American Journal of Industrial Medicine 21S77-885 (1992)
Bladder Cancer and Occupation in Shanghai, 1980-1 984 Wei Zheng, MD, MPH, Joseph K. McLaughlin, PhD, Yu-Tang Gao, MD, Debra T. Silverman, SCD, Ru-Nie Gao, MD, and William J. Blot, PhD
To investigate occupational determinants of bladder cancer in the urban area of Shanghai, occupation and industry information for 1,219 incident bladder cancer cases diagnosed during the period 1980 to 1984 were compared with 1982 census data on employment. Standardized incidence ratios (SIR) for bladder cancer were estimated for occupation and industry classifications. Significant excess risks were observed for plastic products workers (male: SIR = 432; female: SIR = 368); textile bleachers, dyers, and finishers (male: SIR = 169); metal refining and processing workers (male: SIR = 139; female: SIR = 197); petroleum refining workers (male: SIR = 2152); railway engine drivers and firemen (male: SIR = 683); and workers employed in industries of apparel and other textile products manufacturing (female: SIR = 204); paper processing (male: SIR = 146; female: SIR = 226); organic chemical manufacturing (male: SIR = 186); plastic product manufacturing (male: SIR = 218; female: SIR = 272); and metallurgy (male: SIR = 107; female: SIR = 561). This study indicates that many of the industries and occupations that are responsible for increased risk throughout the world are also associated with occupational bladder cancer in Shanghai. 0 1992 Wiley-Liss, Inc.
Key words: bladder cancer, occupation, China, plastics and chemical workers
INTRODUCTION Bladder cancer is one of the most common cancers in the world, particularly in industrialized countries. A number of occupational hazards have been etiologically linked to bladder cancer, especially beta-naphthylamine, benzidine, and 4-aminobiphenyl [Silverman et al., in press]. Information on risks associated with other workplace hazards, however, is less consistent, although it is generally thought that occupational factors play an important role in the etiology of this cancer. The present study links employment data from incident bladder cancer cases with census information for urban Shanghai, the largest industrial city in China, to generate and test hypotheses related to employment and bladder cancer risk. Department of Epidemiology, Shanghai Cancer Institute, Shanghai, China (W.Z., Y.-T.G., R.-N.G.). Epidemiology and Biostatistics Program, National Cancer Institute, EPN-415, Bethesda, MD (W.Z., J.K.M., D.T.S., W.J.B.). Address reprint requests to Dr. Joseph K. McLaughlin, Epidemiology and Biostatistics Program, National Cancer Institute, EPN-415, Bethesda, MD 20892. Accepted for publication October 28, 1991. 0 1992 Wiley-Liss, Inc.
878
Zheng et al.
MATERIALS AND METHODS
Details on the methods used in this occupation-cancer linkage study have been published elsewhere [Gao et al., 1990; Kneller et al., 19901. Briefly, employment information at the time of cancer diagnosis, obtained by direct or next-of-kin interviews, for all incident cases of bladder cancer aged 30 and over, diagnosed during the period 1980 to 1984 among residents of urban Shanghai were compared with occupation data from the 1982 census for the same population. All bladder cancer cases were identified through the Shanghai Cancer Registry (SCR), a population-based cancer registry covering urban Shanghai. Urban Shanghai is composed of 10 districts, with 1,574,948 men and 1,546,126 women 30 years of age and older at the time of the 1982 census. The age-, sex-, and site-specific cancer incidence rates for urban Shanghai were calculated for 1980-1984 based on the cancer incidence data collected by the SCR. To calculate the expected number of cancer cases in a particular occupation or industry group, these incidence rates were applied to the age- and sexspecific person-years estimated for 1980-1984 in each occupation and industry category based on the 1982 census. The observed number of cases in each occupation and industry category was divided by the expected number of cases and multiplied by 100 to obtain standardized incidence ratios (SIRs). Statistical significance was tested, assuming that the observed number of cancer cases followed a Poisson distribution [Breslow and Day, 19871. P values less than 0.10 were considered statistically significant in this exploratory analysis. The standardized coding scheme from the 1982 census was used to code the employment data obtained from the cancer patients [National Bureau of Statistics, 19821. The census code grouped occupation and industry according to three levels of increasing detail: single-digit codes reflect broad occupation and industry categories, whereas two- and three-digit codes provide more-detailed classification of each occupation and industry. Only broad single-digit employment categories with 5 or more expected cases for either men or women, and general two-digit categories with 5 or more (for males) and 3 or more (for females) expected cases were included in the tables. Data for specific three-digit categories are presented in the text to clarify associations observed for broader categories or to evaluate jobs of a priori interest not adequately described at the two-digit level. Twenty bladder cancer patients (16 males and 4 females) were excluded from the analysis because of unclassified occupation, as were 159 patients (20 males and 139 females) classified as houseworkers at the time of diagnosis. RESULTS
Between 1980 and 1984, 1031 men and 188 women, for whom employment classification was available were diagnosed with bladder cancer in urban Shanghai. Table I presents SIRs for the broad one-digit level of occupation and industry. For occupation, clerical and related workers had an excess risk of bladder cancer among both men and women, although the SIR was statistically significant only among men. A significant excess bladder cancer risk was observed among women employed as craftsmen and related workers, but not among men. No statistically elevated risk of bladder cancer was observed among men at the one-digit level of industry categories, whereas women employed in manufacturing and construction industries were found
Occupational Bladder Cancer in Shanghai
879
TABLE I. Standardized Incidence Ratios (SIRs) for Bladder Cancer by Broad (1-digit) Occupation and Industry in Urban Shanghai, 198G1984 ~~
Men Code Occupations 011
2
3 4 5 6 7/819 Industries" 1
4 6 7 8
9 10
11 12 13 14
Category Professional and technical workers, scientists, and research workers Leaders of government agencies, and businesses branches, parties, mass organizations Clerical and related workers Commercial workers (sales workers) Service workers Agricultural workers, foresters, animal hunters, husbandry workers, fishermen Craftsmen and related workers Agriculture, husbandry, forestry and fisheries Manufacturing Construction industry Transportation, postal, and communication services Commercial trade, restaurant and goods supply, and marketing services Housing, public utilities, administration, and personal services Health, athletic, and social welfare services Educational, cultural, and art activities Scientific research and technical services Finance and insurance services Government agencies, parties, and mass organizations -
Women
Cases
SIR
Cases
SIR
143
98
17
106
96
6 6 8 36
151 158 63 127
66 69 85 169
86 103
6 493
64
1
100
114
10 532 33
101
1
106 95
I32 4
57 123a 31Sb
102
88
3
94
181
99
16
77
44
103
10
121
20 45 11 12
122 85 87 109
4 12 2 0
78 123 260
37
94
4
99
132a
55 119b
0
"0.01 < p 5 0.05. b0.05 < p 5 0.10. 'Among men only four cases occurred in electric, power, gas, and water supply and service industry (code 3). No cases occurred in mining (code 2) or geological exploration and survey industries (code 5).
to have significantly elevated SIRs. Because of their opportunity for a large number of exposures in the workplace, the remainder of the analysis will focus on craftsmen and workers in related occupations, and employees in manufacturing and construction. Table I1 presents SIRs for general (two-digit) occupations of craftsmen and related workers. A 39% excess risk of bladder cancer was found among male metal refining and processing workers (p < 0. lo), with the SIR highest among the specific (three-digit) category of metal platers and coaters (8 cases, SIR = 249, p < 0.05). Although chemical processors and related workers had a non-significant elevated SIR of 153, petroleum workers within this general category were at high risk for bladder cancer (3 cases, SIR = 2152, p < 0.01). Rubber and plastics product makers had a significant excess of bladder cancer (SIR = 208, p < 0.05). There were four specific
880
Zheng et al.
TABLE II. Standardized Incidence Ratios (SIRs) for Bladder Cancer by General (2-Digit) Craftsmen and Related Occupations Among Men in Urban Shanghai, 1980-1984 General occupation
Cases
SIR
Metal refining and processing workers Chemical processors and related workers Rubber and plastics product makers Textile workers (spinners, weavers, knitters, dyers, and related workers) Leather and fur processors Tailors Food and beverage processors Wood preparation workers, cabinet makers and bamboo, hemp, rattan, and straw product makers Printers and related workers Toolmakers, machine-tool operators, and blacksmiths Machinery fitters, machine assemblers, and precision instrument makers (except electrical) Electrical fitters and related electrical and electronic workers Plumbers, welders, and sheet metal and structural metal preparers and erectors Painters Other production and related workers Construction workers Stationary engine and related equipment operators Material handling and related equipment operators, dockers, and freight handlers Transportation equipment operators Inspectors and product analysis Other transportation and production workers
34 14 16 45 I 15 8
139” 153 2011b 105 132 73 65
29 9
101 103
44
100
26 16
95 82
16 12 9 21 I
96 120 85 110 60
50 36 14 51
97 15a 99 104
Code 12 73 74 75 16
I1 78 80 82 84 85 86 88 90 91 92 93 94 95 96 99
~~
aO.05 < p bO.O1 < p
5 5
0.10. 0.05.
occupations among rubber and plastic product makers: rubber manufacturing and product makers (5 cases, SIR = 142, p > 0.10); plastic manufacturing and product makers ( 9 cases, SIR = 432, p < 0.01); tire makers, vulcanizers, and repairmen (2 cases, SIR = 317, p > 0.10); and other rubber and plastic workers (0 cases). No excess risk was observed for textile workers overall (45 cases, SIR = 105), but the risks were elevated among textile bleachers, dyers, and product finishers (14 cases, SIR = 169, p < 0.10); textile machinery mechanics (16 cases, SIR = 136, p > 0.10); and knitters (4 cases, SIR = 112, p > 0.10). The SIRs for the specific occupations of weavers, spinners, and winders were all less than 100. Metal workers and mechanics (OCC. 84, 85) were not at increased risk of bladder cancer, with the SIRS for the 9 specific categories within these two general occupations between 66-1 10. Although transportation equipment operators overall were at a significantly low risk, within this occupation, the specific occupations of railway engine drivers and firemen were significantly elevated (3 cases, SIR = 683, p < 0.05), while no increased risk was found among automobile and truck drivers (15 cases, SIR = 101). Table I11 shows SIRs for all general (two-digit) manufacturing and construction industries among men. A significant SIR was observed for rubber and plastics manufacturing, and within this general industry, risk for the specific industry plastics products manufacturing was particularly elevated (13 cases, SIR = 218, p < 0.05).
Occupational Bladder Cancer in Shanghai
881
TABLE 111. Standardized Incidence Ratios (SIRs) for Bladder Cancer by General (2-Digit) Manufacturing and Construction Industries Among Men in Urban Shanghai, 1980-1984 Code
General industry
Cases
SIR
18/19 23 24 25 26 27 28 29 31 34 37 39 4014 1 42/43 44/45 47 48/49 50 52
Food products Textile mill products Apparel and other textile products (sewing manufacturing) Leather, fur, and leather products Lumber and wood products Furniture manufacturing Paper and allied products Stationary, musical instruments, crafts, sporting goods, and printing Chemical industry Rubber and plastic products Nonmetallic minerals products, except insulating materials Metallurgy industry Fabricated metal products Machinery manufacturing (except electrical) Electrical and electronic machinery and equipment Transportation equipment manufacturing and repairing Precision machinery, instruments, and meters Other manufacturing and repairing services Construction services (residential. commercial, heavy construction, etc .)
20 90 16 9 9 9 12 24 31 25 22 35 29 89 41 31 22 8 29
82 94 79 120 136 179 122 91 132
aO.05< p
5
152a
140 107 139 104 102 99 126 102 97
0.10.
Elevated but non-significant risks were also seen for furniture manufacturing (SIR = 179), lumber and wood products (SIR = 176), nonmetalic mineral products (SIR = 140), fabricated metal products (SIR = 139), the chemical industry (SIR = 132) and precision machinery and instruments (SIR = 126). Within the chemical industry, the specific industry of organic chemical production was associated with a significantly elevated risk (14 cases, SIR = 186, p < 0.05). Table IV lists SIRs for all general (two-digit) craftsmen and related occupations and manufacturing and construction industries among women. Two occupations were significantly elevated in risk: paper and paper product workers (10 case, SIR = 206, p < 0. lo), and toolmakers, machine tool operators, and blacksmiths (1 1 cases, SIR = 185, p < 0.10). Non-significant excesses were also seen for women in rubber and plastic products making (SIR = 189), tailoring (SIR = 156), and inspecting and product analysis (SIR = 146). Within the general category of rubber and plastic product makers, the specific occupation of plastic product makers was at a significantly increased risk (6 cases, SIR = 368, p < 0.05). For industry, significant SIRs were found for apparel and other textile products (14 cases, SIR = 204, p < 0.05), paper and allied products (1 1 cases, SIR = 215, p < 0.05), the metallurgy industry (8 cases, SIR = 561, p < O.Ol), and construction services (4 cases, SIR = 373, p < 0.05). Two specific industry categories within the metallurgy category were statistically increased: ferrous metal refining and processing (5 cases, SIR = 496, p < 0.01) and non-ferrous metal refining and processing (3 cases, SIR = 726, p < 0.05). Although the overall SIR for the general rubber and plastic industry was not statistically significant (9 cases, SIR = 189, p > 0. lo), workers in the plastic industry were at a significantly elevated risk (8 cases, SIR = 272, p < 0.05).
882
Zheng et al.
TABLE IV. Standardized Incidence Ratios (SIRS)for Bladder Cancer, by General (2-Digit) Craftsmen and Related Occupations, and Manufacturing, Construction, and Transportation Industries Among Women in Urban Shanghai, 1980-1984 Code Occupations 74 75 77 81 84 86 96 99 Industries 23 24 28 29 34 39 4014 1 42/43 44/45 48/49 50 52
General occupation
Cases
SIR
Rubber and plastics product makers Textile workers Tailors Paper and paper product makers Toolmakers, machine-tool operators, and blacksmiths Electrical fitters and related electrical and electronic workers Inspectors and product analysis Other transportation and production workers
7 27 14 10 11 4 5 7
188 89 156 206= 185a 71 146 110
Textile mill products Apparel and other textile products (sewing manufacturing) Paper and allied products Stationery, musical instruments, crafts, sporting goods, and printing Rubber and plastic products Metallurgy industry Fabricated metal products Machinery manufacturing (except electrical) Electric and electronic machinery and equipment Precision machinery, instruments, and meters Other manufacturing and repairing services Construction services
33 14 11
92 204b 215b 168 189 561' 157 97 125 59 66 373b
11
9 8 11 6 14 2 2 4
"0.05 < p 5 0.10. bO.O1 < p 5 0.05. 'p 5 0.01.
DISCUSSION This occupation-cancer record linkage survey was conducted in Shanghai, the largest industrial city in China. The wide variety of industries enabled us to examine occupational bladder cancer in a newly developing industrial setting. We found several occupations and industries with an increased risk of bladder cancer. Although some findings may be due to chance, as a result of multiple comparisons, most are in agreement with findings from previous epidemiologic studies. The most consistent finding in this study was for workers in plastic manufacturing and processing. Significantly increased SIRS were observed among both men and women. Such internal consistency suggests that this finding may be signalling an occupational hazard, but few studies on the association of bladder cancer risk and plastic workers have been published. In the U.S. National Bladder Cancer Study, a three-fold elevated risk of bladder cancer was observed among women employed in the plastics industry (5 cases and 3 controls), although little or no excess was observed among men [Zahm et al., 19871. A number of studies, however, observed a significant excess risk among workers in the chemical, especially organic chemical industry, which often includes the plastic industry [Malker et al., 1987; Zahm et al., 1987; Silverman et al., 1990; La Vecchia et al., 19901. Workers in the organic chemical industry in our study were also found to have a significant excess risk of bladder cancer.
Occupational Bladder Cancer in Shanghai
883
We observed no significant increased risk of bladder cancer among textile workers overall, but a 69% excess among male bleachers, dyers, and product finishers. Such an excess risk may be due to dye exposure. Although employment in the synthetic dye industry was found to increase risk of bladder cancer nearly a century ago, bladder cancer in dye users has not been widely reported recently [Schulte et al., 19871, and some studies have found no excess risk of bladder cancer among workers exposed to finished dyes [Silverman et al., 1989b; You et al., 19901. On the other hand, a number of jobs with potential dye exposures have been associated with elevated risks of bladder cancer [Schoenberg et al., 1984; Malker et al., 1987; Steenland et al., 1987; Silverman et al., 1989a,b; Schumacher et al,. 1989; Gonzalez et al., 1988, 1989; La Vecchia et al., 19901. Among them, textile workers, painters, and printing and leather workers are most widely reported in the literature. Since aromatic amine-based dyes could be metabolized back to their parent compound, such as benzidine [Lynn et al., 19801, finished as well as intermediate dyes may be involved in bladder carcinogenesis. In addition, our study also found other workers with possible dye exposure, such as painters, printing pressmen, and leather and fur processors, to be at elevated risk, although the SIRS were not statistically significant. Thus, our findings generally support the hypothesis that dye exposures may increase the risk of bladder cancer. We found increased risks of bladder cancer for employment in metal refining and processing and the metallurgy industry. Similar results were seen by Steenland [ 19871 and Silverman et al. [ 1989b] among U.S. workers. We observed increased risks among petroleum refining workers and railway engine drivers and firemen, also consistent with previous findings [Steenland et al., 1987; Silverman et al., 1989bl. No increase in risk was observed among automobile and truck drivers, in contrast to findings in other countries, where truck drivers have been reported to be at significant risk for bladder cancer [Silverman, et al. 1989bl. Our other findings were also consistent with earlier epidemiologic studies. Clerical and related workers had a significantly elevated risk, as was reported in case-control studies in the U.S. [Cole et al., 1971; Schumacher et al., 19891, and in an occupation-cancer record linkage study in Sweden [Malker et al., 19871. The significantly increased risks for women employed in apparel and other textile products and in paper and allied products were consistent with the finding from the Swedish study [Malker et al., 19871, although we found no increase among men. The limitations of this study should be considered when interpreting the results. First, we were not able to adjust for the possible confounding effect of other risk factors, notably cigarette smoking [International Agency for Research on Cancer, 19861. The excess risk in certain occupations may be simply due to a high prevalence of cigarette smoking in these occupations. We have ancillary data to address this issue, however. Based on a survey conducted among 110,000 residents aged 30 and over in the Shanghai urban area in 1982, the prevalence of smoking in the general population was 48.0% for men and 8.1% for women [Gao and Deng, unpublished data]. Smoking prevalences for the occupations at elevated risk of bladder cancer were not significantly different from the sex-specific averages, ranging from 40-57% for men and 743% for women (Table V). Second, only employment at the time of diagnosis was used in this study, which may not be the usual occupation the study subjects held during their lifetime. However, we have less concern about this issue because workers in Shanghai rarely change jobs, especially the work unit or industry.
884
Zheng et al.
TABLE V. Prevalence of Smoking for Selected Occupation Categories in Urban Shanghai, 1982 Code 3 71819 72 726 73 736 74 742 756 952
Occupation Clerical and related worker Production, transportation, and related worker Metal refining and processing worker Metal plater and coater Chemical processor and related worker Petroleum refining worker Rubber and plastic product maker Plastic manufacturing and product maker Textile bleacher, dyer, and product finisher Railway engine driver and fireman
Male
Female
(%)
(%)
40.3 49.4 53.2 53.1 37.3 50.0 51.0 40.4 49.5 43.4
3.7 5.8 6.2 3.4 3.0 7.6 8.1 7.5 NIA” NIA”
aData were not available due to small sample size.
A small validation study was conducted to determine whether or not occupational data obtained for the cancer cases were consistent with employment information obtained in the census. The study showed concordance at the two-digit level of 90.5% and 91.3% for males and females, respectively [Gao et al., 19901. We had no information, however, on the duration of employment in the occupation or industry. Finally, the analysis was limited to occupation and industry titles, with no direct information on exposure agents. In summary, our investigation has identified occupations and industries in Shanghai at increased risk for bladder cancer, consistent with findings from other countries, thereby indicating that occupational factors likely are playing a role in this Chinese industrial city, similar to those observed in other areas of the world for this cancer. ACKNOWLEDGMENTS This work was supported by the U.S. National Cancer Institute and Shanghai Municipal Committee for Science and Technology. We thank M.H. Liu, J.P. Sheng, D.L. Qin, and Y.S. Wang for computer support, and H. Brown, D. Eyler, and N. Connor for manuscript preparation.
REFERENCES Breslow NE, Day NE (1987): “Statistical Methods in Cancer Research. Vol. 11. The Design and Analysis of Cohort Studies.” Lyon: IARC Sci. Pub. 82. Cole P, Hoover R, Friedell GH (1971): Occupation and cancer of the lower urinary tract. Cancer 29: 1250-1260. Gao YT, McLaughlin JK, Gao RN, Kneller RW, Liu MH, Sheng JP, Blot WJ (1990): Investigation of occupational cancer incidence in Shanghai: Background and methods. Shanghai Tumor 10:49-53. Gonzalez CA, Lopez-Abente G, Errezola M, Escolar A, Riboli E, Izarzugaza I, Nebot M (1989): Occupation and bladder cancer in Spain: A multi-center case-control study. Int J Epidemiol 18~569-577. Gonzales CA, Riboli E, Lopez-Abente G (1988): Bladder cancer among workers in the textile industry: Results of a Spanish case-control study. Am J Ind Med 14:673-680. International Agency for Research on Cancer (1986): “Tobacco Smoking.” IARC Monogr 38:244-268.
Occupational Bladder Cancer in Shanghai
885
Kneller RW, Gao YT, McLaughlin JK, Gao RN, Blot WJ, Liu MH, Sheng JP, Fraumeni JF Jr (1990): Occupational risk factors for gastric cancer in Shanghai, China. Am J Ind Med 18:69-78. La Vecchia C, Negri E, D’Avanzo B, Franceschi S (1990): Occupation and the risk of bladder cancer. Int J Epidemiol 2:264-268. Lynn RK, Donielson DW, Ilias AM, Kennish JM, Wong K, Matthews HB (1980): Metabolism of bisazophenyl dyes derived from benzidine, 3,3’-dimethylbenzidine or 3,3’-dimethoxybenzidine to carcinogenic aromatic amines in the dog and rat. Toxic01 Appl Pharmacol 56:248-258. Malker HSR, McLaughlin JK, Silverman DT, Ericsson JLE, Stone BJ, Weiner JS, Malker BK, Blot WJ (1987): Occupational risks for bladder cancer among men in Sweden. Cancer Res 47:6763-6766. National Bureau of Statistics (1982): “Standard Classification of Industries and Occupations Used for the Third National Census.” Beijing; Office of the National Census of the State Council. Schoenberg JB, Stemhagen A, Mogielnicki AP, Altman R, Abe T, Mason TJ (1984): Case-control study of bladder cancer in New Jersey. I. Occupational exposures in white males. J Natl Cancer Inst 72:973-981. Schumacher MC, Slattery ML, West DW (1989): Occupation and bladder cancer in Utah. Am J Ind Med 16:89-102. Schulte PA, Ringen K, Hemstreet GP, Ward E (1987): Occupational cancer of the urinary tract. Occup Med State of the Art Reviews 2:85-107. Silverman DT, McLaughlin JK, Malker HSR, Weiner JS, Ericsson JLE (1989a): Bladder cancer and occupation among Swedish women. Am J Ind Med 16:239-240. Silverman DT, Levin LI, Hoover RN, Hartge P (1989b): Occupational risks of bladder cancer in the United States: I. White Men. J Natl Cancer Inst 81:1472-1480. Silverman DT, Levin LI, Hoover RN (1990): Occupational risks of bladder cancer among white women in the United States. Am J Epidemiol 132:453-461. Silverman DT, Momson AS, Devesa SS (1991): Bladder cancer. In Schottenfeld D, Fraumeni JF Jr (eds), “Cancer Epidemiology and Prevention, ” 2nd edition. Philadelphia: Saunders (in press). Steenland K, Burnett C, Osorio A (1984): A case-control study of bladder cancer using city directories as a source of occupational data. Am J Epidemiol 126:247-257. You XY, Chen JG, Hu YN (1990): Studies on the relation between bladder cancer and benzidine or its derived dyes in Shanghai. Br J Ind Med 47:544-552. Zahm SH, Hartge P, Hoover R (1987): The National Bladder Cancer Study: Employment in the chemical industry. J Natl Cancer Inst 79:217-222.
Bladder Cancer and Occupation in Shanghai, 1980-1 984 Wei Zheng, MD, MPH, Joseph K. McLaughlin, PhD, Yu-Tang Gao, MD, Debra T. Silverman, SCD, Ru-Nie Gao, MD, and William J. Blot, PhD
To investigate occupational determinants of bladder cancer in the urban area of Shanghai, occupation and industry information for 1,219 incident bladder cancer cases diagnosed during the period 1980 to 1984 were compared with 1982 census data on employment. Standardized incidence ratios (SIR) for bladder cancer were estimated for occupation and industry classifications. Significant excess risks were observed for plastic products workers (male: SIR = 432; female: SIR = 368); textile bleachers, dyers, and finishers (male: SIR = 169); metal refining and processing workers (male: SIR = 139; female: SIR = 197); petroleum refining workers (male: SIR = 2152); railway engine drivers and firemen (male: SIR = 683); and workers employed in industries of apparel and other textile products manufacturing (female: SIR = 204); paper processing (male: SIR = 146; female: SIR = 226); organic chemical manufacturing (male: SIR = 186); plastic product manufacturing (male: SIR = 218; female: SIR = 272); and metallurgy (male: SIR = 107; female: SIR = 561). This study indicates that many of the industries and occupations that are responsible for increased risk throughout the world are also associated with occupational bladder cancer in Shanghai. 0 1992 Wiley-Liss, Inc.
Key words: bladder cancer, occupation, China, plastics and chemical workers
INTRODUCTION Bladder cancer is one of the most common cancers in the world, particularly in industrialized countries. A number of occupational hazards have been etiologically linked to bladder cancer, especially beta-naphthylamine, benzidine, and 4-aminobiphenyl [Silverman et al., in press]. Information on risks associated with other workplace hazards, however, is less consistent, although it is generally thought that occupational factors play an important role in the etiology of this cancer. The present study links employment data from incident bladder cancer cases with census information for urban Shanghai, the largest industrial city in China, to generate and test hypotheses related to employment and bladder cancer risk. Department of Epidemiology, Shanghai Cancer Institute, Shanghai, China (W.Z., Y.-T.G., R.-N.G.). Epidemiology and Biostatistics Program, National Cancer Institute, EPN-415, Bethesda, MD (W.Z., J.K.M., D.T.S., W.J.B.). Address reprint requests to Dr. Joseph K. McLaughlin, Epidemiology and Biostatistics Program, National Cancer Institute, EPN-415, Bethesda, MD 20892. Accepted for publication October 28, 1991. 0 1992 Wiley-Liss, Inc.
878
Zheng et al.
MATERIALS AND METHODS
Details on the methods used in this occupation-cancer linkage study have been published elsewhere [Gao et al., 1990; Kneller et al., 19901. Briefly, employment information at the time of cancer diagnosis, obtained by direct or next-of-kin interviews, for all incident cases of bladder cancer aged 30 and over, diagnosed during the period 1980 to 1984 among residents of urban Shanghai were compared with occupation data from the 1982 census for the same population. All bladder cancer cases were identified through the Shanghai Cancer Registry (SCR), a population-based cancer registry covering urban Shanghai. Urban Shanghai is composed of 10 districts, with 1,574,948 men and 1,546,126 women 30 years of age and older at the time of the 1982 census. The age-, sex-, and site-specific cancer incidence rates for urban Shanghai were calculated for 1980-1984 based on the cancer incidence data collected by the SCR. To calculate the expected number of cancer cases in a particular occupation or industry group, these incidence rates were applied to the age- and sexspecific person-years estimated for 1980-1984 in each occupation and industry category based on the 1982 census. The observed number of cases in each occupation and industry category was divided by the expected number of cases and multiplied by 100 to obtain standardized incidence ratios (SIRs). Statistical significance was tested, assuming that the observed number of cancer cases followed a Poisson distribution [Breslow and Day, 19871. P values less than 0.10 were considered statistically significant in this exploratory analysis. The standardized coding scheme from the 1982 census was used to code the employment data obtained from the cancer patients [National Bureau of Statistics, 19821. The census code grouped occupation and industry according to three levels of increasing detail: single-digit codes reflect broad occupation and industry categories, whereas two- and three-digit codes provide more-detailed classification of each occupation and industry. Only broad single-digit employment categories with 5 or more expected cases for either men or women, and general two-digit categories with 5 or more (for males) and 3 or more (for females) expected cases were included in the tables. Data for specific three-digit categories are presented in the text to clarify associations observed for broader categories or to evaluate jobs of a priori interest not adequately described at the two-digit level. Twenty bladder cancer patients (16 males and 4 females) were excluded from the analysis because of unclassified occupation, as were 159 patients (20 males and 139 females) classified as houseworkers at the time of diagnosis. RESULTS
Between 1980 and 1984, 1031 men and 188 women, for whom employment classification was available were diagnosed with bladder cancer in urban Shanghai. Table I presents SIRs for the broad one-digit level of occupation and industry. For occupation, clerical and related workers had an excess risk of bladder cancer among both men and women, although the SIR was statistically significant only among men. A significant excess bladder cancer risk was observed among women employed as craftsmen and related workers, but not among men. No statistically elevated risk of bladder cancer was observed among men at the one-digit level of industry categories, whereas women employed in manufacturing and construction industries were found
Occupational Bladder Cancer in Shanghai
879
TABLE I. Standardized Incidence Ratios (SIRs) for Bladder Cancer by Broad (1-digit) Occupation and Industry in Urban Shanghai, 198G1984 ~~
Men Code Occupations 011
2
3 4 5 6 7/819 Industries" 1
4 6 7 8
9 10
11 12 13 14
Category Professional and technical workers, scientists, and research workers Leaders of government agencies, and businesses branches, parties, mass organizations Clerical and related workers Commercial workers (sales workers) Service workers Agricultural workers, foresters, animal hunters, husbandry workers, fishermen Craftsmen and related workers Agriculture, husbandry, forestry and fisheries Manufacturing Construction industry Transportation, postal, and communication services Commercial trade, restaurant and goods supply, and marketing services Housing, public utilities, administration, and personal services Health, athletic, and social welfare services Educational, cultural, and art activities Scientific research and technical services Finance and insurance services Government agencies, parties, and mass organizations -
Women
Cases
SIR
Cases
SIR
143
98
17
106
96
6 6 8 36
151 158 63 127
66 69 85 169
86 103
6 493
64
1
100
114
10 532 33
101
1
106 95
I32 4
57 123a 31Sb
102
88
3
94
181
99
16
77
44
103
10
121
20 45 11 12
122 85 87 109
4 12 2 0
78 123 260
37
94
4
99
132a
55 119b
0
"0.01 < p 5 0.05. b0.05 < p 5 0.10. 'Among men only four cases occurred in electric, power, gas, and water supply and service industry (code 3). No cases occurred in mining (code 2) or geological exploration and survey industries (code 5).
to have significantly elevated SIRs. Because of their opportunity for a large number of exposures in the workplace, the remainder of the analysis will focus on craftsmen and workers in related occupations, and employees in manufacturing and construction. Table I1 presents SIRs for general (two-digit) occupations of craftsmen and related workers. A 39% excess risk of bladder cancer was found among male metal refining and processing workers (p < 0. lo), with the SIR highest among the specific (three-digit) category of metal platers and coaters (8 cases, SIR = 249, p < 0.05). Although chemical processors and related workers had a non-significant elevated SIR of 153, petroleum workers within this general category were at high risk for bladder cancer (3 cases, SIR = 2152, p < 0.01). Rubber and plastics product makers had a significant excess of bladder cancer (SIR = 208, p < 0.05). There were four specific
880
Zheng et al.
TABLE II. Standardized Incidence Ratios (SIRs) for Bladder Cancer by General (2-Digit) Craftsmen and Related Occupations Among Men in Urban Shanghai, 1980-1984 General occupation
Cases
SIR
Metal refining and processing workers Chemical processors and related workers Rubber and plastics product makers Textile workers (spinners, weavers, knitters, dyers, and related workers) Leather and fur processors Tailors Food and beverage processors Wood preparation workers, cabinet makers and bamboo, hemp, rattan, and straw product makers Printers and related workers Toolmakers, machine-tool operators, and blacksmiths Machinery fitters, machine assemblers, and precision instrument makers (except electrical) Electrical fitters and related electrical and electronic workers Plumbers, welders, and sheet metal and structural metal preparers and erectors Painters Other production and related workers Construction workers Stationary engine and related equipment operators Material handling and related equipment operators, dockers, and freight handlers Transportation equipment operators Inspectors and product analysis Other transportation and production workers
34 14 16 45 I 15 8
139” 153 2011b 105 132 73 65
29 9
101 103
44
100
26 16
95 82
16 12 9 21 I
96 120 85 110 60
50 36 14 51
97 15a 99 104
Code 12 73 74 75 16
I1 78 80 82 84 85 86 88 90 91 92 93 94 95 96 99
~~
aO.05 < p bO.O1 < p
5 5
0.10. 0.05.
occupations among rubber and plastic product makers: rubber manufacturing and product makers (5 cases, SIR = 142, p > 0.10); plastic manufacturing and product makers ( 9 cases, SIR = 432, p < 0.01); tire makers, vulcanizers, and repairmen (2 cases, SIR = 317, p > 0.10); and other rubber and plastic workers (0 cases). No excess risk was observed for textile workers overall (45 cases, SIR = 105), but the risks were elevated among textile bleachers, dyers, and product finishers (14 cases, SIR = 169, p < 0.10); textile machinery mechanics (16 cases, SIR = 136, p > 0.10); and knitters (4 cases, SIR = 112, p > 0.10). The SIRs for the specific occupations of weavers, spinners, and winders were all less than 100. Metal workers and mechanics (OCC. 84, 85) were not at increased risk of bladder cancer, with the SIRS for the 9 specific categories within these two general occupations between 66-1 10. Although transportation equipment operators overall were at a significantly low risk, within this occupation, the specific occupations of railway engine drivers and firemen were significantly elevated (3 cases, SIR = 683, p < 0.05), while no increased risk was found among automobile and truck drivers (15 cases, SIR = 101). Table I11 shows SIRs for all general (two-digit) manufacturing and construction industries among men. A significant SIR was observed for rubber and plastics manufacturing, and within this general industry, risk for the specific industry plastics products manufacturing was particularly elevated (13 cases, SIR = 218, p < 0.05).
Occupational Bladder Cancer in Shanghai
881
TABLE 111. Standardized Incidence Ratios (SIRs) for Bladder Cancer by General (2-Digit) Manufacturing and Construction Industries Among Men in Urban Shanghai, 1980-1984 Code
General industry
Cases
SIR
18/19 23 24 25 26 27 28 29 31 34 37 39 4014 1 42/43 44/45 47 48/49 50 52
Food products Textile mill products Apparel and other textile products (sewing manufacturing) Leather, fur, and leather products Lumber and wood products Furniture manufacturing Paper and allied products Stationary, musical instruments, crafts, sporting goods, and printing Chemical industry Rubber and plastic products Nonmetallic minerals products, except insulating materials Metallurgy industry Fabricated metal products Machinery manufacturing (except electrical) Electrical and electronic machinery and equipment Transportation equipment manufacturing and repairing Precision machinery, instruments, and meters Other manufacturing and repairing services Construction services (residential. commercial, heavy construction, etc .)
20 90 16 9 9 9 12 24 31 25 22 35 29 89 41 31 22 8 29
82 94 79 120 136 179 122 91 132
aO.05< p
5
152a
140 107 139 104 102 99 126 102 97
0.10.
Elevated but non-significant risks were also seen for furniture manufacturing (SIR = 179), lumber and wood products (SIR = 176), nonmetalic mineral products (SIR = 140), fabricated metal products (SIR = 139), the chemical industry (SIR = 132) and precision machinery and instruments (SIR = 126). Within the chemical industry, the specific industry of organic chemical production was associated with a significantly elevated risk (14 cases, SIR = 186, p < 0.05). Table IV lists SIRs for all general (two-digit) craftsmen and related occupations and manufacturing and construction industries among women. Two occupations were significantly elevated in risk: paper and paper product workers (10 case, SIR = 206, p < 0. lo), and toolmakers, machine tool operators, and blacksmiths (1 1 cases, SIR = 185, p < 0.10). Non-significant excesses were also seen for women in rubber and plastic products making (SIR = 189), tailoring (SIR = 156), and inspecting and product analysis (SIR = 146). Within the general category of rubber and plastic product makers, the specific occupation of plastic product makers was at a significantly increased risk (6 cases, SIR = 368, p < 0.05). For industry, significant SIRs were found for apparel and other textile products (14 cases, SIR = 204, p < 0.05), paper and allied products (1 1 cases, SIR = 215, p < 0.05), the metallurgy industry (8 cases, SIR = 561, p < O.Ol), and construction services (4 cases, SIR = 373, p < 0.05). Two specific industry categories within the metallurgy category were statistically increased: ferrous metal refining and processing (5 cases, SIR = 496, p < 0.01) and non-ferrous metal refining and processing (3 cases, SIR = 726, p < 0.05). Although the overall SIR for the general rubber and plastic industry was not statistically significant (9 cases, SIR = 189, p > 0. lo), workers in the plastic industry were at a significantly elevated risk (8 cases, SIR = 272, p < 0.05).
882
Zheng et al.
TABLE IV. Standardized Incidence Ratios (SIRS)for Bladder Cancer, by General (2-Digit) Craftsmen and Related Occupations, and Manufacturing, Construction, and Transportation Industries Among Women in Urban Shanghai, 1980-1984 Code Occupations 74 75 77 81 84 86 96 99 Industries 23 24 28 29 34 39 4014 1 42/43 44/45 48/49 50 52
General occupation
Cases
SIR
Rubber and plastics product makers Textile workers Tailors Paper and paper product makers Toolmakers, machine-tool operators, and blacksmiths Electrical fitters and related electrical and electronic workers Inspectors and product analysis Other transportation and production workers
7 27 14 10 11 4 5 7
188 89 156 206= 185a 71 146 110
Textile mill products Apparel and other textile products (sewing manufacturing) Paper and allied products Stationery, musical instruments, crafts, sporting goods, and printing Rubber and plastic products Metallurgy industry Fabricated metal products Machinery manufacturing (except electrical) Electric and electronic machinery and equipment Precision machinery, instruments, and meters Other manufacturing and repairing services Construction services
33 14 11
92 204b 215b 168 189 561' 157 97 125 59 66 373b
11
9 8 11 6 14 2 2 4
"0.05 < p 5 0.10. bO.O1 < p 5 0.05. 'p 5 0.01.
DISCUSSION This occupation-cancer record linkage survey was conducted in Shanghai, the largest industrial city in China. The wide variety of industries enabled us to examine occupational bladder cancer in a newly developing industrial setting. We found several occupations and industries with an increased risk of bladder cancer. Although some findings may be due to chance, as a result of multiple comparisons, most are in agreement with findings from previous epidemiologic studies. The most consistent finding in this study was for workers in plastic manufacturing and processing. Significantly increased SIRS were observed among both men and women. Such internal consistency suggests that this finding may be signalling an occupational hazard, but few studies on the association of bladder cancer risk and plastic workers have been published. In the U.S. National Bladder Cancer Study, a three-fold elevated risk of bladder cancer was observed among women employed in the plastics industry (5 cases and 3 controls), although little or no excess was observed among men [Zahm et al., 19871. A number of studies, however, observed a significant excess risk among workers in the chemical, especially organic chemical industry, which often includes the plastic industry [Malker et al., 1987; Zahm et al., 1987; Silverman et al., 1990; La Vecchia et al., 19901. Workers in the organic chemical industry in our study were also found to have a significant excess risk of bladder cancer.
Occupational Bladder Cancer in Shanghai
883
We observed no significant increased risk of bladder cancer among textile workers overall, but a 69% excess among male bleachers, dyers, and product finishers. Such an excess risk may be due to dye exposure. Although employment in the synthetic dye industry was found to increase risk of bladder cancer nearly a century ago, bladder cancer in dye users has not been widely reported recently [Schulte et al., 19871, and some studies have found no excess risk of bladder cancer among workers exposed to finished dyes [Silverman et al., 1989b; You et al., 19901. On the other hand, a number of jobs with potential dye exposures have been associated with elevated risks of bladder cancer [Schoenberg et al., 1984; Malker et al., 1987; Steenland et al., 1987; Silverman et al., 1989a,b; Schumacher et al,. 1989; Gonzalez et al., 1988, 1989; La Vecchia et al., 19901. Among them, textile workers, painters, and printing and leather workers are most widely reported in the literature. Since aromatic amine-based dyes could be metabolized back to their parent compound, such as benzidine [Lynn et al., 19801, finished as well as intermediate dyes may be involved in bladder carcinogenesis. In addition, our study also found other workers with possible dye exposure, such as painters, printing pressmen, and leather and fur processors, to be at elevated risk, although the SIRS were not statistically significant. Thus, our findings generally support the hypothesis that dye exposures may increase the risk of bladder cancer. We found increased risks of bladder cancer for employment in metal refining and processing and the metallurgy industry. Similar results were seen by Steenland [ 19871 and Silverman et al. [ 1989b] among U.S. workers. We observed increased risks among petroleum refining workers and railway engine drivers and firemen, also consistent with previous findings [Steenland et al., 1987; Silverman et al., 1989bl. No increase in risk was observed among automobile and truck drivers, in contrast to findings in other countries, where truck drivers have been reported to be at significant risk for bladder cancer [Silverman, et al. 1989bl. Our other findings were also consistent with earlier epidemiologic studies. Clerical and related workers had a significantly elevated risk, as was reported in case-control studies in the U.S. [Cole et al., 1971; Schumacher et al., 19891, and in an occupation-cancer record linkage study in Sweden [Malker et al., 19871. The significantly increased risks for women employed in apparel and other textile products and in paper and allied products were consistent with the finding from the Swedish study [Malker et al., 19871, although we found no increase among men. The limitations of this study should be considered when interpreting the results. First, we were not able to adjust for the possible confounding effect of other risk factors, notably cigarette smoking [International Agency for Research on Cancer, 19861. The excess risk in certain occupations may be simply due to a high prevalence of cigarette smoking in these occupations. We have ancillary data to address this issue, however. Based on a survey conducted among 110,000 residents aged 30 and over in the Shanghai urban area in 1982, the prevalence of smoking in the general population was 48.0% for men and 8.1% for women [Gao and Deng, unpublished data]. Smoking prevalences for the occupations at elevated risk of bladder cancer were not significantly different from the sex-specific averages, ranging from 40-57% for men and 743% for women (Table V). Second, only employment at the time of diagnosis was used in this study, which may not be the usual occupation the study subjects held during their lifetime. However, we have less concern about this issue because workers in Shanghai rarely change jobs, especially the work unit or industry.
884
Zheng et al.
TABLE V. Prevalence of Smoking for Selected Occupation Categories in Urban Shanghai, 1982 Code 3 71819 72 726 73 736 74 742 756 952
Occupation Clerical and related worker Production, transportation, and related worker Metal refining and processing worker Metal plater and coater Chemical processor and related worker Petroleum refining worker Rubber and plastic product maker Plastic manufacturing and product maker Textile bleacher, dyer, and product finisher Railway engine driver and fireman
Male
Female
(%)
(%)
40.3 49.4 53.2 53.1 37.3 50.0 51.0 40.4 49.5 43.4
3.7 5.8 6.2 3.4 3.0 7.6 8.1 7.5 NIA” NIA”
aData were not available due to small sample size.
A small validation study was conducted to determine whether or not occupational data obtained for the cancer cases were consistent with employment information obtained in the census. The study showed concordance at the two-digit level of 90.5% and 91.3% for males and females, respectively [Gao et al., 19901. We had no information, however, on the duration of employment in the occupation or industry. Finally, the analysis was limited to occupation and industry titles, with no direct information on exposure agents. In summary, our investigation has identified occupations and industries in Shanghai at increased risk for bladder cancer, consistent with findings from other countries, thereby indicating that occupational factors likely are playing a role in this Chinese industrial city, similar to those observed in other areas of the world for this cancer. ACKNOWLEDGMENTS This work was supported by the U.S. National Cancer Institute and Shanghai Municipal Committee for Science and Technology. We thank M.H. Liu, J.P. Sheng, D.L. Qin, and Y.S. Wang for computer support, and H. Brown, D. Eyler, and N. Connor for manuscript preparation.
REFERENCES Breslow NE, Day NE (1987): “Statistical Methods in Cancer Research. Vol. 11. The Design and Analysis of Cohort Studies.” Lyon: IARC Sci. Pub. 82. Cole P, Hoover R, Friedell GH (1971): Occupation and cancer of the lower urinary tract. Cancer 29: 1250-1260. Gao YT, McLaughlin JK, Gao RN, Kneller RW, Liu MH, Sheng JP, Blot WJ (1990): Investigation of occupational cancer incidence in Shanghai: Background and methods. Shanghai Tumor 10:49-53. Gonzalez CA, Lopez-Abente G, Errezola M, Escolar A, Riboli E, Izarzugaza I, Nebot M (1989): Occupation and bladder cancer in Spain: A multi-center case-control study. Int J Epidemiol 18~569-577. Gonzales CA, Riboli E, Lopez-Abente G (1988): Bladder cancer among workers in the textile industry: Results of a Spanish case-control study. Am J Ind Med 14:673-680. International Agency for Research on Cancer (1986): “Tobacco Smoking.” IARC Monogr 38:244-268.
Occupational Bladder Cancer in Shanghai
885
Kneller RW, Gao YT, McLaughlin JK, Gao RN, Blot WJ, Liu MH, Sheng JP, Fraumeni JF Jr (1990): Occupational risk factors for gastric cancer in Shanghai, China. Am J Ind Med 18:69-78. La Vecchia C, Negri E, D’Avanzo B, Franceschi S (1990): Occupation and the risk of bladder cancer. Int J Epidemiol 2:264-268. Lynn RK, Donielson DW, Ilias AM, Kennish JM, Wong K, Matthews HB (1980): Metabolism of bisazophenyl dyes derived from benzidine, 3,3’-dimethylbenzidine or 3,3’-dimethoxybenzidine to carcinogenic aromatic amines in the dog and rat. Toxic01 Appl Pharmacol 56:248-258. Malker HSR, McLaughlin JK, Silverman DT, Ericsson JLE, Stone BJ, Weiner JS, Malker BK, Blot WJ (1987): Occupational risks for bladder cancer among men in Sweden. Cancer Res 47:6763-6766. National Bureau of Statistics (1982): “Standard Classification of Industries and Occupations Used for the Third National Census.” Beijing; Office of the National Census of the State Council. Schoenberg JB, Stemhagen A, Mogielnicki AP, Altman R, Abe T, Mason TJ (1984): Case-control study of bladder cancer in New Jersey. I. Occupational exposures in white males. J Natl Cancer Inst 72:973-981. Schumacher MC, Slattery ML, West DW (1989): Occupation and bladder cancer in Utah. Am J Ind Med 16:89-102. Schulte PA, Ringen K, Hemstreet GP, Ward E (1987): Occupational cancer of the urinary tract. Occup Med State of the Art Reviews 2:85-107. Silverman DT, McLaughlin JK, Malker HSR, Weiner JS, Ericsson JLE (1989a): Bladder cancer and occupation among Swedish women. Am J Ind Med 16:239-240. Silverman DT, Levin LI, Hoover RN, Hartge P (1989b): Occupational risks of bladder cancer in the United States: I. White Men. J Natl Cancer Inst 81:1472-1480. Silverman DT, Levin LI, Hoover RN (1990): Occupational risks of bladder cancer among white women in the United States. Am J Epidemiol 132:453-461. Silverman DT, Momson AS, Devesa SS (1991): Bladder cancer. In Schottenfeld D, Fraumeni JF Jr (eds), “Cancer Epidemiology and Prevention, ” 2nd edition. Philadelphia: Saunders (in press). Steenland K, Burnett C, Osorio A (1984): A case-control study of bladder cancer using city directories as a source of occupational data. Am J Epidemiol 126:247-257. You XY, Chen JG, Hu YN (1990): Studies on the relation between bladder cancer and benzidine or its derived dyes in Shanghai. Br J Ind Med 47:544-552. Zahm SH, Hartge P, Hoover R (1987): The National Bladder Cancer Study: Employment in the chemical industry. J Natl Cancer Inst 79:217-222.
