Creatine in Pregnancy: No Trial Has Given It to a Pregnant Woman

This page exists because the question is asked, and because this site writes about searched topics rather than skipping the uncomfortable ones. Its subject is an absence: there is no trial of creatine supplementation in human pregnancy. The useful thing a page can do with an absence is document how thoroughly it holds, in a way a reader can repeat, and name what stands in its place. Nothing here is advice, in either direction, and a decision about a pregnancy belongs with a clinician who can see the individual case.

The census, made reproducible

PubMed, searched 2026-09-20. creatine[tiab] AND (pregnancy[tiab] OR pregnant[tiab]) returns 603 records; 362 are tagged human; 14 carry the clinical-trial publication type.

Fourteen would be a literature. Read one by one, it is not one. The 14 are:

The word doing the damage is creatine kinase — an enzyme, not the supplement — and creatinine, its metabolite. Both match a search for "creatine" and neither has anything to do with taking any. This site has had to separate those three entities before, on what creatine means in bloodwork, and they are the reason a raw count here reads eight times larger than the evidence.

ClinicalTrials.gov, API v2, searched the same day. An interventional search for creatine in pregnancy returns 11 studies. Reading each one's intervention list: three genuinely supplement creatine — on renal function, on serum and urine creatinine, and a 2026 study on cognition and kidney markers in exercise heat — and all three are in non-pregnant populations, matched because pregnancy-related terms appear elsewhere in the record. The remaining eight involve no creatine administration at all.

Zero, twice, by two different methods. Following the rule this network now applies to any zero, it was checked against queries known to return something: the same PubMed phrasing without the trial filter returns 603, and the same registry query for creatine without the pregnancy term returns hundreds. The search works. There is nothing to find.

And the one study that does exist is in a registry most searches never reach

The ISSN's 2025 expert review names it: initial pharmacokinetic and tolerability work is underway, registered as ACTRN12620001373965 — an entry in the Australian New Zealand Clinical Trials Registry, which does not feed ClinicalTrials.gov.

This page reports that registration as the review states it; the ANZCTR site refused automated requests from our tooling on the date of writing, so it is cited at one remove and labelled as such. The lesson underneath is the more durable part, and it is the same one this site applied to GHRP-2 on the same night: a registry census is a census of that registry. Monash University's Ritchie Centre runs this programme; Australian trials are registered in Australia. A researcher's country determines where the record lives, and a search that knows only one registry will report an emptiness that is really a border.

What has been studied, precisely

Fetal sheep. Chronically instrumented fetuses received a continuous intravenous creatine infusion of 6 mg per kg per hour for 13 days, at a stage of cerebral development equivalent to 34–36 weeks of human pregnancy. Circulating creatine rose about five-fold. Reported: no change in cardiovascular or systemic arterial blood-gas parameters through the infusion, no change in fetal body or organ weights at postmortem, and no adverse change in cerebral interstitial lactate, glutamate or hydroxyl free radicals. Transcriptional changes appeared in anti-apoptotic and mitogenic genes in the fetal hippocampus, without corresponding changes in cell death or mitochondrial respiration.

Spiny mice. Pregnant dams were fed creatine at 5% of feed by weight from mid-gestation to delivery, 18 days. Offspring were followed to adulthood with no reported impact on body growth or on the structure or function of kidney, skeletal muscle or diaphragm. A companion study in pregnant and non-pregnant animals found no sustained negative effect on maternal creatine homeostasis, body composition, capacity for endogenous synthesis, or renal excretory function.

Non-pregnant women of reproductive age. A systematic review and meta-analysis pooled 29 studies and 951 consented participants aged 16 to 67, given 1–30 g a day for between 4 and 365 days. No deaths and no serious adverse events, defined as life-threatening events, hospitalisation, persistent disability or events judged medically important. Stratified by dosing regimen, no significant difference in minor adverse events either — gastrointestinal, renal and hepatic blood and urine markers, or weight gain.

That last body of evidence is the one most likely to be repurposed, so its boundary belongs with it: it contains no pregnant participants. It describes what happened to women who were not pregnant.

What the research is for, and the problem its designers state

The programme is not a sports-nutrition programme. Its stated aim is neuroprotection: raising fetal creatine reserves so the newborn brain has more energetic reserve during the oxygen deprivation that can occur around birth, the injury called hypoxic-ischaemic encephalopathy. In the spiny mouse, maternal creatine supplementation across pregnancy has been reported to prevent deficits in skeletal muscle, diaphragm and kidney after birth asphyxia.

The researchers name the design problem themselves, and it is the honest centre of the whole question. Acute hypoxic events in labour are largely unpredictable, so a protective agent would have to be given prophylactically — across the whole antenatal period, to everyone, to benefit the few. In the ISSN review's words, that "does raise questions about potentially exposing healthy fetuses to high creatine concentrations throughout the antenatal period". A prophylactic in a healthy population carries a higher evidential burden than a treatment in a sick one, which is precisely why the tolerability study exists and why it comes first.

Creatine is already there, and that is not a safety argument

Creatine is synthesised in the body from arginine, glycine and methionine and is obtained from meat and fish. Maternal creatine and creatinine excretion in pregnancy was being described in the literature in the 1930s; the placenta's role in creatine metabolism at birth is an active subject in a current prospective cohort of low-risk pregnancies; population survey work has asked whether dietary creatine intake during pregnancy is adequate.

All of that is true and none of it bears on supplementation. A substance being endogenous and dietary says nothing about the safety of a supplemental dose, and the slide from one to the other is the most common error in writing about this topic. Water, iron, vitamin A and iodine are all endogenous or dietary and all have supplemental doses that matter in pregnancy. The two questions are separate and this page keeps them separate.

Who has published a position

• International Society of Sports Nutrition, 2025 (an eighteen-author narrative expert review, including the Monash researcher who runs the pregnancy programme): animal research suggests creatine during pregnancy does not negatively affect mother or offspring; there are no well-designed or executed randomised controlled clinical trials on safety and tolerability during human pregnancy. That is a statement of absence, from the field's own experts, and it is the strongest published line available.
• No obstetric college position specific to creatine supplementation was located in this search — not from ACOG, not from RANZCOG, not from NICE. That is reported as a finding, because a body having published nothing is itself checkable, and because the gap will presumably close when the tolerability data arrive.
• The separate question of breastfeeding does have a dated national position, from the NIH's Drugs and Lactation Database, and it is covered on creatine and breastfeeding.

The honest summary

• No trial has given creatine to a pregnant woman. Two independent censuses, one literature and one registry, return zero.
• The interventional evidence is sheep and spiny mice, at high exposures, with no adverse structural or functional findings reported.
• A tolerability study is registered in Australia and its results are not published.
• The safety meta-analysis people reach for covers 951 women who were not pregnant.
• No obstetric body has published a position on supplementation.
• Creatine is endogenous and dietary. That is a fact about biology, not an argument about supplements.

Limits of this page

The censuses are title-and-abstract classifications of one database and one registry on one date; a differently phrased query returns a slightly different set, and the figure to trust is the shape rather than the last digit. The ANZCTR registration is cited from a secondary source because the registry would not serve an automated request. The absence of a college position is an absence found by searching, not a guarantee that none exists in a language or format this search did not reach. Nothing on this page is medical advice, a recommendation, or a reason to start, continue or stop anything during a pregnancy; that decision requires a clinician who knows the individual case, and this page exists to describe what the evidence contains, not what anyone should do about it.

Sources and dates

• Antonio J, Brown AF, Candow DG, et al. Part II. Common questions and misconceptions about creatine supplementation: what does the scientific evidence really show? J Int Soc Sports Nutr 2025;22(1):2441760 — PMID 39720835 (section 11, and the concluding statement)
• Ostojic SM, Stea TH, Ellery SJ, Smith-Ryan AE. Association between dietary intake of creatine and female reproductive health: evidence from NHANES 2017–2020. Food Sci Nutr 2024;12(7):4893–8 — PMID 39055234
• PubMed, searched 2026-09-20: creatine[tiab] AND (pregnancy[tiab] OR pregnant[tiab]) — 603 records; with "clinical trial"[pt] — 14, classified individually above; "creatine monohydrate"[tiab] AND (pregnancy[tiab] OR pregnant[tiab]) AND "randomized controlled trial"[pt] — 0.
• ClinicalTrials.gov API v2, searched 2026-09-20: query.cond=pregnancy, query.intr=creatine, interventional filter — 11 studies, intervention lists read individually.
• ANZCTR registration ACTRN12620001373965, as cited in the ISSN 2025 review; the registry itself did not serve an automated request on 2026-09-20.