Overestimation of renal function in glucocorticosteroid treated patients

Abstract

Creatinine clearance is commonly used as a parameter for individualization of dosages of drugs primarily excreted by the kidney. Nomograms and equations have been developed for estimating creatinine clearance from serum creatinine concentration, body weight, age and sex. Glucocorticosteroids are said to cause proximal muscle wasting and therefore may be expected to cause a decrease in the creatinine production rate. The purposes of the present investigation were first to evaluate by a computed tomography the effect of long term treatment with prednisone on the mid-thigh muscle area, and second, to establish whether the presumed supposed decrease in muscle mass was associated with a decrease in the urinary creatinine excretion rate, and hence in a systematic error whenever a nomogram is used to predicte creatinine clearance in such subjects. Patients taking prednisone had smaller mid-thigh muscle areas than controls. A linear relationship between the mid-thigh muscle area and the observed urinary excretion of creatinine was found, suggesting that the muscle loss could account for the decrease in the urinary excretion rate of creatinine. The ratio of observed to predicted (by nomogram) urinary creatinine excretion was lower in patients than controls, resulting in a corresponding underprediction of creatinine clearance by nomograms in the patients taking prednisone.