Baseline characteristics of the subjects and controls
A total of 120 children with SCA and an equal number of age- and sex-matched controls were enrolled in this study. There were 67 male and 53 female subjects and controls, for a male:female ratio of 1.3:1. The respondents’ median age was 7 years (IQR 5.0-11.0). Both subjects were similar with respect to body surface area and diastolic blood pressure. However, SCA patients had higher systolic BP and oxygen saturation (Table 1).
Table 1. Comparison of Several Basic Clinical Parameters in SCA Subjects with Age- and Sex-Matched Controls
Parameter
Median (IQR)
|
SCA subjects
|
Hb AA controls
|
p- value
|
Systolic BP (mmHg)
|
100.00 (90-110)
|
90.00 (80-100)
|
0.003*
|
Sat O2 (%)
|
97.00 (93-98)
|
99.00 (98-99)
|
<0.001*
|
Diastolic BP (mmHg)
|
60.00 (50-70)
|
60.00 (50-70)
|
0.13
|
BSA (/m2)
|
0.81 (0.7-1.0)
|
0.85 (0.7-1.0)
|
0.34
|
Keys
IQR- Interquartile range
SCA- Sickle cell anaemia
|
BP- Blood pressure
mmHg- millimetres of mercury
|
BSA- Body surface area
m2- meter squared
*- Statistically significant
|
Clinical and laboratory parameters
Children with SCA were significantly different from the controls with respect to the presence of jaundice (83.3%), pallor (60.0%) and a history of blood transfusion (55.0%) (p<0.001).
SCA patients had significantly lower median haemoglobin and foetal haemoglobin levels than controls (Table 2).
Table 2. Comparison of Selected Haematologic Parameters in SCA Subjects with Age- and Sex-Matched Controls
Parameter
Median (IQR)
|
SCA subjects
|
Hb AA controls
|
p- value
|
Hb F (%)
|
9.50 (5.1-14.4)
|
0.60 (0.4-1.0)
|
<0.001*
|
Hb (g/dl)
|
7.90 (7.2-8.6)
|
11.0 (10.3-11.7)
|
<0.001*
|
Keys
IQR- Interquartile range
SCA- Sickle cell anaemia
|
Hb F- Foetal haemoglobin
Hb- Haemoglobin
|
g/dl- grams per decilitre
*- Statistically significant
|
LV Dimensions and Functional Parameters
The left ventricular internal dimensions in both phases of the cardiac cycle were significantly greater in the systole-2.75 cm (IQR 2.4-3.0) group than in the control-2.22 cm (IQR 2.0-2.4) group (p< 0.001) and in the diastole-4.03±0.5 cm group than in the control group (3.64±0.5 cm; p<0.001). While the interventricular septum in systole was significantly thinner in SCA subjects, the LV posterior wall was significantly thicker in diastole but not significantly thicker in systole (Table 3). ).
LV systolic dysfunction
The median ejection fraction and fractional shortening were also significantly lower in the subjects than in the controls, with values of 62.00% (IQR 55.0-67.0) and 33.00% (IQR 27.3-36.8), respectively, in the subjects and 68.00% (IQR 65.8-73.0) and 37.00% (IQR 35.0-41.0), respectively, in the controls (p<0.001) (Table 3). Left ventricular systolic dysfunction was therefore present in 27% of the SCA patients but only in 1% of the controls (χ² = 26.5, p< 0.001).
Table 3. LV Dimensions and Functional Parameters in Subjects and Controls
Parameter
|
SCA subjects
|
Hb AA controls
|
p- value
|
LVMI (g/m2)
Median (IQR)
|
71.69 (56.7-91.7)
|
41.9(35.8-54.1)
|
<0.001*
|
EF (%)
Median (IQR)
|
62.00 (55.0-67.0)
|
68.00 (65.8-73.0)
|
<0.001*
|
FS (%)
Median (IQR)
|
33.00 (27.3-36.8)
|
37.00 (35.0-41.0)
|
<0.001*
|
LVIDd (cm)
Mean (±SD)
|
4.03 (±0.5)
|
3.64 (±0.5)
|
<0.001*
|
LVIDs (cm)
Median (IQR)
|
2.75 (2.4-3.0)
|
2.22 (2.0-2.4)
|
<0.001*
|
LVPWTs (cm)
Median (IQR)
|
0.89 (0.7-1.0)
|
0.83 (0.7-0.9)
|
0.41
|
LVPWTd (cm)
Median (IQR)
|
0.61 (0.5-0.7)
|
0.55 (0.5-0.7)
|
0.02*
|
IVSTs (cm)
Median (IQR)
|
0.50 (0.4-0.6)
|
0.56 (0.5-0.7)
|
<0.001*
|
IVSTd (cm)
Median (IQR)
|
0.44 (0.4-0.6)
|
0.39 (0.3-0.5)
|
<0.001*
|
RWT (cm)
Median (IQR)
|
0.27 (0.2-0.3)
|
0.27 (0.2-0.3)
|
0.95
|
Keys
SCA- Sickle cell anaemia
Hb- Haemoglobin
LVMI- Left ventricular mass index
EF- Ejection fraction
FS- Fractional shortening
*Statistically significant
|
LVIDd- Left ventricular internal dimension in diastole
LVIDs- Left ventricular internal dimension in systole
LVPWTs- Left ventricular posterior wall thickness in systole
|
LVPWTd- Left ventricular posterior wall thickness in diastole
IVSTs- Interventricular septum thickness in systole
IVSTd- Interventricular septum thickness in diastole
RWT- Relative wall thickness
|
Patient characteristics in relation to LV dysfunction
Spearman’s rank correlation showed a significant negative correlation between EF and age (ρ=-0.25, p= 0.006), BSA (ρ=-0.24, p= 0.008), and systolic (ρ=-0.23, p= 0.022) and diastolic (ρ=-0.31, p= 0.002) BP, i.e., EF decreased significantly with increasing values of these variables (Table 4).
Although SCA patients with LVSD were older (8.5 years, IQR 7.0-12.0) than those without this abnormality (7.0 years, IQR 5.0-10.0), this difference was not statistically significant (p= 0.34). Subjects with LVSD had a significantly greater diastolic BP (70 mmHg; IQR 60-70) than did those without this abnormality (60 mmHg; IQR 50-70) (p=0.04).
The odds of having LVSD decreased by 0.9 for every 1% increase in HbF level (CI= 0.82-0.99, p= 0.03) (Table 5).
Table 4. Correlation of the EF with Clinico-laboratory Parameters
Parameter
|
Spearman’s rho
|
p- value
|
Age
|
-0.25
|
0.006*
|
Body surface area
|
-0.24
|
0.008*
|
Haemoglobin
|
-0.12
|
0.184
|
% Foetal haemoglobin
|
0.09
|
0.336
|
Oxygen saturation
|
-0.02
|
0.871
|
Systolic blood pressure
|
-0.23
|
0.022*
|
Diastolic blood pressure
|
-0.31
|
0.002*
|
*Statistically significant
Table 5. Adjusted odds ratios with 95% confidence intervals for multiple logistic regression analysis to determine predictors of LV systolic dysfunction
Parameters
|
Adjusted odds ratio (95% Confidence interval)
|
p- value
|
HbF
|
0.9 (0.82-0.99)
|
0.03*
|
Hb
|
1.7 (0.94-3.17)
|
0.80
|
Systolic BP
|
0.9 (0.92-1.06)
|
0.69
|
Diastolic BP
|
1.3 (0.96-1.12)
|
0.37
|
Abnormal LV geometry
|
1.1 (0.53-4.13)
|
0.45
|
*Statistically significant
LV geometric patterns
Of the 120 SCA patients studied, 48% had normal LV geometry, and 52% had abnormal LV geometry. Eccentric LVH was the predominant abnormal pattern and was present in 50%, concentric LVH was present in 2%, and none had concentric remodelling (Fig. 1. ).
The presence of abnormal LV geometry was associated with LVSD (Fisher’s exact test, p< 0.001). The odds of having abnormal LV geometry decreased by 0.5 for every 1 g increase in Hb level (CI=0.27-0.87, p=0.02), as shown in Table 6.
Table 6. Adjusted odds ratios with 95% confidence intervals for multiple logistic regression analysis to determine predictors of abnormal LV geometry
Parameters
|
Adjusted odds ratio (95% Confidence interval)
|
p- value
|
Hb
|
0.5 (0.27-0.87)
|
0.02*
|
HbF
|
0.9 (0.89-1.05)
|
0.39
|
Systolic BP
|
1.0 (0.97-1.09)
|
0.30
|
Diastolic BP
|
0.98 (0.92-1.06)
|
0.64
|
*Statistically significant