Professor Chadeau-Hyam holds a chair in Computational Epidemiology and Biostatistics at the School of Public Health, Imperial College, and an honorary position at Utrecht University. He obtained a PhD from Paris University, and while his background was in disease modelling, he has gained expertise over the last 14 years in devising, applying and adapting computationally efficient methods to analyse and integrate multiple OMICs data. He has been involved in numerous large-scale projects as lead statistician, responsible for the analysis and integration of OMICs data. As an active member of the European exposome initiative, he has expertise in the development and application of statistical methods to screen and integrate OMICs profiles and explore the stressor- triggered molecular mechanisms involved in the determination of individual risk profiles. Alongside this applied work, he has focused on the development of dynamic disease models including OMIC variable selection methods to study the dynamics of cancer natural history. Professor Chadeau-Hyam is the director of the MSc in Health Data Analytics at ICL, and leads a series of two Exposome short courses. His team currently includes more than 15 members with a unique diversity of profiles, including medical, biological, and statistical backgrounds, all collaborating for the in-depth analysis of large-scale data from mega-sized and/or deeply phenotypes cohort studies.
Turning open science into practice: causality as a showcaseSee more
23/06 - 14:00Visit the agenda
Title of talk
Exposome Analytics: composite scores, embodiment, and health risk - Evidence for the UK Biobank Study
23/06 - 14:55
Abstract of talk
The concept of embodiment postulates that the human environment, through its physical, chemical and psychosocial stresses, solicits several adaptative processes and leaves a sustainable biological mark. There is limited evidence on the extent to which social factors, measures of its embodiment, socially patterned exposures and behaviours may contribute to these associations, and in particular in relation to incident chronic diseases.
We first used a composite score to measure the physiological wear-and-tear of three key systems (inflammatory, metabolic and cardiovascular) and the functioning of two organs (kidney and liver) based on the measurement of 13 biomarkers in 366 748 UK-Biobank participants who were free of cancer and cardiovascular disease (CVD) when they enrolled. Descriptive analyses showed a strong education gradient in the score that was not fully explained by later-in-life socially patterned behaviours and exposures. We found that the BHS was associated with increased all-cause, cancer and CVD mortality, and cancer and CVD incidence. Despite strong gradients in the BHS across education groups, these associations were only mildly attenuated upon adjustment for education, though larger attenuations were observed while adjusting for other factors, in particular BMI. Our results suggest that composite scores such as ours may act as markers of biological ageing which capture features of social embodiment, as well as biological effects of more proximal behaviours and health risk factors. Our results suggest that biological age complements health behaviours to improve predictions of CVD incidence. Complementary work focused on quantifying the relative contribution of the different domains of the exposome and its internal signature to the prediction of CVD incidence in the same population. Ongoing research aims to identify the possible mediating roles these biomarkers may play in the association between biological ageing markers, behaviours, social determinants and CVD outcomes. Overall, our work highlights the importance of having publicly-available resources to identify the sparse and complementary set of factors involved in the development of chronic conditions and to investigate their interplay.