Cardiovascular Journal of Africa: Vol 26 No 5 (September/October 2015)

37 / 56

CARDIOVASCULAR JOURNAL OF AFRICA • Vol 26, No 5, October/November 2015 AFRICA 35
operating characteristic (ROC) analysis was used to maximise
the sensitivity and specificity of the other anthropometric indi-
ces.
Results: A total of 866 participants aged 18 years and above
were studied, 381 (44.0%) were males and 485 (56.0%) were
females. The prevalence of obesity was 11.2% using a BMI of
≥ 30 kg/m2. Using cut-off values of WHtR ≥ 0.5 irrespective of
gender, WC > 102 cm for men and > 88 cm for women and WHR
> 0.95 for men and > 0.8 for women, the prevalence of obesity
was 75.4, 13.9 and 54.3%, respectively. WHtR had the largest
area under curve (R = 0.862, 95% CI: 0.803–0.923, p < 0.001)
while it was 0.824 (95% CI, 0.764–0.885, p < 0.001) for WC and
0.622 (95% CI, 0.575–0.706, p < 0.001) for WHR.
The prevalence of obesity using the ROC-derived values for
WHtR, WC and WHR was 23.09, 26.79 and 44.23%, respective-
ly. The ROC indicates that in the setting of the study, the WHtR
cut-off value should be 0.6, while WC should be 1.026 times and
WHR 0.919 times the aforementioned cut-off values. Optimal
cut-off values being proposed among adult Nigerians are
WHtR ≥ 0.6, WC > 104 cm and > 90 cm for males and females
and WHR > 0.87 and > 0.73 for males and females, respectively.
Conclusion: WHtR is the best screening tool for diagnosis of
abdominal obesity in our setting. Optimal cut-off values for
WHtR and WC are higher, while WHR is lower compared to
existing values for the European population.
OBESITY INDICES AND PERIPHERAL ARTERY
DISEASE MEASURED BY ANKLE–BRACHIAL INDEX IN
NIGERIAN OUT-PATIENTS
Umuerri Ejiroghene*, Obasohan Austine1
*Delta State University Teaching Hospital, Oghara, Nigeria;
1University of Benin Teaching Hospital, Benin, Nigeria
Introduction: Peripheral artery disease (PAD), an important
component of the cardiovascular triad, has been linked with
obesity as one of the risk factors for its development. The risk
posed by obesity however varies, depending on the indices
measured.
Methods: We compared the relationship of measurements of
central and visceral obesity [waist circumference (WC) and
waist–hip ratio (WHR)] versus that of general obesity [body
mass index (BMI)] in the development of PAD among Nigerians
with hypertension and/or diabetes mellitus. PAD was diagnosed
when the ankle–brachial index (ABI) was < 0.9 in either limb.
Results: A total of 541 patients (194 males and 347 females)
with a mean age of 58.4 (± 0.46) years were studied. The mean
BMI, WC and WHR were 27.8 (± 0.222) kg/m2, 96.8 ( ± 0.515)
cm and 0.941 (± 0.003), respectively. Although the mean BMI,
WC and WHR were higher in patients with PAD than those
without PAD, the difference was only statistically significant
for WC and WHR (p = 0.003 and p = 0.016) but not BMI ( p =
0.151). However, the difference in mean BMI was statistically
significant in patients < 60 years (p = 0.015) but not in those ≥
60 years (p = 0.953).
Conclusion: This study has shown that in Nigerian Africans,
measurement of central and visceral obesity were more related
to the development of PAD than was BMI, which is a measure
of general obesity. This lack of significance was probably due
to the fact that PAD occurred more in older people as there was
a significant relationship with PAD in people younger than 60
years old.
PREVALENCE AND SOCIO-DEMOGRAPHIC CORRE-
LATES OF OBESITY AND OVERWEIGHT IN DELTA
STATE, NIGERIA: A RURAL–URBAN COMPARISON
Umuerri Ejiroghene*, Ayandele Omotola
Delta State University Teaching Hospital, Oghara, Nigeria;
Introduction: Obesity is a lifestyle disease associated with multi-
ple adverse health conditions such as type 2 diabetes mellitus,
cardiovascular disease and some cancers. Its prevalence is on
the increase globally, partly because of urbanisation. This study
explored the differences in prevalence of overweight and obesity
and their association with socio-demographic characteristics in
rural and urban populations in Delta State, Nigeria
Methods: A cross-sectional survey was carried out of house-
holds in Jesse (rural) and Warri (urban) using the modified
WHO Steps.
Results: There was a total of 866 respondents, 44.0 and 56.0%
from rural and urban populations, respectively. The male versus
female distribution was 49.9 vs 50.1% (rural) and 39.4 vs 60.6%
(urban) (χ2 = 9.525, p = 0.002). The mean age (± SD) was 47.1
( ± 19.0) years (rural) and 38.9 (± 12.2) years (urban) (t = 7.332,
95% CI: 6.004–10.396; p < 0.001). The difference in educational
status between rural and urban populations was significant (χ 2
= 308.123; df = 4; p < 0.001). The mean BMI was 23.05 (± 6.9)
kg/m2 (rural) and 24.98 (± 5.6) kg/m2 (urban) (t = 1.936, 95% CI:
1.080–2.792; p = 0.015).
The overall prevalence of obesity and overweight was 11.2 and
20.8%, respectively, with urban being higher than rural (15.9 and
23.7% vs 5.2 and 17.1%, respectively). The prevalence of over-
weight was higher in females than males in both urban and rural
settings (urban: 26.2 vs 19.9%; rural: 17.3 vs 16.8%). Differences
in BMI categories between urban and rural settings was found
only among females (females: χ2 = 29.800, df = 3, p < 0.001; males:
χ2 = 6.191, df = 3, p = 0.103). The prevalence of overweight and
obesity was highest among middle-aged (40–64 years) respond-
ents compared with the young and elderly in both rural (19.7 vs
6.6%) and urban (31.1 vs 20.9%) and the difference in high BMI
( ≥ 25 kg/m2) between urban and rural setting in this age group was
statistically significant (χ 2 = 26.889, df = 3, p < 0.001).
The prevalence of overweight and obesity was higher among
rural participants with secondary education than those with
primary or no formal education. The reverse was the case for
urban participants. The urban–rural differences in the associa-
tion between educational status and prevalence of obesity and
overweight (≤ primary: χ2 = 24.861, df = 3, p < 0.001; secondary:
χ2 = 8.501, df = 3, p < 0.037).