Do GLP-1 Drugs Affect Your Other Medications? What the Labels Report

Four labels say slowed stomach emptying could affect oral medicines. Between them they carry interaction studies on eight drugs. That gap between the general caution and the short tested list is where almost every question about this lands.

One finding runs the opposite way to what people expect. Oral semaglutide raises levothyroxine exposure by a third instead of lowering it, and the reason has nothing to do with gastric emptying.

Do GLP-1 drugs affect other medications?

Labels answer yes, in both directions, for a small number of drugs actually studied. Section 7.2 of the Mounjaro label says "Caution should be exercised when oral medications are concomitantly administered with MOUNJARO."¹ Trulicity states that it "delays gastric emptying and thus has the potential to reduce the rate of absorption" of oral medicines taken at the same time.²

Wegovy's pharmacokinetics section uses the same hedge, noting that "The delay of gastric emptying with semaglutide may influence the absorption of concomitantly administered oral medications."³ The words potential and may are doing visible work in all three sentences. The same mechanism is why an anesthesia team wants to know about these drugs before sedation.

Behind each caution, the tested list is short. Each document lists drug interaction studies covering between one and five named medicines per drug. Everything else a person takes falls outside what was tested.

Which drugs were actually tested with tirzepatide?

Two drugs were tested with tirzepatide. The prescribing information reports interaction studies on acetaminophen and on a combined oral contraceptive containing 0.035 mg ethinyl estradiol with 0.25 mg norgestimate.¹ No other medicine is named in that section's interaction discussion. Two drugs is a small evidence base for a caution addressed to oral medicines in general.

Mounjaro's warning reaches further than its data. It advises caution with oral medicines generally and singles out drugs with a narrow therapeutic index for monitoring, giving warfarin as its example.¹ Warfarin was not among the two drugs studied, so the monitoring example named there has no data behind it.

So the tirzepatide position is an instruction built on a mechanism, with no paired data behind it. That is a defensible way to write a warning and a poor basis for anybody to conclude their own combination is fine.

What happened to acetaminophen on a first dose?

Acetaminophen peak levels halved on a first dose, and then the effect went away. With a single 5 mg tirzepatide dose, the first a participant received, acetaminophen peak concentration fell by 50% and the time to peak was delayed by about an hour.¹ Those are large numbers.

Timing is the whole story. When acetaminophen was given again at week 4, section 12.3 reports no meaningful effect on peak concentration or time to peak, and overall exposure was not influenced.¹ Same drug pair, same measurement, opposite result four weeks apart.

That pattern matches what tirzepatide's prescribing information says about the gastric delay being largest after the first dose. For anyone reading an interaction warning, a single question has two answers depending on where somebody sits in their dosing history. Our guide to tirzepatide and birth control covers the one instruction the label builds on that fact.

Which drugs were tested with injectable semaglutide?

Five, and they include the two that worry prescribers most. The tested set is metformin, an oral contraceptive containing ethinylestradiol and levonorgestrel, warfarin, digoxin and atorvastatin.³ Both of the narrow therapeutic index examples the 2026 review names, warfarin and digoxin, sit in that set.⁴

The reported result is flat. "No clinically relevant drug-drug interactions with semaglutide (Figure 3) were observed based on the evaluated medications," the label states.³ Warfarin and digoxin both sit inside that list, so the two narrow therapeutic index drugs people ask about were examined and came back unremarkable.

Read side by side, the two sit oddly: the drug carrying the stronger oral medication caution studied two medicines, and the drug carrying the weaker one studied five, including the risky pair.

Why does oral semaglutide behave differently?

The formulation is the reason, not the molecule. Semaglutide tablets contain an absorption enhancer called SNAC. SNAC is what allows a peptide to survive the stomach at all, and a 2026 review attributes the tablet's interaction behavior to that enhancer rather than to gastric emptying.⁴

Consequences follow for co-administered drugs. An agent that changes how the stomach handles absorption does not confine that effect to the drug it was added for, so the tablet's interaction profile is its own.

Two products with the same active ingredient therefore carry different interaction findings. Our comparison of the two GLP-1 pills covers how the oral products differ from each other. This page is about what either of them does to the rest of a medicine cabinet.

Does oral semaglutide raise levothyroxine exposure?

Yes, by a third, and the direction is up. Section 7.2 of the semaglutide tablet label states that "Levothyroxine exposure was increased 33% (90% CI: 1.25 to 1.42)" when given with the tablets.⁵ Section 12.3 reports the same finding as total thyroxine, "adjusted for endogenous levels," with AUC "increased by 33%" after a single 600 mcg levothyroxine dose.⁵

Peak concentration did not move. Section 7.2 also records that "Maximum exposure (Cmax) was unchanged," so the extra exposure accumulated over the curve instead of arriving as a spike.⁵

A 2026 review reports the same 33% figure and draws the practical conclusion, "suggesting that thyroid parameters should be monitored."⁴ Monitoring is a prescriber's decision, and our guide to talking to your doctor covers raising it.

What about injectable semaglutide and levothyroxine?

No measured figure exists for that pair. Levothyroxine does not appear among the five medicines Wegovy's interaction studies covered, so the injectable label has nothing on it in either direction.³

Absence of a figure is not a finding. Anybody reading the oral semaglutide number and applying it to an injection is carrying a result across a formulation difference the review attributes to the tablet's absorption enhancer.⁴ Carrying it the other way, and assuming the injection must be safe because the tablet result came from SNAC, assumes the gastric mechanism does nothing.

Both readings run past the evidence. What the documents support is narrower: one formulation has a measured levothyroxine figure and the other has not been studied with it.

Which drugs were tested with dulaglutide?

The dulaglutide label does not name them in its text. Trulicity names only sitagliptin, reporting that co-administration raised dulaglutide exposure by about 38% and peak concentration by about 27%, a shift called "not considered clinically relevant."² That result describes what sitagliptin did to dulaglutide.

For the reverse direction the document is less forthcoming. It states that dulaglutide "did not affect the absorption of the tested orally administered medications to any clinically relevant degree" without naming them in the text.² The list sits in a figure instead.

A conclusion with an unnamed denominator is hard to use. Somebody wanting to know whether their own medicine was among those tested cannot answer it from the sentence that reports the result.

Is there a second kind of interaction besides absorption?

A second kind exists, and it is the one carrying a dose instruction. Everything above concerns absorption, meaning how much of a swallowed drug reaches the blood. A separate interaction acts on blood sugar directly, and the language there is firmer.

On the Mounjaro label, section 7.1 advises prescribers to "consider reducing the dose of concomitantly administered insulin secretagogues," because the combination raises hypoglycemia risk.¹ The semaglutide tablet carries the same instruction for a sulfonylurea or insulin, noting that "Semaglutide stimulates insulin release in the presence of elevated blood glucose concentrations."⁵

Two features separate that from the absorption findings. It comes with a dose action instead of a monitoring suggestion, and it does not depend on what else somebody swallowed or when. Anybody taking insulin or a sulfonylurea alongside a GLP-1 drug is in the one interaction category treated as settled.

What is a narrow therapeutic index drug, and why does it matter here?

Narrow therapeutic index describes a medicine where the gap between too little and too much is small, so a modest change in exposure can cross a line. The 2026 review puts it directly: "These changes may have clinical consequences for medicines with a narrow therapeutic index," and it names digoxin, warfarin and oral contraceptives as the examples.⁴

The wording points the same way. An exact-phrase check confirms that narrow therapeutic index appears in Mounjaro's drug interactions section, where warfarin is the worked example and monitoring is the advice.¹ The semaglutide tablet names the same category, covering other oral drugs with a narrow index or a monitoring requirement.⁵

Our guide to GLP-1 side effects covers what happens to the person, not to their other medicines. Narrow-index drugs are where a 20% or 30% shift stops being academic. Most medicines absorb a change of that size into normal variation. A few do not.

Is warfarin affected?

Warfarin was studied with semaglutide and not with tirzepatide. That split is uncomfortable. Wegovy's interaction studies included warfarin and reported no clinically relevant interaction across the medicines evaluated.³ Mounjaro names warfarin as its monitoring example without reporting a tirzepatide and warfarin study.¹

So the drug with the warning has no data on the pair, and the drug with the data has no warning. Neither of those is a contradiction, since a label can advise caution on mechanism alone. It does mean the two documents cannot be compared as if they said the same kind of thing.

The 2026 review groups warfarin with digoxin and oral contraceptives as drugs where exposure changes could matter.⁴ Nothing in these sources reports a warfarin outcome on either drug.

Are there interactions outside the label lists?

Two such interactions appear in the 2026 review, at different evidence levels. Its overall judgment is that "in most cases, these changes are not of major clinical relevance."⁴ Against that baseline it describes the semaglutide and alectinib interaction as clinically relevant.⁴ The same review cites a case series in which lithium levels rose to toxic levels after semaglutide was started.⁴

One of those two rests on a single case series. A case series is a collection of reports and not a controlled comparison, so it establishes that something happened in those patients, without saying how often it happens.

Those two examples carry a general point about those lists. A drug absent from an interaction study is untested rather than cleared, and the alectinib and lithium reports are what untested territory looks like when somebody eventually looks.

Is birth control affected?

Yes, for tirzepatide, and no other product here carries a comparable instruction. Our guide to tirzepatide and birth control sets out the section numbers, the measured reductions and the four-week windows in full.

In the interaction-study context, the combined oral contraceptive was one of only two medicines tirzepatide studied. The reductions it found were large enough to generate a labeled instruction.¹ Semaglutide's interaction studies also covered an oral contraceptive and produced no comparable instruction.³ Same drug category, two products, one instruction between them.

The 2026 review lists oral contraceptives among its narrow therapeutic index examples, alongside digoxin and warfarin.⁴ A pregnancy is a binary outcome. That is part of why a modest exposure change gets treated seriously there.

Does the timing of a pill matter?

Only the oral product prescribes a separation interval. Patients taking the tablet are told to "wait at least 30 minutes before eating food, drinking beverages" or taking other oral medications.⁵ That is an explicit separation requirement from other medicines, written into the administration instructions instead of the interactions section.

None of the three injectables here carries an equivalent. Mounjaro, Wegovy and Trulicity advise caution or monitoring for oral medicines without naming an interval.¹²³ None tells a patient to space a pill away from an injection by a set time.

Timing still changes the measurement on the injectable side. Acetaminophen peak concentration fell by 50% alongside a single 5 mg first tirzepatide dose, and showed no meaningful effect when given again at week 4.¹ The same drug pair produced two different results four weeks apart. What that implies for any other medicine has not been studied.

Nothing on this page is advice about when to take anything. The two instructions above are addressed to the people they name, and sequencing a specific combination belongs with whoever prescribes both.

What happens when two labels both want an empty stomach?

Oral semaglutide and levothyroxine make competing claims on the same morning slot, and each one wants an empty stomach. The semaglutide tablet must be taken "on an empty stomach in the morning with water," specifically "up to 4 ounces of water," with no other liquids.⁵ A wait of at least 30 minutes then follows before food, drink or other oral medicines.⁵

Levothyroxine asks for the same conditions and more water. Its prescribing information directs a single daily dose "on an empty stomach, one-half to one hour before breakfast," taken "with a full glass of water as the tablet may rapidly disintegrate."⁶ Up to four ounces and a full glass are not the same quantity, and neither instruction leaves room for the other.

Separation language already exists on the levothyroxine side. Levothyroxine's own instructions call for administration "at least 4 hours before or after drugs known to interfere with" its absorption, and they name a four-hour gap from calcium carbonate and ferrous sulfate.⁶ Whether a GLP-1 drug belongs in that category is not something either label answers.

So the practical collision is documented from both directions while the resolution is not. One label puts itself first by 30 minutes and the other wants a clear stomach for up to an hour, and no document cited here reconciles them.

What has not been measured?

Almost nothing in an ordinary medicine cabinet has been measured. Across four labels the interaction studies name eight medicines in total, and dulaglutide's tested list is not named in its text at all.¹³⁵² Antidepressants, thyroid medicine on the injectable products, antibiotics, immunosuppressants and most cardiac drugs do not appear.

Outcomes are the second gap. Every figure on this page is an exposure measurement, a change in how much drug reached the blood. No source cited here reports whether that translated into a clinical event for any pair.¹²³⁴⁵

Dose escalation is the third. Tirzepatide's acetaminophen result differed between a first dose and week 4, and no label says what happens to a co-administered drug when the dose steps up again later.¹

What would settle it?

Paired studies at the points where exposure actually changes would settle it. Nobody has run them. A useful design would take each narrow therapeutic index drug people commonly take and measure it on a first GLP-1 dose, at steady state and after an escalation. Outcomes would be reported alongside exposure.

Formulation would need its own arm. The levothyroxine figure belongs to a tablet with an absorption enhancer, and whether an injection does anything to levothyroxine is unstudied, so one experiment cannot answer both.⁵⁴

Until that exists, the summary is short. Four labels agree the mechanism could matter. One formulation has a measured 33% increase that runs opposite to the usual assumption, and five drugs came back unremarkable with injectable semaglutide. The rest is a conversation with a prescriber, since no figure covers it.

Common questions

Do GLP-1 drugs affect other medications?

Does Ozempic or Mounjaro affect levothyroxine?

Why would a GLP-1 drug increase a medicine's absorption?

Which drugs were tested with tirzepatide?

Does tirzepatide affect acetaminophen?

Is warfarin affected by GLP-1 drugs?

Do GLP-1 drugs interact with insulin or sulfonylureas?

Which GLP-1 interaction do the labels treat most firmly?

What is a narrow therapeutic index drug?

Which drugs were tested with injectable semaglutide?

Which drugs were tested with dulaglutide?

Should I separate my pills from my injection?

Can I take my levothyroxine and oral semaglutide at the same time?

Does any GLP-1 label give a separation interval from other pills?

Are there interactions the labels do not list?

Does my antidepressant interact with a GLP-1 drug?

Does the interaction get worse when my dose goes up?

Has anyone shown a GLP-1 interaction causing harm?

Sources

1. Mounjaro (tirzepatide) injection, US prescribing information, sections 7.1, 7.2 and 12.3. Read at source via the openFDA drug label API (https://api.fda.gov/drug/label.json), drug_interactions and pharmacokinetics fields. Quoted for "Caution should be exercised when oral medications are concomitantly administered with MOUNJARO.", the advice to consider reducing the dose of concomitantly administered insulin secretagogues, the acetaminophen interaction study reporting a 50% peak concentration reduction and about a one hour delay to peak with a single 5 mg first dose and no meaningful effect when coadministered at week 4, and the combined oral contraceptive study. The presence of the phrase narrow therapeutic index in the drug_interactions field was confirmed by exact-phrase query returning seven records under this brand name; the surrounding sentence was not read verbatim and is described here, not quoted. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18 ↩19 ↩20 ↩21 ↩22 ↩23 ↩24 ↩25 ↩26 ↩27
2. Trulicity (dulaglutide) injection, US prescribing information, sections 7.1 and 12.3. Read at source via the openFDA drug label API, drug_interactions and pharmacokinetics fields. Quoted for the statement that the drug "delays gastric emptying and thus has the potential to reduce the rate of absorption of concomitantly administered oral medications", the sitagliptin interaction study reporting roughly a 38% increase in dulaglutide exposure and 27% in peak concentration described as "not considered clinically relevant", and the statement that dulaglutide "did not affect the absorption of the tested orally administered medications to any clinically relevant degree". The medicines tested are shown in a figure and are not named in the label text. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11
3. Wegovy (semaglutide) injection, US prescribing information, section 12.3. Read at source via the openFDA drug label API, pharmacokinetics field. Quoted for "The delay of gastric emptying with semaglutide may influence the absorption of concomitantly administered oral medications", the list of medicines studied for interactions (metformin, an oral contraceptive containing ethinylestradiol and levonorgestrel, warfarin, digoxin and atorvastatin), and "No clinically relevant drug-drug interactions with semaglutide (Figure 3) were observed based on the evaluated medications." The label reports per-drug exposure ratios in a figure that was not readable, so no individual exposure figure is quoted here. Levothyroxine does not appear in this label's interaction study list. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17
4. Zbârcea CE, Marineci CD, Văleanu A, Chiriță C, Șeremet OC. Drug-Drug and Drug-Disease Interactions Across Antidiabetic Drug Classes: A Narrative Review and Practical Recommendations. Diabetology 2026;7(6):108. DOI 10.3390/diabetology7060108. Read at source. Quoted for "GLP-1 RAs slow gastric emptying, which delays the movement of food and oral medications from the stomach", "These changes may have clinical consequences for medicines with a narrow therapeutic index", "levothyroxine exposure increased by about 33%, suggesting that thyroid parameters should be monitored", "in most cases, these changes are not of major clinical relevance", the attribution of the oral semaglutide levothyroxine finding to the SNAC absorption enhancer, the description of the semaglutide and alectinib interaction as clinically relevant, and the cited case series reporting lithium levels rising to toxic levels after semaglutide was started. This is a narrative review across antidiabetic drug classes and not a GLP-1 interaction study. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18
5. RYBELSUS (semaglutide) tablets and OZEMPIC (semaglutide) tablets, for oral use, US prescribing information, sections 7.1, 7.2 and 12.3. Read at source via DailyMed. Quoted for "Levothyroxine exposure was increased 33% (90% CI: 1.25 to 1.42)", the section 12.3 report that total thyroxine "adjusted for endogenous levels" AUC was "increased by 33%" after a single 600 mcg levothyroxine dose, and "Maximum exposure (Cmax) was unchanged." The section 7.2 advice on other oral drugs with a narrow therapeutic index or a monitoring requirement is described here, not quoted, because the fetch truncated that sentence. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17
6. Levothyroxine Sodium tablets, 100 mcg, US prescribing information, Proficient Rx LP (repackager, source NDC 0527-3284), sections 2.1, 7.1 and 17. Read at source via DailyMed. Quoted for the instruction to administer as a single daily dose "on an empty stomach, one-half to one hour before breakfast", the instruction to take it "with a full glass of water as the tablet may rapidly disintegrate", and the instruction to administer "at least 4 hours before or after drugs known to interfere with" its absorption, including the named four-hour separation from calcium carbonate and ferrous sulfate. This is one repackager's label for one strength; levothyroxine is marketed by many labelers and the administration language is standard across them, but only this record was read. ↩1 ↩2 ↩3