- Design
- Drug-target Mendelian randomisation and genetic association analyses
- Population
- 457,242 UK Biobank and 7,058 Singapore participants
- Primary outcome
- Risk of early, advanced and exudative AMD
- Effect
- Advanced AMD OR 1.45 (1.25-1.67); exudative AMD in East Asians OR 3.14 (2.16-4.57)
A drug-target Mendelian randomisation study in Ophthalmology (29 September 2026) used genetic variants mimicking CETP inhibition in 457,242 UK Biobank participants and 7,058 participants in the Singapore Epidemiology of Eye Diseases study.
Greater genetically proxied CETP inhibition was associated with advanced AMD (OR 1.45, 95% CI 1.25-1.67) and early AMD (1.29, 1.19-1.41) in Europeans, and with exudative AMD (3.14, 2.16-4.57), typical neovascular AMD (2.84) and polypoidal choroidal vasculopathy (3.18) in East Asians. The observational genetic score analyses pointed the same way, with a small effect in UK Biobank (HR 1.03 per SD).
CETP inhibitors raise HDL and lower LDL and are under development again for cardiovascular prevention. HDL-related genes have long been implicated in AMD, so the signal is plausible. Mendelian randomisation estimates lifelong genetic exposure rather than years of drug treatment, so this does not prove the drugs cause AMD, but it is a reason to examine the retina in trials and, if approved, to watch patients at high AMD risk.
- Treat CETP inhibitors as carrying a possible retinal safety signal pending trial data
- Ask about AMD history and family history if a patient is enrolled in a CETP inhibitor trial
- Report any new macular disease in patients on investigational lipid drugs
- Note the stronger association in East Asian populations with polypoidal disease
Why it matters
A lipid drug class returning for heart protection may carry an eye risk no cardiovascular trial was designed to detect.
Don't overread it
Genetic proxies, not drug exposure - this is a signal to test, not proof of harm.
The statistics, in plain English
Mendelian randomisation uses inherited variants as a natural experiment, which reduces confounding but assumes the variants act only through CETP. Odds ratios around 3 in East Asians are large but come from a smaller cohort with wider intervals. The UK Biobank hazard ratio of 1.03 per standard deviation is a small per-unit effect.
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