Geographic Analyses of Bone Cancers and Fluoridated Water Supplies

The purpose of this research was to explore the relation between fluoride concentration in drinking water and incidence of bone cancers in humans. The hypothesis of this analysis is that there is geographic clustering of bone cancer incidence rates along with significant correlation between bone cancer and fluoride levels in drinking water. The null hypothesis is that there is no clustering and no significant correlation between bone cancer incidence and fluoridation levels. The study area consisted of all counties in the states of Iowa, Kentucky, Louisiana, New Mexico, and Utah. Bone cancer incidence rates and fluoride levels of water systems in the study area on a county level have been used for the study. The analysis consisted of visual examination to detect trends, cluster analysis using Local Index of Spatial Autocorrelation statistical methods, and univariate regression analysis to determine the significance of the relationship between fluoride and bone cancers.

The initial visual analyses of cancer incidence rates on choropleth maps show few geographic trends. There is an apparent national trend towards the southeastern part of the country. LISA analysis has shown no clustering in any of the states within the study area. Regression analysis shows no statistically significant correlation between cancer incidence and fluoride levels, with a R-squared value of 0.002. Based on these results, the null hypothesis of no geographic clustering and no statistical correlation must be accepted.