Optic neuropathy after semaglutide ramp-up

A case report published in JCEM Case Reports on 29 May 2026 describes a man in his early 30s who developed painless, asymmetric vision loss in both eyes four weeks after his semaglutide dose was escalated to 1 mg weekly^[1]. The presentation was unusual: his optic discs looked normal on examination, yet visual evoked potentials showed markedly delayed signals in both eyes. The authors classify this as retrobulbar optic neuropathy, distinct from the more commonly discussed non-arteritic anterior ischemic optic neuropathy (NAION) that has drawn attention since mid-2024.

What happened clinically

The patient had class III obesity (BMI 47), severe obstructive sleep apnoea on CPAP, prediabetes (HbA1c 6.3%), and dyslipidaemia. He started semaglutide at 0.25 mg weekly, stepped up to 0.5 mg at week four, and reached 1 mg at week eight. Four weeks after the final escalation, he noticed blurred vision^[1].

On examination, his right eye retained 6/6 distance acuity but had a superior visual field defect. His left eye had a central scotoma (mean deviation of minus 5.18 dB on Humphrey perimetry), colour vision dropped to 3 out of 21 Ishihara plates, and a relative afferent pupillary defect was present^[1]. Both optic discs appeared healthy, with no swelling or pallor. OCT confirmed normal optic nerve head and retinal nerve fibre layer thickness. MRI of the brain and orbits was unremarkable.

The key finding was on visual evoked potentials (VEP): amplitudes were reduced and P100 latencies were significantly delayed in both eyes, pointing to dysfunction behind the globe rather than at the optic disc^[1]. Blood tests including inflammatory markers, vitamin B12, and thyroid function were all normal. The team considered demyelinating optic neuritis (MOG/AQP4-associated) and NAION before settling on drug-related retrobulbar optic neuropathy.

How this differs from NAION

Most of the semaglutide-eye-safety discussion since 2024 has centred on NAION, which typically presents with sudden painless vision loss and visible optic disc swelling. A 2024 retrospective study at Massachusetts Eye and Ear reported hazard ratios of 4.28 in patients with type 2 diabetes and 7.64 in patients with overweight or obesity when comparing semaglutide users with non-GLP-1 receptor agonist users^[2]. A 2026 Bayesian meta-analysis of over 1.5 million patients in Neurology calculated a pooled relative risk of 2.52 for NAION with semaglutide^[3].

This new case is different because the optic discs were normal throughout. The damage was retrobulbar (behind the eye), detectable only on electrophysiology. The authors suggest this means the spectrum of semaglutide-associated optic nerve injury may be broader than NAION alone^[1]. For clinicians who screen patients before prescribing GLP-1 receptor agonists, this case raises the question of whether standard fundoscopy (which examines only the disc) is sufficient to catch retrobulbar pathology.

What the consensus says

In May 2026, the North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology published a joint consensus statement in Ophthalmology. They acknowledged the possible association between GLP-1 receptor agonists and NAION but noted that other studies have found no correlation and that the absolute risk remains low^[4]. Their recommendation: shared decision-making between patient and clinician, not blanket avoidance of these drugs.

What we do not yet know

A single case report cannot establish causation. The patient had multiple vascular risk factors (severe obesity, sleep apnoea, prediabetes, dyslipidaemia) that independently predispose to optic neuropathy. Whether semaglutide contributed directly, accelerated a process already underway, or was coincidental is impossible to determine from one patient. No rechallenge was attempted, so we cannot say whether restarting semaglutide would reproduce the problem. Case reports sit at the bottom of the evidence hierarchy, and their value lies in generating hypotheses rather than proving them.

The proposed mechanism (altered optic nerve perfusion and disrupted vascular autoregulation) is plausible but unproven. Whether the rapid dose escalation schedule played a role, or whether slower titration would have been safer, is unknown. It is also unclear whether the retrobulbar pattern seen here shares a common vascular pathway with NAION or represents a separate mechanism entirely. Prospective registries tracking optic events in GLP-1 receptor agonist users would help, but none have reported results yet.

After discontinuation, the patient's vision stabilised but the left central scotoma persisted at follow-up. He regained 3.2 kg over eight weeks without semaglutide^[1]. If you are using semaglutide and notice any change in vision, report it to your prescriber promptly. This article is for educational purposes and does not replace advice from a qualified healthcare professional.