Multi-Level Modeling of Addiction Comorbidity

项目总结/摘要 修订后的NIH个人奖学金（F30）提案旨在培养职业特定技能， 知识的行为贡献和潜在的神经机制相关的共同发生 药物使用障碍（DUD）和其他精神疾病。DUD合并症的诊断是公认的 在临床上，并且倾向于更困难和昂贵的治疗计划和更差的治疗结果， 单独诊断。尽管DUD合并症的患病率和临床意义，很少有研究 描述了环境、行为和神经组织之间的相互作用，这些因素导致了DUD 共病轨迹。与自我信念和自我导向行为相关的认知受到损害 在患有DUD、抑郁症或PTSD的个体中，这些缺陷背后的神经回路改变， 共病个体尚未被研究。因此，拟议项目的总体研究目标是 识别与DUD合并症相关的功能神经网络相关的特征， 确定网络组织的变化如何导致共病患者自我相关认知的缺陷 个体此外，机器学习计算模型将在大数据上进行训练， 对DUD合并症两个不同阶段的全脑网络组织模式进行分类 发展该提案包括为候选人提供严格的培训计划，以获得以下方面的专业知识： 神经成像方法和神经成像数据的高级计算方法（例如， 结构方程建模、机器学习和多变量模式分析（MVPA））。一个日益 将使用12-60岁成人和青少年的大数据样本（n=550+）来测试本研究的目的。 提议在目标1中，控制年龄和性别，环境变量和自我信念将与 DUD共病的表达。使用全脑中介分析，重要的特质将与 大脑激活区域，重点是前扣带皮层（ACC）和背外侧前额叶皮层 （dlPFC）。目标2将使用结构方程模型来测试ACC网络中的性别特异性变化是如何影响的。 与健康个体相比，DUD和DUD合并症个体的自我调节相关。目标3 将使用MVPA衍生的计算模型来对全脑活动模式进行分类， 青春期易患DUD合并症，成年期持续合并症。通过分类 在疾病发展的两个不同阶段，DUD合并症的大脑活动，这个项目将有助于 为未来研究更有效的治疗方法和更好的预防措施铺平道路， 排除DUD合并症。与计算精神病学相关的职业发展里程碑-临床 医学，负责任的研究行为和计算神经科学-将通过这一点实现 指导研究提案。

PROJECT SUMMARY/ ABSTRACT This revised NIH Individual Fellowship Award (F30) proposal seeks to develop career-specific skills and knowledge of the behavioral contributions and underlying neural mechanisms associated with co-occurring drug use disorder (DUD) and other psychiatric illnesses. The diagnosis of DUD comorbidity is well established clinically, and is prone to more difficult and expensive treatment plans and worse treatment outcomes than either diagnosis alone. Despite the prevalence and clinical significance of DUD comorbidity, few studies have characterized the interactions between environment, behavior, and neural organizations that contribute to DUD comorbidity illness trajectories. Cognitions related to self-beliefs and self-directed behaviors are compromised in individuals with DUD, depression, or PTSD, yet the altered neural circuitry underlying such deficits in comorbid individuals has not been studied. The overall research goal of the proposed project is therefore to identify traits associated with functional neural networks underlying DUD comorbidity and determine how changes in network organization lead to deficits in self-related cognitions in comorbid individuals. Additionally, machine-learning computational models will be trained on Big Data to classify brain-wide patterns of network organization at two distinct stages of DUD comorbidity development. The proposal includes a rigorous training plan for the candidate to gain expertise in neuroimaging methodology and advanced computational approaches to neuroimaging data (e.g., structural equation modeling, machine learning, and multivariate pattern analysis (MVPA)). An increasingly large data sample (n=550+) of adults and adolescents 12-60 years old will be used to test the aims of this proposal. In Aim 1, controlling for age and sex, environmental variables and self-beliefs will be related to the expression of DUD comorbidity. Using whole-brain mediation analyses, significant traits will then be related to areas of brain activation, with focus on the anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (dlPFC). Aim 2 will use structural equation modeling to test how sex-specific changes in ACC networks are related to self-regulation in individuals with DUD and DUD comorbidity compared to healthy individuals. Aim 3 will use MVPA-derived computational models to classify whole-brain patterns of activity that characterize susceptibility to DUD comorbidity during adolescence and sustained comorbidity in adulthood. By classifying brain activity underlying DUD comorbidity at two separate stages of disorder development, this project will help pave the way for future research into more effective treatment methods and better preventative efforts to preclude DUD comorbidity. The career development milestones related to computational psychiatry – clinical medicine, responsible conduct of research, and computational neuroscience – will be attained via this mentored research proposal.