Across 226,485 people taking GLP-1 drugs, the mouth problems that went viral on social media mostly did not show up.
A retrospective study published online June 9 in Oral Surgery Oral Medicine Oral Pathology Oral Radiology ↗ set out to test "Ozempic teeth," "Ozempic mouth," and "Ozempic tongue," the shorthand that spread across social feeds and dental waiting rooms for a cluster of complaints (dry mouth, tooth decay, mouth sores, altered taste) blamed on the weight-loss and diabetes injections. Stella Oyewole and Adepitan Owosho pulled the TriNetX research network, matched GLP-1 receptor agonist users one to one against people not on the drugs by age and sex, and excluded anyone with prior oral conditions. Then they counted orofacial diagnoses across four groups: mucosal, salivary, neuropathic, and dental.
For most of those conditions, the drug users were not more likely to have a problem. They were less likely. The study found significant inverse associations across multiple mucosal, salivary, neuropathic, and dental disorders, with odds ratios running from 0.295 to 0.893 (a range that means roughly 70 percent lower to 11 percent lower odds than the matched comparison group). On this measure, being on a GLP-1 drug tracked with fewer mouth diagnoses, not more.
One condition broke the pattern. Gastroesophageal reflux disease, the acid reflux that pushes stomach contents up toward the throat, was both the most common event (5.36 percent of users) and the only one with significantly higher odds: an odds ratio of 1.329 (95 percent confidence interval 1.291 to 1.367), about a third more likely than in controls. That is the one signal a dentist has real reason to watch, because reflux is also the most plausible route to any genuine dental harm here. Stomach acid that reaches the mouth erodes enamel over time. If GLP-1 drugs raise reflux, some of the tooth damage people blame on the drug directly may be the acid taking the long way around.
GLP-1 receptor agonists ↗ are the class that includes semaglutide ↗, sold as Ozempic and Wegovy, and they work by mimicking a gut hormone that slows stomach emptying and blunts appetite. Slower stomach emptying is exactly the mechanism that can push acid back up, which makes the reflux finding mechanistically coherent rather than a fluke of the database. The dry-mouth worry has a plausible story too (these drugs cut food and fluid intake, and less chewing means less saliva flow), but at the scale of a quarter-million users it did not translate into a measurable excess of salivary diagnoses.
The study is observational, and the authors say so. Propensity matching balances age and sex and excludes people with pre-existing oral disease, but it cannot rule out that GLP-1 users differ from non-users in ways a claims database never records: how often they see a dentist, income, how carefully symptoms get coded. Inverse associations that strong (odds cut by half or more for some conditions) are more likely to reflect who ends up on these drugs and who shows up for care than a protective effect of the molecule on the gums. The honest read is narrower than either the viral panic or a clean bill of health. The big harms people fear are not visible in the record, and the one that is, reflux, is the one worth monitoring.
For a section of the audience that heard "Ozempic teeth" before it heard "GLP-1 receptor," that distinction matters. The mouth is not falling apart. The acid is worth a conversation with a dentist.