Abstract

Background. Early identification of diabetic kidney disease (DKD) in type 2 diabetes mellitus (T2DM) remains challenging due to limitations of conventional biomarkers. Body composition analysis using computed tomography (CT) may provide novel insights into DKD risk stratification.

Objective. To investigate the predictive value of abdominal CT-based body composition parameters as potential imaging biomarkers for early DKD risk in T2DM patients.

Methods. This retrospective cohort study enrolled 350 patients with T2DM from the Second Hospital of Ningbo between January 2020 and December 2024. Patients were stratified into the early DKD group (n = 171) and the T2DM control group without DKD (n = 179) based on the results of renal function assessment. Using Slice-O-Matic software, we measured area, index, and radiodensity of skeletal muscle and adipose tissue depots at the L3 vertebral level on abdominal CT images. Spearman correlation analysis evaluated associations between body composition parameters and renal function indicators. Univariate and multivariate logistic regression analyses identified independent risk factors for the development of early DKD. Receiver operating characteristic (ROC) curve analysis was employed to assess the predictive value of body composition parameters for early DKD.

Results. Multivariate logistic regression analysis revealed four independent predictors of early DKD. age (OR=1.04, 95% CI 1.01-1.07, P=0.023), high-sensitivity C-reactive protein (OR=1.02, 95% CI 1.01-1.04, P=0.007), renal sinus fat index (OR=0.48, 95% CI 0.29-0.78, P=0.003), and renal sinus fat density (OR=0.78, 95% CI 0.72-0.83, P<0.001). Multiple linear regression analysis demonstrated that renal sinus fat density maintained significant associations with both the urinary albumin-to-creatinine ratio (β=-1.61, P<0.001) and the estimated glomerular filtration rate (β=0.29, P=0.002) after adjusting for confounding variables. The combined clinical-body composition model (AUC = 0.82, 95% CI 0.78-0.87) and the body composition-only model (AUC = 0.78, 95% CI 0.73-0.82) both demonstrated superior predictive performance compared to the clinical-only model (AUC = 0.66, 95% CI 0.61-0.72).

Conclusions. Reduced renal sinus fat density emerges as a novel independent predictor of early DKD in T2DM patients, demonstrating potential utility as an imaging biomarker for early identification and risk stratification. These findings support the integration of CT-based body composition analysis into comprehensive DKD screening strategies.