Quasi-experimental studies of early risk factors for severe psychopathology

描述（由申请人提供）：意义：犯罪、药物使用问题、自杀和严重的精神疾病，如精神分裂症和双相情感障碍，与巨大的个人痛苦和社会成本相关。确定这些问题的原因，特别是在生命的早期，是一项至关重要的研究工作。研究人员已经确定了许多假定的环境因素，这些因素与这些严重问题的风险增加有关，包括母亲在怀孕期间吸烟、母亲在怀孕期间的压力、母亲的生育年龄和父亲的生育年龄提前。然而，我们不知道这些风险因素是严重精神病理形式的偶然风险因素，还是仅仅是家庭混杂的指标。我们对潜在因果机制的理解有限，部分原因是由于大多数现有的这些风险研究都未能使用可以测试替代假设的研究设计，特别是环境和遗传选择因素的作用。创新：我们建议利用强大的准实验设计，自然实验，可以分离共同发生的风险机制，并解释未测量的遗传和环境过程，探索几个假定的早期风险因素的影响。特别是，我们将比较完全暴露的不同兄弟姐妹，这种方法排除了所有可能解释早期推定风险与后来结果之间统计关联的遗传因素，以及兄弟姐妹共有的所有环境因素。我们还将比较不同暴露的半兄弟姐妹、全兄弟姐妹的后代（全表兄弟）、半兄弟姐妹的后代（半表兄弟）和双胞胎的后代。这些额外的比较将使我们能够为我们的发现提供聚合证据，并确定任何家族选择因素（遗传和/或环境）的来源。对后代、父母和社区层面风险的详细评估将进一步使我们能够确定早期风险因素与严重精神病理相关的机制。方法：我们建议分析一个大型数据集，其中包括1940-1995年在瑞典出生或生活的每个人的信息（nh750万），其中包括早期环境风险、相关家庭和社区风险的测量，以及来自瑞典国家登记处的严重犯罪、药物使用、自杀和精神健康问题的有效指标。由于数据集还包括每个人与其所有家庭成员的生物学相关性信息，我们能够通过比较环境和遗传风险不同的个体，对这些关联进行大量的准实验研究。环境和调查员：我们经验丰富的团队包括与早期产前和围产期因素相关的风险的国际专家；重度精神病理的评估及病因分析使用大型国家登记处；并使用准实验设计来研究假定的环境因素。印第安纳大学和卡罗林斯卡学院提供了良好的环境和相关资源。

DESCRIPTION (provided by applicant): Significance: Crime, substance use problems, suicide, and severe mental illness, such as schizophrenia and bipolar disorder, are associated with enormous personal distress and societal costs. Identifying the causes of these problems, especially very early in life, is a crucial research endeavor. Researchers have identified numerous putative environmental factors that are associated with increased risk for these severe problems, including maternal smoking during pregnancy, maternal stress during pregnancy, maternal age at childbearing, and advancing paternal age at childbearing. We do not know, however, whether these risk factors are casual risk factors for severe forms of psychopathology or are merely indices of family confounding. Our limited understanding of the underlying causal mechanisms is due, in part, to the fact that most existent studies of these risks have failed to use research designs that can test alternative hypotheses, especially the role of environmental and genetic selection factors. Innovation: We propose to utilize powerful quasi-experimental designs, natural experiments that can pull-apart co-occurring risk mechanisms and account for unmeasured genetic and environmental processes, to explore the effects of several putative early risk factors. In particular, we will compare differentially exposed full siblings, an approach that rules out all genetic factors that could account for the statistical association between early putative risks and later outcomes, as well as all environmental factors that siblings share. We will also compare differentially exposed half siblings, offspring of full siblings (full cousins), offspring of half siblings (half cousins), and offspring of twins. These additional comparisons will enable us to provide converging evidence for our findings and identify the source any familial selection factors (genetic and/or environmental). Detailed assessments of offspring, parental, and community- level risks will further enable us to identify the mechanisms through which early risk factors are associated with severe psychopathology. Approach: We propose to analyze a large dataset that includes information on every individual who was born or lived in Sweden from 1940-1995 (NH7,500,000), which includes measures of early environmental risks, correlated familial and community risks, and well-validated indices of severe criminal, substance use, suicide, and mental health problems from Swedish national registries. Because the dataset also includes information on the biological relatedness of each individual with all of their family members, we are able to conduct numerous quasi-experimental studies of the associations by comparing individuals who differ in both their environmental and genetic risk. Environment and Investigators: Our experienced team includes international experts in the risks associated with early prenatal and perinatal factors; the assessment and etiology of severe psychopathology; the use of large, national registries; and the use of quasi-experimental designs to study putative environmental factors. Indiana University and the Karolinska Institutet provide strong environments and relevant resources. PUBLIC HEALTH RELEVANCE: The staggering burden on individuals and society caused by violence, substance use problems, suicide, and severe mental illness has been well documented. Researchers have identified early risk factors that predict these major societal problems, but, it is still unclear whether putative early risks are truly causal or whether part-or most-of the associations with these risks are due to background familial factors. In order to improve prevention programs and our understanding of early risk factors, the current proposal seeks support to analyze a unique resource, a large dataset that includes every individual in Sweden for roughly five decades, relying on natural experiments to rigorously test causal hypotheses.