Interpretation of results and comparison with existing literature
A number of studies have demonstrated that in utero exposure to hyperglycaemia can adversely affect the fetal heart (3, 4, 10). Consistent with this, in our study, we showed that fetuses exposed to GDM have altered heart morphology with more globular hearts compared to that seen in controls and the difference in shape were driven by changes in the right ventricle. By using a variety of echocardiographic modalities we also showed that fetuses of mothers with GDM, compared to controls, have subtle functional cardiac changes which can be identified only by using more advanced imaging modalities. Right and left longitudinal myocardial deformation were reduced in fetuses exposed to GDM compared to controls but diastolic indices were similar between the two groups when analysis was adjusted for differences in maternal characteristics, estimated fetal weight and heart rate. Similar results have been reported before, in some studies (11, 12) but not in all (13). For example, Miranda et al in a combined group of 76 women with pregestational diabetes and GDM demonstrated biventricular diastolic dysfunction in their fetuses (4). Although measurements of diastolic function commonly precede systolic functional changes, these are more difficult to accurately assess in fetal life. In our study, we used speckle tracking analysis to assess the rate of change in the right and left myocardial deformation as well as conventional and tissue Doppler imaging. We followed a strict protocol for image acquisition using high frames per rate as per recent guidelines(14) and performed the analysis without compromising our temporal resolution. It is possible therefore that the noted discrepancies are due to differences in the study population ie inclusion on pregestational diabetic women in Miranda’s study as well as differences in the software used for speckle tracking analysis (15).
The influence of GDM, however, might not be limited to fetal life as observational data suggest that maternal diabetes before or during pregnancy is associated with increased rate of early onset cardiovascular disease from childhood to adulthood (1,2). To date, only few studies have assessed offspring of diabetic mothers spanning from fetal to neonatal life and provided conflicting results about the presence of persistent cardiac changes (6, 16, 17). For instance, Pateyet al, in a group of 21 neonates of mothers with pregestational diabetes or GDM, compared to controls, demonstrated persistent alterations in left ventricular chamber geometry in the perinatal period (5) and Zablahet al, in a retrospective study reported that 75 neonates who were exposed to pregestational diabetes or GDM, compared to controls, had decreased left ventricular systolic and diastolic function in the first week of life (17). In contrast, Mehta et al, documented in 50 newborns of mothers with pregestational diabetes or GDM, that early cardiac changes such as reduced diastolic ventricular function and myocardial hypertrophy are transient and resolve in the first month of life (6). However, cardiac assessment in the neonatal period is also affected by changes in loading conditions, which relate to closure of cardiac shunts and the change from a parallel circulation to one in series as part of the physiological adaptation to postnatal life and this may obscure small differences to become apparent in offspring of women with GDM compared to controls. Therefore, to minimize this confounding effect, in our study we elected to study children after the first few months of life. We showed that infants exposed prenatally to GDM have increased diastolic functional indices and reduced biventricular systolic function compared to controls, whereas no differences in left ventricular mass was noted. Cardiac changes were seen in both ventricles but more apparent in the left ventricle and remained after accounting for differences in maternal characteristics, infant weight gain and interval of cardiac assessment since birth. However, it remains to be established whether the noted cardiac functional changes persist in childhood and contribute to the reported increased cardiovascular disease risk noted in offspring of diabetic mothers.
In the management of GDM, insulin therapy is often added when diet or oral pharmacological treatments fail to establish good glycemic control. Although insulin may have growth stimulating effects on the myocardium (18) this does not cross the placenta and is unlikely to affect directly the fetal heart (19). However, from different hypoglycaemic treatments, it is well described that metformin crosses the placenta and concerns were raised regarding long term programming effects on fetal metabolism as well as its impact on fetal heart with sustained effects in childhood (20, 21). In our study, there was no difference in cardiac indices both in fetal or postnatal life between the treatment groups thus our results would not support such a hypothesis.