Exposure to organic dust and respiratory disorders : an epidemiological study in the animal feed industry

Chapter 1 summarises the background of the study. Nonmalignant respiratory disorders account for a significant part of sick leave diagnoses (19%), disability pension (3.4%) and mortality (7%). The rate of chronic obstructive pulmonary diseases has risen substantially during the last 10 to 20 years. Allthough smoking is a major external cause for the increasing mortality and morbidity, other environmental factors also play an important and probably underestimated role in the aethiology of COPD. Knowledge on occupational dust related respiratory diseases can be found in the work of famous sixteenth and seventeenth century authors, such as Agricola and Ramazzini.Allthough the role of organic dust in this field has never been questionned, scientists (and legislative authorities) have not been able to establish guidelines or limits for occupational exposure.This thesis is an epidemiological study and deals with the relationship between grain dust and endotoxin, and spirometric lung function changes and respiratory symptoms. The study was performed in the animal feed industry, in which approximately 6000 workers are exposed to grain dust. Grain dust was broadly defined as the dust present in working environments in which grain and related products are handled and processed. It is estimated that outside the agricultural sector, 20,000 to 35,000 workers are exposed to grain dust in the Netherlands.Chapter 2 is a discussion on methodological issues concerning epidemiological research on respiratory disorders, and on aethiology and terminology of respiratory disorders.It is estimated that occupational exposure may account for an excess mortality for bronchial carcinoma of up to 10%, and for at least 10% of all clinically relevant CNSLD morbidity. The concept of CNSLD (chronic non specific lung diseases) is used predominantly in the Netherlands, and comprises bronchitis, asthma and emphysema. Other concepts, more generally accepted, divide this group into asthma on the one hand, and chronic bronchitis and emphysema on the other. This latter group is also known as COPD or COLD (chronic obstructive pulmonary (lung) diseases).Spirometric lung function tests, and especially FVC and FEV _{<font size="-2">1</font>} , meet most of the requirements for epidemiological research, in terms of acceptability, simplicity, objectivity and reproducibility. Standardised questionnaires, however, lack reproducibility, and therefore require study groups of sufficient size. The general problem of validity, agreement between 'the disease' and the test value, is overcome by using both lung function tests and questionnaire score as operationally defined effect variables.The etiology of CNSLD can be explained by two different hypothesis. The socalled Dutch hypothesis states that all constituents of CNSLD are interrelated, and that the development of CNSLD is genetically predisposed. The British hypothesis, however, states that exogenous stimuli may lead to chronic imflammatory responses, and thus to bronchitis.Chapter 3 describes the history and present characteristics of the animal feed industry. The industry is relatively young; industrial production did not start until the 1930's and 40's.Generally the production process can be represented in four steps; grinding, mixing, addition of vitamins and additives, and pressing into pellets. The process is highly mechanised.Raw materials may be grains, other vegetable products, and waste products from the food industry. The relative amount of grains has been steadily decreasing since industrial production started. The proportion is now approximately 10% in the Netherlands, allthough it may be considerably more in other EC countries. The production volume in the Netherlands is about 15 million tonslyear, in approximately 350 facilities, employing over 10,000 workers. Due to upscaling and a stabilised production volume, the number of facilities and employees are expected to decrease in the coming years.An overview of the literature on grain dust and health is given in chapter 4. Exposure to grain dust may include exposure to grain; non grain plant matter; fungi, moulds and spores; mycotoxins; bacteria and their chemical components and excreta; mites and other animal matter; pesticides; and inorganic matter. A limit value for exposure to grain dust has not been established in the Netherlands.Non respiratory effects of exposure to grain dust include skin disorders. Mortality and cancer incidence studies give some indications for an increased risk of primary liver cancer, lung cancer, lymphatic and hemopoietic malignancies and bladder cancer. Three respiratory syndromes are more or less clearly defined.Extrinsic allergic alveolitis (EAA) is characterised by respiratory and systemic symptoms, appearing some hours after exposure. Grain fever is characterised by systemic effects during or up to 4-8 hours after exposure. In contrast to EAA the mechanism is not allergy mediated. Bakers asthma is an IgE mediated bronchial asthma, frequently found in bakers and millers.Epidemiological research has revealed that symptoms and spirometric lung function are frequently related to both long term and acute exposure to grain dust. The nature of respiratory effects is not unequivocal, allthough the obstructive effects appear to be most pronounced. The symptoms which point towards chronic bronchitis appear to prevail. The role of immunological mechanisms or bronchial hyperreactivity is not yet clarified.Chapter 5 is an overview of measurements of exposure to dust, endotoxin and fungi in the Dutch animal feed industry. 79 stationary and 530 personal dust samples were taken in eight animal feed production facilities. The stationary sampled showed gravimetric eight hour mean total dust concentrations ranging from 0.2 to 25 mg/m <sup><font size="-2">3</font></SUP>(Geometric Mean 1.3 mg/m <sup><font size="-2">3</font></SUP>). Concentrations of smaller particle fractions were considerably lower. Personal inspirable dust concentrations were considerably higher than stationary concentrations, and were strongly related to job titles. Inspirable dust concentrations ranged from 0.2 to 450 mg/m <sup><font size="-2">3</font></SUP>(GM = 2.4 mg/m <sup><font size="-2">3</font></SUP>). After adjusting for differences between inspirable and total dust, 25% of the measurements exceeded the Dutch MAC for nuisance dust, and 42% exceeded the ACGIH Threshold Limit Value for grain dust (4 mg/m <sup><font size="-2">3</font></SUP>). Endotoxin concentrations ranged from 0.2 to 1870 ng/m <sup><font size="-2">3</font></SUP>inspirable dust (GM = 5.1 ng/m <sup><font size="-2">3</font></SUP>) . Endotoxin was less prevalent in respirable dust than in larger dust fractions. Concentrations in dust appeared to be related to different stages in the production proces. Colony forming units (CFU) of fungi ranged from 130 to 15300 CFU/m <sup><font size="-2">3</font></SUP>(GM = 2300 CFU/m <sup><font size="-2">3</font></SUP>), and were in parallel measurements more strongly related to endotoxin concentrations than to dust concentrations.A cross-sectional analysis of 315 workers in 14 animal feed mills is given in chapter 6. Primary aims were to explore relationships between exposure to organic dust and chronic lung function changes. The study comprised monitoring dust and endotoxin exposure, spirometric lung function measurements (FVC, FEV1, mean midexpiratory flow, and flow volume parameters) and a questionnaire for respiratory symptoms. The exposure was measured in two periods, during spring and autumn. The average 8-hour personal inspirable dust exposure was 9 mg/m <sup><font size="-2">3</font></SUP>grain dust (ranging from 0.2-150 mg/m <sup><font size="-2">3</font></SUP>). and 25 ng/m <sup><font size="-2">3</font></SUP>endotoxin (0.2-470 ng/m <sup><font size="-2">3</font></SUP>) based on the 530 measurements described in chapter 5. On the basis of these measurements and the occupational history of the workers, the number of years 'worked in dust' and an estimate of the cumulative dust and endotoxin exposure was calculated. The prevalence of most chronic respiratory symptoms tended to decrease with increasing years of exposure. The healthy worker effect is probably responsible for this finding. In general, a strong negative association between most of the exposure variables and lung function was found. The endotoxin exposure was more strongly related to decreases in lung function than the dust exposure. The lung function changes occur at endotoxin levels lower than previously reported in field studies.In chapter 7, an analysis is given of symptoms during work and acute lung function changes. Reported respiratory and related symptoms during work were compared between 265 exposed animal feed workers and a control group consisting of 175 external controls and non exposed workers in the animal feed industry. Symptoms indicating respiratory and nasal irritation were significantly increased in the animal feed workers. Prevalences ranged from 9% (cough) to 21 % (sneezing). Reported cough after work was also significantly increased. In 119 workers, a total of 457 cross shift spirometric lung function changes were measured. Almost all lung function variables showed a decrease during the work shift, as could be expected since the circadian rythm is in a downward phase during the measurement period (2 PM - 10 PM). When the workers were grouped into dust and endotoxin exposure categories according to their job title, an exposure- response trend was seen for MMEF and MEF _{<font size="-2">50</font>} . The effect of endotoxin was stronger than that of dust, both in magnitude and significance. For the same lung function variables and for FEV _{<font size="-2">1</font>} and MEF _{<font size="-2">25</font>} a significant cross week change was also detected. The results of this study are in concordance with other studies that indicate acute effects on lung function and elevated prevalences of respiratory symptoms during work caused by exposure to grain dust.A general discussion of the animal feed study is given in chapter 8. Sources of bias may be related to exposure characterisation, selection, ethnicity of the study group and cohort effects. In our study, the latter two are probably of minor importance. Despite the (other) sources of bias, the study clearly indicates exposure related lung function changes and increased symptom prevalences. It is evident that these effects may be related to clinically relevant morbidity and mortality.Using data from our study and from the literature, tentative exposure limits for grain dust are proposed between 1 and 2.5 mg/m <sup><font size="-2">3</font></SUP>, and for endotoxin between 3 and 7.5 ng/m <sup><font size="-2">3</font></SUP>8 hour time weighted average.The results of this study lead to the following summarising conclusions:- Workers in the Dutch animal feed industry are very frequently exposed to dust levels in excess of internationally recommended exposure limits for grain dust and the Dutch maximum allowed concentration for nuisance dust.Endotoxin concentrations in airborne animal feed dust appear to be related to stages in the production process. Airborne endotoxin concentrations are more strongly related to fungal counts than to dust concentrations.- A comparison between animal feed workers and external controls revealed an obvious healthy worker effect.- Within the group of animal feed workers a strong negative association between (present and retrospective) exposure, and lung function was found.- The effects on lung function were more strongly related to endotoxin exposure indices than to dust exposure indices, in terms of statistical strength and size of effects at average exposure levels.- The predominant symptomatic and lung function effects indicate a clinical picture that seems related to chronic bronchitis.- Cross shift changes of sensitive lung function variables were related to dust exposure and, more strongly, to endotoxin exposure.- Primarily irritative respiratory symptoms and cross week lung function decreases were present in animal feed workers.- Tentative health based exposure limits for grain dust and endotoxin can be derived from the results of this study and existing literature. Depending on the safety factor to compensate for methodologically induced downward bias of the exposure-effect relationships, it is estimated that for 90 to 100% of all animal feed workers the risk of respiratory disorders can not be excluded. This would relate to 5400 to 6000 workers.- The value of this study, in terms of exposure-response quantification, can be significantly improved by a follow-up study in the same cohort of animal feed workers.- The present knowledge of respiratory effects of grain dust exposure can be significantly improved by further research into pathophysiological mechanisms and relevant dust components, as well as into clinically relevant respiratory morbidity and mortality.