Urgent symptoms come before reading the statistics
If your eyesight suddenly deteriorates or you lose vision, seek urgent medical assessment. Do not wait for a routine refill message, a review-site response or a symptom to become painful. The UK Medicines and Healthcare products Regulatory Agency specifically advises urgent eye or emergency care for sudden impairment or rapidly worsening vision while using semaglutide. An article cannot identify the cause from a description. The immediate task is assessment; interpreting a medicine’s possible role comes with the clinicians evaluating you.[2]
What NAION means
Non-arteritic anterior ischemic optic neuropathy, shortened to NAION, concerns reduced blood supply to part of the optic nerve. MHRA describes a typical presentation as sudden, painless loss of vision in one eye, sometimes experienced as blurring or cloudiness. That description is a warning to seek care, not a checklist that can confirm or exclude the condition at home. I would not use the absence of pain as reassurance. Also avoid treating every visual change during treatment as the same diagnosis; different eye problems need different assessments.[2]
What the European review concluded
In June 2025, EMA’s safety committee concluded that NAION was a very rare adverse effect of semaglutide medicines and recommended updating their product information. The agency described an estimated additional occurrence of about one case per 10,000 person-years of treatment in the epidemiological evidence it reviewed. A person-year is one person observed for a year. That estimate belongs to the reviewed populations and evidence; it is not a personalized prediction, a lifetime risk or a statement that any one symptom has a particular cause.[1]
Why “twice the risk” needs an absolute number
The EMA communication also described an approximately twofold association in several large epidemiological studies involving adults with type 2 diabetes. A relative comparison tells you how two rates relate; it does not tell you the starting rate on its own. I would show the absolute estimate beside the relative one, rather than using either to alarm or reassure by itself. The table keeps the units visible. In particular, an additional event rate per person-year is not the same measurement as the percentage affected over a multi-year follow-up.[1] For the arithmetic distinction, see absolute versus relative risk before comparing percentages from different reports.
| EMA statement | Unit or scope | What it is not |
|---|---|---|
| Very rare adverse-effect category | May affect up to 1 in 10,000 users | Not a personalized lifetime prediction |
| Approximately twofold association | Relative comparison in reviewed epidemiological studies | Not a 50% chance of vision loss |
| About one additional case | Per 10,000 person-years of treatment | Not the total rate from every possible cause |
EMA June2025 safety communication. Category frequency, relative association and additional annual event rate are different summaries; do not substitute one denominator for another. Source: EMA: NAION classified as a very rare semaglutide side effect.
Why individual studies do not all read the same way
A multinational population-based study by Chou and colleagues did not establish the same increased risk across its comparisons and time periods. That does not erase a subsequent regulator’s broader review. Studies can differ in who receives each treatment, which comparison medicine is used, how an eye diagnosis is recorded and how long people are followed. Observational data require particular care because treatment groups are not created by random assignment. I would look for the population and design before deciding that two headlines genuinely contradict one another.[3] Our real-world versus trial guide explains why a large database alone does not remove differences between groups.
NAION and diabetic retinopathy are separate questions
Our retinopathy coverage concerns a different eye condition and a separate body of evidence. Combining them under a single vision-loss percentage would make both less clear. If you already have an eye diagnosis, bring its actual name and relevant records to the prescriber rather than saying only that you have eye trouble. Ask what follow-up is appropriate for that history. A risk discussion should also account for the condition being treated and the treatment’s established benefits, rather than being reduced to a list of unrelated warnings. Read semaglutide and diabetic retinopathy for that distinct research question.
Product information differs across jurisdictions
MHRA issued its UK update in February 2026. Australia’s TGA published a rare-vision-disorder update in July 2026. These communications show that safety surveillance continues after initial approval. They should be identified by their regulator and location; an overseas notice is not automatically the wording of a US product label. Ask the treating clinician to interpret the information applicable to your product and situation. Do not infer a complete class-wide incidence from a semaglutide-specific estimate, and do not substitute an online forum’s summary for an actual safety communication.[4]
The questions worth recording for your care team
For a non-urgent discussion before treatment, ask how your eye history affects the plan, what changes should prompt urgent care and how to reach the practice between appointments. If a new visual event occurs, clinical assessment takes priority over any documentation task. Once care is arranged, a clear record of the medicine, timing and clinical findings can help the professionals evaluating the event. My standard for useful coverage is straightforward: keep the rare-event numbers in context while making the action for sudden symptoms unmistakable. For ongoing care, keep the urgent contact route somewhere easy to find. A routine online message and an emergency assessment serve different purposes; the convenience of the former should not delay the latter when vision changes suddenly. For reporting after clinical care, our MedWatch guide explains the difference between a suspected adverse-event report and proof of causation.
Continue your research
Sources & scope
Sources checked October 23, 2026. This guide explains published information and consumer research questions; it is not a clinical evaluation.
- EMA: NAION classified as a very rare semaglutide side effect
- MHRA: February 2026 semaglutide vision safety update
- Chou et al.: multinational population-based NAION study
- TGA: July 2026 update on GLP-1 medicines and rare vision disorder
Prices and terms can change. Confirm your specific offer directly with the provider.