Ultra-Processed Foods: Artificial Intelligence’s New Contribution to Nutrition Science – Neuroscience News

Summary: Researchers have developed a machine learning algorithm, FoodProX, that can predict the degree of processing in food products.

The tool rates foods on a scale from zero (minimally or unprocessed) to 100 (highly ultra-processed). FoodProX fills the gaps in existing nutrient databases by providing higher resolution analysis of processed foods.

This development is a significant advance for researchers examining the health impacts of processed foods.

Main aspects:

1. FoodProX is a machine learning tool that predicts the level of processing in a food product.
2. The tool uses nutritional information from the US Department of Agriculture’s Food and Nutrient Database.
3. The AI ​​tool confirmed that more than 73% of the US food system is ultra-processed.

Source: Northeastern University

Northeastern researchers have been busy trying to better understand the links between “ultra-processed foods” and human health through the university-sponsored Foodome Project.

As part of that effort, researchers at the Center for Complex Network Research have now developed a machine learning algorithm that they say accurately predicts the degree of processing of the food items that make up the US food supply.

Their results have been published inNature communicationsin April.

The machine learning classifier, called FoodProX, uses nutrition labeling information from the US Department of Agriculture’s Food and Nutrient Database for Dietary Studies as input to evaluate the level of processing in a given food product.

The algorithm works by producing an output representing the probability that a respective food falls into one of the four categories that are part of the NOVA food classification system, a system developed by researchers at the University of São Paulo, Brazil, which according to researchers is “used extensively in epidemiological studies”.

Users can try the tool by visiting the TrueFood research project website. Users can search for a food to see its food processing score. The algorithm assigns each product a single score ranging from zero (denoting “minimally or unprocessed” food) to 100 (highly ultra-processed food).

Using FoodProX, the researchers were able to fill in the gaps in the nutrient database for dietary studies; classify “complex recipes and mixed foods and meals”; and provide a higher resolution loupe with which to examine processed foods.

As a result, the researchers note that FoodProX provides a sharper understanding of how processed foods are actually an important step for researchers studying the health impacts these foods have.

The researchers note that the NOVA system, which divides foods into four classifications, from “unprocessed or minimally processed” to ultra-processed, is fundamentally limiting because it does not account for the different degrees of processing within each separate category.

‘This perceived homogeneity of NOVA 4 foods limits both scientific research and practical consumer guidance on the health effects of different stages of processing,’ the researchers wrote.

“It also reduces industry incentives to reformulate foods towards less processed offerings, shifting investment from ultra-processed NOVA 4 foods to the less processed NOVA 1 and NOVA 3 categories.”

“What we really do in the paper is say that we believe that nutritional information, so the chemicals that are measured as nutrients in nutritional facts, somehow encode the fingerprint of food processing,” says Giulia Menichetti, senior researcher at the Northeastern Network Science Institute and lead author of the research.

“Because when we process a food, when we change some basic ingredients, we change its chemistry in many different ways.”

That “fingerprint” is how researchers can gather information about how many chemical alterations have been made to a particular food.

“We don’t necessarily know what all the chemical fingerprints are associated, one by one, with each process,” Menichetti told Northeastern Global News. “We can’t even count how many different ways there are to process a food.”

Ultimately, the AI ​​tool confirmed the team’s earlier finding that more than 73 percent of the U.S. food system is ultra-processed, while providing a previously unattainable level of detail. Menichetti says his team is the first to successfully create an AI tool that reliably assesses the chemical content of food.

“It is the first paper in the nutrition and public health space that leverages machine learning to reproducibly and systematically classify foods according to their degree of food processing,” he says.

The team’s work is significant because, as Menichetti says, “there was not much data culture” in the field of nutrition and health science as it relates to food processing, which prompted less scientifically rigorous conversations about what processing.

“When you don’t have a systematic method for examining a food and evaluating its properties, it’s difficult to enable large, comparable studies in other parts of the world,” Menichetti says.

“FPro helps us assess the quality of an individual’s diet, offering predictive power on more than 200 health variables,” says Albert-Lszl Barabsi, Robert Gray Dodge Professor of Network Science at Northeastern and co-author of the study.

“It tells us the impact of replacing processed foods with less processed alternatives of the same item, resulting in personalized dietary changes with minimal effort.”

About this machine learning research news

Author: Tanner Stening
Source: Northeastern University
Contact: Tanner Stening-Northeastern University
Image: The image is credited to Neuroscience News

Original research: Free access.
“Machine learning forecast of the food processing” by Giulia Menichetti et al. Nature communications

---

Abstract

Machine learning prediction of the degree of food processing

Despite the accumulating evidence that increased consumption of ultra-processed foods has negative health implications, it remains difficult to decide what constitutes a processed food.

Indeed, the current processing-based classification of foods has limited coverage and does not distinguish between degrees of processing, hampering consumer choices and slowing research into the health implications of processed foods.

Here we introduce a machine learning algorithm that accurately predicts the degree of processing of any food, indicating that over 73% of the US food supply is ultra-processed.

We demonstrate that an individual diet’s increased dependence on ultra-processed foods is related to higher risk of metabolic syndrome, diabetes, angina, elevated blood pressure, and biological age, and reduces the bioavailability of vitamins.

Finally, we find that replacing foods with less processed alternatives can significantly reduce the health implications of ultra-processed foods, suggesting that access to information on the extent of processing, currently unavailable to consumers, could improve population health.