Coffee and Tea Consumption Linked to Lower Atrial Fibrillation Risk

Coffee and tea consumption patterns affect AF risk

Drinking two to three cups of coffee daily may be linked to a lower risk of atrial fibrillation, according to a large-scale study published in Food & Nutrition Research, News-Medical reported.

The research, conducted by teams from Zhengzhou University, its First Affiliated Hospital, and Peking University in China, analyzed data from nearly half a million adults within the United Kingdom Biobank. Researchers tracked these participants over a median period of 13 years to see how their beverage habits influenced the development of the heart rhythm disorder.

Atrial fibrillation (AF) remains the most common type of arrhythmia, currently affecting approximately 33 million people worldwide. As the condition’s impact on mortality and healthcare systems continues to rise, scientists are looking closely at modifiable lifestyle factors that might help prevent it.

The 2–3 cup sweet spot

The study’s findings suggest that the relationship between caffeine and heart health isn’t a simple linear progression where more is always better. Instead, the data showed that a specific moderate intake level was most strongly associated with lower risk.

For coffee drinkers, the risk of developing AF was estimated to be between 7% and 11% lower than those who consumed no coffee at all. This protective trend was most pronounced in the group consuming 2–3 cups per day. Interestingly, the tea results were less consistent, though those drinking 2–3 cups of tea also saw a 5% lower risk of AF.

The researchers noted that the benefits did not steadily increase as consumption went up. For both beverages, the lowest risk estimates were found at the 2–3 cup mark, rather than at higher volumes.

Beverage Type	Daily Intake Level	Observed AF Risk Association
Coffee	2–3 cups	7% to 11% lower risk
Tea	2–3 cups	5% lower risk
Coffee	0.5–1 or ≥4 cups	Weakened association (after excluding existing CVD)

Observational study findings cannot prove causality

While the scale of the UK Biobank cohort—comprising 495,269 participants with an average age of 57—is massive, the researchers were quick to provide necessary context regarding what this data can and cannot prove.

Because this was an observational study, the findings cannot establish a definitive cause-and-effect relationship. The researchers stated that the study cannot prove that drinking coffee or tea actually prevents AF. Several factors limit the ability to draw a direct line of causality:

Self-reported data: Participants provided their own intake information via touchscreen questionnaires, which can be subject to memory errors.
Baseline assessment: Beverage consumption was only measured at the start of the study, not tracked continuously over the 13-year period.
Lack of standardization: The study did not account for specific cup sizes, brewing methods, beverage strength, or exact caffeine quantification.
Decaffeinated inclusion: Because the coffee intake data included decaffeinated varieties, the researchers noted the results cannot be attributed solely to caffeine itself.

The study also noted that the associations for coffee appeared to weaken among participants who drank either very little (0.5–1 cup) or a great deal (≥4 cups) once those with existing cardiovascular disease were excluded from the analysis.

Medical professionals often emphasize that dietary studies are just one piece of a much larger puzzle in cardiovascular health. For more information on managing heart rhythm, the Centers for Disease Control and Prevention provides extensive resources on arrhythmia and heart disease prevention.

The researchers concluded that while the link between moderate coffee and tea consumption and lower AF risk is promising, further research is required to understand the underlying biological mechanisms and to validate these findings across more diverse global populations.

The study’s demographic was largely White (94.4%) and female (54.6%), which may impact how broadly these specific findings can be applied to the global population.

Worth a look