Harnessing Artificial Intelligence for the Provision of Personalised Nutrition Advice to Population Groups across the UK ^†

Abstract

Personalised nutrition could promote greater adherence to a healthy lifestyle, and thereby potentially improve health outcomes. The principal aim of the PROTEIN project was to develop a mobile application that delivers tailored nutrition advice to adults. In this pilot study, 80 participants were recruited from the general public and sorted into three groups: (i) adults with a poor-quality diet (PQD, n 29), (ii) adults with iron deficiency anaemia (IDA, n 11; Hb < 120 mg/L) and (iii) adults who were overweight (OW, n 40; BMI 25–30 kg/m²). The participants provided baseline anthropometric and general health data, which were inputted into the PROTEIN dashboard, along with their dietary preferences and individual goals, triggering the generation of an individualised 7-day nutrition and activity plan (NAP), which the participants were encouraged to follow. Their interactions with the app were determined through the number of occasions the user would either ‘confirm’ that they had consumed or ‘skipped’ a recommended meal. They were also expected to rate the meals and input their own instead of, or as well as, those recommended. Following 4 weeks of use, the participants were asked to complete online questionnaires on the usability of the app and report their current weight. The data are presented as mean (±SD); the significance was set at p < 0.05. The mean age and BMI were 44.7 ± 16.1 years and 27.7 ± 5.5 kg/m², respectively, for the whole sample. Over 90% of the users did not confirm that they had consumed or skipped a meal, thereby suggesting a lack of user acceptability of the meal plans provided. However, the OW group users who completed the intervention (n 32) reported an average of −1.1 ± 1.4 kg weight loss. The responses to questionnaires from all users suggested that the app increased their ‘motivation to’ and ‘ability to eat a healthy diet’ (n 35 and 41, respectively). Overall, the PROTEIN app could motivate users to improve their lifestyle, in line with previous pilots. Furthermore, the system could accurately define appropriate meal plans and aid its users achieve their personal ‘goals’. Future versions of the mobile app should focus on developing a more user-friendly system to increase interaction.