Occupation and chronic obstructive pulmonary disease (COPD)

Transcription

1 Published Online October 18, 2012 Occupation and chronic obstructive pulmonary disease (COPD) Paul Cullinan * Occupational and Environmental Medicine, Imperial College (NHLI), London, UK Introduction: There is growing interest in preventable, non-smoking causes of chronic obstructive pulmonary disease (COPD), among which are chronic exposures to respiratory irritants in the workplace. Sources of data: Reviews of occupational COPD in specific occupations and industries and in general populations; supplemented with other or more recently published material. Areas of agreement: There is good evidence for an increased risk of COPD from certain specific exposures (coal mine dust, silica, welding fume, textile dust, agricultural dust, cadmium fume). Areas of controversy: Less clear is the causal role of non-specific dusts or fumes/ gases in general populations where the available literature is notably uncritical. Growing points: Other specific exposures, such as diesel fume; interactions between specific exposures and cigarette smoking; the development of safe working limits. Areas timely for developing research: Occupations with large numbers of exposed employees, particularly in low-income countries. Keywords: occupation/copd/vgdf Accepted: September 17, 2012 Introduction *Correspondence address. p.cullinan@ imperial.ac.uk The clinical entity of chronic obstructive pulmonary disease (COPD), one of the progressive and largely irreversible airflow obstruction in the middle and late ages, is predominantly a disease of cigarette smokers. This relationship was established in a series of workplacebased studies directed by Charles Fletcher and Peto in the UK, 1 and has subsequently been confirmed in numerous population-based studies in other countries. Indeed, it is claimed that in high-income countries 75% of cases of and deaths from COPD are directly attributable to smoking, 2 and some have gone as far as stating that it is probably the British Medical Bulletin 2012; 104: DOI: /bmb/lds028 & The Author Published by Oxford University Press. All rights reserved. For permissions, please journals.permissions@oup.com

2 P. Cullinan only chronic disease for which the finger of blame can be pointed at a single risk factor tobacco smoking. 3 While it would be dangerous to deflect attention away from those who manufacture and sell cigarettes, it is important to recognize that there are other likely causes of the disease that require attention. Contemporary studies of lung function, using standardized spirometry in representative populations, suggest that the international variation in the prevalence of COPD cannot be explained solely by differential rates of smoking in the community. 4 Thus, it is suggested, other factors must be aetiologically important, alone or in concert with smoking. Among these have been included exposures to irritant agents in workplace air. The symptoms of COPD are not specific and the disease, certainly in its early stages, produces few if any characteristic signs. Thus, in the absence of quantitative measures of pulmonary function and of radiology, it is essentially impossible to distinguish it from other, longlatency respiratory diseases in which the primary symptom is breathlessness on exertion. This is particularly the case in settings where other such diseases may be relatively common. Charles Thackrah 5 made detailed observations of respiratory symptoms among employees in a very diverse range of trades in 19th century England and even practised what may be the first ever attempt at pre-employment screening of lung function, measured with a pulmometer in a population of Light Dragoons; but it remains impossible to know whether he was describing, as seems likely, what we now recognize as diseases specific to certain occupations ( predominantly, it seems, silicosis and byssinosis). Curiously, John Hutchinson, the inventor of the spirometer in 1846 reported that although he could relate lung function under tidal breathing ( vital capacity ) to height and, more weakly, to age in 1200 men, he found no effect of occupational exposures. 6 Defining and recognising COPD In the absence of any clinical specificity, it fell to the anatomists to provide the first descriptions of a condition that we would recognize today as characteristic of (much) COPD. In 1691, Frederik Ruysch 7 the Dutch anatomist and curator, reported among a series of post-mortem specimens he later sold to Peter the Great, a case of what was probably emphysema in conjunction with severe and irreversible airway narrowing. It has been claimed, on the basis of a high prevalence of apparent emphysema in autopsy series reported in the 18th and 19th centuries, that COPD was common in Europe well before cigarette smoking became prevalent, 8 but in reality it is difficult to know how far these 144 British Medical Bulletin 2012;104

3 Occupation and chronic obstructive pulmonary disease (COPD) observations were an artefact of contemporary methods of preservation after death. Nonetheless and fittingly, it is in the same context that of the 9th Emphysema Conference in Aspen, Colorado in 1966 that the term chronic obstructive pulmonary disease was first used by William Briscoe. 9 By that time the confluence of limitation of airflow through the bronchial tree that, in contrast to asthma, is largely irreversible and generally progressive, of chronic sputum production and, often, of pathological emphysema was clear and an encompassing term was a useful advance. More recently, for both epidemiological and clinical purposes, the definition of COPD has moved to one that is based almost entirely on spirometric measurements of FEV 1 and FVC made after the use of a short-acting, bronchodilating inhaler. 10 Based on a series of fixed cut-offs expressed as percentages of predicted values derived from surveys of more or less normal populations, the disease is categorized into several degrees of increasing severity. This shift has had profound implications. First, the recognition of the disease now strictly requires the use of lung function measurements; second, it is apparent that there is a large proportion of disease that is asymptomatic and thus otherwise unrecognized; third, a far higher proportion of persons exposed to the causes of COPD appear to have the condition than was formerly believed; 11 and fourth, the global market for a condition that was previously of little interest to the pharmaceutical industry has grown enormously. The use of fixed cut-offs has been criticized as being arithmetically inept because the variability of lung function measurements is not proportional to their predicted values based on an individual s age, sex and height. A consequence is that definitions of abnormality based on fixed cut-off values will misclassify a proportion of those whose measurements are outside the mid-range of expected values; and, importantly, will artefactually increase the prevalence of airflow limitation (FEV 1 /FVC,0.70) in older persons since the expected ratio declines with age. 12 More appropriate is the use of lower limits of normal (LLN) based, for example, on the 5th percentile of the approximately normal distribution of expected values of lung function. 13 In studies of populations where the prior probability of abnormal lung function is low as in many contemporary occupational settings it has even been argued that a more stringent (2.5th) percentile is appropriate to reduce the number of false-positive findings in older participants. British Medical Bulletin 2012;

4 P. Cullinan Approaches to the study of occupational COPD Broadly, studies in this area are divisible into those that examine particular workforces or industries, with relatively homogenous exposure(s) and those of community populations with diverse occupations and hence exposures. The former have the advantage, when at their best, of sufficiently large numbers and detailed information on exposure to workplace agents and important confounders. Perhaps more even than in most occupational studies, longitudinal data are preferable because they allow the prospective identification of exposed workers who have accelerated declines in lung function irrespective of where their measurements are in relation to normal ranges (see above). In practice, the majority of such studies have been cross-sectional in design and so open to concerns over cause and effect arising, for example, from healthy worker effects. Studies based in general populations have been of two sorts: the few that have examined, within their confines, the role of particular occupations and the larger number that have used exposure estimates derived from self-reports or, less commonly, expert judgement. These approaches have the advantage of more readily allowing an assessment of the total burden of occupational exposures in a population, a feature that has been heavily emphasized in the many reviews of this field. Indeed, the enthusiastic use of population attributable risks (PAR) is a striking feature of the relevant literature. The first, and most cited, systematic review 14 is replete with references to this index, as is its successor 15 in which the term is used 52 times. While not wishing to diminish any important public health message, it is worth noting that the value of PAR estimates lies in those that relate to exposures that are definitively causal and are thus, conceptually, amenable to prevention; as will be seen later, a second striking feature of the relevant literature is its relatively uncritical approach to the distinction between association and cause. It is not possible, in the space of this review, to discuss in any detail the extensive, industry-based literature but a summary of the areas of agreement is provided below with a discussion of areas that would benefit from further research. More attention will be paid to studies of occupational COPD in general populations, particularly those focused on non-specific exposures to irritant agents in workplace air. Here, despite a large number of published studies, there is considerable remaining uncertainty and a detailed discussion of the areas of doubt is provided. The sources of data are largely drawn from previous reviews with supplementary material that was either not included in them or has been published subsequently. 146 British Medical Bulletin 2012;104

5 Occupation and chronic obstructive pulmonary disease (COPD) Studies of particular workforces or exposures: areas of agreement and areas of uncertainty There is a large literature that describes the rates of COPD variously defined in different occupational groups. It is well reviewed elsewhere 16,17 and, usefully, by a recent, Danish collective 18 that focused on studies that used spirometry in their outcome definitions. Although a very wide variety of occupations have been studied in this way, for most the number of relevant studies is limited and the evidence remains inconclusive. Coal mining is one exception. Here the numbers available for study tend to be large and advantage can be taken of the routine pulmonary surveillance offered in many countries. Thus, for example, Marine et al. 19 examined the information collected from 3400 British miners over a 10-year period. Using detailed measurements of coal mine dust they demonstrated an exposure-related increase in the risk of FEV 1 reduction that was independent of smoking and eventually became the basis for the current statutory compensation offered to coal miners with COPD in the UK. Similar findings have been reported in US miners. 20 Less clear, despite plenty of research investment, is the evidence relating to some other large, occupational groups. The extensive crosssectional literature on welding fume, for example, remains difficult to interpret largely because of concerns that studies of this design preferentially include relatively healthy employees but a recent review of the available longitudinal studies suggests a small increase in the rate of FEV 1 decline among welders in comparison to referent trades, a risk that may be confined to those welders who smoke. 21 Grain handlers, too, appear to have an increased risk of airflow obstruction, with perhaps a similar interaction with smoking. 22,23 For the wider group of agricultural workers, including farmers, the evidence is less clear perhaps because of the diversity of both organic and inorganic exposures encountered in such work Underground miners and others who work with crystalline silica probably face an increased risk of airflow obstruction that is independent of radiological silicosis and potentiated by cigarette smoking. 28 Those exposed to textile dusts notably cotton have long been recognized as at risk for the development of obstructive lung diseases perhaps greatest in those with symptoms of byssinosis and possibly induced by exposures to endotoxin; much recent work in this setting has been done in Chinese factories. Although the evidence base is small, there is convincing evidence that exposure to cadmium fume (in, for example, the disposal British Medical Bulletin 2012;

6 P. Cullinan of batteries) increases, in a dose-dependent manner, the risk of COPD, through the induction of pulmonary emphysema. 32 For some other important occupational groups or specific exposures the evidence is less certain, in some cases perhaps because they entail well-known risks of other types of respiratory disease. Among these are included exposure to diisocyanates for which the evidence is inconsistent, 33,34 coke oven workers exposed to polycyclic aromatic hydrocarbons, 35 work with asbestos 36 and iron/steel processing. 37 Recently, interest has grown in the effects of diesel fume, a common exposure in the transport, construction and mining sectors. 38 Studies in general populations: areas of controversy Studies of general populations have formed the basis of most recent reviews of occupational COPD and are the source of claims that 15% of COPD (and of chronic bronchitis) can be attributed to work-related factors. 15 The majority of the studies that have contributed to such meta-analysis have used measures of exposure derived from self-reports of work with airborne irritants, usually expressed as some combination of vapours, gases, dusts or fumes (VGDF); less often, expert judgement of the same has been employed, with or without the use of explicit job exposure matrices (JEM) developed specifically for the disease. Sources of potential error Before examining these studies in detail, it is worth rehearsing here the potential for bias and other errors which are several and in some cases applicable both to studies of general populations and of particular occupational groups. Confounding by smoking: smoking is believed to be the predominant cause of COPD and requires consideration in any study of other potential aetiologies. In the occupational setting, confounding will arise if smoking habits are related to employment in jobs where there is exposure to, for example, VGDF. The extent of such confounding has been examined most closely within the field of occupational lung cancer. Perhaps unsurprisingly there are substantial differences in smoking rates between blue and white collar workers 39 and close relationships between smoking habits and exposure to a range of occupational hazards. 40 Axelson 41, in the application of data such as these and through the use of simple models, has suggested however that risk ratios from smoking in occupational studies would rarely be.1.5. While this provides some reassurance in the study of occupational carcinogens, where risk estimates are generally higher, the same 148 British Medical Bulletin 2012;104

7 Occupation and chronic obstructive pulmonary disease (COPD) may not be true for COPD, especially where, as is usually the case, estimates have been derived from comparisons with office workers. In any case, it confirms the importance of detailed measurements of (lifetime) smoking habits. Other confounders: far less attention has been paid to other potential confounders of the relationship between occupation and lung function, largely because the available information is seldom if ever available. There is good evidence that adult lung function is related to states and events in early life such as birth weight and respiratory infections in infancy which are, respectively, positively and negatively associated with FEV 1 in adulthood. More generally, socio-economic position across a life course appears to determine lung function decline, even in early middle age, 42 a relationship that may intimately and perhaps irrevocably confound any association with occupational exposures. Unsurprisingly, factors such as these have received little attention in community-based studies of occupational COPD although some authors have used a probable, non-occupational surrogate for socio-economic status in adjusted analyses. 43,44 Exposure-response and latency: by analogy with cigarette smoking, generic occupational exposures to VGDF as a cause of COPD are likely to have an effect that is related directly to cumulative exposure with a latent period of many years. Consequently, the examination of these issues would lend important weight to any causal argument. A high proportion of studies, however, have failed to report this information and in those that have the information is in most cases unconvincing (see below). Most but not all have been of cross-sectional design with exposure estimates based on the current or longest-held job with the attendant concerns over healthy worker effects. Self-reported exposures: self-reported exposures are relatively easy to obtain in epidemiological studies. The accuracy of self-reports of occupational exposure to VGDF has been questioned but while there is certainly some resultant misclassification of exposure, its extent seems to be limited. Table 1, for example, summarizes information derived from a survey of the self-reported working conditions of around 2200 men and women employed in the UK. 48 Those who reported that they had ever been exposed to breathing fumes, dusts or other harmful substances at work were significantly more likely to be employed in occupations where such exposures are plausible; the converse was also true, lending some face validity to the exposure measurement. Similar findings have been described in other settings. 49,50 While it is reassuring that the extent of any random misclassification is probably limited, the proper concern is over systematic error or bias. Worryingly, there is, in studies of asthma and other respiratory conditions, evidence that the discrepancy between self-reported exposures and a more objective assessment is greater among those with than without disease. 51,52 Differential exposure misclassification of this sort usually leads to a bias away from the null, as reflected in studies where risks measured with self-reported exposure are not apparent for those British Medical Bulletin 2012;

8 P. Cullinan Table 1 Occupational groups by proportion of UK workers reporting exposure to fumes, dusts or other harmful substances at work (adapted from ref. 48) Men (% yes) Women (% yes) Relative risk (men and women) Construction Metal processing Other processing Farming, fishing, forestry Electrical processing Other transport and machinery operative Material moving and storing Road transport operatives Hair and beauty Textile processing Repetitive assembly, inspection Security and protective services Managerial Science and engineering All occupations Other personal services Other education and welfare Catering Nursing Cleaning Teaching Professional and related supporting management Selling Literary, artistic and sports Clerical secretarial Care workers assessed more objectively. As with other sources of bias, the extent of any discrepancy arising in this way may be limited but is nonetheless important in studies where risk estimates are generally low. Job-exposure matrices (JEM): it is claimed, with some justification, that the use of JEMs or other expert approaches reduces any bias arising from self-reported exposures since job title is believed to be less subject to distorted recall. In the study of occupational COPD, most JEM-based studies have used one of two disease-specific matrices, the first developed for studies in Europe ( ALOHA ) using job titles from the International Standard Classification of Occupations, the second for studies of US populations developed by Blanc et al. These have each proved to be valuable tools but it is worth noting that both closely resemble matrices used for the study of occupational asthma a disease that largely arises from exposure to respiratory sensitising agents. Dichotomous outcomes ( COPD ): most studies in this field have used a categorical outcome ( COPD ) which has the advantage of more readily providing an index of risk that can be useful in considerations of disease 150 British Medical Bulletin 2012;104

9 Occupation and chronic obstructive pulmonary disease (COPD) prevention. It sacrifices, however, information that is retained by a continuous measure (such as FEV 1 decline ) and, in prospective studies, restricts analysis to those without prevalent COPD at baseline. Both these may result in a loss of sensitivity. Confusion with other respiratory diseases: a related source of error is the possibility of confusion between COPD and other types of airflow limitation, notably asthma in which breathlessness is also a cardinal feature. This can be alleviated by the use of post-bronchodilator measurements of spirometry or by analyses that exclude participants with a prior history of asthma. Few studies have used or had access to the former 53,54 but those that have employed the latter strategy (eg 47) have generally found little difference from analyses of the full population. Studies of exposure to VGDF Studies that have examined self-reported or expert assessment of exposure to VGDF in the workplace form the bulk of the literature on occupational COPD in general populations. This issue is important because such exposures have traditionally been considered nuisance only and regulated correspondingly. The discovery that they play, rather, a causal role in an important disease would have profound implications for the control of workplace risks. For the reasons outlined earlier, this discussion is limited to studies that have included an objective, spirometric measure of COPD. Table 2 summarizes the findings from 10 such studies that relied, in whole or in part, on self-reported exposures to VGDF. The first striking feature is the high frequency of exposure, the main determinant of the high PARs that accompany most reports. In these studies of general populations, some 50% reported exposure at some point to workplace VGDF; it seems likely that, both quantitatively and qualitatively, such exposure will represent different experiences for workers from different eras. The resultant risk estimates are based on the comparison with the remainder who reported no exposure to VGDF, after variable adjustment for some confounding exposures. In all but one instance 55 the resulting, adjusted odds ratios are,2.0 and few are statistically significant. In one study, 46 while no significant association between occupational exposure to dusts or chemicals was found for the dichotomous outcome of incident COPD, there was a relationship with a continuous measure of FEV 1 decline. The lack of any information on unadjusted estimates in all but one study 56 makes any consideration of residual confounding for example by incomplete adjustment for smoking impossible; and the limited information on whether the effects of VGDF exposure are modified by smoking seems inconsistent. British Medical Bulletin 2012;

10 152 British Medical Bulletin 2012;104 Table 2 General population studies of COPD defined by spirometry in relation to self-reported workplace exposures to VGDF Paper Country Population COPD # Cases Exposure Exposed Adjustment Adjusted OR (95% CI) Notes 46 Poland Random sample of adults aged (n ¼ 1769) 64 USA Random sample of adults aged (n ¼ 8515) 65 NZ Random sample of adults aged (n ¼ 1132) 66 Spain Random sample of adults aged (n ¼ 1735) 53 USA Hospital cases/referents (non-smokers) 67 Europe Random sample of adults aged (n ¼ 14855) 54 Spain Random sample of adults aged (n ¼ 497) 55 USA Community sample cases/referent (n ¼ 302) matched on age, sex, race; mean age USA Community sample cases/referent (n ¼ 1709) 56 UK Community/hospital sample cases/referent (n ¼ 374) D, dusts; G, gases; F, fumes. FEV 1,65% p (incidence) 125 D or chemicals 5 years D: 21% chemicals: 17% FEV 1 /FVC, D; G/F D: 31% G/F: 30% D/G/F: 61% Age, smoking (current, ex-, never), baseline FEV 1, education, symptoms Age, city, sex, smoking (total years) FEV 1 /FVC VGDF 52% Age, sex, smoking (current, ex-, never) FEV 1 /FVC, VGDF DGF 46% G/F: 35% Age, sex, area, smoking (smoker/ non-smoker) FEV 1 /FVC,0.70 and FEV 1,80%p (post-bd) GOLD criteria 537 VGDF 43% Age, sex, smoking, pack years, SEC, infection FEV 1 /FVC,0.70 and FEV 1,80%p (post-bd) FEV 1 /FVC,0.7 and FEV 1,80%p Not stated (not statistically significant) D: 1.53 ( ) G/F: 1.15 ( ) DGF: 1.57 ( ) 67 DGF 53% Only non-smokers 1.79 ( ) 36 GDF Men: D: 51% F/G: 30% women D: 37% F/G: 8% Sex, smoking (current, ex-, never) 1202 VGDF 58% Age, sex, race, smoking (current, ex-, never) FEV 1 /FVC, VGDF 48% of cases Age, sex, race, smoking (current, ex-, never) FEV 1 /FVC, VGDF Age, sex smoking (pack years) Dust exposure associated with continuous measure (FEV 1 decline) in men No modification by smoking. Smoking associated with dust exposure. 1.2 ( ) COPD with chronic bronchitis: 3.1 ( ) 1.4 ( ) No modification by smoking Subjects with asthma symptoms excluded FEV 1 /FVC,0.70: 1.08 (0.88 to 1.31) GDF: 1.2 ( ) Unadjusted risks: FEV 1 /FVC,0.70: 1.54 ( ) Stratum-specific OR: Never smokers: 1.6 ( ) Current smokers: 0.5 ( ) ex-smokers: 4.8 (0.5 42) 2.11 ( ) Population attributable fraction 31% (22 39%) 1.6 (0.99 to 2.6) No spirometry on referents 1.40 ( ) Unadjusted OR: 2.00 ( ) Adjusted for age, sex, ever/never smoking: 1.5 ( ) P. Cullinan

11 British Medical Bulletin 2012; Table 3 General population studies of COPD defined by spirometry in relation to expert assessment of workplace exposures to VGDF Paper Country Population COPD # Cases 69 Norway Random sample of adults aged (n ¼ 714) 66 Spain Random sample of adults aged (n ¼ 1735) 57 Australia Random sample of adults aged (n ¼ 1232) 44 Europe Random sample of adults aged (n ¼ 3202 men, 3279 women) 59 USA Community sample cases/referent (n ¼ 356) matched by age, sex, smoking; mean age 66 FEVI/FVC,0 7, FEVJ,80% of predicted values Exposure Exposed Adjustment Adjusted OR (95% CI) 48 Expert assessment: VGDF FEV 1 /FVC, ALOHA JEM (COPD) Mild: FEV 1 /FVC,0.70 Moderate FEV 1 / FVC,0.70 and FEV 1,80%p FEV 1 /FVC,0.70 (incidence) FEV 1 /FVC,LLN or algorithm derived 42 ALOHA JEM (COPD) 414 ALOHA JEM (COPD) 388 Expert assessment: VGDF and sensitizers 39% Age, sex, smoking (undefined) DGF 46% BD: 17% MD: 28% G/F: 35% Any:.60% BD: 39% MD: 32% G/F: 54% All: 44% Men: 52% Women 37% NS Age, sex, area, smoking (smoker/ non-smoker) Age, smoking (pack years) Age, smoking (# cigarettes), length of follow-up Age, sex, smoker, pack years Low: 1.1 ( ) High 1.5 ( ) Low: 1.5 ( ) High: 1.1 ( ) BD: 2.7 ( ) MD: 1.13 ( G/F: 1.63 ( ) Men BD: 1.34 ( ) MD: 0.66 ( ) G/F: 0.44 ( ) Women BD: 2.91 ( ) MD: 2.86 ( ) G/F: 0.00 Any VGD: 1.5 ( ) Metal dust: 1.5 ( ) Mineral dust: 1.7 ( ) Organic dust: 0.99 ( ) Irritant V/G: 1.6 ( ) Notes Includes asthma No modification by smoking subjects with asthma symptoms excluded Men BD: 1.49 ( ) MD: 0.88( ) G/F: 1.19 ( ) Women BD: 7.43 ( ) MD: 1.79 ( ) G/F: 2.73 ( ) A history of asthma was strongly associated with COPD and with most of the occupational exposures Continued Occupation and chronic obstructive pulmonary disease (COPD)

12 154 British Medical Bulletin 2012;104 Table 3 Continued Paper Country Population COPD # Cases 55 USA Community sample cases/referents (n ¼ 302) matched by age, sex, and race; mean age UK Community/hospital sample cases/referent (n ¼ 374) 47 CH Random sample of adults aged (n ¼ 4267) FEV 1 /FVC,0.70 and FEV 1,80%p Exposure Exposed Adjustment Adjusted OR (95% CI) 1202 Blanc JEM (COPD) FEV 1 /FVC, Blanc JEM (COPD) FEV 1 /FVC,0.7 or FEV 1 /FVC,LLN (incidence) 577 or 299 ALOHA JEM (COPD) JEM, job exposure matrix; BD, biological dust; MD, mineral dust; LLN, lower limit of normal. High exposure: 23% NS Any: 45% BD: 23% MD: 21% G/F: 39% Age, sex, race, smoking (current, ex-, never) Age, sex, smoking (pack years) Age, sex, smoker, pack years, BMI, early respiratory infection, education Organic solvents: 1.3 ( ) Diesel exhaust : 1.9 ( ) 2.27 ( ) Intermediate/high exposure: 0.88 ( ) (Cumulative exposure per 10 years) FEV 1 /FVC,0.7 any VGDF: 1.1 ( ) BD: 1.2 ( ) MD: 1.1 ( ) G/F: 1.1 ( ) FEV 1 /FVC,LLN any VGDF: 1.1 ( ) BD: 1.1 ( ) MD: 1.1 ( ) G/F: 1.1 ( ) Notes Unadjusted OR: 1.37 ( ) Excluded participants with history of asthma at baseline Interaction between smoking and exposure to VGDF or MD P. Cullinan

13 Occupation and chronic obstructive pulmonary disease (COPD) Table 3 summarizes information from studies of spirometry that have used an expert or JEM-based assessment of VGDF exposure; most are the same studies as those listed in Table 2. The use of the ALOHA JEM allowed for the separate examination of exposures to biological or mineral dusts, an opportunity taken in three studies. In almost all cases none of the risk estimates was statistically significant, an exception being that relating to biological dust exposures in an Australian population, a risk that appeared to be confined to women 57 which was not, however, confirmed in a subsequent, longitudinal study of a Swiss population. 47 This last, of the SAPALDIA cohort, reported significantly raised incident risk estimates for most exposure categories although the associations were dependent on the definition used for COPD, being less consistent where this was measured in relation to the LLN. The positive associations in this study were restricted to men and to those aged.40 years at baseline, the most common exposed occupational category being agricultural although very few participants were employed in this sector. The third set of studies (Table 4) are the few where exposure was defined by work in one or more of a group of occupations considered by the authors to be of high risk. In the earliest, 58 the risk estimates among non-smokers were,1.0 but in each of the others the reported adjusted odds ratios were.1.0, notably so in one of the few studies of a UK population. 43 Assigning a causal explanation to any of these associations is made more difficult by the general lack of any reported information on even the qualitative relationship between exposure and response. The use of the ALOHA JEM permits such an assessment and has been used so in four studies (Table 5) where exposure concentrations have been variously expressed by estimated intensity or duration or by a combination of the two. With the exception of the prospective study of Mehta et al. 47 in which a cumulative exposure index was used, the results are inconsistent even across just two concentration categories. In some contrast, Jaen et al. in a study of post-bronchodilator spirometry of Spanish adults reported a greater reduction in both FEV 1 and FEV 1 / FVC ratio in those with 15 years or more of self-reported exposure to workplace irritants. 54 Similarly, Weinmann et al. reported effect estimates that were greater for high than low exposures in most of several exposure categories, although these included exposures to respiratory sensitizing agents (more relevant in the study of occupational asthma) and none of the differences was statistically significant. 59 Moreover, they failed to find any relationship between risk and duration of exposure. British Medical Bulletin 2012;

14 156 British Medical Bulletin 2012;104 Table 4 General population studies of COPD defined by spirometry in relation to self-reported exposures to selected occupations Paper Country Population COPD # Cases 58 USA Sample of white adults (n ¼ 1195); analyses restricted to men 71 Italy Random sample of adults aged (n ¼ 1218) 45 Italy Men employed in mandatory, workplace Health Surveillance Program (n ¼ 2019) 43 UK Random sample of adults aged years (n ¼ 845) FEV 1,75%p or FEV 1 /FVC,80% or physician confirmed diagnosis FEV 1 %p or FEV 1 / FVC%,70 FEV 1 /FVC,0.7, and FEV 1,80%p and chronic sputum production (incidence) FEV 1 /FVC,LLN and FEV1,LLN Exposures/exposed occupations Exposed Adjustment Adjusted OR (95% CI) 268 Smelting, foundry work, welding, flour/feed, cotton/jute, chemicals, detergents, rubber, insulation, sandblasting, quarrying, construction, painting, open pit mining, underground mining, farming 171 Silica, asbestos, beryllium, talc, graphite, fibre glass, rock wool, sawdust, molybdenum dust, other dust; freon, smoke, auto exhaust, other fumes; solvents, insecticides, radiation, other 324 Not stated: self-reported exposure to dusts, fumes or vapours confirmed by research group 50 Asbestos, welding, dust/fume, paint spraying, factory, shipyard, mining 57% Age, smoking (ever, never) 34% Age, smoking (pack years) 53% Smoking (current vs ex/never) 50% Sex, weight, smoking (yes/no, pack-years), allergies Age 18þ : 1.3 ( ) Age 18 44: ever smoked: 1.43 ( ) Never smoked: 0.83 ( ) Age 45þ: Ever smoked: 1.28 ( ) Never smoked: 0.75 ( ) Men: 1.45 ( ) 2.62 ( ) 2.99 ( ) P. Cullinan

15 Occupation and chronic obstructive pulmonary disease (COPD) Table 5 Summary of exposure-response findings in general population studies of COPD defined by spirometry in relation to expert assessment of workplace exposures to VGDF Reference Biological dust Mineral dust Gas/fumes Low High Low High Low High Men Women Men Women Men and women (FEV 1 /FVC,0.7) Men and women (FEV 1 /FVC,LLN) DFEV 1 (ml):,15 years þ3; 15 years 280 DFEV 1 /FVC (%):,15 years 20.6; 15 years Effect estimates greater for high than low exposures in most exposure categories; no statistically significant differences No effect of exposure duration Figures are adjusted odds or incidence rate ratios in relation to a referent, unexposed population. LLN, lower limit of normal. Exposure to vapours, gases, dust and fumes (VGDF): biological plausibility Studies that report increased risks of work with VGDF raise an important mechanistic question; is it the case that work with sufficiently high concentrations of any respirable material is capable of inducing airflow obstruction, or do the findings of such studies reflect risks in only specific, but usually unspecified, occupations that entail exposure to particular respiratory toxins? The toxicological literature here, in relation to COPD rather than pneumoconiosis or lung cancer, is sparse and, naturally, limited largely to in vitro and animal experiments. Crystalline silica, coal mine dust and the several constituents of welding fume are pro-inflammatory and their repeated inhalation could plausibly lead to airway epithelial damage 60,61 while cadmium fume is used directly to induce emphysema in animal models of the disease. Airway disease in agricultural and textile workers is often attributed to the inflammatory effects of inhaled endotoxin suggesting perhaps a more generic effect that might be applicable more widely to those who work with biological dusts. Finally, brief but very intense exposures to a wide variety of respiratory irritants that include gases and fumes are known to cause airflow obstruction that is relatively resistant to treatment; 62 a similar effect may follow longer exposures at lower intensities, such as is claimed for those who use domestic biomass fuels. 63 British Medical Bulletin 2012;

16 P. Cullinan Conclusions There seems little doubt that certain occupations enhance the risk of COPD and may do so independently of or in concert with cigarette smoking; the evidence is most coherent for coal mining and work that entails exposure to silica, cotton dust or cadmium fume. For other occupations the evidence is less conclusive but warrants further study, particularly perhaps for those that entail exposure to welding fume, agricultural dusts and diesel fume where globally there are very large numbers of exposed workers. These studies will need to encompass careful measurement of (fixed) airflow limitation within exposed populations, preferably in a longitudinal fashion, with detailed attention to potentially confounding exposures including but not limited to smoking. These undertakings will be neither easy nor cheap; and it is probable that most will be appropriately undertaken in low-income parts of the world where exposures are likely to remain common and require greater regulation. The evidence for more generic workplace exposures to nuisance dusts and the like is less convincing and it remains unclear whether the reported associations, which are weak albeit widely distributed, are truly causal or reflective of one or more confounding explanations. To be illuminating, further population-based study in this area would benefit from more refined measures of relevant exposures, detailed examination of exposure response relationship and, even, an approach to COPD that extends beyond arbitrary categorizations of abnormality. References 1 Fletcher C, Peto R. The natural history of chronic airflow obstruction. Br Med J 1977;1: Lopez AD, Mathers CD, Ezzati M et al. Global burden of disease and risk factors. Washington: The World Bank, Hurd SS, Lenfant C. COPD: good lung health is the key. Lancet 2005;366: Buist AS, McBurnie MA, Vollmer WM et al. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007;370: Thackrah CT. The effects of arts, trades and professions and of civic states and habits of living on health and longevity with suggestions for removal of many of the agents which produce disease and shorten the duration of life. London: Longman, Rees, Orme, Brown, and Green, Hutchinson J. On the capacity of the lungs, and on the respiratory functions, with a view of establishing a precise and easy method of detecting disease by the spirometer. Med Chir Trans 1846;29: Ruysch F. Observationum anatomico-chirurgicarum centuria. Amsterdam: H. & V. Th. Boom, Warren CP. The nature and causes of chronic obstructive pulmonary disease: a historical perspective. The Christie Lecture 2007, Chicago, USA. Can Respir J 2009;16: Petty TL. The history of COPD. Int J Chron Obstruct Pulmon Dis 2006;1: British Medical Bulletin 2012;104

17 Occupation and chronic obstructive pulmonary disease (COPD) 10 Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease (revised 2011), Rennard SI, Vestbo J. COPD: the dangerous underestimate of 15%. Lancet 2006;367: Hnizdo E, Glindmeyer HW, Petsonk EL et al. Case definitions for chronic obstructive pulmonary disease. COPD 2006;3: Culver BH. How should the lower limit of the normal range be defined? Respir Care 2012;57: Balmes J, Becklake M, Blanc P et al. American Thoracic Society Statement: occupational contribution to the burden of airway disease. Am J Respir Crit Care Med 2003;167: Blanc PD, Toren K. Occupation in chronic obstructive pulmonary disease and chronic bronchitis: an update. Int J Tuberc Lung Dis 2007;11: Boschetto P, Quintavalle S, Miotto D et al. Chronic obstructive pulmonary disease (COPD) and occupational exposures. J Occup Med Toxicol 2006;1: Rushton L. Occupational causes of chronic obstructive pulmonary disease. Rev Environ Health 2007;22: Aasen TB, Blanc PD, Brisman J et al. Occupational COPD: correlations between chronic obstructive pulmonary disease and various types of physical and chemical exposures at work. A scientific reference document on behalf of The Danish Working Environment Research Fund, ashx. 19 Marine WM, Gurr D, Jacobsen M. Clinically important respiratory effects of dust exposure and smoking in British coal miners. Am Rev Respir Dis 1988;137: Seixas NS, Robins TG, Attfield MD et al. Longitudinal and cross sectional analyses of exposure to coal mine dust and pulmonary function in new miners. Br J Ind Med 1993;50: Szram J, Schofield S, Cosgrove M et al. Welding, lung function decline and respiratory symptoms: a systematic review of cohort studies. Oxford, net/reports/template/onetextabstract.xml?xsl=template/onetextabstract.xsl&conferenceid= 2501&abstractID= Dosman JA, Cotton DJ, Graham BL et al. Chronic bronchitis and decreased forced expiratory flow rates in lifetime nonsmoking grain workers. Am Rev Respir Dis 1980;121: Pahwa P, Senthilselvan A, McDuffie HH et al. Longitudinal estimates of pulmonary function decline in grain workers. Am J Respir Crit Care Med 1994;150: Eduard W, Pearce N, Douwes J. Chronic bronchitis, COPD, and lung function in farmers: the role of biological agents. Chest 2009;136: Omland O. Exposure and respiratory health in farming in temperate zones a review of the literature. Ann Agric Environ Med 2002;9: Schenker M. Exposures and health effects from inorganic agricultural dusts. Environ Health Perspect 2000;108(Suppl. 4): Szczyrek M, Krawczyk P, Milanowski J et al. Chronic obstructive pulmonary disease in farmers and agricultural workers an overview. Ann Agric Environ Med 2011;18: Hnizdo E, Vallyathan V. Chronic obstructive pulmonary disease due to occupational exposure to silica dust: a review of epidemiological and pathological evidence. Occup Environ Med 2003;60: Khan AJ, Nanchal R. Cotton dust lung diseases. Curr Opin Pulm Med 2007;13: Shi J, Mehta AJ, Hang JQ et al. Chronic lung function decline in cotton textile workers: roles of historical and recent exposures to endotoxin. Environ Health Perspect 2010;118: Wang XR, Zhang HX, Sun BX et al. A 20-year follow-up study on chronic respiratory effects of exposure to cotton dust. Eur Respir J 2005;26: Davison AG, Fayers PM, Newman Taylor AJ et al. Cadmium fume inhalation and emphysema. Lancet 1988;1: Diem JE, Jones RN, Hendrick DJ et al. Five-year longitudinal study of workers employed in a new toluene diisocyanate manufacturing plant. Am Rev Respir Dis 1982;126: Ott MG, Klees JE, Poche SL. Respiratory health surveillance in a toluene di-isocyanate production unit, : clinical observations and lung function analyses. Occup Environ Med 2000;57: British Medical Bulletin 2012;

18 P. Cullinan 35 Wu J, Griffiths D, Kreis IA et al. Lung function changes in coke oven workers during 12 years of follow up. Occup Environ Med 2004;61: Antonescu-Turcu AL, Schapira RM. Parenchymal and airway diseases caused by asbestos 1. Curr Opin Pulm Med 2010;16: Scotti PG, Arossa W, Bugiani M et al. Chronic bronchitis in the iron and steel industry: prevalence study. Med Lav 1989;80: Hart JE, Eisen EA, Laden F. Occupational diesel exhaust exposure as a risk factor for chronic obstructive pulmonary disease. Curr Opin Pulm Med 2012;18: Sterling TD, Weinkam JJ. Smoking characteristics by type of employment. J Occup Med 1976;18: Stellman SD, Boffetta P, Garfinkel L. Smoking habits of 800,000 American men and women in relation to their occupations. Am J Ind Med 1988;13: Axelson O. Confounding from smoking in occupational epidemiology. Br J Ind Med 1989;46: Ramsay SE, Whincup PH, Lennon LT et al. Longitudinal associations of socioeconomic position in childhood and adulthood with decline in lung function over 20 years: results from a population-based cohort of British men. Thorax 2011;66: Melville AM, Pless-Mulloli T, Afolabi OA et al. COPD prevalence and its association with occupational exposures in a general population. Eur Respir J 2010;36: Sunyer J, Zock JP, Kromhout H et al. Lung function decline, chronic bronchitis, and occupational exposures in young adults. Am J Respir Crit Care Med 2005;172: Boggia B, Farinaro E, Grieco L et al. Burden of smoking and occupational exposure on etiology of chronic obstructive pulmonary disease in workers of Southern Italy. J Occup Environ Med 2008;50: Krzyzanowski M, Jedrychowski W, Wysocki M. Factors associated with the change in ventilatory function and the development of chronic obstructive pulmonary disease in a 13-year follow-up of the Cracow study. Risk of chronic obstructive pulmonary disease. Am Rev Respir Dis 1986;134: Mehta AJ, Miedinger D, Keidel D et al. 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. Am J Respir Crit Care Med 2012;185: Jones JR, Hodgson JT, Osman J. Self-reported working conditions in 1995: results from a household survey. London: HSE Publications, Quinlan PJ, Earnest G, Eisner MD et al. Performance of self-reported occupational exposure compared to a job-exposure matrix approach in asthma and chronic rhinitis. Occup Environ Med 2009;66: Blanc PD, Eisner MD, Balmes JR et al. 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: Bakke PS, Hanoa R, Gulsvik A. Relation of occupational exposure to respiratory symptoms and asthma in a general population sample: self-reported versus interview-based exposure data. Am J Epidemiol 2001;154: de Vocht F, Zock JP, Kromhout H et al. Comparison of self-reported occupational exposure with a job exposure matrix in an international community-based study on asthma. Am J Ind Med 2005;47: 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: Jaen A, Zock JP, Kogevinas M et al. Occupation, smoking, and chronic obstructive respiratory disorders: a cross sectional study in an industrial area of Catalonia, Spain. Environ Health 2006;5:2. 55 Blanc PD, Iribarren C, Trupin L et al. Occupational exposures and the risk of COPD: dusty trades revisited. Thorax 2009;64: Darby AC, Waterhouse JC, Stevens V et al. Chronic obstructive pulmonary disease among residents of an historically industrialised area. Thorax 2012, in press. 160 British Medical Bulletin 2012;104

19 Occupation and chronic obstructive pulmonary disease (COPD) 57 Matheson MC, Benke G, Raven J et al. Biological dust exposure in the workplace is a risk factor for chronic obstructive pulmonary disease. Thorax 2005;60: Lebowitz MD. Occupational exposures in relation to symptomatology and lung function in a community population. Environ Res 1977;14: Weinmann S, Vollmer WM, Breen V et al. COPD and occupational exposures: a case control study. J Occup Environ Med 2008;50: Rom WN. Relationship of inflammatory cell cytokines to disease severity in individuals with occupational inorganic dust exposure. Am J Ind Med 1991;19: Castranova V, Vallyathan V. Silicosis and coal workers pneumoconiosis. Environ Health Perspect 2000;108(Suppl. 4): Brooks SM, Weiss MA, Bernstein IL. Reactive airways dysfunction syndrome (RADS). persistent asthma syndrome after high level irritant exposures. Chest 1985;88: Kurmi OP, Semple S, Simkhada P et al. COPD and chronic bronchitis risk of indoor air pollution from solid fuel: a systematic review and meta-analysis. Thorax 2010;65: Korn RJ, Dockery DW, Speizer FE et al. Occupational exposures and chronic respiratory symptoms. A population-based study. Am Rev Respir Dis 1987;136: Fishwick D, Bradshaw LM, D Souza W et al. Chronic bronchitis, shortness of breath, and airway obstruction by occupation in New Zealand. Am J Respir Crit Care Med 1997;156: Sunyer J, Kogevinas M, Kromhout H et al. Pulmonary ventilatory defects and occupational exposures in a population-based study in Spain. Spanish Group of the European Community Respiratory Health Survey. Am J Respir Crit Care Med 1998;157: de Marco R, Accordini S, Cerveri I et al. An international survey of chronic obstructive pulmonary disease in young adults according to GOLD stages. Thorax 2004;59: Blanc PD, Eisner MD, Earnest G et al. Further exploration of the links between occupational exposure and chronic obstructive pulmonary disease. J Occup Environ Med 2009;51: Bakke PS, Baste V, Hanoa R et al. Prevalence of obstructive lung disease in a general population: relation to occupational title and exposure to some airborne agents. Thorax 1991;46: Fishwick D, Barber CM, Bradshaw LM et al. Standards of care for occupational asthma: an update. Thorax 2012;67: Viegi G, Prediletto R, Paoletti P et al. Respiratory effects of occupational exposure in a general population sample in north Italy. Am Rev Respir Dis 1991;143: British Medical Bulletin 2012;