Etiology and Modeling Core

点击翻译按钮获取中文摘要

Abstract: Complex Persistent Pain Conditions: Etiology and Modeling Core. This core will support four functions relevant to subprojects and cores of this program: 1) convene an Internal Scientific Advisory Committee (ISAC) to insure consistency of data collection and analysis strategies for the program's five complex persistent pain conditions (CPPCs); 2) implement consistent study protocols and data collecfion procedures throughout the program; 3) undertake statistical analysis of pooled data from all five of the program's CPPCs; 4) consultation with project Pis in their analysis of aims specific to single CPPCs. During the first year, the ISAC will be devoted to overseeing development of study protocols and data collecfion methods. In subsequent years the ISAC will devote greater effort to refining the program's hypothesized causal model and specifying the data analysis strategy. Data collection procedures will be made consistent through creation of program-wide data management systems. Analysis of pooled data from all five of the program's CPPCs will address the program's specific aims that focus on genetic, physiologic, psychological and environmental exposures contribufing to the risk of one or more CPPCs. Conventional epidemiological methods for analysis of casecontrol studies will be compare odds of exposure between 1,500 CPPC cases and 1,230 controls who have none of the CPPCs. This sample size will allow us to identify risk factors that affect as few as 5% of subjects and that have exposure odds ratios as low as 1.6, while controlling for multiple tesfing (eg. from among 3,000 genetic markers) with an appropriate rate of false discovery. The most promising risk factors identified through these convenfional methods Will be evaluated jointly in multivariate structural equation models that isolate individual effect, interactions and pathways among causes. Investigators in the etiology and modeling care will consult with project Pis in their analysis of specific CPPCs which will follow a similar strategy of odds ratio estimation and detection of interacfion, followed by multivariate causal modeling. For these analyses, 300 cases with a specific CPPC will be contrasted with at least 1,230 controls.