Coffee Drinkers Had 18% Lower Dementia Risk in a 43-Year Study, but Decaf Showed No Benefit
A headline circulating this week says two or three cups of coffee a day are linked to “dramatically” lower dementia risk. The study behind it is one of the longest and largest ever done on the question, and it deserves attention. But the number that should stick with you is 18%, not the much bigger gap the raw figures suggest. For anyone who treats caffeine as part of a focus routine, the details are more useful than the headline.
What the Researchers Did
A Harvard-affiliated team led by Yu Zhang analyzed two long-running U.S. cohorts, the Nurses’ Health Study and the Health Professionals Follow-up Study. Together they included 131,821 people (86,606 women and 45,215 men) who had no dementia, cancer or Parkinson’s disease when they entered the analysis. Participants filled in detailed food questionnaires every two to four years, so the researchers could track what people actually drank across decades instead of relying on a single snapshot. Median follow-up was 36.8 years, and the longest ran to 43. Over that time, 11,033 people developed dementia.
The paper appeared in JAMA in February 2026 and was picked up again by ScienceDaily on November 7, so this isn’t brand-new data.
Why 18% Is the Number to Remember
The eye-catching figures compare 141 dementia cases per 100,000 person-years among the heaviest caffeinated coffee drinkers with 330 among people who drank little or none. Taken alone, that looks like a drop of nearly 60%. Those rates don’t account for the many other ways the two groups differed, though. Once the researchers adjusted for those differences, the highest-intake group had an 18% lower risk than the lowest (hazard ratio 0.82, 95% confidence interval 0.76–0.89).
That is still a meaningful result for something so ordinary, but it’s a different story from “cut your risk in half.” Study author Daniel Wang of Mass General Brigham and Harvard Medical School said as much in coverage of the paper: “While our results are encouraging, it’s important to remember that the effect size is small.”
The Details That Matter
More wasn’t better. The relationship was curved rather than a straight line. Risk was lowest at about two to three cups of caffeinated coffee a day, or one to two cups of tea, and drinking more didn’t add anything further. It also didn’t appear to do harm in this analysis.
Tea held up well. The highest tea drinkers had a 14% lower risk (HR 0.86, 0.83–0.90), and tea was the more consistent performer on cognitive testing. The questionnaires didn’t record what kind of tea people drank, however, so the study can’t say whether caffeine explains it.
Decaf was the outlier. Decaffeinated coffee had no meaningful link to dementia (HR 0.97, 0.93–1.01). This is the most intriguing clue in the paper, since decaf keeps most of coffee’s plant compounds and removes the caffeine. It doesn’t settle the matter. Decaf drinkers actually reported slightly more subjective memory problems (prevalence ratio 1.16). One plausible reading, and it’s ours rather than the authors’, is that some people switch to decaf because of health issues, which would muddy any comparison.
Genetics didn’t change the pattern. Results looked similar in people at high and low genetic risk of dementia, according to the authors.
Put Your Coffee Where It Does the Most Work
If you drink coffee anyway, time it for your hardest work and keep it early in the day. Our free Focus Timer can help you plan those sessions.
Open the Focus Timer →What This Study Can’t Tell You
This is observational research, so it can only show an association. Regular coffee drinkers may differ from non-drinkers in ways no statistical adjustment fully captures. Reverse causation is also possible, since people who are starting to feel unwell may cut back on coffee. Repeated diet measurements over decades reduce that problem but don’t eliminate it.
The effect on actual thinking ability was also small. Objective tests were given only to women over 70 in the nurses’ cohort. The top coffee group scored just 0.11 points higher on a 41-point telephone cognitive test, and the difference on a broader cognition score (0.02) was not statistically significant. Dementia was identified from death records and self-reported physician diagnoses, which will miss some cases. The participants were mostly health professionals, so the results may not apply equally to everyone. And the questionnaires didn’t capture how coffee was brewed, which affects what ends up in the cup.
A Practical Takeaway
If you already enjoy two or three cups a day, this is reassuring. If you don’t drink coffee, it isn’t a reason to start. Aladdin Shadyab of UC San Diego, who was not an author, said in Pharmacy Times’ coverage that the findings don’t necessarily mean people should be encouraged to take up the habit.
What does change for a focus-minded reader is the framing. Coffee is more defensible as a long-term habit than some people fear, but the short-term cost still counts. Our recent look at how caffeine thins out deep sleep found effects that people often don’t notice, even after a full night in bed. So keep the dose moderate and finish your last cup by early afternoon. If caffeine makes you jittery, our guide to pairing it with L-theanine covers one well-studied option. Coffee is also a small factor next to the bigger habits we covered in our posts on the MIND diet and brain aging and structured lifestyle programs.
- Very long follow-up. 131,821 U.S. adults were tracked for up to 43 years, and 11,033 developed dementia.
- The real figure is 18%. Raw rates suggest a gap near 60%, but the adjusted estimate for the highest coffee intake is an 18% lower risk.
- Moderate intake looked best. Risk was lowest around 2–3 cups of caffeinated coffee or 1–2 cups of tea a day, with no added benefit beyond that.
- Decaf showed no link. That hints at a role for caffeine but doesn’t prove it.
- Association, not proof. Differences between drinkers and non-drinkers may explain part of the result, and effects on test scores were small.
Your morning coffee probably isn’t hurting your long-term brain health, and it may be helping a little. That’s a good reason not to feel guilty about it, but a poor reason to treat it as a dementia strategy.
Frequently Asked Questions
Does coffee prevent dementia?
This study can’t show that. It is observational, so it found that people who drank more caffeinated coffee developed dementia less often, not that coffee was the reason. After adjusting for other factors, the highest coffee group had about 18% lower risk than the lowest.
How much coffee was linked to the lowest dementia risk?
Risk was lowest at about 2 to 3 cups of caffeinated coffee a day, or 1 to 2 cups of tea. Drinking more did not appear to lower risk any further.
Does decaf coffee have the same effect?
Not in this study. Decaffeinated coffee was not linked to lower dementia risk or better cognitive scores. That hints caffeine may matter, but people who choose decaf can differ from other coffee drinkers in ways the study can’t fully measure.
Health disclaimer: This article summarizes published research and is for educational purposes only. It is not medical advice. If you are worried about memory changes, talk with a qualified healthcare professional. If you are pregnant, have a heart condition or anxiety, or take medication that interacts with caffeine, check with your doctor about how much is right for you.
FocusWaveHub Editorial Team
Editorial & Research Team
The FocusWaveHub editorial team synthesizes peer-reviewed research on cognitive performance, attention science, and evidence-based productivity. About our editorial approach →