Adolescent peer groups and alcohol use: longitudinal effect and self-selection in a Czech cohort

Abstract Background Peer norms strongly influence adolescent alcohol use, yet few studies contrast multiple peer profiles, including abstaining peers, when predicting later drinking. This study investigates how distinct peer groups at age 15 relate to later alcohol use and what early-life factors predict peer group membership, thus modeling both peer effects and self-selection. Methods Data come from 5,150 Czech adolescents (ELSPAC study). At age 15, adolescents reported whether their peers a) often got drunk or b) abstained from alcohol. Based on their responses, we created four mutually exclusive categories: (1) exclusively drinkers, (2) exclusively abstainers, (3) mixed group (both drinkers and abstainers), and (4) neither (reference). A structural equation model estimated the effect of peer group membership on latent alcohol use at age 18/19, adjusting for sex, maternal education, and parental drinking. Peer group membership was regressed on alcohol use at age 11 and covariates to assess selection processes. Results Structural model showed an adequate fit, χ2(38) = 168.13, p <.001, CFI = .95, RMSEA = .03. Affiliation with abstaining peers at age 15 predicted lower alcohol use at 18/19 (β = -0.22, p < .001), while being in a mixed peer group predicted higher use (β = 0.26, p = .001). Being in a drinker-only group was not significantly associated with later drinking. Early alcohol use (age 11) was negatively associated with abstainer group membership (β = -0.13, p = .001) and positively associated with mixed group membership (β = 0.14, p = .001), supporting self-selection mechanisms. Conclusions Having abstaining peers in mid-adolescence is protective against later drinking, whereas mixed peer groups pose a risk, possibly due to the influence of drinking peers. Exclusive affiliation with drinkers was not a significant predictor, potentially due to limited statistical power. Early drinking shaped later peer context, highlighting the role of self-selection into peer environments. Key messages • Peer-based prevention should support youth affiliation with exclusively abstaining peers, as having some drinking friends is already a risk factor for later alcohol use. • Very early experience with alcohol steers teens toward riskier peer groups later on; early prevention could disrupt this path and reduce later alcohol problems.