Exposure to ‘cocktails’ of food additives and chronic disease risk

ended

Today, our daily diet typically contains dozens of food additives (e.g. colours, emulsifiers, sweeteners: ~350 substances allowed on the EU market). Safety assessment is performed by health agencies to protect consumers against potential adverse effects of each additive, yet such an assessment is only based on current available evidence, i.e., for most additives, only in-vitro/in-vivo toxicological studies and exposure simulations. Meanwhile, the long-term health impact of additives intake and any potential ‘cocktail’ effects remain largely unknown and have become a source of serious concern. Growing evidence link the consumption of ultra-processed foods, containing numerous additives, to adverse health outcomes, in particular our recent results on cancer (Fiolet BMJ 2018). While most additives allowed in the EU are likely to be neutral for health and some may even be beneficial, recent animal and cell-based studies have suggested detrimental effects of several such compounds. In humans, data is lacking. No epidemiological study has ever assessed individual-level exposure to a wide range of food additives and its association with health, hampered by unsuited traditional dietary assessment tools facing the high additive content variability across commercial brands. Hence, a major breakthrough will come from the novel and unique tools I developed with my team, notably within the NutriNet-Santé cohort (n=164,000), collecting precise and repeated data on foods and beverages usually consumed, including names and brands of industrial products. With this unique resource, I propose a project at the forefront of international research to provide answers to a question of major importance for public health. Built as a combination of epidemiological studies and in-vitro/in-vivo experiments, this project will shed light on individual exposure to food additive 'cocktails' in relation to obesity, cancer, cardiovascular diseases and mortality, while depicting underlying mechanisms.