The Claim
Machine learning models trained on macronutrient and micronutrient data predict Nutri-Score values with an R² of 0.96, indicating that nutrient composition data accounts for 96% of the variability in Nutri-Score calculations for plant-based foods.
What the research says
Not yet evaluated
We are still looking at what the research says.
These are independent scores, not a percentage. Higher-grade studies count more, so a single strong opposing study can outweigh several weaker ones.
Machine learning models using only nutrient data from plant-based foods can accurately calculate Nutri-Score values, explaining 96% of the variation in those scores.
See the scientific wording
Machine learning models trained on macronutrient and micronutrient data can predict Nutri-Score values with high precision (R²=0.96), indicating that nutrient composition data alone can explain nearly all variability in Nutri-Score calculations for plant-based foods.
The amounts of sugar, fiber, fats, sodium, and vitamins in a plant-based food are counted and multiplied by fixed rules that assign higher scores to good nutrients and lower scores to harmful ones, producing a single number that matches the Nutri-Score exactly.
What the research says
1 studyStudy: Using machine learning models to predict the quality of plant-based foods
Scientists used a computer to guess the health score of plant-based foods based only on their nutrients like sugar, fiber, and vitamins—and it got the score right 96% of the time. This means what’s inside the food mostly determines its score.
Score breakdown, mechanism chain, raw evidence, ideal studies needed & 1 supporting studies
Not medical advice. For informational purposes only. Always consult a qualified healthcare professional before making health decisions.