Your hippocampus, the seahorse-shaped memory hub deep in the brain, shrinks faster than normal in mild cognitive impairment (MCI). For years the standard advice has been to watch and wait. A randomized, double-blind trial in older adults with MCI now suggests that a year of daily DHA, the omega-3 fat concentrated in brain tissue, can slow that shrinkage on MRI and move several cognitive scores in the right direction. The details matter, though, and so do the caveats.
The trial: who, what and how long
The study, published in the Journal of Alzheimer's Disease in 2017 by Zhang and colleagues, was run in Tianjin, China, and registered as ChiCTR-IOR-15006058.
- Participants: 240 adults aged 65 or older with mild cognitive impairment, randomized to treatment or placebo.
- Completers: 219 finished the trial, 110 on DHA and 109 on placebo.
- Intervention: a DHA supplement at 2 g/day versus corn oil as the placebo.
- Duration: 12 months, with cognitive testing and MRI at baseline, 6 months and 12 months.
- Design: double-blind, placebo-controlled human RCT.
The 2 g/day figure refers to the supplement dose. According to Ergo-Log's reading of the paper, it supplied roughly 900 mg (0.9 g) of actual DHA per day. That distinction matters when you compare products, because the label amount of "fish oil" or "algal oil" is not the same as the DHA content.
The DHA group did absorb the supplement. Serum DHA rose by 3.85% in the DHA arm versus 1.06% in the placebo arm.
The MRI results: slower hippocampal atrophy
The imaging data are the most striking part of the paper. Over 12 months, the DHA group differed significantly from placebo in:
- Left hippocampus (p = 0.016)
- Right hippocampus (p = 0.008)
- Total hippocampus (p = 0.023)
- Global cerebrum volume (p = 0.032)
The direction favored DHA in every case: less shrinkage, not growth. In MCI, some volume loss is expected over a year, so the finding is that DHA slowed the decline relative to placebo. The authors concluded that DHA "slowed the progression of hippocampal atrophy."
One caution: the paper reports some unusually large partial eta-squared effect sizes, such as 0.757 for the right hippocampus. These are worth treating carefully until independent groups replicate the work.
The cognitive results: IQ, Information and Digit Span
Cognition was assessed with Wechsler-based measures. Compared with placebo, the DHA group showed significant improvement in:
- Full-Scale IQ: p = 0.039
- Information subtest (general knowledge and semantic memory): p < 0.001
- Digit Span (working memory and attention): p < 0.001
Ergo-Log's analysis of the data put the DHA group's gains at more than 10% for Full-Scale IQ, about 33% for Information and about 48% for Digit Span. Those percentages come from a secondary interpretation of the paper, not from the abstract itself. They are large for a nutritional intervention, which is one reason replication matters.
Why DHA might protect the brain
DHA is a major structural fat in the brain, especially in neuronal membranes and synapses. Two mechanisms are usually proposed:
- Membrane structure and function. DHA built into neuronal membranes may improve fluidity and the behavior of receptors and signaling proteins. That could support synaptic communication and resilience in aging neurons.
- BDNF support. Brain-derived neurotrophic factor helps neurons survive, grow and form connections. DHA is thought to support BDNF signaling, which may help preserve tissue such as the hippocampus.
These are plausible explanations, but this trial was not designed to prove them. Observational research has linked higher DHA intake to lower Alzheimer's risk, and this RCT adds interventional data in a high-risk group.
How this fits with mixed omega-3 trials
If omega-3s are so good for the brain, why have large trials in healthy older adults been disappointing? Studies such as VITAL and the cognition arm of DO-HEALTH have produced mixed or null results. A few reasons could explain the gap:
- Population: healthy people have little cognitive decline to slow, so benefits are hard to detect. People with MCI are on a steeper trajectory.
- Baseline status: people who already eat fish regularly may gain less than those with low DHA levels.
- Formulation and dose: many trials used mixed EPA/DHA products at lower DHA doses.
- Duration and endpoints: cognitive tests in healthy adults can be insensitive over a short time.
The Tianjin trial does not overturn the mixed evidence. It suggests that DHA's benefit may be concentrated in people who already show early cognitive trouble.
Practical takeaways
- Dose: the trial used 2 g/day of supplement, which delivered about 0.9 g of DHA. Check the label for actual DHA per serving, not just total oil.
- DHA vs EPA: this trial tested DHA specifically. EPA has strong evidence for cardiovascular and mood effects, but the brain-structure rationale here centers on DHA. If memory is your goal, look for a DHA-rich product.
- Algal vs fish oil: both can provide DHA. Algal oil suits vegetarians and vegans and is a direct DHA source. Fish oil works too, but you may need more capsules to reach the same DHA dose.
- Timing: take it with a meal that contains fat, which generally improves absorption and reduces fishy aftertaste.
- Who it may be for: older adults with memory complaints or a diagnosis of MCI, ideally in discussion with a clinician.
- Safety: omega-3s can have mild blood-thinning effects at higher doses. If you take warfarin, antiplatelets or other anticoagulants, or have a bleeding disorder or upcoming surgery, talk to your doctor first.
Honest limitations
- One trial, one site. This was a single-center study in a Chinese cohort, and the results may not generalize to other diets, genetics or health systems.
- MCI only. The findings do not apply to healthy adults or to people with established dementia.
- Placebo choice. Corn oil is a common comparator, but any placebo oil can raise questions about neutrality.
- Surrogate-heavy picture. MRI volume and test scores are informative, but they are not the same as preventing dementia or preserving daily function.
- Replication needed. The authors themselves call for larger, longer confirmatory studies.
Bottom line
In 219 older adults with MCI who completed the trial, 12 months of DHA at about 0.9 g/day slowed hippocampal shrinkage and improved IQ-related scores compared with placebo. The result is promising and biologically plausible, but it comes from a single trial and needs replication. For older adults with memory concerns, a DHA supplement is a reasonable topic to raise with a doctor. It should not replace a proper cognitive evaluation or other brain-healthy habits.
