Personal Exposure to Dust, Endotoxin and Crystalline Silica in California Agriculture

ABSTRACT

Aims: The aim of this study was to measure personal exposure to dust, endotoxin and crystalline silica during various agricultural operations in California over a period of one year.

Methods: Ten farms were randomly selected in Yolo and Solano counties and workers were invited to wear personal sampling equipment to measure inhalable and respirable dust levels during various operations. The samples were analysed for endotoxin using the Limulus Amebocyte Lysate assay and crystalline silica content using X-ray diffraction. In total 142 inhalable samples and 144 respirable samples were collected.

Results: The measurements showed considerable difference in exposure levels between various operations, in particular for the inhalable fraction of the dust and the endotoxin. Machine harvesting of tree crops (Geometric mean (GM) = 45.1 mg/m3) and vegetables (GM = 7.9 mg/m3), and cleaning of poultry houses (GM = 6.7 mg/m3) showed the highest inhalable dust levels. Cleaning of poultry houses also showed the highest inhalable endotoxin levels (GM = 1861 EU/m3). Respirable dust levels were generally low, except for machine harvesting of tree crops (GM = 2.8 mg/m3) and vegetables (GM = 0.9 mg/m3). Respirable endotoxin levels were also low. For the inhalable dust fraction, levels were reduced considerably when an enclosed cabin was present. The percentage of crystalline silica was overall higher in the respirable dust samples than the inhalable dust samples.

Conclusions: Considerable differences exist in personal exposure levels to dust, endotoxin and crystalline silica during various agricultural operations in California agriculture with some operations showing very high levels.

MATERIAL AND METHODS

Dust sampling

Seventeen farms in Solano and Yolo counties, situated in the Central Valley near the capital Sacramento of the State of California, US were randomly selected, stratified for commodities, from our farm operator cohort which is described elsewhere (Nieuwenhuijsen et al., 1996), and ten farms agreed to participate (Table 1). The farms were visited 8 times with periodic intervals over the period April 1995–June 1996, except for farms 1, 6, 8 and 10 which were visited 5, 7, 3 and 9 times respectively. Farm 8 went out of business during the study. Workers on the farms were asked to wear a personal sampler during various operations on the farms. This was either 1) an IOM inhalable dust sampler (SKC West, Fullerton, CA), containing a 25 mm diameter polyvinylchloride (PVC) filter (pore size 5 μm) and connected to a personal sampling pump which run at 2 litres/minute to measure the inhalable dust fraction; or 2) a personal respirable dust cyclone (BGI Inc, Waltham, MA), containing a 37 mm diameter PVC filter (pore size 5 μm), connected to a personal sampling pump run at 2.2 litres/minute (50% cut point diameter = 4 μm) to measure the respirable fraction. The filters were weighed before and after sampling on an six figure ATI Cahn C-35 microbalance (ATI Orion Cahn, Boston, MA), and a dust concentration calculated using the difference in filter weight, after adjustment for blanks, and sampling volume. The results were expressed as mg/m3. For the IOM inhalable dust sampler the detection limit was 0.030 mg per filter and for the cyclone the detection limit was 0.067 mg.

Endotoxin analysis

To measure the endotoxin concentration on the filters, dust was extracted individually in 10 ml of sterile, non pyrogenic water (LAL reagent water; BioWhittaker) in a 50 ml centrifuge tube by rocking at room temperature for 60 minutes (Labquake shaker; Labindustries, Berkeley, CA). The extracts were then centrifuged at 1000 g for 10 minutes, and dilutions of 1 ml of the supernatant fluids were analyzed in duplicate for endotoxins by the kinetic chromogenic modification of the Limulus Amebocyte Lysate (LAL) assay (Kinetic-QCL kit lot numbers 4L3010, 5L2180, 5L3560, 6L0440; Pyrogen-Free water lot numbers 4L2870, 5L1160, 5L3470, 5L4720; Kinetic-QCL reagent lot numbers 4L1760, 5L1780, 5L3030, 5L473G; Certified Standard Endotoxin 4L1370, 4L5090, 5L2110, standards were standardized to the common standard endotoxin, EC-5) (Kinetic-QCL; BioWhittaker). If the supernatant fluids could not be analyzed immediately, then they were frozen at −80°C. The procedure described in the commercial endotoxin kit is followed, and inhibition/enhancement assays were used as required. Results were reported in terms of Endotoxin Units (EU) per cubic meter of air. The limit of detection was 0.05 EU/ml.

Silica analysis

X-ray diffraction was used to measure the crystalline silica content of the dust. To get enough dust for analysis, samples, including the filters, with the same job codes were pooled and then they were placed in 50 ml centrifuge tubes for shipment from Davis to the analysis laboratory in Morgantown. Only pooled samples with more than 1 measurement, a combined sampling volume over 1000 litres or more, and a combined weight of 0.2 mg or more were used for analysis. The filters with water were dried in their containers and tetrahydrofuran (THF) was added to dissolve the filters. After dissolving the filters they were sonicated for 20 minutes and filtered onto tared silver membrane (FWS-B) filters. The dried silver membrane filters were then dried to determine the total particulate weight for each sample. Samples with greater than 3 mg particulate weight were dissolved again in THF, divided and filtered onto silver membrane filters to maintain a particle load of less than 3 mg.

The silver filters were analyzed for quartz and cristobalite using the NIOSH 7500-NMAM method (NIOSH Manual of Analytical Methods, Fourth Edition, 1994). Standards and samples were run concurrently and an external calibration curve was prepared from the integrated intensities rather than using the suggested normalization procedure. Using this technique the limits of detection for quartz and cristobalite were 0.01 mg and 0.02 mg respectively. The results were expressed as percentage crystalline silica in the dust.

Silica analysis

X-ray diffraction was used to measure the crystalline silica content of the dust. To get enough dust for analysis, samples, including the filters, with the same job codes were pooled and then they were placed in 50 ml centrifuge tubes for shipment from Davis to the analysis laboratory in Morgantown. Only pooled samples with more than 1 measurement, a combined sampling volume over 1000 litres or more, and a combined weight of 0.2 mg or more were used for analysis. The filters with water were dried in their containers and tetrahydrofuran (THF) was added to dissolve the filters. After dissolving the filters they were sonicated for 20 minutes and filtered onto tared silver membrane (FWS-B) filters. The dried silver membrane filters were then dried to determine the total particulate weight for each sample. Samples with greater than 3 mg particulate weight were dissolved again in THF, divided and filtered onto silver membrane filters to maintain a particle load of less than 3 mg.

The silver filters were analyzed for quartz and cristobalite using the NIOSH 7500-NMAM method (NIOSH Manual of Analytical Methods, Fourth Edition, 1994). Standards and samples were run concurrently and an external calibration curve was prepared from the integrated intensities rather than using the suggested normalization procedure. Using this technique the limits of detection for quartz and cristobalite were 0.01 mg and 0.02 mg respectively. The results were expressed as percentage crystalline silica in the dust.

RESULTS

Inhalable dust levels

The measurements showed considerable differences in personal exposure levels between various operations (Tables 2-5). The highest inhalable dust levels were measured for machine harvesting of both tree crops (nuts) and vegetables, and scraping poultry houses (Table 2). Lower but still considerable levels were measured for ground preparation operations, mechanical mowing of weeds and feeding poultry. Inhalable dust levels were lower during ground preparation and machine harvesting of dry harvested field crops when an enclosed cabin was present. An analysis of variance showed that operation alone explained 28% of the variation in inhalable dust exposure and that operation, the type of commodity and the presence of enclosed cabin together explained 60% of the variation in inhalable dust exposure.

Inhalable endotoxin levels

The highest average inhalable endotoxin levels were measured during cleaning of poultry houses (Table 3). Livestock related operations in general showed higher inhalable endotoxin levels than field crop, vegetable and fruit and nut related operations, although relatively high inhalable endotoxin levels were measured during machine harvesting of vegetables (tomatoes) and nuts, and mowing of weeds.

Respirable dust levels

Respirable dust levels were generally low, with the highest levels being measured for machine harvesting of tree crops (nuts) and vegetables (Table 4). A large percentage of the respirable measurements were below the limit of detection and the presented geometric means and geometric standard deviations should therefore be interpreted with caution. The respirable dust level during ground preparation was slightly lower when an enclosed cabin was present. An analyses of variance showed that operation alone explained 19% of the variation in inhalable dust exposure and that operation, the type of commodity and the presence of enclosed cabin together explained 41% of the variation in inhalable dust exposure.

Respirable endotoxin levels

The respirable endotoxin levels were generally low, and a large proportion were under the limit of detection (Table 5). The highest respirable endotoxin levels were measured during the machine harvest of vegetables (tomatoes).

Crystalline silica

The percentage of crystalline silica in the dust varied between operation and between the dust fractions (Table 6). Overall, the respirable dust fraction showed a higher percentage of crystalline silica in the dust than the inhalable dust fraction.
