Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Maternal Benzene Exposure during Pregnancy and Risk of Childhood Acute Lymphoblastic Leukemia: A Meta-Analysis of Epidemiologic Studies

  • Yanfeng Zhou ,

    Contributed equally to this work with: Yanfeng Zhou, Shaozun Zhang

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • Shaozun Zhang ,

    Contributed equally to this work with: Yanfeng Zhou, Shaozun Zhang

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • Zhen Li,

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • Jie Zhu,

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • Yongyi Bi,

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • YuE Bai,

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

  • Hong Wang

    Affiliation Department of Occupational and Environmental Health, School of Public Health, Wuhan University, Wuhan, P. R. China

Maternal Benzene Exposure during Pregnancy and Risk of Childhood Acute Lymphoblastic Leukemia: A Meta-Analysis of Epidemiologic Studies

  • Yanfeng Zhou, 
  • Shaozun Zhang, 
  • Zhen Li, 
  • Jie Zhu, 
  • Yongyi Bi, 
  • YuE Bai, 
  • Hong Wang



The prevalence of childhood leukemia is increasing rapidly all over the world. However, studies on maternal benzene exposure during pregnancy and childhood acute lymphoblastic leukemia (ALL) have not been systematically assessed. Therefore, we performed a meta-analysis to investigate the association between maternal solvent, paint, petroleum exposure, and smoking during pregnancy and risk of childhood ALL.


Relevant studies up to September 1st, 2013 were identified by searching the PubMed, EMBASE, Cochrane library and the Web of Science databases. The effects were pooled using either fixed or random effect models based on the heterogeneity of the studies.


Twenty-eight case-control studies and one cohort study were included for analysis, with a total of 16,695 cases and 1,472,786 controls involved. Pooled odds ratio (OR) with 95% confidence interval (CI) for ALL was 1.25 (1.09, 1.45) for solvent, 1.23 (1.02, 1.47) for paint, 1.42 (1.10, 1.84) for petroleum exposure, and 0.99 (0.93, 1.06) for maternal smoking during pregnancy. No publication bias was found in this meta-analysis and consistent results were observed for subgroup and sensitivity analyses.


Childhood ALL was associated with maternal solvent, paint, and petroleum exposure during pregnancy. No association was found between ALL and maternal smoking during pregnancy. Avoidance of maternal occupational and environmental benzene exposure during pregnancy could contribute to a decrease in the risk of childhood ALL.


In recent years, leukemia has become the most common cancer of children under the age of 15 [1]. About 75%–80% of leukemia cases are acute lymphoblastic leukemia (ALL) [2],[3], which is five times more prevalent than acute myelogenous leukemia (AML) [4][8]. The average annual incidence rate of childhood leukemia per million is 16.4 (standard deviation [SD] 13.6) in low-income countries, 36.5 (SD 11.6) in middle-income countries, and 40.9 (SD 6.1) in high-income countries [9]. The incidence of leukemia has been rising rapidly among children under 5 years old [10], [11], while declining among children aged 5–19 years [12]. A number of possible risk factors (e.g., infectious, genetic, or environmental) have been explored in an attempt to determine the etiology of childhood leukemia. To date, some risk factors for ALL including ionizing radiation, specific chromosomal and genetic abnormalities [1], [13], [14] have been well documented. The association between ALL and other risk factors, such as infectious agents, environmental factors, drug use, and chemical exposure are still unclear. [8], [14][16].

In previous studies, benzene has been identified as a carcinogen for leukemia among exposed workers [17]. However, occupational exposure to benzene directly in children has never been reported. Recently, several epidemiological studies have indicated that increasingly environmental benzene exposure is potentially a major cause of childhood leukemia. For example, Freedman et al. [18] have identified elevated risk among children whose mothers were exposed to interior house painting during the year before the children’s birth. Moreover, experimental research had demonstrated the cytotoxicity of hydroquinone on yolk sac hematopoietic stem cells was more noticeable than on bone marrow hematopoietic stem cells [19]. Therefore, it is a reasonable assumption that childhood leukemia may have occurred during the fetal stage of development. Accordingly, maternal exposure to leukemogenic factors in early pregnancy may increase hematopoietic system DNA instability, genetic susceptibility to cancer, and oncogenic lesions of the hematopoietic system during fetal stage, which ultimately results in the development of childhood leukemia in their offspring [9].

Among environmental risk factors, the impact of solvent, paint, petroleum products and cigarette smoking on the development of childhood ALL have been extensively investigated. Studies have shown that cigarette smoking is leading contributor of benzene in non-occupational environments [20][23]. Cigarettes contains a variety of carcinogenic substances including benzene and butadiene, and there are 48 µg benzene and 38.5 µg butadiene per cigarette [24]. Therefore, pregnancies exposed to the benzene are very common, particularly among pregnant mother who smoke. With regard to childhood ALL, some studies revealed a significant positive association whereas an inverse association was reported by others. The aim of our study was to perform a meta-analysis of cohort and case-control studies in a comprehensive overview of all available knowledge to provide a more thorough mathematical assessment summarizing the possible relationships between maternal solvent, paint, petroleum exposure and smoking during pregnancy and childhood ALL.

Materials and Methods

Search strategy

A comprehensive review of the literature was conducted by searching PubMed, Embase, the Cochrane library and Web of Science databases up to September 1st 2013. The search terms were: (childhood OR child OR infant) AND (leukemia OR cancer OR malignancy) AND (environmental factor OR occupational exposure OR benzene OR hydroquinone OR smoking OR paint OR petroleum OR solvent). Search strategies are shown in Table 1. Additionally, we carried out a manual search using reference lists of retrieved original articles and recent reviews. However, we did not search the grey literature.

Table 1. Search Strategy for PubMed (up to September 1st, 2013).

Study selection criteria

Studies were selected for inclusion based on the following criteria: 1) the information to estimate the relationship between benzene exposure and risk for childhood leukemia (effect size) in terms of odds ratio (OR), relative risk (RR) and hazard risk (HR) was available; 2) the study examined maternal exposure during pregnancy; 3) the study design was a cohort or case-control study; 4) current smoking during pregnancy was included; and 5) studies with children who had Down’s syndrome or any neoplastic disease were excluded. When multiple publications on the same study population were identified or study populations overlapped, the most recent and complete study was included in the meta-analysis. Furthermore, if a variable representing “any” solvent or paint or petroleum exposure was reported, these data were used. Only studies published in English were included. For reliability,two co-authors (Zhou and Zhang) reviewed all the identified citations independently, and articles met the criteria of inclusion were finally included in the present meta-analysis. Disagreements were resolved by reaching consensus between the two co-authors.

Data extraction and quality assessment

Information from studies was also extracted independently by two researchers, with disagreements resolved by consensus. The following terms of data were collected: first author’s last name, country studied, study design, study period, age range, matching factors, adjusting factors, exposure definition, the number of cases and controls, effect estimate (OR, RR, HR) with 95% CIs, and assessment of exposure. Most of the time, the effect estimate were extracted from the papers directly, however, the ORs were calculated when the paper only reported frequencies. All literatures analyzed in this meta-analysis are publicly published and are accessible in the database using the citation. The data for each literature are extracted using designed standard data extract form, which are available to all readers as required.

The Newcastle-Ottawa-Scale (NOS) was used to assess the quality of papers [25]. The NOS for cohort and case-control studies includes the following items: 1) representativeness of the exposed cohort/adequacy of case definition; 2) selection of the non-exposed cohort/representativeness of the cases; 3) ascertainment of exposure/selection of controls; 4) demonstration that outcome of interest was not present at start of study/definition of controls; 5) comparability of cohorts on the basis of the design or analysis/comparability of cases and controls on the basis of the design or analysis; 6) assessment of outcome/ascertainment of exposure; 7) sufficiency of follow-up for outcomes to occur/similarity of method of ascertainment for cases and controls; and 8) adequacy of follow-up of cohorts/non-response rate. The full score was 9 stars, and studies with scores of 0–3, 4–6, 7–9 were considered as low, moderate and high quality, respectively. Quality assessment was independently extracted by two co-authors, and any disagreements were resolved by consensus.

Statistical analysis

Meta-analysis was performed using Stata version 12.0 (Stata Corporation, College Station, Texas). ORs, RRs, and HRs with 95% corresponding CIs were used (HR and RR were directly considered as OR) to assess the association between benzene exposure and risk for childhood ALL. Heterogeneity among studies was estimated by chi-squared test and Cochran Q score (reported as I2) with corresponding P-values and the level of significance was set at P = 0.10[26]. If P<0.10 or I2≥50%, the heterogeneity was considered statistically significant, and the DerSimonian and Laird random-effect model was used; Otherwise, the Mantel-Haenszel fixed-effect model was used to calculate the pooled ORs [27].

Publication bias was assessed using the STATA procedure of ‘Metabias’, which is based on two different approaches, Begg funnel plots [28] and Egger’s tests [29]. Moreover, we conducted sensitivity analyses by sequential omission of individual studies under various contrasts to reflect the influence of the individual data to the pooled ORs and evaluate the stability of the results.


Literature search

The results of the search strategy and study selection process are presented in Figure 1. From a total of 7,384 studies identified, 952 were identified as duplicate studies and removed. By screening of titles or abstracts, 6,368 articles irrelevant to the study were excluded. After reading the full text, 20 studies were excluded as they only reported RR for all cancer or overall childhood leukemia. Six studies [30][35] were excluded because they used the same study population as one of the included studies. Five studies were excluded because they did not include data during pregnancy, one study was excluded due to low quality of NOS [36]. We also searched the reference lists of retrieved original articles and recent reviews, no additional studies were found. A total of 29 studies, including twenty eight case-control studies and [1], [2], [18], [37][61] one cohort study [62] were analyzed for this paper.

Figure 1. Literature search results.

NOS, Newcastle-Ottawa-Scale.

Study characteristics

Table 2 and Table 3 summarized the main characteristics of the selected studies for analysis. Among the twenty-nine eligible studies, eleven studies were conducted in North America, three studies in the United Kingdom (UK), three studies in France, two studies in Latin America, two studies in Italy, and two studies in Australia. Other six studies were conducted in Germany, Sweden, Israel, Spain, Greece, United States, respectively Across the studies, twenty-seven reported OR, one study reported RR [51], and one reported HR [62]. Children’s age across the studies was under 15 years of old, except for one study by Sorahan et al. which was under 16 years old. The study period ranged from one year to 20 years. The studies used a variety of factors to adjust their analyses including age, sex, birth weight, study region, maternal age/education/race, family income.

Table 2. Characteristics of Studies on the Association Between Maternal Solvent, Paint, Petroleum Exposure During Pregnancy and Risk of ALL.

Table 3. Characteristics of Studies on the Association Between Maternal Smoking During Pregnancy and Risk of ALL.

Based on the quality assessment of NOS, sixteen studies were in high quality (six studies scored 8 and ten studies scored 7) and thirteen were in moderate quality (nine studies scored 6 and four studies scored 5).

Main analysis

As shown in Figure 2, no statistically significant heterogeneity across studies was found in solvent and petroleum exposure (I2 = 33.0%, P = 0.176; I2 = 14.1%, P = 0.322, respectively), however significant heterogeneity was observed in paint (I2 = 47.6%, P = 0.076). The results in the meta-analysis showed an increased risk of ALL in those who were exposed to solvent or petroleum or paint (OR = 1.25, 95%CI = 1.09–1.45, fixed effect model; OR = 1.42, 95%CI = 1.10–1.84, fixed effect model; OR = 1.23, 95%CI = 1.02–1.47, random effect model, respectively).

Figure 2. Forest plot of the association between maternal exposure during pregnancy and risk of childhood ALL.

The size of each box indicates the relative weight of each study in the meta-analysis; the bars show the 95% confidence intervals (CIs). For paint, the weights are from random effects analysis because there are heterogeneity among studies (P<0.1).

Results for maternal smoking during pregnancy are shown in Figure 3. No significant associations were observed between maternal smoking during pregnancy and risk for childhood ALL (OR = 0.99, 95%CI = 0.93–1.06) in random effects model with heterogeneity (I2 = 41.9%, P = 0.032) among these studies.

Figure 3. Forest plot of the association between maternal smoking during pregnancy and risk of childhood ALL.

Considering study type may be the source of heterogeneity (P = 0.032), studies are divided into two subgroups (seventeen case-control studies (P = 0.134) and one cohort study). For studies of Abadi-Korek et al. [48] and Ferreira et al. [45], Crude odds ratios are calculated from the data provided.

Subgroup analysis, Sensitivity analysis and Publication bias

The results for subgroup analysis of childhood ALL incidence and maternal smoking during pregnancy are presented in Table 4. Consistent results were observed across the subgroup analyses including study period before or after 2000, sample size greater than or less than 1000, study region in Europe or other region, children’s age under five or not, study quality, maternal education, maternal age, household income, and ethnicity groups. In sensitivity analysis, after excluding the studies which potentially affected the pooled results no significant changes were observed (data not shown).

Table 4. Summary of the results for childhood ALL in relation to maternal smoking during pregnancy: subgroup analyses.

Using Egger’s and Begg’s test, no publication bias was observed in this meta-analysis (seen in Figure 4), for solvent (Egger’s test: P = 0.996; Begg’s test: P = 0.764), paint (Egger’s test: P = 0.533; Begg’s test: P = 0.879), petroleum (Egger’s test: P = 0.378; Begg’s test: P = 0.448), and smoking (Egger’s test: P = 0.836; Begg’s test: P = 0.649).

Figure 4. Funnel plot of the association between maternal exposure and the risk of childhood ALL.


Our meta-analysis included data from 29 studies with 16,695 cases and 1,472,786 controls. Based on the present meta-analysis, the overall data demonstrated that the ALL was significantly associated with maternal solvent, paint, or petroleum exposure. For maternal smoking during pregnancy, the same conclusion has been concluded for risk of childhood ALL in previous meta-analysis (OR = 1.03, 95% CI: 0.95–1.12) [13]. However, Different with the previous meta-analysis which only examined the association between maternal smoking and childhood ALL, this study focused the exposure of solvent, paint and petroleum exposure for children ALL. In addition, we updated the literatures in this meta-analysis, assessed the study quality we analyzed, and performed subgroup analysis in this paper.

Exposures to solvent and paint have been suggested as potential risk factors of childhood leukemia [1]. A study of cancer patients in Montréal, Canada estimated that 40% of workers had been occupationally exposed to at least one solvent over their careers [55]. Another study indicated that home exposure to solvents might also be associated with childhood ALL [18]. However, one challenge for the study with parental occupation and childhood ALL is that maternal exposure during pregnancy was rarely assessed [57], and the exposure strength at home is low. Mothers who worked as a painter or were exposed to paints were consistently associated with childhood leukemia [57]. Paint consists of various compositions, but we are not sure whether they influence the identified genetic “hits” or influence the progression of the disease via other means. The toxicity of maternal paint intake may differ according to the kinds of paints. One of the two major groups of paints is latex paints, in which the solvent is water. The other is the alkyd paints or oil-based resin paints, in which the solvent is usually petroleum-based and organic, such as toluene or xylene [1]. Nevertheless, avoiding the use of paint in the house during pregnancy and early childhood would be a prudent measure [1].

Petroleum is another main source of environmental benzene exposure. Studies have shown that childhood leukemia is associated with various exposures to petroleum. A study by Swaen et al. indicated leukemia mortality was associated with gasoline consumption, number of household cars, distance to a gasoline station, and distance to main road as evaluation indexes of petroleum [63]. Considering that the exposure dose among children is difficult to assess, in this meta-analysis, we restricted maternal occupational exposure to petroleum products as the index for children leukemia.

Several studies specifically evaluated childhood ALL risk associated with maternal smoking during pregnancy. However, only three case–control studies [36], [49], [51] reported significant association for children ALL risk (OR: 2.20, 1.40 and 1.24, respectively). In addition, the sole cohort study [62] conducted in Sweden reported a non-statistically significant association in the same population at a later time period than the preceding case–control study [36], with the proportion of maternal heavy smoking during pregnant three times lower than that reported in the earlier Swedish study [36]. Additionally, there are also studies suggesting that tobacco exposure induces in vivo fragile-site expression, which contributes to tumor formation [64]. Although non-statistically significant associations were shown between maternal smoking during pregnancy and ALL in this meta-analysis, parental cigarette smoking should also be strictly prohibited. It is suggested that parental smoking affects many childhood diseases, such as respiratory tract infection, asthma and otitis media, which are much more prevalent than childhood leukemia [45].

To detect the source of heterogeneity, we conducted subgroup analysis by region, year, child’s age, sample size of each study, high or moderate study quality and whether the effect size had adjusted by maternal education, maternal age, household income or ethnicity groups. Both fixed and random effect model were used for each subgroup. No heterogeneity was found across these subgroup variables, and the result of each subgroup was similar to the overall effect size. However, we did find the heterogeneity for types of study design and low quality study in our preliminary analysis. Therefore, we analyzed case-control studies and cohort studies separately and excluded the one low quality study and no heterogeneity was observed in this meta-analysis.

Our study may have the following limitations. First, interviewer bias: 1) The data in most studies we analyzed were collected by personal interview [1], [2], [18], [37][39], [42], [43], [45], [49], [51], [53], [56], [58], [60], [62] or independent telephone interviews [44], [47], [48], [50], [52], [54], [55], [57], [59], [61]. Several studies’ data were collected by self-administered questionnaire [40], [46] or ascertained from the birth certificates [41]. The participants’ attitudes and their understanding of the questions may have differed across populations [65]; 2) interviewers could not be blinded to case-control status, one study used different interviewers for case and control parents [51]. Second, recall bias: 1) parents were asked for the exposure occurred many years ago, which may not be remembered clearly; 2) parents who had a child with leukemia might tend to recall greater or perhaps distorted levels of exposure. Furthermore, we only collected articles published in English, which could bring publication bias, despite there being no significant evidence of publication bias observed using Egger’s test or Begg’s test. In addition, not all of the studies adjusted for family income, while socioeconomic status has been associated with the risk of childhood leukemia [66]. For example, indoor house painting was more common among high-income controls, which suggests that a selection bias. Finally, study design might have influenced the results since only one cohort study was included.

Despite the limitations, there are several strengths of this study. First, twenty-nine studies cover a broad region and long study period. Second, all the eligible studies had high or moderate research design quality. Third, no publication bias was observed based on Egger’s test and Begg’s test. Finally, our eligible literatures had comprehensive matching factors and adjusting factors, thus reducing the corresponding errors.


Based on the present meta-analysis, we concluded that maternal solvent, paint, petroleum exposure during pregnancy are associated with ALL. Avoidance of maternal benzene exposure during pregnancy might contribute to a decrease in the risk of childhood ALL.

This meta-analysis suggested the directions for future study are: 1) conduct large sample cohort studies or high quality case–control studies; 2) investigate new risk factors for childhood leukemia, such as drug use, traffic exhaust, alcohol consumption, infection, etc.; 3) explore the exposure during pregnancy, as well as before pregnancy and childhood exposure; 4) explore dose-response relationships; 5) assess gene-environment interaction, as the mechanism of childhood leukemia is likely to associate with gene–environment interactions [67].

Supporting Information


Sincere acknowledgement to Dr. Guanmin Chen, University of Calgary for his scientific review. Sincere acknowledgement to professor Yuanan Lu from the University of Hawaii and professor Guoxun Chen from the University of Tennessee for revision of manuscript. The authors also extend a sincere thank-you to all of the researchers who kindly responded to their requests for information, including Catherine Metayer, Mike Crommie, Lucia Miligi and Hsu-HueiWeng.

Author Contributions

Conceived and designed the experiments: HW. Performed the experiments: HW YFZ SZZ. Analyzed the data: YFZ SZZ ZL JZ Y. Bai. Contributed reagents/materials/analysis tools: YFZ SZZ ZL JZ Y. Bai. Wrote the paper: HW YFZ SZZ Y. Bi.


  1. 1. Scelo G, Metayer C, Zhang L, Wiemels JL, Aldrich MC, et al. (2009) Household exposure to paint and petroleum solvents, chromosomal translocations, and the risk of childhood leukemia. Environ Health Perspect 117: 133–139.
  2. 2. Infante-Rivard C, Mur P, Armstrong B, Alvarez-Dardet C, Bolumar F (1991) Acute lymphoblastic leukaemia among Spanish children and mothers’ occupation: a case-control study. J Epidemiol Community Health 45: 11–15.
  3. 3. Pui CH, Sandlund JT, Pei D, Rivera GK, Howard SC, et al. (2003) Results of therapy for acute lymphoblastic leukemia in black and white children. JAMA 290: 2001–2007.
  4. 4. Gurney JG, Severson RK, Davis S, Robison LL (1995) Incidence of cancer in children in the United States. Sex-, race-, and 1-year age-specific rates by histologic type. Cancer 75: 2186–2195.
  5. 5. Parkin DM, Stiller CA, Draper GJ, Bieber CA (1988) The international incidence of childhood cancer. Int J Cancer 42: 511–520.
  6. 6. Pui CH, Relling MV, Downing JR (2004) Acute lymphoblastic leukemia. N Engl J Med 350: 1535–1548.
  7. 7. Zipf T, Berg S, Roberts W, Poplack D, Steuber C, et al.. (2000) Childhood leukemias. Clinical Oncology (Abeloff MD, Armitage JO, Lichter AS, Niederhuber JE, eds) 2nd ed Philadelphia: Churchill Livingston: 2402–2429.
  8. 8. Belson M, Kingsley B, Holmes A (2007) Risk factors for acute leukemia in children: a review. Environ Health Perspect 115: 138–145.
  9. 9. Howard SC, Metzger ML, Wilimas JA, Quintana Y, Pui CH, et al. (2008) Childhood cancer epidemiology in low-income countries. Cancer 112: 461–472.
  10. 10. Kinlen LJ (1995) Epidemiological evidence for an infective basis in childhood leukaemia. Br J Cancer 71: 1–5.
  11. 11. Law GR, Parslow RC, Roman E (2003) Childhood cancer and population mixing. Am J Epidemiol 158: 328–336.
  12. 12. Linabery AM, Ross JA (2008) Trends in childhood cancer incidence in the U.S. (1992–2004). Cancer 112: 416–432.
  13. 13. Klimentopoulou A, Antonopoulos CN, Papadopoulou C, Kanavidis P, Tourvas AD, et al. (2012) Maternal smoking during pregnancy and risk for childhood leukemia: a nationwide case-control study in Greece and meta-analysis. Pediatr Blood Cancer 58: 344–351.
  14. 14. Buffler PA, Kwan ML, Reynolds P, Urayama KY (2005) Environmental and genetic risk factors for childhood leukemia: appraising the evidence. Cancer Invest 23: 60–75.
  15. 15. Eden T (2010) Aetiology of childhood leukaemia. Cancer Treat Rev 36: 286–297.
  16. 16. Pyatt D, Hays S (2010) A review of the potential association between childhood leukemia and benzene. Chemico-Biological Interactions 184: 151–164.
  17. 17. Pyatt D, Hays S (2010) A review of the potential association between childhood leukemia and benzene. Chem Biol Interact 184: 151–164.
  18. 18. Freedman DM, Stewart P, Kleinerman RA, Wacholder S, Hatch EE, et al. (2001) Household solvent exposures and childhood acute lymphoblastic leukemia. Am J Public Health 91: 564–567.
  19. 19. Zhu J, Wang H, Yang S, Guo L, Li Z, et al. (2013) Comparison of toxicity of benzene metabolite hydroquinone in hematopoietic stem cells derived from murine embryonic yolk sac and adult bone marrow. PLoS One 8: e71153.
  20. 20. Wallace L (1996) Environmental exposure to benzene: an update. Environ Health Perspect 104 Suppl 6: 1129–1136.
  21. 21. Wallace L, Pellizzari E, Hartwell TD, Perritt R, Ziegenfus R (1987) Exposures to benzene and other volatile compounds from active and passive smoking. Arch Environ Health 42: 272–279.
  22. 22. McNabola A, Broderick B, Johnston P, Gill L (2006) Effects of the smoking ban on benzene and 1,3-butadiene levels in pubs in Dublin. J Environ Sci Health A Tox Hazard Subst Environ Eng 41: 799–810.
  23. 23. Wallace LA (1989) Major sources of benzene exposure. Environ Health Perspect 82: 165–169.
  24. 24. Adam T, Mitschke S, Streibel T, Baker RR, Zimmermann R (2006) Quantitative puff-by-puff-resolved characterization of selected toxic compounds in cigarette mainstream smoke. Chem Res Toxicol 19: 511–520.
  25. 25. Stang A (2010) Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 25: 603–605.
  26. 26. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327: 557–560.
  27. 27. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7: 177–188.
  28. 28. Begg CB, Mazumdar M (1994) Operating characteristics of a rank correlation test for publication bias. Biometrics 50: 1088–1101.
  29. 29. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629–634.
  30. 30. Shu XO, Linet MS, Steinbuch M, Wen WQ, Buckley JD, et al. (1999) Breast-feeding and risk of childhood acute leukemia. J Natl Cancer Inst 91: 1765–1772.
  31. 31. Trivers KF, Mertens AC, Ross JA, Steinbuch M, Olshan AF, et al. (2006) Parental marijuana use and risk of childhood acute myeloid leukaemia: a report from the Children’s Cancer Group (United States and Canada). Paediatr Perinat Epidemiol 20: 110–118.
  32. 32. Metayer C, Zhang L, Wiemels JL, Bartley K, Schiffman J, et al. (2013) Tobacco smoke exposure and the risk of childhood acute lymphoblastic and myeloid leukemias by cytogenetic subtype. Cancer Epidemiol Biomarkers Prev 22: 1600–1611.
  33. 33. Edraki M, Rambod M (2011) Parental smoking and risk of childhood cancer: hospital-based case-control study in Shiraz. East Mediterr Health J 17: 303–308.
  34. 34. Bailey HD, Milne E, de Klerk NH, Fritschi L, Attia J, et al. (2011) Exposure to house painting and the use of floor treatments and the risk of childhood acute lymphoblastic leukemia. Int J Cancer 128: 2405–2414.
  35. 35. McKinney PA, Fear NT, Stockton D (2003) Parental occupation at periconception: findings from the United Kingdom Childhood Cancer Study. Occup Environ Med 60: 901–909.
  36. 36. Stjernfeldt M, Berglund K, Lindsten J, Ludvigsson J (1992) Maternal smoking and irradiation during pregnancy as risk factors for child leukemia. Cancer Detect Prev 16: 129–135.
  37. 37. Metayer C, Zhang L, Wiemels JL, Bartley K, Schiffman J, et al. (2013) Tobacco Smoke Exposure and the Risk of Childhood Acute Lymphoblastic and Myeloid Leukemias by Cytogenetic Subtype. Cancer Epidemiol Biomarkers Prev 22: 1600–1611.
  38. 38. MacArthur AC, McBride ML, Spinelli JJ, Tamaro S, Gallagher RP, et al. (2008) Risk of childhood leukemia associated with parental smoking and alcohol consumption prior to conception and during pregnancy: the cross-Canada childhood leukemia study. Cancer Causes Control 19: 283–295.
  39. 39. Magnani C, Pastore G, Luzzatto L, Terracini B (1990) Parental occupation and other environmental factors in the etiology of leukemias and non-Hodgkin’s lymphomas in childhood: a case-control study. Tumori 76: 413–419.
  40. 40. Menegaux F, Ripert M, Hemon D, Clavel J (2007) Maternal alcohol and coffee drinking, parental smoking and childhood leukaemia: a French population-based case-control study. Paediatr Perinat Epidemiol 21: 293–299.
  41. 41. Okcu MF, Goodman KJ, Carozza SE, Weiss NS, Burau KD, et al. (2002) Birth weight, ethnicity, and occurrence of cancer in children: a population-based, incident case-control study in the State of Texas, USA. Cancer Causes Control 13: 595–602.
  42. 42. Pang D, McNally R, Birch JM (2003) Parental smoking and childhood cancer: results from the United Kingdom Childhood Cancer Study. Br J Cancer 88: 373–381.
  43. 43. Petridou E, Trichopoulos D, Kalapothaki V, Pourtsidis A, Kogevinas M, et al. (1997) The risk profile of childhood leukaemia in Greece: a nationwide case-control study. Br J Cancer 76: 1241–1247.
  44. 44. Rudant J, Menegaux F, Leverger G, Baruchel A, Lambilliotte A, et al. (2008) Childhood hematopoietic malignancies and parental use of tobacco and alcohol: the ESCALE study (SFCE). Cancer Causes Control 19: 1277–1290.
  45. 45. Ferreira JD, Couto AC, Pombo-de-Oliveira MS, Koifman S (2012) Pregnancy, maternal tobacco smoking, and early age leukemia in Brazil. Front Oncol 2: 151.
  46. 46. Milne E, Greenop KR, Scott RJ, Bailey HD, Attia J, et al. (2012) Parental prenatal smoking and risk of childhood acute lymphoblastic leukemia. Am J Epidemiol 175: 43–53.
  47. 47. Slater ME, Linabery AM, Spector LG, Johnson KJ, Hilden JM, et al. (2011) Maternal exposure to household chemicals and risk of infant leukemia: a report from the Children’s Oncology Group. Cancer Causes Control 22: 1197–1204.
  48. 48. Abadi-Korek I, Stark B, Zaizov R, Shaham J (2006) Parental occupational exposure and the risk of acute lymphoblastic leukemia in offspring in Israel. J Occup Environ Med 48: 165–174.
  49. 49. Menegaux F, Steffen C, Bellec S, Baruchel A, Lescoeur B, et al. (2005) Maternal coffee and alcohol consumption during pregnancy, parental smoking and risk of childhood acute leukaemia. Cancer Detect Prev 29: 487–493.
  50. 50. Shu XO, Ross JA, Pendergrass TW, Reaman GH, Lampkin B, et al. (1996) Parental alcohol consumption, cigarette smoking, and risk of infant leukemia: a Childrens Cancer Group study. J Natl Cancer Inst 88: 24–31.
  51. 51. Sorahan T, Lancashire RJ, Hulten MA, Peck I, Stewart AM (1997) Childhood cancer and parental use of tobacco: deaths from 1953 to 1955. Br J Cancer 75: 134–138.
  52. 52. Brondum J, Shu XO, Steinbuch M, Severson RK, Potter JD, et al. (1999) Parental cigarette smoking and the risk of acute leukemia in children. Cancer 85: 1380–1388.
  53. 53. Chang JS, Selvin S, Metayer C, Crouse V, Golembesky A, et al. (2006) Parental smoking and the risk of childhood leukemia. Am J Epidemiol 163: 1091–1100.
  54. 54. Schuz J, Kaletsch U, Meinert R, Kaatsch P, Michaelis J (2000) Risk of childhood leukemia and parental self-reported occupational exposure to chemicals, dusts, and fumes: results from pooled analyses of German population-based case-control studies. Cancer Epidemiol Biomarkers Prev 9: 835–838.
  55. 55. Infante-Rivard C, Siemiatycki J, Lakhani R, Nadon L (2005) Maternal exposure to occupational solvents and childhood leukemia. Environ Health Perspect 113: 787–792.
  56. 56. McKinney PA, Raji OY, van Tongeren M, Feltbower RG (2008) The UK Childhood Cancer Study: maternal occupational exposures and childhood leukaemia and lymphoma. Radiat Prot Dosimetry 132: 232–240.
  57. 57. Reid A, Glass DC, Bailey HD, Milne E, Armstrong BK, et al. (2011) Parental occupational exposure to exhausts, solvents, glues and paints, and risk of childhood leukemia. Cancer Causes Control 22: 1575–1585.
  58. 58. Miligi L, Benvenuti A, Mattioli S, Salvan A, Tozzi GA, et al. (2013) Risk of childhood leukaemia and non-Hodgkin’s lymphoma after parental occupational exposure to solvents and other agents: the SETIL Study. Occup Environ Med 70: 648–655.
  59. 59. Shu XO, Stewart P, Wen WQ, Han D, Potter JD, et al. (1999) Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia in offspring. Cancer Epidemiol Biomarkers Prev 8: 783–791.
  60. 60. Castro-Jimenez MA, Orozco-Vargas LC (2011) Parental exposure to carcinogens and risk for childhood acute lymphoblastic leukemia, Colombia, 2000–2005. Prev Chronic Dis 8: A106.
  61. 61. Slater ME, Linabery AM, Blair CK, Spector LG, Heerema NA, et al. (2011) Maternal prenatal cigarette, alcohol and illicit drug use and risk of infant leukaemia: a report from the Children’s Oncology Group. Paediatr Perinat Epidemiol 25: 559–565.
  62. 62. Mucci LA, Granath F, Cnattingius S (2004) Maternal smoking and childhood leukemia and lymphoma risk among 1,440,542 Swedish children. Cancer Epidemiol Biomarkers Prev 13: 1528–1533.
  63. 63. Swaen GM, Slangen JJ (1995) Gasoline consumption and leukemia mortality and morbidity in 19 European countries: an ecological study. Int Arch Occup Environ Health 67: 85–93.
  64. 64. Glover TW (1998) Instability at chromosomal fragile sites. Recent Results Cancer Res 154: 185–199.
  65. 65. Chen C, Xu T, Chen J, Zhou J, Yan Y, et al. (2011) Allergy and risk of glioma: a meta-analysis. Eur J Neurol 18: 387–395.
  66. 66. Poole C, Greenland S, Luetters C, Kelsey JL, Mezei G (2006) Socioeconomic status and childhood leukaemia: a review. Int J Epidemiol 35: 370–384.
  67. 67. Lightfoot TJ, Roman E (2004) Causes of childhood leukaemia and lymphoma. Toxicol Appl Pharmacol 199: 104–117.