Online Transdiagnostic Emotion Regulation Treatment for Adolescents With Mental Health Problems: A Randomized Clinical Trial

Adolescent mental health problems are prevalent worldwide, often comorbid, and include substantial burden from subthreshold conditions. Access to psychological treatments is limited in primary care due to structural barriers, including insufficient service availability and long wait times. Transdiagnostic, online psychological interventions may address these barriers by being scalable and suitable for comorbid and subthreshold presentations. Maladaptive emotion regulation strategies are implicated across a range of psychopathologies, making emotion regulation a promising transdiagnostic treatment target. Building on prior work with internet-delivered emotion regulation therapies for youth, the authors developed Primary Care Online Emotion-Regulation Treatment (POET), a brief, therapist-guided, online program. The primary aim was to assess the feasibility and acceptability of POET versus an active supportive control among adolescents with mental health problems in primary care. Secondary aims were to examine preliminary effects on clinical outcomes (symptom severity, depression/anxiety, functioning) and the target mechanism (emotion regulation).

The paper situates POET within literature highlighting: (1) high worldwide prevalence and disability from youth mental disorders and increasing concerns over time; (2) frequent comorbidity and burden from subthreshold conditions; (3) treatment gaps in adolescent mental health care; and (4) the promise of online, transdiagnostic psychological interventions to address scalability and comorbidity. Prior treatments for adolescents often target specific disorders and show variable efficacy. Emotion regulation difficulties are transdiagnostically related to psychopathology, supporting interventions that target emotion regulation. Meta-analyses indicate psychosocial interventions can improve emotion regulation and clinical symptoms in youth. The Extended Process Model of Emotion Regulation provides a theoretical framework for skill targets. Prior internet-delivered Emotion Regulation Individual Therapy for Adolescents (ERITA/IERITA) showed feasibility and efficacy in youths with nonsuicidal self-injury, informing POET’s design as a brief, scalable program with limited therapist contact.

Design: Single-blind randomized clinical trial comparing POET to an active supportive treatment in Swedish primary care. Recruitment: October 16, 2022–January 31, 2023; final follow-up July 28, 2023. CONSORT guidelines were followed. Ethics approval obtained; informed consent from adolescents ≥15 years and parents in writing; younger adolescents provided verbal consent with parental written consent.

Participants: Adolescents aged 12–17 years with mental health problems and at least one participating parent. Exclusions: severe mental illness requiring specialized care or CGAS <41; acute suicidality; ongoing psychotherapy; psychopharmacological changes within 2 months; insufficient Swedish proficiency; or life circumstances impeding participation. Referrals came from primary care/schools and self-referrals (including social media).

Randomization and masking: Allocation 1:1 to POET or supportive treatment via independent researcher using a true random number service; sequence concealed in opaque sealed envelopes. One parent per adolescent completed a parallel 6-week online course. Outcomes were assessed at baseline, immediately posttreatment, and 3 months posttreatment via blinded assessor interviews (telephone) and online self-reports. Blinding integrity was checked by assessor guesses.

Interventions: POET is grounded in the Extended Process Model of Emotion Regulation, targeting situational, attentional, cognitive, and response modulation strategies. It comprises 6 adolescent modules with psychoeducation and skills training to enhance adaptive and reduce maladaptive emotion regulation, plus a parallel parent course to train supportive strategies. The active control (supportive treatment) matched format, length, and therapist contact but excluded POET’s active components; it provided information on mental health and prompted reflections on well-being (eg, school, friends). Parents in the control completed weekly reflections on supporting well-being. Both arms used a blended format over 6 weeks with online modules and therapist-guided video sessions, therapist feedback on modules and homework. Therapists were clinical psychologists in Swedish primary care.

Outcomes: Primary (feasibility/acceptability): consent rate; assessment completion; adherence (≥3 modules completed); adolescent CEQ credibility (1–9) and expectancy (0–100%); adolescent and parent CSQ-8 satisfaction (8–32); therapist time; blinding feasibility; adverse events. Clinical outcomes (blinded-rated): CGI-S and CGI-I; self-reported RCADS-47 anxiety/depression; CGAS global functioning. Treatment response defined as CGI-I 1–2. Target mechanism: CERQ adaptive and maladaptive cognitive emotion regulation; PAQ-S alexithymia (adolescent self-report at all time points).

Statistical analysis: Sample size of 30 consistent with CONSORT extension for feasibility trials; a priori within-group power estimate (80% power to detect standardized mean difference 0.6 at α=0.05, assuming 20% dropout). Feasibility/acceptability summarized descriptively; between-group differences in satisfaction and therapist time via independent t-tests; blinding guesses tested with binomial test. Exploratory within-group effects analyzed ITT. For ordinal CGI-S, linear quantile mixed models (lqmm in R 4.4.1) estimated median effects with 1000-bootstrap CIs; effect size calculated as median time effect divided by SD at baseline. For continuous outcomes (RCADS-47, CGAS, CERQ, PAQ-S), linear mixed-effects regressions with random intercept for participant and dummy-coded time; effect sizes as Cohen d for mixed models (β_time divided by baseline SD) with 1000-bootstrap CIs; positive sign indicates improvement. Missing data assumed missing at random. Two-sided tests, P<.05.

Sample: 30 adolescents (mean age 14.2 years; 93% female) randomized to POET (n=15) or supportive treatment (n=15). One POET participant (3%) dropped out before initiation.

Feasibility/acceptability: Consent rate 81% (30/37 eligible). Posttreatment assessment completion 93% (28/30); at 3 months, 87% (26/30). Adherence: Adolescents completed mean (SD) modules—POET 4.6 (1.8)/6; control 5.7 (0.7)/6. Parents: POET 5.1 (1.4), control 5.9 (0.5). Two adolescents in POET (7%) completed ≤3 modules; all parents completed >3. Credibility/expectancy (adolescents): CEQ credibility mean (SD)—POET 5.79 (1.37), control 6.33 (1.10); CEQ expectancy—POET 49.3% (15.4%), control 50.0% (16.5%). Satisfaction (CSQ-8) mean (SD): adolescents—POET 20.6 (4.93), control 22.8 (3.12), P=.20; parents—POET 24.8 (3.56), control 23.1 (3.76), P=.25. Therapist time (minutes, mean [SD]): adolescents—POET 74.7 (57.3), control 64.6 (30.0), P=.55; parents—POET 70.3 (32.1), control 64.1 (29.5), P=.59. Blinding maintained; assessor guesses not different from chance posttreatment (67.8% correct; P=.09) or at 3 months (42.3% correct; P=.56). Adverse events: six total, evenly distributed; increased stress and transient symptom worsening; no serious events.

Clinical outcomes (within-group exploratory effects):

• CGI-S (ordinal): POET showed significant reduction pre to post (effect size 1.30; 95% CI, 0.73–1.86) maintained at 3 months (1.32; 95% CI, 0.76–1.88). Supportive treatment showed no significant change.
• Treatment responders (CGI-I 1–2): post—POET 47% (7/15), control 7% (1/15); 3 months—POET 33% (5/15), control 20% (3/15).
• RCADS-47 anxiety/depression: POET improved pre to post (Cohen d 1.07; 95% CI, 0.37–1.84) and pre to 3 months (d 1.28; 95% CI, 0.51–2.08). Control: no significant effects (post d 0.15; 95% CI, −0.24 to 0.53; 3 months d 0.37; 95% CI, −0.04 to 0.77).
• CGAS global functioning: POET improved pre to post (d 1.26; 95% CI, 0.66–1.85) and pre to 3 months (d 1.54; 95% CI, 0.95–2.14). Control: no significant improvements.

Target mechanism:

• CERQ-maladaptive: POET reduced maladaptive cognitive coping pre to post (d 1.10; 95% CI, 0.52–1.70) and pre to 3 months (d 0.82; 95% CI, 0.22–1.40). Control: no significant changes.
• CERQ-adaptive: no significant effects in either group.
• PAQ-S alexithymia: no significant effects in either group.

The trial demonstrates that a brief, therapist-guided, online, transdiagnostic emotion regulation treatment (POET) is feasible and acceptable for adolescents with mental health problems in primary care, as evidenced by high consent and completion rates, satisfactory adherence, credibility/expectancy, and high satisfaction in both trial arms, with efficient therapist time and preserved blinding. Importantly, adolescents receiving POET showed large within-group improvements in global symptom severity, depression/anxiety symptoms, and global functioning that were maintained at 3 months, alongside reductions in maladaptive cognitive emotion regulation strategies—the proposed target mechanism. The supportive control did not show corresponding improvements, suggesting specific benefits of emotion regulation skill training beyond nonspecific supportive effects. These findings align with prior research on youth emotion regulation interventions and extend them by showing promising effects with a brief, scalable program requiring limited therapist involvement. In the context of substantial treatment gaps in adolescent mental health care, POET may offer a practical approach to increase access and address comorbid and subthreshold presentations in primary care.

A brief, therapist-guided, online transdiagnostic emotion regulation treatment (POET) was feasible and acceptable for adolescents in primary care and was associated with reductions in overall symptom severity, anxiety/depression symptoms, and maladaptive cognitive coping, and with improved global functioning, maintained at 3 months. POET’s brevity and limited therapist time suggest strong potential for scalable implementation to improve access for adolescents. Future work should include larger, more gender-diverse randomized trials powered for between-group comparisons, comparisons with gold-standard treatments and treatment as usual, longer-term follow-up, and implementation studies including noninferiority and health economic evaluations.

Key limitations include: predominantly self-referred sample introducing potential selection bias; modest sample size and lack of stratified randomization, precluding robust between-group effectiveness analyses; use of an active supportive control (rather than treatment as usual or a gold-standard treatment), limiting clinical interpretability and comparative effectiveness conclusions; need to further improve adolescent assessment completion rates; short follow-up (3 months); and generalizability concerns due to the high proportion of female participants.