Abstract Body: Introduction: Atrial fibrillation (AF) is a potent risk factor for stroke, cardiovascular disease (CVD), and death. Data suggest variation in AF across racial and ethnic groups, but little is known among Asian American, Native Hawaiian, and other Pacific Islanders (AANHPI). We evaluated the incidence of AF in disaggregated AANHPI subgroups vs. non-Hispanic Whites (NHW) in California and Hawaii.

Methods: We identified adults between 2012-2022 aged ≥30 years in Kaiser Permanente Northern California and Hawaii integrated healthcare delivery systems with no prior CVD. AANHPI subgroups included Filipino (N=193,327), Chinese (N=182,776), South Asian (N=92,738), Native Hawaiian/other Pacific Islander (NH/PI, N=64,488), Vietnamese (N=50,141), Japanese (N=45,502), other Southeast Asian (N=26,602), and Korean (N=21,989), with 1,975,444 NHW. Incident AF was defined as a hospitalization, emergency department visit, or ≥2 outpatient visits with a diagnosis code for AF through December 2023. We calculated age- and sex-adjusted rates of AF by race and ethnicity, and then used Cox regression to assess the association of AANHPI subgroup with incident AF after adjustment for age, sex, and clinical CVD risk factors.

Results: Among 2,653,007 eligible adults, mean±SD age was 49±15 years and 53% were women. Age- and sex-adjusted incidence (per 1000 person-years) of AF was 6.64 in NHW, and significantly varied in AANHPI subgroups: NH/PI (8.13), Filipino (5.32), other Southeast Asian (4.40), South Asian (3.79), Japanese (3.70), Korean (3.54), Chinese (3.45), and Vietnamese (3.28). Compared to NHW, in the fully-adjusted model, the adjusted hazard ratio (aHR, 95%CI) for AF was modestly higher for NH/PI (aHR 1.07, 1.02-1.13) but variably lower for Filipino (aHR 0.71, 0.69-0.73), other Southeast Asian (aHR 0.64, 0.57-0.71), Vietnamese (aHR 0.54, 0.50-0.59), Korean (aHR 0.54, 0.49-0.60), Chinese (aHR 0.53, 0.52-0.55), South Asian (aHR 0.51, 0.48-0.55) and Japanese (aHR 0.50, 0.48-0.53) (Figure).

Conclusions: Notable variation exists in incident AF across disaggregated AANHPI subgroups living in California and Hawaii, with a higher adjusted rate in NH/PI and variably lower rates for other subgroups compared with NHW that was not explained by demographic and clinical CVD risk factors. Differentiating potential pathophysiological mechanisms vs. variable ascertainment of AF across AANHPI subgroups will help improve personalized care strategies for prevention and treatment of AF.