Search form


French Agency for Food, Environmental and Occupational Health & Safety

Search form

Contrasting perspectives on the conference

Chemical mixtures: challenges for research and risk assessment

Humans are exposed to multiple substances in their surrounding environment including food, water, air, soil and consumer goods. In the last decades the possible effects of chemical mixtures on humans and on the environment have become a major concern in many countries across the world. This is why scientific experts and risk managers are increasingly addressing this challenge both in supporting research programmes and developing new methodologies to assess risks. In this context ANSES, National Food Institute/DTU and BfR have decided to convene this conference on the state of the art on chemical mixtures bringing together researchers and scientists from European and cross Atlantic agencies to stimulate international exchange of practices and research knowledge.


Reflections by 3 of ANSES’s scientists - a social economist, a toxicologist and a specialist in chemical risk - on the two-day conference on exposure to chemical mixtures. The views expressed belong to their authors and do not necessarily reflect the opinions of ANSES.


10 December 2013

Chemical risk specialist

Overview of the first day

Anyone who might need to update their knowledge on health risk assessment in the field of chemical mixtures can take this conference to be a perfect opportunity, and apparently a rare one, from what I have heard during the breaks.  And for those who are not lucky enough to attend, the proceedings and slide shows made available will be a precious resource for accessing the relevant research work, tools and institutional guidance documents, at the cutting edge in the field. 

Soon chemical mixtures will no longer be considered a novel issue involving too many methodological difficulties to enable them to be assessed as a health risk. Of course there are still numerous uncertainties and many questions that remain unanswered.  But it seems clear that a sufficient amount of knowledge has begun to accumulate.

The different health and safety agencies have not yet fully integrated the problem of mixtures into their health risk assessment methods, and the same also goes for the requesters who petition them. The remaining obstacles must still be identified and overcome before this transition can be made.


Social economist

Toxicology, the end of the "substance by substance" approach 

Numerous chemicals surround us all in our everyday life.  Until now, toxicological analyses were performed one substance at a time. The presentations given during this conference highlight how unrealistic such an approach actually is, and invite us to consider the principles of toxicology through the prism of chemical mixtures.  Many studies have been conducted on pesticide intake in particular through the alimentary route.  Special care seems to have been taken to provide realistic experimental conditions in the studies, which is well appreciated. Clearly an effort has been made to avoid creating a rupture between laboratories and the real world. Consequently, this seems to be a good opportunity to extend research to classes of substances present in other environments to which humans are exposed, through inhalation or skin contact, for example. The acquisition of biomonitoring data seems to be a complementary approach for the identification of relevant and realistic mixtures in addition to theoretical exposure. The main challenge will be seeing how these elements can improve the approach to health risk assessment and how at the end of the day they can be used for risk management.



First day: the end of "dose-additivity"?

I attended this first day of the conference with one question in mind: is the default "dose-additivity" approach still an acceptable one (or at least a "lesser evil")? The presentations by Sybille Ermler (Brunel University), Lars Niemann (BfR) and Ulla Hass (DTU) support this position. The effects observed in their studies on chemical mixtures (mainly dealing with pesticides) can be predicted using a dose-additive model, regardless of whether or not the substances taken separately at the same dose produce an effect. Cases of synergistic effects are rare. However, other presentations were made which cast doubt on these findings. Supra-additive effects were reported by Roger Rahmani (INRA) using an in vitro model. A study on rats by Anne-Marie Vingaard (DTU) showed a potentiation effect on concentrations of circulating hormones. But as she concluded, it is still too early to know whether such findings are proof of adverse effects.And the results provided by the "OMICs" (gene expression, metabolomics) in her study and those by Lars Niemann (BfR) and Laurence Gamet-Payrastre (INRA) are at the same stage: the statistical analyses show differences between the substances alone and when mixed, but we do not yet know how to interpret the impact of these changes on health. Are they an early sign of adverse effects or just simple adaptive responses by the body? Certain more worrisome data show that mixtures modify the toxico-kinetic parameters of substances. What is their impact on the development of PBPK models which aim to extrapolate results of mixtures tested in vitro? Clearly, many questions have yet to be answered and much additional research is still needed. The second day of the conference will perhaps provide us with some of the answers...

11 December 2013


Mixture toxicity: a substance-based approach or one based on the exposure profile? 

With regard to the research on mixture toxicity presented during this two-day conference, two approaches seem to be in competition. On the one hand, we have an approach which takes chemical substances and attempts to group together those having a common target or mechanism of action (like the NTP program on polycyclic aromatic hydrocarbons presented by Cynthia Rider of the NIEHS), and on the other hand we have an exposure-based approach (biomonitoring data and the exposome concept presented by Christopher Wild of the IARC, in particular). Personally, I find these approaches to be complementary. Together, they should make it possible to identify tools, criteria and methods for assessing mixture toxicity because it now seems clear that a single approach is not appropriate, and that the Hazard Index has shown limitations.



Social economist

The issue of substance synergies remains a major challenge

Final comments on the conference on exposure to chemical mixtures. 

The various presentations and discussions today showed a glimpse of the difficulties encountered beyond the toxicological phase discussed on the first day with regard to this topic. It is therefore regrettable that no "mixture TRVs" were presented. In addition, the issue of substance synergies remains a major challenge. Risk assessment does currently seem to be based essentially on a cumulative logic (Hazard Index) even though certain elements presented by Richard Hertzberg tend towards an acknowledgement of interactions. The most complete results mentioned during the conference are based on studies conducted on a single class of substances (phthalates, flame retarders) and present the advantage of considering several exposure routes.

Lastly, the regulatory framework was brought up regularly during the round table discussions, both by association representatives and by scientists. And yes, despite the uncertainties which are inherent in risk assessment, several speakers consider that current knowledge with regard to mixtures is sufficiently advanced for legislation to now be updated.


Chemical risk specialist

"We need to take on the challenge of integrating the issue of chemical mixtures into our current health risk assessments"

While the first day of the conference gave the impression that the work on chemical mixtures involved pesticides in foods alone, the second day showed more diverse applications through case studies. 

The work conducted by Hélène Budzinski's team at the University of Bordeaux on the identification of endocrine disruptors in a river polluted by a pharmaceutical plant, and those by the Danish EPA on phthalates were of particular interest to me. Pesticides did however seem to be the preferred topic for work on mixtures, to the dismay of HEAL, an association that deplored the fact that other types of chemicals were somewhat ignored.

As I already suspected yesterday, several of the participants in the final round table – in particular Marco Vighi, a long-time member of the European Commission's scientific committees on ecotoxicology – consider that there is now sufficient knowledge on mixtures for it to be integrated into the regulatory process. HEAL even considers that we were ready as early as 2009. So my conclusion for yesterday still stands: we need to take on the challenge of integrating the issue of chemical mixtures into our current health risk assessments.



Presentation by Cynthia Rider of the NIEHS

As a toxicologist, I especially appreciated the presentation by Cynthia Rider of the NIEHS and more specifically her description of the NTP program on polycyclic aromatic hydrocarbons (PAHs) which intends to compare in vitro assays with 28-day in vivo mouse studies in order to test additivity hypotheses, toxic equivalency factors, and the extrapolation of in vitro to in vivo data (including a study of the toxico-kinentic parameters of PAHs alone or in mixtures). She also mentioned the use of the Tox21 high-throughput screening tool for testing a hundred or so mixtures composed of as many as 62 substances in in vitro assays on several cell lines and comparing these results with the predictions provided by dose-addition or response-addition models. The data generated by these research programmes should enable major progress in the area of mixture toxicity.