Adverse childhood experiences, depressive symptoms, and non-suicidal self-injury in Chinese adolescents: the moderating role of SKA2 gene rs7208505 polymorphism

Abstract

Background

The Hypothalamic-pituitary-adrenal axis (HPA) and its single nucleotide polymorphisms (SNPs) potentially influence depressive symptoms and non-suicidal self-injury (NSSI) among adolescents. Adverse childhood experiences (ACEs) may dysregulate HPA axis functioning, with these complex gene-environment interactions showing significant heterogeneity across individuals. This study examined whether depressive symptoms mediate the relationship between ACEs and adolescent NSSI, and whether HPA axis genetic polymorphisms moderate this indirect pathway.

Methods

For this study, 172 adolescents aged 12 to 18 years were recruited from Xuzhou Oriental Hospital affiliated with Xuzhou Medical University as the NSSI group. Additionally, 58 age-, sex-, residence-, and health-matched volunteers were recruited from the local area as the healthy control group. Demographic information was collected through questionnaires, and adverse childhood experiences, depressive symptoms, and non-suicidal self-injury behaviors were assessed using the Adverse Childhood Experiences Scale (ACEs Scale), the Beck Depression Inventory-II (BDI-II), and the Non-Suicidal Self-Injury Behavior Scale (NSSI-BS). Blood samples were collected for HPA axis genotyping targeting the following loci: SKA2 (rs7208505, rs9911583), SLC1A3 (rs2269272), FKBP5 (rs9470080), and AVPR1B (rs28373064), which were analyzed using TaqMan-PCR. All statistical analyses were performed in SPSS 27.0, including Hardy-Weinberg equilibrium (HWE), χ² test, and Pearson’s correlation coefficient, as well as PROCESS macro-model mediation analysis and simple slope analysis.

Results

Depressive symptoms were found to mediate the association between ACEs and NSSI. The SKA2-rs7208505 polymorphism significantly moderated the relationship between ACEs and depressive symptoms, influencing the indirect pathway from ACEs to depressive symptoms to NSSI. Adolescents with the AA genotype of the SKA2-rs7208505 polymorphism were more susceptible to ACEs and more likely to develop depressive symptoms compared to those with GA and GG genotypes.

Conclusion

Timely identification and interventions targeting depressive symptoms, especially among individuals with AA genotypes of SKA2-rs7208505 who have experienced ACEs, may be crucial for preventing or reducing the risk of NSSI in adolescents.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12993-026-00332-2.

Introduction

Non-suicidal self-injury (NSSI) encompasses deliberate acts of self-directed harm without suicidal intent, manifesting through various forms of tissue damage such as cutting, burning, or hitting oneself [1]. Adolescence represents a critical vulnerability period for the onset of NSSI, with peak emergence typically occurring between ages 13 and 17 [2]. Epidemiological studies have revealed that approximately 22% of children and adolescents report at least one lifetime instance of NSSI, while the rate rises to 39% among university students, contrasting with 5.5% in adults [3, 4]. Global meta-analyses indicate an adolescent NSSI prevalence of 19.5% [4], with specific regional variations, such as a 6-month prevalence rate of 16% observed among Chinese adolescents [5]. Although NSSI behaviors tend to diminish in late adolescence, they may become entrenched as maladaptive coping mechanisms for managing emotional distress. The implications of NSSI extend beyond immediate physical harm, as individuals who cease self-injury may subsequently develop severe substance abuse disorders [6]. Furthermore, NSSI has been identified as a significant risk factor for anxiety disorders and suicidal behaviors [7–9], underscoring the importance of understanding its underlying mechanisms to develop effective interventions.

The etiology of NSSI remains incompletely understood, although several potential mechanisms have been proposed, including inflammatory processes and stress responses [10]. A growing body of evidence suggests that adverse childhood experiences (ACEs) constitute a primary environmental factor influencing NSSI [11, 12]. ACEs encompass a broad spectrum of negative and stressful events experienced during childhood and adolescence [13], including domestic violence, abuse, neglect, family mental illness, substance abuse, bereavement, parental separation, poverty, and social exclusion [13]. These experiences can significantly impact psychological well-being and social functioning throughout the developmental trajectory of adolescents [14]. The prevalence of ACEs among young populations is notably high [15]. Contemporary research indicates that approximately 70% of university students report at least one type of ACEs [16, 17]. A survey conducted among Chinese middle school students documented that 89.4% of participants reported experiencing at least one ACEs, with 43.6% reporting exposure to 5–6 different types of adverse experiences [18]. Prolonged exposure to elevated stress levels can significantly increase the risk of problematic behaviors, with impacts manifesting from childhood through adulthood. The pragmatic hypothesis of the integrated theoretical model of NSSI [1] suggests that adolescents may utilize physical pain as a mechanism to divert or release persistent negative emotional stress stemming from adverse childhood experiences [19]. This theoretical perspective is supported by empirical evidence, as demonstrated in a comprehensive systematic review encompassing 20 cross-sectional studies, which established a positive correlation between childhood maltreatment and NSSI [20]. Further substantiating this relationship, an extensive multinational European investigation revealed that approximately 79% of individuals engaging in self-harm reported experiences of childhood abuse or bullying [21]. The study also indicated that victims of bullying demonstrate significantly elevated risks of developing depressive symptoms and engaging in NSSI in adulthood, with rates approximately 2–3 times higher than those without such experiences [19]. Collectively, these findings underscore the significant role of ACEs as a crucial predictive factor for adolescent NSSI [22, 23].

Drawing from the vulnerability-stress model [24], depressive symptoms may represent a mediating factor through which ACEs influence NSSI. Research indicates that individuals exposed to early adversities such as abuse and household dysfunction are more susceptible to developing depressive symptoms [25], as these experiences disrupt healthy emotional development processes [26]. Yu and Zhang’s [27] 14-year longitudinal study of 5734 adolescents demonstrated that ACEs were significantly associated with sustained elevation in depressive symptoms from adolescence into young adulthood. Furthermore, adolescents with higher levels of depression are more likely to employ NSSI as a maladaptive strategy for managing negative emotions [28], with multiple studies confirming depressive symptoms as a significant predictor of NSSI among adolescents [29], mediating the relationship between ACEs and NSSI.

Although ACEs are a key predictor of depressive symptoms and NSSI in adolescents, their impact varies across individuals. NSSI is determined by both genetic and environmental factors [30]. The hypothalamic-pituitary-adrenal (HPA) axis, as a key component of the stress system [31], regulates stress responses through the release of corticotrophin releasing hormone (CRH) and adrenocorticotrophin hormone (ACTH) [32]. Extensive research has indicated that HPA axis dysregulation, particularly manifesting as attenuated responsiveness in adolescents with NSSI [33], represents a fundamental biological mechanism underlying the development and progression of depressive symptoms [34]. According to the differential susceptibility model [35], different genotypes of HPA axis-related genes (such as AVPR1B, SLC1A3, FKBP5, and SKA2) [36–39] may lead to significant variations in individuals’ sensitivity to environmental stimulus. This gene-environment interaction helps explain why some individuals develop NSSI when facing similar environmental stressors (such as adverse childhood experiences), while others demonstrate greater resilience [40]. Based on the theoretical framework discussed, we posit that genetic polymorphisms in the HPA axis will function as a moderator in the indirect pathway connecting ACEs to NSSI through depressive symptoms among adolescents. The conceptual model is shown in Fig. 1.

Fig. 1.

The conceptual model of the proposed moderated mediation framework

Method

Participants

The NSSI group consisted of Han Chinese adolescents aged 12–18 years who sought treatment for non-suicidal self-injury (NSSI) at the Affiliated Xuzhou Oriental Hospital of Xuzhou Medical University, a tertiary Grade-A psychiatric hospital in Xuzhou, China. Eligibility for inclusion required confirmation by at least two attending psychiatrists that participants had engaged in at least one NSSI episode within the past year. Ultimately, 172 adolescents (28 males and 86 females) who provided consent for blood sample collection were enrolled. However, participants were excluded from the NSSI group if they met any of the following criteria: (1) diagnosis other significant physical or psychiatric conditions, including cardiovascular disease, kidney disease, cerebral ischemia, disruptive mood dysregulation disorder, or persistent mood disorder; (2) substance abuse or dependence; (3) a family history of psychiatric disorders; or (4) inability to independently or assistedly complete study-related questionnaires. At the same time, the healthy control (HC) group comprised 58 age-, sex-, and residence-matched volunteers (24 males and 34 females) recruited from local schools and communities during the same period. HC participants were rigorously screened using the Beck Scale for Suicide Ideation to exclude those with any history of suicidal ideation and self-injurious behaviors. Furthermore, potential control subjects were deemed ineligible if they exhibited evidence of substance abuse or dependence or reported any family history of mental health concerns.

This study was reviewed and approved by the ethics committee of the Xuzhou Center for Disease Control and prevention and Xuzhou Medical University. The procedures used in this study adhere to the tenets of the Declaration of Helsinki. All participants gave their written informed consent before their inclusion in the study. All participants signed a written informed consent form after a fullwritten and verbal explanation of the study.

Measures

Demographic characteristics

A self-designed questionnaire was administered to gather general demographic information from both the NSSI and healthy control groups. The data collected included age (in years), gender, household residence, only-child status, and parental education levels (father’s and mother’s education).

Adverse childhood experiences

The Adverse Childhood Experiences (ACEs) Scale [41] was administered to evaluate participants’ early-life adversities. To distinguish from expanded ACE inventories that include community-level stressors, we utilized the standard 10-item version derived from the original CDC-Kaiser study, which focuses specifically on intrafamilial adversities. This instrument comprises ten dichotomous items divided into two domains: childhood maltreatment (items 1–5, physical, emotional, and sexual abuse, as well as physical and emotional neglect) and household dysfunction (items 6–10, examining parental separation/divorce, domestic violence, household substance abuse, household mental illness, and having an incarcerated household member). Each item was scored binary (yes = 1, no = 0), with affirmative responses indicating the presence of specific adverse experiences. The cumulative score, calculated by summing responses across all items, ranges from 0 to 10, with higher scores reflecting greater exposure to childhood adversities. Internal consistency reliability analysis yielded a Cronbach’s α coefficient of 0.691 for the present sample.

Depressive symptoms

The 21-item Beck Depression Inventory-Second Edition (BDI-II) [42] was employed to evaluate the severity of depressive symptoms in participants. This widely validated self-report instrument assesses cognitive, affective, motivational, and physiological dimensions of depression over the past two weeks. Each item consists of four graduated statements scored on a 4-point Likert scale (0 to 3), reflecting increasing symptom severity. The composite score is calculated by summing the responses to all items, resulting in a total score ranging from 0 to 63, with higher scores indicating greater symptom severity. In the present study, the BDI-II demonstrated excellent internal consistency, with a Cronbach’s α coefficient of 0.969.

Non-suicidal self-injury

The Non-Suicidal Self-Injury (NSSI) Behavior Scale [43] was employed to assess the methods and severity of deliberate self-harm behaviors among adolescents without suicidal intent. This comprehensive tool evaluates 18 distinct forms of intentional self-injury. The frequency of each behavior is categorized into four levels: none (0 occurrences), single episode (1 occurrence), occasional (2–4 occurrences), and frequent (≥ 5 occurrences). Similarly, the severity of physical harm is rated across five gradients: none, mild, moderate, severe, and extreme. To calculate the NSSI score, the frequency and severity ratings for each self-injurious behavior are multiplied, and the resulting products are summed across all behaviors to yield a total score. This cumulative score serves as a measure of the presence and intensity of NSSI behaviors. The scale demonstrated robust psychometric properties with a Cronbach’s α coefficient of 0.934.

Experimental procedures

Peripheral venous blood samples (5 mL) were collected from both the NSSI and healthy control groups using EDTA-K2 anticoagulant tubes and stored at − 80 °C until analysis. Genomic DNA was extracted using the TIANamp Blood DNA Kit, following the manufacturer’s protocol. The DNA concentration and purity were assessed via UV spectrophotometry to ensure sample quality.

Key single nucleotide polymorphisms (SNPs) in hypothalamic-pituitary-adrenal (HPA) axis-related genes were selected based on prior research and data from the NCBI database (http://www.ncbi.nlm.nih.gov/gene). Primer design, synthesis, and validation were performed by Wuhan Qingke Biotechnology. Genotyping was conducted using the TaqMan-PCR allelic discrimination assay, which employs specific primers and fluorescent probes.

Statistical analysis

All statistical analyses were conducted using IBM SPSS Statistics version 27.0. preliminary analyses encompassed descriptive statistics for demographic characteristics. The distribution of genotype frequencies in both NSSI and healthy control groups was evaluated for Hardy-Weinberg equilibrium (HWE). Between-group comparisons of allelic and genotypic frequencies were performed using chi-square (χ²) tests. Bivariate associations among principal variables were examined through pearson correlation coefficients, with means and standard deviations computed for each variable. To elucidate potential mechanisms underlying the relationships among variables, mediation analyses were implemented using PROCESS macro Model 4 [44]. Subsequently, genetic polymorphisms demonstrating significant distributional differences between NSSI and healthy control groups were incorporated as moderating variables in a moderated mediation framework (PROCESS Model 59). Both mediation and moderated mediation analyses utilized 5,000 bootstrap samples to obtain robust standard errors and bootstrap confidence intervals for parameter estimates. A result was considered significant if the confidence interval did not include zero. To better interpret the moderating effects, simple slope analysis was performed to investigate how genetic polymorphisms influence the relationship between ACEs and depressive symptoms.

Result

Demographic information

A total of 172 adolescents participated in this study, including 114 in the NSSI group and 58 in the healthy control group. The NSSI group was younger (14.4 ± 1.44 years), had a higher proportion of females (75.4%), and showed significant differences in permanent residential status, only-child status, and parental education levels compared to the healthy control group. Given these significant demographic disparities, these variables were subsequently incorporated as covariates in further analyses. Specific details are shown in Table 1.

Table 1.

Comparisons of demographic characteristics between the adolescent NSSI group and the healthy control group (n = 172)

Characteristic	NSSI (n = 114)	HC (n = 58)	Z/χ 2	p
Age in years	14.4 (1.437)	15.4 (2.160)	− 2.971	0.003
Gender			5.155	0.024
Female	86 (75.4%)	34 (58.6%)
Male	28 (24.6%)	24 (41.4%)
Permanent residence			5.719	0.017
Urban	61 (53.5%)	42 (72.4%)
Rural	53 (46.5%)	16 (27.6%)
Only child			7.804	0.005
Yes	18 (15.8%)	20 (34.5%)
No	96 (84.2%)	38 (65.5%)
Paternal education			9.314	0.009
Junior high school and below	73 (64.0%)	24 (41.4%)
High school, technical secondary school and vocational school	29 (25.4%)	20 (34.5%)
College or undergraduate	12 (10.6%)	14 (24.1%)
Maternal education			16.657	< 0.001
Junior high school and below	80 (70.2%)	25 (43.1%)
High school, technical secondary school and vocational school	22 (19.3%)	13 (22.4%)
College or undergraduate	12 (10.5%)	20 (34.5%)

Note: Statistically significant values (p< 0.05) are reported in bold

Hardy–Weinberg equilibrium test

Hardy-Weinberg equilibrium (HWE) analyses demonstrated that three polymorphisms (SLC1A3-rs2269272, FKBP5-rs9470080, and SKA2-rs9911583) conformed to HWE in both NSSI and the healthy control group (all p > 0.05). However, AVPR1B-rs28373064 (χ² = 27.740, p < 0.05) and SKA2-rs7208505 (χ² = 20.584, p < 0.05) showed deviations from HWE in the NSSI group, while maintaining equilibrium in the healthy control group (χ² = 4.401, p = 0.111 and χ² = 4.390, p = 0.111). The consistent HWE distribution of all investigated polymorphisms in the healthy control group indicated robust genotyping quality and sample representativeness. Details are provided in Table 2.

Table 2.

Hardy-Weinberg equilibrium test at each point in the NSSI group and the healthy control group

Gene	Locus rs number	NSSI		HC
		χ2	p	χ2	p
SLC1A3	rs2269272	0.895	0.639	0.067	0.967
AVPR1B	rs28373064	27.740	< 0.05	4.401	0.111
FKBP5	rs9470080	0.190	0.909	0.932	0.632
SKA2	rs7208505	20.584	< 0.05	4.390	0.111
	rs9911583	0.834	0.659	2.427	0.296

Genotypic and allelic frequencies

In this study, we investigated the association between several polymorphisms and non-suicidal self-injury (NSSI) behavior in adolescents. The genotypic and allelic frequencies of specific polymorphisms were analyzed to assess their potential role in NSSI risk. For the SKA2-rs7208505 polymorphism (Table 3), significant differences were observed in allele frequencies between NSSI and the healthy control group (χ² = 4.882, p = 0.027). The codominance model (χ² = 8.279, p = 0.016) and recessive model (χ² = 7.804, p = 0.005) demonstrated statistically significant differences in genotype distribution. Binary logistic regression analysis indicated significant associations in the allelic model (G vs. A: OR = 1.685, 95%CI 1.059–2.683, p = 0.028), codominance model (GG vs. AA: OR = 4.190, 95%CI 1.034–16.982, p = 0.045), and recessive model (GA + GG vs. AA: OR = 2.706, 95%CI 1.329–5.510, p = 0.006), suggesting that the G allele might confer increased risk for NSSI behavior.

Table 3.

Association analyses between SKA2 gene polymorphisms (rs7208505 and rs9911583) and non-suicidal self-injury (NSSI) group

SKA2-rs7208505	NSSI	HC	χ2	p	OR (95%CI)	p
Allele
A	123	77	4.882	0.027	1
G	105	39			1.685 (1.059–2.683)	0.028
Codominance model
AA	21	22	8.279	0.016	1
GA	81	33			1.630 (0.432–6.151)	0.471
GG	12	3			4.190 (1.034–16.982)	0.045
Dominant model
AA + GA	102	55	1.384	0.239	1
GG	12	3			2.157 (0.584–7.969)	0.249
Recessive model
AA	21	22	7.804	0.005	1
GA + GG	93	36			2.706 (1.329–5.510)	0.006
Overdominant model
GA	81	33	3.447	0.063	1
AA + GG	33	25			0.538 (0.278–1.039)	0.065
SKA2-rs9911583
Allele
C	146	59	5.541	0.019	1
T	82	57			0.581 (0.369–0.915)	0.019
Codominance model
CC	49	18	5.496	0.064	1
CT	48	23			0.367 (0.155–0.870)	0.023
TT	17	17			0.479 (0.208–1.106)	0.085
Dominant model
CC	49	18	2.308	0.129	1
TT + CT	65	40			0.597 (0.306–1.165)	0.130
Recessive model
CT + CC	97	41	5.025	0.025	1
TT	17	17			0.423 (0.197–0.908)	0.027
Overdominant model
CT	48	23	0.095	0.758	1
CC + TT	66	35			0.904 (0.474–1.721)	0.758

Note: Statistically significant values (p< 0.05) are reported in bold

The SKA2-rs9911583 polymorphism (Table 3) demonstrated significant differences in allele frequencies (χ² = 5.541, p = 0.019) and in the recessive model (χ² = 5.025, p = 0.025). Binary logistic regression analysis revealed significant associations in the allelic model (T vs. C: OR = 0.581, 95%CI 0.369–0.915, p = 0.019), codominance model (CT vs. CC: OR = 0.367, 95%CI 0.155–0.870, p = 0.023), and recessive model (TT vs. CT + CC: OR = 0.423, 95%CI 0.197–0.908, p = 0.027).

For SLC1A3-rs2269272, FKBP5-rs9470080, and AVPR1B-rs28373064 polymorphisms (Supplemental Tables 1–3), no statistically significant differences were observed in genotype distribution between NSSI and the healthy control groups across all genetic models (p > 0.05). Binary logistic regression analyses revealed no significant associations between these polymorphisms and adolescent NSSI behavior in any of the tested inheritance models.

Bivariate correlations

Table 4 presents bivariate correlations between study variables. Results showed that NSSI was positively correlated with ACEs (r = 0.509, p < 0.01) and depressive symptoms (r = 0.748, p < 0.01). For genetic factors, NSSI demonstrated a significant positive correlation with SKA2-rs7208505 (r = 0.154, p < 0.05). Similarly, depressive symptoms showed strong positive associations with ACEs (r = 0.580, p < 0.01) and SKA2-rs7208505 (r = 0.228, p < 0.01). The two SKA2 genetic variants (rs7208505 and rs9911583) were also significantly correlated with each other (r = 0.351, p < 0.01).

Table 4.

Descriptive statistics and correlations between study variables (N = 172)

	1	2	3	4	5	6	7	8	9	10	11
1.rs7208505	1
2.rs9911583	0.351**	1
3.Gender	− 0.088	0.059	1
4.Age in years	− 0.009	− 0.098	− 0.226**	1
5.Only child	− 0.049	− 0.063	0.168*	− 0.118	1
6.Permanent residence	0.115	− 0.048	− 0.120	− 0.158*	0.282**	1
7.Maternal education	− 0.161*	− 0.005	− 0.007	0.018	− 0.391**	− 0.363**	1
8.Paternal education	− 0.174*	− 0.019	− 0.031	0.064	− 0.356**	− 0.382**	0.669**	1
9.ACEs	0.101	− 0.089	0.107	− 0.166*	0.099	0.140	− 0.181*	− 0.160*	1
10.Depressive symptoms	0.228**	0.098	0.197**	− 0.245**	0.234**	0.264**	− 0.348**	− 0.256**	0.580**	1
11.NSSI	0.154*	0.126	0.206**	− 0.275**	0.180*	0.112	− 0.198**	− 0.132	0.509**	0.748**	1

rs7208505 were dummy coded such that 1 = GG = GA, 0 = AA. *p < 0.05, **p < 0.01

Mediating effect of depressive symptoms

The mediating effect of depressive symptoms was examined using the PROCESS macro for SPSS (model 4, with 5000 bootstrapped samples and bias-corrected 95% confidence intervals). After controlling for demographic variables (gender, age, only child, permanent residence, and parental education), the analysis revealed a significant mediation pathway(Supplemental Table 4). Specifically, adverse childhood experiences (ACEs) significantly predicted depressive symptoms (β = 5.062, SE = 0.612, t = 8.267, p < 0.001, 95% CI [3.853, 6.272]), and depressive symptoms subsequently predicted NSSI (β = 1.312, SE = 0.130, t = 10.109, p < 0.001, 95% CI [1.056, 1.568]). Notably, the direct effect of ACEs on NSSI was not significant (β = 2.119, SE = 1.209, t = 1.753, p = 0.082, 95% CI [− 0.269, 4.506]). Furthermore, the bootstrapped indirect effect of ACEs on NSSI via depressive symptoms was significant (Effect = 6.643, Boot SE = 1.055, 95% Boot CI [4.800, 8.893]). These findings indicate that depressive symptoms serve as a full mediating mechanism through which ACEs influence NSSI among adolescents.

Testing for the moderated mediation model

The moderated mediation analysis explored the potential influence of two SKA2 gene polymorphisms (rs7208505 and rs9911583) on the indirect relationship between adverse childhood experiences (ACEs) and non-suicidal self-injury (NSSI) through depressive symptoms .

Significantly, SKA2-rs7208505 demonstrated a significant moderating effect on the indirect pathway from ACE to depressive symptoms and subsequently to NSSI. The interaction between ACE and SKA2-rs7208505 on depressive symptoms was statistically significant (β = − 3.124, p = 0.039), suggesting a buffering effect of this polymorphism for individuals with higher ACEs exposure. However, the interaction between depressive symptoms and SKA2-rs7208505 on NSSI was not significant (β = − 0.098, p = 0.766), as illustrated in Fig. 2A and Supplemental Table 5. Further analysis of the conditional indirect effects showed that the indirect pathway was significant for both the AA genotype (Effect = 10.452, 95% Boot CI [5.178, 17.317]) and the GA/GG genotypes (Effect = 5.676, 95% Boot CI [3.853, 8.038]). The overall index of moderated mediation was − 4.776 (95% Boot CI [− 11.897, 0.965]).

Fig. 2.

The final structural model with standardized coefficients(N = 172). A SKA2-rs7208505 model with standardized path coefficients. B SKA2-rs9911583 model with standardized path coefficients. *p < 0.05, **p < 0.01, ***p < 0.001

To elucidate the nature of this interaction, simple slope analysis was performed (Fig. 3 and Supplemental Table 5). The results indicated that the positive association between ACEs and depressive symptoms was significantly stronger in adolescents with the AA genotype (slope = 7.512, t = 5.487, p < 0.001) compared to those with the GA and GG genotypes (slope = 4.388, t = 6.722, p < 0.001). Although the GA/GG genotypes were identified as risk factors for the presence of NSSI (main effect), the interaction analysis suggests that AA carriers exhibit enhanced environmental sensitivity, showing a more pronounced increase in depressive symptoms in response to adverse childhood experiences.

Fig. 3.

Simple slope plot of SKA2-rs7208505 polymorphism on the relationship between ACEs and depressive symptom

In contrast, for the moderated mediation model involving SKA2-rs9911583, the interaction between ACEs and SKA2-rs9911583 on depressive symptoms was not significant (β = − 1.256, p = 0.398), while the interaction between depressive symptoms and SKA2-rs9911583 on NSSI was marginally significant (β = − 0.625, p = 0.051). Importantly, the overall index of moderated mediation was significant (Index = − 4.201, 95% Boot CI [− 7.583, − 1.152]). Specifically, the conditional indirect effect of ACEs on NSSI via depressive symptoms was significantly stronger for individuals with the TT/CT genotype (Effect = 7.146, 95% Boot CI [5.013, 9.893]) compared to those with the CC genotype (Effect = 2.968, 95% Boot CI [1.111, 5.212]) (Fig. 2B and Supplemental Table 6). These findings indicate that both SKA2-rs7208505 and SKA2-rs9911583 play significant moderating roles in the ACEs-depressive symptoms-NSSI pathway.

Discussion

This study investigated the relationship between Adverse Childhood Experiences (ACEs) and non-suicidal self-injury (NSSI) among adolescents, examining the mediating role of depressive symptoms and the moderating effects of HPA axis-related genetic polymorphisms. Through vulnerability-stress model [45] and differential susceptibility model [46], our findings not only validated the mediating effect of depressive symptoms between ACEs and NSSI but also revealed whether this indirect link was moderated by the HPA axis genetic polymorphisms, particularly under environmental influences.

The results demonstrated that depressive symptoms significantly mediated the relationship between ACEs and adolescent NSSI. Drawing from ecological systems theory [47], childhood represents a critical period for cognitive, emotional, and neurophysiological development, during which adverse experiences can impact psychological well-being through multiple pathways [48, 49]. Our findings align with recent comprehensive systematic reviews and meta-analyses which have robustly established that ACEs serve as profound, transdiagnostic risk factors for suicidal and self-injurious behaviors across different psychiatric conditions. Specifically, a recent meta-analysis by Baldini et al. [50] found that exposure to ACEs nearly doubles the risk of suicidal behaviors in individuals with affective disorders, with sexual abuse showing the highest risk association. Similarly, in schizophrenia spectrum disorders, Baldini et al. [51] confirmed a “highly suggestive” association between ACEs and suicidal outcomes. These studies highlight that early-life trauma exerts a cumulative deleterious impact on mental well-being [50, 51]. By integrating these epidemiological insights with our mediation model, we clarify the specific pathway: ACEs compromise self-esteem and self-efficacy [27], increasing vulnerability to depressive symptoms [25], which in turn elevates NSSI risk through impairments in emotional cognition and social functioning [52].

A crucial finding emerged regarding the genetic moderation of this mediational pathway. The “ACEs → depressive symptoms → NSSI” mediation effect was significantly moderated by SKA2 gene polymorphisms. Specifically, carriers of the SKA2-rs7208505 AA genotype exhibited significantly higher levels of depressive symptoms when experiencing ACEs compared to GA and GG carriers. Additionally, our updated analysis confirmed that the rs9911583 locus also significantly moderated this pathway, with TT/CT carriers exhibiting a stronger conditional indirect effect than CC carriers. The SKA2 gene interacts with the glucocorticoid receptor and influences the HPA axis through DNA methylation patterns, ultimately affecting the body’s production and regulation of cortisol [53, 54]. Our findings are consistent with recent research by Luz I. Valenzuela-García et al. [55], which suggested that the AA genotype of rs7208505 might elevate susceptibility to depressive symptoms and NSSI risk through its effects on CpG site methylation and the interaction between SKA2 protein and glucocorticoid receptors. While previous studies have linked ACEs to altered neuroplasticity in a general sense, our findings pinpoint a specific mechanism involving the HPA axis stress response. It is plausible that the AA genotype is associated with impaired negative feedback of the stress system, rendering carriers biologically less capable of buffering the psychological impact of ACEs compared to those with GA or GG genotypes.

From the perspective of gene-environment interactions [56], our results support the differential susceptibility model. Individuals with the AA genotype of SKA2-rs7208505 exhibit heightened sensitivity (“plasticity”) to environmental stressors. These individuals are more likely to develop depressive symptoms under stressful conditions (e.g., high levels of ACEs) [35]. In contrast, individuals with the GA and GG genotypes show a relatively buffered response, appearing more resilient to the depressogenic effects of trauma. This distinction is vital: while our main effect analysis (Table 3) indicated that the G-allele was associated with a higher baseline presence of NSSI, the interaction analysis revealed that the A-allele creates a specific vulnerability to early-life trauma. This precise gene-environment interaction thereby exacerbates the developmental pathway from childhood adversity to depressive symptoms and subsequent NSSI. Additionally, our study did not find significant associations between other HPA axis-related gene variants (SLC1A3 rs2269272, FKBP5 rs9470080, and AVPR1B rs28373064) and NSSI in adolescents, suggesting that different genetic loci may play distinct roles in the development of NSSI.

Several limitations warrant consideration. First, the relatively small sample size constrains our ability to fully capture the complexity of NSSI. Specifically, the low frequency of the homozygous GG genotype necessitated combining GA and GG groups, which may obscure specific dose-dependent effects of the G allele. Consequently, the risk estimates for these specific genotypes should be interpreted with caution and necessitate replication in larger-scale studies. Additionally, the selected genetic loci may not comprehensively represent all relevant genetic factors influencing NSSI behaviors. Furthermore, the cross-sectional design precludes causal interpretations. Future research should incorporate a broader range of genetic loci, conduct longitudinal studies, and integrate experimental research and gene expression analyses to develop a more comprehensive understanding of how HPA axis genes and gene-environment interactions influence NSSI behaviors.

Despite these limitations, our findings have valuable practical implications. The research highlights ACEs as crucial risk factors for adolescent NSSI, emphasizing the importance of trauma-informed care and routine screening for ACEs in clinical settings. The mediating role of depressive symptoms suggests that emotion regulation training could effectively reduce NSSI risk. Moreover, the identification of genetic differences in ACEs sensitivity, particularly regarding SKA2-rs7208505 AA genotypes, enables clinicians to better identify high-risk individuals who are biologically more sensitive to trauma and implement targeted interventions.

Conclusion

The comprehensive analysis in this study found that Adverse Childhood Experiences (ACEs) are profound risk factors for Non-Suicidal Self-Injury (NSSI) behaviors in adolescents, and depressive symptoms play a pivotal mediating role between ACEs and NSSI. Among the five representative loci of key genes of the Hypothalamic-Pituitary-Adrenal (HPA) axis selected in this study, the rs7208505 polymorphism of the SKA2 gene plays a significant regulatory role in the pathway between ACEs and depressive symptoms. Crucially, our interaction analysis revealed that individuals with the SKA2-rs7208505 AA genotype exhibit significantly higher sensitivity to environmental stressors and are more likely to develop depressive symptoms in the context of ACEs compared to those with GA or GG genotypes. This heightened reactivity in AA carriers subsequently increases their risk of engaging in NSSI behaviors. Our findings also hold potential clinical significance. Identifying adolescents with the SKA2-rs7208505 AA genotype could help in the early screening of high-risk individuals who are biologically more vulnerable to trauma, enabling targeted psychological interventions and stress management strategies to prevent the onset of depressive symptoms and NSSI. Furthermore, the results provide a theoretical foundation for personalized treatment approaches, highlighting the critical role of specific gene-environment interactions in adolescent mental health interventions.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ACEs

Adverse childhood experiences

ACTH

Adrenocorticotrophin hormone

AVPR1B

Arginine vasopressin receptor-1B

BDI-II

Beck Depression Inventory-II

CRH

Corticotrophin releasing hormone

DNA

Deoxyribonucleic acid

FKBP5

FK506 binding protein 5

HC

Healthy Control

HPA

Hypothalamus-pituitary-adrenal axis

HWE

Hardy–Weinberg equilibrium

NSSI

Non-suicidal self-injury

PCR

Polymerase chain reaction

SKA2

Spindle and kinetochore associated complex subunit 2

SLC1A3

Solute carrier family 1, member 3；glial high-affinity glutamate transporter

SNPs

Single nucleotide polymorphisms

Author contributions

Y. W. was responsible for the design and writing of the paper, Q.L. and W.Z. were involved in the collection and organization of data, Y.W. , S.T. and X.H. were responsible for the revision of the paper. J.T. and Y.L. have contributed to data analysis. H.F. and W.W. are the corresponding authors who conceived and designed the study, structured the manuscript, and conducted revisions. They have also read and approved the final manuscript and agreed to be accountable for all aspects of the work.

Funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by Scientific Research Project of Jiangsu Provincial Health Commission (MQ2024002), Science and Technology Special Project of Kunshan City (KS2373) and Xuzhou Municipal Peak Discipline Team (2025DF10).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All participants and their guardians signed an informed consent form before participating in the study. The study was approved by the ethics committee of Xuzhou Medical University and its affiliated hospital (XYFY2020-KL219-01), and the procedures were conducted following the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.