Acute Effects of Coffee Consumption on Health Among Ambulatory Adults

Although coffee is one of the most commonly consumed substances in the world, the great majority of research regarding health effects have been observational and therefore prone to confounding by other factors. The Coffee and Real-time Atrial and Ventricular Ectopy (CRAVE) trial was a randomized, case–crossover assessment of acute health effects attributed to coffee consumption. One hundred volunteers were fit with wearable continuously recording ECG devices, a wrist-worn accelerometer to assess step counts and sleep, and a continuously recording glucose monitor. Their DNA was genotyped for common genetic variants that influence caffeine metabolism. Participants were then randomized via daily text messages to either consume all the caffeinated coffee they wanted versus to avoid all caffeine. Adherence to the randomization assignment was assessed by: 1) real-time participant activation button presses on the ECG monitor to timestamp every cup of coffee (a strategy previously validated by the same group for alcohol consumption compared with wearable alcohol sensors¹); 2) daily surveys regarding the previous day’s coffee consumption; 3) reimbursing all coffee consumption evidenced by a date-stamped receipt of coffee purchased for immediate consumption (whether they were supposed to have coffee that day based on randomization assignment or not); and 4) the Eureka digital research mobile application installed on participants’ smartphones, which was used to “geofence” visits to those coffee shops.

In both intention-to-treat analyses and “as-treated” (based on the amount of coffee actually consumed), no statistically significant differences in the primary outcome — premature atrial contraction (PAC) counts — were observed when individuals were exposed to caffeinated coffee versus no caffeine. In contrast, days randomized to coffee were associated with 51% more premature ventricular contractions (PVCs) (rate ratio [RR], 1.51; 95% CI, 1.18–1.94). In the as-treated analysis, those who consumed more than two cups of coffee experienced more than a doubling of their PVCs (RR, 2.24; 95% CI, 1.06–4.71). On days randomly assigned to coffee, participants took about 1000 more steps (mean difference, 1058 steps; 95% CI, 441–16750) but experienced more than a half-hour less sleep (36 fewer minutes of sleep per night; 95% CI, 25–47). Despite epidemiologic evidence that coffee consumers experience a lower risk of developing diabetes, no differences in serum glucose were observed in any of the analyses comparing exposure to coffee with caffeine avoidance. Two statistically significant interactions by caffeine metabolism–related genotype were observed: faster caffeine metabolizers exhibited an amplified relationship between coffee and an increased frequency of PVCs, whereas the slower caffeine metabolizers experienced the worst sleep when exposed to coffee (mean, −47 min; 95% CI, −69 to −25 min). In fact, the fast caffeine metabolizers experienced no detectable differences in sleep when exposed to coffee.

Taken all together, these data reveal complex acute effects of coffee related to several physiologic processes fundamental to overall health that themselves manifest heterogeneously across individuals. While those concerned about the frequency of PACs (a potent harbinger of atrial fibrillation risk²) should be reassured, those concerned about PVC frequency (and the associated risk of subsequent heart failure³) might consider reducing or eliminating coffee. Although the health benefits of coffee may arise from motivating more physical activity, those struggling with insomnia should incorporate these randomized data to motivate at least a trial of strict coffee avoidance to determine if it might benefit their sleep.

References

1. Marcus GM, Vittinghoff E, Whitman IR, et al. Acute Consumption of Alcohol and Discrete Atrial Fibrillation Events. Ann Intern Med. 2021;174(11):1503-1509. https://www.acpjournals.org/doi/10.7326/M21-0228
2. Dewland TA, Vittinghoff E, Mandyam MC, et al. Atrial Ectopy as a Predictor of Incident Atrial Fibrillation: A Cohort Study. Ann Intern Med. 2013;159(11):721-728. https://www.acpjournals.org/doi/10.7326/0003-4819-159-11-201312030-00004
3. Dukes JW, Dewland TA, Vittinghoff E, et al. Ventricular Ectopy as a Predictor of Heart Failure and Death. J Am Coll Cardiol. 2015;66(2):101-109. https://www.sciencedirect.com/science/article/pii/S0735109715022561