Association between ankle-brachial index and cardiovascular mortality in a cohort from the United States of America

Abstract Background Ankle-brachial index (ABI) is a non-invasive and accessible indicator of peripheral arterial disease (PAD) which shares pathophysiological mechanisms with coronary and cerebral atherosclerosis. Purpose This study investigated associations between ABI and cardiovascular (CV) mortality in a longitudinal cohort study. Methods Ankle-brachial index was measured in 7571 participants of the National Health and Nutrition Examination Survey (NHANES) between 1999 and 2004. Participant data were linked with National Death Index mortality data until December 2015. Participants were categorised as having low ABI, <0.9, indicative of PAD; or normal ABI 0.9–1.4. Cox Proportional Hazards Regression was used to investigate associations between ABI, CV mortality, adjusting for demographic variables (age, sex, educational status, ethnicity & poverty-income ratio) and traditional CV risk factors (HDL-C, LDL-C, triglycerides, systolic blood pressure, body mass index & diabetes) Results Individuals with low ABI were on average older (mean ± SD; 71.1±11.6 v. 59.3±12.8 years) and had higher systolic blood pressure (142.8±24.3 v 131.1±20.8) than those with normal ABI. Low ABI (<0.9) was associated with increased odds of cardiovascular mortality (Odds Ratio (OR) 3.092, 95% CI 1.81–5.28, p<0.001) after adjustment for all variables. Adjusted survival curves are shown in the figure, below. In a subgroup analysis of individuals with nominally normal ABI (0.9–1.4), ABI, treated as a linear variable was negatively associated with cardiovascular mortality, OR 0.185 (95% CI 0.051–0.067, per 1 unit change in ABI, P=0.01) after adjustment for demographic variables. In a subgroup analysis of individuals with low ABI, antihypertensive use was associated with lower odds of CV death (OR 0.246 (95% CI 0.091–0.663, p=0.006) independently of demographic variables, systolic blood pressure, LDL-C and use of statins and antiplatelets. Conclusion We have demonstrated strong associations between ABI and CV mortality. Our results suggest that ABI should be investigated as an additional tool in routine CV risk prediction, and highlight the importance of risk-factor management in PAD patients. Funding Acknowledgement Type of funding sources: None. Survival curves for low and normal ABI