Coffee, tea, dementia risk and thinking skills over 43 years
What it found
People who drank more caffeinated coffee or tea had a lower chance of developing dementia and slightly better thinking skills.
Decaffeinated coffee showed no such link.
What they found
Studies in people
Caffeinated coffee and dementia
Among those drinking the most caffeinated coffee, there were 141 new dementia cases per 100 000 person-years. Among those drinking the least, there were 330 cases per 100 000 person-years.
Tea and thinking skills
Higher tea intake was linked to similar benefits for thinking skills and dementia risk as caffeinated coffee. Decaffeinated coffee was not linked to these outcomes.
Best amounts to drink
The strongest links with lower dementia risk and less self-reported memory trouble were seen at about 2 to 3 cups of caffeinated coffee a day. For tea, the strongest links were at 1 to 2 cups a day.
Objective thinking tests
In the women's group, those drinking the most caffeinated coffee scored slightly higher on a phone-based thinking test. The difference on a broader thinking score was very small and could have been due to chance.
What the authors conclude
- They conclude that drinking more caffeinated coffee and tea was linked to a lower chance of dementia and slightly better thinking skills.
- The authors note that the strongest links appeared at moderate daily amounts.
How it was done
Past research on coffee, tea and brain health has given mixed results. Most earlier work also did not separate regular coffee from decaffeinated coffee.
Researchers followed 86 606 women and 45 215 men in the US for up to 43 years. None had cancer, Parkinson disease or dementia at the start. They asked about food and drink every 2 to 4 years and tracked who developed dementia.
What it can’t tell you
- We summarised it from the abstract only.
- Because this study followed people over time rather than assigning drinks at random, it can only show a link, not prove cause and effect.
- People reported their own diets, so the exact amounts they drank may not be perfectly accurate.
Who paid
- Funding
- Funded by NIDDK NIH HHS; NINR NIH HHS (from the PubMed record).
- Conflicts
- Conflict of Interest Disclosures: Dr Hu reported receiving research support from the Analysis Group. No other disclosures were reported.
- Authors work at
- Harvard University, Massachusetts; Broad Institute, Massachusetts
The paper
- Title
- Coffee and Tea Intake, Dementia Risk, and Cognitive Function
- Type
- Study
- Evidence
- Studies in people
- Summarised from
- Abstract only
- Cite
- Zhang Y, Liu Y, Li Y, et al (2026). Coffee and Tea Intake, Dementia Risk, and Cognitive Function. JAMA. doi:10.1001/jama.2025.27259Free full textPubMed 41661604DOI
Summary written 30 Sep 2026. Check it against the paper before it changes what you eat. How we summarise papers · Report an error