Study Reveals AI Calorie Tracking Apps Underestimate Meal Calories by Up to 345
AI-powered calorie tracking apps have become popular tools for individuals seeking to monitor their nutritional intake, particularly for those managing their health or attempting to lose weight. These apps utilize image recognition technology to estimate the nutritional content of meals based on photographs. However, new research indicates that the calorie and fat estimates provided by these apps may be significantly lower than the actual content of the meals. In a study presented at NUTRITION 2026, researchers from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) found that four widely used photo-based apps consistently underestimated calorie counts by an average of 250 to 345 calories per meal. The study involved a rigorous testing process where researchers prepared 102 meals in a controlled metabolic kitchen, measuring each ingredient with high precision. The meals were then photographed and analyzed using the apps MyFitnessPal, LoseIt!, CalAI, and Appediet. The results showed a notable discrepancy in the calorie and fat estimates, with the apps also underestimating fat content by approximately 30 grams. While MyFitnessPal and LoseIt! demonstrated better accuracy with higher-calorie meals, all four apps struggled more with lower-calorie options. This inconsistency raises concerns about the reliability of these tools for users who may not adjust portion sizes or manually enter food amounts. Following the initial analysis, researchers tested over 200 additional meals to explore factors affecting app accuracy. Preliminary findings suggest that the apps may face challenges in accurately evaluating meals from low-carbohydrate ketogenic diets, which typically contain higher fat content. The researchers recommend that users combine photo-based tracking with traditional methods of assessing food intake for improved accuracy. It is important to note that the results presented are preliminary and have not yet undergone the full peer review process required for publication in a scientific journal. Therefore, the findings should be interpreted with caution until they are confirmed through further research.




