A Typology of Risk for HIV/AIDS Among Malawian Adolescents: Implications for Universal and Targeted Programs

Purpose: Although prevention efforts have been implemented, the rate of new HIV infection among adolescents still remains a major issue in Malawi, and the majority of newly infected individuals are between the ages of 15 and 24 (UNAIDS, 2008). To decrease these rates, it is necessary to identify which individuals are most at risk for infection (Bertozzi, Laga, Bautista-Arredondo, & Coutinho, 2008), and then match them to programs that focus on changing their sexual behaviors. Using latent class analysis, the purpose of this study was to identify a typology of risk using constructs from AIDS behavioral theories (Noar, 2009). Specifically, three questions were addressed: 1) Do patterns of HIV/AIDS risk exist among Malawian adolescents? (2) Do the patterns differ by adolescents' perceptions of their HIV/AIDS status? (3) Are the patterns associated with changes in adolescents' sexual behaviors?

Method: This study analyzed data collected in 2004 by the Malawi Diffusion and Ideation Project. During in-person interviews, participants completed the Role of Informal Conversations on Health and AIDS Behavior survey. The sample for this study was restricted to males and females between the ages of 15 and 24 who reported never having been married and were not missing data on the dependent variable (N = 606). The analyzed sample included 61.2% males; 31.7%, 30.4%, and 38% from Balaka, Mchinji, and Rumphi, respectively.

Latent Class Analysis (LCA) was used to address questions one and two, and was conducted using Mplus. This software was selected because it can include covariates and sampling weights, which is important for this study because the sample was disproportionate by site, gender, and age. To answer question three, a chi-square test was used.

Results: The study results indicate that two classes of risk exist among Malawian youths: High Risk class and At Risk class. Of the sample, the High Risk class comprise 63% and At Risk class comprise 37%. Further, individuals in the High Risk class are significantly more likely than the At Risk class to perceive themselves as currently infected with HIV/AIDS. Last, the High Risk class is significantly less likely than the At Risk class to report changes in their sexual behaviors to avoid HIV/AIDS [÷2 (1, N = 606) = 10.48; p < .01]. For example, when changes in sexual behavior were not reported, it was members of the High Risk class 42.5% of the time and members of the At Risk class 29.3% of the time.

Conclusions and Implications: Results indicate that among the Malawian adolescents in this sample two classes of risk exist, and each class reports at least some risk for contracting HIV/AIDS. Thereby, indicating that universal programs are needed. Results also identify which adolescents may be at a high risk for contracting HIV/AIDS, and indicate that for a subset of adolescents targeted programs are warranted. Policy implications related to the typology and adolescents' sexual behaviors is discussed.