Adequate physical activity (PA) and proper nutrition during pregnancy are important factors contributing to the health of the mother and the fetus. Despite the well-established benefits of PA and proper nutrition during pregnancy, many women do not meet the recommendations, particularly in low-income populations. Discrepancies between the recommended and actual behaviors occur due to various factors, which can be classified into Intrapersonal, Interpersonal, and Environmental variables, constructed according to the Ecological Model. The overall purpose of this dissertation was to examine factors impacting PA and dietary behaviors in low-income pregnant women based on the Ecological Model framework. Specifically, this study assessed the Intrapersonal, Interpersonal, and Environmental factors that may impact levels of household PA (HPA), Job PA (JPA), Leisure-Time PA (LTPA), and diet quality (DQ). A convenience sample of low-income pregnant and postpartum was recruited nationwide using the Research Match online system. Enrolled participants completed a questionnaire comprised of five parts: 1) demographics, 2) correlates of PA, 3) correlates of DQ, 4) the International Physical Activity Questionnaire, and 5) a fruit and vegetable intake survey (FV). The questionnaire was self-administered and conducted online using the Qualtrics online platform. Descriptive statistics were preformed for all demographic data of the participants, all correlates of PA and diet, all PA data and FV score. Data was categorized for all correlated and all outcome variables (HPA, JPA, LTPA, and DQ). For each outcome variable, three correlation matrices were created (one for each of the Intrapersonal, Interpersonal, and Environmental factors). Based on established criteria, significant variables were selected to be included in the confirmatory factor analysis (CFA). A CFA model for each Intrapersonal, Interpersonal, and Environmental latent factor was created and model fit was evaluated. Four structural equation models (SEM) models were created and evaluated for predicting HPA, JPA, LTPA, and DQ from the three latent factors. A convenience sample of 158 women completed the survey with 109 providing complete data for analysis. Of these, median HPA was 28 MET-hrs/wk, median JPA 0.2 MET-hrs/wk, median LTPA 6.0 MET-hrs/wk and median FV intake 1.75 cups/day. While 43% met LTPA guidelines, only 12% met FV guidelines. No significant associations among latent Intrapersonal, Interpersonal, and Environmental factors for HPA, JPA, LTPA, and DQ behaviors were found. It is possible that the latent variables were not well specified by the correlates considered. However, this analysis was the first attempt to examine the Ecological model when using such latent factors in a pregnant population. Our findings suggest the Ecological Model as analyzed in this dissertation might not capture the essence and the complexity of PA and DQ behaviors in the most accurate way. Future studies, examining larger samples and more traditional low-income population of pregnant women could provide more support for the Ecological Model when considering PA and DQ behaviors during pregnancy..
Read
