PURPOSE: To compare the hazard of age-related macular degeneration (AMD) among non-diabetic, weight loss-eligible adults prescribed glucagon-like peptide-1 receptor agonists (GLP-1 RAs; semaglutide or liraglutide) versus other weight loss pharmacotherapies (OWL; phentermine, orlistat, setmelanotide, phentermine-topiramate, or bupropion-naltrexone).
DESIGN: Retrospective cohort study.
PARTICIPANTS: Adults ≥ 50 years without diabetes meeting criteria for weight loss pharmacotherapy and prescribed GLP-1 RAs or OWL medications. Before matching, 60 336 patients were included in the GLP-1 RA group and 21 609 in the OWL cohort; after matching, 20 959 patients remained in each cohort.
METHODS: Cohorts were constructed using deidentified data from the TriNetX Research Network (June 2021 - October 2025). Non-diabetic status was defined as absence of diabetes mellitus diagnosis, hemoglobin A1c ≥ 6.5%, or metformin or insulin use. Patients met criteria for weight loss pharmacotherapy with documentation of an obesity diagnosis, BMI ≥ 30 kg/m, or BMI ≥ 27 kg/mwith dyslipidemia or hypertension, recorded within 1 year before initiation. Inclusion required ≥ 2 prescriptions for a single study medication at least 6 months apart. Patients with recorded medication cross-exposure or outcome diagnoses prior to index were excluded. Cohorts were matched for demographics, established AMD risk factors, covariates influencing treatment allocation, access to ophthalmological care, and proxies for social determinants of health using 1:1 propensity score matching. Outcomes were assessed using Cox proportional hazards models. 2 2
MAIN OUTCOME MEASURES: Hazard ratios of nonexudative AMD, exudative AMD, and any AMD (exudative, nonexudative, or unspecified). Changes in BMI and hemoglobin A1c were analyzed to contextualize findings.
RESULTS: Compared with other weight loss pharmacotherapies, GLP-1 RAs were associated with a lower hazard of nonexudative AMD (HR, 0.47; 95% CI, 0.28-0.78) and any AMD (HR, 0.61; 95% CI, 0.43-0.85), with no difference for exudative AMD (HR, 0.63; 95% CI, 0.30-1.32). BMI and hemoglobin A1c were similar over follow-up.
CONCLUSIONS: Among non-diabetic adults ≥ 50 years old eligible for weight loss pharmacotherapy, prescriptions for GLP-1 RAs were associated with a lower incidence of nonexudative and any (nonexudative, exudative, or unspecified) AMD diagnoses compared with other weight loss medications. The difference in new exudative AMD diagnoses was not statistically significant.
