Vegetarian diet and blood pressure levels: Incidental or causal association?

Vegetarian or causal Lawrence

Barrie

diet and blood association?3

J Beilin,

MD;

M Margetts,

Ian

PhD;

L Rouse,

and Robert

PhD;

pressure Bruce

levels:

KArmszrong,

Vandongen,

MB

incidental BS,

DPhil

Oxon;

MD

blood

pressure-lowering

effect

ofa

lactoovovegetarian

diet

with

reversible

changes

of 5-6

mm

Hg systolic and 2-3 mm Hg diastolic occurring over 6-wk periods. Similar dietary effects in mild hypertensive subjects provides impetus for identifying the responsible nutrients. Am JClin

Nutr

KEY

WORDS

l988;48:806-10. Diet,

hypertension,

blood

The possibility that eating a vegetarian diet might be responsible for lowering blood pressure was first suggested in a study in 1926 (1) in which Donaldson asked vegetarian college students to eat meat and observed a blood pressure rise within 2 wk of their doing so. However, as with many other studies of diet and blood pressure, the experiment was uncontrolled and may have been confounded by the stress experienced by students encouraged to eat meat contrary to their beliefs. Subsequently reports describing lower blood pressures in acculturated vegetarians compared with meat-eating populations (2) have included populations as diverse as Benedictine and Trappist Monks, American macrobiotic vegans (3), Seventh-day Adventist vegetarians in the United States (4) and Australia (5-8), and vegetarians in Israel (9). It will be clear from even these reports that the term vegetarian embraces a wide variety ofeating habits, lifestyles, and social environments. All the groups referred to come from so-called acculturated backgrounds and differ in at least one important aspect other than social milieu from nomadic unacculturated vegetarians whose diets are often characterized by a very low intake ofsodium (10). Even acculturated vegetarians vary widely with respect to diet, social environment, and lifestyle, eg, macrobiotic vegans live in communes and eschew dairy products as well as meat, fish, and poultry; Seventh-day Adventists 806

Am

iC/in

Nuir

pressure,

vegetarian

are more aptly described as lactoovovegetanans (ie, consume dairy products but not meat, fish, or poultry) but also avoid alcohol, caffeine, and nicotine; whereas Israeli vegetarians show no such obligate restrictions but nevertheless tend to drink and smoke less, exercise more, and are less obese than their age-matched meat-eating controls (9). Given the now well-established blood pressureraising effect of obesity and alcohol consumption (11, 12) and the blood pressure-lowering effects ofregular exercise (1 3), as well as the more immediate effects of mental state on blood pressure, it is doubtful whether any of the earlier studies on vegetarians’ blood pressures provide conclusive evidence as to the effect of diet per se rather than effects ofother differences in lifestyle. One interesting exception to the reports oflower blood pressures in vegetarians was the study ofHarris et al (14) in which 3 159 children aged 6-16 y attending Seventhday Adventist schools in California showed blood pressures similar to 4681 children from other schools. The I From the University Department ofMedicine, University of Western Australia Medical School, Royal Perth Hospital, Perth, and the NHMRC Research Unit in Epidemiology and Preventive Medicine, University of Western Australia Queen Elizabeth II Medical Centre, Nedlands, Western Australia, Australia. 2 Supported by the National Health and Medical Research Council of Australia, the Australian Kidney Foundation, and the Royal Perth Hospital Research Foundation. 3 Reprints not available.

1988;48:806-lO.

Printed

in USA.

© 1988 American

Society

forClinical

Nutrition

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

ABSTRACI’ Evidence that nutrients other than the major cations may influence blood pressure levels stems from studies ofacculturated vegetarians and from randomized controlled dietary trials. Earlier studies of vegetarians focused on religious groups and on vegans, making it difficult to know whether their lower blood pressures were due to diet per se or to other aspects oflifestyle. Seventh-day Adventist vegetarians showed significantly less hypertension and lower blood pressures compared with Mormon omnivores, effects which were independent ofdifferences in obesity and not due to altered sodium intake. Subsequently, controlled dietary intervention studies in healthy normotensive omnivores provided more direct evidence for a

VEGETARIAN

(a)

180

DIET

(b)

In

170

160 32’

150

a.,

-f,, 140

130

;

120

4’; ::3j

50J59

o-g

70-79

30-39

FIG 1 . Blood pressure omnivores. Data for men tarians; ---, Narrogin and number in each age ence 7.

40-49

50-59

60-69

-79

(YEARS)

(BP)by age in SDA vegetarians and Narrogim are shown in a, for women in b.-SDA vegeomnivores. Results are mean with SEM and diet group is shown. Adapted from refer-

authors suggested that the potential effects of diet or other aspects oflifestyle might not become evident until later life. Some indication of a specific vegetarian-vs-meat dietary effect came from a brief report by Anholm et al in 1975 (15)(abstract only) ofa comparison ofblood pressures in Seventh-day Adventists lactoovovegetarians vs Mormon omnivores; Anholm et al found lower blood pressures in the vegetarians with the differences increasing with age. However no specific details ofpossible confounding factors such as obesity or salt intake were given. Mormons and Seventh-day Adventists are in several ways excellent controls for one another in that both groups avoid alcohol, caffeine-containing beverages, tobacco, and other stimulants and are deeply religious, but their diets contrast markedly with respect to meat, fish, and poultry. As a result ofthis report and the observations of Armstrong et al (7) that West Australian Seventh-day Adventist vegetarians attending a conference had lower blood pressures and less of a blood pressure rise with age than did the local townsfolk (Fig 1), Rouse et al (8) undertook a detailed comparison of blood pressures and lifestyles of Seventh-day Adventists and Mormons in Perth (Table 1). The object ofthis study was to try and determine the relative importance ofdiet and other aspects oflifestyle in determining the cause of any blood pressure differences between the two groups. All the subjects screened completed a lifestyle questionnaire and had blood pressures measured under carefully standardized conditions. Comparisons were made between 98 Seventh-day Adventist (SDA) lactoovovegetarians (eating meat, fish, or poultry less than once a month)

PRESSURE

807

with 82 (SDA) omnivores and 1 1 3 Mormon omnivores, aged 25-44 y. The SDA vegetarians were significantly slimmer than the Mormon omnivores but mean blood pressure adjusted for age, height, and weight were significantly lower in the vegetarians than meat eaters ( 1 15.6/68.7 and 12 1.2/72.2 mm Hg, respectively, in males and 109. 1/66.7 and 1 14.9/72.6 mm Hg, respectively, in females). These differences were not related to past or current use of alcohol, tobacco, tea and coffee, physical activity, personality characteristics, or degree of religious observance. The entire frequency distribution ofblood pressures was shifted to the left in the SDA vegetarians (Fig 2) and the prevalence of mild hypertension ( 140 mmHg systolic or 90 mmHg diastolic) was 10 and 8.5% in the Mormon and SDA omnivores, respectively, compared with 1-2% in SDA vegetarians. Analysis ofdiet records showed that the vegetarians ate significantly more fiber, polyunsaturated fatty acids, magnesium and potassium, vitamin C, and vitamin E and significantly less total fat, saturated fatty acids, cholesterol, Vitamin B-l2 and iron compared with Mormon omnivores. The SDA omnivores had dietary patterns and blood pressures that were intermediate to those of the SDA vegetarians and Mormon omnivores. In a detailed follow-up study of 47 randomly selected pairs of Adventist vegetarians and Mormon omnivores matched for sex, age, and obesity, home blood pressure readings showed consistent blood pressure differences (Table 2) (16). Moreover mean 24-h urinary excretion of Na was similar in the two groups although the vegetarians excreted significantly more K. These results with urinary Na are in accord with those of Arkwright et al (11) and indicate that the blood pressure differences were unlikely to be due to dietary Na. Although these comparisons between Seventh-day Adventist vegetarians and Mormons provide support for a specific blood pressure-lowering effect of a vegetarian diet, cross-sectional population studies cannot prove cause-and-effect relationships, and it is possible that other unidentified factors were contributing to the blood pressure differences. However the issue was resolved beyond reasonable doubt by randomized controlled dietary intervention trials in Perth that involved introducing a vegetarian diet to groups of regular meat eaters. In the first ofthese trials 59 healthy volunteer Hospital employees were assigned to one ofthree dietary groups after matching for age, sex, and obesity (17). After an initial familiarization period the first group in the trial continued their normal meat diet for two consecutive 6-wk penods; the second group ate a vegetarian diet for 6 wk and then reverted to their normal diets for 6-wk and the third group ate these diets in the reverse order. The volunteers had their two main meals a day prepared in the hospital canteen for 5d/wk and for all other meals were given specific advice about a lactoovovegetarian diet typical of that eaten by the West Australian Seventh-day Adventists. Subjects were also encouraged to maintain other aspects of lifestyle constant for the duration of the study, including total calorie consumption, physical activity,

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

AGE

AND BLOOD

808

BEILIN

TABLE 1 Twenty-four-hour

nutrient

intake

ofAdventist

vegetarians:

Adventist

ET AL

ommivores

and Mormon

omnivores

Men Adventist vegetarians Energy (MJ) (kcal) Totalcarbohydrate(g)

10.96 ± 3.76 2619±8.99 353±126 44.3± 13.3 9 1.7 ± 32.8 99.8 ± 42.6 33.7 ± 17.5 25.1 ± 18.1 0.83 ± 0.54 191±195 162±73 112±50 21.7 ± 6.6

Dietaryfibre(g)

Protein (g) Total fat (g) Saturated fatty acids(g) Polyunsaturatedfattyacids(g) P:S ratio Cholesterol(mg) Sodium(mmol) Potassium(mmol) Magnesium (mmol) SD. From

t Significantly Significantly

reference different different

10.50 ± 2.71 2510±648 332±91 41.4± 17.5 92.7 ± 26.3 95.5 ± 33.3 37.2 ± 16.7 18.6± 11.5 0.57 ± 0.36f 272±2l7f 163±73 112±38 20.6 ± 8.1

Mormon ommivores

Adventist vegetarians

1 1.62 ± 4.29 2777±1025 321±116 24.3± l6.lt 103.6 ± 39. 1 126.9 ± 57.4 50.3 ± 25.3t 18.4± l4.2f 0.4 1 ± 0.36t 398±2llt 155±67 92±37t 15.5 ± 7.9t

8.55 ± 2.63 2040±629 261±87 32.6± 11.9 72. 1 ± 24.0 83.2 ± 34.8 29.2 ± 13.7 19.0± 12.7 0.77 ± 0.62 208±159 131±56 88±29 16.3 ± 5.5

Adventist ommivores

Mormon omnivores

8.00 ± 2.75 1910±657 240±88 27.1 ±8.9 66.9 ± 24.4 79.9 ± 35.6 31.3 ± 16.5 l3.7±8.3 0.48 ± 0.28 230±166 l09±53f 80±26 14.4 ± 4.7

8.43 ± 3.32 2010±793 224±93 19.7± lO.6t 79.4 ± 29.5 92.2 ± 42.2 35.4 ± 19.7 14.2± l0.7 0.46 ± 0.35t 308±161 76±33 1 1.9 ± S.2t

8. from intakes from intakes

in vegetarians, in vegetarians,

p < 0.002 (protected p < 0.05 (protected

salt intake, alcohol drinking, and smoking habits. Mean systolic and diastolic blood pressures fell significantly during the vegetarian periods in both experimental groups and rose again in the group which reverted to the omnivorous diet (Fig 3) whereas the control group showed no change after the initial familiarization period. Adjustment of the blood pressure changes for age, obesity, heart rate, weight change, and initial blood pressure showed a reversible diet-related decrease of the order of 5-6 mmHg systolic and 2-3 mmHg diastolic. Again, the blood pressure changes were unrelated to urinary Na cxcretion or estimated dietary Na intake. Compliance with

z

least significant least significant

difference difference

test between log-transformed test between log-transformed

intakes). intakes).

the diets

was confirmed by measurement of urinary 3methylhistidine, an amino-acid found predominantly in meat protein. Twenty-four hour dietary diaries obtained each week indicated nutrient changes (Fig 4) corresponding to the diets ofthe Seventh-day Adventist vegetarians(l8). Taken in conjunction with the epidemiological data, this intervention study provides strong cvidence for a direct blood pressure-lowering effect of a lactoovovegetarian diet. One possible confounding factor in this trial was that the subjects were not blind to the dietary changes. Thus theoretically the vegetarian diets may have exercised a placebo-like effect on blood pressures. Although such an effect cannot be excluded, the eagerness with which subjects resumed a meat diet suggests greater stress during the vegetarian periods and mitigates against this possibility. Margetts et al (19) in Perth subsequently demonstrated a blood pressure-lowering effect of a vegetarian diet in untreated mild hypertensive subjects with a study design similar to that described above. The results of this

I TABLE Average

2 home

blood

pressures

z

in

vegetarians

and omnivores*

Men

Systolic A$T0uc seu

Omnivores

(n=23)

(n=23)

111.9±7.0 70. 1 ± 6.7

Diastolic

Women

Vegetarians

Vegetarians

Omnivores

(n24)

(n-24)

98.2±7.3 65.8 ± 8.2

116.1 ±7.3t 71.2 ± 8.6

103.8±9.Ot 73.4 ± 6.9f

sITouc (..-..)

-

S

Average

ofduplicate

determinations

(on rising, at lunch time, on arriving FIG 2. Cumulative frequency distributions ofsystolic and diastolic blood pressure (BP) in SDA vegetarians (-) and Mormon omnivores (---) aged 25-44 y. For study details see Rouse et al (8).

arei±

made

on

home from

four

work,

occasions

during

and on retiring).

the

day

Results

SD.

t 5ignificant difference between omnivores

and

vegetarians,

p <

0.05,

1 test.

* Significant

and

vegetarians,

p

0.01,

: test.

difference

between

omnivores

<

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

6

Adventist omnivores

Women

VEGETARIAN

DIET AND BLOOD

0 .2

-2 -4

1

,

C)

:? :

0

E [menta

Period

1

Experimental

Period

2

809

marker of another unidentified dietary component influencing blood pressure. It will be seen from other papers in this symposium that there is still considerable uncertainty as to the nutrients responsible for lower blood pressures in vegetarians. Of some interest is another recent study from Perth in the form ofa randomized double-blind crossover trial of the effects ofmeat vs vegetable protein, with other nutrients held constant. No blood pressure-lowering effect of meat protein could be detected over a l2-wk period in normotensive subjects (20). In summary, there is now strong evidence for a blood pressure-lowering effect of a lactoovovegetarian diet in acculturated subjects; the effect is independent ofNa and energy intake and of other aspects of lifestyle that tend to characterize vegetarian populations. The effect may be important in terms of seeking means of preventing and treating hypertension. The subject warrants further

PROXIMATE

A

>.

Q.L&.

trial and other studies designed to identify dietary components responsible for these blood pressure differences will be reported elsewhere in this symposium. One of the main reasons for difficulty in pinpointing dietary components that are responsible for blood pressure differences between meat eaters and vegetarians is that the vegetarian diets differ from omnivore diets in many respects other than consumption of meat, fish, or poultry. Some of the differences in fiber, fat, minerals, and vitamins were alluded to above and were described in detail elsewhere (7, 17). There are likely to be further substantial differences within and between vegetarian groups with seasonal and local market factors operating. Moreover given the high degree ofcovariability of foods and nutrient intakes when diets are changed to and from vegetarian, it becomes extremely difficult to isolate any one component that might be responsible for the effects on blood pressure. Nevertheless an attempt was made to do this in Rouse et als intervention trial of Rouse et al (18) in normotensive subjects, by principal components or factor analysis. When changes in nutrient intake were examined in this way, three main factors were identified and the resultant factor scores were used as independent predictors ofchanges in blood pressure. Eight-three percent of the variation in nutrient intake could be accounted for by these three factors but only one was sig-

-soS

nificantly

-25

associated

with

blood

pressure

changes.

This

factor represented increase in polyunsaturated fatty acids, fiber, vitamin C, vitamin E, calcium, and magnesium and a decrease in the intake ofprotein. Some caution needs to be exercised in interpreting this analysis because although it may provide some clue as to the nutrients responsible for blood pressure changes, it is equally possible that the strength of the association with these nutrient factors reflects compliance with the vegetarian diet overall; alternatively, the factor may merely be a

I

Co

WTRIENTS

100

is ,,_

< U.

50

U. .

.<

I-

<

I

25

C,)

/ -z

-N;-

-25

a MNERALS

and

VITAMNS

o

100’

75

50

Ms

I

-50 *

-is FIG 4. Percent change in nutrient intakes with change to vegetarian diet. < 0.001, significance of the difference in nutrient intakes on the vegetarian and omnivorous diets, paired z test, n = 36. NS = not significant. From Rouse et al (17).

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

FIG 3. Mean changes in resting blood pressure(BP) measured in the laboratory. #{149}, control group; Y, experimental group 1; U, experimental group 2; ---, vegetarian diet; -, omnivore diet. Vertical bars indicate SEM. P < 0.05 and ** < 0.01 for significance of difference in respomse ofexperimental groups in comparison with control group (protected least-significant-difference test). From Rouse et al (17).

PRESSURE

810

BEILIN

References 1. Donaldson AN. The relation of protein foods to hypertension. CalifWest Med l926;24:328-3l. 2. Salle F. Influence ofvegetarian food on blood pressure. Med aim 198 1;26:929-3 1. 3. Groen JJ, Tijong KB, Koster M, et al. The influence of nutrition and ways oflife on blood cholesterol and the prevalence of hypertension and coronary heart disease among Trappist and Benedictime monks. Am J am Nutr 1962; 10:456-70. 4. Phillips RL. Role oflife-style and dietary habits in risk of cancer among Seventh-day Adventists. Cancer Res 17333-22. S. Webster

1W,

Rawson

OK.

Health

status

of Seventh-day

Adven-

tists. Med J Aust 1979; 1:417-20. 6. Simons L, Gibson 3, Jones A, Baim D. Health day Adventists. Med J Aust l979;2: 148.

status

of Seventh-

7.

Blood

pressure

Armstrong

Seventh-day 444-9.

B,

Van

Adventist

Merwyk

AJ,

Coates

vegetarians.

Am

H.

J Epidemiol

in

1977; 105:

8.

Rouse

IL,

Armstrong

BK,

Beilin

U.

The

pressure to diet and lifestyle in two religious tens l983;l:65-71. 9. Ophir 0, Peer 0, Gilad J, Blum M, Aviram in vegetarians:

the

possible

role

relationship

of blood

populations.

J Hyper-

A. Low blood

of potassium.

Am

pressure Nutr

J Clin

1983;37:755-62. 10. Page LB. Epidemiologic evidence on the etiology ofhuman hypertension and its possible prevention. Am Heart J l976;91:527-34. 1 1 . Arkwright PD, Beilim U, Rouse I, Armstrong BK, Vandongen R. Effects ofalcohol use and other aspects oflife-style on blood pressure levels and prevalence of hypertension in a working population. Circulation 1982;66:60-6. 12. Puddey lB. Beilin U, Vandongen R, Rouse IL, Rogers P. Evidence for a direct effect ofalcohol consumption on blood pressure in normotensive man-a randomised controlled trial. Hypertension l985;7:707-l3. 13. NeLson L, Jennings OL, Esler MD, Korner PL Effect of changing levels of physical activity on blood-pressure and haemodynamics in essential hypertension. Lancet 1986;2:473-6. 14. Harris RD. Phillips RL, Williams PM, KuzmaJW, FraserGE. The child-adolescent blood pressure study: 1 Distribution of blood pressurelevels in Seventh-day Adventist(SDA)and mon-SDA children. Am J Public Health 1981;7l:l342-9. 15. Anholm AC. The relationship of a vegetarian diet to blood pressure. Prey Med l975;4:35(abstr). 16. Rouse IL, Beilim LI, Armstrong BK, Vandongen R. Vegetarian diet, blood pressure and cardiovascular risk. Aust NZ J Med 1984; 14:439-43. 17. Rouse IL, Beilim U, Armstrong BK, Vandongen R. Blood pressure lowering effect ofa vegetarian diet: a oontrolled trial in normotensivesubjects. Lancet l983;l:5-lO. 18. Rouse IL, Beilin U, Mahoney DP, et al. Nutrient intake, blood pressure, serum and urinary prostaglandins and serum thromboxane B in a controlled trial with a lacto-ovo-vegetarian diet. J Hypertens 1986;4:241-50. 19. Margetts BM, Beilin U, Vandongem R, Armstrong BK. Vegetarian diet in mild hypertension: a randomized controlled trial. Br Med J 1986;293:

20.

1468-71.

Prescott SL, Jenner DA, Beilin U, Margetts B, Vandongen R. Isolated effects ofdietary protein on the blood pressure of normotemsives. Clim Exp Pharmacol Physiol(in press). 21. Masarei JRL, Rouse IL, Lynch WL, Robertson K, Vandongem R, Beilin U. Effects ofa lacto-ovo vegetarian diet on serum concentrations of cholesterol, triglyceride, HDL-C, HDL2-C, HDL3-C, apoproteim-B and Lp(a). Am J aim Nutr l984;40:468-79. 22. Rouse IL, Beilin U. Nutrition, blood pressure and hypertensiona critical review ofdietary intervention studies in man. Med J Aust 1983; 12:519-23. 23. Rouse IL, Beilin U. Vegetarian diet and blood pressure. J Hypertens l984;2:231-40 (editorial review).

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

study with particular respect to responsible dietary components and the pathophysiological mechanisms involved. Cardiovascular risk in general is low in people adhering to a lactoovovegetarian diet, not only because their blood pressure are lower and tend to rise less with age, but also because they carry less excess fat and tend to have healthier blood-lipid profiles than do meat eaters (16, 21). These effects are often further compounded by the tendency ofvegetarians to smoke less, drink less alcohol, and remain physically active. Future research attempting to analyze the effects of vegetarian diet on blood pressure and cardiovascular disease must pay careful attention to methodology to distinguish these factors and to avoid the pitfalls inherent in studying dietary changes (22, 23), which are themselves inevitably complex, and blood pressure levels, which are highly variable and susceptible to multiple influences. Other areas ofparticular interest for research includes possible synergistic effects of vegetarian diets with antihypertensive drug therapy in moderate to severe hypertension; the extent of interractions with other nonpharmacological approaches to blood pressure reduction, such as weight reduction, sodium restriction, alcohol moderation, and physical training; and interractions with fish intake and fish oils in terms ofpotential beneficial effects on blood lipids, blood pressure, cardiovacular risk, and outcome. B

ET AL