Validation of daily urinary creatinine excretion measurement by muscle-creatinine equivalence

BACKGROUND: Twenty-four-hour urinary creatinine excretion (24hUCrE) is strongly correlated with skeletal muscle mass (SMM). This study suggests how to exploit the power of the SMM-24hUCrE correlation to assess the accuracy of 24hUCrE measurement. METHODS: Four hundred and sixty-six men, a subgroup of participants in the 2002-2004 follow-up examination of the Olivetti Heart Study, performed a 24-h urine collection to measure 24hUCrE and underwent bioelectrical impedance analysis to evaluate SMM. Linear regression analysis between 24hUCrE and SMM was used to calculate the muscle-creatinine equivalence and to develop an equation to predict the 24hUCrE depending on SMM. The accuracy of the 24hUCrE measurement was assessed using the change in the SMM-24hUCrE correlation coefficient upon variation in the percentage deviation (%D) between the measured and predicted 24hUCrE. RESULTS: The calculated muscle-creatinine equivalence was 1 g of 24hUCrE = 22.73 kg of SMM. The %Ds and the corresponding SMM-24hUCrE correlation coefficients were as follows: %D = 3.0, r = .997; %D = 4.7, r = .989; %D = 8.1, r = .963; %D = 10.5, r = .940; %D = 12.6, r = .909; %D = 18.9, r = .825; %D = 25.8, r = .707; %D = 33.5, r = .595; %D = 41.4, r = .453. CONCLUSION: The increase in %D corresponds to a reduced correlation between muscle mass and creatinine excretion, which indicated a poor performance in the measurement of the 24hUCrE. For studies on single individuals, where small variations in 24hUCrE could be significant, a %D up to 12.6% is suggested; on the other hand, a wider %D interval could be acceptable for population studies.