Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 83 of the 83 references it cites were found in a registry or an index; 30 of the 62 sources its claims cite were checked for support.

Clinical Efficacy and Psychological Mechanisms of an App-Based Digital Therapeutic for Generalized Anxiety Disorder: Randomized Controlled Trial

Open access
Roy, Alexandra, Hoge, Elizabeth A, Abrante, Pablo, Druker, Susan, Liu, Tao, Brewer, Judson A · 2021 · Journal of Medical Internet Research

Published version of a preprint

Abstract

Background: Current treatments for generalized anxiety disorder (GAD) often yield suboptimal outcomes, partly because of insufficient targeting of underlying psychological mechanisms (eg, avoidance reinforcement learning). Mindfulness training (MT) has shown efficacy for anxiety; yet, widespread adoption has been limited, partly because of the difficulty in scaling in-person–based delivery. Digital therapeutics are emerging as potentially viable treatments; however, very few have been empirically validated. Objective: The aim of this study is to test the efficacy and mechanism of an app-delivered MT that was designed to target a potential mechanism of anxiety (reinforcement learning), based on which previous studies have shown concern regarding feedback and the perpetuation of anxiety through negative reinforcement. Methods: Individuals with GAD were recruited using social media advertisements and randomized during an in-person visit to receive treatment as usual (n=33) or treatment as usual+app−delivered MT (Unwinding Anxiety; n=32). The latter was composed of 30 modules to be completed over a 2-month period. Associated changes in outcomes were assessed using self-report questionnaires 1 and 2 months after treatment initiation. Results: We randomized 65 participants in this study, and a modified intent-to-treat approach was used for analysis. The median number of modules completed by the MT group was 25.5 (IQR 17) out of 30; 46% (13/28) of the participants completed the program. In addition, the MT group demonstrated a significant reduction in anxiety (GAD-7) compared with the control group at 2 months (67% vs 14%; median change in GAD-7: –8.5 [IQR 6.5] vs –1.0 [IQR 5.0]; P<.001; 95% CI 6-10). Increases in mindfulness at 1 month (nonreactivity subscale) mediated decreases in worry at 2 months (Penn State Worry Questionnaire; P=.02) and decreases in worry at 1 month mediated reductions in anxiety at 2 months (P=.03). Conclusions: To our knowledge, this is the first report on the efficacy and mechanism of an app-delivered MT for GAD. These findings demonstrate the clinical efficacy of MT as a digital therapeutic for individuals with anxiety (number needed to treat=1.6). These results also link recent advances in our mechanistic understanding of anxiety with treatment development, showing that app-delivered MT targets key reinforcement learning pathways, resulting in tangible, clinically meaningful reductions in worry and anxiety. Evidence-based, mechanistically targeted digital therapeutics have the potential to improve health at a population level at a low cost. © 2021 Journal of Medical Internet Research. All rights reserved.

Overview

CoChat Summary

## TL;DR

In this randomized controlled trial of 65 adults with generalized anxiety disorder, an app-delivered mindfulness training program reduced anxiety more than treatment as usual and showed a very large number needed to treat (1.6). The paper also reports that increases in nonreactivity and decreases in worry statistically mediated the treatment’s effects.

## Key findings

  • Among 65 randomized participants, 61 completed the study and were included in the modified intent-to-treat analysis; the median number of modules completed was 25.5 out of 30, and 46% (13/28) completed the program.
  • At 2 months, the app-delivered mindfulness training group had greater anxiety reduction than treatment as usual: median GAD-7 change of –8.5 (IQR 6.5) vs –1.0 (IQR 5.0), with 67% vs 14% reduction and P<.001; the authors report NNT=1.6 and remission in 64% (18/28) vs 3% (1/33).
  • Mindfulness nonreactivity increased more in the intervention group at 2 months: median increase 7.5 (IQR 6) vs 1 (IQR 6) in controls, with P<.001.
  • Worry decreased more in the intervention group at 2 months: median reduction 15 (IQR 14.3) vs 3 (IQR 6) in controls, with P<.001; mediation analyses showed a significant indirect effect through nonreactivity (βindirect = –0.29; 95% CI –0.68 to –0.04; P=.02) and through worry for anxiety reduction (βindirect = –0.19; 95% CI –0.40 to –0.02; P=.03).

## Methods

  • Parallel-group randomized controlled trial with preregistered analyses and outcome measures; participants were recruited through Facebook/social media ads, screened by phone, and enrolled after an in-person visit at Brown University.
  • Sample: adults age 18+ with GAD-7 ≥10 and smartphone ownership; 65 were randomized to treatment as usual (n=33) or treatment as usual plus app-delivered mindfulness training (n=32).
  • Main analysis used a modified intent-to-treat approach and robust mixed ANOVA for primary outcomes, plus mediation analyses with bootstrap confidence intervals and nonparametric/robust tests.

## Limitations

  • The paper states that standard clinical care in the TAU condition was highly variable, and that future studies using an active comparator, longer follow-up, and multiple sites are needed.
  • The sample was 90% women (57/63), and the authors note that future sex-balanced studies are needed for generalizability.
  • The study assessed outcomes only up to 2 months after treatment initiation, so long-term effects are not established.
  • Evident from the text: participants and the project director were unblinded to group allocation, which the authors rate as medium risk of bias for blinding of participants and personnel.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.2196/26987

Cites

83 of 83 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 (A) Path model for longitudinal causal mediation evaluating if increases in nonreactivity mediate the relationship between MT and reduction in worry. (B) Path model for longitudinal causal mediation evaluating if reduction in worry mediates the relationship between MT and reduction in anxiety. MT: mindfulness training.
Figure 1 (A) Path model for longitudinal causal mediation evaluating if increases in nonreactivity mediate the relationship between MT and reduction in worry. (B) Path model for longitudinal causal mediation evaluating if reduction in worry mediates the relationship between MT and reduction in anxiety. MT: mindfulness training.CC BY · PubMed Central
Figure 2 Participant flow diagram. MT: mindfulness training; TAU: treatment as usual.
Figure 2 Participant flow diagram. MT: mindfulness training; TAU: treatment as usual.CC BY · PubMed Central
Figure 3 (A) Change in GAD-7 scores. (B) Change in nonreactivity scores. (C) Change in PSWQ scores. (D) Change in MAIA scores. GAD-7: Generalized Anxiety Disorder 7-item; MAIA: Multidimensional Assessment of Interoceptive Awareness; PSWQ: Penn State Worry Questionnaire.
Figure 3 (A) Change in GAD-7 scores. (B) Change in nonreactivity scores. (C) Change in PSWQ scores. (D) Change in MAIA scores. GAD-7: Generalized Anxiety Disorder 7-item; MAIA: Multidimensional Assessment of Interoceptive Awareness; PSWQ: Penn State Worry Questionnaire.CC BY · PubMed Central
Figure 4 (A) Longitudinal causal mediation model with standardized regression coefficients illustrates how an increase in nonreactivity mediates the effects of mindfulness training on reduction in worry. (B) Longitudinal causal mediation model with standardized regression coefficients illustrates how a reduction in worry mediates the effects of mindfulness training on reduction in anxiety. *P<.05; **P<.01; ***P<.001.
Figure 4 (A) Longitudinal causal mediation model with standardized regression coefficients illustrates how an increase in nonreactivity mediates the effects of mindfulness training on reduction in worry. (B) Longitudinal causal mediation model with standardized regression coefficients illustrates how a reduction in worry mediates the effects of mindfulness training on reduction in anxiety. *P<.05; **P<.01; ***P<.001.CC BY · PubMed Central

Verification

Claims the paper makes

25 Valid · 1 Not valid · 19 Not checked · 66 Own results and interpretations

40 more claims · 4 with their cited sources checked

Create a free account to see all 111 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 43
  • Probably not a substantial claim. 35
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

2 more claims about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

Citations

Cites

Research works printed but not linked: 3

  1. 2013, Diagnostic and Statistical Manual of Mental Disorders (5th Ed.)
  2. Brewer, JA, 2015, J Fam Med, 2, 1035
  3. Sheehan, DV, 1998, J Clin Psychiatry, 59 Suppl 20, 22

Cited by

See all 41

Metadata

Authors
Roy, A., Hoge, E. A., Abrante, P., Druker, S., Liu, T., & Brewer, J. A.
Affiliations
  • Brown University (US)
  • Georgetown University Medical Center (US)
  • University of Massachusetts Chan Medical School (US)
Corresponding author
Brewer, Judson A
Venue
Journal of Medical Internet Research
Published
2021-12-02
Volume
23
Issue
12
Pages
e26987
ISSN
1438-8871 (ISSN-L)
OpenAlex
W3203508624
Cited by
41
References
80
Citations by year
2026: 10 · 2025: 10 · 2024: 12 · 2023: 4 · 2022: 4 · 2021: 1
FWCI
4.66
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Journal of Medical Internet Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
National Institute of Mental Health (R41MH118130)
Topics
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Anxiety, Depression, Psychometrics, Treatment, Cognitive Processes
  • Social Sciences › Psychology › Clinical Psychology › Mindfulness and Compassion Interventions
Keywords
generalized anxiety disorder, digital therapeutics, mindfulness training, mobile health apps, reinforcement learning, avoidance, negative reinforcement, worry, non-reactivity, medication adherence, psychological mechanisms, GAD-7, Penn State Worry Questionnaire, treatment as usual comparison, number needed to treat
MeSH terms
Anxiety Disorders, Humans, Treatment Outcome, Mobile Applications

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.