Study Discovers How AI Is Revolutionizing Healthy Eating Every Day - Video

A recent study published in Frontiers in Nutrition explored how artificial intelligence (AI) can help create personalized weekly meal plans based on the Mediterranean diet, aiming to improve health and prevent lifestyle diseases like obesity, stroke, and diabetes. The research was led by nutrition experts who wanted to see if AI could offer balanced meal suggestions that fit individual needs, such as allergies, calorie requirements, and food preferences.

To test this, researchers generated 4,000 different user profiles with various dietary restrictions, including allergies to dairy, eggs, nuts, and fish, as well as preferences like halal food. The AI-based nutritional recommendation (AINR) system then created meal plans by filtering foods by season, avoiding allergens, meeting calorie and nutrient targets, and ensuring food variety throughout the week.

The findings showed the system was very accurate—100%—in filtering meals for country, allergies, preferences, and seasonality. However, meal plans could only be generated for about 90% of the profiles. For example, Spanish users allergic to milk or nuts didn’t get plans due to missing alternative options in the database. Turkish men’s plans met calories less accurately but were better for Turkish women.

The researchers said the AI system shows great promise but needs more food options and real-world testing to improve, especially for allergies and certain groups. The next step is to test this AI tool in the SWITCHtoHEALTHY program and eventually expand it to cover meals for entire families.

This study highlighted how AI might soon make it easier for people to eat healthy meals tailored to their individual needs, helping prevent chronic diseases and maintain better overall health.

Reference: Kalpakoglou, K., Calderon-Perez, L., Boque, N., et al. (2025). An AI-based nutrition recommendation system: technical validation with insights from Mediterranean cuisine. Frontiers. doi: https://doi.org/10.3389/fnut.2025.1546107