SGLT2 inhibitors were associated with slower retinopathy progression

Glycaemic control slows diabetic retinopathy, but whether the choice of agent matters beyond its glucose-lowering effect has been unclear. This meta-analysis reviewed 273 studies, including 16 covering 1,785,409 patients, comparing SGLT2 inhibitors against dipeptidyl peptidase-4 inhibitors, GLP-1 receptor agonists, sulfonylureas and other agents.

SGLT2 inhibitor therapy was associated with lower relative risk of retinopathy progression against all comparators pooled (relative risk 0.77, 95% CI 0.72-0.82) and against each subclass individually. Diabetic macular oedema progression showed a similar reduction (0.75, 0.69-0.82). Benefit scaled with baseline risk: in a cohort with 20% retinopathy, the absolute risk difference was around 4 percentage points. The comparative advantage was most pronounced against sulfonylureas, and least against GLP-1 receptor agonists.

The authors' conclusion is appropriately calibrated: in patients with established retinopathy or macular oedema, SGLT2 inhibitors may be preferred over DPP-4 inhibitors and sulfonylureas, while the choice between SGLT2 inhibitor and GLP-1 receptor agonist should rest on other comorbidities. For an ophthalmologist, the practical value is in the letter back to the diabetes team: for a patient with progressing retinopathy still on a sulfonylurea, there is now a reason to raise the agent choice specifically, rather than asking generically for better control.

• Retinopathy progression relative risk 0.77; macular oedema 0.75
• About 4 percentage points absolute benefit where baseline retinopathy is 20%
• Largest advantage over sulfonylureas; smallest over GLP-1 receptor agonists
• Worth naming in correspondence for a patient with progressing retinopathy on a sulfonylurea

The statistics, in plain English

With 1.8 million patients the confidence intervals are very tight, but size does not fix the central problem: these are observational comparisons, and patients prescribed SGLT2 inhibitors differ systematically from those given sulfonylureas in ways no adjustment fully captures. The dose-response with baseline risk and the consistency across every comparator subclass both argue for a real effect, but this is association, and a randomised trial with retinopathy as a primary endpoint does not exist.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.