Nutrition at the crossroads of cultural, economic and ecological factors

Research Director at Inserm and principal investigator of the NutriNet-Santé study, Mathilde Touvier is interested in the causal relationships between nutrition and human health, using a holistic, multidisciplinary approach. Her team's work contributes to the recommendations of the French National Nutrition and Health Program.

For 2022-2023, she has been invited to hold the Public Health Annual Chair at the Collège de France, created in partnership with Santé publique France.

Your work focuses on the relationship between nutrition and health. How did you become interested in this subject ?

Mathilde Touvier : I've always been drawn to the field of human health. At the start of my studies, I thought of studying medicine, but I was more interestedin prevention through research than in clinical care. So I enrolled at AgroParisTech, which specializes in human health and nutrition. Then I took a detour to California, to the University of Davis, where I did an internship involving laboratory experiments on rodent models, focusing on the metabolic pathways involved in the etiology^[1] of obesity and a hormone, leptin. This interlude made me realize that I wanted to work on pathology prevention, but with a disciplinary field that wasn't fixated on a particular metabolic pathway ; with a slightly more holistic approach to the individual, in short, taking into account the factors in his environment, his personal and family history, and his behaviors that influence the risk of chronic disease. These aspects of epidemiology and public health weren't very clear to me at the outset, but became clear to me following this internship, which was very much focused on experimental research and bench work. During my final year at engineering school, I had the chance to carry out my engineering project with the French Food, Environmental and Occupational Health Safety Agency (l'Anses), which inspired me to complete my course with a Master's degree (DEA at the time) in public health. From then on, the dual course of study in nutrition, public health and epidemiology made perfect sense, and I don't regret my choice at all, as it corresponds to me completely in terms of its impact on human health and prevention, with aspects that are at once cultural, socio-economic and socio-ecological - nutrition being at the crossroads of all these factors.

You head up the NutriNet-Santé study, which has been assessing the link between eating habits and health since 2009. How is the data collected and studied ?

The idea is to collect, via the Internet, a vast amount of information on nutritional exposure - everything to do with diet, lifestyle and eating patterns, but also physical activity and sedentariness, notions that are taken into account in the global field of nutrition. We have a set of tools that make this study, at international level, one of the most advanced and detailed on the characterization of these eating habits, with certain emerging behaviors, and others, more traditional. We are working with what is known as " a cohort ". We follow a group of participants over time, and some will develop pathologies such as cancer, diabetes or cardiovascular disease, while others will gain or lose weight, for example. There's a lot going on in terms of health, which we measure by means of a regular questionnaire and a medical committee, which, when a particular event is detected, checks additional information, such as hospitalization reports, to validate all these health events. Our data is also linked to the medico-administrative databases of the French health insurance system. We also collect a great deal of information on lifestyle factors that are not directly related to diet, but which can play a confounding role in the relationship between health and nutrition, such as smoking, sun exposure and medication. The evaluation of these profiles is very comprehensive, with data collected repeatedly throughout the follow-up. This enables us to study the associations between a dietary profile - with its possible nutritional exposure to certain additives or contaminants - and a greater or lesser risk of developing a particular pathology. To date, this large-scale cohort includes more than one hundred and seventy-three thousand participants aged fifteen and over, and is said to be open, i.e. we recruited a large part of the sample in 2009, when the study was launched, but we are still recruiting new volunteers who join us as the study progresses each year.

In thirteen years, what have the data from this study  revealed?

To date, we have published more than two one hundred and fifty publications from the NutriNet-Santé study, notably in major journals such as the British Medical Journal and Nature Food, and more than three one hundred and fifty publications in international congresses. Taking all this work into account is changing public health recommendations. Let's take the example of food additives and ultra-processed foods : concepts we hadn't heard of in France until the first papers from the 2018 NutriNet-Santé study. In these, we showed a link between consumption of so-called " ultra-processed " products - containing additives we could do without such as colorants, emulsifiers, sweeteners or having undergone strong processing, with frying treatments, for example - and an increased risk of cancers, cardiovascular disease, diabetes, obesity, functional digestive disorders, depressive symptoms and mortality in the cohort. The revelation of these relationships has transformed the mentality of consumers, who have developed a certain vigilance with regard to the degree of food processing. Manufacturers began to "clean-label  clean-labelling  since the publication of these studies, manufacturers have begun to "clean-label" by reducing the quantities of additives used. We were then called on several times by the French National Assembly and Senate to reorient public health policy choices in the light of these new findings. The National Nutrition and Health Program (PNNS) has become more aware of the issue, and now recommends that we try to limit the consumption of these foods in the French diet as much as possible. Another result of our work : nutritional labelling, and in particular the Nutri-Score, which was invented and proposed by our team. We know that we shouldn't eat too much fat, too much sugar or too much salt, but how can we get people to take this knowledge on board in practice ? If we have to ask consumers to turn over the packaging and decipher the sometimes complex composition of a product, nobody is going to do it, least of all the most disadvantaged populations. This model would exacerbate the social inequalities linked to nutrition. The aim was therefore to have a very simple, didactic indicator on the front of the packaging, providing information at a glance. Around a hundred publications (fifty of them by our team) from NutriNet-Santé and other studies validate the underlying score and show that people who eat foods that are well classified by the Nutri-Score have a lower risk of developing chronic diseases. Finally, a whole strand of this research is looking at the link between contaminants and the risk of pathologies, as well as the impact of our food on the environment.

Ultra-processed products rightly get a bad press, but continue to be consumed on a massive scale. What needs to change to combat this ?

There's still a lot to be done. Ultra-processed foods form a very heterogeneous group : they range from soft drinks to dishes prepared with emulsifiers. If we want to be able to act to protect consumers, we need to go further and ask ourselves what action needs to be taken in terms of food regulation and targeted public health initiatives. We are currently conducting research to this end, using experimental epidemiological data, and in particular a European-funded project on additives which aims to identify which are associated with the risk of pathologies. One of the first levers for action is at the level of the individual, through education from the earliest age at school, for example. Just as it's shocking to see a teacher smoking outside school, or having a bottle of soda on his desk all day, we hear teachers recommend eating cereal bars before sport, even though they're full of additives and sugars, which makes us wonder about the example we're setting for children. The recommendations of the National Nutrition and Health Program are there to guide citizens, from an early age, but as long as the public budgets devoted to this communication are far lower than those devoted to marketing by the agri-food industry, it will be difficult to get the right messages across. However, we can't blame everything on the consumer ; in other words, we can't give people instructions if, at the same time, they are faced with an uncontrolled supply of food, with vending machines on every street corner loaded with sodas and chocolate bars propelled by very incisive marketing that makes everyone want to eat them... Action needs to be taken on the supply side, for example to lower salt levels in a harmonized way, so that there is no unfair competition. So we need to act at the level of the individual, but also of the supply to which he is subjected and the environment in which he evolves.

How do you get from scientific findings to dietary recommendations ?

In the team I lead, we're lucky enough to be involved in every link of this long and important chain. It starts with the production of scientific knowledge through research, but recommendations are not based on a single study. Studies must be reproducible, and evidence built up through a multidisciplinary approach. In the case of additives, we look at the mixtures to which the population is exposed, and try to determine potential links with pathologies - an epidemiological approach. At the same time, we also collaborate with teams working on mice, in vitromodels , which will mimic these same additive mixtures, in order to test in a causal manner the impact of these mixtures on genotoxicity ^[2], the microbiota, and so on. To provide this evidence in the field of nutrition and health, we build up a body of arguments through a series of epidemiological and experimental studies. All this then goes through a process of collective expertise, steered by bodies such as the French National Cancer Institute, Anses and the French High Council for Public Health. On an international scale, bodies such as theWorld Cancer Research Fund(WCRF) exhaustively list all the scientific literature in the field of nutrition and cancer, using precise criteria to determine levels of evidence. Once these assessments have been completed, the factors whose impact on health (favorable or unfavorable) is judged to be convincing or probable are translated into recommendations for the public, or into changes in regulations for a given substance.

How do you, as a researcher, make your voice heard in the public arena ?

It's a good question, and one that can be asked in many scientific fields, as we've seen in recent years with opinions on Covid given at every turn by non-specialists. This is all the more true in the field of nutrition and health, where everyone is involved ; since we all eat, we're all concerned. These days, with social networks, there's a risk that all voices will be put on the same level. We often see this when we are invited to appear in the media. We like to play the game, because it's a good way of getting the right messages across in the face of a meagre public health communications budget. On TV shows, for example, our opinion is sometimes put on the same level as that of a non-specialist citizen, an industrialist, or someone who has set up a start-up to sell this or that product. There's nothing wrong with hearing different sides, on the contrary, but the problem lies in the levelling of information, because we don't all have the same level of expertise on these subjects. So we always try to keep to a scientific line, to avoid becoming Manichean, with a well-considered discourse that clearly states what we know and what we still don't know.

Have you encountered any other particular difficulties along the way ?

One of the major difficulties is obtaining public research funding, but this is not specific to our field. The other major problem, a little more specific to the type of work we do, is the pressure we come under from the agri-food industry and detractors. When we proposed the Nutri-Score, there was an outcry and, at the time, some industrialists even tried to have our research work banned by writing to government ministries... These are pressures that can be found whenever we want to put public health ahead of the economic interests of large companies... whose anti-Nutri-Score lobbying intensity is inversely proportional to the nutritional quality of their product portfolio. That's what motivates me even more. But that comes with the constraint, for us, of operating solely with public funding to avoid situations of conflict of interest.

Interview by William Rowe-Pirra

Glossary

[1] Etiology : study of the causes and factors of diseases.

[2] Genotoxicity : a substance is said to be genotoxic when it can compromise the physical integrity or functionality of the genetic material (DNA) in cells.