DOI: 10.1001/jamanetworkopen.2026.34846 ISSN: 2574-3805

Racial and Ethnic Disparities in Medication Adherence Among Medicare Advantage Enrollees With Diabetes

Utibe R. Essien, Eric T. Roberts, Dominic A. Ruggiero, Jasmyn J. Tang, Sindhu Raghava, John Z. Ayanian, Renuka Tipirneni
Show PDF Cite

Importance

Appropriate medication management can prevent adverse outcomes, yet medication nonadherence remains common, including among patients with diabetes.

Objective

To examine racial and ethnic disparities in medication adherence among Medicare Advantage (MA) enrollees with diabetes and assess factors contributing to these disparities.

Design, Setting, and Participants

Using the Medicare Master Beneficiary Summary File and MA encounter data, this cohort study included MA enrollees with Part D coverage in 2019 and a documented diabetes diagnosis from January 1, 2017, to December 31, 2018, who filled at least 1 prescription for metformin, insulin, a sodium-glucose cotransporter 2 (SGLT2) inhibitor, or a glucagon-like peptide-1 receptor agonist (GLP-1 RA) in 2019. Analysis was conducted between December 2024 and November 2025.

Main Outcomes and Measures

Unadjusted and adjusted disparities in adherence (proportion of days covered ≥80%) during 2019 were examined for insulin, metformin, and SGLT2 inhibitors and/or GLP-1 RAs, distinguishing within-plan disparities from differences between plans serving proportionately more vs fewer Black or Hispanic enrollees. Oaxaca-Blinder decomposition analysis estimated the contribution of covariates to adherence disparities.

Results

Among 587 716 total beneficiaries (mean [SD] age, 72.11 [8.85] years), analyses of disparities between Black and White enrollees included 80 586 Black (63.3% female) and 233 505 White (50.9% male) beneficiaries; analyses of disparities between Hispanic and White enrollees included 83 116 Hispanic (53.9% female) and 190 509 White (51.1% male) beneficiaries. In unadjusted analyses, adherence was lower among Black beneficiaries than among White beneficiaries: −9.0 percentage points [pp] for SGLT2 inhibitors and/or GLP-1 RAs, −8.0 pp for insulin, and −10.5 pp for metformin (all P  < .001). Adherence was also lower among Hispanic beneficiaries than among White beneficiaries: −8.1 pp for SGLT2 inhibitors and/or GLP-1 RAs, −2.8 pp for insulin, and −3.6 pp for metformin (all P  < .001). By medication type, within-plan disparities accounted for 74.6% to 85.3% of adherence disparities between Black and White enrollees and 39.1% to 66.1% of disparities between Hispanic and White enrollees. Medicaid or Low-Income Subsidy enrollment was associated with reductions in disparities between Black and White enrollees (22.0% for insulin and 32.3% for SGLT2 inhibitors and/or GLP-1 RAs) and Hispanic and White enrollees (63.9% for insulin and 41.3% for SGLT2 inhibitors and/or GLP-1 RAs).

Conclusions and Relevance

This retrospective cohort study of MA enrollees with diabetes found racial and ethnic disparities in diabetes medication adherence. Disparities in adherence to insulin or SGLT2 inhibitors and/or GLP-1 RAs were partially mitigated by Medicaid and prescription drug subsidies but less so by plan-level factors. The findings suggest that additional insurance- and community-level strategies are needed to improve medication adherence among disadvantaged patients and communities.

More from our Archive