Impact of Air Pollution on Stroke Incidence in Texas
Date
2004-12
Authors
Anand, Sowmya
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Stroke is a cardiovascular disease that affects arteries supplying blood to the brain. A stroke is caused when blood vessels in the brain are blocked by a clot or when the blood vessel bursts. Cardiovascular diseases (heart disease and stroke) have been the leading cause of death in Texas since 1940. About 2 out of every 5 deaths are due to cardiovascular disease in Texas. Risk factors for stroke include high blood pressure, tobacco use, diabetes mellitus, atrial fibrillation, high blood cholesterol, physical inactivity, obesity, excessive alcohol intake and increasing age. The objective of this thesis was to examine the relation between air pollution and stroke death rate in Texas in 1999. Air pollutants considered in this study were carbon monoxide, sulfur dioxide, nitrogen oxide, particulate matter (PM) 2.5, PM 10, volatile organic compounds and total criteria air pollutant. The stroke death rate was collected for the year 1999. With the stroke death rate as a dependent variable and carbon monoxide, sulfur dioxide, nitrogen oxide, PM 2.5, PM 10, volatile organic compounds pollutant level and total criteria pollutant as independent variables, a linear regression analysis was done. A linear regression analysis with subsets of the data was also performed. In subset analysis, log of the population was added as a covariate. Analyses revealed significant pollutants for stroke mortality. A similar analysis was done using Poisson regression since stroke mortality is a count variable. The population was used as a covariate in Poisson regression. Linear regression analysis with the entire data implied that air pollutant levels were not a good predictor of the stroke death rate. Linear regression analysis was performed with the subset of data in which, counties with stroke deaths less than ten were removed. Log of the population was added as a covariate. The slope associated with carbon monoxide pollutant was positive, which implied that as carbon monoxide pollutant increased by one ton per square mile, stroke death rate would increase by 0.1911. Linear regression analysis was performed with the second subset of data in which counties with stroke deaths less than 15 were removed. The slope associated with the carbon monoxide pollutant was positive, which implied that as carbon monoxide pollutant increased by one ton per square mile, stroke death rate would increase by 0.2434. Subset three was formed with all the counties in which the stroke deaths were equal to or less than 20. No significant variables were identified in subset three. Subset four was formed with all the counties in which the stroke deaths were greater than 20. The subset four model showed that as the volatile organic compound pollutant increased by one unit, the stroke death rate would increase by 1.201 units. Subset five was formed with all the counties in which stroke deaths were equal to or less than 25. Subset six was formed with all the counties in which stroke deaths were greater than 25. The subset five model did not yield any significant variable and the model was insignificant. The model six showed that if carbon monoxide increased by one unit, stroke death rate increased by 0.2269 units. A Poisson regression analysis was performed with full data. Number of stroke deaths was considered as the dependent variable. Carbon monoxide, sulfur dioxide, nitrogen oxide, PM 2.5, PM 10, volatile organic compounds pollutant level and total criteria pollutant were independent variables which were measured in tons per square miles. Log of the population was also considered as an independent variable, which acted as a covariate. The estimated model was the following: y (stroke death) = Exponential (-4.812 + 0.00328 * Volatile organic compounds + 1.75923 * Log of population + 0.00054 * Carbon monoxide). The slope associated with carbon monoxide was positive, which implies that as carbon monoxide increased by one ton per square mile, stroke death increased at the rate of 1.00054. The slope associated with volatile organic compounds was positive, which implies that as volatile organic compounds increased by one ton per square mile, stroke death increased at the rate of 1.00328. The carbon monoxide pollutant level was a significant factor in three linear regression models and the Poisson regression model. Therefore, higher carbon monoxide pollutant level would increase the risk of deaths from stroke. Volatile organic compounds pollution level was a significant factor in the Poisson regression model and subset four. Further studies should be done to understand the association between air pollutants and stroke death.
Description
Keywords
cerebrovascular disease, health, air pollution, Texas, stroke
Citation
Anand, S. (2004). Impact of air pollution on stroke incidence in Texas (Unpublished thesis). Texas State University-San Marcos, San Marcos, Texas.