High-dose omega-3 reached the brain but didn't help memory in a two-year trial

Omega-3 fish-oil capsules are among the most popular supplements sold for "brain health". A team at the University of Southern California tested a high dose of one omega-3 fat for two years in older adults at higher risk of dementia. The fat did get into the brain, but memory test scores and brain scans showed no benefit over a dummy pill. It was a well-designed randomized controlled trial, but a mid-sized one that lost more than a third of its participants along the way.

DHA (docosahexaenoic acid) is an omega-3 fatty acid found mainly in oily fish. It is a building block of the membranes around nerve cells. People with low DHA in their blood tend to have more memory decline, but that is an association; earlier trials of omega-3 pills in people without dementia gave mixed results, and the team counts only 5 of 24 such trials in a recent review as showing a benefit for thinking.

One open question was whether those trials simply used too little, so that not enough reached the brain. This trial was built to answer that. It also focused on people with APOE ε4, the gene variant that is the strongest genetic risk factor for Alzheimer's disease and that changes how the body handles DHA.

What they did

• Who: 365 adults aged 55 to 80 (average 66) with normal memory and thinking, who ate little DHA (judged from a food questionnaire) and had at least one risk factor such as obesity, high blood pressure, high cholesterol or little exercise. Almost half (47%) carried APOE ε4, and 39% were Hispanic.
• What: half took 2 grams of DHA a day (five capsules), the other half a placebo of corn and soybean oil that looked and tasted the same, for 24 months. Neither the participants nor the researchers knew who got which. Everyone also took a B-vitamin supplement.
• Measuring the brain dose: 181 people agreed to a lumbar puncture, a needle in the lower back that takes a little of the cerebrospinal fluid that bathes the brain and spinal cord. The main goal of the trial was the change, after six months, in the ratio of DHA to another fat (arachidonic acid) in this fluid.
• Brain and memory: MRI scans at the start and after two years measured the hippocampus, a brain region key for memory that shrinks in Alzheimer's disease. A standard set of memory and thinking tests was repeated over the two years.

What they found

• DHA reached the brain. In the 167 people with spinal fluid results, the DHA ratio rose by 0.17 in the DHA group and fell slightly (by 0.02) on placebo. This large effect was the same in carriers and non-carriers of APOE ε4. In red blood cells, the omega-3 index rose by about 6 percentage points with DHA and barely moved with placebo.
• No effect on the hippocampus. It shrank slowly in both groups, by less than 1% a year, with no meaningful difference between DHA and placebo.
• No effect on memory and thinking. Overall test scores rose a little in both groups over two years (by 2.76 points with DHA and 2.67 with placebo), so DHA made no difference. Thinner brain areas linked to Alzheimer's, also measured, didn't differ either.
• Safety. Health problems of any kind during the trial were about equally common (about 40% in both groups). Serious ones affected 5 people on DHA and 13 on placebo; two people died, one in each group, neither because of the capsules.

What to keep in mind

• Many dropouts. 38% of participants left before the end, mainly because of the COVID-19 pandemic, and those who left differed from those who stayed (for example, more were Hispanic and had lower scores at the start). The authors say this may have biased the results and reduced the trial's power to detect changes in memory.
• A healthy, educated group. The participants were relatively young, well educated and showed little decline: memory scores even improved in both groups. A benefit might be hard to see in two years in people who are hardly declining. The trial ran at a single centre.
• Memory was not the main question. The trial was designed to show whether DHA reaches the brain. Brain size was a secondary outcome and memory tests were exploratory, so the "no benefit" result is informative but not the final word.
• One nutrient, one dose. The authors suggest that a single supplement may not be enough when inactivity, blood pressure and other risk factors are all at work, and that programmes combining several lifestyle changes may do better. That is their hypothesis, not something this trial tested.

Why it matters

Earlier omega-3 trials left a loophole: maybe the dose was too low to reach the brain. This trial closes much of it. DHA clearly got into the fluid around the brain, in people with and without the APOE ε4 risk gene, and still did not change memory or brain shrinkage over two years. It fits a broader pattern of null results for omega-3 pills in people without dementia, such as the large VITAL trial. The researchers conclude that the next step is to study how the brain uses and breaks down DHA, rather than to run more supplement trials. It is not a verdict on eating fish, which the trial didn't test, and it is not advice to start or stop any supplement.