NIOSHTIC-2 Publications Search
Occupational cohorts: confounder & effect-modifier models.
Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, R01-OH-008087, 2009 Dec; :1-34
Epidemiologic studies of cohorts exposed occupationally to carcinogens or other toxicants often have several features that complicate statistical analyses of the data: longitudinal data, repeated measures, cumulative exposures; certain covariates are substantially correlated with cumulative exposure; health endpoints of interest often have other potential confounding or effect-modifying exposures; and the effect of exposure measurement errors may be significant. Statistical issues related to several of these features have not been dealt with satisfactorily. In this proposal, several statistical methods were developed to improve upon existing methods for analyzing occupational cohort data by minimizing these limitations. Particularly, we proposed the use of ridge regression as an alternative method when multicollinearity exists among covariates; we examined how different types of measurement errors affect the estimation of risk associated with repeated occupational exposure to radiation, while controlling for possible confounders such as time since exposure. We have showed that random measurement error has a larger impact on the relative risk estimates compared to the error due to the Minimum Detection Level (MDL) of the dosimeter. We developed new statistical methods to design efficient cohort study and methods to estimate and test for effect modification when traditional methods failed. These methods were developed with a focus on the goal of reaching a non-statistical audience therefore can be easily adopted by epidemiologists and clinicians.
Epidemiology; Group-dynamics; Exposure-levels; Exposure-limits; Carcinogens; Toxins; Statistical-analysis; Risk-factors; Environmental-exposure; Hazardous-materials; Models; Mathematical-models; Analytical-models
Xiaonan Xue, Ph.D., Division of Biostatistics,Department of Epidemiology & Population Health, Albert Einstein College of Medicine Bronx, NY 10461
Final Grant Report
National Institute for Occupational Safety and Health
Albert Einstein College of Medicine