PS-C35-8: ESTIMATING A MODEL FOR 24-HOUR URINARY SODIUM EXCRETION FROM SPOT URINE SODIUM AND CREATININE OF BANGLADESHI POPULATION

Abstract

Background: The best estimate of population distribution and the average level of dietary salt intake is provided by measuring 24hour urinary sodium (Na) excretion in a representative sample. Collection of 24hour urine is difficult for community living people, and therefore, use of the spot urine Na concentration for estimation of daily salt intake was proposed as an alternative to 24hour urine collection. Purpose: To develop and validate equations to estimate the population mean levels of 24hour urinary Na excretion from spot urine Na for a Bangladeshi population. Methods: We utilized a dataset in which estimates of urinary Na levels were available from urine samples from 24hour urine collection and from spot urine of 217 participants, aged 40 to 59 years. We developed formulas to estimate 24hour urinary Na (24HUNa) by using 24hour urinary Creatinine (24HUCr), spot urine Na (SUNa), and spot urine creatinine (SUCr). For the reliability and validity tests of developed formulas, we use another dataset of 517 participants, aged 30 to 59 years, that also contains Na levels from urine samples from 24hour urine collection and from spot urine collection. Results: The obtained formulas were as follows: (i) PRCr (mg/day) = − 6.68 × age (year) + 14.68 × weight (kg) + 23.89 × height (cm) − 3198.18; (ii) estimated 24HUNa (mmol/day) = 42.40 × XNa0.270; where PRCr = predicted value of 24HUCr, XNa = SUNa (mmol/L)/SUCr (mg/L) × PRCr (mg/day). The correlation coefficients of Na (mmol/day) between estimated and measured from 24hour urine sample were 0.477, which was statistically highly significant (p < 0.001). The mean difference (bias) between estimated and measured Na (mmol/day) was 2.9, which was not statistically significant (p > 0.05). Intraclass correlation coefficient (ICC) between estimated and measured 24HUNa was 0.502. Conclusions: These formulas are useful for estimating population mean levels of 24hour urinary Na excretion, though this method is not suitable for estimating individual Na excretion.