Respiratory health of workers exposed to swine confinement buildings only or to both swine confinement buildings and dairy barns

G. Respiratory health of workers exposed to swineconfinement buildings onlyor to both swineconfinement buildings and dairy barns. Scand J Work Environ Health 1991;17:269-75. Swinebuilding workers (N = 488) and nonfarming neighborhood referents (N = 216) wereenrolledin thisstudy. Therewasa slight but significant increasein the prevalenceofchronic bronchitis (17.49 versus 11.57 0/0) and more evidence of airflow obstruction (forced expiratory volume in I slforced vital capacity 0.75 versus 0.78) among the swine workers when they were compared with the referents. The subjects who spent more than 3 hid in the swine buildings had a higher prevalence ofchronicbronchitis(21.94 versus 13.25 %) and airflowobstruction(forcedexpiratoryvolumein I slforced vital capacity 0.75 versus 0.76) than those with shorter daily contact. Swine building only workers had no precipitinsto antigens found in their environment and no clinicalevidence of extrinsicallergicalveoli tis. The number of years on the farm, dual exposure with dairy cattle, positive skin prick tests, type of piggery, and type of feeding did not add to the respiratory health impact of swine buildings.

Although numerous studies ha ve been reported o n th e subject , th e respiratory health impact of working in a swine confinement building remains controversial (1). There are large differences in the report ed incidence of respiratory ailments, such as chronic bronchitis, and in th e functional impairment observed. Som e studies ha ve reported a very high prevalence of cough and sputum production for these workers (2)(3)(4), while others found fewer symptomatic subjects (5,6). Similar differences can be found for the number o f subjects with abnormal pulmonary fun ctions (5,7). Some of these differences can be explained by dif fer ences in study populations and differences in the swine building enviro nme nts. While one study reported on worker s o f small hog raising units (7), another involved wor kers of industrial-scale swine production (8). Climatic conditions in the different co untr ies where the studies were done could also have in fluenced the results, In colder climate swine buildings tend to be less well ventilated , while microbial gro wth ma y be more impo rta nt in warmer environments. Oth er confoundi ng vari able s, such as the small size o f the study population a nd th e ab sence or ina dequacy of references, mak e th e interpretation o f published result s so metimes dif ficult (2,3,5,6,8,9).
Many farmers in Canad a and other co untr ies have mixed swine-dairy cattl e expo sure. Both swine confinement buildin gs and dair y barn s contain po tentiall y haz-  4 ardous airborn e contaminants. Previous reports show a n increased prevalenc e of respiratory ailments for both of the se environments (10,11). Data from one study suggest that dairy barns ha ve sign ificantly less negative impact on respiratory health than swine building s (11), while no differences were found regarding the preval ence of chronic bronchitis and lung function impairments between swinery workers and cattle workers in another study (12). Swine confi nement buildings are known to conta in lar ge number s o f microorganisms (13,14), so me of which indu ce precipitating antibodies and extr insic allergic alveolitis (13). The prevalence of hog producers with seru m precipit ins to these specific antigens is currently unknown . Previous studies ha ve reported variable prevalences of immunoglobulin G (lgG) to hog and feed antigens and to swine building dust (5,11,15). Interstitial changes in the lungs of guinea pigs and rabbits raised in swine confinement buildings have been described (16) , and pos sible interstitial abnormalities have been suggested for swine building workers (7). Bronchoalveolar lava ge o f asymptomatic swine show an increase in alveolar lymphocyte counts (17) when compa red with the co unts of asymptomatic dairy farmers (18).
Antigens pre sent in swine buildings include animal danders and ur ine, grain dusts, and a variety of microorgan isms. The po ssible role of immediate type hype rsensitivity to airborne antigens on respiratory symptoms and fun ctional abnormalities remains to be elucidated. Atopy has been reported to increase the risk of respiratory symptoms in farmers (19).
Thi s study reports the results obtained from 488 Quebec swine confinement building workers (164 swine

Subjects and methods
In the winter months of 1988 and 1989 (between 15 January and 15 April) 704 subjects were enrolled into this study. Of these persons, 488 were swine confinement building workers (164 swine only and 324 swine only and 324 both swine and cattle) and 216 nonfarming neighborhood referents. We found some increase both in respiratory symptoms and in the number of subjects with abnormal pulmonary funct ions, but the prevalence of these abnormalities was less than previously reported. Dual building exposure did not increase the health risk . The swine workers had not developed precipitating antibodies to microorganisms present in their work environment. And the presence of immediate skin prick test reaction was not associated with the respiratory symptoms of these workers. and dairy) and 216 were referents. The swine workers were taken from a membership list of the Quebec "Union des producteurs agricoles, section pore" (Swine worker's union). We contacted 526 subjects identified on the list provided by the union. Of these, 117 were no longer involved in swine production and 45 refused participation. A total of 364 farms were therefore retained. An additional 15 farms were subsequently ident ified by the visited farmers. The 488 swine workers were enrolled from these 379 farms. Each enrolled worker identified his or her first nonfarming and not retired neighbor, who was then asked to serve as a referent. Three hundred and twenty referents were identified and solicited. Of these, 104 refused. Each swine worker was visited at home , while the nonfarmer referents were seen at a local social club. The swine farmers were visited in the daytime, while the referents were seen in the evening. After having signed an informed consent form, the subjects were asked to (i) answer a standard questionnaire, (ii) do a forced expiratory maneuver, (iii) have venous blood drawn for precipitin analysis, and (iv) submit to a batter y of skin prick tests with 23 airborne allergens.
The questionnaire was based on the standard suggested by the American Thoracic Society (ATS) (20) with questions on the work environment added. These questions included the number of pigs, the number of hours spent daily in the confinement buildings, the number of years worked in this environment, type of piggeries, type of feedings used (eg, dry versus wet), type of farm (hog only, mixed hog-dairy cattle , etc). Questions related to a history of extrin sic allergic alveolitis (diagnosis confirmed by a qualified physician) or to symptoms suggestive of this disease were also added. All the questionnaires were filled out by the same trained nurse (GB). Ch ronic bronchitis was defined as the presence of cough and sputum production for three months per year for a minimum of two consecutive years. Forced expiratory flows were performed according to a standardized procedure (21) using a compact Vitalograph'" spirometer (Roxon, Buckingham, England). From the best forced expirator y maneuver we obtained the forced vital capacity (FVC), the forced expiratory volume in 1 s (FEV1. 0)' and the maximal midexpiratory flow rate (MMFR). The study was approved by our ethics committee. The  consent form explained in detail all of the four aspects of the study and the study goals. Ten milliliters of venous blood was drawn and kept at 4°C until the following morning when it was centrifuged . The sera were frozen at -70°C until analyzed for serum precipitins. The sera were tested for the presence of precipitating antibodies to Saccharopolyspora rectivirgula, formerly Micropolyspora jaeni (22), and to five other antigens produced from microorganisms identified in the air of four Quebec swine buildings (13). These microorganisms were Aspergillus spp, Enterobacter agglomerans, Mucor spp, Penicillium spp, and Scopulariopsis spp. The analysis for precipitins was done by a modified double diffusion technique based on the method of Ouchterlony (23). The antigens were prepared from live cultures of the aforementioned microorganisms by the method described by Schuyler et al (24). Skin prick tests to 20 common aero allergens and three hog antigens (table 1) were performed as previously described (25) all by the same nurse (OB). A test was considered valid if the histamine positive control reacted to a minimum of 2 mm and the glycerol negative control produced no measurable induration. A mean wheal diameter of~3 mm to any antigen was read as positive for that antigen.
The characteristics of our study population are presented in table 2. Significant differences between the groups included a higher prevalence of smokers and more females in the reference group. The occupations of the referents were varied, 170 were considered as having no significant environmental exposure (eg, health personnel, teachers, office workers , housewives). The other 46 subjects (including woodworkers, general mechanics, grain mill workers, etc) had some potential work-related exposure to different pollutants .
The number of subjects reported under each variable differs. The questionnaires were answered by all the subjects, reproducible forced expiratory maneuvers could not be obtained from 25 workers and seven referents, serum for precipitin analysis was not available from two swine workers, while skin prick tests were either not obtained or invalid for 58 workers and four referents . A large number of workers did not have the skin prick test because we had enrolled 55 of the subjects before we decided to include the skin tests.
A description of the exposure of the swine workers is provided in table 3. Most of these workers spent more than I hid in the swine buildings. These subjects were long-term swine confinement building workers (71 % > 10 years) and almost half had more than 500 pigs. As expected, the swine only workers spent more time in the swine buildings than those with dual work environments (P<O.OOI). The swine only workers also had more pigs, more often had both types of piggeries (farrowing and fattening), and had been in the business of raising pigs for a shorter period of time than the dual environment farmers .
The chi-square test was used to verify any association between gender or smoking status and group of subjects . The significance between the means of the pulmonary function tests was assessed by an analysis of covariance, adjustment being made for age, smoking status, gender, and height, while the significance of differences in the prevalence of chronic bronchitis was assessed by logistic regression, adjustment being made for gender and smoking status (26). All the treatment of the data was performed with an SAS (statistical analysis system) package (27).

Results
The results of the symptoms of chronic bronchitis and pulmonary functions in relation to a variety of variables are presented in tables 4, 5, 6, and 7. The swine building workers had a higher prevalence of chronic bronchitis and more evidence of airflow obstruction (lower FEV\.o/FVC and MMFR) than the referents (table 4). Time spent daily in the swine confinement buildings significantly influenced symptoms and pulmonary functions; workers exposed for more than 3 h daily had more chronic bronchitis and airflow obstruction (table 5). The number of years spent in the swine raising industry did not influence these variables (table 5).
The skin prick tests revealed a higher prevalence of immediate type allergy to hog antigens, while positive reactions to other antigens were similar in all the groups (table 8). A positive skin prick test to common aero allergen and parameters of airflow obstruction (FEV\.o/FVC and MMFR) were not associated with a significantly lower FEV \. 0 (table 6). The prevalence of chronic bronchitis was also not associated with positive skin tests. Allergy to specific swine allergens was not a contributing variable (table 6). The type of piggery (farrowing, fattening, or both) and the type of 272 feeding (dry versus wet) did not influence these results (data not shown).
Comparing for the variable double exposure (dairy cattle-swine), we found no differences between the two Table 7. Prevalence of chronic bronchitis and t he pulmonary function of the swine workers who had contact with swine confinement build ings only (swine only farme rs) and those who had exposu re to both swine buildings and dairy barns (dual en-  Table 8. Results of skin prick tests given as the percentage of the subjects with a positive skin reaction to one or more antigen. There were no differences between the groups for the common aeroallergens (P = 0.56); however, the swine workers had more positive reactions to hog antigens than the referents .
by personal interest (ie, subjects who suspected they may have pulmonary disease), it is possible that we underestimated the health impact of swine building exposure. Again st this possibility is the fact that the prevalence of respiratory symptoms and functional impairments of our referents were as previously reported for " normal" populations (28) . Since there were more smokers among the referents than the farmers, we corrected for this variable in the analysis . The paucity of precipitins for workers who raised pigs only is somewhat surprising. Only farmers with dual contact, cattle and hogs, had significant levels of precipitins . However, swine only workers are exposed to large quantities of microorganisms that are known to produce extrinsic allergic alveolitis (eg, Penicillium sp, Aspergillus sp). The antigens that we tested were obtained from the air of local piggeries and were the most predominant (13). Brouwer et al (5) found an increase in the level of Ig0 4 against pig antigens, but they did not study antibodies to other environmental ant igens. Matson et al (11) did not find IgO antibodies against antigens commonly associated with extrinsic allergic alveolitis. Bronchoalveolar lavages of asymptomatic swine workers do not show an increase in lymphocytes as has been described for dairy farmers (17,18). These findings, the absence of a history of farmer' s lung, and the lack of evidence of symptoms suggestive of extrinsic allergic alveolitis in the swine only workers suggest that such an entity is very a 95 % confidence interval 0.46-1. 25. subgroups (table 7). The only clinical difference between these two groups of farmers was th at a history of farmer 's lung was only reported for five workers exposed to both building environments.
Eight farmers with dual exposure had precipitating antibodie s to S rectivirgula. Only one of the swine only workers had antibodies to this antigen, while none were determined for the referents. One subject had precipitins to Aspergillus sp; he was in the dual environmen t group. There were eight positive reactions to E agg/omerans (three dual environment farmers , two swine only workers, and three referents). No precipitins were identified against Mucor sp, Penicillium sp, or Scopulariopsis sp .

Discussion
The present study supports the results of authors who have shown a moderate increase in respiratory abnormalities in swine confinement building workers (5). We found more abnormalities than some researchers (6,7) but less than what was reported by Donham et al (9). These differences between studies can be explained by the differences in design, differences in contact (ie, types of piggeries, number of hogs, duration of daily contact, etc), and differences in climatic conditions between countries where the studies were performed.
Some of the previous studies had not compared swine worker s with referents (5); others used nonhog farmer s as referents (2,8,9,12). In view of the potential health problems related to, for example , dair y farmin g (10), such referents may be questionable.
In this study we did not select a sample of swine confinement building workers , but solicited all workers from a predefined geographic region (three rural counties south of Quebec City). We had an excellent participat ion rate of 88 070 . The number of solicited subjects who participated was higher for the swine workers than for the referents . This difference is understandable since swine workers were the target population and had personal interests in the study . If some of the referent s who participated were also influenced unlikely related to swine building contact. This result is consistent with those of previous reports (2,17), but different from the results reported by Terho et al (29). The presence of skin allergy in a swine building worker was not associated with a greater prevalence of respiratory symptoms or lower FEV l.oIFVC. This finding is in opposition to the results of Vohlonen et al (19), who found a higher prevalence of chronic bronchitis for atopic subjects. These differences between our two studies could be explained by patient selection and data analysis. Vohlonen et al selected subjects with skin lesions, often eczema; we made no prior selection. Their population was therefore probably more atopic than ours, and it is possible that the increased risk is found only for this subset of subjects. Swine building exposure increases the prevalence of immediate type skin reaction to hog antigens, but not to other common aeroallergens, those with underlying allergies being more susceptible to the swine building environment.
Farmers who had dual exposure (hog and cattle) had a prevalence of chronic bronchitis. and respiratory function abnormalities similar to that of the subjects who worked in swine buildings only; there does not seem therefore to be an added risk for this double exposure. We did not study dairy farmers only; previous studies of workers of that environment have however described respiratory abnormalities resembling those of swine confinement building workers (12,30).