Cow hair allergen concentrations in dairy farms with automatic and conventional milking systems: From stable to bedroom

International Journal of Hygiene and Environmental Health 219 (2016) 79–87

Contents lists available at ScienceDirect

International Journal of Hygiene and Environmental Health journal homepage: www.elsevier.com/locate/ijheh

Cow hair allergen concentrations in dairy farms with automatic and conventional milking systems: From stable to bedroom A. Böhlandt a , R. Schierl a,∗ , J. Heizinger a , G. Dietrich-Gümperlein a , E. Zahradnik b , L. Bruckmaier c , J. Sültz d , M. Raulf b , D. Nowak a a Institute and Outpatient Clinic for Occupational, Social and Environmental Medicine, Clinical Center, Ludwig Maximilians University, Munich, Member of the German Center for Lung Research (DZL), Munich, Germany b Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Ruhr-University Bochum (IPA), Bochum, Germany c Social Insurance for Agriculture, Forestry and Horticulture Niederbayern/Oberpfalz und Schwaben, Landshut, Germany d Occupational and Internal Medicine, Pneumology, Neusäss, Germany

a r t i c l e

i n f o

Article history: Received 21 May 2015 Received in revised form 16 September 2015 Accepted 18 September 2015 Keywords: Cow hair allergen Dairy farm Automatic milking Dust Stable Dwellings

a b s t r a c t Bovine hair and dander are considered to be a notable risk factor for sensitization and allergic symptoms in occupationally exposed cattle farmers due to various IgE binding proteins. Farmers are suspected not only to be exposed during their work inside the stables but also inside their homes as allergens could be transferred via hair and clothes resulting in continued bovine allergen exposure in private areas. In recent years a new sensitive sandwich ELISA (enzyme linked immunosorbent assay) test has been developed to measure the cow hair allergen (CHA) concentration in dust. The aim of the present study was to determine the CHA concentration in airborne and settled dust samples in stables and private rooms of dairy cattle farms with automatic milking systems (AM) and conventional milking systems (CM), also with respect to questionnaire data on farming characteristics. For this purpose different sampling techniques were applied, and results and practicability of the techniques were compared. Dust sampling was performed in the stable, computer room (only AM), changing room, living room and bedroom (mattress) of 12 dairy farms with automatic milking systems (AM group) and eight dairy farms with conventional milking systems (CM group). Altogether, 90 samples were taken by ALK filter dust collectors from all locations, while 32 samples were collected by an ion charging device (ICD) and 24 samples by an electronic dust fall collector (EDC) in computer rooms (AM) and/or changing and living rooms (not stables). The dust samples were extracted and analyzed for CHA content with a sandwich ELISA. At all investigated locations, CHA concentrations were above the limit of detection (LOD) of 0.1 ng/ml dust extract. The median CHA concentrations in dust collected by ALK filters ranged from 63 to 7154 ␮g/g dust in AM farms and from 121 to 5627 ␮g/g dust in CM farms with a steep concentration gradient from stables to bedrooms. ICD sampling revealed median CHA contents of 112 ␮g/g airborne dust in the computer rooms of the AM farms and median CHA loads of 5.6 ␮g/g (AM farms) and 19.8 ␮g/g (CM farms) in the living rooms. Passive dust sampling by EDC was performed only at two locations in the AM group resulting in median CHA values of 116 ␮g/m2 (computer room) and 55.0 ␮g/m2 (changing room). Except for the stable samples the median CHA load was lower in AM farms compared to CM farms. The CHA contents of ALK filter samples were significantly correlated in most locations. Differences between the farming types were not significant. Although allergen transfer to the private area of the farmers has been found and results from several locations were correlated, differences in CHA concentrations were not significant with respect to questionnaire data such as the wearing of stable clothes in living room, free access of pets to stable and home, frequency of hair washing. All sampling techniques seem to being practicable for simple and effective CHA measurement. © 2015 Elsevier GmbH. All rights reserved.

∗ Corresponding author. E-mail address: [email protected] (R. Schierl). http://dx.doi.org/10.1016/j.ijheh.2015.09.004 1438-4639/© 2015 Elsevier GmbH. All rights reserved.

80

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

1. Introduction Dairy farming is a main branch of the agricultural sector in Germany and many other countries. Cattle farmers are occupationally exposed to a variety of bioaerosols of which some components are considerable risk factors for airway diseases. Among them, the lipocalin Bos d 2 is known to be a major respiratory allergen, but several other bovine IgE binding proteins have also been identified in cow hair and dander (Ylonen et al., 1992; Rautiainen et al., 1996; Heutelbeck et al., 2009; Zahradnik et al., 2011a). Clinical symptoms of the exposed workers can reach from asymptomatic sensitization, rhinitis up to severe asthmatic attacks with lung function impairment (Reynolds et al., 2013) leading to a high rate of initial employment disabilities in Germany, as reported by the German Social Insurances for Agriculture, Forestry and Horticulture (LBGs), and other European countries (Kogevinas et al., 1999; Radon et al., 2001; Heutelbeck et al., 2007; Elholm et al., 2010). This is of high public health relevance with respect to the large number of young adults being confronted with grave economic and personal consequences. To reduce the allergen exposure, many farmers introduced personal protective equipment (e.g.; masks, overalls, hair caps) during their work inside the stables and modified their stables (e.g.; installation of automatic milking, changing rooms). Automation of working processes can help to facilitate working steps, and robotic systems like automatic milking systems (also known as robotic milking) or feeding dispersers have already been installed by many dairy farmers. Nevertheless, despite prevention strategies, notable amounts of dust and bovine allergens could be quantified in the working and also in the living environments of cattle farmers suggesting an allergen transfer into the private rooms. This has been shown for the major allergen Bos d 2: Berger et al. (2005) and Hinze et al. (1997) described a strong relationship between the concentration of bovine allergens in dust from farmers’ homes and the sensitization of farmers. Notable concentrations of Bos d 2 in settled dust from sheds but also from the farmers’ dwellings (living rooms, mattresses) were measured by a Rocket immunoelectrophoresis using an anti-Bos d 2 antibody. Meanwhile, this Bos d 2-test has been no longer commercially available and a new and more sensitive sandwich ELISA (enzyme linked immunosorbent assay) test has been developed to measure the exposure to cow hair allergens (CHA) in airborne and settled dust (Zahradnik et al., 2011b). The aim of the study was to determine the CHA concentration in airborne and settled dust samples from dairy cattle farms with automatic milking systems compared to conventionally equipped farms and to investigate the allergen transfer from the working areas to the living areas of farmers, also with respect to questionnaire data on farming characteristics. Three different sampling methods were performed, and the results and the practicability of sampling strategies were compared.

2. Materials and methods

per telephone. Subsequently, all farms were personally visited by the field worker for providing more detailed information. A room between the stable and the living area, where farmers could change their clothes and shoes and had the possibility to wash their hands or take a shower (“changing room”), was necessary for inclusion in the study and existent in all contacted farms. Presence or absence of allergic symptoms of the farmers was no recruitment criterion and was not evaluated, but farmers with a current official announcement of an occupational disease were not contacted. All contacted farmers agreed to participate in the study and informed consent was obtained from all study participants. “Conventional” milking systems comprised/defined other systems than milking robots (e.g.; herringbone parlour, swing-over, side-by-side parlour, milking pipeline) with close animal contact, but detailed information on type and placement of these systems were not collected. The milking robots of the AM farms, however, were either installed in a corner inside or in a compartment directly adjacent to the stable and were operated from a computer room adjoining the cow stable. Farming types and study site characteristics were representative for the region of Bavaria, but not necessarily for Germany in general. Data on study sites’ characteristics and working/living conditions of the AM farms and CM farms (e.g.; number of cows, type of cattle breed, housing system and bedding) were collected via questionnaire and by observation of the field worker during the first visit, and the evaluation of these data is summarized in Table 1. Moreover, data regarding the farmers’ practices, which are likely to contribute to allergen transfer from working to private areas (e.g.; frequency of hair washing, wearing of stable clothes inside private rooms, free access of cats/dogs to stable and living area) were collected via yes/no questions.. 2.2. Sampling strategy Between November 2009 and March 2010 dust sampling was performed at five locations inside the cow stable area (next to milking robot/milking parlour and adjacent computer room) and inside the farmers’ dwellings (changing room, living room and mattress in the farmers’ bedroom). Three different sampling techniques for dust collection were used according to a standardized protocol: (1) ALK filter dust sampler, (2) Ion-charging device Ionic Breeze Quadra (ICD) and (3) electrostatic dust fall collector (EDC). While settled dust sampled by the ALK device was collected from all sampling locations of all participating farms, ICD and EDC sampling of airborne dust could not be performed at all locations due to practicability and capacity reasons: As only two ICD devices were available to run in parallel, only the computer and living room were monitored. EDC sampling was restricted to the computer and the changing rooms, as the amounts of settling dust were expected to be very high in the stables and very low in the sleeping rooms for adequate results. All samples were collected by the same field worker in order to provide comparable sampling conditions. Number of dust samples, sampling technique and sampling locations are shown in Table 2.

2.1. Participants and study sites 2.3. Sampling techniques Altogether, the farmers of 20 dairy cattle stables were contacted, of which 12 farms were equipped with automatic milking systems (“milking robot”) and 8 farms applied “conventional” milking systems (other than milking robots). The 12 farms with automatic milking systems (AM farms) were located in the north-eastern region of Bavaria (South Germany) and telephonically contacted by a fieldworker in cooperation with the regional Social Insurance for Agriculture, Forestry and Horticulture. The 8 dairy cattle farms with “conventional” milking systems (CM farms) from neighbouring Bavarian regions were directly contacted by the field worker

2.3.1. ALK filter sampling device Collection of reservoir dust from floors and other easily accessible surfaces in computer, changing and living rooms (e.g.; desktops, window sills, floors) and from mattresses was performed using a ALK sampling device consisting of a conventional vacuum cleaner (Miele S S624) fitted with ALK filters (ALK Copenhagen, Denmark), as described in detail by Waser et al. (2004) and Berger et al. (2005). The ALK filters were stored at room temperature and transferred within five days to the laboratory of the Institute and Outpatient

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

81

Table 1 Questionnaire data: study site characteristics of the farms with milking robots (AM farms) and with conventional milking systems (CM farms). AM farms (n = 12)

CM farms (n = 8)

Median

Range

Median

Range

Number of persons working in stable Time since installation of milking robot (months) Total number of cows Number of cows in the sampled stable Number of cows per day using milking robot Frequency of stable cloths washing (timespan in days) Frequency of washing bedclothes (timespan in days) Distance from stable to dwelling (m)

3.5 16 142 77 60 7 28 20

3–5 11–68 97–230 57–145 50–95 7–8 14–70 3–50

3 – 96 41 – 5 21 13.5

1–5 – 70–330 33–57 – 1–7 14–28b 1–40

Type of milking robot

50% LELY Type A3 or A4 (n = 6) 41.6% DeLaval/Alfa Laval (n = 5) 8.3% Merlin Lemmer-Fullwood (n = 1)

Type of predominant cattle breed

91.7% German Simmental (n = 10) 8.3% German Simmental & German Holsteins (n = 2)

100% German Simmental (n = 8)

Housing system

50% loose-housing barn (n = 6) 50% combined loose-housing and stanchion barn (n = 6)

25% loose-housing barn (n = 2) 25% stanchion barn (n = 2) 50% combined loose-housing and stanchion barn (n = 4)

Stable floor

8.3% continuous solid floor (n = 1) 16.7% slatted floor (n = 2) 75% combined continuous solid and slatted floor (n = 9)

12.5% continuous solid floor (n = 1) 25% slatted floor (n = 2) 62.5% combined continuous solid and slatted floor (n = 5)

Bedding/Litter

50% only straw (n = 6) 50% straw & chalk (n = 6)

12.5% no litter (n = 1) 87.5% straw (n = 7)

Separate changing room/mudrooma Wearing of stable cloths in living area Pet with stable and living room access Pet with stable and sleeping room access Daily hair washing Inspection round in the evening

Yes: n = 6 Yes: n = 8 Yes: n = 4 Yes: n = 1 Yes: n = 5 Yes: n = 8

a b c

Table 2 Number and kind of dust samples for cow hair allergen (CHA) analysis and sampling locations in farms with automatic (AM) and conventional (CM) milking systems. AM farms

CM farms

Bovine allergen in settled dust (g/g): ALK dust collector 12 Milking parlour/stable 12 Computer room Changing room 12 12 Living room 12 Mattress

8 – 8 8 6a

Bovine allergen in airborne dust (g/g): Quadra Ionic Breeze (ICD) 12 Computer room – Changing room Living room 12

– – 8

Bovine allergen in passively collected dust (g/m2 ): EDCb Computer room 12 12 Changing room

– –

a

Yes: n = 4 Yes: n = 8 Yes: n = 3 Yes: n = 2 Yes: n = 2 Yes: n = 3

No: n = 4 No: n = 0 No: n = 5 No: n = 6 No: n = 5c No: n = 3c

Separate from the dwelling: other building, floor/story, garage, cellar. No data provided in two cases of the CM group. No data provided in one case of the CM group.

Clinic for Occupational, Social and Environmental Medicine (IPASUM), University of Munich for CHA analysis. Inside the cow stables, the dust was manually collected (e.g.; window sills) at a height of 0.5–1.5 m with a spatula into a plastic bag. In the laboratory, the dust was sieved, transferred into a 1.5 ml Eppendorf safe-lock tube (Eppendorf Biopure, Hamburg, Germany) and stored at room temperature until analysis.

b

No: n = 6 No: n = 4 No: n = 8 No: n = 11 No: n = 7 No: n = 4

Two farmers did not agree with sampling. Electrostatic Dust Fall Collector.

2.3.2. Ionic Breeze Quadra Airborne dust collection was performed using the Ionic Breeze Quadra, a commercially available ion-charging device (ICD) which cleans large volumes of air by particle deposition. The ICD Quadra Ionic Breeze system is usually applied as air filter system and is not a sampling device, but it is well suitable for long-term measurements in inhabited rooms due to its silent run and its construction with removable plates for particle collection. This technique has already been successfully applied for quantitative measurement of airborne allergens for several years (Custis et al., 2003). In brief, the device is equipped with negatively charged steel plates. Airborne particles are positively charged and stick to the negatively charged plates. The filter device was positioned at the floor inside the computer room and living room passing air through at a height from 17–57 cm. The flow rate has been estimated to 1.7 (±0.15) m3 /min, which provides a reasonable sensitivity with a 24 h run (Custis et al., 2003). However, as the flow rate and thus the particle deposition efficiency could not be determined exactly, the air concentration of CHA was not calculated in the present study. After the 24 h run (mean 22.5 h, range 19.0–26.5 h), the device was transferred to IPASUM Munich, and particle load was sampled according to Custis et al. (2003), but wiping with PBST (phosphate buffer saline with 0.05% Tween 20) instead of pure water. Dust recovery in the washing solution was measured by weighing and CHA content in dust was determined in ␮g/g dust.

82

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

2.3.3. Electrostatic dust fall collectors Electrostatic dust fall collectors (EDC) enable passive sampling of airborne dust settling on dust-binding cloths. EDC sampling was conducted as previously described (Zahradnik et al., 2011a), using EDCs developed by Noss et al. (2008), and started at the same day as the vacuumed dust sampling. The EDC consisted of 4 electrostatic cloths, each with an area of 0.032 m2 , that were fixed in a 40 cm × 30 cm plastic folder. EDCs were positioned for 14 following days in the computer room and the changing room of the AM group in a horizontal position with the cloths exposed to air, usually at cupboards (or other adequate surfaces approximately 2 m above ground level) far from any ventilation systems, doors and windows. After two weeks, the folders were closed by the farmers, put into the provided envelope and sent by post to the Institute for Prevention and Occupational Medicine of the German Social Accident Insurance (IPA), Ruhr-University Bochum, Germany, for dust extraction and CHA analysis. 2.4. Dust extraction The extraction of ALK filter dust and manually collected dust was performed according to Zahradnik et al. (2011b). Briefly, 50 mg dust was transferred into a screw cap glass bottle containing 5 ml PBST (pH 7.4) and the closed bottle was then shaken for 60 min at 250 U/min. The complete solution was transferred into a 10 ml Sarstedt tube and centrifuged at 3000 × g. Aliquots of 200–500 ␮l were frozen in reaction tubes until CHA determination. Regarding the ICD samples, the dust suspension was transferred into a 12 ml Sarstedt tube, filled up to 10 ml and vortexed for 1 min. Vortexing was repeated after 30 and 60 min, and 200–1000 ␮l aliquots were frozen at −20 ◦ C until allergen analysis. Dry weight of dust was determined from 3 replicates of 100 ␮l suspensions. EDC devices were frozen overnight to eliminate any mite growth on the cloths. Each cloth was extracted by rotation in 20 ml PBST for 1 h at room temperature. After extraction, the cloth was removed, the extract was centrifuged at 3000 × g for 15 min and the supernatant was stored in aliquots at −80 ◦ C until analysis (Zahradnik et al., 2011a).

7 measurements was used for statistical calculations of the CM group percentiles. Due to the small numbers of study sites and samples and due to the heterogeneity of the investigated factors, it was not reasonable to conduct more detailed calculations such as multivariate analysis or factor modelling. Thus, the presentation of the results focused on percentiles and correlations to provide an overview of exposure. 3. Results Overall, 20 dairy farms equipped with automatic or conventional milking systems were sampled at different locations resulting in 90 ALK dust samples, 32 Ionic Breeze Quadra samples and 24 EDC measurements. Cow hair allergen (CHA) levels of all samples were above the LOD of the sandwich ELISA. 3.1. ALK filter sampling device Altogether, 60 settled dust samples from the 12 AM group settings were collected inside the cow stables and the farmers’ dwellings. The CHA levels ranged between 35.1 ␮g/g and 12,500 ␮g/g dust with highest concentrations in stable areas and lowest amounts from the mattresses (Table 3). A steep concentration gradient (Fig. 1) was found from the milking parlour to the computer room and to the dwelling areas. The median CHA load in settled dust from milking parlour/stable (7154 ␮g/g) revealed concentrations in a magnitude of 3–4 orders higher compared to the computer room and more than 60-fold and 100-fold higher than from private rooms, i.e. from the living room (median: 109 ␮g/g) and from mattresses (median: 62.9 ␮g/g). Regarding the CM group, CHA concentrations in ALK dust samples from working and living areas ranged between 37.0 and 14,130 ␮g/g with the highest median concentration (5627 ␮g/g) in the stables (Table 3) being more than 12-fold the levels from the changing rooms (median: 450 ␮g/g dust) and nearly 47-times higher than from the farmers’ mattresses, which were nonetheless still contaminated with median CHA levels of 121 ␮g/g dust. 3.2. ICD Ionic Breeze Quadra

2.5. Cow hair allergen analysis Cow hair allergen concentrations of all dust sample extracts were determined using a Sandwich ELISA as previously described by Zahradnik et al. (2011b). This sandwich ELISA is based on polyclonal antibodies against a mixture of hair extracts from different German cattle breeds and was found to offer a high sensitivity with a detection limit of 0.1 ng/ml dust extract. 2.6. Statistical analysis Statistical calculations were performed using the software package IBM SPSS for Windows, version 21.0 (Armonk, NY). Since data were not normally distributed, percentiles of the CHA concentrations were presented. Correlations were calculated using the Spearman rank correlation test. The Spearman rank correlation test was used for calculating correlations of the different sampling techniques and locations for AM and CM farms in general as well as with respect to the farmers’ practices (yes/no questions) collected via questionnaire. The Mann–Whitney-U-test for independent variables was applied for calculation of significant differences in contamination. At two CM farms, dust collection by vacuuming and Ionic Breeze in the changing and living rooms was performed for 7 days within 3 months. The variation coefficients over this period were 38.8% (changing room, farm 1), 41.4% (living room, farm1), 35.1% (changing room farm 2) and 62.7% (living room, farm2). Within these cases, the arithmetic mean of the

In total, 24 airborne dust samples were taken using the ioncharging device (ICD) Ionic Breeze Quadra for particle collection in the stable area (computer room) and in the living room of the AM farmers and 8 samples were collected in the living rooms of the CM farmers. Median CHA concentrations of 112 ␮g/g and 5.6 ␮g/g were measured in the collected dust of the computer rooms and the living rooms of the AM, respectively, while the allergen levels from the living rooms of the CM group ranged between 3.8 ␮g/g and 472 ␮g/g with a median value of 19.8 ␮g/g (Table 3). 3.3. EDC dust samples EDC sampling was performed in the computer and changing rooms of the AM farms and CHA concentrations ranged between 11.0 ␮g/m2 and 1353 ␮g/m2 (Table 3). Although the median concentration in the changing rooms was half as much as in the computer rooms, the maximum CHA concentration in EDC dust samples was detected in a changing room. 3.4. Correlation of CHA concentrations from different sampling techniques and locations Correlations of CHA concentrations in dust from different sampling techniques and locations of all 20 farms were presented in Table 4: Differences between the two farming types (AM, CM) were not significant, but significant correlations were found for ALK dust

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

83

Table 3 Cow hair allergen (CHA) concentrations of all samplings from AM farms (automatic milking) and CM farms (conventional milking). Sampling location

Number of sampled farms

Median CHA concentration

Range

CHA in settled dust samples (g/g): ALK dust collector AM Stable/Milking parlour CM Computer room AM AM Changing room CM AM Living room CM Mattress AM CM

12 8 12 12 8 12 8 12 6

7154 5627 2164 380 450 109 199 62.9 121

1164–12,500 1786–14,130 279–3628 110–905 127–1524 55.4–833 39.9–1032 35.1–474 37.0–206

CHA in airborne dust samples (g/g): Quadra Ionic Breeze (ICD) AM Computer room AM Living room CM

12 12 8

112 5.6 19.8

6.7–2140 0.8–264 3.8–472

CHA in passive dust samples (g/m2 ): EDCa Computer room AM Changing room AM

12 12

116 55.0

21.9–1264 11.0–1353

a

Type of milking system

Electrostatic dust fall collectors.

for stable/computer room, stable/changing room, stable/mattress, changing/living room and living room/mattress for the overall ALK results and/or for the 12 AM farms. Within the 8 CM farms, CHA concentrations in ALK dust from stable and mattress dust were significantly correlated (r = 0.829, p = 0.042*). Regarding EDC and ICD (Quadra Ionic Breeze) sampling, no correlation was found except for the computer rooms of the AM farms where results of both techniques from were significantly correlated (r = 0.776, p = 0.003**). 3.5. Evaluation of CHA concentrations in ALK dust with respect to questionnaire data The farmers’ practices which could lead to a direct transfer of allergens from the stable to the living area were collected via the questionnaire applying “yes/no questions” (Table 1). As most mea-

surements were performed by the ALK dust sampler and available both from living rooms and from mattresses, CHA loads in ALK dust from the private area (median, ranges) were evaluated with respect to these questionnaire data (Table 5). Comparing the median CHA concentrations in farmers’ dwellings no clear tendency of contamination could be shown for the evaluated issues. However, the trend of decreasing contamination from living to sleeping room could be illustrated again (Table 5). Moreover, correlations of the CHA loads in ALK dust samples were calculated with respect to the information on those farmers’ practices (yes/no questions). Altogether, no significant differences in CHA concentration of the different dust samples were found for the “yes/no”-questions, but correlations were significant for the following issues: - Wearing of stable clothes in private area?

Fig. 1. Cow hair allergen (CHA) concentrations in ALK filter dust (␮g/g) from different locations of 12 farms with automatic milking (AM) systems and 8 farms with conventional milking (CM) systems.

84

Table 4 Correlations of cow hair allergen (CHA) concentrations from different sampling methods and locations. ALK comp. room (ng/mg)

ICD comp. room (ng/mg)

EDC comp. room (ng/m2 )

ALK chang. room (ng/mg)

EDC chang. room (ng/m2 )

ALK liv. room (ng/mg)

ALK comp. room (ng/mg)

r (n)

All AM CM

.657* (12) .657* (12) –

ICD comp. room (ng/mg)

r (n)

All AM CM

−.056 (12) −.056 (12) –

−.182 (12) −.182 (12) –

EDC comp. room (ng/m2 )

r (n)

All AM CM

.014 (12) .014 (12) –

.014 (12) .014 (12) –

.776** (12) .776** (12) –

ALK chang. room (ng/mg)

r (n)

All AM CM

.336 (12) .336 (12) –

.189 (12) .189 (12) –

.245 (12) .245 (12) –

EDC chang. room (ng/m2 )

r (n)

All AM CM

.098 (12) .098 (12) –

−.378 (12) −.378 (12) –

.336 (12) .336 (12) –

.469 (12) .469 (12) –

ALK liv. room (ng/mg)

r (n)

All AM CM

.081 (20) .245 (12) .167 (8)

.259 (12) .259 (12) –

.434 (12) .434 (12) –

.315 (12) .315 (12) –

ICD liv. room (ng/mg)

r (n)

All AM CM

.125 (20) −.070 (12) .524 (8)

−.077 (12) −.077 (12) –

.028 (12) .028 (12) –

−.126 (12) −.126 (12) –

.039 (20) −.042 (12) .190 (8)

−.189 (12) −.189 (12) –

.247 (20) .056 (12) .595 (8)

ALK mattress (ng/mg)

r (n)

All AM CM

.378 (12) .378 (12) –

.476 (12) .476 (12) –

.524 (12) .524 (12) –

.257 (18) .217 (12) .200 (6)

−.168 (12) −.168 (12) –

.470* (18) .713** (12) .029 (6)

r = correlation coefficient; (n) = number of values. * p < 0.05. ** p < 0.01.

.486* (20) .713** (12) .381 (8)

.472* (18) .322 (12) .829* (6)

ICD liv. room (ng/mg)

.259 (12) .259 (12) – .537* (20) .392 (12) .571 (8)

−.140 (12) −.140 (12) –

.051 (18) −.084 (12) .200 (6)

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

ALK stable (ng/mg)

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

85

Table 5 Cow hair allergen (CHA) concentrations in ALK filter dust (␮g/g) from the farmers’ dwellings with respect to the questionnaire data. Yes Sampling location

N

No Median

Range

Wearing of stable clothes in private area? Living room 16 14a Mattress

144 69

39.9–1032 37.0–474

Free access of cats and/or dogs to living room? Living room 7 Mattress 6b

113 75

Free access of cats and/or dogs to sleeping room? 3 Living room 3 Mattress Daily hairwashing?c Living room Mattress Inspection round in the evening?c Living room Mattress a b c

6 6 11 11

N

Median

Range

4 4

119 78

79.1–159 35.1–116

39.9–333 44.1–190

13 12b

155 67

55.4–1032 35.1–474

165 105

39.9–333 78.6–190

17 15a

142 66

55.4–1032 35.1–474

114 69

79.1–333 35.1–116

12 12

148 69

39.9–1032 37.0–474

142 70.8

39.9–833 37.0–474

7 7

158 66.4

78.9–1032 35.1–206

CHA concentration not available in 2 cases. CHA concentration not available in 1 case. In 2 cases no yes/no response given in questionnaire.

The CHA concentrations from the changing rooms were significantly correlated to those from the living rooms (r = 0.526, p = 0.036*) in farms where farmers reported to wear their stable clothes also inside their private living area (n = 16: 8 AM and 8 CM farms). Separate evaluation for the different farm types resulted in significant correlations of CHA concentrations in stable/changing room dust (r = 0.738, p = 0.037*), and in living room/mattress dust (r = 0.786, p = 0.021*) within the AM farms and for stable/mattress samples within the CM farms (r = 0.829, p = 0.042*) when stable clothes were also worn in the private rooms. No correlations were found for those 4 farms (only AM), where stable clothes were not worn in the dwellings. - Free access of cats and/or dogs to living and sleeping room? Regarding the 7 farms (4 AM, 3 CM) with cats and/or dogs having free access both to the stable and the living room, CHA concentrations of the different rooms were not correlated, whereas at the 13 farms without access of pets to the private rooms, CHA amounts were significantly correlated for stable/changing room (r = 0.758, p = 0.003**), changing room/mattress (r = 0.629, p = 0.028*) and living room/mattress (r = 0.657, p = 0.020*). - Daily hairwashing? Within the 12 farms (7 AM, 5 CM), where farmers did not daily wash their hair for elimination of dust residues, the CHA concentrations were correlated for changing and living room samples (r = 0.769, p = 0.003**). For the 6 farms (5 AM, 1 CM), where farmers reported to wash their hair daily, the CHA loads of the stable and changing room were significantly correlated (r = 0.943, p = 0.005**), but no correlation was found for private rooms. - Inspection round in the evening? There were no correlations for those cases where the farmer made an “inspection round” through the stable in the evening without personal protective equipment. However, when no “inspection round” was done (4 AM, 3 CM), the CHA load were correlated for changing/living room (r = 0.786, p = 0.036*) and for changing room/mattress (r = 0.821, p = 0.023*).

4. Discussion Bovine hair and dander are the most important inducers of occupational allergic diseases of cattle farmers. Diverse studies reported grave dust and bovine allergen exposures in livestock farming using various sampling techniques and methodologies. Nevertheless, despite long-term knowledge of these facts no clear downward trends in dust – and consequently in allergen exposure – with the time were observed (Basinas et al., 2013). In recent years, numerous dairy farms were refitted with modern technologies including automatic milking systems. Due to the rationalized working steps one could assume that these modern installations would reduce allergen exposure inside the stables and that farm workers would be less exposed to allergens during their work. We are not aware of any study quantifying CHA concentrations in AM cow stables. 4.1. General The results of all sampling techniques revealed a wide range of CHA concentrations both for the AM and the CM stables. As the different sampling methods resulted in different units (␮g/g and ␮g/m2 ) direct comparison of the results was limited, but a common steep gradient became apparent. As expected, the CHA concentration decreased from the stables to the bedrooms (stable/milking parlour → computer room → changing room → living room → mattress throughout all sampling methods) (Table 3, Fig. 1). This phenomenon has also been described in previous studies quantifying dust and Bos d 2 allergens in dairy cow stables (Hinze et al., 1997; Berger et al., 2005). Regarding CHA, comparison is limited due to the small number of studies on CHA concentrations in dust: Zahradnik et al. (2011a) determined high CHA concentrations by the same EDC methodology in dairy cow stables (median 51,700 ␮g/m2 ) and also reported a sharp decrease to the CHA load in the changing rooms (median 104.5 ␮g/m2 ) and the farmers’ dwellings (median: living room 4.3 ␮g/m2 , bedroom 11.8 ␮g/m2 ). A recently published study, likewise applying the EDC, found a CHA gradient of nearly 5000-fold magnitude comparing CHA levels in dairy stables and bedrooms (Schluenssen et al., 2015). Although there is no use of AM systems described their results add well to our EDC data from computer and changing rooms, which range in between.

86

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

4.2. Comparison of the CHA results from AM and CM farms In this study, median CHA concentrations from the 12 AM farms were generally lower compared to the 8 CM farms at corresponding locations, except for the stables. Median CHA concentration per mg airborne dust (ICD Ionic Breeze) were nearly 20-times lower than per mg reservoir dust (ALK filter sampling device) both for the computer room and the living room of the AM farms. In the living rooms of the CM group, the median airborne CHA level measured by the Ionic Breeze ICD was 10-times lower than the corresponding median CHA concentration in ALK reservoir dust. It is apparent that the changing rooms and the private rooms at AM farms were less CHA contaminated than at CM farms. Probably, the computer room for operating the milking robot serves as “barrier” for allergen transfer or the less close animal contact leads to lower dust exposure and transfer. The median CHA loads of the EDC samples, which were only performed in the AM group, were twice as high in the computer room compared to the changing room while ranges were comparable.

outdoor activities. Interestingly, they found that robotic milking was strongly related to inhalable dust exposure and on average led to 2-fold higher dust exposure of workers when compared with workers from farms with parlour or pipe milking. As reasons they suggest that the reduced number of employees attending an unchanged number of animals in AM stables and the intensified involvement of the employees in other, more dustier tasks (reparations, distribution of bedding, handling of silos etc.), and also different activity patterns due to automatic systems regularly running in parallel (rail feed disperser, etc.) may influence the dust load within the dairy shed and explain the observed increased personal levels of dust exposure. Regarding these findings, one could expect that the CHA concentrations would rise in parallel with the personal dust exposure in AM farms. However, in the present study, only median CHA levels in ALK dust from the AM stables were higher compared to the CM stables while ranges were in the same order of magnitude. In contrast, the allergen load at all other sampling locations was generally lower for the AM farms. Probably the allergen transfer from working to private area is reduced in AM farms as farmers spend less time in the stables.

4.3. Comparison of CHA results with respect to questionnaire data Despite the CHA concentration gradient from stable to bedroom, CHA concentrations in some farmers’ homes were obvious and suggest that the farmers enter their homes in their work clothes (n = 16 of 20, thereof 8 AM and 8 CM farms) leading to allergen transfer from the workplace through clothes and hair of the workers. This was supported by the significant correlation of ALK filter samples from changing room and living room (Table 4), as well as of the ALK samples from stable/changing room and living room/mattress within the AM group and from stable and mattress within the CM group and also by the significant correlations in farms (changing/living room), where farmers reported to wear their stable clothes also inside their dwellings. Allergen transfer via workers’ clothes has already been shown for other species (e.g.; mouse and cat allergens) by previous studies (Karlsson and Renstrom, 2005; Krop et al., 2007). The free access of cats and/or dogs as CHA carrier in the stable and into the dwellings was also suspected to be a source for CHA transfer. At least in 7 of 20 participating farm, cats and/or dogs with access to stables were allowed to enter the living room and in 3 cases also the bedroom. However, probably due to the heterogeneity of the data, vague answers at questionnaires and unknown confounders, neither pets’ access nor frequency of hair washing nor the inspection round in the evening showed a clear tendency for the allergen load in private rooms and CHA concentrations were only correlated in few cases. Moreover, the small number of cases for the different topics must be taken into account. 4.4. Comparison to other studies on dust exposure in AM stables We are not aware of any study on bovine allergen exposure measurements at farms using AM systems. Only a case report by Korinth et al. (2005) referred to a young female farmer with sensitization against cow dust-derived allergens and daily allergic bronchial asthma symptoms associated with allergen exposure by milking with a conventional device. After implementation of a milking robot with reduction of daily allergen exposure from >2 h to 10 min, symptoms and allergy parameters decreased, but no CHA measurement was performed. Basinas et al. (2014) investigated dust and endotoxin concentrations in the inhalable air using personal samplers in 77 subjects from 26 dairy farms in Denmark, of which five performed fully automatic milking and two farms installed this system during the course of the study. Measurements covered a working period of 290 min during summer and 280 min during winter including both indoor and

4.5. Comparison and suitability of the different sampling techniques As personal sampling is comparatively time-consuming and complex, this study focussed on applying easily applicable quantification strategies for CHA exposure. Even if sampling of airborne dust on filters may be regarded as gold standard (Renstrom, 2002), passive dust sampling using EDCs could be a low-cost and simple alternative without producing any dust dispersion and noise. Applying this EDC method, Zahradnik et al. (2011b) found median CHA levels of 51,700 ␮g/m2 in 37 cow stables, between 4.3 and 63.1 ␮g/m2 in the homes of cattle farmers, and 104.5 ␮g/m2 in the changing rooms which is in the range of the present study. With respect to the easy access and long sampling time, the EDC folders may be susceptible to several unknown confounders such as unknown dust-producing events or accidental contamination (e.g.; pets sitting at the folders, etc.), potentially accounting for the two maximum values of 1264 ng/m2 (computer room) and 1353 ng/m2 (changing room). A larger case number and the investigation of more different rooms are needed for definitive assessment. The present study confirmed that dust collection via ALK filter device is easy to perform and provided reliable data. This methodology has been well established in numerous studies before and described as one usual tool for exposure assessment for allergen monitoring (Raulf et al., 2014). For long-term measurements of airborne exposure, the ICD Ionic Breeze appeared to be a silent and comfortable alternative to personal samplers. However, for interpretation of the ICD results, it must be taken into account that overloading of the plates can result in decreased binding efficiency and allergen collection can become less efficient. As the ICD Ionic Breeze Quadra is primarily developed as an air filter system, routine ICD use for airborne allergen sampling should be examined in further and more extensive studies. Summarizing the results, it became apparent that each of the applied methods has its strengths and that a definite interpretation of the results is certainly limited by the small number of samples. Sampling strategies must be adopted depending on the specific aims and scopes of the intended investigation, and existing recommendations for suitable monitoring strategies of different exposure scenarios must be considered and pursued (Raulf et al., 2014). Standardization of sampling and analytical procedures is needed for a better comparison of data and for estimating the impact of the exposure dose on sensitization and allergic symptoms.

A. Böhlandt et al. / International Journal of Hygiene and Environmental Health 219 (2016) 79–87

5. Conclusion A decrease of CHA concentration from the stable to the computer and changing room and to the private rooms was quantified for both dairy farm types with and without milking robots. The installation of AM systems is likely to reduce the CHA exposure in the working and living environment of dairy farmers, at least if no other dust-producing or dispersing tasks increase in parallel. However, in the stables the CHA concentrations were in comparable ranges. Nevertheless, relevant allergen concentrations were present at all sampled sites, and allergens are likely to be transferred from stables to the private rooms via clothes and pets. All sampling methods suggest to being practicable for simple and effective measurement of CHA concentration, but strengths and shortcomings must be regarded. The results demonstrate the relevance of controlling and reducing allergen concentrations at the workplaces and the living environments as an essential prevention strategy. Further monitoring studies on CHA exposure in the farmers’ environment with higher sample numbers would be desirable and standardized sampling strategies and analytical procedures have to be defined. Conflict of interest The authors declare no conflict of interest. Acknowledgement The authors would like to thank all farmers and their families for their participation in the study and their important support. References Basinas, I., Sigsgaard, T., Erlandsen, M., Andersen, N.T., Takai, H., Heederik, D., Omland, O., Kromhout, H., Schlunssen, V., 2014. Exposure-affecting factors of dairy farmers’ exposure to inhalable dust and endotoxin. Ann. Occup. Hyg. 58 (6), 707–723, http://dx.doi.org/10.1093/annhyg/meu024. Basinas, I., Sigsgaard, T., Kromhout, H., Heederik, D., Wouters, I.M., Schlunssen, V., 2013. A comprehensive review of levels and determinants of personal exposure to dust and endotoxin in livestock farming. J. Expo Sci. Environ. Epidemiol., http://dx.doi.org/10.1038/jes.2013.83. Berger, I., Schierl, R., Ochmann, U., Egger, U., Scharrer, E., Nowak, D., 2005. Concentrations of dust, allergens and endotoxin in stables, living rooms and mattresses from cattle farmers in southern bavaria. Ann. Agric. Environ. Med. 12 (1), 101–107. Custis, N.J., Woodfolk, J.A., Vaughan, J.W., Platts-Mills, T.A.E., 2003. Quantitative measurement of airborne allergens from dust mites, dogs, and cats using an ion-charging device. Clin. Exp. Allergy 33 (7), 986–991, http://dx.doi.org/10. 1046/j.1365-2222.2003.01706.x. Elholm, G., Omland, O., Schlunssen, V., Hjort, C., Basinas, I., Sigsgaard, T., 2010. The cohort of young danish farmers – a longitudinal study of the health effects of farming exposure. Clin. Epidemiol. 2, 45–50, org/10.2147/CLEP.S9255. Heutelbeck, A.R.R., Janicke, N., Hilgers, R., Kutting, B., Drexler, H., Hallier, E., Bickeboller, H., 2007. German cattle allergy study (cas): Public health relevance of cattle-allergic farmers. Int. Arch. Occup. Environ. Health 81, 201–208, http://dx.doi.org/10.1007/s00420-007-0207-y. Heutelbeck, A.R.R., Junghans, C., Esselmann, H., Hallier, E., Schulz, T.G., 2009. Exposure to allergens of different cattle breeds and their relevance in

87

occupational allergy. Int. Arch. Occup. Environ. Health 82 (9), 1123–1131, http://dx.doi.org/10.1007/s00420-009-0400-2. Hinze, S., Bergmann, K.C., Lowenstein, H., Hansen, G.N., 1997. Cow hair allergen (bos d 2) content in house dust: correlation with sensitization in farmers with cow hair asthma. Int. Arch. Allergy Immunol. 112 (3), 231–237. Karlsson, A.S., Renstrom, A., 2005. Human hair is a potential source of cat allergen contamination of ambient air. Allergy 60 (7), 961–964, http://dx.doi.org/10. 1111/j.1398-9995.2005.00796.x. Kogevinas, M., Anto, J.M., Sunyer, J., Tobias, A., Kromhout, H., Burney, P.G., 1999. European Community Respiratory Health Survey Study Occupational asthma in europe and other industrialised areas: A population-based study. Lancet 353 (9166), 1750–1754, http://dx.doi.org/10.1016/s0140-6736(98)07397-8. Korinth, G., Broding, H.C., Uter, W., Drexler, H., 2005. Secondary prevention of allergic symptoms in a dairy farmer by use of a milking robot. Clin. Mol. Allergy: CMA 3, 8, http://dx.doi.org/10.1186/1476-7961-3-8. Krop, E.J.M., Doekes, G., Stone, M.J., Aalberse, R.C., van der Zee, J.S., 2007. Spreading of occupational allergens: laboratory animal allergens on hair-covering caps and in mattress dust of laboratory animal workers. Occup. Environ. Med. 64 (4), 267–272, http://dx.doi.org/10.1136/oem.2006.028845. Noss, I., Wouters, I.M., Visser, M., Heederik, D.J.J., Thorne, P.S., Brunekreef, B., Doekes, G., 2008. Evaluation of a low-cost electrostatic dust fall collector for indoor air endotoxin exposure assessment. Appl. Environ. Microbiol. 74 (18), 5621–5627, http://dx.doi.org/10.1128/aem.00619-08. Radon, K., Danuser, B., Iversen, M., Jorres, R., Monso, E., Opravil, U., Weber, C., Donham, K.J., Nowak, D., 2001. Respiratory symptoms in european animal farmers. Eur. Resp. J. 17 (4), 747–754, http://dx.doi.org/10.1183/09031936.01. 17407470. Raulf, M., Buters, J., Chapman, M., Cecchi, L., de Blay, F., Doekes, G., Eduard, W., Heederik, D., Jeebhay, M.F., Kespohl, S., Krop, E., Moscato, G., Pala, G., Quirce, S., Sander, I., Schlunssen, V., Sigsgaard, T., Walusiak-Skorupa, J., Wiszniewska, M., Wouters, I.M., Annesi-Maesano, I., 2014. Monitoring of occupational and environmental aeroallergens - eaaci position paper concerted action of the eaaci ig occupational allergy and aerobiology & air pollution. Allergy 69 (10), 1280–1299, http://dx.doi.org/10.1111/all.12456. Rautiainen, J., Pentikainen, J., Rytkonen, M., LinnalaKankkunen, A., Pelkonen, J., Virtanen, T., Mantyjarvi, R., 1996. Molecular analysis of allergenic proteins in bovine dander. Allergy 51 (6), 378–382, http://dx.doi.org/10.1111/j.1398-9995. 1996.tb04633.x. Renstrom, A., 2002. Exposure to airborne allergens: A review of sampling methods. J. Environ. Monit. 4 (5), 619–622, http://dx.doi.org/10.1039/b202756a. Reynolds, S.J., Nonnenmann, M.W., Basinas, I., Davidson, M., Elfman, L., Gordon, J., Kirychuck, S., Reed, S., Schaeffer, J.W., Schenker, M.B., Schlunssen, V., Sigsgaard, T., 2013. Systematic review of respiratory health among dairy workers. J. Agromed. 18 (3), 219–243, http://dx.doi.org/10.1080/1059924x.2013.797374. Schluenssen, V., Basinas, I., Zahradnik, E., Elholm, G., Wouters, I.M., Kromhout, H., Heederik, D., Bolund, A.C.S., Omland, O., Raulf, M., Sigsgaard, T., 2015. Exposure levels, determinants and ige mediated sensitization to bovine allergens among danish farmers and non-farmers. Int. J. Hyg. Environ. Health 218 (2), 265–272, http://dx.doi.org/10.1016/j.ijheh.2014.12.002. Waser, M., Schierl, R., Von mutius, E., Maisch, S., Carr, D., Riedler, J., Eder, W., Schreuer, M., Nowak, D., Braun-Fahrlander, C., Team, A.S., 2004. Determinants of endotoxin levels in living environments of farmers’ children and their peers from rural areas. Clin. Exp. Allergy 34 (3), 389–397, http://dx.doi.org/10.1111/j. 1365-2222.2004.01873.x. Ylonen, J., Mantyjarvi, R., Taivainen, A., Virtanen, T., 1992. Comparison of the antigenic and allergenic properties of 3 types of bovine epithelial material. Int. Arch. Allergy Immunol. 99 (1), 112–117, http://dx.doi.org/10.1159/000236343. Zahradnik, E., Sander, I., Bruckmaier, L., Flagge, A., Fleischer, C., Schierl, R., Nowak, D., Sueltz, J., Spickenheuer, A., Noss, I., Bruening, T., Raulf-Heimsoth, M., 2011a. Development of a sandwich elisa to measure exposure to occupational cow hair allergens. Int. Arch. Allergy Immunol. 155 (3), 225–233, http://dx.doi.org/ 10.1159/000319839. Zahradnik, E., Sander, I., Bruckmaier, L., Flagge, A., Fleischer, C., Schierl, R., Nowak, D., Sultz, J., Spickenheuer, A., Noss, I., Bruning, T., Raulf-Heimsoth, M., 2011b. Development of a sandwich elisa to measure exposure to occupational cow hair allergens. Int. Arch. Allergy Immunol. 155 (3), 225–233, http://dx.doi.org/ 10.1159/000319839.