A study of nearly half a million adults examines how daily coffee habits, inherited stroke risk, caffeine metabolism and even the addition of sugar relate to long-term stroke risk.

Study: Habitual Coffee Consumption, Genetic Susceptibility, and Incident Stroke: A Prospective Cohort Study in the UK Biobank. Image Credit: Pixel365 / Shutterstock
In a recent study published in the Journal of the American Heart Association, researchers found that moderate coffee consumption was associated with a lower risk of stroke across different genetic susceptibility profiles.
An international team of scientists and clinicians from Chang Gung University, Chang Gung Memorial Hospital, Taipei Medical University, Tzu Chi University, Asia University, and other medical institutions in Taiwan conducted the research, in collaboration with researchers from Zhengzhou University-affiliated Henan Provincial People's Hospital in China.
Stroke is a substantial global public health challenge, accounting for 11.9 million events and 7.3 million deaths per year, according to the Global Burden of Disease (GBD) Study 2021. Because of this burden, even minor changes in dietary habits could have major implications for populations. Coffee is among the most consumed beverages and has been extensively studied in relation to overall cardiovascular risk, but evidence on stroke remains inconclusive.
Genetic susceptibility may influence cardiovascular responses to coffee. Individuals with the slow-metabolizer genotype of the cytochrome P450 1A2 (CYP1A2) gene, which encodes the primary enzyme responsible for caffeine metabolism, have been reported to have substantially higher risks of hypertension and myocardial infarction with heavy coffee intake. The interplay between innate susceptibility to stroke and coffee intake remains largely unexplored.
About the study
The researchers evaluated associations between coffee intake and incident stroke. Participants were drawn from the United Kingdom Biobank (UKB), a prospective cohort of over 500,000 adults. UKB participants with missing coffee or tea intake data or a history of stroke before baseline assessment were excluded. Standardized, self-reported questionnaires were used to examine habitual coffee intake, with consumption averaged across all available assessments to better reflect long-term drinking patterns.
Participants were stratified into four groups based on daily coffee intake: ≤0.5 cups (quartile 1), 0.6 to 1.67 cups (quartile 2), 1.68 to 3 cups (quartile 3), and >3 cups (quartile 4). The study's primary endpoint was a first stroke, determined from self-reported physician diagnosis and linked hospital records and death registries. Participants were followed up from baseline until a stroke occurred, death, loss to follow-up, or December 2023, whichever came first.
Cox proportional hazards models were used to estimate differences in stroke risk after adjusting for demographic factors, socioeconomic indicators, lifestyle behaviors, physiological and cardiometabolic biomarkers, comorbidities, medication use, and tea consumption. Sensitivity analyses accounted for missing data and the competing risk of death and examined baseline coffee intake alone.
The team examined whether genetic susceptibility influenced the relationship between coffee consumption and stroke by grouping participants according to their stroke-specific polygenic risk scores (PRS) and CYP1A2 genotypes. The dose-response relationship was examined using restricted cubic splines. Subgroup analyses assessed associations by caffeine content and sugar addition.
Findings
The study included 490,889 UKB participants. Across consumption groups, the mean age was 55.7 to 57 years, and the median follow-up ranged from 11.5 to 12.8 years. Individuals with higher coffee consumption were more likely to be White and female, with smoking and obesity most common in the highest consumption group. During the follow-up, 22,579 participants experienced a stroke.
In the adjusted model, moderate coffee intake was associated with a lower risk of stroke. Participants consuming 0.6 to 1.67 cups and 1.68 to three cups daily had 14% and 8% lower stroke risk, respectively, than those drinking half a cup or less per day. No statistically significant association was observed among participants consuming more than three cups daily. Sensitivity analyses produced broadly consistent findings, but associations were weaker when only baseline coffee intake was examined.
Moderate coffee intake was generally associated with a lower risk of stroke across different CYP1A2 genotypes and groups of inherited stroke risk, with no statistically significant genetic interaction. This finding does not establish that the relationship is identical across all genetic groups.
A separate dose-response analysis revealed a J-shaped relationship, with the lowest estimated stroke risk at approximately 2.6 cups of coffee per day, corresponding to an estimated risk 21% lower than with no coffee consumption. Beyond this level, the association with lower stroke risk gradually weakened.
In subgroup analyses, the lower risk of stroke persisted among participants consuming 0.6 to 1.67 cups daily who reported adding sugar to coffee, compared with those in the lowest consumption category within the same subgroup. Little difference in stroke risk was observed across consumption categories among participants who did not report adding sugar to coffee. These findings may reflect residual confounding and do not establish a benefit from adding sugar.
Within the same subgroup, consuming more than three cups of decaffeinated coffee daily was associated with a 33% higher estimated risk of stroke than consuming half a cup or less. The association with lower stroke risk was primarily found among those consuming caffeinated coffee. Individuals with existing cardiovascular concerns may switch to decaffeinated coffee, potentially explaining the elevated risk through reverse causation.
The researchers suggested that coffee polyphenols, including chlorogenic acid, may improve vascular function and reduce oxidative stress and inflammation, while excessive caffeine intake may raise blood pressure. The study did not directly test these potential biological mechanisms.
Conclusions
The study identified a J-shaped association between habitual coffee consumption and incident stroke, with the lowest risk estimated at around two to three cups per day. The association was broadly consistent across participants' CYP1A2 genotypes and stroke polygenic risk scores.
“These observational findings should not be interpreted as evidence to prescribe coffee for stroke prevention.”
The study's limitations included its observational design, self-reported coffee intake, residual confounding, healthy volunteer selection bias, and a predominantly White study population. Variations in cup size, preparation methods, and caffeine concentration may have affected estimates of coffee consumption. Averaging coffee intake across later assessments may also have introduced bias arising from the timing of these measurements. The researchers did not distinguish between ischemic and hemorrhagic stroke.
The results provide more detailed context for lifestyle counseling for primary stroke prevention. They may also reassure moderate coffee drinkers that routine genetic screening for caffeine metabolism is not necessary. The findings do not establish causality or justify prescribing coffee for stroke prevention, recommending sugar addition, or discouraging decaffeinated coffee consumption.
Source:
Journal reference:
- Lin C, Hsu P, Lin Y, et al. (2026). Habitual Coffee Consumption, Genetic Susceptibility, and Incident Stroke: A Prospective Cohort Study in the UK Biobank. Journal of the American Heart Association, 15, e052262. DOI: 10.1161/JAHA.126.052262, https://www.ahajournals.org/doi/10.1161/JAHA.126.052262