Long-term mortality study of oil refinery workers: V. Comparison of workers hired before, during, and after World War II (1940-1945) with a discussion of the impact of study designs on cohort results

Occupational exposure to gasoline and the risk of non-Hodgkin lymphoma: a review and meta-analysis of the literature

AIMS

Gasoline comprises over 500 chemicals, including the known or suspected carcinogens benzene, 1,3-butadiene, ethylbenzene and methyl tert-butyl ether (MTBE). To assess whether work in the production, distribution and use of gasoline is associated with non-Hodgkin lymphoma (NHL), we reviewed the published literature on this topic.

METHOD

English-language peer-reviewed articles were identified by keyword searches of bibliographic databases. Twenty-two cohorts and thirteen case-control studies examined the risk of NHL among persons employed in the downstream petroleum industry.

RESULT

No positive associations were found with the exception of one study. The pooled risk estimate from a random-effects meta-analysis was 1.02 (95% confidence interval (CI) 0.94-1.12). Although there were no estimates available, exposure is likely to have varied by occupation, location and time period; there was no evidence however that risk estimates varied by any of these factors. NHL is a heterogeneous disease, yet no data were reported for NHL subtypes.

CONCLUSION

In summary, there is no suggestion across an extensive literature that exposure to gasoline at the levels workers' experience in an occupational setting increases the risk of NHL.

Incidence of lymphohaematopoietic malignancies in a petrochemical industry cohort: 1983-94 follow up

OBJECTIVES

In response to a previous finding of increased mortality from lymphohaematopoietic (LH) malignancies, this study examines incidence of LH malignancy in a petrochemical industry cohort. Emphasis is on chronic lymphocytic leukaemia (CLL) and on comparisons by period of first employment.

METHOD

The study cohort consists of 8942 employees who were active in the period 1970-92 and alive on 31 December 1982. Record linkage with the Louisiana tumour registry (LTR) provided information on cancer for cases occurring between 1983 and 1994. Standardised incidence ratios (SIR), with the south Louisiana population as a comparison, were computed for all cancers, all LH malignancies and specific LH subtypes. Analyses were conducted for sex and race categories, and by period of first employment, job type, duration of employment, and latency.

RESULTS

672 Cases of cancer were identified, including 59 LH malignancies. Women (n=1169) had an overall cancer SIR below unity and four LH malignancies versus 2.28 expected. Among the 7773 men, those first employed before 1950 had no overall cancer excess, a significant 1.4-fold increase in overall LH malignancies (43 observed versus 30.78 expected), and four CLL cases versus 3.27 expected. Findings for men first employed after 1950 are based on fewer cases, but there was no indication of excesses of overall cancer or LH malignancy. Numbers were too small in the group first employed after 1950 for meaningful analysis of LH malignancy subtypes such as CLL (one case).

CONCLUSION

These findings do not suggest a continuing excess of CLL but do suggest a small increase in incidence of overall LH malignancy for workers first employed before 1950. This may reflect associations with earlier workplace conditions, although work related patterns are mixed. Interpretation is limited by the diverse group of diseases within LH malignancies, and the lack of control for non-work factors other than sex, age, race, and period of diagnosis. This study has a major advantage of more complete and reliable cancer ascertainment compared with the mortality investigation, and shows the feasibility and benefits of using cancer registry incidence data in an occupational cohort study.

A 50-year mortality follow-up of a large cohort of oil refinery workers in Texas

To investigate further the possible role of occupational exposures on mortality, an update of a large Texas petroleum refinery cohort was undertaken. Between 1937 and 1987, 6799 deaths were identified among 17,844 employees. Relative to the general population of Texas, the overall standardized mortality ratio (SMR) showed a statistically significant deficit, as did nine other cause-of-death categories. Statistically significant mortality excesses were found for bone cancer (SMR = 207.8: 95% confidence interval [CI], 110.6 to 355.3), acute lymphocytic leukemia (ALL) (SMR = 259.6; 95% CI, 112.1 to 511.5), and benign/unspecified neoplasms (SMR = 194.9; 95% CI, 129.5 to 281.7). However, none of these diseases demonstrated an exposure-response relationship with length of employment. Subcohort mortality analyses by sex and race groups, length of employment, interval since hire, period of hire, and pay status were also performed. Overall, the update findings do not indicate that any excess mortality occurred as a result of employment at the refinery.

Mortality, morbidity, and haematological results from a cohort of long-term workers involved in 1,3-butadiene monomer production

A retrospective mortality analysis and prospective morbidity and haematological analyses were performed for Shell Deer Park Manufacturing Complex (DPMC) male employees who worked in jobs with potential exposure to 1,3-butadiene from 1948 to 1989. 614 employees qualified for the mortality study (1948-89), 438 of those were still employed during the period of the morbidity study (1982-9), and 429 of those had haematological data available for analysis. Industrial hygiene data from 1979 to 1992 showed that most butadiene exposures did not exceed 10 ppm (eight-hour time weighted average (8 hour TWA)), and most were below 1 ppm, with an arithmetic mean of 3.5 ppm. 24 deaths occurred during the mortality study period. For all causes of death, the standardised mortality ratio (SMR) was 48 (95% confidence interval (95% CI) = 31-72), and the all cancer SMR was 34 (95% CI = 9-87). There were only two deaths due to lung cancer (SMR 42, 95% CI = 5-151) and none due to lymphohaematopoietic cancer (expected = 1.2). Morbidity (illness absence) events of six days or more for the 438 butadiene employees were compared with the rest of the complex. No cause of morbidity was in excess for this group; the all cause standardised morbidity ratio (SMbR) was 85 (95% CI = 77-93) and the all neoplasms SMbR was 51 (95% CI = 22-100). Haematological results for the 429 with laboratory data were compared with results for the rest of the complex. No significant differences occurred between the two groups and the distributions of results between butadiene and non-butadiene groups were virtually identical. These results suggest that butadiene exposures at concentrations common at DPMC in the past 10-20 years do not pose a health hazard to employees.

Mortality of workers hired during World War II

There has been some suggestion that men first hired during World War II do not show the typical healthy worker effect and may have been more unhealthy than their counterparts who entered military service. We have studied 41,000 workers at six plants to determine whether men hired during World War II had higher mortality than men hired just before or after WWII. No evidence was found of any "unhealthy WWII worker" effect.

A retrospective cohort mortality study of blending and packaging workers of Mobil Corporation

This retrospective cohort mortality study examined 2,467 workers in lubrication products blending and packaging (B&P) operations at two refineries of Mobil Corporation between January 1, 1945 and December 31, 1978. Ninety-seven percent were male. Compared with U.S. males, there were significantly fewer deaths observed among males due to all causes, external causes, and diseases of the circulatory, respiratory, digestive, and genitourinary systems. Deaths observed from all cancer were fewer than expected, although not statistically significant. No statistically significant excess cause-specific mortality occurred at B&P facilities combined or separately. Nonsignificant increases in mortality were observed for cancers of the stomach, large intestine, prostate, the category of "other lymphatic tissue" cancer, and leukemia and aleukemia. Analyses demonstrated a statistically significant pattern of increasing SMR with employment duration for "other lymphatic tissue" cancer. Within the highest cumulative duration of employment category, the excess was confined to workers after 30 or more years since first employment. Although the interpretation of cancer mortality patterns is limited due to small numbers of deaths, the absence of associations with specific B&P departments is evidence against a causal interpretation.

Mortality patterns among petroleum refinery and chemical plant workers

A historical cohort study was conducted to evaluate the mortality experience of 6,831 employees of the Shell Oil Company, Deer Park, Texas, petroleum refinery and chemical plant with emphasis on cancer mortality. Subjects were all workers with potential plant exposure who were employed for at least 3 months during 1948-72. Vital status was determined as of 12/31/83 for 98% of the cohort and death certificates were obtained for 95.4% of 1,180 observed deaths. The statistical analyses excluded 159 female study members. For all causes of death combined, all cancers combined, and for most of the nonmalignant disease categories examined, there were deficits in mortality among refinery workers, chemical plant workers, and workers with experience in both areas. These deficits were generally most pronounced for chemical plant workers. An analysis of specific cancer sites revealed patterns of increased risk suggestive of a possible relationship between occupational exposures in the refinery and lympho-reticulosarcoma. Patterns of increased risk were also observed among chemical plant workers for a category of lymphopoietic tissue cancers, including multiple myeloma, myelofibrosis, polycythemia vera, and certain non-Hodgkin's lymphomas. Some very limited evidence of a possible workplace association was also found among refinery workers for leukemia and cancers of the central nervous system and biliary passages/liver. No evidence was found of an increased risk for cancer of the respiratory system or stomach or for malignant melanoma. A work history review of all suspect cancer excesses failed to identify any common work areas, job assignments, or exposure potentials, although the lack of detailed data on specific chemical exposures precluded accurate assessments of exposure-response.

Mortality study of cancer risk among oil refinery workers

The mortality experience of 1595 male workers employed in one of the largest Italian refineries in the period from 1949-1982 was examined. From the comparison with national and local death rates, increases in mortality owing to lung and kidney cancers, brain tumors, and leukemias emerged. No definite trends according to duration of exposure and years since first exposure were apparent. The increases regarding cancer of the lung, kidney and brain appeared to be associated with the early period of operations. Analysis by exposure category suggested an association of the increased mortality from leukemias with working in production (observed = 2; expected = 0.61). Kidney cancer mortality was elevated among maintenance workers (obs. = 2; exp. = 0.18). Small numbers prevented firmer conclusions. Workers in the moving department had a significantly increased mortality from all cancers (obs. = 22; exp. = 11.7), and lung cancer (obs. = 11; exp. = 3.6). Confounding by smoking could be excluded as sufficient explanation of the three-fold increase in lung cancer deaths. It was in moving that highest airborne levels of polynuclear aromatic hydrocarbons had been discovered in an independent environmental investigation.

Mortality comparisons of chemical workers hired before, during, and after World War II (1941-1945)

Cause-specific mortality was evaluated by period of hire for 37,682 male chemical workers in order to test the hypothesis that employees hired during World War II (1941-1945) were at increased risk for selected causes of death. One recent study of refinery workers reported that those hired during the war years had experienced greater mortality from external causes (accidents, homicides, and suicides), alcoholism, and cancers of several sites relative to employees hired before or after the war. In the present study, employees were divided into three period-of-hire subcohorts: prior to 1941, 1941-1945 (World War II), and 1946 and after. Comparison of observed mortality among these subcohorts through 1982 was made with expected levels based on age- and calendar year-specific U.S. white male rats. Neither hourly nor salaried employees hired during the war showed evidence of higher mortality from homicide, suicide, alcoholism, or any of the selected cancer types suggested from the refinery study. Hourly, but not salaried, war years new hires experienced excessive mortality from only those accidents involving motor vehicles. Possible reasons for the discrepant findings between this and the earlier study of refinery workers are discussed, with methodological differences being dismissed.

Health experience of workers in the petroleum manufacturing and distribution industry: a review of the literature

A review of 120 papers on the health effects on workers in the petroleum manufacturing and distribution industry revealed inconsistent results. This may be due in part to the variable quality of the epidemiological studies. There is, however, the consistent finding of a deficit for "all causes mortality" and for "all cancers mortality," thus implying that a widespread serious health effect of gasoline exposure seems remote. For specific health effects, the possibility exists that occupational etiological factors may play a part in the pathogenesis of brain cancer and renal disease. The evidence for a link between occupational and other health effects is weak. Further work of better quality is necessary to investigate those diseases in which there remains a suspicion of an occupational etiology.