Recent studies have established that the neonate born to a pregnancy with maternal diabetes or obesity (‘diabesity’) is characterized by increased fat accumulation. The neonatal fat is the result of triglyceride synthesis and deposition stimulated by elevated fetal insulin levels combined with insulin’s mitogenic activity directly stimulating the growth of the white adipocytes. Fetal insulin levels are determined by fetal glucose and some amino acids such as arginine. Although the placenta plays a key role in providing maternally derived nutrients to the growing fetus, there is currently no evidence that it actively contributes to an excessive maternal-to-fetal glucose flux at the end of gestation. Early in gestation, the maternal environment in diabesity, and in particular the glucose–insulin axis, can modify placental growth and development, which may contribute to an enhanced glucose flux to the fetus already early in pregnancy. This may have long-lasting effects on the fetal pancreas and accelerate beta-cell maturation.
The association of fetal and neonatal insulin levels and the proportion of body fat with obesity later in the offspring’s life calls for interventions during pregnancy to prevent or reduce fetal hyperinsulinaemia. Dietary and/or physical activity interventions initiated before or in early pregnancy would likely be most effective. Results from the very few studies with fetal insulin as the outcome are inconsistent. However, there is a major lack of randomized intervention trials on this topic.
The rapid increase in the prevalence of diabesity, that is, obesity, type 2 diabetes and associated complications, is a major problem for global health worldwide. The developmental origins of health and adult disease (DOHaD) concept puts pregnancy and its influences on the mother and the developing fetus in the focus of the transgenerational transmission of diabesity risk. Thus, the maternal–placental–fetal interaction has received increasing research interest.
The placenta is interposed between the maternal and fetal blood stream and thus constitutes the physical link between the two generations. The maternal environment associated with diabetes (type 1, type 2 and gestational) mellitus and/or obesity has an influence on placental development, structure and function, which has been summarized in several articles .
These placental alterations may be protective or adaptive responses to the maternal environment, with the ultimate purpose of allowing the fetus to grow and develop in a stable environment. Alternatively, these changes might be mechanistically linked to the fetal phenotype associated with maternal diabesity. It is not the purpose of this review to summarize once again placental changes associated with diabesity. Rather, we will focus on one particular concept, with the focus on the maternal and fetal glucose/insulin axis and its effects on the placenta and the fetus. This concept may explain the fetal phenotype and some aspects of intrauterine programming of childhood obesity in the wake of maternal diabesity. We further intend to discuss options to prevent this from happening.
The fetal phenotype in diabesity
Much work has been published demonstrating the increased incidence of ‘macrosomia’ and large-for-gestational age neonates, that is, >90th birth-weight centile, born to diabesity pregnancies. This research focus is largely based on the obstetric complications associated with a macrosomic neonate such as an increased rate of caesarean sections, shoulder dystocia and others. However, there is a growing body of evidence showing that the neonate of a diabetic mother, even when born in the appropriate-for-gestational age range (between the 5th or 10th and 90th birth-weight centile), has a phenotype different from those born to normal pregnancies. The major difference and thus the major effect of maternal diabesity on the neonate is on body composition: These neonates are born with a higher proportion of body fat independent of their birth weight or birth-weight category , whereas the fat-free mass appears unaltered ( Fig. 1 ). Interestingly, gestational diabetes and maternal obesity are independent risk factors for neonatal percentage body fat and contribute additively .
The higher neonatal body fat is of key importance because the number of adipocytes for a human being seems to be determined very early in the life cycle if not already in utero . The trajectory of adipocyte number cannot be changed not even when weight, that is, fat, is lost. Thus, once the growing fetus has developed more adipocytes in the wake of the diabesity environment, the neonate may be strongly predisposed to remain on this trajectory and have more adipocytes also in childhood. This is indeed the case as shown by the association of neonatal body fat with that in childhood at the age of 9 years, whereas total body weight is not associated .
The excessive accretion of body fat even in neonates with appropriate-for-gestational-age birth weight constitutes a major problem from the public health perspective. These neonates represent by far the majority of all neonates born to diabesity pregnancies. They are inconspicuous and yet are at a risk of childhood obesity. Therefore, measures meant to prevent the childhood obesity risk should perhaps include all neonates of diabetic mothers, regardless of birth weight.
The role of the placenta in determining the fetal phenotype in diabesity
More than 50 years ago, Jorgen Pedersen explained fetal overgrowth in diabetic pregnancies by his hyperglycaemia–hyperinsulinaemia concept: Maternal hyperglycaemia as in diabetes leads to fetal hyperglycaemia, which stimulates the fetal pancreas to produce and secrete more insulin ultimately resulting in fetal hyperinsulinaemia. This concept has been expanded later to include amino acids and fatty acids (fuel-mediated teratogenesis concept) . In essence, fetal insulin is the main driver for the fetal phenotype, because of its dual activities as a mitogen to stimulate the growth of white adipocytes and as an anabolic hormone to drive triglyceride formation and deposition in fetal white adipocytes. Until now, the concepts have not been refuted and are still the key to our understanding of fetal development in a situation of maternal hyperglycaemia as is associated with diabesity.
Although amino acids, and in particular arginine, can serve as insulin secretagogues and may contribute to fetal hyperinsulinaemia in diabesity, glucose is the most important stimulator of insulin secretion. Thus, this has sparked research interest trying to understand the regulation of maternal-to-fetal glucose transport across the placenta. Despite a number of molecular changes in the placenta at the level of glucose transporters and glucose metabolism, the total glucose flux across the placenta is unaltered in placentas from gestational diabetic pregnancies . Thus, at least at the end of gestation, maternal-to-fetal glucose flux is mostly dictated by the maternal-to-fetal concentration gradient of glucose, perhaps with some contribution of utero-placental and feto-placental blood flow . This has important implications for the clinical care of the obese or diabetic woman: Adequate glycaemic control of the mother after diagnosis of gestational diabetes mellitus (GDM) may not be sufficient to prevent excess fetal fat accretion once the fetal pancreas has already been overstimulated by maternal hyperglycaemia prior to gestational diabetes diagnosis. The level of fetal hyperinsulinaemia has to be taken into account in the clinical decisions of how to manage diabetes in the pregnant women. Directly measuring amniotic fluid insulin levels has been advocated , but not been accepted because of the invasive nature of amniocentesis.
Elevated levels of fetal insulin will have multiple consequences. It can stimulate (1) placental growth and also growth of the fetal heart, at least in non-human primates , (2) glycogen deposition in the placental endothelium (Desoye, Hiden unpublished) and (3) aerobic fetal metabolism. The latter will raise the demand for oxygen in the fetus. At the same time, in diabetic pregnancies, maternal haemoglobin is glycosylated to a greater extent and, thus, has a lower oxygen transport capacity. Along with potentially reduced utero-placental blood flow in maternal diabetes mellitus, this may result in an imbalance between oxygen supply and demand with a net oxygen deficit. This stimulates fetal erythropoietin synthesis and erythropoiesis . In synergy with increased oxygen transport capacity within the fetal circulation, the placental vascular network is expanded and the capillary surface enlarged . Fetal insulin contributes to placental angiogenesis through multiple cellular mechanisms .
One further consequence of fetal hyperinsulinaemia is not only the stimulation of triglyceride synthesis and fat deposition but also aerobic metabolism and stimulation of glucose uptake into fetal tissues. This insulin-induced shift of circulating glucose into tissue will remove glucose from the circulation and thus transiently steepen the concentration gradient between the maternal and fetal circulations, thus increasing the glucose flux until steady-state concentrations have been reached again. Thus, fetal insulin will siphon maternal glucose and direct it to the fetus. The term ‘glucose steal phenomenon’ was coined for this effect, which was shown in rat and sheep and indirectly demonstrated also in humans .
The important consequence for clinical care of the diabetic mother is that late in gestation normalization of maternal glucose levels may not be enough to prevent fat accretion (or ‘macrosomia’), because the fetus is also involved in determining transplacental glucose flux. This will always be the case when the fetal pancreas is already altered by hyperplasia earlier in pregnancy and fetal hyperinsulinaemia has become manifest. This raises the question about the onset of fetal hyperplasia and, thus, hyperinsulinaemia.
Insulin depots in the human fetal pancreas were demonstrated as early as at week 8 or 9 of gestation ; insulin is secreted into the fetal circulation as early as at 11 weeks of gestation and can be measured in the fetal amniotic fluid at week 14 of gestation . Importantly, amniotic fluid insulin in this early period was associated with the risk of birth weight >90th centile. Thus, early fetal hyperinsulinaemia as a result of maternal and fetal hyperglycaemia can have consequences for further fetal development. The importance of the glucose–insulin axis in the first trimester is also highlighted by the correlation of maternal fasting glucose levels at weeks 9–10 of gestation with the risk of fetal macrosomia .
The role of the placenta in determining the fetal phenotype in diabesity
More than 50 years ago, Jorgen Pedersen explained fetal overgrowth in diabetic pregnancies by his hyperglycaemia–hyperinsulinaemia concept: Maternal hyperglycaemia as in diabetes leads to fetal hyperglycaemia, which stimulates the fetal pancreas to produce and secrete more insulin ultimately resulting in fetal hyperinsulinaemia. This concept has been expanded later to include amino acids and fatty acids (fuel-mediated teratogenesis concept) . In essence, fetal insulin is the main driver for the fetal phenotype, because of its dual activities as a mitogen to stimulate the growth of white adipocytes and as an anabolic hormone to drive triglyceride formation and deposition in fetal white adipocytes. Until now, the concepts have not been refuted and are still the key to our understanding of fetal development in a situation of maternal hyperglycaemia as is associated with diabesity.
Although amino acids, and in particular arginine, can serve as insulin secretagogues and may contribute to fetal hyperinsulinaemia in diabesity, glucose is the most important stimulator of insulin secretion. Thus, this has sparked research interest trying to understand the regulation of maternal-to-fetal glucose transport across the placenta. Despite a number of molecular changes in the placenta at the level of glucose transporters and glucose metabolism, the total glucose flux across the placenta is unaltered in placentas from gestational diabetic pregnancies . Thus, at least at the end of gestation, maternal-to-fetal glucose flux is mostly dictated by the maternal-to-fetal concentration gradient of glucose, perhaps with some contribution of utero-placental and feto-placental blood flow . This has important implications for the clinical care of the obese or diabetic woman: Adequate glycaemic control of the mother after diagnosis of gestational diabetes mellitus (GDM) may not be sufficient to prevent excess fetal fat accretion once the fetal pancreas has already been overstimulated by maternal hyperglycaemia prior to gestational diabetes diagnosis. The level of fetal hyperinsulinaemia has to be taken into account in the clinical decisions of how to manage diabetes in the pregnant women. Directly measuring amniotic fluid insulin levels has been advocated , but not been accepted because of the invasive nature of amniocentesis.
Elevated levels of fetal insulin will have multiple consequences. It can stimulate (1) placental growth and also growth of the fetal heart, at least in non-human primates , (2) glycogen deposition in the placental endothelium (Desoye, Hiden unpublished) and (3) aerobic fetal metabolism. The latter will raise the demand for oxygen in the fetus. At the same time, in diabetic pregnancies, maternal haemoglobin is glycosylated to a greater extent and, thus, has a lower oxygen transport capacity. Along with potentially reduced utero-placental blood flow in maternal diabetes mellitus, this may result in an imbalance between oxygen supply and demand with a net oxygen deficit. This stimulates fetal erythropoietin synthesis and erythropoiesis . In synergy with increased oxygen transport capacity within the fetal circulation, the placental vascular network is expanded and the capillary surface enlarged . Fetal insulin contributes to placental angiogenesis through multiple cellular mechanisms .
One further consequence of fetal hyperinsulinaemia is not only the stimulation of triglyceride synthesis and fat deposition but also aerobic metabolism and stimulation of glucose uptake into fetal tissues. This insulin-induced shift of circulating glucose into tissue will remove glucose from the circulation and thus transiently steepen the concentration gradient between the maternal and fetal circulations, thus increasing the glucose flux until steady-state concentrations have been reached again. Thus, fetal insulin will siphon maternal glucose and direct it to the fetus. The term ‘glucose steal phenomenon’ was coined for this effect, which was shown in rat and sheep and indirectly demonstrated also in humans .
The important consequence for clinical care of the diabetic mother is that late in gestation normalization of maternal glucose levels may not be enough to prevent fat accretion (or ‘macrosomia’), because the fetus is also involved in determining transplacental glucose flux. This will always be the case when the fetal pancreas is already altered by hyperplasia earlier in pregnancy and fetal hyperinsulinaemia has become manifest. This raises the question about the onset of fetal hyperplasia and, thus, hyperinsulinaemia.
Insulin depots in the human fetal pancreas were demonstrated as early as at week 8 or 9 of gestation ; insulin is secreted into the fetal circulation as early as at 11 weeks of gestation and can be measured in the fetal amniotic fluid at week 14 of gestation . Importantly, amniotic fluid insulin in this early period was associated with the risk of birth weight >90th centile. Thus, early fetal hyperinsulinaemia as a result of maternal and fetal hyperglycaemia can have consequences for further fetal development. The importance of the glucose–insulin axis in the first trimester is also highlighted by the correlation of maternal fasting glucose levels at weeks 9–10 of gestation with the risk of fetal macrosomia .
The maternal glucose–insulin axis in the first trimester of pregnancy and its effect on the placenta
There is a growing body of evidence that the placenta early in gestation may influence fetal growth and development with manifestations at the end of pregnancy. The placental volume at week 14 of gestation and the rate of placental growth between weeks 14 and 17 of gestation were associated directly with fetal anthropometric parameters such as abdominal circumference . In another study, the placental volume at 19 weeks was associated with neonatal fat mass both absolute and relative to birth weight .
Despite the importance of knowing the regulatory factors for placental growth early in gestation, surprisingly little is known about this, if anything. It is also unclear what the effect of pre-gestational diabetes on the early placenta is. There are conflicting reports on the degree of placental vascularization, that is, more or less, as compared to non-diabetic pregnancies . There is indirect evidence to suggest a delayed placental development in pre-gestational, that is, type I, diabetes: In these pregnancies, fetal growth is delayed . Although placental volume measurements at this early stage in type I diabetic pregnancies are pending, the lower levels of circulating human placental lactogen (hPL) and pregnancy-associated plasma protein-A (PAPP-A) , which are both produced in the placental trophoblast, suggest also a reduced trophoblast, that is, placental, growth. In vitro evidence demonstrates an interaction of hyperglycaemia and higher/lower levels of oxygen levels resulting in a slower trophoblast growth .
Maternal insulin may also have an independent effect on early placental development. It upregulates matrix metalloproteinase 14, one key enzyme, which is involved in growth regulation, invasion, inflammation and also in angiogenesis . The placental levels of this enzyme correlate with the daily insulin dose given to type 1 diabetes mellitus (T1DM) women in the first trimester . Thus, the combined effects of hyperglycaemia and compensatory hyperinsulinaemia may alter the dynamics of placental growth already in the first trimester and have a long-lasting influence on fetal growth and development including body composition.
The mother–placenta–fetus dialogue in diabesity
Based on the above, we propose the following model linking first-trimester events to the childhood diabesity risk ( Fig. 2 ).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
