A claim made on a major podcast in August 2026 — that sudden blindness has doubled because of GLP-1 medicines — travelled a long way in a short time [1]. It is worth separating what the underlying research actually reports from what was said about it.
The short answer: a rare optic-nerve condition called NAION has been reported in connection with semaglutide, regulators have looked at it, and the absolute risk involved is very small. The alarm being circulated is a much larger claim than the evidence supports.
Table of Contents
- What is NAION?
- How common is NAION?
- Who is already at higher risk?
- What did the Harvard study actually find?
- Why those numbers do not transfer to the public
- What do clinical guidelines say?
- Which GLP-1 side effects are better established?
- References
What is NAION?
NAION is short for non-arteritic anterior ischaemic optic neuropathy. The condition arises when the blood supply reaching the optic nerve falls away, which is why it is sometimes described in plain language as a stroke affecting the eye.
Vision loss from NAION arrives abruptly and without pain, generally affects a single eye, and is frequently discovered on waking. There is no proven treatment once it has occurred, which is part of why any suggestion that a widely used medicine might trigger it draws attention so quickly.
How common is NAION?
In people over the age of 50, roughly one person in every 10,000 develops NAION in a given year. That baseline figure is the single most useful number for interpreting everything that follows, because a relative risk means very little until it is anchored to how often something happens in the first place.
The United Kingdom's Medicines and Healthcare products Regulatory Agency puts the incidence at 7.73 to 11.35 cases per 100,000 person-years in adults aged 40 and above, and 10.2 per 100,000 person-years among those aged 50 and above [2]. Its 2026 safety update describes NAION as having been "very rarely reported" in association with semaglutide.
Who is already at higher risk?
What triggers NAION is still not fully understood. Several risk factors are recognised, however: diabetes, smoking, raised blood pressure and raised cholesterol each make the condition more likely.
That list matters enormously when interpreting any study of GLP-1 medicines, because it describes almost exactly the group of people who get prescribed them. Untangling the drug from the metabolic profile of the person taking it is the central difficulty in this entire question.
What did the Harvard study actually find?
The alarm traces back to work led by Joseph Rizzo, Professor of Ophthalmology at Harvard. He observed that patients presenting with NAION appeared to be taking semaglutide more frequently than expected, so his team searched their institution's records [3].
In the type 2 diabetes group, 17 semaglutide patients developed NAION compared with six in the comparison arm. In the overweight and obesity group, the split was 20 against three. Added together, that is 37 NAION cases among people taking semaglutide — drawn from a pool of 16,827 high-risk patients attending a single ophthalmology clinic.
Those are small event counts, and the researchers were explicit about what that permits. Their paper is retrospective, and its conclusion states plainly that the study design did not enable query into a causal relationship between semaglutide and the condition.
One further detail from the same paper deserves attention. When the team manually reviewed the records, 40% of cases carrying an ischaemic optic neuropathy code turned out not to be NAION at all — they were the arteritic form driven by giant cell arteritis, or other ischaemic and non-ischaemic optic neuropathies entirely [3]. Diagnostic coding, in other words, is unreliable in this area.
Why those numbers do not transfer to the public
A specialist neuro-ophthalmology service is not a sample of the general population. Patients arrive there precisely because something has gone wrong with the optic nerve, so the rate of optic-nerve disease inside that building bears no relationship to the rate outside it.
There is also a simple sense check available. Tens of millions of people have now taken GLP-1 medicines. Were the risk anywhere near the scale implied by the podcast claim, a wave of blindness would already be visible in ordinary clinical practice and in national statistics. It is not.
That objection was raised publicly. The data analyst who writes as Crémieux argued that the linked evidence rests on small samples with weak controls, and that the implied population-level consequences have simply never materialised [4]. Sinclair replied, "Are you paid to post this? Read the scientific literature" [5]. His post was subsequently annotated by readers on the platform, noting that the linked studies examine patients at one specialist clinic and do not show sudden blindness has doubled in the general population, and that NAION remains very rare with GLP-1 drugs.
What do clinical guidelines say?
Current clinical reference material describes the picture as unsettled rather than resolved. Studies published after the original report have produced mixed findings — some point towards an association, while others detect no increase in risk at all [6].
Two further points from the same source are worth holding onto. The biological plausibility of a mechanistic link between GLP-1 receptor agonists and optic nerve ischaemia remains unproven. And in other settings, these medicines appear to be neuroprotective rather than harmful to nerve tissue.
Which GLP-1 side effects are better established?
Loss of muscle mass is the effect that warrants more routine attention. It is not unique to GLP-1 medicines — essentially every method of losing weight takes some muscle along with the fat — but it is worth planning around rather than discovering later.
Two measures reduce that loss meaningfully: resistance exercise, and eating enough protein while weight is coming down. Both are within the patient's control and both are more useful than worrying about a very rare eye condition.
Pancreatitis and thyroid cancer were both flagged as potential concerns when these medicines came into wide use. Real-world data gathered since has been reassuring on both counts.
None of this means the NAION question should be dismissed outright. Regulators have examined it and updated their product information accordingly, and anyone who experiences sudden loss of vision while taking semaglutide should be assessed urgently by an ophthalmologist [2]. What the evidence does not support is the idea that sudden blindness has doubled across the population.
