Study wrapper · #421
Assessing the risk of diabetic retinopathy progression with GLP-1 receptor agonists: a systematic review and meta-analysis.
Editor's note
This is a systematic review and meta-analysis (11 studies, RCTs and retrospective cohorts, samples from 137 to 9,463) of whether GLP-1 receptor agonists are associated with progression of diabetic retinopathy — a long-standing safety question for the class, and semaglutide in particular. The pooled result found no statistically significant association overall (RR 1.07, 95% CI 0.90–1.28), with low assessed risk of bias and no detected publication bias. The nuance worth flagging: semaglutide showed the highest point estimates for retinopathy progression in two individual studies (RR 1.76, CI 1.11–2.79; and RR 6.41, CI 0.67–61.46 — the latter's enormous confidence interval signals a tiny, imprecise dataset), contrasting with the null overall. This is a class-level reassurance tempered by an unresolved, compound-specific signal for semaglutide that the authors explicitly say warrants further investigation. Meta-analyses inherit the limits of their inputs — here, heterogeneity is the key caveat.
Plain-language abstract
Diabetic retinopathy is eye damage from diabetes that can harm vision, and there have been questions about whether GLP-1 medicines affect it. This study combined results from 11 earlier studies (including randomized trials and record-based studies, with 137 to over 9,000 participants each) to estimate the overall risk. Combining everything, GLP-1 medicines were not linked to a significantly higher risk of retinopathy getting worse. However, semaglutide specifically showed the highest risk estimates in two of the individual studies — one clearly elevated, and one with such a wide range of uncertainty that it is hard to interpret — which stood out against the reassuring overall picture. The researchers checked their results for bias and found them stable, but noted that the differences between studies, especially involving semaglutide, mean this question needs more research. In short: no clear class-wide increase in risk, but an unresolved question mark around semaglutide.