• Gregory M Marcus, Eric Vittinghoff, Isaac R Whitman, Sean Joyce, Vivian Yang, Gregory Nah, Edward P Gerstenfeld, Joshua D Moss, Randall J Lee, Byron K Lee, Zian H Tseng, Vasanth Vedantham, Jeffrey E Olgin, Melvin M Scheinman, Henry Hsia, Rachel Gladstone, Shannon Fan, Emily Lee, Christina Fang, Kelsey Ogomori, Robin Fatch, and Judith A Hahn.
• University of California, San Francisco, San Francisco, California (G.M.M., E.V., S.J., V.Y., G.N., E.P.G., J.D.M., R.J.L., B.K.L., Z.H.T., V.V., J.E.O., M.M.S., H.H., R.G., S.F., E.L., C.F., K.O., R.F., J.A.H.).
• Ann. Intern. Med. 2021 Nov 1; 174 (11): 150315091503-1509.

BackgroundPatients' self-reports suggest that acute alcohol consumption may trigger a discrete atrial fibrillation (AF) event.ObjectiveTo objectively ascertain whether alcohol consumption heightens risk for an AF episode.DesignA prospective, case-crossover analysis.SettingAmbulatory persons in their natural environments.ParticipantsConsenting patients with paroxysmal AF.MeasurementsParticipants were fitted with a continuous electrocardiogram (ECG) monitor and an ankle-worn transdermal ethanol sensor for 4 weeks. Real-time documentation of each alcoholic drink consumed was self-recorded using a button on the ECG recording device. Fingerstick blood tests for phosphatidylethanol (PEth) were used to corroborate ascertainments of drinking events.ResultsOf 100 participants (mean age, 64 years [SD, 15]; 79% male; 85% White), 56 had at least 1 episode of AF. Results of PEth testing correlated with the number of real-time recorded drinks and with events detected by the transdermal alcohol sensor. An AF episode was associated with 2-fold higher odds of 1 alcoholic drink (odds ratio [OR], 2.02 [95% CI, 1.38 to 3.17]) and greater than 3-fold higher odds of at least 2 drinks (OR, 3.58 [CI, 1.63 to 7.89]) in the preceding 4 hours. Episodes of AF were also associated with higher odds of peak blood alcohol concentration (OR, 1.38 [CI, 1.04 to 1.83] per 0.1% increase in blood alcohol concentration) and the total area under the curve of alcohol exposure (OR, 1.14 [CI, 1.06 to 1.22] per 4.7% increase in alcohol exposure) inferred from the transdermal ethanol sensor in the preceding 12 hours.LimitationConfounding by other time-varying exposures that may accompany alcohol consumption cannot be excluded, and the findings from the current study of patients with AF consuming alcohol may not apply to the general population.ConclusionIndividual AF episodes were associated with higher odds of recent alcohol consumption, providing objective evidence that a modifiable behavior may influence the probability that a discrete AF event will occur.Primary Funding SourceNational Institute on Alcohol Abuse and Alcoholism.

13 keywords...

hide…