People on GLP-1 drugs were about half again as likely to develop a smell or taste problem as similar patients on other diabetes medicines. That is the headline number from a large records study published this week in JAMA Otolaryngology-Head and Neck Surgery ↗, and the more specific finding is that smell took the bigger hit.
The drugs in question are the GLP-1 receptor agonists, the class sold as Ozempic and Wegovy (semaglutide ↗), Victoza and Saxenda (liraglutide ↗), and Trulicity (dulaglutide ↗). They work by copying a gut hormone called GLP-1 that tells the brain you are full and the pancreas to release insulin. Nausea, vomiting, and other gut complaints are the well-known costs. A dulled sense of smell is not one anyone counsels patients about.
What the study found
Researchers used the TriNetX Global Collaborative Network, a database of de-identified electronic health records, and pulled adults with type 2 diabetes between December 2017 and April 2026. They split them into people who started a GLP-1 drug and people who took other diabetes drugs instead. Then they matched the two groups on age, sex, and a long list of clinical and socioeconomic factors, to make the comparison as close to apples-to-apples as records allow. Each group ended up with 438,474 patients, nearly 877,000 in total.
Over a follow-up that ran from three months to two years, the GLP-1 group had a 48 percent higher rate of a new smell or taste disturbance (hazard ratio 1.48, 95 percent confidence interval 1.37 to 1.61). The risk was about 1.5 times the comparison group's. Pulled apart, smell ran ahead of taste. Smell problems were about 81 percent more common (hazard ratio 1.81, interval 1.58 to 2.07). Taste problems were 52 percent more common (hazard ratio 1.52, interval 1.35 to 1.71). The gap held steady across the follow-up rather than spiking early and fading.
Why smell more than taste
The study does not explain the mechanism, and its authors are careful to say so. Smell and taste are wired differently. A drug that changes appetite and gut behavior has several routes to either sense. It could act directly, through receptors in the nose and on the tongue. Or it could act indirectly, through the nausea, dehydration, and sharply reduced food intake the drugs bring. Sorting cause from consequence is exactly the work a records study cannot do.
What it can do is flag a signal big enough to chase. An 81 percent relative increase in smell complaints across nearly half a million treated patients is not noise. Smell loss is also badly under-recorded in routine care, because most people who notice food tastes flat never get a diagnostic code for it, which means the true rate is probably higher than the records show, not lower.
What it does and does not change
This is an association, not proof that the drugs cause the problem. The matching reduces confounding but cannot erase it, and the outcomes are whatever a clinician happened to code, not the result of formal smell and taste testing. Nobody should stop a GLP-1 drug over this.
What changes is the counseling conversation. The GLP-1 receptor that these drugs hit, GLP-1R ↗, is the same target behind their benefits on weight, blood sugar, and increasingly the heart, and the same one being engineered into the next wave of obesity peptides. A side effect that shows up across the whole class is worth naming to patients before they notice their coffee stopped smelling like anything, and worth a prospective study that actually measures smell instead of waiting for someone to complain.