Design: Multi-database active-comparator cohort and self-controlled case series
Population: 227,971 new semaglutide users with type 2 diabetes across 12 databases
Primary outcome: Neovascular age-related macular degeneration
Effect: Self-controlled IRR 0.92 (0.67 to 1.26) and 1.02 (0.76 to 1.36)

A study across 12 databases of the Observational Health Data Sciences and Informatics network, published in June, examined 227,971 new semaglutide users with type 2 diabetes. It used both active-comparator cohorts — against dulaglutide, empagliflozin, sitagliptin and glipizide — and a self-controlled case series.

No comparison showed a significant difference in neovascular age-related macular degeneration. The self-controlled analysis gave incidence rate ratios of 0.92 (95% CI 0.67 to 1.26) and 1.02 (0.76 to 1.36) under two outcome definitions.

This addresses wet macular degeneration only. It does not speak to the separate concern about semaglutide and non-arteritic anterior ischaemic optic neuropathy.

Reassure patients with type 2 diabetes on semaglutide that this large study found no link with wet macular degeneration.
Keep asking about sudden painless visual loss in semaglutide users, given the separate optic neuropathy concern.
Continue routine diabetic retinopathy screening for all patients with type 2 diabetes.
Do not stop semaglutide because of age-related macular degeneration risk on this evidence.

Why it matters

Ophthalmologists are increasingly asked whether GLP-1 receptor agonists harm the eye, and this answers one part of that.

The statistics, in plain English

Incidence rate ratios near 1 with intervals crossing 1 mean no detectable change in risk. Several comparator hazard ratios had very wide intervals because events were few, so small effects in either direction cannot be excluded.