AimsCell-free DNA (cfDNA) is associated with diabetes and cardiovascular diseases. Our study was to evaluate whether serum cfDNA could predict the progression of diabetic kidney disease (DKD).MethodsIn this prospective study, a total of 160 patients with DKD were enrolled, and the kidney function was followed up by measurement of estimated glomerular filtration rate (eGFR) and urinary albumin–creatinine ratio (UACR) for three consecutive years. At baseline, concentrations of serum cfDNA were measured. DKD progression was defined as two-continuous decrease in eGFR and changes of UACR from less than 300 mg/g at baseline to higher than 300 mg/g at last follow-up. Regression models were used to analyze associations of serum cfDNA with the DKD progression.ResultsIn total, 131 patients finished all the follow-up visits. At the end of the study, 64 patients showed decreased eGFR and 29 patients had changes of UACR from less than 300 mg/g at baseline to higher than 300 mg/g at follow-up. At baseline, the progression group had higher serum cfDNA levels than the non-progression group (960.49 (816.53, 1073.65) ng/mL vs 824.51 (701.34, 987.06) ng/mL, p=0.014). Serum cfDNA levels were significantly negatively associated with the 1.5-year eGFR change (r=−0.219 p=0.009) and 3-year eGFR change (r=−0.181, p=0.043). Multivariate logistic analyses showed that after adjustment of age, gender, body mass index, fast plasma glucose, smoking, triglycerides, total cholesterol, duration of diabetes, systolic blood pressure, diabetic retinopathy, eGFR, high sensitivity C-reactive protein, angiotensin receptor blocker/ACE inhibitor usage, with the increase of one SD of serum cfDNA levels, the risk of DKD progression increased by 2.4 times (OR, 2.46; 95% CI 1.84 to 4.89).ConclusionSerum cfDNA is closely associated with DKD, and it might be a predictor of DKD progression in patients with type 2 diabetes.