Muscle cramps are among the most commonly discussed complaints on these drugs and among the least documented. The words cramp, myalgia, arthralgia, musculoskeletal, hypokalemia and electrolyte appear a combined zero times across the current prescribing information for Wegovy, Zepbound, Mounjaro, Ozempic, Saxenda and Rybelsus.[1][2][3] That is not a limitation of the reading: in the same six documents nausea appears 10 to 20 times each, volume depletion six to eight times, and dehydration five to seven. The one hit for spasm is inside bronchospasm, in an allergic reactions paragraph, and the hits for magnesium are all magnesium stearate, a tablet excipient.
A threshold makes the absence mean something
An adverse reaction missing from a label is only informative if the label would have caught it. The liraglutide 3 mg document sets the clearest floor in the class: its adult table lists every reaction occurring in at least 2% of treated patients and more often than placebo, and it goes all the way down to anxiety at 2.0% against 1.6%, and dry mouth at 2.3% against 1.0%, across 3,384 treated adults and 1,941 on placebo.[1] A cramp reaction occurring in one treated adult in fifty would have appeared on that table. None did, in any of the six programs. Whatever is happening to people, it did not separate from placebo at the rate the conversation implies — which is a different statement from saying it does not happen, and the distance between those two is most of this page.
The one musculoskeletal entry is in the adolescent table
There is exactly one place in the six labels where muscle-adjacent terms appear at all, and it is not the adult data. The liraglutide pediatric table, drawn from a 56-week trial, reports pain in extremity in 4.0% of treated adolescents against 2.4% on placebo, and increased blood creatine kinase in 3.2% against 2.4%.[1] Creatine kinase is the enzyme that leaks from damaged muscle, so on its face that is the only signal in the class pointing at muscle tissue directly.
The denominators undo most of it. That trial randomized 125 adolescents to liraglutide and 126 to placebo, so 4.0% against 2.4% is five patients against three, and 3.2% against 2.4% is four against three. A one- or two-patient difference in a trial that size carries no useful precision, and the threshold for that table was 3% rather than the adults’ 2% — a higher bar over a sample twenty-seven times smaller. It belongs on the page because it is the only entry of its kind, and it belongs with its denominator because without one it reads like a finding. The wider adolescent picture is in the adolescent article.
What the musculoskeletal analyses found when they went looking
Two 2026 syntheses examined this class specifically for musculoskeletal harm, and neither produced a cramp result. A disproportionality analysis of FDA adverse event reports from the first quarter of 2004 through the second quarter of 2024 examined 15,052 de-duplicated reports of musculoskeletal and connective tissue disorders and found no signal at the system organ class level for any incretin agent. At the level of individual preferred terms, GLP-1 receptor agonists were linked to back pain, myalgia and neck mass, and tirzepatide to muscle atrophy and neck mass, with a median onset under 30 days.[4] Cramp terms are not among the signals that emerged. A spontaneous reporting database cannot establish a rate, but it is the surface where an unlabeled symptom would surface first, and what surfaced there was pain rather than cramping.
The other synthesis went to the trial literature. It pooled 60 studies covering 1,250,717 individuals — 46 randomized trials, 13 real-world studies and one pharmacovigilance study — across bone, muscle and joint outcomes. It found no effect on bone mineral density or fractures in the most adjusted models, no significant change in WOMAC pain, physical function or stiffness, and a consistent reduction in lean body mass across 28 comparisons (95% CI −0.80 to −0.23), rated low certainty and attributed largely to the weight loss itself.[5] Cramp is not among the outcomes it was able to pool, because the underlying trials did not record one. The lean-mass figures are the subject of the muscle-loss article and the scan substudies behind them are in the body-composition article.
The count of dedicated papers is the same shape. Indexed records pairing semaglutide, tirzepatide or liraglutide with muscle or leg cramps in the title or abstract number one. On the same index, those three drugs return 592 records alongside nausea, 252 alongside pancreatitis, 120 alongside lean mass and 19 alongside dehydration — and the cramp literature is not itself thin, returning 1,808 records on its own. The instrument works and the subject is well studied. The intersection is empty.
The single record is a 2025 case report, and it argues against the reading it would be cited for. A 73-year-old woman with type 2 diabetes collapsed at home with an undetectable serum magnesium below 0.3 mmol/L and hypocalcemia. She had already been taking four other drugs known to lower magnesium — metformin, gliclazide, sitagliptin and esomeprazole — and had stopped semaglutide, at the starting dose of 0.25 mg weekly, two weeks before presenting. Her longstanding muscle cramps resolved after electrolyte repletion. The author attributed the picture to chronic unrecognized hypomagnesemia from polypharmacy, with a contribution from the recent GLP-1 use, and discloses lecture fees from the manufacturer.[6] The cramps in the only cramp paper in this literature were pre-existing, and the drug had been stopped.
The electrolyte reports that exist name another drug
Hypokalemia has been reported, and the reports are worth reading for what else was in the chart. A 2025 case series described two women hospitalized during semaglutide treatment. The first developed nausea and emesis with potassium of 2.7 mmol/L, was readmitted months later at 2.6 mmol/L, and required parenteral nutrition and a feeding tube. The second was taking chlorthalidone 50 mg daily, a thiazide-like diuretic, with a pre-treatment potassium already low at 3.4 mmol/L; after six weeks on semaglutide it fell to 3.1 and then to 2.5 mmol/L. Stopping the diuretic and increasing supplementation brought it to 4.2 mmol/L. The authors state plainly that hypokalemia is not a reported effect of this class and that the cause in both patients may be multifactorial.[7]
That second patient is the practical lesson. A potassium-wasting diuretic plus a drug that suppresses intake and can cause vomiting is a combination with a mechanism, and it is invisible unless someone draws the blood. Anyone on a thiazide, a thiazide-like agent or a loop diuretic who starts cramping has a question with a cheap answer, and it is not an electrolyte powder. The broader fluid-loss picture sits in the dehydration article.
Where cramps were studied with blood drawn, the model reversed
The depletion story has been tested properly in one population — athletes — and it did not survive. A 2026 systematic review screened 14,181 records, included nine longitudinal studies covering 11,274 participants, and ran 36 meta-analyses. Only three variables were significantly associated with exercise-associated cramping. A prior history of cramps conferred roughly 4.4-fold higher odds of another episode, on strong evidence. The other two ran the wrong way: post-race serum potassium and magnesium were higher in the people who cramped, not lower. Strong evidence showed no association with age, body mass or height, and no consistent result emerged across the hydration, metabolic, training or performance domains.[8] The review's own conclusion is that the findings challenge the traditional preventive paradigm.
That is an athletic population under race conditions, not adults losing weight on an incretin, and it does not transfer directly. What it establishes is that the premise every GLP-1 cramp remedy is sold on has been measured where cramping is common enough to measure, and came out pointing the other way. An evidence review in athletic training reached a compatible position: cramping reflects a confluence of intrinsic and extrinsic factors rather than a single cause, acute treatment remains gentle static stretching, and individualized prevention is likely to beat generalized advice such as drinking more fluid.[9] Related ground for anyone training while losing weight is in the athletes article.
The remedy has been randomized, and it splits by population
Magnesium is the intervention most often recommended for this symptom, and it has been tested enough to give a divided answer. A 2026 systematic review and meta-analysis of 13 trials found that in pregnancy-associated cramps, across 4 trials and roughly 364 participants, magnesium significantly reduced cramp frequency against placebo — pooled relative risk 1.35 (95% CI 1.05 to 1.74, P = .02). In nocturnal or persistent leg cramps in non-pregnant adults, across 4 trials and roughly 396 participants, there was no significant effect: mean difference −0.42 cramps per week (95% CI −1.15 to 0.31, P = .26). Calcium and potassium had no supportive evidence at all.[10]
A Cochrane review of 11 trials in 735 people is more pointed about the older adult group. For idiopathic cramps in participants with mean ages of 61.6 to 69.3 years, magnesium against placebo gave a mean difference of −0.18 cramps per week at four weeks (95% CI −0.84 to 0.49; 5 studies, 307 participants; moderate certainty), and the proportion achieving at least a 25% reduction in cramp rate was no different at RR 1.04 (95% CI 0.84 to 1.29; 3 studies, 177 participants) on high-certainty evidence.[11] Its conclusion is that clinically meaningful prophylaxis in older adults is unlikely.
There is a second reason that matters here specifically. In the same review, minor adverse events were more frequent on magnesium than placebo — RR 1.51 (95% CI 0.98 to 2.33) — and were mostly gastrointestinal, with diarrhea reported by 11% to 37% of magnesium recipients against 10% to 14% of controls.[11] Oral magnesium’s characteristic side effect is loose stool, which is the exact loss pathway most likely to be driving the problem in someone on this class of drug. A remedy with no demonstrated benefit in non-pregnant adults and a known tendency to worsen fluid loss is a poor first move. Fluid and electrolyte practice generally is covered in the hydration article.
What a compounded prescription changes about all of this
Every threshold quoted above belongs to an FDA-reviewed document. Most sellers covered on this site dispense compounded semaglutide or tirzepatide, which are not FDA-approved and are not reviewed by the FDA for safety, efficacy or quality before they are dispensed. The consequence for this page is specific: a compounded vial carries no adverse-reaction table, so the census that gives the absence of cramp its meaning cannot be performed on the product actually being shipped. There is no 2% threshold behind it to argue from.
The one systematic look at compounded products analyzed 81,078 GLP-1 reports, 707 of them involving compounded formulations, and found higher reporting odds for abdominal pain at 2.84 (2.29 to 3.49), diarrhea at 1.59 (1.25 to 1.99), suicidality at 6.34 (4.32 to 8.99) and hospitalization at 2.35 (1.94 to 2.83), alongside preparation errors at 48.92 and contamination at 19.00.[12] No musculoskeletal term appears among the signals it reports. That is an absence in a voluntary database rather than a measurement, but the error and contamination figures are the part of that paper which transfers to a buying decision.
What can honestly be said
No label in the class counts this symptom, at thresholds low enough to have caught it at 2%. The only muscle-related entries anywhere in the six documents are two lines in an adolescent table separated by one or two patients. The dedicated literature is a single case report describing pre-existing cramps in someone whose magnesium was being lowered by four other drugs. The depletion model that justifies every remedy on offer ran backwards in the one population where it was measured with blood drawn, and the supplement most often recommended has a null result in non-pregnant adults on moderate-to-high-certainty evidence. The tolerability picture this sits inside is in what the trials recorded, and the seller-level questions are in the provider write-ups.
Two presentations are worth a same-week call rather than a supplement: cramping alongside an inability to keep fluids down, which is volume depletion and has a kidney consequence, and cramping in anyone taking a diuretic, where a single potassium and magnesium draw answers the question outright. Cramping with dark urine, muscle weakness or swelling after unaccustomed exertion is a different problem again and needs assessment the same day. Everything else is a symptom nobody has counted yet, and the honest version of that sentence is more useful than a number invented to fill it.