions which are already being instituted by some organisations, such as increasing job diversity and control could be extended. Other implications suggested are the positive health benefits of developing local community social support networks where these have not formed naturally and, as previously noted, facilities which may act as stress buffers. Evidence of the potential impact of such initiatives is sparse, and at least one study 5~ has been unable to find any advantage conferred by increasing social support. However, this study concerned the impact of increased social support on maternal health and the incidence of low birthweight babies and may involve different mechanisms than those involved in adult health maintenance, suggesting the need for caution in translating epidemiological evidence to any consideration of the impact of health promotion initiatives.

Conclusion Until now the most significant contribution of psychology to health promotion has been the use

SOCIALCOGNITION MODELS

of social cognition models to provide a structure and rationale for complex intervention strategies. These theories have been used as the foundation of a n u m b e r of major health p r o m o t i o n programmes. However, the centrality accorded to such theories has caused other psychological models and methods to be neglected. This, we argue, has been to the detriment of health promotion, not only because other psychological approaches m a y benefit health promotion and population health, but also because interventions which emphasise individual responsibility for health m a y actively foster a climate of 'victim blaming'. If psychology is to maintain its role as one of the p r i m a r y disciplines which inform and support health promotion it will need to address more fully a wider set of issues, and in particular the social and material context of behaviour. The World Health Organisation has advocated a multi-level approach to health promotion which moves b e y o n d the individual to encompass social and physical environments conducive to health. At the m o m e n t m u c h of this work, though theoretically developed, requires application. The use, and further development, of social contextual models m a y provide at least some of the means of achieving this.

Paul Bennett, Simon Murphy and Douglas Carroll References 1. Rawson, D. (1994). Will-power and authentic choice in stopping smoking. Health Care Analysis 2(3), 201-205. 2. Bennett, P. and Hodgson, R. (1992). Psychology and health promotion. In, Health Promotion: Disciplines and Diversity, ed. by R. Burton and G. Macdonald, Routledge, London. 3. Ingham, I. and Bennett, P. (1990). Health psychology in community settings: models and methods. In, Current Developments in Health Psychology, ed. by P. Bennett, J. Weinman and P. Spurgeon, Harwood, Press, London. 4. Bury, M. (1994). Health promotion and lay epidemiology: a sociological view. Health CareAnalysis 2(1), 23-30. 5. Maccoby, N. (1988). The community as a focus for health promotion. In, The Social Psychology of Health, Sage, Newbury Park, California. 6. Puska, P. (ed) (1988). ComprehensiveCardiovascular Community Control Programmes in Europe, Euro Reports and Studies 106, Copenhagen, World Health Organisation.

21

7. Farquhar, J. W., Eortmann, S. P., Maccoby, N., Haskell, W. L., Williams, P. T., Flora, J. A., Taylor, C. B., Brown, B. W., Solomon, D. S. and Holley, S. B. (1985). The Stanford five city project: design and methods. American Journal of Epidemiology 122, 323-334. 8. Sarver, V. T. (1983). Ajzen and Fishbein's theory of reasoned action: a critical assessment. Journalfor the Theory of Social Behaviour 13, 155-163. 9. Ajzen, I. and Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behaviour, Prentice Hall, Englewood Cliffs, NJ. 10. Puska, P., Nissinen, A., Tuomilehto, J., Salonen, J. T., Koskela, K., McAlister, A., Kottke, T. E., Maccoby, N. and Farquhar, J. W. (1985). The community-based strategy to prevent coronary heart disease: conclusions from the ten years of the North Karelia project. Annual Review of Public Health 6, 147-193. 11. Bandura, A. (1977). Social Learning Theory, Prentice Hall, Englewood Cliffs, NJ. 12. Bennett, P., Smith, C., Nugent, Z. and Panter, C. (1991). Pssst ... the really useful guide to alcohol: evaluation of an alcohol education television series. Health Education Research, Theory and Practice 6, 57-64. 13. Winett, R. A., King, A. C. and Altman, D. G. (1989). Health Psychology and Public Health, Pergamon Press, New York. 14. De Foe, J. R. and Breed, W. (1988/9). Consulting to change media contents: two cases in alcohol education. International Quarterly of Community Health Education 4, 257-272. 15. Becker, M. H. (1974). The health belief model and personal health behaviour. Health Education Monographs 2, 324-508. 16. Catford, J. and Parish, R. (1989). 'Heartbeat Wales': new horizons for health promotion in the comm u n i t y - t h e philosophy and practice of Heartbeat Wales. In, Changing Ideas in Health Care, ed. by D. F. Seedhouse and A. Cribb, John Wiley and Sons, Chichester. 17. Murphy, S., Smith, C. and Powell, C. (1993). An Evaluation of the Heartbeat Award Scheme, Technical Report No. 3, HPW, Cardiff. 18. Axelson, M. L. and Brinberg, D. (1989). A Social19. 20. 21. 22.

Psychological Perspective on Food-related Behaviour, Springer-Verlag, New York. Hayes, D. and Ross, C. E. (1987). Concern with appearance, health beliefs, and eating habits. Journal of Health and Social Behaviour 28, 120-130. Hunt, S. M. and Martin, C. J. (1988). Health-related behavioural change---a test of a new model. Psychology and Health 2, 209-230. Bremble, A., Bennett, P. and Morgan, M. (1993). Reasons for choice of health-related behaviours: a qualitative study. Unpublished manuscript. Bennett, P., Moore, L., Murphy, S. and Smith A. (1994). Health locus of control and value for health as predictors of dietary behaviour. Psychology and Health, in press.

22

23. Wallston, K. A., Wallston, B. S. and DeVellis, R. (1978). Development of the multi-dimensional health locus of control. Health Education Monographs 6, 160-170. 24. Antonovsky, A. (1993). The structure and properties of the sense of coherence scale. Social Scienceand Medicine 36, 725-733. 25. Wright, S. J. (1987). Self ratings of health: the influence of age and smoking status and the role of different explanatory models. Psychology and Health 1, 379-397. 26. Abraham, C., Fife-Schaw, C., Ingham, R., Scott, S. and Sheeran, P. (1993). Quantative and qualitative methods in health psychology. Health Psychology Update 13, 15-18. 27. Buchanan, D. (1992). An uneasy alliance combining qualitative and quantative research methods. Health Education Quarterly, 19, 1-18. 28. Graham, H. (1988). Smoking in pregnancy: the attitudes of expectant mothers. Social Science and Medicine 10, 399-405. 29. Jacobson, B. (1981). TheLadykillers: Why Smoking is a Feminist Issue, Pluto, London. 30. Warwick, I., Aggleton, P. and Homans, H. (1988). Constructing commonsense--young people's beliefs about AIDS. Sociology of Health and Illness 10, 213-233. 31. Ingham, R., Woodcock, A. and Stenner, K. (1991). Getting to know you.., young people's knowledge of their partners at first intercourse. Journal of Community and Applied Social Psychology 1, 117-132. 32. Backett, K. (1990). Studying health in families: a qualitative approach. In, Readings in Medical Sociology, ed. by I. Cunningham, S. Burly and N. McKeganey, Routledge, London. 33. Dorn, N. (1983). Alcohol, Youth, and the State, Croom Helm, London. 34. Leathar, D. S. (1981). Lack of response to health guidance amongst heavy drinkers. In, Preventive Nutrition and Society, ed. by M. R. Turner, Academic Press, London. 35. Ashton, J. and Seymour, H. (1988). The New Public Health, Open University Press, Milton Keynes. 36. Carroll, D., Bennett, P. and Davey Smith, G. (1993). Health and social and material circumstances: their origins and implications. Psychology and Health 8, 295-316. 37. Carroll, D., Davey Smith, G. and Bennett, P. (1994). Health and socio-economic status. The Psychologist: Bulletin of the British Psychological Society 7(3), 10-13. 38. Berkman, L. F. and Breslow, L. (1983). Health and Ways of Living: The Alameda County Study, Oxford University Press, Oxford. 39. Hein, H. O., Suadicani, P. and Gyntelberg, F. (1992). Ischaemic heart disease incidence by social class and form of smoking: the Copenhagen male study--17 years' follow-up. Journal of Internal Medicine 231, 477-483. 40. Amler, R. W. and Dull, H. D. (eds) (1987). Closing

R BENNETT,S.MURPHYANDD CARROLL

41. 42.

43. 44.

45.

46.

47.

the Gap. The Burden of Unnecessary Illness, Oxford University Press, Oxford. Blaxter, M. (1990). Health and Lifestyle, Tavistock, London. Marmot, M. G., Davey Smith, G., Stansfeld, D., Patel, C., North, F., Head, J., White, I., Brunner, E. and Feeney, A. (1991). Health inequalities among British civil servants; the Whitehall II study. Lancet 337, 1387-1392. Williams, D. R. (1990). Socio-economic differentials in health: a review and redirection. Social Psychology Quarterly 53, 81-99. Orth-Gomer, K. and Unden, A. L. (1990). Type A behaviour, social support, and coronary risk: interaction and significance for mortality in cardiac patients. Psychosomatic Medicine 52, 59-72. Orth-Gomer, K. and Johnson, J. V. (1987). Social network interaction and mortality: a six year follow-up of the Swedish population. Journal of Chronic Disease 40, 949-957. Schoenbach, V., Kaplan, B., Freedman, L. and Kleinbaum, D. (1986). Social ties and mortality in Evans County, Georgia. American Journal of Epidemiology 123, 577-591. Fox, A. J. and Goldblatt, P. D. (1982). The OPCS

Longitudinal Study: Socio-Demographic Mortality Differentials 1971-75, OHMS, London. 48. Carroll, D. (1992). Health Psychology: Stress, Behaviour and Disease, Falmer Press, London. 49. Alfredsson, L., Spetz, C. L. and Theorell, T. (1985). Type of occupation and near-future hospitalisation for myocardial infarctions and other diagnoses. International Journal of EpidemioIogy 14, 378-388. 50. Villar, J., Farnot, U., Barros, F., Victoria, C., Langer, A. and Belizan, J. M. (1992). A randomized trial of psychosocial support during high-risk pregnancies. The New England Journal of Medicine 327, 1266-1271.

Paper Two: What You Do Is Determined By What You Do lan G r o u n d

Jake Skepsis from the magazine Arts Analysis has an interview with Dr Plurabelle of the Department

of Aesthemiology. Plurabelle:

Ah, yes, the journalist chap, n o w do sit down. Some coffee? Yes? Have a seat dear boy. Splendid. Splendid. N o w what can I do for you?

Skepsis:

Well, Dr Plurabelle, I'm writing a feature for m y arts magazine and y o u r name came u p quite a lot in m y research. And I was w o n d e r i n g if y o u could tell our readers, as simply