The creatinine-adjusted median concentrations (nmol/g Cr) of total DM and total DAP among 100 pregnant women in the Generation R study () were higher than that of NHANES 2001–2002 pregnant women aged 15–44 years (Centers for Disease Control and Prevention, 2008
), higher than that of pregnant women residing in an agricultural community in the Salinas Valley (The Center for the Health Assessment of Mothers and Children of Salinas, CHAMACOS), California (Bradman et al., 2005
), and higher than that of pregnant women who were studied at Mount Sinai Hospital (The Children’s Environmental Health Study, CEHS) in New York City (Wolff et al., 2007
). Urinary DE levels were similar to those reported in other populations. As observed in other populations, DE levels were lower than that of DM. This may be due to less exposure to diethyl organophosphorus pesticides or because their metabolites are less stable and more difficult to detect. The creatinine-adjusted median level of TCPy in the Generation R study, however, was similar to that of NHANES 2001–2002 pregnant women and was much lower (7 fold) than that of women from the CEHS study () (Berkowitz et al., 2003
). The unadjusted median of TCPy in the Generation R study was also lower than in the CHAMACOS cohort (Eskenazi et al. 2007
). This suggests that this urban population studied in the Netherlands was exposed to a different mixture of OP pesticides than subjects in the U.S.
Comparisons of creatinine-adjusted median concentrations of metabolites in urine of pregnant women
Urinary free BPA accounts for only a small percentage of total BPA, and most human biomonitoring studies, including ours, reported urinary concentrations of total BPA (free BPA plus glucuronide/sulfate conjugated BPA). Urinary median BPA concentration in the Generation R study was slightly lower than that in the pregnant women in NHANES 2003–2004, as shown in . The unadjusted BPA levels were much higher in Korean women (not shown in figure because creatinine-adjusted results not available) (Yang et al., 2006
As shown in , the creatinine-adjusted median MEP concentration in the present study was similar to the creatinine-adjusted median for 25 New York pregnant women (Adibi et al., 2003
) and was higher than in pregnant women from NHANES 2001–2002, the pregnant women in the SFF (Study for Future Families) conducted in several cities in the United States (Marsee et al., 2006
), and pregnant women from Taiwan (Huang et al., 2007
). The median level of MnBP (DBP metabolite) in the present study was higher than in NHANES 2001–2002, in the SFF study, and in New York study, but lower than in Taiwan. Pregnant women in the present study had a median level of MiBP that was more than 10 times higher, compared with NHANES 2001–2002 and the SFF study. The concentrations of three DEHP metabolites (MEHP, MEHHP and MEOHP) were similar to that of women in NHANES 2001–2002 and the SFF study, but lower than in Taiwan. Ratios of the median levels of two secondary DEHP metabolites (MEHHP and MEOHP) to that of MEHP (primary metabolite) are indicators of metabolic capacity and vary across populations. The ratios of MEHHP/MEHP (1.9) and MEOHP/MEHP (2.0) were somewhat lower in the present study compared to other studies’ in which ratios were found to be greater than 3 (Barret al., 2003
; Becker et al., 2004
; Fromme et al., 2007
; Koch et al., 2003b
). This finding indicates a difference in metabolism, but still suggests that MEHHP and MEOHP are more sensitive biomarkers of exposure than is MEHP. The median concentration of MBzP, a metabolite of butylbenzyl phthalate which is used mainly as a plasticizer (PVC for vinyl floor tile, vinyl foams, and carpet backing), was slightly lower than in women in NHANES 2001–2002 and the SFF study (Marsee et al., 2006
), but higher than in Taiwan (Huang et al., 2007
The laboratory we used regularly takes part in an external quality assessment scheme for the determination of phthalates and organophosphates. In particular, the analytical laboratory at the National Center for Environmental Health, U.S. Centers for Disease Control and Prevention participated in this scheme and that laboratory did the measurements in all other studies shown in except for the Taiwan study. Thus, the values for phthalate and DAP metabolites are comparable across laboratories (e.g. similar LODs and recoveries) for the present study and for the U.S. studies. However, the interlaboratory variability may partially contribute to the differences in BPA between the Generation R study and NHANES, and the differences in phthalates between western and Taiwanese women.
The differences in OP metabolites may be due to differences in pesticide use among different countries. In 2003, 14.4% of tested foods in the Netherlands had pesticide residues above maximum residue levels (MRLs), the highest percentage of any country in the European monitoring program (European Commission, 2007
). The most frequently detected pesticides were dichlorvos, pirimiphos-methyl, malathion, chlorpyriphos-ethyl, and methamidophos, which are primarily metabolized to DM, supporting our finding that urinary DM metabolite levels were relatively high in the present study. Pesticide residues in foods imported from developing countries are more often higher than MRLs and may be another factor relevant to higher DAP levels (European Commission, 2007
). In the U.S., violative pesticide residues (a residue which exceeds a tolerance or a residue at a level of regulatory significance for which no tolerance has been established in the sampled food) were found in 2.4% of domestic foods and in 6.1% of imported foods in 2003 (Food and Drug Administration, 2005
). The higher concentration of TCPy in the CEHS study and the CHAMACOS study might be because chlorpyrifos and chlorpyrifos-methyl were more commonly used in these areas. Chlorpyrifos-methyl was the one of five most frequently detected chemicals in food items from U.S. markets (Food and Drug Administration, 2005
). The chlorpyrifos degradation product TCPy has also been frequently detected in U.S. food and environmental media (Morgan et al., 2005
). In the Netherlands, however, none of three TCPy parent pesticides (chlorpyrifos, triclopyr and chlorpyrifos-methyl) has been frequently detected in food items (European Commission, 2007
Phthalates, including di-n-butyl phthalates (DBP), diethyl phthalate (DEP) and di-2-ethylhexyl phthalate (DEHP) are used as plasticizers and ingredients in cosmetic products (bath preparations, eye shadows, perfumes and other fragrance preparations, hair sprays, nail polish, and skin care preparations), building materials, household furnishings, clothing, food packaging, cleaning materials, and insecticides (Schettler, 2006
). A survey in 2002 (DiGangi and Norin, 2002
) showed that more than half of cosmetic products in the European market contained more than one type of phthalate and about 40% of the products contained either DEHP or DBP, which were subsequently banned by the European Union in 2002 for use in personal care and cosmetic products. Di-iso-butyl phthalate (DiBP) has similar properties to DBP and has now replaced DBP in many applications including cosmetics. The relatively high levels of MnBP and MiBP in the Generation R women may be related to use of personal care and household products.
Few studies have examined the correlates of exposure to these contaminants and their metabolites. The higher concentrations of some phthalate metabolites among younger women (MnBP), women whose household income was less than 2200 Euros/month (MnBP, MiBP, and MBzP), women with primary education only (MnBP and MBzP), and non- Dutch women (MnBP, MiBP, MBzP, and total DEHP) may be related to more use of personal care products containing phthalates. The higher total DEHP concentrations in urine may also be related to DEHP exposure from diet and exposure to dust and air in the indoor environment. The lower level of the total DAP among women who provided a urine sample after 26 weeks of gestation was consistent with the increasing urine volume during pregnancy (Maikranz et al., 1989
). The higher levels of total DAP among women who had higher education and of TCPy among non-Dutch women may be related to diet or use of pesticides in the home.
The metabolites analyzed in this study were selected in part so as to be comparable to previous studies; however, caution should be used when interpreting these levels as biomarkers of exposure. DAP metabolites are regarded as good indicators of OP exposure although several limitations exist. Measures of DAP metabolites may also reflect exposure to OP breakdown products in the environment and in the food supply and the contribution of some industrial chemicals and drugs (Barr and Needham, 2002
). DAP metabolites have been used for biomonitoring in workers and other exposed populations but the interpretation of DAP metabolite levels in the general population is less clear. DAP metabolites are not toxic, so that the component that consists of "pass through" metabolites via food is, as far as we know at this time, of no toxicological significance. Because phthalates can be contaminants in the laboratory, measurement of parent phthalates is useful only for highly exposed populations (Barr et al., 2003
). Phthalate metabolite concentrations in urine samples are the most frequently used biomarkers for evaluating phthalate exposure in the general population. The selection of metabolites for measurement needs careful attention because some intermediate metabolites can be further metabolized. For example, the simple monoester mono-iso-nonyl phthalate (MINP) has been used for human exposure assessment of di-iso-nonyl phthalate (DINP) (Silva et al., 2004
) and was recently shown to be an unreliable biomarker because it is further metabolized to oxidative metabolites before excretion in urine (Koch et al., 2007
; Silva et al., 2006
). Investigations by two independent groups have shown that oxidative metabolites mono-4-methyl-7-hydroxy-octyl phthalate (7oh-MMeOP), mono-4-methyl-7-oxo-octyl phthalate (7oxo-MMeOP) and mono-4-methyl-7-carboxy-heptyl phthalate (7carboxy-MMeHP) are better biomarkers of DINP exposure (Koch et al., 2007
; Silva et al., 2006
). As observed before (Barr et al., 2003
), concentrations of secondarily oxidized metabolites (MEHHP and MEOHP) were higher than that of MEHP in the present study.
The choice of matrix for biomonitoring depends on pharmacokinetics, access to samples, and analytical capability. For the agents considered in the present study, urine is the most frequently used matrix for human studies because of the ready availability of large amounts of sample and relatively higher concentrations than in blood (Barr and Needham, 2002
). Collection of spot urine is easier and more frequently done than 24-hr or first morning urine samples. Spot urine samples were collected in all previous studies on these compounds with the exception of two studies of phthalates in Germany that used first morning void urine samples (Fromme et al., 2007
; Wittassek et al., 2007
One concern about measurements in spot urine is within-individual variability over time. Significant inter-day variation of urinary phthalate metabolite levels during 8 consecutive days has been reported (intraclass correlation coefficient ranged from 0.2 to 0.5) (Fromme et al., 2007
). Bradman et al. (Bradman et al., 2005
) also found temporal variation of OP metabolites between the first (about 13 weeks) and the second (about 26 weeks) prenatal urine samples, and between prenatal and postpartum samples. Due to temporal variability and short half-lives (<48 hours) (Barr and Angerer, 2006
; Koch et al., 2006
; Tsai, 2006
), a single spot urine sample usually provides only an imprecise estimate of long-term exposure to the target compounds. Thus, measuring metabolites in multiple samples across the entire pregnancy may allow improved exposure assessment, and help to identify critical windows of exposure.
Although we compared analytes among pregnant women in different studies (figure), the difference in time of gestation when urines were collected in these studies need consideration. Creatinine excretion is affected by many factors including pregnancy (Boeniger et al., 1993
). Urinary creatinine level changes with the time of gestation and is about 20% lower in the 3rd
trimester than in the 1st
trimester of gestation (Cherry, 1991
). Urine volume increases by about 25% in pregnancy and is the greatest in the 3rd
trimester (Maikranz et al., 1989
). The magnitude of effect on metabolite concentrations of these and other changes in clearance is unclear. In the present study, the median levels of total DM and total DAP in women who provided a urine sample after 26 weeks of gestation were statistically lower than that in women who provided a sample earlier. In CHAMACOS study, however, the median total DAP concentrations in the first urines (about 13 weeks of gestation) and in the second urines (about 26 weeks of gestation) were similar (Bradman et al., 2005
). The influence of timing may contribute to, but can not fully explain the observed large differences in concentrations of DAP metabolites, TCPy, and several phthalate metabolites (MEP, MnBP, MiBP, and MEHP) across studies. Although urine osmolality or specific gravity may be a better approach to adjust for differences in urine dilution, we chose to focus and report creatinine and non-creatinine adjusted measurements to faciliate comparison with other studies.
The subjects in the present study for whom data were complete had characteristics that were not statistically significantly different from those in the Generation R women overall (Jaddoe et al., 2006
). However, the proportion with missing data in our study was higher than in the Generation R study. This reflects that the analyzed urines were selected from women who had only a third trimester specimen available. Consequently, the present subset contains a relatively large proportion of women who entered the study relatively late during pregnan0063y, for whom the responses to the questionnaires in early and mid pregnancy were not available. Thus, some differences that we could not detect may have been present. Therefore, some caution in extrapolation to all Generation R subjects is in order.
Relatively high levels of OP pesticide metabolites and some phthalate metabolites were found in the Generation R subjects and the reasons merit further study. Nonetheless, because detailed birth outcomes and follow-up information were recorded, the Generation R study provides an opportunity to efficiently address questions regarding reproductive and developmental effects of prenatal exposure to some OP pesticides, BPA, and phthalates.