REGIONAL CONVERGENCE OF CANCER MORTALITY RATES OVER TIME IN THE UNITED STATES, 1940–1960

Abstract

Waggoner, D. E. (National Cancer Institute, NIH, Bethesda, Md. 20014) and Newell, G. R. Regional convergence of cancer mortality rates over time in the United States, 1940–1960. Amer J. Epidem 93: 79–83, 1971.—High negative correlations of mortality from several cancer sites with environmental temperature among regions of the United States along with equally high but positive correlations for the same cancer sites with per capita income and numbers of physicians per unit population have been shown. This implied that much of the geographic differences in cancer mortality was associated with data artifacts and indirect causative factors rather than reflecting a direct meteorological effect on cancer risk. This study confirms the initial impression by demonstrating that relative variance of mortality rates among regions of the United States declined for all cancer sites from 1940 to 1960 regardless of whether the cancer mortality rates were increasing or decreasing, or whether the correlation of mortality with temperature was changing. Among those cancer sites for which correlations with temperature had decreased from 1940 to 1960, their regression slopes had diminished as well. Further, the regions with the lowest cancer death rates in 1940 had the largest per cent increase by 1960 and conversely for regions with higher rates in 1940. These findings suggest that among regions of the United States some cancer mortality rates are in the process of converging toward national averages.