Share
 
Title
Trajectories of physical intimate partner violence among adolescent girls in rural South Africa: Who is at consistently high risk? (HPTN068)
Presenter
A. Pettifor
Authors
S.M. DeLong1, K.A. Powers1, B.W. Pence1, S. Maman2, K.L. Dunkle3, A. Selin4, R. Twine5, F.X. Gómez-Olivé5, C. MacPhail6, K. Kahn5, A. Pettifor1,4,5
Institutions
1University of North Carolina at Chapel Hill, Department of Epidemiology, Chapel Hill, United States, 2University of North Carolina at Chapel Hill, Department of Health Behavior, Chapel Hill, United States, 3South African Medical Research Council, Pretoria, South Africa, 4Carolina Population Center, University of North Carolina at Chapel Hill, Chapel Hill, United States, 5MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt), Faculty of Health Sciences, School of Public Health, University of the Witwatersrand, Johannesburg, South Africa, 6University of Wollongong, New South Wales, Australia

Background: Adolescent girls are at high risk of HIV in South Africa, and intimate partner violence (IPV) is considered an important risk factor. Little is known about how IPV develops differently across individuals over time and what predicts those experiences.
Methods: We identified longitudinal trajectories and related baseline predictors of physical IPV (PIPV) among adolescent girls in the MRC/Wits-Agincourt Health and socio-Demographic Surveillance Site, South Africa. Data came from HPTN068, a randomized controlled trial of conditional cash transfers for HIV prevention. Adolescent girls in grade 8 or 9 in 2011 (baseline visit) who had at least three annual assessments for PIPV (yes/no in the prior 12 months) were included in the analysis (n=907). We used group-based trajectory modeling to identify latent groups with different temporal patterns of PIPV and simple logistic regression to assess baseline predictors of group membership.
Results: We identified two distinct PIPV trajectories among adolescent girls. A “High PIPV” group, comprising an estimated 64.8% of the population, was characterized by an increase in the predicted probability of PIPV from ~12.5% to ~31.6% over the first year, followed by a probability of ~37.5% and ~23.7% thereafter. The “Low PIPV” group, comprising an estimated 35.2% of the population, had a 0% probability of reporting PIPV for the first two years, followed by an increase to ~10% between years 2 and 3 (Figure 1). Girls in the high group were more likely to report at baseline that they had ever had vaginal or anal sex (unadjusted odds ratio [uOR]= 2.78; 95% CI: 1.88, 4.12); had unprotected sex in the prior 3 months (uOR= 2.52; 95% CI: 1.14, 5.57); had ever had transactional sex (uOR= 7.54; 95% CI: 1.70, 33.37); or had an older partner (uOR= 2.84; 95% CI: 1.24, 6.48).
Conclusions: Longitudinal patterns of PIPV varied among adolescent girls, with nearly two-thirds experiencing a consistently high probability of PIPV. Targeted intervention strategies tailored to those at higher risk could maximize reductions in PIPV.


Figure 1. Physical IPV Trajectories among Adolescent Girls in the MRC/Wits-Agincourt HIV and socio-Demographic Surveillance Site, South Africa
[Figure 1. Physical IPV Trajectories among Adolescent Girls in the MRC/Wits-Agincourt HIV and socio-Demographic Surveillance Site, South Africa]