Browsing by Subject "Mediation analysis"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Associations between dietary factors in early life and childhood growth(2014-05) Zhu, Yeyi; Forman, Michele R.Early life factors play important roles in disease susceptibility in later life. However, the relationship between dietary factors in early life on childhood growth, especially linear growth, remains unclear. This research aimed to improve our understanding of the associations between dietary factors in early life (i.e., infant feeding practices and age of introduction of solid foods) and childhood growth, especially using ulnar length as a surrogate measure of length/height, in a cross-sectional study of 1634 mother-child dyads across eight study centers in the National Children's Study Formative Research in Anthropometry in the United States from 2011-2012 (Chapter 1). Chapter 2 described the data acquisition and preprocessing procedures used in this research and provide practical guidelines of data quality control. In Chapter 3, predictive models for exclusive breastfeeding (XBR) initiation and duration was developed. Discriminant analysis revealed maternal sociodemographic factors had greater discriminating abilities to predict XBR initiation and XBR for 6 months, compared to child birth characteristics and maternal perinatal factors. Chapter 4 demonstrated that ulnar length can serve as an accurate and reliable surrogate measure of recumbent length in healthy infants/children aged 0-1.9 years and of height in healthy children aged 2-5.9 years, respectively. Bland-Altman plots and mixed-effects linear regression analyses showed that the three simple and portable tools (i.e., caliper, ruler, and grid) used to measure ulnar length could be used interchangeably in terms of prediction accuracy. Chapter 5 focused on assessing the interplay among gestational weight gain (GWG), birthweight, infant feeding practices, and childhood anthropometrics. Longer duration of breastfeeding reduced the positive associations of GWG and birthweight with weight-for-age z-scores, weight-for-height/length z-scores, and body mass index-for age z-scores in non-Hispanic Whites. These findings underscore the importance of promoting breastfeeding among women with excessive GWG to mitigate childhood obesity. Longer breastfeeding and a later age at introduction of solid foods had positive effects on ulnar length, a linear growth parameter of upper extremity, in Hispanics. Future prospective research aiming to investigate the underlying mechanisms that drive ethnic variation in these associations between early life dietary factors and childhood growth is warranted (Chapter 6).Item Bayesian mediation analysis for partially clustered designs(2013-05) Chu, Yiyi; Beretvas, Susan NatashaPartially clustered design is common in medicine, social sciences, intervention and psychological research. With some participants clustered and others not, the structure of partially clustering data is not parallel. Despite its common occurrence in practice, limited attention has been given regarding the evaluation of intervention effects in partially clustered data. Mediation analysis is used to identify the mechanism underlying the relationship between an independent variable and a dependent variable via a mediator variable. While most of the literature is focused on conventional frequentist mediation models, no research has studied a Bayesian mediation model in the context of a partially clustered design yet. Therefore, the primary objectives of this paper are to address conceptual considerations in estimating the mediation effects in the partially clustered randomized designs, and to examine the performances of the proposed model using both simulated data and real data from the Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K). A small-scale simulation study was also conducted and the results indicate that under large sample sizes, negligible relative parameter bias was found in the Bayesian estimates of the indirect effects and of covariance between the components of the indirect effect. Coverage rates for the 95% credible interval for these two estimates were found to be close to the nominal level. These results supported use of the proposed Bayesian model for partially clustered mediation in conditions when the sample size is moderately large.