Diabetes Medication Adherence Lags for Black and Hispanic Medicare Patients

Black and Hispanic Medicare Advantage (MA) enrollees with diabetes met adherence thresholds less often than White enrollees, according to a cohort study published in JAMA Network Open. The gaps appeared for metformin, insulin, and the combined class of sodium-glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide-1 receptor agonists (GLP-1 RAs). Adherence was below 60% in all 3 racial and ethnic groups for every medication category.¹

The team analyzed a 20% national sample of MA enrollees with Part D coverage in 2019 who had a documented diabetes diagnosis in 2017 or 2018 and filled at least 1 prescription for metformin, insulin, an SGLT2 inhibitor, or a GLP-1 RA that year. The total sample was 587,716 beneficiaries (mean age, 72.11 years). Adherence meant having at least 80% of days covered by medications in a class, calculated from the days' supply on Part D claims.¹

Gaps Appeared Across All 3 Medication Classes

Among SGLT2 inhibitor and GLP-1 RA users, 31.4% of Black beneficiaries met the adherence threshold, compared with 40.4% of white beneficiaries, a difference of 9.0 percentage points (P < .001). For insulin, the figures were 40.9% and 48.9%, a difference of 8.0 percentage points (P < .001). Metformin showed the widest gap between Black and white enrollees at 46.0% adherence compared with 56.5%, or 10.5 percentage points (P < .001).¹

In the separate analysis of Hispanic and white enrollees, Hispanic beneficiaries trailed by 8.1 percentage points for SGLT2 inhibitors and GLP-1 RAs (33.9% and 42.0%), 2.8 percentage points for insulin (47.8% and 50.6%), and 3.6 percentage points for metformin (54.0% and 57.6%). All comparisons reached P < .001.¹

The investigators split the unadjusted disparities into within-plan differences, meaning gaps between racial and ethnic groups enrolled in the same plan, and between-plan differences. Within-plan disparities accounted for 74.6% of the gap between Black and white enrollees for SGLT2 inhibitors and GLP-1 RAs, 85.3% for insulin, and 84.5% for metformin. For Hispanic and white enrollees, the within-plan shares were 62.5%, 39.1%, and 66.1%, respectively.¹

Plan composition still mattered for the higher-cost drugs. In plans in the lowest quartile of Black enrollment, regression-adjusted adherence to SGLT2 inhibitors and GLP-1 RAs was 36.1% among white beneficiaries and 35.8% among Black beneficiaries (P = .97). In the highest quartile, it was 33.6% and 27.0% (P < .001). Hispanic and White enrollees showed the same pattern: 47.2% and 45.5% in the lowest quartile plans (P = .76) and 37.7% and 30.4% in the highest (P < .001). For insulin and metformin, adherence among Black enrollees was lower than among White enrollees in every plan quartile.¹

Low-Income Coverage Narrowed Some Disparities

Black and Hispanic beneficiaries were more likely than white beneficiaries to be dual eligible or enrolled in the Part D Low-Income Subsidy (LIS). Enrollment was 39.2% among Black and 47.7% among Hispanic beneficiaries, compared with 16.2% and 14.2% among the white beneficiaries in each respective analysis.¹

Using an Oaxaca-Blinder decomposition, the authors estimated that Medicaid or LIS enrollment was associated with smaller adherence gaps than would be expected if Black and Hispanic beneficiaries enrolled at the same low rate as white beneficiaries. For SGLT2 inhibitors and GLP-1 RAs, the reduction was 2.9 percentage points (95% CI, –3.2 to –2.5) between Black and white enrollees, or 32.3% of the total disparity, and 3.4 percentage points (95% CI, –3.9 to –2.8) between Hispanic and white enrollees, or 41.3%. For insulin, the reduction was 1.8 percentage points in both comparisons, or 22.0% and 63.9% of the total disparity. For metformin, Medicaid or LIS enrollment explained little of the gap.¹

Plan characteristics played a smaller role. They accounted for 18.6% (95% CI, 14.8%-22.4%) of the gap between Black and white enrollees for SGLT2 inhibitors and GLP-1 RAs and 31.0% (95% CI, 25.1%-36.9%) of the gap between Hispanic and white enrollees. For metformin, plan characteristics explained 2.6% (95% CI, 1.5%-3.7%) of the gap between Black and white enrollees. Higher enrollment of Black and Hispanic beneficiaries in contracts with lower star ratings, and enrollment in health maintenance organization plans rather than preferred provider organization plans, were each associated with greater disparities. The authors said those contributions were small compared with Medicaid or LIS enrollment.¹

Access, Metformin, and Study Limits

The authors linked adherence to access to prescribers and pharmacies and cited prior work documenting structural barriers to medication access, including pharmacy closures. Neighborhood disadvantage, measured with the Area Deprivation Index, explained up to 6.7% of the gap in insulin adherence between Hispanic and White enrollees (95% CI, 1.1%-12.3%) and 3.4% between Black and white enrollees (95% CI, 0.9%-5.8%). It did not explain a statistically significant share of the gaps for SGLT2 inhibitors, GLP-1 RAs, or metformin.¹

The authors called the lower metformin adherence among Black beneficiaries notable, given the drug's relatively low cost. They pointed to prior studies identifying patient concerns about long-term use, adverse effects, and beliefs about effectiveness as reasons for metformin nonadherence, factors the analysis could not measure.¹

"The findings suggest that additional insurance- and community-level strategies are needed to improve medication adherence among disadvantaged patients and communities," the authors wrote.¹

The study was observational and remains susceptible to residual confounding. It included only Black, Hispanic, and white beneficiaries, and the proportion of days covered may not reflect how patients actually take their medications.¹

The adherence findings follow known differences in diabetes prevalence. In a cross-sectional analysis of NHANES data from 2007 to 2020, non-Hispanic Black adults had a 47% higher adjusted prevalence of diabetes than non-Hispanic white adults, and Hispanic adults had a 31% higher prevalence. Adults from low socioeconomic status households had a 37% higher prevalence than those from high socioeconomic status households. Diabetes status in that analysis was self-reported.²

CDC says some racial and ethnic minority groups and people with lower socioeconomic status have historically had higher rates of illness and death from diabetes, and that this gap has not substantially narrowed.³

REFERENCES

1. Essien UR, Roberts ET, Ruggiero DA, et al. Racial and Ethnic Disparities in Medication Adherence Among Medicare Advantage Enrollees With Diabetes. JAMA Netw Open. 2026;9(9):e2634846. Published 2026 Sep 1. doi:10.1001/jamanetworkopen.2026.34846

2. Deng Y, Moniruzzaman M, Rogers B, Hu L, Jagannathan R, Tamura K. Unveiling inequalities: Racial, ethnic, and socioeconomic disparities in diabetes: Findings from the 2007-2020 NHANES data among U.S. adults. Prev Med Rep. 2024;50:102957. Published 2024 Dec 27. doi:10.1016/j.pmedr.2024.102957

3. Centers for Disease Control and Prevention. Advancing health equity: diabetes. Diabetes. May 15, 2024. Accessed September 29, 2026. https://www.cdc.gov/diabetes/health-equity/index.html