Does Creatine Cause Hair Loss? Nobody Checked The Hair Until 2025

Somebody measured a hormone in twenty rugby players in 2009.

Nobody measured anyone’s hair until 2025.

That gap — sixteen years between the study everyone cites and the study that actually tested the claim — is the entire story, and it is worth walking through slowly, because you can watch a myth get built out of one real result and one missing step.

What the 2009 study actually did

At a rugby institute in South Africa, twenty college-aged players took either creatine or a placebo, double-blind, swapping over after a six-week gap [1].

The dose was aggressive: 25 grams a day for a week — five times what anyone takes as a normal daily amount — then 5 g a day for a fortnight.

Their testosterone did not budge. But a related hormone called DHT did. Dihydrotestosterone, if you want the full name; it is what your body converts some testosterone into, and it is a good deal more potent at the receptor it lands on.

DHT went up 56% after the loading week, and was still 40% above where it started two weeks later [1].

Here is the thing about that study, and it is not a criticism of the researchers, who did what they set out to do: they never looked at anybody’s hair. Not with a camera, not with a comb, not with a questionnaire. It was a hormone study. Hair does not appear in it.

Why a hormone reading became a hair scare

The leap is easy to make, and once you see it laid out you will recognise it, because it is the same leap the internet makes every time.

DHT is genuinely central to male pattern baldness. It binds to androgen receptors in hair follicles that happen to be susceptible, and those follicles shrink. That is not folklore — it is why finasteride, the prescription baldness drug, works by blocking the enzyme that makes DHT in the first place.

So you have two true statements. Creatine raised DHT in this study. DHT drives male pattern baldness. Weld them together and you get “creatine causes hair loss,” which sounds like a deduction and is actually a guess.

The 2009 paper even handed the internet its pull quote. Its formal statement of clinical relevance says the mechanisms are “incompletely understood, particularly in relation to dihydrotestosterone, and therefore the long-term clinical safety cannot be guaranteed” [1]. Which is scientists being careful. It reads online as scientists being worried.

Now look at the actual numbers, not the percentage

This is the part that gets left out of every version of this story, and it is the part that should change your mind.

A 56% increase sounds enormous. So what were the numbers?

The creatine group started at 0.98 and peaked at 1.53. The normal range for a healthy adult man is 0.8 to 3.5 [2].

So the frightening percentage describes a move from the very bottom of normal to somewhere in the lower-middle of normal. Every reading, in both groups, at every time point, sat inside the reference range [2].

And notice something else: the creatine group happened to start lower than the placebo group, which began at 1.26 [2]. When your starting number is small, a 56% rise is arithmetically cheap. Percentages are wonderful for making a small absolute change look like an event.

None of this means the 2009 result was wrong. It means the result was always smaller than the headline it produced.

Sixteen years later, somebody finally checked the hair

In 2025 a group ran the study that should have been run in 2010 [2].

Forty-five resistance-trained men aged 18 to 40, randomly assigned to 5 grams of creatine a day or 5 grams of maltodextrin that looked identical, for twelve weeks. Thirty-eight finished. The people assessing the results did not know who was in which group until the analysis was done.

They measured total testosterone, free testosterone and DHT, same as before. And then they did the thing nobody had done: they examined the hair. A trichogram — pulling and examining a small sample of hairs under magnification to see what phase of the growth cycle they are in — plus imaging that counted hair density, follicular units, and total hair thickness. Run with the Iran Skin Research Center supervising.

The result: nothing. No difference between the groups in DHT, in the DHT-to-testosterone ratio, or in any measure of hair [2].

One detail from that trial is a small masterclass in why control groups exist. Over the twelve weeks, total testosterone rose in both groups and free testosterone fell in both groups [2]. Had they run the creatine arm alone, they would have had a headline about creatine and testosterone. With a placebo group, it turned out to be something else entirely — the seasons, the training, who knows — happening to everybody.

Where this verdict could still be wrong

We are rating this Unsupported, so we should be straight about what the 2025 trial does not settle. The authors list these themselves, which we note approvingly.

It did not use a loading phase. The 2009 study gave 25 g a day for a week; the 2025 study gave 5 g a day throughout. The authors name that as one of the differences that could explain why they saw no DHT rise [2]. If you are one of the people who loads, that specific protocol has not been tested against hair.

Twelve weeks is short. Male pattern baldness takes years. The authors say plainly that “a longer term and higher dose creatine study may alter androgens and hair” [2].

They did not record family history. Which matters, because the same paper notes that men whose fathers lost hair are twice as likely to lose it themselves [2]. Thirty-eight men, none screened for genetic susceptibility, is not a test of what happens to the men actually at risk.

And it is one trial. Men only. Blood androgens rather than androgens measured in the hair itself.

So why Unsupported rather than “we cannot tell yet”? Because the claim never had support to lose. In sixteen years, no study has observed hair loss from creatine — not one. The 2009 DHT finding has never been confirmed or refuted by any subsequent research [2], no study has ever linked creatine to hair follicle physiology, and the first study to look directly found nothing. An unsupported claim with a plausible mechanism is still an unsupported claim.

The disclosure you are entitled to

We report funding on every page, and here it matters more than usual.

The 2025 trial was published in the journal of the International Society of Sports Nutrition. One of its authors is the ISSN’s CEO and co-founder — an organisation that, in its own words, “may be sponsored by companies that manufacture, market, and sell dietary supplements” — and is a scientific advisor to creatine brands including Creapure. Another author has received creatine donations from Creapure for research. A third owns a consultancy that sells services to supplement manufacturers [2].

The paper reports no funding for the work itself, and every one of those interests is declared openly in it. But you should know all of it before you decide how much weight to put on a favourable result about a supplement.

Our own read: the conflict is real and the finding still stands, because of how it was done rather than who did it. Randomised, placebo-controlled, assessors blinded, with objective imaging rather than self-report, in collaboration with a dermatology research centre. That is a design that could have produced an inconvenient answer. It did not.

If you want to hold it to a lower weight than an independent replication, that is a reasonable position and the numbers above are all there for you to do it with.

What to actually do

If you are taking creatine and you are worried about your hairline, the honest answer is that nobody has ever caught creatine doing this, and the one group that went looking with a camera came back empty-handed.

If your hair is thinning and you take creatine, the overwhelmingly likely explanation is the one that explains most thinning hair in most men, and it arrived with your genes rather than your supplement tub. That is a conversation for a doctor, not a website — there are actual treatments, and none of them involve stopping creatine.

And if a product is being sold to you specifically as the “hair-safe” alternative, notice what it is charging you for: a fear built out of one hormone reading in twenty rugby players, taken at a dose you almost certainly are not using, that never left the normal range.

We rate creatine Established for safety and have looked separately at whether it does anything for fat loss. This page changes neither.