| This analytic data file set consists of seven files generated for the Chemical Laboratory Workers (CLWs) cohort mortality study by Travis Kubale, National Institute for Occupational Safety and Health, and published in American Journal of Industrial Medicine in 2008. The study cohort consists of chemical laboratory workers at Oak Ridge facilities (K-25, X-10, and Y-12) and the Savannah River Plant in South Carolina. |
| The study evaluates the mortality experience of 6,157 male and female workers employed from 1943 to 1998 at the DOE Oak Ridge, Tennessee, facilities (K-25, X-10, and Y-12) and the Savannah River Plant in South Carolina. Chemicals generally encountered by these laboratory workers are the primary exposures of interest. |
| The study cohort consists of 6,157 CLWs employed for at least 1 day at SRS or the Oak Ridge sites (X-10, Y-12, and K-25 facilities) from January 1, 1943, through December 31, 1998. |
| Follow-up for each worker began with either the start of facility operations (i.e., 1943 for Oak Ridge workers or 1952 for SRS workers) or the first date employed at a study facility, whichever was later, and ended on date of death or December 31, 1998, whichever was earlier. The primary data sources used to ascertain deaths were the National Death Index-Plus (NDI-Plus) and the Social Security Administration Death Master File (SSA-DMF). In addition, the SSA “presumed living” file was used to confirm that workers not found to be deceased by NDI-Plus were alive as of December 31, 1998. |
| Underlying and contributing causes of death were obtained directly from NDI-Plus for workers deceased in 1979 or later. For those who died before 1979, death certificates were obtained from the states in which death occurred. |
| CLWs were also potentially exposed to ionizing radiation while performing duties both in the laboratory and outside the laboratory in process areas. Radiation exposure information for each worker was abstracted from available site exposure records and NIOSH databases maintained from previous studies. External occupational radiation exposure was evaluated as equivalent dose in mSv to the whole body. |
| Standardized Mortality Ratio Analysis. |
| The NIOSH Personal Computer Life Table Analysis System (PC-LTAS, version 1.0d) was used to generate expected numbers for all deaths, all cancer deaths, and causespecific deaths, by race and sex, within 5-year age and 5-year calendar time periods [NIOSH, 2001]. Person-years and observed deaths were accumulated for each of these age and calendar time periods from the worker’s follow-up begin date through the end of follow-up. |
| Expected numbers of deaths were based on U.S. population death rates specific for the race, gender, and 5-year age and calendar time periods. United States expected rates were based on deaths from 1940 forward except for non-Hodgkin’s lymphoma and multiple myeloma, causes for which rates were available only from 1960 forward. For these outcomes, enumeration of observed deaths and person-years at risk began in 1960 or the first date employed at a study facility, whichever came last. Because differences between mortality among members of the study population and that of the United States population could be, in part, the result of regional (non-occupational) differences, mortality for selected cancers was compared with Tennessee state rates, which were also available from 1960 forward. Numbers of deaths observed for each cause were divided by the expected number of deaths to obtain cause-specific SMRs. The precision of each estimated SMR was assessed assuming a Poisson distribution, with two-sided 95% confidence intervals. |
| Internal Analysis |
| Because SMRs are affected by the age structure of the study population, comparison of SMRs for different populations can be misleading. SRRs allow for comparison among populations by weighting observed stratum-specific rates according to a common (internal) standard (Rothman and Greenland, 1998). SRR analyses were performed for the entire cohort to assess the associations between mortality outcomes and employment duration. Duration of employment categories of 0 to <5 years employed, 5 to <10 years employed, 10 to <20 years employed, and 20+ years employed were used in the analyses. For dose-response analysis, a linear trend was calculated in a person-year-weighted regression of directly standardized rates (Rothman and Greenland, 1998). Statistical significance of each trend was determined using a two-tailed z-test with an alpha of 0.05. |
| Modeling Analysis |
| Cox regression analyses were conducted to further evaluate the relation between occupational exposure and the risk of mortality due to selected cancers. Risk sets were defined from the entire cohort of CLWs for each case of interest. Each risk set also included all controls who were under observation and who lived to an age equal to or greater than the age of the case at death. Time was defined as the length of life (attained age) of the case. The cases and controls were matched on attained age because it is a strong predictor of cancer mortality and therefore an important potential confounder of the association of interest in the analyses. All occupational exposures of the controls were truncated at the age attained by the case minus any lag. |
| Conditional logistic regression was calculated using the SAS procedure PHREG. This procedure performs a regression analysis of survival data, the results of which are identical to those provided by the Cox proportional hazards model (SAS Inc., 1999). The probability of death from lymphatic and hematopoietic cancers was modeled as a function of the following candidate variables: duration of employment (surrogate for chemical exposures-the primary exposure of interest), external ionizing radiation exposure, gender, and employment in a process area. A change of plus or minus 10% in the parameter estimate of the primary exposure variable was used as an indicator of confounding in determining the final model. Lag periods of 2 years for leukemia, 10 years for multiple myeloma, and 5, 10, and 20 years for lung cancer were evaluated. Because latency and risk factors may be different for CLL, the leukemia analysis was performed both with and without CLL deaths. |
| The data file set consists of the following files: |
| Work History and Demographic files used in the Life Table Analysis System (LTAS) analysis. |
| Both a work history and demographic file were used in all LTAS analyses, i.e., SMR and SRR analyses. The work history file is MFMCHMA1_d1 and the demographic file is MFMCHMA1_d2. |
| The file, MFMCHMA1_d3, was the source file for the LTAS analysis. |
| There are four files that were used for the regression analyses. |
| For Leukemia, MFMCHMA1_d4 was used for analysis with a duration of employment lagged 2 years. |
| Multiple myeloma was analyzed using MFMCHMA1_d5 with a duration of employment lagged 10 years. |
| Multiple Myeloma re-analysis (MFMCHMA1_d6) was performed with a duration of employment lagged 10 years with individuals employed for at least 1 year. |
| Lung Cancer analysis (MFMCHMA1_d7) was done with a duration of employment lagged 10 years. |
