Pediatr Cardiol (2015) 36:445–453 DOI 10.1007/s00246-014-1033-0
ORIGINAL ARTICLE
Speckle Tracking and Myocardial Tissue Imaging in Infant of Diabetic Mother with Gestational and Pregestational Diabetes Mohammed Al-Biltagi • Osama Abd Rab Elrasoul Tolba Mohamed Ahmed Rowisha • Amal El-Sayed Mahfouz • Mona Ahmed Elewa
•
Received: 23 January 2014 / Accepted: 26 September 2014 / Published online: 7 October 2014 Ó Springer Science+Business Media New York 2014
Abstract The aim of this study was to evaluate the myocardial changes in infants of diabetic mother either with gestational or pregestational diabetes and its relation to maternal diabetic control. The study included 45 infants of diabetic mother (IDMs) and 45 healthy newborn as a control group. IDMs were then categorized into 2 subgroups: twenty infants of mother with pregestational diabetes and twenty-five infants of mothers with gestational diabetes. The studied groups underwent measurement of the maternal and neonatal glycated Hb % (HbA1c), conventional echocardiography, tissue Doppler imaging (TDI) and two-dimensional speckle tracking imaging (STI). The weight, the rate of complications, and the rate of cesarean section were significantly higher in the IDMs group than in the control group. Significant positive correlation was present between the levels of HbA1c of IDMs and HbA1c of their mothers (P \ 0.05). A significant deterioration of both systolic and diastolic functions measured by both conventional echocardiography and TDI was present in IDMs with both pre-gestational and gestational diabetes compared with the control group. Also, the septal/posterior wall ratio (SW/ PW) was significantly higher in pregestational (1.86 ± 0.3) and gestational (2 ± 0.4) groups than in the control group (1 ± 0.06). Two-dimensional STI showed that the cardiac torsion was significantly impaired in pre-gestational (9.66 ± 2.5) and gestational (8.66 ± 3.9) groups when M. Al-Biltagi (&) O. A. R. E. Tolba M. A. Rowisha M. A. Elewa Pediatric Department, Faculty of Medicine, Tanta University, Medical Complex, Tanta, Egypt e-mail: [email protected] A. E.-S. Mahfouz Gynecology and Obstetrics Department, Faculty of Medicine, Tanta University, Tanta, Egypt
compared with the control group (5.4 ± 2.4) [P \ 0.0001]. It also showed that the global strain was significantly impaired in pre-gestational (-10.4 ± 3.2) and gestational (-13.1 ± 4.7) groups when compared with the control group (-19 ± 2) [P \ 0.0001]. However, no significant differences were present among the two patients’ subgroups in echocardiographic data except for a significant decrease of E’/A’ ratio and S wave at tricuspid annulus derived by TDI and impaired global strain derived by STI in infants of mothers with pre-gestational DM than those with gestational DM [P = 0.02]. SW/PW and cardiac torsion were significantly higher in infant of diabetic mother than the normal newborn and on the contrary systolic function and global strain were significantly lower in IDMs especially in infants of mother with pre-gestational diabetes. All the previous TDI findings did not show any significant correlation to neither maternal nor fetal HbA1c. Also, there was no significant correlation between cardiac torsion and the rest of TDI data neither in IDMs group nor in the control group. TDI and two-dimensional STI were efficient and sensitive tools able to early detect cardiac dysfunction in IDMs even in the absence of morphologic cardiac changes. Keywords Gestational diabetes Pre-gestational diabetes Infants of diabetic mothers Speckle tracking imaging Tissue Doppler imaging
Introduction Diabetes mellitus (DM) is a chronic metabolic disorder that affects in some ways all the stages of childhood; from in utero to neonates, infants, children, adolescents, and adults. Diabetes in pregnant women could be either pre-gestational
123
446
or gestational with an incidence of impaired glucose tolerance in pregnancy ranging between 3 and 10 %; which varies according to the average incidence of diabetes in the general population [20]. Its cardiovascular complications impose a significant morbidity risk in such active age groups due to metabolic derangement, the associated dyslipidemia, atherosclerosis, hypertension, and autonomic dysfunction [1]. The fetal heart is a target organ for the congenital effects of pre-existing as well as gestational maternal diabetes. It significantly affects the fetal heart and fetal–placental circulation in both functions and structures; causing cardiac anomalies and myocardial hypertrophy [6]. The reported incidence of these complications was 3.4 % with HbA1c levels lower than 8.5 and 22.4 % with HbA1c levels higher than 8.5 %. Infants born to mothers with an HbA1c level more than 10 % in late pregnancy are more prone to present with neonatal complications [18]. M-mode and 2-D echocardiography can show cardiomegaly (30 %), asymmetric septal hypertrophy, and fetal ventricular walls thickness that simulates idiopathic hypertrophic subaortic stenosis which increases progressively with advancing gestation [24]. Tissue Doppler imaging (TDI) is a relatively new echocardiographic technique useful to study the myocardial functions of fetal and neonatal hearts of abnormal gestation including pregnancies affected by diabetes mellitus [8]. Two-dimensional speckle tracking imaging (STI) is another relatively new echocardiographic modality that can provide non-Doppler, angle-independent, and objective quantification of myocardial deformation and left ventricular (LV) systolic and diastolic dynamics by analyzing the motion of speckles identified on routine 2-dimensional sonograms. By tracking the displacement of the speckles during the cardiac cycle; strain and the strain rate can be rapidly measured offline after sufficient image acquisition [14]. It can be used to measure interventricular septal thickness, posterior LV wall thickness, LV ejection fraction (EF), and LV rotation and torsion in IDMs [12]. The aim of this study was to evaluate the myocardial changes by conventional echocardiography, TDI and two-dimensional STI in infants of diabetic mothers either in pre-gestational or gestational diabetes and their relation to maternal glycemic control.
Patients and Methods The study was done as a case–control study included a group of 45 infants of diabetic mother (IDMs) and 45 healthy neonates with clinically and anatomically normal heart as a control group assigned from Neonatology and Cardiology Units; Pediatric Department, from a tertiary care hospital during the period from March, 2006 to April, 2009. The IDMs group was further divided into 2
123
Pediatr Cardiol (2015) 36:445–453
subgroups according to the type of the maternal diabetes (pre-gestational or gestational). The pre-gestational diabetes is the diabetes which started before the current pregnancy. Each subgroup was further subdivided according to the adequacy of control of maternal diabetes into adequately controlled and not adequately controlled. Maternal gestational diabetes is defined as controlled when fasting capillary blood glucose level of less than 95–105 mg per dL and the postprandial capillary blood glucose level of less than 140 mg per dL at one hour and less than 120 mg per dL at two hours [29]. Infants with intrauterine growth retardation (IUGR), multiple congenital anomalies, hypoxic ischemic encephalopathy (HIE), or chromosomal abnormalities, twins and preterm babies were excluded. Infants of mothers with history of hypertension, preeclampsia, chronic systemic disease, or on chronic medications except for insulin were also excluded from the study. Full history taking including gestational age, maternal name and age, type of maternal diabetes, type of delivery, and maternal blood glucose level was taken. A thorough clinical examination was done by pediatric neonatologist and pediatric cardiologist with special attention to cardiovascular system. Serial blood glucose levels were done in the first day of life and until the infants blood sugar remained stable with normal feedings. Measurements of HbA1c were done from the cord blood. Complete blood count, blood gases, C reactive protein (CR), blood glucose level, level of ionized calcium, renal and liver function tests, and chest X-ray were done when clinically indicated. Conventional echocardiography, two-dimensional STI and TDI were done for all cases within the first 2 postnatal days. To avoid intra-observer variability, two examinations were done by the same operator for each patient within the first 2 days of life, and we considered the average results. All the diabetic mothers with singleton pregnancy, and whom diagnosed as having diabetes during pregnancy following WHO criteria had HbA1c level from 36 week of gestation and onward before delivery. Conventional Echocardiography Echocardiographic examination was done using Vingmed Vivid-7 (General Electric Vingmed, and Milwaukee, Wisconsin, USA). Data acquisition was done with a 5-MHz transducer at a depth of 16 cm in all the standard echocardiographic views according to the recommendation of the American Society of Echocardiography [21]. All neonates were examined while lying quietly in the right anterior oblique position breathing room air. Cardiovascular malformations including patent ductus arteriosus were carefully searched and excluded by all standard views. Left ventricle ejection fraction (LV EF), LV
Pediatr Cardiol (2015) 36:445–453
fractional shortening (LV FS), LV end-diastolic and endsystolic volumes, systolic pulmonary artery pressure (PAP), and LV and right ventricle (RV) diastolic functions were measured according to the guidelines of the American Society of Echocardiography [21]. Tei index, which is a Doppler-derived time interval index that combines both systolic and diastolic cardiac performance, was also calculated as previously described by Tei and colleagues [10]. Hypertrophic cardiomyopathy (HCM) was diagnosed when the absolute values of inter ventricular septum and LV posterior wall diameters were above normal ranges, and in addition, the ratio of inter ventricular septum in diastole to the LV posterior wall in diastole exceeded 1.3. Two-Dimensional Speckle Tracking Imaging (STI) Longitudinal strain in 18 left LV segments using twodimensional STI were obtained using the 3-standard apical views; apical long axis, apical 4-chamber, and apical 2-chamber views. The parameters obtained represented the average of 3 cardiac cycles, with a frame rate of 65 frames/ s, and all the segmental data were represented. We used automated function imaging that enables the assessment of longitudinal strain available in Vivid 7 ultrasound machine to measure average LV global peak systolic strain. Global peak systolic strain was recorded in 3-standard apical views, and segmental peak systolic strain in basal, mid, and apical segments of anteroseptal, anterior, lateral, posterior, inferior, and septal LV walls. Left ventricular torsion was defined as the net difference of LV rotation at the basal and apical planes. For the STI analysis, we used high frame (65 frames/s) with harmonic two-dimensional images [19]. Tissue Doppler Imaging (TDI) Tissue Doppler imaging was done using the same machine and probe at a depth of 16 cm in the parasternal and apical views (standard long axis and two- and four-chamber images). Using pulsed-wave angle-corrected color-coded TDI filters, the baseline was adjusted to a low velocity range (-20–20 cm/s) and Doppler frame rates were varied between 80 and 115 frames/s depending on the sector width of the range of interest with minimal gain setting to reduce background noise and to get the highest quality images. Two millimeters sample volume was placed within the myocardium equidistant from the endocardial and epicardial borders. From the apical four-chamber planes, using pulsed-wave TDI, the myocardial velocity curves of septal mitral valve annulus, lateral mitral valve annulus, and lateral tricuspid valve annulus were recorded. The electrocardiogram and respiration curve monitoring were connected and traced simultaneously to define the timing of cardiac cycle events and its relation to respiratory events.
447
The beginning of QRS complex was used as a reference point. At least 10 cardiac cycles were recorded at a speed of 100 mm/s, and the images were stored electronically. TDI measurements were indexed for children’s body surface area [2]. The mean values for three heart beats during expiration were used for the analysis. The systolic wave (S) reflects the systolic function of either RV or LV. The early/atrial (E’/A’) ratio of tricuspid and mitral valve annulus reflects the diastolic function of the RV and LV, respectively. Isometric contraction time [ICT] (was defined as the time duration between the beginnings of QRS complex in the electrocardiogram to the beginning of TDI S wave), the isometric relaxation time [IRT] (was defined as the interval between the end of S wave and the beginning of the early wave), and myocardial contraction time (CT) were measured by TDI. Tei index (myocardial performance index, MPI) was calculated using the following formula: ICT ? IRT/CT. Both ICT and IRT were corrected for heart rate [16]. All the data were adjusted for gestational age and estimated neonatal weight. The mothers of the studied newborns were informed of the purpose of the study before participating in the study, and all the parents of the included newborns signed a written-informed consent. The study was approved by the Institutional Ethical and Research Review Board of Faculty of Medicine, Tanta University.
Statistics The statistical power of the study was more than 90 % (using a computerized program: Power and Precision V3; www.PowerAnalysis.com). Data were presented as mean (±standard deviation) values. The two-way analysis of variance with repeated measures and chi-square test by SPSS V.16 were used to find statistically significant differences in the different parameters among the groups. For all analyses, a statistical significance of P value less than 0.05 was used. Wilcoxon’s signed-rank test was used to assess the normality of distributions of the data. The Bonferroni correction/adjustment procedure was done to avoid ‘‘significance’’ attributable to chance only in multiple comparisons with echocardiographic parameters. Correlation between maternal HbA1c, neonatal HbA1c, and echocardiographic variables was evaluated using Pearson’s correlation coefficient.
Results Figure 1 showed the flow chart of the study that included 45 IDM neonates and 45 healthy neonates as a control group. Among the IDMs, there were 25 IDMs with gestational DM
123
448
Pediatr Cardiol (2015) 36:445–453
Fig. 1 Flow chart of the studied groups
Table 1 Clinical and laboratory data of IDMs and control group Variables (Mean ± SD)
IDMs n = 45
Controls n = 45
t
P
Gestational age (weeks) (Mean ± SD)
38.8 ± 1.2
39 ± 2
0.57
0.56
Males, n (%)
21 (46.6%)
22 (48.8%)
0.21
0.8
Females, n (%)
24 (53.4%)
23 (51.2%)
0.2
0.83
3.54 ± 0.54
3.23 ± 0.34
3.25
0.0016
Respiratory distress Neonatal jaundice (including physiological jaundice)
9 (20%) 30 (66.6%)
0 15 (33%)
3.2 3.1
0.002 0.002
Hypocalcaemia
11 (24%)
0
3.5
0.0007
Hypoglycemia
6 (13%)
0
2.5
0.01
Hypothyroidism
5 (11%)
1 (2.2%)
1.6
0.09
Sex
Birth Weight (kg) (Mean ± SD) Postnatal problems, n (%):
Mode of delivery: Cesarean section, n (%)
36 (80%)
16 (35.5%)
4.2
0.0001
Vaginal delivery, n (%)
9 (20%)
29 (64.6%)
4.2
0.0001
28.6 ± 4.8
27 ± 5.6
1.45
0.15
Maternal age (years) Mean ± SD P value \ 0.05 is significant
and 20 IDMs with pre-gestational DM. Tables 1 and 2 showed the demographics of the included groups. Table 1 showed significant differences between the control group and the IDMs in gestational age, maternal age and sex while the weight, the rate of complications, and rate of cesarean section were significantly higher in the IDMs than in the control group. Table 2 showed no significant differences between the two subgroups of IDMs (pre-gestational and gestational diabetic mothers) in their demographic data.
123
Table 3 showed echocardiographic data in IDMs of mothers with pre-gestational and gestational diabetes and control groups. The intra-observer agreement K values of the echocardiographic measurements were between 77 and 88 which indicated good agreement between the 2 readings. The table showed the presence of significant deterioration of both LV and RV systolic and diastolic functions measured by both conventional echocardiography and TDI among IDMs of both pre-gestational and gestational
Pediatr Cardiol (2015) 36:445–453
449
Table 2 Clinical and laboratory data of babies and mothers of pre-gestational and gestational diabetes group Variables
Pre-gestational DM group (Group 1) n = 20
Gestational DM group (Group 2) n = 25
P
Gestational age (weeks) Mean ± SD
39 ± 1.2
38.7 ± 1.48
0.45
Males, n (%)
11 (55%)
10 (40%)
0.3
Females, n (%)
9 (45.4%)
15 (60%)
0.33
3.65 ± 0.64
3.44 ± 0.44
0.22
Respiratory distress Neonatal jaundice (including physiological jaundice)
5 (25%) 14 (70%)
4 (16%) 16 (64%)
0.4 0.6
Hypocalcemia
6 (30%)
5 (20%)
0.44
Hypoglycemia
3 (15%)
3 (12%)
0.7
Hypothyroidism
3 (15%)
2 (8%)
0.46
Cesarean section, n (%)
15 (75%)
21 (84%)
0.45
Vaginal delivery, n (%)
5 (25%)
4 (16%)
0.45
28.9 ± 5.07
28.4 ± 4.7
0.7
Sex
Birth weight Kg (Mean ± SD) Postnatal problems, n (%):
Mode of delivery
Maternal age (years) Mean ± SD Maternal glycemic control: Controlled, n (%)
11 (55%)
15 (60%)
0.73
Uncontrolled, n (%)
9 (45%)
10 (40 %)
0.73
P value \ 0.05 is significant
diabetes compared with the control group. Also, the septal/ posterior wall ratio was significantly higher in patients groups than in the control group (P \ 0.0001 for both subgroups). However, no significant differences were present among the two patients subgroups in conventional echocardiographic data, but TDI showed significant decrease of E’/A’ ratio (P \ 0.0001) and S wave (P \ 0.0001) at tricuspid annulus in infants of mothers with pre-gestational DM than in infants of mothers with gestational DM. Two-dimensional STI showed significantly impaired cardiac torsion in two patients groups when compared with the control group (P \ 0.0001 for both subgroups). However, cardiac torsion was more impaired in IDMs of mother with pre-gestational diabetes than those of mothers with gestational diabetes but without statistical significance (P [ 0.05). STI showed also significant impairment of the global strain among IDMs when compared with the control group (P \ 0.0001 for both subgroups). However, the global strain was significantly impaired in infants of mothers with pre-gestational DM than in infants of mothers with gestational DM (P \ 0.05). Tables 4 and 5 showed the absence of significant correlation between any of the echocardiographic parameters (including conventional echocardiography, TDI and twodimensional STI) and neither maternal nor neonatal HbA1c. However, significant positive correlation was present between the levels of HbA1c of IDMs and HbA1c of their
mothers in both pre-gestational (P \ 0.05) and gestational diabetes subgroups (P \ 0.05). No significant differences were present between HbA1c of the mothers with pre-gestational and gestational diabetes and any of the cardiac TDI data including Septal/posterior wall (sw/pw), E’/A’ ratio or S wave of mitral annulus, LV global strain and LV torsion. On the other hand, the infants’ birth weight had significant positive correlation with neonatal (P \ 0.05) and maternal (P \ 0.05) HbA1c value and septal/posterior wall ratio (P \ 0.05) but negatively correlated with the cardiac torsion (P \ 0.05) as shown in Table 6.
Discussion Cardiac functional abnormalities are present in about 30 % of IDMs and include intraventricular septal hypertrophy and cardiomyopathy and 10 % may have frank heart failure. Hypertrophic cardiomyopathy observed in the infant of the diabetic mother is characterized by thickening of intraventricular septum, and to a lesser extent the ventricular free walls [20]. In our study, a significant increase in LV septal/posterior wall ratio was present in the IDMs groups than the control indicating septal thickening. Various echocardiographic modalities showed that the IDMs (with either gestational or pre-gestational DM) had impairment of both systolic and diastolic functions as observed by both
123
450
Pediatr Cardiol (2015) 36:445–453
Table 3 Comparison of Echocardiographic Data in IDMs (pre-gestational and gestational diabetes) and control groups Pre-gestational DM (Group 1) (n = 20) LV FS
E/A ratio MV
Mean ± SD
Mean ± SD
38.04 ± 4.12
0.89 ± 0.07
Gestational DM (Group 2) (n = 25) 38.3 ± 4.4
0.92 ± 0.08
Control Group (n = 45)
t
41.56 ± 5.14
t1
-2.94
0.0052
t2
-2.79
0.007
t3
0.2
1.26 ± 0.13
P value
0.8393
t1
-14.85
\0.0001
t2
-13.5
\0.0001
t3
1.3
0.18
12.8 11.2
\0.0001 \0.0001
Septal/posterior wall (sw/pw)
Mean ± SD
1.86 ± 0.28
2 ± 0.42
1.05 ± 0.06
t1 t2 t3
1.3
E’/A’ wave mitral annulus
Mean ± SD
0.70 ± 0.13
0.76 ± 0.14
1.27 ± 0.14
t1
-15.9
\0.0001
t2
-14.6
\0.0001
S wave mitral annulus (Cm/s))
LV TEI
Pul systolic Pressure (mmHg)
Mean ± SD
Mean ± SD
Mean ± SD
3.2 ± 0.6
0.45 ± 0.061
38.5 ± 5
3.4 ± 0.8
0.43 ± 0.05
36.5 ± 6
6.1 ± 0.5
0.38 ± 0.06
30.5 ± 4
0.18
t3
1.5
0.14
t1
18.9
\0.0001
t2
15.3
\0.0001
t3
-0.96
0.34
t1
4.3
\0.0001
t2
3.7
0.0004
t3
1.2
0.24
t1
6.31
\0.0001
t2
4.47
\0.0001
t3
1.2 36 26.1
0.2 \0.0001 \0.0001
E’/A’ wave tricuspid annulus
Mean ± SD
0.84 ± 0.03
0.87 ± 0.04
1.11 ± 0.03
t1 t2 t3
4.7
\0.0001
S wave tricuspid annulus cm/s)
Mean ± SD
5.3 ± 0.5
5.9 ± 0.4
7.4 ± 0.3
t1
17.4
\0.0001
t2
16.3
\0.0001
t3
4.3
0.0001
t1
6.7
\0.0001
t2
4
t3
1.9
0.0621
t1
7.5
\0.0001
t2
8.7
\0.0001
t3
2.3
0.02
t1
6.5
\0.0001
t2
3.75
0.0006
t3
1.04
0.3
RV TEI
Global strain (GLS)
Torsion (T)
Mean ± SD
Mean ± SD
Mean ± SD
0.42 ± 0.03
-10.4 ± 3.22
9.66 ± 2.46
0.40 ± 0.04
-13.1 ± 4.70
8.66 ± 3.95
0.36 ± 0.04
-19 ± 2.05
5.40 ± 2.41
0.0002
t1 is the comparison between control group and group 1; t2 is the comparison between control group and group 2; t3 is the comparison between Group 1 and group 2 P value \0.05 is significant
conventional echocardiography (lower LV FS, lower E/A ratio at mitral valve, and prolonged Tie index) and TDI (lower S wave and E’/A’ wave at both mitral and tricuspid valve annuli). In our study, LV and RV Tie indices were prolonged in IDMs of mother with either gestational or pregestational DM than in healthy control infants. The Tie index is a simple, noninvasive, simple to estimate, and
123
reproducible index which has close correlation with the widely accepted systolic and diastolic hemodynamic parameters as well as the potential for clinical application in the assessment of overall cardiac performance including isovolumic contraction time (ICT), ejection time (ET), and isovolumic relaxation time [15]. In our study, the pulmonary systolic pressure was significantly higher in IDMs than
Pediatr Cardiol (2015) 36:445–453
451
Table 4 Correlation between glycated Hb of the babies in pre-gestational and gestational diabetic groups and their cardiac tissue Doppler data
Septal/posterior wall (sw/pw)
Glycated Hb of the babies
Babies’ birth weight
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
R
R
R
R
P value 0.411
0.080
0.756
E’/A’ wave mitral annulus
-0.145
0.606
-0.128
0.650
S wave mitral annulus (Cm/s))
-0.555
0.032
-0.376
0.219
0.432
Torsion (T) Glycated Hb of the mothers
-0.440 0.362
P value
P value
P value
0.299
Global strain (GLS)
Table 6 Correlation between babies’ birth weight in pre-gestational and gestational diabetes groups and their glycated Hb, their cardiac Torsion, their SW/PW, and their maternal glycated Hb
Babies’ glycated Hb
0.425
0.011*
0.396
0.047*
Maternal glycated Hb
0.396
0.020*
0.428
0.023*
Cardiac Torsion (T)
-0.357
0.044*
-0.362
0.043*
0.178
Septal/posterior wall (sw/pw)
-0.425
0.019*
0.388
0.028*
0.128
0.649
P value \0.05 is significant
0.001*
0.265
0.341
0.044
0.396
0.021
P value \0.05 is significant
in infants of non-diabetic mothers. This finding agreed with the work of Dimitriu et al. who observed the presence of pulmonary hypertension in 3 cases out of 35 IDMs [4]. Koza´k-Ba´ra´ny et al. showed that infants of mothers with well-controlled pre-gestational or gestational diabetes had impaired LV relaxation. They related this finding to the effects of maternal hyperglycemia during the third trimester and subsequent fetal hyperinsulinemia leading to neonatal cardiac hypertrophy which also agreed with our study findings [11]. However, despite Ren et al. and Hate´m et al. showed septal thickening among IDMs and that the pulsed TDI demonstrated evidence of impaired diastolic function, in fetuses of diabetic mothers compared with fetuses of non-diabetic mothers; yet; they found that this dysfunction was independent of the presence of myocardial hypertrophy [7, 23]. In our study, no significant differences were present in the conventional echocardiographic parameters and TDI between IDMs of mothers with gestational diabetes and those of mothers with pre-gestational diabetes except for presence of lower E’/A’ ratio and S wave at tricuspid annulus by TDI in IDMs of mothers with
pre-gestational diabetes than those of mothers with gestational diabetes. However, TDI has important limitations. TDI measures only the vector of motion that is parallel to the direction of the ultrasound beam and it measures absolute tissue velocity. It is also unable to discriminate passive motion (related to translation or tethering) from active motion (fiber shortening or lengthening) [8]. To overcome these limitations of TDI, we did strain rate imaging, using STI which has the ability to quantify regional wall deformation and ability to differentiate between active and passive motion by quantification of myocardial deformation. It allows an objective and quantitative evaluation of global and regional myocardial function independently from the angle of insonation and from cardiac translational movements [22]. STI-derived measurements have high feasibility and reproducibility and are comparable to that taken by sonomicrometry and tagged MRI [3]. We found that both the global strain (GLS) and the cardiac torsion were significantly impaired in the IDMs than in the control group. Also, we found that GLS was significantly impaired in IDMs of mother with pregestational diabetes than those of mothers with gestational diabetes. However, there was no significant difference in cardiac torsion between the two patients’ subgroups. The result of our study agreed with the previous findings of
Table 5 Correlation between torsion in pre-gestational and gestational diabetes groups and other tissue Doppler data Torsion (T)
Septal/posterior wall (sw/pw)
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
Control group (n = 45)
R
R
P value
R
P value
P value
-0.068
0.809
-0.287
0.300
-0.325
0.225
E’/A’ wave mitral annulus
0.088
0.745
-0.139
0.629
0.152
0.341
S wave mitral annulus (Cm/s) Global strain (GLS)
0.257 0.129
0.356 0.649
-0.545 0.119
0.036* 0.673
0.321 0.155
0.280 0.625
P value \0.05 is significant
123
452
Liao et al. who showed that the cardiac functions may be impaired in newborn infants of mothers with gestational diabetes, with changes in LV shape and abnormalities of LV rotation and torsion even in the presence of normal ventricular ejection fraction. They recommended using two-dimensional STI for early detection of LV dysfunction [12]. Global LV longitudinal systolic strain (GLS) is a sensitive and simple measure of LV mechanics that can be obtained by STI. It takes a few to 10 minutes to be done and has more ability to predict adverse cardiovascular events than LVEF and TDI-derived S wave. It could offer an additional prognostic benefit over clinical, conventional echocardiographic, and TDI systolic parameters [27]. However, it depends on the level of the operator’s experience, needs high-quality images for assessment of global LV strain and is frame rate-dependent. The frame rate of the obtained images in our study was 65 frames/s. A major limitation of STI that can affect the accuracy of imaging process and data obtained is its strict dependency on the frame rate and the need for high-quality two-dimensional images, which are necessary for obtaining an optimal definition of the endocardial border [13]. The reason of impaired GLS in IDMs of mother with pre-gestational diabetes than those of mothers with gestational diabetes may be related to its early effect on the fetus which can start as early as 11–14 weeks of pregnancy. Turan et al. showed that fetuses of poorly controlled diabetic mothers demonstrated significant differences in firsttrimester diastolic myocardial function compared with nondiabetic controls. The decrease in myocardial performance was more marked with increasing HbA1c and appears to be independent of preload and after load [28]. These early cardiac effects of pre-gestational DM that started during the first trimester could affect the early embryonic development and alter the cardiac morphogenesis and placental development. These effects continue through the second and third trimesters to affect the fetal circulation and also can extend to the perinatal and neonatal period exposing the IDMs to the same effects of gestational diabetes including the metabolic derangement [9]. In our study, no significant correlation was present between SW/PW ratio, e’/a’ ratio, S wave, (cardiac torsion, and GLS in IDMs in the two patients subgroups with neither maternal nor neonatal HbA1. This agreed with the findings of Sheehan et al. who confirmed the absence of relationship between echocardiographic evidence of hypertrophic cardiomyopathy and maternal HbA1 levels. Their data did not support the third trimester maternal hyperglycemia as instrumental etiology of cardiomyopathy and other complications observed in IDMs [26]. However, El-Ganzoury et al. found a positive correlation between maternal HbA1c and IVS thickness and impaired LV
123
Pediatr Cardiol (2015) 36:445–453
contractility [5]. Meanwhile, our study showed significant positive correlation between neonatal birth weight and neonatal and maternal HbA1c value and SW/PW ratio but negatively correlated with the cardiac torsion. These finding agreed with the result of El-Ganzoury et al. who found that the birth weight is the best predictor of hypertrophied IVS especially in infants born to sub optimally controlled diabetic mothers [5]. On the other hand, Roodpeyma et al. showed neither the types of maternal diabetes nor the somatic findings of newborns that were related to the occurrence of cardiac complications [25]. The relation between the increase in birth weight and cardiac dysfunction could be attributable to the same mechanism. For example, the insulin resistance-associated hyperinsulinemia that is frequently associated with pregnancy-associated diabetes can induce smooth-muscle cell hypertrophy and hyperplasia and increased extracellular proteins which are an important factor contributing to increase birth weight and at the same time affects the cardiac function [17]. Our study has certain limitations such as few numbers of the studied subgroups and lack of follow-up period. We did not do cardiac magnetic resonance (CMR) to compare STI measurements with it. Further studies are suggested on larger number of IDMs and for longer duration.
Conclusion SW/PW and cardiac torsion were significantly higher in IDMs than the healthy newborn, and on the contrary, systolic and diastolic functions and global strain were significantly impaired in IDMs especially in babies in pregestational diabetes group. All the previous TDI findings show insignificant correlation to neither maternal nor fetal HbA1c. Also, no significant correlation was present between cardiac torsion and the rest of TDI data in infants of both patients’ subgroups. TDI and two-dimensional STI are efficient and sensitive tools able to early detect cardiac dysfunctions in IDMs even in the absence of morphologic cardiac changes.
References 1. Al-Biltagi, M (2014) Cardiovascular effects of diabetes mellitus in pediatric population. In: Research on Diabetes II. ISBN: 978-1922227-26-3 . iConcept Press, Hong Kong 2. Al-Biltagi M, Serag AR, Hefidah MM, Mabrouk MM (2013) Evaluation of cardiac functions with Doppler echocardiography in children with Down syndrome and anatomically normal heart. Cardiol Young 23(2):174–180. doi:10.1017/S1047951112000613 3. Amundsen BH, Helle-Valle T, Edvardsen T et al (2006) Noninvasive myocardial strain measurement by speckle tracking
Pediatr Cardiol (2015) 36:445–453
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
15.
echocardiography: validation against sonomicrometry and tagged magnetic resonance imaging. J Am Coll Cardiol 47(4):789–793. doi:10.1016/j.jacc.2005.10.040 Dimitriu AG, Russu G, Stamatin M, Jita˘reanu C, Streanga˘ V (2004) Clinical and developmental aspects of cardiac involvement in infant of diabetic mother. Rev Med Chir Soc Med Nat Iasi 108(3):566–569 El-Ganzoury MM, El-Masry SA, El-Farrash RA, Anwar M, Abd Ellatife RZ (2012) Infants of diabetic mothers: echocardiographic measurements and cord blood IGF-I and IGFBP-1. Pediatr Diabetes 13(2):189–196. doi:10.1111/j.1399-5448.2011.00811.x Gardiner HM, Pasquini L, Wolfenden J, Kulinskaya E, Li W, Henein M (2006) Increased periconceptual maternal glycated haemoglobin in diabetic mothers reduces fetal long axis cardiac function. Heart 92:1125–1130. doi:10.1136/hrt.2005.076885 Hate´m MA, Zielinsky P, Hate´m DM et al (2008) Assessment of diastolic ventricular function in fetuses of diabetic mothers using tissue Doppler. Cardiol Young 18(3):297–302. doi:10.1017/ S1047951108002138 Ho CY, Solomon SD (2006) A clinician’s guide to tissue Doppler imaging. Circulation 113(10):e396–e398. doi:10.1161/CIRCU LATIONAHA.105.579268 Hornberger LK (2006) Maternal diabetes and the fetal heart. Heart 92(8):1019–1021. doi:10.1136/hrt.2005.083840 Karatzis EN, Giannakopoulou AT, Papadakis JE, Karazachos AV, Nearchou NS (2009) Myocardial performance index (Tei index): evaluating its application to myocardial infarction. Hellenic J Cardiol 50(1):60–65 Koza´k-Ba´ra´ny A, Jokinen E, Kero P, Tuominen J, Ro¨nnemaa T, Va¨lima¨ki I (2004) Impaired left ventricular diastolic function in newborn infants of mothers with pregestational or gestational diabetes with good glycemic control. Early Hum Dev 77(1–2):13–22. doi:10.1016/j.earlhumdev.2003.11.006 Liao WQ, Zhou HY, Chen GC, Zou M, Lv X (2012) Left ventricular function in newborn infants of mothers with gestational diabetes mellitus. Zhongguo Dang Dai Er Ke Za Zhi 14(8):575–577 Mollema SA, Delgado V, Bertini M et al (2010) Viability assessment with global left ventricular longitudinal strain predicts recovery of left ventricular function after acute myocardial infarction. Circ Cardiovasc Imaging 3(1):15–23. doi:10.1161/ CIRCIMAGING.108.802785 Mondillo S, Galderisi M, Mele D et al (2011) Speckle-tracking echocardiography: a new technique for assessing myocardial function. J Ultrasound Med 30(1):71–83 Nearchou N, Tsakiris A, Tsitsirikos M et al (2005) Tei index as a method of evaluating left ventricular diastolic dysfunction in acute myocardial infarction. Hellenic J Cardiol. 46(35–42):57
453 16. Ng AC, Thomas L, Leung DY (2010) Tissue Doppler echocardiography. Minerva Cardioangiol 58:357–378 17. Nigro J, Osman N, Dart AM, Little PJ (2006) Insulin resistance and atherosclerosis. Endocr Rev 27(3):242–259. doi:10.1210/er. 2005-0007 18. Nold JL, Georgieff MK (2004) Infants of diabetic mothers. Pediatr Clin N Am 51:619–637. doi:10.1016/j.pcl.2004.01.003 19. Notomi Y, Lysyansky P, Setser RM et al (2005) Measurement of ventricular torsion by two-dimensional ultrasound speckle tracking imaging. J Am Coll Cardiol 45(12):2034–2041. doi:10. 1016/j.jacc.2005.02.082 20. Opara PI, Jaja T, Onubogu UC (2010) Morbidity and mortality amongst infants of diabetic mothers admitted into a special care baby unit in Port Harcourt, Nigeria. Ital J Pediatr 36(1):77. doi:10.1186/1824-7288-36-77 21. Picard MH, Adams D, Bierig SM et al (2011) American society of echocardiography recommendations for quality echocardiography laboratory operations. J Am Soc Echocardiogr 24(1):1–10. doi:10.1016/j.echo.2010.11.006 22. Pislaru C, Abraham TP, Belohlavek M (2002) Strain and strain rate echocardiography. Curr Opin Cardiol 17(5):443–454 23. Ren Y, Zhou Q, Yan Y, Chu C, Gui Y, Li X (2011) Characterization of fetal cardiac structure and function detected by echocardiography in women with normal pregnancy and gestational diabetes mellitus. Prenat Diagn 31(5):459–465. doi:10.1002/pd.2717 24. Review Gilbert-Barness E (2004) Metabolic cardiomyopathy and conduction system defects in children. Ann Clin Lab Sci. Winter 34(1):15–34 25. Roodpeyma S, Rafieyian S, Khosravi N, Hashemi A (2013) Cardiovascular complications in infants of diabetic mothers: an observational study in a pediatric cardiology clinic in Tehran. J Compr Pediatr 3(3):119–123 26. Sheehan PQ, Rowland TW, Shah BL, McGravey VJ, Reiter EO (1986) Maternal diabetic control and hypertrophic cardiomyopathy in infants of diabetic mothers. Clin Pediatr (Phila). 25(5):266–271 27. Su HM, Lin TH, Hsu PC et al (2013) Global left ventricular longitudinal systolic strain as a major predictor of cardiovascular events in patients with atrial fibrillation. Heart 99(21):1588–1596. doi:10.1136/heartjnl-2013-304561 28. Turan S, Turan OM, Miller J, Harman C, Reece EA, Baschat AA (2011) Decreased fetal cardiac performance in the first trimester correlates with hyperglycemia in pregestational maternal diabetes. Ultrasound Obstet Gynecol 38(3):325–331. doi:10.1002/uog. 9035 29. Turok DK, Ratcliffe SD, Baxley EG (2003) Management of gestational diabetes mellitus. Am Fam Physician 68(9):1767–1772
123
ORIGINAL ARTICLE
Speckle Tracking and Myocardial Tissue Imaging in Infant of Diabetic Mother with Gestational and Pregestational Diabetes Mohammed Al-Biltagi • Osama Abd Rab Elrasoul Tolba Mohamed Ahmed Rowisha • Amal El-Sayed Mahfouz • Mona Ahmed Elewa
•
Received: 23 January 2014 / Accepted: 26 September 2014 / Published online: 7 October 2014 Ó Springer Science+Business Media New York 2014
Abstract The aim of this study was to evaluate the myocardial changes in infants of diabetic mother either with gestational or pregestational diabetes and its relation to maternal diabetic control. The study included 45 infants of diabetic mother (IDMs) and 45 healthy newborn as a control group. IDMs were then categorized into 2 subgroups: twenty infants of mother with pregestational diabetes and twenty-five infants of mothers with gestational diabetes. The studied groups underwent measurement of the maternal and neonatal glycated Hb % (HbA1c), conventional echocardiography, tissue Doppler imaging (TDI) and two-dimensional speckle tracking imaging (STI). The weight, the rate of complications, and the rate of cesarean section were significantly higher in the IDMs group than in the control group. Significant positive correlation was present between the levels of HbA1c of IDMs and HbA1c of their mothers (P \ 0.05). A significant deterioration of both systolic and diastolic functions measured by both conventional echocardiography and TDI was present in IDMs with both pre-gestational and gestational diabetes compared with the control group. Also, the septal/posterior wall ratio (SW/ PW) was significantly higher in pregestational (1.86 ± 0.3) and gestational (2 ± 0.4) groups than in the control group (1 ± 0.06). Two-dimensional STI showed that the cardiac torsion was significantly impaired in pre-gestational (9.66 ± 2.5) and gestational (8.66 ± 3.9) groups when M. Al-Biltagi (&) O. A. R. E. Tolba M. A. Rowisha M. A. Elewa Pediatric Department, Faculty of Medicine, Tanta University, Medical Complex, Tanta, Egypt e-mail: [email protected] A. E.-S. Mahfouz Gynecology and Obstetrics Department, Faculty of Medicine, Tanta University, Tanta, Egypt
compared with the control group (5.4 ± 2.4) [P \ 0.0001]. It also showed that the global strain was significantly impaired in pre-gestational (-10.4 ± 3.2) and gestational (-13.1 ± 4.7) groups when compared with the control group (-19 ± 2) [P \ 0.0001]. However, no significant differences were present among the two patients’ subgroups in echocardiographic data except for a significant decrease of E’/A’ ratio and S wave at tricuspid annulus derived by TDI and impaired global strain derived by STI in infants of mothers with pre-gestational DM than those with gestational DM [P = 0.02]. SW/PW and cardiac torsion were significantly higher in infant of diabetic mother than the normal newborn and on the contrary systolic function and global strain were significantly lower in IDMs especially in infants of mother with pre-gestational diabetes. All the previous TDI findings did not show any significant correlation to neither maternal nor fetal HbA1c. Also, there was no significant correlation between cardiac torsion and the rest of TDI data neither in IDMs group nor in the control group. TDI and two-dimensional STI were efficient and sensitive tools able to early detect cardiac dysfunction in IDMs even in the absence of morphologic cardiac changes. Keywords Gestational diabetes Pre-gestational diabetes Infants of diabetic mothers Speckle tracking imaging Tissue Doppler imaging
Introduction Diabetes mellitus (DM) is a chronic metabolic disorder that affects in some ways all the stages of childhood; from in utero to neonates, infants, children, adolescents, and adults. Diabetes in pregnant women could be either pre-gestational
123
446
or gestational with an incidence of impaired glucose tolerance in pregnancy ranging between 3 and 10 %; which varies according to the average incidence of diabetes in the general population [20]. Its cardiovascular complications impose a significant morbidity risk in such active age groups due to metabolic derangement, the associated dyslipidemia, atherosclerosis, hypertension, and autonomic dysfunction [1]. The fetal heart is a target organ for the congenital effects of pre-existing as well as gestational maternal diabetes. It significantly affects the fetal heart and fetal–placental circulation in both functions and structures; causing cardiac anomalies and myocardial hypertrophy [6]. The reported incidence of these complications was 3.4 % with HbA1c levels lower than 8.5 and 22.4 % with HbA1c levels higher than 8.5 %. Infants born to mothers with an HbA1c level more than 10 % in late pregnancy are more prone to present with neonatal complications [18]. M-mode and 2-D echocardiography can show cardiomegaly (30 %), asymmetric septal hypertrophy, and fetal ventricular walls thickness that simulates idiopathic hypertrophic subaortic stenosis which increases progressively with advancing gestation [24]. Tissue Doppler imaging (TDI) is a relatively new echocardiographic technique useful to study the myocardial functions of fetal and neonatal hearts of abnormal gestation including pregnancies affected by diabetes mellitus [8]. Two-dimensional speckle tracking imaging (STI) is another relatively new echocardiographic modality that can provide non-Doppler, angle-independent, and objective quantification of myocardial deformation and left ventricular (LV) systolic and diastolic dynamics by analyzing the motion of speckles identified on routine 2-dimensional sonograms. By tracking the displacement of the speckles during the cardiac cycle; strain and the strain rate can be rapidly measured offline after sufficient image acquisition [14]. It can be used to measure interventricular septal thickness, posterior LV wall thickness, LV ejection fraction (EF), and LV rotation and torsion in IDMs [12]. The aim of this study was to evaluate the myocardial changes by conventional echocardiography, TDI and two-dimensional STI in infants of diabetic mothers either in pre-gestational or gestational diabetes and their relation to maternal glycemic control.
Patients and Methods The study was done as a case–control study included a group of 45 infants of diabetic mother (IDMs) and 45 healthy neonates with clinically and anatomically normal heart as a control group assigned from Neonatology and Cardiology Units; Pediatric Department, from a tertiary care hospital during the period from March, 2006 to April, 2009. The IDMs group was further divided into 2
123
Pediatr Cardiol (2015) 36:445–453
subgroups according to the type of the maternal diabetes (pre-gestational or gestational). The pre-gestational diabetes is the diabetes which started before the current pregnancy. Each subgroup was further subdivided according to the adequacy of control of maternal diabetes into adequately controlled and not adequately controlled. Maternal gestational diabetes is defined as controlled when fasting capillary blood glucose level of less than 95–105 mg per dL and the postprandial capillary blood glucose level of less than 140 mg per dL at one hour and less than 120 mg per dL at two hours [29]. Infants with intrauterine growth retardation (IUGR), multiple congenital anomalies, hypoxic ischemic encephalopathy (HIE), or chromosomal abnormalities, twins and preterm babies were excluded. Infants of mothers with history of hypertension, preeclampsia, chronic systemic disease, or on chronic medications except for insulin were also excluded from the study. Full history taking including gestational age, maternal name and age, type of maternal diabetes, type of delivery, and maternal blood glucose level was taken. A thorough clinical examination was done by pediatric neonatologist and pediatric cardiologist with special attention to cardiovascular system. Serial blood glucose levels were done in the first day of life and until the infants blood sugar remained stable with normal feedings. Measurements of HbA1c were done from the cord blood. Complete blood count, blood gases, C reactive protein (CR), blood glucose level, level of ionized calcium, renal and liver function tests, and chest X-ray were done when clinically indicated. Conventional echocardiography, two-dimensional STI and TDI were done for all cases within the first 2 postnatal days. To avoid intra-observer variability, two examinations were done by the same operator for each patient within the first 2 days of life, and we considered the average results. All the diabetic mothers with singleton pregnancy, and whom diagnosed as having diabetes during pregnancy following WHO criteria had HbA1c level from 36 week of gestation and onward before delivery. Conventional Echocardiography Echocardiographic examination was done using Vingmed Vivid-7 (General Electric Vingmed, and Milwaukee, Wisconsin, USA). Data acquisition was done with a 5-MHz transducer at a depth of 16 cm in all the standard echocardiographic views according to the recommendation of the American Society of Echocardiography [21]. All neonates were examined while lying quietly in the right anterior oblique position breathing room air. Cardiovascular malformations including patent ductus arteriosus were carefully searched and excluded by all standard views. Left ventricle ejection fraction (LV EF), LV
Pediatr Cardiol (2015) 36:445–453
fractional shortening (LV FS), LV end-diastolic and endsystolic volumes, systolic pulmonary artery pressure (PAP), and LV and right ventricle (RV) diastolic functions were measured according to the guidelines of the American Society of Echocardiography [21]. Tei index, which is a Doppler-derived time interval index that combines both systolic and diastolic cardiac performance, was also calculated as previously described by Tei and colleagues [10]. Hypertrophic cardiomyopathy (HCM) was diagnosed when the absolute values of inter ventricular septum and LV posterior wall diameters were above normal ranges, and in addition, the ratio of inter ventricular septum in diastole to the LV posterior wall in diastole exceeded 1.3. Two-Dimensional Speckle Tracking Imaging (STI) Longitudinal strain in 18 left LV segments using twodimensional STI were obtained using the 3-standard apical views; apical long axis, apical 4-chamber, and apical 2-chamber views. The parameters obtained represented the average of 3 cardiac cycles, with a frame rate of 65 frames/ s, and all the segmental data were represented. We used automated function imaging that enables the assessment of longitudinal strain available in Vivid 7 ultrasound machine to measure average LV global peak systolic strain. Global peak systolic strain was recorded in 3-standard apical views, and segmental peak systolic strain in basal, mid, and apical segments of anteroseptal, anterior, lateral, posterior, inferior, and septal LV walls. Left ventricular torsion was defined as the net difference of LV rotation at the basal and apical planes. For the STI analysis, we used high frame (65 frames/s) with harmonic two-dimensional images [19]. Tissue Doppler Imaging (TDI) Tissue Doppler imaging was done using the same machine and probe at a depth of 16 cm in the parasternal and apical views (standard long axis and two- and four-chamber images). Using pulsed-wave angle-corrected color-coded TDI filters, the baseline was adjusted to a low velocity range (-20–20 cm/s) and Doppler frame rates were varied between 80 and 115 frames/s depending on the sector width of the range of interest with minimal gain setting to reduce background noise and to get the highest quality images. Two millimeters sample volume was placed within the myocardium equidistant from the endocardial and epicardial borders. From the apical four-chamber planes, using pulsed-wave TDI, the myocardial velocity curves of septal mitral valve annulus, lateral mitral valve annulus, and lateral tricuspid valve annulus were recorded. The electrocardiogram and respiration curve monitoring were connected and traced simultaneously to define the timing of cardiac cycle events and its relation to respiratory events.
447
The beginning of QRS complex was used as a reference point. At least 10 cardiac cycles were recorded at a speed of 100 mm/s, and the images were stored electronically. TDI measurements were indexed for children’s body surface area [2]. The mean values for three heart beats during expiration were used for the analysis. The systolic wave (S) reflects the systolic function of either RV or LV. The early/atrial (E’/A’) ratio of tricuspid and mitral valve annulus reflects the diastolic function of the RV and LV, respectively. Isometric contraction time [ICT] (was defined as the time duration between the beginnings of QRS complex in the electrocardiogram to the beginning of TDI S wave), the isometric relaxation time [IRT] (was defined as the interval between the end of S wave and the beginning of the early wave), and myocardial contraction time (CT) were measured by TDI. Tei index (myocardial performance index, MPI) was calculated using the following formula: ICT ? IRT/CT. Both ICT and IRT were corrected for heart rate [16]. All the data were adjusted for gestational age and estimated neonatal weight. The mothers of the studied newborns were informed of the purpose of the study before participating in the study, and all the parents of the included newborns signed a written-informed consent. The study was approved by the Institutional Ethical and Research Review Board of Faculty of Medicine, Tanta University.
Statistics The statistical power of the study was more than 90 % (using a computerized program: Power and Precision V3; www.PowerAnalysis.com). Data were presented as mean (±standard deviation) values. The two-way analysis of variance with repeated measures and chi-square test by SPSS V.16 were used to find statistically significant differences in the different parameters among the groups. For all analyses, a statistical significance of P value less than 0.05 was used. Wilcoxon’s signed-rank test was used to assess the normality of distributions of the data. The Bonferroni correction/adjustment procedure was done to avoid ‘‘significance’’ attributable to chance only in multiple comparisons with echocardiographic parameters. Correlation between maternal HbA1c, neonatal HbA1c, and echocardiographic variables was evaluated using Pearson’s correlation coefficient.
Results Figure 1 showed the flow chart of the study that included 45 IDM neonates and 45 healthy neonates as a control group. Among the IDMs, there were 25 IDMs with gestational DM
123
448
Pediatr Cardiol (2015) 36:445–453
Fig. 1 Flow chart of the studied groups
Table 1 Clinical and laboratory data of IDMs and control group Variables (Mean ± SD)
IDMs n = 45
Controls n = 45
t
P
Gestational age (weeks) (Mean ± SD)
38.8 ± 1.2
39 ± 2
0.57
0.56
Males, n (%)
21 (46.6%)
22 (48.8%)
0.21
0.8
Females, n (%)
24 (53.4%)
23 (51.2%)
0.2
0.83
3.54 ± 0.54
3.23 ± 0.34
3.25
0.0016
Respiratory distress Neonatal jaundice (including physiological jaundice)
9 (20%) 30 (66.6%)
0 15 (33%)
3.2 3.1
0.002 0.002
Hypocalcaemia
11 (24%)
0
3.5
0.0007
Hypoglycemia
6 (13%)
0
2.5
0.01
Hypothyroidism
5 (11%)
1 (2.2%)
1.6
0.09
Sex
Birth Weight (kg) (Mean ± SD) Postnatal problems, n (%):
Mode of delivery: Cesarean section, n (%)
36 (80%)
16 (35.5%)
4.2
0.0001
Vaginal delivery, n (%)
9 (20%)
29 (64.6%)
4.2
0.0001
28.6 ± 4.8
27 ± 5.6
1.45
0.15
Maternal age (years) Mean ± SD P value \ 0.05 is significant
and 20 IDMs with pre-gestational DM. Tables 1 and 2 showed the demographics of the included groups. Table 1 showed significant differences between the control group and the IDMs in gestational age, maternal age and sex while the weight, the rate of complications, and rate of cesarean section were significantly higher in the IDMs than in the control group. Table 2 showed no significant differences between the two subgroups of IDMs (pre-gestational and gestational diabetic mothers) in their demographic data.
123
Table 3 showed echocardiographic data in IDMs of mothers with pre-gestational and gestational diabetes and control groups. The intra-observer agreement K values of the echocardiographic measurements were between 77 and 88 which indicated good agreement between the 2 readings. The table showed the presence of significant deterioration of both LV and RV systolic and diastolic functions measured by both conventional echocardiography and TDI among IDMs of both pre-gestational and gestational
Pediatr Cardiol (2015) 36:445–453
449
Table 2 Clinical and laboratory data of babies and mothers of pre-gestational and gestational diabetes group Variables
Pre-gestational DM group (Group 1) n = 20
Gestational DM group (Group 2) n = 25
P
Gestational age (weeks) Mean ± SD
39 ± 1.2
38.7 ± 1.48
0.45
Males, n (%)
11 (55%)
10 (40%)
0.3
Females, n (%)
9 (45.4%)
15 (60%)
0.33
3.65 ± 0.64
3.44 ± 0.44
0.22
Respiratory distress Neonatal jaundice (including physiological jaundice)
5 (25%) 14 (70%)
4 (16%) 16 (64%)
0.4 0.6
Hypocalcemia
6 (30%)
5 (20%)
0.44
Hypoglycemia
3 (15%)
3 (12%)
0.7
Hypothyroidism
3 (15%)
2 (8%)
0.46
Cesarean section, n (%)
15 (75%)
21 (84%)
0.45
Vaginal delivery, n (%)
5 (25%)
4 (16%)
0.45
28.9 ± 5.07
28.4 ± 4.7
0.7
Sex
Birth weight Kg (Mean ± SD) Postnatal problems, n (%):
Mode of delivery
Maternal age (years) Mean ± SD Maternal glycemic control: Controlled, n (%)
11 (55%)
15 (60%)
0.73
Uncontrolled, n (%)
9 (45%)
10 (40 %)
0.73
P value \ 0.05 is significant
diabetes compared with the control group. Also, the septal/ posterior wall ratio was significantly higher in patients groups than in the control group (P \ 0.0001 for both subgroups). However, no significant differences were present among the two patients subgroups in conventional echocardiographic data, but TDI showed significant decrease of E’/A’ ratio (P \ 0.0001) and S wave (P \ 0.0001) at tricuspid annulus in infants of mothers with pre-gestational DM than in infants of mothers with gestational DM. Two-dimensional STI showed significantly impaired cardiac torsion in two patients groups when compared with the control group (P \ 0.0001 for both subgroups). However, cardiac torsion was more impaired in IDMs of mother with pre-gestational diabetes than those of mothers with gestational diabetes but without statistical significance (P [ 0.05). STI showed also significant impairment of the global strain among IDMs when compared with the control group (P \ 0.0001 for both subgroups). However, the global strain was significantly impaired in infants of mothers with pre-gestational DM than in infants of mothers with gestational DM (P \ 0.05). Tables 4 and 5 showed the absence of significant correlation between any of the echocardiographic parameters (including conventional echocardiography, TDI and twodimensional STI) and neither maternal nor neonatal HbA1c. However, significant positive correlation was present between the levels of HbA1c of IDMs and HbA1c of their
mothers in both pre-gestational (P \ 0.05) and gestational diabetes subgroups (P \ 0.05). No significant differences were present between HbA1c of the mothers with pre-gestational and gestational diabetes and any of the cardiac TDI data including Septal/posterior wall (sw/pw), E’/A’ ratio or S wave of mitral annulus, LV global strain and LV torsion. On the other hand, the infants’ birth weight had significant positive correlation with neonatal (P \ 0.05) and maternal (P \ 0.05) HbA1c value and septal/posterior wall ratio (P \ 0.05) but negatively correlated with the cardiac torsion (P \ 0.05) as shown in Table 6.
Discussion Cardiac functional abnormalities are present in about 30 % of IDMs and include intraventricular septal hypertrophy and cardiomyopathy and 10 % may have frank heart failure. Hypertrophic cardiomyopathy observed in the infant of the diabetic mother is characterized by thickening of intraventricular septum, and to a lesser extent the ventricular free walls [20]. In our study, a significant increase in LV septal/posterior wall ratio was present in the IDMs groups than the control indicating septal thickening. Various echocardiographic modalities showed that the IDMs (with either gestational or pre-gestational DM) had impairment of both systolic and diastolic functions as observed by both
123
450
Pediatr Cardiol (2015) 36:445–453
Table 3 Comparison of Echocardiographic Data in IDMs (pre-gestational and gestational diabetes) and control groups Pre-gestational DM (Group 1) (n = 20) LV FS
E/A ratio MV
Mean ± SD
Mean ± SD
38.04 ± 4.12
0.89 ± 0.07
Gestational DM (Group 2) (n = 25) 38.3 ± 4.4
0.92 ± 0.08
Control Group (n = 45)
t
41.56 ± 5.14
t1
-2.94
0.0052
t2
-2.79
0.007
t3
0.2
1.26 ± 0.13
P value
0.8393
t1
-14.85
\0.0001
t2
-13.5
\0.0001
t3
1.3
0.18
12.8 11.2
\0.0001 \0.0001
Septal/posterior wall (sw/pw)
Mean ± SD
1.86 ± 0.28
2 ± 0.42
1.05 ± 0.06
t1 t2 t3
1.3
E’/A’ wave mitral annulus
Mean ± SD
0.70 ± 0.13
0.76 ± 0.14
1.27 ± 0.14
t1
-15.9
\0.0001
t2
-14.6
\0.0001
S wave mitral annulus (Cm/s))
LV TEI
Pul systolic Pressure (mmHg)
Mean ± SD
Mean ± SD
Mean ± SD
3.2 ± 0.6
0.45 ± 0.061
38.5 ± 5
3.4 ± 0.8
0.43 ± 0.05
36.5 ± 6
6.1 ± 0.5
0.38 ± 0.06
30.5 ± 4
0.18
t3
1.5
0.14
t1
18.9
\0.0001
t2
15.3
\0.0001
t3
-0.96
0.34
t1
4.3
\0.0001
t2
3.7
0.0004
t3
1.2
0.24
t1
6.31
\0.0001
t2
4.47
\0.0001
t3
1.2 36 26.1
0.2 \0.0001 \0.0001
E’/A’ wave tricuspid annulus
Mean ± SD
0.84 ± 0.03
0.87 ± 0.04
1.11 ± 0.03
t1 t2 t3
4.7
\0.0001
S wave tricuspid annulus cm/s)
Mean ± SD
5.3 ± 0.5
5.9 ± 0.4
7.4 ± 0.3
t1
17.4
\0.0001
t2
16.3
\0.0001
t3
4.3
0.0001
t1
6.7
\0.0001
t2
4
t3
1.9
0.0621
t1
7.5
\0.0001
t2
8.7
\0.0001
t3
2.3
0.02
t1
6.5
\0.0001
t2
3.75
0.0006
t3
1.04
0.3
RV TEI
Global strain (GLS)
Torsion (T)
Mean ± SD
Mean ± SD
Mean ± SD
0.42 ± 0.03
-10.4 ± 3.22
9.66 ± 2.46
0.40 ± 0.04
-13.1 ± 4.70
8.66 ± 3.95
0.36 ± 0.04
-19 ± 2.05
5.40 ± 2.41
0.0002
t1 is the comparison between control group and group 1; t2 is the comparison between control group and group 2; t3 is the comparison between Group 1 and group 2 P value \0.05 is significant
conventional echocardiography (lower LV FS, lower E/A ratio at mitral valve, and prolonged Tie index) and TDI (lower S wave and E’/A’ wave at both mitral and tricuspid valve annuli). In our study, LV and RV Tie indices were prolonged in IDMs of mother with either gestational or pregestational DM than in healthy control infants. The Tie index is a simple, noninvasive, simple to estimate, and
123
reproducible index which has close correlation with the widely accepted systolic and diastolic hemodynamic parameters as well as the potential for clinical application in the assessment of overall cardiac performance including isovolumic contraction time (ICT), ejection time (ET), and isovolumic relaxation time [15]. In our study, the pulmonary systolic pressure was significantly higher in IDMs than
Pediatr Cardiol (2015) 36:445–453
451
Table 4 Correlation between glycated Hb of the babies in pre-gestational and gestational diabetic groups and their cardiac tissue Doppler data
Septal/posterior wall (sw/pw)
Glycated Hb of the babies
Babies’ birth weight
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
R
R
R
R
P value 0.411
0.080
0.756
E’/A’ wave mitral annulus
-0.145
0.606
-0.128
0.650
S wave mitral annulus (Cm/s))
-0.555
0.032
-0.376
0.219
0.432
Torsion (T) Glycated Hb of the mothers
-0.440 0.362
P value
P value
P value
0.299
Global strain (GLS)
Table 6 Correlation between babies’ birth weight in pre-gestational and gestational diabetes groups and their glycated Hb, their cardiac Torsion, their SW/PW, and their maternal glycated Hb
Babies’ glycated Hb
0.425
0.011*
0.396
0.047*
Maternal glycated Hb
0.396
0.020*
0.428
0.023*
Cardiac Torsion (T)
-0.357
0.044*
-0.362
0.043*
0.178
Septal/posterior wall (sw/pw)
-0.425
0.019*
0.388
0.028*
0.128
0.649
P value \0.05 is significant
0.001*
0.265
0.341
0.044
0.396
0.021
P value \0.05 is significant
in infants of non-diabetic mothers. This finding agreed with the work of Dimitriu et al. who observed the presence of pulmonary hypertension in 3 cases out of 35 IDMs [4]. Koza´k-Ba´ra´ny et al. showed that infants of mothers with well-controlled pre-gestational or gestational diabetes had impaired LV relaxation. They related this finding to the effects of maternal hyperglycemia during the third trimester and subsequent fetal hyperinsulinemia leading to neonatal cardiac hypertrophy which also agreed with our study findings [11]. However, despite Ren et al. and Hate´m et al. showed septal thickening among IDMs and that the pulsed TDI demonstrated evidence of impaired diastolic function, in fetuses of diabetic mothers compared with fetuses of non-diabetic mothers; yet; they found that this dysfunction was independent of the presence of myocardial hypertrophy [7, 23]. In our study, no significant differences were present in the conventional echocardiographic parameters and TDI between IDMs of mothers with gestational diabetes and those of mothers with pre-gestational diabetes except for presence of lower E’/A’ ratio and S wave at tricuspid annulus by TDI in IDMs of mothers with
pre-gestational diabetes than those of mothers with gestational diabetes. However, TDI has important limitations. TDI measures only the vector of motion that is parallel to the direction of the ultrasound beam and it measures absolute tissue velocity. It is also unable to discriminate passive motion (related to translation or tethering) from active motion (fiber shortening or lengthening) [8]. To overcome these limitations of TDI, we did strain rate imaging, using STI which has the ability to quantify regional wall deformation and ability to differentiate between active and passive motion by quantification of myocardial deformation. It allows an objective and quantitative evaluation of global and regional myocardial function independently from the angle of insonation and from cardiac translational movements [22]. STI-derived measurements have high feasibility and reproducibility and are comparable to that taken by sonomicrometry and tagged MRI [3]. We found that both the global strain (GLS) and the cardiac torsion were significantly impaired in the IDMs than in the control group. Also, we found that GLS was significantly impaired in IDMs of mother with pregestational diabetes than those of mothers with gestational diabetes. However, there was no significant difference in cardiac torsion between the two patients’ subgroups. The result of our study agreed with the previous findings of
Table 5 Correlation between torsion in pre-gestational and gestational diabetes groups and other tissue Doppler data Torsion (T)
Septal/posterior wall (sw/pw)
Pre-gestational DM (Group 1) (n = 20)
Gestational DM (Group 2) (n = 25)
Control group (n = 45)
R
R
P value
R
P value
P value
-0.068
0.809
-0.287
0.300
-0.325
0.225
E’/A’ wave mitral annulus
0.088
0.745
-0.139
0.629
0.152
0.341
S wave mitral annulus (Cm/s) Global strain (GLS)
0.257 0.129
0.356 0.649
-0.545 0.119
0.036* 0.673
0.321 0.155
0.280 0.625
P value \0.05 is significant
123
452
Liao et al. who showed that the cardiac functions may be impaired in newborn infants of mothers with gestational diabetes, with changes in LV shape and abnormalities of LV rotation and torsion even in the presence of normal ventricular ejection fraction. They recommended using two-dimensional STI for early detection of LV dysfunction [12]. Global LV longitudinal systolic strain (GLS) is a sensitive and simple measure of LV mechanics that can be obtained by STI. It takes a few to 10 minutes to be done and has more ability to predict adverse cardiovascular events than LVEF and TDI-derived S wave. It could offer an additional prognostic benefit over clinical, conventional echocardiographic, and TDI systolic parameters [27]. However, it depends on the level of the operator’s experience, needs high-quality images for assessment of global LV strain and is frame rate-dependent. The frame rate of the obtained images in our study was 65 frames/s. A major limitation of STI that can affect the accuracy of imaging process and data obtained is its strict dependency on the frame rate and the need for high-quality two-dimensional images, which are necessary for obtaining an optimal definition of the endocardial border [13]. The reason of impaired GLS in IDMs of mother with pre-gestational diabetes than those of mothers with gestational diabetes may be related to its early effect on the fetus which can start as early as 11–14 weeks of pregnancy. Turan et al. showed that fetuses of poorly controlled diabetic mothers demonstrated significant differences in firsttrimester diastolic myocardial function compared with nondiabetic controls. The decrease in myocardial performance was more marked with increasing HbA1c and appears to be independent of preload and after load [28]. These early cardiac effects of pre-gestational DM that started during the first trimester could affect the early embryonic development and alter the cardiac morphogenesis and placental development. These effects continue through the second and third trimesters to affect the fetal circulation and also can extend to the perinatal and neonatal period exposing the IDMs to the same effects of gestational diabetes including the metabolic derangement [9]. In our study, no significant correlation was present between SW/PW ratio, e’/a’ ratio, S wave, (cardiac torsion, and GLS in IDMs in the two patients subgroups with neither maternal nor neonatal HbA1. This agreed with the findings of Sheehan et al. who confirmed the absence of relationship between echocardiographic evidence of hypertrophic cardiomyopathy and maternal HbA1 levels. Their data did not support the third trimester maternal hyperglycemia as instrumental etiology of cardiomyopathy and other complications observed in IDMs [26]. However, El-Ganzoury et al. found a positive correlation between maternal HbA1c and IVS thickness and impaired LV
123
Pediatr Cardiol (2015) 36:445–453
contractility [5]. Meanwhile, our study showed significant positive correlation between neonatal birth weight and neonatal and maternal HbA1c value and SW/PW ratio but negatively correlated with the cardiac torsion. These finding agreed with the result of El-Ganzoury et al. who found that the birth weight is the best predictor of hypertrophied IVS especially in infants born to sub optimally controlled diabetic mothers [5]. On the other hand, Roodpeyma et al. showed neither the types of maternal diabetes nor the somatic findings of newborns that were related to the occurrence of cardiac complications [25]. The relation between the increase in birth weight and cardiac dysfunction could be attributable to the same mechanism. For example, the insulin resistance-associated hyperinsulinemia that is frequently associated with pregnancy-associated diabetes can induce smooth-muscle cell hypertrophy and hyperplasia and increased extracellular proteins which are an important factor contributing to increase birth weight and at the same time affects the cardiac function [17]. Our study has certain limitations such as few numbers of the studied subgroups and lack of follow-up period. We did not do cardiac magnetic resonance (CMR) to compare STI measurements with it. Further studies are suggested on larger number of IDMs and for longer duration.
Conclusion SW/PW and cardiac torsion were significantly higher in IDMs than the healthy newborn, and on the contrary, systolic and diastolic functions and global strain were significantly impaired in IDMs especially in babies in pregestational diabetes group. All the previous TDI findings show insignificant correlation to neither maternal nor fetal HbA1c. Also, no significant correlation was present between cardiac torsion and the rest of TDI data in infants of both patients’ subgroups. TDI and two-dimensional STI are efficient and sensitive tools able to early detect cardiac dysfunctions in IDMs even in the absence of morphologic cardiac changes.
References 1. Al-Biltagi, M (2014) Cardiovascular effects of diabetes mellitus in pediatric population. In: Research on Diabetes II. ISBN: 978-1922227-26-3 . iConcept Press, Hong Kong 2. Al-Biltagi M, Serag AR, Hefidah MM, Mabrouk MM (2013) Evaluation of cardiac functions with Doppler echocardiography in children with Down syndrome and anatomically normal heart. Cardiol Young 23(2):174–180. doi:10.1017/S1047951112000613 3. Amundsen BH, Helle-Valle T, Edvardsen T et al (2006) Noninvasive myocardial strain measurement by speckle tracking
Pediatr Cardiol (2015) 36:445–453
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
15.
echocardiography: validation against sonomicrometry and tagged magnetic resonance imaging. J Am Coll Cardiol 47(4):789–793. doi:10.1016/j.jacc.2005.10.040 Dimitriu AG, Russu G, Stamatin M, Jita˘reanu C, Streanga˘ V (2004) Clinical and developmental aspects of cardiac involvement in infant of diabetic mother. Rev Med Chir Soc Med Nat Iasi 108(3):566–569 El-Ganzoury MM, El-Masry SA, El-Farrash RA, Anwar M, Abd Ellatife RZ (2012) Infants of diabetic mothers: echocardiographic measurements and cord blood IGF-I and IGFBP-1. Pediatr Diabetes 13(2):189–196. doi:10.1111/j.1399-5448.2011.00811.x Gardiner HM, Pasquini L, Wolfenden J, Kulinskaya E, Li W, Henein M (2006) Increased periconceptual maternal glycated haemoglobin in diabetic mothers reduces fetal long axis cardiac function. Heart 92:1125–1130. doi:10.1136/hrt.2005.076885 Hate´m MA, Zielinsky P, Hate´m DM et al (2008) Assessment of diastolic ventricular function in fetuses of diabetic mothers using tissue Doppler. Cardiol Young 18(3):297–302. doi:10.1017/ S1047951108002138 Ho CY, Solomon SD (2006) A clinician’s guide to tissue Doppler imaging. Circulation 113(10):e396–e398. doi:10.1161/CIRCU LATIONAHA.105.579268 Hornberger LK (2006) Maternal diabetes and the fetal heart. Heart 92(8):1019–1021. doi:10.1136/hrt.2005.083840 Karatzis EN, Giannakopoulou AT, Papadakis JE, Karazachos AV, Nearchou NS (2009) Myocardial performance index (Tei index): evaluating its application to myocardial infarction. Hellenic J Cardiol 50(1):60–65 Koza´k-Ba´ra´ny A, Jokinen E, Kero P, Tuominen J, Ro¨nnemaa T, Va¨lima¨ki I (2004) Impaired left ventricular diastolic function in newborn infants of mothers with pregestational or gestational diabetes with good glycemic control. Early Hum Dev 77(1–2):13–22. doi:10.1016/j.earlhumdev.2003.11.006 Liao WQ, Zhou HY, Chen GC, Zou M, Lv X (2012) Left ventricular function in newborn infants of mothers with gestational diabetes mellitus. Zhongguo Dang Dai Er Ke Za Zhi 14(8):575–577 Mollema SA, Delgado V, Bertini M et al (2010) Viability assessment with global left ventricular longitudinal strain predicts recovery of left ventricular function after acute myocardial infarction. Circ Cardiovasc Imaging 3(1):15–23. doi:10.1161/ CIRCIMAGING.108.802785 Mondillo S, Galderisi M, Mele D et al (2011) Speckle-tracking echocardiography: a new technique for assessing myocardial function. J Ultrasound Med 30(1):71–83 Nearchou N, Tsakiris A, Tsitsirikos M et al (2005) Tei index as a method of evaluating left ventricular diastolic dysfunction in acute myocardial infarction. Hellenic J Cardiol. 46(35–42):57
453 16. Ng AC, Thomas L, Leung DY (2010) Tissue Doppler echocardiography. Minerva Cardioangiol 58:357–378 17. Nigro J, Osman N, Dart AM, Little PJ (2006) Insulin resistance and atherosclerosis. Endocr Rev 27(3):242–259. doi:10.1210/er. 2005-0007 18. Nold JL, Georgieff MK (2004) Infants of diabetic mothers. Pediatr Clin N Am 51:619–637. doi:10.1016/j.pcl.2004.01.003 19. Notomi Y, Lysyansky P, Setser RM et al (2005) Measurement of ventricular torsion by two-dimensional ultrasound speckle tracking imaging. J Am Coll Cardiol 45(12):2034–2041. doi:10. 1016/j.jacc.2005.02.082 20. Opara PI, Jaja T, Onubogu UC (2010) Morbidity and mortality amongst infants of diabetic mothers admitted into a special care baby unit in Port Harcourt, Nigeria. Ital J Pediatr 36(1):77. doi:10.1186/1824-7288-36-77 21. Picard MH, Adams D, Bierig SM et al (2011) American society of echocardiography recommendations for quality echocardiography laboratory operations. J Am Soc Echocardiogr 24(1):1–10. doi:10.1016/j.echo.2010.11.006 22. Pislaru C, Abraham TP, Belohlavek M (2002) Strain and strain rate echocardiography. Curr Opin Cardiol 17(5):443–454 23. Ren Y, Zhou Q, Yan Y, Chu C, Gui Y, Li X (2011) Characterization of fetal cardiac structure and function detected by echocardiography in women with normal pregnancy and gestational diabetes mellitus. Prenat Diagn 31(5):459–465. doi:10.1002/pd.2717 24. Review Gilbert-Barness E (2004) Metabolic cardiomyopathy and conduction system defects in children. Ann Clin Lab Sci. Winter 34(1):15–34 25. Roodpeyma S, Rafieyian S, Khosravi N, Hashemi A (2013) Cardiovascular complications in infants of diabetic mothers: an observational study in a pediatric cardiology clinic in Tehran. J Compr Pediatr 3(3):119–123 26. Sheehan PQ, Rowland TW, Shah BL, McGravey VJ, Reiter EO (1986) Maternal diabetic control and hypertrophic cardiomyopathy in infants of diabetic mothers. Clin Pediatr (Phila). 25(5):266–271 27. Su HM, Lin TH, Hsu PC et al (2013) Global left ventricular longitudinal systolic strain as a major predictor of cardiovascular events in patients with atrial fibrillation. Heart 99(21):1588–1596. doi:10.1136/heartjnl-2013-304561 28. Turan S, Turan OM, Miller J, Harman C, Reece EA, Baschat AA (2011) Decreased fetal cardiac performance in the first trimester correlates with hyperglycemia in pregestational maternal diabetes. Ultrasound Obstet Gynecol 38(3):325–331. doi:10.1002/uog. 9035 29. Turok DK, Ratcliffe SD, Baxley EG (2003) Management of gestational diabetes mellitus. Am Fam Physician 68(9):1767–1772
123
