Estimation of populational 24-h urinary sodium and potassium excretion from spot urine samples: evaluation of four formulas in a large national representative population

Objective: The aim of this study was to assess the validity of the estimation of 24-h urinary sodium (UNa) and urinary potassium (UK) excretion obtained through four formulae based on occasional urine samples. Design and methods: We analysed 2399 individuals (51% females) aged 18 to 96 years representatives of Portuguese population. Tanaka, Kawasaki, INTERSALT and NHANES formulae were used to predict 24-h UNa and UK excretions from spot morning urinary samples (OUrS). We compared it with validated real 24-h urine samples (VUrS) (24-h UNa: 4052 +/- 1432 mg/day, 24-h UK = 2928 +/- 1004 mg/day). We compared observed vs. estimated measurements by examining bias (observed minus predicted UNa and UK), the correlation and intraclass correlation (ICC) coefficients between measurements, and Bland-Altman plots. We analysed the differences between observed and estimated Na and K excretion across subgroups defined by quintiles of observed Na and K excretion and subgroups defined hypertension status and control. The area under the ROC curve was used to assess the discriminatory capacity of formulas between high-intake salt individuals from lowintake individuals, taking the arbitrary values 3000 and 3900 mg/ day for, respectively, Na and K intake. Results: Formulas produced significant mean bias for UNa: Kawasaki - 1277, INTERSALT - 569, NHANES - 116 and for UK Tanaka - 754, Kawasaki - 95 mg/day. Correlation coefficients were less than 0.360 and ICC coefficients were all less than 0.458 for both UNa and UK estimations. Bias varied across quintiles with overestimation of UNa at lower quintiles (by 29-105%) and underestimation at higher quintile (by 7-37%) regardless of formula. The Bland-Altman plots indicated a high dispersion of the estimates biases regardless of the formulae and normotension/hypertension condition particularly at higher levels. All formulas exhibited an area under the receiver operating characteristic curve below 0.67 both in normotensive individuals and hypertensive individuals. Conclusion: We found a poor agreement between estimated and observed measurements of UNa and UK in our large population. All these formulas incur in over-or underestimations of UNa and UK excretion that may be unreliable for clinical evaluation of individual's and mean population daily UNa and UK excretion.