The findings, published this week in the medical journal JAMA, don’t prove caffeine causes these beneficial effects, and it’s possible other attributes protected caffeine drinkers’ brain health. But independent experts said the study adjusted for many other factors, including health conditions, medication, diet, education, socioeconomic status, family history of dementia, body mass index, smoking and mental illness.
The caffeine correlation held regardless of whether people had genetic risk factors for Alzheimer’s or other dementias. The study, funded by the National Institutes of Health, didn’t distinguish between dementia types.
Some previous studies haven’t found cognitive benefits from caffeine, but those studies often had limitations like shorter time periods or one-time assessments of diet, experts said. The new study aligns with a growing body of research “that’s suggested caffeinated coffee may reduce risk of age-related chronic diseases”, Shadyab said.
Researchers followed participants in two long-running studies of medical practitioners: women in the Nurses’ Health Study and men in the Health Professionals Follow-up Study. Typically in their mid-40s or early 50s at the start, participants received repeated surveys about diet, health and lifestyle factors. During that time, 11,033 participants developed dementia, documented with death certificates or physician diagnoses.
Compared with people who consumed virtually no caffeine, people who drank between one and five 8-ounce (237ml) cups of caffeinated coffee had about 20% less dementia risk. Those who drank at least one cup of caffeinated tea daily had about 15% less risk.
But beyond 2½ cups of coffee daily, the advantage plateaued, possibly because humans cannot metabolise any more of the bioactive compounds in coffee and tea, said the study’s senior author, Dr Daniel Wang, an epidemiologist specialising in neurodegenerative diseases at Mass General Brigham health system.
Wang, who drinks about three cups each of coffee and green tea daily, said the study didn’t find anything negative about larger caffeine quantities. But some studies suggest exceeding moderate amounts can harm health by disrupting sleep or exacerbating anxiety, said Dr Fang Fang Zhang, an epidemiologist at Tufts University’s Friedman School of Nutrition Science and Policy, who wasn’t involved in the new research.
In a 2025 study linking caffeine consumption to lower mortality, Zhang said her team also found “you don’t get additional benefits when you go beyond three cups”. Adding more than a little milk or sugar nullified caffeine’s mortality benefits, she said. The dementia study didn’t track milk or sugar.
Beyond dementia, the new study collected assessments of subjective cognitive decline – people’s perception that their memory and thinking is slipping, often an early sign on the path toward dementia. Participants who drank more caffeine were less likely to report subjective cognitive decline.
About 17,000 participants, all women older than 70, also completed periodic cognitive tests. The more caffeinated participants had somewhat better scores for their age, suggesting their cognitive decline was slower by about seven months, Wang said.
Scientists say caffeine might protect brain health because it contains components that reduce neuroinflammation or aid vascular function. Research also suggests it improves insulin sensitivity, protecting against diabetes, a dementia risk factor.
For people who don’t consume caffeine, Shadyab said, the findings “don’t necessarily suggest that we should be encouraging people to drink coffee, but it is reassuring to those who currently drink coffee that it may reduce risk of dementia”.
Zhang said non-caffeine drinkers might “give it a try”, starting with small amounts in case they are sensitive to caffeine.
The correlation between caffeine and lower dementia risk was strongest for people younger than 75. Wang said cognitive impairment develops over decades, “so if you can change your health habits early, before midlife, that will be more beneficial”.
As medical practitioners, the participants might not reflect the general population. But Zhang noted that those who drank more caffeine were more likely to smoke and drink alcohol, suggesting they weren’t exceptionally healthier than typical Americans.
Researchers couldn’t exclude all possible influences on the results. For example, they wrote, were some people drinking decaf for medical reasons that fuelled dementia, “rather than a direct beverage effect?”
They also couldn’t say which might be better, brainwise: Darjeeling or matcha? Sumatra or Colombian? A $6 Americano or that cardboard-tasting free coffee from the office machine?
This article originally appeared in The New York Times.
Written by: Pam Belluck
©2025 THE NEW YORK TIMES
