Document Y9VzBBrdxJ6BK3dzM57ddwRrn

ELSEVIER Available online at www.sciencedirectcom #. / / " . , N . / / ''''"",;# ScienceDirect Regulatory Toxicology and Pharmacology 46 (2006) 149-156 Regulatory Toxicology and Pharmacology www.elsevier.com/locate/yrtph Benzene exposure in the shoemaking industry in China, a literature survey, 1978-2004 Laiming Wang a, Yimei Zhou b, Youxin Liang a,*, Otto Wong c,d,e, Thomas Armstrong b, A. Robert Schnatter b, Qiangen Wu a, linbin Fang a, Xibiao Ye a, Hua Fu a Richard D. Irons f,g a Fudan University School of Public Health, Shanghai, China b ExxonMobil Biomedical Sciences, Inc., Annandale, NJ, USA C Applied Health Sciences, San Mateo, CA, USA d Tulane University, New Orleans, LA, USA e Chinese University of Hong Kong, Hong Kong, China f Sino-US Joint Clinical and Molecular Laboratory (JCML) , Institutes of Biomedical Sciences and Department of Pathology, Fudan University, Shanghai, China g Department of Pharmaceutical Sciences, School of Pharmacy and Department of Pathology, School of Medicine, University of Colorado Denver and Health Sciences Centre, Denver, CO 20262 USA Received 19 April 2006 Abstract This article presents a summary of benzene exposure levels in the shoemaking industry in China reported in the Chinese medical lit- erature between 1978 and 2004. A comprehensive search identified 182 papers reporting such exposure data. These papers could be clas- sified into two categories: benzene poisoning case reports and industrial hygiene surveys. From each paper, the following information was abstracted whenever available: location and year of occurrence, occupation and/or task involved, benzene content in adhesives/sol- vents, work environment, working conditions, working hours, diagnosis, and air monitoring data of benzene. A total of 333 benzene measurements (88 averages, 116 minimums, 129 maximums) in the shoemaking industry were reported in the 182 papers identified. The data were analyzed in terms of geographical location, time period, type of ownership (state, township, or foreign), type of report (benzene poisoning reports vs. industrial hygiene surveys), and job title (work activity) or process. The reported data covered a wide range; some measurements were in excess of 4500 mg/m3 Thirty-five percent of the reported benzene concentrations were below 40 mg /m 3 , which was the national occupational exposure limit (OEL) for benzene between 1979 and 2001. The remaining 65% measure- ments, which exceeded the national OEL in effect at the time, and were distributed as follows: 40-100 mg/m3, II %; 100-300 mg / m 3 , 21%; 300-500 m g/ m 3 , 13%; and 5 0 0+ m g / m 3 , 20%. However, only 24% of the reported measurements after 2002 were below 6 mg / m 3 , i.e., Permissible Concentration-Time Weighted Average (PC-TWA) and 10 mg/m3, i.e., Permissible Concentration-Short Term Exposure Limit (PC-STEL), the newly amended benzene OELs in effect after May 2002. The data demonstrated that the majority of the facilities in the shoemaking industry reported in the literature were not in compliance of the OEL for benzene in effect at the time. Overall, the data show a clear downward trend of benzene exposure levels over the years, particularly after the introduction of the new lower OEL in 2002. Even though substantially lower when compared to levels in the past, current benzene exposure measurements from the literature review suggest that many facilities in the shoemaking industry in China have benzene concentrations that are still above the new OEL. The reported data, stratified by job, year and survey reason, can be used as part of the information and analysis for developing a job- exposure matrix in retrospective exposure assessment and thus may be part of the information used in developing historical exposure estimates in epidemiologic studies of shoe workers. 2006 Elsevier Inc. All rights reserved. * Corresponding author. Fax: +8621 64043069. E-mail address:yxliang@shmu.edu.cn (Y. Liang). 0273-23001$ - see front matter 2006 Elsevier Inc. All rights reserved. doi: 10.1 016/j .yrtph.2006.06.009 SHELL-MCCLURG-060106 150 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 Keywords: Shoemaking; China; Benzene; Adhesives; Glues; Solvents; Benzene poisoning; Occupational exposure limit; Retrospective exposure assessment; Job-exposure matrix 1. Introduction Shoemaking is one of the most economically important industries in China, producing approximately 7 billion pairs of shoes each year in the 2000s. It is estimated that there are 38,000 enterprises for the shoemaking industry in China with approximately 2 million shoe workers employed (Huang, 2000). Shoemaking consists of cutting, fitting, sewing and gluing various parts together. A number of occupational hazards are associated with the shoemaking industry. Shoe workers in China are potentially exposed to a variety of toxic chemicals that are present in adhesives, glues and solvents; such as benzene, toluene, xylene, methyl ethyl ketone, acetone, n-hexane, gasoline, and other organic solvents. Of particular health concern are adhesives or solvents that contain a high concentration of benzene. In addition, the general conditions in many workplaces in the shoe industry in China have been or still are unsatisfactory; often with no or inadequate ventilation and little or no personal protection. The Chinese medical literature is replete with reports of benzene overexposure and benzene poisoning cases (Wong, 2002, 2003a). In this literature survey, we summarize and present benzene exposure data in the shoemaking industry reported in the Chinese medical literature in the last three decades. The reported data represent documented historical benzene exposure levels in the workplaces in the shoemaking industry. These data are useful to demonstrate the impact of regulatory polices of occupational benzene exposure and to provide insights of historical benzene exposure levels associated with benzene-related hematopoietic diseases in China. 2. Materials and methods Primarily, there are two types of articles in the literature that report workplace benzene exposure levels: benzene poisoning case reports and systematic industrial hygiene surveys. A comprehensive online search was performed to identify reports of both types of articles. The online search included the following two databases: Chinese Biomedical Literature Database (1978-1993) and China National Knowledge Infrastructure (19942004). Key words used in the search included the Chinese equivalents of the following technical terms: shoemaking, benzene exposure, benzene poisoning, and occupational health. In addition to technical terms, several commonly used trade names of benzene-containing adhesives/solvents were also included (Chinese names: Ludingjiao, Tiannashui, Xiangjiaoshui). In addition to the online search, a supplemental manual search was also conducted, particularly for information describing the production technology and the history of chemical compositions of adhesives, glues and solvents used in the shoemaking industry. The titles and, whenever warranted, the abstracts of all articles published during 1978-2004 in 25 major Chinese occupational medical journals were reviewed. The online and manual searches resulted in 182 papers of benzene poisoning cases or industrial hygiene surveys published in 25 journals from 1978 to 2004. From each paper, the following information was abstracted whenever available: location and year of occurrence, occupation and/or task involved, benzene content in adhesives/solvents, work environment, working conditions, working hours, diagnosis, and air monitoring data of benzene. A paper might provide one or more sets of benzene measurements with each set representing one or more benzene exposure levels by work locations or job titles (work activity). In the 182 papers identified, benzene measurements were reported as individual samples or in aggregate as averages, minimums, and/or maximums. Arithmetic means were calculated for the reported exposure data: arithmetic means of reported averages, arithmetic means of reported minimums, and arithmetic means of reported maximums. The data were analyzed in terms of geographical location, time period, type of ownership (state, township, or foreign), type of report (benzene poisoning reports vs. industrial hygiene surveys), and job title (work activity) or process. From the 1950s to the 1970s, the occupational exposure limit (OEL) for benzene in China was 50 mgfm3 as a maximum allowable concentration (MAC). The regulatory limit was lowered to 40 mgfm3 (also as a MAC) in 1979, which remained in effect through 2001. In 2002, with the implementation of the new Occupational Diseases Prevention and Control Act, a new benzene OEL was introduced: 10 mg/ m3 as permissible concentration of 15 min short-term exposure limit (PC-STEL) and 6 mgfm3 as permissible concentration of 8-h time-weighted average (PC-TWA) (Liang et aI., 2005). Thus, the reported benzene data from 2002 to 2004 may provide the first indication of the immediate impact of the new OEL on benzene exposure in the shoemaking industry. 3. Results The collected data from the 182 papers covered 26 provinces, municipalities and autonomous regions; representing 84% of the 31 administrative regions in the country. However, the majority of the 182 papers (>80%) were reported from the most developed industrial areas in the coastal regions of eastern China; including Fujian, Zhejian, Shandong, Guangdong provinces and Shanghai municipality (Fig. 1). This distribution by geographical location was to be expected, as these areas are heavily industrialized and many shoemaking facilities are mostly located along coastal China. One of the factors determining benzene air concentration is the benzene content in solvents used. It is estimated Middle"south 12% North-east 9% South-west 2% North-west 4% North China 5% East China 68% Fig. 1. Geographical distribution of published papers reporting benzene exposure levels in the shoemaking industry in China. SHELL-MCCLURG-060107 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 151 that solvents constitute approximately 75-80% (by weight) of most adhesives used in the shoemaking industry in Chi- na. Of the 182 papers included in the literature survey, 28 reported benzene content in the solvents used. The report- ed benzene content in solvents ranged from 27% to 96% (average of 83%). Assuming on the average typical adhe- sives contain 80% solvents, the average benzene content in adhesives would be approximately 66%. However, it has been reported that, in some cases, pure benzene was used during the 1980s. Furthermore, benzene is present in many "non-benzene-based" solvents or adhesives. For example, many toluene-based industrial products, includ- ing toluene-based adhesives, that were manufactured in China contained about 10% benzene; with some as high as 20-30% of benzene as an impurity in the toluene (Chen and Chan, 1999). A total of 333 benzene measurements (88 averages, 116 minimums, 129 maximums) in the shoemaking industry were reported in the 182 papers identified. The reported data covered a wide range; some measurements were in excess of 4500 m g /m 3 . Based on the 333 measurements, the overall arithmetic mean, median and geometric mean were 422.4 mg/m3, 145.2 mg/m3 and 108.7 mg/m3, respec- tively. Thirty-five percent of the reported benzene concen- trations (averages, minimums or maximums) were below 40 mg/m3, which was the national occupational exposure limit (OEL) for benzene between 1979 and 200l. The remaining 65% of the measurements which exceeded the national OEL in effect at the time were distributed as follows: 40-100mg/m3, 11%; 100-300mg/m3, 21%; 300- 500 mg/m3, 13%; and 500+ mg/m3, 20%. The data report- ed in the literature thus suggest that most facilities in the shoemaking industry were not in compliance of the MAC OEL for benzene in effect at the time. The 182 papers reporting benzene exposure measure- ments were published between 1978 and 2004. As shown in Fig. 2, few papers were published before the 1990s. Starting in the early 1990s, with the increased development of the shoemaking industry and the improvement in occu- pational health services, an increasing number of papers were published in the medical journals. Of the 182 papers with benzene data, year of air monitoring or sampling was specified in 177 papers. The temporal trend of reported ~o ~------------------------~----------~ ..,. ., ... ...,.. ., ". "', .,. e - - - '""" '" "" "" '" '" '" '" '" '" '""" "'" '" ." "', ." "'" ...~ ~ ;;m~. z,r'>1 .ry c, ~. Zl cffi <., '"c:t:~ ~ ~ ~ ~ ~ ~~~ ~~ Q <>~l <> N ;~;":, >(O.,j '=' 0;; : Year Fig. 2. Number of published papers reporting benzene exposure levels in the shoemaking industry in China by year. benzene exposure levels in the shoemaking industry is pre- sented graphically in Figs. 3, a and b. Measurements taken in the 1970s and/orl980s were quite high, with results well above 400 mg/m3 (the arithmetic mean of averages) and 2000 mg/m3 (the arithmetic mean of maximums). Even as recently as the early 1990s, reported maximum levels were close to or higher than 1000 mg/m3. Overall, Fig. 3 indicates a tendency of declining benzene exposure in terms of all three exposure variables: averages, minimums and maximums. The entire time span covered by the literature survey (1978-2004) was divided into the following three time periods according to the growth of the industry and regulatory events: A. 1978-1990: Opening policy and the beginning of economic growth, B. 1991-2001: Further industrial development and growth, and C. 2002-2004: Post 2002 Occupational Diseases Prevention and Control Act and the introduction of the new OEL for benzene. Table l,a and b show reported benzene concentrations in the shoemaking industry in China by the three time peri- ods above. The arithmetic means of averages, minimums and maximums were the lowest for 2002-2004, and those for 1978-1990 were the highest. The arithmetic mean of the reported averages for 1978-1990 was 405.3 mg/m3, more than 10 times the maximum allowable concentration (MAC) of 40 mg/m3 in effect at the time. Although the reported averages were much lower after the introduction of the new OEL in 2002, the arithmetic mean of the report- ed averages (69.7 m g/m 3 ) during the interval 2002-2004 was still seven times the new STEL of 10 mg/m3 and more than 11-fold of the new TWA of 6 mg/m3. An analysis of variance indicated that the differences of arithmetic means of the reported averages for the three time periods were statistically significant (F = 8.471, p < 0.05). Although the arithmetic means of both mini- mums and maximums were also different for the three time periods, the differences were not statistically significant at the 5% level. The reported benzene data were stratified by type of report: benzene poisoning cases vs. industrial hygiene sur- veys Table 2. Both maximum and average benzene levels reported in benzene poisoning cases were significantly high- er than those reported in industrial hygiene surveys. The arithmetic means for the reported maximums and averages in benzene poisoning reports (1358 and 470 mg/m3, respec- tively) were more than twice those reported in industrial hygiene surveys (533 and 161 mg/m3, respectively). As most benzene poisoning cases appear to arise from high-ex- posure work environments; it is not surprising therefore those exposure measurements in such specific environments tend to be much higher than others. Information identifying the type of ownership of the industrial enterprises was provided in 84 papers. For SHELL-MCCLURG-060108 152 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 i..~.. Average 'l . jinlJaum A 451)0 ('1 4000 51 3500 "~- ~ 3000 251)0 ,~. 201)0 1500 (,) 10QO i.l u0 500 0 '" ". '" ''"" '.".. '" '" *' '.".. ''"" g... ... ... ... ... ... ,.. ... ,., ... ... ... ...~ 'c"ill 'OJ' ~ :- ::> 0 l> (;: (;: ~ ~ ill N (") ~ ~.., 'Z t- il) II> a'. a'. a" a" 0 :~::~: M 0 '6' 0 II> II> II> ~ 0'< a'. 1:. II> (1', II> II> 0 0 0 0 '" """,.; 'I"'!/: N NNN Year I...:..... min m-0-m ave ~max I a 5000 4500 4000 3500 ~ 3000 ." ~ 2500 ;i 2000 o '-' 1500 1000 / -"- /\ 500 ... - ... - ...00 t- ~ oo ~ "00' ~ u:> 00 ~ t- 00 ~ oo 00 ~ '0"0 ~ 0 '~" '~" ~''"" M '~" '~" ~"'"' u:> ~'" t- '~" oo '~" ''~"" 0 0 0 '" 0 0 '" '0" 0 '" M 0 0 '" 0 0 '" Year Imin. I--'--ave. max. b 5000 4500 4000 ';i 3500 'bD 3000 ~ 2500 2000 g 1500 / / ~~-A-~-.------,/L~----------4-\-~~-~-\I\'~~-/-----~-\-~------------------------------------------------------------------------------------- -... ~....,/- \/ \\ o " 1000 500 o year ---- ................... ~...-' //6 ....,./ ~ -''"" '"0 0 M 000 00 000 00 -'" '" '" '" '" '" Fig. 3. Temporal trend of reported benzene levels in the shoemaking industry in China by year (based on overall data). (a) Temporal trend of reported benzene levels in the shoemaking industry in China by year in IH survey reports. (b) Temporal trend of reported benzene levels in the shoemaking industry in China by year in Case report (Panel a shows a similar temporal trend as was shown by the overall data; Panel b shows a similar trend as for Fig. 3, but is not as certain because of the lack of data for certain time periods.). Table 1 Arithmetic means of reported benzene concentrations (mgfm3) at workplaces by time periods (based on overall data) Time periods No. paper Minimum Averagea No. paper Mean No. paper Mean 1978-1990 (A) 1991-2001 (B) 2002-2004 (C) 24 129 24 17 339.4 12 81 63.0 62 16 31.0 12 405.3 293.1 69.7 Total 177 114 a A vs. B, p > 0.05; A vs. C, p < 0.05; B vs. C, p < 0.05. 86 Maximum No. paper 18 91 18 127 Mean 2074 730.1 490.7 SHELL-MCCLURG-060109 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 Table la Arithmetic means of reported benzene concentrations (mgfm3) at workplaces by time periods in IH survey reports Time periods No. paper Minimum Average" 1978-1990 (A) 1991-2001 (B) 2002-2004 (C) 15 79 19 No. paper 9 55 11 Mean 218.5 26.7 12.9 No. paper 8 39 10 Mean 302.2 181.3 64 Total 113 75 57 a A vs. B, p > 0.05; A vs. C, p < 0.05; B vs. C, p < 0.05. Maximum No. paper 10 58 13 81 153 Mean 1194.6 480.8 295.1 Table Ib Arithmetic means of reported benzene concentrations (mgfm3) at workplaces by time periods in case reports" Time periods No. paper Minimum Average No. paper Mean No. paper Mean 1978-1990 (A) 9 8 229 5 663.9 1991-2001 (B) 48 26 155.8 22 464 2002-2004 (C) 5 5 59.1 2 129.2 Total 62 39 29 a No statistical differences were shown among any compared group. Maximum No. paper 8 33 5 46 Mean 2411.4 1144.4 1086.4 Table 2 Comparison of benzene levels (mg/m3) in benzene poisoning case reports and industrial hygiene surveys Benzene levels Source No. Arithmetic mean Range Median P25 Maximum Average Minimum Case report Ind hyg survey Case report Ind hyg survey Case report Ind hyg survey 46 83 29 57 39 77 1358 533 470 161 85 80 * In most samples, "0" denotes benzene levels below detection limit. 82-4968 2.5-4800 13-1496 2-1000 0*-995 0*-1840 1002 339 370 93 30 11 277 89 167 37 3 2 P 75 2084 582 668 222 54 52 t-test for means 4.88 4.36 0.76 p <0.01 <0.01 >0.01 Table 3 Comparison of benzene levels by ownership of enterprises Type of ownership Comparison Total Average (mg/m3) Aver difference Foreign-invested (A) Town-/village-owned (B) State-owned (C) A vs. B A vs. C B vs. C 131.38 488.67 467.89 352.29 336.51 15.78 p <0.05 <0.05 >0.05 IH survey Average (mg/m3) Aver difference 137.56 243.96 467.89 112.66 336.59 223.93 p >0.05 >0.05 >0.05 Case report Average (mg/m3) Aver difference 73.7 650.1 NA 576.4 p <0.05 analysis, three broad categories of ownership were used: state, township or village, and foreign investors. Table 3 compares reported benzene levels by category of owner- ship. The average for reported benzene levels was much lower in foreign-owned facilities when compared with the averages in state-owned or township-owned enterprises. However, even for foreign-owned enterprises, the arithme- tic mean of reported averages (131.38 m g /m 3 ) was more than threefold of the OEL for benzene in effect at the time (40 mg/m3). There were 1054 cases of benzene poisoning in the shoe- making industry reported in the literature between 1978 and 2004. Typically, in these case reports a description of the work environment was presented along with benzene measurements. The following two examples are typical of benzene poisoning case reports in the Chinese medical literature. In 1995 Yang reported two cases of benzene poisoning at a small workshop making slippers in the province of Zhejiang (Yang, 1995). The adhesive used was a mixture of chloroprene and benzene (1:4). The work area was actually a makeshift workshop inside a farmer's residence, with little or no ventilation. Four heating lamps were used to enhance drying after the application of adhesive. Two women were diagnosed with benzene poisoning after working there 10 h a day for 4 months. Benzene levels measured at the workshop ranged from 302.4 to 1383.8 mg/m3 SHELL-MCCLURG-060110 154 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 Table 4 Comparison of benzene poisoning case by job titles (25-75% intervals added) Work activity No. case Arithmetic mean (mgfm3) Median P25 P 75 Range (mgfm3) Uppers and soles fitting Soles making Uppers making Temporary uppers embedding 418 395 139 102 897 1155 290 774 476 277 1143 5-3117 478 278 1384 103-4010 229 99 512 88-560 574 427 1290 278-1530 Total 1054 (average of 731.3 mg / m 3 ) . A similar episode of benzene poisoning in a small shoe workshop in Suzhou was report- ed by Tao in 2000 (Tao, 2000). The small workshop was described as a room inside a farmer's residence. There was no ventilation system and the windows were often closed. A mixture of chloroprene and benzene (1:4) was used as adhesive. Two charcoal-heated stoves were used for drying. Air benzene concentrations were recorded above 1000 mg/m3. The distribution of benzene poisoning cases by job cat- egory is presented in Table 4. The largest number of cases (418) occurred among workers who assembled the upper parts and the soles of shoes, and the arithmetic mean of the reported averages was 897 mg/m3. Among workers making soles, there were 395 benzene poisoning cases, and the arithmetic mean of the reported average exposures was 1155 mg/m3. By comparison, workers in "temporary uppers embedding" and "uppers making" had fewer poi- soning cases and the exposure associated with "uppers making" was lower (arithmetic mean of 290 m g /m 3 ). The number of work hours varies as well. According to the published reports, many workers in the shoemaking industry worked long hours, far exceeding 40 h a week in many instances, even though "8 work hours per day and 5 work days a week" was specified by the State Council PR China in May 1995. The published reports indicated that many employees frequently worked between 10 and 14 h a day and some worked as many as 20 h a day to meet deadlines of "rush orders" Table 5. According to 52 of the 182 published reports, most workers in the shoemaking industry spent 11-14 h working per day, with 6 and 7 day workweeks not uncommon. The impact of routinely pro- longed work hours is an important aspect to properly con- sider in exposure assessment studies. Finally, an attempt was made to stratify exposure data by job title (work activity) and time period Table 6. Exposure data in this format, (particularly those from the industrial hygiene survey reports) can provide reference for historical exposure estimates used in, for example, epidemiologic stud- Table 5 Work hour for some work activities of shoe making workers Work activity Work hours per day Minimum Maximum Uppers and soles fitting Soles making Temporary uppers embedding Uppers making 7 8 10 9 16 20 17 14 Average 11.5 12.9 14.0 11.7 ies of workers in the shoemaking industry. The discussion section below includes a review of some of the limitations on the use of the data for Job Exposure Matrix (JEM) application. As expected, Table 6 indicates that both job task (work activity) and time period are important parameters in characterizing historical exposures. In terms of maximum exposure, measurements in the early time period 1978-1990 were the highest (~ 1500 mg/m3). However, the middle time period 1991-2001 was associated with the highest average exposures. For example, "temporary uppers embedding" in 1991-2001 entailed the highest average exposure (mean of 534.6 mg/m3) and "soles making" in the same time period the second highest average exposure (466.6 mg/m3). The table also identifies gaps in the exposure data. For example, there are no sufficient data available for "temporary uppers embedding" in 2002-2004. As indicated before, Table 6 shows that exposures in the shoemaking industry have declined significantly during the most recent period, but still remained above the current OEL. 4. Discussion and conclusion The Chinese medical literature is a unique and important source of historical exposure information in workplaces in China. The reported data represent documented exposure levels in the workplaces in the shoemaking industry. The data indicate that benzene exposure levels from both the benzene poisoning case reports and the industrial hygiene survey reports for the shoemaking industry in China were quite high, particularly before the passage of the Occupational Diseases Prevention and Control Act and the introduction of the new OEL for benzene in 2002. The literature survey indicates that benzene levels in excess of 1000 mg/m3 were not uncommon in shoemaking facilities in China, with some were as high as 4500 mg/m3 under certain extreme circumstances. Sixty-five percent of the reported benzene measurements were above the OEL of 40 mg/m3 in effect at the time, with 20% in excess of 500 mg/m3. In addition, shoe workers typically worked 11-14 h a day and some even worked as long as 20 h daily. High benzene exposures, coupled with long working hours, have resulted in many acute and chronic benzene poisoning cases in the shoemaking industry in China. The exposure data may be useful in developing historical exposure estimates in epidemiologic studies of shoe workers. Exposure data (especially from the industrial hygiene survey reports) once stratified by job title and year of exposure, as in Table 6, can be used in constructing a SHELL-MCCLURG-060111 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 155 job-exposure matrix in retrospective exposure assessment. When applying these data to retrospective exposure assessment, the nature and the limitations of the reported data must be taken into consideration. Some of such considerations include: Based on dominant sampling methods, most of the data represent 1 minute grab samples from the work environment. If also given data on where and for how long workers perform tasks, the measurements might be used to construct a task-time weighted average to estimate full shift exposures. Some benzene poisoning investigation reports may represent re-creation of conditions that mayor may not match well with the actual routine conditions. Since most reports are mute on the survey strategy, this is hard to determine. The surveys were undertaken by a range of investigators, for a range ofreasons which were seldom mentioned in the reports. Biases introduced by the survey rationale and by possible alterations in the normal working environment limit the direct applicability of such diverse data to an epidemiologic study and job-exposure matrix. It is unknown if the submittal and publication process biases toward reporting only higher level measurements. It has often been suggested that the occupational epidemiology literature is subject to a bias where more positive than negative or equivocal studies are published. Overall, the data indicate exposure levels associated with benzene poisoning cases were generally higher than those reported in industrial hygiene surveys by a factor of more than two. Most benzene measurements were reported in aggregate as averages, and as minimums or maximums individually. Many of the data were reported without any measure of variability or sampling interval. The majority of measurements were area samples. Limitations notwithstanding, these data can provide valuable information on historical benzene exposures in the shoemaking industry in China. There were three common sampling and analytical methods used for benzene measurements in China. The earliest employed an impinger with solvent through which the air was bubbled, followed by a colorimetric analysis. The next method employed glass syringe grab sampling, with gas chromatographic analysis. This glass synnge method may be the one used for most of the data reported in the reviewed literature, but such details are often not included in the reports. The last method uses a single section charcoal tube with a twenty minute sampling period at a nominal flow rate of 200 ml/min, also with gas chromatographic analysis. Investigation of the comparability of the methods is underway, but is not ready for summary in this current report. The data also reflect temporal trend of exposure levels in relation to the growth of the industry, the type of glues or solvents used and regulatory events (new laws or OELs). SHELL-MCCLURG-060112 156 L. Wang et al. / Regulatory Toxicology and Pharmacology 46 (2006) 149-156 However, due to the limited number of early-era publica- tions, the early data are limited in number. Knowledge of factors specific to the industry sector needs to be consid- ered to help interpret the trend data. For example, the shoemaking industry in China has evolved from manual sewing processes with small production capacity in the 1950s and 1960s to large-scale assembly line facilities of the 1990s and 2000s. Although the gluing process of shoe- making started as early as the 1930s in industrialized coun- tries, the technology was not adopted in China until the late 1960s. Since then the gluing process has developed gradually and has been widely used in the shoemaking industry. Among the solvents commonly used in the shoemaking industry in China, benzene ranked first, followed closely by toluene and xylene. A number of factors contributed to the popularity of the use of benzene in the industry; includ- ing its low price, strong solvation capacity and rapid vapor- ization. These factors resulted in lower cost, higher work efficiency and better quality of the products for the industry. Because of the awareness of the toxic effects of benzene and the increasingly more stringent regulatory occupation- al exposure standards, the use of benzene in the industry in China has declined steadily over the years. Benzene-based adhesives are now banned in China and the national stan- dard for benzene in adhesives is regulated to be less than 0.5% (State General Administration, 2003). Toluene-based or other less toxic solvent-based glues or adhesives have been used to replace benzene-based glues. At the same time, control measures at workplaces have also contributed to the further reduction of benzene exposure. Fig. 3 clearly demonstrates the downward trend of benzene exposure in the shoemaking industry in the last two decades that include the township industries which account for about 70% of the overall measurements. An important milestone in occupational health in China is the passage of the Occupational Diseases Prevention and Control Act in 2002. The 2002 Act is one of the most signif- icant and comprehensive legislatures in occupational health in China (Wong, 2003b). Along with the 2002 Act, new or revised OELs were introduced. The OEL for benzene was reduced substantially. Although it may still be too early to see the full impact of the new reduced OEL, early insights are encouraging. Table 1 shows that benzene expo- sure levels reported in 2002-2004 (mean of 69.7 m g/ m 3 ) were substantially lower than that of proceeding years 1991-2001 (mean of 293.1 mg/m3). However, the data indi- cate that the current benzene exposure levels are still high, exceeding the current OEL by several fold. Although benzene-based solvents are now banned in China and, unlike in the past, pure benzene is no longer used commercially, many Chinese workers are still exposed to relatively high levels of benzene. Today, the most com- monly used Chinese-made adhesives are toluene-based, and the toluene may contain 10% or more benzene as one of the "impurities". According to a national survey, benzene was detected in the air of 76% of workplaces which used toluene- or xylene-based solvents (Yin et aI., 1994). Legislative and enforcement actions may be necessary to deal with the issue of benzene contamination in tolueneor xylene-based solvents. In summary, current benzene exposures in the shoemaking industry in China, even though substantially lower when compared to levels in the past, are still at concentrations greater than the new OEL (PC-TWA or PC-STEL) at many facilities (Liang et aI., 2003). However, additional efforts to construct full-shift exposure estimates, or verify the sample applicability to evaluating a STEL need to be undertaken before drawing conclusions about compliance to the new OEL for benzene. The significant number of references drawn upon precludes listing them all in this report. Copies of the list are available from the corresponding author on request. Acknowledgments We are grateful to the Benzene Health Research Consortium for sponsoring the Shanghai Health Study; to Dr. Jerry Rice at Georgetown University for reviewing the manuscript; and to Professor Xipeng Jin at Fudan University School of Public Health, Professor Shixing Yang at Shanghai Municipal Institute of Public Health Supervision for advising literature survey; and other team workers for the literature search. References Chen, M.S., Chan, A., 1999. China's "market economics in command": Footwear workers' health in jeopardy. International Journal of Health Services 29 (4), 793-81I. Huang, P., 2000. Advances in adhesives used in the shoe industry (in Chinese). Guangdong Chemical Engineering 1, 13-16. Liang, Y.X., Su, Z., Wu, W.A., Lu, B.Q., Fu, W.Z., Yang, L., Gu, J.Y., 2003. New trends in the development of occupational exposure limits for airborne chemicals in China. Regulatory Toxicology and Pharmacology 38, 112-123. Liang, Y.X., Wong, 0., Armstrong, T., Ye, X.B., Miao, L.Z., Zhou, YM., Wu, Q.E., Qian, H.J., Fu, H., 2005. An overview of published benzene exposure data by industry in China, 1960-2003. ChemicoBiological Interactions 153-154, 55-64. State General Administration, 2003. PR China for Quality Supervision and Inspection and Quarantine. National Standards for Adhesives used in Footwear and Suitcase Manufacturing (GB 19340-2003). China Standard Publisher, Beijing. Tao, G.Y, 2000. Some thoughts and reflections of benzene poisoning. Occupation and Health 16 (8), 67-68 (in Chinese). Wong, 0., 2002. Investigations of benzene exposure, benzene poisoning and malignancies in China. Regulatory Toxicology and Pharmacology 35 (1), 126-135. Wong, 0., 2003a. Letter from Shanghai. Occupational and Environmental Medicine 60 (3), 223. Wong, 0., 2003b. Regulation of occupational exposures in China. Regulatory Toxicology and Pharmacology 38, 109-11I. Yang, Y, 1995. A report of two cases of benzene poisoning and aplastic anemia. Occupational Medicine 22 (2), 26-27 (in Chinese). Yin, S.N., Li, G.L. (Eds.). The explanation of the development of the diagnostic criteria and management principles for occupational benzene poisoning (in Chinese). In: Toxicological and Epidemiological Studies on Benzene in China. Institute of Occupational Medicine, Chinese Academy of Preventive Medicine, Beijing, 1994. SHELL-MCCLURG-060113