Risk Predictors of Lower-limb Amputation in Patients with Type 2 Diabetes Mellitus in the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) Study

Introduction: Lower limb amputations associated with type 2 diabetes have major implications for morbidity and mortality. The aim of this analysis was to identify important risk predictors for future lower limb amputation on the basis of a large cohort of 9795 patients with type 2 diabetes mellitus in the FIELD study. Methods: Patients were randomised to receive fenofibrate 200 mg/day or matching placebo over five years, and amputation events (a prespecified tertiary endpoint) were documented at six-monthly intervals. Time to cardiovascular disease events and death according to the predicted risk of amputation was also evaluated. Multivariable proportional-hazards regression analysis using exhaustive-search methods was used to develop predictive models. Results: The main predictors of the first on-study amputation were a history of previous diabetic skin ulcer or nontraumatic amputation (hazard ratio (HR) 5.6), neuropathy (HR 3.0), peripheral vascular disease (HR 2.6), age over 65 years (HR 2.04) and height (HR 1.5 per 10 cm taller) (all P < 0.001). Other significant predictors included smoking, albuminuria, HBA1c, retinopathy and PTCA. Increasing risk of cardiovascular disease events and death was associated with increasing model-predicted amputation risk (P < 0.0001). Conclusion: Classical markers of macrovascular and microvascular risk predicted amputations. We also identified height as a major predictor of diabetic amputations, independent of the presence of neuropathy, confirming a previous report from an observational study. These findings could enable more aggressive targeting of modifiable risk factors among patients at high risk of amputations and cardiovascular events who would benefit most from therapeutic intervention.