Lead: A long-term analysis of U.S. health records involving 131,821 adults found that regularly drinking a modest amount of caffeinated tea or coffee was associated with a lower risk of dementia and slightly better cognitive performance. Researchers who followed participants in the Nurses’ Health Study and the Health Professionals Follow-up Study for up to 43 years report a roughly 15–20% reduced dementia risk among people who habitually consumed two to three cups of caffeinated coffee or one to two cups of caffeinated tea daily, compared with those who avoided these drinks. The study, published in the Journal of the American Medical Association, also found an 18% lower dementia risk for the highest caffeinated coffee consumers and small advantages on some objective cognitive tests. The authors caution the association does not prove that tea or coffee prevents dementia.
Key takeaways
- Sample and follow-up: 131,821 participants from two large U.S. cohort studies were tracked for diet, cognitive testing and dementia diagnoses for up to 43 years.
- Risk reduction: Habitual consumption of 2–3 cups of caffeinated coffee or 1–2 cups of caffeinated tea per day was associated with a 15–20% lower dementia risk versus little or no intake.
- Caffeinated coffee: Those drinking the most caffeinated coffee had an 18% lower dementia risk; the effect appeared to plateau at ~2–3 cups of coffee or 1–2 cups of tea daily.
- Decaffeinated coffee: No clear association between decaffeinated coffee and dementia risk was observed in the analysis.
- Cognitive tests: Regular caffeinated coffee drinkers showed marginally less reported cognitive decline and slightly better performance on several objective cognitive measures.
- Biological plausibility: Researchers highlight caffeine and polyphenols as candidate mechanisms via improved vascular, metabolic or anti-inflammatory effects.
- Limitations: Observational design means residual confounding and reverse causation (for example, people at higher dementia risk avoiding caffeine) cannot be excluded.
Background
Dementia is a growing global public-health challenge as populations age. Public-health authorities and researchers have increasingly focused on modifiable risk factors — such as smoking, high blood pressure, obesity, hearing loss and physical inactivity — that together are estimated to account for a substantial share of dementia cases. Diet and habitual beverage consumption are among the lifestyle factors under active study because they are widespread and potentially modifiable.
Prior research on coffee, tea and cognitive outcomes has produced mixed results: some studies report protective associations, others find null effects, and mechanisms remain speculative. The two cohorts used here — the Nurses’ Health Study and the Health Professionals Follow-up Study — collect repeated, prospectively recorded dietary data and medical outcomes, making them valuable for long-term exposure assessment, though participants are predominantly health professionals and may not represent the general population.
Main event
Investigators pooled data from 131,821 volunteers enrolled in the Nurses’ Health Study and the Health Professionals Follow-up Study and analyzed repeated assessments of beverage intake, cognitive testing scores, and clinician-recorded dementia outcomes across as many as 43 years of follow-up. Intake categories were defined by typical daily cups of caffeinated coffee and caffeinated tea, and analyses adjusted for many established dementia risk factors, including smoking, physical activity, body mass index and medical comorbidities.
The principal finding was a lower dementia risk among habitual consumers of modest amounts of caffeinated coffee or tea: approximately 15–20% lower compared with non-consumers, with the strongest single estimate an 18% reduction among the highest caffeinated coffee drinkers. Importantly, the apparent benefit leveled off at about two to three cups of caffeinated coffee or one to two cups of caffeinated tea per day.
The researchers also compared caffeinated to decaffeinated coffee and found no consistent inverse association for decaf, and they reported small but measurable advantages on some objective cognitive function tests for caffeinated coffee drinkers. Lead author Yu Zhang and colleagues emphasize that the study provides strong observational evidence but cannot establish causation, and they call for further biological and mechanistic studies.
Analysis & implications
Biologically, coffee and tea contain multiple compounds — notably caffeine and polyphenolic antioxidants — that might plausibly slow brain ageing by supporting vascular health, lowering inflammation and reducing oxidative stress. Caffeine is also associated with lower rates of type 2 diabetes, a known dementia risk factor, which could mediate part of the observed association. These pathways provide plausible mechanisms but do not prove that beverage consumption causes the reduced risk.
From a public-health viewpoint, a modest protective association, if causal, could have population-level impact because coffee and tea are widely consumed. However, the magnitude of benefit observed here is comparable to other lifestyle interventions and should not be interpreted as a substitute for established prevention measures such as blood pressure control, smoking cessation and hearing care.
Methodologically, the observational design leaves several alternative explanations intact: healthy-user bias (habitual caffeine consumers may have other protective behaviors), reverse causation (people with early sleep or cognitive problems might avoid caffeine), and residual confounding by socioeconomic or occupational factors. The cohort composition — largely nurses and health professionals in the U.S. — also limits generalisability to other populations and ethnic groups.
Randomized trials that assign individuals to caffeinated versus decaffeinated regimens for decades are impractical, so next steps could include experimental studies of short- to medium-term biological effects, imaging and biomarker work, and genetic (Mendelian randomization) approaches to probe causality.
Comparison & data
| Exposure | Approx. relative risk vs none | Notes |
|---|---|---|
| 2–3 cups caffeinated coffee/day | ~0.82 (18% lower) | Effect plateaued at this range |
| 1–2 cups caffeinated tea/day | ~0.80–0.85 (15–20% lower) | Similar magnitude to coffee |
| Decaffeinated coffee | ~1.00 (no clear link) | No consistent association found |
The table summarizes the study’s principal comparative findings. The risk estimates are adjusted associations from the published analysis and should be interpreted as observational signals rather than confirmed causal effects. The detected plateau suggests more is not necessarily better beyond modest daily intake.
Reactions & quotes
The lead investigator described the results as the strongest observational evidence to date linking regular caffeinated tea and coffee intake with better cognitive outcomes, while urging caution about causal claims.
Yu Zhang, Harvard University (lead author)
An independent cardiometabolic specialist noted that caffeine has mixed effects — potential metabolic and motivational benefits but possible blood-pressure elevation in some people — and said randomized evidence would be needed to estimate net effect definitively.
Naveed Sattar, University of Glasgow (cardiometabolic medicine professor)
Unconfirmed
- Whether the observed associations reflect a true protective effect of caffeine/tea components rather than residual confounding remains unproven.
- Whether the findings apply across different ethnicities, age groups and non–health-professional populations is uncertain because the cohorts studied are not fully representative.
- The specific biological mechanisms linking beverage intake to reduced dementia risk (vascular vs metabolic vs direct neuroprotective action) have not been established.
Bottom line
This large observational study strengthens evidence that modest daily intake of caffeinated coffee or tea correlates with a lower long-term risk of dementia and slightly better cognitive test performance, but it does not prove causation. The most consistent estimates point to roughly a 15–20% lower dementia risk for habitual consumers of about 2–3 cups of caffeinated coffee or 1–2 cups of caffeinated tea per day, with little or no signal for decaffeinated coffee.
Clinically and for public-health guidance, these findings suggest that moderate caffeinated coffee or tea can be part of a brain-healthy lifestyle but should not be viewed as a standalone protective measure. Priority remains on established risk-reduction strategies — blood-pressure control, smoking cessation, physical activity, hearing care and metabolic health — while further studies probe causality and mechanisms.