新北市自閉症與非自閉症青少年之自殺風險因子比較分析

Suicide Risk Factors in Adolescents with and Those without Autism Spectrum Disorder: Comparative Analysis in New Taipei City

衛福部2020年「青少年自殺防治策略及作為等情案」調查報告顯示，青少年族群自殺死亡率連續20年攀升，15至24歲人口群自殺通報更是逐年增加，青少年十大死因中，自殺排名占第二位，僅次於意外，顯示青少年自殺問題、自傷議題的嚴重性。研究指出，身心障礙青少年自殺風險高於一般青少年，其可能原因為自我形象的改變、情緒困擾及壓力、家庭內部衝突、學習困擾及同儕關係等及社會適應等多重因素交互作用下之結果。目前對於身心障礙青少年之自殺相關研究，以零星個案報告為主，對於此族群自殺相關之風險因子，尚未有系統化研究。本研究為回溯性研究，結合新北市自殺防治通報系統與新北市特殊教育通報網之資料庫，收集2019至2022年間，年齡介於9歲至21歲且具有身心障礙身分，或取得特殊教育鑑定之自閉症青少年共92位為ASD組，並隨機抽樣具自殺行為之一般青少年92位作為非ASD組。主要目的在了解兩組之間在自殺發生率、人口背景以及自殺風險因子間是否存在差異性。研究採描述性統計分析，並以卡方檢定與邏輯迴歸模型檢驗中介變項之顯著性與關聯強度。結果指出自閉症青少年有較高的自殺發生率，其中自閉症青少女自殺發生率較男性高，此外，同儕關係衝突及童年逆境此兩項風險因子，在兩組間具顯著差異。本研究作為自閉症青少年之自殺相關風險因子之前導研究，期能持續深入發展，獲取更多的研究數據與證據，進而為第一線自殺防治工作政策擬定提供相關建議。

Rationale & Purpose: The continuous rise in suicide mortality in Taiwan over the past 20 years underscores the severity of adolescent suicide and self-injury concerns. Suicide ranks as the second leading cause of death for Taiwanese adolescents. Given the current global emphasis on mental health, suicide prevention is a key component of Sustainable Development Goal initiatives, which emphasize the need for targeted interventions in vulnerable adolescent populations. Suicide risk is considerably higher in adolescents with disabilities compared with in the general population due to the complex interactions between changes in self-image, physical disabilities, emotional distress, familial conflict, learning difficulties, and challenging peer relationships. A comparatively high suicide rate has been noted among adolescents with autism spectrum disorder (ASD), which is considered an independent risk factor for suicidal behavior. The estimated prevalence of ASD in Taiwan is currently between 1.0% and 1.2%, with a male-to-female ratio of approximately 3:1. Given the relative lack of systematic research on suicide-related risk factors in this vulnerable group, this retrospective study was conducted to investigate suicide rates, demographic characteristics, suicide risk factors, and their interrelationships in adolescents with versus those without ASD in New Taipei City. Evidence-based suggestions for the formulation of frontline suicide prevention policies are provided. Methods: In this retrospective study, data were collected from the New Taipei City Suicide Prevention Notification System and the New Taipei City Special Education Notification Network. The sample comprised 92 adolescents aged 9–21 years with ASD who reported suicidal behavior and were identified through special education systems between 2019 and 2022. A control group of 92 adolescents without ASD who exhibited suicidal behavior was selected through random sampling to match the size, age range, and gender distribution of the experimental group. Regarding the ASD group, individuals for whom repetitive self-stimulatory behavior rather than intentional suicide was noted and those with severe intellectual disability for whom suicidal intent could not be reliably confirmed were excluded. Regarding the control group, adolescents with records or registration indicating a need for special education were excluded. Data from suicide follow-up interview records were anonymized, coded in a binary format, analyzed statistically, and handled confidentially. Interrater reliability analysis were conducted by two experienced mental health practitioners to confirm high consistency, with the overall Cohen’s kappa being 0.95, demonstrating strong agreement. Descriptive statistics, chi-square tests, and logistic regression models were employed to examine the significance and strength of the association between suicidal behavior and eight mediating risk variables (i.e., gender, family background, family history of psychiatric disorders, substance abuse, comorbid psychiatric disorders, peer conflict, academic stress, and adverse childhood experiences [ACEs]) in the two groups. Results/Findings: The suicide rate was markedly higher among adolescents with ASD (1.86%) compared with in the general adolescent population of New Taipei City (average 0.33%). The rate was even higher among female adolescents with ASD (2.87%), being 1.54 times higher than that among male adolescents with ASD (1.47%). Thus, female sex may represent an additional independent risk factor for suicide within the ASD population. Possible explanations for this difference include delayed diagnosis, a greater likelihood of camouflaging or masking social difficulties, and higher vulnerability to social isolation. The age at first suicidal attempt did not differ significantly between the ASD (mean: 13.96 years) and control (mean: 14.42 years) groups. However, the range of ages at first suicidal attempt in the ASD group (9 to 20 years) was wider than that in the control group, suggesting considerable heterogeneity. Two risk factors differed significantly between the groups: peer conflict and ACEs. The rate of peer conflict was significantly higher in the ASD group than in the control group, particularly among male adolescents with versus without ASD. Conversely, a higher proportion of male adolescents without than with ASD reported experiencing ACEs. Factor interaction analysis revealed that comorbid psychiatric disorders amplified the risk of peer conflict in the ASD group, which may be explained by difficulties in interpersonal interaction skill and impaired impulse control. In the control group, a family history of mental illness increased the impact of ACEs, which may be attributable to disruptions in family interaction patterns, family functioning and increased individual psychological vulnerability. These findings are consistent with both the interpersonal theory of suicide (IPTS) and diathesis–stress model. As proposed by the IPTS, social difficulties may cause individuals with ASD to experience“thwarted belongingness”and feel burdensome, and these two feelings constitute key core factors contributing to suicidal behavior. Comorbid conditions represent an internal vulnerability (diathesis), which, when combined with external pressures (stress) such as peer conflict, elevates overall suicide risk. Conclusions/Implications: This study highlights the urgent need to address suicide risk among adolescents with ASD, particularly female adolescents. Peer conflict and ACEs were identified as group-specific risk factors in the ASD and control groups, respectively, indicating that tailored, population-specific prevention strategies are necessary. Given these findings, ASD-specific suicide risk assessment tools should be established that incorporate core clinical characteristics, including social-motivation deficits and the“masking”or camouflaging behaviors frequently observed in female adolescents with ASD, which may contribute to under-recognized distress and missed detection. Additionally, procedures for suicide notification and interview documentation should be standardized to increase data quality, ensure situational consistency, and reduce variability across evaluators. Improved cross-system collaboration among families, schools, and medical services—supported by targeted training for front-line professionals regarding suicide risk in adolescents with ASD—is essential for early detection and timely intervention. Specific prevention programs should also be developed for high-risk female adolescents with ASD that focus on addressing relationship-based bullying, fostering emotional regulation skills, and mitigating the risks associated with social difficulties. Future longitudinal studies with larger samples are warranted to explore the long-term effects of gender differences, family functioning, and psychiatric comorbidities on suicide risk in adolescents with ASD. Continued research is crucial to improve the identification and assessment of suicide risk factors, refine evidence-based intervention strategies, and establish safer, more supportive environments that promote the psychological resilience of adolescents with ASD.