Identifying diabetic kidney disease (DKD) risk before overt renal decline remains a major challenge in diabetes care. A study published in Cardiovascular Diabetology evaluated whether the Cholesterol, High-Density Lipoprotein, and Glucose (CHG) index and its body mass index–adjusted derivative (CHG-BMI) could improve the prediction of DKD development and progression compared with traditional surrogate markers of insulin resistance.
The primary analysis used 10-year longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS, 2011–2020), including 984 adults with diabetes and no baseline kidney injury. Incident DKD was assessed with Cox proportional hazards models, Kaplan-Meier analysis, and restricted cubic spline modeling. External validation was performed using the National Health and Nutrition Examination Survey (NHANES 1999-2018). Progression analyses included an independent cohort of 217 patients with biopsy-confirmed DKD, defined by estimated glomerular filtration rate (eGFR) below 60 mL/min/1.73 m².
In CHARLS, CHG-BMI independently predicted incident DKD in fully adjusted models, with Quartile 3 showing a hazard ratio (HR) of 2.19 (P=0.020). It also improved discrimination versus traditional markers, with significant integrated discrimination improvement (IDI, P=0.030). Subgroup analyses showed stronger associations in non-overweight and normolipidemic individuals (interaction P<0.05). NHANES confirmed these findings and demonstrated J-shaped relationships for CHG and CHG-BMI.
In the biopsy-confirmed cohort, CHG and CHG-BMI were independently associated with lower eGFR (P<0.05), unlike conventional markers. Adjusted area under the curve (AUC) values for predicting DKD progression were 0.733 for CHG and 0.730 for CHG-BMI. Significant interactions were also observed with Renal Pathology Society grades, particularly classes IIa and IIb.
These findings suggest that CHG-based indices may strengthen DKD risk stratification, especially in earlier disease stages and in individuals without obvious metabolic risk features.
