DHA and Memory: What the Trials in Healthy Adults Actually Found

The two biggest trials found nothing

Three and a half thousand adults took omega-3 for five years. Their cognitive scores fell at exactly the same rate as the placebo group.

That was AREDS2, reported in JAMA in 2015. The yearly change in composite cognitive score was −0.19 with long-chain omega-3 and −0.18 without it. Mean age 72.7 years. The gap is a rounding error.

MAPT ran a similar test in France. It enrolled 1,680 adults over 70 who had told a doctor about memory complaints. They took 800 mg DHA plus 225 mg EPA daily for three years.

The difference against placebo was 0.011 on a composite Z score. The confidence interval ran from −0.081 to 0.103, p=0.812. That is not a small benefit. That is nothing, measured carefully in a lot of people.

Any honest piece on DHA and memory starts there. The largest and longest trials came back empty. Everything after this is about the smaller ones, and about why they might have found something the big ones could not.

MIDAS is where the memory claim comes from

Nearly every marketing line about DHA and memory traces back to one 2010 trial. It is worth knowing what it did and did not show.

MIDAS randomised 485 adults aged 55 and over to 900 mg of algal DHA a day or placebo for 24 weeks. Everyone had a memory complaint and a Logical Memory score at least one standard deviation below younger adults. So this was not a general healthy population.

The primary outcome was errors on a paired associate learning task. DHA produced 1.63 fewer errors than placebo, 95% CI −3.10 to −0.14, p=0.03. Verbal recognition memory also improved, immediate and delayed.

Working memory did not move. Executive function did not move.

One outcome moved, at p=0.03, in a selected group. That is a real result and a thin one. It is not a licence to expect a sharper memory.

The meta-analysis splits the room in two

The same lead author pooled the trial evidence in 2015, in PLoS ONE. The split it found is the most useful thing in this literature.

In adults with mild memory complaints, episodic memory improved with DHA, alone or with EPA. Hedge's g gave z=2.86, p=0.004. Above 1 g a day of combined DHA and EPA, episodic memory improved regardless of baseline status, p<0.04.

Now the part supplement copy tends to skip. In adults with no cognitive complaints at baseline, the pooled effect was not significant. Nothing.

Read that alongside AREDS2 and MAPT and the picture stops looking contradictory. Broad enrolment gives you nothing. Selected enrolment gives you a small something.

Low baseline intake changes the answer

The other selection that seems to matter is diet. Stonehouse tested 176 healthy adults aged 18 to 45 who habitually ate very little DHA. Six months, 1.16 g DHA a day.

Reaction times for episodic and working memory improved: −0.18 SD and −0.36 SD against placebo. Then the trial found a sex split. Episodic memory improved in women by 0.28 SD (95% CI 0.08 to 0.48, p=0.006). Reaction time of working memory improved in men by 0.60 SD.

Subgroup findings from one trial are a lead, not a conclusion. But the entry criterion is the interesting part. These people were short of DHA before they started.

If you already eat two portions of oily fish a week, you are not that person. Our guide to closing the DHA gap on a plant-based diet covers who genuinely runs low.

Dose did not decide it

It would be tidy if the null trials had simply used too little. They did not.

EPOCH gave 1,720 mg DHA plus 600 mg EPA daily for 18 months to 403 cognitively healthy Australians aged 65 to 90. That is nearly double the MIDAS dose, for three times as long. It did not maintain or improve cognitive performance on any domain tested.

It went slightly the other way on one. Psychomotor speed showed a small negative main effect, d=0.24, p=0.03. Participants also reported more perceived cognitive mistakes, d=0.24, p=0.003.

So the dose ladder does not sort the results. 800 mg failed in MAPT. 900 mg worked in MIDAS. 1,720 mg failed in EPOCH. Whatever separates these trials, it is upstream of milligrams.

Practically: if you supplement, there is no evidence that pushing past roughly 1 g of DHA a day buys you more cognition. Take it with a fat-containing meal, since absorption depends on it.

What DHA did shift

The Witte trial is often cited as a memory win. It was not one.

Sixty-five healthy adults aged 50 to 75 took 2.2 g a day of long-chain omega-3 for 26 weeks. Executive function improved against placebo, p=0.023. Grey matter volume and white matter integrity changed in frontal, temporal and parietal areas. Carotid intima-media thickness and diastolic blood pressure improved too.

Executive function is not memory. The structural imaging is genuinely interesting, and it is 65 people.

A 2025 dose-response meta-analysis of 58 randomised trials found modest gains per 2,000 mg a day: primary memory SMD 0.87 (95% CI 0.17 to 1.56), global cognition SMD 1.08 (0.73 to 1.44). The authors graded most of that evidence low to moderate certainty and called for longer trials. Wide confidence intervals and low GRADE ratings are the sound of an unsettled question.

What this means if you are 55 and training hard

Start with the regulatory floor, because it is the only claim anyone is allowed to make. The GB nutrition and health claims register permits one statement here: DHA contributes to the maintenance of normal brain function, at 250 mg of DHA a day. Maintenance, not improvement.

Then check your actual intake. Two portions of oily fish a week, the NHS suggestion, gets most people to that threshold and beyond. If you hit that, the trial evidence gives you little reason to expect a cognitive change from a capsule.

If you eat no oily fish — vegan, vegetarian, or simply do not like it — you are closer to the population where trials found something. That is where an algal DHA supplement has the better argument, and algal oil matches fish oil for DHA delivery.

Set the timescale honestly. Red blood cell fatty acids take about four months to settle. The trials that reported anything ran 24 weeks minimum. Nothing you notice in a fortnight is the DHA.

The honest summary

Pooled across more than 3,500 healthy older people, the 2012 Cochrane review found no benefit of omega-3 on cognitive decline over 24 to 40 months. AREDS2 and MAPT agreed, in another 5,000. That is the weight of the evidence, and it points at no effect.

Underneath it sits a narrower finding that keeps recurring: people with low baseline DHA or existing memory complaints sometimes improve on episodic memory tasks, at doses around 900 mg to 1.2 g a day.

Both things are true. Which one applies to you depends on what you are eating now, and that is a question a blood test answers better than an article. Our broader look at DHA and the brain covers the structural biology behind why anyone expected an effect in the first place.

Buy DHA for the reasons the evidence supports — membrane composition, triglycerides, the authorised maintenance claim. Do not buy it expecting a better memory at 60. On current trials, that is a coin toss dressed as a benefit.

References

1. Chew EY, Clemons TE, Agron E, Launer LJ, Grodstein F, Bernstein PS; AREDS2 Research Group (2015). "Effect of omega-3 fatty acids, lutein/zeaxanthin, or other nutrient supplementation on cognitive function: the AREDS2 randomized clinical trial." JAMA, 314(8), 791-801. DOI: 10.1001/jama.2015.9677.
2. Andrieu S, Guyonnet S, Coley N, et al. (2017). "Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): a randomised, placebo-controlled trial." Lancet Neurology, 16(5), 377-389. PMID: 28359749.
3. Yurko-Mauro K, McCarthy D, Rom D, Nelson EB, Ryan AS, Blackwell A, Salem N Jr, Stedman M (2010). "Beneficial effects of docosahexaenoic acid on cognition in age-related cognitive decline." Alzheimer's & Dementia, 6(6), 456-464. PMID: 20434961.
4. Yurko-Mauro K, Alexander DD, Van Elswyk ME (2015). "Docosahexaenoic acid and adult memory: a systematic review and meta-analysis." PLoS ONE, 10(3), e0120391. DOI: 10.1371/journal.pone.0120391.
5. Stonehouse W, Conlon C, Podd J, Hill S, Minihane AM, Haskell C, Kennedy D (2013). "DHA supplementation improved both memory and reaction time in healthy young adults: a randomized controlled trial." American Journal of Clinical Nutrition, 97(5), 1134-1143. DOI: 10.3945/ajcn.112.053371.
6. Danthiir V, Hosking DE, Nettelbeck T, Vincent AD, Wilson C, O'Callaghan N, Calvaresi E, Clifton P, Wittert G (2018). "An 18-mo randomized, double-blind, placebo-controlled trial of DHA-rich fish oil to prevent age-related cognitive decline in cognitively normal older adults." American Journal of Clinical Nutrition, 107(5), 754-762. DOI: 10.1093/ajcn/nqx077.
7. Witte AV, Kerti L, Hermannstadter HM, Fiebach JB, Schreiber SJ, Schuchardt JP, Hahn A, Floel A (2014). "Long-chain omega-3 fatty acids improve brain function and structure in older adults." Cerebral Cortex, 24(11), 3059-3068. PMID: 23796946.
8. Sydenham E, Dangour AD, Lim WS (2012). "Omega 3 fatty acid for the prevention of cognitive decline and dementia." Cochrane Database of Systematic Reviews, Issue 6, CD005379. PMID: 22696350.
9. Shahinfar H, Yazdian A, et al. (2025). "A systematic review and dose response meta analysis of Omega 3 supplementation on cognitive function." Scientific Reports, 15. DOI: 10.1038/s41598-025-16129-8.
10. NHS (2026). "Fish and shellfish." NHS.uk, Live Well - Eat Well - Food types.
11. Great Britain nutrition and health claims register (retained Regulation (EC) No 1924/2006; Commission Regulation (EU) No 432/2012). Authorised claim: DHA contributes to the maintenance of normal brain function, at 250 mg DHA per day.

DHA at the dose the trials used

The trials that reported memory gains used 900 to 1,160 mg of DHA a day, not 250. Check the DHA per capsule on any label — including ours — against the numbers in this article before you buy.

See Omega-3 DHA →