Login

Guides

Memory and focus: what the trials show

Six supplements with real randomized trials behind them for thinking and memory. What each one actually improved, at what dose, over how long, and which popular claims the evidence does not support.

Last scientifically reviewed: November 22, 2026Editorial review: MDS Labs. External scientific review pending.11 minute read

In brief

  • Six supplements here have randomized human trials behind them. Each one improved something specific, and none of them improved everything.
  • Time is the part most people get wrong. Bacopa needed 12 weeks to show an effect and omega-3 built over roughly 12 months. Nothing here works in a week.
  • Most of the strong results come from people who already had memory problems. Evidence in dementia does not transfer automatically to a healthy adult who wants sharper afternoons.
  • Dose and extract matter more than the plant name. The ginkgo evidence is for one standardized extract at 240 mg a day, not for ginkgo in general.
  • The single best question to ask any memory product: which test improved, in whom, and after how long? If the label cannot answer that, the trial it is pointing at probably did not test what you want.

Why these six and not others

Plenty of ingredients are sold for memory. Far fewer have been through randomized controlled trials, where people are assigned at random to the supplement or to a dummy, and fewer still have enough of those trials to pool together into a meta-analysis.

Everything on this page clears that bar. That is the only reason these six are here, and it is not a claim that they are the best six things you could do for your brain. Sleep, exercise and blood pressure control all have stronger evidence than anything below.

Herbal formulas, rather than single herbs

The largest body of evidence in this area is not about one plant. A meta-analysis pooled 57 randomized trials of traditional East Asian herbal formulas in Alzheimer's disease and found they improved cognitive scores compared with conventional medicine 1.

What the trial shows

Two results inside it are worth more than the headline. The benefit concentrated in formulas containing particular herbs: those containing sweet flag rhizome (Acori Graminei Rhizoma) or Cnidii Rhizoma improved both cognitive scales used, while formulas containing ginseng only trended in the right direction without reaching significance 1.

The second is that these formulas worked alongside standard treatment. Herbal medicine added to conventional care improved scores compared with conventional care alone, and the same held when it was added to donepezil, the most commonly prescribed Alzheimer's drug 1.

What it does not show

These are trials in diagnosed Alzheimer's disease, not in healthy adults. And most of this literature comes from Chinese and Korean journals, where trials often compare against an active treatment rather than a placebo, and where positive results are published more readily than negative ones. The direction is consistent across 57 trials, which counts for something; the individual trials are not all strong.

Ginkgo biloba, as EGb 761

Four randomized placebo-controlled trials covering 782 patients with mild dementia were pooled for the standardized extract EGb 761 at 240 mg a day 2.

What the trial shows

It beat placebo on cognition, on global assessment, on activities of daily living, and on quality of life. Side effects were no more common than with placebo. The effects were described as medium to large, which is unusual in this field 2.

What it does not show

This is evidence for one specific extract at one specific dose, in people with mild dementia. A ginkgo capsule that does not say EGb 761 or an equivalent standardization is not the product that was tested, and a healthy person is not the population that was tested.

Bacopa monnieri

Nine randomized placebo-controlled trials in 518 people were pooled. Every included trial used a standardized extract, dosed for at least 12 weeks, with no other medication alongside it 3.

What the trial shows

Bacopa improved two timed tasks: the Trail Making Test B, which measures mental switching, and choice reaction time. The review also rated the included trials as low risk of bias with good reporting, which is rarer here than it should be 3.

What it does not show

This is speed and mental flexibility rather than memory storage. And the 12-week minimum was a condition of entry into the analysis, so there is no evidence here for taking bacopa for a fortnight and expecting anything.

Huperzine A

Twenty randomized trials in 1,823 participants with Alzheimer's disease found improved cognitive scores at 8, 12 and 16 weeks compared with placebo 4.

What the trial shows

A consistent direction of benefit across a large number of trials and a large number of people.

What it does not show

The same review states that most of the included trials had a high risk of bias. That is the honest frame: many studies pointing the same way, most of them methodologically weak. Huperzine A also acts on the same enzyme as prescription Alzheimer's drugs, which is why it is one of the few supplements here with a genuine interaction risk worth asking a pharmacist about.

Ginseng

Fifteen randomized trials in 671 people were pooled 5.

What the trial shows

A small but statistically significant improvement in memory, larger at higher doses 5.

What it does not show

No positive effect on overall cognition, and none on attention 5. Ginseng is widely sold for focus and energy, and this is the evidence against that particular use. A specific, modest, dose-related memory effect is the real finding.

Omega-3

Twenty-four trials covering 9,660 adults without dementia, followed between three months and three years, make this the largest pool of people on the page 6.

What the trial shows

The benefit landed on executive function, the planning-and-switching side of thinking, and it built gradually over roughly the first 12 months 6.

What it does not show

Nothing dramatic, and nothing quickly. A short trial of omega-3 for memory would be expected to find nothing, and several have.

The doses that were actually studied

If you take one thing from this page, take this table. It is the difference between buying the thing that was tested and buying a product that shares a name with it.

WhatWhat improvedHow long it took
Herbal formulasCognitive scores, including added to donepezilVaries across 57 trials
Ginkgo, EGb 761 at 240 mg/dayCognition, daily function, quality of lifeTrials in mild dementia
Bacopa, standardized extractMental switching, reaction time12 weeks minimum
Huperzine ACognitive scores in Alzheimer's8 to 16 weeks
Ginseng, higher dosesMemory onlyVaries across 15 trials
Omega-3Executive functionBuilds over about 12 months

How to read the scores these trials use

Every result above came from a specific test. Knowing the four common ones lets you check a claim yourself instead of taking a label's word for it.

MMSE is a 30-point screening questionnaire covering orientation, recall, attention and language. Higher is better. It is quick and blunt, good at detecting dementia-range impairment and poor at picking up small changes in a healthy person.

ADAS-cog is a longer, more sensitive scale built for Alzheimer's trials. Lower is better on this one, which catches people out when reading results.

Trail Making Test B is a timed task connecting alternating numbers and letters. It measures mental switching. Faster is better.

SMD, or standardized mean difference, expresses an effect when studies used different tests. Roughly, 0.2 is small, 0.5 is moderate, 0.8 is large. Most supplement results in this area sit at the small end, including the ginseng memory result above.

What this adds up to

There is real evidence here, and it is smaller and slower than the marketing suggests. The pattern across all six is that effects are specific: ginseng moves memory and not attention, bacopa moves switching speed and not memory, omega-3 moves executive function over a year. A product claiming to improve everything is not describing any of these trials.

The most interesting signal is the one about formulas. The largest evidence base on this page belongs to combinations rather than single compounds, and the benefit was additive with standard treatment rather than competing with it. That is the opposite of how the Western supplement market has approached these plants.

If you are healthy and want to think more clearly, the honest ranking puts sleep, exercise and treating high blood pressure above everything on this page. These are worth considering after those, not instead of them.

Keep reading

For the research behind the formula finding, see The West Took the Ingredients and Left the Formulas, and for why combinations behave differently from single compounds, Disease Is Multifactorial.

New or worsening memory problems have many causes, several of them treatable, and they are a reason to see a doctor rather than to start a supplement. Huperzine A and ginkgo both interact with prescription medicines, including blood thinners and Alzheimer's drugs.

Back to Top
Added to cart
Your Cart
0
Close

No products in the cart.