On January 13, 2025, a new study revealed GroceryDB, a database that reveals the processing levels of over 50,000 food items available at Walmart, Target, and Whole Foods. This study was published in Nature Food.
Dr. Babak Ravandi from Northeastern University led the research, highlighting the challenge of finding information about how processed food items are. To tackle this, they used a method called the food processing score (FPro) and advanced machine learning techniques. GroceryDB allows users to access detailed insights about different foods and their processing levels.
Each food in the database is assigned an FPro score, which is derived from nutrition labels. This system provides a thorough analysis of ingredient patterns and processing levels. Researchers categorized these findings by grocery store, food type, and price, also examining how over 1,000 ingredients contribute to the ultraprocessing of foods.
Coauthor Dr. Giulia Menichetti emphasized the potential of this tool for public health. She mentioned, “Our goal is to use this data-driven approach to improve public health.” The study suggests that leveraging artificial intelligence and data science could make nutrition research more efficient compared to traditional methods that rely heavily on manual tracking.
While the team notes one author has connections to the health technology industry, the focus remains on how GroceryDB can provide valuable information to consumers and researchers alike.
For more detailed insights, you can view the study’s abstract here.
Source link
Journal,Artificial Intelligence,Ultraprocessed Foods