Risk predictors for lone atrial fibrillation

Atrial fibrillation (AF) is the commonest sustained tachyarrhythmia, with a substantial morbidity and mortality. However, AF is commonly associated with a number of cardiac and non-cardiac risk factors including ischaemic heart disease, cardiac failure, valvular heart disease, hypertension, diabetes, alcohol abuse, thyroid disorders, and pulmonary disease. 1 It has been estimated that up to 80% of people presenting with AF have structural or organic heart disease, 2 and such cases have been referred to as ‘pathologic AF’. Clearly, this prevalence may differ in different study populations, where the so-called ‘pathologic AF’ may be less prevalent (e.g. 50%) among those with paroxysmal AF. Given the common relationship to a ‘sick’ heart, AF has often been considered to be an indicator of affliction by co-morbidities—such as valvular heart disease, heart failure, hypertension, and diabetes mellitus—that alter the substrate and/or haemodynamics (e.g. atrial stretch) sufficiently to precipitate AF. These common clinical disorders also contribute to the complications of AF, especially stroke and thrombo-embolism. What about ‘non-pathologic AF’? It is estimated that in ∼2–10% (and possibly as much as 30%) 3 of people presenting with AF, there is no identifiable aetiology and this subset of patients is often referred to as ‘lone AF’ (LAF). 4 Does ‘LAF’ really exist? Perhaps, this relies on definition and also how hard one looks for associated co-morbidities. There has been little acknowledgement for LAF in the large AF trials and evidence regarding its development, treatment, and prognosis is sparse. Therefore, it is unsurprising that there remains some debate as to whether LAF and pathological AF are two distinct disorders requiring different approaches and management strategies. In recent years, it has even been hypothesized that LAF and pathological AF are different disease processes with contrasting risk predictors. 2 Many demographic, genetic, and anthropometric factors have been implicated as risk predictors for the development of LAF. While pathological AF has an increasing prevalence with age and cardiovascular risk factors, LAF is more likely among younger populations. In one study, the mean age at diagnosis of LAF was 46.8 years. 5 It has been postulated that pathological AF occurs as a consequence of age-related changes and cardiac fibrosis whereas LAF occurs as a result of physiological changes related to autonomic tone, insulin sensitivity, and electrolyte imbalances exerting their effects at a cellular level. 2 The genetic associations with LAF have recently attracted renewed interest. Kato et al.6 studied 196 patients with chronic LAF, who were compared with 873 controls—they found that the T polymorphism of the matrix metalloproteinase 2 ( MMP2 ) gene and the C polymorphism of the interleukin 10 ( IL10 ) gene were significantly associated with the prevalence of AF (false discovery rate of <0.05). In fact, the T allele of the MMP2 polymorphism was a risk factor for the development of AF, whereas the C allele of the IL10 polymorphism was found to be protective for AF. Furthermore, the cardiac sodium-channel SCN5A R558 allele has been identified to be present in more patients with LAF compared with matched controls (LAF vs. controls, 30 vs. 21%; P = 0.002). 7 The identification of familial clustering of LAF has also strengthened these findings, suggesting some genetic pre-disposition for LAF. In an earlier study, the prevalence of LAF in different classes of relatives was compared with age- and sex-matched controls in the general population, and while family members had an increased relative risk of AF compared with the general population, this varied if one was a son (risk ratio 8.1), daughter (9.5), brother (70), sister (34), mother (4.0), or father (2.0). 8 Clearly, some patients with LAF may have been ‘pre-disposed’ to its development. Are there pre-disposing factors beyond genes? Mattioli et al.9 investigated the impact of socio-economic factors in a case–control cohort of 116 patients with LAF. When compared with controls, patients with LAF were more likely to exhibit a Type A behaviour pattern (20% patients with LAF vs. 9% controls, P < 0.001) and had much higher acute life-stress scores. Exercise also seems to be related to LAF, and the relationship appears to be well established, with LAF having an estimated prevalence of around 9% among athletes. 10 Interestingly, this risk appears to be related to more intense exercise, with a three-fold higher prevalence among vigorously exercising men. 11 Of note, a sporting practice of >1500 h was associated with more LAF (OR = 2.87; 95% CI, 1.20–6.91). 12 In the current issue of Europace , Mont et al.13 extend this work further, by investigating the relationship between physical activity, height, and left atrial (LA) size as independent risk predictors for LAF. In an emergency department setting, 107 patients with LAF (<48 h duration) were compared with 107 age- and sex-matched controls. They found that patients with LAF were significantly higher than controls and increasing cumulative physical exertion was associated with a greater risk of developing LAF in both the moderate and heavy physical activity groups. While this is an interesting observation, the cross-sectional design and the recruitment setting (i.e. emergency department) are limitations. Given that one-third of AF patients in primary care have not been under hospital care, 14 and that many AF patients are asymptomatic, the sceptic would argue that patients presenting to hospital would have had ‘something wrong’ with their health—or a pre-disposition to symptoms—sufficient for them to seek medical attention. Essentially, how thoroughly were co-morbidities (or heart disease) looked for? Thus, is LAF really ‘lone’ AF? If such patients are truly free of heart disease or co-morbidities, their prognosis should be good. The prognosis of patients with LAF has been investigated in different cohorts. For example, Jahangir et al.15 followed up 76 patients with LAF over a period of 30 years and found that their overall survival was 92 and 68% at 15 and 30 years, respectively. However, 31% of patients with either paroxysmal or persistent AF at the time of enrolment progressed to developing permanent AF. For their patients who had suffered a cerebrovascular event, it became evident that at least one further risk factor for stroke had developed, especially, hypertension. Thus, as patients get older, associated co-morbidities will intervene—the precise time when the dividing line between LAF and pathological AF changes, remains uncertain, given that many risk factors are asymptomatic. In contrast, Scardi et al.16 followed up a cohort of 145 patients with LAF (all <50 years age at time of diagnosis) and those with paroxysmal LAF had a much better prognosis than those with chronic LAF, with embolic complications occurring in 3 and 16.3% of patients, respectively, suggesting patients with chronic LAF to be at somewhat higher risk. However, older patients with LAF had a higher incidence of sudden death, raising the possibility that some did have underlying heart disease. Of note, the Framingham study 19 did show a 4–5 fold increase in stroke compared to controls, but 56% were aged >70 and 32% had a history of hypertension. In the Paris Prospective study, 17 even idiopathic LAF increased was an independent risk factor for cardiovascular mortality. Although LAF is increasingly considered as a specific disorder among patients with AF, a working clinical definition is needed. Many studies have led to the implication of independent risk predictors of LAF, such as genes, lower age, taller stature, and greater physical exertion, which differ from the risk factors associated with pathological AF; indeed, other risk predictors such as male gender and lower body mass index have also been identified 2 , 15 . For many, LAF is really a diagnosis of exclusion, where AF is associated with no obvious pre-disposing factor on thorough clinical history and examination, with a structurally normal heart on echocardiography, normal ECG (except for AF), normal blood tests, and normal chest X-ray. Essentially, do we look hard enough? Even subclinical and subtle changes in thyroid function tests may predispose patients to AF. 18 Clearly, larger studies and clinical trials are required to further define the LAF population, evaluate the risk predictor profile associated with LAF, as well as optimal treatment strategies and ultimate prognosis which may lead to changes in its identification and management. Conflict of interest: G.L. has received funding for research, educational symposia, consultancy, and lecturing from different manufacturers of drugs used for the treatment of AF and thrombosis. He was Clinical Adviser to the Guideline Development Group writing the United Kingdom National Institute for Health and Clinical Excellence (NICE) Guidelines on AF management ( www.nice.org.uk ), and is on the writing committee for the American College of Chest Physicians Guidelines on Antithrombotic Therapy.