Identifying disparities in air pollution-attributable cardiovascular health outcomes within a quasi-experimental framework

项目总结/摘要 空气污染-包括暴露于颗粒物（PM）-是发病率和死亡率的主要原因， 美国和全世界。然而，现有的研究检查心血管健康的影响 PM2.5暴露有几个局限性，包括相关的空间和时间的残留混杂 特征，以及有限或相互矛盾的证据，关于PM2.5之间的关联程度 健康结果因个人和社区特征而异。此外，少数国家 在美国，关于PM2.5对心血管影响的研究包含了一系列健康结果， 从收缩压（SBP）的变化到住院和死亡率。检查 空气污染与心血管疾病（CVD）结果之间的关联有助于 我们对因果机制的理解，并揭示了潜在的干预机会。 这个K 01奖的目标是解决文献中的差距，同时也支持候选人的目标， 过渡到一个独立的学者，专注于确定和解决上游环境 CVD差异的决定因素。通过正式的课程，研讨会，指导研究和出席 在科学会议上，Titus博士将获得污染引起的CVD发病机理方面的专业知识 结果，并将追求因果推理，多层次的EHR为基础的分析方法的培训， 机器学习方法。基于泰特斯博士先前作为社会流行病学家的培训， 互补的培训和研究目标将使泰特斯博士阐明复杂的，交叉的模式， 易受空气污染对健康影响的脆弱性，并将为她提供一系列方法工具， 这可以灵活地应用于未来的研究问题，在环境健康，心血管， 流行病学和健康差异。 这项拟议中的研究将利用来自美国退伍军人的大型全国性队列的EHR数据， 在一项新的准实验性研究中，长期PM2.5暴露与CVD差异之间的关联 框架，利用基于住宅流动性的PM2.5暴露的个人水平变化。拟议 目的1）检查PM2.5与心血管疾病相关结果（收缩压变化、住院治疗）之间的关联 和死亡率），2）使用假设驱动的方法来检查协会如何变化 根据社会经济地位和城市化程度，以及3）使用机器学习方法来探索 基于个人和地区水平特征的PM2.5可归因于CVD结局的交叉模式。 这项工作符合NHLBI的战略目标，即阐明CVD中人口差异的驱动因素， 寻求大规模环境和EHR数据的新联系。通过提供工资支持、培训和 正式的导师结构，这个奖项将促进博士泰特斯的过渡到一个独立的职业生涯，重点是 环境驱动的CVD差异的研究。

PROJECT SUMMARY/ABSTRACT Air pollution – including exposure to particulate matter (PM) – is a leading cause of morbidity and mortality in the United States and worldwide. However, existing research examining the cardiovascular health effects of PM2.5 exposure has several limitations, including residual confounding by correlated spatial and temporal characteristics, and limited or conflicting evidence regarding the extent to which associations between PM2.5 and health outcomes vary according to individual and neighborhood characteristics. Moreover, few national studies in the U.S. on the cardiovascular effects of PM2.5 have encapsulated a spectrum of health outcomes, ranging from changes in systolic blood pressure (SBP) to hospitalizations and mortality. Examining associations between air pollution and a continuum of cardiovascular disease (CVD) outcomes contributes to our understanding of causal mechanisms and sheds light on potential intervention opportunities. The goal of this K01 award is to address gaps in the literature while also supporting the candidate’s goal of transitioning to an independent scholar focused on identifying and addressing upstream environmental determinants of CVD disparities. Through formal coursework, workshops, mentored research, and attendance at scientific conferences, Dr. Titus will gain expertise in the pathogenesis of pollution-attributable CVD outcomes, and will pursue methodological training in causal inference, multi-level EHR-based analyses, and machine learning approaches. Building on Dr. Titus’s prior training as a social epidemiologist, these complementary training and research aims will allow Dr. Titus to elucidate complex, intersectional patterns of vulnerability to the health effects of air pollution and will provide her with a combination of methodological tools that can be flexibly applied to future research questions at the nexus of environmental health, cardiovascular epidemiology, and health disparities. The proposed research will leverage EHR data from a large, national cohort of U.S. veterans to examine associations between long-term PM2.5 exposure and CVD disparities within a novel quasi-experimental framework, exploiting individual-level changes in PM2.5 exposure based on residential mobility. The proposed aims 1) examine associations between PM2.5 and CVD-related outcomes (changes in SBP, hospitalizations and mortality) among all veterans, 2) use hypothesis-driven approaches to examine how associations vary according to socioeconomic status and urbanicity, and 3) use machine learning methods to explore intersectional patterns of PM2.5-attributable CVD outcomes based on individual and area-level characteristics. This work is in line with NHLBI’s strategic goals to elucidate drivers of population disparities in CVD and to pursue novel linkages of large-scale environmental and EHR data. By providing salary support, training, and a formal mentorship structure, this award will facilitate Dr. Titus’s transition to an independent career focused on the study of environmentally driven CVD disparities.