Can lipid indices aid in predicting diabetic kidney disease? Findings from a cross-sectional, matched case-control study

Abstract

Background This study explored the association between lipid indices, including AC, TG/HDL-C ratio, AIP, and LCI, and diabetic kidney disease (DKD). Methods A cross-sectional, matched case-control study was conducted involving patients with type 2 diabetes (T2D), divided into two groups based on the presence of DKD. The groups were matched for age and duration of diabetes. The association between lipid indices and DKD was assessed using RCS, multivariable logistic regression, and ROC curve analysis. Results The study included 2940 individuals with T2D, with 1470 in each group. A nonlinear association was observed between all lipid indices and the presence of DKD. These lipid indices demonstrated relatively high predictive ability for DKD, with all AUC values higher than 0.707. The AIP and TG/HDL-C ratio had the highest AUCs of 0.717 and 0.713, respectively. Both indices also exhibited the highest sensitivity at 68%, while LCI showed the highest specificity at 79%. After adjusting for potential confounders, all lipid indices were significantly associated with DKD in the multivariable logistic regression analysis. Non-linear associations were found between lipid indices and components of DKD. All lipid indices demonstrated significant relationships with uACR ≥ 30 mg/g, whereas only AIP showed a significant association with eGFR < 60 mL/min/1.73m2. According to the ROC curve analysis, AIP was the most effective at identifying reduced eGFR (AUC = 0.676 [0.637–0.712]), and LCI was the best performer for detecting elevated uACR (AUC = 0.741 [0.701–0.783]). Conclusions Lipid indices may serve as valuable, non-invasive tools for the early detection of DKD, potentially leading to effective diabetes management and reducing the burden of DKD.