DOES RELIGIOUSNESS EXPLAIN REGIONAL DIFFERENCES IN ALCOHOL USE IN FINLAND?

Torsten Winter,*, Sakari Karvonen,1 and Richard J. Rose,2

Department of Public Health, University of Helsinki,
1 The Unit for Promotion of Well-Being and Health, National Research and Development Centre for Welfare and Health (STAKES), Helsinki, Finland and
2 Department of Psychology, Indiana University, Bloomington, Indiana, USA

Received 17 August 2001; in revised form 14 December 2001; accepted 3 January 2002


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Aims: Because religiousness, a protective factor for alcohol use, is much more prevalent in rural regions, we examined its importance in explaining the differences in adolescent alcohol use found in the rural and the urban regions. In rural Ostrobothnia (hereafter referred to as the rural region), alcohol use is at the lowest level in all of Finland, whereas in Uusimaa, the urban region that surrounds Helsinki (hereafter referred to as the urban region), alcohol use is at the highest level. Methods: We analysed cross-sectional questionnaire data collected from Finnish adolescents and their mothers, during 1991–1995. Results: Abstinence was more prevalent, drinking less frequent, and religiousness higher in the rural region. In the urban region, there was but a negligible correlation between alcohol use and religiousness, whereas in the rural region, the correlation was clear, especially when abstainers were included. In modelling the relationship between region and adolescent abstinence, we found an interaction between mothers' religiousness and region: high religiousness among mothers was more protective of abstinence in the rural region. Conclusions: Our results show the importance of religiousness in explaining differences in adolescent alcohol use in regions with different religious traditions, but further studies will be required to explain why mothers' religiousness affects rates of adolescent abstinence differently in the two regions.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Large regional variation in alcohol use has been found in many countries, including the USA (Williams and Debakey, 1992Go), Britain (Braddon et al., 1988Go; Duncan et al., 1993Go) and Finland (Simpura and Lahti, 1988Go; Karvonen, 1997Go). Regional variation in alcohol consumption can be attributed to both structural and cultural differences between regions. Structural factors relevant to alcohol use include such characteristics as the level of urbanization of regional populations, their socioeconomic status, their level of education, and their age distribution (Blaxter, 1990Go; Kelleher et al., 1992Go). Cultural factors, such as attitudes and norms related to behaviour, religious beliefs, and political and other ideologies may also regulate alcohol use differently in different regions (Simpura and Partanen, 1987Go).

Earlier studies on regional differences in alcohol use in Finland focused on structural factors, while largely ignoring cultural factors. Among the latter, religiosity is especially important, because the degree of religiosity varies across regions. Until recently, religiosity was given scant attention in explaining adolescent use of alcohol (Manson and Windle, 2001Go). However, religious adolescents use less alcohol than do others in Finland (National Board of Health, 1990Go), a finding consistent with evidence from other protestant countries (Francis and Mullen, 1993Go). A study of the religious effect on alcohol use in different regions is important, because the sociocultural system may affect the relationship between personal religiousness and the use of alcohol (Cochran and Akers, 1989Go).

In the present study, we analyse the extent to which regional variation in adolescent alcohol use can be attributed to regional differences in religiousness. Our study is focused on two regions in Finland with differing religious traditions and different patterns of alcohol use. We compare alcohol use among adolescents living in a rural and religious region (the province of Ostrobothnia) with those living in the most urbanized and secularized region of the country (the province of Uusimaa).

Finland is a homogeneous country in relation to religious matters. In 1995, 85.6% of Finns belonged to the Lutheran church, ~1% was Orthodox, and ~2% belonged to other religious denominations (Heino et al., 1997Go), with the remainder being non-affiliated. There are five main religious revival movements in Finland. The movements are regionally separated, but the rural region (Ostrobothnia) is a regional point of intersection for three of the five movements. The religious revival movements have had effects on alcohol consumption, although the magnitude of their effect may have been exaggerated. Most of these religious movements have, however, been regarded as rather conservative (Huotari, 1981Go).

Most Finns who are not affiliated as members of the Lutheran Church or other religious organizations live in the urban region (Uusimaa) or in Häme, another southern province (Heikkilä et al., 1993Go). The contrary is true in the rural region (Ostrobothnia), the region with the smallest number of Finns who do not belong to any formal religious organization. During a recent 15-year period (1980–1995), resignation from the Lutheran Church in the Helsinki diocese was twice as common as in the dioceses (Oulu and Lapua) of the rural region (Ostrobothnia) (Heino et al., 1989Go, 1993Go, 1997Go).

A recent study that compared adolescents from the two regions found differences in their religiousness; young people in the urban region less frequently attended Church services, prayed less frequently, and expressed less belief in God than did youth from the rural region. And, if urban adolescents profess belief, they do not believe according to the teaching of the Lutheran Church (Helve, 1997Go). Further, in a series of studies of young people's values, those in the capital area have been more individualistic and disappointed with religious and political affiliation than those from the rural region, who appear to have more traditional, family-oriented values. Still, most young people are not interested in religion or politics, and they usually mix different belief systems (Helve, 1994Go, 1995Go, 1998Go).

These findings also suggest that the social context of adolescent religiousness differs between the two regions. Given the high levels of individualism and estrangement from society that characterize adolescents from the urban region, it can be hypothesized that in their personal life, as well, these adolescents experience more fragile social relationships. In extreme cases, this could lead to a state of severe alienation and lack of trust from other people and may enhance the relationship between religiousness and alcohol use. The two regions also have structural differences. New jobs in industry and service are concentrated in the southern area of the country, especially in the Helsinki urban region. In an earlier study, differences in alcohol use between the rural and urban regions could not be explained solely by structural factors (Karvonen, 1995Go). But that study did not take into account whether religiousness contributed to the regional variation in drinking among young people. Further, the study included only young people's drinking frequency, and other aspects of alcohol use were not analysed. Here, we compare the rural, religious and ‘dry’ rural region to the secular and ‘wet’ urban region and assess whether religiousness can explain why adolescents drink less in the rural region than in the rest of the country. As indicators of use of alcohol, we employed abstinence, weekly use of alcohol, and intoxication. Because social alienation has an important effect on the use of alcohol, and because there are structural differences (socioeconomic and cultural) between the regions that may affect social alienation, we also studied their effects on adolescents' abstinence.


    SUBJECTS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Subjects and general procedures
Data were collected during 1991–1995, in a population-based twin study (‘FinnTwin16’) of five consecutive nationwide birth cohorts; the baseline assessment included postal questionnaires sent to the twins and their parents. All twins received their questionnaires within 60 days of their 16th birthday. The Swedish-speaking minority was sent questionnaires in its own language. The response rates were 92% for girls, 87% for boys, and 86% for mothers. The mothers were, on average, 44 years old. In 81% of the families, the twins' parents lived together. Here, we report a secondary analysis of data collected in FinnTwin16, details of which are presented elsewhere (Rose et al., 1999Go).

Baseline questionnaires were returned by 2262 individual twins who resided in the urban (Uusimaa) or rural (Ostrobothnia) region. Of data from the parents, we chose to use only those from mothers, because divorced fathers lived infrequently with their children. We excluded families for which we had incomplete data from twins or from their mothers, and we excluded adolescents who were not living with their mothers (n = 80), those not able to answer the questionnaire by themselves (n = 3), and those who omitted more than five REL (Wiggins' Religious Fundamentalism) items (n = 52) or more than four Social Alienation items (n = 18) or who had omitted alcohol frequency or alcohol intoxication items (n = 23). Since there was some overlap among the missing answers, there remained 2104 adolescents with complete answers in the two regions. Incomplete data from mothers consisted of those who did not answer necessary items (n = 59), or whose socioeconomic status was missing (n = 27), or who omitted more than five REL (n = 18) or the use of alcohol item (n = 9). Excluding families in which data were missing from the mother, we had data from 989 families, containing 1904 adolescents and 989 mothers.

Measure of religiousness
Religiousness was assessed with the REL scale (Wiggins, 1966Go). The REL scale consists of 12 dichotomous Minnesota Multiphasic Personality Inventory (MMPI) items; the scale was formed by content analysis of the entire MMPI item pool (Dahlstrom et al., 1972Go). The REL Scale includes items of a fundamentalist character, such as ‘Everything is turning out just like the prophets of the Bible said it would’ and ‘I believe in the second coming of Christ'. In Finnish culture, such items can be regarded as measuring a traditional religiousness (Huotari, 1981Go). However, the REL scale also contains items (such as ‘I believe there is a God’ and ‘I believe in a life hereafter’) that measure general Christian beliefs.

The REL scale can be regarded as reasonably consistent: Cronbach alphas were 0.82 for girls, 0.81 for boys, and 0.82 for mothers. Only one of the items (‘I have no patience with people who believe there is only one true religion’) correlated <0.20 with the rest of the scale. Although a correlation of 0.20 has been considered as a threshold for retaining items in a scale (Streiner and Norman, 1989Go), that item was not dropped; we wanted to use the complete Wiggins' REL scale, in order to retain comparability with other studies. A test–retest reliability of 0.95 for males and 0.93 for females has been reported (Wiggins, 1966Go).

Alcohol use measures
Alcohol use among the adolescents was studied with structured questions that derive from the Finnish Adolescent Health and Lifestyle Survey (AHLS; Lintonen et al., 2000Go). One question was: ‘How often do you use alcohol? Try to include also those times you consumed only small amounts of alcohol’. The nine response alternatives were ordered from ‘Daily’ to ‘I do not use alcohol’. Those giving the response, ‘I do not use alcohol’ (21% of all respondents), were classified as abstainers. Respondents who endorsed any of the remaining eight alternatives were classified as drinkers. The time frame was lifetime alcohol use, and we regarded the answer ‘Once a year or less frequently’ as use of alcohol. The three responses, ‘Daily’, ‘A couple of times a week’ and ‘Once a week’ were collapsed and coded as weekly drinkers (14% of the drinking respondents, once abstainers were excluded).

To measure intoxication, we used the structured question: ‘How often do you use alcohol until you are really drunk?’. The four response alternatives were ‘Once a week or more often’, ‘Once or twice a month’, ‘Less frequently’ and ‘Never’. The first two alternatives were considered as regularly drinking to the point of intoxication (18% of the respondents after abstainers were excluded). In correlational analysis, all four intoxication alternatives were used.

Causal factors underlying the initiation of drinking (the dichotomy of abstaining/drinking) differ from those underlying the frequency of drinking, once initiated (Maes et al., 1999Go; Rose et al., 1999Go). In the initiation of drinking, shared environmental factors play a major role, whereas in frequency of drinking, once initiated, genetic variation assumes some importance. Among relevant environmental factors, religiosity has been shown to moderate genetic influences. In a Dutch study, genetic influences on alcohol initiation accounted for 39% of the variance among girls without a religious upbringing, compared to 0% for girls with a religious upbringing (Koopmans et al., 1999Go). For such reasons, we performed analyses in which abstinent adolescents were included and others from which they were excluded.

Psychosocial and socioeconomic measures
Not only religiousness but other psychosocial characteristics, such as alienation from social relationships, have been associated with alcohol use (Albas et al., 1978Go), especially to adverse consequences of alcohol use (Thomas and Hsieu, 1993Go). Although adolescent alienation (the estrangement of individuals from themselves and others) is a major problem, it has rarely been studied, perhaps because the meaning of the term has many definitions (Calabrese, 1987Go). As there were empirical reasons to believe that alienation may moderate the relationship of religiousness to alcohol use, we included it in our analysis. It was also important to include social alienation in the analysis as a control variable, as the adolescents in our study were twins, and their companionship with one another could have different effects on alienation in urban and rural regions.

To assess alienation, we used the Harris–Lingoes social alienation scale (Lingoes, 1960Go). Elsewhere (Winter, 1998Go), it has been shown that the scale has good psychometric properties, and there are good reasons to describe it as an alienation scale. The 9-item scale measures psychological feelings of social alienation. The items include: ‘I am sure I get a raw deal from life’ and ‘No one seems to understand me’. The 9-item social alienation scale combines two dimensions, which have been named ‘self-esteem and loneliness’ and ‘cynicism and suspiciousness’ (Winter, 1998Go).

To study how different degrees of social alienation and religiousness affected the use of alcohol in logistic regression, we divided the social alienation scale and the REL scale into quintiles. Those who were least religious were used as a reference category. We chose to divide subjects into quintiles to obtain a descriptive understanding of the association between the use of alcohol in the different social alienation and REL groups.

In Finland, weekly use of alcohol among 16–18-year olds is related to their parents' social status: children of farmers and workers drink less than do those from white-collar employees' families (Karvonen, 1995Go). As the earlier study showed that some of the regional difference in alcohol use between the two regions can be explained by socioeconomic variations, these differences were taken into account. Accordingly, we used twins' mothers' socioeconomic status (when their parents are not living together, Finnish adolescents usually reside with their mother) as an index of their social background, as most of the 16-year-olds were in school (all but 2% were attending school at the baseline assessment) and living at home. A second reason to choose mothers' socioeconomic status was the high percentage of women working full-time in Finland and, by international standards, a low percentage of part-time working women (Kivelä and Lahelma, 2000Go). The classification was based on the question ‘What is your occupation, or if you are not at work, your former occupation?’ Information on present employment (employee vs self-employed) and education was also used for the measure of mothers' socioeconomic status (Statistics Finland, 1989Go). There were 17 mothers classified into ‘other’ socioeconomic status. These (housewives, students and socioeconomic status unknown) were included in the self-employed group. The classification of family structure was based on adolescents' reports. During the time the survey was made, Finland was in a deep recession with high unemployment. Because parents' unemployment may effect their children's use of alcohol, we included mothers' unemployment among background variables. Mothers' employment status was classified as employed, unemployed and on pension or other employment status. The adolescents' own socioeconomic status, defined by education, was not informative, because nearly half the sample of adolescents was still in compulsory school.

Regional classification
Finland was divided regionally according to geographic divisions made by the Statistical Office of the European Union (EU) (Statistics Finland, 1998aGo). We used level 3 of the EU system, the traditional regional division. In the rural region, we included Coastal Vaasa, Central Ostrobothnia and North Ostrobothnia. The urban region consists of Uusimaa proper and East Uusimaa. Swedish-speaking Finns make up ~10% of the population in both the rural and urban regions (Statistics Finland, 1996Go) and are known to differ in some of their behaviour patterns (Kannas et al., 2000Go). The response rates were identical for the two language groups.

Table 1Go summarizes the sociodemographic characteristics of the participating adolescents and their mothers, by region. There were more white-collar employees (upper-level and lower-level employees) and fewer manual workers and farmers among the mothers in the urban, than in the rural, region. This socioeconomic difference corresponds to the different educational levels in the two regions. In the urban region, 17% of mothers aged 35–54 years have an academic degree, compared to only 8% in the rural region (Statistics Finland, 1998bGo). Throughout all of Finland in the end of 1995, 18% of families with children under age 17 years were one-parent families (Statistics Finland, 1996Go). In the urban region, our sample approximates that number, but in the rural region, one-parent families were less common, although the difference was not significant. According to the classification into rural and urban municipalities, there were 49% urban municipalities in the rural region. The classification is, however, an administrative one and is not based on degree of urbanization. The unemployment rate among the mothers in this study was lower than in general during the time 1991–1995 (Statistics Finland, 1996Go). People with families are less often unemployed, which partly explains the lower figures. The unemployment figures are also lower in both regions so the small differences from official unemployment statistics should not affect the results in this study.


View this table:
[in this window]
[in a new window]
 
Table 1. Sociodemographic characteristics of mothers (n = 989) to 16-year-olds in rural and urban regions
 
Statistical procedures
Statistical analyses were performed by using Stata version 6 (Stata Corporation, 1999Go). In cross-tabulations, the statistical significance of the differences between categories of variables was tested separately for girls and boys and between the two regions. The multivariate analyses (abstinence, weekly drinking and intoxication) were done using logistic analysis regression (n = 1904 for abstinence and n = 1475 for weekly drinking and intoxication). To correct for the likelihood that observations from twin pairs are correlated, we used the complex survey data methods in Stata to estimate correct confidence intervals in all analyses.

When modelling abstinence, we used 1904 adolescents, of whom 429 were abstainers. When modelling weekly drinking and intoxication, we used only the 1475 adolescents drinking alcohol. Of these 1475 adolescents, 200 were weekly drinkers and 268 reported being intoxicated once a month. The dependent variables were dichotomous and measured abstinence among all subjects, and weekly drinking and regular intoxication only among those using alcohol.

The purpose of the sequential modelling process was to elaborate how the odds ratio for abstinence, weekly drinking and intoxication for the two regions changed comparatively when the background variables, religiousness and social alienation, were added into the model. The background variables were considered as control variables causally preceding the relationship between region and adolescent alcohol use. Own and mothers' religiousness and social alienation were regarded as intervening variables, which could explain the different use of alcohol in the two regions.

Measuring religiousness at age 16 years is appropriate in Finland, since >90% of Finns attend confirmation school at age 15 years (Heino et al., 1997Go). We included in the analyses mothers' REL scores, as well as those of their adolescents, because family environment has proved to have an essential association with adolescents' religiosity (Winter et al., 1999Go).

The modelling process was done sequentially, so that explaining adolescents' use of alcohol began with a bivariate model including only the regional variable (urban–rural). The background variables were then added into the model, followed by own religiousness and mothers' religiousness. As the last main effect, social alienation was put into the model to explore whether alienation modifies the relationship between religiousness and alcohol use. Interactions between the regional variable and the other variables (background variables, own, mothers' religiousness and social alienation) were tested on top of a model with all the variables included.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Drinking was more common and abstinence less prevalent in the urban than in the rural region (Table 2Go). In the rural region, there were almost four times as many abstainers as weekly drinkers, whereas in the urban region that difference was much smaller. There were significant differences in use of alcohol between both boys ({chi}2 = 112.06, df = 2, P < 0.001) and girls ({chi}2 = 59.42, df = 2, P < 0.001) in the two regions. When analysing only those drinking alcohol, there were no significant differences between the regions in intoxication frequency.


View this table:
[in this window]
[in a new window]
 
Table 2. Use of alcohol among adolescents (16-year-olds) in rural and urban regions (n = 1904 of whom 1475 use alcohol) with {chi}2-test
 
Adolescents' religiousness differed in two ways: boys were less religious than girls, and adolescents living in the rural region were more religious than those living in the urban region (Table 3Go). There were significant differences both between sexes and between regions. Of the 12 REL items (combined sample, boys and girls together), 11 had higher endorsement percentages in the rural region. Ten of these items were significantly (P < 0.05) higher in the rural region (data not shown here, but see also Winter et al., 1999Go).


View this table:
[in this window]
[in a new window]
 
Table 3. Mean scores on Wiggins' Religious Fundamentalism Scale (12 items) among adolescents (16-year-olds) (n = 1904) and their mothers (n = 989) in rural and urban regionsa
 
Mothers from the rural region, followed by girls from the same region, were the most religious (Table 3Go). Religiousness was lowest among adolescents residing in the urban region. Religiousness differed more among mothers residing in the two regions than among their adolescent children.

Among boys in the rural region, but not among those in the urban region, religiousness was related to use of alcohol, so that religiousness correlated significantly (negatively) with the frequency of use of alcohol (Table 4Go). Among girls, the relationship became much weaker when limited to those using alcohol.


View this table:
[in this window]
[in a new window]
 
Table 4. Correlations (Spearman) between three measures of adolescents' use of alcohol and religiousness (own and mothers') among 16-year-old boys and girls in the rural and urban regions
 
The correlation between mothers' religiousness and their children's use of alcohol was significant in the rural region for all subjects (no major differences between the sexes, data not shown). However, when abstainers were excluded from the analysis, those correlations were no longer significant.

Religiousness and abstinence have historically been interrelated in the Nordic countries (i.e. in Norway, see Pedersen and Kolstad, 2000Go). Accordingly, we studied possible multi-colinearity between own religiousness and mothers' abstinence and between mothers' religiousness and mothers' abstinence. The correlations between mothers' abstinence and the adolescents' religiousness were 0.27 (Pearson's) and 0.48 (Gamma), and those between mothers' abstinence and mothers' religiousness were 0.36 (Pearson's) and 0.61 (Gamma). The differences in the Pearson's and Gamma correlations suggest a non-linear relationship between the two variables (Hellevik, 1987Go). The high Gamma correlation between mothers' religiousness and abstinence may indicate problems of colinearity. In the presence of multi-colinearity, estimates of partial regression coefficients become less precise (Tacq, 1997Go); that is why mothers' abstinence was not included in the regression, given that the focus in this study was religiousness and adolescents' abstinence.

The next step in the analysis was the elaboration of the relationship between adolescents' abstinence and region (Tables 5 and 6GoGo). In an unadjusted model (model 1), the odds ratio for abstinence was 2.67 for the rural region compared to the urban region. When adjusting for the background variables (model 2), the odds ratio for abstinence remained nearly the same. The significant background variables were family structure (single parent compared to two parent families), language (Swedish compared with Finnish) and mothers' employment (unemployment compared with employment).


View this table:
[in this window]
[in a new window]
 
Table 5. Factors explaining abstinence in 16-year-old adolescents in two regions according to logistic regression analysis
 

View this table:
[in this window]
[in a new window]
 
Table 6. Factors explaining abstinence on 16-year-old adolescents in two regions according to logistic regression analysis
 
We next included in the model the adolescents' own religiousness and then the mothers' religiousness. The odds ratio for abstinence for the rural region compared to the urban region fell when adjusting for own REL (model 3). Of the REL quintiles, only the highest fifth compared to the lowest fifth was significant. When mothers' REL was added to this model, the odds ratio for abstinence for the rural region compared to the urban region fell still further, though it remained significant (model 4). In this model, the highest quintiles of both own and mothers' REL were significant, mothers' odds ratio being the larger of the two. The last main variable, social alienation, had only a very limited effect on regional odds ratio explaining adolescent abstinence (model 5), although the two highest social alienation quintiles had a significant effect on adolescents' abstinence.

When all the main variables were included in the model (model 5), the odds ratio for abstinence for the rural region was about twice as high as in the urban region. The next step in the modelling process was to study possible interactions between region and the other variables.

We found a significant two-way interaction between region and mother's religiousness. According to the interaction, the estimates for the two regions in the highest quintile of mothers' REL scores differed significantly from the lowest (the reference category), suggesting that mothers' religiousness has a greater effect on their children's abstinence in the rural region. The effect can be illustrated by cross-tabulating mothers' REL with adolescents' abstinence in the two regions. In both regions, 14% of the adolescents whose mothers are in the lowest REL quintile reported abstaining, but when the mother belonged to the highest quintile, 24% of adolescents in the urban region and 53% in the rural region reported abstaining. The different relationship between mothers' REL (categorized in quintiles) and adolescents' abstinence in the two regions could also be seen in gamma correlations, which were 0.14 in the urban region and 0.47 in the rural region.

We then studied the odds ratio for weekly drinking and intoxication in two separate logistic regression models (n = 1475) with the same main variables as in the final model for abstinence. The significant variables for weekly drinking were sex (odds ratio 0.67, girls drinking less often), family structure (odds ratio for family with mother but not father 1.56), own religiousness (odds ratio for the highest fifth 0.37) and social alienation (odds ratio the fourth fifth 2.05 and the highest fifth 2.54). The significant variables for intoxication were own religiousness (odds ratio for the highest fifth 0.53) and social alienation (odds ratio the fourth fifth 1.96 and the highest fifth 3.03). When using the same modelling process as before (abstinence), region was not significant in either model in any stage (Results not shown).


    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
The aim of our study was to assess whether cultural factors, especially religiousness, contribute to regional variation in alcohol use. The study suggests that religiousness explains some of the difference observed in adolescent drinking behaviour between rural and urban regions of Finland. After adjusting for an adolescent's own religiousness and that of the adolescent's mother, the regional difference in abstinence decreased, though it remained significant. In analyses of weekly drinking and intoxication, analyses from which abstainers were excluded, no significant differences between the two regions were found.

These results are consistent with previous research: abstinent people are more religious than others (Fillmore et al., 1998Go) and family environment has important effects on abstinence among adolescents (Heath et al., 1999Go). After adolescents start to drink alcohol, influences of family environment decrease, and genetic and other extra-familial environmental factors assume greater influence on individual differences in drinking frequency (Maes et al., 1999Go). Similarly, in a study among rural and small-town adolescents, attendance at religious services was related to age at first drink, but not to frequency or amount of alcohol use, once initiated (Gibbons et al., 1986Go; see also D'Onofrio et al., 1999Go).

One important characteristic of the family environment is family structure. In our data, family structure had an effect on onset of adolescent drinking, but not on weekly drinking. This result differs from an earlier Finnish study (Karvonen, 1995Go), where family structure had a significant effect on weekly drinking among 16- and 18-year-olds; but in that study, both age groups were considered together, and family structure was classified into three groups, and these differences in analysis may explain the differences found. Family environment has a crucial influence on abstinence among adolescents, but the effect varies by region. In Northern Finland, 16-year-olds were 2.7 times more likely to be abstinent if their parents were abstinent, but within the Helsinki area, only 1.4 times more likely than those of drinking parents (Rose et al., 1999Go).

The correlations between religiousness and use of alcohol indicated possible interactions, which could be confirmed in the logistic regression analyses between mothers' religiousness and region. This interaction showed that the mothers' religiousness has a rather small effect in the urban region, whereas the correlation between mothers' religiousness and adolescents' abstinence was rather stronger in the rural region. But the interaction could be seen only in the highest religiousness quintile. The interaction shows at the same time the importance of both region and religiousness. Religiousness alone does not explain the differences between the regions in abstinence among the adolescents; it requires a particular environment (the rural region) to have an effect. This conditional relationship between mothers' religiousness and her children's abstinence could not, however, be found between the adolescent's own religiousness and abstinence, since it was not dependent on the region.

These results can be interpreted in the context of the ‘hellfire hypothesis' studies (for a summary see Cochran and Akers, 1989Go). According to the hellfire hypothesis, religion ensured conformity and deterred deviance through the threat of hellfire for sinners. This relationship may, however, depend on the social context, religious affiliation and offence type (Cochran, 1991Go). Our results among mothers support findings of an interaction between individual religiosity and community context, as mothers' religiousness had an effect on their children's abstinence in a more religious region (the rural region), but not in a more secular region (the urban region). On the other hand, we found no interaction between the adolescent's own religiousness and region. This may be attributed to mothers being more integrated to the local society and more influenced by its religious traditions, than are their children (Welch et al., 1991Go).

The validity of our findings can be assessed externally by comparing our study with another covering representatively the whole country, the AHLS study (Lintonen et al., 2000Go). Both studies used identical questions to measure alcohol use, yet the abstinence rate in our study is clearly higher (23%) than those from the AHLS in which ~12–15% of 16-year-olds were abstaining in comparable study years (Rimpelä et al., 1999Go). Our higher rates can probably be explained by the subjects of the FinnTwin16 study being almost 6 months younger than those of the AHLS. As abstinence decreases extremely rapidly at this age, the fact that the AHLS respondents received their questionnaires 6 months after their 16th birthday (Lintonen et al., 2000Go), whereas the FinnTwin16 respondents within a month of the birthday, most likely accounts for the difference.

The adolescents in the present study were twins. That does not constrain inferences, because twin studies regarding alcohol use are comparable with those of non-twins, and other results from FinnTwin16 have been compared to results from other Finnish studies, for menarche (Kaprio et al., 1995Go), intensity of physical activity (Aarnio et al., 1997Go) and body mass index (Pietiläinen et al., 1999Go). These comparisons show negligible differences between twins and adolescents in general.

On an individual level, abstinence is related to religiousness, but there may also be indirect social links between religiousness and abstinence. Municipalities in Finland have the right to decide about retail sales of alcohol, and, during the 1970s, some municipalities in religiously influenced areas declined to give permission for retail sales outlets. During the 1980s, attitudes toward alcohol sales became more liberal, but in the early 1990s when these data were collected, there remained municipalities in the rural region where retail sales of medium-strength beer were not allowed (Sairanen and Sulkunen, 1998Go; Valli, 1998Go). When such sales were allowed in one of these municipalities (Pietarsaari) in 1993, an increase in medium-strength beer consumption among adolescents (aged 13–17 years) followed, but at the same time, wine consumption decreased, so that the total consumption of alcohol did not change. Thus, easy access to alcohol may not raise alcohol consumption in the same way among adolescents as among adults (Valli, 1998Go).

That we find Swedish-speaking adolescents to be less abstinent and to drink more frequently than those who are Finnish-speaking is not new (Kannas et al., 2000Go). The different drinking habits have been attributed to different cultural factors. Use of alcohol has another role in Swedish-speaking communities, where it is more common in family celebrations than within the Finnish-speaking culture. Social alienation (the Harris–Lingoes scale) correlated significantly with use of alcohol in both the rural and urban regions. Social alienation also proved to be a significant risk factor in all three logistic regression analyses explaining different kinds of use of alcohol. Social alienation did not, however, explain the differences between the regions, because it was equally important in both regions. According to Helve (1998)Go, young people in the urban region are more individualistic and frustrated politically than young people from the rural region, where family values are stronger. Because alienation is expressed by feelings of estrangements, one could have expected differences in social alienation between the two regions affecting the abstinence rate between the two regions under study.

In our data, a mother's socioeconomic status had no significant effect on her adolescents' alcohol-drinking behaviour. These results differ from earlier studies (Rahkonen and Ahlström, 1989Go; Karvonen, 1995Go) where adolescents from (upper) white-collar families drank most and farmers' children least. Compared with these studies, however, three differences in the definition of the socioeconomic situation for the family and the adolescents characterize our study. We used mothers' employment status to control the possible effect of unemployment, whereas the two earlier studies did not. Using mothers' socioeconomic status, rather than status of fathers, may affect the results. The third reason for the different results may be the age of the adolescents. As the adolescents were 6 months younger and many still in compulsory school, we did not use their own educational level as a measure of socioeconomic status as did Karvonen (1995)Go.

Our study, in common with all studies, has both strengths and limitations. A particular strength is the sample, which is population-based and is characterized by very high response rates. Because the adolescent participants answered within 2–3 months of their 16th birthdays, age restriction was tight. A methodological limitation of our study may be in our measure of religiousness. Religiousness is a multi-dimensional construct, and different dimensions of religiousness (i.e. spirituality, beliefs, practices, religious affiliation and attendance) may affect the observed relationship between religiousness and use of alcohol (D'Onofrio et al., 1999Go). Our measure was based on Wiggins' REL scale, which is a true– false scale, measuring especially traditional religiousness (also religious fundamentalism). Our results should therefore be interpreted as indicative of this kind of religiousness.

In summary, this study documents the importance of religiousness in explaining differences in adolescent alcohol use in different regions characterized by different religious traditions. The association of religiousness with alcohol use is most evident in the predictive association of religiousness with abstinence, and this relationship varies and is modified according to different cultural and social contexts. Once adolescents initiate drinking, individual differences in frequency, quantity and density of their use appear to reflect the influence of other factors.


    ACKNOWLEDGEMENTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
FinnTwin16 is supported by the National Institute of Alcohol Abuse and Alcoholism (AA 08315) and from the Academy of Finland (grant nos. 42044 and 44069); preparation of this paper was supported by the Research Institute of the Evangelical Lutheran Church of Finland, the Finnish Foundation for Alcohol Studies, and the Indiana Alcohol Research Center (AA 07611).


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
* Author to whom correspondence should be addressed at: Department of Public Health, University of Helsinki, P. O. Box 41 (Mannerheimintie 172), 00014 Helsinki, Finland. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Aarnio, M., Kujala, U. M. and Kaprio, J. (1997) Associations of health-related behaviors, school type and health status to physical activity patterns in 16-year old boys and girls. Scandinavian Journal of Social Medicine 25, 156–167.[ISI][Medline]

Albas, D., Albas, C. and McCluskey, K. (1978) Anomie, social class and drinking behavior. Journal of Studies on Alcohol 39, 910–913.[ISI][Medline]

Blaxter, M. (1990) Health and Lifestyles. Routledge, London.

Braddon, F., Wadsworth, M., Davies, J. and Cripps, H. A. (1988) Social and regional differences in food and alcohol consumption and their measurement in a national birth cohort. Journal of Epidemiology and Community Health 42, 341–349.[Abstract]

Calabrese, R. L. (1987) Adolescence: a growth period conducive to alienation. Adolescence 22, 927–938.

Cochran, J. K. (1991) The effects of religiosity on adolescent self-reported frequency of drug and alcohol use. Journal of Drug Issues 22, 91–104.[ISI]

Cochran, J. K. and Akers, R. L. (1989) Beyond hellfire: an exploration of the variable effects of religiosity on adolescent marijuana and alcohol use. Journal of Research in Crime and Delinquency 26, 198–225.[ISI]

Dahlstrom, W. G., Welsh, G. S. and Dahlstrom, L. E. (1972) MMPI Handbook, Vol. I, Clinical Interpretation. University of Minnesota Press, Minneapolis.

D'Onofrio, B. M., Murrelle, L., Eaves, L. J., McCullough, M. E., Landis, J. L. and Maes, H. H. (1999) Adolescent religiousness and its influence on substance use: preliminary findings from the Mid-Atlantic School Age Twin Study. Twin Research 2, 156–168.[Medline]

Duncan, C., Jones, K. and Moon, G. (1993) Do places matter? A multi-level analysis of regional variations in health-related behaviour in Britain. Social Science and Medicine 37, 725–733.[ISI][Medline]

Fillmore, K. M., Golding, J. M., Graves, K. L., Kniep, S., Leino, E. V., Romelsjö, A., Shoemaker, C., Ager, C. R., Allebeck, P. and Ferrer, H. P. (1998) Alcohol consumption and mortality. Characteristics of drinking groups. Addiction 93, 183–204.[ISI][Medline]

Francis L. J. and Mullen, K. (1993) Religiosity and attitudes towards drug use among 13–15 year olds in England. Addiction 88, 665–672.[ISI][Medline]

Gibbons, S., Wylie, M. L., Echterling, L. and French, J. (1986) Patterns of alcohol use among rural and small-town adolescents. Adolescence 21, 887–900.[ISI][Medline]

Heath, A. C., Madden, P. A. F., Grant, J. D., McLaughlin, T. L., Todorov, A. A. and Bucholz, K. K. (1999) Resiliency factors protecting against teenage alcohol use and smoking: influences of religion, religious involvement and values, and ethnicity in the Missouri Adolescent Female Twin Study. Twin Research 2, 145–155.[Medline]

Heikkilä, M., Heino, H., Holm, N. S., Korelin, A., Rusama, J. and Sidoroff, M. (1993) Uskonnolliset yhteisöt (Religious communions). In Suomen kartasto (Atlas of Finland) Vol. 3. Yhteiskunta. Kulttuuri. Elinolot (Society. Culture. Living conditions). Ursin, H. ed., pp. 17–27. Karttakeskus, Helsinki, Finland.

Heino, H., Kauppinen, J. and Ahonen, R. (1989) Suomen evankelis-luterilainen kirkko vuosina 1984–1987 (The Finnish Lutheran Church), The Research Institute of the Evangelical Lutheran Church of Finland. Gummerus, Jyväskylä.

Heino, H., Kauppinen, J. and Ahonen, R. (1993) Suomen evankelis-luterilainen kirkko vuosina 1988–1991 (The Finnish Lutheran Church), The Research Institute of the Evangelical Lutheran Church of Finland. Gummerus, Jyväskylä.

Heino, H., Salonen, K., Rusama, J. and Ahonen, R. (1997) Suomen evankelis-luterilainen kirkko vuosina 1992–1995 (The Finnish Lutheran Church), The Research Institute of the Evangelical Lutheran Church of Finland. Gummerus, Jyväskylä.

Hellevik, O. (1987) Forskningsmetoder i sociologi och statsvetenskap (Research methods in Sociology and Politics), pp. 212–213. Natur och Kultur, Lund.

Helve, H. (1994) The role of religion in postmodern society: a longitudinal study of the values of contemporary young Finns. Temenos 30, 81–93.[ISI]

Helve, H. (1995) Religion, world views and gender ideology. Temenos 31, 77–92.[ISI]

Helve, H. (1997) Nuorten muuttuvat arvot ja maailmankuvat — empiirinen tutkimusnuorten maailmankuvista (The Changing World View of Young People — an Empirical study). In Arvot, maailmankuvat, sukupuoli (Values, World Views, Gender), Helve, H. ed., pp. 140–176. University Press, Helsinki.

Helve, H. (1998) What happened to young ‘Humanists', ‘Individualists' and ‘Traditionalists'? A comparative study of changing value-worlds of young people in the framework of postmodernity. Council of Europe — European youth centres youth research and documentation. Research papers. (http://www.coe.fr/youth/research/helena.htm, 26 January 1999.)

Huotari, V. (1981) Kirkkomme herätysliikkeet tänään (The revivalist movements in our Church). Kirjapaja, Pieksämäki.

Kannas, L., Tynjälä, J. and Brunell, V. (2000) Tobak, alkohol och narkotika (tobacco, alcohol and drugs). In Subjektiv hälsa, hälsovanor och skoltrivsel (Subjektiv Health, Health Habits and School Comfort), Kannas, L. and Brunell, V. eds, pp. 94–220. Terveystieteiden laitoksen julkaisusarja 10 (Health Sciences), Jyväskylä.

Kaprio, J., Rimpelä, A., Winter, T., Viken R. J., Rimpelä, M. and Rose, R. J. (1995) Common genetic influences on BMI and age of menarche. Human Biology 67, 739–753.[ISI][Medline]

Karvonen, S. (1995) Regional differences in drinking among Finnish adolescents. Addiction 90, 57–64.[ISI][Medline]

Karvonen, S. (1997) The regional context of health behaviour among Finnish adolescents. STAKES (National Research and Development Centre for Welfare and Health). Research Reports 76, Gummerus, Jyväskylä.

Kelleher, K. J., Richert, V. I., Hardin, B. H., Pope, S. K. and Farmer, F. I. (1992) Rurality and gender. Effects on early adolescent alcohol use. American Journal of Diseases of Children 146, 317–322.[Abstract]

Kivelä, K. and Lahelma, L. (2000) Ansiotyön ja perheen yhdistäminen: kaksinkertainen etu vai taakka naisten hyvinvoinnille ja terveydelle? (Combining work and family: double burden or double buffert for women's health and well-being?). Sosiaalilääketieteellinen aikauslehti (Journal of Social Medicine) 37, 40–52.

Koopmans, J. R., Slutske, W. S., van Baal, G. C. and Boomsma, D. I. (1999) The influence of religion on alcohol use initiation: evidence for genotype x environment interaction. Behavior Genetics 29, 445–453.[ISI][Medline]

Lingoes, J. (1960) MMPI factors of the Harris and Wiener subscales. Journal of Consulting Psychology 24, 74–83.[ISI]

Lintonen, T., Rimpelä, M., Ahlström, S., Rimpelä, A. and Vikat, A. (2000) Trends in drinking habits among Finnish adolescents from 1977 to 1999. Addiction 95, 1255–1263.[ISI][Medline]

Maes, H. H., Woodard, C. E., Murrelle, L., Meyer, J. M., Silberg, J. L., Hewitt, J. K., Rutter, M., Simonoff, E., Pickles, A., Carbonneau, R., Neale, M. C. and Eaves, L. J. (1999) Tobacco, alcohol and drug use in eight- to sixteen-year-old twins: the Virginia Twin Study of adolescent behavioral development. Journal of Studies on Alcohol 60, 293–305.[ISI][Medline]

Manson, W. A. and Windle, M. (2001) Family, religious, school and peer influences on adolescent alcohol use: a longitudinal study. Journal of Studies on Alcohol 62, 44–53.[ISI][Medline]

National Board of Health (1990) Uskonto ja terveys (Religiosity and health). Series Statistics and Reviews, 3/1990. National Board of Health, Helsinki.

Pedersen, W. and Kolstad, A. (2000) Adolescent alcohol abstainers: traditional patterns in new groups. Acta Sociologica 43, 219–233.[ISI]

Pietiläinen, K. H., Kaprio, J., Rissanen, A., Winter, T., Rimpelä, A., Viken, R. J., and Rose, R. J. (1999) Distribution and heritability of BMI in Finnish adolescents aged 16y and 17y: a study of 4884 twins and 2509 singletons. International Journal of Obesity 23, 107–115.[ISI][Medline]

Rahkonen, O. and Ahlström, S. (1989) Trends in drinking habits among Finnish youth from 1973 to 1987. British Journal of Addiction 84, 1075–1083.[ISI][Medline]

Rao, J. N. K. and Scott, A. J. (1984) On chi-squared test for multiway contingency tables with cell proportions estimated from survey data. Annals of Statistics 12, 46–60.[ISI]

Rimpelä, A., Vikat, A., Rimpelä, M., Lintonen, T., Ahlström, S. and Huhtala, H. (1999) Nuorten terveystapatutkimus 1999 — tupakoinnin ja päihteiden käytön muutokset (The Juvenile Health Habit Study 1999 — changes in smoking and use of drugs). Stakes aiheita 18/1999, Stakesin monistamo, Helsinki.

Rose, R. J., Kaprio, J., Winter, T., Koskenvuo, M. and Viken, R. J. (1999) Familial and socio-regional environmental effects on abstinence from alcohol at age sixteen. Journal of Studies on Alcohol 13, 63–74.

Sairanen, S. and Sulkunen, P. (1998) Från centralstyrd till kommunal alkoholpolitik? Kommunerna, alkoholpolitiken och decentraliseringen (From centralized to local alcohol policy? Administrative decentralization and the role of local municipalities in Finnish alcohol policy). Nordisk alkohol-& narkotikaskrift(Nordic studies on alcohol and drugs) 15(2), 75–92.

Simpura, J. and Lahti, M.-L. (1988) Juomatapojen alueellinen vaihtelu Suomessa vuonna 1984 (The regional variation of drinking habits in Finland 1984). Alkoholipolitiikka 53, 114–121.

Simpura, J. and Partanen, J. (1987) Drinking habits in a changing society. In Finnish Drinking Habits — Results from Interview Surveys Held in 1968, 1976 and 1984. Simpura, J. ed., pp. 210–222. The Finnish Foundation for Alcohol Studies, Vol. 35, Jyväskylä.

Stata Corporation (1999) Intercooled Stata 6.0 for Windows. Stata corporation 702 University Drive East College Station, TX 77840, USA.

Stata Reference Manual Release 6 (1999) Vol. 4, Su–Z. Stata Press, College Station, TX.

Statistics Finland (1989) Classification of Socio-economic Groups. Statistics Finland, Helsinki.

Statistics Finland (1996) Statistical Yearbook 1996. Gummerus, Jyväskylä.

Statistics Finland (1998a) Kunnat 1998 — Alueluokitukset Käsikirjoja no. 28 (Municipalities — Regional Classification Handbook) Statistics Finland, Helsinki.

Statistics Finland (1998b) SUOMI CD (Finland CD). Visual Data, OY.

Streiner, D. L. and Norman, G. R. (1989) Health Measurement Scales. A Practical Guide to their Development and Use. Oxford University Press, Oxford.

Tacq, J. (1997) Multivariate Analysis in Social Science Research. From Problem to Analysis, pp. 128–129. Sage, London.

Thomas, B. S. and Hsieu, L. T. (1993) The role of selected risk factors in predicting adolescent drug use and its adverse consequences. International Journal of the Addictions 28, 1549–1563.[ISI][Medline]

Valli, R. (1998) Förändringar i ungdomars alkoholvanor när mellanölet släpptes fritt. Fallet Jakobstad (Changes in young people's alcohol consumption with improved availability of medium strength beer. The case of Pietarsaari). Nordisk alkohol-& narkotikaskrift(Nordic Studies on Alcohol and Drugs) 15(3), 168–175.

Welch, M. R., Tittle, C. P. and Petee, T. (1991) Religion and deviance among adult catholics: a test of the ‘moral communities’ hypothesis. Journal for the Scientific Study of Religion 30, 159–172.[ISI]

Wiggins, J. S. (1966) Substantive dimensions of selfreport in the MMPI item pool. Psychological Monographs 80 (whole no. 630).

Williams, G. D. and Debakey, S. F. (1992) Changes in levels of alcohol consumption: United States, 1983–1988. British Journal of Addiction 87, 643–648.[ISI][Medline]

Winter, T. (1998) Att mäta social alienationskänsla med Harris– Lingoes skala (Measuring social alienation with the Harris–Lingoes scale). Sosiaalilääketieteellinen aikakauslehti(Finnish Journal of Social Medicine) 4, 267–278.

Winter, T., Kaprio, J., Viken, R. J., Karvonen, S. and Rose, R. J. (1999) Individual differences in adolescent religiousness in Finland: familial effects are modified by sex and region of residence. Twin Research 2, 108–114.[Medline]





This Article
Abstract
FREE Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (2)
Request Permissions
Google Scholar
Articles by Winter, T.
Articles by Rose, R. J.
PubMed
PubMed Citation
Articles by Winter, T.
Articles by Rose, R. J.