A Multiple‑Baseline Evaluation of Acceptance and Commitment Therapy Focused on Repetitive Negative Thinking for Comorbid Generalized Anxiety Disorder and Depression

Complete Reference

Ruiz, F. J., Luciano, C., Flórez, C. L., Suárez-Falcón, J. C., & Cardona-Betancourt, V. (2020). Efficacy of a brief transdiagnostic acceptance and commitment therapy protocol for comorbid anxiety and depression: A single-case series study. Frontiers in Psychology, 11, 356. https://doi.org/10.3389/fpsyg.2020.00356

Study Type

Single-case experimental design with delayed multiple-baseline across participants (DMBD).

Context and Objectives

Comorbid generalized anxiety disorder (GAD) and depression are highly prevalent in clinical populations and are associated with worse therapeutic outcomes, greater functional disability, and reduced responsiveness to conventional treatments. Previous research suggests that brief interventions targeting this specific comorbidity have demonstrated limited efficacy. In this context, a transdiagnostic perspective has emerged identifying common psychological processes underlying both disorders, particularly repetitive negative thinking (RNT) and experiential avoidance. Acceptance and Commitment Therapy focused on RNT (ACT-RNT) offers an innovative approach grounded in Relational Frame Theory (RFT), conceptualizing psychological flexibility as the central mechanism of change.

The primary objective of this study was to evaluate the efficacy of a brief ACT protocol focused on RNT (three sessions) administered individually to adults with severe comorbid GAD and depression, using a delayed multiple-baseline design. The researchers hypothesized that this brief yet structured protocol would produce clinically significant reductions in emotional symptoms, pathological worry, and dysfunctional psychological processes (experiential avoidance, cognitive fusion, repetitive negative thinking), while increasing alignment with personal values.

Method

Participants

The sample consisted of 6 adults (4 women and 2 men; mean age = 31.7 years, SD = 11.5) recruited through social media in Bogotá, Colombia. All participants met simultaneous diagnostic criteria for both GAD and major depression, confirmed via the MINI (Mini International Neuropsychiatric Interview). Additionally, participants were required to score ≥10 on the Patient Health Questionnaire-9 (PHQ-9), indicating moderate to severe depression, and ≥10 on the Generalized Anxiety Disorder-7 (GAD-7), indicating moderate to severe generalized anxiety. These inclusion criteria ensured that the sample represented a clinical group with significant comorbidity and severe symptomatology requiring specialized intervention.

Design

A delayed multiple-baseline design (DMBD) across participants was employed, which is a rigorous variant of single-case experimental design (SCED) frequently used in clinical psychology research. This design allows for establishing causality through replication within cases versus across cases, controlling for time-related variables (maturation, historical events). Each participant had a baseline period of variable duration (minimum 5 data points) before receiving the intervention, thus permitting assessment of pre-intervention trends. Weekly assessments were conducted during the baseline and intervention phases, followed by biweekly assessments during a three-month follow-up period, providing considerable temporal density of data for within-participant analysis.

Intervention

A three-session ACT protocol focused on RNT was administered, consistent with prior research (Ruiz et al., 2019). The first session lasted approximately 90 minutes, while the second and third sessions lasted approximately 60 minutes each. The protocol is grounded in the Relational Frame Theory conceptualization of psychological flexibility as a transdiagnostic mechanism. Interventions included acceptance techniques, cognitive defusion techniques specifically targeting repetitive negative thinking, personal value clarification, and training in commitment to actions consistent with identified values. The protocol was administered by a single therapist trained in ACT, ensuring consistency in implementation but limiting generalization of findings regarding protocol efficacy versus therapist-specific characteristics.

Outcome and Process Measures

Primary outcome measure: The 21-item Depression, Anxiety, and Stress Scale (DASS-21) served as the primary outcome measure, providing assessments of general emotional symptoms including depression, anxiety, and stress. Both the DASS-Total and specific subscales were utilized.

Secondary measure: The abbreviated 11-item Penn State Worry Questionnaire (PSWQ-11) assessed pathological worry and rumination, constructs central to generalized anxiety disorder conceptualization.

Process measures (mechanisms of change): The 7-item Acceptance and Action Questionnaire (AAQ-II) assessed experiential avoidance; the 7-item Cognitive Fusion Questionnaire (CFQ) assessed cognitive fusion or thought literality; the 10-item Repetitive Thinking Questionnaire (PTQ) assessed repetitive negative thinking; and the Values Questionnaire (VQ) assessed alignment with personal values, with subscales for Progress (VQ-Progress) and Obstruction (VQ-Obstruction).

Statistical Analyses

The Theil-Sen slope was employed to evaluate baseline trends, allowing identification of pre-intervention changes that could confound interpretation of treatment effects. For within-participant effect analyses, Bayesian JZS+AR (Jarrett, Zero-inflated, Single-case + Autoregressive) testing was utilized, which provides Bayes factors (BF) to evaluate the strength of evidence favoring the hypothesis that change occurred between baseline and intervention phases. The authors employed the standard interpretation scale where BF > 10 indicates "substantial evidence" in favor of the alternative hypothesis. Clinically significant change (CSC) was calculated using the Jacobson and Truax approach, determining whether participants achieved statistically significant changes and transitioned to functional ranges. Design-comparable standardized mean difference effect sizes (d-statistic) were calculated to evaluate magnitude of change across cases at 12-week follow-up.

Results

Baseline analyses: Theil-Sen trend analysis revealed no significant pre-intervention trends on any primary measures, strengthening the interpretation that subsequent changes were attributable to the intervention rather than natural time-related variables.

Visual analysis and within-participant Bayesian analyses: Visual analysis of time-series graphs demonstrated that the protocol was effective in reducing DASS-21 and PSWQ-11 in all six participants. Bayesian analyses with JZS+AR revealed that five of six participants showed at least substantial evidence (BF > 10) for change in DASS-Total, and all six participants showed substantial to extreme evidence for change in PSWQ-11. Regarding mechanisms of change, all six participants showed evidence in the expected direction for AAQ-II (experiential avoidance), CFQ (cognitive fusion), and PTQ (repetitive negative thinking).

Clinically significant change: Five of six participants demonstrated clinically significant change in DASS-Total and PSWQ-11. Regarding process measures, five of six participants showed CSC in CFQ, while four of six demonstrated CSC in AAQ-II. For values-based living (VQ), two to three of six participants showed CSC in Progress and Obstruction subscales, suggesting that while symptoms and cognitive fusion improved substantially, change in values-consistent behavior was more variable.

Effect sizes: Design-comparable standardized mean difference effect sizes at 12-week follow-up were very large. For the primary outcome measure (DASS), the following were observed: DASS-Total d = 3.34, DASS-Depression d = 2.37, DASS-Anxiety d = 1.90, and DASS-Stress d = 3.49. For the secondary measure, PSWQ-11 demonstrated an extraordinarily large effect size (d = 4.52). For process measures: AAQ-II d = 3.46, CFQ d = 3.90, PTQ d = 4.52, VQ-Progress d = 0.92, and VQ-Obstruction d = 1.98. All these effect sizes are considered very large or extremely large according to clinical psychology research standards (d > 0.80 is considered large; d > 2.0 is considered very large).

Three-month follow-up: Therapeutic effects were maintained or improved at three-month follow-up, with most participants continuing their improvement trajectory, suggesting that changes were not transitory but represented sustained change.

Discussion and Conclusions

The findings of this study provide robust additional evidence for the efficacy of brief ACT protocols focused on RNT for comorbid GAD and depression. The observed effect sizes were consistent with prior single-case experimental design (SCED) studies evaluating similar interventions, suggesting replicability and potential generalization of effects. The intervention was notably brief (only three sessions, approximately 4.5 hours of total therapeutic contact), which is especially important given mental health costs and treatment accessibility in resource-limited contexts such as primary care or middle-income countries.

The authors discuss that change in general emotional symptoms (DASS) and pathological worry (PSWQ) was robust in nearly all participants, aligning with the transdiagnostic hypothesis that RNT and experiential avoidance are common processes underlying both disorders. Particularly notable was that change in cognitive fusion (CFQ) was among the largest observed, suggesting that defusion specifically targeting negative repetitive thinking patterns may be a central mechanism of change.

However, change in values-consistent living (VQ) was more variable, with only 2-3 participants demonstrating clinically significant change. This suggests that while the protocol effectively reduces symptoms and process symptoms, consolidation of values-oriented behavioral changes may require additional intervention or extended time. The authors suggest this is consistent with the perspective that acceptance and defusion are more immediate changes, while behavioral commitment to values may develop more slowly.

The authors conclude that these findings provide empirical justification for conducting randomized controlled trials (RCTs) comparing this brief ACT protocol focused on RNT with active control conditions (such as treatment-as-usual, standard cognitive-behavioral therapy, or non-specific psychological attention). Such comparison is necessary to determine whether observed effects are due to specific ACT-RNT components or non-specific factors such as therapeutic attention, change expectations, or alerting variables.

Recognized Limitations

The authors identify several important limitations that circumscribe generalization of findings: (1) measurement was exclusively via self-report, without objective behavioral observations; (2) complete diagnostic interview was not conducted at follow-up, limiting capacity to confirm diagnostic remission; (3) the delayed multiple-baseline design, while rigorous, has limitations relative to true RCTs with control groups; (4) the protocol was administered by a single therapist, not allowing dissociation of treatment effects from therapist-specific characteristics; (5) mediation analyses were not conducted, limiting capacity to understand precisely which mechanisms of change were responsible for observed effects.

Significance and Contribution

This study makes significant contributions to the literature on transdiagnostic treatments for comorbid anxiety and depression. Its primary contribution is demonstrating that a very brief protocol (only three sessions) can produce very large changes in symptoms and transdiagnostic psychological processes. This has important practical implications for resource-limited contexts, primary care, and public health. Additionally, the study provides empirical evidence supporting subsequent research via RCTs, which is the necessary methodological next step in intervention development. The research also adds to the body of literature supporting Relational Frame Theory's conceptualization of transdiagnostic processes and their clinical application in ACT.

Empirical Verification Checklist

Fecha de resumen/Summary date: 27 de marzo de 2026 / March 27, 2026

Resumen preparado por/Summary prepared by: Claude (Anthropic)