Bio-Social Pathways to Poor Mental Health in The UK Population: Using Blood Samples to Index Neurobiological Factors

心理健康不良是整个生命过程中遗传和环境因素之间复杂相互作用的结果，并可由应激性生活事件(SLEs)触发。全面了解这些相互作用有助于确定风险和保护因素。我们保持弹性的能力可能会受到发生在儿童和青春期大脑发育和成熟过程中的基因-环境相互作用(GxE)的影响。此外，一些在血液中测量的表观遗传学和蛋白质组生物标记物可能通过它们测量神经生物学因素的能力来掌握我们在给定时刻的弹性。本论文将整合遗传、表观遗传学和蛋白质组生物标记物，以及SLEs，以了解这些生物标记物如何在通向不良心理健康和恢复能力的道路上相互作用。本论文将产生三篇研究论文。前两篇论文将通过与精神健康症状的关联来评估遗传、表观遗传学和蛋白质组生物标记物的表面有效性，第三篇论文将把重要的生物标记物整合到一个更全面的纵向模型中，该模型结合了SLEs以及保护和风险因素。论文1：将根据心理健康结果对一系列GxE进行建模。对于“E”，将使用三种早期生活暴露：社会经济地位(母亲的教育和父亲的职业)和家庭破裂(完整的家庭、父母搬出去或父母去世)。对于“G”，将产生28倍的多基因指数来替代大脑的结构、功能和分子测量：6倍的皮质下区域体积与抑郁症状、双相情感障碍和/或神经质有关，19倍的功能连通性特征与精神分裂症、双相情感障碍、严重的抑郁障碍和/或交叉障碍特征有关，以及3倍基于人脑组织中糖皮质激素受体(一种关键的应激反应蛋白)的基因表达研究。这些GxE代表发生在童年和青春期的神经发育过程。主要结果将是基于超过GHQ-12或SF12-MCS的既定阈值的两种衡量标准。论文2：DNA甲基化(DNaM)数据和蛋白质组数据将被用来得出一系列生物标记物：a)与心理社会应激相关的五个研究良好的基因的dNaM生物标记物：两个应激反应蛋白NR3C1和FKBP5，5-羟色胺转运体(SLC6A4)，催产素受体(OXTR)和脑源性神经营养因子(BDNF)，以及b)两个蛋白质组生物标记物：创伤后应激障碍(PTSD)风险评分和海马体体积评分。这些生物标记物将被测试与前述精神健康结果的前瞻性关联，以及与第一篇论文中任何重要的GxE交互作用因素的关联。这些分析可能表明，dNaM和蛋白质组生物标记物在神经发育因素和精神健康结果之间起中介作用，并可能表明它们作为压力暴露的衡量标准或作为风险生物标记物的效用。论文3：路径建模将纳入前两篇论文中确定的具有重大早期生活暴露和生物标记物(遗传、表观遗传学和蛋白质组)的心理健康结果、SLES的纵向测量以及风险和保护因素，其中可能包括收入、社会支持、慢性压力源(工作质量、邻里水平因素)、锻炼、教育和亲社会行为。通过对这些途径进行建模，本文旨在帮助我们理解心理健康状况不佳的恢复力和脆弱性，包括遗传因素，以及可能在政策层面和个人层面上可以改变的因素。它还旨在强调一系列神经生物学因素在精神健康状况不佳中的重要性，并证明它们在外周血液中是可测量的。

Poor mental health results from a complex interplay between genetic and environmental factors across the life course, and can be triggered by stressful life events (SLEs). A comprehensive understanding of these interactions could help identify risk and protective factors. Our ability to remain resilient may be influenced by the gene-environment interaction (GxE) which takes place during brain development and maturation in childhood and adolescence. Furthermore, some epigenetic and proteomic biomarkers measured in the blood may hold clues as to how resilient we are at a given moment, by their ability to measure neurobiological factors. This thesis will integrate genetic, epigenetic and proteomic biomarkers, along with SLEs, to understand how these interact on the pathway to poor mental health and resilience. The thesis will generate three research papers. The first two papers will assess the face validity of genetic, epigenetic and proteomic biomarkers by their association with mental health symptoms, and the third will integrate significant biomarkers into a more comprehensive longitudinal model which incorporates SLEs and protective and risk factors. Paper 1: A range of GxE will be modelled in relation to mental health outcomes. For "E", three early life exposures will be used: socioeconomic position (maternal education and father's occupation) and family disruption (intact family, parent moved out, or parent died). For "G", 28x polygenic indices which proxy structural, functional and molecular brain measurements will be generated: 6x subcortical brain regional volumes which are genetically correlated with depressive symptoms, bipolar disorder and/or neuroticism, 19x functional connectivity traits which are genetically correlated with schizophrenia, bipolar disorder, major depressive disorder and/or a cross-disorder trait, and 3x based on gene expression studies of the glucocorticoid receptor (NR3C1, a key stress response protein) in human brain tissue. These GxEs represent neurodevelopmental processes which take place during childhood and adolescence. The primary outcomes will be binary measures based on exceeding established thresholds for GHQ-12 or SF12-MCS. Paper 2: DNA methylation (DNAm) data and proteomic data will be used to derive a range of biomarkers: a) DNAm biomarkers of five well-studied genes related to psychosocial stress: two stress response proteins NR3C1 and FKBP5, serotonin transporter (SLC6A4), oxytocin receptor (OXTR), and brain-derived neurotrophic factor (BDNF), and b) two proteomic biomarkers: a post-traumatic stress disorder (PTSD) risk score, and a hippocampal volume score. These biomarkers will be tested for prospective associations with the aforementioned mental health outcomes, and for associations with any significant GxE interacting factors from the first paper. These analyses could suggest that the DNAm and proteomic biomarkers act as mediators between neurodevelopmental factors and mental health outcomes, and may indicate their utility as measures of stress exposure, or as risk biomarkers. Paper 3: Path modelling will incorporate mental health outcomes with significant early life exposures and biomarkers (genetic, epigenetic and proteomic) identified in the first two papers, longitudinal measures of SLEs, and risk and protective factors, which may include income, social support, chronic stressors (work quality, neighbourhood-level factors), exercise, education and prosocial behaviour. By modelling these pathways, this paper aims to aid our understanding of resilience and vulnerability to poor mental health, including genetic factors, and factors which may be modifiable at the policy level and at the individual level. It also aims to underscore the importance of a range of neurobiological factors in poor mental health, and to demonstrate their measurability in peripheral blood.