Autori: G. Ardissino 1, S. Ghiglia 1, M. Perrone, F 1. Tel, A. Mezzopane 1, A. Piantanida 1, T. Lettera 1, F. Mehmeti 2, P. Savina 1, G. Grimaldi 1, L. Xaiz 1, G. Ippolito 1, F. Napolitano 1, F.L. De Luca 1, P. Salice 1, A. Giussani, I. Cortinovis 1, S. Milani 1, D. Consonni 1
Affiliazioni: 1 Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milano, 2 ASST Valle Olona, Ospedale di Busto Arsizio
The estimate of Sodium (Na) intake is important for monitoring hypertensive patients on Na restriction. 24 hours urinary collection (24-hrsUC) is inaccurate because of errors in time and/or volume measuring. In addition, the urinary collection is unpractical and often done during weekends thus is not representative of the usual Na intake. Since spot urine sampling (SUS) is not affected by any of the above mentioned sources of error, we hypothesized that the mean urinary Sodium-to-urinary Creatinine ratio (uNa-to-uCrR) of multiple SUS, collected in different days, is more precise and accurate for estimating the average Na excretion compared to 24-hrsUC.
Methods: A total of 301 urine samples (1 for each voiding) and the related 52 24-hrsUC performed in different days in 10 heathy subjects (age range 3-65, 6 females, up to 11 voidings per day) were collected. For the purpose of comparison, uNa excretion in mEq/Kg/day from each sample was derived multiplying by 2 each uNa-to-uCrR (the mean conversion constant obtained from all ratios as determined from 24-hrsUCs). We calculated Lin’s correlation coefficient, mean bias, and 95% limits of agreement (LOA), of uNa for: 1) The single 24-hrsUC, using the individual overall 24-hrsUC average as reference; 2) 1000 random samples of means of 4 spot urine samples (1 per subject in 4 different days) using the individual overall 24-hrsUC average as reference. Statistical analysis was performed using Stata 15.
Results: 1) The single 24-hrs-UC urine collections showed a Lin’s coefficient of 0.78, with 95% limits of agreement (LOA) of +1.39 mq/Kg/die (difference between upper and lower LOA: 2.78). 2) In the 1000 random samples the average Lin’s coefficient was 0.82. Difference between upper and lower LOA was smaller than 2.78 in 58.2% of samples.
Conclusion: Sodium excretion can be estimated, on average, more precisely and accurately as the mean of 4 uNa-to-uCrR in different days, then with a single 24-hrsUC. In details the mean of 4 spot samples (expressed as mean ((uNa-to-uCr) x 2) will provide a more reliable estimate (in mEq/kg/day) in as many as 58% of determinations.