At the BPH, we hosted two major days dedicated to Inserm’s EXPOSOME initiative. During this first edition, held on 18 and 19 September 2025, many experts from different disciplines came together to discuss the theme of
‘Social inequalities and the exposome’

 

 

 

 

The major event of the Inserm Exposome programme in 2025 was the international symposium on ‘Social Inequalities and the Exposome’ organised in Bordeaux by Inserm, the BPH and the University of Bordeaux. This was an opportunity for this interdisciplinary exposomics network to discuss methods and approaches already developed or in the future, while strengthening national and international connections through collaborative projects.

 

 

The BPH/UMR1219 has developed a strong research axis on the exposome in public health, involving many of its researchers and academic partners. The main structuring projects cover methodological, fundamental and applied dimensions in the analysis of environmental and societal exposures and their impact on public health.

 

 

 

 

The concept of the ‘exposome’ was described in 2005 by Christopher Wild and is defined as the history of environmental exposures to which an people is subjected throughout their life, including external physical exposures, psychological and social context, and internal environmental regulations with regard to various stresses and long-term effects, as well as those that may be multigenerational.

 

The concept of the exposome allows for a better understanding of all the links between the environment and health, including all the environmental exposures we face throughout our lives. It helps the scientific community to understand these complex interactions as a whole and assists in the modelling of certain processes.

 

 

 

 

“There are many types of exposure that cause concern,” says Cécilia Samieri, researcher and co-leader of the ELEANOR-BPH team.

 

“For 20 years, individual risk factors linked to behavioural habits have been studied more than contextual factors in communities (access to green spaces, air pollution, etc.) and social aspects. There are probably multiple exposome models specific to certain pathologies or situations” she explains.

‘The social exposome is an important component of the exposome, comprising key factors, i.e. initial factors, that influence many other exposure” according Cécilia.

 

 

 

 

Studying the exposome could enable scientists to identify the most significant vulnerabilities that should be targeted by prevention policies.

Vulnerable populations have been identified as being most at risk from worrying exposures. In addition to social vulnerability, which is particularly significant, there are also biological vulnerabilities that are currently being studied, such as female gender in relation to an increased risk of certain diseases such as Alzheimer’s, an individual’s intrinsic inflammatory or immunological predisposition, etc.

 

 

 

On the sidelines of this symposium, a workshop entitled “Causal Inference in High-Dimensional Setting” was held the following day, organised by Quentin Clairon, a researcher in the BPH-SISTM team.

 

Discussions focused on new analytical methods aimed at making better use of these complex data.

Indeed, analysing environmental exposures throughout an individual’s lifetime requires statistical methods capable of handling large-scale data composed of different types of measurements, based on chemical, biological, physical, social and other factors, which are often complexly correlated. This data is collected on numerous subjects and is constantly evolving over time, according to Quentin Clairon.

 

Various statistical inference tools can be used to explore and analyse relationships in research data. These approaches help to better understand the mechanisms and factors under study, can reveal existing links, help to determine cause-and-effect relationships, or describe how phenomena evolve over time.

 

The various statistical tools discussed are based on:

• learning exposure networks when measurements are incomplete or difficult to correlate
• creating simplified causal structures adapted to the data actually available
• estimating models that are more robust in the face of cohort data inaccuracies
• analysing mediation over time using data collected at discrete intervals
• developing mechanistic models capable of handling a large number of variables.

 

With regard to the use of AI in research, Quentin Clairon believes that computational and objective limitations prevent the direct use of these tools.

Today, there are still many challenges , ethical and technical issues to overcome, and it is this entire network of actors around EXPOSOME that is organising itself to accelerate research that could lead to better prevention and action on social inequalities in health and public policy.

 

 

 

---

---

 

 

 

Read the post of Inserm Nouvelle-Aquitaine : Exposome : cartographie des expositions d’une vie – Inserm pro