Can ChatGPT provide appropriate meal plans for NCD patients?

Abstract

Dietary habits have a significant impact on health condition and are closely related to the onset and progression of non-communicable diseases. Consequently, a well-balanced diet plays an important role as a treatment to lessen the effects of various disorders, including non-communicable diseases. To propose healthy and nutritious diets, several AI recommendation systems have been developed, with most of them using expert knowledge and guidelines to provide tailored diets and encourage healthier eating habits. On the other hand, new advances on Large Language Models (LLMs) such as ChatGPT, with their ability to produce human-like responses, has led several individuals to search for advice in several tasks, including diet recommendation. This work comprises the first study on the ability of ChatGPT models to generate appropriate personalized meal plans for patients with obe sity, cardiovascular diseases and Type-2 diabetes. Using a state-of-the-art knowledge-based recommendation system as a reference, this work assesses the meal plans generated by two LLM models in terms of energy intake, nutrient accuracy and meal variability. Experimental results with different user profiles reveal the potential of ChatGPT models to provide personalized nutritional advice, however additional supervision and guidance by nutrition experts or knowledge-based systems is required to ensure meal appropriateness for users with non-communicable diseases.