Statistical Modeling of Human Fecundity

这一项目的动机是在该研究所进行和正在进行的各种怀孕研究。其中包括纽约州垂钓者队列前瞻性妊娠研究，生理和感知心理社会压力对妊娠时间的影响以及生育和环境纵向调查（LIFE）研究（PI，流行病学分部G. Buck Louis博士）。

This project is motivated by various pregnancy studies that have been conducted and are being conducted at DESPR. These include the New York State Angler Cohort Prospective Pregnancy Study, The Impact of Physiologic and Perceived Psychosocial Stress on Time to Pregnancy and the Longitudinal Investigation of Fertility and the Environment (LIFE) Study (PI, Dr G. Buck Louis of Epidemiology Branch). Fecundability (Gini, 1926) is defined as the probability of conception in a menstrual cycle for couples having regular sexual intercourse without contraception. Reproductive epidemiologists are interested in models for human fecundity to identify the determinants and causes of reproductive dysfunction. There is also suspicion that various environmental factors negatively affect the human fertility. Despite extensive efforts (ex., Weinberg and Gladen (1986), Zhou and others (1996), Scheike and Jensen (1997), Dunson and Zhou (1997), Dunson and Stanford (2005) to name a few), modeling fecundability data remains a statistically challenging task. Indeed, as modeling of fecundability stepped from demography and population-based contexts to reproductive biology and treatment of infertility, the new situation created a strong call for: 1) flexibility, because of the important and ever increasing quantity of information researchers collect; and 2) robustness, because of the potential crude interpretations of the estimated effects and the resulting important decisions in everyday medical practice. Two main sources of heterogeneity effecting fecundability are biological and behavioral. This distinction does not preclude potential interactions between these two factors. For example, both womans length of window of fertility and mans length of life of sperm, which are biological factors, modify the effect of intercourse-pattern on the probability of conception. Several elements have been studied and confirmed as factors of impaired fecundability, such as advanced age or exposure to a toxic substance. However, most of the biological heterogeneity remains unexplained. Similarly, the explained part of the effect of intercourse pattern on fecundability includes the number and timing of intercourse on particular days in relation to ovulation, the quality of cervical mucus and the use of local contraceptives. Nevertheless, the magnitude of the unexplained heterogeneity may be at least as important as the effects of the covariates. The research focus of Dr Sundaram and her collaborators are to account for various measurement errors encountered in observing biological factors and to develop joint modeling approach for intercourse pattern and fecundability as assuming the intercourse pattern to be fixed may not be reasonable and they both share common risk factors, like advanced age, parity (previous child birth) and may be influenced by underlying hormonal levels in both male and female. Some of the projects that have been completed or are in completion stage are listed below: