Ankle brachial index as an independent predictor of mortality in anticoagulated atrial fibrillation

Background: An abnormal ankle brachial index (ABI, the ratio of the ankle and the brachial systolic blood pressure) (≤ 0·90 or ≥ 1·4) suggests the presence of peripheral arterial disease (PAD) and has proposed as a marker of cardiovascular risk. We hypothesised that the ABI would predict mortality and adverse events in anticoagulated chronic nonvalvular AF patients.

Methods: We recruited 287 consecutive anticoagulated outpatients with permanent or paroxysmal nonvalvular AF who were stabilised for 6 months on oral anticoagulation (Oral anticoagulation; INR 2·0-3·0). ABI was performed following a standard technique. Cox models were used to determine the association between ABI, and bleeding, cardiovascular events and mortality.

Results: Median ABI was 1·09 (0·93-1·23) and 78 (27%) had an abnormal ABI. Abnormal ABI was associated with diabetes, heart failure and ischaemic heart disease (P = 0·006, 0·019 and 0·009, respectively), and a CHADS(2) score ≥ 2 (P = 0·016). Median follow-up was 861 (718-1016) days, during 21(7%) presented an adverse cardiovascular event, 23 (8%) major bleeding events and 18 (6%) died. ABI was an independent predictor for all-cause mortality, even after adjusting for CHADS(2) score (Cox multivariable regression analysis, HR 2·76(1·08-7·06), P = 0·033). Abnormal ABI was significantly associated with major haemorrhagic events [HR: 2·47(1·01-6·04); P = 0·047], even after adjustment for HAS-BLED score.

Conclusion: Abnormal ABI is common in AF patients, and ABI was an independent predictor for all-cause mortality, even after adjusting for CHADS(2) score. ABI was an independent predictor for major bleeding, even after adjusting for the HAS-BLED score. ABI could be a useful tool for improving risk stratification of anticoagulated AF patients.