Comparison of cystatin C- and creatinine-based estimated glomerular filtration rates for predicting all-cause mortality in Japanese patients with type 2 diabetes: the Fukuoka Diabetes Registry.

ABSTRACT

BACKGROUND: There is little information about the predictive ability of cystatin C-based estimated glomerular filtration rates (eGFRCys) for all-cause mortality in Asian populations. We compared the discriminatory ability of eGFRCys for all-cause mortality with that of creatinine-based estimated glomerular filtration rates (eGFRCr) in Japanese patients with type 2 diabetes. METHODS: A total of 4869 participants were classified into four categories (eGFR ≤29, 30-59, 60-89, and ≥90 ml/min/1.73 m2) by eGFRCr and eGFRCys, and followed up for a median of 3.3 years. RESULTS: 150 deaths were identified. The multivariable-adjusted risk of all-cause mortality was significantly increased in eGFRCr ≤29 ml/min/1.73 m2 compared with eGFRCr ≥90 ml/min/1.73 m2 [hazard ratio (HR) 2.4 (95 % confidence interval (95 % CI) 1.2-5.0)], whereas it was significantly increased in eGFRCys 59 ml/min/1.73 m2 or lower [30-59 ml/min/1.73 m2, HR 1.9 (95 % CI 1.1-3.5); ≤29 ml/min/1.73 m2, HR 5.8 (95 % CI 2.8-12.0)]. Comparing eGFRCys with eGFRCr, the proportions of participants reclassified to lower and higher eGFR stages were 6.3 and 28.8 %, respectively. The multivariable-adjusted HRs for all-cause mortality were 1.8 (95 % CI 1.1-2.9) and 0.7 (95 % CI 0.4-1.1), respectively. The C statistic of the model including eGFRCys and other risk factors was significantly increased compared with the model including eGFRCr. The net reclassification improvement and the integrated discrimination improvement were significantly positive. CONCLUSIONS: Our findings suggest that eGFRCys has a stronger association with all-cause mortality and is superior to eGFRCr for predicting all-cause mortality in Japanese patients with type 2 diabetes.