Precautionary Sexual Behavior Among US Teenagers: A Latent Class Analysis of Psychosocial Vulnerability

Background: Adolescents engage in risky behaviors that increase their vulnerability to sexually transmitted infections including HIV. Scholars have previously attributed psychosocial vulnerability factors to HIV risk. Although current prevention measures address adolescence sexual risk including illicit drug use, very few studies examine psychosocial vulnerability among high school students. Therefore, the purpose of the study is to address the following research questions: (1) Are there distinct patterns of psychosocial vulnerability present among US high school adolescents? (2) If so, what are the distributions of race and precautionary sexual behavior by class membership?

Methods: This study used the Youth Risk Behavior Surveillance System (YRBSS), the prime US national survey on health-risk behaviors among high school students (n=13,589). Teenagers between 14-18 years were included in the analysis, of whom, about 51.2% boys and 48.7% girls. The age range was 14-18. Predictor variables of psychosocial vulnerability included were cocaine use, methamphetamine use, marijuana use, binge drinking, forced sex, alcohol use, depression, and suicidal ideation. The dependent variable condom use was used as a proxy variable for precautionary sexual behavior. Descriptive analyses were conducted to understand the sample. Latent Class Analysis (LCA), a person-centered statistical procedure was used to identify unique groups of adolescents based on various co-occurring psychosocial vulnerabilities. A Chi-squared test was used to determine whether class membership of psychosocial vulnerability was associated with race and reported precautionary sexual behavior.

Results: From the sample of adolescents, LCA resulted in a four-class solution (AIC=80905.213, BIC=81168.282, Entropy=.77, LMR p<.0001) using variables important to psychosocial vulnerability. Based on patterns of vulnerabilities, the four classes were “illicit drug use” (Class 1, n=13,583,5.7%), “mental health problems” (Class 2, n=13,583,11.9%), “normative” (Class 3, n=13,583,47.2%) and “alcoholism” (Class 4, n=13,583,35.1%). Class membership was significantly different across races (χ² [df=12, n=13,262]=245.94,p<0.001). Particularly, African-Americans had the lowest percentage of mental health problems (9.1%) compared to Hispanics (11%), Whites (11.3%), Others (11.5%), while multiracial had the highest percentage (14.6%).  Also, multiracial individuals reported the highest proportion of illicit drug use (7.2%) compared to other races who reported less than 5%. Additionally, class membership of psychosocial vulnerability was significantly different with precautionary sexual behavior (χ² [df=3, n=4,564]=99.57,p<0.001). Specifically, adolescents in the alcoholism class were more likely to engage in precautionary sexual behavior (53.8%) compared to their counterparts with normative behavior (28.5%), mental health problems (11.6%), or illicit drug use (6.1%).

Conclusion and Implications: Our study contributes to a methodological gap by advancing the understanding of psychosocial vulnerability among adolescents using a nationally representative sample. Our findings suggest that practitioners and policy makers should consider heterogeneity among adolescents in the US when designing prevention and intervention programs geared towards sexual behavior prevention. Since psychosocial vulnerability is multidimensional, comprehensive prevention interventions ought to consider the heterogeneity of the student population irrespective of their risk levels. Also, health workers should consider structural factors of psychosocial vulnerability identified in the current study to assess and encourage adolescents to engage in precautionary sexual behaviors.