In a new study using Artificial Intelligence (AI) and machine learning, EPFL researchers discovered that both what we eat and how regularly we eat it are key factors influencing gut health.
O microbiota intestinal is a community of microorganisms, including bacteria, viruses, fungi and other microbes, that inhabit the digestive system. Some of these microbes contribute to health, while others can be harmful.
According to , previous investigations have long demonstrated that Diet strongly influences the composition of the intestinal microbiota. Diets rich in fruits, vegetables, fiber and nuts are consistently associated with greater microbial diversity and better digestive health.
Now, EPFL investigators discovered for the first time that the consistency with which we follow a healthy diet is as vital to intestinal health as the amount or type of food we eat.
In a recent study in Nature Communications, scientists from EPFL’s Digital Epidemiology Laboratory — part of the School of Computer and Communication Sciences and the School of Life Sciences — confirmed previous findings that specific foods, such as fruits and vegetables, promote a more diverse intestinal microbiota.
In collaboration with researchers at the University of California, San Diego, they also discovered new and remarkable knowledge.
Their analysis revealed that maintaining regular intake of fruits, vegetables and cereals is crucial for the development of a healthy intestinal microbiota. While it has long been assumed that eating nutritious foods consistently is beneficial, supported by messages like “five fruits and vegetables a day,” this study provides concrete evidence of this belief.
“This research clearly shows that you cannot eat too many vegetables on a healthy day and then eat unhealthy for the rest of the week or month,” said the Marcel SalathéAssociate Professor, director of the Digital Epidemiology Laboratory and co-director of the EPFL AI Center.
“In fact, our study suggests that irregular consumption of healthy foods negates many of their beneficial effects on the intestinal microbiota. This is a real encouragement so that future studies are not limited to analyzing what people are eating, but rather the patterns of what they are eating over time.”
Secondly, the team was also able to demonstrate that a person’s gut bacteria and what they eat can predict each other with an accuracy of up to 85%.
With just a simple stool sample — an increasingly common component of medical research — advanced machine learning techniques can predict a person’s diet using their microbiota and vice versa.
“To our collaborators in San Diego, who are some of the world’s leading experts in microbiome research, this was exciting”, explained Salathé.
“Obtaining this data from a stool sample is relatively easy, but understanding a person’s diet is notoriously difficult, data that has been difficult to collect.”
The study was made possible thanks to the use of nutritional information details of around 1000 participants who were part of the “Food & You” cohort.
High-resolution dietary data was collected through the app MyFoodRepodeveloped by the same EPFL laboratory, which allowed users record your meals in real timetaking photographs or scanning barcodes.
The app’s AI analyzed these entries for nutritional content, later verified by human reviewers.
“Historically, nutritional research has relied on food frequency questionnaires and 24-hour food records. In theory, we could ask someone to write down everything they eat, but in practice, this is not done because it is practically impossible. Now, AI is so good that we can do this data collection on a large scale,” he said. Rohan SinghPhD assistant at the Digital Epidemiology Laboratory and main author of the article.
“Our study was particularly interesting because when we look at lifestyle-related gastrointestinal diseases, these often develop gradually. Since nutrition is one of the main factors contributing to these diseases, analyzes like ours may be able to assess what can be improved in a person’s diet. AI can then help people adjust their food intake accordingly,” he continued.
Salathé believes the study’s findings suggest that current dietary guidelines may need to be updated to emphasize not only the types and quantities of food, but also the regularity of healthy eating behaviors.
And although this research project has ended, the application MyFoodRepo continues to be used by the Digital Epidemiology Laboratory team for other investigations. They are currently involved in a pilot project on nutrition and cognitive performance, studying potential links between the two.
Additionally, using barcoded dietary data from the Food & You study, researchers are investigating the link between food additives, such as emulsifiers, found in ultra-processed foods, and the gut microbiota.
“There is a strong hypothesis that some of these additives can have a negative impact on the microbiota and we have some early indications that this may, in fact, be the case. We are still in the analysis phase, but we are very excited about the first results”, said Salathé.
In general, they are satisfied with the fact that the application MyFoodRepo be now open the door to important nutrition studies at a global level.
“From the beginning, we knew we needed something that was extremely consumer-friendly and easy to use, while still providing the data we needed. We built it to serve our own research needs, but also in a way that others found useful — and it is now being used in many other nutrition studies around the world,” concluded Salathé.
Teresa Oliveira Campos, ZAP //
