Portal de Periódicos da CAPES

Cotton and Chronic Lung Disease

1984; Elsevier BV; Volume: 85; Issue: 5 Linguagem: Inglês

10.1378/chest.85.5.587

1931-3543

Robert N. Jones,

Occupational exposure and asthma

Contributors to the public health literature have recently recommended a disease-specific approach to compensation for chronic respiratory illness in cotton workers.1Wegman DH Levenstein C Greaves IA Byssinosis: a role for public health in the fare of uncertainty.Am J Public Health. 1983; 73: 188-192Crossref Scopus (6) Google Scholar, 2Merchant JA Byssinosis: progress in prevention.Am J Public Health. 1983; 73: 137-139Crossref Scopus (4) Google Scholar The writers seem to favor a Brown Lung Law drawn on lines similar to the existing Black Lung Law. This is certain to provoke a sharp debate. If the Black Lung Law gives any indication, billions of dollars could be at stake. With the Black Lung Law as an example, there is every reason to doubt that the dollars would be wisely spent. But before the contending parties clash over this issue, there is an antecedent question that should be considered. Can we reasonably conclude, on the basis of present evidence, that workers exposed to cotton dust are at increased risk of developing chronic respiratory disease? Put another way, is chronic obstructive pulmonary disease (COPD) an occupational illness in these workers? At first glance, this seems like a strange question. Ill health has been imputed to textile work since the observations of Ramazzini in the seventeenth century. There is certainly no shortage, in the older medical literature, of persuasive reports linking work in cotton textile mills to disabling respiratory disease and early death. Futhermore, the symptom complex “byssinosis” is still present in textile workers, indicating that something in the mills can cause a distinctive acute respiratory syndrome. These facts seem entirely adequate to raise the question of a potential risk of chronic disabling respiratory disease. The question before us, however, is not whether past conditions caused disability and death, but whether current conditions do so. The answer would doubtless be “yes” if there had been no change in textile mills in the past 120 years (since 1863, when Jesse Leach reported debilitating disease and early death in Lancashire cotton operatives3Leach J Surat cotton, as it bodily affects operations in cotton mills.Lancet. 1883; 2: 648-649Google Scholar). But Merchant has recently pointed to “. . . the remarkable progress the cotton textile industry has made in controlling cotton dust exposures . . .”2Merchant JA Byssinosis: progress in prevention.Am J Public Health. 1983; 73: 137-139Crossref Scopus (4) Google Scholar In 1974, the authors of the NIOSH cotton dust criteria document,4U.S. Department of Health, Education and WelfareCriteria for a recommended standard: occupational exposure to cotton dust.HEW Publication No. (NIOSH) 75-118. U.S. Government Printing Office, Washington, DC1974Google Scholar reviewing the surveys of byssinosis prevalence from 1908 onward, concluded that prevalence had declined. If there has been improvement in the working environment and reduction in the prevalence of acute illness, are these reflected in the indicators and predictors of chronic disease? In 1981, two reports of mortality studies documented no excess respiratory disease deaths of cotton workers, in comparison to the general population. In one of these, Merchant and Ortmeyer5Merchant JA Ortmeyer C Mortality of employees in two cotton mills in North Carolina.Chest. 1981; 79 (suppl): 6S-11SAbstract Full Text Full Text PDF PubMed Google Scholar wrote, “Of particular interest in this study is the apparent lack of an overall increase in mortality attributed to chronic lung diseases.” In the other, Berry and Molyneux6Berry G Molyneux MKB A mortality study of workers in Lancashire cotton mills.Chest. 1981; 79 (suppl): 11S-15SAbstract Full Text Full Text PDF Google Scholar wrote “. . . there is no evidence of an excess of deaths from any particular cause.” So, too, with the study of Henderson and Enterline,7Henderson V Enterline PE An unusual mortality experience in cotton textile workers.J Occup Med. 1973; 15: 717-719PubMed Google Scholar published in 1973, and the unpublished study of Daum et al, as reported by Merchant and Ortmeyer5Merchant JA Ortmeyer C Mortality of employees in two cotton mills in North Carolina.Chest. 1981; 79 (suppl): 6S-11SAbstract Full Text Full Text PDF PubMed Google Scholar. A striking contrast to the situation in 1863, when Leach wrote, “A carder seldom lives in a card-room beyond 40 years of age . . .”3Leach J Surat cotton, as it bodily affects operations in cotton mills.Lancet. 1883; 2: 648-649Google Scholar What, then, of the evidence of chronic respiratory disease morbidity? Turning to the NIOSH criteria document, one reads that “the eventual fate of workers with byssinosis, including both active workers and those who have left the industry, is poorly defined.” (Reference 4, p 31, emphasis added.) If the health outcome is “poorly defined,” even for those workers identified as having had respiratory symptoms, it is difficult to argue that the same evidence proves an adverse outcome for the entire workforce. Admittedly, the NIOSH document writers devoted most of their efforts to reviewing the acute effects of cotton dust, on which their recommended exposure limit of 200 µg/cu m was based. Most of the available epidemiologic studies were cross-sectional, and that type of study may underestimate chronic disease risk if large numbers of affected workers have already quit the industry. But if clear evidence of increased COPD risk has not emerged from mortality or cross-sectional morbidity studies, we must seek it in longitudinal studies. The ideal prospective longitudinal study would follow a cohort of workers from first hire until death, charting the development of chronic illness and the causes and timing of demise. Because we want answers in our own lifetime, and because no one has the resources for such a study, we settle for shorter periods of follow-up, and for health measurements that predict, rather than document, morbidity and mortality. The most widely accepted predictor of the eventual development of COPD is an accelerated decline in expiratory flow rates, particularly the forced expiratory volume in one second (FEV1). The major difficulty is that expected annual declines in FEV1 are small (about 30 ml/year) in relation to the measurement error of the test (standard deviation of about ± 125 ml). This means that follow-up periods of one or two years are generally inadequate, owing to poor precision of estimates of longitudinal change. In addition, more than two longitudinal data points are desirable in order to assess for potential systematic bias or error associated with any one point. These principles have become widely accepted only in the past several years, but by these criteria some of the published longitudinal studies of textile workers are defective. The best of the longitudinal studies is that of Berry and colleagues,8Berry G McKerrow CB Molyneux MKB Rossiter CE Tombleson JBL A study of the acute and chronic changes in ventilatory capacity of workers in Lancashire cotton mills.Br J Industr Med. 1973; 30: 25-36PubMed Google Scholar who measured lung function in 595 subjects as many as six times over a three-year period. After deletion of persons with fewer than three data points, the mean annual FEV1 decline in 490 cotton mill workers was 54 ml, compared to 32 ml in 81 workers from two synthetic fiber mills, a difference significant at the level p<.05. The two synthetic fiber mills, however, had quite different mean annual declines: 52 and 14 ml. Workers in two of six medium cotton mills and five of eight coarse cotton mills had mean annual declines less severe than 52 ml. In other words, workers in half the cotton textile mills had smaller mean declines than workers in half the control mills. Other noteworthy features of the study were a lack of relationship, within cotton workers, of annual FEV1 decline to either dust concentration or a measure of dust bioactivity; and lack of relationship of annual FEV1 decline to byssinosis or to FEV1 decline over the Monday work shift. Fox and colleagues9Fox AJ Tombleson JBL Watt A Wilkie AG A survey of respiratory disease in cotton operatives.Br J Industr Med. 1973; 30: 42-47PubMed Google Scholar carried out two large cross-sectional studies in the late 1960s, in which 886 blow and card room workers were examined on both occasions, about two years apart. The authors do not furnish the grand mean FEV1 change, but from their Table 71 have estimated it, assuming equal numbers of men and women in each cell, as -46.6 ml in approximately two years, or about -23 ml/year. There was no analysis presented for the possible influence of exposure variables on longitudinal change. All of the authors’ conclusions about effects of dust exposure on lung function were based on cross-sectional data. Merchant and co-workers published a study10Merchant JA Lumsden JC Kilburn KH O’Fallon WM Copeland K Germino VH et al.Intervention studies of cotton steaming to reduce biological effects of cotton dust.Br J Industr Med. 1973; 31: 261-274Google Scholar of the effects of cotton steaming, in the course of which 199 subjects were surveyed twice, ten months apart. Annual FEV1 declines computed from these two points were large, but the follow-up period was clearly too short to allow much confidence in the estimates of lung function change. Investigators in India recently published a longitudinal study11Kamat SR Kamat GR Satpekar VY Lobo E Distinguishing byssinosis from chronic obstructive lung disease.Am Rev Respir Dis. 1981; 124: 31-40PubMed Google Scholar showing large functional declines in cotton workers. Their control subjects also showed implausibly large functional declines, and a spirometer of different make was used after the initial visit. The latest longitudinal study was reported in 1982 by Beck et al12Beck GJ Schacter EN Maunder LR Schilling RSF A prospective study of chronic lung disease in cotton textile workers.Ann Intern Med. 1982; 97: 645-651Crossref PubMed Scopus (63) Google Scholar who concluded that active and retired cotton workers showed significantly greater annual FEV1 declines than did control subjects from another community. The control subjects were obtained from a precisely defined community population, by active recruitment directed at all its eligible members. The textile workers’ longitudinal cohort was drawn from a group of which 78 percent (540/692) were recruited by uncontrolled referral and/or self-selection, from an undefined total population. Such a group probably includes a disproportionate number of persons who volunteered because they suspected (or knew) that they were unwell. Beck et al compared the health status of the two groups, but such comparisons are not valid because of the selection bias. This review, although brief, covers the literature on longitudinal lung function changes in cotton textile workers. To summarize: (1) the recent mortality studies are negative; (2) the authoritative NIOSH review of the literature (through 1974) concludes that the eventual fate of symptomatic cotton workers is “poorly defined;” (3) the published longitudinal studies do not make a persuasive case for a prediction of excess COPD prevalence. Thus, the weight of available evidence is insufficient to prove or disprove the contention that cotton textile mill work now leads to a chronic disabling respiratory disease. Arguments for or against a Federal approach to the compensation problem are premature. We should be reasonably sure of the existence and scope of this problem, before undertaking to solve it.