• mattress;
  • mite allergens;
  • pig farmers;
  • RAST


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Background: The aim of the present study was to determine the distribution of mite allergens in pig-farming environments in comparison to urban homes and the relationship between exposure to mite allergens and sensitization to the respective allergens in 100 pig farmers with work-related respiratory symptoms.

Methods: The concentration of storage mite (Lep d 2) and house-dust-mite (Der p 1, and Der f 1, Der 2) allergens in dust collected from five different sampling sites (floor, wall, grain mill, transit floor, and farmers' mattresses) was studied in relationship to the respective sensitization rates. Allergen concentrations in the mattresses were compared to those determined in mattresses from 22 urban dwellers.

Results: Median concentrations of Der p 1 and Der 2 in the mattresses of the farmers were significantly higher than in the urban dwellers' samples (53.4 μg/g dust vs 1.05 μg/g dust, P=0.001; 19.6 μg/g dust vs 2.2 μg/g dust, P<0.0001, respectively). Allergen concentrations in the transit areas were strongly related to bedroom exposure. In a multiple logistic regression model, a weak but significant relationship between Der p 1 exposure and sensitization to Der p 1 was found. Despite these findings, the prevalence of sensitization to mite allergens in the farmers (18%) was comparable to the prevalence in the general population.

Conclusions: Allergen exposure at the work place is strongly related to the concentration of allergens in farmers' beds. Exposure to domestic mite allergens should be taken into account when assessing occupational exposure to allergens and the respiratory health of farmers.

Farmers who work in barns or with animals frequently develop obstructive airway diseases ( 1, 2). In randomly selected farming populations, the prevalence of sensitization to storage mites is from about 5% ( 3) to 6% ( 4), with considerably higher rates in asthmatic farmers ( 3). Studies involving farmers from Sweden ( 5), Finland ( 6), and Denmark ( 3) have found Lepidoglyphus destructor to be the most important species inducing storage mite allergy, and comparable results have been reported in the UK ( 7, 8). High numbers of storage mites are present in crops ( 3, 9), but storage mites have also been found in the farmers' mattresses, with a median number of 60 mites (Acarus siro, Tyrophagus putrescentiae, L. destructor, or any storage mite) per gram dust ( 3).

Besides being rather time-consuming, counting mites as an index of allergen exposure may not optimally reflect the biologically relevant allergen concentrations at the workplace. The study of Campbell et al. ( 10) was the first to investigate the prevalence of a number of potential allergens in barns. The major findings were that L. destructor allergens were abundantly present, and that in some barns allergens reactive with antisera to Dermatophagoides sp. could be demonstrated. No such data, however, are available regarding the distribution of these allergens in animal confinement houses where pigs or poultry are kept.

The main L. destructor allergen, Lep d 2, has been purified and characterized ( 11, 12), and it has become possible to measure Lep d 2, as well as group 1 and group 2 allergens of Dermatophagoides ( 13–17), by ELISA.

This study aimed

  • to assess the concentrations of L. destructor (Lep d 2) and Dermatophagoides (Der p 1, Der f 1, and Der 2) allergens in farming environments

  • to study the relationship between allergen concentration and sensitization to the respective allergens in pig farmers with work-related respiratory symptoms.

Additionally, we determined the antigen concentrations in the mattresses of the farmers and of 22 urban dwellers.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Selection of farms

All pig farmers claiming compensation for suspected occupational airway disease at the local farm organization in Lower Saxony (northern Germany) in the years 1994–8 participated in a large-scale study of agricultural airway disease ( 2). The respective 100 pig farmers complained of work-related respiratory symptoms (69% shortness of breath, 76% cough without cold, and 44% wheezing), but no selection was made as to the presence or absence of atopy. The mean age of the farmers (86 men, 14 women) was 47±11 years (age±SD). The farms were quite representative of commercial pig farms, with an average of 423±343 pigs per farm and 118±90 pigs in an individual pig house.

Dust-sampling sites

Dusts were collected from five different sites on each farm by the same technician. In the swine confinement houses, samples of approximately 20 g were taken from the floor, wall, and grain mill with a metal spoon. In the transit area, an enclosed walkway or room usually located between the confinement house and the farmer's kitchen, floor dust was collected with a 700-W vacuum cleaner on a paper filter (pore size 6 μm, ALK-Laboratories, Hørsholm, Denmark). Likewise, bedroom dust was collected by vacuuming a 1 m2 area of the mattress for 2 min by standard methods ( 18). These sampling sites were chosen in order to get representative samples of the allergen exposure on the farm. The dust was stored at −20°C in sealed plastic bags until extraction and allergen analysis.

Additionally, 22 dust samples from the mattresses of randomly selected urban dwellers were taken by the same method.

Dust extraction and analysis of allergen content

The dust samples were sieved through a 1-mm mesh sieve, and dust was extracted for 2 h at room temperature in 0.1 M phosphate buffer containing 0.01% thimerosal and 0.05% bovine serum albumin. The extracts were centrifuged (4000 rpm, 15 min), passed through a 0.45-μm pore size filter (Acrodisc GHP, Gelman Sciences, Rossdorf, Germany), and stored in aliquots at −20°C. Extracts were mailed on dry ice to the Research Department of Alergia e Immunologı´a Abelló, Madrid, Spain, and the measurements were performed on dust extracts by monoclonal antibody ELISA for Lep d 2, Der p 1, Der f 1, and Dermatophagoides group 2 allergens (Der 2) ( 13, 19). Allergen concentrations were expressed as micrograms per gram of sieved dust. Detection limits were 0.002 μg/ml for Lep d 2, and 0.005 μg/ml for Der p 1, Der f 1, and Der 2, which was equivalent to 0.02 μg/g dust for Lep d 2 and 0.05 μg/g dust for Der p 1, Der f 1, and Der 2 when at least 1 g of dust was available.


Specific IgE antibody measurements on sera from the farmers were made to Dermatophagoides pteronyssinus (d1), D. farinae (d2), L. destructor (d71), swine epithelium (e83), cow dander (e4), horse dander (e3), dog epithelium (e2), cat dander (e1), chicken feathers (e85), Aspergillus fumigatus (m3), Penicillium notatum (m1), and Cladosporium herbarum (m2) with RAST (Pharmacia and Upjohn AB, Uppsala, Sweden). The Phadebas RAST reference system (Pharmacia) was included in all assays, allowing expression of results in Phadebas RAST units (PRU). The assays were performed in duplicate according to the manufacturer's recommendations. A positive result was defined as ≥0.35 PRU/ml.

Statistical methods

Computations were completed with the aid of a statistical package for personal computers (Statistica© 5.1, StatSoft, Tulsa, OK, USA). Because the data were not normally distributed, differences between groups were analyzed by the Mann–Whitney U-test. Because of the wide range of the mite concentrations, the square roots of the crude data were used. Additionally, logistic regression was used to test dose-response relationships between the square root of the allergen concentrations and sensitization to the respective allergens. In order to estimate the quality of the models, chi-square values (−2 * log[L0]–log [L1]) for three (L. destructor, D. farina), and four (D. pteronyssinus) degrees of freedom, as well as the corresponding P values, were calculated.


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Distribution of mite allergens

In the swine confinement buildings and transit areas, Lep d 2 was ubiquitously distributed with the highest concentrations in the grain mills ( Table 1). The concentrations of house-dust mites (Der p 1, Der f 1, and Der 2) were low at all sampling sites in the confinement houses. Only in the transit area was Der p 1 detected in most (80%) of the dust samples, and more than 50% of these samples contained Der 2 mites.

Table 1.  Mite allergen concentration in farming and urban environments
Allergen concentration (μg/g)Confinement houseTransit areaBedroom (mattress)
median 25th, 75th percentile range Floor% pos Wall% pos Grain mill% pos Floor% pos Farmer% pos Urban resident % pos
  1. % pos: percentage of samples above detection limit of assay.

Lep d 20.06 0.07 0.52 0.17 0.04 0.0
 0.02, 0.53 0.0, 0.36 0.12, 3.69 0.07, 0.40 0.0, 0.12 0.0, 0.06
Der p 10.0 0.0 0.0 0.2 53.40 1.05
 0.0, 0.0 0.0, 0.0 0.0, 0.0 0.07, 1.00 13.6, 190.0 0.0, 10.0
Der f 10.0 0.0 0.0 0.0 0.2 1.68
 0.0, 0.0 0.0, 0.0 0.0, 0.0 0.0, 0.0 0.0, 0.9 0.18, 17.0
Der 20.0 0.0 0.0 0.07 19.64 2.2
 0.0, 0.0 0.0, 0.0 0.0, 0.11 0.0, 0.33 3.9, 57.7 0.9, 8.5

In the mattresses of the farmers, the median concentration of Lep d 2 was 0.04 μg/g, compared to a median of 0 μg/g in the group of urban dwellers (P=0.15, Table 1). The median concentrations of Der p 1 and Der 2 in the mattresses of the urban population were significantly lower than that of the farmers' group. Der f 1 was found in 82% of urban beds with a median of 1.7 μg/g dust, a finding which was not significantly different from that of farmers' beds (P=0.14).

A univariate regression model was carried out to get information about the relationship between allergen concentrations in the transit areas and in the farmers' mattresses. As shown in Fig. 1, the concentrations of house-dust mites and storage mites in the transit areas were significantly correlated with the concentrations of these mites in the beds of the respective farmers (P≤0.01). This relationship was strongest for Der p 1.


Figure 1. Spearman rank correlation between mite concentrations in transit areas and mite concentrations in farmers' beds.

Download figure to PowerPoint

RAST results

RAST results were available for 99 farmers, and the distribution of IgE antibodies to house dust and storage mites is shown in Fig. 2. Of the 37 sensitized farmers, 14 had IgE antibodies to both D. pteronyssinus and D. farinae. In total, 37% of the farmers were sensitized to at least one allergen ( Table 2). Of the 15 farmers with IgE to at least one animal allergen, only two were positive to pig epithelium. One of the farmers was sensitized to pig epithelium alone, while the other also had specific IgE to C. herbarum. Forty-one percent of the sensitized farmers reacted to mold, and six of them were also sensitized to animals, while three farmers were also positive to house dust or storage mites.


Figure 2. Distribution (number, %) of positive RAST to dust mites.

Download figure to PowerPoint

Table 2.  Number and percentage of RAST-positive sera to different allergens tested
 RAST results (n=99)
AllergensClass 1Class 2Class 3Class 4Sum
Dermatophagoides pteronyssinus (d1)2 (2%)7 (7%)3 (3%)4 (4%)16 (16%)
D. farinae (d2)3 (3%)5 (5%)3 (3%)3 (3%)14 (14%)
Lepidoglyphus destructor (d71)1 (1%)2 (2%)9 (9%)012 (12%)
Swine epithelium (e83)1 (1%)01 (1%)02 (2%)
Cow dander (e4)1 (1%)1 (1%)002 (2%)
Horse dander (e3)3 (3%)0003 (3%)
Dog epithelium (e2)2 (2%)0002 (2%)
Cat dander (e1)1 (1%)1 (1%)002 (2%)
Chicken feathers (e85)6 (6%)0006 (6%)
Aspergillus fumigatus (m3)2 (2%)2 (2%)004 (4%)
Penicillium notatum (m1)7 (7%)1 (1%)008 (8%)
Cladosporium herbarum (m2)9 (9%)0009 (9%)

Relationship between allergen concentration and sensitization

The median Lep d 2 concentrations in dust samples from the confinement house were slightly higher in farmers sensitized to L. destructor than in farmers not sensitized to L. destructor, a finding which was statistically significant only for the floor samples (0.56 μg/g vs 0.05 μg/g; P<0.05). The median Der p 1 allergen concentrations in the mattresses of the farmers sensitized to D. pteronyssinus was about three times higher than in samples from RAST-negative subjects (110.3 μg/g vs 40.4 μg/g, P=0.09). There was no significant difference in allergen concentration between sensitized and nonsensitized farmers in samples taken from other locations. No differences could be observed when comparing farmers sensitized to D. farinae and farmers not sensitized (D. farinae-positive vs D. farinae-negative subjects: 0.14 vs 0.26 μg/g dust). The Der 2 content of mattresses of farmers sensitized to D. pteronyssinus or D. farinae did not differ significantly from mattresses of farmers who were not sensitized (28.6 vs 18.7 μg/g dust) ( Table 3).

Table 3.  Mite allergen concentration in farming environments for farmers not sensitized to mites in comparison to sensitized farmers
Allergen concentration (μg/g)Swine confinement houseTransit areaBedroom (mattress)
Median (range)FloorWallGrain millFloorFarmer
  1. P<0.05, Mann–Whitney U-test.

Lep d 2
RAST (L. destructor) negative 0.05 (0.0–54.1)0.06 (0.0–48.6)0.48 (0.0–25.7)0.19 (0.0–62.4)0.04 (0.0–2.3)
RAST (L. destructor) positive 0.56 (0.0–29.1)*0.27 (0.0–13.2)0.98 (0.0–10.5)0.13 (0.0–1.7)0.02 (0.0–0.6)
Der p 1
RAST (D. pteronyssinus) negative 0.00 (0.0–2.6)0.00 (0.0–0.0)0.00 (0.0–3.3)0.20 (0.0–10.0)40.4 (0.0–774.0)
RAST (D. pteronyssinus) positive 0.00 (0.0–0.1)0.00 (0.0–0.0)0.00 (0.0–0.0)0.32 (0.0–7.0)110.3 (0.8–722.0)
Der f 1
RAST (D. farinae) negative 0.00 (0.0–0.1)0.00 (0.0–0.5)0.00 (0.0–0.0)0.00 (0.0–0.6)0.26 (0.0–116.5)
RAST (D. farinae) positive 0.00 (0.0–0.0)0.00 (0.0–0.0)0.00 (0.0–0.0)0.00 (0.0–0.4)0.14 (0.0–2.5)
Der 2
RAST (D. pteronyssinus) negative 0.00 (0.0–0.4)0.00 (0.0–300.0)0.00 (0.0–0.6)0.08 (0.0–2.2)18.7 (0.1–312.0)
RAST (D. pteronyssinus) positive 0.00 (0.0–0.3)0.00 (0.0–0.1)0.00 (0.0–0.3)0.00 (0.0–8.9)28.6 (0.0–247.0)

In addition to the univariate comparison between mite concentrations at different sampling sites of sensitized and nonsensitized farmers, logistic regression models were carried out (see Table 4 for mite concentrations in farmers' beds). In these models, no dose-response relationship between storage mite concentrations or Der f 1 concentrations and sensitization to either Lep d 2 or Der f 1 could be confirmed. Both models failed to reach the level of significance. Der p 1 concentrations in farmers' beds were shown to be significant predictors of sensitization to Der p 1 in the respective farmers (OR [95% CI]: 1.30 [1.01–1.67]; P<0.05).

Table 4.  Multiple logistic regression models using RAST results as dependent variable and mite concentration in bed of farmer, age, and sex as independent variables
OR (95% CI)nChi square (DF+) PSquare root Lep d 2 Square root Der p 1 Square root Der f 1 Square root Der 2 SexAge
  1. +Degrees of freedom: three for L. destructor and D. farinae; four for D. pteronyssinus.

RAST (L. destructor) 931.710.630.570.640.97
positive (1) vs negative (0)   (0.04, 8.17)   (0.07, 5.91)(0.92, 1.03)
RAST (D. pteronyssinus) 9410.230.041.300.690.400.95
positive (1) vs negative (0)    (1.01, 1.67) (0.44, 1.06)(0.04, 3.76)(0.90, 1.00)
RAST (D. farinae) 944.280.230.570.570.97
positive (1) vs negative (0)     (0.21, 1.52) (0.06, 5.25)(0.91, 1.02)


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The four main findings of this study were as follows:

  • The median concentrations of Der p 1 were 50-fold higher in farmers' beds than in the beds of the group of urban dwellers (53.40 μg/g vs 1.05 μg/g, P=0.001), and higher than concentrations reported from many other places around the world ( 20–24).

  • Mite concentrations in the transit areas were shown to be related to dust and storage mite concentrations in farmers' beds.

  • In a multivariate model, a dose-response relationship was seen between Der p 1 concentrations in farmers' beds and sensitization to D. pteronyssinus.

  • Despite high Der p 1 concentrations in dust samples taken from farming and home environments, farmers did not show a higher prevalence of sensitization than the general population in Germany (18%).

L. destructor was found to be abundantly distributed. Since storage mites feed on molds, they can flourish on moldy grain. Our findings are consistent with this observation, since we found the highest concentrations of Lep d 2 (median 0.52 μg/g) in the grain mill. Echechipia et al. ( 19) detected Lep d 2 in 23% of house-dust samples in an urban population, with low levels (0.05–0.20 μg/g) in 10%, moderate levels (0.21–2.00 μg/g) in 6%, high levels (2.01–10.00 μg/g) in 3%, and very high levels (>10 μg/g) in 4%. While Lep d 2 in our study was detected in 41% of the urban beds, none of the samples were above 2 μg/g. Forty-one percent were found to have low concentrations, and 9% had moderate concentrations of L. destructor. In contrast, 16% of farmers' beds had concentrations of Lep d 2 of 0.2–2 μg/g, but only one farmer's mattress had high concentrations (>10 μg/g).

The results for Der p 1 agree with the investigations of Iversen et al. ( 3), who also found very high concentrations of house-dust mites in the mattresses of farmers (median count was 1480 living mites/g dust, D. pteronyssinus, D. farinae, any other house-dust mite). Exposure to more than 2 μg/g of group 1 mite allergen or 100 mites per gram dust is considered to increase the risk of sensitization and symptoms; exposure to more than 10 μg/g or 500 mites per gram dust is thought to increase the risk of acute asthma attacks ( 18, 25). Wahn et al. ( 26) proposed even lower thresholds. They suggest a low allergen exposure below concentrations of 0.4 μg Der p 1/g dust and a significant allergen exposure in the range of 0.4–2 μg Der p 1/g dust. In our study, only 13% of the mattresses were below the level of 2 μg/g. Antigen levels equal to and above 2 μg/g but below 10 μg/g were measured in 11% of the mattresses. A possible explanation could be the way farmers were recruited for the study. All subject farmers complained of work-related respiratory symptoms. Thus, one might speculate that the less “hygienic” conditions causing a higher concentration of mites are overrepresented in the sample. Furthermore, it is possible that D. pteronyssinus finds better growth conditions in a rural environment than in urban homes. Indoor temperature and humidity have the greatest influence on the growth of house-dust mites, and reports suggest that 7 g/kg is the level of absolute humidity above which excess mite growth will occur ( 27). Homes with a higher number of occupants, those located on the first floor, and private buildings tend to have a higher concentration of Der p 1 and Der 2 antigens ( 28–30). Older mattresses, use of a cover or under blanket, a higher weight of sampled dust, and a higher ratio of inhabitants per m2 were significantly associated with increased concentrations of Der f 1. On the other hand, lower Der f 1 concentrations were found when interior spring mattresses were used ( 31). However, Der f 1 was the only allergen which tended to be higher in urban mattresses than in farmers' mattresses. Which of these factors were responsible for our results will be elucidated in future studies.

We could show a significant relationship between mite concentrations in transit areas and farmers' beds. Thus, among farmers, mite concentrations at occupational locations are related to bedroom exposure. Allergen avoidance in the workplaces of symptomatic farmers alone may be insufficient.

Only two serum samples contained IgE antibodies to L. destructor alone. The majority of samples (10%) contained IgE antibodies to all three mites tested. This may be due to cross-reactivity between these two families of Acaridae mites, as shown by Luczynska et al. ( 32) and Griffin et al. ( 33), who performed RAST inhibition studies. However, van Hage-Hamsten et al. ( 5) suggested that each of the storage mites and D. pteronyssinus possess unique allergens. Ventas et al. ( 11) also found that the major allergen of Lep d 2 is not responsible for allergenic cross-reactivity between L. destructor and D. pteronyssinus. One explanation is that most individuals were exposed to both storage mites and house-dust mites and became sensitized to both. Our findings demonstrate the complexity of the immunologic responses to the different mite species. Nevertheless, there was a tendency for higher Lep d 2 concentrations in the swine confinement houses in farmers sensitized to L.destructor that was significant for the floor samples. In contrast, concentrations of Der p 1, Der f 1, and Der 2 tended to be higher mostly in the mattresses of sensitized farmers. Due to the different habitat of the mite species, it might be that the sites of sensitization to house-dust mites and storage mites are different. However, when the logistic regression model was repeated, using sensitization to L. destructor as dependent variable and mite concentrations in the confinement houses as predictor variables, no dose-response relationship was seen (data not shown). One reason might be the low number of sensitized farmers in the present study.

A relationship between exposure and sensitization to Der p 1 has been shown in many other investigations ( 25, 26, 34) and was confirmed in a multiple logistic regression model in our study. Because of the high Der p 1 and Der 2 levels in farmers' beds, it might have been expected that these farmers would have a higher sensitization rate than the urban population; however, the prevalence of sensitization to house-dust mites among the farmers was not significantly higher than that of a randomly selected urban population ( 35) in the same region of Germany (farmers vs urban dwellers aged 20–44: 24%vs 19%). In general population samples in the European Health Survey, Burney et al. ( 36) reported that 7–35% of adults aged 20–44 were sensitized to house-dust mites, whereas 24% of the farmers in this age group were sensitized to house-dust mites.

Our results are compatible with observations on sensitization rates among children growing up on farms. Children living on a farm had a significantly lower sensitization rate than children living in a rural environment but not on a farm ( 37, 38). Therefore, another possible explanation for the lack of an association between sensitization and exposure in our study might be a protective effect of dietary or environmental factors which seem to be relevant to farming environments, such as endotoxin. It may well be that “lifestyle” factors responsible for an increase in sensitization to common allergens in urban dwellers, such as spending most of the time in well-insulated buildings, may not be true of farmers and their children growing up on farms.

In conclusion, in farmers sensitized to any mite allergen, avoidance of workplace exposure may not be sufficient to control symptoms due to domestic exposure because of house-dust mites in mattresses. Further studies are needed on indoor factors associated with high Der p 1 and Der 2 antigen concentrations in farmers' mattresses, and the reason for similar sensitization rates in farmers exposed to higher allergen levels than those of the total population.


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

We thank the farmers for their participation. This study was supported by Deutscher Akademischer Austauschdienst and the British Medical Research Council (313-ARC-VII-79), the European Commission (BMH 1-CT94-1554), Hochschulamt Hamburg (VM94/010), and Hannoversche Landwirtschaftliche Berufsgenossenschaft.


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References