MON-729 Association Between Meal Recording And Weight Change Using An Ai-enabled Smartphone App: A Longitudinal Analysis Of Individuals Eligible For Specific Health Guidance

Abstract Disclosure: M. Kometani: asken.inc. R. Kitajima: None. Y. Noda: None. M. Kadono: None. S. Konishi: None. T. Yoneda: None. Background: In Japan, the “Specific Health Guidance (SHG)” programs were introduced in 2008 to prevent lifestyle-related diseases and their progression. These initiatives target individuals aged 40-74 enrolled in public health insurance. SHG provides lifestyle support for those at high risk of metabolic syndrome, identified through physical measurements, blood tests, and questionnaires. This study analyzed dietary intake and weight change among participants who voluntarily engaged in SHG and used the AI-powered smartphone app, which analyzes food photos to estimate and visualize 14 nutrients including calories, protein, fat, and carbohydrates. Objective: To investigate the association between nutrient intake and weight change at 3, 6, and 12 months among SHG participants who recorded meals using an AI-enabled smartphone application. Methods: This study included 99 individuals (mean age: 50.9 years; 26 women) who had received SHG in a prior study and initiated use of the application. All participants met SHG eligibility criteria (metabolic syndrome or pre-metabolic syndrome) and recorded dietary intake and body weight for an average of 195 days using the app, which employs AI-based image analysis. Pearson correlation coefficients were calculated between estimated daily intake of protein, fat, carbohydrates, and energy and weight change at 3, 6, and 12 months. Comparisons were made between weight loss and non-weight loss groups. Results: Weight data were available from 86 participants at 3 months, 63 at 6 months, and 20 at 12 months. Of the 86 participants at 3 months, 65 (75.6%) experienced weight loss, with an average reduction of 2.2 kg. Most weight changes occurred within the first 3 months. A significant negative correlation was observed between weight change and protein intake at 3 months (r = -0.25, p = 0.020), with similar trends at 6 months (r = -0.23, p = 0.060) and 12 months (r = -0.29, p = 0.21). No significant correlations were found for energy, fat, or carbohydrate intake. Protein intake at breakfast was significantly associated with weight loss (r = -0.22, p = 0.030). The weight loss group had significantly higher protein intake (p = 0.006) and lower energy and carbohydrate intake (p = 0.011 and p = 0.018, respectively). The protein ratio in macronutrient balance was slightly higher in this group. A significantly higher proportion of women was also observed among those who lost weight (p = 0.044). Conclusion: Among individuals participating in SHG, an association was observed between dietary protein intake—especially at breakfast—and weight change over time. These findings suggest the potential importance of nutrient composition and intake timing in weight management. Further studies are needed to evaluate whether such dietary patterns contribute causally to weight change and how digital dietary tools may support health behavior tracking and assessment. Presentation: Monday, July 14, 2025