Abstract
BACKGROUND: Peripheral arterial disease (PAD) and arterial calcification (AC) are frequent yet underdiagnosed vascular complications in individuals with type 2 diabetes mellitus (T2DM). The ankle-brachial index (ABI) is a widely used, non-invasive too for detecting these conditions. However, differences in ABI calculation methos can impact diagnostic accuracy and prevalence estimates. OBJECTIVE: To determine the prevalence of PAD and AC based on three ABI calculation methods in patients with T2DM attending a public hospital in Peru. METHODOLOGY: We conducted a cross-sectional study using data from the At-Risk Foot Program of the Endocrinology Department at Hospital María Auxiliadora (2015-2020). ABI was calculated for each lower limb using the highest, average or lowest systolic ankle pressure (SAP) from either the dorsalis pedis or posterior tibial artery as the numerator, divided by the highest brachial systolic pressure as the denominator. We applied a hierarchical classification: PAD was identified first (ABI < 0.9 in either limb), and among those without PAD, AC was identified (ABI > 1.3 in either limb); the rest were classified as normal. Prevalences estimates were calculated with 95% confidence intervals, and associations with clinical characteristics were explored using Poisson regression with robust variance. RESULTS: We included 643 subjects with a mean age of 61.4 years, 69.8% female. The prevalence of PAD was 7.8% (95% CI: 5.8-10.1), 15.4% (95% CI:12.7-18.4), and 28.2% (95% CI 24.7-31.7) using the highest, average, or lowest SAP as the numerator in the ABI, respectively. Conversely, the prevalence of AC was 18.2% (95% CI: 15.3-21.4), 11.0% (95% CI: 8.7-13.7), and 16.2% (95% CI:13.4-19.3). In all three methods, PAD was associated with older age (p < 0.05) and AC was associated with longer duration of diabetes (p < 0.01). CONCLUSIONS: Among patients with T2DM, PAD prevalence varied substantially (7.8% - 28.2%) depending on the ABI calculation method, while AC was present in up to 18.2%. The lowest ankle pressure method increased sensitivity and may be preferred in high-risk populations where avoiding missed diagnoses is critical. The highest ankle pressure method, which maximizes specificity, may be more suitable for general screening and comparability with existing literature, whereas the average pressure approach could be useful in research or prognostic modeling. The hierarchical classification strategy allowed PAD and AC to coexist in the same individual, although this was rare. Given the variability in prevalence across methods, local validation studies are needed to determine which approach optimally balances sensitivity, specificity, and clinical applicability in Peruvian diabetic populations.