There have long been controversies in different ways to assess sodium intake. Several formulas recognized internationally are created by foreign researchers. There is one equation, first proposed in China, created by Ningling Sun’s team, Peking University People's Hospital, China using the spot urine during afternoon to predict 24-h urine excretion. However, the participants in this study were hypertension population, which makes its promotion limited. [19] We used it in our study, but the result was not ideal.
Our study used the spot urine collections at all 4 times (morning, afternoon, evening and overnight) to estimate the 24-hour urinary sodium excretion. According to the Bland-Altman plots, in summer, the bias with the INTERSALT equation was the smallest. 94.5 ~ 95.4% of the individuals were distributed within the consistency boundary. In view of the autumn survey, predicted values with Tanaka equation with morning samples and the Kawasaki equation with overnight samples were most accurate. 93.7% and 94.5% of the individuals were distributed within the consistency boundary. Although the estimated results of the four methods were fluctuant, it can be seen that the predicted values with the Kawasaki and the Sun’s equations were overestimated and the INTERSALT method was underestimated. For the predicted values with the Tanaka equation, all results were overestimated without the overnight samples in both seasons. The amount of salt intake maybe an important reason of the estimated deviation. It is possible that people in different countries with different sodium intakes need different methods to estimate the 24-hour sodium intake[21].
From the view of correlation coefficient, the four methods we used in estimating the 24-hour urinary sodium excretion with spot urine samples also fluctuated. In the summer survey, the predicted values with Kawasaki equations had the highest correlation (0.612), which may due to the large amount of physical activities during daytime and perspiration, and less sweating at rest at night. However, in the autumn, the highest correlation was 0.635 and 0.616 respectively when Tanaka equation was used with afternoon and evening urine samples. Therefore, if the time point urine estimation method is adopted, overnight urine samples are the best in summer. For one reason that the overnight collection has a higher volume [22], it may be a better predictor. But in other seasons, it is recommended to use the afternoon or evening collection with the Tanaka equation, for the high correlation with the gold standard.
Although Kawasaki used the second urine sample in the morning of the Japanese population and obtained a correlation coefficient of 0.73 [18], Tanaka used the random urine samples of the Japanese population with a similar estimated method only obtained the correlation of 0.54 [16]. In a US study, all values estimated by time-point urine samples with prediction equations were moderately correlated with measured 24-h urinary sodium excretion, 0.40–0.60 [21]. In our study (in autumn), the correlation coefficient of Kawasaki method was 0.52, the INTERSALT method was 0.58 ~ 0.60, Tanaka method was 0.48 ~ 0.64, and Sun’s method was 0.48. The difference in correlations indicates that when using estimated methods to predict the 24-h sodium excretion with different random urine samples, it will fluctuate due to ethnic differences or the different time of urine collection.
In this study, we also explored the applicability of the 24 h urine collection (gold standard) in different seasons. The results showed that compared with autumn, the daily urine sodium excretion reduced about 520.7 mg (1.3 g synthetic salt). The dietary survey showed that there was no significant difference in sodium intake between the two seasons. Therefore, when using the 24-h urine collection to estimate sodium intake, we should avoid summer. If it has to in summer, an undervalued value should be added.
Our study had potential limitations. First, we just collected the 24-h urine for one day in each study. It is likely that the correlations observed are different if we had been collected on a different day. Second, we used the overnight urine sample instead of the second morning urine sample, which may result in bias for the predicted sodium excretion with Kawasaki formulae. Finally, we purposefully selected participants aged from 18 to 25 year-old which cannot represent the whole population in china. Thus, it is not known whether our results apply to younger or older persons and populations.