Birk, Thomas Dipl rer soc; Mundt, Kenneth A. PhD; Guldner, Karlheinz PhD; Parsons, William MS; Luippold, Rose S. MS
Chronic inhalation of dust containing sufficient concentrations of free crystalline silica has been linked for decades with disabling lung diseases such as silicosis and silico-tuberculosis. Studies also have suggested that crystalline silica may increase risk of lung cancer as well as some other diseases. Exposure to crystalline silica has long been recognized as a significant health hazard for workers engaged in mining, quarrying, stone cutting and crushing operations, sandblasting, foundry works, construction, ceramics manufacturing and other occupations where significant dust exposures have occurred. Risks to workers appear to vary according to quantity and type of crystalline silica polymorph, with greater risks associated with freshly fractionated quartz and with specific crystalline structures such as crystoballite.1
In 1997, the International Agency for Research on Cancer (IARC) found “sufficient evidence” in both humans and in animal studies to classify occupational exposure to crystalline silica in the form of quartz or crystoballite as carcinogenic in humans.2 However, carcinogenicity in humans was not evident under all industrial exposure circumstances. Risk appeared to be stronger among individuals previously diagnosed with silicosis, raising a basic question of whether lung cancer risk was also elevated among individuals without silicosis, including those less heavily exposed. The scientific debate continues well after the 1997 IARC evaluation,3–5 and newer research reports and reviews have generated additional, divergent evidence and conclusions.6–11
In addition to the relationship between crystalline silica exposure, silicosis and lung cancer risk, several recent studies have suggested relationships between crystalline silica exposure and non-malignant renal disease (NMRD) and kidney cancer.6,12–18 To date the total number of cases studied has been relatively small, precluding firm conclusions regarding their association with crystalline silica exposure.
In this paper, we report on the mortality experience of nearly 18,000 German porcelain workers potentially exposed to crystalline silica and participating in a medical surveillance program in the mid-1980's that included periodic x-ray evaluations as well as nearly complete documentation of employment history.
The German Porcelain Industry
From the 18th century until around 1920, large porcelain factories were constructed in Bohemia, Thuringia, Upper Franconia and Upper Palatinate with a basic technology that did not change substantially over the years. In 1929 in Germany, the 2nd Ordinance on Occupational Diseases made “Severe Silicosis” in combination with jobs in specific industry sectors, a compensable occupational disease.19 Among qualifying industries was porcelain manufacturing, where silicosis was known as “porcelain workers' disease.” By 1938 the Berufsgenossenschaft der keramischen-und Glas-Industrie (BGGK)—the official institution for worker accident insurance/compensation and prevention in the ceramics and glass industry—implemented systematic preventive medical checkups for early detection of silicosis among dust-exposed employees of its member companies, which since that time has been mandatory for all member companies. The results of preventive medical checkups have been stored electronically since the mid-1980s, when participation in the preventive medical program by blue-collar workers had reached approximately 85%.
In addition, since 1951 but more frequently beginning in the 1960's, the BGGK conducted numerous dust measurements among their member companies. Since 1972 and until 2006, the occupational exposure limit (OEL) in Germany had been 0.15 mg/m3 Time Weighted Average as respirable silica. In 2006, however, many established OELs in Germany were suspended—including the limit for crystalline silica—if they were not based on health data as required by the 2005 German Ordinance on Hazardous Substances.
The cohort of German porcelain industry production employees, which has never been evaluated epidemiologically, presents a rare opportunity to evaluate health effects of occupational exposures to crystalline silica generally (but with clear exceptions) around or below the OEL and in contrast with much higher levels associated with porcelain production in the first half of the 1900's. Furthermore, half of the available study population are women, allowing for many causes of death including lung cancers evaluation of silica-related mortality risks among women.
This paper, the first of several investigations underway on this cohort, presents the overall study approach as well as the results of a basic standardized mortality ratio (SMR) analysis. For this report, we focus on diseases of recent interest including lung cancer, silicosis, renal disease, and kidney cancer. Because silicosis often is not fatal, a separate evaluation of silicosis risk will be presented in a subsequent paper, based on the x-ray film evaluations of the cohort over time, and incorporating a quantitative exposure assessment based on the extensive IH data available. However, silicosis mortality results are considered in this report.
The source population for this study consisted of employees from more than 100 plants of the German porcelain and fine ceramics (hereafter referred to as “porcelain”) industry. Most of the porcelain manufacturing plants are located in two regions of the German State of Bavaria—Upper Palatinate and Upper Franconia—mainly due to the proximity to large kaolin deposits in these areas. Kaolin, along with quartz and feldspar, are the main ingredients of porcelain products. Other porcelain plants that are included in this study are scattered throughout the Western States of Germany. Because the BGGK conducted examinations as well as exposure measurements in the plants in the Eastern States of Germany only after reunification in 1990, the study population is restricted to employees of plants located in the Western States (former West Germany).
The study population was initially defined as all employees who participated in the preventive medical surveillance program for early identification of silicotic signs between January 1, 1985 and December 31, 1987, according to the German regulation “Berufsgenossenschaftlicher Grundsatz G 1.1 ‘Quarzhaltiger Staub.'” This time window was chosen for two reasons. First, the BGGK maintains an electronic database on all preventive medical checkups since 1985, including demographic information and the results of radiological examinations according to International Labor Organization (ILO) criteria (1980). Second, the German Ordinance for Hazardous Substances required employers to conduct preventive medical examinations of employees exposed to quartz (ie, crystalline silica) dust. Under these rules, if a threshold level for a workplace hazardous substance has been exceeded in a workplace subject to this ordinance, employees are only allowed to work in such workplaces if they undergo special periodic preventive medical examinations. For crystalline silica, medical examinations are required every 3 years. Therefore the 3-year time window of 1985 to 1987 was selected to facilitate identifying nearly all exposed porcelain workers at this time.
Not included in the electronic database of the BGGK, however, is information regarding the employees' work history (except the specific plant where the employee worked at the time of the medical examination), smoking habit, medical examination results (apart from the radiological results) and home address. Much of this information is recorded in the paper records maintained by the BGGK. Based on the electronic database, we identified 20,039 employees of the German porcelain industry, of which 48% were males and 52% were females.
In retrieving the paper records for the cohort, it was determined that records were not available for all members of the study population. Due to storage space limitations, the general policy of the BGGK was to destroy paper records of persons older than 75 years, provided they had no signs of silicosis (defined as ILO 1980 1/0 or greater) as of their last examination. However, this rule had not been consistently applied. When the cohort was created early in 2005, paper records were located for 408 (32%) of the 1295 members of the study population who were older than 75 years at that time. Among those with records were 88 persons with some sign of silicosis based on the BGGK x-ray film evaluations. Additionally, folders for a few companies which closed many years ago were destroyed, and some missing files are believed to have been destroyed accidentally. Therefore, we redefined the eligible cohort as those employees for whom work histories were available from paper records (N = 18,000). Further exclusion of individuals employed for less than 6 months (cumulative) in the porcelain industry resulted in a final analytical cohort of 17,644 employees. This population was followed through December 31, 2005 for mortality (and for silicosis morbidity based on periodic x-ray films, to be published separately). Because the majority of porcelain plants are located in the German State of Bavaria, we defined a sub-cohort of Bavarian porcelain workers comprised of 15,045 employees (85.3% of the Full Cohort). Figure 1 summarizes the numbers in each cohort grouping, as well as the Full Cohort and the Bavarian Subcohort by sex, vital status, and person-years of follow-up.
Vital Status Ascertainment and Determination of Cause of Death
Several available data sources were consulted to determine the vital status for each cohort member, including the BGGK internal information as well as company health insurance and pension fund data. However, most of the vital status information was determined through direct written enquiries to the community registration offices for the location in which each employee was last known to reside, or by searching a central registration database for Bavaria. Vital status of the cohort ultimately was determined for 94%, with 1610 (9%) of the cohort members reported to be deceased. For the Bavarian Subcohort, follow-up was truncated at age 75, as many records on cohort members over age 75 and with no signs of silicosis had been destroyed. Approximately 50% of employees from the original study population who were 75 years of age or older at the end of the follow-up period had no paper record available and therefore are not part of the analytical cohort. We investigated the possible influence on the risk estimates due to restricting follow-up to the age of 75 years. The result of this subanalysis showed that very few cohort members contributed person-years at risk in the ages older than 75 (0.3% of all person-years) leading to only a very small increase in the SMRs for this population. We subsequently used this cohort with follow-up truncated at age 75 for all further analyses of the Bavarian Subcohort.
Subjects with unknown vital status at the end of the follow-up period contributed person time up to the date last know to be alive. Although quantitative exposure estimates are being developed, some basic surrogates for higher exposure were derived, such as decade of hire and hire before or after the end of 1960, given that earlier years in this industry were dustier; and ever having worked in the preparation area, as this is the area with greatest potential exposure historically to respirable crystalline silica. Follow up time for individual subjects was divided into single calendar years and 5-year age categories to correspond with the available mortality referent rates.
The underlying cause of death was sought on each decedent from the official death certificate, which in Germany usually is stored at the community health department in the town or city where the death occurred. However, minimum record retention times differ from state to state in Germany, ranging, for example, from 30 years in Bavaria to only a few years in other States such as Rhineland-Palatinate. Therefore, cause of death determination based on death certificates was more complete for the Bavarian Subcohort (95%) compared with the Full Cohort (93%). For those cohort members who died in Bavaria between 2000 and 2005—approximately one third of all decedents—the official cause of death code was directly provided by the Statistical Office of Bavaria, according to the tenth revision of the International Classification of Diseases (ICD-10).20 All other death certificates were coded by a professional nosologist from an official German State Statistical Office, also according to the ICD-10.
A detailed employment history was reconstructed for all cohort members, using available files at the BGGK. Files included records of employment before starting work in a porcelain plant, including dates, name, location, and type of plant as well as the specific job. Year of hire, department and/or job in the porcelain plant, and the respective start date of the job were recorded in the forms documenting the initial and all follow-up medical surveillance examinations. Therefore, all potential changes of job or department over time could be reconstructed. Using these data and historical industrial hygiene measurements, we are currently developing quantitative Job-Exposure Matrices for use in dose-response analysis. For this report, however, detailed work history data were used to derive indicators of exposure differences like duration of employment, decade of hire (to evaluate potential for survival bias among cohort members hired before the start of follow-up) and ever having worked in the preparation department. This department is the area believed to have the highest consistent crystalline silica exposure potential, and which, until recent years, consistently had average monitored exposures exceeding the former German OEL of 0.15 mg/m3 as a Time Weighted Average of respirable crystalline silica.
Cause-specific mortality patterns were examined by calculating SMRs for all-cause and selected groupings of cause of death, comparing observed number of deaths in the cohort with expected numbers based on two sets of reference rates. For the Full Cohort, rates for the former Western Germany population (1985–1997) and the total German population (1998–2005) were derived based on mortality and population numbers made available by the Statistical Office of Germany. A second set of reference rates was derived based on the Bavarian population (numbers of deaths and population counts were provided by the Statistical Office of Bavaria). Higher all cause, all cancer, and CVD mortality rates in Upper Palatinate and Upper Franconia, the two sub-regions in which a majority of Bavarian Subcohort members worked, compared with other Bavarian regions are reported in official Bavarian health status reports.21 Because of the greater variability in rates for these sub-regions (especially for rarer causes of death), however, we decided to present SMR results based on the Bavarian referent only. Rates for both reference groups were calculated by gender, 5-year age group (from 15 to 84 and 85+) and calendar year, although analyses of the Bavarian Subcohort were limited to the cohort experience up to age 75.
For the Full Cohort, SMRs and 95% confidence intervals (95% CIs) were calculated for all causes combined and approximately 40 different cause-of-death categories (Appendix). A subset of causes of death (principally those reflecting the key study research questions) was examined for the Bavarian Subcohort. All analyses were conducted using SAS v9.1 (SAS Institute Inc, Cary, NC). We calculated exact CI's using Byar's approximation.22 Because of considerable differences in the risk and distribution of causes of death, and because of the large number of women included in this cohort, SMR analyses are stratified by gender.
For the Bavarian Subcohort, SMRs stratified by date of hire before/after 1960 were estimated to evaluate the potential impact survival bias might have on the cohort results. Because follow-up of the cohort started on January 1, 1985 at the earliest, those starting employment many years prior to this had to survive until the start of follow-up, and the patterns of mortality among this subset may be different from those hired closer to the start of follow-up and not subject to (or at least as strongly to) potential survival bias. We also stratified Bavarian Subcohort results by ever/never having worked in the preparation area—a surrogate for the highest likely exposures to crystalline silica. No further stratification by work area or estimated exposure to crystalline silica was performed at this stage, as a detailed quantitative exposure assessment is underway.
The possible influence of the record retention policy of the BGGK on the risk estimates was evaluated two ways: by excluding those cohort members aged 75 years or more on January 1, 2005 and with signs of silicosis; and by excluding all cohort members older than 75 years as of this date.
The full cohort of 17,644 workers employed in porcelain production for at least 6 months generated 338,495 person-years of observation at risk, 156,713 person-years among men and 181,782 person-years among women. Figure 1 shows the numbers of individuals, person-years observed and numbers of death for the Full Cohort and the Bavarian Sub-Cohort, by sex. Bavarian men comprised approximately 84% (by count and person-time up to age 75) and Bavarian women approximately 86% (by count and person-time up to age 75) of the Full Cohort. Truncating follow-up at age 75 for all in the Bavarian Sub-Cohort resulted in the loss of only 22 deaths (and 496 person-years) among men and 6 deaths (and 301 person-years) among women, so only the truncated Bavarian Sub-Cohort results are presented.
The average age at start of follow up for men was about 35 and about 34 for women (Table 1). At the end of follow-up, 14,935 (85%) were known to be alive, 1610 (9%) were determined to be deceased, and vital status remained unknown for 1099 (6%) (Table 2). Among the 1126 male decedents (13.6% of men), all cancers accounted for 32.1% of all deaths, all circulatory system diseases 32.9%, all respiratory diseases 5.2% and other causes of death 29.8%. Among the 484 female decedents (5.2% of women), all cancers accounted for 39.3% of all deaths, all circulatory system diseases 25.8%, all respiratory diseases 3.7% and other causes of death 31.2%. Follow-up status as well as numbers of deaths in each of these broad categories, by full or Bavarian sub-cohort are presented in Table 2. Duration of employment was similar for both the full and Bavarian groups (Table 1).
Table 3 presents SMR estimates for the full cohort, by sex, for 39 categories of cause of death, using the German population as the referent. The SMR estimate for all causes among men was not statistically different from unity. However, we observed a significantly decreased SMR for women (0.85, 95% CI = 0.78 to 0.93), compatible with a Healthy Worker Effect. Significantly elevated SMRs for liver and pancreatic cancer, and for silicosis, (1.99, 95% CI = 1.29 to 2.93; 1.71, 95% CI = 1.18 to 2.41; 7.20, 95% CI = 2.32 to 16.8) were estimated for men. Only the SMR for diabetes was significantly elevated among women (1.74, 95% CI = 1.07 to 2.65). We observed no statistically significant increased risk of mortality due to lung cancer, renal cancer, or NMRD (identified a priori as causes of interest); however, the point estimate for renal disease among men was elevated (1.32, 95% CI = 0.63 to 2.43).
Using the Bavarian sub-cohort and the Bavarian population as referent resulted in increases for the SMRs calculated for broad categories of cause of death as well as specific causes of interest (Table 4), suggesting that Bavarian mortality rates are generally lower than comparable German rates. The SMR estimate for lung cancer increased from 0.71 to 0.98 (95% CI = 0.75 to 1.27) for men, and from 0.72 to 0.91 (95% CI = 0.52 to 1.48) for women. SMRs for circulatory system diseases increased from 1.00 to 1.17 (95% CI = 1.05 to 1.31) for men, and from 0.83 to 1.03 (95% CI = 0.85 to 1.23) for women. The estimate for renal disease among men remained elevated, but not significantly so (1.65, 95% CI = 0.75 to 3.14).
Analyses were repeated, first excluding the 88 cohort members with signs of silicosis and those older than 75 years at beginning of the study in 2005, and second by restricting the cohort to those younger than 75 years. No remarkable changes in any of the risk estimates were seen (results not shown) except for silicosis, for which the number of silicosis death was reduced from 5 to 3. This led to reduced SMRs for silicosis mortality as follows: Full cohort: SMR 4.67 (95% CI = 0.94 to 13.66) and SMR 6.41 (95% CI = 1.29 to 18.74) for each alternative analysis, respectively; Bavarian sub-cohort: SMR 6.95 (95% CI = 1.40 to 20.32) and SMR 7.95 (95% CI = 1.60 to 23.23), respectively.
When comparing Bavarian men hired through 1960 to those hired after 1960, the excess risk of death due to cancer and respiratory disease appeared to be confined to those hired in the earlier period (Table 5). SMRs for both time periods were 1.24 (95% CI = 1.06 to 1.44) and 1.02 (95% CI = 0.85 to 1.23) for all-cancers, and 1.38 (95% CI = 0.94 to 1.96) and 1.06 (95% CI = 0.61 to 1.73) for respiratory diseases, respectively. Among Bavarian women, the point estimate for COPD appeared to be considerably higher among those hired in the later periods: 1.20 (95% CI = 0.32 to 3.07) versus 2.12 (95% CI = 0.78 to 4.62), although the wide confidence intervals indicate these estimates were not very precise. SMR estimates for lung cancer were similar for those hired through 1960 versus those hired after 1960 and close to unity, for both men and women. All five silicosis deaths (plus additional three which were coded as contributing cause of death) were among men in the Bavarian sub-cohort hired before the end of 1960, resulting in a very high and statistically significant but unstable SMR.
Although no excess of lung cancer deaths was identified, we stratified results by ever having worked in the materials preparation department—the work area known to have the highest potential crystalline silica exposures over time—to determine whether elevated risks might be obscured by combining these workers with those in other work areas. No deaths for any of these selected causes were observed among women ever having worked in preparation, who generated only about 6% of the person-years among women. Table 6 demonstrates that among men, the SMR for lung cancer is not elevated in the subgroup ever having worked in the preparation area. However, SMRs for all causes combined, all cancers and all circulatory system diseases were higher (not significantly) for the preparation workers than for those never working in this area. Interestingly, three of the five silicosis deaths occurred among men in this subgroup, two of whom were part of the 88 individuals whose records were retained beyond age 75 due to x-ray film evidence of silicosis.
This paper presents the first results of a large epidemiological study of workers in the modern (ie, post-WWII) German porcelain industry. Included were a large majority of employees from nearly all manufacturing facilities in Western Germany operating during 1985–1987, the time window in which the study cohort was defined due to the availability of electronic information and the triennial individual occupational medical examinations. The study population was identified by mandatory (those with potential for some level of exposure) and voluntary (those with potential for only low level of exposure) participation in the BGGK preventive medical surveillance program, and therefore is not representative of workers with only very low levels of crystalline silica (white-collar workers, decoration workers). However, we have determined that approximately 85% of all blue-collar employees of the porcelain manufacturing facilities at this time participated in this program. Therefore, the study population is likely to be highly representative of all porcelain production workers in Germany.
There are several noteworthy advantages provided by this study cohort. First, this study group represents one of the largest cohorts of porcelain (including ceramics or pottery) workers or any other type of silica exposed workers, with over 8200 men and over 9300 women. Other epidemiological cohorts have tended to include smaller numbers of men only. For example, the recent studies of British and Chinese pottery cohorts included 511523 and 454724 men, respectively. Earlier cohort studies have included 2055 white men employed in three ceramic plumbing fixture factories in the US25; 2480 men employed in the ceramics industry in Italy26; and 1794 male ceramic workers in the Netherlands.27 In fact, other than from China, we are unaware of any cohort study including substantial numbers of women employed in the porcelain industry.
Second, because the cohort was defined based on participation in the BGGK preventive medical surveillance program, all cohort members have a “baseline” x-ray film examination made before starting work in the porcelain industry and most cohort members have many follow-up x-ray film evaluations through the end of the individual follow-up. Although not taken into account in this preliminary SMR analysis, the x-ray films can demonstrate the presence or absence of lung cancers and signs of silicosis at start of follow-up for each cohort member, as well as the date of the first x-ray film demonstrating signs of silicosis or lung cancer. Participation in the program also assures documentation of each cohort member's work history—including job held prior to entering the porcelain workforce—and, for over two thirds of the cohort, information on smoking status. All of these data resources will be considered in subsequent evaluations of the cohort.
Third, this cohort's exposure to crystalline silica overall—in contrast to individuals employed in the first half of the 1900's—is possibly more relevant to modern industrial conditions and related questions regarding health effects at more modest levels of exposure. Specifically, the 1929 Ordinance on Occupational Diseases led to substantial efforts to reduce employees' exposure to quartz dust for the first time (eg, by governmental or BGGK regulations, medical surveillance and dust prevention programs). By 1952, the 5th Ordinance on Occupational Diseases determined that all silicosis cases may be compensated by the BGs. This resulted in a dramatic rise in the number of compensated cases, and fuelled considerable efforts by the porcelain companies and the authorities to reduce exposures by replacement of rough wooden floors, improving dust controls, introduction of ventilation equipment, training employers and employees regarding health risks, monitoring of exposures, and comprehensive and systematic medical surveillance by the BGGK. Between 1955 and 1965 additional substantial technological changes took place, for example, replacement of intermittent (loaded and unloaded) kilns with gas-fired tunnel kilns.28
Although exposures to crystalline silica would have been reduced over time, parts of the present cohort still had the potential for considerable exposure. Depending on specific work area and time period, respirable crystalline silica levels obtained since the 1960's ranged from below detection level to several times the most recent German OEL of 0.15 mg/m3 (as noted above this has been suspended), with most measured concentrations between the detection limit and the OEL.
Where silica exposures are both likely and substantial for at least some work areas and time periods represented by this cohort, there is relatively low potential for exposures to other known or suspected lung carcinogens in the porcelain industry. Crystoballite, a specific structural form of crystalline silica, historically occurred in the firing area. Until 1950 only intermittent kilns (round kilns) were used for firing of the porcelain and fireclay boxes were repeatedly re-used as firing auxiliaries for the biscuit firing. Quartz was partially transformed to crystoballite within these firing auxiliaries and may have generated crystoballite dusts. Measurements in other work areas during the early 1960s did not detect crystoballite. Workplace air measurements recorded in the MEGA database, a comprehensive chemical workplace exposure database maintained by the Institute for Occupational Safety and Health of the German Statutory Accidence Insurance in St. Augustin suggest no or only low levels of several other lung carcinogens. However, historical asbestos exposure cannot be excluded in the area of the kilns where asbestos-containing insulation was used in and around the kilns, on the tunnel kiln cars or associated with other asbestos containing materials and equipment. Therefore, for the silica-related diseases of interest (especially lung cancer and silicosis) the porcelain industry offers an environment in which historical risks associated with crystalline silica, typically in the form of quartz dusts, may be evaluated.
Although the cohort is still rather young, mortality is approaching 10% overall (greater for men in the Bavarian Sub-Cohort), and already more than 90 lung cancer deaths have been observed during the follow-up period. Because of the relatively young cohort and short follow-up, the observed numbers are too low to produce highly stable SMR estimates for many specific causes of death. Kidney cancers and renal diseases had 16 and 14 deaths among the full cohort, respectively, numbers too small to precisely estimate SMRs for relatively small excesses. Small observed numbers were especially clear for silicosis mortality, for which only 5 deaths were observed. However, as part of the evaluation of this cohort, over 500 cohort members have x-ray films that originally were read by the BGGK physicians according to the ILO as 1/0 or higher, ie, with possible signs of silicosis. All x-ray films are currently being re-read by two expert B-readers to verify the original film readings, and the results will be used in conjunction with the exposure assessment to quantitatively evaluate risks of silicosis morbidity associated with the levels of crystalline silica occurring in the porcelain industry since WWII, and risks of lung cancer associated with silicosis.
Based on the current SMR results, the overall mortality patterns observed in the study cohort were similar to both the German national and the Bavarian referent populations, with some exceptions. The slightly higher SMRs obtained when the regional Bavarian reference rates were used suggest that the porcelain workers, although predominately from Bavaria, experience mortality rates for broad categories of cause of death that lie between Bavarian rates and the German national rates (as defined in the Methods section), which might, however, be caused by regional rate variability—leading to variability in SMR results unrelated to occupational exposures.
Nevertheless, although the purpose of this study was to evaluate mortality patterns for specific causes of death previously associated with crystalline silica exposure (lung cancer, silicosis, kidney cancer, and renal disease), we calculated SMRs for many additional categories of cause of death. Although the SMRs for most of these causes of death were not different from unity, some deviations from expected were seen.
Among men, statistically significant excesses were seen for liver/gallbladder cancers as well as for pancreatic cancer, and this did not materially change in supplemental analyses using Bavarian reference rates (not shown). Risk factors for liver cancer include chronic Hepatitis B or C infection, liver flukes, aflatoxin exposure and alcohol consumption, but apparently not crystalline silica.29 Of these, it is possible that the porcelain workers consume higher levels of alcohol, although men did not have an excess risk of cirrhosis of the liver (generally associated with alcohol consumption) compared with German men overall. Liver/gallbladder cancer SMRs for women were slightly but not statistically significantly increased, but this was based on only 10 observed cases.
Risk factors for pancreatic cancer include smoking, diabetes, diet, obesity, and pancreatitis, K-ras gene mutation, and possibly occupational chemical exposures.29 However, Ojajärvi et al30 reported a non-significant meta-relative risk of 1.4 for silica dust in a 2000 review, and two other papers were located in which pancreatic cancer appeared to be associated with dusts containing crystalline silica: a study of Finnish asphalt workers, who would be exposed to numerous compounds in addition to mineral dusts31; and a population-based occupational case-control study of pancreatic cancers in Finland, which generated an odds ratio of 2.0 (95% CI = 1.2 to 3.5) for exposure to inorganic dusts containing crystalline silica.32 Analyses based on the Bavarian sub-cohort and using Bavarian reference rates were similar (not shown), suggesting less variability by region. In future analyses the possible role of tobacco smoking may be evaluated; however, the male porcelain workers mortality patterns (especially for lung cancer and other smoking-related cancers) do not suggest that this group has a disproportionately high smoking prevalence. In contrast, the SMR for pancreatic cancer among women was not elevated.
Among women but not men, the SMR for diabetes mellitus was significantly elevated using German reference rates, and was slightly higher using Bavarian reference rates. No occupational exposures (including crystalline silica) have been consistently associated with diabetes. However, mortality from complications of diabetes might reflect low utilization of diagnostic and preventive medical services, as well as poor case management.
A primary focus of our study was evaluation of lung cancer mortality in this cohort of porcelain workers exposed to moderate levels of crystalline silica. However, based on 94 observed cases (and over 130 expected, based on German rates) there is no indication of an excess lung cancer risk among the German porcelain workers. Based on the Bavarian sub-cohort, the SMRs were closer to unity, but still provided no evidence for an excess rate of lung cancer among porcelain workers. Non-malignant respiratory disease results also were unremarkable.
Research since IARC's 1997 evaluation has not resolved the debate surrounding lung cancer risk in the absence of silicosis and/or at low to moderate levels of exposure. A recent, large multi-country, community based case-control study10 reported increased risk of lung cancer associated with occupational exposure to crystalline silica (OR 1.37, 95% CI = 1.14 to 1.65), and estimates were increased for both men (OR 1.32, 95% CI = 1.10 to 1.59), and for women (2.07, 95% CI = 0.91 to 4.74), although the number of women with lung cancer was small. Risk was elevated for the two upper tertiles of cumulative exposure: OR 1.47 (95% CI = 1.04 to 2.06) for 35 to 200 mg/m3-hrs, and OR 2.08 (95% CI = 1.49 to 2.90) for >200 mg/m3-hrs, suggesting a dose-response. However, this study was not able to identify whether there is a threshold for lung cancer risk. Unknown silicosis status for this cohort precludes any conclusions about lung cancer risk in the absence of silicosis.
Pelucchi et al11 have published an extensive review of the occupational epidemiological literature published since the IARC evaluation. Most of the reviewed studies evaluated cohorts with known silicosis, or undefined silicosis status. The authors concluded that the evidence supports an association with lung cancer among known silicotics and reported pooled RRs from cohort studies of 1.34 (95% CI = 1.25 to 1.45) overall, and 1.69 (95% CI = 1.32 to 2.16) in cohort studies of silicotics only. Only two reviewed studies assessed lung cancer risk in the absence of silicosis. A cohort study by Checkoway et al33 reported an overall SMR of 1.19 (95% CI = 0.87 to 1.57) among non-silicotics, based on 48 cases, and an SMR of 1.57 (95% CI = 0.43 to 4.03) among silicotics, based on only 4 cases. Exposure response analysis, however, showed elevated risk for non-silicotics for the highest tertile of cumulative exposure (≥5 mg/m3-yrs, SMR 2.40, 95% CI = 1.24 to 4.20). The second, a case-control study34 on a cohort of non-silicotic, stone, quarry, and ceramic workers (247 cases, 795 controls) failed to show an association between lung cancer and occupational exposure to crystalline silica. Several exposure measures were analyzed, including peak, time-weighted average, and cumulative exposure, yielding non-significant ORs ranging from 0.85 to 1.02. Pelucchi et al11 concluded that any association between lung cancer and exposure to crystalline silica, in the absence of silicosis, remains unclear.
An additional case-control study on Chinese mine and factory workers,35 published after the Pelucchi review, reported increased risk of lung cancer among pottery workers exposed to silica (OR 3.4, 95% CI = 1.45 to 8.66 for the highest quintile of exposure, 10.1 to 72.4 mg/m3-yr), which disappeared after adjustment for relevant occupational confounders (OR 0.9, 95% CI = 0.19 to 4.32).
Our results do not indicate an increased risk of lung cancer among porcelain workers in Germany, though complete silicosis status was not available. Future plans include analyses incorporating detailed exposure assessment and linkage of workers to x-ray films and medical records indicating first signs of silicosis, as well as the point at which silicosis is diagnosed (as ILO 1/1). This future work should help clarify the carcinogenic role of crystalline silica.
NMRD and Kidney Cancer
The SMRs for renal disease were slightly but not significantly elevated among men, but not for women, based on either the full cohort or the Bavarian sub-cohort. These results, however, are limited by small numbers of observed deaths (n = 14 and n = 12, respectively) in this category. For 24 additional deceased, renal disease was mentioned as contributing cause of death on the death certificate. However, there is no direct way to evaluate this using the SMR approach. It is possible that analyses using internal comparison groups could evaluate whether these cases were among employees more likely exposed at higher levels of crystalline silica than the rest of the cohort.
Several studies of occupational cohorts exposed to crystalline silica have reported excess risks of morbidity and mortality due to NMRD,12–18 whereas some have not.36,37 In a recent report, Steenland12 suggests that kidney disease may present a higher risk than mortality from silicosis or lung cancer, and may be more important for regulatory purposes. He summarizes results from earlier investigations18 reporting an SIR of 1.97 (95% CI = 1.25 to 2.96) for end stage renal disease (ESRD), based on 23 cases identified through the US ESRD registry, as well as a positive exposure-response trend for ESRD (based on 18 cases). Similarly, he reports a strong exposure-response (based on 50 cases) for mortality using data from a 2002 study of three cohorts of industrial sand workers, granite workers, and diatomaceous earth workers.12,16 Another recent report13 also describes excess mortality from renal disease (SMR 2.80, P < 0.001) based on 18 cases in an updated study of 2670 employees of the sand industry. In this study, renal disease mortality was not related to cumulative exposure, though the authors cite small numbers as limiting statistical power.
Kidney cancer SMRs were low for both men and women regardless of which analytical cohort was examined. Again results were limited by small numbers observed, with a total of 16 in the full cohort and 14 in the Bavarian sub-cohort, which resulted in substantial deficits for each gender-cohort grouping (Tables 4 and 5). Renal cancer has much less frequently been associated in the published literature with crystalline silica exposure. Only two recent studies have reported elevated risk estimates. Attfield and Costello6 reported a small, not significant excess of mortality (SMR 1.37) in granite workers. McDonald et al13 reported an SMR of 2.02 (P = 0.03) based on 10 cases, though they characterize the finding as “unforeseen.”
Summary and Conclusions
This cohort study represents one of the largest available to date, particularly noteworthy in that large numbers of men and women potentially exposed to various levels of crystalline silica over several decades have been followed as part of a preventive medical surveillance program offered throughout the glass and ceramics industry. Primary research questions included whether the cohort provides evidence supporting increased risks of lung cancer, silicosis, renal disease, and kidney cancer among porcelain workers potentially exposed to crystalline silica at concentrations from below to well above the former German OEL (ie, 0.15 mg/m3).
Unexpected findings include a small excess of pancreatic cancer among men but not women, and elevated liver and gallbladder cancer risks among men with some suggestion of an increase among women as well. Kidney cancer was not increased in this cohort, and renal disease was only slightly (and not statistically significantly) elevated among men; however, small numbers limited the precision of these estimates.
Based on the first SMR results presented in this paper, it appears that some increased risk of silicosis mortality is seen (but based on very small numbers), especially among porcelain workers first employed before the end of 1960 and among those having worked in the materials preparation area where the highest crystalline silica exposures in the industry consistently have been documented. Despite this direct but preliminary evidence of substantial crystalline silica exposure potential, the mortality experience of this cohort, to date, suggests no excess risk of lung cancer.
The authors thank the invaluable technical assistance of the BGGK, especially including the many employees who spent long hours extracting a tremendous amount of data from paper records and in conducting the vital status and cause of death follow-up. The authors thank many companies, company health insurance firms and pension funds, as well as trade associations, for facilitating our study in various ways including technical assistance and funding.
The authors thank the Scientific Advisory Group, specifically Drs. Lesley Rushton (chair), Peter Morfeld, Dirk Taeger, and Frank Bochmann, all of whom provided helpful suggestions on the study methods, analysis, and on drafts of this report.
The authors also thank the following colleagues who provided valuable scientific advice on various aspects of the study: Leopold Miksche MD; Annette Bachand, PhD; Robert Adams, CIH, CSP; Linda Dell, MS; and Laura Carlton MPH.
The project was sponsored by the Berufsgenossenschaft der keramischen und Glas-Industrie (BGGK), the Steinbruchs-Berufsgenossenschaft (StBG), and by EUROSIL, the European Association of Industrial Silica Producers, with additional support from other trade associations and individual companies.
1. Cocco P. The long and winding road from silica exposure to silicosis and other health effects. Occup Environ Med. 2003;60:157–158.
2. IARC. Silica, Some Silicates, Coal Dust and Para-Aramid Fibrils. Lyon, France: IARC Press; 1997:506.
3. Hessel PA, Gamble JF, Gee JB, et al. Silica, silicosis, and lung cancer: a response to a recent working group report. J Occup Environ Med. 2000;42:704–720.
4. Checkoway H, Franzblau A. Is silicosis required for silica-associated lung cancer? Am J Ind Med. 2000;37:252–259.
5. Soutar CA, Robertson A, Miller BG, Searl A, Bignon J. Epidemiological evidence on the carcinogenicity of silica: factors in scientific judgement. Ann Occup Hyg. 2000;44:3–14.
6. Attfield MD, Costello J. Quantitative exposure-response for silica dust and lung cancer in Vermont granite workers. Am J Ind Med. 2004;45:129–138.
7. Graham WG, Costello J, Vacek PM. Vermont granite mortality study: an update with an emphasis on lung cancer. J Occup Environ Med. 2004;46:459–466.
8. Brown TP, Rushton L. Mortality in the UK industrial silica sand industry: 2. A retrospective cohort study. Occup Environ Med. 2005;62:446–452.
9. Steenland K, Mannetje A, Boffetta P, et al. Pooled exposure-response analyses and risk assessment for lung cancer in 10 cohorts of silica-exposed workers: an IARC multi-centric study. Cancer Causes Control. 2001;12:773–784.
10. Cassidy A, Mannetje A, van Tongeren M, et al. Occupational exposure to crystalline silica and risk of lung cancer: a multicenter case-control study in Europe. Epidemiology. 2007;18:36–43.
11. Pelucchi C, Pira E, Piolatto G, Coggiola M, Carta P, La Vecchia C. Occupational silica exposure and lung cancer risk: a review of epidemiological studies 1996–2005. Ann Oncol. 2006;17:1039–1050.
12. Steenland K. One agent, many diseases: exposure-response data and comparative risks of different outcomes following silica exposure. Am J Ind Med. 2005;48:16–23.
13. McDonald JC, McDonald AD, Hughes JM, Rando RJ, Weill H. Mortality from lung and kidney disease in a cohort of North American industrial sand workers: an update. Ann Occup Hyg. 2005;49:367–373.
14. Calvert GM, Steenland K, Palu S. End-stage renal disease among silica-exposed gold miners. A new method for assessing incidence among epidemiologic cohorts. JAMA. 1997;277:1219–1223.
15. Rapiti E, Sperati A, Miceli M, et al. End stage renal disease among ceramic workers exposed to silica. Occup Environ Med. 1999;56:559–561.
16. Steenland K. Ten-year update on mortality among mild-steel welders. Scand J Work Environ Health. 2002;28:163–167.
17. McDonald AD, McDonald JC, Rando RJ, Hughes JM, Weill H. Cohort mortality study of North American industrial sand workers. I. Mortality from lung cancer, silicosis and other causes. Ann Occup Hyg. 2001;45:193–199.
18. Steenland K, Sanderson W, Calvert GM. Kidney disease and arthritis in a cohort study of workers exposed to silica. Epidemiology. 2001;12:405–412.
19. Guldner K. Entwicklung der Silikose in der keramischen und Glas-Industrie. Gefahrstoffe—Reinhaltung der Luft. 1999;59:189–191.
20. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. (ICD-10) Tabular List. Geneva, Switzerland: World Health Organization; 1992.
21. Bayerisches Landesamt für Gesundheit und Lebensmittelsicherheit. Regionale Unterschiede der Sterblichkeit in Bayern. Gesundheitsberichterstattung für Bayern 1. Erlangen: 2004.
22. Breslow NE, Day NE. Statistical Methods in Cancer Research. The Design and Analysis of Cohort Studies. Vol. 2. Lyon, France: IARC Press; 1987:404.
23. Cherry NM, Burgess GL, Turner S, McDonald JC. Crystalline silica and risk of lung cancer in the potteries. Occup Environ Med. 1998;55:779–785.
24. Chen W, Hnizdo E, Chen JQ, et al. Risk of silicosis in cohorts of Chinese tin and tungsten miners, and pottery workers (I): an epidemiological study. Am J Ind Med. 2005;48:1–9.
25. Thomas TL. Lung cancer mortality among pottery workers in the United States. IARC Sci Publ. 1990;97:75–81.
26. Cavariani F, Di PA, Miceli M, et al. Incidence of silicosis among ceramic workers in central Italy. Scand J Work Environ Health. 1995;21(suppl 2):58–62.
27. Meijers JM, Swaen GM, Slangen JJ. Mortality and lung cancer in ceramic workers in The Netherlands: preliminary results. Am J Ind Med. 1996;30:26–30.
28. Fibir F, Kaufman G. Technische Silikosebekämpfung in der keramischen Industrie. Berufskrankheiten in der keramischen und Glas-Industrie, Heft 15. Würzburg: Hrsg: Berufsgenossenschaft der keramischen und Glas-Industrie; 1963.
29. Schottenfeld D, Fraumeni JF. Cancer Epidemiology and Prevention. New York: Oxford University Press; 2006:1392.
30. Ojajarvi IA, Partanen TJ, Ahlbom A, et al. Occupational exposures and pancreatic cancer: a meta-analysis. Occup Environ Med. 2000;57:316–324.
31. Kauppinen T, Heikkila P, Partanen T, et al. Mortality and cancer incidence of workers in Finnish road paving companies. Am J Ind Med. 2003;43:49–57.
32. Kauppinen T, Partanen T, Degerth R, Ojajarvi A. Pancreatic cancer and occupational exposures. Epidemiology. 1995;6:498–502.
33. Checkoway H, Hughes JM, Weill H, Seixas NS, Demers PA. Crystalline silica exposure, radiological silicosis, and lung cancer mortality in diatomaceous earth industry workers. Thorax. 1999;54:56–59.
34. Ulm K, Waschulzik B, Ehnes H, et al. Silica dust and lung cancer in the German stone, quarrying, and ceramics industries: results of a case-control study [see comments]. Thorax. 1999;54:347–351.
35. Chen W, Bochmann F, Sun Y. Effects of work related confounders on the association between silica exposure and lung cancer: a nested case-control study among Chinese miners and pottery workers. Int Arch Occup Environ Health. 2007;80:320–326.
36. Calvert GM, Rice FL, Boiano JM, Sheehy JW, Sanderson WT. Occupational silica exposure and risk of various diseases: an analysis using death certificates from 27 states of the United States. Occup Environ Med. 2003;60:122–129.
37. Rosenman KD, Moore-Fuller M, Reilly MJ. Kidney disease and silicosis. Nephron. 2000;85:14–19.
Table. List of ICD-9...Image Tools