Predictors and Correlates of Depression and Anxiety Symptom Trajectories in a Large Digital Mental Health Provider: Retrospective Analysis of Data From Rula Health

Abstract

Background: Depression and anxiety are highly prevalent and burdensome, yet many individuals, especially those with subclinical symptoms, remain underserved by traditional care models. While digital mental health interventions (DMHIs) have improved access, few integrate high-frequency measurement-based care (MBC) or evaluate outcomes across the full spectrum of symptom severity in real-world settings.

Objective: The purpose of this study was to examine the effects of participation in a commercial MBC DMHI, Rula Health, on changes in depression and anxiety over time in both subclinical and clinical patients. We aimed to (1) explore the trajectories of anxiety and depression symptoms and (2) examine the impact of demographics and primary diagnosis on depression and anxiety trajectories.

Methods: We retrospectively analyzed longitudinal symptom data from adults receiving therapy through Rula Health, an MBC-based DMHI. Depression, via the Patient Health Questionnaire-9 (PHQ-9), and anxiety symptoms, via the Generalized Anxiety Disorder-7 (GAD-7), were measured before each visit over 12 therapy visits. Linear spline mixed-effects models with a knot placed at 5 visits (based on visual inspection) were used to evaluate symptom trajectories and identify moderators of treatment response, including demographic characteristics and primary diagnosis.

Results: A total of 365,741 adults (mean age 37.03, SD 11.81 years; 238,839/360,388, 66.27% female; 87,758/145,947, 60.13% White) with 2,685,103 therapy visits were included in the sample. Baseline depression (ie, PHQ-9) and anxiety (ie, GAD-7) scores averaged 9.41 (SD 6.61) and 9.45 (SD 5.65), respectively, decreasing to 6.37 (SD 5.83) and 6.50 (SD 5.01) within 12 visits. Depression (β=-0.72, P<.001) and anxiety (β=-0.72, P<.001) symptoms decreased significantly over the first 5 visits, and to a lesser degree over visits 6-12 (depression: β=-0.02, P<.001; anxiety: β=-0.0078, P=.004). Faster improvements over visits 1-5 occurred in younger patients (βPHQ=0.0031, P<.001; βGAD=0.0027, P<.001), and those identifying as Black/African American (βPHQ=-0.089, P<.001; βGAD =-0.042, P<.001), American Indian/Alaska Native (βPHQ=-0.14, P<.001; βGAD=-0.11, P<.001), and Native Hawaiian/Pacific Islander (βPHQ=-0.12, P<.001; βGAD=-0.069, P=.01). Patients with an anxiety (βGAD=-0.071, P<.001) or trauma-related (βGAD=-0.021, P=.03) disorder had faster improvements in GAD-7.

Conclusions: This study contributes to the growing evidence base from commercial DMHIs by demonstrating significant improvements in depression and anxiety symptoms across both clinical and subclinical populations using real-world data from a large, national provider. Symptom reductions were most rapid in the first 5 visits, with continued improvements through session 12, especially among historically underserved groups. These findings highlight Rula Health's ability to deliver early, sustained, and equitable outcomes through an MBC model.

Keywords: clinical outcomes; digital mental health interventions; measurement-based care; mental health; real-world evidence; teletherapy; web-based therapy.

Figure 1.. (A) Raw mean Patient Health Questionnaire-9 (PHQ-9) and (B) Generalized Anxiety Disorder-7 (GAD-7) score trajectories over 12 visits. 95% CIs were estimated but were too thin to visualize and report.