Q J Med 2000; 93: 163-168
© 2000 Association of Physicians
Diastolic blood pressure is related to urinary sodium excretion in hypertensive Chinese patients
From the Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
Received 16 March 1998 and in revised form 12 January 2000
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
We studied 70 Hong Kong Chinese patients with untreated hypertension and 47 normotensive controls. Blood pressure measurements and 24-h urine collection were performed for each patient, and were repeated 12 weeks later in 14 hypertensive patients who remained untreated. Twenty-two hypertensive patients underwent ambulatory blood pressure monitoring. The primary hypothesis tested was a correlation between diastolic blood pressure and 24-h urinary sodium excretion. In the hypertensive patients, diastolic blood pressure correlated with 24-h urinary sodium excretion (r=0.41, p<0.001), even after adjustment for age, gender, body mass index, ethanol intake and season (r=0.34, p=0.02). In normotensive controls, diastolic blood pressure did not correlate with sodium excretion (r=0.21, p=0.16). A correlation between diastolic blood pressure and sodium excretion was also observed in the patients who underwent ambulatory blood pressure monitoring (r=0.47, p=0.026), and in repeat measurements in untreated patients (r=0.60, p=0.02). Systolic blood pressure did not correlate with sodium excretion, although it increased with patient age (0.6±0.1 mmHg/year, p<0.001). In a multiple regression analysis with diastolic blood pressure as the dependent variable, the regression coefficient was 0.06±0.02 mmHg/mmol Na. The regression coefficients for ambulatory diastolic blood pressure and diastolic pressure repeated at 12 weeks were 0.07±0.03 and 0.09±0.04 mmHg/mmol Na, respectively. Urinary sodium excretion was related to diastolic blood pressure in our hypertensive patients, accounting for 17% of the variance of diastolic blood pressure.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The relationship between sodium chloride and blood pressure is controversial. Guyton postulated that in essential hypertension, increased blood pressure is needed to maintain renal sodium excretion and sodium balance.1 Most prolonged studies of salt restriction in man have shown a change in blood pressure, especially in ‘salt-sensitive’ patients.2 Moreover, a direct relationship between salt intake and blood pressure has been demonstrated recently in the chimpanzee.3 However, the relationship between blood pressure and salt is weak in most communities studied in Intersalt,4 unless it is adjusted for regression dilution bias and not adjusted for body mass.5 In the Far East, where the average salt intake is high, the relationship between blood pressure and salt intake was found to be strong.4 Therefore, we investigated whether salt might be an important contributory factor for hypertension in a prospective study of Hong Kong Chinese patients with untreated hypertension.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
We studied 117 Hong Kong Chinese men and women; 47 were healthy controls without hypertension and not on any medications, 70 were patients referred to the hypertension out-patient clinic of a university teaching hospital with a diagnosis of hypertension, and not on any antihypertensive drug treatment. The inclusion criteria were: (i) age between 18 and 75 years, inclusive; (ii) essential hypertension newly diagnosed or previously diagnosed according to WHO/International Society of Hypertension criteria (diastolic blood pressure
90 mmHg or systolic blood pressure 140 mmHg on more than three separate occasions) but not treated with medications. The exclusion criteria were: (i) significant cardiac disease, including valvular heart disease, arrhythmia, previous myocardial infarction, angina and heart failure; (ii) known or suspected renal or renovascular disease; (iii) liver disease; (iv) oedema from whatever cause; (v) pregnancy; (vi) any serious concomitant disease; (vii) patients who were on a special diet. A full medical history was obtained, physical examination and urine analysis were performed in all. Histories of hypertension, diabetes mellitus, cardiac, hepatic and renal disease, and alcohol intake were sought. Height and weight were measured, from which body mass index was calculated. Blood was taken for a full blood count (haemoglobin, haematocrit, white blood count, differential white count and platelet count), erythrocyte sedimentation rate (ESR), renal and liver function tests (sodium, potassium, urea, creatinine, glucose, total bilirubin, total protein, albumin, ALT and alkaline phosphatase). Urinalysis for glucose, blood and protein was done in each patient after overnight fasting. The 24-h urinary sodium, potassium, albumin and creatinine excretion were determined from 24-h urine collections. An electrocardiogram and a posteroanterior chest X-ray were done in every hypertensive patient. Blood pressure was measured according to a strict protocol under standardized conditions in a temperature-controlled environment. The blood pressure of each patient was measured using a mercury sphygmomanometer and a cuff of the appropriate size around the right arm after the patient had been sitting in a chair, at rest, for 15 min. Blood pressure was measured four times at 5-min intervals. The first blood pressure measurement was to familiarize the patient with the procedure and the sensation of the inflated cuff. The three subsequent systolic and diastolic blood pressure readings were recorded to the nearest 2 mmHg. The mean of three readings was used in data analysis. For hypertensive patients, blood pressure measurements obtained at the third visit to the research clinic were used in the data analysis. Each hypertensive patient was offered 24-h ambulatory blood pressure monitoring (ABPM-630, Colin Electronics). The equipment was set to measure blood pressure hourly. Day and night were defined as 06.00–24.00 and 0.00–6.00, respectively. Ambulatory blood pressure data were available in 22 patients.
Blood pressure and sodium excretion were measured again 12 weeks later in 14 patients in whom antihypertensive drug treatment had not been started.
Statistical analysis of data was done using SPSS for Windows, version 6.1. Spearman's rank correlation was used, as normal distribution of data was not assumed. Stepwise multiple regression was used to investigate the relationship between diastolic blood pressure and independent variables. Regression coefficients are in original units (mmHg/mmol Na) for ease of interpretation. Logarithmic transformation of 24-h urinary sodium excretion resulted in slightly higher R2 values and did not affect the conclusions. The primary hypothesis was a correlation between diastolic blood pressure and 24-h urinary sodium excretion. p<0.05 was considered significant.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Baseline characteristics of the 117 patients and controls are shown in Table 1
. The ethanol intake was low. Among hypertensive patients, 71% were abstainers, while the two highest reported consumption figures among the others were 6 units/week and 42 units/week. Ethanol intake differed between men and women (p=0.01); 12 men drank alcohol occasionally and five drank regularly, whereas only two women drank occasionally and none drank regularly.
|
In hypertensive patients, diastolic blood pressure correlated with 24-h urinary sodium excretion (r=0.41, p<0.001) (Table 2
, Figure 1), even after controlling for age, gender, body mass index, ethanol intake and season (r=0.34, p=0.02). Diastolic blood pressure correlated with sodium excretion in men (r=0.38, p=0.01) and women (r=0.43, p=0.03). Urinary potassium excretion tended to be negatively associated with diastolic blood pressure (r=-0.24, p=0.05). In normotensive controls, diastolic blood pressure did not correlate with sodium excretion (r=0.21, p=0.16). In hypertensive patients, ambulatory diastolic blood pressure correlated with 24-h urinary sodium excretion (r=0.47, p=0.026) (Figure 2), which was mainly due to the correlation between urinary sodium excretion with daytime ambulatory diastolic blood pressure (r=0.53, p=0.01) rather than nocturnal ambulatory diastolic blood pressure (r=0.21, p=0.35). Correlation between blood pressure and urinary sodium excretion was also observed 12 weeks later in those patients in whom antihypertensive drug treatment had not been started (r=0.60, p=0.02) (Figure 3). The correlation between systolic blood pressure and sodium excretion was weak and non-significant, with or without correcting for age, gender, body mass index, potassium excretion, ethanol intake and season (Table 2). However, mean arterial pressure correlated with urinary sodium excretion (r=0.31, p=0.02), and systolic blood pressure increased with age of patient (0.60±0.14 mmHg/year, p<0.001).
|
|
|
|
Body weight correlated with diastolic blood pressure (r=0.36, p=0.003), and was weakly related to 24-h urinary sodium excretion (r=0.23, p=0.06), and so might explain part of the correlation between urinary sodium excretion and diastolic blood pressure in hypertensive patients. However, when body weight was controlled for, the correlation between urinary sodium excretion and diastolic blood pressure remained significant (r=0.30, p=0.01).
The 24-h urine collection depended on the co-operation of patients, and so urine volume could be a confounding factor. However, it did not correlate with diastolic blood pressure (r=-0.17, p=0.18) or sodium excretion (r=0.16, p=0.23).
Urinary sodium excretion, diastolic blood pressure and 24-h creatinine clearance were inter-related, as diastolic blood pressure correlated with creatinine clearance (r=0.26, p=0.04) and creatinine clearance correlated with urinary sodium excretion (r=0.51, p<0.001). These correlation coefficients remain significant if creatinine clearance is adjusted for body surface area.
In a multiple regression analysis with diastolic blood pressure as the dependent variable, the regression coefficient (slope of the regression line) was 0.057±0.018 mmHg/mmol Na. Hence, a sodium intake 100 mmol higher was associated with a diastolic blood pressure 6 mmHg higher. Age, gender, body mass index, urinary potassium, ethanol intake and season were rejected as predictor variables in stepwise regression analysis. Inclusion of these variables did not alter the regression coefficient for sodium excretion appreciably. The regression coefficients for male and female patients were 0.046±0.021 mmHg/mmol Na and 0.076±0.034 mmHg/mmol Na, respectively. The regression coefficient in those who had ambulatory blood pressure monitoring was 0.068±0.029 mmHg/mmol Na, similar to the regression coefficient based on blood pressure measurements in the research clinic. The regression coefficient in those whose blood pressure was measured again at 12 weeks was also similar, at 0.091±0.035 mmHg/mmol Na, despite considerable variation in blood pressure and sodium excretion during this period. In contrast, the regression coefficient in normotensive controls was 0.034±0.025 mmHg/mmol Na.
As additional analyses, we related blood pressure to urine sodium : potassium ratio and urine sodium : creatinine ratio which are used by some investigators. Urine sodium : potassium correlated with diastolic pressure in hypertensive patients (r=0.45, p=0.002). There was no correlation between urine sodium : creatinine and diastolic blood pressure. Another measure of sodium excretion favoured by renal physiologists is the fractional excretion of sodium, defined as the ratio of the clearance of sodium to the glomerular filtration rate (approximately the creatinine clearance). It is independent of urine volume and plasma creatinine concentration. In our hypertensive patients, there was a relationship between the fractional excretion of sodium and diastolic blood pressure (r=0.28, p=0.05).
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
This was a prospective study in untreated hypertensive patients to investigate a novel hypothesis that diastolic blood pressure is related to sodium intake. There is no precedent in the literature for hypothesizing that sodium affects diastolic rather than systolic blood pressure. Our unconventional hypothesis was formulated after we observed in our hypertensive patients that diastolic blood pressure appeared to be related to urinary sodium excretion, while systolic blood pressure was strongly related to age.
The relationship between blood pressure and salt is weak in most communities studied in Intersalt.4 Even within a single country, studies of the association between salt intake and blood pressure can differ in their conclusions.6,7 Hence, a causal link between salt and blood pressure has been contested by the food industry.8,9 Our results are consistent with the unexpected relationship between diastolic blood pressure and sodium excretion found in Nanning and Tianjin in China in the Intersalt study,4,10 and provide further evidence that this is likely to be a genuine phenomenon in this part of the world.
The number of subjects in this preliminary study was small, due to the scarcity of untreated hypertensive patients. Hypertensive patients on drug treatment were excluded from this study, because drug treatment of hypertension is a serious confounding factor, and antihypertensive drugs may alter the blood pressure for any given level of salt intake. In particular, diuretics and angiotensin-converting enzyme inhibitors tend to increase salt excretion while lowering blood pressure.
Nevertheless, our conclusions are robust for a number of reasons. Firstly, the correlation between diastolic blood pressure and sodium excretion was strong before controlling for variables such as age, gender, body mass index, ethanol intake and season, and remained strong after controlling for these variables. Secondly, the relationship between diastolic blood pressure and sodium excretion was strong even without correction for regression dilution bias. Thirdly, ambulatory blood pressure readings and blood pressure readings repeated at 12 weeks showed the same correlation, with very similar correlation and regression coefficients. Moreover, observer bias was eliminated by ambulatory blood pressure measurement.
The relationship between diastolic blood pressure and sodium excretion was much less strong in normotensive controls. This was expected, because not everyone with a high salt intake will become hypertensive. The regression coefficient, 3.4 mmHg per 100 mmol increase in sodium excretion, is very close to that found in Nanning, Tianjin and Taiwan in the Intersalt Study.4 This degree of association attains conventional statistical significance when there are 100 subjects or more.
In this study, emphasis was placed on the accurate measurement of blood pressure. The office blood pressure readings on the first two visits of these patients to our hypertension research clinic were not used in the analysis. This allowed patients to get used to the surroundings, the staff and the measurement of blood pressure. In our experience, this would give a truer reading of blood pressure than in a one-off visit.
Systolic blood pressure did not correlate significantly with urinary sodium excretion in our study. A study larger than the present one might be able to demonstrate a significant but weak relationship between systolic blood pressure and urinary sodium excretion. Nevertheless, the weak relationship between systolic blood pressure and urinary sodium excretion in our study contrasted markedly with the strong and consistent relationship between diastolic blood pressure and urinary sodium excretion. It is noteworthy that systolic blood pressure correlated strongly with age and increased markedly with age in our patients. This is in agreement with the increment in systolic blood pressure with age, especially in populations with high salt intake, observed in the INTERSALT study.4,5 The rise in systolic blood pressure with age has also been demonstrated convincingly in a survey of 7730 individuals randomly selected from the whole population in Hong Kong.11
The mean salt intake in our patients was high, and was comparable to the salt intake found in the survey of 7730 Hong Kong men and women.11 The preference for marinated meat, soya sauce and monosodium glutamate in food are probably responsible for the high sodium intake. On the other hand, the intake of ethanol is low both in our hypertensive patients and in the whole population.11 Ethanol is therefore not a major factor influencing the development of hypertension in this population. The combination of low ethanol but high salt intake in this population may allow sodium to emerge as a major factor influencing hypertension. Interestingly, urinary potassium excretion, unlike sodium excretion, did not correlate with diastolic blood pressure. It would be of interest to study these subjects further in a metabolic ward on controlled diets to measure accurately their plasma renin activity and salt sensitivity.
Although the relationship between blood pressure and salt intake is almost flat in normotensive persons without a family history of hypertension, an increase in blood pressure in response to increased salt intake may be a pathological response in hypertensive patients and may reflect an underlying abnormality in essential hypertension.1 In view of the correlations among urinary sodium excretion, diastolic blood pressure and creatinine clearance, it is tempting to speculate that changes in salt intake affect diastolic blood pressure and glomerular filtration rate which then adjusts the urinary sodium excretion to maintain sodium balance in a Guytonian manner.
We have observed from the family history of 323 hypertensive patients that there is strong familial aggregation of hypertension.12 Although the prevalence of hypertension in adults in Hong Kong is 10%,11 91% of our patients have a hypertensive first-degree relative, 71% have at least one hypertensive parent, and 41% of patients have one or more hypertensive siblings. Shared environment and dietary habits may partly account for the familial aggregation of hypertension, but there may also be an inherited susceptibility. It is tempting to speculate that the genes predisposing to hypertension may code for salt sensitivity in this population.
In conclusion, our results show that sodium intake is strongly related to diastolic blood pressure in our hypertensive patients, accounting for 17% of the variance of diastolic blood pressure. Long-term reduction in salt intake may be indicated, at least in these hypertensive patients, if not the whole population.
|
Acknowledgments |
|---|
Drs Cheung and Lau are members of the Institute of Cardiovascular Science and Medicine, University of Hong Kong. They received support for a study on sodium intake and blood pressure from the Health Services Research Committee Grant, Hong Kong. The study of the prevalence of hypertension in relatives of hypertensive patients was supported by a Committee on Research and Conference Grant, University of Hong Kong.
|
Notes |
|---|
Address correspondence to Dr B.M.Y. Cheung, University Department of Medicine, Queen Mary Hospital, Pokfulam, Hong Kong. e\|[hyphen]\|mail: mycheung{at}hkucc.hku.hk
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Guyton AD, Coleman TG, Cowley AW Jr, Scheel KW, Manning RD, Norman RA. Arterial pressure regulation: overriding dominance of the kidneys in long-term regulation and in hypertension. Am J Med 1982; 52:584–94.
2. Law MR, Frost CD, Wald NJ. By how much does dietary salt reduction lower blood pressure? Br Med J 1991; 302:819–24.
3. Denton D, Weisinger R, Mundy NI, Wickings EJ, Dixson A, Moisson P, et al. The effect of increased salt intake on blood pressure of chimpanzees. Nature Med 1995; 1:1009–16.[Web of Science][Medline]
4. Intersalt Cooperative Research Group. Intersalt: an international study of electrolyte excretion and blood pressure. Results for 24 hour urinary sodium and potassium excretion. Br Med J 1988; 297:319–28.
5.
Elliott P, Stamler J, Nichols R, et al. for the Intersalt Cooperative Research Group. Intersalt revisited: further analyses of 24 hour sodium excretion and blood pressure within and across populations. Br Med J 1996; 312:1249–53.
6. Smith WCS, Crombie IK, Tavendale RT, Gulland SK, Tunstall-Pedoe. Urinary electrolyte excretion, alcohol consumption and blood pressure in the Scottish heart health study. Br Med J 1988; 297:329–30.
7.
Beard TC, Blizzard L, O’Brien DJ, Dwyer T. Association between blood pressure and dietary factors in the dietary and nutritional survey of British adults. Arch Intern Med 1997; 157:234–8.
8.
Hanneman RL. Intersalt: hypertension rise with age revisited. Br Med J 1996; 312:1283–4.
9.
MacGregor GA, Sever PS. Salt—overwhelming evidence but still no action: can a consensus be reached with the food industry? Br Med J 1996; 312:1287–9.
10. Stamler R, Liu LS, Nichols R, Huang DX, Long ZP, Xie JX, Elliott P. Blood pressure and life style in the People's Republic of China: three samples in the INTERSALT Study. J Hum Hypertens 1993; 7:429–35.[Web of Science][Medline]
11. Janus ED, ed. The Hong Kong Cardiovascular Risk Factor Prevalence Study 1995–1996. Hong Kong, 1997.
12. Cheung BMY, Ho SPC, Lau CP. Prevalence of hypertension in close relatives of patients with essential hypertension. Hong Kong Practitioner 1998; 20:2(suppl.):4.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S.-C. Tiu, C.-H. Choi, C.-C. Shek, Y.-W. Ng, F. K. W. Chan, C.-M. Ng, and A. P. S. Kong The Use of Aldosterone-Renin Ratio as a Diagnostic Test for Primary Hyperaldosteronism and Its Test Characteristics under Different Conditions of Blood Sampling J. Clin. Endocrinol. Metab., January 1, 2005; 90(1): 72 - 78. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T.O. Cheng Systolic and diastolic blood pressures and urinary sodium excretion in mainland China QJM, August 1, 2000; 93(8): 557 - 558. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||