Introduction: Food insecurity is a potentially important barrier to the success of antiretroviral therapy (ART) programs in resource-limited settings. We undertook a longitudinal study in rural Uganda to estimate the associations between food insecurity and HIV treatment outcomes.
Design: Longitudinal cohort study.
Methods: Participants were from the Uganda AIDS Rural Treatment Outcomes study and were followed quarterly for blood draws and structured interviews. We measured food insecurity with the validated Household Food Insecurity Access Scale. Our primary outcomes were: ART nonadherence (adherence <90%) measured by visual analog scale; incomplete viral load suppression (>400 copies/ml); and low CD4 T-cell count (<350 cells/μl). We used generalized estimating equations to estimate the associations, adjusting for socio-demographic and clinical variables.
Results: We followed 438 participants for a median of 33 months; 78.5% were food insecure at baseline. In adjusted analyses, food insecurity was associated with higher odds of ART nonadherence [adjusted odds ratio (AOR) 1.56, 95% confidence interval (CI) 1.10-2.20, P < 0.05], incomplete viral suppression (AOR 1.52, 95% CI 1.18-1.96, P < 0.01), and CD4 T-cell count less than 350 (AOR 1.47, 95% CI 1.24-1.74, P < 0.01). Adding adherence as a covariate to the latter two models removed the association between food insecurity and viral suppression, but not between food insecurity and CD4 T-cell count.
Conclusions: Food insecurity is longitudinally associated with poor HIV outcomes in rural Uganda. Intervention research is needed to determine the extent to which improved food security is causally related to improved HIV outcomes and to identify the most effective policies and programs to improve food security and health.