Regression Analysis of the Percentage of Respiratory-related Deaths using Respiratory Covariates and OLS

Dublin Core

Title

Regression Analysis of the Percentage of Respiratory-related Deaths using Respiratory Covariates and OLS

Subject

OLS analysis of covariates predictors of the percentage of deaths from 2013-2017.

Description

This map presents a subsequent analysis of the "% of Deaths due to Respiratory Diseases" Covariate and seeks to identify what factors, if any, can serve as significant predictors of the % of deaths linked to respiratory disease. Prior to analysis, respiratory disease-linked deaths were found to be a significant predictor of Years of Potential Life Lost (YPLL) in the Four Corners Region.

A number of variables including average PM2.5 levels, lung cancer death rates, median household income, obesity rates, background radon exposure, and the % of adults who smoked were identified as likely variables that could be tested. Data on average PM2.5 levels, median household income, the % of adults who smoked, and obesity rates were obtained from 2017 datasets produced by the University of Wisconsin Population Health Institute's 'County Health Rankings' project for counties in Arizona, Colorado, New Mexico, and Utah covering the years 2011-2013. A 5-year average for lung cancer deaths from 2013-2017 was obtained from the CDC's USCS website. Radon exposure values were coded from an EPA produced radon-zone map at the County Level. Each dataset was cleaned and joined by FIPS code to an existing shapefile before analysis in ArcPro 2.6.1. All counties with suppressed data were excluded from the analysis.

OLS Diagnostics indicate these regression results was safe to use. Spatial Autocorrelation was assessed with Moran's I with no significant difference found indicating that variation is spatially random. The OLS analysis computed a R-squared value of 0.44. Average PM2.5 Levels, Obesity and Median Household Income were found to be significant at p < 0.001. The Radon Zone classification was found to be significant at p < 0.05. Smoking was the only covariate that was not found to be statistically significant. Taken together, these findings suggest that, while the percentage of deaths due to respiratory disease are spatially sensitive, the actual causes are not. Environmental conditions such as PM2.5 levels and radon were found to be significant as expected and diseases such as obesity and lung cancer were found to be explanatory.

This is by no means an exhaustive list of covariates that may affect the percentage of deaths due to respiratory illness, but these findings could provide stepping points for future additions to the Four Corners Health Atlas.

Creator

David Gunther

Source

County Health Rankings Data - AZ
County Health Rankings Data - CO
County Health Rankings Data - NM
County Health Rankings Data - UT

CDC - United States Cancer Statistics

EPA Radon Zone Classification - Previous research conducted by Dr. Ruth Kerry, Associate Professor of Geography at Brigham Young University; taken from the generalized EPA Radon Zone Map created in 1993.

Country Shapefiles - ESRI Living Atlas
U.S. State Shapefiles - ESRI Living Atlas
U.S. County Shapefiles - ESRI Living Atlas
U.S. State Capitals - ESRI Living Atlas

Publisher

Website maintained by Chantel Sloan, Associate Professor in the Public Health Department at Brigham Young University.

Date

County Health Rankings Data – 2017 covering data from 2011 – 2013

Cancer Statistics – 5-year summary from 2013-2017

EPA – Original source map published in 1993

Contributor

David Gunther

Rights

All shapefiles are ESRI proprietary data

Format

[PDF]

Language

[English]

Type

Map file generated using ArcGIS Pro 2.6.1

Coverage

This is a series of 1:6,500,00 maps of the Four Corners region centered on Arizona, Colorado, New Mexico, and Utah.

Files

OLS_Respiratory.pdf
RespiratoryOLS_SpatialAutocorrealtion_Report.pdf
Lung_Mort.pdf

Reference

Regression Analysis of the Percentage of Respiratory-related Deaths using Respiratory Covariates and OLS, David Gunther, Website maintained by Chantel Sloan, Associate Professor in the Public Health Department at Brigham Young University.,

County Health Rankings Data – 2017 covering data from 2011 – 2013

Cancer Statistics – 5-year summary from 2013-2017

EPA – Original source map published in 1993