Bladder Cancer and the Rubber Industry : An Epidemiological Review


This review has been written at the request of the Ontario Ministry of Labour to assist the Occupational Disease Panel in its deliberations concerning the probable connection between cancer of the bladder and working in the rubber industry.

The review is based primarily on the results of epidemiological studies of bladder cancer in relation to the rubber industry. No attempt has been made to review the toxicological evidence concerning the carcinogenicity of specific chemicals used in the rubber industry, or the measurement of the degree of exposure to such chemicals.

The paper begins with a general description of bladder cancer, its descriptive epidemiology and the risk factors, other than work in the rubber industry, which have been identified. Methods of early detection are also discussed briefly.

Available information on the rubber industry, in particular its importance in Canada, is then summarized. The hazards associated with working in the industry, other than bladder cancer, are also mentioned.

In the main section of the paper the epidemiological studies relating the mortality and incidence of bladder cancer to work in the rubber industry are reviewed in depth. The most important types of study are cohort studies (in which rubber workers are followed up and the incidence of, or mortality from bladder cancer determined), and case-control studies (in which cases of bladder cancer and controls are compared with respect to history of working in the rubber industry). These "analytical" studies provide estimates of the risk of bladder cancer among rubber workers relative to that of people who are not rubber workers, the "risk ratio" or "relative risk" (RR). The degree to which the observed RR is greater than unity is a measure of the possible causal importance of the exposure in relation to the disease.

The precision of the estimate of the RR from any particular study is low : in cohort studies because the incidence of bladder cancer is small, and in case-control studies because the proportion of the population working in the rubber industry is small. For each type of study the estimates of RR from the individual studies (with sufficient data) have been combined by weighting the estimates of the RR by their precision.

Finally, the results of the review are summarized and some conclusions drawn with respect to present knowledge and the need for further research.

Bladder cancer

"Bladder cancer" is a convenient abbreviation for the more precise specification in the International Classification of Diseases: malignant neoplasm of bladder (ICD 188). This category includes malignant neoplasms other than carcinoma, but these are rare in the bladder. The majority (94%) of bladder cancers are transitional cell carcinomas, the remainder being squamous cell carcinoma (4%) and adenocarcinoma of urachal origin (1%) (Hirayama, Waterhouse and Fraumeni, 1980). The distinction between a benign papilloma and a low-grade papillary transitional cell carcinoma is not easy to make (del Regato and Spjut, 1977). This can lead to difficulties in interpretation of incidence statistics. Some epidemiological studies include benign papilloma in the case group. The transitional cell epithelium of the bladder is continuous with that of the ureter and renal pelvis, and the risk factors for carcinomas of these parts of the kidney are similar to those for bladder cancer, so these sites are sometimes combined in epidemiological reviews (Morrison and Cole, 1982).

The World Health Organization estimates that, globally, approximately 170,000 new cases of bladder cancer occur each year, two thirds of them occurring in developed countries and three quarters in males. The cases can be subdivided into two broad categories based on etiology : those caused by tobacco and industrial carcinogens, and those caused by schistosomiasis. The former are predominantly transitional cell carcinomas and are common in industrialized countries. The latter are more likely to be squamous cell carcinomas and are found chiefly in Mediterranean and African countries. Both types are thus preventable in principle. (Korolthchouk et al, 1987)

In Canada, it is estimated that there will be 4,500 new cases of bladder cancer in 1996, i.e. about 3% of the global total and 5% of those in Developed countries. In keeping with the global figures, 3,300 (73%) of the new cases in Canada will occur in males. It is estimated that bladder cancer will cause 1,320 deaths in Canada in 1996, 920 (70%) in males (National Cancer Institute of Canada, 1996).

In the period 1984-88 the incidence of bladder cancer in Canada (rates per 100,000 standardized to the world standard population) was 21.0 for men and 5.7 for women (Band et al, 1993). These rates lie towards the upper end of the spectrum of international incidence rates for about the same period, which range from 2.8 to 27.3 for men, and 0.9 to 7.4 for women (Whelan, Parkin and Masuyer, 1990). In general, incidence rates are highest in North America and lowest in East Europe and Asia (Hirayama, Waterhouse and Fraumeni, 1980).

In international comparisons of mortality from bladder cancer Canada is not so highly placed. Among 31 Developed countries in 1978-79 the Canadian rates of 4.9 in males and 1.6 in females ranked 23 and 15 respectively (Kurihara, Aoki and Tominaga, 1984).

Within Canada age standardized incidence rates are highest in Quebec and Ontario in both sexes, and lowest in British Columbia and the North (Band et al, 1993). A similar, but less pronounced, pattern is seen for age-standardized mortality rates (Statistics Canada, 1989). The five-year relative survival rate for bladder cancer in Ontario in the period 1978-87 was 80% for males and 74% for females (National Cancer Institute of Canada, 1991).

Since 1969 the age-standardized incidence rate for bladder cancer among Canadian males increased gradually to a plateau in the mid-1980's and then declined. The upward secular trend is consistent with that observed between 1940 and 1979 in Connecticut, where the age-standardized incidence rate increased by 155% in males and 90% in females over that period. A detailed analysis of the Connecticut rates showed that the secular increase was best explained by an age-cohort model, the risk increasing in male cohorts beginning in those born around 1870 with a peak in those born around 1930, and the risk increasing in female cohorts beginning in those born around 1900 and continuing to the most recent cohort born around 1940. Similar patterns were found for lung cancer, and are probably attributable to trend in cigarette smoking (Roush et al, 1987). It seems likely that the secular trends in Canada could also be explained in this way.

The literature on the risk factors for bladder cancer has been the subject of several in-depth reviews in the last 15 years (Matanoski and Elliott, 1981; Alderson, 1982; Morrison and Cole, 1982; Hill, 1984; Roush et al, 1987; Silverman et al, 1992; Cohen and Johansson, 1992). Occupational risk factors have also been reviewed by several authors (Cartwright, 1983; Conso, 1983; Alderson, 1986; Steineck et al, 1990; Jones et al, 1992). Table 1 (taken from Hill, 1984), has been modified to take into account more recent studies.

The risk factors are grouped by the type of known, or postulated, carcinogenic agent : biological, physical and chemical. The evidence for a causal role in human bladder cancer is graded as follows :

A = association consistently positive in analytical studies, including at least one cohort study, with evidence of dose response or experimental validation;

B = association consistently positive in two or more analytical studies but no definite dose response or experimental validation;

C = association positive in only one analytical study or inconsistent finding in multiple analytical studies;

D = association suggested by descriptive studies, case reports or laboratory findings not confirmed by analytical studies.

Other than occupational exposures, the only risk factors for which a causal relationship can be deemed definitively established (grade A evidence) are ionizing radiation and tobacco. The association with ionizing radiation is only established for atomic bomb survivors and for women treated for cancer of the cervix or ovary. Without doubt, tobacco use, especially cigarette smoking is the dominant causal factor for bladder cancer in many countries. In a recent Canadian study, the population attributable risk for cigarette smoking was about 47% in males and 33% in females (Burch et al, 1989).

Among the occupational risk factors, in addition to chemical, gas and rubber workers, the evidence for aluminum workers is now definitive, based on Canadian studies (Theriault et al, 1981; Rockette and Arena, 1983; Theriault et al, 1984; Gibbs, 1985).

Screening for bladder cancer, as an attempt to control the disease among workers exposed to aromatic amines, has been used for 70 years (Cartwright, 1984, 1990). Earlier methods, such as routine cystoscopy and microscopic examination of the urine for the presence of hematuria, were quickly displaced after the introduction of malignant cell cytology (MCC - Papanicolou method) in the 1950's. MCC has been used extensively in the British chemical and rubber industries, and also in the United States and Japan. The MCC screening test has satisfactorily high sensitivity and specificity. However, despite its routine use in the United Kingdom, its effectiveness in reducing mortality or morbidity has not been evaluated in controlled studies. The comparative studies which have been reported suggest that screening has no effect on life expectancy (Fox and White, 1976; Cartwright, 1984).

Rubber industry

The manufacture of rubber products began in Canada in 1854 and increased rapidly after the introduction of the automobile (McDougall, 1988). In 1985 there were 148 establishments in Canada engaged in rubber manufacture, employing 25,366 people. Sixty per cent of the employees were in Ontario (Statistics Canada, 1990), and about the same proportion were involved in the manufacture of tires (McDougall, 1988). The size of the industry has probably declined in the last decade. However, workers leaving the industry will remain at risk for occupational bladder cancer, since the clinical latency following exposure to aromatic amines can be as long as 48 years (Goldblatt, 1947).

Manufacture of rubber products involves the following processes:

(a) raw materials handling, mixing and weighing, in which the raw polymer (natural or synthetic) is mixed with a variety of chemical additives;

(b) milling, to blend the rubber and chemicals;

(c) extruding and calendering, where the re-heated rubber is shaped into sheets, strips or pellets;

(d) component assembly and buidling, to create the eventual product (tires, rubber boots etc.);

(e) vulcanizing, using sulphur compounds and accelerators;

(f) inspection and finishing;

(g) storage and dispatch.

During these processes workers are exposed to many chemicals, including carcinogens of several types : aromatic amines, polycyclic aromatic hydrocarbons, nitrosamines and halogenated hydrocarbons (International Agency for Research on Cancer, 1982).

The cohort mortality studies of rubber workers, initiated primarily to study the risk of bladder cancer (see below), have identified increased risks for other cancer sites (International Agency for Research on Cancer, 1982; Nutt, 1983; Alderson, 1986). The results of these studies were combined by Alderson (1986) to give the following estimates for cancer sites with statistically significant increased risks:

Cancer site RR 95% CI
large intestine            
1.1, 1.3
1.1, 1.3
1.0, 1.1
1.0, 1.4

Increased risks of mortality from other chronic diseases were not consistently found in these studies. In one of them a significant increase in mortality from diabetes was reported (McMichael et al, 1976).

Epidemiological studies of bladder cancer in rubber workers

Epidemiological studies of occupational cancer are of four main types :

(a) occupational mortality (incidence) studies i.e. comparisons of occupation-specific mortality (or incidence) rates in the population, calculated from statements of occupation on the death certificates (or cancer registration) and census records;

(b) ecological studies i.e. correlations of the mortality (or incidence) rates for geographical regions with the proportions of the populations in the regions employed in different occupations;

(c) case-control studies i.e. comparisons of the occupational histories of patients (alive or dead) with those of controls;

(d) cohort studies i.e. follow-up studies of people in a given occupation to determine their subsequent mortality (or incidence) in comparison to that of other occupations or the population at large.

The first two types of study are often called "descriptive". They are used to generate hypotheses, or provide initial support for hypotheses arising from clinical observation. Due to their cross-sectional nature, which makes them prone to selection bias, descriptive studies do not provide definitive conclusions. Standardized mortality ratios (SMR), which estimate the RR, can be obtained from occupational mortality studies, but estimates of the RR cannot be obtained from ecological correlations. The third and fourth types of study fall under the heading of "analytical studies" and, if rigorously performed, can provide reliable risk estimates.

All epidemiological studies can be misleading due to the presence of three types of bias : un-representative selection in the choice, or loss to follow-up, of subjects, inaccurate information on occupation or the diagnosis of cancer, and the presence of confounding factors, i.e. factors which are related to both occupation and the incidence of cancer. It is generally believed that cohort studies are the least prone to selection and information bias. However, information on confounding factors, such as smoking, is rarely available.

Descriptive studies

The longest series of occupational cancer mortality studies is that published decennially by the Registrar General of England and Wales. This series of reports has been collated by Logan (1982). Unfortunately the numbers of deaths from bladder cancer among rubber workers were too small to permit estimates of the SMR. This was also the case for occupational mortality studies in the Netherlands (Versluys, 1949) and British Columbia (Gallagher et al, 1989).

Two occupational incidence studies of bladder cancer have been published recently. In a study of 701 cases in the medical insurance records of the Paris region a RR of 1.4 based on 39 cases was found for workers in the rubber sector (Di Menza et al, 1992). In a much larger study of 1,219 incident cases from the Shanghai Cancer Registry in 1980 to 1984, the standardized incidence ratio for men working as rubber and plastics product makers was 208, based on 16 cases, and that for women was 188, based on 7 cases (Zheng et al, 1992).

Atlases of cancer mortality have been published in several countries. In Canada, for mortality by census division in the period 1966-76, high mortality rates for bladder cancer in males were seen in census divisions in Ontario bordering Lake Ontario and in the south west peninsular, and in the Montreal region of Quebec (Health and Welfare Canada and Statistics Canada, 1980). No formal correlations with respect to occupation or other risk factors were made with these data. More recently an atlas of cancer incidence in Ontario in 1984-88 has been published (Mills and Semenciw, 1992). Significantly high incidence ratios were found in the areas around Brantford and Windsor.

Atlases of cancer mortality in the period 1950-1980 by county or state economic area in the United States have also been published (Mason et al, 1975; Pickle et al, 1987). High rates for bladder cancer among males were found in the North-east, around the Great Lakes and in southern Louisiana. Among females high rates were seen in New York and New England. Ecological correlations of the bladder cancer rates in the period 1950-69 showed significantly higher rates for counties with high levels of white male employment in chemical and printing industries, but not in rubber manufacture (Blot and Fraumeni, 1978).

An atlas of cancer mortality for the European Economic Community has been published recently (Smans, Muir and Boyle, 1992). Areas of risk for bladder cancer among males were seen in some, but not all heavily industrialized regions, especially those with high concentrations of chemical industries. No formal correlations were performed.

Ecological studies of mortality from bladder cancer in England and Wales (Dolin, 1992) and Japan (Yamaguchi et al, 1992) have also been published recently. In neither study was a correlation between bladder cancer mortality and the proportion of people working in the rubber industry found.

Analytical studies

(a) Cohort mortality studies

The fact that chemical workers have an increased risk of bladder cancer was first noticed clinically (Rehn, 1875), as is often the case where workers are at very high risk for a rare event. Over the next 50 years many occurrences of bladder cancer in chemical workers exposed to aromatic amines were documented in several countries (Goldblatt, 1947), but the first formal cohort mortality study of these workers did not begin until 1948 in England and Wales (Case et al, 1954; Case and Pearson, 1954). The results of this study have been summarized by Alderson (1986) as follows:

         Agent    Deaths RR   95% CL
  Obs       Exp    
2-Naphthylamine           26 0.30                 86.7        56.6, 127.0
Benzidine 10 0.72 13.9 6.6, 25.5
1-Naphthylamine   6 0.70   8.6 3.1, 18.7
Mixed exposures 81 1.48 54.7 43.5, 68.0
Auramine   3 0.13 23.1 4.6, 67.4
Magenta   6 0.45 13.3 4.9, 29.0

The fact that workers in the rubber industry were also at increased risk of bladder cancer was an incidental finding from the ancillary work connected with the study of chemical workers :

"Quite early in the survey of chemical workers an almost fortuitous discovery was made. As part of the epidemiological survey of the general population, the County of Birmingham was selected for study because it was a large industrial city without any large-scale dyestuff industry....

It at once became apparent that an undue number of patients who suffered from bladder tumour had worked in one large rubber works...At the same time it came to light that one class of compounds called antioxidants which was put into rubber to stop it perishing was made from alpha- and beta-napthylamine and that some cases of bladder tumour had occurred amongst the men making it." (Case, 1966)

Following upon this finding, Case and his colleagues mounted a study of mortality from bladder tumour in England and Wales based on the death certificates from 1921 to 1951 ( UK(a) = Case and Hosker, 1954). This study is usually referred to in the literature as a cohort study but this is not strictly true. Unlike the study of chemical workers there was no nominal roll of rubber workers available. Observed deaths were attributed to working in the rubber industry based on the statement of occupation on the death certificate and man-years at risk in the rubber industry were estimated from census statistics. Thus the study was really an extended version of the occupational mortality analyses published decennially by the Registrar General (see above). However, the study has been included with true cohort studies in this review. A second feature of the study which is worthy of note is that, as in the study of chemical workers, deaths attributed to papilloma of the bladder were included as well as those due to carcinoma of the bladder. Since the outcome measured was mortality rather than incidence this may not have made much difference to the risk estimates.

Although the British rubber manufacturers acted quickly to remove the suspect antioxidant, they, and the Ministry of Labour, were urged to mount further studies to quantify the risk (Lancet editorial, 1965). This resulted in two cohort studies among British rubber workers. The first (UK(b) = Fox and Collier, 1976; Baxter and Werner, 1980) was based on a census of men working in the rubber and cable-making industry in Great Britain in 1967, with subsequent linkage to death records. The second (UK(c) = Parkes et al, 1982) was based on a nominal roll of male employees who first started work in 13 rubber factories between 1946 and 1960 with follow-up to death prior to 1976. Clearly, there was some overlap between these two studies. Alderson (1986) states that 18 per cent of the men involved in the first were also included in the second.

In the early 1970's three cohort mortality studies were initiated in rubber workers in the United States. The first two (US(a) = McMichael et al, 1974; US(b) = Anjelkovic, Taulbee and Symons, 1976) were conducted by researchers at the University of North Carolina, and the third (US(c) = Delzell and Monson, 1981) by researchers at Harvard University.

Cohort mortality studies of rubber workers have also been reported from Switzerland (Bovet and Lob, 1980), Sweden (Gustavsson, Hogstedt and Holmberg, 1986), Italy (Negri et al, 1989) and Canada (Choi and Nethercott, 1991). In the Canadian study, which involved workers in a tire manufacturing plant in Ontario, a proportional mortality analysis was made, in which the expected number of deaths was calculated based on the proportion of all deaths in Ontario which were attributed to bladder cancer. Presumably this was done because it was not possible to estimate the person-years at risk among the rubber workers.

The results of the British, American and Swiss studies were collated by Alderson (1986). This analysis has been up-dated in Table 2 to include the subsequent Swedish, Italian and Canadian studies. The method used to combine the studies is described in the appendix. Some studies included women but only the results for men are included Table 2.

The combined estimate of the RR is 1.17, which differs significantly from unity (chisquare = 5.30, 1 degree of freedom, p < 0.025). The 95% confidence limits for the combined RR are 1.02, 1.31. However the chisquare for homogeneity (17.99, 9 degrees of freedom) is significant at the 5% level. If the two outliers (UK(c) and Sweden) are omitted, the chisquare for homogeneity (7.26, 7 degrees of freedom) is not significant. The combined estimate of the RR from the remaining 8 studies is 1.23, with confidence limits 1.03, 1.31, which is significantly different from unity (chisquare = 7.69, 1 degree of freedom, p<0.01).

Some of the studies provide information on the RR among the occupational groups within the rubber industry. Checkoway et al (1981) compared the RR within the combined studies in the United States and found above average RR for milling, calendaring and the final inspection of tires. In the British study reported by Parkes et al (1982) the RR was highest for extruding and calendaring. A value of RR greater than 2 was found for workers in the weighing and mixing departments of the Swedish study (Gustavsson, Hogstedt and Holmberg, 1986). In the Italian study (Negri et al, 1989) the RR was above average for milling and for "various services".

(b) Cohort incidence studies

Results have been reported from three cohort incidence studies of rubber workers : in the United Kingdom (Veys, 1969); Finland (Kilpikari et al, 1982) and Sweden (Gustavsson, Hogstedt and Holmberg, 1986). The results are shown in Table 3. The combined estimate of RR is 1.45, with 95% confidence limits 0.92, 1.98. However the heterogeneity among the three estimates is greater than would be expected by chance.

(c) Case-control studies

Working in the rubber industry was mentioned in 18 case-control studies of bladder cancer reported over the last 25 years : Argentina (Iscovich et al, 1987); Belgium (Schifflers, Jamart and Renard, 1987); Canada ((a) Howe et al, 1980, (b) Miller et al, 1978, (c) Risch et al, 1988)); Denmark (Jensen et al, 1987); France ((a) Cordier et al, 1993, (b) Hours et al, 1994); Germany (Claude, Frentzel-Beyme and Kunze, 1988); Italy ((a) Vineis and Magnani, 1985, (b) La Vecchia et al, 1990); Spain (Gonzalez et al, 1989); United Kingdom (Anthony and Thomas, 1970); United States ((a) Cole, Hoover and Friedell, 1972, (b) Wynder and Goldsmith, 1977, (c) Najem et al, 1978, (d) Silverman et al, 1989, (e) Schumacher, Slattery and West, 1989).

Of these studies 11 provided estimates of RR and 95% confidence limits, adjusted (or matched) for other variables. These estimates are shown in Table 4, and have been combined using a method described in the appendix. The test for heterogeneity is not significant. The combined estimate of RR is 1.51, significantly different from unity (p < 0.05), with 95% confidence limits 1.03 to 2.20.

None of the studies was large enough to examine the interaction of smoking and employment in the rubber industry. However in the recent Spanish study (Gonzalez et al, 1989) the interaction of smoking with employment in any high risk occupation was examined, with the following estimates of RR:

high risk
ex-smoker            smoker
                  ratio   1.13     3.65 1.80

These findings indicate a multiplicative relationship between smoking and occupational exposure.

Summary and conclusions

It is estimated that there will be 4,450 new cases of bladder cancer in Canada in 1996, about three quarters of them in men. The incidence of the disease among Canadian men increased until about 10 years ago and is now decreasing slightly. Cancer of the bladder is more common in Quebec and Ontario than in the rest of the Canada. Although occupational exposure to carcinogenic chemicals could have played a role in these trends, most of the differences are probably due to changes in the prevalence of cigarette smoking, which is by far the most important risk factor for bladder cancer.

The overall survival rate for bladder cancer is about 80 per cent, but is much lower for cases presenting with late stage disease. Presymptomatic cases can be detected by cytological examination of the urine so that early detection is a possible way to reduce mortality. Cytological screening of men in some high risk occupations has been used routinely in Britain for several decades, but there is no evidence that this has reduced mortality or morbidity due to the disease.

About 25,000 people were employed in the Canadian rubber industry in 1985, over half of them in Ontario. The size of the industry is probably declining, but the effects of exposure would persist for as long as 40 years. Rubber workers are exposed to many chemicals, some of which are known to be carcinogenic. There is evidence from epidemiological studies that they are at increased risk of cancer at various sites, including stomach, large intestine, lung and leukemia, as well as the bladder.

The evidence in favour of an increased risk of bladder cancer among rubber workers comes from a large number of epidemiological studies, both cohort studies and case-control studies, including some in Canada. The evidence from each individual study is rarely persuasive, but taken together they point to a small but significant excess risk for men, probably not greater than 50 per cent. The attributable fraction among those exposed, which is usually accepted as a measure of the probability that a cancer arising in a rubber worker is due to his occupation, is thus quite small, about a third. However there is some evidence that the risks of bladder cancer from smoking and working in high risk occupations is multiplicative, so that the attributable fraction among smokers in the industry might be higher than this figure.

Even after combining the different studies from several countries the estimate of the increased risk of bladder cancer among rubber workers is still imprecise. Further studies in Canada, especially in Ontario, would help to reduce this uncertainty. Only one, rather small, cohort study in Ontario has been reported, and the possibility of broadening this study should be explored. The Canadian cancer registries, in collaboration with the Laboratory Centre for Disease Control of Health Canada, have recently started a large study of newly registered cases of several types of cancer, including bladder cancer. This will include smoking and occupational histories of cases and controls, and should provide more accurate estimates of the risks for rubber workers and the interaction of the risk with that for smoking.


The results of individual studies of the same type can be combined using the following approach:

Let Ri, i = 1, 2....N, be the estimates of the RR from N studies, and assume that they are distributed by chance about R, the true value of the RR, with variances Vi. Then it can be shown that Re, the estimate of R with minimum variance, is the weighted mean of the Ri, the weights Wi being the reciprocals of the Vi.

Thus Re = SWiRi / SWi, where Wi = 1/Vi, and S denotes summation over i = 1 to N.

The estimate of the variance of Re is Ve = 1/ SWi, so that the 95% confidence limits are Re +/- 1.96 sqrt(Ve).

A test of the homogeneity of the individual estimates is obtained by comparing Xh = SWiRi2 - (SWiRi)2/SWi with chisquare on N-1 degrees of freedom. A test of the null hypothesis that R = 1 is obtained by comparing Xo = (Re - 1)2/Ve with chisquare on one degree of freedom.

In cohort studies the estimates Ri = Oi/Ei, where Oi is the observed number of deaths (or cases) among the rubber workers and Ei is the expected number of deaths (or cases) based on the rates in the general population. Assuming the Oi have Poisson distributions, and the Ei are fixed numbers then Wi = Ei/R, giving Re = SOi/SEi, Ve = SOi/(SEi)2, Xh = (SOi2/SEi)/Re - SOi, Xo = (SOi - SEi)2/SOi.

The results of case-control studies are presented in various ways depending on the degree of matching between cases and controls. Common to all studies is an estimate of the RR (odds ratio) R, with confidence limits L,U. These estimates are derived by exponentiating estimates of the natural logarithm of the odds ratio, and an estimate of its variance which depends on the study design.

To combine the estimates from several studies values of Yi = ln(Ri), and Wi = 3.92/{ln(Ui) - ln(Li)} are used in the general formulae : Ye = SWiYi/SWi, Re = exp(Ye) with confidence limits exp(Ye +/- 1.96 sqrtVe), Xh = SWiYi2 - (SWiYi)2/SWi, Xo = Ye2/Ve.

References Of Gerald Hill study as transmitted 96-07-10

Alderson M. The causes of cancer. In: Alderson M (ed.) The prevention of cancer. London: Edward Arnold, 1982; 20-79

Alderson M. Occupational cancer, London: Butterworths, 1986

Andjelkovic D, Taulbee J, Symons M. Mortality experience of a cohort of rubber workers, 1964-1973. J Occup Med 1976; 18: 387-394

Anthony HM, Thomas GM, Cole P. Tumors of the urinary bladder: an analysis of the occupations of 1,030 patients in Leeds, England. J Natl Cancer Inst 1971; 46: 1111-1113

Band PR, Gaudette LA, Hill GB, Holowaty EJ, Huchcroft SA, Johnston GM, Makomaski Illing EM, Mao Y, Semenciw R. The making of the Canadian Cancer Registry: Cancer incidence in Canada and its regions, 1969 to 1968. Ottawa: Minister of Supply and Services Canada, 1993.

Baxter PJ, Werner JB. Mortality in the British rubber industries 1967-1974. London: Her Majesty's Stationery Office, 1980

Blot WJ, Fraumeni JF Jr. Geographic patterns of bladder cancer in the United States. J Natl Cancer Inst 1978; 61: 1017-1023

Bovet P, Lob M. Mortality from malignant tumours in rubber workers in Switzerland: epidemiological study 1955-1975. Schweiz Med Wochenschr 1980; 110: 1277-1287

Burch JD, Rohan TE, Howe GR, Risch HA, Hill GB, Steele R, Miller AB. Risk of bladder cancer by source and type of tobacco exposure: a case-control study. Int J Cancer 1989; 44: 622-628

Cartwright RA. Historical and modern epidemiological studies on populations exposed to N-substituted aryl compounds. Environ Health Perspect 1983; 49:13-19

Cartwright RA. Screening for bladder cancer with particular reference to industrial groups. In: Prorok PC, Miller AB (eds.) Screening for cancer. I - General principles on evaluation of screening for cancer and screening for lung, bladder and oral cancer. Geneva: Union Internationale Contre le Cancer, 1984; 144-160

Cartwright RA. Bladder cancer screening in the United Kingdom. J Occup Med 1990; 32:878-880

Case RAM, Pearson JT. Tumours of the urinary bladder in workmen engaged in the manufacture and use of certain dyestuff intermediates in the British chemical industry. Part II. Further considerations of the role of aniline and of the manufacture of auramine and magenta (fuchsin) as possible causative agents. Br J Indust Med 1954; 11: 213-216

Case RAM, Hosker ME. Tumour of the urinary bladder as an occupational disease in the rubber industry in England and Wales. Br J Prev Soc Med 1954; 8: 39-50

Case RAM, Hosker ME, McDonald DB, Pearson JT. Tumours of the urinary bladder in workmen engaged in the manufacture and use of certain dyestuffs intermediates in the British chemical industry. Part I. The role of aniline, benzidine, alpha-naphthylamine and beta-naphthylamine. Br J Indust Med 1954; 11: 75-104.

Case RAM. Tumours of the urinary tract as an occupational disease in several industries. Ann R Coll Surg Engl 1966; 39: 213-235

Checkoway H, Smith AH, McMichael AJ, Jones FS, Monson RR, Tyroler HA. A case-control study of bladder cancer in the United States rubber and tyre industry. Br J Ind Med 1981;38:240-246

Choi BC, Nethercott JR. A proportionate mortality study on the risk of bladder cancer among rubber workers. Cancer Detect Prev 1991; 15:403-406

Claude JC, Frentzel-Beyme RR, Kunze E. Occupation and risk of cancer of the lower urinary tract among men. A case-control study. Int J Cancer 1988; 41: 371-379

Cohen SM, Johansson. Epidemiology and etiology of bladder cancer. Urol Clin North Am 1992; 19:421-428

Cole P, Hoover R, Friedell GH. Occupation and cancer of the lower urinary tract. Cancer 1972; 29: 1250-1261

Conso F. Amines aromatiques en milieu de travail et tumeurs urotheliales. Sem Hop 1983; 59:2103-2106

Cordier S, Clavel J, Limasset JC, Boccon-Gibod L, Le Moual N, Mandereau L, Hemon D. Occupational risks of bladder cancer in France: a multicentre case-control study. Int J Epidemiol 1993; 22: 403-411

del Regato JA, Spjut HJ. Cancer diagnosis , treatment, and prognosis. Fifth edition. St. Louis: C.V. Mosby, 1977

Delzell E, Monson RR. Mortality among rubber workers III. Cause specific mortality 1940-1978. J Occup Med 1981; 23:677-684

Di Menza L, Baron JC, Vieillefond A, Choudat D, Boccon-Gibod L, Zummer K. Facteurs de risque des tumeurs de vessie. Etude epidemiologique chez 701 sujets en Ile-deFrance. Presse Med 1992; 21: 153-156

Dolin PJ. A descriptive study of occupation and bladder cancer in England and Wales. Br J Cancer 1992; 65: 476-478

Editorial. Occupational bladder tumours and the control of carcinogens. Lancet 1965; 1: 306-307

Fox AJ, Collier PF. A survey of occupational cancer in the rubber and cablemaking industries: Analysis of deaths occurring in 1972-74. Br J Ind Med 1976; 33:249-264

Fox AJ, White GC. Bladder cancer in rubber workers; do screening and doctor's awareness distort the statistics? Lancet 1976; 1: 1009-1011

Gallagher RP, Threlfall WJ, Band PR, Spinelli JJ. Occupational mortality in British Columbia 1950-1984. Vancouver, B.C.: Cancer Control Agency of British Columbia, 1989

Gibbs GW. Mortality of aluminum reduction plant workers, 1950 through 1957. J Occup Med 1985; 27: 761-770

Goldblatt MW. Occupational cancer of the bladder. Br Med Bull 1947; 4: 405-417

Gonzalez CA, Lopez-Abente G, Errezola M, Escolar A, Riboli E, Izarzugaza I, Nebot M. Occupation and bladdercancer in Spain: a multicentre case-control study (in press in 1991)

Gustavsson P, Hogstedt C, Holmberg B. Mortality and incidence of cancer among Swedish rubber workers 1952-1981. Scand J Work Environ Health 1986;12:538-544

Health and Welfare Canada and Statistics Canada. Mortality atlas of Canada, volume 1: cancer. Ottawa: Minister of National Health and Welfare and President of Treasury Board, 1980

Henry SA, Kennaway NM, Kennaway EL. The incidence of cancer of the bladder and prostate in certain occupations. J Hyg 1931; 31:125-137

Hill GB. Epidemiologic considerations. In: Javadpour N (ed.) Bladder cancer. Baltimore: Williams and Wilkins, 1984; 1-36

Hirayama T, Waterhouse JAH, Fraumeni JF Jr (eds). Cancer risks by site. Geneva: International Union Against Cancer, 1980

Hours M, Danache B, Fevotte J, Bergeret A, Ayzac L, Cardis E Etard JF, Pallen C, Roy P, Fabry J. Bladder cancer and occupational exposures. Scand J Work Environ Health 1994; 20:322-330

International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risk of Carcinogens to Humans, vol 28: The Rubber Industry. Geneva: World Health Organization, 1982

Iscovich J, Castelletto R, Esteve J, Munoz N, Colanzi R, Coronel A, Deamezola I, Tassi V, Arslan A. Tobacco smoking, occupational exposure and bladder cancer in Argentina. Int J Cancer 1987; 40: 734-740

Jensen OM, Wahrendorf J, Knudsen JB, Sorensen BL. The Copenhagen case-referent study on bladder cancer. Risks among drivers, painters and certain other occupations. Scand J Work Environ Health 1987;13:129-134

Jones RF, Debiec-Rychter M, Wang CY. Chemical carcinogenesis of the urinary bladder - a status report. J Cancer Res Clin Oncol 1992; 118: 411-419

Kilpikari I, Pukkala E, Lehtonen M, Hakama M. Int Arch Occup Environ Health 1982; 51: 65-71

Koroltchouk V, Stanley K, Stjernward J, Mott K. Bladder cancer: approaches to prevention and control. Bull World Health Org 1987;65:513-520

Kurihara M, Aoki K, Tominaga S (eds.) Cancer mortality statistics in the world. Nagoya, Japan: University of Nagoya Press, 1984

La Vecchia C, Negri E, D'Avanzo B, Franceschi S. Occupation and the risk of bladder cancer. Int J Epidemiol 1990; 19:264-268

Logan WPD. Cancer mortality by occupation and social class 1851-1971. London: Her Majesty's Stationery Office (Studies on Medical and Population Subjects No. 44), and Lyon: International Agency for Research on Cancer (IARC Scientific Publications No. 36), 1982

Mason TJ, McKay FW, Hoover R, Blot WJ, Fraumeni JF Jr. Atlas of cancer mortality for U.S. counties: 1950-1969. Bethesda MD: U.S. Department of Health, Education, and Welfare, DHEW Publication No. (NIH) 75-780, 1975

Matanoski GM, Elliott EA. Bladder Cancer Epidemiology. Epidemiol Rev 1981;3:203-229

McDougall AJ. Rubber products industry. In: Marsh JH (ed). The Canadian Encyclopedia. Second edition. Edmonton, Alberta: Hurtig Publishers, 1988; 1898-1899

McMichael AJ, Spirtas R, Gamble JF, Tousey PM. Mortality among rubber workers; relationship to specific jobs. J Occup Med 1976;18:178-185

McMichael AJ, Spirtas R, Kupper LL. An epidemiological study of mortality within a cohort of rubber workers, 1964-1974. J Occup Med 1974; 16: 458-464

Miller CT, Neutel CI, Nair RL, Marrett LD, Last JM, Collins WE. Relative importance of risk factors in bladder carcinogenesis. J Chron Dis 1978; 31: 51-56

Mills C, Semenciw R. Great Lakes Region Cancer Incidence. Atlas 1984-1988: Ontario. Ottawa: Health and Welfare Canada, 1992

Morrison AS, Cole P. Urinary tract. In: Schottenfeld D, Fraumeni JF Jr (eds.) Cancer epidemiology and prevention. Philadelphia: W.B. Saunders, 1982; 925-937

Najem GR, Louria DB, Seebode JJ, Thind IS, Prusakowski JM, Ambrose RB, Fernicola AR. Int J Epidemiol 1982; 11: 212-217

National Cancer Institute of Canada. Canadian Cancer Statistics 1991. Toronto, Canada, 1991

National Cancer Institute of Canada. Canadian Cancer Statistics 1996. Toronto, Canada, 1996

Negri E, Piolatto G, Pira E, Decarli A, Kaldor J, La Vecchia C. Cancer mortality in a northern Italian cohort of rubber workers. Br J Ind Med 1989; 46: 624-628.

Nutt A. Rubber work and cancer - Past, present and perspectives. Scand J Work environ Health 1983; 9 Suppl 2:49-57

Parkes HG, Veys CA, Waterhouse JAH, Peters A. Cancer mortality in the British rubber industry. Br J Ind Med 1982; 39: 209-220

Pickle LW, Mason TJ, Howard N, Hoover R, Fraumeni JF Jr. Atlas of U.S. cancer mortality among whites: 1950-1980. Washington, DC: U.S.Department of Health and Human Services. DHHS publication no. (NIH) 87-2900, 1987

Risch HA, Burch JD, Miller AB, Hill GB, Steele R, Howe GR. Occupational factors and the incidence of cancer of the bladder in Canada. Brit J Ind Med 1988; 45: 361-367

Rockette HE, Arena VC. Mortality studies of aluminum reduction plant workers: Potroom and carbon department. J Occup Med 1983; 25: 549-557

Roush GC, Holford TR, Schymura MJ, White C. Cancer risk and incidence trends. The Connecticut perspective. Washington, DC: Hemisphere Publishing, 1987.

Schifflers E, Jamart J, Renard V. Tobacco and occupation as risk factors in bladder cancer. A case-control study. Int J Cancer 1987; 39:287-292

Schumacher MC, Slattery ML, West DW. Occupation and bladder cancer in Utah. Am J Ind Med 1989; 16: 89-102

Silverman DT, Hartge P, Morrison AS, Devesa SS. Epidemiology of bladder cancer. Hematol Oncol Clin North Am 1992; 6: 1-30

Silverman DT, Levin LI, Hoover RN et al. Occupational risks of bladder cancer in the United States I. White men. J Natl Cancer Inst 1989; 81: 1472-

Smans M, Muir CS, Boyle P (eds.) Atlas of cancer mortality in the European Economic Community. IARC Scientific Publications No. 107. Lyon: International Agency for Research on Cancer, 1992

Statistics Canada. Canada Year Book 1990. Ottawa: Minister of Supply and Services, 1990; Tables 16.6, 16.12

Statistics Canada. The Nation. Cat. No. 93-327. Ottawa: Minister of Industry, Science and Technology, 1993

Statistics Canada. Cancer in Canada 1984. Health Reports 1989; 1 (suppl)

Steineck G, Plato N, Norell SE, Hogstedt C. Urothelial cancer and some industry-related chemicals: an evaluation of the epidemiological literature. Am J Ind Med 1990;17:371-391

Theriault G, DeGuire L, Cordier S. Reducing aluminum: An pccupation possibly associated with bladder cancer. Can Med Ass J 1981; 124: 419-425

Theriault G, Tremblay C, Cordier S, Gingam S. Bladder cancer in the aluminum industry. Lancet 1984; 1: 947-950

Versluys JJ. Cancer and occupation in the Netherlands. Br J Cancer 1949; 3: 161-185.

Veys CA. Two epidemiological inquiries into the incidence of bladder tumours in industrial workers. J Natl Cancer Inst 1969; 43: 219-226

Vineis P, Magnani C. Occupation and bladder cancer in males: a case-control study. Int J Cancer 1985;35:599-606

Whelan SL, Parkin DM, Masuyer E (eds). Patterns of cancer in five continents. Lyon: International Agency for Research on Cancer, 1990

Wynder EL, Goldsmith R. The epidemiology of bladder cancer: A second look. Cancer 1977; 40: 1246-1268

Yamaguchi N, Watanabe S, Okubo T, Takahashi K. Work-related bladder cancer risks in male Japanese workers. Estimation of attributable fraction and geographical correlation analysis. Jpn J Cancer Res 1991; 82:624-631

Zheng W, McLauglin JK, Gao Y-T, Silverman DT, Gao R-N, Blot WJ. Bladder cancer and occupation in Shanghai, 1980-1984. Am J Ind Med 1992; 21: 877-885

Table 1. Risk Factors for Bladder Cancer 1
Risk Factor Evidence2             Risk factor Evidence
Biological               Chemical (contd)  
Allergy/immunity C             Occupation  
Bacterial infection C             - Leather B
Balkan nephropathy D             - Machinery B
Blood group C             - Medicine B
Schistosomiasis B             - Paint B
Tryptophan metabolites D             - Petroleum C
Viral infection C             - Photography C
Vitamin deficiency C             - Plumbing B
                - Rubber & cable A
Physical               - Seafaring B
                - Textiles B
Ionizing radiation A             - Trucking B
                - Wood C
Air pollution D             - Bracken fern D
Water treatment C             - Fat & protein D
Occupation:               - Sweeteners C
- Agriculture C    
- Aluminum A             Lifestyle:  
- Chemicals & dyes A             - Coffee B
- Coal gas & tar A             - Hair dye D
- Cooking /food C             - Opium D
- Glass C             - Tobacco A
- Hairdressing B    
                - Azathioprine D
                - Chlornaphazine B
                - Cyclophosphamide D
                - Isoniazid D
                - Phenacetin C
1 Modified from Hill (1984)
2 Grade of evidence (A highest, see text for definitions)


Table 2. Cohort Mortality Studies
Study1      Observed deaths       Expected deaths       Risk ratio
Canada   6          4.6        1.31
Italy 16          8.7        1.83
Sweden   4          5.3        0.75
Switzerland   4          1.1        3.62
UK (a) 35       24.4        1.43
UK (b) 73       57.6        1.27
UK (c) 36       43.0        0.84
US (a)  9       12.3        0.73
US (b) 21       18.1        1.16
US (c) 60       51.5        1.17
                   Total 264         226.6  
Combined RR = 1.165, 95% confidence limits 1.025, 1.306
Chisquare (RR = 1) = 5.30, 1 df, p<0.025
Chisquare (homogeneity) = 17.99, 9 df, p<0.05
Excluding studies UK (c) and Switzerland:
Combined RR = 1.227, 95% confidence limits 1.067, 1.388
Chisquare (RR = 1) = 7.69, 1 df, p<0.01
Chisquare (homogeneity) = 7.26, 7 df, p>0.25
1 See text for references


Table 3. Cohort Incidence Studies
Study1 Observed cases      Expected cases      Risk ratio
Finland  2        0.3       6.7 
Sweden 17      16.2       1.05
UK (d) 10        3.5       2.86
                 Total 29      20.0  
Combined RR = 1.450, 95% confidence limits 0.922, 1.978
Chisquare (RR = 1) = 2.79, 1 df, p>0.05
Chisquare (homogeneity) = 12.20, 2 df, p<0.005
1 See text for references


Table 4. Case-control Studies
Study1 Risk Ratio         95% Confidence Limits
Canada (a) 5.0               0.6        236.5
Canada (c) 1.27               0.67            2.44
Denmark 0.98               0.18            5.38
France (a) 1.17               0.50            2.76
France (b) 2.62               0.3          25.1    (includes females)
Germany 2.5               1.2            5.1
Italy (a) tyres 1.2               0.6            2.4
other 2.5               1.0            6.0
Spain 0.94               0.4            2.3
US (a) 1.57               1.04            2.36
US (d) 1.3               0.8            2.1
Combined RR = 1.510, confidence limits 1.034, 2.204
Chisquare (RR = 1) = 4.56, 1 df, p < 0.05
Chisquare (homogeneity) = 3.66, 10 df, p > 0.90
1 See text for references