Patterns of Maternal Smoking During Pregnancy

University Of Chicago, Chicago IL

Investigators

Linked publications & trials

Abstract

Studies of maternal smoking during pregnancy have traditionally conceptualized it as a relatively stable behavior, based on the assumption that few changes will occur after the transition to pregnancy. Those few studies that have examined changes over the course of the pregnancy have examined them categorically (e.g., third trimester smoking or not). In contrast, we present preliminary evidence of substantial fluctuation in maternal smoking over the course of the pregnancy, including repeated changes in overall status (i.e., smoking or not) and in categorical status (e.g., light to moderate). While these data provide evidence of individual fluctuation, group patterns of maternal smoking during pregnancy have not been empirically identified. Lack ; of empirical knowledge about patterns of maternal smoking seriously impedes scientific progress for two reasons. First, prenatal exposure to cigarettes has serious consequences, including accruing evidence of long term consequences such as increased risk of disruptive behavior disorders. Identifying the role that maternal smoking plays in the etiology of complex, multifactorial child outcomes will require more precise specification of exposure. Second, prevailing methods of prenatal cessation intervention are frequently limited to the first prenatal visit. Classification of specific patterns of smoking behavior may inform the development of targeted interventions. The proposed project is designed to classify patterns of maternal smoking during pregnancy utilizing sophisticated methods of trajectory analysis. We propose to conduct secondary data analysis of the Maternal-Infant Smoking Study of East Boston (MISSEB), a prospective population-based study with repeated measures of maternal smoking throughout the course of the pregnancy (n=873). Group patterns of maternal smoking trajectories (e.g., late pregnancy relapse, cycling between cessation and relapse) will be modeled using a semi-parametric mixed-model approach. Specific aims of the project are to: (1) characterize patterns of maternal smoking during pregnancy, using both self-reported and biochemical measures of maternal smoking and, (2) examine the explanatory power of these patterns for predicting adverse perinatal outcomes.