Using Mendelian Randomization to investigate the effect of maternal intra-uterine exposures on fetal growth

Abstract

The aim of my thesis is to identify modifiable maternal exposures that causally affect offspring birth weight during pregnancy. Birth weight is a major correlate of both neonatal mortality and morbidity as well as adult health outcomes. Most exposures identified by observational studies to be correlated with birth weight are shown not to be causally related to birth weight in randomised controlled trials (RCTs). An alternative method to identify maternal exposures that causally affect offspring birth weight is Mendelian Randomisation, where genetic variants shown to be associated with an exposure are used as instruments to measure the effect on the outcome. I therefore used Mendelian Randomisation to investigate the effects of proposed maternal intrauterine exposures on offspring birth weight. Chapter 1 introduces details on the underlying biology and epidemiology of birth weight, how variation in birth weight is associated with adult health outcomes, and maternal exposures proposed to affect birth weight during pregnancy. It will also discuss the methodology behind Mendelian Randomisation, and the latest research using Mendelian Randomisation to investigate potential causal exposures effect on birth weight. Chapter 2 uses Mendelian Randomisation to investigate the effect of maternal circulating gestational 25(OH)D and calcium on offspring birth weight. The Mendelian Randomisation analysis is triangulated with instrumental variable analysis applied to RCTs of vitamin D and calcium supplementation during pregnancy. Chapter 3 uses Mendelian Randomisation to investigate the effect of maternal metabolically favourable adiposity and BMI on offspring birth weight. Recently discovered genetic variants are associated with higher body fat percentage yet a lower risk of cardiovascular disease and type II diabetes, a so called metabolically “favourable” adiposity phenotype. Chapter 4 investigates the effect of fetal genetic predisposition to higher adult metabolically favourable adiposity on own birth weight. This is the only chapter in the thesis not directly using Mendelian Randomisation, but is inspired by novel findings found whilst performing the analyses for Chapter 3. Chapter 5 uses Mendelian Randomisation to investigate the effect of maternal morning plasma cortisol on offspring birth weight. This chapter is unlike the other chapters in that it performs Mendelian Randomisation using a single independent genetic locus, whilst the other chapters use several loci. Finally, in Chapter 6 I demonstrate how this research has increased knowledge regarding maternal intra-uterine exposures on offspring birth weight, and discuss future avenues of research.