背景：哀伤（grief），又称悲痛，是个体在丧亲（bereavement）事件发生后出现的，与个体精神活动密切相关的一系列心理生理反应。多数类型哀伤属于自然反应，是不需要予以处理的一种自然状态，但一些特殊类型的异常哀伤则会显著影响到个体的健康水平和社会功能。延长哀伤障碍（prolonged grief disorder, PGD）是一种独特的、可识别的异常哀伤。从临床现象学角度考察，在除外急性死亡案例外，从丧亲预备产生哀伤反应到丧亲事件发生后出现PGD的整个过程，是一个连续谱。截至目前，国内外哀伤相关研究领域的纵向研究数量有限，已有的纵向研究多数限于对丧亲后哀伤发展规律的观察，欠缺对丧亲前哀伤的应有关注，尤其是在同一个队列中连续观察丧亲前、后哀伤的发展规律，这并不利于学界进一步深入理解延长哀伤障碍疾病全病程发展规律。

目的：在特定纵向队列中，初步探索丧亲前、后哀伤发展轨迹及其与后继精神障碍（包括而不限于PGD）的相关性。此外，本研究将同时关注抑郁、焦虑、应激等重要精神症状的发展轨迹及其与后继精神障碍的相关性。

方法：本研究是一项探索性单中心前瞻性队列研究，计划纳入60名目前在北京某三甲综合医院住院治疗/社会招募来源的临终者核心家属。在基线期、丧亲前每1月1次、丧亲后6个月内每3个月1次重要精神症状以及安全性评估。分别采用预期性哀伤量表（anticipatory grief scale, AGS）/延长哀伤障碍量表修订版（Prolonged Grief-13 Revised, PG-13-R）、汉密尔顿抑郁量表17项（17-item Hamilton Rating Scale for Depression, HAMD-17）、汉密尔顿焦虑量表（Hamilton Anxiety Scale, HAMA）、事件影响量表修订版（Impact of Event Scale-Revised, IES-R）评估哀伤、抑郁、焦虑、应激症状严重程度；采用综合生理-心理-社会评估问卷（Union Physio-Psycho-Social Assessment Questionnaire，UPPSAQ-70）和简明幸福与生活质量满意度问卷（Quality of Life Enjoyment and Satisfaction Questionnaire, Short Form, Q-LES-Q-SF）评估整体心身健康状态和生活质量；采用哥伦比亚-自杀严重程度评定量表（Columbia-Suicide Severity Rating Scale, C-SSRS）和简明精神病问卷（Brief Psychiatric Rating Scale, BPRS）评估安全性；采用生活事件量表（Life Event Scale, LES）评估研究过程中其它生活事件影响。在第6个月末末次访视时，对受试者进行延长哀伤障碍诊断评估（基于ICD-11，一名受过ICD-11诊断系统训练的精神科执业医师操作，一名受过ICD-11诊断系统训练的精神科执业医师复核），同时对受试者进行其它精神障碍筛查（基于简明国际神经精神访谈, Mini-International Neuropsychiatric Interview 7.0.2, MINI 7.0.2）。基于丧亲前第3个月末、丧亲前第2个月末、丧亲前第1个月末分析丧亲前重要精神症状的变化轨迹，基于丧亲、丧亲后第3个月末、丧亲后第6个月末分析丧亲后重要精神症状变化轨迹。采用群组轨迹模型（group-based trajectory model, GBTM）识别具有相似轨迹的潜在簇。

结果：本研究于2023年10月份至2023年12月份期间，通过社会招募途径筛选临终者核心家属共86人，26人因不符合入组标准筛败。研究最终纳入60位临终者核心家属，平均年龄（52.82 ± 10.18）岁，年龄范围为25岁至74岁，女性32人（53.3%），有躯体疾病既往史32人（53.3%），与临终者关系多为父母（44 [73.3%]）、配偶（6 [10.0%]），VAS-亲密度量表平均得分（6.82 ± 1.68）分，医学判断临终者预期剩余寿命中位数为5（4, 6）个月。研究分析发现：（1）丧亲前AGS量表评分的变化大体呈渐增趋势，丧亲后PG-13R量表评分的变化整体呈逐渐下降趋势。类似地，丧亲前HAMD-17、HAMA、IES-R量表评分的变化均呈渐增趋势，丧亲后HAMD-17、HAMA、IES-R量表评分的变化均呈渐减趋势。（2）丧亲前哀伤水平发展轨迹组包括丧亲前哀伤持续低水平组（n=51 [85%]）和丧亲前哀伤渐增组（n=9 [15%]）；2组在“临终者目前临床主要诊断系统分类”（χ² = 13.916, P = 0.031）、“医学判断临终者预期剩余寿命”（Z = -2.945, P = 0.003）以及“VAS-亲密度量表评分”（t = -2.882, P = 0.006）存在显著统计学差异。哀伤持续低水平组占比最高的主要诊断系统分类为脑血管系统（35.3%）；哀伤渐增组占比最高的主要诊断系统分类为呼吸系统（33.3%）。相较于哀伤渐增组，哀伤持续低水平组的“医学判断临终者预期剩余寿命”更高、“VAS-亲密度量表评分”更低。（3）丧亲后哀伤水平发展轨迹组包括丧亲后哀伤较高-持续下降组（n = 9 [15%]）、丧亲后哀伤中等-持续下降组（n = 26 [43.33%]）、丧亲后哀伤较低-持续下降组（n = 5 [25%]），以及丧亲后哀伤持续较高组（n = 10 [16.67%]）；4组在“受教育程度”（χ² = 21.474, P = 0.044）、“与临终者关系”（χ² = 21.882, P = 0.001）、“临终者目前临床主要诊断系统分类”（χ² = 32.313, P = 0.020）、“医学判断临终者预期剩余寿命”（Z = 27.039, P < 0.001）以及“VAS-亲密度量表评分”（[F_{(3, 56)} = 27.181, P < 0.001]）变量间存在显著统计学差异。在较高-持续下降组和中等-持续下降组中，以受教育程度为高中/中专的受试者占比最高，分别为33.3%和38.5%；较低-持续下降组和持续较高组则分别以大学及以上（40.0%）、初中（40.0%）占比最高。在较高-持续下降组、中等-持续下降组以及持续较高组中，均以与临终者关系为父母的受试者占比最高，分别达到77.8%、96.2%和60.0%；而在较低-持续下降组则以与临终者关系为其他（养父母、养儿女、与临终者长期共同生活的时间≥5年）居多（46.7%）。脑血管系统是较高-持续下降组、中等-持续下降组最多见的目前临床主要诊断分类，占比分别达到55.6%、42.3%；心血管系统和呼吸系统则分别是较低-持续下降组和持续较高组最多见的目前临床主要诊断分类，占比均为40.0%。4组中，医学判断临终者预期剩余寿命中位数最低的是持续较高组[3.00 (2.00, 3.25)岁]，最高的是中等-持续下降组[5.50 (5.00, 6.00)岁]。每2组间VAS-亲密度量表评分均有显著统计学差异，持续较高组（8.80 ± 1.14）最高，较低-持续下降组（4.87 ± 1.51）最低。（4）丧亲前抑郁和应激均有2种发展轨迹，且在基本特征上与丧亲前哀伤的2种轨迹相仿，即一种以持续的低水平为特征，而另一种以较高的起始水平并持续上升为特征。不同的是，丧亲前焦虑的发展轨迹除上述2种发展轨迹外，还存在一种同以较高的症状起始水平并持续上升为特征的发展轨迹。丧亲后抑郁、焦虑和应激的发展轨迹表现出更大的变异性：即使哀伤和应激均可归类出4种发展轨迹，而抑郁和焦虑均可归类出3种发展轨迹，但难以基于轨迹特征作以症状轨迹间类比。抑郁、焦虑、应激症状发展轨迹每两组间一般人口信息学和临床特征均无显著统计学差异。（5）丧亲前哀伤轨迹与丧亲后哀伤轨迹间存在显著相关性（Cramer’s V = 0.695, P < 0.001），二者均未发现与其它症状轨迹间存在显著相关性。（6）在丧亲后第6个月末，丧亲前哀伤渐增组的各疾病发生率均显著高于哀伤持续低水平组（PGD: 77.8% vs 5.9%, χ² = 28.471, P < 0.001; MDD: 55.6% vs 9.8%, χ² = 11.529, P < 0.001; GAD: 100.0% vs 17.6%, χ² = 24.706, P < 0.001）；丧亲后哀伤持续较高组各疾病发生率均显著高于其它丧亲后哀伤轨迹组。医学判断临终者预期剩余寿命以及VAS-亲密度量表评分在任一种疾病发生与否的组间均有显著差异。具体地，各疾病发生组的医学判断临终者预期寿命均显著短于未发生组，而VAS-亲密度量表评分均显著高于未发生组。

结论：丧亲事件带给个体的影响是广泛、涉及多维度的。符合特定哀伤发展规律的居丧者更易发生延长哀伤障碍等后继精神障碍，具体地，在丧亲前表现为哀伤水平持续上升或在丧亲后表现为哀伤水平持续较高的居丧者。特定一般人口学信息和临床特征，如居丧者丧亲预备时间较短、临终者罹患疾病具有强消耗性和病情波动大的特点、居丧者与临终者具有更紧密的情感联结/依附等，可以帮助靶定风险人群。本研究支持在预备丧亲人群中进行含括丧亲前、后全过程的长期纵向研究的文化可接受性，未来可进行大规模多中心研究以验证本研究发现。

Background: Grief involves a range of psychophysiological responses following the loss of a loved one. Although most grief reactions are considered normal and do not require clinical intervention, certain forms can impair health and social functioning. Prolonged grief disorder (PGD) is recognized as a distinct, diagnosable condition. Excluding cases of sudden death, the progression from pre-bereavement grief responses to PGD after bereavement appears to follow a continuous spectrum rather than discrete stages. The limited number of longitudinal studies that examine both pre- and post-bereavement trajectories has hindered our understanding of PGD.

Objective: To identify the longitudinal trajectories of grief before and after bereavement and to assess their association with PGD and other adverse mental health outcomes in the same cohort. In addition, we evaluated the trajectories of depression, anxiety, and stress, and their associations with subsequent mental disorders.

Methods: In this single-center, prospective cohort study, 60 immediate family caregivers of terminally ill patients were enrolled from Peking Union Medical College Hospital or via community outreach. Participants were assessed at baseline, monthly during the pre-death period, and every three months for six months after bereavement. Grief symptoms were measured using the Anticipatory Grief Scale (AGS) and the Prolonged Grief Disorder-13 Revised (PG-13-R). Depressive, anxiety, and stress symptoms were assessed with the 17-item Hamilton Rating Scale for Depression (HAMD-17), the Hamilton Anxiety Scale (HAMA), and the Impact of Event Scale-Revised (IES-R), respectively. Overall physical, psychological, and social well-being, as well as quality of life, were evaluated using the Union Physio-Psycho-Social Assessment Questionnaire (UPPSAQ-70) and the Quality-of-Life Enjoyment and Satisfaction Questionnaire, Short Form (Q-LES-Q-SF). Safety was monitored with the Columbia-Suicide Severity Rating Scale (C-SSRS) and the Brief Psychiatric Rating Scale (BPRS), and other life changes were documented using the Life Event Scale (LES). At the final visit, PGD was diagnosed according to ICD-11 criteria by two independent psychiatrists trained in ICD-11. Other psychiatric disorders were screened using the Mini-International Neuropsychiatric Interview version 7.0.2 (MINI 7.0.2). Group-based trajectory modeling (GBTM) was used to identify latent symptom trajectories.

Results: Between October and December 2023, 86 individuals were screened; 26 were excluded for not meeting the inclusion criteria, leaving 60 participants (mean [SD] age, 52.82 years [range, 25–74]; 53.3% women; 53.3% with a history of physical illness). Most participants were offspring of terminally ill patients (44 [73.3%]), followed by spouses (6 [10.0%]). The mean (SD) VAS-Closeness Scale score was 6.82 (1.68), and the median (IQR) estimated life expectancy of the terminally ill patients was 5 (4–6) months. (1) Pre-death AGS scores increased over time, whereas post-death PG-13R scores decreased. Similarly, pre-death HAMD-17, HAMA, and IES-R scores increased, while these scores declined during the post-death period. (2) Pre-death grief trajectories included a persistently low-level group (n = 51 [85%]) and an increasing group (n = 9 [15%]). There were statistically significant differences between these two groups in terms of the primary diagnostic system classification of the terminally ill patient (χ² = 13.916, P = 0.031), medically estimated remaining life expectancy (Z = -2.945, P = 0.003), and the VAS-Closeness Scale scores (t = -2.882, P = 0.006). For the low-level group, the primary diagnostic system classification with the highest proportion was cerebrovascular disease (35.3%), while for the increasing group, it was respiratory disease (33.3%). Participants in the low group had a higher medically estimated remaining life expectancy and lower VAS-Closeness Scale scores than the increasing group. (3) Post-death grief trajectories included high/decreasing (n = 9 [15%]), moderate/decreasing (n = 26 [43.33%]), low/decreasing (n = 15 [25%]), and high/prolonged (n = 10 [16.67%]). Significant differences were found in education level (χ² = 21.474, P = 0.044), relationship to the terminally ill patient (χ² = 21.882, P = 0.001), primary diagnostic classification (χ² = 32.313, P = 0.020), estimated remaining life expectancy (Z = 27.039, P < 0.001), and VAS-Closeness Scale scores (F_[3,56] = 27.181, P < 0.001). In the high/decreasing and moderate/decreasing groups, most participants had a high school or technical secondary education (33.3% and 38.5%, respectively); in contrast, the low/decreasing group mainly had a college education or higher (40.0%), and the high/prolonged group had mostly a junior high school education (40.0%). Furthermore, most participants in the high/decreasing (77.8%), moderate/decreasing (96.2%), and high/prolonged groups (60.0%) were children of the terminally ill patient, whereas the low/decreasing group primarily included individuals in other relationships (46.7%). Cerebrovascular disease was most frequent in the high/decreasing (55.6%) and moderate/decreasing (42.3%) groups, while cardiovascular and respiratory diseases were most common in the low/decreasing and high/prolonged groups (each 40.0%). Participants in the high/prolonged group had the lowest estimated remaining life expectancy (median, 3.00 months; IQR, 2.00–3.25) and the highest VAS-Closeness Scale scores (mean, 8.80 [SD, 1.14]). (4) Both pre-death depression and stress exhibited two trajectories, like the two types of pre-death grief trajectories—one characterized by a consistently low level, and the other by a relatively high initial level with a continued increase. The difference was that pre-death anxiety also had a third trajectory featuring a high initial level that continued to rise. While both grief and stress could be categorized into four trajectories, and depression and anxiety could each be categorized into three trajectories, making direct comparisons across trajectories based on their characteristics was difficult. There were no statistically significant differences in general demographic or clinical characteristics between any two trajectories of depression, anxiety, or stress. (5) The pre-death grief trajectory was correlated with the post-death grief trajectory (Cramer’s V = 0.695, P < 0.001). Neither grief trajectories were correlated with the trajectories of depression, anxiety, or stress. By the end of the study (6 months post-bereavement), apart from significant differences in the incidence of PGD, MDD, and GAD across different grief trajectories, no significant differences were observed across the trajectories of other symptoms. (6) The incidence rates of each illness in the increasing group were higher than those in the low-level group (PGD: 77.8% vs 5.9%, χ² = 28.471, P < 0.001; MDD: 55.6% vs 9.8%, χ² = 11.529, P < 0.001; GAD: 100.0% vs 17.6%, χ² = 24.706, P < 0.001). In the high/prolonged group, the incidence rates were all significantly higher than in the other post-death grief groups. Participants developed any kind of mental disorders had a significantly lower medically estimated remaining life expectancy and a higher VAS-Closeness Scale scores compared to those who did not.

Conclusion: Individuals with increasing pre-death grief or persistently high-level post-death grief were more likely to develop PGD and other mental disorders. Limited preparatory time for bereavement, the chronicity and debilitation of the terminally ill patient’s condition, and greater emotional closeness may help identify these vulnerable populations. This study supports the cultural acceptability of long-term longitudinal research on the bereaving population. Larger-scale, multicenter studies are needed in the future to further validate these findings.