JMIR Publications

ABSTRACT

Background:

Anxiety is highly prevalent among young adults, yet access to timely treatment remains limited. Mobile-based digital therapeutics, particularly Ecological Momentary Assessment and Intervention (EMA·EMI), offer a promising approach for delivering context-responsive care. Integrating wearable-derived digital phenotypes may enhance the precision and personalization of such interventions, but empirical evidence from real-world settings remains scarce.

Objective:

This study examined the real-world applicability and effectiveness of an EMA·EMI–based digital therapeutic, ANZEILAX-Green, for anxiety reduction in young adults. It further investigated how personalized feedback generated by combining EMA·EMI data with wearable-based digital phenotypes supports symptom improvement, self-management, and user engagement.

Methods:

An eight-week exploratory experimental study conducted with young adults aged 19–39 years. Participants used a smartphone app integrating EMA·EMI modules along with a smartwatch that continuously collected digital phenotyping data, including heart rate, sleep, and activity. Psychological outcomes were assessed using the Generalized Anxiety Disorder 7-item scale (GAD-7) and Patient Health Questionnaire-9 (PHQ-9) at baseline, weeks 2, 4, 6, and 8, and Rosenberg Self Esteem Scale (RSES) at baseline and week 8. Quantitative changes were analyzed using paired t tests and repeated-measures analysis of variance, and qualitative interviews were thematically analyzed to identify user experience mechanisms and design implications.

Results:

Twenty-nine participants were included in the final analysis. Significant improvements were observed in GAD-7 scores (mean difference –4.79, 95% CI –6.56 to –3.02, P<.001, Cohen d=1.03), PHQ-9 (mean difference –3.86, 95% CI –5.79 to –1.93, P<.001, Cohen d=0.76), RSES (mean difference +5.21, 95% CI 3.47 to 6.85, P<.001, Cohen d=1.14). Consistent improvements in GAD-7 scores were observed across assessment points, with statistically significant reductions from week 4 onward. Participants demonstrated high adherence, completing 80% of EMA·EMI prompts. Digital phenotype trends showed increased deep sleep and reduced mobility, and sleep indicators significantly correlated with depressive symptoms. Interview findings revealed five mechanisms: enhanced self-awareness, data-driven reflection, behavioral change, intrinsic motivation, and extension of treatment into daily life.

Conclusions:

EMA·EMI–based digital therapeutics integrated with wearable-derived digital phenotypes are feasible and effective for reducing anxiety in young adults. The combination of real-time emotional self-reporting and passive physiological monitoring enables more personalized, timely, and context-aware interventions. These findings provide empirical and design insights for future automated, personalized mental health intervention systems.