An idiographic comparison of ACT focused on repetitive negative thinking versus non-directive therapy in child generalized anxiety

Title

An Idiographic Comparison of Acceptance and Commitment Therapy Focused on Repetitive Negative Thinking versus Non-directive Therapy in Child Generalized Anxiety

Complete Reference

Barajas, H. N., Farfán, N., Rodríguez, L. J., Peña-Tomas, B., Larrea-Rivera, Y. A., Gutiérrez, Y., Iriarte-Becerra, S., Rodríguez, P. A., Moreno-Méndez, J. H., & Ruiz, F. J. (2025). An idiographic comparison of ACT focused on repetitive negative thinking versus non-directive therapy in child generalized anxiety. Journal of Contextual Behavioral Science, 38, 100944.

Study Type

Empirical idiographic study: randomized two-sample single-case experimental design (SCED) with controlled comparison of two brief psychological interventions delivered online.

BACKGROUND AND OBJECTIVES

Anxiety is the most common mental health disorder in childhood. Generalized Anxiety Disorder (GAD) is particularly prevalent in the child population, characterized by excessive, persistent, and uncontrollable worry occurring for at least 6 months. In many Latin American countries, there is a scarcity of studies examining GAD prevalence separately, with anxiety recognized as a relevant mental health concern in childhood.

Repetitive Negative Thinking (RNT) is a central feature of anxiety and depressive disorders in children. Although Cognitive-Behavioral Therapy (CBT) has demonstrated efficacy in treating childhood GAD, there is significant variability in remission rates and a substantial proportion of children do not respond adequately. Additionally, in low- and middle-income countries, access to specialized psychological treatments is limited.

Acceptance and Commitment Therapy (ACT) is a practical application of Contextual Behavioral Science that has expanded its empirical support over the past decade. ACT seeks to promote psychological flexibility through six core processes. It has been proposed that a version of ACT integrating Relational Frame Theory (RFT) and focused specifically on RNT could be particularly effective in addressing RNT in children with GAD.

Non-directive Therapy (NDT) is frequently used as a psychological placebo control in clinical trials. It is an unstructured intervention allowing the child to lead the process, with the therapist maintaining an empathic stance. Although some studies suggest NDT can be effective in treating emotional problems, evidence regarding its effectiveness is mixed.

The objective of this study was to compare the efficacy of a brief RNT-focused ACT protocol versus an NDT intervention in Colombian children with GAD using a randomized two-sample SCED. It was hypothesized that the RNT-focused ACT intervention would be more effective than NDT in reducing anxiety symptoms in childhood.

METHOD

Participants

A total of 232 potential participants were recruited through public announcement on social media. Inclusion criteria were: (a) age 8-12 years, (b) perceived excessive worry difficulties, (c) principal diagnosis of GAD based on the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID). Exclusion criteria included: (a) psychotic psychopathology or other severe psychotic disorder, (b) high suicide risk.

Of the 232 assessed, 214 were excluded for not meeting inclusion criteria (n=155), not responding to phone calls (n=43), not maintaining interest (n=9), or not attending the interview (n=8). Eighteen participants were randomized. Seventeen participants were included in the final analysis: 9 assigned to RNT-focused ACT (8 received allocated intervention; 1 withdrew during baseline) and 9 assigned to NDT (all received allocated intervention). In analyses: n=8 in ACT and n=9 in NDT. Of the 17 participants, 12 were girls with a mean age of 9.41 years (SD = 1.27).

During the initial MINI-KID assessment, 12 participants were identified as presenting comorbid conditions, including other anxiety disorders, depression, Attention-Deficit/Hyperactivity Disorder (ADHD), Obsessive-Compulsive Disorder (OCD), and Oppositional Defiant Disorder (ODD).

Design

The study employed a randomized two-sample single-case experimental design (SCED). Participants had a 4-week baseline period before being randomly assigned to either RNT-focused ACT or NDT intervention. Random assignment was 1:1 using the Research Randomizer web-based tool.

The independent variable was the type of intervention received by participants. The primary dependent variable was emotional symptoms (as reported by participants), and pathological worry was defined as the process/mediator variable. Multiple outcome measures were recorded every 3 days throughout the study.

Intervention

RNT-focused ACT Protocol: The protocol designed by Salazar et al. (2020) was adapted for delivery via videoconference. The ACT protocol consisted of three 45-minute sessions conducted via videoconference, with one session scheduled each week.

Session 1 aimed to: (a) discriminate between psychological inflexibility and psychological flexibility; (b) recognize psychologically flexible and inflexible behaviors in daily life; (c) establish commitments for the next session. A projected image was used allowing the child to develop the ability to identify all emotional symptoms, memories, and thoughts they might choose. Different examples of children exhibiting "wise king" and "slave" behaviors were presented.

Session 2 was dedicated to discriminating against RNT and its effect on valued action. Hierarchical frame training was used presenting private events and providing regulatory functions. A "spinning around" exercise was employed where the therapist presented an important event and the child chose to behave as a "wise king" or "slave."

Session 3 introduced experiential exercises to identify RNT triggers and practices to distance oneself from them. A fantasizing and worrying exercise was used where the child was asked to fantasize about something pleasant, followed by a sound, and asked to notice that it was a thought.

NDT Protocol: The objective of the NDT protocol was to establish space for active, reflective, and empathic dialogue with the participant. The protocol consisted of three sessions of approximately 45 minutes each. During these sessions, the therapist redefined problem-solving by suggesting new behaviors, assuming the role of expert, asking open-ended questions about the child's program life. The therapist engaged in open dialogue with the participant, refraining from making judgments or offering advice regarding the child's concerns.

Measurement Instruments

MINI-KID (Mini International Neuropsychiatric Interview for Children and Adolescents): Structured diagnostic interview designed to assess primary psychiatric disorders based on DSM-IV-TR and ICD-10 criteria.

Intensive measures of emotional symptoms: Three items were selected from each emotional symptom type measured by the Anxiety-Depression Scale for Children (ADSC), Child Depression Scale (CDS), and Depression Symptom Index (DSI). Participants responded to these original ADSC/CDS items in slightly modified version (to report child symptoms). All were responded to on a 5-point Likert-type scale.

Intensive measures of worry: Three items were selected from the Penn Worry Questionnaire for Children (PWQC), based on applicability and relevance for intensive measurement. The PWQC was utilized by Barrios & Ruiz (2025) and showed good psychometric properties. Items were slightly modified to report child symptoms. All were responded to on a 5-point Likert-type scale.

Data collection: An in-path procedure was conducted where questionnaires were sent to parents daily via email every 3 days. If the questionnaire was not completed on the scheduled day, a reminder was sent 30 minutes later. Measurements were collected every 3 days throughout the study.

Data Analysis

Longitudinal factorial invariance of intensive measures: The longitudinal factorial invariance of emotional symptom and pathological worry measures, reported by children and parents, was first assessed. This step was necessary to ensure that the primary outcome (dependent variable) was measured reliably throughout the study.

Visual trend analysis and selection of statistical analyses: Statistical tests appropriate for the observed data patterns were selected. The trend analysis confirmed there were no statistically significant trends in the baselines of RNT-focused ACT and NDT participants. Visual stability of scores was examined; some exhibited greater variability than others.

Within-case and between-case analyses: A non-overlapping Tau-U analysis was conducted as a within-case approach for SCED analysis. The Tau-U method provides both an effect size and a p-value. DC-SMD (Design-Comparable Standardized Mean Difference) values were computed as an overall estimate of intervention effect size for each condition. Visual adjustments were made to analyses for the statistical analyses representing trends observed during intervention phases.

Baseline trend analysis: The visual approach suggested by Parker et al. (2011) using segmented regression models was employed. A trend was considered to exist if the 95% confidence interval did not overlap.

Within-subject mediation analysis: A within-subject mediation analysis suggested by Vuorre & Bolger (2018) using the R package bmm was conducted to examine whether the change in pathological worry mediated the effect of intervention on emotional symptoms.

RESULTS

Longitudinal Factorial Invariance of Intensive Measures

Table 3 shows that all measures demonstrated longitudinal factorial invariance, as indicated by Comparative Fit Indices (CFI) and Tucker-Lewis Indices (TLI) exceeding 0.05, and Root Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Square Residual (SRMR) below 0.05. All values were above 0.80, demonstrating satisfactory psychometric consistency.

Trend Analysis

Baseline trends: Analysis confirmed there were no statistically significant trends in the baselines of RNT-focused ACT condition participants. However, some NDT participants exhibited slight improvement trends. Specifically, P7 (Tau-U = −0.60, p < .05) and P19 (Tau-U = −0.65, p < .001) showed improvement in emotional symptoms. Additionally, P2 (Tau-U = −0.63, p < .05) and P9 (Tau-U = −0.78, p < .001) showed improvement trends in worry as reported by children. P15 (Tau-U = −0.59, p < .01) also demonstrated improvement trends in emotional symptoms.

Intervention trends: Regarding the RNT-focused ACT condition, most participants showed significant improvement trends. P1 (Tau-U = −0.43, p < .05), P5 (Tau-U = −0.45, p < .001), P10 (Tau-U = −0.45, p < .001), and P17 (Tau-U = −0.63, p < .001) exhibited improvement trends in worry. Additionally, P1 (Tau-U = −0.50, p < .05) and P10 (Tau-U = −0.30, p < .05) demonstrated improvement trends in emotional symptoms as reported by parents. Furthermore, P5 (Tau-U = −0.37, p < .05) and P10 (Tau-U = −0.35, p < .05) and P17 (Tau-U = −0.55, p < .001) demonstrated improvement trends in worry, while P5 (Tau-U = −0.60, p < .001) exhibited deterioration trends in emotional symptoms as reported by parents. Both P16 (Tau-U = 0.38, p < .05) and P11 (Tau-U = −0.53, p < .001) experienced a deterioration trend in self-reported worry.

In the NDT condition, four participants showed consistent improvement (P3, P7, P12, P16), while two showed contradictory results regarding changes in emotional symptoms and pathological worry. P3 showed contradictory changes: reported increased symptoms while parents reported decreased symptoms. P7 showed similar contradictory results, with parents noting increased symptoms while the child reported decreased symptoms.

Within-Case Analysis

Emotional symptoms: Figure 2 shows emotional symptom scores in the RNT-focused ACT condition, as reported by children and parents. Table 4 indicates that, according to Tau-U vs. B-trend, all participants in the ACT condition experienced statistically significant improvements in emotional symptoms according to children's own reports (P5, P6, P8, P10, P17). Additionally, six participants showed changes according to parent observations (P1, P5, P8, P10, P12, P17). Table 4 also shows the evolution of emotional symptoms in NDT conditions.

Figure 2 also shows the evolution of emotional symptoms in the NDT condition. Two participants (P3 and P7) and three participants (P8, P17, P16) demonstrated improvements in emotional symptoms as reported by children and parents, respectively. Additionally, according to parent reports, Table 5 showed contradictory changes with one participant (P7) showing increased emotional symptoms while parents reported decreased symptoms.

Pathological worry: Figure 4 shows pathological worry scores in the RNT-focused ACT condition. According to children's reports, five participants experienced statistically significant decreases in worry (P5, P6, P8, P10, P17) (see Table 5). Again, P12 was the only participant who did not show a clear trend, likely due to consistently low worry throughout the study. P1 showed contradictory results; reported increased symptoms while parents reported decreased symptoms.

Figure 4 also shows the evolution of pathological worry in the NDT condition. P1 demonstrated contradictory results; reported increased emotional symptoms while parents observed decreased symptoms. P16 reported increased symptoms while parents reported increased emotional symptoms.

Summary: Figure 6 provides a summary of within-case results. Differences between conditions are evident. According to children's reports, there is a 40.3% difference in intervention efficacy in reducing emotional symptoms and pathological worry, favoring the RNT-focused ACT condition (62.5% vs. 22.2% across conditions). According to parent reports, differences were higher: 75% vs. 33.3% in emotional symptoms and 75% vs. 11.1% in pathological worry.

Between-Case Analysis

Table 5 shows weighted Tau-U vs. B-trend and DC-SMD values. All weighted effect sizes were statistically significant for the RNT-focused ACT condition, compared with DC-SMD for children's emotional symptoms (g = −0.64; worry: g = −0.50) and very large for parents (emotional symptoms: g = −1.02; worry: g = −1.03). Effect sizes were statistically significant for NDT, but effect sizes were small (emotional symptoms: g = −0.36; worry: g = −0.42). However, effect sizes were not statistically significant for children's variables, as reported by children, with insignificant DC-SMD values (emotional symptoms: g = −0.09; worry: g = −0.05).

Indeed, 83.4% of confidence intervals did not overlap for any weighted Tau-U vs. B-trend value, indicating that the RNT-focused ACT condition demonstrated significantly higher efficacy than the NDT condition. Regarding DC-SMD, there were statistically significant differences between conditions on all variables except worry as reported by children.

Mediation Analysis

Given the small effect of the NDT protocol, mediation analysis was conducted only for the RNT-focused ACT intervention. Table 6 shows estimates of indirect effects, confidence intervals, and percentage of total effect. Regarding children's reports, change in pathological worry mediated the effect of intervention on emotional symptoms for P1, P5, P17, and P18. However, mediation of pathological worry on emotional symptoms for P1 was non-significant because the change experienced increases in both worry and emotional symptoms. The overall indirect effect was significant, although the percentage of total effect mediated by change in worry was 57%. In contrast, worry reduction mediated the effect of intervention on emotional symptoms for all participants, as reported by parents. The overall indirect effect was significant, accounting for 64% of the total intervention effect.

DISCUSSION AND CONCLUSIONS

This study compared the efficacy of a brief RNT-focused ACT intervention versus an NDT intervention in treating childhood GAD. The comparison was conducted using SCED, allowing conclusions to be reached at both the individual case level and the group level.

Results indicated that the majority of participants (87.5%) in the RNT-focused ACT condition experienced significant improvements in emotional symptoms and pathological worry, as reported by children and primary caregivers. P12 showed mixed results; the child reported significant improvement, but the mother reported minimal effect on emotional symptoms. Although the child had been stressed due to competitions in the past, it did not affect their performance or well-being as it had previously. The mother reported a reduction in interference in the child's life.

In contrast, fewer than half of the participants (44.4%) in the NDT condition showed statistically significant improvements in emotional symptoms or pathological worry as reported by the child or their primary caregivers. Changes were less clear than in the RNT-focused ACT condition. Only one participant (P3) showed significant and coherent changes across most variables.

An important finding of this study was the mediation analysis conducted at the individual level. According to the RNT-focused ACT model, reduction in worry should be a key process in this intervention for improving emotional symptoms. This was partially confirmed. For parent reports across all participants, however, it was a significant mediator for only three children according to their reports. This discrepancy may be attributed to greater reliability in parent reports.

An important contribution of this study was the mediation analysis conducted at the individual level. According to the RNT-focused ACT model, reduction in worry should be a key process in this intervention for improving emotional symptoms. This was partially confirmed. For parent reports across all participants, however, it was a significant mediator for only three children according to their reports. This discrepancy may be attributed to greater reliability in parent reports.

Group-level analyses revealed that the RNT-focused ACT intervention was significantly more effective than NDT across all variables. It is important to note that NDT is often used as a form of psychological placebo in previous ACT-focused RNT studies where interventions were conducted in conditions that did not control for non-specific components of psychotherapy (Larrea et al., 2025; Salazar et al., 2020). This present study provides evidence that the principles outlined hypothesized that brief RNT-focused ACT interventions could be effective for child depression as a psychological placebo intervention.

An important finding of this study was that the RNT-focused ACT intervention was more effective than other types of psychological interventions. The results of this study do not support NDT as a particularly effective intervention, at least in brief form. This is especially relevant in view of the fact that CBT has not demonstrated greater effectiveness than psychological placebo interventions in interventions with adults with depression (Cuijpers et al., 2012, 2024). Consequently, this is the first empirical study demonstrating that RNT-focused ACT interventions could be more effective than other types of psychological interventions. The results of this study do not support NDT as a particularly effective intervention, at least in brief form.

An important contribution was that brief RNT-focused ACT interventions were successfully implemented via videoconference. This is significant because this type of delivery modality can increase accessibility and reduce access costs and save time for caregivers. Considering that access to therapies in low- and middle-income countries is limited, it is important to develop psychological interventions that are effective, brief, and accessible. This approach to creating accessible interventions via videoconference is an advantage over in-person therapies.

Some limitations of the present study are worth noting. First, we did not control the effect of the RNT-focused ACT intervention in the NDT condition. Future studies should analyze this effect similarly using the study conducted by Riley & Gaynor (2014). Second, the sample size is relatively small; SCED favored conducting idiographic analyses. Third, although within-subject mediation analysis was conducted, future studies could analyze this effect in more depth, including possible additional mediators not measured, such as psychological flexibility. Finally, we only measured changes in one possible process mechanism of RNT-focused ACT interventions (worry). Future studies could include measures of additional mediators such as psychological flexibility.

In conclusion, this study offers promising evidence for the efficacy of brief RNT-focused ACT interventions delivered via videoconference in treating childhood GAD. The intervention demonstrated superior therapeutic benefits in relation to NDT, as reported by children and their primary caregivers. Consequently, it is concluded that brief RNT-focused ACT interventions are a promising and effective treatment for childhood GAD, providing significant evidence that their therapeutic benefits exceed those of the non-specific psychological placebo intervention.

SIGNIFICANCE AND CONTRIBUTION

This study provides empirical evidence for the efficacy of a brief acceptance and commitment therapy intervention for childhood generalized anxiety disorder. The research demonstrates that specific ACT processes—cognitive defusion, acceptance of private events, and commitment to values—can be effectively taught through brief experiential exercises delivered via videoconference. The results contribute to the field by showing that a three-session protocol can produce clinically meaningful improvements in anxiety symptoms, with effects maintained at follow-up. The mediation analysis indicates that reduction in pathological worry operates as a mechanism of change in anxiety reduction. The study also demonstrates the accessibility and viability of delivering ACT interventions via videoconference, addressing an important gap for populations with limited access to specialized services.