Annals of Occupational Hygiene Advance Access originally published online on August 14, 2008
Annals of Occupational Hygiene 2008 52(8):685-694; doi:10.1093/annhyg/men052
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Variability and Determinants of Wood Dust and Resin Acid Exposure during Wood Pellet Production: Measurement Strategies and Bias in Assessing Exposure–Response Relationships
1 Department of Occupational and Environmental Medicine, Örebro University Hospital, 701 85 Örebro, Sweden
2 Department of Occupational and Environmental Medicine, University Hospital of Umeå, 901 85 Umeå, Sweden
3 Public Health and Clinical Medicine, Occupational Medicine, Umeå University, 901 87 Umeå, Sweden
* Author to whom correspondence should be addressed. Tel: +46-19-6022492; fax: +46-19-120404; e-mail: katja.hagstrom{at}orebroll.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Production of wood pellets is a relatively new and expanding industry in which the exposure profiles differ from those in other wood-processing industries like carpentries and sawmills where there are lower levels of wood dust. Sixty-eight personal exposure measurements of wood dust (inhalable and total dust) and resin acids were collected for 44 participants at four production plants located in Sweden. Results were used to estimate within- and between-worker variability and to identify uniformly exposed groups and determinants of exposure. In addition, overexposure, whether the risk of the long-term mean exposure of a randomly selected worker exceeding the occupational exposure limit is acceptably low, was calculated as well as the underestimation of the exposure–response relationship (attenuation). Greater variability in exposure between work shifts than between workers was observed with the within-worker variation accounting for 57–99% of the total variance in the individual-based model. Several uniformly exposed groups were detected but were mostly associated with a between-worker variation of zero which is an underestimation of the between-worker variation but an indication of uniformly exposed groups. Cleaning was identified as a work task that increases exposure slightly; so reducing workers' exposure during this operation is advisable. The levels of wood dust were high and were found to pose unacceptable risks of overexposure at all plants for inhalable dust and at three out of four plants for total dust. These findings show that exposure to dust needs to be reduced in this industry. For resin acids, the exposure was classed as acceptable at all plants. According to an individual-based model constructed from the data, the level of attenuation was high, and thus there would be substantial bias in derived dose–response relationships.
Keywords: attenuation • determinants • overexposure • resin acids • variability • wood dust
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Wood pellet production from compressed shavings and sawdust of pine (Pinus sylvestris) and spruce (Picea abies) wood is an expanding industry in various countries. Sweden, in which
2000 workers are employed in the industry, is the second biggest producer in the world (SVEBIO, 2004). This is a relatively new industry and the exposure profile at wood pellet production plants differs from that at other wood-processing sites, e.g. carpentries and sawmills, where there are generally lower levels of wood dust but higher levels of monoterpenes (Edman et al., 2003; Svedberg et al., 2004; Hagström et al., 2008). Thus, the associated health risks may differ too. Workers at wood pellet production plants are also exposed to resin acids (Hagström et al., 2008). Dust from pine and spruce wood has been reported to cause acute health effects, e.g. eye irritations, and chronic health effects, e.g. cancer and reductions in lung function (IARC, 1995; Eriksson and Liljelind, 2000). Furthermore, resin acids are the main components of colophony, a technical product that has been associated with contact dermatitis and chronic health effects such as asthma (Sadhra et al., 1994; Färm, 1996; Keira et al., 1997). It is well known that exposure levels to various substances can vary substantially within- and between workers (Rappaport, 1991). In order to assess this variability, repeated measurements are required and information regarding the variation can be used for measurement strategies, in risk management, in risk assessment and in epidemiological studies. Knowledge of within- and between-worker variability also enables people with similar exposure levels to be grouped in uniformly exposed groups (Rappaport et al., 1995), which is useful when formulating monitoring programs. In risk management so-called determinants of exposure, factors that increase or decrease the exposure, can be identified by using a mixed-effect analysis of variance model, taking into account correlations in repeated within-worker measurements. Such factors may include work environment characteristics (Peretz et al., 2002; Vermeulen et al., 2004), work practices (Rappaport et al., 1999; McClean et al., 2004; Blanco et al., 2005), specific work operations (Houba et al., 1997; van Tongeren and Gardiner, 2001), ventilation (Rappaport et al., 1999) and the type of material used in a process (Nylander-French et al., 1999; McClean et al., 2004). This knowledge can be used to identify changes that could be done for a better work environment.
To assess the risk associated with exposure, the probability of overexposure can be calculated. Overexposure is defined as the situation when the long-term mean exposure level of a randomly selected worker within a group exceeds a certain value, e.g. the occupational exposure limit (OEL), and it is tested to see if the risk of overexposure is acceptably low (Rappaport et al., 1995; Lyles and Kupper, 1996; Lyles et al., 1997; Tornero-Velez et al., 1997; Rappaport et al., 1999). In occupational epidemiology, some key tasks are to identify hazardous substances or workplaces and to characterize the effects of exposure to them on workers’ health. The inherent variability of exposure leads to tendencies to underestimate the strength of exposure–response relationships, i.e. attenuation, but degrees of attenuation can be estimated with knowledge of within- and between-worker variation (Kromhout and Heederik, 1995; Nieuwenhuijsen et al., 1995; Kromhout et al., 1996; Vinzents et al., 2001; Liljelind et al., 2003; Loomis and Kromhout, 2004).
Since high wood dust exposures during the production of wood pellets have been found (Edman et al., 2003), an extended project was initiated to address gaps in our knowledge concerning levels of exposure to wood dust and other agents in the industry and their health effects. The objectives of the project were to evaluate, in a series of studies air exposure to wood dust (expressed as total dust and inhalable dust), monoterpenes, resin acids, volitile organic compounds, CO and diesel exhaust (Hagström et al., 2008); skin exposure to resin acids (Eriksson et al., 2008); the variability in exposure between and within workers and both chronic and acute health effects related to the exposure. This article reports the findings concerning the variability and determinants of exposure to wood dust (inhalable and total dust) and resin acids during wood pellet production. In addition, the results were used to identify uniformly exposed groups, to determine the risk of overexposure and to assess underestimation of potential exposure–response relationships (attenuation) derived from both individual-based and group-based models. Data acquired concerning health effects related to the exposure will be presented in forthcoming papers.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Study design
The data analysed are from a study in which personal exposure to inhalable dust, total dust and resin acids [7-oxodehydroabietic acid (7OXO) and dehydroabietic acid (DHAA)] was monitored at four Swedish production plants, in which similar processes and tasks were performed, from 2003 to 2004 (Hagström et al., 2008). The operators, asked to participate voluntarily, were those working in shifts at pre-decided measurement dates.
The production process is described in detail elsewhere (Edman et al., 2003) but, briefly, the steps include drying, milling, compression, cooling and finally bagging or storage. Work tasks performed by the monitored workers included monitoring from the control room, cleaning (sweeping and vacuuming), cleaning with compressed air, maintenance, welding, repairs, bagging and truck driving. During the measurement periods, the participants were asked to record both the time and duration of each working task they undertook on a work record sheet. Determinants analysed were time spent cleaning, time cleaning with compressed air and work time in the control room, expressed as percentages of the monitored work time, obtained from the work record sheets. These tasks were chosen since they were easy for the workers to report and for the occupational hygienist to interpret from their work record sheets. However, these work operations only accounted for part of the workers’ workdays and varied between workers.
Inhalable dust and total dust were collected by sampling air at 2 l min–1 using Institute of Occupational Medicine (IOM) samplers and 25-mm (open-faced) samplers, respectively, with PVC-membrane filters (pore size: 5µm). The dust was gravimetrically determined (detection limit, 0.001 mg). Each total-dust sampler was stored in an airtight container at –20°C until extraction (with 3 ml methanol). The resin acids were then analysed by high performance liquid chromatography with electrospray ionisation and mass spectrometry as described by Hagström et al. (2008).
Statistical analysis
The geometric mean (GM) and geometric standard deviation (GSD) were calculated as descriptive statistics. For measurements under the limit of quantification (LOQ), the levels were recorded as LOQ/
2 if the GSD was below three and otherwise as LOQ/2 (Hornung and Reed, 1990). The GSDs of inhalable dust and total dust were both 2.6, while those of 7OXO and DHAA were 4.4 and 3.4, respectively. Differences in exposure between plants were tested using a mixed model with worker as random factor nested within plant for inhalable dust, total dust and resin acids.
Estimation of variance components.
Within- and between-worker variance components were estimated by a one-way random effect analysis of variance model, with restricted maximum likelihood estimation and with worker as random factor (Rappaport, 1991; Rappaport et al., 1993). Thus, if we denote the exposure of the ith worker (i = 1, ..., N) at the jth measurement occasion (j = 1, ..., n) Xij, our model for the exposure measurement is
 | (1) |
ij is the random error of the jth measurement of the ith worker. The variance of βi is the between-worker variance, 
, and the random error variance (ij) is the within-worker variance, 
. The total variability in model (1) is 
.
The model was expanded to a two-way random effects analysis of variance model in order to estimate the variance between production plants, a between-group variability, with workers nested within production plant. Thus, the model for the exposure of the hth group (h = 1, ..., g), the ith worker (i = 1, ..., N) at the jth measurement occasion (j = 1, ..., n) Xhij was
 | (2) |
h is the random effect of the hth group (production plant), which is assumed to be normally distributed with variance 
, the between-group variation. The total variability in this model is 
.
Uniformly exposed groups.
To characterize between-worker variance components, fold ranges can be used. The 95% fold range is the ratio of the 97.5th percentile to the 2.5th percentile of the log-normal distribution of the random effect of workers, βi. The between-worker fold range is calculated from the variance estimates according to the formula given by Rappaport (1991):
 | (3) |
A between-worker fold range of two suggests that 95% of the mean exposure of workers within a group lies within a 2-fold range, and a between-worker fold range of 
2 
has been defined as the criterion for a uniformly exposed group (Rappaport, 1991; Rappaport et al., 1993).
Identification of determinants.
The effects of determinants on the exposure level were estimated using the mixed-effect model
 | (4) |
1, ..., m corresponding to the fixed effects of the determinants, R1j, ..., Rmj. The determinants of interest here are the proportions of the working day the workers’ spent in the control room, time cleaning and part of the day cleaning with compressed air. Thus, each determinant has a value between 0 and 100%.
In the same manner, model (2) was expanded to include fixed effects for the determinants, while accounting for the correlation between workers within the same production plant:
 | (5) |
Determinants are sometimes used as a basis for forming uniformly exposed groups, testing overexposure given a certain level of the determinant or to decrease attenuation by adjusting for the determinants. However, a prerequisite for such calculations is that the workers can be assigned a particular value of the determinant. In our study, the levels varied from day to day and consequently that could not be done. Therefore, the main use of the determinants in this study is as explanatory factors for exposure.
Overexposure.
A rough estimate of the probability of the long-term mean exposure exceeding the OEL for an individual at plant j was calculated as
 | (6) |
{x} is the probability of a standard normal variable having a value below x and 
BW,j is the estimated between-worker variance at plant j (Lyles et al., 1997).
It was also tested if the 
j was <10% using the test procedure proposed by Weaver et al. (2001) using the hypotheses as follows:
 |
The first step in the procedure is to test if the within-worker variances can be pooled, using a likelihood ratio test. The main test is then based on a test statistic, denoted 
, which can be constructed in two ways depending on whether or not 
. In both cases, is compared to a critical value, C (level of significance = 0.05). The exposure level is considered unacceptable unless H0 can be rejected, i.e. if < C (indicating acceptable exposure level). Details on the test procedure and the calculation of and C are given by Weaver et al. (2001).
The levels used for comparison were the Swedish OEL for wood dust as inhalable dust, 2 mg m–3 (AFS, 2005); the old Swedish OEL for wood dust as total dust, 2 mg m–3 (AFS, 2000) and the British OEL for rosin-based solder flux fumes, 50 µg m–3 (COSHH, 2005), for comparison to the detected levels of resin acids since rosin-based solder flux fumes consist of resin acids.
Attenuation.
Due to the known within-worker (Rappaport, 1991) and random variation in exposure, there is an underestimation of the regression coefficient in an exposure–effect relationship, i.e. an attenuation (Kromhout and Heederik, 1995; Nieuwenhuijsen et al., 1995; Kromhout et al., 1996; Vinzents et al., 2001). The attenuation is related to the ratio of the within-worker and between-worker variation [
here, according to model (1)] and the number of repeated measurements per worker (n) as follows (Cochran, 1968; Kromhout et al., 1996):
 | (7) |
is the estimated regression coefficient and B the true regression coefficient. The attenuation is expressed as one minus the ratio between the estimated exposure effect and the true effect, consequently it can span between zero and one, since < B, where values close to zero indicate low bias and values close to one high bias. The attenuation was also estimated using a group-based approach and applying the equation (Tielemans et al., 1998).
 | (8) |
were also used to estimate the numbers of measurements needed for an attenuation of 10%, with k as 2, 5 and 9 in equation (7). All variance components and regression coefficients obtained in this study were estimated using PROC MIXED in SAS Release 9.1.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The maximum level of exposure of inhalable dust was 12 mg m–3, and 24 of the 68 individual measurements exceeded the Swedish OEL of 2 mg m–3 (Table 1). For total dust, the GM was 0.62 mg m–3 and nine of the samples exceeded the old OEL of 2 mg m–3. The GM for resin acids was 1.1 µg m–3 (Table 1). Statistically significant differences were seen between the plants in levels of inhalable dust (P = 0.0001), total dust (P = 0.0001) and resin acids (P = 0.01). Working operations recorded in the work diaries included (inter alia) monitoring from the control room, welding, maintenance, repairs, loading raw material and wood pellets, bagging, sweeping and cleaning with compressed air. On average, the participants spent 6% (0–50%) of the day cleaning, 0.7% (0–26%) of the time cleaning with compressed air and 15% (0–87%) of the work day in the control room.
|
The within-worker variance accounted for 57, 91 and 99% of the total variance (
) of exposure to inhalable dust, total dust and resin acids, respectively, according to model (1), see Table 2. The between-worker variance estimates were negative for total dust and the resin acids when between-group variance was also accounted for, in model (2). The between-group variance estimates accounted for 34, 30 and 15% of the total variance (
) for inhalable dust, total dust and resin acids, respectively, and the corresponding values for within-worker variance (
) were 50, 70 and 85% (Table 2). For the individual plants, the within-worker variance also accounted for most of the variance for inhalable dust, total dust and resin acids (51–100, 67–100 and 83–100%, respectively; Table 1). Several uniformly exposed groups (R0.95, BW 2) were identified: one comprising the whole group with respect to resin acids (Table 2), one consisting of the workers in one plant with respect to inhalable dust and two consisting of workers in two plants with respect to total dust and resin acids (Table 1).
|
All determinants were tested simultaneously in model (4), and the regression coefficients were found to be statistically significant (P
0.05) for all determinants for exposure to inhalable dust. Cleaning and work in the control room were significant factors for total dust, while only cleaning was significant for resin acids (Table 3). After adjusting for the average exposure level at each production plant [model (5)], significant determinants were the proportions of the day spent cleaning and time cleaning with compressed air for inhalable dust, while for total dust and resin acids only the amount of time spent cleaning was significant (Table 3). Cleaning and cleaning with compressed air were commonly positively correlated with exposure, while time spent in the control room was negatively correlated.
|
The likelihood ratio test of equal within-worker variance between production plants yielded non-statistically significant results for exposure to all examined substances (data not shown), and consequently data acquired at the different plants were pooled, in accordance with the test procedure of Weaver et al. (2001) (Table 1). The tests of the probability of overexposure showed that it could not be inferred to be <10% at any of the four plants for inhalable dust, and thus exposure to this substance was classified as unacceptable (Table 1). For total dust, the exposure was classed as unacceptable at plants 1, 2 and 4 and acceptable in plant 3, while the exposure to resin acids was found to be acceptable at all plants.
Based on the estimated variance components, the attenuation according to the individual-based model varied between 40 and 98% with two repeated measurements (equation 7). The numbers of measurements needed per participant to obtain an attenuation of 10% varied from 12 to 1600 (Table 2). For the group-based model, attenuation could only be estimated for inhalable dust and was 7% for two repeated measurements (equation 8). The numbers of measurements needed for an attenuation of 10% were six, three and two with two, five and nine participants per group, respectively (Table 2).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The results show that levels of wood dust in the examined plants were high; 35% of the inhalable dust measurements exceeded the Swedish OEL and 15% of the total dust measurements exceeded 2 mg m–3, the old OEL for wood dust in Sweden. Both inhalable dust and total dust were included in the analysis since we wanted to assess the difference, if any, in results from the calculations for the two different dust fractions.
Within-worker variability is often associated with factors linked to the facility and organization of the work, while between-worker variability can be linked to individuals’ work practices as well as differences between plants. Our results show that there were larger variations in exposure between work shifts than between workers, indicating that the work practices of the individual workers affect the variability less than the differences in exposure between days in the investigated wood pellet production plants. Within-worker variability has often been found to be greater than between-worker variability in various work environments (Kromhout et al., 1993; Nieuwenhuijsen et al., 1995; Heederik and Attfield, 2000; Burdorf and Van Tongeren, 2003; Symanski et al., 2006). However, higher between-worker variability was observed in 17 of the 27 studies of exposure to particulates listed in Table 4, although it should be noted that this list does not include all of the literature related to this research field. By using models with a production plant effect, we have also shown that a fairly large proportion of the variation can be attributed to plant-specific factors, so it is important to take plant effects into account in the calculations.
|
Large within-worker variation has also been seen in mobile groups working outdoors (Peretz et al., 1997; Burdorf and Van Tongeren, 2003), in groups working on intermittent processes, groups working without local exhaust ventilation (Burdorf and Van Tongeren, 2003) and those working in environments where there are mobile sources of exposure (Peretz et al., 1997). The tasks involved in the production of wood pellets meet all of these criteria. The possibility has also been raised that measurements on consecutive days may lead to autocorrelation, which in this case would reduce between-worker variability. However, relatively weak autocorrelation was observed in the cited studies (Francis et al., 1989; Kumagai et al., 1993; Symanski and Rappaport, 1994).
Our estimates of between-worker variance were zero for exposure to total dust at plant 1, exposure to resin acids at plants 2 and 3 and exposure to both inhalable dust and total dust at plant 4 (Table 1) as well as for total dust and resin acids in model (2), see Table 2. This may occur when the within-worker variability is much larger than the between-worker variability and/or the sample size is small (Brown and Prescott, 1999). Both of these criteria apply to our plant-specific datasets, and thus we have an underestimation of those between-worker variances. The between-worker variation was much larger for inhalable dust than total dust, while the within-worker variance was larger for total dust, resulting in higher attenuation for total dust than for inhalable dust. This difference in variation distributions is probably related to findings that levels of inhalable dust were 3.2-fold higher (range: 0.67–17 times higher) than the levels of total dust in the plants (Hagström et al., 2008), in accordance with results from other studies (Davies et al., 1999; Lidén et al., 2000; Tatum et al., 2001; Harper and Muhler, 2002). These indicate that it is important to measure the right dust fraction since there are clearly differences between them.
The total group of workers could be defined as a uniformly exposed group with respect to resin acids (Table 2), but in further analyses workers at only two of the plants could be defined as uniformly exposed groups (Table 1). For inhalable dust, the workers at only one of the plants formed a uniformly exposed group and for total dust the workers at two of the plants each formed uniformly exposed groups. As mentioned above, the between-worker variation was estimated to zero for all five of these defined uniformly exposed groups at individual plants. As also mentioned, this is almost certainly partly due to underestimation of the between-worker variation, but it does at least provide a strong indication that between-worker variation in exposure to the respective substances is low at the respective plants, and thus the designation of uniformly exposed groups is justified. For the other groups, the between-worker fold range varied between 3.0 and 13, indicating that further grouping would be needed to obtain uniformly exposed groups for the substances at these plants if further sampling is to be done from uniformly exposed groups in this industry (Table 1). However, this would not be straightforward since there are few workers at each plant, and even if they have nominally specific work titles the production processes are highly dynamic and the work done by the operators varies, depending on the plants’ needs at specific times. Thus, grouping may not be practical in this industry and workers may need to be individually monitored in order to obtain information regarding their personal exposure profiles.
Time cleaning and time cleaning with compressed air were positively correlated to exposure, in accordance with the findings of a real-time exposure monitoring study indicating that these tasks are associated with high exposure (Edman et al., 2003). Conversely, time spent in the control room was negatively correlated with exposure, in accordance with area measurements indicating that there were low levels (0.18 mg m–3) of total dust in the control rooms (Edman et al., 2003; Hagström et al., 2008). Measures that could be beneficial include cleaning with central vacuum cleaners instead of sweeping and cleaning with compressed air. Exposures would also probably be reduced if sawdust was automatically transported from storage areas to production stations instead of being moved on manually loaded trucks. This would also lead to more work time being spent in the control room.
Compliance testing, in occupational exposure contexts, assesses whether individual measurements exceed relevant OELs. However, when sample sizes are small and there is a high risk of measurements exceeding the OEL, as in this study, since 35% of the individual measurements were above the OEL and the sample sizes were quite small, it has been shown that compliance testing can underestimate chronic health risks (Tornero-Velez et al., 1997). In contrast, chronic health effect evaluations are more concerned with long-term, cumulative exposures (Rappaport, 1991; Rappaport et al., 1995; Tornero-Velez et al., 1997), and in the context of this study the most suitable indicators of unacceptable exposure levels were deemed to be assessments of the risk of overexposure.
The results of this study show that levels of exposure to inhalable dust were not acceptable at any of the plants, yielding estimated overexposure probabilities of 13–97%, and the highest probabilities were for plant 1 (Table 1). However, it should be noted that the estimates of the probability of overexposure are biased and should only be regarded as rough indications, while the test of the probabilities is more reliable (Lyles et al., 1997). For total dust, the exposure was only acceptable at plant 3. The results show that there is a need to reduce dust exposure in this industry to protect the workers’ health. For resin acids, the exposure was acceptable at all four plants in comparison to the British OEL for rosin-based solder flux fumes. However, since the overexposure is calculated by comparing detected levels to a given OEL, the results would be altered if the OEL was changed. Ideally, OELs should comfortably be below levels at which there are significant health risks for workers. But this may not always be the case since economic factors can be taken into consideration when setting an OEL as in Sweden (AFS, 2005). Clearly, if an OEL is set too high, there may be significant implications for workers’ health, even if no measurements exceed it.
Higher levels of wood dust and lower levels of monoterpenes have been seen in this industry than in other wood-processing industries, warranting investigations of the worker's health risks, since they could differ from those in other wood-handling industries. The individual-based model we obtained indicated that the level of attenuation was high, implying that exposure–response relationships derived from the data would be subject to substantial bias, complicating attempts to draw definitive conclusions in an epidemiological study. The numbers of repeated measurements needed to reduce the attenuation to 10% were high, ranging for the whole-study group from 12 measurements for inhalable dust, through 88 for total dust to 1600 measurements for resin acids (equation 7). The numbers of measurements needed to obtain an attenuation of 10% in previous studies, on particulates in other industries, has varied greatly (from 2 to 42 measurements), but generally they have been much lower, and only measurements of inhalable dust have fallen within this range (Table 4).
Attenuation can be decreased by maximizing the differences between workers’ exposure levels in relation to the within-worker variation (Liu et al., 1978; Kromhout and Heederik, 1995) or by using a grouping strategy rather than an individual-based strategy (Nieuwenhuijsen et al., 1995; Kromhout et al., 1996; Schlünssen et al., 2004; Teschke et al., 2004; Tjoe Nij et al., 2004; Mwaiselage et al., 2005), as seen in this study, for inhalable dust, where the attenuation decreased from 40 to 7% with two repeated measurements per person. However, a group-based strategy will give less precision in estimates of dose–response relationships (Kromhout et al., 1996), and thus require a larger number of groups. Taking measurements for a large number of groups to increase the precision would have to be weighed against taking a large number of repeated measurements for each individual in order to reduce the bias to an acceptable level.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The levels of wood dust were high and considered to be unacceptable at all plants for inhalable dust and at three out of four plants for total dust, indicating that the wood dust exposure should be reduced at these plants to protect the workers’ health. For resin acids, the exposure was classed as acceptable. Greater within-worker variation than between-worker was seen indicating that the work practice between individuals affects the variability less than the difference in exposure between days. Time cleaning was identified as a determinant that increases the exposure for all measured substances while more work time in the control room seemed to lower the exposure. The individual-based model derived from the data indicated that there was a high level of attenuation.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Swedish Council for Working Life and Social Research (2003–0084); Departments of Occupational and Environmental Medicine of Örebro University Hospital and Departments of Occupational and Environmental Medicine of University Hospital of Northern Sweden in Umeå.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
We would first like to thank all the production plants and workers involved for participating in the study. We would also like to thank Helena Arvidsson, Sara Axelsson, Krister Berg, Ing-Liss Bryngelsson, Britt-Marie Isaksson, Carin Norberg and Mona Svensson for skilful assistance with the fieldwork, laboratory analysis and statistical calculations.
Received December 19, 2007; in final form July 2, 2008
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
AFS. Occupational exposure limits and measures against air contaminants (2000) Solna, Sweden: Swedish Work Environment Authority.
AFS. Occupational exposure limit values and measures against air contaminants Swedish Work Environment Authority (2005) Available at http://www.av.se/dokument/inenglish/legislations/eng0517.pdf. Accessed 10 June 2008.
Blanco LE, Aragon A, Lundberg I, et al. Determinants of dermal exposure among Nicaraguan subsistence farmers during pesticide applications with backpack sprayers. Ann Occup Hyg (2005) 49:17–24.
Brown H, Prescott R. Applied mixed models in medicine (1999) Chichester, UK: John Wiley & Sons Ltd.
Burdorf A, Van Tongeren M. Commentary: variability in workplace exposures and the design of efficient measurement and control strategies. Ann Occup Hyg (2003) 47:95–9.
Burstyn I, Kromhout H. Are the members of a paving crew uniformly exposed to bitumen fume, organic vapor, and benzo(a)pyrene? Risk Anal (2000) 20:653–63.[CrossRef][Web of Science][Medline]
Cochran WG. Errors of measurements in statistics. Technometrics (1968) 60:637–66.
COSHH. Workplace exposure limits (EH40/2005) [homepage on the Internet]. (2005) Health and Safety Executive, UK. Available at http://www.hse.gov.uk/coshh/table1.pdf. Accessed 3 June 2007.
Davies HW, Teschke K, Demers PA. A field comparison of inhalable and thoracic size selective sampling techniques. Ann Occup Hyg (1999) 43:381–92.
Edman K, Löfstedt H, Berg P, et al. Exposure assessment to alpha- and beta-pinene, delta(3)-carene and wood dust in industrial production of wood pellets. Ann Occup Hyg (2003) 47:219–26.
Eriksson K, Liljelind I. Consensus report for wood dust. In: Scientific basis for Swedish occupational standards XXI—Montelius J, ed. (2000) Stockholm, Sweden: National Institute of Working Life. 51–71.
Eriksson K, Hagström K, Axelsson S, et al. Tape-stripping as a method for measuring dermal exposure to resin acids during wood pellet production. J Environ Monit (2008) 345–52.
Färm G. Contact allergy to colophony and hand eczema. A follow-up study of patients with previously diagnosed contact allergy to colophony. Contact Dermatitis (1996) 34:93–100.[CrossRef][Web of Science][Medline]
Francis M, Selvin S, Spear R, et al. The effect of autocorrelation on the estimation of workers’ daily exposures. Am Ind Hyg Assoc J (1989) 50:37–43.[Web of Science][Medline]
Hagström K, Axelsson S, Arvidsson H, et al. Exposure to wood dust, resin acids and volatile organic compounds during production of wood pellets. J Occup Environ Hyg (2008) 5:296–304.[CrossRef][Web of Science][Medline]
Harper M, Muhler BS. An evaluation of total and inhalable samplers for the collection of wood dust in three wood products industries. J Environ Monit (2002) 4:648–56.[CrossRef][Web of Science][Medline]
Heederik D, Attfield M. Characterization of dust exposure for the study of chronic occupational lung disease: a comparison of different exposure assessment strategies. Am J Epidemiol (2000) 151:982–90.
Hornung RW, Reed LD. Estimation of average concentration in the presence of nondetectable values. Appl Occup Environ Hyg (1990) 5:46–51.
Houba R, Heederik D, Kromhout H. Grouping strategies for exposure to inhalable dust, wheat allergens and alpha-amylase allergens in bakeries. Ann Occup Hyg (1997) 41:287–96.
IARC. Wood dust and formaldehyde (1995) Lyon, France: International Agency for Research on Cancer.
Keira T, Aizawa Y, Karube H, et al. Adverse effects of colophony. Ind Health (1997) 35:1–7.[CrossRef][Web of Science][Medline]
Kromhout H, Symanski E, Rappaport SM. A comprehensive evaluation of within and between-worker components of occupational exposure to chemical agents. Ann Occup Hyg (1993) 37:253–70.
Kromhout H, Heederik D. Occupational epidemiology in the rubber industry: implications of exposure variability. Am J Ind Med (1995) 27:171–85.[Web of Science][Medline]
Kromhout H, Tielemans E, Preller L, et al. Estimates of individual dose from current measurements of exposure. Occup Hyg (1996) 3:23–39.
Kumagai S, Matsunaga I, Kusaka Y. Autocorrelation of short-term and daily average exposure levels in workplaces. Am Ind Hyg Assoc J (1993) 54:341–50.[Web of Science][Medline]
Lidén G, Melin B, Lidblom A, et al. Personal sampling in parallel with open-face filter cassettes and IOM samplers for inhalable dust-implications for occupational exposure limits. Appl Occup Environ Hyg (2000) 15:263–76.[CrossRef][Medline]
Liljelind I, Rappaport S, Eriksson K, et al. Exposure assessment of monoterpenes and styrene: a comparison of air sampling and biomonitoring. Occup Environ Med (2003) 60:599–603.
Liu K, Stamler J, Dyer A, et al. Statistical methods to assess and minimize the role of intra-individual variability in obscuring the relationship between dietary lipids and serum cholesterol. J Chronic Dis (1978) 31:399–418.[CrossRef][Web of Science][Medline]
Loomis D, Kromhout H. Exposure variability: concepts and applications in occupational epidemiology. Am J Ind Med (2004) 45:113–22.[CrossRef][Web of Science][Medline]
Lyles RH, Kupper LL. On strategies for comparing occupational exposure data to limits. Am Ind Hyg Assoc J (1996) 57:6–15.[Web of Science][Medline]
Lyles RH, Kupper LL, Rappaport SM. A lognormal distribution-based exposure assessment method for unbalanced data. Ann Occup Hyg (1997) 41:63–76.
McClean MD, Rinehart RD, Ngo L, et al. Inhalation and dermal exposure among asphalt paving workers. Ann Occup Hyg (2004) 48:663–71.
Mwaiselage J, Bratveit M, Moen B, et al. Variability in dust exposure in a cement factory in Tanzania. Ann Occup Hyg (2005) 49:511–9.
Nieuwenhuijsen MJ, Lowson D, Venables KM, et al. Flour dust exposure variability in flour mills and bakeries. Ann Occup Hyg (1995) 39:299–305.
Nylander-French LA, Kupper LL, Rappaport SM. An investigation of factors contributing to styrene and styrene-7,8-oxide exposures in the reinforced-plastics industry. Ann Occup Hyg (1999) 43:99–105.
Peretz C, Goldberg P, Kahan E, et al. The variability of exposure over time: a prospective longitudinal study. Ann Occup Hyg (1997) 41:485–500.
Peretz C, Goren A, Smid T, et al. Application of mixed-effects models for exposure assessment. Ann Occup Hyg (2002) 46:69–77.
Preller L, Kromhout H, Heederik D, et al. Modeling long-term average exposure in occupational exposure-response analysis. Scand J Work Environ Health (1995) 21:504–12.[Web of Science][Medline]
Rappaport SM. Assessment of long-term exposures to toxic substances in air. Ann Occup Hyg (1991) 35:61–121.
Rappaport SM, Kromhout H, Symanski E. Variation of exposure between workers in homogeneous exposure groups. Am Ind Hyg Assoc J (1993) 54:654–62.[Web of Science][Medline]
Rappaport SM, Lyles RH, Kupper LL. An exposure-assessments strategy accounting for within- and between-worker sources of variability. Ann Occup Hyg (1995) 39:469–95.
Rappaport SM, Weaver M, Taylor D, et al. Application of mixed models to assess exposures monitored by construction workers during hot processes. Ann Occup Hyg (1999) 43:457–69.
Rappaport SM, Goldberg M, Susi P, et al. Excessive exposure to silica in the US construction industry. Ann Occup Hyg (2003) 47:111–22.
Sadhra S, Foulds IS, Gray CN, et al. Colophony-uses, health effects, airborne measurement and analysis. Ann Occup Hyg (1994) 38:385–96.
Scheeper B, Kromhout H, Boleij JS. Wood-dust exposure during wood-working processes. Ann Occup Hyg (1995) 39:141–54.
Schlünssen V, Sigsgaard T, Schaumburg I, et al. Cross-shift changes in FEV1 in relation to wood dust exposure: the implications of different exposure assessment methods. Occup Environ Med (2004) 61:824–30.
Svedberg UR, Högberg HE, Högberg J, et al. Emission of hexanal and carbon monoxide from storage of wood pellets, a potential occupational and domestic health hazard. Ann Occup Hyg (2004) 48:339–49.
SVEBIO. Bioenergy: small-scale use of biofuels (in Swedish). (2004) Available at: http://svebio.agriprim.com/attachments/33/104.pdf. Accessed 11 August 2008.
Symanski E, Rappaport SM. An investigation of the dependence of exposure variability on the interval between measurements. Ann Occup Hyg (1994) 38:361–72.
Symanski E, Sällsten G, Barregård L. Variability in airborne and biological measures of exposure to mercury in the chloralkali industry: implications for epidemiologic studies. Environ Health Perspect (2000) 108:569–73.[Web of Science][Medline]
Symanski E, Maberti S, Chan W. A meta-analytic approach for characterizing the within-worker and between-worker sources of variation in occupational exposure. Ann Occup Hyg (2006) 50:343–57.
Tatum VL, Ray AE, Rovell-Rixx DC. The performance of personal inhalable dust samplers in wood-products industry facilities. Appl Occup Environ Hyg (2001) 16:763–9.[Medline]
Teschke K, Spierings J, Marion SA, et al. Reducing attenuation in exposure-response relationships by exposure modeling and grouping: the relationship between wood dust exposure and lung function. Am J Ind Med (2004) 46:663–7.[CrossRef][Web of Science][Medline]
Tielemans E, Kupper LL, Kromhout H, et al. Individual-based and group-based occupational exposure assessment: some equations to evaluate different strategies. Ann Occup Hyg (1998) 42:115–9.
Tjoe Nij E, Hohr D, Borm P, et al. Variability in quartz exposure in the construction industry: implications for assessing exposure-response relations. J Occup Environ Hyg (2004) 1:191–8.[CrossRef][Web of Science][Medline]
Tornero-Velez R, Symanski E, Kromhout H, et al. Compliance versus risk in assessing occupational exposures. Risk Anal (1997) 17:279–92.[CrossRef][Web of Science][Medline]
van Tongeren M, Gardiner K, Calvert I, et al. Efficiency of different grouping schemes for dust exposure in the European carbon black respiratory morbidity study. Occup Environ Med (1997) 54:714–9.
van Tongeren MJ, Kromhout H, Gardiner K. Trends in levels of inhalable dust exposure, exceedance and overexposure in the European carbon black manufacturing industry. Ann Occup Hyg (2000) 44:271–80.
van Tongeren MJ, Gardiner K. Determinants of inhalable dust exposure in the European carbon black manufacturing industry. Appl Occup Environ Hyg (2001) 16:237–45.[CrossRef][Medline]
Vermeulen R, Li G, Lan Q, et al. Detailed exposure assessment for a molecular epidemiology study of benzene in two shoe factories in China. Ann Occup Hyg (2004) 48:105–16.
Vinzents PS, Schlünssen V, Feveile H, et al. Variations in exposure to inhalable wood dust in the Danish furniture industry. Within- and between-worker and factory components estimated from passive dust sampling. Ann Occup Hyg (2001) 45:603–8.
Weaver MA, Kupper LL, Taylor D, et al. Simultaneous assessment of occupational exposures from multiple worker groups. Ann Occup Hyg (2001) 45:525–42.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||