The University of Edinburgh
Effects of in-utero exposure to metformin on child development and the underlying epigenetic mechanism
Abstract
dc:description.abstractObese pregnant women are at increased risk of developing pregnancy complications and having adverse pregnancy outcomes. The offspring born from obese pregnant women are also at increased risk of becoming obese themselves at a young age and developing cardiovascular disease at their adulthood. Metformin is a drug used as a treatment for diabetes in pregnancy. Studies on metformin treatment during obese pregnancy report that metformin has no effect on birth weight, but lowers gestational weight gain and fasting glucose levels. There are limited follow-up data of consequences for offspring exposed to metformin in-utero during obese pregnancy. In this thesis, the children born from women who participated in the EMPOWaR trial (obese (BMI≥ 30 kg/m²) pregnant women without gestational diabetes randomised to metformin or placebo during pregnancy) were recruited at age 4 to 7 years old (n= 40; metformin= 19, placebo= 20). The children’s body composition, adiposity and cardiovascular function were measured. The children’s cognitive, diet behaviour and activity levels (children themselves and family activity levels) were assessed by using questionnaires. The protocol of using MRI to measure the cardiac function of young children (aged 4 to 8 years old) was also examined by recruiting healthy volunteers (n= 5). No differences were found between metformin and placebo groups in the children’s body composition or adiposity measurements. For the cardiovascular measurements, no differences were found between metformin and placebo groups, and metformin was not associated with cardiovascular outcomes after adjustment for confounding factors. Regarding the questionnaire data, no differences were found in children’s cognitive, eating behaviour, own activity levels, nor family activity levels according to metformin exposure. Using MRI to measure the cardiac function of young children was shown to be workable in the study with good data acquisition. Metformin has been reported to delay the ageing of the tissue, and there is a lack of data about the effects of metformin on placental ageing. In this thesis, placenta samples (n= 103) collected from the EMPOWaR trial were analysed using machine learning models (placental clocks) to calculate the predicted placental DNA methylation age (DNAm gestational age) and placental ageing. The predicted DNAm gestational age was significantly associated with the actual gestational age (coefficient= 0.63, p= 3.69 x 10-13). The estimated DNAm gestational age was significantly lower than the actual GA by 1.40 weeks (38.31(1.34) weeks vs 39.66(1.39) weeks, p= 2.28 x 10-11). Predicted GA was significantly higher in boys than girls (38.59 vs 38.03 weeks, p= 0.03). No effects of metformin or smoking were found on placental ageing. The mechanisms of how metformin in-utero exposure affects fetal growth are still unclear. Epigenetic modifications, such as DNA methylation, have been proposed as one of the potential mechanisms linking to the growth of the offspring. In this thesis, the placental methylation array data from the EMPOWaR trial were used to investigate the effect of metformin on DNA methylation in placentas collected from obese pregnant women exposed to metformin or placebo. Methylation array findings were replicated by pyrosequencing. The body size data (weight, length and ponderal index) of their offspring were used to examine the associations between methylation alterations with body size changes at birth and at age 3 months. No associations were found between metformin treatment and placental DNA methylation. DNA methylation of short chain acyl-CoA dehydrogenase (ACADS) was positively associated with ponderal index at birth (beta = 0.36, p= 0.004) and weight at age 3 months (beta= 0.24, p= 0.03). The gene expression of ACADS was negatively associated with ponderal index at age 3 months (beta= -0.29, p= 0.01). In conclusion, in our cohort, metformin had no effect on body growth, adiposity, and cardiovascular risk of children exposed to metformin in-utero during obese pregnancy. Metformin also had no effect on the methylation changes in the obese placenta or on placental ageing. The gene ACADS could potentially be used as a biomarker for child adiposity changes in postnatal growth.
Degree
thesis:*- Grantor dc:publisher
- The University of Edinburgh
- Year dc:date.issued
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Yang, Liu
- Advisors dc:contributor.advisor
-
- Reynolds, Rebecca
- Drake, Mandy
Subjects
dc:subject × 4Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- http://dx.doi.org/10.7488/era/1821