Genetic and environmental pathways to adolescent sexual initiation

Introduction. Previous studies of sexual initiation have focused on psychological, familial, and sociological factors to explain the timing of sexual onset in adolescence. Relatively few studies have investigated potential biological influences such as pubertal timing, androgen hormone levels, and the dopamine-reward system. This series of three studies (1) estimated the overall contribution of genetic and environmental factors to sexual onset, (2) examined the potential role of genetic markers for the regulation of dopamine, and (3) developed several models for the relationship between the timing of pubertal maturity and sexual onset.;Method. In the first study, participants included 799 adopted and non-adopted siblings from the Colorado Adoption Project. This study investigated the relative contributions of genetic and environmental influences on AFS using a biometrical survival model which accounted for censored information and for differences in subjects' sexual initiation by age, adoption status, and sex. The second study included 3451 participants from the Center for Antisocial Drug Dependence (CADD) who were siblings from adoptive and non-adoptive families and twin pairs. Alleles at several dopaminergic markers were examined for their relationship with AFS using family-based association tests and Cox proportional hazard models accounting for censorship in the data. In the final study, participants included 422 female and 420 male siblings in the Longitudinal Twin Study who were assessed for pubertal status from ages 9 through 16 and for AFS. This study used an overall measure of pubertal development as well as specific indices of pubertal maturation. In two different approaches, pubertal trajectories were evaluated using growth models and assessed for their relationship with AFS. Correlations and survival analyses were used to evaluate the relationship between AFS and these measures of pubertal timing and tempo. For those measures that demonstrated some relationship, this study investigated the degree to which genetic and environmental factors played a role using the lower diagonal factorization (Cholesky model).;Results. In the first study, point estimates of variance components from an ACE model, including additive genetic (A), shared environmental (C), and non-shared environmental (E) influences, were 28%, 24%, and 48%, respectively, though a CE model provided the best fit to the data by AIC. In the candidate gene study, the within-family analysis using PBAT revealed a promising relationship between AFS and the 7-repeat (7R) exon-3 D4 receptor polymorphism, but this was not significant after Bonferroni correction. A follow-up cohort analysis was not able to replicate the association using a 4R versus 7R allele comparison. In the third study, linear and quadratic slopes of pubertal trajectories were significantly related in survival analyses to AFS for females, and marginal relationships were found for males. Body hair, an indicator of adrenarche, was significantly predictive of AFS in both males and females. Biometrical analyses suggested a model with additive genetic and nonshared environmental variance explaining the relationship between pubertal growth parameters and AFS.;Conclusions. Implications of the relatively low heritability estimate for AFS from the first study are discussed in terms of social attitudes and behaviors, changing sexual behaviors, and measurement problems. Based on the sum of the existing literature, genetic associations with AFS are to be expected, and similar studies involving dopaminergic markers should focus on a denser and more comprehensive array of polymorphisms across these gene regions as well as an expansion of the search region to include genes for hormones related to sexual and affiliative behaviors. The third study indicated that measures of pubertal timing and a measure of adrenarche were predictive of AFS. Potential explanations for differential findings for indicators of adrenarche and gonadarche, including measurement issues and limitations, are discussed. (Abstract shortened by UMI.).