Creatine’s brain grade just dropped

LONGEVITY LATEST ISSUE 36 · 11 NOVEMBER 2026

LONGEVITY LATEST

The Evidence-Based Edge on Living Longer and Better

Issue 36 · The Cheapest A · 11 November 2026

Welcome

In April, Issue 07 gave creatine for the brain a B. This week I’m taking that B back. No new trial overturned it. The statistical problem was already in front of me in April: I flagged it and kept the grade. That was the wrong call.

Creatine is still one of the best-tested supplements this newsletter has graded, at about 12p a day for plain powder. Cheap and safe is exactly what makes every other claim about it easy to believe, so those claims need checking harder, not less.

Inside: six claims graded, where the brain evidence holds up best (in sleep-deprivation lab trials), a two-year bone trial, and a lab test in which four of six gummy brands failed.

SPOTLIGHT

Six creatine claims, graded

My reading: the strongest claim is the least glamorous. Creatine does its best work next to a barbell.

Xu et al. Front Nutr (2024) · Prokopidis et al. Nutr Rev (2023) · Sandkühler et al. BMC Med (2023) · Gordji-Nejad et al. Sci Rep (2024) · Gordji-Nejad et al. Nutrients (2026) · Chilibeck et al. Open Access J Sports Med (2017) · Smith et al. Alzheimer’s Dement TRCI (2025) · Chilibeck et al. Med Sci Sports Exerc (2023) · SuppCo. Creatine lab testing report (2025)

THIS WEEK’S ANALYSIS

Top 3 Interventions Under the Microscope

Each grade covers the named outcome only. A = consistent randomised trials; B = positive but few, small or narrow trials; C = mixed or preliminary; D = mostly lab and animal data; F = claims far outrun the data.

1. Creatine for the brain | C rested, B after a night awake

Down from B in Issue 07. The correction is to my judgement, not to the data.

What changed. Issue 07 leaned on Xu et al., a 2024 meta-analysis of 16 trials. Tom Citherlet’s April commentary shows it counted several memory subtests from the same people as separate results, which makes the finding look far more precise than it is. I knew about that problem and kept the B. Separately, when the same correction was applied to an earlier review, Prokopidis et al., the memory effect held only in older adults, where the trials disagreed sharply.

The dose people actually take. The cleanest test is a preregistered 2023 crossover trial: 123 adults, 5 g a day for six weeks. Memory narrowly missed significance, reasoning didn’t move, and the authors concluded creatine might have a small benefit. Vegetarians did no better than meat-eaters.

Where it holds best. Sleep deprivation, in the lab. In 2024 a Jülich team kept 15 healthy adults awake overnight and gave one dose of 0.35 g per kg (about 25 g at 70 kg) or placebo. Processing and short-term memory held up better on creatine for several hours. In April the group repeated it in 29 people at about 14 g. The benefit was smaller, and after correcting for the number of tests, only one logic measure clearly held. Both trials were small, from one lab, at three to five times the daily dose. The authors advised against trying it at home.

If you are 40 to 60. On 3–5 g a day and sleeping normally, the best trial points to a small, uncertain memory effect at most. A 2026 trial in 45- to 65-year-olds on 10 g reported some cognitive improvements, but its authors call them exploratory. Take creatine for your muscles; treat a brain benefit as possible but unproven.

Which of six reader types you are, and the blood test creatine can fool: read the companion →

Xu et al. Front Nutr (2024) · Citherlet. Front Nutr (2026) · Prokopidis et al. Nutr Rev (2023) · Sandkühler et al. BMC Med (2023) · Gordji-Nejad et al. Sci Rep (2024) · Gordji-Nejad et al. Nutrients (2026) · Chun et al. J Int Soc Sports Nutr (2026)

2. Creatine with resistance training | Grade A

The question: does it add anything to the training itself?

How it works. Creatine tops up phosphocreatine, the fast-access energy muscles burn in the first seconds of a hard set. A fuller store buys a rep or two more per set, and over months that extra work adds up. A bigger fuel tank, not a better engine.

The evidence. Philip Chilibeck and colleagues pooled 22 randomised trials in 721 lifting adults with average ages of 57 to 70. Those also taking creatine gained an extra 1.37 kg of lean tissue and small but significant strength gains. Later reviews in 2021 and 2025 landed in the same place.

Fair warning. Lean tissue isn’t the same as new muscle: part of that gain is water held inside muscle, which scans count as lean. Several 2021 review authors disclose ties to creatine makers. And without training the evidence is thinner. A 2026 Texas A&M trial in 45- to 65-year-olds saw lean tissue and leg-press gains at 10 g a day without exercise, but with 13 people finishing in each of those groups and a creatine maker supplying the product. Interesting, not settled.

Chilibeck et al. Open Access J Sports Med (2017) · Forbes et al. Nutrients (2021) · Sharian et al. Eur Rev Aging Phys Act (2025) · Chun et al. J Int Soc Sports Nutr (2026)

3. Creatine for bone after menopause | Grade C

Bone density, no. Bone shape, possibly.

The trial. Can creatine protect bone after menopause? The best test comes from Saskatchewan: Philip Chilibeck, Darren Candow and colleagues gave 237 postmenopausal women, average age 59, creatine (0.14 g per kg a day, about 10 g at 70 kg) or placebo for two years, while everyone lifted three times a week and walked most days.

Hip density, the main outcome, fell slightly in both groups, with no significant benefit from creatine; the spine looked the same. What differed was geometry: the creatine group held on to the estimated bending strength of the femoral neck while placebo’s fell. They also covered an 80-metre walk about 1.5 seconds faster.

What holds it back. About 56% of doses were taken and 61% of sessions attended. Hip geometry is a scan estimate, not a fracture count, and the lean-tissue gain appeared only among women who completed the trial. If bone is your reason, lifting is the intervention. Creatine is a passenger.

Chilibeck et al. Med Sci Sports Exerc (2023)

HYPE CHECK

“Same creatine, now in a sweet”

The claim. Creatine gummies: no scooping, no grit, just chew. Create’s 90-gummy tub, a month at 4.5 g a day, costs $67 ($58 on subscription).

The evidence. In June 2025, SuppCo, an app that rates supplements, bought 11 creatine products off the shelf and sent them to an ISO 17025-accredited lab. All five powders landed within 98–101% of their label. Of six gummies, four failed and two passed. Two of the failures held traces: you’d need about 800 of one brand, or 2,000 of another, for a single 5 g dose.

The other two failures contained no measurable creatine.

Why it misleads. Creatine breaks down into creatinine when kept warm, wet and acidic, a fair description of a gummy. Even the two that passed, Create and Force Factor, carried high creatinine. One caveat: SuppCo sells a paid membership and didn’t name its lab, and this was one test of six brands, so I’d like a second lab’s numbers. The powders sailed through.

Our verdict. Don’t assume a gummy holds what its label says. If you genuinely can’t face powder, buy only a brand that publishes batch tests. Otherwise, 500 g of plain monohydrate costs £12–18 and lasts 100 days.

SuppCo. Creatine lab testing report (2025) · Create. 90-count gummies listing (2026) · PriceSpy UK. 500 g monohydrate price history (2026)

IN BRIEF

“Will it make my hair fall out?” The fear traces to a 2009 study in which 20 rugby players took 25 g a day for a week and their DHT, the hormone behind male-pattern hair loss, rose 56%. Nobody measured their hair. Last year a trial did: 38 trained men took 5 g a day or placebo for 12 weeks, with no differences in DHT, hair density, follicle counts or thickness. Twelve weeks isn’t a decade, but the myth rests on a study that never looked at a single hair.

van der Merwe et al. Clin J Sport Med (2009) · Ashtary-Larky et al. J Int Soc Sports Nutr (2025)

THIS WEEK’S DEEP DIVE

The Brain Is Last in the Queue

Your brain makes most of its own creatine and lets little in from the blood. In supplement studies, brain levels rose about half as much as muscle, and the clearest effects came at four to five times the usual dose.

So is your 5 g a day doing anything for your head? The companion follows creatine from bloodstream to neuron, explains why a few kinds of reader look different in the data, and opens with a short decision map. Most readers fit more than one row. It also covers the routine blood test creatine can make look like kidney trouble.

Read the companion: The Brain Is Last in the Queue →

BIOHACKING CORNER · THE PLAN LIVES HERE

Buy the boring version. Plain creatine monohydrate powder, ideally batch-tested. It’s what nearly every trial used; pricier forms haven’t beaten it.

Take 3–5 g a day. Any time, with food if your stomach complains. Skip the loading week: in a 1996 trial of 31 men, 3 g a day reached the same muscle levels as 20 g a day, in four weeks rather than six days.

Pair it with lifting. Twice a week, as in Issues 34 and 35. The A belongs to the combination, not the tub.

Ignore the first fortnight on the scale. Early gain is mostly water. Judge it on what you lift at 12 weeks.

Tell whoever orders your blood tests. Creatine can raise one kidney marker without any change in your kidneys; the Deep Dive explains why.

Hultman et al. J Appl Physiol (1996) · Chilibeck et al. Open Access J Sports Med (2017)

READER PULSE

Pick one: where do you stand on creatine?

I take it daily

On and off

I tried it and stopped

Never: I don’t see the point

Never: I’m worried about kidneys or hair

If the two ‘worried’ and ‘stopped’ answers outnumber the daily takers, the kidney question gets its own Hype Check before Christmas.

Production: replace this list with the live single-choice beehiiv poll before sending.

CLOSING

Know which claim you’re buying

Creatine’s evidence beats its marketing, as long as you know which claim you’re paying for. For muscle, it’s an A with a barbell attached. For the brain, a narrow B for lab-tested sleepless nights and an open question otherwise. Next week: NAD+ boosters, three years and several human trials after Issue 02, and why the IV drip still costs what it does.

Stay curious and stay healthy!
— Christian Thomsen, Editor

Longevity Latest is published weekly by FrontWave Media Ltd. Educational content, not personal medical advice. If you have kidney disease, diabetes affecting your kidneys, or are pregnant or breastfeeding, speak to your GP before taking creatine. Creatine can raise blood creatinine, which is used to estimate kidney function; tell your doctor you take it before blood tests. The high single doses in the sleep-deprivation trials were given under supervision and are not a home protocol. Check with your doctor before taking supplements alongside prescribed medicines.

Reading notes

Evidence checked through 8 November 2026; each claim links to its source beside it. Xu et al. (2024) is the meta-analysis Citherlet critiqued; in it, one trial contributed at least seven memory subtests. The older-adult result comes from a reanalysis of a separate review, Prokopidis et al. (2023). Sandkühler et al. (2023) was a preregistered, double-blind crossover; its Bayesian analysis favoured a small benefit. The 2026 Jülich trial reported several improvements, but only the pooled logic result passed its correction for multiple comparisons. The Texas A&M trial (Chun et al., 2026) randomised creatine versus placebo within self-selected exercise and non-exercise groups; its cognitive results were labelled exploratory, a creatine manufacturer supplied the product, and the senior author chairs that company’s scientific advisory board. Lean-tissue figures are pooled mean differences versus placebo with the same training. The Alzheimer’s pilot was single-arm and assessed feasibility, brain creatine and cognitive scores, not disease progression. In the bone trial, femoral-neck density fell by about 2% in both groups; the walking and section-modulus results were secondary, and the lean-tissue result came from a completer analysis. The gummy test was commissioned by SuppCo and covered one purchase per product for most brands. Prices: US and UK retail listings, November 2026.