Behavioral clusters in a nationwide mobile healthcare program and their associations with glycemic control and quality of life in older adults

Abstract

Mobile health (mHealth) interventions delivered through public health centers have emerged as scalable strategies for chronic disease management in aging societies. However, participant engagement with these programs is heterogeneous, and little is known about how different behavioral clusters relate to health outcomes among older adults. We conducted a retrospective cohort analysis of 2,012 adults aged ≥60 years who participated in the 2023 nationwide mHealth program in South Korea. Engagement indicators included daily steps, weekly exercise duration, meal logging frequency, diet quality scores, and goal adherence. Latent cluster analysis identified behavioral subgroups. Outcomes were six-month changes in hemoglobin A1c (HbA1c), body mass index (BMI), and health-related quality of life (HRQoL, EQ-5D-5L and HINT-8). Multivariable regressions adjusted for sociodemographic and clinical covariates. Results: Four distinct clusters were identified: Exercise-oriented (31.8%), Diet-focused (27.4%), Low-adherence (24.9%), and Balanced (15.9%). Compared with the Low-adherence group, the Balanced cluster achieved the largest improvements (ΔHbA1c –0.7%, ΔBMI –1.2 kg/m², ΔEQ-5D-5L +0.06, all p<0.01). Exercise-oriented participants demonstrated greater BMI reductions (β –0.80, p<0.001), while Diet-focused participants achieved meaningful HbA1c improvements despite modest weight loss (β –0.38, p<0.001). Subgroup analyses revealed stronger HbA1c benefits among women in the Diet-focused cluster and greater QoL gains among participants aged ≥70 years. Conclusions: Engagement in community-based mHealth programs is heterogeneous and meaningfully associated with metabolic and QoL outcomes among older adults. Exercise, diet, and balanced behavioral clusters each conferred distinct benefits, while low adherence yielded minimal improvements. These findings underscore the need for tailored digital health strategies that leverage behavioral clustering and AI-driven personalization to optimize chronic disease management in aging populations.