Substantially motivated by analytic needs from a NIH/EPA co-funded P01 project, this project develops a set of novel statistical models and efficient algorithms to evaluate, interpret, and predict impacts of early life exposures to environmental risk factors on children health and developmental outcomes. The P01 project has collected massive complex data from multiple longitudinal birth cohorts for a systematic study of the hypothesis that environmental exposures (e.g. BPA, phthalates and heavy metals) and during pre- and/or post-natal life influence developmental plasticity, thereby altering susceptibility to adult chronic diseases. This project will be accomplished through the following specific aims:
Aim 1 : Develop statistical models and algorithms to evaluate delayed somatic growth and sexual maturation during the adolescent period driven by prenatal and/or postnatal exposures. These methods will help us assess if, and which toxicant sets, and in which form, modify children growth patterns and tempo of sexual maturation.
Aim 2 : Develop semi-parametric stochastic models to evaluate the functional rate of growth changes and its potential exposure-driven alterations during the child period of 0-5 years old. This new class of stochastic differential equations allows analyze both growth velocity and acceleration as functions of subject's characteristics and environmental exposures to better understand, evaluate and predict child growth patterns.
Aim 3 : Develop statistical methods to assess the validity of merging longitudinal cohort data and to perform joint analysis of merged longitudinal data. Merging longitudinal data sets from multiple cohorts is complicated by the underlying heterogeneity across cohorts, and proper methods are needed to validate the data merging as well as to perform a valid joint analysis of merged data.
Aim 4 : Develop, test, distribute, and support freely available implementations of the proposed methods in this application. The developed statistical tools can be applied to analyze data form longitudinal cohort studies of similar scale and complexity. The advanced and expanded statistical models and methods can provide adequate answers to the most benefit of disease prevention and health practice as well the best interest of well beings.
Emerging areas of research in environmental health sciences (EHS) have led to the development of new hypotheses and technologies to collect massive complex data in the EHS that present great challenges to the data analysis. The existing statistical methodologies are not inadequate to address several major analytic challenges that complicate the data analysis in many EHS studies. Substantially motivated by analytic needs from the ELEMENT projects, the primary goal of this project is to develop a set of novel statistical models and efficient algorithms to evaluate, interpret, and predict impacts of early lie exposures to environmental risk factors to children health and developmental outcomes. This new toolbox will advance and expand the statistical models and methods in EHS that provide adequate answers to the most benefit of disease prevention and health practice as well the best interest of well beings.
