Occupational exposures have been shown to be risk factors for chronic obstructive pulmonary disease (COPD) among never-smokers. In a Danish population-based cohort, we analysed this association and the population attributable fraction. The study population (N=1575) was aged 45–84, COPD was defined by lung function measurements and the method of lower limit of normal (LLN), and occupational exposure was assessed by questionnaire and expert judgement. Furthermore, the estimates additionally were provided according to the Global Initiative for Chronic Obstructive Lung Diseases. More than a threefold increased risk (LLN OR=3.69 (95% CI 1.36 to 10.04) was found for occupational exposure to vapour, gas, dust and fumes (predominantly organic dust) in this never-smoking population, with a corresponding 48% (95% CI 30% to 65%) population attributable fraction among never-smokers.
Statistics from Altmetric.com
What this paper adds
Occupational exposure is an essential risk factor for chronic obstructive pulmonary disease (COPD) among never-smokers.
The method of lower limit of normal (LLN) assessing COPD is recommended by the European Respiratory Society (ERS) and the American Thoracic Society (ATS), but only used in a few studies among never-smokers.
In this Danish population-based cohort of 1575 never-smokers, we estimated a population attributable fraction (PAF) of 48% for occupational exposure to vapour, gas, dust or fumes, which is higher than that estimated in earlier studies.
PAF seems to be higher when the LLN approach was used compared with the Global Initiative for Chronic Obstructive Lung Diseases (GOLD) approach assessing COPD, although wide and overlapping CIs were given.
Occupational exposure to vapour, gas, dust and fumes seems to have an even higher impact on COPD among never-smokers than was previously estimated.
Occupational exposures to vapour, gas, dust and fume (VGDF) have been shown to be risk factors for chronic obstructive pulmonary disease (COPD) among never-smokers. Published data suggest a population attributable fraction (PAF) for COPD caused by occupational exposure to be 26–43% among never-smokers.1 ,2 We analysed for occupational COPD among never-smokers in a Danish population-based study from the cross-sectional North Jutland COPD Prevention Study (NCPS),3 including an expert assessed job exposure.
In 2004–2006, the baseline NCPS study recruited by random sampling an age-stratified and sex-stratified, mixed urban/rural population aged 45–84 from 155 general practitioners (GP). Data consisted of a self-administered questionnaire, pulmonary function tests, height and age. Pulmonary function tests by spirometry (forced expiratory volume 1 s (FEV1) and forced vital capacity (FVC)) were performed by the GP or a trained member of the practice staff with the GP’s own spirometer. Volume and time calibration of the spirometers was performed before the study start and every 6 months by trained staff using a 1 L syringe. Adequate spirometry test instructions followed the statement from the European Respiratory Society (ERS) and the standard from the American Thoracic Society (ATS).4 ,5 Never-smokers accounted for 33% (N=1577) of whom two were excluded due to prior lung cancer. COPD was defined by spirometry according to the method of lower limit of normal (LLN) as prebronchodilator FEV1/FVC z-score <2 SDs and FEV1 z-score <2.6 Data were additionally presented based on the Global Initiative for Chronic Obstructive Lung Diseases (GOLD) criteria with a fixed FEV1/FVC ratio<0.70 and FEV1<80% of the predicted value (GOLD 2+).7 Both methods of defining COPD reflect moderate airway obstruction and thus comparable severity of obstruction. COPD defined by GOLD was estimated using the postbronchodilator values if the prebronchodilator FEV1/FVC ratio was <0.70; otherwise, the prebronchodilator values were used. Only participants with an FEV1/FVC ratio <0.70 were projected to perform this reversibility test to minimise costs and time consumption, although postbronchodilator values are recommend by GOLD to exclude reversible obstruction in asthmatics. The Global Lung Function 2012 Equations8 in the ‘GLI-2012 Desktop Software for Large Data Sets V.1.3.4 build 3’ was used as the reference population. The occupational exposure assessment was based on a self-administered questionnaire, where the participants were asked for job titles and duration of employment (0, ≤5, 5–9, 10–14, 15–19, and ≥20 years of employment) in jobs with organic dust, inorganic dust, fume/gas and vapour exposure, respectively. For each exposure category, the participants could state their three longest held job titles and appertaining durations. Blinded to COPD status, jobs were coded using the Danish version of The International Standard Classification of Occupations (DISCO-88)9 and expert-derived assessment by two specialists in occupational medicine was used to identify jobs with known exposure to VGDF. A final decision was agreed on by the two specialists. The total period of employment was calculated by adding the mean years from each 5-year span of each stated employment. The total duration of employment per exposure was intentionally calculated to define four categories of cumulated duration of exposure: no exposure (0 years), low exposure (<5 years), medium exposure (5–14) and high exposure (≥15 years). However, owing to the small number of cases and exposed participants at each level, we had to dichotomise the exposure category to ever or never. An overall exposure was summarised in a VGDF exposure. The participants history of smoking habits and possible prior asthma were obtained through a questionnaire.
The McNemar test for matched data was used to compare the two methods of assessing COPD; LLN and GOLD were as defined earlier. Data were analysed in a univariate and mixed random effect logistic regression model, with GP practice as the random variable, and sex and age as fixed effects. The PAF for COPD was estimated as the proportion of cases exposed×(OR-1)/OR (OR used as proxy for relative risk).10 Values for the individual PAF equations can be accessed from table 1 (eg, VGDF exposure and LLN: exposed cases/all cases×adjusted OR-1/adjusted OR=(15/15+8)×(3.69−1)/3.69=48%). The significance level was set at 5%. The 95% CIs were calculated using a normal approximation. Statistical analyses were conducted in Stata V.12.1 (StataCorp LP, 2011). The NCPS study was performed in accordance with the Helsinki Declaration and approved by the Danish Scientific Ethics Committee (VN2003/62) and the Danish Data Protection Agency (updated in 2007 before follow-up: 2007-41-1576). Written informed consent was obtained from all participants.
Of 372 DISCO-88 codes, 72 were identified with VGDF exposure and occupational exposure was present in 658 (42%) participants in between 1 (72%) and 5 jobs; the most frequent job codes are presented in table 2. Organic dust exposure was the dominating exposure (86%) while exposure to vapour, gas/fume and inorganic dust was less common at 5%, 16%, and 21%, respectively. The occupational exposures and the age distribution are presented in online supplementary table S1, stratified by sex and combined. There was a statistical difference between gender according to age and any occupational exposure, p<0.001. COPD, defined by LLN, was present in 26 participants, equal to a prevalence of 1.7%, while the prevalence was 3.4% when COPD was defined by the GOLD criteria (n=53), with a COPD discordance among 31 participants between the two definitions, p<0.05. Table 1 shows the prevalences of COPD in the different exposures of VGDF and organic dust, and of crude along with adjusted associations between the occupational exposures and COPD. Participants exposed to VGDF and organic dust had an increased prevalence and risk of COPD. Using the GOLD definition of COPD increased the prevalence and slightly decreased the association with the exposures compared with the LLN definition of COPD. Specific inorganic dust, fume/gas and vapour occupational exposures showed no evidence of an association with COPD (data not shown). Excluding 145 never-smokers with prior self-reported asthma (reported as never or ever) in the LLN estimation provided similar associations; VGDF: OR=2.64 (95% CI 0.70 to 9.92), organic dust: OR=3.43 (95% CI 0.86 to 13.70). Another important confounder is passive smoking, but only a few (n=45) in this population-based study in the age group of 45–84 had never experienced an exposure to passive smoking either at home or at work. Passive smoking was not associated with COPD and additional adjustment for passive smoking changed only to a minor extent the estimates. When COPD was defined by LLN the study PAFs for COPD caused by occupational exposure was 48% (95% CI 30% to 65%) for VGDF exposure and 41% (95% CI 19% to 62%) for organic dust exposure among never-smokers. Defining COPD by the GOLD criteria the corresponding PAF was 44% (95% CI 29% to 58%) and 40% (95% CI 23% to 57%), respectively.
In the present study, the risk for COPD in never-smokers was increased more than three times when occupationally exposed to VGDF and increased threefold when occupationally exposed to organic dust despite the low prevalence of COPD (1.7%). However, the calculated associations between occupational exposure and COPD have wide CI due to the few cases in each stratum, as expected in a population-based setting. Our prevalence estimate was low compared with the prevalence in never-smokers from the BOLD (Burden of Obstructive Lung Disease) study (0–11%).11 COPD was defined by GOLD criteria stage II or higher in a slightly younger population. When converting our prevalence estimate based on LLN criteria to GOLD criteria, the prevalence of COPD increased to 3.4%, reducing the difference in the prevalence between the studies. The LLN approach is recommended concurrently by the ERS and ATS,12 as opposed to the GOLD fixed method, which is somewhat biased by the participant’s age, sex and height.13 We have found the highest PAF among studies published in never-smokers of 48% (95% CI 30% to 65%),1 ,2 which is higher than the PAF of 43% (95% CI 0% to 68%) from the study by Weinmann et al,14 a study with a similar design to this study (defining COPD by LLN and expert assessed occupational exposure). However, a big Chinese population-based study including 6648 never-smokers found no significant association between occupational exposure and COPD defined by LLN when using a self-reported occupational exposure assessment among never-smokers, OR=1.29 (95% CI 0.92 to 1.81).15 This might be a result of non-differential misclassification in the exposure assessment. Recall bias of occupational exposure might be introduced, tending to overestimate the association between exposure and disease, although the risk was considered low, as the questionnaires were filled out before the GP examination. We think that only a few Danes aged 45–84 have knowledge of an association between occupational exposure to VGDF and COPD, and owing to the a priori selected DISCO-88 codes and expert management of each job title into DISCO-88, without awareness of COPD status, the exposure assessment has further reduced the risk of bias. The prevalence of COPD was the same (p=0.13) among participants who have answered questions on work exposure and among those who did not. Information bias of exposure was reduced by using the specialist assessed exposure on the basis of job titles, instead of the commonly used self-reported exposure assessment. A validation of the self-reported exposures showed that job titles often were connected to the wrong exposure and the specialists assessed many of the job titles as having no occupational exposure. As a consequence, a significant discordance in the four exposure groups was observed when comparing the dichotomised expert assessed exposure and the self-reported exposure in the unrestricted data set (N=4717).
The external study validity is considered to be high in a Western world setting with similar occupational distribution and occupational exposures, although the major contribution of organic dust exposure might reflect a special Danish occupational exposure scenario. The enrolled study population included more young women and fewer participants in the oldest group than among non-responders. This might introduce an age-dependent healthier study population tending to underestimate the associations. Spirometry error measurements have been managed by regular calibration, but the variations of brands among GP practices were neglected for the benefit of a local experienced operator. However, the internal biological variability in lung function was addressed by requiring three sufficient measurements as recommended by the ERS and the ATS.12 A possible misclassification would be of a non-differential nature and tend to underestimate the association. Moreover, when using a spirometrically defined COPD, some patients with COPD with compliance difficulties might be excluded from the analysis, resulting in false low associations.
Occupational exposure to VGDF and organic dust significantly increased the risk of COPD, corresponding to a high PAF, indicating that occupational exposures contribute substantially to the burden of COPD in never-smokers. The major contribution of organic dust exposure among the VGDF exposures in this study might reflect a special Danish occupational exposure scenario.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Contributors ETW was involved in the data management, conversion of occupational description into DISCO-88, statistics and manuscript writing. VS was involved in the project planning, selection of DISCO-88 codes, conversion of occupational description into DISCO-88, statistics and manuscript review. THM contributed in the data management, statistics and manuscript review. JGH contributed in the project planning, project accomplishment, data management and manuscript review. ØO contributed in the Project planning, selection of DISCO-88 codes, data management, conversion of occupational description into DISCO-88, statistics and manuscript review. All authors read and approved the final manuscript.
Competing interests None.
Patient consent Obtained.
Ethics approval Danish Scientific Ethics Committee (VN2003/62).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.