The Effect of the Shikamana Peer-and-Provider Intervention on Depressive Symptoms, Alcohol Use, and Other Drug Use among Gay, Bisexual, and Other Men Who Have Sex with Men (GBMSM) in Kenya

Nicholas Graff | 2018

Advisor: Susan M. Graham

Research Area(s): Global Health, Psychiatric Epidemiology, Public Health Practice


Kenyan gay, bisexual, and other men who have sex with men (GBMSM) face unique stigma and discrimination, which may contribute to adverse mental health symptoms, and could limit antiretroviral therapy (ART) adherence among GBMSM living with HIV. This secondary analysis of data from a randomized pilot study evaluated whether the Shikamana peer-and-provider intervention was associated with changes in psychosocial factors over time. Psychosocial measures were taken at three quarterly study visits by audio computer-assisted self-interview (ACASI). Generalized estimating equations (GEE) were used to evaluate whether rates of change in depressive symptom severity, alcohol use, or other drug use differed by allocation group. We used linear regression to further examine when these differences were observed and to identify other predictors of changes in depressive symptoms. At baseline, both groups were comparable, but the intervention group reported greater depressive symptomatology on average (p=0.09). The intervention was associated with a 1.6 point greater (95% CI: 0.2 – 3.1 points) reduction in depressive symptoms per visit over the 6-month period compared to the standard of care. There were no significant changes in scores for other psychosocial factors. In an exploratory analysis, baseline HIV stigma was associated with greater reductions in depressive symptoms. For each one-point increase in baseline HIV stigma in the intervention group, participants had a 0.12-point (95% CI: 0.02 – 0.23 points) monthly reduction in depressive symptoms. We found that the Shikamana intervention was associated with a reduction in depressive symptoms, and baseline HIV stigma may predict response to the intervention to some extent. Additional research is required to understand factors that influence this peer-and-provider intervention’s effects and whether depressive symptoms, stigma, or other psychosocial factors could be affected in a larger trial.