Reported racial discrimination, trust in physicians, and medication adherence among inner-city African Americans with hypertension

Abstract

Objectives: We sought to determine if reported racial discrimination was associated with medication nonadherence among African Americans with hypertension and if distrust of physicians was a contributing factor.

Methods: Data were obtained from the TRUST project conducted in Birmingham, Alabama, 2006 to 2008. All participants were African Americans diagnosed with hypertension and receiving care at an inner city, safety net setting. Three categories of increasing adherence were defined based on the Morisky Medication Adherence Scale. Trust in physicians was measured with the Hall General Trust Scale, and discrimination was measured with the Experiences of Discrimination Scale. Associations were quantified by ordinal logistic regression, adjusting for gender, age, education, and income.

Results: The analytic sample consisted of 227 African American men and 553 African American women, with a mean age of 53.7 ± 9.9 years. Mean discrimination scores decreased monotonically across increasing category of medication adherence (4.1, 3.6, 2.9; P = .025), though the opposite was found for trust scores (36.5, 38.5, 40.8; P < .001). Trust mediated 39% (95% confidence interval = 17%, 100%) of the association between discrimination and medication adherence.

Conclusions: Within our sample of inner city African Americans with hypertension, racial discrimination was associated with lower medication adherence, and this association was partially mediated by trust in physicians. Patient, physician and system approaches to increase "earned" trust may enhance existing interventions for promoting medication adherence.

FIGURE 1—

Mediation triangle depicting the association between discrimination and medication adherence with trust as a mediator. Note. Total effect: c = –0.145. Direct effect: c’ = –0.089. Mediated proportion: (c- c’)/c = 39%. Effects are from standardized parameter estimates. Bias-corrected and accelerated bootstrap 95% confidence interval for mediated proportion = 17%, 100%.