STATISTICAL METHODS IN HUMAN DEVELOPMENT/CLINICAL STUDIES

点击翻译按钮获取中文摘要

This project was initiated to develop new statistical methodologies for problems in human development and to apply existing methodologies to those problems in novel ways. Work has begun in two areas: (1) statistical modeling of serum concentrations of reproductive hormones in peripubertal children and (2) statistical modeling of the distribution of birth weights in human populations. Secretion of pituitary gonadotropins and gonadal hormones is pulsatile; serum profiles exhibit distinct peaks with irregular spacings. With an appropriate statistical model, one could describe these profiles with a few parameters, allowing comparisons among subjects with normal and abnormal pubertal development. Such comparisons would provide insight into mechanisms of dysfunctional development. We are currently evaluating models used for adult hormone profiles to see whether they also provide adequate descriptions in prepubertal children. These models are limited to one hormone at at a time; our long term goal is to develop models for suites of hormones as interacting systems. Development of a satisfactory statistical model for the distribution of birth weights in a population will facilitate a refined understanding of the relationship of birthweight to infant mortality. We have developed a model which allows a parsimonious description of a birth weight distribution and which can be used to investigate social and environmental influences on birthweight distributions.