PROXIMATE DETERMINANTS OF LBW AND INFANT MORTALITY

点击翻译按钮获取中文摘要

Reducing low birth weight (LBW) and infant mortality (IM) are prominent objectives of the United States Public Health Service. The proposed research will contribute new information outlining the parameters and causes of LBW and IM. This research is guided by the Mosley-Chen analytic framework. It focuses on identifying the proximate determinants of LBW and IM by maternal age and their racially, ethnically, and economically associated biosocial and biobehavioral roots. Data from the National Health and Nutrition Examination Survey, 1976-1980 and the Hispanic Health and Nutrition Examination Survey, 1982-1984 will be analyzed to estimate prevalences of smoking and other specific biobehavioral risk factors for poor pregnancy outcomes among women of childbearing age. These prevalence rates will be estimated by age, parity, race, Hispanic ethnicity, and socioeconomic factors. The odds of smoking and of having specific diseases will be modeled as functions of social and demographic factors using log-linear modeling techniques. The hypothesis that reproductive health status deteriorates at differential rates for black, white, and Hispanic women will be tested, and the influence of fertility patterns (maternal age, birth spacing, parity) on these deterioration rates assessed. Rates of LBW and IM by parity and maternal age will be estimated for racial, Hispanic ethnic, and socio-economic groups using linked birth and infant death certificate data tapes for the 1985 California, New York, South Carolina, and Missouri birth cohorts. To model the odds of LBW and IM as functions of maternal age, race, ethnicity, socioeconomic factors, parity, prenatal care, medical conditions, smoking, and access to neonatal intensive care technology, and to estimate the relative risk of LBW and IM within and between strata, log-linear modeling techniques will also be applied to these data. Under the Mosley-Chen model, the proposed research will go beyond the demographic identification of high risk populations, and beyond the biomedical focus on the etiology of LBW and IM through disease processes on the individual level. Rather, disease processes will be connected to populations. Causal modeling will be used to locate the social and behavioral correlates of disease processes and identify their consequences for birthweight distributions and infant mortality rates among populations at large.