The goal of this clinical trial is to evaluate the effectiveness and cost-effectiveness of a digital mindfulness-based intervention in adults (aged 18-65) diagnosed with emotional disorders like depression or anxiety. The main questions it aims to answer are:

• Does adding a digital mindfulness intervention to usual care help people recover from emotional disorders faster and more sustainably over two years?
• Is this combined approach more cost-effective than usual care alone? Researchers will compare the group receiving the digital mindfulness intervention plus their usual treatment to the group receiving only their usual treatment to see if the intervention leads to better long-term recovery and represents good value for money.

Participants in the intervention group will:

• Attend eight weekly 2-hour online group mindfulness sessions.
• Use a WeChat mini-program for 49 days of guided mindfulness exercises and daily tasks.
• Patients who have not achieved reliable recovery after group retraining voluntarily participate in individual UP&MIED counseling.
• Complete regular questionnaires and interviews over two years to track their progress.

All participants will continue to receive their usual medical care from their doctors throughout the study.

The prevalence of emotional disorders remains high, with depression and anxiety disorders being the most common (Freeman, 2022; Huang et al., 2019). Although existing treatments can alleviate symptoms in the short term, emotional disorders often follow a chronic and recurrent course, making long-term recovery difficult for patients (Mulder, 2015) and resulting in a persistent and heavy socioeconomic burden (Amos et al., 2018; GBD, 2022). Furthermore, individuals with emotional disorders often exhibit significant diagnostic instability and high comorbidity rates, making single-diagnosis interventions inadequate for real-world treatment needs (Bullis et al., 2019). Consequently, developing intervention strategies applicable to high comorbidity and capable of delivering long-term recovery benefits represents an urgent requirement in current emotional disorder treatment.

Transdiagnostic interventions are considered a key approach to meeting this need (Barlow et al., 2020). Grounded in the shared etiology and common mechanisms underlying psychopathology, these interventions enable practitioners to address diverse disorders using a single set of methods. Previous meta-analyses indicate that transdiagnostic interventions, exemplified by Mindfulness-Based Interventions (MBIs), yield moderate improvements in depressive and anxiety symptoms (Cuijpers et al., 2023). Specifically, Mindfulness-Based Stress Reduction (MBSR) has demonstrated potential as an alternative to medication in treating anxiety disorders (Hoge et al., 2023). For preventing depressive relapse in recurrent depression patients, Mindfulness-Based Cognitive Therapy (MBCT) proved as effective as maintenance antidepressant medication (Kuyken et al., 2015). The "Mindfulness Intervention for Emotional Distress" (MIED) employed in this study is a transdiagnostic psychological intervention developed based on MBSR and the "Unified Protocol for Cross-Diagnostic Treatment of Emotional Disorders" (UP). This program proposes a transdiagnostic psychopathology diamond model for emotional disorders, identifying four core pathological mechanisms hindering recovery: low distress tolerance, cognitive inflexibility, excessive emotional behavior, and abnormal engaging in life (Liu, 2024). MIED targets these four mechanisms through mindfulness practices combined with intervention strategies from UP (e.g., cognitive reappraisal, interoceptive exposure). This improvement of underlying pathological mechanisms is considered crucial for facilitating patients' transition from short-term symptom relief to robust, long-term recovery.

At the same time, the application of digital technology and group formats in psychological interventions continues to expand, providing low-threshold, scalable pathways for integrating transdiagnostic mindfulness interventions into routine healthcare systems (Schaeuffele et al., 2024). The digital MIED program has demonstrated promising effects in improving emotional distress and accelerating symptom improvement in emotional disorders across preliminary short-term studies (Ju et al., 2022; Li et al., 2024; Li et al., 2025; Wang et al., 2024). The high-frequency interaction compensates for the lack of proactive intervention tools in current clinical care. However, the efficacy of transdiagnostic interventions in promoting long-term recovery remains under-validated. Most transdiagnostic interventions, including mindfulness-based approaches, have primarily focused on single diagnostic dimensions (Barlow et al., 2017) or non-clinical samples (Lindegaard et al., 2021) , and generally lack large-sample support and long-term follow-up (Cuijpers et al., 2023; Newby et al., 2015; Schaeuffele et al., 2024) , making it difficult to reflect the long-term benefits for individuals with emotional disorders within real-world clinical pathways. Furthermore, the potential cost-effectiveness of such interventions remains under-examined (Schaeuffele et al., 2024). Against the backdrop of prolonged course and high comorbidity in emotional disorders, these evidence gaps constrain the effective integration of cross-diagnostic mindfulness interventions within routine healthcare systems.

Beyond this, the current limited understanding of the mechanisms of action also constrains improvements in treatment outcomes (Kazdin, 2007). While numerous theories have proposed hypotheses regarding these mechanisms (Baer, 2003; Brown et al., 2007; Garland et al., 2015; Liu, 2024; Shapiro et al., 2006), research on the mechanisms underlying MBIs remains lacking in systematic validation when measured against the criteria for establishing psychological intervention mechanisms proposed by Kazdin (Kazdin, 2007, 2009). No variables have been found to satisfy most mechanism validity criteria, with particularly insufficient evidence meeting experimental manipulation, temporal priority, and consistency criteria. More rigorous, high-quality research is needed to examine MBI mechanisms.

To address these evidence gaps, we plan to conduct a large-scale, multicenter, pragmatic randomized controlled trial evaluating the long-term efficacy and cost-effectiveness of a transdiagnostic digital mindfulness intervention as an adjunct to treatment as usual (TAU). This study will also incorporate an examination of MBI mechanisms of action, aiming to facilitate further optimization of intervention protocols.

Inclusion Criteria:

1. Age: 18-65 years.
2. Diagnosed with an emotional disorder by a outpatient psychiatrist, including depressive disorders, anxiety disorders (e.g., generalized anxiety disorder, panic disorder, agoraphobia, social anxiety disorder), obsessive-compulsive disorder, post-traumatic stress disorder, and eating disorders (e.g., anorexia nervosa, bulimia nervosa).
3. Symptom severity meeting the threshold: PHQ-9 score ≥10 or GAD-7 score ≥8.

Exclusion Criteria:

1. Current diagnosis of psychotic disorders or bipolar disorder.
2. Current organic mental disorders, pervasive developmental disorders, severe cognitive impairment, or substance use disorders.
3. Current suicide risk (PHQ-9 item 9 score >2).
4. Antisocial personality disorder.
5. Severe medical illnesses that may affect intervention participation or require recent hospitalization.
6. Previous participation in a systematic 8-week mindfulness course.
7. Inability to access the internet.

Given that the data in this study involves sensitive mental health information of patients and ethical restrictions, this research plan does not share individual-level research data with the public. If there is a reasonable academic need, an application can be made through the corresponding author, and after evaluation by the research team and the ethics committee, a decision will be made on whether to provide the necessary data summary or analytical support.

Data analysts (as part of the investigator team) are masked during the primary analysis phase, working with group codes (e.g., A/B) and unblinding only after completion of the main analysis to minimize bias .