Caffeine ingestion: yet another wake-up call?

By Thomas B. Graboys and Susanna E. Bedell
Copyright 1998 American Heart Journal
October 1998



   The idea that “we are what we eat” is as old as the hills, and it is continually straining to find the rewards and hazards of the dinner table. The history of these attempts reduces science to witty entertainment. Food and drinks such as coffee, alcohol, ginseng, oat bran, and even margarine come and go as they are “proven” by varied studies to be useful, or poisonous for those who prefer a long life to a happy one.
   One of America’s most controversial foods is coffee, as essential as the alarm clock to get America going in the morning. According to a recent Consumer Reports on Health article, “it’s America’s favorite drug—should you just say no?”
   The public continues to be fascinated, indeed obsessed, with issues regarding caffeine. Comments in both the printed and visual media raise questions on potential ills induced by caffeine-containing beverages. Clinicians embrace the sympathetic limb of the autonomic nervous system as a trigger for arrhythmia and then acknowledge that caffeine stimulates the release of catecholamines from sympathetic nerve endings. The understandable inference is that caffeine causes atrial and ventricular arrhythmia or even life-threatening dysrhythmia after ingestion of high dosages. We reviewed this subject more than 15 years ago, noting that there has never been any evidence that caffeine induces consistent reproducible significant changes in the heart rhythm. Indeed, a study by DeBacker et al found that abstinence from caffeine, even when combined with a regular exercise program over a 6-week period, did not alter the prevalence of ventricular arrhythmias among subjects with known ventricular extrasystoles.
   More recently we evaluated the effects of caffeine ingestion on ventricular arrhythmia among 50 consecutive hospitalized patients admitted for symptomatic nonsustained or sustained ventricular tachycardia. Ninety percent of this patient group had repetitive ventricular arrhythmia and nearly 60% had salvos of ventricular tachycardia during telemetry observation. We found no differences between the caffeine and decaffeinated trials among this group of patients with manifest malignant ventricular arrhythmias despite achieving significant caffeine blood levels.
   In this issue of the Journal, Donnerstein et al investigated the acute effects of moderate caffeine ingestion on signal-averaged P wave and QRS complexes. They observed small changes in the signal-averaged QRS and no change in P-wave duration. They conclude that “further prolongation in the duration of the QRS complex may be a factor in the arrhythmias associated with caffeine toxicity.” Other studies have not supported the speculation that higher doses of caffeine may affect cardiac conduction. Chelsky et al observed no changes in inducibility of ventricular arrhythmias among patients with ventricular tachycardia undergoing electrophysiologic study before and after caffeine ingestion. This finding was underscored in an ambulatory electrocardiographic monitoring study by Newcombe et al. In the case report cited by Donnerstein et al detailing ventricular fibrillation after toxic amounts of caffeine ingestion, the authors fail to note that ventricular fibrillation occurred in association with a potassium level of 1.8 mEq/L in the case report.
   Both atrial and ventricular extrasystoles are ubiquitous in the population. This is true regardless of whether the patient has structural heart disease or not. Within the cardiologic community we have become much more sanguine about “benign arrhythmias,” concentrating exclusively on individuals with symptomatic sustained arrhythmia or patients with atrial fibrillation, regardless of symptomatology. Preliminary work from our center suggests that when surveyed, <20% of patients with so-called “lone” atrial fibrillation drew a relation between caffeine-containing beverages and the onset of dysrhythmia. Alcohol was far more likely to be a provocateur of atrial fibrillation in this population (47%) than caffeine. Extrapolating data from healthy subjects to the population at large is frequently fraught with methodologic problems. Furthermore, the subjects involved in this study were ages 21 to 26 and there is no mention of possible other factors that could affect cardiac conduction such as alcohol, electrolyte imbalance, drugs, or nicotine.
   There are a number of questions that one needs to assess before drawing conclusions from studies that may have no relevance to our or our patients’ day-to-day activities, including ingestion of caffeinated beverages. First is whether the data presented are methodologically rigorous. Second, are the findings in the population studied relevant to individuals with and without structural heart disease? Third is whether the data are merely a statistically significant observation. Fourth, how will this information reach the public, and what would its impact be on managing patients with cardiac dysrhythmia?
   Is there a “bottom line” take-home message from this study? We think nothing different from what has already been published. Minimal changes on signal-averaged electrocardiographic studies among a small number of patients receiving a moderate amount of caffeine should in no way alter our management of patients. There are individuals who are caffeine sensitive, and occasionally a patient will find causality between caffeinated beverages and some perturbation in the heart rhythm. Indeed, common sense once again dictates that if coffee bothers you, don’t drink it; otherwise why should we physicians, as we do all too often, modify life’s pleasures without a sound basis of scientific accuracy to do so?

Comments on this posting?

Click here to post a public comment on the Trash Talk Bulletin Board.

Click here to send a private comment to the Junkman.


Material presented on this home page constitutes opinion of Steven J. Milloy.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
 1