Evaluating Function Availability and User Satisfaction with Nutrition Apps: User- Generated Content Analysis and Text Mining Study

Background Nutrition management is essential for sub-health and chronic diseases control. As key digital aids for self-nutrition management, whether nutrition app functionalities align with task requirements, especially factors driving user satisfaction, remains unclear. Objective To evaluate the function availability and user satisfaction of nutrition apps. Methods Apps were selected from mainstream iOS/Android stores in China and the US, collecting user-generated ratings and reviews. Function availability was assessed by aligning app functions with clinical guideline requirements. Topic clustering used Latent Dirichlet Allocation (LDA), with DeepSeek-R1 naming factors. Tobit regression and Wald test analyzed the asymmetry of factors’ impact, followed by KANO classification. A user satisfaction index system was developed using the Analytic Hierarchy Process (AHP). Results We included 537 apps (US: n = 406; China: n = 131) and 724,545 reviews (US: 455,245; China: 269,300). Chinese apps had higher coverage in target monitoring (n = 111, 84.73%) and education support (n = 101, 77.10%), while US apps had higher coverage in dietary behavior intervention (n = 364, 89.66%). LDA identified 13 topics in Chinese reviews and 12 in US reviews. In China, the basic factor causing dissatisfaction was reliability (β = 2.852, P < 0.001) and the primary attractive factor boosting satisfaction was fee (β = 0.133, P < 0.001). In the US, the basic factor causing dissatisfaction was food database (β = 0.896, P < 0.001) and the primary attractive factor boosting satisfaction was weight management function (β = 0.053, P < 0.001). Overall, Chinese users had higher comprehensive satisfaction (0.8901/1) than US users (0.8372/1). Conclusions This study revealed that all factors influencing satisfaction exhibited asymmetric effects. All Chinese factors were charm attributes, whereas US factors included charm and essential attributes. Both countries showed functional gaps in personalized services. Distinguishing between basic and attractive factors, along with assessing function availability, can help optimize the design and adoption of nutrition apps.