Risky Sexual Behaviors and Peer Network Health in Urban African American Adolescents

Introduction: African American adolescents tend to have higher rates of risky sexual activity compared to Caucasian youth, and this is especially true for African American males, putting them at high risk of contracting sexually transmitted infections, including HIV. Peer norms and behaviors are strongly associated with risky sexual behavior, thus, peer networks can be either risk-promoting or protective. In this study, we measured peer network health (PNH), asking adolescents about both protective and risky activities of three friends, and examined how PNH was related to risky sexual behaviors in this population of urban African American youth.

Methods: Participants were 244 primarily African American adolescents, ages 13-14, from urban U.S. neighborhoods. Adolescents reported on their PNH and on risky sexual behaviors, utilizing Youth Risk Behavior Survey items. Zero-inflated Poisson (ZIP) regression estimated with restricted maximum likelihood was used to model the count of youth risky sexual behavior at baseline. Useful for data with high rates of zeros, the ZIP model has two parts, a logit model for predicting the excess zeros and a Poisson count model. Health of youth peer networks was standardized and included as a predictor of the probability to be in the zero class (i.e., not sexually active) and participant gender as a predictor of the count of risky sexual behaviors.

Results: Most youth (83%) did not report having sexual intercourse at the time of the baseline assessment. The average number of risky behaviors (e.g., sex without a condom, sex while under the influence of drugs), of those sexually active, was 3.8 (SD=1.4, minimum=1, maximum=7). Peer network health was a significant predictor of being in the zero class (estimate = 0.44, SE = .180, t-value = 2.46, p = .014); higher peer health increased the probability of being in the zero class. Participant gender was a significant predictor of the count of risky sexual behaviors (estimate = 0.43, SE = .141, t-value = 3.07, p-value = .002). For males, the number of risky sexual behaviors increases by a multiple of 1.54 compared to females.

Conclusions: Peer network health is related to risky sexual behaviors, and may be protective, as participants with healthier peer networks were less likely to be engaging in risky sex. Consistent with previous research, males reported engaging in more risky sex compared to females. Interventions that target peer networks may be effective in preventing risky sexual behaviors in African American youth.