Report to the Workers' Compensation Board on Stomach Cancer in Ontario Gold Miners
August, 1996

Occupational Disease Panel
ODP Report No. 16
Toronto, Ontario

Relevant Links

Gold Mining
Uranium Mining
Hardrock Mining Industry
Addendum to Hardrock Mining

Occupational Disease Panel

In 1985, the Ontario legislature established the Industrial Disease Standards Panel to investigate and identify diseases related to work. The Panel is independent of both the Ministry of Labour and the Workers' Compensation Board. At the end of each fiscal year the WCB reimburses the Ministry for the Panel's expenditures. In 1995, the name was changed to the Occupational Disease Panel (ODP).

The Panel's authority flows from section 95 of the Workers' Compensation Act and its functions are set out as follows:


(a) to investigate possible industrial diseases;

(b) to make findings as to whether a probable connection exists between a disease and an industrial process, trade or occupation in Ontario;

(c) to create, develop and revise criteria for the evaluation of claims respecting industrial diseases; and

(d) to advise on eligibility rules regarding compensation for claims.

Decisions of the Panel are made by its members who represent labour, management, scientific, medical and community interests. Once the Panel makes a finding, the WCB is required to publish the Panel's report in the Ontario Gazette and solicit comments from interested parties. After considering the submissions the WCB Board of Directors decide if the Panel's recommendations are to be implemented, amended or rejected.

To assist with its work, the Panel has a small staff of researchers, analysts and support people. In addition to its own staff, the Panel relies heavily on the advice of outside experts in science, medicine and law, as well as input from parties of interest.

Additional copies of this publication are available by writing:

Occupational Disease Panel
69 Yonge Street, Suite 1004
Toronto, Ontario M5E 1K3
(416) 327-4156

Panel Membership

Panel Members Appointment
Ms. Nicolette Carlan (Chair) May 16, 1991 to May 15, 1997
Mr. James Brophy January 23, 1992 to January 22, 1998
Dr. Carol Buck June 1, 1988 to June 16, 1997
Mr. William Elliott November 7, 1991 to November 6, 1997
Mr. John Macnamara November 7, 1991 to November 6, 1997
Mr. Homer Seguin May 28, 1989 to May 27, 1998
Dr. Michael Wills November 7, 1991 to November 6, 1997

Panel Staff

Panel Staff  
Carolyn Archer Senior Research Officer
Robert Chase Medical Consultant
Francis Macri Policy Analyst
Cara Melbye Policy Analyst
Anne Rekenye Data Entry Clerk
Tracy Soyka Administrative Co-Ordinator
Barry Spinner Policy Analyst
Salima Storey Administrative Officer
George Tomlinson Biostatistician
Jason Tung Industrial Hygienist


Letter of Transmittal
Panel Members
Panel Staff

Chapter 1. The Panel's investigation

a) ODP mandate and terms of reference
b) The issue and how it arose
c) The focus of the investigation
d) The investigative process
e) Submissions
f) Other Panel initiatives
g) Other jurisdictions

Chapter 2. The evidence

a) Background information about stomach cancer
b) Preliminary considerations
c) The questions

1) What is the evidence concerning stomach cancer mortality among non-gold miners worldwide and in Ontario?
i) International studies - non-goldmining
ii) Provincial studies - non-goldmining
iii) Conclusions

2) What is the evidence concerning stomach cancer mortality among goldminers worldwide and in Ontario?
i) International studies - goldmining
ii) Provincial studies - goldmining
iii) Discussion
iv) Conclusions

3) What is known about potential occupational causes of stomach cancer?
i) Dr. Stock's review
ii) Conclusions

4) Are any of these and other potential causes relevant in the Ontario gold mining environment?
i) Airborne dusts
a. Arsenic
b. Asbestos
c. Silica
d. Chromium
ii) Polycyclic aromatic hydrocarbons(PAHs)
a) Oil mist
b) Diesel emission
iii) Nitroso compounds
iv) Conclusions

5) What, if any, are the differences between gold mining and other kinds of hardrock mining in Ontario?
i) Backfill
ii) Aluminum powder
iii) Underground working environment
iv) Conclusions

6) Are there other risk factors that could affect the pattern of stomach cancer?
i) Diet
ii) Cigarette smoking and alcohol consumption
iii) Ethnicity and place of birth
iv) Conclusions

Chapter 3. Summary of conclusions

Chapter 4. The Bradford Hill criteria for a causal association

Chapter 5. Findings and recommendations


Appendix A: Summary of epidemiologic studies of gastric cancer and occupation

Appendix B: Report on analysis of stomach cancer among Ontario gold miners.

Figure 1.
The stomach

Figure 2.
Stomach cancer in non-gold mining (world)

Figure 3.
Stomach cancer in non-gold mining (Ontario)

Figure 4.
Gold mining and mixed gold mining

Figure 5.
Gold mining, Muller 1983, and stomach cancer

Figure 6.
(Post 1945 start) and stomach cancer mortality among Goldminers

Figure 7.
Stomach cancer and age at death

Figure 8.
Stomach cancer among gold miners and birthplace

Figure 9.
Fitted SMRs from Poisson regression

Figure 10.
Northern Ontario male population compared to all Ontario males

Figure 11.
Miners compared to the Northern Ontario male population

Figure 12.
Miners compared to all Ontario males

Figure 13.
Flow diagram of gold mining process

Table 1.
Summary of evidence of gold mining and stomach cancer mortality


For the purposes of this report, the Panel has adopted the following definitions:

Miner: refers to any person employed underground; in shaft sinking; and surface crushing; grinding; milling and tailings operations. All other surface work is excluded.

Goldminer: a miner with at least 85 % of mining time or 51 months in the foregoing employment whichever is less of which two weeks must have been in Ontario.

Stomach Cancer: means primary cancer of the stomach (International Classification of Diseases,7th, 8th & 9th Revisions, code#151).


a) The ODP mandate and terms of reference

The Occupational Disease Panel (ODP) is mandated by the legislation to investigate possible occupational diseases and, when appropriate, make findings of "probable connection" between disease and work (103).

The Panel's specific responsibilities are set out in section 95 of the Workers' Compensation Act which reads as follows:


(a) to investigate possible industrial diseases;

(b) to make findings as to whether a probable connection exists between a disease and an industrial process, trade or occupation in Ontario;

(c) to create, develop and revise criteria for the evaluation of claims respecting industrial diseases; and

(d) to advise on eligibility rules regarding compensation for claims.

After weighing all of the evidence, the Panel decides whether or not a probable connection exists between occupation and disease. If the results of the investigation do not indicate the existence of an association, the Panel will also report those findings.

When a probable connection is identified, depending on its strength, the Panel may recommend that the Workers' Compensation Board (WCB) enact guidelines for adjudication of claims on a case-by-case basis, or it may recommend that the disease and the relevant exposure(s) be added to Schedule 3 or 4 of the Act.

When a worker suffers from one of the diseases listed in the "Disease" column of Schedule 3 and can show that he or she was exposed to the associated substance or industrial process shown in the "Process" column, the Act states that the disease "shall be deemed to have been due to the nature of that employment unless the contrary is proved".

A Schedule 4 presumption, however, is termed "irrebuttable". Created by 1984 amendments to the Act, this presumption is only used in cases where the evidence of a work-relationship is indisputable. At present, only asbestosis, mesothelioma and cancer of the nasal cavities or paranasal sinuses are in Schedule 4. This means, for example, that if a worker has worked in a process involving asbestos listed in Column 2 and has a diagnosis of either mesothelioma or asbestosis, the disease is presumed caused by work and benefits are automatic without further investigation.

Readers are welcome to contact the ODP at (416) 327-4156 to receive more detailed information about these decision-making principles used by the Panel.

b) The issue and how it arose

In 1983, Dr. Jan Muller et al. of the Ministry of Labour, published a report entitled "Study of Mortality of Ontario Miners, 1955-1977, Part I" sponsored jointly by the Ontario WCB and the Atomic Energy Control Board of Canada. Although focusing primarily on lung cancer, some analysis was done on stomach cancer. The authors found a statistically significant 48% increase in stomach cancer among underground goldminers (61).

In 1984, Muller and colleagues described the results from an extension of the follow-up of their cohort of uranium miners through 1981. This study found an excess of stomach cancer among uranium miners who had prior gold experience compared to the no gold portion of the cohort (62). This finding was later confirmed in 1989 in an updated report (65).

In 1986, in conjunction with colleagues from the WCB and the University of Toronto, Dr. Muller provided Part II of the earlier study, taking a more detailed look at the excess of stomach cancer. Their new analysis found that goldminers experienced a 57% increase in stomach cancer (63). In response to queries from the WCB and others, an explanatory Addendum was also published in 1987 (64).

These findings prompted the United Steelworkers of America (USWA) in 1986 to request that the government authorize compensation of the victims' families from the WCB, in the belief that the cancers were work-related (81). When the Steelworkers made their request the ODP had just been created and was not yet functioning. For that reason the WCB decided to undertake an investigation of its own and asked Dr. A. B. Miller and his colleagues at the University of Toronto to investigate the pattern of lung and stomach cancer among Ontario goldminers. The Miller mandate was to evaluate the questions originally posed to the ODP.

In a letter dated August 18, 1986, the WCB had requested the ODP(1) to respond to a number of questions about the occurrence of lung and stomach cancer among goldminers (53). Regarding stomach cancer specifically, the Board asked:

"Does there exist sufficient evidence for the WCB to provide compensation benefits to gold or mixed ore miners who have subsequently contracted stomach cancer? If the answer to the [above] question is yes, then what criteria should the Board employ to compensate this category of workers?"

The Panel began an investigation to answer the Board's questions and struck a special scientific sub-panel, chaired by Dr. Harry Shannon to conduct an epidemiological review of the lung and stomach cancer mortality experience of various mining cohorts in Ontario. Both Dr. Miller and Dr. Shannon employed data from the Mining Master File (MMF) as had Dr. Muller.

The ODP reported the results of both scientific investigations. Dr. Miller found a 36% increase in stomach cancer among goldminers. While this increase in incidence of stomach cancer could represent an occupational role in disease, Miller suggested reasons for scepticism: no dose-response relationship; mixed ore miners showed no increased risk of stomach cancer; and few other studies of goldminers showed an increase in mortality from stomach cancer (57).

Shannon assembled a cohort which was larger and included more multi-ore miners than the Miller study. The Shannon sub-panel reported an overall stomach cancer excess of 42% among goldminers. Similar to Miller's study the highest standardized mortality ratio (SMR) was found for those with first exposure after 1950 (83).

In April 1987, based on the findings of these two studies the IDSP issued the final report to the WCB, (36). In that Report , it concluded that at best, the relationship between stomach cancer and gold mining was considered inconclusive. The Panel stated its intention "to review the evidence when the Muller report is updated and requests that as full as possible an update be carried out as soon as possible".

In its submission to the WCB following release of the IDSP report, the USWA suggested that the causative agent for stomach cancer for those miners who worked after 1945 might be found in the mixture of new chemical agents such as aluminum dust, blasting agents and diesel fumes (95).

The topic of stomach cancer and goldminers re-appeared on the ODP agenda when a further study was published in 1993. Kusiak et al. reported a 52% excess of stomach cancer among Ontario goldminers(44). This last study initiated the current investigation by the Panel.

c) The focus of the investigation

To answer the questions originally posed by the WCB in 1986, the Panel asked itself the following:

1. What is the evidence concerning stomach cancer mortality among non-gold miners worldwide and in Ontario?

2. What is the evidence concerning stomach cancer mortality among goldminers worldwide and in Ontario?

3. What is known about potential occupational causes of stomach cancer?

4. Are any of these and other potential causes relevant in the Ontario gold mining environment?

5. What, if any, are the differences between gold mining and other kinds of hardrock mining in Ontario?

6. Are there other risk factors that could affect the pattern of stomach cancer?

d) The investigative process

As the first step in its most recent investigation the ODP commissioned Dr. Susan Stock, Medical Consultant, Occupational and Environmental Health Unit of the Montreal Department of Public Health, to conduct a review of the world literature relating stomach cancer and occupation. That paper, entitled Gastric Cancer and Occupation: A Review of the Literature (1993), was reviewed by Dr.J.M. Peters (78), Dr. M. Neuberger (71) and Drs. M. Isler and M. Caron (39). Dr. Stock revised her work in accordance with the comments in 1994 (87). Appendix A of this report sets out the studies she evaluated together with her references.

e) Submissions

Following the review the Panel made copies available to interested stakeholders. Notice of the Panel's investigation was published in August 1993 in the Ontario Gazette. The following chose to comment:

f) Other Panel Initiatives

g) Other Jurisdictions

British Columbia, Manitoba, Newfoundland, Saskatchewan and Ontario have policies dealing with stomach and/or gastrointestinal cancer with a clear history of exposure to asbestos.

Alberta, New Brunswick, Northwest Territories, Nova Scotia, Prince Edward Island, Quebec and the Yukon adjudicate such claims on their individual merits with no specific policies in place.

The Ontario Board reports that it has received claims for stomach cancer from gold miners, but that none have been allowed due to gold mining.(50)


a) Background information about stomach cancer

Stomach cancer: what is known in general

Types of gastric cancers

Please refer to Figure 1 for the structure of the human stomach.

Approximately 85-90% of gastric cancers are gland like malignant tumours (adenocarcinomas), and according to the Lauren Classification based on histology, they are divided into two types: intestinal or diffuse (47).

Intestinal types are marked by cohesive abnormal cells forming gland like tubular structures, believed to have a long pre-cancerous process. Most ulcerating tumours fall into this category, and these tumours occur most often in the antrum. The intestinal type is more common in men and patients over 60 years of age, and appears to be more affected by diet/environmental factors than the diffuse type. The intestinal type is also

Figure 1.

The Stomach

associated with a pre-cancerous stage of gastric atrophy and metaplasia(2) (28).

Diffuse types lack cell cohesion. These predominate in younger patients and develop throughout the stomach (including cardia), resulting in a "leather-bottle appearance", with a far more ominous prognosis (86). The incidence of diffuse gastric carcinoma appears to be similar in most populations worldwide (52).

b) Preliminary considerations

In the following pages we will discuss the mortality from stomach cancer among miners.

The following explanation of Standard Mortality Ratios may assist the reader:

"Standardized Mortality Ratios" (SMR)

Results are measured in terms of a "standardized mortality ratio" ("SMR"), which is an estimate computed by comparing the number of deaths observed among miners with the number of deaths which are expected based upon a comparison group of the same age and sex, during the same time period (then multiplying by 100):

                  "observed" deaths among miners         
SMR =   ------------------------------------------------ X 100
             "expected" deaths among comparison group  

For example, if among 1000 miners three died of stomach cancer, whereas two of 1000 individuals in the comparison group of the same age and sex died of stomach cancer:

         (observed) 3		    
SMR is -----------------   X 100 = 150.

An SMR greater than 100 would suggest an excess risk of stomach cancer. Epidemiologists evaluate the statistical significance of an SMR by using a 95% confidence interval, a range of numbers in which the true SMR would fall 95% of the time. If the lower end of the 95% confidence interval is above 100, the likelihood that the excess mortality is due to chance is less than 5% (or 1 out of 20). In this Report, confidence intervals, where available from the original paper, are noted. When not available, they were calculated using the following formula:

Lower limit = [Square root of Observed events - (1.96 x 0.5)]2

Upper limit = [Square root of Observed events + (1.96 x 0.5)]2

It is also important to note that the SMRs may not accurately reflect the excess mortality or true risk for the following reasons:

Healthy worker effect

Most epidemiological studies compare workers with the general population. Since the general population includes people who do not or cannot work due to illness or disability, a working population is usually healthier and is, therefore, expected to have a lower mortality rate for most causes of death. The influence of these factors on the results of studies is known as the "healthy worker effect". It results in lower SMRs than would occur if more similar groups had been compared and may conceal a real increase in deaths among workers. In other words, the phenomenon results in an underestimation of the excess caused by work exposures. Comparisons with another group of "healthy" workers, rather than to the general population, are therefore more likely to provide accurate statistical estimates of occupational risks.

Whether or not the healthy worker effect influences cancer mortality ratios is controversial. In a previous Report of Findings [HWE report], the IDSP published comments on the healthy worker effect solicited from nine experts. The Panel's review of those opinions led it to conclude that the healthy worker effect must be taken into account when interpreting epidemiological studies of mortality or morbidity from any cause, including cancer.

Accuracy of the data

Because epidemiological cohort studies follow subjects over a long period of time many people may be difficult to trace. A portion of these subjects may die outside Canada and their deaths would not be accurately recorded.

The ODP has concluded that the SMR for stomach cancer has been underestimated in all of the studies of Ontario miners primarily because of the lack of social insurance numbers (SINs) to accurately ascertain mortality. In fact, the vital status of 25% of all miners without SINs could not be determined, compared to the ascertainment of the vital status of all but 5% of those miners with SINs (45).

Estimating the extent of disease when comparing working populations to the general population

When the SMR is computed, the general population, on which the expected number of deaths is based, includes the observed number of deaths from the study population. This is because these deaths contribute to both populations: i.e., the deaths are counted in both the observed and the expected, thus underestimating the real excess. If deaths from a disease such as nasal cancer are caused almost always by occupational factors, the SMR will thus be even more underestimated.

The preceding factors constitute the framework within which reports of SMRs need to be considered. The following account of the SMRs found in mining cohort studies are affected by questions of data accuracy, the healthy worker effect and population comparisons. They should be evaluated with these factors taken into account.

c) The questions

1) What is the evidence concerning stomach cancer mortality among non-gold miners worldwide and in Ontario?

i) International studies - Non-gold Mining

Figures 2 and 3 on the following pages show the SMRs from studies of stomach cancer among non-goldminers worldwide and in Ontario respectively.(3)

Statistically significant elevated stomach cancer rates were found in copper mining in China (13,15) and iron ore in the U.S. and France (48,79).

Although not statistically significant, excesses were found likewise in tin miners in the U.K (22,30), iron ore miners in the U.K.(42) and France(12); and uranium miners in the U.S.(97), N.W.T. and Saskatchewan(67), and Newfoundland (59).

Four studies, two among mixed ore miners in Italy (11,14) and two among iron ore miners in China and France (14,66) do not show increases in stomach cancer mortality.

It should be noted that although there are a large number of studies on stomach cancer among coal miners, they are not reviewed here but are referred to later in other sections of the text. Such studies also formed part of the review on the worldwide literature conducted by Dr. Stock.

ii) Provincial studies - Non-gold Mining

In Ontario the rates of stomach cancer among all miners including gold, nickel, uranium and mixed miners have been considered by Muller and Kusiak during three different periods of follow-up: 1955-77; 1955-1981 and 1955-86.

In 1983 Muller et al. reported an SMR of 108 for nickel/copper miners (61)

and in 1993 Kusiak et al. found an SMR of 101 (44). Uranium miners, other ore and mixed ore but non-goldminers likewise showed no statistically significant increase. The 1985 study by Nair et al. was on refinery workers extracting radium from uranium ore and who overall exhibited a strong healthy worker effect (67).

iii) Conclusions

As cited in Figure 2, the evidence of an increase in the rates of stomach cancer among non-gold miners worldwide is inconsistent. However, the literature supports the potential work-relatedness of stomach cancer among certain occupational groups as described in the review by Susan Stock (87).

There is no evidence of a statistically significant increase in stomach cancer mortality among non-goldminers in Ontario.

2. What is the evidence concerning stomach cancer mortality among goldminers worldwide and in Ontario?

i) International studies - Gold Mining

There are three studies of non-Ontario cohorts which examine the mortality from stomach cancer among goldminers; two of these studies show elevated rates. Wyndham reports an elevated but non-statistically significant excess of stomach cancer (SMR 157, CI 78-281) when white South African goldminers were compared to the white South African male population (107). In a study of Russian workers who were exposed to non-fibrous mineral dusts, Katsnelson and Mokronosova reported a relative risk (RR) 5.7 for stomach cancer for miners working in gold mines (41). The only cohort which did not show an excess was a study conducted of the goldminers in South Dakota (8). The SMR for that cohort was 57. This unusual finding is in keeping with the unique experience in South Dakota which also reports lower than expected SMRs for lung cancer, in contrast to most other studies of goldminers.

Figure 4, on the next page, illustrates the mortality experience of underground goldminers worldwide, including Ontario. More detailed information on the cohorts follows on Table 1. (Although other studies exist, the Panel has routinely excluded studies which report fewer than five deaths for the cause of interest and that practice has been adopted here, except for the sub-comparisons among gold mining camps within Ontario).

ii) Provincial studies - Gold Mining

All of the studies of Ontario goldminers confirm elevated rates of stomach cancer. The five analyses of the Ontario cohort (Muller et al., 1983; Muller et al.,1986; Shannon 1987; Miller 1987; and Kusiak et al.,1993) used the same gold mining cohort with modifications and different follow-up durations from the MMF maintained by the Ministry of Labour. For the various studies the cohort has been described as follows:

Muller et al, 1983, defined a goldminer as a miner who (a)spent at least 60 months in non-uranium mines, of which 85% was in gold with a minimum of two weeks in an Ontario mine (b)was alive between January 1, 1955 and December 31, 1977 and (c)underwent a chest x-ray each year as required for a mining

Muller et al (1983) 7,542 1955-77 (23 years) Ontario males N. Ontario - -non-uranium, 5 yrs mining, 2 weeks Ont mine 85% gold mining by 1977, at death, or becoming uranium miner

-annual chest x-ray after 1954

-underground and dusty surface jobs

148 (113-188) 60/40.4 - Ont.

148 (103-205)-35/23.7-Timmins

121 (63-211)-12/9.9-Kirkland Lake

Ft. William


-highest mortality <50 years old (SMR=393, 9/2.3)
Muller et al (1986, 1987) 7.057 1955-77 (23 years) Ontario males


N. Ontario - -non-uranium, 5yrs+ mining, 2 weeks + Ont mine, 85% gold mining by 1977, at death or becoming uranium miner

-annual chest x-ray

-must be employed anytime 1955-1977 2 weeks

-FT if 50% underground

-underground and dusty surface jobs

157 (118-121) 54/34.3

154 - born in Canada

136 - born outside Canada

238 (no CI available)

-immigrant status not a satisfactory explanation of excess
Wyndham et al (1986) 3.971 white males 1970-79 (10 years) white S. African males S. Africa -miner's lifestyle (diet, smoking, alcohol) -born between 1916-1930 and alive 1970

-followed for nine years

157 (78-281) 11/7 -rejects healthy worker effect

-miners lifestyle less healthy than pop.?

-too labourious to collect smoking histories

-no arsenic found in certain mines

-dust level index counted

Brown et al (1986) 3,328 white males 1940-65 (26 years) S. Dakota State males S. Dakota, U.S. -arsenic (to 1952) confounding effects from amphibole fibres or silica -FT underground at least 1 year, 1940-65

-non-uranium exposure

57 (18-133) 5/8.8 -rates possibly underestimated due to standard pop. used

-smoking prohibited in mines until 1952

-negligible trace metal quantities in air

-dust level index exposure created

-reduced dust exposure underground due to reduced working hours from 1920 to 1986

Miller et al (1987) retrospective mortality cohort 7,057 1955-77 (23 years) Ontario males N. Ontario -ethnicity -5 yrs. + gold mining, 2 weeks + Ont. gold mine between 1955-77

-underground & dusty surface jobs

-mixed ore & gold (non-uranium)

-annual chest x-ray

-85% gold

136 (105-171) 66/48.5

113 (81-145) <1945

221 (159-283) >1945

-cigarette smoking weak factor

-healthy worker effect?

-low colo-rectal mortality

-heaviest dust <1945

Kusiak et al (1993) 11,097 ever gold never uranium

never gold

1955-87 (33 years) Ontario males N. Ontario -metal dust


-5 yrs + mining after 1954

-2 weeks +, Ont. mining

-annual chest x-ray

-85% gold

152(125-185) 104/68.2

160 (115-218) 41/25.6 -born outside N. America

102 (88-128)

-no increase for non-gold miners

-rejects arsenic, diesel emissions, mineral fibre, aluminum powder

-5-19 yrs latency

-no dietary confounders considered

-excess greater in <60 age group

-weighting to dust

-<60 intestinal rather than diffuse gastric cancer

-no direct exposure measurement

-best fit to dates chromium and/or dust

Shannon et al (1987) 10,185 ever gold

6,786 pure gold

1955-77 (23 years) Ontario males

N.Ontario males

N. Ontario -labour intensive efforts as indicated by job -5 yrs+ gold mining

-underground & dusty surface jobs

-2 weeks + Ont gold, between 1955-77

-mixed ore & gold (non-uranium)

142 (109-180) 63(44.4)

114 (88-146) 63 (55.3) N. Ont.

-10-19 yrs latency

-inverse dose response

-excess higher for native born Canadian/US

-quantifying exposure and levels of contamination not available

-reduced dust over time and varied mine to mine

Katsnelson et al (1979) 1948-74 (27 years) -corresponding town in Russia males Russia -silica

-PAH in tobacco smoke

-nonfibrous industrial dust

-underground and surface broken into occupational groups RR=5.7 underground

RR=1.1 surface

RR=4.4 underground minus silicosis persons


Muller et al.,1986, restricted the cohort to require that an individual had to actually be working in a gold mine(i.e. not retired from gold mining) when reporting for his annual chest x-ray (63).

Shannon specified 60+ months as a goldminer with a minimum of two weeks in an Ontario gold mine (83), while Miller 1987 (57) used Muller's 1986 cohort definition.

Kusiak et al., 1993, broadened the cohort by dropping the 85% rule and defined a goldminer as someone who has ever worked in a gold mine but has spent 60 months (five years) or more as a non-uranium miner, with two weeks or more in an Ontario mine, as well as having an annual chest x-ray taken (44). This definition could therefore include mixed ore mining, and underground and dusty surface jobs.

The overall rates of stomach cancer have been identified as follows:

Gold Mining and Stomach Cancer

1993 Kusiak (Ont) n=104 SMR 152 CI 125-185
1987 Miller (Ont) n=66 SMR 136 CI 105-171
1987 Shannon (Ont) n=63 SMR 142 CI 109-180
1986 Muller (Ont) n=54 SMR 157 CI 118-202
1983 Muller (Ont) n=60 SMR 148 CI 113-188

As part of his analysis Shannon compared the goldminers to the general Ontario male population and also to the Northern Ontario male population The SMR for stomach cancer was reduced from 142 to 114 (not statistically significant) when the Northern Ontario population was used as the control group. ODP staff conducted a person-years analysis and confirmed that the stomach cancer deaths among the gold miners formed less than five percent of the total number of deaths from stomach cancer in Northern Ontario counties (89). Thus there was little bias caused by including mining-exposed deaths in the Northern Ontario reference rates. However, Muller et al., 1986, had compared the goldminers to what was perhaps the most ideal control group, the nickel/copper miners. In this comparison, Muller found that the SMR for the goldminers was a statistically significant 238 in those under the age of 70. Kusiak et al., 1993, also observed that although empirically,

" mortality from stomach cancer for gold miners ought to have been compared with that of men living in Northern Ontario......many of the men in Northern Ontario worked in the mines and no excess gastric carcinoma relative to the male population in the whole of Ontario was evident in northern Ontario miners who mined nickel and copper or uranium only."

Figure 5 reflects the efforts of Muller et al., in their seminal 1983 study, to search for any differences in mortality between gold camps which were located as much as 500 miles apart. In all gold camps, an excess of stomach cancer mortality was found, although not always reaching statistical significance because of small numbers. However when placed in one gold cohort, statistical significance is achieved.

In addition to the research chiefly focused on the variously defined gold mining cohorts, other evidence produced concurrently or as off-shoots of other studies needs to be considered.

Muller et al., 1983 found excess stomach cancer among mixed ore miners only if their mining experience had included gold. In mixed camps with no gold, there was no excess (Fig.3). In 1984, Muller et al. extended the follow-up period for uranium miners in their original cohort from 1977 to 1981. They found an excess of stomach cancer (SMR=157, CI 84-268) only among uranium miners who had also mined gold. This finding was confirmed in 1989 as a statistically significant excess (SMR =196, CI 118-306) again only among uranium/goldminers. This evidence is shown in Figure 4 as mixed gold mining. In both cases the uranium only miners showed no excess with SMRs of 94 and 95, respectively.

In an unrelated but relevant study, also shown in Figure 4, Finkelstein et al.,1987 (20) found a statistically significant excess of stomach cancer among compensated silicotic miners (SMR=188, CI 120-290). It can be reasonably assumed that the vast majority of these miners were goldminers because they received compensation for silicosis before 1970. The only other miners likely to have developed silicosis are uranium miners who, in most cases, would have had insufficient exposure prior to 1970 to have qualified for workers' compensation benefits.


Calendar date of first employment

Excesses of stomach cancer are found in all time periods studied. The highest SMRs are observed among gold miners first hired after 1945. Most authors, excluding Kusiak et al., provided findings for those workers who commenced employment pre and post 1945 (Shannon,1987; Muller et al.,1987; Miller,1987).(4)

In addition, Shannon grouped those men first exposed from 1950 onwards and found that they had the highest SMR (SMR=330, CI 178-570).

The findings are shown in the table below and illustrated in Figure 6.

Year of First Employment (Post-1945 Start) and Stomach Cancer Mortality
Miller et al., 1987 n=19 SMR 322 CI 240-405
Shannon et al., 1987 n=22 SMR 286 CI 179-418
Muller et al., 1987 n=19 SMR 322 CI 194-483


The 1987 studies of Muller et al., Shannon et al., and Miller et al. all observed that younger goldminers are more at risk of stomach cancer than older ones.

Muller et al. found the highest SMRs among the goldminers in those under the age of 50 (SMR 393, CI 180-746). The risk decreased with age.

Regression analysis found that the effect of age was statistically significant.

Miller et al., looking at age at death among gold and mixed ore miners with stomach cancer, found a statistically significant increase only among

those miners employed in 1945 or later and under the age of 50. For reasons that are explained beginning on Page 33 of this report, Miller's results are displayed in Figure 7 without regard for period of hire.

The 1993 study of Kusiak et al. confirmed these earlier findings. The excess of stomach cancer was higher in gold miners under the age of 60 (SMR 167, CI 122-223) than in gold miners aged 60-74 (SMR 143, CI 109-184). As mentioned on p. 8, the incidence of stomach cancer normally increases with increasing age.

Time since first exposure

Shannon, 1987 found stomach cancer mortality peaked 10-19 years since first exposure to gold mining. After 30 years since first exposure no elevated rates of stomach cancer were observed. This pattern was confirmed by Kusiak et al., 1993, who found that statistically significant excess stomach cancer deaths appeared five to 19 years after the miners

began gold mining in Ontario.

Duration of exposure

In his 1987 study, Miller et al. found no relationship between duration of exposure and stomach cancer mortality.

Shannon et al. found that stomach cancer mortality was most increased in men exposed for between 10 and 14 years while durations over 25 years showed only a slight increase. When Shannon attempted to weight the exposure by year of first employment an inverse dose response was found, i.e. the SMRs decreased for those periods of longest weighted duration of exposure to dust. Shannon did not perform an age-specific analysis.

Kusiak et al., 1993, who reported that the average duration of employment in the gold mines was 10.8 years, found a statistically significant relationship between time-weighted duration of exposure to dust and/or chromium in gold mines and the SMRs for stomach cancer in gold miners under the age of 60. The association was strongest for weighted duration of exposure to chromium. For miners between 60 and 74, the association was not statistically significant.


Although it is well recognized that ethnic origin affects the incidence of stomach cancer in populations, the studies of Muller et al. (1986) and Shannon (1987) found that immigrant status was not a satisfactory explanation for the excess of stomach cancer among the gold miners since the SMRs were higher among those gold miners born in North America than those born elsewhere.

Kusiak et al.,(1993) found that with a latency of less than 20 years since starting gold mining in Ontario, there was no difference in the SMR for stomach cancer in miners born in North America (SMR 255, CI 139-428)or outside North America (SMR 270, CI 140-472). However, for a latency in excess of 20 years, only gold miners born outside North America showed a statistically significant increase in the SMR for stomach cancer (SMR 160, CI 115-218). (See Figure 8).


Using records from the Ontario Cancer Treatment and Research Foundation (OCTRF), Kusiak and colleagues compared histological types of stomach cancer in miners and controls. The table below is based on data in Kusiak et al's 1993 paper, analyzed by ODP staff (90). Stomach cancers in each age/exposure group are classified as either intestinal or diffuse according to OCTRF records. In all but the young gold miners, intestinal-type cancers make up about two-thirds of the stomach cancers. Among the young gold miners, 90% of the cancers are of the intestinal type. However, the numbers among the younger gold miners are small. For example, if two of their intestinal cancers were diffuse, then this group would exhibit essentially the same pattern as the other two groups.

Using logistic regression, a formal test of whether the younger gold miners had a higher proportion of intestinal-type cancers gave a p-value of 0.13, indicating that the observed difference is not beyond what would be seen by chance 13% of the time.

In the table below, the first number in each cell is the observed count of cancers and the second number is the corresponding percentage in the age/exposure category.
Gold Miners
Other Miners
Hospital Controls
Cancer Type
Cancer Type
Cancer Type
Age Int. Diff. Total Int. Diff Total Int. Diff Total
< 60
















> 60
































iii) Discussion

A dose-response trend is shown when an increase in the "dose" (exposure level including intensity and duration) corresponds to an increase in the "response" (usually a health outcome and in this paper a death from stomach cancer). A dose-response relationship is shown when more disease occurs after greater exposure to a substance. Epidemiologists usually demand verification of dose/response before they are willing to indicate that a causal relationship exists.

In the studies of the Ontario goldminers and stomach cancer, the only example of a dose/response comes from the Kusiak analysis which shows a statistically significant association with the time weighted exposure to chromium and/or dust in gold miners under the age of 60.

Coupled with the lack of a traditional dose response is the repeated finding that those who started working in gold mining after 1945 are at greater risk for stomach cancer than those who mined in earlier years. This finding has perturbed the investigators because it has always been assumed that dust levels and working conditions were more hazardous prior to 1945. On the basis of this assumption researchers expected that the rates of disease would also be higher among those workers who began employment before 1945.

There are several possible explanations for this phenomenon: No Causal Relationship
There is no occupational association and hence no causal relationship between work in a gold mine in Ontario and excess mortality from stomach cancer. Social Insurance Numbers.
The lack of a dose response in the Miller and Shannon studies may have occurred, in part, because of an under ascertainment problem among miners who started before 1945 and stopped working before 1965 when social insurance numbers (SINs)were introduced. This would mean that many deaths of miners employed during the dustier exposures for the longest periods of times may not have been captured. Most of the Ontario studies report higher SMRs for those gold miners who started working in the mines after 1945.

As noted by Kusiak et al., in their 1993 study of lung cancer in uranium miners , SINs were available for only 63.1% of the men in that cohort largely because the numbers were not issued until 1965. For the uranium miners with SINs the SMR was 225, whereas for those without, the SMR was 135. The social insurance number is a valuable aid in determining vital status. Twenty-five percent of those miners without SINs could not be identified as dead or alive versus all but five percent of those with SINs.

Muller et al., in their 1983 study among goldminers, had encountered the same problem. SINs were available for only fifty percent of his cohort approximately. Even using manual resolution methods available to them, Muller and colleagues were forced to reject almost fifty percent of the possible links.

An analysis conducted by ODP staff however, looked at the percentage of gold miners with SINs hired before and after 1945 (91). The post-1945 hires still included a fairly large proportion without SINs. The statistical analysis indicates that differential death ascertainment in the two groups (i.e., pre and post 1945) can only explain about 10% of the difference in the SMRs between these two groups. Factors other than lack of SINs must be considered to account for the 42% lower SMR for stomach cancer in those hired before 1945. Cohort age structure.
As illustrated in Figure 6, apparent SMRs for stomach cancer mortality are higher in gold miners first hired after 1945. However, Kusiak et al., 1993, has observed :

"The association between the excess of stomach cancer and starting gold mining after 1945 can be interpreted as an artefact of the strong associations between the relative excess of mortality from stomach cancers and both age and time since first gold mining. The mortality in the younger age groups of the miners who first worked in gold mines before 1945 would be only a small part of the mortality from stomach cancer seen in these miners."(p.124)

A regression analysis performed by ODP staff supports the concept that comparing SMRs for pre-and post-1945 hires is dangerous, as the age structures are drastically different for the two groups (92). Therefore, the SMRs of 322 and 123 in the 1987 paper of Muller et al.,are misleading


Because the pre-1945 gold miners are older, the expected number of deaths, as with most cancers, is higher. Since excess risk is limited to the younger age groups, the excess of observed deaths over expected is small.

In contrast, post 1944 hires tend to be younger and are therefore in age groups where the expected number of deaths from stomach cancer is low. Among younger miners, the relative risk of stomach cancer due to gold mining is high. The potential effect of this differing age structure on the SMRs is shown below with hypothetical numbers:

Pre Young 10 5 2
1945 Old 70 70 1
Total 80 75 1.2
Post Young 20 10 2
1945 Old 10 10 1
Total 30 20 1.5

By ending the follow-up of the gold-miners cohort at various times, we can see that the pre/post distinction decreases as both groups reach older ages where baseline rates are higher and mining-associated relative risks are lower. This is illustrated in the table below:

SMR by period of hire

End in year Pre-1945 hires Post-1945 hires
1965 1.64 3.85
1970 1.32 3.01
1975 1.32 4.03
1980 1.31 3.13
1985 1.25 2.60

The Effect of Age - Regression Analysis
To further examine the effect that age and other factors have on the SMR for stomach cancer, Poisson regression was performed on the data grouped by era of hire, age, age at hire, number of years of mining gold, and active versus inactive status (according to whether a miner was listed as employed or not in the MMF). The work was done in collaboration with Professor Paul Corey of the University of Toronto and full details of this analysis are reported in Appendix B of this report and an ODP memo (93).

The main findings of this regression analysis are summarized below and are based on what was determined to be the simplest model that gave a good fit to the data.

(1) When age, age at hire, years of gold mining and active status are controlled for, there is not a significant difference in risk between the pre- and post-1945 hires (p=0.30)

2) Active miners have a much lower risk of stomach cancer (p<0.0001), a fact that probably reflects how unlikely it is that a miner with stomach cancer continues to work, but also reflecting the generally better health of those who are able to continue full time work.

(3) Risk for stomach cancer increases with age at hire into gold mining (p=0.0009). It is unclear whether this represents an actual effect of later onset of exposure or represents some other unidentified characteristic of those who start gold mining at a later age.

(4) Years of gold mining and attained age are not independent risk factors (p=0.006). For younger miners, relative risk for stomach cancer increases with number of years of gold mining, whereas for the oldest miners, there appears to be no relationship between years of mining and relative risk - see Figure 9. There is no single age where risk drops off: the model assumes a continuous decline in risk with age for any particular exposure level. For example, the relative risk for 25 versus 2.5 years of gold mining drops from over 6 at 45 years of age to less than 1 at 70 years of age. For different gold exposures, these numbers would be different. Overall, though, according to this model, the upward trend in risk with years of gold mining is statistically significant below age 62 .

Choice of the Reference Population
As mentioned previously, Shannon, 1987, analyzed stomach cancer mortality among gold miners with the Northern Ontario male population as the control group and observed a lower SMR in both the North American-born and non-North American-born gold miners from that observed when the Ontario male population was used as controls.

To see if stomach cancer in the gold miners had a different pattern with age or calendar time from that in the Northern counties, three comparisons were made:

The results, shown graphically in Figures 10-12, indicate that:

In other words, stomach cancer in the gold miners does not follow the same trends as that in the males of Northern Ontario when both are compared to the Ontario male population. The gold miners' relative risk of stomach cancer decreases with increasing age, while the relative risk among the Northern males compared to Ontario males remains constant with age.

Thus Shannon's 1987 comparison using the Northern population as a reference group would have the effect of reducing the SMR, but the present age-stratified analyses provide strong evidence that the excess of stomach cancer among gold miners is due to something other than merely being Northern residents.

Other Considerations
The make-up of the cohort also reflects the following:

The gold miners had to have been alive and working in 1955 or later to be in the cohort. Those who fell ill, died, retired, or switched employment to a position not requiring an annual x-ray before 1955 could all have been lost to the cohort, even though they might ultimately have died of stomach cancer.

With a relatively uncommon cancer such as stomach, under ascertainment of deaths could have an even more profound effect on the SMR than was apparent in Kusiak's later study of lung cancer among uranium miners.

In Kusiak et al., 1993, the miners were followed until the end of 1986. Included in the analysis was any miner who had 60 months of experience and had a chest x-ray between 1955 and 1977. While miners continued to be added to the cohort it is unlikely that there would be sufficient follow-up to examine the real rates of mortality for those miners who commenced employment after 1967.

Detailed histological typing for the stomach cancers among the gold miners was obtained for the first time in the 1993 study of Kusiak et al. Given the apparent absence of a true dose response and the indication of high rates of intestinal cancers among younger workers, (see analysis on p. 32), it may be that it is only intestinal cancer which is associated with work. Although admittedly speculative, a dose response based on both histological types could be difficult to demonstrate looking at the entire cohort.

iv) Conclusions

3. What is known about potential occupational causes of stomach cancer?

i) Dr. Stock's Review

In a 1994 Panel commissioned review of the literature on stomach cancer, Stock addressed the above question by summarizing that:

"In reviewing this literature one is struck by the huge number of studies demonstrating a statistically significant association between occupation and stomach cancer despite enormous variation in methods and approach and the inherent difficulty in demonstrating such associations in occupational cancer epidemiology....

While no one study is methodologically strong enough to provide unequivocal evidence of a causal relationship between stomach cancer and the occupation under study and none has quantitative exposure measurements, there are now numerous examples of excess risk of stomach cancer in various occupational groups supporting the hypothesis that occupational exposures can contribute to the development of stomach cancer....

Although there are probably at least 100 studies which have examined a link between occupation and stomach cancer, few articles in the literature have attempted to explain potential mechanisms for occupational gastric carcinogenesis. The main theories found in the literature have arisen in an effort to explain stomach cancer in coal miners and were described by Ames who suggested the following hypotheses: (emphasis added)

1. Poly-aromatic hydrocarbons (PAHs) especially benzopyrene may be absorbed by coal mine dust. Absorbed precarcinogens may be activated by the (cytochrome)p-450 oxidase enzyme system and lead to intragastric production of carcinogens that attach (to) mucosal tissue and induce gastric carcinoma. PAHs such as benzopyrene are known to transform epithelial cells and initiate cancer. Combustion products of coal have been shown experimentally to be carcinogenic to the gastric epithelium in mice and in tissue culture (Weinberger et al). A similar PAH hypothesis is proposed by Nutt as a likely explanation of the stomach cancer found in rubber workers. PAHs are also found in lubricating oils and are a possible explanation of increased stomach cancers found in bearing plant workers found by Park et al and Silverstein et al. Interestingly the sharp decline in gastric cancer in many developed countries parallels the drop in the use of coal burning in homes and this cancer is high in those countries where coal burning for cooking is still common.

2. Exposure to dust in general may cause gastric cancer. Respirable dust may be subject to macrophage scavenging, brought up through mucociliary clearance and swallowed as with other inhaled particles. Coal mining dust exposures include iron dust, iron oxides from welding, other metals found in coal, rock dust, particulate from blasting, asbestos insulation, coal shuttle brake linings dust and diesel particulate. Coal miners may also have been exposed to other non-occupational environmental sources of dusts such as soot from coal combustion from cooking and/or heating or lived near nuisance dusts or soil erosion dust. This hypothesis is supported by studies of quarrymen and some air pollution studies. Pham et al propose an interesting mechanism for the carcinogenesis of general dust exposure in their efforts to explain the high rate of stomach cancer in iron ore miners. They suggest that the inhaled iron oxide dust may have acted by production of free radicals arising from surface activity by a mechanism of oxygen reduction which is greatly increased in freshly generated dust. This physicochemical surface activity has also been found in silica and asbestos fibre dust (Costa et al 1989). It may also be relevant in the high rates of stomach cancer found in those in grinding occupations in bearing plants.

3. Gastric cancer risk may be associated with exposure to nitrosamines, their precursors, or to nitrosation facilitators. Coal miners were exposed to nitrosamines and nitrogen oxides released from coal by pyrolysis; nitrates from inhalation of particulate after blasting (nitroglycerin or ammonium nitrate); dietary sources in smoked or salted fish or bacon (smoked foods also contain benzopyrene); drinking water sources from wells contaminated with high levels of nitrates (1981 studies by Armijo et al suggest a co-factor of arsenic and selenium to explain nitrate exposure and gastric cancer); nitrosation enhancement from chewing tobacco and/or bacterial contamination from sewage polluted water, poor dental health, and poor preservation of and storage of food. Similarly nitrosamines may be implicated in stomach cancer in rubber workers (Nutt et al) - exposure was very high when N-nitrosodiphenylamine was in use. Nitrosamines are also found in soluble oil and synthetic cutting fluids and may be relevant to the high stomach cancer rates found in bearing plant workers.

Stock continues....

Another hypothesis was proposed by Meyer et al. to explain stomach cancer in coal miners but was later refuted by Ames and Gamble and not supported by subsequent studies. Meyer et al. suggested that stomach cancer risk is increased when particle retention in the lungs is decreased, i.e., particles are more rapidly cleared to the mouth and swallowed when there is no pulmonary impairment; gastric cancer risk would be lower when lung clearance is reduced because of pulmonary function impairment.

These hypotheses are of interest and could potentially explain a high proportion of the observations of increased mortality from stomach cancer in various occupational groups. Alternative mechanisms are obviously possible and would be relevant to explain the Kusiak et al suggestion that exposure to chromium is the main cause of the high rate of stomach cancer among Ontario goldminers observed in that study." (5)

In addition to nitrosamines, there is strong experimental evidence to show an association between other nitroso compounds (e.g. nitrosoureas) and stomach cancers (99). In fact, some of these compounds are the chemicals of choice to induce cancer of the glandular stomach in experimental animals. These compounds appear to act as precursors that are easily converted to the ultimate carcinogens.

As in nitrosamines, nitrosoureas may be produced from nitrite and nitrate by way of chemical reactions that are believed to be possible in a low acid stomach environment. Interestingly, this environment is commonly observed in the stomachs of individuals with stomach cancer. It has thus been suggested that any factor that increases exposure to nitrates and other nitrogen compounds would also increase the risk of stomach cancer (3).

ii) Conclusions:

4. Are any of these and other potential causes relevant in the Ontario gold mining environment?

Gold is found throughout the earth's crust but usually very sparsely, ranging from 0.002 to 0.005 grams per ton. This metal frequently occurs in a free state or as an alloy with other metals such as silver, copper, aluminum, platinum, zinc, palladium and tellurium. Gold commonly occurs in veins among rocks and ores of various other metals such as copper, iron, lead, zinc, antimony and arsenic.

With a low degree of chemical reactivity, gold is generally stable under most environmental conditions. It is not known to be toxic to humans or animals when ingested or inhaled. Occupational health hazards associated with gold are attributed to its extraction by underground mining (as described below), largely as a result of worker exposure to airborne contaminants generated by the mining and milling processes.

Epidemiological studies have reported a strong relationship between stomach cancer and gold mining. Figure 13 illustrates the typical processes carried out in a gold mine. The 1994 review by Stock found that these studies were not designed in such a way as to be able to implicate any specific occupational substance (87) and the evidence of a specific causal agent continues to be speculative. Nevertheless, a number of agents she suggested as possible causes of stomach cancer can be found in the Ontario gold mining environment.

i) Airborne Dusts

Studies of various occupations have suggested an association between dust exposure and increased stomach cancer mortality (85,69). In particular, elevated risks have been reported for coal miners (2,7,19,51)) foundry workers (21) and cement workers (54). The combination of dust and PAHs was implicated in some cases (70), while in others it was thought to be the production of free radicals from cellular activities to reduce oxygen in freshly generated dust (79). The latter has been suggested as a factor in the etiology of carcinogenesis for silica and asbestos (17), as well as for certain metal dusts (87). Metal dusts were implicated in the development of stomach adenocarcinoma in male workers of dusty occupations in Los Angeles, California (106). Some of the metal dusts include beryllium, chromium and nickel.

Wright also found an elevated risk of stomach adenocarcinoma in workers in dusty jobs (proportional incidence ratio = 1.3, 95% CI = 1.2-1.4) after controlling for ethnicity, immigration and socioeconomic status but not cigarette smoking (105). It was suggested that abrasive particulates in the stomach may disrupt the mucosal barrier, permitting either acid or mechanical damage, or cellular contact with carcinogens that would not ordinarily occur in the presence of an intact barrier. A similar hypothesis for mechanical irritation from particulates in food stuff has been offered as a cause of stomach cancer (78).

Dust in underground gold mining is produced by such operations as drilling, blasting, scaling, loading, crushing and moving of broken rocks (mucking). While water is used for dust suppression in many of these operations, this practice has not always been applied in the past. Airborne dust condition in underground mines was made worse in the past by the lack of adequate ventilation.

According to personal accounts from retired goldminers, excessive dust exposure occurred in such ore handling operations as crushing or dumping of ores from one level down to another through ore passes. Dust records for Ontario gold mines were not available until 1958, when konimeter measurements of airborne insoluble particulates were made as an indication of underground dust levels. Average levels from early konimeter measurements were as high as 460 ppcc, well above the recommended guideline of 300 ppcc (26). Since then, levels have gradually decreased (by almost two fold to 250 ppcc between 1960 and 1975) due to improved underground ventilation in 1960's.

High inhalation exposure to airborne dust was made worse by heavy breathing as a result of intensive physical work and by not using respiratory protection of any kind. The practice of using respiratory protective equipment has never been made mandatory in underground gold mining and is not common even today. For individuals working in a dusty environment, some of the inhaled dust particles would be cleared from the lungs and swallowed, increasing the risk for stomach cancer (56,83). Hand-to-mouth contact was another likely route through which dust from gold mining can enter the stomach.

Apart from dust in general, there are also specific contaminants of gold mining dust which may be associated with an elevated rate of stomach cancer. These are separately discussed below:

a) Arsenic

IARC considers arsenic a Group 1 carcinogen with sufficient evidence of carcinogenicity in humans (32). It does not exclude any specific forms or compounds of arsenic in its evaluation and includes miners among those occupations potentially exposed to arsenic.

Recognized mainly as a lung and skin carcinogen in humans, arsenic has also been shown to be associated with cancers of other sites. Although inconclusive, there is evidence to suggest that exposure to inorganic arsenic compounds may be related to the development of stomach cancer. In a study of Cornish tin miners, elevated rate of stomach cancer (SMR = 890) was reported for surface workers involved in the roasting operation of the trivalent arsenic trioxide (30). A higher rate was also noted in underground miners but the increase was not significant and did not appear to be associated with dust exposure underground. Arsenic trioxide was also suggested as a possible cause associated with an increased risk of stomach cancer in a group of Swedish glassblowers (relative risk = 6.4, CI = 3.0-14.0) (102).

Arsenic is found naturally with gold deposits, generally in the form of pentavalent arsenopyrite and often in high concentrations. Although it is found in most Ontario hard rock mines, inorganic arsenic in gold mines is commonly found in greater concentrations. Therefore, the likelihood of arsenic exposure to Ontario gold miners is not in doubt, the question is on the bioavailability of the inhaled inorganic arsenic. Pentavalent arsenopyrite is relatively inert when undisturbed but may become more soluble, and more bioavailable, when disturbed by mining process, natural leaching, microbial action, blasting and contact with water. Urinary levels in Ontario goldminers do suggest that at least a portion of the inhaled inorganic arsenic is bioavailable (40).

It is also reported that arsenic trioxide, as a by-product of gold processing, has also been stored in large quantities underground (10). To what extent this contributes to the exposure of Ontario goldminers to arsenic is not known. However, there is anecdotal evidence indicative of significant exposure in goldminers, as numerous symptoms commonly related to arsenic exposure such as dryness, itching, burning and cracking of the skin have been observed (10). Arsenic related abnormalities such as pitting of the nails, thickening of the soles and palms, linear markings, hyperpigmentation of the hands, feet and legs, and hyperkeratosis were also noted.

On the basis of this study and in conjunction with other findings, stakeholders and the Ontario Ministry of Labour agreed during the consultative process for the designation of arsenic that this substance in underground hardrock mining is bioavailable to a certain extent (101).

b) Asbestos

IARC has designated asbestos as a Group 1 human carcinogen not only in lungs but also in various sites including the gastrointestinal tract (32).

Muller in 1986 postulated the association between the presence of mineral fibres, including asbestos, and lung cancer. Although this work did not look at other cancer sites, a link between asbestos and gastrointestinal cancer has been identified (104). In its report of 1990, the Industrial Disease Standards Panel confirmed the finding by the Royal Commission on Asbestos that occupational exposure to this substance is causally associated with cancer of the gastrointestinal tract, including the stomach (37).

Asbestos may be present in the mining environment both intrinsically and extrinsically. Veins of tremolite, a form of asbestos in the group amphiboles, have been identified in underground mines (88). In one U.S. gold mine, mean total fibre concentration in underground air as measured by transmission electron microscopy was 4.8 fibre/cm3 (range 0.7-11.8 fibre/cm3) (25). Over 80% of the fibres identified were of the amphibole type.

The presence of asbestos in the natural state may not present much of a hazard to non-asbestos miners. However, once disturbed and made airborne, its presence in dust can be a hazard to underground miners. Asbestos can also be brought into the mining environment in its processed form. It is used in the brakes of motorized mining equipment and in insulating materials around pipes.

c) Silica

According to IARC, crystalline silica is a probable human carcinogen (Group 2a) with limited evidence of association to lung cancer in miners and other workers (32). With respect to stomach cancer, it has been linked to industries related to silicosis in one study (41) although other studies were unable to confirm the association (8, 84).

Silica is present in the Ontario gold mining environment, with a concentration of 15-35% free crystalline silica in gold ores. Since similar concentrations were found in total dust samples collected from Ontario gold mines, total dust was measured historically as an indicator of potential silica exposure in these mines.

Since 1960, dust levels in Ontario gold mines have gradually decreased so that the 1975 level was about half of that measured in 1960 (26). In the 1970's, air sampling measurements made on respirable crystalline silica showed average levels consistently at or below the current TWAEV of 0.1 mg/m3.

d) Chromium

IARC recognizes chromium in its hexavalent state as a Group 1 carcinogen with sufficient evidence of carcinogenicity in humans, while the evidence for metallic and trivalent chromium compounds is insufficient (34).

Like arsenic, chromium is widely distributed in Ontario ore bodies associated with gold. It commonly occurs as fuchsite (a chromian mica) or chromite (a chrome oxide), neither of which involves chromium in its hexavalent form. There is no evidence to suggest that the hexavalent state can be converted from the other forms within the human body (5). It remains to be determined if such conversion can take place in the gold mining environment. As with other agents discussed in this section, the possibility that chromium, perhaps in combination with other factors, can contribute to the increased risk of stomach cancer in Ontario gold miners cannot be dismissed.

In fact, chromium has been suspected as a cause of the excess of stomach cancer amongst Ontario gold miners (44). This conclusion was made on the basis that stomach cancer mortality was more closely associated with chromium exposure than with any other factors examined. Interestingly, a review of the mines which showed high chromium content revealed that most of them began operation only after 1938 and not before. If prolonged and excessive exposure to chromium were indeed a contributing factor in the development of stomach cancer, this would help to explain the finding that the highest excess mortality from stomach cancer is observed in younger miners who worked in Ontario gold mines after 1945.

ii) Polycyclic Aromatic Hydrocarbons (PAHs)

Certain PAHs such as benzo[a]pyrene are known to be potent human carcinogens associated with various types of cancer in the body (32). Some may act as precarcinogens which become activated in the body and lead to the formation of carcinogens that attack the stomach mucosal tissue and induce gastric carcinoma (87).

In the mining environment, PAHs are found in diesel emission, blasting smoke and in oil mist generated from the use of straight lubricating oils. These contaminants may be adsorbed onto airborne particulates which can then be inhaled. Some of these inhaled particulates are cleared from the lungs and are swallowed, allowing the PAHs to enter the digestive tract. In addition, PAHs can also be delivered to the digestive tract through hand-to-mouth contact as discussed later in section 5, iii.

a) Oil Mist

IARC has designated oil mists into two groups: untreated and mildly treated oils (Group 1), and highly refined oils (Group 3) (31). In underground mining, oil mist refers to airborne particulates generated from oil used to lubricate drills and other equipment. The predominant lubricant for drills and motorized equipment used in underground mining is of the first type.

The introduction of diesel equipment into the mining environment in the 1960s complicates the measuring of oil mists. They are hard to distinguish from oil in diesel emissions. Any dust sample could contain both diesel soot and oil mist (23). Nevertheless, a study of Finnish sulphide ore miners reported that the average airborne concentration of oil mist during drilling was as high as 3 mg/m3 in the 1970s, ranging from 0.1 to 17 mg/m3. The authors estimated that underground mining exposure to oil mist between 1950 and 1960 could have been double that of the 1970s (1).

The Ontario Ministry of Labour's bipartite Occupational Exposure Limits Task Force conducted a review recently of the current exposure limits for 101 substances and recommended lowering the current time-weighted average exposure value (TWAEV) for mineral oil mist from 5 mg/m3 to 1 mg/m3, and the current short term exposure value (STEV) from 10 mg/m3 to 3 mg/m3. Some of the other Task Force recommendations have been implemented, but those pertaining to mineral oil mist have not. The members eventually agreed to maintain the current guidelines due to the potential economic impact of reducing them (73,77).

The United Auto Workers (UAW) have petitioned the Occupational Safety and Health Administration (OSHA) in the US to consider a reduction of the time-weighted average limit for worker exposure to oil mist to 0.5 mg/m3 from the current 5mg/m3 (6). Unlike the auto manufacturing sector in North America, in which exposure levels as low as 0.2 mg/m3 are now achieved, the underground mines and particularly the more confined environment of many Ontario gold mines continue to be associated with much higher oil mist levels.

b) Diesel Emission

IARC has found that "Diesel engine exhaust is probably carcinogenic to humans"(Group 2A) (33). This emission is a complex mixture of particulates, gases and vapours that include PAHs and other potential carcinogens.

Diesel equipment was first introduced to Ontario underground mining in the early 1960's, and its use underground has been a continuous health concern among mine workers and management. As a result, diesel equipment was gradually replaced in some mines by electrical powered machinery during the late 1970's and 1980's. Today, many Ontario gold mines still use diesel equipment underground and are absorbing the higher cost of improved ventilation. It has been postulated that improved ventilation may not help reduce exposure as was once assumed, since PAHs remain in the mine air longer than particulate and are not removed completely with the ventilation (100).There are presently no Ontario standards for occupational exposure to diesel emission. This issue is currently before the Ontario Mining Legislative Review Committee which is recommending an exposure level of 1.5 mg/m3.

iii) Nitroso Compounds

Nitrate and nitrite are commonly found in the environment, foods (additives and preservatives) and in various occupational environments. As mentioned earlier, salts of nitrate and nitrite may be converted to nitrosamines and nitrosoureas by chemical reactions with amines in a low acid stomach environment. Many of these nitroso compounds have been shown to be carcinogens and a few can induce stomach cancers in experimental animals.

In underground mining, nitrate salts can be found in the airborne mixture of gases, vapours and particulates generated after blasting with explosives. These explosives were either nitrogen-glycerin based (before the 1950's) or a mixture of ammonium nitrate (94%) and fuel oil (6%) called ANFO used after the 1950's. In some mines, water gel explosives are also used.

The largest consumption of explosives is for development and production blasting. These operations were initially carried out in two stages during normal workshifts. In the first stage, the centre of the rock face is blasted and miners would often be excessively exposed to smoke, dust and gases from the first blast as they walked through this mixture to prepare the remainder of the rock face for secondary blasting. In gold mines, such a mixture could not be easily cleared owing to the confined work space and poor ventilation in the past.

Few miners are now exposed to such blasting since the introduction of central blasting, a practice by which the initial blast takes place between shifts or when workers are either on the surface or confirmed to be in safe, unaffected areas. With adequate ventilation, it takes about 3 to 4 hours for the concentration of gases to return to pre-blast levels. However, secondary blasting still occurs today and has not changed much over the years. This type of blasting (for making raises, muck spillage or clearing obstructions) could take place any time during a shift and often at close proximity to miners working underground.

iv) Conclusions:

5. What, if any, are the differences between gold mining and other kinds of hardrock mining in Ontario?

Further to the agents discussed above, the Panel also investigated other aspects of Ontario gold mining which may be related to stomach cancer. These focused on differences between gold mining and other types of hardrock mining in Ontario. In her report to the Panel, Stock indicated that:

"In general, the gold mines began operating considerably earlier than other mines (1910-1949) and closed down much earlier (1950-1970). Thus goldminers may have tended to be older than the other miners. The mines were known to have had extremely high dust levels prior to 1950 with a silicosis SMR of approximately 850, high rates of silicotuberculosis and measured dust levels in the 1930's and 40's often above 1000 particles/ml. Although dust levels in gold mines and non-gold mines may have been comparable at any given period, dust exposure levels were likely much higher among miners who worked prior to 1950 and thus the cumulative exposure levels were likely highest among goldminers..."

The Panel's investigation also revealed there was in general a notable difference in mining methods between gold and other hardrock mining. Namely, gold which usually occurs in small veins necessitates underground miners to often work in narrow workplaces and confined spaces. The methods used for mining gold are of the low-tonnage type and are less mechanized than bulk mining methods used to extract massive (e.g. nickel, copper and iron) or tabular ore bodies (uranium) (88). Such low-tonnage methods require work that is more physically demanding and close to the rock face, thereby resulting in a greater potential for worker exposure to airborne contaminants (e.g. oil mist, diesel emission and dust) than would be with other methods.

Ontario goldminers would also be more exposed to airborne crystalline silica and inorganic arsenic than other miners. The amount of crystalline silica in the respirable fraction of airborne dust in mining operations is directly related to its content in the ore and host rock being broken (26). Higher concentration of crystalline silica as quartz occurs in gold ore (15 to 35 per cent) than in nickel or iron ore (up to 10%) or uranium ores in the Bancroft area (5 to 15%). The uranium ores in the Elliot Lake area contain 60-70% crystalline silica. Since the average length of work for uranium miners is 1.5 years, compared to 10.8 in Ontario goldminers, cumulative silica exposure in these Elliot Lake workers would in general still be far less than the exposure experienced by Ontario goldminers.

Arsenic concentration in Ontario gold mines varies widely from undetectable levels to levels as high as 1.6% of the host rock (43). Arsenopyrite is also found combined with nickel ore but generally in lower concentration than in gold ore. Arsenic levels in Sudbury ore range from 0.096% to as low as 0.0004% (96).

Other notable differences between gold mining and other hardrock mining in Ontario include:

I) Backfill

Backfill refers to the process by which waste material is used to fill the void created by mining an ore body. This material can consist of gravel, waste rocks, cement, mill tailings or a combination of these mixed with sand. It is used mostly in 'cut-and fill' operations and was first introduced in Ontario gold mines in the 1940's.

Although backfill is not unique to gold mining and is also used in base metal mines (e.g. nickel) in Ontario, only the gold mines are known to have used liquid mill waste to backfill (82). Even after washing and other treatments, such mill waste could still contain significant amounts of harmful contaminants. For example, 2.41 mg/l and 12.8 mg/l of arsenic were found in recent effluent water samples from the backfill and tailing, respectively, of an Ontario gold mine (18). The Ontario effluent limit for total arsenic is 0.5 mg/l (76).

At first, backfill from the surface was brought to the desired location underground by dumping it down vertical chutes between levels. This process was extremely dusty since the backfill was dry going down the chutes. After the 1960's, hydraulic backfill was introduced which allowed this material to be transported in pipelines. This material contains 40 - 60% water and is transported in the form of a slurry; thus very little dust is generated.

ii) Aluminum Powder

Many Ontario goldminers were exposed to large quantities of aluminum powder as a prophylactic agent against the development of silicosis. This powder (a finely ground mixture of mostly aluminum oxide, metallic aluminum and iron oxide) was dispensed as an aerosol which miners were compelled by working procedures to inhale each day before their shift. In the lungs, this powder was believed to be capable of binding to and coating crystalline silica to prevent responses of the lungs that would otherwise elicit the cellular changes which can lead to silicosis.

In Ontario, the practice of dispensing aluminum powder to miners began in 1943 and was rapidly adopted by the mines. The greatest application in Ontario was in gold mines (37 gold mines by the end of 1944), although six uranium and one silver mine also adopted the practice by 1958 (80). There was no record of usage in nickel, copper or other Ontario mines. The recommended regimen based on effective dose was deep mouth breathing for 10 to 20 minutes at an airborne concentration of 35.6 mg/m3, which resulted in significant respiratory (estimated alveolar burden of 375 mg/year) as well as gastric exposure. Although there was limited evidence of harmfulness, a scientific task force concluded that further studies were required on the biological effects of this powder in humans. Since the effectiveness of aluminum powder against silicosis was itself doubtful and inadequate by scientific standard at the time, the practice was halted in 1979.

Aluminum does not appear to have an essential physiological role in humans and its toxicology is not well understood. This element is poorly absorbed in the body and is efficiently excreted. In the human stomach, aluminum oxide acts as a buffer against the acidity and forms ions which can react with other substances to form aluminum complexes. It is not known what other effects these complexes may have on the stomach and the remaining gastrointestinal tract.

Kusiak dismissed aluminum powder as a causal agent of stomach cancer in Ontario gold miners, on the basis that no apparent increase in this cancer was seen in uranium miners who were also exposed to the powder(44). However, the average duration of employment for uranium miners is 1.5 years while that for goldminers is 10.8years. Furthermore, in Ontario the use of aluminum powder in gold mines commenced 15 years earlier and halted later than in uranium mines. Therefore, this conclusion may be inappropriate since the exposure to aluminum oxide in uranium miners could have been insufficient to permit this substance to elicit an effect similar to that observed in Ontario goldminers.

iii) Underground Working Environment

While gradual efforts have been made to lessen mining hazards, other unhealthy working conditions long tolerated by underground goldminers have changed very little. For example, underground washroom facilities and potable water were only recently made available to Ontario gold mines and, in general, later than in other metal mines. For example, Ontario uranium mines, most of which opened in the 1950's, made provision for potable water and washing facilities underground as part of their initial operational plans. These facilities soon became available in Ontario nickel mines but long before they were introduced into the gold mines.

In a mine that has been extensively excavated, underground refuge stations and washroom facilities are often far away (up to a mile) from the work areas and, because of the lengthy travelling time needed, are not always used. Instead, it is customary for underground miners to rest and have meals where they work, where the level of dust and other airborne contaminants may be considerable. A recent tour of Ontario gold mines also showed that these miners are routinely covered in dirt and grease from work and their hands are often dirty while they have their meals. The likely ingestion of particulates from food stuff in these conditions (poor personal hygiene, dusty environment and few washing facilities) may affect the development of stomach cancer.

The lack of washing facilities may be significant since it has been postulated that the normal mucosal barrier may be overcome by abrasives or irritants (16). Peters noted that the highest risk of stomach cancer is observed to occur in the antrum-pylorus area of the stomach where the tissue lining is regularly under muscle pressure (78). The forcing of fluid containing particulates through this area could create a direct mechanical effect. He speculated that increasing dietary cleanliness may have contributed to the downward trend in stomach cancer in industrialized countries. In goldminers, irritation of the stomach lining could be attributed to ingestion of particulates such as aluminum powder, dust in general and contaminants in foods.

Before the 1970's, potable water was not available underground except in containers which underground miners carried with them to their work areas. Even now, it is only available in refuge stations which are few and often far from areas where miners are working.

Miners were also known to drink water intended for mining processes such as drilling or dust suppression - water that was not considered potable.

Samples of process water from a few Ontario gold mines were recently collected by the Ontario Ministry of Labour (MOL) and analyzed by the Ontario Ministry of Environment (MOE). Most of the parameters examined (e.g. PAHs, mercury, cyanide and metals) were well within standards for Ontario drinking water, while others such as nitrite/nitrate and arsenic exceeded their respective standards in a few of the samples. It should be noted that these results do not reflect an on-going monitoring process nor water quality in the past.

iv) Conclusions:

6. Are there other risk factors that could affect the pattern of stomach cancer?

i) Diet

Diet appears to play an important role in the development of stomach cancer. For example, studies have shown that increased risks of stomach cancer may be associated with a high intake of smoked, salted, fried or starchy foods (28,72,49,58,55), as well as exposure to high levels of nitrite and nitrate in foods and drinking water (3). It is hypothesized that excess salt in the diet may irritate the stomach lining, while nitrite and nitrate in the stomach can form N-nitroso compounds, many of which are carcinogenic in animals and a few are human stomach carcinogens (see nitroso compounds under 2B.3).

On the other hand, fruits, vegetables and specific substances such as vitamins A, C and E found in these foods appear to have a protective effect against stomach cancer (52,58,55) and other cancers in the gastrointestinal tract (46). It is suggested that some of these substances may act by interfering with the formation or reaction of chemical carcinogens or pre-carcinogens such as N-nitroso compounds.

While a deficiency in dietary vegetables and fruits could certainly be a contributing factor in the increased stomach cancer rates of the general Northern Ontario population, this does not explain why elevated stomach cancer mortality is specifically confined to goldminers. It would be difficult to argue that the diet of goldminers in rural Northern communities differed substantially from that of other Northern mining groups, none of whom showed any excess mortality from stomach cancer.

ii) Cigarette smoking and alcohol consumption

In coal miners, light smokers (one to ten cigarettes a day) were found to have significantly higher stomach cancer rates than non-smokers (56). This led some to postulate that light smoking may stimulate the lungs' clearance process (in contrast to heavy smoking which would impair the process) thereby introducing more carcinogens into the stomach (56,24).

A recent population based case control study analyzed lifetime tobacco use and occupational histories (9). Subjects who had ever smoked cigarettes had a statistically significant increase in stomach cancer (OR=1.5, CI=1.2 to 1.8) and the risk increased with increasing pack years. The risk was highest among black men and white women and not significantly increased in white men (OR=1.2, CI=.9-1.6). Increased risk among the white men was associated with the occupations of farmer, mechanics, assemblers, or machine repairmen. The authors speculate about the potential role of diesel exhaust and/or asbestos.

However, Kusiak et al., 1991 (43) found no material difference in smoking habits between gold and nickel/copper miners. Therefore, cigarette smoking alone cannot explain the elevated risk of stomach cancer in Ontario goldminers. In fact, the evidence regarding cardiovascular disease among miners in general, does not show the typical pattern of increased risk characteristic of heavily smoking populations.

Results of studies of alcohol and cancers of the stomach have been inconsistent, but generally not supportive of a causal association (46). Ontario miners' mortality from cirrhosis of the liver shows a statistically significant deficit (SMR=77, CI=66-88) from this cause and suggests that alcohol consumption in Ontario miners is less than that of other men in Ontario (45). A deficit of morbidity from cirrhosis of the liver was also confirmed by Muir and Julian in a study of cancer morbidity among nickel/copper workers (60).

iii) Ethnicity and place of birth

World-wide, the rate of stomach cancer varies as much as ten-fold, with some of the highest rates occurring in eastern Europe and Japan (98). The rate is found to decrease slowly in populations moving from a high-risk to a low-risk country; thus suggesting a strong environmental component in the cause of stomach cancer.

Place of birth, which is sometimes used as a surrogate for ethnicity, may also be an indicator of elevated stomach cancer risks. Newman and Spengler showed that in comparison to Ontario men born in Canada, there are excesses of stomach cancer in Ontario residents born in Great Britain(SMR 159, CI 145-174), Italy (SMR 189, CI 155-228), The Netherlands(SMR 228, CICI 159-309), Poland (SMR 183, CI 151-222) and the Soviet Union(SMR 161, CI 131-196) (72). As Stock points out however, "the role of ethnicity is confusing....the risk of stomach cancer may be higher in immigrants who may have lower socioeconomic status, poorer diets and/or use coal for cooking and heating rather than due to ethnicity per se" (87).

However, several studies on the Ontario gold mining population have reported higher rates of stomach cancer among those born in Canada versus foreign born (63,83,44). In particular, Kusiak and colleagues report a statistically significant excess in both groups. When these groupings are further analyzed by age however, an interesting pattern becomes apparent. The excess of stomach cancer among foreign born miners over age 60 is similar to that seen throughout Ontario among those of similar ethnic background. The excess among all gold miners under age 60 is elevated regardless of place of birth.

iv) Conclusions






  • Goldmining is a dusty occupation;
  • Goldminers are likely occupationally exposed to many known and suspected human carcinogens. These include arsenic, asbestos, silica, chromium, PAHs, oil mist, diesel emission). Of these, asbestos and hexavalent chromium are known to be potential gastrointestinal carcinogens.
  • Goldminers could also be exposed to nitroso compounds or precursors of such compounds. There is experimental evidence that some of these nitroso compounds are associated with stomach cancers.
  • Ingestion is a likely route of entry for any carcinogens to which Ontario goldminers may be exposed.



    CHAPTER FOUR. THE BRADFORD HILL CRITERIA FOR A CAUSAL ASSOCIATION - The Panel's Analysis of the Connection Between Stomach Cancer and Gold Mining in Ontario.

    Sir Austin Bradford Hill articulated nine factors to be weighed when considering whether an environmental factor has caused disease (29). These criteria are now widely accepted. As in previous reports, the Panel's analysis is based upon these considerations.

    1. Strength of association: The degree of increase in the risk of disease after exposure to a substance or process (for example, a high SMR). A statistically significant increase is usually required to confirm an excess risk of disease; however, a strong association may be relevant even if statistical significance has not been shown, particularly when the failure to show statistical significance is because a small population is being studied.

    All of the studies discussed have been retrospective cohort studies, likely to provide the strongest evidence feasible.

    All but one analysis of the mortality experience of Ontario gold miners have shown statistically significant increases in stomach cancer mortality. The overall SMRs range from 136 to 157. The one exception was the study of Shannon, 1987 which compared the mortality from stomach cancer among gold miners to that in the Northern Ontario male population as well as to that in the provincial male population. The SMR decreased from a statistically significant 142 with the Ontario male population as the control group to a non-statistically significant SMR of 114 when the Northern male population was used as the control group. When Muller and colleagues did a similar comparison with the Ontario male population as controls (SMR=157), and then compared stomach cancer mortality among gold miners using nickel/copper miners as the population control, the SMR increased from 157 to 248, giving strong evidence, most recently confirmed by Tomlinson, 1995 (89) that the higher rates seen among gold miners are not merely a reflection of the higher rates seen among Northern Ontario males generally.

    Kusiak et al.,1993 found the highest reported SMR of 463 among gold miners under age 60 for the time-weighted exposure to chromium (44).

    Extensions to the period of follow-up have strengthened the observation of excess stomach cancer risk.

    2. Consistency: Consistency is shown when several different studies produce similar findings. Consistency is particularly persuasive if the studies are of various designs, and are large and carefully conducted.

    Regardless of changing inclusion criteria and length of follow-up, the Ontario cohort studies of stomach cancer in gold miners are consistent. Consistency in increased SMRs is also shown between individual gold mining camps (61).

    Furthermore, confirmatory evidence is provided by the studies of Finklestein (20) among silicotic gold miners, and by Muller et al., (61,62,65) who confirmed that apparent excesses of stomach cancer among mixed ore or uranium miners were in fact due to their prior gold mining experience.

    Two of the three non-Ontario studies show elevated results. One study reports a negative finding. All are retrospective mortality studies of various size and quality.

    3. Specificity: Specificity is shown when exposure to a particular substance is associated with one particular disease. Specificity must be weighed with caution because it does not apply if a certain substance is capable of causing more than one disease, such as asbestos or lead. Specificity must be weighed together with all the other evidence.

    4. Temporality: Temporality means that the exposure took place before the disease occurred.

    Exposure pre-dates disease onset. Studies of Ontario gold miners indicate that the excess for stomach cancer appeared five to 19 years after beginning gold mining in Ontario (44) and were associated with higher risk than latencies of greater than 20 years.

    5. Biological Gradient (Dose-Response): A dose-response trend is shown when an increase in the exposure ("dose") corresponds with an increase in the rate of disease or death ("response"). Conversely, a dose-response trend will also occur if the incidence of disease decreases with a corresponding decrease in the exposure.

    The greatest weakness in the Ontario studies has been the apparent failure to demonstrate a dose-response effect, with the exception of that found by Kusiak et al.,1993 who demonstrated a statistically significant association between the excess of stomach cancer in miners under the age of 60 and the time-weighted index of exposure to chromium (44).

    In fact, an inverse relationship of length of employment and onset of the disease has been found (83). However, there is now an explanation for this apparent absence of a dose-response. The results of the regression analysis included as Appendix B of this report show:

    . After controlling for employment status and age at hire, duration of employment is more strongly associated with stomach cancer at younger ages than at higher ages.

    . As age at first gold mining increases, risk for stomach cancer increases.

    . In a statistical model that includes age at risk, age at hire, duration of employment and employment status, there is no effect of calendar time, year of hire, or being born outside of North America.

    As discussed previously on page 36 of this report, years of gold mining and attained age are not independent risk factors. The statistically significant and declining effect of age on the relative risk of stomach cancer among the Ontario gold miners means that observations on the whole cohort, instead of just on those at highest risk, i.e. the youngest, tend to obscure patterns of risk. The effect of age had originally been noted by Muller et al., in their 1987 Addendum, but not followed-up with additional analyses at the time.

    Other, less important factors, which could affect a dose-response pattern include:

    6. Biological plausibility: An association is biologically plausible when the suspected connection between the exposure and the disease is consistent with what we already know about biological and chemical patterns.

    The stomach is one of the organs that can have significant exposure to occupational hazards through inhalation and subsequent swallowing of material that has either been coughed up, enters the mouth and nose, or is ingested as contaminants on workers hands when they eat. Workers in dusty conditions are known to swallow varying amounts of dust, especially when performing strenuous work resulting in mouth breathing.

    As consultant Dr. Stock notes:

    " The epidemiological studies of occupational stomach cancer have not been designed in such a way as to be able to implicate any specific occupational hazardous substance. The studies are consistent with theories previously proposed in the literature by Ames including poly-aromatic hydrocarbons; nitrosamines; and freshly generated dust. Chromium exposure remains a possible explanation of excess gastric cancer in Ontario gold miners but alternative occupational exposures are also possible explanations...."

    As summarized by Dr. Stock in her review, the various theories of occupational carcinogenesis and the consistent demonstration of associations found between workers in occupations exposed to dust, poly-aromatic hydrocarbons and/or nitroso compound precursors in working populations in different countries make biologic as well as epidemiologic sense.

    7. Coherence: Coherence is shown when the evidence as a whole makes sense and is not contradicted by what we already know about the disease process. This may simply be another way of describing biological plausibility.

    There is no conflict between what is known about the natural history of stomach cancer and a possible occupational connection. Early bacterial conditions, ulcers, susceptibility and even diet do not rule out a possible occupational contribution. The decreasing incidence in the general population could be associated with cleaner working conditions in general, however there is no indication of a decreasing incidence in the susceptible gold mining population in Ontario.

    8. Experiment: Experimental evidence is also relevant, if it is available. For ethical reasons, scientists cannot conduct experiments on human beings, so such evidence is rarely available; however, often animal studies provide evidence about the effects of exposures.

    Animal and in vitro experiments have linked some of the substances present in the gold mining environment with stomach cancer. These substances include PAHs, nitrite, and machining oils.

    9. Analogy: An analogy is shown if the same exposure in another setting is also associated with an increased risk of the same disease.

    Stomach cancer excess is found repeatedly in such dust-exposed workers as coal miners, construction, granite and foundry workers. It is also found in workers exposed to machining fluids, such as bearing plant workers. Both agents are found in the gold mining environment. In addition, Muller et al. studied different gold mining camps in different areas of Ontario and found an excess in all of them.



    The existing international literature supports the concept that

    stomach cancer can indeed be work-related. Although a specific substance in gold mining has not been identified as the causal agent for stomach cancer, there are several theories concerning dust, PAHs, and nitrosation reactions that render an association biologically plausible.

    The peculiar excess of stomach cancer among Ontario gold miners

    has historically been accepted as real, although not readily explainable.

    The most complete mortality data available, that of Kusiak et al., 1993, reconfirms earlier findings of a statistically significant excess of stomach cancer among gold miners showing an overall SMR of 152.

    The repeated finding of an excess of stomach cancer in gold miners, especially those under age 60, who seem to have a propensity to develop a type of stomach cancer (intestinal) that is more common over age 60, but which is more often associated with environmental factors, provides strong evidence for an occupational cause.

    Gold miners show excess stomach cancer mortality more than two and one half times that of nickel/copper miners, arguably the best possible control group. The unusual statistically significant relationship of decreasing relative risk for stomach cancer with age distinguishes gold miners from the male population in Northern Ontario, and indeed from the mortality experience from stomach cancer in all Ontario, and speaks of a unique set of circumstances.

    The Panel is convinced that the only plausible explanation for this excess is that work in gold mining contributed significantly to the excess seen in these workers.


    The Panel finds a probable connection between work in Ontario gold mining and the occurrence of primary cancer of the stomach.

    As outlined in the Panel's occasional paper entitled "Decision-Making: Theory and Practice" (38) p.66-68, the Panel weighs the sum total of the evidence in deciding whether to recommend entry into Schedule 3, 4, or adjudication by guidelines on a case-by-case basis. Diseases should be entered into Schedule 3 when there is a "strong connection" between work and disease. The Panel is of the unanimous opinion that the stomach cancer found among gold miners in Ontario is work-related. However, it recognizes the possibility that some unidentified, but unique feature of Ontario gold mining characterises this work-association.

    Members of the Panel discussed at length the relative merits of entry into Schedule 3 versus specific guidelines but were unable to reach a consensus. A majority of the Panel is of the opinion that the relationship between work and disease is strong enough in this instance to support the following:


    Primary stomach cancer and the associated process, trade or occupation of gold mining should be added to Schedule 3 of the Act.

    Although stomach cancer can and does occur in the non-occupationally exposed population, the peculiar characteristics of the excess among Ontario gold miners should facilitate the development of fair and equitable grounds for rebutting the presumption with due consideration being given regarding the age of the claimant. The Panel would be willing, if requested, to assist the Workers' Compensation Board in their development. This recommendation was endorsed by Panel members

    J.Brophy, C.Buck, N.Carlan, H.Seguin, and M.Wills.


    Although Panel members are unanimous in their opinion that there is a probable connection between stomach cancer and work in Ontario gold mines, there is a difference of opinion as to how the finding should be implemented. Panel members W.Elliott and J.Macnamara recommend that guidelines should be used to assist adjudicators in assessing the merits of claims for stomach cancer from gold miners. This recommendation is based on their consideration of the facts before the Panel, primarily the peculiar dose response pattern and the lack of an identified causal agent.


    1. Ahlman,K.; Koskela,R-S.; et al. Mortality among sulfide ore miners. American Journal of Industrial Medicine. Vol.19(1991). p.603-17.

    2. Ames,R.G.; Gamble,J.F. Lung cancer, stomach cancer, and smoking status among coal miners. A preliminary test of a hypothesis. Scandinavian Journal of Work, Environment and Health. Vol.9(1983). p.443-448.

    3. Armijo,R.; Gonzales,A.; et al. Epidemiology of gastric cancer in Chile: II - nitrate exposures and stomach cancer frequency. International Journal of Epidemiology. Vol.10,no.1(1981). p.57-62.

    4. Battista, G.; Belli, S.; et al. Mortality among pyrite miners with low-level exposure to radon daughters. Scandinavian Journal of Work, Environment and Health. Vol.14(1988). p.280-85.

    5. Beliles R.P. Metals. In: Casarett L.J. and Doull J.,eds. Toxicology: The Basic Science of Poisons. New York: MacMillan Publishing Co. Inc., 1975. p.454-502.

    6. Bieber,O. [Letter to Hon. Joe Dear, Assistant Secretary of Labor for OSHA, requesting a change in the PEL limit for machining fluids]. December 9, 1993.

    7. Bridbord K.; Costello,J.; et al. Occupational safety and health implications of increased coal utilization. Environmental Health Perspectives. Vol.33(1979). p.285-302.

    8. Brown, D.P.; Kaplan, S.D.; et al. Retrospective cohort mortality study of underground gold mine workers. In: Proceedings of the Third NCI/EPA/NIOSH Collaborative Workshop: Progress on Joint Environmental and Occupational Cancer Studies. Cincinatti: National Institute of Occupational Health and Safety, 1984. p.7-55.

    9. Burns,P.B.; Swanson,M.G. Stomach cancer risk among black and white men and women: the role of occupation and cigarette smoking. Journal of Occupational and Environmental Medicine. Vol.37,no.10(October,1995). p.1218-23.

    10. Carnow, B. and Conibear, S.A. Health survey of arsenic-exposed Dickenson Mines employees. Unpublished Report. Chicago, 1978.

    11. Carta, P.; Cocco, P.; Picchiri, G. Lung cancer mortality and airways obstruction among metal miners exposed to silica and low levels of radon daughters. American Journal of Industrial Medicine. Vol.25(1994). p.489-506.

    12. Chau, N.; Benamghar, L.; et al. Mortality of iron miners in Lorraine (France): relations between lung function and respiratory symptoms and subsequent mortality. British Journal of Industrial Medicine. Vol.50(1993). p.1017-31.

    13. Chen, R.; Hu, M. A cohort study of cancer mortality in copper miners. In: Sakurai. H. et al., eds. Occupational Epidemiology. China: Elsevier Science Publishers B.V. (Biomedical Division), 1990. p. 75-78.

    14. Chen, S.Y.; Hayes, R.B.; et al. Mortality experience of haematite mine workers in China. British Journal of Industrial Medicine. Vol.47(1990). p.175-81.

    15. Chen, R.; Wei, L.; Huang, H. Mortality from lung cancer among copper miners. British Journal of Industrial Medicine. Vol.50(1993). p.505-9.

    16. Correa,P.; Haenszel,W.; et al. A model for gastric cancer epidemiology. The Lancet. Vol.2,no.7924 (July 12, 1975). p.58-60.

    17. Costa, D.; Guignard, J.; et al. Production of free-radicals arising from the surface activity of minerals and oxygen. Part I: Iron mine ores. Toxicology and Industrial Health. Vol.5,no.6(Dec. 1989). p.1061-78.

    18. Dickenson Mine, Red Lake. Effluent sampling data. 1993.

    19. Enterline, P. A review of mortality data for American coal miners. Annals of the New York Academy of Sciences. Vol.200(1972). p.260-77.

    20. Finkelstein, M.; Liss, G.M.; et al. Mortality among workers receiving compensation awards for silicosis in Ontario 1940-85. British Journal of Industrial Medicine. Vol.44(1987). p.588-94.

    21. Fletcher, A.C. The mortality of foundry workers in the United Kingdom. In: Goldsmith, D.F., ed. Silica, Silicosis and Cancer: Controversy in Occupational Medicine. New York: Praeger, 1986. p.385-401.

    22. Fox, A.J.; Goldblatt, P.; Kinlen, L.J. A study of the mortality of Cornish tin miners. British Journal of Industrial Medicine. Vol.38(1981). p.378-80.

    23. French,Ian W. & Associates. Health implications of exposure of underground mine workers to diesel exhaust emissions: an update. Markham: Ian W. French and Associates Ltd., April 1984.

    24. Gamble, J. F.; Ames, R.G. The role of the lung in stomach carcinogenisis: a revision of the Meyer hypothesis. Medical Hypotheses. Vol.11(1983). p.359-64.

    25. Gillam, J.D.; et al. Mortality patterns among hard rock gold miners exposed to an asbestiform mineral. Annals of the New York Academy of Sciences. Vol.271(1976). p.336-44.

    26. Ham, J. Report of the Royal Commission on the Health and Safety of Workers in Mines. Toronto: Ministry of the Attorney General, 1976.

    27. Harrison, T.; Braunwald, E.; et al. Harrison's Principles of Internal Medicine. 11th ed. 2 Vols. New York: McGraw-Hill Book Co., 1987.

    28. Higginson,J.; Muir,C.; Munoz,N. Stomach. In: Higginson,J.,Muir,C.,Munoz,N.,eds. Human Cancer: Epidemiology and Environmental Causes. Cambridge: Cambridge University Press, 1992. p.273-82.

    29. Hill, A.B. The environment and disease: association or causation? Proceedings of the Royal Academy of Medicine. Vol.58(1965). p.295-300.

    30. Hodgeson, J.T.; Jones, R.D. Mortality of a cohort of tin miners 1941-86. British Journal of Industrial Medicine. Vol.47(1990). p.665-76.

    31. International Agency for Research on Cancer (IARC). Polynuclear aromatic compounds, Part2, Carbon blacks, mineral oils and some nitroarenes. (Vol 33). Lyon: International Agency for Research on Cancer, 1984.

    32. International Agency for Research on Cancer (IARC). Overall evaluations of carcinogenicity: an updating of IARC monographs Volumes 1 to 42. (Supplement 7). Lyon: International Agency for Research on Cancer, 1987.

    33. International Agency for Research on Cancer (IARC). Diesel and gasoline engine exhaust and some nitroarenes. (Vol.46). Lyon: International Agency for Research on Cancer, 1989.

    34. International Agency for Research on Cancer (IARC). Chromium, nickel and welding. (Vol. 49). Lyon: International Agency for Research on Cancer, 1990.

    35. Industrial Disease Standards Panel (IDSP). Report to the Workers' Compensation Board on the Healthy Worker Effect. (IDSP Report No. 3). Toronto: Industrial Disease Standards Panel, July 1988.

    36. Industrial Disease Standards Panel (IDSP). Report to the Workers' Compensation Board on the Ontario Gold Mining Industry. (IDSP Report No. 1). Toronto: Industrial Disease Standards Panel, April, 1987.

    37. Industrial Disease Standards Panel (IDSP). Second Report to the Workers' Compensation Board on Certain Issues Arising from the Report of the Royal Commission on Asbestos. (IDSP Report No.7). Toronto: Industrial Disease Standards Panel, April 1990.

    38. Industrial Disease Standards Panel (IDSP). Decision Making: Theory and Practice. (IDSP Occasional Paper). Toronto: Industrial Disease Standards Panel, December 1994.

    39. Isler, M; Caron, M. [A review of the paper presented by S.R. Stock, for IDSP]. December, 1993.

    40. Kabir, H. and Biligi, C. Ontario gold miners with lung cancer -occupational exposure assessment in establishing work-relatedness. Journal of Occupational Medicine. Vol.35,no.12(1993). p.1203-7.

    41. Katsnelson, B.A.; Mokronosova, K.A. Non-fibrous mineral dusts and malignant tumors. Journal of Occupational Medicine. Vol.21,no.1(January, 1979). p.15-20.

    42. Kinlen, L.J.; Willows, A.N. Decline in the lung cancer hazard: a prospective study of the mortality of iron ore miners in Cumbria. British Journal of Industrial Medicine. Vol.45(1988). p.219-24.

    43. Kusiak, R.A.; Springer, J.; et al. Carcinoma of the lung in Ontario gold miners: possible etiologic factors. British Journal of Industrial Medicine. Vol.48(1991). p.808-17.

    44. Kusiak, R.A.; Ritchie, A.C.; et al. Mortality from stomach cancer in Ontario miners. British Journal of Industrial Medicine. Vol.50(1993). p.117-26.

    45. Kusiak,R.A.; Ritchie,A.C.; et al. Mortality from lung cancer in Ontario uranium miners. British Journal of Industrial Medicine. Vol.50, no.10(October, 1993). p.920-28.

    46. Thomas,D.B. Cancer. In: Last, J.M., ed. Maxcy-Rosenau Public Health and Preventive Medicine. 12 ed. Norwalk, Connecticut: Appleton-Century-Crofts, 1986. p.1133-58.

    47. Lauren, P. The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma; an attempt at histo-clinical classification. Acta Pathologie et microbiology, Scandinavia. Vol 64(1965). p.31-49.

    48. Lawler, A.B.; Mandel, J.S.; et al. A retrospective cohort mortality study of iron ore (hematite) miners in Minnesota. Journal of Occupational Medicine. Vol.27,no.7(July, 1985). p.507-17.

    49. Lee, J-K.; Park, B-J.; et al. Dietary factors and stomach cancer: a case-control study in Korea. International Journal of Epidemiology. Vol. 24,no.1(1995). p.33-41.

    50. Luce, C. [Letter to C. Archer]. June, 1996.

    51. Matalo, N.M.; Klauber,M.R.; Gorishek,W.M. High incidence of gastric carcinoma in a coal mining region. Cancer. Vol.29,no.3(1972). p.733-37.

    52. Mayer, R.J. Gastrointestinal cancer. In: Rubenstein,E.; Federman,D.D.,eds. Scientific American Medicine Series. Vol.3. Chapter 12 (Oncology), Section VIII. New York: Scientific American Inc., 1986.

    53. McDonald,H.B. [Letter to J. Ham requesting further information on lung and stomach cancer in gold miners for WCB]. August 18, 1986.

    54. McDowall, M.E. A mortality study of cement workers. British Journal of Industrial Medicine. Vol.41 (1984). p.179-82.

    55. Mettlin, C. Epidemiologic studies in gastric adenocarcinoma. In: Douglass,H.O.,ed. Gastric Cancer. New York: Churchill Livingstone, 1988.

    56. Meyer, M. B.; Luk, G.D.; et al. Hypothesis: the role of the lung in stomach carcinogenesis. American Review of Respiratory Disease. Vol.121(1979). p.887-92.

    57. Miller, A. Appendix C. Report of the Scientific Panel on "Mortality from Cancer among Ontario Gold Miners 1955-1977". In: Report to the Workers' Compensation Board on the Ontario Gold Mining Industry. (IDSP Report No.1). Toronto: Industrial Disease Standards Panel, April 1987. p.64-86.

    58. Miller, A.B. Planning cancer control strategies. Chronic Diseases in Canada. Vol.13, no.1,Suppl. (Jan.-Feb.1992).

    59. Morrison, H.I.; Semenciw, R.M.; et al. Cancer mortality among a group of fluorspar miners exposed to radon progeny. American Journal of Epidemiology. Vol.128,no.6(1988). p.1266-75.

    60. Muir,D.M.; Julian,J.A. A Study of Cancer Incidence in Ontario Nickel Workers. Report prepared for the Occupational Disease Panel. January 15, 1996.

    61. Muller, J.; Wheeler, W.C.; et al. Study of Mortality of Ontario Gold Miners, 1955-1977. Part I. Toronto: Ontario Ministry of Labour, May 1983.

    62. Muller, J.; Wheeler, W.C.; et al. Study of mortality of Ontario miners. In: Stocker,H.,ed. In: Occupational Radiation Safety in Mining. Toronto: Canadian Nuclear Association, 1985. p.335-43.

    63. Muller, J.; Kusiak, R.A.; et al. Study of mortality of Ontario gold miners, 1955-1977. Part II. Toronto: Ontario Ministry of Labour, July 1986.

    64. Muller, J.; Kusiak, R.A.; et al. Study of mortality of Ontario gold miners, 1955-1977. Part II, Addendum. Toronto: Ontario Ministry of Labour, February 1987.

    65. Muller,J.; Kusiak,R.; Ritchie,A.C. Factors modifying lung cancer risk in Ontario uranium miners, 1955-1981. Toronto: Ontario Ministry of Labour, July 1989.

    66. Mur, J-M.; Meyer-Bisch, C.; et al. Risk of lung cancer among iron ore miners: a proportional mortality study of 1,075 deceased miners in Lorraine, France. Journal of Occupational Medicine. Vol.29,no.9(September, 1987). p.762-68.

    67. Nair, R.C.; Abbatt, J.D.; et al. Mortality experience among workers in the uranium industry. In: Stocker,H.,ed. In: Occupational Radiation Safety in Mining. Toronto: Canadian Nuclear Association, 1985. p.354-64.

    68. National Cancer Institute of Canada (NCIC). Canadian Cancer Statistics 1995. Toronto: National Cancer Institute of Canada, 1995. Tables 4,9,10.

    69. Neuberger, M.; Kundi, M.; et al. Cancer mortality of dust workers and controls--results of a prospective study. In: Prevention of Occupational Cancer. (Occupational Safety and Health Series 46). Geneva: International Labour Office, 1982. p.235-41.

    70. Neuberger, M.; Kundi, M.; et al. The Viennese dusty worker study. In: Goldsmith, D.F.; Winn, D.M.; Shy, C.M.; eds. Silica, Silicosis, and Cancer. New York: Praeger, 1986. p.415-22.

    71. Neuberger, M. Gastric cancer and occupation. [A review of the paper presented by S.R. Stock, for IDSP]. December, 1993.

    72. Newman, A.M.; Spengler, R.F. Cancer mortality Among immigrant populations in Ontario, 1969-73. Canadian Medical Association Journal. Vol.130(1984). p.399-405.

    73. Office of the Minister, Ontario Ministry of Labour. Notice of proposed changes to occupational exposure limits of 101 substances. August 4, 1992.

    74. Ontario Cancer Treatment and Research Foundation (OCTRF). North Counties and Ontario Cancer Death Frequencies. Toronto: Ontario Cancer and Treatment Research Foundation, October 23, 1992.

    75. Ontario Cancer Treatment and Research Foundation (OCTRF). Cancer Incidence in Ontario. Toronto: Ontario Cancer Treatment and Research Foundation, 1995.

    76. Ontario Ministry of the Environment. Best Available Pollution Control Technology. [Report prepared by Kilborn,Inc. and The Environmental Applications Group Ltd.]. Toronto: Ministry of the Environment, December, 1991.

    77. Ontario Ministry of Labour. Notice to retain the current Ontario occupational exposure limits of seven substances. November 24, 1994.

    78. Peters, J.M. [Letter to S. Meuer providing comments of the literature review of S.R. Stock]. December 28, 1993.

    79. Pham, Q.T.; Chau, N. et al. Prospective mortality study among iron miners. Cancer Detection and Prevention. Vol.15,no.6(1991). p.449-54.

    80. Rifat, S.L. Evidence Regarding Effects of Exposure to Mcintyre Powder. PhD Thesis submitted to University of Toronto. 1990.

    81. Seguin, H. WCB for Lung and Stomach Cancer of Ontario Gold and Mixed Ore Miners. [Brief presented to the IDSP by the United Steelworkers of America]. November 14, 1986.

    82. Seguin, H. [Personal communication to J. Tung]. January,1996.

    83. Shannon, H. Appendix A. Report of the Special Panel on Ontario Gold Mining Industry. In: Report to the Workers' Compensation Board on the Ontario Gold Mining Industry. (IDSP Report No.1). Toronto: Industrial Disease Standards Panel, April 1987. p.12-55.

    84. Simonato, L.; Moulin, J.J.; et al. A retrospective mortality study of workers exposed to arsenic in a gold mine and refinery in France. American Journal of Industrial Medicine. Vol.25(1994). p.625-33.

    85. Spengler, R.F.; Price, J. Male cancer mortality by occupations, Ontario 1973-79. (Cancer Epidemiology Series, no.1). Toronto: The Ontario Cancer Treatment and Research Foundation (OCTRF), March 3, 1982.

    86. Stemmerman, G.N.; Hayashi,T. The pathology of gastric cancer. In: Douglass, H.O., ed. Gastric Cancer. New York: Churchill Livingstone, 1988.

    87. Stock, S.R. Gastric Cancer and Occupation: a Review of the Literature. [Unpublished Report for the Industrial Disease Standards Panel]. Revised, September, 1994

    88. Northern Miner. Mining Explained: A Guide to Prospecting and Mining. Toronto: Northern Miner Press, 1990.

    89. Tomlinson,G. Analyses of Stomach Cancer Data from IDSP Files. [Memorandum to C. Archer]. June, 1995.

    90. Tomlinson, G. Further analysis of histologic types of stomach cancer among gold miners. [Memorandum to C. Archer]. 1995.

    91. Tomlinson, G. Effect of lack of SINs on SMRs for stomach cancer in different eras. [Memorandum to C. Archer]. January, 1996.

    92. Tomlinson, G. Regression models for gold miners' stomach cancer data. [Memorandum to C. Archer]. May, 1996.

    93. Tomlinson, G.; Corey,P. Regression models for gold miners' stomach cancer data. [Memorandum to C. Archer]. June, 1996.

    94. Tung, J. Report to the Occupational Disease Panel on Visits to Northern Ontario Gold Mines. December 8, 1994.

    95. United Steelworkers of America (USWA). Submission to the Workers' Compensation Board on the Findings of the Industrial Disease Standards Panel on Gold Miners and the Gold Mining Industry. August 7, 1987.

    96. Warner,S. Estimated nickel exposures of INCO's Sudbury miners. INCO Limited. December, 1986.

    97. Waxweiller, R.J.; Roscoe, R.J.; et al. Mortality follow-up through 1977 of the white underground uranium miners cohort examined by the United States Public Health Service. In: Gomez, M. ed. Radiation Hazards in Mining: Control, Measurement and Medical Aspects. New York: Society of Mining Engineers of American Institution of Mining, Metallurgical and Petroleum Engineers Inc., 1981. p.823-30.

    98. Weinberg, G.B.; Kuller, L.H.; Stehr, P.A. A case-control study of stomach cancer in a coal mining region of Pennsylvania. Cancer. Vol.25(1985). p.703-13.

    99. Weisberger,J.H. Chemical carcinogenesis. In: Casarett L.J. and Doull J., eds. Toxicology: The Basic Science of Poisons. New York: MacMillan Publishing Co. Inc., 1975. p.333-78.

    100. Westaway, K.C. Polynuclear aromatic hydrocarbon and mutagens in the mine environment. In: Myers,D.K.;Osborn,R.V.;eds. Workshop on Research Needs in Toxicology - Chalk River. November 3-4, 1987. Chalk River: Atomic Energy of Canada Ltd., 1988.

    101. Wilson, J. Memorandum. November 24, 1987.

    102. Wingren, G.; Axelson, O. Mortality pattern in a glass producing area in SE Sweden. British Journal of Industrial Medicine. Vol.42(1985). p.411-14.

    103. Workers' Compensation Act, R.S.O. 1990, c.W.11.

    104. World Health Organization. Asbestos and other natural mineral fibres. (WHO Environmental Health Criteria no.53). Geneva: World Health Organization, 1986.

    105. Wright, W.E.; Bernstein, L.; et al. Adenocarcinoma of the stomach and exposure to occupational dust. American Journal of Epidemiology. Vol.128,No.1(1988). p.64-73.

    106. Wu-Williams, A.H.; Yu, M.C.; Mack, T.M. Life-style, workplace and stomach cancer by subsite in young men of Los Angeles County. Cancer Research. Vol.50(May 1, 1990). p.2569-76.

    107. Wyndham, C.H.; Bezuidenhout, B.N.; et al. Mortality of middle aged white South African gold miners. British Journal of Industrial Medicine. Vol.43(1986). p.677-84.



    Kusiak1 2 3 4 5 retrospective cohort mortality 54,128 Ontario miners (gold, nickel, copper, uranium, and other ores) followed by provincial chest clinics after 1955 who died between 1955-87 (exposed to arsenic, chromium, mineral fibre, diesel, emissions, aluminum powder) Ontario mortality data age, gold vs non-gold miners, born in N Am or not Gold miners SMR 152 (CI 125-185) for stomach cancer; no increased SMR for non-gold miners (uranium, copper, nickel) except those born outside N Am (SMR 160 95% CI 105-233);
    -gold miners born in N Am SMR 133; born out of N Am 177; SMR greater in those who were <60yr (167 **vs 145 miners 60-74) and born out of N Am (160); association statistically significant between gastric cancer mortality and time weighted exposure to dust in gold mines, chromium (but not arsenic or mineral fibre)
    no mention of denominator figures ie total number of deaths for each group of miners; no direct exposure measurements; difficult to assess power to detect increased stomach cancer in various gps of miners without knowing total number of deaths in each gp
    Hodgeson6 retrospective cohort mortality study 3072 UK (2 Cornwall mines) tin miners who worked at least 12 months between Jan 1/41 and May 1/84 excluding those born before 1880 or over age 60 when began mining exposed to radon, silica, arsenic (13 miners);(851 deaths) national (England, Wales) mortality rates calendar period, age, total underground exposure, lime since last exposure SMR Cancer of stomach 1.41 n.s. (27 cases) only 28% of cohort had died during the study period
    Lawler7 retrospective cohort mortality study 10,403 white male Minnesota iron-ore miners (45% underground, 55% above ground) employed for at least 12 months prior to Jan 1/66 followed from Jan 1/35 to Dec 31/78 (4699 deceased, 4478 death certificates obtained) US white males mortality rates; St Louis county mortality rates age, calendar time, above vs underground mining miners SMR 167; above ground 181, if total cohort SMR stomach cancer 172 (p<.01); underground (underground started exposures earlier, hired 1937-65 SMR=223, if native born SMR=155; but SMR=101 if controls=St Louis county; foreign born underground SMR=186, foreign born above ground=234; when compared to St Louis county no statistically signif increase except for Finnish born miners SMR 173 differences between underground and above ground miners worked longer, >3x more were foreign born 39.3% vs 11.7%)
    Atuhaire8 30 year prospective(?) mortality cohort study 7939 male miners and non-miners in the Rhondda Fach, Wales followed for 30 yrs death rates for England and Wales for the same period; comparison between miners, ex-miners and non-miners age, time period Miners/non-miners SMR=148 p<.05); non-miners SMR=131 (p>.05); no statistically signif differences between miners and non- miners % who died of total cohort not stated; very few non-miners were age > 45 (404/2000=20%); only 926 of 6182 miners were over age 65
    Swaen9 case control gastric cancer incidence study 323 male cases of gastric cancer diagnosed between Jan 1/73-Dec 31/83 in southern Netherlands 323 controls from same pathology dept matched for date of birth within 30 days age, coal mining or not no signif difference between cases and controls wrt occupation as a coal miner very low numbers, ?appropriateness of controls
    Swaen10 case control gastric cancer incidence study 683 male cases of gastric cancer diagnosed between Jan 1/73-Dec 31/83 in 5 pathology depts in southeastern Netherlands 683 controls diagnosed in the same pathology dept matched for date of birth age OR 1.15 (95% CI .89-1.47) as above
    Meljers11 retrospective pilot cohort mortality study 165 deaths of 334 Dutch coal miners randomly selected from 5400 Dutch coal miners diagnosed as having pneumoconiosis in routine cross-sectional x-ray examination between 1956-60 and registered with the General Mining Fund followed to Dec 31/83 cause, age, calendar time specific mortality rates for total Dutch male popn age, time period, smoking SMR stomach cancer =401 (95% CI 2.3-6.5) not based on a full occupational cohort; high mortality rate (49.4%), long follow up period (52% had >30 yr employment, 87.5%>20 yrs)
    Pham12 prospective proportional mortality study 13,801 French Iron miners from the Lorraine basin alive on Jan 1/82 followed from Jan 1/82-Dec 31/86 (1,283 deaths ie <10% cohort); <20% cohort >60 yr proportion of deaths of French male popn age, calendar time, smoking PMR stomach cancer 2.31 (p<.001); surface workers PMR 2.18 (p<.05); underground workers PMR 2.37 (p<.001); underground workers working 1-19 yr PMR=2.63, 20-29 yr PMR=2.66 (p<.001), >30 yr PMR=2.01 (p <.01); (51 deaths of stomach cancer; 80% of those who died of stomach cancer were 60-80 yr old) limitations of PMR study; small % died; relatively young pop'n for mortality study esp for stomach cancer multiple methods of ascertainment of cause of death of study population
    Weinberg13 case control 228 deaths 1978-80 for 4 counties in Penn (from state vital stats with death certificate after gastric cancer identified in v.s. 178 next of kin interviewed by tele (107 males, 71 females) 3 controls/case (107M/71F digestive cancer, 107/71 ASHD death, and 80/58 live neighbourhoodcontrol (dig-same sex, race, county, 5-yr age; ASHD- next death following case of same sex, race, county, within 5 yr age) tobacco chewing, smoking, age, SES (neighbour-hood; income), ethnicity (East European, foreign born); gender East European origin OR: 1.57 dig cancer 2.05 ASHD 1.88 neighbourhood; occupation n.s. increase in coal miners and farmers and wives of c miners; gas cooking protective poor study; info from next of kin, SES info poor; death certificates requested for cases only; neighbourhood controls problematic; data on occupation poor; power low
    Berger14 retrospective cohort mortality study all 4908 male Hamburg gas plant workers employed for 10+ yrs between Jan 1/00 - Dec 31/89 who died between 1952- 1989 (2699 total deaths); sub cohorts: 789 gas furnace workers with heavy exposure to coal tar gas (572 deaths); 3401 other blue collar workmen in other parts of the plant (1708 deaths) calendar period, age, and cause specific death rates for German males; subcohort of 718 white collar workers internal controls (419 deaths) age, calendar period, smoking SMR stomach cancer gas workers = 177 (95% CI (120-251); other workmen SMR=163 (88-142); white collar workers SMR=57 (27-105); when gas workers compared to white collar workers as standard SMR=284 (95% CI 205-385); smokers had higher SMR compared to non-smokers; strong healthy worker effect for most other causes of death multiple methods of ascertainment of cause of death of study pop'n including asking family members; no controls of ethnic identity, immigrant status, SES
    Siemiatyckl15 case control 156 Montreal incident cases of stomach cancer from multiple hospitals 1524 cases of other cancers age, ethnicity (French or not), SES, smoking, salubrity of work history OR 1.7 for exposure to wood dust with 5 stratifying variables (n=35) (95% CI 1.1-2.8) well designed study (poor wrt ethnic variable); limits of exposure coding
    Milham16 Proportional mortality study 2689 deaths of active or retired union members identified from listings in the Pulp, Sulfite and Paper Mill Workers' Journal 1935-64 (2113 death certificates found, 79%); 68 stomach cancer deaths age, calendar year specific US proportional mortality rates age, calendar year PMR stomach cancer = 176 (p < .05) significant methodologic problems; no attempt to determine what proportion of total cohort deaths included in journal listings; Canadian workers made up 25-30% but treated as US in PMR calculations
    Roscoe17 retrospective cohort mortality study 2294 white male wood model makers who had worked at least one month at one plant of each of the 3 major US auto makers between 1940 and 1980, followed 1940- 84; 706 deaths (31%) Wayne County, Mich mortality rates; US mortality rates age, calendar yr, place of birth national mortality rates overall stomach cancer compared to case control study 1940-80 SMR = 1.6 (95% CI 0.9-2.6), (17 cases); using Wayne County mortality rate SMR 1960-80 1.6 (0.9- 2.7), (15 cases); US mortality rate comparison 1960-80 SMR = 2.2 (1.2-3.6); difficulty with exposure measures in
    Svirchev18 Proportional mortality study of BC Wood and Pulp workers BC wood and pulp workers aged 20-65, 1950-78 BC mortality data not stated PMR stomach cancer carpenters 155 (95% CI 126-191); cabinet makers 150 (40-384); woodworkers misc 179 (89-320; pulp mill workers 115 (42-250); no increases for other wood/pulp workers methods not described in this paper (letter)
    Olsen19 Proportional incidence ratio (SPIR) study cases of stomach cancer notified to the Danish Cancer Registry 1970-84 for ages 16-66 employed after Apr 1/64 linked to personal employment histories on file in a nationwide Pension Fund for those in wood working occupations proportion of stomach cancers among all Danish employees age (5 yr gps), time (1 yr periods) SPIR stomach cancer 204 (p<.01) basic wood industries; 102 wood manufacturing industries; 139 (p<.01) agriculture, forestry, fishing; 101 manufacturing study unable to detect cancers in former Danish workers >age 66; limits of PMR
    Toren20 case control 33 men who died between 1961 and 1985 of stomach cancer from 3 parishes near 2 paper mills 550 men over age 40 who died in the same parishes of non- malignant non-respir diseases (no indiv matching age no statistically signif increase risk of exposure among stomach cancer cases very low power, low numbers
    Finkelstein21 retrospective "cohort" mortality study (see comments) 276 Ontario silica exposed non-miners in various industries (with 212 deaths - 204 death certificates located) and unspecified number of Ontario miners (with 905 deaths) identified from a WCB cohort of workers awarded compensation for silicosis between 1940 and 1985 age, sex, time specific Ontario mortality rates smoking, decade of initial award, industry type, birthplace miners with silicosis SMR stomach cancer 188 (CI 120-290); silica exposed non-miners SMR stomach cancer 366 (CI 147-754) high proportion died; limitations of accuracy of WCB files; not a true historical cohort study in which cohort is identified on the basis of exposure but rather mortality study of a cohort of cases for another diagnosis which suggests high silica exposure
    Coggon22 case control stomach cancer incidence study 95 cases of stomach cancer newly diagnosed in pathology reports or in the endoscopy, radiology and surgical depts of the local hospitals of Stoke-on- Trent and Newcastle- under-Lyme county districts in those less than 70 yrs of age between Jan/85 and Aug/87 2 controls/case (190) matched for age (within 2 yrs), sex from the county district popn registered with NHS GPs age, sex, diet (salad vegetables, fruit, salt) Relative Risk (RR) of cases holding a manual job in one of the 4 local dusty industries=1.4 (95% CI 0.8-2.4); RR manual work in rubber industry controlling for diet=2.5 (95% CI 1.0- 6.4) and when controlling for social class as well RR=3.3 (1.2-9.2) low numbers; minimal exposure information; no accounting for duration of exposure
    Neuberger23 24 prospective cohort mortality study 1630 male workers in multiple industries > 40 yr exposed to silica and "inert" dusts screened by the Vienna Occupational Health Care Unit in the 1950's followed to Dec/1985 by death certificates (87% had died) death rates in general Viennese population; internal control: 1650 male Viennese workers without occupational dust exposure (84% died) selected from the same screening examination records matched for age (yr of birth), yr of first examination, smoking status at time of first exam; age, smoking SMR stomach cancer in all exposed=166 (p<.01); in foundry workers 177 (p<.01); other metal workers 158 (p,.05); Internal controls SMR=90 occupational history limited to occupation at time of first examination; very crude exposure information; no way of ensuring that internal controls not exposed during subsequent employment (bias here would have underestimated association)
    Bross25 case control cancer incidence study 14,000 cases of cancer admitted to Roswell Park Memorial Institute, NY 1956-65 with occupational histories taken patients admitted in the same period with a non-neoplastic disease and patients employed as clerks sex, age, yrs in occupation, smoking, country of origin For stomach cancer RR of Carpenters >2 (p <.05); shoemakers RR=16.5 (p<.05); millwrights RR=5.5 (p<.05); when only those born in US included the following occupations RR>2, p<.05: blacksmiths, brickmasons, lumber industry, millwrights, paper industry operators, textile industry operators limits of case control study; crude occupational information
    Sorahan26 retrospective cohort mortality study 10,491 male UK steel foundry workers who worked for 1 or more yrs 1946-65 and followed 1946-85 (3110 died, cause known, 29%) mortality rates for England and Wales age, sex, calendar year, duration of employment, age at starting employment, factory, yr of follow up All steel foundry workers SMR stomach cancer = 137 (p<.01); furnace workers SMR=265 (p<.01);furnace repair SMR=359 (p<.001); highest SMR's for those working 10-19 yrs only 29% of cohort dead by end of follow up, therefore difficult to look at dose response relationships
    McDowell27 retrospective cohort mortality study (SMR) 607 men known to be cement workers in 1939 in north Kent, UK, and alive in 1948 identified by reviewing National Health Service Central Register records followed from Jan/1948 to Dec/1981 (419 deaths, 22 stomach cancer deaths) male mortality rates for England and Wales age, social class, calendar yr of death Stomach cancer SMR=175 (p<.05); packing or loading labourers (high dust exposure) SMR=321 (p>.05); SMRs above that for stomach cancer in social class IV or V 69% of cohort had died; no information on duration of employment or level of exposure except by job title; ? accuracy of occupational information
    Amandus28 retrospective cohort mortality study (SMR) 4231 white male workers employed in 23 US non-asbestos cement plants for at least 5 years between 1950-1960 (5292 men met inclusion criteria but lack f vital status or incomplete occupational or other history on 1061; death certificates available for 31% of cohort (32.6% cohort died; 135 stomach cancer deaths) age-year specific US white male mortality rates age, calendar year, latency intervals, years of employment, country/continent of birth (US/non-US) SMR for stomach cancer overall 135 (not statistically significant); for those with >20 yr latency SMR=141 (p>.05); for those born <1900 followed up 1965-9 SMR=491 (p<.05), those born >1900 and followed 1965-9 SMR=221 (p>.05); for other periods of follow up (<1964, >1970) no signif SMRs low power - negative results unconvincing because of study limitations; less than a third of cohort had died; high proportion of original 8714 workers enumerated excluded: 1791 because began working before 1950, 828 excluded because work history before 1960 unavailable - exclusion of more older workers limiting potential of finding a statistically signif increased SMR; no attempt to check accuracy of stomach cancer diagnoses particularly in early decades of study
    Wu-Williams29 case control 137 incident male cases of stomach cancer under age 55 identified through the L A County Surveillance Program (cancer registry) interviewed between 1975-1984 137 age-race-sex matched neighbourhood controls social class, nativity, dietary factors, smoking, alcohol Metal dust exposure RR (conditional logistic regression) =1.7 (95% CI 1.0-3.0); metal dust exposure 1-19 yr RR = 1.4 (0.7- 2.5); 20+ yr exposure RR= 7.4 (1.6-33.5) Despite low numbers and controlling for other confounders occupational metal dust exposure show very high association with stomach cancer; large number of cases' occup info came from surrogate
    Dubrow30 PMR study of Rhode Island jewellery workers 1334 female and 1807 male RI jewellery workers listed in RI death certificate records 1968-78 all other decedents listed in RI death certificate records 1968-78 standardized PMR for gender,race (B,W), age, yr of death PMR 174 (CI 114-265) among female jewel, workers; neg for males (different jobs) limitations of PMR method, of death certificate data; exposure based on only broad occup categories not specific exposures; no control for SES, immigrant status or ethnicity
    Sparks31 PMR study of Attleboro, Massachusetts jewelry workers; nested case control 931 male who died in Attleboro, Mass. between 1956 and 1975 and were listed as jewelry workers on death certificates; 14 stomach cancer deaths US white male mortality data; case control internal comparison: all non- stomach cancer deaths age, calendar year, ethnicity (French Canadian) Overall PMR for stomach cancer = 1.2; PMR for polishers = 3.9; Odds ratio comparing stomach cancer to all other causes of death in polishers = 4.4 (p<.01) Small numbers; only one p-value cited - ? statistical significance of other results, weak study
    Garabrant32 PMR with nested case control PMR: 78 female and 758 male leather workers; 767 F and 1195 M shoe workers in 3 Massachusetts communities listed in Mass death certificates 1954-74 all non cancer decedents listed in the above death certificates of 3 Mass. communities standardized PMR for gender, age, yr of death Male leather workers PMR stomach cancer 1.69 (CI 1.04-2.73); F leather PMR 2.80 ns (2 obs deaths); case control study did not show statistically signif increase in exposure in those with stomach cancer limitations of PMR method, of death certificate data; PMR exposure based on only broad occup categories not specific exposures; no control for SES, immigrant status or ethnicity
    Edling33 case control cancer mortality study looking at exposure to tannery occupation 19 male cases of stomach cancer were identified between 1951-1979 in a local parish death registry in SE Sweden which had a very large tannery between 1873- 1960 388 male controls who had died of causes other than cancer during the same period in that parish age SMR for stomach cancer and exposure to tanning (as occupation on death registry) = 1.5; Mantel-Haenzel rate ratio point estimate = 1.6 (95% CI 0.6-4.0) low numbers, low power; plant closed 1960, those who switched jobs after that may be listed in registry with another occupation
    Park34 PMR mortality study of a employees from a Connecticut ball bearings plant; case control study for specific disease outcomes and exposure categories (exposures to grinding dust, abrasives, water- based cutting fluids and straight cutting oils) PMR: all hourly employees with 10 or more yrs of service who died between Jan 1/69 and July 31/82 = 768 deaths (702 had death certificates & job histories - 565 white males, 20 non- white males, 116 white females, 1, non-white female) Case control: stomach cancer deaths from above pop'n (11 stomach cancers in white males) PMR: US proportional mortality rates Case control: all deaths excluding other GI cancer sites, lung cancer and non- malignant GI disease age, sex, race, calendar yr, Central European origin PMR: for stomach cancer in white males PMR=3.1, p<.002; females PMR=3.1, p=.15; men in grinding PMR=3.8, p=.006; grinders with 30+ yr latency PMR=4.2, p=.03); Case control: crude mortality OR for stomach cancer and grinding=5.9, p=.06; water-based cutting fluids OR=6.2, p=.05; no association found for straight oils limits of PMR analysis; attention to exposure more detailed than most studies; good discussion of threats to validity of this study; case control results suggest stomach cancer risk in grinding is likely due to additives in fluids not present in straight oils (eg petroleum sulfonates, amine soaps, nitrite corrosion inhibitors) abrasive aerosols, or oils with abrasive debris
    Silverstein35 Proportional mortality study, nested case control PMR: 1532 white male bearing plant unionized workers whose year of termination was 1950 or later and who died between Jan 1950 and June 1982 (exposures to cutting fluids, abrasives, grinding dust; Case control: 35 stomach cancer cases PMR: national US proportional mortality rates; Case control: all non- cases of stomach cancer age, yr of death, Central European origin, French Canadian origin PMR: overall PMR for stomach cancer = 1.97 (p < .001); PMR for those with 10 or more years exposure to grinding (comprehensive) = 3.39 (p < .001); PMR Tool grinders = 5.2 (p=.04); no excess found in machining occupational categories; Case control: Crude Mortality Odds Ratio 10 or more yrs Comprehensive grinding 2.37 (p=.03); logistic regression controlling for non-US origin Adjusted Odds Ratio for comprehensive grinding = 2.25 (p=.024) Study appears to have begun as a retrospective cohort mortality study but been analyzed as a PMR study. Cause of death and job information available on 87% of 2027 deaths identified - 1532 in white males; exposure data based on years in specific job groupings with knowledge of hygiene values - more sophisticated exposure estimates than most other studies
    Sorahan36 37 retrospective mortality cohort study 36,691 male UK rubber workers in 13 factories who worked for minimum of 1 yr 1946-60 followed 1946-85 by the National Health Service Central Register or Scottish General Register Office (11,765 dead, 32.1%; 359 stomach cancer deaths); internal controls of different occup gps mortality rates of England and Wales or Scotland age, calendar yr, duration of employment, factory, yr of starting employment, age at starting employment, yr of follow up SMR stomach cancer= 113 (p<.05); age <50 yr SMR=139; age >50 yr SMR=110; greatest increases in SMR for those with latency of 10-19 yrs; SMR higher for those first employed 1956-60 (SMR=133, p<.05) than for those first employed 1946-50 (SMR=117, p<.01); suggestion that risk associated more with exposure to dust than fumes or solvents but test statistic not statistically signif exposure data limited to yrs of employment in industry; no measurement of ethnicity or immigrant status; considerable care, concern with methods
    Monson38 retrospective mortality cohort (SMRs) study 13,571 white male union rubber workers in any of 3 rubber plants of one company in Akron, Ohio who worked 5 yrs or more followed Jan/40-Jun/74 (5079 deaths, 37%) age-time-cause specific mortality rate for US white males and for Ohio males age, calendar period no increased stomach cancer risk for all company workers; processing workers (exposure to high levels dusts, fumes including carbon black) SMR for stomach cancer = 190 (18 cases); age >74 at death SMR for GI cancer = 250 access to company and union records; concerns with methodologic issues discussed; no specific exposure measurements (but departments info available for each worker); exposures following termination unavailable for 1515 workers
    McMichael39 retrospective cohort mortality study 4 cohorts of male rubber workers in tire plants in 4 companies in Akron, Ohio followed Jan/64- Dec/74 aged 40-84 at inception (5106 deaths, 27%) 1968 race-age-cause specific US male death rates age,race Stomach cancer SMR ages 40-84 = 148; SMR ages 4-64 = 183, ages 65-84 = 136 calendar year not taken into account; no specific exposure measurements presented; limited analyses presented
    McMichael40 retrospective cohort mortality study 6678 male hourly rubber workers aged 40-84 Jan 1/64 at a tire plant in Akron, Ohio followed Jan/64 to Dec/72 (1783 deaths, 27%) 1968 age-cause specific US male death rates age Stomach cancer SMR age 40-84 = 187 (p<.001); age 40-64 SMR = 219 (p < .01) calendar year not taken into account; no specific exposure measurements presented; limited analyses presented
    Kneller41 case control 13,489 gastric cancer cases from the urban Shanghai cancer registry in persons over age 30 during 1980-84; census derived age-sex specific cancer incidents rates calculate for each occupational group age-sex-site specific cancer incidence rates calculated for urban Shanghai age, sex, occupation Multiple statistically signif elevated SIR (standardised incidence rates) found among males: e.g. grain farmers SIR = 402; wood processing workers 176; cabinetmakers 147; sanitation workers 157; rickshaw drivers 153; leather products 150; blacksmiths 156; machine tool setters 305; sheet metal workers 140; metal grinders 141; dock workers 154; railway engine drivers and firemen 363; among females: grain farmers 414; wood processing workers 308; machinery/engine mechanics 231; sanitation workers 193; rickshaw drivers 172; plastics manufacture 160; knitters 148 case control limits; multiple comparisons; diet, smoking, home coal cooker exposures not taken into account
    Wright42 case control stomach cancer proportional incidence ratio study 1342 male cases of adenocarcinoma of the stomach in a popn based cancer registry, LA County, Calif aged 20-64 diagnosed between 1972-82 (657 cases of stomach cancer not included in study because 307 not adenocarcinomas, 350 inadequate info on occupation) distribution of cancers in the LA County Cancer Surveillance Program age, race, SES (based on education, income of adults in census tract of residence), immigrant status PIR stomach cancer of exposed/PIR of unexposed = 1.3 (95% CI 1.2-1.4); for antrum/pylorus 1.8 (1.6-2.0); mineral dust 1.6 (1.0-2.5); organic dust 1.5 (1.0-2.3) exposure to dust inferred from occupational title; occupation based only on job at time of diagnosis; men > 64 not studied because occupation not available, likely leading to underestimate of cancer and occupation link; limits of proportional incidence study; no multivariate analysis when controlling for multiple confounders
    Heldaas43 retrospective cohort cancer incidence study 2391 male Norwegian magnesium plant workers who worked more than 1 year between 1951 and 1974 followed 1953 to 1984 (exposures to volatile coal tar pitch products, chlorinated hydrocarbons (eg hexachlorbenzene), asbestos, chlorine gas, magnesium oxide and chloride dust, strong electrical magnetic fields); 21 cases of stomach cancer national Norwegian 5 year, age specific cancer incidence rates age, calendar year RR for stomach cancer. for overall study population = 1.6 (95% CI 1.0-2.4); stomach cancer showed highest rates in those with short and medium employment experience Information on years of employment and age distribution of cohort not provided; methodologically sound study


    1.Kusiak RA, Ritchie AC, Springer J, Muller J. Mortality from stomach cancer in Ontario miners. Br J Indust Med 1993, 50, 117-126

    2.Kusiak RA, Springer J, Ritchie AC, Muller J. Carcinoma of the lung in Ontario gold miners: possible aetiological factors. Br J Indus Med 1991, 48, 808-817

    3.Muller J, Kusiak RA, Suranyi G, Ritchie AC. Study of Mortality of Ontario Gold Miners 1955-1977. Feb, 1987, Ontario Ministry of Labour, Workers' Compensation Board of Ontario, Atomic Energy Control Board of Canada

    4.IDSP. Report to the Workers' Compensation Board on the Ontario Gold Mining Industry. IDSP Report No 1, Toronto, Ont, Apr, 1987

    5.Muller J, Wheeler WC, Gentleman JF, Suranyi G, Kusiak RA. Study of Mortality of Ontario Miners 1965-1977. Part I. April, 1983, Ontario Ministry of Labour, Workers' Compensation Board of Ontario, Atomic Energy Control Board of Canada

    6.Hodgeson JT, Jones RD. Mortality of a cohort of tin miners 1941-86. Br J Indust Med 1990, 47,665-676

    7.Lawler AB, Mandel J, Schuman LM, Lubin JH. A retrospective cohort mortality study of Iron ore (hematite) miners in Minnesota. J Occup Med 1985, 27, 507-517

    8.Atuhaire LK, Campbell MJ, Cochrane AL et al. Gastric cancer in a south Wales valley. Br J Indust Med 1986, 43, 350-52

    9.Swaen GMH, Aerdts WHM, Sturmans F, et al. Gastric cancer in coal miners: a case control study in a coal mining area. Br J Indus Med 1985, 42, 627-630

    10.Swaen GMH, Aerdts, Slangen JJM. Gastric cancer in coal miners: final report. Br J Indust Med 1987, 44, 777-779

    11.Meijers JMM, Swaen GMH, Slangen JJM, et al. Long-term mortality in miners with coal workers' pneumoconiosis in the Netherlands: a pilot study. Am J Indust Med 1991, 19, 43-50

    12.Pham QT, Chau N, Patris A et al. Prospective mortality study among iron miners. Cancer Detection and Prevention 1991, 15, 449-454

    13.Weinberg GB, Kuller LH, Stehr PA. A case control study of stomach cancer in a coal mining region of Pennsylvania. Cancer 1985, 56, 703-713

    14.Berger J, Manz A. Cancer of the stomach and colon-rectum among workers in a coke gas plant. Am J Indust Med 1992, 22, 825-834

    15.Siemiatycki J, Richardson L, Gerin M, et al. Associations between several sites of cancer and nine organic dusts: results from an hypothesis-generating case control study in Montreal, 1979-83. Am J Epidemiol 1986, 123,235-49

    16.Milham S, Demers RY. Mortality among pulp and paper workers. J Occup Med 1984, 844-6

    17.Roscoe RJ, Steenland K, McCammon CS, et al. Colon and stomach cancer mortality among automotive wood model makers. J Occup Med 1992, 34, 759-770

    18.Svirchev LM, Gallagher RP, Band PR et al Gastric Cancer and lymphosarcoma among wood and pulp workers. J Occup Med 1986, 28, 264-5

    19.Olsen JH, Moller H, Jensen OM. Risks for respiratory and gastric cancer in wood-working occupations in Denmark, J Cancer Res Clin Oncol 1988, 114,420-424

    20.Toren K, Sallsten G, Jarvholm B. Mortality from asthma, COPD, respiratory system cancer, and stomach cancer among paper mill workers: a case-referent study. Am J Indust Med 1991, 19, 729-737

    21.Finkelstein M, Liss GM, Krammer F, Kusiak RA. Mortality among workers receiving compensation award for silicosis in Ontario 1940-85, Br J Indust Med 1987, 44, 588-594

    22. Coggon D, Barker DJP, Cole RB. Stomach cancer and work in dusty industries. Br J Indust Med 1990, 47, 298-301

    23.Neuberger M, Kundi M. Occupational dust exposure and cancer mortality - results of a prospective cohort study. In Fletcher AC, Saracci R, Thomas TL (ed): Occupational Exposure to Silica and Cancer Risk. Lyon, International Agency on Cancer, 1990, 65-73

    24.Neuberger M, Kundi M, Westphal G, Grundorfer W. The Viennese Dusty Worker Study. In: Silica, Silicosis and Cancer, 1986, 415-422

    25.Bross ID, Viadana E, Houten L. Occupational cancer in men exposed to dust and other environmental hazards. Arch Environ Health 1978, 33,300-307

    26.Sorahan T, Cooke MA. Cancer mortality in a cohort of United Kingdom steel foundry workers; 1946-85. Br J Indust Med 1989, 46, 74-81

    27.McDowell ME. A mortality study of cement workers. Br J Indust Med 1984, 41, 179-182

    28.Amandus HE. Mortality from stomach cancer in US cement plant and quarry workers, 1950-80. Br J Indust Med 1986, 43, 526-528

    29.Wu-Williams AH, Yu MC, Mack TM. Lifestyle, workplace and stomach cancer by subsite in young men of Los Angeles County. Cancer Research 1990, 50, 2569-76

    30.Dubrow R, Gute DM. Cause-specific mortality among Rhode Island jewellery workers. Am J Indust Med 1987, 12, 579-593

    31.Sparks PJ, Wegman DH. Cause of death among jewelry workers. J Occup Med 1980, 22, 733-736

    32.Garabrant DH, Wegman DH. Cancer mortality among shoe and leather workers in Massachusetts. Am J Indust Med 1984, 5, 303-314

    33.Edling C, Kling H, Axelson O. Cancer mortality among leather tanners. Br J Indust Med 1986. 43, 494-496

    34.Park RM, Wegman DH, Silverstein MA, et al. Causes of death among workers in a bearing manufacturing plant. Am J Indust Med 1988, 13, 569-680

    35.Silverstein M, Park R, Marmor M et al. Mortality among bearing plant workers exposed to metalworking fluids and abrasives. J Occup Med 1988, 30, 706-714

    36.Sorahan T, Parkes HG, Veys CA, et al. Mortality in the British rubber industry 1946-85. Br J Indust Med 1989, 46, 1-11

    37. Sorahan T, Parkes HG, Veys CA, Waterhouse JAH. Cancer mortality in the British rubber industry 1946-80. Br J Indust Med 1986, 43, 363-373

    38.Monson RR, Nakano KK. Mortality among rubber workers: 1. White male union employees in Akron, Ohio. Am J Epidemiology 1976, 103, 284-296

    39.McMichael AJ, Andjelkovic DA, Tyroler HA. Cancer mortality among rubber workers: an epidemiologic study. Annals New York Academy of Sciences 1976, 125-137

    40.McMichael AJ, Spirtas R, Kupper LL. An epidemiologic study of mortality within a cohort of rubber workers, 1964-72, J Occup Med 1974, 16, 458-464

    41.Kneller RW, Gao YT, McLaughlin JK, et al. Occupational risk factors for gastric cancer in Shanghai, China. Am J Indus Med 1990, 18, 69-78

    42.Wright WE, Bernstein L, Peters JM. Adenocarcinoma of the stomach and exposure to occupational dust. Am J Epidemiology 1988, 128, 64-73

    43.Heldaas SS, Langard S, Andersen. Incidence of cancer in a cohort of magnesium production workers. Br J Indust Med 1989, 46, 617-623


    1.Parkin D, Laaea E, Muir C. Estimates of the worldwide frequency of 16 major cancers in 1980. Int J Cancer 1988, 41, 184-97

    2.Page HS, Asire AJ. Cancer rates and risks, Third edition. 1985. NIH publication no. 85-691, US Department of Health and Human Services

    3.Neugut AI, Wylie P. Occupational cancers of the gastrointestinal tract. Occupational Medicine: State of the Art Reviews 1987, 2, 109-135

    4.Nutt A. Rubber work and cancer - past, present and perspectives. Cancer Mortality and Morbidity 1983, 9, 49-57

    5.IDSP. Report to the Workers' Compensation Board on the Ontario Gold Mining Industry, IDSP Report No 1, Toronto, Ont, April 1987

    6.Sackett DL, Haynes RB, Tugwell P. Clinical Epidemiology: a Basic Science for Clinical Medicine, Little, Brown and Co., Toronot, 1985

    7.Ames, RG. Gastric cancer in coal miners: some hypotheses for investigation. J Soc Occup Med 1982, 32, 73-81

    8.Nutt A. Rubber work and cancer - past, present and perspectives. Scand J Work Environ Health 1983, 9, suppl 2, 49-57

    9.Kraus AS, Levin ML, Gerhardt PR. A study of occupational associations with gastric cancer. Am J Public Health 1957, 47, 961

    10.Hagstrom RM, Sprague HA, Landau E. The Nashville air pollution study. VII. Mortality from cancer in relation to air pollution. Arch Environ Health 1967, 15, 237

    11.Winkelstein W Jr, Kantor S. Stomach cancer. Positive association with suspended particulate air pollution. Arch Environ Health 1969, 18 544

    12.Armijo R, Gonzales A, Orlellana M et al. Epidemiology of gastric cancer in Chile: II - nitrate exposures and stomach cancer frequency. Intl J Epidemiology 1981, 10, 57

    13.Meyer MB, Luk GD, Sotelo JM, et al. Hypothesis: the role of the lung in stomach carcinogenesis. Am Rev Respir Dis 1980, 121, 887-892

    14.Gamble JF, Ames RG. The role of the lung in stomach carcinogenesis: a revision of the Meyer hypothesis. Medical Hypotheses 1983, 11, 359-364

    15.Ames RG, Gamble JF. Lung cancer, stomach cancer, and smoking status among coal miners: a preliminary test of a hypothesis. Scand J Work, Environ, Health 1983, 9,443-48


    Report on Analysis of Stomach Cancer Among Ontario Gold Miners

    1. Rationale for Analysis

    Previous analysis of stomach cancer among Ontario gold miners (Muller et al., 1983, 1987) has shown that era of hire is a major factor in risk: those hired prior to 1945 had an SMR of 132 and those hired in 1945 and later had an SMR of 322. The pattern of increased risk in more recent hires is unexpected in light of the general improvement in gold mining conditions over the century and has been seen as evidence against an occupational role for gold-mining in the development of stomach cancer.

    Given that the earlier hires are necessarily an older cohort and that excess risk appears to be higher in the younger age groups (as found here and that by Muller, 1987), the above comparison of SMRs is problematic. This is a classic example of the difference in age structure of two cohorts confounding the comparison of SMRs (Breslow and Day, 1987).

    Poisson regression analysis of followup to 1977 (Muller et al., 1987) showed that this difference by era of hire continued even after adjusting for age and calendar year. There, the relative risk for the later hire period was about 2 and was statistically significant. A test for interaction between age and period of hire was not significant. Kusiak et al (1993) state, without presenting evidence, that differences in stomach cancer mortality between the pre and post-1945 hires are an artefact of the age distribution, the differences in time since hire and differences in the content of the ores mined in the different periods. In particular, Kusiak et al believe that chromium is responsible for the elevated stomach cancer in miners and have evidence that the major chromium-containing mines did not open until after 1945.

    Since year of hire and age are closely related to the issue of eligibility for compensation for stomach cancer, the Occupational Disease Panel decided that these variables warranted further investigation. Preliminary results of this investigation were shown to Professor Paul Corey of The University of Toronto, who suggested that extending the analysis to include duration of employment and age at hire would give a clearer picture of the work-relatedness of this disease. Kusiak (1993) found that age was related to risk of stomach cancer in miners born in North America (lower risk in older workers), but that no such pattern existed in foreign-born workers. For this reason, place of birth was added to the list of factors to be investigated. After a review of the literature on the healthy worker effect (see section 2.4.7) , a variable representing active working status was also included.

    In summary, the analyses presented here were motivated by an attempt to address the effects of age and year of hire as they pertain to stomach cancer risk in gold miners. As the work progressed, the scope of investigation was enlarged to include variables that spoke directly to an exposure-response relationship as well as those which might act as confounders of this relationship. This report refers only to the analysis in its final stage. Furthermore, this report details only the methodological approach and the results of this statistical analysis. There is no attempt at interpretation of the results in terms of what is known about stomach cancer.


    2.1 Data

    Data originated in the Mining Master File (MMF), the origins of which have been documented elsewhere. The MMF is actually a set of files linked by the unique identification number assigned to each miner. One of these files contains personal information about each miner, such as his birth date and country of birth, which is used for linkage to the CMDB at Statistics Canada to determine date and cause of death. The Ontario Ministry of Labour has so obtained mortality follow-up to the end of 1986 .

    The file containing employment histories has multiple records for each miner, one for each period of employment, listing the ore, the mine, the occupation, the start date and number of months of work during that period. Below is a section of this file showing two employment histories. Notice that for the second employee, there is a substantial gap in the employment record between early 1963 and mid-1979. For this cohort, there is a substantial number of employees with gaps of varying lengths which necessitated careful calculation of duration of employment. This topic will come up again in the section on allocation of person-years (section 2.5).

    Table 1: Sample MMF work histories.

    A002131 1 1956 11 16 8-2 32 1.5 8 A002131 2 1957 1 1 8-2 32 7.5 8 A002131 3 1957 8 16 8-2 12 4.5 8 A002131 4 1958 1 1 8-2 12 8.5 8 A002131 5 1958 9 16 8-2 31 3.5 8 A002131 6 1959 1 1 8-2 31 12.0 8 A002131 7 1960 1 1 8-2 31 12.0 8 A002139 1 1956 2 16 8-2 32 10.5 8 A002139 2 1957 1 1 8-2 32 12.0 8 A002139 3 1958 1 1 8-2 32 12.0 8 A002139 4 1959 1 1 8-2 32 12.0 8 A002139 5 1960 1 1 8-2 32 12.0 8 A002139 6 1961 1 1 8-2 32 12.0 8 A002139 7 1962 1 1 8-2 32 12.0 8 A002139 8 1963 1 1 8-2 32 2.1 8 A002139 9 1979 7 2 8-5 32 6.0 8 A002139 10 1980 1 1 8-5 32 4.0 8 A002139 11 1981 4 11 8-1 14 5.0 8

    Data in the MMF was not always complete for every entry for a miner. In order that as much data as possible could be used, without introducing biases, the following approaches were taken for missing data:

    Table 2: Treatment of Missing Data
    Missing Treatment
    Start day or start month Not needed, since person-years were allocated on a whole-year basis
    Start year Never missing
    Ore Observation could not be used
    Length of Employment Assigned average of rest of MMF, according to ore type and number of periods of employment in year
    Birth date Subject could not be used

    2.2 Reference rates

    The reference rates used in calculating expected numbers of deaths by age and calendar period were those for stomach cancer (ICD9 codes 151-151.9) among Ontario males, obtained from Stats Canada in 1991 by the (then) IDSP for analysis of mortality among Metro fire fighters. Fortunately, the code for stomach cancer has not changed through the various ICD revisions, so it was not necessary to translate the ICD codes in effect at the time of death to the 9th revision ICD.

    2.3 Cohort definition:

    The study carried out by the IDSP and by Muller et al. examined both lung cancer and stomach cancer among gold miners, so the cohort was defined to exclude time after exposure to agents known a priori to cause lung or stomach cancer (in particular, asbestos for both stomach and lung cancer and uranium for lung cancer). The present analysis focuses exclusively on stomach cancer, so no exclusion based on lung cancer was necessary.

    The criteria for a miner's inclusion in the cohort were

    (1) Some gold-mining experience; and
    (2) No asbestos mining; and
    (3) Employment on or after Jan 1, 1955: Anyone who attended the chest X-ray clinics after the start of 1955 was entered into the MMF. However, some of these employees had retired prior to 1955. Since this self-selected group may have exhibited different mortality patterns from those who did not continue to go in to the clinics, and who never entered the MMF, they were not included.

    Entry into the cohort was defined as the either the date of first mining gold or January 1, 1955, whichever came later.

    Person-years were accrued from time of entry until the earlier of:

    (1) Death from stomach cancer, counted as an event;

    or, all counted as censoring:

    (2a) Death from some other cause;

    (2b) The end of the follow-up period: December 31, 1986;

    (2c) Date at which a miner turned 75. Due to demonstrable problems with ascertainment in older miners, previous analysis of this cohort has ended follow-up at this age.

    2.4 Reduction of Bias and Choice of Stratification Variables:

    2.4.1 Attained age: Previous analysis of this cohort has shown a consistent trend to decreasing relative risk for stomach cancer with attained age. Age groups were defined by 5 year intervals between age 35 and age 75. No deaths were observed (and only 0.6 were expected) below age 35, so person time for these strata was omitted. Including them as the "under 40" group would not have changed the results, but might have led to misunderstanding about exactly which ages experienced relatively high mortality.

    2.4.2 Calendar year: No trends with calendar time were found in previous analysis of this cohort, but it was necessary to verify this under the new cohort definition.

    2.4.3 Cumulative duration of employment: As a measure of exposure to potential carcinogens, the quality of duration of employment in gold mining is unknown. Heterogeneity of exposure over time, across mines and even within mines may be responsible for duration of employment being a poor measure of exposure to whatever may be causing the excess of stomach cancer. In general, this type of measurement error biases estimates of relative risk downwards. However, without additional information about working conditions in each mine over the study period, no better measure is available. Since gold miners typically did not work continuously in gold between their first and last employment in gold mining (see section 4.1 for details), it was important to accumulate duration only during active work periods.

    2.4.4 Age at hire: Although there was no prior evidence that this was a risk factor for stomach cancer, it is suggested in the literature (Checkoway et al, 1989) that several time-related variables be examined in addition to age and calendar time. Professor Corey also recommended inclusion of this variable.

    2.4.5 Year of hire (pre and post 1945): As noted above, that the risk was higher in recent years has been somewhat controversial in analysis of this cohort.

    2.4.6 Ethnicity: Stomach cancer rates around the world fluctuate widely and appear to be strongly related to local diet. Apprixmately 32% of the gold miners were born outside North America, predominantly in Europe, where some countries have stomach cancer rates higher than those in Ontario. In addition to their potentially higher background rates of stomach cancer, these immigrant miners may have been assigned to the dirtier jobs in the mine, at least early in their careers, thus weakening the association between duration of emplyoment and actual exposure. Inclusion of ethnicity as a stratification variable should reduce this potential bias.

    2.4.7 Employment Status: Healthy Worker Survivor Effect

    The healthy worker survivor effect (HWSE) refers to the observation that long-term workers experience lower mortality than would be expected and that short-term workers experience higher mortality than would be expected, all other things being equal. The appearance is that those with the longest exposure to the workplace environment benefit from the exposure. If there is no work-related increase in mortality, this manifests as lower mortality rates among long-term workers than among short-term workers. If there is an association between length of employment and mortality, the healthy worker survivor effect will tend to obscure it. Since a dose-response relationship is one of the important criteria for assessing causation in occupational epidemiology, it is essential to reduce bias that directly affects this relationship.

    The HWSE has been the topic of much discussion in the literature (Steenland, 1991 and 1996; Fox and Collier, 1976; Arrighi et al., 1994 for review and further references) and there have been several suggestions as to how to adjust for it in data analysis. Consensus appears to be that the methods of Robins (1987) are the only ones that directly address the rather complex relationships that exist between exposure, continued employment and mortality. However, software to implement these methods is not generally available.

    Among the other proposed methods, one that has been shown to give reasonable results under certain assumptions, (Steenland, 1996) and the one that is used here, is to classify time at risk as either active or inactive. The rationale for this can be seen by studying the simple example in Table 3, below.

    Table 3: Demonstration of Healthy Worker Survivor Effect
    Active Time Inactive Time Combined Death Rate
    Exposure Pyrs Deaths Death Rate Pyrs Deaths Death Rate
    Low 10000 10 1.0/1000 30000 60 2.0/1000 1.75/1000
    Medium 8000 8 1.0/1000 8000 16 2.0/1000 1.5/1000
    High 6000 6 1.0/1000 4000 8 2.0/1000 1.2/1000

    In this example, although there is no exposure-related mortality trend in either the active (rate = 1/1000) or inactive (rate=2/1000) person-time, the combined person-time shows a decreasing trend. This is a result of two things:

    (1) The higher overall mortality rate among inactive person-time; and
    (2) The high proportion of inactive person-time in the low-exposure group.

    The inactive group is heterogeneous: some of them choose other types of work; some of them are unable to handle the present work and are forced to find other types of work; some of them suffer ill health as a result of the present work and are unable to continue work. Other possibilities exist. The key point is that the inactive group contains all of those at the highest risk for mortality of many kinds: the unemployed, the sick and the sickly. People who leave employment early, but who remain "on study", contribute all their time at risk to the low-exposure group. In the example above, the low exposure group contains the highest proportion of these inactive/high risk person-years, so exhibits the highest overall mortality rate, close to that of the entire inactive group.

    People with high exposures have relatively less inactive person-time once their careers are over, so in the highest exposure categories, relatively more person-time is active time and the overall mortality rate is closer to that in the active group. A similar example could be worked up to show how a true dose-response could be obsfucated by the HWSE.

    There is some disagreement as to which causes of death are affected by this bias, especially with regard to mortality from cancer. One report, based on a summary of several cohorts thought to have no exposure-related mortality, suggests that cancer causes of death do exhibit an overall healthy worker effect, but the HWSE is not specifically mentioned (Park et al. 199?). But the general feeling has been that the healthy worker survivor effect in general does not operate as strongly on cancer causes of death, as evidenced in Steenland (1991). The argument seems to be that since cancer is not a chronic disabling disease, but rather a silent, acute and fatal one, it is less likely to cause self-selection out of the workplace . This may be an incomplete view: once workers are selected out of the workplace, for whatever reason, those entering into unemployment or worse-paying jobs may be more susceptible to diseases of lifestyle as a result of worsened financial situations.

    2.5 Assignment of person-years to the various strata

    Numbers of observed deaths, person-years and expected deaths were cross-classified by all of the variables mentioned above. Table 4 below shows the strata in detail.

    Table 4: Strata Definition
    Variable Groupings
    calendar year 1955 to 1984 by 5 years, and then 1985-1986
    age 35-40, 40-44,..., 70-74
    duration 0-5, 5-10,..., 20-25, 25+ years
    employment status mining/not mining
    start year <1945/1945 +
    birth place North America/outside North America
    age at hire 15-23, 23-32, 33+

    Calendar year, age, duration of gold-mining and employment status were treated as time-dependent variables, so that subjects "passed through" strata as time on study elapsed. Other variables were static, so that subjects remained in only one stratum.

    The treatment of duration of employment deserves some explanation, in light of the sporadic nature of gold-mining. For each miner, each calendar year of follow-up was divided into time spent mining gold, time spent mining other ores and time that was not accounted for in the MMF. Only time spent mining gold counted toward cumulative gold-mining exposure; time spent in any kind of mining was classified as active time; and time not accounted for in the MMF was classified as inactive time. For a given miner (with fixed variables that do not change over the follow-up period), each full year of follow-up was classified according to the following time-changing variables: (1) calendar year; (2) age at the end of the year; and (3) cumulative gold exposure at the end of the year. Within the stratum defined by these three variables, the person-year was subdivided according to the fraction that was spent actively mining and the fraction that was spent not mining.

    Incompatibilities in the MMF made it infeasible to use exact dates for allocation of person-years. Lengths of employment were given in months and decimal fractions of months that could not easily be added to year-month-day start dates to give end-of-employment dates. In many cases, a calculated end-of-employment date exceeded the next start date by an amount that was larger than could be explained by the variation in month length. Complicating matters further, start dates and length of employment periods were sometimes missing. All of these issues were resolved by taking the approach outlined above, which ignored ordering of employment periods within a year.

    3. Statistical Methods

    3.1 Calculation of observed and expected deaths Within each stratum defined by cross-tabulation of the above variables, the number of person-years at risk was multiplied by the reference rate for Ontario as a whole to give the expected number of deaths under the hypothesis of no excess risk. The observed deaths were classified into these same strata according to a miner's characteristics at time of death: age, year, duration of employment and so on. SAS (SAS Institute Inc, Cary, NC) , was used to create stratified tables of observed and expected.

    3.2 Poisson regression: Poisson regression was used to test for associations between risk of stomach cancer and the stratification variables listed above. Details of this method can be found in Breslow and Day (1987), but a brief summary follows.

    First, it is necessary to understand what a Poisson distribution is. For any given average number of deaths, the Poisson distribution describes the relative probabilities for the numbers of deaths we might observe. If the real average is 1.3 deaths, for example, the Poisson distribution says how often we should expect to observe 0, 1, 2, 3 deaths and so on.

    In Poisson regression, the observed number of deaths for those at risk with a certain set of characteristics is assumed to come from a Poisson distribution. The average for this Poisson distribution is a sum of the average based on the reference population and some factors based these characteristics. For example, if we have a regression model with only the expected number and age, the regression equation is:

    log(OBSERVED) = log(EXPECTED) + B x AGE

    If the number of observed is essentially the same as the number expected, no matter what the age, then age has no effect, so it appears that B is zero. If the number of observed gets smaller and smaller than what is expected as age increases, then it appears that B is less than zero. The test of whether B is equal to zero or not is a hypothesis test about the effect of age. To test a hypothesis about this particular B, we fit a model with and without age and examine how much more variation is explained by the model with age than the one without it. The Poisson distribution lets us know just how large "more" needs to be for us to make a decision: if the variation accounted for by age is larger than would be expected by chance, then we conclude that B is not really zero. The whole process is called a likelihood ratio test (LRT).

    One major advantage of Poisson regression over simple tabulation of SMRs by category is that it allows simultaneous analysis of many more variables. A full cross-classification of even three variables with five categories each would yield 125 sets of observed and expected. Only the most common causes of death in large cohorts would provide enough observed to give sizeable numbers in each cell of a 5 x 5 x 5 table. Poisson regression is not limited to the same extent by the sparsity of data in a multi-way table.

    This report used Poisson regression in 2 ways:

    (i) To test hypotheses about variables of interest, namely age, ethnicity, year of hire and duration of employment.

    (ii) To investigate the roles of other variables that may be acting as confounders or independent risk factors, namely calendar year, age at hire and employment status.

    A variable was deemed significant in (i) or kept in the model in (ii) if the p-value was less than 0.05 for the likelihood ratio test of the models with and without the variable in question.

    The best Poisson regression model for this data, was one in which all the variables were significant and which omitted no significant variables from among the ones listed above in section 2.4.

    The S-Plus software package (Statistical Sciences, 1993) was used to fit the regression models described in this report.

    4 Results

    4.1.1 Distribution of Person-Years

    Table 5 shows the distribution of person-years at risk by age and calendar period for those miners hired before 1945 and those hired in 1945 and later. Any comparisons between these two groups are comparisons of groups with entirely different age structures.

    Table 5: Distribution of Person-Years by Year of Hire, Age and Calendar Period.

    Hired in 1944 and before
    AGE      1955-59 1960-64 1965-69 1970-74 1975-79 1980-84 1985-87   Total
    35-39       2296    1031     105       0       0       0       0    3432
    40-44       4204    2257    1004     103       0       0       0    7568
    45-49       4863    4106    2208     970      96       0       0   12243
    50-54       5505    4660    3985    2114     938      89       0   17291
    55-59       4688    5234    4390    3737    2009     895      76   21029
    60-64       2228    4369    4781    3981    3366    1819     455   20999
    65-69        904    2005    3775    4113    3426    2998     757   17978
    70-74        296     755    1620    3078    3388    2865    1144   13146
    Total      24984   24417   21868   18096   13223    8666    2432  113686
    Hired in 1945 and later
    AGE      1955-59 1960-64 1965-69 1970-74 1975-79 1980-84 1985-87   Total
    35-39       7085   11282   13186   13210    9390    6743    2489   63385
    40-44       4044    7869   11687   13364   13252    9508    2826   62550
    45-49       1993    4409    8047   11735   13280   13220    4452   57136
    50-54        722    2120    4526    7927   11477   12994    5361   45127
    55-59        373     779    2164    4394    7548   10988    4992   31238
    60-64        145     391     785    2043    4046    7031    3826   18267
    65-69         77     143     366     701    1818    3617    2035    8757
    70-74         20      77     121     312     583    1561    1089    3763
    Total      14459    27070  40882   53686   61394   65662   27070  290223

    Since the early-hire cohort cannot recruit new members, it steadily ages over the period of the study and many members reach the age of 75 before the end of 1986. The age distribution of the post-1944 hires changes to a much smaller degree over time. Coupled with a declining relative risk with age, this difference in age structure means that a simple comparison of SMRs by era of hire may be seriously confounded by age.

    4.1.2 Overall summaries

    There were 403909 person-years at risk, giving rise to 112 deaths from stomach cancer. 82.0 were expected, giving an overall SMR of 1.37 (95% C.I.: 1.13-1.64).

    Employment was often not continuous between first and last dates in gold, with 34% of miners having at least a one-year absence from mining during their careers. The average length of employment within a calendar year for men who mined only gold within that year was 10.5 months. For those with more than one type of mining experience within a year, the average was considerably less. A miner's total duration of gold-mining divided by the time between first and last gold mining gives the proportion of this time spent mining gold. Overall, this proportion was less than 80% in about 60% of miners. Clearly, the measures taken to carefully calculate duration of empployment were justified.

    As mentioned above, further results will concentrate on the Poisson regression analysis.

    4.2 Poisson Regression Results:

    Active versus terminated status was a big determinant of mortality, as were age at hire and age at risk. These factors were always statistically significant alone or in combination with other variables, so are included in all regression models presented here.

    There was no statistically significant difference between using three age-at-hire groups and using a linear term (straight line) through the midpoints of each age category. Similarly, there was no significant difference between using 8 age groups and using a linear term through the midpoints of each age group. The LRT X2 for linear terms versus group terms was 3.06 on 7 degrees of freedom (p=0.86). For simplicity's sake, a linear term was used for each of these two variables. The base model, with p-values from LRTs for the model with and without each term is shown in Table 6

    Table 6: Base Model in Poisson Regression
    Variable Rel. Risk LRT X2 df p-value
    Age 0.95/Yr 15.4 1 < 0.0001
    Employment Status (Inactive vs Active) 3.72 27.1 1 <0.0001
    Age at Hire 1.045/Yr 9.2 1 0.0024

    Next, duration of employment was added to this model, first as a six-level categorical variable, then as a linear term with values at the midpoints of the duration categories. The p-value for the difference between these was 0.75 (X2 = 1.9 on 4 degrees of freedom), so a linear term was added. In this new model (Table 7), all terms were still significant.

    Table 7: Base Model Plus Duration of Employment
    Variable Rel. Risk LRT X2 df p-value
    Age 0.94/Yr 19.6 1 <0.0001
    Employment Status

    (Inactive vs Active)

    4.5 31.7 1 <0.0001
    Age at Hire 1.058/Yr 12.9 1 0.0003
    Duration of Employment 1.026/Yr 4.56 1 0.0326

    When year of hire was added to this model, the variables in Table 7 remained significant, but although the post-1944 hires had a slightly higher risk than the pre-1945 hires, the effect was not statistically significant (LRT X2=2.6, p =0.1074 ).

    When calendar year was added to this model the variables in Table 7 remained significant, but difference in risk according to year was not signifcant (LRT X2=6.6 on 6 degrees of freedom p =0.35, for year as a group variable). There was no discernible pattern of risk with the calendar periods, and there was certainly no linear trend (LRT X2=0.05, on 1 degree of freedom, p=0.82 for year as a linear term ).

    When country of birth was added to this model the variables in the Table 7 remained significant, but the higher risk among those born outside North America was not statistically significant(LRT X2=0.55 on 1 degree of freedom , p=0.46).

    Among these 4 variables, there are 6 possible 2-way interactions, such that the level of one variable affects the behaviour of the other. Only one interaction was statistically significant: age and duration of employment interact in such a way that younger miners show a larger increased risk with duration of gold mining than do older miners.

    The final model for this data is shown in Table 8, below. The "N/A" entries are a logical necessity: there is no single relative risk for age or for duration of employment, since they interact as outlined above. Furthermore, since the interaction describes the joint effect of age and duration beyond their individual effects, the individual effects must be in the model to establish a baseline. It does not make sense to test the model below against a model without age or duration so no LRT statistics are given.

    The age-duration relative risk is given in units of risk per year per year. The interpretation is that for each 1 year increase in age, the risk per year of gold mining goes down by a factor of 0.996. The actual risk for any age-duration combination is calculated from the regression model equation.

    The age-at-hire relative risk is 1.06 per year, so that a miner who started work ten years later than another would have an estimated (1.06)10 =1.79 times the risk of stomach cancer. Inactive workers have an enormous 5.57 times the risk of active workers.

    Table 8: Final Regression Model
    Variable Rel.Risk LRT X2 df p-value
    Age N/A N/A N/A N/A
    Employment Status

    (Inactive vs Active)

    5.57 38.5 1 < 0.0001
    Age at Hire 1.06/Yr 13.8 1 0.0002
    Duration of Employment N/A N/A N/A N/A
    Age-Duration 0.996/Yr/Yr 8.7 1 0.0032

    5. Discussion

    5.1 Employment Status: Only 14 deaths occurred among active workers, workers who had some mining experience in their calendar year of death. The other 98 occurred among workers who were inactive in their calendar year of death; the inactivity, as outlined in section 2.4.7, could have happened for a number of reasons.

    Table 9, below, shows that SMR among inactive workers is higher under the age of 65 than over the age of 65. Over age 65, both the reference population and the inactive miners are more alike, in that they are both largely made up of those who have retired due to social convention, so the SMR is near 1. Under 65, the inactive miners are a mix of those who chose work in other jobs and those who were not able to continue mining. Compared to the average under 65 population, which probably has a larger proportion of workers, these younger non-mining workers experience higher stomach cancer mortality.

    Table 9: SMRs by Employment Status and Retirement Age
    Age at Risk
    Employment Status <65 > 65 Total
    Inactive O=65 E=28.8


    O=33 E=28.9


    O=98 E=57.7


    Active O=12 E=20.66


    O=2 E=3.60


    O=14 E=24.26


    Total O=77 E=49.46


    O=35 E=32.5


    O=112 E=82


    Of the 98 deaths among inactive workers, a full 40 occurred within the two years after cessation of mining of any kind, suggesting that in many cases, illness may have been a cause of retirement. The phenomenon of high mortality in the first few years after retirement is well-recognized. While it would certainly artificially inflate the relative risk due to inactive employment status, it is not clear exactly how it would affect the use of this variable to address the issue of the healthy worker survivor effect. This may be worthy of further study.

    5.2 Age at Hire: That there is an increasing relative risk with age at hire into gold-mining is indisputable. What is uncertain is whether it represents an actual effect of later onset of exposure or whether it represents some other characteristic of those who start gold mining at a later age. The groups (15-23/23-32/33+) were chosen on the basis of percentiles of the data, not on any physiological basis, but they do precede the ages at which stomach cancer becomes more prevalent by 20 to 30 years, so provide ample latency between hire and onset of cancer.

    5.3 Duration of Employment/Age at Risk: It is well-established that younger gold miners in this cohort experience higher risk of stomach cancer than older gold miners. This analysis extends this finding to show that the positive association between risk and duration of gold-mining is higher for younger miners than for older ones. Where absolute risk is highest - in the older miners- relative risk is lowest.

    Breslow and Day (1987) suggest that before including such interactions in a model, the specification of the model be further examined. In particular, they suggest an additive risk (rather than relative risk) model might be a better fit. Consider the situation where the additional risk from exposure to gold is added to the baseline risk, rather than multiplied by it. At a low baseline rate, 10 say, an additive element R will give a relative risk (R+10)/10 = 1 + R/10. At a high baseline risk, 100 say, the same element R will give a relative risk of (R+100)/100 = 1 + R/100. This would give the illusion of a decreasing risk with age, when in fact what happened is that a true additive risk relationship was wrongly analyzed with a relative risk model. Unfortunately, the available software was not able to analyze this dataset with an additive risk model due to the sparsity of the data.

    The effect of the interaction is shown in Table 10, below. Miners with 20-25 years and 0-5 years of gold-mining are compared at various ages. All other factors are held constant and the findings are presented as relative risks for the two duration-of-employment groups (not SMRs for comparison to the population).

    Table 10: Decreasing Relative Risk for Gold Mining With Increasing Age
    Age at Risk Relative Risk

    for 25 vs. 2.5 Years of Gold-Mining

    40 11.41
    50 5.03
    60 2.22
    70 0.98

    It should be remembered that the terms in age and duration are linear, so that in the extreme, beyond the range of the data in either direction, the relative risk will get extremely large and extremely small. On the basis of the data, it was not possible to distinguish between this simple model and a more reaslistic one in which risk does not continuously decrease with increasing age.

    6. Conclusions

    .The major findings of this analysis can be qualitatively presented as below:

    (1) After controlling for employment status and age at hire, duration of employment appears to be more strongly associated wth stomach cancer at younger ages than at higher ages.

    (2) As age at first gold-mining increases, risk for stomach cancer increases.

    (3) In a statistical model that includes age at risk, age at hire, duration of employment and employment status, there is no effect of calendar time, year of hire, or being born outside of North America.

    Quantitatively, it would be a mistake to use the model developed in this report to pick an age at which gold mining does not increase risk of stomach cancer. Risk is modeled to decline continuously with age, so there is no single age which separates high-risk workers from low-risk workers. While this is the nature of the linear terms in the regression model, it is most likely close to the truth in that continuous variables like age probably act continuously.

    Further work in the area of the healthy worker survivor effect is needed to determine exactly when it is not appropriate to stratify on the basis of employment status and what effect this stratification has when work-leaving appears to be caused by illness.


    Arrighi, HM, Hertz-Picciotto, I . The evolving concept of the healthy worker survivor effect. Epidemiology, 1994; Vol. 5, Number 2.

    Breslow NE, Day NE. Statistical Methods in Cancer Research. vol 2. The Design and Analysis of Cohort Studies. IARC Scientific Pub. No. 82. Lyon: International Agency for Research on Cancer, 1987

    Checkoway H, Pearce N, Crawford-Brown DJ. Research Methods in Occupational Epidemiology. Oxford University Press. Oxford, 1989.

    Fox, AJ, Collier PF. Low mortality rates in industrial cohort studies due to selection for work and survival in the industry. Br J Prev Soc Med 1976; 30:225-230

    Kusiak RA, Ritchie AC, Springer J, Muller J. Mortality from stomach cancer in Ontario miners. Br J Ind Med 1993; 50 (2): 117-126.

    Muller J, Wheeler WC, Gentleman JF, Suranyi G,Kusiak RA Ritchie AC. Study of mortality of Ontario gold miners 1955-1977, Part 1. Toronto: Ontario Ministry of Labour, Ontario Workers' Compensation Board, Atomic Energy Control Board, 1983.

    Muller J, Kusiak RA, Suranyi G, Ritchie AC. Study of mortality of Ontario gold miners 1955-1977 addendum. Toronto: Ontario Ministry of Labour, Ontario Workers' Compensation Board, Atomic Energy Control Board, 1987.

    Park RM, Maizlish NA,Punnett L, Moure-Eraso R, Silverstein MA. A comparison of PMRs and SMRs as estimators of occupational mortality. Epidemiology 1991; 2(1)49-59.

    Robins, J A graphical approach to the identification and estimation of causal parameters in mortality studies with prolonged exposure periods. J Chron Dis (suppl 2) 1987; 40:139S-161S

    Steenland K, Stayner L. The importance of employment status in occupational cohort studies. Epidemiology 1991; 2(6) 418-423.

    Steenland K, Deddens J, Salvan A, Stayner L. Negative bias in exposure-response trends in occupational studies: modeling the healthy worker survivor effect. Am J Epidemiol 1996; 143(2) 202-210.

    August 29, 1996

    Mr. Glen Wright
    Workers' Compensation Board
    200 Front Street West, 18th floor
    Toronto, Ontario
    M5V 3J1

    Dear Mr. Wright:

    I enclose a copy of the Panel's "Report to the Workers' Compensation Board on Stomach Cancer in Ontario Gold Miners."

    I would be pleased to discuss the Report with you. Please let me know when it would be convenient to do so.


    Nicolette Carlan



    1. Then named the Industrial Disease Standards Panel (IDSP). Both are used interchangeably in this report.

    2. Atrophic gastric mucosa, especially when associated with intestinal metaplasia, probably increases the risk of gastric cancer. The resultant loss of gastric acidity from the atrophy may encourage the presence of certain bacteria which can facilitate the formation of carcinogenic compounds in the stomach (27,52).

    3. Throughout this report, all SMRs are expressed with a 95% confidence interval, unless otherwise noted.

    4. Throughout this report, pre-1945 refers to hires before January 1, 1945. Post-1945 refers to hires during 1945 and after. The pre and post 1945 distinction originated with Muller et al., 1986, who were examining lung cancer risks.

    5. Ames and Gamble empirically tested the Meyer hypothesis by conducting a case-control study on 46 coal miners who died from stomach cancer. Their results did not confirm the hypothesis. They subsequently postulated a revised version which incorporated some of Meyer's original theories (2).