To the editor

Bidirectional relationships have been suspected to exist between atrial fibrillation (AF) and heart failure, thromboembolism, renal dysfunction, and coronary artery disease (CAD).1 Also, CAD and AF share similar clinical risk factors including obesity, smoking, low physical activity, hypertension, diabetes, and sleep apnea.2 Previous experimental animal studies indicated that acute ischemia is associated with higher atrial vulnerability contributing to AF development.3 Nevertheless, we could not confirm this hypothesis in our study and found that AF prevalence was higher in patients with coronary artery sclerosis rather than CAD (Supplementary material, Figure S1).4 We also found that neither CAD origin nor its extent were associated with prevalent AF in our cohort.

With great interest, we read the article by Tomaszuk-Kazberuk et al5 confirming our findings. However, there are some issues that we feel should be highlighted and discussed here. First, it was intriguing and slightly confusing to us that the authors showed a lower risk of chronic coronary syndromes (CCS) in patients with hypertension, diabetes, hyperlipidemia, heart failure, and renal dysfunction despite higher prevalence of these comorbidities observed in CCS. Second, it is unclear how medical CCS treatment with statins, angiotensin-converting enzyme inhibitors, or angiotensin II receptor blockers modulated the findings because of their pleiotropic effects.5 In our study, we found that drug use was significantly higher in patients with clinically relevant CAD compared with those with normal vessels or noncritical CAD (59.1%, 22.8%, and 18.1%, respectively).4 Finally, in contrast to the study by Tomaszuk-Kazberuk et al,5 we analyzed differences in AF prevalence among 3 groups: individuals with normal (unobstructed) coronary vessels, coronary artery sclerosis, and clinically relevant CAD (defined as coronary artery stenosis ≥75%). Despite our hypothesis of a (monotonic) relationship between AF and angiographic coronary artery status, we found the highest AF prevalence in patients with coronary artery sclerosis, and—paradoxically—in CAD.4

Nevertheless, both analyses from the Leipzig Heart Study and the Bialystok Coronary Project reported the same key findings.4,5 We agree with the authors that the main results could be partly explained by the similarity of the clinical presentations of CAD and AF. Chest pain, dyspnea, anxiety—or even elevated levels of biomarkers—are common in both CAD and AF paroxysms. We suspect that these clinical signs and symptoms in patients with AF biased physicians’ decision, favoring the indication for coronary angiography, which eventually showed normal or only sclerotic coronary arteries. Importantly, both conditions—CAD and AF—share not only various clinical risk factors but are also supposed to have common underlying pathomechanisms such as inflammation. Nevertheless, despite higher interleukin-6 levels in patients with critical CAD in our study, we observed that adjustment for inflammatory markers did not change the results regarding the association between CAD and AF.

In summary, two large observational studies reported similar results contradicting previous hypotheses that AF prevalence is higher in patients with CAD. Further longitudinal studies with deeper phenotyping, eg, assessment of myocardial vitality, should further clarify the relationship between CAD and AF.