Cardiovascular Journal of Africa: Vol 27 No 6 (November/December 2016)

59 / 76

CARDIOVASCULAR JOURNAL OF AFRICA • Volume 27, No 6, November/December 2016 AFRICA 389
Results
In Table 1, the black African group portrayed a more vulnerable
cardiometabolic profile than the Caucasians. They consumed
more alcohol, had higher BP and mean pre-diabetes (HbA1c)
levels, lower BDNF levels (p < 0.001), larger cortisol:BDNF
ratios (p = 0.012) and a higher mean 24-hour hypertensive state.
Caucasians were more physically active compared to the Africans.
Both ethnic groups’ cortisol levels were within the normal range of
138–635 nmol/l but the Africans’ cortisol levels showed a tendency
towards lower levels (p = 0.093) than their Caucasian counterparts.
In Table 2, ANCOVA analyses, considering a priori covariates,
showed that the cortisol level was lower in African men,
while BDNF was lower in African women compared to their
Caucasian counterparts. The African gender groups showed
increased hyperglycaemia, low-grade inflammation, 24-hour BP
values and heart rate compared to their Caucasian counterparts.
Pearson correlations showed inverse associations between
silent ischaemia and cortisol:BDNF (r = 0.34; p = 0.001) in the
African male cohort but not in any of the other ethnic gender
groups (data not shown). When considering a priori covariates
(Tables 3, 4), forward stepwise linear regressions confirmed
similar trends with a stronger association between silent ischaemia
[Adj R2 0.22; β 0.40 (0.2–0.6); p < 0.01] and cortisol:BDNF ratio
in African men, but not in any of the other ethnic gender groups.
Cortisol level was positively associated with HbA1c level, and
24-hour BP with a tendency for silent ischaemia (p = 0.07) in the
African men only. No change in the outcome was demonstrated
after adjustment for HIV-positive status, hypertension and
diabetes medication use.
Discussion
Our objectives were to investigate associations between cortisol,
the cortisol:BDNF ratio and cardiometabolic risk markers,
Table 1. Characteristics of a South African bi-ethnic gender cohort
Variables
Africans
(n = 197)
Caucasians
(n = 209) p-values
Confounders
Age (years) 44.4 ± 8.2 45.0 ± 10.9 0.49
Body mass index (kg/m2) 30.1 ± 7.0 27.6 ± 5.9 < 0.001
Body surface area (m2) 1.9 ± 0.2 2.0 ± 0.3 < 0.001
Physical activity (kcal/day) 2670 ± 794.4 3112 ± 1596.5 < 0.001
Cotinine (ng/ml) 27.5 ± 61.3 22.71 ± 77.5 0.5
γ -Glutamyl transferase (U/l) 66.3 ± 83.0 26.91 ± 33.9 < 0.001
Potential cardiometabolic risk markers
Cortisol (nmol/l) 358.03 ± 151.63 384.11 ± 159.9 0.093
BDNF (pg/ml) 1411.6 ± 652.3 1687.3 ± 888.1 < 0.001
Cortisol:BDNF ratio 126.6 ± 114.4 102.7 ± 71.6 0.012
C-reactive protein (mg/l) 8.55 ± 10.56 3.1 ± 3.88 < 0.001
Cholesterol (mmol/l) 4.6 ± 1.16 5.5 ± 1.28 < 0.001
HbA1c (%) 6.1 ± 1.2 5.5 ± 0.42 < 0.001
24-h SBP (mmHg) 133 ± 16 124 ± 12 < 0.001
24-h DBP (mmHg) 83 ± 11 77 ± 8 < 0.001
12-lead ECG HR (bpm) 68 ± 13 66 ± 11 0.045
Silent ischaemic events 6.0 ± 15.56 2.5 ± 5.94 0.003
Hypertension, n (%) 43 (26.54) 18 (9.33) < 0.001
Medications
Hypertensive treatment, n (%) 69 (35.03) 27 (12.92) < 0.001
Statins, n (%) 2 (1.23) 9 (4.67) 0.05
Diabetes medication, n (%) 10 (5.08) 2 (0.96) 0.01
CRP > 3 mg/l, n (%) 106 (65.35) 39 (20.97) < 0.001
HIV status, n (%) 19 (9.5) 0 < 0.001
Values presented as arithmetic mean ± SD.
BDNF, brain-derived neurotrophic factor; HbA1c, glycated haemoglobin; 24-h
hypertension (SBP ≥ 130 mmHg and/or DBP ≥ 80 mmHg); HR, heart rate;
CRP, C-reactive protein; physical activity, 24-h total energy expenditure, consid-
ering resting metabolic rate; n, prevalence (%).
Table 2. Comparing differences in cardiometabolic risk markers in ethnic male and female groups
Risk markers African men (n = 99) Caucasian men (n = 101) African women (n = 98) Caucasian women (n = 108)
Unadjusted cardiometabolic risk markers
γ -Glutamyl transferase (U/l) 84.83 (70.1–99.5) 34.83 (20.2–49.4)** 46.9 (35.9–57.9) 19.9 (9.6–30.2)**
Cholesterol (mmol/l) 4.64 (4.4–4.9) 5.63 (5.4–5.9)** 4.4 (4.1–4.7) 5.57 (5.3–5.8)**
C-reactive protein,(mg/l) 5.93 (4.6–7.2) 1.71 (0.4–3.0)** 11.14 (9.4–12.8) 4.41 (2.8–6)**
HbA1c (%) 6.29 (6.1–6.5) 5.60 (5.4–5.8)** 5.85 (5.7–6.1) 5.40 (5.2–5.6)**
Adjusted cardiometabolic risk markers
24-h SBP (mmHg) 140 (137–142) 125 (123–128)** 128 (126–131) 121 (119–123)**
24-h DBP (mmHg) 89 (87–91) 78 (77–80)** 79 (77–80) 74 (72.7–76)**
24-h heart rate (bpm) 79 (77–82) 72 (70–74)** 80 (78–82) 76 (73.7–77)**
Silent ischaemic events, score 10.3 (6.9–13.6) 1.1 (0.02–4.4)** 2.8 (1.7–4.0) 3 (1.9–4.0)
Cortisol (nmol/l) 364.38 (334.59–394.2) 410.41 (380.86–440.0)* 343.95 (308.5–379.4) 368.87 (335.7–402.1)
BDNF (pg/ml) 1250.41 (1095.8–1405.1) 1426.28 (1272.9–1579.7) 1599.62 (1429–1770.3) 1925.77 (1765.3–2086.3)**
Cortisol:BDNF ratio 142.8 (118.2–167.4) 139.67 (115.3–164.1) 96.16 (82.9–109.4) 80.43 (67.9–92.9)
Values depicted as mean (± 95% confidence interval) and proportions as n (%). Adjustments were made for age, body surface area, log physical activity, log cotinine and
log γ-GT. HbA1c, glycated haemoglobin; BDNF, brain-derived neurotrophic factor; DBP, diastolic blood pressure; SBP, systolic blood pressure.
*p ≤ 0.05; **p ≤ 0.01.
Table 3. Independent associations between cardiometabolic risk
markers, cortisol as well as cortisol:BDNF in a South African cohort
South African cohort (n = 406)
HbA1c Silent ischaemia
β (95% CI) p-value β (95% CI) p-value
Adjusted R2 0.16 0.10
Cortisol 0.1 (0.0–0.2) 0.03 0.26 (0.2–0.4) < 0.01
Ethnicity –0.34 (–0.4– –0.2) < 0.01 –0.13 (–0.2–0.0) < 0.01
Gender –0.09 (–0.2–0.0) 0.08 –
Age 0.15 (0.1–0.2) < 0.01 0.11 (0.0–0.2) 0.02
Body surface area 0.17 (0.1–0.3) < 0.01 –
Adjusted R2 0.15 0.10
Cortisol:BDNF – 0.26 (0.2–0.4) < 0.01
Ethnicity – –0.13 (–0.2–0.0) < 0.01
Age – 0.11 (0.0–0.2) 0.02
HbA1c, glycated haemoglobin. Additional covariates included: log physical
activity, log cotinine levels, log gamma glutamyl transferase. Where ethnicity
(1 = African, 2 = Caucasian); gender (1 = men, 2 = women).