Two to three cups of coffee a day was linked to lower dementia risk over up to 43 years

Two to three cups of caffeinated coffee a day was associated with lower dementia risk in a Harvard-led study that followed 131,821 American adults for up to 43 years. Of those participants, 11,033 developed dementia. The paper appeared in JAMA in 2026, and decaffeinated coffee did not show the same pattern.

Help yourself and the planet: The Cool Down makes it easy to help yourself while helping the planet, with smart, simple changes to save more and waste less in your everyday life.

Get The Cool Down →

We may earn a commission if you sign up.

Daniel Wang, the senior author, cautioned that the results are encouraging but the effect size is small, and he described coffee or tea as one piece of a larger puzzle of habits that protect thinking skills. Lead author Yu Zhang reported that coffee or caffeine appeared equally beneficial for people with high and low genetic risk of developing dementia. ScienceDaily carried a write-up by Tori DeAngelis on Nov. 7, though the JAMA paper dates to Feb. 9.

Nurses’ Health Study and Health Professionals Follow-Up Study data

The Mass General Brigham release on EurekAlert says the work drew on two long-running cohorts, the Nurses’ Health Study and the Health Professionals Follow-Up Study, with Mass General Brigham, the Harvard T.H. Chan School of Public Health and the Broad Institute involved. Diet, dementia, subjective cognitive decline and objective cognitive function were reassessed repeatedly across the decades. The study was observational and prospective, and a long list of National Institutes of Health grants supported it, with the release stating that the funder had no role in design, analysis or the decision to publish.

The paper, Coffee and tea intake, dementia risk, and cognitive function, is in JAMA, volume 335, issue 11. The Harvard Gazette’s account reports that people with the highest caffeinated coffee intake had 18% lower dementia risk than people with little or none, in both men and women. Two to three cups a day showed the strongest association, and drinking more did not add benefit. The release publishes no hazard ratios or confidence intervals, so the 18% is as precise as the public summaries get.

The follow-up length is the study’s selling point. Earlier work on coffee and dementia has been inconsistent, in part because earlier studies had limited follow-up and less detailed diet data. Here, dietary questionnaires and cognitive measures were repeated over decades, which lets the analysis track a habit as it was actually practiced instead of relying on a single snapshot.

Rates per 100,000 and the subjective decline gap

A second set of numbers comes from the American Psychological Association’s Monitor, which presents the findings as a short numbers feature written by Tori DeAngelis. It lists about 141 dementia cases per 100,000 people per year among higher caffeinated coffee drinkers versus 330 per 100,000 among those who rarely or never drank it. Simple division would make that gap look much larger than 18%. The sources do not explain how the two figures relate, so the 18% reduction is the safer summary of the study’s main result.

Subjective cognitive decline, meaning a person’s own sense that their memory or thinking is slipping, was reported by 7.8% of higher caffeinated coffee drinkers and 9.5% of low or non-drinkers, according to the ScienceDaily post. On objective cognitive tests, coffee drinkers did better only on some measures, in the release’s words.

Decaf, tea and the caffeine question

Decaffeinated coffee showed no comparable benefit. The authors read that as pointing toward caffeine as the likely active ingredient, though they say the idea needs further validation, and the release also floats polyphenols as a possible contributor without confirming either one. Tea followed a similar pattern, with the strongest associations at one to two cups a day, but the study did not separate caffeinated from noncaffeinated tea, so it cannot say whether caffeine drives that result.

The standard caution applies in full. People who drink coffee differ from people who do not in ways the study may not capture, and earlier research on coffee and dementia has been inconsistent, partly because of shorter follow-up and thinner diet data. Wang’s own framing is the most honest summary: a modest association in a very large group, consistent across genetic risk, with no randomized evidence that adding a cup of coffee changes anyone’s chances.

For a reader deciding what to pour tomorrow morning, the practical reading is modest. The association peaked at two to three cups and flattened beyond that, so the data offer no reason to chase a fourth or fifth cup, and they offer no case for starting a caffeine habit to protect memory. Nothing in the sources addresses people who avoid caffeine for other medical reasons, and the cohorts were made up of nurses and health professionals, a group that may not mirror the wider public.

Two details keep the reported numbers in proportion. The 11,033 dementia cases are about 8% of the 131,821 participants, so most people in the study never developed the disease whatever they drank. And the reported benefit is relative, a shift in risk among those who did, which is a different thing from a guarantee for any one drinker.

The open question for the next study is mechanism. Without one, a mug of coffee remains a marker that travels with other habits, and the data cannot say whether it carries the benefit itself.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

Help yourself and the planet: Get The Cool Down →

We may earn a commission if you sign up.