This meta-analysis pooled 16 studies covering 1,785,409 patients with diabetes, comparing SGLT2 inhibitors against DPP-4 inhibitors, GLP-1 receptor agonists, sulfonylureas and other glucose-lowering drugs, with retinopathy and macular oedema progression as the outcomes.
SGLT2 inhibitors were associated with a relative risk of 0.77 for retinopathy progression against all comparators together, and 0.75 for macular oedema progression. The benefit scaled with baseline risk: in a cohort where 20% had retinopathy, the absolute difference was around 4 percentage points. The advantage was clearest against sulfonylureas and least clear against GLP-1 receptor agonists.
This is observational data, and confounding by indication is the obvious problem — the patients started on an SGLT2 inhibitor differ from those started on a sulfonylurea in renal function, cardiovascular risk, affordability and how closely they are followed. None of that is fully removed by adjustment.
What it supports is a conversation, not a switch. When a diabetologist and an ophthalmologist are both involved in a patient with established retinopathy, the choice of glucose-lowering agent is worth raising explicitly rather than assumed to be settled elsewhere. In Indian practice, where sulfonylureas remain widely prescribed on cost grounds, that comparison is the one most likely to come up.
- In a patient with established retinopathy, ask what glucose-lowering agent they are on before assuming the eye is the only variable.
- The clearest contrast was against sulfonylureas, which is the relevant comparison where cost drives prescribing.
- Do not switch an agent on this evidence alone: it is observational and confounded by indication.
- Against GLP-1 receptor agonists the difference was least clear; let other comorbidity decide there.
- Absolute benefit scales with baseline risk, so it means most in a clinic with a high retinopathy burden.
The statistics, in plain English
A relative risk of 0.77 means about 23% fewer patients progressed, and the interval of 0.72 to 0.82 is tight because the pooled population is very large. Size of population is not the same as strength of evidence, though: these are observational cohorts, so the comparison is between people who were prescribed different drugs for reasons the data cannot fully capture. The absolute figure is the more useful one for a patient — about 4 fewer progressions per 100 in a cohort with 20% retinopathy — because a relative risk sounds identical whether the underlying risk is 2% or 40%. The authors are appropriately careful in the conclusion, preferring SGLT2 inhibitors over sulfonylureas and DPP-4 inhibitors but leaving the GLP-1 comparison open.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for ophthalmology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free