Feasibility of Assessment of Habitual Salt Intake Using a 24-h Urinary Salt Excretion Self-Measuring Device

High-salt diets are a major cause of non-communicable diseases worldwide [1] and a risk factor for cardiovascular diseases related to hypertension [2, 3]. The World Health Organization guidelines [4] set a target salt intake of <5 g/day for adults. According to the Japanese National Health and Nutrition Survey [5], salt intake in the Japanese population is approximately twice the recommended amount. The target in the national health promotion policy is 8 g/day, and targets for hypertension prevention in the Dietary Reference Intakes (DRIs) for Japanese [6] are 7.5 g/day for males, 6.5 g/day for females, and <6 g/day for preventing the progression of hypertension and chronic kidney disease.

Salt intake is indirectly assessed based on dietary intake or urinary excretion. The measurement of urinary sodium excretion using the 24-h urine collection method is considered the gold standard [7, 8], although this method is highly accurate, daily repetition is difficult, and single measurements are often used [7]. When measuring habitual salt intake, intraindividual variation (day-to-day) must be considered. In studies of Japanese participants, the number of survey days required to estimate an individual’s habitual sodium intake has been reported to be 8 [9] or 13–14 days [10], allowing for a 20% error margin. Although there are few reports on salt intake status in each season, seasonal variations in salt intake have been reported [11‒14]. Furthermore, some reports [15, 16] have suggested that urinary salt excretion varies even when individuals consume a diet with a fixed salt content. Therefore, some studies have suggested that multiple measurements using spot urine instead of 24-h urine collection may be suitable for assessing habitual salt intake at an individual level [17, 18]. Thus, an ideal measurement method would be one that is highly accurate and easy to repeat.

A simple salt intake measurement device (KME-03; Kohno ME Institute, Kanagawa, Japan) developed by Yamasue et al. [19] estimates daily salt intake using nocturnal urine (urine from nighttime to the next morning). The device measures urine volume and conductivity with a sensor, predicts sodium concentrations (using a formula derived from the relationship between conductivity and sodium concentrations), multiplies this by the urine volume to obtain the sodium volume, and finally estimates 24-h urinary salt excretion. This method is recommended for patients to evaluate their salt intakes according to the Guidelines for the Management of Hypertension [20] by the Japanese Society of Hypertension. The effect of salt reduction or lowering blood pressure by self-measurement of salt intake using the device has been reported [21‒24]. The use of nocturnal urine alone reduces the burden associated with the traditional method. However, the main problems are errors due to individual differences in the ratio of nocturnal urine volume to 24-h urine volume and overestimation of the salt content in individuals with low urinary Na/K because sodium and potassium are not measured separately.

In this study, the device (KME-03) was improved and validated for the self-measurement of 24-h urinary salt excretion. We aimed to contribute to nutrition epidemiology research and public health by improving the accuracy of nutritional assessment based on habitual salt intake and supporting self-management.

The previous and new versions of the KME-03 device are compared in online supplemental Figure 1 (for all online suppl. material, see https://doi.org/10.1159/000540797). The new version of the self-measuring device (hereafter referred to as the salt monitor) measures the amount of salt excreted per urination and sums these values to provide an accurate estimate of urinary salt excretion over 24 h. The capacity of the measuring cup was changed from 1 L to 500 mL to increase portability and to allow measurements of smaller volumes of urine. The number of urine volume sensors was increased to improve the accuracy of urine volume measurement. In addition, the equation for urinary sodium chloride concentration estimation was improved by collecting additional conductivity-Na concentration data from individuals with low urinary Na/K ratios.

Participants in our study were required to accurately and repeatedly perform 24-h urinary salt excretion measurements using two methods and dietary records. This required a high level of cooperation and compliance from the participants. For this reason, voluntary participants were recruited from the students of the dietitian training facility where the study was conducted. The sample size was estimated to be 20 participants for whom a significant correlation would be detected, assuming a correlation coefficient of 0.9 between the salt monitor and 24-h urine collection. Due to limited resources, the maximum number of participants was set at approximately 10 per year.

Ultimately, 20 healthy Japanese female volunteers (11 participants between May 2021 and February 2022 and 9 participants between May 2022 and February 2023) participated in the study. The eligibility criteria were not receiving treatment for hypertension, no kidney disease, and no special dietary requirements. Consent to participate was obtained from 21 participants; however, one participant relocated and was therefore excluded. Age, height, weight, alcohol consumption, smoking, and exercise habits were verified at the beginning of the study using a self-administered questionnaire.

Surveys were conducted in the spring (May), summer (July and August), fall (October and November), and winter (January and February) for two nonconsecutive days, excluding weekends and national holidays (Fig. 1). The interval between the two measurements was 2–10 days, excluding menstrual periods. On each survey day, measurements using the new salt monitor, 24-h urine collection, and diet recordings were performed. During the study period, no special instructions such as salt restrictions were provided. Participants were only asked to avoid days when special events involving eating or drinking occurred, when they were sick, or when they exercised intensely.

The appearance of the new device and the procedure for its use are shown in online supplemental Figure 2. The revised equation for estimating the urinary NaCl concentration is Y = 0.4289(X)^1.116

Y: NaCl concentration using the ion electrode method.

X: NaCl concentration using the conductivity method.

On the day of the measurement, the first urine sample in the morning was discarded, and measurements were taken from the second urination to the first urine sample the next morning. After each measurement, the urine was replaced with the proportional urine collector described below and a 1/50 volume was collected. The measurement results and time were recorded. After each day’s measurements, we received the salt monitor and records and cross-checked the recorded content with the values recorded in the instrument’s memory. The 24-h adjusted urine volume was obtained from the actual urine storage time based on the records [25], and the 24-h urinary salt excretion (g/day) was calculated as 24-h urinary salt concentration (g/L) × 24-h adjusted urine volume (L).

After each measurement with the salt monitor, urine was stored at 1/50 of the voided volume using a proportional urine collector (Urinemate P; Sumitomo Bakelite Co., Ltd., Tokyo, Japan), which was validated using whole urine collection [26]. The samples were immediately brought to the laboratory and transported to an external laboratory (SRL, Inc., Tokyo, Japan). Sodium and potassium levels were measured using electrode methods, and creatinine concentrations were measured using enzymatic methods.

The total urine volume was obtained by multiplying the weight of urine in the proportional urine collector by 50 and dividing it by the specific gravity of urine. Based on this, the total urine volume was corrected to the 24-h equivalent, and the 24-h urinary sodium, salt equivalent, and potassium excretion were determined.

To evaluate the completeness of urine storage objectively, the creatinine index was obtained from the 24-h creatinine excretion using Jossens’ formula, and a creatinine index of ≥0.7 indicated successful urine storage [27]. Days when the urine storage was incomplete were excluded from the analyses.

The participants recorded the type and quantity of food and beverages consumed and the time of intake on the day of the 24-h urine measurement. They recorded the weight or volume measured using a digital cooking scale, spoon, or cup, along with the unit of measurement. For foods purchased or consumed outside the home, the approximate quantities of all foods, the name of the product, and the company or restaurant were recorded. For meals recorded in approximate quantities, participants provided food images captured using a smartphone to improve the accuracy of the food intake estimation.

The recorded dietary contents were verified by student staff with training in dietary research and by a nutritionist with expertise in the field. Nutrient intake was calculated using the Standard Tables of Food Composition in Japan (eighth revised edition).

Agreement between the 24-h urine collection and salt monitor for all measurement dates was examined using the Bland-Altman plot [28]. The Bland-Altman plot is a graphical technique used to compare two different measurements intended to measure the same quantity. The plot displays the difference between the two measurements (y-axis) versus the mean of the two measurements (x-axis); a line representing the mean difference between the two methods is drawn horizontally in the center to indicate if there is a fixed error between the two measurements. In addition, another two horizontal lines are drawn from the mean difference to ±1.96 times the standard deviation of the difference. Points outside the lines, known as limits of agreement, indicate outliers where there is significant disagreement between the two measurements. Agreement between the 24-h urine collection method and the salt monitor was evaluated by the percentage of data that fell within the limits of agreement. The fixed error between the two methods was examined by a paired t test, and the proportional error was examined by Pearson’s correlation to show the difference between the two methods and the mean value.

In the assessment of habitual salt intake for each participant, the annual average urinary salt excretion by 24-h urine collection (8 days in four seasons) was used as the reference, and values measured using the salt monitor were compared. The ratio of urinary salt excretion to salt intake was set to 0.86 [29], and urinary excretion was converted to an equivalent intake for comparison with the intake reference value. We evaluated the ability of the salt monitor to classify participants according to the target values indicated by the Japanese Society of Hypertension and the 2020 DRIs for Japanese.

Values of p < 0.05 in two-tailed tests were considered significant. Statistical analyses were performed using IBM SPSS Statistics (version 26.0; IBM Corp., Armonk, NY, USA).

Data for 157 days were analyzed, excluding data for 3 days from two participants owing to incomplete urine collection. After confirming that there were almost no differences in the basic attributes of the 11 participants from May 2021 to February 2022 and the 9 participants from May 2022 to February 2023 and no difference in average outdoor temperatures during the study days in the study facility area (data not shown), data for all 20 participants were analyzed together. Basic characteristics and urinary and dietary salt contents are listed in Table 1. The mean salt intakes for 8 days were 8.50 ± 1.98 g/day using the 24-h urine collection method and 8.29 ± 2.28 g/day using the salt monitor.

Figure 2 shows the agreement between 24-h urine collection and the salt monitor measurements, examined by Bland-Altman plots. Over 157 days, 94% of data (147 days) were within the limits of agreement (Fig. 2a). The mean bias of urinary salt excretion based on the salt monitor compared with 24-h urine collection was −0.17 g/day (7.31 ± 2.67 g vs. 7.15 ± 2.61 g, p = 0.098) and no fixed errors were observed. There was also no significant proportional error based on the correlation between the difference in the two methods and the mean (r = −0.054, p = 0.504). All differences in mean salt excretion for individuals were within the limits of agreement (Fig. 2b). The mean bias of urinary salt excretion based on the salt monitor compared with 24-h urine collection was −0.18 g/day (7.31 ± 1.70 g vs. 7.14 ± 1.96 g, p = 0.334). There were no significant proportional errors between the two methods (r = 0.325, p = 0.162).

As shown in Figure 3, there was a strong correlation (r = 0.912, p < 0.001) between habitual salt intake using the two methods. In 19 of 20 participants, the habitual intake was ≥6 g, consistent with their salt monitor values. At the 6.5 g reference value for adult females defined by the DRIs for the Japanese population, sensitivity was 100% and specificity was 67%.

We developed and validated a new device for the self-measurement of urinary salt excretion. As an alternative to 24-h urine collection, spot urine samples [30, 31] are widely used in medical institutions. A recent report revealed a maximum difference of 3–4 g between 24-h urinary salt excretion estimated from various types of spot urine and actual measurements, suggesting that the ability to measure 24-h urinary salt excretion at the individual level is inadequate [32, 33]. Using the new device, all urine samples were analyzed throughout the day. Therefore, the 24-h urinary salt excretion can be accurately measured without being affected by the ratio of one excretion to 24-h urinary salt excretion.

The new salt monitor could determine the number of days of salt intake ≥6 g with high accuracy with the potential for repeated measurements. A recent survey on habitual nutritional intake among the Japanese population [34] reported that the 10th percentile of salt intake for women was 7.2 g in the 16-day survey. Therefore, the fact that few of the study participants achieved the recommended value of less than 6 g is considered credible. A recent study report indicates that estimating salt intake using urinary Na/K ratios effectively reflects changes in salt intake [35]. However, the Na/K ratio monitoring device is expensive and does not quantify daily salt intake; it can only detect changes in sodium intake when potassium intake levels remain stable.

This newly developed method for the measurement of 24-h urinary salt excretion without urine storage expands the options for assessing salt intake in nutritional epidemiology studies. This method also provides secondary information, such as urine frequency, volume, and temperature. Although there is a burden on the person taking the measurement, the individual does not have to carry a container of urine as is the case with conventional 24-h urine collection, and the individual can check the amount of salt excreted without bringing samples to a laboratory. This method shares advantages with previous self-measurement tools [19, 36] and is anticipated to be beneficial for home nutrition management.

This study had several limitations. First, the 24-h urine storage method uses proportional urine collection, rather than whole urine collection. Second, this study used only 8-week-day measurements to assess habitual intake, and holidays were excluded. In previous studies [9, 10, 13, 14], dietary surveys were conducted over 3–4 days in each season. Finally, the most serious limitation of the study was the small number of participants, which was limited to female nutrition students. Moreover, the sample size was small, with very few cases with salt intake <6 g and >10 g. According to a previous study of Japanese women [37], the average urinary sodium excretion for those in their 20s corresponded to a salt intake of 9.8 g/day, approximately 1 g higher than that in the national dietary survey at the time. Our study participants had a salt intake of 8.5 g based on the 24-h urine collection, which was lower than this previous estimate and more similar to the estimate in the dietary survey. In addition, since the urine collection and dietary survey were conducted on the same day, the measurement results may include “reactivity” [38] owing to survey implementation; we cannot rule out the possibility that the knowledge and enthusiasm of the participants, who were nutrition students, determined their urinary behavior and diet content. However, sex and age limitations allowed for comparisons with reference values. According to a recent National Health and Nutrition Survey [5], the average height, weight, and salt intake of Japanese women in their 20s are 157.5 cm, 52.0 kg, and 8.3 g, respectively, suggesting that the population was representative, despite the small sample size. Additionally, compliance with the measurements was high, contributing to the demonstration of the application of the new device. Although this study lacks male participants, we believe that the validity of the new method of assessing urinary salt excretion presented in this study can be generalized to some extent for the following reasons. In Japan, the average man eats more than women and accordingly consumes more salt. The salt monitor used in this study was a modified version of a conventional type. The conventional type has high measurement accuracy at an intake of approximately 10 g [39], which is a higher range than the range observed in this study. This may be supported by the fact that the participant with the highest salt excretion (upper right dot in Fig. 3) showed high agreement.

Large-scale studies in populations with different salt intakes, ages, sexes, and other characteristics are needed to compare salt intake assessment methods. In conclusion, this study confirmed the accuracy of the new device for measuring urinary salt excretion and showed that it can be used to assess habitual salt intake at the individual level. The device provides an alternative to traditional 24-h urine collection for repeated surveys to determine habitual salt intake and for self-management of daily salt intake.

We would like to thank all study participants and the staff of the Research Promotion Division of our institution.

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Ethics Committee of Jissen Women’s University (Approval No. H2021-05). Written informed consent was obtained from all participants.

The authors declare that they have no conflict of interest.

This work was supported by JSPS KAKENHI, Grant No. JP21K11579.

N.M. and K.Y. designed and conducted research and analyzed data. K.Y. and O.T. developed a salt monitor and improved its accuracy. N.M. wrote the manuscript and had primary responsibility for the final content. All authors have read and approved the final manuscript.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from corresponding author, N.M., upon reasonable request.