Ann. occup. Hyg., Vol. 48, No. 2, pp. 105-116, 2004
© 2004 British Occupational Hygiene Society
Published by Oxford University Press
Detailed Exposure Assessment for a Molecular Epidemiology Study of Benzene in Two Shoe Factories in China
1 Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS, 6120 Executive Boulevard, Rockville, MD 20892, USA; 2 National Institute of Occupational Health and Poison Control, China CDC, Beijing, China; 3 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; 4 School of Public Health, University of California, Berkeley, CA, USA; 5 Radiation Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS, Rockville, MD, USA; 6 Tianjin Center for Health and Diseases Prevention, Tianjin, China
Received 20 May 2003; in final form 14 August 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objectives: We carried out a detailed exposure assessment of benzene and toluene in two shoe factories in Tianjin, China. Our goal was to identify workers with a broad range of benzene exposures, for an epidemiologic study relating exposure to early biologic effects. Methods: A comprehensive exposure survey program was initiated. Over a period of 16 months, 2783 personal solvent exposure samples were collected in two workplaces from 250 workers. Mixed-effects models were used to identify factors affecting exposure. Principal component analyses (PCA) and subsequent regression analyses on the scores of the identified principal components were used to relate potential co-exposures to various exposure sources present in the workplace. Results: The mean benzene exposure level was 21.86 p.p.m. (10th–90th percentiles 5.23–50.63 p.p.m.) in the smaller shoe factory (factory A) and 3.46 p.p.m. (10th–90th percentiles 0.20–7.00 p.p.m.) in the larger shoe factory (factory B). Within-factory exposure levels differed among job titles and were higher for subjects directly involved in handling glues. In contrast, mean toluene levels were relatively similar in the two factories (factory A, 9.52 p.p.m.; factory B, 15.88 p.p.m.). A seasonal trend was identified for both benzene and toluene in factory B. This could be explained in part by changes in air movement and ventilation patterns occurring during the year. A seasonal trend was not present in the smaller shoe factory, where general ventilation was absent. Supplemental analysis showed that exposure levels to other hydrocarbons were low (
5 p.p.m.), less than 5% of their respective ACGIH threshold limit values, and generally comparable in the two factories. PCA showed that co-exposures in factory B could largely be explained by glue sources that were used in distinct areas in the workplace. Conclusions: We demonstrated the occurrence of a broad range of benzene exposure levels in two shoe manufacturing factories in Tianjin, China. Benzene and toluene exposures were determined in part by the degree of contact with glues, the benzene and toluene content of each glue, air movement and ventilation patterns. The availability of long-term monthly personal monitoring data provides an excellent opportunity to estimate individual exposures at different times during the 1 yr period of observation.
Keywords: benzene; China; exposure assessment; glues; molecular epidemiology; principal component analysis; shoe manufacturing; toluene
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
By 1981, the evidence linking benzene exposure with increased risk of acute non-lymphocytic leukemia (ANLL) was considered sufficient to classify benzene as a Group I carcinogen (IARC, 1982). Subsequent studies have confirmed that benzene causes ANLL (Aksoy, 1985; McCraw et al., 1985; Bond et al., 1986; Rinsky et al., 1987; Wong, 1987; Yin et al., 1987, 1996; Austin et al., 1988; Swaen and Meijers, 1989) and myelodysplastic syndrome (MDS) (Hayes et al., 1997) and may play a role in other leukemias, non-Hodgkin’s lymphoma and lung cancer (Savitz and Andrews, 1997). Despite extensive research, questions remain regarding the mechanisms by which benzene exerts its effects, the influence of interindividual variation in benzene metabolism and the risk benzene presents at low doses.
In a retrospective cohort study of 74 828 workers exposed to benzene and 35 805 unexposed workers in China, Hayes et al. (1997) reported an excess risk of ANLL and MDS at occupational exposures below 10 p.p.m., providing some of the strongest evidence to date that benzene poses health risks in this exposure range. In addition, we previously documented hematotoxicity (Rothman et al., 1996) and elevated levels of specific classes of chromosomal aberrations that are relevant for hematological malignancies (Smith et al., 1998; Zhang et al., 1998) in workers exposed to a wide range of benzene concentrations in Shanghai, China. To follow up these findings, we conducted another molecular epidemiology study to evaluate hematological, cytogenetic and molecular end-points in workers exposed to benzene, with a particular focus on exposures below 10 p.p.m.. As hematological and other end-points in this study are related to exposures in the more recent past (weeks to months) (Rothman et al., 1996), the exposure survey was primarily designed to assess individual exposures to benzene and toluene over a 16 month period of observation.
This paper describes the two Chinese factories selected for the molecular epidemiology investigation and presents results of the extensive monitoring survey to characterize personal exposures to benzene, toluene and other organic hydrocarbons at work and in the workers’ homes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Factories and subjects
Our goal was to study at least 200 workers exposed to benzene, particularly at levels below 10 p.p.m. Eligible factories had to be in the same industry and geographical area and not to have used genotoxic or hematotoxic chemicals other than benzene. The workforce needed to have been very stable, to have worked for at least 5 yr in the same factory and to have had little or no task rotation. Two shoe factories in Tianjin, China were selected based on the above criteria. Factory A is a small facility consisting of one workshop with 58 production workers producing
70 000 ‘uppers’ for luxury shoes per year (Table 1). The production process is characterized by a low level of mechanization and so requires a skilled labor force. No general ventilation is present (with only doors able to be opened) and no personal protective equipment (PPE) is used. Factory B is a large shoe manufacturing facility consisting of five workshops with about 560 workers producing 1 400 000 shoes per year. The level of mechanization is higher than in factory A, with shoe components transported on conveyor systems to machines and work positions. All gluing activities are performed manually. Drying ovens are equipped with local exhaust ventilation systems while ceiling and wall fans provide air movement and general ventilation in the workshops. No PPE is used.
|
The shoe production process of both factories can be divided into five main production functions: (i) ‘cutting’, (ii) ‘modeling’, (iii) ‘fitting’, (iv) ‘finishing’ and (v) ‘packing’. In ‘cutting’, leather or other shoe materials are sorted and prepared, then cut out either by hand or with automatic presses. In ‘modeling’, the different components of the ‘uppers’ are sewn or glued together. In ‘fitting’, the uppers and the heels, inner soles and outer soles are glued and sewn together. Shoes are cleaned, waxed and polished (‘finishing’) and, finally, inspected and packaged (‘packing’). Subjects were enrolled in the study through random selection, based on the expected range of exposures defined by their job titles and preliminary exposure data. Thirty-seven workers from factory A (64% of the workforce) were enrolled in the study. In factory B, 213 workers were enrolled (38% of the workforce). The study was approved by Institutional Review Boards at the NCI and the National Institute of Occupational Health and Poison Control, China CDC, Beijing. All recruited study subjects were informed of the nature of the study, the potential benefits and potential risks and provided written informed consent.
Information on the production process and work practices was obtained during several walk-through surveys and interviews with workers and management. Data on the amounts of glues used were abstracted from factory records.
Glue analyses
Bulk samples of glues used in factory B were collected in June 2001. No bulk samples of glues used in factory A could be obtained, as permission for collection of source samples could not be obtained from management. Samples were collected in gas-tight tubes, sealed and stored in a protective container. Bulk samples were analyzed for benzene and toluene content via static headspace gas chromatography (GC-MS) by US Environmental Protection Agency (EPA) method 8260. The percentages of benzene and toluene in the headspace of the glue were estimated as the ratio between the peak areas of benzene or toluene divided by the sum of all peak areas.
Personal exposure sampling
Workplace sampling
Personal sampling employed 3MTM organic vapor monitors (OVMs), which were attached to the worker’s lapel over the full work shift. After collection, the monitors were individually sealed and transported to the laboratory of the National Institute of Occupational Health and Poison Control, China CDC, Beijing, for analysis of benzene and toluene. In 15% of the samples a duplicate monitor was worn on the other lapel, which was sent to a commercial laboratory in the USA for analysis. Measurements in factory A were carried out on one random day each month between March and June 2000. Biological samples were collected from the 37 participating workers in June. In May 2001, air sampling was repeated for nine of these workers who provided a repeat biological sample the same month. Exposure measurements in factory B were carried out on 1–2 days per month between March 2000 and June 2001. In June 2000, exposure was measured only for workers who provided a biological sample in the pilot phase of the study (n = 20). Samples were collected and processed by the same personnel working in the same laboratories and using essentially the same protocols throughout the duration of the study. In total, 2783 exposure measurements (not including duplicate monitors) were collected, providing data on 250 workers (number of samples per subject: median 12, range 2–15).
Home sampling
Personal exposure measurements to benzene and toluene during non-working hours were taken once during the winter (December 2000) and once or twice during the spring (March/April 2001) using OVMs. Samplers were placed on the workers’ lapels at the end of the work shift with instructions to wear them until they returned to work the next day (14–16 h). The sampler was to be placed next to the bed at night. Duplicate monitors were worn for 18% of the measurements. In total, 595 solvent exposure measurements (not including duplicate monitors) were collected.
Analyses
Analyses of the monitors for benzene and toluene were conducted by GC-FID according to NIOSH method 1501 (NIOSH, 1994). The limit of detection (LOD) was 0.20 p.p.m. for benzene and 0.30 p.p.m. for toluene. A subset of duplicate OVMs from the workplace (n = 66), collected between March and June 2001, were analyzed for other organic hydrocarbons (n = 29) according to OSHA method 7 (OSHA, 2001) (GC-FID) to identify potential exposures to 1,1,1-trichloroethane, 1,2,4-trimethylbenzene, 1,2-dichloroethane, 1,4 dioxane, acetone, benzene, butyl acetate, chlorobenzene, cyclohexanone, decane, ethyl acetate, ethyl benzene, heptane, hexane, isooctane, methyl ethyl ketone (MEK), methyl isobutyl ketone (MIBK), methylene chloride (DCM), octane, pentane, propyl acetate, styrene, tetrachloroethylene, tetrahydrofuran, toluene, trichloroethylene, m-xylene, o-xylene and p-xylene. For the analyses of other hydrocarbons, OVMs were selected to represent all workplace activities and glue sources, in order to identify potential co-exposures to organic hydrocarbons.
Statistical analyses
There was a very high correlation between the logged air levels of benzene in the monitors analyzed at the CDC in China and in the duplicate monitors analyzed at the commercial laboratory in the USA (Pearson r = 0.93). In addition, the slope of the least squares regression line (forced through the origin) between the two sets of measurements was close to unity (slope = 1.10). Similar results were obtained from the analysis of the duplicate monitors for toluene (r = 0.95, slope = 0.99). These results indicate good agreement between measurements of benzene and toluene by the two laboratories. Subsequently, the results will be based on the analyses performed by the China CDC.
Workplace measurements that had an effective sampling duration of <240 min were excluded from the analyses (n = 152, 5.2%). All concentrations below the LOD [n = 370 (13.3%) for benzene and n = 238 (8.6%) for toluene] were replaced with values equal to the LOD divided by 
2 (Hornung and Reed, 1990). Exposure levels are summarized as averages (AM), geometric means (GM) and geometric standard deviations (GSD) and the 10th and 90th percentile of the exposure distributions.
Normal probability plots indicated that exposures by factory could best be described by log-normal distributions. Therefore, natural logarithms of exposure concentrations were used in the statistical analyses. To evaluate determinants of exposure in factory B, mixed-effects models were constructed (model 1). Potential exposure determinants (e.g. season, ventilation, amount of benzene and toluene used per day and distance to exposure source) were treated as fixed effects, whereas the worker and error were treated as random effects.
where Yij is the natural logarithm of the exposure concentration measured on the jth day for the ith worker, µ is the true underlying mean of log-transformed exposure averaged over all strata, ß1, ß2,..., ßk are fixed effects of k exposure determinants, 
i is the random effect for the ith worker and 
ij is the random error (within-worker day-to-day variation).
The model assumes that all i and ij are mutually independent and normally distributed with 0 means and variances BW
y2 and WWy2, representing between- and within-worker variability, respectively. These variances are estimated as between-worker (BWSy2) and day-to-day (WWSy2) covariance components.
To quantify the contribution of the fixed effects to the BWSy2 and WWSy2 covariance components, we compared the values of the covariance components obtained under model 1 with those from a mixed-effects model without the inclusion of fixed effects (e.g. one-way random effects model).
To examine whether the interrelationships between the different hydrocarbons could be attributed to distinct exposure sources present in the workplace, principal component analyses (with VARIMAX rotation) were performed (Fisher and van Belle, 1993). Only principal components (PCs) with eigenvalues > 1 were considered significant (i.e. explaining more multiple correlation than a single hydrocarbon concentration). Numerical indices representing each PC were generated as principal component scores and were used as dependent variables in subsequent regression analyses with handling of particular glues as independent variables. All analyses were carried out using SAS version 8.0 software (SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Glues
Factory A used two glues in the production process (total use
18 000 kg/yr) (Table 2). Glue A was used in ‘fitting’ and ‘finishing’, while glue B was used in ‘modeling’. In total, 17 study subjects were involved in gluing activities. In factory B, seven glues were used on a regular basis, accounting for almost all of the glue used per year (220 000 kg/yr). Glues C and E were used in ‘modeling’, glues D, G, H and I were predominantly used in ‘fitting’ and glue F was used solely in ‘packing’. In total, 43 study subjects from factory B were involved in gluing activities. Estimated benzene content of the glues ranged from 0.6 to 34%, with all but one glue having benzene levels 2%. Estimated toluene content ranged from 3 to 46%, with glues used in ‘fitting’ having a somewhat higher toluene content.
|
Workplace sampling: benzene and toluene
The average benzene exposure level in factory A was 21.86 p.p.m., ranging from 9.04 p.p.m. among workers in ‘modeling’ to >20 p.p.m. among workers who worked in ‘finishing’ or worked directly with glues (Table 3). As there was no physical separation between the operations in factory A, workers who did not handle any glues were still exposed on average to
10 p.p.m. benzene. Toluene exposure levels showed a similar pattern, but were about half of those for benzene. Overall, benzene and toluene exposure levels were highly correlated (r = 0.91, P < 0.0001).
|
The average benzene level in factory B was 3.46 p.p.m., ranging from 0.45 p.p.m. in cutting to 24.18 p.p.m. for subjects who glued packaging materials. Average toluene exposure levels ranged from 1.49 p.p.m. in ‘cutting’ to 58.83 p.p.m. for subjects involved in gluing of the outer sole. Workers directly involved in gluing had higher benzene and/or toluene exposure levels (depending on the particular glue used) than subjects within the same job titles who did not handle glues. Due to different sources and physical separation between ‘cutting’ (workshop 1), ‘modeling’ (workshops 2 and 3), ‘fitting’, ‘finishing’ and ‘packing’ processes (workshops 4 and 5), the overall correlation between benzene and toluene was lower (r = 0.46, P < 0.0001) than in factory A.
Benzene and toluene exposure levels in factory A did not show a clear seasonal trend over the follow-up period. In contrast, exposure levels in factory B showed a clear seasonal pattern with the lowest exposures during the summer months, followed by the fall and spring and higher exposures during the winter months (Fig. 1). Mixed-effects modeling, with season as the fixed effect, showed that the different seasons explained 17 and 24% of the day-to-day variance (WWSy2) for benzene and toluene, respectively. The use of ceiling and wall fans and opening of windows largely coincided with the different seasons and changes in ambient temperatures and therefore these factors could not be used as fixed effects in one single model. As the model fit criterion, Akaike’s information criterion, favored the model with the use of fans and opening of windows as the fixed effect and, as these are more ‘objective’ determinants of exposure, the final model presented is based on these parameters (Table 4). Benzene and toluene exposure increased with the amount of benzene or toluene used per day at the nearest glue source. Direct handling of glues (distance = 0 m) resulted in the highest exposures, which decreased with increasing distance from the source for subjects that did not directly handle glues. The use of ceiling and wall fans reduced benzene and toluene exposure by 43 [100 x (1 – exp–0.57)] and 47%, respectively. Opening the windows in addition to the use of fans resulted in total reductions of 78 and 84% for benzene and toluene exposure, respectively, compared with no general ventilation. The mixed-effects model explained 52% of the between-worker variance (BWSy2) for both benzene and toluene exposure and explained slightly more of the day-to-day variance for toluene (27%) than for benzene (21%). The explained day-to-day variance in benzene and toluene exposures could largely be attributed to changes in air movement and ventilation (95 and 71%, respectively) with absolute effects similar to the observed seasonal effects, corroborating the conclusion that the observed seasonal pattern was largely due to changes in air movement and ventilation patterns (data not shown).
|
|
Workplace sampling: co-exposures to organic hydrocarbons
In total, 11 different hydrocarbons were detected in the selected samples from factory A, of which nine were detected in >25% of the samples (Table 5). Benzene, toluene, pentane, ethyl benzene, hexane and m-xylene were detected in all samples, followed by p-xylene (83.3%), 1,1,1-trichloroethane (83.3%) and heptane (66.7%). In factory B, 16 different hydrocarbons were detected in the selected samples, of which eight could be detected in over 25% of the samples. Only toluene was detected in all samples, followed by benzene (93.3%), pentane (81.7%), hexane (75.0%), MEK (71.7%), ethyl acetate (61.7%), acetone (60.0%) and heptane (36.7%). Exposure levels of hydrocarbons other than benzene and toluene were low in both factories (generally <5 p.p.m.) and were less than 5% of the current ACGIH threshold limit value (TLV). None of the individual measurements exceeded the respective TLV, except those for benzene.
|
Principal component (PC) analysis of multiple correlations among levels of hydrocarbons measured in factory B revealed that four sources of variability accounted for 78% of the multiple correlations among the measured compounds (Table 6). The first PC was positively associated with pentane, hexane, heptane, 1,1,1-trichloroethane and octane; the second PC with toluene, MEK, ethyl acetate, 1,2-dichloroethane and acetone; the third PC with m-xylene, MIBK, ethyl benzene and DCM; the fourth PC with benzene. Subsequent regression analyses, using the scores of the PCs as dependent variables, revealed that PC1, PC2 and PC4 were strongly related to certain glues (Table 7). However, no relationship was found between PC3 and any documented glue source. Analyses with job titles and other production characteristics did not reveal any clear association with the identified PC (not shown), although all positive samples were detected in a subset of samples from subjects fitting the outer sole.
|
0.70)
|
Home sampling: benzene and toluene
Personal monitoring of subject exposures to benzene and toluene during non-working hours revealed no detectable levels of these substances outside the workplace in either winter or spring. This result was confirmed in duplicate monitors analyzed at the US laboratory (n = 105), which had somewhat lower limits of detection for benzene (LOD = 0.03 p.p.m.) and toluene (LOD = 0.16 p.p.m.). This indicates that home exposures did not contribute significantly to benzene and toluene exposures among these workers.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We assessed benzene, toluene and other solvent exposures in two shoe factories in China. Benzene and toluene exposures were estimated from passive air monitors analyzed at the Chinese CDC laboratory. Previous studies utilizing the same laboratory had implemented detailed quality control procedures to ensure high quality analyses (Qu et al., 2002). In this study, results from the CDC correlated well with duplicate analyses performed at a commercial laboratory in the USA.
Benzene exposure levels in the smaller shoe factory (factory A) were substantially higher than those in the larger shoe factory (factory B). However, benzene exposure distributions did overlap between the low exposed job titles of factory A and high exposed job titles in factory B. Since ‘fitting’ and ‘finishing’ processes in factory A showed the highest benzene exposure levels, it seems likely that the treatment glue used in factory A contained a significant amount of benzene. Due to the high source strength, poor ventilation, close vicinity and absence of any physical separation of the processes in factory A, all workers in this facility were exposed to relatively high benzene levels (10 p.p.m.).
There was more variability in benzene and toluene exposure levels in factory B than in factory A, both between workers (benzene, BWSy2 = 0.77 versus 0.10; toluene, BWSy2 = 1.27 versus 0.14) and within workers (benzene, WWSy2 = 0.88 versus 0.60; toluene, WWSy2 = 0.85 versus 0.36). The observed difference in day-to-day variance between the two factories was not influenced by the shorter observation period for factory A, as WWSy2 for the exact same observation period in factory B was essentially the same (WWSy2 = 0.89) as WWSy2 for the full observation period. The temporal variation in factory B was partly due to differences in air movement and general ventilation during the year. Only crude measures of air movement and general ventilation (i.e. fan use and open windows) could be used in the model. Measurements of the actual ventilation or of outdoor wind speed and direction, which could have had a significant role in the effectiveness of the general ventilation on a given day, were not available. However, the strong repetitive nature of the seasonal trend over the 16 month time period clearly indicates an important influence of seasonal differences in air movement and general ventilation patterns, as was confirmed with mixed-effects modeling using season as a covariate.
Differences in benzene exposure between job titles in factory B could be ascribed largely to the use of different glues in various stages of the production process and the proximity of workers to these sources. In the beginning of the production process (‘cutting’) no glues were used and benzene levels in this part of the factory were generally <1 p.p.m. Workers in ‘modeling’ and ‘fitting’ who were not actively involved in gluing procedures were generally exposed to benzene levels between 1 and 5 p.p.m. Workers in ‘modeling’ and ‘fitting’ who also performed gluing activities were exposed to benzene levels up to 10 p.p.m. Even higher benzene exposures occurred among workers who handled benzene glue in ‘packing’ or worked in close proximity to this activity (‘finishing’).
Although benzene and toluene were clearly the dominant exposures, vapors of 18 other hydrocarbons were detected in selected personal samples from the two factories. Exposure levels were generally low (5 p.p.m.) for these other hydrocarbons, representing less than 5% of their respective ACGIH TLVs. We used principal component analysis to link the presence of different hydrocarbons to specific glues used in factory B, where most hydrocarbon data were obtained. As specific combinations of glues were used within particular production processes (see Table 3), the principal component analyses clustered glues that were used in these processes. Interestingly, one principal component (PC3) could not be linked to the use of specific glues or other production characteristics. Given the low exposure levels of the hydrocarbons associated with this PC and the apparently sporadic occurrence of these hydrocarbons, this exposure is most likely associated with a source that was not used routinely in the shoe production process, possible a specialty glue or an adhesive remover.
Because the identified glues were used in different parts of the production process, it is unlikely that hydrocarbon exposures associated with these glues would confound possible relationships between benzene exposure and particular biological end-points. This conclusion is further supported by the fact that exposures to hydrocarbons other than toluene were generally very low and are not suspected of being hematotoxic or genotoxic or of influencing benzene metabolism at these levels. Finally, the identification of these distinct sources of co-exposures does facilitate the quantitative estimation of co-exposures for all study subjects, as the location of each study subject in relation to the major exposure sources (e.g. glues) is known and as glue formulations appeared to be constant over time, based on the observed ratios between benzene and toluene over the observation period (data not shown).
Toluene has been shown to competitively inhibit benzene metabolism and therefore might alter the dose–response relationships between benzene and biological end-points (Medinsky et al., 1994). The correlation between benzene and toluene was high for factory A (r = 0.91) and was modest for factory B (r = 0.46). When all measurements were combined (n = 2783) the correlation between benzene and toluene among the total study population was 0.44. Given this modest level of correlation, it should be possible to determine if any observed dose–response relationship between benzene and a particular biological end-point is modified by toluene exposure.
No exposure to benzene and toluene was detected during non-working hours. The lowest exposure level that could possibly have been detected was 0.03 p.p.m. for benzene and 0.16 p.p.m. for toluene (for a sampling duration of 14–16 h). Although these detection limits were much greater than air concentrations of benzene and toluene that have been reported indoors and outdoors in Asia (Guo et al., 2003), they indicate that sources outside the workplace should not significantly add to the occupational exposures of workers in these factories. Future analyses of urinary benzene and its metabolites and benzene-related protein adducts in blood should further corroborate this finding and, in conjunction with questionnaire data, should shed light on the contribution of environmental and active tobacco smoking on internal benzene exposure levels.
As the study factories were selected based on several a priori eligibility criteria, it is unclear how representative these two shoe factories are of the contemporary Chinese shoe manufacturing industry. However, this study showed that benzene exposure levels are still rather high in at least some parts of the Chinese shoe manufacturing industry. It is noteworthy that 53% of the benzene measurements in factory A were above the Chinese exposure limit at the time of surveying (then 12.3 p.p.m., currently 1.9 p.p.m.), while 4% of the measurements in factory B were above the exposure limit. Given the known health risks of benzene at high levels of exposure, it is imperative that control measures be implemented in these two shoe factories. Based on measurement results collected during this survey, we made specific recommendations to improve the work environment. Since the survey was completed, factory B has installed local exhaust ventilation on almost all workstations where glues are used. In addition, the high benzene-containing glue has been replaced with a glue that has a lower benzene content.
A dose-dependent decline in peripheral blood cells is a well-known indicator of benzene exposure (Rothman et al., 1996; Qu et al., 2002). Preliminary analyses of the hematological data, collected as part of the hematological, cytogenetic and molecular end-points in this study, showed a clear dose-dependent decline between the benzene exposure levels described here and peripheral white blood cells (Lan et al., 2003). This observation seems to be consistent with a recently published study on hematological effects among Chinese workers with a broad range of benzene exposures (Qu et al., 2002) and are an additional indication that exposures were assessed accurately.
In summary, we identified two shoe factories in the same geographical region of China that had very stable workforces. Over the course of the 16 month survey, there was essentially no task rotation in either workplace. We demonstrated the presence of a broad range of benzene and toluene exposure levels in personal air samples among these workers and identified production processes with distinct exposure patterns. Benzene and toluene exposure levels could be largely explained by the amount of benzene and toluene handled per day, the average distance to the nearest exposure source and air movement and ventilation patterns. Preliminary analyses of dermal exposure data collected as part of the current study indicate that this route of exposure did not contribute substantially to the total benzene and toluene doses received (unpublished data). Exposures to other measured hydrocarbons were quite low in both workplaces and were generally attributable to distinct glues. Overall, these data support the appropriateness of combining exposure and biological end-point data from the two factories, which, from an exposure perspective, differ primarily with regard to benzene exposure levels. The availability of long-term monthly personal monitoring data provides an excellent opportunity to estimate individual exposures at different times during the 1 yr period of observation and enables us to define subgroups of individuals who experienced relatively low levels of benzene exposure throughout the year.
Acknowledgements—The authors are indebted to all workers who took part in exposure survey. M.T.S. and L.Z. are supported by NIH grants RO1 ES06721, P30 ES01896 and P42 ES04705. S.M.R. is supported by NIH grant P42ES05948.
|
FOOTNOTES |
|---|
* Author to whom correspondence should be addressed. Tel: +1-301-402-9850; fax: +1-301-402-1819; e-mail: vermeulr{at}mail.nih.govThe first two authors contributed equally to this work
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ACGIH. (2002) TLVs and BEIs; threshold limit values for chemical substances and physical agents and biological exposure indices. Cincinnati, OH: American Conference of Governmental Industrial Hygienists.
Aksoy M. (1985) Malignancies due to occupational exposure to benzene. Am J Ind Med; 7: 395–402.[Web of Science][Medline]
Austin H, Delzell E, Cole P. (1988) Benzene and leukemia. A review of the literature and a risk assessment. Am J Epidemiol; 127: 419–39.
Bond GG, McLaren EA, Baldwin CL, Cook RR. (1986) An update of mortality among chemical workers exposed to benzene. Br J Ind Med; 43: 685–91.[Web of Science][Medline]
Fisher LD, van Belle G. (1993) Principal component analysis and factor analysis. In Biostatistics: a methodology for the health sciences. New York: Wiley Interscience. pp. 692–762.
Guo H, Lee SC, Li WM, Cao JJ. (2003) Source characterization of BTEX in indoor microenvironments in Hong Kong. Atmos Environ; 37: 73–82.
Hayes RB, Yin SN, Dosemeci M et al. (1997) Benzene and the dose-related incidence of hematologic neoplasms in China. Chinese Academy of Preventive Medicine–National Cancer Institute Benzene Study Group. J Natl Cancer Inst; 89: 1065–71.
Hornung R, Reed LD. (1990) Estimation of average concentration in the presence of nondetectable values. Appl Occup Environ Hyg; 5: 46–51.
IARC. (1982) IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans: some industrial chemicals and dyestuffs: benzene. Lyon: IARC. pp. 93–148.
Lan Q, Li G, Zhang L et al. (2003) Hematoxicity among workers exposed to benzene in China. Redefining the frontiers of science, Proceedings of the 94th annual meeting of the American Association for Cancer Research. Philadelphia, PA: AACR. p. 1284.
McCraw DS, Joyner RE, Cole P. (1985) Excess leukemia in a refinery population. J Occup Med; 27: 220–2.[Web of Science][Medline]
Medinsky MA, Schlosser PM, Bond JA. (1994) Critical issues in benzene toxicity and metabolism: the effect of interactions with other organic chemicals on risk assessment. Environ Health Perspect; 102 (suppl. 9): 119–24.
NIOSH. (1994) Hydrocarbon aromatics: method 1501. In NIOSH Manual of Analytical Methods, 4th edn. Cincinnati, OH: National Institute of Occupational Safety and Health. Available online at http://www.cdc.gov/niosh/nmam/.
NIOSH. (2003) NIOSH pocket guide to chemical hazards, 3rd edn. Cincinnati, OH: National Institute of Occupational Safety and Health.
OSHA. (2001) Organic vapors, method 07. In Sampling and Analytical Methods. Washington, DC: US Department of Labor. Available online at http://www.osha-slc.gov/dts/sltc/methods.
Qu Q, Shore R, Li G et al. (2002) Hematological changes among Chinese workers with a broad range of benzene exposures. Am J Ind Med; 42: 275–85.[CrossRef][Web of Science][Medline]
Rinsky RA, Smith AB, Hornung R et al. (1987) Benzene and leukemia. An epidemiologic risk assessment. N Engl J Med; 316: 1044–50.[Abstract]
Rothman N, Li GL, Dosemeci M et al. (1996) Hematotoxocity among Chinese workers heavily exposed to benzene. Am J Ind Med; 29: 236–46.[CrossRef][Web of Science][Medline]
Savitz DA, Andrews KW. (1997) Review of epidemiologic evidence on benzene and lymphatic and hematopoietic cancers. Am J Ind Med; 31: 287–95.[CrossRef][Web of Science][Medline]
Smith MT, Zhang L, Wang Y et al. (1998). Increased translocations and aneusomy in chromosomes 8 and 21 among workers exposed to benzene. Cancer Res; 58: 2176–81.
Swaen GM, Meijers JM. (1989) Risk assessment of leukaemia and occupational exposure to benzene. Br J Ind Med; 46: 826–30.[Web of Science][Medline]
Wong O. (1987) An industry wide mortality study of chemical workers occupationally exposed to benzene. II. Dose response analyses. Br J Ind Med; 44: 382–95.[Web of Science][Medline]
Yin SN, Li GL, Tain FD et al. (1987) Leukaemia in benzene workers: a retrospective cohort study. Br J Ind Med; 44: 124–8.[Web of Science][Medline]
Yin SN, Hayes RB, Linet MS et al. (1996) A cohort study of cancer among benzene-exposed workers in China: overall results. Am J Ind Med; 29: 227–35.[CrossRef][Web of Science][Medline]
Zhang L, Rothman N, Wang Y et al. (1998) Increased aneusomy and long arm deletion of chromosomes 5 and 7 in the lymphocytes of Chinese workers exposed to benzene. Carcinogenesis; 19: 1955–61.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Q. Lan, L. Zhang, M. Shen, W. J. Jo, R. Vermeulen, G. Li, C. Vulpe, S. Lim, X. Ren, S. M. Rappaport, et al. Large-scale evaluation of candidate genes identifies associations between DNA repair and genomic maintenance and development of benzene hematotoxicity Carcinogenesis, January 1, 2009; 30(1): 50 - 58. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Hagstrom, C. Lundholm, K. Eriksson, and I. Liljelind Variability and Determinants of Wood Dust and Resin Acid Exposure during Wood Pellet Production: Measurement Strategies and Bias in Assessing Exposure-Response Relationships Ann. Hyg., November 1, 2008; 52(8): 685 - 694. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. McHale, Q. Lan, C. Corso, G. Li, L. Zhang, R. Vermeulen, J. D. Curry, M. Shen, R. Turakulov, R. Higuchi, et al. Chromosome Translocations in Workers Exposed to Benzene J Natl Cancer Inst Monographs, July 1, 2008; 2008(39): 74 - 77. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kim, R. Vermeulen, S. Waidyanatha, B. A. Johnson, Q. Lan, M. T. Smith, L. Zhang, G. Li, M. Shen, S. Yin, et al. Modeling Human Metabolism of Benzene Following Occupational and Environmental Exposures. Cancer Epidemiol. Biomarkers Prev., November 1, 2006; 15(11): 2246 - 2252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. SYMANSKI, S. MABERTI, and W. CHAN A Meta-Analytic Approach for Characterizing the Within-Worker and Between-Worker Sources of Variation in Occupational Exposure Ann. Hyg., June 1, 2006; 50(4): 343 - 357. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kim, R. Vermeulen, S. Waidyanatha, B. A. Johnson, Q. Lan, N. Rothman, M. T. Smith, L. Zhang, G. Li, M. Shen, et al. Using urinary biomarkers to elucidate dose-related patterns of human benzene metabolism Carcinogenesis, April 1, 2006; 27(4): 772 - 781. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Vermeulen, Q. Lan, L. Zhang, L. Gunn, D. McCarthy, R. L. Woodbury, M. McGuire, V. N. Podust, G. Li, N. Chatterjee, et al. Decreased levels of CXC-chemokines in serum of benzene-exposed workers identified by array-based proteomics PNAS, November 22, 2005; 102(47): 17041 - 17046. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Lan, L. Zhang, M. Shen, M. T. Smith, G. Li, R. Vermeulen, S. M. Rappaport, M. S. Forrest, R. B. Hayes, M. Linet, et al. Polymorphisms in Cytokine and Cellular Adhesion Molecule Genes and Susceptibility to Hematotoxicity among Workers Exposed to Benzene Cancer Res., October 15, 2005; 65(20): 9574 - 9581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Fustinoni, D. Consonni, L. Campo, M. Buratti, A. Colombi, A. C. Pesatori, M. Bonzini, P. A. Bertazzi, V. Foa, S. Garte, et al. Monitoring Low Benzene Exposure: Comparative Evaluation of Urinary Biomarkers, Influence of Cigarette Smoking, and Genetic Polymorphisms Cancer Epidemiol. Biomarkers Prev., September 1, 2005; 14(9): 2237 - 2244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B van Wendel de Joode, R Vermeulen, J J van Hemmen, W Fransman, and H Kromhout Accuracy of a semiquantitative method for Dermal Exposure Assessment (DREAM) Occup. Environ. Med., September 1, 2005; 62(9): 623 - 632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Lan, L. Zhang, G. Li, R. Vermeulen, R. S. Weinberg, M. Dosemeci, S. M. Rappaport, M. Shen, B. P. Alter, Y. Wu, et al. Hematotoxicity in Workers Exposed to Low Levels of Benzene Science, December 3, 2004; 306(5702): 1774 - 1776. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. BURSTYN Principal Component Analysis is a Powerful Instrument in Occupational Hygiene Inquiries Ann. Hyg., November 1, 2004; 48(8): 655 - 661. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||