alt Hacker NewsThe article made up the claim it’s not from the paper itself.
There was some improvement in cognitive scores, but no placebo group. Without a placebo group, there are a lot of explanations for the data.
Replies
Yes, I can't find a 30%-slowdown number either.
I'll add to this: the referenced trial occurred over 8 weeks, so even if we stipulate that the improvements in cognition (which are dubious, as tgv points out in this comment: https://news.ycombinator.com/item?id=48347906) are due to treatment rather than some other effect, we don't know that the effect is disease-modifying as opposed to symptomatic. As with acetylcholinesterase inhibitors, it may just be having a cognition-enhancing effect which, nevertheless, does not alter the underlying disease trajectory (i.e. just shifting the declining trajectory up vertically by a constant amount), and might revert shortly after discontinuing use of the drug.
A controlled trial, over a much longer duration, and ideally with a wash-out period, would be necessary to identify a disease-modifying effect.
FWIW creatine is "one of the most studied supplements for muscle and strength".
But at the same time "creatine’s brain benefits aren’t as exciting as social media makes them out to be. The research at this point just doesn’t support the hype".
Source: https://physiqonomics.com/creatine-cognitive-performance/
I think you're right.
Confusingly they reference a 2026 article (which isn't included in their citation list) that allegedly includes "placebo-controlled trial", but I think it might just be [0], which is based on the same single-arm trial. If they do have a paper using a placebo-controlled trial, they should definitely include that citation.
> Without a placebo group...
You just needed to read the next paragraph:
"The 2026 multicenter placebo-controlled trial extending this work..."
They also quote a follow up study that sounds more compelling:
> The 2026 multicenter placebo-controlled trial extending this work enrolled 240 participants with early Alzheimer’s... The intervention group showed slower decline on standard cognitive scales by about 30% versus placebo.
But there's no such study in the references section. Not sure what's going on there but I want to see the data before I believe this.
Normally I wouldn't ask this, but having seen the effects of Alzheimer's I must ask: is there any evidence that taking creatine will harm my brain?
Let's be real: I take a lot of edibles. I smoke sometimes on a hike. Every once in a while I get a Guiness and a gyro. My health is by no means perfect, and if I'm willing to take in literal poison (yes, alchohol is that dangerous that I say such things even about my precious Guiness).
Anyways, for years I've been mostly skeptical of vitamins. I've heard a megadose of Vitamin C can shorten the symptoms of a cold, but a "megadose" is relatively small enough that just some OJ can do that, no need for pills.
But if I'm not giving up my Guiness and gyros, I probably should be willing to be more flexible about my "no vitamins or weird supplements rule".
So TL;DR: Let's flip this around: What are the risks of creatine, presuming a safe supply chain? (It's legal where I am as far as I know, which I'm a fan of for most things since then you can get a receipt and there will be some authority that investigates if you're unlucky to get a "bad batch".)
Lots of creatine gummy dropshippers trying to gain ground.
I can't speak for the Alzheimer piece, but there are other studies about supporting the brain.
Worth looking at the range of studies as well.
>> Recently, a pilot study (single-arm) by Smith et al., recruited 20 patients (73 years of age) with AD and provided them with 20 grams/day of CrM for 8 weeks [20]. Serum creatine levels were increased at weeks 4 and 8 (p < 0.001), and total brain creatine levels (as measured by H-MRS) increased by 11% (p < 0.001). Clinically, there were demonstrated improvements in cognition on global (p = 0.02) and fluid composites (p = 0.004), as well as List Sorting (p = 0.001), Oral Reading (p < 0.001) and Flanker tests (p = 0.05).
Yeah 20 patients is not a lot. I'm inferring this is a pre-post test. However some of those p-values are pretty good (.001 on reading and and sorting). Very promising pilot study but not conclusive imo.