Research Watch: What Recent Studies Say About Supplements for Memory and Cognitive Aging

Welcome to the fourth Research Watch. Every Saturday I read the recent supplement research that crossed my desk and write up the studies worth knowing about. This week the topic is memory and cognitive aging, and the pattern that jumped out is almost the opposite of what the marketing would predict. The supplement with the most robust randomized evidence for memory is the cheapest, least glamorous one on the shelf, while a celebrated antioxidant and even a famous brain diet came up empty in head-to-head trials. Here is what the recent peer-reviewed evidence actually shows.

Quick Answer

A plain daily multivitamin has the strongest randomized evidence for protecting memory in older adults right now. Three large trials pooled in the COSMOS program found it improved global cognition and episodic memory, roughly equal to rolling back two to three years of cognitive aging. The more marketed options did worse: cocoa flavanols showed no benefit in the same trial program, and the MIND diet did not beat a control diet in a three-year randomized study. Creatine, usually sold for muscle, showed a surprising memory benefit in older adults. The honest summary is that the strength of the marketing and the strength of the evidence often point in opposite directions, and the unglamorous basics are winning.

What This Week's Reading Covered

Memory supplements are a high-margin, high-confidence category, and very little of that confidence comes from controlled trials. I stayed in recent randomized controlled trials and meta-analyses from 2023 to 2026, and I leaned on the largest and most rigorous studies available, including a Harvard-run trial program of more than 21,000 older adults and a New England Journal of Medicine diet trial. Six studies anchor this post, each linked to PubMed so you can read the abstract yourself. Where a study was funded by the company selling the ingredient, I say so, and I have included the trials that found nothing, because in this category the nulls are as informative as the wins.

A Daily Multivitamin Improved Memory in a 3,500-Person Trial

The most important single study this week is COSMOS-Web, a randomized, double-blind, placebo-controlled trial of 3,562 older adults published in the American Journal of Clinical Nutrition in 2023. Participants took either a standard daily multivitamin (Centrum Silver) or placebo and were tested with an internet-based battery of neuropsychological tasks each year for three years. The prespecified primary endpoint was change in episodic memory after one year.

The multivitamin group did significantly better on immediate recall at one year (p equals 0.025) and across the full three years on average (p equals 0.011). The authors estimated the benefit was equivalent to about 3.1 years of age-related memory change. There was no significant effect on the secondary measures of object recognition or executive function, so this is a memory signal specifically, not a global cognitive transformation. But it is a clean, large, placebo-controlled result for a supplement most people already have in a kitchen cabinet.

Source: Yeung LK et al. American Journal of Clinical Nutrition. 2023. PubMed 37244291

Pooling Three Trials Confirmed the Multivitamin Signal

One trial is a data point, not a verdict, so the same group ran a meta-analysis combining three separate cognitive substudies inside COSMOS: the in-person clinic cohort, the telephone-tested COSMOS-Mind, and COSMOS-Web. Published in the American Journal of Clinical Nutrition in 2024, it pooled non-overlapping participants across all three for a much more robust estimate.

Daily multivitamin-mineral supplementation produced a clear benefit on global cognition (mean difference 0.07 standard deviation units, 95 percent confidence interval 0.03 to 0.11, p equals 0.0009) and on episodic memory (mean difference 0.06 units, p equals 0.0007). The authors put the global cognition effect at roughly two years of reduced cognitive aging. When the same intervention beats placebo across three independently run trials using different testing methods, that is about as convincing as nutrition evidence gets. The effect is modest in absolute terms, and a multivitamin is not a treatment for dementia, but the direction and consistency are real.

Source: Vyas CM et al. American Journal of Clinical Nutrition. 2024. PubMed 38244989

Cocoa Flavanols: The Marketed Antioxidant That Did Nothing

Here is the contrast worth sitting with. The same COSMOS program tested cocoa extract, a concentrated source of flavanols heavily marketed for brain and vascular health, in the same kind of older adults. The clinic subcohort, 573 participants who completed detailed in-person neuropsychological testing, took 500 mg of cocoa flavanols per day or placebo for two years, reported in the American Journal of Clinical Nutrition in 2023.

The result was a clean null. Cocoa extract had no significant effect on global cognition, episodic memory, or executive function versus placebo. The only hint of benefit was in an uncorrected subgroup analysis among people who started with poorer diet quality, which is hypothesis-generating at best. So the antioxidant with the glossier story and the bigger price tag lost to the boring multivitamin tested in the very same program. That is the cleanest possible illustration of why mechanism and marketing are not evidence.

Source: Vyas CM et al. American Journal of Clinical Nutrition. 2023. PubMed 38070683

Creatine: A Muscle Supplement With a Surprising Memory Signal

A 2023 systematic review and meta-analysis in Nutrition Reviews pooled randomized trials of creatine, usually thought of as a strength and power supplement, on memory in healthy people. Across eight trials, creatine modestly improved memory overall (standardized mean difference 0.29, 95 percent confidence interval 0.04 to 0.53).

The interesting part was the age split. In older adults aged 66 to 76, the effect was large (standardized mean difference 0.88), while in younger people aged 11 to 31 it was essentially zero. The likely explanation is bioenergetic: the aging brain is under more energetic strain, and creatine supports the brain's energy buffering, so there is more room to help where reserves are lower. The caveats are real. Heterogeneity between trials was high, the older-adult subgroup rests on a small number of studies, and doses and durations varied widely. This is a promising signal in older adults that deserves larger dedicated trials, not a settled conclusion.

Source: Prokopidis K et al. Nutrition Reviews. 2023. PubMed 35984306

Marine Omega-3 Peptides Helped the People Declining Fastest

A 2026 randomized, double-blind, placebo-controlled trial in the American Journal of Clinical Nutrition targeted a specific group: 53 healthy older adults aged 60 to 73 who were already declining faster than their peers, identified by a memory test rather than by complaints. For three months they took a fish hydrolysate (1 g of low-molecular-weight marine peptides plus 30 mg of long-chain omega-3 fatty acids) or placebo.

The supplement group significantly improved on a paired-associates episodic memory test (p equals 0.003), completed more difficult patterns, and showed partial gains in spatial working memory. Their blood omega-3 levels rose and a marker of inflammation, C-reactive protein, fell. Two honest caveats: the trial was small, and it was run by Abyss Ingredients, the company that makes the fish hydrolysate, so independent replication matters. It is also a reminder that who is studied changes what you find. Targeting people already declining may be exactly why a three-month nutrition study could move the needle.

Source: Chataigner M et al. American Journal of Clinical Nutrition. 2026. PubMed 42162916

An Honest Null: The MIND Diet Did Not Beat a Control Diet

Diet is upstream of every supplement, so the strongest test of the whole "eat for your brain" idea belongs here. A 2023 randomized controlled trial in the New England Journal of Medicine assigned 604 cognitively healthy older adults with a family history of dementia and a suboptimal diet to either the MIND diet (a Mediterranean-DASH hybrid built around brain-protective foods) or a control diet, both with mild calorie restriction, for three years.

Both groups improved their cognition over three years, and the difference between them was not statistically significant (mean difference 0.035 standardized units, 95 percent confidence interval minus 0.022 to 0.092, p equals 0.23). Brain MRI measures were similar too. This does not mean diet is irrelevant; observational studies linking the MIND pattern to slower decline are consistent, and the control group also ate reasonably well and lost weight, which may have narrowed the gap. But it is a sharp lesson in why randomized trials matter. A dietary pattern with years of encouraging association data did not, when actually tested against an active comparison, produce a measurable cognitive edge.

Source: Barnes LL et al. New England Journal of Medicine. 2023. PubMed 37466280

What This Means for You

• If you want the option with the best evidence, it is unglamorous: a plain daily multivitamin-mineral has the most consistent randomized support for memory in older adults, replicated across three trials. It is cheap, low-risk, and roughly equivalent to a couple of years of memory aging. Manage expectations: this is a modest edge, not a cure.
• Be skeptical of the heavily marketed antioxidant: cocoa flavanols, tested in the same rigorous program, did nothing for cognition. A compelling mechanism and a premium price are not evidence. Ask what the placebo-controlled trial showed.
• Creatine is worth watching, especially with age: the cognitive signal was real in adults in their late sixties and seventies and absent in the young. It is inexpensive and well studied for safety, but the brain-specific trials are still small, so treat it as a reasonable experiment rather than a proven memory aid.
• Food first, then fill gaps: the multivitamin benefit likely reflects correcting subtle nutrient shortfalls that become more common with age and lower appetite. A varied diet remains the foundation; supplements close gaps, they do not replace it.

What the Evidence Does Not Support

A few claims run well ahead of the data. Antioxidant supplements sharpen memory. The flavanol arm of COSMOS is the direct counterexample: no cognitive benefit over two years. A "brain diet" guarantees slower decline. The MIND diet did not beat a control diet in a three-year randomized trial, even though observational data look favorable. A supplement that helps one group helps everyone. The marine peptide trial worked in people already declining and was industry funded, which tells you little about a healthy fifty-year-old. These are promising threads, not settled facts.

What I Did Not Cover This Week

I focused on randomized trials and meta-analyses in generally healthy older adults. I did not cover B vitamins and homocysteine in depth, where observational data and the older VITACOG trial in mild cognitive impairment are interesting but the broad-population trials have been mixed; vitamin D and cognition; lion's mane and other mushroom nootropics, where human data are still thin; or citicoline and phosphatidylserine. I also left magnesium L-threonate out of this edition because its cognition-and-sleep trial appeared in the sleep Research Watch, and the no-repeat rule keeps each study to one post. We will return to these as the literature builds.

This article is for informational purposes and is not medical advice. Memory concerns that are new, worsening, or interfering with daily life deserve a real conversation with a qualified professional rather than a supplement. If you take prescription medication, have a health condition, or are pregnant or breastfeeding, talk to a healthcare practitioner before adding any supplement. Creatine in particular can affect people with kidney conditions, and high-dose nutrients can interact with medications.

Sources

1. Yeung LK et al. Multivitamin Supplementation Improves Memory in Older Adults: A Randomized Clinical Trial (COSMOS-Web). American Journal of Clinical Nutrition. 2023.
2. Vyas CM et al. Effect of multivitamin-mineral supplementation versus placebo on cognitive function: COSMOS clinic subcohort and meta-analysis of 3 COSMOS cognitive studies. American Journal of Clinical Nutrition. 2024.
3. Vyas CM et al. Effect of cocoa extract supplementation on cognitive function: results from the clinic subcohort of the COSMOS trial. American Journal of Clinical Nutrition. 2023.
4. Prokopidis K et al. Effects of creatine supplementation on memory in healthy individuals: a systematic review and meta-analysis of randomized controlled trials. Nutrition Reviews. 2023.
5. Chataigner M et al. Dietary fish hydrolysate improves episodic memory in healthy elderly with lower level of memory performance: a double-blind, randomized, placebo-controlled clinical study. American Journal of Clinical Nutrition. 2026.
6. Barnes LL et al. Trial of the MIND Diet for Prevention of Cognitive Decline in Older Persons. New England Journal of Medicine. 2023.

Related reading

• Research Watch: Supplements for Sleep
• Research Watch: Supplements for Stress and Cortisol

Shop Magnesium L-Threonate