20 / 66
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 25, No 6, November/December 2014
270
AFRICA
correlating with a higher blood pressure (BP) in adults and
resulted in a low treatment rate for HTN due to monetary issues.
22
Stress, in addition, was another factor related to HTN
prevalence, especially in Africa.
23
It has been shown that
psychosocial stress affects the L-arginine/nitric oxide (NO)
system, with a higher susceptibility in black Africans, which in
turn contributes to a higher risk of CVD in those individuals.
24
Therefore, a multiplicity of factors may be associated with
and contributing to a high prevalence of HTN among Africans.
The current study was undertaken to determine and quantitate
the prevalence of HTN in two countries in western sub-Saharan
Africa, namely, the Gambia and Sierra Leone.
Methods
This was a population-based, cross-sectional study performed
in the Gambia and Sierra Leone. The data were collected from
the Gambia in 2000 and from Sierra Leone from 2001 to 2003
and in 2009. The Gambia is a small country, about 11 000 km
2
in 2007, with a population of 1 705 000 by 2009.
25,26
Sierra Leone
is a larger country, about 72 000 km
2
in 2007, with a population
of 5 696 000 by 2009.
25,26
This study took place as part of the annual visit to countries
in western Africa sponsored by a medical delegation from
California. In the Gambia, the visit was to specific areas within
the capital city of Banjul, including Serrekunda, Latrikunda
and Fajikunda. In Sierra Leone, the medical delegation visited
Freetown, Kenema, Lunsar, Bonthe, Bo, Jui and Makeni.
People waited in queues to be examined in a clinic by the
team.
27
Subjects underwent a history and general physical
examination, had their blood pressure checked, and were given
medications depending on the health issues they discussed with
the healthcare providers. The current study focused on the BP
readings collected for adults aged
≥
18 years.
People coming for general examinations stayed in a waiting
area in front of the clinic to be triaged by a nurse before
being checked by a physician. BPs were measured using a
sphygmomanometer. Patients whose BP fell in the hypertensive
range (SBP
≥
140 mmHg, or DBP
≥
90 mmHg) had their BP
measured again once or twice by the physician, depending on the
initial BP. If more than one BP was recorded, an average value
was determined.
In the Gambia and Sierra Leone, one of the additional
procedures performed was echocardiography using a hand-
carried ultrasound (HCU) to assess left ventricular hypertrophy
(LVH) to prioritise HTN treatment.
27
LVH was previously found
in 65% of people with HTN.
27
Statistical analysis
All the data collected during these visits, including BP
measurements, medications prescribed, and diagnostic tests, were
recorded on a paper form and were later entered in a computerised
data spreadsheet and then de-identified. The study was reviewed
and certified by the institutional review board (IRB).
Data were analysed statistically using the
χ
2
-test, and the
p
-values calculated were classified based on
p
<
0.05 as considered
of statistical significance. Other statistical tests included the
Fisher’s exact test, Cochran–Armitage trend test, Wilcoxon rank
sum test, Student’s
t
-test and ANCOVA multivariable-model
test. The data were analysed by country prior to and following
the combination of both data sets.
Data from Sierra Leone were available for the years 2001–2003
and 2009. Differences in SBP and DBP means were assessed
across the years by analysis of covariance (ANCOVA) models.
The preliminary model was a two-way full factorial model with
factors gender and year and the gender-by-year interaction, and
age was the covariate.
In the SBP model, the gender-by-year interaction term was
significant (
p
=
0.011), so separate one-way ANCOVA models
were assessed in females and males, with age as the covariate.
In the DBP model, the gender-by-year interaction term was not
significant (
p
=
0.17); however, for comparison, separate one-way
ANCOVA models were assessed in females and males, with age
as the covariate. The least-squares means (LSmeans) for SBP and
DBP were used to present the findings.
The data were divided into three categories: all adults with
and without known recorded age (
n
=
2 615), only adults with
known age
≥
18 years old (
n
=
2 348) and only adults with known
age
≥
20 years old (
n
=
2 247). There was one female who did not
have a recorded DBP.
The first classification was used to have general demographics
for the whole population tested. The second and third
classifications were used to observe trends of SBP, DBP and
HTN prevalence with age decade, starting with 20-year-old
patients. For all results including age decade analyses, the
indications
≥
70s and
+
70s stand for the age decade 70 years and
above, which were combined together with patients over 80 years
due to the small sample size in these older groups.
Results
In total, there were 2 615 adult participants: 46.5% males (
n
=
1
215) and 53.5% females (
n
=
1 400). Because one female lacked
a recorded DBP, the total number of individuals analysed based
on SBP, DBP and HTN prevalence were 2 615, 2 614 and 2 614
individuals, respectively.
Of the overall population studied, 44.8% were hypertensive,
while mean SBP was 133.6
±
29.2 mmHg and mean DBP was
83.7
±
17.0 mmHg. For females, mean SBP was 134.3
±
29.7
mmHg and mean DBP was 84.5
±
17.5 mmHg, while 46.2%
were hypertensive. For males, mean SBP was 132.8
±
28.5
mmHg and mean DBP was 82.8
±
16.2 mmHg, while 43.2% were
hypertensive.
The
t
-test showed no significant difference in mean SBP
between males and females (
p
=
0.18). However, for mean DBP,
the
t
-test indicated a significant difference between males and
females (
p
=
0.008), with females having a higher mean DBP.
Regarding HTN prevalence, the
χ
2
-test showed that there was
no significant difference between males and females, and the
Fisher’s exact test confirmed this insignificance (
p
=
0.119 and
p
=
0.124, respectively).
From the total number of subjects in the study (
n
=
2 615), a
large proportion (
n
=
2 348) represented individuals with known
age
≥
18 years old. The demographics of this subpopulation
(Table 1) were compared across gender in terms of age, SBP and
DBP means using the
t
-test.
For mean age, males were older on average (
p
=
0.018).
For mean SBP, there was no evidence that SBP differed across
gender; 133.5 mmHg for females and 132.8 mmHg for males (
p