Use of a Culturally and Personally Tailored Cooking Application is Associated with Improved Glycemic Control in Type 2 Diabetes Mellitus

Background: Digital health interventions may support diabetes self-management, but evidence for fully automated, culturally tailored dietary applications without direct human support remains limited. Objective: To examine the association between use of the Oishi-kenko mobile cooking application and changes in glycemic control, body weight, and dietary behaviors among individuals with type 2 diabetes mellitus. Materials and Methods: This 12-week observational study was conducted entirely online. Participants were recruited via the Oishi-kenko website, internet advertisements, and Tsukuba City public relations channels. A total of 24,671 website visitors, 214 installed the application, and 65 were included in the final analysis; glycated hemoglobin (HbA1c) data were available for 24 participants. The application provided personalized, culturally tailored recipe suggestions based on user profiles and dietary standards. Body weight and HbA1c were assessed at baseline and every 4 weeks. Dietary habits were evaluated using the Brief-type Self-administered Diet History Questionnaire (BDHQ). Results: Among the 24 participants with HbA1c data, HbA1c decreased from 6.40% at baseline to 6.12% after 12 weeks (−0.28%, P < 0.05). Among participants with baseline HbA1c > 7%, the reduction was −0.63%. Body mass index (BMI) declined from 23.47 to 23.28 overall, with a greater reduction among those with baseline BMI ≥25. More frequent application-based cooking and higher baseline HbA1c were associated with greater reduction. BDHQ results suggested favorable dietary changes, although none reached statistical significance. More than 80% of participants reported improved dietary habits. Conclusion: Use of the Oishi-kenko application was associated with modest improvements in glycemic control, body weight, and dietary behaviors. Controlled trials are needed because of the observational design, small sample size, self-reported outcomes, and high attrition. The trial registry number: UMIN-CTR, ID: R000053861.