Blood Pressure in Relation to Interactions Between Sodium Dietary Intake and Renal HandlingNovelty and Significance
Jump to

Abstract
There is abundant evidence that sodium intake is related to blood pressure. However, the relationship varies between individuals and is probably determined by renal sodium handling. We investigated clinic and ambulatory blood pressure in relation to interactions between sodium dietary intake and renal handling, as assessed by 24-hour urinary sodium excretion and endogenous lithium clearance, respectively. We calculated fractional excretion of lithium and fractional distal reabsorption rate of sodium, as markers of proximal and distal sodium handling, respectively. The 766 subjects included 379 men and 478 ambulatory hypertensive patients. They were never treated (n=697) or did not take antihypertensive medication for ≥2 weeks (n=69). In adjusted analyses, none of the associations of urinary sodium excretion, fractional excretion of lithium, and fractional distal reabsorption rate of sodium with clinic or ambulatory blood pressure were statistically significant (P≥0.09). However, there was significant (P=0.01) interaction between urinary sodium excretion and fractional excretion of lithium in relation to nighttime diastolic blood pressure. In tertile 3 but not tertiles 1 and 2 of fractional excretion of lithium, nighttime diastolic pressure was positively associated with urinary sodium excretion (P=0.03). However, nighttime diastolic pressure was higher in tertile 1 than tertile 3 of fractional excretion of lithium (+2.0 mm Hg; P=0.01), especially in the bottom tertile of urinary sodium excretion (+4.9 mm Hg; P<0.001). Similar trends were observed for nighttime systolic pressure and clinic and 24-hour diastolic pressure. In conclusion, sodium dietary intake and proximal tubular handling interact to be associated with blood pressure.
Introduction
Excessive sodium in the human body, as a consequence of either increased dietary intake or decreased urinary excretion, is a well-established risk factor of hypertension.1,2 In between-population cross-sectional studies, populations with higher dietary sodium intake had higher average levels of blood pressure and higher prevalence of hypertension.3,4 In within-population prospective studies, participants with higher dietary sodium intake had higher risk of incident hypertension5 and cardiovascular events.6–8 In spite of strong evidence on the detrimental effect of high dietary intake of sodium, a number of questions remain unresolved.
One of the major questions is the interindividual variability in the blood pressure response to dietary sodium intake. For instance, even within a population of a similar modern lifestyle, people may have quite different levels of blood pressure and different risks of hypertension. Among the complex mechanisms, renal sodium handling must play a major role in the determination of this interindividual variability.9,10 We hypothesize that sodium dietary intake and renal handling interact to influence blood pressure and believe that illustration of this relationship would be clinically relevant for the prevention and management of hypertension by sodium restriction.
However, renal sodium handling has been difficult to evaluate noninvasively. By measuring endogenous lithium clearance, sodium reabsorption in the proximal and distal tubules can be estimated.11 Lithium ions are freely filtered through the glomerulus and reabsorbed in the proximal tubule in the same proportion of sodium and water. There could be some reabsorption of lithium in the loop of Henle on some extreme conditions.12 However, there is almost no reabsorption of lithium in the distal tubule.13 Lithium clearance and the difference between lithium and sodium clearances are, therefore, indicators of sodium reabsorption in the proximal and distal tubules, respectively. High lithium clearance or a small difference between lithium and sodium clearances indicates a better ability to excrete sodium from the corresponding tubule, but does not depend on pressure natriuresis.
We recently measured serum and 24-hour urinary concentrations of sodium and lithium in a group of patients who were referred for clinic and 24-hour ambulatory blood pressure monitoring and did not take any medication of blood pressure–lowering action. In the present analysis, we investigated clinic and ambulatory blood pressure in relation to interactions between sodium dietary intake and renal handling, as assessed by 24-hour urinary sodium excretion and endogenous lithium clearance–derived indexes, respectively.
Methods
Study Population
From December 2008 to December 2011, we invited 1758 consecutive patients who were not on antihypertensive medication for ≥2 weeks and were referred for ambulatory blood pressure monitoring to the specialized outpatient clinic of hypertension, Ruijin Hospital, Shanghai, China. Of those invited, 971 provided written informed consent (participation rate 55.2%). The Ethics Committee of Ruijin Hospital, Shanghai Jiaotong University School of Medicine approved the study protocol.14
We excluded from the present analysis 205 subjects, because of no (n=95) or inappropriate 24-hour urine collection (n=67),15 or no (n=31) or invalid 24-hour ambulatory blood pressure recording (n=7), or because their serum or urinary lithium concentration was ≥1.0 or 20 μmol/L, respectively (n=5), which was suggestive of external contamination.16 Thus, the total number of patients in the present analysis was 766.
Clinic and Ambulatory Blood Pressure Measurements
An experienced physician measured each participant’s clinic blood pressure 3 times consecutively by the use of the Omron HEM-7051 device (Omron Healthcare, Kyoto, Japan) after the study subject had rested for ≥5 minutes in the sitting position. These 3 blood pressure readings were averaged for analysis.
Validated oscillometric SpaceLabs 90217 monitors (SpaceLabs Inc, Redmont, WA) were programmed to obtain blood pressure readings at 20-minute intervals from 6:00 am to 10:00 pm and at 30-minute intervals from 10:00 pm to 6:00 am. All recordings covered >20 hours and included ≥10 readings during the awake period and 5 readings during sleep. The 24-hour blood pressure means were weighed for the time interval between consecutive readings. Daytime and nighttime were defined as the short clock time intervals from 8:00 to 18:00 and from 23:00 to 5:00, respectively.
Questionnaire, Anthropometry, and Blood and Urine Sampling
The same physician administered a standardized questionnaire, inquiring into each subject’s medical history, intake of medications, and smoking and drinking habits. Nurses measured body height to the nearest 0.5 cm. Participants wore light indoor clothing without shoes for body weight measurement. Body mass index was weight in kilograms divided by the height in meters squared. Body surface area was calculated according to the Stevenson’s formula in Chinese.17
Venous blood samples were taken after overnight fasting for the measurement of plasma glucose concentration and for measurements of serum concentrations of sodium, potassium, creatinine, and lithium. In the same period (n=614) or within 2 weeks after (n=152) 24-hour ambulatory blood pressure recording, a 24-hour urine was collected for measurements of sodium, potassium, creatinine, and lithium, and a spot urine for measurement of albumin-to-creatinine ratio.
Estimated glomerular filtration rate (eGFR) was calculated according to the modified MDRD (Modification of Diet in Renal Disease) equations in Chinese.18 We defined mild and moderate/severe renal dysfunction as an eGFR <60 and 30 mL/min per 1.73 m2, respectively; diabetes mellitus as a fasting plasma glucose of ≥7.0 mmol/L or as the use of antidiabetic treatment19; and microalbuminuria and macroalbuminuria as an albumin-to-creatinine ratio of 30 to 299 and ≥300 mg/g, respectively.
Serum and Urinary Sodium and Lithium Measurements
Serum and urinary sodium concentration was measured using an automatic analyzer (Hitachi-7060C, Tokyo, Japan) and lithium concentration by inductively coupled plasma mass spectrometry (7500CE; Agilent Technologies Inc, Santa Clara, CA). The intra- and interassay coefficients of variation of lithium concentration were 5.4% and 16.4%, respectively, for the serum and 2.5% and 17.1%, respectively, for the urine.
Creatinine, sodium, and lithium clearances (C) were calculated with the formula Cx=Ux·V/Px and standardized by body surface area,17 where Ux and Px were urinary and plasma concentrations of the solute x, and V was the urine flow rate in mL/min. Fractional excretions of sodium (FENa) and lithium (FELi) were calculated as sodium and lithium clearances divided by creatinine clearance, respectively. FELi is a marker of proximal tubular sodium handling. A higher FELi indicates that less sodium and water is reabsorbed in the proximal tubule and hence more sodium and water is cleared from the proximal tubule to the distal tubule. Fractional distal reabsorption rate of sodium (FDRNa) was calculated by dividing the difference between lithium and sodium clearances by lithium clearance. FDRNa is a marker of the proportion of sodium escaping reabsorption in the proximal tubule that is not eliminated in the urine. A high FDRNa indicates that more sodium and water is reabsorbed in the distal tubule and hence less sodium and water is cleared from the distal tubule to the urine.
Statistical Methods
We used SAS version 9.13 (SAS Institute, Cary, NC) for database management and statistical analyses. Comparisons of means and proportions relied on the Student t test and Fisher exact test, respectively. We performed stepwise linear regression analyses to identify correlates of renal sodium handling indexes (FELi, FENa, and FDRNa) with P value set at 0.10 for variables to enter and stay in the model. We performed multiple linear regression analyses to study the association of urinary sodium excretion with clinic and 24-hour ambulatory blood pressure, before and after subdivision of the study participants according to FELi. We also performed analysis of covariance to study other associations of interest, while controlling for covariables.
Results
Characteristics of the Study Participants
The 766 participants included 379 (49.5%) men, 478 (62.4%) ambulatory hypertensive patients, 39 (5.1%) diabetic patients, 31 (4.0%) patients with albuminuria, and 2 (0.3%) patients with mild renal dysfunction. They were never treated (n=697) or did not take antihypertensive medication for ≥2 weeks (n=69). Table 1 shows the characteristics of the study participants by gender. Men, compared with women, had significantly higher clinic and ambulatory systolic and diastolic blood pressure (P<0.0001), urinary sodium excretion (P<0.0001), and FENa (P=0.0006), and lower FDRNa (P=0.03). However, men and women had similar FELi (P=0.27).
Characteristics of the Study Subjects by Sex (n=766)
Correlates of Renal Sodium Handling Parameters
In a multiple stepwise regression model, age, sex, body mass index, current smoking, and alcohol intake were considered as potential correlates of renal sodium handling parameters. FENa was significantly (P≤0.003) higher (+0.066%) in men versus women and with older age (+0.023% for each 10-year increase in age). FELi tended to be higher with older age (+0.849% for each 10-year increase in age; P=0.09). FDRNa was lower in men versus women (−0.388%; P=0.03).
Blood Pressure in Relation to Urinary Sodium Excretion, FELi, and FDRNa
After adjustment for age, sex, body mass index, current smoking, and alcohol intake, none of the associations of urinary sodium excretion, FELi, or FDRNa with each of the clinic and ambulatory blood pressure components were statistically significant (P≥0.09; Table 2). Further adjustment for urinary potassium excretion did not change the results (P≥0.08). However, in a similar adjusted continuous analysis, there was significant (P=0.01) interaction between urinary sodium excretion and FELi in relation to nighttime diastolic blood pressure (Table 3).
Blood Pressure in Relation to 24-Hour Urinary Sodium Excretion, FELi, and FDRNa (n=766)
Blood Pressure in Relation to Urinary Sodium Excretion According to FELi (n=766)
We, therefore, performed analyses after subdivision of the study participants according to tertile distributions of FELi. Both before (Figure 1) and after (Table 3) adjustment for the abovementioned confounding factors, nighttime diastolic blood pressure was associated with urinary sodium excretion in tertile 3 (P≤0.03) but not tertiles 1 and 2 of FELi (P≥0.14). The multivariate adjusted regression coefficient (±standard error) for nighttime diastolic blood pressure associated with each 1 standard deviation increment in urinary sodium excretion (65 mmol/day) was −0.70±0.59 (P=0.23), 0.90±0.62 (P=0.14), and 1.37±0.65 mm Hg (P=0.03) in tertiles 1, 2, and 3, respectively (Table 3). However, the level of nighttime diastolic blood pressure was slightly but significantly lower in tertile 3 than tertile 1 of FELi (71.9±0.6 versus 73.9±0.6 mm Hg; P=0.01).
Scatter plots on the association between nighttime diastolic blood pressure and urinary sodium excretion within each tertile of fractional excretion of lithium (FELi). Regression line was drawn with 95% confidence limits of the mean. P values for the regression coefficient are given for each tertile of FELi.
After further subdivision of the study participants within each tertile of FELi according to urinary sodium excretion, we noticed that only in the bottom tertile of urinary sodium excretion nighttime diastolic blood pressure was significantly different across tertile distributions of FELi (Figure 2). Indeed, in the bottom tertile of urinary sodium excretion, nighttime diastolic blood pressure was lower in tertiles 2 and 3 than tertile 1 of FELi by 2.8 (P=0.048) and 4.9 mm Hg (P<0.001), respectively.
Nighttime diastolic blood pressure in relation to interactions between urinary sodium excretion and fractional excretion of lithium (FELi). Values are means, adjusted for age, sex, body mass index, current smoking, and alcohol intake, according to the tertile distributions of FELi ( tertile 1,
tertile 2,
tertile 3) and urinary sodium excretion. Vertical lines denote SE. P values for trend are given for each tertile of FELi.
Similar trends were observed for clinic and 24-hour diastolic blood pressure and nighttime systolic blood pressure (Table 3 and Table S1 in the online-only Data Supplement). In addition, we performed sensitivity analyses on nighttime systolic and diastolic blood pressure according to the presence and absence of overweight (body mass index 25.0 to 29.9 kg/m2) and obesity (≥30.0 kg/m2) or in 614 subjects in whom the urine collection was done in the same period as 24-hour ambulatory blood pressure monitoring. The results of these sensitivity analyses were confirmatory (Table S2).
Discussion
Using the ambulatory blood pressure monitoring technique, our study has shown that sodium dietary intake and proximal tubular handling interact to be associated with blood pressure. At the usual levels of dietary sodium intake (ie, from 100 to 250 mmol/day; sodium chloride 5.8–14.6 g/day), those individuals with a lower FELi, which reflects higher proximal tubular sodium reabsorption, had higher blood pressure, whereas those individuals with a high FELi, which reflects lower proximal tubular sodium reabsorption, showed a positive association between blood pressure and dietary sodium intake. Accordingly, at low levels of dietary sodium intake, those with higher proximal tubular sodium reabsorption had 3.6, 3.1, 4.9, and 4.7 mm Hg higher clinic, 24-hour, and nighttime diastolic blood pressures, and nighttime systolic blood pressure, respectively.
Previous studies without accounting for dietary sodium intake investigated in the other way around—whether hypertensive patients had lower FELi than normotensive subjects—and produced inconsistent results.16,20–24 Weder first reported that 14 untreated hypertensive patients had a lower FELi than 31 normotensive subjects (13.96% versus 17.75%; P<0.01).20 In a later study, Burnier et al found that 13 untreated white-coat hypertensive patients but not 53 untreated hypertensive patients had significantly lower FELi than 48 normotensive subjects (11.6% versus 15.4% versus 17.0%).21 Two studies of similar size did not show any difference in FELi between hypertensive and normotensive subjects.22,23 Other studies even showed that hypertensive patients had a higher FELi than normotensive subjects.16,24 If a similar analysis would be performed in our study (data not shown), subjects in the highest tertile of nighttime diastolic blood pressure (mean systolic/diastolic blood pressure 125.5/83.5 mm Hg) had lower FELi than those in the lowest tertile (mean systolic/diastolic blood pressure 103.1/62.3 mm Hg) in unadjusted (19.5% versus 21.8%; P=0.01) as well as adjusted analyses (19.8% versus 21.7%; P=0.058). The contradictory results of these studies are not understood. However, taken the results of previous studies and our research together, we believe that the role of renal sodium handling in the determination of blood pressure can only be properly studied after accounting for the amount of dietary sodium intake.
The amount of dietary sodium intake might also explain the variance of our finding from an earlier observation in 340 black subjects from 76 pedigrees living in the Seychelles Islands on a free dietary sodium intake.9 In these black subjects, dietary sodium intake was positively associated with systolic blood pressure in subjects with low FELi but not in those with high FELi. The mean urinary sodium excretion per day was ≈50 mmol in these black subjects9 and 163 mmol in our study subjects. The relationship between blood pressure and sodium renal handling might differ from low to high dietary sodium intake. Nonetheless, the potential difference in the genetic background between the 2 different populations might also play a role.
The explanation on the more prominent association with nighttime than daytime blood pressure could be straightforward. According to Guyton pressure natriuresis theory, elevation in perfusion pressure in the renal artery would lead to a rapid increase in sodium and water excretion by the kidney.25 When the dietary sodium intake is too high to be sufficiently excreted by the transiently elevated blood pressure, sustained high blood pressure develops.25 Because the normal circadian blood pressure rhythm follows a rule of nocturnal dipping from daytime usually in a sodium-insensitive individual of low sodium intake,26 blood pressure elevation attributable to high sodium dietary intake or proximal tubular reabsorption is, therefore, most likely to initiate in the sleep hours and show a nondipping pattern. Nocturnal nondipping in blood pressure is associated with sodium sensitivity.27 Both sodium-sensitive humans9,10 and Dahl rats28 showed a lower FELi, indicating higher proximal sodium reabsorption. Nonetheless, because clinic blood pressure measured in the daytime and after ≥5 minutes’ rest in the sitting position showed similar trends, other factors, such as physical exercise, might have masked or attenuated the potential difference during the physically active daytime hours.
Why the association with diastolic blood pressure tended to be stronger than with systolic blood pressure can be explained by the predominant diastolic pressure elevation in the present study. Indeed, in 478 patients with ambulatory hypertension, 201 (42.1%) had isolated diastolic hypertension, 236 (49.4%) had systolic/diastolic hypertension, and only 41 (8.5%) had isolated systolic hypertension. The mean age of our study subjects was 51.4 years, and 78.1% were ≤60 years. In this relatively younger population, volume expansion attributable to increased dietary sodium intake or increased proximal tubular sodium reabsorption may lead to unbalanced increase in systolic and diastolic pressure.
Our study should be interpreted within the context of its strengths and limitations. A major strength was that we performed 24-hour ambulatory blood pressure monitoring in a relatively large number of subjects who were suspected of hypertension by measuring clinic blood pressure and not on any antihypertensive medication. In addition, the endogenous lithium clearance technique is considered a noninvasive and reliable evaluation of renal fractional sodium handling in human subjects. However, in addition to the cross-sectional design, several other limitations are noteworthy. First, our study subjects were consecutive patients instead of randomly selected population subjects. Second, we collected a single 24-hour urine sample for measurements of lithium and electrolytes. In spite of correction for creatinine clearance, the accuracy might be less than multiple collections of urine. In addition, the 24-hour urine collection was not timed for the day and night and hence does not allow testing of the hypothesis that nighttime blood pressure is dependent on the ability of the kidney to excrete sodium during the day.29 Finally, FELi had much larger interindividual variability than FENa, which strongly suggested that many factors might influence the kidney to handle water and sodium excretion at various levels of dietary sodium intake.
In conclusion, sodium dietary intake and proximal tubular handling interact to be associated with blood pressure. High dietary sodium intake, high proximal tubular sodium reabsorption, and especially both factors in combination are associated with high blood pressure, particularly during sleeping hours. In contrast, low dietary sodium intake and low proximal tubular sodium reabsorption are associated with low blood pressure, which in the long run may protect kidneys and other organs. These findings may have potential clinical implications for the diagnosis of renal tubular dysfunction9,10 and nocturnal hypertension30 by measuring lithium clearance and ambulatory blood pressure, respectively, and for the cardiovascular protection and prevention by dietary sodium restriction, and these should be confirmed by prospective follow-up studies.
Perspectives
Dietary sodium restriction is currently recommended as a nonpharmacological approach at population level for blood pressure lowering. However, probably, not only the blood pressure response to but also the health outcome of sodium restriction varies between individuals, depending on proximal tubular sodium handling. This is apparently an important topic of future longitudinal research and can be studied using the endogenous lithium clearance technique.
Acknowledgments
We gratefully acknowledge the voluntary participation of all study subjects and the expert technical assistance of Jing-Ling Han and Wei-Zhong Zhang (The Shanghai Institute of Hypertension, China).
Sources of Funding
The present study was financially supported by grants from the National Natural Science Foundation of China (grants 30871360, 30871081, 81170245, and 81270373) and the Ministry of Education (NCET-09-0544), Beijing, China; the Shanghai Commissions of Science and Technology (the Rising Star program 11QH1402000) and Education (the Dawn project 08SG20); the Shanghai Bureau of Health (XBR2011004); and Shanghai Jiaotong University School of Medicine (a grant of Distinguished Young Investigators to Yan Li).
Disclosures
None.
Footnotes
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.111.00776/-/DC1.
- Received December 10, 2012.
- Revision received January 10, 2013.
- Accepted July 20, 2013.
- © 2013 American Heart Association, Inc.
References
- 1.↵
- 2.↵
- Brown IJ,
- Tzoulaki I,
- Candeias V,
- Elliott P
- 3.↵Intersalt Cooperative Research Group. Intersalt: an international study of electrolyte excretion and blood pressure. Results for 24 hour urinary sodium and potassium excretion. BMJ. 1988;297:319–328.
- 4.↵
- Elliott P,
- Stamler J,
- Nichols R,
- Dyer AR,
- Stamler R,
- Kesteloot H,
- Marmot M
- 5.↵
- Forman JP,
- Scheven L,
- de Jong PE,
- Bakker SJ,
- Curhan GC,
- Gansevoort RT
- 6.↵
- 7.↵
- 8.↵
- Umesawa M,
- Iso H,
- Date C,
- Yamamoto A,
- Toyoshima H,
- Watanabe Y,
- Kikuchi S,
- Koizumi A,
- Kondo T,
- Inaba Y,
- Tanabe N,
- Tamakoshi A
- 9.↵
- 10.↵
- Chiolero A,
- Maillard M,
- Nussberger J,
- Brunner HR,
- Burnier M
- 11.↵
- Magnin JL,
- Decosterd LA,
- Centeno C,
- Burnier M,
- Diezi J,
- Biollaz J
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- Seidlerová J,
- Staessen JA,
- Maillard M,
- Nawrot T,
- Zhang H,
- Bochud M,
- Kuznetsova T,
- Richart T,
- Van Bortel LM,
- Struijker-Boudier HA,
- Manunta P,
- Burnier M,
- Fagard R,
- Filipovský J
- 17.↵
- Stevenson PH
- 18.↵
- Ma YC,
- Zuo L,
- Chen JH,
- Luo Q,
- Yu XQ,
- Li Y,
- Xu JS,
- Huang SM,
- Wang LN,
- Huang W,
- Wang M,
- Xu GB,
- Wang HY
- 19.↵
- Sacks DB,
- Arnold M,
- Bakris GL,
- Bruns DE,
- Horvath AR,
- Kirkman MS,
- Lernmark A,
- Metzger BE,
- Nathan DM
- 20.↵
- 21.↵
- Burnier M,
- Biollaz J,
- Magnin JL,
- Bidlingmeyer M,
- Brunner HR
- 22.↵
- 23.↵
- 24.↵
- Weinberger MH,
- Smith JB,
- Fineberg NS,
- Luft FC
- 25.↵
- Guyton AC
- 26.↵
- Uzu T,
- Ishikawa K,
- Fujii T,
- Nakamura S,
- Inenaga T,
- Kimura G
- 27.↵
- Castiglioni P,
- Parati G,
- Brambilla L,
- Brambilla V,
- Gualerzi M,
- Di Rienzo M,
- Coruzzi P
- 28.↵
- Roos JC,
- Kirchner KA,
- Abernethy JD,
- Langford HG
- 29.↵
- Bankir L,
- Bochud M,
- Maillard M,
- Bovet P,
- Gabriel A,
- Burnier M
- 30.↵
- Li Y,
- Wang JG
Novelty and Significance
What Is New?
Our study has shown that sodium dietary intake and proximal tubular handling interact to be associated with blood pressure. When proximal tubular sodium reabsorption is high, blood pressure is also high at the usual range of sodium intake. When proximal tubular sodium reabsorption is low, blood pressure is higher with higher dietary sodium intake.
What Is Relevant?
Our study provided evidence on the association between sodium dietary intake, proximal tubular sodium handling, and blood pressure.
Summary
There was interaction between sodium dietary intake and proximal tubular sodium handling in relation to blood pressure. At the usual levels of dietary sodium intake, a positive association between blood pressure and dietary sodium intake was observed in those individuals with lower proximal tubular sodium reabsorption, although those individuals with higher proximal tubular sodium reabsorption had higher blood pressure in general and at low levels of dietary sodium intake in particular.
This Issue
Jump to
Article Tools
- Blood Pressure in Relation to Interactions Between Sodium Dietary Intake and Renal HandlingNovelty and SignificanceJun Zou, Yan Li, Chong-Huai Yan, Fang-Fei Wei, Lu Zhang and Ji-Guang WangHypertension. 2013;62:719-725, originally published September 11, 2013https://doi.org/10.1161/HYPERTENSIONAHA.111.00776
Citation Manager Formats
Share this Article
- Blood Pressure in Relation to Interactions Between Sodium Dietary Intake and Renal HandlingNovelty and SignificanceJun Zou, Yan Li, Chong-Huai Yan, Fang-Fei Wei, Lu Zhang and Ji-Guang WangHypertension. 2013;62:719-725, originally published September 11, 2013https://doi.org/10.1161/HYPERTENSIONAHA.111.00776