Age-Dependent Contributions of Rare and Common Genetic Variation in Atrial Fibrillation.

Abstract

Background: Genetic variation contributes to atrial fibrillation (AF), but its impact may vary with age. The All of Us Research Program contains whole-genome sequencing of data from 100 574 adult participants with linked electronic health records.

Methods: We assessed clinical, monogenic, and polygenic associations with AF in a cross-sectional analysis, stratified by age: <45 years (n=22 290), 45 to 60 years (n=26 805), and >60 years (n=51 659). AF was defined as ≥2 Systematized Nomenclature of Medicine-Clinical Terms codes on separate days. We identified pathogenic/likely pathogenic variants in 145 cardiac genes with dominant inheritance and calculated a previously established polygenic risk score. Adjusted for known clinical factors, multivariable analysis quantified associations between monogenic and polygenic factors and AF in each age group.

Results: Among 100 574 participants (mean age 59±16 years), 7811 (7.8%) had AF, while 92 763 (92%) did not. Monogenic pathogenic/likely pathogenic variants were associated with AF across all age groups, most strongly in participants aged <45 years (odds ratio, 2.1 [95% CI, 1.2-3.2]; P=0.007). In contrast, the polygenic risk score was not associated with AF in this youngest group (odds ratio, 1.0 [95% CI, 0.9-1.2]; P=0.650) but was in older groups (odds ratio 1.3 [95% CI, 1.2-1.4]; P<0.001 for both ages 45-60 and >60 years). Clinical factors were significantly associated with AF (C-index, 0.84 [0.83-0.84]; P<0.001), with marginal improvement when monogenic and polygenic data were added (C-index, 0.86 [0.86-0.87]; P<0.001). In hazard-based time-to-event analysis, monogenic variants were associated with earlier onset, whereas the polygenic risk score was not associated with age of onset.

Conclusions: In this large cross-sectional study, monogenic variants were associated with AF throughout life, particularly in younger participants, whereas polygenic risk was associated with AF only in older participants. While genetic information added only marginal improvements to AF risk discrimination beyond existing clinical risk factors, monogenic variants were associated with an earlier age of onset in participants with AF.