Relationship between salt excretion and blood pressure in various regions of China

Bulletin of the World Health Organization, 64 (5): 729-733 (1986) i World Health Organization 1986 Relationship between salt excretion and blood pressure in various regions of China: Part 2* L. S. Liu1 & S. H. LAI2 Overnight urine samples were collected on three consecutive daysfrom 2483 persons in 12 regions of China and the mean 9-hour overnight sodium and potassium levels sub- sequently determined. Analysis of the data by univariate regression, simple correlation, and multiple regression indicated that there was a positive correlation between the mean level of sodium in urine and blood pressure. During 1979-80, a nationwide survey of blood pressure was carried out in China according to a uniform study protocol (1). The northern and north- eastern provinces had a higher prevalence of hyper- tension than the southern provinces. Geological, socioeconomic, and racial factors may account in part for these differences; however, different patterns of food consumption and cooking practices, especially salt intake, may also contribute to the observed variations. In order to determine whether there is a relation- ship between salt intake and the prevalence of hyper- tension in various regions of China and also whether sodium itself affects blood pressure, a collaborative study was carried out in 1981. Individuals of similar group mean age from 12 regions of the country where the population was of Han nationality were selected for the study; patients under treatment were excluded (5). MATERIALS AND METHODS Random population samples of men and women of group mean age range 41.0-49.5 years were selected from communities that had been surveyed for blood pressure in 1979-80 (5). A subset of data from this survey formed the basis of the present study, and a total of 2483 persons participated (Table 1). Overnight urine samples were collected from each patient on three consecutive days. The urine was analysed for sodium and potassium by flame * From the Cardiovascular Institute and Fu Wai Hospital, Chinese Academy of Medical Sciences, Beijing, China. For Part 1, see Bulletin of the World Health Organization, 62: 255-260 (1984) and 63: 963 (1985). ' Chief, Department of Cardiology and Hypertension. 2 Postgraduate student, Department of Epidemiology. photometry at local laboratories, without standard- ization of the method used. All the data were included in the subsequent statistical analysis; com- pleteness was ascertained primarily by carefully questioning the participants about the possibility of missed or extra specimens. The height and weight of each participant were measured for subjects in light clothes without shoes. Blood pressures were measured in the right brachial artery after 15 minutes' rest in a quiet room, with the subject in the sitting position and his or her arm resting on a table at heart level. Pressures were recorded at the appearance of systolic and fifth-phase diastolic Korotkoff sounds. Measurements were made in the morning of each of the three days when urine was collected. On each occasion, readings were made until the difference between two successive dia- stolic pressures was less than 4 mmHg, and for each individual the average of the two lowest readings was taken as the blood pressure of that day. The mean of six readings (the two lowest daily pressures for three days) was used in the analysis. Of the 2483 persons in the study, 107 (4.3%) had definite hypertension (systolic pressure > 160 mmHg (21.3 kPa) or dia- stolic pressure > 95 mmHg (12.7 kPa)). Statistical methods Initially, data were analysed by interpopulation unifactorial regression and simple correlation of group means. Multiple regression analysis was then performed, using as independent variables the mean value of age, body mass index (weight (kg)/height2 (m2)), and level of sodium and potassium in urine; with systolic and diastolic pressure as the dependent variables. Standardized multiple regression coeffi- cients were obtained by multiplying the calculated 4716 -729- L. S. LIU& S. H. LAI Table 1. Characteristics of the participants in the urine sodium survey, China, 1981 Mean body mass index Sex Mean age (years) (kg/M2) No. of No. with Region participants hypertension M F M F M F North China Beijing (urban) 334 27 181 153 43.7 42.5 23.1 24.0 (5.2)' (4.7) (2.7) (3.5) Beijing (rural) 111 8 65 46 44.8 42.2 22.6 23.7 (6.1) (4.7) (2.9) (3.4) Shanxi 45 0 9 36 41.0 42.1 22.5 22.6 (6.7) (5.7) (2.4) (2.9) Lanzhou 209 17 209 - 44.3 22.4 (5.2) - (2.8) - Shaanxi 52 16 23 29 47.5 48.2 20.5 21.3 Hanzhung (4.5) (4.2) (2.7) (2.0) Shijiazhang 230 0 115 115 43.7 42.1 23.2 23.6 Hobei (6.0) (6.5) (2.0) (2.5) Middle China Zhenjiang 48 1 15 33 46.5 46.0 22.6 21.5 (5.0) (5.1) (3.2) (3.0) Shanghai 232 7 53 179 42.3 44.7 20.2 20.9 (21.2) (16.5) (2.6) (2.3) Daishan 226 7 34 192 49.5 43.1 21.5 20.1 (14.9) (14.5) (3.7) (2.0) Zhoushan 58 0 58 - 43.3 - 21.1 (6.1) - (1.7) - South China Guangdong 215 5 112 103 44.2 44.2 20.1 19.9 (6.7) (5.8) (1.7) (1.8) Guangxi 723 9 667 56 43.6 43.3 20.5 20.9 (5.6) (5.3) (2.7) (3.1) Meanb _- - - 44.5 43.8 21.7 21.9 (2.3) (2.0) (1.2) (1.5) Figures in parentheses indicate standard deviations. b Represents the mean and standard deviation of the data shown. regression coefficients by the standard deviations of the variables. RESULTS The means and standard deviations for blood pressures and other variables are shown in Table 2. For males the mean 9-hour overnight sodium level ranged from 51 to 144 mmol/l, while that of potass- ium ranged from 7 to 15 mmol/l; for females the corresponding levels ranged from 48 to 126 mmol/l and from 6 to 13 mmol/l, respectively. The results of the interpopulation simple cor- relation and regression analyses are shown in Table 3 and Fig. 1 and 2, while those of the multiple linear regression analyses are given in Table 4. For females, sodium levels correlated significantly with both systolic and diastolic blood pressure; however, for males there was correlation only with systolic press- ure. The level of potassium in urine did not correlate significantly with blood pressure for either sex. DISCUSSION That arterial hypertension can be caused by an excess of sodium chloride in the diet was postulated 730 SALT EXCRETION AND BLOOD PRESSURE IN CHINA Table 2. Means, standard deviations of pressure, China, 1981 9-hour overnight levels of sodium and potassium in urine, and blood Urine sodium Urine potassium Systolic pressure Diastolic pressure (mmol/l)' (mmol/l) (mmHg) ' b (mmHg) b Region M F M F M F M F North China Beijing (urban) 118.05 101.53 14.42 13.11 118.05 117.11 78.68 77.07 (48.05) (74.85) (5.67) (5.04) (16.00) (17.62) (11.37) (11.79) Beijing (rural) 96.03 79.89 10.20 9.83 124.34 117.84 80.88 76.97 (37.50) (33.62) (4.33) (4.96) (16.96) (13.27) (10.74) (9.49) Shanxi 96.24 92.34 11.56 11.10 120.00 114.67 78.11 74.74 (36.96) (31.31) (3.73) (3.99) (7.97) (7.52) (7.98) (4.69) Lanzhou 69.78 - 13.56 - 113.17 - 77.53 (27.03) - (6.33) - (14.78) - (11.88) - Shaanxi 144.06 126.17 12.07 10.69 128.35 144.66 85.30 87.69 Hanzhung (34.25) (32.62) (4.68) (4.74) (30.69) (32.27) (17.76) (15.43) Shijiazhang 100.28 91.05 14.68 13.19 116.37 119.56 72.60 72.99 Hobei (46.59) (38.85) (11.40) (5.68) (14.72) (19.33) (9.96) (10.65) Middle China Zhenjiang 73.19 60.42 7.23 6.31 115.53 110.48 74.67 73.58 (23.21) (26.66) (2.37) (2.78) (10.84) (9.77) (5.14) (7.07) Shanghai 95.19 69.29 9.31 7.12 114.15 113.42 69.92 67.29 (45.54) (25.70) (4.95) (3.06) (14.59) (19.34) (8.83) (10.05) Daishan 112.66 99.75 7.79 7.34 121.33 113.71 73.00 69.55 Zejiang (48.40) (46.96) (3.59) (3.64) (28.73) (21.32) (13.15) (10.45) Zhoushan 97.21 - 9.07 - 107.38 - 66.10 (40.85) - (4.03) - (7.28) - (5.50) - South China Guangdong 51.22 47.62 7.60 7.64 114.93 109.86 73.10 70.29 (18.46) (20.73) (3.25) (3.29) (14.33) (12.37) (8.20) (7.39) Guangxi 63.69 53.15 8.18 8.48 113.08 107.07 71.62 67.29 (24.37) (21.59) (4.07) (4.91) (35.17) (18.32) (8.50) (10.05) Meanc 93.13 82.12 10.47 9.48 117.22 116.84 75.13 73.74 (25.68) (24.68) (2.72) (2.49) (5.63) (10.51) (5.22) (6.06) ' Mean value shown. Figures in parentheses are standard deviations. b Blood pressure is given in mmHg: 100 mmHg = 13.3 kPa. ' Represents the mean and standard deviation of the data shown. as early as 1904 by Ambard and Beaujard. Epidemio- logical evidence of such an association has come from studies of Japanese, American, and Eskimo populations (2-4). In China, there are differences in the prevalence of hypertension between populations in the north and south of the country (Table 1). In this respect, there are also climatic and body-build differences, as well as different patterns of food consumption. To what extent is salt intake responsible for these blood pressure differences? Whether dietary sodium levels should be reduced has become an important issue in China, especially since a restriction in intake of sodium has been proposed as a primary preventive measure in developed countries (S. C. Tao, unpub- lished observations). We have previously reported a positive intra- population correlation between the level of sodium in urine or the ratio sodium in urine/potassium in urine and blood pressure in the study regions (5). This correlation applies also to the subset of data used in the present study. However, when each region was treated as a homogeneous unit, no significant inter- population correlation was found, presumably because of confounding factors. In order to minimize interference from age, race, and blood viscosity, we 731 L. S. LIU & S. H. LAI Table 3. Correlation coefficients (r) between population characteristics, urine sodium and potassium levels, and blood pressure, China, 1981 Body mass Urine Urine Diastolic Systolic index sodium potassium pressure pressure Males Age -0.0742 0.3203 -0.2745 0.2635 0.4663 Body mass index 0.1460 0.57962 0.2815 0.1260 Urine sodium 0.4332 0.4519 0.6231 Urine potassium 0.4390 0.2017 Diastolic pressure 0.8469 Females Age -0.4807 0.1547 -0.3766 0.4597 0.5651 Body mass index 0.3576 0.7974b 0.3775 0.1679 Urine sodium 0.5906 0.7384b 0.8062 Urine potassium 0.4895 0.4212 Diastolic pressure 0.8926 a P<0.05. p0.01 . selected 12 male and 10 female population groups of similar age, having the same nationality, and who lived at similar altitudes. Age was kept within a narrow range (41-49 years), and hence the correlation coefficients between age and blood pressure were not significant. Three of the four correlation coefficients between sodium level and blood pressure were statis- tically significant (for sodium and systolic pressure, r=0.6231 for males and 0.8062 for females; while for females, r= 0.7384 for sodium and diastolic pressure). For either sex, levels of urinary potassium were not significantly related to either systolic or dia- stolic pressure. Based on the populations surveyed, residents of southern China are shorter and thinner, as indicated by their mean body mass indices, than those in other areas of the country. However, in the simple regression-correlation analyses described here, no significant association was found between body mass index and blood pressure. For males, multiple regression analysis of age, body mass index, and sodium level showed that the regression coefficient between sodium and systolic blood pressure was significant over the range a a 80 90 100 110 120 130 140 150 Socdium level (mmol/l) Fig. 1. Linear regression plot of mean systolic pressure among males in the 12 population samples and their mean 9-hour sodium excretion level (r=0.6231, P< 0.05). Table 4. Standardized regression coefficients from a multiple regression analysis of blood pressure, popu- lation characteristics, and urine sodium levels Urine Body mass sodium Age index Males Systolic pressure 2.8941 a 1.7195 0.4258 Diastolic pressure 1.9003 0.8558 1.2888 Females Systolic pressure 6.7574b 5.9620c 2.1878 Diastolic pressure 2.7370° 3.8954c 3.1421 a a P<0.10. b p<0.01. c P<0.05. 'a 130 I E E o 120 to 0O,o a 50 60 70 140. r- 732 SALT EXCRETION AND BLOOD PRESSURE IN CHINA 733 EE 130- Z.20- =110 3040 S0 eo 10 eo 90 100 110 120 130 10 Sodium level (mmol/l) Fig. 2. Linear regression plot of mean systolic pressure among females in 10 of the 12 population samples and their mean 9-hour sodium excretion level (r=0.8062, P< 0.005). 0.05<P<0.10; however, the regression coefficients between the other variables and blood pressure, either systolic or diastolic, were not statistically sig- nificant. For females, in contrast, sodium level and age were both significantly related to systolic pressure (P<0.01 and P<0.05, respectively); for diastolic pressure, sodium level and body mass index had borderline significance (0.05<P<0.10), while age was significant (P<0.05). The results reported here indicate that differences in the prevalence of arterial hypertension in the north and south of China correlate with salt intake. For the age group 41-49 years, females appear to be more sensitive in this respect than males, with the systolic pressure of both sexes showing closer correlation than diastolic pressure with salt intake. ACKNOWLEDGEMENTS The collaborative group taking part in the study included the Shanghai Institute of Hypertension, the Guangdong Provincial Institute of Cardiovascular Diseases, the Beijing Jiuxianchiao Hospital, the Second People's Hospital, Lanzhou Medical College, the Sanxi Provincial People's Hospital, the Hanzhung Municipal People's Hospital, the Zhengjiang Region People's Hospital, the Guangxi Medical College, the Cardiovascular Disease Institute, Chinese Academy of Medical Sciences, and the Capital Iron and Steel Complex Hospital. We are grateful to Professor J. Stamler and Professor K. Liu, Northwestern University Medical School, Chicago, IL, USA, for their valuable comments in the preparation of this article. RPSUME RELATION ENTRE LA NATRIURIE ET LA TENSION ARTERIELLE DANS DIFFERENTES REGIONS DE CHINE: Partie 2 En Chine, il existe des differences de pr6valence de 1'hypertension arterielle entre le Nord et le Sud du pays. Afin d'6tudier plus en detail ces differences, nous avons examine la relation entre la prise de sel et la pr6valence de l'hypertension dans chacune des 12 regions de Chine. Pen- dant 3 jours cons&utifs, on a recueilli les urines de la nuit (a la premiere miction de la journ&e) de 2483 hommes et femmes, la moyenne d'age des differents groupes etant de 41 a 49,5 ans. Chez les hommes, le taux de sodium moyen sur 9 heures etait de 51 a 144 mmol / et le taux de potassium de 7 a 15 mmol/l; chez les femmes, les taux correspondants 6taient de 48 a 126 mmol/l et de 6 a 13 mmol/l. Les donn6es ont ete soumises a une analyse statistique par regression simple, correlation simple (sur les moyennes des groupes) et regression multiple. Les r6sultats montrent une correlation entre les differences de prevalence de l'hyper- tension entre le Nord et le Sud du pays et l'absorption de sel. Les femmes semblent plus sensibles A cet effet que les hommes, et c'est la pression systolique qui chez les deux sexes est la plus etroitement correlee avec la prise de sel. REFERENCES 1. Lu, C. C. ET AL. Nationwide hypertension screening in China during 1979-80, Chinese journal of cardio- vascular diseases, 8: 165 (1980). 2. DAHL, L. K. Possible role of salt intake in the develop- ment of essential hypertension. In: Bock, K. D. & Cottier, P. T., ed. Essential hypertension. Berlin, Springer, 1960. 3. GLEIBERMANN, L. Blood pressure and dietary salt in human populations. Biology offood and nutrition, 2: 143 (1973). 4. KAY-TEEKHAW & ROSE, G. Population study of blood pressure and associated factors in St. Lucia, West Indies. International journal of epidemiology, 11: 372 (1982). 5. Liu, L. S. ET AL. Relationship between salt excretion and blood pressure in various regions of China. Bulletin of the World Health Organization, 62: 255-260 (1984).