Does Creatine Cause Hair Loss? What the Evidence Says

Type “does creatine cause hair loss” into a search bar and you’ll get a wall of confident answers, most of them built on a single three week study of twenty rugby players from 2009. That study never counted a hair. Seventeen years later, exactly one trial has, and it found nothing, though it has real limitations that its cheerleaders tend to skip.

So here’s the short version. The evidence that creatine causes hair loss is weak and indirect. The evidence that it doesn’t is better, but it’s still thin, it comes mostly from one research community, and it hasn’t been tested in the people most likely to worry about it. If you’re already losing hair or your father went bald young, the honest answer is that nobody has studied you yet.

What The Evidence Actually Shows

Here is the whole literature in four sentences. In 2009, a small crossover trial in South African rugby players reported that a week of high dose creatine loading raised blood levels of dihydrotestosterone, the hormone most tied to pattern baldness, by about 56 percent. Over the following decade, a dozen other trials measured testosterone or related hormones during creatine use and found no meaningful change, though almost none of them measured DHT directly. In 2025, the first randomized trial to actually examine scalp hair gave 5 grams a day to resistance trained men for twelve weeks and found no difference from placebo in DHT, in the DHT to testosterone ratio, or in any of eight hair measurements. A planned six month trial at a US university that would have added longer follow up was registered and then withdrawn.

That’s it. Everything else you read on this topic is interpretation layered on top of those facts, and I’ll try to keep my own interpretation clearly labeled as such.

The 2009 Rugby Study That Started Everything

The paper is by van der Merwe, Brooks and Myburgh in the Clinical Journal of Sport Medicine, and it deserves a fair reading because it’s usually either waved away or treated as a smoking gun. It was a randomized, double blind, placebo controlled crossover trial in twenty college aged male rugby players during their competitive season. The creatine arm took 25 grams a day of creatine monohydrate plus 25 grams of glucose for seven days, then 5 grams a day of creatine plus 25 grams of glucose for fourteen days. The placebo arm took glucose alone.

Total testosterone didn’t change. DHT did. After the loading week it was 56 percent above baseline, and after the two maintenance weeks it was still about 40 percent above baseline. The ratio of DHT to testosterone rose 36 percent after loading and stayed 22 percent elevated during maintenance. The authors framed this as a shift in how much testosterone was being converted to DHT, which happens through an enzyme called 5 alpha reductase, and that’s the mechanism people have been worrying about ever since.

Two things get lost in the retelling. First, every DHT value at every time point in both arms sat inside the normal reference range for healthy adult men, which is roughly 0.8 to 3.5 nmol/L. Second, the study measured no hair outcome of any kind. Not density, not shedding, not photographs, not a questionnaire. The link from this paper to baldness is an inference, and it was never the authors’ claim.

What The 56 Percent Actually Means

Percentages hide a lot. The 2021 review by Antonio and colleagues in the Journal of the International Society of Sports Nutrition pointed out something I think every article on this topic should mention and almost none do. At baseline, before anyone took anything, the creatine group’s DHT was 23 percent lower than the placebo group’s, 0.98 versus 1.26 nmol/L. The creatine group then rose by 0.55 nmol/L after loading and by 0.40 nmol/L at three weeks. The placebo group drifted down by 0.17 and then 0.20 nmol/L.

So the headline figure is a modest rise from a low starting point, measured against a comparison group that was falling. That doesn’t make the finding wrong, and I don’t want to overcorrect into dismissing it. A statistically significant change in the DHT to testosterone ratio in a controlled crossover design is a real observation. But it’s a small absolute change in a small group over three weeks, and the same review noted that hard resistance training can itself nudge androgen levels around. Rugby players in season are not a resting population.

The other thing worth saying plainly is that this result has never been reproduced. Nobody has run the same loading protocol and measured DHT again in the sixteen years since. That’s not the same as it being refuted, and I’d push back on anyone who says it has been. It’s simply an unreplicated finding, which in a field this large should make you nervous in both directions.

Twelve Studies That Looked For A Hormone Signal And Did Not Find One

The same 2021 review counted twelve other trials that measured androgen hormones during creatine supplementation and reported no meaningful increase in total testosterone, free testosterone or DHT. A few found small, clinically trivial fluctuations in total testosterone. The review’s summary was that the existing evidence does not indicate creatine raises DHT or causes hair loss, and I think that’s a fair reading of the studies it looked at.

Here’s my caveat, and it matters. Most of those twelve trials were designed to look at strength, body composition or training adaptation, and they measured testosterone as a secondary marker. Very few measured DHT itself. A null result for testosterone tells you almost nothing about DHT, because the 2009 study also found no change in testosterone. What it found was a change in the conversion ratio. So “twelve studies found nothing” is true but it’s answering a slightly different question than the one the rugby study raised. The evidence base on DHT specifically, before 2025, was one positive study and a handful of nulls. If you want a sense of how thin that is, consider that creatine has been studied in several hundred human trials for muscle and performance. The hormone question got a footnote in most of them, and the hair question got none.

I say this not to resurrect the scare but because the site you’re reading exists to report what studies found rather than what we’d like them to have found. The hormone evidence is reassuring. It isn’t as overwhelming as the “twelve studies” framing suggests.

The First Trial To Actually Count Hairs

In April 2025, Lak and colleagues published a twelve week randomized controlled trial in the same journal, and it’s the single most important paper on this question because it did the thing nobody had done. It measured hair. Forty five resistance trained men aged 18 to 40 were randomized to 5 grams a day of creatine monohydrate or 5 grams of maltodextrin placebo, double blind, for twelve weeks. Thirty eight finished. Blood was drawn at baseline and week twelve for total testosterone, free testosterone and DHT. Hair was assessed by board certified dermatologists at the Iran Skin Research Center using a trichogram and the FotoFinder imaging system, focused on the vertex of the scalp, which is where pattern thinning usually shows first.

The hair panel was thorough. Hair count, hair density per square centimeter, percentage of hairs in the growth phase, percentage in the resting phase, total follicular units, percentage of terminal hairs, percentage of fine vellus hairs, and cumulative thickness. Eight outcomes. None of them differed between creatine and placebo, and none changed over time in either group. DHT didn’t move. The DHT to testosterone ratio didn’t differ between groups. The authors ran the analysis three ways, on raw data, on ranked data and with outliers removed, and got the same answer each time.

There was an odd wrinkle. Total testosterone rose and free testosterone fell in both groups over the twelve weeks, independent of what they were taking. The authors speculate about seasonal variation, since the study ran from late summer into autumn, and are candid that they don’t know why. It’s the kind of detail that gets cut from the summaries, and it’s a reminder that hormones in free living people drift for reasons that have nothing to do with a supplement.

The Criticism That Trial Deserves

I like this study. I also think its own discussion section is more careful than most of the coverage it got, and the criticism it has drawn is worth taking seriously rather than brushing aside. The American Hair Loss Association published a response within a week arguing that the trial was far more limited than the headlines implied, and several of their points are just correct.

Thirty eight completers is small. The study didn’t screen for genetic predisposition or ask about family history of baldness, which the authors themselves flag as a limitation, noting that men whose fathers lost their hair are about twice as likely to lose their own. It excluded anyone with diagnosed hair loss conditions, so it tells you nothing about people already thinning. It measured DHT in blood rather than at the scalp, and pattern hair loss is a local process driven by follicle sensitivity, not just circulating hormone levels. And twelve weeks is a short window for a process that unfolds over years.

The association also raised conflicts of interest, and the disclosures back them up. Several authors advise creatine brands or run consulting businesses for supplement manufacturers, one is the chief executive of the society that publishes the journal, and the creatine itself was donated through a supplement company’s distributor. None of that means the data are wrong. Blinded outcome assessment by dermatologists is a real safeguard, and the statistical approach was transparent. But if you’d discount an industry funded trial that found a benefit, you should apply the same discount to an industry connected trial that found no harm. Consistency is the whole game.

The Six Month Trial That Never Happened

There’s a registry entry I keep coming back to. In 2020, Texas Tech University registered a six month randomized trial of creatine, androgens and hair under Grant Tinsley, who was later a co author on the 2025 paper. The plan was 5 grams a day for six months in free living men aged 18 to 35, with blood work for testosterone and DHT plus standardized global photography and hair loss questionnaires at baseline and at the end. That design would have addressed the two biggest gaps in what we have, duration and a patient reported outcome.

The record now lists the study as withdrawn. I don’t know why, and I’m not going to guess. But it means the twelve week trial is the longest direct look at hair we have, and that a longer one was on the table and didn’t run. When you see articles claiming the question is settled, remember that the researchers themselves apparently thought a six month study was worth doing.

Why Blood DHT Is Not The Whole Story

A quick primer, because the mechanism matters for judging the evidence. Androgenetic alopecia, the medical name for pattern hair loss, happens when hair follicles that are genetically sensitive to DHT gradually shrink and produce finer, shorter hairs. DHT is made from testosterone by 5 alpha reductase, and it binds to androgen receptors more strongly than testosterone does. Two men can have identical blood DHT and completely different hairlines, because what matters is the receptor density and enzyme activity inside the follicle, both of which are largely inherited.

This cuts both ways for the creatine debate. It means a 56 percent rise in serum DHT in the rugby players wouldn’t necessarily translate to anything at the scalp, especially with every value inside the normal range. It also means a null serum DHT result in the 2025 trial doesn’t fully close the question, because a supplement could in theory alter local enzyme activity without moving blood levels much. The 2025 authors say as much when they suggest future studies measure androgens in hair itself. As far as I can find, nobody has.

I’ll add my own opinion here. The idea that creatine upregulates 5 alpha reductase is a hypothesis with one supporting data point and no proposed biochemical mechanism I’ve seen worked out anywhere. Creatine is a nitrogenous compound involved in cellular energy recycling. There’s no obvious reason it should touch steroid metabolism, and the burden is on that hypothesis, not against it.

Loading Doses Versus What People Actually Take

This is where a lot of the confusion lives, so let’s be exact. The rugby study’s DHT rise appeared after a 25 gram per day loading week, five times the standard dose. Even then, the 40 percent elevation during the 5 gram maintenance phase was measured only two weeks into maintenance, with the loading week immediately behind it. Nobody has measured what DHT does after months at 5 grams following a load.

The 2025 trial skipped loading entirely and used 5 grams a day, which is what most people take, for twelve weeks, and found no DHT change. The authors themselves point to the lack of a loading phase and the longer duration as the most likely reasons their result differed from 2009. That’s a reasonable reading, and it suggests a practical takeaway I’m comfortable stating in general terms about creatine as an ingredient. If the only DHT signal in the literature came from a high dose loading protocol, and you’re worried about DHT, there’s no strong reason to load. Loading gets muscle stores saturated in a week instead of three or four, and that’s the only thing it does. The 2017 ISSN position stand on creatine safety describes 3 to 5 grams a day as sufficient to saturate stores over time without a loading phase.

I want to be careful not to imply that skipping the load is proven to protect your hair. It isn’t proven to do anything about hair, because hair has only been measured once. It’s simply that the one worrying hormone result came from a protocol most people don’t need.

Women, Family History, And Who The Findings Cover

Every human data point on creatine and DHT comes from men. Both the 2009 and 2025 studies enrolled only males, and the 2025 authors explicitly list this as a limitation. Women do experience pattern hair loss, and androgens play a role in some cases, but their baseline testosterone and DHT are far lower and the biology is different. There is no study of creatine and hair in women. Articles that reassure women on the basis of these trials are extrapolating, and articles that warn women on the basis of the rugby study are extrapolating harder.

Family history is the other gap. The 2025 paper cites a 2004 study by Chumlea and colleagues finding that men whose fathers had hair loss were roughly twice as likely to have it themselves, and then acknowledges it didn’t collect that information. So the trial can’t tell you whether creatine affects the subgroup that actually matters, men with the genetic sensitivity that makes DHT relevant in the first place. If that’s you, the correct summary of the evidence is that a small trial in a general population of trained men found nothing over twelve weeks, and nobody has looked specifically at people like you.

Age matters too. Both trials studied men under 40. Older adults, who are a growing share of creatine users because of the research on muscle and cognition in aging, haven’t been studied for hair at all. Neither have adolescents, vegetarians with low baseline creatine, or anyone taking creatine alongside the prescription drugs that alter DHT, and those last two groups would be the most informative to study.

What Regulators Say, And What They Do Not

In the United States, creatine is sold as a dietary supplement under the Dietary Supplement Health and Education Act, which means the FDA doesn’t approve it or review its safety before sale. Creatine monohydrate has been the subject of a Generally Recognized as Safe notification, and the FDA has not, to my knowledge, issued any communication about creatine and hair. The FTC, which polices advertising, would treat any claim that a creatine product protects hair or that it definitely doesn’t affect hair as a claim needing substantiation, and I’d argue neither claim has it.

The European picture is slightly different and worth knowing because EU labels sometimes get quoted as though they apply everywhere. The European Food Safety Authority’s 2011 opinion on creatine and physical performance supported a health claim for short term high intensity exercise at 3 grams a day, which is why EU products can carry that wording and US products can’t in the same form. In 2024, the Spanish food safety agency published a risk assessment of creatine supplements in EFSA’s journal. I haven’t found a hair related warning from EFSA, the UK’s Food Standards Agency, or any national regulator. That’s not evidence of safety, since regulators generally respond to adverse event signals rather than to theoretical mechanisms, but it’s the honest state of play.

On general safety, the ISSN position stand reviewed studies at doses up to 30 grams a day for as long as five years in clinical populations and found no consistent adverse effects in healthy people. The 2025 trial also tracked creatinine and estimated kidney filtration rate and saw no change. None of that is hair specific, but it’s the context in which the hair question sits.

If You Are Losing Hair And Taking Creatine

Let me address the reader who found this article because they noticed more hair in the drain after starting creatine, since that’s the person these search results mostly fail. Shedding that begins within a few weeks of a new supplement is far more likely to be telogen effluvium, a temporary shed triggered by stress, illness, dieting, a new training program or a dozen other things, than androgenetic thinning, which is slow and usually takes months to years to notice. Starting creatine often coincides with starting a hard new training block or a cut, and either of those is a plausible trigger on its own.

The self experiment people run, stopping creatine for a month and watching, is nearly worthless, and I’d say that gently. Hair cycles are long. A hair that sheds today entered its resting phase two to three months ago. Whatever you see in the shower a month after quitting reflects what was happening in your follicles before you quit. You’d need something closer to six months and ideally photographs under consistent lighting to learn anything, and even then you’d have no control group.

If the thinning is at the temples or crown and there’s pattern baldness in your family, the likeliest explanation is that you were going to thin regardless and the timing is coincidence. That’s not what anyone wants to hear. A dermatologist can distinguish diffuse shedding from pattern loss with a scalp exam and can talk you through options, and that conversation is worth more than any supplement decision. I’m not in a position to give medical advice and neither is anyone else writing about this online.

What Would Change The Picture

I’ll finish with what I’d need to see before saying this question is closed, because I don’t think it is, and I’d rather tell you that than pretend.

A replication of the 2009 protocol with DHT measured directly would settle whether the loading effect is real. It’s a cheap study and it’s strange nobody has run it. A longer hair trial, six or twelve months, with family history recorded, standardized photography, a shedding questionnaire and ideally scalp androgen measurement, would tell us whether the twelve week null holds in the people who matter. Any hair data in women at all would be new. And a trial run by a group without ties to the creatine industry would remove an objection that, fairly or not, hangs over the 2025 result.

Until then, here’s where I land, clearly marked as opinion. The claim that creatine causes hair loss rests on one small unreplicated hormone study that didn’t measure hair and whose values never left the normal range. The claim that it doesn’t rests on a dozen indirect hormone studies and one direct twelve week trial with real design gaps and industry connections. The second body of evidence is better than the first, and it’s consistent with the absence of any mechanism, any regulatory signal, or any clinical case series after decades of mass use. If I were advising a friend with no family history, I’d tell them the hair worry is low on the list. If they had a strong family history and were anxious about it, I’d tell them the science hasn’t looked at people like them, that skipping the loading phase removes the only dose that ever produced a signal, and that a dermatologist is a better use of their attention than a forum. Does creatine cause hair loss? On current evidence, probably not, and I’ll update that the moment someone runs the study that would actually tell us.