Comparison of cystatin C- and creatinine-based estimation of glomerular filtration rate according to glycaemic status in Type 2 diabetes


Tae Seo, Sohn, Uijeongbu St Mary’s Hospital, Geumo-dong, Uijeongbu-si, Gyeonggi-do, 480-717, Korea. E-mail:


Diabet. Med. 29, e121–e125 (2012)


Aims  The influence of hyperglycaemia on the performance of glomerular filtration rate (GFR) estimating equations remains to be determined. We compared the performance of creatinine-based GFR with cystatin C-based GFR in patients with Type 2 diabetes according to glycaemic status.

Methods  In a cross-sectional study of 210 patients with Type 2 diabetes, we staged glycaemic status by HbA1c tertiles [HbA1c≤ 75 mmol/mol (9.0%) (n = 70), HbA1c 76–95 mmol/mol (9.1–10.8%) (n = 70), HbA1c >95 mmol/mol (10.8%) (n = 70)] and measured GFR. Isotopic GFR was measured using renal dynamic imaging with 99mTc-diethylene-triamine-penta-acetic acid. Estimated GFR (eGFR) was measured using creatinine-based formulae (Cockcroft–Gault-eGFR, the Modification of Diet in Renal Disease equation-eGFR and the Chronic Kidney Disease Epidemiology Collaboration formula-eGFR) and a cystatin C-based formula (cystatin C-eGFR).

Results  The isotopic GFR of all patients was 93.1 ± 34.1 ml min−1 1.73 m−2. All methods for estimating GFR underestimated isotopic GFR [Cockcroft–Gault-eGFR (68.8 ± 38.6 ml min−1 1.73 m−2) (P < 0.05), Modification of Diet in Renal Disease-eGFR (74.8 ± 31.3 ml min−1 1.73 m−2) (P < 0.05), Chronic Kidney Disease Epidemiology Collaboration-eGFR (72.9 ± 26.6 ml min−1 1.73 m−2) (P < 0.05) and cystatin C-eGFR (83.5 ± 33.2 ml min−1 1.73 m−2) (P < 0.05)]. In all patient groups, cystatin C-eGFR was less biased and more accurate than the creatinine-based formulae, especially in the group with HbA1c > 95 mmol/mol (10.8%) where there was no difference between cystatin C-eGFR and isotopic GFR.

Conclusions  Performance of cystatin C-eGFR was superior to creatinine-based GFR in patients with Type 2 diabetes with HbA1c > 95 mmol/mol (10.8%).