OBJECTIVE: To provide nationwide data on health status and health behaviours among young adults in Switzerland, and to illustrate social and regional variations.
METHODS: Data came from the Swiss Federal Surveys of Adolescents, conducted in 2010/11. The sample consisted of 32,424 young men and 1,467 young women. We used logistic regression models to examine patterns of social inequality for three measures of health status and three measures of health behaviour.
RESULTS: Among men, lower self-rated health, overweight and lower physical fitness levels were associated with lower educational and fewer financial resources. Patterns were similar among young women. Unfavourable self-rated health (odds ratio [OR]: men 0.83, women 0.75) and overweight (OR: men 0.84, women 0.85; p >0.05) were less common in the French- than in the German-language region. Low physical fitness was more common in the French- than in the German-language region. In both sexes, daily smoking was associated with fewer educational resources, and physical inactivity was associated with lower educational and fewer financial resources. Males from the Italian-language region were three times more likely to be physically inactive than their German-speaking counterparts (OR 2.95). Risk drinking was more widespread among males in the French- than in the German-speaking language region (OR 1.47).
CONCLUSIONS: Striking social and moderate regional differences exist in health status and health behaviours among young Swiss males and females. The current findings offer new empirical evidence on social determinants of health in Switzerland and suggest education, material resources and regional conditions to be addressed in public health practice and in more focused future research.
Summary and discussion
We found striking differences in the social distributions of health and health behaviour among young Swiss adults. For both genders, the likelihood of daily smoking was strongly associated with lower educational resources; physical inactivity was significantly more common in those with lower educational attainment and with less family financial resources. Many of the social inequality effects conformed to a social gradient pattern.
The largest regional variation in health behaviours was in physical activity. Men from the Italian-language region were almost three times as likely to be physically inactive than their German-speaking counterparts. In men, risk drinking was more widespread in the French-language region of Switzerland and daily smoking was more prevalent in the Italian-language region. These regional differences in physical activity, risk drinking and daily smoking are consistent with, and corroborate, previous findings [28–32]. A more recent study, however, did not find evidence of regional differences in risk drinking between the German- and the French-language regions [33]. In summarising our findings on health behaviours it should be noted that the current data on the regional distribution of risk behaviours are basically of a descriptive nature. Reasons for such variations may be found in differences in the infrastructure conditions (particularly for physical activity) or in culture-based factors such as values and norms. More focused studies are needed to explain those variations. Our descriptive findings can provide starting points for such future explanatory studies.
As for health status, among young Swiss men, lower self-rated health, overweight and less physical fitness were all associated with lower educational and fewer financial resources. The pattern was similar for young women, although social determinant effects in our female sample were weaker. The gradient effect in overweight we found in our data is consistent with previous findings [10, 31, 34]. As for the prevalence rates, our study updates older national data on overweight in young Swiss adults. Earlier findings showed an increase in the prevalence of overweight among young Swiss men between 1993 (10.9%) and 2003 (14.8%) [8], and our current results show a continuation of this trend, with the prevalence rising again between 2003 and 2011 to 22.4%. The same data showed that, during the same time period, overweight in young Swiss women was less prevalent and rather stable. Differences in health status by language region showed that unfavourable self-rated health in both genders and overweight in men were less common in the French-language region than in the German-language region. Young men from the French-language region most often reported low physical fitness, followed by men from the German-language region, and then men from the Italian-language region. Unfavourable self-rated health was associated with urban residence, but the effect size was small. We found no systematic variation associated with urban vs. rural residence in the other health status measures. There may be no statistically significant associations for the latter, or our measures of urban/rural residence might not be specific enough to detect such variations in health status.
Social gradient effects are most evident in the link between own education and smoking among males, and own education and physical inactivity among females. Parents’ education also demonstrates gradient effects on overweight and physical inactivity in both genders. Our results on alcohol use are in line with those of earlier studies that showed systematic associations between health behaviour and social class; behaviours tend to be worst at the bottom rungs of the class ladder [14, 32, 35, 36]. In our sample of young males, smoking was more frequent among those in vocational training, a result which is in line with previous findings [32]. Also, risk drinking was most frequent among those with less education. This finding corroborates that of other Swiss studies, which found problem drinking was most prevalent in young adults in vocational training [33, 37]. However, we also found that risk drinking was less frequent among young men whose parents had only mandatory education. This points towards an inverse social gradient for parental education. A similar inverse social gradient was reported in a French study of young adults, which also explored effects of parental class on alcohol use [38]. These combined findings suggest a need for future research to consider different effect patterns of own and parental social class on alcohol use among young people. The likelihood of risk drinking also decreased in correspondence to household financial resources. Poorer young people may be unable to purchase alcohol as easily as their wealthier counterparts, which may explain the latter finding.
Our research and analyses resulted in some specific methodological and conceptual insights into studies of youth health. Among social factors, respondents’ own education was the measure that showed the most, and on average the strongest, statistical associations with all outcome variables. The effects of parental education on health and health behaviours among young adults mostly ran parallel to those of own education and showed weaker effect sizes on average. Thus, educational status of the respondent may appear to be the “best” indicator for social inequality in this age group [see also 36], even though many young people have not yet completed their education. However, we found considerable effects of parental education on youth health behaviours that would not have been captured by simply measuring an individual’s own educational status. For instance, risk drinking among men is less likely among those whose parents have only mandatory education. The effect of educational resources in the family might not have been apparent if we had looked only at individuals’ own education. Parental education shows a stronger effect on overweight and physical inactivity in young women than does own education. These examples indicate that for youth health, family educational resources can be relevant, especially for health factors that relate to health behaviours like eating, drinking and physical activity.
We used a limited number of indicators for social and regional differences, and for health and health behaviours. Other social determinants, such as occupational status, and outcome measures like chronic health conditions deserve to be analysed in future studies. Our sampling procedure excluded non-Swiss residents and people with severe disabilities, and so we could not analyse patterns in the distribution of health and health behaviours among those populations. Injured and sick men are likely to postpone their conscription. Therefore, our sample is likely not to include individuals who were severely ill on the day of the survey. Thus, prevalence of poor health status might be underestimated in our male sample. Self-reports are susceptible to bias from social desirability. Still, validity and reliability of self-reports were demonstrated for most of our outcome indicators, for example, for self-assessed health [23–25], physical fitness [26] and smoking [39]. Underreporting might be an issue in the amount of weekly alcohol consumption [40] and BMI [41]. The sample of women was small, and this led to low frequencies in the cells for the Italian-language region and in the category of mandatory parent education. We classified urban/rural residence according to the official classification of the Swiss Federal Statistical Office, which is based on a very heterogeneous typology of communes, especially in suburban areas. It is possible that if we used an alternative classification, more or stronger differences between urban and rural areas would be evident.
The descriptive findings presented here on the regional differences in health behaviours can serve as starting points for future studies on the determinants of such variations. Merging descriptive and explanatory findings will then allow to produce more comprehensive data for the development and implementation of focused public health interventions.
Our finding that parental education is relevant to health behaviours and health status in young people can serve as a starting point for concept development and formulation of more specific research questions. In particular, our finding that parents’ education shows social gradient effects on overweight and physical inactivity in both genders, combined with our knowledge that overweight and physical inactivity are causally related, suggest that future studies on the social determinants of youth health should seek to capture these bivariate associations, and also to formulate and to answer questions about the effect of social resources on the interplay between different health behaviours.