To overcome the limited availability of individual-level socio-demographic information in the national vital statistics mortality database and to obtain self-reported racial/ethnic and socioeconomic data, the NCI?s Surveillance Research Program (SRP) will participate in the proposed extension and expansion of the U.S. National Longitudinal Mortality Study (NLMS). The NLMS is a continuous long-term intergovernmental study that follows prospectively several cohorts of American men and women (totaling approximately 3.1 million) drawn from the CPS between 1973 and 2002 2006 for their mortality experience from 1979 onwards. In the NLMS, which first started in 1983, data from death certificates on the fact of death (e.g., age at death and place of death) and the cause of death (ICD-9 and ICD-10 underlying cause-of-death codes) are linked to the individual-level socioeconomic and demographic data from the Current Population Surveys (CPS) by means of the National Death Index (NDI). The NDI, a listing of all deaths in the US since 1979, is maintained by NCHS. The NLMS-Phase I has a population base of 2.42 million persons, whose mortality follow-up from 1979 through 1998 has yielded approximately 256,000 deaths, including 60,000 deaths from cancer. The NLMS is currently the largest federal study that permits examination of socioeconomic and occupational inequalities in cause-specific mortality in the United States. The Phase II component of the NLMS calls for the inclusion of eight additional CPS cohorts (including the 1999 and 2006 CPS Tobacco Use Supplements) and the extension of the mortality follow-up by eight years to include all deaths through 2006. The Phase II database development is expected to be completed in 2010. The expansion would bring the total number of individuals being followed in the NLMS to 3.6 million and the identified mortality to 495,000 deaths from all causes and approximately 120,000deaths from cancer over the 28-year follow-up period, 1979 to 2006. The objective of this multi-year record linkage project is to use the many additional socioeconomic and demographic variables available from the CPS, to eliminate numerator/denominator differences that characterize vital statistics based mortality rates, and to estimate differentials in cancer mortality by major sites, according to the preferred self-reported race/ethnicity, marital status, living arrangement, educational attainment, income, occupation, industry, residence, nativity/immigrant status, smoking, health status, and availability of health insurance.
