Aims

To employ techniques for identifying latent structure in genome-wide DNA methylation data, characterised by high dimensionality, in order to increase statistical power to detect effects due to prenatal alcohol exposure (both maternal and paternal).

Objectives:

1. To identify latent structures in genome-wide cord-blood DNA methylation

2. To correlate these with maternal and paternal alcohol use before and during pregnancy (in search for both maternal-intrauterine and paternal-line effects)

3. To investigate to what extent confounding could explain these associations

4. To investigate the persistence of the latent structures into childhood and adolescence

5. To investigate whether the relationship between exposures and latent structure persist in later childhood and adolescence

Hypothesis

DNA methylation is thought to be one possible mediator of the feto-toxic effects of alcohol use in pregnancy, given the remarkable correspondence between the most alcohol-sensitive gestational periods and the occurrence of major epigenetic events (including erasure of methylation marks around gastrulation). This is further supported by various animal experiments. Further evidence also points at the role of paternal alcohol use around conception on offspring DNA methylation marks, in particular affecting (paternally) imprinted genes.