Skip main navigation

Twenty-Four–Hour Urinary Sodium and Potassium Excretion and Their Associations With Blood Pressure Among Adults in China

Baseline Survey of Action on Salt China
Originally publishedhttps://doi.org/10.1161/HYPERTENSIONAHA.120.15238Hypertension. 2020;76:1580–1588

Abstract

This study aimed to assess current level of sodium and potassium intake and their associations with blood pressure (BP) using the 24-hour urinary data in a large sample of China. Data from participants aged 18 to 75 years were collected as the baseline survey of Action on Salt China in 2018. Of 5454 adults, 5353 completed 24-hour urine collection. The average sodium, potassium excretion, and sodium-to-potassium molar ratio were 4318.1±1814.1 mg/d (equivalent to 11.0±4.6 g/d of salt), 1573.7±627.1 mg/d, and 5.0±2.1, respectively. After adjusting for potential confounding factors and correcting for regression dilution, each 1000-mg increase in sodium excretion was associated with increased systolic BP (1.32 mm Hg [95% CI, 0.92–1.81]) and diastolic BP (0.34 mm Hg [95% CI, 0.09–0.60]). Each 1000-mg increase in potassium excretion was inversely associated with systolic BP (−3.19 mm Hg [95% CI, −4.38 to −2.20]) and diastolic BP (−1.56 mm Hg [95% CI, −2.29 to −0.90]). Each unit increase in sodium-to-potassium molar ratio was associated with an increase of systolic BP by 1.21 mm Hg (95% CI, 0.91–1.60) and diastolic BP by 0.44 mm Hg (95% CI, 0.24–0.64). The relationships between sodium and BP mostly increase with the rise of BP quantiles. Potassium shows the opposite trend. The current sodium intake in Chinese adults remains high and potassium intake is low. Sodium and sodium-to-potassium ratio were positively associated with BP, whereas potassium was inversely associated with BP.

Registration—

URL: https://tinyurl.com/vdr8rpr; Unique identifier: ChiCTR1800017553. URL: https://tinyurl.com/w8c7x3w; Unique identifier: ChiCTR1800016804. URL: https://tinyurl.com/s3ajldw; Unique identifier: ChiCTR1800018119.

Introduction

Evidence has shown that dietary sodium intake and potassium intake are associated with blood pressure (BP).1,2 Although there are reports available on sodium and potassium consumption in different populations, precise quantification of dietary sodium and potassium intake in population is difficult due to the high variability of dietary intake patterns. Multiple complete 24-hour urine collections were regarded as the most accurate method for sodium and potassium assessment, which overcomes the limitations of dietary measures and accounts for intraindividual variability. When measuring sodium intake in a population, an adequate sample size with only one 24-hour urine collection was considered a robust estimate of sodium intake at the population level.3,4 However, collecting complete 24-hour urine samples remains challenging, as this requires high compliance from the participants.

A large-scale 24-hour urine collection took place in 2018 as part of an ongoing salt reduction program, Action on Salt China. With the aim of reducing salt intake by 15% by 2021, Action on Salt China consists of 4 cluster randomized controlled trials (RCTs) in 6 provinces of China to develop and evaluate different salt reduction strategies in primary school children and their families (AppSalt-based salt reduction in primary school children and their families [AIS]), home cooks and their families (Home cook salt reduction intervention study [HIS]), whole communities (Community-based comprehensive salt reduction intervention study [CIS]), and restaurants (Restaurant-based salt reduction intervention study).5,6 Except for Restaurant-based salt reduction intervention study, which targets restaurants, AIS, HIS, and CIS target individuals and share a common primary outcome, that is, the change in salt intake (as measured by 24-hour urinary sodium) from baseline to end of trial. In this study, we pooled the baseline data of these 3 RCTs to assess the most up-to-date sodium and potassium intake of Chinese adults and determine their associations with BP, which will add further evidence to the sodium and potassium intake level and risk factors of BP with the accurate measurement method in large samples.

Methods

The overall design of Action on Salt China6 and the specific design and data collection of the 3 RCTs (CIS,7 HIS,8 and AIS9) have been described previously. There is no pre-prepared analysis plan for the present study. The data that support the findings of this study are available from Prof Puhong Zhang upon reasonable request after the publication of major results of 3 RCTs and in compliance with the pertinent regulations of data management and data sharing in China.

Participants

Study sites of CIS, HIS, and AIS were located in different counties of 6 provinces (Qinghai, Hebei, Heilongjiang, Sichuan, Jiangxi, and Hunan) representing a broad range of geographic locations, economic levels, and dietary habits. In all the 3 RCTs, local residents aged 18 to 75 years without plans of moving in the next 24 months were eligible to be recruited as adult participants. Individuals who were unable or ineligible to provide 24-hour urine samples due to pregnancy, lactation, or other relevant conditions were excluded from the studies. Finally, a total of 2688 participants were enrolled from 48 towns (clusters) of 12 counties in CIS; 1560 participants were included from 60 communities/villages (clusters) of 6 counties in HIS; and in AIS, 1188 adults and 594 children were enrolled from 54 schools (clusters) of 3 counties.

Definition of Variables

The duration of urine collection was required to be 24 hours, with both starting and finish times recorded by the research team. If the duration was not exactly 24 hours, adjusted 24-hour urine volume was calculated as total volume divided by collection time and multiplied by 24. Urine volume (mL) was assumed to be equal to urine weight (g). Twenty-four–hour urinary sodium/potassium (mg/d) excretion was calculated as the concentration (mmol/L) multiplied by molecular weight (sodium, 23 mg/mmol; potassium, 39.1 mg/mmol) and adjusted 24-hour urine volume (L/d). Sodium-to-potassium ratio was calculated as sodium concentration (mmol/L) divided by potassium concentration (mmol/L). Salt intake (g/d) was calculated as sodium excretion (mg/d, estimate of sodium intake) multiplied by 2.54 and divided by 1000. According to the World Health Organization recommendation, if a sodium intake was >2000 mg/d, it was defined as an excess intake of sodium; if a potassium intake was <3510 mg/d, it was defined as an insufficient intake of potassium.10,11 According to previous studies, urine data were defined as incomplete if the 24-hour urine volume was <500 mL or creatinine <4.0 mmol for women or <6.0 mmol for men.12–14 If the urine collections last <20 hours or >28 hours, it will also be excluded.

BP was calculated from the average of the last 2 of the 3 measurements. Hypertension was defined as mean systolic BP (SBP) ≥140 mm Hg or mean diastolic BP (DBP) ≥90 mm Hg or self-reported use of antihypertensive medication in the last 2 weeks. Body mass index (BMI) was calculated as body weight (kg) divided by the square of height (m2). According to BMI cutoffs specific to the Chinese population, individuals were considered underweight if their BMI was <18.5 kg/m2, normal weight if it was between 18.5 and 23.9 kg/m2, overweight if it was between 24 and 27.9 kg/m2, and obese if it was ≥28 kg/m2.15 Participants were considered physically active if they reported engaging in moderate physical activity for at least 30 minutes and at least 3 times a week. Regions were classified as north-rural, north-urban, south-rural, and south-urban according to the geographic and urban-rural division of study sites.

Statistical Analysis

Mean and SD were used to describe the excretion levels of sodium, potassium, and the sodium-to-potassium ratio by age group, sex, and region, for all 3 combined and for each trial separately. The sodium and potassium excretions of the 2 consecutive complete 24-hour urine data of AIS were averaged beforehand. If one day’s urine collection is incomplete, then only the data of the complete day will be used in AIS. Factors related with potassium excretion and sodium-to-potassium ratio were analyzed by multivariate analysis. In view of the possible heterogeneity of subjects across the 3 RCTs and the potential aggregation of participants within a community, we fitted multivariate mixed models to examine the associations of sodium, potassium, and sodium-to-potassium ratio with SBP and DBP by specifying the random effects of RCT and community levels. Model 1 adjusted age group and sex; model 2 adjusted factors in model 1 plus region, education, smoking, drinking, and physical activity; model 3 adjusted all factors in model 2 plus BMI and antihypertensive medication. Coefficients of sodium, potassium, and sodium-to-potassium ratio were corrected by using the repeated measurement of AIS samples (n=1121, complete 24-hour urinary data for both 2 measurement days) as data for reliability to correct the regression dilution bias caused by measurement variability of the corresponding variables in each model.16 Missing values were directly excluded from the multivariate models in consideration of the few numbers. All the models examining sodium controlled for potassium simultaneously and vice versa. The dose unit for sodium and potassium was 1000 mg/d and 1 for sodium-potassium molar ratio in the models. To further understand the relationships at different BP levels, quantile regression was performed to estimate the relationship of SBP and DBP with sodium, potassium, and sodium-to-potassium ratio under various quantiles of SBP and DBP (quantile from 0.1 to 0.9 with 0.1 increment).

Sensitivity analyses were conducted to (1) include all urine collections to estimate levels of sodium, potassium, and their ratio; (2) exclude participants taking antihypertensive medications to examine the relationships of SBP and DBP with sodium, potassium, and their ratio; and (3) analyze the relationships in each RCT independently and then perform random-effect meta-analysis of the 3 RCTs. The interactions between the sodium, potassium, or sodium-to-potassium ratio and all other covariates including age group were tested in the multivariate models with BP as the dependent variable. All analyses were 2 sided, and P<0.05 was considered significant. SAS Enterprise Guide 7.1 was used for analyses except R 3.6.2 was applied for regression dilution bias correction.

Ethical Approval

All trials have been approved by Queen Mary Research Ethics Committee in the United Kingdom (AIS, QMERC2018/13; HIS, QMERC2018/15; CIS, QMERC2018/16) and the Institutional Review Boards of China including Peking University (AIS, IRB00001052-18051), Chinese Center for Disease Control and Prevention (HIS, No. 201801), and National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention (CIS, No. 201807). Written consents have been obtained from all participants according to well-established practices.

Results

Characteristics of Participants

A total of 5454 adult participants were recruited in the 3 RCTs. After excluding 101 incomplete urine samples, the sample size was 5353 (98.1% of the recruited participants; Figure 1).

Figure 1.

Figure 1. Flowchart of participants. AppSalt-based salt reduction in primary school children and their families (AIS), Community-based comprehensive salt reduction intervention study (CIS), and Home cook salt reduction intervention study (HIS) are the 3 independent salt reduction trials in Action on Salt China program.

Of these 5353 participants, 2094 (39.1%) and 1420 (26.5%) came from north-rural and south-rural regions, respectively, and 396 (7.4%) and 1443 (27.0%) from north-urban and south-urban regions, respectively. Forty-nine percent of them were men, and the mean age was 49.3 (SD=12.8) years. Average SBP and DBP were 125.4 (SD=19.2) and 79.0 (SD=11.2) mm Hg, respectively, and 1767 (33.0%) of the participants were defined as hypertensive. Other detailed characteristics of overall and RCT-specific participants are shown in Table 1.

Table 1. Characteristics of Participants in the Study

VariableOverall (n=5353)AIS (n=1180)CIS (n=2463)HIS (n=1530)
Region, n (%)
 North-rural2094 (39.1)01316 (49.8)778 (50.8)
 North-urban396 (7.4)396 (33.6)00
 South-rural1420 (26.5)0996 (37.7)424 (27.7)
 South-urban1443 (27.0)784 (66.4)331 (12.5)328 (21.4)
Sex: male, n (%)2621 (49.0)549 (46.5)1314 (49.7)758 (49.5)
Age, y; n (%)
 18–442063 (38.5)700 (59.3)1141 (43.2)222 (14.5)
 45–591877 (35.1)202 (17.1)967 (36.6)708 (46.3)
 ≥601413 (26.4)278 (23.6)535 (20.2)600 (39.2)
Education years, n (%)
 0–62136 (39.9)223 (18.9)1117 (42.3)796 (52.0)
 7–91801 (33.6)300 (25.4)1018 (38.5)483 (31.6)
 ≥101416 (26.5)657 (55.7)508 (19.2)251 (16.4)
Smoking: yes, n (%)*1450 (27.1)278 (23.6)791 (29.9)381 (24.9)
Drinking, n (%)*
 No3179 (59.4)661 (56.0)1562 (59.1)956 (62.5)
  Occasionally1655 (30.9)424 (35.9)828 (31.3)403 (26.3)
 Frequently518 (9.7)95 (8.1)252 (9.5)171 (11.2)
Physical activity: active, n (%)1897 (35.4)379 (32.1)1065 (40.3)453 (29.6)
BMI, n (%)
 Underweight100 (1.9)26 (2.2)46 (1.7)28 (1.8)
 Normal weight2140 (40.0)461 (39.1)1110 (42.0)569 (37.2)
 Overweight2091 (39.1)454 (38.5)1017 (38.5)620 (40.5)
 Obesity1019 (19.0)239 (20.3)468 (17.7)312 (20.4)
 Missing3 (0.0)0 (0.0)2 (0.1)1 (0.1)
Hypertension: yes, n (%)1767 (33.0)258 (21.9)858 (32.5)651 (42.6)
Antihypertensive medication: yes, n (%)869 (16.2)113 (9.6)398 (15.1)358 (23.4)
SBP, mm Hg; mean (SD)125.4 (19.2)118.6 (17.2)126.1 (19.3)129.3 (19.1)
DBP, mm Hg; mean (SD)79.0 (11.2)76.9 (10.4)79.3 (11.6)79.9 (11.0)
Urinary collection time, h; mean (SD)23.9 (0.4)24.0 (0.6)23.9 (0.4)23.8 (0.4)
Urinary volume, mL/24 h; mean (SD)1602.8 (661.3)1562.1 (639.1)1616.1 (647.6)1611.1 (699.6)
Urinary creatinine, mmol/24 h; mean (SD)10.6 (3.3)10.8 (3.3)10.7 (3.3)10.3 (3.1)

AppSalt-based salt reduction in primary school children and their families (AIS), Community-based comprehensive salt reduction intervention study (CIS), and Home cook salt reduction intervention study (HIS) are the 3 independent salt reduction trials in Action on Salt China program. BMI indicates body mass index; DBP, diastolic blood pressure; and SBP, systolic blood pressure.

* One participant had missing values in smoking and drinking.

† Two participants (CIS and HIS, 1 each) had missing values in blood pressure measurement.

Twenty-Four–Hour Urinary Sodium, Potassium Excretion, and Sodium-to-Potassium Ratio Levels in Participants

Overall, average sodium excretion was 4318.1 (SD=1814.1) mg/d, which corresponds to an average salt intake of 11.0 (SD=4.6) g/d, with men consuming 11.6 (SD=4.8) g/d and women 10.4 (SD=4.3) g/d. Average potassium excretion was 1573.7 (SD=627.1) mg/d (1555.3±633.7 mg/d in men and 1591.4±620.2 mg/d in women). Mean sodium-to-potassium ratio of participants was 5.0 (SD=2.1; 5.3±2.2 in men and 4.7±2.0 in women; Table 2). RCT-specific results are shown in Tables S1 through S3 in the Data Supplement. World Health Organization upper limit of 2000 mg/d was exceeded by 93.5% of the participants, and the potassium excretion of nearly all the participants (99.0%) was below the World Health Organization lower limit of 3510 mg/d (Table S4). Over half of the participants (2744; 51.3%) had sodium intake >4000 mg/d (≈10 g/d salt) and 2772 (51.8%) had potassium intake <1500 mg/d.

Table 2. Twenty-Four–Hour Urinary Sodium Excretion, Potassium Excretion, and Sodium-to-Potassium Ratio of Participants

VariableSodium Excretion, mg/d; Mean (SD)Potassium Excretion, mg/d; Mean (SD)Sodium-to-Potassium Ratio, mol/mol; mean (SD)
OverallMenWomenOverallMenWomenOverallMenWomen
n=5353n=2621n=2732n=5353n=2621n=2732n=5353n=2621n=2732
Age, y
 18–444350.7 (1783.8)4549.4 (1880.4)4176.7 (1676.4)1549.2 (609.2)1508.1 (609.8)1585.3 (606.6)5.2 (2.3)5.5 (2.3)4.8 (2.2)
 45–594429.8 (1847.2)4697.2 (1956.1)4177.5 (1701.2)1623.2 (657.6)1585.3 (660.5)1659.0 (653.1)5.0 (2.0)5.4 (2.1)4.6 (1.9)
 ≥604122.1 (1799.2)4381.9 (1853.4)3830.7 (1690.8)1543.7 (607.5)1579.5 (627.9)1503.5 (581.6)4.9 (2.1)5.1 (2.2)4.6 (1.9)
Region
 North-rural4877.0 (2001.6)5238.5 (2084.8)4530.3 (1854.3)1684.1 (693.1)1682.8 (694.6)1685.4 (692.0)5.3 (2.3)5.7 (2.4)4.9 (2.1)
 North-urban4272.1 (1729.2)4805.5 (1839.6)3799.6 (1475.9)1670.6 (653.5)1682.1 (727.9)1660.5 (581.3)4.7 (2.1)5.3 (2.2)4.2 (1.8)
 South-rural4138.0 (1668.0)4215.0 (1716.8)4056.5 (1613.7)1467.7 (551.4)1431.4 (559.8)1506.1 (540.1)5.1 (2.1)5.3 (2.1)4.9 (2.0)
 South-urban3696.8 (1402.8)3813.6 (1390.7)3592.7 (1406.2)1491.2 (554.8)1461.3 (535.7)1537.9 (570.3)4.5 (1.8)4.8 (1.8)4.3 (1.8)
Overall4318.1 (1814.1)4553.0 (1902.9)4092.7 (1694.6)1573.7 (627.1)1555.3 (633.7)1591.4 (620.2)5.0 (2.1)5.3 (2.2)4.7 (2.0)

Multivariate analysis of factors related with sodium, potassium excretion, and sodium-to-potassium ratio showed that sodium excretion was higher in younger age groups than in elders aged over 60 years. Men had a higher sodium excretion than women (P<0.001). Northern and rural regions had relatively higher sodium excretion than southern and urban regions (P<0.001). Factors related with potassium excretion and sodium-to-potassium ratio were similar in general and are detailed in Table S5.

Association Between Sodium, Potassium Excretion, and Sodium-to-Potassium Ratio With SBP and DBP

Table 3 presents the association between sodium, potassium, and their ratio with SBP and DBP. After adjusting for all other sociodemographic and individual behavior factors and correcting for regression dilution bias of sodium and potassium measurement, every 1000-mg/d increase of sodium excretion was associated with a 1.32-mm Hg ([95% CI, 0.92–1.81] P<0.0001) increase in SBP and a 0.34-mm Hg ([95% CI, 0.09–0.60] P≤0.0001) increase in DBP; each 1000-mg/d increase of potassium excretion was associated with a 3.19-mm Hg ([95% CI, −4.38 to −2.20] P<0.0001) decrease in SBP and a 1.56-mm Hg ([95% CI, −2.29 to −0.90] P<0.0001) decrease in DBP; and each unit increase in sodium-to-potassium molar ratio was associated with a 1.21-mm Hg ([95% CI, 0.91–1.60] P<0.0001) increase in SBP and a 0.44-mm Hg ([95% CI, 0.24–0.64] P<0.0001) increase in DBP. Random-effects results showed that the variance at RCT-level accounted for <0.3% of the whole variance and was insignificant. In contrast, the contribution of community level variance was 3% to 4% and significant (Table S6).

Table 3. Association Between Sodium, Potassium, and Sodium-Potassium Ratio and Blood Pressure (n=5353)

ModelsSBPDBP
Uncorrected*CorrectedUncorrected*Corrected
β95% CIβ95% CIβ95% CIβ95% CI
Sodium
 Model 11.561.27 to 1.852.261.74 to 2.690.640.46 to 0.820.930.62 to 1.16
 Model 21.431.14 to 1.722.071.64 to 2.540.570.38 to 0.750.810.55 to 1.10
 Model 30.940.66 to 1.221.320.92 to 1.810.250.08 to 0.430.340.09 to 0.60
Potassium
 Model 1−2.35−3.16 to −1.55−2.99−4.26 to −1.88−1.08−1.59 to −0.57−1.38−2.10 to −0.69
 Model 2−2.38−3.19 to −1.57−3.04−4.19 to −1.90−1.12−1.63 to −0.60−1.43−2.18 to −0.77
 Model 3−2.46−3.22 to −1.70−3.19−4.38 to −2.20−1.19−1.68 to −0.70−1.56−2.29 to −0.90
Sodium-to-potassium ratio
 Model 11.140.92 to 1.351.551.21 to 1.900.470.33 to 0.610.640.43 to 0.85
 Model 21.080.87 to 1.301.481.13 to 1.850.440.30 to 0.580.600.37 to 0.80
 Model 30.890.69 to 1.091.210.91 to 1.600.320.19 to 0.450.440.24 to 0.64

Mixed models were performed with RCT and community nested within RCT as random factors. Model 1 adjusted age group and sex; model 2 adjusted age group, sex, region, education, smoking, drinking, and physical activity; model 3 adjusted all factors of model 2 plus BMI and antihypertensive medication. All the models examining sodium controlled for potassium simultaneously and vice versa. Sodium and potassium are 1 g/d dose unit; sodium-potassium molar ratio is 1 dose unit. P<0.0001 in all β-coefficients. BMI indicates body mass index; DBP, diastolic blood pressure; RCT, randomized controlled trial; and SBP, systolic blood pressure.

* Direct results of mixed models.

† Results of regression dilution bias correction.

Figure 2 shows the coefficients of sodium, potassium, and sodium-to-potassium ratio in quantiles of SBP and DBP from 0.1 to 0.9 with 0.1 increment. The SBP and DBP distribution showed 0.7 quantiles of SBP and DBP were 134 and 84 mm Hg, 0.8 quantiles of SBP and DBP were 141 and 88 mm Hg, respectively (Table S7). In general, the coefficients of sodium with SBP and DBP rise with the increase of quantiles except for fluctuation of 0.9 quantile in SBP and 0.8 quantile in DBP. The coefficients with SBP were significant in all the quantiles and those with DBP were significant after 0.4 quantile. The slopes of potassium with SBP and DBP decline with the increase of quantiles. Sodium-to-potassium ratio presents similar increasing trend with that of sodium. The values of coefficients and the 95% CI can be found in Tables S8 through S10.

Figure 2.

Figure 2. Coefficients of sodium (Na), potassium (K), and Na-to-K ratio in different quantiles of systolic blood pressure (SBP) and diastolic blood pressure (DBP). The blue shadow indicated the 95% CI of coefficients.

Sensitivity Analysis

In the sensitivity analysis with all participants included, average sodium and potassium excretion would be, respectively, 38.60 and 16.60 mg/d lower than the current 5353 participants dropping those unqualified data. The mean sodium-potassium ratio remained 5.0 in the all-participant sample (Table S11). The multivariate results remained robust when the analysis was restricted in 4484 participants without taking hypertensive medicines during the last 2 weeks (Table S12). Furthermore, the multivariate results are consistent with random-effect meta-analysis based on the independent coefficients of 3 RCTs (Tables S13 and S14). There was no significant interaction between the sodium, potassium, or sodium-to-potassium ratio and most other covariates including age group in multivariate models with BP except the interaction of sodium excretion with BMI in the DBP model, which suggested the association of sodium and DBP in overweight adults appeared 0.40 mm Hg ([95% CI, 0.02–0.79] P=0.041) and 0.46 mm Hg ([95% CI, 0.04–0.89] P=0.033) higher than those in normal-weight and obese adults, respectively.

Discussion

Principal Findings

This study is the largest to date in measuring urinary sodium and potassium excretion using the most accurate method of 24-hour urine collection in China. The results showed that the average sodium excretion was over 4300 mg/d (equivalent to 11.0 g/d salt), whereas the average potassium excretion was under 1600 mg/d. That is, over half participants had sodium intake >4000 mg/d (≈10 g/d salt) and potassium intake <1500 mg/d. As a result, the mean sodium-to-potassium ratio was as high as 5. A recent meta-analysis reviewing the published 24-hour urinary sodium and potassium data in China over the past 4 decades showed that salt intake was 11.06 g/d and potassium intake 1.42 g/d.17 The China Health and Nutrition Survey, using series of 3 consecutive 24-hour dietary recalls, reported that in 2015, average salt intake was estimated to be 10.3 g/d and average potassium intake 1.5 g/d, resulting in a sodium-to-potassium molar ratio of ≈4.6.18 The consistency of the sodium and potassium intake presents the sustained high sodium and low potassium situation in China.

Our results showed that a higher sodium excretion, lower potassium excretion, and a greater sodium-to-potassium ratio were associated with a higher SBP and DBP, which is consistent with numerous previous studies.19–21 The difference, however, was the magnitude of the association with BP. This study found an increase of 1.32 mm Hg in SBP and 0.34 mm Hg in DBP per gram increase in sodium excretion, a decrease of 3.19/1.56 mm Hg in SBP/DBP per gram increase in potassium excretion, and an increase of 1.21/0.44 mm Hg in SBP/DBP per unit increase of sodium-to-potassium molar ratio. In National Health and Nutrition Examination Survey 2014 study, the change was 4.58/2.25 mm Hg for 1-g sodium increase, −3.72 mm Hg in SBP for 1-g potassium increase, and 1.72 mm Hg in SBP for 0.5-unit increase of sodium-to-potassium molar ratio.19 In the Prospective Urban Rural Epidemiology China study, the change was 1.70/0.49 mm Hg, −1.10/−0.91 mm Hg, and 4.33/1.54 mm Hg for 1-g increase of sodium, 1-g increase of potassium, and 1 unit of sodium-to-potassium ratio in gram, respectively.20 In another study in China, the dosage was correspondingly 2.36/1.60, −10.7/−6.86, and 0.97/0.65 mm Hg (per 1 unit of molar ratio).21 Our study appeared to find relatively lower dose-response rate of sodium than other studies. However, due to the discrepancy of the participants and investigation methods in different studies, the difference would be reasonable. Evidence has shown that the relationship between sodium, potassium, and their ratio with BP was associated with the hypertension status, sodium and potassium intake level, BMI, ethnicity, and other anthropometric characteristics.1,2,22 In addition, the correction method in the analysis may alter the slopes with SBP and DBP to different degrees. In the International Cooperative Study on Salt, Other Factors, and Blood Pressure study, the coefficients could be 44% to 50% larger after correction, and the values varied with measurement of reliability.23 Our corrected coefficients were 30% to 40% larger than uncorrected values.

Although the positive relationship between sodium intake and BP has been supported by extensive evidence, some studies reported the relationship was significant only in hypertensive patients and there might be negative effect of large sodium reduction in short duration.24,25 Our results showed increasing trend of positive relationships among the quantiles of SBP and DBP although there was small fluctuation in the upper quantiles of SBP and the relationships of DBP became significant after 0.4 quantiles. The overall result was consistent with other studies showing the relationships between sodium and BP are larger in hypertensive populations than in normotensive ones.26 It is worth noting that compared with sodium, this study found robustly stronger associations between potassium and BP, which implies the importance of potassium elevation in the prevention of hypertension. In addition, although sodium-to-potassium ratio was similarly associated with BP as sodium, sodium-to-potassium ratio was regarded in some studies as a superior metric than sodium or potassium in relation to BP, with higher sodium-to-potassium ratio reflecting both higher sodium intake and lower potassium intake and its independent measurement of the urine collection method.27 The current high sodium-to-potassium ratio in the Chinese population highlighted the urgency of comprehensive strategies in salt reduction and potassium elevation across the country to prevent and control hypertension that is highly prevalent in China with 1 in 3 or 4 adults experiencing high BP.28 The present study will also provide reference on diet and BP research for other developing countries with similar eating habits as China.

Strengths and Limitations of the Study

Our study has 3 main strengths. First, we collected data from an unprecedented over 5300 participants across 6 provinces of China. Although this study is not a national survey with a universal sampling frame, with a large sample size covering various typical regions of China, it is reasonable to infer that the sodium and potassium estimates in this study may reflect the national level closely. Second, we used the most accurate assessment method of 24-hour urine collection, which provided robust estimate of sodium and potassium intake and minimized bias in determining associations of sodium and potassium with BP.29 Third, we collected the urine data with high quality under the support of a bespoke electronic data capture system. Over urine collection or under urine collection rarely occurred in this study, and incomplete urine collection accounted for only 1.9% of all participants, which is lower than the previous 2 studies (with 4.7%21 and 25.0%19 incomplete rate, respectively) irrespective of various completeness criteria of 24-hour urine collection.

The limitations of this study are as follows: first, most samples collected only one 24-hour urine sample, which does not reflect the individual day-to-day variation in sodium and potassium excretion. Nonetheless, the 1-day collection is still capable of estimating the group level as shown in other studies.3,4 We also performed correction using the reliability data from repeat measurement of 1 RCT in multivariable regression analysis. Second, it has previously been reported that the urinary excretion of sodium and potassium represented 80% to 95% and 63% to 77% of their dietary intakes, respectively.3031 However, in view of the large uncertainty ranges, we decided not to use any conversion factor to estimate sodium and potassium intake from their urinary excretions, which could result in underestimations. Third, as the baseline data used in this present study are of cross-sectional nature, no causal relationship can be inferred.

Perspectives

Using the most accurate method of 24-hour urine collection in a large sample of China, this study has shown that the current sodium intake in Chinese adults remains high and potassium intake is low. Sodium and sodium-to-potassium ratio were positively associated with BP and potassium inversely with BP. Although China has promoted salt reduction over the past decade, for instance, the government-led China Healthy Lifestyle for All campaign since 200732 and a pilot salt reduction project in Shandong province in 2011 to 2016.33 The persistently high salt intake and low potassium levels across the country indicated that the efforts of salt reduction and potassium elevation in China were insufficient. Urgent action is needed to reduce sodium and especially increase potassium intake to prevent high BP in China, which would save thousands of deaths from cardiovascular diseases.34 The Healthy China 2030 plan has been facilitating the public health promotion and research in salt reduction toward salt intake goal of <5 g/d. Together with the promotion of salt reduction, further efforts should also be put into the issue of low potassium intake, which was paid less attention than sodium reduction. Healthy diets including the consumption of potassium-rich foods such as fruits and vegetables should be more strongly recommended, and potassium-enriched salt substitutes should be recommended to the general population.

Acknowledgments

We thank the staff from local Centers for Disease Control and Prevention, community health service centers, schools, and participants for contributing to the data collection. Coauthors contributed to the following: P. Zhang, Xinhua Li, G.A. MacGregor, and F.J. He conceived the overall design of Action on Salt China (ASC). Xinhua Li and P. Zhang were responsible for the overall survey governance and implementation of ASC. J. Wu, J. Xu, and M. Liu led the survey of CIS. J. Ma and X. Zhang organized the survey of HIS. P. Zhang, Y. Li, R. Luo, Y. Sun, and F.J. He contributed to the survey of AIS. Y. Li drafted the first version of the manuscript. X. Li and M. Tan confirmed the statistical analysis. All the authors contributed to the review and approved the final manuscript.

Footnotes

*These authors contributed equally to this work.

The Data Supplement is available with this article at https://www.ahajournals.org/doi/suppl/10.1161/HYPERTENSIONAHA.120.15238.

Correspondence to Xinhua Li, Chinese Center for Disease Control and Prevention, No. 155 Changbai Rd, Changping District, Beijing 102206, China. Email

References

  • 1. Mente A, O’Donnell MJ, Rangarajan S, McQueen MJ, Poirier P, Wielgosz A, Morrison H, Li W, Wang X, Di C, et al; PURE Investigators. Association of urinary sodium and potassium excretion with blood pressure.N Engl J Med. 2014; 371:601–611. doi: 10.1056/NEJMoa1311989CrossrefMedlineGoogle Scholar
  • 2. Binia A, Jaeger J, Hu Y, Singh A, Zimmermann D. Daily potassium intake and sodium-to-potassium ratio in the reduction of blood pressure: a meta-analysis of randomized controlled trials.J Hypertens. 2015; 33:1509–1520. doi: 10.1097/HJH.0000000000000611CrossrefMedlineGoogle Scholar
  • 3. Campbell NRC, He FJ, Tan M, Cappuccio FP, Neal B, Woodward M, Cogswell ME, McLean R, Arcand J, MacGregor G, et al. The International consortium for quality research on dietary sodium/salt (TRUE) position statement on the use of 24-hour, spot, and short duration (<24 hours) timed urine collections to assess dietary sodium intake.J Clin Hypertens (Greenwich). 2019; 21:700–709. doi: 10.1111/jch.13551CrossrefMedlineGoogle Scholar
  • 4. McLean RM. Measuring population sodium intake: a review of methods.Nutrients. 2014; 6:4651–4662. doi: 10.3390/nu6114651CrossrefMedlineGoogle Scholar
  • 5. He FJ, Zhang P, Li Y, MacGregor GA. Action on salt China.Lancet. 2018; 392:7–9. doi: 10.1016/S0140-6736(18)31138-3CrossrefMedlineGoogle Scholar
  • 6. Zhang P, He FJ, Li Y, Li C, Wu J, Ma J, Zhang B, Wang H, Li Y, Han J, et al. Reducing salt intake in China with “Action on Salt China” (ASC): protocol for campaigns and randomized controlled trials.JMIR Res Protoc. 2020; 9:e15933. doi: 10.2196/15933CrossrefMedlineGoogle Scholar
  • 7. Xu J, Tang B, Liu M, Bai Y, Yan W, Zhou X, Xu Z, He J, Jin D, Sun J, et al. A town level comprehensive intervention study to reduce salt intake in China: protocol for a cluster randomised controlled trial.BMJ Open. 2020; 10:e032976. doi: 10.1136/bmjopen-2019-032976CrossrefMedlineGoogle Scholar
  • 8. Zhang X, Hu X, Ma J, Zhang P, Li Y, Luo R, He FJ, MacGregor GA, Wang J, Yin Z. Cluster randomised controlled trial of home cook intervention to reduce salt intake in China: a protocol study.BMJ Open. 2020; 10:e033842. doi: 10.1136/bmjopen-2019-033842CrossrefMedlineGoogle Scholar
  • 9. He FJ, Zhang P, Luo R, Li Y, Chen F, Zhao Y, Zhao W, Li D, Chen H, Wu T, et al. An Application-based programme to reinforce and maintain lower salt intake (AppSalt) in school children and their families in China.BMJ Open. 2019; 9:e027793. doi: 10.1136/bmjopen-2018-027793CrossrefMedlineGoogle Scholar
  • 10. World Health Organization. Guideline: Sodium Intake for Adults and ChildrenGenevaWorld Health Organization (WHO);2012Google Scholar
  • 11. World Health Organization. Guideline: Potassium Intake for Adults and ChildrenGenevaWorld Health Organization (WHO);2012Google Scholar
  • 12. He FJ, Wu Y, Feng XX, Ma J, Ma Y, Wang H, Zhang J, Yuan J, Lin CP, Nowson C, et al. School based education programme to reduce salt intake in children and their families (school-EduSalt): cluster randomised controlled trial.BMJ. 2015; 350:h770. doi: 10.1136/bmj.h770CrossrefMedlineGoogle Scholar
  • 13. Harris RM, Rose AMC, Hambleton IR, Howitt C, Forouhi NG, Hennis AJM, Samuels TA, Unwin N. Sodium and potassium excretion in an adult Caribbean population of African descent with a high burden of cardiovascular disease.BMC Public Health. 2018; 18:998. doi: 10.1186/s12889-018-5694-0CrossrefMedlineGoogle Scholar
  • 14. Stolarz-Skrzypek K, Kuznetsova T, Thijs L, Tikhonoff V, Seidlerová J, Richart T, Jin Y, Olszanecka A, Malyutina S, Casiglia E, et al; European Project on Genes in Hypertension (EPOGH) Investigators. Fatal and nonfatal outcomes, incidence of hypertension, and blood pressure changes in relation to urinary sodium excretion.JAMA. 2011; 305:1777–1785. doi: 10.1001/jama.2011.574CrossrefMedlineGoogle Scholar
  • 15. Chen C, Lu FCDepartment of Disease Control Ministry of Health, PR China. The guidelines for prevention and control of overweight and obesity in Chinese adults.Biomed Environ Sci. 2004; 17suppl1–36MedlineGoogle Scholar
  • 16. Berglund L. Regression dilution bias: tools for correction methods and sample size calculation.Ups J Med Sci. 2012; 117:279–283. doi: 10.3109/03009734.2012.668143CrossrefMedlineGoogle Scholar
  • 17. Tan M, He FJ, Wang C, MacGregor GA. Twenty-four-hour urinary sodium and potassium excretion in China: a systematic review and meta-analysis.J Am Heart Assoc. 2019; 8:e012923. doi: 10.1161/JAHA.119.012923LinkGoogle Scholar
  • 18. Du S, Wang H, Zhang B, Popkin BM. Dietary potassium intake remains low and sodium intake remains high, and most sodium is derived from home food preparation for Chinese adults, 1991-2015 trends.J Nutr. 2020; 150:1230–1239. doi: 10.1093/jn/nxz332CrossrefMedlineGoogle Scholar
  • 19. Jackson SL, Cogswell ME, Zhao L, Terry AL, Wang CY, Wright J, Coleman King SM, Bowman B, Chen TC, et al. Association between urinary sodium and potassium excretion and blood pressure among adults in the United States: National Health and Nutrition Examination Survey, 2014.Circulation. 2018; 137:237–246. doi: 10.1161/CIRCULATIONAHA.117.029193LinkGoogle Scholar
  • 20. Yin L, Deng G, Mente A, Sun Y, Liu X, Zhang X, Wang X, Wang Y, Bo J, Chen H, et al. Association patterns of urinary sodium, potassium, and their ratio with blood pressure across various levels of salt-diet regions in China.Sci Rep. 2018; 8:6727. doi: 10.1038/s41598-018-25097-1CrossrefMedlineGoogle Scholar
  • 21. Xu J, Chen X, Ge Z, Liang H, Yan L, Guo X, Zhang Y, Wang L, Ma J. Associations of usual 24-hour sodium and potassium intakes with blood pressure and risk of hypertension among adults in China’s shandong and Jiangsu provinces.Kidney Blood Press Res. 2017; 42:188–200. doi: 10.1159/000475486CrossrefMedlineGoogle Scholar
  • 22. Stamler J, Chan Q, Daviglus ML, Dyer AR, Van Horn L, Garside DB, Miura K, Wu Y, Ueshima H, Zhao L, et al; INTERMAP Research Group. Relation of dietary sodium (salt) to blood pressure and its possible modulation by other dietary factors: the INTERMAP study.Hypertension. 2018; 71:631–637. doi: 10.1161/HYPERTENSIONAHA.117.09928LinkGoogle Scholar
  • 23. Dyer AR, Elliott P, Shipley M. Urinary electrolyte excretion in 24 hours and blood pressure in the INTERSALT Study. II. Estimates of electrolyte-blood pressure associations corrected for regression dilution bias. The INTERSALT cooperative research group.Am J Epidemiol. 1994; 139:940–951. doi: 10.1093/oxfordjournals.aje.a117100CrossrefMedlineGoogle Scholar
  • 24. Graudal N, Hubeck-Graudal T, Jürgens G, Taylor RS. Dose-response relation between dietary sodium and blood pressure: a meta-regression analysis of 133 randomized controlled trials.Am J Clin Nutr. 2019; 109:1273–1278. doi: 10.1093/ajcn/nqy384CrossrefMedlineGoogle Scholar
  • 25. Graudal NA, Hubeck-Graudal T, Jurgens G. Effects of low sodium diet versus high sodium diet on blood pressure, renin, aldosterone, catecholamines, cholesterol, and triglyceride.Cochrane Database Syst Rev. 2017; 4:CD004022. doi: 10.1002/14651858.CD004022.pub4CrossrefMedlineGoogle Scholar
  • 26. Huang L, Trieu K, Yoshimura S, Neal B, Woodward M, Campbell NRC, Li Q, Lackland DT, Leung AA, Anderson CAM, et al. Effect of dose and duration of reduction in dietary sodium on blood pressure levels: systematic review and meta-analysis of randomised trials.BMJ. 2020; 368:m315. doi: 10.1136/bmj.m315CrossrefMedlineGoogle Scholar
  • 27. Iwahori T, Miura K, Ueshima H. Time to consider use of the sodium-to-potassium ratio for practical sodium reduction and potassium increase.Nutrients. 2017; 9:700. doi: 10.3390/nu9070700CrossrefMedlineGoogle Scholar
  • 28. Wang Z, Chen Z, Zhang L, Wang X, Hao G, Zhang Z, Shao L, Tian Y, Dong Y, Zheng C, et al. China hypertension survey I. Status of hypertension in China: results from the China hypertension survey, 2012-2015.Circulation. 2018; 137:2344–2356. doi: 10.1161/CIRCULATIONAHA.117.032380LinkGoogle Scholar
  • 29. Espeland MA, Kumanyika S, Wilson AC, Reboussin DM, Easter L, Self M, Robertson J, Brown WM, McFarlane MTONE Cooperative Research Group. Statistical issues in analyzing 24-hour dietary recall and 24-hour urine collection data for sodium and potassium intakes.Am J Epidemiol. 2001; 153:996–1006. doi: 10.1093/aje/153.10.996CrossrefMedlineGoogle Scholar
  • 30. Iwahori T, Miura K, Ueshima H, Tanaka-Mizuno S, Chan Q, Arima H, Dyer AR, Elliott P, Stamler JINTERSALT Research Group. Urinary sodium-to-potassium ratio and intake of sodium and potassium among men and women from multiethnic general populations: the INTERSALT Study.Hypertens Res. 2019; 42:1590–1598. doi: 10.1038/s41440-019-0263-1CrossrefMedlineGoogle Scholar
  • 31. Lucko AM, Doktorchik C, Woodward M, Cogswell M, Neal B, Rabi D, Anderson C, He FJ, MacGregor GA, L’Abbe M, et al; TRUE Consortium. Percentage of ingested sodium excreted in 24-hour urine collections: a systematic review and meta-analysis.J Clin Hypertens (Greenwich). 2018; 20:1220–1229. doi: 10.1111/jch.13353CrossrefMedlineGoogle Scholar
  • 32. Shao S, Hua Y, Yang Y, Liu X, Fan J, Zhang A, Xiang J, Li M, Yan LL. Salt reduction in China: a state-of-the-art review.Risk Manag Healthc Policy. 2017; 10:17–28. doi: 10.2147/RMHP.S75918CrossrefMedlineGoogle Scholar
  • 33. Xu A, Ma J, Guo X, Wang L, Wu J, Zhang J, Bai Y, Xu J, Lu Z, Xu Z, et al. Association of a province-wide intervention with salt intake and hypertension in shandong province, China, 2011-2016.JAMA Intern Med. 2020; 180:877–886. doi: 10.1001/jamainternmed.2020.0904CrossrefMedlineGoogle Scholar
  • 34. Zhou M, Wang H, Zeng X, Yin P, Zhu J, Chen W, Li X, Wang L, Wang L, Liu Y, et al. Mortality, morbidity, and risk factors in China and its provinces, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017.Lancet. 2019; 394:1145–58. doi: 10.1016/S0140-6736(19)30427-1CrossrefMedlineGoogle Scholar

Novelty and Significance

What Is New?

  • This study assessed up-to-date level of sodium and potassium intake and their associations with blood pressure using the most accurate method of 24-hour urine collection in a large sample of China collected in 2018.

  • The average sodium, potassium excretion, and sodium-to-potassium molar ratio of Chinese adults were 4318.1±1814.1 mg/d (equivalent to 11.0±4.6 g/d of salt), 1573.7±627.1 mg/d, and 5.0±2.1, respectively.

What Is Relevant?

  • Sodium and sodium-to-potassium ratio were positively associated with blood pressure and potassium inversely with blood pressure.

  • The relationships between sodium and blood pressure mostly increase with the rise of blood pressure quantiles. Potassium shows the opposite trend. Sodium-to-potassium ratio presents similar increasing trend with that of sodium.

Summary

  • The current sodium intake in Chinese adults remains high and potassium intake is low.

  • Urgent action is needed to reduce sodium and increase potassium intake simultaneously to prevent high blood pressure in China.

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.