Login to Access Video or Poster Abstract: MP01-10
Sources of Funding: Not applicable

Introduction

Approximately half of 24-hour urine collections in prior studies have been found to be outside the accepted cutoffs with urinary creatinine to weight ratio. Our study objective was to evaluate the relationship between measured urinary creatinine and lean body mass calculated 24-hour creatinine excretion based to improve accuracy for the evaluation of 24-hour urine collection adequacy.

Methods

This was a retrospective evaluation of 24-hour urine collections for 1319 unique nephrolithiasis patients. An established formula (figure 1c) previously to estimate urinary creatinine based on lean body mass was applied all patients in our cohort (Yu et al). We divided our cohort into two equal partitions (training and validation datasets) using a random number generator. Linear regression was used to quantify the relationship between the calculated and measured urinary creatinine in our training dataset. We then applied this relationship to our validation dataset. Two standard deviations from the expected value was considered an &[Prime]abnormal&[Prime] 24-hour urine collection. This regression-based approach was then compared with the standard method of verifying 24-hour urine adequacy.

Results

As demonstrated in Figure 1 for the validation cohort, there was a strong relationship between expected 24-hour urine creatinine based on lean body mass and the measured urinary creatinine (r²=0.7, p<0.01), which was stronger than the relationship between measured urinary creatinine and weight (r²=0.5, p<0.01). Using the traditional metric of Cr/Kg, 38% of patients in our cohort were considered to have an inadequate 24-hour urine collection (Figure 1a), and there more dispersion among inadequate specimens. Using the regression model on the validation set (Figure 1b), 15% of specimens were considered inadequate, and the majority were due to under collection.

Conclusions

Urinary creatinine is more strongly correlated to an individuals muscle mass rather than body weight. Our regression model demonstrated the utility of a lean body mass based 24-hour creatinine estimator that improves the ability to determine the adequacy of a 24-hour urine collection.

Funding

Not applicable