Meta-Analysis Finds High Uptake but Moderate Adherence in Clinical Trials of Depression and Anxiety Apps

Claudia Liu, PhD

Credit: LinkedIn

A new systematic review and meta-analysis of 79 randomized clinical trials (RCTs) reports that mental health smartphone apps for depression and anxiety demonstrate high rates of initial uptake but only moderate adherence, with substantial variability in attrition across studies. Findings, published in JAMA Psychiatry, also identify design and trial features associated with improved user retention.

The analysis included 92 app conditions and 78 control conditions spanning trials of depression, generalized anxiety, mixed presentations, social anxiety, panic disorder, and specific phobias. Investigators evaluated uptake (initial activation), adherence (meeting the trial’s predefined engagement threshold), and attrition at posttest and follow-up.

High Uptake Across Trials but Broad Variability

Across 56 app conditions, the meta-analytic uptake rate—defined as log-in or activation at least once—was 92.4% (95% prediction interval [PI], 48.4%-99.4%). Uptake estimates were similar when restricted to lower–risk-of-bias trials and when outliers were excluded. Uptake by condition-specific subgroups ranged from 89.5% in specific phobia trials to 96.6% in generalized anxiety trials.

Adherence Reached 61.8% Across Trials Reporting Data

Only 20 trials reported adherence outcomes, most commonly defined as completion of all required modules. The pooled adherence rate across 21 app conditions was 61.8% (95% PI, 53.2%-69.7%). When limited to module-completion definitions, adherence was 58.7% (95% PI, 14.6%-92.2%). Condition-specific adherence ranged from 55.1% (general anxiety) to 66.7% (specific phobia).

Exploratory stratified analyses indicated descriptively higher adherence in trials that included personalization features (69.3%), symptom monitoring tools (66.9%), recommended-use instructions (67.4%), or cognitive behavioral therapy (CBT)-based content (66.5%).

Posttest Attrition Averaged 18.6%; Human Contact Associated With Lower Dropout

From 95 app conditions, the pooled posttest attrition rate was 18.6% (95% PI, 2.4%-68.3%). At follow-up, attrition increased to 28.4% (95% PI, 3.0%-83.7%). Disorder-specific posttest attrition ranged from 10.9% in specific phobia trials to 19.5% in mixed depression–anxiety samples.

Subgroup analyses identified lower attrition in trials that included human contact—for example, guided onboarding or interaction with study staff—compared with those in which participants enrolled independently (14.8% vs 35.2%; P<.001).

Apps with symptom monitoring features showed modestly lower attrition (16.3%) compared with apps without such features (21.5%), although this difference was not statistically significant. Trials offering engagement reminders exhibited somewhat lower attrition than those without reminders (16.3% vs 24.2%).

Attrition Higher for Apps vs Wait-list Controls but Not Other Comparators

In 36 direct comparisons, attrition was higher in app conditions than in wait-list controls (risk ratio, 1.49; 95% PI, 0.34-6.48). This pattern remained significant in several disorder-specific comparisons, including generalized anxiety and social anxiety, although sample sizes were small. No significant differences in attrition emerged when apps were compared with usual care, placebo controls, or active therapies.

Findings Support Refinement of App Features and Trial Protocols

The authors note that typical engagement patterns in controlled trials may not generalize to real-world use, where uptake can be substantially lower. Nonetheless, the benchmarks generated in this review may guide future trial design, sample size planning, and intervention development.

The study also highlights the need for standardized reporting of adherence metrics and more consistent definitions of app engagement to support meaningful synthesis across future digital mental health trials.

Reference: Liu C, Torous J, Fuller-Tyszkiewicz M, et al. Uptake, Adherence, and Attrition in Clinical Trials of Depression and Anxiety Apps: A Systematic Review and Meta-Analysis. JAMA Psychiatry. Published online November 19, 2025. doi:10.1001/jamapsychiatry.2025.3439