Variations in the prevalence of atrial fibrillation, and in the strength of its association with ischemic stroke, in countries with different income levels: INTERSTROKE case–control study

Background: The contribution of atrial fibrillation (AF) to the etiology and burden of stroke may vary by country income level. Aims: We examined differences in the prevalence of AF and described variations in the magnitude of the association between AF and ischemic stroke by country income level. Methods: In the INTERSTROKE case–control study, participants with acute first ischemic stroke were recruited across 32 countries. We included 10,363 ischemic stroke cases and 10,333 community or hospital controls who were matched for age, sex, and center. Participants were grouped into high-income (HIC), upper-middle-income (subdivided into two groups—UMIC-1 and UMIC-2), and lower-middle-income (LMIC) countries, based on gross national income. We evaluated the risk factors for AF overall and by country income level, and evaluated the association of AF with ischemic stroke. Results: AF was documented in 11.9% (n = 1235) of cases and 3.2% (n = 328) of controls. Compared to HIC, the prevalence of AF was significantly lower in UMIC-2 (aOR 0.35, 95% CI 0.29–0.41) and LMIC (aOR 0.50, 95% CI 0.41–0.60) on multivariable analysis. Hypertension, female sex, valvular heart disease, and alcohol intake were stronger risk factors for AF in lower-income countries, and obesity a stronger risk factor in higher-income countries. The magnitude of association between AF and ischemic stroke was significantly higher in lower-income countries compared to higher-income countries. The population attributable fraction for AF and stroke varied by region and was 15.7% (95% CI 13.7–17.8) in HIC, 14.6% (95% CI 12.3–17.1) in UMIC-1, 5.7% (95% CI 4.9–6.7) in UMIC-2, and 6.3% (95% CI 5.3–7.3) in LMIC. Conclusion: Risk factors for AF vary by country income level. AF contributes to stroke burden to a greater extent in higher-income countries than in lower-income countries, due to a higher prevalence and despite a lower magnitude of odds ratio.