ToxSci Advance Access originally published online on January 18, 2007
Toxicological Sciences 2007 97(2):241-252; doi:10.1093/toxsci/kfm005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Published by Oxford University Press 2007.
Can Exposure Characterization Explain Concurrence or Discordance between Toxicology and Epidemiology?
* Department of Environmental Biology, University of Guelph, Guelph, Ontario, Canada
Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Ontario, K1A 0K9, Canada
1 To whom correspondence should be addressed. Fax: 519-837-3816. E-mail: lritter{at}uoguelph.ca.
Received January 4, 2007; accepted January 10, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
The importance of reliable exposure assessment, as a key component of the overall risk assessment process, has been well described for some considerable time. Yet, despite this widely accepted tenet, many studies conclude significant adverse health effects, with associated public policy implications, in the absence of adequate or, in some cases, even rudimentary, exposure quantification. Moreover, it appears that epidemiological studies in humans and toxicological studies in experimental animals may both suffer from inadequate exposure assessment. In this review, we discuss the nature and quality of the exposure assessment in both epidemiologic and toxicologic studies using examples from the pesticides and phthalate literature. Each type of study has its strengths and weaknesses in how exposure is assessed and often the strength of one is also a weakness. It would appear that insufficient or incomplete information about differences in exposure assessment could explain, at least in some cases, the differences in outcome between toxicological and epidemiological studies. Research efforts should focus on improving the feasibility of including biomonitoring in both animal and human studies to facilitate comparisons between animal and human models and improve exposure assessment in epidemiologic studies. Animal and human studies should measure the same biomarkers, where possible, to facilitate human health risk assessment.
Key Words: epidemiology; toxicology; phthalates; pesticides; exposure assessment; risk assessment; biological dose; administered dose.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Toxicological and epidemiological studies differ significantly in how they define and measure exposure. One obvious difference between the two types of exposure assessment is that toxicological studies draw on relatively precise dosing information and create dose-response curves, whereas epidemiologic studies must rely on other sources of exposure information such as self-reports, environmental monitoring, or body burden residues (if such data are even available) to measure exposure. Prior to examining a specific research question, a few dose measurements of interest need to be defined. The available dose is a measure of the total concentration of the chemical in the personal environment of the individual from all exposure pathways as measured in the various media (e.g., air, drinking water, food). The administered dose is the chemical concentration that actually contacts the barrier of the body by various routes (i.e., dermal, inhalation, and ingestion), whereas, the absorbed dose is the concentration that actually enters the body. The active dose (or biologically effective dose) is the amount of chemical (parent or metabolite) that reaches the target sites (organs, tissues, cells) where toxicity can occur. The administered dose is often not the appropriate dose metric for risk assesment. The exposure biomarker is a measure of body burden, which may or may not accurately reflect the absorbed or active dose, depending on the tissue or body fluid sampled and the toxicokinetics of the chemical. There have been only a limited number of environmental pollutants regulated by the U.S. Environmental Protection Agency (U.S. EPA) on the basis of epidemiology studies, largely because there is no formal process for epidemiologic risk assessment (Calderon, 2000
). In order to inform the risk assessment process, epidemiologic studies need to be performed and analyzed in a way that is suitable for risk assessment (Rice, 2005). This requires information on the shape of the relationship between exposure and effect and on whether there is evidence of a threshold for the effect. The objective of this paper is to use a few examples from the literature to compare the results of experimental animal laboratory studies with observational epidemiologic studies, with a specific focus on how exposure was assessed in these studies and whether this analysis might explain concurrence or discordance between the two types of studies.
Table 1 provides an overview of how toxicological and epidemiological studies assess exposure, which is directly related to the study design—experimental or observational. In toxicological studies, exposure to the study agents is controlled and well characterized in terms of chemical species, constant dose, source, vehicle and route of administration, and timing and duration of treatment. As environmental epidemiologic studies are by their nature observational, there can be multiple sources and routes of exposure of varying intensity, timing, and duration to various forms of the chemical of interest. In reality, humans are exposed to the chemical agents of interest within a mixture. While toxicological studies have fixed and limited numbers of exposure groups with preset sample sizes set a priori, epidemiological studies estimate dose based on environmental sampling (i.e., ad libitum exposure to contaminated air, drinking water, food), consumption patterns, and body weight or on biomarkers of exposure (i.e., concentration of chemical in available body tissues or fluids). Because of the high costs of collecting and analyzing biological specimens for environmental chemicals, epidemiologic studies incorporating body burdens generally only have samples collected for individuals at one point in time. The suitability of this snapshot of exposure will depend on the critical developmental stage of interest, consistency of exposure over time period, and the half-life of the chemical. Depending on what biological tissue or fluid is sampled and the toxicokinetics of the chemicals, it can be difficult to compare an animal-dosing regimen with a biomarker from one point in time; this may be less of a concern for environmental monitoring data (e.g., arsenic levels in drinking water). Our current technical inability to document contaminant levels during critical windows of exposure has hampered the ability of epidemiologic studies to measure exposure. Other difficulties in using exposure biomarkers include dealing with concentrations below the limits of detection, undersampling of key strata, time delays between exposure and sampling, limited variation across individuals as well as high variability within individuals (Koo et al., 2002), potential inconsistencies between sophisticated laboratory methods, potential contamination of samples, and the ethical, legal, and social issues involved in conducting these types of studies.
|
The epidemiologist must use selection and restriction criteria for selecting the study population and/or for data analysis, in an attempt to have a range of exposures with sufficient sample sizes within exposure categories and as homogenous a population as possible. The toxicologist can select animals that are genetically homogenous and largely control any nutritional or environmental factors which could affect the results of the study.
While statistical analysis of toxicological studies is relatively straightforward, a number of factors inherent in the epidemiological design make their statistical analysis more complicated. Often, continuous exposure data derived from environmental or biological sampling are not normally distributed and must be transformed in some manner before application in statistical models. As is evident from a recent analysis of blood lead levels and children's IQ (Rothenberg and Rothenberg, 2005), statistical issues such as what if any transformation of the often non-normal distribution of contaminant concentrations is used (e.g., logarithmic, square root) and the choice of statistical model (e.g., linear regression, cubic spline generalized additive models) can dramatically influence the results and conclusions. Each analytical method will also have a detection limit, and different statistical approaches can be used to handle values below the detection limit of the laboratory method. If the data are to be categorized, the selection of the cut points may be influenced by what is known from the published literature, policy (reflecting current guidelines or regulatory action levels), and population distributions (adequate sample sizes) within categories. Furthermore, the choice of the statistical model for characterizing the dose-response curve can affect the interpretation of the results. In the analysis to determine the shape of the relationship between body burden of mercury, for example, and adverse outcome, the National Research Council chose a linear model for its assessment of mercury; however, there was evidence that the slope was actually steeper at lower body burdens than at higher ones (Rice, 2004). Moreover, as accurate rate constants are not available for humans for all the necessary compartments (maternal blood, fetal blood, fetus, and fetal organs) (Rice, 2004), a one-compartment model was used to convert body burden to maternal intake, which was a significant simplification of the pharmacokinetics (PK) of methylmercury in the maternal body and maternal-fetal unit.
In addition, the complexity of exposure to environmental chemicals for humans is difficult to compare with the dose administered in an experimental animal study. Issues such as duration, magnitude, timing, routes of exposure, and variability in exposure, as well as genetic diversity and other endogenous factors are typically not encountered in animal studies (Calderon, 2000). Toxicologic studies are rarely able to mimic the human exposure experience, with the result that assumptions must be made of constant lifetime exposure to usually unrealistic doses.
Human biomonitoring surveys such as the U.S. Third National Report on Human Exposure to Environmental Chemicals (http://www.cdc.gov/exposurereport) have revealed that human populations have measurable body burdens of multiple chemicals, most of which have accumulated in the body over extended periods and at different times. In a study which measures only one contaminant, potentially more important contaminants for the population under investigation may be missed. In the laboratory animal environment, exposure to environmental chemicals other than the dosing chemicals rarely occurs and hence need not typically be considered or measured. However, as evidenced from several reports, uncontrolled biological or environmental factors such as a change in water supply to an animal facility (Sharpe et al., 1998) or leaching of chemicals from cages or plastic bottles (Hunt et al., 2003) could have pronounced effects in control animals, which could mask the effect induced by the agent under study.
A critical issue in epidemiologic studies is careful evaluation of the posited relation between the exposure and health effect controlling for any potential confounding factors which if uncontrolled could bias the result and lead to incorrect inferences. A confounding factor must be associated with the exposure under study in the population from which the diseased cases derive, must be a risk factor for the disease even among people who lack the exposure, and cannot represent a step in the causal chain between exposure and disease (Rothman, 1986). Another issue rarely considered in toxicology is effect modification (interaction between chemicals) and how best to represent this in a statistical model (although those few studies that do incorporate mixtures use statistical models). Measuring and considering other chemicals in the statistical analysis might account for additional variance and thereby increase the study's power to detect an effect of exposure (Rice, 2005). Current statistical models do not adequately characterize the behavior of several chemicals that are highly correlated, making it particularly difficult to differentiate effects produced by these contaminants in a population. This often results in epidemiologic investigators only analyzing and reporting the effect of a single contaminant, even though multiple chemicals were measured.
While both types of studies require institutional research ethics approvals (human or animal), epidemiological studies must all consider privacy laws, confidentiality, informed consent, and what, if any, information to communicate to the subjects.
Differences in PK, particularly variability in half-lives of drugs are known to exist between adults and children and among children between various age groups (e.g., premature and term births, 1 week, 1 week to 2 months, 2–6 months, 6 months to 2 years, 2–12 years, and 12–18 years) (Hattis et al., 2003). For example, in an analysis of the PK of 45 drugs, premature and full-term neonates tended to have drug half lives that were three to nine times longer half-lives than adults for the drugs examined; whereas, beyond 6 months of age, half-lives were shorter than in adults for some drugs and pathways (Ginsberg et al., 2002). This increases the uncertainty in making extrapolations of adult dosimetry estimates for environmental contaminants to children, especially for infants (Ginsberg et al., 2002). As risk assessments increasingly rely on physiologically-based pharmacokinetic (PBPK) models to adjust for PK differences between laboratory animals and humans, more attention should focus on exposure during very narrow critical developmental stages postnatally, in addition to the attention given to prenatal windows.
|
PESTICIDES |
|---|
|
TOPABSTRACTINTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Pesticides, a generic term used to describe a broad range of structurally unrelated compounds with varying modes of action, biological targets and target pests, are likely among the most intensively regulated of all chemical substances in modern commerce. Given their widespread use in every aspect of modern life, from food production to industrial, home, and garden and ornamental uses, and the potential of pesticide use to threaten both environmental and human health, such intensive regulation is both laudable and warranted. The controversy appears, however, to arise not from any debate as to the need for such intensive regulation but rather from the frequent disagreement, indeed contradiction, in risks characterized by classical epidemiological and toxicological studies. Noting that laboratory toxicological studies can only be viewed as predictive, at best, and are often the sole toxicological basis supporting the registration decision in most regulatory jurisdictions, and noting that epidemiological studies are almost always carried out long after a pesticide has been in use for many years, the importance of concordance becomes self-evident if regulatory authorities are to continue to rely on predictive laboratory studies to assess and predict safety in human populations.
Almost 500 years ago, Paracelsus, considered by many to be the father of modern toxicology, observed that all substances have the potential to be toxic and that only the dose differentiates a safe chemical from a toxic one. Of course, today, we understand that a "safe" chemical can only be defined in terms of the expected exposure in some susceptible population and that the concept of exposure must necessarily include the frequency, duration, and intensity of that exposure, as well as relevant windows of sensitivity and susceptibility, if a meaningful and scientifically robust risk assessment is to be conducted.
Four hundred years later, in 1965, in his landmark address to the British Royal Society, Sir Austin Bradford Hill (Hill, 1965) articulated a disciplined framework by which investigation in the life sciences should be conducted and, in so doing, laid the modern day foundation for our understanding of the importance of exposure assessment in assessing and predicting risks in human populations from exposure to potentially toxic chemicals. Almost 20 years later, in 1983, the U.S. National Research Council articulated its framework for risk assessment (National Research Council, 1983), noting the critical importance of each step of the paradigm, including the assessment of the expected exposure and, recognizing the principle first articulated by Paracelsus almost 500 years earlier, an analysis of the dose-response relationship. The paradigm has now become entrenched in regulatory practice worldwide and provides a disciplined basis for the assessment of risks associated with the use, and misuse, of a broad array of chemicals in modern life. Yet, despite almost 500 years of experience in the development of our understanding of the importance of exposure assessment in reducing uncertainty in the risk assessment process, the lack of robust exposure assessment continues to be a major impediment to our understanding of real, versus perceived, risks to human populations. Indeed, discordance in study outcomes, when comparing laboratory- and population-based studies, may simply reflect the uncertainty inherent in the exposure component of different study designs and the consequent impact on the risk assessment process. Typically, while dosing is relatively precise in laboratory studies, the methods used to assess exposure in epidemiological studies are often imprecise employing indirect surrogates, with greater precision only possible with direct measurement of pesticide metabolites in biological samples of those involved in pesticide application or those who may use pesticide-treated areas. Assessment of environmental and biological concentrations is understood to offer far greater reliability for estimating exposure; however, the assessment of exposure based on individual assessments in study subjects is simply not practical or feasible in large-scale epidemiology studies (Stewart, 1999).
Biomonitoring of Pesticide Exposure
The importance of robust exposure assessments as a critical component of the risk assessment process continues to attract the attention of many researchers. The dilemma posed by inadequate exposure assessment has been described by many workers, and the resulting quagmire is well described in the scientific literature. Leiss and Savitz (1995) investigated the relationship between household pesticide use and risk of childhood cancer and concluded that the major weakness of their study, as with previous studies of home pesticides and childhood cancer, was the crudeness of the exposure measures. The authors went on to note that future studies should aim for more specific measures of exposure in terms of age, duration, intensity, and particular chemical agents. Similarly, Daniels et al. (1997) evaluated epidemiologic studies published between 1970 and 1996 in order to investigate the possible association between pesticides and the risk of childhood cancer. The authors noted, inter alia, that while the reported relative risk estimates were modest, risk estimates appeared to be stronger when pesticide exposure was measured in more detail. Having said this, the authors also noted that methodological limitations, including indirect exposure classification, constrained the strength of their conclusions. In both cases, potential childhood exposures to pesticides, and subsequent impact on cancer outcome, were assessed through administration of a questionnaire to parents of children who had previously been diagnosed with childhood cancer; no direct assessment of exposure was made. The authors concluded that an etiologic relationship between pesticide exposure and childhood cancer is far from proven.
In contrast, exposure in rodent carcinogenicity studies is typically assessed through ongoing and frequent analysis of the dosage form to ensure that administered doses are in accordance with study protocol, thereby providing direct analytical evidence of exposure has occurred in the test species. Notwithstanding, animal toxicology studies such as feeding studies and inhalation studies often do not assess internalized dose.
Pesticides and Cancer
The dichotomy in outcomes between laboratory animal studies and population studies in humans is, perhaps, well characterized by conflicting results that have been reported for 2,4-dichlorophenoxyacetic acid (2,4-D). In this case, there is a well-described literature that has frequently concluded increased cancer risks in humans, particularly in children (Ontario College of Family Physicians, 2004; Sears et al., 2006; Zahm and Ward, 1998), associated with use of this herbicide in home and garden settings, contrasting sharply with the position taken by governments in both Canada and the United States (PMRA, 2005; USEPA, 2005), that have concluded that when used properly, and in accordance with label directions, the use of 2,4-D in these settings poses little or no risk to users or bystanders, including children.
In 1998, Zahm and Ward (1998) published the results of their meta-analysis of pesticides and childhood cancer and concluded that many of the cancers associated with pesticides among children were similar to those repeatedly associated with pesticide exposure among adults, suggesting a plausible role in childhood cancer. Notwithstanding, Zahm and Ward, however, also noted that most of the studies included in their analysis were based on crude exposure information with little specificity in pesticide type or amount and echoed the concerns of other investigators regarding the need to study and better quantify these exposures.
The reliability of reporting on lifestyle and agricultural factors in order to assess exposure in human subjects has also been investigated in the Agricultural Health Study (Alavanja et al., 1996) as well. Blair et al. (2002) have studied exposure assessment in self-completed interviews reported a year apart in 4088 Iowa pesticide applicators. The authors noted that agreement was generally quite high (70–90%) for ever/never use of specific pesticides, but much lower (50–60%) for duration, frequency, or decade of first use—information that is critical for a scientifically robust dose-response assessment that forms the basis for reliably predicting risk from pesticide use. Ritter (2002) has also reflected on the importance of reliable exposure assessment as a basis for meaningful investigation of pesticide exposure–related adverse health outcomes. In his analysis of many of the studies included in the Zahm and Ward (1998) assessment, Ritter noted that only 7% of the studies included by these workers in their meta-analysis relied on any direct exposure measure, underscoring the concerns of many investigators regarding the reliability of risk assessments that relied on inadequate exposure estimates. Moreover, the reliability of questionnaire data in many epidemiologic studies, particularly noting the limitations imposed by possible recall bias and small sample size, appear to further support the plea by Zahm and Ward, and others, for studies that better quantify exposure.
Exposure Surrogates
The challenge appears to have been taken up, at least in part, by several investigators. Almost a decade ago, Arbuckle et al. (2004) undertook a large-scale exposure study of farmers utilizing phenoxy herbicides in order to directly investigate pesticide exposure in applicators, their spouses, and farm children residing on these farms and, perhaps more importantly, to compare the validity of questionnaire data for estimating exposure confirmed through biomonitoring. In designing their study, these authors observed that their approach of directly comparing epidemiologic questionnaire data with biomonitoring data in children did not appear to have ever been reported elsewhere. The authors opined that exposure determinants could be more readily measured in a smaller scale study and that these determinants could then be applied in larger scale epidemiological studies, where direct exposure assessment is not practical, in order to refine the validity of the surrogate measure. The study, which investigated exposure in 92 children, revealed that approximately 70% of the children residing on farms that reported use of either 2,4-D or (4-chloro-2-methylphenoxy) acetic acid (MCPA) did not have detectable concentrations in their urines and, of the 30% that did, concentrations were low (microgram range). While the relatively limited exposure in children reported by these workers is comforting, a comparison of the exposure estimate assessed by questionnaire, when compared to the biomonitoring data, was more disturbing. To wit, these workers reported that a comparison of the two approaches revealed that the sensitivity of questionnaire data in predicting whether a child was exposed to the herbicide (as measured by detection in the urine) was only 47% for 2,4-D and 91% for MCPA, while the specificity of the questionnaire information in predicting whether a child was truly unexposed was 72% for 2,4-D and only 30% for MCPA. The authors concluded that questionnaire data related to living on a farm, or living on a farm when a specific pesticide is used, is not enough to classify children's exposures. The authors also opined that given the potential for exposure misclassification revealed by their study, incorporation of biomonitoring studies in at least population subsets was important to at least estimate the extent of misclassification.
In a similar study which examined questionnaire data as a surrogate for directly assessing exposure in epidemiological studies, Coble et al. (2005) investigated the reliability of important exposure determinants, such as application methodology and use of personal protective equipment, as predictors of applicator exposure. Similar to Arbuckle and her coworkers, these authors speculated that it might be possible to add significant precision to surrogate exposure assessments from questionnaires by utilizing quantitative data from smaller scale exposure studies along with detailed descriptions of factors that might affect exposure levels in order to better identify and quantify exposure determinants. Coble et al. then developed an algorithm to calculate pesticide exposure intensity based on responses to questions about pesticide handling procedures and application methods in a self-administered questionnaire. The authors reported that urinary concentrations across three categorical exposure groups for both MCPA and 2,4-D, based mostly on reported use of personal protective equipment, provided a reasonable estimate of exposure intensity for the pesticide applicators. However, the study also revealed that the statistically significant correlation between the algorithm score and the urinary concentrations worked less well for MCPA than for 2,4-D, suggesting that additional refinements of the algorithm model are necessary to improve the validity of the approach for a broad range of pesticides and application methodologies.
Drawing on the considerable strength provided by the Agricultural Health Study and its database of pesticide exposures in more than 58,000 applicators in North Carolina and Iowa, Dosemeci et al. (2002) have reported on a quantitative method to estimate pesticide exposure. In their study, Dosemeci et al. noted that most epidemiological studies tend to consider pesticides as a group, rather than as chemically specific entities, and without characterization of chemical-specific exposures. Moreover, while some epidemiological studies have evaluated specific health outcomes as a function of chemical-specific exposures (Baris et al., 1998), frequency of exposure, duration, or the intensity of exposures to individual pesticides have received little attention. Indeed, Dosemeci et al. noted that most studies relied entirely on surrogate measures of intensity, such as number of acres treated, number of days of pesticide application, crop type, and information on the extent of use of personal protective equipment. Relying mostly on results of different exposure measurements from monitoring studies that used different pesticides for the same variables, Dosemeci et al. then developed a weighting factor for each of the variables. The authors concluded that the distribution patterns of all basic exposure indices (intensity, duration, and cumulative exposure—in this case to 2,4-D and chlorpyrifos) in both their full cohort and a subcohort were virtually identical, suggesting that a relatively small subcohort of applicators may be a reasonable representation of the much larger full cohort in terms of estimating exposure.
|
PHTHALATES AND MALE REPRODUCTIVE HEALTH |
|---|
|
TOPABSTRACTINTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Another research question that has received recent attention is whether phthalate exposure during pregnancy plays an etiological role in male reproductive disorders. Phthalates are a group of chemicals frequently used in consumer and personal care products, plastics, medications, and medical devices. Testicular dysgenesis syndrome (TDS) has come to refer to a collection of disorders of male newborns or young adults with a common origin in fetal life (Skakkeback, 2002
). Examples include cryptorchidism (undescended testes) and hypospadias (two of the more common congenital malformations prevalent at birth), testicular germ cell cancer, and low sperm counts (Sharpe, 2005).
Phthalate administration to laboratory animals during pregnancy has been shown to induce TDS-like disorders. In rat studies, certain phthalates (di-n-butyl, diethylhexyl phthalate, or butyl benzyl phthalate) administered during the period of sexual differentiation resulted in a collection of disorders in the male offspring similar to TDS disorders in human males (Fisher et al., 2003; Mylchreest et al., 2000). Phthalate exposure also resulted in associated suppression of hormone production (testosterone, insulin-like factor 3) by the fetal testis (Parks et al., 2000; Thompson et al., 2004; Wilson et al., 2004). Phthalate-exposed pups exhibited a reduction in anogenital distance (Ema and Miyawaki, 2001; Gray et al., 2000; Mylchreest et al., 2000; Zhang et al., 2004), which was a direct reflection of growth-stimulating actions of androgens such as testosterone on the perineum in fetal life (Sharpe, 2005). Although lifelong exposure to di-2-ethylhexyl phthalate (DEHP) has also been shown to result in increased rates of liver and testicular cancer in rats (Voss et al., 2005), the Leydig cell tumors observed in the rat testis are not related to the germ cell tumors observed in humans.
From the epidemiologic perspective, the data are less clear; however, maternal phthalate exposure has been associated with subtle developmental effects in male babies (Swan et al., 2005) and inhibition of testis function as measured by alterations of reproductive hormones (Main et al., 2006). In addition to developmental effects, phthalate ester exposure has also been correlated with reduced semen quality (Duty et al., 2003).
Biomonitoring of Phthalates
Measuring phthalate exposure in humans is problematic for a number of reasons. Traditionally, phthalate exposure has been estimated based on spot urine levels of phthalate monoester metabolites of the parent diesters. In addition to concerns about whether or not to adjust for creatinine levels (Barr et al., 2005), little is known about how urinary levels vary over time within the same individual. One study which analyzed two consecutive first morning urine specimens from women reported that phthalate levels did not differ between the two sampling days, which would suggest that women's patterns of exposure may be sufficiently stable to assign an exposure level based on one first morning void measurement; however, the correlation did vary by metabolite with the highest correlation for mono-n-butyl phthalate (MBP) (0.8) and the lowest for mono-benzyl phthalate (MBzP) (0.5) (Hoppin et al., 2002). Furthermore, a study which collects urine on only 2 days may be inadequate to conclude that women's patterns of exposure are stable over time.
Identifying the appropriate biomarker of exposure to various phthalate compounds is still a challenge. Traditionally, mono-isononyl phthalate (MINP) has been measured as the biomarker for di-isononyl phthalate (DINP); however, a recent study has concluded that the prevalence of human exposure to DINP is underestimated using only this urinary biomarker and that the oxidative metabolites better reflect exposure (Silva et al., 2006). Similarly, mono-2-ethylhexyl phthalate (MEHP) was the common urinary metabolite of DEHP measured; however, secondary oxidized DEHP metabolites like mono-(2-ethyl-5-hydroyhexyl)phthalate (MEHHP), mono-(2-ethyl-5-oxohexyl)phthalate (MEOHP), mono-(2-ethyl-5-carboxypentyl)phthalate (5cx-MEPP), and mono-[2-(carboxymethyl)hexyl]phthalate (2cx-MMHP) represent the major proportion of DEHP metabolites excreted in urine and may be the ultimate developmental toxicants (Koch et al., 2006).
The collection of different biological specimens at different points in development makes it difficult to compare and interpret both epidemiologic and toxicology studies.
Although it is difficult to compare the phthalate dose administered to animals with the concentration of urinary phthalate metabolites measured in humans, humans are generally considered to be exposed to levels lower than those capable of affecting fetal testis hormone production and masculinization in rats (Sharpe, 2005). The results of epidemiologic studies would challenge this assumption. Infrequently, experimental animal studies have measured urinary concentrations of phthalate metabolites following dosing. In one study where rats were dosed with di-n-octyl phthalate (DOP) by gavage, the authors concluded that measurement of mono-n-octyl phthalate (MnOP) alone would have significantly underestimated DOP exposure as urinary levels of one of the oxidative metabolites mono-(3-carboxypropyl) phthalate (MCPP) were about 560-fold higher than MnOP (Calafat et al., 2006a). The results suggested that the monoesters may be poor biomarkers of exposure to their precursor phthalates.
Phthalate Toxicokinetics
Are rats the best animal model for predicting human toxicity of phthalates? Di-n-butyl phthalate (DBP) is metabolized in female pregnant rats to MBP and its glucuronide, monohydroxybutylphthalate and its glucuronide, and butanoic acid phthalate and its glucuronide (Fennell et al., 2004). Rodents metabolize phthalate diesters to monoesters extensively in the gut following oral administration; however, primates do not (Foster et al., 2000). If the monoester is the active metabolite causing the developmental toxicity in male rodents and humans produce very low levels of the monoester from environmental exposure to the diester, some investigators have postulated that reproductive or developmental toxicity would be unlikely via the oral route in humans (Foster et al., 2000). The metabolism of DEHP to its metabolite MEHP has been studied in pigs, with the conclusion that MEHP consistently reached the systemic circulation when DEHP was administered orally; however, the kinetic pattern of DEHP was more difficult to characterize (Ljungvall et al., 2004). A study of DINP and DEHP toxicity in cynomolgus monkeys would suggest that rodents are not the best animal model for predicting hepatic toxicity of phthalates in humans (Pugh et al., 2000). Similarly, after studying the toxicokinetics of DINP in rats, one group of investigators has concluded that results of rodent studies of high–molecular weight phthalates may not be very useful in assessing potential human risks (McKee et al., 2002).
Silva et al. (2005) have reported that when DOP was orally administered to adult female rats (300 mg/kg), in addition to phthalic acid, MnOP, and MCPP, a further five urinary DOP oxidative metabolites were measured, again highlighting the need to know the toxicokinetics of the chemical in order to confidently estimate the body burden based on urinary concentrations. Similarly, Calafat et al. (2006a) have also noted that high–molecular weight phthalates such as DEHP metabolize to hydrolytic monoesters which can be further transformed to the more water soluble oxidative metabolites, resulting in more metabolites than the lower molecular weight phthalates.
The absorption of phthalates likely differs by dose and route of administration, as well as between species. For example, while at least 50% of orally administered DEHP is absorbed by rats, absorption of high–molecular weight phthalates is considerably less in primates (Astill, 1989; Rhodes et al., 1986). In addition, the phthalates absorbed by primates may not be distributed to the target organs identified in rodent studies, with the result that at equivalent external exposure levels, testicular doses in rodents may be significantly higher (Kessler et al., 2004; Kurata et al., 2003; McKee et al., 2004). The conclusions of the Toxicology Research Task Group of the Phthalates Esters Panel of the American Chemistry Council are that the new information suggests that humans are unlikely to be more sensitive than rodents and may in fact be substantially less sensitive to the effects of phthalates (McKee et al., 2004). However, further work remains including completion of PBPK models and further assessment of selected subgroups.
Are humans exposed to the parent phthalates or the metabolites, and how might this affect toxicity and comparison between animal and human studies? Is it also to be expected that the toxicokinetics of phthalates will differ between fetuses, infants, and adults? A study of MBP and monobutyl phthalate glucuronide PK in pregnant rats has suggested that neither the chemical (DBP vs. MBP), vehicle (oil vs. aqueous), dose level, and route (oral vs. iv) substantially affected maternal PK (Kremer et al., 2005).
Several authors have also reported large interindividual differences in phthalate metabolite concentrations (Main et al., 2006; Silva et al., 2004) as well as large temporal variations within the same individual, although generally consistent within low-, median-, and high-exposure groups (Hauser et al., 2004). Phthalate metabolite concentrations in a spot urine sample represent only a snapshot in time of exposure for that individual, given the current belief that the majority of metabolites are excreted within 24 h (Lottrup et al., 2006). In rats, more than 90% of the dose of DBP was excreted in urine within 48 h, with the majority excreted within the first 24 h (Tanaka et al., 1978).
After a single oral dose of 48.1 mg of DEHP to a male volunteer, the major metabolite in serum was MEHP, while the major metabolite in urine was MEHHP, followed by MEOHP and MEHP, with 47% of the DEHP dose excreted in urine after 44 h (Koch et al., 2004). In a follow-up study where 0.35 (4.7 µg/kg), 2.15, and 48.5 mg of D4-ring–labeled DEHP were administered orally to a male volunteer, two new metabolites 5cx-MEPP and 2cx-MMHP were measured in urine (Koch et al., 2005). In the latter study, about 75% of the administered DEHP dose was excreted in urine after 2 days and the authors suggested that these secondary oxidized DEHP metabolites and not DEHP or MEHP may be the ultimate developmental toxicants.
Comparison of Animal and Human Exposures to Phthalates
While the exposure route is known for the animal studies, the most important routes and media for phthalate exposure for pregnant women is largely unknown but likely includes air, water, diet, and personal care products. At present, the proportional contribution of the various sources and routes of exposure to phthalates is unknown (Hauser and Calafat, 2005).
In addition, while experimental animal studies are able to carefully control the environment and co-exposures of the pregnant animals, epidemiologic studies are by their nature observational, and exposure to any individual phthalate occurs simultaneously with exposure to other phthalates, as well as other environmental chemicals. Most toxicologic studies evaluate only one chemical at a time and are thus unable to detect adverse outcomes that arise from interactions between concurrent exposures. These co-exposures could affect the dose at which effects are observed as evidenced by a toxicology study where combinations of chemicals produced cumulative, dose-additive effects on tissues (Gray et al., 2006). It is only by the accumulation of evidence from well-designed studies that conclusions can be made about possible etiologic risk factors, ruling out confounding factors that may also be associated with the exposure and health outcome. Other factors that differ between experimental animal studies and epidemiologic studies are the control over the timing and duration of exposure. While it is known that critical windows for reproductive development exist for which animal studies can control, in humans it is often difficult to precisely assess when exposure occurred during pregnancy and its duration.
Exposure in epidemiologic studies does not occur in a controlled environment. In this regard, there is evidence from at least one study that phthalates may interact with other ubiquitous chemicals, such as polychlorinated biphenyls, to produce greater than additive effects on male reproductive health (Hauser et al., 2005). Similarly, antioxidant vitamin supplementation (vitamins C and E) during DEHP treatment has been shown to have a therapeutic effect on DEHP-induced aspermatogenesis in mice (Ablake et al., 2004). In laboratory animal studies, the sampling strategy for selecting pups from each litter for in-depth examination can also influence the outcome of studies, particularly, when the end point is highly variable and only one pup is randomly selected from each litter (Elswick et al., 2000).
In one of the few studies which have measured phthalates in the urine of pregnant women and compared these values with levels in personal air samples, significant correlations were found for diethyl phthalate (DEP) and monoethyl phthalate (MEP), DBP and MBP, and butyl benzyl phthalate (BBzP) and MBzP (Adibi et al., 2003). This study also demonstrated that pregnant women are exposed to considerable levels of phthalates and inhalation was identified as an important route of exposure. In another study, a German group calculated the daily intake of the parent phthalates using urinary metabolite excretion factors and determined a median intake for DEHP of 13.8 µg/kg body weight/day, with 12% of the subjects exceeding the tolerable daily intake (TDI) of 37 µg/kg/day and 31% having values higher than the reference dose of 20 µg/kg/day (Koch et al., 2003b). In a Japanese biomonitoring study, MBP and MEHP were measured in urine samples and the daily intake estimated at 0.22–4.5 µg/kg/day (DBP) and 0.37–7.3 µg/kg/day (DEHP) based on urinary levels of MBP of < 1.8–280 µg/l and MEHP of 0.76–25 µg/l (Itoh et al., 2005).
The current U.S. EPA reference doses for phthalates are 0.1 (DBP), 0.8 (DEP), 0.2 (BBzP), and 0.02 (DEHP) mg/kg/day (USEPA, 1990, 1991, 1993a,b). Marsee et al. (2006) have recently estimated that the exposures observed by Swan et al. (2005) in their epidemiologic study were two orders of magnitude lower than the reference doses assumed to be protective by the U.S. EPA. In one study of breast milk residues of phthalate monoesters, the sum of the three most abundant metabolites (MBP, MEHP, and MINP) was five times the recommended daily dose (Main, IN; Grandjean and Toppari, 2006). A biomonitoring study of Korean women and children has estimated that male children were exposed to 9.9 µg/kg/day DEHP and that the women were exposed to 41.7 µg/kg/day, which exceeded TDI of 37 µg/kg/day established by the European Union Scientific Committee for Toxicity, Ecotoxicity, and the Environment based on reproductive toxicity (Jung-Koo and Mu-Lee, 2005).
While a comprehensive knowledge of the toxicokinetics of the phthalate is important to ensure that all relevant metabolites are included in analysis, uncertainty can also result from the substantial day-to-day and month-to-month variability in an individual's urinary phthalate metabolite levels, with a single urine sample only moderately predictive (sensitivities ranged from 0.56 to 0.74, depending on the metabolite) of a person's exposure over 3 months (Hauser et al., 2004). In addition, both the degree of between- and within-subject variability differed among the phthalate metabolites.
Significant variability has also been reported in infants exposed to DEHP-containing medical devices in neonatal intensive care units, who can have more than five times higher urinary levels of MEHP (median 86 ng/ml) than infants in the lowest exposure group (Green et al., 2005).
How Well Are We Measuring Human Exposure to Phthalates?
Previous sections have identified a number of sources of uncertainty and misclassification of exposure that must be considered when designing or evaluating an epidemiological study of phthalate toxicity and comparing its results to animal studies. Has exposure been measured during the critical window and was the timing of collection standardized? What is the extent of intra-individual variability in exposure and how well would a spot urine sample represent the critical exposure? Given that population distributions of several phthalate metabolites vary by time of day when collected (Silva et al., 2004), collecting a first morning urine void may miss relevant exposure opportunities (Hauser and Calafat, 2005) and result in an underestimate of exposure. Is the study population exposed to the parent phthalates or the metabolites and how might this affect toxicity and comparison between animal and human studies? Are we collecting the appropriate matrix in humans to assess exposure during the critical time period? For example, prenatal exposure to phthalates can be estimated based on measurements in maternal urine, amniotic fluid, and meconium. Based on toxicokinetics, which parent or metabolites should be measured and do these represent the totality of exposure?
To accurately assess human exposure to phthalates, it is essential to have an understanding of how phthalates are metabolized. For example, for DEHP, urinary levels of MEHP are often lower than what would be expected given its high volume of use; however, two other metabolites formed by oxidative metabolism of MEHP (MEOHP and MEHHP) were found in one survey to be four times higher than those of MEHP (Barr et al., 2003). In another survey, the percentages of free and conjugated monoesters excreted in urine differed for various phthalates, with from 6 to 16% of the more lipophilic monoesters (MBP, MBzP, MEHP) excreted and 71% of the most hydrophilic monoester MEP excreted (Silva et al., 2003). Serum levels of MEOHP and MEHHP were comparatively lower than those in urine (Kato et al., 2004). In another study of urine samples from the general population, concentrations of primary and secondary DEHP metabolites were shown to vary by up to three orders of magnitude from phthalate to phthalate and subject to subject (Koch et al., 2003a). These data suggest that for exposure assessment it is critical to measure the most sensitive biomarker, otherwise incorrect conclusions will be drawn about the dose-response relationship.
As MEHP is an environmental contaminant of its own, it is impossible to distinguish between exposure and secondary cross contamination during sampling, handling and processing; however, all secondary oxidized metabolites are not susceptible to contamination (Koch et al., 2006). MEHP also has the shortest half-life of elimination of all metabolites investigated; therefore, basing exposure estimates exclusively on MEHP likely underestimates DEHP exposure (Koch et al., 2006). Also of concern is that the measured amounts of phthalate metabolites may be affected by the degree of esterase activity in the matrix. For example, esterases present in milk (Calafat et al., 2004) or serum (Kato et al., 2003) may hydrolyze phthalates to their hydrolytic monoesters. Furthermore, the samples may be contaminated during collection, storage, or analysis if plastic materials are used, requiring knowledge of the chemical properties of the compounds of interest and composition of the matrix and its potential effects on concentrations of selected analytes prior to developing biomonitoring protocols (Calafat et al., 2006b).
Phthalate Summary
Although it is difficult to compare the phthalate dose administered to animals with the concentration of urinary phthalate metabolites measured in humans, humans are generally considered to be exposed to levels lower than those capable of affecting fetal testis hormone production and masculinization in rats (Sharpe, 2005). The results of epidemiologic studies could challenge this assumption. Based on the biomonitoring studies conducted to date, there are subpopulations who are exposed to phthalates at levels above the reference dose. While animal studies examine one phthalate at a time (generally the parent compound), humans are exposed to mixtures of phthalates, along with a myriad of other xenobiotics. Recent research suggests that measuring only the phthalate monoesters in urine (and ignoring the oxidized metabolites) would substantially underestimate human exposure and therefore human risks. Potential effects of human exposure to mixtures of phthalates along with other environmental chemicals must also be considered when comparing results between epidemiological and toxicological studies.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
It will be clear that a number of issues need to be considered in order to evaluate the quality of the exposure information from a toxicological or epidemiological study for human health risk assessment. Critical issues include whether exposure has been measured at the appropriate time relevant to the health outcome under study and if the assessment of exposure has accounted for all relevant exposures in the study design and risk assessment in both population and laboratory studies. Other considerations include the appropriateness of the media and markers measured in biomonitoring studies, potential contamination of the sample during administration or collection, the stability of the exposure over time, and the reliability of one measurement reflecting the likely exposure during the critical period. Where significant within-subject variability in body burden is observed, multiple biomonitoring collection times must be performed. Interspecies differences in toxicokinetics of specific subgroups (e.g., age, gender), relative sensitivity to chemical mixtures, and the latency period between exposure and observation of effect may also contribute to observed differences between rodent and human studies.
Research efforts should focus on improving the feasibility of including biomonitoring in both animal and human studies to facilitate comparisons between animal and human models and improve exposure assessment in epidemiologic studies. Animal and human studies should measure the same biomarkers, where possible, to facilitate human health risk assessment.
Toxicological studies can contribute to human health risk assessment by identifying the potentially toxic agents in mixtures and the mechanism of action of the agent and its toxicokinetics, while epidemiological studies examine human risks under real-life exposure scenarios. Each type of study has its strengths and weaknesses in how exposure is assessed—the strength of the toxicological dosing regimen can also be a weakness for human risk assessment because it may not reflect real-life human exposures. Exposure assessment in epidemiologic studies while conducted in the target population may be limited in its scope. Insufficient critical examination of the differences in how exposure is measured could explain, at least in some cases, the discordance between toxicologic and epidemiologic study results.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPESTICIDESPHTHALATES AND MALE REPRODUCTIVE...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Ablake M, Itoh M, Terayama H, Hayashi S, Shoji S, Naito M, Takahashi K, Suna S, Jitsunari F. Di-(2-ethylhexyl) phthalate induces severe aspermatogenesis in mice, however, subsequent antioxidant vitamins supplementation accelerates regeneration of the seminiferous epithelium. Int. J. Androl. (2004) 27:274–281.[CrossRef][Web of Science][Medline]
Adibi JJ, Perera FP, Jedrychowski W, Camann DE, Barr D, Jacek R, Whyatt RM. Prenatal exposures to phthalates among women in New York City and Krakow, Poland. Environ. Health Perspect. (2003) 111:1719–1722.[Web of Science][Medline]
Alavanja MCR, Sandler DP, McMaster SB, Hoar Zahm S, McDonnell CJ, Lynch CF, Pennybacker M, Rothman N, Dosemeci M, Bond AE, Blair A. The Agricultural Health Study. Environ. Health Perspect. (1996) 104:362–369.[Web of Science][Medline]
Arbuckle TE, Cole DC, Ritter L, Ripley BD. Farm children's exposure to herbicides: Comparison of biomonitoring and questionnaire data. Epidemiology (2004) 15:187–194.[CrossRef][Web of Science][Medline]
Astill B. Metabolism of DEHP: Effects of prefeeding and dose variation, and comparative studies in rodents and the cynomolgus monkey (CMA studies). Drug Metab. Rev. (1989) 21:35–53.[Web of Science][Medline]
Baris D, Zahm SH, Cantor KP, Blair A. Agricultural use of DDT and risk of non-Hodgkin's lymphoma: Pooled analysis of three case-control studies in the United States. Occup. Environ. Med. (1998) 55:522–527.
Barr DB, Silva JM, Kato K, Reidy JA, Malek NA, Hurtz D, Sadowski M, Needham LL, Calafat AM. Assessing human exposure to phthalates using monoesters and their oxidized metabolites as biomarkers. Environ. Health Perspect. (2003) 111:1148–1151.[Web of Science][Medline]
Barr DB, Wilder LC, Caudill SP, Gonzalez AJ, Needham LL, Pirkle JL. Urinary creatinine concentrations in the U.S. population: Implications for urinary biological monitoring measurements. Environ. Health Perspect. (2005) 113:192–200.[Web of Science][Medline]
Blair A, Sandler D, Lynch CF, Roland A, Wintersteen W, Steen WC, Samanic C, Dosemeci M, Alavanja MCR. Reliability of reporting on lifestyle and agricultural factors by a sample of participants in the agricultural health study from Iowa. Epidemiology (2002) 13:94–99.[CrossRef][Web of Science][Medline]
Calafat AM, Silva MJ, Reidy JA, Gray LE, Samandar E, Preau JL, Herbert AR, Needham LL. Mono-(3-carboxypropyl) phthalate, a metabolite of di-n-octyl phthalate. J. Toxicol. Environ. Health A (2006a) 69:215–227.[CrossRef][Web of Science][Medline]
Calafat AM, Slakman AR, Silva MJ, Herbert AR, Needham LL. Automated solid phase extraction and quantitative analysis of human milk for 13 phthalate metabolites. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. (2004) 805:49–56.[CrossRef][Web of Science][Medline]
Calafat AM, Ye X, Silva MJ, Kuklemyik Z, Needham LL. Human exposure assessment to environmental chemicals using biomonitoring. Int. J. Androl. (2006b) 29:166–171.[CrossRef][Web of Science][Medline]
Calderon RL. Measuring risks in humans: The promise and practice of epidemiology. Food Chem. Toxicol. (2000) 38:S59–S63.[CrossRef][Web of Science][Medline]
Coble J, Arbuckle T, Lee W, Alavanja M, Dosemeci M. The validation of a pesticide exposure algorithm using biological monitoring results. J. Occup. Environ. Hyg. (2005) 2:194–201.[CrossRef][Web of Science][Medline]
Daniels JL, Olshan AF, Savitz DA. Pesticides and childhood cancer. Environ. Health Perspect. (1997) 105:1068–1077.[Web of Science][Medline]
Dosemeci M, Alavanja M, Rowland A, Mage D, Zahm S, Rothman N, Lubin J, Hoppin J, Sandler D, Blair A. A quantitative approach for estimating exposure to pesticides in the agricultural health study. Ann. Occup. Hyg. (2002) 46:245–260.
Duty SM, Silva MJ, Barr DB, Brock JW, Ryan L, Chen Z, Herrick RF, Christiani DC, Hauser R. Phthalate exposure and human semen parameters. Epidemiology (2003) 14:269–277.[CrossRef][Web of Science][Medline]
Elswick BA, Welsch F, Janszen DB. Effect of different sampling designs on outcome of endocrine disruptor studies. Reprod. Toxicol. (2000) 14:359–367.[CrossRef][Web of Science][Medline]
Ema M, Miyawaki E. Adverse effects on development of the reproductive system in male offspring of rats given monobutyl phthalate, a metabolite of dibutyl phthalate, during late pregnancy. Reprod. Toxicol. (2001) 15:189–194.[CrossRef][Web of Science][Medline]
Fennell TR, Krol WL, Sumner SC, Snyder RW. Pharmacokinetics of dibutylphthalate in pregnant rats. Toxicol. Sci. (2004) 82:407–418.
Fisher JS, Macpherson S, Marchetti N, Sharpe RM. Human testicular dysgenesis syndrome: A possible model based on in-utero exposure of the rat to dibutyl phthalate. Hum. Reprod. (2003) 18:1383–1394.
Foster PM, Cattley RC, Mylchreest E. Effects of di-n-butyl phthalate (DBP) on male reproductive development in the rat: Implications for human risk assessment. Food Chem. Toxicol. (2000) 38(Suppl.):S97–S99.[CrossRef][Web of Science][Medline]
Ginsberg G, Hattis D, Sonawane B, Russ A, Banati P, Kozlak M, Smolenski S, Goble R. Evaluation of child/adult pharmacokinetic differences from a database derived from the therapeutic drug literature. Toxicol. Sci. (2002) 66:185–200.
Grandjean P, Toppari J. General discussion: Possible effects of phthalate exposure in doses relevant for humans. Int. J. Androl. (2006) 29:181–185.[CrossRef][Web of Science]
Gray LE Jr, Ostby J, Furr J, Price M, Veeramachaneni DN, Parks L. Perinatal exposure to the phthalates DEHP, BBP, and DINP, but not DEP, DMP, or DOTP, alters sexual differentiation of the male rat. Toxicol. Sci. (2000) 58:350–365.
Gray LE Jr, Wilson VS, Stoker T, Lambright C, Furr J, Noriega N, Howdeshell K, Ankley GT, Guillette L. Adverse effects of environmental antiandrogens and androgens on reproductive development in mammals. Int. J. Androl. (2006) 29:96–104.[CrossRef][Web of Science][Medline]
Green R, Hauser R, Calafat AM, Weuve J, Schettler T, Ringer S, Huttner K, Hu H. Use of di(2-ethylhexyl) phthalate-containing medical products and urinary levels of mono(2-ethylhexyl) phthalate in neonatal intensive care unit infants. Environ. Health Perspect. (2005) 113:1222–1225.[Web of Science][Medline]
Hattis D, Ginsberg G, Sonawane B, Smolenski S, Russ A, Kozlak M, Goble R. Differences in pharmacokinetics between children and adults—II. Children's variability in drug elimination half-lives and in some parameters needed for physiologically-based pharmacokinetic modeling. Risk Anal. (2003) 23:117–142.[CrossRef][Web of Science][Medline]
Hauser R, Calafat AM. Phthalates and human health. Occup. Environ. Med. (2005) 62:806–818.
Hauser R, Meeker JD, Park S, Silva MJ, Calafat AM. Temporal variability of urinary phthalate metabolite levels in men of reproductive age. Environ. Health Perspect. (2004) 112:1734–1740.[Web of Science][Medline]
Hauser R, Williams P, Altshud L, Calafat AM. Evidence of interaction between polychlorinated biphenyls and phthalates in relation to human sperm motility. Environ. Health Perspect. (2005) 113:425–430.[Web of Science][Medline]
Hill AB. The environment and disease: Association or causation. Proc. R. Soc. Med. (1965) 58:295–300.[Web of Science][Medline]
Hoppin JA, Brock JW, Davis BJ, Baird DD. Reproducibility of urinary phthalate metabolites in first morning urine samples. Environ. Health Perspect. (2002) 110:515–518.[Web of Science][Medline]
Hunt PA, Koechler KE, Susiarjo M, Hodges CA, Ilagan A, Voigt RC, Thomas S, Thomas BF, Hassold TJ. Bisphenol A exposure causes meiotic aneuploidy in the female mouse. Curr. Biol. (2003) 13:546–553.[CrossRef][Web of Science][Medline]
Itoh H, Yoshida K, Masunaga S. Evaluation of the effect of government control of human exposure to two phthalates in Japan using a urinary biomarker approach. Int. J. Hyg. Environ. Health (2005) 208:237–245.[CrossRef][Web of Science][Medline]
Jung-Koo H, Mu-Lee B. Human monitoring of phthalates and risk assessment. J. Toxicol. Environ. Health A (2005) 68:1379–1392.[CrossRef][Web of Science][Medline]
Kato K, Silva MJ, Brock JW, Reidy JA, Malek NA, Hodge CC, Nakazawa H, Needham LL, Barr DB. Quantitative detection of nine phthalate metabolites in human serum using reversed-phase high-performance liquid chromatography-electrospray ionization-tandem mass spectrometry. J. Anal. Toxicol. (2003) 27:284–289.[Web of Science][Medline]
Kato K, Silva MJ, Reidy JA, Hurtz D III, Malek NA, Needham LL, Nakazawa H, Barr DB, Calafat AM. Mono(2-ethyl-5-hydroxyhexyl) phthalate and mono-(2-ethyl-5-oxohexyl) phthalate as biomarkers for human exposure assessment to di-e(2-ethylhexyl) phthalate. Environ. Health Perspect. (2004) 112:327–330.[Web of Science][Medline]
Kessler W, Numtip W, Grote K, Csanady G, Chahoud I, Filser J. Body burden of di(2-ethylhexyl) phthalate (DEHP) and its primary metabolite mono(2-ethylhexyl) phthalate (MEHP) in pregnant and non-pregnant rats and marmosets. Toxicol. Appl. Pharmacol. (2004) 195:142–153.[CrossRef][Web of Science][Medline]
Koch HM, Bolt HM, Angerer J. Bi(2-ethylhexyl) phthalate (DEHP) metabolites in human urine and serum after a single oral dose of deuterium-labeled DEHP. Arch. Toxicol. (2004) 78:123–130.[CrossRef][Web of Science][Medline]
Koch HM, Bolt HM, Preuss R, Angerer J. New metabolites of di(2-ethylhexyl) phthalate (DEHP) in human urine and serum after single oral doses of deuterium-labelled DEHP. Arch. Toxicol. (2005) 79:367–376.[CrossRef][Web of Science][Medline]
Koch HM, Drexler H, Angerer J. An estimation of the daily intake of di(2-ethylhexyl) phthalate (DEHP) and other phthalates in the general population. Int. J. Hyg. Environ. Health (2003b) 206:77–83.[CrossRef][Web of Science][Medline]
Koch HM, Preuss R, Angerer J. Di(2-ethylhexyl)phthalate (DEHP): Human metabolism and internal exposure—An update and latest results. Int. J. Androl. (2006) 29:155–165.[CrossRef][Web of Science][Medline]
Koch HM, Rossbach B, Drexler H, Angerer J. Internal exposure of the general population to DEHP and other phthalates—Determination of secondary and primary phthalate monoester metabolites in urine. Environ. Res. (2003a) 93:177–185.[Medline]
Koo JW, Parham F, Kohn MC, Masten SA, Brock JW, Needham LL, Portier CJ. The association between biomarker-based exposure estimates for phthalates and demographic factors in a human reference population. Environ. Health Perspect. (2002) 110:405–410.[Web of Science][Medline]
Kremer JJ, Williams CC, Parkinson HD, Borghoff SJ. Pharmacokinetics of monobutylphthalate, the active metabolite of di-n-butylphthalate in pregnant rats. Toxicol. Lett. (2005) 159:144–153.[CrossRef][Web of Science][Medline]
Kurata Y, Makinodan F, Okada M, Kawasuso T, David R, Gans G, Regnier J, Katoh M. Blood concentration and tissue distribution of 14C-di(2-ethylhexyl) phthalate (DEHP) in juvenile and adult common marmoset. Toxicologist (2003) 72:384–385.
Leiss JK, Savitz DA. Home pesticide use and childhood cancer: A case-control study. Am. J. Public Health (1995) 85:249–252.
Ljungvall K, Tienpont B, David F, Magnusson U, Torneke K. Kinetics of orally administered di(2-ethylhexyl) phthalate and its metabolite mono(2-ethylhexyl) phthalate in male pigs. Arch. Toxicol. (2004) 78:384–389.[Web of Science][Medline]
Lottrup G, Andersson A-M, Leffers H, Mortensen GK, Toppair J, Skakkebaek NE, Main KM. Possible impact of phthalates on infant reproductive health. Int. J. Androl. (2006) 29:172–180.[CrossRef][Web of Science][Medline]
Main KM, Mortensen GK, Kaleva MM, Boisen KA, Damgaard IN, Chellakooty M, Schmidt IM, Suomi A-M, Virtanen HE, Petersen JH, et al. Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age. Environ. Health Perspect. (2006) 114:270–276.[Web of Science][Medline]
Marsee K, Woodruff TJ, Axelrad DA, Calafat AM, Swan SH. Estimated daily phthalate exposures in a population of mothers of male infants exhibiting reduced anogenital distance. Environ. Health Perspect. (2006) 114:805–809.[Web of Science][Medline]
McKee RH, Butala JH, David RM, Gans G. NTP center for the evaluation of risks to human reproduction reports on phthalates: Addressing the data gaps. Reprod. Toxicol. (2004) 18:1–22.[CrossRef][Web of Science][Medline]
McKee RH, El-Hawari M, Stoltz M, Pallas F, Lington AW. Absorption, disposition and metabolism of di-isononyl phthalate (DINP) in F-344 rats. J. Appl. Toxicol. (2002) 22:293–302.[CrossRef][Web of Science][Medline]
Mylchreest E, Wallace DG, Cattley RC, Foster PM. Dose-dependent alterations in androgen-regulated male reproductive development in rats exposed to Di(n-butyl) phthalate during late gestation. Toxicol. Sci. (2000) 55:143–151.
National Research Council. Risk Assessment in the Federal Government: Managing the Process (1983) Washington, DC: National Academy Press.
Ontario College of Family Physicians. Pesticides Literature Review (2004) Available at: http://www.ocfp.on.ca/English/OCFP/Communications/CurrentIssues/Pesticides/default.asp?s=1.
Parks LG, Ostby JS, Lambright CR, Abbott BD, Klinefelter GR, Barlow NJ, Gray LE Jr. The plasticizer diethylhexyl phthalate induces malformations by decreasing fetal testosterone synthesis during sexual differentiation in the male rat. Toxicol. Sci. (2000) 58:339–349.
Pest Management Regulatory Agency (PMRA). Re-Evaluation of the Lawn and Turf Uses of 2,4-D. PACR 2005-01 (2005) PMRA (2005) = USEPA (2005) = Available at: http://www.pmraarla.gc.ca/english/pdf/pacr/pacr2005-01-e.pdf.
Pugh G Jr, Isenberg JS, Kamendulis LM, Ackley DC, Clare LJ, Brown R, Lington AW, Smith JH, Klaunig JE. Effects of di-isononyl phthalate, di-2-ethylhexyl phthalate and clofibrate in cynomolgus monkeys. Toxicol. Sci. (2000) 56:181–188.
Rhodes C, Orton T, Pratt I, Batten P, Bratt H, Jackson SJ, Elcombe CR. Comparative pharmacokinetics and subacute toxicity of di(2-ethylhexyl) phthalate (DEHP) in rats and marmosets: Extrapolation of effects in rodents to man. Environ. Health Perspect. (1986) 65:299–308.[Web of Science][Medline]
Rice DC. The US EPA reference dose for methylmercury: Sources of uncertainty. Environ. Res. (2004) 95:406–413.[Medline]
Rice DC. Assessing the effects of environmental toxicant exposure in developmental epidemiological studies: Issues for risk assessment. Neurotoxicology (2005) 26:483–489.[CrossRef][Web of Science][Medline]
Ritter L. Cancer and exposure characterization: Application of the risk assessment paradigm. In Proceedings of the 223rd National Meeting of the American Chemical Societ,April, 2002, Orlando, Florida (2002).
Rothenberg SJ, Rothenberg JC. Testing the dose-response specifications in epidemiology: Public health and policy consequences for lead. Environ. Health Perspect. (2005) 113:1190–1195.[Web of Science][Medline]
Rothman KJ. Modern Epidemiology (1986) Boston, MA: Little, Brown and Company.
Sears M, Walker RC, Van der Jagt RHC, Claman P. Pesticide assessment: Protecting public health on the home turf. Paediatr. Child Health (2006) 11:229–234.
Sharpe RM. Phthalate exposure during pregnancy and lower anogenital index in boys: Wider implications for the general population? Environ. Health Perspect. (2005) 113:A504–A505.[Web of Science][Medline]
Sharpe RM, Turner KJ, Sumpter JP. Endocrine disruptors and testis development. Environ. Health Perspect. (1998) 106:A220.[Web of Science][Medline]
Silva MJ, Barr DB, Reidy JA, Kato K, Malek NA, Hodge CC, Hurtz D III, Calafat AM, Needham LL, Brock JW. Glucuronidation patterns of common urinary and serum monoester phthalate metabolites. Arch. Toxicol. (2003) 77:561–567.[CrossRef][Web of Science][Medline]
Silva MJ, Barr DB, Reidy JA, Malek NA, Hodge CC, Caudill SP, Brock JW, Needham LL, Calafat AM. Urinary levels of seven phthalate metabolites in the U.S. population from the National Health and Nutrition Examination Survey (NHANES) 1999-2000. Environ. Health Perspect. (2004) 112:331–338.[Web of Science][Medline]
Silva MJ, Kato K, Gray EL, Wolf C, Needham LL, Calafat AM. Urinary metabolites of di-n-octyl phthalate in rats. Toxicology (2005) 210:123–133.[CrossRef][Web of Science][Medline]
Silva MJ, Reidy JA, Preau JL Jr, Needham LL, Calafat AM. Oxidative metabolites of di-isononyl phthalate as biomarkers for human exposure assessment. Environ. Health Perspect. (2006) 114:1158–1161.[Web of Science][Medline]
Skakkeback NE. Endocrine disruptors and testicular dysgenesis syndrome. Horm. Res. (2002) 57(Suppl. 2):43.[Web of Science][Medline]
Stewart PA. Challenges to retrospective exposure assessment. Challenges to retrospective exposure assessment. Scand. J. Work Environ. Health (1999) 25:505–510.[Web of Science][Medline]
Swan SH, Main KM, Liu F, Stewart SL, Kruse RL, Calafat AM, Mao CS, Redmon JB, Ternand CL, Sullivan S, et al. Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. (2005) 113:1056–1061.[Web of Science][Medline]
Tanaka A, Matsumoto A, Yamaha T. Biochemical studies on phthalic esters. III Metabolism of dibutyl phthalate (DBP) in animals. Toxicology (1978) 9:109–123.[CrossRef][Web of Science][Medline]
Thompson CJ, Ross SM, Gaido KW. Di(n-butyl) phthalate impairs cholesterol transport and steroidogenesis in the fetal rat testis through a rapid and reversible mechanism. Endocrinology (2004) 145:1227–1237.
U.S. Environmental Protection Agency (USEPA). Integrated Risk Information System: Dibutyl Phthalate (1990) Available at: http://cfpub.epa.gov/iris/quickview.cfm?substance_nmbr=0038.
U.S. Environmental Protection Agency (USEPA). Integrated Risk Information System: Di(2-ethylhexyl)Phthalate (DEHP) (1991) Available at: http://cfpub.epa.gov/iris/quickview.cfm?substance_nmbr=0014.
U.S. Environmental Protection Agency (USEPA). Integrated Risk Information System: Diethylphthalate (1993a) Available at: http://cfpub.epa.gov/iris/quickview.cfm?substance_nmbr=0226.
U.S. Environmental Protection Agency (USEPA). Integrated Risk Information System: Butyl Benzyl Phthalate (1993b) Available at: http://cfpub.epa.gov/iris/quickview.cfm?substance_nmbr=0293.
U.S. Environmental Protection Agency (USEPA). Reregistration Eligibility Decision for 2,4-D. 738-R-05-002 (2005) USEPA (2005) = http://www.epa.gov/REDs/24d_red.pdf.
Voss C, Zerban H, Bannasch P, Berger MR. Lifelong exposure to di-(2-ethylhexyl) phthalate induces tumors in liver and testes of Sprague-Dawley rats. Toxicology (2005) 206:359–371.[CrossRef][Web of Science][Medline]
Wilson VS, Lambright C, Furr J, Ostby J, Wood C, Held G, Gray LE Jr. Phthalate ester-induced gubernacular lesions are associated with reduced insl3 gene expression in the fetal rat testis. Toxicol. Lett. (2004) 146:207–215.[CrossRef][Web of Science][Medline]
Zahm SH, Ward MH. Pesticides and childhood cancer. Environ. Health Perspect. (1998) 106(Suppl. 3):893–908.[CrossRef][Web of Science][Medline]
Zhang Y, Jiang X, Chen B. Reproductive and developmental toxicity in F1 Sprague-Dawley male rats exposed to di-n-butyl phthalate in utero and during lactation and determination of its NOAEL. Reprod. Toxicol. (2004) 18:669–676.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E MacFarlane, D Glass, and L Fritschi Is farm-related job title an adequate surrogate for pesticide exposure in occupational cancer epidemiology? Occup. Environ. Med., August 1, 2009; 66(8): 497 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lebailly, V. Bouchart, I. Baldi, Y. Lecluse, N. Heutte, A. Gislard, and J.-P. Malas Exposure to Pesticides in Open-field Farming in France Ann. Hyg., January 1, 2009; 53(1): 69 - 81. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||