Risk factors, theoretical models, and biological mechanisms of nonsuicidal self-injury: a brief review

Abstract

Nonsuicidal self-injury (NSSI) refers to the direct, deliberate infliction of harm to one’s body tissue without the intention to die. The prevalence of NSSI has increased significantly globally in recent years and has become an important public health problem affecting the health of people, especially adolescents. The occurrence of NSSI in adolescents is the result of the interaction of different factors. Many scholars have proposed various theoretical models to explain the mechanism of NSSI behavior based on previous research on the influencing factors of NSSI. Moreover, advances have been made in genetic and neuroimaging mechanisms related to NSSI. Understanding the genetic and neuroimaging mechanisms of NSSI is important for both describing and treating the disorder. This literature review discusses the progress made on the risk factors, theoretical models, and biological mechanisms of NSSI.

Introduction

Nonsuicidal self-injury (NSSI) refers to the behavior that an individual deliberately and repeatedly changes or injures his or her body tissue without explicit suicide intention¹. There are different methods of NSSI, including cutting, carving, burning, hitting the head, stabbing body parts with needles or sharp objects, and causing obvious tissue damage². According to a meta-analysis conducted in 2012, the lifetime prevalence of NSSI among teenagers internationally was 17.2%³. A recent meta-analysis that included 686,672 participants showed the global lifetime prevalence of NSSI in children and adolescents was 22.1%, and the 12-month prevalence was 19.5%⁴. There was an upward trend in NSSI prevalence that needs attention. An increase in NSSI was especially noted during the coronavirus disease 2019 pandemic^5,6. In clinical settings, patients with mental disorders had a higher risk of NSSI than patients with somatic diseases (relative risk > 5)⁷.

A retrospective nested case-control study was conducted in Hong Kong to investigate the incidence of self-harm among patients hospitalized with 7 distinct types of mental health issues between 2000 and 2010 in public hospitals. The study findings revealed that all patient groups exhibited a markedly elevated risk of self-harm. The risk in the patients with substance misuse or dependence was ~10 times higher than the control group, followed by personality disorder, alcohol dependence, depression disorder, bipolar disorder, schizophrenia, and anxiety disorder⁸. A meta-analysis conducted by another research team reported a greater risk of NSSI among patients diagnosed with emotional disorders compared with those without such diagnoses. The findings also revealed that there was no discernible variation in NSSI risk across different subtypes of emotional disorders⁹. These results underscore the importance of considering NSSI in patients with mental disorders, particularly those with emotional disorders.

NSSI is closely associated with suicide and has been identified as a robust predictor of future suicidal behaviors¹. Moreover, NSSI has been found to increase the risk of suicide by up to seven-fold, independent of other known risk factors^10,11. According to the interpersonal theory of suicide, although NSSI is considered to be an effective way to resist suicide ideation (SI) or suicide attempt (SA) in the short term, repeated NSSI may make an individual less pain sensitive and thus enhance the ability to transform suicidal thoughts to behavior^12,13. A previous study using ecological momentary assessment real-time monitoring showed that patients’ SI was significantly reduced after the onset of NSSI. Lastly, a meta-analysis showed that the most common outcome of patients with self-harm was SA (47.8%), followed by death (40.5%) and suicide concept (11.6%) during a period of 4–5 years¹⁴. Thus, NSSI plays a potential role in indicating suicide risk.

NSSI has emerged as a significant mental health concern that necessitates prompt attention. Leading experts have recommended that NSSI be recognized as a distinct disorder and have called for additional research in this area. To enhance our comprehension of the etiology and persistence of NSSI, this review endeavors to explicate the risk factors, theoretical models, and biological mechanisms, such as genetic biology and neuroimaging, that underlie NSSI.

Risk factors

The occurrence of NSSI in adolescents is the result of interactions between numerous factors. An illustration of this phenomenon is reflected in the modification of neural pathways, which signifies the biological foundation and prerequisite for an individual’s engagement in self-injury. Subsequently, when an individual is exposed to specific environmental factors, particularly acute stressors (such as peer victimization, bullying, and loss of interpersonal relationships), the probability of self-injury may escalate^15,16. This is demonstrated by the findings of a meta-analysis that identified numerous personal and environmental factors as significant predictors of NSSI¹⁷. These factors encompass sex, physical symptoms (eg, disabilities and sleep problems), mental disorders, low health literacy, adverse childhood experiences, bullying, and problem behaviors.

Regarding personal factors, a mental disorder diagnosis may heighten the risk of NSSI¹⁸. Specifically, NSSI was observed to be significantly more widespread among patients with personality disorders, trailed by those with eating disorders, bipolar disorder, and mood disorders¹⁸. NSSI is also seen in patients with anxiety disorder, substance abuse, or dependence^19–21. Moreover, depression and anxiety symptoms are common emotional symptoms in adolescents and are influencing factors of NSSI. Individuals with depressive symptoms are more likely to choose self-injurious behaviors to regulate their emotions compared with those without depression^22–24. Lastly, comorbid anxiety disorder constitutes a risk factor for NSSI among individuals with depression^25,26.

There is also evidence that internet addiction is related to NSSI²⁷. Tang et al²⁸ (2020) showed that adolescents aged 15–17 with internet addiction were significantly associated with NSSI, but it was difficult to infer the causal relationship between internet addiction and NSSI. Based on a study conducted in Hungary, pathologic network use was not found to be a risk factor for NSSI, except when it co-occurred with certain mental disorders, such as mood disorders, psychoactive substance abuse, and dependence²⁹.

With regard to personality traits, individuals with NSSI tend to exhibit elevated levels of neuroticism compared with those without NSSI³⁰. This observation highlights neuroticism as a noteworthy risk factor for NSSI. In the five-factor model of personality, neuroticism represents the tendency of individuals to experience negative emotions³¹. Individuals with heightened levels of neuroticism may be predisposed to a range of difficulties, such as daily issues, conflicts with friends, unstable familial relationships, and other disharmonious social interactions^32,33. In addition, they may be more vulnerable to negative stressors and experience a greater burden of mental or physical health-related problems. Individuals in the cluster with high neuroticism have significantly higher NSSI frequency compared with other clusters³⁴.

Negative environmental factors exert a notable influence on the incidence of NSSI, including adverse childhood experiences, negative life events, poor family functioning, and inadequate social support. Life events can serve as short-term and long-term risk factors for NSSI^35,36. In the short term, life events may lead to abnormal functions of the immune system and stress response system and lead to individuals being vulnerable to stress^37,38. In the longer term, the number of life events predicts the occurrence of self-injury in the next year; therefore, life events may have a significant impact on NSSI³⁹. Childhood abuse was believed to cause intra or interpersonal vulnerabilities that predispose individuals to use NSSI to modulate affective or social dysregulation⁴⁰. Meta-analysis showed a significant correlation between childhood abuse and NSSI⁴¹. In addition, due to the important role of parenting methods in the development of psychopathology, family factors were considered to have a significant impact on NSSI⁴². The research found that family dysfunction was associated with the risk of NSSI⁴³. In terms of social support, high-level social support was a protective factor for NSSI, whereas low-level social support was associated with an increase in NSSI risk^44,45. Thus, a correlation has been established between a suboptimal environment and the emergence of NSSI, whereas an improved environment has been linked to a decrease in self-injury⁴⁶.

Despite the identification of numerous risk factors, a comprehensive understanding of NSSI necessitates an integrative approach. To significantly enhance predictive accuracy beyond individual risk factors, it is imperative to integrate a range of NSSI risk factors and examine how they interact. Adopting a perspective that accounts for the interplay between personal and environmental factors could offer valuable insights into the onset and maintenance of NSSI.

Theoretical models

NSSI can be more systematically understood from the aspects of occurrence conditions, motivations, promotion factors, and results. Previous studies have proposed various theoretical models to further the understanding of NSSI (Table 1).

Table 1.

The functional models and etiological models of NSSI.

Model	Authors	The Core of the Model
Functional models
1. The four-function model (2004)36	Nock and Prinstein	The 4 primary functions of NSSI include 2 dimensions: automatic regulation versus social regulation, and positive reinforcement versus negative reinforcement
2. Seven function models (2007)12	Klonsky	Seven functions include affect regulation, antidissociation, antisuicide, interpersonal boundaries, interpersonal-influence, self-punishment, and sensation-seeking
Etiological models
1. The developmental psychopathology model (2004)47	Yates	NSSI is a compensation strategy for relational adaptation and regulatory adaptation based on trauma-induced adaptive functional defects
2. The experiential avoidance model (2006)48	Chapman et al	NSSI is primarily maintained by negative reinforcement in the form of escape from, or avoidance of, unwanted emotional experiences
3. The emotional cascade model (2008)49	Selby et al	High levels of rumination may cause extremely intense states of negative affect, which result in dysregulated behaviors (NSSI) that can distract from rumination, and thus reduce that state of negative affect
4. The cognitive-emotional model (2017)50	Hasking et al	The model is based on (1) the process model of emotion regulation, (2) the model of difficulties in emotion regulation, (3) the experiential avoidance model, and (4) the emotional cascade model. Factors will increase the risk of NSSI including (1) a propensity towards emotional reactivity, (2) negative self-schemas, (3) outcome expectancies which hold that performing NSSI will achieve a desired mental state, (4) a belief in the ability to self-injure, or limited belief in the ability to resist NSSI, and (5) poor emotion regulation strategies
5. The benefits and barriers model (2018)51	Hooley and Franklin	NSSI carries many powerful benefits including (1) self-punishment benefit, (2) affective benefit, (3) communication benefit, and (4) affiliation benefit. Most people do not access these benefits because there are several barriers that motivate them to avoid self-injury including (1) awareness barrier, (2) aversion barrier, (3) pain barrier, (4) positive self-barrier, and (5) social norms barrier
6. The integrated theoretical model (2009)35	Nock	NSSI can serve several intrapersonal and interpersonal functions. The risk of NSSI is increased by general factors that contribute to problems with affect regulation or interpersonal communication and by specific factors that influence the decision to use NSSI rather than other behavior to serve these functions
7. The integrated motivational-volitional model (2012)52,53	Connor et al	The model is based on (1) the arrested flight model, (2) the diathesis-stress hypothesis, and (3) a dominant model of health behavior. NSSI develops through three phases: the premotivational phase, the motivational phase, and the volitional phase. The occurrence of NSSI depends on motivational moderators and volitional moderators, including a range of factors
8. The behavioral-functional analysis model (2020)54	Minlu Liang et al	The possible causes, susceptibility, process, and maintenance factors are important parts of the development and progression of NSSI

NSSI indicates nonsuicidal self-injury.

Functional models

The functions were conceptualized as motivating or reinforcing the behaviors¹². Nock and Prinstein (2004)³⁶ proposed a four-function model to evaluate the functions of self-mutilative behavior (SMB, the same as NSSI) that includes 2 dimensions: automatic regulation versus social regulation and positive reinforcement (triggering favorable stimulus) versus negative reinforcement (eliminating aversive stimulus). Specifically, automatic-negative reinforcement means individuals reducing tension or other negative emotional states through NSSI. Automatic-positive reinforcement means individuals consider NSSI as a way to generate emotion and create an ideal state, such as experiencing stimulation or pain. Social-negative reinforcement means individuals using NSSI to avoid interpersonal task demands and adjust their social environment, such as avoiding punishment or doing something unpleasant. Social-positive reinforcement means that an individual obtain the attention of others through NSSI and changes interpersonal relationships, for example, letting others know about his negative emotions or trying to get a response from someone. The model shows that NSSI can be maintained and strengthened through these 4 processes. Different treatment methods should be considered according to the different functions of NSSI, and NSSI should be replaced with behaviors that have the same function.

Klonsky (2007)¹² presented a taxonomy of seven-functional models to delineate the diverse functions of NSSI. These models encompass affect regulation, sensation seeking, antidissociation, interpersonal boundaries, interpersonal influence, antisuicide, and self-punishment. The affect-regulation function shows that self-injury is a method to relieve negative emotions or stimulate positive emotions. The sensation-seeking function believes that self-injury can be used to generate exhilaration or excitement. The antidissociation function, also called feeling generation, shows that self-injury makes individuals feel real or alive again by generating emotional and physical feelings. The interpersonal boundaries function suggests that self-injury is a sign of the difference between oneself and others, or it can be used to emphasize identity or autonomy. The interpersonal-influence function considers that self-injury is used to influence the interpersonal environment or control others. The antisuicide function regarded NSSI as a strategy for resistance to SI or SA in the short term. The self-punishment function regards self-injury as a way to derogate oneself or express anger towards oneself. There is a partial overlap between the seven-function model and the four-function model, and the 2 models can complement each other in clinical practice⁵⁵.

Integrative model

Multiple theoretical models explain why people engage in self-injury. Yates (2004)⁴⁷ constructed a developmental psychopathology model based on the developmental organization theory. In the developmental psychopathology model, individuals gradually develop 5 abilities in the process of growth, including motivational competence, attitudinal competence, instrumental competence, emotional competence, and relational competence. Through these abilities, individuals can effectively deal with the problems they face in current and future development, thus generating positive adaptation to society. The traumatic experiences in childhood, especially those from family care relationships, will affect the development of individuals’ abilities. The development of the psychopathology model emphasizes that based on trauma-induced adaptive functional defects, self-injury is a compensation strategy for relational adaptation and regulatory adaptation. This model provided theoretical support for some studies exploring the relationship between traumatic events and NSSI^56–58.

Experience avoidance pertains to the behavioral pattern of individuals to evade any negative internal experiences or prevent the emergence of such experiences, including unpleasant or distressing thoughts, emotions, bodily sensations, or other internal experiences⁵⁹. This behavior is predominantly reinforced by negative reinforcement. Intentional self-injury is a kind of behavior resulting from an ineffective emotional response, which can help individuals avoid the internal experience of thought, memory, body feeling, or other aversions, despite obvious negative consequences. According to the experience avoidance model, intentional self-injury is maintained and strengthened through avoidance regulation and negative reinforcement⁴⁸. In essence, an emotional arousal event will trigger an aversive emotional response⁴⁸. The individual has an impulse to avoid unpleasant experiences and has intentional self-injury behavior, which will reduce or eliminate emotional arousal, thus strengthening the intentional self-injury behavior through the negative reinforcement process. In the cycle of maladaptive behavior, recurrent negative reinforcement reinforces the association between negative emotions and deliberate self-injury, thereby facilitating the automaticity of the escape response of deliberate self-injury^60,61. This model was also applied by studies about moderated mediation models about NSSI^62,63.

The emotional cascade model (ECA) was developed and tested by structural equation modeling and may elucidate the mechanism of NSSI alleviating negative emotions⁴⁹. In an emotional cascade, repeated thinking about negative emotions can increase the level of negative emotions, which in turn strengthens the individual’s attention to emotional stimuli, leading to more rumination⁴⁹. According to the ECA, the pain and visual stimuli caused by NSSI can shift attention from rumination to strong physical sensations related to NSSI; thereby the emotional cascade process is weakened⁴⁹. This model provided the theoretical reference for the study of the relationship between rumination, negative emotions, and NSSI⁶⁴.

Based on the experience avoidance model and the ECA, other researchers further combined models of emotion regulation––for example, process model of emotion regulation and difficulties in emotion regulation model––and developed a cognitive-emotional model to explain the onset and maintenance of NSSI from the perspectives of emotion regulation and cognition⁵⁰. According to the cognitive-emotional model, if individuals are highly reactive to negative emotions, encounter difficulties in regulating these emotions, and recognize NSSI as an effective strategy for regulating these emotions, then they are more likely to engage in NSSI. On the contrary, those who believe they can endure pain, resist the urge to NSSI, and use other strategies to manage emotions are less likely to engage in NSSI⁶⁵. The benefits and barriers model holds a similar opinion that NSSI may provide a range of benefits, including self-punishment benefit, affective benefit, communication benefit, and affiliation benefit, but most people are unable to obtain benefits from NSSI because of barriers, including awareness, aversion, pain, positive self, and social norms⁵¹. This model also supports that distal risk factors generate proximal risk factors and further lower the barriers to NSSI.

Based on the proposed four-functional model of NSSI, Nock (2009)⁴⁰ further integrated findings from different fields, summarized the factors that increase the risk of NSSI, hypothesized about the choice of NSSI instead of other regulatory strategies, and developed the integrated theoretical model. The model proposes that certain individuals may develop vulnerability in their internal or interpersonal relationships during their developmental process, rendering them susceptible to emotional or social dysfunctions in the face of challenging or stressful events. NSSI or other extreme behaviors may then serve as a regulatory mechanism for such dysfunctions. These vulnerability factors may be further compounded by more remote risk factors. Researchers have suggested several potential explanations for why individuals experiencing stress events may opt for NSSI over other coping strategies, such as the social learning hypothesis, social signal hypothesis, and self-punishment hypothesis.

The integrated motivation-volitional model is a tripartite framework, including biological, psychological, and social factors. It integrates the main contents of a variety of models (eg, diathesis-stress hypothesis, dominant model of health behavior, and cry of pain model), and provides an explanation for NSSI^52,53. The integrated motivation-volitional model posits that self-injury occurs in 3 stages: the premotivation stage, the motivation stage, and the volitional stage. To assess the relevance of established risk factors for self-harm across these stages, researchers investigated and found that the model can effectively predict stage-specific risk factors during the occurrence of NSSI⁵³.

Liang et al (2020)⁵⁴ provided a summary of the etiological factors of NSSI and elucidated them by using the “behavioral-functional analysis model.” The possible causes, susceptibility, process, and maintenance factors of NSSI were analyzed in the model. The potential etiological factors exert an impact on vulnerable individuals, leading to responses in four domains, including cognition, emotion, body, and behavior, ultimately resulting in self-injurious behavior. The result of self-injury behavior strengthens the behavior and is the maintenance factor for its persistence and recurrence. However, this model has been comparatively underutilized in other investigations.

Further theoretical models focus on different contents. The functional models primarily explain the regulation function of NSSI in individual emotion and social interpersonal and the theory of behavior maintenance and strengthening^12,36. The four-function model was widely applied in functional research of NSSI^66–70. As for the integrative model, the early models tried to explain the occurrence and maintenance of NSSI from the aspects of personal vulnerability factors and environmental exposure factors or the aspect of emotional regulation, whereas the newly proposed model in recent years integrated various previous models to explain the occurrence and development of NSSI from a more comprehensive perspective^{40,47–50,53}.

To conclude, these models suggest that individual susceptibility or vulnerability, negative stress events, and remission states after NSSI are crucial factors in the occurrence, maintenance, and reinforcement of this behavior. During this process, emotional dysfunction and cognitive-related changes play an important role. There are various existing theoretical models, and the development of theoretical models is at the summary stage. Future research should prioritize selecting suitable models that can guide clinical practice rather than solely adhering to the theoretical level.

Genetic mechanism

Based on the examination of the factors that influence NSSI, it has been determined that this behavior is impacted by mental disorders. The field of genetic research on NSSI is still in its early stages, although limited evidence suggests that genetic factors may play a role in the development of this behavior. One study that utilized a sizable twin sample found that genetic factors were responsible for the majority of variability in NSSI, whereas the remaining variance was attributed to nonshared environmental effects and measurement errors⁷¹.

Candidate genes

Studies have shown that genetic factors play a role in the development of intentional self-harm behavior. One study tested 6 serotonergic gene polymorphisms in 129 subjects with intentional self-harm and 329 control subjects⁷². They found that intentional self-harm was related to allelic variation in the tryptophan hydroxylase gene (TPH A779C) (odds ratio = 1.38, 95% CI = 1.02~1.88; P = 0.03). However, the molecular mechanism of A779C polymorphism-mediated intentional self-harm has not yet been clearly explained.

Another study explored the relationship between the catechol-O-methyltransferase (COMT) gene and suicidal behavior in a sample of patients with affective disorders⁷³. The results showed that rs737865, rs6269, and rs4633 polymorphic loci of the COMT gene were significantly correlated with NSSI. Though the P values did not resist multiple testing corrections, they had a certain implication for the exploration of genetic susceptibility loci of NSSI.

Solute carrier family 1 (glial high-affinity glutamate transporter), member 3 (SLC1A3) is one of the important transporters. A study found that the allele frequency of rs2269272 of the SLC1A3 gene was significantly different between the NSSI group and the control group (χ2 = 4.208, P = 0.040), which indicated that SLC1A3 might be associated with NSSI in adolescents⁷⁴.

A study conducted by Zheng and colleagues (2020) found that the methylation in the promoter region of Proopiomelanocortin in depressive patients with NSSI was abnormal. It pointed out that the methylation of CpG1 can be regarded as an epigenetic marker for depressive adolescents with NSSI⁷⁵. While it was difficult to differentiate whether the methylation irregularity was a result of depression or NSSI, this implied the possibility of exploring epigenetic markers related to NSSI in this area. Another study reported that in individuals with NSSI, the methylation level at the cytosine-guanine dinucleotide 5 site of the silent information regulator 2 related enzyme 1 gene promoter region was significantly elevated compared with the control group, which may assist in identifying molecular markers in adolescents with NSSI⁷⁶.

Genome-wide association study

The candidate gene approach is based on a prior hypothesis, whereas the genome-wide association study (GWAS) is an exploratory study driven by data. The GWAS can simultaneously investigate the associations between a phenotype and hundreds of thousands to millions of single nucleotide polymorphisms (SNPs). A study that used data from UK Biobank conducted a GWAS on self-injurious thoughts and behaviors regardless of suicidal intent⁷⁷. They identified a locus significantly associated with self-injurious thoughts and behavior phenotypes at the genome-wide level. SNP heritability (h ² _snp ) estimated was ~10%. Another study showed that the h ² _snp of NSSI was 0.132⁷⁸. This study revealed a low degree of overlap in the genetic architecture of NSSI and related phenotypes, suggesting that NSSI may manifest independently of mental illnesses. Nevertheless, the relationship between NSSI and mental disorders is intricate and multifaceted^8,9,79. Lim et al (2020)⁸⁰ used the polygenic score of 24 risk factors to predict NSSI through UK Biobank data and found that major depressive disorder (MDD) can significantly forecast the occurrence of NSSI (odds ratio = 1.20, 95% CI = 1.16–1.23, P _adjust = 1.33×10⁻³⁵), further Mendelian randomization analysis found that the most likely causal risk factor for self-injury was MDD. Yet, a previous genetic risk prediction study found that individuals with a genetic predisposition to depression were at greater risk for suicidal thoughts and behaviors, but there was no evidence for a higher risk of NSSI⁸¹.

Interaction between gene and environment

In recent years, an increasing number of researchers have advocated for the inclusion of environmental factors in investigations of how gene-environment interactions influence human behavior through the brain. A study investigated retrospective reports of childhood emotional environment and analyzed its interaction with the BDNF Val66Met polymorphism possibly associated with NSSI⁸². Emotional abuse was significantly and directly associated with self-injurious behavior in individuals with 2 Val alleles. However, this correlation was not shown in Met carriers. Another study found that the coinheritance of at least one 5-HTTLPR short allele with long-term interpersonal stress confers a higher risk for NSSI⁸³. In addition, a study showed that the monoamine oxidase A and COMT can modulate the effect of childhood maltreatment on NSSI, even though they were not significantly and directly related to NSSI⁸⁴. Liu et al (2021)⁸⁵ found that high levels of harsh parenting had a relation to severe NSSI, and the COMT Val158Met polymorphism moderated the association between them.

Another theoretical model that explains the interplay between genetics and environment in shaping human development is known as gene-environment correlation. This model posits that genes can influence an individual’s surrounding environment, thereby impacting their psychology and behavior. A study involving 9526 pairs of twins discovered that self-injurious behavior and high-risk trauma exposure shared common genetic factors, which were primarily attributed to heritable traits influencing vulnerability to trauma and psychopathology, including neuroticism, antisocial and other personality traits, and difficulties with emotional regulation⁸⁶. These genetic factors influenced the occurrence of self-injurious behavior through phenotypic causality. In phenotypic causality, genetic characteristics can affect exposure to traumatic events. Traumatic experience makes individuals more likely to have self-injurious behavior. On the contrary, individuals with self-injurious behavior are more likely to be exposed to traumatic events.

In conclusion, the etiology of NSSI may be attributed to the combined influence of multiple genetic variations; however, the available evidence is inconclusive. There is heterogeneity in the candidate gene research, such as variations in the participants’ ethnicity. Moreover, those studies lack a control group comprising patients without NSSI^74–76,86. Determining the underlying causes of observed differences in research outcomes between NSSI and related disorders can be challenging, as it remains unclear whether the differences are primarily attributable to NSSI, the presence of concurrent disorders, or the interaction between these factors. Further research is required to explore the reliability of research findings. In the event of conflicting results, meta-analysis can be used as a strategy to resolve inconsistencies, provided sufficient studies are available in the future.

Neuroimaging mechanism

The rapid progress of neuroimaging provides important biological evidence for exploring the mechanisms of NSSI. Neuroimaging studies found that adolescents with NSSI had abnormal brain structure and brain functions⁸⁷. This biological vulnerability is thought to necessitate alterations to neural circuits to facilitate engagement in self-injurious behavior⁸⁷. On this basis, experiencing some environmental exposure, especially acute stress, such as peer victimization and bullying, the occurrence of self-injury may increase^15,16,87.

Imaging biomarkers of brain structure

Currently, there is limited evidence suggesting the existence of specific structural changes in the brains of individuals with NSSI that are linked to emotional processes. A previous study found that the higher frequency of self-injury in the past 6 months was related to the smaller volume of the left anterior cingulate gyrus⁸⁸. Beauchaine et al (2019)⁸⁹ reported that the gray matter volume in the right inferior frontal cortex of adolescents with NSSI was reduced compared with healthy people. Several studies focusing on the insula found that the volume of insula gray matter in the self-injury group was reduced compared with the healthy control^89,90. Due to the lack of psychiatric samples without NSSI in the mentioned studies, we cannot determine whether the effects are specific to NSSI.

Imaging biomarkers of brain functional connectivity

Several studies have explored brain functional changes in patients with NSSI using functional magnetic resonance imaging⁸⁷. One study tested amygdala function in female adolescent patients with NSSI, using both resting-state and task-functional magnetic resonance imaging⁹¹. The task used a negative emotional face match that could activate the amygdala. The researchers found that NSSI patients showed significant differences in resting-state functional connectivity (rsFC) between the amygdala and the frontal and occipital lobes compared with healthy controls, as did rsFC in the supplementary motor area (SMA) and the dorsal anterior cingulate cortex. Moreover, NSSI patients had greater amygdala-frontal task functional connectivity differences during the performance of the task than at rest, and the same differences were found in amygdala-occipital task functional connectivity compared with healthy controls. After regressing the severity of depressive symptoms as a covariate, the amygdala-SMA rsFC abnormalities remained significant between the two groups. The decreased connectivity between the amygdala and frontal lobe may represent the difficulty of negative emotion regulation and dependence on NSSI as a regulation strategy. The increased connectivity between the amygdala and SMA may lead to changes in the connection between negative affect and planning of movements, thereby increasing the likelihood of NSSI.

Another study implemented an N-acetylcysteine intervention in female adolescents with NSSI and compared rsFC changes in the amygdala and nucleus accumbens before and after the intervention⁹². It was found that the decrease in NSSI frequency was accompanied by decreased rsFC in the left amygdala-right SMA, right nucleus accumbens-left superior medial frontal cortex and increased rsFC in the right amygdala-right inferior frontal cortex. Another interventional study implementing psychotherapy found abnormal alterations in rsFC in the amygdala-medial prefrontal cortex (mPFC) in adolescents with NSSI compared with healthy controls. In addition, the stronger the amygdala-prefrontal negative connectivity before the intervention, the greater the improvement in NSSI after psychotherapy⁹³. The results indicate that the activation of the amygdala in NSSI adolescents was abnormal, and there were changes in functional connectivity between the amygdala and frontal lobe, cingulate cortex, or SMA. The amygdala is a key limbic region that initiates the threat response and plays a central role in negative affect⁹¹. As NSSI is often used to regulate negative emotion, widespread amygdala circuitry anomalies in adolescents with NSSI may be the key neural mechanism for emotional dysfunction and can be used as neural targets for treatment or predictive indicators of treatment effectiveness^69,91–93.

NSSI patients also showed abnormal activation of some brain regions in different tasks. In the emotional processing task, compared with healthy adolescents, the activation of the bilateral amygdala, hippocampus, and anterior cingulate gyrus increased under the stimulation of negative pictures and self-injury-related pictures, but the difference between groups can be partly explained by depression⁹⁴. In social processing tasks, compared with depressed adolescents, depressed patients with NSSI showed activation of mPFC, ventrolateral PFC, and parahippocampal gyrus in social exclusion and social acceptance tests⁹⁵. The activation of PFC was linked to ineffective emotion regulation, self-referential processes, and evaluation of negative feeling states^96,97. The study pointed out that there may be deficiencies in recognizing and neural processing social contexts, which seemed unique to NSSI⁹⁵. In the reward processing task, the reward expectation of putamen, orbitofrontal cortex, and amygdala of self-injured adolescents was weaker than that of healthy controls, and the activation of bilateral putamen in high-risk adolescents was higher than that in the low-risk group^98,99. The orbitofrontal cortex, amygdala, and putamen all play important roles in the evaluation of reward¹⁰⁰. The decreased activation of these brain regions may indicate that self-injuring adolescents experienced limited hedonic value from general rewards and thus engaged in increasingly frequent reward-seeking behaviors such as self-injury. The specificity of alterations in reward processing to NSSI remains uncertain.

At present, the neuroimaging findings of NSSI mainly include the amygdala, cingulate gyrus, and PFC in the limbic system, involving emotional perception processing and emotional regulation. It also includes many brain regions in the default network and salience network, such as the anterior insular lobe, dorsal anterior cingulate gyrus, posterior cingulate gyrus, mPFC, and so on, which are related to social cognitive processing. However, some of the studies included suicidal self-injurious behaviors together with NSSI behaviors or only compared patients with a mental illness accompanied by NSSI with normal controls and lacked the same kind of illness patients without NSSI behaviors with control groups^{88,89,91,95,98}. Distinguishing the potential influence of psychiatric disorders on the observed results is challenging, and the replicability of the findings is also uncertain.

Conclusion

With the rapid development of the social economy, social pressure and competition are becoming increasingly prominent. The incidence of NSSI in adolescents and early adulthood is increasing yearly, especially in patients with various emotional and personality disorders, which seriously damages the prognosis of patients. In the research on the influencing factors of NSSI, individual vulnerability traits, environmental factors, and various biological factors significantly affect its occurrence. Despite the emergence of numerous novel findings, the neural and genetic mechanisms underlying NSSI remain uncertain. In the current neuroimaging research, the focuses are on the amygdala, cingulate gyrus, prefrontal lobe, and other brain regions related to emotion regulation, pain processing, and reward processing, among others. The focus of existing genetic studies is still the association between genes and suicide, and NSSI is just one of the secondary outcomes concerned by the research. To advance early prevention and intervention efforts and seek breakthroughs in treatment, it is crucial to obtain a comprehensive understanding of the current state of NSSI in adolescents, including its clinical characteristics, risk factors, and potential mechanisms. Such knowledge would provide significant clinical value and research significance. Future research needs to focus on the transformation of theoretical models into clinical practice. In addition, the research foundation of neurobiology needs to be further expanded to discover more potential therapeutic targets. The above possible studies will be helpful for breakthroughs in the treatment of NSSI.

Funding sources

This study was supported by the National Natural Science Foundation of China (81825009); National Key R&D Program of China (2021YFF1201103); Academy of Medical Sciences Research Unit (2019-I2M-5-006); Chinese Institute for Brain Research at Beijing (2020-NKX-XM-12); PKUHSC-KCL Joint Medical Research (BMU2020KCL001); Major science and technology projects of Henan Province (201300310200).

Conflict of interest

None.