Estimation of salt intake from spot urine samples in patients with chronic kidney disease

Makoto Ogura, Ai Kimura, Koki Takane, Masatsugu Nakao, Akihiko Hamaguchi, Hiroyuki Terawaki, Tatsuo Hosoya

ABSTRACT: BACKGROUND: High salt intake in patients with chronic kidney disease (CKD) may cause high blood pressure and increased albuminuria. Although, the estimation of salt intake is essential, there are no easy methods to estimate real salt intake. METHODS: Salt intake was assessed by determining urinary sodium excretion from the collected urine samples. Estimation of salt intake by spot urine was calculated by Tanaka's formula. The correlation between estimated and measured sodium excretion was evaluated by Pearson's correlation coefficients. Performance of equation was estimated by median bias, interquartile range (IQR), proportion of estimates within 30 % deviation of measured sodium excretion (P30) and root mean square error (RMSE).The sensitivity and specificity of estimated against measured sodium excretion were separately assessed by receiver-operating characteristic (ROC) curves. RESULTS: A total of 334 urine samples from 96 patients were examined. Mean age was 58 +/- 16 years, and estimated glomerular filtration rate (eGFR) was 53 +/- 27 mL/min. Among these patients, 35 had CKD stage 1 or 2, 39 had stage 3, and 22 had stage 4 or 5. Estimated sodium excretion significantly correlated with measured sodium excretion (R = 0.52, P < 0.01). There was apparent correlation in patients with eGFR <30 mL/min (R = 0.60, P < 0.01). Moreover, IQR was lower and P30 was higher in patients with eGFR < 30 mL/min. Estimated sodium excretion had high accuracy to predict measured sodium excretion, especially when the cutoff point was >170 mEq/day (AUC 0.835). CONCLUSIONS: The present study demonstrated that spot urine can be used to estimate sodium excretion, especially in patients with low eGFR.