Does semaglutide need to be discontinued in a patient who develops Non-Arteritic Anterior Ischemic Optic Neuropathy

Comment by InpharmD Researcher

Current guidance on semaglutide use after non-arteritic anterior ischemic optic neuropathy (NAION) remains mixed and is based primarily on pharmacovigilance and observational evidence. While the World Health Organization recommends discontinuing semaglutide once NAION is confirmed as a precaution, available data are inconsistent and do not clearly demonstrate a consistent increase in risk attributable to therapy (see Tables 1-3). Due to this, some experts do not support routine discontinuation in asymptomatic patients; instead, they emphasize counseling on the overall favorable benefit-risk profile and vigilance for rare adverse events, with heightened caution in patients with predisposing optic nerve anatomy.

Background

A 2025 World Health Organization (WHO) safety alert highlighted concerns regarding a potential association between semaglutide and non-arteritic anterior ischemic optic neuropathy (NAION). Following a review of nonclinical studies, clinical trials, post-marketing surveillance data, and published literature, the European Medicines Agency's Pharmacovigilance Risk Assessment Committee (PRAC) concluded that NAION is a very rare adverse effect of semaglutide, occurring in up to 1 in 10,000 patients. The WHO Advisory Committee on Safety of Medicinal Products subsequently recommended that NAION be included as a potential risk in semaglutide risk management plans. The WHO advised that patients experiencing sudden vision loss or rapidly worsening vision during semaglutide therapy should undergo prompt evaluation and that semaglutide should be discontinued if NAION is confirmed. This recommendation was based on pharmacovigilance and safety review data rather than studies evaluating outcomes after continuation versus discontinuation of semaglutide following a NAION diagnosis. [1]

A 2026 systematic review evaluated nine retrospective cohort studies involving more than 3 million patients and found conflicting evidence regarding an association between semaglutide and NAION. Four studies reported an increased risk of NAION, with hazard ratios ranging from 1.76 to 4.28, while four studies found no significant association and one study reported a protective effect. The authors noted that studies demonstrating increased risk generally used broad comparator groups, whereas studies using active comparators with similar metabolic risk profiles consistently found no association. The review concluded that current evidence does not support a causal relationship between semaglutide and NAION and that the conflicting findings are most likely explained by confounding by indication, as patients prescribed semaglutide often have more severe diabetes, obesity, and cardiovascular disease, which are independent risk factors for NAION. Even under the highest reported risk estimates, the absolute risk remained very low and was substantially outweighed by the established cardiovascular and renal benefits of semaglutide. The authors also stated that current evidence does not support routine NAION screening or treatment discontinuation in asymptomatic patients and emphasized that clinicians should counsel patients regarding the overall favorable benefit-risk profile of semaglutide while remaining vigilant for rare adverse events. Of note, the review did not discuss whether semaglutide should be discontinued in patients who develop symptomatic or confirmed NAION, nor did it evaluate outcomes after continuation versus discontinuation of therapy. [2]

Another systematic review, published in 2025, similarly evaluated nine studies examining the potential association between semaglutide and NAION. Findings were inconsistent, with several cohort and pharmacovigilance studies reporting a modest increase in NAION risk, while other matched cohort and real-world studies found no significant association after adjustment for confounding factors. Heterogeneity in study design, patient populations, exposure definitions, outcome ascertainment, and diagnostic criteria limited comparability across studies and precluded meta-analysis. Proposed mechanisms remain speculative, although vascular effects on optic nerve perfusion and rapid glycemic improvement have been suggested. Overall, the authors concluded that the absolute risk of NAION appears to be low and that semaglutide's established metabolic and cardiovascular benefits likely outweigh potential ocular risks for most patients. Current evidence does not support widespread changes to clinical practice; however, clinicians should exercise caution in patients with predisposing ocular anatomy, such as crowded optic nerve heads, small Bruch's membrane opening, or optic disc drusen, and additional prospective ophthalmology-focused studies are needed to clarify whether a causal relationship exists. [3]

References: [1] World Health Organization. The use of semaglutide medicines and risk of non-arteritic anterior ischemic optic neuropathy (NAION). Published June 27, 2025. Accessed June 17, 2026. https://www.who.int/news/item/27-06-2025-27-06-2025-semaglutide-medicines-naion
[2] Eisa N, Barood O. Semaglutide and Non-arteritic Anterior Ischemic Optic Neuropathy: A Systematic Review and Narrative Synthesis. AACE Endocrinol Diabetes. 2026;13(2):259-269. Published 2026 Jan 17. doi:10.1016/j.aed.2026.01.001
[3] Hidalgo Ramos RA, Ortiz M, Dufner Krieger S, Secades D. Semaglutide and Non-arteritic Anterior Ischemic Optic Neuropathy: A Systematic Review. Cureus. 2025;17(8):e89656. Published 2025 Aug 8. doi:10.7759/cureus.89656
Literature Review

A search of the published medical literature revealed 3 studies investigating the researchable question:

Does semaglutide need to be discontinued in a patient who develops Non-Arteritic Anterior Ischemic Optic Neuropathy

Level of evidence

C - Multiple studies with limitations or conflicting results  Read more→



Please see Tables 1-3 for your response.


 

Use of semaglutide and risk of non-arteritic anterior ischemic optic neuropathy: A Danish–Norwegian cohort study
Design

Danish–Norwegian cohort study

N= 61,377 (Denmark: 44,517; Norway: 16,860)

Objective To investigate the putative association between semaglutide and non-arteritic anterior ischemic optic neuropathy (NAION)
Study Groups

Semaglutide users (n= 61,377)

SGLT-2 inhibitors users (n= 119,967)

Inclusion Criteria New users of semaglutide and SGLT-2is in Denmark and Norway from 2018 to 2024 (Denmark) or 2022 (Norway), restricted to those using these drugs for the treatment of type 2 diabetes
Exclusion Criteria Previous history of anterior ischemic optic neuropathy (AION) and migrations within 2 years before cohort entry
Methods Data from national health registries in Denmark and Norway were used. A target trial approach was applied to mimic a hypothetical RCT. Propensity scores were used for confounder adjustment. Hazard ratios were pooled using a meta-analysis approach. A supplementary self-controlled analysis was conducted using the sequence symmetry design
Duration 2018 to 2024 (Denmark) 2018 to 2022 (Norway)
Outcome Measures

Incidence of non-arteritic anterior ischemic optic neuropathy (NAION)

Baseline Characteristics   Semaglutide (n= 60,734)* SGLT-2is (n= 118,967)*
Male sex 54% 64%
Age <50 21% 12%
Age 50–64 41% 35%
Age 65–79 32% 42%
Age 80+ 5.1% 11%
Obesity 22% 9.9%
Heart failure 4.3% 11%
Statins 62% 69%
*Unweighted cohort
Anticoagulants 10% 14%
Results Incidence rate per 10,000 py (events) Semaglutide SGLT-2is Hazard ratio (95% CI)
Denmark 2.18 (24) 1.02 (11.6) 2.17 (1.20–3.92)
Norway 2.90 (8) 0.40 (1.1) 7.25 (2.34–22.4)
Pooled - - 2.81 (1.67–4.75)
Pooled IRD per 10,000 py - - +1.41 (+0.53 to +2.29)
Adverse Events Not specifically detailed in the provided content
Study Author Conclusions The use of semaglutide for managing type 2 diabetes is associated with an increased risk of NAION compared with the use of SGLT-2is. However, the absolute risk remains low.
Critique The study benefits from the use of comprehensive national health registries and a relevant active comparator. However, the relatively low number of NAION events limits detailed subgroup analyses. The observational nature of the study may introduce residual confounding, and the diagnostic codes used may not perfectly distinguish between NAION and AAION. The findings may not be generalizable to more diverse populations outside Denmark and Norway.
References:
[1] [1] Simonsen E, Lund LC, Ernst MT, et al. Use of semaglutide and risk of non-arteritic anterior ischemic optic neuropathy: A Danish-Norwegian cohort study. Diabetes Obes Metab. 2025;27(6):3094-3103. doi:10.1111/dom.16316

 

Semaglutide and Nonarteritic Anterior Ischemic Optic Neuropathy
Design

Retrospective study across 14 databases (6 administrative claims and 8 electronic health records)

N= 37.1 million

Objective To investigate the potential association between semaglutide and NAION in the Observational Health Data Sciences and Informatics (OHDSI) network
Study Groups

New semaglutide users (n= 810,390)

New dulaglutide users (n= 326,282)

New exenatide users (n= 25,936)

New empagliflozin users (n= 715,802)

New sitagliptin users (n= 493,563)

New glipizide users (n= 832,295)

Inclusion Criteria Adults with T2D taking semaglutide, other GLP-1RA (dulaglutide, exenatide), or non–GLP-1RA medications (empagliflozin, sitagliptin, glipizide) from December 1, 2017, to December 31, 2023
Exclusion Criteria Patients with traumatic optic neuropathy or 2 diagnosis codes of giant cell arteritis
Methods The study used an active-comparator cohort design and a self-controlled case-series (SCCS) analysis. The cohort design used propensity score–adjusted Cox proportional hazards models to estimate hazard ratios (HRs). The SCCS used conditional Poisson regression models to estimate incidence rate ratios (IRRs). Network-wide HR and IRR estimates were generated using a random-effects meta-analysis model
Duration December 1, 2017, to December 31, 2023
Outcome Measures

Incidence proportion and rate of NAION

Baseline Characteristics   Semaglutide (n= 43,620) Dulaglutide (n= 14,923) Exenatide (n= 1,414) Empagliflozin (n= 43,302) Sitagliptin (n= 19,581) Glipizide (n= 44,092)
Age ≤29 453 (1%) 168 (1%) 23 (1%) 165 (0%) 29 (0%) 174 (0%)
Age 30-49 8,110 (19%) 3,095 (21%) 315 (22%) 5,122 (12%) 875 (4%) 4,478 (10%)
Age 50-69 24,473 (56%) 8,238 (55%) 832 (59%) 21,614 (50%) 7,292 (37%) 20,330 (46%)
Age ≥70 10,586 (24%) 3,426 (23%) 248 (18%) 16,401 (38%) 11,389 (58%) 19,111 (43%)
Female 26,699 (61%) 8,343 (56%) 797 (56%) 17,872 (41%) 10,683 (55%) 20,587 (47%)
Male 16,921 (39%) 6,580 (44%) 617 (44%) 25,430 (59%) 8,898 (45%) 23,505 (53%)
Results
In this study, the incidence of non-arteritic anterior ischemic optic neuropathy (NAION) among semaglutide users was 14.5 cases per 100,000 person-years. Compared with active comparators (empagliflozin, sitagliptin, and glipizide), semaglutide was not associated with a significantly increased risk of NAION when a sensitive outcome definition was used. Using a more specific NAION definition, a significantly increased risk was observed only when semaglutide was compared with empagliflozin (HR 2.27, 95% CI 1.16–4.46). In a self-controlled case series analysis, semaglutide exposure was associated with a modest increase in NAION risk (incidence rate ratio 1.32, 95% CI 1.14–1.54), although the absolute risk remained low.
Adverse Events No specific adverse events were reported in the study.
Study Author Conclusions Results of this study suggest a modest increase in the risk of NAION among individuals with T2D associated with semaglutide use, smaller than that previously reported, and warranting further investigation into the clinical implications of this association.
Critique The study's large sample size and use of multiple databases are strengths, providing a comprehensive analysis of the association between semaglutide and NAION. However, limitations include the retrospective design, potential for case misspecification due to reliance on diagnosis codes, and lack of access to detailed ophthalmic examination data. Additionally, the study could not evaluate dose-dependent effects or include drug dose in the analysis, which may impact the findings.
References:
[1] Cai CX, Hribar M, Baxter S, et al. Semaglutide and Nonarteritic Anterior Ischemic Optic Neuropathy. JAMA Ophthalmol. 2025;143(4):304-314. doi:10.1001/jamaophthalmol.2024.6555

 

The Effect of Semaglutide and GLP-1 RAs on Risk of Nonarteritic Anterior Ischemic Optic Neuropathy
Design

Retrospective matched cohort study

N= Approximately 120,000 patients with semaglutide prescription and 220,000 prescribed any GLP-1RA

Objective To determine the risk of NAION and ischemic optic neuropathy (ION) in patients prescribed GLP-1RAs
Study Groups

Semaglutide (n= 120,000)

Any GLP-1RA (n= 220,000) Matched T2DM controls High BMI controls

Inclusion Criteria Patients ≥12 years old with type 2 diabetes (T2DM) and considered overweight or obese (high BMI), with at least one ophthalmology or neurology visit
Exclusion Criteria Patients with the outcome prior to the index event
Methods Patients prescribed semaglutide or any GLP-1RA were compared with those on non-GLP-1RA medications. Populations were propensity matched (1:1) on various demographic and risk factors to balance baseline cohorts. Outcomes were evaluated at various time periods after the index event
Duration Outcomes in the T2DM cohort were evaluated at 1-, 3-, and 5-years. For the overweight/obesity group, outcomes were evaluated at 1- and 2-years
Outcome Measures

Cumulative incidence and risk of NAION and ION

Baseline Characteristics   Semaglutide eligible Cohorts No. (%) Semaglutide cohorts After Matching No. (%)
Current Age, Mean ( ± SD) 60.0 ± 12.6 60.2 ± 12.6
Race - White 61761 (57.00%) 61306 (56.90%)
Race - Black or African American 24327 (22.40%) 24162 (22.40%)
Race - Hispanic or Latino 10127 (9.30%) 10101 (9.40%)
Sex - Female 61000 (56.30%) 60488 (56.20%)
BMI (25-30 kg/m2) 30486 (28.10%) 30426 (28.30%)
BMI ( > 30 kg/m2) 77256 (71.30%) 76534 (71.10%)
Essential (primary) hypertension (I10) 88452 (81.60%) 87774 (81.50%)
Hyperlipidemia, unspecified (E78.5) 72897 (67.20%) 72275 (67.10%)
Sleep apnea (G47.3) 54256 (50.10%) 53534 (49.70%)
Other hyperlipidemia (E78.4) 34061 (31.40%) 33759 (31.40%)
Atherosclerotic heart disease of native coronary artery (I25.1) 24636 (22.70%) 24578 (22.80%)
Chronic kidney disease (CKD) (N18) 22122 (20.40%) 22062 (20.50%)
Acute pancreatitis (K85) 2184 (2.00%) 2183 (2.00%)
Malignant neoplasm of thyroid gland (C73) 1071 (1.00%) 1054 (1.00%)
Other chronic pancreatitis (K86.1) 804 (0.70%) 804 (0.70%)
Alcohol-induced chronic pancreatitis (K86.0) 55 (0.10%) 55 (0.10%)
Family history of multiple endocrine neoplasia [MEN] syndrome (Z83.41) 10 (0.00%) 10 (0.00%)
Multiple endocrine neoplasia [MEN] type IIA (E31.22) 10 (0.00%) 10 (0.00%)
Multiple endocrine neoplasia [MEN] type IIB (E31.23) 0 (0.00%) 0 (0.00%)
Sildenafil (136411) 10359 (9.60%) 10211 (9.50%)
Tadalafil (358263) 6400 (5.90%) 6271 (5.80%)
Amiodarone (703) 3298 (3.00%) 3294 (3.10%)
Vardenafil (306674) 1038 (1.00%) 1027 (1.00%)
Avanafil (1291301) 162 (0.10%) 158 (0.10%)
Results
 

In patients with T2DM, semaglutide use was not associated with a statistically significant increase in risk of .NAION or ischemic optic neuropathy (ION) over follow-up periods of up to 5 years compared with matched controls. The relative risk for NAION was actually lower with semaglutide (RR 0.70, 95% CI 0.523–0.937), while the risk of ION was not significantly different (RR 0.788, 95% CI 0.609–1.102). Similarly, across all glucagon-like peptide-1 receptor agonists (GLP-1 RAs), there was no significant difference in NAION (RR 0.887, 95% CI 0.735–1.071) or ION (RR 0.969, 95% CI 0.813–1.154) compared with controls.

In subgroup analyses of patients with high body mass index, semaglutide and GLP-1 RA exposure also did not confer an increased risk of either outcome. Absolute event rates were very low, with a cumulative 5-year risk in semaglutide-treated T2DM patients of 0.065% for NAION and 0.08% for ION. In high-BMI semaglutide users, the 2-year risk was 0.038% for NAION and 0.404% for ION

Adverse Events No significant increase in risk of NAION or ION in patients taking semaglutide or GLP-1RAs compared to T2DM or high BMI controls.
Study Author Conclusions There was no significant increase in risk of NAION or ION in patients taking semaglutide or GLP-1RAs compared to T2DM or high BMI controls.
Critique The study's large sample size and use of propensity score matching are strengths, providing robust data. However, reliance on ICD coding may lead to misclassification, and the lack of direct clinical confirmation of NAION or ION diagnoses is a limitation. Additionally, the study's retrospective nature and potential socioeconomic disparities in GLP-1RA prescription may affect generalizability.
References:
[1] [1] Abbass NJ, Nahlawi R, Shaia JK, et al. The Effect of Semaglutide and GLP-1 RAs on Risk of Nonarteritic Anterior Ischemic Optic Neuropathy. Am J Ophthalmol. 2025;274:24-31. doi:10.1016/j.ajo.2025.02.025