Document n96byyMXQN7KZRn3507daQpJR

AMERICAN JOURNAL OF INDUSTRIAL MEDICINE 34~393-398(1 998) A Nested Case-Control Study of Bladder Cancer Incidence in a Steel Manufacturing Plant KatherineMallin, PhD* Background A surveillance study of bladder cancer incidence in northwestern Illinois detected a number of cases who had worked in a large steel manufacturing plant. To investigate these cancers further; a nested case-control study of bladder cancer was conducted at thisplant. Methods Casesfrom the surveillance study were matched to company personnel files. Four controls per case were selectedfrom companyfiles andfrequency matched on age {casesand controls were all white males).Employment histories were abstractedfrom company records and odds ratios calculated according to job titles and departments worked. Cases and controls who held only white collarjobs were excludedfrom the analysis. Results Resultsfor 16 cases and 75 controls showed elevated odds ratiosfor heaters (OR= 21.1, 95%cCI = 2.2-205.8)based on three cases. Heaters monitored electric reheatfurnuces and may have been exposed to polycyclic aromatic hydrocarbons. Previous studies have shown moderate risk for bladder cancer amongfumace operators, but most of these studies were also based on a small'numberof cases. ConclusionsAdditional studies offumace operators' exposures and cancer risk are needed to investigate these resultsfurther.Am. J. h d . Med. 34~393-398,1998. o 1998 Wley-Liss,Inc. KEY WORDS: bladder neoplasms; steel manufacturing; furnace operators; polycyclic aromatic hydrocarbons INTRODUCTION Medical record data collected for a surveillance study of bladder cancer incidence in northwestern Illinois revealed that a number of bladder cancer cases had worked at a single large steel and wire manufacturing plant [Mallin, 19901. Although epidemiologic studiesof steelworkershave primarily detected excess lung cancer risks Lloyd, 1971; Blot et al., 1983;Finkelstein and Wilk, 1991; Beaumont et al., 1987; Steenland and Beaumont, 19891, elevated bladder cancer risks have been found in industries that might share similar Division of Epidemiology-Biostatistics, University of Illinoisat Chicago, School of Public Health, Chicago, Illinois Contract grant sponsor: State Cancer Pian of the IllinoisCancer Center *cMTespMldence to: Katherine Mallin, Division of EpidemMgy-Biitistics, University of Illinois at Chicago, School of Public Health, 2121 W. Taylor St, Chicago, IL 60612 Accepted 29 April 1998 exposures with steelworkers, including metal workers [Schifflers et al., 1987;Houten et al., 1977;Silverman et al., 19831, machinists [Anton-Culver et al., 1992; Silverman et al., 1983; Houten et al., 1977; Howe et al., 1980; Maker et al., 19871,heaters and furnace operators [Hansen, 1992; Silverman et al., 1989a; Zahm 1987; Steenland, 1987; Houten et al., 19771, welders [Howe et al., 19801, foundry workers [Egan-Baum et al., 1981; Sherson et al., 1991; Hansen, 1991; Cordier, 19931, coal carbonization and gasification workers [Doll et al., 1972; Henry et al., 19311, and aluminum smelter workers [Thkriault et al., 1984; Armstrong et al., 19861. These workers may be exposed to a variety of agents, including solvents, cutting oils, metal fumes and dusts, polycyclic aromatic hydrocarbons (PAH), combustion gases, and vapors. The steel plant under study was founded in 1879 as a barbed wire and field fence manufacturer. Currently, the plant includes a building with three large electric furnaces for converting raw scrap to molten steel, a casting building for producing semifinished steel, a steel rod mill, two o 1998Wiley-Lis, Inc. 394 Mallin structural steel mills, and a wire mill that produces field fence and packaged nails. The rod and structural mills each have gas- andor oil-fired reheat furnaces that reheat the semifinished steel before entry into the rolling mills. The electric furnaces have the capability to produce up to two millions tons of raw steel annually. At its peak, the company employed approximately 4,000 workers but currently employs approximately 2,500 hourly workers. Employee turnover averages 5 4 % per year; many workers have been employed at the plant for their entire working careers. The plant is a major employer in a small town of 15,000people. Because of the number of different processes used, more than 1,OOO different job titles have been used at thls plant. In order to further investigate the bladder cancer risks in this plant, a nested case-control study of bladder cancer was conducted, to assess whether cases had a different plant work and exposure history than other workers at the plant. MATERIALS AND METHODS A list of 210 male bladder cancer cases obtained in the surveillance study [Mallin, 19901 were submitted to the company for matching to their personnel files. These cases were diagnosed between 1978 and 1985 and lived in the county in which the steel plant was located or in four adjacent counties. Only cases with pathologic confirmation were included. In order to test the accuracy of the matching procedure, 13 cases already known to have worked at the plant were included in the list of cases submitted to the company. The personnel files that were matched were computerized in 1977 and include all employees with 10 or more years of service. Cases were matched on name, birth date, and Social Security number. Death certificates for deceased employees were also obtained for all vested employees who died between 1977 and 1989.The list of bladder cancer cases was also matched to these death certificate names to identify plant employees among the cases. In addition to the 13 cases already known to have worked at the plant, six more cases were identified as company employees, for a total of 19 bladder cancer cases among plant employees. All these cases were employed for at least 10 years at the plant. Ten were deceased, and nine were still living at the time the study was conducted. A systematic random sample of four controls were selected from the plant personnel files for each of the 19 cases. Controls were frequency matched on age within 1 year. Since all cases were white males, only white male controls were selected. Controls also had to be alive within 6 months of the case's diagnosis date of bladder cancer. For one case with no matches for exact age, age criteria were increased to within 2 years of age. Work histories were obtained from paper personnel records for the 19 cases and 76 controls. Information abstracted from personnel records includedjob title, department and/or plant number, and starting and ending month and year for each job. Jobs lasting less than 1 month were not included in the analysis. These would include layoff periods when employees were assigned to departments needing temporary help. Jobs were analyzed using both company job titles and Bureau of the Census occupational codes [Bureau of the Census, 19821. Company job descriptions were requested for jobs that could not be assigned census occupational codes. Information from these job descriptions was then used to assign an occupational code. In order to adjust for the effects of smoking, smoking history questionnaires were also mailed to each case and control, or for deceased subjects, to their next of kin. Next of kin were identified through death certificates, employee records, pension and insurance records, obituaries, and telephone directories. Two mailings were sent and, in many cases, more than one next of kin was contacted if the initial next of kin failed to return the questionnaire. Statistical Analysis Age-adjusted odds ratios [Mantel-Haenszel, 19591 and 95 percent test based confidence intervals [Miettinin, 19761 were calculated for job titles and departments worked in. Smoking adjusted odds ratios were also calculated. RESULTS Two cases and one control who held only white collar occupations were excluded from the analysis. One case and one control who had no information on job titles or departmentswere also excluded, resulting in 16cases and 74 controls for analysis. Analyses were conducted using company job titles and department names, and according to census defined occupational groups if at least three cases were employed in that occupation, department, or plant. Owing to the small number of cases, age-adjusted odds ratios were calculated for two age groups only, 65 and younger, and older than 65 years. About one-half of the cases and one-half of the controls comprised each of these two age groups. As shown in Table I, odds ratios were significantly elevated for the company job title of heaters (odds ratios [OR] = 21.1; 95% confidence interval [95%CI] = 2.2206.8) and the census job title of heat-treating equipment operators (OR = 29.7; 95%CI = 5.1-173.2). Confidence intervals were wide because of the small number of cases in these jobs. The three cases employed as heaters worked as heaters or chargers for their entire employment at the plant Bladder Cancer in a Steel Plant 395 TABLE 1. Age-Adjusteda Bladder Cancer Odds Ratios, Number of Exposed Workers, and 95% Confidence Intervals, by Plant, Company Occupational and Census Code Occupational TitleCfor Ever Versus Never Employed Job We1 plant No. of Odds exposed raib cases 95%CI Company job title Heater 21.ldne 3 2.2-205.8 Laborer 0.9 4 0.3-2.8 Machineoperator/operatorlearner 1 .l 3 0.34.4 Census job title (Code) Heat-treating equipment operator (472) 29.7 4 5.1-173.2 Machine operators (755,769,777,779) 1.1 8 0.3-3.6 Plant number Plant 1 0.7 10 0.2-2.0 Plant 2 0.7 6 0.2-2.2 Plant 3 1.2 3 0.3-5.1 TABLE II. Bladder Cancer Odds Ratios and 95% Confidence Intervals for Heaters, Among Smokers Only,aAdjusted for Ageb Job We Odds No.ofexposed ratios ca88s,comls %%CI Heater (Verbatim) 8.4c 2 , o 0.77-92.2 Heat-treating equipment operator (Censuscode 472) 13.9 3,1 1.3-1 51.4 #Eversmoked 100 or more cigarettes. bAge 5 6 5 vs >65 years. c ~ o geistimate of odds ratio. smoking history, so ORs for employment as a heater or annealer were calculated for smokers only (10 cases, 24 controls). ORs were still elevated for heaters (verbatim) and among the census-definedgroup of heaters (Table II). DISCUSSION `Age 5 6 5 vs >65 years. bForjobdplants with three or more cases ever employed. cBureauof the Census [1982] occupational code in parentheses. dLogRestimate of relative fisk. `p c 0.01. for periods of 20-40 years, and the case who worked as an annealer was employed in that occupation for 35 years. One control worked as an annealer helper for 4 years; no control had worked as a heater. Thus, ORs for heaters with 20 or more years of service compared with those with less than 20 years of service would be the same as those presented. Other jobs with three or more cases ever employed included laborers and machinists/operators. ORs for these jobs were not elevated. ORs adjustedfor 10or more, and for 20 or more years of service produced similar results. No elevated ORs were found for any departments or for any of the three separateplant buildings. Smoking Histories Smoking histones were obtained for 10of the 16cases, and 29 of the 74 controls. On the basis of these data, all 10 cases with smoking information had ever smoked 100 cigarettes or more, compared with 83% of controls (p > 0.10). Among smokers, information on the number of years smoked and number of cigarettes smoked per day was available for 21 controls and eight cases. Mean pack-years was slightly higher for cases than controls (17,467 vs 16,057),but the differences were not significant. Among the census occupational group of heaters, smoking information was available for three of the four cases and the single control, and all had smoked at least 100 cigarettes. There were too few nonsmokers to adjust for A limitation of this study is the relatively small number of cases. Given the large number of different jobs in this plant, the probability of cases having worked in similarjobs is small. Despite this, the number of cases who worked as heaters or annealers was significantly higher than expected. Another limitation is that the personnel records used to i d e n m cases only included workers with 10 or more years of service. However, given the relatively low employee turnover at this plant, it is not likely to affect the results. The results would also be biased if any cases who had worked at the plant were not identified in the matching process. However, because the 13 cases already known to have worked at the plant were correctly matched, it is likely that the matching process used was reliable. It was not possible to adjust completely for the effect of smoking on the results, owing to missing data for cases and controls. However, given the data available, smoking did not appear to confound the association between heater/annealer occupations and bladder cancer. Heaters/annealers could be exposed to a number of substances,especiallyPAH,including benzo(a)pyrene (BaP). Electric furnaces have been used at the plant since 1936.BaP exposure in steelworks with electric furnaces have been found to be moderately elevated in Sweden (range 0.4-4-2.6 mg/m3 mean BaP) [Lindstedt and Sollenberg, 19821 and Japan (range 0.43-0.58 mg/m3)[Tanimura, 19681. The three heaters spent their entire employment at the plant as heaters or chargers, and the annealer similarly spent his entire employment in this job. Heaters monitor the "reheat" furnaces that reheat the semifinished steel billets produced from the electric furnaces. This heated steel is then put through one of the rolling mills. Historically, the reheat furnaces were both gas and oil fired but are now predomi- 396 Mallin TABLE 111. Results from Case-ControlStudies ofBladder Cancer for Occupations of Heater, Stationary Fireman, Furnace Operator, or Related Occupations No. of cases, contrds In this job OR, MCI Smoking adjusted? Doserespon= tested? Rbfersnce Metal Heaters 534 1.9 (0.4-8.7) No No Silverman et al. [1989a] Stationary fireman or furnace operator 57,80 1.3 (0.4-1.9) Yes Yes; significanttrend Silverman et al. 11989al Forgeman, hammerman, roller finisher 11,13 1.4 (0.6-3.4) Yes No Silvermanet al. [1989a] Blacksmith Heat-associatedoccupationsin plastics 7,6 Not given 2.4 (0.7-8.2) 2.8 (0.5-15.3) Yes No Yes No Zahm et al. [1987 and rubber industry Furnaceman,smelterman, pourer 3,14 0.6 (0.1-22) Yes No Silverman et al. [1989b] Metal molder 2,14 0.4 (0.1-1.7) Yes No Silvermanet al. [1989b] Stationary fireman or furnace operator 5,lO 1.4 (0.4-5.1) Yes No Silverman et al. [1989b] Furnace workers 18,17 1.5 (0.8-3.2) Yes No Bums and Swanson [1991] Steelfoundry workers Blacksmiths, stokers, workers exposed 993 2.9 (0.8-10.9) Yes No 9,12 2.3 (0.9-6.2) Yes No Cordier et al. [1993] Schmers et al. [1987] to hot metal Thermal power stationworker and 13,6 2.2 (0.8-5.6) No Yes, p = 0.32 smoking adj. Kunze et al. [1992] furnaceman Gas worker Furnace, forge, foundry rolling mill 6,1 6.0 (0.4-38.5) No No 12,17 RR<1 (CINA) Yes No Kunze et al. [1992] Anthony and Thomas [1970] workers CI,confldenceinterval;NA, not available. nantly gas fired. Chargers operate machines to push billets into the reheat furnace, while annealers operate industrial heating units for annealing wire. Heaters, chargers, and annealers are all are exposed to heat, gas, and hot materials. Several case-control studies have detected elevated bladder cancer risks among workers in heat-associated occupations, including furnace operators, although most of the elevated risks were not statistically significant (Table III). One study found no elevated risk among furnacemen/ smeltermadpourers [Silverman et al., 1989b1, while three studies including the latter found risks ranging from 1.3 to 2.2 for stationary firemen or furnace operators [Silverman et al., 1989a,b; Kunze, 19921. Varying definitions of furnace operators as well as differences in exposures may account for these differentresults. Mortality studies of bladder cancer have also found elevated risks among furnace operators [Steenland et al., 19871,and stokers [Hansen, 1992; Henry et al., 1931](Table IV). Most of these studies were based on a small number of furnace operators that limited the ability to detect significantly elevated risks, particularly if the risks were small. Exposure to PAH, particularly BaP exposure has been associated with bladder cancer risks in studies of workers in aluminum reduction plants exposed to BaP [Thkriault et al., 19841 and in aluminum smelters with increasing tar expo- sures [Gibbs, 1985; Spinelli et al., 1991; Risch et al., 1988; Armstrong et al., 19861, and in case-control studies among workers determined to have frequent PAH exposure [Clavel et al., 1994; Bonassi et al., 19891. With one exception [Gibbs, 19851, all these studies were smoking adjusted. A Montreal case-control study did not detect elevated bladder cancer associated with coal tar pitch or benzo(a)pyrene exposurebut did find a threefold excess risk among workers with substantial exposure to natural gas combustion products [Nadon et al., 19951. A Swedish record linkage study found slightly eleveated relative risks for bladder cancer among workers exposed to combustion gases from wood or oil [Steineck et al., 19891. A recent literature review concluded that occupational exposures to PAH increase the risk of both lung and bladder cancers [Mastrangelo et al., 19961. Results from this study are consistent with results from other case-control studies of furnace operators and workers with PAH exposure. Since furnace operators generally make up a small proportion of workers within a plant, studies that have analyzed risks among these workers and have had limitedpower to detect significantly elevatedrisks. Documentation of PAH and other exposures among furnace workers in steel and other industries is needed to further assess any dose-related bladder cancer risks among these workers. Bladder Cancer in a Steel Plant 397 TABLE IV. Results from Bladder Cancer Mortality Studiesfor Occupations of Heater, Stationary Fireman/Fumace Operator and Related Occupations Job title No. of erposed cases,source ofocwpation OWSMRa 95% CI Smoking adjusted? DOW ~SspOrn tastsd? Reference Furnace operators US. death certificates 6 exposed deaths OR = 6.0 p < 0.01 No No Steenlandet al. [1987] CI NA U.S. C i Directories OR = 2.50 No Yes p < 0.01 for 20+ 13exposed deaths p < 0.05 yr employment CI NA Oven workers (exposedto wood soot) Swedish census RR = 1.1 No No Steineck et al. [1989] (0.5-2.1) Stokers in metal industry Danish census SMR = 147 No No Hansen [1992] Gas stoker 14exposed deaths (80-246) EnglanMales Census SMR = 156 No No Henry et al. [1931] 7 exposed deaths Brick kiln/oven men EnglandiWales Census SMR = 200 No No 4 exposed deaths CI. confidenceintelval;OR, odds ratio; SMR, standardizedmottalityratio. ACKNOWLEDGMENTS The author thanks the steel company and Union for their assistance in this study. Ms. Mary Kobialka and MS. Candice =Ora abstracted hospital and data for this study, and provided coding assistance. John ~ h ph.D~., at t~he un,iversity of mnOisat chicago, provided helpful comments. The State Health Registry of Iowa provided cancer incidence data for Illinois residents diagnosed in Iowa. Thefollowing hospitals provided cancer incidence data: Community General Hospital, Sterling, Morrison Community Hospital, Momson, Savanna City Hospital, Savanna, Freeport Memorial Hospital, Free- port, Rochelle Community Hospital, Rochelle, Katherine Shaw Bethea Hospital, Dixon, Galena-Stauss Hospital, Galena, Sycamore Community Hospital, Sycamore, High- land Hospital, Belvidere, St. Joseph's Hospital, Belvidere, Rockford Memorial Hospital, Rockford, Swedish American Hospital, Rockford, and St. Anthony's Hospital, Rockford. REFERENCES Anthony HM,Thomas GM (1970): Tumors of the urinary bladder: An analysis of the occupations of 1030 patients in Leeds, England. J Natl Cancer Inst 45:879-895. Anton-CulverH, Lee-FeldsteinA, TaylorT (1992):Occupation and bladder cancer risk. Am J Epidemiol 136:89-94. Armstrong BG, Tremblay CG, Cyr D, Th6riault GP (1986): Estimating the relationship between exposure to tar volatiles and the incidence of bladder cancer in aluminum smelter workers. Scand J Work Environ Health 12:486493. Beaumont J, Leveton H, Know K, Bloom T, McQuiston T, Young M, Goldsmith R, Steenland K, Brown D, Halperin W (1987): Lung cancer mortality in workers exposed to sulfuric acid mists and other acid mists in steel pickling operations. J Natl CancerInst 79911-921. Blot WJ, Moms LM, Pottern LM, Stone BJ, Fraumeni Jr. JF (1983): Lung cancer among long-term steel workers. Am J Epidemiol117:706-716. Bonassi S, Merlo F, Pearce N, Puntoi R (1989): Bladder cancer and occupational exposure to polycyclic aromatic hydrocarbons. Int J Cancer 44.64-51. Bureau of the Census (1982): "Alphabetic Index of Industries and Occupations (PHC8O-R3)," Washington DC: US. Department of Com- merce. Bums PB, Swanson GM (1991): Risk of urinary bladder cancer among Blacks and Whites: The role of cigaretteuse and occupation. CancerCauses and Control 2:371-379. Clavel J, Mandereau L, Limasset JC, Hemon D, Cordier S (1994): Occupational exposure to polycyclic aromatichydrocarbons and the risk of bladder cancer: AFrench case-controlstudy.Int J Epidemiol23:1145-1153. Cordier S, Clavel J, Limasset JC, Boccon-Gibod L, Le Moual N, Mandereau L, Hemon D (1993): Occupational risks of bladder cancer in France: A multicentre case-control study. Int J Epidemiol22:403411. DollR,Vessey MF',Beasley RWR,BuckleyAR,FearEC,FisherRFW,Gammon EI,GUMW, Hughes GO, Lee K, N~rman-SmithB (1972): Mortality Of gasworkers-finalreport of a prospective study. Br J Ind Med 29:394406. Egan-Baum E, Miller BA, Waxweiler RJ (1981): Lung cancer and other mortality patterns among foundrymen. Scand J Work Environ Health ~ ( S U P P4~):147-155. Finkelstein M, Wilk N (1991): Investigation of a lung cancer cluster in the melt shop of an Ontario steel producer.Am J Ind Med 17:483491. Gibbs GW (1985): Mortality of aluminum reduction plant workers, 1950-1977. J &CUP Med 27:761-770. 398 Mallin Hansen E (1991): Cancer mortality among Danish molders. Am J Ind Med 20:401409. Hansen E (1992): A mortality study of Danish stokers. Br J Ind Med 49:4&52. Henry SA, Kennaway NM, Kennaway EL (1931):The incidence of cancer of the bladder and prostate in certain occupations. J Hygiene 31:9-137. Houten L, Bross IDJ, Viadana E, Sonnesso G (1977): Occupational cancer in men exposed to metals. Adv Exp Med Biol91:93-102. Howe GR, Burch JD, Miller AB, Cook GM, Esteve J, Momson B, Gordon P, Chambers LW, Fodor G, Winsor GM (1980): Tobacco use, occupation, coffee, various nutrients, and bladder cancer. J Natl Cancer Inst 64:701713. Kunze E, Chang-Claude J, Frentzel-Beyme R (1992): Life style and occupational risk factors for bladder cancer in Germany. Cancer 69:17761790. Lindstedt G, Sollenberg J( 1982): Polycyclic aromatic hydrocarbons in the occupational environment. Scand J Work Environ Health 8:l-19. Lloyd JM (1971): Long-term mortality study of steelworkers. V. Respiratory cancer in coke plant workers. J Occup Med 1353-68. Malker HSR, McLaughlin JK, Silverman DT, Ericsson JLE,Stone BJ, Weiner JA, Maker BK, Blot WJ (1987): Occupational risks for bladder cancer among men in Sweden. Cancer Res 47:6763-6766. Mallin K (1990): Investigation of a bladder cancer cluster in Northwestern Illinois. Am J Epidemiol 132(suppl):S96-S106. Mantel N, Haenszel W (1959): Statistical aspects of the analysis of data from retrospective studiesof disease. J Natl Cancer Inst 22:719-748. Mastrangelo G, Fadda E, Marzia V (1996): Polycylic aromatic hydrocarbons and cancer in man. Environ Health Persp 104:1166-1170. Miettinen OS (1976): Estimability and estimation in case-referent studies. Am J Epidemiol 103:226-235. Nadon L, Siemiatycki J, Dewar R, Krewski D, Gerin M (1995):Cancerrisk due to occupational exposure to polycyclic aromatic hydrocarbons. Am J Ind Med 28:303-324. Risch HA, Burch JD, Miller AB, Hill GB, Steele R, Howe GR (1988): Occupational factors and the incidence of cancer of the bladder in Canada. Br J Ind Med 45:361-367. Schifflers E, Jamart J, Renard V (1987): Tobacco and occupation as risk factors in bladder cancer: A case-control study in southern Belgium. Int J Cancer 39:287-292. Sherson D, Svane 0, Lynge E (1991): Cancer incidence among foundry workers in Denmark.Arch Environ Health 4675-81. Silverman DT, Hoover RN, Albert S, Graff KM (1983). Occupation and cancer of the lower urinary tract in Detroit. J Natl Cancer Inst 70:237-245. Silverman DT, Levin LI, Hoover RN, Hartge P (1989a): Occupational risks of bladder cancer in the United States: I. White men. J Natl Cancer Inst 19:1472-1480. Silverman DT, Levin LI, Hoover RN (1989b): Occupational risks of bladder cancer in the United States. II. Nonwhite men. J Natl Cancer Inst 19:1480-1 483. Spinelli JJ, Band P,Svirchev, Gallagher RF' (1991): Mortality and cancer incidence in aluminum reduction plant workers. J Occup Med 33:11501155. Steenland K, Beaumont J (1989): Further follow-up and adjustment for smoking in a study of lung cancer and acid mists. Am J Ind Med 16347-354. Steenland K, Bumett C, OsorioAM (1987): Acase-control study of bladder cancer using city directories as a source of occupational data. Am J Epidemiol 126247-257. Steineck G, Plat0 N, Alfredsson L, Norell S (1989); Industry-related urothelial carcinogens:Application of a job-exposurematrix to census data. Am J Ind Med 16:209-224. Tanimura H (1968): Benzo(a)pyrene in an iron and steel works. Arch Environ Health 17:172-177. Thdriault G, Tremblay C, Cordier S, Gingras S (1984): Bladder cancer in the aluminum industry.Lancet Vol. 1 (8383) 947-950. Zahm SH, Hartge P, Hoover R (1987): The national bladder cancer study: Employmentin the chemical industry. J Natl Cancer Inst 79:217-221.