INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Confining large numbers of animals in relatively small spaces has become standard practice of swine production in most industrialized countries. The air in swine confinement buildings contains large quantities of dust, bacteria and fungi, endotoxin and toxic gases, with NH₃ and H₂S being the most abundant (1). In cold countries like Canada, adequate ventilation required to assure proper elimination of these air contaminants is economically nonfeasible, particularly during the winter months.

Farmers working in these buildings have a high prevalence of chronic bronchitis (6), airway hyperresponsiveness (16- 18), and cross shift falls in their lung functions (19). Naive individuals who spend a few hours in a swine building develop an acute response, with cough, febrile reactions, and an increased airway responsiveness (20, 21). In these subjects, the number of neutrophils markedly increases in the nasal and lung bronchoalveolar lavages and in peripheral blood during this short period of time (20, 21). The increase in blood leukocytes could be explained by the release of proinflammatory cytokines (22, 23). The exact component(s) in swine building air responsible for these reactions is (are) currently unknown. Although the air of swine buildings contains gram-negative bacteria, the level of endotoxin likely does not explain all these effects (4, 24, 25). It is possible that a combination of particles and gases found in this highly contaminated environment results in the dysfunctions that have been observed.

Humans may alter nose versus mouth breathing depending upon the circumstances. At rest most breathe only through the nose; if this were also true in swine confinement building workers, there would be no point in questioning the effect of route of breathing on response to this environment. However, because nasal resistance is relatively high (26), breathing often increases with increasing ventilation and in the presence of impediment to nasal airflow, both conditions that may be associated with working in swine confinement buildings. In our previous studies, subjects complained of nasal obstruction as a consequence of swine building exposure, and inflammation at this level was documented by nasal lavage (21). The nasal passages and sinuses filter inhaled air in an efficient manner (27, 28). The size of the particles influences their deposition in the respiratory tract. Particles  5 µm in aerodynamic diameter are classified as respirable and those  5 µm as nonrespirable. Respirable particles are more likely to penetrate the terminal bronchioles and enter the alveolar level, whereas larger nonrespirable particles are more easily filtered either in the nasal and sinus passages or in the proximal airways (28). Exclusive nose breathing would be expected to trap some of the soluble gases and most of the nonrespirable particles prior to their gaining access to the airways (27, 28). However, if the subject breathes exclusively through the mouth, it would be expected that such particles and gases would enter the airways. Small particles that reach the periphery of the lung (small airways and alveoli) should not be significantly affected by nasal or oral routes of breathing.

Depending on gas solubility and particle size, failure to filter all the inspired air by the nose and sinuses could be responsible for the airway's responses to swine building exposure. Studies in allergic asthma have evaluated the effect of antigen challenge through the nose only, or through both nose and mouth (29, 30). The results are conflicting, but they point to a protective role of nasal breathing. It does not seem reasonable to extrapolate findings observed in an allergic condition to those of an irritant toxic effect of inhaling swine building air.

We postulated that looking at the effects normal breathing versus exclusive nasal or oral breathing during exposure to swine confinement buildings would help us estimate the size of particles involved and verify if the nasal inflammation itself, with its cytokine release, has a role in the response. We predicted that during nose breathing direct effects of nonrespirable particles on the lower airways should be eliminated, or at least significantly decreased, compared with breathing only through the mouth. Therefore it is possible to speculate that cross-shift decline in FEV₁ and increase in airway responsiveness could be eliminated or at least reduced by nasal breathing. In contrast, all lung or systemic effects caused by respirable particles would not be expected to be modified by the route of breathing. On the other hand, nasal occlusion with exclusive oral breathing would be expected to prevent any nasal inflammation and cytokine release at this level. Any effect mediated by cytokine release in the nasal passages should therefore be decreased. The present study was designed to address these issues. Young normal volunteers were exposed to swine confinement buildings and allowed to breathe normally or exclusively through the nose or through the mouth. Whether or not the route of breathing can modify toxic responses to swine building contaminants could change our thinking on the methods of environmental interventions and on uses and types of respiratory protective devices that would be more appropriate for this setting.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

Nine normal young (23 to 37 yr of age) male nonsmokers who had never been exposed to swine confinement buildings prior to their participation in our projects were recruited for this study. Each subject was first screened with spirometry and skin prick tests to common aeroallergens, including hog antigens. The subject was eligible if he was normal in these aspects and had no history of lung disease and was not receiving any medication. Once enrolled the subjects were studied according to the following protocols (Table 1). In both protocols each subject was studied three times: once before and after each of two exposures to the same swine building. On exposure days the subject spent 5 h in the swine building. For the first protocol, on one of the exposure days, he was allowed to breathe normally, whereas on the other day his mouth was taped shut; the sequence was randomized. Mouth occlusion was achieved by tightly taping it with multiple strips of plastic wrapped around the neck and face up to the base of the nose. To verify occlusion the subject was asked to try and inhale through the mouth while pinching his nostrils closed. We always assured that the subject could not inhale through the mouth before each swine confinement exposure when mouth occlusion was required by the protocol. Eight subjects were enrolled in this part of the study. The second protocol had a similar design, the only exception was that on one of the swine exposure days the nose was occluded. Four subjects participated in this protocol; three of them had participated in Protocol 1 and one was a new volunteer. Both protocols were done in the same swine building in the winter months of 1995 and 1996, one year separating the two protocols. The project was approved by our institution's internal review board, and each volunteer signed a written informed consent form.

Swine Confinement Building

The building chosen was selected to represent an average type in Quebec. It housed about 400 swine between 2 and 5 mo of age; this corresponds to the fattening period before market. The 20-yr-old building was ventilated by eight temperature-controlled fans placed on each side at 6 feet above the floor level. During the winter months when the experiments were performed, the inside temperature was set at 16° C, and at this temperature the fans seldom functioned on cold winter days.

Spirometric Measurements

Forced expiratory flows and volumes were obtained using a Vitalograph spirometer (Roxon, Buckingham, UK). Complete flow curves, from which the FVC, FEV₁ and the ratio FEV₁/FVC were derived, were obtained at baseline, at the end of the swine building exposure, and before each methacholine challenge. The tests were performed according to the ATS standard procedure (31). For the methacholine challenges only FEV₁ measurements were obtained.

Methacholine Challenges

Standard procedure as described by Juniper and colleagues (32) was used in this study. Briefly, doubling doses of methacholine were delivered via a calibrated Wright nebulizer and inhaled for 2 min every 5 min until there was a drop in FEV₁ of at least 20% or a maximal dose of 256 mg/ml had been reached. Provocation concentration giving a 20% fall in FEV₁ (PC₂₀) was obtained by intrapolation on a semilogarithmic scale.

Nasal Lavage

For this sampling, after blowing and wiping his nose, the subject was instructed to occlude his posteriopharynx by positioning the tongue against the soft palate. He then tilted his head backward and 5 ml of 0.9% saline was instilled into each nostril. The saline wash was kept in the nasal cavities for approximately 20 s while the subject held his breath and maintained a closed posteriopharynx and a backward tilted head position. The subject than titled his head forward and blew the nasal wash into a clean dry flask (33).

The nasal fluid recovered was measured and centrifuged at 500 g for 10 min at 4° C. The supernatant was aliquoted and frozen at 70° C for subsequent cytokine analysis. The cell pellet was resuspended in 100 to 1,000 µl of Hanks' balanced saline solution (HBSS) for total cell count. Differential counts were performed on Diff-Quik (Baxter Health Care Corp., Miami, FL) stained glass coverslip slides (34). Cell viability was verified by trypan blue exclusion.

Bronchoalveolar Lavage

Under local anesthesia, a 5.5-mm O.D. fiberoptic bronchoscope was advanced through the mouth into the trachea and wedged into a segmental or subsegmental bronchus. A different lung region was lavaged at the two consecutive bronchoscopies done 10 d apart. The wedged lung segment was lavaged with five 60-ml aliquots of 0.9% saline; the fluid was gently aspirated after each aliquot. The recovered fluid was processed as described for the nasal lavage to obtain total cell and differential counts. BAL fluids were concentrated 15-fold by centrifugation using a 10,000 cutoff centricon filter (Amicon, Beverly, MA), and aliquots were frozen at 70° C for cytokine measurement.

Blood Samples

Ten milliliters of heparinized venous blood were withdrawn. Total and differential counts were obtained electronically with a cell counter STKS (Coulter Electronics, Hialeah, FL). Another 10 ml of nonheparinized blood were withdrawn and allowed to clot, and the serum was separated by centrifugation. Serum was kept frozen at 70° C until processed for cytokine determination.

Cytokine Measurements

Levels of cytokines in nasal lavage fluids, BAL fluids, and sera were assayed by high sensitivity immunoassay kits from R&D Systems, Inc. (Minneapolis, MN) for TNF- and IL-6, and from PerSeptive Diagnostics (Cambridge, MA) for IL-8.

Data Analysis

Depending on the data, Student's paired t test or Wilcoxon's signed test were made for all measured variables, including changes in lung function and airway response to methacholine, nasal, lung, and bronchial inflammation, and systemic response. With the small number of subjects studied only large differences could be detected.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung Function Tests

All swine building exposures, regardless of the route of breathing, resulted in a decrease in forced expiratory airflow. Mean and SEM data for the mouth occlusion protocol are presented in Table 2. Airflow obstruction had partially resolved spontaneously by the next morning as shown by the premethacholine values (Table 2); FEV₁, however, remained significantly lower before the methacholine challenge after the mouth occlusion exposure (p < 0.05). As expected, swine building exposures resulted in an increased methacholine bronchoresponsiveness. Individual and mean data for PC₂₀ before exposure, after normal breathing exposure, and after exposure with mouth or nasal occlusions are presented in Figures 1 and 2. In the mouth occlusion protocol, both types of swine building exposures (i.e., normal breathing or mouth occluded) resulted in significant decreases in PC₂₀ (p = 0.02). The sequence of exposure was more important in determining the increased responsiveness than was the route of breathing, the second exposure resulting in a much greater drop (Figure 1B). We only have PC₂₀ values for three subjects in the nasal occlusion group. These results show effects of swine building exposure similar to those in the first protocol.

View larger version (18K): [in this window] [in a new window]   Figure 1.   PC20 methacholine before and after each of two exposures to a swine confinement building (Protocol 1). Results are presented for the route of breathing (A) and the sequence of the exposures (B). Mean (lines) identified by different letters are significantly different. Note that the route of breathing had no influence, but the sequence of exposure did because the drop in PC20 was significantly greater for Visit 2.

View larger version (15K): [in this window] [in a new window]   Figure 2.   PC20 methacholine for three of the four subjects studied in Protocol 2. Two of these three dropped their PC20 after each exposure.

Nasal Wash Analysis (Cells and Cytokines)

For both protocols, a significant increase (p = 0.0001) in inflammatory cells, mainly neutrophils, was observed in nasal lavage fluids during normal breathing and after exposure to the swine confinement building (Figure 3): an increase from 2.14 ± 0.78 × 10⁴ to 30.78 ± 14.2 × 10⁴ cells/ml for the first protocol, and from 0.85 ± 0.62 × 10⁴ to 11.22 ± 9 × 10⁴ in the second protocol. Higher levels of inflammatory cytokines IL-6, IL-8, and TNF- were also measured in nasal lavage fluids (Figure 4) after exposure, the most enhanced being IL-8 (all significant p values < 0.001). No increase in cellularity (0.41 ± 0.23 × 10⁴ cells/ml) and cytokines was obtained in subjects exposed with their noses occluded (Figures 3A and 4). Subjects exposed with their mouths occluded showed similar nasal cellular inflammatory responses to the ones obtained when the subjects were breathing normally (Figure 3B).

View larger version (30K): [in this window] [in a new window]   Figure 3.   Results (mean ± SEM) for nasal wash total cells (dotted columns) and neutrophils (shaded columns) for both the mouth (A) and the nasal (B) occlusion protocols. Columns marked by different letters are significantly different.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Individual and mean data (horizontal lines) for nasal wash levels of IL-6, IL-8, and TNF- before and after swine building exposure. Increases after exposure reached significant levels during normal breathing and with mouth occlusion. As expected, nasal occlusion blunted these increases at this level.

BAL Analysis (Cells and Cytokines)

A significant increase in BAL inflammatory cells was observed after swine building exposure. Similar levels of cells (total and differential) were detected whether the subjects were breathing normally, through the nose, or through the mouth (Figure 5A and B). BAL IL-6, IL-8, and TNF- increased after exposure; the route of breathing had no effect on their levels (Figure 6).

View larger version (20K): [in this window] [in a new window]   Figure 5.   Results of bronchoalveolar lavage cell recovery at baseline and after swine building exposures for the two separate protocols: closed columns = total cells; open columns = alveolar macrophages; hatched columns = lymphocytes; shaded columns = neutrophils. Exposure resulted in large increases in total cells, neutrophils being the subset with the greatest increase. Significantly more alveolar macrophages and fewer neutrophils were seen with nasal occlusion than with normal breathing; otherwise, the route of breathing had no effects on the cellular response to exposure. All significant p < 0.0001.

View larger version (12K): [in this window] [in a new window]   Figure 6.   Levels of IL-6, IL-8, and TNF- measured in BAL fluids under all studied conditions. Swine confinement building exposure, regardless of the route of breathing, resulted in a significant increase of these cytokines (all significant, p < 0.0001).

Blood Sample Analysis (Cells and Cytokines)

Total white blood cells increased after swine dust exposure during normal breathing on both protocols (Figure 7A and B). Neither nose occlusion nor mouth occlusion had a significant effect on this increase. IL-6 serum levels were enhanced after exposure for all routes of breathing (p = 0.0001) (Figure 8). IL-8 levels in sera did not change significantly except after nasal occlusion when the levels were significantly higher than at baseline (p = 0.001) (Figure 8). TNF- in the serum was not significantly increased after all exposures except with normal breathing in Protocol 2 (p < 0.05).

View larger version (23K): [in this window] [in a new window]   Figure 7.   Total white blood cell counts before and after swine building exposures. The significant increases observed (p < 0.0001) were mostly due to an increase in neutrophils (data not shown). Route of breathing did not modify this response to exposure.

View larger version (12K): [in this window] [in a new window]   Figure 8.   Levels of serum IL-6, IL-8, and TNF- for all study conditions. IL-6 was increased after all exposures; however, in Protocol 2, IL-8 was increased only after swine building exposure with the nose occluded (p = 0.001), whereas TNF- was increased only after normal breathing.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has shown that the route of breathing does not modify most of the lung, airway, and white blood cell response to exposure in a swine building, suggesting that respirable-size particles and/or toxic gases are responsible for the observed effects. Because of the complexity of the study and because we wanted to verify if nasal filtration could offer significant protection in swine confinement exposures, we studied only a small number of subjects. The eight subjects studied by mouth occlusion were sufficient to confirm that nasal breathing would not prevent the major effect of this exposure. The mouth occlusion protocol verified the effect of exclusive nasal breathing; therefore, we needed the nasal occlusion protocol to have a total bypass of the nasal filtration. The major significant difference observed was the expected absence of response at the nasal level with nasal occlusion. The other differences were: slightly higher numbers of alveolar macrophages, fewer BAL neutrophils, higher serum IL-8, and higher serum TNF- after exposure with normal breathing in Protocol 2. Because only four subjects were studied in this second group, the significance of these differences has to be interpreted carefully. The limits of this interpretation are supported by the discordance between IL-8 and TNF- responses; especially since the increase in peripheral blood neutrophils was not altered by this route of breathing. This small number of subjects, we believe, however, was sufficient to support the main objective of doing this group, that is, to verify the possible effect of cytokines released in the nasal passages. The data support the conclusion that cytokines produced locally in the nasal passages are of little significance in the overall response to swine building exposure.

Results of this study also further confirm the acute effects of swine building exposure in normal volunteers. The cytokine profile obtained by exposure to this environment supports what was previously described, except for serum TNF- (21, 22). This difference may be related to the low levels of this cytokine found in the serum in both studies. Previous studies in different models where inflammation was induced by endotoxin have also failed to show increases in TNF- (35). TNF- is an acute-phase cytokine, and the sampling may have occurred too late to record its maximum levels.

Besides supporting our previous observation that the response to this environment was reproducible with repeated exposure 10 d later, the results of the current study show that the response to a swine confinement building is reproducible 1 yr later. All exposures for this study were done in the same swine building, it may therefore not be surprising that similar results are observed between two consecutive winters. We believe the building selected for this study represented the average for this industry in Quebec. Further studies will be needed to verify if exposure to more modern production units would result in similar inflammatory responses.

Because the two separate protocols were conducted 1 yr apart and that some of the subjects were different, no direct comparisons can be made between the two. However, the nonobstructed breathing resulted in similar responses in both protocols, suggesting that the exposures were similar in both studies. No modifications had been made in the swine building or in the operation of the swinery between the two studies. As both protocols were carried out in the winter months, minimal ventilation was in place at the time of both studies.

PC₂₀ did not drop in two subjects with mouth occlusion and in one with the open mouth exposure. This lack of response was related to the visit sequence, suggesting as previously described, that a residual or additive effect persisted after a prior exposure 1 wk previously (36). This visit-related difference seemed to be related only to airway responsiveness since such differences were not seen in the other parameters evaluated.

In conclusion, the absence of effect of mouth occlusion and the minimal effect of nasal occlusion on the lung and systemic responses to swine confinement building exposures suggest that local nasal filtration and inflammation do not contribute to the overall response and that small, respirable particles and/ or nonabsorbable gases are responsible for the observed effects of acute exposure to this environment. We have recently demonstrated that canola oil spray blunted most of this response (37). This technique is known to decrease dust levels (38), but it does not affect in a major way the levels of toxic gases. If this is the case, the results of this study and those of previous reports strongly suggest that substances in small respirable particles are responsible for the acute effect of short-term exposure to swine confinement buildings.