Longitudinal Personalized Dynamics Among Anorexia Nervosa Symptoms, Core Dimensions, and Physiology Predicting Suicide Risk

项目总结/摘要 神经性厌食症（AN）是一种严重的精神疾病，在所有精神疾病中死亡率最高， 自杀是第二大死因。尽管自杀率极高，但自杀的风险因素 在这个高风险人群中的想法（SI）和行为/尝试（SA）还没有得到很好的理解。虽然 有证据表明，威胁反应，压力反应，过度觉醒，情绪失调，和激动有助于 到自杀风险，这些过程之间的动态关系还没有在一个 全面的，瞬间的基础上。我们的科学前提，从我们过去的工作发展而来，是应用 概念到行动和网络理论的结合将使我们能够识别动态的纵向相互作用 在芯尺寸（例如，唤醒，威胁），AN症状和个体之间和个体内的SI/SA。 我们的研究目标是：（1）识别AN和SI/SA并存的症状和维度风险相互作用 人与人之间和人与人之间，（2）区分哪些风险因素预测SI vs SA，（3）测试这些风险因素是否 预测SI/SA的发作。这些目标将最终确定新的预防措施应针对哪些因素 和治疗努力。我们将采用多单元的分析方法，结合新颖、前沿的 自杀和网络科学的进步我们将收集关于AN和自杀行为的密集的实时数据， 焦虑、过度兴奋、情绪调节和激动使用移动的技术，以及 情绪调节（通过心率变异性）和唤醒（通过 皮肤电活动表征过度觉醒和加速表征睡眠-觉醒周期），从 230例诊断为AN/非典型AN（AAN）的患者。在1个月、6个月和1年随访时， 测试个体风险因素是否预测SI/SA。我们预计在整个研究期间将有35 - 58名参与者患有SA。 具体目标是（1）测试跨时间和人与人之间的症状和维度 共病SI/SA和AN症状，（2）开发个性化的网络模型，以确定哪些自杀和AN 特征预测个体内的SI/SA和探索性目标（3）以测试AN之间是否存在差异 和AAN。拟议的研究使用高度创新的方法，结合密集的纵向数据 收集方法，通过可穿戴传感器技术测量生理数据，以及 网络科学来回答以前无法解决的问题， 会导致自杀这项研究具有临床意义。如果我们识别出 有助于自杀风险，这些数据将提供一个模型的个性化医疗的整个领域 精神病学，以及提供新的干预目标，以预防和治疗AN谱疾病。 此外，我们开发的算法可以用于（a）临床医生友好的软件，以确定治疗 目标是防止SI/SA和（B）可穿戴警报设备，可以在SA发生之前破坏SA。

PROJECT SUMMARY/ABSTRACT Anorexia nervosa (AN) is a severe mental illness with the highest mortality rate of any psychiatric disorder, with suicide as the second leading cause of death. Despite extremely high rates of suicide, risk factors for suicidal ideation (SI) and behaviors/attempts (SA) in this high-risk population are not well understood. While there is evidence that threat reactivity, stress-response, over-arousal, emotion dysregulation, and agitation contribute to suicide risk, the dynamic relations among these processes have not been characterized on a comprehensive, momentary basis. Our scientific premise, developed from our past work, is that the application of ideation-to-action and network theories will enable the identification of dynamic longitudinal interactions among core dimensions (e.g., arousal, threat), AN symptoms, and SI/SA both between and within individuals. Our study goals are to (1) identify symptom and dimension risk interactions of co-occurring AN and SI/SA between and within persons, (2) differentiate which risk factors predict SI vs SA and (3) test if these risk factors predict onset of SI/SA. These goals will ultimately identify which factors should be targeted in novel prevention and treatment efforts. We will use a multiple units of analysis approach, combined with novel, cutting-edge advances in suicide and network science. We will collect intensive real-time data on AN and suicide behaviors, anxiety, over-arousal, emotion regulation, and agitation using mobile technology, as well as psychophysiological assessment of emotion regulation (via heart-rate variability) and arousal (via electrodermal activity characterizing over-arousal and acceleration characterizing the sleep-wake cycle), from 230 individuals with a diagnosis of AN/Atypical AN (AAN). At 1-month, 6-month, and one year follow-up we will test if individual risk factors predict SI/SA. We expect 35-58 participants will have SA across our study period. Specific aims are to (1) test which symptoms and dimensions across time and between-persons maintain comorbid SI/SA and AN symptoms, (2) develop personalized network models to identify which suicide and AN features predict SI/SA within individuals and an exploratory aim (3) to test if there are differences between AN and AAN. The proposed research uses highly innovative methods, combining intensive longitudinal data collection methods, measurement of physiological data via wearable sensor technology, and novel advances in network science to answer previously unresolvable questions pinpointing which individual risk factors contribute to suicide outcomes. The proposed research has clinical impact. If we identify patterns that contribute to suicide risk, these data will provide a model of personalized medicine for the entire field of psychiatry, as well as providing novel intervention targets to prevent and treat AN spectrum illnesses. Additionally, the algorithms we develop can be used in both (a) clinician friendly software to identify treatment targets to prevent SI/SA and (b) in wearable alert devices that can disrupt SA before it occurs.