802-P: Barriers to Medication Taking in African American Patients with Comorbid Diabetes and Hypertension: The Potential Role of Mild Cognitive Impairment

Both diabetes and hypertension increase risk for mild cognitive impairment (MCI) and both rely on medications for control. MCI may impede medication adherence, but this has not been well studied. We examined the association between MCI, defined as missing ≥3 items on the validated Six-Item Screener (SIS), and 19 self-reported barriers to medication taking. Negative binomial logistic regression adjusted for demographic characteristics, physical and mental health, social support, depressive symptoms, and total medication count. Participants in a cluster randomized trial (n=1243; all African Americans with persistently uncontrolled hypertension; 62% female; 43% with diabetes) testing interventions to improve BP control, were studied at baseline. Patients with diabetes were older (59 vs. 56 yrs.), took more medications (14 vs. 9), and were significantly more likely to report too many daily doses (20% vs. 14%; p<0.01) and difficulty affording medications (29% vs. 22%; p<0.01) than those without diabetes. The prevalence of MCI was similar in those with (15.2%) and without (13.4%) diabetes (p=0.35). Regardless of MCI status, 29% of patients with diabetes reported medication side effects and 32% reported allowing medications to run out. In patients with diabetes, those with MCI were significantly more likely to report concern over too many daily doses (30 vs. 18%; p=0.02) and not knowing what their medications were for (25% vs. 12%; p<0.01). In fully adjusted models in patients with diabetes, those with MCI were at high risk for not knowing what their medications were for (OR 2.06; 95% CI 1.09-3.89) and reported modestly more medication barriers although this did not achieve statistical significance (PR = 1.12; 95% CI 0.92-1.36, p=0.26). These findings suggest that patients with diabetes report multiple barriers with medications and those with MCI, which may not be clinically apparent, may require more assistance to optimize adherence. Disclosure D.M. Cummings: Research Support; Self; Novo Nordisk Inc. S. Singh: None. A. Adams: None. S.P. Patil: Research Support; Self; Novo Nordisk Inc. S. Hart: None. A. Cherrington: None. J. Halladay: None. J.B. Ringel: None. E. Baquero: None. M. Safford: None. Funding National Heart, Lung, and Blood Institute (4UH3HL130691)