What the evidence says, and how strong it is
20 dossiers. Every claim carries its own grade, its population and its preparation, and several of them are graded as evidence that does not support the claim. Nothing here is for sale and no page links to a seller.
What actually comes first
Fuel opens on diet because that is the honest ranking, not a matter of taste. Anything in the supplements section sits below all four rungs above it, and most of those entries end at “not in people like you”.
- Sleep and recoveryThe one input where a change is reliably noticed within a week, and the one this module keeps pointing back at.
- Nutritional adequacyEating enough, regularly, and not being short of anything. Adequacy first; optimisation is a later argument.
- Physical activityIts evidence is not in this module, and it still outranks everything below it.
- How you learn, train and where you workPractice design and environment. Free, unglamorous, and further up this list than any capsule.
- A targeted supplement, where the evidence justifies oneLast, narrow, and conditional on the four above being handled. Most entries in this module end at "not in people like you".
Brain Diet
Food, drink and timing. Every claim graded, and the best-evidenced line in the whole module is on the caffeine page and costs nothing.
- Sleep and diet
Two dietary decisions have measured effects on sleep: the last caffeine and the last meal. Both make it worse. Nothing cited in this section makes it better, and the entry says so rather than filling the gap.
- Caffeine as a tool
The acute effect on alertness and simple speed is about as well established as anything in this section, and it is narrow. The bill is real: a dose six hours before bed still disrupts measured sleep, and withdrawal has a documented timetable.
- Adequacy before optimisation
Enough food, often enough, before anything clever. The honest evidence here is thin and short-term: a small memory advantage from eating breakfast, nothing settled about what that breakfast should contain, and a protein literature that measures muscle rather than thinking.
- Dietary pattern, not brain foods
The unit of evidence is the whole pattern, not a hero ingredient. One randomised trial in older adults found an advantage; the largest and most recent found none; the association with ultraprocessed food is real and observational. Nobody has run the trial in adults under fifty.
- Hydration
Two meta-analyses, one finding impairment above two per cent body-mass loss and one finding none. Neither studied a normal day at a desk, and neither tested drinking more than thirst asks for. The harm on the other side of that advice is real.
- Meal timing and the afternoon dip
The afternoon dip is real and it is not caused by lunch — it appears in people who skipped the meal and in people who cannot see a clock. A heavy, carbohydrate-rich lunch deepens it. Fasting schedules have nine low-quality trials behind them and no settled cognitive finding.
- Omega-3 from food
Two independent syntheses find no cognitive benefit from omega-3 supplements in people who are not deficient. The advice to eat oily fish stands, on other grounds, from national guidance — and this page grades it against a cognitive outcome, where it has none.
- Micronutrients: who should be tested
The largest review in this section found no vitamin or mineral strategy that maintains cognition in healthy adults. The one trial here that found an effect studied women who were deficient. That difference is the whole entry: correcting a shortfall is not the same claim as topping up.
- What Brain Diet is not
No miracle food, no cleanse, no neuro-branded aisle. Each refusal here is a graded claim with a source behind it rather than a statement of tone — and one refusal the section could not source is left unmade.
Supplements
Nothing here is for sale and nothing here links to a seller. Read these for what the evidence is, which is often less than the label implies.
- Mediterranean and MIND dietary patterns
Two randomised trials of the same idea reach opposite answers, and the difference is what they were compared against. Against ordinary advice the pattern won on two tests; against a matched control diet it won on nothing.
- Caffeine
The best-established acute effect in this section, and the clearest cost. It reliably sharpens alertness and reaction time when you are flagging; it reliably disrupts sleep six hours later.
- Omega-3 (EPA and DHA) supplements
Three high-quality trials, six to forty months, and no cognitive benefit in well-nourished older adults. This is the clearest published null in the section, and the largest safety signal.
- Vitamin D
Two systematic reviews reach different answers about older adults, and neither of them is about a well-nourished thirty-year-old topping up. The useful question here is whether you are low, and that needs a blood test rather than a page.
- Vitamin B12
Sixteen randomised trials in people who are not deficient found nothing. The honest content here is who should have their level checked, and that is a conversation with a clinician rather than a purchase.
- Iron
One good trial in women whose iron was actually low, one Cochrane review saying the cognitive evidence is too conflicting to pool, and a gastrointestinal harm that is better evidenced than the benefit. Nothing here applies without a blood test.
- Magnesium, by form
Two small trials of magnesium L-threonate, both funded by companies that sell it, both with real nulls inside them. For glycinate, citrate and oxide the cognitive literature is not weak — it is missing.
- Creatine monohydrate
Two meta-analyses, one signal, and a subgroup line that decides whether the signal is about you. In the pooled trials the memory effect sits in adults aged 66 to 76 and is flat in adults aged 11 to 31.
- L-theanine with caffeine
Three small acute trials agree that the pair does something to attention-switching that caffeine alone does not, and that L-theanine on its own does nothing useful. Nobody has tested taking it every day.
- Choline sources — citicoline, alpha-GPC and dietary choline
One decent sponsored trial of citicoline in older adults with memory complaints, no adequate trial of alpha-GPC in healthy young adults, and a stroke association in twelve million people that the category does not mention.
- Lion's Mane, erinacine A and Erinamax
Four small human trials, four different preparations, and no trial at all of the branded product most readers are about to buy. This is the dossier where the difference between an ingredient and a product does all the work.
The rest of the evidence hub, and what is planned and not written yet.
How a grade is assigned, and why nothing on these pages is for sale.
The four families, and the question each mode asks first.