Background: Minority race/ethnicity is generally associated with antiretroviral therapy nonadherence in US-based studies. Limitations of the existing literature include small samples, subjective adherence measures, and inadequate control for potential confounders such as mental health and substance use, which have been consistently associated with poorer adherence.
Methods: Individual-level data were pooled from 13 US-based studies employing electronic drug monitoring to assess adherence. Adherence was operationalized as percent of prescribed doses taken from the first 12 (monthly) waves of data in each study. Depression symptoms were aggregated from several widely used assessments, and substance use was operationalized as any use of cocaine/stimulants, heroin/opiates, ecstasy, hallucinogens, or sedatives in the 30–365 days preceding baseline.
Results: The final analytic sample of 1809 participants ranged in age from 18 to 72 years and was 67% male. Participants were 53% African American, 14% Latino, and 34% White. In a logistic regression adjusting for age, gender, income, education, and site, race/ethnicity was significantly associated with adherence (P < 0.001) and persisted in a model that also controlled for depression and substance use (P < 0.001), with African Americans having significantly lower adherence than Latinos [odds ratio (OR) = 0.72, P = 0.04] and whites (OR = 0.60, P < 0.001). Adherence did not differ between whites and Latinos (OR = 0.84, P = 0.27).
Conclusions: Racial/ethnic differences in demographics, depression, and substance abuse do not explain the lower level of antiretroviral therapy adherence in African Americans observed in our sample. Further research is needed to explain the persistent disparity and might examine factors such as mistrust of providers, health literacy, and inequities in the health care system.