For nearly 300 years, work in the textile industry has been recognized as hazardous. Ramazzini (1964), in the early 18th century, described a peculiar form of asthma among those who card flax and hemp. The “foul and poisonous dust” which he observed “makes the workmen cough incessantly and by degrees brings on asthmatic troubles”. That such symptoms did in fact occur in the early textile industry was illustrated by Bouhuys and colleagues (1973) in physiological studies at Philipsburg Manor (a restoration project of life in the early Dutch colonies in North Tarrytown, New York, in the United States). While numerous authors throughout the 19th and early 20th centuries in Europe described the respiratory manifestations of work-related illness in textile mills with increasing frequency, the disease remained essentially unrecognized in the United States until preliminary studies in the middle of the 20th century under the direction of Richard Schilling (1981) indicated that, despite pronouncements to the contrary by both industry and government, characteristic byssinosis did occur (American Textile Reporter 1969; Britten, Bloomfield and Goddard 1933; DOL 1945). Many subsequent investigations have shown that textile workers around the world are affected by their work environment.

Historical Overview of Clinical Syndromes in the Textile Industry

Work in the textile industry has been associated with many symptoms involving the respiratory tract, but by far the most prevalent and the most characteristic are those of byssinosis. Many but not all vegetable fibres when processed to make textiles may cause byssinosis, as discussed in the chapter Respiratory system. The distinguishing feature of the clinical history in byssinosis is its relationship to the work week. The worker, typically after having worked a number of years in the industry, describes chest tightness beginning on Monday (or the first day of the work week) afternoons. The tightness subsides that evening and the worker is well for the remainder of the week, only to re-experience the symptoms on the following Monday. Such Monday dyspnoea may continue unchanged for years or may progress, with symptoms occurring on subsequent workdays, until chest tightness is present throughout the work week, and ultimately also while away from work on weekends and during vacation. When the symptoms become permanent, dyspnoea is described as effort dependent. At this stage a non-productive cough may be present. Monday symptoms are accompanied by across-shift decreases in lung function, which may be present on other workdays even in the absence of symptoms, but the physiological changes are not so marked (Bouhuys 1974; Schilling 1956). Baseline (Monday pre-shift) lung function deteriorates as the disease progresses. The characteristic respiratory and physiological changes seen in byssinotic workers have been standardized into a series of grades (see table 1) which currently form the basis of most clinical and epidemiological investigations. Symptoms other than chest tightness, particularly cough and bronchitis, are frequent among textile workers. These symptoms probably represent variants of the airway irritation brought on by dust inhalation.

Table 1. Grades of byssinosis

Source: Bouhuys 1974.

There is unfortunately no simple test capable of establishing the diagnosis of byssinosis. The diagnosis must be made on the basis of worker symptoms and signs as well as on the physician’s awareness of and familiarity with the clinical and industrial settings in which the disease is likely to occur. Lung function data, although not always specific, may be very helpful in establishing the diagnosis and in characterizing the degree of impairment.

In addition to classic byssinosis, textile workers are subject to several other symptom complexes; in general, these are associated with fever and not related to the initial day of the work week.

Mill fever (cotton fever, hemp fever) is associated with fever, cough, chills and rhinitis which occurs with the worker’s first contact with the mill or with return after a prolonged absence. Chest tightness does not appear to be associated with this syndrome. The frequency of these findings among workers is quite variable, from as low as 5% of the workers (Schilling 1956) to a majority of those employed (Uragoda 1977; Doig 1949; Harris et al. 1972). Characteristically, symptoms subside after a few days despite continued exposure in the mill. Endotoxin in vegetable dust is thought to be a causative agent. Mill fever has been associated with an entity now commonly described in industries using organic materials, the organic dust toxic syndrome (ODTS), which is discussed in the chapter Respiratory system.

“Weaver’s cough” is primarily an asthmatic condition characteristically associated with fever; it occurs in both new and senior workers. The symptoms (unlike mill fever) can persist for months. The syndrome has been associated with materials used to treat the yarn—for example, tamarind seed powder (Murray, Dingwall-Fordyce and Lane 1957) and locust bean gum (Vigliani, Parmeggiani and Sassi 1954).

The third non-byssinotic syndrome associated with textile processing is “mattress maker’s fever” (Neal, Schneiter and Caminita 1942). The name refers to the context in which the disease was described when it was characterized by an acute outbreak of fever and other constitutional symptoms, including gastrointestinal symptoms and retrosternal discomfort in workers who were using low-grade cotton. The outbreak was attributed to contamination of the cotton with Aerobacter cloacae.

In general, these febrile syndromes are thought to be clinically distinct from byssinosis. For example, in studies of 528 cotton workers by Schilling (1956), 38 had a history of mill fever. The prevalence of mill fever among workers with “classic” byssinosis was 10% (14/134), compared to 6% (24/394) among workers who did not have byssinosis. The differences were not statistically significant.

Chronic bronchitis, as defined by medical history, is very prevalent among textile workers, and in particular among non-smoking textile workers. This finding is not surprising since the most characteristic histological feature of chronic bronchitis is mucous gland hyperplasia (Edwards et al. 1975; Moran 1983). Chronic bronchitis symptomatology should be carefully distinguished from classic byssinosis symptoms, although byssinotic and bronchitic complaints frequently overlap and in textile workers are probably different pathophysiological manifestations of the same airway inflammation.

Pathology studies of textile workers are limited, but reports have shown a consistent pattern of disease involving the larger airways (Edwards et al. 1975; Rooke 1981a; Moran 1983) but no evidence suggestive of destruction of lung parenchyma (e.g., emphysema) (Moran 1983).

Clinical Course of Byssinosis

Acute versus chronic disease

Implicit in the grading system given in table 1 is a progression from acute “Monday symptoms” to chronic and essentially irreversible respiratory disease in workers with byssinosis. That such a progression occurs has been suggested in cross-sectional data beginning with the early study of Lancashire, United Kingdom, cotton workers, which found a shift toward higher byssinosis grades with increasing exposure (Schilling 1956). Similar findings have since been reported by others (Molyneux and Tombleson 1970). Moreover, this progression may begin relatively soon after employment (e.g., within the first few years) (Mustafa, Bos and Lakha 1979).

Cross-sectional data have also shown that other chronic respiratory symptoms and symptom complexes, such as wheeze or chronic bronchitis, are much more prevalent in older cotton textile workers than in similar control populations (Bouhuys et al. 1977; Bouhuys, Beck and Schoenberg 1979). In all cases the cotton textile workers have displayed more chronic bronchitis than the controls, even when adjusting for sex and smoking status.

Grade 3 byssinosis indicates that, in addition to symptoms, textile workers demonstrate changes in respiratory function. The progression from early byssinosis (grade 1) to late byssinosis (grade 3) is suggested by the association of lung function loss with the higher grades of byssinosis in cross-sectional studies of textile workers. Several of these cross-sectional studies have given support to the concept that across-shift changes in lung function (which correlate with the acute findings of chest tightness) are related to chronic irreversible changes.

Underlying the association between acute and chronic disease in textile workers is a dose-response relationship in acute symptoms, which was first documented by Roach and Schilling in a study reported in 1960. These authors found a strong linear relation between biological response and total dust concentrations in the workplace. Based on their findings they recommended 1 mg/m³ gross dust as a reasonably safe level of exposure. This finding was later adopted by the ACGIH and was, until the late 1970s, the value used as the threshold limit value (TLV) for cotton dust in the United States. Subsequent observations demonstrated that the fine dust fraction (<7 μm) accounted for practically all of the prevalence of byssinosis (Molyneux and Tombleson 1970; Mckerrow and Schilling 1961; McKerrow et al. 1962; Wood and Roach 1964). A 1973 study by Merchant and colleagues of respiratory symptoms and lung function in 1,260 cotton, 803 blend (cotton-synthetic) and 904 synthetic-wool workers was undertaken in 22 textile manufacturing plants in North Carolina (United States). The study confirmed the linear association between byssinosis prevalence (as well as decrements in lung function) and concentrations of lint-free dust.

The validation of changes in respiratory function suggested by cross-sectional studies has come from a number of longitudinal investigations which complement and extend the results of the earlier studies. These studies have highlighted the accelerated loss of lung function in cotton textile workers as well as the high incidence of new symptoms.

In a series of investigations involving several thousand mill workers examined in the late 1960s over a 5-year span of time, Fox and colleagues (1973a; 1973b) found an increase in byssinosis rates which correlated with years of exposure, as well as a sevenfold greater annual decrease in forced expired volume in 1 second (FEV₁) (as a per cent of predicted) when compared to controls.

A unique study of chronic lung disease in textile workers was initiated in the early 1970s by the late Arend Bouhuys (Bouhuys et al. 1977). The study was novel because it included both active and retired workers. These textile workers from Columbia, South Carolina, in the United States, worked in one of four local mills. The selection of the cohort was described in the original cross-sectional analysis. The original group of workers consisted of 692 individuals, but the analysis was restricted to 646 whites aged 45 years or older as of 1973. These individuals had worked an average of 35 years in the mills. The control group for the cross-sectional results consisted of whites aged 45 years and older from three communities studied cross-sectionally: Ansonia and Lebanon, Connecticut, and Winnsboro, South Carolina. In spite of geographic, socio-economic and other differences, the community residents did not differ in lung function from textile workers who held the least dusty jobs. Since no differences in lung function or respiratory symptoms were noted between the three communities, only Lebanon, Connecticut, which was studied in 1972 and 1978, was used as the control for the longitudinal study of textile workers studied in 1973 and in 1979 (Beck, Doyle and Schachter 1981; Beck, Doyle and Schachter 1982).

Both symptoms and lung function have been extensively reviewed. In the prospective study it was determined that the incidence rates for seven respiratory symptoms or symptom complexes (including byssinosis) were higher in textile workers than in controls, even when controlling for smoking (Beck, Maunder and Schachter 1984). When textile workers were separated into active and retired workers, it was noted that those workers retiring during the course of the study had the highest incidence rates of symptoms. These findings suggested that not only were active workers at risk for impairing respiratory symptoms but retired workers, presumably because of their irreversible lung damage, were at continuing risk.

In this cohort, loss of lung function was measured over a 6-year period. The mean decline for male and female textile workers (42 ml/yr and 30 ml/yr, respectively) was significantly greater than the decline in male and female controls (27 ml/yr and 15 ml/yr). When classified by smoking status, the cotton textile workers in general still had greater losses in FEV₁ than did the controls.

Many authors have previously raised the potential confounding issue of cigarette smoking. Because many textile workers are cigarette smokers, it has been claimed that the chronic lung disease associated with exposure to textile dust can in large part be attributed to cigarette smoking. Using the Columbia textile-worker population, this question was answered in two ways. One study by Beck, Maunder and Schachter (1984) used a two-way analysis of variance for all lung function measurements and demonstrated that the effects of cotton dust and smoking on lung function were additive—that is, the amount of lung function loss due to one factor (smoking or cotton dust exposure) was not changed by the presence or absence of the other factor. For FVC and FEV₁ the effects were similar in magnitude (average smoking history 56 pack-years, average mill exposure 35 years). In a related study, Schachter et al. (1989) demonstrated that using a parameter which described the shape of the maximum expiratory flow volume curve, angle beta, distinct patterns of lung function abnormalities could be shown for a smoking effect and for a cotton effect, similar to conclusions reached by Merchant earlier.

Mortality

Studies of cotton-dust exposure on mortality have not consistently demonstrated an effect. Review of experience in the late 19th and early 20th centuries in the United Kingdom suggested an excess of cardiovascular mortality in older textile workers (Schilling and Goodman 1951). By contrast, review of the experience in New England mill towns from late in the 19th century failed to demonstrate excess mortality (Arlidge 1892). Similar negative findings were observed by Henderson and Enterline (1973) in a study of workers who had been employed in Georgia mills from 1938 to 1951. By contrast, a study by Dubrow and Gute (1988) of male textile workers in Rhode Island who died during the period 1968 to 1978, showed a significant increase in proportionate mortality rate (PMR) for non-malignant respiratory disease. The elevations in PMR were consistent with increased dust exposure: carding, lapping and combing operatives had higher PMRs than did other workers in the textile industry. An interesting finding of this and other studies (Dubrow and Gute 1988; Merchant and Ortmeyer 1981) is the low mortality from lung cancer among these workers, a finding that has been used to argue that smoking is not a major cause of mortality in these groups.

Observations from a cohort in South Carolina suggest that chronic lung disease is indeed a major cause (or predisposing factor) for mortality, since among those workers aged 45 to 64 who died during a 6-year follow-up, lung function measured as residual FEV₁ (observed-to-predicted) showed marked impairment at the initial study (mean RFEV₁ = -0.9l) in male non-smokers who died during the 6-year follow-up (Beck et al. 1981). It may well be that the effect of mill exposure on mortality has been obscured by a selection effect (healthy worker effect). Finally, in terms of mortality, Rooke (1981b) estimated that of the average 121 deaths he observed annually among disabled workers, 39 had died as a result of byssinosis.

Increased Control, Decreased Disease

Recent surveys from the United Kingdom and the United States suggest that the prevalence as well as the pattern of lung disease seen in textile workers has been affected by the implementation of stricter air-quality standards in the mills of these countries. In 1996, Fishwick and his colleagues, for example, describe a cross-sectional study of 1,057 textile spinning operatives in 11 spinning mills in Lancashire. Ninety-seven per cent of the workforce was tested; the majority (713) worked with cotton and the remainder with synthetic fibre). Byssinosis was documented in only 3.5% of the operatives and chronic bronchitis in 5.3%. FEV₁, however, was reduced in workers exposed to high dust concentrations. These prevalences are much reduced from those reported in earlier surveys of these mills. This low prevalence of byssinosis and related bronchitis appears to follow the trend of decreasing dust levels in the United Kingdom. Both smoking habits and cotton dust exposures contributed to the lung function loss in this cohort.

In the United States, results of a 5-year prospective study of workers in 9 mills (6 cotton and 3 synthetic) was conducted between 1982 and 1987 by Glindmeyer and colleagues (1991; 1994), where 1,817 mill workers who were employed exclusively in cotton yarn manufacturing, slashing and weaving or in synthetics were studied. Overall, fewer than 2% of these workers were found to have byssinotic complaints. Nevertheless, workers in yarn manufacturing exhibited a greater annual loss of lung function than workers in slashing and weaving. The yarn workers exhibited dose-related lung function loss which was also associated with the grade of cotton used. These mills were in compliance with then current OSHA standards, and the mean airborne lint-free respirable cotton dust concentrations averaged over 8 hours were 196 mg/m³ in yarn manufacture and 455 mg/m³ in slashing and weaving. The authors (1994) related across-shift changes (the objective lung function equivalent of byssinotic symptoms) with longitudinal declines in lung function. Across-shift changes were found to be significant predictors of longitudinal changes.

While textile manufacture in the developed world appears now to be associated with less prevalent and less severe disease, this is not the case for developing countries. High prevalences of byssinosis can still be found worldwide, particularly where governmental standards are lax or non-existent. In his recent literature survey, Parikh (1992) noted byssinosis prevalences well above 20% in such countries as India, Cameroon, Ethiopia, Sudan and Egypt. In a study by Zuskin et al. (1991), 66 cotton textile workers were followed in a mill in Croatia where mean respirable dust concentrations remained at 1.0 mg/m³. Byssinosis prevalences doubled, and annual declines in lung function were nearly twice those estimated from prediction equations for healthy non-smokers.

Non-Respiratory Disorders Associated with Work in the Textile Industry

In addition to the well-characterized respiratory syndromes which can affect textile workers, there are a number of risks that have been associated with working conditions and hazardous products in this industry.

Oncongenesis has been associated with work in the textile industry. A number of early studies indicate a high incidence of colorectal cancer among workers in synthetic textile mills (Vobecky et al. 1979; Vobecky, Devroede and Caro 1984). A retrospective study of synthetic textile mills by Goldberg and Theriault (1994a) suggested an association with length of employment in the polypropylene and cellulose triacetate extrusion units. Other associations with neoplastic diseases were noted by these authors but were felt to be “not persuasive” (1994b).

Exposure to azo dyes have been associated with bladder cancer in numerous industries. Siemiatycki and colleagues (1994) found a weak association between bladder cancer and work with acrylic fibres and polyethylene. In particular, workers who dye these textiles were found to be at an increased risk. Long-term workers in this industry presented a 10-fold excess risk (marginal statistical significance) for bladder cancer. Similar findings have been reported by other authors, although negative studies are also noted (Anthony and Thomas 1970; Steenland, Burnett and Osorio 1987; Silverman et al. 1989).

Repetitive-motion trauma is a recognized hazard in the textile industry related to high-speed manufacturing equipment (Thomas 1991). A description of carpal tunnel syndrome (Forst and Hryhorczuk 1988) in a seamstress working with an electrical sewing machine illustrates the pathogenesis of such disorders. A review of hand injuries referred to the Regional Plastic Surgery Unit treating Yorkshire wool workers between 1965 and 1984 revealed that while there was a fivefold decrease in employment in this industry, the yearly incidence of hand injuries remained constant, indicating increased risk in this population (Myles and Roberts 1985).

Hepatic toxicity in textile workers has been reported by Redlich and colleagues (1988) as a result of exposure to the solvent dimethylformanide in a fabric-coating factory. This toxicity was recognized in the context of an “outbreak” of liver disease in a New Haven, Connecticut, factory that produces polyurethane-coated fabrics.

Carbon disulphide (CS₂) is an organic compound used in the preparation of synthetic textiles which has been associated with increased mortality from ischemic heart disease (Hernberg, Partanen and Nordman 1970; Sweetnam, Taylor and Elwood 1986). This may relate to its effects on blood lipids and diastolic blood pressure (Eyeland et al. 1992). Additionally, this agent has been associated with peripheral neurotoxicity, injury to sensory organs and disturbances in hormonal and reproductive function. It is generally believed that such toxicity results from long-term exposure to concentrations in excess of 10 to 20 ppm (Riihimaki et al. 1992).

Allergic responses to reactive dyes including eczema, uticaria and asthma have been reported in textile-dyeing workers (Estlander 1988; Sadhro, Duhra and Foulds 1989; Seidenari, Mauzini and Danese 1991).

Infertility in men and women has been described as a result of exposures in the textile industry (Rachootin and Olsen 1983; Buiatti et al. 1984).

Back