Aims/hypothesis  

The Modification of Diet in Renal Disease (MDRD) equation has recognised limitations when using estimated GFR in persons at risk of chronic kidney disease. Equations based on cystatin C provide an alternative method. We compared performance of the MDRD equation with a selection of cystatin C-based formulae for estimation of GFR in normoalbuminuric patients with type 2 diabetes.

Methods  

Estimated GFR was calculated using the MDRD equation and the cystatin C formulae proposed by several investigator teams. Isotopic GFR was measured using plasma clearance of ⁵¹Cr-EDTA.

Results  

We studied 106 participants, of whom 83 (78%) were men with the following characteristics, mean (SD): age 61 (9) years, HbA_1c 7.10 (1.27)%, creatinine 89.0 (12.7) µmol/l, cystatin C 0.859 (0.234) mg/l and isotopic GFR 104.5 (20.1) ml min⁻¹ 1.73 m⁻². MDRD estimated GFR was 77.4 (13.6) ml min⁻¹ 1.73 m⁻² (p < 0.05 for difference from isotopic GFR). Cystatin C-based calculations of estimated GFR were: Perkins 124.5 (31.8), Rule 90.0 (30.0), Stevens (age) 96.0 (30.4) and Stevens (creatinine) 85.6 (19.0) ml min⁻¹ 1.73 m⁻² (p < 0.05 for difference with isotopic GFR). For Arnal’s, MacIsaac’s and Tan’s formulae cystatin-C estimated GFR were 101.7 (34.8), 102.1 (27.0) and 101.6 (27.8) ml min⁻¹ 1.73 m⁻², respectively (p = NS for difference with isotopic GFR). Cystatin C-based formulae were less biased and, with the exception of Perkins’ formula, more accurate to within 10% of isotopic GFR than MDRD.

Conclusions/interpretation  

Performance of cystatin C equations was superior to MDRD in normoalbuminuric patients with type 2 diabetes. These results support further evaluation of cystatin C for estimation of GFR in persons at risk of chronic kidney disease.