Prognosis in atrial fibrillationThe opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.

The public health importance of a disease is related to the number of people affected and to the mortality and mor- bidity that it causes. This metric underscores the clinical importance of atrial fibrillation (AF). AF is the most common of the supraventricular arrhythmias. Recent data suggest that there are 2.3 million individuals with AF in the USA alone.1 Also, there is a strong association of AF with advancing age. About 1% of individuals between 60 and 69 years of age have AF but this rises to 5% for individ- uals older than 69 years and to 9% for individuals in their ninth decade.2,3 With the known, dramatic aging of the population, the number will climb to more than five million by mid-century.1 Aging of the population alone would increase AF prevalence. There is also a suggestion that the age-adjusted prevalence of AF is also increasing,2 although the underlying mechanisms have not been defined. The clinical sequelae of AF are several, for some of which there is a clear mechanistic connection between AF and events. Stroke is likely related to stasis of blood in the non-contracting atria and occurs annually in 4.5% of un-anticoagulated patients.3 Significant progress has been made in the treatment of stroke. A series of major trials has demonstrated that anticoagulation with warfarin reduces the stroke rate by almost 70%.4-6 Aspirin also reduces the risk of stroke in these patients but is less effec- tive than warfarin.7 Patients with rapid ventricular response to their AF may experience dyspnoea with exertion because of non-physiologic reduction in filling time. While not as ominous as stroke, the impact on exercise tolerance does have a significant effect on quality-of-life and much of the day-to-day treatment of AF revolves around optimizing rate control or maintaining sinus rhythm to minimize exer- tional dyspnoea and/or palpitations. In patients with other haemodynamic compromise, new onset of AF or inadequate ventricular rate control can result in heart failure. A number of studies have found an association between AF and reduced survival. For instance, in a multivariable analysis of data on more than 5000 Framingham patients,8 AF was a risk factor for reduced survival independent of stroke or transient ischaemic attack, myocardial infarction, hypertension, age, smoking, ECG diagnosed left ventricular hypertrophy, heart failure, or valvular heart disease. Interestingly, the mode of death was similar to control patients. This multivariable analysis cannot take into account the time course and severity of the various variables and did not account all of the risk factors now known to be important (such as renal function). Other studies have not found an independent association of AF and survival. Further, the prognosis of younger (less than 60 years old) AF patients without cardiovascular disease appears similar to that of the general population, suggesting that it may be the underlying disease or non-cardiac causes, rather than AF, per se, driving prognosis.9 Attempts to maintain sinus rhythm in the AFFIRM trial10 were associated with a trend towards increased mortality, rather than a reduction, but this may have been medication-related. Finally, in general, the risk factors for AF are also risk factors for major adverse cardiovascular events and distinguishing an independent effect of AF from the effect of these risk factors is difficult.