CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 9, October 2012
AFRICA
503
scale (Nagata bw-1122h). Participants were asked to remove
footwear and only wore their school physical education outfit
that included shorts and a short-sleeve T-shirt.
Seated resting heart rate, measured to the nearest beat per
minute (bpm). Resting blood pressure, measured to the nearest
millimetre of mercury (mmHg) were recorded after a resting
period of ten minutes.
Body fat percentage was determined using the four-site
skinfold method.
22
Triceps, biceps, supra-iliac and sub-scapular
skinfolds were measured on the right side of the body using
Harpenden
©
(
West Sussex, UK: Quality Measurement, Ltd)
skinfold calipers. Each site was measured twice to the nearest
millimetre and the mean value was recorded. Circumferences
at the waist (narrowest part of the torso) and hip (level of
maximum extension of the buttocks) were measured to the
nearest millimetre with a tape measure and the waist-to-hip ratio
was calculated.
23
CRF was assessed using the 20-m multi-stage shuttle run test
that predicts an individual’s maximal aerobic capacity (VO
2
max
).
24
This test has been shown to be an appropriate predictor of CRF
for the age groups participating in the study.
25
Statistical analysis
The Shapiro-Wilk algorithm was used to assess whether the body
composition, cardiovascular and CRF variables, and salivary
CRP concentration and secretion rates demonstrated a normal
distribution. Both salivary CRP concentration and secretion rate
were found to have skewness and kurtosis and so the values were
log transformed before analysis.
Mean and standard deviations were calculated for the
log-transformed CRP values as well as the demographic, body
composition, cardiovascular and CRF variables. One-way
ANOVA examined the differences in body composition,
cardiovascular andCRFvariables, and salivaryCRPconcentration
and secretion rates by BMI categories. Tukey’s
post hoc
analysis
was completed when appropriate.
Binary logistic regression was performed to determine the
unadjusted and adjusted odds ratios for the prediction of elevated
salivary CRP concentration and secretion rate by specific risk
factors. The unadjusted odds ratio indicates the combined
contribution of the three risk factors to elevated salivary CRP,
while calculation of the adjusted odds ratio adjusts for the inter-
relationships between them.
The choice of these risk factors was based on an
a priori
decision. The non-log transformed salivary CRP concentration
and secretion rate data as well as the risk factors were divided
into quartiles. The data for each child was then coded depending
on whether they were positive or negative for the presence of
each of the risk factors (defined below) and either positive or
negative for the presence of a salivary CRP concentration or
secretion rate
≥
75
th percentile.
Elevated salivary CRP was entered as the dependant variable
with the risk factors entered into the model as covariates, using
the enter method (unadjusted odd ratios) backward stepwise
selection technique (adjusted odds ratios), based on likelihood
ratios with entry and exit probabilities set to 0.05 and 0.1,
respectively. The risk factors selected included the following:
•
Not meeting percentage body fat recommendations of
≤
25%
for boys and
≤
32%
for girls. The body fat recommendations
are based on the Fitnessgram Health Fitness Zone standards
for body composition.
25
•
Being overweight/obese:
≥
85
th percentile for BMI, which
was calculated as mass (kg) divided by height (m) squared.
Growth charts have been published by the Center for Disease
Control and Prevention (CDC) for BMI in boys and girls,
two to 20 years old. These charts are percentiles showing the
distribution of BMI at a given age and can be used to identify
children who are at risk of being overweight (BMI > 85th
percentile) or obese (BMI > 95th percentile).
25
According
to these CDC BMI-for-age standards, the participants were
grouped into the following CDC BMI-for-age categories:
normal weight (
<
85
th percentile), overweight (
≥
85
th
percentile to
<
95
th percentile), and obese (
≥
95
th percen-
tile).
26
These cut-off points are unchanged from the 1998
expert committee recommendations and CDC and Institute of
Medicine recommendations.
24
•
Demonstrating poor CRF (
≤
50
th percentile) for predicted
VO
2
max
:
≤
24.66
ml/kg/min. There are currently no standard-
ised VO
2
max
data that can be used to categorise the CRF of
children.
27
However, for the present study, the authors divided
the VO
2
max
data into quartiles and used VO
2
max
values
≤
50
th
percentile to represent poor CRF in the children.
Statistical analysis was performed using SAS (version 9.2,
Research Triangle, NC, USA) and SPSS 19.0 (SPSS) software.
Significance level was set at
p
<
0.05.
Results
Demographic, body composition, cardiovascular and CRF data
and logged salivary CRP concentration and secretion rates for
the children were divided according to three BMI categories
(
normal weight, overweight and obese). They are presented as
means
±
standard deviations.
The outcomes of the one-way ANOVAs examining the
differences by BMI categories are indicated in Table 1. There
were significant differences in age (F
=
3.37,
df
=
2, 167,
p
=
0.037),
height (F
=
9.19,
df
=
2, 167,
p
<
0.001),
mass (F
=
127.52,
df
=
2, 167,
p
<
0.001),
BMI (F
=
248.89,
df
=
2, 167,
p
<
0.001),
body fat percentage (F
=
185.14,
df
=
2, 167,
p
=
0.001),
systolic blood pressure (SBP) (F
=
19.94,
df
=
2, 167,
p
<
0.001),
diastolic blood pressure (DBP) (F
=
23.95,
df
=
2, 167,
p
<
0.001),
VO
2
max
(
F
=
17.76,
df
=
2, 167,
p
<
0.001),
salivary CRP
concentration (F
=
5.89,
df
=
2, 167,
p
=
0.0034)
and salivary
CRP secretion rate (F
=
5.90,
df
=
2, 167,
p
=
0.0033)
between
children of different BMI categories.
Tukey’s
post hoc
analyses revealed that obese children had
significantly (
p
<
0.05)
higher body fat percentages, SBP and
DBP, as well as a significantly (
p
<
0.05)
lower aerobic capacity
(
VO
2
max
)
than both normal weight and overweight children. In
addition, there was a significant difference in salivary CRP
concentration and salivary CRP flow rate between normal
weight and obese children (
p
<
0.05).
There were no significant
differences between the normal versus overweight and overweight
versus obese categories for salivary CRP concentration or
secretion rate.
The non-log transformed salivary CRP concentrations ranged
from 217.99 to 24562.94 pg/ml (median
=
700.10
pg/min and
interquartile range
=
546.45–1372.14
pg/ml) and salivary CRP
secretion rate ranged from 113.90 to 20694.28 pg/min (median
