OUP user menu

Of hammers and screws: renin–angiotensin–aldosterone system inhibition to prevent atrial fibrillation in patients with hypertension

Paulus Kirchhof, Larissa Fabritz
DOI: http://dx.doi.org/10.1093/eurheartj/ehu068 1169-1171 First published online: 23 February 2014

This editorial refers to ‘Antihypertensive treatment and risk of atrial fibrillation: a nationwide study’, by S.C.W. Marott et al., on page 1205

This editorial discusses the observation that monotherapy of uncomplicated hypertension with an ACE inhibitor or a sartan is associated with a lower risk of new-onset atrial fibrillation than treatment with a calcium channel blocker, diuretic, or beta-blocker. This association, found in a Danish nationwide database over a mean follow-up duration of 6–7 years, is put in context with the outcomes of controlled clinical trials testing similar interventions in patients without hypertension. In summary, the data suggest that different types of atrial fibrillation require different forms of preventive treatment, and that such ‘pathophysiological types of atrial fibrillation’ can be differentiated in patients using clinical characteristics supplemented by blood and ECG biomarkers, potentially opening the way towards a more personalized therapy of atrial fibrillation.

Although atrial fibrillation (AF) is rarely dangerous in the acute setting, the long-term consequences are grim. Even the most modern forms of management, including adequate stroke prevention and rate control therapy,1 are not sufficient to prevent premature cardiovascular deaths in patients with AF.2,3 Furthermore, the majority of AF patients remain symptomatic,4 and many are hospitalized for AF.5,6 After its manifestation, early rhythm control treatment of AF, i.e. thorough prevention of recurrent AF after the first episode, may be useful.2 Although we do not have formal proof of this concept, preventing the first episode of AF could be an even better way to prevent these complications.

How can we prevent the first occurrence of atrial fibrillation?

The mechanisms by which AF perpetuates itself have been studied in great detail in animal models and in patients in the last two decades, yielding important insights.7,8 The mechanisms that cause the first episode of AF are much less well understood. While some overlap between the vicious circles that maintain AF and the signalling processes that promote its first occurrence is likely, the mechanisms causing the first episode of AF are likely to differ from those that cause recurrent AF (Figure 1). Thus, we may need a different set of tools to fix the defects that cause the first episode of AF.

Figure 1

Possible mechanisms that cause atrial fibrillation, distinguished by the causes of the first episodes (left part of the figure) and causes of recurrent AF (right part). Dysregulation of ion channel function (shown in red), altered intracellular calcium handling (blue), and dysregulation of fibroblasts (green) are all likely to contribute to AF. We have learned a lot about the vicious circles that cause AF-induced atrial changes (right part of the figure). We still need to find out more (left part of the figure) about the atrial changes that cause the first episode of AF. AF, atrial fibrillation; IL-6, interleukin-6; RAAS, renin–angiotensin–aldosterone system; TGFβ, transforming growth factor β.

Histological alterations in patients with established AF include activation of fibroblasts, increased deposition of extracellular matrix proteins, and interstitial fibrosis which can uncouple cardiomyocytes and alter their electrical function. Another important change found in many, but not all, models of AF is cellular hypertrophy, suggesting altered intracellular calcium homeostasis and afterdepolarizations.7,8 One of the strongest signalling pathways causing both cardiomyocyte hypertrophy and formation of interstitial cardiac fibrosis is local and systemic activation of the renin–angiotensin–aldosterone system (RAAS). Indeed, its inhibition by angiotensin-converting enzyme inhibitors (ACE inhibitors) or angiotensin receptor blockers (sartans) can prevent recurrent AF when combined with amiodarone therapy,9 and reduce atrial fibrosis in dogs with pacing-induced heart failure.10 Furthermore, ACE inhibitors or sartans can prevent new-onset AF in patients with pre-existing heart failure or left ventricular hypertrophy.11,12

Moratt and colleagues now add important information to this evidence base: in their analysis of Danish prescription and hospitalization data, they found that monotherapy of uncomplicated arterial hypertension with ACE inhibitors, sartans, or a calcium channel blocker (excluding verapamil) decreased the risk of subsequent AF compared with monotherapy with either a beta-blocker, a calcium channel antagonist, or a diuretic.13 Importantly, the authors excluded patients with known heart failure, coronary artery disease, diabetes, chronic kidney disease, prior stroke, or hyperthyroidism. The patients share clinical characteristics with early-onset AF patient cohorts: Mean CHADS2VA2Sc score was 2.7 and the mean age was 63 years. The study cohort is one of the few cardiovascular studies comprising more women than men (63%).

Marott and colleagues compared the incidence of newly diagnosed AF over a mean follow-up time of 5.9–6.8 years: newly diagnosed AF was found in 0.5–1% of patients treated with ACE inhibitors, sartans, or calcium channel blockers, in ∼2% of patients treated with a diuretic, but in a staggering 5–6% of patients treated with a beta-adrenoreceptor blocker. The increased incidence of AF in the latter group only became apparent after several years of follow-up (see figure 2 in Marott et al.13). The strength of the study is its size, with 20 000–70 000 patients per group, and the long follow-up duration. This strength is balanced by lack of granularity: simple clinical parameters such as baseline blood pressure, blood pressure reduction on treatment, or information on valvular heart disease or ventricular hypertrophy unfortunately are not available. Even the main outcome of the study, newly diagnosed AF, is only indirectly detected through interventions and medications. More importantly, good physicians may have selected a beta-blocker in patients who are deemed at risk for AF or in patients with palpitations. A diuretic may have been prescribed in those with signs or risk of fluid overload. These prescription patterns, consistent with good clinical practice, generate confounders that cannot be adequately accounted for. Despite these limitations, the large size of the study population allowed several sensitivity analyses that are supportive of the main findings of the study.

How can we make sense of this conflicting information?

At first sight, the report of Marott and colleagues seems at odds with the outcome of two large, well-conducted, randomized trials that tested whether a sartan can prevent recurrent AF.14,15 It may even be perceived at odds with an older clinical trial suggesting that metoprolol prevents recurrent AF after cardioversion.16 Some have suggested that the antifibrillatory effects of sartans and ACE inhibitors require a few years of therapy, consistent with the time lag in the analysis by Marott et al. Possibly more importantly, AF is a syndrome that has different aetiologies at its onset.

  1. AF can be a consequence of concomitant cardiovascular conditions (heart failure, diabetes, ‘complex’ AF).

  2. Monogenic AF is found in patients with inherited arrhythmogenic diseases.

  3. A subtle genetic predisposition, e.g. located on chromosome 4q25, may contribute more to incident than to recurrent AF, i.e. ‘polygenic’ AF.

  4. Patients with severe mitral or aortic valve disease (valvular AF) will exhibit specific atrial arrhythmogenic mechanisms.

  5. Post-operative AF and AF in the elderly are other examples where intuition suggests different underlying mechanisms.17

  6. Based on the findings by Marott et al., AF in hypertensive patients may suffer from a biologically distinct type of AF.

Although the findings of Marott et al. may just be caused by different levels of blood pressure lowering or by treatment bias, it is tempting to speculate about the mechanisms: ACE inhibitors and sartans can interrupt the direct proarrhythmic effects of angiotensin, prevent cellular hypertrophy and intracellular calcium and sodium overload, or modulate the function of fibroblasts and their interaction with cardiomyocytes. Whatever the molecular mechanisms, it seems likely that inhibition of the RAAS is only beneficial for patients with a pathological activation of this signalling cascade, as will be the case in hypertension. In patients without hypertension, in contrast, in whom the RAAS may not be activated, pharmacological interruption of this signalling pathway does not seem to convey an antiarrhythmic effect.14,15 A hammer is the right tool to bang in a nail, but is useless for tightening screws. We may have to consider ‘atrial fibrillation caused by hypertension’ to differentiate nails and screws.


This work was supported by European Union (EUTRAF 251057) and the British Heart Foundation (FS/13/43/30324).

Conflict of interest: P.K. reports grants from the EU (EUTRAF) and the British Heart Foundation, during the conduct of the study; grants from several companies, and personal fees from several companies outside the submitted work. L.F. has no conflicts to declare.


  • The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.

  • doi:10.1093/eurheartj/eht507.


View Abstract