238
Blood Pressure in Four Remote Populations
in the INTERSALT Study
Jairo J. Mancilha Carvalho, Roberto G. Baruzzi, Peter F. Howard, Neil Poulter,
Michael P. Alpers, Laercio J. Franco, Luiz F. Marcopito, Veronica J. Spooner,
Alan R. Dyer, Paul Elliott, Jeremiah Stamler, and Rose Stamler
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
Four remote population samples (Yanomamo and Xingu Indians of Brazil and rural populations
in Kenya and Papua New Guinea) had the lowest average blood pressures among all 52
populations studied in INTERSALT, an international cooperative investigation of electrolytes
and blood pressure. Average systolic blood pressure was 103 versus 120 mm Hg in the
remaining INTERSALT centers; diastolic blood pressure in these four population samples
averaged 63 versus 74 mm Hg in the 48 other centers. There was little or no upward slope of
blood pressure with age; hypertension was present in only 5% of the rural Kenyan sample and
virtually absent in the other three centers. Also in marked contrast with the rest of the centers
was level of daily salt intake, as estimated by 24-hour urinary sodium excretion. Median salt
intake ranged from under 1 g to 3 g daily versus more than 9 g in the rest of INTERSALT
populations. Average body weight was also low in these four centers, with no or low average
alcohol intake, again unlike the other centers. The association within these four centers between
the above variables and blood pressure was low, possibly reflecting their limited variability.
While several other INTERSALT centers also had low average body weight or low prevalence
of alcohol drinking, when this was accompanied by much higher salt intake (7-12 g salt or
120-210 mmol sodium daily), hypertension prevalence ranged from 8% to 19%. These findings
confirm previous reports that in populations with a low salt intake, there is little or no
hypertension or rise of blood pressure with age. While the contributory role of other
characteristics of these populations must also be considered, the results are consistent with the
view that a certain minimum salt intake is essential for rise in blood pressure with age in adults
and a high frequency of hypertension in populations. {Hypertension 1989; 14:238-246)
ndividual epidemiological studies in several isolated populations have, over many years,
reported low average blood pressures and
little variation in blood pressure with age.1-13 An
opportunity to assess these observations in simultaneous investigations using standardized methods,
as well as to investigate possible explanatory factors,
was provided by the INTERSALT study. INTERSALT, a multicenter international cooperative crosssectional study focusing on relations between electrolytes and blood pressure, was comprised of 52
I
samples from 32 countries in North and South America, Europe, Africa, Asia, and the Pacific.14 Four
remote population samples (Yanomamo and Xingu
Indians of Brazil, rural populations in Kenya and
Papua New Guinea) had the lowest average blood
pressures among all populations studied in INTERSALT, as well as little or no increase of blood
pressure with age.
The present report describes the blood pressure
pattern and other characteristics of each of these
From the Hypertension Ambulatory Unit, Central Navy Hospital, Rio de Janeiro, Brazil (J.J.M.C); the Department of
Preventive Medicine, Escola Paulista de Medicina, Sao Paulo,
Brazil (R.G.B., L.J.F., L.F.M.); the Papua New Guinea Institute of Medical Research, Goroka, Papua New Guinea (P.F.H.,
M.P.A., V.J.S.); the Department of Community Medicine, University College and Middlesex School of Medicine, London,
England (N.P.); the Department of Community Health and
Preventive Medicine, Northwestern University School of Medicine, Chicago, Illinois (A.R.D., J.S., R.S.); Department of
Epidemiology and Population Sciences, London School of
Hygiene and Tropical Medicine, London, England (P.E.).
Supported by grants from the Wellcome Trust (United
Kingdom); the National Heart, Lung, and Blood Institute
(United States); the International Society of Hypertension; the
World Health Organization; the Heart Foundations of Canada,
Great Britain, Japan, and The Netherlands; the Chicago Health
Research Foundation; the FWGO-FMRS (Belgian National
Research Foundation); and the ASLK-CGER (Parastatal Insurance Co., Brussels).
Address for correspondence: Professor Rose Stamler, Northwestern University Medical School, Suite 1102, 680 N. Lake
Shore Drive, Chicago, IL 60611.
Received February 23, 1989; accepted April 24, 1989.
Carvalho et al Blood Pressure in Four Remote Populations
four samples, including urinary electrolytes, body
mass index (BMI), pulse, and alcohol.
Participants and Methods
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
Population Samples
The Yanomamo sample was composed of all
adults from three villages located on the Surucucu
plateau in northern Brazil. These villages were 8
hours walking distance from a Government Health
Station in the area. The Yanomamo Indians are one
of the most unacculturated native tribes in South
America and possibly in the world. They live in an
area of about 200,000 km2 located along the border
of Brazil and Venezuela. The total population consists of approximately 18,000 individuals, scattered
throughout the Amazon rain forest in 200 or so
villages composed of 40 to 250 people each. The
Yanomamo are generally seminomadic, slashand-burn agriculturalists who live on a diet of
locally produced crops and game supplemented by
wild fruits and insects. Dietary staples consist of
cooked banana and manioc (cassava). In most villages there is little if any access to salt, refined
sugar, alcohol, milk, or other dairy products.
The Xingu sample was obtained by random sampling from 10 native Indian tribes in the "Parque
Indigena do Xingu" (PK). The PEX occupies an
area of 22,000 km2 in the central region of Brazil.
The 10 Indian tribes involved in the study have their
own villages of approximately 150 inhabitants each.
Each tribe still preserves its own language and,
although less isolated than the Yanomamo, do not
maintain regular contact with non-Indians. The diet
of Xingu Indians is principally based on manioc
(cassava) and fish. The consumption of meat from
game animals is free in some tribes and restricted by
taboos in others. The diet also includes, to a lesser
extent, corn, sweet potatoes, cara, peanuts, bananas,
and wild fruit. Salt and other foods sold in stores are
not used regularly.
The sample from Papua New Guinea was obtained
by random sampling from the rural villages of
Kamus and Gimisave in the Asaro Valley of the
Eastern Highlands Province of Papua New Guinea.
The Kamus group consists of 569 persons living in
five small hamlets that are approximately a 60minute walk from the highlands highway and are
located at altitudes ranging from 6,000 to 6,600 feet
above sea level. The economy is based on smallholder coffee gardens and subsistence agriculture to
provide food. The main staple is the sweet potato.
Health services and a school for primary education
are provided in another village at a 30-minute
walking distance. Of the two groups, the Kamus are
more traditional than the Gimisave; they rely more
on subsistence agriculture and have fewer men in
paid jobs and fewer women who have attended
school. The Gimisave villages (population 594) are
closer to the highway and the inhabitants have
greater contact with a more modern lifestyle. There
239
is less reliance on subsistence agriculture, consumption of "store" foods, beer, and cigarettes is higher,
and betel nut chewing is more common than among
the Kamus group.
The Kenyan sample was obtained by random
selection of households. The population from which
the sample was drawn consisted of the 1,500 inhabitants and 320 households in the rural villages of
Rambugu and Ndori, which are located just north of
Lake Victoria in western Kenya. About 85% of
adults engage in subsistence farming and are exclusively from the Luo tribe.
Each center was asked to recruit 200 men and
women aged 20-59 years, 25 in each of eight age and
sex groups. In each center a number of participants
were excluded, most due to incomplete urine collection (based on self-reported incompleteness or
because total urine volume was less than 250 ml). A
small number were excluded because of pregnancy.
Data Collection
The Study's two Coordinating Centers, in London and Chicago, were responsible for development
of standardized procedures and training local investigators. Details of INTERSALT methods have
been published.15 Standardized techniques were
used for blood pressure measurement and for collection of 24-hour and casual urine samples.
For these four centers, interviews were conducted with the help of interpreters. Where participants did not know their ages, criteria for estimation of age were physical appearance, number and
age of children, personal knowledge of the interpreters, and calendars of local events. Two blood
pressure measurements were made, with the participant in a sitting position, by observers previously
trained and certified in INTERSALT procedures. A
random zero sphygmomanometer and the bell of a
Littmann stethoscope were used. The mean of the
two readings was the blood pressure of record.
Participants refrained from strenuous activity or
eating for at least 30 minutes before measurement;
they emptied their bladders and then sat quietly for
5 minutes. A range of blood pressure cuffs was
available to allow for differences in arm circumference. Systolic blood pressure (SBP) was recorded
as the appearance of sound and diastolic blood
pressure (DBP) as the disappearance of sound.
Height and weight were also measured twice, using
a stadiometer and a portable scale, calibrated daily.
For the 24-hour urine collection, standard 1 liter,
wide-mouthed plastic jars, containing boric acid as
preservative, were supplied (along with a funnel for
women). Participants were carefully instructed on
the need to collect all urine passed during 24 hours.
Aliquots of urine samples were refrigerated at 4° C
within 24 hours and frozen at -20° C within 7 days.
During transportation and shipment to the Central
Laboratory in Belgium the aliquots of urine were
maintained frozen inside styrofoam boxes containing dry ice.
240
Hypertension Vol 14, No 3, September 1989
TABLE 1.
Blood Pressure Patterns by Sex in Four Remote INTERSALT Populations
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
BP variable by sex
Systolic (mm Hg)'
Men
Mean
(SD)
Women
Mean
(SD)
Diastolic (mm Hg)*
Men
Mean
(SD)
Women
Mean
(SD)
Percent
hypertensivet
Slope of BP with age
(mm Hg/10 yr)t
Systolic
Diastolic
Yanomamo
(n = 195)
Xingu
(n = 198)
Papua New
Guinea
(n = 162)
Kenya
(n = 176)
101.3
(9.3)
103.3
(12.0)
109.6
(11.6)
114.3
(14.5)
90.6
(7.8)
96.2
(9.0)
106.4
(12.1)
107.9
(15.7)
64.7
(8.5)
65.2
(7.7)
64.4
(9.6)
67.0
(15.2)
56.5
(7.5)
59.2
(7.8)
61.0
(10.3)
64.7
(15.1)
0.0
1.0
0.8
5.0
-0.8
+0.6
+0.6
-0.4
-1.0
-0.4
+2.4
+ 1.3
'Values are mean±SD. BP, blood pressure.
tHypertension defined as systolic blood pressure a 140 or diastolic blood pressure &90 mm Hg.
^Corrected for body mass index and sex.
All electrolyte analyses were carried out by the
Central Laboratory, Department of Epidemiology,
St. Rafael University, Leuven, Belgium. Sodium
and potassium were analyzed by emission flame
photometry,16 chloride by the Cotlove method,16
calcium and magnesium by atomic absorption flame
photometry,17 and creatinine by the Jaffe method.18
In addition to internal laboratory controls, quality
control was also monitored in the London Coordinating Center by comparison of results on anonymous split samples.
Individual electrolyte excretion was computed as
the product of electrolyte concentration in the urine
and urinary volume corrected to 24 hours. BMI was
calculated as weight (kg) divided by height2 (m2).
Results
Descriptive Statistics
Blood pressure. Mean SBPs were low in all four
centers, particularly among the two Indian population samples in Brazil (101 and 103 mm Hg in men,
91 and 96 mm Hg in women) (Table 1). The combined average for the four centers was 103 mm Hg.
In contrast, mean SBP in the remaining 48 INTERSALT centers was 120 mm Hg.14 Mean DBPs were
also low in all four centers (57-67 mm Hg) (Table
1). Again, these levels were considerably lower
than the average observed in other INTERSALT
samples (74 mm Hg).14 In all four samples, with
average age approximately 39 years, mean levels
for men were higher than for women in both systolic
and diastolic pressure.
Hypertension, denned as SBP 140 mm Hg or
greater or DBP 90 mm Hg or greater, was com-
pletely absent among the Yanomamo, virtually
absent among the Xingu and Papua New Guineans
(0.8-1.0%) and low among the Kenyans (5%) (Table
1). The distribution of blood pressure values was
similar among the four centers (Figure 1), except
that in the Kenya sample, this distribution was
shifted upward compared with the others and, as
noted, there were a few persons (nine) in that
Diastolic
< 40 45 55 85 75 85 05 1OS 115 125 135
2/40
Blood Pressure (mm Hg)
Yanomamo Sample
Xingu Sample
Papua New Guinea Sample
Kenya Sample
FIGURE 1. Line graph of blood pressure distributions in
the four populations; percent at given levels of diastolic
and systolic pressure.
Carvalho et al
110
I 2
I *> 100
00
900)
It)
80
miimnmmn^^
50-
Urinary electrolytes. Mean 24-hour urinary
20-29
30-39
40-49
Age (Years)
50-59
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
• Yanomamo Sample
' Xingu Sample
Papua New Guinea Sample
Kenya Sample
FIGURE 2. Line graphs of systolic and diastolic pressures by 10-year age groups in the four populations.
sample at hypertensive levels. Variation in blood
pressure values, as reflected both in the magnitude
of the standard deviations (Table 1) and the shape of
the distribution curves (Figure 1), was generally
smaller than observed in the 48 other INTERSALT
centers.14 Again, variation in the Kenya sample was
the largest among the four remote samples. For
example, the coefficient of variation (standard deviation/mean x 100) was almost twice as great for
diastolic pressure in the Kenya sample as in the two
Brazil samples (22% vs. 13%) and almost 50%
TABLE 2.
241
greater for systolic pressure. Slope (increase) of
blood pressure with age was also small in three of
the centers and was actually negative in four of the
eight estimates (Table 1 and Figure 2). In the Kenya
sample, the positive slope with age was smaller than
the average in the remaining INTERSALT centers
(about 2 mm Hg in 10 years in systolic pressure in
Kenya vs. a mean of 5 mm Hg in the rest of the
INTERSALT centers).14 Similarly, increase of diastolic pressure with age was negligible, with the
largest rise (1+ mm Hg/10 yr in Kenya) less than
the average for the remaining INTERSALT centers
(close to 4 mm Hg).
120-|
c
Blood Pressure in Four Remote Populations
sodium in the Yanomamo sample (0.8 mmol in men
and 1.0 mmol in women) was at the lower limit of
measurability by the flame photometer (Table 2).
Only nine persons in that sample had values higher
than 5 mmol (the maximum was 27 mmol), and
median output was 0.2 mmol. The Xingu also had
very low mean 24-hour sodium excretion (11 mmol
for women and 14 mmol for men) as did the Papua
New Guinean sample (32 mmol for women, 43
mmol for men). Median levels were 5.8 and 26.8
mmol in these two samples. Although daily sodium
excretion in Kenya (means of 53 and 60 mmol and
median of 51.3) was higher than in the other three
small-village population samples, it was still only
about a third of the average level in the rest of
INTERSALT sample (mean of 166 and median of
160 mmol/24 hr).14
Urinary potassium excretion in three of the centers was relatively high (60-96 mmol/24 hr) (Table
2). These levels exceeded the mean potassium value
of 54 mmol observed in the 48 other INTERSALT
centers and used here for comparison. The level in
Kenya, 32 mmol in men and 35 mmol in women,
Urinary Variables in Four Remote INTERSALT Populations
Yanomamo (n = 195)
Variables
Urinary Na (mmol/24 hr)
Urinary K (mmol/24 hr)
Urinary Na/K
ratio
Urinary Cl (mmol/24 hr)
Urinary Ca (mmol/24 hr)
Urinary Mg (mmol/24 hr)
Urinary Cr (mg/24 hr)
Urinary vol (1/24 hr)
Xingu (n-198)
Papua New Guinea
( n - •162)
Kenya (n = 176)
Men
(n=99)
Women
(n=96)
Men
(»=99)
Women
(n=99)
Men
(*=88)
Women
(n=74)
Men
(n=90)
Women
(n=86)
0.8
(1.8)
66.9
(35.4)
0.02
(0.03)
6.9
(7.4)
0.5
(0.1)
3.3
(1.9)
4.5
(2.3)
1.41
(0.60)
1.0
(3.1)
13.6
(25.2)
60.3
(36.7)
34.7
(20.1)
1.94
(1.04)
53.4
(28.3)
95.6
(40.8)
0.20
(0.53)
31.1
(31.6)
1.6
(1.2)
5.6
(2.3)
11.6
(3.3)
42.8
(34.7)
78.7
(43.6)
0.70
(0.65)
41.8
(35.0)
31.6
(26.6)
59.7
(28.1)
11.0
(17.0)
78.5
(46.0)
0.19
(0.31)
24.6
(29.9)
0.9
(0.9)
0.02
(0.01)
5.5
(4.8)
0.4
(0.5)
3.3
(1.7)
4.0
(2.0)
0.73
(0.22)
Values are mean±(SD). Cr, creatine; vol, volume.
1.86
(0.77)
4.0
(1.9)
7.2
(2.8)
1.30
(0.77)
0.8
(0-7)
5.5
(2.1)
9.1
(3.4)
62.0
(26-4)
0.53
(0.43)
29.1
(23.8)
0.6
(1.0)
4.1
(1.8)
6.0
(5.1)
0.77
(0.72)
0.53
(0.35)
68.2
(39.6)
1.6
(1.5)
4.5
(2.3)
8.8
(2.9)
1.14
(0.29)
32.0
(16.3)
1.89
(1.07)
57.9
(30.8)
1.6
(1.6)
3.6
(1.5)
6.7
(2.1)
1.00
(0.57)
242
Hypertension
TABLE 3.
Vol 14, No 3, September
1989
Height, Weight, Body Mass Index, and Pulse in Four Remote 1NTERSALT Populations
Yanomamo I(#, = 195)
Variables
Height (cm)
Weight (kg)
BMI (kg/m 2 )
Pulse (beats/min)
Xingu (n = 198)
Papua New Guinea
(#1 = 162)
Kenya( #1 = 176)
Men
(#,=99)
Women
(#i=96)
Men
("=99)
Women
(#.=99)
Men
(#i=88)
Women
("=74)
Men
(»i=90)
Women
(#i=86)
152.3
(4.4)
50.0
(6.1)
21.5
(2.0)
142.1
(4.2)
42.1
(4.3)
20.8
(1.7)
84.2
(10.9)
161.5
(5.6)
63.2
(8.6)
24.2
(2.7)
149.7
(5.2)
50.7
(8.3)
22.6
(3.2)
72.1
(8.2)
161.8
(6.5)
58.3
(6.9)
22.2
(1.8)
67.6
(11.1)
152.8
(4.7)
49.9
(6.0)
21.3
(2.0)
71.7
(12.4)
172.9
(6.3)
61.2
(9.2)
20.6
(2.7)
68.4
(11.1)
162.3
(5.2)
55.5
(7.9)
21.0
(2.7)
73.9
(10.7)
77.8
(10.7)
68.9
(9.2)
Values are mean±(SD). BMI, body mass index.
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
was below that of the other remote population
samples and of the average INTERSALT value for
potassium.
Reflecting the generally low level of sodium and
high level of potassium excretion (except for Kenya),
the sodium/potassium ratio was low (Table 2). In
three of these centers, the ratio was below 1.0. In
no other INTERS ALT center was a ratio less than
1.0 observed. In Kenya, this ratio was approximately 1.9; in the remaining 48 centers, mean
sodium/potassium ratio was 3.4.
Levels of urinary chloride were similar to those
of sodium: lowest in the Yanomamo (about 6 mmol),
highest in the Kenya sample (58 and 68 mmol)
(Table 2), but all lower than the mean level found in
the remaining 48 samples (171 mmol). Excretion of
calcium and magnesium was also low. Urinary
creatinine was lower than in almost all other centers, and was lowest in the Yanomamo who are
generally of small body build. In the Xingu and
Kenya samples and among Yanomamo men, urinary volume was similar to that of other centers,
but was lower than average in Papua New Guinea
and among Yanomamo women. This could reflect
low fluid intake, sweat loss, or incomplete collections in some individuals.
In almost all urinary measurements, values in
women were lower than those among men.
Body mass index, pulse, and alcohol. Mean BMI
was low in all groups (20.6-22.6) except for Xingu
men (24.2) (Table 3). In the remaining 48 centers,
mean BMI was 25.2, markedly higher than in three
of the small-village centers. Even among the shorter
Yanomamo, very low body weight (42-50 kg)
resulted in a low BMI. Except for Kenya, mean
BMI was higher in men than in women. Mean pulse,
however, was higher in women in all four samples.
In regard to alcohol (not shown), neither the
Yanomamo nor the Xingu drink alcoholic beverages, and the proportions of individuals reporting
consumption of alcohol were low in the Papua New
Guinea (8.7%) and Kenya samples (30.7%). In the
remaining INTERSALT centers, 53% reported consumption of alcohol. Among drinkers, the median
reported intake was 85 ml alcohol/wk in Papua New
Guineans; among Kenyan consumers of alcohol,
median weekly intake was 122 ml alcohol, not very
different from the median level among drinkers in
the remaining centers (131 ml).14
Relation of Sodium Excretion to Blood Pressure
In view of the limited variation in both mean
sodium excretion and blood pressure, it was not
likely that there would be a significant association
within each center between the two variables.
Although the confidence intervals around the
regression coefficients for these centers generally
included the value of the overall INTERSALT
coefficient, only a low-order association was seen
in regression analysis between sodium and blood
pressure when controlled for age-sex or alcohol,
BMI, and potassium. However, when the blood
pressure means in these four small-village samples
were plotted against the centers' mean 24-hour
urinary sodium excretions, blood pressure
increased in a pattern consistent with increased
sodium means, being lowest in the Yanomamo and
highest in the Kenya sample (Figure 3). This
relation was stronger for systolic than diastolic
pressure. The Kenya sample, highest in sodium
and lowest in potassium, also showed a significant
association between the sodium/potassium ratio of
individuals and their systolic pressure.
Relation of Other Factors to Blood Pressure
Within these centers, BMI was positively associated with blood pressure in seven of the eight
analyses (i.e., of SBP and DBP in the four centers),
significantly so in the Yanomamo (both SBP and
DBP) and Kenyans (DBP). Pulse, like BMI, was
positively associated with blood pressure (in all
eight analyses), significantly so in the Yanomamo
and also in the Xingu sample (DBP). Alcohol, low
or not used in these centers, was not associated
significantly with blood pressure.
Regression coefficients of blood pressure on
urinary potassium were negative in seven of the
eight within-center analyses, significantly so only
for DBP in the Yanomamo. All remaining coefficients of blood pressure with urinary electrolytes
Carvalho et al
=106-
1
<D CO
95
L_
CO
CO
CD
85
°-
75
•o
O o
£ f, 65CD
S
§ O 550
10 20 30 40 50 60
Mean Sodium Excretion
(mmol/24 hr)
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
• Yanomamo Sample
• Xingu Sample
o Papua New Guinea Sample
• Kenya Sample
FIGURE 3. Graphs of mean systolic and diastolic pressures by mean level of sodium excretion in the four
populations.
(chloride, calcium, magnesium) were small and
nonsignificant.
Discussion
In the INTERSALT study, it was possible to
examine patterns of blood pressure and the factors
possibly associated with these patterns in four
remote populations and to compare them with other
population samples examined contemporaneously
and using the same methods. The key findings in the
four remote population samples were low average
blood pressures, absence or near absence of hypertension or rise of blood pressure with age. These
patterns were in marked contrast with findings in
the remaining 48 INTERSALT centers.
Also in marked contrast was the level of sodium
intake, as assessed in carefully timed 24-hour urine
collections. When the sodium values are converted
to salt, those individuals studied in the Yanomamo
villages consumed virtually no salt, while the Xingu
Indians had a median intake of about one third gram
salt daily. Those in the sample from Papua New
Guinea had a median consumption of 1.5 g salt
daily. The highest consumers among these four
populations, the Kenya sample, had a median intake
under 3 g/day. This contrasts with the median level
of more than 9 g salt in the rest of INTERSALT.
In Kenya, where migration to the cities has an
increasing impact on those remaining in the rural
communities,12 intake of sodium was much higher
than in the other three remote population samples,
potassium was lower, and the sodium/potassium
ratio was markedly greater. Parallel with these
differences were the differences in blood pressure in
the Kenya sample: higher mean levels of SBP and
DBP, wider variation in the distribution, 5% with
Blood Pressure in Four Remote Populations
243
hypertension (compared with 0-1% in the other
three centers), and a positive (though still small)
increase of blood pressure with age.
Epidemiological data indicate that excess body
weight and high alcohol intake are also factors
contributing to higher blood pressures.19-23 In these
characteristics as well, the four samples differed
from the 48 other INTERSALT samples. Body
weights were low, and alcohol intake, if any, was
also lower than in other centers. However, among
INTERSALT samples overall, low body weight
was not in itself observed to be a sufficient assurance of low prevalence of hypertension; several
other INTERSALT centers (e.g., Colombia, South
Korea, Taiwan, and six of the eight centers in
Japan, China, and India) with low mean BMI similar to those in the four remote population samples
(20.0-23.1) but with high salt intake (8-12 g daily)
had prevalence rates of hypertension of 8-18%. Nor
was low alcohol intake, in itself, a sufficient assurance against hypertension; five INTERSALT centers reported prevalence of drinking at or lower
than the proportion in Kenya (30.7%, highest among
the remote population samples). However, these
five centers also had salt intake of 7-12 g daily and
prevalence rates of hypertension of 8-19%.
The conclusion from the overall INTERSALT
study, as well as from the findings in the present
report, that habitual high salt intake is a critical
environmental factor contributing to rise in blood
pressure and high prevalence rates of hypertension
in populations, is supported by data from other
research methodologies. These include animal experimentation with hypertension induction by high salt
feeding, early trials on severe sodium restriction in
hypertensive patients, and more recent trials with
more moderate salt restriction.24-30
Overall in the INTERSALT study, a significant
positive association within centers was found
between 24-hour urinary sodium excretion, BMI,
excess alcohol intake, and systolic blood pressure
of individuals and a significant negative association
of urinary potassium with blood pressure. Within
the four isolated centers, such significant associations were generally not found; it is the authors'
judgment that this could be a consequence of the
small variation within the centers both in the characteristics reflecting lifestyle and in blood pressure.
In these populations, therefore, the role of such
factors as low sodium intake, high potassium intake,
low alcohol, and low body weight is chiefly shown
in the ecological finding of little or no hypertension
and little or no rise in blood pressure with age. This
is not the first observation of these patterns of blood
pressure in such populations,1-13 but the standardized methods used across all INTERSALT centers
and the ability to compare contemporaneous measurements over a wide range of the variables studied add strength and depth to the earlier findings.
It may well be that other factors not measured in
this study also have relevance to the blood pressure
244
Hypertension
Vol 14, No 3, September 1989
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
findings. These four have other characteristics that
may also relate to blood pressure level: high intake
offiber,31-32low intake of saturated and total fats,33-34
and relatively high level of physical activity.35 Their
social structure and daily life patterns are also
distinctive. The findings of Page et al6 may have
relevance here: in comparing blood pressure levels
among six Solomon Islands populations similar in
such aspects of lifestyle, they found that hypertension was virtually absent in all but the one population using salty inlet water for cooking.
Some investigators have suggested that the
absence of blood pressure rise in such populations
may be due to chronic disease or malnutrition.36-37
However, adults in the four remote population
samples were physically active and generally
appeared healthy, and no physical signs of evident
malnutrition or protein deficiency were found.
Truswell et al3 and Page et al6 also found generally
good nutritional status among similar populations
with little or no hypertension or rise of pressure
with age.
Based on the study of these four centers and in
their comparison with the remaining 48 centers, we
conclude that, with a habitual low salt Intake over
the life span, there is little or no hypertension or rise
of blood pressure with age. Although other characteristics of these four populations may have also
played a contributory role, the findings in these
centers are consistent with the view that a certain
minimum intake of salt is required to produce a high
frequency of hypertension in populations.38 As
reported earlier, in the remaining 48 ENTERSALT
centers with higher salt intake, a positive association was found between sodium excretion and blood
pressure of individuals, as well as rise of pressure
with age.14 Those overall findings indicate the potential for lower average population pressure with
lowered salt intake, even where such intake is
above that of the four very low sodium populations
reported on here.
Acknowledgments
The authors and investigators in the four INTERSALT samples described above want to thank the
people in the local villages who helped as participants, translators, or facilitators and thereby made
possible this important part of the INTERSALT
study. We wish also to acknowledge specific individuals and groups for their support of the study in
the four populations. Xingu: The agreement of the
Fundacao Nacional do Indio and the Escola Paulista
de Medicina, which gave access to the Indian
population; CNPq Grant #40.6495/85. The following persons aided greatly in the field work: Douglas
A. Rodrigues, Sofia B. Mendonca, Paula Zucchi,
Marilia Louvison, and Nicanor R.S. Pinto. Kenya:
We thank the Wellcome Trust for special assistance
to field work in Kenya, and express appreciation to
Drs. J. Cavanagh and R. Nieman for their aid in
their work. Papua New Guinea: We acknowledge,
with thanks, the help of Danny Namarope and
Bikus Buni in the field work and of Travers Jenkins
in preparation of the population census. Yanomamo:
Thanks are expressed to the Fundacao Nacional do
Indio for their assistance in contact with the local
population and to Francisco Bezzera for his aid as
well.
INTERSALT was launched under the auspices
of the Council on Epidemiology and Prevention of
the International Society and Federation of Cardiology (ISFC).
Appendix
INTERSALT Cooperative Research Group
Members of the Executive Committee were Professors Geoffrey Rose and Jeremiah Stamler (Principal Investigators), Professor Rose Stamler, Dr.
Paul Elliott (Coordinator), Professor Michael Marmot, Professor Kalevi Pyorala (Council on Epidemiology and Prevention, ISFC), Professors Hugo
Kesteloot and Josef Joossens (Central Laboratory),
Professors Lennart Hansson and Giuseppe Mancia
(Council on Hypertension, ISFC), Professors Alan
Dyer, Daan Kromhout, and Ulrich Laaser, and Dr.
Susana Sans.
Participating centers and investigators were:
Argentina (Buenos Aires), Drs. E.C. Balossi, J.
Hauger-Klevene; Belgium (Charleroi), Professor M.
Kornitzer, M-P Vanderelst, M. Dramaix; Belgium
(Ghent), Dr. G. De Backer, I. De Craene, P.
Vannoote; Brazil (Yanomamo Indians), Drs. J.J.
Mancilha Carvalho, R. de Ohveira, R.J. Esposito;
Brazil (Xingu Indians), Professor R. Baruzzi, Drs.
L.J. Franco, L.F. Marcopito; Canada (Labrador
and St. John's), Professor J.G. Fodor, Dr. M.
Baikie, M. Webb, Dr. J.R. Martin, Dr. G. Mohacsi,
C. Bursey; Colombia (Tuquerres), Drs. P. Correa,
G. Montes; Denmark (Glostrup), Drs. K. Klarlund,
M. Schroll; Finland (Joensuu), Dr. P. Pietinen, U.
Uusitalo, Dr. A. Nissinen; Finland (Turku), Drs. O.
Impivaara, A. Aromaa, J. Maatala; FRG (Bernried), Drs. H. Hofmann, C. Bothge, S. Haselwarter; FRG (Heidelberg), Professor U. Laaser,
Dr. M. Siegel, Professor F. Luft; GDR (Cottbus),
Professor L. Heinemann, Drs. W. Barth, E.
Schueler; Hungary (Porcsalma village), Dr. J. Kishegyi; Iceland (Reykjavik and district), Dr. J. Ragnarsson, Dr. G. Sigurdsson, T. Karlsdottir; India
(Ladakh and New Delhi), Drs. K. Srinath Reddy,
M. Vijay Kumar, T. Norboo; Italy (Bassiano),
Professor G. Urbinati, Drs. F. Angelico, M. Del
Ben, A. Calvieri; Italy (Gubbio), Drs. M. Laurenzi,
L. Matarazzi, M. Panfili; Italy (Mirano), Professor
C. Dal Palu, Dr. S. Zamboni, G.B. Ambrosio, V.
Urbani, Dr. L. Mazzucato; Italy (Naples), Drs. E.
Farinaro, F. Jossa, M. Trevisan, Professor M.
Mancini; Japan (Osaka), Drs. H. Ueshima, S. Baba,
K. Mikawa, H. Ozawa; Japan (Tochigi prefecture),
Professor T. Hashimoto, Drs. Y. Fujita, S.
Maezawa; Japan (Toyama), Professor S. Kagami-
Carvalho et al
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
mori, Drs. H. Nakagawa, Y. Naruse; Kenya (Rambugu and Ndori villages), Drs. N. Poulter, J. Cavenagh, R. Nieman; Malta (Dingli village), Drs. J.M.
Cacciottolo, A. Amato Gauci; Mexico (Tarahumara
Indians), Professor W. Connor, Dr. M. McMurray,
M. Cerqueira, Dr. D. Leaf; The Netherlands (Zutphen), Professor D. Kromhout, Drs. M. Drijver, L.
Spliet-van Laar; Papua New Guinea (Asaro valley),
Drs. M. Alpers, P. Howard, V. Spooner; People's
Republic of China (Beijing), Professor Huang Da
Xian, Dr. Gong Wei Ru; People's Republic of China
(Nanning), Dr. Long Zupeng; People's Republic of
China (Tianjin), Drs. Liu Lisheng, Xie Jinxiang,
Hui Rutai; Poland (Krakow), Professor J. Sznajd,
Drs. G. Nowacki, A. Pajak, R. Konarska; Poland
(Warsaw), Professor S. Rywik, Drs. G. Broda, M.
Polakowska; Portugal (Cartaxo village), Drs. J.G.
Forte, J.M. Pereira Miguel; South Korea (Pusan),
Dr. B. Park, Dr. J. Lee, Dr. S. Lee, R. Struyven;
Soviet Union (Moscow), Professors R. Oganov, A.
Britov, Drs. N. Elisseeva, A. Deev; Spain (Manresa), Dr. S. Sans, Dr. J. Borras, I. Balaguer; Spain
(Torrejon), Professor M. Luque Otero, Drs. M.
Martell-Claros, F. Pinilla; Taiwan (San Chilo village
area), Professor Wen-Ping Tseng; Trinidad and
Tobago (Plymouth-Bethesda), Dr. A. Patrick: United
Kingdom (Belfast), Dr. G. Scally, Dr. A. Evans, G.
Keenan; United Kingdom (Birmingham), Dr. D.G.
Beevers, R. Hornby; United Kingdom (South
Wales), Dr. P.C. Elwood, S. Rogers, M. Lichtenstein; United States (Chicago), Professor J. Stamler, Professor R. Stamler, G. Crvinelli, C. McMillan,
C. Westbrook; United States (Goodman, two centers), Drs. S.A. Johnson, D.A. Frate; United States
(Hawaii), Drs. J.D. Curb, S. Knutsen, R. Knutsen;
United States (Jackson, two centers), Professor H.
Langford, Dr. R. Watson, J. Barr; Zimbabwe
(Harare), Dr. J. Matenga, S. Mukumba.
Members of the London Coordinating Center
were Professor G. Rose, Professor M. Marmot, D.
P. Elliott, M.J. Shipley, S. Tulloch, L. Cohvell, B.
Peachey, and L. Tudge.
Members of the Chicago Coordinating Center
were: Professor J. Stamler, Professor R. Stamler,
Professor A. Dyer, and G. Crvinelli.
Members of the Central Laboratory (Leuven)
were: Professor H. Kesteloot, Professor J. Joossens, and J. Geboers (laboratory coordinator).
References
1. Lowenstein FW: Blood pressure in relation to age and sex in
the tropics and subtropics. A review of the literature and an
investigation in two tribes of Brazil Indians. Lancet 1961;
1:389-392
2. Dahl LKJ Salt and hypertension. Am J Clin Nutr 1972;
25:231-244
3. Truswell AS, Kennelry BM, Hansen JDL, Lee RB: Blood
pressure of !Kung bushmen in Northern Botswana. Am
Heart J 1972;84:5-12
4. Shaper AG: Cardiovascular disease in the tropics. III. Blood
pressure and hypertension. Br Med J 1972;3:8O5-8O7
5. Gleibermann L: Blood pressure and dietary salt in human
populations. Ecology Food Nutr 1973;2:143-156
Blood Pressure in Four Remote Populations
245
6. Page LB, Damon A, Moellering RC: Antecedents of cardiovascular disease in six Solomon Islands societies. Circulation 1974;49:1132-1146
7. Oliver WJ, Cohen EL, Neel JV: Blood pressure, sodium
intake, and sodium related hormones in the Yanomamo
Indians, a "no-salt" culture. Circulation 1975;52:146-151
8. Freis ED: Salt, volume and prevention of hypertension.
Circulation 1976;53:589-595
9. Baruzzi RG, Marcopito LF, Serra MLC, Souza FAA, Stabile C: The Kren-Akorore: A recently contacted indigenous
tribe, in: Health and Disease in Tribal Societies, CIBA
Foundation Symposium 49 (new series). Amsterdam, Elsevier Science Publishing Co, Inc, 1977, pp 179-211.
10. Baruzzi RG, Franco LJ: Amerindians of Brazil, in Trowel!
HC, Burkitt DP (eds): Western Diseases: Their Emergence
and Prevention. London, Edward Arnold, 1981, pp 138-153
11. Elliott P, Marmot M: International studies of salt and blood
pressure. Ann Clin Res 1984;16(suppl 43):67-71
12. Poulter N, Khaw KT, Hopwood BE, Mugambi M, Peart
WS, Rose G, Sever PS: Blood pressure and associated
factors in a rural Kenyan community. Hypertension 1984;
6:810-813
13. Mancilha-Carvalho JJ: Estudo da pressao arterial de indios
Yanomami (thesis). Universidade Federal do Rio de Janeiro,
Rio de Janeiro, Brazil, 1986, pp 1-110
14. INTERSALT Cooperative Research Group: INTERSALT:
An international study of electrolyte excretion and blood
pressure. Results for 24-hour urinary sodium and potassium
excretion. Br MedJ 1988;297:319-330
15. Elliott P, Stamler R: Manual of Operation for "INTERSALT", an International Cooperative Study on the Relation
of Sodium and Potassium to Blood Pressure. Controlled Clin
Trials 1988;9(suppl):l-118
16. American Association of Clinical Chemists: Standard Methods of Clinical Chemistry. New York, Academic Press,
1961, vol 3, pp 81-92
17. Trudeau DL, Freier EF: Determination of calcium in urine
and serum by atomic absorption spectrophotometry. Clin
Chem 1967;l3:101-114
18. Jaffe M: Uber den Niederschlag welchen Pikrinsaure in
normalen Harn erzeug und uber eine neue Raction des
Creatinins. Z Physiol Chem 1986;10:391-400
19. Chiang BN, Perlman LV, Epstein FH: Overweight and
hypertension. A review. Circulation 1969;39:403-421
20. Berchtold P: Clinical and epidemiologic evidence relating
caloric intake, obesity, and hypertension, in Horan MJ,
Blaustein M, Dunbar JB, Kachadorian W, Kaplan NM,
Simopoulos AP (eds): NIH Workshop on Nutrition and
Hypertension. New York, Biomedical Information Corporation, 1984, p 189
21. Kannel WB, Dawber TR: Hypertensive cardiovascular disease: The Framingham study, in Onesti G, Kim KE, Moyer
JH (eds): Hypertension, Mechanisms, and Management.
New York, Grune and Stratton, 1973, p 93
22. Stamler J: Preventive cardiology, in Cheng TO (ed): The
International Textbook of Cardiology. New York, Pergamon
Press, 1986, pp 1231-1255
23. Dyer AR, Stamler J, Paul O, Berkson DM, Shekelle RB,
Lepper MH, McKean H, Lindberg HA, Garside D, Tokich T:
Alcohol, cardiovascular risk factors and mortality: The Chicago experience. Circulation 1981;64(suppl III):in-20-m-27
24. Ambard L, Beaujard E: Causes de l'hypertension arte'rielle.
Arch Gin Mid 1904;l:520-533
25. Allen FM: Arterial hypertension. JAMA 1920;74:652-655
26. Kempner W. Treatment of kidney disease and hypertensive
vascular disease with the rice diet: 1. N Carolina Med J
1944;5:125-133
27. Watkin DM, Froeb HF, Hatch FT, Gutman AB: Effect of
diet in essential hypertension. II. Results with unmodified
Kempner rice diet in fifty hospitalized patients. Am J Med
1950;9:441-493
28. MacGregor GA, Markandu ND, Best FE, Elder DM, Cam
JM, Sagnella GA, Squires M: Double blind randomised
246
29.
30.
31.
32.
33.
Hypertension
Vol 14, No 3, September 1989
crossover trial of moderate sodium restriction in essential
hypertension. Lancet 1982;l:351-355
Markandu ND, Singer DRJ, Sagnella GA, Cappuccio FP,
Sugden AL, MacGregor GA: A double blind study of three
sodium intakes, and the long term effects of sodium restriction in essential hypertension (abstract). / Hypertens 1988;
6(suppl 4):S742
Beard TC, Cooke HM, Gray WR, Barge R: Randomised
controlled trial of a no added-sodium diet for mild hypertension. Lancet 1982;2:455-458
Sacks FM, Rosner B, Kass EH: Blood pressure in vegetarians. Am J Epidemiol 1974;100:390-398
Mendeloff AI: Dietary fiber and hypertension, in Horan MJ,
Blaustein M, Dunbar JB, Kachadorian W, Kaplan NM,
Simopoulos AP (eds): NIH Workshop on Nutrition and
Hypertension. New York, Biomedical Information Corporation 1984, p 325
Puska P, Iacono JM, Nissinen A, Korhonen JH, Vartiainen
E, Pietinen P, Dougherty R, Leino U, Mutanen M, Moisio S,
34.
35.
36.
37.
38.
Huttunen J: Controlled randomised trial of the effect of
dietary fat on blood pressure. Lancet 1983;l:l-5
Iacono JM, Puska P, Dougherty RM: Studies on the effect of
dietary fat on blood pressure. Ann Clin Res 1984;
16(suppl 43):116-125
Paffenbarger RS Jr, Hyde RT: Exercise as protection against
heart attack. N Engl J Med 1980;302:1026-1027
Burns-Cox CJ, Maclean JD: Splenomegaly and blood pressure in an Orang Asli community in West Malaysia. Am
Heart J 1970;80:718-719
Maddocks I, Vine AP: The influence of chronic infection on
blood pressure in New Guinea males. Lancet 1966;2:262-264
Denton D: The Hunger for Salt. New York, SpringerVerlag, 1982, pp 542-629
KEY WORDS
• low sodium
pressure with age
blood pressure • slope of
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
Blood pressure in four remote populations in the INTERSALT Study.
J J Carvalho, R G Baruzzi, P F Howard, N Poulter, M P Alpers, L J Franco, L F Marcopito, V J
Spooner, A R Dyer and P Elliott
Hypertension. 1989;14:238-246
doi: 10.1161/01.HYP.14.3.238
Downloaded from http://hyper.ahajournals.org/ by guest on June 16, 2017
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1989 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/14/3/238
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/