Refining Borderline Personality Disorder Using Flexible Latent Variable Modeling

描述（由候选人提供）：边缘型人格障碍（BPD）是一种复杂的精神障碍，与主观痛苦、生活质量差和慢性自杀倾向有关，估计自杀完成率高达10%。由于BPD患者存在攻击性、自杀和自残行为，其治疗利用率明显高于心境障碍或其他人格障碍患者，导致治疗成本显著。虽然这种疾病影响了1-2%的人口，但BPD患者占所有门诊患者的10-20%和住院患者的15%。尽管与BPD相关的显著的社会负担、主观痛苦和不良的精神预后，但相对于其他形式的精神病理，这种表型还没有得到很好的理解。在一定程度上，由于理论概念和DSM诊断标准中反映的相当大的异质性，BPD的构建有效性的提高受到了阻碍。现有的诊断标准定义了一个有意义但分散的结构，如果仅使用这些标准来选择患者，未来的研究工作将受到阻碍。本研究旨在通过探索其他临床和实验标记物，以及使用灵活的潜在变量建模和统计模拟来明确BPD的分类，从而完善BPD的表型。这项工作将导致经验衍生的BPD分类，包括特征，亚型，或两者兼而有之，以及基于数据的阈值，以确定临床显著的边缘性病理。本研究有可能识别更多同质的BPD患者群体，这将提高未来实证工作的分辨力和可重复性。此外，明确BPD患者之间的异质性将改善未来的治疗研究，因为临床策略可以更好地针对个体患者。具体目的是：1)使用蒙特卡罗模拟方法评估现有分析方法在解决BPD潜在结构方面的效率，2)使用整合连续和分类概念的混合潜在变量模型阐明BPD内部异质性，以及3)通过探索其他临床和实验标记物来完善BPD表型。这项拟议的研究将使用两个主要的流行病学数据集（国家共病调查复制和国家酒精及相关疾病流行病学调查）和从先前的五项研究中汇总的616名参与者的大样本进行，其中约25%患有BPD。此外，一种实验性的精神病理学排斥敏感性测量方法将被开发和验证，使用较小的168名成人参与者样本，该测量方法将被整合到分类模型中。我们期望DSM-IV的BPD特征可以通过一个连续的严重程度维度来最好地描述，但是该障碍的亚型将在其他标记的基础上被识别，包括攻击性、不信任、专注依恋和排斥敏感性。竞争的维度，分类和综合模型将被测试。

DESCRIPTION (provided by candidate): Borderline Personality Disorder (BPD) is a complex psychiatric disorder associated with subjective misery, poor quality of life, and chronic suicidality, with estimates of suicide completion as high as 10%. Because of aggressive, suicidal, and self-injurious behaviors, treatment utilization among BPD patients is significantly higher than patients with mood disorders or other personality disorders, resulting in significant treatment costs. Although this disorder affects 1-2% of the population, individuals with BPD account for 10-20% of all outpatients and 15% of inpatients. Despite the significant societal burden, subjective distress, and poor psychiatric outcomes associated with BPD, the phenotype is not well understood relative to other forms of psychopathology. Improving the construct validity of BPD has been impeded, in part, by the considerable heterogeneity reflected in theoretical conceptualizations and the DSM diagnostic criteria. Existing diagnostic criteria define a meaningful, but diffuse, construct and future research efforts will be impeded if patients are selected using these criteria alone. The proposed research seeks to refine the BPD phenotype by exploring additional clinical and experimental markers, as well as using flexible latent variable modeling and statistical simulation to clarify the classification of the disorder. This work will result in an empirically derived taxonomy of BPD consisting of traits, subtypes, or both, as well as data-based thresholds for identifying clinically significant borderline pathology. This research has the potential to identify more homogeneous groups of BPD patients, which will increase the resolving power and replicability of future empirical work. Additionally, clarifying the heterogeneity among BPD patients will improve future treatment research, as clinical strategies can be tailored better to individual patients. The specific aims are: 1) to assess the efficiency of extant analytic methods for resolving the latent structure of BPD using Monte Carlo simulation methods, 2) to clarify within-BPD heterogeneity using hybrid latent variable models that integrate continuous and categorical conceptions, and 3) to refine the BPD phenotype by exploring additional clinical and experimental markers. This proposed research will be conducted using two major epidemiological datasets (the National Comorbidity Survey Replication and the National Epidemiologic Survey on Alcohol and Related Conditions) and a large sample of 616 participants aggregated from five prior studies, approximately 25% of whom have BPD. In addition, an experimental psychopathology measure of rejection sensitivity will be developed and validated using a smaller sample of 168 adult participants, and this measure will be integrated into classification models. We expect that the DSM-IV BPD features are best described by a single continuous dimension of severity, but that subtypes of the disorder will be identified on the basis of additional markers, including aggression, mistrustfulness, preoccupied attachment, and rejection sensitivity. Competing dimensional, categorical, and integrative models will be tested. PUBLIC HEALTH RELEVANCE: Borderline Personality Disorder (BPD) is a complex psychiatric disorder associated with aggressive, suicidal, and self-injurious behaviors that often require intensive long-term treatment, resulting in significant treatment costs. Although this disorder is disproportionately represented in inpatient and outpatient settings relative to its prevalence in the population, patients diagnosed with this condition are so heterogeneous that it has been difficult to identify its causes or to design effective treatments. The proposed research seeks to refine the prevailing conception of BPD using sophisticated statistical models and epidemiological samples, which will permit better classification of BPD, resulting in improved treatment outcomes for patients.