CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 2, March 2012
AFRICA
79
socio-economic status (living area, household composition,
income and educational level).
16
These women were recruited
from among the employees of a government institution. The
dietician in this institution recruited the subjects while taking the
initial study design into account.
An attempt was made to include Caucasian women from
the same institution, but since only a very small percentage of
women employed by the institution were Caucasian, we had
to include women from other institutions as well. The research
nurse who was involved in the second phase of the study
recruited the Caucasian women, based on the characteristics of
the African women, namely, for each urban African woman who
was, for example, 30 years of age, with a BMI of 25 kg/m
2
and
employed, a similar Caucasian woman with these characteristics
was sought.
All participants lived in the Potchefstroom district, South
Africa. Exclusion criteria were pregnancy, lactation, diabetes
mellitus or an oral temperature above 37°C. CRP readings were
only available for 101 African subjects.
Following the introduction of the subjects to the experimen-
tal set up and an explanation of the procedures used, they each
signed an informed consent form. The Ethics Committee of the
North-West University approved the study, and all procedures
followed were in accordance with institutional guidelines. Each
subject received a participant sheet that guided her through the
different research stations, and was signed at each station.
All anthropometric measurements (except height and weight
measurements) were taken during the course of the evening,
after which all participants received an identical supper at
20:00, which excluded alcohol or caffeine. All subjects were
asleep before 23:00 and fasted overnight. At 6:00 the follow-
ing morning, weight, height and blood pressure measurements
were obtained, followed by blood sampling at 08:00. They then
received breakfast and personal information sheets regarding
their own blood pressure, blood glucose levels, etc, and indicat-
ing where further testing and/or treatment were necessary. All
subjects were given a small financial compensation and were
transported back to their places of work.
All data were collected from the African subjects during
March and April 2003, and the samples were assayed during
August 2003. Data collection on the Caucasian subjects was
done during August 2004 and the assays were completed during
October 2004.
Height (stature), weight and waist circumference of the
subjects were measured by a qualified anthropometrist with
calibrated instruments using standard methods (Precision Health
scale, A & D Company, Japan; Invicta Stadiometer, IP 1465,
UK; Holtain unstretchable metal tape; John Bull callipers). All
measurements were standardised and taken in triplicate.
A seven-minute resting continuous measurement of cardio-
vascular parameters was taken while in the supine position using
the Finometer device (Finapres Medical Systems
®
, Amsterdam,
Netherlands). The Beatscope
®
version 1.1 software further
calculated an integrated age-dependent aortic flow curve from
the surface area beneath the pressure/volume curve, determin-
ing each subject’s heart rate (HR), systolic (SBP) and diastolic
(DBP) blood pressure, stroke volume (SV), cardiac output (CO),
total peripheral resistance (TPR) and arterial compliance (Cw)
of the small and large arteries (Finapres Medical Systems
®
,
Amsterdam, Netherlands). Duplicate blood pressure readings
were taken using a single-headed stethoscope and a mercury
sphygmomanometer (model ALPK2) both before and after the
Finometer measurements. The first and fifth Korotkoff phases
were recorded as the SBP and DBP, respectively.
hs-CRP levels were determined using blood serum samples
that were analysed with a high-sensitivity C-reactive protein
kit from Immage
®
Immunochemistry Systems (Cat no. 474630,
Beckham Coulter, Inc). Serum cortisol was measured with
125
I
RIA Coat-a-count kit (Diagnostic Products Corporation, Cat no.
TKC01). The intra- and inter-assay coefficients of variation for
cortisol were, respectively, 7.7 and 9.8%.
Statistical analyses
Data were analysed using the software computer package
STATISTICA 9.0. Departure from normality was evaluated
through Shapiro-Wilk’s analyses and hs-CRP and cortisol were
log transformed. All data was corrected for age, smoking and
alcohol consumption. Independent
t
-tests were performed to
compare the two ethnic groups in terms of age, anthropometric,
cardiovascular and biochemical variables. Partial correlations
were performed to determine the correlations of both CRP and
cortisol with various anthropometric and cardiovascular varia-
bles, while adjusting for age, smoking and alcohol consumption.
The subjects of each ethnic group were stratified into
normotensive (NT) and hypertensive (HT) groups according to
WHO recommendations to determine their hypertensive status.
17
Significant differences in each ethnic group were determined
by ANCOVA analyses for both cardiovascular and anthropo-
metric variables, while adjusting for age, smoking and alcohol
consumption. Stepwise forward regression analyses were used in
each ethnic group to predict the relationship between cardiovas-
cular and anthropometric variables and hypertension.
Results
Table 1 represents the cardiovascular and anthropometric char-
acteristics of the African and Caucasian women. Waist circum-
ference was the only anthropometric variable that exhibited
significantly higher values in Caucasian women compared to
African women. Conversely, SBP and the vascular markers
TPR and DBP were increased in African women compared
to their Caucasian counterparts. In the Caucasian women the
cardiac markers CO and SV, and cortisol values were increased
compared to African women. Both ethnic groups exhibited
elevated CRP levels (
>
3.0 mg/l).
Table 2 represents the overall mean characteristics of the
normotensive/hypertensive Caucasian and African women
subjects after adjusting for age, smoking and alcohol usage.
There was an overall trend of a significant increase in mean
values of WC and BMI in the African women once the hyperten-
sive range (
≥
140/90 mmHg) was approached.
In Fig. 1, increased vascular responses (increased TPR and
decreased Cw) were demonstrated in African women compared
to Caucasian women. The central cardiac pattern (increased SV
and CO) seen in Fig. 2 showed a propensity to significantly
increase (
p
≤
0.05) from normotensive to hypertensive values in
Caucasian women.
In Table 3, hypertensive African women exhibited significant
positive associations between log CRP and both WC and BMI.
