There is substantial interest in determining the etiology of preterm delivery (PTD). Despite much effort, the cause remains elusive and effective prevention measures do not exist. Uteroplacental vascular compromise (UPVC) via inflammation, thrombosis, or atherosis is a biologically plausible cause of preterm delivery, albeit not adequately explored. We propose to test his hypothesis by conducting a prospective, epidemiologic study of UPVC and by integrating information about known risk factors for PTD. Placental histopathologic examination and morphometric analysis of the basal plate will be done to assess compromise of placental vessels. We will explore novel, possible antecedents of such compromise, dyslipidemia and insulin resistance, using nuclear magnetic resonance analysis of lipid subclasses and fasting insulin-glucose ratios. Given the inaccessibility of the uteroplacental vasculature in ongoing gestations at midpregnancy, we will utilize non-invasive measures of UPVC, Doppler velocimetry of the uterine artery, and maternal serum alpha fetoprotein to indirectly evaluate vascular function. In addition, we will carefully evaluate tobacco and cocaine use, nutrition, and changes in vaginal microflora within our cohort. The data will enable us to thoroughly assess whether UPVC constitutes a distinct etiologic pathway for PTD and help to identify modifiable risk factors. We will utilize cohort and case-cohort techniques, refined in our present research, to answer these questions. Blood, urine and vaginal fluid are collected twice between 15 and 20 weeks and between 24 and 29 weeks gestation. Hair will be collected after delivery. All subjects will complete two telephone interviews and two self administered questionnaires regarding various behaviors, dietary intake, physical activity, and psychosocial stressors. Placentas will be collected at the time of delivery and histopathologic analysis will be completed by an experienced perinatal pathologist for cases and a non-case subgroup. Nuclear magnetic resonance measurement of lipoprotein subclasses will be done to assess dyslipidemia. Insulin glucose ratios will be measured from fasting blood samples. We expect to enroll a cohort of 1800 women with 250 preterm deliveries and a randomly selected non-case subgroup (n=500). We will analyze the relationship between UPVC and PTD using logistic regression. Given 1) the size of the study, 2) thorough histopathologic assessment of the placenta, 3) extensive questionnaire data, 4) biologic markers of exposure to bacterial vaginosis, insulin resistance, dyslipidemia, and cocaine use, and 5) the careful assessment of potential confounding factors, this study promises to markedly advance our knowledge of the potential role of UPVC in the etiology of PTD.
Showing the most recent 10 out of 43 publications
