# Digital Mental Health Coaching in Clinically Diverse Populations: Controlled Engagement and Outcomes Study

## TL;DR

Digital mental health coaching improves outcomes for diverse, clinically severe populations compared to app-only use.

## Contribution

Demonstrates effectiveness of digital coaching in clinically severe and diverse samples, challenging prior assumptions.

## Key findings

- Coaching users showed greater symptom reduction in anxiety and stress compared to app-only users.
- Significant improvements in depression, anxiety, and stress were observed in coaching users with severe symptoms.
- Engagement with app content did not differ between groups, indicating coaching's independent impact.

## Abstract

Digital coaching programs, offering virtually delivered mental health care by coaches and companion apps, are an increasingly popular care model designed to increase accessibility and reduce strain on traditional mental health care systems. Initial studies suggest these programs can produce a range of positive mental health outcomes; however, methodological limitations and a focus on homogeneous, subclinical populations have constrained conclusions about their effectiveness, especially in diverse and clinically severe samples.

This study aimed to evaluate the impact of an evidence-based digital mental health coaching program in a clinically and demographically diverse sample. The study compared engagement with app-based content and changes in depression, anxiety, and stress symptoms, over the course of 1 month, among users who received coaching versus those who used the app alone (controls).

Program users (N=64) were categorized as coaching users (attending at least 1 session) or controls (app-only users). Depression, anxiety, and stress symptoms were assessed using the Depression Anxiety Stress Scale-21 at baseline and after 30 days. Engagement with app content was also measured. Between-group differences were analyzed using t tests and mixed multivariate analysis of covariance models, with follow-up sensitivity analysis of covariance analyses (controlling for age).

Participants were diverse in terms of demographics and clinical severity, with half reporting severe to extremely severe depression and nearly half reporting severe to extremely severe anxiety or stress at baseline. A repeated-measures multivariate analysis of covariance revealed a significant group-by-time interaction (P=.02), indicating greater symptom reduction among coaching users, primarily driven by changes in anxiety and stress. Follow-up analyses of covariance exploring symptom-specific patterns, excluding participants with subclinical baseline symptoms, yielded significant group-by-time interactions across depression (P=.04), anxiety (P=.003), and stress (P=.03). Engagement with app-based content did not significantly differ between the groups (P=.20), suggesting coaching’s effectiveness was not contingent on differential app usage.

This study demonstrates that digital mental health coaching can significantly improve clinical outcomes, even in diverse and clinically severe populations. These findings challenge the notion that coaching is only effective for subclinical or high-functioning individuals and highlight its potential to extend the reach of mental health care to underserved communities.