Dissertations and Theses

Date of Degree

9-30-2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy (Ph.D.)

Department

Epidemiology and Biostatistics

Advisor(s)

Katarzyna Wyka

Committee Members

Deborah Balk

Terry T. Huang

Glen Johnson

Subject Categories

Epidemiology | Public Health

Keywords

obesity, small area estimates, spatial error model, geographically weighted regression

Abstract

Objectives: The purpose of this dissertation was to create small area estimates to describe and explore the county-level spatial variation of extreme obesity among adults in the United States and to assess the county-level association and spatial variation of extreme obesity and mortality.

Methods: The Behavioral Risk Factor Surveillance System was used in conjunction with data from the Census Bureau to estimate county-level model-predicted prevalence of extreme obesity using multilevel regression and poststratification. Spatial dependence of estimates was assessed using global Moran’s Index, and local Moran’s Indices were used to identify clusters of higher and lower rates of extreme obesity and to map significant clusters of counties. Then, the study examined the association between extreme obesity and age-adjusted all-cause mortality using ordinary least squares regression, the spatial error model, and geographically weighted regression. Throughout the study moderate obesity was assessed for comparison.

Results: County-level prevalence of extreme obesity ranged from 1.3% to 15.7%, showing more variability than evident from state-level analysis. Moran’s Index for extreme obesity was 0.35, indicating the distribution of prevalence of extreme obesity was spatially clustered. There were significant clusters of high prevalence of obesity in several regions including the Mississippi Delta region and the Southeastern Coastal Plains, and significant clusters of low prevalence in the Rocky Mountain region and the Northeast. Both extreme and moderate obesity were positively associated with mortality rates after controlling for covariates and the association was stronger for extreme obesity. One unit rise in prevalence of extreme obesity was associated with increased 8.4 mortality rate (SE=1.07, p

Conclusions: County-level prevalence estimates of extreme obesity indicated substantial variation across the United States and demonstrated spatial dependence. Geographical prevalence patterns were similar for moderate and extreme obesity though many individual counties had an uneven distribution of prevalence by obesity group. Hot spots were identified indicating clustering of high prevalence of extreme obesity. Extreme obesity was more strongly associated with mortality than was moderate obesity and the association displayed significant spatial heterogeneity. This study highlights the importance of disaggregation by obesity class and local geographies in ongoing obesity research.

Included in

Epidemiology Commons

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.