Background: The ankle--brachial index (ABI) is a simple, noninvasive, widely used test that detects peripheral arterial disease (PAD). In patients with diabetes, the ABI is notoriously unreliable and this is usually attributed to medial calcinosis, which stiffens the arteries and renders them poorly compressible. However, the distribution of atherosclerotic lesions in those with diabetes is different as well: lesions predominantly reside in below-the-knee (BTK) arteries. To what extent this contributes to the unreliability of the ABI is unknown. The aims of this study were (1) to confirm the notion that the ABI poorly predicts PAD in the diabetic foot, (2) to determine whether arterial calcifications can be blamed, and (3) to establish the role of the distribution of atherosclerotic lesions.
Methods: We studied 187 lower extremities with a diabetic foot that had an intra-arterial angiography and an ABI. The extent of atherosclerosis on angiography was rated by scoring all arterial segments from the aorta to the foot conforming to the Joint Vascular Societies reporting standard. Arterial calcification was assessed using a 4-level severity scale based on the number and length of calcified arterial segments as seen on plain X-ray. The ABI was calculated using four arterial pressures (two brachial, dorsal pedal, and tibial posterior arteries). To prove that the standard ABI is ill-designed to detect atherosclerosis located in BTK arteries, we postulated that an ABI that incorporated more information about these arteries would correlate better with angiographic atherosclerotic disease. To this end, we compared the standard ABI (which only uses the highest distal pressure available and neglects the pressure in other BTK arteries) with two alternatively calculated ABIs: one using the lowest pressure as numerator, and one with the average of both pressures while assuming a pressure of 0 for arteries that were not found by Doppler.
Results: The ABI could be determined in 123 cases (65.7%), mean ABI was 0.92. Analysis of the angiographies showed that atherosclerotic lesions had a high predilection for BTK arteries. The correlation between ABI and angiographic PAD was weak (Pearson correlation coefficient r = -0.487). Arterial calcification was absent in 36.9%, and considered to be light (5.4%), moderate (17.1%), or heavy (40.6%). Patients with moderate or heavy calcifications were grouped together and considered to have calcified arteries. These individuals had a lower ABI (p < 0.013), more advanced atherosclerotic disease (p < 0.001), and a higher interbrachial pressure differential (p < 0.017). Of all the ABIs tested, the one that used both distal pressures correlated best (r = -0.554) with atherosclerotic disease. Irrespective of how the ABI was calculated, patients without calcifications consistently had a better correlation with atherosclerotic disease than those with calcified arteries.
Conclusion: The ABI underestimated angiographic atherosclerotic disease in this population of patients that we thought had PAD causing tissue loss. Calcifications were frequently present and indeed can be partly blamed for the unreliability of the ABI, but the distribution of atherosclerotic disease can be held responsible as well: the ABI is hampered by design because it is meant to detect a drop in pressure caused by the additive effect of serially located lesions in the arterial tree. In the diabetic foot, where lesions tend to be situated in BTK arteries (which lie parallel to each other), the pressure measured in one distal artery is less representative of atherosclerotic disease in the lower extremity.
Copyright © 2011 Annals of Vascular Surgery Inc. Published by Elsevier Inc. All rights reserved.