26 / 76
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 28, No 1, January/February 2017
24
AFRICA
excluded from the study. Women with known history of chronic
hypertension or diabetes mellitus were also excluded from the
study.
Endothelial function was assessed using the EndoPAT 2000
technique, which measures PAT using the reactive hyperaemia
index (RHI, arbitrary units). Briefly, after 20 minutes’ rest in
a chair inclined at an angle of about 45° at room temperature,
a blood pressure cuff was placed on the non-dominant upper
arm (study arm), while the other arm served as the control. The
hands were placed on armchair supports with the palm side
down, such that the fingers hung freely.
The EndoPAT probes were then placed on the tip of each
index finger of both hands. The probes were prevented from
touching any other finger or object, and were then electronically
inflated. The PAT signal was continuously recorded on a
personal computer during the test. Baseline pulse amplitude was
measured from each fingertip for five minutes. After baseline
recording of five minutes on each arm, arterial flow was then
interrupted in the experimental arm by rapidly inflating the
cuff to occlusion pressure of 200 mmHg or 60 mmHg plus
systolic blood pressure (whichever was higher). After exactly five
minutes’ occlusion, the cuff pressure was rapidly deflated, and
post-occlusion recording continued for another five minutes in
the experimental arm as well as the control arm.
Pulse amplitude response to hyperaemia was automatically
calculated from the hyperaemia in the finger of the experimental
arm as a ratio of post-deflation average pulse amplitude to the
baseline average pulse amplitude (i.e.
A
h
/A
h,
with
A
representing
pulse amplitude,
h
denoting hyperaemic finger). This result
was divided by the corresponding ratio from the contralateral,
control hand (i.e.
A
c
/A
c
, with
c
denoting the control finger) to
obtain the RH–PAT ratio or PAT ratio.
The EndoPAT 2000 not only measured endothelial function
with the RHI but also assessed arterial stiffness by measuring
the peripheral augmentation index (PAIx) from the radial pulse-
wave analysis. PAIx was automatically calculated as the ratio of
the difference between the early and late systolic peaks of the
waveform relative to the early peak (P
2
–P
1
/P
1
), expressed as a
percentage.
Statistical analysis
Graphpad Prism 5 (GraphPad software Inc, San Diego,
California) software was used for data analysis. Normality of
the data distribution was evaluated by the Shapiro–Wilk and
Kolmogorov and Smirnov normality tests. Data are summarised
as means
±
standard error of the mean (SEM) for normally
distributed data and medians (interquartile range, IQR) for
non-normally distributed data. The two-sample Student’s
t
-test
was used to compare means, while the Mann–Whitney
U
-test
was used to compare medians. Spearman’s correlation and
multiple regression analyses were used to determine relationships
between RHI, PAlx, baseline pulse-wave amplitude (BPWA) and
maternal blood pressure.
Secondary analysis was carried out based on whether the
cases had early- or late-onset pre-eclampsia and whether cases
and controls were HIV positive or negative. Kruskal–Wallis and
one-way ANOVA were used to compare means between the cases
and controls. Statistical significance was set at
p
<
0.05.
Results
The general characteristics of the participants are as laid out in
Table 1. As expected, the cases had significantly higher systolic,
diastolic, mean arterial and pulse pressure compared with the
controls. A significantly lower baseline heart rate was observed
in the cases compared to the controls (81.5
±
15.4 vs 87.9
±
10.8
bpm;
p
<
0.001). There were significantly more mothers with a
previous history of pre-eclampsia among the cases compared to
the controls (21.4 vs 4.5%;
p
<
0.001). There was no difference
in the percentage of nulliparous women between the cases and
controls (
p
>
0.05).
Women with pre-eclampsia were found to have significantly
lower RHI [1.70 (1.04–3.61) vs 1.81 (1.18–4.62) au;
p
<
0.05]
and log-transformed RHI [0.31 (–0.03–1.24) vs 0.48 (0.00–1.87)
au;
p
<
0.01) compared to normotensive controls. Augmentation
index at 75 bpm [12.42 (–35.79–81.76) vs 2.76 (–33.17–23.86)%;
p
<
0.0001] and BPWA [543.66 (23.44–1939.8) vs 450.56 (16.12–
1359.4) au;
p
<
0.01] were found to be higher among women with
pre-eclampsia compared to the normotensive controls, as shown
in Table 2.
Relationship between RHI, PAlx and BPWA with
maternal risk factors
On bivariate correlation analysis, there was a significant inverse
relationship between RHI and diastolic blood pressure, parity
and mean arterial pressure, and no relationship with maternal age,
body mass index (BMI), systolic blood pressure, or total number
Table 2. Vascular reactivity characteristics of the two groups
Characteristic
Controls
(
n
=
110)
Cases
(
n
=
105)
p
-value
Reactive hyperaemia index
(RHI) (au)
1.81
(1.18–4.62)
1.70
(1.04–3.61)
0.0269
Log-transformed RHI
(F-RHI)
0.48
(0.00–1.87)
0.31
(–0.03–1.24)
0.0034
Baseline pulse-wave
amplitude (au)
450.56
(16.12–1359.4)
543.66
(23.44–1939.8)
0.0021
Augmentation index
@75 (%)
2.76
(–33.17–23.86)
12.42
(–35.79–81.76)
0.0000
Table 1. General characteristics of the study population
Characteristic
Controls
(
n
=
110)
Cases
(
n
=
105)
p
-value
Maternal age (years)
25.4
±
0.5 27.0
±
0.8 0.089
Gestational age (weeks)
32.3
±
0.4 30.8
±
0.4 0.017*
BMI (kg/m
2
)
30.7
±
0.5 33.1
±
0.8 0.010*
Systolic BP (mmHg)
112.3
±
1.3 140
±
1.8 0.0000*
Diastolic BP (mmHg)
64.2
±
0.9 84.2
±
1.6 0.00008
Mean arterial pressure (mmHg)
79.1
±
1.2 102.8
±
1.5 0.0000*
Pulse pressure (mmHg)
48.2
±
1.4 55.9
±
1.4 0.0001*
Baseline heart rate (bpm)
87.9
±
1.0 81.5
±
1.5 0.0004*
Parity
0.95
±
0.1 1.42
±
0.2 0.033*
History of previous pre-eclampsia,
n
(%)
5 (4.5)
22 (21.4) 0.0003*
Nulliparity,
n
(%)
37 (33.3) 42 (40.8) 0.3221
HIV
+
,
n
(%)
28 (25.2) 38 (36.9) 0.076
Family history of HPT,
n
(%)
31 (28)
37 (36) 0.251
Family history of DM,
n
(%)
25 (23)
15 (14.6) 0.162
BMI, body mass index; HPT, hypertension; DM, diabetes mellitus.