Cancer mortality in 180 countries of the world investigated through spatial autocorrelation models. Is there a clustered pattern of cancer?
Abstract
Abstract: Cancer is a leading cause of death worldwide. Purpose of the present study is to investigate the spatial autocorrelation of cancer deaths in 180 countries worldwide. It aims at answering whether neighboring countries tend to have similar cancer deaths or not and how these 180 countries could be clustered according to their spatial autocorrelation.
Methods: Data of total cancer deaths and cancer deaths for both genders per 100,000 people were selected from World Health Organization database. Cancer deaths were adjusted to age and selected for 180 countries worldwide. The whole analysis was carried out in the GIs Arc map and spatial statistics. Moran’s I was applied in order to detect the spatial autocorrelation.
Results: Positive spatial autocorrelation was proved as far as cancer mortality is concerned, worldwide. Total Moran’s I= 0.10, men’s Moran’s I= 0.13 and women’s Moran’s I = 0.05 (p=0.01). There is a spatial clustering of the 180 countries which measures their correlation and indicates that neighbouring countries tend to have similar values of cancer mortality.
Conclusions: This study imports new methods of analysis of spatial data in Epidemiology. Therefore, it highlights the possible applications of these methods as well as their suitability in capturing epidemiological issues.
Key words: cancer mortality, spatial autocorrelation, Global Moran’s I, spatial clusters, spatial epidemiology
Methods: Data of total cancer deaths and cancer deaths for both genders per 100,000 people were selected from World Health Organization database. Cancer deaths were adjusted to age and selected for 180 countries worldwide. The whole analysis was carried out in the GIs Arc map and spatial statistics. Moran’s I was applied in order to detect the spatial autocorrelation.
Results: Positive spatial autocorrelation was proved as far as cancer mortality is concerned, worldwide. Total Moran’s I= 0.10, men’s Moran’s I= 0.13 and women’s Moran’s I = 0.05 (p=0.01). There is a spatial clustering of the 180 countries which measures their correlation and indicates that neighbouring countries tend to have similar values of cancer mortality.
Conclusions: This study imports new methods of analysis of spatial data in Epidemiology. Therefore, it highlights the possible applications of these methods as well as their suitability in capturing epidemiological issues.
Key words: cancer mortality, spatial autocorrelation, Global Moran’s I, spatial clusters, spatial epidemiology
Full Text:
PDFRefbacks
- There are currently no refbacks.