Social network and inequalities in smoking amongst school-aged adolescents in six European countries.

OBJECTIVES: Smoking contributes to socio-economic health inequalities; but it is unclear how smoking inequalities emerge at a young age. So far, little attention has been paid to the role of friendship ties. We hypothesised that the combination of peer exposure and friendship social homophily may contribute to socio-economic inequalities in smoking at school. METHODS: In 2013, a social network survey was carried out in 50 schools in six medium-size European cities (Namur, Tampere, Hanover, Latina, Amersfoort, and Coimbra). Adolescents in grades corresponding to the 14-to-16 age group were recruited (n = 11.015, participation rate = 79.4 %). We modelled adolescents' smoking behaviour as a function of socio-economic background, and analysed the mediating role of social homophily and peer exposure. RESULTS: Lower socio-economic groups were more likely to smoke and were more frequently exposed to smoking by their close and distant friends, compared with adolescents of higher SES. The smoking risk of the lowest socio-economic group decreased after controlling for friends smoking and social homophily. CONCLUSIONS: Smoking socio-economic inequalities amongst adolescents are driven by friendship networks.

A social network analysis of substance use among immigrant adolescents in six European cities.

BACKGROUND: Social integration and the health of adolescents with a migration background is a major concern in multicultural societies. The literature, however, has paid little attention to the wider determinants of their health behaviours, including the composition of their social networks. The aim of this study was to describe the composition of adolescents' social networks according to migration background, and to examine how social networks are associated with substance use. METHOD: In 2013, the SILNE study surveyed 11,015 secondary-school adolescents in 50 schools in six European cities in Belgium, Finland, Germany, Italy, the Netherlands, and Portugal, using a social network design. Each adolescent nominated up to five of their best and closest friends. Migration status was defined as first-generation migrants, second-generation migrants, and speaking another language at home. We computed two groups of network structural positions, the centrality of individual adolescents in networks, and the homophily of their social ties regarding migration (same-migration). Multilevel logistic regression was used to model the association between network structural position and smoking, alcohol use, and cannabis use. RESULTS: Compared with non-migrant adolescents, adolescents with migration backgrounds had similar relationship patterns. But almost half their social ties were with same-migration-background adolescents; non-migrants had few social ties to migrants. For adolescents with a migration background, a higher proportion of social ties with non-migrants was associated with increased use of cannabis (OR = 1.07, p = 0.03) and alcohol (OR = 1.08, p < 0.01), but not with increased smoking (p = 0.60). Popular migrant adolescents were at less risk of smoking, alcohol use, and cannabis use than popular non-migrant adolescents. CONCLUSION: Homophily of social ties by migration background is noticeable in European schools. The tendency of migrant adolescents to have same-migration social ties may isolate them from non-migrant adolescents, but also reduces their risky health behaviours, in particular cannabis and alcohol use.

Smoking in school-aged adolescents: design of a social network survey in six European countries.

BACKGROUND: In Western countries, smoking accounts for a large share of socio-economic inequalities in health. As smoking initiation occurs around the age of 13, it is likely that school context and social networks at school play a role in the origin of such inequalities. So far, there has been little generic explanation of how social ties at school contribute to socio-economic inequalities in smoking. The SILNE (Smoking Inequalities - Learning from Natural Experiments) survey was designed to test the hypothesis that a combination of peer effect, homophilous social ties, and school context may explain how smoking inequalities are magnified at school - a theory known as network-induced inequality. In this paper, the survey theory and design are presented. FINDINGS: The social network survey was carried out in 2013 in six medium-sized European cities with average incomes similar to the national average: Namur (Belgium), Tampere (Finland), Hannover (Germany), Latina (Italy), Amersfoort (The Netherlands), and Coimbra (Portugal). In each city, 6 to 8 schools were selected in a stratified sampling procedure. In each school, two grades in secondary education, corresponding to 14-16-year-olds, were selected. All adolescents in these two grades were invited to participate in the survey. Social ties were reported using the roster approach, in which each adolescent had to nominate up to 5 friends from a directory. The survey collected information from 11,015 adolescents in 50 schools, out of a total of 13,870 registered adolescents, yielding a participation rate of 79%. The SILNE survey yielded 57,094 social ties, 86.7% of which referred to friends who also participated in the survey. DISCUSSION: The SILNE survey was designed to measure the association between adolescents' social ties at school, their socio-economic background, and their smoking behaviour. Two difficulties were encountered, however: legal privacy constraints made it impossible to apply the same parental consent procedure in all countries, leading to somewhat lower participation rates in two cities: Hannover and Latina. It was also difficult to match the 6 cities in terms of both age and type of education. The SILNE survey provided a comparable database for the study of smoking inequalities across European cities from a social network perspective.