GLP-1 drugs and blindness: why no one knows how often it happens

Nobody can yet say how often GLP-1 drugs are followed by blindness or pancreatitis, because the main tracking system relies on unverified, mostly voluntary reports. That gap sits at the heart of litigation brought by patients such as Todd Engel, whose doctor prescribed Ozempic for his diabetes in early 2023.

One patient, two eyes

On New Year’s Eve that winter, Engel woke up feeling as if a patch covered his right eye. “I couldn’t see anything. It was all dark. It was black,” he recalled. His doctor diagnosed NAION, a form of optic neuropathy that cannot be operated on, in which tiny strokes starve the eye of blood. When the doctor listed the known causes, his wife, Shelley Engel, asked specifically about Ozempic. She recalls the doctor answering, “We have heard nothing about that.”

Seven months after that first episode, the journal JAMA Ophthalmology published research finding higher NAION rates among diabetic patients on semaglutide, the active ingredient in Ozempic. Neither Engel nor his doctor knew about it, so the weekly injections continued until his left eye also went dark.

Engel is now suing Novo Nordisk, arguing the company did not warn patients and doctors about NAION. Shelley Engel says more side-effect warnings might have helped her husband avoid blindness. The stakes reach well beyond one household: a recent Gallup analysis found more than one in 10 Americans currently take a GLP-1 for weight loss, and one in seven has used them at some point.

Drugmakers, regulators and plaintiffs disagree

Novo Nordisk told NPR that later research casts doubt on a causal link with NAION, that the cases lack merit, and that NAION and other serious conditions were exceedingly rare in clinical trials. Eli Lilly, which sells tirzepatide as Mounjaro and Zepbound, said by email it would vigorously defend against the claims. The FDA has requested no warning-label changes; it says it is running its own study of NAION risk and will decide on regulatory action once that assessment is finished.

Attorney Jonathan Orent leads the court-appointed legal team in the two major multi-district suits. He argues that drugmakers should now update their warnings to doctors and patients, as has already happened in Europe, even though the conditions are rare. In his view, trials are structurally limited: “you’re never going to have the population size to detect rarer events in a clinical trial; you’re also never going to have the diversity of your real world population.” He says his caseload covers 247 plaintiffs with NAION, plus over 4,000 others with serious digestive illnesses such as pancreatitis or gallstones.

Two eye-care groups, the American Academy of Ophthalmology and the North American Neuro-Ophthalmology Society, issued a consumer clarification about the JAMA Ophthalmology study, saying it is not the type of study that can show the treatment caused NAION. They acknowledge a potential link that needs more research and do not advise patients to stop taking their medication.

What FDA reports can and cannot show

Those counts come from an NPR analysis of the FDA’s adverse events database, as reported by WJCT. Most of the nearly 190,000 reports concerned minor digestive trouble, dosing mistakes or redness where the shot went in. The database is one of the few channels open to doctors, patients and companies, but filing is largely optional and the agency does not check submissions, so a tally of reports is not a measure of frequency.

Some cases never get filed. Mohammed Ali, a diabetes specialist at Emory University, treated three patients who developed problems like pancreatitis or gallstones on GLP-1s and reported none, because he could not be sure the drugs were responsible despite the suspicious timing. “They may have had subclinical disease,” he said.

Why cause is so hard to prove

Observational data can reveal that a complication appears more often among users, not that the drug produced it; the courts need that harder answer. Obesity and diabetes complicate it further, since both are metabolic diseases that strain many organs and raise the odds of other illnesses on their own.

Photo: Pavel Danilyuk / Pexels