Sociodemographic characteristics and admission laboratory biomarkers associated with lower-limb amputation in hospitalised patients with diabetic foot disease: a single-centre retrospective cohort study
Year: 2026
Type: article
Study design: Observational Study
Abstract: Background Lower-limb amputation (LLA) is a disabling complication of diabetic foot disease, but the prognostic value of routinely available admission measurements remains poorly characterised among Chinese inpatients. Methods We built a retrospective adult cohort from the electronic medical records of all diabetic foot admissions to one tertiary hospital. Cases underwent LLA of any level during the index admission; controls did not. Associations were examined by logistic regression with collinearity control, restricted cubic splines, and LASSO, and performance was assessed by bootstrap validation, calibration, and decision-curve analysis; the models were also benchmarked against tree ensembles using the same evaluation protocol. Results Of 1,757 adults, 389 (22.1%) underwent LLA (158 minor, 231 major). Male sex (adjusted odds ratio: 1.77; 95% CI: 1.34–2.34), married status (adjusted odds ratio: 1.93; 95% CI 1.29–2.89), higher white blood cell count (1.07 per 1 × 10 9 /L), lower haemoglobin (0.78 per 10 g/L), lower uric acid (0.97 per 10 µmol/L), lower HDL cholesterol (0.60 per 1 mmol/L), higher D-dimer (adjusted odds ratio: 1.04 per 1 µg/mL; 95% CI, 1.00–1.07) and lower admission glucose (0.95 per 1 mmol/L) were independently associated with LLA; fibrinogen was not significantly associated after adjustment (adjusted odds ratio: 1.08, p = 0.096). In the sensitivity analysis restricted to case measurements explicitly documented before the first amputation operation, the haemoglobin association was attenuated and nonsignificant (0.95 per 10 g/L; 95% CI: 0.88–1.03), while fibrinogen strengthened (adjusted odds ratio: 1.24; 95% CI: 1.13–1.35). The apparent AUC was 0.733 (optimism-corrected: 0.705), the calibration slope was 0.75, and the Brier score was 0.15. Discrimination was moderate, mild overfitting was present, and clinical applicability is therefore limited pending external validation. Conclusions In this single-centre cohort, routinely measured admission characteristics—male sex, leucocytosis, lower haemoglobin, lower uric acid, lower HDL cholesterol, higher D-dimer, and lower admission glucose—were associated with in-hospital amputation. Given the observational design, the unavailability of the strongest clinical determinants of amputation, and the absence of external validation, these findings should guide heightened vigilance and the selection of variables for future validation studies, not individual clinical decision-making.
Source: Frontiers in Endocrinology
Authors: Weidi Wang, Hongkun Guo, Wenlong Zhang, Qiaoyi Wu
Language: English
Subfield: Endocrinology, Diabetes and Metabolism
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,801
OpenAlex ID: W7234754788
