Cardiovascular Journal of Africa: Vol 28 No 2 (March/April 2017)

29 / 76

CARDIOVASCULAR JOURNAL OF AFRICA • Volume 28, No 2, March/April 2017 AFRICA 95
regression models) and gender-specific AUC (from logistic
regression models) between a model with BMI + WC, BMI +
PI or BMI + WHtR to a model with BMI alone, as described
by Tuan et al.28 A change in prevalence ratio or AUC of ≥ 10%
was used as the criterion for a significant contribution of WC,
PI, or WHtR to the prediction of hypertension using BMI.
This criterion of ≥ 10% was adopted as it is commonly used to
determine a notable confounding factor.26
Data analysis was conducted using Stata and MedCalc
statistical packages. Model 1 was crude while model 2 was
adjusted for factors, such as age, smoking, alcohol consumption
and physical activity.
Results
A total of 912 individuals aged 17 years and older from the three
major cities participated and provided informed consent for the
study; 32.89% of the respondents were from Awka, 33% from
Onitsha and 34.10% from Nnewi. The overall crude prevalence
of hypertension and prehypertension in the study population
was 22.81 and 42.54%, respectively.
Tables 1 and 2 show the general characteristics of our study
population. The mean values of all the anthropometric indices
analysedweresignificantlyhigherinthewomen,withtheexception
of weight, height and WHR, when compared to the men. The
mean values of all anthropometric indices studied also increased
from normotensive participants, through prehypertensive
subjects and peaked in the hypertensive participants in both male
and female categories, with hypertension showing the highest
mean values for all anthropometric indices studied.
Table 3 presents the results of the correlation between the
anthropometric indices with blood pressure. All anthropometric
indices correlated significantly with systolic and diastolic blood
pressure. BMI had the highest correlation coefficient, while CI
had the lowest.
Table 4 lists the results of the predictive potentials of each
individual anthropometric index in discriminating between
hypertension and normotension, and between prehypertension
and normotension. For hypertension, BMI, WHtR and WC had
the strongest/highest predictive potential in both the male and
female categories (WHtR was slightly higher). BMI and WHtR
also performed relatively well in all age categories except for age
category 2 (21–25 years) for BMI, and age categories 1 and 2 (≤
20 years and 21–25 years) for WHtR. PI also showed a strong
predictive power (AUC) in this regard but was lower than that
Table 3. Correlation between blood pressure, age and anthropometric
variables
SBP (r) r2 p-value DBP (r) r2 p-value
BAI 0.18 0.03 0.0000 0.12 0.01 0.0004
BMI 0.33 0.11 0.0000 0.29 0.08 0.0000
WHtR 0.25 0.06 0.0000 0.21 0.04 0.0000
WHR 0.15 0.02 0.0000 0.19 0.04 0.0000
PI 0.27 0.07 0.0000 0.25 0.06 0.0000
WC 0.27 0.07 0.0000 0.22 0.05 0.0000
HC 0.24 0.06 0.0000 0.16 0.02 0.0000
CI 0.105 0.01 0.0014 0.08 0.01 0.0410
SBP, systolic blood pressure; DBP, diastolic blood pressure; BAI, body adipos-
ity index; BMI, body mass index; WHtR, waist–height ratio; WHR, waist–hip
ratio; PI, ponderal index; WC, waist circumference; HC, hip circumference; CI,
conicity index.
Table 4. Analysis of the predictive power of each index for
hypertension and prehypertension
Anthropometric
measures
Hypertension
AUC p-value
Prehyperten-
sion AUC p-value
BAI# Age category
1 0.535 0.8185 0.542 0.5523
2 0.534 0.5274 0.525 0.4775
3 0.583 0.0229 0.500 0.9964
4 0.626 0.0034* 0.613 0.0642
BAI# Gender
Male 0.625 0.0002* 0.581 0.0009*
Female 0.594 0.0019* 0.564 0.0329*
BMI Age category
1 0.727 0.0001* 0.563 0.3575
2 0.542 0.4923 0.571 0.0399*
3 0.686 0.0001* 0.622 0.0008*
4 0.589 0.0434* 0.674 0.0133*
BMI Gender
Male 0.698 0.0001* 0.659 0.0001*
Female 0.622 0.0001* 0.609 0.0002*
WHtR Age category
1 0.511 0.9172 0.559 0.3915
2 0.556 0.2618 0.515 0.6764
3 0.631 0.0002* 0.560 0.1080
4 0.601 0.0209* 0.615 0.1408
WHtR Gender
Male 0.682 0.0001* 0.636 0.0001*
Female 0.624 0.0001* 0.572 0.0163*
WHR# Age category
1 0.623 0.3645 0.614 0.0839
2 0.502 0.9693 0.501 0.9817
3 0.643 0.0001* 0.619 0.0012*
4 0.525 0.5685 0.662 0.0876*
WHR# Gender
Male 0.645 0.0001* 0.562 0.0531*
Female 0.570 0.0208* 0.554 0.0753*
WC Age category
1 0.508 0.9433 0.591 0.1634
2 0.550 0.3858 0.556 0.1038
3 0.607 0.0031* 0.551 0.1645
4 0.542 0.3421 0.580 0.3115
WC Gender
Male 0.692 0.0001* 0.645 0.0001*
Female 0.616 0.0001* 0.584 0.0046*
PI# Age category
1 0.680 0.0008* 0.652 0.0223*
2 0.525 0.6617 0.547 0.1818
3 0.679 0.0001* 0.607 0.0043*
4 0.642 0.0008* 0.670 0.0132*
PI# Gender
Male 0.670 0.0001* 0.639 0.0001*
Female 0.619 0.0001* 0.599 0.0008*
HC# Age category
1 0.530 0.8371 0.649 0.0193*
2 0.552 0.3242 0.543 0.2095
3 0.565 0.0866 0.507 0.8576
4 0.535 0.4372 0.567 0.3003
HC# Gender
Male 0.646 0.0001* 0.602 0.0008*
Female 0.592 0.0036* 0.584 0.005*
CI# Age category
1 0.618 0.2350 0.556 0.3980
2 0.503 0.950 0.520 0.5602
3 0.527 0.4571 0.520 0.5915
4 0.539 0.3878 0.541 0.6106
CI# Gender
Male 0.592 0.0082* 0.541 0.2032
Female 0.558 0.0528 0.533 0.2773
*Statistically significant at p < 0.05; #AUC significantly different from that of
BMI. Age category 1 = ≤ 20 years, 2 = 21–25 years, 3= 26–40 years, 4 = ≥ 41
years. BAI, body adiposity index; BMI, body mass index; WHtR, waist–height
ratio; WHR, waist–hip ratio; WC, waist circumference; PI, ponderal index;
HC, hip circumference; CI, conicity index.