For more than four decades, researchers tracked what over 130,000 American nurses and health professionals ate, drank, and how their minds held up as they aged. The payoff from that patience arrived in a study published in JAMA in June 2026: participants who consistently drank two to three cups of coffee a day had a 35 percent lower risk of developing dementia compared with those who skipped coffee entirely. The protective link was strongest among adults younger than 75, suggesting that what you drink in midlife may matter more for your brain than what you drink in old age.
The findings land at a moment when dementia projections are worsening worldwide and researchers are hunting for modifiable risk factors that don’t require a prescription. Coffee, consumed daily by roughly 62 percent of American adults according to the National Coffee Association, would be among the most accessible.
What the study actually measured
The research drew on two of the longest-running dietary cohorts in American medicine: the Nurses’ Health Study, launched in 1976, and the Health Professionals Follow-up Study, launched in 1986. Both are run by Harvard-affiliated investigators. The lead corresponding author on the JAMA paper is Dr. Changzheng Yuan, an epidemiologist at Zhejiang University who has collaborated extensively with the Harvard cohort teams.
All participants were dementia-free at enrollment and completed repeated dietary questionnaires over follow-up periods stretching up to 43 years. That repeated measurement is critical. Rather than asking someone once in 1990 how much coffee they drink and assuming the answer still applies in 2020, the study captured shifts in habits across multiple life stages.
Dementia outcomes were identified through a combination of medical records, cognitive testing, and validated screening instruments, not self-report alone. The researchers measured both subjective cognitive complaints, the kind of mental slippage people notice in daily life, and objective neuropsychological test performance. Because those two lenses don’t always agree, using both reduces the chance that results hinge on a single measurement quirk.
The central result: two to three cups of coffee per day were linked to a 35 percent reduction in dementia risk after adjustments for smoking, alcohol use, diet quality, physical activity, and other lifestyle factors. The association was most pronounced among adults under 75. Tea consumption showed a parallel pattern of slower cognitive decline, though it was analyzed separately from coffee.
Independent data points in the same direction
The Harvard cohort findings don’t exist in isolation. A separate analysis from the UK Biobank, a population study of roughly 500,000 British adults, reported that moderate coffee and tea consumption is associated with slower cognitive decline in that cohort as well.
The UK Biobank draws from a different country, a different recruitment era, and a different study design. When two independent datasets using different methods and populations point in the same direction, the signal gains credibility. It doesn’t confirm that the size of the effect is identical, and neither dataset can rule out confounding. But the overlap makes it harder to dismiss the pattern as a quirk of one particular group of American health workers.
Why this doesn’t prove coffee prevents dementia
Observational studies, no matter how large or long, cannot prove that coffee itself is what’s protecting the brain. People who drink moderate amounts of coffee may differ from non-drinkers in ways that are hard to capture in a questionnaire: sleep architecture, social routines, occupational complexity, or genetic makeup. The researchers adjusted for known confounders, but residual confounding is always a concern in epidemiology.
The age-75 threshold deserves particular scrutiny. The stronger association in younger participants could mean that coffee’s bioactive compounds, including caffeine and polyphenols, exert their effects on vascular and neuronal health during midlife, when the brain is still building cognitive reserve. But it could also reflect survival bias. People who remain healthy enough to drink coffee regularly into their late 70s and 80s may already be a resilient group, which would dilute any apparent benefit in the oldest age bands.
Brewing method is another open question. Filtered drip, espresso, French press, and instant coffee deliver different concentrations of cafestol, kahweol, and chlorogenic acids, compounds with distinct biological profiles. The JAMA paper references preparation methods in its metadata, but publicly available summaries don’t break out results by brewing type. That gap matters for anyone trying to translate the findings into specific kitchen choices.
The study also doesn’t distinguish between caffeinated and decaffeinated coffee in its headline result, a detail that matters because caffeine and the non-caffeine compounds in coffee may contribute to brain health through different pathways. Some prior research, including smaller studies and Mendelian randomization analyses using genetic proxies for caffeine metabolism, has produced mixed results on whether caffeine itself is the active ingredient or merely a bystander.
No randomized trial has tested whether prescribing coffee to middle-aged adults would reduce dementia incidence decades later, and such a trial would be extraordinarily difficult to run.
What coffee drinkers and non-drinkers should take from this
The practical takeaway is narrow but grounded. Adults who already drink two to three cups of coffee a day have no reason to stop on cognitive grounds. The data, drawn from over 130,000 people across more than four decades and echoed by an independent British cohort, suggest they may be doing their brains a favor.
Adults who don’t drink coffee should not start solely to prevent dementia. The evidence is associational, individual responses to caffeine vary widely, and people with anxiety disorders, cardiac arrhythmias, or chronic sleep disturbances may find that caffeine’s downsides outweigh any speculative long-term brain benefit.
What the study does accomplish is shift coffee from the “guilty pleasure” column closer to the “plausibly protective habit” column, at least for brain health. The 35 percent figure should be understood as the upper range of the observed association in a specific dose window, not as a guaranteed personal shield against cognitive decline. But for a beverage that most of its drinkers were going to have anyway, the reassurance is worth something.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
