Scientific deep-dive

Semaglutide vs Tirzepatide: Which Has Higher NAION Risk?

Data lean toward a smaller NAION signal with tirzepatide than semaglutide, but the evidence is observational and no trial compares them. What cohorts and adverse event reports show, why report databases can't measure the gap, and what to ask a prescriber.

By Eli Marsden · Founding Editor
Editorially reviewed & fact-checked against primary sources · How we verify contentLast reviewed
8 min read·22 citations

Published data lean toward a smaller NAION signal with tirzepatide than with semaglutide, but no randomized trial or semaglutide-only cohort has settled the question. The strongest comparative estimate is a hazard ratio of 0.45 (95% CI 0.27 to 0.86) for tirzepatide versus GLP-1 drugs without GIP activity, in a retrospective cohort of adults with type 2 diabetes[1]. Spontaneous report databases point the same way, with a reporting odds ratio of 15.77 for optic ischemic neuropathy on semaglutide versus tirzepatide[4], though that author says the findings do not measure comparative clinical risk. NAION, a sudden loss of blood flow to the front of the optic nerve, stays rare on either drug.

About this article

This page covers one question: does the NAION signal differ between the two drugs? For what NAION is, the original semaglutide cohorts and the class-wide debate, see our main NAION evidence review.

The studies that compare the drugs

Hong and colleagues give the closest thing to a direct answer. They matched 102,590 adults with type 2 diabetes starting tirzepatide to the same number starting a non-GIP GLP-1 drug, using a nationwide electronic health record network from June 2022 to June 2025[1]. Tirzepatide was associated with less NAION (HR 0.45), and also with less diabetic retinopathy (HR 0.79) and fewer retinopathy-related vision-saving procedures (HR 0.65)[1]. Three limits matter. The comparison group mixes non-GIP GLP-1 drugs, so this is not a pure semaglutide comparison. The abstract gives no NAION event counts. And the population has diabetes, so the result does not transfer automatically to people using Zepbound or Wegovy for obesity alone. The authors themselves say confirmation in prospective studies is warranted[1].

Every eye outcome in that study moved in tirzepatide’s favor. That could reflect a real drug difference or unmeasured differences in who gets prescribed a newer drug, and a retrospective design cannot separate the two.

The one cohort that compared tirzepatide with semaglutide by name did so in people with obesity, not diabetes. Among 25,060 matched pairs with a BMI of at least 27, Huang and colleagues reported that no differences emerged across ocular outcomes[2]. NAION is not among the primary outcomes the abstract lists, which were broader categories such as visual disturbances and visual issues with blindness, so this is not evidence of equal NAION risk[2].

A third large study is often quoted in this debate and cannot answer it. Wang and colleagues found a hazard ratio of 1.76 (95% CI 1.01 to 3.07) for NAION, with 0.04% versus 0.02% affected over two years, but they pooled semaglutide and tirzepatide into a single exposure group of 79,699 patients[3]. Its data ended in January 2023[3], and tirzepatide was approved only in May 2022[8], so by our reading the pooled group probably leaned heavily toward semaglutide.

Evidence that bears on semaglutide versus tirzepatide. HR is a cohort hazard ratio; ROR is a reporting odds ratio, which does not measure risk.
StudyDesign and populationComparisonNAION or optic ischemic resultWhat it can tell you
Hong 2026Retrospective EHR cohort, 102,590 per arm, type 2 diabetesTirzepatide vs non-GIP GLP-1 drugsHR 0.45 (95% CI 0.27 to 0.86)Closest estimate; comparator is not semaglutide alone
Huang 2025Retrospective TriNetX cohort, 25,060 pairs, BMI 27 or higherTirzepatide vs semaglutideNo differences across ocular outcomes; NAION not a listed primary outcomeDirect comparison, but not of NAION
Wang 2025Target trial emulation, 79,699 per arm, type 2 diabetesSemaglutide or tirzepatide, pooled, vs other diabetes drugsHR 1.76 (95% CI 1.01 to 3.07)Cannot separate the two drugs
Arslan 2026FAERS reports, 60,423 semaglutide and 143,126 tirzepatideSemaglutide vs tirzepatideROR 15.77Reporting pattern, not risk

Why the report databases can’t measure the gap

The widest gaps come from pharmacovigilance data. In the FDA Adverse Event Reporting System, Kakkar and colleagues calculated a reporting odds ratio of 94.45 (95% CI 83.02 to 107.45; 355 reports) for semaglutide and 2.94 (95% CI 2.20 to 3.93; 49 reports) for tirzepatide[5]. Arslan’s direct comparison gave 15.77, with no confidence interval in the abstract[4]. In European reports, semaglutide also signaled against tirzepatide on optic ischemic neuropathy (IC025 0.17 in health professional reports and 0.13 in the full data set)[7]. A signal for tirzepatide does exist on its own terms: an earlier analysis through June 2025 identified 28 cases and a reporting odds ratio of 2.599[9], while one covering reports through September 2024 found tirzepatide significantly associated only with diabetic retinopathy in FAERS[8].

A ratio of 94 versus 3 does not mean semaglutide carries 30 times the risk. These databases have no count of how many people took each drug. Their mix of reporters also differs sharply: health professionals filed 5.04% of tirzepatide reports and 23.42% of semaglutide reports in one analysis[6]. Publicity matters too. The July 2024 semaglutide cohort prompted widespread media coverage, and 48% of 71 news articles studied contained misleading claims[10]. Attention of that kind can plausibly raise reporting for the named drug, though no study here measures that effect for NAION. Arslan states that all findings reflect spontaneous reporting rather than comparative clinical risk or causality[4].

A 2026 network meta-analysis of NAION across GLP-1 drugs does not help here. It reports semaglutide (HR 1.538, 95% CI 1.305 to 1.816) and liraglutide (HR 1.251, 95% CI 1.080 to 1.450) against non-GLP-1 treatment, and its abstract gives no tirzepatide estimate[11].

Absolute risk puts the comparison in scale

Whichever drug has the larger signal, the numbers involved are small, and the class-wide studies below do not break out tirzepatide in their abstracts. In a Swedish registry, anterior ischemic optic neuropathy affected 0.04% of GLP-1 users and 0.02% of SGLT-2 inhibitor users at 1 year (RR 1.93, 95% CI 1.00 to 3.73), and the gap shrank substantially among patients on metformin at baseline, leading the authors to suggest residual confounding[12]. In US commercial claims, 18-month risk was 8.5 versus 5.5 per 10,000 against SGLT-2 inhibitors, a number needed to harm of 3,333[13]. In UK primary care records the 1-year figures were 18.5 versus 7.2 per 100,000 against DPP-4 inhibitors (RR 2.56, 95% CI 1.44 to 4.86)[14].

Randomized data on the class are thinner and do not resolve it. A pooled analysis of Novo Nordisk’s liraglutide and semaglutide trials, with company employees among its authors, found about 3 confirmed cases per 100,000 participant-years on drug and about 6 on placebo[15], whereas EMA read clinical trial data as pointing to a slightly higher risk on semaglutide than placebo[17]. A separate meta-analysis of 20 trials reported an odds ratio of 1.50 (95% CI 0.49 to 4.63) for ischemic optic neuropathy, with the caveat that trials logged these events as adverse events and not as prespecified outcomes[16]. We could not find an equivalent pooled NAION analysis of the tirzepatide trials, so the randomized comparison between these two drugs remains missing.

How regulators treat each drug

Regulatory action so far names semaglutide. In June 2025 the European Medicines Agency’s safety committee concluded NAION is a very rare side effect of semaglutide, affecting up to 1 in 10,000 people, and estimated about one additional case per 10,000 person-years of treatment[17]. One European analysis notes that the tirzepatide product information does not specify optic ischemic neuropathy, even though it appears in tirzepatide reports (0.28%)[7]. In the US, neither the Zepbound label revised in August 2026[18] nor the Wegovy label revised in June 2026[19] mentions NAION or ischemic optic neuropathy. A warning on one drug and silence on the other may reflect how much evidence has accumulated and when, and it is not proof that tirzepatide is free of the problem. Case reports of NAION on tirzepatide exist[22]; single cases cannot show cause.

What may matter more than the molecule

Personal risk factors are better documented than any drug-to-drug difference. A 2026 review calls a small cup-to-disc ratio, the so-called disc at risk, the most robust anatomical predictor, and lists diabetes, hypertension, hyperlipidemia, sleep apnea and end-stage kidney disease as added susceptibility[21]. In the UK cohort, risk was higher in the first 6 months of use and in men, ever-smokers, people under 50 and those whose A1c fell by at least 1 point[14]. The US claims study found larger differences in men and in people aged 50 or older[13], so the age finding is inconsistent across data sets.

What to raise with a prescriber

The North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology describe the overall magnitude of risk as low and call for shared decision-making on whether to start, continue or stop[20]. The current evidence does not show that switching from semaglutide to tirzepatide lowers anyone’s NAION risk, because no study has followed switchers for this outcome in the abstracts we read. Useful questions for a prescriber or eye doctor:

  • If you have had NAION before, or have sleep apnea or a known crowded optic disc, does that change which drug makes sense? The 2026 review suggests screening for disc-at-risk anatomy and considering prior NAION before starting therapy may be warranted[21].
  • If you have diabetes, how fast should your A1c come down, given the signal tied to larger reductions[14]?
  • Do you need a baseline eye exam?

Sudden vision loss needs urgent attention

EMA advises people on semaglutide who experience sudden loss of vision or rapidly worsening eyesight to contact their doctor without delay[17]. The same symptom on tirzepatide deserves the same urgency. It is different from the transient blurred vision linked to blood-sugar shifts.

Frequently Asked Questions

It may be, but that is unproven. One large retrospective cohort in type 2 diabetes found a hazard ratio of 0.45 for tirzepatide versus GLP-1 drugs without GIP activity, and report databases show a much smaller signal for tirzepatide. No randomized trial has compared the two drugs on NAION, and report databases cannot measure risk.
Causation has not been established. Tirzepatide has its own signal in FDA adverse event reports, with a reporting odds ratio of about 2.6 to 2.9 across two analyses, and case reports exist. The US Zepbound label revised in August 2026 does not mention NAION. Any sudden vision loss on the drug needs prompt medical attention.
The evidence does not support switching for this reason alone, since no study in the abstracts reviewed here tracked NAION in people who switched, and absolute risk on semaglutide is low. European regulators estimate about one additional case per 10,000 person-years. Discuss your own risk factors with your prescriber first.
It is rare. European regulators classify it as very rare on semaglutide, affecting up to 1 in 10,000 people. Class-wide cohorts report figures such as 0.04% at 1 year in Sweden. The studies reviewed here gave no tirzepatide-specific absolute rate.

References

  1. 1.Hong AT, Lin F, Keenan JD, et al. Ocular Outcomes with Tirzepatide versus Glucagon-like Peptide-1 Receptor Agonists in Type 2 Diabetes. Ophthalmol Retina. 2026. PMID: 41655764.
  2. 2.Huang YN, Chen JC, Li PH, et al. Comparative ocular outcomes of tirzepatide versus other anti-obesity medications in people with obesity. Commun Med (Lond). 2025. PMID: 40750822.
  3. 3.Wang L, Volkow ND, Kaelber DC, et al. Semaglutide or Tirzepatide and Optic Nerve and Visual Pathway Disorders in Type 2 Diabetes. JAMA Netw Open. 2025. PMID: 40788646.
  4. 4.Arslan T. Comparative adverse-event reporting patterns of semaglutide and tirzepatide: a disproportionality analysis of FAERS. Endocrine. 2026. PMID: 42955519.
  5. 5.Kakkar AK, Payra S, Charaya A. Optic Ischaemic Neuropathy in Incretin-Based Therapy: A Comparative Analysis of Real-World Safety Data. Diabetes Obes Metab. 2026. PMID: 42259643.
  6. 6.Liu X, Wang W. Real-World Adverse Event Reporting Patterns of Tirzepatide Versus Semaglutide: A Stratified Study of General and Healthcare Professional Cohorts in FAERS. Diabetes Obes Metab. 2026. PMID: 42768828.
  7. 7.Popa Ilie IR, Ghibu S, Butuca A, et al. Incretin-Based Drugs for Obesity: Common and Drug-Specific Reporting Patterns of Adverse Drug Reactions-A Comparative Disproportionality Analysis Using EudraVigilance Reports Integrating SmPC Data. Pharmaceuticals (Basel). 2026. PMID: 42356493.
  8. 8.Lakhani M, Kwan ATH, Mihalache A, et al. Association of Glucagon-Like Peptide-1 Receptor Agonists With Optic Nerve and Retinal Adverse Events: A Population-Based Observational Study Across 180 Countries. Am J Ophthalmol. 2025. PMID: 40383360.
  9. 9.Castellana E, Chiappetta MR. Exploring the Potential Link Between Tirzepatide and Non-Arteritic Anterior Ischemic Optic Neuropathy (NAION): Evidence from FAERS and Google Trends. Hosp Pharm. 2026. PMID: 41509894.
  10. 10.Mahiny D, KamaliZonouzi S, Jafari S, Micieli JA. Media Coverage of Nonarteritic Anterior Ischemic Optic Neuropathy (NAION) After Semaglutide Use: A Content Analysis of News Reporting Accuracy and Framing. J Neuroophthalmol. 2026. PMID: 41870413.
  11. 11.Sun H, Xin X, Yan J, Huang T. Comparative Risk of Nonarteritic Anterior Ischemic Optic Neuropathy Among Glucagon-Like Peptide-1 Receptor Agonists: An Integrated Safety and Mechanistic Analysis. Diabetes Obes Metab. 2026. PMID: 42768849.
  12. 12.Ueda P, Svanström H, Söderling J, et al. Glucagon-like Peptide-1 Receptor Agonists and Risk for Anterior Ischemic Optic Neuropathy : A Nationwide Cohort Study. Ann Intern Med. 2026. PMID: 42441962.
  13. 13.Reynolds KR, O'Malley KM, Roy JA, Dave CV. Glucagon-like Peptide-1 Receptor Agonists and Risk for Ischemic Optic Neuropathy : A Target Trial Emulation. Ann Intern Med. 2026. PMID: 42441967.
  14. 14.Noh Y, Yin H, Ben Ghezala I, et al. Glucagon-Like Peptide 1 Receptor Agonists and Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients With Type 2 Diabetes. Diabetes Care. 2026. PMID: 41701611.
  15. 15.Vilsbøll T, Arun Kumar D, Aiello LP, et al. Non-arteritic anterior ischaemic optic neuropathy incidence in placebo-controlled clinical trials of liraglutide or semaglutide. Br J Ophthalmol. 2026. PMID: 42049287.
  16. 16.Li HY, Chan TK, Co Shih K, et al. GLP-1 Receptor Agonists and Risk of Optic Nerve or Vision-Threatening Events in Patients With Type 2 Diabetes or Cardiometabolic Diseases: A Meta-analysis of Randomized Controlled Trials. Diabetes Care. 2026. PMID: 41587563.
  17. 17.European Medicines Agency, PRAC. PRAC concludes eye condition NAION is a very rare side effect of semaglutide medicines Ozempic, Rybelsus and Wegovy. European Medicines Agency. 2025. https://www.ema.europa.eu/en/news/prac-concludes-eye-condition-naion-very-rare-side-effect-semaglutide-medicines-ozempic-rybelsus-wegovy
  18. 18.Eli Lilly and Company. ZEPBOUND (tirzepatide) injection prescribing information, revised August 2026. DailyMed. 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=487cd7e7-434c-4925-99fa-aa80b1cc776b
  19. 19.Novo Nordisk. WEGOVY (semaglutide) injection and tablets prescribing information, revised June 2026. DailyMed. 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=ee06186f-2aa3-4990-a760-757579d8f77b
  20. 20.DeParis SW, Oke I, Gaier ED, et al. Glucagon-Like Peptide-1 Receptor Agonists and the Risk of Non-Arteritic Anterior Ischemic Optic Neuropathy: A Consensus Statement by the North American Neuro-Ophthalmology Society and the American Academy of Ophthalmology. Ophthalmology. 2026. PMID: 42132708.
  21. 21.Ruan F, Gnana B, Mei C, et al. Current review of glucagon-like peptide 1 receptor agonists and risk of nonarteritic anterior ischemic optic neuropathy. J Endocr Soc. 2026. PMID: 42857401.
  22. 22.Valášková J, Bondarchuk D, Kadlic P, et al. Acute Unilateral NAION following Tirzepatide Therapy: A Case Report. Case Rep Ophthalmol. 2026. PMID: 42569780.

GLP-1 and Glaucoma: NAION Signal, IOP, and Dry Eye Evidence

The Hathaway 2024 JAMA Ophthalmology study flagged a NAION signal with semaglutide. We review the actual data, the glaucoma + IOP literature, the dry-eye pattern during dose escalation, and the ophthalmology follow-up protocol.

11 min read

GLP-1, Ozempic & Vision Loss: The NAION Evidence

Does Ozempic cause blindness? Studies link semaglutide to NAION (sudden one-eye vision loss) — relative risk is real (HR ~1.8–8) but absolute risk is low (≤1 in 10,000 per EMA). Causality unproven; diabetes itself harms the optic nerve. Know the one red-flag symptom.

10 min read

GLP-1 Allergic & Injection-Site Reactions (2026)

GLP-1 injection-site reactions (redness, itching, nodules) are usually benign: 1.4% in Wegovy, 6-8% in Zepbound trials. True allergy is rare. Benign vs ER.

10 min read

GLP-1 and Anesthesia Aspiration: The Evidence (2026)

What the case reports, gastric-ultrasound cohorts, and 2024 multisociety guidance actually show about GLP-1 drugs and pulmonary aspiration under anesthesia.

13 min read

GLP-1 and Cosmetic Surgery: Hold & Recovery (2026)

GLP-1s before cosmetic surgery: the real aspiration/hold question (~4.5× retained gastric contents) vs the theoretical wound-healing worry from lean-mass loss.

11 min read

GLP-1 and Dental Sedation: Do You Stop First? (2026)

Do you need to hold Ozempic, Wegovy, Mounjaro or Zepbound before dental IV or deep sedation? What the ASA and 2024 multisociety guidance say about aspiration risk.

10 min read

Where to get tirzepatide (Mounjaro / Zepbound) safely: vetted online providers

Vetted telehealth providers that prescribe online. We compare pricing, form, and states served.

No insurance needed · vetted by our editors

WeightLossRankings.org is reader-supported. When you buy through links on our site, we may earn an affiliate commission. How we make money

7.1

AgelessRx

Longevity-focused GLP-1 care with microdosing and brand-name options

Read full review →

From $209/mo (2-month plan rate)

Get started →
7.2

HumeCare+

GLP-1 weight loss with real-time body-composition tracking to protect muscle

Read full review →

From $299/mo

Get started →
8.3

Collective

The lowest flat medication price if you can absorb an annual membership fee

Read full review →

From $69/mo (+ required membership)

Get started →