Calibration of Creatinine and Cystatin C assays in NHANES

Calibration of filtration markers to reference methods and the commutability across surveys are essential to the correct use of NHANES data for kidney function and for generating estimates of CKD prevalence.

Creatinine Calibration of the creatinine assays was performed in 2006 at the Cleveland Clinic Research Laboratory (CCRL).   We selected random samples of 200 stored surplus specimens from persons aged 60 years or older from each survey (NHANES III, 1999-200, 2001-2002, 2003-2004) for a total sample size of 800 and assayed them on the Roche enzymatic assay using the Roche Hitachi 737 analyzer.   The equations below shows the calibration factors needed to apply to each survey in order to obtain creatinine values traceable to reference materials.

Standard creatinine= -0.184 + 0.96 x [NHANES III calibrated serum creatinine]

Standard creatinine= 0.147 + 1.013 x [NHANES 1999-2002 uncalibrated serum creatinine]

Standard creatinine= [no calibration needed] NHANES 2001-2002 uncalibrated serum creatinine

Standard creatinine= [no calibration needed] NHANES 2003-2004 uncalibrated serum creatinine

Cystatin C The calibration of the serum cystatin C assay was performed in 2009 at the University of Minnesota (UMN). We selected random samples of 200 stored surplus specimens from persons aged 60 years or older from each survey (NHANES 1988-1994 or 1999-2002) for a total sample size of 400 and the measurements at the UMN were conducted using the same PENIA reagents run on the Siemens ProSpec analyzer (Siemens Healthcare Diagnostics acquired Dade-Behring).  Results were then multiplied by 1.12, the factor we had previously derived to calibrate the Simens assay at UMN to reference methods put forth by the International Federation of Clinical Chemists (IFCC).   As such, the equations to be used for standardization of serum Cystatin C in existing data from NHANES are as follows:

IFCC-standardized SCysC = 1.12 × [0.022 + 0.80 × (NHANES III–reported SCysC)]

IFCC-standardized SCysC = 1.12 × [(NHANES 1999-2002−reported SCysC) − 0.12]

These equations both eliminate the bias in the originally reported measurements that resulted from drift in calibration of the measurement procedures and standardize the assay to reference methods.

Last Updated: July 2014