The impact of working alliance in managing youth anxiety and depression: a scoping review

Working alliance (WA), also termed therapeutic alliance, is a multidimensional construct encompassing the bond between patient and therapist, agreement on therapy goals, and consensus on tasks. Adult-focused reviews show that WA predicts engagement, adherence, and symptom reduction, while ruptured alliances undermine effectiveness. In youth, especially within individual and low-intensity therapies (e.g., behavioural activation, psychoeducation, problem-solving therapy), the importance and mechanisms of WA are not well understood. Given that most mental health problems onset during youth, understanding WA is key to maximising prevention, treatment, and ongoing management of anxiety and depression. This study aimed to explore how WA affects anxiety and depression outcomes among young people aged 14–24 via a scoping review complemented by stakeholder consultations and validation workshops.

Prior systematic reviews and meta-analyses (primarily in adults) consistently identify WA as essential across psychotherapies, predicting uptake, engagement, adherence, and symptom reduction. In youth-focused evidence, meta-analyses indicate that certain CBT features (goal setting, parental involvement, relapse prevention, booster sessions) explain limited variance in outcomes, implying additional factors (including WA) contribute to effects. Strong WA predicts positive outcomes in family-involved youth treatments, but the role of WA in individual and low-intensity youth therapies remains unclear. Evidence on mechanisms suggests WA may influence outcomes via improvements in interpersonal functioning, self-esteem, coping, optimism, adherence, and emotional regulation; however, heterogeneity and potential confounding (e.g., reverse causality, third variables) complicate causal inference.

The initiative comprised: (1) a scoping review following PRISMA-ScR; (2) semi-structured stakeholder consultations; and (3) validation workshops with young people (YP). Research questions asked whether better WA improves outcomes for youth with anxiety/depression; which WA elements (bond, goal, task) influence outcomes; which factors (patient characteristics, therapy format/delivery) influence WA–outcome relations; and whether ruptured WA can be harmful. Eligibility: quantitative studies (RCTs, cross-sectional, cohort, case-control) reporting on WA in YP aged 14–24 with anxiety and/or depression; studies included if ≥50% of participants were in the 14–24 range; English-language only; excluded systematic reviews, editorials, qualitative studies, case studies, protocols. Information sources: PubMed, CINAHL, Scopus, PsycINFO, Africa-Wide Information from inception to August 2021; grey literature via Google Scholar; forward and backward citation searches; authors contacted when only abstracts were available. Selection process: duplicates removed; titles/abstracts screened; full texts retrieved; independent dual extraction and cross-checking; disagreements resolved with senior researcher input. Data items: author, year, age group, outcomes (WA, anxiety, depression), and secondary outcomes (relationships, coping). Synthesis: qualitative narrative by study design and objectives. Stakeholder consultations: purposive sample of lay health counsellors (n=6), psychologists (n=2), occupational therapists (n=2), psychiatrists (n=2), and YP with lived experience (n=20); interviews shifted to phone due to SARS-CoV-2; real-time note-taking, inclusion of verbatim quotes; thematic analysis used for coding and theme development. Validation workshops: YP co-created a visual mechanistic framework of WA pathways; community advisory group consulted; final workshop refined insights. YP involvement spanned project design, literature searching, data collection, analysis, and dissemination.

- Study selection: 274 records identified; 139 after duplicates; 70 full texts screened; 27 studies included. - Settings and modalities: 26/27 in high-income countries; US 12/27; therapies mostly individual (18/27) vs group (5/27); only 2 studies used digital platforms; others in-person (25/27). CBT was most common (16/27). DSM was the most-used diagnostic tool (10/27). - WA measurement: Working Alliance Inventory (WAI) most used (12/27). WA assessed at the beginning (14/27) or both beginning and end (17/27). Clients were primary assessors (15/27); therapist ratings recorded in two studies. - Associations: 24/27 studies found positive associations between WA and clinical outcomes (reduced anxiety/depression, improved functioning); 2 null; 1 negative association. - Mechanisms/secondary outcomes: Strong WA associated with improved interpersonal relationships, self-esteem, positive coping strategies, optimism, treatment adherence, and emotional regulation. - Stakeholder insights (N=32): YP emphasised conducive environment, regular engagements, confidentiality, collaborative goal setting, and involvement in tasks; client factors (willingness, politeness, expressiveness) and therapist attributes (trustworthiness, communication, empathy, non-judgmental stance, boundary setting) were key. Potential barriers: cultural background, communication patterns, dress, therapist age, gender, religion; many YP preferred same-gender therapists and valued experience. Clinicians highlighted personal connection, trust, boundaries, confidentiality, empathy, and shared goal setting; motivation and repeated goal review supported WA formation. - Delivery format: Evidence suggests comparable WA and outcomes across in-person and videoconference CBT; digital guided self-help showed high WA yet mixed WA–outcome associations. - Group vs individual: Mixed findings; group therapy linked to increased self-esteem and reduced depression, but severe pre-treatment interpersonal problems predicted poorer WA and dropout in group settings relative to individual therapy. - Setting: WA–outcome associations observed in both outpatient and inpatient settings. - Timing and components: Early WA often predicted subsequent symptom improvements; task agreement within CBT frequently showed the strongest link with outcomes after controlling for expectancy and prior change.

Findings address whether WA improves outcomes for youth with anxiety/depression by demonstrating that stronger WA is generally associated with better clinical outcomes, corroborated by both empirical studies and stakeholder perspectives. The results underscore WA as an active ingredient across psychotherapies and formats. Mechanistically, WA appears to enhance engagement, adherence, self-esteem, coping, optimism, emotional regulation, and interpersonal functioning, creating a beneficial feedback loop with symptom reduction. However, mechanisms are not fully delineated due to methodological heterogeneity and possible confounding (e.g., expectancy effects, prior symptom change). Task agreement may be a particularly salient component in CBT. Moderators and contextual factors likely shaping WA–outcome relations include client characteristics (motivation, attachment style, relational difficulties), therapist attributes (empathy, communication skill, professional experience), therapy format (individual vs group), delivery mode (in-person vs digital), setting (inpatient vs outpatient), and timing (early vs later alliance). While evidence suggests early WA is beneficial, findings are mixed and may depend on measurement timing and method. Digital modalities can support effective WA, but measures adapted for online contexts and repeated assessments are needed to clarify associations. Group therapy may benefit those with impaired relational experiences but can be challenging for individuals with severe interpersonal problems. Overall, WA’s significance in youth mirrors adult literature, but youth-specific contexts, cultural factors, and low-intensity therapies warrant further study.

WA is a key active ingredient in psychotherapies for managing anxiety and depression among young people aged 14–24. Stronger alliances correlate with better clinical outcomes and beneficial psychosocial processes. Stakeholder input highlights actionable elements for cultivating WA: safe and confidential environments, regular contact, collaborative goal and task setting, empathy, clear boundaries, and effective communication. Routine, multi-timepoint assessment of WA from both client and therapist perspectives is recommended, alongside efforts to enhance WA across therapeutic approaches and delivery formats. Future research should employ longitudinal and well-powered experimental designs, repeated WA measurements, and advanced modelling to disentangle mechanisms, within- vs between-person effects, timing, and moderators, and should prioritise diverse and low-resource settings and the validation of WA measures across modalities.

- No formal risk of bias assessment conducted due to scoping review design. - Evidence base dominated by high-income countries; only one study from a low-income context (Kenya), limiting generalisability. - Few studies focused exclusively on the 14–24 age group; some included broader age ranges. - Heterogeneity in WA measurement (instruments, timing, assessor) complicates comparisons and causal inference; need for psychometric evaluation and standardisation across client, observer, and therapist perspectives. - Limited long-term and digital-format studies; potential confounding (e.g., expectancy, prior symptom change) and reverse causality not fully addressed.