Data analysis and integration
Developing and applying novel analytical approaches to identify lifestyle and psychosocial determinants of healthy growth and ageing.
This theme will develop novel approaches to integrate the wealth of information made available to the Mohn Centre to identify determinants of healthy growth and development in children and young people. Using our holistic approach, we will employ trusted techniques we have developed to identify if and which exposures from the environment, psychosocial and lifestyle domains are affecting the growth and development of children.
We are continuously developing novel approaches and methodologies to data analysis including clustering techniques, as seen in the EXPANSE project below. We define childhood ‘expotypes’ (a set of environmental exposures during childhood) by sets of exposures for certain population partitions and interpreted as latent exposure profiles. This allows us to identify those most at risk of developing obesity or mental health outcomes.
Our approaches involve investigating the association of all exposures and will rely on informing models about the prior grouping of risk factors according to the exposome domains they contribute to. This approach allows:
- identification of which domain contributes to the healthy development of children,
- within each domain, identification of which features most contribute to the differential development of children, and
- quantification of the marginal effect each of these factors has.
This analysis informs the prioritisation of targets for risk stratification and health interventions.
Our methodological approach also uses molecular data to investigate the biological signatures of relevant sets of exposures and clinical outcomes, respectively. Using graphical models, we can visualise complex correlation structures among the exposure and/or disease related biomarkers to infer possible mechanisms involved in the embodiment of these insults and subsequent health consequences.
Leveraging our current work in the LongITools and Healthy Choices projects, we extend current approaches in a longitudinal context to explore the dynamics of change in these biomarkers in early-life and whether changes in these may then differentially affect child development.
This theme will build upon exposure surfaces that have been created within the EXPANSE project to provide (via residential address linkage) time-resolved environmental exposures (including pollution, noise, urban environment, green/blue space etc.) to perform life course exposome analyses.
The LongITools project investigates the interactions between the environment, lifestyle and health in determining people’s risks of developing chronic cardiovascular and metabolic diseases. Environmental exposures (e.g. noise and air pollution) interact with genetic factors and can increase risks to developing certain conditions.
Much of Europe’s population live and interact with urban environments which consist of social and environmental factors. All of the factors that make up an urban environment are known as the urban exposome, all which have an impact on a person’s health and provide important targets to improve population health. EXPANSE seeks to understand how these environmental drivers have an effect on our health.
Funded by the Norwegian Research Council, this project will explore and characterise the social exposome. This includes measuring an individual’s exposures and the impact of these on health and ageing trajectories.
Integrating molecular data
By integrating molecular data into our analysis we can investigate the biological signatures of exposures and associated clinical outcomes. For example, the underlying biological association between childhood obesity and depression is unclear. Using graphical models we can visualise the complex correlation among exposure and disease biomarkers, which can reveal possible health consequences.
Professor Marc Chadeau-Hyam
Personal detailsProfessor Marc Chadeau-Hyam Professor in Computational Epidemiology and Biostatistics
Marc Chadeau-Hyam is a Professor in Computational Epidemiology and Biostatistics in the School of Public Health, Imperial College London. His expertise is in the analyses and integration of complex and high-dimensional data to derive reproducible and interpretable determinants of health.
Dr Oliver Robinson
Dr Oliver Robinson
Lecturer in Molecular Epidemiology
Dr Dragana Vuckovic
Dr Dragana Vuckovic
Lecturer in Computational Epidemiology and Biostatistics
Research Assistant (PhD Candidate) in Computational Epidemiology and Biostatistics