Abstract
Aims: Although etiologic factors for diabetes-related lower-extremity amputation (LEA) have been extensively examined, the role of incident diabetic foot ulcer (DFU) has not been prospectively evaluated. We investigated the independent effect of incident DFU, among other limb- and person-level factors, on LEA risk in a cohort of veterans with diabetes at one Department of Veterans Affairs general internal medicine clinic.
Methods: We prospectively followed 1458 male U.S. veterans with 2893 lower limbs without DFU between 1990 and 2002 (mean follow-up = 4.9 years). Potential risk factors were evaluated in proportional hazards and flexible parametric survival models and collected through interviews (demographic, lifestyle, and diabetes characteristics), physical examination (blood pressure, weight, Charcot deformity, visual acuity), and blood (albumin, HbA1c, and eGFR [estimated Glomerular filtration rate]) and neurovascular (10-g monofilament, TcPO2, and ankle-brachial index [ABI]) testing.
Results: During follow-up, 227 (7.8%) DFUs occurred and 72 (2.5%) LEAs. Significant predictors (p < 0.05) of LEA in the final model adjusted for time-updated measurements (hazard ratio, 95% CI) included age ≥ 70 years (0.29, 0.08-0.97), prior amputation (2.31, 1.13-4.74), lower eGFR (1.16, 1.02-1.32), higher systolic blood pressure (1.40, 1.11-1.78), ABI ≤ 0.5 (3.94, 2.03-7.62), and incident DFU (10.44, 6.01-18.15).
Conclusions: In this 12-year prospective study, incident DFU emerged as having a substantial independent power in predicting LEA among other person- and limb-specific risk factors. Preventing DFU occurrence is critical for reducing the burden of LEA.