Prediction of non-suicidal self-injury in adolescents at the family level using regression methods and machine learning

The mediating roles of anxiety, loneliness, stress, and depression in the relationship between cyberbullying and non-suicidal self-injury: propensity score matching and causal mediation analysis

BACKGROUND

Non-suicidal self-injury (NSSI) not only causes direct physical harm but also severely impacts mental health, increasing the risks of psychological disorders, behavioral issues, and suicidal tendencies. With the widespread use of the internet and electronic communication tools on smartphones, such as social media and instant messaging applications, cyberbullying has become increasingly prevalent. However, research on the relationship between cyberbullying and NSSI remains relatively limited at present.

METHODS

This research was conducted in Nantong City, Jiangsu Province, from July to September 2024, utilizing a convenience sampling method with 1,751 participants. Propensity Score Matching (PSM) was employed to control for confounding factors, while causal mediation analysis was used to assess the mediating effects of anxiety, depression, stress, and loneliness between cyberbullying and NSSI.

RESULTS

In this study, 557 participants (31.8%) reported having experienced cyberbullying, Cyberbullying significantly increases NSSI odds through direct and indirect effects, mediated by anxiety (ACME = 0.08, ADE = 0.22, mediated proportion = 0.28), loneliness (ACME = 0.04, ADE = 0.26, mediated proportion = 0.14), stress (ACME = 0.05, ADE = 0.24, mediated proportion = 0.18), and depression (ACME = 0.06, ADE = 0.24, mediated proportion = 0.19).

CONCLUSION

This study explores the relationship between cyberbullying, mental health issues (depression, anxiety, stress, loneliness), and NSSI. Findings show cyberbullying directly and indirectly increases NSSI risk, with loneliness as a key mediator. Despite insights, the study has limitations, suggesting a need for more rigorous designs and comprehensive data.

Family Functioning and NSSI Urges among Chinese Adolescents: A Three-wave Chain Multiple Mediation Model

Family functioning is used to evaluate the functioning of a family system, which exerts a huge influence on non-suicidal self-injury (NSSI). However, little is known about how family functioning relates to NSSI among adolescents, and existing research mainly depends on cross-sectional design, which cannot capture dynamic changes between variables. Additionally, NSSI urge has been established as a robust predictor of NSSI engagement and is clinically important. Guided by the integrated model of NSSI, the current study investigated two potential mediators (emotion regulation and self-esteem) of the relations between family functioning and NSSI urges to address these gaps. Participants were 1298 Chinese adolescents (50.5% male; Mage at Wave 2 = 15.08 years) from a three-wave longitudinal study with data spanning ten months. The results supported the chain multiple mediation model. Specifically, valid family functioning prevented NSSI urges through three indirect paths: (a) higher level of self-esteem; (b) more cognitive reappraisal to higher self-esteem; (c) less expressive suppression to higher self-esteem. The prediction of family functioning on cognitive reappraisal was weaker among girls, while the prediction of self-esteem on NSSI urges was stronger. These findings indicate that high family functioning is a powerful protective factor of NSSI urges, and NSSI interventions should aim to improve adolescents' family functioning, with attention to reinforcing emotion regulation and self-esteem, especially for girls.

The association of family functioning and suicide in children and adolescents: positive behavior recognition and non-suicidal self-injury as sequential mediators

Background

Suicide is a major behavioral issue among children and adolescents, and family functioning potentially influencing its occurrence. Furthermore, positive behavior recognition, as a key component of positive youth development, may act as a mediating factor in the relationship between family functioning and suicide. Non-suicidal self-injury (NSSI) often co-occurs with suicide and may also mediate the impact of family functioning on suicide. Therefore, the aim of this study is to examine the role of family functioning in child and adolescent suicide, with positive behavior recognition and NSSI serving as sequential mediating factors.

Methods

The participants in this study were recruited from the Chengdu Positive Child Development (CPCD) cohort study. The analysis was based on the second round of cross-sectional data from the CPCD cohort. They were required to complete questionnaires that included measures of family functioning, suicide, positive behavior recognition, and NSSI. SPSS 26.0 and Mplus 8.3 were used for descriptive statistical analysis, correlation analysis and mediation effect analysis.

Results

A total of 8,124 participants completed the questionnaires, with an average age of 11.00 ± 2.18 years. The sample comprised 4,195 male and 3,929 female participants. The findings indicate that 17.93% of children and adolescents reported suicide thoughts, 10.06% had formulated suicide plans, and 8.64% had attempted suicide. Poor family functioning shows a significant positive correlation with suicide (r = 0.322, p < 0.01). The multiple mediation effect of positive behavior recognition and NSSI in the association between family functioning and suicide was statistically significant (β = 0.034, 95% CI = 0.027, 0.042).

Conclusion

This study found that poor family functioning is a risk factor for suicide in children and adolescents, with positive behavior recognition and non-suicidal self-injury acting as sequential mediating factors.

Longitudinal Relationship between Aggressive Behavior and Non-Suicidal Self-Injury among Adolescent Boys and Girls: The Mediating Role of Peer Victimization

Although a large body of research has found associations between aggressive behavior and non-suicidal self-injury (NSSI), evidence for the directionality, underlying mechanisms, and potential gender differences in their associations remain unclear. To address the gaps, this study investigated the bidirectional relationship between aggressive behavior and NSSI, the mediating role of peer victimization (physical and relational victimization), and gender differences in these associations among a sample of Chinese adolescents. Using a longitudinal design, a total of 1394 Chinese adolescents (Mage = 13.39, SD = 0.59, 43.3% girls) completed surveys across three waves, with intervals of nine and six months, respectively. The results revealed that within the total sample, T1 aggressive behavior positively predicted T2 physical victimization, and T2 physical victimization positively predicted T3 NSSI. T1 NSSI positively predicted T2 relational victimization, and T2 relational victimization positively predicted T3 aggressive behavior. However, the indirect effect of T1 aggressive behavior on T3 NSSI mediated by T2 physical victimization was significant only for boys. The indirect effect of T1 NSSI on T3 aggressive behavior mediated by T2 relational victimization was significant only for girls. These findings highlighted the importance of considering the gender-specific process underlying the relationship between aggressive behavior and NSSI, thus guiding the development of gender-informed prevention and intervention strategies.

Non-suicidal self-injury in inpatient and outpatient adolescents: disentangling psychopathology and interactive family dynamics

Background

Non-suicidal self-injury (NSSI) is defined as a transdiagnostic phenomenon that has well increased in the latest years, especially in the adolescent population. It has been associated with suicidality, alexithymia, emotion dysregulation, and psychosocial impairment, as well as family issues. The choice of level of care (i.e., hospitalization versus outpatient visit) depends on a number of factors that relate not only to suicidal risk but also to severity of individual's psychosocial functioning, the ability of family environment to support treatment choices and to contain child, as well as the need for ongoing monitoring of the young patient. A scarcity of studies has compared outpatients with inpatients, both of them engaging in NSSI.

Methods

the current study aimed to further expand knowledge regarding features that characterize young self-harmers who receive different levels of care, with particular attention on psychopathological, family, and NSSI-related characteristics, as well as suicidality. The current research included 56 inpatients and 56 outpatients with NSSI, paired for gender, age, and psychiatric diagnosis. Instruments investigating psychopathology, emotion dysregulation, alexithymia, psychosocial functioning, and interactive family dynamics were administered. Descriptive statistics, parametric and non-parametric inferential statistics were applied.

Results

study findings highlighted that inpatients engaging in NSSI reported lifetime suicidality, clinical level of externalizing and internalizing problems, more severe alexithymia, emotion dysregulation, and impaired psychosocial functioning compared to outpatients engaging in self-harming. Furthermore, when compared to outpatients' families, inpatients' families were more capable of adhering to rules and time of the family play situation (the Lausanne Trilogue Play procedure) and fixing interactive mistakes through activities. On the contrary, in the inpatient group, global performance, role implication, parental scaffolding, child's involvement and self regulation tend to decline, while parental conflicts tend to rise over the four part scenario of the family play.

Conclusion

these findings confirmed a more severe global picture of young inpatients engaging in self-harming, suggesting that NSSI may be the expression of this larger psychopathological picture. In addition, the study highlighted the need for a multi-informant and multimethod clinical assessment, which should include evaluation of family context and co-parenting system, especially for hospitalized young patients engaging in self-harm.

Family Economic Hardship and Non-Suicidal Self-Injury Among Chinese Adolescents: Relative Deprivation as a Mediator and Self-Esteem as a Moderator

The literature provides empirical evidence that family economic hardship can increase the likelihood of adolescents engaging in non-suicidal self-injury (NSSI). However, the mechanisms underlying this relationship remain unclear. Guided by the risky families model, this study aimed to investigate whether relative deprivation mediates the link between family economic hardship and adolescent NSSI, and whether self-esteem moderates this indirect link. A combined 673 adolescents (45.9% female; Mage = 12.81 years) completed measures of family economic hardship, relative deprivation, self-esteem, and NSSI. The results verified that relative deprivation mediated the positive link between family economic hardship and NSSI. Additionally, self-esteem moderated the effect of family economic hardship on relative deprivation. Adolescents possessing high self-esteem reported lower relative deprivation compared to those possessing low self-esteem, regardless of their level of family economic hardship. Notably, the mitigating effect of high self-esteem diminished with high family economic hardship. These findings provide deeper insights into the mechanisms by which family economic hardship affect adolescent NSSI, and have practical implications for prevention and intervention strategies targeting this behavior.

Effectiveness of individualized rTMS under sMRI guidance in reducing depressive symptoms and suicidal ideation in adolescents with depressive disorders: an open-label study

Background

Major Depressive Disorder (MDD) is occurring at a progressively younger age, and suicide is now the second leading cause of death among adolescents with MDD. Studies have shown that structural magnetic resonance imaging (sMRI) can improve the positioning accuracy and anti-depressant effects of repetitive transcranial magnetic stimulation (rTMS), thereby reducing suicidal ideation.

Objective

To compare the efficacy of sMRI-guided rTMS combined with pharmacotherapy, surface 5-cm rTMS positioning combined with pharmacotherapy, and pharmacotherapy alone on reducing depressive symptoms and suicidal ideation (SI) in MDD adolescents.

Methods

This was an open-label study of adjustable-dose pharmacotherapy combined with rTMS for the treatment of depressive symptoms and suicidal ideation in MDD adolescents. The three study groups were as follows: sMRI navigation for individualized rTMS coordinates targeting the dorsolateral prefrontal cortex (DLPFC) and in combination with pharmacotherapy for 10 rTMS sessions over two weeks; surface 5-cm positioning for DLPFC in combination with pharmacotherapy for 10 rTMS sessions over two weeks; pharmacotherapy. All patients received only one type of SSRIs anti-depressant. A total of 123 Chinese adolescents aged 13-18 with MDD were enrolled, and psychological parameters were evaluated in the first and second weeks of treatment.

Results

Following treatment, the clinical symptoms improved in all three groups. The sMRI navigation group exhibited significantly more improvement in depressive symptoms and suicidal ideation, without severe adverse reactions.

Conclusion

Ten sessions of rTMS treatment are feasible and effective in improving depressive symptoms and reducing SI in MDD adolescents. The combination of sMRI navigation rTMS and pharmacotherapy was found to yield the best outcomes.

Clinical trial registration

https://www.medicalresearch.org.cn/index, identifier MR-33-24-030536.

Predicting the Risk of Loneliness in Children and Adolescents: A Machine Learning Study

BACKGROUND

Loneliness is increasingly emerging as a significant public health problem in children and adolescents. Predicting loneliness and finding its risk factors in children and adolescents is lacking and necessary, and would greatly help determine intervention actions.

OBJECTIVE

This study aimed to find appropriate machine learning techniques to predict loneliness and its associated risk factors among schoolchildren.

METHODS

The data were collected from an ongoing prospective puberty cohort that was established in Chongqing, Southwest China. This study used 822 subjects (46.84% boys, age range: 11-16) followed in 2019. Five models, (a) random forest, (b) extreme gradient boosting (XGBoost), (c) logistic regression, (d) neural network, and (e) support vector machine were applied to predict loneliness. A total of 39 indicators were collected and 28 predictors were finally included for prediction after data pre-processing, including demographic, parental relationship, mental health, pubertal development, behaviors, and environmental factors. Model performance was determined by accuracy and AUC. Additionally, random forest and XGBoost were applied to identify the important factors. The XGBoost algorithm with SHAP was also used to interpret the results of our ML model.

RESULTS

All machine learning performed with favorable accuracy. Compared to random forest (AUC: 0.87 (95%CI: 0.80, 0.93)), logistic regression (AUC: 0.80 (95%CI: 0.70, 0.89)), neural network (AUC: 0.80 (95%CI: 0.71, 0.89)), and support vector machine (AUC: 0.79 (95%CI: 0.79, 0.89)), XGBoost algorithm had the highest AUC values 0.87 (95%CI: 0.80, 0.93) in the test set, although the difference was not significant between models. Peer communication, index of general affect, peer alienation, and internet addiction were the top four significant factors of loneliness in children and adolescents.

CONCLUSIONS

The results of this study suggest that machine learning has considerable potential to predict loneliness in children. This may be valuable for the early identification and intervention of loneliness.

Predicting non-suicidal self-injury among Chinese adolescents: The application of ten algorithms of machine learning

Background and aims

High non-suicidal self-injury (NSSI) prevalence among adolescents is a global health issue. However, current prediction models for adolescent NSSI rely on a limited set of algorithms, resulting in biased predictions. Therefore, the aim of this study is to develop multiple machine learning models to enhance prediction accuracy and mitigate biases among Chinese adolescents.

Methods

A total of 4487 junior and senior high school students in China were recruited. Multiple algorithms were included, such as logistic regression, decision tree, support vector machine, Naive Bayes, multi-layer perceptron, K-nearest neighbors, and ensemble learning algorithm like random forest, bagging, AdaBoost, and stacking to build predictive models. Data processing techniques, including standardization and the synthetic minority oversampling technique, were employed to optimize the predictive model. The model was trained on 70 % of the data, reserving 30 % for testing.

Results

The ten prediction models achieved a good performance, with area under the receiver operating characteristic curve (AUC) scores above 0.700 in the test set. The stacking and random forest models achieved AUC scores of 0.904 and 0.898, respectively. The prediction performance of the Naive Bayes model was relatively poor. The top five important variables were resilience, bully, suicidal ideation, internet addiction, and depression.

Conclusions

The ensemble machine learning algorithm showed promising results predicting NSSI among adolescents. Such algorithms should be recommended for future NSSI research to enhance predictive accuracy. Identification of important features in NSSI prediction can help develop screening protocols and lay a foundation for clinical diagnosis and intervention in adolescent populations.

Psychological pain and sociodemographic factors classified suicide attempt and non-suicidal self-injury in adolescents

This study aimed to utilize machine learning to explore the psychological similarities and differences between suicide attempt (SA) and non-suicidal self-injury (NSSI), with a particular focus on the role of psychological pain. A total of 2385 middle school students were recruited using cluster sampling. The random forest algorithm was used with 25 predictors to develop classification models of SA and NSSI, respectively, and to estimate the importance scores of each predictor. Based on these scores and related theories, shared risk factors (control feature set) and distinct risk factors (distinction feature set) were selected and tested to distinguish between NSSI and SA. The machine learning algorithm exhibited fair to good performance in classifying SA history [Area Under Curves (AUCs): 0.65-0.87] and poor performance in classifying NSSI history (AUC: 0.61-0.68). The distinction feature set comprised pain avoidance, family togetherness, and deviant peer affiliation, while the control feature set included pain arousal, painful feelings, and crisis events. The distinction feature set slightly but stably outperformed the control feature set in classifying SA from NSSI. The three-dimensional psychological pain model, especially pain avoidance, might play a dominant role in understanding the similarities and differences between SA and NSSI.