Synthetic wollastonite

Epidemiological data

Administrative data

Endpoint:

epidemiological data

Type of information:

experimental study

Adequacy of study:

key study

Study period:

1941-1994

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: Study in accordance with generally accepted scientific standards with minor restrictions (inadequate evaluation of smoking behavior)

Data source

Reference

Reference 1

Reference Type:

study report

Title:

Unnamed

Year:

1992

Reference 2

Reference Type:

publication

Title:

Dose-response associations of silica with nonmalignant respiratory disease and lung cancer mortality in the diatomaceous earth industry

Author:

Checkoway H, Heyer NJ, Seixas NS, Welp EA, Demers PA, Hughes JM, Weill H

Year:

1997

Bibliographic source:

Am J Epidemiol.; 145(8):680-8

Reference 3

Reference Type:

publication

Title:

Re: "Dose-response associations of silica with nonmalignant respiratory disease and lung cancer mortality in the diatomaceous earth industry"

Author:

Gibbs GW

Year:

1998

Bibliographic source:

Am J Epidemiol. 148(3):307

Reference 4

Reference Type:

publication

Title:

Radiographic evidence of silicosis risk in the diatomaceous earth industry

Author:

Hughes JM, Weill H, Checkoway H, Jones RN, Henry MM, Heyer NJ, Seixas NS, Demers PA

Year:

1998

Bibliographic source:

Am J Respir Crit Care Med; 158: 807-14

Reference 5

Reference Type:

publication

Title:

Crystalline silica exposure, radiological silicosis, and lung cancer mortality in diatomaceous earth industry workers

Author:

Checkoway H, Hughes JM, Weill H, Seixas NS, Demers PA

Year:

1999

Bibliographic source:

Thorax. 54(1):56-9; PMID: 10343633

Reference 6

Reference Type:

review article or handbook

Title:

FINAL REPORT - SILICA, SILICOSIS, AND LUNG CANCER IN DIATOMITE WORKERS (R01 OH03126)

Author:

Checkoway H, Heyer NJ, Seixas NS, Welp, EAE, Demers PA, Hughes JM & Weill H

Year:

1996

Bibliographic source:

NIOSH R01 OH03126

Materials and methods

Study type:

cohort study (retrospective)

Endpoint addressed:

repeated dose toxicity: inhalation

Test guideline

Qualifier:

no guideline available

Principles of method if other than guideline:

A retrospective cohort mortality study was performed on 2342 white male workers from a plant in the diatomaceous mining and processing industry which were employed between 1942 and 1994. Information on vital status was obtained from National Death Index.
Quantitative air monitoring data were obtained from the company.
Standard mortality ratios (SMRs) were calculated based on national and local mortalities as well as cause specific rates using the NIOSH life table program. Lung cancer mortality rates for white male residents of local counties in southern California were used to compute a regionally based standardized mortality ratio for lung cancer. Dose-response estimation for lung cancer and non-malignant respiratory diseases was conducted using Poisson regression modelling.
Five categories of cumulative exposure were defined such that there was an equal number of deaths from all causes in each category. Rate ratios were estimated treating the lowest cumulative exposure category as the referent. Covariates included: age, calendar year, duration of follow-up, Hispanic ethnicity, and cumulative asbestos exposure.

GLP compliance:

no

Test material

Test material information

Constituent 1

Reference substance name:

Diatomaceous earth

IUPAC Name:

Diatomaceous earth

Details on test material:

- Name of test material (as cited in study report): diatomaceous earth (DE)

Method

Type of population:

occupational

Ethical approval:

no

Details on study design:

HYPOTHESIS TESTED (if cohort or case control study)
To gain information on the mortality ratios and causes of diatomaceous earth (DE)-exposed workers in comparison to unexposed population taking into account occupational workplace measurements.

METHOD OF DATA COLLECTION
- Type: Record review / Work history
- Details: The study cohort was defined as workers employed for at least 12 months cumulative service at the plant and employed for at least 1 day between January 1, 1942 and December 31, 1987. The cohort for whom data are presented here was was composed of 2342 white men. Vital status was determined for the years 1942-1987 using several data sources, including the National Death Index (NDI), state driver's license bureaus, and a commercial credit bureau. Vital status information for the years 1988 -1994, inclusive, was determined from an updated NDI search. Workers known to be alive in 1979 who did not match the NDI files were assumed to be alive as of the end of follow-up; this assumption is supported by the previously noted high sensitivity of the NDI for death identification. Copies of death certificates, obtained from state vital statistics offices, were coded by a trained nosologist according to the International Classification of Diseases, Fifth through Ninth Revisions, codes in effect at the times of death.

STUDY PERIOD: from January 01 1942 to December 31 1987

SETTING: The study was conducted among the workforces of a diatomaceous earth (DE) mining and processing operation in Lompoc, California located roughly 80 km north of Santa Barbara. DE was discovered in the Lompoc area in the late nineteenth century, and mining and milling began at the plant in 1902.

STUDY POPULATION
- Total population (Total no. of persons in cohort from which the subjects were drawn): 2961
- Selection criteria: only white males due to much lower numbers of people of the other gender and race
- Total number of subjects participating in study: 2342
- Sex/age/race: males; Americans and Mexican-Americans; year of birth (mean ±SD): 1927 (range: 1881-1966)
- Smoker/non-smoker: smoker and non-smoker; Distinction between ever and never smokers were available for 1171 members of the main study cohort (50%). The smoking data did not include detailed information on amount smoked or duration of smoking.

COMPARISON POPULATION
- Type: Regional registry / National registry
- Details: Observed deaths in the cohort were compared with expected numbers based on rates in the United States population for 1942-94 for white males. Lung cancer mortality rates for white male residents of local counties in southern California were obtained from the University of Pittsburgh, in Pittsburgh, Pennsylvania.

HEALTH EFFECTS STUDIED
- Disease(s): Cancer sites, Death categories
- ICD No.: 6-9

Exposure assessment:

measured

Details on exposure:

TYPE OF EXPOSURE: Inhalation

TYPE OF EXPOSURE MEASUREMENT: Quantitative air monitoring data, including 5,709 measurements made during 1962-1988, were obtained from the company. These data were supplemented with earlier occupational hygiene data for the years 1948 -1962 (686 measurements) that were discovered subsequently in the company's archives. All of the pre-1962 data were based on particle counts, whereas nearly half of the post-1962 data were measured as gravimetric quantities of total dust (17 percent) or respirable dust concentrations (32 percent).
Conversion of the data from the older units of million particles per cubic foot to modem gravimetric units in milligrams per cubic meter (mg/m3) was performed by linear regression modelling on companion sampling data for both measurement methods.
Job-specific exposures for years before measurement data were available (pre-1948) were estimated by regression modelling extrapolation based on observed temporal changes and knowledge of dates when dust exposure reduction interventions occurred.
The respirable crystalline silica estimates data on the percentages of crystalline silica in the various product mixes and estimated job-specific fractions of exposure times to these products. The estimates of crystalline silica content were 1 percent for uncalcined DE (quartz content of the ore), 10 percent for calcined DE, and 20 percent for flux-calcined DE. The crystalline silica percentages were provided by knowledgeable industry personnel and were based on measured concentrations in bulk product samples rather than on data from measurements of airborne dust, which were not available.

EXPOSURE LEVELS:
See Table 1

Statistical methods:

Cause-specific mortality rate comparisons were made against rates for US white males for the years 1942-1994. Standardized mortality ratios and 95 percent confidence intervals were calculated using the National Institute for Occupational Safety and Health life table program.
Lung cancer mortality rates for white male residents of local counties in southern California, obtained from the University of Pittsburgh, in Pittsburgh, Pennsylvania, were used to compute a regionally based standardized mortality ratio for lung cancer. Person-time accumulation for each worker began at the later of January 1, 1942 or the date when 1 year of cumulative service was attained. Dose-response estimation for lung cancer and non-malignant respiratory diseases was conducted using Poisson regression modelling.
These internal analyses involved mortality rate comparisons among subcohorts classified according to cumulative exposures to respirable dust and respirable crystalline silica. Five categories of cumulative exposure were defined such that there was an equal number of deaths from all causes in each category. Rate ratios were estimated treating the lowest cumulative exposure category as the referent. Covariates included: age, calendar year, duration of follow-up, and Hispanic ethnicity, as a binary variable. Cumulative exposure to asbestos (fibers/ml-years) was included as a covariate to assess potential confounding. Asbestos categories were defined as non-exposed and two strata containing equal numbers of total deaths.
Exposure-response analyses were conducted with cumulative exposures of the silica and asbestos indices lagged by 0 and 15 years to accommodate disease latency effects. Poisson regression modelling was performed using the AMFIT Version 2.0 program of the Epicure statistical package.

Results and discussion

Results:

Observed mortality from all causes combined was nearly identical to expectation (SMR = 1.02), and only a small excess was detected for all cancers (SMR = 1.06); for details see table 2.
There was a prominent mortality excess of NMRD (SMR = 2.01) and a smaller excess for lung cancer (SMR = 1.29). The NMRD category included 27 deaths with underlying causes indicative of pneumoconiosis or silicosis listed as: "silicosis" (7), "diatomaceous earth pneumoconiosis" (3, "silicosis and asbestosis" (1), "pneumoconiosis" not otherwise specified (9), "pulmonary fibrosis" (4), and "idiopathic pulmonary fibrosis" (1). There were no deaths from silicotuberculosis. The lung cancer SMR increased to 1.44 (95% CI 1.14-1 30) in comparison with local county rates.
As shown in table 3, there were consistently strong NMRD mortality gradients with respect to the dust exposure indices. The association with cumulative exposure to respirable crystalline silica was stronger than with the less-specific respirable dust index, irrespective of lag interval. The most pronounced trend was seen for respirable crystalline silica, lagged 15 years; the rate ratio for the highest exposure stratum (25.0 mg/m3-years) compared with the lowest exposure category (<0.5 mg/m3 -year) reached 5.35 (95% CI 2.23-12.8). The trend slopes were unchanged when adjustments were made for cumulative exposure to asbestos.
Dose-response trends for lung cancer (table 4) were considerably weaker than those detected for NMRD. Despite the absence of consistent monotonically increasing gradients, there is evidence for a positive dose-response relation for lung cancer, particularly with cumulative exposure to respirable crystalline silica. The rate ratio for the 25.0 mg/m3-year stratum was 2.15 (95% CI 1.08-4.28).

Confounding factors:

We found no evidence that asbestos exposure in the DE industry confounded the observed associations for NMRD and lung cancer. Statistical adjustment for asbestos exposure had no influence on the dose-response trends for NMRD and lung cancer.
It is very unlikely that confounding by cigarette smoking was the sole or predominant explanation for the observed associations with NMRD and lung cancer. Our ability to assess potential confounding by smoking was limited by incomplete, crude data. However, the patterns of smoking prevalence and rate ratios for lung cancer (and NMRD) mortality with respect to dust exposure were dissimilar.

Strengths and weaknesses:

Vital status could not be ascertained for 9 percent of the cohort, which may have caused inflated standardized mortality ratios for the cohort as a whole because person-year accumulation was truncated for unknowns at the dates that they were last known to be alive. However, the internal analyses of dose-response trends for NMRD and lung cancer may have been biased slightly downward by incomplete vital status tracing because vital status was more complete for workers hired since 1960 (98 percent) than for those hired in earlier years (87 percent), when exposures and disease risks were highest.
The validity of our dose-response estimates is largely dependent on the quality and completeness of exposure data. Available data on dust monitoring spanned the years 1948-1988, with the majority of measurements made since the early 1960s. Consequently, extrapolation models were necessitated to estimate exposure concentrations for earlier years. Uncertainties in the conversion factors for dust particle counts to gravimetric units, the respirable fractions of dust concentrations, and the relative amounts of crystalline silica in the dust are other potential sources of measurement error.
Exposure to asbestos could be documented from personnel records for workers assigned to two small, dust-mixing operations during 1952-1977; however, asbestos exposures may also have been experienced by workers who had temporary, yet unrecorded, assignments in those departments, as well as episodically by maintenance workers involved in lagging and kiln relining. Therefore, misclassification of asbestos exposure may have hindered our ability to control for asbestos as a potential confounder.

Any other information on results incl. tables

Table 2: Standardized mortality ratios for selected causes of death, 1942-1994

Cause of death	Obs.	Exp.^a)	SMR	95% CI^b)
All causes	749	737	1.02	0.94-1.09
All cancers	181	171	1.06	0.91-1.22
- buccal cavity, pharynx	4	4.62	0.87	0.24-2.22
- esophagus	1	4.22	0.24	0.01-1.32
- stomach	7	6.83	1.03	0.41-2.11
- intestines, except rectum	14	15.5	0.90	0.49-1.52
- rectum	3	3.97	0.76	0.16-2.21
- liver	4	3.97	1.01	0.27-2.58
- pancreas	10	8.66	1.15	0.55-2.12
- larynx	4	2.32	1.73	0.47-4.42
- trachea, lung, bronchus	77	59.9	1.29	1.01-1.61
- prostate	11	12.6	0.88	0.44-1.59
- kidney	3	4.31	0.70	0.14-2.04
- bladder	2	4.46	0.45	0.05-1.62
- skin	2	3.53	0.57	0.07-2.05
- brain and nervous system	7	5.06	1.38	0.56-2.85
- leukemia and aleukemia	5	6.47	0.77	0.25-1.81
- other hematologic malignancies	5	5.88	0.85	0.28-1.99
Diabetes mellitus	8	11.3	0.71	0.31-1.40
Ischaemic heart disease	191	232	0.82	0.71-0.95
Cerebrovascular disease	34	39.5	0.86	0.60-1.20
Digestive disease	21	34.6	0.61	0.38-1.93
Genitourinary disease	10	9.47	1.06	0.51-1.94
Non-malignant respiratory disease (NMRD)	91	50.9	1.79	1.44-2.20
- Pneumonia	22	16.5	1.33	0.83-2.01
- Emphysema	14	8.56	1.64	0.89-2.75
- NMRD except pneumonia and infections diseases	67	33.4	2.01	1.56-2.55
Nervous system diseases	7	8.63	0.81	0.33-1.67
Accidents	50	45.7	1.10	0.81-1.44

^a)       Based on rates of US white males

^b)       95% confidence interval for SMR

 

Table 3: Non-malignant respiratory disease mortality trends by cumulative exposure to respirable dust and cumulative exposure to respirable crystalline silica, 1942-1994

Cumulative exposure [mg/m3 years]	Exposure lag [years]
	0			15
	Deaths	RR^a)	95% CI	Deaths	RR^a)	95% CI
Respirable dust
<1.9	10	1.00		13	1.00
1.9 - <4.0	8	0.94	0.37-2.40	9	1.26	0.51-3.15
4.0 - <7.4	9	1.18	0.47-2.98	8	1.25	0.48-3.25
7.4 - <18.3	14	1.37	0.59-3.20	11	1.31	0.53-3.24
≥18.3	26	2.63	1.16-5.94	26	3.39	1.49-7.70
Trend slope		1.02	1.00-1.03		1.02	1.01-1.03
Trend slope adjusted for asbestos exposure		1.02	1.00-1.03		1.02	1.00-1.03
Respirable crystalline silica
<0.5	7	1.00		10	1.00
0.5 - <1.1	8	1.52	0.55-4.20	9	2.04	0.77-5.45
1.1 - <2.1	10	1.98	0.75-5.22	8	1.96	0.71-5.43
2.1 - <5.0	12	2.34	0.91-6.00	13	3.17	1.25-8.05
≥5.0	30	4.79	2.01-11.9	27	5.35	2.23-12.8
Trend slope		1.08	1.03-1.13		1.08	1.03-1.14
Trend slope adjusted for asbestos exposure		1.08	1.02-1.14		1.08	1.03-1.14

^a) RR: Rate ratio, adjusted for age, calendar year, duration of follow-up, and ethnicity

 

Table 4: Lung cancer mortality trends by cumulative exposure to respirable dust and cumulative exposure to respirable crystalline silica, 1942-1994

Cumulative exposure [mg/m3 years]	Exposure lag [years]
	0			15
	Deaths	RR^a)	95% CI	Deaths	RR^a)	95% CI
Respirable dust
<1.9	15	1.00		18	1.00
1.9 - <4.0	13	1.01	0.48-2.13	14	1.31	0.62-2.76
4.0 - <7.4	11	0.99	0.45-2.19	15	1.64	0.78-3.47
7.4 - <18.3	20	1.50	0.75-2.99	14	1.43	0.66-3.12
≥18.3	18	1.67	0.77-3.63	16	2.05	0.92-4.56
Trend slope		1.01	1.00-0.03		1.01	1.00-1.03
Trend slope adjusted for asbestos exposure		1.01	1.00-1.03		1.01	1.00-1.03
Respirable crystalline silica
<0.5	17	1.00		22	1.00
0.5 - <1.1	14	1.07	0.53-2.18	12	0.96	0.47-1.98
1.1 - <2.1	7	0.55	0.23-1.32	9	0.77	0.35-1.72
2.1 - <5.0	15	1.19	0.59-2.41	14	1.26	0.62-2.57
≥5.0	24	2.11	1.07-4.11	20	2.15	1.08-4.28
Trend slope		1.06	1.01-1.11		1.05	0.99-1.11
Trend slope adjusted for asbestos exposure		1.06	1.01-1.11		1.05	0.99-1.11

^a) RR: Rate ratio, adjusted for age, calendar year, duration of follow-up, and ethnicity

 

Applicant's summary and conclusion

Conclusions:

The findings for NMRD indicating strong, monotonically increasing exposure-response associations with crystalline silica are consistent with other studies of cohorts exposed to silica.
A reasonably strong, but not monotonically increasing, dose-response trend was detected for lung cancer. Excess risk was predominantly concentrated in the highest cumulative exposure stratum of either respirable dust or respirable crystalline silica.
The findings for NMRD strongly indicate an etiologic role of crystalline silica. Additionally, the dose-response analyses provide support for the hypothesis that crystalline silica is a human lung carcinogen.

Executive summary:

A historical cohort mortality study of 2,342 male workers exposed to crystalline silica, predominantly cristobalite, was conducted in a diatomaceous earth mining and processing facility in. During the years 1942-1994, mortality excesses were detected for non-malignant respiratory diseases (NMRD) (standardized mortality ratio = 2.01, 95% confidence interval (CI) 1.56-2.55) and lung cancer (standardized mortality ratio = 1.29, 95% CI 1.01-7.61). NMRD mortality rose sharply with cumulative exposure to respirable crystalline silica; allowing for a 15-year latency, the rate ratio for the highest exposure stratum (≥5.0 mg/m3-years) was 5.35 (95% CI 2.23-12.8). The rate ratio for lung cancer reached 2.15 (95% CI 1.08-4.28) in the highest exposure category. These associations were unlikely to have been confounded by smoking or asbestos exposure. The findings indicate a strong dose-response relation for crystalline silica and NMRD mortality. The lung cancer results, although less convincing, add further support to an etiologic role for crystalline silica.