Reliability of Two Ankle-Brachial Index Methods to Predict Silent Lacunar Infarcts: A Population-Based Study in Stroke-Free Older Adults (the Atahualpa Project).

Individuals with an abnormal ankle-brachial index (ABI) are four times more likely to have a silent lacunar infarct (SLI), but reliability of ABI in predicting the presence of these lesions has not been estimated yet. We compared two methods of calculating ABI to assess their reliability in predicting SLIs. Stroke-free Atahualpa residents aged ≥ 60 years underwent MRI of the brain and ABI determinations. Persons with ABI ≥ 1.4 were excluded. Using receiver operator characteristic curve analysis, we calculated the reliability of the traditional as well as an alternative ABI method to identify individuals with SLI. The traditional ABI uses the higher systolic pressure of either the dorsalis pedis or the posterior tibial arteries as the numerator, whereas the alternative ABI uses the lower pressure. Of the 247 participants, 38 (15%) had traditional and 95 (38%) had alternative ABIs ≤ 0.9. Twenty-one individuals had SLI. Traditional and alternative ABIs ≤ 0.9 identified 9 and 13 individuals with SLI, respectively. The traditional ABI had sensitivity of 42.9% (22.6-65.6%) and specificity of 87.2% (81.9-91.1%). The alternative ABI had sensitivity of 61.9% (38.6-81%) and specificity of 63.7% (57-69.9%). The area under the curve for the predictive value of SLI was 0.65 (0.54-0.76) for the traditional and 0.63 (0.52-0.74) for the alternative ABI ≤ 0.9. The ABI is moderately reliable for identifying candidates for MRI screening in studies assessing the burden of SLI in older adults. The traditional ABI seems to be more suitable for this purpose.