Glaucoma patients taking statins needed 36% to 60% fewer procedures in a UC San Diego records study

The absolute effect on new cases is small. The two five-year rates differ by 0.3 percentage points ^[15]. With about 210,000 people in each arm ^[3], that is roughly 4,830 diagnoses among statin users against 5,460 among non-users, or about 630 fewer ^[16]. If the drug caused the whole gap, about 333 people would need to take a statin for five years to avoid one diagnosis ^[17]. Dividing the rounded rates gives a relative reduction near 12% ^[14], a little below the reported 15% ^[1].

The procedure figures are larger and harder to size. In patients who already had glaucoma, the study tied statin use to lower risk across three procedures: down 36% for incisional surgery, 55% for minimally invasive glaucoma surgery and 60% for selective laser trabeculoplasty ^[7]. The release does not give the number of patients in each arm who had a procedure, so the absolute benefit cannot be calculated from it.

One design choice rules out an easy artefact. Because both groups had to have documented eye exams on record ^[8], statin users cannot look protected simply because they saw an eye doctor less often and went undiagnosed. Matching across a federated database of 165 million records ^[4] then balances the groups on whatever those records contain. It cannot balance what they leave out. People who keep filling a daily prescription for years may differ from non-users in ways no chart captures.

There is a plausible biological route. Glaucoma destroys retinal ganglion cells and optic nerve fibres, and current treatment works almost entirely by lowering eye pressure, which does not always stop the damage ^[9]. The researchers point to statins' known effects on ocular blood flow, inflammation and ganglion-cell death ^[10]. According to the release, smaller earlier trials had reached conflicting conclusions ^[11].

The write-up states the finding more strongly than the senior author does. It says the study "provides robust population-scale evidence that statins protect against both the development and the clinical progression of glaucoma" ^[19]. Robert Weinreb, director of UC San Diego's Hamilton Glaucoma Center ^[13], put it more carefully. "If this association is confirmed in future studies, it could have implications for a very large number of people," he said ^[18].

Weinreb's word, association, is the accurate one for a records study. I think the evidence justifies a randomized trial with eye endpoints. It does not yet justify a prescription. The procedure results make the stronger case for running that trial, but only if the full paper shows the absolute rates behind them are large enough to matter.