Regional Cancer Clusters Across Northern Europe: A Geographic Analysis Using Shift Share, Spatial Autocorrelation, and Multiple Regression
Date
Authors
Sanchez, Michael
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
There is perhaps no disease in the modern era as geographically ubiquitous as cancer. It proliferates across borders and ecosystems to every corner of the planet. Despite the universal proliferation of the disease, there is some geographic variability in the type of cancer that spreads. Liver cancer, for example, has a low rate of incidence in the United States, but is one of the highest incidence cancers in parts of the developing world such as Southeast Asia (Singh 2015). In this thesis, I examine the possibility of commonalities or even clusters of cancer incidence across borders using the statistical methods of Moran’s I and Getis Ord Gi*. At the same time, I introduce a new application of shift share analysis to breakdown cancer incidence. Additionally, regression analysis illustrates the shifting relationship of increased cancer incidence, and socio-behavioral/healthcare variables. Regions of Denmark, Southern Sweden, and Wales were found to be central to several high incidence clusters including melanoma and bladder cancer. Unsurprisingly, regions closer to the artic (ie: those in northern Scandinavia) formed clusters of low incidence for Melanoma due to lower sunlight exposure. The application of shift share analysis to cancer incidence yielded a variety of noteworthy results such as a gender disparity in the average increase of cancer between men (1.55%) and women (6.35%). Finally, a gender disparity was also revealed for pancreatic cancer in which disease and alcohol consumption was shown to be significantly correlated with high incidence in men but not in women.
Description
Keywords
Cancer, Geography, Cancer clusters, Spatial autocorrelation, Europe, Shift share