GLP-1 weight loss drugs have long been linked to irksome effects on the stomach. But users have also begun reporting other strange symptoms, such as hair loss, detaching fingernails, dental changes and even a rare eye condition that can lead to permanent vision loss.
Are GLP-1 drugs actually causing this constellation of problems? The question highlights a critical challenge for pharma: identifying rare safety effects that may not show up until a drug is widely used and then determining what’s to blame.
In the case of the rare eye condition — non-arteritic anterior ischemic optic neuropathy or eye stroke — the FDA found that there may be increased risk associated with GLP-1 drugs. On Nov. 9, it required GLP-1 makers to add a NAION warning to their labels.
NAION occurs when blood flow to the optic nerve is disrupted, and only affects an estimated 2.3 to 10.2 patients out of 100,000 over age 50 each year, according to the FDA.
Risk factors for the condition are also common among GLP-1 users, including high blood pressure, diabetes and sleep apnea. Those connections made it a challenge for FDA officials to determine what was driving higher rates of the condition, which is also linked to other medications, including erectile dysfunction drugs like Viagra and Cialis, and the heart drug amiodarone.
But for other reported side effects the picture is even muddier. Hair loss is already listed as an adverse reaction on Wegovy and Zepbound’s labels but the role of the drugs themselves in reported hair, nail and dental changes can be difficult to isolate.
These effects may be due to factors such as nutritional deficiencies, self-reported survey data or even the process of rapid weight loss itself, which can complicate research findings. One published review found no evidence that semaglutide directly affects teeth, but suggested potential indirect effects from GLP-1 usage — such as dry mouth, nausea or vomiting — could be playing a role.
The NAION decision shows how the FDA could approach untangling these questions.
Investigating side effects
Clinical trials prior to drug approval generally pick up on the most common drug side effects, but are often too small to detect rarer effects, defined as those that occur in anywhere from 1 in 1,000 to 1 in every 10,000 people. Very rare side effects affect fewer than 1 in every 10,000 patients, according to drugs.com.
Research studies often don’t enroll enough participants to see those side effects before approval. A phase 3 trial, for example, typically includes 300 to 3,000 patients — large enough to assess whether the drug works, but often too small to reliably detect an adverse effect that occurs in 1 in every 10,000 patients. But those effects can show up when a drug rolls out to the masses.
FDA approval doesn’t end a drugmaker’s responsibility to ensure safety. Companies are required to continue tracking adverse events and to report potential problems to the FDA. The agency also collects and tracks safety reports through its new Adverse Event Monitoring System, which consolidated formerly separate databases for drugs, devices and other products, and through the MedWatch system.
The FDA launched an extensive investigation into NAION after a July 2024 study in JAMA Ophthalmology suggested an association. The regulator looked at clinical trial results, observational and animal studies, post-marketing data and patient health records. The main challenge for the FDA was determining whether patients who took GLP-1s were really at higher risk for the disorder or if the condition was actually driven by independent factors that are common among GLP-1 users.
An analysis of 126 clinical trials found a slight potential increased risk of the condition, 6.4 cases per 100,000 person-years among GLP-1 users versus 3.8 among people taking placebo or a non-GLP-1 diabetes drug, although the overall number of cases was very small.
In a separate real-world analysis of type 2 diabetes, there was no statistically significant increase in NAION among GLP-1 users compared with people taking another type of diabetes drug, SGLT-2 inhibitors. There was, however, an increased risk in GLP-1 users when compared to those taking a third type of diabetes drug, DPP-4 inhibitors.
The FDA ultimately concluded there is a potential increased risk, even though it’s still unclear why the drugs might have that effect.
A familiar challenge for FDA
GLP-1s aren’t the first medications to face new safety concerns after approval. The pain drug Vioxx and the statin Baycol were both pulled after serious risks emerged, including heart-related effects and a potentially fatal breakdown of muscle tissue.
In other instances, however, post-marketing approval monitoring hasn’t confirmed suspected safety concerns.
The human papillomavirus vaccine Gardasil, for example, has been subjected to intense post-approval monitoring amid reports of potential serious effects to the FDA’s safety monitoring system. So far, agencies including the FDA and CDC haven’t found any links to unexpected safety issues. But the surveillance did find that people sometimes faint after receiving the injection.
As GLP-1 uptake continues to rise, reports of new symptoms likely will too. But sorting out whether it’s the drug or other factors causing them will pose an ongoing challenge.