Investigation into DNAm and brain structural and connectomic covariance: a life course approach

Abstract

Early life environmental stress, indexed by perinatal factors such as birth weight and gestational age, is associated with differences in brain structure and connectivity in early life, as well as being associated with a range of neurodevelopmental, psychiatric and cognitive outcomes, from childhood and into adulthood. While studies have investigated how these early life factor influence brain structure in early life, few have investigated this in older age. The molecular underpinnings of these relationships are not well understood. DNA methylation is an epigenetic mechanism that regulates gene expression; it is developmentally dynamic and responsive to environmental factors, making it a promising candidate for providing mechanistic insight into how early life stressors exert their effects. The aims of this thesis are as follows: to better characterise the associations between birth weight and brain structure and connectivity in later-life; to evaluate the evidence that DNAm is implicated in brain structure and function; to investigate the impact of gestational age at birth on the neonatal methylome and its association with brain connectivity. In the first study, I investigated the associations between variation in normal birth weight and measures of brain structure and connectivity in participants aged 73 years from the Lothian Birth Cohort 1936. Larger birth weight was associated with larger brain volume, and with regional cortical surface area, but not with white matter microstructure. This relationship between birth weight and brain size did not appear to be related to the degree of atrophy that had taken place. Early life growth is likely to be associated with brain tissue reserve, in evidence in later life. In the second study, I conducted a systematic review to evaluate the evidence linking DNA methylation to brain structure and function across the life-course. Sixty studies, encompassing both health and disease contexts, were identified. Together, these studies indicated that differential DNAm is associated with brain structure and function for 8 categories of disease across the life course, although uncertainties remain. Modest consistency between DNAm and neuroimaging features precluded the possibility of quantitative synthesis. I identified potential sources of bias in existing literature, enabling the development of guidelines that could reduce methodological heterogeneity in imaging-DNAm studies. Finally, I identified a DNAm signature of gestational age in neonatal saliva samples and tested its association with brain white matter microstructure. Participants were neonates, born preterm or term, recruited to the Theirworld Edinburgh Birth Cohort. There was widespread differential methylation associated with gestational age at birth, at term equivalent age. Several genes were identified that have previously been implicated in association with gestational age in cord blood, and with disorders known to contribute to the aetiology of preterm birth, such as pre-eclampsia. An epigenome-wide variable of the DNAm signature was associated with white matter microstructure, suggesting that DNAm contributes to white matter dysconnectivity in the neonatal period. This thesis provides evidence that early life exerts an effect on brain structure into later life, that DNAm and MRI neuroimaging are associated across the life-course and in a range of health and disease contexts, and that DNAm is profoundly altered in association with variation in gestational age and that this may contribute to white matter connectivity in the neonatal period.