The World Health Organization's 2026 dementia guidelines recommend adopting a healthy diet as part of reducing dementia risk.
They are noticeably more specific about what they do not recommend.
Dietary pattern, rather than any single nutrient.
The recommendation is about how you eat overall, and it overlaps almost entirely with advice for cardiovascular health. WHO also recommends managing cardiometabolic conditions including diabetes and high cholesterol, which is the same territory approached from another direction.
the gap between the evidence and the marketing is unusually wide here, and it is worth understanding why.
Observational studies often find that people with higher intake of a nutrient have better outcomes. When those nutrients are then given as supplements in controlled trials, the benefit generally does not appear. The most likely explanation is that eating fish is a marker for a whole pattern of living, and a capsule is not.
Supplements are also regulated far more loosely than medicines, which means the incentive to publicise a promising observational finding is large and the obligation to confirm it is small.
None of which means a supplement is harmful. It means that if you are taking one to protect your memory, the current evidence does not support that specific purpose.