Hide and seek: does the toe-brachial index allow for earlier recognition of peripheral arterial disease in diabetic patients?

Abstract

Arterial calcification may render the ankle-brachial index (ABI) unreliable in diabetic patients. Although guidelines recommend the toe-brachial index (TBI) for patients with falsely elevated ABI arbitrarily defined as an ABI > 1.4, arterial calcification is also common among diabetic patients with an ABI ≤ 1.4. This could result in a "falsely normalized" ABI and under-diagnosis of peripheral arterial disease (PAD). We investigated whether diabetes invalidates the ABI as opposed to the TBI, and if the TBI may therefore be more suitable for detecting PAD in diabetic patients. The difference between ABI and TBI was compared between diabetic and non-diabetic patients with an ABI ≤ 1.4 referred to the vascular laboratory. A Bland-Altman plot was constructed to assess whether ABI-TBI differences were dependent on the magnitude of the measurements. Subgroup analyses were performed for patients with a normal ABI, and for patients with critical ischemia. The population comprised 161 diabetic (252 limbs) and 160 non-diabetic (253 limbs) patients (mean age 67). Median ABIs (0.79 vs. 0.80) were similar, while median TBI was 0.07 higher in diabetics (p = 0.024). The ABI-TBI difference in diabetics and non-diabetics was similar (0.32 vs. 0.35; p = .084), and was also similar for patients with a normal ABI. Moreover, ABI-TBI differences in diabetic- and non-diabetic patients overlapped, irrespective of the magnitude of the measurements. Diabetes was not associated with larger differences between ankle and toe pressures (mean difference -0.9 mmHg, 95% confidence interval -15 to 13 mmHg) among patients with critical ischemia. No evidence was found that the TBI may overcome the potentially invalidated ABI in diabetic patients with an ABI ≤ 1.4. ABI and TBI are strongly associated, and this relationship is not influenced by diabetes. Therefore, the TBI does not allow for earlier detection of ischemia in diabetes.