DOI: 10.1001/jamanetworkopen.2026.40781 ISSN: 2574-3805

GLP-1 RA Prescribing in Patients With Type 2 Diabetes and Cancer

Xin Hu, Changchuan Jiang, Xuesong Han, Lu Zhang, Piaopiao Li, Francisco J. Pasquel, Jane L. Meisel, Jacqueline B. Vo, Jithin Sam Varghese, Hui Shao, Ilana Graetz
Show PDF Cite

Importance

Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are increasingly used to treat type 2 diabetes, yet little is known about its utilization patterns in patients with cancer history, given limited oncology-specific guidance.

Objective

To evaluate trends and factors associated with GLP-1 RA use among Medicare beneficiaries with type 2 diabetes, comparing those with and without cancer history.

Design, Setting, and Participants

This retrospective cohort study used Surveillance, Epidemiology, and End Results–Medicare linked data to identify patients diagnosed with 7 common cancers in 2010 to 2021 and preexisting type 2 diabetes before cancer diagnosis. A noncancer cohort with type 2 diabetes was identified from a 5% random sample of Medicare beneficiaries and matched 1:1 to patients with cancer based on sociodemographic characteristics; controls were assigned a pseudo–index date corresponding to their matched patient’s cancer diagnosis date. Beneficiaries were followed up annually until death or December 31, 2022, for TM or December 31, 2021, for MA. Data were analyzed from December 2025 to July 2026.

Exposures

Diagnosis of 1 of 7 common cancers.

Main Outcomes and Measures

Any GLP-1 RA use. Proportions of beneficiaries with any GLP-1 RA use annually were stratified by traditional Medicare (TM) and Medicare Advantage (MA). Multivariable logistic regression examined (1) differences in GLP-1RA use between cancer and noncancer cohorts with and without adjusting for sociodemographic characteristics and comorbidities and diabetes-related complications; (2) cancer-specific factors associated with GLP-1RA use among the cancer cohort after 2018.

Results

A total of 373 278 Medicare beneficiaries (mean [SD] age, 76.1 [6.6] years; 188 746 [50.6%] female) with cancer and diabetes and 373 278 noncancer controls (153 358 unique beneficiaries) were included. GLP-1 RA use increased from 85 beneficiaries (0.5%) to 9083 beneficiaries (8.2%) and from 138 beneficiaries (0.9%) to 9300 beneficiaries (6.2%) in the cancer and noncancer cohorts, respectively, from 2010 to 2022 under TM; a similar increase was observed under MA. Despite higher unadjusted GLP-1 RA use among the cancer cohort than the noncancer cohort (probability difference, TM: 1.19 [95% CI, 0.94 to 1.44] percentage points [pp]; P  < .001; MA: 0.45 [95% CI, 0.21 to 0.68] pp, P  < .001), adjusted analysis showed no significant differences in GLP-1 RA use under TM (adjusted probability difference, 0.16 [95% CI, −0.10 to 0.41,] pp; P  = .22) and lower use among cancer cohort under MA (−0.52 [95% CI, −0.76 to −0.28] pp; P  < .001). Complications or comorbidities, including obesity (adjusted probability difference, TM: 3.35 [95% CI, 3.10 to 3.61] pp; P  < .001; MA: 2.42 [95% CI, 2.19 to 2.64] pp; P  < .001), dyslipidemia (adjusted probability difference, TM: 2.26 [95% CI, 2.03 to 2.50] pp; P  < .001; MA: 1.80 [95% CI, 1.57 to 2.03] pp; P  < .001), neuropathy (adjusted probability difference: TM: 3.91 [95% CI, 3.64 to 4.19] pp; P  < .001; MA: 3.17 [95% CI, 2.94 to 3.41] pp; P  < .001), retinopathy (adjusted probability difference, TM: 2.02 [95% CI, 1.78 to 2.27] pp; P  < .001; MA: 1.84 [95% CI, 1.63 to 2.06] pp; P  < .001), and chronic kidney disease (adjusted probability difference, TM: 2.01 [95% CI, 1.76 to 2.26] pp; P  < .001; MA: 1.63 [95% CI, 1.40 to 1.85] pp; P  < .001) were associated with greater likelihood of GLP-1RA use. Within the cancer cohort, breast (adjusted probability difference, TM: 1.05 [95% CI, 0.43 to 1.66] pp; P  = .001; MA: 0.89 [95% CI, 0.37 to 1.41] pp; P  = .001), gynecologic (adjusted probability difference, MA: 1.13 [95% CI, 0.45 to 1.81] pp; P  = .001), and prostate (adjusted probability difference, TM: 1.30 [95% CI, 0.72 to 1.88] pp; P  < .001; MA: 1.07 [95% CI, 0.56 to 1.58] pp; P  < .001) cancers were associated with higher GLP-1 RA use than colorectal cancer, while pancreatic cancer was associated with lower GLP-1 RA use (adjusted probability difference, TM: −2.55 [95% CI, −3.17 to −1.93] pp; P  < .001; MA: −1.64 [95% CI, −2.23 to −1.05] pp; P  < .001).

Conclusions and Relevance

In this retrospective study of Medicare beneficiaries with type 2 diabetes, GLP-1 RA use increased substantially and primarily was associated with metabolic risk profile, with limited associations with cancer status. Variation across cancer types and stages suggests oncologic context may influence prescribing.

More from our Archive