Creatine on GLP-1s: What the Evidence Actually Shows

The creatine conversation exists because the lean-mass numbers are genuinely notable. The best body-composition data come from DXA substudies embedded in the pivotal trials. In STEP 1, the semaglutide 2.4 mg trial, a DXA substudy found that roughly 40% of the total loss was lean tissue (Wilding et al., Journal of the Endocrine Society, 2021) — a later case-series analysis compiled the figures as 6.9 kg of lean soft tissue lost against 10.4 kg of fat over 68 weeks (Tinsley & Nadolsky, SAGE Open Med Case Rep 2025). In SURMOUNT-1, the tirzepatide trial, the DXA substudy recorded 5.6 kg lean against 15.9 kg fat over 72 weeks — roughly 26% (Look et al., Diabetes, Obesity and Metabolism, 2025).

The 2026 meta-analysis confirms the pattern at obesity doses with more granularity: across seven RCTs and 821 patients, pooled lean-mass loss was −1.74 kg (95% CI −3.04 to −0.45), with semaglutide arms averaging −5.44 kg and tirzepatide −10.9% versus −2.0% on placebo in SURMOUNT-1, all figures as reported by the meta-analysis (Laverde et al., Int J Obes 2026). The proportion of weight lost as lean tissue varied by drug — 14–22% for liraglutide, up to 45% for semaglutide, roughly 26% for tirzepatide. Different analyses of the same trials produce somewhat different proportions because substudy sizes, measurement methods, and calculation choices differ; an honest page reports the range rather than cherry-picking the scariest or gentlest number.

One more piece of context the commerce pages never include: in that same meta-analysis's treated groups, lean mass as a share of total body weight was slightly higher (+1.81% as reported), because fat loss so far outpaced lean loss that body composition improved in relative terms. Lean-mass loss on a GLP-1 is real, measurable, and worth managing — and it occurs inside a larger picture of substantially improved composition. Both halves of that sentence are what the data says.

Creatine is not a peptide and not a hormone. It is a small nitrogen-containing compound made naturally by the body from amino acids, stored mostly in skeletal muscle, and also obtained from food — meat and fish in particular. Its established biochemical role is energy buffering: muscle converts creatine into phosphocreatine, which can rapidly donate a phosphate group to regenerate ATP, the cell's immediate energy currency, during short high-intensity effort. In simple terms, phosphocreatine is a recharge buffer that helps muscle cells resupply energy during bursts of demanding work.

That mechanism is why creatine's research footprint centers on strength, power, and training adaptation: more available phosphate buffering can, in some contexts, support harder work in the gym, and harder work over time is what drives muscle retention and growth. It is also why the exercise context is inseparable from the creatine conversation — the mechanism pays off through mechanical work, not through the compound sitting in muscle doing nothing. This distinction will matter repeatedly below, because the reviews that discuss creatine for lean mass consistently pair it with resistance training rather than treating it as a standalone intervention.

Creatine is among the most studied supplements in sports nutrition, with a literature spanning decades and populations from athletes to older adults. None of that longevity automatically transfers to a GLP-1 context: evidence from other populations does not establish effectiveness during GLP-1-induced weight loss, and the classic creatine trials were not designed to examine the defining condition of that context — reduced intake during rapid weight loss.

The paper at the center of the current wave of content is "Creatine supplementation for preventing muscle atrophy in patients taking incretin-based pharmacotherapies: a perspective" (Ribeiro et al., The Journal of Nutritional Physiology, April 2026; DOI 10.1016/j.jnphys.2026.100028). Eight authors. Its thesis, stated by the authors themselves, is a proposal: that creatine supplementation, particularly alongside resistance training, could help mitigate fat-free-mass loss in patients on incretin therapies — and that this possibility deserves formal investigation.

The article type is the whole story. A perspective in a journal is a structured argument — a case for what future research should test. It contains no randomized participants, no creatine arm, no GLP-1 arm, no outcome measurements of any kind in the combination. When the authors conclude that creatine may have a role, they are hypothesizing on the record, which is a legitimate and useful scientific act. What it is not, under any reading, is evidence that creatine preserves muscle in GLP-1 users. The distance between those two things — a published hypothesis and a tested result — is exactly the distance supplement marketing elides.

To its credit, the perspective's very existence signals that researchers take the lean-mass problem seriously enough to propose countermeasures. The honest takeaway is double-edged: the idea is plausible enough that creatine researchers bothered to write it up, and untested enough that no published trial exists to cite. A reader who understands both halves of that sentence understands this topic better than most of page one.

The review, meta-analysis, and case series below converge on the same absence. The Acta Diabetologica review of muscle loss and GLP-1 receptor agonist use surveys nutritional countermeasures — protein, amino acids, creatine among them — and states plainly that definitive randomized trials specifically designed to test supplement strategies in GLP-1RA users are still lacking (Rossi et al., 2025/2026). The 2026 meta-analysis of lean-mass outcomes at obesity doses never mentions creatine at all, because there was nothing to pool (Laverde et al., Int J Obes 2026). And the SAGE case series that documented the STEP 1 and SURMOUNT-1 substudy numbers notes its own patients happened to take creatine, but could not analyze it separately — with the authors stating these practices have been minimally examined in this context (Tinsley & Nadolsky, 2025).

That is the gap between what studies measured and what users ask. Users ask: "should I take creatine on my GLP-1?" The literature has measured: how much lean mass GLP-1 users lose (well), whether exercise attenuates it (suggestively, mostly outside GLP-1 populations), and whether creatine helps other populations (partially, mostly alongside training). The specific question — creatine, in a GLP-1 user, during GLP-1-induced weight loss — has zero trial answers. Every confident public answer to it is either borrowed from a different population or projected from a hypothesis.

This gap is not a reason for despair; it is the reason a page like this one exists. A question with no direct evidence has exactly one honest treatment: state the indirect evidence, label it indirect, and refuse to upgrade it. The next section covers what the indirect evidence actually shows — including a result the supplement aisle will not be quoting.

The strongest general statement comes from the Acta Diabetologica review's summary of the creatine literature: creatine supplementation combined with resistance training may increase lean mass and muscle strength — while the evidence for creatine's effects on muscle mass in the absence of exercise remains inconclusive (Rossi et al., 2025/2026). The review documents the regimen used across that literature — a loading phase of roughly 20 g/day for five to seven days followed by 3–5 g/day maintenance — as a description of what prior studies employed, not a recommendation for any population, least of all one where no trial has run. The same review emphasizes protein intake in its cited expert recommendations (1.2–1.6 g/kg of adjusted body weight per day) and resistance training as the cornerstones, with periodic body-composition and strength monitoring during GLP-1 therapy.

The closest weight-loss-context datapoint is sobering: a pilot study in women after bariatric surgery found that creatine supplementation before strength training was not superior to standard nutritional recommendations for preventing muscle mass loss (Diaz-Pizarro et al., Obes Surg 2024). Bariatric surgery is not GLP-1 therapy, one pilot settles nothing, and the finding cannot be projected onto the GLP-1 context with any confidence — but its existence matters because it demonstrates that the cheerful assumption "creatine plus training obviously preserves muscle during weight loss" has already failed to hold in at least one tested weight-loss population.

On the encouraging side sits the SAGE case series: three patients on GLP-1 or GLP-1/GIP therapy, all resistance training three to five days per week with adequate protein, of whom two actually gained lean soft tissue while losing 26.8% and 13.2% of body weight, and the third lost only 8.7% of his total weight loss as lean tissue — versus the 26–40% share-of-loss figures from the trial substudies (Tinsley & Nadolsky, 2025). Two of the three happened to take creatine; the authors could not isolate its contribution. The series has no control group and self-reported diet, and says so. What it demonstrates is not that creatine works — it is that aggressive lean-mass preservation during incretin therapy looks achievable in principle, with training as the unambiguous common denominator.

On the specific question users search — does creatine interfere with semaglutide, tirzepatide, or other GLP-1 receptor agonists — the reviewed sources contain no reports of a pharmacological interaction, for the unglamorous reason that the combination has not been studied. The 2026 meta-analysis does not mention creatine; neither review reports interaction concerns; the perspective treats the pairing as an open question rather than a hazard; and no regulator guidance on the pairing appears in any of them. Absence of reports in an unstudied combination is not evidence of safety — it is evidence of silence, and this page will not dress it up as either reassurance or alarm.

What the clinical commentary does emphasize is a different risk frame. The Mass General Grand Rounds summary on preserving lean body mass during GLP-1 treatment (Mass General Advances in Motion, June 2025) makes three points worth carrying: lean-mass loss accompanies weight loss generally and is not established as uniquely worse with GLP-1s; the best-supported protective strategy reported was combining adequate protein with consistent exercise from the start of treatment, with patients who exercised throughout doing best; and measured lean-mass loss does not necessarily equal lost muscle function — an area the piece describes as under active investigation. It also flags bone health, particularly in older women and frailer patients, as a concern that may deserve more attention than the muscle numbers get. Notably, the piece does not mention creatine at all.

The practical reading: nothing in the reviewed literature flags creatine as a hazard alongside GLP-1 therapy, and nothing establishes it as beneficial either — the combination is a blank tile. Anyone taking a prescription medication who is considering any supplement raises a real question for a prescribing clinician, who can weigh it against a full medical record. That sentence is where the interaction question honestly stands today.

Search results for creatine and GLP-1s are dominated by pages that sell one of the two things being discussed — supplement retailers on one side, GLP-1 telehealth clinics on the other — with a handful of clinic education pages and community threads carrying the actual-question answers. That structure predicts the content. When the seller is a creatine company, the story runs: GLP-1s eat your muscle (the lean-tissue numbers above are real, though lean tissue is not the same thing as skeletal muscle), creatine builds muscle (true in other contexts, with training), therefore creatine on your GLP-1 (the inference no trial supports). When the seller is a telehealth clinic, creatine typically arrives bundled inside a protocol the clinic is positioned to prescribe. Neither framing is lying, exactly; both are profiting from the gap between two bodies of evidence and quietly retiring the sentence that would bridge them honestly: "this has not been tested."

Three moves let a reader grade any such page in under a minute. First, trace the citation to its type: if the headline claim rests on a perspective, a review of other populations, or a case series, the page is borrowing evidence across contexts — legitimate for background, fraudulent as proof. Second, watch for population laundering: "clinically studied" is true of creatine in athletes and older adults and says nothing about a person in a semaglutide-induced caloric deficit, a condition the classic trials did not model. Third, notice where dosing appears: a page handing you a gram number as advice has already decided something no clinician could have — the studied regimen is a fact about past trials in other populations, not a recommendation for this one. The pages that survive all three checks tend to be short and hedged, which is what an unstudied question looks like when it is being answered honestly.

PeptaHub sells neither creatine nor GLP-1 treatment, carries no affiliate links to either, and has no telehealth protocol to enroll you in — which is precisely why this page can afford to lead with the research gap instead of burying it below a buy button. The neutral read is the differentiator here: the most useful sentence on this topic is the one no vendor has an incentive to print, and it fits in one line. No reviewed source identifies a trial of creatine in people on GLP-1s; everything else is context.

The verified state of play, compressed: GLP-1 therapy at obesity doses reliably costs some lean tissue alongside much larger fat loss — roughly a quarter to 40%-plus of weight lost depending on drug and analysis, with the largest reported per-drug lean losses in the pooled data on semaglutide (Laverde et al., 2026; Tinsley & Nadolsky, 2025; Rossi et al., 2025/2026). The best-supported countermeasures in the reviewed literature are the unglamorous ones clinicians keep naming: resistance training and adequate protein, begun with the medication rather than after the weight comes off. Creatine is a plausible adjunct in that picture — proposed on the record by researchers in April 2026, supported by general-population evidence when combined with training, untested in GLP-1 users, and null in the one adjacent weight-loss pilot found. That is the entire honest summary, and no section of it is optional.

Two things this page asks of readers. Treat any source that converts the hypothesis into advice as having failed a disclosure test, whatever else it gets right — the conversion is the tell, because the missing trial is not a secret; it is the first thing every review on the topic states. And route the personal version of the question — your medication, your kidneys, your training history, your diet — through a clinician who can see you, not through content. Educational framing is not a legal footnote here; it is the accurate description of what a page without trial evidence, without your medical record, and without any way to observe your outcome can responsibly do.

When a randomized trial of creatine in GLP-1 users does land, this page will be updated with its numbers and its evidence grade — that is the standing commitment. Until then, the graded table above is the whole picture: real problem, borrowed evidence, honest gap.