Vegetarian diet in mild hypertension: effects of fat and ...

Vegetarian

diet in mild hypertension:

Barrie MMargetts, PhD; Lawrence andRobert Vandongen, MD ABSTRACI was sure

J Beilin,

Recently,

estimated

a

to substantially

reduction

is about

the

MD;

effects

Bruce

KArmstrong,

of fat and fiber1’2 DPhil;

relatively small reduction in systolic blood pressure(’5 mm Hg) reduce the numbers of major coronary events. The blood pressame

as the

difference

seen

between

typical

ovolactovegetarians

and omnivores. This paper reviews the evidence for the blood pressure-lowering effects of a vegetarian diet on those with elevated blood pressure. It also reviews whether the effect on blood pressure ofa vegetarian diet can be attributed either to elevation ofthe dietary P:S ratio or to fIber intake alone. Am J C/in Nutr l988;48:80l-5. WORDS

Vegetarian,

diet,

hypertension,

P:S

to a vegetarian trial yet published. The diet et al’s study was not vegetarian; blood pressures fell on this regime and were similar to levels achieved by drug therapy. We recently completed a randomized controlled trial of the effect of an ovolactovegetarian diet on the blood pressure ofuntreated mildly hypertensive subjects. The diet tested was that of typical long-term vegetarians living in Western Australia. The study was described in detail elsewhere (6). In 1983 a representative sample ofthe population of Perth, Western Australia, aged 25-64 y was screened for coronary risk factors, including blood pressure. All subjects who were currently untreated for high blood pressure and whose blood pressures were between 150 and 200 mm Hg systolic or between 90 and 1 15 mm Hg diastolic were asked to return for a repeat measurement. Subjects with uncomplicated (as assessed by clinical examination) persistent mild hypertension on two further visits were asked to participate in the trial. Of 76 eligible subjects, 60 entered the randomization phase and ofthese 58 subsequently completed the trial. The study design is summarized in Table 1. Before commencement of the test diet subjects completed a 7-d dietary record in which portion sizes were estimated. Subjects were then randomly allocated to one of three groups. The guidelines for subjects when on the vegetarian diet are shown in Table 2; subjects usual dietary pracin

It was estimated recently that a relatively small reduction in systolic blood pressure (5 mm Hg) could substantially reduce the numbers of major coronary events (1). The blood pressure reduction is about the same as the difference seen between typical ovolactovegetarians (vegetarians) and omnivores (2). Although there has been considerable research on the pharmacological ways oflowering blood pressures in those with elevated blood pressures, relatively little has been done on the blood pressure lowering effect ofdiet in general and vegetarian diets in particular (2, 3). This paper reviews the evidence for the blood-pressure-lowering effects of a vegetarian diet on those with elevated blood pressure and reviews whether the effect on blood pressure ofa vegetarian diet cam be attributed either to elevation ofthe dietary ratio of polyunsaturated to saturated fatty acids (P:S) or to fiber intake alone. Vegetarian

diet in mild

hypertension

The evidence for a blood-pressure-lowering effect of a vegetarian diet in subjects with blood pressures in the normal range was recently reviewed (4). To date little work has been done to establish whether those with blood pressures above the normal range would show a similar blood-pressure-lowering effect upon changing to a vegetarian diet. Most nonpharmacological trials in subjects with mild hypertension have focused on the effect on blood pressure of reduction in salt or potassium intake. Dodson et al (5) showed a blood pressure lowering effect of a low-fat, low-salt, and high-fiber diet in hypertensive diabetic patients and so far as we are aware this J C/in

fiber

is the closest

Introduction

Am

ratio,

Nuir

l988;48:80l-S.

Printed

in USA.

© 1988 American

Society

Dodson

I From the tive Medicine, tre, Nedlands, of Medicine,

NHMRC Research Unit in Epidemiology and PrevenUniversity Department ofMedicime, QEII Medical CenWestern Australia 6009 and the University Department Royal Perth Hospital, Perth, Western Australia 6000,

Australia. 2

Reprints

for aimial

not available.

Nutrition

801

Downloaded from www.ajcn.org by guest on July 10, 2011

KEY

802

MARGETTS

ET AL 160

TABLE I Study design .

.

.

Familiarization period

Group

155

Test period

.

I

SYSTOLIC

2

BP 1 2 3

Usual Usual Usual

Duration

2 wk

C

Test

diets

used:

Study

diet diet diet

Usual diet Test diet Usual diet

Usual diet Usual diet Test diet

6 wk

6 wk

1 , ovolactovegetariam

subjects; study 2, high-P:S-ratio diet study 3: high-fiber diet in normotemsive

150

mmHg 145

diet in hypertensive

in normotensive subjects.

subjects;

140 and 100

TABLE Dietary

2 guidelines

for the vegetarian-diet

DIASTOLIC 95.

BP mmHg

90

85’ FIG I. Vegetarian diet in untreated mild hypertension: change in mean blood pressures by group during the trial. Key: #{149} Group 1,#{149} Group 2, and A Group 3. Broken line indicates period on vegetarian diet.

not clear and could not be determined. We completed randomized trials of the effect on blood pressure in normotensive subjects of changing the P:S ratio and fiber intake alone and these results will be summarized in the following sections ofthis review. was

Ratio

of polyunsaturated

to saturated

fatty

acids

There have been a number of recent reviews of the effect of fat intake on blood pressure (7-9). Although most trials have tended to show a decrease in blood pressure associated with either a reduction in total fat and/or elevation ofthe P:S ratio, it has not always been possible to determine whether this decrease was due to the dietary change under investigation or some other confounding factor (dietary or other). Iacono and Puska completed a number of trials in which the dietary P:S ratio was increased either by lowering the saturated fatty acid intake (10, 1 1) or by increasing the polyunsaturated fatty acid intake (12, 13). Mean

trial

Guidelines Avoid eating meat, fish, and poultry Use only whole-grain cereals and bread Double the fruit intake Increase vegetable consumption Use only spreads and cooking oils composed of polyunsaturated fatty acids Do not alter salt intake or egg and cheese consumption Do not alter usual total energy intake

TABLE Percent

3 ch anges

in nutrient

Nutrient P:Sratio VitaminB-12 VitaminC Fiber Protein

intakes

on changing

to a vegetarian

Percent

change +80 -62 +56 +35 -21

diet


tices were maintained during other periods. Blood pressures were measured every 2 wk throughout the study by use of an automatic recorder (DINAMAP). Dietary records and 24-h urine specimens were collected during the test periods primarily to check compliance with the dietary guidelines. Base-line characteristics were similar in each group. When subjects ate a vegetarian diet, systolic but not diastolic blood pressure fell significantly both in absolute terms and relative to change in groups not on the vegetarian diet (Fig I). The dietary P:S ratio, vitamin C, and fiber intake increased substantially and intake of protein and vitamin B-l2 fell substantially during the vegetarian diet (Table 3). Urinary 3-methylhistidine excretion, a marker of meat intake, also fell significantly when subjects changed to the vegetarian diet. Change in blood pressure was not significantly related to change in body weight or salt intake. Change in other possibly confounding factors, such as salt use, exercise, smoking, alcohol use, and weight, were monitored during the study and when included with dietary factors in a multiple-regression analysis they did not alter the significant association found between consumption of a vegetarian diet and reduction in blood pressure. It was concluded that an ovolactovegetarian diet could lower the systolic blood pressure of subjects with mildly elevated blood pressure. This confirmed the effect previously reported in normotensive subjects. The dietary change primarily responsible for the blood pressure effect

VEGETARIAN TABLE 4 Home blood

pressure

measurements

averaged

over the whole

DIET

IN MILD

day during

HYPERTENSION

high-P:S-ratio

803

trial Group

1 (n=

Systolic blood pressure (mm Hg) End familiarization period Endtestperiodl Endtestperiod2 Diastolic blood pressure (mm Hg) End familiarization period End test period 1 Endtestperiod2 C

j: SD. Period

on high-P:S-ratio

2 (n=20)

16)

3 (n=

18)

1 13.2 ± 12.0 113.9±11.2 105.3 ± 19.8

1 13.0 ± 10.1 110.5±13.0 108.1 ± 14.2

107.9 ± 10.9 108.8±11.1 106.2 ± 11.3

72.0 ± 8.5 72.4 ± 7.7 66.7± 11.6

74.9 ± 14.8 7 1.2 ± 1 1.0 73.1 ± 8.5

70.8 ± 8.2 68. 1 ± 8.2 72.2± 12.0

diet underlined.

ther 7-d diet record was kept. Blood and 24-h urine samples were also collected at the end of the familiarization period and the end of test periods 1 and 2. Plasma fatty acids were measured to assess compliance with the dietary guidelines. Base-line levels ofdietary intakes, blood pressures, and blood and urinary measures were not statistically significantly different between groups. When subjects were on a high-P:S-ratio diet, mean levels of major nutrients, except the type of fat, were not substantially different from base line. When subjects were on the high-P: S-ratio diet, P:S ratios rose to 1 .0 and dietary change was confirmed by changes in plasma fatty acids. Home blood pressure measurements (Table 4) were not significantly lowered relative to other groups when subjects were on the high-P:S-ratio diet. The pattern seen here with home blood pressure measurement was similar to that previously reported by us (10) for blood pressures measured at our research clinic. Adjusting blood pressure changes for change in other possibly confounding variables (such as plasma and urinary electrolytes and lifestyle changes) by multiple-regression analysis did not uncover any relationship between dietary P:S-ratio change and blood pressure. The results of this study do not support the view that the blood pressure-lowering effect of a vegetarian diet is due solely to an elevation in the dietary P:S ratio. The aim ofthe present study was to keep total fat intake and intake of all other nutrients constant. This was achieved and it is therefore possible that change in P:S ratio may affect blood pressure when, for example, total fat intake ‘

is reduced

at

the

same

time.

It is also

possible

that

the

base-line level of P:S ratio (0.3) in our subjects was already above the level that produces a blood pressure reduction. Vegetarians normally have a P:S ratio of --0.8, considerably above 0.3 (the usual level). If there is a threshold below 0.3 beyond which blood pressure change does not occur, then increasing the P:S ratio to 1 .0 (or 0.8) would not be expected to have any further effect. However the blood pressure reduction achieved by


blood pressures fell in all these studies. These dietary P:S ratio changes were however also associated with other possibly confounding dietary changes. More recently these workers showed that a reduction in total fat intake, irrespective of the dietary P:S ratio, lowered blood pressure (14). With respect to a vegetarian diet, typical vegetarians not only eat less total fat but they also eat more polyunsaturated fatty acids. Some workers (10) suggested that a high P:S ratio diet lowers blood pressure by altering prostaglandin metabolism although up to the present there is little evidence in humans to support this suggestion. We (15) recently completed a randomized controlled trial of the effect on blood pressure of increasing the P:S ratio in the diet of normotensive volunteers. The study design was, again, as summarized in Table 1. Fifty-four healthy normotensive volunteers aged 20-59 y were randomly allocated to groups so that they were matched by sex and 5-y age group. The aim of the dietary intervention was to increase the P:S ratio while maintaining the rest of the diet, including total fat intake, at preexperimental levels. All subjects were placed on a low-fat diet by removal of all visible fat from the diet. This quantity of fat was then replaced as a supplement in a doubleblind fashion so that the P:S ratio was 1 .0 when on the high-P:S-ratio diet or 0.3 (usual level) when on the usual diet. The supplement consisted ofa margarine and biscuits in which the predominant fatty acids were palmitic acid and stearic acid (low-P:S-ratio diet) or linoleic acid (high-P:S-ratio diet); no fish oils were used. Blood pressures were measured in several different ways; here we describe the blood pressures measured at home by the subjects themselves using an automatic reader (Ueda). Subjects were given careful instructions about how to record their blood pressures; recordings were made after 5-mm seated rest twice first thing in the morning, twice at lunch time, and twice in the evening. The values in Table 4 represent averages over the whole day. Before the familiarization period subjects kept an estimated 7-d dietary record. During each test period a fur-

804 change change

MARGETTS

to a vegetarian diet still could in dietary P:S ratio.

not be ascribed

to

ET AL TABLE S Blood pressure

change

during

high-fiber

trial Change period

Fiber Vegetarians population. ally shown

Systolic blood pressure (mm Hg) Test effectt Period effect Diastolic blood pressure (mm Hg) Test effect Period effect *

Groups

1 and 2 combined,

calculated

1.37 (+0.52 to -2.96) -2.53 (-0.5 1 to -3.99) -

-

1.06 (+0.54 to -2.6) 1.82 (-0.2 1 to -3.4)

by method

ofHillsand

Arm-

itage(lS).

t Effect ofhigh-fiber t Mean

change

diet, groups 2 and 3 only. with 95% confidence interval about

change

in paren-

theses.

and magnesium) rose significantly. In those on the low-fiber diet, fiber intakes remained low throughout. Multiple-regression analysis was used to determine if there was any independent effect of high-fiber intake on blood pressure after adjustment for possible confounding effects ofblood and urinary electrolytes as well as lifestyle variables(like exercise and alcohol intake)that may have also changed during the trial. No variables in the analysis accounted for any significant proportion of the variation in blood pressure. The present findings offer no support for the view that a vegetarian diet affects blood pressure by virtue of its higher fiber content. General

conclusions

An

ovolactovegetarian diet can lower the systolic ofthose with mild hypertension. Two of the notably largest dietary changes that occur on introduction of a vegetarian diet are increases in the dietary P:S ratio and fiber intake. When intakes of these nutrients were increased (in isolation from other dietary changes) to levels seen in vegetarians, no effects on blood pressure attributable to the change were seen. It is possible that a combination ofthese factors may cause blood pressure to fall or that some other aspect of the diet is responsible for the effect on blood pressure reduction. B blood

pressure

References 1. Wilkins JR. Calabrese FJ. Health implications of a 5 mmHg increase in blood pressure. In: Calabrese EJ, Tuthill RW, Condie L, eds. Imorgamics in drinking water and cardiovascular disease. Princeton: Princeton Scientific Publishing Co, 1985:85-100. 2. Rouse IL, Beilim U. Vegetarian diet and blood pressure. J Hypertension l984;2:23 1-240 (editorial review). 3. Beilin U, Margetts BM. Vegetarian diet and blood pressure. In: Blaufox MD, Lamgford HG, eds. Bibliotheca cardiologica-mon pharmacologic therapy of hypertension. Vol 41. Basal: Karger, 1987:85-105.


have a higher fiber intake than the general Cross-sectional studies (16-1 8) have generno relationship between fiber intake and blood pressure. To date there have been relatively few studies that have assessed the effect on blood pressure of altering the fiber intake alone (19-23). Studies by Kelsay et al (1 9), Stasse-Wolthuis et al (20), and Fehily (23) did not show any effect on blood pressure of elevating the fiber intake whereas studies by Wright et al (21) and Anderson (22) did show a blood pressure-lowering effect of a high-fiber diet. In all these studies it is difficult to isolate the effect ofthe fiber effect from other possibly confounding variables. It is unclear how these studies relate to the effect on blood pressure ofthe vegetarianfiber complex, a very complex mix of many different components each with different physiological effect (24). We (25) conducted a randomized controlled trial where we altered the usual fiber intake of typical omnivores to that eaten by vegetarians to assess whether this fiber complex explains the lower blood pressure in vegetarians. Eighty-eight healthy normotensive volunteers aged 20-59 y were recruited for the study. Subjects were randomly allocated to groups matched by sex and 5-y age group. All subjects were given the same dietary guidelines. They were placed on a low-fiber diet and told to eat specially prepared biscuits so that the fiber content was either that of typical omnivores (low 1 5 g/d) or that oftypical vegetarians (-50 g/d). The biscuits contained white flour, sugar, and milk (low-fiber diet) or bran, whole-meal flour, banana, apricot, sugar, and milk (highfiber diet). In addition, when on the high-fiber diet, subjects were given 100 g dried apricots to eat each day. As in the previous studies dietary intake was assessed before the study began and also during each test period. Blood pressures were measured every 2 wk throughout the study with an automatic recorder(the DINAMAP). Here we have compared only the values in the two test groups taken at the end of each test period. Change in these blood pressures was assessed using the method of Hills and Armitage (26). This method assesses change on the test diet in relation to any period effect occurring in the group not on the test diet. This allows an assessment of the independent effect of the high-fiber diet on blood pressure. Base-line data were comparable in each group. Blood pressure fell (not statistically significantly) when subjects ate a high-fiber diet but this test effect was similar to the decrease in the group (period effect) not on the high-fiber diet (Table 5). That is, the blood pressure decrease could not be attributed to the high-fiber diet. When subjects were on the high-fiber diet, estimated intakes of dietary fiber (and coincidentally potassium

from end test 1 to end test period 2*

VEGETARIAN

DIET

IN MILD

4. Beilin LI. Vegetarian approach to hypertension. Cam J Physiol Pharmacol l986;64:852-5. S. Dodson PM, Pacy PJ, Bal P, et al. A controlled trial ofa high fibre, low fat and low sodium diet for mild hypertension in type 2 (mominsulin dependent)diabetic patients. Diabetologia l984;27:522-6. 6. Margetts BM, Beilin U, Vandongen R, Armstrong BK. Vegetarian diet in mild hypertension: a randomised controlled trial. Br Med J l986;293: 1468-71. 7. Weinsier RL, Norris D. Recent developments in the etiology and treatment of hypertension: dietary calcium, fat, and magnesium. AmJ aim Nutr l985;42:l33l-8. 8. Margetts BM. Recent developments in the etiology and treatment ofhypertension: dietary calcium, fat, and magnesium. Am J aim Nutr l986;44:704-5. 9. Margetis BM, Beilin LI, Armstrong BK. Fat, fibre and carbohydrate and blood pressure. In: Wahlqvist ML, Truswell AS, eds. Recent advances in clinical nutrition 2. London: John Libbey, 1986: 21-9.

805

15. Margetts BM, Beilim LI, Armstrong BK, et al. Blood pressure and dietary polyunsaturated and saturated fats: a oomtrolled trial. aim Sci l985;69:l65-75. 16. Morris JN, Marr JW, aaton DO. Diet and heart: a postscript. Br MedJ l977;2:1307-l4. 17. Silman AJ. Dietary fibre and blood pressure. Br Med J 1980; 1:2S0. 18. Fehily AM, Milbank JE, Yarnell JWG, et al. Dietary determinants of lipoproteins, total cholesterol, viscosity, fibrinogen and blood pressure. Am J Gin Nutr 1982; 36:890-6. 19. Kelsay JL, Behall KM, Prather ES. Effect of fiber from fruits and vegetables on metabolic responses ofhuman subjects. I Bowel tramsit time, number ofdefecations, fecal weight, urinary excretions of energy and nitrogen and apparent digestibilities ofemergy, nitrogen amdfat. AmJClin Nutr l979;3l:l 149-53. 20. Stasse-Wolthuis M, Albers HFF, van Jeveren JOC, et al. Influence ofdietary fiber from vegetables and fruits, bran or citrus pectin on serum lipids, fecal lipids, and colonic function. Am J Cam Nutr l980;33: 1745-56. 21. Wright A, Burstyn P0, Gibney Mi. Dietary fibre and blood pressure. BrMedJ 1979;2:l54l-3. 22. Anderson JW. Plant fiber and blood pressure. Ann Intern Med l983;98:842-6. 23. Fehily AM, Burr ML, Butland BV, Eastham RD. A randomized controlled trial to investigate the effect ofa high fibre diet on blood pressure and plasma fibrinogen. J Epidemiol Community Health l986;40:334-7. 24. Stasse-Wolthuis M. Influence ofdietary fibre on cholesterol metabolism and colonic function in healthy subjects. World Rev Nutr Diet 198l;36:lOO-40. 25. Margetis BM, Beilim U, Vandongen R, Armstrong BK. A randomised controlled trial ofthe effect ofdietary fibre on blood pressure. Clin Sci l987;72:343-S0. 26. Hills M, Armitage P. The two-period cross-over clinical trial. Br J aim Pharmacol 1979;8:7-20.


10. IaconoJM, Judd JT, Marshall MW, et al. The role ofdietary essential fatty acids and prostaglandins in reducing blood pressure. In: Holman RT, ed. Progress in lipid research. Vol 20. Oxford: PergamonPress, 1981:349-64. 1 1. Puska P, Nissinen A, Vartiaimen E, et al. Controlled randomised trial ofthe effect ofdietary fat on blood pressure. Lancet 1983; 1: 1-S. 12. Iacomo JM, Puska P, Dougherty RM, et al. Effect of dietary fat on blood pressure in a rural Finnish population. Am J aim Nutr l983;38:860-9. 13. Pietinen P. Dougherty R, Mutanen M, et al. Dietary intervention study among 30 free-living families in Finland. J Am Diet Assoc l984;84:313-8. 14. Puska P, Tuomilihto J, Nissinen A, et al. Dietary fat and blood pressure: an intervention study on the effects ofa low fat diet with two levels ofpolyunsaturated fat. Prey Med 1985; 14:573-84.

HYPERTENSION