The effects of group nursing intervention based on emotion regulation training on depression and nonsuicidal self-injury in adolescents

Abstract

Objective

This retrospective cohort study aimed to assess the effectiveness of a group nursing intervention based on emotion regulation training in improving depressive symptoms, emotional regulation ability, and the frequency and severity of NSSI behaviors among adolescents. Additionally, the intervention’s safety and patient satisfaction were evaluated.

Methods

A total of 240 adolescents diagnosed with depressive disorder and NSSI who received treatment across departments of a tertiary veterans’ hospital in Baoding, China, from January 2022 to June 2024 were retrospectively analyzed. Patients were grouped post hoc on the basis of the type of nursing care documented in their records: 120 patients received standard pharmacological treatment combined with supportive nursing care (control group), and 120 patients additionally received a group-based emotion regulation training program (intervention group). The intervention lasted 8 weeks, with weekly 90-minute sessions including modules on emotional awareness, cognitive restructuring, group sharing, and breathing/mindfulness relaxation.

Results

After 8 weeks, the CDI score (from 29.1 ± 4.9 to 13.2 ± 3.6) was significantly lower in the intervention group than in the control group (from 28.7 ± 5.2 to 19.5 ± 4.4; t = 6.45, P < 0.001). The total ERQ score increased more substantially in the intervention group (from 25.1 ± 4.5 to 33.8 ± 3.7) than in the control group (from 25.4 ± 4.3 to 29.5 ± 3.9; t = 7.02, P < 0.001). Weekly NSSI episodes decreased from 2.4 ± 0.7 to 0.5 ± 0.3 in the intervention group, whereas NSSI episodes in the control group decreased from 2.2 ± 0.6 to 1.3 ± 0.4 (t = 5.12, P < 0.001). NSSI severity scores decreased from 1.8 ± 0.6 to 0.4 ± 0.3 in the intervention group and from 1.7 ± 0.5 to 0.9 ± 0.4 in the control group (t = 6.11, P < 0.001). Nursing satisfaction was significantly greater in the intervention group (4.6 ± 0.4) than in the control group (3.8 ± 0.5; t = 8.17, P < 0.001). No serious adverse events occurred in either group.

Conclusion

Grouping nursing interventions based on emotion regulation training was associated with alleviation of depressive symptoms, greater emotional regulation capacity, and lower frequency and severity of NSSI behaviors among adolescents, suggesting a potential beneficial role that warrants further prospective validation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-025-03750-5.

Introduction

Adolescent depression, characterized by persistent low mood and anhedonia, has become an increasing public health concern, with the prevalence of major depressive episodes (MDEs, defined as DSM-diagnosed episodes of clinically significant depressed mood and related symptoms) in youths showing an increasing trend in recent years [1, 2]. Nonsuicidal self-injury (NSSI), defined as deliberate self-harm without suicidal intent, remains common among adolescents with depression [3]. Neuroimaging studies have indicated that comorbid depression and NSSI in youth are associated with alterations in neural circuitry: functional connectivity between limbic regions and the prefrontal cortex—particularly between the right orbitofrontal cortex and insula—correlates negatively with NSSI frequency [4, 5].

Current interventions for adolescent depression and NSSI present notable limitations. Although cognitive behavioral therapy (CBT) is widely implemented, completion rates in group-based CBT drop by 23% when social support is inadequate.

[6]. Antidepressant medications carry a risk of increased suicidality in young patients, and the short‑term benefits of CBT often dissipate by the two‑year follow‑up [7, 8]. Moreover, most existing group interventions focus primarily on symptom management; of the psychological treatments targeting NSSI, only six have demonstrated efficacy, with Treatment of Self-Injurious Behaviors (T-SIB) supported by the most robust randomized controlled evidence [3].

A compelling body of evidence now positions emotion dysregulation as a critical transdiagnostic precursor to both adolescent NSSI and depression. Longitudinal studies have confirmed that difficulties in emotion regulation precede the onset of depressive symptoms and NSSI, establishing it as a key predictive risk factor [9, 10]. Meta-analytic evidence further reinforces this link, revealing a strong association between emotion dysregulation and NSSI [11]. This provides the crucial theoretical basis for selecting emotion regulation as the intervention target. Emotional regulation training (ERT) has emerged as a transdiagnostic approach that employs goal setting, cognitive reappraisal, and other strategies to enhance emotional processing [12, 13]. Additionally, it has been suggested that components such as cognitive reappraisal within ERT play crucial roles in modulating self-injurious behaviors by addressing underlying emotional dysregulation, which in turn impacts the neural mechanisms associated with these behaviors, including disruptions in the prefrontal cortex and limbic system [14, 15]. Multiple studies have demonstrated ERT’s feasibility and potential effectiveness in addressing both internalizing and externalizing problems in adolescents [16, 17]. Functional MRI data reveal that successful ERT strengthens prefrontal modulation of limbic structures, complementing the orbitofrontal–limbic disconnect observed in depressed youth [4, 18]. These ERT modules (e.g., cognitive reappraisal and cognitive restructuring) aim to increase orbitofrontal and dorsolateral prefrontal regulatory control over insular and limbic reactivity; such top-down modulation provides a plausible pathway by which ERT could influence the OFC–insula functional connectivity implicated in NSSI [19]. Furthermore, group nursing interventions leverage peer modeling to activate social learning processes; cost–benefit analyses estimate $2.30 in societal return for every $1 invested [20].

The COVID‑19 pandemic has exacerbated these challenges, increasing adolescent depression rates by 31%, with female adolescents disproportionately affected (23.4%) [2, 21]. This escalation underscores the urgent need for interventions that integrate neurobiological mechanisms and social environmental factors. Group interventions grounded in emotion regulation may fill this gap.

Accordingly, the present study innovatively incorporates ERT modules into a group nursing framework. In this study, “group nursing intervention” refers to a nurse-led, structured group program that integrates ERT modules with nursing tasks (assessment, health education, care coordination, family/school liaison and ongoing monitoring), which is distinct from therapist-led group psychotherapy in its emphasis on nursing continuity and practical care components [22]. The present study aims to evaluate the effectiveness and safety of a nurse-led group nursing intervention based on ERT in reducing depressive symptoms and NSSI among adolescents.

Methods

Study design and participants

This single-center, retrospective cohort study reviewed the medical records of adolescents diagnosed with depressive disorder and NSSI who received treatment across departments of a tertiary veterans’ hospital in Baoding, China, between January 2022 and June 2024. Eligible participants were inpatients or outpatients aged 12–18 years with a DSM-5 diagnosis of depressive disorder and documented NSSI at admission or during their first psychiatric evaluation. To be included, patients required complete nursing documentation and follow-up data. We excluded individuals with comorbid schizophrenia, bipolar disorder, or severe cognitive impairment; those with a history of suicide or homicide attempts; patients concurrently enrolled in other structured psychological interventions (e.g., specialized CBT programs); and those with incomplete records or who discontinued treatment. A total of 240 patients met the inclusion criteria and were grouped post hoc on the basis of the type of nursing care documented in their medical records (control group n = 120; intervention group n = 120). The study protocol was approved by the Ethics Committee of a tertiary veterans’ hospital in Baoding, China. Given its retrospective design and the use of fully anonymized data, the requirement for informed consent was waived. All procedures adhered to the principles of the Declaration of Helsinki and relevant medical ethics guidelines.

Sample size statement

This study was a retrospective analysis, and the sample size was determined by the number of eligible cases available during the study period (n = 240; 120 per group). To assess the statistical detectability of clinically relevant effects, a post hoc power evaluation was performed. With the CDI change as the primary outcome and a two-sample t test (two-sided α = 0.05, power = 80%), under the assumption of within-group SD σ ≈ 6 and an expected mean difference δ ≈ 3, the required sample size would be approximately 63 per group. Therefore, the present sample (n = 120 per group) provides adequate power to detect such a moderate effect. This is a post hoc assessment of detectability and does not replace a priori sample size calculations for prospective trials.

Nursing interventions

Control group

The participants in the control group received a tertiary veterans’ hospital in Baoding, China’s standard nursing care for adolescent depression, comprising four core components. First, pharmacotherapy management was guided by clinical practice guidelines: patients were prescribed selective serotonin reuptake inhibitors (SSRIs) or alternative antidepressants, and nursing staff conducted weekly evaluations of dosage, adherence, and adverse effects in close collaboration with psychiatrists to dynamically adjust treatment plans. Second, emotional monitoring and documentation involved daily completion of mood diaries under nurse supervision to record subjective experiences of anxiety, sadness, and other affective states; these records were supplemented by brief interviews during rounds to facilitate early detection of mood fluctuations or self-harm urges. Third, a crisis alert and intervention mechanism was established: when warning signs of suicide or self-harm risk emerged, a three-tiered crisis response protocol was activated, entailing intensified observation, rapid incident reporting, and referral to a specialized psychiatric crisis intervention team as needed. Finally, daily living support was provided, including structured guidance on sleep hygiene, encouragement of social engagement, and recommendations for appropriate recreational activities, all of which were intended to help adolescents establish healthy sleep and exercise routines and thereby improve their overall quality of life and emotional stability. For transparency, the package of “standard nursing care” described above reflects routine practice at our tertiary veterans’ hospital and is comparatively intensive (regular pharmacotherapy coordination, daily mood monitoring and a formal crisis protocol); readers should note this when interpreting between-group differences.

Intervention group

In addition to the standard pharmacological treatment and individualized supportive care provided to the control group, the intervention group participated in an 8-week group emotion regulation training program, with one 90-minute session per week. The program was co-facilitated by two senior psychiatric nurses and one psychiatrist, each of whom completed at least 12 h of training on the emotion regulation intervention manual and passed a simulation-based assessment. Sessions were delivered in accordance with the Adolescent Emotion Regulation Group Nursing Manual, ensuring strict adherence to predefined objectives, content outlines, and interactive formats. Each group consisted of 8–10 participants who convened in the child and adolescent psychiatry activity room, which was equipped with a projector, whiteboard, and meditation mats. The curriculum was divided into four modules:

Weeks 1–2: Emotional Awareness Training, including theoretical introduction to the components of emotion and hands-on practice using “emotion diaries” and an “emotion thermometer.”

Weeks 3–4: Cognitive restructuring, comprising case-based instruction and group exercises following the “ABCDE” framework (activating event, belief, consequence, dispute, effective new belief) [23].

Weeks 5–6: Group Sharing and Peer Support, featuring structured sharing sessions, “praise loops,” and active listening exercises.

Weeks 7–8: Breathing and mindfulness relaxation, including progressive muscle relaxation, guided mindfulness breathing, and a closing ritual in which each participant set a personalized emotional regulation goal for the next week.

To monitor fidelity and facilitate subsequent quality review, attendance and participation levels were recorded at every session. After the fourth session, all participants completed abbreviated versions of the CDI and ERQ, and the facilitators adjusted the focus of later sessions on the basis of this interim feedback. All session materials, manuals, and participant logs were systematically archived by the nursing department (Table 1).

Table 1.

Content and key implementation elements of Group-Based nursing intervention featuring emotion regulation training

Component	Specific Content
Intervention Team & Training	Led jointly by 2 senior psychiatric nurses and 1 psychiatrist; ≥12 h of manual-based training and simulation exercises; follows the Adolescent Emotion Regulation Group Nursing Manual.
Intervention Duration & Venue	8-week program, once per week, 90 min per session; group size: 8–10 participants; location: psychiatric group activity room equipped with projector, whiteboard, and meditation mats.
Emotion Awareness Training (Weeks 1–2)	Introduction to emotional components and physiological signals; practical exercises using “emotion diary” and “emotion thermometer” self-assessments.
Cognitive Restructuring (Weeks 3–4)	Case-based teaching of negative thought scenarios; instruction in the “ABCDE” model (Adversity–Belief–Consequence–Disputation–Effect) with group role-play.
Group Sharing & Peer Support (Weeks 5–6)	Structured sharing of emotional distress and positive feedback; practice of social skills through “compliment circles” and “active listening.”
Breathing & Mindfulness Relaxation (Weeks 7–8)	Progressive muscle relaxation (from head to feet); 20-minute full-body scan mindfulness breathing with background natural sounds; setting weekly emotional self-management goals at session close.
Intervention Monitoring	Attendance tracked and participation levels assessed each session; abbreviated CDI and ERQ assessments after Session 4 to adjust focus; all materials (slides, manuals, logs) archived in the nursing department for quality review and dissemination.

Observational indices

To comprehensively evaluate the efficacy and safety of the group nursing intervention, this study employed a multidimensional set of outcome measures encompassing changes in psychometric scale scores, behavioral assessments, and patient satisfaction, alongside systematic monitoring of adverse events. All outcome scores were completed independently by two uniformly trained clinical psychology nurses. To minimize information bias, the raters were provided with deidentified medical record summaries and scale completion forms with treatment/care allocation information removed so that the raters were blinded to the study group during the scoring process (the raters were not involved in the clinical treatment and care of the patients and had no access to care allocation records). However, because the source data were historical medical records, complete blinding to group allocation cannot be guaranteed and is acknowledged as a limitation.

Primary outcomes

Children’s Depression Inventory (CDI) score change [24]: The Chinese version of the CDI, consisting of 27 items scored 0–2 (total range 0–54), was used to quantify depressive symptoms, with higher scores indicating greater severity. The scale has excellent reliability (Cronbach’s α = 0.85–0.90) in Chinese adolescents. CDI scores at T0, T1, and T2 were compared to calculate the ΔCDI (post-prechange), thereby reflecting the degree of symptom improvement.

(2) Emotion regulation capacity was assessed via two standardized instruments:

① Emotion Regulation Questionnaire (ERQ) [25]: This 10- item comprises two subscales—cognitive reappraisal and expressive suppression—each item is rated on a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree). Higher total scores denote more frequent use of adaptive emotion regulation strategies. The Chinese ERQ has demonstrated sound reliability (total scale Cronbach’s α ≈ 0.78; subscales 0.72–0.81). We calculated the ERQ total and subscale scores at each time point and derived ΔERQ subscale changes to capture dynamic shifts in regulation strategy utilization.

② Difficulties in Emotion Regulation Scale (DERS) [26]: This 36-item instrument evaluates obstacles across six dimensions: nonacceptance, goals, impulse, awareness, clarity, and strategies. Items are rated 1 (almost never) to 5 (almost always), yielding a total score range of 36–180, where higher values indicate greater difficulty. The Chinese DERS exhibits robust internal consistency (Cronbach’s α >0.85). Total and subscale DERS scores were tabulated at each time point, and ΔDERS (total and dimension‑specific) was computed to assess changes in emotion regulation difficulties.

(3) NSSI behavior frequency and severity: Drawing on patient self-injury diaries and nursing records, the average weekly NSSI frequency over 8 weeks was calculated. NSSI severity was assessed via an event-level rating scale developed by our research team on the basis of clinical experience and relevant literature [27] (see Supplementary Table S1). The scale integrates wound depth, associated bleeding, and the degree of impairment to daily functioning, classifying each NSSI event on a 1–5 scale: 1 = superficial epidermal scratches with no bleeding or functional impairment; 2 = shallow cuts with minimal bleeding and transient, minor functional impact; 3 = moderate injury reaching the dermis with noticeable bleeding, typically requiring local outpatient care; 4 = deep injury extending into subcutaneous tissue or approaching muscle, usually requiring suturing or outpatient surgical management and associated with marked functional limitation; 5 = critical injury involving muscle, tendon, or adjacent vital structures, or multiple severe wounds necessitating hospitalization or operative repair and resulting in severe or long-term functional impairment. Two trained raters (a study nurse and a psychiatrist) independently scored each event; discrepant ratings were adjudicated by a third senior rater.

Secondary outcomes

(1) Nursing Satisfaction: A bespoke “Adolescent Depression Care Satisfaction Scale” comprising eight items that assess dimensions such as information delivery, interactive experience, environmental support, and overall impressions was employed. Each item is rated on a 5-point Likert scale (1 = very dissatisfied to 5 = very satisfied), yielding a total score ranging from 8 to 40. Following pilot testing and refinement, the scale demonstrated excellent internal consistency (Cronbach’s α = 0.88). Satisfaction was measured only at T2 via an anonymous electronic questionnaire, and scores were aggregated independently by the evaluation team.

(2) Adverse events: Any negative psychological or physiological reactions arising during the intervention were recorded, including (i) deterioration of mood (defined as a CDI score increase ≥ 5 points compared with the previous assessment); (ii) aggravation of NSSI behaviors (increased frequency or higher severity rating); and (iii) other treatment-related adverse reactions (e.g., dizziness, exacerbated insomnia). All events were graded according to the Common Terminology Criteria for Adverse Events (CTCAE v5.0) severity scale [28], ranging from Grade 1 (mild) to Grade 5 (death). The incidence rates and distributions across severity grades were calculated to evaluate intervention safety.

Data collection

Baseline demographic and clinical data, intervention adherence details, and outcome measures were extracted from the hospital’s electronic medical record system and standardized nursing log forms. The investigators entered this information into a uniform case report form, and two independent reviewers cross-checked all the entries to ensure data accuracy.

Statistical analysis

Statistical analyses were conducted via IBM SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables were first assessed for normality via the Shapiro–Wilk test and are presented as the means ± standard deviations (x̄ ± s). Homogeneity of variance was verified via Levene’s test. Between-group comparisons at baseline and at each time point utilized independent-samples t- tests, whereas within-group longitudinal changes were examined via paired-samples t-tests. Categorical data are expressed as counts and percentages, with between-group comparisons performed via χ² tests. NSSI severity ratings, as ordinal data, were analyzed via the Mann–Whitney U test. For all primary between-group comparisons, we report mean differences with 95% confidence intervals and Cohen’s d (pooled SD) as an index of effect size. Cohen’s d was computed via the pooled standard deviation. The interpretation of d follows conventional thresholds (small ≈ 0.2, medium ≈ 0.5, large ≥ 0.8). To control familywise type I error where multiple pairwise comparisons were performed, we applied the Bonferroni correction (adjusted α = 0.05/number of comparisons) or reported adjusted P values; where many comparisons made the Bonferroni correction overly conservative, we reported both unadjusted and false discovery rate (FDR)-adjusted P values. All tests were two-tailed, and a P-value < 0.05 was considered statistically significant.

Results

Baseline characteristics

A total of 240 adolescents with depression and comorbid non-suicidal self-injury (NSSI) were enrolled, with 120 patients each in the control and intervention groups. As summarized in Table 2, there were no significant baseline differences between the groups. The mean ages were 15.2 ± 1.8 years in the control group and 15.1 ± 1.9 years in the intervention group, with comparable sex distributions (50.0% male vs. 48.3% male). The body mass index (BMI) values were similar (20.5 ± 2.3 kg/m² vs. 20.7 ± 2.5 kg/m²), indicating comparable nutritional status. The rates of family history of depression (15.0% vs. 16.7%) and single-parent households (18.3% vs. 20.0%) did not differ significantly. Furthermore, the proportions of SSRI use in the six weeks preceding enrollment were high in both groups (85.0% vs. 86.7%), and the mean illness durations were 6.5 ± 3.1 months and 6.3 ± 3.0 months, respectively, with no significant intergroup differences.

Table 2.

General clinical characteristics of patients

Characteristic	Control group (n = 120)	Intervention group (n = 120)	χ²/t	P
Age (years)	15.2 ± 1.8	15.1 ± 1.9	t = 0.36	0.72
Gender, n (%)
Male	60 (50.0%)	58 (48.3%)	χ²=0.07	0.79
female	60 (50.0%)	62 (51.7%)
BMI (kg/m²)	20.5 ± 2.3	20.7 ± 2.5	t = 0.48	0.63
Family history of depression, n (%)	18 (15.0%)	20 (16.7%)	χ²=0.15	0.7
Single-parent families, n (%)	22 (18.3%)	24 (20.0%)	χ²=0.10	0.75
SSRIs usage, n (%)	102 (85.0%)	104 (86.7%)	χ²=0.18	0.67
Course of the disease (months)	6.5 ± 3.1	6.3 ± 3.0	t = 0.45	0.65

BMI: Weight (kg)/Height ² (m²); SSRIs use refers to those who have continuously used selective serotonin reuptake inhibitors within 6 weeks before the intervention. A family history of depression refers to a confirmed history of depression among first or second-degree relatives. A single-parent family refers to a situation where one parent has guardianship or the guardian is not the parent

Changes in depressive symptoms and emotion regulation

Both groups exhibited reductions in depressive symptoms after intervention, with the intervention group showing greater improvement, indicating an association between participation in the group intervention and symptom improvement. At Week 4 (T1), the Children’s Depression Inventory (CDI) score of the intervention group decreased from 29.1 ± 4.9 to 20.2 ± 4.1, whereas it decreased from 28.7 ± 5.2 to 24.1 ± 4.9 in the control group (between-group t = 4.75, P < 0.001). By week 8 (T2), the scores had further declined to 13.2 ± 3.6 in the intervention group versus 19.5 ± 4.4 in the control group (t = 6.45, P < 0.001). The mean CDI reduction at T1 (–8.9 ± 4.3) was nearly double that of the control group (–4.6 ± 3.2; t = 6.12, P < 0.001), and the disparity persisted at T2 (–15.9 ± 5.0 vs. − 9.2 ± 3.7; t = 6.45, P < 0.001), indicating an added benefit of emotion regulation training (Fig. 1).

Fig. 1.

Comparison of CDI scores before and after intervention in the two groups of patients

Emotional regulation difficulties, as measured by the DERS, improved in both groups but to a greater extent in the intervention cohort. At T1, the mean DERS total scores were 85.2 ± 12.8 in the intervention group versus 92.5 ± 13.9 in the control group (Intervention − Control = − 7.30; 95% CI − 10.70 to − 3.90; t = 4.23, P < 0.001; Cohen’s d = − 0.55); by T2, the gap widened further (78.4 ± 11.3 vs. 90.1 ± 12.5; Intervention − Control = − 11.70; 95% CI − 14.73 to − 8.67; t = 7.01, P < 0.001; Cohen’s d = − 0.98). The intervention group’s mean DERS reduction (–19.5 ± 6.3) was substantially greater than that of the control group (–8.2 ± 5.1; ΔT2–T0, Intervention − Control = − 11.30; 95% CI − 12.76–9.84; t = 14.30, P < 0.001). Dimension--specific analyses demonstrated significantly greater declines in nonacceptance, goals, impulses, awareness, strategies, and clarity scores for the intervention group (all P < 0.001), reflecting broad-spectrum improvements in emotion regulation (Table 3).

Table 3.

Comparison of emotional regulation ability of the two groups of patients before and after intervention

Indicators/Dimensions	Control group (n = 120)	Intervention group (n = 120)	t	P	Mean diff (I − C), 95% CI	Cohen’s d
DERS total score
T0	98.3 ± 14.7	97.9 ± 15.2	0.18	0.86	−0.40 (− 4.20 to + 3.40)	−0.03
T1	92.5 ± 13.9	85.2 ± 12.8	4.23	<0.001	−7.30 (− 10.70 to − 3.90)	−0.55
T2	90.1 ± 12.5	78.4 ± 11.3	7.01	<0.001	−11.70 (− 14.73 to − 8.67)	−0.98
ΔT2–T0	–8.2 ± 5.1	–19.5 ± 6.3	14.3	<0.001	−11.30 (− 12.76 to − 9.84)	−1.97
Non-acceptance
T0	16.4 ± 3.2	16.2 ± 3.4	0.42	0.67	−0.20 (− 1.04 to + 0.64)	−0.06
T1	15.6 ± 3.1	13.7 ± 3.0	4.9	<0.001	−1.90 (− 2.68 to − 1.12)	−0.62
T2	14.8 ± 2.9	11.2 ± 2.7	9.22	<0.001	−3.60 (− 4.31 to − 2.89)	−1.28
Goals
T0	15.8 ± 3.6	15.5 ± 3.5	0.63	0.53	−0.30 (− 1.20 to + 0.60)	−0.08
T1	15.0 ± 3.4	12.9 ± 3.1	4.9	<0.001	−2.10 (− 2.93 to − 1.27)	−0.65
T2	14.2 ± 3.2	10.3 ± 2.9	8.56	<0.001	−3.90 (− 4.68 to − 3.12)	−1.28
Impulse control
T0	17.6 ± 4.1	17.8 ± 4.0	0.38	0.7	+ 0.20 (− 0.83 to + 1.23)	0.05
T1	17.0 ± 3.9	15.1 ± 3.5	4.7	<0.001	−1.90 (− 2.84 to − 0.96)	−0.51
T2	16.1 ± 3.7	12.4 ± 3.2	7.45	<0.001	−3.70 (− 4.58 to − 2.82)	−1.07
Awareness
T0	14.9 ± 3.4	15.1 ± 3.6	0.44	0.66	+ 0.20 (− 0.69 to + 1.09)	0.06
T1	14.4 ± 3.3	12.8 ± 3.2	3.9	<0.001	−1.60 (− 2.39 to − 0.81)	−0.44
T2	13.9 ± 3.1	10.8 ± 2.8	6.52	<0.001	−3.10 (− 3.86 to − 2.34)	−0.96
Strategies
T0	18.2 ± 4.7	18.0 ± 4.5	0.29	0.77	−0.20 (− 1.07 to + 0.67)	−0.04
T1	17.2 ± 4.5	14.9 ± 4.1	4.4	<0.001	−2.30 (− 3.19 to − 1.41)	−0.53
T2	16.3 ± 4.2	12.0 ± 3.8	8.12	<0.001	−4.30 (− 5.20 to − 3.40)	−1.07
Clarity
T0	15.4 ± 3.9	15.3 ± 4.0	0.13	0.9	−0.10 (− 1.00 to + 0.80)	−0.03
T1	15.0 ± 3.8	13.5 ± 3.6	3.3	0.001	−1.50 (− 2.44 to − 0.56)	−0.41
T2	14.4 ± 3.6	11.7 ± 3.3	5.81	<0.001	−2.70 (− 3.58 to − 1.82)	−0.78
ERQ total score
T0	25.4 ± 4.3	25.1 ± 4.5	0.42	0.67	−0.30 (− 1.42 to + 0.82)	−0.07
T1	27.3 ± 4.0	29.8 ± 3.8	4.2	<0.001	+ 2.50 (+ 1.51 to + 3.49)	0.64
T2	29.5 ± 3.9	33.8 ± 3.7	7.02	<0.001	+ 4.30 (+ 3.33 to + 5.27)	1.13
ΔT2–T0	+ 4.1 ± 1.8	+ 8.7 ± 2.1	7.02	<0.001	+ 4.60 (+ 4.10 to + 5.10)	2.35
Reappraisal
T0	16.8 ± 3.2	16.6 ± 3.4	0.3	0.76	−0.20 (− 1.04 to + 0.64)	−0.06
T1	18.1 ± 3.0	20.5 ± 2.8	4.15	<0.001	+ 2.40 (+ 1.66 to + 3.14)	0.83
T2	19.8 ± 2.7	23.1 ± 2.5	6.8	<0.001	+ 3.30 (+ 2.64 to + 3.96)	1.27
ΔT2–T0	+ 3.0 ± 1.5	+ 6.5 ± 1.8	7	<0.001	3.50 (+ 3.08 to + 3.92)	2.11
Suppression
T0	8.6 ± 1.5	8.5 ± 1.6	0.4	0.69	−0.10 (− 0.49 to + 0.29)	−0.06
T1	9.2 ± 1.4	9.3 ± 1.3	0.47	0.64	+ 0.10 (− 0.24 to + 0.44)	0.07
T2	9.7 ± 1.3	10.7 ± 1.2	3.2	0.002	+ 1.00 (+ 0.68 to + 1.32)	0.8
ΔT2–T0	+ 1.1 ± 0.8	+ 2.2 ± 1.0	4.00	<0.001	+ 1.10 (+ 0.87 to + 1.33)	1.21

DERS Difficulties in Emotion Regulation Scale, with a total score ranging from 36 to 180 points. The higher the score, the more severe the difficulty in emotion regulation. ERQ Emotion Regulation Questionnaire, with a total score of 10–70 points, including cognitive reassessment (6 items, 6–42 points) and expression suppression (4 items, 4–28 points). The higher the score, the better the ability of emotion regulation. Δ T2-T0 represents the change from baseline over 8 weeks

The ERQ results paralleled the DERS findings: baseline ERQ did not differ (25.4 ± 4.3 vs. 25.1 ± 4.5, P = 0.67), but the intervention group showed greater increases at T1 (29.8 ± 3.8 vs. 27.3 ± 4.0; I − C + 2.50, P < 0.001) and T2 (33.8 ± 3.7 vs. 29.5 ± 3.9; I − C + 4.30, P < 0.001), with greater 8-week gains (ΔT2–T0 + 8.7 ± 2.1 vs. +4.1 ± 1.8; P < 0.001). Improvements were observed in both the ERQ subscales—cognitive reappraisal (Δ + 6.5 vs. +3.0) and suppression (Δ + 2.2 vs. +1.1)—which was consistent with increased use of adaptive regulation strategies after the group intervention (Table 3).

Taken together, the retrospective data indicate that, compared with standard care, participation in emotion regulation–based group nursing interventions was associated with significantly greater improvements in both depressive symptoms and multiple domains of emotion regulation.

Frequency and severity of NSSI

At baseline, the weekly NSSI frequency of the groups was comparable (control 2.2 ± 0.6 vs. intervention 2.4 ± 0.7; t = 1.64, P = 0.10). By week 4 (T1), the frequency was lower in the intervention arm (0.9 ± 0.4 vs. 1.6 ± 0.5; t = 7.53, P < 0.001) and remained lower at week 8 (T2: 0.5 ± 0.3 vs. 1.3 ± 0.4; t = 5.12, P < 0.001). The 8-week change favored the intervention (ΔT2–T0 − 1.9 ± 0.4 vs. −0.9 ± 0.3; mean difference − 1.00; 95% CI − 1.09 to − 0.91; t = 7.53, P < 0.001; Cohen’s d = − 2.83), indicating a large effect on reducing NSSI frequency.

Severity scores were also similar at baseline (1.7 ± 0.5 vs. 1.8 ± 0.6; t = 1.00, P = 0.32) but declined more in the intervention group at T1 and T2 (T1: 0.9 ± 0.4 vs. 1.4 ± 0.5; T2: 0.4 ± 0.3 vs. 0.9 ± 0.4; both P < 0.001). The 8-week reduction in severity was greater with the intervention (ΔT2–T0 − 1.4 ± 0.3 vs. −0.8 ± 0.2; mean difference − 0.60; 95% CI − 0.66 to − 0.54; t = 5.34, P < 0.001; Cohen’s d = − 2.35) (Table 4). Collectively, these findings indicate that participation in emotion regulation–based group nursing interventions was associated with a lower frequency and severity of NSSI behaviors among adolescents.

Table 4.

Comparison of the frequency and severity of NSSI behavior before and after intervention between the two groups of patients

Indicator/time point	Control group (n = 120)	Intervention group (n = 120)	t	P	Mean diff (I − C), 95% CI	Cohen’s d
NSSI frequency (times per week)
T0	2.2 ± 0.6	2.4 ± 0.7	1.64	0.1	+ 0.20 (0.03 to 0.37)	0.31
T1	1.6 ± 0.5	0.9 ± 0.4	7.53	<0.001	−0.70 (− 0.82 to − 0.58)	−1.55
T2	1.3 ± 0.4	0.5 ± 0.3	5.12	<0.001	−0.80 (− 0.89 to − 0.71)	−2.26
ΔT2–T0	–0.9 ± 0.3	–1.9 ± 0.4	7.53	<0.001	−1.00 (− 1.09 to − 0.91)	−2.83
NSSI severity score
T0	1.7 ± 0.5	1.8 ± 0.6	1.00	0.32	+ 0.10 (− 0.04 to 0.24)	0.18
T1	1.4 ± 0.5	0.9 ± 0.4	6.11	<0.001	−0.50 (− 0.62 to − 0.38)	−1.10
T2	0.9 ± 0.4	0.4 ± 0.3	6.11	<0.001	−0.50 (− 0.59 to − 0.41)	−1.41
ΔT2–T0	–0.8 ± 0.2	–1.4 ± 0.3	5.34	<0.001	−0.60 (− 0.66 to − 0.54)	−2.35

T0 represents the baseline time, T1 is at 4 weeks of intervention, and T2 is at 8 weeks of intervention. ΔT2 - T0 represents the average change from baseline (T0) at 8 weeks (T2)

Adverse events and nursing satisfaction

In this retrospective cohort, adverse events were uncommon and did not differ significantly between groups: control 5/120 (4.2%) versus intervention 3/120 (2.5%) (χ² = 0.23, P = 0.63). The specific events in the control group were mood deterioration (2, 1.7%), increased NSSI (1, 0.8%) and other reactions (2, 1.7%); the intervention group reported mood deterioration in one patient (0.8%) and two other reactions (1.7%), with no aggravated NSSI observed (Table 5).

Table 5.

Comparison of adverse events and nursing satisfaction between the two groups of patients

Indicator	Control group (n = 120)	Intervention group (n = 120)	χ²/t	P	Mean diff (I − C), 95% CI	Cohen’s d
Adverse events
Deterioration of mood	2 (1.7%)	1 (0.8%)	χ²=0.34	0.56	—	—
NSSI behavior intensifies	1 (0.8%)	0 (0.0%)	χ²=1.00	0.32	—	—
Others	2 (1.7%)	2 (1.7%)	χ²=0.00	1	—	—
Overall incidence rate	5 (4.2%)	3 (2.5%)	χ²=0.23	0.63	—	—
Nursing satisfaction
Average score	3.8 ± 0.5	4.6 ± 0.4	t = 8.17	<0.001	+ 0.80 (0.68 to 0.92)	1.77

The nursing satisfaction was measured using a 5-point Likert scale (1 = very dissatisfied, 5 = very satisfied), which was filled out anonymously by the patients at the end of the intervention (T2)

Nursing satisfaction (5-point Likert scale) was greater in the intervention arm at T2 (4.6 ± 0.4 vs. 3.8 ± 0.5; t = 8.17, P < 0.001), with a mean difference of + 0.80 (95% CI 0.68–0.92, Cohen’s d = 1.77), indicating a large effect (Table 5). Overall, the addition of emotion-regulation–based group nursing was not associated with increased adverse events and was linked to substantially greater patient satisfaction.

Discussion

Adolescence represents a critical period for both the development of emotion regulation skills and the emergence of mental health challenges. Rates of depression and NSSI have been increasing in this age group [29, 30]. Emotional regulation is defined as the integrated capacity to identify, understand, and modulate one’s emotional experiences and responses, and deficits in these skills are implicated in the onset and persistence of both depression and NSSI [31, 32]. Although dialectical behavior therapy (DBT) and emotion regulation group therapy (ERGT) have demonstrated robust efficacy in adult and borderline personality disorder populations [33], there remains a lack of empirical research embedding these systematic emotion regulation approaches within routine adolescent nursing care in a group format. Traditional nursing largely emphasizes medication adherence and symptom monitoring and falls short in equipping patients with proactive emotion management and impulse control strategies [34, 35]. This study therefore sought to develop and evaluate a modular group nursing intervention focused on emotion regulation, assessing its combined impact on alleviating depressive symptoms, enhancing emotion regulation capacity, and reducing NSSI behaviors in adolescents. By integrating structured skills training into standard care, we aimed to address the gap between skill-building and conventional management in clinical practice.

The Children’s Depression Inventory (CDI) is a widely accepted instrument for gauging depressive severity in youth, encompassing emotional, cognitive, and functional domains [24]. Our findings revealed that the emotion regulation group intervention significantly accelerated symptom relief and amplified the magnitude of improvement. After eight weeks, the intervention group’s decline in CDI scores from baseline far exceeded that of the control group, reaching a mean CDI score of 13.2. According to established clinical cutoffs, a CDI score below 14 indicates minimal depressive symptoms in adolescents, and the observed reduction surpasses the minimal clinically important difference (MCID) reported in prior pediatric studies [36]. Similarly, an NSSI severity mean of 0.4 corresponds to infrequent, low-intensity self-injurious behavior, suggesting clinically meaningful improvements in functional outcomes and daily functioning [37]. The cognitive restructuring module enables participants to recognize and challenge negative automatic thoughts—thereby curtailing the duration of dysphoric states [38]—a conclusion supported by meta-analytic data affirming cognitive restructuring, behavioral activation, and CBT as effective treatments for adult depression [39]. Concurrently, the emotional awareness and mindfulness components increase sensitivity to physiological and psychological cues, effectively interrupting the cycle of depressive reactivity [40]. Indeed, multiple studies have shown that mindfulness enhances awareness and recovery from negative affect, while cognitive restructuring reshapes maladaptive interpretations, bolstering tolerance for and regulation of negative emotions [41, 42].

Parallel improvements were observed in emotion regulation capacity: both the Difficulties in Emotion Regulation Scale (DERS) and the Emotion Regulation Questionnaire (ERQ) demonstrated markedly greater gains in the intervention group than in the control group, underscoring the intervention’s multidimensional impact. Notably, the scores of the ERQ suppression subscale slightly increased. While suppression is generally linked to maladaptive outcomes, in the context of guided emotion regulation training and structured coping strategies, this slight elevation may reflect adaptive use of controlled suppression rather than unregulated emotional avoidance, although caution is warranted in interpreting this change [43, 44]. Shilyansky et al. reported that pharmacotherapy alone failed to enhance cognitive reappraisal skills [45]; our data confirm that structured emotion regulation training effectively fills this gap. Dimension-specific analyses revealed particularly pronounced improvements in nonacceptance, goals, and impulse control, further validating the relevance and practicality of our module designs.

NSSI frequency and severity ratings reflect the intensity of self-injurious impulses and the resulting tissue damage, and these behaviors are often closely linked to emotional dysregulation and poor impulse control [46]. In our study, reductions in both NSSI frequency and severity were significantly greater in the intervention group than in the control group, indicating that participation in the group nursing intervention based on emotion regulation training was associated with fewer self-harm urges. This finding aligns with previous research suggesting that nonsuicidal self‐injury among adolescents with severe depression often involves emotion‐regulation and interpersonal functions and is associated with deficits in emotion regulation, alexithymia, and histories of childhood trauma [47]. Such evidence underscores the importance of targeting these underlying psychological mechanisms to reduce NSSI. A randomized trial demonstrated that progressive muscle relaxation (PMR) and deep breathing significantly decrease electrodermal activity and other physiological stress markers, thereby promoting psychophysiological relaxation [48], and that PMR can rapidly alleviate anxiety, agitation, and aggression in psychiatric inpatients [49]. Integrating PMR and breath‐control exercises into the emotion regulation module may therefore decrease physiological arousal during self‐harm urges, extending the window for inhibitory control and enhancing overall self‐regulation.

Furthermore, during the group-sharing component, participants recount personal experiences and emotional struggles and receive empathy, understanding, and feedback from peers. This process strengthens social support and a sense of belonging, which have been shown to facilitate cognitive shifts and behavior change through group norms and collective efficacy [50]. In our study, the intervention group also exhibited greater nursing satisfaction and better adherence. The participants reported that sharing personal experiences within the group helped them feel understood and supported, reinforcing their engagement and motivation for self-regulation [51]. Moreover, the group format offers superior accessibility and cost‐effectiveness, as it does not rely on intensive individualized psychological resources. Nurses can feasibly implement ERT in resource-limited settings by adjusting training schedules and session durations; for example, interventions can be delivered in 30–60 min weekly sessions over 8–12 weeks, allowing integration with routine care [52, 53]. This scalability makes it well suited for implementation in schools, community centers, or other resource‐limited settings, thereby paving the way for routine, large‐scale deployment of emotion regulation–focused interventions.

This study innovatively integrated structured emotion regulation training into routine adolescent nursing care and empirically demonstrated its multidimensional benefits in alleviating depressive symptoms, enhancing emotion regulation skills, and reducing NSSI behaviors. Our primary contribution lies in developing a standardized training manual and implementation protocol that combines pharmacological and non-pharmacological techniques, providing actionable guidance for community hospitals, schools, and health centers.

Limitations and prospects

The limitations of this study include the following: (1) it was a single-center, retrospective, nonrandomized design, with group allocation based on clinical records, which may introduce selection bias and unmeasured confounding; (2) the assessments and outcome data were derived from medical records and patient self-reports, which may be susceptible to information bias; (3) adherence to the intervention was not monitored (e.g., by recording completion of each session or using procedural checklists), and the integrity of the intervention was not verified; (4) potential variations in medication dosage during the study period may have influenced the outcomes; (5) follow-up was limited to eight weeks, precluding evaluation of long-term intervention effects; and (6) the sample was drawn from a veterans’ hospital in northern China, which may limit generalizability to civilian populations or adolescents from different cultural backgrounds. Moreover, progressive muscle relaxation (PMR) may assist in modulating arousal during self-injurious impulses; however, future research should incorporate physiological or neurocognitive measures (e.g., heart rate variability, cortisol levels, or fMRI assessments) to validate these effects. We also acknowledge that the comparator in this study represents a relatively intensive standard of nursing care, which may limit its generalizability to settings with less comprehensive routine nursing services. Additionally, although both male and female adolescents were included, the study was not powered for gender-stratified analyses. Given the known gender differences in depression and NSSI, future studies should examine gender as a potential moderator of intervention effects. Future research should evaluate interventions in multicenter or pragmatic studies across settings with differing baseline nursing intensities, using stratified, cluster, stepped-wedge, or randomized designs as appropriate. Such studies should also include longer follow-up periods, objective adherence and fidelity monitoring, and physiological or neurocognitive measures to strengthen causal inference and external validity.

Conclusion

In this retrospective cohort, participation in a group nursing intervention involving emotion regulation training was associated with greater reductions in depressive symptoms, improvements in emotion regulation measures, and a lower frequency and severity of NSSI behaviors than standard care. The causal relationship and wide applicability need to be further verified in prospective, randomized controlled trials.

Abbreviation	Full Term	Description
AEs	Adverse Events	Negative psychological or physiological reactions occurring during the intervention.
CDI	Children’s Depression Inventory	A 27-item self-report scale assessing depressive symptoms in children and adolescents.
CI	Confidence Interval	Statistical range indicating the precision of an estimated effect.
CTCAE (v5.0)	Common Terminology Criteria for Adverse Events (version 5.0)	Standardized classification and severity grading scale for adverse events.
DERS	Difficulties in Emotion Regulation Scale	A 36-item questionnaire assessing multiple dimensions of emotion regulation difficulties.
DSM-5	Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition	The fifth edition of the classification system for mental disorders by the American Psychiatric Association.
Δ (Delta)	Change or Difference	Represents score changes between time points (e.g., ΔT2–T0).
ERQ	Emotion Regulation Questionnaire	A 10-item measure assessing cognitive reappraisal and expressive suppression strategies.
NSSI	Non-Suicidal Self-Injury	Self-inflicted physical injury without suicidal intent.
P	P Value	Statistical measure indicating the probability of obtaining the observed results by chance.
SD	Standard Deviation	Measure of data dispersion around the mean.
SPSS	Statistical Package for the Social Sciences	Statistical software used for data analysis (version 26.0 in this study).
SSRIs	Selective Serotonin Reuptake Inhibitors	A class of antidepressant medications commonly used in depression treatment.
T0, T1, T2	Time Points	Baseline (T0), Week 4 (T1), and Week 8 (T2) assessment time points.

Authors’ contributions

Qian Wang wrote the main manuscript text. ChunAi Lu prepared figure. Haihua Cui prepared tables. Ran An analyzed the results. All authors reviewed the manuscript.

Funding

The work was not funded by any funding.

Data availability

The experimental data used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Ethics Committee of Hebei Third Veterans’ Preferential Hospital (Approval No. 2022004). Given its retrospective design and the use of fully anonymized data, the requirement for informed consent was waived. All procedures adhered to the principles of the Declaration of Helsinki and relevant medical ethics guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.