130,000 people, 43 years, one striking finding
This is one of the largest observational studies ever conducted on coffee and brain health. Published in the Journal of the American Medical Association (JAMA) in 2026 and led by researchers from Harvard T.H. Chan School of Public Health and Mass General Brigham, it followed more than 130,000 participants over decades.
The finding: people who drank 2 to 3 cups of caffeinated coffee daily had an 18% lower risk of dementia compared to those who consumed little or none. Caffeinated tea (1-2 cups) showed similar benefits. Decaffeinated coffee showed no significant effect.
Why caffeine looks like the active ingredient
The fact that decaf produced no measurable effect is a meaningful signal. It points toward caffeine — not coffee's other compounds like polyphenols — as the likely factor behind the cognitive protection.
The proposed mechanisms include reduced brain inflammation and protection against cellular damage. But at this stage, those are hypotheses. This is an observational study: it shows association, not direct causation.
What it actually means — and what it doesn't
An 18% risk reduction is statistically significant, but modest at the individual level. Factors like physical activity, sleep, overall diet, and genetic predisposition have a much larger effect on dementia risk than coffee consumption does.
What the study doesn't say: that drinking more than 3 cups increases the benefit further. Or that coffee alone is a shield against cognitive decline. What it does suggest: for someone already drinking coffee in moderation, there's no reason to stop on brain health grounds.
Practical angle for active people
If you already use caffeine to boost performance — which most athletes do, consciously or not — this study adds another dimension to its potential benefit. Caffeine is one of the very few supplements with a solid evidence base for both acute performance and, now, long-term brain health.
Keep consumption reasonable (2-3 cups daily), avoid late-day intake that disrupts sleep, and the data is on your side.
Sources: JAMA — Zhang et al., 2026, Harvard Gazette