The Role of Religiosity on Racial/Ethnic Disparities in Non-adherence to Antihypertensive Medications
摘要
Some studies have reported an association between religiosity (i.e. adherence to religion) and non-adherence to antihypertensive medications (AHMs), which is disproportionally higher among Black/African Americans (BAAs) as compared to non-Hispanic Whites (nHWs). Given that religiosity varies by race/ethnicity, we sought to answer the question, “does religiosity explain the disparities in AHM nonadherence between BAA and nHW populations?”.
MethodsCounty-level measures of antihypertensive medication (AHM) non-adherence, religiosity, healthcare access, and social determinants of health factors were derived by linking three datasets: the CDC Interactive Atlas of Heart Disease and Stroke (2014–2016), the County Health Rankings and Roadmaps (2016), and the U.S. Religion Census Data (2010). We applied a Hierarchical Linear Model (HLM) to model the association between religiosity (proportion of adherents to any religion per county) and BAA–nHW disparity in AHM non-adherence, defined as the ratio (prevalence ratio) of county-level prevalence of non-adherence among BAA to that among nHWs. We adjusted the HLM models by several healthcare access and social determinants of health factors. We stratified the analysis by the top 10 larger religious denominations to examine differences by adherence to these denominations.
ResultsThe mean prevalence of AHM non-adherence was significantly higher among BAAs (34.5%, 95% CI: 34.4, 34.6) as compared to nHWs (25.4%, 95% CI: 25.3, 25.5), yielding a prevalence ratio of 1.36. While religiosity was significantly associated with AHM non-adherence overall, it was not associated with the observed BAA–nHW disparities. The mediation analysis, however, showed that religiosity may be associated indirectly with BAA–nHW disparities in AHM non-adherence through indirect pathways mediated by food insecurity and physical distress.
Conclusion and RelevanceWhile religiosity may not be directly associated with the observed BAA-nHW disparities, it appears that religiosity may have an indirect association with this disparity via modifiable factors such as food insecurity and physical distress.