Objective
We sought to determine if fetal hypertrophic cardiomyopathy (HCM) or cardiac dysfunction is associated with elevated maternal or neonatal insulin-like growth factor (IGF)-I levels in women with diabetes.
Study Design
In a prospective cohort study, fetal echocardiogram findings at 36 weeks’ gestation in women with pregestational diabetes mellitus were compared to those in women without diabetes mellitus. HCM was defined as septal or free wall thickness ≥5 mm and cardiac dysfunction as a modified myocardial performance index ≥0.43. Cord serum IGF-I levels at delivery were measured with enzyme-linked immunosorbent assay. Neonates with abnormal fetal echocardiogram were followed up until resolution or 6 months of life.
Results
In all, 75 participants completed fetal echocardiography (55 diabetics and 20 controls). In the diabetic group, 33 of 55 (60%) had abnormal fetal echocardiograms with cardiac dysfunction in 21 of 55 (38.2%) and HCM in 8 of 55 (14.5%) and both in 4 of 55 (7.3%). At 6 months of age, 1 of 12 (8%) had persistent HCM. None in the comparison group had abnormal findings. There were no significant clinical differences in those diabetic women with normal vs abnormal fetal echocardiograms. However, among diabetic women, mean neonatal IGF-I was significantly higher in fetuses with HCM (80 ± 16 ng/mL) as compared to those without HCM (61 ± 18 ng/mL), ( P < .001).
Conclusion
Elevated neonatal IGF-I appears to be associated with fetal HCM in fetuses of diabetic women.
Maternal pregestational diabetes complicates 2-8% of pregnancies in the United States, potentially exposing >300,000 infants to maternal diabetes each year. Despite dramatic improvements in maternal glycemic control, infants born to diabetic mothers are still at a 5-fold increased risk of stillbirth (2.5%) and perinatal mortality (2.9%) than infants born to nondiabetic mothers. Of these stillborn fetuses, 22% will have some degree of cardiomegaly in the absence of other congenital anomalies, suggesting that cardiac overgrowth may be a factor in some of these fetal deaths.
Hypertrophic cardiomyopathy (HCM), a condition characterized by thickening of the ventricular septum and/or hypertrophied ventricular muscle, can be induced by fetal or neonatal hyperinsulinemia. However, both cardiac dysfunction and HCM are evident even with excellent maternal glycemic control. Notably, 30-40% of infants of diabetic mothers demonstrate the ventricular and/or septal hypertrophy that characterizes HCM ( Figure ). Strict maternal glucose control reduces the severity of the disease but does not appear to alter the incidence of HCM in infants of diabetic mothers. These findings suggest that fetal glucose and insulin both contribute to HCM but are not solely responsible for its development. The pathways governing fetal cardiac development and the mechanisms that promote fetal HCM are as yet unknown.
In animals models, insulin-like growth factor (IGF)-I signaling pathways are implicated in the development of HCM. In mice, overexpression of IGF-I promotes increased heart size via cardiomyocyte proliferation. In a rapidly hypertrophying rat model, an increase in the synthesis of IGF-I by cardiomyocytes is an early event in the development of cardiomyopathy. In sheep, intrafetal infusion of IGF-I in late gestation promotes cardiomyocyte enlargement and proliferation via phophoinositol-3 kinase–mediated pathways. Taken together, these observations from animal models implicate IGF-I–mediated cardiomyocyte growth in the development of HCM. In 1 small clinical study, significant elevation of maternal IGF-I at the time of delivery was noted in 6 diabetic mothers whose infants had HCM. This observation suggests that elevated maternal IGF-I may be predictive of fetal cardiovascular overgrowth. Our study aims to further explore the association between IGF-I and the development of fetal HCM. While neonatal serum levels of IGF-I have been evaluated with regard to overall fetal growth and elevations are tightly associated with fetal macrosomia, to our knowledge, the association with fetal HCM has not been reported. The objective of this study was to evaluate the incidence of fetal cardiovascular overgrowth and cardiac dysfunction in patients with pregestational diabetes, and to determine if abnormal cardiac development or function are associated with elevated maternal or neonatal serum IGF-I levels.
Materials and Methods
After institutional review board approval was granted (EH11-241: fetal cardiovascular overgrowth and maternal diabetes), a prospective cohort study was conducted. Eligible women were those with either type 1 or type 2 diabetes mellitus with a singleton pregnancy and a nonanomalous fetus at ≤36 weeks’ gestation. Nondiabetic women with a singleton pregnancy and a nonanomalous fetus at ≤36 weeks with a normal 1-hour glucose tolerance test served as our comparison group. Based on the reported rate of HCM in the diabetic population, we assumed that there would be 20 cases of HCM in our cohort and chose to do a 1:1 ratio of HCM cases to controls. As maternal obesity can influence fetal growth, the patients in the control group were matched by body mass index to the patients in the diabetes group. No other matching was performed. All recruitment occurred between November 2011 and May 2013.
A comprehensive fetal echocardiogram was performed between 36 and 36 6/7 weeks’ estimated gestational age to assess fetal cardiac function and development as the fetus approached full term. All fetal echocardiograms were performed by a pediatric cardiologist (L.Y.) or a maternal fetal medicine fellow who received specialized training (A.B.G.). Fetal echocardiograms were performed according to the American Society of Echocardiography guidelines. All cardiac images were reviewed by a single pediatric cardiologist (L.Y.) to assess for quality and accuracy of measurements. HCM was defined as septal or free wall thickness of ≥ 5 mm and cardiac dysfunction was described as a myocardial performance index (MPI) of ≥0.43. The MPI is an indicator of both systolic and diastolic myocardial function. It is based on the measurement of cardiac time intervals, and is defined as the sum of the isovolumetric contraction time and the isovolumetric relaxation time divided by the ejection time. Within the index, the isovolumetric contraction time mainly reflects systolic cardiac function and the isovolumetric relaxation time, diastolic function. The MPI was obtained to assess both the left and right ventricular function.
Maternal serum was obtained at the time of the ultrasound evaluation and 10 mL of venous umbilical cord blood was collected at the time of delivery. All specimens were collected in serum tubes, centrifuged, aliquotted, and stored at –80°C until time of analysis. Maternal serum and neonatal cord blood serum levels of IGF-I were measured by enzyme-linked immunosorbent assay as per manufacturers instruction (R&D Systems, Minneapolis, MN).
After birth, a comprehensive neonatal echocardiogram was performed between 24-48 hours of life on all neonates with antenatally diagnosed abnormalities and images were reviewed by a single pediatric cardiologist (L.Y.). Neonates with persistent abnormal findings received follow-up with pediatric cardiology until either resolution or 6 months of life.
Statistical analysis
Pilot data from a subset of our patients suggested a 20-ng/mL increase in maternal serum IGF-I levels among women with fetus with HCM. To find a difference of 20 ng/mL in maternal or neonatal serum IGF-I levels with an 80% power and a 95% confidence interval we estimated the need to recruit 20 cases of HCM and 20 controls. Given the reported rate of HCM in the diabetic population of 30-40%, we estimated we would need to recruit approximately 80 patients.
The statistical analysis consisted of t test and analysis of variance for comparison of means in normally distributed data. The Mann-Whitney U test was utilized for nonparametric data and χ 2 analysis and Fisher exact test, when appropriate, were used to determine the difference in proportion across groups. Data were stratified by maternal diabetic status to evaluate the association between fetal HCM and neonatal IGF-I levels. Pearson correlation coefficient was used to assess the relationship between serum umbilical cord IGF-I levels and fetal ventricular and septal wall thickness. A P value of < .05 was considered statistically significant. Software (SPSS, version 19; IBM Corp, Armonk, NY) was used for all statistical analyses.
Results
We recruited 74 women with diabetes and 20 patients without diabetes for the comparison group. Seventeen diabetic patients delivered prior to scheduled echocardiography with either spontaneous or indicated preterm delivery. Fetal echocardiogram could not be completed on 2 diabetic women due to poor acoustic windows and incomplete visualization of the cardiac anatomy. The remainder of the diabetic participants (n = 55) served as our study cohort, and the 20 women without diabetes as our comparison group. All of the diabetic women were comanaged by the maternal-fetal medicine group in conjunction with the endocrinology department. Women underwent diet and lifestyle counseling by a certified dietician at their first visit and their blood sugars were managed with insulin therapy. Within the study cohort, 1 intrauterine fetal demise diagnosed at 36 4/7 weeks occurred in a woman with a hemoglobin A1C of 6%. The fetal echocardiogram performed 4 days prior to the demise demonstrated an elevated left MPI of 0.62 without evidence of HCM. The autopsy confirmed normal septal and ventricular wall measurements. The only positive postmortem finding was partial sacral agenesis.
Maternal demographic characteristics for each group are displayed in Table 1 and obstetric outcomes in Table 2 . Women in our diabetic group tended to be older, initiated prenatal care earlier, had a higher weight and were more likely to be of Hispanic ethnicity compared to the group of women without diabetes ( Table 1 ). With regard to obstetric outcomes, women in the diabetic group were more likely to deliver via cesarean and to have neonates of a higher birthweight as compared to those women without diabetes ( Table 2 ).
| Characteristic | Diabetic (n = 55) | Nondiabetic (n = 20) | P value |
|---|---|---|---|
| Age, y | 33 ± 5 | 29 ± 7 | < .05 |
| Nulliparity | 16 (29) | 6 (30) | NS |
| Ethnicity | < .05 | ||
| Caucasian | 17 (31) | 1 (5) | |
| Hispanic | 28 (51) | 9 (45) | |
| African American | 7 (13) | 8 (40) | |
| Other | 3 (5) | 2 (10) | |
| Public aid | 39 (71) | 18 (90) | NS |
| Estimated gestational age at first visit, wk | 9.6 (8.0, 12.4) | 13.5 (9.6, 26.4) | < .001 |
| Prepregnancy body mass index, kg/m 2 | 32.2 ± 7.2 | 29.8 ± 4.4 | NS |
| Prepregnancy weight, kg | 82.8 ± 24 | 67.2 ± 27 | < .001 |
| Weight at 36 wk, kg | 94.7 ± 23 | 79.1 ± 11 | < .01 |
| A1C–first, % | 6.7 ± 1.2 | – | – |
| A1C–second, % | 5.8 ± 0.8 | – | – |
| A1C–third, % | 6.2 ± 0.9 | – | – |
| Outcomes | Diabetic (n = 55) | Nondiabetic (n = 20) | P value |
|---|---|---|---|
| Estimated gestational age at delivery, wk | 38.6 (37.5, 39.1) | 39.3 (38.6, 40.2) | < .01 |
| Route of delivery | |||
| Normal vaginal | 25 (45) | 16 (84) | < .01 |
| Operative | 2 (4) | 0 | NS |
| Cesarean | 28 (51) | 3 (16) | < .01 |
| Birthweight, g | 3655 (3255, 3909) | 3310 (3150, 3505) | < .05 |
| 5-min Apgar <7 | 3 (6) | 2 (10) | NS |
| Hypoglycemia | 11 (20) | 0 | – |
| Fetal demise | 1 (1.8) | 0 | – |
With regard to the fetal echocardiogram findings, in the diabetic group, 33 of 55 (60%) had abnormal fetal echocardiogram results. Cardiac dysfunction occurred in 21 of 55 (38.2%), HCM alone in 8 of 55 (14.5%), and HCM with concomitant dysfunction in 4 of 55 (7.3%). None of the patients in the comparison group had abnormal fetal echocardiograms. The ventricular and septal wall thicknesses were significantly greater in fetus in the diabetic group as compared to those in the group without diabetes ( Table 3 ). However, there were no significant differences in MPI measurements between the 2 groups. Comparison of those diabetic women with fetal HCM (n = 12) as compared to those diabetic women without fetal HCM (n = 43), demonstrated a higher hemoglobin A1C in the second trimester, although this difference did not reach statistical significance ( Table 4 ). The third-trimester hemoglobin A1C was significantly higher among diabetic women with fetal HCM and birthweight was also significantly higher in those neonates of diabetic mothers who were found to have fetal HCM as compared to those without fetal HCM ( Table 4 ). No significant differences were found among the other clinical variables compared as displayed in Table 4 . When diabetic women with fetal cardiac dysfunction (n = 25) were compared to those without (n = 32), no significant maternal or neonatal differences in clinical characteristics were identified ( Table 5 ).
| Echocardiogram findings | Diabetic (n = 55) | Nondiabetic (n = 20) | P value |
|---|---|---|---|
| Left ventricular wall, mm | 3.8 ± 1.3 | 3.0 ± 0.4 | < .001 |
| Right ventricular wall, mm | 3.9 ± 1.2 | 3.2 ± 0.5 | < .01 |
| Septal wall, mm | 3.8 ± 1.3 | 2.8 ± 0.5 | < .001 |
| Left MPI | 0.37 ± 0.2 | 0.34 ± 0.05 | NS |
| Right MPI | 0.30 ± 0.2 | 0.33 ± 0.1 | NS |
| Characteristics | Fetal HCM (n = 12) | No Fetal HCM (n = 43) | P value |
|---|---|---|---|
| Age, y | 31.3 ± 6.2 | 33.7 ± 4.8 | NS |
| Public aid | 10 (83) | 29 (67) | NS |
| Type 2 diabetes mellitus | 9 (75) | 33 (77) | NS |
| Estimated gestational age at first visit, wk | 9.7 (8.1, 15.8) | 9.6 (8.0, 11.6) | NS |
| Prepregnancy body mass index, kg/m 2 | 32.3 ± 7.5 | 32.2 ± 7.0 | NS |
| Prepregnancy weight, kg | 86 ± 21 | 81 ± 22 | NS |
| Weight at 36 wk, kg | 101 ± 25 | 92.9 ± 22 | NS |
| A1C–first, % | 7.4 ± 1.4 | 6.6 ± 1.1 | NS |
| A1C–second, % | 6.6 ± 1.2 | 5.7 ± 0.6 | .064 |
| A1C–third, % | 7.0 ± 1.2 | 5.9 ± 0.7 | < .05 |
| Estimated gestational age at delivery, wk | 38.1 (37.1, 39.1) | 39.0 (38.3, 39.2) | NS |
| Birthweight, g | 4033 (3696, 5150) | 3542 (3127, 3798) | < .001 |
| 5-min Apgar <7 | 0 (0) | 3 (7) | NS |
| Hypoglycemia | 4 (33) | 7 (16) | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
