A daily multivitamin may improve memory slightly in adults over 60. In an analysis that pooled three COSMOS trials and 5,203 people, the pill gave small gains in memory and overall thinking. The authors liken the thinking gain to about 2 years less age-related decline (Meta-analysis).
The pills have not been shown to prevent dementia, though. In one trial, they did not clearly lower new cases over three years (Study). Two years of them also had no effect on blood markers of brain damage (Study).
What Is a Multivitamin?
A multivitamin is one pill that packs many vitamins and minerals together. Most hold doses close to the daily amounts adults are advised to get. In a 2011 to 2014 national survey, about 31% of US adults said they took one (Recommendation).
The biggest recent test of one is COSMOS, the Cocoa Supplement and Multivitamin Outcomes Study. It enrolled 21,442 US adults: women aged 65 and up and men aged 60 and up (Trial). A coin flip decided who got a daily multivitamin and who got a dummy pill. A second coin flip did the same for a cocoa extract, so both were tested at once.
Every COSMOS study used one product, Centrum Silver. Pfizer supplied the pills and the matching dummy pills (Trial). Pfizer’s consumer health arm, which helped supply them, is now Haleon (Study). A grant from Mars Edge, part of the company Mars, paid for the main trial, with help from the NIH (Trial). The National Institute on Aging paid for COSMOS-Mind, the first memory study (Study).
The COSMOS-Mind paper says the NIH, Mars and Pfizer played no part in the design or analysis. They also had no say in the writing or the decision to publish (Study). The pooled analysis says the same for every funder. One of its authors reports research support from Mars Edge, Pfizer and another supplement company, Pure Encapsulations (Meta-analysis).
The cocoa half of COSMOS is covered in our look at cocoa flavanols.
What the COSMOS Trials Found
Three studies inside COSMOS tested memory: one by phone, one online and one in person.
COSMOS-Mind came first. It followed 2,262 people, average age 73, with a phone test once a year for three years (Study). The multivitamin group did better on global cognition, a combined score for overall thinking. They also did better on memory and on executive function. These are the skills you use to plan, focus and switch between tasks.
The trial was mainly built to test cocoa, which had no effect. The multivitamin result answered a planned second question (Study). Only 77% of people took all three yearly tests.
The benefit looked bigger in people who had heart disease in the past. But only 197 people were in that group, and their result was too uncertain to lean on (Study).
COSMOS-Web tested 3,562 people online, once a year for three years (Study). The main test was a list of 20 words to recall right away. After one year, the multivitamin group went from about 7.1 to 7.8 words remembered. The placebo group went from about 7.2 to 7.7 words.
The placebo group improved through practice, since people get better at a test they have seen before. The authors found that the multivitamin gain went beyond that (Study).
Averaged over all three years, the gap between the groups held up (Study). In years 2 and 3 on their own, though, it was small enough that chance could explain it. The pill did not help people recall the words after a delay. It also did not help on tests of object recognition or focus (Study).
COSMOS-Clinic brought people in for detailed tests in person. Of 573 tested at the start, 492 were tested again about two years later (Study). So about 1 in 7 dropped out or could not be reached.
Memory improved on the multivitamin. But memory was a side question in this study, and the authors say such results need confirming. Overall thinking was the main outcome. It rose too, but by an amount small enough that chance could explain it. Executive function showed no benefit.
Pooling all three studies, with each person counted once, gave 5,203 people (Meta-analysis). Memory scores came from all three trials. Overall thinking came from two, since COSMOS-Web had not planned to report it. Both came out a little better on the multivitamin, and the three studies agreed with each other. The results also passed a strict check that allows for testing several outcomes at once. But the same research group did the pooling, with the same pill, so it is not an independent review (Meta-analysis).
How Much a Multivitamin Improves Memory
The COSMOS-Web authors call the effect small. They say it “may not be noticeable to all individuals” (Study).
After one year, the multivitamin group recalled about a quarter of a word more, out of 20. The other trials used scores that do not turn into words, but their gains were small too (Meta-analysis).
The “3.1 years” and “2 years” figures in the headlines come from a yardstick the researchers built. Nobody measured aging directly.
At the start of COSMOS-Web, older people recalled slightly fewer words than younger people. The team worked out how big that gap was for each year of age (Study). Then they divided the quarter-word gain by that yearly gap. The answer was about 3.1 years.
That yardstick compares different people of different ages at a single moment. It does not track how any one person’s memory changed as they aged. The figure also uses only the first year’s result.
COSMOS-Mind used the same method and got 1.8 years. Its authors wrote that the pills “appeared to have slowed aging by 1.8 years, or by 60%” over three years. They also called this yardstick imprecise (Study). They added that practice from taking the same tests again can distort this kind of estimate.
COSMOS-Clinic got 4.8 years for memory. That memory gain, though, was only just too large for chance to explain (Study). The pooled analysis got 2 years for overall thinking (Meta-analysis). Those authors list this “simplified approach” as a limitation. Each study used its own tests, so the figures do not line up.
A “2 years” figure does not mean your brain will act 2 years younger. It means the pill group’s average gain matched the gap in scores between people about 2 years apart in age. In daily life, you may not notice a change at all.
What a Multivitamin Did Not Do
Preventing dementia is the bigger hope. Here the evidence is much thinner.
COSMOS-Mind also tracked mild cognitive impairment (MCI). That is a decline in memory or thinking beyond normal aging, but short of dementia. Over three years there were 110 new MCI cases and 14 new dementia cases (Study). The rates on the multivitamin and the placebo were close enough that chance could explain the gap.
That is too few cases to settle the question. With only 110 MCI cases, the authors worked out, the trial had about a 1-in-5 chance of spotting a real 20% benefit (Study). It would have needed about 631 cases. Dementia can also take many years to develop, far longer than three. So a multivitamin has not been shown to prevent dementia. That is different from showing it cannot.
The authors did see one hopeful sign. People on the multivitamin who did develop MCI had higher scores, and they had slipped less since the year before (Study). The authors call this possible “cognitive resilience,” not prevention. They also say this check was not planned in advance, so it needs confirming.
Blood tests showed no benefit. A 2026 study measured neurofilament light, a protein that leaks out when nerve cells are damaged. It also measured a piece of tau and small clumps of amyloid, two proteins tied to Alzheimer’s disease. About 500 COSMOS-Clinic volunteers gave samples (Study). Over two years, all three markers changed about the same way on the multivitamin as on the placebo. So the small memory gain seen in this group (Study) came with no sign of less brain damage in the blood.
An older trial found no memory gain at all. The Physicians’ Health Study II gave Centrum Silver or a placebo to 5,947 male doctors aged 65 and up (Trial). They took up to four phone tests over about 12 years, with an average of 8.5 years to the last one. Men on the pill and men on the placebo scored about the same on memory and overall thinking at all four check-ins. Their scores also changed at the same pace. The authors said the trial was big enough to spot an effect.
The two trials also used different versions of the pill. The COSMOS version had more vitamin D and K, less vitamin A and iron, and a few other changes (Study).
Living longer is a separate question. A 2024 study followed 390,124 generally healthy US adults for up to 27 years (Study). Daily multivitamin users were no less likely to die than nonusers. In the first half of the follow-up, their risk of dying was about 4% higher. People chose for themselves whether to take a pill, so this kind of study shows a link, not cause.
COSMOS itself found no clear effect on cancer or heart disease either. Over about 3.6 years, there were 518 cancers on the multivitamin and 535 on the placebo. Major heart events came to 429 against 437 (Trial). Deaths were not clearly lower.
In 2022 the US Preventive Services Task Force, an expert panel that rates preventive care, looked at the question. It found too little evidence to judge the benefits and harms of multivitamins for preventing heart disease or cancer (Recommendation). Memory was not part of that review.
Who Might Benefit Most
The trial teams raise one idea: a multivitamin may top up nutrients that run low in later life.
Vitamin B12 is the best-known example. In large US and UK surveys, about 6% of people aged 60 and up were deficient. Close to 20% had borderline levels (Review). In older adults the main cause is usually trouble absorbing B12 from food, not a poor diet (Review). The COSMOS-Mind authors note that certain medicines can also cut B12 absorption (Study).
B vitamins have their own trial record. In VITACOG, 271 people aged 70 and up with MCI were followed for two years. They took either high doses of folic acid, B6 and B12, or a placebo (Trial). In the 168 who had two brain scans, brains shrank about 30% more slowly on the vitamins. Some people had high homocysteine, a blood marker that tends to rise when B vitamins run low. In them, shrinkage was 53% slower. People with the lowest homocysteine saw no effect. The authors say these group-by-group results need confirming.
Memory followed a similar pattern. Across everyone, only one thinking test held up better on the vitamins. In the half of people with higher homocysteine, memory and overall thinking declined less (Trial). VITACOG was small. It also tested only three B vitamins, at high doses, in people who already had memory problems.
Did COSMOS raise nutrient levels? In 399 volunteers, two years of the pill raised blood beta-carotene by 47% and vitamin E by about 30% (Study). That study reported nothing on B12 or folate. A broader look at 168 blood chemicals, such as fats, found only small shifts. Those shifts did not hold up under stricter checks (Study). So the nutrient-gap idea makes sense, but it is unproven.
The idea fits the doctors’ trial too. Its authors suggested their men may have been “too well-nourished to benefit,” or the doses too low (Trial). Testing also began about 2.5 years after the pills started, so early effects could have been missed.
The heart disease finding may fit the same idea. The COSMOS-Mind authors noted that people with heart disease tend to have lower nutrient levels (Study). But they did not measure nutrient levels, so the idea is untested. In COSMOS-Web, the bigger effect in people with heart disease showed up only in a check made after the fact (Study).
Standard doses look fairly low risk. The COSMOS authors reported no safety concerns (Trial). The multivitamin group did report a bit more gut bleeding. It also reported less stomach upset, rash and bruising.
One caution matters for smokers. The Task Force recommends against beta-carotene and vitamin E supplements (Recommendation). Its evidence review found that beta-carotene pills raised lung cancer risk (Review). The clearest evidence came from trials in smokers and asbestos workers. The harm showed up at 20 to 30 mg a day (Recommendation). If you smoke, check the label and skip high-dose beta-carotene.
Every multivitamin trial here tested Centrum Silver, and no trial has compared brands. The pooled analysis authors say they do not know whether their results apply to other multivitamins (Meta-analysis). They add that COSMOS had little racial and ethnic diversity. A label may say a formula was clinically studied. That only tells you a trial happened. It does not show that one bottle works better than a store brand with similar doses.
Frequently Asked Questions
Is Multivitamin Good for Memory?
It may be, a little, for adults over 60. In three COSMOS trials, a daily multivitamin gave small gains on memory tests compared with a placebo (Meta-analysis). The gain was small enough that not everyone may notice it (Study). It has not been shown to prevent dementia.
Does Multivitamin Help With Memory?
In older adults, the best trials so far suggest it helps slightly. In COSMOS-Web, people on a multivitamin recalled about a quarter of a word more from a 20-word list after one year (Study). But in 5,947 male doctors followed for about 12 years, it made no difference (Trial). The gap may come down to how well fed the doctors already were, or to the different pill formulas.
What Multivitamin Is Good for Memory Loss?
No trial has compared brands, so no product has been shown to be the best. The COSMOS trials all used Centrum Silver. The authors do not know whether the results apply to other products (Meta-analysis). If memory loss is real, see a doctor before reaching for a pill. A doctor can check for treatable problems, such as low B12, which is common in later life (Review).
Is Multivitamin Better Than Individual Vitamin?
No trial has compared the two head to head. COSMOS’s whole formula gave a small memory gain (Meta-analysis). In the VITAL vitamin D trial, vitamin D alone did not slow the decline in thinking scores. That result covered 4,218 adults aged 60 and up over two to three years (Trial). In VITACOG, high-dose B vitamins slowed brain shrinkage, mainly in people with high homocysteine (Trial).
Key Takeaways
- A small memory edge. Across three COSMOS trials and 5,203 people, a daily multivitamin gave small gains in memory and overall thinking (Meta-analysis).
- The “years” figure is a yardstick. The 2-year and 3.1-year numbers compare people of different ages at one moment. They do not track anyone’s own aging (Study).
- One brand, donated pills. Every COSMOS trial tested Centrum Silver, with pills and dummy pills from Pfizer. The papers say the funders had no role in the design or analysis (Trial).
- Not dementia prevention. The pills did not clearly lower new MCI or dementia cases, though the trial was too small to be sure (Study). They had no effect on blood markers of brain damage (Study).
- No gain in doctors, no longer life. A trial of about 12 years in 5,947 male doctors found no benefit (Trial). In 390,124 adults, daily users had no lower risk of death (Study).
- People low on nutrients may gain most. That idea is still unproven. About 6% of people aged 60 and up lack B12, and close to 20% have borderline levels (Review).
A Small, Cheap Edge for Memory
If you are over 60, a standard daily multivitamin may be a cheap, low-risk way to nudge your memory a little. No trial has compared brands, so a famous name does not mean a better pill. If you smoke, pick one without high-dose beta-carotene.
Keep the pill in proportion, though. Staying active, sleeping well, keeping blood pressure in check and treating hearing loss all still count.
If you or someone close to you has real memory loss, see a doctor first. Ask about a B12 test while you are there. Low levels are common after 60 and easy to check.
The memory edge is small and the pills are cheap. The dementia question is still open, and the next trials are worth watching.
This article is for educational purposes and is not medical advice. Talk to a qualified clinician before changing your health regimen.
Leave a Reply