Annals of Occupational Hygiene Advance Access originally published online on June 15, 2006
Annals of Occupational Hygiene 2006 50(7):705-715; doi:10.1093/annhyg/mel027
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Large Organic Aerosols in a Dynamic and Continuous Whole-Body Exposure Chamber Tested on Humans and on a Heated Mannequin
1 Department of Applied Environmental Science (ITM), Stockholm University Stockholm, Sweden
2 Department of Medicine, Occupational and Environmental Dermatology Karolinska Institutet, Stockholm Sweden
3 Stockholm Centre for Public Health, Stockholm County Council Stockholm, Sweden
4 Department of Medicine, Karolinska Institutet Stockholm, Sweden
*Author to whom correspondence should be addressed. Tel: +46-8-674 7648; fax +46-8-674 7638; e-mail: lennart.lundgren{at}itm.su.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Exposure to large airborne organic aerosols may cause respiratory and skin symptoms. The use of human exposure chambers permits safe mechanistic studies of the effect of inhalation or dermal deposition of such particles. The performance of a dynamic and continuous whole-body human exposure chamber using turbulent air mixing during exposure to these organic aerosols of humans and of a new heated mannequin was evaluated. Variability of temporal and spatial distribution of the airborne particle concentration, and aerodynamic aerosol size distribution of the inhalable fraction, were evaluated. The temporal and spatial distribution of these aerosols close to the breathing zone during an exposure session was typically
10%, which is low for airborne particles of this size. In a larger section around a human, only slightly higher spatial variation was found. Variability between exposure sessions was also low (<10%). Only limited effect of relative humidity for the organic aerosols was observed. The aerodynamic particle size distribution curves differed slightly, but some were comparable to those in occupational environments. The outcome of the performance tests as measured with the heated mannequin was almost the same as with humans, indicating that the mannequin could be used in preparatory tests in this type of chamber.
Keywords: exposure chamber • large aerosols • mannequin • organic • particle size • performance • spatial variability • temporal variability
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Exposure to airborne organic aerosols can induce sensitization, respiratory symptoms and bronchial hyper-responsiveness, and can also cause contact dermatitis. Besides exposure level, the risks associated with the exposure are related to the size of the aerosol, the chemical and physical characteristics of the aerosol, and the morphology, e.g. fibers or other elongated shapes. Particle size is crucial, since it governs both the deposition rate and site in the respiratory tract. The use of a chamber for exposing humans allows mechanistic studies of the effect of inhalation or dermal deposition of airborne particles from a well-defined exposure in a safe setting. Such studies can lead to a determination of a cause and effect (dose–response) relationship while avoiding confounding variables that almost always occur in occupational field studies. The ideal experimental set-up should deliver an aerosol in a concentration and route similar to those in workplaces.
For inhalation challenge studies in exposure chambers, a uniform and homogenous distribution of the aerosol is essential, since non-uniform distribution might result in high exposure variability. The temporal and spatial distribution of the aerosols is critical (Byrne et al., 1995), since it may affect the delivery of the dose to the subject. Results from studies of effects in the lower respiratory tract of exposure to different large organic dust particles have been reported from one such dynamic whole-body exposure chamber (Gripenbäck et al., 2003, 2005; Grunewald et al., 2003). A few other dynamic human chambers designed for smaller/finer aerosols have also been described (Søstrand et al., 1997; Ghio et al., 2000; Taylor et al., 2000). Dynamic whole-body exposure chambers may also be used when developing new measuring devices and exposure assessment techniques. Such an approach has recently been described for a device that removes particles from the skin (Lundgren et al., 2006).
The technical performance of a chamber is crucial for the interpretation of an exposure study. It may vary between individual chambers depending on whether a continuous and dynamic system is used, on the homogeneity of the particles, on the airborne particle concentration, on the particle size, on the chemical composition, on the electric charge of the particles, and on the temperature and humidity in the air during exposure. The present aim was to investigate the performance of a dynamic and continuous whole-body exposure chamber constructed for turbulent air mixing during generation of organic particulates of pinewood, bakery dust and cornstarch, which hereafter will be addressed as large organic aerosols. This chamber has been used in different exposure studies and will be used in future studies of aerosols. The outcome of the performance of this chamber will be valuable when judging and in interpretation of clinical results found in those studies. Temporal variability during and between exposure sessions, and spatial distribution in the chamber during an exposure session, were determined. In addition, the effect of humidity on the temporal variability was derived from data from five different exposure studies. Further, the aerodynamic particle size distribution was determined as was the relationship between measurements with different ‘size-selective’ samplers. Since the set-up of exposure conditions for previously untested compounds needs high accuracy and is time-consuming, an additional aim was to investigate the utility of a new heated mannequin as a surrogate for humans in initial and preparatory chamber studies.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental design
Five exposure series were performed in a human whole-body exposure chamber at an airborne concentration not uncommon in occupational environments. Exposure characteristics such as temporal and spatial variability and aerosol size distribution were determined close to a subject's or a heated mannequin's breathing zone. When exposing humans, they all moved into the exposure chamber after the dust had reached the targeted and stable concentration level. The humans sat calmly in the middle of the chamber during exposure, not performing any activity. Summarized exposure data from three of the exposure series have previously been published, when reporting clinical results (Gripenbäck et al., 2003, 2005; Grunewald et al., 2003). All the participants gave their informed consent, and the Regional Ethics Committee at Karolinska Hospital and Karolinska Intitutet, Solna, Sweden, approved the studies.
Exposure chamber
The exposure chamber, designed and built as a dynamic system for continuous generation of aerosols, has been described previously (Lidén et al., 1998). This chamber was totally dismantled and rebuilt at a new location, using most of the old materials when possible. A new frequency converter (CS 300, ABB, Sweden) was installed for a continuous and smoother regulation of the ventilation system. A ventilation hood was added to the whole exposure set-up, in where the generation equipments could be placed and thus avoiding any leakage from the generator out to the connecting room. The chamber was also more thoroughly checked and adjusted for leakages than before the removal. A photograph of the chamber at its new location as well as a schematic drawing of it are shown in Fig. 1.
|
In general, filtered room air flows continuously through the chamber (floor 1.8 x 1.5 m2, height 2.1 m; volume 5.7 m3), at 1.1–1.6 m3 min–1 (corresponding to
12–17 air exchanges per hour). The airflow from the aerosol generator contributes 20% of total flow. To create a turbulent mixing, the airflow from the aerosol generator passes through a conical duct in the ceiling of the chamber. The air is removed through a valve in one of the walls to a connecting sluice, and from the sluice through a funnel in the ceiling. By keeping the exposure room and the sluice at appropriate pressure differences, contamination by ambient air into the chamber and from the chamber to the ambient air is eliminated. The original valve between the chamber and the sluice was replaced with an adjustable one in the new set-up, making it a lot easier to achieve the appropriate pressure differences. Temperature and humidity during exposure cannot be changed or targeted, since neither heating nor humidity units are installed.
Dust generation
A rotating brush generator, RBG-1000, from Palas GmbH, Karlsruhe, Germany, was used in all tests. The airflow through the generator head gives a secondary flow between 2 and 5 m3 h–1, and can be reduced to 0.5 m3 h–1 by changing dispersion covers inside the generator. To eliminate potential static charges in the dust, the aerosol stream from the particle generator passes through a stationary neutralizer (a radioactive 85Kr source) to the inlet air delivery system. Parameters such as feed rate, brush speed and airflow through the generator were optimized for each separate challenge.
Depending on the targeted aerosol concentration, the different dust reservoirs were loaded with 3–30 g of dust, sufficient for 2 h generation including a start-up period and a 60 min exposure session. For a longer exposure period, the reservoir has to be refilled. It then takes 30 min for the concentration to reach the desired level again.
A new heated mannequin
A life-sized heated mannequin was constructed by modifying a commercially available doll (POLYFORM GmbH & Co. KG, Rinteln, Germany). The mannequin is heated to 36°C (surface temperature) with a 100 W DC-powered regulated heater, placed inside the chest region of the mannequin. The heat is distributed inside the mannequin with two small fans with one that forces the heat through a tube up to its head. During an initial 30 min exposure session, the temperature was measured every second with eight sensors placed around the upper parts of the mannequin or of a human subject in the chamber. The temperature rose on an average of 1.4°C (range 1.3–1.5°C) around the human subject and 0.2°C (range 0.1–0.2°C) around the mannequin. Humidity as measured with a stationary humidity and temperature transmitter (Model 62102; Hygrocontrol, Hanau, Germany) placed close to the subject increased by 0.2% for a human and decreased by 0.9% for the mannequin. Since the tests were performed during different periods on the same day, the slight differences might be due to external factors.
Dust specification
Wheat flour was selected from two batches (wheat flour 1 and 2) of standard quality grade used at a large Swedish bakery. For a separate exposure study (wheat flour 3), material from the second batch was mixed with small amounts of fungal 
-amylase (Fungamyl 2500 BG, Novo Nordisk A/S, Bagsvaerd, Denmark). Wheat flour consists mainly of starch and proteins, with a protein content of 10–15%. The targeted airborne concentration of wheat flour was 5 mg m–3 in all studies, and the -amylase level 40 ng m–3 in the third study.
Glove powder (AbsorboTM) from National Starch and Chemical Corporation, Bridgewater, NJ, USA, was used. According to the producer, it was prepared by processing cornstarch in accordance with the requirements of the US Pharmacopeia Monographs on Absorbable Dusting Powder as well as the British Pharmacopeia Monographs on Steralisable Maize Starch. It was analyzed by us with low-temperature ashing and contained 97% organic materials. Magnesium oxide (MgO) was detected in the inorganic fraction using X-ray diffraction. The targeted airborne concentration of glove powder in the chamber was 6 mg m–3.
Cornstarch (Maizena, Unilever Bestfoods, Sweden) is a pure homogeneous powder. The targeted airborne concentration of the cornstarch in the chamber was 8 mg m–3.
Pinewood dust was collected from an exhaust box connected to a sanding machine at a wood flooring factory (Rappgo AB, Braås, Sweden). The dust was sieved (<100 µm) by us and stored in argon before being used. The targeted airborne concentration of pinewood dust in the chamber was 6 mg m–3.
Chamber performance measurements
Sampling
After a start-up period of 30 min, the airborne dust concentration was measured and evaluated for temporal distribution during an exposure of at least 30 min, typically 60 min. Measurements were performed with a direct-reading instrument based on infrared light scattering, Casella AMS950 (Casella SIMPEDS, Casella London Ltd, London, UK), with the sensor placed close to the subject's breathing zone. This instrument was checked daily with the calibration element supplied with the equipment, and calibrated for each dust using basic gravimetric analytical techniques. The Casella AMS950 is an instrument suitable and sensitive for fine particles but after calibration works quite well also for larger particles with a substantial fraction of small aerosols. Continuous signals from the direct-reading dust instrument as well as the hygrometer (Model 62102; Hygrocontrol, Hanau, Germany) was recorded online and in real-time using a LabView program (National Instruments Corporation, USA) running on a laptop computer that stored data every 2 s for later calculations and displayed a plot of the airborne concentration versus time on the computer screen. No extra smoothing of the signal than the instruments normal time constant has been used.
The variability between exposure sessions when aiming at a targeted concentration was evaluated using data from exposure challenges with humans and the mannequin. The targeted concentration was kept by adjusting the aerosol generator based on real-time results from the direct-reading instrument. This was done by manual adjustments of the reservoir feed rate controller when needed. The variability was assessed from measurements with conventional air sampling devices for inhalable (INH) dust, for open-faced total (OPFT) dust and for respirable (RESP) dust. The choice of samplers was related to current Swedish occupational exposure limits (OEL) and their recommended sampling methods. However, for both wheat flour and pinewood dust, the OEL changed from OPFT to INH dust during the period of these studies. INH dust (EN481 1993) was sampled with IOM samplers (SKC Ltd, Dorset, UK) and OPFT dust was sampled in 37 mm open-faced cassettes (Millipore, Bedford, Massachusetts, USA). For both these samplers the flow rate was 2 liters min–1 (LPM). RESP dust (EN481 1993), defined as airborne particles that can penetrate into the gas-exchange regions of the human lungs, was sampled with cyclones (Casella SIMPEDS, Casella London Ltd, London, UK), at a flow rate of 1.9 LPM. Dusts were collected on membrane filters made of mixed esters of cellulose, pore size 0.8 µm for OPFT and INH dust and 8.0 µm for RESP dust (Millipore, Bedford, MA). Both filters have adequate particle collection efficiency for dust of this size. For sampling and analysis of -amylase, membrane filters made of pure cellulose acetate were used.
Since subjects may change position during an exposure session, the characteristics of the spatial distribution of the aerosol are important. Sampling for the analysis of the spatial distribution was performed with OPFT dust cassettes. Two different experimental set-ups were investigated. The first was based on two measuring points very close to the breathing zone on each side of the subject or mannequin. The other experimental set-up used eight to nine different measuring points placed within a section of 1.5 m3 around the upper part of the subject or mannequin (Fig. 2). The duration of each test was normally 60 min.
|
For the sampling of the aerodynamic particle size distribution of the inhalable fraction, cascade impactors PIDS (personal inhalable dust spectrometer) (Gibson et al., 1987) were used and run at 2 LPM. The impactor plates were coated with 10% apiezone in toluene.
Analysis
Before and after sampling, the membrane filters and impactor plates were weighed on a microweighing balance MT5 (membrane filters) and on a high-precision weighing balance AT261 (impactor plates), both from Mettler-Toledo AG, Greifensee, Switzerland, in a controlled weighing room (relative humidity 50% and temperature 21°C). The aerodynamic particle size distribution curve based on sampling with the PIDS impactor was determined by assuming the distribution curve to be bimodal and calculated using the model suggested by Lidén et al. (2000). The analysis and the methods used for the determination of -amylase were as suggested by Lillienberg et al. (2000).
Presentation of results and statistics
The results for the different airborne dust concentration are presented in graphs containing quantile box plots. The outer sides of the rectangular box express the 25 and 75% quantiles (quartiles), the horizontal line inside the box, the median, and the vertical line represents the range (minimum and maximum). The skewed rhomboid shows the 95% confidence intervals, with the average value at the widest point of the rhomboid. The temporal and spatial variation, during and between exposure sessions, expressed as coefficient of variations (CVs), are the standard deviation of direct-reading measurement or the filter measurements divided by the actual average of those measurements.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Temporal distribution of airborne dust during an exposure session
The temporal distribution during at least 30, typically 60 min of exposure after the 30 min warm-up period, expressed as the CV, is presented in Fig. 3. Since no difference was observed between the three wheat flours, all these data are plotted together. The temporal distribution did not differ between the different dusts or whether measurements were performed around a human or the mannequin.
|
Variability in exposure between exposure sessions
Variability in OPFT and RESP dust between different challenges of humans with wheat flour, glove powder and pinewood dust was generally low (Fig. 4). Too few measurements were available for an evaluation of INH dust. For the low level of
-amylase (target value of 40 ng m–3) in wheat flour 3, the corresponding variation measured as OPFT dust was 16% (N = 5).
|
The variability between exposure sessions for the different dust samplers (INH, OPFT and RESP dust) was evaluated for pinewood dust and is illustrated in Fig. 5. The variability for RESP dust was low (CV 8.8%) despite the low airborne concentration.
|
Spatial distribution of airborne dust
The spatial distribution of OPFT dust between two measuring points on each side of the subject's or mannequin's breathing zone is shown in Fig. 6. It demonstrates similar variability for the different dust types regardless of whether measurements were performed with a human or the mannequin. Also, when seven to nine different sampling points in a section of
1.5 m3 around the upper part of the subject or mannequin were analyzed, there was low spatial OPFT dust variability (Table 1). For the low level of -amylase (target value of 40 ng m–3) in wheat flour 3, CV was 21% (N = 5) for the first experimental set-up.
|
|
Effect of relative humidity
The effect of relative humidity in the chamber on the temporal distribution of these types of organic aerosols during an exposure can be seen in Fig. 7, where data are based on values for both mannequin and humans, as there was no difference observed between them.
|
Airborne particle size
The aerodynamic particle size for all different types of organic aerosols as measured with the PIDS impactors is shown as mass frequency distribution curves in Fig. 8. The aerosol concentration during these tests was close to the targeted levels.
|
The concentration fractions found between the different ‘size-selective’ samplers (INH, OPFT and RESP dust samplers) for wheat flour, glove powder and pinewood dust are shown in Table 2.
|
The particle size and the morphology of airborne wheat flour, glove powder and pinewood dust is shown in Fig. 9.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The technical performance of a human whole-body exposure chambers was studied during continuous exposure to four different large organic aerosols: wheat flour, glove powder, cornstarch and pinewood dust. The temporal and spatial distribution of the aerosols in the chamber showed very small variability, which was similar for the different dust types, and reproducibility between exposure sessions was good. The effect of humidity on temporal variation was limited, despite the absence of climate-control equipment in the chamber. The use of a newly constructed heated mannequin as a surrogate for a human subject was demonstrated to be useful in the time-consuming set-up of exposure conditions for human studies.
The exposure chamber for human challenge studies was designed to keep a homogenous and continuous distribution of dust during exposure, and is based on a turbulent mixing procedure of the dust and airflow. Dust from an aerosol generator mixes with the general airflow and can easily be adjusted to obtain a desired exposure level. The chamber has a neutralizer for eliminating static charges in the dust.
The homogeneity of the airborne dust distribution in the chamber was investigated for both temporal and spatial variation during each exposure session with either human subjects or the mannequin. Further, the variability in exposure level between different exposure sessions days or months apart, when a targeted air concentration of a certain level was chosen, was also evaluated.
To avoid the effect of high exposure peaks, low temporal variation during exposure is essential in studies of effects in humans. Independently of dust type, the temporal variation during an exposure session averaged <8%. Similar variability (7–11%) during exposure with wheat flour has previously been reported for this chamber before it was dismantled and rebuilt (Lidén et al., 1998), indicating that the characteristics of the chamber did not change after rebuilding. The variability in this chamber was also lower than what has previously been reported for smaller inhalation systems. Fabries et al. (2000) reported a temporal variation of 12% for wheat flour in a small inhalation chamber (0.14 m3), and O'Shaughnessy et al. (2004) found a temporal variation of 13.3% for grain dust generated with a Wright dust feeder (Wright, 1950) in a hooded human exposure system. Thus, the reliability of exposure data from the current chamber will be high.
When interpreting data from provocation studies, it is important to keep similar exposure characteristics from one exposure session to another. In the present studies, the variability between sessions for OPFT dust (CV < 10%) was almost as low as that during one exposure session, and for the very low fraction of airborne -amylase (0.0008% in the OPFT dust) the temporal variation was only 16%. INH dust, containing larger particles, displayed a slightly larger variability (CV 15% for pinewood dust). The variability of the respirable fraction was low (10%) for cornstarch and pinewood dust, but considerably larger for wheat flour dust (20–36%). Such a high variability might be explained by non-homogeneity of wheat flour, with large variation in the content of small particles. Consequently, the respirable fraction might vary considerably despite similar INH and OPFT exposure levels.
Since subjects may change position during an exposure session, although instructed to remain seated, the characteristics of the spatial distribution of the aerosol are of great importance. The spatial distribution based on two measuring points close to the breathing zone on each side of the subjects/mannequin showed an average variation of <10%, but with a larger interval than for temporal variation. For the very low fraction of airborne -amylase, a spatial variation of only 21% was found. We consider that a spatial variation >20% for the main fraction of dust must be regarded as unsatisfactory for two such close measuring points in the breathing zone, and that such large variation may affect the interpretation of data.
Even though the chamber lacked the possibility to control and regulate humidity and temperature, it was possible to evaluate the effects on technical performance in a wide range (5–55%) of relative humidities. It could be expected that lower humidity would increase the exposure variability, but no such effect was found. Since the evaluated humidity interval was quite large, the results indicate that the humidity was not a crucial parameter in the performance of this chamber and for these aerosols. For more hygroscopic dusts, greater influence from humidity may be expected.
In studies of the respiratory effects of aerosol exposure, knowledge of the airborne particle size is essential, since this will determine the deposition site of particles in the respiratory tract. The aerodynamic particle size distribution curves of the inhalable fraction were similar for all studied organic aerosols in the chamber. With the exception of glove powder, these were bimodal with average particle sizes of 10 µm for the main fraction and with the other maximum between 20 and 30 µm for wheat flour and pinewood dust and at 50 µm for cornstarch. We have previously reported data on airborne wheat flour particle size sampled with PIDS impactor in the same exposure chamber and from a through mixing process at a bakery (Lidén et al., 1998) which conflicts with the current result. After publication of that study, we discovered that the material of the impactor plates was hygroscopic, which had a critical and large influence on the analytical result. The previous wheat flour particle size curves have not been reproduced since the hygroscopic effect had been attended to and minimized. On the contrary, the wheat flour distribution curve presented here has been reproduced on several occasions. We are not aware of reports on airborne particle size distribution of wheat flour in other human whole-body exposure chambers, but in a much smaller inhalation chamber (Fabries et al., 2000), a corresponding aerodynamic particle size distribution curve was presented and the curve was found to be almost unimodal, with a maximum of 5 µm. In that study, the aerosol generation system was based on a cyclone that blocked larger particles from reaching the chamber, resulting in that 98.5% of the wheat flour particles being <10 µm. In our chamber, by contrast, 52–59% of the inhalable dust was <10 µm as was evaluated from the cumulative distribution curve. But we have to be aware that using a cyclone or other pre-treatment of the test powder may change the dust characteristics. Wheat flour in air may after such treatment contain a higher fraction of proteins since these seem to be predominately found in the finer fraction of the powder. Thus, the choice of aerosol generating system is crucial, and the aerosol particle size distribution must be considered in interpretation of outcome measurements in exposed subjects.
In provocation studies aiming to evaluate occupational risks, it is important that the particle size distribution is similar to what occurs naturally. However, few particle size curves from real working environments have been published. A particle size distribution curve of the inhalable pinewood dust fraction from a sanding process in a Swedish joinery has been published (Lidén et al., 2000), which shows close similarity with the one found in the present chamber. Data on airborne wheat flour particle size in Swedish bakeries have been presented (Burdorf et al., 1994; Lillienberg and Brisman, 1994), but the authors interpreted the dust collected on the upper impactor plate in a different way, which makes a comparison with our data unsuitable. In a study from a UK bakery, 80–85% of the airborne particles were >14.8 µm (Sandiford et al., 1994), but since the dust was sampled by a Sierra Marple personal impactor, the data are not directly comparable with our result, either. However, it is clear that wheat flour aerosols in bakeries may include a substantial fraction of particles larger than what we have found in our exposure chamber. We know of no reports on particle size distribution for naturally occurring glove powder or cornstarch.
A complementary way to characterize the airborne particle size is to calculate the relationship between measurements with different ‘size-selective’ INH, OPFT and RESP dust samplers, which have differing collection efficiency depending on particle size. The INH and RESP dust sampler are well categorized (EN481, 1993), while the OPFT dust sampler air flow and filter size affects the particle sampling efficiency. The fraction between the concentration of INH and OPFT dust sampler is of particular interest, not only because the INH dust sampler is to replace the OPFT dust sampler in occupational exposure sampling, but also since the ratio gives an indication of the airborne particle size. A high fraction, i.e. similar values from the INH and OPFT dust samplers, shows that the majority of particles are inhalable, while a lower fraction indicates a larger particle size. Our data with 80–90% of the INH dust collected with the OPFT dust sampler indicates that the particle size distribution was not extremely large. Results from Swedish occupational environments (Lidén et al., 2000) using the same type of samplers show fractions of 40–60% for different types of organic aerosols (e.g. wheat flour, dust from textile manufactories, sanding of wood, dust from filter paper mills, starch dust in food industries), indicating a higher proportion of large particles than we found in our chamber. The ratio between the RESP and the INH or OPFT dust samplers indicate the presence of small particles. In a Scottish survey of bakery dust (HSE, 1999), an average of 4.7% of the inhalable dust was respirable (independently of working tasks), which is quite close to what we found for wheat flour 3 in our chamber. Also, Lillienberg and Brisman (1994) found a corresponding fraction of 9% during dough mixing. We know of no reports on such ratios in occupational environments for the other organic aerosols used in this study.
A wider and higher temporal and spatial variability in aerosol concentration might have been expected when exposing humans than the mannequin, since humans may make slight movements that influence the distribution of the aerosol. However, we found no difference in the temporal distribution between the mannequin and humans, and only a slightly lower temporal variation was observed when neither humans nor mannequin occupied the chamber. Further, our data show no difference between humans and the mannequin in spatial variability measured close to the breathing zone. This indicates that the turbulent air mixing is working satisfactory. Thus, we consider that use of a mannequin is appropriate in preparatory studies for human challenges in this chamber. However, spatial distribution variability increased slightly when evaluated in a larger section around the subject/mannequin compared with only two measuring points close to the breathing zone. This need to be considered for exposure studies when subjects are supposed to move around in the chamber performing a task.
In conclusion, the technical performance of this human whole-body exposure chamber confirms that it is appropriate for use in human provocation studies with these types of aerosols, giving low temporal and spatial distribution with limited effect of relative humidity and with airborne aerodynamic particle sizes resembling those in occupational environments. It is also interesting that this type of chamber set-up after a complete dismantling and rebuilding at a new location did not seem to change its performance characteristics as was indicated with wheat flour. Rigorous preparatory work is always needed when introducing a new compound in the chamber, and the use of a heated mannequin instead of humans facilitates such studies.
Received December 29, 2005; in final form March 9, 2006
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Burdorf A, Lillienberg L, Brisman J. (1994) Characterization of exposure to inhalable flour dust in Swedish bakeries. Ann Occup Hyg 38:67–78.
Byrne MA, Goddard AJH, Lange C, et al. (1995) Stable tracer aerosol deposition measurements in a test chamber. Ann Occup Hyg 26:645–53.
EN481. (1993) Workplace atmospheres—size fraction definitions for measurement of airborne particles. Brussels.
Fabries JF, Choudat D, Wrobel R, et al. (2000) Computerized equipment for the delivery of inhaled doses of solid particles in specific bronchial challenge. J Aerosol Med 13:1–10.[Web of Science][Medline]
Ghio AJ, Kim C, Devlin RB. (2000) Concentrated ambient air particles induce mild pulmonary inflammation in healthy human volunteers. Am J Respir Crit Care Med 162:981–8.
Gibson H, Vincent JH, Mark DA. (1987) Personal inspirable aerosol spectrometer for applications in occupational hygiene research. Ann Occup Hyg 31:463–79.
Gripenbäck S, Lundgren L, Eklund A, et al. (2003) Wheat flour exposure results in recruitment of inflammatory cells in the lungs of healthy individuals. Am J Ind Med 44:75–82.[CrossRef][Web of Science][Medline]
Gripenbäck S, Lundgren L, Eklund A, et al. (2005) Accumulation of eosinophils and T-lymphocytes in the lungs after exposure to pinewood dust. Eur Respir J 25:118–24.
Grunewald J, Eklund A, Katchar K, et al. (2003) Lung accumulations of eosinophil granulocytes after exposure to cornstarch glove powder. Eur Respir J 21:51.
HSE. (1999) Flour dust—Risk assessment document. (HSE Books, Sudbury, UK).
Lidén C, Lundgren L, Skare L, et al. (1998) A new whole-body exposure chamber for human skin and lung challenge experiments—the generation of wheat flour aerosols. Ann Occup Hyg 42:541–7.
Lidén G, Melin B, Lidblom A, et al. (2000) Personal sampling in parallel with open-face filter cassettes and IOM samplers for inhalable dust—implications for occupational exposure limits. Appl Occup Environ Hyg 15:263–76.[CrossRef][Medline]
Lillienberg L and Brisman J. (1994) Flour dust in bakeries—a comparison between methods. Ann Occup Hyg 38:571–75.
Lillienberg L, Baur X, Doekes G, et al. (2000) Comparison of four methods to assess fungal alpha-amylase in flour dust. Ann Occup Hyg 44:427–33.
Lundgren L, Skare L, Lidén C. (2006) Measuring dust on skin with a small vacuuming sampler—a comparison with other sampling techniques. Ann Occup Hyg 50:95–103.
O'Shaughnessy PT, Mehaffy J, Watt J, et al. (2004) Characterization of a hooded human exposure apparatus for inhalation of gases and aerosols. J Occup Environ Hyg 1:161–6.[CrossRef][Web of Science][Medline]
Sandiford CP, Nieuwenhuijsen MJ, Tee RD, et al. (1994) Determination of the size of airborne flour particles. Allergy 49:891–3.[Web of Science][Medline]
Søstrand P, Kongerud J, Eduard W, et al. (1997) A test chamber for experimental hydrogen fluoride exposure in humans. Am Ind Hyg Assoc J 58:521–5.[Web of Science][Medline]
Taylor L, Reist PC, Boehlecke BA, et al. (2000) Characterization of an aerosol chamber for human exposures to endotoxin. Appl Occup Environ Hyg 15:303–12.[CrossRef][Medline]
Wright BM. (1950) A new dust-feed mechanism. J Scient Inst Phys Ind 27:12–15.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  W. Eduard, K. Kruse, A. Skogstad, T. Nilsen, O. Synnes, J. Kongerud, and K. K. Heldal Generation and Homogeneity of Aerosols in a Human Whole-Body Inhalation Chamber Ann. Hyg., August 1, 2008; 52(6): 545 - 554. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||