Occupational Exposure to Dusts, Gases, and Fumes ... - ATS Journals

Occupational Exposure to Dusts, Gases, and Fumes and Incidence of Chronic Obstructive Pulmonary Disease in the Swiss Cohort Study on Air Pollution and Lung and Heart Diseases in Adults Amar J. Mehta1,2,3, David Miedinger4,5, Dirk Keidel1,2, Robert Bettschart6, Andreas Bircher7, Pierre-Olivier Bridevaux8, Ivan Curjuric1,2, Hans Kromhout9, Thierry Rochat8, Thomas Rothe10, Erich W. Russi11, Tamara Schikowski1,2, Christian Schindler1,2, Joel Schwartz3, Alexander Turk12, ¨ nzli1,2, and the SAPALDIA Team1,2 Roel Vermeulen9, Nicole Probst-Hensch1,2, Nino Ku 1

Swiss Tropical and Public Health Institute, Basel, Switzerland; 2University of Basel, Basel, Switzerland; 3Harvard School of Public Health, Boston, Massachusetts; 4Occupational Medicine, Suva, Luzern, Switzerland; 5Department of Internal Medicine, University Hospital, Basel, Switzerland; 6 Lungen Zentrum, Hirslanden Klinik, Aarau, Switzerland; 7University Hospital, Basel, Switzerland; 8University Hospitals of Geneva, Geneva, Switzerland; 9Institute for Risk Assessment Sciences, Utrecht, Netherlands; 10Zuercher Hoehenklinik Davos, Davos Clavadel, Switzerland; 11University Hospital Zurich, Zurich, Switzerland; and 12Zu¨rcher Ho¨henklinik Wald, Faltigberg-Wald, Switzerland

Rationale: There is limited evidence from population-based studies demonstrating incidence of spirometric-defined chronic obstructive pulmonary disease (COPD) in association with occupational exposures. Objectives: We evaluated the association between occupational exposures and incidence of COPD in the Swiss Cohort Study on Air Pollution and Lung and Heart Diseases in Adults (SAPALDIA). Measurements and Main Results: Prebronchodilator ratio of forced expiratory volume in 1 second over forced vital capacity (FEV1/FVC) was measured in 4,267 nonasthmatic SAPALDIA participants ages 18–62 at baseline in 1991 and at follow-up in 2001–2003. COPD was defined by the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criterion (FEV1/FVC , 0.70) and Quanjer reference equation (FEV1/FVC , lower limit of normal [LLN]), and categorized by severity (>80% and ,80% predicted FEV1 for stage I and stage II1, respectively). Using a job-exposure matrix, self-reported occupations at baseline were assigned exposures to biological dusts, mineral dusts, gases/fumes, and vapors, gases, dusts, or fumes (VGDF) (high, low, or unexposed as reference). Adjusted incident rate ratios (IRRs) of stage I and stage II1 COPD were estimated in mixed Poisson regression models. Statistically significant (P , 0.05) IRRs of stage II1 GOLD and LLN-

(Received in original form October 30, 2011; accepted in final form March 13, 2012) Supported by Swiss Accident Insurance Fund (Suva); the Swiss National Science Foundation (grants no 33CSCO-108796, 3247BO-104283, 3247BO-104288, 3247BO-104284, 3247-065896, 3100-059302, 3200-052720, 3200-042532, and 4026-028099); the Federal Office for Forest, Environment, and Landscape; the Federal Office of Public Health; the Federal Office of Roads and Transport; the canton’s government of Aargau, Basel-Stadt, Basel-Land, Geneva, Luzern, Ticino, Valais, Zurich, the Swiss Lung League; and the canton’s Lung League of Basel Stadt/Basel Landschaft, Geneva, Ticino, Valais, and Zurich. Author Contributions: A.J.M., D.M., R.B., A.B., P.-O.B., T. Rochat, T. Rothe, E.W.R., C.S., J.S., A.T., N.P.-H., and N.K. were involved in the conception, hypothesis delineation, and design of the study. A.J.M., D.M., D.K., R.B., A.B., P.-O.B., I.C., H.K., T. Rochat, T. Rothe, E.W.R., T.S., C.S., J.S., A.T., R.V., N.P.-H., and N.K. were involved in the acquisition of the data or the analysis and interpretation of the data. A.J.M., D.M., D.K., R.B., A.B., P.-O.B., I.C., H.K., T. Rochat, T. Rothe, E.W.R., T.S., C.S., J.S., A.T., R.V., N.P.-H., and N.K. were involved in writing the manuscript or had substantial involvement in its revision prior to submission. Correspondence and requests for reprints should be addressed to Amar Mehta, Sc.D., M.P.H., Harvard School of Public Health, Landmark Ctr., West 415, 401 Park Dr., Boston, MA 02215. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.www.atsjournals.org Am J Respir Crit Care Med Vol 185, Iss. 12, pp 1292–1300, Jun 15, 2012 Copyright ª 2012 by the American Thoracic Society Originally Published in Press as DOI: 10.1164/rccm.201110-1917OC on April 6, 2012 Internet address: www.atsjournals.org

AT A GLANCE COMMENTARY Scientific Knowledge on the Subject

There is limited evidence from population-based studies demonstrating incidence of spirometric-defined chronic obstructive pulmonary disease (COPD) in association with occupational respirable exposures. What This Study Adds to the Field

The present findings from the Swiss Cohort Study on Air Pollution and Lung and Heart Diseases in Adults indicate that high levels of occupational exposures to biological dusts, mineral dusts, gases/fumes, and either vapors, gases, dusts, or fumes (VGDF), as determined from a general population-based job-exposure matrix, were associated with increased incidence of COPD of at least moderate severity (stage II1 COPD). The findings were similar when stage II1 COPD was defined according to either the Global Initiative for Chronic Obstructive Lung Disease (GOLD) or lower limit of normal criteria, although the magnitude of risk was moderately higher when the GOLD definition was applied.

COPD, indicating risks between two- and fivefold, were observed for all occupational exposures at high levels. Occupational exposureassociated risk of stage II1 COPD was observed mainly in males and ages >40 years, and remained elevated when restricted to nonsmokers. Conclusions: In a Swiss working adult population, occupational exposures to biological dusts, mineral dusts, gases/fumes, and VGDF were associated with incidence of COPD of at least moderate severity.

Chronic obstructive pulmonary disease (COPD) is an important and increasing cause of morbidity and mortality worldwide. Although chronic irritation of the airways from cigarette smoke is the most important single causal factor for developing COPD, a recent statement from the American Thoracic Society (ATS) concluded that there was sufficient epidemiologic evidence to infer a causal relationship between occupational respirable exposures and development of COPD (1). It has been estimated that approximately 15–20% of the population burden of COPD is attributable to occupational exposures (2), but this estimate

Mehta, Miedinger, and Keidel: Occupational Exposures and Incidence of COPD in SAPALDIA

1293

TABLE 1. DISTRIBUTION OF BASELINE OCCUPATIONAL EXPOSURES IN COPD CASES AND NONCASES GOLD* Stage II1 (n ¼ 57)

1,819 (54.8)

291 (56.0)

19 (33.3)

2,037 (54.7)

138 (55.7)

19 (37.5)

618 (18.6) 134 (4.0)

95 (18.3) 20 (3.9)

16 (28.1) 5 (8.8)

685 (18.4) 158 (4.2)

51 (20.6) 8 (3.2)

14 (27.5) 3 (5.9)

20.8 (27.8) 12 (5, 23)

27.5 (36.8) 15 (6, 31)

33.0 (33.8) 17 (10, 42)

23.1 (30.8) 13 (6, 25)

23.2 (37.0) 10 (5, 21)

26.2 (26.5) 15 (8, 36)

496 (14.9) 172 (5.2)

97 (18.7) 26 (5.0)

10 (17.5) 8 (14.0)

574 (15.4) 202 (5.4)

41 (16.5) 12 (4.8)

11 (21.6) 6 (11.8)

26.8 (33.4) 16 (6, 30)

31.3 (36.5) 20 (9, 34)

45.7 (47.1) 22.5 (14, 96)

26.1 (31.7) 15 (7, 31)

42.9 (41.2) 30 (15, 42)

173 (33.3) 31 (6.0)

25 (43.9) 10 (17.5)

84 (33.9) 11 (4.4)

21 (41.2) 8 (15.7)

19.5 (25.5) 12 (6, 24)

33.6 (31.1) 24 (14, 42)

85 (34.3) 25 (10.1)

18 (35.3) 14 (27.5)

Noncases (n ¼ 3,321) Unexposed‡, n (%) Exposure to biological dusts‡, n (%) Low High Cumulative exposure (years)x Mean (SD) Median (IQR) Exposure to mineral dusts‡, n (%) Low High Cumulative exposure (years)x Mean (SD) Median (IQR) Exposure to gases/fumes‡, n (%) Low High Cumulative exposure (years)x Mean (SD) Median (IQR) Exposure to VGDF‡, n (%) Low High Cumulative exposure (years)x Mean (SD) Median (IQR)

LLN*

†

1,069 (32.2) 193 (5.8) 21.2 (27.2) 13 (5, 25) 1,104 (33.2) 398 (12.0) 25.9 (32.9) 15 (6, 28)

Stage I (n ¼ 520)

25.0 (28.9) 18 (10, 28.5)

†

35.4 (35.3) 23 (11, 44)

168 (32.3) 61 (11.7)

18 (31.6) 20 (35.1)

33.2 (40.1) 20 (9, 33)

48.8 (43.3) 33 (14, 80)

Noncases (n ¼ 3,724)

29.3 (36.1) 18 (7, 31.5) 1,204 (32.3) 226 (6.1) 23.0 (28.8) 15 (6, 27) 1,221 (32.8) 466 (12.5) 28.5 (35.8) 16 (6, 30)

Stage I (n ¼ 248)

26.7 (37.2) 12 (6, 25)

Stage II1 (n ¼ 51)

43.4 (39.7) 29 (12.5, 70)

Definitions of abbreviations: COPD ¼ chronic obstructive pulmonary disease; GOLD ¼ Global Initiative for Chronic Obstructive Lung Disease; IQR ¼ interquartile range; LLN ¼ lower limit of normal; VGDF ¼ vapors, gases, dusts, or fumes. * COPD is defined according to the GOLD fixed cut-off criterion (ratio of forced expiratory volume over 1 second over forced vital capacity [FEV1/FVC] , 0.70) and the LLN criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population. y Based on the Quanjer prediction equation, severity of COPD is classified as stage I if FEV1 value > 80% of the predicted value, and as stage II or more (stage II1) if FEV1 value , 80% of the predicted value. z Exposure level assignment is determined from a general population-based job exposure matrix, called ALOHA, in combination with current job title reported in the baseline survey; unexposed participants were assigned no exposure to biological dusts, mineral dusts, gases/fumes, and VGDF. x Restricted to exposed participants, cumulative exposure (unit, years) is a multiplicative product of exposure level and years worked in current occupation, and is weighted by exposure level (1, 2 for low, high, respectively) on an exponential scale.

varies between smokers and nonsmokers; between 15 and 19% of COPD in smokers (2) and as much as 32% in never-smokers (3) may be attributable to occupational exposures. The combined effects of smoking and occupational exposures on COPD have also been observed (4, 5). A wealth of evidence supporting a causal relationship between occupational exposures and development of COPD comes from population-based studies. However, most of the earlier population-based studies are either cross-sectional (3, 6, 7) or case-control (4, 8) by design; few, if any, of the previous studies have prospectively evaluated incidence of spirometric-defined COPD. Little is known about the burden of COPD among working individuals or on exposure to airway irritants in the workplace in Switzerland. A recent analysis of the Swiss Cohort Study on Air Pollution and Lung and Heart Diseases in Adults (SAPALDIA), a large population-based prospective cohort study of adults ages 18–62 from eight regions in Switzerland, estimates that 10.0% of its participants had airflow obstruction according to spirometry (9). The original aim of SAPALDIA was to investigate the chronic respiratory health effects of air pollution exposure (10). In this analysis, we used standardized occupational information available in SAPALDIA, in combination with a general population–based job-exposure matrix (GPJEM) that allows for semiquantitative exposure assessment for multiple occupational exposures (7), to assess whether preceding occupational exposure was associated with incidence of COPD defined by spirometry. We also evaluated whether the observed associations

were modified by smoking status, symptoms consistent with chronic bronchitis, and age. Some of the results presented in this analysis were previously reported in the form of an abstract (11).

METHODS More detailed description of the study population and design are summarized in previous publications (10, 12).

Study Population SAPALDIA is a multicenter, population-based prospective cohort study consisting of a random sample of 9,561 participants of ages between 18 and 62 years old that were recruited from eight regions in Switzerland, and were administered medical examinations, prebronchodilator spirometry testing, and a detailed health questionnaire, at baseline in 1991 (13). The follow-up survey (SAPALDIA 2) was conducted from 2001–2003, of which 8,047 of the original study participants were present. This analysis was restricted to 4,267 nonasthmatic participants who completed spirometry testing at both surveys, and who were with complete information on occupation and important covariates. See online supplement for additional details describing the selection of participants included in this analysis. The SAPALDIA cohort study complies with the Declaration of Helsinki. Written informed consent was obtained from participants in both surveys. The study was approved by the central ethics committee of the Swiss Academy of Medical Sciences and the respective Cantonal Ethics Committees of the eight study regions.

1294

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 185

2012

Figure 1. Adjusted incident rate ratios (IRRs) for chronic obstructive pulmonary disease (COPD) as a function of occupational exposure according to COPD definition and severity, and exposure type. IRRs for exposure levels (high, low, unexposed as reference) of each exposure type are estimated in Poisson regression with random intercept for area in separate models after adjustment for the following baseline covariates: age, body mass index, sex, respiratory infection at early age, parental asthma, smoking status, pack-years smoked, environmental tobacco smoke exposure, and education level; see online supplement for detailed description of covariates. COPD is defined according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) fixed cut-off criterion (ratio of forced expiratory volume over 1 s over forced vital capacity [FEV1/FVC] , 0.70; solid circles) and the lower limit of normal (LLN) criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population (shaded circles). zBased on the Quanjer prediction equation, severity of COPD is classified as stage I if FEV1 value > 80% of the predicted value, and as stage II or more (stage II1) if FEV1 value , 80% of the predicted value; participants absent of COPD at follow-up are considered noncases in all models. xOrdinal trend tests for biological dusts, mineral dusts, gases/fumes, vapors/gases/dusts/fumes in stage II1 GOLD- and LLN-COPD models are statistically significant (P , 0.05). exp. ¼ exposure.

Spirometry Testing

Occupational Exposure Assessment

The spirometry protocol was equivalent to that of the European Community Respiratory Health Survey (ECRHS) (14) and complied with ATS criteria (15). No bronchodilation was applied. See online supplement and previous publications (16, 17) for additional details describing methods for spirometry testing.

Current job titles reported by participants at the baseline survey (1991) were standardized according to the International Standard Classification of Occupations (ISCO-88) code’s four-digit classification (20). The ISCO-88 classifications were linked to a GPJEM for COPD (6, 7), called ALOHA, and assigned scores for high, low, or no exposure (2, 1, or 0, respectively) to biological dusts, mineral dusts, gases/fumes, and either vapors, gases, dusts, or fumes (VGDF); VGDF was a composite variable based on assigned scores of the other three exposures. A uniformed unexposed comparison group comprised of participants whose ISCO-88 occupational classification was assigned a value of “0” for all exposure types. Cumulative exposure (unit, years) was also estimated for each exposure type by multiplying exposure level and the number of years worked in the specific job, and was weighted by exposure level on an exponential scale.

COPD Definition and Severity COPD was defined according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) fixed cut-off criterion (ratio of forced expiratory volume in 1 s over forced vital capacity [FEV1/FVC] , 0.70) (18), and according to the fifth percentile of the FEV1/FVC ratio distribution (FEV1/FVC , lower limit of normal [LLN]) based on the European Coal and Steel Community (ECSC) equations as predicted by Quanjer and colleagues (19). Using the prediction equations by Quanjer and coworkers, incident COPD cases were categorized according to stage I (predicted FEV1 > 80%) or stage II or higher (stage II1, predicted FEV1 , 80%) severity. Of the 4,267 study participants, 369 and 244 participants were identified with COPD at baseline according to the GOLD and LLN definitions, respectively, and thus were excluded from this analysis.

Statistical Analysis Incident rate ratios (IRRs) of COPD were estimated as discrete (high, low, unexposed) and continuous (.0 yr for exposed, 0 yr for unexposed) functions of each occupational exposure in mixed Poisson regression models with a random intercept for study area using the

Mehta, Miedinger, and Keidel: Occupational Exposures and Incidence of COPD in SAPALDIA TABLE 2. ADJUSTED IRRS* FOR STAGE II1 COPD AS A CONTINUOUS FUNCTION OF CUMULATIVE EXPOSURE Stage II1 GOLD-COPD† Cumulative exposure (per 10 yr)‡ Biological dusts Mineral dusts Gases/fumes VGDF

IRR 1.15 1.10 1.10 1.11

95% CI 1.04 1.01 1.02 1.04

1.26 1.20 1.19 1.17

Stage II1 LLN-COPD† IRR 1.11 1.08 1.06 1.07

95% CI 0.99 0.98 0.98 0.99

1.24 1.19 1.15 1.15

Definitions of abbreviations: COPD ¼ chronic obstructive pulmonary disease; GOLD ¼ Global Initiative for Chronic Obstructive Lung Disease; IRR ¼ incident rate ratio; LLN ¼ lower limit of normal; VGDF ¼ vapors, gases, dusts, or fumes. Bold and bold italics fonts indicate P , 0.05 and P , 0.10, respectively. * IRRs for cumulative exposure (unit, per 10 yr) are estimated in multiple Poisson regression with random intercept for area for each exposure type in separate models after adjustment for the following baseline covariates: age, body mass index, sex, respiratory infection at early age, parental asthma, smoking status, pack-years smoked, environmental tobacco smoke exposure, and education level. y COPD is defined according to the GOLD fixed cut-off criterion (ratio of forced expiratory volume over 1 s over forced vital capacity [FEV1/FVC] , 0.70) and the LLN criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population; based on the Quanjer prediction equation, severity of COPD is classified as stage II or higher (stage II1) if FEV1 value , 80% of the predicted value; participants absent of COPD at follow-up are considered noncases in all models. z Cumulative exposure was a multiplicative product of exposure level and years worked in occupation and was weighted by exposure level (0, 1, 2 for unexposed, low, high, respectively) on an exponential scale; cumulative exposure was scaled per 10 yr.

xtpoisson routine in STATA release 10 (StataCorp, College Station, TX). All models were repeated according to COPD definition (GOLD, LLN) and severity (stage I, stage II1), and exposure of interest. IRRs were adjusted for covariates ascertained at baseline including age, smoking status (current, former, never-smoker as reference), cumulative pack-years smoked, gender (male as reference), and other important covariates; see online supplement for additional details on other covariates. All models were stratified by gender and baseline characteristics including smoking status (ever, never), age (>40, ,40 yr), and chronic bronchitis in single nested models; two-way interaction terms between occupational exposure and smoking status, ages >40 years, and chronic bronchitis were tested separately to evaluate whether they underlie susceptibility to the exposure response. Population attributable fractions (PAFs) for the association between each occupational exposure and COPD were also computed separately in smokers and nonsmokers with standard errors adjusted for intra–study area correlation using the aflogit routine in STATA (21–23) (see online supplement for additional details).

Sensitivity and Secondary Analyses Because a substantial number of the participants at follow-up did not complete spirometry testing (22.3%), the Poisson regression models were weighted by the inverse probability of completing spirometry at follow-up. Poisson regression models were repeated using COPD definition (GOLD, LLN) and severity (stage I, stage II1) according to prediction equations by Hankinson and colleagues (24) and Bra¨ndli and coworkers (25, 26) that were derived from the third National Health and Nutrition Examination Survey (NHANES III) and SAPALDIA study populations, respectively. See online supplement for description of additional sensitivity analyses.

RESULTS In total, there were 520 and 57 incident stage I and stage II1 GOLD-COPD cases, respectively, and 3,321 noncases (Table 1). As expected, there were fewer LLN-COPD cases, more so for stage I (n ¼ 248) severity than stage II1 severity (n ¼ 51). Relative to noncases and stage I COPD cases, mean age and body mass index were higher in stage II1 GOLD- and LLN-COPD

1295

cases, and stage II1 COPD cases also had larger proportions of participants who were males, current smokers, exposed to environmental tobacco smoke, and less educated and who reported symptoms of chronic bronchitis (Table E1 in online supplement). Seventeen stage II1 COPD cases were defined according to the GOLD, but not LLN criterion. The mean age difference between these 17 GOLD-COPD and the stage II1 LLN-COPD cases was approximately 4 years, although this difference was not significantly different (49.6 vs. 45.5 yr, respectively, P ¼ 0.24); similar distributions in gender and occupational exposures were observed (data not shown). The distributions of occupational exposures were very similar between stage I COPD cases and noncases, but stage II1 GOLD- and LLN-COPD cases had moderately higher proportions of participants with occupational exposures at high levels (Table 1). Restricting to exposed participants, the median values of cumulative exposure for each exposure were moderately higher in stage II1 GOLD- and LLN-COPD cases relative to noncases; median differences were largest for gases/fumes and VGDF. Exposure to gases/fumes was present in the majority of current occupations reported at baseline that were assigned at least one occupational exposure type according to the ALOHA JEM, but most of these occupations included multiple exposure types (Table E2). Relative to females, males were more likely to report occupations with high levels of occupational exposures dusts, gases, or fumes (Table E3). The majority of male and female participants reported working in white-collar occupations; among the blue-collar occupations, the metal and health care industries were most prevalent for males and females, respectively (Table E3). Among participants with high level of occupational exposure to VGDF, the most prevalent ISCO-88 classifications were “crop and animal producers” (10.9%) and “motor vehicle mechanics and fitters” (10.3%) (Table E4); participants assigned these ISCO-88 classifications described themselves as farmers and auto mechanics, respectively (data not shown). High levels of occupational exposures to biological dusts, mineral dusts, gases/fumes, and VGDF at baseline were significantly associated (P , 0.05) with increased incidence of stage II1 GOLD- and LLN-COPD, after adjustment for all covariates (Figure 1, Table E6). Relative to unexposed participants, the risk of stage II1 GOLD- and LLN-COPD in association with occupational exposures at high levels were between twoand fivefold, with the strongest associations observed between stage II1 GOLD-COPD and VGDF. Magnitude of association between occupational exposure and stage II1 COPD varied by definition; the adjusted IRRs of stage II1 COPD were moderately higher for GOLD- than LLN-defined COPD. Characterized as cumulative exposure, all exposures were significantly associated with stage II1 GOLD-COPD (Table 2); per 10-year increase in cumulative exposure, the risk of stage II1 GOLDCOPD increased approximately 10–15% depending on type of exposure. Associations of slightly lower magnitude and marginal statistical significance (P , 0.10) were observed between stage II1 LLN-COPD and cumulative exposure to biological dusts and between stage II1 LLN-COPD and cumulative exposure to VGDF. The adjusted IRRs of GOLD- and LLN-COPD for each occupational exposure after stratification by smoking status and chronic bronchitis at baseline are summarized in Tables 3 and 4, respectively. Although interaction was not observed, the risk of stage II1 GOLD-COPD in association with each occupational exposure was more elevated in nonsmokers than in ever smokers. Each occupational exposure was associated with at least a threefold risk of stage II1 GOLD-COPD in nonsmokers, and was statistically significant for gases/fumes and

1296


VOL 185

2012

TABLE 3. ADJUSTED IRRS* FOR COPD AS A FUNCTION OF OCCUPATIONAL EXPOSURE AFTER STRATIFICATION BY SMOKING STATUS AT BASELINE IN SEPARATE MODELS BY COPD SEVERITY AND DEFINITION, AND EXPOSURE GOLD-COPD† Ever-Smokers Exposure

LLN-COPD† Nonsmokers

n Cases

IRR (95%CI)

n Cases

55 159

0.90 (0.66–1.23) Ref

60 132

70 159

1.02 (0.76–1.38) Ref

113 159

Ever-Smokers

IRR (95%CI)

Nonsmokers

n Cases

IRR (95%CI)

n Cases

IRR (95%CI)

1.07 (0.79–1.46) Ref

29 84

0.91 (0.59–1.39) Ref

30 54

1.30 (0.83–2.04) Ref

53 132

1.31 (0.94–1.82) Ref

26 84

0.75 (0.47–1.19) Ref

27 54

1.65 (1.02–2.68) Ref

0.95 (0.74–1.22) Ref

91 132

1.18 (0.90–1.55) Ref

54 84

0.85 (0.60–1.21) Ref

41 54

1.28 (0.85–1.94) Ref

123 159

0.89 (0.70–1.13) Ref

106 132

1.11 (0.86–1.44) Ref

58 84

0.80 (0.57–1.13) Ref

52 54

1.34 (0.91–1.97) Ref

15 15

2.40 (1.15–5.02) Ref

6 4

3.14 (0.88–11.24) Ref

15 16

2.76 (1.32–5.75) Ref

2 3

1.56 (0.26–9.40) Ref

13 15

1.60 (0.73–3.52) Ref

5 4

3.22 (0.84–12.36) Ref

15 16

1.56 (0.73–3.31) Ref

2 3

1.90 (0.31–11.63) Ref

24 15

1.50 (0.76–2.97) Ref

11 4

3.94 (1.23–12.58) Ref

24 16

1.52 (0.78–2.97) Ref

5 3

2.55 (0.60–10.84) Ref

27 15

1.54 (0.80–2.99) Ref

11 4

3.28 (1.03–10.41) Ref

27 16

1.57 (0.83–2.99) Ref

5 3

2.17 (0.51–9.14) Ref

†

Stage I Biological dusts High or low Unexposed Mineral dusts‡ High or low Unexposed Gases/fumes High or low Unexposed VGDF‡ High or low Unexposed Stage II† Biological dusts High or low Unexposed Mineral dusts High or low Unexposed Gases/fumes High or low Unexposed VGDF High or low Unexposed

Definitions of abbreviations: COPD ¼ chronic obstructive pulmonary disease; GOLD ¼ Global Initiative for Chronic Obstructive Lung Disease; IRR ¼ incident rate ratio; LLN ¼ lower limit of normal; Ref = reference; VGDF ¼ vapors, gases, dusts, or fumes. Bold and bold italics fonts indicate P , 0.05 and P , 0.10, respectively. * Stratum-specific IRRs were estimated in single nested multiple Poisson regression models, for each exposure type in separate models, with random intercept for area after adjustment for the following baseline covariates: age, body mass index, sex, respiratory infection at early age, parental asthma, smoking status, pack-years smoked, environmental tobacco smoke exposure, and education level. y COPD is defined according to the GOLD fixed cut-off criterion (ratio of forced expiratory volume over 1 s over forced vital capacity [FEV1/FVC] , 0.70), and the LLN criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population; based on the Quanjer prediction equation, severity of COPD is classified as stage I if FEV1 value > 80% of the predicted value, and as stage II or more (stage II1) if FEV1 value , 80% of the predicted value; participants absent of COPD at follow-up are considered noncases in all models. z Two-way interactions between smoking status and mineral dusts, and between smoking status and VGDF, were statistically significant (P , 0.05) in stage I LLNCOPD models.

VGDF. With the exception of biological dusts, a similar pattern was observed to a lesser extent between smokers and nonsmokers for stage II1 LLN-COPD. Chronic bronchitis at baseline was observed to be a significant effect modifier on the associations between all occupational exposures and stage I GOLD-COPD, and between biological dusts and stage I LLN-COPD (Table 4); elevated risks of stage I GOLD- and LLN-COPD in association with occupational exposures were observed only with chronic bronchitis present. In contrast, risk of stage II1 GOLD- and LLN-COPD were increased in participants absent of chronic bronchitis. The strong associations between occupational exposures and stage II1 GOLD- and LLN-COPD, as presented in Figure 1 and Table E6, were observed mainly in ages >40 years at baseline, and in males (Tables E8 and E9, respectively). The PAFs of the associations between each occupational exposure and stage II1 GOLD- and LLN-COPD are presented in Table 5. In smokers, the PAFs of stage II1 COPD were highest for biological dusts, ranging between 31 and 32% depending on definition. In nonsmokers, the estimated PAFs for each occupational exposure varied considerably according to COPD definition; the estimated PAFs ranged between 43 and 56% depending on

exposure for GOLD-defined stage II1 COPD, and between 11 and 38% for LLN-defined stage II1 COPD. Sensitivity and Secondary Analyses

Poisson models that were weighted by the inverse probability of having complete spirometry at the follow-up survey produced very similar effect estimates for each occupational exposure as presented in Figure 1 and Table 2 (Table E7). The main effects of occupational exposure were consistent when COPD definition and severity was based on NHANES and SAPALDIA prediction equations (Tables E5 and E6). However, the magnitude of the observed IRRs of stage II1 GOLD- and LLNCOPD for high-level occupational exposures were moderately lower compared with those based on the Quanjer-ECSC prediction equations. Exclusion of participants who reported wheezing without cold yielded similar effect estimates for occupational exposures as presented in Figure 1 and Table E7, with slightly stronger associations for cumulative exposure (Figure E1 and Table E10). Presence of interaction was not observed between either asthma or wheezing without cold (at baseline) and any of the occupational exposures (data not shown). Relative to unexposed participants who were never smokers, the


1297

TABLE 4. ADJUSTED IRRS* FOR COPD AS A FUNCTION OF OCCUPATIONAL EXPOSURE AFTER STRATIFICATION BY CHRONIC BRONCHITIS AT BASELINE IN SEPARATE MODELS BY COPD SEVERITY AND DEFINITION, AND EXPOSURE GOLD-COPD† Chronic Bronchitis Exposure

LLN-COPD† No Chronic Bronchitis

‡

Chronic Bronchitis

No Chronic Bronchitis

n Cases

IRR (95%CI)

n Cases

IRR (95%CI)

n Cases

IRR (95%CI)

2.31 (1.28–4.20) Ref

92 270

0.86 (1.28–4.20) Ref

14 13

2.20 (1.03–4.70) Ref

45 125

0.91 (0.64–1.28) Ref

27 21

2.12 (1.19–3.78) Ref

96 270

1.00 (0.78–1.29) Ref

12 13

1.52 (0.68–3.37) Ref

41 125

0.95 (0.65–1.39) Ref

34 21

1.76 (1.02–3.04) Ref

170 270

0.96 (0.78–1.17) Ref

19 13

1.51 (0.74–3.07) Ref

76 125

0.92 (0.68–1.24) Ref

38 21

1.74 (1.02–3.97) Ref

191 270

0.90 (0.75–1.10) Ref

22 13

1.55 (0.78–3.08) Ref

88 125

0.90 (0.68–1.20) Ref

4 5

1.77 (0.46–6.85) Ref

17 14

2.62 (1.27–5.43) Ref

4 5

1.90 (0.50–7.13) Ref

13 14

2.60 (1.19–5.67) Ref

2 5

0.51 (0.09–2.71) Ref

16 14

2.31 (1.07–5.00) Ref

3 5

0.69 (0.16–2.95) Ref

14 14

1.92 (0.87–4.22) Ref

9 5

1.20 (0.38–3.83) Ref

26 14

2.13 (1.08–4.22) Ref

9 5

1.16 (0.37–3.66) Ref

20 14

1.77 (0.87–3.62) Ref

9 5

1.13 (0.36–3.58) Ref

29 14

2.07 (1.07–4.02) Ref

9 5

1.08 (0.34–3.39) Ref

23 14

1.81 (0.91–3.60) Ref

n Cases

IRR (95%CI)

23 21

†

Stage I Biological dusts‡,x High or low Unexposed Mineral dusts‡ High or low Unexposed Gases/fumes‡ High or low Unexposed VGDF‡ High or low Unexposed Stage II1† Biological dusts High or low Unexposed Mineral dusts High or low Unexposed Gases/fumes High or low Unexposed VGDF High or low Unexposed

Definitions of abbreviations: COPD ¼ chronic obstructive pulmonary disease; GOLD ¼ Global Initiative for Chronic Obstructive Lung Disease; IRR ¼ incident rate ratio; LLN ¼ lower limit of normal; Ref ¼ reference; VGDF ¼ vapors, gases, dusts, or fumes. Bold and bold italics fonts indicate P , 0.05 and P , 0.10, respectively. * Stratum-specific IRRs were estimated in single nested multiple Poisson regression models, for each exposure type in separate models, with random intercept for area after adjustment for the following baseline covariates: age, body mass index, sex, respiratory infection at early age, parental asthma, smoking status, pack-years smoked, environmental tobacco smoke exposure, education level, and chronic bronchitis. y COPD is defined according to the GOLD fixed cut-off criterion (FEV1/FVC , 0.70) and the LLN criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population; based on the Quanjer prediction equation, severity of COPD is classified as stage I if FEV1 value > 80% of the predicted value, and as stage II or more (stage II1) if FEV1 value , 80% of the predicted value; participants absent of COPD at follow-up are considered noncases in all models. z Two-way interactions between chronic bronchitis and biological dusts, mineral dusts, gases/fumes, and VGDF, were statistically significant (P , 0.05) in stage I GOLD-COPD models. x Two-way interaction between chronic bronchitis and biological dusts was statistically significant (P , 0.05) in stage I LLN-COPD model.

magnitude of risk of stage II1 LLN-COPD was highest in exposed ever-smokers, ranging between two- and fourfold in risk depending on the exposure type (Table E11). Finally, elevated

risks of stage II1 LLN-COPD in association with occupational exposures were also observed, with wide confidence intervals, in participants who continued employment in the same occupation

TABLE 5. ADJUSTED PAFS* AND 95% CONFIDENCE INTERVALS FOR THE ASSOCIATIONS BETWEEN OCCUPATIONAL EXPOSURES AND INCIDENCE OF STAGE II1 COPD ACCORDING TO COPD DEFINITION AND SPIROMETRIC REFERENCE EQUATION Stage II1 GOLD-COPD† Exposure Biological dusts Mineral dusts Gases/fumes VGDF

Ever-Smokers 0.31 0.15 0.21 0.24

(0.15–0.44) (0.00–0.36) (0.00–0.38) (0.003–0.41)

Nonsmokers 0.43 0.45 0.56 0.51

(0.00–0.71) (0.07–0.68) (0.00–0.84) (0.00–0.83)

Stage II1 LLN-COPD† Ever-Smokers 0.32 0.16 0.21 0.23

(0.06–0.50) (0.00–0.33) (0.00–0.40) (0.00–0.44)

Nonsmokers 0.11 0.25 0.38 0.32

(0.00–0.66) (0.00–0.77) (0.00–0.89) (0.00–0.87)

Definitions of abbreviations: COPD ¼ chronic obstructive pulmonary disease; GOLD ¼ Global Initiative for Chronic Obstructive Lung Disease; LLN ¼ lower limit of normal; PAF ¼ population attributable fraction; VGDF ¼ vapors, gases, dusts, or fumes. * PAFs were adjusted for the following baseline covariates: age, body mass index, sex, respiratory infection at early age, parental asthma, smoking status, pack-years smoked, environmental tobacco smoke exposure, and education level; 0.00: negative values truncated to zero; 95% confidence intervals calculated on log(1 2 PAF) scale. y COPD is defined according to the GOLD fixed cut-off criterion (ratio of forced expiratory volume over 1 s over forced vital capacity [FEV1/FVC] , 0.70) and the LLN criterion (FEV1/FVC , LLN) that is based on the Quanjer prediction equation using the European Coal and Steel Community reference population; based on the Quanjer prediction equation, severity of COPD is classified as stage II or more (stage II1) if FEV1 value , 80% of the predicted value.

1298


throughout follow-up; no interaction was present between continued employment and occupational exposure (Table E12).

DISCUSSION The findings from the follow-up survey of SAPALDIA indicate that occupational exposures to biological dusts, mineral dusts, gases/fumes, and VGDF at high levels were associated with incidence of stage II1 GOLD COPD. Statistically significant increases in risk of stage II1 GOLD-COPD per 10 years of cumulative exposure were also identified. Associations were generally observed in male participants, and in ages >40 years at baseline. The overall findings were similar for stage II1 LLNCOPD, but with smaller magnitude of risks observed. Risk of stage II1 COPD in association with occupational exposures also remained elevated when restricted to nonsmokers. Occupational exposures were also associated with incidence of stage I GOLD- and LLN-COPD, but only in presence of chronic bronchitis symptoms. The present findings add to the existing epidemiologic literature supporting a causal relationship between occupational exposures and development of COPD (1, 27). However, there is limited evidence from population-based studies demonstrating incidence of spirometric-defined COPD in association with occupational exposures. A previous Italian population-based study of males in a work surveillance program identified an association between self-reported VGDF confirmed by expert assessment and COPD, which was defined by presence of chronic bronchitis symptoms, FEV1/FVC , 70%, and predicted FEV1 , 80% (5). The adjusted IRRs for occupational exposures were moderately higher for stage II1 COPD when the GOLD definition was applied which requires further discussion. Previous analyses of the Burden of Obstructive Lung Disease (BOLD) and NHANES III study populations identified age-related increases in prevalence of stage II1 GOLD-COPD relative to stage II1 LLN-COPD (28, 29). The adjusted IRRs of stage II1 COPD for occupational exposures in this analysis were more similar between the two COPD definitions after restricting to ages >40 years at baseline. This finding suggests that the difference in estimated IRRs for occupational exposures between stage II1 GOLD- and LLN-COPD may possibly be age-related, and that age stratification may be a method to reduce potential overestimation of the effect of occupational exposure on GOLD-COPD. The choice of spirometric reference equation also appears to influence the magnitude of IRR for the associations between stage II1 COPD and occupational exposures, as the estimated IRRs from reference equations from Hankinson and coworkers (24) and Bra¨ndli and coworkers (25, 26) were moderately lower relative to prediction equations from Quanjer and colleagues (19). A primary limitation of Quanjer’s reference values was that predictions were developed for FEV1/VC rather than FEV1/ FVC. Additionally, reference values from Quanjer and coworkers are lower than Hankinson and coworkers, resulting in substantially less stage II1 COPD cases with Quanjer’s prediction. Overall, the choice of reference equation does not change the interpretation of the observed associations between occupational exposures and stage II1 COPD. The main effects of occupational exposures were predominantly observed in male participants, which is also of interest. The ALOHA GPJEM does not produce separate estimates for males and females for each ISCO-88 job classification. Females have been shown to perform work with less respiratory hazards than males, even within same occupation and industry (30). Additionally, we may have had insufficient power to identify an association between occupational exposures and incidence of stage II1 COPD in females, as far fewer females, relative to

VOL 185

2012

males, reported occupations assigned a high level of occupational exposure. The main effects of occupational exposures on stage II1 COPD also remained elevated when restricted to nonsmokers, but we did not observe a synergistic effect between smoking and occupational exposures on a multiplicative scale. Alternatively, the joint effect models of occupational exposure and smoking indicate that the risk of stage II1 LLN-COPD was highest in exposed ever-smokers in comparison to unexposed never-smokers, and that the joint effect of occupational exposure and smoking appeared more than additive for biological dusts, mineral dusts, and VGDF. Previous studies have observed a combined effect between occupational exposure to VGDF and smoking on COPD on an additive scale (4, 5), but the findings from the joint effect models in this analysis should be interpreted with caution as the confidence intervals between the different categories of occupational exposure and smoking widely overlap each other. In this study, the associations for occupational exposures were observed mainly with more severe COPD. Additionally, elevated risks of stage I and stage II1 COPD in association with occupational exposures were observed in participants with and without chronic bronchitis symptoms, respectively. The inconsistent findings are unlikely associated with COPD definition as the findings were similar for GOLD- and LLN-COPD as Figure 1 and Table 4 illustrate. When we restrict to participants with high level of exposure to VGDF at baseline, a crude comparison indicates that median weighted cumulative exposure during period of follow-up was marginally higher in participants absent of symptoms compared with those with symptoms present (39.3 vs. 24.0 yr, respectively, Kruskal-Wallis P ¼ 0.06). Although it may be hypothesized that symptoms of chronic bronchitis may be a susceptible factor of stage I COPD, presence of symptoms may possibly affect subsequent exposure and risk of more severe COPD. We believe a systematic analysis to evaluate symptom-related exposure selection and COPD-related outcomes in population-based studies would help to illuminate these findings. Little is known of the population burden of COPD attributable to occupational exposures in Switzerland. In this study, the PAF of at least moderate COPD attributable to VGDF is estimated between 23 and 24% in smokers, and between 32 and 51% in nonsmokers, depending on COPD definition. Previous studies have reported PAFs of between 29 and 32% in nonsmokers with definitions analogous to stage II1 GOLD-COPD (3, 31). The higher PAFs reported for nonsmokers in this study is likely due to a combination between the high prevalence of VGDF exposure in stage II1 cases (73.3%) and the threefold magnitude in risk, in addition to other study design and population characteristics. The PAFs for occupational exposures in smokers were similar between the COPD definitions, but the discrepancy observed for nonsmokers suggests that the PAFs for occupational exposures in nonsmokers may be overestimated or underestimated depending on COPD definition. Prior to this analysis, it was less understood how the distribution of relevant occupational exposures for COPD in Switzerland compare with that of neighboring countries. The estimated prevalences of exposures to biological dusts, mineral dusts, gases/ fumes, and VGDF as shown in Table E3 were not considerably different in comparison to observations from the ECRHS cohort (32). The majority of exposed participants reported occupations were also assigned multiple types of exposures according to the ALOHA GPJEM, which may minimize the specificity of the exposure-based models, particularly with mineral dusts. Given the extensive overlap between different exposures for many of the occupations, multiple testing of the same hypothesis may also be of concern in this analysis. However, the motivation for the JEM is to translate jobs into exposures to prevent the multiple


testing concerns of the occupation itself (ISCO-88 occupations assigned in study population, n ¼ 285). The use of the VGDF, a composite variable based on scores of the three specific exposures, also partially addresses the multiple testing concerns. Although this study has a number of strengths, including prospective study design to evaluate incidence of spirometry-based COPD, use of a GPJEM with semiquantitative estimates for exposure, and extensive control for confounding, there are a number of limitations to be considered. The use of prebronchodilator spirometric measurements to define COPD may have misclassified participants with asthma, whose obstruction is fully reversible, as having COPD. It is likely that the extent of the bias is minimal because participants who reported ever having asthma, which was more sensitive at identifying individuals with asthma than physician’s diagnosis of asthma, at either survey were excluded from the main analysis. Exclusion of participants who reported wheezing without cold 12 months previous to either survey in a sensitivity analysis yielded similar patterns of associations. Furthermore, if misclassification of COPD is random relative to exposure assignment, downward bias would likely result in an underestimation of the true association. Self-reported exposure to VGDF is widely used as an occupational exposure metric in epidemiological studies of chronic respiratory diseases (33–35) and is observed to have at least moderate agreement with JEMs (36, 37). For this study, we chose to use the ALOHA GPJEM because it produces semiquantitative estimates for multiple occupational exposures. However, the use of the ALOHA GPJEM may result in misclassification of exposure because the JEM does not account for differences in exposure levels observed between individuals with the same ISCO-88 classification for their reported job at baseline (38, 39). This may be more problematic for jobs with low levels of occupational exposures, in which we only observed an association with biological dusts. Additionally, current occupation at baseline may not be an appropriate proxy for historical exposure for all jobs that occurred before the baseline survey, particularly for older participants in SAPALDIA who reported short duration of employment for current job. The use of current job at baseline may also understate the effect of exposure after the baseline survey should continued exposure in the same occupation during follow-up affect susceptibility to increased risk, which the results from this analysis partially suggest. Because the ALOHA GPJEM avoids recall bias, any bias from exposure misclassification is likely to be independent of COPD status, and result in an underestimate of the true association. In conclusion, long-term exposure to occupational exposures to biological dusts, mineral dusts, gases/fumes, and VGDF at high levels were associated with increased incidence of moderate COPD in a population of Swiss working adults, particularly in males and in ages >40 years. The observed findings support existing evidence of a causal relationship between occupational exposures and development of COPD. Estimation of the population burden of COPD attributable to occupational exposures in nonsmokers may be sensitive to choice of COPD definition.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14. 15. 16.

Author disclosures are available with the text of this article at www.atsjournals.org. Acknowledgment: The authors thank the whole SAPALDIA Team for their contribution to the study. Additionally, the study could not have been done without the help of the study participants, technical and administrative support, and the medical teams and field workers at the local study sites. See online supplement for detailed description of the SAPALDIA Team.

References 1. Eisner MD, Anthonisen N, Coultas D, Kuenzli N, Perez-Padilla R, Postma D, Romieu I, Silverman EK, Balmes JR; Committee on Nonsmoking COPD, Environmental and Occupational Health Assembly. An official American Thoracic Society public policy statement:

17.

18.

19.

1299

novel risk factors and the global burden of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2010;182:693–718. Balmes J, Becklake M, Blanc P, Henneberger P, Kreiss K, Mapp C, Milton D, Schwartz D, Toren K, Viegi G. American Thoracic Society statement: occupational contribution to the burden of airway disease. Am J Respir Crit Care Med 2003;167:787–797. Hnizdo E, Sullivan PA, Bang KM, Wagner G. Association between chronic obstructive pulmonary disease and employment by industry and occupation in the US population: a study of data from the Third National Health and Nutrition Examination Survey. Am J Epidemiol 2002; 156:738–746. Blanc PD, Iribarren C, Trupin L, Earnest G, Katz PP, Balmes J, Sidney S, Eisner MD. Occupational exposures and the risk of COPD: dusty trades revisited. Thorax 2009;64:6–12. Boggia B, Farinaro E, Grieco L, Lucariello A, Carbone U. Burden of smoking and occupational exposure on etiology of chronic obstructive pulmonary disease in workers of southern Italy. J Occup Environ Med 2008;50:366–370. Sunyer J, Kogevinas M, Kromhout H, Anto JM, Roca J, Tobias A, Vermeulen R, Payo F, Maldonado JA, Martinez-Moratalla J, et al. Pulmonary ventilatory defects and occupational exposures in a populationbased study in Spain: Spanish Group of the European Community Respiratory Health Survey. Am J Respir Crit Care Med 1998;157:512–517. Matheson MC, Benke G, Raven J, Sim MR, Kromhout H, Vermeulen R, Johns DP, Walters EH, Abramson MJ. Biological dust exposure in the work-place is a risk factor for chronic obstructive pulmonary disease. Thorax 2005;60:645–651. Weinmann S, Vollmer WM, Breen V, Heumann M, Hnizdo E, Villnave J, Doney B, Graziani M, McBurnie MA, Buist AS. COPD and occupational exposures: a case–control study. J Occup Environ Med 2008;50:561–569. Bridevaux PO, Probst-Hensch NM, Schindler C, Curjuric I, Felber Dietrich D, Braendli O, Brutsche M, Burdet L, Frey M, Gerbase MW, et al. Prevalence of airflow obstruction in smokers and neversmokers in Switzerland. Eur Respir J 2010;36:1259–1269. Ackermann-Liebrich U, Kuna-Dibbert B, Probst-Hensch NM, Schindler C, Felber Dietrich D, Stutz EZ, Bayer-Oglesby L, Baum F, Bra¨ndli O, Brutsche M, et al.; SAPALDIA Team. Follow-up of the Swiss Cohort Study on Air Pollution and Lung Diseases in Adults (SAPALDIA 2) 1991–2003: methods and characterization of participants. Soz Praventivmed 2005;50:245–263. Mehta AJ, Miedinger D, Keidel D, Bettschart R, Bircher A, Bridevaux PO, Curjuric I, Kromhout H, Probst-Hensch N, Rothe T, et al.; SAPALDIATeam. Occupational exposure to dusts, gases, and fumes, and incidence of COPD in SAPALDIA [abstract]. Eur Respir J 2011;38:329s. Probst-Hensch NM, Curjuric I, Pierre-Olivier B, Ackermann-Liebrich U, Bettschart RW, Bra¨ndli O, Brutsche M, Burdet L, Gerbase MW, Kno¨pfli B, et al. Longitudinal change of pre-bronchodilator spirometric obstruction and health outcomes: results from the SAPALDIA cohort. Thorax 2010;65:150–156. Martin BW, Ackermann-Liebrich U, Leuenberger P, Ku¨nzli N, Stutz EZ, Keller R, Zellweger JP, Wu¨thrich B, Monn C, Blaser K, et al. SAPALDIA: methods and participation in the cross-sectional part of the Swiss Study on Air Pollution and Lung Diseases in Adults. Soz Praventivmed 1997;42:67–84. Burney PG, Luczynska C, Chinn S, Jarvis D. The European Community Respiratory Health Survey. Eur Respir J 1994;7:954–960. American Thoracic Society. Standardization of spirometry, 1994 update. Am J Respir Crit Care Med 1995;152:1107–1136. Ku¨nzli N, Ackermann-Liebrich U, Keller R, Perruchoud AP, Schindler C. Variability of FVC and FEV1 due to technician, team, device and subject in an eight centre study: three quality control studies in SAPALDIA. Swiss Study on Air Pollution and Lung Disease in Adults. Eur Respir J 1995;8:371–376. Ku¨nzli N, Kuna-Dibbert B, Keidel D, Keller R, Bra¨ndli O, Schindler C, Schweinzer KM, Leuenberger P, Ackermann-Liebrich U; SAPALDIA Team. Longitudinal validity of spirometers—a challenge in longitudinal studies. Swiss Med Wkly 2005;135:503–508. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease [Internet]. The Global Initiative for Chronic Obstructive Lung Disease [updated 2010; accessed 2011 June 17]. Available from: www.goldcopd.org. Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows. Report Working Party

1300

20. 21.

22.

23. 24.

25.

26.

27. 28.

29.

30.


Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society. Eur Respir J Suppl 1993;16:5–40. International Labour Office. International standard classification of occupations: ISCO-88. Geneva: International Labour Organization; 1990. Huber PJ. The behavior of maximum likelihood estimates under nonstandard conditions. In: Leycam L, Neyman J, editors. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Vol. 1. Berkeley, CA: University of California Press; 1967. pp. 221–233. White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 1980;48:817– 830. Rogers WH. Regression standard errors in clustered samples. Stata Tech Bull 1993;13:19–23. Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general US population. Am J Respir Crit Care Med 1999;159:179–187. Bra¨ndli O, Schindler C, Ku¨nzli N, Keller R, Perruchoud AP. Lung function in healthy never smoking adults: reference values and lower limits of normal of a Swiss population. Thorax 1996;51:277–283. Bra¨ndli O, Schindler C, Leuenberger PH, Baur X, Degens P, Ku¨nzli N, Keller R, Perruchoud AP. Re-estimated equations for 5th percentiles of lung function variables. Thorax 2000;55:173–174. Blanc PD. Occupation and COPD: a brief review. J Asthma 2012;49:2–4. Vollmer WM, Gı´slason T, Burney P, Enright PL, Gulsvik A, Kocabas A, Buist AS. Comparison of spirometry criteria for the diagnosis of COPD: results from the BOLD study. Eur Respir J 2009;34:588–597. Hnizdo E, Glindmeyer HW, Petsonk EL, Enright P, Buist AS. Case definitions for chronic obstructive pulmonary disease. COPD 2006;3: 95–100. Eng A, Mannetje A, McLean D, Ellison-Loschmann L, Cheng S, Pearce N. Gender differences in occupational exposure patterns. Occup Environ Med 2011;68:888–894.

VOL 185

2012

31. Mak GK, Gould MK, Kuschner WG. Occupational inhalant exposure and respiratory disorders among never-smokers referred to a hospital pulmonary function laboratory. Am J Med Sci 2001;322:121–126. 32. Sunyer J, Zock JP, Kromhout H, Garcia-Esteban R, Radon K, Jarvis D, Toren K, Ku¨nzli N, Norba¨ck D, d’Errico A, et al.; Occupational Group of the European Community Respiratory Health Survey. Lung function decline, chronic bronchitis, and occupational exposures in young adults. Am J Respir Crit Care Med 2005;172:1139–1145. 33. Korn RJ, Dockery DW, Speizer FE, Ware JH, Ferris BG Jr. Occupational exposures and chronic respiratory symptoms: a population based study. Am Rev Respir Dis 1987;136:298–304. 34. Bakke P, Eide GE, Hanoa R, Gulsvik A. Occupational dust or gas exposure and prevalences of respiratory symptoms and asthma in a general population. Eur Respir J 1991;4:273–278. 35. Fishwick D, Bradshaw LM, D’Souza W, Town I, Armstrong R, Pearce N, Crane J. Chronic bronchitis, shortness of breath, and airway obstruction by occupation in New Zealand. Am J Respir Crit Care Med 1997;156:1440–1446. 36. Quinlan PJ, Earnest G, Eisner MD, Yelin EH, Katz PP, Balmes JR, Blanc PD. Performance of self-reported occupational exposure compared to a job-exposure matrix approach in asthma and chronic rhinitis. Occup Environ Med 2009;66:154–160. 37. Blanc PD, Eisner MD, Balmes JR, Trupin L, Yelin EH, Katz PP. Exposure to vapors, gas, dust, or fumes: assessment by a single survey item compared to a detailed exposure battery and a job exposure matrix. Am J Ind Med 2005;48:110–117. 38. Peters S, Vermeulen R, Cassidy A, Mannetje A, van Tongeren M, Boffetta P, Straif K, Kromhout H; INCO Group. Comparison of exposure assessment methods for occupational carcinogens in a multi-centre lung cancer case-control study. Occup Environ Med 2011;68:148–153. 39. Kromhout H, Symanski E, Rappaport SM. A comprehensive evaluation of within- and between-worker components of occupational exposure to chemical agents. Ann Occup Hyg 1993;37:253–270.