Cardiovascular Journal of Africa: Vol 27 No 3 (May/June 2016)

57 / 102

CARDIOVASCULAR JOURNAL OF AFRICA • Volume 27, No 3, May/June 2016 AFRICA 179
and LDL-C concentrations than white women, which remained
significant after adjusting for age and FMI.
Table 3 shows the partial correlations (adjusted for age
and FMI) between body fatness and its distribution and
cardiometabolic risk factors for the black and white women
individually and combined. In summary, age-adjusted total body
fat, as defined by FMI, was positively associated with plasma
glucose concentrations, measures of IR (fasting insulin and
HOMA-IR), TG and LDL-C concentrations, and negatively
associated with HDL-C concentrations in both black and white
women. FMI was positively associated with TC concentrations
in white women only.
In black and white women, increased central FM and
reduced lower-body fat correlated with measures of IR (fasting
insulin and HOMA-IR). In black women only, greater trunk
FM and lower leg FM were associated with increased fasting
plasma glucose concentrations. Notably these associations
with glucose concentrations and trunk FM were significantly
different between black and white women. In white women only,
increased abdominal SAT was associated with reduced fasting
glucose concentrations, and this association differed significantly
between black and white women.
In both black and white women, reduced lower-body FM,
characterised by leg FM, was associated with TG concentrations.
In the black women, higher trunk FM, and in the white women,
higher VAT was associated with TG concentrations. In addition,
in both the black and white women, trunk FM was associated
with reduced HDL-C concentrations. By contrast, higher leg
FM was associated with increased HDL-C concentrations in
both the black and white women. In white women only, increased
VAT and lower-leg FM were associated with increased LDL-C
concentrations. There were no associations between arm FM and
any metabolic risk factor in both black and white women.
We further explored the confounding effects of various
lifestyle factors, including contraceptive use, smoking, physical
activity and alcohol consumption on metabolic risk. In summary,
in black women, IR [2.3 (1.3–3.7) vs 1.8 (1–3.2) mU/l, p = 0.04]
was higher and HDL-C concentrations [1.1 (0.9–1.4) vs 1.3 (1.1–
1.6) mmol/l, p < 0.01] were lower in women on contraception
than those not on it. Despite few black women consuming
alcohol, consumption was positively associated with serum
HDL-C concentrations (r = 0.20, p < 0.05) in black women. In
white women, TC [5 (4.5–5.5) vs 4.5 (4.1–5.2) mmol/l, p = 0.01],
TG [1 (0.7–1.4) vs 0.8 (0.6–1.2) mmol/l, p < 0.01] and HDL-C
[1.7 (1.5–2) vs 1.5 (1.3–1.8) mmol/l, p < 0.01] concentrations
were higher in women on contraception than those not on it,
while lower MVPA was associated with higher fasting insulin
concentrations ( r = –0.19, p < 0.05) and HOMA-IR ( r = –0.18,
p < 0.05).
In separate models for black and white women, we then
used backward stepwise regression to determine the factors
that accounted for the greatest variance in cardiometabolic risk
Table 1. Body composition and body fat distribution
of black and white women
n
Black women
median (IQR) n
White women
median (IQR)
p-value
adjusted
for age
Body composition
Height (m) 288 1.60 (1.56–1.64) 197 1.67 (1.60–1.70) < 0.001
Weight (kg) 288 80.4 (60.9–96.2) 197 73.9 (62.0–94.1) 0.02
BMI (kg/m 2) 288 31.7 (23.6–37.2) 197 26.6 (22.4–33.2) < 0.001
Fat (kg) 288 33.7 (19.4–44.3) 197 26.7 (17.1–40.3) < 0.001
Fat (%) 288 42.1 (32.7–46.9) 197 36.5 (28.9–43.9) < 0.001
FMI (kg/m 2) 288 13.3 (7.6–17) 197 9.9 (6.4–14.5) < 0.001
Body fat distribution
Waist (cm) 288 94.8 (77.3–108.6) 197 88 (78–101.5) < 0.001
Trunk FM (kg) 288 14.1 (7.4–20.6) 197 12.2 (7.4–20.3) 0.01
Trunk FM (% FM) 288 42.1 (36.7–46.7) 197 45.5 (40.7–49.7) < 0.001
Leg FM (kg) 288 13.7 (9.3–18.01) 197 10.7 (7.5–15.9) < 0.001
Leg FM (% FM) 288 44.3 (39.5–49.4) 197 41.4 (37.5–45.7) < 0.001
Trunk FM/leg FM 288 0.95 (0.74–1.2) 197 1.1 (0.90–1.3) < 0.001
Arm FM (kg) 288 3.8 (1.9–4.9) 197 2.9 (1.8–4.6) < 0.001
Arm FM (%) 288 10.7 (9.3–12.3) 197 10.8 (9.7–11.9) 0.9
VAT (cm 2) 222 71 (47–102) 153 80 (60–124) 0.04
SAT (cm 2) 220 442 (212–577) 150 297 (169–460) < 0.001
VAT/SAT 220 0.20 (0.14–0.27) 150 0.31 (0.23–0.42) < 0.001
Values presented as median and interquartile range (IQR);
p- values for one-way ANCOVA adjusting for age.
BMI, body mass index; FMI, fat mass index; WC, waist circumference; FM, fat
mass; VAT, visceral adipose tissue; SAT, subcutaneous adipose tissue.
Table 2. Cardiometabolic risk factors of black and white women
n
Black women
median
(IQR) n
White women
median
(IQR)
p-value
adjusted
for age
p-value
adjusted
for age
and FMI
Glucose (mmol/l) 280 4.5 (4.2–4.9) 196 4.7 (4.4–4.9) 0.08 < 0.001
Insulin (mU/l) 287 9.8 (5.6–16.6) 197 6.9 (4.6–10.8) < 0.001 0.27
HOMA-IR 279 2.1 (1.1–3.4) 196 1.5 (1.0–2.2) < 0.001 0.59
TC (mmol/l) 274 3.9 (3.3–4.5) 197 4.7 (4.1–5.3) < 0.001 < 0.001
TG (mmol/l) 274 0.7 (0.5–1.0) 197 0.9 (0.6–1.2) < 0.001 < 0.001
HDL-C (mmol/l) 273 1.2 (1.0–1.6) 197 1.6 (1.4–1.9) < 0.001 < 0.001
LDL-C (mmol/l) 273 2.2 (1.7–2.8) 197 2.6 (2.1–3.3) < 0.001 < 0.001
Values presented as median and interquartile range (IQR);
p-values for one-way ANCOVA adjusting for age and age and FMI.
HOMA-IR, homeostasis model for insulin resistance; TC, total cholesterol; TG,
triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density
lipoprotein cholesterol.
Table 3. Correlations between body fatness and its distribution
and cardiometabolic risk factors in black and white women
and the combined sample
Glucose Insulin
HOMA-
IR TG TC HDL-C LDL-C
FMI (kg/m2 ) B 0.26∆ 0.53∆ 0.54∆ 0.23∆ 0.035 –0.31∆ 0.15*
W 0.23∆ 0.61∆ 0.61∆ 0.35∆ 0.26∆ –0.39∆ 0.36∆
All 0.25∆ 0.59∆ 0.60∆ 0.30∆ 0.13∆$ –0.34∆ 0.25∆$
Trunk FM (kg) B 0.34∆ 0.30∆ 0.34∆ 0.30∆ –0.060 –0.23∆ –0.039
W 0.069 0.29∆ 0.29∆ 0.099 0.043 –0.18∆ 0.12
All 1.04∆$ 1.16∆ 1.23∆ 0.92∆ –0.09$ –0.87∆ 0.09$
VAT (cm2) B 0.12 0.24∆ 0.27∆ 0.078 –0.034 –0.099 0.0011
W –0.039 0.21∆ 0.19* 0.21* 0.10 –0.25∆ 0.17*
All 0.10$ 0.33∆ 0.33∆ 0.20∆ –0.04$ –0.28∆ 0.05$
SAT (cm2) B 0.094 –0.023 0.019 –0.11 –0.039 0.15* –0.071
W –0.25∆ –0.0002 –0.017 –0.016 0.15 0.16 0.079
All 0.13$ 0.18 0.20 –0.06 –0.20$ –0.06 –0.19$
Leg FM (kg) B –0.16∆ –0.33∆ –0.34∆ –0.23∆ –0.011 0.15* –0.015
W –0.12 –0.25∆ –0.26∆ –0.16* –0.097 0.19∆ –0.15*
All –0.38∆ –0.67∆ –0.69∆ –0.52∆ –0.09 0.39∆ –0.15
Values are presented as partial correlation coefficients adjusted for age and FMI
(except for FMI); ∆p < 0.01 and * p < 0.05. ‘All’ values are presented as beta-coeffi-
cients adjusted for age, FMI and ethnicity. $ Ethnic × body composition interaction.
FMI, fat mass index; VAT, visceral adipose tissue; SAT, subcutaneous adipose
tissue; FM, fat mass; HOMA-IR, homeostasis model of insulin resistance; TG,
triglycerides; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol;
LDL-C, low-density lipoprotein cholesterol.