Geographic concentrations of prostate cancer incidence in Texas and their relationship to socioeconomic conditions

Abstract

<p>Prostate cancer is a dangerous and elusive disease. The disease is the most diagnosed cancer and the second highest leading cause of death from cancer in Texas men. Although prostate cancer is highly prevalent in male populations worldwide as well as in Texas, the environmental risk factors for the disease are relatively unknown. For these reasons, research about the environmental risk factors for prostate cancer is of great importance and urgency.</p> <p>This thesis addresses four research questions: (1) Are there any statistically significant spatial clusters of prostate cancer incidence in Texas at the census tract level?; (2) is there any statistically significant association between prostate cancer incidence clusters and socioeconomic status at the census tract level in Texas?; (3) is there any significant association between prostate cancer incidence clusters and rural place of residence in Texas?; and (4) is socioeconomic status a significant risk factor for prostate cancer incidence in Texas at the census tract level? To answer these questions, several research objectives were met.</p> <p>The first objective was to investigate the geographic distribution of prostate cancer incidence in Texas at the census tract level. This objective was achieved using a spatial scan statistic cluster test developed by Kulldorff (1997). The scan statistic was applied to prostate cancer incidence data from the Texas Cancer Registry of the Texas Department of Health and population data from the U.S. Census for three race-ethnicity categories (white, black, and Hispanic), and four age groups (18-24, 25-44,45-64,65+). A statistically significant most likely cluster was detected in north San Antonio and 23 significant secondary clusters were observed in other sections of Texas.</p> <p>The second objective was to investigate socioeconomic status and place of residence as risk factors for prostate cancer incidence clusters and prostate cancer incidence rates. Logistic regression and Poisson regression analyses were employed to reach this objective. Median household income and census tract education were chosen as the indicators of socioeconomic status and people per square mile was the indicator of rural or urban areas. High income, high education, and urban place of residence were all found to be significant predictors of counties containing prostate cancer incidence clusters. Further, high income, high education, and urban place of residence were found to be risk factors for prostate cancer incidence rates independent of clusters. In more precise terms, higher socioeconomic status and living in an urban area make one more prone to developing prostate cancer.</p> <p>This research contributes to the established literature of environmental spatial analysis and spatial epidemiology. For the first time, this research demonstrated that: (1) there is a significant cluster of prostate cancer incidence in northern San Antonio at the census tract level in Texas, and (2) socioeconomic status and urban place of residence are risk factors in both prostate cancer incidence clusters and prostate cancer incidence rates at the census tract level in Texas. Medical researchers and public health planning officials may benefit from this research by using the results found here to further focus prostate cancer research.</p>