Document EqLe3VryZYod0YLe1ojOXgL9n

G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS Chemico-Biological Interactions xxx (2009) xxxxxx Contents lists available at ScienceDirect Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification Otto Wong a,b,c,, Fran Harris a,d, Thomas W. Armstrong e, Fu Hua c a Applied Health Sciences, San Mateo, CA, USA b University of Hong Kong, Hong Kong, China c Fudan University School of Public Health, Shanghai, China d University of California School of Medicine, San Francisco, CA, USA e TWA8HR Occupational Hygiene Consulting, NJ, USA article info Article history: Available online xxx Keywords: Acute myeloid leukemia World Health Organization classification Epidemiology Case-control study Risk factors Environmental and occupational exposures China abstract The objectives were: (1) to investigate potential environmental and occupational risk factors of acute myeloid leukemia (AML), and (2) to explore the relationships between risk factors and AML subtypes according to the World Health Organization (WHO) classification. The investigation was a hospital-based case-control study consisting of 722 newly diagnosed AML cases (August 2003 through June 2007) and 1444 individually genderage-matched patient controls at 29 hospitals in Shanghai. A 17-page questionnaire was used to obtain information on demographics, medical history, family history, lifestyle risk factors, employment history, residential history, and occupational and non-occupational exposures. Certain occupations of interest triggered a second questionnaire, which was occupation-specific and asked for more details about jobs, tasks, materials used and work environment. Exposure assessments were based on the questionnaires, on-site workplace investigations, data published in the Chinese literature, historical exposure measurements maintained by government health agencies, and expert opinions of a panel of local scientists who were familiar with workplaces in Shanghai. Risk estimates (odds ratios and 95% confidence intervals) of individual risk factors were calculated using conditional logistic regression models. A number of potential environmental and occupational risk factors were associated with an increased risk of AML (all subtypes combined) and/or individual subtypes; including home or workplace renovation, living on a farm, planting crops, raising livestock or animals, farm workers, metal workers, rubber and plastic workers, wood and furniture workers, printers, loading and unloading workers, automobile manufacturing, general construction, and food and beverage industry (restaurants and other eateries). Exposures associated with an increased risk of AML (all subtypes combined) and/or individual subtypes included benzene, diesel fuel, metals, insecticides, fertilizers, glues and adhesives, paints and other coatings, and inks and pigments. Multivariate models were used to adjust for potential confounding exposures, and several potential risk factors were subsequently eliminated. The results of the investigation indicated that some risk factors applied to all or most subtypes (e.g., living on a farm and overall AML and several subtypes), while others to specific subtypes only (e.g., raising livestock and AML with multilineage dysplasia). Thus, some risk factors were subtype-specific. The difference in risk by subtype underscores the importance of the etiologic commonality and heterogeneity of AML subtypes. 2009 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Acute myeloid leukemia (AML) is the most common type of leukemias in the United Sates and other western countries. The estimated number of newly diagnosed AML in the US for 2008 is 13,290, Corresponding author at: Applied Health Sciences, P.O. Box 2078, San Mateo, CA 94401, USA. E-mail address: ottowong@aol.com (O. Wong). with a male-to-female ratio of 1.2:1.0 [1]. Based on the US National Cancer Institute's (NCI) Surveillance Epidemiology and End Results (SEER) data for 20012005, the adjusted annual incidence rates for males and females are 4.5 per 100,000 and 2.9 per 100,000, respectively [2]. There are some variations in incidence by ethnicity and geographical location. For example, in the US the rates for males and females classified as "Asian/Pacific Islander" are 3.7 per 100,000 and 2.5 per 100,000, respectively, which are slightly lower than those for their white counterparts (4.6 per 100,000 and 3.0 per 100,000). Generally, the incidence of AML in adults is higher in developed 0009-2797/$ see front matter 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.cbi.2009.10.017 GARABRANT/BERRYMAN 00362 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 2 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx countries than in less developed nations. Incidence rates of AML in Asian countries such as China and Japan are generally lower than those in the US and European nations [3]. Most early epidemiologic studies treated leukemia as a single diagnostic category, partly because of the lack of specific diagnostic information and partly because of the limited number of patients by subgroups of leukemia in individual studies. However, starting in the 1980s, following the clinical and pathological recognition of the heterogeneity of leukemias, epidemiologists began to appreciate the differences among the various subgroups within the broad category of malignancies collectively known as "leukemia," and an increasing number of investigations focusing on specific major subgroups of leukemia (acute and chronic myeloid leukemias, and acute and chronic lymphocytic leukemias) began to appear in the literature [4]. A previous comprehensive review has identified a large number of potential risk factors of AML reported in epidemiologic investigations; including personal and family medical histories (such as blood transfusion, alkylating drugs for cancer treatment, diagnostic X-rays, rheumatoid arthritis, and family history of blood disorders), lifestyle (such as tobacco, alcohol, and hair dyes), environmental exposures (living on a farm, living near electrical power transmission lines), occupations and industries (such as farmers, painters, shoe and leather workers, chemical workers, printers, and grain workers), and exposures to chemical, physical or biological agents (such as benzene, solvents, radiation, and retroviruses) [3]. In particular, high levels of benzene have been linked to an elevated AML risk. Some recent studies have also reported a positive association between anthropometric measurements (such as weight, height or body mass index) and lymphatic and hematopoietic malignancies including AML [58]. The findings of AML risk factors reported in epidemiologic studies, however, have not always been consistent. For example, while epidemiologic evidence generally suggests an increased risk of AML among smokers, no association between cigarette smoking and AML was found in some case-control studies [911]. Other studies have reported that cigarette smoking seems to affect certain subtypes of AML more than the others [12,13]. This observation of difference in risk by subtype underscores the fact that the diagnostic category of AML actually consists of several distinct subtypes and, hence, the need for epidemiologic studies of AML to treat these subtypes as separate diagnostic entities [3,14]. At present, only a few epidemiologic studies have systematically investigated the effects of personal, lifestyle, environmental and occupational risks of individual AML subtypes. Additional studies to investigate the etiologic commonality and heterogeneity of AML subtypes are needed. The objectives of the present study are 2-fold: (1) to investigate potential environmental and occupational risk factors of AML, and (2) to explore the relationships between risk factors and specific AML subtypes according to the new World Health Organization (WHO) classification of myeloid neoplasms [15,16]. A previous analysis of the data in the study has identified several potential risk factors associated with personal characteristics or lifestyle, including low-level education, alcohol consumption, home or workplace renovation, living on a farm, planting crops, and raising livestock or animals; whereas the use of traditional Chinese medicines was associated with a reduced risk [14]. In the current paper we will report the results of an analysis based on occupational or industrial categories and specific exposures. 2. Materials and methods We conducted a hospital-based case-control study of AML in Shanghai. The study was one of several research projects of the Shanghai Health Study (SHS) program, a collaborative research effort between investigators in the US and China. Participants of and contributors to the SHS program included Fudan University in Shanghai, Shanghai Center for Disease Control and Prevention (CDCP), Shanghai Municipal Institute of Public Health Supervision (IPHS), Shanghai District Institutes of Public Health Supervision, University of Colorado, Applied Health Sciences, ExxonMobil Biomedical Sciences, and 29 hospitals in Shanghai. The study protocols were approved by Chinese and US Institutional Review Boards of respective organizations. In designing the study, it was estimated that a sample size of approximately 500600 would provide adequate statistical power to detect a modest risk of AML (e.g., approximately 2-fold) resulting from benzene exposure (one of the chemicals of primary interest). It should be noted that even though benzene was of primary interest in any investigation of lymphohematopoietic cancers including our case-control study, other risk factors (personal, lifestyle, environmental and occupational) were also taken into consideration in the study protocol as reflected in the design of the questionnaire. Based on crude hospital admission data in Shanghai, it was anticipated that the targeted size sample could be reached in 45 years. Potential cases were defined as patients aged 18 or older and newly diagnosed with AML ("provisional diagnosis") at any of the 29 participating hospitals in Shanghai between August 2003 and June 2007. The WHO 2001 classification of hematopoietic and lymphoid neoplasms was used in the diagnosis [15]. The WHO classification system utilizes morphologic, cytogenetic, immunophenotypic, and molecular information as well as clinical features of the patients [16]. To provide the equipment and facilities needed for the WHO diagnostic procedures, a new laboratory "the Joint Sino-US Clinical and Molecular Laboratory (JCML)" was built on the campus of Fudan University in Shanghai, staffed with scientists from both Fudan University and the University of Colorado. The JCML functioned as the centralized diagnostic laboratory for the participating hospitals in Shanghai and served as the clinical arm to provide diagnostic information to the research projects in the SHS program. At each participating hospital, a designated clinical coordinator was responsible for identifying and recruiting eligible patients (i.e., patients with a provisional diagnosis of AML and aged 18 or older) for study participation. Each participant was asked to sign an informed consent according to the Declaration of Helsinki of 1975. Peripheral blood, bone marrow aspirates, tissue and core biopsies were collected in conjunction with diagnostic procedures and were sent to JCML for analysis. Details of diagnostic procedures at JCML have been described elsewhere [17,18]. The clinical coordinators were also responsible for recruiting controls. For each case, two individually matched controls were randomly selected from patients admitted to the same hospital. Patients with any malignant or non-malignant diseases of the lymphatic and hematopoietic system were excluded from control selection. Matching criteria included gender and age. For each case, the clinical coordinators at the hospitals were asked to recruit two patients of the same gender within 5 years of age of the case. For some cases, suitable controls within 5 years of age were not available and the age requirement was relaxed. Because cases and controls were enrolled around the same time, hospital admission dates of the case and controls within each matched set (matched triplet) were quite similar. To obtain relevant information from study participants, faceto-face interviews at the hospitals were conducted by trained interviewers from Fudan University School of Public Health. To minimize recall bias, neither the patients nor the interviewers were informed about the specific objectives or hypotheses of the study. Furthermore, the interviewers were not informed of the case/control status of the patients. A 17-page primary questionnaire was used to obtain information on: demographics, medical GARABRANT/BERRYMAN 00363 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx Table 1 Distribution of demographic and personal variables of acute myeloid leukemia (AML) cases and controls. Variable Number Male Female Cases 722 406 316 100.00% 56.23% 43.77% Mean age (standard deviation) in years 49.92 (16.85) Marital status Married Divorced Widowed Never married Missing data 605 83.80% 10 1.39% 35 4.85% 72 9.97% 0 0.00% Education None Primary school Middle school High school University or higher Missing data 57 7.89% 137 18.98% 227 31.44% 183 25.35% 115 15.93% 3 0.42% Body mass index (BMI) (standard deviation) 22.6 (3.3) Controls 1444 812 632 49.98 1204 15 70 147 8 60 199 448 393 340 4 23.1 3 100.00% 56.23% 43.77% (16.49) 83.38% 1.04% 4.85% 10.18% 0.55% 4.16% 13.78% 31.02% 27.22% 23.55% 0.28% (3.7) history, family history, lifestyle risk factors, employment history, residential history, and occupational and non-occupational exposures. Certain occupations of interest trigger a second questionnaire, which was occupation-specific and asked for more details about jobs, tasks, materials used and work environment. Some examples of jobs requiring a second questionnaire included shoe workers, farmers, mechanics, machinists, electricians, printers, and painters. Occupations and industries reported by the patients were coded according to the official Chinese standard classification systems [19,20]. Exposure assessment was performed by a committee consisting of local experts who were familiar with workplace exposures in Shanghai. The senior committee members included a retired official from the Shanghai Municipal IPHS who had inspected many industrial facilities in Shanghai, a professor of toxicology at Fudan University who had conducted benzene research for decades, and a professor of occupational medicine at Fudan University who served for many years as the Chair of the Chinese National Committee for Occupational Exposure Standards. The data sources utilized in exposure assessment included: the Shanghai Municipal CDCP database of historical exposure measurements taken at factories in Shanghai (consisting of 50,000+ measurements dating back to the 1950s), measurements maintained at District IPHS, measurement records and facility histories maintained at factories, on-site investigations including ad hoc walk-through surveys and measurements conducted by the exposure assessment team, reports on technology changes over time in key industrial sectors, and measurement data reported in the Chinese medical literature. For exposure assessment purposes, a study subject's employment was separated into job segments or work events. Exposure classification was carried out on a job-by-job basis. The committee members were blinded with regard to the case/control status. A US-based industrial hygienist with experience in retrospective exposure assessment provided periodic reviews and support to the Shanghai team. Because of the interest in benzene, an extensive exposure assessment was conducted. Exposure to benzene was categorized into five exposure groups (EG): EG0 (non-exposure), EG1 (<1 mg/m3), EG2 (110 mg/m3), EG3 (10100 mg/m3), and EG4 (>100 mg/m3). The ranges of these EGs were intended as a guideline for assigning jobs to these exposure categories and not meant to imply any numerical precision of the estimates. Exposure group scores (EGS) with values of 04 were assigned to the corresponding exposure groups. For other substances, a dichotomous classification of "ever/never" was used. A detailed description of the procedures and results of exposure assessment can be found elsewhere [21,22]. Data from completed questionnaires were entered (double entry) in an Oracle database at the Fudan University computer system in Shanghai, with a periodically updated backup system at the University of Colorado in Denver. The two databases were maintained and managed by the JCML staff at Fudan University and the University of Colorado. After data collection of all study subjects has been completed, relevant diagnostic and questionnaire data were extracted from the JCML database and converted to SAS files for analysis in the case-control study. Data analysis was carried out with the SAS statistical software [23]. Conditional logistic regression models taking into account the matching between cases and controls (gender and age) were used to calculate odds ratios (ORs) and 95% confidence intervals (95% CIs). Multivariate models were used to adjust for potential confounding exposures, including lifestyle and non-occupational risk factors identified in a previous analysis [14]. 3. Results Twenty-nine hospitals in the Shanghai metropolitan area participated in the case-control study. The majority of the hospitals were municipal or university-affiliated hospitals and most of the remaining ones were district hospitals. In total, these 29 hospitals covered a sizable percentage of the cancer patient population of the city. Between August 2003 and June 2007, of all the patients with a provisional diagnosis of AML referred to JCML, 741 were subsequently confirmed with a diagnosis of AML according to the WHO 2001 criteria. Nineteen patients, who were without informed consent forms or with incomplete interviews, were excluded from the case-control study. Thus, the case-control study consisted of 722 confirmed AML patients with informed consents and completed questionnaires, representing a participation rate of 97%. For every AML case, 2 individually matched controls from the same hospital were subsequently recruited, resulting in a total of 1444 control patients with a variety of diagnoses (the most frequent categories being diseases of the circulatory system, endocrine, respiratory system, digestive system and cancer). The interviewees (i.e., persons answering the questions during the interviews) could be (1) the patients themselves, (2) the patients plus family members, or (3) family members only. However, a great majority of the patients (94.60% cases and 99.03% controls) participated in the interviews (for case, 68.56% patients only and 26.04% patients plus family GARABRANT/BERRYMAN 00364 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 4 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx Table 2 Distribution of acute myeloid leukemia (AML) cases by subtype according to the WHO 2001 classification. Subtype ICD-Oa code AML with recurrent cytogenetic abnormalities AML with t(8;21)(q22;q22),(AML1/ETO) AML with inv(16)(p13q22) or t(16;16)(p13;q22),(CBF/MYH11) Acute promyelocytic leukemia (AML with t(15;17)(q22;q12),(PML/RAR) and variants) AML with 11q23 (MLL) abnormalities 9896/3 9871/3 9866/3 9897/3 Subtotal AML with multilineage dysplasia AML with multilineage dysplasia, with or without prior MDS 9895/3 AML and MDS, therapy related Alkylating agent related; Topoisomerase II inhibitor related 9920/3 AML not otherwise categorized AML, minimally differentiated AML without maturation AML with maturation Acute myelomonocytic leukemia Acute monoblastic and monocytic leukemia Acute erythroid leukemia Acute megakaryoblastic leukemia Acute basophilic leukemia Acute panmyelosis with myelofibrosis Myeloid sarcoma Subtotal 9872/3 9873/3 9874/3 9867/3 9891/3 9840/3 9910/3 9870/3 9931/3 9930/3 Acute leukemia of ambiguous lineage Acute leukemia of ambiguous lineage 9805/3 Other uncategorized cases Undifferentiated acute leukemia AML, not otherwise specified AML (total) a International Classification of Diseases for Oncology. Frequency 64 23 124 33 244 186 5 26 64 48 60 42 19 7 1 2 0 269 11 3 4 722 Percentage 8.86% 3.19% 17.17% 4.57% 33.80% 25.76% 0.69% 3.60% 8.86% 6.65% 8.31% 5.82% 2.63% 0.97% 0.14% 0.28% 0.00% 37.26% 1.52% 0.42% 0.55% 100.00% members; for controls, 89.20% patients only and 9.83% patients plus family members). The reason for a higher percentage of "patients plus family members" among the cases was that more AML patients than control patients were too weak to complete the entire interview (average 45 min) and needed assistance from family members. Table 1 presents demographic and other personal characteristics of the cases and controls. The similarities in gender and age between the cases and controls indicate that matching was successful (56.23% males in both cases and controls, and mean ages of 49.92 for cases and 49.98 for controls). The distribution of the 1444 controls by their age differences comparing with the cases was as follows: within 5 years (n = 1255 or 87%), 68 years (n = 143 or 10%), 910 years (n = 34 or 2%), and 1113 years (n = 12 or 1%). The cases and controls were remarkably similar with respect to the most common marital status; close to 84% in both groups reported as "married." Based on a comparison in Table 1, slightly more controls than cases attended high schools (27.22% vs. 25.35%) and significantly more controls than cases received university or higher education (23.55% vs. 15.93%). Several anthropometric measures were included in the questionnaire, including height and weight (before the current illness), from which BMI were calculated. The mean BMI of the controls (23.1) was slightly higher than that of the cases (22.6). The distribution of the 722 AML cases by WHO subtypes is given in Table 2. In the WHO classification, three unique subgroups are recognized: (1) AML with recurrent cytogenetic abnormalities (AML-RCA), (2) AML with multilineage dysplasia (AML-MD), and (3) therapy-related AML (t-AML) and myelodysplastic syndromes (MDS) [15,16]. Cases that do not fit into any of these three categories or no genetic data are available are classified in a fourth residual category "AML, not otherwise categorized (AML-noc)," which represents a heterogeneous group of subtypes of AML. It should be noted that some of the subtypes in the AML-noc group in the WHO classification resemble those in the previous French-AmericanBritish (FAB) classification [15,16]. There were 244 (33.80%) cases of AML-RCA, with approximately half of them (n = 124, 17.17%) being categorized as "acute promyelocytic leukemia" (APL). Classified in the category of AML-MD were 186 (25.76%) patients, whereas only five (5) patients were categorized as t-AML. In the residual category AML-noc were 269 (37.26%) cases, of which four subtypes had more than 40 patients in each: AML without maturation (n = 64, 8.86%), AML with maturation (n = 48, 6.65%), acute myelomonocytic leukemia (n = 60, 8.31%), and acute monoblastic and monocytic leukemia (n = 42, 5.82%). Table 3 shows the distribution of demographic and personal characteristics of all AML cases (AML-total) and by selected major WHO subtypes (AML-RCA, APL, AML-MD and AML-noc). The male-to-female ratio for AML-RCA was slightly higher than that for AML-total and even higher when compared with the other two major subtypes (AML-MD or AML-noc). AML-RCA patients (especially APL) were younger (mean ages of 42.56 and 40.52, respectively) than other subtypes. This was not unexpected as recurring chromosomal abnormalities tend to be more prevalent in younger age groups. On the other hand, the AML-MD patients were the oldest on the average (mean age of 55.97). Table 4 presents the risk analysis of AML-total and major subtypes for more than 30 occupational groups. These occupational groups were chosen based on a consideration of potential exposures, findings from previous studies reported in the literature and frequency in our study. A significantly increased risk of AML-total GARABRANT/BERRYMAN 00365 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx 5 Table 3 Distribution of demographic and personal variables of AML cases by selected major WHO subtypes. Variable AML-total AML-RCA APL AML-MD AML-noc Number Male Female 722 100.00% 244 100.00% 124 100.00% 186 100.00% 269 100.00% 406 56.23% 145 59.43% 70 56.45% 99 53.23% 145 53.90% 316 43.77% 99 40.57% 54 43.55% 87 46.77% 124 46.10% Mean age in years (standard deviation) 49.92 (16.85) 42.56 (14.49) 40.52 (13.43) 55.97 (16.68) 52.85 (16.45) Education None Primary School Middle school High school University or higher Missing data 57 7.89% 137 18.98% 227 31.44% 183 25.35% 115 15.93% 3 0.42% 9 3.69% 36 14.75% 77 31.56% 78 31.97% 43 17.62% 1 0.41% 4 3.23% 15 12.10% 45 36.29% 40 32.26% 20 16.13% 0 0.00% 21 11.29% 48 25.81% 50 26.88% 38 20.43% 27 14.52% 2 1.08% 27 10.04% 51 18.96% 92 34.20% 62 23.05% 37 13.75% 0 0.00% Marital status Married Divorced Widowed Never married 605 83.80% 200 81.97% 101 81.45% 155 83.33% 231 85.87% 10 1.39% 6 2.46% 3 2.42% 1 0.54% 3 1.12% 35 4.85% 4 1.64% 1 0.81% 18 9.68% 11 4.09% 72 9.97% 34 13.93% 19 15.32% 12 6.45% 24 8.92% Body mass index (BMI) (standard deviation) 22.6 (3.3) 22.9 (3.3) 23.3 (3.5) 22.8 (3.4) 22.3 (3.1) AML = acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized. was associated with the category "general farm workers, grains" (OR = 1.71, 95% CI = 1.322.22) and similar risks were observed for the subtypes as well. According to the description in the official Chinese Standard Classification of Occupations, farm workers in this category grow grains such as rice, wheat, corn, mullet, bean, potatoes, etc. [19]. A similar significant excess of risk for AML-total was also found for the broader group "general farm workers, grains and other products" (OR = 1.76, 95% CI = 1.392.22). This group includes "general farm workers, grains" and those who grow crops such as vegetables, fruits, tea, cotton, etc. "Farm workers (all types)," which include the service sector as well, experienced a similar elevated risk for AML-total (OR = 1.61, 95% CI = 1.272.03) and increased risks were noted across the major subtypes. A significant risk for AML-RCA was found for "metal workers," which included workers in smelting, metal forging, and metal heat processing (OR = 2.86, 95% CI = 1.097.51). The increased risk appeared to come from the subtype APL (OR = 14.00, 95% CI = 1.72113.77). A borderline significant risk for AML-total was observed for "rubber and plastic workers" (OR = 2.50, 95% CI = 0.996.33), and the excess came from two subtypes: AML-RCA and AML-noc. A significantly increased risk of AML-noc was associated with "wood and furniture workers" (OR = 5.09, 95% CI = 1.3519.15), but other AML subtypes were not affected. Both "printers and related workers" and "painters" were at an increased risk of AML-total (OR = 3.20 and 1.93, respectively) and major subtypes, although the ORs for some subtypes were based on relatively small numbers. Significant increases in risk were found among "loading and unloading workers" at train stations, truck depots, airports, marine terminals and warehouse loading docks for AML-total (OR = 2.11, 95% CI = 1.223.64) and for AML-RCA (OR = 3.00, 95% CI = 1.078.43). Analysis by industrial categories is presented in Table 5. A significantly increased risk of approximately 50% associated with the industrial category "crop plantation" was found for AML-total (OR = 1.54, 95% CI = 1.221.94) and for all three major WHO subtypes: AML-RCA (OR = 1.51, 95% CI = 1.012.27), AMLMD (OR = 1.77, 95% CI = 1.132.78), and AML-noc (OR = 1.49, 95% CI = 1.032.17). The slightly broader category "agriculture," which included the service sector as well, showed a similar risk increase. A significantly elevated APL risk was reported for "textile and other fabric manufacturing" (OR = 2.50, 95% CI = 1.105.66), but there was no increased risk overall for AML-total or other subtypes. Similarly, "furniture manufacturing" was a significant risk factor for AML-noc (OR = 6.99, 95% CI = 1.4533.66), but did not affect other subtypes; and "paper manufacturing" was associated with a significantly increased risk for AML-MD (OR = 3.50, 95% CI = 1.0311.96) but not with other subtypes. Other significant associations were found for the following combinations: "metal milling and processing" and AML-RCA (OR = 2.60, 95% CI = 1.145.93), "automobile manufacturing" and AML-total (OR = 2.50, 95% CI = 1.175.34), "general construction" and APL (OR = 2.28, 95% CI = 1.035.05), and "food and drink industry" and both AML-total (OR = 1.61, 95% CI = 1.022.56) and AML-RCA (OR = 2.62, 95% CI = 1.245.53). The "food and drink" industry was also associated with a non-significant risk increase of APL (OR = 2.11, 95% CI = 0.765.89). Table 6 presents the results of exposure to specific substances (ever vs. never) by major AML subtypes. Exposure to benzene was associated with a significant increase in risk of AML-total (OR = 1.43, 95% CI = 1.051.93). The subtypes most strongly related to benzene exposure were AML-RCA (OR = 1.61, 95% CI = 1.002.61) and the more specific category APL (OR = 1.95, 95% CI = 0.983.88). AMLMD appeared to be least affected by benzene exposure (OR = 1.16, 95% CI = 0.602.24). Solvents were associated with non-significant increases in all major subtypes of AML except AML-noc. In particular, exposure to solvents resulted in a borderline excess of AML-RCA (OR = 2.50, 95% CI = 0.996.33). No significant increase of any major subtypes were reported for exposures to paint thinners, toluene, xylene, or any of the petroleum fuels, with the exception of diesel fuel and AML-noc (OR = 1.92, 95% CI = 1.033.57). Increases in risk of all subtypes were found for metals (general) and metals at smelters or steel mills; with significant ORs for metals (general) and APL (OR = 2.68, 95% CI = 1.017.09), and for metals at smelter or steel mills and AML-total (OR = 2.19, 95% CI = 1.094.39) or AML-RCA (OR = 4.67, 95% CI = 1.2118.05). Exposure to insecticides resulted in elevated risks for all subtypes, with significant increases for AML-total (OR = 1.53, 95% CI = 1.162.04) and AMLnoc (OR = 1.75, 95% CI = 1.112.78). Similar patterns of increases were observed for herbicides and fertilizers. Exposures to paints and other coatings were associated with a significantly elevated OR of 3.12 (95% CI = 1.128.65) for AML-RCA and a non-significant OR of 2.94 (95% CI = 0.959.10) for APL. Glues and adhesives raised the risk in every major subtype of AML; with significant ORs of 3.96 (95% CI = 2.137.35) for AML-total, 3.50 (95% CI = 1.0311.96) for AMLMD, and 8.67 (95% CI = 2.4730.41) for AML-noc. A significant OR of 3.50 (95% CI = 1.478.34) for AML-total and a borderline significant OR of 5.00 (95% CI = 0.9725.77) for AML-noc were associated with exposures to inks and pigments. GARABRANT/BERRYMAN 00366 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 6 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx GARABRANT/BERRYMAN 00367 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 Table 4 Odds ratios and 95% confidence intervals for AML and selected major WHO subtypes by occupational categories. Occupational category AML-total AML-RCA APL AML-MD AML-noc OR (95% CI) Ca Co OR (95% C I) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co Science and engineering 0.27 (0.080.91) 3 22 0.22 (0.031.75) 1 9 0 4 0 4 0.44 (0.102.06) 2 9 researchers Engineering technicians 0.67 (0.421.04) 28 81 0.82 (0.391.76) 10 24 0.60 (0.172.18) 3 10 0.44 (0.181.04) 7 29 0.73 (0.331.62) 9 24 Health-related workers 0.82 (0.491.37) 21 51 1.00 (0.422.38) 8 16 1.33 (0.384.73) 4 6 0.92 (0.342.48) 6 13 0.50 (0.191.33) 5 20 Physicians and nurses 0.68 (0.371.25) 14 41 0.85 (0.322.27) 6 14 1.60 (0.435.96) 4 5 1.00 (0.342.93) 5 10 0.38 (0.111.29) 3 16 Financial workers 0.98 (0.671.43) 42 86 1.33 (0.712.48) 17 26 1.96 (0.864.51) 12 13 1.41 (0.593.37) 10 15 0.60 (0.321.14) 13 42 Teachers 0.81 (0.551.20) 41 98 1.29 (0.612.74) 14 23 1.16 (0.324.22) 4 7 0.55 (0.241.23) 8 28 0.70 (0.371.31) 16 43 Managers 0.79 (0.611.03) 100 241 0.76 (0.491.18) 35 87 0.78 (0.411.49) 18 43 0.96 (0.581.59) 29 60 0.71 (0.451.12) 32 84 Clerks 0.73 (0.481.10) 32 87 0.80 (0.431.49) 14 35 0.63 (0.261.49) 7 22 0.58 (0.211.59) 5 17 0.69 (0.351.36) 12 34 Police officers, firefighter and 0.89 (0.631.27) 53 117 0.97 (0.521.79) 19 39 0.94 (0.422.14) 10 21 0.59 (0.281.25) 11 34 1.00 (0.581.72) 22 44 soldiers Retail sales persons 1.14 (0.821.58) 64 114 1.09 (0.621.90) 22 41 0.86 (0.401.85) 11 25 1.08 (0.562.07) 15 28 1.32 (0.792.21) 27 42 Building workers (primarily 1.55 (0.952.55) 30 40 1.80 (0.575.69) 6 7 2.00 (0.2814.20) 2 2 1.64 (0.683.95) 9 11 1.59 (0.773.27) 15 20 janitorial) Chefs and other food 1.35 (0.802.28) 24 36 1.64 (0.683.95) 9 11 1.60 (0.435.96) 4 5 1.55 (0.514.69) 6 8 1.24 (0.503.07) 8 13 preparation workers General farm workers, grains 1.71 (1.322.22) 131 172 1.69 (1.062.69) 39 51 1.23 (0.622.46) 17 29 2.20 (1.363.57) 41 43 1.51 (0.992.30) 50 74 General farm workers, grains 1.76 (1.392.22) 184 251 1.70 (1.122.56) 56 77 1.42 (0.802.52) 28 44 2.04 (1.313.18) 56 70 1.68 (1.142.48) 69 98 and other products Farm workers (all types) 1.61 (1.272.03) 187 275 1.53 (1.022.29) 57 85 1.22 (0.692.15) 28 49 1.86 (1.192.91) 56 77 1.56 (1.072.29) 71 107 Metal workers 1.52 (0.872.66) 23 31 2.86 (1.097.51) 10 7 14.00 (1.72113.77) 7 1 0.73 (0.182.91) 3 8 1.22 (0.512.90) 9 15 Chemical workers 0.75 (0.291.92) 6 16 0.80 (0.164.12) 2 5 0.67 (0.076.41) 1 3 0.50 (0.064.47) 1 4 0.86 (0.223.32) 3 7 Rubber and plastic workers 2.50 (0.996.33) 10 8 5.99 (0.6257.52) 3 1 0 0 0 4 4.67 (1.2118.05) 7 3 Fabric workers 0.66 (0.331.31) 11 33 0.44 (0.102.06) 2 9 1 0 0.34 (0.071.59) 2 11 1.08 (0.422.78) 7 13 Leather workers 1.80 (0.734.43) 9 10 1.20 (0.295.02) 3 5 0.67 (0.076.41) 1 3 2.67 (0.6011.91) 4 3 2.00 (0.2814.20) 2 2 Fabric sewing and cutting 1.15 (0.662.00) 20 35 1.24 (0.503.07) 8 13 1.38 (0.365.29) 4 6 1.33 (0.384.73) 4 6 1.00 (0.432.34) 8 16 workers Wood and furniture workers 1.53 (0.773.04) 16 22 1.13 (0.353.70) 5 9 0.84 (0.203.49) 3 7 0.50 (0.112.36) 2 8 5.09 (1.3519.15) 9 5 Printers and related workers 3.20 (1.059.77) 8 5 3.00 (0.5017.93) 3 2 2.00 (0.1331.97) 1 1 2.00 (0.409.91) 3 3 20 Locksmiths 1.75 (0.853.59) 14 16 3.00 (0.8510.63) 6 4 4 0 0.50 (0.064.47) 1 4 1.75 (0.644.83) 7 8 Metal cutting and grinding 0.80 (0.501.28) 25 62 1.06 (0.482.34) 10 19 1.11 (0.373.32) 5 9 1.07 (0.452.52) 8 15 0.48 (0.211.12) 7 28 workers Machinery mechanics and 0.94 (0.581.53) 25 53 0.52 (0.171.60) 4 15 2.00 (0.409.91) 3 3 0.71 (0.281.79) 6 17 1.45 (0.712.98) 14 20 repairers Welders and sheet metal 1.33 (0.762.35) 21 32 0.91 (0.322.62) 6 13 2.00 (0.508.00) 4 4 1.78 (0.694.61) 8 9 1.56 (0.584.18) 7 9 workers Painters 1.93 (0.983.79) 17 18 2.80 (0.898.81) 7 5 2.00 (0.586.91) 5 5 1.50 (0.346.70) 3 4 1.61 (0.584.49) 7 9 Masonry and plastering 1.38 (0.643.00) 12 18 1.00 (0.234.35) 3 6 0.77 (0.134.48) 2 5 1.23 (0.275.75) 3 5 1.77 (0.506.24) 5 6 workers Loading and unloading workers 2.11 (1.223.64) 27 26 3.00 (1.078.43) 9 6 2.67 (0.6011.91) 4 3 1.71 (0.585.10) 6 7 1.74 (0.763.96) 11 13 Automobile and other drivers 1.36 (0.902.06) 40 60 1.48 (0.812.69) 19 26 1.60 (0.713.61) 11 14 1.36 (0.543.40) 8 12 1.20 (0.522.76) 10 17 Product and chemical testing 0.97 (0.651.46) 37 76 0.78 (0.411.50) 13 33 0.51 (0.211.27) 6 23 1.05 (0.492.27) 11 21 1.17 (0.552.49) 12 21 workers Packaging and storage workers 1.09 (0.751.60) 45 83 1.44 (0.782.66) 20 29 1.74 (0.763.96) 12 15 1.30 (0.553.07) 9 14 0.85 (0.461.58) 16 37 AML = acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized; OR = odds ratio; 95% CI = 95% confidence interval; Ca = exposed cases; Co = exposed controls. G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx GARABRANT/BERRYMAN 00368 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 Table 5 Odds ratios and 95% confidence intervals for AML and selected major WHO subtypes by industrial categories. Industrial category AML-total AML-RCA APL AML-MD AML-noc OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co Crop plantation Agriculture Electricity production and supply Cotton textile manufacturing Textile and other fabric manufacturing Garment manufacturing Shoe manufacturing Leather and leather products Furniture manufacturing Paper manufacturing Printing industry Toy manufacturing Agricultural chemicals manufacturing Organic chemicals, chemical fibers and other products Paint manufacturing and paint factories Drug manufacturing Rubber and plastic manufacturing Concrete manufacturing Glass and related product manufacturing Metal milling and processing Boiler, steam engine and other equipment Electronics manufacturing Automobile manufacturing General construction Transportation (railroad, highway, water, air) Railroad transportation Highway transportation Water transportation Domestic commercial business Food and beverage industry Beauty salon Health care facilities Hospitals Education Higher education Middle education Lower level education Military Government 1.54 (1.221.94) 198 1.57 (1.241.97) 200 1.17 (0.462.96) 7 0.81 (0.511.28) 29 0.90 (0.651.24) 62 1.12 (0.681.85) 1.75 (0.644.83) 1.71 (0.724.06) 1.64 (0.683.95) 1.38 (0.642.96) 1.83 (0.814.16) 2.00 (0.586.91) 1.50 (0.346.70) 24 7 10 9 11 11 5 3 1.27 (0.742.20) 21 1.67 (0.515.46) 5 1.50 (0.633.56) 9 0.78 (0.401.50) 13 1.88 (0.953.73) 16 0.71 (0.281.79) 6 1.40 (0.882.25) 31 0.97 (0.731.30) 80 0.62 (0.390.97) 2.50 (1.175.34) 1.17 (0.801.72) 1.25 (0.861.81) 26 15 46 49 1.00 (0.333.04) 1.46 (0.912.34) 0.82 (0.411.63) 1.05 (0.801.38) 1.61 (1.022.56) 1.00 (0.333.04) 0.89 (0.541.44) 0.92 (0.551.52) 0.76 (0.561.04) 0.57 (0.370.89) 0.67 (0.381.17) 1.20 (0.672.16) 0.79 (0.521.22) 0.84 (0.571.23) 5 30 12 94 34 5 24 23 67 28 18 20 37 39 304 304 12 70 136 43 8 12 11 16 12 5 4 33 6 12 33 17 17 45 164 82 12 80 80 10 41 29 180 43 10 54 50 167 92 52 34 89 92 1.51 (1.012.27) 61 1.55 (1.042.32) 62 0.80 (0.164.12) 2 1.77 (0.724.31) 10 1.40 (0.782.52) 21 1.06 (0.472.42) 1.00 (0.185.46) 1.71 (0.427.05) 0.33 (0.042.77) 0.33 (0.042.77) 2.67 (0.6011.91) 0.50 (0.064.47) 9 2 4 1 1 4 1 0 0.89 (0.272.89) 4 3.00 (0.5017.93) 3 0.67 (0.076.41) 1 2.38 (0.658.74) 6 1.78 (0.694.61) 8 0 2.60 (1.145.93) 13 1.05 (0.631.73) 27 0.77 (0.391.51) 2.40 (0.737.86) 1.53 (0.842.81) 0.97 (0.511.83) 12 6 21 16 0.44 (0.044.61) 0.94 (0.412.18) 0.92 (0.342.48) 1.18 (0.771.83) 2.62 (1.245.53) 1.71 (0.427.05) 1.06 (0.462.47) 1.14 (0.482.67) 0.97 (0.561.71) 0.77 (0.361.69) 0.87 (0.352.14) 1.00 (0.293.50) 1.10 (0.532.25) 0.51 (0.231.14) 1 8 6 38 17 4 9 9 23 11 8 4 16 8 93 93 5 12 31 17 4 5 6 6 3 4 2 9 2 3 6 9 8 10 52 31 5 29 33 4 17 13 66 14 5 17 16 47 27 18 8 30 30 1.21 (0.692.12) 31 55 1.77 (1.132.78) 57 82 1.49 (1.032.17) 77 121 1.21 (0.692.12) 31 55 1.77 (1.132.78) 57 82 1.54 (1.062.24) 78 121 0 1 4.00 (0.3644.11) 2 1 1.00 (0.254.00) 3 6 3.56 (0.8814.48) 6 4 0.60 (0.231.57) 6 19 0.63 (0.321.23) 13 39 2.50 (1.105.66) 15 14 0.56 (0.281.11) 12 40 0.89 (0.541.44) 29 64 1.12 (0.363.49) 1.00 (0.166.42) 1.00 (0.0911.03) 0.50 (0.064.47) 0.67 (0.076.41) 0.67 (0.076.41) 5 0 2 1 1 1 1 0 9 1.75 (0.644.83) 7 2 8.00 (0.8971.56) 4 4 2.00 (0.508.00) 4 2 1.00 (0.0911.03) 1 4 3.50 (1.0311.96) 7 3 1.00 (0.185.46) 2 31 2 0 8 0.94 (0.412.18) 8 17 1 0.67 (0.076.41) 1 3 4 1.33 (0.227.98) 2 3 2 6.99 (1.4533.66) 7 2 4 1.20 (0.295.02) 3 5 4 1.60 (0.435.96) 4 5 0 5.99 (0.6257.52) 3 1 2 20 1.33 (0.227.98) 2 3 0.86 (0.332.23) 6 14 2.00 (0.834.81) 10 10 2.00 (0.2814.20) 2 2 0 1 1.33 (0.227.98) 2 3 1.00 (0.0911.03) 1 2.73 (0.2333.00) 2 2 2.67 (0.6011.91) 4 3 1.00 (0.254.00) 3 6 2 0.12 (0.020.95) 1 15 1.09 (0.402.95) 6 11 0.86 (0.223.32) 3 0 7 3.00 (0.5017.93) 3 2 1.50 (0.346.70) 3 2 1.67 (0.515.46) 5 4 1.00 (0.185.46) 2 6 4 2.33 (0.786.94) 7 6 0.75 (0.252.21) 5 13 1.22 (0.582.55) 12 20 1.17 (0.552.49) 12 21 0.81 (0.461.42) 21 50 1.14 (0.721.82) 31 55 0.61 (0.241.58) 6 19 0.62 (0.241.58) 6 19 0.48 (0.221.09) 8 31 2.00 (0.1331.97) 1 1 5.00 (0.9725.77) 5 2 1.50 (0.346.70) 3 4 2.28 (1.035.05) 15 15 1.49 (0.703.16) 14 20 0.70 (0.341.47) 10 28 1.34 (0.603.00) 11 17 1.23 (0.572.65) 11 18 1.44 (0.782.66) 19 27 0.59 (0.057.43) 1 3 1.00 (0.185.46) 2 4 2.00 (0.2814.20) 2 2 1.00 (0.342.93) 5 10 1.40 (0.533.68) 7 10 2.00 (0.904.45) 12 12 2.00 (0.508.00) 4 4 1.33 (0.227.98) 2 3 0.60 (0.191.89) 4 13 1.18 (0.682.06) 23 40 0.90 (0.471.71) 16 35 1.08 (0.711.65) 39 73 2.11 (0.765.89) 8 8 1.18 (0.453.07) 7 12 1.13 (0.502.55) 9 16 2.73 (0.2333.00) 2 2 0 0 0.40 (0.053.42) 1 5 1.48 (0.444.95) 5 7 1.00 (0.342.93) 5 10 0.63 (0.281.42) 8 25 1.77 (0.506.24) 5 6 1.00 (0.342.93) 5 10 0.72 (0.321.64) 8 22 0.52 (0.211.30) 7 24 0.68 (0.351.31) 14 39 0.62 (0.371.02) 25 74 0.55 (0.181.68) 5 16 0.48 (0.161.48) 4 16 0.45 (0.230.89) 11 46 0.33 (0.042.77) 1 6 0.48 (0.131.76) 3 12 0.53 (0.191.48) 5 18 0 4 1.10 (0.383.19) 6 11 1.23 (0.512.97) 9 15 1.33 (0.473.78) 9 15 0.84 (0.361.93) 9 21 0.52 (0.251.08) 11 38 0.42 (0.131.34) 4 17 1.10 (0.532.30) 12 22 0.76 (0.411.40) 15 39 AML = Acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = Acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized. OR = odds ratio; 95% CI = 95% confidence interval; Ca = exposed cases; Co = exposed controls. 7 G Model CBI-6040; No. of Pages 17 8 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx GARABRANT/BERRYMAN 00369 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 Table 6 Odds ratios and 95% confidence intervals for acute myeloid leukemia and selected major WHO subtypes by occupational exposures. AML-total AML-RCA APL AML-MD AML-noc OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co OR (95% CI) Ca Co Benzene 1.43 (1.051.93) 78 113 1.61 (1.002.61) 33 43 1.95 (0.983.88) 18 20 1.16 (0.602.24) 16 Solvents 1.50 (0.952.71) 26 33 2.50 (0.996.33) 10 8 1.67 (0.515.46) 5 6 1.65 (0.634.32) 8 Paint thinners 2.00 (0.795.04) 9 9 3.33 (0.8013.95) 5 3 2.67 (0.6011.91) 4 3 1.50 (0.346.70) 3 Toluene 2.00 (0.795.04) 9 9 3.33 (0.8013.95) 5 3 1.33 (0.227.98) 2 3 2.00 (0.2814.20) 2 Xylene 1.44 (0.712.95) 13 18 2.00 (0.586.91) 5 5 1.00 (0.185.46) 2 4 2.00 (0.409.91) 3 Petroleum fuels 1.16 (0.851.59) 69 121 1.16 (0.691.95) 25 44 1.17 (0.552.49) 12 21 0.92 (0.441.88) 13 Kerosene 0.94 (0.392.25) 8 17 1.14 (0.343.90) 4 7 2.00 (0.2814.20) 2 2 0.40 (0.053.42) 1 Gasoline 1.07 (0.721.61) 40 75 0.91 (0.762.07) 16 35 1.07 (0.442.64) 8 15 1.09 (0.412.87) 7 Diesel fuel 1.23 (0.811.89) 36 59 0.83 (0.361.96) 8 19 0.79 (0.242.61) 4 10 0.75 (0.291.92) 6 Cutting and 0.86 (0.561.34) 32 73 0.75 (0.331.72) 9 23 2.00 (0.508.00) 4 4 0.55 (0.221.40) 6 lubricating oils Diesel or gasoline 0.57 (0.122.75) 2 7 1.00 (0.0911.03) 1 2 01 0 engine exhaust Metals (general) 1.47 (0.952.28) 36 50 1.49 (0.713.12) 13 18 2.68 (1.017.09) 10 8 1.50 (0.633.56) 9 Metals at smelters 2.19 (1.094.39) 17 16 4.67 (1.2118.05) 7 3 3.33 (0.8013.95) 5 3 1.20 (0.295.02) 3 and steel mills Welding 1.49 (0.743.02) 14 19 1.48 (0.444.95) 5 7 4.00 (0.7321.84) 4 2 2.00 (0.586.91) 5 Heavy metals 1.43 (0.454.50) 5 7 1.50 (0.346.70) 3 4 4.00 (0.3644.11) 2 1 0 Insecticides 1.53 (1.162.04) 106 152 1.44 (0.882.35) 34 51 1.26 (0.652.44) 18 30 1.39 (0.802.43) 28 Herbicides 1.83 (0.993.38) 21 24 1.94 (0.635.92) 7 8 0 6 2.68 (0.749.68) 6 Fertilizers 1.64 (1.232.19) 104 141 1.61 (0.952.73) 31 43 1.41 (0.702.82) 17 26 1.34 (0.792.27) 29 Wood and/or wood 1.57 (0.673.67) 10 13 1.00 (0.234.35) 3 6 1.00 (0.185.46) 2 4 0.67 (0.076.41) 1 dust Paint and other 1.71 (0.933.16) 20 24 3.12 (1.128.65) 10 7 2.94 (0.959.10) 8 6 1.20 (0.295.02) 3 coatings Glues and 3.96 (2.137.35) 31 17 2.44 (0.976.15) 11 10 1.40 (0.464.30) 6 9 3.50 (1.0311.96) 7 adhesives Inks and pigments 3.50 (1.478.34) 14 8 2.67 (0.6011.91) 4 3 0 1 3.33 (0.8013.95) 5 Fumigants, 1.64 (0.683.95) 9 11 3.00 (0.5017.93) 3 2 4.00 (0.3644.11) 2 1 0.40 (0.053.42) 1 sterilizers and preservatives Radiation 0.67 (0.076.41) 1 3 10 00 0 Electromagnetic 5.99 (0.6257.52) 3 1 2 0 1 0 0 field 28 1.38 (0.832.28) 27 40 10 0.79 (0.302.08) 6 15 4 1.00 (0.0911.03) 1 2 2 0.50 (0.064.47) 1 4 3 0.80 (0.252.55) 4 10 28 1.35 (0.832.20) 30 46 5 1.28 (0.246.89) 3 5 13 1.31 (0.682.51) 16 25 16 1.92 (1.033.57) 21 23 21 1.11 (0.592.12) 16 29 1 0.50 (0.064.47) 1 4 12 1.43 (0.71-2.91) 14 20 5 1.83 (0.635.29) 7 8 5 1.14 (0.343.90) 4 7 0 1.33 (0.227.98) 2 3 44 1.75 (1.112.78) 42 54 5 1.46 (0.593.62) 8 11 46 1.89 (1.173.03) 41 50 3 3.00 (0.8510.63) 6 4 5 1.18 (0.453.07) 7 12 4 8.67 (2.4730.41) 13 3 3 5.00 (0.9725.77) 5 5 2.50 (0.679.31) 5 2 4 2 1 01 10 AML = acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized; OR = odds ratio; 95% CI = 95% confidence interval; Ca = exposed cases; Co = exposed controls. G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx 9 Table 7 Odds ratios and 95% confidence intervals for selected combinations of acute myeloid leukemia (AML) subtypes and occupational or industrial categories and exposures. Combinations of AML subtypes and occupational risk factors OR (95% CI) p-Value Ca Co AML with t(8;21)(q22;q22),(AML1/ETO) General farm workers, grains and other products Farm workers (all types) Crop plantation Agriculture Food and beverage industry Glues and adhesives 2.26 (0.995.17) 2.26 (0.995.17) 1.97 (0.884.41) 1.97 (0.884.41) 8.58 (0.9874.91) 10.00 (1.1785.59) 0.05 0.05 0.10 0.10 0.05 0.04 16 18 16 18 16 20 16 20 52 51 AML, minimally differentiated Product and chemical testing workers 8.00 (0.8971.56) 0.06 41 AML without maturation Police officers, firefighters and soldiers Military Government Insecticides Fertilizers 0.25 (0.051.16) 0.16 (0.021.34) 0.27 (0.061.17) 2.06 (0.954.47) 2.26 (0.995.17) 0.08 0.09 0.08 0.07 0.05 2 14 1 10 2 15 16 18 15 16 AML with maturation Managers 0.32 (0.101.02) 0.05 6 24 Acute myelomonocytic leukemia Financial workers Cutting and lubricatiing oils Metal (general) Glues and adhesives 0.12 (0.020.88) 14.00 (1.72113.77) 3.50 (1.0311.96) 5.00 (0.9725.77) 0.04 0.01 0.05 0.05 1 17 71 74 52 Acute monoblastic and monocytic leukemia General farm workers, grains General farm workers, grains and other products Farm workers (all types) Crop plantation Agriculture Transportation (railroad, highway, water, air) 3.18 (0.9211.00) 6.60 (1.3731.76) 4.64 (1.2217.64) 5.00 (1.4920.25) 5.00 (1.4920.25) 4.33 (0.8222.81) 0.07 0.01 0.02 0.01 0.01 0.08 75 10 8 11 10 13 11 13 11 53 Only combinations with p < 0.10 are presented. OR = odds ratio; 95% CI = 95% confidence interval; Ca = exposed cases; Co = exposed controls. A similar analysis by occupational or industrial categories and exposures was carried out for other rarer AML subtypes. We will not present all the numerical results in their entirety, as most ORs were based on small numbers (hence, unstable) and a presentation of the ORs, 95% CIs and numbers of exposed cases and controls will take up too much space. Only ORs and 95% CIs for combinations of selected AML subtypes and risk factors with a p-value less than or equal to 0.10 (to include all significant as well as "suggestive" associations) are presented in Table 7. Elevated ORs were reported for several farm-related variables. In particular, a significantly elevated risk of acute monoblastic and monocytic leukemia was found for "general farm workers, grains and other products" (OR = 6.60, 95% CI = 1.3731.76) and "farm workers (all types)" (OR = 4.64, 95% CI = 1.2217.64). "Crop plantation" showed a significantly increased risk for acute monoblastic and monocytic leukemia (OR = 5.00, 95% CI = 1.4920.25) and a non-significant increase for AML with t(8;21)(q22;q22),(AML1/ETO). The "food and beverage" industry was associated with a borderline significant risk (OR = 8.58, 95% CI = 0.9874.91) for AML with t(8;21)(q22;q22),(AML1/ETO), but it should be noted that the number of exposed controls was only 2 and the 95% CI was extremely large. Interestingly, for the subtype AML with maturation, the data suggested a negative association for the industrial categories "military" or "government," but the underlying numbers of exposed cases were extremely small. Exposures to glues and adhesives were associated with a significant risk of AML with t(8;21)(q22;q22),(AML1/ETO), with an OR of 10.00 (95% CI = 1.1785.59), and a borderline significant increase was reported for acute myelomonocytic leukemia (OR = 5.00, 95% CI = 0.9725.77). It should be noted that these two ORs were based on small numbers of exposed controls. Finally, because of the interest in benzene and various subtypes of A ML, in addition to the results presented in Table 7, ORs (95% CIs, exposed cases:exposed controls) for other sub- types with more than 15 patients, regardless of the p-value are reported below (not shown in Table 7): 1.16 (0.562.83, 10:16) for AML with t(8;21)(q2;q22), 1.82 (0.2811.95, 3:4) for AML with inv(16)(p13q22) or t(16;16)(p13;q22), 1.33 (0.237.98, 2:3) for AML with 11q23 abnormalities, 1.00 (0.185.46, 2:4) for AML minimally differentiated, 1.00 (0.333.04, 5:10) for AML without maturation, 2.40 (0.737.86, 6:5) for AML with maturation, 1.61 (0.584.49, 7:9) for acute myelomonocytic leukemia, 1.13 (0.353.70, 5:9) for acute monoblastic and monocytic leukemia, and 2.00 (0.1331.97, 1:1) for acute erythroid leukemia. Because many study subjects were exposed to multiple risk factors, multivariate models were employed to adjust for confounding factors (personal characteristics, lifestyle, residential, employment, and occupational and non-occupational exposures). The choice of variables in the multivariate models was based on a consideration of results of individual effects of risk factors identified among the subjects in the current study, the frequencies of the variables in the current study, and findings in literature [3,4,14]. The variables included in the multivariate models are listed in Table 8, which shows the results of the multivariate analysis. Variables with a pvalue greater than 0.20 were eliminated from the models, and only ORs of variables remaining in the models are presented. Because of the interest in the association between AML and benzene exposure, the latter was always in the models regardless of the p-value. The results based on multivariate analysis for major subtypes are presented in Table 8. A similar multivariate analysis was also performed for other rarer subtypes, and significant results will be reported in the text below. The multivariate analysis in Table 8 indicated that low-level education was a risk factor for not only AML-total (OR = 1.57, 95% CI = 1.192.08) but also all major subtypes, especially AML-MD (OR = 2.01, 95% CI = 1.203.39) and AML-noc (OR = 1.64, 1.032.60). The use of traditional Chinese medicines was associated with a GARABRANT/BERRYMAN 00370 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 10 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx Table 8 Ratios (95% confidence intervals) of risk factors of acute myeloid leukemia (AML) by subtypes based on conditional multivariate logistic regression models. Risk factorsa AML-total AM-RCA APL AML-MD AML-noc Low-level education Traditional Chinese medicines Alcohol Home/workplace renovation Living on a farm Planting crops Raising livestock or animals Farm workers Metal milling and processing Benzene Solvents Toluene Xylene Petroleum fuels Kerosene Gasoline Agricultural industry Insecticides Herbicides Fertilizers Glues and adhesive Inks and pigments 1.57 (1.192.08) 0.68 (0.451.02) 1.29 (0.971.73) 1.29 (0.951.76) 1.47 (1.181.85) 1.36 (0.922.01) 1.45 (0.892.38) 1.20 (0.841.71) 2.76 (1.405.46) 2.39 (0.965.98) 1.65 (0.972.80) 1.91 (1.183.09) 1.63 (1.132.36) 2.28 (0.985.35) 1.37 (0.762.50) 2.34 (0.767.18) 1.78 (0.853.74) 2.01 (1.014.00) 1.58 (0.952.62) 1.85 (0.873.94) 2.01 (1.203.39) 0.42 (0.180.99) 1.64 (1.012.67) 2.32 (1.075.03) 0.95 (0.451.98) 0.55 (0.271.13) 2.89 (0.6013.78) 1.64 (1.032.60) 2.00 (1.193.35) 1.55 (1.072.23) 0.99 (0.511.94) 0.17 (0.012.24) 1.56 (0.842.89) 7.67 (1.9121.48) 3.86 (0.6921.48) AML = acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized. a All risk factors (variables) listed in the column were included in the multivariate models. Variables with p-value greater than 0.20 (except benzene) were eliminated from the models. Only the ORs (95% CIs) of variables remaining in the models are presented. borderline significant reduction in risk of AML-total (OR = 0.68, 95% CI = 0.451.02) and a significant risk reduction of almost 60% of AML-MD (OR = 0.42, 95% CI = 0.180.99). In contrast, the use of traditional medicines resulted in significant elevated risks of AML-noc (OR = 2.00, 95% CI = 1.193.35) and, in particular, the more specific subtype AML with maturation (OR = 2.91, 95% CI = 1.028.30, not shown in Table 8). Alcohol increased the risk of AML-noc (OR = 1.55, 95% CI = 1.072.23). Residential exposure to home (or workplace) renovation was associated with a significant increase in risk of AML-RCA (OR = 1.91, 95% CI = 1.183.09) and, more specifically, APL (OR = 2.01, 95% CI = 1.014.00). Living on a farm increased the risk of AML-total (OR = 1.47, 95% CI = 1.181.85), AML-RCA (OR = 1.63, 95% CI = 1.132.36), and AMLMD (OR = 1.64, 95% CI = 1.012.87). Although living on a farm was not associated with an increased risk of AML-noc in general, it significantly increased the risk of the specific subtype AML with maturation (OR = 3.54, 95% CI = 1.269.97, not shown in Table 8). In multivariate models, planting crops was not found to be a risk factor of AML. A significantly elevated risk of AML-MD (OR = 2.32, 95% CI = 1.075.03) was associated with raising livestock or animals. The multivariate analysis indicated that, when other variables were considered, the occupational category "farm workers" was no longer a significant risk factor for AML or subtypes. Among farmers a 5-fold increase of the specific subtype AML with t(8;21)(q22;q22) was detected (OR = 5.40, 95% CI = 0.9331.32, not shown in Table 8), although the increase was not significant. The multivariate analysis indicated that employment in the agricultural industry or exposure to agricultural chemicals (pesticides, herbicides, or fertilizers) was not related to increased risks of AML-total or major subtypes. However, a significant risk of more than 8-fold for acute monoblastic and monocytic leukemia was associated with the agricultural industry (OR = 8.56, 95% CI = 1.5746.73, not shown in Table 8). In the multivariate models exposure to benzene was no longer a significant risk factor for AML-total or the major subtypes, although non-significant increases were observed for AML-total (OR = 1.20, 95% CI = 0.841.71), AML-RCA (OR = 1.37, 95% CI = 0.762.50), and APL (OR = 1.85, 95% CI = 0.873.94). When other rarer subtypes were examined, benzene exposure was associated with a significant increase in risk of AML with t(8;21)(q22;q22), with an OR of 4.26 (95% CI = 1.0117.96, based on 10 exposed cases, not shown in Table 8). The multivariate analysis indicated that benzene exposure was not related to AML-MD (OR = 0.95, 95% CI = 0.451.98) or AML-noc (OR = 0.99, 95% CI = 0.511.94). In multivariate models, exposures to solvents, toluene, xylene, petroleum fuels, gasoline or kerosene were not related to increased risks of AML-total or any subtypes. Exposures to glues and adhesives increased the risks of AML-total by more than 2-fold (OR = 2.76, 95% CI = 1.405.46) and AML-noc by more than 7-fold (OR = 7.67, 95% CI = 1.9121.48). No other subtype was affected by exposures to glues and adhesives. Table 9 shows the analysis by benzene exposure variables: ever/never exposed, length of exposed jobs, maximum EG category, and decade of first exposure. Included in Table 9 are the major AML subtypes and AML with t(8;21)(q22;q22), because the OR for the latter was significantly elevated in the multivariate analysis. For AML-total, the increased risk appeared to concentrate in patients exposed to benzene for 10 years or less (OR = 1.99, 95% CI = 1.293.07), whereas there was a non-significant deficit among those exposed for more then 20 years (OR = 0.74, 95% CI = 0.391.39). There was a significant upward trend of risk of AMLtotal by maximum benzene exposure category (p-trend = 0.01). Among individuals in the EGS3 and 4 categories, the OR was 2.05 (95% CI = 1.053.98). In terms of first exposure, a significantly elevated OR of 4.18 (95% CI = 1.5611.15) of AML-total was reported in the group first exposed in 2000 or later. The pattern of risk by exposure variables of AML-RCA was similar to that of AML-total. Similar to AML-total, ORs of AML-RCA was significantly elevated for those who were exposed 10 years on their jobs, whose maximum exposures were in the EGS3 and 4 categories, and who were first exposed in or after 2000. APL seemed to share a similar risk pattern for length of maximum exposure (OR = 5.77 for 10 years) and time of first exposure (OR = 7.13 for first exposure in 2000 and after), but in terms of maximum exposure the highest risk was observed for the lowest EG (OR = 2.68 for EGS1). There was no pattern between maximum exposure and AML with t(8;21)(q22:q22). There was a non-significant increase of AML with t(8;21)(q22;q22) for exposure 10 years (OR = 2.17, 95% CI = 0.835.63). Of the 10 AML with t(8;21)(q22;q22) patients exposed to benzene, first exposure occurred in 2000 or later in six cases, with a significant GARABRANT/BERRYMAN 00371 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx GARABRANT/BERRYMAN 00372 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 Table 9 Odds ratios and 95% confidence intervals for acute myeloid leukemia (AML) and selected major WHO subtypes by benzene exposure variables. Variable AML-total AML-RCA AML with t(8;21)(q22;q22) APL AML-MD Benzene (ever vs. never) Length of exposed jobs (years) Non-exposed 10 years >10 to <20 years >20 years OR (95% CI) 1.43 (1.051.93) 1.00 (0.981.02) 1.00 Reference 1.99 (1.293.07) 1.44 (0.822.51) 0.74 (0.391.39) Ca Co OR (95% CI) 78 113 1.61 (1.002.61) 78 113 1.00 (0.971.03) 644 1331 1.00 Reference 43 45 2.70 (1.425.14) 21 30 1.01 (0.382.67) 14 38 0.71 (0.232.22) Ca Co OR (95% CI) 33 43 1.26 (0.562.83) 33 43 0.97 (0.901.04) 211 445 1.00 Reference 23 18 2.17 (0.835.63) 6 13 4 12 0.50 (0.064.47) Ca Co OR (95% CI) 10 16 1.95 (0.983.88) Ca Co OR (95% CI) 18 20 1.16 (0.602.24) 10 16 1.01 (0.971.05) 18 20 0.99 (0.951.03) 54 112 1.00 Reference 106 228 1.00 Reference 9 8 5.77 (1.8617.92) 13 6 1.56 (0.643.81) 0 4 0.89 (0.233.42) 3 8 1.95 (0.636.09) 1 4 0.67 (0.143.30) 2 6 0.16 (0.021.39) p-Trend = 0.84 p-Trend = 0.89 p-Trend = 0.33 p-Trend = 0.79 p-Trend = 0.59 AML-noc OR (95% CI) Ca Co Ca Co 16 28 1.38 (0.832.28) 27 40 16 28 1.00 (0.981.03) 27 40 170 344 1.00 Reference 9 12 1.35 (0.603.06) 6 6 2.00 (0.795.04) 1 10 1.02 (0.422.47) p-Trend = 0.95 242 498 10 15 99 8 16 Maximum exposure (EGS) Non-exposed EGS1 EGS2 EGS3 and 4 1.25 (1.061.46) 78 113 1.36 (1.031.80) 33 43 1.36 (0.852.17) 1.00 Reference 1.18 (0.791.76) 1.63 (0.902.94) 2.05 (1.053.98) p-Trend = 0.01 644 1331 1.00 Reference 211 445 1.00 Reference 40 70 1.55 (0.862.79) 21 28 1.00 (0.372.74) 20 25 1.05 (0.392.82) 6 12 1.00 (0.372.74) 18 18 4.00 (1.0015.99) 6 3 p-Trend = 0.03 p-Trend = 0.20 10 16 1.29 (0.881.89) 18 20 1.22 (0.891.67) 16 28 1.17 (0.911.50) 27 40 54 112 1.00 Reference 6 12 2.68 (1.146.34) 2 4 0.69 (0.143.36) 2 0 2.00 (0.409.91) p-Trend = 0.20 106 228 1.00 Reference 13 10 0.58 (0.211.65) 2 7 1.56 (0.475.15) 3 3 2.40 (0.737.86) p-Trend = 0.22 170 344 1.00 Reference 5 17 1.08 (0.532.19) 5 6 2.58 (0.966.94) 6 5 1.22 (0.433.47) p-Trend = 0.22 242 498 12 23 97 6 10 First exposure (decades) Non-exposed 1.00 Reference 644 1331 1.00 Reference 211 445 1.00 Reference 19401959 1.33 (0.543.26) 8 12 2.97 (0.5017.80) 3 2 2.00 (0.5017.80) 19601979 0.97 (0.571.62) 22 46 0.90 (0.372.19) 7 16 19801999 1.57 (1.002.46) 36 49 1.44 (0.702.94) 14 22 0.97 (0.253.81) 2000 4.18 (1.5611.15) 12 6 6.31 (1.7923.46) 9 3 11.97 (1.4499.64) 54 112 1.00 Reference 106 228 1.00 Reference 1 1 2 0 0 7 1.14 (0.343.90) 4 7 0.66 (0.231.92) 3 7 1.79 (0.694.61) 9 12 3.23 (1.188.87) 6 1 7.13 (0.7270.61) 3 1 170 344 1.00 Reference 242 498 0 4 1.69 (0.515.53) 56 5 14 1.28 (0.572.87) 10 16 11 8 1.07 (0.472.43) 9 17 0 2 5.99 (0.6257.52) 3 1 AML = acute myeloid leukemia; AML-RCA = AML with recurrent cytogenetic abnormalities; APL = acute promyelocytic leukemia; AML-MD = AML with multilineage dysplasia; AML-noc = AML not otherwise categorized; OR = odds ratio; 95% CI = 95% confidence interval; Ca = exposed cases; Co = exposed controls; EGS = exposure group score (see text). 11 G Model CBI-6040; No. of Pages 17 12 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx Table 10 Odds ratios and 95% confidence intervals for acute myeloid leukemia (all subtypes) by education level in farm workers and non-farm workers. Education level Farm workers None Primary school Middle school (reference) High school University or higher Missing OR (95% CI) 2.93 (1.515.70) 1.77 (1.122.79) 1.00 Reference 0.84 (0.451.59) 0.58 (0.181.88) p-Trend < 0.01 Ca 35 69 60 18 4 1 Co 28 78 115 41 13 0 Non-farm workers None Primary school Middle school (reference) High school University or higher Missing 1.67 (0.913.06) 1.24 (0.861.78) 1.00 Reference 0.92 (0.711.20) 0.65 (0.490.87) 22 68 167 165 111 2 32 121 333 352 327 4 p-Trend < 0.01 OR = odds ratio; 95% CI = 95% confidence interval; Ca = cases; Co = controls. OR of 11.97 (95% CI = 1.4499.64, one exposed control). Most of the benzene-exposed AML-MD patients were first exposed in 19801999 (OR = 3.23, 95% CI = 1.188.87), whereas for AML-noc a non-significant increase was observed among individuals first exposed in 19801999 (OR = 5.99, 95% CI = 0.6257.52). Finally, to investigate the potential confounding between "farming" and "education level," we performed a logistic regression analysis of AML-total by education level stratified by the occupation "farm workers" (Table 10). In both farm and non-farm workers, a clear inverse relationship between education level and AML risk was observed (p-trend < 0.01), although in non-farm workers the gradient was less steep. Thus, low-level education appeared to be an independent risk factor of AML-total in our study. 4. Discussion This study represents an attempt to investigate the relationships between environmental and occupational risk factors and AMLtotal and AML subtypes according to the WHO 2001 classification of myeloid neoplasms. The case-control study design allowed us to systematically examine a wide variety of lifestyle, environmental and occupational risk factors in relation to specific AML subtypes. Had we chosen the cohort study design, exposures would have been limited to one specific group only (such as a single industry). Furthermore, unless the cohort size is very large, the numbers of cases in most subtypes would not have been sufficient for analysis. On the other hand, a general concern with this type of case-control study is the so-called "mass significance" (or multiple comparison) problem [24]. Because of the large number of ORs calculated (for the large number of combinations of risk factors and AML subtypes), some ORs could have been "statistically significant" by chance alone. Therefore, in interpreting the results, we must take consistency into consideration and isolated findings must be viewed with caution. Certain findings need to be confirmed in future studies. In discussing our findings below, we will compare our results with those reported by other investigators in the literature. It should be noted that our study differed from previous studies of AML in a number of areas. First, we used the new WHO 2001 classification of AML, whereas most previous studies relied on the FAB classification, which was introduced more than three decades ago. In the WHO classification, the blast threshold for the diagnosis of AML was reduced from 30% to 20% blasts in the blood or marrow. In addition, patients with certain recurrent cytogenetic abnormalities, regardless of the blast percentage, are considered to have AML [16]. Therefore, the new WHO diagnostic criteria of AML are different from those used in the FAB classification, and some of the cases in our study would not have been classified as AML using the older FAB system. Second, occupations with the same titles may be quite different in terms of exposure between China and western countries. Many aspects of the same job (such as work practice, work environment and materials used) may be very different in China. At least in the past, exposures at workplaces in China have not always been closely monitored and exposure standards not been strictly enforced. For example, the national occupational exposure limits for benzene were 40 mg/m3 for 19792002 and 10 mg/m3 since 2002. Yet the Chinese medical journals are replete with reports of benzene overexposure (many in excess of 1000 mg/m3) as well as reports of overt benzene poisoning cases in recent years [2527]. According to a publication prepared by scientists at the Shanghai Municipal IPHS, the Shanghai Municipal CDCP and the Fudan University School of Public Health, based on exposure measurements taken at workplaces in Shanghai, the average benzene levels were 138.6 mg/m3 in 19651979, 120.1 mg/m3 in 19801984, and 112.9 mg/m3 in 19851989, and some 250 measurements were above 3000 mg/m3 [28]. Obviously, the exposures of, for example, painters in China and those in western countries could be markedly different. In comparing our results and those from other studies, these differences must be noted. We will first discuss our findings of "farm workers" and "education." One of the most consistent results in a previous analysis in the study was the inverse relationship between "education" and AML [14]. Based on a comparison to "middle school," the OR for "education, none" for AML-total was 2.66, whereas a significantly reduced OR was found for university or higher education (OR = 0.64). In our study, farm workers were found to have an increased risk of AMLtotal (OR = 1.61, 95% CI = 1.272.03) as well as several subtypes. Because farm workers generally had lower education, the observed increased risk of AML associated with low-level education could be an indirect association through farm workers. However, Table 10 shows that in both the farm and non-farm workers, a clear inverse relationship between education level and AML risk was observed. Thus, low-level education was an independent risk factor of AMLtotal in our study. The most consistent finding in terms of occupations or industries in our study was that for farm workers or the agricultural industry, which was in agreement with many previous studies reporting an increased risk of either leukemia in general or, more specifically, AML in the farming industry [3,29]. In a case-control study of AML in Novi Sad (Yugoslavia) and London, "farmers/gardeners" were found to have a significantly increased risk of AML (OR = 5.46, 95% CI = 1.1126.69, based on five exposed cases and five exposed controls) [29]. Similarly, a death-registry based case-control study in Nebraska reported a significant increase of AML among farmers who died at age 65 or under (OR = 1.98), but no increase for those dying after age 65 [30]. On the other hand, in a large case-control study in the US and Canada no increase of AML associated with farming was reported (OR = 0.7, 95% CI = 0.51.2) [31]. One advantage of our study was the large number of farm workers (187 AML cases and 275 controls classified as farm workers) and the 95% CIs were quite narrow. Another advantage of our study is the analysis by WHO subtypes. We are not aware of any other epidemiologic study that examined AML risk in farmers by WHO subtypes. As reported previously, a significantly increased risk of AML was also detected for "living on a farm" (OR = 1.67, 95% CI = 1.372.03) in our study in Shanghai [14]. Many of the "farm workers" in our study were actually village farmers who worked and lived on small family-owned farms in the rural areas in the large Shanghai metropolitan area or former farmers who migrated from inland farming regions of China to the coastal city. Of the 897 patients (cases and controls) with a history of farm residence in our study, GARABRANT/BERRYMAN 00373 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx 13 452 (50.39%) were also farm workers, and were thus exposed to "farm environment" through both work and residence. We were interested in separating occupational exposures from environmental exposures in farms in our study. In other words, did persons who lived on a farm but who were not farmers by occupation experience an increased risk of AML? To do so, we took out study subjects who were farm workers by profession, and re-analyzed the remaining data using a conditional logistic regression model. The new analysis yielded a significantly increased risk of AML for "living on a farm" (OR = 1.51, 95% CI = 1.121.97), which was slightly smaller than the original OR = 1.67. In the conditional logistic regression analysis, the removal of farm workers from the data created some matched triplets with either no case or no controls, and these matched triplets were excluded from the conditional logistic regression model (matched analysis). To investigate whether the exclusion of these matched triplets had an impact on the result, we also performed an unconditional logistic regression analysis with age and gender as covariates in a dataset that excluded only individual farmers (but not matched triplets). The unconditional logistic regression model showed a similar result (OR = 1.50, 95% CI = 1.151.83). This analysis demonstrated that persons living on a farm, even if they did not work as farmers, were at an increased risk of AML. In other words, "living on a farm" was an independent risk factor for AML, apart from direct occupational exposures. The multivariate analysis confirmed that "living on a farm" remained a significant risk factor for AML-total and several subtypes, even with the variable "farm workers" in the models. Our result of "living on a farm" was consistent with the finding of a large prospective cohort study in Iowa, which reported a significant risk ratio of 1.91 (95% CI = 1.193.05) associated with "living on a farm" [32]. Also at an increased risk of AML and major subtypes were several occupations; including metal workers, rubber and plastic workers, wood and furniture workers, printers, painters, and loading and unloading workers (Table 4). For example, among metal workers the risk of AML-RCA was significantly elevated (OR = 2.86, 95% CI = 1.097.51), and the risk for APL was even higher (OR = 14.00, 95% CI = 1.72113.77, wide 95% CI duly noted). Printers experienced a significantly elevated risk of AML-total (OR = 3.20, 95% CI = 1.059.77), whereas painters were at a borderline significant risk of AML-total (OR = 1.98, 95% CI = 0.983.79). There was a 5fold increased risk of AML-noc among wood and furniture workers (OR = 5.09, 95% CI = 1.3519.15), and a more than 4-fold increased risk of AML-noc among rubber and plastic workers (OR = 4.67, 95% CI = 1.2118.05). These occupations, especially printers and painters, have previously been reported to be at an increased risk of AML in the literature [3]. For example, a case-control study of AML patients in Novi Sad (Yugoslavia) and London reported a significant AML risk for painters (OR = 4.57, 95% CI = 1.2916.14) [29]. A large cohort study of Chinese workers exposed to benzene and other chemicals in a variety of industries reported an elevated relative risk of 2.2 among painters and other coating application workers [33]. A number of exposures were common in these occupations; including organic solvents, benzene, toluene, xylene, degreasers, adhesives and glues containing organic solvents; which naturally led to the suspicion that one or more of these exposures were associated with the increased risk. With respect to the current analysis of industries, several highrisk categories were identified. Some high-risk industries were consistent with the corresponding findings based on occupations. For example, the elevated risks found for "agriculture," "furniture manufacturing," "metal milling and processing" and "general construction" (which included painters) were consistent with findings from the corresponding occupational groups. However, from the exposure point of view, industrial titles were generally less specific than occupational titles. Therefore, findings based on industries may be more difficult to interpret. For example, significantly elevated risks of AML-total and AMLRCA were observed in the "food and beverage" industry, and a non-significant 2-fold excess risk of APL was found for the same industry. For AML with t(8;21)(q22;q22),(AML1/ETO), the risk was more than 8-fold (OR = 8.58, 95% CI = 0.9874.91). According to the official definition of the "food and beverage" industry, included in this category are restaurants, eateries, cafes, canteens, cafeterias, noodle shops, cold drink shops, teahouses, and bars. Persons employed in the "food and beverage" industry, regardless of differences in job duties or exposures (such as manager, bookkeepers, waiters, waitresses or servers, chefs or cooks, bakers, dish washers, janitors, drivers, security guards, electrical equipment and facility maintenance workers) were treated as a single group. Obviously, such a heterogeneous group was not specific enough from the exposure point of view. The likely causes for the increase for the "food and beverage" industry could be viral (contact with people or meat) or chemical (solvents, cooking fuels, or cleaning solutions) depending on the specific job in the industry. Exposure-specific analyses by AML subtype in this industry in future investigations are warranted. One interesting finding in our study was the more than 3-fold increased risk of AML-MD for the "paper manufacturing" industry (OR = 3.5, 95% CI = 1.0311.96), although the increase for AML-total was not significant (OR = 1.38, 95% CI = 0.642.96). No increased risk of AML overall (OR = 0.8) was reported for paper or pulp mills in the US and Canadian case-control study cited above [31]. Exposures to a variety of chemicals are encountered in the paper industry; including wood dust, wood extracts, sulfur compounds, talc, formaldehyde, combustion products, acid mists, dyes, and chlorinated organic compounds. Furthermore, maintenance workers at paper mills, such as machinist and mechanics, may be exposed to solvents, degreasers, and lubricants as well. We are not aware of any studies of pulp and paper workers reporting an increased risk of AML. It should be noted that the increased risk in our study was limited to the subtype AML-MD only and it could be a chance finding. We are not aware of any other investigations of AML-MD and the paper industry in the literature. Our result needs further investigations. To assess the risk of specific chemicals, we developed both a structured primary questionnaire to prompt recall of jobs, tasks and hazards and more than 60 second occupation-specific questionnaires to solicit detailed exposure information from individual study participants having certain occupations of interest. These occupation-specific questionnaires were derived from a system of questionnaires developed by researchers at NCI for collecting detailed occupational information in community-based case-control studies [34]. Detailed information obtained from study subjects included the specific tasks that he or she performed in a job, the frequencies and durations of tasks, the chemicals used, and a general description of the work environment. Thus, specific exposure assignment in our study was based on a consideration of the specific tasks in individual jobs of the patient and not based on generic occupational titles. A large number of previous studies examined the relationship between benzene and AML. There is little doubt that exposure to benzene above certain level can result in an increased risk of AML. The primary focus of epidemiologic studies on benzene and AML (overall) in the last two decades has been on the exposure-response relation [33,35,36]. Our study confirmed that benzene exposure was significantly associated with an increased risk of AML-total (OR = 1.43, 95% CI = 1.051.93). Furthermore, we detected a significant upward trend of AML-total with increasing maximum exposure (p-trend = 0.01), and the high exposure groups EG3 and 4 were associated with a 2-fold increase (OR = 2.05, 95% CI = 1.053.98). The excess of AML-total seemed to concentrate among individuals with the shortest exposure (10 years, OR = 1.99, 95% CI = 1.293.07) and among individuals with most GARABRANT/BERRYMAN 00374 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 14 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx recent exposure (first exposed in or after 2000, OR = 4.18, 95% CI = 1.5611.15). We also experimented with other benzene exposure variables in an effort to best characterize the pattern of exposure. The challenge was to maintain an appropriate balance between maximizing exposure information in the data on one hand and not over-quantifying the historical estimates on the other. For example, we constructed an exposure variable "cumulative EGS-years" = {EGS years of exposure} (summed over jobs and adjusted for frequency of exposure). No trend was detected for AML-total or any subtype by cumulative EGS-years (for example, p-trend = 0.38 for AML-total, and p-trend = 0.85 for AML with t(8,21)(q22;q22)). Because of the wide range of the EGS categories (especially EG3 and EG4) and the inherent limitations of such historical estimates, exposure variables based on EGS must be viewed with caution. In the multivariate analysis in our study, the OR for AMLtotal associated with benzene exposure was reduced to 1.20 (95% CI = 0.841.71). This reduction in risk of AML-total when other exposures were considered simultaneously was consistent with a previous case-control study of leukemia published in 1991 [37]. The 1991 study consisted of 217 leukemia cases (123 AML) at hospitals in several districts in Shanghai. When exposures were examined individually, a 2-fold risk of leukemia was reported (OR = 2.0, 95% CI = 1.13.7). In a multivariate model taking other risk factors into account, the leukemia risk was reduced and no longer significant (OR = 1.58, p > 0.05). Thus, in both our study and the 1991 investigation in Shanghai, the risk associated with benzene exposure for AML-total (or leukemia) diminished in magnitude when other risk factors were simultaneously considered. A more recent casecontrol study of 236 patients diagnosed with AML in Shanghai 19871989 reported a non-significant OR of 1.4 (95% CI = 0.82.3) and the risk increased to 2.9 (95% CI = 1.27.0) for exposure 15 years [38]. On the contrary, in our study, the highest risk of AMLtotal was associated with the group with the shortest length (10 years) of benzene exposure (OR = 1.99, 95% CI = 1.293.07). Few previous studies have examined the relationship between benzene exposure and different subtypes of AML, mainly because of the relatively small number of patients in individual subtype. In our study, there were three large groups: 244 of AML-RCA, 186 of AML-MD, and 269 in the heterogeneous category RCA-noc. Our analysis indicated that the group affected by benzene exposure most was AML-RCA (OR = 1.61, 95% CI = 1.002.61). The largest subgroup within the category of AML-RCA was APL (n = 124). A borderline significant risk of almost 2-fold was found between benzene exposure and APL (OR = 1.95, 95% CI = 0.983.88). However, there was no clear pattern of APL risk in relation to benzene exposure variables: the highest risks concentrated in the group with the shortest length of exposure (OR = 5.77) or the group with the lowest maximum exposure (OR = 2.68). APL in the WHO classification corresponds to M3 subtype in the FAB classification. A previous case-control study consisting of 38 APL (FAB M3) diagnosed at hospitals in Rome, Bologna and Pavia, Italy, reported a significant OR of 6.3 (95% CI = 1.331.1) among shoemakers, and the authors suggested that the increase was most likely related to the use of benzene, even though no exposure assessment was carried out [39]. Similar to the APL patients in our study, the authors of the Italian study also noted that the APL patients were younger than other AML patients. In a national study of leukemia and aplastic anemia in China, 19861988, a subtype-specific analysis based on 171 patients diagnosed with FAB M3 indicated a non-significant OR of 1.42 associated with benzene exposure [40]. The multivariate analysis in our study indicated that benzene exposure was significantly associated with the specific subtype AML with t(8;21)(q22;q22), with an OR of 4.26 (95% CI = 1.0117.96). The risk of AML with t(8;21)(q22;q22) appeared to be associated most strongly with recent first exposure in or after 2000 (OR = 11.97, 95% CI = 1.4499.64). We are not aware of any epidemiologic study of benzene exposure and the WHO subtype AML with t(8;21)(q22;q22). The translocation t(8;21)(q22;q22) is one of the most common structural aberrations in AML and is often found in patients with FAB M2 [41,42]. The Chinese study cited above reported a significant OR of 1.58 for patients diagnosed with FAB M2 (n = 193) associated with benzene exposure [40]. In a small clinical study of 43 healthy workers exposed to benzene in Shanghai, the level of t(8;21) was increased 15-fold in the 22 workers with a medium exposure >31 ppm (range 1328 ppm) when compared to non-exposed controls (n = 44) [42]. However, in a subsequent study of 57 benzene-exposed workers and 31 non-exposed controls in Tianjing, China, the same group of investigators reported that no individuals were positive with t(8;21) and attributed the negative finding to the rarity of the translocation in normal population [43]. In a case-control study of acute myeloid leukemia and smoking in the UK, results by cytogenetic subgroup were reported [13]. There were 32 AML patients with t(8;21), and the OR for smoking was significantly elevated (OR = 4.77, 95% CI = 1.7712.85). No increased risk was found for any other cytogenetic subgroup. The authors of the UK study suggested that the increased risk of AML with t(8;21) might be associated with benzene in tobacco smoke, although they cautioned that "The mechanism by which smoking may cause AML and in particular t(8;21) positive AML is far from clear." In our study, a previous analysis indicated that smoking was not associated with an elevated risk of AML with t(8;21), the OR being 1.33 (95% CI = 0.612.90) [14]. Thus, smoking did not play any significant role in AML with t(8;21) in our study. Taken as a whole, the results in our study appeared to lend some support to the hypothesis that benzene plays a role in the etiology of AML with t(8;21)(q22;q22). Additional studies focusing on benzene exposure and this specific cytogenetic subtype are needed to confirm our finding. As mentioned earlier, the subgroups in the category AML-noc in the WHO classification are similar to the FAB classification. In a hospital-based case-control study of AML in Texas, results of environmental exposures were reported by FAB subgroups [44]. No environmental exposures were associated with FAB M1 or M2, but a significant risk was reported for FAB M4 (acute myelomonocytic leukemia) associated with benzene exposure (based on the longest job) (OR = 11.4, 95% CI = 1.683.8, based on two exposed cases and one exposed control). In our study there were 60 cases of acute myelomonocytic leukemia, and no increased risk was related to benzene exposure (for example, based on multivariate analysis, OR = 0.45, 95% CI = 0.092.11, 7 exposed cases). It appeared that our study does not support the finding of acute myelomonocytic leukemia in the earlier investigation from Texas. Several previous case-control studies investigated the relationship between solvents in general and AML [9,38]. In a case-control study of AML in Sweden, a small non-significant increase was associated with exposure to solvents (OR = 1.2) [9]. In the case-control study in Shanghai cited earlier, the OR for AML associated with exposure to organic solvents was 1.1 (95% CI = 0.81.7) [38]. One of the problems in comparing results of exposure to solvents from different studies is the heterogeneous properties of the large group of chemicals known loosely as "solvents." In our study, "solvents" included mineral spirits, Stoddard solvent, VM&P naphtha, paint thinner, chlorinated solvents, perchloroethylene, trichloroethylene, toluene, xylene and other unspecific solvents. Exposure to "solvents" was not related to any significantly increased risk of AML-total or specific subtypes, although some of the ORs were elevated. The OR for AML-RCA associated with exposure to solvents was borderline significant (OR = 2.50, 95% CI = 0.996.33). When the data were analyzed in multivariate models, the OR for AML-RCA associated with solvent exposure was 2.34 (95% CI = 0.767.18). Because exposures to individual solvents are generally highly GARABRANT/BERRYMAN 00375 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx 15 correlated, several specific solvents and petroleum fuels were included in the multivariate analysis. With respect to specific solvents, neither toluene nor xylene was associated with an increased risk of AML-total or any subtype in our study. Similarly, no increased risk was found for any of the petroleum fuels. In particular, the OR for AML-total associated with gasoline exposure was 1.07 (95% CI = 0.721.61). The lack of an association between gasoline exposure and AML was consistent with a large cohort study of more than 18,000 petroleum workers exposed to gasoline in the US (mortality ratio = 1.17, 95% CI = 0.691.85) and a subsequent nested case-controls study with detailed analysis by gasoline exposure indexes (for example, for cumulative exposure, OR = 1.00) [45,46]. The relationship between exposure to pesticides and AML has been investigated in several epidemiologic studies, and the results have not been always consistent. For example, in a recent review, the summary AML rate ratio based on cohort studies of pesticide applicators was 2.14 (95% CI = 1.393.31), but the authors noted significant heterogeneity among the studies [47]. In the same review, the summary AML OR for exposure to pesticides based on casecontrol studies was 1.14 (95% CI = 0.751.73). In a case-control study of AML patients in Novi Sad, Yugoslavia and London, exposure to pesticides was not associated with AML (OR = 1.15, 95% CI = 0.472.80) [29]. In the case-control study in Shanghai cited earlier, no association was found between exposure to pesticides and AML (OR = 1.0, 95% CI = 0.52.0) [38]. On the other hand, in the Chinese national study of AML discussed earlier, a significant OR of 2.15 for FAB M2 was associated with exposure to fertilizers and insecticides [40]. In our study, when exposures were analyzed individually, exposure to insecticides (including organic phosphate, carbamates and other unspecific insecticides) increased the risk of not only AML-total (OR = 1.53, 95% CI = 1.162.04) but also AML-noc (OR = 1.75, 95% CI = 1.112.78). Exposures to herbicides or fertilizers showed a similar pattern of excess. Furthermore, other farm-related variables (such as "farm workers" and "living on a farm") were also associated with increased risks of AMLtotal and other subtypes. In the multivariate analysis, several key farm-related variables or exposures were analyzed simultaneously (Table 8). The multivariate analysis indicated that the only independent farm-related risk factor was "living on a farm." With respect to the association between exposure to fertilizers and insecticides and FAB M2 (AML with maturation in the WHO classification) in the Chinese study cited above, we did not find a similar association between insecticides and AML with maturation in our study (OR = 1.09, 95% CI = 0.402.97). To clarify the relationship between insecticides and AML-total or AML subtypes, future studies with specific information of subtype classification and types of insecticides will be needed. Residential exposure to home renovation was related to increased risks of AML-RCA (OR = 1.91, 95% CI = 1.183.09) and the more specific subtype APL (OR = 2.01, 95% CI = 1.014.00). An increased risk of acute leukemia associated with home renovation has been reported previously in China. In a case-control study of 250 acute leukemia patients in Luzhou, China, a significant OR of 2.276 was associated with home renovation [48]. In another case-control study of 122 acute leukemia (80 AML), a significant OR of 1.307 (95% CI = 1.0371.647) was found for home renovation [49]. Similarly, in a case-controls study of 114 patients diagnosed with hematological malignancies at 4 hospitals in Nanjing, China, elevated risks for hematological cancers (OR = 1.717, 95% CI = 1.2422.374) and for leukemia (OR = 1.477, 95% CI = 1.0042.172) were associated with home renovation [50]. In a case-control study of adult leukemia in Ningxia, China, a significant increase of leukemia was associated with "moving in after home renovation" (OR = 4.44, 95% CI = 1.3015.09) [51]. Thus, our result was in agreement with previous studies in China. Potential exposures from new or newly renovated homes or workplaces included a variety of chemi- cals; such as paints, adhesives, glues, solvents, preservatives, dust, treated fabrics and other building materials that might contain potentially hazardous chemicals. Activities associated with building renovations have not always been under strict regulation or monitoring in China. For example, in China benzene levels in the range of several hundred mg/m3 associated with commercial painting have been reported [21,2527]. With no mandatory regulations, the renovation of private homes could likely be even worse. Elevated levels of formaldehyde, benzene, toluene, xylene and other volatile organic chemicals (VOC) in newly renovated homes have been reported in China [52,53]. Given the elevated risk found in our investigation, exposure to new or newly renovated homes or workplaces can be a serious public health issue in China. Approximately 10% (n = 141) of the controls in our study were exposed to the risk factor "home/workplace environment," indicating a sizable number in the general population potentially at risk. As many cities in China, especially large metropolitan areas such as Shanghai, have recently undergone both economic and construction booms (both commercial buildings and residential homes), this public health problem will continue to grow unless regulations are more strictly enforced and public awareness more widely promoted. An interesting finding based on multivariate analysis in our study was the use of traditional Chinese medicines and reduced risks for AML-total (OR = 0.68, 95% CI = 0.451.02) and the subtype category AML-MD (OR = 0.42, 95% CI = 0.180.99), but a significantly increased risk of AML-noc (OR = 2.00, 95% CI = 1.193.35). Traditional Chinese medicines used by subjects in our study (longer than 1 month) included several common varieties such as bezoar of ox, chiretta and angelica root. We searched the Chinese literature but were unable to find any investigation on AML and traditional Chinese medicines. One possible explanation is that indeed some traditional Chinese medicines have a beneficial effect on certain AML subtypes such as AML-MD. Another explanation is that individuals in our patient population who regularly relied on traditional Chinese medicines might have avoided or minimized the use of some western medicines (such as penicillin and other antibiotics, and aspirin and other pain relievers), which have been linked to an increased risk of AML [3,50,54]. On the other hand, the positive association between traditional Chinese medicines and AML-noc might be a direct effect. One of the problems in evaluating the effects of medications is the potential confounding effect of the underlying diseases that prompt the use of medications. Furthermore, the term "traditional Chinese medicines" used in this investigation included a wide range of pharmacologically distinct formulations of medicinal herbs as well as animal parts. Obviously, the relationships between specific types of traditional Chinese medicines and subtypes of AML, regardless whether it is a direct or indirect one, require additional investigations. Several limitations of the study should be noted. First, in an interview-based case-control study such as ours, recall or reporting bias is always a potential concern. To minimize such bias, we used structured questionnaires. Further, in our study neither the patients nor the interviewers were informed about the specific objectives or hypotheses of the study. The interviewers were not informed of the case/control status of the patients and the exposure assessment team was also blinded. Second, in some matched sets the age requirement for the controls (within 5 years of the case) was relaxed, because of the lack of eligible controls. Thirteen percent (13%) of controls had an age difference of more than 5 years when compared to their corresponding cases. Nevertheless, the mean ages of all the cases (49.92 years) and controls (49.98) were almost identical (Table 1). In an analysis restricted to matched sets satisfying the original matching criterion (age difference <5 years), the results were similar to the results based GARABRANT/BERRYMAN 00376 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 16 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx on the entire dataset, except for some situations in which the ORs were no longer significant because of reduced sample sizes. Third, our study was hospital-based and consisted of patients from 29 hospitals in Shanghai. Thus, the patients in our study might not be a representative sample of the entire patient population in Shanghai. Moreover, potentially eligible cases (patients with a provisional diagnosis of AML) at these 29 participating hospitals were referred by the clinical coordinators at the hospitals to our clinical laboratory JCML for diagnostic confirmation. Although we believe that most eligible patients were referred to JCML, there was no practical way to validate the number. Fourth, because of the large number of combinations of risk factors by AML subtypes, a large number of risk estimates were calculated and some might be statistically significant simply by chance alone. Therefore, consistency must be taken into consideration in the interpretation of the results, and, furthermore, our findings need to be replicated by other investigations in the future. Fifth, some of the risk estimates were based on small numbers and the results need to be interpreted with caution. Finally, even though the exposure assessment was based on several sources of historical exposure data unique to the study setting, the historical benzene exposure estimates must be viewed in the context of a community case-control study. In particular, given the retrospective nature of the assessment procedure and the uncertainties of the underlying data sources, the assignments of jobs to specific exposure groups (EG) necessarily involved subjective judgment, and analysis based on EG might have been affected by misclassification. On the other hand, our study has a few strengths as well. First, the participation rate among eligible AML patients in our study was extremely high (97%). Second, unlike some previous studies of leukemia in general, our study focused on only one specific type of leukemia, AML. Third, the number of AML patients was large not only in the overall disease category (all AML subtypes combined) but also in some major subtypes of AML. Fourth, detailed diagnosis based on the new WHO classification of myeloid neoplasms was available, which allowed us to classify patients by individual AML subtypes based on a consideration of not only morphologic findings but also genetic, immunophenotypic, biologic, and clinical features of the patients. Many of the new WHO subtypes have never been investigated in an epidemiologic setting. With the large study size and the detailed diagnostic information, we were able to examine a wide spectrum of potential risk factors including personal characteristics, lifestyle, and environmental and occupational exposures by subtype according to the new WHO classification. We believe our study is the first large-scale epidemiologic investigation of AML using the WHO classification. 5. Conclusion In summary, we found several potential risk factors of AML (all subtypes combined) and/or individual subtypes associated with environmental or occupational exposures. Some risk factors applied to all or most subtypes (e.g., living on a farm and overall AML and several subtypes), while others were limited to one or two specific subtypes (e.g., raising livestock and AML with multilineage dysplasia). The difference in risk by subtype underscores the importance of the etiologic commonality and heterogeneity of AML subtypes. Conflict of interest statement Funding of the study was provided by the Benzene Health Effects Consortium. In the past O. Wong has acted as a consultant/expert to some of the member companies of the Consortium. Acknowledgments This case-control study is part of the Shanghai Health Study (SHS) program, a collaborative effort between investigators in the US and China, involving several organizations in both countries. First and foremost, we would like to express our gratitude to the 29 participating hospitals in Shanghai and the patients at these hospitals who consented to participate in our study. This study could not have been possible without the contributions to the study from these hospitals or patients. We are indebted to the Joint Sino-US Clinical and Molecular Laboratory (JCML) team for the diagnostic and questionnaire information (in particular, Dr. Richard Irons, Dr. Sherilyn Gross, Gail Jorgensen and Chen Xiaobao) and the exposure assessment team for exposure information (in particular Dr. Jin Xipeng, Dr. Liang Youxin, Zhou Yimei and Zhang Chi). We are grateful to the data collection team at the Fudan University School of Public Health (in particular, Dr. Ye Xibiao and Wang Yiying). We would like to thank our collaborators in Shanghai for their generosity in sharing their exposure data with us: the Shanghai Municipal Center for Disease Control and Prevention, the Shanghai Municipal Institute of Public Health Supervision, and other District Institutes of Public Health Supervision. We would like to express our sincere thanks to the Scientific Review Panel and the Ethics Review Panel of the SHS program for their encouragement and guidance throughout the project. Finally, we would like to express our appreciation to the Benzene Health Effects Consortium for supporting the study and to the staff at the American Petroleum Institute (in particular, Dr. Russell White) for administrative supports in coordinating our interactions with the Scientific Review Panel, the Ethics Review Panel and the sponsor. References [1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu, T. Murray, M.J. Thun, Cancer statistics, 2008, CA Cancer J. Clin. 58 (2008) 7196. [2] L.A.G. Reis, D. Melbert, M. Krapcho, D.G. Stinchcomb, N. Howlader, M.J. Horner, A. Mariotto, B.A. Miller, E.J. Feuer, S.F. Altekruse, D.R. Lewis, L. Clegg, M.P. Eisner, M. Reichman, B.K. Edwards (Eds.), SEER Cancer Statistical Review, 19752005, National Cancer Institute, Bethesda, MD, http://seer.cancer.gov/csr/19752005/, accessed March 22, 2009. [3] M.S. Linet, R.A. Cartwright, The leukemias, in: Cancer Epidemiology and Prevention, 2nd ed., Oxford University Press, 1996. [4] M.S. Linet, The Leukemias: Epidemiologic Aspects, Oxford University Press, 1985. [5] S.C. Larsson, A. Wolk, Overweight and obesity and incidence of leukemia: a meta-analysis of cohort studies, Int. J. Cancer 122 (2008) 14181421. [6] A. Engeland, S. Tretli, S. Hansen, T. Bjorge, Height and body mass index and risk of lymphohematopoietic malignancies in two million Norwegian men and women, Am. J. Epidemiol. 165 (2007) 4452. [7] K. Kasim, P. Levallois, B. Abdous, P. Auger, K.C. Johnson, Lifestyle factors and the risk of adult leukemia in Canada, Cancer Causes Control 16 (2005) 489500. [8] J.A. Ross, E. Parker, C.K. Blair, J.R. Cerhan, A.R. Folsom, Body mass index and risk of leukemia in older women, Cancer Epidemiol. Biomarkers Prev. 13 (2004) 18101813. [9] U. Flodin, M. Fredriksson, O. Axelson, B. Persson, L. Hardell, Background radiation, electrical work, and some other exposures associated with acute myeloid leukemia in a case-referent study, Arch. Environ. Health 41 (1986) 77 84. [10] G.C. Kabat, A. Augustine, J.R. Herbert, Smoking and adult leukemia: a casecontrol study, J. Clin. Epidemiol. 41 (1988) 907914. [11] M.R. Spitz, J.J. Fueger, G.R. Newell, M.J. Keating, Leukemia and cigarette smoking, Cancer Causes Control 1 (1990) 195196. [12] J.M. Pogoda, S. Preston-Martin, P.W. Nichols, R.K. Ross, Smoking and risk of acute myeloid leukemia: results from a Los Angeles County case-controls study, Am. J. Epidemiol. 155 (2002) 546553. [13] A.V. Moorman, E. Roman, R.A. Cartwright, G.J. Morgan, Smoking and the risk of acute myeloid leukemia in cytogenetic subgroups, Br. J. Cancer 86 (2002) 6062. [14] O. Wong, F. Harris, Y. Wang, H. Fu, A hospital-based case-controls study of acute myeloid leukemia in Shanghai: analysis of personal characteristics, lifestyle and environmental risk factors by subtypes of the WHO classification, Regul. Toxiocol. Pharmacol. 55 (2009) 340352. [15] E.S. Jaffe, N.L. Harris, H. Stein, J.W. Vardiman, World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Haematopoietic and Lymphoid Tissues, International Agency for Research on Cancer Press, Lyon, 2001. GARABRANT/BERRYMAN 00377 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017 G Model CBI-6040; No. of Pages 17 ARTICLE IN PRESS O. Wong et al. / Chemico-Biological Interactions xxx (2009) xxxxxx 17 [16] J.W. Vardiman, N.L. Harris, R.D. Brunning, The World Health Organization (WHO) classification of myeloid neoplasms, Blood 100 (2002) 22922302. [17] L. Bao, X. Wang, J. Ryder, M. Ji, Y. Chen, H. Chen, H. Sun, Y. Yang, X. Du, P. Kerzic, S.A. Gross, L. Yao, L. Lv, H. Fu, G. Lin, R.D. Irons, Prospective study of 174 de novo acute myelogenous leukemias according to the WHO classification: subtypes, cytogenetic features and FLT3 mutations, Eur. J. Haematol. 77 (2006) 3545. [18] S.A. Gross, X. Zhu, L. Bao, J. Ryder, A. Le, Y. Chen, X.Q. Wang, R.D. Irons, A prospective study of 728 cases of non-Hodgkin lymphoma from a single laboratory in Shanghai, China, Int. J. Hematol. 88 (2008) 165173. [19] Chinese National Bureau of Census, Standard Classification of Occupations, 1982. [20] Chinese National Bureau of Census, Standard Classification of Industries, 1982. [21] O. Wong, H. Fu, Exposure to benzene and non-Hodgkin lymphoma, an epidemi- ologic overview and an ongoing case-control study in Shanghai, Chemico-Biol. Interact. 153154 (2005) 3341. [22] T. Armstrong, Y. Liang, Y. Zhou, S. Bowes, O. Wong, H. Fu, A. Schnatter, C. Zhang, Exposure assessment for case-control and case-case epidemiology studies based in Shanghai China: summary of methods and results, Chemico-Biol. Interact., doi:10.1016/j.cbi.2009.11.008, in press. [23] SAS Institute, SAS Release Version 9.1, SAS Institute Inc., 2004. [24] O. Wong, Use of administrative databases in occupational and environmental epidemiologic research, Ann. Epidemiol. 11 (2001) 281285 (editorial). [25] O. Wong, Investigation of benzene exposure, benzene poisoning, and malignancies in China, Regul. Toxiocol. Pharmacol. 35 (2002) 126135. [26] O. Wong, Letter from Shanghai, Occup. Environ. Med. 60 (2003) 223. [27] O. Wong, Regulation of occupational exposures in China, Regul. Toxiocol. Pharmacol. 38 (2003) 109111 (commentary). [28] S. Zhu, W. Lu, S. Xue, J. Xu, Y. Jiang, Job-exposure matrix of Shanghai: the analysis of the intensity of occupational benzene exposure during 19531989, J. Environ. Occup. Med. 21 (2004) 5254 (text in Chinese, with an English abstract). [29] D. Lazarov, H.A. Waldron, D. Pejin, Acute myeloid leukemia and exposure to organic solvents a case-control study, Eur. J. Epidemiol. 16 (2000) 295301. [30] A. Blair, D.W. White, Leukemia cell types and agricultural practices in Nebraska, Arch. Environ. Health 40 (1985) 211214. [31] P.D. Terry, D.L. Shore, G.H. Rauscher, D.P. Sandler, Occupations, hobbies, and acute leukemia in adults, Leuk. Res. 29 (2005) 11171130. [32] P.J. Sinner, J.R. Cerhan, A.R. Folsom, J.A. Ross, Positive association of farm or rural residence with acute myeloid leukemia incidence in a cohort of older women, Caner Epidemiol. Biomarkers Prev. 14 (2005) 24462448. [33] R.B. Hayes, S.N. Yin, M. Dosemeci, G.L. Li, S. Wacholder, L.B. Travis, C.Y. Li, N. Rothman, R.N. Hoover, M.S. Linet, Benzene and the dose-related incidence of hematologic neoplasms in China, J. Natl. Cancer Inst. 89 (1997) 10651071. [34] P.A. Stewart, W.F. Stewart, J. Siemiatycki, E.F. Heineman, M. Dosemeci, Questionnaires for collecting detailed occupational information for community-based case controls studies, Am. Ind. Hygiene Assoc. J. 59 (1998) 3944. [35] O. Wong, Risk of acute myeloid leukemia and multiple myeloma in workers exposed to benzene, Occup. Environ. Med. 52 (1995) 380384. [36] R.A. Rinsky, R.W. Hornung, S.R. Silver, C.Y. Tseng, Benzene exposure and hematopoietic mortality: a long-term epidemiologic risk assessment, Am. J. Ind. Med. 42 (2002) 474480. [37] G.W. Lin, P.D. Lin, Q. Wang, X.J. Ding, A survey of the incidence of leukemia in the Jing'an and Jingshan districts of Shanghai and a case-control study of 217 cases of leukemia, Chin. J. Med. 14 (1991) 497500 (text in Chinese, with an English abstract). [38] O.J. Adegoke, A. Blair, X.O. Shu, M. Sanderson, F. Jin, M. Dosemeci, C.L. Addy, W. Zheng, Occupational history and exposure and the risk of adult leukemia in Shanghai, Ann. Epidemiol. 13 (2003) 485494. [39] A. Mele, M.A. Stazi, A. Pulsoni, G. Visani, B. Monarca, G. Castelli, L. Rocchi, G. Avvisati, F. Mandelli, Epidemiology of acute promyelocytic leukemia, Haematologica 80 (1995) 405408. [40] Chinese Epidemiologic Study Group (corresponding author: C.L. Yang), Analysis of risk factors of leukemia and aplastic anemia in China, Acta Acad. Med. Sin. 14 (1992) 185189 (text in Chinese, with an English abstract). [41] G. Flandrin, Classification of acute myeloid leukemia, in: Atlas of Genetics, Cytogenetics in Oncology and Haematology, 2002, http://AtlasGeneticsOncology. org/Anomalies. [42] M.T. Smith, L. Zhang, Y. Wang, R.B. Hayes, G. Li, J. Wiemels, M. Dosemeci, N. Titenko-Holland, L. Xi, P. Kolachana, S. Yin, N. Rothman, Increased translocations and aneusomy in chromosomes 8 and 21 among workers exposed to benzene, Cancer Res. 58 (1998) 21762181. [43] C.M. McHale, Q. Lan, C. Corso, G. Li, L. Zhang, R. Vermeulen, J.D. Curry, M. Shen, R. Turakulov, R. Higuchi, S. Germer, S. Yin, N. Rothman, M.T. Smith, Chromosome translocations in workers exposed to benzene, J. Natl. Cancer Inst. Monogr. 39 (2008) 7477. [44] M.M. Crane, J.E. Godwin, J.F. Annegers, M.J. Keating, Is histological subtype a marker for environmental exposures in acute myelogenous leukemia? Caner Epidemiol. Biomarkers Prev. 1 (1992) 183188. [45] O. Wong, F. Harris, T.J. Smith, Health effects of gasoline exposure. II. Mortality patterns of distribution workers in the United States, Environ. Health Perspect. 101 (Suppl. 6) (1993) 6376. [46] O. Wong, L. Trent, F. Harris, Nested case-control study of leukemia, multiple myeloma, and kidney cancer in a cohort of petroleum exposed to gasoline, Occup. Environ. Med. 56 (1999) 217221. [47] G. Van Maele-Fabry, S. Duhayon, D. Lison, A systematic review of myeloid leukemias and occupational pesticide exposure, Cancer Causes Control 18 (2007) 457478. [48] J.Y. Wang, H. Jia, Analysis of environmental risk factors of acute leukemia, Modern Prev. Med. 33 (2006) 809810 (text in Chinese, with an English abstract). [49] H. Yuan, Y. Yao, Y. Jin, Logistic regression analysis on risk factors of acute leukemia, Chin. J. Publ. Health 20 (2004) 13281329 (text in Chinese, with an English abstract). [50] Y.P. Pu, L.H. Yin, H. Yuan, Case-control study on relationship between hair dye use and risk of hematological cancers, J. Environ. Occup. Med. 20 (2003) 415417 (text in Chinese, with an English abstract). [51] W.Q. Ding, L. Bao, Z.Z. Li, X.Z. Pan, The risk factors of adult leukemia: a casecontrol study, J. Environ. Health 23 (2006) 195197 (text in Chinese, with an English abstract). [52] Y.L. Chen, C.X. Sha, J. Zhang, L.D. Tian, An investigation of indoor air pollution of organic chemicals, Chin. J. Health Inspect. 9 (2002) 8486 (in Chinese). [53] J.Z. Liu, Y.X. Tao, T.Y. Wen, Q. Cui, H. Liu, X.B. Jin, Decreasing characteristics of formaldehyde and benzene concentrations in indoor air after decoration, J. Environ. Health 19 (2002) 387388 (text in Chinese, with an English abstract). [54] X.O. Shu, Y.T. Gao, M.S. Linet, L.A. Briton, R.M.N. Gao, F. Jin, J.F. Fraumeni, Chloramphenicol use and childhood leukemia in Shanghai, Lancet 2 (1987) 934937. GARABRANT/BERRYMAN 00378 Please cite this article in press as: O. Wong, et al., A hospital-based case-control study of acute myeloid leukemia in Shanghai: Analysis of environmental and occupational risk factors by subtypes of the WHO classification, Chem. Biol. Interact. (2009), doi:10.1016/j.cbi.2009.10.017