Do GLP-1 drugs cause blindness? What does the FDA warning mean?

Do GLP-1 Drugs Cause Blindness? What Does the FDA's New Warning Really Mean?

First of all, what is NAION disease, which is mentioned in the FDA's new warning?

NAION, or non-arteritic anterior ischemic optic neuropathy, is a disease that results from reduced blood flow to the anterior portion of the optic nerve, potentially causing sudden and usually painless vision loss. In some patients, the vision loss may be permanent (FDA, 2026). However, there is an important point here.

NAION is not a newly discovered disease. It is a disease that we have known about long before GLP-1 drugs became available, and it has been described in medical literature for a long time.

We know that the risk is higher in people with type 2 diabetes, obese individuals, those with hypertension, smokers, those with sleep apnea, and those with certain vascular diseases. Studies also show a link between obesity and metabolic syndrome and NAION (Liu et al., 2021; Kohli et al., 2022).

The point I want to make is that a significant portion of patients using GLP-1 medications already have diabetes, obesity, or cardiovascular risk factors. This patient group is therefore already more susceptible to NAION. So, a patient developing NAION while using GLP-1 doesn't necessarily mean the disease was caused by the medication.

To make this distinction, we need to look at scientific data. The FDA's review, published on November 9, 2026, examined 126 randomized clinical trials involving a total of 134,253 participants and covering a follow-up period of 288,879 person-years. Do you know how many potential NAION cases were observed in total?

A total of 15.

Ten of these cases were observed in individuals receiving GLP-1 treatment, and five in the comparison group (FDA, 2026). Now, how do we interpret these numbers? First, we need to consider what other risk factors these individuals might have had. Did they have hypertension? Did they smoke? How long did they have diabetes? Did they have sleep apnea? We cannot learn all the individual risk characteristics of these 15 patients from the published FDA summary. Randomized trials are designed to reduce such differences, but it is not easy to definitively isolate all possible effects in an analysis with only 15 cases. We know that individual risks can vary significantly in such small case numbers. So, for example, were all of these 15 cases smokers? Were all of them both diabetic and hypertensive? We don't know.

Therefore, this data points to a potential increase in risk, however There is no definitive proof that GLP-1 drugs directly cause NAION.

Here, instead of focusing on sensationalist headlines generated by social media for clickbait, we need to look at scientific facts. In the comparison group not using GLP-1, there were 3.8 NAION cases per 100,000 people per year, while in the GLP-1 group, this figure was 6.4. This means that per 100,000 people per year of follow-up, we are talking about approximately 2.6 additional NAION cases compared to the comparison group (FDA, 2026).

On the other hand, in April 2026, another study involving approximately 97,000 participants was published in the British Journal of Ophthalmology. Researchers examined the results of randomized, placebo-controlled clinical trials using semaglutide and liraglutidine, the active ingredients in drugs like Ozempic and Saxenda (Vilsbøll et al., 2026). And guess what they found?

In approximately 65,000 people using GLP-1, only three confirmed cases of NAION were observed, while in approximately 32,000 people using placebo, a total of five NAION events were detected in four patients.

Based on follow-up periods, the GLP-1 group experienced approximately three cases per 100,000 people per year, compared to approximately six in the placebo group. Therefore, this study showed fewer NAION cases among people using GLP-1. In fact, if you only look at the numbers from this study, you might even think that GLP-1 drugs could be protective against NAION. But that's not how science is done.

Because it is not correct to say that the drug is protective based on a study in which only a few cases were observed. In this study, all patients in whom the diagnosis of NAION was confirmed had at least one of the known risk factors such as advanced age, obesity, hypertension, smoking, or high cholesterol (Vilsbøll et al., 2026).

This reminds us once again that NAION is already associated with certain risk factors.

Of course, since this study only included eight confirmed NAION events, we cannot definitively say that GLP-1 drugs either reduce or increase the risk. However, similarly, it is not scientifically accurate to pick one study and ignore another and say "GLP-1 drugs cause blindness."

So the results of the studies and the estimates regarding the magnitude of the risk are not entirely consistent.

In short, what we know today is that NAION is a very rare disease, but one that can have serious consequences. Its possible association with GLP-1 drugs is still being investigated. The additional absolute risk observed in the combined data from current randomized trials is quite small.

Now let's assume the difference the FDA found in these studies is entirely accurate, and that weight-loss injections are indeed responsible for these additional 2.6 cases per 100,000 person-years. What does this mean to you?

We're talking about an additional absolute NION risk of approximately 0.0026 percent per year. So, if you are obese, have type 2 diabetes, and hypertension, what are the long-term risks these conditions pose to your health? Your risk of premature death? Your risk of heart attack? Your risk of stroke?

We have known for years that these risks are far more widespread and serious. According to a large meta-analysis published in The Lancet in 2016, which evaluated data from more than 10 million people, compared to normal-weight individuals, the all-cause mortality rate was 45% higher in people with first-degree obesity, 94% higher in those with second-degree obesity, and 176% higher in those with third-degree obesity (Global BMI Mortality Collaboration, 2016). These are not absolute probabilities of death, but relative increases in risk. However, they show how serious the risk of death can become as the degree of obesity increases. In other words, the mortality rate in severely obese individuals is almost three times higher than in normal-weight individuals. Moreover, according to another large study, life expectancy was approximately 2 to 4 years shorter in individuals with a BMI between 30 and 35, and approximately 8 to 10 years shorter in those with a BMI between 40 and 45 (Prospective Studies Collaboration, 2009).

So, are we only using these medications to help people lose weight?

No.

The SELECT study, published in the New England Journal of Medicine in 2023, evaluated 17,604 individuals with pre-existing cardiovascular disease and who were overweight or obese. Over approximately 3.3 years of follow-up, the composite endpoint of heart attack, stroke, or cardiovascular death decreased from 8% in the placebo group to 6.5% in those receiving semaglutide. This represents a 20% relative risk reduction and a 1.5 percentage point absolute risk reduction for serious cardiovascular events (Lincoff et al., 2023).

In 2024 Alzheimer's & Dementia A large study published in the journal showed that the use of semaglutide in patients with type 2 diabetes was associated with a lower risk of developing Alzheimer's disease. The study evaluated data from more than 1 million eligible patients and found that those using semaglutide had a lower risk of receiving their first Alzheimer's diagnosis compared to other diabetes medications (Wang et al., 2024). This is a very important development because the potential effects of GLP-1 drugs on neuroinflammation, metabolic disorders, and neurodegenerative processes in the brain have long been investigated. However, it should also be noted that two large randomized clinical trials published in 2026 failed to demonstrate that semaglutide slowed the progression of Alzheimer's disease in individuals who had already developed the disease (Cummings et al., 2026). Therefore, the possibility of reducing the risk of developing Alzheimer's is not the same as treating existing Alzheimer's disease. The possibility of a protective effect is still worth investigating and is a promising development. There is also the issue of inflammation.

In 2026 Circulation A sub-analysis of the SELECT study published in the journal showed that individuals using semaglutide experienced approximately a 38% reduction in high-sensitivity C-reactive protein (hsCRP) levels, a key indicator of inflammation, after 104 weeks compared to placebo (Plutzky et al., 2026). Furthermore, this reduction began before significant weight loss and was observed in some patients who did not experience significant weight loss. Therefore, there is evidence suggesting that the effects of these drugs on inflammation may not be explained solely by weight loss.

We also have quite strong data regarding stroke, or paralysis. In 2024... International Journal of Stroke A meta-analysis of 11 randomized clinical trials involving more than 82,000 participants, published in the journal, showed that GLP-1 drugs reduced the risk of stroke by approximately 15 percent and the risk of non-fatal stroke by 13 percent (Adamou et al., 2024).

And recently, some very interesting research on breast cancer has begun to be published. June 2026 JCO Oncology Practice A study published in the journal examined data from more than 111,000 overweight or obese women. It found that women using GLP-1 had approximately 30 percent less chance of developing breast cancer than those who did not. Furthermore, this association persisted even after accounting for factors such as age, body mass index, breast density, and diabetes (McDonald et al., 2026).

So on one hand, we're talking about approximately 2.6 additional NAION cases per 100,000 people per year, which are thought to be drug-related but may not be. On the other hand, there are the much more common, cumulative, and potentially fatal complications of obesity, type 2 diabetes, and cardiovascular disease. So, When discussing the potential side effects of a drug, we cannot ignore the serious health problems that the drug might prevent.

In medicine, no treatment is completely risk-free. What matters is how significant that risk is, how accurately it is known, and how it compares to the benefits of the treatment. The health risks posed by obesity and uncontrolled type 2 diabetes are considerable and well-documented (Global BMI Mortality Collaboration, 2016).

Before resorting to sensationalist headlines to grab a few extra clicks on social media, it's important to consider the consequences of unnecessarily scaring people who genuinely need these treatments and could reap significant health benefits. What patients need is accurate and balanced scientific information, not frightening headlines.

Let's stop the fearmongering and start talking about the numbers.

Dr. Ahmet ÖzyiğitMD, FAAMM, ABAARM
Longevity Physician

References:

1. Adamou, A., Barkas, F., Milionis, H. and Ntaios, G. (2024) 'Glucagon-like peptide-1 receptor agonists and stroke: A systematic review and meta-analysis of cardiovascular outcome trials', International Journal of Stroke, 19(8), pp. 876–887.
DOI: 10.1177/17474930241253988.

2. Cummings, JL et al. (2026) 'Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): two phase 3, randomized, placebo-controlled trials', The Lancet, 407(10544), pp. 2167–2179. DOI: 10.1016/S0140-6736(26)00459-9.

3. Food and Drug Administration (FDA) (2026) 'FDA requires safety labeling changes for glucagon-like peptide-1 receptor agonists (GLP-1 RAs) due to potential increased risk of rare but serious eye condition', FDA Drug Safety Communication, 9 November. Available at: FDA Official Safety Communication (Accessed: 13 November 2026).

4. Global BMI Mortality Collaboration (2016) 'Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents', The Lancet, 388(10046), pp. 776–786. DOI: 10.1016/S0140-6736(16)30175-1.

5. Kohli, D., Wu, KY, White, LJ, Hodge, DO, Chen, JJ and Roddy, GW (2022) 'Metabolic syndrome and its components are associated with non-arteritic anterior ischaemic optic neuropathy', BMJ Open Ophthalmology, 7(1), e001111. DOI: 10.1136/bmjophth-2022-001111.

6. Lincoff, AM et al. (2023) 'Semaglutide and cardiovascular outcomes in obesity without diabetes', New England Journal of Medicine, 389(24), pp. 2221–2232. DOI: 10.1056/NEJMoa2307563.

7. Liu, B., Yu, Y., Liu, W., Deng, T. and Xiang, D. (2021) 'Risk factors for non-arteritic anterior ischemic optic neuropathy: A large scale meta-analysis', Frontiers in Medicine, 8, 618353. DOI: 10.3389/fmed.2021.618353.

8. McDonald, ES et al. (2026) 'GLP-1 agonists are associated with a significant reduction in breast cancer incidence in women', JCO Oncology Practice, 22(7), pp. 1341–1348. DOI: 10.1200/OP-26-00485.

9. Plutzky, J. et al. (2026) 'Effect of semaglutide on the inflammatory biomarker high-sensitivity CRP in patients with established cardiovascular disease and overweight or obesity in SELECT: A prespecified secondary analysis', Circulation, 154(11), pp. 976–991. DOI: 10.1161/CIRCULATIONAHA.125.074482.

10. Prospective Studies Collaboration (2009) 'Body-mass index and cause-specific mortality in 900 000 adults: Collaborative analyzes of 57 prospective studies', The Lancet, 373(9669), pp. 1083–1096. DOI: 10.1016/S0140-6736(09)60318-4.

11. Vilsbøll, T., Arun Kumar, D., Aiello, L.P., Rönnbäck, C., Stellfeld, M. and Larsen, L.M. (2026) 'Non-arteritic anterior ischaemic optic neuropathy incidence in placebo-controlled clinical trials of liraglutide or semaglutide', British Journal of Ophthalmology, published online 28 April. DOI: 10.1136/bjo-2025-328729.

12. Wang, W., Wang, Q., Qi, X., Gurney, M., Perry, G., Volkow, ND, Davis, PB, Kaelber, DC and Xu, R. (2024) 'Associations of semaglutide with first-time diagnosis of Alzheimer's disease in patients with type 2 diabetes: Target trial emulation using nationwide real-world data in the US', Alzheimer's & Dementia, 20(12), pp. 8661–8672. DOI: 10.1002/alz.14313