Multiple, random spot urine sampling for estimating urinary sodium excretion

The measurement of sodium intake may be important for the management of hypertension. Dietary surveys and 24-h urinary collection are often unreliable and/or impractical. We hypothesized that urinary sodium excretion can be accurately estimated through multiple spot urine samples from different days. All enrolled subjects were children of the coauthors of the study. Fifty-two 24-h urinary collections (4 per subject) for measuring sodium excretion and the 297 related urinary samples (1 per voiding) were collected for calculating the urinary sodium/urinary creatinine ratio in 13 children. The mean of 4 measured sodium excretions served as the individual “gold standard”. Twenty-four urinary collections were used to generate the equation predicting the mean measured sodium excretion from the mean of 4 urinary sodium/urinary creatinine [= 0.016 × urinary sodium (mmol/L) / urinary creatinine (mmol/L) ratio + 3.3)]; the remaining 28 urinary collections and 153 urinary samples were used for the external validation. All subjects underwent an additional validation procedure involving 12 urinary samples randomly collected on different days 6 months apart. The performance of sodium excretion calculated from a total of over 22,000 possible means of 4 out of all the available urinary samples, randomly taken on different days, was analyzed as to precision (by means of the coefficient of variation) and as to accuracy (by means of the P30). The coefficients of variations of measured vs. calculated sodium excretion were 25.3% vs. 25.8%, and the P30 of calculated sodium excretion was 100%. The excellent performance of calculated sodium excretion was confirmed both by external validation and by samples collected 6 months apart with mean P30s, all between 86 and 100%.