“It is easier to change a man’s religion than to change his coffee”
– Georges Courteline
The history of coffee runs parallel to the history of humankind. The plant, originating in Ethiopia, was first used as a beverage in Yemen after its beans were roasted. From the 15th century, its use spread throughout Muslim countries as a means of enhancing concentration during prayer and, from the 16th century, it became an urban custom in Europe.
By the 17th century, coffee houses could be found in the Netherlands, UK and Germany, where coffee became not only a traditional social habit, but also the most widely consumed physiological stimulant. The international term derives from an etymological root denoting, simultaneously, the colour black, (dark) wine, strength and lack of appetite.1
Over time, a range of qualities and, in particular, vices have been attributed to coffee, largely based on empirical considerations. One such misconception is the erroneous attribution of a chronic hypertensive effect. Its widespread use and the various positive or negative attributes assigned to coffee have, in recent years, prompted increased scientific scrutiny, including randomised controlled trials. The conclusions of these studies must be interpreted with caution, given the differences in quantity and preparation methods across geographical regions.
Coffee’s reputation as a stimulant is supported by a recent meta-analysis suggesting an ergogenic effect and improved exercise endurance when it is consumed approximately 1 hour before physical activity.2 Early studies documented anti-atherosclerotic and anti-diabetic effects of coffee, with an associated reduction in all-cause mortality, irrespective of the European population studied.3 Nevertheless, metabolic effects may vary according to preparation time and method; data from the Tromsø study demonstrate a greater increase in serum cholesterol with espresso compared with other brewing techniques.4;
An analysis of UK Biobank data, assessing long-term outcomes associated with different coffee preparation methods, concluded that both decaffeinated and caffeinated coffee consumption are associated with reductions in overall and cardiovascular mortality. However, only ground and instant coffee exhibited antiarrhythmic properties.5
Indeed, the principal concern regarding coffee consumption relates to its potential arrhythmogenicity, likely stemming from the fact that palpitations are the most frequently reported unpleasant effect. Despite this, a prospective study using Mendelian randomisation in a Biobank cohort revealed an inverse association between habitual coffee consumption and arrhythmias.6 Separate analyses of polymorphisms, including the slow caffeine metaboliser variant of CYP1A2, did not alter these conclusions.
A similar study evaluating nearly 35,000 individuals reached concordant results, emphasising a lower risk of AF independent of caffeine metabolism.7 Interestingly, coffee appears to exert superior cognitive effects in elderly patients with AF who are at the increased risk of stroke and dementia.8 This positive effect correlates with reduced inflammatory markers. Chronic coffee consumption may also confer beneficial cerebral effects in depression, Parkinson’s disease and Alzheimer’s disease.9
The beneficial effects of coffee consumption appear to be conditional on the timing of intake. A large study including 40,725 adults enrolled in the National Health and Nutrition Examination Survey demonstrated an inverse dose–response relationship (for 1–3 cups) between morning coffee consumption and all-cause, cardiovascular and cancer-related mortality, an association not observed with non-morning consumption.10 This circadian effect seems to correlate with higher morning concentrations of inflammatory markers.
In clinical practice, however, the issue remains open, as more than one in four patients consider coffee the second most common trigger for AF (after alcohol). Some recent observational studies and contemporary guidelines tend to endorse the notion of coffee abstinence to avoid AF onset or symptom exacerbation, despite a recent randomised controlled trial that refutes these hypotheses.11–13 Furthermore, N-of-1 studies have demonstrated that, with the exception of alcohol, purported AF triggers are likely matters of personal belief.14
It is against this background that the Does Eliminating Coffee Avoid Fibrillation? (DECAF) was initiated in the US, Canada and Australia, focusing specifically on AF, the arrhythmia associated with the highest hospitalisation rate.15 The study enrolled 200 patients (out of 1,965 screened) who were regular coffee drinkers who had AF (or atrial flutter with a history of AF) and were scheduled for electrical cardioversion. Patients were randomised 1:1 to either continued consumption of caffeinated coffee or abstinence from any product containing coffee or caffeine. Monitoring was performed using ECG, Holter recordings or implantable devices as appropriate.
The primary endpoint was the clinical detection of AF or atrial flutter lasting at least 30 seconds. Secondary endpoints included major adverse cardiovascular events, emergency department visits, hospitalisation or death. The study was powered to detect a reduction in AF risk. No significant inter-group differences were observed in baseline patient characteristics, and CHA2DS2-VASc scores were comparable and >2 in both groups. Coffee consumption in the active group was modest (approximately one cup per day) and remained constant throughout follow-up.
The study demonstrated a 39% reduction in the primary endpoint among coffee consumers compared with abstainers (p=0.01). This effect was comparable when AF alone was analysed. No other significant differences were observed between groups with respect to adverse outcomes.
The findings are encouraging for regular coffee consumers but the study is subject to several limitations. These include a relatively low willingness to participate (potentially leading to underrepresentation of individuals in whom coffee precipitates AF). Despite a balanced use of diagnostic methods overall, wearable ECG devices were more frequently used in the active group. Additional limitations include non-blinded exposure to coffee consumption, reliance on self-reported coffee intake and suboptimal adherence in the abstinence group (69%). Because only patients with persistent AF were enrolled, AF burden could not be assessed; paroxysmal AF may respond differently. Finally, the results apply to consumers averaging one cup per day and cannot be extrapolated to those with higher intakes exceeding three cups per day (>400 mg caffeine).
Nothing prevents those who prefer to avoid coffee from continuing to abstain; however, individuals accustomed to their morning coffee may rest assured that they can safely preserve this daily ritual.