探测潜在抑郁症脆弱者的脑动态模型研究

Identifying individuals who are potentially vulnerable to.depression is an important step toward prevention of depression. There is an.increasing interest in investigating underlying neural mechanisms of depression.vulnerability to identify biomarkers for vulnerable individuals. So far, family.history, risk genes, or early life stress (ELS) are typically used to define.depression vulnerability. However, it is well-known that some individuals with ELS.can develop resilience to depression. Therefore, using neural responses through.functional magnetic resonance imaging (fMRI) to directly identify depression.vulnerable individuals would be ideal and objective. However, because of large.individual differences in fMRI responses, one can hardly use fMRI responses for.individualized diagnosis. To compensate this weakness, we propose to observe neural.responses to a repeated stressor dynamically and use the changing trend in neural.activation within each subject to identify depression vulnerable individuals. In.response to a repeated stressor, such as meeting deadlines for multiple grant.submissions, some people habituate and neural responses in their emotional.processing system decrease quickly (habituated emotional response, HER). But.others may have prolonged emotional responses (PER) and even become sensitized to.the stressor such that brain responses get amplified over time. We hypothesize that.HER vs. PER in response to a repeated stressor differentiates those who are resilient.to depression from who are vulnerable to depression..Our pilot study revealed that decreased activation in an emotion regulatory.region, the ventromedial prefrontal cortex (vmPFC), and increased activation in an.emotion processing region, the amygdala, in later stressful trials relative to early.trials could be an index of PER; whereas continuously heightened activation in the.vmPFC but quickly dropped activation in the amygdala could be the feature of HER..In this proposed project, we will examine 200 nondepressed college students to.further investigate neural activation patterns (PER vs. HER) in response to a.repeated stressor and their underlying phenotypes. We will follow up all subjects.for two years to validate the PER and HER patterns. We will also examine differences.in neural circuits as a function of depression vulnerability or resilience as.measured by ELS and personality traits (rumination and mindfulness). Genotypes will.also be examined as an exploratory aim..The major innovative elements of this effort are: 1) setting up a new model.for detecting depression vulnerability at an individual level which even can be.considered as a new way for fMRI in clinical application; 2) translating personality.factors and exposure to early life stress into individual differences in depression.vulnerability and resilience.

在正常人群中发现潜在的抑郁症、焦虑症易患者是早期预防心理疾病的关键。近年来，对抑郁症脆弱者脑内变化的研究虽得到了更多的关注，但大多数研究是依据家族史或幼年精神创伤史等来定义抑郁症脆弱性的。事实上，不是所有的有家族史或幼年精神创伤史的都是抑郁症脆弱者，有些人也可成为抑郁症抵御者。依据大脑对刺激的动态变化来定义抑郁症脆弱性应该是一个更客观的测量指标。本项目将利用反复的精神压力诱发的脑内动态变化建立抑郁症脆弱性的新模型，同时研究与这一生物标记相关的心理因素与环境因素，并验证这一生物学标记的可信度。我们将采集血样本，为日后研究基因与环境对脑功能的影响打下基础。.因我们将依据脑内对压力刺激的动态反应变化来观察判断每个人的结果，这不仅可以检测潜在的抑郁症焦虑症易患者，又可避免因脑功能成像信号的个体差异大而难以用来做个体化诊断的弊病。该方法将为脑功能成像在个体化诊断的临床应用上提供一个新方向。

本项目依据大脑对精神压力刺激的动态反应变化来探究个体对精神压力的耐受性，以在个体化层面检测潜在的抑郁症易患者。为达到研究目标，本项目在经典的压力课题基础上进行修改，设计了适合功能磁共振（fMRI）扫描的压力课题（计算课题、演讲课题）与放松课题（冥想课题），收集了62例受试的可用数据（包括三个fMRI数据及行为学数据），探究了大脑活动在不同的精神压力和放松状态下的差异，分析了可以预测抑郁脆弱性的脑影像学指标，并进一步解明了抑郁脆弱性的神经机理。本项目，1）探索精神压力刺激影响脑网络活动的机理，明确与压力相关的脑网络基础，重点探究了不同情绪状态下，内源网络和外源网络之间地交互变化；2）利用动态网络分析技术，提取数据的动态特性，探究个体在压力刺激后的恢复能力及个体差异，探索利用动态影像学标记预测个体的抑郁脆弱性的可行性；3）利用关联分析，寻找计算和演讲这两种不同压力刺激给大脑造成的相同影响，帮助理解压力影响大脑活动的神经机理。通过上述研究，主要结论如下：1）在压力环境时，大脑的注意力资源倾向于外源网络，则内源网络的注意力资源减少，且定义的注意力偏差指数的改变与抑郁程度相关，这部分研究结果与项目负责人之前在杜克大学采集的五羟色胺（抑郁症的另一重要影响因素）减少的数据结果一致；2）抑郁指数较高的个体的动态恢复指数较低，与假设一致，说明了利用动态信息预测抑郁脆弱性的可能，为早期筛查抑郁脆弱者提供一种新的途径；3）在正式实验中，演讲课题与计算课题均能引起右背侧前岛叶、右背外侧前额叶的活动的升高，说明这两个脑区在抑郁症的产生和发展中具有重要的作用。本项目探究了不同情绪对于大脑活动的调控机制，理解压力这一抑郁症重要因素对大脑活动的影响，为抑郁症的早期预防提供理论基础，推动fMRI应用于临床诊断。本项目一共发表了5篇会议摘要，在国际磁共振年会（ISMRM）、脑图谱组织年会（OHBM）等国际会议中进行了展示，其中，一篇口头报告被选为2016年韩国磁共振年会（KSMRM）的最佳摘要。