青春期全程的生活事件轨迹与自杀意念
青春期全程的生活事件轨迹与自杀意念：一项来自德国、法国、爱尔兰和英国的前瞻性队列研究
Life event trajectories and suicidal ideation throughout adolescence: a prospective cohort study from Germany, France, Ireland, and the UK
——《柳叶刀·精神病学》第13卷，第3期，2026年3月——
【摘要】背景：青春期是一个敏感时期，自杀意念在此期间可能达到峰值。尽管创伤性生活事件已被公认为自杀意念的一个既定风险因素，但跨越家庭、自主性、性等多个维度的累积性日常琐碎生活事件所扮演的角色，目前仍未得到充分的探索。至关重要的是，目前尚无证据表明此类事件的累积暴露在多维度上的发展轨迹，与整个青春期期间的自杀意念之间存在何种关联。本研究旨在考察青春期期间累积性及特定领域的生活事件暴露轨迹，与自杀意念之间存在何种关联。方法：在这项多国参与的前瞻性队列研究中，我们利用了来自IMAGEN队列的纵向数据；这些数据于2008年1月至2010年1月期间，在德国、法国、爱尔兰和英国的八个中心收集而成。我们利用这些数据对青春期期间的生活事件暴露轨迹进行了建模，并量化了不同轨迹组别与自杀意念之间的关联。参与者分别在14、16、19和23岁时接受了评估；主要纳入标准包括：具备提供知情同意或知情认可的能力，且其父母已完成关于参与者围产期及发育史的问卷填写。研究采用《生活事件问卷》（LEQ）对参与者在七个维度（包括家庭、意外事故、心理困扰、自主性、越轨行为、性相关事件及搬迁）上对39项生活事件的暴露情况进行了评估；该问卷在上述每一个研究时间点均进行了施测。此外，研究还利用《童年创伤问卷》对参与者的童年期受虐待经历进行了评估；该问卷仅在参与者19岁时施测一次。本研究的主要结局指标为：在不同生活事件暴露轨迹组别中，出现自杀意念的参与者人数；该指标通过LEQ问卷中专门针对自杀意念设置的条目进行评估，评估时间点涵盖前三个研究时间点。此外，我们还对参与者在前三个研究时间点通过《发育与福祉评估》（DAWBA）所填报的关于自伤行为及自杀意念的应答结果进行了比对分析。研究采用“潜在轨迹分析”方法，根据多次评估结果将累积性生活事件暴露情况划分为不同的类别（即轨迹组别）：高暴露组、中高暴露组、中低暴露组及低暴露组。随后，研究利用逻辑混合模型回归分析方法，对不同轨迹组别与自杀意念之间的关联进行了量化；在进行回归分析时，已对童年期受虐待经历及人口学协变量进行了统计学校正。值得注意的是，在本次研究的设计、实施及成果报告过程中，未曾邀请任何具有自杀经历或精神健康障碍经历的个体参与其中。研究结果：在基线期，共有2161名于2008年1月至2010年1月期间入组的青少年被纳入该队列研究（其中女性1106人 [51.2%]，男性1055人 [48.8%]）。后续各波随访的样本量分别为：第一次随访（2011年1月至2012年1月进行）时为1581人；第二次随访（2013年1月至2015年1月进行）时为1474人；第三次随访（2016年1月至2019年1月进行）时为1331人。在每一个时间点上，“高轨迹组”（占全部2161名参与者的1125人 [52.1%]）均拥有最高比例的具有自杀意念的参与者（具体比例分别为：基线前1125人中有90人 [8.0%]；14岁时1125人中有127人 [11.3%]；16岁时759人中有102人 [13.4%]；19岁时709人中有78人 [11.0%]）。在控制了人口学特征及童年期受虐待经历等混杂因素后，该“高轨迹组”在每一个时间点上仍表现出最高的自杀意念预测概率。在预测青少年自杀意念方面，家庭维度的生活事件所具有的预测力（边际 R?GLMM = 0.078）高于所有其他维度生活事件的预测力之和（边际 R?GLMM = 0.073）。研究解读：本研究发现，在多个维度上累积暴露于日常普通生活事件的程度越高，青少年期出现自杀意念的风险也就越高；其中，发生在家庭环境内的生活事件所带来的风险最高。这些结果提示我们，在评估青少年自杀风险时，不仅应关注童年期创伤或受虐待经历，还必须重视对不同类型日常普通生活事件暴露情况的全面评估。
[Summary] Background: Adolescence is a sensitive period when suicidal ideation can peak. Although traumatic life events are a well established risk factor for suicidal ideation, the role of cumulative ordinary life events across multiple dimensions, such as family, autonomy, and sexuality, remains underexplored. Crucially, no evidence is available on how multidimensional trajectories of cumulative exposure to such events relate to suicidal ideation throughout adolescence. We aimed to examine how cumulative and domain-specific trajectories of ordinary life events during adolescence relate to suicidal ideation. Methods: In this multinational, prospective cohort study, we used longitudinal data from the IMAGEN cohort, gathered from eight centres in Germany, France, Ireland, and the UK between January, 2008, and January, 2010, to model trajectories of exposure to life events during adolescence, and to quantify the association between trajectory group and suicidal ideation. Participants were assessed at ages 14, 16, 19, and 23 years, and key inclusion criteria included having the capacity to provide informed assent or consent, and having parental completion of perinatal and developmental history. Exposure to 39 life events across seven dimensions (family, accidents, distress, autonomy, deviance, sexuality, and relocation) were assessed using the Life Events Questionnaire (LEQ), administered at each of the study timepoints. Childhood maltreatment was also assessed, using the Childhood Trauma Questionnaire, administered once at age 19 years. The primary outcome was the number of participants with suicidal ideation across trajectories, assessed via a dedicated item in the LEQ at the first three timepoints. We also compared self-harm and suicidal ideation responses from the Developmental and Well-Being Assessment at the first three timepoints. Latent trajectory analysis was used to identify classes of cumulative life event exposure across repeated assessments: high, mid-high, mid-low, and low. The association between trajectory group and suicidal ideation was quantified using logistic mixed model regressions, adjusting for childhood maltreatment and demographic covariates. No individuals with lived experience of suicidality or mental health conditions were involved in the design, conduct, or reporting of this research. Findings: At baseline, 2161 adolescents, enrolled between January, 2008, and January, 2010, were included in the cohort (1106 [51·2%] female and 1055 [48·8%] male). Sample sizes at subsequent waves were 1581 at first follow-up (which took place between January, 2011, and January, 2012), 1474 at second follow-up (between January, 2013, and January, 2015), and 1331 at third follow-up (between January, 2016, and January, 2019). At each timepoint, the high trajectory group (1125 [52·1%] of 2161 participants) had the highest proportion of participants with suicidal ideation (90 [8·0%] of 1125 at pre-baseline, 127 [11·3%] of 1125 at age 14 years, 102 [13·4%] of 759 at age 16 years, and 78 [11·0%] of 709 at age 19 years). After controlling for demographics and childhood maltreatment, the high trajectory group had the highest predicted probability of suicidal ideation at each timepoint. Life events in the family dimension predicted adolescent suicidal ideation (marginal R2GLMM=0·078) more than those in all remaining dimensions combined (marginal R2GLMM=0·073). Interpretation: We found that higher cumulative exposure to ordinary life events across multiple dimensions is associated with higher suicidal ideation in adolescence, with life events occurring within the family environment carrying the highest risk. These results point to the importance of evaluating different types of everyday life event exposure, not just childhood trauma or maltreatment, when assessing suicidal risk in adolescents.
论文原文：Pei-Jung Chen, Andrew J Lawrence, Xuemei Ma, et al. (2026). Life event trajectories and suicidal ideation throughout adolescence: a prospective cohort study from Germany, France, Ireland, and the UK. The Lancet Psychiatry, Volume 13, Issue 3, Pages190-199. March 2026.
https://doi.org/10.1016/S2215-0366(25)00359-1
（翻译兼责任编辑：MARY）
（需要英文原文的朋友，请联系微信：millerdeng95或iacmsp）
