Developmental relations and patterns of change between alcohol use and number of sexual partners from adolescence through adulthood

Abstract

We explored two unanswered questions about the role of alcohol use in sexual behavior. First, we considered whether alcohol use temporally precedes and predicts changes in sexual behavior assessed as the number of sexual partners, whether the reverse pattern holds, or whether the association reflects a common, external cause. Second, we assessed whether associations between these behaviors change as adolescents transition to adulthood. These questions were addressed with a bivariate dual change latent difference score model. Drinking frequency and number of yearly sex partners were assessed 8 times across a 13-year period in a sample of 553 individuals. Assessment began when participants were in the 9th grade (age: M = 15.11 years, SD = 0.43). In addition to an association between the individual growth trajectories of these behaviors, alcohol use was a leading indicator of changes in number of sex partners throughout adolescence, but the reverse pattern was not supported. Of importance, the predictive association could not be explained by individual differences in impulsivity, excitement seeking, conduct problems, hostility/aggression, conventional attitudes, gender, or divorce. Finally, in a developmentally meaningful pattern, alcohol use ceased to significantly predict changes in the number of sexual partners as adolescents transitioned to adulthood.

Figure 1

Latent Difference Score Model of Four Measurement Occasions: Dual Change Variant.

Figure 2

Conditional Bivariate Dual Change Latent Difference Score Model evaluating the relations between alcohol use and number of sexual partners. The manifest variable time series at the top of the figure is denoted A for alcohol use, and the manifest variable time series at the bottom of the figure is denoted S for number of sexual partners. Due to space limitations, some representations were omitted from the figure such as covariances among residuals (e.g.,σ_{εS1, εA1}), covariances among intercepts (I) and growth factors (G) of alcohol use and sex, and residual variances of the difference scores, Δa and Δs.