OR29-07 A Novel Estimate of Creatinine Excretion to Determine Adequacy of 24-Hr Urine Collection

24-hour urine collections are used to assess excretion of various analytes. Although concomitant measurement of creatinine excretion adjusted for body weight (BW) is utilized to determine adequacy of collection, no gold standard exists for determining completeness of a collection. Use of current reference ranges for daily creatinine excretion/BW of 15–20 mg creatinine/kg BW/day in women and 20–25 mg creatinine/kg/d in men established prior to the rising prevalence of obesity has resulted in a large proportion of contemporary individuals appearing to have “incomplete” urine collections. Our objective was to evaluate the range of creatinine excretion in accurately collected urine specimen from adults with a wide spread of age and BW, and to generate an equation that accurately predicts 24-hour urine creatinine excretion rate from readily available clinical parameters to aid in assessing adequacy of a 24-hour urine collection. We analyzed data from participants who completed two consecutive 24-hour urine collections while consuming a fixed metabolic diet during inpatient research admissions. Results from participants with 24-hour urine creatinine excretion rate differing by >10% between the two consecutive collections were excluded. In the initial 115 pairs of inpatient 24-hour urine collections (50 female, 65 male) participants, creatinine excretion/BW fell outside the currently accepted reference ranges in >50% of collections. The proportion below the reference range increased with higher BMI. In this derivation dataset, linear regression models were then constructed to predict 24-hr urine creatinine excretion from race, sex, age, weight and height. Reliable prediction of observed 24-hr urine creatinine excretion was confirmed in a validation dataset that included 50 pairs of 24-hour urine samples similarly collected in an inpatient research setting. This new prediction model performed significantly better than the currently used reference ranges in a large outpatient dataset including 1,399 pairs of 24-hour urine collections. In women, actual creatinine excretion fell within the 95% prediction intervals for our derived equation in 90% of cases using the new interval vs 46% using the current reference range. The corresponding values were 90% and 33% in men. In both genders, the superiority of the new prediction over the current reference range was more pronounced at higher BMI. We therefore propose revision of currently used criteria to define adequacy of 24-hour urine collection to account for the impact of obesity. The proposed equation incorporates readily available demographic parameters to predict urine creatinine excretion. These findings have wide implications on patient care and research studies, and future studies should test this equation in different settings, diets, and populations.