Social determinants of health and a grounded mixed-methods approach to explain declining life expectancy in Eastern Kentucky : 1980 – 2014.

Geography;mixed-methods;health;regression;life expectancy;Kentucky

William Renda

University:University of Louisville

Department:Geography and Geosciences

关键词: Geography;    mixed-methods;    health;    regression;    life expectancy;    Kentucky;   

Others  :  https://ir.library.louisville.edu/cgi/viewcontent.cgi?article=4213&context=etd

美国|英语

来源: The Universite of Louisville's Institutional Repository

【 摘 要 】

Trends in life expectancy for the United States from 1980-2014 suggest general improvement overall. However, eastern Kentucky stands out with a cluster of counties experiencing declining life expectancy. Given that life expectancy is an excepted indicator of overall population health, this anomaly in eastern Kentucky warrants investigation. This thesis uses a grounded mixed methods approach to explore this trend across the greater Appalachian region. Content and discourse analysis of interviews with public health and medical professionals in the study area revealed key themes perceived as being related to declining life expectancy, which informed variable selection for regression analysis. Regression models included ordinary least squares, spatial lag and geographically weighted regression (GWR) and were used to explore relationships between key themes from the interviews and declining life expectancy. Results suggest certain causes of death, including cancer, cardiovascular disease, cirrhosis and other liver disease, mental health and substance abuse, and other non-communicable diseases, are statistically significant related to the eastern Kentucky anomaly. Further, overdose mortality rate, average poverty rate, mining/manufacturing jobs and change in college-educated adults stood out as the most powerful explanatory variables. Moreover, GWR revealed nonstationarity in the relationships between life expectancy and the explanatory variables whereby the regression model performed best in eastern Kentucky, but had less explanatory power elsewhere, particularly in the norther Appalachian region. Ultimately, place-based interpretations of health in Appalachia and the mixed methods approach provided deeper insight into life expectancy trends across the Appalachian region and specifically eastern Kentucky were socioeconomic and cultural forces moderate health and engagement with the healthcare system.