Conference Coverage

CRAVE: Drinking coffee not linked to atrial arrhythmias



A novel trial using real-time monitoring found that drinking coffee did not increase atrial arrhythmias but was associated with more premature ventricular contractions.

Coffee cup with ECG and heart drawn in foam S_Bachstroem/Getty Images

There was no increase in premature atrial contractions (PACs) or supraventricular tachycardia (SVT) with coffee consumption, and, in fact, there was less SVT in per protocol analyses.

Coffee consumption was also linked to a “clinically meaningful increase in physical activity as well as a clinically meaningful reduction in sleep,” coprincipal investigator Gregory M. Marcus, MD, of the University of California, San Francisco, reported at the American Heart Association scientific sessions.

Although some professional society guidelines warn against caffeine consumption to avoid arrhythmias, he noted that the data have been mixed and that growing evidence suggests coffee consumption may actually lower the risk for arrhythmias, diabetes, and even mortality. The exact relationship has been hard to prove, however, as most coffee studies are observational and rely on self-report.

The Coffee and Real-time Atrial and Ventricular Ectopy (CRAVE) trial took advantage of digital health tools to examine the effect of caffeine consumption on cardiac ectopy burden in 100 healthy volunteers using an N-of-1 design. The primary outcomes were daily PAC and premature ventricular contraction (PVC) counts.

Participants consumed as much coffee as they wanted for 1 day and avoided all caffeine the next, alternating the assignment in 2-day blocks over 2 weeks. They used a smartphone app to receive daily coffee assignments and reminders and wore a continuous recording electrocardiography monitor (ZioPatch, iRhythm Technologies); a continuous glucose monitor (Dexcom); and Fitbit Flex 2, which recorded step counts and sleep duration.

At baseline, 21% of participants drank six to seven cups of coffee per month, 29% drank one cup per day, 21% drank two to three cups per day, and 3% drank four to five cups per day. The U.S. Food and Drug Administration has cited 400 mg per day, or about four or five cups of coffee, as generally safe for healthy adults.

To assess adherence, participants were asked to press the button on the ZioPatch for every coffee drink and were queried daily regarding actual coffee consumption the previous day. Date-stamped receipts for coffee purchases were reimbursed, and smartphone geolocation was used to track coffee shop visits. The great majority of times, participants followed their assignment by all measures, Dr. Marcus said.

ITT and per protocol analyses

ZioPatch data collected over a median of 13.3 days showed a daily median of 12.8 PACs, 7.5 PVCs, 1 nonsustained SVT, and 1 nonsustained ventricular tachycardia.

In intention-to-treat (ITT) analyses, there was no evidence of a relationship between coffee consumption and daily PAC counts (RR, 1.09; 95% confidence interval, 0.98-1.20; P = .10).

In contrast, participants had an average of 54% more PVCs on days randomized to coffee by ITT (RR, 1.54; 95% CI, 1.19-2.00; P = .001), and, per protocol, those consuming more than two cups of coffee per day had a doubling of PVCs (RR, 2.20; 95% CI, 1.24-3.92; P = .007).

No relationship was observed with coffee consumption and SVT episodes in ITT analyses (RR, 0.84; 95% CI, 0.69-1.03; P = .10), but, per protocol, every additional coffee drink consumed in real time was associated with a 12% lower risk for an SVT episode (RR, 0.88; 95% CI, 0.79-0.99; P = .028).

No significant relationships were observed with VT episodes, which were admittedly rare, Dr. Marcus said.

In ITT analyses that adjusted for day of the week, participants took an average of 1,058 more steps on days they drank coffee (95% CI, 441-1,675 steps; P = .001) but slept 36 fewer minutes (95% CI, 22-50 minutes; P < .001).

Per protocol, every additional coffee drink was associated with 587 more steps per day (95% CI, 355-820 steps; P < .001) and 18 fewer minutes of sleep (95% CI, 13-23 minutes; P < .001).

No significant differences in glucose levels were observed. Genetic analyses revealed two significant interactions: fast coffee metabolizers had a heightened risk for PVCs and slow metabolizers experienced more sleep deprivation, Dr. Marcus said.


Next Article: