Characteristics associated with prevalent atrial fibrillation and risk profile for future atrial fibrillation among those in sinus rhythm in an elderly population in the ELSA Brasil Study

Abstract Background Atrial fibrillation (AF) is an arrhythmia that can lead to limiting symptoms and increases the risk of stroke and heart failure (HF). Purpose To evaluate the association of clinical, electrocardiographic, and echocardiographic parameters with prevalent AF or flutter (AFF), and the risk profile for future AFF according to Charge-AF and EHR scores in an elderly population. Methods The ELSA-Brasil is a prospective cohort study designed to investigate cardiovascular disease and diabetes in 15,105 men and women, civil servants from universities or research institutions in six cities of Brazil. We included all participants in ELSA-Brasil aged 60 and over whose diagnosis of AFF could be assessed and had echocardiography performed at the study's baseline. From the initial sample with participants ≥ 60 years (n = 3,263), 400 participants were excluded due to missing data about AFF, and 775 who did not have echocardiographic studies. Results There were 88 (4.2%) participants with AFF at baseline among the 2,088 (65 ± 4.1 years; 53% women) included in this analysis. Those with AFF were older (66.8 ± 4.4 vs 65 ± 4.0 y-o) and had a higher prevalence of HF (16% vs 3.1%), previous myocardial infarction (14.8% vs 3.8%), left bundle branch block (LBBB: 4.6% vs 0,9%), prolonged QT interval (7.9% vs 2.0%), supraventricular extrasystoles (4.6% vs 1.2%), and sinus bradycardia (11.3% vs 2.9%). AFF individuals also presented larger left atrial volume index - LA (33.3 mL/m2 versus 27 mL/m2), left ventricular end-diastolic diameter (4,63 ± 0,6 cm versus 4,47 ± 0,4 cm) and lower left ventricular ejection fraction (LVEF: 63.6 ± 9.1% vs 67.4 ± 6.4%) that non-AFF (all p-values <0.05). In multivariable analysis, prevalent HF, LBBB, larger LA, and lower LVEF were independently associated with the presence of AFF. Among those in sinus rhythm at study baseline, both risk models showed that most individuals in our sample were at low risk. In the Charge-AF, 63% of individuals were identified as presenting low risk (<2.5%), and 12% of the sample was categorized as high risk, with a risk of five-year AFF greater than 5%. Similarly, in the EHR, 67% of participants were classified as low risk, and 13% were considered high risk. Conclusion We found that AFF was present in 4.2% of the subset of older participants in a cohort composed of free-living individuals in Brazil. The presence of AFF was associated with HF history, altered cardiac electrical activity, LA dilation, and worse LV systolic function. Among those with sinus rhythm, 12-13% were considered at high risk of developing AFF (> 5% in five years) based on two clinical scores. Clinical surveillance and monitoring of electrocardiography and echocardiography parameters may help the early identification of risk factors for AFF.