CoChat

Behavioral Mechanisms That Mediate Mental and Physical Health Improvements in People With Chronic Pain Who Receive a Digital Health Intervention: Prospective Cohort Pilot Study

Open access

Published version of a preprint

The DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index; 14 of the 21 sources its claims cite were checked for support.

Abstract

Background: Preliminary evidence suggests that digital mental health intervention (Wysa for Chronic Pain) can improve mental and physical health in people with chronic musculoskeletal pain and coexisting symptoms of depression or anxiety. However, the behavioral mechanisms through which this intervention acts are not fully understood. Objective: The purpose of this study was to identify behavioral mechanisms that may mediate changes in mental and physical health associated with use of Wysa for Chronic Pain during orthopedic management of chronic musculoskeletal pain. We hypothesized that improved behavioral activation, pain acceptance, and sleep quality mediate improvements in self-reported mental and physical health. Methods: In this prospective cohort, pilot mediation analysis, adults with chronic (≥3 months) neck or back pain received the Wysa for Chronic Pain digital intervention, which uses a conversational agent and text-based access to human counselors to deliver cognitive behavioral therapy and related therapeutic content. Patient-reported outcomes and proposed mediators were collected at baseline and 1 month. The exposure of interest was participants’ engagement (ie, total interactions) with the digital intervention. Proposed mediators were assessed using the Behavioral Activation for Depression Scale–Short Form, Chronic Pain Acceptance Questionnaire, and Athens Insomnia Scale. Outcomes included Patient-Reported Outcomes Measurement Information System Anxiety, Depression, Pain Interference, and Physical Function scores. A mediation analysis was conducted using the Baron and Kenny method, adjusting for age, sex, and baseline mediators and outcome values. P<.20 was considered significant for this pilot study. Results: Among 30 patients (mean age 59, SD 14, years; 21 [70%] female), the mediation effect of behavioral activation on the relationship between increased intervention engagement and improved anxiety symptoms met predefined statistical significance thresholds (indirect effect –0.4, 80% CI –0.7 to –0.1; P=.13, 45% of the total effect). The direction of mediation effect was generally consistent with our hypothesis for all other proposed mediator or outcome relationships, as well. © 2023 The Author(s).

Overview

CoChat Summary

## TL;DR

This prospective cohort pilot study found preliminary evidence that behavioral activation may partially mediate the relationship between engagement with Wysa for Chronic Pain and improved anxiety symptoms in adults with chronic neck or back pain. Other proposed mediators—pain acceptance and sleep quality—showed directionally consistent but weaker evidence.

## Key findings

  • Among 30 patients (mean age 59, SD 14; 21/30 [70%] female), 28/30 (93%) used the digital intervention at least once during the 1-month study period, and 9/30 (30%) used the human coach at least once.
  • The mediation effect of behavioral activation on the relationship between intervention engagement and anxiety met the study’s predefined threshold (indirect effect –0.4; 80% CI –0.7 to –0.1; P=.13), and it accounted for 45% of the total effect.
  • In the full cohort, the largest mean change among proposed mediators was in behavioral activation (BADS-SF mean change 4.5 points; 80% CI 2.9 to 6.1).
  • Clinically meaningful average improvements were observed for PROMIS Anxiety (–3.2; 80% CI –4.3 to –2.0) and PROMIS Pain Interference (–3.5; 80% CI –4.8 to –2.2).

## Methods

  • Design: Single-site, single-arm, prospective cohort pilot study conducted at a tertiary care academic medical center in the United States; participants were enrolled from January 2022 through May 2022.
  • Sample/data: Adults (18+ years) with chronic (≥3 months) neck or back pain presenting for orthopedic spine care; 30/85 (35%) eligible patients enrolled, and all 30/30 (100%) completed follow-up.
  • Analysis: Mediation analysis using the Baron and Kenny method, with intervention engagement (total interactions over 1 month; log transformed) as the exposure; models adjusted for age, sex, baseline mediators, and baseline outcome values. The study used P<.20 and 80% CIs as the pilot significance threshold.

## Limitations

  • The paper states the study was limited by small sample size, lack of a comparison arm, short 1-month follow-up, and minimal reported depression symptoms at baseline.
  • The authors note the study was not powered to detect prespecified mediation effects, and the multiple comparisons produced large CIs.
  • The paper also states that mediation and outcome measures were collected at only 2 time points, so it was not possible to assess whether mediator changes preceded outcome changes.

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

Integrity signals

Registry

Crossref · 10.2196/51422

Cites

50 of 50 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Inclusion flowsheet of patients who sought orthopedic care for chronic neck or back pain between January 2022 and May 2022.
Figure 1 Inclusion flowsheet of patients who sought orthopedic care for chronic neck or back pain between January 2022 and May 2022.CC BY · PubMed Central
Figure 2 Mediation effect of behavioral activation on the relationship between intervention engagement and anxiety symptoms over a 1-month time interval among 30 patients who presented to an orthopedic clinic for management of chronic neck or back pain between January 2022 and May 2022. “a” represents the slope estimate between the exposure (intervention engagement) and proposed mediator (behavioral activation). “b” represents the slope estimate of the proposed mediator (behavioral activation) in the full model that includes the exposure (intervention engagement) and all covariates. Parentheses depict 80% CIs. Covariates include age, sex, and baseline BADS-SF and PROMIS Anxiety scores. c’: total direct effect; c: total effect; BADS-SF: Behavioral Activation for Depression Scale–Short Form; PROMIS: Patient-Reported Outcomes Measurement Information System.
Figure 2 Mediation effect of behavioral activation on the relationship between intervention engagement and anxiety symptoms over a 1-month time interval among 30 patients who presented to an orthopedic clinic for management of chronic neck or back pain between January 2022 and May 2022. “a” represents the slope estimate between the exposure (intervention engagement) and proposed mediator (behavioral activation). “b” represents the slope estimate of the proposed mediator (behavioral activation) in the full model that includes the exposure (intervention engagement) and all covariates. Parentheses depict 80% CIs. Covariates include age, sex, and baseline BADS-SF and PROMIS Anxiety scores. c’: total direct effect; c: total effect; BADS-SF: Behavioral Activation for Depression Scale–Short Form; PROMIS: Patient-Reported Outcomes Measurement Information System.CC BY · PubMed Central

Verification

Claims the paper makes

7 Valid · 0 Not valid · 0 Not checked · 32 Own results and interpretations

14 more claims

Create a free account to see all 39 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 10
  • Probably not a substantial claim. 11

Create a free account to see these claims.

Citations

Cites

Show all 50 references

Research works printed but not linked: 1

  1. Imai, H, 2020, Innov Clin Neurosci, 17, 21

Cited by

See all 10

Metadata

Authors
Cheng, A. L., Agarwal, M., Armbrecht, M. A., Abraham, J., Calfee, R. P., & Goss, C. W.
Affiliations
Washington University in St. Louis (US)
Corresponding author
Cheng, Abby L
Venue
JMIR Formative Research
Published
2023-11-17
Volume
7
Pages
e51422
ISSN
2561-326X (ISSN-L)
OpenAlex
W4388788550
Cited by
10
References
48
Citations by year
2026: 4 · 2025: 6
FWCI
2.58
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • JMIR Formative Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Digital Commons@Becker (Washington University School of Medicine) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Institutes of Health (UL1TR002345)
  • National Institute of Mental Health (P50MH122351)
  • National Cancer Institute (P30CA091842)
  • National Institute of Arthritis and Musculoskeletal and Skin Diseases (K23AR074520)
  • National Center for Advancing Translational Sciences
Topics
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Health Sciences › Health Professions › Physical Therapy, Sports Therapy and Rehabilitation › Musculoskeletal pain and rehabilitation
  • Health Sciences › Health Professions › General Health Professions › Mobile Health and mHealth Applications
Keywords
chronic musculoskeletal pain, digital mental health interventions, behavioral activation, pain acceptance, sleep quality, anxiety, depression, pain interference, physical function, mediation analysis, chatbots, cognitive behavioral therapy, neck pain, back pain, orthopedic management, intervention engagement, insomnia, patient-reported outcomes, self-compassion

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.