Purpose: To identify risk and protective factors for initiation of sexual intercourse before age 16 years at the level of the individual, family, and school.
Methods: A longitudinal study based on a cohort of 1020 people born in Dunedin, New Zealand in 1972/73 and followed up to age 21 years. Demographic characteristics of the sample were similar to the New Zealand population of that age, except that a smaller proportion (3%) were Maori or Pacific Island Polynesian. Information on individual, family and school factors was collected by interview with parents at ages 3, 5, 7 and 9 years and then by postal questionnaire two-yearly up to 15 years. Subjects were assessed two-yearly from age 3 years and interviewed about their behaviours and ambitions at ages 11, 13, and 15 years. Questions about age at first intercourse were asked by computer at age 21 years. Multivariate logistic regression was used to model associations with age of first intercourse less than 16 years.
Results: Data on age at first intercourse were available for 926/1020 (91%) of surviving members of the cohort assembled at age 3 years. Overall 27.5% of males and 31.7% of females reported sexual intercourse before age 16 years. In multivariate analyses the independent predictors for early sexual initiation for males were: not having outside home interests at age 13 years, no religious activity at age 11 years, not being attached to school at age 15 years, a low reading score, and a diagnosis of conduct disorder in early adolescence. For females, independent predictors were: socioeconomic status in the middle range, mother having her first child before age 20 years, IQ in the middle range, not being attached to school, being in trouble at school, planning to leave school early, cigarette smoking and higher self-esteem score.
Conclusions: Individual and school factors appear to be more important than family composition or socioeconomic status in the decision to have sexual intercourse before age 16 years. The lowering of age at first intercourse may be partly a cohort effect related to high rates of teenage childbearing in the mothers' generation, and to changes in social acceptability of early sexual behaviour.