Do multivitamins improve memory—or does age change the answer?

COSMOS found small cognitive benefits in older US adults, while an earlier long trial was neutral. Separate memory-test changes from dementia prevention and advice for younger adults.

Prepared by PoshanSense · AI-assisted educational writing
Our sources and editorial approach · Not clinically reviewed

Complementary human studies · Published 2013 & 2023–2024

You forget why you opened a cupboard, then see a headline promising that a daily multivitamin keeps the brain young. Does that make the bottle a sensible memory strategy? The best answer starts by separating an encouraging research signal from the much broader promise on the shelf.

There is a small benefit worth understanding

The 2024 COSMOS-Clinic report found a small advantage in episodic memory, the ability to recall information and experiences. Its global-cognition estimate favoured multivitamins, but the confidence interval included no difference. Not every cognitive measure improved clearly.

Combining three nonoverlapping COSMOS cognitive samples—5,203 people—gave a global-cognition difference of 0.07 standard-deviation units. That is a small average shift, not a 7% improvement or a promise that a particular person will remember appointments better.

What about the “two years younger” claim?

The authors compared the pooled cognitive difference with the relationship between age and test scores. That statistical translation can make a result easier to picture. It does not mean that a brain reversed two years of ageing, that life expectancy increased, or that dementia was postponed by two years.

Think of it as a scale comparison: researchers placed a score difference alongside another score difference associated with age. Both belong to the analysis. Neither is a birthday that disappeared. Keeping the original outcome in view prevents a helpful explanation from becoming a biological promise.

Why an older trial matters

The Physicians’ Health Study II followed cognitive performance in 5,947 US male physicians aged 65 or older. Long-term multivitamin assignment did not show a benefit in cognitive change. These well-educated men and their testing schedule differed from COSMOS; the initial cognitive assessment also occurred after supplementation had begun.

That does not automatically cancel the newer finding. It does stop us from describing the evidence as uniformly positive. Two trials can disagree because participants, baseline nutrition, formulations, timing or measurements differ. Without a direct comparison, choosing one explanation as the proven cause of the disagreement would go beyond the data.

Memory scores are not dementia cases

A separate COSMOS-Mind analysis did not find a statistically clear reduction in new mild cognitive impairment or probable dementia over three years. There were few dementia events, so the analysis could not rule out smaller or longer-term effects. It equally could not establish prevention.

This is why “promising for some cognitive scores” and “prevents dementia” are different statements. The first has trial support in a particular older population. The second needs evidence about diagnosed outcomes, with enough events and follow-up to answer it reliably.

Would this apply in Europe, India or elsewhere?

The question is globally relevant, but the main cognition trials discussed here were conducted in the US. An older American volunteer's diet, fortified foods, education and access to care may differ from those of a younger reader or an older person elsewhere. A country name alone cannot capture those differences.

The findings do not prove the same benefit for someone in their twenties, someone with a restrictive diet, or someone with an untreated deficiency. They also do not test every multivitamin sold worldwide. A matching word on the label is less informative than matching the formulation and the people actually studied.

Use the evidence without making the pill the whole plan

The NIH supplement guidance explains that multivitamins cannot replace a varied diet and that ingredients and amounts differ. Combining products can duplicate nutrients; some ingredients interact with medicines. Those practical details remain relevant even when a cognitive trial sounds encouraging.

A food journal can help you describe recurring meals and gaps in variety, but it cannot diagnose a deficiency or explain a memory symptom. If memory changes are persistent, worsening or disrupting everyday life, a supplement headline is not a diagnosis. Seek assessment rather than treating the article as a substitute.

For example, someone looking for a faster workday memory boost is asking a different question from the older volunteers completing scheduled tests. Matching the question first helps avoid buying a promising result that was never tested for your goal. It also leaves room to discuss real symptoms, sleep, diet and medicines without assuming a bottle explains everything.

The answer to keep

COSMOS adds evidence for a small cognitive benefit in some older adults. It does not establish a universal memory upgrade, and it should sit beside the earlier neutral trial and the uncertain clinical-outcome evidence. A useful decision considers the actual person and product instead of borrowing certainty from the phrase “brain health.”

Evidence context: COSMOS received public research support, Mars Edge support and study-product support from Pfizer Consumer Healthcare, now Haleon; author research relationships are disclosed. PHS II reports public support and donated study agents. These disclosures accompany the findings, not an endorsement of a brand.

Published 11 October 2026 · Sources checked 11 October 2026. Suggest a correction.