Biological embedding of childhood adversity - a multi-omics perspective on stress regulation

Abstract The experience of adversity in childhood can have life-long consequences on health outcomes. In search of mediators of this relationship, alterations of bio-behavioral and cellular regulatory systems came into focus, including those dealing with basic gene regulatory processes. Systems biology oriented approaches have been proposed to gain a more comprehensive understanding of the complex multiple interrelations between and within layers of analysis. We used co-expression based, supervised and unsupervised single and multi-omics system approaches to investigate the influence of childhood adversity on gene expression, protein expression and DNA methylation in CD14+ monocytes of healthy adults before and after exposure to an experimental psychosocial stress protocol. Childhood adversity explained some variance at the single analyte level and within gene and protein co-expression structures. A single-omic, post stress gene expression model differentiated best between participants with a history of childhood adversity and controls in supervised analyses. In unsupervised analyses, a multi-omics based model showed best performance but separated participants based on sex only. Multi-omics analyses are a promising concept but might yield different results based on the specific approach taken and the omic-datasets supplied. Here, stress associated gene-expression pattern were most strongly associated with childhood adversity, and integrating multiple cellular layers did not results in better discriminatory performance. Currently, the capacity and yield of different omics-profiling methods might limit the full potential of integrative approaches.