Abstract 15607: Prevalence of Atrial Fibrillation (AF) Risk Factors and Lone AF in a General Healthcare Population

Abstract

Background: Evidence is increasing that genetic factors play a major role in AF predisposition, yet known variants explain only a small percentage of its polygenetic basis. Large AF populations enriched for genetic factors are needed to address this heritability gap. AF occurring at a young age in the absence of traditional risk factors, i.e., “lone” AF, represents a promising model to explore undiscovered genetic pathways. However, the prevalence of AF risk factors and lone AF in general healthcare populations is unknown.

Methods: Our aim was to determine, in the Intermountain Healthcare population, the prevalence of AF, AF risk factors (RF), and lone AF subjects, potentially available to explore the genetic basis for AF. The Intermountain electronic medical record (eMR) warehouse was searched between 1994-2012 for an initial ICD-9 code diagnosis of AF among subjects ≥18 yo and for associated demographics and RF, including age, sex, and diagnoses of valvular heart disease (VHD), hypertension (HTN), diabetes (DM), coronary heart disease (CHD), heart failure (HF), thyrotoxicosis (TTX), and obstructive sleep apnea (OSA). Body mass index (BMI) was available in a subset. RF were assessed overall and in AF subsets by age (early onset, ≤45 yo; late onset, ≥70 yo), and the prevalence of AF in the absence of RF was determined.

Results: The eMR search identified 94,289 subjects with a diagnosis of AF; age averaged 72 y (range 18-106), 51.3% were female; 90.7% were white; 19.2% had VHD; 59.4%, HTN; 23.4%, diabetes; 38.0%, CHD; 35.9%, HF; 2.2%, TTX; and 3.6%, OSA. Multivariable logistic regression identified HTN (OR=6.29), CHD (OR=4.85), female sex (OR=2.07), DM (OR=1.84), HF (OR=1.77), OSA (OR=0.65), TTX (OR=0.58), and obesity (OR=0.39) as predictors (all p<0.001) of late (n=60,178) vs. early onset (n=4062) AF. One or more RF was present in 83.1% of late versus 40.2% of early onset AF (99% vs. 85% including obesity), all p<0.001. Overall, 20,412 (21.6%) subjects had no AF risk factors, and 2,429 (2.6%) had no RF and were ≤45 yo.

Conclusion: AF generally is associated with age and co-morbid RF; however, sizable subsets occur in the absence of RF both overall (21.6%) and for age ≤45 y (2.6%, n=2427). These represent enriched target populations for genetic and familial studies.