Memory Supplement Guide: What The Evidence Shows

The complaint is familiar enough to be a cliche: a name that will not come, a room entered for a reason that has evaporated, a thread lost halfway through a sentence. Most people reaching for a capsule are between mildly irritated and quietly worried, and the category is priced for the second feeling.

The first thing worth separating is ordinary from otherwise. Forgetting where the keys are is ordinary. Forgetting what the keys are for is not. Getting lost somewhere familiar, losing common words rather than proper nouns, or a family member noticing before you do are all reasons for an assessment rather than a purchase.

That matters commercially as well as clinically. Sleep apnoea, untreated hearing loss, thyroid function, blood pressure, alcohol, depression and a list of common medicines all move memory, all are found by asking, and several are fixable. A year of capsules does not rule a single one of them out.

Open almost any nootropic panel and the same four or five names appear: arginine in one or two salts, citrulline in one or two salts, sometimes beta-alanine, occasionally a nitrate source. None of those is a brain chemical. All of them feed one pathway.

Nitric oxide is made from arginine by nitric oxide synthase, it relaxes the smooth muscle in vessel walls, and vessels that relax carry more blood. That is settled physiology, and a 2024 review of endothelial nitric oxide at the neurovascular unit describes roles beyond flow regulation as well.

The commercial leap is the next sentence rather than that one. From "nitric oxide regulates cerebral blood flow" to "this capsule will sharpen your memory" are two large steps: the capsule has to raise nitric oxide meaningfully, and the raised flow has to change cognition. Each step has been tested.

The first step has been tested and it works, at a price. Schwedhelm’s dosing study established the amount that moves plasma arginine at all: 3 g of citrulline twice daily. Below that the curve is flat, which is the least often printed number in this category.

The second step is where the category runs out of road. Wightman’s crossover study raised dietary nitrate acutely and did move cerebral blood flow parameters and some cognitive performance measures, which is the result every stack in this aisle is built on. Then a thirteen-week randomised trial in older adults gave incremental doses of nitrate-rich beetroot juice and changed neither cognitive function nor cerebral blood flow.

Read those two together and the honest position is that an acute signal exists and a sustained one has not been demonstrated in the population these products are sold to. That is not the same as "it does nothing". It is a reason to hold the mechanism loosely and to look hard at amounts.

Arginine is the substrate. Give the enzyme more of it and, in principle, it makes more nitric oxide. The literature is large, old and mostly cardiovascular.

A meta-analysis of arginine and endothelial function in people with cardiovascular or metabolic disease found the markers move, and a 2020 review of arginine and the endothelium covers the biology behind it. This is a real ingredient with a real mechanism and it is not snake oil.

The amounts are the problem. VINTAGE MI, the best-known cardiac trial, ran 3 g three times a day, which is 9 g. The arginine alpha-ketoglutarate pharmacokinetic and training study ran 4 g three times a day, which is 12 g. Those are teaspoons rather than capsules.

For cognition specifically the evidence is thinner and newer. A 2026 pilot in older adults with cognitive impairment gave 6.4 g a day for thirty days, randomised against placebo, and reported favourably. It is a pilot, in one hospital, with thirty-five completers, and it is the closest human result this ingredient has to a memory claim.

There is one acute result worth naming because it is often cited as though it were the same thing. An inositol-stabilised arginine silicate study improved cognitive outcomes in healthy adults at 1.5 g, measured over an hour. A different salt, a smaller amount and a one-hour window: interesting, and not transferable to a daily capsule taken for a month. Below about a gram a day of arginine, nothing in the published record is relevant at all, and labels in this category rarely reach that floor.

Citrulline is the counter-intuitive one. Swallowed arginine is largely destroyed in the gut and the liver before it reaches the circulation. Citrulline slips past both, is converted to arginine in the kidney, and raises plasma arginine more reliably than arginine itself does.

That is why the better-designed products use it, and the pharmacokinetic work that established it is also where the dose floor comes from: 3 g twice daily. The same paper is the reason nobody serious recommends arginine powder any more.

The outcome literature sits where you would expect. A trial in hypertensive postmenopausal women improved arterial blood flow and muscle oxygenation during handgrip exercise at 10 g a day for four weeks. A 2025 crossover in middle-aged and older adults with type 2 diabetes improved microvascular function and muscle strength at 6 g a day. A 2025 meta-analysis of fifteen trials covers blood pressure across the same dose range.

Citrulline malate is a different row and deserves to be read as one. A randomised crossover comparing the two forms directly used 5.3 g of citrulline with 2.7 g of malate, and a three-day repeated-sprint trial ran 8 g a day. Both are sports studies, which is where this form lives.

For memory the human evidence is essentially absent. The one suggestive result is a mouse study of citrulline and spatial memory in an Alzheimer's model, which is a reason to run a human trial rather than a reason to buy a capsule. It is cited here because leaving it out would be incomplete.

One safety note belongs with both amino acids and it is the same note. A review covering oral arginine and citrulline together summarises their effect on blood pressure, and it is downwards. Anyone already taking something for blood pressure should mention either ingredient to whoever prescribes it, and that is a conversation rather than a warning.

Beta-alanine is not a vasodilator and it does not belong to the nitric-oxide story at all. It is the rate-limiting precursor of carnosine, a dipeptide that accumulates in muscle and in brain tissue, and it reached this category from sports nutrition.

The International Society of Sports Nutrition position stand sets the standard protocol at 4 to 6 g a day for two to four weeks. A Bayesian dose-response meta-analysis models how muscle carnosine rises against that, and a comparison of two dosing protocols ran 6 g against 12 g in a sustained-release form. Every one of those figures is in grams.

The reason it appears on memory labels is more interesting than most ingredient stories in this aisle. A double-blind randomised trial in older adults gave 2.4 g a day for ten weeks and reported effects on cognitive function, mood and physical function. A follow-up imaging study by the same group reported improved fractional anisotropy in the hippocampus and amygdala in 60 to 80-year-olds. That is a cognition result, in the right population, with a plausible tissue mechanism behind it.

The counterweight is equally real and belongs in the same paragraph rather than a footnote. A fourteen-week trial combining beta-alanine with high-intensity interval training at 6.4 g a day improved neither Stroop performance nor serum brain-derived neurotrophic factor. A high-dose short-duration study at 12 g a day before simulated military operational stress is a third data point in a third direction.

A 2024 meta-analysis of histidine-containing dipeptides and delayed recall is the broadest summary available and is cautiously positive. This is the most promising cognition story in the category and it is still a handful of small trials pulling in different directions.

A systematic risk assessment covers safety. The characteristic effect is paraesthesia, the harmless skin tingling that arrives with a gram-scale dose and passes, and it is the clearest signal in this category that a real amount has been swallowed. Its absence is informative.

Niacin looks like filler on a cognition panel. It is the row with the smallest number and the least exciting name, and it is also the only ingredient in this category whose cognition literature was collected in ordinary older people going about their lives.

The Chicago Health and Aging Project followed dietary niacin intake against incident Alzheimer's disease and cognitive decline in a population cohort. It is observational, it measured niacin eaten as food over years rather than swallowed as a capsule, and it cannot establish cause. It is also the kind of evidence that no amino acid on these panels has at all.

The mechanism is not mysterious. Niacin is the precursor of NAD, the coenzyme every cell uses to run energy metabolism, and a randomised trial of nicotinamide riboside gave 1 g a day to older adults with mild cognitive impairment on exactly that reasoning. A different precursor at a far larger amount: the shape of the hypothesis rather than support for a vitamin row.

Two things are worth knowing before treating niacin as harmless because it is a vitamin. A systematic review of its tolerable upper intake exists because high-dose nicotinic acid has dose-dependent effects, the flushing response is a prostaglandin-mediated event that arrives at hundreds of milligrams, and a case report of acute hepatotoxicity leading to emergency liver transplantation is the extreme end of what gram-scale niacin has done.

None of that is about a vitamin-sized row. A supplement providing something under half the Daily Value is a nutrient amount rather than a pharmacological one, and neither the flush nor the liver signal arises there. It is worth knowing because niacin is one of the few ingredients in this aisle where more is genuinely worse: AIM-HIGH and HPS2-THRIVE are the reason cardiology stopped prescribing the high-dose form. The unit is also worth a sentence, because niacin is printed in equivalents to account for the amount the body makes from tryptophan, and the conversion ratio study is where that arithmetic comes from.

Calcium turns up on cognition panels for reasons that are usually formulation rather than pharmacology: it is a flow agent, a bulking agent and an easy row to add. Where it appears at 1 or 2 per cent of the Daily Value it is doing a manufacturing job.

There is a small literature to be honest about. A Gothenburg population study found an association between calcium supplementation and dementia risk in women with cerebrovascular disease, which is a signal running in the unhelpful direction and is worth knowing rather than hiding. It concerns supplemental doses measured in hundreds of milligrams.

Recker’s classic absorption study gave a 250 mg calcium load and is the reference point for how calcium carbonate is taken up at all. A row at a small fraction of that sits below the amount the absorption literature was built to measure, which is the shape of the whole problem. The rule a shopper can take away applies well beyond calcium: a row supplying a small percentage of a Daily Value is present for a formulation reason, and reading it as an active is a mistake the panel layout encourages.

Dietary supplements in the United States are not approved before sale. A product may make a structure and function claim, meaning that it supports a normal body function, and may not claim to diagnose, treat, cure or prevent a disease. Alzheimer's disease and every other dementia are diseases, so no supplement may claim to touch them.

That single rule explains most of the strange language in this aisle. "Supports memory and cognitive function" is what a company writes when the sentence it wants is unavailable to it. The phrasing is a legal ceiling rather than clumsiness.

The corollary is the cheapest quality filter a buyer has. Any product promising to restore memory, reverse decline or prevent dementia has stepped outside that ceiling, and a company casual about the claims rule is not a safe bet on sourcing either.

It also reframes the interesting question. Every label here claims roughly the same thing, so the claim carries almost no information. What carries information is the panel: which names, at what amounts, in which salts, against which trials.

• Get a memory change looked at once. It answers a different question from the one a capsule answers, and several of the answers are treatable.
• Judge a panel on its amounts, not its ingredient count. Seven names is not seven pieces of evidence.
• Prefer a label that prints every figure. A proprietary blend cannot be checked against any of the research above, and the checking is the only power a buyer has here.
• Hold the vasodilator names to a gram-scale standard. The dose floor for citrulline is published and it is 3 g twice daily.
• Do not dismiss the small rows. The one ingredient in this category with cognition data in older adults is a B vitamin, and it is always the least advertised row on the panel.
• Give anything here eight weeks and write down one concrete task at the start, because memory judged by memory is the least reliable measurement in the category.

The companion page turns that into seven checks with a pass or fail against each, and then applies all seven to Mind Honey Pro, the product this website sells. It fails two of them. The scorecard is here.