Patterns of ovarian cancer and uterine cancer mortality and incidence in the contiguous USA

Abstract

The main objective is to investigate the geographical variation in ovarian cancer and uterine cancer mortality, and to test associations between some risk factors and these cancer types in the contiguous US for mortality and for incidence. The modern disease surveillance software SaTScan™ was used for a spatial cluster analysis to assess any observable geographical variation in ovarian and uterine cancer mortality rates and to identify and test for spatial clusters with elevated relative risk. The analyses were first completed using age adjusted cancer rates for ovarian cancer and for uterine cancer. The cancer data was then adjusted for the risk factors (or covariates) obesity rate, smoking rate, urban, poverty rate, college education rate, race, opioids mortality rate, and for arsenic intake from well water rate. All used data for cancer mortality were for 2000-2014 while incidence data were for 2011-2015. There exist seven significant mortality clusters of ovarian cancer, with large clusters in NW, NE and SE of the US, and there exist two large mortality clusters of uterine cancer in NE and Central US. Most risk factors studied for mortality and for incidence were significant at significance levels much lower than 0.05 for either of the two cancer types, except race for ovarian cancer mortality and arsenic for ovarian incidence. This study has identified several important factors, and these findings could be used for a more effective search for cancer prevention for uterine and ovarian cancer. LIMITATIONS OF THE STUDY: The accuracy of the data could not be controlled as data were downloaded from websites. While the mortality data was complete, the incidence data had counties with missing data. The data were obtained at the county resolution. No data were available on women who had one type of cancer and then had the second type of cancer later in life. Only purely spatial clusters were studied and no temporal analysis was done.

Keywords: Cancer cluster; Cluster analysis; Epidemiology; United States.