Personalized Nutrition & Citizen Science: A Healthy Diet Through Algorithms

Food & You is one of the first large-scale citizen science research project in Switzerland. This project is building on the results of Zeevi et al. study, which demonstrated in 2015 for the first time with a cohort of 800 participants that there is a high interpersonal variability in postprandial (postmeal) glycemic responses (PPGR). In a second stage, the group developed an algorithm that predicts PPGR and gives personalized diet recommendation to better control the blood sugar level. These results have only been shown once, if they turn out to be correct, it will have a very strong impact on the field of nutrition. Other studies have since then also shown the variability in glucose response, leading to the concept of individual "glucotypes". The most important particularity of Food & You is that it will be a citizen science project. Citizen science is a method that involves volunteers in scientific research activities. It increases scientific knowledge, raises people's awareness of scientific questions, and allows scientists to share enthusiasm and expertise. Furthermore, to enable any citizen to take part, the study will be entirely coordinated digitally. It means that unlike conventional studies, no face-to-face interviews or medical appointments will be organized. All instructions will be given by a website and emails / text messages sent to the participants. The study material will be sent by post and the data will be collected via the website and the MyFoodRepo app. The participants will have to track during 14 days:

• Their dietary intake with the myFoodRepo app.
• Their physical activity and sleep with a smartphone app or an activity tracker.
• Their glucose level wearing the FreeStyle Libre Sensor.
• Collect a stool sample. The microbiome composition will be analyzed by sequencing the 16s rDNA extracted from the stool sample.
• Answer anthropometric, demographic, and health online questionnaires.

With the Food & You project, the research team will build a new algorithm that will predict PPGR with input that is readily available through such a "digital cohort" (i.e. low-burden on user, and affordable) in order to test whether this can scale to tens or hundreds of thousands of people. The results of this project may have substantial consequences in dietary guidelines, personalized diet and on health conditions related to dietary intake.