Kliewer E V
National Centre for Epidemiology and Population Health, Australian National University, Canberra, ACT.
Int J Epidemiol. 1992 Jun;21(3):442-9. doi: 10.1093/ije/21.3.442.
In order to examine the impact of migrants on regional variations in stomach and colon cancer, standardized mortality ratios (SMRs) were calculated for the total, nonmigrant (born and dying in same state), and migrant (born out of state) White residents of each of the 11 western states in the United States (US). The SMRs were derived from the National Center for Health Statistics' Mortality Detail Files for 1979-1981 and the 1980 Census Public Use Microdata 5-Percent Sample tapes. Migrants in the western US accounted for 79% of all stomach and colon cancer deaths. There was no consistent relationship between the SMRs of migrants and nonmigrants, with the migrant SMRs being higher in some states and lower in others. As a consequence of this differential impact, and their substantial numbers, migrants obscured the underlying regional patterns of mortality risk observed in the nonmigrants. The states of high or low risk were more contiguous in the analysis based on nonmigrants than the total population, and the interstate ranges in mortality were greater for nonmigrants. In areas with high in-migration, mortality atlases based on the total population may give an inaccurate portrayal of regional mortality risks, and spurious correlations may arise between the distributions of diseases and environmental characteristics of the regions. Regional mortality patterns of nonmigrants may more precisely reflect the factors which are influencing these cancers and thus provide a greater potential in providing clues to their aetiologies.
