Hierarchical statistical modeling and causal inference approaches to elucidate exposure pathways underlying health disparities

分层统计模型和因果推理方法阐明健康差异背后的暴露途径

基本信息

批准号：
10589163
负责人：
Li Luo
金额：
$ 16.25万
依托单位：
UNIVERSITY OF NEW MEXICO HEALTH SCIS CTR
依托单位国家：
美国
项目类别：
财政年份：
2015
资助国家：
美国
起止时间：
2015-08-01 至 2025-03-31
项目状态：
未结题

来源：
https://reporter.nih.gov/project-details/10589163
关键词：
Air Pollutants American Asthma Bayesian Method Behavior Behavioral Biological Cause of Death Chemical Exposure Cheyenne Chronic Chronic Disease Communities Community Health Complex Crows Data Data Set Databases Development Diabetes Mellitus Diet Dimensions Disease Disparity Ensure Environmental Exposure Environmental Hazards Environmental Health Environmental Pollutants Exposure disparity Exposure to Frequencies Future General Population Genetic Predisposition to Disease Health Health Disparities Research Hypertension Incidence Individual Infectious Agent Infrastructure Intervention Intervention Studies Kidney Diseases Knowledge Life Style Machine Learning Malignant Neoplasms Malignant neoplasm of liver Manuscripts Mediation Metal exposure Metals Mining Modeling National Cancer Institute National Health and Nutrition Examination Survey Native Americans Navajo Not Hispanic or Latino Nutrient Obesity Outcome Pathway Analysis Pathway interactions Physical activity Play Policy Making Population Preparation Prevalence Psychosocial Stress Public Health Reduce health disparities Renal carcinoma Reporting Research Research Project Grants Resource Allocation Resources Risk Risk Factors Role Sioux Indians Socioeconomic Status Solid Waste Disposals Source Statistical Data Interpretation Statistical Methods Statistical Models Structure Techniques Uranium War Water behavioral economics cancer type causal model cold temperature cost effective data management data quality data reuse dietary disparity reduction exposure pathway frontier health determinants health disparity health equity high risk improved innovation large datasets lifestyle factors malignant stomach neoplasm member modifiable risk novel response social social determinants sociodemographics socioeconomics statistics stressor transmission process tribal community tribal lands wasting

项目摘要

Summary RP3 Hierarchical statistical modeling and causal inference approaches to elucidate exposure pathways underlying health disparities The health disparity between the Native American population and the US general population arises from the complex interplay between multiple socio-demographic, behavior, lifestyle and genetic susceptibility factors. Environmental contaminants are increasingly acknowledged to play an important part in explaining health disparity through their combined or interaction effects with other factors. Proximities of Native American communities to abandoned uranium mines (AUM) have been of particular health concern. These chronic exposures to AUM waste related metal mixtures pose higher risk for developing chronic and fatal diseases including hypertension, diabetes, kidney disease, and types of cancer in Native American populations compared to the US population. The hypothesis of this project is that the three Native American tribal communities included in this study (Navajo Nation, Crow, and Cheyenne River Sioux) encounter great risk of exposures to environmental hazards (mine waste related metal mixture exposures, unregulated water resources, and illegal dumping, etc.). These hazardous exposures along with socioeconomic status, psychosocial stress, behavior/lifestyle factors influence multiple biological pathways to produce health disparities in Native American communities. The complex set of exposure variables including dietary nutrients, physical activity, infectious agents, air pollutants and metal exposures at both the individual and community levels are acknowledged as contributors to health disparities, however, their relative contributions of the potential causal factors have not been well studied. The objective of this project is to employ data-driven and modeling approaches to understand the relative contribution of different environmental, behavior, and socioeconomic determinants of the health disparities between the native population and the US national population. We will use innovative modeling approaches such as decomposition analyses and structural causal models to estimate the effects of risk factors at the individual and community level on the health disparities. In Aim 1, we will collect data and summarize the frequency distributions for major chronic and fatal diseases in the Native American communities. In Aim 2, we will employ novel hierarchical modeling approaches to estimate the relative contribution of different risk factors at the individual level and community level to the health disparities. In Aim 3, we will implement frontier causal pathway analyses to illustrate the intermediate mechanisms explaining the health disparity. Aim 4 is to examine the complex correlation structure among multi-dimensional exposures, intermediate biological responses, and health endpoints using frontier statistical approaches. We expect this project will identify major contributing factors that explain a large proportion of the health disparity, and in addition elucidate the intermediate causal pathway that the effects are transmitted to the health disparity endpoints. These findings have the potential to inform policymaking on the cost-effective resource allocation to maximally reduce disparity and improve community health.

概括 RP3分层统计建模和因果推理方法以阐明暴露健康差异的潜在途径美洲原住民与美国普通人口之间的健康差异来自多个社会人口统计学，行为，生活方式和遗传敏感性因素之间的复杂相互作用。环境污染物越来越多地承认在解释健康方面起着重要的作用通过与其他因素的结合或相互作用效应的差异。美国原住民的接近废弃铀矿（AUM）的社区特别关注健康。这些慢性暴露于AUM废物相关的金属混合物造成较高的慢性和致命疾病的风险包括高血压，糖尿病，肾脏疾病和类型的癌症与美国人口相比。该项目的假设是三个美国原住民部落本研究中包括的社区（纳瓦霍民族，乌鸦和夏安河苏族）遇到了很大的风险暴露于环境危害（矿废物相关的金属混合物暴露，不管制的水资源和非法倾销等）。这些危险的暴露以及社会经济地位，社会心理压力，行为/生活方式因素会影响产生健康的多种生物学途径美国原住民社区的差异。一组复杂的暴露变量，包括饮食营养素，个人和社区的体育锻炼，传染剂，空气污染物和金属暴露但是，水平被认为是对健康差异的贡献者，但是，它们的相对贡献潜在的因果因素尚未得到很好的研究。该项目的目的是采用数据驱动和建模方法以了解不同的环境，行为和社会经济的决定因素是土著人口与美国国家之间的健康差异人口。我们将使用创新的建模方法，例如分解分析和结构性因果模型，以估计个人和社区层面上风险因素对健康的影响差异。在AIM 1中，我们将收集数据并总结主要慢性和致命的频率分布美国原住民社区的疾病。在AIM 2中，我们将采用新颖的分层建模估计各个层面和社区不同风险因素的相对贡献的方法水平与健康差异。在AIM 3中，我们将实施边境因果途径分析以说明中间机制解释了健康差异。目标4是检查复杂的相关结构在多维暴露，中间生物学反应和使用边界的健康终点中统计方法。我们预计该项目将确定主要的促成因素，以解释大型健康差异的比例，此外，阐明了影响是传输到健康差异终点。这些发现有可能告知有关的政策具有成本效益的资源分配，以最大程度地降低差异并改善社区健康。