Pregnancy outcomes in type 1 diabetes remain suboptimal despite improvements in glycaemic control in recent years. For instance, the Continuous Glucose Monitoring in Women with Type 1 Diabetes in Pregnancy Trial (CONCEPTT) found that 60% of infants were still large for gestational age (LGA). This suggests that processes other than glucose metabolism may be involved in this complication because these women had improved antenatal glucose control. It is likely that LGA also involves altered carbohydrate, protein and fat metabolism, and it is influenced by factors including maternal obesity, gestational weight gain, dietary quality and maternal lipids, as well as glycaemia.
Neonatal hypoglycaemia affects 25% of offspring from type 1 pregnancies and is associated with foetal hyperinsulinism. It has been attributed to suboptimal maternal glycaemia during late pregnancy and birth, although this may also occur earlier in pregnancy. Meanwhile, pregnant women with type 1 diabetes have a five times higher risk of pre-eclampsia than healthy women and this can be predicted with first trimester biomarkers, including leptin and glucose, and protein-related biomarkers at 28 weeks. However, the relative contributions of carbohydrate, lipid and protein metabolism to the development of pre-eclampsia in type 1 pregnancies are unclear.
Claire Meek, Zoe Stewart and other members of the CONCEPTT team have used samples from the trial to examine the pathophysiology associated with these complications in more detail. They looked at metabolomic and lipidomic changes associated with modifiable risk factors for suboptimal pregnancy outcomes. These risk factors were hyperglycaemia, maternal BMI, gestational weight gain and habitual diet, while the adverse pregnancy outcomes were LGA, offspring adiposity, neonatal hypoglycaemia, offspring hyperinsulinism and pre-eclampsia.
CONCEPTT involved 225 women with type 1 diabetes recruited in early pregnancy or when planning a pregnancy, 174 of whom gave additional serum samples at 12, 24 and 36 weeks to create the biorepository on which this study is based.
Metabolomic and lipidomic changes
Maternal hyperglycaemia
Maternal time above range was associated with increased lipids in maternal blood suggestive of de novo lipogenesis rather than other sources of fatty acids. Given that de novo lipogenesis is a physiological pathway associated with using up excess glucose, it is highly likely to be active in diabetes in pregnancy.
Cord blood was enriched in carnitines, which are cofactors in β- oxidation of fatty acids – a process of generating ATP in conditions of low glucose availability. The authors consider this a surprising finding worthy of further investigation because β- oxidation generally occurs after, not before, birth. However, they note that few studies have looked at foetal fuel utilisation in pregnancies affected by diabetes.
The findings also demonstrate a link between maternal hyperglycaemia and adiposity in the offspring, associated with the presence of carnitines, which may indicate the foetus dissipating excess fuel received from the maternal circulation.
Offspring size and adiposity
During the first trimester, there were no metabolomic changes in maternal blood associated with LGA, but by 24 weeks, a pattern emerged that features multiple metabolites likely to promote or sustain excessive foetal growth. There was evidence of increased steroid hormone production and increased activity of the tricarboxylic acid cycle and β- oxidation pathway, all of which suggests abundant energy production. But neonatal adiposity showed dramatically different metabolic profiles from LGA, suggesting that increased size overall (LGA) and adiposity are not the same thing in the context of type 1 diabetes pregnancy. Further exploration of the underlying mechanisms is warranted.
Neonatal hypoglycaemia and hyperinsulinism at birth
Both neonatal hypoglycaemia and hyperinsulinism at birth – marked by the presence of cord C-peptide – were associated with metabolite changes during the period of foetal pancreatic development. Early pregnancy maternal triacyl glycerol abundance was linked with neonatal hypoglycaemia, while both lipids and phenols are linked to cord C-peptide.
Previous research has suggested that lipid storage is likely to be prioritised over lipid mobilisation in early pregnancy. The authors believe that the increased abundance of lipids in early pregnancy could be the result of insufficient dietary intake (hyperemesis gravidarum), insufficient insulin dosing or in response to certain dietary patterns, such as a low-carbohydrate diet, which would promote lipogenesis, or the direct impact of a high-fat diet. Meanwhile, saccharin and structurally related phenols were consistently related to cord C-peptide. These findings suggest that maternal diet in early pregnancy is a key determinant of offspring metabolic health at birth. They also raise the possibility that offspring beta cell function – amount of C-peptide – could be modulated by maternal diet in pregnancy.
Pre-eclampsia
Pre-eclampsia was associated with increased abundance of phosphatidylethanolamines (PEs), which are likely to have come from vascular cell membranes during the first two trimesters. This suggests that the endothelial damage that leads to pre-eclampsia may occur from early in pregnancy.
Meanwhile, the pattern of multiple metabolic changes in pre-eclampsia suggests high energy requirements throughout pregnancy, with utilisation of lipid energy sources during the first trimester and protein during the second and third trimesters. Catabolism of protein provides branched-chain amino acids and aromatic amino acids, which can act as substrates for gluconeogenesis. The involvement of branched-chain fatty acids is particularly marked in this metabolomic analysis, although it is not clear why they would be involved in gluconeogenesis. Their presence has been associated with insulin resistance in other studies. It is possible that placental insufficiency induces insulin resistance to improve the glucose supply to the foetus.
There were relatively few women with pre-eclampsia in this study, but previous work with women without diabetes has shown some similarities in metabolic changes. However, phosphatidylserines were most predictive of pre-eclampsia in women without diabetes, while PEs were more prominent in this study.
Clinical implications
Type 1 diabetes pregnancy is associated with altered carbohydrate, lipid and protein metabolism from the first trimester and this strongly predicts perinatal complications. Inadequate insulin, as a master regulator of carbohydrate, protein and lipid metabolism, is a likely unifying cause, according to the authors.
This study demonstrates, for the first time, a possible metabolic link between maternal hyperglycaemia and offspring adiposity mediated partially though altered maternal and offspring lipid metabolism. Increases in the abundance of many classes of lipids were common in all analyses. These were likely to be caused by insufficient insulin, which is the key pathophysiological feature of type 1 diabetes.
These findings suggest that, despite improved access to continuous glucose monitoring and insulin pumps, current efforts to address hyperglycaemia are not fully correcting the underlying abnormalities leading to complications.
We know that maternal diet influences insulin dose requirements, but the study also found an association between cord C-peptide and various metabolic changes in the first trimester that may be linked with maternal dietary intake of sugar substitutes, high-glycaemic-index carbohydrates and phenol compounds. This suggests that maternal diet might be a key modifiable determinant of offspring health, independent of maternal glycaemia. Future research on optimising insulin dosing from the first trimester, with better matching of insulin to dietary intake and insulin resistance, would improve maternal and neonatal health.
Traditionally, approaches to identifying and treating women at risk of LGA, neonatal hypoglycaemia and pre-eclampsia have focused on mid-to-late pregnancy. This new study suggests that early pregnancy is also important in determining outcomes. New strategies should therefore be developed to optimise access to care in the first trimester.
To read this paper, go to: Meek CL, Stewart ZA, Feig DS, Furst S, Neoh SL, Koulman A, Murphy HR on behalf of the CONCEPTT collaborative group. Diabetologia online 24 August 2023. http://doi.org/10.1007/s00125-023-05989-2
To learn more, enrol on the EASD e-Learning course ‘Pregestational diabetes in pregnancy’:
Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.