Women encounter organic solvents in many occupational settings, including various industries and health and personal services.1 These solvents are also present in household products, cosmetics, and drugs,2 and they contaminate waste sites and water resources.3 Many women in a wide variety of occupational and nonoccupational circumstances may thus be exposed to these compounds or their by-products.
Several solvents are developmental toxicants for animals.4 Epidemiologic studies have reported both increased risks––and, less frequently, no excess risk––of congenital malformations (mainly oral clefts, neural tube defects, and cardiac defects) among children of women working in occupations that include exposure to solvents.5–16 The evidence nonetheless remains inconclusive because most of these studies did not identify the specific chemicals used. Moreover, because most studies of congenital malformations have collected exposure information only after the pregnancy outcome was known, recall bias cannot be ruled out.
We therefore organized a cohort of pregnant women to investigate the hypothesis suggested by two of our team’s case–control studies11,15––that prenatal exposure to oxygenated glycol ethers and chlorinated solvents affects the risk of birth defects, especially oral clefts. In a previous publication based on this cohort, we reported an increased risk of congenital malformations with increasing level of exposure to solvents during pregnancy assessed by indirect methods (self-report and based on a job-exposure matrix).17 Here, we present the results of a nested case–control analysis within the same cohort that incorporated urinary biomarkers of solvent exposure.
The PELAGIE cohort included pregnant women from the general population of three districts in Brittany (northwestern France), recruited at prenatal care visits before 19 weeks’ gestation by 39 private and hospital practitioners, about 30% of all those in the study area. Participants provided informed consent, and the INSERM (National Institute of Health and Medical Research) ethics committee approved the study procedures. At inclusion, women completed a questionnaire about their sociodemographic characteristics, education, present occupation, occupational exposure to solvent-containing products, medical and obstetric history, dietary habits, and alcohol and tobacco use. They sent the completed questionnaire to our laboratory, together with a first-morning-void urine sample in a vial containing an acidic stabilizer. The samples were frozen at −20°C until analysis (eAppendix 1, http://links.lww.com/EDE/A610).
From April 2002 through February 2006, 3421 women returned the inclusion questionnaire before 19 weeks of gestation (estimated participation rate = 80%). Pregnancy outcomes (delivery or medical termination, length of gestation, newborn’s anthropometric measurements, and health status) were obtained from maternity hospital records for 3399 pregnancies (99%).
Assessment of Congenital Malformations
Staff pediatricians of the maternity wards (or neonatology or pediatric critical care units) completed forms reporting the results of their routine clinical examination for congenital malformations in live-born infants, together with specific questions about oral clefts and male genital anomalies. Malformations in fetal deaths or medically indicated abortions were determined by pathology and karyotype investigations. Mentions of male genital anomalies were later validated by surgery reports within 2 years of follow-up.
Ninety-seven major malformations (in 94 fetuses or newborns) were identified and grouped according to the European Registration of Congenital Anomalies guidelines (eTable 1, http://links.lww.com/EDE/A610).18 All male genital malformations requiring surgical repair (including undescended testes and glandular hypospadias) were considered major malformations.
Solvent Exposure Assessment
Solvent exposure at work was defined by (1) the self-administered inclusion questionnaire, which asked about the last occupation held since pregnancy began and the frequency of contact at work with 11 classes of products considered to contain solvents (exposure to solvents: none, occasional, regular) and (2) use of a job-exposure matrix19 combining occupation and industrial activity (exposure to solvents: no, medium, high) (eAppendix 2, http://links.lww.com/EDE/A610). The inclusion questionnaire asked about nonoccupational exposure to solvent-containing products during hobbies in the 3 months before inclusion.
Nested Case-control Study for Urinary Chemical Analyses
In a nested case–control study, chemical analyses of samples from all 79 cases of nonchromosomal nongenetic major malformations were compared with those from 580 controls randomly drawn from the 3269 newborns without malformations in the cohort. Ten solvent urinary metabolites were measured by gas chromatography mass spectrometry.20,21 We focused on eight alkoxycarboxylic acids as the main urinary metabolites of glycol ethers (methoxyacetic acid [MAA], ethoxyacetic acid [EAA], butoxyacetic acid [BAA], n-propoxyacetic acid [PAA], phenoxyacetic acid [PhAA], methoxy ethoxyacetic acid [MEAA], ethoxy ethoxyacetic acid [EEAA], and 2-methoxypropionic acid [2-MPA]), and trichloroacetic acid (TCAA) and trichloroethanol (TCOH), the main urinary metabolites of tetrachloroethylene and trichloroethylene (eAppendix 1, http://links.lww.com/EDE/A610).
Urinary measurements of alkoxycarboxylic acids were obtained in 73 cases (92%) and 580 controls. The presence of hydrochloric acid (HCl) in some samples, used as a stabilizer at the beginning of the study, prevented determination of TCAA and TCOH levels for 22 cases (28%) and 121 controls (21%). TCAA and TCOH measurements were therefore available for 51 cases and 459 controls.
We studied the influence on urinary measurements of sampling conditions, including transportation time at room temperature, duration of storage at −20°C, and type of acidic stabilizer (HCl or nitric acid [HNO3]), and of individual parameters (urinary creatinine level and gestational age at inclusion). This investigation included an experiment studying the stability of the metabolites for various values of transportation time (up to 10 days), temperature (room temperature, 4°C), and acidic stabilizer, and an a posteriori analysis of the influence of sampling conditions and individual parameters on urinary levels of solvent metabolites (eAppendix 1, http://links.lww.com/EDE/A610).
The BAA detection rate decreased with the number of days at ambient temperature when HCl (but not HNO3) was used as a stabilizer. Therefore, results for BAA are presented only for women whose samples were stabilized with HNO3.
First, using logistic regression adjusted for maternal age at inclusion, tobacco and alcohol use at inclusion, folic acid supplementation, and educational level, we estimated the odds ratios (ORs) of major congenital malformations (and their confidence intervals [CIs]) associated with indirect assessment of exposure to solvents (self report, job-exposure matrix) in the nested case–control study. This analysis was restricted to working women.
Each metabolite with a detection rate less than 50% was categorized in two classes (<limit of detection [LOD] and ≥LOD). Other metabolites were categorized in three classes (<25th percentile, 25th–74th percentile, ≥75th percentile). We then estimated ORs and CIs of congenital malformations associated with the urinary level of each metabolite, using logistic regression models including the above-listed adjustment factors plus district of residence, year of inclusion, and the sampling condition covariates potentially associated with detection in urine. Additional potential risk factors were preterm birth (for the study of male genital malformations) and parity, sex, oligohydramnios, and fetal presentation (for the analysis of limb defects). The small number of cases led us to select covariates according to a criterion of at least a 10% change in the OR estimate.
Correlations between the urinary metabolites were studied using chi-square or Fisher’s exact tests or Spearman rank correlation depending on the level of detection. Additionally, we studied associations between metabolites detected (for those identified in <50%) or levels (for those identified in >50%), and both products handled at work and indirect methods of exposure assessment using logistic regression or Tobit regression. SAS software version 9.2 (SAS, Inc., Cary, NC) was used for data analysis.
The women were mostly of European origin and living with a partner (Table 1). Their average age was 30 years, and over 60% had at least 14 years of schooling. At inclusion, 29% of controls reported smoking and 13% alcohol use. Folic acid supplementation was slightly more frequent among controls (17%) than cases (11%). Regular occupational exposure to solvents was reported by 29% of our working population (mainly cleaners and helpers, nurses, nurses’ aides, hairdressers, and chemists/biologists). The job-exposure matrix classified 18% of working women in the medium-exposure category (mainly nurses, nurses’ aides, hairdressers, and some cleaners) and 3% in the high-exposure category (mainly chemists/biologists and some production workers). Exposure to solvents during hobbies was reported by 13% of controls.
The risk of major malformations increased linearly with exposure, assessed either by self-report or by the job-exposure matrix (Table 1). As in the whole cohort,17 ORs were elevated for oral clefts (OR = 4.3 [95% CI = 1.0–18.2] for regularly exposed vs. nonregularly exposed by self-report; 12 [2.3–60] for women classified as exposed by the job-exposure matrix vs. nonexposed), urinary tract malformations (2.2 [0.6–7.3] and 3.0 [0.9–9.5], respectively), and male genital malformations (3.6 [1.1–12] and 2.11 [0.6–7.3]). Solvent exposure during hobby activities was not associated with the risk of major malformations.
The detection of alkoxycarboxylic acids among controls ranged from 4% for EAA to 93% for PhAA; TCAA was detected among 7% and TCOH among 6% (eAppendix 3, http://links.lww.com/EDE/A610). A higher risk of major malformations was associated with the detection of EEAA (in particular, oral clefts and limb malformations), 2-MPA (in particular, urinary tract defects), TCAA (in particular, limb malformations), and TCOH (limb malformations) (Table 2).
Correlations were observed between the detection of some metabolites, in particular between MAA and MEAA, TCAA and TCOH, EEAA and PhAA, and PAA with 2-MPA and BAA. The first two correlations were expected, as they arise from the same metabolic pathways. The others involved metabolites of glycol ethers known to be components of cleaning agents and cosmetics. When EEAA, TCAA, and TCOH entered jointly into the model, the association between the risk of limb malformations and TCOH was much reduced (OR = 1.7 [95% CI = 0.3–8.4]), although increased risks were still observed for EEAA (3.5 [1.1–10.8]) and TCAA (5.8 [1.4–23.4]).
Our data also showed good agreement between the indirect methods of exposure assessment and urinary metabolites: urinary EAA, EEAA, BAA, TCAA, and TCOH were associated with regular occupational use of detergents and cleaning agents. Metabolites such as EAA, EEAA, BAA, and PhAA were frequently found when occupational cosmetic use was reported (eAppendix 3, http://links.lww.com/EDE/A610).
Our results, based on three methods of solvent exposure assessment, provide evidence of a link between exposure to some classes of solvents in early pregnancy and the risk of major malformations in the offspring.
We used a prospective birth cohort based on a population of homogeneous ethnic origin, high education level, and low tobacco and alcohol consumption during pregnancy. Major congenital malformations previously analyzed in the whole cohort17 include only those identified by hospital discharge and confirmed later (male genital). During the normal 2-year follow-up of the cohort, we identified seven additional major malformations (four cardiac, two urinary tract, one limb, in addition to the 6, 15, and 22, respectively, diagnosed at birth); prenatal solvent exposure was likely for four of them. These additional malformations (false-negatives) were not included in the analyses. We believe, however, that because the specificity of our case ascertainment at birth was high (that is, a negligible number of false-positives) and unrelated to exposure status, they would not have biased our results but could have resulted in a loss of statistical power for the cardiac malformations.
Exposure to solvents during pregnancy was assessed through self-report of occupational exposures, a job-exposure matrix, and urinary biomarkers, all collected in early pregnancy. Self-reports of occupational exposures have been shown to be accurate when familiar agents are listed as done here.22 Although a generic job-exposure matrix has a moderate positive predictive value for specific solvents,23 it is probably most accurate for the solvents present in common products, such as cleaning products, glues, or paints.
Measurement of urinary alkoxycarboxylic acids and of TCAA and TCOH is considered the method of choice for monitoring occupational exposure to glycol ethers and chlorinated solvents, respectively. Our biomonitoring, using a single urine sample during pregnancy, is likely to have identified mainly chronic exposures and higher exposure levels. The intra-individual variability of these markers is not well characterized, but the single sample was taken in the first half of pregnancy, when most women were still working and at a period close to the window of susceptibility for induction of congenital malformations. In general, the frequency of detection of solvent metabolites in this study is consistent with what is known about glycol ether use during the study period.24 Simultaneous detection of TCAA and TCOH in most samples suggests that some chlorinated solvents, metabolized into these two molecules, were the main sources of exposure here.
Taken together, our results suggest associations between occupational solvent exposure and the risk of congenital malformations. The additional information provided by urinary biomarkers, not previously studied, is unfortunately limited by the rarity of the outcome and the low detection rates, which result in unstable risk estimates. Nonetheless, their correlation with indirect exposure measures and the consistency of the associations found with the products handled show that they provide both objective evidence of exposure during the relevant time window and clues about potentially meaningful exposures.
Some urinary metabolites (or their precursor compounds) are known or suspected as developmental toxicants25: EAA, mainly derived from ethylene glycol ethyl ether (EGEE); 2-MPA, derived from the minor β−isomer of propylene glycol methyl ether; or EEAA, mainly derived from diethylene glycol ethyl ether present in biocides, cleaning agents, and alcohol-based products.24 These compounds are metabolized by alcohol dehydrogenase,25 and possession of the slow-metabolizing variant of the alcohol dehydrogenase 1C (ADH1C) gene by either mother or fetus has been shown to increase the fetus’s vulnerability to alcohol-related oral clefts.26,27 TCOH and TCAA are the principal metabolites of tetrachloroethylene and trichloroethylene, present in a variety of compounds common in occupational use. Several animal studies report adverse developmental effects of trichloroethylene or TCAA, mainly cardiac malformations28 or delayed ossification in rat whole-embryo culture.29
Although the precursors mentioned above are plausible suspects in the induction of some classes of congenital malformations, these biomarkers may represent exposure to solvent-containing mixtures that we did not trace. In our study population, five occupations explained most of the solvent exposure: nurses and related occupations; nurses’ aides; chemists, biologists and related occupations; cleaners and helpers; and hairdressers and beauticians. The description of the classes of solvents present in these work environments,9,30–32 as well as the correlations we observed between urinary biomarkers and products handled,33 point to two groups of products (cleaning agents and cosmetics) and at least one chemical class of solvents (oxygenated solvents comprising alcohols, ethers, ketones, and esters) as common to several of these work environments and thus likely to explain the classification as exposed to solvents by indirect methods. Moreover, the consistent association we found between the use of cleaning agents and urinary detection of TCOH or TCAA leads us to suspect the presence of some chlorinated compounds in cleaning agents34,35 or the possible generation of halogenated volatile compounds from bleach-containing cleaning products, either alone or through interaction with other agents.36 These results identify work situations that require further investigation.
We are grateful to the gynecologists, obstetricians, ultrasonographers, midwives, pediatricians, and biochemists who participated in the study, and to the regional medical associations (ADEPAFIN, CGMO) for their collaboration. We thank the staff of all maternity hospitals and clinics in Brittany for their participation in the study. We thank De Vigan from the Paris registry for her help in classifying the congenital malformations.
1. . NOES (National Occupational Exposure Survey) CDC-NIOSH. Available at: http://www.cdc.gov/noes/
. Accessed May 24, 2011
2. Singer BC, Destaillats H, Hodgson AT, Nazaroff WW. Cleaning products and air fresheners: emissions and resulting concentrations of glycol ethers and terpenoids. Indoor Air. 2006;16:179–191
3. . IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol 84 Some Drinking-water Disinfectants and Contaminants, Including Arsenic. 2004 Lyon WHO
4. Bruckner JV, Anand SS, Warren DAKlaassen CD. Toxic effects of solvents and vapors. In: Casarett and Doull’s Toxicology: The Basic Science of Poisons. 20087th ed New York McGraw-Hill Medical:981–1052
5. Saavedra D, Arteaga M, Tena M. Industrial contamination with glycol ethers resulting in teratogenic damage. Ann N Y Acad Sci. 1997;837:126–137
6. García AM, Fletcher T. Maternal occupation in the leather industry and selected congenital malformations. Occup Environ Med. 1998;55:284–286
7. Rylander L, Axmon A, Torén K, Albin M. Reproductive outcome among female hairdressers. Occup Environ Med. 2002;59:517–522
8. Zhu JL, Vestergaard M, Hjollund NH, Olsen J. Pregnancy outcomes among female hairdressers who participated in the Danish National Birth Cohort. Scand J Work Environ Health. 2006;32:61–66
9. Wennborg H, Magnusson LL, Bonde JP, Olsen J. Congenital malformations related to maternal exposure to specific agents in biomedical research laboratories. J Occup Environ Med. 2005;47:11–19
10. Zhu JL, Knudsen LE, Andersen AM, Hjollund NH, Olsen J. Laboratory work and pregnancy outcomes: a study within the National Birth Cohort in Denmark. Occup Environ Med. 2006;63:53–58
11. Cordier S, Bergeret A, Goujard J, et al. Congenital malformation and maternal occupational exposure to glycol ethers. Occupational Exposure and Congenital Malformations Working Group. Epidemiology. 1997;8:355–363
12. Khattak S, K-Moghtader G, McMartin K, Barrera M, Kennedy D, Koren G. Pregnancy outcome following gestational exposure to organic solvents: a prospective controlled study. JAMA. 1999;281:1106–1109
13. Brender J, Suarez L, Hendricks K, Baetz RA, Larsen R. Parental occupation and neural tube defect-affected pregnancies among Mexican Americans. J Occup Environ Med. 2002;44:650–656
14. Shaw GM, Nelson V, Iovannisci DM, Finnell RH, Lammer EJ. Maternal occupational chemical exposures and biotransformation genotypes as risk factors for selected congenital anomalies. Am J Epidemiol. 2003;157:475–484
15. Chevrier C, Dananché B, Bahuau M, et al. Occupational exposure to organic solvent mixtures during pregnancy and the risk of non-syndromic oral clefts. Occup Environ Med. 2006;63:617–623
16. Aguilar-Garduño C, Lacasaña M, Blanco-Muñoz J, Borja-Aburto VH, García AM. Parental occupational exposure to organic solvents and anencephaly in Mexico. Occup Environ Med. 2010;67:32–37
17. Garlantézec R, Monfort C, Rouget F, Cordier S. Maternal occupational exposure to solvents and congenital malformations: a prospective study in the general population. Occup Environ Med. 2009;66:456–463
18. Eurocat. . Eurocat Guide 1.3 and reference documents. Instructions for the Registration and the Surveillance of Congenital Anomalies. 2005 Northern Ireland, UK Eurocat Available at: http://www.eurocat.ulster.ac.uk/pdf/EUROCAT-Guide-1.3.pdf
. Accessed March 24, 2008
19. Labat L, Humbert L, Dehon B, et al. Dosage des métabolites urinaires des éthers de glycol par chromatographie en phase gazeuse couplée à la spectrométrie de masse [in French]. Ann Toxicol Anal. 2008;20:227–232
20. Dehon B, Humbert L, Devisme L, et al. Tetrachloroethylene and trichloroethylene fatality: case report and simple headspace SPME-capillary gas chromatographic determination in tissues. J Anal Toxicol. 2000;24:22–26
21. Ferrario F, Continenza D, Pisani P, et al.Hogstedt C, Reuterwall C Description of a job-exposure matrix for sixteen agents which are or may be related to respiratory cancer. In: Progress in occupational Epidemiology. 1988 Amsterdam, Netherlands Elsevier Science Publishers:379–382
22. Joffe M. Validity of exposure data derived from a structured questionnaire. Am J Epidemiol. 1992;135:564–570
23. Tielemans E, Heederik D, Burdorf A, et al. Assessment of occupational exposures in a general population: comparison of different methods. Occup Environ Med. 1999;56:145–151
24. Multigner L, Droz PO, Dutertre H, et al. Ethers de Glycol Synthèse des connaissances sur les expositions de la population française générale et professionnelle [in French] French Agency for Environmental and Occupational Health Safety (AFSSET). 2008 ISBN 979-2-11-097837-0
25. Catala M, Cordier S, Delaforge M, et al. Éthers de glycol. Quels risques pour la santé? [in French]. Expertise collective INSERM. Paris: INSERM. 1999 ISBN 2-85598-761-X
26. Chevrier C, Perret C, Bahuau M, et al. Interaction between the ADH1C polymorphism and maternal alcohol intake in the risk of nonsyndromic oral clefts: an evaluation of the contribution of child and maternal genotypes. Birth Defects Res Part A Clin Mol Teratol. 2005;73:114–122
27. Boyles AL, DeRoo LA, Lie RT, et al. Maternal alcohol consumption, alcohol metabolism genes, and the risk of oral clefts: a population-based case-control study in Norway, 1996-2001. Am J Epidemiol. 2010;172:924–931
28. Johnson PD, Goldberg SJ, Mays MZ, Dawson BV. Threshold of trichloroethylene contamination in maternal drinking waters affecting fetal heart development in the rat. Environ Health Perspect. 2003;111:289–292
29. Saillenfait AM, Langonné I, Sabaté JP. Developmental toxicity of trichloroethylene, tetrachloroethylene and four of their metabolites in rat whole embryo culture. Arch Toxicol. 1995;70:71–82
30. Delclos GL, Gimeno D, Arif AA, et al. Occupational risk factors and asthma among health care professionals. Am J Respir Crit Care Med. 2007;175:667–675
31. Charles LE, Loomis D, Demissie Z. Occupational hazards experienced by cleaning workers and janitors: A review of the epidemiologic literature. Work. 2009;34:105–116
32. Rylander L, Källén B. Reproductive outcomes among hairdressers. Scand J Work Environ Health. 2005;31:212–217
33. Garlantézec R, Multigner L, Labat L, et al. Urinary biomarkers of exposure to glycol ethers and chlorinated solvents during pregnancy: determinants of exposure and comparison with indirect methods of exposure assessment. Occup Environ Med. 2012;69:62–70
34. ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological Profile for Tetrachloroethylene. 2006 Atlanta, GA: ATSDR
35. ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological Profile for Trichloroethylene. 2006 Atlanta, GA: ATSDR
36. Odabasi M. Halogenated volatile organic compounds from the use of chlorine-bleach-containing household products. Environ Sci Technol. 2008;42:1445–1451