Skip to content
GLP Loss
← Research
Safety

GLP-1s and Dental Work: What the Sedation Warning Covers

Every labeled warning in this class is scoped to general anesthesia or deep sedation, and almost no dentistry reaches either. At the depth dental offices actually use, oxygen desaturation did not rise — the odds of needing an antiemetic nearly doubled.

Hana Brennan9 min read
The warning names two depthsMost dental work sits below both of themLocal anesthetic onlyOutside every label warning and every study hereNitrous oxide or oral anxiolysisMinimal sedation, also outside the labeled scopeModerate sedationTwo studies here: no airway signal, more nauseaDeep sedation or general anesthesiaThe only depth the labeled warning addressesUnder mild-to-moderate sedation across 45,636 visits,desaturation did not rise but antiemetic use nearly doubled.No label in this class mentions a dentist anywhere.

The medical history form at a dental office now asks about these drugs, and the answer often produces a phone call and a rescheduled appointment. The warning behind that call is real, it is printed on every approved label, and it is scoped to a depth of anesthesia that most dentistry never reaches. Sorting out which procedures the warning covers is most of the work, and it is worth doing before a filling gets postponed for a month. The wider operating-room version of this question is in the surgery article.

What the labels say, and what they say it about

The section is titled Pulmonary Aspiration During General Anesthesia or Deep Sedation. Its text describes rare postmarketing reports of pulmonary aspiration in patients receiving GLP-1 receptor agonists undergoing elective surgeries or procedures requiring general anesthesia or deep sedation, who had residual gastric contents despite reported adherence to preoperative fasting recommendations.[1][2]

Then the sentence that governs the rest of this page. Available data are insufficient to inform recommendations to mitigate the risk, including whether modifying preoperative fasting recommendations or temporarily discontinuing the drug would reduce the incidence of retained gastric contents.[1][2] The single instruction either label gives a patient is to tell health care providers about any planned surgery or procedure.

Two things follow. The warning is bounded by an anesthetic depth, and it stops short of telling anyone to stop the drug. A sweep of all four approved labels finds no mention of a dentist, a dental procedure or an oral finding anywhere in any of them.

Most dentistry is below the line that warning draws

Sedation is a continuum with named steps. A restoration, a cleaning, a crown preparation, a root canal and the great majority of extractions are done under local anesthetic with the patient fully awake, protective airway reflexes intact and no fasting requirement. Nitrous oxide and oral anxiolysis sit one step up, at minimal sedation. Neither is general anesthesia and neither is deep sedation.

Deep sedation and general anesthesia in dentistry are real but narrow: oral and maxillofacial surgery, full-mouth rehabilitation, treatment for patients with severe dental phobia or significant special needs, and much pediatric operative dentistry. Those appointments are the ones the labeled warning is written about, and they are the ones where preoperative fasting instructions already exist for other reasons.

A 2025 review written for dental anesthesia providers reaches the same boundary from the other side. It walks through the 2024 multisociety perioperative guidance, and what it takes from it is risk stratification of the individual patient, weighing the risks of holding against continuing, and shared decision making between the sedation provider, the prescribing physician and the patient — rather than a blanket rule.[3]

What the multisociety guidance actually stratifies on

That document’s position is that therapy may be continued before a procedure in patients without elevated risk of delayed gastric emptying and aspiration, and it defines elevated risk rather than leaving it to instinct: the escalation phase rather than maintenance, a higher dose, weekly rather than daily dosing, gastrointestinal symptoms, and conditions that delay gastric emptying independently of the drug.[4]

Three of those five are things a patient can report accurately and a dentist cannot look up: which molecule, what dose, how recently it went up, and whether nausea or reflux is present this week. That is the practical content of the conversation, and it is worth having at the booking rather than in the chair.

One depth down, holding the dose made a large difference

The strongest evidence in this area is now randomized, and it comes from endoscopy rather than dentistry. The OCULUS trial randomized adults on a stable GLP-1 or GLP-1/GIP agonist undergoing elective upper endoscopy under moderate sedation or monitored anesthesia care either to continue the drug or to hold one dose.[5]

It was stopped early. At the preplanned interim analysis of 60 patients, clinically significant residual gastric volume — a composite of contents that prevented examination, forced termination or intubation, or produced an aspiration event — occurred in 3.1% of those holding a dose against 25.0% of those continuing, an absolute difference of 21.9 percentage points (90% CI, 7.0 to 36.7; P = .003). In the endoscopy-only subgroup the split was 5.0% against 46.7% (P = .001). Continuing did not increase other adverse events.[5]

The subgroup that complicates it is the most useful part. Among the 25 patients having endoscopy plus colonoscopy — who were on clear liquids the day before — no patient in either arm had clinically significant residual gastric volume. The authors conclude that clear liquids the day prior may mitigate the risk regardless of GLP-1 use.[5] The diet preparation, not the drug hold, erased the problem in that group.

At the depth dentistry uses, the numbers point elsewhere

Two studies have looked at mild-to-moderate sedation specifically, and both used ophthalmic surgery, the commonest outpatient procedure done at that depth.

The larger is a retrospective cohort of 45,636 visits from 30,328 patients between 2022 and 2024, of which 1,596 visits were by GLP-1 users. After adjustment, GLP-1 use was not associated with oxygen desaturation below 90%, at an odds ratio of 1.22 (95% CI, 0.38 to 3.88; P = 0.74). It was associated with postoperative nausea or vomiting requiring an antiemetic, at an odds ratio of 1.93 (95% CI, 1.29 to 2.89; P = 0.001).[6]

That is the finding worth carrying into a dental chair. The outcome everyone is bracing for did not move, with a confidence interval spanning 1.0 in both directions. The outcome nobody warns about — needing something for nausea afterward — nearly doubled. It is an observational comparison and the usual confounding applies, but the direction of the two results is the opposite of the one the pre-op phone call implies.

A smaller retrospective review supports the airway half. Among 155 surgical records of patients actively taking semaglutide presenting for elective eye surgery, 151 received moderate sedation and 120 had taken a dose within one week of the procedure. There were no cases of aspiration, no significant respiratory events and no significant change in postoperative oxygen saturation.[7] A series with no events cannot estimate a rate, and a study designed to find nothing is easy to publish. Read alongside the 45,636-visit cohort, it is consistent rather than decisive.

The dental literature is reviews, not studies

Nobody has run a study in a dental operatory. What exists is guidance written by dental faculties reading the medical evidence across. A 2026 review for dental practitioners itemizes what a patient on semaglutide may present with — xerostomia, halitosis, enamel erosion tied to vomiting, and altered taste — and frames all of them as secondary manifestations of systemic effects rather than direct oral toxicity.[8]

Enamel erosion is the one on that list a dentist can act on independently. It is a mechanical consequence of acid reaching teeth, which is why it belongs with reflux and vomiting rather than with the drug, and both are covered in the reflux article. The incidence figures behind dry mouth, and the surprisingly weak chain that links reduced saliva to decay, are in the dry mouth article rather than repeated here. Taste change has its own labeled numbers in the taste article.

Gums and implants: the direction is favorable and the evidence is thin

The periodontal question runs the other way from the sedation one, and it is easy to overstate. A 2025 review assembling everything published on GLP-1 receptor agonists and periodontal or peri-implant health found 10 in vitro studies, nine animal studies and one clinical study. The laboratory and animal work is consistently positive: reduced periodontal inflammation, less alveolar bone resorption, improved implant osseointegration in diabetic models.[9]

The one clinical study is a retrospective cohort comparing peri-implant marginal bone loss across glucose-lowering drug classes, which found significantly less clinical and radiographic bone loss in the GLP-1 group than in the insulin and metformin groups (P < 0.01).[9] The review’s own conclusion is that the role of these drugs looks more like preserving periodontal and peri-implant health in type 2 diabetes than treating periodontitis. A 2026 periodontology review reaches the same place in its own words: human data are limited and mostly observational, with confounding by metabolic status, smoking and nutrition, and GLP-1 therapy should be regarded as a contextual modifier of periodontal risk rather than a therapy.[10] Nine in ten of those studies were not done in people, and all of the clinical signal is in diabetes.

What to tell a dentist, and when

Five facts do the work, and all five are things only the patient has: which molecule, what dose, when the last dose was given, whether the dose has changed in the last few weeks, and whether nausea, reflux or vomiting is present now. Those map directly onto the risk factors the multisociety document names, which means a dentist who has them can stratify and a dentist who does not cannot.[4]

The sixth is the depth of the planned anesthesia, and the patient is entitled to ask. A procedure under local anesthetic raises none of the questions above. A procedure under deep sedation or general anesthesia raises all of them, and whether to hold a dose for it is a decision for the sedation provider and the prescribing clinician together. One thing the OCULUS result does support is asking about the pre-procedure diet instruction as well as the drug.[5]

Why a compounded prescription makes that harder

Every item on that list assumes the patient can name what they are taking. A compounded vial often arrives labeled by a pharmacy rather than under a brand name, at a concentration that is not on the container, with a dose expressed as a volume or a number of units rather than a milligram figure. Compounded drugs are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed, and no reviewed insert accompanies them for a dental office to consult.

That is a practical reason to favor a seller that names the molecule, the dose in milligrams and the dispensing pharmacy, which is recorded in the seller write-ups against the criteria in the methodology. What changes about the numbers when a person moves from a pen to a vial is in the switching article.

The summary a reader can use is short. A routine dental appointment is not the procedure the labeled warning is about. The evidence at moderate sedation does not show an airway problem and does show more post-procedure nausea. One depth further down, holding a dose mattered a great deal, and so did the day-before diet. Which of those situations applies is a question about the appointment rather than about the drug, and the dentist is the one who knows the answer.

Frequently asked

Do I need to stop my GLP-1 before a dental appointment?
The labeled warning is scoped to elective procedures requiring general anesthesia or deep sedation, which almost no routine dentistry involves. Both labels also state that available data are insufficient to recommend whether discontinuing the drug would reduce retained gastric contents. Whether to hold a dose before a procedure under deep sedation is a decision for the sedation provider and the prescribing clinician, not a default.
Is sedation dentistry safe on semaglutide or tirzepatide?
The evidence at moderate sedation is reassuring on the airway and less so on nausea. A retrospective cohort of 45,636 ambulatory visits under mild-to-moderate sedation found no association between GLP-1 use and oxygen desaturation, at an odds ratio of 1.22 with a confidence interval of 0.38 to 3.88, but a near-doubling of postoperative nausea or vomiting requiring an antiemetic, at 1.93 with a confidence interval of 1.29 to 2.89.
Does holding one dose before a procedure help?
At endoscopy depth it mattered a great deal. The OCULUS randomized trial was stopped early after clinically significant residual gastric volume occurred in 3.1% of patients who held one dose against 25.0% who continued. Its most instructive subgroup was the one on clear liquids the day before, in which no patient in either arm had a clinically significant volume.
Do these drugs cause cavities?
No study has measured caries incidence on a GLP-1. The plausible routes are indirect: reduced saliva, and acid erosion from vomiting or reflux rather than from the drug itself. Dental reviews written for practitioners list xerostomia, halitosis, enamel erosion and altered taste as secondary manifestations of systemic effects rather than direct oral toxicity.
What should a dentist be told?
Which molecule, what dose, when the last dose was taken, whether the dose has changed recently, and whether nausea or reflux is present. Those map onto the risk factors the 2024 multisociety guidance uses to stratify patients, and none of them is visible on a chart. It is also fair to ask what depth of anesthesia the appointment will use, because that is what decides whether any of it applies.

Sources

  1. [1] Novo Nordisk Pharmaceutical Industries, LP (2026). WEGOVY (semaglutide) injection and tablets — Pulmonary Aspiration During General Anesthesia or Deep Sedation DailyMed, U.S. National Library of Medicine. Source
  2. [2] Eli Lilly and Company (2026). ZEPBOUND (tirzepatide) injection, solution — Pulmonary Aspiration During General Anesthesia or Deep Sedation DailyMed, U.S. National Library of Medicine. Source
  3. [3] McKenzie C, DeBernardo A, Schwartz P (2025). Implications of GLP-1 Agonists on Office-Based Sedation and General Anesthesia for Dentistry. Anesth Prog. PMID 40657828
  4. [4] Kindel TL, Wang AY, Wadhwa A, et al. (2024). Multisociety clinical practice guidance for the safe use of glucagon-like peptide-1 receptor agonists in the perioperative period. Surg Obes Relat Dis. PMID 39482213
  5. [5] Ahmad AI, Garg S, Jacobs J, et al. (2026). Holding vs Continuing GLP-1/GIP Agonists Before Upper Endoscopy: The OCULUS Randomized Clinical Trial. JAMA Intern Med. PMID 41837981
  6. [6] Hyung B, Ye XY, Chandrakumar M, et al. (2026). Perioperative Desaturation and Postoperative Nausea or Vomiting Requiring Antiemetic Administration in Glucagon-Like Peptide-1 Receptor Agonist Users Undergoing Ambulatory Ophthalmic Surgery Under Mild-To-Moderate Sedation: A Large Retrospective Cohort Study. Diabetes Obes Metab. PMID 42259623
  7. [7] Zapp C, Downer Ii DM, Levine A, et al. (2026). The Risk of Aspiration Is Low With Continuing Semaglutide During Elective Eye Surgery When Patients Receive Only Moderate Sedation. Cureus. PMID 41798554
  8. [8] Kofman K, Ouanounou A (2026). Oral Health Considerations and Dental Management Guidelines for Semaglutide Medications. Can J Diabetes. PMID 41967795
  9. [9] Ahmad P, Estrin N, Farshidfar N, et al. (2025). Glucagon-Like Peptide 1 Receptor Agonists (GLP-1RAs) Improve Periodontal and Peri-Implant Health in Type 2 Diabetes Mellitus. J Periodontal Res. PMID 40348599
  10. [10] Sufaru IG, Vasiliu BC, Hancianu M, et al. (2026). GLP-1 Receptor Agonists in Periodontology: Mechanisms, Clinical Evidence, and Implications for Care. Biomolecules. PMID 42352323

More in Safety