Mobile Health App Acceptance in Japan's Aging Society: Multigroup Structural Equation Modeling Based on the Extended Unified Theory of Acceptance and Use of Technology and eHealth Literacy Frameworks.

BACKGROUND: Mobile health (mHealth) technologies are increasingly promoted as tools for chronic disease management and healthy aging, yet adoption remains persistently uneven across demographic groups. Japan, where 29.1% of the population is 65 years or older-the highest proportion globally-exemplifies the challenges of mHealth promotion in super-aging societies. Despite high smartphone penetration (90.1%) and active national digital transformation initiatives, only 21.6% of Japanese adults report regular mHealth app use, with marked disparities by age and sex. OBJECTIVE: This study examined determinants of mHealth acceptance by extending the unified theory of acceptance and use of technology to incorporate eHealth literacy, self-efficacy, perceived risk, distrust, and health-related factors (health status and health interest). Age- and sex-specific differences in acceptance mechanisms were also investigated using multigroup structural equation modeling (SEM). METHODS: We conducted a cross-sectional online survey in November 2023 with 960 Japanese adults sampled across 7 age strata (aged 18-27 years to aged ≥78 years). SEM tested hypothesized relationships among 9 constructs. Health status and health interest were included as observed covariates. Multigroup SEM with configural, metric, and structural invariance testing examined age- and sex-specific differences, and binary logistic regression identified predictors of current mHealth app use. RESULTS: /df=2.06; comparative fit index 0.953; Tucker-Lewis index 0.945; root mean square error of approximation 0.047) and explained 71.6% of the variance in behavioral intention. Effort expectancy (β=0.404), facilitating conditions (β=0.349), and performance expectancy (β=0.188) were the primary proximal predictors of behavioral intention. Social influence exerted strong upstream effects on effort expectancy (β=0.811), eHealth literacy (β=0.507), and self-efficacy (β=0.422). Health interest positively influenced performance expectancy (β=0.133), whereas neither health interest nor health status showed a significant direct effect on distrust. Distrust did not directly predict behavioral intention in the overall sample. Multigroup analyses identified 5 significant age differences and 5 sex differences. eHealth literacy increased distrust among young adults but reduced perceived risk among middle-aged and older adults. Self-efficacy negatively predicted performance expectancy among young adults yet positively predicted it among middle-aged and older adults. Distrust significantly reduced behavioral intention only among middle-aged adults. CONCLUSIONS: mHealth acceptance in Japan's aging society is characterized by stable proximal determinants of behavioral intention alongside heterogeneous upstream belief formation processes that vary systematically by age and sex. Health interest, rather than health status, emerged as the key contextual driver of perceived usefulness. At the theoretical level, this study clarifies how eHealth literacy, self-efficacy, and distrust function as age- and sex-contingent antecedents within an extended unified theory of acceptance and use of technology framework. At the practical level, these findings highlight the need for trust-centered, demographically tailored, and literacy-sensitive strategies to promote equitable mHealth adoption in rapidly aging societies.