[Prediction of 24-hour urinary creatinine excretion from age, body weight and height of an individual and its application]

Abstract

Equations to predict 24-hour urinary creatinine excretion (UcrV) from age, body weight (W) and body height (H) were developed based on the data for UcrV determined by collecting 24-hour urine for 3-5 days in 256 male and 231 female clinically healthy subjects, aged from 20 to 84 years. By regression analysis a regression equation was developed. Multiple regression analysis revealed that age, W and H contributed most to the measured value of UcrV. The regression equations to predict UcrV from age, W and H were as follows: Predicted value of UcrV (PRCr) = -12.63 x Age + 15.12 x W + 7.39 x H - 79.90 (male: mg/day) PRCr = -4.72 x Age + 8.58 x W + 5.09 x H - 74.50 (female: mg/day) The multiple correlation coefficients were 0.87 for males and 0.73 for females, respectively. The difference between PRCr and measured UcrV (residual) was +0.81 +/- 154 (SD) mg/day on the average. An analysis of residuals by means of plotting of standardized residuals and predicted standardized dependent variables showed no remarkable deviations in both sexes, suggesting that these equations are applicable to the prediction of UcrV. These equations were applied to 38 Japanese men and women who were asked to collect 24-hour urine for 14 days on average, and also to 47 foreigners (American, Nepalese and Italian). The average residual in the former was +27 +/- 87 mg/day and that in the latter -3 +/- 85 mg/day, suggesting that these equations can be applied not only to the Japanese but also to the foreigners as a useful index in predicting 24-hour urinary creatinine excretion.