Association of STI/HIV Infection with Reported Behavior Change and Concurrency among Rural Youth in South Africa and Malawi Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 20, 2019
  • Levandowski, Brooke Amanda
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • HIV prevalence in rural southern Africa is increasing towards urban levels, highlighting intervention needs of rural youth. Being faithful is a common HIV prevention message as concurrency, or having more than one sexual partnership at one time, has been associated with STI transmission. Two surveys concerning sexual behaviors and STI prevalence conducted with youth in South Africa (SA) and Malawi were used. Analyses were stratified by gender and country and restricted to rural, sexually experienced youth 15-24 years old. We studied the association between reporting behavior change since hearing of HIV and HIV infection. Many youth reported changed behavior, ranging from 48.2% among Malawian females to 70.6% among SA females. HIV prevalence was 19.0% (95%CI 14.9, 23.1), 4.5% (95%CI 2.6, 6.3), 4.7% (95%CI 3.0, 6.5) and 1.9% (95%CI 0.6, 3.1) for SA women, Malawian women, SA men, and Malawian men, respectively. We found little to no association between behavior change and HIV infection among youth. Analyses stratified by specific behavior changes uncovered that some behaviors were protective while others were riskier, leading to an attenuated effect when all behaviors were combined into a single measure. SA women who reported that they changed their behavior to abstinence were 2.90 (95%CI 1.01, 8.38) times more likely to be HIV+ than those who reported another behavior change. We hypothesized that rural Malawian youth who reported concurrency were more likely to have STIs. Concurrency was reported by women (5.0%) and men (16.0 %). The overall prevalences of STIs and HIV were 10.4% and 4.4% for women and 0.3% and 2.1% for men, respectively. Concurrency was not associated with STIs for either gender or with HIV infection among males. Using multiple imputation methods and controlling for confounders, women reporting concurrency were 6.08 (95% CI 1.23, 30.16) times more likely to be HIV+ than women not reporting concurrency. The cross sectional data may mask the temporal relationship between behavior change or concurrency and STI infection. Further research needs to focus on better measurement of type and duration of prevention behaviors. Innovative HIV prevention methods, highlighting the risk associated with concurrency, are needed to reach rural youth.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Behets, Frieda
  • Open access

This work has no parents.