Factors influencing non-suicidal self-injury in adolescents with mood disorders: Focusing on gender differences and only-child vs non-only-child disparities

Abstract

BACKGROUND

Non-suicidal self-injury (NSSI) refers to intentional, self-inflicted damage to body tissues without suicidal intent and for non-socially sanctioned purposes. It has become a pressing global public health issue, particularly among adolescents with mood disorders.

AIM

To investigate the prevalence of NSSI in adolescents with mood disorders and identify key influencing factors, with a focus on gender and sibling status (only-child vs non-only-child).

METHODS

A cross-sectional study was conducted among 620 adolescents (12-18 years) diagnosed with mood disorders, who were recruited from psychiatric clinics. Sociodemographic, clinical, psychological, and family functioning data were collected using standardized questionnaires. Univariate analyses and binary logistic regression were applied to explore factors associated with NSSI.

RESULTS

The overall prevalence of NSSI was 72.26%. Multivariate regression identified female gender [odds ratio (OR) = 1.87, 95% confidence interval (CI): 1.24-2.82], only-child status (OR = 1.54, 95%CI: 1.03-2.29), longer mood disorder duration (OR = 1.08, 95%CI: 1.04-1.12), comorbid obsessive-compulsive symptoms (OR = 2.13, 95%CI: 1.38-3.29), low self-warmth (OR = 0.92, 95%CI: 0.89-0.94), and poor family cohesion (OR = 0.94, 95%CI: 0.91-0.96) as independent predictors. Gender and sibling status also moderated the frequency and methods of NSSI: Females and only-children exhibited a higher NSSI frequency and preferred cutting/scratching, while males favored hitting/head-banging.

CONCLUSION

NSSI is highly prevalent in adolescents with mood disorders. Gender and sibling status are significant sociodemographic predictors, with distinct psychosocial mechanisms contributing to the risk of NSSI. Tailored interventions that address these factors may help reduce the burden of NSSI in this population.

Key Words: Non-suicidal self-injury; Adolescents; Mood disorders; Gender difference; Sibling status

Core Tip: This study investigated non-suicidal self-injury in 620 adolescents with mood disorders, highlighting the roles of gender and sibling status. Female and only-child adolescents exhibited higher non-suicidal self-injury prevalence and frequency, with distinct psychological and familial predictors. Findings emphasize the importance of gender- and sibling-specific screening and interventions, including trauma-informed care, self-compassion training, cognitive reappraisal, and family support enhancement, to prevent and mitigate self-injurious behaviors in this vulnerable population.

INTRODUCTION

Non-suicidal self-injury (NSSI) refers to intentional, self-inflicted damage to body tissues without suicidal intent and for non-socially sanctioned purposes[1]. It has become a pressing global public health issue, particularly among adolescents with mood disorders. Epidemiological studies show that the lifetime prevalence of NSSI in the general adolescent population ranges from 17.2% to 22.3%, but this rate surges to 34.0%-62.9% in adolescents diagnosed with mood disorders, including major depressive disorder and bipolar disorder[2]. This elevated risk carries severe consequences: Adolescents with mood disorders who engage in NSSI face a 9-fold higher chance of subsequent unnatural death, including suicide and fatal substance poisoning[3]. They also experience long-term adverse outcomes such as persistent mental health impairments, tobacco dependence, and illicit drug use, highlighting the urgency of addressing NSSI in this vulnerable group[4].

Mood disorders in adolescence are marked by core symptoms of emotional dysregulation, including persistent depressed mood, anxiety, and affective instability[5]. For these adolescents, NSSI often serves as a maladaptive coping strategy to alleviate overwhelming negative emotions or to regain a perceived sense of control. However, the multifactorial mechanisms underlying NSSI in this population are not fully understood[6]. Existing research has identified a range of potential influencing factors across multiple domains. Individual psychological factors such as low self-compassion and poor emotional regulation are closely linked to NSSI - adolescents with mood disorders who struggle to regulate emotions or lack self-warmth are more likely to engage in self-injurious behaviors. Clinical factors, including the severity of depressive symptoms and comorbid obsessive-compulsive symptoms, also play a role; greater symptom severity and comorbidity increase NSSI risk. Behavioral factors such as problematic mobile phone use and substance use further contribute, as excessive phone dependence and substance consumption are associated with heightened psychological distress and impulsive behaviors that precede NSSI[7]. Environmental factors, such as adverse childhood maltreatment and poor family functioning, are additional key correlates - experiences of abuse or neglect in childhood and strained family relationships exacerbate emotional vulnerability and increase the likelihood of NSSI[8].

Two demographic factors demand specific attention due to their contextual significance and clinical relevance: Gender and sibling status (only-child vs non-only-child). Gender disparities in NSSI are consistently observed, with female adolescents with mood disorders showing higher NSSI prevalence (46.3%-80.6%) than males (19.4%-53.7%)[9]. This difference stems from multiple sources, including differential emotional socialization - females tend to exhibit greater emotional reactivity and rely on internalizing coping strategies like NSSI to manage distress, while males more frequently use externalizing behaviors[10]. Neurobiological differences in stress response systems and societal norms that restrict emotional expression further widen this gap. Females also report more frequent NSSI episodes and a broader range of self-injury methods, such as cutting and scratching, compared to males[11].

Sibling status, a unique contextual factor in regions with historical family planning policies, introduces another layer of complexity. Only children with mood disorders often face heightened parental expectations, a lack of sibling interaction, and fewer opportunities to develop social skills[12]. These experiences can exacerbate psychological distress and emotional dysregulation, which are key precursors to NSSI. Preliminary findings indicate that only children with mood disorders report higher NSSI frequency and more diverse self-injury methods than non-only children[13]. In contrast, non-only children may benefit from sibling relationships that provide emotional support and opportunities for developing conflict-resolution skills, potentially buffering against NSSI risk[14].

Despite these insights, few studies have systematically integrated the analysis of gender, sibling status, and other key factors - such as childhood maltreatment, emotional regulation, and family functioning - to disentangle their independent and interactive effects on NSSI in adolescents with mood disorders. Given the heavy burden of NSSI in this group, clarifying these relationships is essential for developing targeted prevention and intervention strategies. The present cross-sectional study aims to address this gap by comprehensively investigating the factors influencing NSSI, including gender, sibling status, psychological distress, emotional regulation, childhood maltreatment, and family functioning, among adolescents with mood disorders. By identifying these correlates, this study seeks to provide empirical evidence to inform tailored approaches that reduce NSSI risk and improve long-term outcomes for this vulnerable population.

MATERIALS AND METHODS

Study design and participants

This cross-sectional study was conducted in the Suzhou Guangji Hospital and Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine from 2023 to 2025. Participants were adolescents aged 12-18 years with a primary diagnosis of mood disorders, in accordance with the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria for major depressive disorder or bipolar disorder[15]. Eligibility criteria included: (1) Age 12-18 years; (2) Confirmation of mood disorder diagnosis by two attending psychiatrists via structured clinical interviews; (3) Ability to complete self-report questionnaires independently (or with minimal assistance from researchers for reading difficulties, without altering responses); and (4) Written informed consent from both the adolescents and their legal guardians. Exclusion criteria were: (1) Comorbid psychotic disorders (e.g., schizophrenia), intellectual disability, or organic brain disease; (2) History of electroconvulsive therapy within 3 months prior to enrollment; and (3) Incomplete clinical or questionnaire data.

Sample size was determined based on the primary outcome (presence of NSSI) and key predictors (gender and sibling status). Using a two-sided α of 0.05, power of 0.80, and an estimated NSSI prevalence of 75.6% among adolescents with mood disorders[16], the minimum required sample size was calculated as 558 using the formula for cross-sectional studies. Accounting for a potential 15% attrition rate, a total of 657 participants were initially recruited, with 620 completing all assessments (effective response rate: 94.4%); among these, 448 were classified into the NSSI group.

This study was approved by the Ethics Committee of Suzhou Guangji Hospital (approval No. SGH-ER-2023-069). Informed consent was obtained from the legal guardians/close relatives of all the participants, and all adolescent participants also provided assent prior to participation. All methods were carried out in accordance with Declaration of Helsinki.

Data collection instruments

Sociodemographic and clinical questionnaire: A structured questionnaire was developed to collect sociodemographic data, including age, gender, sibling status (only-child vs non-only-child), family monthly income, parental marital status, parental education level, and residential area (urban vs rural). Clinical data included duration of mood disorder, comorbidities (e.g., obsessive-compulsive symptoms), and history of psychiatric treatment.

Assessment of NSSI: NSSI was evaluated using the Ottawa Self-Injury Inventory (OSI)[17], a validated tool to assess the presence, frequency, methods, and functional motives of NSSI in adolescents. The OSI captures NSSI behaviors (e.g., cutting, scratching, hitting, burning) occurring in the past month and past year, with frequency categorized as “0 times”, “1-4 times”, or “≥ 5 times”. Participants who reported ≥ 1 NSSI episode in the past year were defined as the “NSSI group”; those with no NSSI in the past year were the “non-NSSI group”. The OSI has demonstrated good internal consistency (Cronbach’s α = 0.84) in previous adolescent samples with mood disorders.

Psychological and emotional functioning assessments: Self-Compassion Scale (SCS): A 26-item scale measuring self-compassion across six dimensions, with scores aggregated into two factors: Self-warmth (adaptive self-compassion) and self-coldness (maladaptive self-criticism)[18]. Responses are rated on a 5-point Likert scale (1 = “almost never” to 5 = “almost always”), with higher self-warmth scores indicating stronger adaptive self-compassion. Cronbach’s α for the SCS was 0.91 in this study.

Emotion Regulation Questionnaire (ERQ): A 10-item scale assessing two emotion regulation strategies: Cognitive reappraisal (6 items) and expressive suppression (4 items). Responses are scored on a 7-point Likert scale (1 = “strongly disagree” to 7 = “strongly agree”), with higher cognitive reappraisal scores indicating more frequent use of adaptive emotion regulation. The ERQ showed good reliability (Cronbach’s α = 0.88) in the present sample.

Kessler Psychological Distress Scale (K-10): A 10-item scale measuring psychological distress over the past month, with responses ranging from 1 (“none of the time”) to 5 (“all of the time”)[19]. Total scores range from 10 to 50, with higher scores indicating greater distress. Cronbach’s α for the K-10 was 0.89.

Childhood Trauma Questionnaire (CTQ): A 28-item scale evaluating five types of childhood maltreatment (emotional abuse, physical abuse, sexual abuse, emotional neglect, physical neglect)[20]. Responses are rated on a 5-point scale (1 = “never true” to 5 = “very often true”), with total scores ranging from 25 to 125. Higher scores indicate more severe maltreatment. Cronbach’s α for the CTQ was 0.83.

Family and social environment assessment

The Family Adaptation and Cohesion Evaluation Scales III (FACES III) was used to measure family functioning[21], focusing on cohesion (emotional bonding among family members) and adaptability (family’s ability to adjust to changes). The scale includes 20 items rated on a 5-point Likert scale (1 = “almost never” to 5 = “almost always”), with higher scores indicating healthier family functioning. Cronbach’s α for FACES III was 0.85.

Data collection procedures

Trained research assistants (psychology graduate students or psychiatric nurses) administered the questionnaires in a quiet clinical setting. Participants completed self-report measures independently; for those with reading difficulties, assistants read items aloud without offering interpretive guidance. Clinical data (e.g., diagnosis, comorbidities) were extracted from electronic medical records and verified by attending psychiatrists. All data were double-entered into Epidata 3.1 and cross-checked for accuracy.

Statistical analysis

Data were analyzed using SPSS 26.0 and Mplus 8.3. Categorical variables were described as n (%), and continuous variables as mean ± SD for normal distribution or median and interquartile range for non-normal distribution. χ² tests were used to compare categorical variables between NSSI and non-NSSI groups, while independent samples t-tests or Mann-Whitney U tests were applied for continuous variables, with Bonferroni correction adjusting the significance level to 0.0083 for multiple comparisons. Spearman’s rank correlation was conducted to examine bivariate associations between NSSI frequency and key factors. Binary logistic regression was performed to identify independent factors influencing NSSI, with NSSI presence as the dependent variable; potential confounders and key predictors were entered via a stepwise method, and interaction terms were included to explore effect modification. A two-sided P less than 0.05 was considered statistically significant.

RESULTS

Univariate analysis of NSSI behavior in adolescents with mood disorders

A total of 620 adolescents with mood disorders were included, with 448 (72.258%) in the NSSI group and 172 (27.742%) in the non-NSSI group. Univariate analysis showed significant differences between the two groups in sociodemographic characteristics, clinical factors, psychological assessments, and family functioning (all P < 0.05). Detailed results are presented in Table 1.

Table 1 Univariate analysis of non-suicidal self-injury behavior in adolescents with mood disorders, n (%)/mean ± SD.

Variables	NSSI group (n = 448)	Non-NSSI group (n = 172)	t/χ2	P value
Sociodemographic characteristics
Age (years)	15.36 ± 1.82	16.12 ± 1.75	-4.58	< 0.001
Gender			12.63	< 0.001
Male	126 (28.13)	68 (39.53)
Female	322 (71.82)	104 (60.47)
Sibling status			8.95	0.003
Only-child	296 (66.07)	92 (53.49)
Non-only-child	152 (33.93)	80 (46.52)
Family monthly income			15.72	0.001
< 5000	108 (24.11)	22 (12.79)
5000-10000	212 (47.32)	86 (50.00)
> 10000	128 (28.57)	64 (37.21)
Parental marital status			10.37	0.001
Married	364 (81.2500)	150 (87.2093)
Divorced/remarried	84 (18.7500)	22 (12.7907)
Parental education level			14.89	< 0.001
≤ High school	204 (45.5357)	58 (33.7209)
> High school	244 (54.4643)	114 (66.2791)
Residential area			1.70	0.184
Urban	312 (69.6429)	126 (73.2558)
Rural	136 (30.3571)	46 (26.7442)
Clinical factors
Duration of mood disorder (months)	14.82 ± 6.35	11.26 ± 5.87	6.73	< 0.001
Comorbid obsessive-compulsive symptoms	186 (41.5179)	42 (24.4186)	15.28	< 0.001
Psychological assessments
SCS self-warmth	31.85 ± 7.62	37.42 ± 8.15	-7.84	< 0.001
SCS self-coldness	51.63 ± 8.24	45.38 ± 7.96	8.36	< 0.001
ERQ cognitive reappraisal	20.76 ± 6.43	25.89 ± 6.71	-9.21	< 0.001
ERQ expressive suppression	17.58 ± 5.12	16.24 ± 4.87	3.02	0.002
K-10 score	36.82 ± 7.45	29.35 ± 6.98	11.54	< 0.001
CTQ total score	53.76 ± 10.82	45.21 ± 9.76	8.97	< 0.001
Family functioning
FACES III cohesion	42.38 ± 8.56	48.65 ± 9.12	-7.53	< 0.001
FACES III adaptability	38.74 ± 7.93	44.21 ± 8.45	-7.01	< 0.001

NSSI: Non-suicidal self-injury; SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; K-10: Kessler Psychological Distress Scale; CTQ: Childhood Trauma Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

Multivariate logistic regression analysis of NSSI behavior in adolescents with mood disorders

Binary logistic regression analysis was performed with NSSI presence as the dependent variable and variables with P < 0.05 in univariate analysis as independent variables (stepwise entry method). The model was statistically significant (χ² = 218.64, df = 12, P < 0.001), with a Nagelkerke R² of 0.587, demonstrating good explanatory power. The results showed that female gender [odds ratio (OR) = 1.87, 95% confidence interval (CI): 1.24-2.82], only-child status (OR = 1.54, 95%CI: 1.03-2.29), longer duration of mood disorder (OR = 1.08, 95%CI: 1.04-1.12), comorbid obsessive-compulsive symptoms (OR = 2.13, 95%CI: 1.38-3.29), lower SCS self-warmth (OR = 0.92, 95%CI: 0.89-0.94), higher SCS self-coldness (OR = 1.05, 95%CI: 1.02-1.08), lower ERQ cognitive reappraisal (OR = 0.90, 95%CI: 0.87-0.93), higher K-10 score (OR = 1.10, 95%CI: 1.07-1.13), higher CTQ total score (OR = 1.03, 95%CI: 1.01-1.05), and lower FACES III cohesion (OR = 0.94, 95%CI: 0.91-0.96) were independently associated with NSSI in adolescents with mood disorders (all P < 0.05). Parental marital status, parental education level, family monthly income, and ERQ expressive suppression were not significantly associated with NSSI (all P > 0.05). Detailed results are presented in Table 2.

Table 2 Multivariate logistic regression analysis of non-suicidal self-injury behavior in adolescents with mood disorders.

Variables	B	SE	Wald χ2	df	P value	OR (95%CI)
Gender (female = 1, male = 0)	0.629	0.214	8.643	1	0.003	1.876 (1.245-2.823)
Sibling status (only-child = 1, non-only-child = 0)	0.433	0.208	4.267	1	0.039	1.542 (1.036-2.298)
Duration of mood disorder (months)	0.081	0.020	15.32	1	< 0.001	1.084 (1.041-1.128)
Comorbid obsessive-compulsive symptoms (yes = 1, no = 0)	0.757	0.231	10.89	1	0.001	2.135 (1.384-3.298)
SCS self-warmth	-0.081	0.014	33.06	1	< 0.001	0.921 (0.894-0.949)
SCS self-coldness	0.055	0.013	17.18	1	< 0.001	1.056 (1.029-1.085)
ERQ cognitive reappraisal	-0.101	0.017	34.57	1	< 0.001	0.903 (0.872-0.935)
ERQ expressive suppression	0.021	0.018	1.334	1	0.248	1.021 (0.985-1.059)
K-10 score	0.097	0.015	41.89	1	< 0.001	1.102 (1.071-1.134)
CTQ total score	0.037	0.009	14.98	1	< 0.001	1.038 (1.019-1.058)
FACES III cohesion	-0.058	0.013	21.14	1	< 0.001	0.942 (0.918-0.967)
Parental marital status (divorced/remarried = 1, married = 0)	0.321	0.246	1.70	1	0.192	1.379 (0.854-2.223)
Parental education level (> high school = 1, ≤ high school = 0)	-0.284	0.201	2.002	1	0.157	0.752 (0.513-1.098)
Family monthly income (ref: < 5000 RMB)
5000-10000 RMB	-0.142	0.2387	0.3645	1	0.546	0.866 (0.538-1.394)
> 10000 RMB	-0.312	0.2543	1.4987	1	0.221	0.731 (0.449-1.192)
Constant	-12.84	1.876	46.53	1	< 0.001	2.6741 × 10-6

OR: Odds ratio; CI: Confidence interval; SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; K-10: Kessler Psychological Distress Scale; CTQ: Childhood Trauma Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

Gender comparison of adolescents with mood disorders and NSSI and univariate analysis of NSSI characteristics between male and female adolescents

Within the NSSI group (n = 448), there were 126 males (28.12%) and 322 females (71.88%). Univariate analysis revealed significant gender differences in family monthly income, parental education level, K-10 score, CTQ total score, SCS self-warmth, SCS self-coldness, ERQ cognitive reappraisal, and FACES III cohesion (all P < 0.05). Females were more likely than males to have lower family monthly income and parental education level, higher K-10 scores and CTQ total scores, lower SCS self-warmth, higher SCS self-coldness, lower ERQ cognitive reappraisal, and lower FACES III cohesion. No significant differences were observed in age, sibling status, parental marital status, residential area, duration of mood disorder, comorbid obsessive-compulsive symptoms, or ERQ expressive suppression (all P > 0.05). Detailed results are presented in Table 3.

Table 3 Univariate analysis of non-suicidal self-injury characteristics between male and female adolescents with mood disorders, n (%)/mean ± SD.

Variables	Male (n = 126)	Female (n = 322)	t/χ2	P value
Sociodemographic characteristics
Age (years)	15.48 ± 1.76	15.31 ± 1.84	0.89	0.371
Sibling status			2.13	0.143
Only-child	82 (65.08)	214 (66.46)
Non-only-child	44 (34.92)	108 (33.54)
Family monthly income			10.24	0.006
< 5000	22 (17.46)	86 (26.71)
5000-10000	64 (50.79)	148 (45.96)
> 10000	40 (31.75)	88 (27.33)
Parental marital status			1.56	0.210
Married	104 (82.54)	260 (80.75)
Divorced/remarried	22 (17.46)	62 (19.25)
Parental education level			8.97	0.002
≤ High school	48 (38.10)	156 (48.76)
> High school	78 (61.90)	166 (51.24)
Residential area			0.98	0.321
Urban	88 (69.84)	224 (69.57)
Rural	38 (30.16)	98 (30.43)
Clinical factors
Duration of mood disorder (months)	14.62 ± 6.51	14.91 ± 6.28	-0.45	0.646
Comorbid obsessive-compulsive symptoms	52 (41.27)	134 (41.61)	0.006	0.933
Psychological assessments
SCS self-warmth	33.96 ± 7.28	30.82 ± 7.65	4.12	< 0.001
SCS self-coldness	49.25 ± 8.03	52.58 ± 8.21	-4.01	< 0.001
ERQ cognitive reappraisal	22.58 ± 6.17	20.03 ± 6.45	4.23	< 0.001
ERQ expressive suppression	17.24 ± 5.03	17.72 ± 5.15	-0.98	0.3245
K-10 score	34.52 ± 7.18	37.68 ± 7.41	-4.35	< 0.001
CTQ total score	50.83 ± 10.26	54.87 ± 10.89	-3.84	< 0.001
Family functioning
FACES III cohesion	44.62 ± 8.31	41.45 ± 8.59	3.67	< 0.001
FACES III adaptability	39.21 ± 7.86	38.52 ± 7.95	0.89	0.372

SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; K-10: Kessler Psychological Distress Scale; CTQ: Childhood Trauma Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

Comparison of NSSI frequency and methods between male and female adolescents

Gender differences were observed in NSSI frequency and methods. Females were more likely to engage in high-frequency NSSI (≥ 5 times in past year) compared with males (χ² = 9.87, P = 0.007). Regarding NSSI methods, females showed higher rates of cutting, scratching, and interfering with wound healing, while males had higher rates of hitting and head-banging (all P < 0.05) (Figure 1). No significant gender differences were found in burning or hair-pulling (P > 0.05). Detailed results are presented in Table 4.

Figure 1 Sankey diagram of non-suicidal self-injury method preferences by gender. The diagram illustrates the proportional flow from participant gender (left nodes) to the specific non-suicidal self-injury methods they reported (right nodes). Band width corresponds to the relative frequency of endorsement. NSSI: Non-suicidal self-injury.

Table 4 Comparison of non-suicidal self-injury frequency and methods between male and female adolescents, n (%).

Variables	Male (n = 126)	Female (n = 322)	χ2	P value
NSSI frequency (past year)			9.874	0.007
1-4 times	58 (46.03)	102 (31.68)
≥ 5 times	68 (53.97)	220 (68.32)
NSSI methods
Cutting	64 (50.79)	202 (62.73)	6.784	0.009
Scratching	56 (44.44)	189 (58.70)	9.215	0.002
Hitting	65 (51.58)	148 (45.96)	1.568	0.210
Burning	22 (17.46)	58 (18.01)	0.042	0.835
Head-banging	49 (38.89)	102 (31.68)	2.894	0.089
Hair-pulling	32 (25.40)	82 (25.47)	0.001	0.982
Interfering with wound healing	38 (30.16)	138 (42.86)	8.124	0.004

NSSI: Non-suicidal self-injury.

Multivariate logistic regression analysis of NSSI frequency (≥ 5 times vs 1-4 times) by gender

Binary logistic regression was performed separately for males and females, with NSSI frequency (≥ 5 times = 1, 1-4 times = 0) as the dependent variable and factors showing statistical significance in the gender-stratified univariate analysis (Table 3) as independent variables. For males, higher K-10 score (OR = 1.1234, 95%CI: 1.0567-1.1942) and lower FACES III cohesion (OR = 0.9218, 95%CI: 0.8765-0.9692) were independently associated with high-frequency NSSI (all P < 0.05). For females, lower SCS self-warmth (OR = 0.9035, 95%CI: 0.8672-0.9413), higher CTQ total score (OR = 1.0456, 95%CI: 1.0187-1.0732), and lower ERQ cognitive reappraisal (OR = 0.8879, 95%CI: 0.8465-0.9317) were independently associated with high-frequency NSSI (all P < 0.05). Detailed results are presented in Table 5.

Table 5 Gender-specific multivariate logistic regression analysis of non-suicidal self-injury frequency.

Variables	Male (n = 126)				Female (n = 322)
	B	SE	OR (95%CI)	P value	B	SE	OR (95%CI)	P value
K-10 score	0.116	0.032	1.123 (1.056-1.194)	< 0.001	0.021	0.0207	1.021 (0.981-1.063)	0.298
CTQ total score	0.020	0.018	1.021 (0.984-1.058)	0.256	0.044	0.013	1.045 (1.018-1.073)	< 0.001
SCS self-warmth	-0.031	0.024	0.969 (0.922-1.017)	0.205	-0.101	0.018	0.903 (0.867-0.941)	< 0.001
SCS self-coldness	0.032	0.025	1.0333 (0.983-1.085)	0.192	0.028	0.019	1.029 (0.991-1.068)	0.136
ERQ cognitive reappraisal	-0.041	0.028	0.959 (0.906-1.016)	0.152	-0.119	0.023	0.887 (0.846-0.931)	< 0.001
FACES III cohesion	-0.081	0.031	0.921 (0.876-0.969)	0.001	-0.024	0.017	0.975 (0.942-1.010)	0.166
Family monthly income (> 10000 vs < 5000 RMB)	-0.214	0.321	0.807 (0.446-1.461)	0.498	-0.187	0.205	0.822 (0.556-1.215)	0.328
Parental education level (> high school vs ≤ high school)	-0.321	0.287	0.725 (0.428-1.223)	0.235	-0.215	0.189	0.807 (0.556-1.172)	0.258
Constant	-5.874	1.896	0.002 (0.000-0.038)	< 0.001	-2.145	1.023	0.116 (0.013-1.032)	0.035

OR: Odds ratio; CI: Confidence interval; K-10: Kessler Psychological Distress Scale; CTQ: Childhood Trauma Questionnaire; SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

Comparison of only-child and non-only-child adolescents with mood disorders and NSSI univariate analysis of NSSI characteristics between only-child and non-only-child adolescents

Within the NSSI group (n = 448), only-children showed significant differences compared with non-only-children in parental marital status, K-10 score, SCS self-warmth, SCS self-coldness, ERQ cognitive reappraisal, and FACES III cohesion between the two groups (all P < 0.05). Only-children had a higher proportion of parental divorce/remarriage, higher K-10 scores, lower SCS self-warmth, higher SCS self-coldness, lower ERQ cognitive reappraisal, and lower FACES III cohesion than non-only-children (Figure 2). No significant differences were observed in age, gender, family monthly income, parental education level, residential area, duration of mood disorder, comorbid obsessive-compulsive symptoms, ERQ expressive suppression, CTQ total score, or FACES III adaptability (all P > 0.05). Detailed results are presented in Table 6.

Figure 2 Comparison of psychological and family factors between only-child and non-only-child adolescents. A: Parental divorce rate; B: Psychological distress; C: Family cohesion; D: Self-warmth; E: Self-coldness; F: Cognitive reappraisal. Statistical annotations on the bars denote significant intergroup differences. ^aP < 0.05. SCS: Self-Compassion Scale; FACES: Family Adaptation and Cohesion Evaluation Scales; ERQ: Emotion Regulation Questionnaire.

Table 6 Univariate analysis of non-suicidal self-injury characteristics between only-child and non-only-child adolescents, n (%)/mean ± SD.

Variables	Only-child (n = 296)	Non-only-child (n = 152)	t/χ2	P value
Sociodemographic characteristics
Age (years)	15.32 ± 1.85	15.43 ± 1.78	-0.625	0.533
Gender			1.037	0.309
Male	82 (27.70)	44 (28.95)
Female	214 (72.30)	108 (71.05)
Family monthly income			4.216	0.122
< 5000	72 (24.32)	36 (23.68)
5000-10000	144 (48.65)	68 (44.74)
> 10000	80 (27.03)	48 (31.58)
Parental marital status			6.895	0.009
Married	229 (77.36)	135 (88.81)
Divorced/remarried	67 (22.64)	17 (11.18)
Parental education level			3.126	0.077
≤ High school	136 (45.95)	68 (44.74)
> High school	160 (54.05)	84 (55.26)
Residential area			0.587	0.444
Urban	206 (69.59)	106 (69.74)
Rural	90 (30.41)	46 (30.26)
Clinical factors
Duration of mood disorder (months)	14.91 ± 6.42	14.65 ± 6.21	0.439	0.661
Comorbid obsessive-compulsive symptoms	124 (41.89)	62 (40.79)	0.068	0.795
Psychological assessments
SCS self-warmth	30.75 ± 7.58	33.92 ± 7.26	-4.587	< 0.001
SCS self-coldness	52.86 ± 8.15	49.21 ± 8.03	4.321	< 0.001
ERQ cognitive reappraisal	19.85 ± 6.37	22.46 ± 6.12	-4.218	< 0.001
ERQ expressive suppression	17.62 ± 5.08	17.51 ± 5.17	0.214	0.830
K-10 score	37.95 ± 7.38	34.62 ± 7.15	4.785	< 0.001
CTQ total score	54.12 ± 10.76	53.15 ± 10.89	0.824	0.410
Family functioning
FACES III cohesion	40.85 ± 8.47	44.92 ± 8.23	-5.214	< 0.001
FACES III adaptability	38.42 ± 7.89	39.25 ± 7.96	-1.021	0.307

SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; K-10: Kessler Psychological Distress Scale; CTQ: Childhood Trauma Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

Comparison of NSSI frequency and methods between only-child and non-only-child adolescents

Sibling status differences were observed in NSSI frequency and methods, with only-children being more likely to engage in high-frequency NSSI (≥ 5 times in past year) compared with non-only-children (P < 0.001). Only-children were more likely to use cutting, scratching, and interfering with wound healing, whereas non-only-children had higher rates of hitting (all P < 0.05). No significant differences were found in burning, head-banging, or hair-pulling between the two groups (all P > 0.05). Detailed results are presented in Table 7.

Table 7 Comparison of non-suicidal self-injury frequency and methods between only-child and non-only-child adolescents, n (%).

Variables	Only-child (n = 296)	Non-only-child (n = 152)	χ2	P value
NSSI frequency (past year)			11.24	< 0.001
1-4 times	82 (27.70)	63 (41.45)
≥ 5 times	214 (72.30)	89 (58.55)
NSSI methods
Cutting	191 (64.53)	75 (49.34)	10.87	0.001
Scratching	176 (59.46)	69 (45.39)	8.652	0.003
Hitting	148 (49.93)	83 (54.61)	0.897	0.343
Burning	52 (17.57)	28 (18.42)	0.058	0.809
Head-banging	98 (33.11)	53 (34.87)	0.128	0.720
Hair-pulling	78 (26.35)	36 (23.68)	0.428	0.512
Interfering with wound healing	134 (45.27)	42 (27.63)	12.58	< 0.001

NSSI: Non-suicidal self-injury.

Multivariate logistic regression analysis of NSSI frequency (≥ 5 times vs 1-4 times) by sibling status

Binary logistic regression was conducted separately for only-children and non-only-children, with NSSI frequency as the dependent variable and factors showing statistical significance in the sibling-status–stratified univariate analysis (Table 6) as independent variables. For only-children, lower SCS self-warmth (OR = 0.897, 95%CI: 0.856-0.941), higher K-10 score (OR = 1.118, 95%CI: 1.057-1.183), and lower FACES III cohesion (OR = 0.912, 95%CI: 0.874-0.952) were independently associated with higher NSSI frequency (all P < 0.05). For non-only-children, only higher K-10 score (OR = 1.097, 95%CI: 1.028-1.170) was an independent factor (P < 0.05). Detailed results are presented in Table 8.

Table 8 Sibling status-specific multivariate logistic regression analysis of non-suicidal self-injury frequency.

Variables	Only-child (n = 296)				Non-only-child (n = 152)
	B	SE	OR (95%CI)	P value	B	SE	OR (95%CI)	P value
K-10 score	0.111	0.028	1.118 (1.057-1.183)	< 0.001	0.093	0.032	1.097 (1.028-1.170)	0.004
SCS self-warmth	-0.108	0.021	0.897 (0.856-0.941)	0.0000	-0.042	0.029	0.958 (0.903-1.017)	0.152
SCS self-coldness	0.031	0.020	1.032 (0.991-1.074)	0.132	0.028	0.026	1.029 (0.977-1.083)	0.276
ERQ cognitive reappraisal	-0.038	0.022	0.962 (0.920-1.006)	0.084	-0.021	0.028	0.978 (0.925-1.035)	0.456
FACES III cohesion	-0.091	0.025	0.912 (0.874-0.952)	< 0.001	-0.032	0.031	0.967 (0.910-1.028)	0.298
Parental marital status (divorced/remarried vs married)	0.287	0.214	1.332 (0.894-1.987)	0.1658	0.312	0.328	1.367 (0.735-2.541)	0.318
Constant	-4.215	1.287	0.015 (0.002-0.121)	< 0.001	-2.874	1.568	0.055 (0.004-0.721)	0.028

OR: Odds ratio; CI: Confidence interval; K-10: Kessler Psychological Distress Scale; SCS: Self-Compassion Scale; ERQ: Emotion Regulation Questionnaire; FACES: Family Adaptation and Cohesion Evaluation Scales.

DISCUSSION

This cross-sectional study investigates factors influencing NSSI in 620 adolescents with mood disorders, focusing on gender and sibling status (only-child vs non-only-child). The results reveal a high NSSI prevalence (72.26%) in this clinical group, aligning with global findings of elevated self-harm rates among adolescents with depressive or bipolar disorders. Multivariate logistic regression identifies female gender and only-child status as independent sociodemographic predictors of NSSI, alongside clinical, psychological, and family-related factors.

Female adolescents with mood disorders were 1.87 times more likely to engage in NSSI than males, aligning with meta-analytic evidence that females exhibit higher NSSI prevalence in clinical samples. Beyond prevalence, gender-specific patterns emerged: Females had higher frequencies (68.32% with ≥ 5 episodes/year vs 53.97% in males) and favored subtle, repetitive methods (cutting, scratching, interfering with wound healing), while males predominated in overt, forceful behaviors (hitting, head-banging). These differences may be interpreted through gender role socialization theory, which posits that females are encouraged to internalize distress and attend to relational harmony, whereas males are socialized toward externalizing behaviors and action-oriented coping[22,23]. Emotion regulation theory further supports this, suggesting that females’ preference for subtle NSSI methods reflects reliance on cognitive and affective strategies (e.g., self-criticism, rumination) to manage overwhelming negative emotions[24,25]. Gender also moderated predictors of recurrent NSSI[26]. For males, psychological distress (K-10 score) and poor family cohesion were the sole independent factors, suggesting external stressors and familial support deficits drive persistent self-harm. In contrast, females were vulnerable to low self-warmth, high childhood trauma, and impaired cognitive reappraisal - aligning with frameworks of self-critical cognitive schemas, which propose that females are more likely to internalize trauma and develop maladaptive self-referential beliefs that undermine adaptive emotion regulation[27]. The gender-specific predictors highlight the need for tailored interventions: Males may benefit from family-focused therapies to enhance cohesion, while females may require trauma-informed care and training in cognitive reappraisal to mitigate self-critical tendencies.

Only-child adolescents with mood disorders had 1.54 times higher odds of NSSI compared to non-only-children, a finding salient in contexts with historical one-child policies. Within the NSSI group, only-children exhibited more severe self-harm: 72.30% reported ≥ 5 episodes/year (vs 58.55% in non-only-children) and preferred methods similar to females (cutting, scratching). Psychologically, only-children had lower self-warmth, higher self-coldness, and poorer cognitive reappraisal, indicating reduced adaptive self-relating and emotion regulation capacities. Family functioning also differed: Only-children were more likely to experience parental divorce/remarriage and lower family cohesion, consistent with family systems theory and social buffering models, which posit that sibling presence provides emotional support, observational learning, and opportunities for co-regulation, thereby reducing stress reactivity and internalizing symptoms[28,29]. Sibling status further moderated NSSI frequency predictors. For only-children, low self-warmth, high psychological distress, and poor family cohesion independently predicted recurrent NSSI - reflecting a confluence of intrapersonal and familial vulnerabilities. In contrast, non-only-children’s NSSI frequency was solely predicted by psychological distress, underscoring the protective role of siblings as buffers that may facilitate adaptive coping, provide social feedback, and moderate the impact of stress on self-injurious behavior[30,31]. The heightened vulnerability of only-children underscores the need for targeted social support interventions to compensate for the absence of sibling bonds[32].

Beyond gender and sibling status, several factors consistently predicted NSSI. Longer mood disorder duration (OR = 1.08) underscores that unremitted emotional dysregulation over time amplifies self-harm risk, as chronic dysphoria erodes adaptive coping resources. Comorbid obsessive-compulsive symptoms (OR = 2.13) reflect overlapping psychopathology - obsessive-compulsive disorder-related intrusive thoughts and impulsivity may exacerbate self-injurious urges, consistent with findings of shared neurobiological substrates (e.g., altered prefrontal-striatal connectivity) between mood disorders, obsessive-compulsive disorder, and NSSI[33].

Psychologically, low self-warmth (OR = 0.92) and high self-coldness (OR = 1.05) highlight the role of self-compassion deficits. Adolescents who lack self-kindness and exhibit harsh self-criticism may use NSSI to punish themselves or cope with feelings of inadequacy, a mechanism supported by neuroimaging studies linking self-criticism to reduced activity in the ventral medial prefrontal cortex (a region implicated in self-referential processing). Integrating emotion regulation theory, these findings suggest that deficits in self-compassion impede the cognitive reappraisal of negative events, particularly in females, thereby increasing reliance on NSSI as an immediate emotion regulation strategy. Poor cognitive reappraisal (OR = 0.90) further confirms ineffective emotion regulation as a core mechanism - adolescents unable to reframe negative events turn to NSSI for immediate distress relief, as self-harm induces transient analgesia via endorphin release[34].

Childhood trauma (OR = 1.04) and low family cohesion (OR = 0.94) emphasize contextual roots of NSSI. Trauma disrupts hypothalamic-pituitary-adrenal axis function and emotional reactivity, while weak family cohesion reduces access to social support - both factors amplifying self-harm risk. Notably, these variables interacted with gender and sibling status: Only-children’s vulnerability was exacerbated by poor family cohesion, while females’ risk was amplified by trauma exposure, highlighting the multilevel interplay of individual, familial, and sociocultural influences on NSSI, consistent with interpersonal support models[35].

This study has limitations. Cross-sectional design precludes causal inference; longitudinal studies are needed to clarify whether gender- or sibling-related vulnerabilities precede NSSI or result from it. Self-report measures may be prone to recall bias, particularly for childhood trauma and NSSI frequency. Additionally, the single-center sample limits generalizability to community-dwelling adolescents or those with milder mood symptoms. Strengths include a large sample size (n = 620) with high response rate (94.4%), comprehensive assessment of sociodemographic, clinical, psychological, and family factors, and gender- and sibling-specific analyses that enhance granularity. The focus on sibling status addresses a gap in literature, particularly relevant in contexts with high only-child populations. Clinical practice should integrate gender and sibling status into NSSI screening and intervention. For females, trauma-informed care, self-compassion training, and cognitive reappraisal skills building are critical. Males may benefit from family therapy to improve cohesion and stress management. Only-children require enhanced social support systems (e.g., peer mentorship, group therapy) to compensate for sibling absence, alongside family interventions to strengthen cohesion. Routine screening for NSSI in adolescents with mood disorders should include assessments of gender, sibling status, childhood trauma, self-compassion, and family functioning to identify high-risk groups early.

CONCLUSION

In conclusion, gender and sibling status are key determinants of NSSI in adolescents with mood disorders, with distinct mechanisms driving risk in each subgroup. These findings highlight the importance of personalized approaches to NSSI prevention and treatment, ultimately improving outcomes for this vulnerable population.