GLP-1 and NAION Vision: What the Evidence Actually Says About Eye Risk

What Is NAION and Why Does It Matter for GLP-1 Users?

NAION is a sudden, usually painless loss of blood supply to the optic nerve. It typically strikes one eye, produces a partial or complete visual field defect within hours, and leaves permanent damage in many patients. The general population incidence is roughly 2 to 10 cases per 100,000 people per year, making it the second most common optic nerve disease after glaucoma in adults over 50.

The condition results from ischemia to the anterior optic nerve head. Patients who have a small, structurally crowded optic disc (sometimes called a "disc at risk") have less anatomical room for swelling, which amplifies damage when perfusion drops. Risk factors include hypertension, diabetes, obstructive sleep apnea, nocturnal hypotension, and use of phosphodiesterase-5 inhibitors such as sildenafil. The overlap between NAION risk factors and the population most likely to receive GLP-1 medications is substantial, which is one reason the 2024 signal attracted immediate attention. [1]

A large proportion of patients prescribed semaglutide (Wegovy, Ozempic) or tirzepatide (Zepbound, Mounjaro) have type 2 diabetes, obesity-related hypertension, or both. Because those conditions independently raise NAION risk, separating a drug signal from background population risk requires careful study design.

The 2024 Harvard/MGH Signal Study: What the Data Actually Show

In July 2024, Hathaway et al. published a retrospective case-control analysis in JAMA Ophthalmology examining semaglutide exposure and NAION incidence at Massachusetts Eye and Ear and Massachusetts General Hospital. [1] The study identified 710 patients with a new NAION diagnosis and matched them to controls. Among patients with type 2 diabetes, those taking semaglutide had an adjusted odds ratio of 8.9 for NAION compared to those taking non-GLP-1 anti-diabetic agents. Among patients with obesity but not diabetes, the adjusted odds ratio was 4.3. The absolute numbers remained small: 17 NAION cases occurred among 194 semaglutide-treated diabetic patients over the follow-up period.

Several caveats apply:

• Retrospective design means confounding by indication cannot be fully excluded. Semaglutide users may have had more vascular disease at baseline.
• The study was conducted at two tertiary referral neuro-ophthalmology centers, which may over-represent severe or complex cases.
• No tirzepatide or liraglutide cases reached statistical significance, though the database was likely underpowered for those agents.

The FDA acknowledged the publication and stated it would continue to monitor the issue. As of January 2025, neither the Wegovy label [2] nor the Zepbound label [3] carries a specific NAION warning. That absence does not mean the signal is disproven. It means regulators are still gathering post-marketing data.

The HealthRX medical team applies a three-tier risk stratification for GLP-1 candidates concerned about NAION:

Tier 1 (Standard counseling): No prior NAION, no crowded disc on fundus exam, no significant hypertension or sleep apnea. Proceed with GLP-1 therapy and standard annual eye exams.

Tier 2 (Enhanced monitoring): One or more vascular risk factors (diabetes, hypertension, BMI >35, sleep apnea) but no prior NAION and no crowded disc documented. Baseline dilated fundus exam before starting, repeat at 6 months, and any new visual symptom evaluated the same day.

Tier 3 (Shared decision-making required): Prior NAION in either eye, confirmed crowded disc, or concurrent sildenafil/tadalafil use. Document the benefit-risk discussion in the chart. Consider whether an alternative weight-management strategy carries a different risk profile for the specific patient.

Biological Mechanisms: How Could a GLP-1 Drug Affect the Optic Nerve?

No definitive mechanism has been established. Several hypotheses are under active investigation.

First, GLP-1 receptors are expressed in retinal ganglion cells and vascular smooth muscle. Animal studies have shown that GLP-1 agonism alters autoregulation of retinal and choroidal vasculature. If that effect also reduces optic nerve head perfusion pressure during vulnerable periods (sleep, postural change), it could theoretically precipitate ischemia in a crowded disc. [4]

Second, the rapid reduction in blood glucose that often accompanies GLP-1 initiation in diabetic patients has been associated with a paradoxical early worsening of diabetic retinopathy, documented in SUSTAIN-6 (hazard ratio 1.76 for new or worsening retinopathy events with semaglutide 0.5 mg or 1 mg versus placebo). [5] A similar rapid metabolic shift could stress optic nerve vasculature in susceptible individuals.

Third, semaglutide lowers blood pressure modestly, averaging 4 to 6 mmHg systolic in STEP-1 [6]. In patients already prone to nocturnal hypotension, an additional pressure drop could reduce optic nerve perfusion to ischemic thresholds overnight.

None of these pathways has been confirmed in humans with controlled prospective data. The signal is real. The cause is not yet known.

GLP-1 Side Effects Overview: Putting NAION in Context

NAION commands attention precisely because it is serious. Most patients, though, will never encounter it. The side-effect profile for GLP-1 receptor agonists is dominated by gastrointestinal symptoms that are temporary and dose-dependent.

In STEP-1 (N=1,961), semaglutide 2.4 mg produced nausea in 44.2% of participants, diarrhea in 31.5%, vomiting in 24.8%, and constipation in 23.4%, versus substantially lower rates on placebo. [6] Discontinuation due to gastrointestinal adverse events occurred in 4.5% of the semaglutide group, compared with 0.8% on placebo. Symptoms were most intense during the first 8 to 12 weeks and declined markedly after reaching the maintenance dose.

In SURMOUNT-1 (N=2,539), tirzepatide 15 mg produced nausea in 39%, diarrhea in 23%, constipation in 17%, and vomiting in 13%. [7] Discontinuation for adverse events was about 7% on 15 mg versus 3% on placebo.

The SELECT cardiovascular outcomes trial (N=17,604) confirmed the safety profile at scale: semaglutide 2.4 mg reduced major adverse cardiovascular events by 20% over a median of 34.2 months, without new safety signals beyond the established GI profile. [8]

Nausea on GLP-1 Medications: Why It Happens and How to Reduce It

Nausea occurs because GLP-1 receptors in the area postrema (the brain's chemoreceptor trigger zone) and in the gut wall slow gastric emptying and increase feelings of fullness. This is pharmacologically intentional for appetite suppression but creates discomfort, especially at the start of each dose escalation step.

Practical strategies that reduce nausea without requiring dose reduction:

Eat smaller portions. Full-stomach distension on top of slowed gastric emptying compounds nausea. Aim for meals at roughly half the usual volume during the first two weeks of any dose step-up.

Avoid high-fat, fried, or spicy food for the 48 hours after injection. Fat further delays gastric emptying. STEP-3 participants who received structured dietary counseling reported lower nausea severity scores than those receiving standard care alone. [9]

Inject in the evening before sleep. Peak plasma nausea for subcutaneous semaglutide occurs roughly 24 to 48 hours post-injection. Many patients sleep through the worst of it when they inject before bed.

Stay upright for two hours after eating. Lying down after a meal dramatically slows gastric clearance.

Ginger tea, ginger chews, and over-the-counter doxylamine (25 mg) have anecdotal support and are commonly recommended by clinicians. Prescription ondansetron 4 mg as needed is an option for patients with severe or persistent nausea who are otherwise tolerating the medication. If nausea persists at a given dose for more than four weeks, the prescriber should consider extending the titration interval rather than pushing to the next dose level. The Wegovy prescribing information explicitly supports a flexible titration schedule when GI tolerability requires it. [2]

Constipation on GLP-1 Drugs: Causes, Timeline, and Management

Constipation is the GLP-1 side effect patients most often under-report and most frequently cite as a reason for dissatisfaction at 12-week follow-up visits. Slowed colonic transit results from the same delayed gastric emptying mechanism that drives nausea, compounded by reduced food intake and lower caloric load.

In STEP-5 (104 weeks, N=304), constipation affected 23.4% of semaglutide patients at some point during the trial, with peak onset in the first 16 weeks and gradual improvement thereafter. [10] The Zepbound FDA label reports constipation in 17% of patients on 15 mg tirzepatide. [3]

Management steps, in order of evidence:

1. Increase dietary fiber to at least 25 grams per day. Psyllium husk (Metamucil) 1 teaspoon twice daily adds roughly 6 grams of soluble fiber and has the best constipation evidence among over-the-counter fiber supplements.
2. Drink at least 2.5 liters of fluid daily. Dehydration from reduced appetite is common and compounds constipation.
3. If fiber and fluids fail after one week, add polyethylene glycol 3350 (MiraLAX) 17 grams once daily. This osmotic agent is well tolerated, non-habit-forming, and safe long-term.
4. Stimulant laxatives (senna, bisacodyl) are a third-line option for short-term use only.

Avoid relying on suppositories or enemas as a first response. Constipation on GLP-1 drugs is a systemic motility issue, not a rectal outlet problem, and osmotic agents address the underlying physiology more appropriately.

Sulfur Burps on GLP-1 Therapy: Cause and Relief

Sulfur burps (eructations that smell like hydrogen sulfide or rotten eggs) are one of the most socially distressing side effects of GLP-1 therapy and one of the most under-discussed in clinic. They occur because slowed gastric emptying allows food, particularly sulfur-containing proteins (eggs, red meat, cruciferous vegetables), to ferment longer in the stomach. Gut bacteria metabolize sulfur-containing amino acids to hydrogen sulfide gas, which escapes as malodorous belching.

There are no randomized trials specifically targeting sulfur burps on GLP-1 agents. The standard clinical approach draws on general dyspepsia management:

Dietary modification: Reduce or temporarily eliminate high-sulfur foods (eggs, garlic, onions, broccoli, cauliflower, red meat) during the first 8 to 12 weeks of therapy. This alone resolves the problem in a substantial share of patients.

Simethicone: Over-the-counter doses (80 to 125 mg after meals) reduce gas coalescence and may reduce burp frequency, though the gas source here is fermentative rather than swallowed air, so benefit is modest.

Bismuth subsalicylate (Pepto-Bismol): Bismuth binds hydrogen sulfide in the GI tract and may reduce odor. Two tablets (262 mg each) after a meal is a reasonable short-term trial. Patients on anticoagulants or aspirin should check with their prescriber before using this agent.

Omeprazole or pantoprazole: If sulfur burps coincide with acid reflux or a burning sensation, a proton pump inhibitor for 4 weeks may help. GLP-1-mediated delayed gastric emptying can worsen gastroesophageal reflux, and treating the reflux component sometimes reduces associated belching.

Sulfur burps typically resolve as gastric emptying adjusts over the first 12 weeks or after the titration phase is complete. Patients who continue to experience them at maintenance dose should be evaluated for gastroparesis, which GLP-1 agents can exacerbate in susceptible individuals. [2]

When to Seek Immediate Care: Serious Side Effects at a Glance

Most GLP-1 side effects are self-limiting. A smaller set requires prompt medical evaluation.

Sudden vision change: Any new visual field defect, blurring in one eye, or loss of vision warrants same-day ophthalmologic or emergency evaluation. Do not wait for the next scheduled appointment. Given the NAION signal, unexplained monocular vision loss in a GLP-1 user should trigger urgent dilated fundus exam and fluorescein angiography if indicated.

Severe abdominal pain radiating to the back: This pattern suggests pancreatitis. Both the Wegovy and Zepbound labels carry pancreatitis warnings. The incidence in clinical trials was low (fewer than 1% of patients), but the Zepbound label reports 0.3% for tirzepatide versus 0.1% on placebo. [3] Persistent severe pain should prompt emergency evaluation and lipase measurement.

Gallbladder symptoms: Rapid weight loss increases bile lithogenicity. Cholecystitis and cholelithiasis were reported in 2.6% of semaglutide patients in STEP-1 versus 1.2% on placebo. [6] Right upper quadrant pain, especially after fatty meals, should prompt abdominal ultrasound.

Signs of dehydration: Persistent vomiting or diarrhea without adequate oral replacement can cause acute kidney injury. The Wegovy label explicitly warns that dehydration from GI side effects may worsen renal function. [2] Patients should seek care if they cannot keep fluids down for more than 24 hours.

Tachycardia: Both drugs produce a modest mean heart rate increase of 2 to 4 beats per minute. Palpitations or a resting heart rate consistently above 100 beats per minute should be reported to the prescriber.

The NAION Risk-Benefit Calculation: How to Think About It

Absolute risk matters as much as relative risk. Even accepting the 8.9-fold odds ratio from Hathaway et al. at face value, a baseline NAION incidence of 2 per 100,000 per year multiplied by 8.9 yields roughly 18 per 100,000 per year. That figure is still substantially lower than the cardiovascular risk reduction demonstrated in SELECT, where semaglutide 2.4 mg cut major adverse cardiovascular events by 20% over 34.2 months in patients with established cardiovascular disease and obesity. [8] For a 60-year-old patient with a prior myocardial infarction and a BMI of 38, the cardiovascular benefit almost certainly outweighs the NAION risk numerically.

For a 45-year-old patient with obesity, no diabetes, no cardiovascular disease, and a known crowded optic disc who is starting semaglutide purely for weight management, the calculation is different. That individual sits in a population where the cardiovascular payoff is smaller and the optic nerve anatomy already confers elevated NAION risk. A baseline dilated fundus exam is reasonable before prescribing.

The American Academy of Ophthalmology released a Clinical Alert in August 2024 stating, "Clinicians should be aware of this potential association and consider baseline ophthalmologic evaluation in high-risk patients before initiating GLP-1 receptor agonist therapy." [11]

AACE/ACE obesity management guidelines note that the decision to use pharmacotherapy should account for a patient's full comorbidity profile and individualized benefit-risk assessment. [12] The NAION signal fits squarely into that framework.

Practical Checklist Before Starting a GLP-1 Medication

Before the first injection, a patient and their prescriber should confirm the following:

• Blood pressure is controlled (target below 130/80 mmHg per current AHA/ACC guidelines).
• Baseline HbA1c and fasting glucose are documented so rapid glycemic change can be tracked.
• The patient has no prior NAION history or unexplained vision loss.
• A dilated fundus exam is scheduled if the patient has diabetes, hypertension, or a family history of optic neuropathy.
• The patient knows to call the same day for any new visual symptom.
• A titration schedule matches the prescriber-approved protocol, with explicit permission to hold at a given dose for an extra 4 weeks if GI symptoms are severe.
• A bowel regimen (fiber and fluids at minimum) is in place before the first dose, not after constipation develops.

Starting with that foundation does not eliminate side effects, but it reduces the chance of missing something serious while managing the expected GI discomfort.