A memory supplement guide for people who want the honest version

Grouped by what the research looks like rather than by what the packaging claims.

The row most people find surprising is the last one. The cheapest and least fashionable product on the shelf has the best evidence in the aisle, and it is the one nobody markets as a memory supplement because there is no margin in it.

The row above it is where a good deal of the current category sits, including the product this website sells. Real trials, real amounts, in a field that is not memory.

Almost every product in this aisle has a mechanism, and most of the mechanisms are real. Nitric oxide really does relax blood vessels. NAD really is required for metabolism. Carnosine really is present in brain tissue.

A mechanism tells you why something might work. It does not tell you whether it does, at what amount, in whom, or by how much, and the history of medicine is largely a list of excellent mechanisms that failed when somebody finally measured them.

The niacin story on this label is a good illustration in both directions. Severe deficiency causes a dementia, the observational data associate higher dietary intake with lower risk (the Chicago Health and Aging Project analysis), and high-dose niacin as a cardiovascular drug was tested properly and the trial was stopped for harm (the HPS2-THRIVE trial). Same nutrient, three different questions, three different answers.

The practical rule for a shopper: when a page describes a mechanism at length and a result briefly or not at all, the mechanism is doing the selling.

A capsule is a small container. The product this website sells fits 665.5 mg of listed ingredients into one, by its own panel, and that total is shared across every ingredient in the formula, with the flow agents that get the powder through a filling machine on top.

Set that against the amounts in the published trials of the ingredients this aisle uses: 12 g a day of arginine AKG, 8.8 g of citrulline HCl, 2.4 g of beta-alanine, 2 g of carnosine. One of those does not fit in a capsule, let alone four of them together.

Which is why a formula's ingredient count is a poor guide to its strength and often an inverse one. Six names in one capsule means six small fractions; two names might mean two meaningful ones. The panel is where you find out which you have: on the product this website sells it shows five amino acids sharing 640 mg, each between 40 and 200 mg.

Niacin is the honourable exception and worth understanding as the pattern to look for. Its useful amount is milligrams and the US upper limit is 35 mg a day (the 2017 niacin upper-intake review), so a capsule can carry a real dose of it comfortably. Vitamins and minerals generally can. Gram-scale amino acids generally cannot.

Modest, slow and hard to attribute. The multivitamin trial that is the strongest result in this aisle reported an improvement in memory scores over three years (the COSMOS-Web multivitamin trial), which is a real finding and is nothing like the transformation the category's advertising implies.

A reader who expects to notice a supplement working the way they notice caffeine working will be disappointed by everything in this aisle, including the products that do something.

The honest framing is that these are small additions with long horizons, judged over months, on top of the things that actually move the needle. Anyone selling you a different framing is selling you the framing rather than the product.

And if what prompted the shopping trip was a change somebody has noticed, the first step is not this aisle at all. MedlinePlus sets out what is worth reporting.