Adaptation of a Self-Guided Cognitive Behavioral Therapy App for Middle-Aged and Older Adults with Anxiety or Depression: Preliminary Acceptability, Feasibility, and Efficacy

Introduction

Major depressive disorder and anxiety disorders are prevalent and debilitating in middle-aged and older adults. The risks associated with these disorders are reduced with adequate treatment, but only a minority of individuals with major depression or anxiety disorders receive any intervention. While self-guided mobile apps can increase access to care, existing apps often omit key therapeutic elements, have low rates of user engagement, and are rarely tailored for older adults. We developed the Maya cognitive behavioral therapy (CBT) app to address these problems by delivering a comprehensive array of CBT skills via an interactive interface informed by principles of behavioral economics. Although Maya was originally designed for younger adults with anxiety, its highly modifiable format allows for the creation of new programs tailored for specific audiences.

Methods

We adapted Maya for middle-aged and older adults with anxiety or depression by modifying language and examples throughout the app to be more relevant to the target population. Adults aged 40 and older with depressive or anxiety disorders used the modified Maya app for two sessions per week, 20 minutes per session, for six weeks. Major assessments were conducted at baseline, week 3, week 6 (end of intervention), and week 12 (follow up). Acceptability was measured by User Version of the Mobile Application Rating Scale (uMARS) score at weeks 1, 3, and 6. Feasibility was measured by the total number of sessions completed by week 6. Preliminary efficacy was measured in terms of change in anxiety (Hamilton Rating Scale for Anxiety, HAM-A) and depression (Montgomery-Asberg Depression Rating Scale, MADRS) from baseline to week 6. Qualitative feedback questionnaires were collected to inform future iterations of the app.

Results

Data collection is still ongoing; updated results from the full sample will be presented at the meeting. Preliminary results indicate good app acceptability (mean uMARS scores of 3.93 to 4.36 out of 5) and feasibility (mean session completion rate of 9.5 out of 12 sessions at the end of the intervention period). Feedback at the end of the intervention was generally positive, with participants indicating that they were "likely" or "extremely likely" to continue using the app after the intervention period. Both anxiety and depression showed trend-level decreases from baseline to the end of the intervention.

Conclusions

These preliminary findings suggest that Maya, when modified for middle-aged and older adults with anxiety or depression, may be acceptable and feasible for use in the intended population and may decrease clinical symptoms. If our findings are replicated, Maya could represent a useful tool for improving access to care among middle-aged and older adults with anxiety or depression.