CARDIOVASCULAR JOURNAL OF AFRICA • Vol 21, No 5, September/October 2010
270
AFRICA
tions per minute for 10 minutes, separated and frozen at –80°C
until analysis.
An overnight (eight-hour) collected fasting urine sample was
used for the assessment of microalbuminuria to indicate target
end-organ damage (TOD). Measurement of albumin excretion
followed by means of immune precipitation, enhanced by poly-
ethylene glycol at 450 nm (reference range of creatinin:albumin
ratio
=
0–2.9 mg/
µ
mol
17
). The sequential multiple analys-
er computer (Konelab
TM
, Vantaa, Finland) calculated fasting
glucose, triglycerides, cholesterol and HDL levels.
HIV/AIDS screening was done with antibody tests, namely
the First Response
®
kit (RPM Plus, Colonia, New Jersey, USA)
and the Pareekshak test (Bhat Biotech, India) to determine the
participants’ status.
Statistical analysis
Data were analysed using the Statistica 8.0 software (Statsoft
Inc, Tulsa, USA). Normal distributions of the variables deter-
mined with the Kolmogorov-Smirnov test revealed symmetrical
data, but microalbuminuria data were not evenly distributed and
therefore logarithmic transformation was performed.
T
-tests
followed to determine significant differences in age, BMI,
alcohol consumption, smoking and PAI between coping style
groups. Chi-square tests (
χ
2
) determined the prevalence of each
factor. Analysis of covariance (ANCOVA), adjusting for alcohol
consumption as co-factor, was used to compare MS data between
AC and PC African men.
According to the International Diabetes Federation (2006),
central obesity should be present together with two or more of
the following factors, in order to diagnose the MS: increased
triglyceride levels (
>
1.7 mmol/l); reduced HDL cholesterol (
<
1.03 mmol/l in men and
<
1.29 mmol/l in women); raised fasting
plasma glucose concentrations (
>
5.6 mmol/l); and hypertension
(
>
130/
>
80 mmHg).
13
Partial correlations (adjusted for alcohol) were performed
between MS indicators and markers of target end-organ damage.
Multiple regression analyses using MS indicators as independent
variables were performed to explain microalbuminuria and carot-
id intima–media thickness of the far wall (CIMT
f
). For CIMT
f
,
additional adjustments for BP and lipids were done. Significant
values were noted as
p
≤
0.05,
r
≥
0.35, and adjusted
r
2
≥
0.25.
All statistical analyses, i.e. analysis of covariance and partial
correlations were repeated after exclusion of participants treated
with hypertensive, diabetic and statin medications. No signifi-
cant differences prevailed and therefore discussion on these
results is not warranted.
Results
In the SABPA study, 53 urban black African men were included
and stratified into AC and PC groups. Table 1 compares the char-
acteristics of the African men and MS indicators of the urban AC
and PC men. Only PC men showed higher alcohol consumption
and cholesterol values.
In Fig. 1, in AC men, waist circumference, as an essential
prerequisite of the MS (IDF, 2006), correlated positively (
p
≤
0.05) with SBP (
r
=
0.57), DBP (
r
=
0.49), glucose (
r
=
0.39) and
triglyceride levels (
r
=
0.46). Additionally, highly significance
correlations (
p
≤
0.05) were observed in AC men between WC
and TOD, respective correlations for microalbuminuria (
r
=
0.46)
and CIMT
f
(
r
=
0.53). However, in PC men (Fig. 2), correlations
existed between only WC and blood pressure.
In Table 2, in AC men, the HDL levels correlated significantly
(
p
≤
0.05) with SBP (
r
=
–0.45), DBP (
r
=
–0.40) and CIMT
f
(
r
=
–0.49). Strong correlations (
p
≤
0.05) also existed between
microalbuminuria and glucose (
r
=
0.50) as well as triglycer-
ide levels (
r
=
0.56). Another significant (
p
≤
0.05) correlation
existed between glucose levels and CIMT
f
(
r
=
0.63). In Table 3,
in PC men, HDL and triglyceride levels correlated significantly
(
r
=
–0.53;
p
≤
0.05). Correlations also existed between micro-
albuminuria and SBP as well as DBP, with significant
p
-values
(
p
≤
0.05).
In Table 4, multiple regression analyses revealed that an AC
style in urban African men was associated with most MS indica-
tors, as well as TOD, depicted by CIMT
f
and microalbuminu-
ria. Furthermore, in PC men, significant associations existed
between social support (
r
=
0.531) and augmented CIMT
f
values,
whereas SBP (
r
=
0.596) and lower WC (
r
=
–0.408) were associ-
ated with microalbuminuria.
TABLE 1. DESCRIPTIVE STATISTICSANDANCOVAS
COMPAREDACAND PC CHARACTERISTICS INCLUDING MS
AND TARGET END-ORGAN DAMAGE INDICATORS,
INDEPENDENT OFALCOHOL CONSUMPTION (MEAN
±
95% CI)
19
Urbanised men (
n
=
101)
Active coping
(
n
=
30)
Passive coping
(
n
=
23)
p
-value
Age (years)
41.77
(34.59, 48.94)
41.52
(34.22, 48.82)
0.90
Body mass index (kg/m
2
)
28.08
(22.06, 34.11)
27.08
(22.97, 31.19)
0.50
Alcohol,
n
(%)*
7 (23.33)
a
12 (52.17)
a
0.03
Smoking,
n
(%)*
8 (26.67)
8 (34.78)
0.52
High physical activity,
n
(%)*
3 (10.00)
1 (4.35)
0.56
AIDS,
n
(%)*
5 (16.67)
2 (8.70)
0.40
Anti-hypertensive drugs,
n
(%)*
4 (13.33)
b
0 (0.00)
b
0.07
Anti-diabetic drugs,
n
(%)*
1 (3.33)
0 (0.00)
0.38
Social support,
n
(%)*
14 (46.67)
8 (34.78)
0.38
Cholesterol (mmol/l)
§
4.36
(3.96, 4.76)
c
5.09
(4.63, 5.55)
c
0.02
Systolic BP (mmHg)
§
138.33
(131.72, 144.94)
133.48
(125.88, 141.08)
0.35
Diastolic BP (mmHg)
§
88.68
(84.57, 92.79)
84.20
(79.48, 88.92)
0.17
MS indicators
Hypertension,
n
(%)*
23 (76.67)
15 (65.22)
0.36
Waist circumference (cm)
§
93.65
(88.56, 98.74)
92.68
(86.83, 98.53)
0.81
Gucose (mmol/l)
§
6.09 (5.33, 6.85) 5.40 (4.53, 6.27) 0.25
Triglycerides (mmol/l)
§
1.60 (1.14, 2.05) 1.51 (0.99, 2.04) 0.81
HDL (mmol/l)
§
1.04 (0.91, 1.17) 1.07 (0.92, 1.22) 0.74
Target end-organ damage
Microalbuminuria (
µ
g/l/min)
§
1.04 (0.92, 1.16)
d
0.87 (0.73, 1.01)
d
0.08
CIMT
f
(mm)
§
0.68 (0.62, 0.74) 0.71 (0.64, 0.78) 0.49
§
ANCOVA: analysis of covariance; MS: metabolic syndrome; AC: active
coping; PC: passive coping;
±
95% CI: 95% confidence intervals;
n
: number
of participants; independent
t
-test values; *Chi-squares; HDL: high-density
lipoproteins; CIMT
f
: carotid intima–media thickness far wall. Values in bold
with identical alphabetical superscripts differ significantly,
p
≤
0.05, and in
bold italics are borderline significant.
