Non-Suicidal Self-Injury (NSSI) refers to the behavior of directly and deliberately harming bodily tissue (such as cutting or burning) without suicidal intent, which is a significant public health problem (1). A meta-analysis reported that the lifetime prevalence and 12-month prevalence of NSSI among children and adolescents were 22% and 23.2% globally between 2010 and 2021 (2). Deng et al. reported that the NSSI prevalence among adolescents and adults during the COVID-19 pandemic was 32.40% and 15.70% (3). Qu et al. estimated the lifetime prevalence of NSSI among the Chinese youth population to be alarmingly high at 24.7% (4). A number of demographic, personal, and social factors are found to be significantly correlated with NSSI behavior (5). Although NSSI behavior frequently co-exists with conditions such as depression (6), eating disorders, and borderline personality disorder (7), it can also manifest independently (8), such as being linked to emotional instability and compromised social functioning (9); it is also predictive of suicidal tendencies (10). Recognizing its high comorbidity and cross-diagnostic features, NSSI has been proposed as a standalone diagnosis in Section III of the DSM-5 (11). Research has indicated that NSSI behavior is most prevalent among adolescents (7) and is often associated with adverse outcomes such as dropping out of school and difficulties in interpersonal relationships (12), causing immense pain for both the adolescent and their family. Six Specific Psychotherapeutic Interventions (SPI) were found to specifically and significantly reduce NSSI in adolescents: Developmental Group Psychotherapy (DGP), Therapeutic Assessment (TA), the Cutting Down Program (CDP), Emotional Regulation Individual Therapy for Adolescents (ERITA), Treatment for Self-Injurious Behaviors (T-SIB) and Intensive Contextual Treatment (ICT) (13).

In recent years, clinicians and researchers have increasingly emphasized the need for a more multidimensional approach to psychiatric diagnosis and psychopathology research, highlighting the significant heterogeneity within traditional diagnostic (14). Consequently, the search for effective and objective biomarkers has become imperative. However, research exploring the underlying neurobiological systems of NSSI among adolescents has been limited so far. In the past, some studies reported the metabolic alterations in adolescents with major depressive disorder (MDD) exhibiting non-suicidal self-injury (NSSI) using magnetic resonance spectroscopy (MRS), as Zhang suggested that abnormal metabolic changes of N-acetylaspartate (NAA) and choline-containing compounds (Cho) were present in the thalamus in adolescents with depressive disorder and self-injurious behavior (15). Yan et al.’s research identified that elevated Cho and creatine (Cr) levels in the right thalamus might be correlated with increased NSSI risk in male adolescents during major depressive episodes (MDE) (16). However, few studies just focused on NSSI without a depressive episode. Previous functional imaging studies reported specific brain regions associated with Non-Suicidal Self-Injury (NSSI). Some evidence suggests that individuals engaging in self-harm exhibit enhanced activation in the prefrontal regions (including the orbitofrontal cortex and other areas within the prefrontal cortex) during emotional, social, and reward processing (17–19). Findings from previous research have suggested that adolescents engaging in non-suicidal self-injury (NSSI) have significantly reduced gray matter volume in the anterior cingulate cortex (ACC) and multiple prefrontal-limbic brain regions compared to healthy adolescents (20, 21). Additionally, during their resting state, adolescents with NSSI behavior display enhanced connectivity between the amygdala and ACC compared to healthy controls (HC) (22). A study utilizing multimodal magnetic resonance imaging (MRI) revealed significant differences in the temporal gyrus (TG) between Major Depressive Disorder (MDD) patients with a history of NSSI and those without (23). A study using functional near-infrared spectroscopy (fNIRS) also indicated weaker functional connectivity between the prefrontal and temporal lobes in MDD patients with Non-Suicidal Self-Injury (NSSI), as well as abnormal functional connectivity in the frontal lobe associated with NSSI (24). However as interest in multidimensional and cross-diagnostic research continues to grow, so does the need for studies exploring the potential pathophysiological mechanisms underlying NSSI behaviors.

There is a growing need for simple, non-invasive, and reliable scientific techniques to investigate the biological mechanisms of adolescents with NSSI behavior. This would aid in early and accurate diagnosis of the disorder and facilitate the implementation of effective treatment approaches. Functional near infrared spectroscopy (fNIRS) is a non-invasive brain functional imaging technology that indirectly reflects neural activity by monitoring changes in blood flow within the cerebral cortex (25). This technique offers a balance between temporal and spatial resolution, coupled with robust anti-noise capabilities, low cost, high tolerance, and ease of operation (26–28). As such, it is highly suitable as a tool for investigating neural activities in patients with mental illnesses. Further, the Verbal Fluency Task (VFT) is a classic cognitive test commonly used to assess executive functions (29), including inhibitory control, working memory, and cognitive flexibility. It is widely used in fNIRS research to induce activation in the Prefrontal Cortex (PFC) (30). Previous studies have shown that individuals with NSSI have difficulty in inhibitory control and cognitive flexibility (31). Mürner-Lavanchy et al.’s studies have found that during response inhibition tasks, adolescents with NSSI have slightly higher PFC oxygenation than the control group (32).

The primary objective of this study is to investigate the activation of the prefrontal and temporal cortex in adolescents with NSSI and interpret its relationship with NSSI behavior.

Our study aimed to investigated the differences of cortical activation between adolescents with NSSI and a healthy control group during VFT using fNIRS. Furthermore, we examined whether the task-based fNIRS activity correlated with NSSI behavior. The results of the study revealed that the degree of cortical activation in the dorsolateral prefrontal cortex (DLPFC) was significantly higher in the NSSI group compared to the healthy control group, and the level of DLPFC activation was strongly positively correlated with the social communication and expression subscales. This study adds to the evidence regarding the candidate neural correlates of NSSI in adolescents.

Results indicate distinct personality patterns in adolescents with NSSI based on the Eysenck Personality Questionnaire (EPQ). These adolescents exhibit obvious emotional instability, frequently experiencing intense emotions and overreacting to stimuli. This makes it difficult for them to restore emotional balance after arousal. Additionally, compared with healthy controls, the NSSI group has slightly higher scores on the psychoticism subscale. This indicates slightly higher levels of aggressive tendencies, oppositional feelings, and reduced adaptability. Furthermore, these adolescents tend to be introverted, preferring solitude over social interaction, which may exacerbate feelings of loneliness and emotional distress. These findings are consistent with recent studies. Kang et al. (2021) identified comparable patterns of emotional instability and introversion in college students with NSSI, emphasizing that emotional instability is central to maintaining self-injurious tendencies (39). Building on these findings, Peng et al. (2022) observed similar traits in adults with depressive disorders and NSSI (40). The study noted that patients with MDD and NSSI presented higher levels of EPQ-psychoticism and more immature defense styles compared to healthy controls (40). This reduced defensive capacity may be attributed to developmental factors, such as age-related differences in emotional regulation and varying degrees of socialization, which could impair the NSSI group’s ability to cope with interpersonal stress effectively (41). Collectively, these findings highlight the importance of integrating emotional regulation training and social skills interventions into therapeutic approaches for people with NSSI, particularly to address their heightened emotional instability and introverted tendencies.

Further, NSSI is often triggered by emotional outbursts (42), and individuals with NSSI consistently report poorer emotion regulation abilities compared to their non-self-injuring peers (43). Studies have demonstrated a mutual influence between NSSI and emotion regulation (44). Individuals with NSSI tend to use cognitive reappraisal strategies less frequently and use more emotional suppression strategies (45).

The DLPFC is a region highly associated with cognitive control, emotional regulation, expression, and response inhibition (46–48).Functional magnetic resonance imaging (fMRI) studies indicate that downregulation of negative emotions under cognitive reappraisal conditions is associated with stronger activation of the DLPFC (49). Our study found that adolescents with NSSI exhibited increased cortical activation in the DLPFC compared to the healthy control group, which is consistent with previous fMRI Mürner-Lavanchy et al., demonstrated that adolescents with non-suicidal self-injury (NSSI) exhibit marginally higher prefrontal cortex oxygenation levels during inhibitory control tasks compared to healthy controls (32), however, Koenig et al. (50)observed significantly decreased PFC oxygenation during resting state, which is different from our study; they proposed that activity of the PFC might switch from under-(adolescence) to over- (adult) activation as a function of age and development. Importantly, most of the existing studies have been conducted in relatively small samples and most studies are task-based. Furthermore, in MDD patients with NSSI behaviors, weaker functional connectivity was observed between the left dorsolateral prefrontal cortex and left premotor cortex (LDLPFC-LPMC) relative to healthy populations (24).

No significant between-group differences were observed in other ROIs (including Frontopolar area, Subcentral area, Orbitofrontal area, Middle Temporal gyrus area and pars triangularis Broca’s area), except for the DLPFC. This suggests that NSSI behavior may have specific functional impairments in the dorsolateral prefrontal cortex. Most of the fNIRS studies focused on the DLPFC, and while current fMRI studies used task- based fMRI approach, they mainly investigated the networks or connectivity of the amygdala. For example, Nam et al. demonstrated that the NSSI group had increased activity relative to the control group in the inferior parietal lobe, inferior temporal gyrus, calcarine, insula, and thalamus in response to positive self-referential stimuli. They also showed greater activation in the calcarine and reduced activation in the inferior frontal gyrus in response to negative self-referential stimuli compared with the control group (51). Ho et al. found that diminished coherence between the default mode and salience networks and higher connectivity between the central executive and default-mode networks (52). Factors such as task difficulty, complexity, and presentation methods can all influence the activation levels of brain regions. In the future, it is necessary to combine multiple paradigms or multimodal imaging to further validate the results.

We did not observe a direct correlation between cortical activation in the DLPFC and the frequency of NSSI. Few studies focused on the presence or frequency of NSSI and cortical activation. Zahid et al. observed that participants who engaged in NSSI exhibited some deactivation of the DLPFC when faced with more difficult cognitive challenges (53). These findings suggest that NSSI behavior is an intricate behavior associated with multiple psychological, social, and biological factors. It is not solely attributable to the activation of a single brain region but rather the consequence of intricate interactions among multiple systems and networks. Moving forward, we could consider incorporating mediating or moderating variables into studies examining the frequency and severity of NSSI among adolescents.

Furthermore, within the NSSI group, we identified a robust positive correlation between cortical activation in the dorsolateral prefrontal cortex (DLPFC) and social communication and expression score of NSSI-AT. This underscores the DLPFC’s pivotal role in facilitating social interaction and emotional expression. Given that social communication and expression critically influence interpersonal relationships, and considering the integrated theoretical model of NSSI which categorizes risk factors into interpersonal and intrapersonal domains (54), this finding holds significant relevance. Interpersonal risk factors in adolescents may motivate or trigger NSSI engagement (55), while intrapersonal risk factors can increase vulnerability to stress coping via emotional dysregulation (56), potentially leading to NSSI. Although adolescents with NSSI often behaviorally avoid social interaction, they paradoxically seek social support, attention, or assistance (57), reflecting a defense mechanism of reaction formation and indicating an underlying need to be understood and to establish healthy interpersonal relationships. Successful social interaction necessitates continuous environmental assessment, interpretation of others’ emotions and behaviors, and generation of appropriate responses. The observed DLPFC activation likely reflects the neural demand associated with processing this complex social information to facilitate engagement. This correlation suggests that while NSSI behavior may not be directly linked to the degree of DLPFC activation, it could be associated with deficits in social competence, emotional expression, or emotion regulation strategies. Consequently, this insight offers potential avenues for NSSI intervention and treatment by targeting the enhancement of social skills and emotional expression capabilities, thereby potentially reducing the occurrence of NSSI behaviors.

There are several limitations in the current study. First, the sample size of adolescents with NSSI is relatively small. Secondly, the age and educational years between the two groups were shown to be statistically different, thus we added these data as covariates when analyzed. However, this preliminary study is the first study examining the cortical activation in Chinese adolescents. Given that adolescence represents a critical period for DLPFC development, age disparities may contribute to the observed group differences. We will continue to include adolescents with NSSI and also recruit healthy controls who are matched in terms of demographic data to validate our preliminary results.

The specific manifestation of DLPFC cortical activation in adolescents with NSSI may serve as a candidate neural correlate for NSSI behavior. Targeted improvement of individual social skills and emotional expression abilities is expected to reduce NSSI behaviors. This discovery not only confirms the link between NSSI behavior and brain function abnormalities but also provides new insights and evidence for understanding the underlying neural mechanisms of NSSI in adolescents, laying a theoretical foundation for further clinical interventions and treatment strategies.