Multidimensional Profiles of Adolescent Social-Ecological Risk and Protective Factors and Young Adult Sexual Behavior

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Sex, Gender

Abstract

Research has documented how a range of social-ecological risk and protective factors at multiple levels (e.g., family, peer, neighborhood) are associated with sexual behaviors (DiClemente et al., 2005; Kirby, 2007; Poundstone et al., 2004; Zimmer-Gembeck & Helfand, 2008). However, most research on this topic has used variable-centered approaches that examine effects of risk and protective factors separately. A person-centered approach, in contrast, can examine particular patterns of risk and protective factors to better understand which combinations of factors are particularly important in predicting sexual behavior outcomes. In this study we apply a person-centered approach to uncover latent subgroups of adolescents marked by their patterns of risk and protective factors and examine how subgroups were associated with young adult sexual behaviors and sexually transmitted infections (STIs).
Method. Latent class analysis (LCA) was applied to nationally representative data from the National Study of Adolescent to Adult Health (N = 5,114). We identified latent classes (at Wave 1) using 13 indicators of adolescents’ social-ecological contexts across 7 domains: family structure, family processes, romantic relationships, peers, school, religion, and neighborhood. We then examined how these classes predicted having multiple past-year sexual partners and past-year STI diagnoses in young adulthood (at Wave 4) using the “BCH 3-step approach” (e.g, Bakk & Vermunt, 2016).
Results. We selected a model with 5 latent classes (Table 1): Socially and Ecologically Protected (28.9% prevalence), Limited Connections to Peers (17.2%), In Relationship and Connected to Peers (19.7%), Low-Income, Single Parent Household (21.2%), and In Relationship but Socially Disconnected (13.0%). Class membership was a significant predictor of having multiple past-year sexual partners, with adolescents who were Socially and Ecologically Protected having the lowest odds of multiple partners later on (Figure 1). Similar results were found for STIs, although no pairwise tests showed significant differences.
Discussion. Findings suggest that adolescents are growing up in environments with heterogeneous patterns of risk and protective factors for sexual behavior. The largest class was marked by high probabilities of protective factors and lower probabilities of risk factors. Other classes included ones marked primarily by risks related to social connectedness, as well as one class marked primarily by structural risks. Belonging to a class marked by multiple social and ecological protective factors was associated with lower odds of having multiple sexual partners in the past year, suggesting the importance of early contexts on later behavior. However, the other classes did not differ from each on adult outcomes, suggesting the combination of multiple protective factors is particularly important in influencing risky sexual behavior. Additional analyses will examine the roles of gender, race/ethnicity, sexual orientation, and age, as well as examine associations between class membership and concurrent sexual behavior outcomes.

Authors

• Sara A. Vasilenko, Syracuse University
  Presenting Author

• Caitlin S. Smith, Syracuse University
  Non-Presenting Author

• Bethany C. Bray, University of Illinois at Chicago
  Non-Presenting Author