J Relig Health DOI 10.1007/s10943-017-0389-x ORIGINAL PAPER
Smoking Prevalence Among Users of Primary Healthcare Units in Brazil: The Role of Religiosity Edson Zangiacomi Martinez1 • Fla´via Masili Giglio1 • Natalia Akemi Yamada Terada1 • Anderson Soares da Silva1 Miriane Lucindo Zucoloto1
•
Ó Springer Science+Business Media New York 2017
Abstract The objective of this cross-sectional study is to examine the association between religious involvement and tobacco use in a large representative sample of users of primary healthcare units of Ribeira˜o Preto, Southeast Brazil. Current and past smoking habits were determined among 1055 users of primary healthcare units. Participants’ religiosity was measured using the DUREL questionnaire. The prevalence of smoking among men was 16.8% [95% confidence interval (CI) 12.0–22.5] and among women was 12.6% (95% CI 10.4–15.0). Among the current smokers, 40.9% were light smokers, 24.6% were moderate smokers, and 34.5% were heavy smokers. The mean number of cigarettes smoked per day was 13.5. Respondents who have a religion had a lower smoking prevalence than people who had no religion. Current smoking prevalence tended to be higher among people who do not practice their religion than people who practice their religion. Smoking status is also associated with self-reported religiosity, organizational religious activity and some aspects of intrinsic religiosity. Religiosity is an important factor in influencing the smoking behavior in Brazilian users of the public health services. Keywords Smoking Religion Religiosity Public health services
Introduction Cigarette smoking and alcohol consumption are well-known risk factors for non-communicable chronic diseases (Linardakis et al. 2013). Despite the known health problems associated with cigarette smoking, its prevalence remains high and remains a worldwide problem. In a study that used data from 187 countries, the overall prevalence of daily
& Edson Zangiacomi Martinez
[email protected] 1
Ribeira˜o Preto Medical School, University of Sa˜o Paulo (USP), Av. Bandeirantes 3900, Monte Alegre, Ribeira˜o Preto 14049-900, Brazil
123
J Relig Health
smoking for the years 1980, 1996, 2006 and 2012, standardized by age, were respectively estimated as 41.2, 38.5, 32.6 and 31.1% for male individuals, and 10.6, 8.3, 6.8 and 6.2% for female individuals (Ng et al. 2014). This study showed that the prevalence of daily smoking in 2012 is similar among men living in developed and developing countries (30.1 and 32.0%, respectively). However, it was observed a deep difference when considering only the female population, with a prevalence of daily smoking estimated at 17.2% for women living in developed countries and 3.7% among those living in developing countries (Ng et al. 2014). Prevalence estimates of smoking for the Brazilian population presented by these authors were based on the Telephone-based Surveillance of Risk and Protective Factors for Chronic Diseases (VIGITEL) study. Estimates for the male population were 25.9, 21.3, 18.8, and 16.5% for the years 1980, 1996, 2006, and 2012, respectively, and 15.6, 14.4, 12.0, and 11.0% for the female population, considering the same years (Ng et al. 2014). In the same study, it was estimated for these 187 countries an average daily consumption of 17.7 cigarettes per smoker. Considering the Brazilian population, this value was estimated at 22.8 cigarettes per smoker in 2012. This average is higher than that estimated for the population of Argentina (19.7), Bolivia (3.1), Chile (9.2), Colombia (10.1), Ecuador (14.8), Mexico (11.8), Peru (4.0), and Venezuela (19.6) and similar to those estimated for Paraguay (23.2) and Uruguay (25.4). Also, the 2008 National Household Sample Survey (PNAD) interviewed more than 250,000 individuals and estimated the prevalence of daily smoking in Brazil 15.1% (Barros et al. 2011). This prevalence was higher in men, people aged 40 and 59 years, and people with lower household income, and it is similar to that estimated in 2012 for the US population, where 18.1% of adults were smokers and 14.2% smoked daily (Agaku et al. 2014). In the Study of Cardiovascular Risk Factors in Adolescents (ERICA), more than 74 thousand adolescents living in Brazilian cities with more than 100,000 inhabitants were interviewed about their habits regarding smoking. Among these adolescents, 18.5% had smoked at least once in life, 5.7% were smoking at the time of the study, and 2.5% smoked regularly (Figueiredo et al. 2016). Other studies showed that the prevalence of cigarette experimentation among Brazilian adolescents was 30.2% in the city of Cuiaba´ (Silva et al. 2008) (Midwestern Brazil), 16.1% in Salvador (Neto et al. 2010) (Northeastern Brazil) and 11.6% in Santa Maria (Horta et al. 2001) (Southern Brazil). Some of the main variables related to cigarette experimentation are low maternal level of education (Silva et al. 2008), low socioeconomic level (Silva et al. 2008; Horta et al. 2001), studying in a public school (Silva et al. 2008; Horta et al. 2001), having divorced parents (Silva et al. 2008; Mak et al. 2012), alcohol consumption (Neto et al. 2010; Mak et al. 2012), paternal smoking (Neto et al. 2010), having friends who smoke (Silva et al. 2008; Neto et al. 2010; Horta et al. 2001; Malta et al. 2015) and having a sibling or boyfriend/girlfriend who smoke (Silva et al. 2008; Neto et al. 2010). Because most adult smokers begin to smoke in adolescence (Horta et al. 2001; McVicar 2011), this age group seems particularly important for interventions and prevention strategies. A study published in 1997 found that 24.6% of the scholar adolescents living in Ribeira˜o Preto (Southeastern Brazil) with age group from 13 to 15 years old had already experimented cigarettes (Muza et al. 1997). This percentage increased to 43.1% among adolescents 16–17 years old and to 50.2% among those aged 18–19 years (Muza et al. 1997). These numbers seems to be quite high when compared with those in studies reported in the literature (Silva et al. 2008; Neto et al. 2010; Horta et al. 2001), which makes this population interesting for new epidemiological studies about smoking and related diseases. The relationship between religiosity and smoking has been studied by several authors in different contexts, but the effects of religious involvement on the smoking can be
123
J Relig Health
inconsistent in different cultural and ethnic groups (Nollen et al. 2005; Yong et al. 2009). It is known that more religious women appear to be less likely to use tobacco during pregnancy (Mann et al. 2007), and religiosity may play an important role in the prevention of substance use among adolescents (Gryczynski and Ward 2011; Alexander et al. 2015). However, differences in patterns of religiosity and the different religious affiliations that characterize a given population should be considered, so that the role of religion on the health practices should be studied without neglecting these specificities. Brazil has a large population with many different religious groups, including Catholics, Protestants, Evangelicals, and a variety of other spiritualist religions, but it has a relatively small number of Buddhists, Taoists, Muslims, and Jews when compared with other populations (http:// www.ibge.gov.br/home/estatistica/populacao/censo2010/). In this context, more and more research is focusing on the relationship between religious involvement and cigarette smoking in Brazil (Bezerra et al. 2009). As an example, a study (Silva et al. 2011) based on a sample of high school students of the State of Pernambuco, Brazil, showed that, regardless of gender, exposure to alcohol consumption and smoking was inversely associated with religious affiliation and religious practices. However, most of these studies have been focused primarily on adolescent populations and little research has focused on users of health services (Queiroz et al. 2015). The importance of conducting studies about the health habits of people who use the primary healthcare services involves aspects that include the early identification of smoking exposure and consequent preventive actions, planning health promotion strategies and campaigns for the cessation of smoking addiction. The present study was undertaken to examine the association between religious involvement and tobacco use in a large representative sample of users of primary healthcare units of Ribeira˜o Preto, Southeast Brazil. The objectives of this study were (1) to describe the prevalence of smoking consumption and identify the social and demographic factors associated with smoking among the users of primary healthcare units and (2) to explore associations between religiosity and tobacco use.
Methods This sectional study considers a stratified sample of users of primary care health units of Ribeira˜o Preto, State of Sa˜o Paulo (21°100 41.500 S 47°480 39.000 W), Brazil. Ribeira˜o Preto is one of the largest municipalities of the State, with approximately 600,000 inhabitants. Its per capita income is one of the highest in the country, and its economy is based on agribusiness. At the time of data collection, the primary healthcare in this municipality was composed of 41 units divided into five health districts (North, South, Central, East and West) with diverse demographic characteristics. For the sampling purposes of this study, the units were classified using the Sa˜o Paulo Social Vulnerability Index (IPVS), based on their predominant areas of coverage. This index, proposed by the Brazilian State Data Analysis System Foundation, classifies geographical areas into six categories of social vulnerability (Ferreira et al. 2006). The units were thus grouped by health district and IPVS classification to form 12 strata. After stratifying the population, it was randomly selected one unit from each stratum for the interviews. The sample size determination used a stratified sampling design (Scheaffer et al. 2011), and it was based on the estimates of the mean number of monthly health visits in each unit. It was thus considered a 95% confidence level and a 3% absolute precision to estimate the smoking prevalence in this
123
J Relig Health
population. This prevalence was considered to be 50% in each stratum to maximize the variance, and it was thus calculated that a minimum sample size of 1055 individuals is required for this study. Data were collected through personal interviews conducted by three trained interviewers from August 2015 to May 2016. Participants were invited to participate in the study while they waited for medical care in the health units. Individuals with less than 18 years of age or who do not live in Ribeira˜o Preto were not included in this study. Socioeconomic status was assessed using the classification proposed by the Brazilian Association of Market Research Institutions, which is based on the accumulation of material goods and the schooling of the head of the household. More privileged socioeconomic classes are labeled by A, while the intermediate classes are labeled by B and less privileged classes as classified on an ordinal scale by C1, C2, D and E. The educational level was classified into eight groups: insufficient (including illiterates, people who have never attended school and people who did not complete the fundamental I level), complete fundamental I, incomplete and complete fundamental II, incomplete and complete medium school, and incomplete and complete higher education. In the Brazilian educational system, ‘‘fundamental I’’ corresponds to first to fifth year of the elementary education while ‘‘fundamental II’’ corresponds to sixth to ninth year, ‘‘medium school’’ is equivalent to the high school and ‘‘higher education’’ corresponds to the college, undergraduate schools or the university level courses. People with self-reported diabetes were defined as those who responded ‘‘yes’’ to the question: ‘‘Have you ever in your life, had a doctor that ever told you have diabetes?’’ Persons with self-reported high blood pressure are identified in a similar manner. Quality of life was measured using the 15-item Flanagan Quality of Life Scale (QOLS) (Flanagan 1982), modified by adding the 16th item on independence, as recommended by Burckhardt et al. (1989). A higher score indicates a higher quality of life. The Alcohol Use Disorder Identification Test (AUDIT) (Reinert and Allen 2002) was used to evaluate the pattern of alcohol use. AUDIT classifies the participants into four risk zones, according to the score obtained: non-drinker, non-hazardous drinking, hazardous drinking and harmful drinking. These categories are, respectively, labeled as I, II, III and IV. Participants were classified as current cigarette smokers, ex-smokers, and lifelong nonsmokers from their responses to questions about smoking habits. Current smokers were categorized into three levels of tobacco consumption, based on the reported average number of cigarettes smoked per day: light smoker (1–9 cigarettes), moderate smoker (10–19 cigarettes), and heavy smoker (20 or more cigarettes). This classification is based on that used by Alati et al. (2004). Religiosity was measured using the Duke University Religion Index (DUREL) questionnaire (Koenig and Bu¨ssing 2010). The DUREL is a five-item questionnaire that assesses three subscales of religiosity: organizational religious activity (ORA), non-organizational religious activity (NORA), and intrinsic religiosity (IR). ORA is assessed with the question ‘‘How often do you attend church or other religious meetings?’’ using a sixpoints scale ranging from ‘‘more than once a week’’ to ‘‘never.’’ NORA is the behavior that occurs apart from the organized religious communities, such as prayer, Scripture study, watching religious TV or listening to religious radio, and it is assessed with the question ‘‘How often do you spend time in private religious activities, such as prayer, meditation or Bible study?’’, with responses also ranging from ‘‘more than once a week’’ to ‘‘never’’ (Koenig and Bu¨ssing 2010). IR is a three-item subscale that assesses degree of personal religious commitment or motivation. The three IR items use a five-point scale, and the total IR score is calculated by summing the scores of these items that ranged from 3 to 15. DUREL is a reliable and valid measure that is widely used in different cultures and a
123
J Relig Health
previous validation of the questionnaire in the context of Brazilian users of primary healthcare units also provided satisfactory results of reliability and validity (Martinez et al. 2014). Other dimensions or aspects of individual religiosity considered in this study include the religious affiliation and the self-reported religiosity (the possible answers are ‘‘very religious,’’ ‘‘moderately religious,’’ and ‘‘a little or nothing’’). Continuous variables are described by means and standard deviations (SD), and categorical variables by frequencies and percentages. Exact 95% confidence intervals (95% CI) for prevalence of smoking were calculated using the binomial distribution. A multinomial logistic regression was fitted to the data (Hosmer et al. 2013), using as dependent variable a three-level indicator classifying the individuals as current smokers, ex-smokers and people who never smoked. Odds ratios (OR) were used as measures of association between each independent variable and the smoking status. In order to control the possible confounding effect of sex and age, the logistic regression analysis of the association between the smoking status and the variables related to religiosity included sex and age groups as covariates. Ninety-five percent confidence intervals for the OR were obtained. Intervals that do not include 1 are interpreted to mean that the respective independent variable is associated with the smoking status at the 0.05 significance level (similar to usual p \ 0.05). A regression model based on the beta-binomial distribution (Martinez et al. 2015) was used to compare the means for ORA, NORA, and IR subscales of the DUREL, according to sex and levels of tobacco consumption. These models are appropriate for analyzing data that follow a discrete distribution (variables that take only integer values) rather than the usual normal distribution (Martinez et al. 2015). A two-way ANOVA followed by Bonferroni post hoc comparison was used to compare the means of the Flanagan QOLS between sex and smoking groups. The SAS 9.3 software was used for all statistical analyses.
Results The sample was composed of 214 (80%) men with mean age of 39.5 years (SD 14.9) and 841 (20%) women with mean age of 45.1 years (SD 15.3). Of these participants, 142 reported being current smokers (13.4%), 216 ex-smokers (20.5%), and 697 never smoked (66.1%). The prevalence of smoking among men is 16.8% (95% CI 12.0–22.5) and among women is 12.6% (95% CI 10.4–15.0). Among the current smokers, 58 (40.9%) are light smokers, 35 (24.6%) are moderate smokers, and 49 (34.5%) are heavy smokers. The mean number of cigarettes smoked per day was 13.5 (SD 10.3). Considering both current and past smokers, the mean age of smoking initiation was 15.5 years among the men (SD 4.1) and 17.3 years among the woman (SD 5.7). The ORs with their respective 95% CI showed in the last two columns of Table 1 evidence that the smoking status is associated with sex, age groups, educational level, marital status, self-reported hypertension, self-reported diabetes, and AUDIT risk level. In Table 1, it can be observed that the prevalence of smoking is higher among the age group of 51–60, people with lower education, divorced individuals, hypertensive participants and hazardous and harmful drinkers. The prevalence of smoking is also higher in lower socioeconomic groups (D and E), but the 95% CIs for the respective ORs did not evidenced a significant association (Table 1). Table 2 shows the prevalence of current and past smoking for categories of the variables related to religiosity. All ORs showed in this table were adjusted by sex and age groups, given the association between these variables and the smoking status. Those who have a
123
J Relig Health Table 1 Prevalence of current and past smoking by demographics and clinical characteristics in users of primary health care, Ribeira˜o Preto, Southeast Brazil n
Smoking status (%) Current cigarette smokers
Exsmokers
Never smoked
Smokers versus never smoked OR1 (95% CI)
Ex-smokers versus never smoked OR2 (95% CI)
Sex Women
841
12.6
19.3
68.1
Reference
Reference
Men
214
16.8
25.2
58.0
1.60 (1.03–2.40)*
1.54 (1.07–2.22)*
Age group (years) Until 25
188
9.6
7.4
83.0
Reference
Reference
26–30
132
12.9
13.6
73.5
1.52 (0.74–3.09)
2.07 (0.98–4.34)
31–40
250
13.2
19.2
67.6
1.69 (0.91–3.13)
3.16 (1.67 – 5.96)*
41–50
173
14.4
23.7
61.9
2.02 (1.05–3.89)*
4.27 (2.21–8.21)*
51–60
164
19.5
28.1
52.4
3.22 (1.71–6.08)*
5.96 (3.09–11.45)*
[60
148
11.5
33.1
55.4
1.80 (0.87–3.67)
6.65 (3.46–12.76)*
Complete higher education
67
10.5
16.4
73.1
Reference
Reference
Incomplete higher education
59
11.9
20.3
67.8
1.22 (0.39–3.78)
1.34 (0.53–3.35)
Complete medium school
405
11.1
14.6
74.3
1.05 (0.44–2.45)
0.87 (0.42–1.78)
Incomplete medium school
112
11.6
21.4
67.0
1.21 (0.45–3.26)
1.42 (0.64–3.17)
Complete fundamental II
117
13.7
25.6
60.7
1.58 (0.60–4.12)
1.88 (0.86–4.11)
Incomplete fundamental II
104
16.4
20.2
63.5
1.80 (0.69–4.68)
1.42 (0.62–3.21)
61
21.3
24.6
54.1
2.76 (0.99–7.64)
2.02 (0.82–4.95)
130
18.5
33.9
47.7
2.71 (1.08–6.81)*
3.16 (1.47–6.76)*
648
11.6
20.8
67.6
Reference
Reference
53
13.2
35.9
50.9
1.51 (0.63–3.60)
2.28 (1.23–4.23)*
Single
247
14.6
15.0
70.4
1.21 (0.78–1.87)
0.69 (0.46–1.03)
Divorced
107
22.4
23.4
54.2
2.42 (1.41–4.13)*
1.40 (0.84–2.32)
Educational level
Complete fundamental I Insufficient Marital status Married Widowed
Socioeconomic status A or B
355
11.2
18.9
69.9
Reference
Reference
C1
368
14.1
21.5
64.4
1.36 (0.86–2.13)
1.23 (0.85–1.79)
C2
235
14.0
21.7
64.3
1.36 (0.81–2.24)
1.25 (0.82–1.89)
97
17.5
19.6
62.9
1.73 (0.91–3.25)
1.15 (0.64–2.06)
D or E
Self-perception of health
123
J Relig Health Table 1 continued n
Smoking status (%) Current cigarette smokers
Exsmokers
Never smoked
Smokers versus never smoked OR1 (95% CI)
Ex-smokers versus never smoked OR2 (95% CI)
Good
735
12.4
19.0
68.6
Reference
Reference
Regular
280
15.7
23.2
61.1
1.43 (0.95–2.12)
1.37 (0.97–1.93)
40
17.5
27.5
55.0
1.76 (0.73–4.25)
1.80 (0.85–3.80)
Poor
Have health insurance Yes
154
10.4
18.8
70.8
Reference
Reference
No
901
14.0
20.7
65.3
1.45 (0.83–2.55)
1.19 (0.76–1.86)
Self-reported hypertension No
823
12.8
17.5
69.7
Reference
Reference
Yes
232
15.9
31.0
53.0
1.64 (1.07–2.51)*
2.33 (1.65–3.29)*
Self-reported diabetes No
971
13.7
19.0
67.3
Reference
Reference
Yes
84
10.7
38.1
51.2
1.03 (0.49–2.16)
2.64 (1.62–4.30)*
AUDIT risk level I
889
9.8
20.4
69.9
Reference
Reference
II
136
30.2
22.1
47.8
4.5 (2.8–7.1)*
1.6 (0.9–2.5)
30
46.7
16.7
36.7
9.1 (3.9–20.6)*
1.5 (0.5–4.5)
III or IV
OR odds ratio, CI confidence interval Confidence intervals that do not include 1 are marked with an asterisk (similar to p \ 0.05)
religion have a lower prevalence of current smoking (12%) than people who have no religion (23%). Among the participants who have a religion, it is noticeable that the prevalence of current smoking varies across different types of religious affiliations. The lowest smoking prevalence is observed among Evangelical people (7.2%) and the higher prevalence is observed among Spiritists (21%). Table 2 also shows that the current smoking prevalence tended to be higher among people who do not practice their religion (17.1%) than people who practice their religion (10.1%). Growing up in a religious home does not seem to be associated with smoking status (95% CIs for the respective ORs include the value 1). The results in Table 2 indicate an association between ORA and smoking status, where the prevalence of current smoking is higher for respondents with lower frequency in religious meetings. However, the analysis did not indicate a significant association between NORA and the smoking status. Among the three items of the IR subscale, only the third item is associated with the smoking status (‘‘I try hard to carry my religion over into all other dealings in life’’). Table 3 shows means and standard deviations for ORA, NORA and IR subscales of the DUREL, according to sex and levels of tobacco consumption, with ORA and NORA measures ranging from 1 to 6 and IR ranging from 3 to 15, so higher values indicate higher religiosity. According to the results of the regression analysis based on a beta-binomial distribution, there is evidence that women who are heavy smokers or light smokers have lower mean ORA and IR values than women who do not smoke. However, there is no evidence that the mean values of NORA are different among the levels of tobacco
123
J Relig Health Table 2 Prevalence of current and past smoking by variables associated with religiosity, including the five items of the DUREL, Ribeira˜o Preto, Southeast Brazil n
Smoking status Current cigarette smokers
Exsmokers
Never smoked
Smokers versus never smoked OR1 (95% CI)a
Ex-smokers versus never smoked OR2 (95% CI)a
Have a religion Yes
916
12.0
21.6
66.4
Reference
Reference
No
139
23.0
13.0
64.0
2.47 (1.53–3.98)*
0.90 (0.51–1.57)
Religious affiliationb Evangelic
332
7.2
22.0
70.8
Reference
Reference
Catholic
495
14.3
21.4
64.3
2.20 (1.32–3.65)*
0.98 (0.68–1.40)
Spiritist
62
21.0
25.8
53.2
3.77 (1.72–8.23)*
1.36 (0.69–2.65)
Others
27
7.4
11.1
81.5
1.14 (0.24–5.24)
0.55 (0.15–1.95)
Practice their religionb Yes
665
10.1
20.6
69.3
Reference
Reference
No
251
17.1
24.3
58.6
2.01 (1.30–3.10)*
1.41 (0.98–2.04)
Are you a religious person?b Very religious
215
7.0
21.4
71.6
Reference
Reference
Moderately religious
559
13.0
21.5
65.5
2.35 (1.29–4.28)*
1.37 (0.91–2.06)
A little or nothing
142
15.5
22.5
62.0
3.32 (1.60–6.90)*
1.85 (1.06–3.22)*
Growing up in a religious home No
280
11.8
20.7
67.5
Reference
Reference
Yes
775
14.0
20.4
65.6
1.17 (0.76–1.81)
0.88 (0.61–1.26)
How often do you attend church or other religious meetings? (ORA) More than once/week
269
8.2
21.2
70.6
Reference
Reference
Once a week
314
10.8
20.1
69.1
1.46 (0.82–2.60)
1.03 (0.67–1.57)
A few times a month
119
19.3
20.2
60.5
2.95 (1.53–5.69)*
1.16 (0.65–2. 04)
A few times a year
178
16.3
22.5
61.2
2.48 (1.34–4.58)*
1.35 (0.83–2.19)
Once a year or less
84
19.1
16.7
64.3
2.69 (1.31–5.55)*
0.98 (0.49–1.93)
Never
91
19.8
19.8
60.4
3.03 (1.49–6.14)*
1.16 (0.61–2.18)
How often do you spend time in private religious activities, such as prayer, meditation, or Bible study? (NORA) More than once/week
184
12.0
22.8
65.2
Reference
Reference
Once a week
629
13.4
22.3
64.4
1.13 (0.67–1.91)
1.00 (0.66–1.51)
A few times a month
88
9.1
17.1
73.9
0.67 (0.28–1.61)
0.68 (0.34–1.34)
A few times a year
52
17.3
15.4
67.3
1.31 (0.54–3.15)
0.62 (0.26–1.48)
Once a year or less
42
11.9
9.5
78.6
0.86 (0.29–2.50)
0.36 (0.11–1.13)
123
J Relig Health Table 2 continued n
Smoking status Current cigarette smokers
Never
60
23.3
Exsmokers
11.7
Never smoked
65.0
Smokers versus never smoked OR1 (95% CI)a
Ex-smokers versus never smoked OR2 (95% CI)a
1.96 (0.90–4.25)
0.53 (0.21–1.31)
In my life, I experience the presence of the Divine (i.e., God) (IR) Definitely true of me
965
13.5
20.8
65.7
Reference
Reference
Tends to be true
79
11.4
16.5
72.2
0.88 (0.42–1.85)
0.93 (0.48–1.77)
Unsurec
11
9.1
27.3
63.6
1.87 (0.57–6.06)
1.59 (0.48–5.25)
Tends not to be truec
1
–
100.0
–
–
–
Definitely not truec
8
37.5
–
62.5
–
–
My religious beliefs are what really lie behind my whole approach to life (IR) Definitely true of me
773
12.3
21.0
66.8
Reference
Reference
Tends to be true
197
15.2
21.3
63.4
1.54 (0.96–2.48)
1.41 (0.93–2.13)
48
18.8
14.6
66.7
1.71 (0.78–3.76)
0.83 (0.35–1.98)
7
28.6
14.3
57.1
2.74 (0.46–16.1)
0.85 (0.08–8.24)
30
20.0
13.3
66.7
1.83 (0.69–4.84)
0.88 (0.28–2.71)
Unsure Tends not to be true Definitely not true
I try hard to carry my religion over into all other dealings in life (IR) Definitely true of me
610
9.0
22.3
68.7
Reference
Reference
Tends to be true
262
15.3
17.9
66.8
2.11 (1.33–3.34)*
1.03 (0.70–1.53)
Unsure
110
22.7
20.0
57.3
3.74 (2.12–6.59)*
1.47 (0.85–2.55)
Tends not to be true
17
29.4
5.9
64.7
5.55 (1.75–17.6)*
0.50 (0.06–4.09)
Definitely not true
56
30.4
17.9
51.8
6.00 (2.99–12.0)*
1.52 (0.69–3.32)
ORA organizational religious activity, NORA non-organizational religious activity, IR intrinsic religiosity, OR odds ratio, CI confidence interval Confidence intervals that do not include 1 are marked with an asterisk (similar to p \ 0.05) a
Odds ratios adjusted by sex and age
b
Considering the n = 916 respondents who reported having a religion
c
In the regression logistic analysis, these classes were combined into a single class due to small sample size in each group
consumption. Among men, no significant differences were found between the levels of tobacco consumption regarding the mean ORA and IR values. However, the small sample size for men may have limited the ability of the model to detect real differences. There is
123
J Relig Health Table 3 Means and standard deviations for ORA, NORA and IR subscales of the DUREL, according to sex and levels of tobacco consumption Levels of tobacco consumption
Women n
Man Mean (SD)
n
Mean (SD)
ORA No smokers
735
4.41 (1.53)
178
3.84 (1.70)
Light
46
3.57 (1.53)*
12
3.33 (1.37)
Moderate
30
4.20 (1.52)
5
3.40 (1.14)
Heavy
30
3.77 (1.77)*
19
3.74 (1.94)
NORA No smokers
735
4.73 (1.18)
178
4.42 (1.42)
Light
46
4.17 (1.66)
12
4.58 (1.24)
Moderate
30
4.50 (1.46)
5
4.20 (1.79)
Heavy
30
4.97 (0.85)
19
4.37 (1.61)
IR No smokers
735
13.88 (1.81)
178
13.69 (2.11)
Light
46
12.67 (3.00)*
12
13.17 (2.44)
Moderate
30
13.70 (1.49)
5
13.20 (1.48)
Heavy
30
13.43 (1.19)*
19
12.37 (3.18)*
ORA organizational religious activity, NORA non-organizational religious activity, IR intrinsic religiosity, SD standard deviation Statistically significant differences from the reference group (no smokers) are marked with an asterisk
only some evidence that men who are heavy smokers have lower mean IR values than men who do not smoke. Figure 1 shows a boxplot of the Flanagan QOLS by levels of tobacco consumption, where one observation was missing due to participant non-response. A two-way ANOVA (factors: sex and levels of tobacco consumption) followed by Bonferroni multiple comparison do not indicate any significant differences of the means of the Flanagan QOLS between sexes or smoking groups (all p values higher than 0.05, including for the interaction term). A sensitivity analysis excluding the two outliers observed in Fig. 1 was also performed, showing very similar results.
Discussion Although the prevalence of smoking in Brazil has shown an important reduction in the last decades (Ng et al. 2014; Silva et al. 2008), largely due to government policies for tobacco control (Cavalcante 2005), the monitoring and surveillance are still essential to guide public policies for health promotion and prevention of risk factors for chronic diseases, given the known hazards to health caused by cigarette smoking. In the present study, the prevalence of self-reported hypertension was 22% and the prevalence of self-reported diabetes was 8%. These values are quite near to those observed in the VIGITEL study (Schmidt et al. 2009) considering the Brazilian population (21.6 and 5.3%, respectively),
123
Flanagan Quality of Life Scale (QOLS)
J Relig Health
100
80
60
40
n Mean SD
Heavy
Moderate
Light
No smokers
49 83 8.6
35 87.4 10.8
58 88.6 9.8
912 86.5 10.5
Fig. 1 Boxplot of the Flanagan QOLS by levels of tobacco consumption
showing that these conditions remain quite prevalent in the country. The aging population has been increasing rapidly in Brazil, accompanied by a continued growth of the prevalence of non-communicable chronic diseases and increased use of health services (Veras 2012). Therefore, studies about the modifiable risk factors for chronic conditions are becoming important sources of information for the optimal allocation of resources to prevention and treatment programs. In this sense, studies involving the population who use the public health services are essentials for a better understanding of their habits and attitudes toward substance use and other health behaviors, and thus to promote healthenhancing public policies. Within this scenario, the present study showed that religiosity is an important factor in influencing the smoking behavior in Brazilian users of the public health services. The simple fact of having a religion seems to be sufficient for promoting an decrease in the prevalence of current smoking, since the prevalence of current smoking among participants who have a religion is 12% and among those who do not have a religion is 23% (Table 2). The influence of affiliation indicates that being Evangelical induces lower levels of smoking since only 7.2% of the Evangelical participants are current cigarette smokers against 14.3% of the Catholics and 21% of the Spiritists participants. Persons affiliated to Evangelical religious groups tend to have a proper lifestyle and they are more disposed to assume restrictive moral attitudes, which include disapproval of smoking and drinking. The 2010 Brazilian Demographic Census showed that 22.2% of the whole population of the country are adherents of Evangelical faiths, and in the present study it was showed that 31.5% of the participants are Evangelicals. In addition to these results, individuals who reported to practice their religion and those who claim to be very religious have the lower smoking frequency (Table 2). The frequency in which people attend religious meetings (ORA) also shown to be an important contributor to reducing smoking. Among subjects that attend religious services
123
J Relig Health
once a week or more than once a week, the smoking prevalence is lower than among those who infrequently or never attend these institutions or groups. This suggests a hypothesis that the social involvement motivated by the participation in activities linked to religion can be important to influence moral attitudes that include the non-acceptance of the consumption of substances as alcohol and tobacco. From another point of view, the social involvement can also promote a well-being state and positive feelings toward himself, and this can be sufficient to prevent the use of tobacco. The results of the logistic regression analysis do not allow evidencing an association between NORA and smoking status (Table 2), although the smoking prevalence among the people who never spend time in private religious activities is very high (23.3%) when compared with the other categories. Moreover, NORA is not associated with the intensity of smoking (Table 3). Also, the two first items of IR subscale are not associated with smoking status (Table 2), but the third item shows a strong association. There is a low smoking prevalence among people who declared they had try to carry their religion over into all other dealings in life (9.0%), and a very high smoking prevalence among the people who declared that this is not true for them (30.4%). This result reinforces that a lifestyle motivated by religious concerns and beliefs tends to strongly influence smoking habits. In this study, the smoking prevalence was 12.6% for women and 16.6% for men. These values are quite close to that obtained in the VIGITEL study (Ng et al. 2014) for the year 2012, considering the Brazilian population (11.0% for women and 16.5% for men). However, the mean number of cigarettes smoked per day (13.5 cigarettes) among the participants of the present study was less than that found in the VIGITEL study for the whole country (22.8 cigarettes) (Ng et al. 2014). Although other Brazilian studies have reported that tobacco consumers tend to earn low salaries and have less schooling than someone who does not consume tobacco (Bazotti et al. 2016; Menezes et al. 2008), the present study did not show a clear relationship between smoking, socioeconomic status and educational level (Table 1). Nonetheless, it is important to conduct anti-smoking campaigns among low-income people given the situation of vulnerability of these groups of people. Increased spending on tobacco can critically decrease spending on other critical expenditures, including basic needs (Wang et al. 2006). Among other variables not related to religiosity, it was found in the present study that smoking is more prevalent among divorced (22.4%) than married (11.6%), widowed (13.2%), or single (14.6%) respondents. The respective OR (2.42) evidences a significant association since their 95%CI do not include 1 (Table 1). This association was still significant when adjustment was made for age and sex using a logistic regression model (adjusted OR 2.19, 95% IC: 1.25–3.82). Similar results also were found in a Brazilian study involving bank employees (Griep et al. 1998) and in a large Portuguese survey of population coverage (Machado et al. 2009), but the possible role of the marital status on the smoking status is not yet clear in the literature. The well-known positive association between smoking and alcohol use also was shown in this study. A future study should address more specifically the possible associations between use of alcohol and religiosity in this population. One of the strengths of this study is the sample size and representativeness of the sample. A population-based study published in 2008 showed that 52% of the urban population of Ribeira˜o Preto had exclusive use of the public healthcare resources and 80% used these services at some time (https://www.ribeiraopreto.sp.gov.br/ssaude/pdf/fatoresrisco.pdf). Thus, the sample of 1,055 participants used in the present study can be considered representative of the municipality’s population. A limitation of this study is the cross-sectional nature of data, which does not allow establishing a true causal relationship. Important information such as previous religious affiliations and how long the respondents
123
J Relig Health
have been professing their current religion are not captured using this approach. Despite these shortcomings, the present study provides new insights into the role of the religiosity in the smoking habits of an important part of the population. Funding This work was supported by grants from Brazilian agencies CNPq (Grant Number 305942/20123) and FAPESP (Grant Number 14/14020-6). Compliance with Ethical Standards Conflict of interests None declared. Ethical Approval The study was approved by the Research Ethics Committee of Hospital das Clı´nicas and Ribeira˜o Preto Medical School (protocol 931.952).
References Agaku, I. T., King, B. A., Dube, S. R., & Centers for Disease Control and Prevention (CDC). (2014). Current cigarette smoking among adults—United States, 2005–2012. MMWR. Morbidity and Mortality Weekly Report, 63(2), 29–34. Alati, R., Kinner, S., Najman, J. M., Fowler, G., Watt, K., & Green, D. (2004). Gender differences in the relationships between alcohol, tobacco and mental health in patients attending an emergency department. Alcohol and Alcoholism, 39(5), 463–469. Alexander, A. C., Robinson, L. A., Ward, K. D., Farrell, A. S., & Ferkin, A. C. (2015). Religious beliefs against smoking among black and white urban youth. Journal of Religion and Health, 55, 1907–1916. Barros, A. J., Cascaes, A. M., Wehrmeister, F. C., Martı´nez-Mesa, J., & Menezes, A. M. B. (2011). Tobacco smoking in Brazil: Regional inequalities and prevalence according to occupational characteristics. Cieˆncia & Sau´de Coletiva, 16(9), 3707–3716. Bazotti, A., Finokiet, M., Conti, I. L., Franc¸a, M. T. A., & Waquil, P. D. (2016). Smoking and poverty in Brazil: An analysis of the profile of the smoking population based on the 2008-09 Brazilian government Family Budget Survey. Cieˆncia & Sau´de Coletiva, 21(1), 45–52. Bezerra, J., Barros, M. V. G., Teno´rio, M. C. M., Tassitano, R. M., Barros, S. S. H., & Hallal, P. C. (2009). Religiousness, alcohol consumption and smoking in adolescence. Revista Panamericana de Salud Pu´blica, 26(5), 440–446. Burckhardt, C. S., Woods, S. L., Schultz, A. A., & Ziebarth, D. M. (1989). Quality of life of adults with chronic illness: A psychometric study. Research in Nursing & Health, 12(6), 347–354. Cavalcante, T. M. (2005). Tobacco control in Brazil: Advances and challenges. Archives of Clinical Psychiatry (Sa˜o Paulo), 32(5), 283–300. Ferreira, M. P., Dini, N. P., & Ferreira, S. P. (2006). Espac¸os e dimenso˜es da pobreza nos Municı´pios do Estado de Sa˜o Paulo: I´ndice Paulista de Vulnerabilidade Social – IPVS. Sa˜o Paulo em Perspectiva, 20, 5–17. Figueiredo, V. C., Szklo, A. S., Costa, L. C., Kuschnir, M. C. C., Silva, T. L. N. D., Bloch, K. V., et al. (2016). ERICA: Smoking prevalence in Brazilian adolescents. Revista de Sau´de Pu´blica, 50(Suppl), 1. Flanagan, J. C. (1982). Measurement of quality of life: Current state of the art. Archives of Physical Medicine and Rehabilitation, 63(2), 56–59. Griep, R. H., Cho´r, D., & Camacho, L. A. (1998). Cigarette smoking among workers of a Bank. Revista de Sau´de Pu´blica, 32(6), 533–540. Gryczynski, J., & Ward, B. W. (2011). Social norms and the relationship between cigarette use and religiosity among adolescents in the United States. Health Education & Behavior, 38, 39–48. Horta, B. L., Calheiros, P., Pinheiro, R. T., Tomasi, E., & Amaral, K. C. D. (2001). Tobacco smoking among teenagers in an urban area in Southern Brazil. Revista de Sau´de Pu´blica, 35(2), 159–164. Hosmer, D. W., Jr., Lemeshow, S., & Sturdivant, R. X. (2013). Applied logistic regression. Hoboken: Wiley. Koenig, H. G., & Bu¨ssing, A. (2010). The Duke University Religion Index (DUREL): a five-item measure for use in epidemological studies. Religions, 1(1), 78–85. Linardakis, M., Smpokos, E., Papadaki, A., Komninos, I. D., Tzanakis, N., & Philalithis, A. (2013). Prevalence of multiple behavioral risk factors for chronic diseases in adults aged 50 ? , from eleven European countries—the SHARE study (2004). Preventive Medicine, 57(3), 168–172.
123
J Relig Health Machado, A., Nicolau, R., & Dias, C. M. (2009). Consumo de tabaco na populac¸a˜o portuguesa: ana´lise dos dados do Inque´rito Nacional de Sau´de 2005/2006 (Vol. 4, pp. 5–74). Lisboa: Instituto Nacional de Sau´de Doutor Ricardo Jorge - Departamento de Epidemiologia. Mak, K. K., Ho, S. Y., & Day, J. R. (2012). Smoking of parents and best friend—independent and combined effects on adolescent smoking and intention to initiate and quit smoking. Nicotine & Tobacco Research, 14(9), 1057–1064. Malta, D. C., Oliveira, T. P., Luz, M., Stopa, S. R., Silva Junior, J. B. D., & Reis, A. A. C. D. (2015). Smoking trend indicators in Brazilian capitals, 2006-2013. Cieˆncia & Sau´de Coletiva, 20(3), 631–640. Mann, J. R., McKeown, R. E., Bacon, J., Vesselinov, R., & Bush, F. (2007). Religiosity, spirituality, and tobacco use by pregnant women. Southern Medical Journal, 100(9), 867–872. Martinez, E. Z., Achcar, J. A., & Aragon, D. C. (2015). Parameter estimation of the beta-binomial distribution: An application using the SAS software. Cieˆncia e Natura, 37(3), 12–19. Martinez, E. Z., Alves, A. C., Carneiro, A. F. T. M., Jorge, T. M., Carvalho, A. C. D. D., & Zucoloto, M. L. (2014). Assessment of the psychometric properties of the Duke Religious Index scale in the context of the Public Health research. Cadernos Sau´de Coletiva, 22(4), 419–427. McVicar, D. (2011). Estimates of peer effects in adolescent smoking across twenty six European countries. Social Science and Medicine, 73(8), 1186–1193. Menezes, A., Minten, G. C., Hallal, P. C., Victora, C. G., Horta, B. L., Gigante, D. P., et al. (2008). Smoking prevalence in the 1982 birth cohort: From adolescence to adult life, Pelotas, Southern Brazil. Revista de Sau´de Pu´blica, 42, 78–85. Muza, G. M., Bettiol, H., Muccillo, G., & Barbieri, M. A. (1997). The consumption of psychoactive substances by adolescents in schools in an urban area of Southeastern region of Brazil. I-Prevalence by sex, age and kind of substance. Revista de Sau´de Pu´blica, 31(1), 21–29. Neto, M., de Souza, A., Andrade, T. M., Napoli, C., Abdon, L. C. S. D. L., Garcia, M. R., et al. (2010). Determinants of smoking experimentation and initiation among adolescent students in the city of Salvador Brazil. Jornal Brasileiro de Pneumologia, 36(6), 674–682. Ng, M., Freeman, M. K., Fleming, T. D., Robinson, M., Dwyer-Lindgren, L., Thomson, B., et al. (2014). Smoking prevalence and cigarette consumption in 187 countries, 1980–2012. JAMA, 311(2), 183–192. Nollen, N. L., Catley, D., Davies, G., Hall, M., & Ahluwalia, J. S. (2005). Religiosity, social support, and smoking cessation among urban African American smokers. Addictive Behaviors, 30(6), 1225–1229. Queiroz, N. D. R., Portella, L. F., & Abreu, A. M. M. (2015). Association between alcohol and tobacco consumption and religiosity. Acta Paulista de Enfermagem, 28(6), 546–552. Reinert, D. F., & Allen, J. P. (2002). The alcohol use disorders identification test (AUDIT): A review of recent research. Alcoholism, Clinical and Experimental Research, 26(2), 272–279. Scheaffer, R. L., Mendenhall, W., III, Ott, R. L., & Gerow, K. G. (2011). Elementary survey sampling. Boston: Cengage Learning. Schmidt, M. I., Duncan, B. B., Hoffmann, J. F., Moura, L. D., Malta, D. C., & Carvalho, R. M. S. V. D. (2009). Prevalence of diabetes and hypertension based on self-reported morbidity survey, Brazil, 2006. Revista de Sau´de Pu´blica, 43, 74–82. Silva, M. P., Silva, R. M. V. G. D., & Botelho, C. (2008). Factors associated with cigarette experimentation among adolescents. Jornal Brasileiro de Pneumologia, 34(11), 927–935. Silva, G. A., Valente, J. G., & Malta, D. C. (2011). Trends in smoking among the adult population in Brazilian capitals: A data analysis of telephone surveys from 2006 to 2009. Revista Brasileira de Epidemiologia, 14, 103–114. Veras, R. P. (2012). International experiences and trends in health care models for the elderly. Cieˆncia & Sau´de Coletiva, 17(1), 231–238. Wang, H., Sindelar, J. L., & Busch, S. H. (2006). The impact of tobacco expenditure on household consumption patterns in rural China. Social Science and Medicine, 62(6), 1414–1426. Yong, H. H., Hamann, S. L., Borland, R., Fong, G. T., & Omar, M. (2009). Adult smokers’ perception of the role of religion and religious leadership on smoking and association with quitting: A comparison between Thai Buddhists and Malaysian Muslims. Social Science and Medicine, 69(7), 1025–1031.
123
