The impact of weight-loss drugs on eye health

May 2, 2026 Dr Arvind Gupta

Glucagon-like peptide-1 receptor agonists and risk of neovascular AMD

Shor R, Mihalache A, Noori A, et al
JAMA Ophthalmol, 2025;143:587-594

 

Review: This population-based retrospective study, based on a review of Ontario Health Insurance records of patients between January 2020 and November 2023, examined the risk of neovascular age-related macular degeneration (nAMD) in patients on GLP-1 receptor agonists (GLP-1RAs). Patients aged 66 years or older, with GLP-1 RA exposure of at least six months and a follow-up of 12 months, were included. In total, 46,334 patients exposed to GLP-1RAs were identified and twice as many age-matched controls were selected. Systemic comorbidities associated with any kind of AMD and socioeconomic status were also matched.

 

The incidence of nAMD was higher in the exposed cohort than in the unexposed cohort. Cox proportional hazard models, both unadjusted (crude) and adjusted, estimated hazard ratios for nAMD development of greater than 2.0 among patients exposed to GLP-1RAs (exposed, 0.2% vs unexposed, 0.1%; HR, 2.11; 95% CI, 1.58–2.82; adjusted: HR, 2.21; 95% CI, 1.65–2.96). In this cohort study, the use of GLP-1RAs among patients with diabetes was associated with double the risk of incident nAMD development compared with similar patients with diabetes who did not receive a GLP-1RA.

 

Comment: This is the first study looking at the risk of nAMD in patients with exposure to GLP-1RAs. It is too early to conclude on the true causal association. Theoretically, GLP-1RAs would be expected to reduce the risk of AMD, since they reduce insulin resistance, and increased insulin resistance is linked to an increased risk of AMD. Hence, studies are needed to establish the pathophysiological mechanism underlying this association.

 

Risk of retinal vein occlusion between GLP-1RAs and dipeptidyl peptidase-4 inhibitors in type 2 diabetes: a retrospective cohort study

Pan S-Y, Lin H-J, Weng C-H, Wang I-J, Tien P-T, Chou C-C, Tsai S-F, Lin J-F

Ophthalmol Sci. 2025 Feb 7;5(4):100734

 

Review: A multinational, retrospective cohort study across 21 countries was conducted to evaluate the association between GLP-1RAs and the risk of retinal vein occlusion (RVO) compared with dipeptidyl peptidase-4 (DPP-4) inhibitors in patients with type 2 diabetes mellitus (T2DM). There were 79,486 eligible participants, with a 1:1 ratio between those on GLP-1RAs and DPP-4 inhibitors. The propensity score matched the two groups for all the possible variables. GLP-1RA use was associated with a lower risk of central RVO (hazard ratio [HR], 0.73; 95% confidence interval [CI], 0.54–0.98) and branch RVO (HR, 0.62; 95% CI, 0.41–0.95) over five years compared with DPP-4 inhibitor use. This association is consistent among patients aged ≥50 years, African Americans, those prescribed human-analogue GLP-1RAs and those with baseline HbA1c ≥8%, BMI ≥30kg/m² and diabetes duration ≥3 years. The authors concluded GLP-1RAs may have a protective effect, in contrast to DPP-4 inhibitors.

 

Comment: This is the first large-scale study demonstrating the beneficial effect of GLP-1RAs over the commonly used medication in diabetic patients. This preventive effect goes beyond the benefits of mere blood glucose reduction. This could be related to the anti-inflammatory properties of GLP-1RAs. However, some smaller studies have suggested an increased risk of central RVO with GLP-1RAs due to their thrombotic properties. More studies are required to understand the true association.

 

Association of GLP-1RAs with optic nerve and retinal adverse events: a population-based observational study across 180 countries

Lakhani M, Kwan ATH, Mihalache A, Kertes PJ, et al

Am J Ophthalmol. 2025 Sep;277:148-168

 

Review: This global observational pharmacovigilance study investigated the association of GLP-1RAs and common ocular adverse events (AEs), particularly ischaemic optic neuropathy (ION) and diabetic retinopathy (DR) and maculopathy.

The US FDA Adverse Event Reporting System (FAERS) database (via OpenVigil 2.1) and WHO’s VigiBase (via VigiAccess) for optic nerve and retinal AEs associated with semaglutide and tirzepatide, covering the period from their approval dates – December 2017 for semaglutide and May 2022 for tirzepatide – through September 2024, were analysed. In FAERS, all other drugs were compared, while in VigiBase, metformin, empagliflozin, dulaglutide and insulin served as controls. Disproportionality metrics included reporting odds ratios (RORs) with 95% confidence intervals (CI). Semaglutide showed significantly higher reporting odds of ION (FAERS: ROR 11.12, 95% CI 8.15–15.16; VigiBase: ROR 68.58, 95% CI 16.75–280.67) and DR (FAERS: ROR 17.28, 95% CI 13.62–21.91; VigiBase: ROR 7.81, 95% CI 5.60–10.90), as well as retinal/vitreous detachment, retinal/vitreous haemorrhage and retinal tear. In FAERS, the reporting odds ratios for these latter events ranged from 2.44 to 5.89 (95% CI 1.70–8.97; all P<0.001) versus all other drugs, while in VigiBase they ranged from 5.49 to 20.91 (95% CI 2.71–90.11; all P≤0.0001) versus metformin.

 

Comment: This is the largest real-world study to demonstrate a very strong association between GLP-1RAs and ION (10-fold or more) and DR (seven-fold or more) compared with other drugs. However, smaller studies have raised doubts about these associations. The landmark Sustain-6 phase 3 trial showed a very small association between DR progression and GLP-1 RA therapy. It is well known that rapid lowering of blood glucose can worsen DR. Hence, patients planned to start GLP-1RA therapy are recommended to take a baseline assessment of DR status and close follow-up is essential.

 

 

Dr Arvind Gupta is a consultant at Health New Zealand Counties Manukau and Auckland, one of the directors at Eye Doctors, Auckland and specialises in cataract surgery with premium IOL, neuro-ophthalmology and medical retina.