Testing the biobehavioral regulation of negative emotion as a mechanism of change in transdiagnostic youth psychotherapy: study protocol for a randomized controlled trial Authors: Madison Aitken, Brendan F. Andrade, Claire E. Krasinkiewicz, Sabrina W. S. Chan, Sera P. Gandhi, Jessica W. Pun, Andrew Hamilton-Wright, Wei Wang, Leah Sack, John R. Weisz, Robert Bancroft, Jo Henderson, Kristel Thomassin Published: 2026 ABSTRACT BackgroundYouth (i.e., child and adolescent) mental health difficulties are a prevalent concern, with anxiety, depression, and disruptive behavior disorders being the most common presentations. Even though psychotherapy is often recommended to help youth and families manage mental health difficulties, recent meta-analyses suggest that youth psychotherapy is only moderately effective, highlighting a need for further improvement and innovation. Emotion dysregulation is a transdiagnostic risk factor across childhood emotional and behavioral disorders, yet despite the important connection between emotion regulation and psychopathology, little research has been conducted on emotion regulation as a potential mechanism of change during psychotherapy. This study will test the biobehavioral regulation of negative emotion as a transdiagnostic mechanism of change in youth psychotherapy using the Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct Problems (MATCH). MATCH is a well-researched therapy program for youth that is suitable for testing transdiagnostic mechanisms of treatment response.MethodsThis protocol describes a two-site randomized controlled trial that aims to recruit 202 youth between the ages of 8 to 15 years with anxiety, depression, and/or disruptive behavior. Participants are randomized to the MATCH intervention condition or a waitlist control condition. Youth and their parent(s) in both conditions complete in-lab assessments and online questionnaires at the start of the study, every 3 months (i.e., quarterly), and at post-test (i.e., following the intervention/waitlist period). Physiological measures of emotion regulation such as heart rate variability and skin conductance are acquired during lab-based tasks. Youth symptoms and emotion regulation are monitored weekly for both conditions. The primary outcome is change in youth symptoms of psychopathology at post-treatment, and whether this change is mediated by change in behavioral and physiological emotion regulation. Secondary outcomes include parental functioning, parenting, family functioning, impairment, and additional measures of youth psychopathology.DiscussionFindings from the study are expected to enhance the understanding of processes that drive therapeutic change, ultimately leading to better therapy personalization and effectiveness.Trial registrationClinicalTrials.gov NCT05637320. Prospectively registered on November 15, 2022. https://clinicaltrials.gov/study/NCT05637320. Administrative information Note: the numbers in curly brackets in this protocol refer to SPIRIT checklist item numbers. The order of the items has been modified to group similar items (see http://www.equator-network.org/reporting-guidelines/spirit-2013-statement-defining-standard-protocol-items-for-clinical-trials/). Title {1a}Testing the biobehavioral regulation of negative emotion as a mechanism of change in transdiagnostic youth psychotherapy: study protocol for a randomized controlled trialTrial registration {1b and 4}ClinicalTrials.gov: NCT05637320https://clinicaltrials.gov/study/NCT05637320Protocol version {2}Version 1.4, August 2024Funding {7a}This trial is funded by the Canadian Institutes of Health Research via a Project Grant (CIHR #437267) awarded to MA, BA, and KT and a Walter Katkovsky Research Grant awarded to KT by the American Psychological Foundation. KT is also supported by an Early Researcher Award from the Ontario Ministry of Colleges and Universities.Author details {3a}Madison Aitken,1,2 Brendan F. Andrade,3,4 Claire E. Krasinkiewicz,5 Sabrina W. S. Chan,3 Sera P. Gandhi,3 Jessica W. Pun,5 Andrew Hamilton-Wright,6 Wei Wang,7,8,9 Leah Sack,5 John R. Weisz,10 Robert Bancroft,11 Jo Henderson,3,4 and Kristel Thomassin51Department of Psychology, York University, Toronto, Ontario, Canada2Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, Canada3Margaret and Wallace McCain Centre for Child, Youth & Family Mental Health, Centre for Addiction and Mental Health, Toronto, Ontario, Canada4Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada5Department of Psychology, University of Guelph, Guelph, Ontario, Canada6School of Computer Science, University of Guelph, Guelph, Ontario, Canada7Centre for Addiction and Mental Health, Toronto, Ontario, Canada8Department of Biostatistics and Data Science, University of South Florida, Tampa, FL, USA9Department of Psychological Clinical Science, University of Toronto Scarborough, Toronto, Ontario, Canada10Department of Psychology, Harvard University, Cambridge, MA, USA11Child, Youth and Emerging Adult Program, Centre for Addiction and Mental Health, Toronto, Ontario, CanadaName and contact information for the trial sponsor {3b}Not applicable (there is no sponsor for this trial)Role of sponsor {3c}The funders had no role in the design of the study protocol and have no role in the conduct or reporting of the trial. Title {1a}Testing the biobehavioral regulation of negative emotion as a mechanism of change in transdiagnostic youth psychotherapy: study protocol for a randomized controlled trialTrial registration {1b and 4}ClinicalTrials.gov: NCT05637320https://clinicaltrials.gov/study/NCT05637320Protocol version {2}Version 1.4, August 2024Funding {7a}This trial is funded by the Canadian Institutes of Health Research via a Project Grant (CIHR #437267) awarded to MA, BA, and KT and a Walter Katkovsky Research Grant awarded to KT by the American Psychological Foundation. KT is also supported by an Early Researcher Award from the Ontario Ministry of Colleges and Universities.Author details {3a}Madison Aitken,1,2 Brendan F. Andrade,3,4 Claire E. Krasinkiewicz,5 Sabrina W. S. Chan,3 Sera P. Gandhi,3 Jessica W. Pun,5 Andrew Hamilton-Wright,6 Wei Wang,7,8,9 Leah Sack,5 John R. Weisz,10 Robert Bancroft,11 Jo Henderson,3,4 and Kristel Thomassin51Department of Psychology, York University, Toronto, Ontario, Canada2Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, Canada3Margaret and Wallace McCain Centre for Child, Youth & Family Mental Health, Centre for Addiction and Mental Health, Toronto, Ontario, Canada4Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada5Department of Psychology, University of Guelph, Guelph, Ontario, Canada6School of Computer Science, University of Guelph, Guelph, Ontario, Canada7Centre for Addiction and Mental Health, Toronto, Ontario, Canada8Department of Biostatistics and Data Science, University of South Florida, Tampa, FL, USA9Department of Psychological Clinical Science, University of Toronto Scarborough, Toronto, Ontario, Canada10Department of Psychology, Harvard University, Cambridge, MA, USA11Child, Youth and Emerging Adult Program, Centre for Addiction and Mental Health, Toronto, Ontario, CanadaName and contact information for the trial sponsor {3b}Not applicable (there is no sponsor for this trial)Role of sponsor {3c}The funders had no role in the design of the study protocol and have no role in the conduct or reporting of the trial. ClinicalTrials.gov: NCT05637320 https://clinicaltrials.gov/study/NCT05637320 Madison Aitken,1,2 Brendan F. Andrade,3,4 Claire E. Krasinkiewicz,5 Sabrina W. S. Chan,3 Sera P. Gandhi,3 Jessica W. Pun,5 Andrew Hamilton-Wright,6 Wei Wang,7,8,9 Leah Sack,5 John R. Weisz,10 Robert Bancroft,11 Jo Henderson,3,4 and Kristel Thomassin5 1Department of Psychology, York University, Toronto, Ontario, Canada 2Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, Canada 3Margaret and Wallace McCain Centre for Child, Youth & Family Mental Health, Centre for Addiction and Mental Health, Toronto, Ontario, Canada 4Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada 5Department of Psychology, University of Guelph, Guelph, Ontario, Canada 6School of Computer Science, University of Guelph, Guelph, Ontario, Canada 7Centre for Addiction and Mental Health, Toronto, Ontario, Canada 8Department of Biostatistics and Data Science, University of South Florida, Tampa, FL, USA 9Department of Psychological Clinical Science, University of Toronto Scarborough, Toronto, Ontario, Canada 10Department of Psychology, Harvard University, Cambridge, MA, USA 11Child, Youth and Emerging Adult Program, Centre for Addiction and Mental Health, Toronto, Ontario, Canada Introduction Background and rationale {9a} Depression, anxiety, and disruptive behavior disorders have been identified as priority conditions impacting approximately 20% of children and adolescents (referred to here collectively as youth) globally [1–8]. Youth mental health difficulties elevate the risk of a range of adverse outcomes that can persist into adulthood, such as lower educational attainment, substance misuse, chronic mental health issues, and increased risk of suicide. These challenges highlight the need for effective interventions to address youth mental health. Psychotherapy in particular plays an essential role in providing support and promoting well-being. Hundreds of randomized controlled trials (RCTs) have evaluated the effectiveness of psychological therapies for internalizing disorders, such as anxiety and depression, as well as externalizing disorders, including disruptive behavior and attention deficit hyperactivity disorder (ADHD), among youth [9–14]. Recent meta-analyses indicate that psychotherapies for youth mental health problems are beneficial, albeit with moderate effects (mean Hedges’s gs = 0.46–0.50) [15, 16]. Therefore, there is an opportunity to refine therapies and achieve greater treatment impact. Most clinically referred youth present with multiple, overlapping mental health challenges that fluctuate significantly over time and across episodes of care [6]. This variability highlights the need for personalized and flexible approaches [17–19]. Experts have advocated for the adoption of transdiagnostic and transtheoretical treatments to more effectively address the dynamic, personalized, and dimensional aspects of youth mental health [20–23]. The empirically supported principles of change (ESPC) framework One promising framework that has been proposed to address the complexity of youth mental health is the use of empirically supported principles of change (ESPCs) [24–31]. The ESPC framework emphasizes core therapeutic principles shown to be effective across various disorders and treatment approaches, moving beyond disorder-specific treatments [32]. ESPCs offer a flexible, empirically grounded foundation that supports a client-centered approach, aligning with a shift in youth psychotherapy toward interventions that address individual needs rather than adhering rigidly to protocols [32]. The ESPC approach holds promise for addressing co-occurring difficulties, offering greater adaptability than disorder-specific protocols by allowing therapists to tailor interventions to the unique needs of youth [33, 34]. Fitzpatrick et al. [35] conducted a systematic review and meta-analysis of hundreds of RCTs over the last 60 years to examine the feasibility and utility of focusing on ESPCs in youth psychotherapy. They identified five core ESPCs: (1) calming; (2) increasing motivation; (3) changing unhelpful cognitions; (4) solving problems; and (5) practicing the positive opposite. Their analysis revealed that, across the 348 treatment-control comparisons, most psychotherapies featured one (n = 106) or two (n = 102) ESPCs, followed by three (n = 79), with a significant decline in treatments utilizing four (n = 42) or five (n = 19). Notably, treatments incorporating all five ESPCs achieved effect sizes nearly twice as large as those with fewer principles, especially in treating anxiety. Although there is substantial evidence for simple therapies such as single-session treatments [36, 37], the findings presented by Fitzpatrick et al. [35] suggest that psychotherapies targeting multiple ESPCs may yield more pronounced benefits than their simpler counterparts. Emotion regulation as a primary ESPC in youth psychotherapy The five ESPCs identified by Fitzpatrick et al. [35] arguably all relate to some aspect of emotion regulation, underscoring its foundational role in youth psychotherapy. Emotion regulation refers to the strategies and processes used to manage the intensity, duration, and expression of emotions [38, 39]. We propose youth emotion regulation as a central, transdiagnostic ESPC. Calming reduces tension and emotional arousal and, through behavioral regulation, helps with emotion regulation [40]. Increasing motivation involves using environmental contingencies to make desirable behaviors more rewarding than undesirable ones, thereby supporting behavioral regulation through changes in parent/caregiver behavior that reinforce positive actions [40]. Changing unhelpful cognitions supports emotion regulation through cognitive regulation of negative emotion [40]. Solving problems involves building skills in sequential problem-solving, which helps with regulation by promoting adaptive thinking and actions, thus supporting cognitive and behavioral regulation [40]. Finally, practicing the positive opposite involves engaging in activities that directly counter unhelpful behaviors, thus behavioral regulation supports emotion regulation [40]. The role of emotion regulation in youth mental health is supported by a robust body of literature [41–45]; however, rigorous investigations exploring the role of emotion regulation as a primary ESPC in youth psychotherapy are scarce. In addition, critical biobehavioral aspects of emotion regulation have not been systematically integrated into treatment response models. A recent meta-analysis of 67 studies examined the relationship between emotion regulation and symptom change in adults undergoing treatment for various psychological disorders [46]. The findings revealed a significant reduction in emotion dysregulation and maladaptive regulation strategies, such as rumination and avoidance, across many treatment protocols [46]. However, only three of the 67 studies specifically tested emotion dysregulation as a mechanism of change in psychotherapy [46]. Moreover, no studies explored emotion regulation as a mechanism of change at a biobehavioral level or employed multimodal measurement approaches within youth psychotherapy contexts [46]. In addition, a recent meta-analysis on emotion in adult psychotherapy identified emotion regulation as a significant and promising mechanism associated with therapeutic improvement [47]. Similarly, a meta-analysis of psychosocial interventions for children and adolescents reported a moderate effect size for improvement in emotion regulation [48], positioning emotion regulation as an important mechanism to test in youth psychotherapy [49]. At present, research on emotion regulation as a central ESPC in youth psychotherapy is limited, and most existing research does not account for important measurement and analytical considerations required to robustly test ESPCs in therapy. This protocol details a rigorous approach to testing the biobehavioral regulation of negative emotion as a primary ESPC over the course of treatment in a RCT following the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2025 [50]. Recognizing that emotion regulation operates across physiological, behavioral, cognitive, and experiential dimensions [51], the proposed trial employs an integrated measurement approach—including youth and parent reports, physiological measures, and experimental tasks—to provide a comprehensive understanding of emotion regulation. By incorporating multiple treatment sessions with youth symptom tracking and emotion regulation monitoring on a frequent basis, the study will capture the dynamic nature of treatment progression. This approach will enhance our understanding of when and how therapeutic benefits emerge in relation to changes in emotion regulation. Objectives {10} The primary objective of the proposed trial is to test the biobehavioral regulation of negative emotion as an ESPC in psychotherapy for youth with anxiety, depression, and/or disruptive behavior. A secondary objective is to apply machine leaning techniques to develop a data-driven algorithm reflecting the relation between emotion regulation and treatment response. Trial design {12} The “Big Feelings” study is a two-site superiority RCT with two arms—a treatment arm and a waitlist control arm. The treatment arm uses a transdiagnostic and modular psychotherapy program for youth (i.e., the Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct Problems (MATCH)) [52]. As soon as services become available, they will be offered to waitlisted families, and families in the waitlist condition will not be delayed in starting treatment once it is available. All research assessments will be conducted on-site. Methods: participants, interventions, and outcomes Study setting {13} Both clinical sites (the Child, Youth, and Emerging Adult Program at the Centre for Addiction and Mental Health, CAMH, in Toronto, Canada; and the Maplewoods Centre for Family Therapy and Child Psychology at the University of Guelph in Guelph, Canada) provide evidence-based outpatient mental health services for youth and families and are well-equipped to support research protocols and the implementation of evidence-based psychotherapeutic interventions. Both sites have waitlists of approximately 6 months, allowing for the ethical randomization of families to the waitlist control condition. Eligibility criteria for participants {14a} This study employs broad inclusion criteria to ensure that results will be generalizable to clinic-referred youth with emotional and behavioral disorders, alongside minimal exclusion criteria to maintain feasibility and generalizability. Inclusion criteria are as follows: (1) youth aged 8–15 years; (2) proficiency in English (spoken and written); (3) seeking psychosocial treatment for concerns related to anxiety, depression, or disruptive behavior (based on information provided at the time of referral to the clinic); and (4) at least one parent/caregiver (referred to throughout as “parent”) aged 18 years or older with sufficient English proficiency to provide consent and participate in study activities. Exclusion criteria include (1) youth meets criteria for psychosis, schizophrenia spectrum disorders, eating disorders, or autism spectrum disorder; (2) youth has low cognitive functioning that would prevent active participation in research tasks; (3) youth reports active suicidal ideation requiring acute care or hospitalization; or (4) the youth does not provide assent to participate. Eligibility criteria for individuals who will deliver the intervention {14b} Clinicians in the trial must either be licensed mental health practitioners or doctoral-level clinical psychology trainees supervised by a licensed clinical psychologist. Who will take informed consent? {32a} Informed consent will be obtained from parent(s), and assent will be obtained from youth prior to enrollment. Parent(s) will be provided with the informed consent form via the Research Electronic Data Capture (REDCap) [53] platform, and a trained research assistant will review it with them. Some components of the assessment will be video and audio recorded for transcription, qualitative coding, and fidelity checks. Participants will be free to decline being audio or video recorded. Youth will be asked to provide assent during the initial assessment by a trained research assistant with their parent(s) present. If the youth is not able to write, the research assistant will obtain assent orally. Participants may withdraw at any time without consequences to their access to clinical care. Additional consent provisions for collection and use of participant data and biological specimens {32b} Consent for future use of their study data in other research projects is included in the description of the study. Interventions Explanation for the choice of comparators {9b} Given the primary objective of the trial is to test emotion regulation as a mediator of therapeutic change, a waitlist control is an appropriate comparator to account for potential threats to internal validity (e.g., maturation) [54]. Intervention description {15a} There is strong empirical support for the effectiveness of transdiagnostic psychotherapies for youth. MATCH [52], a modular treatment program for youth aged 5–15, is one of the most researched transdiagnostic psychotherapy programs for children and adolescents. Two RCTs have demonstrated MATCH’s effectiveness against usual care (Cohen’s ds from 0.47 to 0.55) [55, 56] (though see also [57, 58], in which MATCH produced significant symptom improvements but was not superior to usual care). MATCH was designed in response to several treatment barriers clinicians face in real-world mental health care, including comorbid symptom presentations and difficulty establishing treatment targets [19]. The program contains a menu of 33 evidence-based treatment modules that are common in cognitive behavioral therapy for anxiety (e.g., graduated exposure), depression (e.g., behavioral activation), and trauma (e.g., trauma narrative) and behavioral parent training (e.g., effective instructions). The modules are arranged, according to their evidence base, into core protocols targeting primary problem areas of anxiety, depression, trauma, and disruptive behavior. Flowcharts guide clinicians through the decision-making process of arranging treatment modules into a personalized course of treatment to best fit the youth’s needs [52]. MATCH is an ideal and highly generalizable psychotherapy to test transdiagnostic mechanisms of treatment response given its 33 thoroughly vetted, empirically supported modules. MATCH is delivered primarily as individual psychotherapy when youth present with anxiety and depression. To target youth disruptive behaviors, MATCH takes a behavioral parent training approach, working primarily with parents. MATCH does not prescribe a specific number of sessions; rather, treatment continues until symptoms are addressed, though published trials report an average of 11 to 22 sessions [55, 59]. Sessions will typically be held weekly. Criteria for discontinuing or modifying allocated interventions {15b} Participation in the trial is voluntary. Families may choose to withdraw from the trial at any time. Families may also be withdrawn from the trial by the research team in response to a serious adverse event; the availability of services for waitlisted families; a worsening or change in symptoms that requires a different psychotherapeutic approach (e.g., the emergence of symptoms of a disorder that is not well addressed by MATCH, or of a severity of symptoms that cannot be addressed in an outpatient context); or if participants express active suicidal ideation requiring acute care or immediate hospitalization. In the case of worsening symptoms or change in clinical need, families will be referred to other appropriate services. Any decision to discontinue an allocated intervention due to worsening or change in symptoms will be made based on clinician judgment, in consultation with the supervising psychologist, if applicable. Strategies to improve adherence to interventions {15c} All therapists providing treatment in the intervention arm will receive training in MATCH following established procedures by an expert MATCH trainer. Training will consist of approximately 4 days of group instruction over video conference describing the MATCH modules in detail and discussing clinical decision-making in MATCH. Video examples, clinical vignettes, role plays, and discussion will be used to enhance clinician learning. Training will be held annually during the trial. Where clinicians join the project between training sessions, they will be trained through review of video recorded training sessions and virtual experiential activities. Therapists will also participate in ongoing consultation on the implementation of the MATCH program. Master’s level therapists who are still in training toward their doctoral degrees will receive weekly clinical supervision by registered clinical psychologists with training in MATCH. Therapist fidelity in the MATCH condition will be measured by coding two randomly selected session audio-recordings per client following a similar approach used in previous trials of MATCH [56]. Relevant concomitant care permitted or prohibited during the trial {15d} Youth are not to be receiving other psychotherapy services during their participation in the trial. They may be taking psychotropic medications during the trial, and medication use will be monitored throughout their participation and reported descriptively in the trial manuscript. Medication use will also be controlled for statistically in analyses of outcomes and mediation. Provisions for post-trial care {34} All therapists will be either licensed, independent practitioners or be supervised by licensed clinical psychologists, and thus will have the requisite competencies to discharge families, and as needed, arrange for referrals and post-treatment care. Outcomes {16} Primary outcome The primary outcome will be youth symptoms of psychopathology at quarterly (3-month) intervals after pre-test, and from pre- to post-treatment. The Child Behavior Checklist (CBCL) is a 113-item parent-rated questionnaire that generates a broadband Total Problems score, with T-scores based on youth age and sex. Additionally, the parallel Youth Self-Report questionnaire will be completed by children ages 11 and older [60, 61]. As CBCL scores will be available for children of all ages in the sample, parent CBCL ratings will serve as the primary outcome. The Total Problems scores have shown high levels of internal consistency in previous samples (CBCL α = 0.90 and 0.94, YSR α = 0.80 and 0.89) [61, 62]. Mediating variables Youth emotion regulation will be assessed via parent- and youth-report questionnaires, performance on experimental tasks, and physiological indices, including heart rate and electrodermal activity. Child and adolescent dispositions scale Youth (7 items) and parent (9 items) ratings on the Negative Emotionality Scale will be used to measure youth regulation of negative emotion [62, 63]. The Negative Emotionality Scale has shown adequate item factor loadings, test-retest reliability, and concurrent validity [62, 63]. Experimental tasks Youth heart rate variability and electrodermal activity will be measured with a wearable BIOPAC unit during four established in-lab tasks. Prior to the tasks, youth will sit quietly for 5 min to establish baseline heart rate and electrodermal activity. Then, youth will participate in four experimental tasks and complete the Positive and Negative Affect Schedule before and after each task. On the Positive and Negative Affect Schedule, youth will rate 20 single-word items (10 positive and 10 negative emotion words; αs = 0.84 to 0.90 in previous samples) [64, 65]. The four experimental tasks include:The modified Trier Social Stress Test for Children, in which the youth is given 2 min to prepare a 5-min speech on a topic given by the research assistant (e.g., friendship). The youth is then asked to deliver their speech facing a camera and is told the speech will be evaluated by experts [66]. Speech performance is coded observationally using the validated Evaluated Speech Performance Measure [67];The Piñata Task, in which Piñatas appear on a computer screen in succession, each having a different reward value (i.e., varying number of stars). A measure of sensitivity to reward is derived based on youths’ reaction times on targets with low versus high reward values [68];The Cognitive Reappraisal Task, in which the youth views a series of emotionally arousing images, and for each image, is instructed to either attend to the image or to use cognitive strategies to decrease the intensity of their affect [69]. Reappraisal will be measured by comparing the change in affect on trials requiring the youth to attend versus trials requiring the youth to reappraise.The Conflict Discussion Task, in which the parent and youth each rate the presence and intensity of their conflict on common sources of parent-youth disagreement on the Issues Checklist (e.g., chores, homework, screentime) [70, 71]. The research assistant will then select the overlapping item with the highest conflict rating and instruct the parent and youth to discuss for 5 min and attempt to resolve the conflict [72]. The research assistant will remain close by (in an adjacent room or hallway) and monitor the discussion to ensure that the conflict does not escalate. Should any concerns arise about parent or youth conflict, the research assistant will immediately contact the on-call clinician. Video-recordings of the conflict discussion will be coded observationally for parent-youth interaction quality, parenting, and youth emotion regulation. The modified Trier Social Stress Test for Children, in which the youth is given 2 min to prepare a 5-min speech on a topic given by the research assistant (e.g., friendship). The youth is then asked to deliver their speech facing a camera and is told the speech will be evaluated by experts [66]. Speech performance is coded observationally using the validated Evaluated Speech Performance Measure [67]; The modified Trier Social Stress Test for Children, in which the youth is given 2 min to prepare a 5-min speech on a topic given by the research assistant (e.g., friendship). The youth is then asked to deliver their speech facing a camera and is told the speech will be evaluated by experts [66]. Speech performance is coded observationally using the validated Evaluated Speech Performance Measure [67]; The Piñata Task, in which Piñatas appear on a computer screen in succession, each having a different reward value (i.e., varying number of stars). A measure of sensitivity to reward is derived based on youths’ reaction times on targets with low versus high reward values [68]; The Piñata Task, in which Piñatas appear on a computer screen in succession, each having a different reward value (i.e., varying number of stars). A measure of sensitivity to reward is derived based on youths’ reaction times on targets with low versus high reward values [68]; The Cognitive Reappraisal Task, in which the youth views a series of emotionally arousing images, and for each image, is instructed to either attend to the image or to use cognitive strategies to decrease the intensity of their affect [69]. Reappraisal will be measured by comparing the change in affect on trials requiring the youth to attend versus trials requiring the youth to reappraise. The Cognitive Reappraisal Task, in which the youth views a series of emotionally arousing images, and for each image, is instructed to either attend to the image or to use cognitive strategies to decrease the intensity of their affect [69]. Reappraisal will be measured by comparing the change in affect on trials requiring the youth to attend versus trials requiring the youth to reappraise. The Conflict Discussion Task, in which the parent and youth each rate the presence and intensity of their conflict on common sources of parent-youth disagreement on the Issues Checklist (e.g., chores, homework, screentime) [70, 71]. The research assistant will then select the overlapping item with the highest conflict rating and instruct the parent and youth to discuss for 5 min and attempt to resolve the conflict [72]. The research assistant will remain close by (in an adjacent room or hallway) and monitor the discussion to ensure that the conflict does not escalate. Should any concerns arise about parent or youth conflict, the research assistant will immediately contact the on-call clinician. Video-recordings of the conflict discussion will be coded observationally for parent-youth interaction quality, parenting, and youth emotion regulation. The Conflict Discussion Task, in which the parent and youth each rate the presence and intensity of their conflict on common sources of parent-youth disagreement on the Issues Checklist (e.g., chores, homework, screentime) [70, 71]. The research assistant will then select the overlapping item with the highest conflict rating and instruct the parent and youth to discuss for 5 min and attempt to resolve the conflict [72]. The research assistant will remain close by (in an adjacent room or hallway) and monitor the discussion to ensure that the conflict does not escalate. Should any concerns arise about parent or youth conflict, the research assistant will immediately contact the on-call clinician. Video-recordings of the conflict discussion will be coded observationally for parent-youth interaction quality, parenting, and youth emotion regulation. Physiological measures Throughout all tasks, youth heart rate and electrodermal activity will be acquired via the BIOPAC MP160 data acquisition system. Electrocardiogram electrodes will be placed on the youth in a Lead II configuration and electrodermal activity electrodes will be placed on the youth’s palm [73, 74]. Secondary outcomes Secondary outcomes will be examined as change scores from pre- to post-treatment (unless otherwise specified) on the measures below, all of which have demonstrated reliability and validity and sensitivity to intervention. Secondary outcomes are classified into areas of youth psychopathology, parental functioning, parenting, and family functioning. Additional measures of psychopathology Parents and youth will complete idiographic and weekly measures to provide further information on youth psychopathology symptoms. Additional subscales from the CBCL and YSR will also be analyzed. Top problems assessment This brief clinical interview is completed separately with parents and youth. At baseline, interviewees list up to three problems that concern them the most, which are written down in their own words and rated in severity (0 = Not a problem; 4 = A very big problem). The problems are then rank ordered based on severity. At subsequent timepoints, the severity of the top problems identified at baseline is re-rated [60, 75, 76]. The Top Problems Assessment has demonstrated test-retest reliability (rs 0.69 to 0.91), convergent validity with established measures (rs 0.25 to 0.48), and sensitivity to treatment effects [76]. Behavior and feelings survey The Behavior and Feelings Survey is a 12-item questionnaire with parallel parent- and youth-report versions designed to monitor weekly change during youth psychotherapy. It generates Internalizing, Externalizing, and Total Problems scores, all of which have demonstrated high internal consistency (α = 0.87-0.94) [77]. Youth symptoms of psychopathology will be measured weekly to investigate trajectory of change during treatment. Using weekly measurement provides precise information on the trajectory of youth emotion regulation and psychopathology symptoms across treatment and allows the measurement schedule to match different treatment durations. We have successfully used this approach to measurement in previous trials [56], supporting its feasibility in the present sample. CBCL and YSR Additional broadband scales from the CBCL and YSR [61], including the broadband Internalizing and Externalizing scores, will be analyzed. Impairment Parents will complete the Brief Impairment Scale, a 23-item questionnaire covering three domains of youth impairment: Interpersonal, School/Work, and Self-Fulfillment (α ≥ 0.60 for all scales) [78, 79]. Parental functioning Parents will complete the following self-report measures. Brief symptom inventory This 18-item questionnaire will be used to measure parents’ own psychopathology symptoms in the domains of Somatization, Depression, and Anxiety, all of which have demonstrated good internal consistency (αs = 0.82, 0.87, and 0.84, respectively) [80–82]. Difficulties in emotion regulation scale The 18-item short form will be used to measure parents’ own emotion regulation [83–85]. The total score has demonstrated high internal consistency (α = 0.91) [86]. Parenting stress index The 36-item short form will be used to measure stress specific to the parent-youth relationship [87, 88], categorized into subscales (parental distress, parent-child dysfunctional interaction, and difficult child), all of which have demonstrated acceptable internal consistency (αs 0.71 to 0.87) [89]. Parenting Parenting behaviors will be measured through parent self-report of their likelihood of 6 different responses in 12 hypothetical scenarios on the Coping with Children’s Negative Emotions Scale [90, 91]. Responses are grouped in two overall scores: Supportive (Expressive Encouragement, Problem-Focused, and Emotion-Focused Reactions) and Unsupportive (Punitive, Minimizing, and Distress Reactions). The Supportive and Unsupportive scores have shown adequate internal consistency in previous samples (α = 0.65 and 0.75, respectively) [92]. Family functioning Parents will complete the following measures of family functioning. Family functioning style A 26-item measure of family strengths and ability to deal with stress [93, 94], with items categorized into five subscales (interactional patterns, family values, coping strategies, family commitment, and resource mobilization). The total score has demonstrated good internal consistency (α = 0.81) [95]. Confusion, hubbub, and order scale A 15-item questionnaire measuring disorder and confusion within the home [96–98]. The total score has demonstrated adequate internal consistency (α = 0.79) [97]. Covariates Demographic information Parents will provide basic demographic information using a locally developed form. Child and adolescent trauma screener Parents will complete a checklist of 15 potentially traumatic events at baseline. If one or more event is endorsed, 20 traumatic stress symptoms items will be rated by the parent at baseline and subsequent visits [99, 100]. Tanner stages assessment Parent and youth will report on the youth’s pubertal development using the Tanner Stages Assessment [101] because puberty can influence physiological measures of emotion regulation. Medication tracking form Parents will report on their child’s use of medication for issues related to attention, mood, anxiety, or disruptive behavior at each study visit. Service access form Parents of youth in the waitlist condition will report on any external services accessed while on the waiting list using the Service Access Form. Session attendance Information on session attendance will be collected from participants’ health records. Participant timeline {18} MATCH Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct ProblemsaTreatment continues until symptoms are addressed (published trials report an average of 11–22 sessions)bFamilies will be offered services when they become available. Wait times are approximately 6 months at both sitescQuarterly assessments will be completed while participants are in the study (intervention or waitlist) dRe-administered at quarterly and post assessments if trauma exposure is endorsed at baseline MATCH Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct ProblemsaTreatment continues until symptoms are addressed (published trials report an average of 11–22 sessions)bFamilies will be offered services when they become available. Wait times are approximately 6 months at both sitescQuarterly assessments will be completed while participants are in the study (intervention or waitlist) dRe-administered at quarterly and post assessments if trauma exposure is endorsed at baseline MATCH Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct Problems aTreatment continues until symptoms are addressed (published trials report an average of 11–22 sessions) bFamilies will be offered services when they become available. Wait times are approximately 6 months at both sites cQuarterly assessments will be completed while participants are in the study (intervention or waitlist) dRe-administered at quarterly and post assessments if trauma exposure is endorsed at baseline Sample size {19} A Monte Carlo simulation was conducted with 10,000 replications to determine the required sample size for the primary mediation analysis. The simulation assumed a moderate total treatment effect size (Cohen’s d = 0.50), based on findings from previous MATCH trials [55, 56]. Notably, these previous trials have compared MATCH to an active treatment condition, meaning the use of this effect size in the present calculations (in which MATCH will be compared to waitlist) is a conservative approach. To achieve an adequate statistical power of 80%, a sample size of 202 can detect the proportion in therapy outcome mediated as small as 25%. We parameterized the power analysis using the proportion mediated [102], defined as the proportion of the total effect explained by the indirect pathway. We adopted this parameterization to provide a clinically interpretable measure of the extent to which the intervention effect operates through emotion regulation. The simulation further assumed a moderate effect size from the treatment to emotion regulation [103], a type I error level of 0.05, and a 20% attrition rate based on similar studies in the field and the anticipated challenges of longitudinal follow-up. The machine learning analysis is exploratory; however, our proposed sample size would be sufficient to detect small-to-medium effects (based on a χ2 goodness of fit model, α = 0.05, β = 0.8, and 4 association categories, though this should not be considered a formal power analysis and the sample size determination was based on the primary mediation analysis). Recruitment {20} The flow of participants is shown in Fig. 1. Participants can self-refer or can be referred from another provider such as a physician or pediatrician. Participants who meet the inclusion criteria will be contacted by a research assistant. The research assistant will conduct initial screening to confirm their eligibility and introduce the study process and the treatment protocol. Once the consent process is completed, the family will attend a baseline assessment.Fig. 1CONSORT diagram of participant flow from enrollment to study completion CONSORT diagram of participant flow from enrollment to study completion Assignment of interventions: allocation Sequence generation {21a, 21b} Randomization will follow a 1:1 allocation for treatment:waitlist stratified by youth age (8–10, 11–13, and 14–15 years) and gender (boy, girl, non-binary) given the potential age and gender differences in referral rates for anxiety, depression, and disruptive behavior disorders. Randomization will not be stratified by site. An allocation table with all necessary parameters and a random seed will be generated by the trial statistician and provided to REDCap. Concealment mechanism {22} To maintain blinding, the statistician will label treatment assignments as A and B without knowledge of the corresponding intervention conditions. The trial coordinator will then link these labels to the actual treatment conditions. The allocation table in REDCap is locked once the project is in production, such that no one can view or alter the randomization list. The ability to randomize participants will be restricted to staff who do not perform any client-interacting activity. Implementation {23} Once randomization is activated, REDCap randomly assigns participants according to the allocation table and stratification design. The randomization process will be performed via REDCap randomization upon completion of the initial assessment. Once participants are randomized, the designated research assistant will follow up with the family, informing them of the randomization decision and next steps. Assignment of interventions: blinding Who will be blinded {24a, 24b} All client interactions, including screening and in-person assessments, will be conducted by research assistants who are blinded to participant condition. All randomization information on the participant log and research binder will be blacked out to ensure blinding is maintained. The statistician and data analysts will be blinded to treatment allocation throughout the conduct of the primary analyses. Randomization codes will be replaced with masked labels. Participants and clinicians will be aware of their client’s assigned condition. Procedure for unblinding if needed {24c} It is unlikely that unblinding will be needed because participants are aware of their assigned condition. If such a situation arises (e.g., in the case of an adverse event), the research assistant responsible for randomization will obtain participants’ condition from the randomization log. Data collection and management Plans for assessment and collection of outcomes {25a} Data collection will involve a combination of questionnaire, experimental, physiological, and interview measures. Assessments will be completed at pre-treatment, quarterly during the treatment and/or waitlist phase, and at post-treatment. A subset of questionnaires will be completed weekly during the treatment and waitlist phase (see Participant timeline for measurement schedule). Prompts to complete weekly surveys will be sent to participants by email or text message, based on their preference. Plans to promote participant retention and complete follow-up {25b} To promote participant retention, youth and parents will receive honoraria in the form of gift cards following each study visit and for the completion of weekly measures, as follows: parents will receive $40 (pre-test and quarterly visits) and $60 (post-test), and youth will receive $15 (pre-test and quarterly visits) and $25 (post-test). In addition, both parents and youth will receive $2 per weekly survey completed plus a $5 bonus for every 5 surveys completed. Other retention strategies will include the use of reminder calls and messages for assessments, sending thank you and birthday cards to families, and having families complete questionnaires online. Data management {26} Electronic data capture (REDCap) will be used for all questionnaire measures [53, 104]. Physiological data will be acquired during experimental tasks and saved on a secure server immediately after each session. In addition, physiological data will be examined and cleaned following recommended best practices prior to analysis. Video and audio recordings will be transferred from recording devices to a secure internal server as soon as possible following research visits, after which they will be deleted from the recording device. All data will be stored on a secure, encrypted, password-protected server in Canada with dual factor authentication and will only be accessible to the research team. Regular backups are conducted by the host institution to minimize any potential for data loss. The second phase of the study, which will involve data mining and machine learning, will take place at the University of Guelph. All data collected from CAMH clinics will be de-identified and transferred to the custody of co-investigators at Guelph via secure, encrypted networks. Personal health information collected at CAMH will be stored on CAMH’s secure servers for 25 years. Personal health information collected at the University of Guelph will remain at the University of Guelph and will be stored securely following their institution’s storage requirements. Data access is managed by a Research Collaboration Agreement executed by both clinical sites, and study investigators will have access to the full trial data. The agreement states that data are to be shared between the two clinical sites, in compliance with all applicable laws and research ethics. Confidentiality {33} Participants will be assigned a study identifier, and their identifying information will be stored separately from their study data. A key file will be used to store identifying information and link participant identity to their study identifier. Consent and assent forms containing names and signatures will be stored separately from study data and study identifier. Computer databases are encrypted and password protected, and only study investigators and approved members of the research team will be given access. When transferring data between the two sites, a secure file transfer system with end-to-end encryption will be used, and for larger files, transfers will occur with an encrypted hard drive with keypad lock. Statistical methods Statistical methods for primary and secondary outcomes {27a} Analytic plan for primary and secondary outcomes Preliminary analyses will test for significant baseline differences in demographic factors between the MATCH and waitlist groups, with any significant factors included as covariates in our models. An exploratory factor analysis will assess whether observed emotion regulation variables load onto a broad latent construct. These variables include the Child and Adolescent Dispositions Scale parent and youth report, performance on the Trier Social Stress Test, Piñata Task, Cognitive Reappraisal Task, and Conflict Task, and physiological signals including heart rate and electrodermal activity. Variables that do not load onto the same factor will be analyzed independently within our mediation models. Latent growth curve mediation modeling [105, 106] will test whether improvements in emotion regulation mediate symptom and impairment improvements resulting from MATCH treatment, compared to those in the waitlist group. This approach will account for the nested structure of the data (i.e., data nested within individuals and individuals nested within therapists) as well as potentially unequally spaced time points. We will conduct two types of mediation analyses based on our operationalization of treatment response—pre-to-post change and trajectory of change. First, emotion regulation (latent biobehavioral regulation factors identified by the exploratory factor analysis) will be modeled for all participants as a latent change from baseline to the 3-month quarterly assessment, while change in symptoms and impairment will be modeled as a latent change from baseline to the post-assessment. The mediated effect will be tested using the bootstrap-based product-of-coefficients method [107], which will provide bias-corrected 95% confidence intervals using the recommended asymmetric confidence interval method. Second, to examine trajectories of change during MATCH treatment and for those in the waitlist group, we will use latent growth curves to model latent slopes of change in emotion regulation (weekly Child and Adolescent Dispositions Scale) and symptoms (weekly Behavior and Feelings Survey) and explore the temporal relationships between these trajectories. This analysis will evaluate whether the growth characteristics of emotion regulation (e.g., slope of the latent growth trajectories) mediates the treatment effect of MATCH on the growth characteristics of the behavior outcome trajectories. This analysis will further help us understand the temporal order and potential bidirectionality of changes in regulation and symptoms. In both modeling approaches, we will include youth age, sex, medication use at baseline, puberty status, and ethnicity as an a priori set of covariates. Most previous research has focused on sex differences in prevalence of internalizing and externalizing problems and treatment response, and sex differences in physiological measures have also been identified [108]. However, given the potential for differences in emotion expression and behavior related to socially constructed gender roles [109], we will also conduct a sensitivity analysis with gender (instead of sex) as a covariate. The bootstrap-based product-of-coefficients method [107] will also be used to test the latent biobehavioral regulation factor as a mechanism of change in secondary outcomes including parental functioning and emotion-related parenting. While medication use can vary over time, adjustment for baseline medication status was selected to balance confounding control with model stability. Similarly, the sample size and population will not allow detailed analyses of medication class or dose. Unmodeled time-varying medication effects and variations in medication type and dose remain limitations of the proposed design. Machine learning Association mining techniques will be applied to extract characteristic patterns from the multimodal emotion regulation data that are relevant to treatment response. Association mining techniques have been used extensively for problems in which a deep understanding of the relationships discovered by the algorithm is necessary for their safe and effective application. This has led to extensive use of these techniques in healthcare [110–115]. Our interest in these techniques rests on their production of lists of descriptive patterns, each of which is discovered through the identification of a statistically significant relationship among values in the dataset under observation. By using a rigorous discovery model, the efficacy and association confidence with the application of rules discovered by these algorithms are immediately amenable to simple mathematical analysis. For this reason, these techniques are transparent to a posteriori analysis. This is required for our overarching knowledge translation goal, which is to use the produced rules outside of any computational system in the form of a decision tree tool that may be used clinically. To explore the power of these patterns, we will use information theory to score each one, describing how much support each pattern has within the overall distribution of the data. This allows us to distinguish patterns based on richly described regions of the data distribution from those with sparse support using a transparent and interpretable, statistically based association mining algorithm. The result is a confidence value associated with assertions made about treatment response using the patterns, indicating whether there is high confidence (based on well characterized regions of the data) or poor confidence (based on rare or conflicting regions of the data). An additional benefit of these techniques is their ability to identify negative patterns [116]. A negative pattern describes associations within a data set that occur more rarely than random chance would suggest. Counter-intuitively, such patterns contain important predictive information as they will uncover relationships between variables where an observed value in one variable restricts the value of another, which can be particularly important when describing how a particular emotion regulation pattern may relate to treatment outcomes. Results will be reported following recommendations for machine learning in clinical research [117]. Interim analyses {28b} As this is not an efficacy or effectiveness trial, stopping guidelines will not apply and, therefore, no interim analysis of outcome data is planned. Methods for additional analyses (e.g., subgroup analyses) {27d} Additional analyses, specifically linear mixed models, will examine whether family functioning and parenting moderate youth symptom outcomes and trajectory of change in youth symptoms. To evaluate the robustness of results under alternative adjustment strategies, we will conduct a sensitivity analysis using inverse probability of treatment weighting based on propensity scores. The propensity score model will include the same covariate set as the primary analyses: youth age, sex, baseline medication use, pubertal status, and ethnicity. Covariate balance before and after weighting will be assessed using standardized mean differences. Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data {27b, 27c} We will use an intent-to-treat approach and apply full information maximum likelihood estimation [118] to minimize potential bias caused by missing data. Plans to give access to the full protocol, participant-level data, and statistical code {5, 6} Individuals may request the full protocol and statistical analysis plan from the authors. As well, datasets and related statistical code may be deposited in an open access data repository. Other documents (participant information sheet, questionnaires, etc.) can be requested from the authors. Oversight and monitoring Composition of the coordinating center and trial steering committee {3d} Principal investigators (KT, BFA, MA) will meet biweekly with research assistants involved in data collection to ensure that established protocols are followed and to identify any relevant issues (e.g., recruitment rate, measure completion) that may need to be addressed. In addition, weekly to biweekly consultation meetings with trial therapists delivering the MATCH intervention will be used to encourage fidelity, monitor participant safety, and identify relevant issues (e.g., retention). Composition of the data monitoring committee, its role and reporting structure {28a} As this trial is focused on identifying a principle of change in an established, evidence-based treatment, there is no need to monitor changes in clinical equipoise through accumulated data. In addition, as both waitlist and the focal intervention follow established approaches to treatment, risk to participants is minimal; therefore, a formal data monitoring committee is not needed. Adverse event reporting and harms {17} The potential for adverse events will be monitored through biweekly meetings with research personnel involved in data collection and with trial clinicians. In addition, participants will be provided with the principal investigators’ and the respective research ethics board (REB) chairs’ contact information, should they wish to express any concerns about the study. All adverse events or harms will be reported to the REBs at both institutions. Frequency and plans for auditing trial conduct {29} Auditing of study protocols and processes will be conducted annually at the time annual reviews are submitted to the REBs and may result in adjustments to protocols that facilitate study processes and enhance adherence and data quality. In addition, audits on data completion, enrollment, and attrition will be completed every 3 to 6 months. Plans for communicating important protocol amendments to relevant parties (e.g., trial participants, ethical committees) {31} Principal investigators (KT, BFA, MA) will be responsible for decisions regarding protocol amendments. Any amendments to the protocol will be reviewed by the REBs at both institutions. Amendments will also be made to the trial registry, which tracks the history of changes to the protocol for transparency. Dissemination plans {8} Results will be disseminated via peer-reviewed publications and presentations at national and international conferences. Knowledge exchange meetings will involve the sharing of results to clinician and leadership stakeholders at the agencies involved in the study. Authorship eligibility will be guided by established recommendations in the field [119]. Sub-project ideas will be presented to a publication committee for approval. Plans for patient or public involvement {11} Clinician partners were involved in the design of the study from the outset as knowledge users. Results will be disseminated to community audiences, including family partners as well as clinicians and trainees. Discussion Extensive empirical research supports emotion dysregulation as a transdiagnostic risk factor for child and adolescent emotional and behavioral disorders [1–5, 7, 42, 120], and a growing body of literature suggests that improvements in emotion regulation account for improvements in therapy [40, 48]. Results from this trial will provide a clearer picture of how improvement in emotion regulation as an ESPC can effectively promote therapeutic change. In addition, an expected outcome of our machine learning approach will be statistically based, observable rule-based algorithms of emotion regulation components that account for variations in treatment response. Together, these algorithms are the initial step toward a clinical decision-making framework. Through further development and evaluation in subsequent studies, these algorithms may be able to inform patient-oriented care pathways that tailor combinations of interventions to patient mental health needs, ultimately advancing both theory and clinical practice. Trial status Protocol version 1.4, August 2024. Enrollment began on 2022-12−12 and is expected to continue until 2027-09.