Prevalence and determinants of hypertension in the urban population of Jaipur in western India

C

2218 Original article

Prevalence and determinants of hypertension in theAlgerian SaharaMohamed Temmara, Carlos Labatb, Salim Benkheddac, Meriem Charifid,Frederique Thomase, Mohamed Tahar Bouafiad, Kathy Beane,Bernadette Darnef, Michel E. Safarf and Athanase Benetosb,g

Background In-Salah is a city-oasis located in the middle

of the Algerian Sahara, a desert area whose drinking

water has a high sodium content. No cardiovascular

epidemiological studies have ever been conducted in

this region.

Methods A randomized sample of 635 men and

711 women, aged 40–99 years, was studied. Blood

pressure measurements, combined with a clinical

questionnaire that included educational and

socio-economic data, and standard blood samples for

the detection of dyslipidemia and diabetes mellitus,

were collected.

Results The mean age was 55 W 12 years. The prevalence

of hypertension was 44% and was highly influenced by age,

sex, skin colour, educational status, obesity and metabolic

parameters. The higher prevalence of hypertension among

black individuals was independent of socio-economic and

educational levels, and of metabolic parameters. The

presence of antihypertensive treatment was three times

more frequent in women than in men, and there was no

difference according to skin colour. Among treated subjects,

opyright © Lippincott Williams & Wilkins. Unautho

0263-6352 � 2007 Wolters Kluwer Health | Lippincott Williams & Wilkins

25% were well controlled, and this percentage was similar

among both black and white individuals.

Conclusion Epidemiological studies in such an emergent

population indicate that hypertension is a major public

health problem. The high sodium content in drinking water

in this region could play a major role in the development of

hypertension. J Hypertens 25:2218–2226 Q 2007 Wolters

Kluwer Health | Lippincott Williams & Wilkins.

Journal of Hypertension 2007, 25:2218–2226

Keywords: age, Algerian Sahara, cardiovascular risk factors, high bloodpressure, NaCl intake, skin colour

aCardiology Centre, Ghardaia, Algeria, bINSERM U684, Nancy, France,cCardiology Department, University Hospital Mustafa, Alger, dCardiologyDepartment, University Hospital of Blida, Blida, Algeria, eIPC Center, Paris,fHotel Dieu Hospital, Paris and gGeriatric Department, University Hospital ofNancy, Nancy, France

Correspondence to Athanase Benetos, MD, PhD, INSERM Unit 684, Universityof Nancy, 54511 Nancy-les-Vandoeuvre, FranceE-mail: [email protected]

Received 14 November 2006 Revised 29 May 2007Accepted 10 June 2007

IntroductionHypertension is a major cardiovascular disorder in Africa

[1–9]. Parallel to this, the incidence of diabetes mellitus

is on the rise [10], thus increasing the severity of renal and

cardiac damage caused by any given blood pressure level.

Most of the existing data on hypertension and diabetes

mellitus in Africa has been collected in sub-Saharan

countries. Hypertension is also a problem in supra-

Saharan nations such as Egypt [11].

Algeria is made up of two different geographical

regions. The northern region is situated primarily

along the Mediterranean sea. The southern region

represents the Algerian portion of the Sahara, between

Ghardaia in the north and Tamanrasset in the south. In-

Salah, a city located in the middle of this Saharan

country, is an oasis in a desert area where temperatures

range between 258 and 498C in summer (mean 388)and between 28 and 288C in winter (mean 158). The

population is composed of 50% black and 50% white

individuals.

The aim of this study was to estimate the prevalence of

hypertension and assess the determinants of blood pres-

sure levels in a random population over 40 years old living

in In-Salah city.

MethodsThe population sample was randomly selected and was

composed of 1346 subjects (635 men and 711 women,

aged 40–99 years). All subjects were aware of the time

and the purpose of the visit several weeks before. All

subjects had been asked to be under fasting conditions

the day of the examination. If an individual had not

respected the fasting conditions, the investigators

returned a few days later for the blood tests.

The total the population of In-Salah is approximately

30 000 inhabitants. According to the city records, 5210

individuals were 40 years of age or older (approximately

17% of the entire population) at the time this study was

undertaken. The city is divided into 38 districts; each

rized reproduction of this article is prohibited.

C

Prevalence of hypertension in the Algerian Sahara Temmar et al. 2219

district is subdivided into several neighbourhoods. For

the purpose of this study, 25% of the total population over

40 years of age was included. A sample of 1400 subjects

was drawn proportionally from each district. Among these

subjects, 54 were not included (39 were absent at the time

of the examination and 16 did not give their consent to

participate in the study). The study duration was 6 weeks,

from April to May. All subjects were examined between

0700 and 1230 h.

During the study, outdoor temperatures were recorded

using professional thermometers provided by the

National Meteorological Society (Meterologie Natio-

nale). Measurements were taken during short periods

of time in order to avoid major differences in temperature

during the study.

A questionnaire, which included life habits, personal and

family medical history as well as data for the socio-

economic and educational level, was completed. Subjects

were asked if they currently smoked more than five

cigarettes per day. Subjects who gave a negative answer

to this question were considered non-smokers. Alcohol

intake was not assessed with this questionnaire. In this

traditional Muslim area where alcohol intake is prohib-

ited by religious rules (even though not officially prohib-

ited by law), subjects might not be truthful and therefore

we judged that the answers to the question would be very

difficult to interpret.

Three categories of socio-economic level were defined

according to the following question included in the

questionnaire: (i) high socio-economic level: house with

refrigerator, air conditioning, and a maximum of two

individuals per room; (ii) mid level: either a refrigerator

or air conditioning; (iii) low level: neither of these two.

We chose these criteria because in this very hot area of

the world they constitute a priority for families. In the

Results section, we combined groups of the high and

mid level.

Educational level was defined as follows: (i) low level: no

school at all or only elementary school (4–5 years); (ii)

mid level: high school level (þ3 years); (iii) high level:

college, university. As the large majority of subjects

belonged to the first group, mid and high level subjects

were grouped together.

Three blood pressure measurements were taken at 2-min

intervals, after a 5-min rest in the sitting position, using

an electronic blood pressure monitor (Omron 705

CP; Omron Healthcare Europe BV, Amsterdam, the

Netherlands) [12]. The mean of the last two measure-

ments was used for statistical analysis. Hypertension was

defined as systolic blood pressure (SBP) of 140 mmHg

or greater or diastolic blood pressure (DBP) of 90 mmHg

opyright © Lippincott Williams & Wilkins. Unauth

or greater, or both, or the presence of chronic antihyper-

tensive drug treatment. Height, weight and waist and

hip circumference were also measured the same day.

Blood chemistry [plasma cholesterol, high-density lipo-

protein (HDL)-cholesterol, triglycerides and glucose]

was measured under fasting conditions according to stan-

dard techniques. The metabolic syndrome was defined

according the National Cholesterol Education Program

2001 definition [13].

For the present analysis, subjects were divided into four

subgroups according to sex and skin colour. Skin pig-

mentation and general characteristics were the main

criteria the investigators of the present study used to

classify the examined subjects as ‘black’ or ‘white’. When

these criteria were not conclusive (in less than 10% of the

cases), the examined subjects were asked if they defined

themselves as white or black. Following this procedure,

all subjects were classified as black or white.

For the statistical evaluation, mean values� standard

deviation are presented. The prevalence between differ-

ent groups was compared using a chisquared test. Black

men and women and white men and women were studied

using a two-way analysis of variance, involving a sex

effect, a skin colour effect and their interaction. Multiple

regression analyses were performed using a two-tailed

test. A P value of 0.05 or less was considered significant.

ResultsDescription of the populationTable 1 shows the main clinical and biological charac-

teristics of the population, as well as their distribution

according to sex and skin colour.

Age was similar in the four groups of subjects. SBP, DBP

and mean arterial pressure were significantly higher in

black than in white subjects (P< 0.001; Table 1). This

difference was less important for pulse pressure

(P¼ 0.045). There was no sex effect on SBP, DBP, mean

arterial pressure and pulse pressure levels. As expected,

heart rate was markedly higher (P< 0.001) in women

than in men. Body mass index and waist circumference

were also significantly higher in women than in men

(P< 0.001).

In both sexes, plasma glucose did not differ in black and

white subjects. Total and HDL-cholesterol were signifi-

cantly higher in blacks than in whites (P< 0.01 for total

cholesterol and P< 0.05 for HDL-cholesterol) and in

women than in men (sex effect P< 0.05). The contrary

was found for plasma triglyceride levels, which were

higher in men than in women (P< 0.001) and higher

in white than in black subjects (P< 0.02). No significant

sex–skin colour interaction was observed for these

parameters.

orized reproduction of this article is prohibited.

C

2220 Journal of Hypertension 2007, Vol 25 No 11

Table 1 In-Salah population sample: clinical and biological characteristics according to skin colour and sex

All

Blacks Whites ANOVA (P)

Men Women Men Women Skin colour effect Sex effect Interaction

Number 1346 349 372 286 339White skin colour (%) 46.3%Women (%) 52.9%Age (years) 55.1�12.2 55.6�11.9 55.6�12.6 54.3�11.8 54.6�12.2 0.09 0.91 0.81SBP (mmHg) 137.0�28.5 139.8�28.5 139.1�29.7 135.9�26.0 132.7�28.7þ <0.001 0.19 0.41DBP (mmHg) 81.4�14.1 82.6�14.9 83.2�14.4 80.4�13.4þ 79.2�13.0þ <0.001 0.71 0.26PP (mmHg) 55.6�18.8 57.3�17.7 55.9�20.1 55.6�17.0 53. 5�19.6 0.045 0.10 0.67MAP (mmHg) 99.9�18.0 101.7�18.7 101.8�18.5 98.9�16.8 97.0�17.4þ <0.001 0.39 0.30Heart rate (bpm) 79.3�13.2 77.5�12.8 81.9�13.8M 76.4�12.8 80.7�12.7M 0.12 <0.001 0.99BMI (kg/m2) 24.9�5.5 23.8�4.3 25.6�6.2M 23.6�4.2 26.6�6.0Mþ 0.12 <0.001 0.05Waist circumference (cm) 83.9�12.89 83.8�12.2 83.8�14.5 83.2�11.5 84.6�12.9 0.053 0.17 0.15Hip circumference (cm) 96.7�10.56 94.8�8.4 97.4�11.4M 95.2�10.1 98.8�11.5M 0.12 <0.001 0.37Cholesterol (g/l) 1.81�0.54 1.80�0.56 1.89�0.56M 1.74�0.51 1.79�0.51þ <0.02 <0.05 0.48HDL-cholesterol (g/l) 0.49�0.18 0.47�0.15 0.53�0.20M 0.48�0.18 0.48�0.16þ <0.05 <0.005 <0.01LDL-cholesterol (g/l) 1.08�0.51 1.09�0.51 1.14�0.53 1.01�0.49 1.07�0.50 <0.02 0.06 0.69Triglycerides (g/l) 1.18�0.71 1.20�0.68 1.09�0.61M 1.32�0.88þ 1.17�0.68M <0.02 <0.001 0.56Glucose (g/l) 1.05�0.50 1.04�0.45 1.03�0.44 1.04�0.47 1.10�0.63 0.21 0.39 0.19Metabolic syndromea 26% 17% 32%M 19% 33%Mþ 0.58 <0.001 0.94Smokers (current) 9% 18% 0%M 20% 0%M 0.60 <0.001 0.62Low educational level (%) 86.8% 73.5% 98.6%M 75.6% 96.7%M 0.99 <0.001 0.27Low socio-economic status (%) 27.8% 23.6% 32.6%M 24.5% 29.7% 0.73 <0.01 0.45

ANOVA, Analysis of variance; BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MAP, mean arterialpressure; PP, pulse pressure; SBP, systolic blood pressure. Values are mean�SD. MP<0.05 versus men and þP<0.05 versus black. a For the definition of the metabolicsyndrome, we respected the NCEP 2001 definition [13].

Educational level (P< 0.001) and socio-economic status

(P< 0.01) were markedly higher in men than in women.

Interestingly, there were no differences for these

parameters according to skin colour. Tobacco consump-

tion was declared exclusively among the men; none of the

women declared smoking.

Prevalence of hypertension and blood pressure controlwith antihypertensive treatmentFor the entire population, the prevalence of hypertension

(�140/90 mmHg or the presence of antihypertensive

treatment) was 44%, with a higher prevalence among

women than men (47 versus 41%; P< 0.05) and a higher

prevalence among black than white subjects (47 versus

41%; P< 0.01). As shown in Table 2, the prevalence of

opyright © Lippincott Williams & Wilkins. Unautho

Table 2 Prevalence of hypertension and effects of antihypertensive tre

All

Blacks

Men Wome

Number 1346 349 372Prevalence of hypertension 594 (44%) 151 (43%) 194 (52Antihypertensive therapy among hypertensives 179 (30%) 23 (15%) 81 (42SBP in untreated (mmHg) 134�27 138�28 134�2SBP in treated (mmHg) 159�29 169�25 156�2DBP in untreated (mmHg) 80�13 82�14 81�1DBP in treated (mmHg) 91�14 97�15 90�1PP in untreated (mmHg) 54�18 56�17 53�1PP in treated (mmHg) 68�20 72�17 66�2BP control among treated

SBP <140, DBP <90 45 (25%) 1 (4%) 25 (31SBP <140, DBP �90 6 (3%) 1 (4%) 1 (1%SBP �140, DBP <90 49 (28%) 7 (31%) 21 (26SBP �140, DBP �90 79 (44%) 14 (61%) 34 (42

ANOVA, Analysis of variance; DBP, diastolic blood pressure; PP, pulse pressure. Val

hypertension was 52% in black women, 43% in black

men, 41% in white women, and 39% in white men.

Table 3 shows the results of the multivariate logistic

regression analyses describing the variables influencing

the presence of hypertension. Age and increased waist

circumference, high triglyceride levels and glycemia

were all strongly associated with hypertension. In

addition, black skin and low educational levels were

independently associated with hypertension. The logistic

regression analysis showed that women had a 20% higher

prevalence of hypertension, but this sex effect was

not significant.

In both black and white subjects, antihypertensive drug

treatment was markedly more frequent in women than in

rized reproduction of this article is prohibited.

atment according to skin colour and sex

Whites ANOVA (P)

n Men Women Skin colour effect Sex effect Interaction

286 339%)M 111 (39%) 138 (41%)þ <0.01 <0.05 0.19%)M 13 (12%) 62 (45%)M 0.13 <0.001 0.698 135�25 127�26Mþ <0.001 <0.001 0.198 163�39 157�28 0.71 0.08 0.554 80�13 77�12Mþ <0.001 <0.05 0.155 92�18 89�13 0.22 0.08 0.569M 55�16 50�18Mþ <0.05 <0.001 0.360 71�24 69�21 0.71 0.25 0.66

%)M 4 (31%) 15 (24%) 0.24 0.23 <0.05) 0 (0%) 4 (6%) 0.90 0.63 0.17

%) 3 (23%) 18 (29%) 0.80 0.93 0.54%) 6 (46%) 25 (41%) 0.39 0.19 0.50

ues are mean�SD. MP<0.05 versus men and þP<0.05 versus black.

C

Prevalence of hypertension in the Algerian Sahara Temmar et al. 2221

Table 3 Variables associated with presence of hypertensionfollowing a logistic multivariate regression

OR 95% CI Z P

Sex (women vs men) 1.18 0.90–1.54 1.20 0.23Age (1 year) 1.06 1.05–1.07 9.907 <0.001Skin colour (black/white) 1.41 1.09–1.82 2.66 <0.01Education level (low vs others) 1.68 1.08–2.61 2.29 <0.05Waist (1 cm) 1.03 1.02–1.05 6.17 <0.001Triglycerides (1 g/l) 1.32 1.08–1.60 2.76 <0.01Glucose (1 g/l) 1.57 1.19–2.06 3.24 <0.001

CI, Confidence interval; OR, odds ratio.

men (sex effect P< 0.001), with no difference according

to the skin colour (Table 2). Among treated subjects, 25%

were well controlled (SBP < 140 and DBP < 90 mmHg),

and this percentage was similar among both black and

white individuals. Forty-two per cent of the treated white

subjects and 46% of the treated black subjects had

uncontrolled SBP and DBP, whereas 27 and 28%,

respectively, presented with uncontrolled SBP but had

controlled DBP under treatment. Finally, 5% of the

white and 2% of the black subjects had high DBP and

controlled SBP under treatment. Blood pressure values in

the four groups, according to sex and skin colour, are also

illustrated in Table 2. These data show a better control of

blood pressure in black women than in black men. No

opyright © Lippincott Williams & Wilkins. Unauth

Fig. 1

Black men

White men

y = 0.9187x + 89.432R2 = 0.1492

50

100

150

200

250

300

110 years90705030

y = 0.6964x + 97.666R2 = 0.1021

50

100

150

200

250

300

110 years90705030

SB

P (

mm

Hg

) S

BP

(m

mH

g)

Univariate correlation of age versus systolic blood pressure (SBP) in the In

other significant influence of sex or skin colour on blood

pressure control under treatment was observed, but as a

result of the small number of treated men the statistical

power is relatively weak.

As shown in Table 2, in all groups, treated subjects had

higher blood pressure levels than untreated subjects

(P< 0.001). Among untreated subjects, SBP and DBP

were higher in men than in women (P< 0.001 for SBP

and P< 0.05 for DBP). The same trend was observed

among treated subjects, but the differences between men

and women did not reach levels of significance (P¼ 0.08

for both SBP and DBP). Pulse pressure was also higher

in untreated men compared with untreated women

(P< 0.001); no such sex difference was observed in

treated subjects. Among untreated individuals, SBP

(P< 0.001), DBP (P< 0.001) and pulse pressure

(P< 0.05) were higher in black than in white subjects.

No skin colour differences in blood pressure levels were

observed among treated subjects.

Factors associated with systolic, diastolic and pulsepressureThe evolution of SBP (Fig. 1), DBP (Fig. 2) and pulse

pressure (Fig. 3) with age in black men, black women,

white men and white women, was examined. As

orized reproduction of this article is prohibited.

Black women

White women

y = 0.8502x + 91.676R2= 0.1305

50

100

150

200

250

300

110 years90705030

y = 0.9584x + 81.243R2 = 0.1603

50

100

150

200

250

300

110 years90705030

SB

P (

mm

Hg

) S

BP

(m

mH

g)

-Salah population sample of black and white men and women.

C

2222 Journal of Hypertension 2007, Vol 25 No 11

Fig. 2

Black men

DB

P (

mm

Hg

)

y = --0.0079x2 + 1.1984x + 41.521R2 = 0.0426

40

80

120

160

110 years90705030

White men

DB

P (

mm

Hg

)

y = --0.0175x2 + 2.1426x + 17.917R2 = 0.049

40

80

120

160

110 years90705030

Black women

DB

P (

mm

Hg

)

y = --0.0121x2 + 1.583x + 34.373R2 = 0.0314

40

80

120

160

110 years90705030

White women

DB

P (

mm

Hg

)

y = --0.0073x2 + 1.0885x + 42.728R2 = 0.0479

40

80

120

160

110 years90705030

Univariate correlation of age versus diastolic blood pressure (DBP) in the In-Salah population sample of black and white men and women.

expected, SBP (Fig. 1) and pulse pressure (Fig. 3) showed

strong positive correlations with age in all four groups of

subjects (P< 0.001).

Sex and skin colour did not interact with the age/SBP and

age/pulse pressure relationship. Actually, the age�sex and

the age�colour interactions were not statistically signifi-

cant. The relationships between age and SBP or pulse

pressure were also found in the multivariate analyses after

taking into account sex, skin colour and several other

parameters (Table 4). As shown in Fig. 2 a non-linear

relationship was found between DBP and age in the

four groups of subjects; DBP increases with age up

until 65–70 years, and then decreases.

The multivariate analysis showed that age and male sex

were associated with higher SBP, DBP and pulse pres-

sure (Table 4). Skin colour influenced SBP and DBP

levels (black >white) but not pulse pressure levels. Waist

circumference and glycemia were also significantly

associated with SBP, DBP and pulse pressure levels.

Triglyceride levels were positively associated with

opyright © Lippincott Williams & Wilkins. Unautho

levels of SBP and pulse pressure, whereas cholesterol

levels were correlated with DBP levels. The presence of

antihypertensive treatment was associated with higher

SBP, DBP and pulse pressure levels. The age�treatment

interaction was not significant.

DiscussionThe study showed that in In-Salah, a city located in the

middle of the Algerian Sahara, the prevalence of hyper-

tension is approximately 44.2%. The highest prevalence

was observed in black women, followed by black men,

and then white women. The lowest prevalence was

observed in white men. Other than sex and skin colour,

age and educational level were also associated with the

presence of hypertension.

Effects of age on systolic, diastolic and pulse pressureIn the past, many genetic and environmental factors have

been said to characterize the elevated incidence of

hypertension in African countries [1–14]. The most

established factors are those derived from the concepts

of acculturation and migration, which have been widely

rized reproduction of this article is prohibited.

C

Prevalence of hypertension in the Algerian Sahara Temmar et al. 2223

Fig. 3

Black men

White men

Black women

White women

y = 0.7025x + 17.889R2 = 0.2212

0

40

80

120

160

110 years90705030

y = 0.6099x + 21.571R2 = 0.1807

0

40

80

120

160

110 years90705030

y = 0.7522x + 13.569R2 = 0.2191

0

40

80

120

160

110 years90705030

y = 0.783x + 10.725R2 = 0.2371

0

40

80

120

160

110 years90705030

PP

(m

mH

g)

PP

(m

mH

g)

PP

(m

mH

g)

PP

(m

mH

g)

Univariate correlation of age versus pulse pressure (PP) in the In-Salah population sample of black and white men and women.

applied to several populations in Africa, primarily via

cross-sectional investigations [1]. Most previous studies

were, however, performed at a period during which the

duration of life was relatively short. Populations were

composed mainly of individuals aged 70 years and

younger. Consequently, the classic reduction in DBP

with age was difficult to evaluate. In the present study,

this age-related decrease in DBP was clear, especially

after the age of 70 years. This is an important point to take

into consideration because the age-related decrease in

DBP is an obligatory component of the age-related

increase in arterial stiffness and pulse pressure, which

represents a significant cardiovascular risk factor especi-

ally after the age of 55 years [14,15].

Sodium intake and hypertensionAs shown in several other populations, blood pressure

levels are primarily influenced by age, obesity and meta-

bolic factors such as blood sugar, cholesterol and trigly-

ceride levels. In the population studied, the presence of

metabolic disorders was low compared with what is

observed in developed countries, and obesity was practi-

cally non-existent in men. Despite this low-risk profile,

the presence of hypertension and an increase in SBP and

pulse pressure with age were observed in this population.

Clinical studies have shown that the age-dependent

opyright © Lippincott Williams & Wilkins. Unauth

increase in SBP and arterial stiffness is absent in societies

having a low sodium (NaCl) intake [1,2,16]. The high

prevalence of hypertension and the important age-related

changes in SBP and pulse pressure observed in the In-

Salah population indicate that high NaCl levels in drink-

ing water could be the major determinant of hypertension

and arterial stiffening with age in this population [17].

The sodium concentration is actually much higher in

In-Salah than in other Algerian cities and also in other

countries (Fig. 4). The contribution of sodium intake to

the mechanisms of hypertension has been widely inves-

tigated [9], particularly in black American individuals

[9,18,19]. Sodium levels in circulating blood cells are

elevated in certain African individuals, especially in south

and central African hypertensive blacks. A depression of

the sodium pump has been widely observed [20–23].

Further studies in In-Salah are needed to evaluate the

possible role of the high NaCl content in drinking water

in the prevalence of hypertension, arterial ageing and

cardiovascular events.

Sex, skin colour and blood pressure levelsIn this middle-aged population, the prevalence of hyper-

tension was higher in women than in men. Interestingly,

among subjects with hypertension, the percentage of

treated individuals was three times higher in women

orized reproduction of this article is prohibited.

Copyright © Lippincott Williams & Wilkins. Unautho

2224 Journal of Hypertension 2007, Vol 25 No 11

Table 4 Multivariate analysis: variables associated with changes insystolic, diastolic and pulse pressure expressed in mmHg

ParametersRegressioncoefficient R2 T P

SBPAge (1 year) 0.80�0.06 0.096 12.33 <0.001Sex (women vs men) �6.11�1.59 0.009 �3.85 <0.001Skin colour (black/white) 3.48�1.46 0.004 2.39 <0.02Education level (mid or high vs low) �4.38�2.41 0.002 �1.81 0.07Antihypertensive treatment (yes/no) 18.40�2.15 0.046 8.56 <0.001Waist circumference (1 cm) 0.25�0.06 0.012 4.27 <0.001Triglycerides (g/l) 3.06�1.09 0.005 2.81 <0.005Glucose (g/l) 5.69�1.46 0.010 3.91 <0.001Cholesterol (g/l) 1.98�1.39 0.001 1.43 0.16Total model 0.266

DBPAge (1 year) 0.11�0.03 0.008 3.26 <0.002Sex (women vs men) �2.53�0.84 0.007 �3.02 <0.005Skin colour (black/white) 2.60�0.77 0.008 3.38 <0.001Education level (mid or high vs low) �1.60�1.27 0.001 �1.25 0.21Antihypertensive treatment (yes/no) 8.96�1.13 0.046 7.90 <0.001Waist circumference (1 cm) 0.16�0.03 0.020 5.19 <0.001Triglycerides (1 g/l) 0.92�0.57 0.002 1.60 0.11Glucose (1 g/l) 2.26�0.77 0.006 2.94 <0.005Cholesterol (1 g/l) 2.51�0.73 0.009 3.43 <0.001Total model 0.148

PPAge (1 year) 0.69�0.04 0.162 16.41 <0.001Sex (women vs men) �3.57�1.03 0.007 �3.48 <0.001Skin colour (black/white) 0.87�0.94 0.001 0.93 0.36Education level (mid or high vs low) �2.73�1.56 0.002 �1.75 0.08Antihypertensive treatment (yes/no) 9.48�1.39 0.028 6.81 <0.001Waist circumference (1 cm) 0.09�0.04 0.003 2.33 <0.02Triglycerides (1 g/l) 2.14�0.70 0.006 3.04 <0.005Glucose (1 g/l) 3.43�0.94 0.008 3.65 <0.001Cholesterol (1 g/l) �0.52�0.90 0.000 �0.58 0.57Total model 0.301

DBP, Diastolic blood pressure; PP, pulse pressure; SBP, systolic blood pressure.

Fig. 4

0

100

200

300

400

500

600

Na+

(m

g/l

)

0

100

200

300

400

500

600

Do

mes

tic

dis

trib

uti

on

syst

em

Dri

llin

g

Pu

rifi

edw

ater

Sai

da

(Alg

eria

)

Ifri

(Alg

eria

)

El G

olé

a

Ca+

(m

g/l

)

OMS limit

CEE limit

OMS limitCEE limit

Sodium and calcium concentrations (mg/l) in water in In-Salah and other citiworld; CEE, threshold limit for Europe. & In-Salah waters; mineral water

than in men. The low percentage of treated subjects

among men (< 15%) explains the significantly higher

blood pressure values among men without treatment

compared with untreated women. These observations

explain why in the entire population, despite a higher

prevalence of hypertension in women, SBP and DBP

levels were not higher in women than in men.

The findings observed in In-Salah suggest that skin

colour is an important determinant of blood pressure

severity and the prevalence of hypertension. Interest-

ingly, contrary to what is found in the United States and

other populations [8,9], in In-Salah there were no socio-

economic and educational differences between black and

white individuals. Therefore, although educational level

was strongly associated with hypertension, the observed

differences according to skin colour cannot be attributed

to such social indicators. Moreover, in the present study,

the metabolic parameters associated with high blood

pressure cannot explain the higher blood pressure levels

in blacks because body mass index and glycemia were no

different according to skin colour, whereas triglyceride

levels were lower and HDL-cholesterol was higher in

black compared with white subjects. Therefore, the

association between skin colour and blood pressure could

most likely be related to a greater sensitivity to salt

among black individuals in this area in which drinking

water contains very high levels of NaCl.

rized reproduction of this article is prohibited.

Ain

sai

ss(M

oro

cco

)

(Alg

eria

)

Evi

an(F

ran

ce)

Vit

el(F

ran

ce)

Co

ntr

ex(F

ran

ce)

Dec

amte

(US

A)

Aq

uel

a(S

wit

zerl

and

)

es and countries within the world. OMS, Threshold limit for totality of thes.

C

Prevalence of hypertension in the Algerian Sahara Temmar et al. 2225

Blood pressure control with antihypertensive treatmentThe present analysis shows that among subjects receiving

antihypertensive treatment, 25% were well controlled,

and this percentage was similar in both black and white

individuals. Half the treated subjects had high levels of

both SBP and DBP, followed by those with high SBP and

controlled DBP, and the percentage of subjects with

uncontrolled DBP and controlled DBP was very low.

This clearly confirms that, as measured in a clinical

setting, the control of blood pressure, especially SBP,

is not obtained in a large majority of individuals. Data

from several countries have demonstrated similar results,

showing that among treated hypertensive individuals, the

proportion of those who are well controlled is less than

30% [24–27]. Interestingly, education level, sex and skin

colour had no significant influence on blood pressure

control through treatment.

Limitations of the studyOne limitation is that the present study can only suggest

but not confirm that the high prevalence of hypertension

in the population could be related to the high levels of salt

in the drinking water. Unfortunately, we do not have

urinary sodium measurements, and even if we did, it

would not be possible to detect the effects of high salt

levels in drinking water on blood pressure levels, because

in that area almost the entire population uses this water,

with very few exceptions who drink mineral or

‘filtered’ water.

The fact that blood pressure was measured three times in

only one visit is also a limitation, but we would like to

make the following comments. First, as mentioned in the

Methods section, blood pressure was measured using an

automatic device, after a 5-min rest period, and the

reported values were the mean of the second and third

measurements. As in most observational studies, subjects

had only one visit. In a subgroup of 89 subjects with high

blood pressure, measurements were repeated the day

after in order to evaluate the evolution of blood pressure.

The mean values of blood pressure in the 89 subjects

during the first visit were 164/96 mmHg. The mean

values (mean of the three measurements) during the

second visit were 160/93 mmHg. In 81 out of the 89

patients (> 90%), the diagnosis of hypertension was con-

firmed during the second visit. Although those with

normal blood pressure levels during the first visit did

not have a second visit, one might suggest that because of

the normal variability of blood pressure levels some of the

low blood pressure patients could have shown higher levels

during the second visit. The 3–4 mmHg differences in

blood pressure levels between the two visits observed in

the subgroup of 89 patients could be responsible for a

minor overestimation of the number of hypertensive

individuals in this population, but cannot explain the

high prevalence of hypertension in this population.

opyright © Lippincott Williams & Wilkins. Unauth

Clinical implicationsThe results of this study, which show a very high preva-

lence of hypertension in a Saharan population, confirm

results of other epidemiological studies that have shown

that not only is hypertension a frequent disease in devel-

oped countries, but it is also a major public health

problem in these developing countries [28,29]. These

results may also point out the harmful effects of high

sodium intake, especially in drinking water, and suggest

that the entire population, independently of age, sex,

skin colour and socio-economic level, is concerned.

AcknowledgementsThe authors would like to thank the public health direc-

tion of Ghardaia and Tamanrasset for their important

contribution.

This study was made possible with the help of INSERM

(Institut National de la Sante et de la Recherche Med-

icale), Groupe de Pharmacologie et d’Hemodynamique

Cardiovasculaire, Caisse Nationale d’Assurance Maladie

(CNAM), and the Caisse Primaire d’Assurance Maladie

de Paris (CPAM–Paris), France.

There are no conflicts of interest.

References1 Opie LH, Seedat YK. Hypertension in sub-Saharan African populations.

Circulation 2005; 112:3562–3568.2 Seedat YK. Hypertension in developing nations in sub-Saharan Africa.

J Hum Hypertens 2000; 14:739–747.3 Sever PS. Blood-pressure and its correlates in urban and tribal Africa.

Lancet 1980; 2:60–64.4 Kaminer B, Lutz WP. Blood pressure in Bushmen of the Kalahari Desert.

Circulation 1960; 22:289–295.5 Mbanya JC, Minkoulou EM, Salah JN, Balkau B. The prevalence of

hypertension in rural and urban Cameroon. Int J Epidemiol 1998; 27:181–185.

6 Edwards R, Unwin N, Mugusi F, Whiting D, Rashid S, Kissima J, et al.Hypertension prevalence and care in an urban and rural area of Tanzania.J Hypertens 2000; 18:145–152.

7 Cooper R, Rotimi C, Ataman S, McGee D, Osotimehin B, Kadiri S, et al. Theprevalence of hypertension in seven populations of west African origin.Am J Public Health 1997; 87:160–168.

8 Cooper R, Rotimi C. Hypertension in blacks. Am J Hypertens 1997;10:804–812.

9 Saunders E. Hypertension in African-Americans. Circulation 1991;83:1465–1467.

10 De Courten M, McCarty D, Zimmet P. In: Hitman G, editor. Diagnosis: thescale of the problem of furure risks, 1999 in type II diabetes, prediction andprevention. Chichester: John Wiley & Son. pp. 17–36.

11 Ibrahim MM. Hypertension surveys in the developing world. Lessons fromthe Egyptian National Hypertension Project (NHP). J Hum Hypertens1997; 11:709–726.

12 O’Brien E, Mee F, Atkins N, Thomas M. Evaluation of three devices for self-measurement of blood pressure according to the revised BritishHypertension Society Protocol: the Omron HEM-705CP, Philips HP5332,and Nissei DS-175. Blood Press Monit 1996; 1:55–61.

13 Executive Summary of the Third Report of the National CholesterolEducation Program (NCEP) Expert Panel on Detection, Evaluation, andTreatment of High Blood Cholesterol in Adults (Adult Treatment Panel III).JAMA 2001; 285: 2486–2497.

14 Franklin SS, Wilkinson IB, Cockcroft JR. Brachial and central pulsepressure, and cardiovascular risk. In: Safar ME, O’Rourke MF, editors.Arterial stiffness in hypertension. Sydney: Elsevier; 2006. pp. 225–240.

15 Stehouwer CD, Ferreira I. Diabetes, lipids and other cardiovascular riskfactors. In: Safar ME, O’Rourke ME, editors. Arterial stiffness inhypertension. Sydney: Elsevier; 2006. pp. 427–458.

orized reproduction of this article is prohibited.

C

2226 Journal of Hypertension 2007, Vol 25 No 11

16 Kaplan NM, Lieberman E. Clinical hypertension. 4th Ed. Williams andWilkins: Baltimore; 1986. pp. 71–72.

17 Worthington MG, Wendt MC, Opie LH. Sodium transport in hypertension:assessment of membrane-associated defects in South African black andwhite hypertensives. J Hum Hypertens 1993; 7:291–297.

18 Cooper R, Rotimi C. Hypertension in populations of West African origin: isthere a genetic predisposition? J Hypertens 1994; 12:215–227.

19 Burt VL, Cutler JA, Higgins M, Horlan MJ, Labarthe D, Whelton P, et al.Trends in the prevalence, awareness, treatment, and control ofhypertension in the adult US population. Data from the health examinationsurveys, 1960 to 1991. Hypertension 1995; 26:60–69.

20 M’Buyamba-Kabangu JR, Lepira B, Lijenen P, Tshiani K, Fagard R, Amery A.Intracellular sodium and the response to nitrendipine or atenolol in Africanblacks. Hypertension 1988; 11:100–105.

21 Touyz RM, Milne FJ, Reinach SG. Platelet and erythrocyte Mg2þ, Ca2þ,Naþ, Kþ and cell membrane adenosine triphosphatase activity in essentialhypertension in blacks. J Hypertens 1992; 10:571–578.

22 Hoosen S, Seedat YK, Bhigjee AI. A study of urinary and intracellularsodium and potassium, renin, aldosterone, and hypertension in blacks andIndians in Natal. Cardiovasc Drugs Ther 1990; 4 (Suppl 2):363–365.

23 Touyz RM, Milne FJ, Reinach SG. Racial differences in cell membraneATPases and cellular cation content in urban South African normotensiveand hypertensive subjects. Am J Hypertens 1993; 6:693–700.

24 Burt VL, Cutler JA, Higgins M, Horan MJ, Labarthe D, Whelton P, et al.Trends in the prevalence, awareness, treatment, and control ofhypertension in the adult US population. Data from the Health ExaminationSurveys, 1960 to 1991. Hypertension 1995; 26:60–69.

25 Marques-Vidal P, Tuomilehto J. Hypertension awareness, treatment andcontrol in the community: is the ‘rule of halves’ still valid? J Hum Hypertens1997; 11:213–220.

26 Colhoun HM, Dong W, Poulter NR. Blood pressure screening,management and control in England: results from the health survey forEngland 1994. J Hypertens 1998; 16:747–752.

27 Benetos A, Thomas F, Bean K, Gautier S, Smulyan H, Guize L. Prognosticvalue of systolic and diastolic pressure in treated hypertensive men.Arch Intern Med 2002; 162:577–581.

28 Douglas JG, Bakris GL, Epstein M, Ferdinand KC, Ferrario C, Flack JM,et al. Management of high blood pressure in African Americans: consensusstatement of the Hypertension in African Americans Working Group of theInternational Society on Hypertension in Blacks. Arch Intern Med 2003;163:525–541.

29 Cruickshank JK, Mbanya JC, Wilks R, Balkau B, Forrester T, Anderson SG,et al. Hypertension in four African-origin populations: current ‘Rule ofHalves’, quality of blood pressure control and attributable risk ofcardiovascular disease. J Hypertens 2001; 19:41–46.

opyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.