Article Text

Download PDFPDF

Atrial fibrillation prevention: an appraisal of current evidence
  1. Giuseppe Boriani1,
  2. Marco Proietti2,3
  1. 1 Cardiology Division, Department of Diagnostics, Clinical and Public Health Medicine, University of Modena and Reggio Emilia, Policlinico di Modena, Modena, Italy
  2. 2 Department of Neuroscience, IRCCS–Istituto di Ricerche Farmacologiche Mario Negri, Milan, Italy
  3. 3 Institute of Cardiovascular Sciences, University of Birmingham, Birmingham, UK
  1. Correspondence to Professor Giuseppe Boriani, Cardiology Division, Department of Diagnostics, Clinical and Public Health Medicine, University of Modena and Reggio Emilia, Policlinico di Modena, 41124 Modena, Italy; giuseppe.boriani{at}


Atrial fibrillation (AF), which is the most diagnosed arrhythmia, is becoming a significant issue for health policy-makers. In particular, more and more attention is being devoted to AF prevention. Indeed, several studies recently published point out how targeted interventions could be useful in reducing the risk of AF occurrence (or recurrence). In this review, we briefly summarised the role of the major risk factors associated with the incidence of AF, as well as the effectiveness of interventions aimed at controlling these risk factors. Several general risk factors, such as alcohol consumption, physical activity, smoking habit, as well as specific cardiovascular risk factors as diabetes mellitus, hypertension and obesity have a relevant impact in determining the occurrence of AF, along with a strong clinical evidence of a dose–effect response mechanism for most of the factors examined. Specific interventions aimed at controlling risk factors have been showed to clearly reduce the risk of AF in several cohorts. Even more importantly, integrated programmes aimed at controlling for multiple risk factors would be more efficient in terms of reducing risk of AF, in particular whena stricter control is observed. AF prevention requires a series of initiatives focused on the many risk factors that we reviewed, as well as a more integrated approach, which should involve many stakeholders at different levels. In this light and also considering the constantly changing epidemiology, AF prevention may constitute a future ‘win–win’ strategy for all the stakeholders.

  • atrial fibrillation
  • stroke
  • epidemiology
  • cardiac risk factors and prevention
  • obesity

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Atrial fibrillation (AF) is the most common sustained arrhythmia and constitutes an emerging epidemiological threat, with increasing epidemiological burden worldwide.1 2 Traditional epidemiological studies performed in North America and Europe have found AF burden to be dependent on age, leading to an increasing prevalence in the last decades of  life,as the population progressively ages.3 More recently the increasing burden of AF has been confirmed in both developed and developing countries.4 Currently, around 33.5 million people worldwide are affected by AF (0.5% of the world population) with 5 million new cases diagnosed annually.1 2

The consequences of developing AF are important, given its strong association with major adverse outcomes.1 In this perspective AF is considered one of the thromboembolic conditions that leads to an increase in mortality risk, since one in four deaths occurred worldwide in 2010 has been attributed to a thrombotic/thromboembolic disease.5 Indeed, the strong relationship between AF and an increased risk of cardiovascular and all-cause death clearly emerged in the last years,6–10 as well as for stroke and thromboembolic events. Indeed, every year patients with AF are estimated to account for 15% of the 15 million strokes occurring worldwide.1

In table 1, we briefly summarised some estimates for morbidity and mortality risks associated with the presence of AF. In a large and comprehensive systematic review of cohort studies, it was pointed out how AF increases the risk for occurrence of a large spectrum of events.8 Indeed, it was found that the presence of AF carries a relative risk increase ranging from 31% for peripheral arterial disease to 96% for the occurrence of major cardiovascular events, and up to more than twofold and fivefold increase in risk for any stroke and heart failure (table 1). Moreover, a twofold increase in cardiovascular death and 46% increase in relative risk of all-cause death were found. Also, a 40% increase in relative risk for dementia and cognitive impairment has been reported.11

Table 1

Increased morbidity and mortality in patients affected by AF

Taking into account the epidemiological profile of AF and the ongoing progressive ageing of the general population, an impressive increase of patients at risk of AF or affected by AF is expected to occur in the next decades.1 2 12 Within this complex scenario, prevention of AF becomes an important priority for any healthcare system.

Traditional stages of prevention and AF care

Traditionally, in the original definition stated by the WHO, disease prevention is a strategy targeted at implementing interventions not only to prevent the occurrence of disease, such as reduction of risk factors but also to arrest its progression and reduce its consequences, once they are established.13 In this context, AF prevention could be approached considering the same stages of disease prevention that have been generally described.14

Primordial prevention is a relatively new concept, based on actions at a society and community level, targeted at inhibiting the establishment and diffusion of all factors (environmental, economic, social, behavioural, cultural) known to increase the risk of a disease.14 Limitation of alcohol use, as well as the promotion of sports and healthy food at school in order to reduce obesity in the community, are measures that can be classified as ‘primordial prevention’ and that, by promoting a healthy lifestyle, may have a positive impact on reducing the risk of AF at a population level.15

Primary prevention aims to reduce the exposure to risk factors at an individual level, thus minimizing the risk of developing a disease. In the clinical context of AF, the epidemiological remark about the large known amount of asymptomatic patients,16 17 coupled with the shared risk profile between incidence of AF and occurrence of AF-related major adverse events, creates some difficulties in applying a strict distinction between the different prevention stages. In particular, primary and secondary prevention would frequently overlap with each other. Notwithstanding, primary prevention of AF is targeted at reducing the first occurrence of a diagnosed AF and is currently emerging as a primary goal of policy-makers.15 18 19

Secondary prevention is traditionally defined as slowing the progression of a disease or its sequelae.14 In the case of AF, this strategy is used for the prevention of AF recurrences using rhythm control strategies.

Tertiary prevention applies once a disease has been clinically detected, assessed and treated. It has the aim of reducing the impact of disease on a patient’s quality of life, avoiding subsequent hospitalizations and improving long-term outcomes.14 In the case of AF, tertiary prevention is considered as the long-term care of patients with AF that currently also includes patients’ involvement, education and self-management.20–22

In the present review, we will focus on the main clinical factors associated with an increased risk of AF occurrence and, subsequently, on strategies aimed at reducing this risk. The decision to analyse only the most relevant, in terms of epidemiology and strength of evidence, clearly somewhat limit our analysis, notwithstanding that the most relevant evidence in terms of clinical impact have been adequately addressed. We will take into account either a patient-targeted or a community-based approach, centred on both individual or general interventions.

Clinical risk factors and risk for incidental AF

The relationship between several risk factors and incidence of AF has been widely discussed in several published studies. Recently, a large intersocietary consensus document reviewed the role of several risk factors associated with AF occurrence.15 Similarly, another a comprehensive review was recently published addressing the role of the main factors associated with AF.19

We summarised the main of evidence of the most relevant factors associated with AF occurrence in table 2.

Table 2

Association between main risk factors and risk of incident AF15 19

Clinical factors associated with the risk of AF can be considered in two different groups: (1) modifiable lifestyle risk factors and (2) cardiovascular risk factors. As regards the former, alcohol consumption and physical activity appear to be the most relevant ones. The role of alcohol in determining the occurrence of AF has been investigated quite earlier than compared to other lifestyle risk factors. We reported about 17 studies analysing relationship between alcohol consumption and AF, with almost 2 50 000 subjects enrolled and a long-term follow-up largely ranging up to more than 50 years. All the studies were consistently uniform in reporting that alcohol consumption increases the risk of AF occurrence and that this risk appears to be proportionally higher with increasing amount of alcohol.15 19 In particular, two large meta-analyses, even if they were different studies, concluded that there is a linear increase in AF risk with the increasing amount of alcohol intake. Kodama and colleagues reported an 8% cumulative risk for AF occurrence for every 10 g of alcohol per day,23 while Larsson and colleagues reported a similar 8% risk increase for every drink increment per day.24

The role of physical activity in determining AF has been largely debated. Several studies, with different designs, have been performed to assess whether having a regular physical activity would have been beneficial in decreasing the risk of AF. Up to date, the studies seem to agree on the dual role of physical activity. On one side, leisure exercise consistently reduces the risk of AF, even when progressively increasing in intensity; on the other side both retrospective and prospective studies in the context of athletes and endurance training setting reported that risk of AF constantly increases over time.15 19

Few studies investigated the role of emotions in determining AF occurrence, and despite the reduced attention that the issue attracted, these studies seem to consistently indicate that all those emotions related to an increased emotional strain (anger, tension, anxiety, panic) are associated with an increased risk of AF, even though this risk appears quite restrained.15 19

The role of smoking habit conversely seems to be quite controversial. Several studies reported a significant association between smoking and risk of AF, also found to be higher in subjects currently smoking compared with those formerly smoking and that recognise a dose–effect relationship (table 1). Conversely, several analyses, with similar characteristics, failed in proving an independent relationship between smoking and AF occurrence, in particular when smoking habit was considered to be one of the large multivariable models.15 19 This leads us hypothesise that in the context of patients with multiple risk factors smoking habit would not be relevant, but among younger and healthier subjects its role in determining AF risk would still be determinant.

Moving to cardiovascular risk factors, several factors have been found to be associated with risk of AF. Among them, diabetes mellitus and hypertension have been both extensively investigated. In particular, the relationship between diabetes mellitus and AF have been analysed in more than 1.5 million subjects in 12 studies with a follow-up spanning up until 38 years. The increase in risk ranged between 14% and 40%, according to the various studies. Also, a dose–effect relationship was found between HbA1c and AF risk, with a 13% increase in AF risk for every 1% increase.15 19 Moreover, it has been also found that a longer disease history depict a higher risk for AF occurrence.15 19

Similarly, hypertension has been strongly associated with AF risk. Both systolic and diastolic blood pressure increase was found to be associated with AF. According to the studies, uncontrolled blood pressure was associated with a higher risk of AF, which progressively increases as much as it is more uncontrolled. Also, a dose–effect association was reported both for systolic and diastolic blood pressure with up to 16%–17% increase in risk for each 10 mm Hg increase.15 19

The relationship between obesity and AF is strong and investigated in a number of studies.25 Indeed, at least 10 studies (table 1) reported an increased risk of AF occurrence in overweight and obese patients. This risk appears to be proportionally higher in obese patients than those who are overweight. Also, a dose–effect relationship was found, with a 45% increase in risk for every 5 kg/m2 increase in body mass index.15 19

Lastly, a strong role in determining AF occurrence was found for obstructive sleep apnoea syndrome (OSAS). Indeed, even if it was investigated only in two studies, with a relatively shorter follow-up (compared with the other risk factors examined), OSAS appears to be strongly related to the incidence of AF with its risk ranging between 55% and up to more than twofold (table 1).

Other cardiovascular risk factors have been investigated, as the type of diet, lipid and n–3 polyunsaturated fatty acid levels and air pollution but the results of various studies were inconsistent in supporting a role in determining an increased risk of AF.15 19

As recently pointed out, despite a lot of work done to analyse the epidemiological link between risk factors and incidence of  AF, scarce data and evidence are still available for the arrhythmogenic effect of these risk factors.18 Those limited available data seem to indicate that the presence of risk factors contribute at several levels, that is, in causing structural abnormalities, conduction abnormalities and early/delayed after depolarisations.18

Interventions to control risk factors and effects on AF burden

As we have shown above, a significant role of several risk factors has been acknowledged in determining the occurrence of AF. We could then hypothesise that specific interventions to control those risk factors would reduce the AF burden. As we reported in table 3, several interventions have been tested in order to reduce the incidence of AF or recurrent AF. Among them, those interventions aimed at increasing the physical activity and to reduce weight seem to be the most successful in reducing the burden of AF. Indeed, the findings of the CARDIO-FIT study show that an increase and gain in cardiorespiratory fitness is highly associated with a reduction in risk for AF recurrence.26

Table 3

Interventions to control risk factors and effectiveness to reduce incident and recurrent AF15 19

Regarding interventions aimed at reducing obesity, the same group developed the LEGACY trial, in which obese patients with AF underwent a specific intervention aimed at significantly reducing weight, after AF management based on a rhythm control approach.27 In this study, those patients showing the larger reduction in weight reported a significantly lower risk for AF recurrence and also those patients who constantly and steadily lost weight (ie, with less fluctuations in weight throughout the follow-up observation) had a greater reduction in recurrence than those with a larger variability in weight reduction.27 Another study reported that a structured weight management programme leading to improvement of multiple cardiometabolic risk factors (obesity, diabetes, high blood pressure, smoking, alcohol consumption and sleep apnoea) was effective in reducing the AF burden in highly symptomatic patients.28

Also for diabetes mellitus and hypertension, specific interventions showed a certain effectiveness in reducing AF burden.15 19 Despite that, the results seem to be less convincing. In patients with diabetes, for example, a significant role of metformin was found in reducing the risk of AF occurrence, but intensive glucose control failed to show effectiveness in reducing AF events.15 19 Conversely, in hypertensive patients the role of blood pressure-lowering drugs, in particular ACE inhibitors and angiotensin receptor blockers, seem to be inconsistent in reducing the risk of AF. Indeed, if they seem to work in those patients with a structural heart disease, in whom the role of the aforementioned drugs could also be relevant in other terms (ie, cardiac remodelling), they failed to reduce the risk of AF in patients on primary prevention (table 3).

Furthermore, the same authors,26 27 in another study, suggested an interesting approach. In the ARREST-AF study, an aggressive risk factor control management, aimed at controlling at the most all the major known cardiovascular risk factors,29 reduced the risk for AF recurrence in obese patients, with at least one patient showing a reduced recurrence risk for a cardiovascular risk factor after undergoing an ablation procedure.30 This approach seems to suggest that to reduce the risk of AF occurrence to the maximum, a stricter risk factor control is needed, in order to combine all the beneficial effects of all the possible interventions.

Indeed, already in the ARIC study more than one-half of incident AF events proved to be attributable to elevated or borderline levels of risk factors for AF. Consequently, borderline and optimal risk factor control was shown to progressively reduce the risk of AF occurrence.31 Therefore it can be hypothesised that a substantial number of AF events and AF-related complications could be avoided by prevention of the development of these cardiovascular risk factors.32

The evidence coming from these studies, investigating how the overall risk factors control impact the incidence of AF, are highly clinically relevant. Indeed, projecting the expected reduction of AF risk (based on the results from the ARREST-AF30 and ARIC studies31) on the expected risk (from the Rotterdam Study33), it clearly appears how relevant it would be to implement these strategies in the current management of subjects at risk of developing AF, in terms of reduction of absolute numbers of patients affected, in particular in the older subjects (figure 1). It is also clinically evident the stricter the risk factor control is, the higher would be the impact in terms of reduction of AF occurrence (figure 1).

Figure 1

Expected incidence of atrial fibrillation (AF) without any prevention strategy and implementing the suggested global management of risk factors. Details about risk factors control strategies can be found in Pathak et al30 for ARREST-AF and Huxley et al31 for ARIC.

Some evidence suggests that after the implementation of some preventive strategies, in the occurrence of AF the risk for adverse events is reduced in those patients randomised to the preventive strategy.34 35 More data to elucidate the impact of preventive strategies in those patients who develop AF are still needed.

The approach to AF: need for a paradigm shift with more focus on prevention

‘Prevention is better than cure’, wrote the Dutch humanist and scholar Desiderius Erasmus, known as Erasmus of Rotterdam during the 15th century. Even if this quote is well known, we can say that does not fit with the primary goals of our healthcare systems, which are mostly organised for providing care to subject with an already established disease, often after the appearance of some complications. In the case of AF, the ultimate complication is the occurrence of ischaemic stroke, and the fact that AF is detected, in some cases, just after an ischaemic stroke occurred36 is the proof of how much could still be done in terms of prevention.

The frequent occurrence of asymptomatic AF makes necessary to consider initiatives of AF screening (opportunistic or systematic) as an essential step, since identification of those subjects already on AF is essential for primary prevention of stroke in subjects presenting a relevant concurrent risk.37 38

Preventive measures can be applied at different levels and at different times according to the natural history of the disease. AF should be considered as a final pathway that would develop as the consequence of action due to one or more promoting, determining or facilitating factors.39 The final effect on AF development may depend on a variable combination of individual or genetic predispositions, extent of atrial structural or electrical remodelling, as well as ageing, underlying cardiac disease or associated comorbidities (figure 2).

Figure 2

The complex interplay between risk factors, atrial and cardiac substrates, ageing and predisposing factors that is object of interventions for prevention of atrial fibrillation and of stroke/systemic embolism.

The AF epidemic is becoming a worldwide threat and, in this complex and evolving scenario, it is worth considering a paradigm shift for healthcare interventions aimed at reducing the clinical and economic consequences of AF. This could imply overcoming the conventional ‘reactive’ approaches to healthcare and invest resources for more ‘proactive’ initiatives, in the context of ‘primordial’ prevention, to be applied at a community level, as well as in the context of primary prevention to be applied at an individual level.

Due to availability of limited data for some of the clinical factors reported, and also taking into account other possible risk factors not yet investigated enough, more evidence is still needed both in terms of epidemiological data reflecting association of clinical factors with incident AF and, more relevant, about reduction of AF burden in patients receiving specific interventions aimed at controlling the risk of new-onset AF.

In conclusion, AF prevention requires a series of initiatives focused on the many risk factors that we reviewed and should involve many stakeholders. Thus, it may not be easy to decide to dedicate efforts and resources to prevention strategies. For these reasons, until now AF prevention has been frequently applied in the context of a ‘fiction’ rather than a ‘fact’. However, the major threats linked to the growing impact of AF epidemiology in the coming years, as well as the important benefit that can be achieved through minimization of AF in the community and in individual patients calls for a new ‘alliance’ constituting synergetic interactions with the aim to reduce harms and costs for the community and individuals. In this perspective and with appropriate organisation, AF prevention may constitute a future ‘win–win’ strategy for all the stakeholders.



  • Contributors Both authors participated substantially to the work and meet all of the following conditions: (1) substantial contributions to the conception and design, acquisition of data or analysis and interpretation of data; (2) drafting the article or revising it critically for important intellectual content; and (3) final approval of the version to be published.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.