Impaired neural activity and functional connectivity in the hippocampus of adolescents with non-suicidal self-injury addiction

Abstract

Background

Non-suicidal self-injury (NSSI) addiction is prevalent among adolescents, but its underlying neural mechanisms remain unclear. This study aims to investigate the neural activity and functional connectivity characteristics associated with NSSI addiction using resting-state functional magnetic resonance imaging (rs-fMRI).

Methods

A prospective collection of 62 adolescents was completed for this study, including 33 adolescents with self-injury behaviors and 29 age-, gender-, and education-matched healthy controls. The addiction component of the Ottawa Self-Injury Inventory (OSI) was used to assess the degree of NSSI addiction. Amplitude of low-frequency fluctuation (ALFF) analysis was employed to detect changes in local neural activity. Differential brain regions were considered regions of interests (ROIs). Whole-brain functional connectivity (FC) analysis based on ALFF was used to further explore potential changes in functional connections between ROIs and other brain areas in the NSSI group, and to analyze the relationship between these neural changes and addiction characteristics.

Results

ALFF analysis revealed decreased ALFF values in the bilateral hippocampus and increased ALFF values in the right supplementary motor area of NSSI adolescents compared to healthy controls. Significantly reduced FC values was observed between the left hippocampus and the left precuneus, right middle temporal gyrus, and right inferior temporal gyrus, and between the right hippocampus and the right middle temporal gyrus. Additionally, increased FC values was observed between the left hippocampus and the left thalamus. Furthermore, ALFF values in the bilateral hippocampus were negatively correlated with the total score of addiction characteristics in NSSI adolescents.

Conclusions

This study highlights reduced local neural activity and functional connectivity in the bilateral hippocampus of NSSI adolescents, and demonstrates that these alterations are associated with heightened addictive features in self-injuring individuals.

Trial registration

A study of positive psychological group interventions in adolescents with non-suicidal self-injury (registration date: 03/01/2024; registration number: ChiCTR2400079412).

Introduction

Non-Suicidal Self-Injury (NSSI) refers to the act of intentionally harming one’s own body tissue without suicidal intent, such as cutting, hitting, stabbing, and burning [1]. Adolescents, due to the imbalance between rapid mental maturation and relatively lagging emotional regulation capabilities, have become a high-risk group for NSSI [2]. The global prevalence of NSSI among adolescents is approximately 18% [3], with a male-to-female ratio of 1:3. The detection rate of NSSI among Chinese adolescents is 21.9%, making it a significant threat to the physical and mental health of Chinese adolescents in recent years [4]. Currently, the fifth edition of the American Diagnostic and Statistical Manual of Mental Disorders (DSM-5) has officially listed NSSI as an independent clinical diagnosis. NSSI is not only associated with various mental disorders such as depression, anxiety, bipolar disorder, and borderline personality disorder [5], but it is also one of the important predictive indicators for suicide [6].

Repetitive NSSI is considered an addictive behavior. Its addictive characteristics may be related to the interaction mechanisms of behavior reinforcement and the emotional, cognitive, and behavioral mechanisms of susceptible individuals. Many adolescents who become entangled in this behavior find it difficult to extricate themselves, which may lead to more severe mental disorders [7, 8]. Studies have shown that individuals with addictive characteristics have a higher frequency of NSSI, more methods and parts of the body harmed, and more daily impulsivity [9].Mechanisms associated with the addictive nature of NSSI behavior involve neurobiological mechanisms that are similar to those involved in substance addiction, such as the dopamine reward system and the endogenous opioid system [10]. The hippocampus plays a key role in learning and memory and is also closely related to addictive behaviors, playing an important regulatory role in the formation and maintenance of addictive behaviors. Studies have found that the hippocampus, amygdala, and corpus callosum are brain structures sensitive to early negative stress experiences, and damage or dysfunction in these areas may increase the incidence of NSSI [11]. The hippocampus also plays a key role in addiction-related memories associated with drugs [12].

Resting-state functional magnetic resonance imaging (rs-fMRI) is a technique for measuring spontaneous neural activity in the brain at rest and is an important research method for brain function studies. Amplitude of low-frequency fluctuation (ALFF) and resting-state functional connectivity (rs-FC) analysis are widely used and effective rs-fMRI analysis methods. ALFF analysis is used to detect changes in local neural activity. Rs-FC analysis aims to discover the correlated fluctuations of blood oxygen-level dependent (BOLD) signals in distant brain regions. Previous MRI studies on NSSI adolescents have been limited, with some studies finding brain regions associated with NSSI. Studies have shown that compared with adolescents without NSSI, those with NSSI have changes in brain activity and corresponding brain circuits related to visual and emotional regulation functions, with significant abnormalities in the ALFF values of the lingual gyrus and fusiform gyrus, as well as parts of the cingulate gyrus and left frontal superior gyrus [13]. Other studies have found significant differences in the amygdala, cingulate gyrus, orbital part of the frontal gyrus, precuneus, putamen, and striatum in adolescents with NSSI and major depressive disorder compared to healthy control groups [14–17]. The above studies have found that brain regions in the default mode network (DMN) are mainly related to NSSI. In addition, studies on NSSI behavior in adolescents with depression have found that FC changes related to NSSI may include the bilateral nucleus accumbens, right putamen, and bilateral cingulate gyrus in the reward circuit [18]. Most previous studies have been based on exploration of ALFF or FC, lacking integrated analysis of ALFF and FC and research on the addictive characteristics of NSSI adolescents. To more comprehensively explore the addictive characteristics of NSSI adolescents, this study applies an integrated ALFF and rs-FC analysis method to deeply analyze the neural activity and functional connectivity in the brains of NSSI adolescents, aiming to reveal the intrinsic connections between these changes and NSSI addictive characteristics. Through this study, we hope to provide reference strategies based on brain regions for the precise treatment of NSSI in adolescents in the future, promoting innovation and development of clinical intervention methods in this field.

Methods

Subjects

A total of 33 diagnosed Han Chinese, right-handed adolescent NSSI patients (NSSI group) were recruited from the People’s Hospital of Henan Province. They met the diagnostic criteria for “non-suicidal self-injury” suggested by the fifth edition of the American Diagnostic and Statistical Manual of Mental Disorders (DSM-5). According to the Ottawa Self-Injury Scale (OSI), they had more than five self-injury incidents in the past year and at least one self-injury behavior in the month before enrollment. No patients had taken any psychotropic medications for at least 4 weeks before participating in the study. None had ever attempted suicide. Subjects with a history of drug abuse, brain trauma, other serious diseases, or MRI contraindications were excluded. In addition, 29 right-handed healthy adolescents (HC group) were recruited through advertisements. Exclusion criteria included any past self-harm behavior, drug abuse history; brain trauma or other serious disease history; any past mental health disorder history; any contraindications for MRI scanning.

This study follows the Declaration of Helsinki and was registered and reviewed by the China Clinical Trials Center (Date of Registration: 2024-01-03, Clinical Trial Number: ChiCTR2400079412). The study was approved by the Ethics Committee of Zhengzhou University (No. 2023 − 275). All participants were informed about the study in detail in the presence of their parents or legal guardians and participated voluntarily, signing informed consent forms.

Clinical assessment

NSSI was assessed by two experienced psychiatrists through clinical interviews related to NSSI behavior, based on the diagnostic criteria recommended by DSM-5, to determine whether patients had NSSI, regardless of whether they or their parents disclosed NSSI. To obtain as much information about patients’ NSSI behavior as possible, we asked each recruited patient to complete the Ottawa Self-Injury Inventory (OSI) assessment [19]. It comprises two validated subscales: the Functional Dimension and Addictive Dimension [9].The Functional Dimension evaluates the psychological or social purposes served by NSSI behaviors, containing 30 items. Each item is rated on a 5-point Likert scale (0="Never” to 4="Always”), with total scores ranging from 0 to 120. Higher scores indicate stronger functional dependence on NSSI behaviors.The Addictive Dimension Quantifies craving intensity, loss of control, and withdrawal symptoms associated with NSSI, containing 7 symptom-frequency items. Items are scored from 0 (“Never”) to 4 (“Always”), with total scores ranging from 0 to 28. The scale demonstrates good reliability in Chinese adolescent populations [20].

Additionally, the Generalized Anxiety Disorder-7 (GAD-7) [21] and the Patient Health Questionnaire-9 (PHQ-9) [22] were used to assess patients’ anxiety and depression status. The Chinese versions of the above scales have good reliability and validity in adolescents. The self-assessment and entire clinical assessment process for NSSI adolescents were conducted with the participation of experienced psychiatrists to ensure that the self-assessment data were relatively accurate, objective, and comprehensive.

Acquisition of rs-fMRI data

Within 36 h after the clinical assessment, all recruited adolescents underwent head MRI examinations. Since most adolescents had not previously undergone head MRI examinations, they were asked to rest for at least 30 min between the clinical assessment and MRI examination. Additionally, they were required to arrive at the waiting area of the examination site 30 min in advance to adapt to the environment, thereby reducing the impact of the first examination. During the waiting period, they were instructed to avoid any activities that could lead to excessive physical or mental exertion.

MRI images were acquired using the MAGNETOM Prisma 3.0 T MRI scanner (Siemens Healthcare, Erlangen, Germany) and a 64-channel head coil at the medical imaging center of our hospital. Appropriate-sized foam pads were used to fix the head, reduce unconscious head movement, and comfortable earplugs were worn to reduce equipment noise, thereby minimizing the interference of noise on the examination results. The scanning range was from the top of the skull to the base of the skull. Conventional sequence scanning was performed first, including axial T1WI (TR 25 ms, TE 2.4 ms, FOV 230 mm×230 mm, number of slices 8, slice thickness 6 mm, slice gap 0.5 mm, matrix 256 × 256, excitation number 1), T2WI (TR 500 ms, TE 96 ms, FOV 230 mm×230 mm, number of slices 18, slice thickness 6 mm, slice gap 1.2 mm, matrix 230 × 230, excitation number 1), and fluid-attenuated inversion recovery sequence (TR 7500 ms, TE 81 ms, FOV 230 mm×230 mm, number of slices 18, slice thickness 6 mm, slice gap 1.2 mm, matrix 230 × 230, excitation number 1). After excluding intracranial organic lesions with conventional scanning, rs-fMRI scanning was performed, during which subjects were required to clear their minds, keep their eyes open, and not engage in any thinking activities. Rs-fMRI data acquisition used the EPI sequence (TR 1500 ms, TE 30 ms, FOV 244 mm×244 mm, number of slices 72, slice gap 2 mm, matrix 112 × 112, flip angle 60°, and 260 time points collected).

Preprocessing of rs-fMRI data

The raw image data of each subject were quality-controlled by two experienced imaging professionals to ensure that there was no excessive head movement or abnormal signals in the images. The scanning data were then classified and organized for preprocessing. Preprocessing of all subjects’ rs-fMRI images was performed using the DPARSF software (http://rfmri.org/DPARSF) on the MATLAB 2022b platform. The preprocessing included the following steps: (1) converting DICOM format images to NIFTI format; (2) deletion of the first 10 time-point images, in order to avoid the effects of subjects’ discomfort with the environment at the beginning of the examination and the initial instability of the magnetic field; and (3) slice timing correction. Correcting the BOLD images collected within the same repetition time to the same time point; (4) head movement correction. Data with head rotation exceeding 2° or any direction exceeding 2 mm of maximum displacement were excluded [23]; (5) using the DARTEL method, registering individual high-resolution T1 anatomical images to fMRI data, then obtaining aligned T1-weighted images, and converting them to the standard Montreal Neurological Institute space to reduce the impact of individual head shape differences; (6) Gaussian smoothing with a full width at half maximum of 6 mm×6 mm×6 mm to improve statistical power and reduce the impact of affine transformation; (7) removing linear trends to reduce signal drift caused by machine heating or pulse; (8) regression of nuisance factors included head-motion effect (Friston-24 model), white matter, cerebrospinal fluid and global signals. 1 HC was excluded for excessive head motion (translational movements exceeding 2 mm or rotation exceeding 2° in any direction during scanning). The average frame-wise displacement (FD) across time points of each participant was computed [24].

ALFF analysis

After preprocessing the data, DPABI software was used for ALFF analysis to calculate ALFF values. The time signal was transformed into a frequency signal, and a band-pass filter was applied to extract signals in the 0.01 ~ 0.08 Hz range to reduce the impact of low-frequency drift and high-frequency noise. The filtered data were subjected to Fourier transformation, and the power spectrum of the transformed signal was calculated, and the square root of the power at each voxel frequency was extracted from the power spectrum. The whole-brain average ALFF value was normalized to obtain z-ALFF values. These data were used as the final ALFF values for statistical analysis.

Seed region selection

After the ALFF analysis, regions with statistical differences in brain activity were considered ROIs. Seed regions were set in the ROIs to further investigate changes in brain functional networks. Spheres with a radius of 6 mm were drawn centered on the peak voxel coordinates of each positive cluster, which were considered seed regions for rs-FC analysis.

ALFF-based whole-brain rs-FC analysis

DPABI software was used again to study changes in brain functional networks through ALFF-based whole-brain FC analysis. First, a band-pass filter was applied after preprocessing to extract signals in the 0.01 ~ 0.08 Hz range. Second, the Pearson correlation coefficient between the average time series of each ROI and every voxel in the whole brain was calculated, and the value was the FC strength. Fisher Z transformation was used to improve the normality of the correlation coefficients, and further statistical analysis was performed.It should be noted that the seed regions used in the FC analysis of this study were defined based on regions showing inter-group differences in ALFF. This non-independent selection approach carries the risk of overstating the observed FC associations.

Statistical analysis

To explore the differences in demographic and clinical characteristics between the NSSI and HC groups, we used SPSS 27.0 software for two-sample t-tests and chi-square tests to analyze continuous and categorical variables, respectively. The DPABI tool package was used for statistical analysis of ALFF and FC. We used two-sample t-tests to assess the differences in ALFF between the NSSI and HC groups. With gender, age, and years of education as covariates, we evaluated the significantly different ALFF values between the two groups, with the significance level set at P < 0.05. To correct for multiple comparisons, we used a family-wise error correction (FWE correction), with regions set to be statistically different at the voxel level of P < 0.001 and at the cluster level of P < 0.05. We also used a two-sample t-test to assess differences in FC between the NSSI and HC groups. Gender, age, and years of education were used as covariates to assess significantly different FC values between the two groups, and the significance level was set at P < 0.05. We also used the FWE correction for the two-sample t-test for FC. Finally, after extracting the z-ALFF and z-FC with significant differences between the two groups, we performed partial correlation analysis to study the relationship between these abnormal regions and NSSI behavior clinical scales, controlling for the influence of anxiety and depression status, with P < 0.05 considered statistically significant.Cohen’s d effect sizes were calculated using IBM SPSS Statistics version 27.0 software.

Power analysis​​

A priori power analysis was conducted using G*Power 3.1 [25]. Based on prior studies reporting ALFF/FC differences between clinical populations and healthy controls, we assumed a medium-to-large effect size (Cohen’s d = 0.6) [26, 27]. With α = 0.05 (two-tailed) and power = 0.8, the required sample size was 26 per group. Our final sample (NSSI group: n = 33; controls: n = 29) exceeds this requirement.

Results

Demographic data and clinical characteristics

The NSSI group consisted of 33 adolescents (8 male and 25 female subjects) with an average age of 14.88 ± 1.673 years and an average age of onset of 12.15 ± 0.972 years. The HC group had 29 adolescents (9 male and 20 female subjects) with an average age of 15.10 ± 1.819 years. There were no significant differences between the two groups in terms of gender (P = 0.557), age (P = 0.616), and years of education (P = 0.815). Compared to the HC group, the NSSI group had higher scores on the GAD-7 and PHQ-9, with statistically significant differences between the two groups. Statistical data are shown in Table 1.

Table 1.

Demographic and clinical characteristics of adolescents with NSSI and healthy adolescents

	NSSI Group	HC Group	statistical value	P-value
Gender			χ2 = 0.348	0.557
Male (n,%)	8	9
Female (n,%)	25	20
Age(years)	14.88 ± 1.673	15.10 ± 1.819	t = 0.504	0.616
Education (years)	7.94 ± 1.600	8.17 ± 1.733	t = 0.235	0.815
Age at onset (years)	12.15 ± 0.972	-	-	-
GAD-7 score	15.64 ± 3.740	0.72 ± 1.137	t=−22.426
PHQ-9 score	18.94 ± 4.776	1.28 ± 2.031	t=−19.515
NSSI frequency
Past 1 month	1.42 ± 0.502	-	-	-
Past 6 month	2.06 ± 1.088	-	-	-
Past 1 year	1.82 ± 0.808	-	-	-
Function Ottawa Self-Injury scale	55.36 ± 27.537	-	-	-
Addictive Ottawa Self-Injury scale	12.21 ± 4.098	-	-	-

Abbreviations: NSSI non-suicidal self-injury, HC healthy control, GAD-7 generalised anxiety scale, PHQ-9 depression screening scale 

The chi-square test was used for gender comparisons

Two-sample t tests were used to compare age, years of education, and GAD-7 and PHQ-9 scores

ALFF Results

The brain regions with statistically significant differences in ALFF values between the NSSI and HC groups are shown in Table 2 and Figure 1. Compared to the HC group, the NSSI group had decreased ALFF values in the right hippocampus(HIP.R) and left hippocampus (HIP.L) and increased ALFF values in the right supplementary motor area (SMA.R). All differences were statistically significant (Table 2, Figure 1). The brain regions with differences in ALFF values between the two groups were used as ROIs for further rs-FC analysis. A sphere with a radius of 6 mm was defined, centered on the peak voxel coordinates of the ROIs, as the seed region for further analysis.

Table 2.

Significant differences in ALFF values of brain regions between the two groups brain regions

Cluster	Region	Side	Cluster size (voxels)	MNI coordinate			t-value	Cohen’s d	d_adj
				x	y	z
NSSI
Cluster 1 (ROI 1)	HIP	L	118	−21	−24	−3	−4.827	−0.996	−1.003
Cluster 2 (ROI 2)	HIP	R	80	24	−27	3	−4.557	−1.117	−1.139
NSSI > HC
Cluster 3 (ROI 3)	SMA	R	36	9	−12	60	3.967	0.830	0.860

Abbreviations: ALFF amplitude of low-frequency fluctuations, NSSI non-suicidal self-injury, HC healthy control, MNI Montreal Neurological Institute, R right, L left, HIP hippocampus, SMA supplementary motor area

Voxels indicate the size of local brain regions, MNI coordinates are the spatial 3D coordinates of local brain regions, t-values indicate the size of the difference, and the absolute maximum of t-values was selected as the location as the coordinates of the difference brain regions

d_adj: Covariate-adjusted Cohen’s d. FWE corrected, voxel level P < 0.001, cluster level P < 0.05

Fig. 1.

Significantly Different Brain Regions with Elevated or Decreased ALFF Values in the NSSI and HC Groups

ALFF-based rs-FC analysis

Based on the ALFF analysis, spherical ROIs with a radius of 6 mm centered on the peak voxel coordinates of HIP.L, HIP.R, and SMA.R were used as seed regions for ALFF-based rs-FC analysis (referred to as ROI1, ROI2, and ROI3, respectively). Compared to the HC group, HIP.L had reduced FC with the left precuneus (PCUN.L), right middle temporal gyrus (MTG.R), and right inferior temporal gyrus (ITG.R), and increased FC with the left thalamus (THA.L). HIP.R had reduced FC with MTG.R. SMA.R had reduced FC with MTG.R, right postcentral gyrus (PoCG.R), left dorsolateral prefrontal cortex (SFGdor.L), and right superior parietal gyrus (SPG.R) (Table 3).

Table 3.

Brain areas with differences in whole brain FC values between the two groups using ALFF difference brain areas as seed points

Cluster	Region	Side	Cluster size (voxels)	MNI coordinate			t-value	Cohen’s d	d_adj
				x	y	z
HIP.L
NSSI
Cluster 1	PCUN	L	166	−12	−51	42	−4.879	−1.087	−1.087
Cluster 2	MTG	R	37	57	−57	9	−4.197	−1.049	−1.081
Cluster 3	ITG	R	30	60	−3	−30	−4.251	−0.959	−1.006
NSSI > HC
Cluster 4	THA	L	35	−18	−24	0	8.469	0.212	0.220
HIP.R
NSSI
Cluster 1	MTG	R	87	54	−72	6	−3.898	−0.943	−0.966
SMA.R
NSSI
Cluster 1	MTG	R	66	54	−33	0	−3.735	−0.829	−0.824
Cluster 2	PoCG	R	78	63	−9	36	−3.989	−0.938	−0.978
Cluster 3	SFGdor	L	84	−9	45	48	−4.032	−0.901	−0.918
Cluster 4	SPG	R	93	24	−57	63	−4.154	−0.939	−0.937

Abbreviations: ALFF amplitude of low-frequency fluctuations, rs-FC static functional connectivity, NSSI non-suicidal self-injury, HC, healthy control, MNI Montreal Neurological Institute, ROI region of interest, R right, L left, PCUN precuneus, MTG middle temporal gyrus, ITG inferotemporal gyrus, THA thalamus, PoCG postcentral gyrus, SFGdor dorsal lateral superior frontal gyrus, SPG superior parietal gyrus

Voxels indicate the size of local brain regions, MNI coordinates are the spatial 3D coordinates of local brain regions, t-value indicates the size of the difference, and the absolute maximum of t-value is selected as the location as the coordinates of the difference brain regions

 d_adj: Covariate-adjusted Cohen’s d. FWE corrected, voxel level P < 0.001, cluster level P < 0.05

Correlation between the ALFF value and addictive features of NSSI

In the self-injury group, decreased ALFF values in the bilateral hippocampus were negatively correlated with score of addiction features of NSSI.(r = −0.376、−0.453, P = 0.031、0.008, Fig. 2.

Fig. 2.

2 Pearson's correlation between between the ALFF value and addictive features of NSSI. Abbreviations: ALFF low frequency fluctuation amplitude; R: right; L: left; NSSI:non-suicidal self-injury; HIP: hippocampus

Discussion

In the current study, we using ALFF and FC analysis methods to explore the differences in brain activity and functional connectivity between NSSI adolescent patients and healthy control groups, with a particular focus on the bilateral hippocampus region. By comparing the ALFF values and functional connectivity differences between NSSI adolescent patients and healthy control groups, we obtained several interesting findings. The study results showed that compared to the HC group, the ALFF values of the bilateral hippocampus in the NSSI group were significantly decreased, and this decrease was negatively correlated with the addictive characteristics of NSSI. In addition, the ALFF values of the SMA.R in the NSSI group were significantly increased. In terms of functional connectivity, the FC values between the HIP.R and the PCUN.L and MTG.R were significantly decreased. The FC values between HIP.L and the ITG.R were also significantly decreased. The FC values between the SMA.R and the MTG.R , between the SMA.R and the SPG.R were significantly decreased. These findings are of great significance for understanding the neurobiological mechanisms of NSSI behavior. Firstly, we observed that compared to the healthy control group, the NSSI group had weakened local neural activity in the bilateral hippocampus. The hippocampus, as part of the limbic system, plays an important role in regulating emotions, spatial memory, and endocrine regulation[28–32]. It serves as a critical structure for processing information about events and contextual elements, known as episodic memory, and is highly sensitive to stress, drugs, and aging[33]. First, the hippocampus modulates the activity of reward and punishment systems by regulating the release of neurotransmitters such as dopamine and serotonin[34]. This regulation is crucial for addictive behaviors. Additionally, the hippocampus is involved in enhancing drug-induced reward and reinforcement sensitivity. Hippocampal neurogenesis is closely associated with addictive behaviors, and some key addiction-related behaviors are promoted by deficits in hippocampal plasticity[35]. The hippocampus plays a pivotal role in memory formation and emotional regulation. Studies have shown that hippocampal hyperactivity may contribute to depression and nicotine addiction[32, 36]. Structural and functional abnormalities in the hippocampus may be linked to the development and maintenance of addictive behaviors. Drug exposure can suppress hippocampal neurogenesis, and insufficient neurogenesis is associated with impaired learning and memory functions, heightened stress responses, and the emergence of negative emotional states[37].

In this study, the decrease in ALFF values of the bilateral hippocampus may reflect obstacles in emotional processing and memory functions in NSSI adolescents[38]. Studies have also shown that reduced hippocampus activation is closely related to emotional dysregulation[39]. Additionally, studies have shown a connection between NSSI and abnormal activation of brain regions responsible for cognitive reward processing [40]. It is noteworthy that we also found a negative correlation between the ALFF values of the bilateral hippocampus and the total score of addictive characteristics in NSSI adolescents, suggesting that low activity in the bilateral hippocampus may be related to the addictive characteristics of NSSI behavior. Addictive characteristics, such as compulsivity, craving, and withdrawal symptoms, may play an important role in NSSI behavior, and abnormal activity in the hippocampus may provide a neurobiological basis for this.

Furthermore, we analyzed the functional connectivity between the bilateral hippocampus and other brain regions. The results showed that the functional connectivity between the HIP.L and the left precuneus, MTG.R, and right inferior temporal gyrus was significantly decreased. The precuneus is one of the key nodes of the default mode network and is closely related to functions such as self-awareness and episodic memory [41]. The middle temporal gyrus and inferior temporal gyrus, on the other hand, are involved in various aspects of emotion processing, hearing and language comprehension[42]. The decreased functional connectivity between these regions and the hippocampus may imply functional impairments in NSSI adolescents in terms of emotional regulation, memory consolidation, and language comprehension[13].

In addition, this study also found some interesting phenomena worth further exploration. For example, although some brain regions (such as the precuneus, middle temporal gyrus, and inferior temporal gyrus) showed significantly decreased functional connectivity with the hippocampus, no significant changes in ALFF values of these brain regions themselves were observed. This may mean that in NSSI behavior, the functional abnormalities of some brain regions may be more manifested as abnormalities in connections with other brain regions rather than as local activity abnormalities. When interpreting these findings, other possible factors also need to be considered. For example, NSSI behavior may be influenced by a variety of factors, including genetic, environmental, and psychological factors [43].

In summary, this study used ALFF analysis and functional connectivity analysis to explore the differences in brain activity between NSSI adolescent patients and healthy control groups. We found decreased ALFF values in the bilateral hippocampus, which were negatively correlated with the total score of addictive characteristics in NSSI adolescent patients. At the same time, the functional connectivity between the HIP.L and the left precuneus, MTG.R, and right inferior temporal gyrus was significantly decreased and negatively correlated with the total score of addictive characteristics. These findings provide an in-depth understanding of the neurobiological basis of NSSI behavior and potential targets for future intervention and treatment. However, these results still need further verification and deepening to reveal more details about the neurobiological basis of NSSI behavior. At the same time, future studies should more comprehensively explore the connection patterns between different brain regions and how these connection patterns are related to NSSI behavior.

This study has certain limitations, mainly in terms of sample representativeness and research depth. Specifically, this study is a cross-sectional study, and therefore, causal inferences cannot be made. In addition, although this study identified brain regions with functional differences, the underlying neural mechanisms behind these differences remain unclear, which also points out an important direction for our subsequent research. It should be noted that the seed regions were defined based on the differences in ALFF within the same sample, and this approach may introduce circularity. Therefore, it is necessary for future studies to use independent datasets or pre-defined anatomical ROIs to verify the current research findings.Furthermore, as mentioned in the Methods, the use of ALFF-derived ROIs for seed-based FC analysis introduces a risk of inflated associations due to non-independent selection.

Conclusions

This study used ALFF and FC analysis methods to explore the differences in brain function between NSSI adolescents and healthy adolescents, revealing not only the specific changes in brain function in NSSI adolescents but also preliminarily elucidating the intrinsic connection between the bilateral hippocampus and self-injury behavior and addictive symptoms.

Abbreviations

NSSI

Non-self-injurious self-injury

DSM-5

Diagnostic and Statistical Manual of Mental Disorders

OSI

Ottawa Self-Injury Inventory

ALFF

Amplitude of Low Frequency Fluctuations

ROIs

Regions of Interests

FC

Functional Connectivity

rs-FC

Resting-state Functional Connectivity

BOLD

Blood Oxygen-level Dependent

rs-fMRI

Resting-state Functional Magnetic Resonance Imaging

GAD-7

General Anxiety Disorder-7

PHQ-9

Patient Health Questionnaire

R

Right

L

Left

HIP

Hippocampus

SMA

Supplementary Motor Area

PCUN

Precuneus

MTG

Middle Temporal Gyrus

ITG

Inferotemporal Gyrus

THA

Thalamus

PoCG

Postcentral Gyrus

SFGdor

Dorsal Lateral Superior Frontal Gyrus

SPG

Superior Parietal Gyrus

Authors’ contributions

Xinbei Lin and Y.L. contributed the central idea. Xinbei Lin, and Y.S. analyzed most of the data. Xinbei Lin wrote the initial draft of the paper. Y.S., Y.L., Y.G., K.C.and Y.C. helped in revising the manuscript. YH, ZZ, ZL, XW, FY, and S.D. contributed to the data collection. T.L., Xue Lv, C.S., Z.G. and Y.L.contributed to the opinion refinement, supplementary analysis, and finalization of this paper. All authors reviewed the manuscript.

Funding

This study was supported by the Central Plains Science and Technology Innovation leading talent Program (244200510016); Henan Provincial Key R&D Program (251111312900, 251111312800); Henan Provincial Medical Science and Technology Research Program (LHGJ20240050); Henan Provincial Science and Technology Research Program (242102311018); Key Science and Technology Projects of Henan Province (241100310300).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Studies involving human participants were reviewed and approved by the Zhengzhou University Ethics Committee (No.2023 − 275). Written informed consent to participate in this study was provided by the participants’ legal guardians/next of kin. Written informed consent was obtained from the legal guardian/next of kin of the minors for the release of any potentially identifiable images or data contained herein.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.