Ibuprofen and naproxen are the drugs a person reaches for without telling anybody. They are on a supermarket shelf, they get taken for a headache or a knee, and they are the single most common accompaniment to a new prescription that nobody records. Two separate questions follow from starting a GLP-1 while taking them, and they have very different answers: one has never been studied at all, and the other has been studied thoroughly under a different name.
The absorption question has never been asked
Searching PubMed for semaglutide together with ibuprofen returns one record, and it is a case series about fixed drug eruptions in primary care.[1] Semaglutide with naproxen returns none. Tirzepatide with ibuprofen returns none. Broadening the search to semaglutide against the whole category returns six records, not one of which is a pharmacokinetic study. No NSAID has ever appeared in a GLP-1 drug interaction trial, and none appears in the systematic review that assembled this literature — 22 reports and six prescribing sheets, none involving an anti-inflammatory.[2]
What can be said comes from a neighbor. Paracetamol has been used for decades as the marker drug for gastric emptying, and a population analysis measured it alongside semaglutide: the absorption rate constant fell 53%, which translated into an additional five-minute delay in the drug leaving the absorption compartment.[3] Five minutes. Acetaminophen and ibuprofen are both small, soluble, permeable analgesics taken for an effect that arrives within the hour, and the class pattern for such drugs is a reduced or unchanged peak with a later time to peak and no meaningful change in total exposure.[2] The general form of that pattern is set out in the oral medications article.
So the prediction is that a painkiller works slightly later and no less. That prediction has never been tested on an NSAID, and a prediction is what it remains.
The question that has been studied belongs to the kidney
Both weight-management labels carry a warning with the same structure. The semaglutide label records “postmarketing reports of acute kidney injury, in some cases requiring hemodialysis,” states that “the majority of the reported events occurred in patients who experienced gastrointestinal adverse reactions leading to dehydration such as nausea, vomiting, or diarrhea,” and instructs that renal function be monitored in patients reporting such reactions, “especially during dosage initiation and escalation.”[4] The tirzepatide label carries the same warning in almost the same words.[5]
The prescription NSAID label describes the identical event from the other end. Renal toxicity “has also been seen in patients in whom renal prostaglandins have a compensatory role in the maintenance of renal perfusion,” where an NSAID reduces prostaglandin formation and “secondarily, in renal blood flow, which may precipitate overt renal decompensation.” It then lists who is at greatest risk: “those with impaired renal function, dehydration, hypovolemia, heart failure, liver dysfunction, those taking diuretics, ACE inhibitors or the ARBs, and the elderly,” and instructs prescribers to “correct volume status in dehydrated or hypovolemic patients prior to initiating” the drug.[6]
One label says this drug can leave people volume-depleted. The other says do not start this drug in people who are volume-depleted. Neither names the other drug class anywhere — the words NSAID, nonsteroidal, ibuprofen and naproxen appear zero times across the semaglutide and tirzepatide labels. The interaction is not pharmacological at all. It is a state one drug creates episodically and the other drug is contraindicated during.
The size of the volume-depletion effect, in a triad already assembled
The combination of a diuretic, a renin-angiotensin system blocker and an NSAID has its own literature and its own name. A 2025 systematic review and meta-analysis of five case-control studies covering 410,130 participants found the risk of acute kidney injury within twelve months significantly higher with that triple exposure than without it — odds ratio 2.01 (95% CI 1.30 to 3.10), pooled across the four studies that reported it, with heterogeneity of 90%, low certainty of evidence and no significant excess in all-cause mortality.[7]
Two of those three legs are already standard in the population buying these drugs. In the 17,604-participant cardiovascular outcome trial for semaglutide, 74% of participants were on an ACE inhibitor or angiotensin receptor blocker at baseline.[4] The third leg is the one sold without a prescription. The NSAID label makes the point itself, instructing that during concomitant use with an ACE inhibitor or ARB in patients who are volume-depleted, prescribers should “monitor for signs of worsening renal function” and that “patients should be adequately hydrated.”[6]
Where the outcome was measured directly, it moved the other way
The label warning is built on spontaneous postmarketing reports, which record events without a denominator. The nearest thing to a controlled measurement comes from bariatric surgery, a setting that deliberately combines fasting, fluid restriction and gastrointestinal upset. A retrospective cohort matched 5,480 patients in each arm by prior GLP-1 exposure before elective laparoscopic surgery. Acute kidney injury within 90 days occurred in 1.6% of the exposed group and 2.3% of the unexposed — hazard ratio 0.68 (95% CI 0.52 to 0.89, P = 0.005), an absolute risk reduction of 0.73% and a number needed to treat of 137. The 30-day hazard ratio was 0.58. Mortality, intensive care admission and sepsis did not differ.[8]
That is the opposite direction from the one the label implies, in exactly the circumstance the label describes. It does not refute the warning: surgical patients are prepared, hydrated and monitored, which is the thing an unattended dose escalation at home is not, and the drug’s longer-run kidney effects are a separate matter covered in the kidney disease article. What it establishes is that the population-level relationship between this drug class and acute kidney injury is not simply additive, and that a page reporting only the label’s half would describe a picture the one controlled dataset contradicts.
The stomach question the labels do not raise
NSAIDs damage the gastric mucosa, and the over-the-counter Drug Facts panel is explicit about it: the product “may cause severe stomach bleeding,” with the chance higher in people aged 60 or older, those with prior ulcers, those on an anticoagulant or a steroid, and those taking more than directed or for longer than directed.[9] A drug that slows gastric emptying increases the time a tablet sits against that mucosa, which is a plausible mechanism and an entirely unmeasured one. No study has compared NSAID-associated gastrointestinal bleeding rates with and without a GLP-1, and neither product label raises the question.
What complicates any attempt to read it from symptoms is that the two drugs produce overlapping complaints. Nausea, upper abdominal pain and reflux are common early on a GLP-1 and are also the presenting symptoms of NSAID gastropathy, which is part of why reflux on a GLP-1 is difficult to attribute. An absent study and a confounded symptom is not an argument that nothing is happening; it is an argument that nobody would currently know.
What this changes at a cash intake
Nothing here supports banning a painkiller. Occasional ibuprofen in a well-hydrated person on a stable dose is not what either label is describing. The instruction both labels converge on is narrower and time-bound: the days during and after a bout of vomiting or diarrhea, and the weeks of each dose increase, are when an NSAID should not be added and when kidney function is worth checking. Those are the same days covered by the sick-day article.
Carrying that instruction out requires somebody to know three things at once: that the GLP-1 exists, that an over-the-counter analgesic is being taken, and that a blood pressure prescription supplies the third leg of the triad. An intake form that asks for a height, a weight and a list of prescriptions captures at most one of the three, because the NSAID is not a prescription and nobody thinks to list it. That gap — a labeled instruction addressed to a clinician nobody in the transaction is — is the recurring pattern described in the telehealth article.
Compounded semaglutide and tirzepatide are not FDA-approved, and the agency does not evaluate their safety, efficacy or quality before a pharmacy dispenses them. The kidney warning quoted above is a property of the branded labels, not of the molecule’s paperwork: a compounded vial arrives without prescribing information, so the sentence instructing renal monitoring during dose escalation reaches the buyer only if a prescriber says it out loud. How claims here are established is set out in the methodology.