2.1. Cohort studies (Tables 2.1 and 2.2)

Table 2.1

Overview of cohort studies of firefighters.

Table 2.2

Cohort studies of cancer among firefighters.

Among municipal firefighters, five studies evaluated the incidence of cancer and 15 assessed mortality – some with multiple reports. None of these was adjusted for smoking, or other potential confounders.

Mastromatteo (1959) conducted a cohort mortality study of all firefighters employed with the city fire department in Toronto, Ontario, Canada during 1921–1953. The cohort consisted of 1039 active and retired firefighters. A total of 325 firefighters (31%) were lost to follow-up. The followed cohort accrued 25 918 person–years of observation; 271 deaths were recorded. Causes of death were determined by the examination of death certificates. Comparison was made to age-specific death rates for the Province of Ontario. Because most of the firefighters were urban dwellers, the author calculated death rates for the City of Toronto. There was no excess mortality from cancer at all sites combined (34 observed, 30 expected), and no site-specific analysis of cancer mortality.

Musk et al. (1978) conducted a cohort mortality study in 5655 firefighters with three or more years of service in Boston, USA, during 1915–1975. Firefighters were identified from employment records. The observed cause of death as stated on the death certificates of 2470 deceased firefighters was compared with expected numbers based on rates for the male population of Massachusetts. Cancer risks were only presented by organ system, and no statistically significant increases were seen.

Eliopulos et al. (1984) conducted a cohort mortality study of all males employed as full-time firefighters by the Western Australian Fire Brigade during 1939–1978. The cohort consisted of 990 male firefighters and follow-up was 98.3% complete through examination of death certificates. The cohort accrued a total of 16 876 person–years of follow-up; 116 deaths were recorded. Mortality from all causes was less than expected (standardized mortality ratio [SMR], 0.80; 95% confidence interval [CI]: 0.67–0.96). For those firefighters already employed at the study start date, mortality was a little higher (SMR, 0.84) than for those who started later (SMR, 0.74). There was no tendency for rates to rise with increasing duration of employment. SMR for all malignancies was 1.09 (95% CI: 0.74–1.56). Standardized proportional mortality ratios (SPMRs) were calculated for cancers, primarily by organ system, and no statistically significant excesses were seen.

Vena and Fiedler (1987) examined all full-time employees of the City of Buffalo, USA, who worked at least 5 years during 1950–1979. A total of 1867 Caucasian male firefighters employed for at least one year as a firefighter were studied. Vital status was determined for 99% of the cohort, resulting in 470 observed deaths. Significantly elevated SMRs were found for benign neoplasms (SMR, 417), cancer of the colon (SMR, 183), and cancer of the bladder (SMR, 286). Cause-specific mortality was presented by the number of years employed, calendar year of death, year of hire, and latency. Cancer mortality was significantly higher in the long-term firefighters, and risk of mortality from all malignant neoplasms tended to increase with increasing latency. Statistically significant excesses of colon and bladder cancer were observed among firefighters employed for 40 or more years.

Beaumont et al. (1991) calculated mortality rates for 3066 firefighters employed during 1940–1970 at the San Francisco Fire Department, USA. Vital status was ascertained through to 1982, and observed and expected rates were computed using United States death rates. About 3% of the population was lost to follow-up. Mortality was examined by duration of employment as a firefighter. The total number deceased (1186) was less than expected (risk ratio [RR] = 0.90), and there were fewer cancer deaths than expected (RR = 0.95). However, there were significant excess numbers of deaths from oesophageal cancer (12 observed, six expected). A statistically significant excess of biliary and related cancer was observed among firefighters employed for 30 or more years.

Grimes et al. (1991) conducted a proportionate mortality study involving all male firefighters with at least one year of service in the fire department of the City of Honolulu, USA. The observed percentage of firefighter deaths from each cause from 1969–1988 was compared statistically to the expected numbers of deaths for all males aged over 20 years in Hawaii's general population. The proportionate risk ratio (PRR) for all malignant neoplasms was 1.19 (95% CI: 0.96–1.49). Significant increases in risk of death were found for brain cancer (PRR, 3.78), prostate cancer (RR, 2.61), and cirrhosis of the liver (PRR, 2.3). [The Working Group noted that it does not appear as though PRRs were standardized by age and calendar period as is standard practice for this type of analysis.]

Heyer et al. (1990) examined the mortality among 2289 firefighters from Seattle, Washington, USA employed during 1945–1980. Subsequently, Demers et al. (1992a) examined the mortality of 4546 firefighters who were employed by the cities of Seattle and Tacoma (Washington, USA), and Portland (Oregon, USA) for at least one year during 1944–1979. Demers et al. (1992b) also examined the cancer incidence in 4528 firefighters from Seattle and Tacoma during 1944–1979. Mortality in these firefighters was compared to United States national mortality rates and to mortality rates of police officers from the same cities. Mortality was examined by the duration of employment as a firefighter (i.e., actually controlling fires) rather than as an inspector or a support person. This mortality was then compared to a reference group of police from the same cities. Complete follow-up was achieved for 98% of the firefighters. During 1945–1989 (the cohort was the same as Demers et al. [1992a] but the follow-up lasted until 1989), 1169 deaths occurred in the study population, and 1162 death certificates (99%) were collected. There was no excess risk of overall mortality from cancer. Excesses of brain tumours (SMR, 2.1; 95% CI: 1.2–3.3) and lymphatic and haematopoietic cancers (SMR, 1.3; 95% CI: 0.9–1.8) were found. Younger firefighters (< 40 years of age) showed an excess risk of cancer (SMR, 1.45; 95% CI: 0.8–2.39), primarily due to brain cancer (SMR, 3.75; 95% CI: 1.2–8.7). The risk of lymphatic and haematopoietic cancers was greatest for men with at least 30 years of exposed employment (SMR, 2.1; 95% CI: 1.1–3.6), especially for leukaemia (SMR, 2.6; 95% CI: 1.0–5.4).

Demers et al. (1994) further examined the incidence of cancer in a subcohort of 2447 male firefighters who were employed for at least one year during 1945–1979 in Seattle and Tacoma, who were still alive on January 1^st 1974. Incident cancer cases were ascertained through the Cancer Surveillance System of the Fred Hutchinson Cancer Research Center, a population-based tumour registry. The follow-up period was from 1974 to 1989. Cancer incidence in firefighters was compared with local rates and with the incidence among 1878 policemen from the same cities. The overall risk of cancer among firefighters was found to be similar to that of both the police (SIR, 1.0; 95% CI: 0.8–1.3) and the general male population (SIR, 1.1; 95% CI: 0.9–1.2). No excesses were observed for the most common organ sites. An elevated risk of prostate cancer was observed relative to the general population (SIR, 1.4; 95% CI: 1.1–1.7), but was less elevated compared with rates in policemen (incidence density ratio [IDR], 1.1; 95% CI: 0.7–1.8), and was not related to duration of exposure. The risk of colon cancer, although only slightly elevated relative to that of the general population (SIR, 1.1; 95% CI: 0.7–1.6) and the police (IDR, 1.3; 95% CI: 0.6–3.0), appeared to increase with duration of employment.

Giles et al. (1993) conducted a cancer incidence study of 2855 male firefighters employed by the fire brigade in Melbourne, Australia, during 1917–1988. All were operational personnel, who would more than likely have been called to combat fires. The follow-up period was from 1980 to 1989, and was estimated to have been 95% complete. To determine cancer incidence during the follow-up period, fire brigade employment records were linked to the Victorian Cancer Registry. SIRs were calculated by the direct method using the population of the State of Victoria as the reference group. The cohort accrued a total of 20 853 person–years, and 50 firefighters developed cancer during the period of observation. The SIR for all cancer sites and all ages combined was 1.13 (95% CI: 0.84–1.48). For firefighters under the age of 65 years, the all-site SIR was 0.84 (95% CI: 0.56–1.20); for those above 65 years of age, the all-site SIR was 2.14 (95% CI: 1.32–2.37). The only site-specific cancer that was elevated in the age group of 65 and older was colorectal cancer, with an SIR of 3.65 (95% CI: 1.13–7.94). The SIR for all other cancers in the age group 65 and above after removing colorectal cancer remained elevated, with a residual SIR of 1.83 (95% CI: 1.03–3.02).

Guidotti (1993) examined the mortality by cause of death for two cohorts totaling 3328 firefighters active during 1927–1987 in Edmonton and Calgary, Alberta, Canada. Associations were examined by cohort (before and after the 1950s) and by years of service weighted by exposure opportunity. The study attained 96% follow-up of vital status and over 64 983 person–years of observation; 370 deaths were recorded. Excesses were observed for all malignant neoplasms (SMR, 1.3; 95% CI: 1.0–1.6), and for cancers of the lung (SMR, 1.4; 95% CI: 0.9–2.1), bladder (SMR, 3.2; 95% CI: 0.9–8.1), kidney and ureter (SMR, 4.1; 95% CI: 1.7–8.5), colon and rectum (SMR, 1.6; 95% CI: 0.9–2.7), pancreas (SMR, 1.6; 95% CI: 0.5–3.6), and leukaemia, lymphoma and myeloma (SMR, 1.3; 95% CI: 0.6–2.3). The lung cancer excess was confined to Edmonton; there was no consistent association with duration of employment, exposure opportunity, or decade of entry into the cohort (before or after the 1950s) except that the highest risk was observed among Edmonton firefighters with over 35 weighted years of service. Urinary tract cancer excess was observed mostly among firefighters entering service after 1950, and appeared to increase with the length of service and exposure opportunity, and was observed in both cities.

Aronson et al. (1994) conducted a retrospective cohort mortality study of all male employees of the six fire departments within metropolitan Toronto, Ontario, Canada (n = 5995). The study population consisted of all male firefighters who had worked for at least 6 full months in metropolitan Toronto at any time during 1950–1989. Mortality was ascertained through computerized record linkage and compared to that of the male Ontario population specific to cause, age, and calendar period during 1950–1989. The cohort accrued 114 008 person–years and the average duration of follow-up was 21 years. Mortality was examined by duration of exposure. The SMR for all malignant neoplasms was 105 (95% CI: 91–120), for brain tumours, 201 (95% CI: 110–337), and for “other” malignant neoplasms, 238 (95% CI: 145–367). Non-significant increases in risk were observed for some other sites, in particular rectum (SMR, 171), larynx (SMR, 140), and testis (SMR, 252).

Tornling et al. (1994) conducted a cohort mortality study of all male fire fighters employed for at least 1 year in the City of Stockholm, Sweden during 1931–1983 (n = 1116). The population was identified from annual employment records. Follow-up for mortality was from 1951 until 1986, and for cancer incidence from 1958 to 1986. Except for four persons who had emigrated from Sweden, follow-up was 100% complete. To assess the occupational exposure as a firefighter, an index of participation in number of fires was calculated for each individual based on the number of reports on all fires in Stockholm that had been maintained since the beginning of the twentieth century. The all-site cancer mortality in 1958–1986 was equal to the expected (SMR, 100; 95% CI: 83–119). An excess of stomach cancer incidence (SIR, 192; 95% CI: 114–304; 18 observed versus 9.37 expected) was observed. There was also a tendency for higher incidence and mortality in stomach and brain cancers with increasing number of fires. Four brain cancer cases were observed compared to 0.8 expected (SIR, 496; 95% CI: 135–1270) in the highest exposure category.

Deschamps et al. (1995) investigated all professional male members of the Brigade des Sapeurs-Pompiers de Paris (n = 830) who served for a minimum of 5 years as of January 1^st, 1977. They were monitored for a 14-year period, with follow-up terminating on January 1^st, 1991. Cause-specific mortality rates in these firefighters were compared with national mortality data provided by the Institut National de la Santé et de la Recherche Médicale. To assess the occupational exposure as a firefighter, data were collected on duration of employment as an active duty firefighter (as opposed to office work). These 830 firefighters accrued a total of 11 414 person–years of follow-up. Follow-up appears to have been 100% complete. There were 32 deaths in the cohort during the 14-year period of follow-up. When compared to the average French male, they were found to have a far lower overall mortality (SMR, 0.52 [95% CI: 0.35–0.75]). None of the cause-specific SMRs was significant. However, a greater number of deaths than expected was observed for genito-urinary cancer (SMR, 3.29) [based on one bladder cancer, and one testicular cancer], and digestive cancer (SMR, 1.14).

Baris et al. (2001) conducted a retrospective cohort mortality study among 7789 firefighters in Philadelphia, Pennsylvania, USA, on males employed during 1925–1986. Vital status was ascertained up until 1986. SMRs and 95% CI were calculated with expected numbers of deaths in the United States white male population, as the overwhelming majority of firefighters were white. Occupational exposure histories were abstracted from detailed records maintained by the Philadelphia Fire Department, and a job–exposure matrix was created for each firefighter. To estimate exposure–response relationships, the study used this matrix to compare mortality among groups of firefighters defined by the estimated number of career runs. There were 2220 deaths and a total of 6.2% of the cohort was lost to follow-up. In comparison with white males in the United States, firefighters had a similar mortality from all causes of death combined (SMR, 0.96), and all cancers (SMR, 1.10). Statistically significant excess risks were observed for colon cancer (SMR, 1.51). The risks of mortality from colon cancer (SMR, 1.68), kidney cancer (SMR, 2.20), non-Hodgkin lymphoma (SMR, 1.72), multiple myeloma (SMR, 2.31), and benign neoplasms (SMR, 2.54) were increased in firefighters with at least 20 years of service.

Bates et al. (2001) conducted a historical cohort study of mortality and cancer incidence in all remunerated New Zealand firefighters, who served during 1977–1995. Ascertainment of employment was through a registry maintained by the United Fire Brigades Association of New Zealand. The final cohort comprised 4221 male firefighters. To assess the occupational exposure as a firefighter, data were collected on duration of employment. The 4221 male firefighters in this cohort accrued a total of 58 709 person–years of follow-up. Follow-up was successful in tracing 93.5%. There were 117 deaths up until 1995. Cancer incidence was ascertained during 1977–1996. The SIR for all cancers was 0.95. For most sites, no excesses were observed. The only cancer for which this study provided evidence of an increased risk was testicular cancer. Eleven testicular cancers were observed versus 7.1 expected (SIR, 1.55; 95% CI: 0.8–2.8). For the years 1990–1996, the SIR for testicular cancer was 3.0 (95% CI: 1.3–5.9).

Ma et al. (2005) examined age- and gender-adjusted mortality rates of 36 813 professional firefighters employed during 1972–1999 in Florida, USA, and compared those with that of the Florida general population. The study population consisted of 34 796 male and 2017 female professional firefighters. The racial/ethnic composition was caucasian (90.1%), hispanic (7%), and black (6.5%). Employment as a firefighter was ascertained from employment records in the Florida State Fire Marshall Office. Surrogate information on occupational exposures in firefighting was collected by examining the year of certification and duration of employment as a firefighter. No information was collected on smoking histories. A total of 1411 male and 38 female deaths with known causes were identified in this cohort. In male firefighters, a deficit of overall mortality from cancer was observed (SMR, 0.85). Excess risks were observed for male breast cancer (SMR, 7.41; 95% CI: 1.99–18.96), and thyroid cancer (SMR, 4.82; 95% CI: 1.30–12.34), each based on four cases. Mortality from bladder cancer was increased and approached statistical significance (SMR, 1.79; 95% CI: 0.98–3.00). Female firefighters had similar overall cancer mortality patterns to Florida women (SMR, 1.03), but the numbers were small for specific cancer sites.

In a further analysis of the same cohort, Ma et al. (2006) determined the relative cancer risk for firefighters in the State of Florida compared with the Florida general population. Employment as a firefighter was ascertained from employment records in the Florida State Fire Marshall Office. Cancer incidence was determined through linkage to the Florida Cancer Data System, a statewide cancer registry estimated to capture 98% of cancers in Florida residents. No pathological verification of cancer diagnoses was undertaken. A total of 970 male and 52 female cases of cancer were identified; 6.7% of the cohort were lost to follow-up. Male firefighters had significantly increased incidence rates of cancers of the bladder (SIR, 1.29; 95% CI: 1.01–1.62), testis (SIR, 1.60; 95% CI: 1.20–2.09), and of the thyroid (SIR, 1.77; 95% CI: 1.08–2.73). Female firefighters had significantly increased incidence rates of overall cancer (SIR, 1.63; 95% CI: 1.22–2.14), cervical (SIR, 5.24; 95% CI: 2.93–8.65) and thyroid cancers (SIR, 3.97; 95% CI: 1.45–8.65), and Hodgkin disease (SIR, 6.25; 95% CI: 1.26–18.26).

2.2. Case–control studies

Case–control studies have been used to examine the risk of firefighting and its association with various types of cancers. In all but one of these studies, ten or fewer firefighters were included in the case and/or control group. Several studies combined broad occupational categories with heterogeneous exposures such as firefighter and fireman, with the latter not necessarily working as a firefighter. These types of studies may result in exposure misclassification. Even within specific occupational groups such as firefighters, all would not have the same intensity or type of exposures. The magnitude of this form of misclassification is unknown, but it is likely that the resulting misclassification will be non-differential with regard to cases and controls. Another limitation to case–control studies is that cases may be more likely than controls to remember jobs of shorter duration. Those jobs in the more distant past may be more likely recalled by cases than controls resulting in differential bias away from the null. Alternatively, in several of the reported studies, cases were more likely than the controls to provide proxy interviews by their survivors rather than by the cases themselves. Because of the relatively few studies available for individual organ sites, the studies were grouped into four categories including urogenital, brain and central nervous system, larynx and lung, and other.

2.2.1. Cancers of the urogenital system

Four cancers of the urogenital organs in relation to employment as a firefighter were considered (Tables 2.3 and 2.6).

Table 2.3

Case–control studies of the urogenital system.

Table 2.6

Case–control studies of other or multiple sites.

Delahunt et al. (1995) examined pathologically confirmed incident renal cell carcinomas the New Zealand Cancer Registry during the period 1978–1986. The registry included 95% of those patients diagnosed and treated in both the public and private sector. At time of registration, the current or most recent occupation was recorded. Additional information collected included age, and smoking habits. Renal cell carcinomas with an ICD-9 code of 189.0 (malignant neoplasm of the kidney, excluding the renal pelvis) were evaluated. The control groups were a random sample of registrations drawn from all cases over 20 years of age, having primary tumours from sites other than the urinary tract registered during the same time period. There were a total of 710 male cases and 12 756 controls. There were 52 cases and 737 controls under the occupational classification of “Service” which included firefighters and five other occupational groups. The relative risk for firefighters was 4.7 (95% CI: 2.5–8.9).

Bates (2007) (see Table 2.6) conducted a registry-based case–control study using the California Cancer Registry. Anonymized records of all male cancers for the period 1988–2003 were collected. To identify firefighters, the occupation and industry fields were searched for titles including fire, firefighter, fire fighter, fireman, fire man, and fire chief. If the subjects indicated that they did not carry out firefighting activities, they were not considered. A total of 16 cancer organ sites were examined including kidney, bladder, prostate, and testis. For each analysis, all other cancers were used as controls except for those cancers shown in the initial analysis that had demonstrated a firefighting etiology; these included cancers of the lung, bronchus, bladder, prostate, colorectum, and skin melanomas. Analysis was limited to males aged 21–80 at time of diagnosis. There were 3659 firefighters and 800 448 controls in the analysis after exclusion of 13% of the files (n = 140 000) with no recorded occupation or industry. Logistic regression analyses were performed for each cancer type for which there had been more than 50 cancers recorded in firefighters. This was not done for for cancer of the thyroid (n = 32 cases) or multiple myeloma (n = 37 cases) as these two were based on prior hypotheses.

Logistic regression analyses adjusted for 5-year age categories, 4-year categories from date of diagnosis, five ethnic categories and five categories of an indicator of socioeconomic status. A total of 101 firefighters with a diagnosis of cancer of the kidney or renal pelvis were assessed, and the OR was 1.07 (95% CI: 0.87–1.31), adjusted for age, calendar period of diagnosis, race, and an indicator of socioeconomic status for the census block of residence.

Krstev et al. (1998) investigated incident prostate cancer cases in the USA using population-based cancer registries for Atlanta Georgia and Detroit Michigan, and for ten counties in the state of New Jersey during 1986–1989. Histologically confirmed cases were identified from pathology and outpatient records at hospitals included in these registries. Cases were selected by random sampling among the total number of cases identified in each age–race category. [Three additional cancer sites were investigated but not reported including oesophagus, pancreas, and multiple myeloma, and no published articles were located regarding these cancers.] Control subjects were proportional to the age, sex, and race distribution of the cases. Controls younger than 65 years of age were selected through random-digit dialling. Older controls were systematically selected by random sampling from computerized records of the Health Care Financing Administration stratified by three age groups, and race (african american or caucasian for each geographic area). Cases and controls were interviewed in person. There were 981 cases and 1315 controls analysed using unconditional logistic regression adjusted for age (< 60, 60–69, 70+), study site, and race. A total of ten cases and five controls were classified as firefighting (SOC 512.3). The overall adjusted OR for prostate cancer was 3.85 (95% CI: 1.34–11.10), for caucasians only (nine cases and three controls) 4.75 (1.26–18.00), and for african americans (three cases and two controls), 2.64 (0.43–16.20).

Bates (2007) evaluated 1144 firefighters diagnosed with cancer of the prostate (cohort described above for cancer of the kidney), and found an adjusted OR of 1.22 (95% CI: 1.12–1.33).

Stang et al. (2003) examined the risk of testicular cancer or extragonadal germ cell tumours during 1995–1997 in five German regions. Cases were reported through an active reporting system. A pathologist derived histological evaluations for 95% of the cases. Interviews were conducted with 269 of the 353 eligible cases, with a response rate of 78% including the two surrogate interviews. Controls were randomly selected from mandatory registries of residence. Approximately two controls were age- and region-matched for the cases between the ages of 15–34 years. Four controls were matched for those cases aged 35–69 to increase study power related to the fewer number of cases in this older age group. The response rate in the controls was 57%, with 918 interviewed (eight surrogate) of 1982 eligible subjects. Each job held for at least 6 months was recorded including job tasks and hours per week worked. These jobs were coded according to the International Standard Classification of Occupation. Conditional logistic regression models were calculated with matching factors including 5-year age groups, and geographic region. The adjusted ORs for ‘ever’ versus ‘never’ employed as a firefighter were 4.3 (95% CI: 0.7–30.5, four cases and three controls); for working as a firefighter ≥ 10 years, 3.0 (95% CI: 0.2–45.5, two cases and two controls); and for employment ≥ 5 years before the ‘reference’ date [date of diagnosis], 3.1 (95% CI: 0.4–24.4, three cases and three controls).

Bates (2007) also evaluated 70 firefighters diagnosed with cancer of the testis (SEER code 28020, cohort described above for cancer of the kidney), and found an adjusted OR of 1.54 (95% CI: 1.18–2.02).

Gaertner et al. (2004) reported on incident cases of bladder cancer with a histological confirmation, identified through the National Enhanced Cancer Surveillance System programme in seven Canadian provinces. The cases were adults aged 20–74, identified during 1994–1997 and interviewed 2–5 months after diagnosis. Random selections of population controls were included in the programme by frequency-matching age and gender to all cancer cases. Random digit dialling was used during the 1996 calendar year to recruit controls living in Newfoundland and Alberta, while all other provinces used a random sample from the provincial health insurance database. Native Indians and subjects in the military were excluded from the study. Mailed questionnaires with telephone follow-up, as necessary, were used to gather data regarding sociodemographics, occupational history, smoking history, dietary habits, and specific agent exposures. The response rates for the male and female bladder cancer cases were 66% and 72%, respectively, and for the controls, 59% and 65%, respectively. The overall analysis included 887 cases and 2847 controls. In the analysis of firefighters, eight male cases and 13 male controls were considered. the Standardized Occupational Classification system was used to code occupations, with up to 12 occupations coded per person. Data analysis also included demographic information provided from the interviews. An unconditional logistic regression analysis was used adjusting for age, province, race, smoking, ex-smoking, and consumption of fruit, fried food, and coffee. For the analysis of ‘ever’ or ‘never’ worked as a firefighter for more than one year, an elevated OR of 1.51 (95% CI: 0.59–3.84) was found. When stratified by duration of employment as a firefighter, the ORs were: 2.0 (95% CI: 0.43–9.49) for > 1–5 years (three cases and four controls); 0.86 (95% CI: 0.708–8.93) for > 5–15 years (one case and three controls); and 1.36 (95% CI: 0.36–5.16) for > 15 years (four cases and six controls).

Bates (2007) assessed 174 firefighters diagnosed with cancer of the bladder (SEER code 29010, cohort described above for cancer of the kidney and Table 2.6), and found an adjusted OR of 0.85 (95% CI: 0.72–1.00).

2.2.2. Cancer of the brain

Four studies on brain cancer in relation to firefighting were considered, all from the USA (Tables 2.4 and 2.6).

Table 2.4

Case–control studies of the brain.

Brownson et al. (1990) evaluated brain cancers using the Missouri Cancer registry. Cancer cases from public and private hospitals have been collected since 1972, and reporting has been mandated since 1984. The group of cases comprised Caucasian males diagnosed with histologically confirmed brain and other central nervous system cancers (ICD codes 191 and 192). Four controls were randomly selected and frequency-matched from all Caucasian male patients diagnosed with cancers during the same time period. Control group cancers included cancers of the lip, oral cavity and pharynx, digestive organs and peritoneum, respiratory system, skin, bones and connective tissue, genitourinary system, and leukaemia, lymphoma, multiple myeloma, and other sites. Of the initially eligible cases, occupational information was lacking in 34% of the cases, and 38% of the controls. Analysis combined industries with Unites States census code related to justice, public order and safety which included firefighters, and for occupations combining police and fire protection services. Age- and smoking-adjusted ORs were elevated and reported as 2.1 (95% CI: 0.9–4.8, ten cases and 19 controls) for the industry of justice, public order and safety, and 2.2 (95% CI: 1.0–4.7, 12 cases and 22 controls) for police and fire protection workers. This excess risk among police and fire protection workers was confined to the astrocytic cell series (OR, 2.3; 95% CI: 1.0–5.1). The OR for firefighters examined separately was 2.0 (95% CI: 0.4–9.6), with an unknown number of cases and controls.

Carozza et al. (2000) conducted a population-based case–control study among adults in the San Francisco Bay area during 1991–1994. Lifetime job histories were available for this study. Using the Northern California Cancer Center population, 603 incident cases of gliomas among adults aged 20 years or older were identified with histological confirmation (ICD 9380–9481). Interviews were completed with 492 cases (82%), and 476 were analysed after additional exclusions. Using random-digit dialling, controls were frequency-matched by 5-year age groups as well as by gender and race/ethnicity. There were 754 potential controls identified with 22 removed because of their residence, insufficient level of English or some relationship to the cases. Of the 732 controls meeting the eligibility criteria, 462 (63%) interviews were completed. The job history data for cases and controls were provided by proxy for 45.6% and 0.9%, respectively. For each job reported, the following information was collected: name and location of the company, description of daily work activities, starting date and duration of job including hours worked per week. Jobs were coded using Standardized Occupational Classification 1980 and Standard Industrial Codes 1987 without knowledge of the case–control status. Duration of all jobs held for at least 6 months was analysed; the most recent 10 years were excluded to allow for a hypothesized 10-year latency period between the exposure and the clinical recognition of the disease. Subjects who were not employed in the occupational category being evaluated served as the ‘unexposed’ referent group. Multiple logistic analyses were used adjusting for age, gender, years of education and race (caucasian, noncaucasian). Astrocytic tumours were evaluated including glioblastoma, multiforme, and astrocytoma. The adjusted OR for ‘ever’ versus ‘never’ employed as a fireman was 2.7 (95% CI: 0.3–26.1), and for being diagnosed with having an astrocytic tumour, 3.6 (95% CI: 0.4–36, three cases, 1 control).

Krishnan et al. (2003) conducted a follow-up study to the one designed by Carozza et al. (2000). This follow-up study examined incident glioma cases diagnosed during both 1991–1994 and 1997–1999. All adults newly diagnosed with glioma during these time periods were ascertained using the Northern California Cancer Center's rapid ascertainment programme. Controls were ascertained through random-digit dialling and matched to cases by age, race, and gender. There were 1129 eligible cases with 81% (n = 896) completing full interviews. In-person interviews were conducted for 98%, and there were 879 cases with complete information available for analysis. Of the eligible controls, 66% (n = 864) completed a full interview. In the analysis of ‘ever’ employed as a firefighter, the OR was 2.85 (95% CI: 0.77–10.58, nine cases and three controls). Analysis by the longest-held job resulted in an OR of 5.88 (95% CI: 0.70–49.01, six male cases and one male control). In the analysis of astrocytic cases, the OR was 6.31 (95% CI: 0.73–54.4, five cases and one control), and for the non-astrocytic cases, 9.27 (95% CI: 0.55–155.27, one case and one control). [These two studies are very similar with more cases and controls available in the Krishnan report. The Krishnan report, however, did not carry out analyses by 10-year latency period, and therefore both studies may be relevant.]

Bates (2007) also evaluated brain cancers (SEER code 31010) in firefighters as described above under kidney cancer and Table 2.6. There were 71 firefighters with brain cancer. The adjusted OR was 1.35 (95% CI: 1.06–1.72).

2.2.3. Cancers of the larynx and lung

One case–control study of cancer of the larynx and two studies of cancer of the lung were considered by the Working Group (Tables 2.5 and 2.6).

Table 2.5

Case–control studies of cancers of the larynx and lung.

Muscat and Wynder (1995) conducted a case–control study of cancer of the larynx in New York City, USA, recorded during 1956–1965. Caucasian men from seven hospitals newly diagnosed with histologically confirmed cancer of the larynx were interviewed. Control subjects were also caucasian men frequency-matched to the cases by hospital of diagnosis, age (within 5 years), and year of interview. Eligibility as a control also required a hospital admission for a condition unrelated to an etiology associated with tobacco exposures including cancer of the prostate, lymphomas, benign prostatic hypertrophy, and various non-malignant conditions. All subjects were interviewed by personnel who were not blinded to the case–control status of subjects, with a 90% response rate. The questionnaire included information on smoking status (never, current or ex-smoker, number of cigarettes, pipe and cigars smoked, and alcohol intake). Data were collected on lifetime occupations and self-reported exposures to chemicals, metals, exhaust, asbestos, and other occupational substances. There were 235 cases and 205 controls. The cases compared to controls were most likely to be: current cigarette smokers, (66.4% and 24.4%, respectively), heavy cigarette smokers (> 31 cigarettes/day), (55.1% and 22.8%, respectively), and drink more than 7 ounces of alcohol per day (29.4% and 11.2%, respectively). Analyses were adjusted for current smoking status.

Among the cases, two were employed as firefighters. Of those occupations which self-reported exposure to diesel exhaust, including truck drivers, firefighters, road workers, and mine workers (5.5 cases and 4.4 controls), the adjusted OR was 1.47 (95% CI: 0.5–4.1). For those occupations which self-reported exposure to diesel ‘fumes', firefighter was not listed amongst them The authors noted that the self-reported exposure to diesel exhaust or diesel fumes may reflect uncontrolled confounding with cigarette smoking and alcohol consumption as almost all patients who reported diesel exposure were also heavy cigarette smokers, and consumed large amounts of alcohol.

Elci et al. (2003) examined the link between occupations and risk of lung cancer by histological types in Turkey. Cases were identified from an oncology treatment centre at one of the largest cancer hospitals, including treatment for workers, in Istanbul. After admission to hospital, all patients completed a standardized questionnaire administered by trained interviewers. There were 1354 male lung cancer cases with complete interview information identified during 1979–1984. An oncologist reviewed hospital records for diagnostic verification and classification of cancer types. When there were four or more cases per cancer type, histopathology and morphological type was examined. Patient controls “with the same sociodemographic background as the cases” were selected having the following diagnoses: cancers of the skin (non-melanoma), testis, bone, male breast, Hodgkin disease, soft-tissue sarcoma, and non-cancer patients. Of the 27 occupations, firefighting (n = 10 cases) had an excess risk of lung cancer, with an age- and smoking-adjusted OR of 6.8 (95% CI: 1.3–37.4). In firefighters, for squamous-cell carcinoma (n = 4), the age- and smoking-adjusted OR was 6.2 (95% CI: 0.8–46.2), and for peripheral tumours including bronchus and parenchyma (n = 9), the age- and smoking-adjusted OR was 7.0 (95% CI: 1.3–39.1).

Bates (2007) investigated cancers of the lung and bronchus in firefighters as described above under kidney cancer and in Table 2.6. There were 495 firefighters with these cancers. The adjusted OR was 0.98 (95% CI: 0.88–1.09).

2.3. Descriptive studies

Several descriptive studies have provided results for firefighters. These have varied in their design including cohort studies based on record linkage, and studies based solely on death certificate or registry data. In some cases, these have been investigations specifically directed at firefighters. They are described in more detail below and in Tables 2.7 and 2.8.

Table 2.7

Cohort and linkage studies of firefighters.

Table 2.8

Descriptive studies with results on firefighters.

2.4. Case reports

Individual firefighters have applied for, and sometimes received, workers' compensation for cancer. An apparent cluster of cancer among firefighters was reported in an investigation of a chemical waste dump fire by NIOSH (Hrubec et al., 1992). However, the authors concluded it was not likely to have been related to firefighting. [Given the limitations of these reports and the large number of descriptive, cohort, and case–control studies with data on firefighters, the Working Group did not believe that case reports would contribute to the evaluation.]

2.5. Meta-analyses

Two meta-analyses of studies of firefighters and cancer have been conducted (Howe & Burch, 1990; LeMasters et al., 2006). The most recent meta-analysis included a great majority of the studies considered by the Working Group (LeMasters et al., 2006). Cancer risk was significantly elevated for ten of the 21 cancer types analysed (stomach, colon, rectum, skin, prostate, testis, brain, non-Hodgkin lymphoma, multiple myeloma, and malignant melanoma). With the exception of testicular cancer (summary RR = 2.02), the summary relative risk estimates were moderate, ranging from 1.21 for colon to 1.53 for multiple myeloma. For four of these sites (prostate, testis, non-Hodgkin lymphoma, and multiple myeloma), findings were consistent across study designs and the types of study available. However, since that analysis, two additional large studies of cancer in firefighters had been published (Ma et al., 2006; Bates, 2007). Therefore, another meta-analysis was performed by the Working Group to assess the impact of these recent studies.

Inclusion criteria for studies in this meta-analysis were reported estimates of relative risk with corresponding 95% confidence intervals or information that allowed their computation by the Working Group for ‘ever’ versus ‘never’ exposure to firefighting or employment as a firefighter. For those studies that did not report for this category, the relative risks and 95% confidence intervals were estimated by the Working Group from strata-specific relative risk and corresponding number of cases, assuming a normal distribution when possible. Studies that only reported point estimates without confidence intervals were not included. Proportionate mortality studies were also excluded. Statistical heterogeneity among studies was tested with the Q statistic. Summary relative risk estimates were obtained from random-effect models for prostate cancer (Q = 32.816, P = 0.005), and fixed-effect models for testicular cancer (Q = 3.928, P = 0.560), and non-Hodgkin lymphoma (Q = 6.469, P = 0.486). All statistical analyses were performed using STATA (version 9.0; StataCorp, College Station, TX).

Based on the Working Group's meta-analysis, three of the four sites remained statistically significant. Testicular cancer was evaluated based on six studies and 409 cases (Giles et al., 1993; Aronson et al., 1994; Bates et al., 2001; Stang et al., 2003; Ma et al., 2006; and Bates, 2007). The results demonstrated an approximate 50% increased risk (1.47, 95% CI: 1.20–1.80, fixed effects). Prostate cancer was evaluated using 16 available studies and 1764 cases (Aronson et al., 1994; Baris et al., 2001; Bates et al., 2001 ; Bates, 2007; Beaumont et al., 1991; Demers et al., 1994; Firth et al., 1996; Giles et al., 1993; Grimes et al., 1991; Guidotti 1993; Krstev et al., 1998; Ma et al., 1998; Ma et al., 2006; Pukkala, 1995; Tornling et al., 1994; and Vena & Fiedler, 1987). The results showed a 30% elevated risk (1.30; 95% CI: 1.12–1.51, random effects). Non-Hodgkin lymphoma was evaluated based on seven studies and 312 cases, and had a 21% elevated risk estimate (1.21; 95% CI: 1.08–1.36, fixed effects) (Baris et al., 2001; Bates, 2007; Giles et al., 1993; Ma et al., 1998, 2006; Pukkala, 1995; and Sama et al., 1990).

2.6. References

• Aronson KJ, Tomlinson GA, Smith L. Mortality among fire fighters in metropolitan Toronto. Am J Ind Med. 1994;26:89–101. [PubMed: 8074127] [CrossRef]
• Baris D, Garrity TJ, Telles JL, et al. Cohort mortality study of Philadelphia firefighters. Am J Ind Med. 2001;39:463–476. [PubMed: 11333408] [CrossRef]
• Bates MN. Registry-based case-control study of cancer in California firefighters. Am J Ind Med. 2007;50:339–344. [PubMed: 17427202] [CrossRef]
• Bates MN, Fawcett J, Garrett N, et al. Is testicular cancer an occupational disease of fire fighters? Am J Ind Med. 2001;40:263–270. [PubMed: 11598972] [CrossRef]
• Beaumont JJ, Chu GS, Jones JR, et al. An epidemiologic study of cancer and other causes of mortality in San Francisco firefighters. Am J Ind Med. 1991;19:357–372. [PubMed: 2008922] [CrossRef]
• Berg JW, Howell MA. Occupation and bowel cancer. J Toxicol Environ Health. 1975;1:75–89. [PubMed: 1185818] [CrossRef]
• Blair A, Walrath J, Rogot E. Mortality patterns among U.S. veterans by occupation. I. Cancer. J Natl Cancer Inst. 1985;75:1039–1047. [PubMed: 3865009]
• Brownson RC, Reif JS, Chang JC, Davis JR. An analysis of occupational risks for brain cancer. Am J Public Health. 1990;80:169–172. [PMC free article: PMC1404624] [PubMed: 2297060] [CrossRef]
• Burnett CA, Halperin WE, Lalich NR, Sestito JP. Mortality among fire fighters: a 27 state survey. Am J Ind Med. 1994;26:831–833. [PubMed: 7892834] [CrossRef]
• Carozza SE, Wrensch M, Miike R, et al. Occupation and adult gliomas. Am J Epidemiol. 2000;152:838–846. [PubMed: 11085395] [CrossRef]
• Decoufle P, Lloyd JW, Salvin LG. Mortality by cause among stationary engineers and stationary firemen. J Occup Med. 1977;19:679–682. [PubMed: 903842] [CrossRef]
• Delahunt B, Bethwaite PB, Nacey JN. Occupational risk for renal cell carcinoma. A case-control study based on the New Zealand Cancer Registry. Br J Urol. 1995;75:578–582. [PubMed: 7613791] [CrossRef]
• Demers PA, Checkoway H, Vaughan TL, et al. Cancer incidence among firefighters in Seattle and Tacoma, Washington (United States). Cancer Causes Control. 1994;5:129–135. [PubMed: 8167259] [CrossRef]
• Demers PA, Heyer NJ, Rosenstock L. Mortality among firefighters from three northwestern United States cities. Br J Ind Med. 1992a;49:664–670. [PMC free article: PMC1039313] [PubMed: 1390274]
• Demers PA, Vaughan TL, Checkoway H, et al. Cancer identification using a tumor registry versus death certificates in occupational cohort studies in the United States. Am J Epidemiol. 1992b;136:1232–1240. [PubMed: 1476145]
• Demers PA, Vaughan TL, Koepsell TD, et al. A case-control study of multiple myeloma and occupation. Am J Ind Med. 1993;23:629–639. [PubMed: 8338527] [CrossRef]
• Deschamps S, Momas I, Festy B. Mortality amongst Paris fire-fighters. Eur J Epidemiol. 1995;11:643–646. [PubMed: 8861847] [CrossRef]
• Dubrow R, Wegman DH. Setting priorities for occupational cancer research and control: synthesis of the results of occupational disease surveillance studies. J Natl Cancer Inst. 1983;71:1123–1142. [PubMed: 6581356]
• Dubrow R, Wegman DH (1984) NIOSH Research report: Occupational characteristics of cancer victims in Massachussetts 1971–1973. Cincinnati: National Institute for Occupational Safety and Health.
• Elci OC, Akpinar-Elci M, Alavanja M, Dosemeci M. Occupation and the risk of lung cancer by histologic types and morphologic distribution: a case control study in Turkey. Monaldi Arch Chest Dis. 2003;59:183–188. [PubMed: 15065312]
• Eliopulos E, Armstrong BK, Spickett JT, Heyworth F. Mortality of fire fighters in Western Australia. Br J Ind Med. 1984;41:183–187. [PMC free article: PMC1009280] [PubMed: 6722044]
• Feuer E, Rosenman K. Mortality in police and firefighters in New Jersey. Am J Ind Med. 1986;9:517–527. [PubMed: 3488681] [CrossRef]
• Finkelstein MM. Occupational associations with lung cancer in two Ontario cities. Am J Ind Med. 1995;27:127–136. [PubMed: 7900730] [CrossRef]
• Firth HM, Cooke KR, Herbison GP. Male cancer incidence by occupation: New Zealand, 1972–1984. Int J Epidemiol. 1996;25:14–21. [PubMed: 8666482] [CrossRef]
• Gaertner RR, Trpeski L, Johnson KC. Canadian Cancer Registries Epidemiology Research Group, author. A case-control study of occupational risk factors for bladder cancer in Canada. Cancer Causes Control. 2004;15:1007–1019. [PubMed: 15801485] [CrossRef]
• Gallagher RP, Threlfall WJ, Band PR et al. (1989) Occupational Mortality in British Columbia 1950–1984. Vancouver: Cancer Control Agency of British Columbia.
• Giles G, Staples M, Berry J. Cancer incidence in Melbourne Metropolitan Fire Brigade members, 1980–1989. Health Rep. 1993;5:33–38. [PubMed: 8334236]
• Grimes G, Hirsch D, Borgeson D. Risk of death among Honolulu fire fighters. Hawaii Med J. 1991;50:82–85. [PubMed: 2061032]
• Guidotti TL. Mortality of urban firefighters in Alberta, 1927–1987. Am J Ind Med. 1993;23:921–940. [PubMed: 8328477] [CrossRef]
• Guralnick L (1963) Mortality by occupation and cause of death among men 20 to 64 years of age, United States, 1950. Washington, DC: Public Health Service, National Vital Statistics Division, (Vital Statistics–Special reports 53, No. 1–5).
• Gute DM (1981). The association of occupation and industry with mortality in Rhode Island (1968–1972). Providence: Rhode Island Department of Health (Technical report No. 23, Health Planning and Development).
• Hansen ES. A cohort study on the mortality of firefighters. Br J Ind Med. 1990;47:805–809. [PMC free article: PMC1035284] [PubMed: 2271386]
• Heyer N, Weiss NS, Demers P, Rosenstock L. Cohort mortality study of Seattle fire fighters: 1945–1983. Am J Ind Med. 1990;17:493–504. [PubMed: 2327416] [CrossRef]
• Howe GR, Burch JD. Fire fighters and risk of cancer: an assessment and overview of the epidemiologic evidence. Am J Epidemiol. 1990;132:1039–1050. [PubMed: 2260535]
• Hrubec, Z., Blair, A.E., Rogot,E. & Vaught, J. (1992) Mortality risks by occupation among U.S. Veterans of known smoking status. NIH Publication No. 92–3407.
• Krishnan G, Felini M, Carozza SE, et al. Occupation and adult gliomas in the San Francisco Bay Area. J Occup Environ Med. 2003;45:639–647. [PubMed: 12802217] [CrossRef]
• Krstev S, Baris D, Stewart P, et al. Occupational risk factors and prostate cancer in U.S. blacks and whites. Am J Ind Med. 1998;34:421–430. [PubMed: 9787845] [CrossRef]
• LeMasters GK, Genaidy AM, Succop P, et al. Cancer risk among firefighters: a review and meta-analysis of 32 studies. J Occup Environ Med. 2006;48:1189–1202. [PubMed: 17099456] [CrossRef]
• Ma F, Fleming LE, Lee DJ, et al. Mortality in Florida professional firefighters, 1972 to 1999. Am J Ind Med. 2005;47:509–517. [PubMed: 15898094] [CrossRef]
• Ma F, Fleming LE, Lee DJ, et al. Cancer incidence in Florida professional firefighters, 1981 to 1999. J Occup Environ Med. 2006;48:883–888. [PubMed: 16966954] [CrossRef]
• Ma F, Lee DJ, Fleming LE, Dosemeci M. Race-specific cancer mortality in US firefighters: 1984–1993. J Occup Environ Med. 1998;40:1134–1138. [PubMed: 9871891]
• Mastromatteo E. Mortality in city firemen. II. A study of mortality in firemen of a city fire department. AMA Arch Ind Health. 1959;20:227–233. [PubMed: 14422193]
• Milham S (1976) NIOSH Research report: Occupational mortality in Washington state 1950–1971. Volume 1–3. Cincinnati: National Institute for Occupation Safety and Health [DHEW Publication No. (NIOSH) 76–175].
• Milham S (1997) Occupational Mortality in Washington State: 1950–1989. Cincinnati, Ohio: US DHHS (NIOSH) Publication No 96–133.
• Morton W, Marjanovic D. Leukemia incidence by occupation in the Portland-Vancouver metropolitan area. Am J Ind Med. 1984;6:185–205. [PubMed: 6475965] [CrossRef]
• Muscat JE, Wynder EL. Diesel exhaust, diesel fumes, and laryngeal cancer. Otolaryngol Head Neck Surg. 1995;112:437–440. [PubMed: 7870446] [CrossRef]
• Musk AW, Monson RR, Peters JM, Peters RK. Mortality among Boston firefighters, 1915-1975. Br J Ind Med. 1978;35:104–108. [PMC free article: PMC1008362] [PubMed: 656333]
• OPCS (1971) Occupational mortality. The Registrar General's decennial supplement for England and Wales, 1961. London: Her Majesty's Stationary Office.
• OPCS (1972) Supplement on cancer. The Registrar General's statistical review of England and Wales for the two years, 1966–1967. London: Her Majesty's Stationary Office.
• OPCS (1975) Supplement on cancer. The Registrar General's statistical review of England and Wales for the three years, 1968–1970. London: Her Majesty's Stationary Office.
• OPCS (1978) Occupational mortality. The Registrar General's decennial supplement for England and Wales, 1970–1972. London: Her Majesty's Stationary Office.
• OPCS (1995) Occupational Health: Decennial Supplement. The Registrar General's decennial supplement for England and Wales. Series DS No. 10. London: Her Majesty's Stationary Office.
• Petersen GR, Milham S (1980) NIOSH Research report: Occupational mortality in the state of California 1959–1961. Cincinnati: National Institute for Occupation Safety and Health [DHEW Publication No. (NIOSH) 80–104].
• Pukkala E (1995) Cancer risk by social class and occupation: a survey of 109 000 cancer case's among Finns of working age. In: Contributions to Epidemiology and Biostatistics: vol. 7. M.A., Klingberg, editor.
• Sama SR, Martin TR, Davis LK, Kriebel D. Cancer incidence among Massachusetts firefighters, 1982–1986. Am J Ind Med. 1990;18:47–54. [PubMed: 2378369]
• Stang A, Jöckel KH, Baumgardt-Elms C, Ahrens W. Firefighting and risk of testicular cancer: results from a German population-based case-control study. Am J Ind Med. 2003;43:291–294. [PubMed: 12594776] [CrossRef]
• Tornling G, Gustavsson P, Hogstedt C. Mortality and cancer incidence in Stockholm fire fighters. Am J Ind Med. 1994;25:219–228. [PubMed: 8147394] [CrossRef]
• Vena JE, Fiedler RC. Mortality of a municipal-worker cohort: IV. Fire fighters. Am J Ind Med. 1987;11:671–684. [PubMed: 3605104] [CrossRef]
• Walrath J, Rogot E, Murray J et al. (1985) Mortality among U.S. veterans by occupation and smoking status. NIH Publication No. No 85-2756. Washington, DC: U S Government Printing Office :1-311.
• Williams RR, Stegens NL, Goldsmith JR. Associations of cancer site and type with occupation and industry from the Third National Cancer Survey Interview. J Natl Cancer Inst. 1977;59:1147–1185. [PubMed: 903993]