Longitudinal association between internalized HIV stigma and antiretroviral therapy adherence for women living with HIV: the mediating role of depression

Abstract

Objective: We investigated whether internalized HIV-related stigma predicts adherence to antiretroviral therapy (ART) longitudinally in women living with HIV in the United States, and whether depression symptoms mediate the relationship between internalized stigma and suboptimal ART adherence.

Design: Observational longitudinal study utilizing data from the Women's Interagency HIV Study cohort.

Methods: A measure of internalized HIV-related stigma was added to the battery of Women's Interagency HIV Study measures in 2013. For current analyses, participants' first assessment of internalized HIV-related stigma and assessments of other variables at that time were used as baseline measures (Time one or T1, visit occurring in 2013/14), with outcomes measured approximately 2 years later (T3, 2015/16; n = 914). A measure of depression symptoms, assessed approximately 18 months after the baseline (T2, 2014/15), was used in mediation analyses (n = 862).

Results: Higher internalized HIV-related stigma at T1 predicted lower odds of optimal ART adherence at T3 (adjusted odds ratio = 0.61, P = 0.001, 95% confidence interval [0.45, 0.82]). Results were similar when ART adherence at T1 was added as a control variable. Mediation analysis revealed a significant indirect effect of internalized HIV stigma at T1 on ART adherence at T3 through depression symptoms at T2 (while controlling for depression symptoms and ART adherence at T1; B = -0.05, SE = 0.03, 95% confidence interval [-0.11, -0.006]).

Conclusion: These results provide strong longitudinal support for the hypothesis that internalized HIV-related stigma results in suboptimal ART adherence in a large sample of women living with HIV in the United States, working through the pathway of increased depression symptoms.

Figure 1.

Depression symptoms at T2^a mediate the effect of internalized HIV-related stigma at T1 on ART adherence at T3 (n=862) Note. Associations are presented as path coefficients (unstandardized). Depression symptoms and ART adherence at T1 (as well as the covariates race, age, time on ART, illicit drug use, income, education, and region) were also entered as control variables. ^a Fifty two participants did not have data on depression symptoms at T2 or T1. These 52 participants did not differ from the participants included in the mediation analysis, on internalized stigma or any of the covariates (all p values > .10), and listwise deletion was used. ^b When depression at T2 is in the model. * p < .05; ^** p < .01