Correlates of sexual initiation among European adolescents.

Background: Sexuality is a physiological component of adolescent development, though early initiation is associated with reproductive health risk. This study aimed at identifying correlates and predictors of sexual initiation in a large multinational cohort of European adolescents.

Methods: A questionnaire addressing socio-demographics, behaviours, mental health and sexual activity, was delivered to 11,110 adolescents recruited from 168 randomly selected schools in 10 European countries between 2009 and 2011. A follow-up questionnaire was delivered after 12 months. The longitudinal association of baseline risk behaviors, psychological attributes and contextual vulnerabilities, with sexual initiation during follow-up was evaluated through simple and multivariable age/sex stratified logistic regression. Multinomial logistic regression measured the association between predictors and sexual initiation with or without coexisting reproductive risk factors, such as multiple partners or infrequent condom use.

Results: Baseline sexual experience was reported by 19.2% of 10,757 respondents (median age 15; IQR 14-15; females 59.6%). This was significantly more frequent among pupils older than 15 (41%) and males (20.8%). Of 7,111 pupils without previous experience who were available at follow-up (response rate 81.8%), 17% reported sexual initiation, without differences between females and males. Baseline smoking (age/sex adjusted odds ratio [aOR] 3.63), alcohol use (aOR 2.95), illegal drugs use (aOR 2.72), and poor sleep (aOR 1.71) predicted sexual initiation. Stratified analyses showed a particularly strong association in case of younger and female pupils, and, among girls, when initiation was reported together with multiple partners and/or infrequent condom use. Externalizing (i.e. conduct and hyperactivity) symptoms independently predicted sexual initiation. Internalizing difficulties (i.e. emotional and peer problems) were negatively associated with early and risky sexual initiation among boys. Significant predictors included also being bullied, fighting, truancy, and low parental involvement.

Conclusions: Adolescent sexual behaviours are related to non-sexual risk behaviours, psychological difficulties and contextual vulnerabilities. While gateway effects explain some associations, a comprehensive model is needed to understand adolescent sexual behaviours, their physical, mental, and social health outcomes, and their potential positive effects on wellbeing. Tailored interventions may need to consider younger girls as a particularly vulnerable group in view of a strong association between non-sexual and sexual behaviors.