學童自殺意念之變化趨勢及相關因素

Trends and Correlates of Suicide Ideation among School Children

江宜珍(Yi-Chen Chiang)

指導教授 ： 李蘭

國立臺灣大學/公共衛生學院/衛生政策與管理研究所/博士(2007年)

https://doi.org/10.6342/NTU.2007.02120

摘要

本研究目的在瞭解學童自殺意念於2001∼2005年之變化趨勢及相關因素。利用兒童與青少年行為之長期發展研究（簡稱CABLE計畫）於2001年開始追蹤調查的國小四年級世代樣本的資料，進行次級資料分析。五年均有參與調查且未漏答自殺意念題目之學童為本研究樣本，共1,532人。採用趨勢分析、比率Z檢定、多項logit模型、存活分析、多層級比例勝算模式等方式完成統計分析。
結果顯示，2001∼2005年學童自殺意念之終生盛行率由18.86%上升至51.17%，且具有顯著的線性趨勢。學童最近一個月自殺意念盛行率在五個年度分別為8.29%，6.91%，6.20%，17.95%及17.37%，顯示在國中階段有躍升的現象。2002∼2005年學童自殺意念之發生率分別為10.70%，10.81%，13.84%及12.31%。整體而言，學童的自殺意念不論是終生盛行率、最近一個月盛行率或發生率，不僅隨年度上升，且在性別和地區上呈現顯著差異。
2001∼2005年的五年間，學童自殺意念之變化可分為以下六種類型：（1）五年均無自殺意念；（2）五年內曾有自殺意念但均屬輕微程度；（3）從輕微程度變成嚴重程度；（4）呈現起起伏伏的現象；（5）從嚴重程度變成輕微程度；（6）均屬嚴重程度，所佔比率分別為48.83%，40.67%，2.35%，5.22%，2.87%及0.07%。
學童首次出現自殺意念的時間，在其就讀國小四、五、六年級時依序有18.86%，8.68%及7.83%；在就讀國中一、二年級時有8.94%和6.85%。以存活分析模式探討與首次出現自殺意念有關的因素，發現居住在台北市、憂鬱程度越高或變嚴重、社交孤寂程度越高或變嚴重、較具情感變異性人格、物質使用情形越多或變嚴重、母親教育程度越高、與家人互動程度越低或變少、自覺受到父母懲罰程度越高或變嚴重之學童，其較早出現自殺意念的可能性較高。
以多層級比例勝算模式探討學童自殺意念之嚴重度變化發現，無論學童於國小四年級自殺意念之嚴重度為何，均會隨時間惡化。此外，也發現憂鬱程度越高或變嚴重、社交孤寂程度越高或變嚴重、暴力傾向越多或變嚴重、物質使用情形越多或變嚴重、與家人互動程度越低或變少、自覺父母支持程度越低或變少、自覺父母懲罰程度越高或變嚴重、父母衝突程度越高或變嚴重之學童，其自殺意念變嚴重的可能性越高。加入社會環境因素之變動性指標後發現，若學童居住在離婚率或失業率較高或變高的地區，其自殺意念變嚴重的可能性越高。
根據本研究結果，提出以下建議：1.正視兒童有自殺意念且會隨著成長而增加或變嚴重的問題，及早採取預防和輔導措施。2.家長應多花時間與孩童互動並給予支持，少用懲罰及避免配偶間的衝突。3.學童之心理（憂鬱、社交孤寂）及行為（如物質使用和暴力傾向增多）問題，可作為預測學童自殺意念的指標。4.國中教師應注意新入學之一年級學生對新環境的調適情形，適時給予關懷與協助。

關鍵字

自殺意念 ； 兒童 ； 長期追蹤 ； 趨勢分析 ； 存活分析

並列摘要

The purpose of this study was to investigate the trends and correlates of suicide ideation among school children over a five year period. Data from the CABLE study (Child and Adolescent Behaviors in Long-term Evolution) from 2001 (grade 4) to 2005 (grade 8) were used for secondary data analysis. Our study sample included the 1,532 students followed for five years who had completed the question on suicide ideation each year. We used trend analysis, proportional Z-test, multi-logit model, survival analysis, and multilevel analysis to analyze our data.
We found that the lifetime prevalence rates of suicide ideation among children from 2001 to 2005 significantly increased from 18.86% to 51.17%. The one-month prevalence rates were 8.29%, 6.91%, 6.20%, 17.95% and 17.37%, for the years 2001 to 2005 respectively. The period prevalence rate almost tripled in grade 7 (2004). The incidence rates of suicide ideation for the years 2002 to 2005 were 10.70%, 10.81%, 13.84% and 12.31%, respectively. The lifetime and one-month prevalence rates and the incidence rates during this 5 year period all showed significant urban-rural disparity and gender differences.
There were 6 patterns of change in children’s suicide ideation: (1) never had suicide ideation during the 5 years (48.83%); (2) ever had suicide ideation during the 5 years that remained mild in degree (40.67%); (3) ever had suicide ideation that worsened in degree over time (2.35%); (4) ever had suicide ideation with undulating severity over time (5.22%); (5) ever had suicide ideation that lessened in degree over time (2.87%); (6) ever had suicide ideation that remained severe (0.07%).
The proportion of students reporting their first experience of suicide ideation from grade 4 to grade 6 were 18.86%, 8.68% and 7.83%, respectively. In grade 7 and grade 8, 8.94% and 6.85% of students reported thinking about suicide for the first time respectively. From the extended Cox model, risk factors for the earlier development of suicide ideation were “living in Taipei City,” “having more depressive symptoms or depressive symptoms increasing in severity,” “having a higher level of social loneliness or having a worsening level,” “having a more emotionally labile personality,” “heavy substance abuse or worsening substance abuse,” “having a mother with higher education,” “less family interaction or a worsening level” and “punished more by parents or a worsening level of punishment”.
Multilevel analysis showed that regardless of the initial degree of severity of suicide ideation among children in grade 4, suicide ideation worsened over time. Depressive symptoms, social loneliness, violent tendency, substance abuse, low family interaction, low parental support and high parental punishment were all risk factors for suicide ideation among children. After controlling for various social factors, it was found that children whose suicide ideation became more severe were more likely to live in an area with higher divorce or unemployment rates.
These findings underline the seriousness of suicide ideation among children and the importance of facing this issue. Particularly, as we know that the prevalence and incidence rates will get higher with age, prevention strategies and support at younger ages are necessary. As lack of family interaction and increased punishment are both risk factors for suicide ideation, parents need to be encouraged to spend more time with their children and to avoid excessive punishment. Parents and teachers should be made aware of characteristics of at risk children such as social isolation, depressive symptoms, substance abuse and violent behavior. In addition, as the period prevalence of suicide ideation seems to markedly increase in grade 7, the first year of junior high school, teachers at these schools should be made aware of this phenomenon and be encouraged to assist grade 7 students in adapting to their new school environment.

並列關鍵字

suicide ideation ； children ； longitudinal data ； trend analysis ； survival analysis

參考文獻

黃鈞蔚、江弘基、李明濱、戴傳文（2006）。青少年的自殺防治。台灣醫學，10(3)，343-352。

蘇宗偉、李明濱（2006）。自殺的流行病學。台灣醫學，10(3)，329-338。

陳建仁（1999）。流行病學：原理與方法。台北市：聯經。

江宜珍（2005）。台灣北部國小四年級學童曾經發生自殺意念之相關因素研究。台灣衛誌，24(6)，471-482。

李秀珠、彭玉賢、蔡佳如（2002）。新傳播科技對台灣新聞媒體之影響—從新聞內容之區位談起。新聞學研究，72，27-54。