Cognitive Supplement Guide: What The Evidence Shows

Four ingredient families are covered here, and they are the four that turn up on amino-acid cognitive formulas: arginine and its alpha-ketoglutarate salt, citrulline in its two common salts, beta-alanine and the carnosine it builds, and the B3 and NAD route.

Five families are deliberately left alone: bacopa, ginkgo, lion’s mane, citicoline and phosphatidylserine. Each has a literature and none of it was verified against a live source in the session that assembled this site’s references. Writing about their evidence would mean citing from memory, and a citation written from memory is the one failure mode this desk refuses outright.

So this guide is narrower than a complete survey of the category and it is checkable line by line, which is the trade worth making. A reader comparing a botanical formula should look for a guide whose author verified those sources.

One organising idea runs through all four sections. In this category the interesting question is almost never whether an ingredient does anything. It is whether the amount on the label is anywhere near the amount that did it.

L-arginine is the direct substrate for nitric oxide synthase. Nitric oxide relaxes the smooth muscle around a blood vessel, and a relaxed vessel carries more blood at lower pressure. That is the mechanism the whole family rests on and it is not in dispute.

The dose at which swallowing arginine does that is the useful number. A 2022 dose-response meta-analysis in Advances in Nutrition pooled the randomised trials and found blood pressure moved at four grams a day and above, with nothing detectable below it and nothing further above about nine grams. A 2026 reassessment using ambulatory blood pressure monitoring revisited the question with better measurement.

Citrulline reaches the same chemistry by a better route. Swallowed arginine is largely stripped out by arginase before it reaches the circulation; citrulline passes that checkpoint and is converted to arginine in the kidney. Schwedhelm's 2008 pharmacokinetic study tested a ladder of doses and found that only the top rung, three grams twice a day, lifted the plasma arginine trough.

The alpha-ketoglutarate salt has a thinner and older file. The eight-week pharmacokinetic and safety study in trained men used twelve grams a day and the seven-day blood-flow trial used the same, while three grams taken acutely and 3,700 mg per serving both failed to move anything.

Citrulline malate is the performance form. The meta-analysis of repetitions during strength training covers six to eight grams taken forty to sixty minutes beforehand, and the review of perceived exertion and soreness reports eight grams as the commonest dose. Against all of that, eight days of eight grams a day of either amino acid did not improve a swimmer’s time trial.

Two further findings are worth having. A gram and a half of citrulline a day for a month improved erection hardness in men with mild erectile dysfunction, and 800 mg paired with trans-resveratrol is the lowest citrulline dose with a positive human result in this search. Both describe vascular endpoints, which is what this family is.

The one study that reaches cognition is an acute trial of inositol-stabilised arginine silicate, which reported improved cognitive outcomes in healthy adults at 1.5 g of a stabilised compound. It is a real result, it is one study, and the compound is not the free base that appears on most labels.

Beta-alanine has nothing to do with nitric oxide. It is the rate-limiting precursor for carnosine, a dipeptide stored in muscle and, in smaller quantities, in brain tissue, where it acts as a buffer and an antioxidant.

The International Society of Sports Nutrition position stand sets the working dose at four to six grams a day for at least four weeks, and notes that the tingling most people associate with the ingredient is managed by splitting intake into 1.6 g portions. A comparison of two loading protocols found twelve grams a day of a sustained-release form accelerated carnosine loading, and a Bayesian dose-response analysis modelled the relationship across the literature.

The cognitive results are genuinely mixed and both directions deserve printing. A double-blind randomised study in older adults at 2.4 g a day reported changes in cognitive function, mood and physical function; an imaging follow-up found altered fractional anisotropy in the hippocampus and amygdala at the same dose; and an executive-function trial in middle-aged adults used 2.4 g for twenty-eight days.

Against those, a trial in combat soldiers found tactical performance improved while cognitive function did not, with a follow-up confirming muscle carnosine had risen. Twelve grams a day for a fortnight was tested before simulated operational stress, and fourteen weeks alongside interval training moved neither serum BDNF nor Stroop performance.

The strongest thread in this whole guide runs through carnosine itself rather than its precursor. A 2024 meta-analysis of histidine-containing dipeptides found improved delayed recall, and the NEAT trial gave two grams of carnosine daily for up to twelve weeks with improvements in the younger participants.

That distinction is the practical lesson of this family. Supplying carnosine directly and supplying the precursor that builds it over weeks are different propositions with different evidence, and a label carrying beta-alanine is doing the second.

Niacin is converted into nicotinamide adenine dinucleotide, the coenzyme that carries electrons through cellular energy production. Neurons are metabolically expensive and their demand for it is continuous, which is the honest reason B3 appears on cognitive labels.

The cognition research in this area does not use niacin. It uses nicotinamide riboside, a different compound, at doses between one and two grams a day. A randomised placebo-controlled trial in older adults with mild cognitive impairment escalated to a gram a day for ten weeks. A 2025 trial in subjective cognitive decline used the same gram a day for eight weeks. A long-COVID trial used two grams.

Niacin’s own large literature is cardiovascular and also runs in grams. HPS2-THRIVE gave 25,673 adults two grams a day of an extended-release form, and an FDA Mini-Sentinel assessment examined prescription-strength niacin in routine use.

At food-level amounts none of that applies and neither does the flush. The 2024 review of the flushing mechanism describes a prostaglandin-mediated response to doses in the hundreds of milligrams, so a capsule carrying single-figure milligrams produces nothing a person will notice.

The reasonable reading is that a niacin row on a cognitive label is a nutritional contribution rather than an active intervention. That is a modest thing and a real one, and it is the row most likely on any such label to be at an amount the reference figures actually support.

Small mineral rows turn up constantly on this shelf and are almost never doing nutritional work. Their practical significance is whether they interfere with something else you take.

Calcium is the one that matters, because calcium carbonate binds levothyroxine in the gut. The threshold is what a reader needs and it is available: the work that first established the interaction ran at 1,200 mg of the element daily, and a later comparison across three calcium salts halved that to 500 mg and still found it.

That gives a reader a usable threshold rather than a vague caution. A label carrying tens of milligrams of calcium is far below anything that has been tested for interference; a label carrying hundreds is in the territory the studies covered and is worth separating from a thyroid dose.

The general lesson applies beyond calcium. When a mineral row appears at a fraction of a Daily Value, the right question is not what it contributes but whether it collides with anything, and the answer is usually available from the dose in the interaction literature rather than from the label.

• Find the serving size first. Every other figure on the panel is per serving, and a three-capsule serving makes a panel look three times stronger than a one-capsule one.
• Look for a proprietary blend. A blend total with no individual amounts means nothing below it can be compared with anything, and the rest of this exercise stops there.
• Read the form, not just the name. Arginine base and arginine AKG are different rows with different literatures, and an extract weight is not an active weight.
• Add the actives. One number, in milligrams, for the whole serving. It is usually smaller than the front of the pack implies.
• Compare it with the middle column above. One ingredient at a trial dose is worth more than eight at a gesture.
• Read the other-ingredients line. The capsule shell decides whether the product suits a vegetarian diet, and it is the line most often skipped.

Six steps, none of which needs any specialist knowledge, and between them they sort most of this shelf. The one that does the most work is the fourth, because a total expressed in milligrams can be held against a figure expressed in grams without any further thought.

The companion checklist turns the same idea into seven scored questions and applies them to a real label.

Dietary supplements in the United States may make structure and function claims. They may say a product supports cognitive function; they may not say it treats, prevents or cures anything, including any form of cognitive decline.

That boundary is why the phrases on this shelf repeat. Cognitive function, mental clarity, focus and healthy brain function are the vocabulary available, and a product claiming more than that has stepped outside what its category permits.

Every such claim carries the same disclaimer. These statements have not been evaluated by the Food and Drug Administration, and the product is not intended to diagnose, treat, cure or prevent any disease. No supplement in this category is FDA-approved, because none of them is a drug.

The practical consequence for a reader is that the claims on the front of two packs carry almost no information, since both are drawn from the same narrow permitted vocabulary. The panel on the back carries all of it.

A reader who takes only one thing from this guide should take the last line. Ingredient lists in this category are easy to admire and almost never the thing that matters.

The checklist applies that to a real label, and the ingredients page on this site prints the trial dose beside every amount on the product this website sells.