CARDIOVASCULAR JOURNAL OF AFRICA • Vol 22, No 2, March/April 2011
AFRICA
77
Statistical analysis
Data were presented as means
±
standard deviation. Comparison
of continuous variables between the group of sickle cell disease
patients and the control group was made with the independent
Student’s
t
-test. In order to examine the effect of anaemia on the
variables, the subjects were classified, based on the haematocrit
values, into four classes in accordance with the World Health
Organisation classification of anaemia as follows: class 1,
normal (haematocrit
≥
36%); class 2, mild anaemia (haematocrit
30–35.9%); class 3, moderate anaemia (haematocrit 21–29.9%);
class 4, severe anaemia (haematocrit 18–20.9%).
9
Inter-class differences in clinical, electrocardiographic and
echocardiographic parameters in the patients were compared by
one-way analysis of variance and
post hoc
multiple compari-
son of means using the Tukey’s honestly significant difference
test. Intra-class differences in parameters between patients and
controls in the same haematocrit class were analysed using the
independent Student’s
t
-test. The relationship between P-wave
dispersion and echocardiographic indices of left ventricular
function (while controlling for the effect of anaemia) (haemato-
crit) was examined using the partial correlation analysis.
Results
The mean ages of the patients and controls were 28.27
±
5.58
(range 18–44) and 28.37
±
5.91 (range 18–45) years, respec-
tively. There were no statistically significant age and gender
differences between patients and controls. The patients had
statistically significant lower mean values than the controls in
the measurement of height, body mass index and body surface
area (
p
<
0.001, Table 1).
P-wave duration and P-wave dispersion were significantly
higher in patients than controls (Table 2). Significant correlation
was demonstrated between P-wave dispersion and age in the
patients (
r
=
0.387;
p
=
0.031). When subsets of sickle cell anae-
mia patients and controls with comparable haematocrit values
(30–35%) were compared, the patients were found to have
significantly higher P-wave duration and dispersion than the
controls. The P-wave duration in patients and controls, respec-
tively, was 111.10
±
14.53 ms and 89.14
±
16.45 ms (
t
=
3.141;
p
=
0.006). P-wave dispersion was 64.44
±
15.86 ms in the patients
and 36.43
±
10.35 ms in the controls (
t
=
2.752;
p
=
0.013).
In order to evaluate the effect of degree of anaemia on P-wave
dispersion in the patients, the electrocardiographic parameters
were compared among the haematocrit categories, as shown
in Table 3. The haematocit values had no effect on the P-wave
duration or dispersion.
The result of a multivariate Pearson’s correlation analysis of
P-wave dispersion and echocardiographic indices of left ventric-
ular function are presented in Table 4. Of all the parameters
evaluated, significant negative correlation was found between
P-wave dispersion and left ventricular transmitral E/A ratio (
r
=
0.289;
p
=
0.023). The correlation was still significant after
controlling for the effect of anaemia (
r
=
0.285;
p
=
0.027). Such
correlation was not observed in the normal controls (
r
=
0.025;
p
=
0.859).
Discussion
This study revealed a significant increase in P-wave disper-
sion in sickle cell anaemia patients. The finding of a negative
correlation between P-wave dispersion and left ventricular E/A
ratio suggests that left ventricular diastolic function might be
deranged in patients with increased P-wave dispersion. Reduced
left ventricular relaxation and alteration in left ventricular
chamber compliance are the haemodynamic abnormality for left
TABLE 1. AGE, GENDERANDANTHROPOMETRIC DATA
Parameters
SCA
mean (SD)
Controls
mean (SD)
t
-test
p
-value
Age (years)
28.27 (5.58) 28.37 (5.91) 0.987 0.924
Gender (frequency (%))
Male
Female
Total
31 (50)
31 (50)
62
31 (50)
31 (50)
62
0.00 1.00
a
Weight (kg)
54.97 (10.61) 67.35 (8.37) 7.20
<
0.001*
Height (m)
1.62 (0.14)
1.72 (0.07) 4.960
<
0.001*
Body surface area (m
2
)
1.62 (0.03)
1.78 (0.14) 3.723
<
0.001*
Body mass index (kg/m
2
) 20.47 (2.73) 23.87 (3.22) 6.181
<
0.001*
*Statistically significant,
a
Chi-square, SCA
=
sickle cell anaemia.
TABLE 2. COMPARISON OF ELECTROCARDIOGRAPHIC
CHARACTERISTICS OF PATIENTSAND CONTROLS
Variables
Values; mean (SD)
t
-test
p
-value
SCA
Controls
Heart rate (beat/min)
80.61 (12.79) 68.98 (4.24) 6.327
<
0.001*
P-wave duration (ms) 128.0 (14.15) 90.30 (14.84) 14.189
<
0.001*
P-wave dispersion (ms) 65.7 (16.09) 34.7 (17.41) 9.014
<
0.001*
*Statistically significant, SCA
=
sickle cell anaemia.
TABLE 3. ELECTROCARDIOGRAPHIC PARAMETERS IN
SICKLE CELLANAEMIA; EFFECT OF HAEMATOCRIT LEVELS
Parameters
Haematocrit levels; mean (SD)
F
-statistic
p
-value
Mild Moderate Severe
Heart rate (beat/min)
83.56
(12.34)
77.65
(12.85)
83.27
(11.76)
1.391 0.297
P-wave duration (ms) 111.10
(14.50)
136.50
(17.31)
109.11
(16.40)
0.228 0.797
P-wave dispersion (ms) 64.42
(12.86)
64.04
(17.10)
60.02
(15.50)
0.245 0.784
*Statistically significant.
TABLE 4. MULTIVARIATE PEARSON’S CORRELATION
OF P-WAVE DISPERSIONAND ECHOCARDIOGRAPHIC
PARAMETERS IN PATIENTSAND CONTROLSWHILE
CONTROLLING FOR HAEMATOCRIT
Variables
Patients
Controls
Pearson’s
r p
-value Pearson’s
r p
-value
LVMI
–0.0489 0.712 –0.0538 0.680
Fractional shortening –0.1507 0.250 –0.1988 0.124
Ejection fraction
0.0489 0.711
0.0510 0.696
VCS
0.0607 0.645 –0.0541 0.679
E/A ratio
–0.285
0.027* 0.025
0.859
IVRT
0.1659 0.205 –0.0735 0.573
EDT
–0.01049 0.425
0.0005 0.997
Cardiac index
–0.2369 0.066 –0.0825 0.531
*Statistically significant; LVM
=
left ventricular mass, VCS
=
velocity of
circumferential shortening, IVRT
=
isovolumic relaxation time, EDT
=
E-wave deceleration time.
