Return to search

Perinatal smoking and its related factors

Indiana University-Purdue University Indianapolis (IUPUI) / The smoking rate of low-income pregnant women is almost 4 times the rate for higher-income women. A better understanding of smoking within the low-income population is needed. The purpose of this dissertation was to study smoking and related factors for pregnant and postpartum women living in poverty. The first component used Rodger’s evolutionary concept analysis method and uncovered three attributes, four antecedents, and three consequences for smoking cessation. The second (N = 1,554) and third (N = 71,944) components were a secondary data analysis of first-pregnancy Medicaid-eligible women enrolled in the Nurse-Family Partnership program from 2011-2016. The second component explored patterns of smoking and depression and their associations. Eight distinct patterns of smoking and depression were found. Smokers were more likely than nonsmokers to have depressive symptoms at the end of pregnancy (OR = 1.37 [1.04, 1.81] and 12 months post-delivery (OR = 1.93 [1.47, 2.51]. The third component investigated covariates present during early pregnancy and their relationships with smoking status and sought to find best fitting predictive models. Multivariable logistic regression showed cigarette use in the 3 months prior to pregnancy and at program intake were significant predictors for smoking status at the end of pregnancy and 12 months post-delivery. Interactive Matrix Language, Structured Query Language, and iterations of logistic regression identified 5 covariates (high school education, cigarette use prior to pregnancy, smoking status at pregnancy baseline, depression, and self-mastery) for the best fitting model at the end of pregnancy and three additional covariates (post-secondary education, marital status, and race) for the 12 months post-delivery model. The area under the receiver operator characteristic curve was 0.9681 for the end of pregnancy model and 0.9269 for 12 months post-delivery model, indicating excellent prediction ability of the models. Results can be integrated in smoking prevention education, screening, and cessation intervention programs.

Identiferoai:union.ndltd.org:IUPUI/oai:scholarworks.iupui.edu:1805/17758
Date12 July 2018
CreatorsJones, Ashley
ContributorsShieh, Carol, Staten, Lisa, Carter-Harris, Lisa, Stiffler, Deborah, Macy, Jon
Source SetsIndiana University-Purdue University Indianapolis
Languageen_US
Detected LanguageEnglish
TypeDissertation

Page generated in 0.0016 seconds