Abstract: Predicting Age of Sexual Initiation: A Theory-Guided Survival Analysis (Society for Prevention Research 22nd Annual Meeting)

166 Predicting Age of Sexual Initiation: A Theory-Guided Survival Analysis

Schedule:
Wednesday, May 28, 2014
Columbia A/B (Hyatt Regency Washington)
* noted as presenting author
Kari Gloppen, MPH, PhD Candidate, University of Washington, Seattle, WA
Richard F. Catalano, PhD, Professor and Director, University of Washington, Seattle, WA
Jennifer A. Bailey, PhD, Research Scientist, University of Washington, Seattle, WA
Kevin P. Haggerty, PhD, Assistant Director, Social Development Research Group, Seattle, WA
Lisa E. Manhart, PhD, Associate Professor, University of Washington, Seattle, WA
Introduction: Early sexual initiation is associated with increased risk for sexually transmitted infection/HIV and unintended pregnancy.  However, little is known about how social processes influence age of sexual initiation. This theoretically driven study describes the social mechanisms that may affect age of sexual initiation, tests whether constructs from this framework predict age of sexual initiation among adolescents, and assesses whether the influence of these constructs varies across time or by gender.

Methods: Using data from Raising Healthy Children, a study of the etiology of risk with a randomized trial of a preventive intervention nested within it, we assessed time-fixed and time-varying predictors of sexual initiation drawn from surveys administered in 4th through 12th grade.  The sample was recruited as 1st and 2nd grade students at 10 public schools (N=1040) surveyed annually.  Discrete time complementary log-log survival analysis was used to predict the probability of initiating sex at each time point.  Interactions between the predictor variables and year, and the predictor variables and gender were examined to assess whether the influence of the predictors varied across years or by gender.  The timing of entry to risk was considered age 14 (grade 8), the earliest time point at which sexual initiation was measured.

Results: Family factors measured in late childhood, bonding and family rules, were significantly associated with a later age of initiation in unadjusted models, but lost significance after accounting for statistically significant school and peer influences at later ages. This suggests the effect of family influences on sexual decision-making may be mediated by school and peer factors such as school bonding and involvement, and having pro-social friends. Several predictor-by-time interactions were significant, suggesting that their influence on sexual decision making changes across development.  For example family rules and parental supervision became less influential over time. Belief in the future reduced the likelihood of initiating sex in 9th grade but was not a significant influence in later years. None of the predictor by gender interactions was significant.

Conclusions: It is important to consider the changing developmental contexts of young people and how influences on adolescent sexual decision-making change.  While family influences are critically important in predicting childhood behavior, as young people move into adolescence, school and peer influences become increasingly significant in adolescent behavior, including the decision to have sex. It is essential that preventive interventions targeting reductions in STI, HIV, and unintended pregnancy incorporate strategies that reflect this developmental etiology.