Document Nek8dV4v4R4Kozx9ma96GdbZE

CinriTtiocxalicRoleovgiyews http://mfQmahearthrare.com/txc ISSN: 1040-8444 (print), 1547-6898 (electronic) Crit Rev Toxicol, 2015; 45(7): 531-641 2015 informa Healthcare USA, Inc. 001:10.3109/10408444.2015.1043976 informa healthcare REVIEW ARTICLE A review of epidemiologic studies of low-level exposures to organophosphorus insecticides in non-occupational populations Richard Reiss1, Ellen T. Chang2-3, Rudy J. Richardson4,5, and Michael Goodman6 1Exponent, Alexandria, VA, USA, 2Exponent, Menlo Park, CA, USA, 3Division of Epidemiology, Department of Health Research and Policy, Stanford University School of Medicine, Stanford, CA, USA,4Department of, Environmental Health Sciences, 5Department of Neurology, University of Michigan, Ann Arbor, Mi, USA and 6Department o f Epidemiology, Emory University, Rollins School of Public Health, Atlanta, CA, USA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Abstract This paper systematically reviews epidemiologic studies related to iow-level non-occupational exposures to organophosphorus (OP) insecticides. Many o f the studies evaluate levels o f mater nal OP metabolites and subsequent health outcomes in offspring. The studies focused primarily on birth outcomes (e.g,, infant body weight or head circumference) and neurodevelopmental (e.g., mental and psychomotor) testing results. The evidence from these studies was reviewed under the Bradford Hill guidelines. Most o f the studies assessing exposure based on urinary lev els of OP insecticide metabolites used only one or two measurements during pregnancy. The potential for exposure misclassification with this method is largely due to (1) preformed metabo lites that are ingested with food, (25 the short elimination half-life o f OP insecticides, and (3) lack of specificity to particular OP insecticides for many o f the metabolites. For birth outcomes, the majority of reported results are not statistically significant, and the associations are inconsistent w ithin and across studies.There is more within-study consistency for some o fth e neurodevelop mental testing results, although few associations were examined across several studies. These associations are generally weak, have been replicated only to a limited extent, and require fur ther confirmation before they can be considered established,The OP insecticide levels measured in the epidemiologic studies are too low to cause biologically meaningful acetylcholinesterase Inhibition, the most widely used metric for OP insecticide toxicity. Overall, the available evidence does not establish that low-level exposures to OP insecticides cause adverse birth outcomes or neurodevelopmental problems In humans. Keywords Bradford Hill, epidemiology, insecticides, organophosphorus, pesticides History Received 13 November 2014 Revised 14 April 2015 Accepted 19 April 2015 Published online 7 July 2015 Table of Contents Abstract................................................................................................ 531 Introduction........................................................................................ 532 Scope o f review................................................................................... 532 Overarching issues. ... ... ............ ... ... ... ............ ... ... ... .......... ... 533 Biomarker validity... ... ............ ... ... ... ............ ... ... ... .......... ... 533 Direct formation o f OP metabolites in food ... ... ... .......... ... 533 Rapid metabolism o f OP insecticides........................................ 534 DAP metabolites are non-specific............................................. 534 Intra-individual variability in urinary DAP levels....................... 534 Potential Impact of exposure misclassification........................ 534 Dose-response................................................................................ 534 Confounding and bias ................................................................... 536 Review o f epidemiologic studies .......... ... ... ... .......... ... ... ... ... 537 Birth outcom es .......... ... ... ... .......... ... ... ... .......... ... ... ... ... 537 Columbia center for children's environment and health, ... ... 537 Mount Sinai children's environmental cohort s tu d y ............... 546 Center for the health assessment o f mothers and children ofSalinas....................................................................................... 566 New jersey birth coh ort.............................................................. 568 Shanghai birth cohort................................................................. 568 Address for correspondence: Richard Reiss, Exponent, 1800 Diagonal Road, Suite 500, Alexandria, VA 22310, USA. Tel: + (571) 227-7228. E-mail: rreiss @exponent.com Health outcomes and measures o f the environment study ... 568 Zhejiang birth coh ort.................................................................. 569 Bradford Hill evaluation o f weight of evidence....................... 569 Strength.................................................................................... 569 Consistency.............................................................................. 569 Temporality.............................................................................. 570 Biological gradient........... ... ... ... .......... ... ... ... ... 570 Plausibility... ... ... ........ ... ... ... .......... ... ... ... ... 570 Coherence... ... ... ........ ... ... ... .......... ... ... ... ... 570 Specificity, experiment, and analogy..................................... 571 Neurodevelopmental outcomes.................................................... 571 Columbia center for children's environment and health......... 571 Mount Sinai children's environmental cohort s tu d y ............... 625 Center for the health assessment o f mothers and children ofSalinas....................................................................................... 627 Health outcomes and measures o f the environment study ... 629 Children pesticide survey ... .......... . ... ... ... .......... . ... ... ... ... 629 National health and nutrition examination survey.. ... ... ... ... 630 Shanghai cross-sectional study.................................................. 630 Canadian health measures survey............................................. 631 Early life exposed in Mexico to environmental toxicants study 631 Shenyang birth cohort ............................................................... 631 Bradford Hill evaluation o f weight of evidence....................... 632 Strength.................................................................................... 632 Consistency. ... ... ........ ... ... ... .......... ... ... ... ... 632 Temporality. ... ... ........ ... ... ... .......... ... ... ... ... 634 BlflHTS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00001 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 532 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 Biological gradient................................................................... 634 Plausibility and coherence...................................................... 634 Specificity, experiment,and analogy..................................... 636 Discussion............................................................................................ 636 Conclusions ........................................................................................ 638 Acknowledgements........................................................................... 638 Declaration o f interests..... . ... ... ............. . ... ... ............. . ... ... ... 638 References.. ... ... ... .......... ... ... ... .......... ... ... ... .......... ... ... ... 639 found that the epidemiologic literature did not support causal effects for pesticides, and that effects found in toxicology stud ies were generally seen at doses similar to or higher than points of departure used in regulatory risk assessments. This review is the first to address potential effects of all OP insecticides from epidemiologic studies with low-level exposures. To evaluate the scientific evidence for a conclusion regard Introduction ing causality, we used die Bradford Hill guidelines, including strength of association, consistency, temporality, biological gra Organophosphorus (OP) insecticides, or their oxon metabo lites, persistently inactivate acetylcholinesterase (AChE), an enzyme involved in neurotransmission in insects as well as humans and other animals. OP insecticides are used widely around the world. Most studies of the adverse human health effects of exposure to OP insecticides have focused on occupa tional or other high-dose exposures, including acute poisoning. Acute clinical effects result from AChE inhibition at synapses in the central nervous system, autonomic nervous system, and neuromuscular junction (Eddleston et al. 2008). Over the last decade, a number of epidemiologic studies dient, plausibility, coherence with toxicological evidence, spec ificity, experiment, and analogy (Hill 1965). The manuscript also includes a detailed evaluation of the validity of the urinary biomarkers used in the epidemiologic studies, and reviews the plausibility of the associations by comparing ani mat and limited human toxicology data with the OP insecticide levels observed in the epidemiologic studies. Potential confounding and bias are also evaluated. The data are then assembled to assess over all evidence for and against a causal relationship between lowlevel exposure to OP insecticides and adverse birth outcomes or neurodevelopmental problems in humans. have been published that evaluate the potential health effects of OP insecticides in populations with little or no occupational Scope of review exposure. These epidemiologic studies have most frequently To identify the relevant studies on low-level OP metabolites and evaluated birth outcomes, such as infant body weight and head human health outcomes, we used PubMed to search MEDLINE circumference, or results of neurodevelopmental tests that using keywords and keyword roots, including organophosph*, measure mental and psychomotor function. The primary expo specific metabolites (e.g., dialkyIphosphate* or dialkyl phos sure pathways for subjects in these studies likely include diet, phate), specific OP insecticides (e.g,, chlorpyrifos, diazinon, residential use, and in some cases, proximity to agricultural malathion, parathion, orphosmet), and various age groups (e.g., operations. Exposures in these studies are estimated primar child*, infan*, toddler*, birth*, men, women, or adult*). Based ily by measuring OP insecticide biomarkers or degradation on a review' o f titles and abstracts, we excluded more than 1500 products (referred to in this article as "OP metabolites") in articles that presented animal and in vitro studies, biomonitor urine and blood. In several study populations, such markers ing studies, and other non-epidemiologic studies, including case have been measured at levels that are sufficiently low to indi reports, commentaries, and reviews (some o f which were exam cate that exposure to OP insecticides originates predominantly ined to identify references missed by the electronic search). from dietary sources (Berman et al. 2013, Lu et al. 2008). After reviewing full-text articles, we further excluded 40 studies Risk assessments in Europe and the United States have of occupational or para-occupational (i.e., take-home) exposure concluded that dietary exposure to OP insecticides appears to OP insecticides, exposure by poisoning, exposure by pedicu generally to be safe (Boon et al. 2008, Claeys et al. 2008, losis treatment, exposure by aerial residential or illegal indoor Jensen et al. 2003, Jensen et al. 2009, Nougadere et al. 2012). residential spraying, exposure to pesticides or insecticides not Nevertheless, several recent epidemiologic studies that mea specific to OP compounds, and paraoxonase 1 (PON!) geno sured OP metabolites in blood or urine suggest associations type or PON1 enzyme activity without specific evaluation of between low-dose exposure to OP insecticides and adverse OP insecticide exposure. We further excluded 31 studies that human health effects. Most of these studies have focused on estimated OP exposure based on self-reported or geographic OP insecticide metabolite levels in utero, which is believed data, and those that estimated associations with health catego to be the critical exposure period for human neurological ries that were evaluated in fewer than three independent studies, development (Rice and Barone 2000) and is, by definition, thereby providing an insufficient basis for a weight-of-evidence the only relevant exposure period for birth outcomes. Given evaluation. Based on this last consideration, the two endpoint the widespread use of OP insecticides and consumption of categories of interest in this review are birth outcomes and OP-treated foods, understanding the potential human health results of neurodevelopmental testing. We ultimately included impact of low-dose exposure to OP insecticides is important 31 epidemiologic studies-- 11 studies of birth outcomes and from a public health and regulatory standpoint. We undertook 20 studies of neurodevelopmental outcomes--in this review. this systematic review of epidemiologic studies of low-level Study characteristics--including study name, location, OP metabolites to evaluate the existing evidence on associa design, description and number of subjects, follow-up time, tions with adverse human health outcomes. A few previous exposure assessment methods, outcome assessment methods, papers have reviewed the epidemiologic literature specific to point and interval estimates of association between specific chlorpyrifos for neurobehavioral outcomes (Li et al. 2012) exposures and outcomes of interest, and adjustment factors-- and fetal growth outcomes (Mink et al. 2012), and found no were abstracted from each relevant study, and independently compelling evidence of effects. Burns et al. (2013) reviewed checked by another reviewer for accuracy, individual stud animal toxicology and epidemiologic data for neurodevelop ies were evaluated with respect to strength of study design, mental outcomes and all classes of pesticides. The researchers exposure and outcome assessment, potential for confounding SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00002 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 533 and bias, role of random error or chance, and interpretation compounds, DAPs are virtually non-toxic (Chen et al. 2013). of results. To evaluate the overall weight of epidemiologic Some studies used chemical-specific urinary metabolites, evidence, we used the framework of the Bradford Hill guide including malathion dicarboxylic acid (MDA, a metabolite of lines (Hill 1965). The Bradford Hill guidelines are one of the malathion) and 3,5,6-trichloro-2-pyridinol (TCP, a metabo most common and established methods of assessing evidence lite of chlorpyrifos), as OP insecticide exposure biomarkers. for a causal relationship between an exposure and a disease These chemical-specific metabolites also form directly in food (Gordis 2013). We assessed the study results relative to each (Morgan et al, 2011, Chen et al. 2012) and are also relatively of the Bradford Hill guidelines, separately for birth outcomes non-toxic compared with their parent compounds or active and neurodevelopment. These aspects were used as consider metabolites (Chen et al. 2012, Eaton et al. 2008) C ations, but not as strict criteria in a checklist fashion, to guide Several studies have demonstrated that most dietary expo our evaluation of causality. sures are actually to the OP metabolites and not to the parent Overarching issues compounds. Zhang et al. (2008) measured OP and DAP levels on 153 produce samples known to be contaminated with OP Before proceeding to a review of the individual studies, three overarching issues need to be discussed. First, most studies use urinary levels of OP insecticide metabolites to classify expo sures. Therefore, we discuss the validity o f exposure assessment using these urinary biomarkers. Second, the OP insecticide exposure levels of the study subjects are generally lower than those previously identified as harmful. Therefore, we briefly review the extensive animal toxicology and limited human toxicology data to evaluate die plausibility of the associations observed in epidemiologic studies. Third, in any epidemiologic study, confounding and bias should be considered as potential explanations for an observed result (Gordis 2013). insecticides. The mean concentrations of OP insecticides and DAP residues were 1.2 and 2.0 nmol/g, respectively. On a molar basis, more than 60% of the total residues were DAPs, in addition, 60% of the samples contained higher DAP than OP residues. The mole fraction of DAPs across the samples varied widely, ranging from 0.02 to 0.99. Zhang et al. (2008) found that the mole ratio of DAPs to parent OP insecticides was both produce-specific and chemical-specific, with higher ratios for diazinon, phosmet, chlorpyrifos, azinphos-methyl, and malathion. When measured on strawberries, the ratio of DAPs to parent insecticide (malathion) increased with time since application, indicating continuous transformation. The mole ratio of DAPs to malathion was 1.4 one day after applica Biomarker validity tion, and increased to 8.7 after 9 days. Morgan et al. (2011) measured chlorpyrifos and TCP levels Most epidemiologic studies that use biomarkers of OP insec in food from homes and daycare centers of 127 Ohio preschool ticide exposure rely on urinary measurements of OP metabo lites. The metabolites include six dialkylphosphates (DAPs): dimethylphosphate (DMP), dimethylthiophosphate (DMTP), dimethyldithiophosphate (DMDTP), diethylphosphate (DEP), diethylthiophosphate (DETP), and diethyldithiophosphate children. The mean chlorpyrifos residues w;ere 0.4 ng/g in homes (n = 125) and 0.2 ng/g in daycare centers (n = 29). The mean TCP residues were 2.6 ng/g in homes (n = 127) and 2.8 ng/g in daycare centers (n = 29). Thus, the TCP residues were significantly higher than the chlorpyrifos residues. Moreover, (DEDTP). The fust three are commonly grouped as DMPs, the Pearson correlation coefficient for dietary chlorpyrifos and the latter three are commonly grouped as DEPs. In a few studies, metabolites of specific OP insecticides (usually chlorpyrifos or malathion) were measured. Use of these metabo and excreted urinary TCP was only 0.30, meaning that dietary chlorpyrifos exposure explained only about 9% of the vari ability in excreted urinary TCP. lites as exposure biomarkers has the potential for exposure misclassification, for the following reasons: OP metabolites formed directly on or in food are well absorbed orally and cannot be distinguished from those formed following absorption. Rapid metabolism of OP insecticides and their metabolites results in high intra-individual variation in levels, such that single samples may not reflect past or long-term average exposure. DAP metabolites are not specific to individual OP insecti cides, and there is a vast range of toxicity across different compounds. Variability in exposure measurements across studies dimin ishes the comparabili ty of results. Chen et al. (2012) measured malathion and its transfor mation products, including the DAPs, MDA, and malathion monocarboxylic acid (MMA), in 157 produce samples. The samples had been confirmed previously to contain detectable malathion, but no detectable levels of other OP insecticides. The mean malathion residue was 0.60 nmol/g, and the mean preformed metabolite residue was 3.29 nmol/g. The mole fraction of preformed metabolites (DAP + MMA + MDA) ranged from 0.41 to 1.00. The mole ratio of total metabolites to malathion parent ranged from 0.70 to 333. In summary, by demonstrating that most of the DAPs, MDA, and TCP are formed on food items, these studies indicate that the metabolite concentration measured in urine may be due to direct exposure to these relatively non-toxic compounds, rather than to the parent OP insecticide. The substantial vari Direct formation o f OP metabolites infood ability in the metabolite-to-parent ratio reduces the value of excreted metabolites as markers of OP insecticide exposure. DAPs are products of OP hydrolysis. The metabolism of OP insecticides in plants and humans is similar. The DAPs detected in human urine may therefore have been ingested with food or formed in the body following absorption of OP insec ticides (e.g., Zhang et al. 2008). Compared with their parent 'To be precise, when formed in the environment, the DAPs are not "tnetabolites" formed by enzymatic transformations, but rather are degradation prod ucts formed by hydrolysis or photolysis. However, we use the term `"DAP metabolites" in the paper for brevity. m \a h i b v Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00003 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 534 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 Rapid metabolism o f OP insecticides The epidemiologic studies of prenatal OP exposure typi cally include either one or two urinary measurements of OP metabolites that are intended to represent the exposure of the mother during pregnancy. However, many OP insecticides are metabolized relatively rapidly. Most OP insecticides are typically excreted within 24--48 h (World Health Organiza tion [WHO] 1996). Some human exposure data suggest even faster rates of metabolism for particular OP insecticides. For example, Garfitt et al. (2002) reported that a single oral dose of diazinon has a urinary elimination half-life of 2 h. In a similar study, Bouchard et al. (2003) estimated a 4-h half-life for malathion. Given the rapid elimination of OP insecticides, any spot measurement will reflect only recent exposure. If the relevant exposure period of interest is an average over pregnancy, a single measurement may be inadequate. There is no biological basis to specify a particular exposure period during pregnancy as especially relevant for neonatal or childhood outcomes examined in this review. However, if the exposure period of interest is a short time window during pregnancy, then a spot measurement taken outside that window may not he etiologically relevant. DAP metabolites are non-specific variability (Griffith et al. 2011, Sexton and Ryan 2012, Bradman et al. 2013, Attfseld et al. 2014). For example, withinchild variability in one study was 2-11 times greater than that observed across the study population (Attfield et al. 2014). Griffith et al. (2011) found similar results for children living in an agricultural community in central Washington State. Sexton and Ryan (2012) measured the intraclass correlation coefficient for urinary DAP among elementary school children in Minneapolis, and observed "only modest correlations" in siblings from the same household. Because the associations estimated in the epidemiologic studies are based on one, or at most two, DAP measurements, a higher level of intra- than inter-individual variability can lead to considerable exposure misclassification. Attfield et al. (2014) illustrated this problem by assigning subjects with multiple available OP metabolite measures to four exposure categories based on the mean values of 1-4 randomly selected samples. If the metric under study is reliable, the grand means of the four resulting exposure categories are expected to increase monotonic ally. In this study, however, the resulting grand means w;ere monotonic only 14-15% of the time for MDA and 19-32% for TCPy, when the exposure assessment was based on only one sample per subject. When two samples were used, the resulting grand means for MDA and TCPy were monotonic 31-32% and 34-41% of the time, respectively. Multiple OP insecticides are metabolized into each of the six DAPs (Duggan et al. 2003, Sudakin and Stone 2011). Some OP insecticides (e.g., malathion and disulfoton) are converted to as many as three different DAPs, whereas others (e.g., dichlorvos and tetrachlorvinphos) metabolize to only a single DAP. Moreover, acephate and methamidophos do not metabolize to DAPs at all (Solecki 2002). There are substantial differences in toxicity across the OP insecticides. The U.S. Environmental Protection Agency (EPA) estimated chronic exposure benchmark doses using 10% brain AChE inhibition threshold (BMD,0) for all regis tered OP insecticides. AChE inhibition is the widely recog nized mechanism of action for OP toxicity (Mileson et al. 1998). The BMD,0 values in the EPA assessment, based on rat laboratory studies, ranged from 0.04 milligrams per kilogram body weight per day (mg/kg/day) for dicrotophos to 313.9 mg/kg/day for malathion (IJSEPA 2002), a nearly 8000-fold difference. Even among the most widely used OP insecticides, the toxicity varies over orders of magnitude (see next section). Such large differences in toxicity across OP insecticides, combined wdth the lack of specificity for DAPs, significantly limit the ability of DAP urinary levels to provide an informative measure of toxic exposure. Intra-individual variability in urinary DAP levels Potential impact of exposure misclassification It is important to consider the potential impact of misclassification of OP insecticide exposure on the results of epidemiologic studies, it is often said that if exposure misclassification is non differential (i.e., independent of health status), bias is expected to produce an attenuated measure of association (Cantor et al. 1992). However, it is plausible that short-term variability in dietary patterns and other influences on OP insecticide expo sure differ by health status. For example, diet is associated with birth outcomes and neurodevelopment (Abu-Saad and Fraser 2010, Millichap and Yee 2012, Smithers et al. 2013), and changes in diet are commonly triggered by health status. Consequently, if diseased individuals altered their dietary habits more frequently than non-diseased individuals, then the degree of exposure misclassification would differ by health status, leading to an unknown degree or direction of bias. Even if exposure misclassification is 11011-differential by health out come, it does not necessarily result in a predictable direction of bias. Additional conditions, such as independence of classi fication errors, must be met for non-differential misclassification of a binary exposure to result in bias toward the null, and even then the tendency applies only to the expectation of the estimated association, not to the value of the estimate from any single study (Jurek et al. 2008, Jurek et al. 2005). Moreover, for Studies with repeated measures of urinary DAP concentra tions offer useful information on intra-individual variability. Bradman et al. (2013) found that spot DAP measurements in exposures with multiple levels, non-differential misclassifica tion results in bias of unpredictable direction and magnitude (Sorahan and Gilthorpe 1994, Wacholder et al. 1995). children changed up to tw'o orders of magnitude over a week or even within a day. In 24-h urine samples, the DAP levels Dose-response differed by as much as an order of magnitude for samples col OP insecticides or their active metabolites inhibit the enzyme lected three days apart. AChE, which breaks down the neurotransmitter acetylcho /A number of researchers have reported that within- line. Neurotoxicity results from excessive accumulation of child variability in DAP levels is higher than between-child acetylcholine in cholinergic synapses. Thus, inhibition of 8 IQ H I S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00004 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 535 nervous system AChE is generally regarded as the primary risk assessment based on earlier data (U.S. EPA 2005b). All toxic mode of action for OP insecticides (Mileson et ah 1998, of the BMD10 val ues are for exposure to rat pups on postnatal U.S. EPA 2000). Accordingly, the U.S. EPA regulates OP day 11 and are the lowest BMD, 0 estimates observed in pups, insecticide safety by setting exposure levels to he sufficiently adults, and pregnant dams. The rat pups were exposed directly low that excessive AChE inhibition will not occur (U.S. EPA on postnatal day 11 and prenatally through exposure from the 2000). It is possible that developmental neurotoxicity may dam. The rats in these studies w;ere generally w'ell nourished, result from mechanisms other than AChE inhibition (Yang which may lead to uncertainty in applying the results to poorly et ah 2011). However, the U.S. EPA requires developmental nourished human populations. neurotoxicity studies for OP insecticides and has found that It is useful to estimate exposures associated with DAP lev AChE inhibition is protective of developmental neurotoxicity els measured in the epidemiologic studies so that AChE inhi effects, it is acknowledged that developmental neurotoxicity bition associated with those DAP levels can be estimated. This studies in animals may not be sensitive enough to detect all can be roughly accomplished by back-calculating an exposure developmental neurotoxicity-related effects; research in this based on the DAP level and urine volume, acknowledging the area continues. uncertainties in the calculation. Curl et al. (2003) provides a In humans and other mammals, AChE exists in both the ner simple equation to estimate the dosage associated with a uri vous system (brain, spinal cord, and peripheral plexuses and nary DAP measurement: nerves) and the red blood cells (RBCs) with varying amounts in plasma in some species. Another type of cholinesterase, hutyrylcholinesterase (BChE), is found in plasma and other Dosaee = --D--A--P----X--V----X--M---W--BW tissues (Li et ah 2005). Inhibition of blood cholinesterase, either in RBCs or where [DAP] is the total molar DAP concentration, V is plasma, is generally regarded as a marker of exposure, but the daily urine volume, MW is the molecular weight, and BW not necessarily a toxic effect (ILS. EPA 2000). Nevertheless, is the body weight. We assume a normal urine volume of because data on AChE activity in peripheral nervous system 20 mL/kg/day (Gonzales and Bauer 1999). The urinary levels tissues may he lacking in animal studies and data on periph are corrected for DAPs formed on food items by assuming eral nervous system tissues and/or brain is usually lacking in that 38% of the urinary DAP levels are from exposure to the humans, the EPA regards AChE inhibition in blood as a surro pesticide, based on data from Zhang et al. (2008). Use of gate for peripheral nervous system AChE inhibition in animals the above equation to estimate the dosage of OP insecticide and brain AChE inhibition in humans. Given that the relevant associated with DAP measurements in the epidemiologic- target for toxicity is nervous system AChE and extensive data studies has important limitations. The DAPs originate from are available on inhibition of brain AChE in rats and other different OP compounds, but to apply the dose-response mod non-human species, the focus of the analysis described below els, we need to assume that all DAPs originate from exposure is on brain AChE inhibition. to one OP insecticide. In addition, data on DAPs formed on It is useful to examine these data relative to the OP bio food items are not available for all OP insecticides and com marker levels in non-occupational settings to determine the modities. The DAP measurements in the epidemiologic stud potential for brain AChE inhibition at the exposure levels ies are typically spot samples, yet the equation estimates full- found in the epidemiologic studies. As part of its risk assess day exposures. Despite these limitations, the models provide ments for registration review, the U.S. EPA has developed a useful approximation to assess AChE inhibition for dosages AChE dose-response models for brain AChE for OP insecti corresponding to the urinary metabolite levels found in the cides used in the United States. The dose-response models are epidemiologic studies. based on the benchmark dose for 10% inhibition (BMD10) of Among participants in the 2000-2004 National Health and brain AChE in animal studies. The BM D,() represents the dose Nutrition Examination Survey (NIIANES), the geometric that, on average across the animals, causes 10% AChE inhibi mean of urinary DAP concentrations was 68 nmol/L, and the tion and is considered by the U.S. EPA to be a '`response level corresponding 75th percentile was 186 nmol/L (Bouchard close to the background cholinesterase" (U.S. EPA 2002). The et al. 2010). The NIIANES data represent a sample of the dose-response models are based on an exponential decline of general non-institutionalized U.S. population. For the Cen AChE activity with dose. ter for the Health Assessment of Mothers and Children of We reviewed the U.S. Department of Agriculture (USDA) Salinas (CHAMACOS) epidemiologic study, Bradman et al. Pesticide Data Program database to identify the OP insecti (2005) reported median and 90th percentile levels of 103 and cides most commonly detected in food. The latest data are 732 nmol/L for the first prenatal sample, 107 and 422 nmol/L from 2012 (USDA 2014). Four OP compounds-- dimethoate, for the second prenatal sample, and 227 and 1349 nmol/L omethoate, malathion, and chlorpyrilbs--account for nearly for the postpartum sample, respectively. The most recent 80% of the 663 detections. In the U.S. EPA risk assessments (2007-2008) NHANES data on urinary DAPs show' that the (U.S. EPA 2005, 2009a, 2011), the lowest BM D,0 values 50th, 75th, 95th, and 99th percentiles across 2564 samples were 1.4 mg/kg (4.1 nmol/kg) for chlorpyrilbs, 1.5 mg/kg were 48, 155, 587, and 1406 nmol/L, respectively, assuming (6.6 nmol/kg) for dimethoate, and 23.6 mg/kg (71.5 nmol/kg) half the limit of detection for non-detects (CDC 2014). The for malathion. For omethoate, we used the BMD10 of 0.14 mg/ 75th percentile of 155 nmol/L is somewhat lower than the 75th kg (0.68 nmol/kg) based on a cholinesterase study conducted percentile reported by Bouchard et al. (2010) for the 2000- after the last U.S. EPA risk assessment (Reiss 2012). U.S. EPA 2004 NHANES data. The 98th percentile in the 2007-2008 used a slightly higher value of 0.18 mg/kg in its last dimethoate NHANES data set is about 2000 nmol/L, which corresponds U IQHTS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00005 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 536 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 to exposures o f less than about 8-14 pg/kg/day, depending on course of pregnancy) exposures. The dose-response analysis the molecular weight of the OP compound. was done with acute exposures, because the DAP urinary Based on the dose-response models assuming that all expo measurements correspond to short-term exposures. Bradman sures are from a single OP insecticide, at 2000 nmol/L, the et al. (2013) showed that there is a large variability in DAP estimated brain AChE inhibition was 0.002% for malathion measurements for individuals over one week. Thus, urinary and 0.001% for chlorpyrifos. While malathion has a higher DAP levels may not be appropriate for chronic dose-response BMD]0, the chlorpyrifos data were fit to a different dose- assessment, unless steady state has been reached between dose response model that has a low-dose shoulder, limiting inhibi rate and biotransformation/elimination, resulting in a plateau tion at low doses. At higher doses, the models diverge, and steady-state level of metabolite(s). malathion is estimated to cause less inhibition than chlorpy While most agree that OP toxicity is mediated through rifos. The estimated brain AChE inhibition at 2000 nmol/L is AChE inhibition, some have argued that toxicity from OP 0.03% for dimethoate and 0.2% for omethoate. insecticides occurs at doses lower than those required to cause These low levels of brain AChE inhibition are highly AChE inhibition (e.g., Slotkin and Seidler 2007). However, unlikely to be clinically detectable, particularly considering for many studies that have reached this conclusion, subsequent the variety of factors that may affect AChE activity. For exam observations indicate that the AChE activity measurements ple, solanaceous glycoalkaloids found in potatoes cause AChE from the inhibition tests were conducted long after the initial inhibition (Krasowski et at. 1997); so does huperzine, another exposure. This allowed time for the AChE activity to recover, natural product derived from club moss, which is used in the missing the point of maximum inhibition, and resulting in an treatment of dementia (Ozarowski et al. 2013). The inhibition underestimate of AChE inhibition (Eaton et al. 2008). Many of AChE activity associated with huperzine is hypothesized of these studies were done with chlorpyrifos. It was also noted to result in improvements in long-term memory (Ozarowski that the doses used in several of these studies ranged from 1 to et al. 2013). Lefkowitz et al. (2007) evaluated baseline RBC 5 mg/kg chlorpyri fos admini stered subcutaneously to rat pups, AChE activity for 46 workers over an average of 20 years of or prenatally (Eaton et al. 2008). For 20 mL/kg/day of urine employment. The mean coefficient of variance for RBC AChE volume (Gonzales and Bauer 1999), assuming that 38%' of was 3.9%. Ferioli and Maroni (2011) report inter-individual DAPs are from exposure to chlorpyrifos (Zhang et al. 2008), variations in RBC AChE of 10-18% and intra-individual the estimated DAP levels associated with 1-5 mg/kg of chlo variations of 3-7%. This baseline variance for individuals is rpyrifos dose are approximately 375 000-1 900 000 nmol/L, higher than the estimated /AChE inhibition at upper percentiles levels that are well above those measured in the epidemiologic of the doses reported in the epidemiologic studies. Moreover, studies discussed in this section. these data are for RBC AChE, which adds uncertainty, because Some recent studies have also pointed to OP-mediated RBC AChE activity serves as a surrogate measure of brain enzyme inhibition in the endoc annabinoid system, which is AChE function. important in nervous system development, and suggested that The estimates from the U.S. EPA dose-response models these effects occurred at doses that do not cause AChE inhibi are for the mean response in rats and do not account for intra- tion (e.g., Carr et al. 2013). However, at this time, the meaning individual variability or the potential for increased sensiLi vity of these effects is unclear. in humans. There are limited data to directly compare animal Overall, there are no toxicological data to suggest that del and human sensitivity to OP compounds, although the mecha eterious effects could occur as a result of the low-level OP nism is considered similar. There was no RBC cholinesterase insecticide exposures experienced by subjects in the epide inhibition in a single-dose study of humans at malathion doses miologic studies. as high as 15 mg/kg (Giles and Dickson 2000). This is higher than the 7.6 mg/kg estimate (95th percentile lower limit of the BMDi0) for malathion-induced RBC cholinesterase inhi Confounding and bias bition based on an acute dose to rats (U.S. EPA 2009b), Tim- OP exposure in non-occupationally exposed populations is chalk et al. (2002) developed physiologically based pharma likely driven by diet and residential pesticide use (Krieger et al. cokinetic and pharmacodynamic (PBPK/PD) models for rats 2012). Both diet/nutritional status and residential pesticide use and humans for chlorpyrifos and found similar differences in may, in turn, be associated with other factors that affect health, chlorpyrifos sensitivity between rats and humans for RBC thereby potentially resulting in confounding bias. In addition, AChE inhibition. Even with a 100-fold uncertainty factor, the selection bias can occur if study completion rates (in cohort estimated AChE inhibition levels are low. At a DAP urinary studies) or participation rates (especially in case-control and level o f 2000 nmol/L and assuming a 100-fold uncertainty cross-sectional studies) vary according to OP exposure, and factor, the AChE inhibition is estimated to be 0.2% for mala health outcome. thion, 2.1%: for chlorpyrifos, 2.4% for dimethoate, and 21%: For example, maternal body mass index (BMI), smoking, for omethoate. While the omethoate estimate is above 10%, it and nutrition can influence urinary DAP levels (see Figure 1 was derived conservatively by assuming that all DAPs come developed from CDC 2014 data; other data from CDC 2014 from omethoate consumption, in addition to the 100-fold show that smokers have lower urinary DAP levels), as well as safety factor. birth outcomes (Marshall and Spong 2012, Mason et al. 2012, The available dose-response models are for acute exposures. Andres and Day 2000). These factors, along with childhood There is no biological basis to determine whether the possible nutrition and BMI, which is inversely associated with urinary effects found in the epidemiologic studies are caused by acute DAP levels (see Figure 2 developed from CDC 2014 data), (during a small window of pregnancy) or chronic (over the can also influence neurodevelopmental outcomes in children SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00006 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 537 also influence OP in ternal dose and DAP levels, thereby lead ing to confounding. Selection bias may also occur if these factors influence study participation or completion rates. The full scope of determinants of OP and DAP exposure levels and of birth and neurodevelopmental outcomes is not known, and potentially vast. Even if statistical models adjust for several behavioral factors, residual confounding may occur due to omission of important variables or imprecise classification of those that are included. Review of epidemiologic studies Birth outcomes i l l ! ! ! 20 30 40 50 60 70 Body Mass index Figure 1. Urinary DAP (nmoi/L) versus BMI for adults (> 1 8 years of age) in the 2007-2008 NHANES dataset. Note: Graph truncated at 1000 nmol/L DAP concentration, which is about the 97th percentile. Red line produced with a LOESS smoothing function in the R programming language (R Core Team, 2014). Eleven studies in seven birth cohorts have investigated associations between OP metabolites and birth outcomes (Barr et al. 2010, Berkowitz et al. 2004, Eskenazi et al. 2004, Harley et al. 2011, Perera ei al. 2003, Rauch et al. 2012, Wang et al. 2012, Whyatt et al. 2005, Whyatt et al. 2004, Wfickerham et al. 2012, Wolff et al. 2007) (Table 1). All studies evaluated OP or OP metabolite levels in maternal prenatal or perinatal biospecimens and/or umbilical cord blood, in relation to stan dard measures of size and gestational age at birth ascertained (Bliddal et al. 2014, Sandjaja et al. 2013, Neggers et al. 2003, from medical records, a computerized hospital perinatal data Burkhalter and Hillman 2011, Anjos et al. 2013). Another issue is that PON1, an enzyme that detoxifies some base, and/or hospital delivery fogs. Table 2 summarizes the analyses in the studies evaluating birth outcomes. OP insecticides and that could therefore play an important role in mediating their toxic effects, may influence health outcomes independently of its effects on bioavailable OP levels---for Columbia Centerfor Children's Environment and Health example, through an antioxidant mechanism (Macharia et al. The first study, based at the Columbia Center for Children's 2014). PQN1 activity has a myriad of endogenous and envi ronmental influences, including diet and lifestyle, as well as Environment and Health (CCCEH), followed healthy, non smoking, pregnant Dominican and African American women genetic determinants (Aviram and Vaya 2013, Schrader and who had lived for at least one year in northern Manhattan or Rimbach 2011). Thus, PON1 activity level could also con found apparent associations between DAP levels and health the South Bronx, New York, from < 2 0 weeks o f gestation through delivery (Table 1) (Perera et al. 2003, Whyatt et al. outcomes through a DAP-independent pathway. In summary, numerous environmental and endogenous factors can affect birth outcomes and neurodevelopment, and many of these factors-- including PON1 activity levels-- may 2005, Whyatt et al. 2004). Study enrollment took place between 1998 and 2006. Chlorpyrifos, diazinon, and other pesticides were measured in maternal plasma samples collected within two days postpartum and in umbilical cord blood collected at delivery. Over the study period, average OP insecticide metabolite concentrations progressively declined. The mean concentration of chlorpyrifos was 7.1 pg/g in maternal plasma and 7.6 pg/g in cord plasma in an earlier study (Perera et al. 2003), but fell to 3.9 pg/g (standard deviation [SD] = 4.8) in maternal plasma and 3.7 pg/g (SD = 5.7) in cord plasma with extended enrollment (Whyatt et al. 2005). In the latter study, the mean concentration of diazinon was 1.3 pg/g (SD = 1.8) in maternal plasma and 1.2 pg/g (SD = 1.4) in cord plasma. OP insecticide levels were also measured in personal ambient air samples collected by mothers, who were asked to wear a backpack air sampling pump during the day and to place the monitor near the bed at night for two consecutive days dur ing the third trimester of pregnancy. Mean air concentrations were 14.3 ng/m3 (SD = 30.7) for chlorpyrifos and 99.5 ng/m3 (SD = 449.8) for diazinon (Whyatt et al. 2005). In multivariate adjusted linear regression models based on Figure 2. Urinary DAP (nmol/L) versus BMI for children (< 19 years of age) in the 2007-2008 NHANES dataset. Note: Graph truncated at 1000 nmol/L DAP concentration, which is about the 97th percentile. Red line produced with a LOESS smoothing function in the R programming language (R Core Team. 2014). 263 mother-newborn pairs and with natural logarithm (ln)transformed outcomes, maternal perinatal plasma chlorpyri fos levels (pg/g) were significantly inversely associated with birth weight (beta = --0.04 In-g, P = 0.01) and birth length (beta = --0.03 In-cm, P = 0.04), but not head circumference 8 IQ H I S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00007 538 R. Reiss et al Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 1. Design of epidemiologic studies of organophosphorus insecticide biomarkers. Reference(s) Perera et al. (2003), Whyatt et al. (2004, 2005), Ran h et al. (2006, 2011, 2012), Lovasi et al. (2011), Horton tal. (2012) Study name Columbia Center for Children's Environmental Health Location Study design Study subjects Study dates N ew York City, N ew York, United States Prospective birth. cohort Pregnant Dominican and African-American women aged 18-35 years, residing for > 1 year before pregnancy in Washington Heights, Central Harlem, or South Bronx, New York, registered at one o f two obstetrics and gynecology clinics by the 20th week of pregnancy, and without diabetes, hypertension, known HIV, or current use of tobacco or illicit drugs; 725 mother-child pairs enrolled, with 70% participation as o f 2002; 83% retention rate at 3-vear follow-up, 82% retention rate at 7-year follow-up. Rauh et ai. (2012) further restricted to children with no/very low prenatal environmental tobacco smoke exposure and low prenatal airborne polycylic aromatic hydrocarbon exposure 1 9 9 8-2006 up to age 7 -1 1 years Exposure assessment Exposure concentrations* Outcome assessment Chlorpyrifos and diazinon (and other pesticides) measured in maternal plasma collected within 2 days postpartum and umbilical cord plasma collected at delivery; regressionderived maternal values were used in analyses when cord levels were unavailable Chlorpyrifos, diazinon, malathion, and methyl parathion (and other pesticides) measured with personal air monitor worn during daytime hours for 2 consecutive days and placed near bed at night during third trimester of pt egnancv Chlorpyrifos and diazinon levels combined by converting diazinon levels to chlorpyrifos levels based on ratio o f relative potency factors (6-1 for chlorpyrifos to diazinon) calculated by the U.S. Environmental Protection Agency (2002) Questionnaire administered at home during third trimester and annually thereafter Maternal perinatal plasma (pg/g): Chlorpyrifos (Perera et al. 2003): mean - 7.1,98% detectable Chlorpyrifos (Whyatt et ai. 2005): mean SD = 3.9 4.8 Diazinon (Whyatt et ai. 2005): mean SD --1.3 1.8 Umbilical cord plasma (pg/g): Chlorpyrifos (Perera et ai. 2003): mean - 7.6, 94% detectable Chlorpyrifos (Whyatt et al. 2005): mean SD 3.7 5.7 Diazinon (Whyatt et al. 2005): mean SD = 1 ,2 * 1.4 Maternal prenatal personal air (ng/ m 3) (Whyatt et al. 2005): Chlorpyrifos: mean S D -1 4 .3 :;: 30.7 Diazinon: mean :t: SD = 99.5 449.8 Spearman correlation for maternal and cord plasma levels of chlorpyrifos = 0.6, P < 0.001 (Perera el al. 2003) and 0.79, F < 0.001 ("Whyatt et al. 2005); diazinon ^ 0.69, F < 0 .0 0 1 (Whyatt et al. 2005) Spearman correlation for maternal plasma and maternal air levels of chlorpyrifos ^ 0.21, F < 0.001; diazinon --0.004, P-value NR (Whyatt et ai. 2005) Birth outcom es information and pregnancy and delivery characteristics obtained from mothers' and infants' medical records following delivery Bayley Scales of infant Development, 2nd Edition (Mental Development Index and Psychom otor D evelopm ent index), administered at 12, 24, and 36 months Child Behavior Checklist for ages 1.5-5 years, including syndrome scale scores, internalizing and externalizing scores, and Diagnostic and Statistical Manual o f Mental Disorders. 4th Edition-oriented scales, com pleted at 36 months Child Behavior Checklist for ages 6 -1 8 years and Wechsler Intelligence Scale for Children, 4th Edition (Verbal Comprehension index. Perceptual Reasoning index, Working Memory Index, and Processing Speed Index, combined for FullScale intelligence Quotient), completed at 7 years Brain morphology assessed using high-resolution, T lweighted magnetic resonance im aging at 5 .9 -1 1 .2 years Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00008 m m i ; m N ;! d o i Low-ievei o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Berkowitz et al. (2004), Engel et al. (2007), W olff et al. (2007), Engel et al. 2011 Mount Sinai Children's Environmental Cohort Study N ew York City, New York, United States Prospective birth cohort Consecutive primiparous pregnant women entering prenatal care with a singleton pregnancy at 26 w eeks o f gestation, without serious chronic disease or serious pregnancy complication, not consuming > 2 alcohol beverages per day or using illegal drugs, in a m ulti-ethnic, urban population; excluding infants with congenital malformation or severe prematurity (< 1,500 g or < 32 weeks of gestation); 479 (33%) participants of 1,450 eligible women; 404 included in analysis after excluding 75 (16%) of 479 due to m edical com plications, prematurity, congenital defect, lack, o f prenatal specim ens, change o f h ospital or residence, or refusal (low er follow -u p for younger and less-educated mothers) 1998-2001 up to age 6 -9 years TCPy, MX3A, and six DAP metabolites (DMPs: dimethylphosphate, dimethyithiophosphate, and direethylditbiophosphale; DEPs: diethy[phosphate, diethylthiophospbate, and diethyldithiophosphate) measured in maternal urine collected during third trimester P O N l,9r P 0 N 1 55, PON1_909> P O N 1_j62, and P O N l_ m genotypes, PON1 activity (measured against phenylacetate) and butyrylcholinesterase activity (measured against butyrylthiocholine) assessed in third-trimester maternal blood and urn biSical cord blood Prenatal questionnaire administered during third trimester Median (IQR for TCPy; range for others) in maternal prenatal urine (Berkowitz et al. 2004, Wolff et al. 2007): TCPy: 7.6 (1 .6-32.6) jxg/L, 11.5 (1 .8 -3 5 .4 ) pg/g creatinine MDA: limit o f detection ( < 0.3 jig/L) ( < 0.3-15.8); 20.5% detectable DAPs: 75.9 (0-1 ,9 8 7 ) nrnol/L, 88.6 (0 -2 ,1 0 6 ) nm ol/g creatinine, DMPs: 42.2 (0-4,903) nmol/L, 55.4 (0 -2 ,0 7 1 ) nm ol/g creatinine DEPs: 18.8 (0 -4 2 9 ) nmol/L, 22.1 (0 -1 ,0 0 2 ) nm ol/g creatinine Birth outcom es information and delivery characteristics obtained from hospital computerized perinatal database Brazelton Neonatal Behavioral Assessment Scale administered before hospital discharge (age ^ 5 days) and scored according to seven clusters developed by Lester et al. Bayley Scales of infant Development, 2nd Edition (Mental Development and Psychomotor Development Indices), administered at --T2 and 24 months W echsler Preschool and Primary S cale o f Intelligence, 3rd Edition (if age < 7 years), or W echsler Intelligence Scale for Children, 4th Edition (if age 7 -9 years) administered between ages 6 and 9 years (Continued) Tier 3/4 Low-level OP insectide exposures 539 ED 002061 00046431-00009 m m i ; m 540 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 1. (Continued ) Reference(s) Eskenazi et al. (2004, 2007, 2010), Young et al. (2005), Marks et al. (2010), Bouchard et ai. (2011), Harley et al. (2011), Quiros-Aicala et al. (2011) Study name Center for the Health Assessment of Mothers and Children of Salinas (CHAMACOS) Location Study design Study subjects Study dates Salinas Valley, California, United States Prospective birth cohort Pregnant women aged 22 18 years, entering prenatal care at < 20 weeks o f gestation, English- or Spanish-speaking, eligible for Medi-Cal, planning to deliver at the county hospital, in a primarily Latino, lowincome, farmworker population; 601 (53.2'%) participants of 1,130 eligible; -5 3 0 [variably reported] followed through delivery of a live-born infant 1 9 9 9-2000 up to age 7 years m m i ; m Exposure assessment Exposure concentrations* S is DAP metabolites measured in maternal and child spot urines collected at interviews: dimethylphosphate, dimethyithiophosphate, and directhyiditbiophosphate (combined as BM Ps), dietbyiphosphate, diethylthiophospbate, and diethyldithiophosphate (combined as DEPs) Seven pesticide-specific metabolites measured in maternal spot urines collected at interviews: MDA, PNP, TCPy; also 2-diethylamino-4-hydroxyb-reethylpyrircidine, 2-isopropyl4-rnethyi-6- hydroxypyrimi dine, 3-cb lorO"4-methy3"7"hydroxycoumann, and. 5-chioro-l-isopropyl-3hydroxytriazole (detectable in < 11%) Cholinesterase and butyrylcholinesterase measured in maternal hiood/plasina taken at second interview during pregnancy and. before delivery and in umbilical cord blood/ plasma Maternal, cord, and child blood specimens genotvped for P O N 1,0. and maternal post-delivery, umbilical cord, and 24-month child blood samples tested for PON1 enzym e quantity (arylesterase activity against phenylacetate) and enzyme activity (paraoxonase activity against paraoxon) Median (range) in maternal urine (Eskenazi et ai. 2004; Young et ai. 2005): DAPs (nmol/L): 136 (10-6,854) prenatal, 222 (7-2 1 ,8 6 7 ) postdelivery DM Ps (nmol/L): 101 (5-6,587) prenatal, 160 (5-2 1 ,8 5 7 ) postdelivery DEPs (nmol/L): 22 (2 -6 8 0 ) prenatal. 27 (2-666) post-delivery M DA (pg/L): 0.2 (0 .2 -2 8 ,9 ) prenatal TCPy (jxg/L); 3.3 (0.2-56.1) prenatal PNP (l-ig/L): 0.5 (0 .1 -3 4 .7 ) prenatal Geometric mean (95% C l) in child urine (nmol/L S(Eskenazi et al. 2007; Marks et al. 2010): DAPs: 45.5 (39.6-52.3) at 6 months, 59.5 (51.7 -6 8 .5 ) at 12 months, 70.9 (61.4 -8 1 .9 ) at 24 months, 77.5 (65.4 -9 1 .9 ) at 3.5 years, 92.6 (87.6-1 0 9 .0 ) at 5 years DMPs: 23.8 (20.4 -2 7 .8 ) at 6 months, 32.9 (27.8 38.9) at 12 months, 48.6 (41.8 -5 6 .6 ) at 24 months, 62.5 (52.2 -7 4 .7 ) at 3.5 years, 72.4 (6 1 ,0 -8 6 ,0 ) at 5 years DEPs: 10.6 (8.9-11.9) at 6 months, 1 5 .2 (1 3 .5 -1 7 .2 ) at 12 months, 10.5 (8.8-1 2 .6 ) at 24 months, 7.0 (5 .8 -8 .3 ) at 3.5 years, 7.2 (6.0-8.7) at 5 years Interviews at --13--I4 weeks of gestation, - 2 6 - 2 7 weeks of gestation, - 1 week after delivery, and when children were - 6 and 12 months and 2, 3.5, 5, and 7 years old Outcome assessment Birth outcomes information obtained from hospital delivery logs and medical records Brazelton Neonatal Behavioral Assessment Scale administered at or before 62 days, with seven clusters developed by Lester et a!. Bayley Scales of Infant Development, 2nd Edition (Mental Development and Psychomotor Development Indices), administered at 6, 12, and 24 months Autonomic nervous system reactivity protocol administered using social, physical, emotional, and cognitive (at 3.5 and 5 years) challenges, with measurement of heart rate, respiratory sinus arrhythmia, and pre-ejection period, at 6 months and 1, 3.5, and 5 years Child Behavior Checklist for ages 1.5-5 years (attention problems syndrome, ADHD, and pervasive developmental disorder scales) completed by mothers at 2, 3.5, and 5 years NEPSY'^-K visual attention subtest administered at 3.5 years Conners' Kiddie Continuous Performance Test (for reaction time, accuracy, and impulse control) administered and. Hillside Behavior Rating Scale (for motor activity and distractibility) completed by psychometricians at 5 year's Wechsler intelligence Scale for Children, 4th edition, administered at 7 years Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00010 S* d o i Low-ievei o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Lizard! et al. 2008 Children Pesticide Survey Yuma County, Arizona, United States Mew Jersey, United States Crosssectional Prospective birth cohort Schoolchildren (mean a g e 7 years) from an agricultural community near the U.S.-M exico border, previously participating in a pesticide screening study and selected for further study based on the, absence (N --23) or presence (/V ~ 28) of urinary organophosphate pesticide metabolites in the original urine specimen 2002 Convenience sample of 150 women with a singleton pregnancy and non-anomalous fetus scheduled for an elective cesarean birth, at term fze 37 weeks of gestation) with hemoglobin level 8 mg/dL, excluded if evidence for labor or rupture of membranes at time of operative delivery or if using medications that could potentially interfere with metabolism or environmental chemicals; 2 maternal blood and 2 umbilical cord blood samples excluded due to processing errors 2 0 0 3 -2 0 0 4 to birth Six DAP metabolites measured in first-void urine sample collected from each child on the, day o f the cognitive assessment: dimethylphosp'nate, dim eth y ith io p h o sph a te, dim eth y ld ith io p h o sph a te, diethyiphosphate, diethylthiophosphate, and diethyldithiophosphate; lower limit of detection --25 ug/L Structured Interview at home with parents Chlorpyrifos and other pesticides measured in maternal blood obtained prior to placement o f intravenous and bladder catheters before cesarean section or in extra maternal blood specim ens available from preoperative testing A lso measured in umbilical cord blood obtained within 15 minutes of delivery Seif-administered questionnaire distributed to pregnant women Mean :T SD dimethylphosphate (pg/L) in child urine = 65.5 78 (95% Cl " 43, 88), 100% detectable Mean (95% Cl) urinary DAPs (,ug/L) in originally exposed group --110 (83-139); mean in originally unexposed group 49 (36-63), P < 0.01, after excluding one high outlier from each group (In original screening urine sample, 25 children had detectable and 23 had undetectable DAPs) Chlorpyrifos in maternal serum (ng/g): 98.6% detectable, mean SD = 0,09 0.87, median (range. IQR) = 0.0007 (0.0007-10.09, 0.0007-0.0007) Chlorpyrifos in umbilical cord serum (ng/g): 62.8% detectable. mean SD = 0,55 0.73, median (range, IQR) - 0.0007 (0.0007-1.84, 0.0007-1.32) Pearson's correlation for chlorpyrifos in maternal and cord serum - 0.12 Cognitive performance assessed using Wechsler Intelligence Scale for Chiklren-Third Edition Short Form, Childreris Memory Scale, W isconsin Card Sorting Test, and Trail Making Test A and B, completed by child at school or, if not possible, at home during a second visit Behavioral performance assessed using Child Behavior C hecklist/4-18 (completed at home by parents) and Teacher Report Form (completed at school by teachers) Birth outcomes information and pregnancy characteristics obtained from medical records prior to hospital discharge (Continued) Tier 3/4 Low-level OP insectide exposures 541 ED 002061 00046431-00011 m m i ; m 542 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 1. (Continued)______________________ Reference^) Study name Location Bouchard et al. (2010) National Health and Nutrition Examination Survey (NHANES) United States Wang etai. (2011) Shanghai, China m m i ; m Study design Study subjects Study dates Crosssectional Prospective birth cohort Population-based health survey data from non-institutionalized children aged 8-15 years seiected using multi-stage probability sampling, with oversampling of certain subgroups, to be representative of the general U.8. population; ADHD assessed in 3,998 participants, urinary DAP metabolite data available for 1,481 (37%) based on 50% sampling rate for ages 6 -! 1 years and 33% for ages 12-15 years in 2000-2002, and 33% sampling rate in 2003-2004; further excluded children who received NICU or premature nursery care and those with birth weight < 2,500 g, urinary creatinine < 20 mg/dL, outlier urinary DAP concentrations, or missing covariate data Pregnant women aged 18-45 years attending one o f two major obstetric hospitals, with no gestational or pre-existing diabetes, hypertension, HIV/ AIDS, or use o f illegal drugs in the preceding '/ear, with singleton infants free, o f severe neonatal illness; 187 (96.9%) participants o f 193 eligible 2000-2004 2 0 0 6-2007 to birth Exposure assessment Six DAP metabolites measured in spot urine samples collected during physical examinations at mobile study centers; dimethyiphosphate, dimethyithiophosphate, dimethyldithiophosphate (combined as DMPs), diethylphosphate, diethylthiophosphate, and diethyldithiophosphate (combined as DEPs) interview conducted in respondent's home Exposure concentrations* Geometric mean (range and IQR) in child urine (nmol/L): DAPs; 68.3 (6.0-10,195, 24.4-186.0) DMPs: 41.3 {4.5-10,068, 10.1-1 30.7) DEPs: i 1.0 (0.8-5905, 2.1-35,0) Dimethyiphosphate: 10.7 (2.8-1324, 2.8-39.0) Dimethyithiophosphate: 13.7(0.9-9929, 1.9-58.8} Dimethyldithiophosphate: 1.7(0.3-7006, 0.4-7.3) Diethylphosphate: 4.7 (0.4-5902, 0.9-28.1} Diethylthiophosphate: 2,0 (0.3-650, 0.4-7.6) Diethyldithiophosphate: 0.5 (0.2-36, 0.3-0.3) Outcome assessment ADH D and ADH D subtypes in previous year assessed based on Diagnostic Interview Schedule for Children IV based on slightly modified criteria from Diagnostic and Statistical Manual o fMental Disorders, Fourth Edition, based on a telephone interview with the mother or another caretaker 2-3 weeks after physical examination A DH D defined in study as meeting diagnostic criteria of ADHD or regularly taking ADH D medication during the previous year Eive DAP metabolites measured in maternal spot urines collected at the onset o f labor; dimethyiphosphate, dim eth y ith io p h o sph a te, diethylphosphate, diethylthiophosphate, and diethyldithiophosphate interview conducted during pregnancy Geometric mean (range and IQR) in maternal prenatal urine (Jg/Ly Dimethyiphosphate: 17.19 (undetectable-269.15; 7.02-53.70) Dimethyithiophosphate: 8.01 (undetectable-109.65; 3.53-20.06) Diethylphosphate: 6.03 (undetectable109,65; 3.55-11.17) Diethylthiophosphate: 6.31 (undetectable-i31.83; 3.36-11.98) Diethyldithiophosphate: NR because 5.34% detectable (undetectable-5.1; undetectable-undetectable) Birth outcomes information and pregnancy and delivery characteristics obtained from mothers' and infants' medical records Geometric mean (range and IQR) in maternal prenatal urine (jug/g creatinine): Dimethyiphosphate: 25.75 (0.81-588.84; 12.25-72.86) Dimethyithiophosphate: 11.99 (0.56-123.02; 5.45-28.40) Diethylphosphate: 9.03 (0.58-89.13; 5.13-16.54) Diethylthiophosphate: 9.45 (0.4793.33; 4.53-18.30) Diethyldithiophosphate: NR because 5.34% detectable (0.31-9.33; 0.94-2.43) Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00012 .m i s* D o i Lovv-ievei o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Guodong et al. ( 2012) Shanghai, China Rauch et al. (2012), Yokon et al. (2013) Health Outcomes and Measures of the Environment (HOME) Study Cincinnati, Ohio, United States m m i ; m Crosssectional Children aged 2 3 -2 5 months attending routine physical check ups at departments of child and adolescent health care at two community hospitals, with no intrauterine distress, pathological jaundice, intrauterine infection, intracranial infection, congenital disease, or current cold or fever, and able to complete the neurodeveloprcental assessment; 301 (97.1%) participants of 310 eligible 2008 Prospective birth. cohort Pregnant women aged > 1 8 years attending one of seven prenatal clinics, living in a home built before 1978, < 19 weeks of gestation, HIV-negative, living within five surrounding counties in a socioeconomically diverse area, and not receiving thyroid or seizure medications, or chemotherapy or radiation treatments; 468 (37.1%) participants of 1,263 eligible; 389 followed through delivery of a live-born singleton infant (9 followed through delivery of twins, 3 followed through stillbirth) 2 0 0 3 -2 0 0 6 up to age -5 weeks Five DAP metabolites measured in spot urine, collected on the. day of study assessment: dirnethylphosphate, dimethyithiophosphate (combined as DMPsS, diethylphosphate, dieihyltbiophosphate, and diethyidithiophosphate (combined as DBFs) interview conducted with mothers Six DAP metabolites measured in maternal spot urines collected at '-16 and - 2 6 weeks o f gestation (averaged for analysis) and within 24 hours o f delivery: dirnethylphosphate, dimethyithiophosphate, and dimethyidithiophosphate (combined as DM.Ps), diethylphosphate. dieihylt biophosphate, and diethyldithiophospate (combined as DBFs) Umbilical cord blood genotyped for PONl,n mdPONl_m Interviews during pregnancy ( - 2 0 weeks o f gestation) and after delivery (-5 weeks) Geometric mean (range and 1QR) in child urine. (pg/L): Dirnethylphosphate: 2.52 ( < 2,0 [limit o f detection]-186.99; < 2 .0 -3 .4 1 ) Dimethyithiophosphate: 1.56 (< 1.0-80.8!; < 1.0-1.63) Diethylphosphate: 1.78 ( < 1 .0 32.19; < 1.0-2.89) Diethylthiophosphate: 3.18 (< 1.0-55.40; < 1 .0 -7 .2 6 ) Diethyidithiophosphate: NR because 2.7% detectable ( < 1.0-3.80; < 1.0-< 1.0) Geometric mean (range and iQR) in child urine (jig/g creatinine): Dirnethylphosphate: 11.27 (1 .53-729.27; 4.33-24.02) Dimethyithiophosphate: 6.99 (1.08-481.50; 3.09-13.12) Diethylphosphate: 7.96 (1.14170.96; 3.84-16.36) Di ethylth iop hosp hate : 14.19 (1.10-980.58; 5.30-37.15) Diethyidithiophosphate: 4.55 (1.08-73.14; 2.49-7.70) Median (IQR) in maternal prenatal urine (nmol/L) (Rauch et al. 2012): DAPs: 81.3 (41.7-220.0) DMPs: 56.9 (26-185) DBFs: 17.7 (8-37) 10-fold increase in DAPs 15th percentile (29.5 nmol/L) to 85th percentile (318.0 nmol/L) Gesell Developmental Schedules for 0- to 3-year-old children administered to evaluate neurological and intellectual development using four main categories of functioning: motor behavior, adaptive behavior, language behavior, and personal and social behavior Birth weight abstracted from medical records; gestational age calculated from mother's selfreported dale of last menstrua! period or based on ultrasound (N --7) or Ballard examination at delivery (N = 3) N1CU Network. N'eurobebaviora! Scale administered in home at '-5 weeks (mean --34 days, range --17-47); 13 dimensions: habituation (omitted due to small number completed), attention, arousal, self-regulation, need for special handling by examiner, quality of movement, excitability, lethargy, non-optimal reflexes, asymmetrical reflexes, hypertomcity, bypotorncity, and stress/abstinence (Continued) Tier 3/4 Low-level OP insectide exposures 543 ED 002061 00046431-00013 .& 544 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table h (Continued) Reference(s) Wickerham et al. (2012) Study name - Location Zhejiang, China On [bote and Bouchard (2013) Canadian Health Measures Survey, cycle 1 Canada Study design Study subjects Prospective birth cohort Consecutive pregnant women with a healthy, uncomplicated, singleton pregnancy recruited from a single hospital at 36 weeks of gestation, excluding those with chronic diseases, complicated pregnancies, or hereditary or metabolic diseases; 116 participants with infants born at:> 37 weeks of gestation and umbilical cord blood pesticide levels (-'-99.6% participation rate); excluded 3 with missing data and 1 highly influential outlier Crosssectional Population-based health survey data from children selected using multi-stage probability sampling, with oversampling of certain, subgroups, to be representative of the general Canadian population; 1,081 children aged 6--11 years among 5,600 participants aged 6 -7 9 years; 1,030 (95%) with most urinary pesticide metabolite levels and behavioral assessment, 779 (72%) after exclusion of those with missing covariate data Study dates 2009 to birth 2007-2009 Exposure assessment Exposure concentrations* Eight organophosphate pesticides (and other pesticides) measured in umbilical cord serum at delivery: chlorpyrifos, diazinon, fonoibs, malathion, parathion-ethyl, parathion-methyl, profenofos, and terbufos % detectable, median, and 90th percentile in umbilical cord serum at delivery (ng/mL): Chlorpyrifos: 23.3%, < 0.05 (limit: of detection), 0.17 Diazinon: 14.7%, < 0.05 (limit of detection), 0.27 Fonofos: 16.4%, < 0.05 (limit of detection), 0.30 Malathion: 25,9%, < 0 .5 0 (limit of detection), 3.13 Six DAP metabolites measured Parathion-ethvl: 2.6%, < 0 .0 5 (limit of detection), < 0 .0 5 Parathion-methyl: 28.5%, < 0 .0 5 (limit o f detection), 1.43 Profenofos: 25.0%, < 0.50 (limit of detection), 0.68 Terbufos: 31.0%, < 0 .0 5 (limit of detection), 0.27 Median (1QR) in child urine in spot urine samples collected during physical examinations at mobile examination centers within 2 weeks of survey questionnaire completion: dirnethy[phosphate, dim etb y lth io p b o sph a te, dimethyidithiophosphate (combined as DMPs}, diethyiphosphate. dieihyltbiophosphate, and (nmol/L): DAPs: 99.2 (34.3-273.3) DMPs: 62.0(18.7-192.8} DEPs: 25.0 (10.5-51.3) Dimethylphosphate: 34.6 (10.8-91.9) Dimethylthiophosphate: 17,6 (< 4 .2 [limit of detection]-75.4) Dimethyidithiophosphate: < 1.9 (limit of detection) (< 1.9-5.6} diethyldithiophosphate (combined as DBFs) Diethyiphosphate: 19.6 (8,5-42,0) Diethyithiophospbate: < 3 .5 (limit of Household survey conducted in respondent's home detection) (< 3.5-6.9) Diethyldithiophosphate: < 1.6 (limit o f detection) ( < 1 .6 - < 1.6) Outcome assessment: Birth, outcomes information and pregnancy characteristics obtained from patient charts Beh avioral problems assessed using parent version of the Strengths and Difficulties Questionnaire, including scales for emotional symptoms, conduct problems, hyperactivity/ inattention, peer problems, prosocial behavior, and total difficulties (sum of all dimension scales except prosocial behavior), categorized into high vs. low/ borderline using authorrecommended cutoff scores Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00014 m m i ; m d o i Low-ievei o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Fortenberry et al. Early Lite M exico City, Prospective Mother-child pairs irom three 1994-1997, TCPy measured in maternal Geometric mean (95% Cl and IQR) Conners' Parent Rating (2014) Exposures M exico birth. sequentially-enrolled cohorts of 1997-2000, or third-trimester morning void urine TCPy in maternal prenatal urine (ng/ Scales-R evised (A DH D Index, in M exico to cohort pregnant women enrolled during 2001-2005- specimens inL) - 1.76 (1.55-2.02, 0.91-3.57) Global Restlessness/Lnpulsivity Environmental Toxicants (ELEMENT) pregnancy or at delivery from a general hospital or afli hated clinics in a low- to moderate-income setting, excluding women with plans to leave the area within five years, daily alcohol consumption, 2007-2011 (ages 6-11 years) intraclass correlation among 21 subjects with measured levels in all three trimesters of pregnancy --0.41 without correction for specific gravity, 0.29 with correction Index, and Hyperactivity/ Impulsivity, Inattention, and Combined ADHD scales based on guidelines from D iagnostic and Statistical Manual of Mental Disorders, 4th Edition) addiction to illegal drugs, completed by parents continuous use of prescription drugs, diagnosis of multiple pregnancy, pre-eclampsia, renal or heart disease, gestational diabetes, high-risk pregnancy, or seizures requiring medical treatment, or history of infertility, diabetes, Behavior Assessment System for Children-Parental Rating Scales completed by parents Conners' Continuous Performance Test completed by children or psychosis; 187 (23%) of 827 participants re-invited from second and third cohorts with child psychometric assessment and third- trimester maternal urine, including Zhang et 1014) Shenyang, China Prospective 21 with urine in all three trimesters Healthy pregnant women recruited 2 0 1 1 -2 0 1 2 to age Five DAP metabolites measured Geometric mean (range and IQR.) in Neonatal Behavioral birth from a single hospital, living in 3 days in maternal prenatal urine (tinting maternal prenatal urine (jig/L): Neurological Assessment cohort Shenyang for > 3 years, without not specified): dimethylphosphate, Dimethylphosphate: 18.03 2 [limit performed at age 3 days, with hypertension, diabetes, thyroid dime! hylthiophosphate (combined of detection j-334.02, 7.83-39.43) five scales: behavior, passive hypofunction, heart disease, or as DMPs), diethylphosphate, Dimethylthiophosphate: 8.53 ( < 1 [limit tone, active tone, primary other chronic diseases before diethylthiophosphate, and of detection]-137.95, 3.4-15.67) reflexes, and general assessment, pregnancy, without serious diethyldithiophosphate (combined as Diethylphosphate: 7.14 ( < 1 [limit of combined as summary score pregnancy complications, and DEPs) detection]-!67.06, 3.54-17.17) without family or medical history of mental retardation. In-person interviews with mothers Diethylthiophosphate: 5.64 ( < 1 [limit of detection]--133.00, 2.34-13.55) phenylketonuria, or Pompe's Diethyldithiophosphate: < 1 (limit of syndrome- for sel f or spouse; detection) ( < 1-6.61, < l - < 1} also excluding infants with disorders associated with adverse neurodevelopment; 249 (81.1%) participants of 307 eligible Geometric mean (range and IQR) in maternal prenatal urine (jig/g creatinine): Dimethylphosphate: 24.02 (0.19- 453.04, 9.72-60.1) Dimethylthiophosphate: 11.29 (0.10-305.92,4.07-29.49) Diethylphosphate: 9.49 (0.36-125.09, 4.29-20.45) Diethylthiophosphate: 7.58 (0.32- 102,21,2,93-19.95) m Diethyldithiophosphate; 0.78 (0.03-15.41, 0.46-1.34) m *Vaiues shown are reported in the earliest available publication from each study cohort, except for the Columbia cohort, where values changed substantially over time and are shown from multiple publications. i AD HD attention deficity/hyperactivity disorder confidence interval, D AP dialkyl phosphate, DEP diethyl phosphate, D M P dimethyl phosphate, IQR interquartile range, MDA rnalathion dicarboxyhc acid, N1CU neonatal intensive care unit, NR not reported, PN P 4-ni.trophenoi, PON1 paraoxonase 1, SD standard deviation, TCPy 3,5,6-irichloro-2-pyridinol. m Tier 3/4 Low-level OP insectide exposures 545 ED 002061 00046431-00015 v Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 546 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 (beta = --0.005 In-cm, P = 0.82) (Table 2) (Perera et al. confounding is difficult to predict, and the magnitude is prob 2003). The inverse association with birth weight was statisti ably limited by die adjustment for several major influences on cally significant among African Americans but not Domini birth outcomes. Fourth, because numerous hypotheses were cans, whereas the reverse race/ethnicity pattern was observed tested, at least some statistically significant associations are for birth length. In subsequent analyses based on 314 mother- expected due to chance. Neither this study nor any other study newborn pairs, cord plasma chlorpyrifos levels (ln-pg/g) of birth outcomes described in this review made statistical cor were also significantly inversely associated with birth weight rections for multiple comparisons. Although such corrections (beta = --42.6 g, 95% confidence interval [Cl] = --81.8, are not standard in traditional epidemiology, authors who do - 3.8) and birth length (beta = - 0.24 cm, 95% Cl = - 0.47, not correct for multiple comparisons should report the number --0.01), but not head circumference (beta = --0.01 cm, 95%: and nature of all associations tested, how certain associations Cl = - 0 .1 3 , 0.11) (Whyatt et al. 2004). Slightly stronger were selected for reporting, and the probable effect of such inverse associations w'ere observed with cord plasma chlo selection on the results (Rothman et al. 2012). As evidence rpyrifos and diazinon levels combined, but diazinon itself was in other areas of research, particularly genetic epidemiol not significantly associated with any of the three outcomes. ogy, numerous exploratory analyses almost inevitably lead Maternal prenatal personal air levels of chlorpyrifos, diazinon, to false-positive results and recent methodological literature and both OPs combined also were not significantly associated includes several practical ways of dealing with this problem with any of the three birth outcomes. The inverse associations (Wacholder et al. 2004, Stromberg et al. 2008, Weitkunat et al. betw'een cord plasma chlorpyrifos and birth weight and length 2010, Wakefield 2007). w'ere restricted to newborns born before January1 ,2001, when Finally, the completeness of follow-up from enrollment the IJ.S. EPA instituted regulations to phase out residential use through delivery was not reported, but if follow-up varied by of these insecticides; exposure levels were substantially lower uncontrolled factors, such as diet, that might be associated and no associations were detected in newborns born in 2001 or with maternal OP exposure and birth outcomes, then an unpre later. Similar findings were reported in a slightly larger group dictable degree of selection bias could have occurred. Cohort of mother-newborn pairs with cord plasma measures of chlo participation rates also were not reported (but were stated as rpyrifos and diazinon (Whyatt et al. 2005). Specifically, birth 70% in an earlier publication [Whyatt et al. 2002]), and could weight was 67.3 g lower (95% Cl = --116.6, --17.8), and birth have been a source of a moderate degree of selection bias if length w as-0 .4 3 cm shorter (95% C I= --0.73, --0.14) for participation were related to OP exposure and birth outcomes. each one-unit (ln-pg/g) increase in cord plasma chlorpyrifos (Participation bias is usually considered not to be a major con among 237 newborns born before January 1,2001, but no such cern in prospective cohort studies, because outcomes occur association was detected among 77 newborns born after that after cohort entry, but with relatively short-term follow-up, it date (beta for birth weight = 30.7 g, 95% Cl 108.6, 169.9; is conceivable that participation could be associated with risk beta for birth length = 0.07 cm, 95% Cl = --0.65, 0.79). of adverse birth outcomes.) Substantial strengths of the CCCEH study include the use In summary, results in the CCCEH cohort suggest an of objective, individually measured metabolites to characterize inverse association of maternal perinatal plasma levels of exposure to OP insecticides (a strength of all studies discussed chlorpyrifos, but not diazinon, with birth weight and birth in this review), the availability of information on numerous length, but not with head circumference, in an urban, low'- potential confounders, and the prospecti ve design, with mater income, minority population. The observation of associations nal interviews and personal air monitoring conducted during only among newborns born before 2001, when exposure lev the third trimester of pregnancy, prior to the health outcomes els were higher, suggests a possible exposure threshold below of interest. which chlorpyrifos is not associated with birth outcomes. The Some methodological limitations of the CCCEH study detection of certain associations only in African Americans should be noted. First, a single maternal blood sample was but not Dominicans, or vice versa, indicates that the observed collected from each subject at or shortly after deli very. Normal associations may he attributable to excessive stratification. fetal growth is approximately linear between 18 and 37 weeks The lack of associations with maternal prenatal personal air of gestation, after which it plateaus; thus, maternal plasma OP levels of chlorpyrifos and diazinon raises the question of levels at delivery may not reflect levels in past weeks or months, whether route of exposure is an effect modifier of associations and may be etiologically irrelevant to fetal growth. Although between OP exposure and birth outcomes. Furthermore, the maternal air samples were obtained in the third trimester of different results for chlorpyrifos and diazinon suggest that OP pregnancy, it is unknown whether a single sample collected insecticides should be analyzed separately, not combined, with over two days is representative of exposure at other time points. respect to birth outcomes, although this approach raises the Second, the number of participants was modest, especially problem of multiple comparisons. Overall, the inconsistent after stratification by race/ethnicity or birth date, resulting in results by outcome, racial/ethnic group, and exposure metric several statistically unstable estimates of association (i.e., wide render the findings difficult to interpret, and do not provide confidence bounds). Third, given the many potential influences compelling evidence to support an adverse effect of chlorpyri- on chlorpyrifos and diazinon levels in peripheral blood and air, fos or diazinon on fetal growth. as well as on birth outcomes, uncontrolled confounding by diet and other factors may partially explain some of the observed results. However, without detailed knowledge of established Mount Sinai Children's Environmental Cohort Study predictors of chlorpyrifos and diazinon exposure and of birth Another birth cohort based in New York City, the Mount Sinai outcomes in this study population, the direction of potential Children's Environmental Cohort Study (CECS), enrolled SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00016 DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. Results of epidemiologic studies of organophosphorus insecticide biomarkers and birth outcomes. Reference Perera et al. (2003) Outcome Birth weight (g), natural log scale Exposure Maternai perinatal plasma chlorpyrifos (pg/g) Number of subjects/ events 263 total 116 African American 146 Dominican Estimate of association (95% Cl) Beta - - 0.04. P - 0.01 Beta - - 0.05, P - 0.04 Beta = 0.02. P = 0.26 Perera et al. (2003) Birth length (cm), natural log scale Perera et al. (2003) Whyatt et al. (2004) Head circumference (cm), natural log scale Birth weight (g) Cord plasma chlorpyrifos (pg/g, natural log scale) 263 total 116 African American 146 Dominican 263 total 116 African American 146 Dominican 314 total 237 born before 1 January 2001 77 bom before 1 January 2001 Beta = - 0.02, P = 0.04 Beta - --0.01, P = 0.15 Beta - - 0.02, P - 0.002 Beta - - 0.005, P - 0.28 Beta - -0 .0 0 3 , P = 0.70 Beta - -0 .0 0 5 , P - 0.31 Beta - -4 2 .6 (- 8 1 .8 , - 3 .8 ) Birth before 1 January 2001 beta - - 67.3 ( - 116.6, - 17.8) Birth after 1 January 2001 beta- 3 0 .7 ( - 108.6, 169.9) Group 2 vs. i beta - 39.2 ( - 107.3, 185.7) Group 3 vs. 1 beta ----50.9 ( - 188.2, 86.3) Group 4 vs. 1 beta - - 150.1 (-2 8 7 .7 , - 12.5) Whyatt et al. (2004) Whyatt et al. (2004) Whyatt et al. (2004) Cord plasma chlorpyrifos + diazinon (pg/g, natural log scale) Cord plasma diazinon (pg/g, natural log scale) Maternal prenatal persona] air chlorpyrifos, diazinon, or chlorpyrifos + diazinon (ng/in3, natural log scale) Beta - -4 9 .1 (- 9 1 .3 , - 6 .9 ) Birth before 1 January 2001 beta - - 72.5 ( - 125.0, - 20.0) Birth after 1 January 2001 beta- 0 .6 ( - 144.7, 145.9) Group 2 vs. 1 beta -- 78.5 (-225.5,68.5) Group 3 vs. 1 beta ----33.1 ( - 173.7, 107.4) Group 4 vs. 1 beta ----186.3 (-327.2, -45.4) Beta - - 44.2 (- 1 1 9 .5 ,3 1 .0) Chlorpyrifos beta ----17.7 ( - 64.2, 28.9) Diazinon beta --13.8 (--23.2, 50.8) Chlorpyrifos f diazinon beta --- 5.1 (--50.7, 40.4) Adjustment factors Maternal body mass index, parity, cotinine, infant sex, gestational age, and maternal prenatal airborne polycyclic aromatic hydrocarbon levels Gestational age, maternal prepregnancy weight, maternal weight gain during pregnancy, newborn sex, parity, race/ ethnicity, environmental tobacco smoke in home, and season of delivery No change after additional adjustment for cord plasma 2-isopropoxyphenol levels (results NR) Comments No significant interactions were observed between chlorpyrifos and polycyclic aromatic hydrocarbons, although numbers were limited (results NR) Except for plasma diazinon levels, insecticide levels decreased substantially for infants bom after 1 January 2001 (after phase out of residential use by U.S. Environmental Protection Agency regulatory action), despite no significant change in self-reported pesticide use habits Group i: < limit of detection (31% of cord plasma chlorpyrifos levels, 48%) of cord plasma diazinon levels); groups 2, 3, and 4: tertiles of detectable levels Group 1: < limit of detection (31 % of cord plasma chlorpyrifos levels, 48 % of cord plasma diazinon levels); groups 2, 3, and 4: tertiles of detectable levels _ Associations with maternal persona] air samples remained non-significant after stratification by birth before or after 1 January 2001 (results NR) {Continued) m Tier 3/4 Low-lcwl OP USECtdE E x p o SUTES 547 ED 002061 00046431-00017 s* Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 548 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 2. (Continued) Reference Why alt et al. (2004) Whyatt et al. (2004) Whyatt et al. (2004) Whyatt et ai. (2004) Whyatt et al. (2004) m Outcome Birth length (cm) Exposure Cord plasma chlorpyrifos (pg/g, natural log scale) Cord plasma chlorpyrifos + diazinon (pg/g, natural log scale) " Cord plasma diazinon (pg/g, natural log scale) " Maternal prenatal personal air chlorpyrifos, diazinon, or chlorpyrifos + diazinon (ng/m3, natural log scale) Head circumference (cm) Cord plasma chlorpyrifos, diazinon, or chlorpyrifos 4- diazinon (pg/g, natural log scale) Number of subjects/ events 309 total 237 born before 1 January 2001 77 bom after 1 January 2001 298 total Estimate of association (95% Cl) Beta = - 0.24 ( - 0.47, - 0.01) Birth before 1 January 2001 beta = --0.43 (-0 /7 3 , -0 .1 4 ) Birth after 1 January 2001 beta = 0.07 (-- 0,65, 0.79) Group 2 vs. 1 beta = 0.17 (-0 .7 0 , 1.0) Group 3 vs, 1 beta ----0.21 ( - 1.0. 0.61) Group 4 vs. 1 b e ta :::: 0.75 ( - 1.6,0.06) Beta = - 0.27 ( - 0.52, - 0.02) Birth before 1 January 2001 beta - - 0.46 ( - 0.77, - 0.14) Birth after 1 January 2001 beta - - 0.07 ( - 0.82, 0.67) Group 2 vs. 1 beta = --0.06 (-0 .9 3 , 0.81) Group 3 vs. 1 beta ----0.005 ( - 0.84, 0.82) Group 4 vs. 1 beta -== - 0.80 ( - 1.6,0.02) Beta ----0.32 (--0.75, 0.11) Adjustment factors Comments Group 1: < limit of detection (31 % of cord plasma chlorpyrifos levels, 48%) of cord plasma diazinon levels); groups 2, 3, and 4: tertiles of detectable levels Group 1: < limit of detection (31 % of cord plasma chlorpyrifos levels, 48% of cord plasma diazinon levels); groups 2, 3, and 4: tertiles of detectable levels Chlorpyrifos beta ----0.02 ( - 0.28, 0.25) Diazinon beta -==0.07 ( - 0.14, 0.28) Chlorpyrifos F diazinon beta ----0.01 (-0 .2 7 , 0.25) Chlorpyrifos beta = --0.01 (-0 .1 3 , 0.11) Diazinon beta ----0.07 (--0.30, 0.16) Chlorpyrifos t diazinon beta = - 0.02 (-0 .1 5 , 0.1! ) Associations with maternal personal air samples remained non-significant after stratification by birth before or after 1 January 2001 (results NR) Gestational age, maternal pre pregnancy weight, maternal weight gain during pregnancy, newborn sex, parity, race/ ethnicity, environmental tobacco smoke in home, season of delivery, and cesarean section delivery No change after additional adjustment for cord plasma 2-isopropoxyphenol levels (results NR) m i ; m : CritRevTcocoL 2015; 45(7): 531-641 ED 002061 00046431-00018 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. d o i Low-ievei o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Whyatt et ai. (2004) Whyatt et al. (2005) Whyatt et ai. (2005) Whyatt et al. (2005) Whyatt et al. (2005) Whyatt et ai. (2005) Whyatt et al. (2005) m m i ; m Birth weigl4 (g) Birth length (cm) Maternal prenatal personal air chlorpyrifos, diazinon, or chlorpyrifos + diazinon (ng/m3, natural log scale) Cord plasma chlorpyrifos (pg/g, natural log scale) 237 born before 1 January 2001 77 born after 1 January 2001 Cord plasma chlorpyrifos + diazinon (pg/g, natural log scale) 237 born before 1 January 2001 77 bom after 1 January 2001 Maternal prenatal personal air chlorpyrifos, diazinon, or chlorpyrifos + diazinon (ng/m3, natural i1U(gy tCaJ.^C)1 Cord plasma chlorpyrifos (pg/g, natural log scale) Cord plasma chlorpyrifos + diazinon (pg/g, natural log scale) Maternal prenatal personal air chlorpyrifos, diazinon, or chlorpyrifos + diazinon (ng/m3, natural log scale) 237 bom before 1 January 2001 77 born after 1 January 2001 237 born before 1 January 2001 77 born after 1 January 2001 237 born before 1 January 2001 77 bom after 1 January 2001 Chlorpyrifos beta - --0.04 ( - 0,18, 0.10) Diazinon beta = - 0.03 (--- 0.14, 0.09) Chlorpyrifos + diazinon beta= --0.03 (-0 .1 7 , 0.11) Beta ----67.3 (--116,6,--17.8) Beta = 30.7 ( - 108.6, 169.9) Beta = - 72.5 ( - 125.0, - 20.0) Beta = 0.6 ( - 144.7, 145.9) Group 4 vs, 1 b eta= 215.1 (-3 8 4 .7 , --45.5) "No association" (results NR) Associations with maternal personal air samples remained non-significant after stratification by birth before or after 1 January 2001 (results NR) Gestational age, maternal pre-pregnancy weight, maternal weight gain during pregnancy, newborn gender, parity, ethnicity, environmental tobacco smoke in home, and season of delivery No change after additional adjustment for cord plasma 2-isopropoxyphenol levels (results NR) 34% of newborns bom before 1 January 2001 and 1,5% of those born after had cord plasma levels of chlorpyrifos + diazinon in the top fertile of detectable levels (P < 0.001) Group 1: < limit of detection; groups 2, 3, and 4: tertiies of detectable levels Beta = - 0.43 ( - 0.73, - 0.14) Beta = 0.07 ( - 0.65, 0.79) Beta = - 0.46 ( - 0.77, - 0.14) Beta = --0.07 ( - 0.82. 0.67) "No association" (results NR) (Continued) Low-level OP insectide exposures 549 ED 002061 00046431-00019 550 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued) Reference Why alt et al. (2005) Outcome Head circumference (cm) Exposure Cord plasma or maternal prenatal personal air chiorpyrifos, diazinon, or chlorpyrifos 4 diazinon (natural log scale) Number of subjects/ events Estimate of association (95% Cl) "No association" (results NR) Berkowitz et al. (2004) Birth weight (g) Maternal prenatal urinary TCPy (pg/L) 216 < 11.0 pg/L (limit of detection) 171 > 11 .0 pg/L Mean SD - 3,284 4 4 ! Mean SD = 3,296 434 P > 0.05 Berkowitz et al. (2004) Berkowitz et al. (2004) Berkowitz et al. (2004) m m i ; " Birth length (cm) Maternal prenatal urinary TCPy (|-tg/L) by maternal PON! activity (tertile) Maternal prenatal urinary TCPy (pg/L) Maternal prenatal urinary TCPy (pg/L) by maternal PON 1 activity (tertile) 76 < 11.0 pg/L, low PON 1 62 < 11.0 pg/L, medium PON1 71 < 11.0 pg/L, high PON 1 47 > 11.0 pg/L, low PON! 57 > 11.0 pg/L, medium PON 1 55 > 11.0 pg/L, high PON 1 216 < 11.0 pg/L (limit of detection) 171 > 1 1 .0 pg/L 75 < 11.0 pg/L, low PON! 62 < 11,0 pg/L, medium PON 1 71 < 11.0 pg/L, high PON 1 46 > 11,0 pg/L, low PON 1 57 > 11.0 pg/L, medium PON1 55 > 1 1 .0 pg/L, high PON 1 Mean SD = 3,237 456 Mean SD = 3,255 436 Mean SD --.3,337 444 P-trend > 0.05 Mean SD = 3,278 395 Mean SD - 3,327 406 Mean SD = 3,270 409 P-trend > 0.05 Mean SD = 50.4 2.4 Mean SD - 50.8 2.4 P > 0.05 Mean SD = 50.3 2.3 Mean S D -5 0 .1 2.2 Mean SD --50.3 2.3 P-trend > 0.05 Mean SD - 50.9 2.3 Mean SD = 51.0 2.3 Mean SD - 50.8 2.4 P-trend > 0.05 Adjustment factors Gestational age, maternal pre-pregnancy weight, maternal weight gain during pregnancy, newborn gender, parity, ethnicity, environmental tobacco smoke in home, season of delivery, and delivery by cesarean section Race/ethnicity, infant sex, and gestational age No difference after additional adjustment for active and passive cigarette smoking, pre-pregnancy body mass index, maternal weight gain, blood lead levels, and cesarean section delivery Comments Results for TCPy not reported by infant PON1 activity or maternal or infant PON1 genotype Tier 3/4 CritRevTcocoL 2015; 45(7): 531-641 ED 002061 00046431-00020 ^4 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. d o i Low-level o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Berkowitz et al. (2004) Berkowitz et ai. (2004) Berkowitz et al. (2004) Wolff et al. (2007) W olffetal. (2007) Wolff et ai. (2007) Head circumference (cm) " Gestational age (weeks) Birth weight (g) Maternal prenatal urinary TCPy (pg/L) Maternal prenatal urinary TCPy (pg/L) by maternal PON 1 activity (fertile) Maternal prenatal urinary TCPy (pg/L) Maternal prenatal urinary DAPs (mrtol/L or nmol/g creatinine, logl0 scale) Maternal prenatal urinary DMPs (nraoi/L or nmol/g creatinine, io g ,p scale) Maternal prenatal urinary DEPs (nmol/L or nmol/g creatinine, logl0 scale) 216 <11.0 pg/L (limit of detection) 171 > 1 1 .0 ug/L 76 < 11.0 pg/L. low PONI 62 < 11.0 pg/L, medium PONI 70 < .0 pg/L, high PONI 47 > 11,0 pg/L, low PONI 57 > .0 pg/L, medium PONI 55 > 1 1 .0 pg/L, high PONI 216 <11.0 pg/L (limit of detection) 171 > 11 .0 pg/L 318 327 318 Mean SD = 33.8 1.7 Mean SD = 33.8 1.7 P > 0.05 Mean SD = 33.6 1.8 Mean SD = 33.7 1.7 Mean SD = 34.1 1.7 u No difference after additional adjustment for birth weight or birth length, stratification by raee/ethnicity, or excluding preterm births Test for interaction among TCPy level, PON1 activity, and head circumference was not statistically significant (P > 0.05) P-trend > 0.05 Mean S D -3 3 .3 1.5 Mean SD = 34.0 1.5 Mean S D = 34.1 1 .6 P-trend = 0.014 Mean SD - 39.3 1.8 Mean SD - 39.3 1 .7 P > 0.05 Beta SE = -- 25 34, P - 0.47 (not creatinine-adjusted) Beta SE = - 27 34, P = 0.43 (creatinine-adjusted) Beta SE = 1.9 29. P ------0.95 (not creatinine-adjusted) Beta SE = --2.7 29, P --0.92(creatinine-adjusted) Beta SE = - 5 2 3 2 ,' P = 0.099 (not creatinineadjusted) Beta SE = -- 56 32, P ----- 0.082 (creatinine-adjusted) Race/ethnicity and infant sex No difference after additional adjustment for active and passive cigarette smoking, pre-pregnancy body mass index, maternal weight gain, blood lead levels, and cesarean section delivery Race/ethnicity, maternal PON 1 activity, infant sex, and gestational age Value of 0.5 was added to urinary DAP before log -transformation; 25 samples with < 20 mg/dL creatinine were excluded (Continued) m m i ; m ^4 Low-level OP insectide exposures 5 5 1 ED 002061 00046431-00021 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 552 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 2. (Continued) Reference W olffetal. (2007) Outcome " W olffetal. (2007) " Wolff et ai. (2007) Wolff et al. (2007) Birth length (cm) Wolff et al. (2007) m i m Exposure Maternal prenatal urinary DEPs a- vs. < median by maternal PONI activity (fertile) Maternal prenatal urinary DEPs a vs, < median by maternal PONI ,02 genotype Maternal prenatal urinary MDA > 0.3 vs. < 0.3 pg/L (limit of detection) Maternal prenatal urinary DAPs (ninol/L or nmol/g creatinine, iog10 scale) Maternal prenatal urinary DMPs (nmol/L or nmol/g creatinine, logl0 scale) Number of subjects/ events 60 DEPs < median. low PONI 53 DEPs < median, medium PON 1 45 DEPs < median, high PONI 53 DEPs a median, low PON I 51 DEPs > median. medium PONI 56 DEPs > median, high PONI 39 DEPs < median, PO N li9, RR 84 DEPs < median. PONI j 02 RQ 33 DEPs < median. PONI m QQ 55 DEPs a median, PONI ,92 RR 66 DEPs a- median, PO N I,,, RQ 42 DEPs a median, PO N li92 QQ 318 327 Estimate of association (95% Cl) Mean SE --3305 53 Mean SE = 3348 57 Adjustment factors Infant race, sex, gestational age, and creatinine level Mean SE --3396 64 Mean SE = 3233 56, P = 0.323 within PONI Mean SE - 3282 57. P - 0.392 within PONI Mean SE = 3279 54. P 0.138 within PON 1 P for interaction term in model --0,878 P ~ 0.042 for high PONI/low DEP vs. low PONI/high DEP Mean SE = 3346 69 Mean SE = 3278 4 6 Mean SE - 3453 60 Mean SE = 3254 63. P = 0.291 within PONI 192 Mean SE = 328550, P - 0.907 within PONI 192 Mean SE - 3232 52, P -----0.005 within PONI 192 P for interaction term in model = 0.0755 P - 0.020 for PON 1192 QQ/low DEP vs. PON1192 RR/high DEP Beta SE = 3 9 52, P = 0.46 (not creatinine-adjusted) Beta SE --59 A 53, P --0.27 (creatinine-adjusted) Beta SE - --0,13 1 9 , P --0.49 (not creatinineadjusted) Beta S E = - 0.13 1 9 , P = 0.49 (creatinine-adjusted) Beta SI 0,121:0.16. P ----- 0.44 (not creatinineadjusted) Beta SE - --0,12 0.16, P --0.44 (creatinine-adjusted) Race/ethnicity, maternal PON 1 activity, infant sex, and gestational age Comments Lowest PONI tettile slowg highest PONI tettile = fast PONI ,92 RR - slow, P O N I, QQ ~ fast Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00022 ED 002061 00046431-00023 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA W olffetal (2007) Wolff et al. (2007) Wolff et al. (2007) m m i ; m Tat ^4 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Maternal prenatal urinary DMPs Sr vs. < median by maternal PON 1 activity (fertile) Maternal prenatal urinary DMPs sr vs. < median by maternal PON1 ;9, genotype Maternal prenatal urinary DEPs (nmol/L or nmol/g creatinine, logJp scale) 60 DMPs < median, low PON 1 53 DMPs < median, medium PON1 45 DMPs < median, high PON 1 53 DMPs Sr median, low PON1 51 DMPs a median, medium PON1 56 DMPs > median, high PON1 39 DMPs < median, PO N 1lg, RR 84 DMPs < median, P O M I9, RQ 33 DMPs < median, p o m }92 QQ 55 DMPs Sr median, PO N 1192 RR 66 DMPs Sr median, PON110? RQ 42 DMPs a median, P O M !0, QQ 318 Mean SE --51.1 0.3 Mean SE - 50.3 0.3 Infant race, sex, gestational age, and creatinine level Mean SE = 50.4 0.3 Mean SE --50.2 0.3, P ----- 0.032 within PON1 Mean SE = 50.7 0.3. P = 0.258 within PON1 Mean SE = 50.8 0.3, P = 0.418 within PON1 P for interaction term in model --0.036 P = 0.549 for high PONl/low DMP vs. low PONl/high DMP Mean SE = 50.6 0.4 Mean SE - 50.4 0.3 Mean SE - 51.0 0.3 Mean SE = 49.9 0.3, P --0.164 within PON 1192 Mean SE --50.7 0.3, P ----- 0.158 within PON1192 Mean SE = 50.8 0.3. P = 0.695 within PON1192 P for interaction term in model --0,230 P ------0.019 for PON1192 QQ/low DMP vs. PON1192 RR/high DMP Beta S E = -~ 0.02 0.18, P ------0.93 (not creatinineadjusted) Beta SE = 0.017 0.18, P --0.924 (creatinine-adjusted) Race/ethnicity, maternal PON 1 activity, infant sex, and gestational age (Continued) L o w -l V C l O P inSCCtlE expOSUTES 5 5 3 DOI Low-level OP insectide exposures Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 554 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 2. (Continued) Reference W olffetal. (2007) Outcome Exposure Maternal prenatal urinary MDA > 0.3 vs. < 0.3 jig/L (limit of detection) Wolff et al. (200 /) Ponderal index (g/cm3) Maternal prenatal urinary DAPs (nmol/L or nmol/g creatinine, logl0 scale) Wolff et al. (2007) Maternal prenatal urinary DMPs (nmol/L or nmol/g creatinine, logl0 scale) W olffetal. (2007) Maternal prenatal urinary DEPs (nmol/L or nmol/g creatinine, logJp scale) W olffetal. (2007) Maternal prenatal urinary MDA > 0.3 vs. < 0.3 ug/L (limit of detection) Wolff et al. (2007) Head circumference (cm) Maternal prenatal urinary DAPs (nmol/L or nmol/g creatinine, log10 scale) W olffetal. (2007) Maternal prenatal urinary DMPs (nmol/L or nmol/g creatinine, logl0 scale) Wolff et al. (2007) Maternal prenatal urinary DEPs (nmol/L or nmol/g creatinine, log10 scale) W olffetal. (2007) m i Maternal prenatal urinary MDA > 0.3 vs. < 0.3 ' Ug/L (limit of detection) Number of subjects/ events 318 27 318 330 318 327 318 330 Estimate of association (95% Cl) Beta S E = -0 .1 6 0.28, P --0.56 (not creatinineadjusted) Beta SE - - 0.032 30, P --0.91 (creatinine-adjusted) Beta SE - -0 .0 0 2 0.023, P --0.93 (not creatinineadjusted) Beta SE = - 0.003 0.023, P --0.91 (creatinine-adjusted) Beta SE = 0.01 0.02, P --0.47 (not creatinineadjusted) Beta SE - 0.01 0.02, P --0.48 (creatinine-adjusted) Beta SE - - 0.04 0.02, P --0.087 (not creatinineadjusted) Beta SE = - 0.04 0.02, P ~ 0.077 (creatinine-adjusted) Beta SE = 0.039 0.035, P --0.27 (not creatinineadjusted) Beta' SE --0.035 0.036, P --0.33 (creatinine-adjusted) Beta SE - - 0 .2 6 0.13, P --0.045 (not creatinineadjusted) Beta SE = --0.25 0.13, P --0.056 (creatinine-adjusted) Beta S E - --0.16 0.11, P --0.14 (not creatinineadjusted) Beta SE - -0 .1 5 0.11, P --0.16 (creatinine-adjusted) Beta SE - - 0.067 0.12, P --0.57 (not creatinineadjusted) Beta SE - -0 .0 5 2 0.12, P --0.67 (creatinine-adjusted) Beta S E -0 .1 5 0.19, P --0.44 (not creatinineadjusted) Beta SE - 0.23 0.20, P --0.25 (creatinine-adjusted) m Adjustment factors Comments No significant interactions between DAPs and PON 1 were detected for ponderai index (results NR) No significant interactions between DAPs and PON 1 were detected for head circumference (results NR) CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00024 DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Wolff etal. (2007) Gestational age Maternal prenatal urinary 318 (weeks) DAPs (nmol/L or nmol/g creatinine, log10 scale) Wolff etal. (2007) Maternal prenatal urinary DMPs (nmol/L or nmol/g creatinine, log10 scale) Wolff etal. (2007) Maternal prenatal urinary 318 DEPs (nmol/L or nmol/g creatinine, logJp scale) Wolff etal. (2007) Maternal prenatal urinary 330 MDA > 0.3 vs. < 0 .3 ' ug/L (limit of detection) Eskenazi et al. (2004) Length of gestation (weeks) Maternal prenatal urinary DAPs, DMPs. or DEPs (nmol/L, log10 scale) 485 with DAPs 485 with DMPs 486 with DEPs Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) m m i m Maternal prenatal urinary MDA (ug/L) Maternal prenatal urinary TCPy (ug/L) Maternal prenatal urinary PNP (ug/L) 233 undetectable 74 detectable < median 75 detectable te median 41 undetectable 220 detectable < median 221 detectable - median 124 undetectable 179 detectable < median 179 detectable > median Beta SE --0.03 0.14, P --0.81 (not creatinineadjusted) Beta SE = 0.03 0.14, P --0.83 (creatinine-adjusted) Beta SE = - 0.029 d. 12, P ~ 0.80 (not creatinineadjusted) Beta SE - -0.030 0.12, P --0.80 (creatinine-adjusted) Beta SE --- 0.006 0.13, P --0.996 (not creatinineadjusted) Beta SE = - 0.004 0.13, P ~ 0.97 (creatinine-adjusted) Beta SE = --0.28 0.21, P --0.18 (not creatinineadjusted) Beta SE - - 0.30 0.22, P --0.16 (creatinine-adjusted) Beta - - 0.20 ( - 0.55, 0.15) Beta - - 0.41 (-0 .7 5 , - 0.07) Beta - - 0.16 ( - 0.53. 0.22) Beta --referent Beta - - 0.13 (--0.55, 0.30) B e ta - -0.21 ( - 0.62. 0.20) Beta --referent Beta - - 0.17 ( - 0.74, 0.40) Beta - - 0.06 (--0.63, 0.51) Beta --referent Beta - --0.37 ( - 0.76, 0.02) Beta-0.18 (-0.21,0.57) Race/ethnicity, maternal PON1 activity, and infant sex Timing of urine collection, timing of entry into prenatal care, maternal age, parity, maternal country of birth, and poverty level Gestational age based on medical record; results similar when based on maternal self-reported date of last menstrual period Results persisted when metabolite levels were controlled for creatinine Inverse association with DMPs was most apparent for specimens collected after 22 weeks of gestation Associations of DEAMPY, TMPY, CMHC, and CIT with birth outcomes not analyzed due to small percentage of women with detectable levels (Continued) Tier 3/4 Low-lcwl OP USECtdE EXpOSUTES 555 ED 002061 00046431-00025 V 556 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued) Reference Eskenazi et al. (2004) Outcome Eskenazi et al. (2004) Eskenazi et al. (2004) Birth weight (g) Exposure Maternal/cord blood cholinesterase (.tmol/ min/niL) Maternal/cord plasma butyryichoiinesterase (pmoi/min/mL) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log1() scale) Number of subjects/ events 340 maternal blood, prenatal 357 maternal blood, delivery 292 cord blood 340 maternal plasma, prenatal 357 maternal plasma, delivery 292 cord plasma 485 with DAPs 485 with DMPs 486 with DEPs Estimate of association (95% Cl) Beta = 0.01 ( - 0.15, 0.17) Beta - 0.09 ( - 0.04, 0.23) B eta-0 .3 4 (0.13,0.55) B e ta - 0.2 ( - 0.64. 0.27) Beta - --0.1 ( - 0.48, 0.36) Beta - - 0.2 ( - 0.78, 0.32) B eta- 4 2 ( - 4 6 , 131) Beta - 4 1 ( - 4 0 , 122) B eta- 5 2 ( - 4 0 , 144) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) m i m Maternal prenatal urinary MDA (jg/L) Maternal prenatal urinary TCPy (pg/E) Maternal prenatal urinary PNP (p.g/L) Maternal/cord blood cholinesterase (pmol/ min/mL) Maternal/cord plasma butyryichoiinesterase (pmol/min/ml.,) 233 undetectable 74 detectable < median 75 detectable > median 41 undetectable 220 detectable < median 221 detectable 3; median 124 undetectable 179 detectable < median 179 detectable > median 340 maternal blood, prenatal 357 maternal blood, delivery 292 cord blood 340 maternal plasma, prenatal 357 maternal plasma, delivery 292 cord plasma Beta --referent Beta - - 4 5 (- 1 5 4 ,6 3 ) Beta - 56 ( - 4 9 , 161) Beta - referent Beta - - 6 ( - 138, 126) Beta --27 (--106, 159) Beta --referent Beta - 34 ( - 57, 125) B eta- 4 9 ( - 42, 140) Beta --8 ( - 35, 52) Beta --6 (--30, 43) B eta- 12 ( - 4 6 . 70) Beta - 56 ( - 67, 179) Beta - - 90 ( - 206, 25) Beta - 111 (- 3 5 .2 5 7 ) Adjustment factors Comments When gestational age was based on maternal self-reported date of last menstrual period, beta for lower cholinesterase in maternal blood - 1.1 days, P 0.04 Timing of urine collection, timing of entry into prenatal care, maternal age, parity, infant sex, maternal country of birth, pregnancy weight gain, body mass index, poverty level, gestational age, and gestational age squared Tier 3/4 Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00026 m. ED 002061 00046431-00027 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Eskenazi et al. (2004) Eskenazi et ai. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. 2004 Eskenazi et ai. (2004) Eskenazi et al. (2004) Eskenazi et al. 2004 m 1 m Body length (cm) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, loglf) scale) 485 with DAPs 485 with DMPs 486 with DEPs B eta-0 .5 2 (- 0 .0 1 , 1.05) Beta - 0 .4 2 (- 0 .0 7 , 0.91) B eta- 0 ,4 0 ( -0 .1 5 , 0.94) Maternal prenatal urinary MDA (jig/L) Maternal prenatal urinary TCPy (j.ig/1.) Maternal prenatal urinary PNP (p.g/L) Maternal/cord blood cholinesterase (pmoi/ min/mL) Maternal/cord plasma butyrylcholinesterase (pmol/min/ml.) Head circumference (cm) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log1(l scale) 233 undetectable 74 detectable < median 75 detectable > median 41 undetectable 220 detectable < median 221 detectable - median 124 undetectable 179 detectable < median 179 detectable > median 340 maternal blood, prenatal 357 maternal blood, delivery 292 cord blood 340 maternal plasma, prenatal 357 maternal plasma, delivery 292 cord plasma 485 with DAPs 485 with DMPs 486 with DEPs Beta --referent Beta - -0 .5 3 ( - 1.18, 0.11) Beta = 0.14 ( - 0.48, 0.76) Beta --referent B e ta - 0.09 ( - 0.70,0.87) Beta - 0.44 ( - 0.35, 1.22) Beta --referent Beta - 0.60 (0.06, 1.13) B eta-0 .4 1 (- 0 .1 3 , 0.94) B eta-0 .0 5 ( 0.20, 0.29) B eta-0 .0 5 (- 0 .1 7 , 0.27) Beta - - 0.01 ( - 0.35, 0.34) Beta - 0.07 ( - 0.63, 0.78) Beta - 0.05 ( - 0.65, 0.75) Beta = 0.23 (- 0 .6 5 , 1.12) B eta-0 .3 2 (0.03, 0.62) B eta-0 .2 5 ( - 0.02, 0.52) Beta-0 .2 8 (- 0 .0 2 , 0.59) Maternal prenatal urinary MDA (ug/L) Maternal prenatal urinary TCPy (pg/L) 233 undetectable 74 detectable < median 75 detectable ^ median 41 undetectable 220 detectable < median 221 detectable = median Beta --referent B e ta - - 0 .1 6 (-0 .5 2 , 0.19) B eta-0 .1 1 (- 0 .2 4 , 0.46) Beta --referent Beta = 0.06 ( - 0.37, 0.49) Beta - 0.04 ( - 0.39, 0.47) V No association when metabolite levels were controlled for creatinine Positive association with DAPs did not vary substantially by week of prenatal urine collection Results persisted when metabolite levels were controlled for creatinine Positive association with DAPs did not vary substantially by week of prenatal urine collection (Continued) Low-level OP insectide exposures 5 5 7 d o i Low-level o p insectide exposures 558 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued ) Reference Eskenazi et ai. (2004) Outcome Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et ai. (2004) Ponderai index (g/cm3) Eskenazi et ai. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Eskenazi et ai. (2004) m Preterm delivery Exposure Maternal prenatal urinary PNP (pg/L) Maternal/eord blood cholinesterase (prnol/ rain/niL) Maternal/cord plasma butyrylcholinesterase (umol/min/niL) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log10 scale) Maternal prenatal urinary MDA (pg/L) Maternal prenatal urinary TCPy (pg/L) Maternal prenatal urinary PNP (pg/L) Maternal/cord blood cholinesterase (pmol/ min/mL) Maternal/cord plasma butyrylcholinesterase (prnoi/niin/mL) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,() scale) Number of subjects/ events 124 undetectable 179 detectable < median 179 detectable > median 340 maternal blood, prenatal 357 maternal blood, delivery 292 cord blood 340 maternal plasma, prenatal 357 maternal plasma, delivery 292 cord plasma 485 with DAPs 485 with DMPs 486 with DEPs 233 undetectable 74 detectable < median 75 detectable - median 41 undetectable 220 detectable < median 221 detectable te median 124 undetectable 179 detectable < median 179 detectable > median 340 maternal blood, prenatal 357 maternal blood, delivery 292 cord blood 340 maternal plasma, prenatal 357 maternal plasma, delivery 292 cord plasma 32 (6.6%) preterm Estimate of association (95% Cl) Beta = referent Beta --0.18 (- 0 .1 2 , 0.48) Beta --0.29 (- 0 .0 1 , 0.58) Beta = 0.06 ( - 0.09, 0.21) Beta = - 0.07 (--0.19, 0.05) Beta - - 0.04 f - 0.23. 0.14) Beta = 0.12 (--0.31,0.56) Beta - - 0.07 ( - 0.45, 0.31) Beta - -0 .0 3 ( - 0.50. 0.45) Beta - - 0.04 ( - 0.12, 0.04) Beta - - 0.03 (-0 .1 0 , 0.04) Beta - --0.01 ( - 0.09, 0.07) Beta --referent Beta = 0.05 (- 0 .0 5 , 0.14) Beta - 0.02 (- 0 .0 7 , 0.12) Beta --referent Beta - -0 .0 1 (-0 .1 2 , 0.11) Beta - - 0 .0 4 ( - 0.16. 0.08) Beta --referent Beta - - 0.08 (- 0 .1 6 , 0.0) Beta - - 0.03 (--0.11,0.05) Beta - 0.00 ( - 0.03, 0.03) Beta - 0.00 ( - 0.03, 0.03) Beta - 0.02 ( - 0.03, 0.07) Beta --0.03 ( --0.06, 0.12) Beta - --0.07 ( - 0.16, 0.03) B eta-0 .0 5 (- 0 .0 7 , 0.17) "not associated" (results NR) Adjustment factors Comments NR Preterm delivery: birth at < 37 completed weeks of gestation Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00028 s* Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Eskenazi et al. (2004) Eskenazi et ai. (2004) Eskenazi et al. (2004) Eskenazi et al. (2004) Harley et al. (2011) Harley et al. (2011) Harley et al. (2011) Harley et al. (2011) m 1 m Low birth weight Small for gestational age birth Gestational age (weeks) Maternal/cord blood cholinesterase (pmol/ rain/niL; per unit decrease) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log10 scale) Cord blood cholinesterase (umoEnurPniL; per unit decrease) Maternal prenatal urinary DAPs, DMPs. or DEPs (nmol/L, log10 scale) Maternal prenatal urinary DAPs (nmol/L, loglr, scale) by child genotype NR preterm with maternal prenatal blood NR preterm with cord blood 18 (3.7%) low birth weight 11 low birth weight with cord blood 23 (48%) small for gestational age birth 76 PON1 ___igs TT 225 P O N 1 - ,0B CT 131 PON1 --`]p8 CC 108 P O N i ,9} QQ 222 P O N l,91 QR 106 P O N lm RR Odds ratio = 1.6 (1.0, 2.5) Odds ratio = 2,3 (1.1, 4.8) "not associated" (results NR) Odds ratio = 4,3 (1.1, 17,5) "not associated" (results NR) Beta ==--0.8 ( - 2.0, 0.2) Beta = - 0.3 (--0.8, 0.3) Beta = 0.1 ( - 0 .7 , 0.8) P-interaction = 0,36 Beta = - 1 . 0 ( - 2.0, 0.0) Beta = - 0.2 ( - 0.7, 0.3) Beta = 0.2 (--0.7, 1.1) P-interaction = 0.21 " Maternal prenatal urinary 76 P O N I __ ,p8 TT Beta = --0.6 (--1.6, 0.4) DMPs (nmol/L, log,0 scale) by child genotype 225 P O N I CT 131 P O N I - ^ g CC Beta = - 0.3 ( - 0.8, 0.2) Beta = 0.0 ( - 0.7, 0.7) P-interaction = 0.49 108 P O N I ,9, QQ Beta = 0.7 ( - 1.7, 0.2) 222 P O N l , ^ QR Beta = --0.3 ( - 0 .7 ,0 .! ) 106 P O N ! 9 7 RR Beta = 0,3 ( - 0 .5 , 1.2) P-interaction = 0.25 " Maternal prenatal urinary 76 P O N I __ m TT Beta = --1.0 (--2.1,0.1) DEPs (nmol/L, log10 225 P O N I CT Beta = - 0 .2 ( - 0 .8 , 0.3) scale) by child genotype 131 P O N I - ^ g CC Beta = 0.6 ( - 0.2, 1.4) P-interaction = 0.09 108 P O N I ,9, QQ Beta = - 1 . 0 (- 2 .1 ,0 .2 ) 222 P O N l , ^ QR Beta = 0.1 ( - 0 .4 , 0.6) 106 P O N ! 9 7 RR Beta = - 0 .3 ( - 1.2, 0.6) P-interaction = 0.17 " Maternal prenatal urinary PONI quantity: DAPs (nmol/L, iog10 108 tettile 1 Beta = 0.3 ( - 0 .5 , 1.2) scale) by cord blood 108 tettile 2 Beta = - 0.4 ( - 1 .2 , 0.4) PON! quantity 108 tertile 3 Beta = - 0.2 ( - 0.9, 0.5) P-interaction = 0.17 NR NR Low birth weight: < 2,500 g NR Six of 11 infants with low' birth weight w'ere also preterm NR Timing of urine collection. timing of entry into prenatal care, maternal age, parity, maternal country of birth, and household income Small for gestational age birth: birth weight < 10th percentile for gestational age according to ethnicity, parity, and infant sex Arylesterase activity (a marker of PON1 enzyme quantity) was lowest in mothers and infants with PON1 !()s I T (but highest in mothers with P O N li9, QQ) Paraoxonase activity (a marker of PON! enzyme activity) was lowest in mothers and infants with either PON1 _ m TT or P O N lm QQ, especially both (Continued) m. L o w -lV l OP if tS C tid XpO SU T S ED 002061 00046431-00029 559 ED 002061 00046431-00030 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued ) Reference Harley et al. (2011) Outcome Harley et al. (2011) Harley et al. (2011) Birth weight (g) Harley et al. (2011) Harley et al. (2011) Harley et al. (2011) Harley et al. (20*11) m Exposure Maternal prenatal urinary DMPs (nmol/L, log10 scale) by cord blood PON1 quantity Maternal prenatal urinary DEPs (nmol/L, logl(t scale) by cord blood PON1 quantity Maternal prenatal urinary DAPs (nmol/L, logUj scale) by child genotype Maternal prenatal urinary DMPs (nmol/L, log,0 scale) by child genotype Maternal prenatal urinary DEPs (nmol/L, log10 scale) by child genotype Maternal prenatal urinary DAPs (nmol/L, log10 scale) by cord blood PON1 quantity Maternal prenatal urinary DMPs (nmol/L, log,0 ' scale) by cord blood PQN1 quantity Number of subjects/ events PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 76 P O N I _ m TX 225 P O N I CT __ ]p s 131 P O N I __ Jm CC 108 P 0 N 1 I Q 2 Q Q 222 P O N I ,9, QR 106 P O N m RR 76 P O N I igs TX 225 P O N I __ ,0, CX 131 P O N I _ ] S C C 108 P O N I ,9} QQ 222 P O N l jg-, QR 106 P O N 1 9 2 RR 76 P O N I igs TX 225 P O N I __ ,0, CX 131 P O N I _ ] S C C 108 P O N I ,9} QQ 222 P O N l , 9 \ QR 106 P 0 N 1 9 2 RR PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 Estimate of association (95% CI) B eta- 0 .5 ( - 0 .3 , 1.2) Beta - - 0 .4 ( - 1.1,0.4) Beta - - 0.4 ( - 1 .0 , 0.3) P-interaction --0.16 Beta - - 0. (--1.2, 0.4) B e ta - 0.7 ( - 1.5, 0.1) B eta- 0 .5 ( - 0 .3 , 1.2) P-interaction --0.69 Beta - - 131.3 ( - 393.3, 130.8) Beta - 147.2 (18.5, 275.7) Beta = 22.1 ( - 182.0, 226.3) P-interaction --0.06 Beta ----60.2 (--266.3, 145.9) B eta-7 9 .4 (- 4 8 .5 , 207.3) Beta - 142.3 ( - 114.6, 399.3) P-interaction --0.20 Beta - - 135.2 ( - 373.8, 103.3) Beta = 134,6 (14.9, 254,2) B eta-4 6 .8 ( - 132.5,226.1) P-interaction --0,05 Beta - - 80.6 ( - 269.2, 107.9) B eta-8 9 .9 (- 2 7 .8 , 207.5) Beta = 72.9 ( - 169.2,315.0) P-interaction --0.16 Beta - - 55.3 ( - 320.5. 209.8) Beta - 120,8 ( - 14.8,256.4) Beta = 45.4 ( - 174.6, 265.4) P-interaction --0,35 Beta - 20.5 ( - 210.5, 251.5) Beta - 6 7 .2 (- 6 3 .3 , 197.5) Beta-258.8 (23.9,493,6) P-interaction --0.30 Beta - --14.6 ( - 263.4, 234.3) B eta-6 3 .8 ( - 131.1,258.7) Beta = 92.2 ( - 106.6, 291.1) P-interaction --0,39 Beta - 14.8 (-2 1 6 .2 ,2 4 5 .7 ) Beta-8 3 .9 ( - 94.8,262.6) B e ta - 60.2 ( - 120.2,240.7) P-interaction --0.29 Adjustment factors Timing of urine collection, timing of entry into prenatal care, maternal age, parity, maternal country of birth, household income, prepregnancy body mass index, maternal weight gain during pregnancy, infant sex, gestational age, and gestational age squared s* Comments CritRevTcocoL 2015; 45(7): 531-641 560 R. Reiss et al. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Harley et al. (2011) Hailey et al. (2011) Harley et al. (20*11) Harley et al. (20*11) Harley et al. (20*11) Harley et al. (2011) Harley et al. (2011) Barr et al. (2010) m m Barr et al. (2010) i m Maternal prenatal urinary DEPs (ntnol/L, iog10 scale) by cord blood PON1 quantity Head circumference (cm) Maternal prenatal urinary DAPs (nmol/L, log1P scale) by child genotype Maternal prenatal urinary DMPs (nmol/L, log,0 ' scale) by child genotype Maternal prenatal urinary DEPs (nmol/L, log,0 scale) by child genotype Maternal prenatal urinary DAPs (nmol/L, ioglp scale) by cord blood PQN1 quantity Maternal prenatal urinary DMPs (nmol/L, log,(l scale) by cord blood PON1 quantity Maternal prenatal urinary DEPs (nmol/L, iogJ(1 scale) by cord blood PON1 quantity Birth weight (g) Head circumference (cm) Maternal, pre/perinatal or cord serum chlorpyrifos (ng/g) > 75th vs. rat 75th percentile PONI quantity: 108 tettile 1 108 tertile 2 108 tertile 3 76 PONI igs TT 225 PONI ,9f, CT 131 PONI _ )m CC 108 PONI ,9} QQ 222 P O N lJtg' QR 106 P O N l192 RR 76 PONI _ ,0H TT 225 PONI CT 131 PONI __m CC 108 PONI ,9, QQ 222 P O N m QR 106 P O N l,92 RR 76 PONI _ ,0H TT 225 PONI CT 131 PONI __m CC 108 PONI ,9, QQ 222 P O N ll92 QR 106 P O N l,92 RR PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 PONI quantity: 108 tertile 1 108 tertile 2 108 tertile 3 138 materna! serum 148 cord scruni 138 maternal serum 148 cord serum Beta - - 55.2 ( -295.6, 185.2) Beta = 48.7 ( - 134.4, 231.8) Beta = 231.4 (19.1,443.6) P-interaction = 0.31 Beta = 0.1 ( - 0.6, 0.9) Beta = 0.2 (--0.2, 0.6) Beta = 0.6 ( - 0 .1 , 1.3) P-interaction = 0,08 Beta = - 0 .3 ( - 1.0, 0.4) Beta = 0.2 ( - 0.2, 0.6) Beta = 0.7 (--0.1, 1.5) P-interaction = 0.01 Beta = 0.2 (--0.5, 0.8) Beta = 0.1 ( - 0 .3 , 0.5) Beta = 0.5 ( - 0 .2 , 1.1) P-interaction = 0.12 Beta = - 0 .4 ( - 1.0, 0.2) Beta = 0.2 (--0.2, 0.5) Beta = 0.5 ( - 0 .3 , 1.3) P-interaction = 0,01 Beta = - 0.1 (--0.8, 0.7) Beta = 0.2 ( - 0.2, 0.6) Beta = 0.6 ( - 0 .2 , 1.4) P-interaction = 0.19 Beta = 0.1 ( - 0 .7 , 0.9) Beta = 0.1 (--0.3, 0.5) Beta = 0.7 (0.0, 1.5) P-interaction = 0,27 Beta = --0.2 (--1.0, 0,5) Beta = 0.3 ( - 0 .3 , 0.9) Beta = 0.8 (0.1, 1.4) P-interaction = 0.32 Beta = -0 .1 ( - 0 .8 , 0.5) Beta = 0.1 ( - 0 .4 , 0.7) Beta = 0.7 (0.1, 1.3) P-interaction = 0.36 Beta = - 0 .5 ( - 1.2, 0.2) Beta = 0.4 ( - 0 .2 , 1.0) Beta = 0.7 ( - 0 .1 , 1.4) P-interaction = 0.48 Mean SD = .3053 111 vs. 3548 448, P = 0.268 Mean SD = 3581 422 vs. 3544 4 3 3 ,/' = 0.408 Mean SD = 33.4 0.6 vs. 35.0 1.3, P = 0.229 Mean SD = 34.1 1.4 vs. .35.0 1.2, P = 0.989 Maternal age, primigravida, race, pre-pregnancy body mass index, infant sex, and gestational age Results were similar when pesticide levels were dichotomized at the 90th percentile (results NR) (Continued) Tier 3/4 Low-lcwl OP iSE C tdE EXpOSUTES 561 ED 002061 00046431-00031 V 562 R. R(?iss t al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued) Reference Barr et al. (2010) Barr et al. (2010) Outcome Abdominal circumference (in) Birth length (cm) Exposure " " Wang et al. (2011) Length of gestation (weeks) Maternal perinatal urinary DAPs (log scale) [unit (nmol/L or nmol/g creatinine) and log base not specified] Wang et al. (2011) Wang et al. (2011) Birth weight (g) m m i ; m Number of subjects/ events 138 maternal serum 148 cord serum 138 maternal serum 148 cord serum 187 91 infant girl s 187 Estimate of association (95% Cl) Mean SD - 29.2 0.8 vs. 32.0 2.7, P = 0.201 Mean SD = 32.5 2.3 vs. 32.0 2.7, P = 0.346 Mean SD = 49.8 0.2 vs. 51.3 .3.0, P --0,686 Mean SD - 50,9 1.7 vs. 51.4 3.1, P --0.318 Dimethylphosphate beta - --0.05 ( - 0.52-0.33) Dimethylthiophosphate beta --0.15 (--1.21-1.03) Diethylphosphate beta = 0.11 ( - 1.27-0.52) Diethylthiophosphate beta - 0.13 (-- 0.92-0.64) Diethyldithiophosphate beta = --0.03 (--0.04--0.06) DAPs beta - 0.04 ( - 0.35-0.59) Dimethylphosphate beta = 0.39 (--0.13, 0.63) Dimethylthiophosphate beta --0.31 (-0 .0 8 -0 .6 3 ) Diethylphosphate beta -- 1.79 ( - 2.82 to - 0.76) [Boys: diethylphosphate beta - 0.17, P --0.164] Diethylthiophosphate beta - 0.72 (-0 .2 8 , 1.16) Diethyldithiophosphate beta = 0.09 ( - 0.35-0.53) DAPs beta = --0.03 (--0.81 -- 0.61) Dimethylphosphate beta = --18 ( - 151-109) Dimethylthiophosphate beta --84 ( - 50-304) Diethylphosphate beta --135 ( - 143-402) Diethylthioph ospha te beta ----112 (-3 1 8 -1 5 9 ) Diethyldithiophosphate beta - 4 (-161-313) DAPs beta --69 ( --74-212) Adjustment factors Comments Maternal height, pregnancy weight gain, and family income Some apparent reporting errors (e.g,, missing " signs) are corrected here based on reported P-values Results were unchanged when preterm infants were excluded (results NR) Gestational age, maternal height, pregnancy weight gain, and family income Tier 3/4 Grit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00032 last ED 002061 00046431-00033 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Wang et al. (2011) " Wang et ai. (2011) Body length (cm) Wang et al. (2011) " Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. " 91 infant girls Dimethylphosphate beta -- 48 (-- 192-218) Dimethylthiophosphate beta = 166 (--40-473) Diethylphosphate beta --174 (-287-529) Diethylthiophosphate beta = - 272 ( - 499-208) Diethyldithiophosphate beta = 45 (-278-412) DAPs beta ----6 (--286-240) " 187 Dimethylphosphate beta - --0.01 Gestational age, maternal (-- 0.67-0.61) height, pregnancy weight gain, Dimethylthiophosphate and family income beta = --0.04 (--0.67-1.07) Diethylphosphate beta --0,12 ( - 0.65-2.02) Diethylthiophosphate beta - - 0.16 (-2 .0 3 -0 .3 1 ) Diethyldithiophosphate beta = - 0.01 (--1.22-1.10.) DAPs beta - 0.03 ( - 0.47-0.73) " 91 infant girls Dimethylphosphate beta = - 0 .0 6 ( - 0.73-0.48) Dimethylthiophosphate beta --0.24 (-0 .1 1 -1 .4 0 ) Diethylphosphate beta --0.33 (-1.00-1.41) Diethylthiophosphate beta - - 0.16 ( - 1.54-0.55) Diethyldithiophosphate beta - - 0.09 ( - 1.40-0.64) DAPs beta = --0.17(--1.84- 0.21) m m :i ; m ip: V' (Continued) Low-level OP insectide exposures 563 doi Lovv-ievei op insectide exposures 564 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 2. (Continued ) Reference Rauch et al. (2012) Outcome length of gestation (weeks) Exposure Maternal prenatal urinary DAPs (nmol/JL, log,0 scale, creatininestandardized) Number of subjects/ events 306 total 93 black mothers 213 white mothers Estimate of association (95% Cl) Beta = 0.5 ( - 0 .8 , - 0 .1 ) Beta - - 0.1 (--0.9. 0,6) Beta - - 0.7 ( - 1 .1 , --0.3) P-interaction by race :::: 0.10 Rauch et al. (2012) Rauch et al. (2012) Rauch et al. (2012) m Rauch et al. (2012) 1 Birth weight (g) m Maternal prenatal urinary DMPs (nmol/L, log,0 scale, creatininestandardized) Maternal prenatal urinary DEPs (nmol/L,logl0 scale, creatininestandardized Maternal prenatal urinary DAPs (nmol/L, iog10 scale, creatininestandardized) 55 P O N I,.,, RR 107 PONI 192 QR 111 PONI ,97 QQ 118 PONI -JOS 106 PONI - os CT 46 PONI _- m TM 306 total 93 black mothers 213 white mothers 306 total 93 black mothers 213 white mothers 306 total 93 black mothers 213 white mothers Beta ::: - 0.3 ( - 1.2, 0.5) Beta = 0.9 ( - 1.6, - 0.3) Beta = - 0.5 (--1.1, 0.0) P --interaction by genotype --0.04 for QR, 0. for QQ Beta = 0.3 ( - 0.9, 0.3) Beta = - 1.0 ( - 1 .6 , - 0 .4 ) Beta - - 0.1 ( - 1.0, 0.8) P --interaction by genotype = 0.04 for CT, 0.: for TT Beta :::: 0.4 i - 0.7, 0.0) Beta -==0.0 ( - 0.7. 0.6) Beta = 0.6 ( - 0.9, - 0.2) P-interaction by race = 0.09 Beta ::: - 0.2 ( - 0 .5 , 0.1) Beta = 0.1 ( - 0.8, 0.5) Beta = - 0.3 (--0.7, 0.0) P-interaction by race - 0.47 Beta = 151 ( - 2 8 7 , - 16) Beta - - 188 ( - 395. 19) Beta - - 118 1 - 296, 60) P-interaction by race :::: 0.46 Adjustment factors Mother's age, mother's race (unless stratified), household income, marital status, parity category, log10-transformed blood lead, and log10transformed cotinine " Comments Results were similar when excluding mothers with abruptio placenta, placenta previa, chorioamnionitis, pre-eclampsia, or pregnancyinduced hypertension, or when excluding mothers with urinary creatinine < 20 mg/dl. Results were similar or attenuated (beta for DAPs and gestational age = --0.2 [--0.4, 0.01; beta for DAP and birth weight = --0.88 [--213, 371) when restricted to fullterm births Results did not vary significantly by child sex (P-interaction > 0.3) Results were "modestly attenuated" when based on non-creatinineadjusted DAPs (beta for gestational age = - 0.3 [- 0.7, 0.01; beta for birth weight ----100 [--232, 32]), and "slightly attenuated" when based on individual urine specimens from 16 or 26 weeks of gestation When stratified by race, effect estimates were also stronger in heterozygous groups Tier 3/4 CritRevTcocoL 2015; 45(7): 531-641 ED 002061 00046431-00034 V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. doi Low-ievei op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Rauch et ai. (2012) Rauch et al. (2012) Rauch et al. (2012) Rauch et al. (2012) Birth weight, adjusted for gestational age (g) Rauch et al. (2012) Rauch et al. (2012) Wickerham et ai. (2012) Birth weight (g) Wickerham et al. (2012) Maternal prenatal urinary DMPs (nmol/L, log,p scale, creatininestandardized) Maternal prenatal urinary DEPs (nmol/L, log|p scale, creatininestandardized Maternal prenatal urinary DAPs (nmol/L, ioglp scale, creatininestandardized) 306 total 93 black mothers 213 white mothers 306 total 93 black mothers 213 white mothers 306 total 93 black mothers 213 white mothers 55 PON1,9, RR 107 PON1 ,'92 QR 111 p o n i 197 q q 118 PON1 __,m CC 106 PON1 igs CT 46 PON1 _ i0gTT Maternal prenatal urinary DMPs (nmol/L, lo g ,p ' scale, creatininestandardized) Maternal prenatal urinary DEPs (nmol/L, logJp scale, creatininestandardized Number of organophosphate pesticides (of 8 tested) detected in cord serum Chlorpyrifos, diazinon, fonofos, maiathion, parathion-ethyi, parathion-methyl, profenofos, or terbufos in cord serum, detectable vs. non-detectabie or 3-level ordinal variables 306 total 93 black mothers 213 white mothers 306 total 93 black mothers 213 white mothers 112 Beta = --124 ( - 2 4 5 , - 2 ) Beta ----142 (--.3.3.3, 50) Beta - -- 96 (--254,62) P-interactionby race --0.46 Beta - --65 (--180, 51) Beta ----162 (--340, 16) Beta - - 3 9 ( - 189, 111) P-interaction by race --0.39 Beta ----40 ( --146, 65) Beta - - 158 ( - 2 9 7 , - 18) Beta - 60 ( - 84, 204) P-interaction by race - 0.02 Beta - - 7 1 (-.384,242) Beta= --454(--707,-201) Beta - - 2 ( - 231, 228) P --interaction by genotype --0.02 for QR, 0.76 for QQ vs. RR B e t a - - 1 1 9 ( - 3 4 0 , 103) Beta - - 299 ( - 520, - 7 8 ) Beta --85 (--361, 530) P - interaction by genotype --0.15 for CT, 0.12 for I T vs, CC Beta ----38 ( --133, 56) B e t a - - 1 3 9 ( - 2 6 7 , - 10) B eta- 4 9 ( - 7 8 , 177) P-interaction by race - 0.02 Beta - - 9 ( - 9 9 , 80) B e t a - - 1 3 1 ( - 2 5 1 , --11) Beta - 4 1 ( - 7 8 , 160) P-interaction by race --0.08 B eta-6 .5 9 ( - 2 1 0 , 222) No significant associations (results NR) When stratified by race, effect estimates were also stronger in heterozygous groups Gestational age, maternal age, maternal body mass index at early pregnancy, and maternal hemoglobin at delivery Of 20 pesticides measured, the mean SD number detected in cord serum per subject was 4.6 1.9, with a maximum of 10; 98.3% had at least one pesticide detected *" D A P dialtyl phosphate, D E P diethyl phosphate, D M P dimethyl phosphate, M D A maiathion dicaiboxylic acid, N R not reported, PNP 4-nitrophenol, PON1 paraoxonase 1, SD standard deviation, SE standard error, 8 TCPy 3,5,6-trichloro-2-pyridinol. 1 m m. SA Low-level O P insectide exp o su res 565 ED 002061 00046431-00035 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 566 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 404 consecutive healthy, primiparous pregnant women with a (P = 0.019). However, such interactions apparently were not singleton pregnancy at < 2 6 weeks of gestation in 1998 -2001 detected (or at least were not reported) for total DAPs or for (Table 1) (Berkowitz et al. 2004, Wolff et al. 2007). OP other birth outcomes. metabolites, including TCPy (a metabolite of chlorpyrifos and Notable strengths of the Mount Sinai CECS are comparable chlorpyrifos methyl), MDA (a metabolite of malathion), and to those of the CCCEH study, and include the personal mea six DAP metabolites were measured in maternal urine col surement of OP metabolites, the detailed characterization of lected during the third trimester. Median concentrations were the study cohort, and the prospective design, with exposures 7.6 pg/L (below the limit of detection [LOD] of 1 1.0 lig/L) measured during the third trimester of pregnancy. Participants (interquartile range [IQR] = 1.6-32.5) for TCPy (Berkow were enrolled during a narrower time window, preventing itz et al. 2004), < 0.3 itg/L (LOD) (range <0.3-15.8) tor evaluation of associations before and after 2001, but informa MDA, 75.9 nmol/L (range = < 1-4 [LODJ-4987) for DAPs, tion on P O N ] genotypes and PONl and BChE activity enabled 42.2 nmol/L (range < 1-4-4903) for DMPs, and 18.8 nmol/L the assessment of putatively susceptible subgroups. Several (range < 1-4-429) for DEPs (Wolff et al. 2007). In addi limitations are also shared between the cohorts, including the tion, five genetic polymorphisms (Q192R, L55M, C-909 G, collection of single biospecimens for exposure assessment, the A-162 G, and C-108 T) in the PON1 gene, PON1 enzymatic small number of subjects and multiple hypothesis testing in activity against phenyl acetate, and butyrylcholinesterase stratified analyses, and possible modest confounding. Thirty- (BChE) enzymatic activity against biityrylthioeholine were three percent of eligible women consented to participate, and assessed in the third-trimester maternal blood and umbili 74% of participants were ultimately included in the analysis cal cord blood, PON1 acts as a detoxifying enzyme for OP after exclusions due to medical issues, lack of biospecimens, metabolites, and higher-activity alleles (e.g,, P0N 1192 QQ change of hospital or residence, or refusal, with shorter and PONl_l0S CC) and higher enzyme levels are hypothesized follow-up for younger and less-educated mothers. If participa to protect against potential adverse health effects of OP expo tion and/or follow-up were associated with both OP exposure sure. A recent paper has found that PON 1 activity with phenyl and subsequent birth outcomes, then selection bias could have acetate as a substrate may not be a reliable index of the quan distorted the results in an unpredictable direction. Maternal tity of PON 1 protein, because the hydrolysis o f phenyl acetate prenatal urinary TCPy and MDA levels were analyzed as is not independent of genotype (McDaniel et al. 2014). dichotomous variables (detectable or non-detectable), preclud When maternal prenatal urinary TCPy concentration was ing analyses of exposure-response trends. As described earlier dichotomized at the LOD (11.0 pg/L), no significant associa in this paper, urinary DAP concentrations are unlikely to accu tion was observed with birth weight, birth length, or head cir rately reflect long-term exposure to OP insecticides. They also cumference after multivariate adjustment, based on 387 sub do not enable identification of associations with specific OP jects (Table 2) (Berkowitz et al. 2004). Moreover, none of the insecticides, which differ substantially in acute toxicity. three birth outcomes differed significantly by the presence of Taken together, the results in the Mount Sinai CECS cohort maternal prenatal urinary TCPy within strata of low, medium, show no association between maternal prenatal levels of TCPy or high maternal PON1 activity. Log ^-transformed concentra or MDA and birth outcomes. Higher maternal prenatal DAP tion of maternal prenatal urinary DAPs, with or without crea levels were associated with smaller head circumference, but tinine adjustment, was not significantly associated with birth not birth weight, birth length, ponderal index, or gestational weight, birth length, ponderal index, or gestational age, but age, while maternal prenatal DMP and DEP levels were not it was significantly inversely associated with head circumfer significantly associated with any of these outcomes in the ence (beta = - 0.26 cm, standard error [SE] = 0.13, P --0.045 combined cohort. The association with head circumference without creatinine adjustment; similar results with creatinine alone could be interpreted as indicative of a neurotoxic effect adjustment) (Wolff et al. 2007). Log10-transformed concentra or, alternatively, a chance for finding amid predominantly null tions of maternal prenatal urinary DMPs and DEPs also were results. Some interactions in the expected direction (assum not significantly associated with birth weight, birth length, ing greater susceptibility in those with lower PONl activity) ponderal index, head circumference, or gestational age, nor were detected between prenatal DMPs or DEPs and maternal was maternal prenatal urinary MDA concentration, dichoto PON1 activity or PONl ]92 genotype in association with birth mized at the LOD (0.3 llg/L), significantly associated with any weight or birth length. However, no explanation was provided of these outcomes. There was some evidence that maternal for why DMPs would interact with PON l activity and geno prenatal urinary DEPs interacted with maternal PON1 activity type in relation to birth length but not weight or other birth and PON1 l92 genotype. Birth weights were significantly lower outcomes, whereas DEPs would interact with PON1 activity among those with higher DEP concentrations and lower PON1 and genotype in relation to birth weight but not length or other activity or the PON1 m RR (low-activity) genotype, compared birth outcomes. The internally inconsistent findings indicate with those with lower DEP concentrations and higher PON1 no clear relationship between prenatal exposure to OP insecti activity (P = 0.042) or the PONl 192 QQ (high-activity) geno cides and fetal growth or gestational age. type (P = 0.020). Also, birth length was significantly shorter among those with higher maternal prenatal urinary DMP con centrations than among those with lower DMP concentrations within the stratum of lower maternal PON 1 activity (P = 0.032), C enter fo r the H ealth A ssessm ent o f M others and Children o f Salinas and among those with higher maternal prenatal urinary DMP The CHAMACOS prospectively enrolled pregnant women concentrations and the PONl m RR genotype, compared with entering prenatal care at < 2 0 weeks of gestation between those with lower prenatal DMP and the PONl 197 QQ genotype 1999 and 2000 in a primarily Latino, low-income, farm- SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00036 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 567 worker population in the Salinas Valley of California (Table 1) significantly associated with longer gestation, and lower levels (Eskenazi ei al. 2004, Harley et al. 2011). Six DAP metabolites were associated with a significantly higher odds of preterm were measured in maternal spot urine specimens collected at birth and low birth weight. However, activities in maternal approximately 13-14 weeks and 26-27 weeks of gestation prenatal and delivery blood were not significantly associated and averaged over the two time points. In addition, seven OP with length of gestation, birth weight, or the other outcomes insecticide metabolites, including TCPy, MDA, 4-nitrophenol assessed. Activities of BChE in maternal prenatal plasma, (PNP; a metabolite of methyl parathion, parathion, and other maternal plasma at delivery, and cord plasma at delivery also chemicals), 2-diethylamino-4-hydroxy-6-methylpyrimidine were not significantly associated with any of the outcomes (a metabolite of pirimiphos methyl), 2-isopropyl-4-methyl-6 examined. The authors did not have baseline AChE data; thus, -hydroxypyrimidine (a metabolite of diazinon), 3-chloro-4- AChE and BChE inhibition was not measured. Furthermore, methyl-7-hydroxycoumarin (a metabolite of coumaphos and AChE and BChE levels may vary significantly across time due coumaphos methyl), and 5-chloro-1-isopropyl-3-hydroxytri to changes in OP insecticide exposure and/or natural variabil azole (a metabolite of isazophos and isazophos methyl), were ity. In stratified analyses by PON] genotype or PON1 acti vity, measured in the same maternal prenatal urine specimens. The maternal prenatal urinary DEP levels were associated with last four were detected in fewer than 11%of subjects and there shorter gestational age only among infants with the P O N 1 _ 108 fore were not studied further. Median concentrations (range) TT (low-activity) genotype (Pinleraction= 0.09; Table 2) (Har in maternal prenatal urine were 136 nmol/L (10-6854) for ley et al. 2011). Maternal prenatal urinary DAP and DMP DAPs, 101 nmol/L (5-6587) for DMPs, 22 nmol/L (2-680) levels were (non-significantly) associated with higher birth for DEPs, 3.3 pg/L (0.2-56.1) for TCPy, 0.2 ig/L (0.2-28.9) weight only among those with the PON1 !08 CT genotype for MDA, and 0.5 (0.1-34.7) for PNP (Eskenazi et al. 2004). interaction = 0*06 and 0,05, respectively), whereas associations AChE and BChE activities were measured in maternal blood with birth weight were statistically non-significant among taken at -26-27 weeks of gestation and before delivery, and in those with the TT genotype. Positive associations with head umbilical cord blood taken at delivery. In addition, P O N 1 ,9? circumference were detected only among those with PON!_!l}8 and PON1 wg polymorphisms w;ere genotyped in maternal CT or P O N l m RR (low-activity) genotype. Cord blood and cord blood specimens, and PQN1 arylesterase activity PON1 arylesterase and paraoxonase activity levels were not against phenyl acetate (as a measure of PON1 quantity) and significant modifiers of the associations between maternal pre paraoxonase activity against paraoxon (as a measure of PON 1 natal DAP, DMP, or DEP concentrations and birth outcomes, activity) were measured in maternal post-delivery and cord although a significant positive association between prenatal blood specimens. DEPs and birth weight was detected only among those with After multivariate adjustment in 485 mother-newborn high cord blood levels of PON1. pairs, a 10-fold increase (i.e., 1 log,0-nmol/L increase) in The CHAMACOS study has several methodological maternal prenatal urinary DAP concentration was positively, strengths, including its relatively large size, evaluation of but not significantly, associated with birth length (beta = numerous potential confounders, and collection of several 0.52 cm, 95% Cl ----0.01, 1.05) and was significantly posi individual-level OP metabolites around the beginning of the tively associated with head circumference (beta = 0.32 cm, second and third trimesters of pregnancy. 95% Cl = 0.03, 0.62) (Table 2) (Eskenazi et al. 2004). After Limitations of CHAMACOS include use of on only two urinary DAP levels were controlled for creatinine, the asso averaged biospecimens to characterize exposure and the other ciation with birth length was no longer evident; however, the concerns identified above for the CCCEH and CECS studies, as result for head circumference did not change. Maternal prena well as the inherent shortcomings of DAP metabolites as bio tal urinary DAP concentrations were not significantly associ markers of OP insecticide exposure. Additionally, the 53,2% ated with length of gestation, preterm delivery (birth at < 37 participation rate (with a follow-up rate of approximately 88%) w'eeks of gestation), birth weight, low birth weight (< 2 5 0 0 raises concerns about selection bias, although the direction and g), ponderal index, or small size for gestational age at birth magnitude of such bias cannot be quantified. The main analy ( < 10th percentile for birth weight at gestational age), nor were ses in the whole CHAMACOS cohort indicate that maternal maternal prenatal urinary DMP or DEP concentrations signifi prenatal levels of DAPs, DMPs, DEPs, TCPy, and MDA, and cantly associated with these outcomes, other than an inverse activities of AChE or BChE, were not associated with most association between prenatal DMPs and length of gestation birth outcomes examined. The only exceptions were the posi (beta = --0.41 weeks, 95% Cl = --0.75, --0.07). In analyses tive association between DAPs and head circumference and by timing of prenatal DMP measurement, the latter association the inverse association between DMPs and length of gestation, appeared to be stronger after 22 weeks of gestation. Maternal especially when DMP concentrations were measured after the prenatal urinary TCPy and MDA levels were not significantly- midpoint of pregnancy. The associations of maternal prenatal associated with any outcome evaluated. When maternal pre PNP levels wi th shorter gestation and greater body length were natal urinary PNP levels were categorized as undetectable, not consistent with a monotonic exposure-response trend, and detectable, and below the median, or detectable and at or above the findings for AChE activities in cord blood were not con the median, newborns in the middle category of exposure, but sistent with the findings for acti vities in maternal prenatal and not the highest category, had a shorter length of gestation and perinatal blood. The observed associations were not modified longer birth length than those in the lowest category. However, by PON1 quantity or activity, but some evidence of modifi the authors cautioned that "PNP may derive from compounds cation by PON1 genotype was found, albeit with somewhat other than parathion" (Eskenazi et al. 2004). AChE activities contradictory patterns (e.g., positive associations with birth in cord blood (beta = 0.34 weeks, 95% Cl = 0.13, 0.55) were weight in P O N 1 l0S CT carriers, but positive associations with SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00037 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 568 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 head circumference in PON1197 RR carriers). The authors not 6.31 fig/L (range --< LOD-131.83) for DETP, and undetect only interpreted the inverse association between DMPs and able (range --< LOD-5.1; 5.34% detectable) for DEDTP. length of gestation as being consistent with a stimulatory In multivariate adjusted models including all 187 newborns, effect of OP insecticides on uterine contraction, but also noted no significant association was detected between any mater that the 6.4% rate of preterm delivery in this population was nal prenatal urinary DAP metabolite or all DAPs com lower than the U.S. average (Eskenazi et al. 2004). Informa bined and length of gestation, birth weight, or body length tion is lacking on effects of OP insecticides on uterine smooth (Table 2). Among the 91 girls, log-transformed DEP con muscle. However, in mouse uterus, regulation of acetylcholine centration was significantly inversely associated with length levels is dominated by BChE and activity changes in excess of gestation (beta ----1.79 weeks, 95% Cl ----2.82, --0.76; of 50%, which can occur during the estrus cycle, appear to log scale not specified), but no such association was observed be required to cause substantial changes in uterine contractile among boys (beta --0.17 weeks, P --0.164). No other signifi activity (Medina et al. 1993). In light of the scattered positive cant associations were reported among girls or boys only. associations with some but not all indicators of fetal growth The strengths and limitations of the Shanghai birth cohort and the internally inconsistent associations with length of ges study have been described in the context of oilier prospective tation, the overall results do not demonstrate an adverse effect birth cohort studies. The high participation rate (stated as 97% o f prenatal exposure to OP insecticides on birth outcomes. among eligible subjects) is a strength, although the derivation of this rate (i.e., the definition of the eligible source popula New Jersey birth cohort tion) is not clear. In a convenience sample of 150 New Jersey women with a non-anomalous singleton pregnancy scheduled for an elec tive cesarean birth at > 37 weeks of gestation in 2003-2004, chlorpyrifos and other pesticides were measured in preop erative maternal serum and umbilical cord serum (Table 1) (Barr et al. 2010). The mean chlorpyrifos concentration was 0.09 ng/g (SD = 0.87, range = 0.0007-10.09) in maternal serum and 0.55 ng/g (SD = 0.73, range = 0.0007-1.84) in cord serum. After multivariate adjustment, mean birth weight, head circum ference, abdominal circumference, and birth length did not dif fer significantly between newborns with maternal prenatal or cord serum chlorpyrifos concentrations > 75th versus < 75th percentile (0.0007 ng/g), nor did they differ significantly when the cutoff was set at the 90th percentile (Table 2). Strengths and limitations of the New Jersey birth cohort As stated earlier, DAP metabolite levels measured at the time of labor may not reflect earlier exposure levels, which may be more relevant to fetal development. In addition, the limitations of DAP metabolites for OP exposure assessment were discussed previously. The authors noted that DAP metab olite levels observed in this study were substantially higher than those reported among pregnant or postpartum women in the United States (Bradman et al. 2005) and the Netherlands (Ye et al. 2008), yet only one statistically significant associa tion w'as detected. The authors did not hypothesize why levels of DEP, but not other DAP metabolites, might plausibly be related to shorter length of gestation, but not other birth out comes, only among girls. With at least 54 associations tested, chance appears to be a more reasonable explanation for this single statistically significant result. study are comparable to those described above for other pro spective birth cohort studies. Chlorpyrifos levels measured in Health Outcomes and Measures of the Environment Study maternal blood just before cesarean section may not be accu rate indicators of earlier prenatal levels, which are probably more etiologically relevant to fetal growth. Because subjects were recruited by convenience sampling, the participation rate was not stated, and selection bias is a possibility if participa tion was related to factors associated with both chlorpyrifos exposure and birth outcomes (e.g., socioeconomic status, diet, and place of residence). It is unclear whether results for newborns delivered by elective cesarean section would be expected to differ from those for newborns delivered vaginally or by unplanned cesarean section. Finally, the scope of the study with regard to OP insecticides was limited by the mea surement of only chlorpyrifos. As a whole, the results of this study do not demonstrate an association between detectable chlorpyrifos in maternal perinatal serum and birth outcomes. In the Health Outcomes and Measures of the Environment (HOME) Study, based in Cincinnati, Ohio, 389 healthy preg nant women living in a home built before 1978 were enrolled at < 1 9 weeks of gestation and followed through delivery of a live-born singleton infant in 2003-2006 (Table 1) (Rauch et al. 2012). Six DAP metabolites were measured in mater nal spot urine samples collected from 344 participants at approximately 16 and 26 weeks of gestation (averaged for analysis) and within 24 h of delivery. Median concentrations were 81.3 nmol/L (1QR --41.7-220.0) for DAPs, 56.9 nmol/L (IQR - 26-185) for DMPs, and 17.7 nmol/L (IQR - 8-37) for DEPs. In addition, umbilical cord blood was genotyped for the PON]ig, and PONl'_ m polymorphisms. Statistically significant inverse associations were detected in multivariate adjusted models between log ^-transformed, Shanghai birth cohort creatinine-standardized maternal prenatal urinary DAP or DMP concentrations and length of gestation (beta for DAPs -- Among 187 healthy women in .Shanghai with an uncompli --0.5 weeks, 95% Cl ----0.8, --0.1; beta for DMPs ----0.4 cated singleton pregnancy in 2006-2007, five DAP metabo weeks, 95% Cl = --0.7, 0.0) and birth weight (beta lites were measured in maternal spot urine specimens col for D A P s - - 1 5 1 g, 95% C l- - 2 8 7 , - 16; beta for lected at the onset of labor (Table 1) (Wang et al. 2012). DMPs - - 124 g, 95% Cl - - 245, - 2) (Table 2) (Rauch Geometric mean concentrations were 17.19 jig/L (range = < et al. 2012). However, the associations with birth weight LOD-269.15) for DMP, 8.01 ig/L (r a n g e -< L O D -109.65) were substantially attenuated and not statistically signifi for DMTP, 6.03 pg/L (range--< L Q D -109.65) for DEP, cant after adjustment for gestational age. Maternal prenatal 8 IQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00038 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 569 urinary DEP concentrations were not significantly associated pesticides) were measured in umbilical cord serum at delivery, with either outcome. After stratification by race, the inverse including chlorpyrifos, diazinon, fonofos, malathion, parathion, associations of prenatal DAPs and DMPs with length of ges methylparathion,profenofos,andterbufos(Tablel)(Wickerham tation were detected only for white mothers (P . = 0 .1 0 et al. 2012). The proportion of serum samples with detectable and 0.09, respectively), whereas the inverse associations of levels (LOD = 0.05 ng/mL except for malathion andprofenofos, prenatal DAPs and DMPs with birth weight were detected where LOD = 0.50 ng/mL) were 23.3% for chlorpyrifos (90th only for black mothers and only after additionally adjusting percentile = 0.17 ng/mL), 14.7% for diazinon (90th percen for gestational age (P4 = 0.02 and 0.02, respectively). tile = 0.27 ng/mL), 16.4% for fonofos (90th percentile = 0.30 In models stratified by genotype, maternal prenatal urinary ng/mL), 25.9% for malathion (90th percentile = 3.13 ng/mL), DAP concentrations were inversely associated with length of 2.6% for parathion (90th percentile <0.05 ng/mL), 25.0% gestation only among new'borns with the PONI 192 QR or QQ for profenofos (90th percentile = 0.68 ng/mL), and 31.0% (not the low-activity RR) genotype (Piniera,,tion = 0.04 and 0.09, for terbufos (90th percentile = 0.27 ng/mL). After multivari respectively) or the PONI WH CT (not the high-activity CC ate adjustment, no significant associations with birth weight or low-activity XT) genotype (Pimeraction = 0.04). The inverse were detected for any of these pesticides, whether analyzed as association between maternal prenatal urinary DAP concen detectable versus non-detectable, three-level ordinal variables, tration and birth weight was observed only among newborns or the total number detected (Table 2). with the PONI 1Q2 QR genotype (Pimeraciion = 0.02) or the In general, some of the strengths and limitations of the p o n 1-I08 CT genotype = -5)- Models stratified by both race and genotype were not only based on small num Zhejiang birth cohort study are similar to those of other birth cohort studies. Additional strengths include the nearly 100% bers and were therefore statistically unstable, with wide CIs, participation rate among eligible subjects (although the basis but also suggested stronger associations among heterozygotes. for calculating this rate is not clear) and the measurement of Results were modestly attenuated when restricted to full-term specific OP insecticides rather than non-specific DAP metabo births, based on non-creatinine-standardized DAP concentra lites, countered by limitations, including the cross-sectional tions, or based on DAP concentrations from either 16 or 26 measurement of pesticide levels in umbilical cord serum col weeks only. lected at delivery, the analysis of pesticide exposure as simpli Besides the previously noted strengths and limitations of fied categorical variables, and the evaluation of only a single prospective birth cohort studies, this study benefits from the birth outcome (weight). In summary, the results of this study measurement of maternal urinary DAP metabolite levels at two suggest no association between concurrent exposure to any of time points near the beginning and end of the second trimester. eight different OP insecticides and birth outcomes. The 37.1% participation rate among 1263 eligible women, com bined with the 86% follow-up rate through delivery (including twins and stillbirths) and the 88% biospecimen availability rate, raises the possibility of selection bias, wdtli an unknown direction and magnitude of influence. Only DAP metabolites, not specific OP insecticides, were measured, and results were reported only for two birth outcomes. Several anomalous results were reported, including the detection of some associations only among white mothers and others only among black moth ers; the detection of significant inverse associations between maternal prenatal urinary DAP and DMP levels and birth weight only in the absence of adjustment for gestational age, but the detection of these associations among black mothers only with adjustment for gestational age; and the stronger observed associations among PONI 192and PONl_m heterozygotes than among low-activity homozygotes. Overall, the results suggest possible inverse associations between maternal prenatal urinary DAP and DMP (but not DEP) concentrations and length of gestational age and birth weight, perhaps restricted to specific racial groups or those with intermediate (but not low or high) Bradford Hill evaluation of weight of evidence Strength. The strength of observed associations between OP metabolites and birth outcomes cannot be compared easily across studies, given differences in the unit of exposure mea surement, the logarithmic base used for transformation (if any), and the format in which results were presented (e.g., as regres sion betas or adjusted means). The distinction between a weak and a strong association, especially with a continuous outcome such as birth weight or length of gestation, is also subjective and hard to define. Nevertheless, most reported associations involved birth weight differences of < 100 g, birth length and head circumference differences of < 0 .5 cm, and gestational length differences of < 5 days (< 0.7 weeks), in association with exposures classified on various scales (e.g., detectable, natural log, or log,0). In general, weak associations are more likely than strong associations to be explained by confounding, bias, or chance. Furthermore, the majority of reported results were not statistically significantly different from the null value. PONI activity. As in other studies with internally inconsistent associations, these findings may be explained at least in part Consistency. An examination of the consistency of associa by chance or bias, especially given the numerous hypotheses tions with specific birth outcomes reveals mostly null findings, tested, and do not offer convincing support for an adverse effect with no uniformity of positive or inverse associations across of OP insecticides on birth outcomes. (as well as within) studies. In particular, associations with birth weight were inconsistently reported as inverse (Perera 'Zhejiang birth cohort et al. 2003, Rauch et al. 2012, Whyatt et al. 2005, Whyatt et al. 2004), positive (Eskenazi et al. 2004, Harley et al. 2011), In rural Zhejiang Province, 116 consecutive healthy women or in most cases, null (Barr et al. 2010, Berkowitz et al. 2004, with a healthy, uncomplicated, singleton pregnancy at Eskenazi et al, 2004, Harley et al. 2011, Perera et al. 2003, 36 weeks of gestation, eight OP insecticides (and other Wang et al. 2012, Whyatt et al. 2005, Whyatt et al. 2004, Wolff SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00039 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 570 R. Reiss et al. et al. 2007) (Wickerham et al, 2012), Associations with birth length were also heterogeneous, including inverse (Perera et al. 2003, Whyatt et al. 2005, Whyatt et al. 2004), positive (Eskenazi et al, 2004), and mostly null findings (Barr et al. 2010, Rerkowitz et al. 2004, Eskenazi et al. 2004, Perera et al. 2003, Wang et al. 2012, Whyatt et al. 2005, Whyatt et al. 2004, Wolff et al. 2007), Likewise, head circumference was variously inversely associated (Berkowitz et al. 2004, Wolff et al. 2007), positively associated (Eskenazi et al. 2004, Harley et al. 2011), and most often not significantly associated (Barr et al. 2010, Berkowitz et al. 2004, Eskenazi et al. 2004, Harley et al. 2011, Perera et al. 2003, Whyatt et al. 2005, Whyatt et al. 2004, Wolff et al. 2007) with OP metabolite levels. The reported associa tions of individual OP metabolites with ponderal index were statistically null in both studies of this outcome (Eskenazi et al. 2004, Wolff et al. 2007). Although inverse associations between a few selected OP metabolites and length of gestation were reported in multiple studies (Eskenazi et al, 2004, Harley et al. 2011, Rauch et al. 2012, Wang et al. 2012), these associations were inconsistent across participant subgroups by race, sex, and genotype, with no cogent biological explanation for the observed heterogene ity. For example, a signi ficant inverse association with maternal prenatal urinary DAPs was detected among white mothers but not black mothers in the HOME .Study (Rauch et al. 2012), and an inverse association with maternal prenatal urinary DEPs was detected among infant girls but not boys in the Shanghai birth cohort study (Wang et al. 2012). Moreover, most OP metabolites measured in these and other studies (Berkowitz et al. 2004, Wolff et al. 2007) were not significantly associated with length of gestation. Thus, the most consistent findings were statistically null, and the lack of consistency of signifi cant associations between OP metabolites and specific birth outcomes does not support a causal interpretation of the few' statistically significant associations observed. Crit Rev Toxicol. 2015; 45(7): 531-641 tionship, without testing the appropriateness of this model. Only two studies examined exposure-response gradients by categorizing exposures into at least three ordinal groups (Eskenazi et al. 2004, Whyatt et al. 2004), (This observa tion also highlights the problem of inconsistency of analytic approaches among studies.) In the CCCEH study, w'here cord plasma concentrations of chlorpyrifos (and chlorpyrifos plus diazinon, but not diazinon alone) were inversely associated with birth weight and birth length, the strength of the inverse associations increased across tertiles of detectable levels com pared with non-detectable levels, a pattern consistent with a monotonie exposure-response gradient (Whyatt et al. 2004). In the CHAMACOS study, maternal prenatal urinary levels of MDA, TCPy, and PNP, which were categorized as undetect able, detectable below the median, or detectable at or above the median, did not show evidence of a monotonie association with length of gestation, birth weight, birth length, or pon deral index (Eskenazi et al. 2004). The middle category of PNP appeared to be inversely associated with length of gesta tion and ponderal index and positively associated with body length; however, the results for the highest category were not significantly different from the null value. Some evidence of a positive exposure-response trend was observed between PNP and head circumference. Although significant linear regression coefficients may be consistent with a monotonie biological gradient, the dearth of information on the shape of exposureresponse relationships between OP metabolites and birth out comes preven ts a thorough evaluation of such gradients. The commonly applied mechanism for OP toxicity is AChE inhibition. As discussed earlier, the levels in the epidemiologic studies are orders of magnitude below what would result in clinically meaningful AChE inhibition. There are a few other postulated mechanisms for non-cholinergic OP toxicity, but effects at the levels observed in the epidemiologic studies have not been established for these mechanisms either. Temporality. As discussed above, an assessment of the tempo Plausibility. The biological plausibility of the associations ral relationship of measured OP and DAP metabolite levels in is not established. While OP insecticides are known to cause prenatal or perinatal maternal biospecimens in relation to fetal neurotoxicity in mature subjects at doses higher than reported growth and other birth outcomes is limited, for several rea in the epidemiologic studies, the mechanism of OP-induced sons. First, exposures measured soon before birth are unlikely neurodevelopmental toxicity has yet to be established. to have a major influence on fetal growth over the course of 40 weeks o f gestation. Second, because these metabolites have Coherence. In the evaluation of the coherence of evidence, a short biological half-life and vary considerably within indi another important consideration is whether observed interac viduals, one or tw'o samples are unlikely to reflect past or long tions with PON1 activity levels or genotypes are consistent term average exposure for a given person. Third, it is unknown with the hypothesis of increased susceptibility to potential whether the time points selected for blood, urine, or personal adverse health effects of OP insecticides in those with lower air collection in various studies are etiologically relevant or PON1 activity. In the three studies that evaluated these inter whether exposures earlier or later in gestation have a greater actions--the Mount Sinai CECS (Berkowitz et al. 2004, Wolff influence on fetal growth or length of gestation. Consequently, et al. 2007), CHAMACOS (Harley et al. 2 0 1 1), and the HOME the measurement of metabolite levels prior to birth does not Study (Rauch et al. 2012)-- results were variable. One study necessarily strengthen the evidence in favor of a causal inter reported the expected stronger inverse associations, albeit only pretation of observed associations, especially if measurements between selected metabolites and birth outcomes, in those with were taken only hours or minutes before birth, but even if they homozygous low-activity PON1 genotypes or low measured were made months in advance. PON1 activity (Wolff et al, 2007). /Another study found mostly no apparent heterogeneity by PON l genotype, level, or activ Biological gradient. Few studies explicitly evaluated the shape ity, but some evidence of stronger positive associations, again of the biological gradient between OP metabolites and birth between only selected metabolites and birth outcomes, in outcomes; instead, using linear regression models, nearly all those with heterozygous or homozygous high-activity PON 1 investigators assumed a log-linear exposure--outcome rela- genotypes or higher PON1 levels, and an inverse association SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00040 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 571 between DEPs and gestational age among those homozygous Diagnostic and Statistical Manual o f Mental Disorders, 4th for the low-activity PONl_108 genotype (Harley et ah 2011). Edition (DSM-IV), to assess recent behavioral problems at age In another study, stronger inverse associations between DAPs 36 months; the Child Behavior Checklist for ages 6-18 years to and birth outcomes were observed among PON1 heterozy assess recent behavioral problems at age 7 years; the Wechsler gotes than among low- or high-activity homozygotes (Rauch Intelligence Scale for Children, 4th Edition, including the Ver et al. 2012). Finally, one study found no evident heterogeneity- bal Comprehension Index, the Perceptual Reasoning Index, in associations by PON1 activity (Berkowitz et al. 2004). As a the Working Memory Index, and the Processing Speed Index, whole, these mixed results are not coherent with a protective which were combined to determine the Full-Scale Intelligence effect of high PON 1 detoxifying activity against adverse effects Quotient (IQ), at age 7 years; and magnetic resonance imaging of OP insecticides on fetal growth and other birth outcomes. for brain morphology at ages 5.9-11.2 years (Table 1) (Horton et al. 2012, Lovasi et al. 2011, Rauh et al. 2011, Rauh et al. Specificity, experiment, and analogy. The other Bradford Hill 2006, Rauh et al. 2012). guidelines-- specificity, experiment, and analogy-- are less After multivariate adjustment, significant or borderline informative for the evaluation of causality. Especially in light significant inverse associations were observed between the of the non-specificity of DAP metabolites, the many influences highest detectable tertile (> 6 .1 7 pg/g) versus lower levels on birth outcomes, and the numerous associations tested, no of cord plasma chlorpyrilbs and the Bayley Mental Devel specific relationship has emerged between any particular OP opment Index (beta ----3.327, SE --1.76, P --0.06) and the insecticide and any particular birth outcome. Relevant quasi- Psychomotor Development Index (beta ----6.46, SE = 2.18, experimental evidence in humans, such as a study of birth P --0.003) at age 36 months (Table 2) (Rauh et al. 2006). outcomes in women who adhere to an organic diet, is unavail The inverse association with the Mental Development Index able. Drawing analogies with other prenatal exposures that at 36 months was observed only among African American cause adverse birth outcomes (e.g., ethanol, methylmercury, children (beta = --6.34), and not among Dominican children certain prescription medications, and dietary factors) is not (beta = --1.70), whereas the inverse association with the Psy warranted, just like existence of numerous exposures shown to chomotor Development Index at 36 months was observed in be safe cannot be used to refute a causal association between both groups (beta = --7.15 and --5.18, respectively). No inter OP insecticides and adverse birth outcomes. On balance, such actions were detected with other covariates tested. When the analogies do not sway the evaluation of causality. Bayley indices were dichotomized at 85 points (one SD below the mean) to indicate developmental delay, cord plasma chlo- Neurodevelopmental outcomes rpyrifos levels in the highest detectable tertile were associated Twenty studies in ten study populations have examined associ ations between OP or OP metabolites and neurodevelopmental outcomes (Bouchard ei al. 2010, Bouchard et ai. 2 0 1 1, Engel et al. 2007, Engel et al. 2011, Eskenazi et al. 2010, Eskenazi et al. 2007, Fortenberry et al. 2014, Guodong et al, 2012, Horton et al. 2012, Lizardi et ai, 2008, Lovasi et al. 2011, Marks et al. 2010, Oulhote and Bouchard 2013, Quiros-Alcala et al. 2011, Rauh et al. 2011, Rauh et al. 2006, Rauh et al. 2012, Yolton et al. 2013, Young et al. 2005, Zhang et al. 2014). Most studies were conducted in birth cohorts enrolled prior to delivery-- including four cohorts described earlier in the section on birth outcomes---whereas other studies were crosssectional in design. Measures of neurodevelopment varied among studies, with several using standard clinical scales or questionnaires, and others using measurement tools that were not used by any other studies reviewed, although all studies reported some degree of validation of the assessment tools used. Table 3 summarizes the analyses in tire studies evaluat ing neurodevelopmental outcomes. with a significantly increased odds of mental delay (odds ratio [OR] = 2.37, 95% C l -1 .0 8 , 5.19) and psychomotor delay (OR - 4.52, 95% Cl = 1.61, 12.70) at 36 months. However, cord plasma chlorpyrifos was not significantly associated with the Mental Development Index at 12 or 24 months (beta at 12 months ----0.344, SE --1.66; beta at24 months ----1.480, SE --2.03) or with the Psychomotor Development Index at either time point (beta at 12 months ----3.30, SE --2.11; beta at 24 months = 1.17, SE --1.98), nor were significant associa tions detected with mental or psychomotor delay at those ages. At 36 months, significant associations were detected between elevated chlorpyrifos levels and Child Behavior Checklist measures of attention problems (OR --11.26, 95% Cl --1.79, 70.99), attention deficit/hyperactivity disorder (ADHD; OR --6.50, 95% C l --1.09, 38.69), and pervasive devel opmental disorder (O R -5 .3 9 , 95% C l - 1.21, 24.11), but not externalizing behavior problems (unadjusted P --0.426) or internalizing behavior problems (unadjusted P --0.444). A subsequent analysis of neighborhood characteristics based on U.S. census data for poverty, education, race, language, and Columbia Centerfor Children \s Environment and Health housing showed no substantial confounding ( < 10% change in beta) or modification (P > 0.20) of the associations between in the CCCEH birth cohort study, which was described earlier cord plasma chlorpyrifos and the Bayley Mental Develop with respect to birth outcomes, childhood neurodevelopmen ment and Psychomotor Indices at 36 months (Table 2) (Lovasi tal outcomes were measured using the Bayley Scales of Infant et al. 2011). Development, 2nd Edition, including the Mental Development Based on results of the Wechsler Intelligence Scale test Index and the Psychomotor Development Index, to assess ing at age 7 years, with outcomes analyzed on the natural cognitive and psychomotor development at ages 12, 24, and fog scale, no significant adjusted associations were detected 36 months; the mother-reported Child Behavior Checklist between cord plasma chlorpyrifos levels and Wechsler Full- for ages 1.5-5 years, including syndrome scale scores, inter Scale IQ, Verbal Comprehension, Perceptual Reasoning, or nalizing and externalizing scores, and scales oriented to the Processing Speed, nor were any significant interactions with SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00041 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. Results of epidemiologic studies of organophosphorus insecticide biomarkers and neurodevelopmental outcomes. U-i w Number of Reference Rauh et al. (2006) Outcome Bayley Mental Development Exposure Cord plasma eiilorpyrifos subjects/events 254 total Estimate of association (9o% Cl) Adjustment factors Comments >5 Beta SE at 12 Prenatal environmental tobacco Cutoff for cord plasma chlorpyrifos was 5wa months ---0.344 A 1.66, P - 0.836 smoke, race/ethnicity, infant gender, set at the highest fertile of detectable ZOc3 Index al 12, 24, or 36 months > 6.17 vs. < 6.1 7 pg/g 229 at 12 months Beta SE at 24 maternal intelligence quotient by Test levels because "the only group fir months ---1.480 2.03, P = 0.466 of Nonverbal Intelligence (Second which mean 36-month [Bayley Scales & 225 at 24 months Beta SE at 36 Edition), maternal education, and of Infant Development 11] scores were months ---3.327 1.76, P = 0.060 Home Observation for Measurement significantly lower was the group with 228 at 36 months Beta at 36 months, African of the Environment core the highest exposure level (> 6.17 Americans - -6.34, P < 0.05 Beta at 36 months, Dominicans ---1.70, P > 0.05 `Interaction terms for the interaction of chlorpyrifos exposure wdth the other exposure and sociodemographic variables were tested in the full model, and none was significant." Generalized linear models showed no significant within-subject association with chlorpyrifos over age groups (P = 0.23) Pg/g)" Bayley scores are standardized to a mean A SD of 100 A 15, with scores < 8 5 indicating developmental delay (minimum score --50, maximum score = 150) "When administered at 3 years of age, the [Bayley Scales of Infant Development II] demonstrates only moderate predictive power for subsequent intelligence and school performance but is clinically useful for children performing in the subnormal range." Rauh et al. Bayley Psychomotor Beta SE at 12 months -- " (2006) Development - 3 .3 0 A 2 .il, P = 0.12 Index al 12, 24, or Beta A SE at 24 36 months months --1.17 A 1.98, P --0.56 Beta A SE at 36 months - -6.46 A 2.18, P --0.003 Beta at 36 months, African Americans = -7 .1 5 ,P < 0.05 Beta at 36 months, Dominicans ---5.18, P <0.05 "All interaction terms for the interaction of chlorpyrifos exposure with the other exposure and sociodemographic variables were tested in the full model, and none was significant." Generalized linear models showed a significant within-subject Q association with chlorpyrifos over s age groups (P = 0.01), with a difference emerging between 24 m and 36 months (p --0.003) m i ; : ED 002061 00046431-00042 DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Rauh et al. (2006) Rauh et al. (2006) Rauh et al. (2006) Rauh et al. (2006) Rauh et al. (2006) Rauh et al. (2006) Rauh et al. (2006) Bayley mild/ significant mental delay at 12, 24, or 36 months Bayley mild/ significant psychomotor delay at 12, 24, or 36 months Child Behavior Checklist attention problems at 36 months Child Behavior Checklist ADHD problems at 36 months Child Behavior Checklist pervasive developmental disorder problems at 36 months Child Behavior Checklist externalizing behavior problems at 36 months Child Behavior Checklist internalizing behavior problems at 36 months " " " 228 " " " " " Odds ratio at 12 months = 1.22 (0.48, 3.06) Odds ratio at 24 months = 1.75 (0.86, 3.60) Odds ratio at 36 months - 2.37 (1,08, 5.19) Odds ratio at 12 months - 1.88 (0,78, 4.53) Odds ratio at 24 months = 1.01 (0.37, 2.76) Odds ratio at 36 months = 4.52 (1.61, 12.70) Odds ratio = 11.26 (1.79,70.99) Odds ratio = 6.50 (1.09, 38.69) Odds ratio = 5.39 (1.21, 24.11) % > 6 . 1 7 - 10.6% % < 6.1 7 = 8.6% P = 0.426 % > 6 .1 7 = 14.9% % <6.17=13.0% P = 0.444 Child Behavior Checklist collects information on behaviors occurring ir the past 2 months, with the cutoff for borderline or clinical problems set at 98th percentile None (Continued) Tier 3/4 L(}W~lVCl OP USECtide Exposures 573 ED 002061 00046431-00043 m m i ; m Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference larvasi et al. ( 2011) Outcome Bayley Mental Development Index at 36 months Exposure Cord plasma chlorpyrifos > 6.17 vs. < 6,1 7 pg/g Number of subjects/events 266 larvasi et al. ( 2011) Rauh et al. (2011) Bayley Psychomotor Development Index at 36 months Wechsler full-scale intelligence quotient at 7 years, natural log scale Cord plasma 265 chlorpyrifos (pg/g) Rauh et ai. (2011) m m i m Wechsler verbal comprehension at 7 years, natural log scale Ui Estimate of association (95% Cl) Model 1 beta = -3 .2 (-5.1, -1.3) Model 2 beta - -3.4 (-5.2, -1.5) Adjustment factors Infant gender, gestational age, Dominican ethnicity, maternal Comments >2 >3 Residential neighborhoods characterized by mothers' self-report Z03c Model 3 beta - -3.2 (-5.0, -1.5) Model 4 beta ---3.1 (-4.8, -1.3) education, maternal intelligence quotient, prenatal environmental and IJ.S, Census data within geocoded network buffers & Model 5 beta - -3.0 (-4.8, -1.2) Model 6 beta = -3.2 (-5.1, -1.3) tobacco smoke exposure, and index of Neighborhood poverty did not building disrepair plus: significantly modify the association Model 1: none additional of chlorpyrifos exposure with Bayiey Model 2: neighborhood % poverty at Mental Development Index (P --0.2) % high school graduates Model 3: neighborhood % African American Model 4: neighborhood % linguistic isolation Model 5: neighborhood % crowded household Model 6: neighborhood inadequate plumbing and % vacant housing Model 1 beta = -6 .9 (-11.1, -2.7) Neighborhood poverty did not Model 2 beta - -7.0 (-11.0, -2.9) significantly modify the association Model 3 beta - -7.3 (-11.5, -3.0) of chlorpyrifos exposure with Bayiey Model 4 beta = -7.2 (-11.3, -3.0) Psychomotor Development Index Model 5 beta = -6.9 (-11.1, -2.8) (P'= 0.4) Model 6 beta ---7.1 (-11.4, -2.7) Parsimonious model beta ---0.003 Parsimonious model (least absolute Full-scale intelligence quotient is the (-0.006, 0.001) shrinkage and selection operator): sum of four composite indices; Fully adjusted beta = -0.003 (-0.006, maternal education, maternal mean SD = 100 15 0.000) Change per SD (4.61 pg/g) increase in exposure = -1.4% "No significant interactions" between chlorpyrifos and any covariates intelligence quotient, and Home Observation for Measurement of the Environment score Fully adjusted model: child sex, race/ethnicity, maternal intelligence P --0.08 for smoothed cubic spline model vs. linear model for chlorpyrifos (unadjusted) quotient, maternal education, household income, child age at testing, prenatal environmental tobacco smoke exposure, and prenatal polycyclic aromatic hydrocarbons exposure Q Parsimonious model beta = none; Verbal comprehension index measures s chlorpyrifos dropped from, model verbal concept formation and Fully adjusted beta ---0.002 (-0.005, predicts school readiness; mean 0.001) SD --100 15 "No significant interactions" between chlorpyrifos and any covariates P = 0.07 for smoothed cubic spline model vs. linear model for chlorpyrifos (unadjusted) Tier 3/4 ED 002061 00046431-00044 V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Dot Low-level op insedile exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Rauh et al. (2011) Rauh et al. (2011) Wechsler perceptual reasoning at 7 years, natural log scale Wechsler processing speed at 7 years, natural log scale Rauh et al. (2011) Wechsler working memory at 7 years, natural log scale Horton et al. (2012) Wechsler working Cord plasma 335 memory at 7 years chlorpyrifos ipg/g, natural log scale) m m i ; m last ^4 Parsimonious model beta--none; chlorpyrifos dropped from model Fully adjusted beta = -0.002 (-0.006, 0.002) "No significant interactions" between chlorpyrifos and any covariates Parsimonious model beta = none; chlorpyrifos dropped from model Fully adjusted beta --0.001 (-0.004, 0.005) "No significant interactions" between chlorpyrifos and any covariates Parsimonious model beta = -0.006 (-0.009, -0.002) Fully adjusted beta = -0.006 (-0.010, -.002) Change per SD (4.61 pg/g) increase in exposure ---2.8% "No significant interactions" between chlorpyrifos and any covariates Model 0 beta, males = - 2.382 ( - 3.88, --0.88) Model 0 beta, females - - 0.524 ( - 1.90, 0.85) Model 1 beta = -1.451 (-2 .265, - 0.438) Model 2 beta - - 1.355 ( - 2.368, - 0.341) Model 3 beta = -1 .4 7 8 (-2 .496, - 0.459) No change after additional adjustment for log-transformed Wechsler General Ability Index or Child Behavior Checklist problem scales or exclusion of imputed chlorpyrifos values based on maternal prenatal plasma levels Model 0: None Family income, maternal education, and child sex plus: Model 1: total Home Observation for Measurement of the Environment score Model 2: parental nurturance composite scale of the Home Observation for Measurement of the Environment score Model 3: environmental stimulation composite scale of the Horne Observation for Measurement of the Environment score Perceptual reasoning index measures nonverbal and fluid reasoning; mean SD --100 15 P --0.08 for smoothed cubic spline model vs. linear model for chlorpyrifos (unadjusted) Processing speed index assesses ability to focus attention and quickly scan, discriminate, and sequentially order visual information; mean SD - 1001:15 P = 0.59 for smoothed cubic spline model vs, linear model for chlorpyrifos (unadjusted) Working memory index assesses ability to memorize new information, hold it in short-term memory, concentrate, and manipulate information; mean SD --100 15 P --0.40 for smoothed cubic spline model vs. linear model for chlorpyrifos (unadjusted) Home Observation for Measurement of the Environment score based on evaluation of child home environment at age 3 years Parental nurturance: sum of z-scores of responsivity, modeling, and acceptance subscales, which measure such maternal behaviors as attentiveness, displays of physical affection, encouragement of delayed gratification, limit setting, and the ability of the mother to control her negative reactions Environmental stimulation: sum of z-scores of learning materials, language stimulation, academic stimulation, and variety subscales, which measure the availability of intellectually stimulating materials in the home and the mother's encouragement of learning (Continued) Low-level OP insectide exposures 575 ED 002061 00046431-00045 576 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Horton et al. ( 2012) Outcome " Rauh et al. (2012) Rauh et al. (2012) Overall brain size (cm3) at 5.9-11.2 years Morphology of cerebral surface (enlargement) at 5.9-11.2 years Exposure Cord plasma chiorpyrifos (pg/g, natural log scale) with interaction terms Number of subjects/events " Cord plasma chiorpyrifos > 4.39 vs. < 4.3 9 pg/g 20 4.39 pg/g 20 < 4.39 pg/g Estimate of association (95% Cl) Model 2a beta = -0.553 ( - 1.943, 0.836) Model 2a interaction beta = --1,714 (--3.753, 0.326) for chiorpyrifos X child sex Model 2b beta = - 1.248 ( - 2.270, 0.227) Model 2c beta = - 1.354 (-2 .3 6 9 , - 0.339) Model 2c interaction beta = 0.024 (--0.690, 0.738) for chiorpyrifos X parental nurturance Mean SD = 1,265.1 17.7 Mean SD = 1,242.1 16.8 P = 0.37 Adjustment factors Family income, maternal education, child sex, and parental nurturance composite score plus: Model 2a: chiorpyrifos X child sex interaction Model 2b: parental nurturance X child sex interaction Mode! 2c: chiorpyrifos X parental nurturance interaction Age, sex, and height Significant enlargement, especiallly of white matter, of superior temporal, posterior middle temporal, and inferior postcentral gyri bilaterally; supramarginal gyrus and inferior pariental lobule of right hemisphere; supramarginal gyrus and inferior parietal lobule of right hemisphere; and superior frontal gyrus, gyrus rectus, cuneus, and precuneus along mesial wall of right hemisphere in those with < 4.39 vs. < 4.39 pg/g Significant positive dose-response relationship between chiorpyrifos and enlargement of mesial surface of superior frontal gyrus bilaterally among those with > 4 .3 9 pg/g Age and sex, with or without overall brain size Comments - Cutoff for cord plasma chiorpyrifos was set at the highest tertile (4.39 pg/g) P-values were corrected for multiple comparisons using a false discovery rate P < 0.05 Tier 3/4 CritRevTcocoL 2015; 45(7): 531-641 ED 002061 00046431-00046 m m i ; m Tat ^4 ED 002061 00046431-00047 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Rauh et al. ('2012) Rauh et al. (2012) Rauh et al. (2012) Rauh et al. (2012) m i m Cord plasma ehlorpyrifos te 4.39 vs. < 4.39 pg/g with interaction ternis with full-scale intelligence quotient Cord plasma ehlorpyrifos te 4.39 vs. < 4.39 pg/g with interaction terms with sex 19 St4.39 pg/g 18 < 4.39 pg/g Morphology of cerebral surface (deformation) at 5.9-11.2 years Cortical thickness at 5.9-11.2 years Cord plasma ehlorpyrifos - 4.39 vs. < 4.3 9 pg/g " 20 < 4.39 pg/g 20 < 4.39 pg/g " Significant interaction between chlorpyrilbs and intelligence quotient on surface measures in superior temporal, inferior frontal, inferior precentral, and inferior postcentral gyri bilaterally, and precuneus of left hemisphere, with positive correlation with intelligence quotient among those with <4.39 pg/g but no correlation among those with --4.39 pg/g Significant interaction between ehlorpyrifos and intelligence quotient on surface measures in right fusiform gyrus, with inverse correlation w'ith intelligence quotient among those with <4.39 pg/g but positive correlation among those with > 4.39 pg/g Significant interaction between ehlorpyrifos and sex on surface measures in right inferior parietal lobule, right superior marginal gyrus, and right mesial superior frontal gyrus, "reflecting disruption of normal, female-larger-thanmale sex differences in the right parietal lobe and a reversal of normal, male-larger-than-female differences in the right mesial superior frontal gyrus" ehlorpyrifos and sex on surface measures in right dorsal parietal lobe, with positive correlation with ehlorpyrifos in girls but inverse correlation in bovs Inward deformations in dorsal and mesial surfaces of left superior frontal gyrus in group with > 4.39 pg/g "Scattered reductions" in cortical thickness in dorsal parietal and frontal cortices in group with > 4.3 9 vs. < 4.3 9 pg/g Inverse dose-response relationship between ehlorpyrifos and cortical thickness in frontal pole, dorsal parietal, and orbitofrontal cortices in those with > 4.3 9 pg/g Age and sex (Continued) Low-level OP insectide exposures 577 doi Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Engel et ai. (2007) Engel et ai. (2007) Outcome Brazelton habituation cluster before hospital discharge Brazelton orientation cluster before hospital discharge Exposure Maternal prenatal urinary DAPs (nmoi/L, log,,, scale) or MDA (detectable vs. nondetectabie) Engel et al. (2007) Brazelton motor cluster before hospital discharge Number of subjects/events 144 with DAPs 153 with DMPs 144 with DEPs 148 with MDA 233 with DAPs 244 with DMPs 233 with DEPs 240 with MDA 249 wdth DAPs 260 with DMPs 249 with DEPs 257 with MDA Engel et al. (2007) Engel et al. (2007) Engel et al. (2007) Engel et al. (2007) m m i m Brazelton range of state cluster before hospital discharge Brazelton regulation of state cluster before hospital discharge 253 with DAPs 264 with DMPs 253 with DEPs 256 with MDA 253 with DAPs 264 with DMPs 253 with DEPs 256 with MDA Brazelton autonomic stability cluster before hospital discharge 253 with DAPs 264 with DMPs 253 with DEPs 256 with MDA Brazelton number of abnormal reflexes before hospital discharge Maternal prenatal urinary DAPs (nmoi/L, log.,, scale or by quartile) or MDA (detectable vs. nondetectabie) 239 with DAPs 250 with DMPs 239 with DEPs 242 with MDA DAPs quartile 2 DAPs quartile 3 DAPs quartile 4 DMPs quartile 2 DMPs quartile 3 DMPs quartile 4 DEPs quartile 2 DEPs quartile 3 DEPs quartile 4 Estimate of association (95% Cl) Adjustment factors Comments >3 Beta - 0.168 (-0.230, 0.566) Beta - -0.024 (-0.335, 0.288) Drug use during pregnancy, examiner, Habituation --ability to respond to and PON 1 enzyme activity, and urinary inhibit discrete stimuli while asleep >W3 Z03c Beta - 0.08 (-0.300, 0.460) Beta = 0.440 (-0.145, 1.025) creatinine; all models other than for No significant associations with DAPs, DEPs also adjusted for overdispersion DMPs, or DEPs categorized by quartile & Beta - -0.106 (-0.414, 0.201) Beta - 0.018 (-0.249, 0.285) Beta ---0.028 (-0.336, 0.279) Beta - -0.100 (-0.597, 0.405) Beta - 0.049 (-0.077, 0.174) Beta = 0.039 (-0.068, 0.146) Beta - 0.048 (-0.078, 0.174) Beta - -0.050 (-0.23.3, 0.156) Beta - 0.0.35 (-0.120, 0.189) Beta - 0.035 (-0.096, 0.167) Beta = 0.015 (-0.140, 0.169) Beta = -0.040 (-0.281, 0.199) Beta = -0.047 (-0.300, 0.207) Beta = -0.072 (-0.283, 0.138) Beta - -0.026 (-0.279, 0.227) Beta - -0.090 (-0.480, 0.303) Beta - -0.154 (-0.382, 0.075) Beta - 0.000 (-0.192, 0.193) Beta = -0.106 (-0.334, 0.122) Beta = 0.090 (-0.274, 0.463) Relative risk --1.49 (1.12, 1.98) Relative risk - 1.1.3 (0,90, 1.41) Relative risk - 1.32 (0.99, 1.77) Relative risk - 2.24 (1.55, 3.24) Relative r i s k - 1.91 (1.12. 3.28) Relative risk --1.22 (0.70, 2.11) Relative risk --1.58 (0.96, 2.58) Relative risk --1.58 (0.94, 2.65) Relative risk - 1.46 (0.83, 2.54) Relative risk - 1.62 (0.98. 2.66) Relative risk --1.29 (0.71, 2.33) Relative risk = 2.59 (1.54, 4.35) Relative risk --1.53 (0.88, 2.66) Pre-pregnancy body mass index, Orientation --attention to visual and examiner, neonatal jaundice, auditory stimuli and quality of overall PON 1 enzyme activity, and urinary alertness creatinine; all models other than for No significant associations with DAPs, DEPs also adjusted for overdispersion DMPs, or DEPs categorized by quartile Infant age at examination, caffeine M otor--motor performance and consumption during pregnancy, drug equality of movement and tone use during pregnancy, examiner, PON! enzyme activity, and urinary creatinine; all models other than for No significant associations with DAPs, DMPs, or DEPs categorized by quartile DEPs also adjusted for overdispersion Infant age at examination, examiner, Range of state --measure of infant PON 1 enzyme activity, and urinary arousal and state lability creatinine; all models other than for No significant associations with DAPs, DEPs also adjusted for overdispersion DMPs, or DEPs categorized by quartile Maternal education, examiner, Regulation of state = ability to regulate PON1 enzyme activity, and urinary state in the face of increasing levels of creatinine; all models other than for stimulation DEPs also adjusted for overdispersion No significant associations with DAPs, DMPs, or DEPs categorized by quartile Infant age at examination, examiner, Autonomic stability = signs of stress smoking during pregnancy, PON 1 related to homeostatic adjustments of enzyme activity, and urinary the central nervous system creatinine; all models other than for No significant associations with DAPs, DEPs also adjusted for overdispersion DMPs, or DEPs categorized by quartile Examiner, anesthesia during delivery, In exploratory analyses of specific PONi enzyme activity, and urinary abnormal reflexes, detectable MDA creatinine; all models other than for levels were significantly associated with DEPs also adjusted for overdispersion abnormal "crawling" and "resist arms" reflexes, and higher DEP levels were associated with an abnormal "crawling" reflex Q s Tier 3/4 ED 002061 00046431-00048 V ED 002061 00046431-00049 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Engel et al. (2007) Engel et al. (2007) m m i ; m Brazelton 2 abnormal reflexes before hospital discharge Maternal prenatal urinary DAPs (nmol/L, log,0 scale) or MDA (detectable vs. nondetectable) 120 with DAPs at age 1 day llS w ith DAPs at age 2 + days 126 with DMPs at age 1 day 123~with DMPs at age 2 + days 129 with DEPs at age 1 day 118 with DEPs at age 2 -1- days 120 with MDA at age 1 day 12fw ith MDA at age 2 -1- days Maternal prenatal NR urinary DAPs (nmol/L, log,0 scale) by maternal PONT expression level Relative r i s k - 1.15 (0.80, 1.63) Relative risk - 1.69 (1.11, 2.59) P-interaction > 0.10 by age Relative risk - 1.00 (0.75, 1.32) Examiner, anesthesia during delivery, PONI enzyme activity, and urinary crevaatinine; all models other than for DEPs also adjusted for overdispersion Relative risk - 1.44 (1.02, 2.03) P-interaction < 0.10 by age Relative risk - 1.39 (0.96, 2.01) Relative risk - 1.60 (0.98, 2.60) P-interaction > 0.10 by age Relative risk - 2.51 (1.61,3.90) Relative risk - 1.34 (0.72, 2.49) P-interaction rs 0.10 by age DAPs, low PON1: relative risk --2.38 (1.37,4.15) DAPs, medium PON 1: relative r i s k - 1.75 (0.96, 3.17) DAPs, high PON 1: relative risk = 0.76 (0.48, 1.20) P-interaction of low and medium vs. high PON 1 - < 0.05 and > 0.05 DMPs, low PON1: relative r i s k - 1.96 (1.27, 3.03) DMPs, medium PON1: relative risk - 1.66 (1.03, 2.65) DMPs, high PON 1: relative risk --0.73 (0.56, 0.96) P-interaction of low and medium vs. high PON1 = 0.002 and 0.001 DEPs, low PON 1: relative risk --1.78 (1.01, 3.14) DEPs, medium P O N i: relative r i s k - 1.42 (0.85. 2.35) DEPs, high PONI: relative risk --T.56 (1.01,2.39) P-interaction a 0.05 Examiner, anesthesia during delivery, and urinary creatinine; all models other than for DEPs also adjusted for overdispersion s* (Continued) Low-level OP insectide exposures 579 doi Low-level op insectide exposures Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 580 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. (Continued ) Reference Engel et ai. (2011) Outcome Bayley Mental Development Index at 12 months Engel et al. (2011) m m i ; m Exposure Maternal prenatal urinary DAPs (mnol/L, log,,, Scale) Maternal prenatal urinary DMPs (mnol/L, log,0 scale) Number of subjects/events Estimate of association (95% Cl) 149 total 'fertile 1 adj. mean, total --97.0 111 blacks/Elispanics (9.3.7, 100.3) 38 whites fertile 2 adj. mean, to tal--95.8 (92.5, 99T) Tertile 3 adj. mean, total = 96.1 (93.1,99.0) Beta, total - -1.00 (-3.28, 1.28) Tertile 1 adj. mean, blacks/ Hispanic's = 96.2 (92.9, 99.4) Tertile 2 adj. mean, blacks/ Hispanics = 94.4 (91.2, 97.5) Tertile 3 adj. mean, blacks/ Hispanics = 91.5 (88.3, 94.7) Beta, blacks/Hispanics - -3.29 (-5.88, -0.70) Tertile 1 adj. mean, whites = 92.0 (85.4, 98/7) Tertile 2 adj. mean, whites --95.9 (90.6, 101.3) Tertile 3 adj. mean, whites = 103.7 (98.5, 108.8) P-interaction by race < 0.001 Beta, w hites-4 ,7 7 (0.69, 8.86) P-interaction by race - 0.001 149 total Tertile 1 adj. mean, total --96.8 111 blacks/ (93.5--100.0) Hispanics Tertile 2 adj. mean, total --96.1 38 whites (92,9-99.3) Tertile 3 adj. mean, total - 96.1 (93.4-99.0) Beta, total = -1.12 (-3.14-0.89) Tertile 1 adj. mean, blacks/ Hispanics = 96.3 (93.0-99.5) Tertile 2 adj. mean, blacks/ Hispanics = 94.2 (91.0-97.4) Tertile 3 adj. mean, blacks/ Hispanics = 92.1 (89.0-95.2) Beta, blacks/Hispanics ---3.35 (-5.64 to -1.06) Tertile 1 adj. mean, whites 92.2 (85.6-98/7) Tertile 2 adj. mean, whites = 97.2 (91.1-102.6) Tertile 3 adj. mean, whites - 103.3 (97.9-108.7) P-interaction by race < 0.01 Beta, whites --4.45 (0.82--8.08) P-interaction by race < 0.001 Adjustment factors Maternal age at enrollment, child sex, examiner, maternal PON1 enzymeactivity, season of urine collection, laboratory batch. Home Observation lor Measurement of the Environment score, alcohol consumption during pregnancy, urinary creatinine, and race/ethnicity (if not stratified; also adjusted for biomarker X race/ ethnicity if stratified) Comments Mental Development Index rates cognitive ability in areas including memory, habituation, problem-solving, early number concepts, generalization, classification, vocalizations, language, and social skills; age-standardized to mean of 100 and SD of 15 Distinct patterns by race/ethnicity at 12 months were also observed by public vs. private housing (results NR) No significant interactions (P > 0.20) were detected between metabolite levels and PON1 L55M or -108C > T polymorphisms or with PON 1 enzyme activity on neurodevelopment at any age (data not shown) CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00050 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. doi Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Engel et al. (2011) Maternal prenatal urinary DEPs (nmoi/L, log] 0 scale) 149 total 111 blacks/ Hispanics 38 whites Engel et ai. (2011) Bayley Mental Development Index at 24 months Maternal prenatal 208 urinary DAPs (nmol/L, log,,, Scale) Engel et ai. (2011) Engel et al. (2011) m m i m Maternal prenatal 208 urinary DMPs (nmol/L, !og,0 scale) Maternal prenatal 208 urinary DEPs (nmol/L, log10 scale) Tettile 1 adj. mean, total = 95.9 (92.9. 98.9) Tertile 2 adj. mean, total = 95.4 (92.3, 98i>) Tertile 3 adj. mean, total = 97.5 (94.3, 100.6) Beta, total = 0.03 (-2.23, 2.29) Tertile 1 adj. mean, blacks/ Hispanics = 94,3 (90.9, 97.6) Tertile 2 adj. mean, blacks/ Hispanics = 93.8 (90.4, 97.1) Tertile 3 adj. mean, blacks/ Hispanics = 95.2 (91.9, 98.6) Beta, blacks/Hispanics = -0 .3 3 (-3.00, 2.35) Tertile 1 adj. mean, whites = 97.3 (91.8, 102.7) Tertile 2 adj. mean, whites = 96.8 (90.8, 102.9) Tertile 3 adj. mean, whites = 100.6 (94.6, 106.5) P-interaction by race = 0.82 Beta, whites = 0.86 (-3.16, 4.87) P-interaction by race = 0.62 Tertile 1 adj. mean = 9 3.6 (89.1, 98.0) Tertile 2 adj. mean = 90.8 (86.3, 95.3) Tertile 3 adj. mean = 90.3 (85.9, 94.7) Beta = -2.08 (--4.60, 0.44) Tertile 1 adj. mean = 92.5 (88.0, 96.9) Tertile 2 adj. mean = 92.9 (88.6, 97.1) Tertile 3 adj. mean = 91.1 (86.9, 95.3) Beta = -0.93 (-3.11, 1.25) Tertile 1 adj. mean = 92.6 (88.2, 97.0) Tertile 2 adj. mean = 90.5 (86.1, 94.9) Tertile 3 adj. mean = 91.2 (86.6, 95.7) Beta = -1.47 (-3.99, 1.04) Maternal age at enrollment, child sex, examiner, maternal education, maternal PON 1 enzyme activity, season of urine collection, laboratory batch, Home Observation for Measurement of the Environment score, alcohol consumption during pregnancy, urinary creatinine, and race/ethnicity Results were not heterogeneous by exact age at 24-month testing (results NR) (Continued) V Low-level OP insectide exposures 581 ED 002061 00046431-00051 582 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Engel et al. (2011) Outcome Bayley Mental Development Index at 12 or 24 months Engel et al. (2011) Engel et al. (2011) m m i m Exposure Maternal prenatal urinary DAPs (nmoi/L, log10 scale) by maternal P O N l192 genotype Number of subjects/events 28 blacks/Hispanics with PON1 m QQ, 12 months 82 blacks/Hispanics with PONljp, QR/RR, 12 months 57 all races with PON1192 QQ, 24 months 140 all races with PO N 1 ,(/, QR/RR, 24 months Maternal prenatal urinary DMPs (nmol/L, iog,0 scale) by maternal PON1192 genotype 28 blacks/Hispanics with P O N l19? QQ, 12 months 82 blacks/Hispanics with PO N 119, QR/RR, !2 ` ` months 57 all races with P O M I9, QQ, 24 months 140 all races with P O N l,92 QR/RR, 24 months Maternal prenatal urinary DEPs (nmoi/L, log,0 scale) by maternal P O N l192 genotype 28 blacks/Hispanics with P O N ] 192 QQ, 12 months 82 blacks/Hispanics with P O N l19~ QR/RR, 12 months 57 all races with PON1192 QQ, 24 months 140 all races with P O N lm QR/RR, 24 months Estimate of association (95% Cl) Beta = 5.72 (-0.48, 11.92) Beta = -4.94 (-7.81,-2.07) P-interaction by genotype < 0.0! Beta = -1.04 (-6.06, 3.99) Beta = -1.27 (-4.40, 1.84) P-interaction by genotype - 0.93 Beta = 3.69 (-0.97, 8.36) Beta = -1.95 (-5.36, 1.47) P-interaction by genotype - 0.06 Beta = -0.55 (4 .7 9 ,3 .7 0 ) Beta = -0.15 (-3.51,3.21) P-interaction by genotype = 0,88 Beta = 2.76 (-2.44, 7.97) Beta = -4.47 (-7.05,-1.89) P-interaction by genotype = 0,0.02 Beta = 0.12 ( 4 . 1 7 , 4.42) Beta = -0.48 (-3.27, 2.30) P-interaction by genotype - 0.81 Adjustment factors Maternal age at enrollment, child sex, examiner, Home Observation for Measurement of the Environment score, alcohol consumption during pregnancy, laboratory batch, season of urine collection, urinary creatinine, and bioraarker X genotype interaction; 24-month model also adjusted for maternal race/ethnicity Comments Interactions between DAPs, DMPs, and DEPs and PON1,92 genotype were detected among blacks and Hispanics at 12 months, but not at 24 months (results NR) Results were similar when stratified by child genotype (available for 57% of subjects; results NR) Tier 3/4 Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00052 V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Doi Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Engel et al. (2011) Engel et al. (2011) m m i Bayley Psychomotor Development Index at 12 months Maternal prenatal urinary DAPs (nmoi/L, lo g ,, scale) Maternal prenatal urinary DMPs (mnol/L, log,,, scale) 149 total 111 blacks/ Hispanics 38 whites 149 total 111 blacks/Hispanics 38 whites Tertile 1 adj. mean, total = 95.3 (90.9, 99A) Tertile 2 adj. mean, total = 96.6 (92.1, 101.1) Tertile 3 adj. mean, total = 92.5 (88.5, 96.6) Beta, total = -0.52 (-3.66. 2.62) Tertile 1 adj. mean, blacks/ Hispanics = 97.7 (93.1, 102.4) Tertile 2 adj. mean, blacks/ Hispanics = 97.5 (93.0, 102.1) Tertile 3 adj. mean, blacks/ Hispanics - 94.2 (89.5, 98.9) Beta, blacks/Hispanics = -1.52 (-5.21, 2.16) Tertile 1 adj. mean, whites = 90.0 (80.5, 99.6) Tertile 2 adj. mean, whites = 97.0 (89.2, 104.7) Tertile 3 adj. mean, whites 90.8 (83.3, 98.2) P-interaction by race = 0.65 Beta, whites = 2.07 (-3.83, 7.96) P-interaction by race = 0.31 Tertile 1 adj. mean, total = 95.3 (91.2, 99.5) Tertile 2 adj. mean, t o t a l 94.5 ( 9 0 .1 ,9S',9) Tertile 3 adj. mean, total = 93.6 (89.3, 98.0) Beta, total = -0.20 (-3.28, 2.87) Tertile 1 adj. mean, blacks/ Hispanics = 97.7 (93.1, 102.4) Tertile 2 adj. mean, blacks/ Hispanics = 95.9 (91.2, 100.6) Tertile 3 adj. mean, blacks/ Hispanics = 95.6 (91.0, 100.2) Beta, blacks/Hispanics = -0.48 (-4.11, 3.16) Tertile 1 adj. mean, whites = 92.1 (84.6, 99.6) Tertile 2 adj. mean, whites 94.4 (86.0, 102.7) Tertile 3 adj. mean, whites = 91.7 (83.5, 99.9) P-interaction by race = 0.25 Beta, whites = 0.46 (-5.12, 603) P-interaction by race = 0.78 Maternal age at enrollment, child sex, examiner, maternal PON1 enzyme activity, season of urine collection, laboratory batch. Home Observation for Measurement of the Environment score, alcohol consumption during pregnancy, urinary creatinine, and race/ethnicity (if not stratified; also adjusted for biomarker X race/ ethnicity if stratified) Psychomotor Development Index rates fine and gross motor coordination; age-standardized to a mean of 100 and S D o f 15 Metabolites were not associated with Psychomotor Development Index at 24 months (results NR) (Continued ) V Low-level OP insectide exposures 583 ED 002061 00046431-00053 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 584 R. R eiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. (Continued ) Reference Engel et al. (2011) Outcome Exposure Maternal prenatal urinary DEPs (nmol/L, log,0 scale) Engel et al. (2011) Bayley Psychontotor Development Index at 24 months Maternal prenatal urinary DAPs (mnol/L, log,0 Scale) Engel et ai. ('2011) Engel et ai. (2011) m i m Maternal prenatal urinary DMPs inmoi/L, log.,, scale) Maternal prenatal urinary DEPs (nniol/L, log,0 scale) Number of subjects/events Estimate of association (95% Cl) 149 total Tertile 1 adj. mean, total = 95.1 11 blacks/Hispanics (90.7, 99.5) 38 whites Tertile 2 adj. mean, total = 93.7 (89.3, 9841) Tertile 3 adj. mean, total = 94.5 (90.6, 98.5) Beta, total = -0.92 (-3.68, 1.85) Tertile 1 adj. mean, blacks/ Hspanles - 97.8 (93.2, 102.4) Tertile 2 adj. mean, blacks/ Hspanles = 94.5 (90.1, 99.0) Tertile 3 adj. mean, blacks/ Hispanics = 96.4 (92.0, 100.8) Beta, blacks/Hispanics = -1.81 (-5.07, 1.45) Tertile 1 adj. mean, whites = 92.5 (84.9, 100.2) Tertile 2 adj. mean, whites = 94.4 (86.7, 102.1) Tertile 3 adj. mean, whites = 89.5 (80.2, 98.8) P-interaction by race = 0.83 Beta, whites = 1.36 (-3.83, 6.56) P-interaction by race = 0.31 210 Tertile 1 adj. mean =94.8 (90.5, 99.1) Tertile 2 adj. mean = 94.5 (90.2, 98.8) Tertile 3 adj. mean = 95.1 (90.9, 99.2) Beta = 0.9.3 (-1.41, 3.28) 210 Tertile 1 adj. mean = 9 4.7 (90.5, 98.9) Tertile 2 adj. mean = 94.9 (90.6, 99.1) Tertile 3 adj. mean = 94.8 (90.5, 99.1) Beta = 0.36 (--1,70, 2.43) 210 Tertile 1 adj. mean = 94.8 (90.5, 99.0) Tertile 2 adj. mean = 95.4 (91.4, 99.5) Tertile 3 adj. mean = 94.2 (90.2, 98.1) Beta = 0.67 (-1.72, 3.06) Adjustment factors Maternal age at enrollment, child sex, examiner, maternal education, maternal PON 1 enzyme activity, season of urine collection, laboratory batch, Home Observation for Measurement of the Environment score, alcohol consumption during pregnancy, urinary creatinine, and race/ethnicitv Comments Results were not heterogeneous by exact age at 24-month testing ('results NR) CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00054 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Engel et al. (2011) Wechsler full-scale intelligence quotient at 6-9 years Maternal prenatal urinary DAPs (mnol/L, log,,. scale) 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years Beta = -1.39 (-4.54, 1.77) Beta ---1.10 (-5.01, 2.81) Beta = -1.14 (-4.55, 2.28) Engel et al. (2011) Engel et ai. (2011) Engel et al. (2011) Engel et al. (2011) Engel et ai. (2011) Engel et al. (2011) Engel et al. m (2011) m i m Wechsler perceptual reasoning at 6-9 years Maternal prenatal urinary DAPs (nmol/L, log10 scale) by PON1 ,9? genotype Maternal prenatal urinary DMPs (nmol/L, log,,, scale) Maternal prenatal urinary DMPs (nmol/L, log,,, scale) by PON 1,9, genotype Maternal prenatal urinary DEPs (nmol/L, log,,, Scale) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) by PON 1,9, genotype Maternal prenatal urinary DAPs (nmol/L, iog,0 scale) Maternal prenatal urinary DAPs (nmol/L, log,,, scale) by P O N l,92 genotype 101 PON02 QR/RR 39 PON92 QQ 142 ages 6-9 years 115 ages 7-9 years 98 age 6 years 101 PONQ2 QR/RR 39 P 0 N 92QQ 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years 101 PON, 0} QR/RR 39 PONjgi 'QQ 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years 101 PO/V,9, QR/RR 39 P O N QQ Beta = -0.66 (-4.33, 3.00) Beta = -2.33 (-8.40,3.74) P-interaction --0.64 Beta = -0.46 (-3.17,2.26) Beta - -0.39 (-3.64, 2.86) Beta ---0.56 (-3.68,2.56) Beta = 0.28 (-2.89, 3.44) Beta - -1.79 (-6.83, 3.25) P-interaction = 0.49 Beta = -2.89 (-6.15, 0.36) Beta - -3.15 (-7.19,0.89) Beta = -1.40 (-5.27, 2.47) Beta ---2.32 (-6.49, 1.86) Beta --3.1.3 (-8.21, 1.96) P-interaction - 0.80 Beta = -2.36 (-6.04, 1.31) Beta = -2.39 (-6.97, 2.19) Beta = -2.07 (-5.66, .52) Beta = -0 .5 6 (--4.80, 3.68) Beta - -7.52 (-14.53, -0.51) P-interaction - 0.09 Sex, race/ethnicity, maternal education, language in the home, alcohol consumption during pregnancy, laboratory batch, season of urine collection, urinary creatinine, Wechsler version (if combined), and maternal PON 1 enzyme activity (unless stratified by genotype) Subtests of Wechsler Preschool and Primary Scale of Intelligence, 3rd Edition: Block Design, Information, Matrix Reasoning, Vocabulary, Picture Concepts, Symbol Search, Word Reasoning, and Coding Subtests of Wechsler Intelligence Scale for Children, 4th Edition: Block Design, Similarities, Digit Span, Picture Concepts, Coding, Vocabulary, Letter-Number Sequence, Matrix Reasoning, Comprehension, and Symbol Search Associations with Wechsler outcomes were not heterogeneous by race/ ethnicity (results NR) " (Continued) V L o w -lV l OP i f t S C t i d XpO S U T S ED 002061 00046431-00055 585 ED 002061 00046431-00056 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Engel et al. (2011) Outcome Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Wechsler verbal comprehension at 6-9 years Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Wechsler processing speed at 6-9 years Exposure Maternal prenatal urinary DMPs (nmol/L, login scale) Maternal prenatal urinary DMPs (nmol/L, log,Q scale) by P O N l102 genotype Maternal prenatal urinary DEPs (nmol/L, log10 scale) Maternal prenatal urinary DEPs (nmol/L, log,0 scale) by PG N i, 92 genotype Maternal prenatal urinary DAPs (nmol/L, log,0 scale) Maternal prenatal urinary DAPs (nmol/L, log10 scale) by PON1 ,9? genotype Maternal prenatal urinary DMPs (nmol/L, log,,, scale) Maternal prenatal urinary DMPs (nmol/L, log,0 scale) by PON 1,9, genotype Maternal prenatal urinary DEPs (nmol/L, log,,, Scale) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) by PON 1,9, genotype Maternal prenatal urinary DAPs (nmol/L, log,0 scale) Number of subjects/events 142 ages 6-9 years 11 5 ages 7-9 years 98 age 6 years 101 PO N ,g, QR/RR 39PONm QQ 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years 101 PON0 QR/RR 39 PON]9' q Q 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years 101 PONj02 QR/RR 39 PON,92 QQ 142 ages 6-9 years 115 ages 7-9 years 98 age 6 years 101 PONQ2 QR/RR 39 PONj92QQ 140 ages 6-9 years 114 ages 7-9 years 96 age 6 years 101 PON, 0} QR/RR 39 PON, 92 QQ 114 ages 7-9 years 96 age 6 years Estimate of association (95% Cl) Beta - -1.15 (-4.31,2.02) Beta = -1.24 (-5.05, 2.57) Beta = -1.46 (-4.74, 1.83) Beta - 0.71 (-2.96, 4.38) Beta ---6.15 (-11.99, -0.31 ) P-interaetion = 0.05 Beta ---3.51 (-7.31,0.30) Beta = -4.37 (-9.10, 0.36) Beta = -1.59 (-5.68, 2.50) Beta = -3.24 (-8.11, 1.62) Beta ---4.80 (-10.73, 1.13) P-interaction --0.68 Beta ---0.42 (-.3.45, 2.62) Beta - 0.56 (-3,11, 4.23) Beta = -1.16 (--4.59, 2.27) Beta ---0.33 (-3.87,3.20) Beta = 0.73 (-5.12, 6.59) P-interaction = 0.76 Beta - -0.05 (-2.64, 2.54) Beta - 0.39 (-2.65, 3.42) Beta ---0.52 (-3.67, 2.62) Beta --0.12 (-2.93, 3.16) Beta = 0.24 (-4,60, 5.09) P-interaction = 0.97 Beta = -1.20 (-4.35, 1.96) Beta - -0.08 (-3.91,3.76) Beta = -2.27 (-6.14, 1.60) Beta = -0.45 (-4,51,3.60) Beta - -1.20 (-6.13, 3.74) P-interaction - 0.81 Beta = -1.05 (-5.57,3.46) Beta ---1.22 (-5.12, 2.67) Adjustment factors s* Comments Crit Rev Toxicol. 2015; 45(7): 531-641 5 8 6 R. R(?iss t ill. Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. d o i Low-level o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. (2011) Engel et al. ('2011) Young et al. (2005) Young et al. (2005) Young et al. (2005) m i m Wechsler working memory at 7-9 years Brazelton habituation cluster at < 2 months Maternal prenatal urinary DMPs (nmoi/L, log.,, Scale) Maternal prenatal urinary DEPs (nmol/L, login scale) Maternal prenatal urinary DAPs (nmoi/L, log.,, scale) Maternal prenatal urinary DMPs (nmoi/L, log,0 Scale) Maternal prenatal urinary DEPs (nmoi/L, log,,, scale) Maternal prenatal urinary DAPs (mnol/L, log,0 scale) Maternal prenatal urinary DMPs (nmoi/L, log,,, scale) Maternal prenatal urinary DEPs (nmoi/L, log,,, scale) 115 ages 7-9 years Beta = -0.79 (-4.52, 2.94) 98 age 6 years Beta - -0.84 (--4.35, 2.67) 114 ages 7-9 years Beta - -2.11 (-6.81, 2.59) 96 age 6 years Beta - - 1 .8 5 (-6.25, 2.56) 114 Beta = -0.53 (-4.24,3.18) 115 Beta = 0.29 (-2.81, 3.38) 114 Beta ---.3.48 (-7.29,0.34) 175 total 107 age < 3 days 66 age > 3 days 175 total 107 age < 3 days 66 age > 3 days 175 total 107 age < 3 days 66 age > 3 days Beta - 0.03 (-0.34, 0.40) Beta, age < 3 days = 0.10 (-0.40, 0.60) " Beta, age > 3 days - 0.06 (-0.54, 0.66) " No association with maternal post-delivery urinary metabolites (results NR) Beta - -0.06 (-0.39, 0.27) Beta, age < 3 days ---0.04 (-0.49, 0.40) ' Beta, age > 3 days - 0.04 (-0.50, 0.58) " No association with maternal post-delivery urinary metabolites (results NR) Beta - 0.33 (-0.06, 0.72) Beta, age < 3 days = 0.47 (-0.05, 0.99) " Beta, age > 3 days - 0.20 (-0.43, 0.83) " No association with maternal post-delivery urinary metabolites (results NR) Age at assessment, smoking, alcohol, method of delivery, minutes since fed at assessment, and interviewer No change after additional adjustment for maternal post-delivery urinary metabolite levels, birth weight, or gestational age; some attenuation after adjustment for creatinine (results NR) Habituation cluster includes light, rattle, bell, and pin-prick Median age at assessment: 3 days (IQR: 1-26) "Urinary metabolite levels measured at the two points during pregnancy were not significantly correlated with each other or with the post-delivery measurement, with all estimated correlations below' 0.1 for total DAP, dimethyl, and diethylphosphate metabolite levels." (Continued) Low-level OP insectide exposures 587 ED 002061 00046431-00057 588 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Young et ai. (2005) Outcome Brazelton orientation cluster at < 2 months Exposure Maternal prenatal urinary DAPs (nmol/L, log,,, scale) Young et al. (2005) Young et aS. (2005) Maternal prenatal urinary DMPs (nmol/L, log,,, scale) Maternal prenatal urinary DEPs (nmol/L, log,0 scale) Young et aS. (2005) Brazelton motor performance cluster at < 2 months Maternal prenatal urinary DAPs (nmol/L, log,0 scale) Young et ai. (2005) Young et ai. (2005) Maternal prenatal urinary DMPs (nmol/L, log,,, scale) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) Number of subjects/events 379 total 197 age < 3 days 182 age > 3 days 379 total 197 age < 3 days 182 age > 3 days 379 total 197 age < 3 days 182 age > 3 days 381 total 197 age > 3 days 184 age > 3 days 381 total 19/ figs --3 days .184 age > 3 days 381 total 19/ age --3 days 184 age > 3 days Estimate of association (95% Cl) Adjustment factors Comments Beta = - 0 .17 (-0.50. 0.17) Age at assessment, interviewer, and Orientation cluster includes inanimate Beta, age < 3 days - -0.02 (-0.53, 0.49) number of prenatal care visits visual, inanimate auditory, inanimate Beta, age > 3 days =-0.13 (-0.54,0.27) No change after additional adjustment visual-auditory, animate visual, animate No association with maternal for maternal post-delivery urinary auditory, animate visual-auditory, and post-delivery urinary metabolites metabolite levels, birth weight, or alertness (results NR) gestational age; some attenuation after adjustment for creatinine (results NR) Beta = -0.12 (-0.43, 0.19) Beta, age < 3 days =-0.08 (-0.54,0.39) Beta, age > 3 days = 0.01 (-0.37, 0.38) No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.32 (-0.66, 0.03) Beta, age < 3 days = -0.11 (-0.65, 0.43) Beta, age > 3 days - -0.33 (-0.73, 0.08) No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.03 (-0.19, 0.14) Beta, age < 3 days = 0.04 (-0.20, Age at assessment, poverty level, Motor performance cluster includes gestational age at initiation of prenatal tonus, maturity, puii-to-sit, defense, and 0.28) care, and interviewer activity Beta, age > 3 days = -0.07 (-0.28, 0.15) No change after additional adjustment No association with maternal for maternal post-delivery urinary post-delivery urinary metabolites metabolite levels, birth weight, or (results NR) gestational age; some attenuation after adjustment for creatinine (results NR) Beta = -0.05 (-0.20, 0.10) Beta, age< 3 days = 0.03 (-0.19, 0,24)' Beta, age > 3 days = -0.11 (-0.31, 0.09) No association with maternal post-delivery urinary metabolites (results NR) Beta = 0.10 (-0.06, 0.27) Beta, age < 3 days = 0.08 (-0.17, 0,33)' Beta, age > 3 days = 0.17 (-0.05, 0.38) No association with maternal post-delivery urinary metabolites (results NR) m Tier 3/4 CritRevTcodcoL 2015; 45(7): 531-641 ED 002061 00046431-00058 s* Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Young et al. (2005) Young et al. (2005) Young et al. (2005) Young et al. (2005) Young et al. (2005) m m i ; m Brazeiton range of state cluster at < 2 months Maternal prenatal urinary DAPs (nmoi/L, log.,, Scale) Maternal prenatal urinary DMPs (nmol/L, login scale) Maternal prenatal urinary DEPs (nrnol/L, login scale) Brazeiton regulation of state cluster at < 2 months Maternal prenatal urinary DAPs (nrnol/L, login scale) Maternal prenatal urinary DMPs (nmoi/L, log,0 scale) 381 total 197 age Sr 3 days 184 age > 3 days 381 total 197 age > 3 days 184 age A .3 days 381 total 197 age > 3 days 184 age A .3 days 381 total 197 age > 3 days 184 age A .3 days 381 total 197 a g e> 3 days 184 age > 3 days Beta = 0.09 (-0.16, 0.34) Beta, age S 3 days - 0.11 (-0.21, 0.43)" Beta, age > 3 days - -0.02 (-0.44, 0.40)" No association with maternal post-delivery urinary metabolites (results NR) Beta - 0.08 (--0.15, 0.32) Beta, age < 3 days = 0.17 (-0.12, 0.46) Beta, age > 3 days = -0.12 (-0.51, 0.27)" No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.02 (-0.27, 0.24) Beta, age < 3 days = -0.21 (-0.54, 0.12) Beta, ase > 3 days = 0.20 (-0.21, 0.62)" No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.07 (-0.39, 0.24) Beta, age < 3 days = -0.07 (-0.50, 0.36) Beta, age > 3 days = -0.10 (-0.58, 0.37) No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.05 (-0.34, 0.24) Beta, age < 3 days = -0.06 (-0.45, 0.33) Beta, age > 3 days = -0.06 (-0.50, 0.39) No association with maternal post-delivery urinary metabolites (results NR) Age at assessment, number of prenatal care visits, gestational age at initiation of prenatal care, aicohoi, and interviewer Range of state cluster includes peak of excitement, rapidity of build-up, irritability, and lability of state No change after additional adjustment for maternal post-delivery urinary metabolite levels, birth weight, or gestational age; some attenuation after adjustment for creatinine (results NR) Age at assessment, pre-pregnancy body mass index, infant sex, parity, caffeine use, and interviewer Regulation of state cluster includes cuddl iness, consol abil ity, self-quietinj and hand-to-mouth No change after additional adjustment for maternal post-delivery urinary metabolite levels, birth weight, or gestational age; some attenuation after adjustment for creatinine (results NR) (Continued) Low-lcwl OP USECtdE E x p o SUEES 589 ED 002061 00046431-00059 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 590 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. (Continued ) Reference Young et al. (2005) Outcome Exposure Maternal prenatal urinary DEPs (nmol/L, log,0 scale) Young et ai. (2005) Brazelton autonomic stability cluster at < 2 months Maternal prenatal urinary DAPs (nmol/L, iog,0 scale) Young et al. (2005) Maternal prenatal urinary DMPs (nmol/L, log,0 scale) Young et al. (2005) Maternal prenatal urinary DEPs (nmol/L, log,0 scale) Young et al. (2005) Brazelton reflexes cluster at < 2 months Maternal prenatal urinary DAPs (nmol/L, login scale) H Young et al. ** (2005) 1 m Maternal prenatal urinary DMPs (nmol/L, log.,, scale) Number of subjects/events 381 total 197 a g e> 3 days 184 age > 3 days 381 total 19/ --3 days .184 age > 3 days 381 total 197 age > 3 days 184 a g e> 3 days 381 total 197 age > 3 days 384 age A 3 days 381 total 197 age > 3 days 184 age ^ 3 days 381 total 197 age > 3 days 384 age ' 3 days Estimate of association (95% Cl) Adjustment factors Comments Beta = -0.1 5 (-0.47, 0.17) Beta, age 3 days - -0.08 (-0.52. 0.37) Beta, age > 3 days = -0.24 (-0.72, 0.24) No association with maternal post-delivery urinary metabolites (results NR) Beta = -0.16 (-0.36. 0.05) Beta, age 3 days = -0.09 (-0.38, 0,20)' Beta, age > 3 days ---0.19 (-0.49, 0.12) No association with maternal post-delivery urinary metabolites (results NR) Beta ---0.17 (-0.35, 0.02) Beta, age 3 days ---0.15 (-0.42, 0.11) Beta, age > 3 days - -0.14 (-0.43, 0.14) No association with maternal post-delivery urinary metabolites (results NR) Beta --0.06 (-0.15, 0.27) Beta, age 3 days --0.31 (0.01, 0.61) Beta, age > 3 days ---0.16 (-0.47, 0.14)' No association with maternal post-delivery urinary metabolites (results NR) Beta = 0.23 (0.05, 0.41) Beta, age 3 days = -0.01 (-0.24, 0.22) Beta, age > 3 days - 0.53 (0.23, 0.82) No association with maternal post-delivery urinary metabolites (results NR) Age at assessment, infant sex, parity, vitamin use, minutes since fed at assessment, interviewer, and illicit drug use during pregnancy No change after additional adjustment for maternal post-delivery urinary metabolite levels, birth weight, or gestational age; some attenuation after adjustment for creatinine (results NR) Age at assessment, maternal age at delivery, smoking, vitamin use, interviewer, and mean diastolic and systolic blood pressure No change after additional adjustment for maternal post-delivery urinary metabolite levels, birth weight, or gestational age; some attenuation after adjustment for creatinine (results NR) Autonomic stability cluster includes tremors, startles, and skin color Reflex cluster includes plantar, Babinsid, ankle clonus, rooting. sucking, glabella, passive resistance of legs, passive resistance of arms, palmar, placing, standing, walking, crawling, incurvation, tonic deviation of head and eyes, nystagmus, tonic neck reflex, and rnoro reflex Beta = 0.18 (0.02, 0.34) Beta, age 3 days ---0.00 (-0.21, 0.20) Beta, age > 3 days - 0.41 (0.12, 0.69) No association with maternal post-delivery urinary metabolites (results NR) CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00060 m. ED 002061 00046431-00061 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Young et al. (2005) Young et al. (2005) Young et al. (2005) Young et al. (2005) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) Brazeiton > 3 abnormal reflexes at > 3 days to < 2 months Maternal prenatal urinary DAPs (nmol/L, log10 scale): 1.07-1.65 1.65-1.83 1.83-2.07 2.08-2.30 2.31-3.17 Maternal prenatal urinary DMPs (nmol/L, log,,. scale): 0.68-1.48 1.48-1.69 1.70-1.95 1.95-2.19 2.19-3.15 Maternal prenatal urinary DEPs (nmol/L, log,,, scale): 0.51-0.90 0.90-1.10 1.11-1.27 1.28-1.58 1.58-2.35 381 total 197 age > 3 days 184 age > 3 days 3 of 37 5 of 37 6 of 37 5 of 37 12 of 36 3 of 37 5 of 37 6 of 37 9 of 37 8 of 36 6 of 37 2 of 37 6 of 37 5 of 37 12 of 36 Beta - 0.22 (0.04, 0.40) Beta, age< 3 days --0.08 (-0.16, 0,32)' Beta, age > 3 days --0.37 (0.09, 0.6' No association with maternal post-delivery urinary metabolites (results NR) Proportion = 8% Proportion - 14% Proportion --16% Proportion - 14% Proportion --33%: P-trend = 0.01 Odds ratio per unit increase - 4.9 (1.5, 16.1) No association with maternal post-delivery urinary metabolites (results NR) Proportion - 8% Proportion --14% Proportion = 16% Proportion - 24%' Proportion --22% P-trend - 0.03 Odds ratio per unit increase --3.2 (1.1, 9.8) No association with maternal post-delivery urinary metabolites (results NR) Proportion - 16%' Proportion --5% Proportion - 16% Proportion --14% Proportion = 33 % P-trend - 0.05 Odds ratio per unit increase --.3.4 (1.2, 9.9) No association with maternal post-delivery urinary metabolites (results NR) m m i ; m (Continued) Low-level OP insectide exposures 591 d o i Low-level o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Eskenazi et al. (2007) Outcome Bayley Mental Development Index at 6, 12, or 24 months Exposure Maternal or child urinary DAPs (nmol/L, log,0 scale) Eskenazi et al. (2007) Eskenazi et al. (2007) Eskenazi et al. (2007) Maternal or child urinary DMPs (nmol/L, log,0 scale) Maternal or child urinary DEPs (nmol/L, log,0 scale) Maternal urinary MDA (pg/L) Number of subjects/events 395 at 6 months 393 at 12 months 369 at 24 months 395 at 6 months 393 at 12 months 369 at 24 months 395 at 6 months 393 at 12 months 369 at 24 months 39% detectable Ui ls> Estimate of association (95% Cl) Adjustment factors Comments >3 Prenatal beta = -1.15 (-2.89, 0.59) Psychometrician, location, exact age Mean SD age (months) at child Child beta = -0.17 (-1.23, 0.90) at assessment, sex, breast-feeding assessments: 6.6 1.1, 12.8 1.6, and Prenatal beta = -1.34 (-3.59. 0.92) duration, score on Infant-Toddler 24.6 1.1 Child beta = 1.36 (-0,05, 2.78) Home Observation for Measurement Bayley Scales of Infant Development Prenatal beta = -3.54 (-6.59, -0.49) of the Environment instrument, are standardized by age to a mean Child beta = 2.37 (0.50, 4.24) household income above poverty SD of 100 15; scores < 85 indicate threshold, parity, and maternal possible developmental delay Peabody Picture Vocabulary Test score Longitudinal analyses of DAPs and Bayley scores produced similar findings (not reported) Prenatal beta = -0.95 (-2.52, 0.62) Child beta = -0.31 (-1,28, 0.67) Prenatal beta = -1.06 (-3.12, 0.99) Child beta = 0.75 (-0,44, 1.93) Prenatal beta = -3.64 (-6.36, -0.91) Child beta = 2.01 (0.24, 3.78) Prenatal beta = -0.16 (-1.96, 1.65) 11 .... Child beta = 0.24 (-0.78, 1.25) Prenatal beta = --1.14 (--3.51. 1.22) Child beta = 1.89 (0.21,3.58) Prenatal beta = -0,85 (-3.98, 2.27) Child beta = 1.02 (-0,52, 2.57) At 6 months: - Undetectable: beta - referent Detectable < median: beta = 0.98 (-0.85, 2.81) Detectable > median: beta = -0.25 (-2.10, 1.60) At 12 months: Undetectable: beta = referent Detectable < median: beta = 0.95 (-1.55, 3.46) Detectable ar median: beta = 2.40 (-0.13, 4.94) At 24 months: Undetectable: beta = referent Detectable < median: beta = -1.09 (-4.51, 2.32) Detectable a median: beta = 0.24 Q (-3.03, 3.52) s Tier 3/4 ED 002061 00046431-00062 m i ; *4 m ED 002061 00046431-00063 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com bv Kathi Lohmann on 07/20/15 For personal use only. Eskenazi et al. Maternal urinary 91% detectable At 6 months: " (2007) TCPy Uig/L) Undetectable: beta = referent Detectable < median: beta = 0.24 (-2.12, 2.61) Detectable > median: beta - 0.08 (-2.29, 2.44) At 12 months: Undetectable: beta --referent Detectable < median: beta ---0.45 (-3.67, 2.76) Detectable a median: beta = -0.65 (-3.88, 2.58) At 24 months: Undetectable: beta --referent Detectable < median: beta - - 1 .0 2 (-5.34, 3.31) Detectable a median: beta - -1.94 (-6.26, 2.37) Eskenazi et al. Bayley Psychomotor Maternal or child 396 at 6 months Prenatal beta = -0.71 (-3.28, 1.86) (2007) Development urinary DAPs Child beta = 0.39 (-1.18, 1.97) Index at 6, 12, or 24 months (nmoi/L, log,,. scale) 392 at 12 months Prenatal beta - -0,60 ( - 3 .//. 2 .51) Child beta = 1.22 (-0,78, 3.21) 371 at 24 months Prenatal beta - -1,28 (-4.01, 1.46) Child beta = 1.06 (-0.62, 2.74) Eskenazi et al. Maternal or child 396 at 6 months Prenatal beta = -0.55 (-2.88, 1.771 (2007) urinary DMPs Child beta --0.28 (-1.17. 1.72) (nmol/L, log,,. 392 at 12 months Prenatal beta = -1.15 (-4.03, 1.74) scale) Child beta = 0.46 (-1.22, 2.13) 371 at 24 months Prenatal beta = -1,24 (-3.70, 1,21) Child beta - 1.01 (-0.58, 2.60) Eskenazi et ai. Maternal or child 396 at 6 months Prenatal beta = 0.02 (-2.63, 2.67) " (2007) urinary DEPs Child beta = 0.60 (-0.89. 2.09) (nmoi/L, log,,, 392 at 12 months Prenatal beta = 0.30 (-3.03, 3.63) scale) Child beta = 1.91 (-0.46. 4.27) 371 at 24 months Prenatal beta = -0.86 (-3.64, 1.92) Child beta = 0.30 (-1.07, 1.67) m m i ; m V' __ .... - (Continued) 593 OP H S C t i d 8XpOSUVES L o w -lV C l DOI Low-evel OP insecide exposures 594 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Eskenazi et ai. (2007) Outcome " Exposure Maternal urinary MDA (pg/L) ' Eskenazi et al. (2007) Maternal urinary TCPy (pg/L)' Eskenazi et al. (2007) Child Behavior Checklist attention problems syndrome score at 24 months Maternal or child urinary DAPs (mnol/L, log,0 scale) m m i ; m Number of subjects/events Estimate of association (95% Cl) 39% detectable 91% detectable 30 (8.4%) borderline At 6 months: Undetectable: beta --referent Detectable < median: beta = 0.42 (-2.34, 3.18) Detectable > median: beta = -1.45 (-4.21, 1.32) At 12 months: Undetectable: beta - referent Detectable < median: beta ---0.53 (-4.05, 3.00) Detectable a median: beta - 0.75 (-2.81,4.31) At 24 months: Undetectable: beta --referent Detectable < median: beta = -0.73 (-3.87, 2.41) Detectable ar median: beta --0.33 (-2,68, 3.35) At 6 months: Undetectable: beta --referent Detectable < median: beta = -0.56 (-4.03, 2.91) Detectable ar median: beta ---0.21 (-3,69, 3.27) At 12 months: Undetectable: beta --referent Detectable < median: beta ---0.70 (-5.26, 3.86) Detectable > median: beta - -1.62 (-6.20, 2.96) At 24 months: Undetectable: beta --referent Detectable < median: beta ---2.65 (-6.50, 1.21) Detectable a median: beta ---2.72 (-6.57, 1.12) Prenatal odds ratio = 0.77 (0,27, 2.24) Child odds ratio --1.41 (0.75, 2.64) Adjustment factors Sex, exact age at assessment, breast-feeding duration, score on Infant-Toddler Home Observation for Measurement of the Environment instrument, household income above poverty threshold, parity, maternal Peabody Picture Vocabulary Test score, and maternal depression Comments "Borderline" score > 93rd percentile "Clinical" score > 97th percentile (N = 7, 2.0%) Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00064 ^4 doi Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Eskenazi et al. (2007) Eskenazi et ai. (2007) Eskenazi et al. (2007) Eskenazi et al. (2007) Eskenazi et al. (2007) Eskenazi et ai. (2007) Eskenazi et al. (2007) Eskenazi et al. (2007) Eskenazi et ai. (2007) Eskenazi et al. (2007) Eskenazi et al. (2007) Child Behavior Checklist ADHD score at 24 months " Child Behavior Checklist pervasive developmental disorder score at 24 months " Maternal or child urinary DMPs (nmoi/L, log,,, scale) Maternal or child urinary DEPs (nmol/L, log,,, scale) Maternal urinary MDA or TCPy (Pg/L) Maternal or child urinary DAPs (nmol/L, log, 0 scale) Maternal or child urinary DMPs (nmoi/L, log,,, Scale) Maternal or child urinary DEPs (nmol/L, log,,, scale) Maternal urinary MDA or TCPy (Pg/L) Maternal or child urinary DAPs (nmol/L, log, ,, scale) Maternal or child urinary DMPs (nmol/L, log,,, scale) Maternal or child urinary DEPs (nmol/L, log,0 scale) Maternal urinary MDA or TCPy (jg/L) 30 (8.4%) borderline Prenatal odds ratio = 0.78 (0.31, 1.96) Child odds ratio = 1.54 (0.85, 2.76) 30 (8.4%) borderline Prenatal odds ratio - 0.78 (0,26, 2.31) Child odds ratio = 1.02 (0.61. 1,71) 30 (8.4%) borderline "no significant associations" (results NR) 34 (9.6%) borderline Prenatal odds ratio = 1.34 (0.50, 3.59) Child odds ratio M l (0.61, 2.03) 34 (9.6%) borderline Prenatal odds ratio --1.27 (0.53, 3.04) Child odds ratio = 1.10 (0.63, 1.94) 34 (9.6%) borderline Prenatal odds ratio = 0.59 (0,21, 1.68) Child odds ratio = 1.18 (0.72. 1,94) 34 (9.6%) borderline "no significant associations" (results NR) 51 (14.4%) clinical Prenatal odds ratio = 2.25 (0.99, 5,16) Child odds ratio = 1.71 (1.02, 2.87) 51 (14.4%) clinical Prenatal odds ratio = 2.19 (1,05, 4.58) Child odds ratio = 1.52 (0.94. 2,45) 51 (14.4%) clinical Prenatal odds ratio = 0.88 (0,37, 2.07) Child odds ratio = 1.72 (1.12, 2.64) 51 (14.4%) clinical "no significant associations" (results NR) m m i ; "Borderline" score > 93rd percentile "Clinical" score > 97th percentile (N = 10, 2.8%) "Borderline" score > 93rd percentile (N = 105, 29.6%) "Clinical" score > 97th percentile (Continued) Tier 3/4 Low-level O P insectide exp o su res 595 ED 002061 00046431-00065 ^4 R. R e iss e t al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Eskenazi et ai. (2010) Outcome Bayley Mental Development Index at 24 months Eskenazi et al. (2010) Eskenazi et al. (2010) Eskenazi et al. (2010) Eskenazi et al. (2010) Exposure Maternal prenatal urinary DAPs (nmof/L, log.,, scale) by child genotype Number of subjects/events 111 PON1 ,nHCC 179 PON1 ;o,C T 74 PONl_ws XT 94 PON1 ]02 RR 188 P O N ltg, QR 86 PON1 ;92 QQ Maternal prenatal urinary DMPs (nniol/L, log,0 scale) by child genotype 111 PONl ,m CC 119 PONI CT 74 PONl_I0S I T 94 PO N l ,9,, RR 188 P O N lm QR 86 PONl m QQ Maternal prenatal urinary DEPs (nrnol/L, log,0 scale) by child genotype 111 PONl ,m CC 179 PONl ,/,C T 74 PONl_I0S I T 94 PONl ,9, RR 188 P O N lm QR 86 PONl m QQ Maternal prenatal urinary DAPs (nniol/L, log10 scale) by cord blood PONI quantity or activity PONl quantity: 89 tertile 1 88 tertile 2 88 tertile 3 PONl activity: 91 tertile 1 85 tertile 2 87 tertile 3 Maternal prenatal urinary DMPs (nmol/L, !og,0 scale) by cord blood PON 1 quantity or activity PONl quantity: 89 tertile 1 88 tertile 2 88 tertile 3 PONl activity: 91 tertile 1 85 tertile 2 87 tertile 3 Estimate of association (95% Cl) Beta - -3.2 (-9.8, 3.5) Beta = -3.7 (-8.0, 0.6) Beta = -5.5 (-11.1, 0.1) P-interaction --0.98 Beta = -6.5 (-15.6, 2.6) Beta - -1.2 (-5.2, 2.9) Beta = -6.9 (-12.8, -0.9) P-interaction = 0.33 Beta = -2.2 (-8.0, 3.6) Beta = --3.4 (-7.4, 0.6) Beta = -5.9 (-1 1.1, -0.6) P-interaction - 0.91 Beta = -4.4 (-12.4, 3.6) Beta = -1.3 (-4.9, 2.4) Beta = -7.4 (-13.0, -1.9) P-interaction = 0.38 Beta = -0.3 (-7.2, 6.7) Beta = -1.7 (-6.3, 3.0) Beta - -3.4 (-8.8, 2.1) P-interaction = 0.84 Beta = 1.4 (-8.4, 11.1) Beta = -1.1 (-5.2, 3.0) Beta = -2.5 (-8.7, 3.6) P-interaction = 0.47 Beta = -5.4 (-1 1.9, 1.1) Beta = -4.3 (-11.6, 3.0) Beta = -1.2 (-8.7, 6.4) P-interaction = 0.89 Adjustment factors Age at assessment, sex, parity, breastfeeding duration, Infant-Toddler Home Observation for Measurement of the Environment score, maternal Peabody Picture Vocabulary Test score, household poverty status, psychometrician, and testing location Comments Interactions were statistically nonsignificant whether genotype was coded as a categorical or ordinal variable Associations were "similar, albeit weaker" when stratified by maternal genotype (not shown here) Associations across maternal PONl enzyme levels and activities at delivery were "similar to those for cord blood enzyme levels" (not shown here) Beta - -6.6 (-12.9, -0.2) Beta = -1.0 (-7.9, 5.9) Beta - -5.8 (-13.9, 2.2) P-interaction --0.72 Beta = -5.4 (-11.4, 0.5) Beta ---4.5 (-11.2, 2.3) Beta = -0.5 (-7,0, 6.1 ) P-interaction = 0.72 Beta = -6.7 (-12.6, -0.8) Beta = -0.9 (-7.2, 5.4) Beta ---4.2 (-11.2, 2.9) P-interaction = 0.97 m Tier 3/4 Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00066 s* ED 002061 00046431-00067 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Eskenazi et al. " Maternal prenatal PON1 quantity: " (2010) urinary DEPs 89 iertile 1 Beta = -3.2 (-9.7, 3.3) (nmol/L, log,0 88 tertiie 2 Beta = 0.2 (-6.8, 7.2) scale) by cord 88 tertiie 3 Beta = 2.7 (-5.2, 10.5) blood PON 1 P-interaction =:0.23 quantity or activity PON 1 aetivity: 91 tertiie 1 Beta = -3.0 (-10.0, 3.9) 85 tertiie 2 Beta = -1.7 (-7.8, 4.5) 87 tertiie 3 Beta = -0.5 (-8.5, 7.6) P-interaction =:0.68 Eskenazi et al. Bayley Psychomotor Maternal prenatal 111 PONl_,m CC Beta = -2.3 (-7.8, 3.3) " (2010) Development urinary DAPs 179 P O N lSm CT Beta = -0.8 (-4.8, 3.3) index at (nmol/L, log.n 74 PONI ,08 TT Beta = -1.0 (-7.1, 5.1) 24 months scale) by child P-interaction = 0.89 genotype 94 PONI m RR Beta = -1.7 (-8.7, 5.4) 188 PON'lo, QR Beta = 0.1 (-3.5, 3.8) 86 PONIlg'Q Q Beta = --5.1 (-11.1, 1.0) P-interaction = 0.53 Eskenazi et al. " Maternal prenatal 111 PONI CC Beta = -1.6 (-6.4, 3.3) " (2010) urinary DMPs 179 P O N l\m CT Beta = -0.3 (-4.0, 3.4) (nmol/L, log,0 74 PONI 2 TT Beta = -1.2 (-6.9, 4.4) scale) by child P-interaction = 0.87 genotype 94 P O N I RR Beta = -2.1 (-8.3, 4.0) 188 PONl0g? QR Beta = 0.7 (-2.6, 4.0) 86 PONI ]92 'QQ Beta = -4.7 (-10.4, 1.0) P-interaction = 0.36 Eskenazi et al. " Maternal prenatal 111 PONI m CC Beta = 0.9 (-4.9, 6.8) " (2010) urinary DEPs (nmol/L, logf0 179 PO N I _ m CT 74 P O N l _ TT Beta = -2.2 (-6.5, 2.1) Beta = -1.5 (-7.3, 4.2) scale) by child P-interaction = 0.66 genotype 94 P O N !]9,,RR Beta = 4.5 (-2.9, 11.9) 188 PONI ",92 QR Beta = -1.9 (-5.6, 1.8) 86 PONI m QQ Beta = -3.8 (-9.9, 2.3) P-interaction =:0.14 Eskenazi ei al. " Maternal prenatal PONI quantity: " (2010) urinary DAPs 89 tertiie 1 Beta = -2.4 (-8.3, 3.4) (nmol/L, log,0 88 tertiie 2 Beta = --1.8 (-8.3, 4.6) scale) by cord 88 tertiie 3 Beta = 1.2 (-5.7, 8.1) blood PON 1 P-interaction = 0.46 quantity or activity PON 1 aetivity: 91 tertiie 1 Beta = -4.7 (-10.6, 1.3) 85 tertiie 2 Beta = 0.0 (-6.6, 6.7) 87 tertiie 3 Beta = 1.5 (-5.4, 8.4) P-interaction = 0.69 Eskenazi et al. " Maternal prenatal PON 1 quantity: " (2010) urinary DMPs 89 tertiie 1 Beta = --1.1 (-6.5, 4.3) (nmol/L, log jq 88 tertiie 2 Beta = -3.2 (-9.2, 2.8) scale) by cord 88 tertiie 3 Beta = 1.7 (-4,3, 7.7) blood PON 1 P-interaction = 0.41 quantity or activity PON 1 aetivity: 91 tertiie 1 Beta = -4,1 (--9.6, 1.4) 85 fertile 2 Beta = 0.4 (-5.7, 6,5) 87 tertiie 3 Beta = 1.4 (-4,6, 7.4) P-interaction = 0.59 Interactions were statistically non significant whether genotype was coded as a categorical or ordinal variable Associations were "similar, albeit weaker" when stratified by maternal genotype (not shown here) Associations across maternal PON1 enzyme levels and activities at delivery were "similar to those for cord blood enzyme levels'' (not shown here) _ (Continued) Low-level O P insectide exp o su res 597 doi Low-level op insectide exposures 598 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Eskenazi et al. (2010) Outcome Eskenazi et al. (2010) Child Behavior Checklist pervasive developmental disorder score at 24 months Eskenazi et al. " (2010) Eskenazi et al. " (2010) Eskenazi et ai. " (2010) m m i m Exposure Maternal prenatal urinary DEPs (mrrol/L, log,0 scale) by cord blood PON 1 quantity or activity Number of subjects/events PONI quantity: 89 trtile 1 88 tettile 2 88 trtile 3 PONI activity: 91 trtile 1 85 tettile 2 87 tettile 3 Maternal prenatal urinary DAPs (nmoi/L, log.,, scale) by child genotype 111 PONI , m CC 179 PONI CT 74 PONl_m TT 94 PONI J02 RR 188 PONI ,,r , QR 86 PONI l92 QQ Maternal prenatal urinary DMPs (nntol/L, log,0 scale) by child genotype 111 PONI , o s CC 179 PONI , 0H CT 74 PONI lgg TT 94 PONI ,9, RR 188 P O N lm QR 86 PONI m QQ Maternal prenatal urinary DEPs (nntol/L, log,0 scale) by child genotype 111 PONI , o s CC 179 PONI , 0H CT 74 PONI m TT 94 PONI ,9, RR 188 P O N lm QR 86 PONI m QQ Maternal prenatal urinary DAPs (nniol/L, log,Q scale) by cord blood PON1 quantity or activity PONI quantity: 89 tettile 1 88 tettile 2 88 fertile 3 PONI activity: 91 tedile 1 85 tedile 2 87 tedile 3 Estimate of association (95% CI) Beta = -5.0 (-10.7, 0.7) Beta = 1.6 (-4.5, 7.8) Beta = 1.6 (-5,6, 8.8) P-interaction = 0.42 Adjustment factors " Comments .... Beta = -4.7 (-11.1, 1.6) Beta = -2.5 (-8.3, 3.4) Beta = 3.7 (-3.1, 10.5) P-interaction = 0.35 0 = 4.2(0.5,36.8) Odds ratio = 2.0 (0.6, 6.0) Odds ratio = 1.9 (0.3. 10.4) P-interaction = 0.91 Odds ratio = 5.4 (0.7, 44.0) Odds ratio = 1.2 (0.4, 3.6) Odds ratio = 5.2 (0.8, 35.1) P-interaction = 0.29 Odds ratio = 3.3 (0.5, 21.3) Odds ratio = 2.2 (0.8, 5.9) Odds ratio = 1.9 (0.4, 9.8) P-interaction = 0.94 Odds ratio = 4.8 (0.8, 31.1) Odds ratio = 1 .2 (0.5, 3.3) Odds ratio = 6.1 (1.0, 39.3) P-interaction = 0.20 Odds ratio = 7.4 (0.6, 93.9) Odds ratio = 0.8 (0.2, 2.8) Odds ratio = 0.8 (0.1, 4.3) P-interaction = 0.44 Odds ratio = 1 .0 (0.1, 8.2) Odds ratio = 0.8 (0.2, 2.6) Odds ratio = 1.2 (0.2, 7.7) P-interaction = 0.97 Odds ratio = 3.7 (0.5, 30.8) Odds ratio = 2.5 (0.3, 25.0) Odds ratio = 1.8 (0.3, 11.8) P-interaction = 0.48 Age at assessment, sex, parity, breastfeeding duration, Infant-Toddler Home Observation for Measurement of the Environment score, maternal Peabody Picture Vocabulary Test score, household poverty status, and maternal depression Interactions were statistically nonsignificant whether genotype was coded as a categorical or ordinal variable Associations were "similar, albeit weaker" when stratified bv maternal genotype (not shown here) " - " - " Associations across maternal PONi enzyme levels and activities at delivery were "similar to those for cord blood enzyme levels" (not shown here) Odds ratio = 13.2 (1.4, 128.3) Odds ratio = 0.4 (0.0, 3.9) Odds ratio = 4.7 (0.5, 41.6) P-interaction = 0.88 Tier 3/4 CritRevTcocoL 2015; 45(7): 531-641 ED 002061 00046431-00068 V DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Eskenazi et al. (2010) Eskenazi et ai. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) m m i m Child Behavior Checklist attention problems borderline at 3.5 years Maternal prenatal urinary DMPs (mnoi/L, log,0 scale) by cord blood PON 1 quantity or activity PONI quantity: 89 tettile 1 88 tettile 2 88 tortile 3 PONI activity: 91 tettile 1 85 tettile 2 87 tortile 3 Maternal prenatal urinary DEPs (nmol/L, log,,, scale) by cord blood PONi quantity or activity PONI quantity: 89 tettile 1 88 tettile 2 88 tortile 3 PONI activity: 91 tettile 1 85 tettile 2 87 tortile 3 Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,0 scale) 17/330 (5.2%) total 12/151 (7.9%) boys 5/179 (2.8%) girls Odds ratio ==3.1 (0.5, 20.2) Odds ratio = 2.8 (0.3, 24.3) Odds ratio = 2.0 (0.4, 11.4) P-interaction - 0.43 Odds ratio = 7.3 (0.9, 56.9) Odds ratio = 0.8 (0.1, 6.2) Odds ratio --4.5 (0.7, 30.4) P-interaction - 0.90 Odds ratio --1.7 (0.3, 11.4) Odds ratio - 1.3 (0.2, 10.6) Odds ratio = 0.9 (0.1, 8.2) P-interaction = 0.24 Odds ratio = 7.0 (0.8, 58.0) Odds ratio - 0.4 (0.1, 3.1) Odds ratio = 0.7 (0.1,6.6) P-interaction = 0.1 5 DAPs odds ratio = 3.0 (0.7, 11.7) DAPs odds ratio, boys - 4.1 (0.8, 22.2) DAPs odds ratio, girls --2.1 (0.2, 29.9) P-interaction by sex = 0.68 DMPs odds ratio - 3.2 (0.9, 11.3) DEPs odds ratio ^ 2.1 (0,6, 7.0) Child urinary DAPs, DMPs, or DEPs (nmol/L, log,,, Scale) 17/289 (5.9A> DAPs odds ratio = 1.6 (0.8, 3.5) DMPs odds ratio --1.6 (0.8, 3.3) DEPs odds ratio ^ 1.9 (0.9, 3.9) Child Behavior Checklist attention problems continuous score at 3.5 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,0 scale) 330 total 151 boys 179 girls DAPs beta --0.3 (-0.2, 0.7) DAPs beta, boys = 0.7 (0.0, 1.4) DAPs beta, girls - -0.1 (-0.7, 0.5) P-interactionby sex = 0.05 DMPs beta = 0.3 (-0.1, 0.7) DEPs beta = 0.0 (-0.5, 0.4) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels "Borderline" score > 93rd percentile "Clinical" score > 97thpercentile (Continued) Tier 3/4 Low-O wl OP iSECtdE E x p o SUEES 59 9 ED 002061 00046431-00069 V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. (Continued ) Reference Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) m m i ; m Outcome Child Behavior Checklist ADHD borderline at 3.5 years Child Behavior Checklist ADHD continuous score at 3.5 years Exposure Child urinary DAPs, DMPs, or DEPs (nmoi/L, iogjQ scale) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log,0 scale) Child urinary DAPs, DMPs, or DEPs (nmol/L, !ogif) scale) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, login scale) Child urinary DAPs, DMPs, or DEPs (nmol/L, log,0 scale) Number of subjects/events 289 18/329 (5.5%) total 12/151 (7.9%) boys 6/176 (3.4%) girls 17/288 (5.9%) 329 total 151 boys 176 girls 288 Estimate of association (95% Cl) DAPs beta = 0.1 (-0.2, 0.4) DMPs beta = 0.1 (-0.2, 0.3) DEPs beta = 0.2 (0.0, 0.5) DAPs odds ratio = 3.1 (0.8, 11.5) DAPs odds ratio, boys = 6.4 (1.1, 39.0) DAPs odds ratio, girls - 1.0 (0.1, 11.2) P-interaction by sex = 0.21 DMPs odds ratio = 1.3 (0.4, 4.4) DEPs odds ratio = 2.8 (0.9, 8.9) DAPs odds ratio = 1.4 (0.7, 3.1) DMPs odds ratio = 1.4 (0.7, 3.0) DEPs odds ratio = 1.0 (0.5, 2.2) DAPs beta = 0.5 (-0.3, 1.3) DAPs beta, boys = 1.3 (0.1, 2.5) DAPs beta, girls = -0.2 (-1.2, 0.8) P-interaction by sex = 0.06 DMPs beta = 0.6 (-0.1, 1.3) DEPs beta = -0.2 (-0.9, 0.6) DAPs beta = 0.1 (-0.3, 0.6) DMPs beta = 0.1 (-0.3, 0.6) DEPs beta = 0.2 (-0.3, 0.7) Adjustment factors Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal, depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Comments Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychornetrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychornetrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs "Borderline" score > 93rd percentile "Clinical" score > 97th percentile - C ritR evToe col. 2015; 45(7): 531-641 ED 002061 00046431-00070 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) m m i ; m NEPSY-II visual attention continuous score at 3.5 years Maternal prenatal urinary DAPs, DMPs, or DEPs (mnol/L, log,,, scale) 319 total 143 boys 176 girls Child urinary DAPs, 'I'll DMPs, or DEPs (nmoi/L, log,,, Scale) DAPs beta = 0.2 (-0.5, 0.8) DAPs beta, bovs --0.2 (-0.8. 1.1) DAPs beta, girls - 0.2 (-0.7, 1.2) P-interaction by sex --0.99 DMPs beta - oil (-0.5. 0.6) DEPs beta = -0.2 (-0.8, 0.5) DAPs beta = -0.1 (-0.5, 0.3) DMPs beta = -0.1 (-0.5, 0.3) DEPs b e t a - -0.1 (-0.5, 0.3) Child Behavior Checklist attention problems borderline at 5 years Maternal prenatal urinary DAPs, DMPsl or DEPs (nmoi/L, log10 scale) 13/322 (4.0%) total 10/154 (6.5%) boys 3/168 (1.8%) girls DAPs odds ratio - 0.8 (0.2, 3.8) DAPs odds ratio, boys --1.0 (0.2, 6.0) DAPs odds ratio, girls - 0.6 (0.0, 17.3) P-Interaction by sex - 0.77 DMPs odds ratio --2.0 (0.5, 8.5) DEPs odds ratio - 0.7 (0.2, 2.8) Child urinary DAPs, DMPs, or DEPs (nmoi/L, log,0 scale) 13/319 (4.1%) DAPs odds ratio --1.0 (0.4, 2.4) DMPs odds ratio --0.9 (0.4, 2.1) DEPs odds ratio --1.8 (0.8, 3.9) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs NEPSY-II visual attention subtest is scaled to an age-standardized mean SD of 10 3 "Borderline" score > 93rd percentile "Clinical" score > 97th percentile (Continued) Low-lcwl OP iSECtdE E x p o SUEES 601 ED 002061 00046431-00071 602 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Outcome Child Behavior Checklist attention problems continuous score at 5 years Child Behavior Checklist ADHD borderline at 5 years Exposure Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log,0 scale) Child urinary DAPs, DMPs, or DEPs (mnol/L, log,0 scale) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,0 scale) Child urinary DAPs, DMPs, or DEPs (nmoi/L, log.,, scale) Number of subjects/events 322 total 154 boys 168 girls 319 23/322 (7.1%) total 14/154 (9.1%) boys 9/168 (5.4%) girls 22/319 (6.9%) Estimate of association (955 Cl) Adjustment factors DAPs beta = 0.7 (0.2, 1.2) DAPs beta, boys = 0.9 (0.2, 1.7) DAPs beta, girls = 0.4 (-0.2, 1.0) P-interaction by sex = 0.28 DMPs beta = 0.6 (0.2, 1.0) DEPs beta = 0.4 (-0.1, 0.9) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score DAPs beta = 0.0 (-0.3, 0.2) DMPs beta = -0.1 (-0.3. 0.2) DEPs beta = 0.0 (-0.2, 0.3) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score DAPs odds ratio =1.1 (0.3. 3.5) DAPs odds ratio, boys = 4.9 (0.7, 33.0) DAPs odds ratio, girls = 0.3 (0.0, 2.2) P-interaction by sex = 0.1 8 DMPs odds ratio = 1.3 (0.4, 4.0) DEPs odds ratio = 1.1 (0.4, 3.2) DAPs odds ratio = 0.6 (0.3, 1.2) DMPs odds ratio = 0.5 (0.3, E l) DEPs odds ratio = 0.9 (0,5, 1.7) Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Comments ... "Borderline" score > 93rd percentile "Clinical" score > 97th percentile m m i ; m Tier 3/4 Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00072 last Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) m m i ; m Child Behavior Checklist ADHD continuous score at 5 years Conners markedly atypical % omissions at 5 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log10 scale) Child urinary DAPs, DMPs, or DEPs (nmol/L, 1og,0 scale) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,(l scale) Child urinary DAPs, DMPs, or DEPs (nmol/L, log,fl scale) 322 total 154 boys 68 girls DAPs beta --1.3 (0.4, 2.1) DAPs beta, boys = 1.9 (0.6, 3.2) DAPs beta, girls = 0.6 (-0.5, 1.6) P-interaction by sex = 0.13 DMPs beta --1.1 (0.3, 1.9) DEPs beta --0.7 (-0.2, 1.5) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinine- adjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels 319 DAPs beta - 0.0 (-0.5, 0.5) Maternal total DAPs, DMPs beta - 0.0 (-0,5, 0.4) psychometrician, age at assessment, DEPs beta = 0.1 (-0.3, 0.6) sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 59/312 (18.9%) total 21/148 (14.2%) boys 38/164 (23.2%) girls DAPs odds ratio = 1.5 (0.7, 3.3) DAPs odds ratio, boys - 1.7 (0,4, 6.4) DAPs odds ratio, girls --1.4 (0.5, 4.0) P-interaction by sex --0.90 DMPs odds ratio = 1.9 (0.9, 4.1) DEPs odds ratio = 1.3 (0.6, 2.8) Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 58/309 (18.8%) DAPs odds ratio = 1.0 (0.6, 1.6) DMPs odds ratio = 0.9 (0.6, 1.5) DEPs odds ratio = 1.5 (1.0, 2.2) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead level s Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Conners' Kiddie Continuous Performance Test is scaled to an agestandardized mean SD of 50 10, with score > 65 considered "markedly atypical" (Continued) Vat Low-lVCl OP IHSECtide EXpOSUTES 603 ED 002061 00046431-00073 ED 002061 00046431-00074 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Marks et al. (2010) Outcome Conners markedly atypical % commissions at 5 years Marks et al. (2010) Marks et al. (2010) Conners markedly atypical hit reaction time at 5 years Marks et al. (2010) m m i ; m Exposure Maternal prenatal urinary DAPs, DMPs, or DEPs (nnrol/L, log,0 scale) Child urinary DAPs, DMPs, or DEPs (nmoi/L, log,,, scale) Number of subjects/events Estimate of association (95% Cl) Adjustment factors 54/312 (17.3%) 24/148 (14.2%) boys 30/164 (18.3%) girls DAPs odds ratio = 1.0 (0.5, 2.2) DAPs odds ratio, boys = 0.9 (0.2, 3,2) DAPs odds ratio, girls = 1.2 (0.4, 3.3) P-interaction by sex = 0.89 DMPs odds ratio = 1.2 (0.6, 2.7) DEPs odds ratio = 0.8 (0.4, 1.6) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test 53/309 (17.2%) DAPs odds ratio =1.1 (0.7, 1.7) DMPs odds ratio =1.1 (0.7, 1,8) DEPs odds ratio = 0.9 (0.6, 1.4) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log,,, scale) Child urinary DAPs, DMPs, or DEPs (nmoi/L, log,,, scale) 20/311 (6.4%) total 7/147 (4.8%) boys 13/164 (7.9%) girls 19/308 (6.2%) Results were similar with creatinine- adjusted DAPs and without adjustment for maternal DAPs DAPs odds ratio = 1.6 (0.5, 5.2) Psychonretrician, age at assessment, DAPs odds ratio, boys = 1.2 (0.1, 11.5) sex, child care, breast-feeding, DAPs odds ratio, girls = 1.7 (0.4, 7.4) maternal education, maternal P-interaction by sex = 0.72 depressive symptoms, and maternal DMPs odds ratio =1.1 (0.3, .3.6) Peabody Picture Vocabulary Test DEPs odds ratio = 1.5 (0.5, 4.6) score DAPs odds ratio =1.1 (0.5, 2.3) DMPs odds ratio = 1.0 (0.5, 2.0) DEPs odds ratio = 1.3 (0.7, 2.4) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychonretrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs ^4 Comments CritRevToocoL 2015; 45(7): 531-641 604 R. Reiss et al. Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Marks et al. (2010) Marks et ai. (2010) Marks et al. (2010) Marks et al. (2010) m m i ; m ADHD Confidence Index > 70th percentile at 5 years ADHD Confidence Index continuous score al 5 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/'L, login scale) Child urinary DAPs, DMPs, or DEPs (nrnol/L, log,0 scale) Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log,,, Scale) Child urinary DAPs, DMPs, or DEPs (mnol/L, log,0 scale) 25/297 (8.4%) total 14/140 (10.0%) boys 11/157 (7.0%) girls 24/294 (8.2%) DAPs odds ratio = 5,1 (1.7, 15.7) DAPs odds ratio, boys --10.1 (1.6,65.3) DAPs odds ratio, girls = 3.3 (0.6, 17.0) P-interaction by sex - 0.41 DMPs odds ratio = 6.6 (2.2, 19,3) DEPs odds ratio = 3.2 (1.2. 8.9) DAPs odds ratio - 1.3 (0.7, 2.5) DMPs odds ratio = 1.2 (0.7, 2.3) DEPs odds ratio = 1.5 (0.8, 2,8) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 297 total 140 boys 157 girls DAPs beta --3.4 (-1.8, 8.7) DAPs beta, boys = 6.3 (-0.5, 13,3) DAPs beta, girls = 0,5 (-7.2, 8,3) P-interaction by sex = 0.39 DMPs beta = 0 (-2.8. 6.9) DEPs beta = 3.4 (-1.7, 8.6) Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinine- adjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels 294 DAPs beta = -0.7 (-3.8, 2.3) Maternal total DAPs, DMPs beta = -1.0 (-3.9, 1.9) psychometrician, age at assessment, DEPs beta = 2.2 (-0.5, 5.0) sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs ADHD Confidence Index score on Conners' Kiddie Continuous Performance Test is scaled to a range of 0-100, with > 70th percentile considered as clinical ADHD __ (Continued ) Low-lcwl OP iSECtdE E x p o SUEES 605 ED 002061 00046431-00075 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 606 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Table 3. (Continued ) Reference Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Marks et al. (2010) Outcome Hillside Behavior Rating Scale Attention sc 7 of 12 at 5 years Exposure Maternal prenatal urinary DAPs, DMPs, or DEPs (nrnol/L, log,0 scale) Child urinary DAPs, DMPs, or DEPs (nrnol/L, log,0 scale) Composite ADHD indicator at 5 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log,,, Scale) Child urinary DAPs, DMPs, or DEPs (nmol/L, iog,0 scale) Number of subjects/events Estimate of association (95% Cl) Adjustment factors 23/322 (7.1%) total 14/153 (9.2%) boys 9/169 (5.3%) girls DAPs odds ratio = .3.0 (0.9, 9.8) DAPs odds ratio, boys = 7,9 (1.4, 46.0) DAPs odds ratio, girls = 1.0 (0.2, 5.9) P-interaction by sex = 0.14 DMPs odds ratio = 2.3 (0.7, 7.4) DEPs odds ratio = 2.9 (1.0, 8.5) Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 23/319 (7.2%) DAPs odds ratio = 1.4 (0,7, 2.8) DMPs odds ratio = 1.1 (0.6, 2,1) DEPs odds ratio = 1.4 (0.8, 2.6) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 27/319 (8.5%) total 19/150 (12.7%) boys 8/169 (4.7%) girls DAPs odds ratio = 3.5 (1.1, 10.7) DAPs odds ratio, boys =11.1 (1.8,66.5) DAPs odds ratio, girls = 1.1 (0.2, 7.1) P-interaction by sex = 0.13 DMPs odds ratio = 1.7 (0.5, 5.5) DEPs odds ratio = 3.0 (1.1, 8.2) Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score 25/316 (7.9%) DAPs odds ratio = 1.0 (0.5, 2.0) DMPs odds ratio = 0.8 (0.4, 1.5) DEPs odds ratio = 2.0 (1.1, 3.6) Results were similar with creatinineadjusted DAPs and after additional adjustment for birth weight, gestational age, and blood lead levels Maternal total DAPs, psychometrician, age at assessment, sex, child care, breast-feeding, maternal education, maternal depressive symptoms, and maternal Peabody Picture Vocabulary Test score Results were similar with creatinineadjusted DAPs and without adjustment for maternal DAPs Comments Hillside Behavior Rating Scale score is scaled to a range of 0-12, with score a 7 (< 10% of children) considered as displaying "a higher degree of attention problems" Composite ADHD indicator is based on at least two of the following: Child Behavior Checklist ADHD scale = borderline range, Conners' Kiddie Continuous Performance Test ADHD Confidence Index s- 60%, and Hillside ADHD scale >75% m m i ; m Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00076 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Bouchard et al. (2011) Wechsler working memory at 7 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmoi/L, log.,, scale) 267 first half 279 second half 298 averaged Bouchard et al. (2011) Bouchard et al. (2011) Child urinary DAPs (nrnol/L, log,,, scale) 265 at 6 months 274 at 12 months 274 at 24 months 231 at 42 months 273 at 60 months 245 at all ages Wechsler processing speed at 7 years Maternal prenatal urinary DAPs, DMPs", or DEPs (nmol/L, log,,, scale) 268 first half 280 second half 298 averaged Bouchard et al. (2011) Child urinary DAPs 266 at 6 months (nmoi/L, log,,, 274 at 12 months scale) 274 at 24 months 231 at 42 months 273 at 60 months 246 at all ages DAPs beta, first half of pregnancy = -1.6 (-4.2, 1.0) DAPs beta, second half of pregnancy = -3 .0 (-6.4, 0.4) DAPs beta, pregnancy average ---4.3 (-7.7, -0.9) DMPs beta, pregnancy average = -4 .0 (-7.1, -1.0) DEPs beta, pregnancy average - -0,4 (-3.5, 2.7) B e t a - -1.7 (-3.9, 0.5) Beta - 0 .9 (-1.4, 3.2) B e t a - -0.4 (-2.7, 1.9) B eta- 0 .8 (-1.7, 3.3) Beta-2 .0 (-0.1,4.0) Beta for area under curve --1.6 (-2.2, 5.4) DAPs beta, first half of pregnancy ---1.5 (-3.9, 0.9) DAPs beta, second half of pregnancy - -2.6 (-5.9, 0.7) DAPs beta, pregnancy average ---3.4 (- 6 .8, - 0. 1) DMPs beta, pregnancy average - -1.8 (-4,8, 1.2) DEPs beta, pregnancy average ---4.0 (-7.0, -1.0) B e t a - -0.3 (-2.5, 1.8) Beta - 1.6 (-0.6, 3.8) Beta - -2.0 (-4.3, 0.2) Beta ---1.1 (--3,6, 1.3) B eta- 0 .7 (-1.3, 2.7) Beta for area under curve --- .3 (-4.9, 2.3) Infant -Toddler Home Observation for Measurement of the Environment score at 6 months, maternal education, and maternal intelligence No difference after additional adjustment for maternal levels of polybrominated diphenyl ethers, polychlorinated biphenyls, dichiorodiphenyitrichloroethane/ dichlorodiphenyldichloroethylene, and lead, prenatal DAPs (in analyses of child DAPs), use of creatinineadjusted DAPs, stratification by sex, or restriction to children tested in Spanish Working memory --Digit Span and Letter-Number Sequencing subtests; scores standardized against U.S. population-based norms for Englishand Spanish-speaking children Estimates of association did not differ significantly (P - 0.10) between prenatal and postnatal DAP concentrations Area under the curve --cumulative DAP level between 6 and 60 months Processing speed --Coding and Symbol Search subtests; scores standardized against U.S. population-based norms for English- and Spanish-speaking children Estimates of association did not differ significantly (P --0.24) between prenatal and postnatal DAP concentrations; interaction term between mean prenatal DAP level and AUC was not statistically significant (P > 0 .1 5 ) (Continued) m m i ; m ^4 Low-lcwl OP iSECtdE E x p o SUEES 607 ED 002061 00046431-00077 608 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Tables. (Continued) Reference Bouchard et al. (201) Outcome Wechsler verbal comprehension at 7 years Exposure Maternal prenatal urinary DAPs, DMPs, or DEPs (mnol/L, log,^ scale) Number of subjects/events 291 first half 309 second half 329 averaged Bouchard et al. (2011) Bouchard et al. (2011) Child urinary DAPs (nmol/L, logl0 scale) 294 at 6 months 303 at 12 months 303 at 24 months 259 at 42 months 302 at 60 months 271 at all ages Wechsler perceptual reasoning at 7 years Maternal prenatal urinary DAPs, DMPs, or DEPs (nmol/L, log,,, Scale) 292 first half 309 second half 329 averaged Bouchard et al. (2011) m a 1 HI m Child urinary DAPs (nmol/L, log,,, Scale) 294 at 6 months 303 at 12 months 303 at 24 months 259 at 42 months 302 at 60 months 271 at all ages Estimate of association (95% Cl) Adjustment factors Comments DAPs beta, first half of pregnancy ---2.6 (-5.1, -0.1) DAPs beta, second half of pregnancy ---3.1 (-6.4, 0.2) DAPs beta, pregnancy average ---5.3 (--8.6, --2.0) DMPs beta, pregnancy average = -4 .8 (-7.8, -1.9) DEPs beta, pregnancy average - -2.0 (-5.0, 1.1) Beta = 0.8 (-1.4, 3.0) Beta = 2.9 (0.7, 5.2) Beta = -0.8 (-3.1, 1.5) Beta = 0.2 (-2.2, 2.6) Beta = 0.4 (-1.6, 2.5) Beta for area under curve = 0.8 (-3.0, 4.6) DAPs beta, first half of pregnancy = -1.2 (-4.1, 1.7) DAPs beta, second half of pregnancy = -2 .4 (-6.3, 1.4) DAPs beta, pregnancy average = -4,0 (-7.9,-0.1) DMPs beta, pregnancy average = -3.3 (-6.7, 0.2) DEPs beta, pregnancy average = -2.1 (-5.6, 1.5) Beta = -2.4 (-4.9, 0.1) Beta = 1.9 (-0.8, 4.5) Beta = -0.7 (-3.4, 2.0) Beta = -0.3 (-3.0, 2.5) Beta = 2.3 (-0.1, 4.7) Beta for area under curve = 0.5 (-3.8, Infant-Toddler Home Observation for Measurement of the Environment score, maternal education, maternal intelligence, and language of assessment No difference after additional adjustment for maternal levels of polybrominated diphenyl ethers, polychlorinated biphenyls, dichlorodiphenyitrichloroethane/ dichSorodiphenyidichSoroethylene, and lead, prenatal DAPs (in analyses of child DAPs), use of creatinineadjusted DAPs, stratification by sex, or restriction to children tested in Spanish Infant-Toddler Home Observation for Measurement of the Environment score, maternal education, and maternal intelligence No difference after additional adjustment for maternal levels of polybrominated diphenyl ethers, polychlorinated biphenyls, dichlorodiphenyltrichloroethane/ dichlorodiphenyldichloroethylene, and lead, prenatal DAPs (in analyses of child DAPs), use of creatinineadjusted DAPs, stratification by sex, or restriction to children tested in Spanish Verbal comprehension = Vocabulary and Similarities subtests; scores standardized against U.S. populationbased norms for English- and Spanish speaking children Estimates of association differed significantly (P = 0.01) between prenatal and postnatal DAP concentrations Perceptual reasoning = Block Design and Matrix Reasoning subtests; scores standardized against U.S. populationbased norms for English- and Spanishspeaking children Estimates of association did not differ significantly (P = 0.19) between prenatal and postnatal DAP concentrations 4-g) Tier 3/4 CritRevToe dcoL 2015; 45(7): 531-641 ED 002061_00046431-00078 v DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Bouchard et al. (2011) Wechsler full-scale intelligence quotient at 7 years Maternal prenatal urinary DAPs, DMPs", or DEPs (nmol/L, log,0 scale) 266 first half 279 second half 297 averaged Bouchard et al. (2011) Quiros-Alcala "et al. (2011) Child urinary DAPs (nrnol/L, log,0 scale) 265 at 6 months 273 at 12 months 273 at 24 months 231 at 42 months 272 at 60 months 245 at all ages Respiratory sinus arrhythmia, resting (index) at 6 months and 1,3.5, and 5 years Maternal prenatal urinary DAPs (nrnol/L, log,0 scale) 142 at 6 months 149 at 1 year 95 at 3.5 years 270 at 5 years Quiros-Alcaia et ai . (2011) Quiros-Alcaia "et al. (2011) Quiros-Alcaia tal. (2011) Quiros-Alcaia et al. (2011) " Maternal prenatal 142 at 6 months urinary DMPs 149 at 1 year (mnol/L, log,,, 95 at 3.5 years Scale) 270 at 5 years " Maternal prenatal 142 at 6 months urinary DEPs 149 at 1 year (nrnol/L, log,0 95 at 3.5 years scale) 270 at 5 years Child urinary DAPs 142 at 6 months (nrnol/L, log10 149 at 1 year scale) 95 at 3.5 years 270 at 5 years " Child urinary DMPs 142 at 6 months (mnol/L, log,0 149 at 1 year scale) 95 at 3.5 years 270 at 5 years DAPs beta, first half of pregnancy = -2.4 (-4.9, 0.2) DAPs beta, second half of pregnancy = -3.5 (-6.9, -0.1) DAPs beta, pregnancy average ---5.6 (-9.0, -2.2) DMPs beta, pregnancy average = -4 .7 (-7.7, -1.6) DEPs beta, pregnancy average = -2.8 (-5.6, 0.3) Beta = -0.9 (-3.2, 1.3) Beta = 2.7 (0.3, 5.1) Beta - -1.5 (-3.9, 0.9) Beta = 0.2 (-2.4, 2.8) Beta = 1.7 (-0.4, 3.9) Beta for area under curve = 0.6 (-3.2, 4.4) Beta - -0.05 (-0.33, 0.24) Beta = -0.11 (-0.43,0.21) Beta = 0.19 (-0.35, 0.73) Beta 0.14 (-0.22, 0.49) Beta = -0.07 (-0.34, 0.19) Beta = -0.10 (-0.39, 0.20) Beta 0.13 (-0.39, 0.64) Beta = 0.02 (-0.30, 0.34) Beta - -0.06 (-0.37, 0.25) Beta -0.13 (-0.47,0.21) Beta = 0.22 (-0.30, 0.74) Beta = 0.17 (-0.17, 0.51) Beta = -0.27 (-0.48, -0.06) Beta = -0.06 (-0.28, 0.16) Beta = -0.13 (-0.46, 0.20) Beta = -0.11 (-0.34,0.12) Beta = -0.24 (-0.42, -0.05) Beta = -0.06 (-0.26, 0.13) Beta = -0.15 (-0.46,0.17) Beta - -0.13 (-0.34,0.09) Infant-Toddler Eiome Observation for Measurement of the Environment score, maternal education, maternal intelligence, and language of assessment No difference after additional adjustment for maternal levels of polybrominated diphenyl ethers, polychlorinated biphenyls, dichiorodiphenyitrichloroethane/ dichlorodiphenyldichloroethylene, and lead, prenatal DAPs (in analyses of child DAPs), use of creatinineadjusted DAPs, stratification by sex, or restriction to children tested in Spanish Estimates of association differed significantly (P -----0.03) between prenatal and postnatal DAP concentrations Sex, exact age at assessment, Resting conditions at 6 months and breast-feeding duration, location of 1 year: listening to digitally recorded assessment, psychometrician, and both lullabies prenatal and child DAPs Resting conditions at 3.5 and 5 years: Results based on creatinine-adjusted listening to a story read aloud metabolite levels were "similar ... although some associations were attenuated" (results NR) For resting measures, the "only significant association in both the unadjusted and creatinine-adjusted models was for child [DEP] concentrations and high [pre election period] testing measures (less sympathetic activation) in 1-year-olds." (Continued) Tier 3/4 Low-lcwl OP USECtdE CXpOSUTES 609 ED 002061 00046431-00079 v ED 002061 00046431-00080 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Outcome " Exposure Child urinary DEPs (nmoi/L, log10 scale) Heart rate, resting (beats per minute) at 6 months and 1, 3.5, and 5 years Maternal prenatal urinary DAPs (nmol/L, log,0 scale) Number of subjects/events 142 at 6 months 149 at 1 year 95 at 3.5 years 270 at 5 years 143 at 6 months 147 at 1 year 95 at 3.5 years 271 at 5 years Estimate of association Beta = -0.1.3 (-0.34, 0.09) Beta = -0.0.3 (-0.29, 0.24) Beta = -0.05 (-0.37, 0.27) Beta = 0.03 (-0.17, 0.23) Beta = -1.09 (-4.96,2.78) Beta = -2.50 (-6.72, 1.73) Beta ---3.82 (-8. /4, 1.10) Beta = 1.36 (-1.89, 4.60) Adjustment factors Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et!. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) m Quiros-Alcala m et!. (2011) T " " " " Pre-ejection period, resting (milliseconds) at 6 months and 1, 3.5, and 5 years " Maternal prenatal urinary DMPs (nmoi/L, log.,, Settle) Maternal prenatal urinary DEPs (nmol/L, log,0 scale) Child urinary DAPs (nmol/L, log,0 scale) Child urinary DMPs (nmol/L, iog,0 scale) Child urinary DEPs (nmol/L, login scale) Maternal prenatal urinary DAPs (nmol/L, log,0 Settle) Maternal prenatal urinary DMPs (nmol/L, login scale) Maternal prenatal urinary DEPs (nmoi/L, log10 scale) 143 at 6 months 147 at 1 year 95 at 3.5 years 271 at 5 years 143 at 6 months 147 at 1 year 95 at 3.5 years 271 at 5 years 143 at 6 months 147 at 1 year 95 at 3.5 years 27! at 5 years 143 at 6 months 147 at 1 year 95 at 3.5 years 271 at 5 years 143 at 6 months 147 at 1 year 95 at 3,5 years 271 at 5 years 136 at 6 months 141 at 1 year 94 at 3.5 years 269 at 5 years 136 at 6 months 141 at 1 year 94 at 3.5 years 269 at 5 years 136 at 6 months 141 at 1 year 94 at 3,5 years 269 at 5 years Beta = -1.19 (-4.71,2.33) Beta = -2.48 (-6.39, 1.43) Beta = -3.11 (-7.81, 1.60) Beta = 1.96 (-0.93, 4.85) Beta = -0.38 (-4.51,3.75) Beta = -1.63 (-6.12, 2.86) Beta = -3.78 (-8.51,0.95) Beta = -0.77 (-3.87,2.33) Beta = 1.39 (-1.43. 4.21) Beta = 0.38 (-2.58, 3.34) Beta = 2.17 (-0.83, 5.16) Beta = -1.14 (-3.21, 0.93) Beta = 1.41 (-1.11, 3.94) Beta = 0.44 (-2.14, 3.02) Beta = 2.28 (-0.58, 5.14) Beta = -1.1.3 (-3.09,0.84) Beta = 0.62 (-2.23, 3.47) Beta = -0.02 (-3.57, 3.53) Beta = 1.58 (-1.38, 4.5.3) Beta = -0.14 (-1.92, 1.65) Beta = -0.67 (-4.11, 2.76) Beta = 3.72 (-0.09, 7.5.3) Beta = 1.27 (-1.92, 4.47) Beta = -0.86 (-3.11, 1.39) Beta = -0.77 (-3.90,2.35) Beta = 3.77 (0.21, 7.33) Beta = 1.04 (--2.01, 4.09) Beta = -1.18 (-3.18,0.83) Beta = 1.25 (-2.46, 4.96) Beta = 2.74 (-1.20, 6.68) Beta = 1.03 (-2 .0 1 ,4 0 7 ) Beta = 0.39 (-1.75, 2.53) Comments No significant associations were found between cumulative measures of prenatal or childhood metabolite levels (based on area under the concentrationtime curve calculations) and resting or reactive measures at age 5 years, except between creatinine-unadjusted cumulative prenatal DEP levels and resting heart rate (beta = -3.19 -6.29. -0.09]) - - - ... __ - ... .... CritRevToocoL 2015; 45(7): 531-641 610 R. Reiss et al. Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Quiros-Aicala et al. (2011) Quiros-Aicala et al. (2011) Quiros-Aicala et al. (2011) Quiros-Aicala tal. (2011) " " Respiratory sinus arrhythmia, reactive (index) at 6 months and 1, 3.5, and 5 years Child urinary DAPs (nmol/L, log,0 scale) Child urinary DMPs (nnrol/L, log,0 scale) Child urinary DEPs (mnol/L, log,,, scale) Maternal prenatal urinary DAPs (nmoi/L, log10 scale) 136 at 6 months 141 at 1 year 94 at 3.5 years 269 at 5 years 136 at 6 months 141 at 1 year 94 at 3.5 years 269 at 5 years 136 at 6 months 141 at i year 94 at 3.5 years 269 at 5 years 141 at 6 months 147 at 1 year 95 at 3,5 years 270 at 5 years Beta = -0.31 (-2.83,2.22) Beta = 1.27 (-1.39, 3.92) Beta = 0.57 (-1.38, 2.51) Beta - 0.35 (-1.09, 1.79) Beta ---0.06 (-2.27, 2.16) Beta = 0.34 (-1,99, 2.68) Beta --0.74 (-1.11, 2.59) Beta = 0.27 (-1.10, 1.63) Beta = -0.59 (-3.15, 1.98) Beta --4.33 (1.24, 7.42) Beta = -0.96 (--2.87, 0.94) Beta = 0.70 (-0.53, 1.93) Beta ---0.17 (-0.36, 0.03) Beta = 0.24 (0.03, 0.46) Beta = 0.06 (-0.23, 0.34) Beta - -0.08 (-0.25, 0.08) Quiros-Aicala et ai. (2011) Quiros-Aicala ~et al. (2011) Quiros-Aicala et al. (2011) Quiros-Aicala et al. (2011) Quiros-Aicala tal. (2011) m i m " Maternal prenatal 141 at 6 months Beta = -0.15 (-0.33, 0.03) urinary DMPs 147 at 1 year Beta = 0.25 (0.05, 0.45) (nmoi/L, log,,, 95 at 3.5 years Beta --0.07 (-0.21, 0.34) Scale) 270 at 5 years Beta - -0.04 (-0.18, 0.11) " Maternal prenatal 141 at 6 months Beta - -0.09 (-0.31, 0.12) urinary DEPs 147 at 1 year Beta - 0.01 (-0.22, 0.24) (nmoi/L, log,0 95 at 3.5 years Beta = 0.02 (-0.25, 0.29) scale) 270 at 5 years Beta ---0.01 (-0.17,0.14) Child urinary DAPs 141 at 6 months Beta - -0.10 (-0.24, 0.05) (nmoi/L, log,0 147 at 1 year Beta = 0.01 (-0.14, 0.16) scale) 95 at 3.5 years Beta ---0.03 (-0.20,0.14) 270 at 5 years Beta = 0.06 (-0.04, 0.16) " Child urinary DMPs 141 at 6 months Beta = -0.08 (-0.21,0.05) (nmol/L, iog,0 147 at 1 year Beta ---0.01 (-0.14,0.12) scale) 95 at 3.5 years Beta = -0.02 (-0.19, 0.14) 270 at 5 years Beta - 0.06 (-0,04, 0.16) " Child urinary DEPs 141 at 6 months Beta = 0.00 (-0.14, 0.15) (nmol/L, log,0 147 at 1 year Beta = 0.09 (-0.09, 0.27) scale) 95 at 3,5 years Beta = 0.01 (-0.16,0.18) 270 at 5 years Beta = -0.02 (-0.11,0.07) " ... " - " ... Sex, exact age at assessment. breast-feeding duration, location of assessment, psychometrician, and both prenatal and child DAPs "Associations between reactivity scores and creatinine-adjusted prenatal [DMP] and DAP levels were similar" (results NR) Challenging conditions at 6 months and 1 year: watching a jack-in-the-box wound up and jumping out of the box (social/startle), listening to a digitally recorded sick baby crying (emotion), and feeling a vibrator on the leg (physical) Challenging conditions at 3.5 and 5 years: answering questions (social), watching a scary video clip (emotion), tasting concentrated lemon juice on the tongue (physical), and repeating a series of numbers (cognitive) - - - ... __ (Continued) Low-lcwl OP USECtdE E x p o SUEES 611 ED 002061 00046431-00081 ED 002061 00046431-00082 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala et al. (2011) Quiros-Alcala (g tal. (2011) S "g Quiros-Alcala 1 tal. (2011) m Outcome Heart rate, reactive (beats per minute) at 6 months and 1, 3.5, and 5 years " " " " Pre-ejection period, reactive (milliseconds) at 6 months and 1,3.5, and 5 years " " " Exposure Maternal prenatal urinary DAPs (nmol/L, log,,, scale) Maternal prenatal urinary DMPs (nmol/L, log,,. scale) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) Child urinary DAPs (nmol/L, log,,, Scale) Child urinary DMPs (nmol/L, log,0 scale) Child urinary DEPs (nmol/L, log,,. scale) Maternal prenatal urinary DAPs (nmol/L, log,,, scale) Maternal prenatal urinary DMPs (nmol/L, log,,. scale) Maternal prenatal urinary DEPs (nmol/L, log,,, scale) Child urinary DAPs (nmol/L, log,0 scale) Child urinary DMPs (nmol/L, log,,, scale) Child urinary DEPs (nmol/L, log,,, scale) Number of subjects/events 142 at 6 months 145 at 1 year 95 at 3.5 years 27! at 5 years 142 at 6 months 145 at 1 year 95 at 3,5 years 271 at 5 years 142 at 6 months 145 at 1 year 95 at 3.5 years 271 at 5 years 142 at 6 months 145 at 1 year 95 at 3.5 years 271 at 5 years 142 at 6 months 145 at 1 year 95 at 3.5 years 271 at 5 years 142 at 6 months 145 at 1 year 95 at 3.5 years 271 at 5 years 135 at 6 months 137 at 1 year 94 at 3,5 years 269 at 5 years 135 at 6 months 137 at 1 year 94 at 2).5 yeais 269 at 5 years 135 at 6 months 137 at 1 year 94 at 3.5 years 269 at 5 years 135 at 6 months 137 at 1 year 94 at 3,5 years 269 at 5 years 135 at 6 months 137 at 1 year 94 at 3.5 years 269 at 5 years 135 at 6 months 137 at 1 year 94 at 3.5 years 269 at 5 years Estimate of association (95% Cl) Beta = 0.62 (-1.37. 2.62) Beta = -0.20 (-2.38, 1.98) Beta ---0.51 (-2.31, 1.28) Beta = 0.42 (-0.87, 1.72) Beta = 0.52 (-1.31,2.36) Beta = -0 .3 2 (-2.34, .70) Beta = -0.44 (-2.16, 1.27) Beta = 0.19 (-0.96, 1.35) Beta = 0.44 (-1.69, 2.57) Beta = 0.13 (-2.16, 2.42) Beta = -0.69 (-2.41, 1.02) Beta = 0.22 (-1.01, 1.45) Beta = 1.20 (-0.26, 2.67) Beta = -0.36 (-1.88, 1.15) Beta = 0.08 (-1.01, 1.17) Beta = -0.09 (--0.91,0.73) Beta = 0.78 (-0.54, 2.09) Beta = -0.13 (-1.46, 1.19) Beta = -0.02 (-1.07, 1.02) Beta --0.08 (--0.71, 0.86) Beta = 1.23 (-0.24, 2.71) Beta = -1.09 (-2.89, 0.71) Beta = 0.00 (-1.07, 1.08) Beta = -0.30 (-1.00, 0.41) Beta = 1.23 (-0.07, 2.54) Beta = -1.07 (-2.56. 0.41) Beta = 0.27 (-0.67, 1.21) Beta = -0.35 (-0.85.0.16) Beta = 1.21 (0.03, 2.40) Beta = -1.00 (-2.39, 0.38) Beta = 0.23 (-0.67, 1.13) Beta = -P.32 (-0.77, 0.14) Beta = 0.07 (-1,37, 1.51) Beta = -0.08 (-1.66, 1.49) Beta = 0.18 (-0.72, 1.07) Beta = -0.26 (-0.74, 0.22) Beta = -0.03 (-1.00, 0.93) Beta = 0.64 (-0.40, 1.67) Beta = -0.23 (-0.80, 0.34) Beta = 0.18 (-0.14, 0.50) Beta = 0.05 (-0.80. 0.89) Beta = 0.37 (-0.53, 1.28) Beta = -0.20 (-0.74,0.35) Beta = 0.15 (-0,16, 0.46) Beta = -0.11 (-1.10,0.88) Beta = 0.75 (-0.50, 2.00) Beta = -0.21 (-0.77, 0.36) Beta = 0.23 (-0.05, 0.50) Adjustment factors " " " " " " " " V ON ls> Comments >2 >3 - Z03c & __ - - .... __ .... ... __ Q s - - Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Quiros-Alcala et a!. (2011) Autonomic nervous system profile at 6 months and 1, 3.5, and 5 years Maternal prenatal urinary DAPs (mnol/L) 6 months: 22 coactivation 43 coinhibition 41 reciprocal parasympathetic activation 20 reciprocal sympathetic activation 1 year: 35 coactivation 33 coinhibition 43 reciprocal parasympathetic activation 21 reciprocal sympathetic activation 3.5 years: 11 coactivation 26 coinhibition 14 reciprocal parasympathetic activation 40 reciprocal sympathetic activation 5 years: 47 coactivation 75 coinhibition 4! reciprocal parasympathetic activation 99 reciprocal sympathetic activation None Geometric mean - 198.3 (143.6, 273.8) Resulls were similar when using Geometric mean = 110.0 (69 9, 173.1) creatinine-adjusted prenatal Geometric m ean--160,8 (lLi.8, 22/.3) concentrations" (NR) Geometric mean = 110.0 (69.9, 173.1) F = 1.53, P = 0.21 Geometric mean = 216.4 (157.0, 298.4) Geometric ntean= 141.8 (100.8, 199.5) Geometric mean --173,4 (125.2, 240.3) Geometric mean = 143.1 (77.4, 264.6) F --1.12, P --0.34 Geometric mean = 198.0 (101.9, 384.7) Geometric m ean= 185.9 (117.0, 295.4) Geometric mean =128.3 (72.4, 227.5) Geometric mean = 121.2 (96.1, 152.9) F = 1.58, P = 0.20 Geometric mean = 138.0 (102.0, 186.6) Geometric mean= 132.2 (106.3, 164.3) Geometric mean = 115.0 (85.8, 154.1) Geometric mean = 149.7 (126.4, 1/7.3) F = 0.83, P = 0.48 Coactivation profile: activation of both sympathetic and parasympathetic nervous systems during challenge tasks compared with rest Coinhibition profile: inhibition of both sympathetic and parasympathetic nervous systems during challenge tasks compared with rest Reciprocal parasympathetic nervous system activation and sympathetic nervous system withdrawal Reciprocal sympathetic nervous system activation and parasympathetic nervous system withdrawal Frequencies of coactivation and coinhibition at 6 months are taken from. Table 4b of manuscript (inconsistent with Table 4a) (Continued) Tier 3/4 L(}W~lVCl OP HSECid expOSUTES 613 ED 002061 00046431-00083 m m i ; m last ^4 614 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Quiros-Alcala et ai. (2011) Outcome Exposure Child urinary DAPs (nmoi/L) Number of subjects/events 6 months: 22 coactivation 43 coinhibition 41 reciprocal parasympathetic activation 20 reciprocal sympathetic activation 1 year: 35 coactivation 33 coinhibition 43 reciprocal parasympathetic activation 21 reciprocal sympathetic activation 3.5 years: 11 coactivation 26 coinhibition 14 reciprocal parasympathetic activation 40 reciprocal sympathetic activation 5 years: 47 coactivation 75 coinhibition 41 reciprocal parasympathetic activation 99 reciprocal sympathetic activation Lizardi et al. (2008) m m i ; m Trail Making Test B (seconds) at - 7 years Child urinary DAPs a 25 vs. < 25 pg/L in original screening sample 24 detectable ( a 25 gg/L) 22 non-detectable (< 25 gg/L) Estimate of association (95% Cl) Geometric mean = 46,3 (28.2, 76.2) Geometric mean = 58.8 (39.5, 87.3) Geometric mean = 38.8 (24.9, 60.5) Geometric mean = 88,4 (40.3, 193.9) Adjustment factors None F = 3.18, P --0.03 for creatinineadjusted child urinary DAPs and reciprocal sympathetic activation with parasympathetic withdrawal at 6 months ! 1.79, P --0.15 Geometric mean --34.5 (20.4, 58.2) Geometric mean = 58.0 (32.2, 104.5) Geometric mean = 68.6 (42.2, 111.8) Geometric mean = 46.9 (21.0, 105.0) ! 1.25, P -----0.29 Geometric mean = 93.3 (23.2,376.0) Geometric mean = 77.2 (45.2, 1.31.8) Geometric mean = 73,6 (37.1, 146.3) Geometric mean = 132.1 (78.9, 221.3) F = 0.87, P -----0.46 Geometric mean = 99.4 (66.2, 149.4) Geometric mean = 110.0 (79.2, 153.0) Geometric mean = 124.81(7.3.7, 211.4) Geometric mean = 90.4 (68.7, 119.0) F = 0.57, F = 0.63 Mean = 283 (224, 341) Mean = 204 (172. 236) F = 0.01 None Comments No significant differences were observed in autonomic nervous system profiles between children with consistently high (top 10%) vs. consistently low' (bottom 10%) prenatal and/or childhood DAP level s One child in each exposure group with a significantly higher urinary DAP level (519 gg/L and 850 gg/L) was excluded from analysis Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00084 ist Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. d o i Low-level o p insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Lizardi et al. (2008) Lizardi et al. (2008) Lizardi et al. (2008) Wechsler Intelligence Scale for Children-Third Edition Short Fonn, Children's Memory Scale, Wisconsin Card Sorting Test, Trail Making Test , Child Behavior C b e e k lis t/4 -1 8, and Teacher Report Form at - 7 years Wisconsin Card Sorting Test measures at ~ 7 years Child urinary DAPs (pg/L) in contemporaneous sample Wechsler Intelligence Scale for Children-Third Edition Short Form, Children's Memory Scale, Trail Making Test A and B, Child Behavior C h e c k lis t/4 -1 8, and Teacher Report Form at ~ 7 vears "No significant effects" (results NR) 48 46 after exclusion of outliers " Number of errors made: correlation = 0.31, P = 0.03 Number of perseverati ve responses: correlation - 0.34, P = 0.01 Number of perseverative errors: correlation = 0.35, P --0.01 Conceptual level responses provided: correlation = 0.38, P = 0.01 Failure to maintain set: correlation = 0.38, P = 0.02 After exclusion of one child in each exposure group with a significantly high urinary DAP level: "no significant correlations" (results NR) "No significant correlations (p < .05)" (results NR) m m i ; m Nat ^4 One child in each exposure group with a significantly higher urinary DAP level (519 ttg/L and 850 .tg/L) was excluded from analysis (Continued) Low-level OP insectide exposures 615 ED 002061 00046431-00085 616 R. R(?iss t ill. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Bouchard et al. (2010) Bouchard et al. (2010) Bouchard et al. (2010) Bouchard et al. (2010) Outcome ADHD by diagnostic criteria at 8-15 years ADHD by diagnostic, criteria or medication use at 8-15 years Exposure Child urinary DAPs, DMPs, or DEPs (nmoi/L, log.,, scale) Child urinary dimethylthiophosphate (nmol/g creatinine) Child urinary DAPs, DMPs, or DEPs (nmoi/L, log10 scale) Child urinary dimethylthiophosphate (nmol/g creatinine) Number of subjects/events 119/1139 (10.4%) Estimate of association (95% Cl) DAPs odds ratio = 1.21 (0.97, 1.51) DMPs odds ratio = 1.55 (1.14, 2.10) DEPs odds ratio - 0.94 (0,69, 1.28) 407 below detection limit 366 < median (30.4 nmol/g creatinine) 366 s- median (30.4 nmol/g creatinine) Odds ratio Odds ratio Odds ratio referent 1.05 (0.57, 1.95) 1.93 (1.23,3.02) 148/1139 (13.0%) DAPs odds ratio = 1.35 (1.10, 1.67) DMPs odds ratio = 1.72 (1.31, 2.28) DEPs odds ratio = 0.80 (0.60, 1.05) 407 below detection limit 366 < median (30.4 nmol/g creatinine) 366 - median (30.4 nmol/g creatinine) Odds ratio Odds ratio Odds ratio referent 1.22(0.65,2.27) 2.12(1.32,3.41) Adjustment factors Comments Gender, age, race/ethnicity, ratio of family income to poverty level, fasting duration, and logarithmically transformed urinary creatinine concentration No change after additional adjustment for year of data collection, blood lead concentration, maternal age at birth, or maternal smoking pregnancy, or after exclusion of children taking ADHD medication Gender, age, race/ethnicity, ratio of family income to poverty level, and fasting duration Diagnosis of ADHD is based on the presence during previous 12 months of symptoms related to inattention, hyperactivity, and impulsivity, with significant impairment in a settings (e.g., at school and at home); no requirement that symptoms occur without another neuropsychiatrie disorder or that symptoms were present before 7 years of age Results were similar when using creatinine-adjusted DAP, DMP, and DEP concentrations (results NR) No change after additional adjustment for year of data collection, blood lead concentration, maternal age at birth, or maternal smoking during pregnancy, or after exclusion of children taking ADHD medication Gender, age, race/ethnicity, ratio of family income to poverty level, fasting duration, and logarithmically transformed urinary creatinine concentration No change after additional adjustment for year of data collection, blood lead concentration, maternal age at birth, or maternal smoking during pregnancy Gender, age, race/ethnicity, ratio of family income to poverty level, and fasting duration No change after additional adjustment for year of data collection, blood lead concentration, maternal age at birth, or maternal smoking during pregnancy m m i ; m Tier 3/4 Crit Rev Toxicol. 2015; 45(7): 531-641 ED 002061 00046431-00086 last Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Bouchard et al. (2010) Hyperactive/ impulsive ADHD subtype at 8--1.5 years Child urinary DAPs, DMPs, orDEPs (nmol/L, iog,0 scale) 21/1139 (1.8%) Bouchard et al. (2010) Bouchard et al. (2010) Guodong et al. (2012) Inattentive ADHD subtype at 8-15 years Combined ADHD subtype at 8-15 years Gesell motor behavior at 23-25 months Guodong et ai, (2012) Gesell adaptive behavior at 23-25 months 69/1139 (6.1%) 29/1139 (2.5%) Child urinary DAPs (nmol/g creatinine, log|p scale) 301 300 normal, 1 (0.3%) with developmental delay 301 297 normal, 4 (1.3%) with developmental delay Guodong et al. (2012) Gesell language behavior at 23-25 months Guodong et al. (2012) Gesell social behavior at 23-25 months 301 282 normal, 19 (6.3%) with developmental delay 301 293 normal, 8 (2.7%) with developmental delay m i m DAPs odds ratio - 1.85 (1.04, 3.27) DMPs odds ratio - 2.13 (1.08, 4.20) DEPs odds ratio - 2.15 (1.06, 4.40) DAPs odds ratio - 1.14 (0.81, 1.61) DMPs odds ratio - 1.47 (0.99, 2.19) DEPs odds ratio --0.70 (0.49, 1.01) DAPs odds ratio - 1.05 (0.51, 2.16) DMPs odds ratio - 1.30 (0.48, 3.48) DEPs odds ratio --1.22 (0.59, 2.50) DAPs beta - 0.30 (-- 1.40, 1.99) DMPs beta - - 1.25 ( - 2,98, 0.47) DEPs beta - 0.32 ( - 1.37, 2.01) DAPs b e ta - 1.71 ( - 1.15, 4.57) DMPs beta - 2.53 ( - 0.05, 5,10) DEPs beta - - 0.41 (-- 3.22, 2.39) DAPs beta - 2,79 ( - 1.01, 6.60) DMPs beta --2.83 (--0.60, 6.26) DEPs beta - - 0.29 (- 4 .0 2 , 3.44) DAPs beta - - 0.66 ( - 2.12, 0.79) DMPs beta - - 0,48 (-- 1.93, 0.97) DEPs beta - - 0.93 ( - 2.40, 0.54) Gender, age, race/ethnicity, ratio of family income to poverty level, fasting duration, and logarithmically tran sforated urinary creatinine concentration No change after additional adjustment for year of data collection, blood lead concentration, maternal age at birth, or maternal smoking pregnancy, or after exclusion of children taking ADHD medication Child sex, maternal education level, and household income Motor behavior includes locomotion, reaching, balance, comprehension, drawing, and hand control Standardized to mean SD of 100 15, w ith< 85 indicating developmental delay Adaptive behavior includes hand-eye coordination, imitation, object recover comprehension, discriminative performance, perception, completion, and number conception Standardized to mean SD of 100 15, w ith< 8 5 indicating developmental delay Language behavior includes vocabulary, word comprehension, conversation, and word production Standardized to mean SD of 100 15, w ith< 85 indicating developmental delay Personal and social behavior includes reactions to people, personal habits, initiative and independence, play responses, and acquired information Standardized to mean SD of 100 15, w ith< 85 indicating developmental delay (Continued) m. Sc Low-lcwl OP USECtdE EXpOSUTES 617 ED 002061 00046431-00087 618 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Yolton et al. (2013) Yolton et al. (2013) Yolton et ai. (2013) Yolton et al. (2013) Yolton et al. (2013) Yolton et al. (201.3) Outcome NICU Network attention subscale at 5 weeks NICU Network lethargy subscale at 5 weeks NICU Network hypotonia subscale at 5 weeks NICU Network autonomic, stress subscale at 5 weeks All other NICU Network subscales at 5 weeks NICU Network profile at 5 weeks Exposure Maternal 16- and 26-week average prenatal urinary DEPs (nmol/g creatinine, log, scale) Maternal 16-week prenatal urinary DEPs (nmol/g creatinine, logscale) Maternal 26-week prenatal urinary DAPs (nmol/g creatinine, log2 scale) All other maternal prenatal urinary DAPs, DMPs, and DEPs (nmol/g creatinine, log, scale) at 16 weeks, 26 weeks, or averaged Maternal prenatal urinary DAPs (nmoi/g creatinine, log, scale) Number of subjects/events 350 Estimate of association (95% Cl) Beta --0.066, SE --0.033, P < 0.05 Beta - -0.069, SE - 0.034, P = 0.04 Beta - -0.101, SE - 0.045, P - 0.03 Beta - -0.010. SE - 0.004, P - 0.01 Not statistically significant (P > 0.05) 157 (45%) social/ easy-going 83 (31%) higharousal/difficult 110 (24%) hypotonic Odds ratio --referent Odds ratio, 16- and 26-week mean - 1.14 (0.98, 1.32) Odds ratio, 16-week - 1.02 (0.91, 1.15) Odds ratio, 26-week --1.13 (0.99, 1.27) Odds ratio, 16- and 26-week mean - 1.02 (0.87. 1.19) Odds ratio, 16-week --0.90 (0.79, 1.03) Odds ratio, 26-week --1.13 (0.99, 1.29) Adjustment factors Infant age at exam and race Infant age at exam, race, birth weight, and maternal consumption of fresh fruits and vegetables Infant age at exam, race, and maternal body mass index Infant age at exam, race, birth weight, and blood lead level NR Infant age at exam, race, maternal weight gain during pregnancy, and maternal body mass index Comments Results are presented only for statistically significant associations Positive coefficient --increased attention Negative coefficient --decreased lethargy Negative coefficient --decreased hypotonia Negative coefficient --decreased autonomic stress Profiles identified using latent profile analysis of patterns across NICU Netw'ork. Neurobehavioral Scale dimensions m m i ; m Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00088 Tat Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Dot Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Yolton et al. (2013) Yolton et al. (2013) Oulhote and Bouchard (2013) Oulhote and Bouchard (2013) Oulhote and Bouchard (2013) m m i m Strengths and Difficulties Questionnaire total difficulties score - 17 at ages 6-11 years Strengths and Difficulties Questionnaire conduct problems score s- 4 at ages 6-11 years Strengths and Difficulties Questionnaire emotional symptoms score s- 5 at ages 6--1.1 years Maternal prenatal urinary DMPs (nmol/g creatinine, log2 scale) Maternal prenatal urinary DEPs (nmol/g creatinine, log, scale) Child urinary DAPs, DMPs, or DEPs (ntnol/L, log,0 scale) Child urinary DAPs (nmol/L, iog10 scale) 157 (45%) social/ easy-going 83 (31%) higharousal/difficuit 110 (24%) hypotonic 157 (45%) social/ easy-going 83 (3%) higharousal/difficult 110 (24%) hypotonic 69 (6.8%) overall 48 boys 21 girls 779 subjects in analysis 78 (8.0%) overall 53 boys 25 girls 779 subjects in analysis Odds ratio --referent Odds ratio, 16- and 26-week mean - 1.11 (0,97, 1.26) Odds ratio, 16-week - 1.00 (0,90, 1. 10) Odds ratio, 26-week - 1.12 (1.00, 1.25) Odds ratio, 16- and 26-week mean = 0.99 (0.86, 1.13) Odds ratio, 16-week --0.90 (0.80, 1.00) Odds ratio, 26-week --1.12 (0.99, 1.26) Odds ratio --referent Odds ratio, 16- and 26-week mean = 1.03 (0,92, 1.15) Odds ratio, 16-week = 0.98 (0.89, 1.08) Odds ratio, 26-week = 1.03 (0.95,1.12) Odds ratio, 16- and 26-week mean = 0.96 (0.86, 1.09) Odds ratio, 16-week = 0.89 (0.81, 0.99) Odds ratio, 26-week = 1.03 (0.94, 1.13) DAPs odds ratio, total - 0.6 (0,3, 1.3) Sex, age, race/ethnicity, income, DAPs odds ratio, boys = 0.5 (0.2, 1.8) parental education, maternal smoking DAPs odds ratio, girls = 0.6 (0.3, 1.6) during pregnancy, birth weight, blood P-interaction by sex = 0.97 DMPs odds ratio --0.8 (0.4, 1.6) lead levels, urinary creatinine, body mass index, and fasting status DEPs odds ratio = 0.3 (0.1, 1.8) No change when using creatinine - standardized metabolite concentrations, not adjusting for blood lead, or not weighting to account for survey design DAPs odds ratio, total - 0.6 (0,3, 1.3) DAPs odds ratio, boys = 0.7 (0.3, 1.7) DAPs odds ratio, girls - 0,4 (0.1, 2.0) P-interaction by sex = 0.58 Questionnaire is designed for screening of mental and behavioral difficulties and strengths in population surveys; each dimension scale is scored on a scale of 0-0, and total difficulties are calculated by summing four dimension scales (total of 0--40) Prosocial behavior not analyzed because "too few' children had high scores" 97 (9.1 %) overall 47 boys 50 girls 779 subjects in analysis DAPs odds ratio, total =1.1 (0.5, 2.2) DAPs odds ratio, boys = 1.0 (0.4, 3.0) DAPs odds ratio, girls = 1.0 (0.5, 2.0) P-interaction by sex = 0.96 (Continued ) Low-level OP insectide exposures 619 ED 002061 00046431-00089 V 620 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued ) Reference Oui hote and Bouchard (2013) Oulhote and Bouchard (2013) Fortenberry et al. (2014) Outcome Strengths and Difficulties Questionnaire hyperactivity/ inattention score a 7 at ages 6-11 years Strengths and Difficulties Questionnaire peer problems score a 4 at ages 6-11 years Conners parent-rated ADHD index at 6-11 years Exposure Maternal prenatal TCPy (ng/raL), textiles 2 and 3 vs. 1 Fortenberry et al. (2014) Conners parent-rated hyperactivity/ impulsivity ADHD at 6-11 years m m i m Number of subjects/events 109 (11.1%) overall 76 boys 33 girls 779 subjects in analysis Estimate of association (95% Cl) DAPs odds ratio, total = 0.8 (0.3, 2.0) DAPs odds ratio, boys = 0.9 (0.4, 2.4) DAPs odds ratio, girls - 0,4 (0.1, 1.8) P-interaction by sex - 0.21 Adjustment factors Comments 71 (7.3%) overall 43 boys 28 girls 779 subjects in analysis 187 total 80 males 97 females DAPs odds ratio, total = 0.8 (0.3, 2.0) DAPs odds ratio, boys --0.8 (0.3, 2.3) DAPs odds ratio, girls --0.6 (0.2, 2.7) P-interaction by sex = 0.75 Tertile 2 beta, total - 2.61 ( - 1.54. 6.75) Tertile 3 beta, total = 4.00 (-0 .9 1 , 8.90) P-trend, total --0.11 Tertile 2 beta, males --2.32 ( --2.55, 7.20) Tertile 3 beta, males = 5.55 ( - 0.19, 11.3) P-trend, males --0.06 Tertile 2 beta, females = 1.63 ( - 5.55, 8.82) Tertile 3 beta, females = 0.17 ( - 8.28, 8.63) P-trend, females = 0.96 Tertile 2 beta, total = --0.56 ( - 5.03, 3.91) Tertile 3 beta, total - --0.51 ( - 5.80.4.78) P-trend, total - 0.84 Tertile 2 beta, males -- - 0.17 (--6.63, 6.29) Tertile 3 beta, males --1.25 ( - 6.36. 8.87) P-trend, males = 0.76 Tertile 2 beta, females --0.33 (--6.44, 7.10) Tertile 3 beta, females = --3.81 ( - 11.8.4.16) P-trend, females = 0.35 Child sex, maternal intelligence quotient, maternal education, income, child age at testing, specific gravity, season, breast feeding, blood lead, delivery length, and delivery head circumference No significant differences in geometric. mean TCPy concentrations were detected between trimesters, but significant within-person variability was detected across trimesters (intraclass correlation = 0.29-0.32 for specific-gravity-corrected TCPy, 0.41 for uncorrected TCPy) Higher score on Conners' Parental Rating Scales-Revised ADHD Index indicates an elevated level of concern for risk of ADHD, with a score of 4059 being average and < 40 displaying fewer concerns Various tests are used to "assess ADHD-related symptoms and are not designed as diagnostic tools, but rather for screening" Scale based on Diagnostic and Statistical Manual o f Mental Disorders, 4th Edition, with scores ranging between 0 and 9 and scores > 6 suggesting a possible diagnosis Tier 3/4 CritRevToocoL 2015; 45(7): 531-641 ED 002061 00046431-00090 V ED 002061 00046431-00091 Tier 3/4 Sierra Club v. EPA 18cv3472 NDCA Fortenberry et al. Conners parent-rated (2014) inattention ADHD at 6-11 years Fortenberry et al. Conners parent-rated (2014) combined ADHD at 6-11 years Fortenberry et al. (2014) Conners parentrated global restlessness/ impulsivity index at 6-11 years m m i m V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Tettile 2 beta, total = 2.37 ( - 1.79, 6.53) Tettile 3 beta, total --2.45 (--2.47, 7.37) P-trend, total --0.31 Tertile 2 beta, males = 2.33 ( --2.36, 7.02) Tertile 3 beta, males --2,63 (--2.89, 8.16) P-trend, males --0.32 Tertile 2 beta, fernales = 1.19 ( - 6.09, 8.47) Tertile 3 beta, fernales = --0,07 ( - 8.64, 8.50) P-trend, fernales - 0.99 Tertile 2 beta, total = 1.23 (-2 .8 9 . 5.35) Tertile 3 beta, total = 1.10 ( - 3.77, 5.98) P-trend, total = 0.64 Tertile 2 beta, males = 0.80 ( --4.48, 6.09) Tertile 3 beta, males = 2.06 ( --4.17, 8.29) P-trend, males = 0.51 Tertile 2 beta, fernales - 1.64 ( - 5.17. 8.45) Tertile 3 beta, fernales = --1.83 (-9 .8 4 , 6.19) P-trend, fernales = 0.66 Tertile 2 beta, total = - 0.15 (--4.57, 4.27) Tertile 3 beta, total = 0.38 (--4.85, 5.61) P-trend, total = 0.89 Tertile 2 beta, males = 0.49 ( --5.71, 6.68) Tertile 3 beta, males = 3,78 (--3.52, 11. 1) P-trend, males = 0.32 Tertile 2 beta, fernales = --0.48 (-7 .1 0 , 6.14) Tertile 3 beta, fernales = --4,90 ( - 12.7, 2.89) P-trend, fernales - 0.22 Scale based on Diagnostic and Statistical Manual o f Mental Disorders, 4th Edition, with scores ranging between 0 and 9 and scores > 6 suggesting a possible diagnosis Scale based on Diagnostic and Statistical Manual o f Mental Disorders, 4th Edition, with scores ranging between 0 and 9 and scores Sr 6 suggesting a possible diagnosis Higher score on Conners' Parental Rating Scales-Revised Global Restiessness/Impulsivity Index indicates an elevated level of concern for tendencies toward hyperactivity and inattention, with a score of 40-59 being average and < 40 displaying fewer concerns (Continued) Low-level OP insectide exposures 621 Dot Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Table 3. (Continued ) Reference Fortenberry et al. (2014) Outcome Behavioral Assessment System for Children attention problems at 6-11 years Fortenberry et al. (2014) Behavioral Assessment System, for Children hyperactivity at 6--11 years Fortenberry et al. Conners clinical (2014) ADHD index at 6-11 years m m i m Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Exposure " " " Number of subjects/events " " " Estimate of association (95% Cl) Tettile 2 beta, total = 1.79 (--2.66, 6.24) Tettile 3 beta, total = 3.46 (--1.81, 8.73) P-trend, total = 0.19 Tettile 2 beta, males ----0.37 ( - 7.02, 6.27) Tettile 3 beta, males --5,59 (- 2.24, 13.4) P-trend, males = 0.18 Tettile 2 beta, females = 5.81 (-0 .7 5 , 12.4) Tettile 3 beta, females = 1.82 (-- 5.91, 9.55) P-trend, females = 0.62 Tettile 2 beta, total = - 3.69 (--7.88, 0.50) Tettile 3 beta, total = --3.35 (--8.31, 1.60) P-trend, total = 0.17 Tettile 2 beta, males = --5.00 ( - 12.0, 2.00) Tettile 3 beta, males = --3.49 (-1 1 .7 , 4.73) P-trend, males = 0.36 Tettile 2 beta, females = - 0.005 ( - 5.17, 5.16) Tettile 3 beta, females = --2,77 (-8.84,3.31) P-trend, females = 0.37 Tettile 2 beta, total = --3.97 (--12.5, 4.51) Tettile 3 beta, total = 2.19 (--8.11, 12.5) P-trend, total = 0.73 Tettile 2 beta, males = --4.29 ( - 15.8, 7.18) Tettile 3 beta, males = 0,84 (--12.8, 14.5) P-trend, males = 0.95 Tettile 2 beta, females = 0.42 ( - 13.2, 14.0) Tettile 3 beta, females = 8.55 ( - 7.83, 24.9) P-trend, females = 0.31 Adjustment factors O' N> Comments Higher score on Behavioral Assessment for Children-Parental Rating Scales indicates elevated level of concern, with scores a 59 indicating increased levels of attention/hyperactivity problems Higher score on Behavioral Assessment for Children-Parental Rating Scales indicates elevated level of concern, with scores a 59 indicating increased levels of attention/hyperactivity problems Conners' Continuous Performance Test clinical index measures the likelihood of an ADHD diagnosis, with a high sensitivity (83-90%) but poorer specificity (59-61%) when compared with clinical ADHD diagnosis n 5 Tier 3/4 ED 002061 00046431-00092 V Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. doi Low-level op insectide exposures Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 Fortenberry et al.. Conners hit reaction (2014) time block change at 6-11 years Zhang et al. (2014) Neonatal Behavioral Neurological Assessment summary score at 3 days Maternal prenatal 249 total urinary DAPs 138 boys (pg/L, logj qscale) 111 girls Zhang et al. (2014) Zhang et al. (2014) Zhang et al. (2014) Zhang et al. (2014) Zhang et al. (2014) Neonatal Behavioral Neurological Assessment behavior score at 3 days Maternal prenatal urinary DMPs (pg/L, iog10 scale) Maternal prenatal urinary DEPs (pg/L, iog10 scale) Maternal prenatal urinary DAPs (pg/L, iog10 scale) 249 total 138 boys 111 girls 249 total 138 boys 111 girls 249 total 138 boys 111 girls Maternal prenatal urinary DMPs (pg/L, log10 scale) g 249 total 138 boys 111 girls 249 total 138 boys 111 girls m i m Tertile 2 beta, total = - 4.59 (--9.55, 0.36) Tertile .3 beta, total.----5.10 (--11.1, 0.91) P-trend, total = 0.09 Tertile 2 beta, males = --5.10 ( - 13.1,2.92) Tertile 3 beta, males ----6.86 (-- 16.4, 2.68) P-trend, males --0.14 Tertile 2 beta, females = --3.79 ( - 10.6, 2.98) Tertile 3 beta, females ----2,33 ( - 10.5, 5.82) P-trend, females - 0.55 Beta = -1.78 (-2.12,-1.45) Beta = -1.47 (-1.93,-1.01) Beta = -2.03 (-2.55, -1.52) Beta = -0.96 (-1.35,-0.57) Beta - -0.9.3 (-1.45, -0.40) Beta ---1.22 (-1.89,-0.55) Beta = -0.88 (-1.30, -0.47) Beta - -0.61 (-1.15,-0.07) Beta ---0.98 (-1.58,-0.39) Beta ---0.65 (-0.85,-0.45) Beta - -0.50 (-0.76, -0.23) Beta ---0.84 (-1.15.20.52) Not significant (results NR) Maternal age, education, gestational age, prenatal body mass index, and cord blood lead concentration Results were similar using creatinineadjusted DAP concentrations Hit reaction time block change is the variability of reaction time for correct responses across blocks or sections of the Conners' Continuous Performance Test, and has been indicated as a measure of vigilance or sustained attention Neonatal Behavioral Neurological Assessment summary score is based on 20 items, each score from 0-2, with > 37 considered as well developed, < 34 considered as abnormal, and 34-37 considered as acceptable No evidence of departure from linearity was observed in analyses bv quintile of DAPs Neonatal Behavioral Neurological Assessment behavior scale includes six items, each scored from 0-2, for a maximum of 12 (higher = better) No evidence of departure from linearity w'as observed in analyses bv quintile of DAPs Beta - -0.59 (-0.79, -0.40) Beta ---0.42 (-0.67,-0.17) Beta = -0.83 (-1.15, -0.53) (Continued) Low-level OP insectide exposures 623 ED 002061 00046431-00093 624 R. Reiss et al. Sierra Club v. EPA 18cv3472 NDCA Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. Table 3. (Continued) Reference Zhang et al. (2014) Zhang et al . (2014) Zhang et al . (2014) Zhang et al. (2014) Zhang et al. (2014) Zhang et al. (2014) Zhang et al. (2014) Outcome Neonatal Behavioral Neurological Assessment passive tone score at 3 days Exposure Maternal prenatal urinary DAPs (pg/L, logj qscale) Number of subjects/events 249 total 138 boys 111 girls Estimate of association (95% Cl) Beta = -0.22 (-0.34, -0.10) B e ta - -0.21 (-0.36,-0.02) Beta - -0.21 (-0.40, -0.02) Neonatal Behavioral Neurological Assessment active tone score at 3 days Maternal prenatal urinary DMPs (pg/L, iog10 scale) Maternal prenatal urinary DEPs (pg/L, logle scale) Maternal prenatal urinary DAPs (pg/L, log10 scale) 249 total 138 boys 111 girls 249 total 138 boys 111 girls 249 total 138 boys 111 girls Beta - -0.22 (-0.33, -0.11) Beta - -0.19 (-0.35, -0.07) Beta - -0.18 (-0.35, -0.01) Not significant (results NR) Beta - -0.48 (-0.66. -0.30) Beta - -0.46 (-0.72, -0.21) B e ta - -0.51 (-0.76,-0.25) Neonatal Behavioral Neurological Assessment primary reflexes score at 3 days Maternal prenatal urinary DMPs (pg/L, log10 scale) Maternal prenatal urinary DEPs (pg/L, log,,, scale) Maternal prenatal urinary DAPs (pg/L, log jqscale) 249 total 138 boys 111 girls 249 total 138 boys 111 girls 249 total 138 boys 111 girls B e ta - -0.41 (-0.57,-0.29) Beta - -0.34 (-0.58, -0.11) B e t a - -0.41 (-0.65,-0.18) Not significant (results NR) Beta - -0.36 (-0.51, -0.21) Beta - -0.34 (-0.55, -0.13) Beta ---0.39 (-0.61,-0.17) Zhang et al. (2014) Zhang et al. (2014) Maternal prenatal urinary DMPs (pg/L, logjQ scale) Maternal prenatal urinary DEPs (pg/L, log,0 scale) 249 total 138 boys 111 girls 249 total 138 boys 111 girls B e ta - -0.30 (-0.44.20.17) Beta --not significant (NR) Beta - -0.347-0.54, -0.14) Beta --not significant (NR) Beta - -0.28 (-0.48, -0.09) Beta --not significant (NR) Adjustment factors " Comments Neonatal Behavioral Neurological Assessment passive tone scale includes four items, each scored from 0-2, for a maximum of 8 (higher - better) No evidence of departure from linearity was observed in analyses bv quintile of DAPs Neonatal Behavioral Neurological Assessment active tone scale includes four items, each scored from 0-2, for a maximum of 8 (higher --better) No evidence of departure from linearity was observed in analyses by quintile of DAPs " Neonatal Behavioral Neurological Assessment primary reflexes scale includes three items, each scored from. 0-2, for a maximum of 6 (higher --better) No evidence of departure from linearity was observed in analyses by quintile of DAPs A D H D attention deficity/hyperactivity disorder confidence interval , DAP dialkyl phosphate, D E P diethyl phosphate, DMP dimethyl phosphate, 1QR interquartile range, MDA malathion dicarboxylic acid, NICU neonatal intensive care unit, NR not reported, PON1 paraoxonase 1, S D standard deviation, SE standard error, T C 'P y 3,5,6-trichloro-2-pyridinol. m i m Tier 3/4 CritRevToxicoL 2015; 45(7): 531-641 ED 002061 00046431-00094 m. SE Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 625 chlorpyrifos detected (Table 2) (Rauh et al. 2011). However, a a concern, because the completeness and accuracy of report significant inverse association was detected with the Wechsler ing may have varied by lifestyle factors related to maternal Working Memory Scale (parsimonious model beta = --0.006, prenatal chlorpyrifos exposure. Additional limitations are the 95% Cl = --0.009, --0.002; no substantial change after dichotomization of cord plasma chlorpyrifos levels in several further adjustment). This association was not substantially analyses, which precluded exposure-response analyses, and confounded (change in beta < 10% ) by childhood home envi the focus on a single OP insecticide. ronment at age 3 years, based on composite indices (total Taken together, the results of neurodevelopmental studies Home Observation for Measurement of the Environment or in the CCCEH cohort suggest associations between prenatal HOME score, Environmental Stimulation Scale, and Paren chlorpyrifos exposure and selected adverse neurodevelopmen tal Nurturance Scale) derived from observational interview tal outcomes, with some as-yet-unexplained heterogeneity by data (Table 2) (Horton et al, 2012). Additionally, no appar subgroups and numerous statistically null associations. For ent interaction was observed between chlorpyrifos and the instance, an inverse association between cord plasma chlorpy Parental Nurturance Scale. However, the association between rifos levels and lower scores on the Bayley Mental Develop chlorpyrifos and Wechsler Working Memory varied by child ment Index was detected at 36 months among African Ameri sex, with a significant inverse association detected only among can children, but not among Dominican children and not in boys (beta - - 2.382, 95% Cl - - 3.88, - 0.88) and not girls either group at 12 or 24 months. In the absence of a priori (beta ---- - 0.524, 95% Cl ------ - 1.90, 0.85). hypotheses, it is unclear why prenatal chlorpyrifos exposure Forty children aged 5.9-11.2 years in the CCCEH cohort might be associated with attention problems and pervasive with low prenatal exposure to environmental tobacco smoke developmental disorder but not externalizing or internalizing (based on maternal self-report and cotinine levels < 1 5 ng/inL behavior problems as assessed by the Child Behavior Check in cord plasma) and polycyclic aromatic hydrocarbons (based list, or with working memory among boys but not overall IQ, on maternal third-trimester personal air monitoring levels verbal comprehension, perceptual reasoning, or processing below the median of 2.26 ng/rn3), including 20 children in the speed as assessed by the Wechsler Intelligence Scale for Chil highest fertile of cord plasma chlorpyrifos (> 4.39 pg/g) and dren. Given the large number of outcomes tested, at least some 20 below the highest tertile, participated in a study of brain of the observed associations are almost certainly due to chance. morphology using Tl-weighted high-resolution magnetic res Again, neither this study nor any other study of neurodevelop onance imaging (Table 2) (Rauh et al. 2012). Significant dif mental outcomes described in this review adjusted for multiple ferences between chlorpyrifos exposure groups that involved comparisons. The observed associations wi th brain morphology primarily white matter included bilateral enlargement of the are notew'orthy, but multiple comparisons are again a concern, superior temporal, posterior middle temporal, and inferior especially given the exclusive reporting of anatomic regions postcentral gyri; right-hemisphere enlargement of the supra where associations with chlorpyrifos exposure were observed, marginal gyrus, inferior parietal lobule, and superior frontal but not those without any such associations. Overall, the results gyrus, gyrus rectus, cuneus, and precuneus along the mesial suggesting an adverse neurodevelopmental effect of prenatal wall; and inward deformations in the dorsal and mesial surfaces chlorpyrifos exposure cannot reliably be interpreted as causal of the left superior frontal gyrus. No significant difference was due to methodological limitations and internal inconsistency, found in overall brain size by chlorpyrifos level. Wechsler and require independent confirmation in other study settings. Full-Scale IQ at age 7 years was positively correlated with surface measures in the bilateral superior temporal, inferior frontal, inferior precentral, and inferior postcentral gyri and Mount Sinai Children's Environmental Cohort Study the left precuneus, and inversely correlated with surface mea The Mount Sinai CECS, described earlier, administered the sures in the right fusiform gyrus, among children with lower Brazelton Neonatal Behavioral Assessment Scale to evaluate cord plasma chlorpyrifos levels, but not those in the higher- 28 behavioral items and 18 primitive reflexes, grouped into exposure group. Normal sex differences in the right inferior seven clusters, in 311 neonates prior to hospital discharge at parietal lobule, superior marginal gyrus, and mesial superior or before 5 days (Engel et al. 2007). Subsequently, the Bayley frontal gyrus were reversed among children with higher chlo Scales of Infant Development, 2nd Edition, were adminis rpyrifos levels. "Scattered reductions" in cortical thickness in tered at 12 months (n = 200) and 24 months in = 276), and dorsal and parietal and frontal cortices were also associated the Wechsler Preschool and Primary Scale of Intelligence, w'ith higher chlorpyrifos levels. 3rd Edition, or the Wechsler Intelligence Scale for Children, The major strengths and limitations of the CCCEH cohort 4th Edition, was administered at ages 6-9 years (n = 169) study were discussed above in the context of analyses of birth (Table 1) (Engel et al, 2011). The Brazelton scale evaluates outcomes, and apply also to the analyses of neurodevelop- 28 behavioral items and 18 primitive reflexes, which can be mental outcomes, except that selection bias due to differen scored into seven clusters: habituation, orientation, motor, tial participation rates of mothers by childhood neurological range of state, regulation of state, autonomic stability, and outcomes at age 7 years is improbable, though not impos number and type of abnormal reflexes (including plantar, sible. For example, risk factors such as a personal or family Babinski, ankle clonus, rooting, sucking, glabella, passive history of neurological problems might influence the deci resistance of legs, passive resistance of arms, palmar, placing, sion to participate. However, selection bias due to differential standing, walking, crawling, incurvation, tonic deviation of dropout rates is a greater concern in studies with relatively head and eyes, nystagmus, tonic neck reflex, and Moro reflex). long follow-up. Bias in analyses of parent-reported outcomes, The Wechsler Preschool and Primary Scale of Intelligence is such as those based on the Child Behavior Checklist, is also used to derive composite Verbal Comprehension, Perceptual SI QHI S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00095 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 626 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 Reasoning, Processing Speed, and Full-Scale IQ scores; the 24 months (including after stratification by race/ethnicity or Wechsler Intelligence Scale for Children was described above PON1192 genotype) or the Bayley Psychomotor Development for the CCCEH study. Index at 12 or 24 months (including after stratification by In adjusted models, maternal prenatal urinary levels of race/ethnicity). Moreover, DAP, DMP, and DEP levels were DAPs, DMPs, and DEPs (classified as linear on the log10 scale not significantly associated with any Wechsler Intelligence or into quartiles) and detectable MDA were not significantly Scale measures, including Full-Scale IQ, Perceptual Reason associated with the Brazelton habituation, orientation, motor, ing, Verbal Comprehension, Processing Speed, and Work range of state, regulation of state, or autonomic stability ing Memory (assessed at ages 7-9 years only) at 6, 7-9, or clusters (Table 2) (Engel et al. 2007). However, a logi()-unit 6-9 years. Only after stratification by PON1 /p2 genotype were increase in DEP levels was associated with a significantly significant inverse associations detected between maternal higher number of abnormal reflexes (relative risk [RR] = 1.49, prenatal urinary' DAP and DMP levels and the Wechsler Per 95% Cl = 1.12, 1.98), and total DAP levels were also mar ceptual Reasoning Index (e.g., beta per log]0-unit increase in ginally associated with abnormal reflexes (RR = 1.32, 95% DAPs = - 0 .5 6 , 95% Cl = " 4.80, 3.68 for PON1 m QR/RR Cl = 0.99, 1.77), whereas DMP levels were not significantly carriers; beta = --7.52,95% Cl = --14.53, --0.51 for PON1192 associated (RR ==1.13,95% Cl = 0.90,1.41). Detectable MDA QQ carriers). No significant interactions with PON1 m geno levels in maternal prenatal urine were also associated with a type were observed for the Wechsler measures of Full-Scale significantly higher number of abnormal reflexes (RR = 2.24, IQ or Verbal Comprehension. 95% Cl = 1.55, 3.24). When levels of DAPs, DMPs, and DEPs Key strengths and limitations of the Mount Sinai CECS were categorized into quartiles, some positive associations were delineated earlier and apply equally to the analyses of with number of abnormal reflexes were still detected, but not neurological outcomes. Selection bias due to unequal enroll in a monotonic exposure-response pattern. When the number ment rates may not have a major influence on associations o f abnormal reflexes was dichotomized as > 2 or < 2 and anal with long-term childhood neurological outcomes, unless yses were stratified by infant age, associations with maternal participation varied by strong neurological risk factors, but prenatal urinary DAPs, DMPs, and DEPs were stronger for selection bias due to unequal follow-up is a reasonable con those aged> 2 days, whereas the association with detectable cern. For example, 311 (77%) of 404 eligible infants com MDA was stronger for those aged 1 day. Statistically significant pleted the Brazelton Neonatal Behavioral Assessment Scale interactions between maternal prenatal plasma PON1 expres before hospital discharge, excluding those admitted to the sion levels and urinary DAP and DMP metabolite levels were Neonatal Intensive Care Unit (NICU), those delivered and detected with risk of > 2 abnormal reflexes as the outcome. discharged over a weekend, those whose parent refused, those Specifically, the RR per-unit increase in prenatal DAPs was who were not testable, and those for whom study personnel 2.38 (95%: C I= 1.37, 4.15) tor those in the lowest tertile of were unavailable; thus, selection bias could have occurred if PON1 expression level versus 0.76 (95% Cl = 0.48, 1.20) for exclusions were associated with both DAP metabolite levels those in the highest tertile, and the RR for prenatal DMPs was and neonatal behavioral outcomes. Multiple comparisons I. 96 (95%: CI = 1.27, 3.03) for those in the lowest tertile of potentially leading to chance findings are a particular con PON1 expression level versus 0.73 (0.56,0.96) for those in the cern in these analyses, given the large number of outcomes highest tertile. Associations with prenatal DEPs did not vary and subgroups examined, along with the apparent lack of a significantly by PON1 expression. priori hypotheses regarding why some but not other neuro In analyses using the Bayley Scales at 12 and 24 months, logical outcomes might be associated with OP metabolites, maternal prenatal urinary DAP and DMP metabolite levels (but or why associations might be observed in some but not other not DEP levels) were associated with significantly lower scores subgroups by age and race/ethnicity. Consequently, although on the Mental Development Index at 12 months among blacks the associations of maternal prenatal urinary levels of DAP and Hispanics (beta per logu)-unit increase in DAPs = -3 .2 9 , and DEP metabolites and detectable MDA with abnormal 95% Cl = - 5 .8 8 , " 0.70; beta for DMPs = - 3 . 3 5 , 95% neonatal reflexes were noteworthy, the absence of a mono Cl = " 5.64, ---1.06), but significantly higher scores among tonic exposure-response pattern, along with the absence of whites (beta for DAPs = 4.77, 95%; Cl = 0.69, 8.86; beta for association with motor performance, autonomic stability, and DMPs = 4.45, 95% Cl = 0.82, 8.08) (Table 2) (Engel et al. other neurological outcomes, detracts from the coherence of 2011). When analyses of the Bayley Mental Development these findings. Likewise, the persuasiveness of the inverse Index at 12 months were stratified by maternal P 0 N 1 192 associations of maternal prenatal urinary levels of DAP and genotype, interactions were observed among blacks and His DMP metabolites with mental development at 12 months in panics, with significantly lower scores among those carrying blacks and Hispanics, especially in PON119? QR/RR car the PON1192 QR or RR genotype (i.e., heterozygotes and low- riers, is undermined by the positive associations in whites, activity homozygotes) than QQ homozygotes (e.g., beta per the absence of any association at 24 months, and the lack of ]og10-unitincrease in DAPs = --4.94,95% Cl = --7.87, --2.07 any interaction with other PON1 genotypes or PON 1 activity for PON1 i92 QR/RR carriers; beta = 5,72, 95% C I= --0.48, levels. The stronger inverse association of prenatal DAP and II. 92 for PON1,92 QQ carriers). However, no significant DMP levels with perceptual reasoning in 6- to 9-year-olds in interactions were found with the PON1L59Mor PONl_10gr T PON1,9? QQ carriers also runs counter to expectation. Con polymorphism, or with PON1 enzymatic activity for any neu- sequently, the few observed significant associations among rodevelopmental outcome assessed. No significant associations a large number of statistically null associations, without a were detected between maternal prenatal urinary DAP, DMP, discern able pattern, cannot reliably be interpreted as causal, or DEP levels and the Bayley Mental Development Index at and require confirmation in independent studies. U IQHTS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00096 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 627 Center for the Health Assessment of Mothers and Children o f Salinas The basic methods of the CHAMACOS birth cohort study were described earlier, follow-up for neurodevelopmental outcomes continued through age 7 years (Table 1) (Bouchard et al. 2011, Eskenazi et al. 2010, Eskenazi et al. 2007, Marks et al. 2010, Quiros-Alcala et al. 2011, Young et al. 2005). Geometric mean urinary DAP metabolite levels measured in children increased with age: 45.5 nmol/L (95% 1 = 39.6, 52.3) at 6 months, 59.5 = nmol/L (51.7, 68.5) at 12 months, 70.9 nmol/L (61.4, 81.9) at 24 months, 77.5 nmol/L (65.4, 91.9) at 3.5 years, and 92.6 nmol/L (78.6, 109.0) at 5 years (Eskenazi et al. 2007, Marks et al. 2010). Neurodevelop mental outcomes were measured using the Brazelton Neo natal Behavioral Assessment Scale administered by 62 days (2 months); the Bayley Scales of Infant Development, 2nd Edition, administered at 6, 12, and 24 months; an autonomic nervous system reactivity protocol that measured heart rate, respiratory sinus arrhythmia, and pre-ejection period follow ing social, physical, and emotional challenges (and cognitive challenges for older children) administered at 6 months and 1, 3.5, and 5 years; the mother-completed Child BehaviorChecklist for ages 1.5-5 years administered at 2, 3.5, and 5 years; the NEPSY visual attention subtest, 2nd Edition, administered at 3.5 years; the Conners' Kiddie Continuous Performance Test, which assesses reaction time, accuracy, and impulse control for ADHD screening using an interactive computer program; the Hillside Behavior Rating Scale, which assesses motor activity and distractibility for ADHD screen ing, administered at 5 years; and the Wechsler Intelligence Scale for Children, 4th Edition, administered at 7 years. Among infants assessed at or before age 2 months, no significant association was observed between maternal pre natal average urinary levels of DAPs, DMPs, or DEPs and the Brazelton habituation, orientation, motor performance, range of state, or regulation of state cluster, either overall or among neonates assessed at age < 3 days or > 3 days (Table 2) (Young et al. 2005). Maternal prenatal urinary DEP metabolite levels, but not DAPs or DMPs, were significantly associated with a higher score on the autonomic stability cluster, which includes tremors, startles, and skin color, at age < 3 days (beta per log10-unit increase = 0.31, 95% Cl = 0.01,0.61), but not at age > 3 days (beta = --0.16,95% Cl = --0.47,0.14) or overall. By contrast, maternal prenatal urinary DAP, DMP, and DEP metabolite level s were all significantly associ ated wi th a higher number of abnormal reflexes, especially at age > 3 days (beta for DAPs = 0.53, 95% Cl = 0.23, 0.82; beta for DMPs = 0,41, 95% Cl = 0.12, 0.69; beta for DEPs = 0.37, 95% Cl = 0.09, 0.64), but not at age < 3 days (beta for DAPs = - 0.01, 95% Cl = - 0.24, 0.22; beta for DMPs = - 0.00, 95% Cl = - 0,21, 0.20; beta for DEPs = 0.08, 95% Cl = - 0 .16, 0.32). When the number of abnormal reflexes w;as dichotomized at > 3 versus < 3, maternal prenatal urinary DAP, DMP, and DEP levels were categorized into quintiles, and the analysis was restricted to neonates aged > 3 days at assessment, statistically significant positive exposure-response trends were observed for each metabolite type. The OR for > 3 abnormal reflexes per log10-unit increase in metabolite concentration was 4.9 (95% C I= 1.5, 16.1) for DAPs, 3.2 (95% C I= 1.1, 9.8) for DMPs, and 3.4 (95% Cl = 1.2, 9.9) for DEPs. No associations with any neonatal neurodevelopmental outcome were detected with maternal post-delivery urinary metabolite levels. A significant inverse association was detected between maternal prenatal urinary DAP and DMP levels and the Bayley Mental Development Index at 24 months (beta per log10unit increase in DAPs = --3.54, 95% Cl = --6.59, --0.49; beta for DMPs = - 3.64, 95% Cl = - 6.36, - 0.91) (Table 2) (Eskenazi et al. 2007), By contrast, child urinary DAP and DMP levels at 24 months were positively associated with the Mental Development Index (beta for DAPs = 2.37, 95% Cl = 0.50, 4.24; beta for DMPs = 2.01, 95% Cl = 0.24, 3.78). Child urinary DEP levels at 12 months were also positively associated with the Mental Development Index at that age (beta = 1.89, 95% 1 = 0.21, 3.58). Otherwise, associations of maternal prenatal and child urinary DAP, DMP, and DEP metabolites, as well as maternal prenatal MDA and TCPy levels, with the Mental Development Index at 6, 12, and 24 months were statistically non-significant, and all associa tions w'ith the Psychomotor Development Index at those ages were non-significant. Maternal prenatal and 24-month child urinary levels of DAPs, DMPs, DEPs, MDA, and TCPy were not significantly associated with a clinically borderline score ( > 93rd percentile) for attention problems or ADHD as assessed by the Child Behavior Checklist at 24 months. However, maternal prenatal urinary levels of DAPs and DMPs were at least marginally significantly associated with a higher odds of clinical pervasive developmental disorder (> 97th percentile) as assessed by the Child Behavior Checklist at 24 months (OR for DAPs = 2.25, 95% Cl = 0.99, 5.16; OR for DMPs = 2.19, 95% Cl = 1.05,4.58; OR for DEPs = 0.88,0.37,2.07), as were all three types of metabolites in children (OR for DAPs = 1.71, 95% Cl = 1.02, 2.87; OR for DMPs = 1.52, 95% Cl = 0.94, 2.45; OR for DEPs = 1.72, 1.12, 2.64). Maternal prenatal uri nary MDA and TCPy levels were not significantly associated with pervasive developmental disorder at 24 months. When associations between maternal prenatal urinary DAP, DMP, and DEP levels and the Bayley Mental and Psychomo tor Development indices and Child Behavior Checklist perva sive developmental disorder score were stratified by child or maternal PON1 genotype, or by umbilical cord or maternal blood PON l activity or quantity, no statistically significant interactions were detected, and stronger associations were not consistently detected among those with lower-activity geno types (i.e., PON1 !92 RR and PON1 ,Q8 TT) or lower enzyme levels (Table 2) (Eskenazi et al. 2010). At age 3.5 years, maternal prenatal urinary DAP, DMP, and DEP levels were not significantly associated with attention problems or ADHD as assessed by the Child Behavior Check list, whether the outcomes were analyzed as continuous or cat egorical variables dichotomized at clinically borderline scores (> 93rd percentile) (Table 2) (Marks et al. 2010). However, several OR point estimates were around 3.0, with wide 95% CIs due to the small number of borderline scores at that age. Prenatal urinary DAP, DMP, and DEP metabolite levels were also unassociated with the NEPSY-II visual attention score at age 3.5 years. Associations with Child Behavior Checklist measures of attention problems and ADHD at age 5 years were attenuated and statistically non-significant in analyses SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00097 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 628 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 o f dichotomized scores at age 5 years, but significant positive increase in DAPs = --0.27, 95% Cl = --0.48, --0.06; beta associations were detected with scores analyzed as continu for DMPs = - 0 .2 4 , 95% Cl = - 0 .4 2 , -0 .0 5 ; beta for); ous outcomes (beta per 10-fold increase in DAPs = 0.7, 95% between maternal prenatal DMPs and child DEPs and resting Cl = 0.2, 1.2 for attention problems; beta = 1.3, 95%: Cl = 0.4, pre-ejection period at 1 year (beta for prenatal DMPs = 3.77 2.1 for ADHD). Whereas no significant associations were milliseconds, 95% Cl = 0 .2 1 ,7.33; beta for child DEPs = 4.33 detected between maternal prenatal urinary DAP, DMP, or milliseconds, 95% Cl = 1.24,7.42); between maternal prenatal DEP metabolite levels and markedly atypical scores for omis DMPs and reactive pre-ejection period at 6 months (beta = 1.2 sions, commissions, or hit reaction time on the Conners' Kid milliseconds, 95% Cl = 0.03, 2.40); between maternal prena die Continuous Performance Test or the ADHD Confidence tal DAPs and DMPs and reactive respiratory sinus arrhythmia Index analyzed as the continuous variable at 5 years, the odds at 1 year (beta for DAPs = 0.24, 95% Cl = 0.03, 0.46; beta for of having an ADHD Confidence Index above the 70th per DMPs = 0.25, 95% Cl = 0.05, 0.45); and between cumulative centile (OR per 10-fold increase in DAPs = 5.1, 95% Cl = 1.7, maternal prenatal DEPs and resting heart rate (beta = --3.19 15.7), a Hillside Behavior Rating Scale attention problems beats per minute, 95% Cl = --6.29, --0.09). Otherwise, all score > 7 out of 12 (OR for DAPs = 3.0, 95% Cl = 0.9, 9.8), tested associations were statistically non-significant, and esti or a positive composite ADHD indicator (OR for DAPs = 3.5, mated coefficients showed no consistent direction of associa 95%: Cl = 1.1, 10.7) were all at least marginally significantly tion. When basic measures o f autonomic nervous system func increased in association with higher prenatal metabolite con tion at 6 months, 1 year, 3.5 years, and 5 years u ere combined centrations. Some heterogeneity was detected by sex, with into four profiles (coactivation of both sympathetic and para boys generally showing stronger associations than girls. Asso sympathetic nervous systems; coinhibition of both nervous ciations with child urinary OP metabolite levels were weaker systems; reciprocal activation of parasympathetic and inhibi and not statistically significant. tion of sympathetic nervous systems; or reciprocal activation At age 7 years, significant inverse associations were of sympathetic and inhibition of parasympathetic nervous found between maternal prenatal urinary levels of DAPs, systems), no significant differences in geometric mean urinary DMPs, and DEPs and Wechsler Intelligence Scale measures DAP concentrations were found based on maternal prenatal o f Working Memory (e.g., beta per log]0-unit increase in or child specimens, nor were these profiles associated with DAPs averaged from the first and second halves of preg consistently high versus low urinary DAP metabolite levels in nancy = --4.3, 95%; Cl = --7.7, - 0 .9 ) , Processing Speed gestation, childhood, or both. (beta for averaged DAPs = --3.4, 95% Cl = --6.8, --0.1), As with other prospective birth cohort studies described Verbal Comprehension (beta for averaged DAPs = --5.3, earlier in the section on birth outcomes, the main strengths 95% Cl = --8.6, --2.0), Perceptual Reasoning (beta for aver and limitations of the CHAMACOS study were discussed aged DAPs = - 4 . 0 , 95% Cl = - 7 .9 , - 0 .1 ) , and Full-Scale above, with perhaps a lower probability of selection bias with IQ (beta for averaged DAPs = - 5 . 6 , 95% Cl - 9.0, - 2 .2 ) respect to enrollment rates but a higher probability in terms (Table 2) (Bouchard et al. 2011). When maternal averaged of follow-up rates, and greater concerns about multiple com prenatal urinary DAP levels were categorized by quintile, parisons due to the larger number of neurodevelopmental inverse exposure-response trends were observed for all five risk factors assessed. Overall, the neurodevelopmental results outcomes, with an average difference of 7.0 Full-Scale IQ from CHAMACOS varied by outcome, metabolite type, age points between the highest and lowest quintiles of prenatal group, and timing of exposure assessment. The associations DAPs. Estimates of association did not differ substantially of maternal prenatal urinary DAPs, DMPs, and DEPs with a among maternal early prenatal, late prenatal, and postna higher number of abnormal reflexes at age > 3 days were fairly tal urinary DAP concentrations, nor were marked changes consistent across metabolites and outcome classifications, but observed after additional adjustment for other environmen they should be balanced against the null associations with tal contaminants, standardization by creatinine, stratifica abnormal reflexes at < 3 days and with other neonatal behav tion by sex, or restriction to Spanish-speaking children. ioral outcomes (except autonomic stability at < 3 days, which However, child urinary DAP levels at 6, 12, 24, 42, or was positively associated with maternal prenatal urinary DEP 60 months, or at all ages taking the area under the concentration- levels). Another noteworthy finding is the positive association time curve, were not significantly associated with any o f the of maternal prenatal DAP and DMP levels and child DAP, Wechsler Intelligence Scale measures at age 7 years. DMP, and DEP levels (but not MDA or TCPy levels) with the Associations of maternal prenatal and child urinary DAPs, risk of pervasive developmental disorder score above the clini DMPs, and DEPs with both resting and reactivity measures cal cutoff at 24 months. These results are notable and warrant of respiratory sinus arrhythmia, heart rate, and pre-ejection further evaluation, although the reliance on mother-reported period were tested at ages 6 months, 1 year, 3.5 years, and symptoms to classify this outcome leaves open the possibility 5 years (Table 2) (Quiros-Alcala et al. 2011). In addition, of misclassification, whether non-differential or differential cumulative measures of prenatal (14-week and 26-week) by exposure status. The inverse associations of prenatal DAP and childhood (6 months to 5 years, based on area under the and DMP levels with the Bayley Mental Development Index at concentration-time curve calculations) urinary DAP, DMP, 24 months are less compelling, given the opposite associations and DEP metabolite levels were analyzed with respect to with child metabolite levels and the statistically null associa resting and reactivity measures at age 5 years. Among the tions with prenatal DEP, MDA, and TCPy levels as exposures, numerous associations tested, significant associations were and with the Mental Development Index at 6 and 12 months as found only between child DAPs and DMPs and resting respi outcomes. The results for autonomic nervous system function ratory sinus arrhythmia score at 6 months (beta per logi()-unit were consistently null. SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00098 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 629 Two sets of striking associations with adverse neurode- of being classified as hypotonic, compared with social/easy- velopmental outcomes were reported in CHAMACOS. The going, was detected among infants whose mothers had higher first were the positive associations between prenatal DAP and creatinine-standardized urinary DEP levels at 16 weeks (OR DMP levels and continuous Child Behavior Checklist scores per log,-unit increase = 0.89,95% Cl = 0.81,0.99). Otherwise, for attention problems and ADHD and dichotomized indica no significant associations were observed between maternal tors for ADHD Confidence Index, Hillside Behavioral Rating prenatal urinary DAPs, DMPs, or DEPs at any time point Scale attention problems, and a composite ADHD indicator and the odds of being classified as hypotonic or high-arousal/ at 5 years. Significant associations were not detected with difficult, although several borderline significant associations dichotomized Child Behavior Checklist scores for attention w'ere found in both directions, with no apparent consistency problems and ADHD, dichotomized atypical scores on the by exposure or outcome. Conners' Kiddie Continuous Performance Test, and continu The main strengths and limitations of the HOME Study ous ADHD Confidence Index, although point estimates were W'ere discussed earlier and previous comments also apply to generally in the positive direction. However, associations were this analysis. The use of only three profiles to classify neu attenuated and mostly non-significant for child urinary DAP, robehavior in secondary analyses may be an oversimplifica DMP, and DEP levels. The somewhat inconsistent findings tion of a complex neurobehavioral scale (Lester et al. 2004). raise the question o f whether some measures are more valid Although several results were generally in the same direction, than others for capturing ADHD risk, and the heterogeneity with higher maternal prenatal urinary DAP or DEP levels of associations between boys and girls-- w'ith some inverse being associated with better neurobehavioral outcomes (i.e., and mostly nearly null point estimates among girls--is not increased attention, decreased lethargy, decreased hypotonia, readily explained by known biological mechanisms. The other and decreased autonomic stress), these significant associations salient results in CHAMACOS were the inverse associations w'ere selected among many others that were tested and found to of maternal prenatal (but not child) urinary DAP, DMP, and be null. Thus, these findings cannot reliably be interpreted as DEP levels with all five Wechsler Intelligence Scales at age demonstrating a beneficial causal effect of prenatal OP insecti 7 years. The consistency of these findings is unlikely to be cide exposure on behavioral neurodevelopment in infants. due to chance. The methodological limitations of this study, especially with regard to OP insecticide exposure assessment, prevent a causal interpretation of these findings, but the robust Children Pesticide Survey associations with impaired behavioral and cognitive develop From the Children Pesticide Survey, a cross-sectional study ment in school-aged children in CHAMACOS warrant atten of children living in an agricultural community in southern tion and replication in independent studies. Arizona in 1998-2000, a subgroup of 25 school-aged chil dren was selected for analysis based on detectable DAP levels Health Outcomes and Measures of the Environment Study ( > 25 pg/mL; metabolite not specified) in an initial urine sam ple, and 23 other children were selected who had undetectable The HOME Study, described earlier, included 350 mothers levels (Table 1) (Lizardi et al. 2008). Subsequently, urinary who provided urine specimens at 16 4 and 26 4 weeks of DAPs were re-measured in a first-void urine sample, and a gestation, and whose infants completed the NICU Network cognitive assessment was conducted on the same day using Neurobehavioral Scale at home at approximately 5 weeks of a short form of the Wechsler Intelligence Scale for Children age (Table 1) (Yolton et al. 2013). The scale covers 13 dimen Third Edition, the Children's Memory Scale, tire Wisconsin sions: habituation (excluded from analysis because it w'as Card Sorting Test, and the Trail Making Test A and B. In addi omitted for sleeping infants), attention, arousal, self-regula tion, the Child Behavior Checklist 4-18 and the Teacher Report tion. need for special handling from the examiner, quality of Form were used to assess behavioral outcomes. Based on the movement, excitability, lethargy, non-optimal reflexes, asym urine samples collected on the day of the cognitive assess metrical reflexes, hypertonicity, hypotonicity, and stress/absti- ment, all 48 children had detectable levels of DMP, although nence. In multivariate regression models, significant associa average levels remained significantly higher in the originally tions were detected between creatinine-standardized maternal designated "exposed" group (mean = 1 1 0 pg/L, 95% Cl = 83, prenatal urinary DEP levels averaged over 16 and 26 weeks 139) than in the originally designated "unexposed" group and increased attention (beta per log2-unit increase = 0.066, (mean = 49 pg/L, 95% Cl = 36, 63) after excluding one outlier SE = 0.033); between DEP levels at 16 weeks and decreased from each group (519 pg/L in the "exposed" group and 850 lethargy (beta = --0.069, SE = 0.034) and decreased hypoto pg/L in the "unexposed" group). nia (beta = --0.101, SE = 0.045, with hypotonia dichotomized Although children in the "exposed" group took significantly as none vs. any); and between DAP levels at 16 weeks and longer time (mean = 283 seconds, 95%; Cl = 224, 341) to com decreased autonomic stress (beta = --0.010, SE = 0.004) plete the Trail Making Test B than children in the "unexposed" (Table 2). No other significant associations were detected group (mean = 204 seconds, 95% Cl = 172, 236), none of the between maternal prenatal urinary DAPs, DMPs, or DEPs at other cognitive or behavioral measures differed significantly 16 weeks, 26 weeks, or the average of the two, and any of the between the groups in unadjusted analyses, excluding the two other eight dimensions assessed. In secondary analyses, latent outliers (Table 2). Concurrent urinary DAP levels (analyzed as profile analysis was used to group infants together based on the sum of all six metabolites) were modestly and statistically NICU Network Neurobehavioral Scale scores into one of three significantly correlated with some measures of the Wiscon patterns: social/easy-going (n = 157), hypotonic (n = 110), or sin Card Sorting Test (p = 0.31-0.38, P ^ 0 .0 3 ), but not after high-arousal/difficult (n --83). A significantly decreased odds exclusion of the tw'o outliers. Moreover, no significant correla- 8 IQ H I S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00099 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 630 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 lions were detected with the other cognitive measures, includ 95% Cl = 1.06, 4.40 for DEPs), whereas only DMPs were ing the Wechsler Intelligence Scale, the Children's Memory marginally significantly associated with the inattentive sub- Scale, and both Trail Making Tests. Correlations between type of ADHD (n = 69 children; OR = 1.47, 95% Cl = 0.99, concurrent urinary DAP levels and behavioral measures were 2,19) and no metabolites were significantly associated with not estimated. the combined hyperactive/impulsive and inattentive subtype A key limitation of this study is its cross-sectional design: of ADHD (n = 2 9 children; OR = 1.30, 95% Cl = 0.48, ?'.48 because exposures and outcomes were measured on the same for DMPs). day, a cause-and-effect relationship cannot be established. This study is strengthened by its population-based sample Reverse causality due to an influence of childhood behavior selection and by the availability of detailed interview and on diet, as the major source o f OP exposure, is plausible. physical examination data to adjust for potential confounders Even without such an effect, it seems unlikely that DAP (albeit not diet [Millichap and Yee 2012]). metabolites are etiologically relevant to cognitive perfor Major methodological limitations are the cross-sectional mance measured on the same day. Other limitations include design and the measurement of urinary DAPs at a single the lack of adjustment for any con founders, the small study point in time. Outcomes were classified based on parent- or size (resulting in unstable estimates and, possibly, insuffi caretaker-reported symptoms, w'hich could have been differen cient statistical power to detect any associations), the large tially misclassified if, for example, accuracy of reporting var number of outcomes tested (resulting in the expectation of ied by dietary patterns or oilier lifestyle characteristics related several chance findings), and the use of a single sample of to OP insecticide exposure. As mentioned by the authors, the urinary DAP levels. In particular, the fact that the originally observed associations might be due to reverse causality--i.e., designated "unexposed" group had detectable urinary DAP ADHD-related behaviors, such as dietary changes (Millichap levels at the second assessment underscores the intra-individ and Yee 2012)-- that could result in higher exposure to OP ual variability of these metabolites. Participation rates were insecticides and their metabolites. Participation rates among not reported, precluding an assessment of potential selection ADHD and non-ADHD children were unknown, given that bias. These predominantly null results add little insight into ADHD diagnosis was predicated on participation; therefore, possible adverse neurodevelopmental effects o f exposure to the potential for selection bias could not be assessed. The OP insecticides. authors did not suggest a mechanism to explain the stronger associations of DAP metabolites with the hyperactive/impul National Health and Nutrition Examination Survey sive subtype of ADHD than others. Overall, the results o f this study indicate a positive association between DAP metabolite The NHANES is a continuous series of population-based health levels and the prevalence of ADHD, but causal inference about surveys designed to assess the health and nutritional status the effects of OP insecticide exposure is limited by the cross- of approximately 5000 representative, non-institutionalized sectional study design. adults and children in the United States each year (Table 1) (Bouchard et al. 2010). In 2000-2004, NHANES data on six urinary DAP metabolites and ADHD were available for 1,139 Shanghai cross-sectional study children (119 with ADHD) aged 8-15 years, w'here ADHD In a cross-sectional study of 301 healthy 2-year-olds diagnostic status during the previous year was assessed based on symptoms reported by the mother or another caretaker in recruited in 2008 from two community hospitals in Shang hai, urinary levels o f five DAP metabolites were measured in a telephone interview' using the Diagnostic Interview Sched spot urine samples on the same day on which a neurological ule for Children IV (using slightly modified criteria from the DSM-1V), or based on reported use of ADHD medica assessment of motor behavior, adaptive behavior, language behavior, and personal and social behavior was completed tion. Geometric mean urinary levels, which were measured using the Gesell Developmental Schedules for 0- to 3-year- 2-3 weeks before the interview', were 68.3 nmoI/L (IQR = 24.4-186.0) for DAPs, 41.3 nmol/L (fQR = 10.1-130.7) for DMPs, and 11.0 nmol/L (IQR - 2.1-35.0) for DEPs. old children (Table 1) (Guodong et al. 2012). Geometric mean urinary levels were 2.52 fig/L (IQR = < 2 .0 [LOD]3.41) for DMP, 1.56 flg/L (IQR 1.0-1.63) for DMTP, A 10-fold increase in urinary DAP or DMP metabolite lev els was associated with a significantly increased prevalence 1.78 flg/L (IQR = < 1.0-2.89) for DEP, and 3.18 flg/L (IQR = < 1.0-7.26) for DETP; DEDTP was detected in only of ADHD as defined based on diagnostic interview criteria or 2,7% of subjects. No significant associations were observed ADHD medication use (adjusted O R = 1.35, 95% C I= 1.10, 1.67 for DAPs; OR = 1.72, 95% Cl = 1.31, 2.28 for DMPs), between a log10-unit increase in creatinine-adjusted urinary DAPs, DMPs, or DEPs and any of the four Gesell Devel or based on diagnostic interview criteria alone (Table 2). opmental Schedule scores, and estimated coefficients were A positive exposure-response gradient was observed across undetectable, below-median, and above-median urinary lev not consistently above or below zero across metabolites or outcome measures (Table 2). els of DMPs. Urinary DEP metabolite levels were not sig The high participation rate in this study (97%) mini nificantly associated with prevalent ADHD (OR = 0.80, 95% 1 = 0.60, 1.05). However, urinary DAPs, DMPs, and DEPs mizes selection bias, but the cross-sectional design and reliance on a single biospecimen remain the major limita were all significantly associated w'ith a higher prevalence of tions. information on confounders was somewhat limited, the hyperactive/impulsive subtype of ADHD (n = 21 chil dren; OR per 10-fold increase = 1.85, 95% Cl = 1.04, 3.27 for although several covariates were included in multivariate models, and the direction and magnitude of any uncon DAPs; OR = 2.13,95% Cl = 1.08,4.20 for DMPs; OR = 2.15, trolled confounding are unpredictable. Again, the lack of SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00100 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 631 prospectively collected serial OP metabolites prevents this study from fully assessing the associations between expo sure to OP insecticides and neurodevelopmental behavioral outcomes in young children. Canadian health m easures survey In the first cycle (2007-2009) of the cross-sectional Cana dian Health Measures Survey, the Canadian counterpart to NHANES, 1081 children aged 6 - 1 1 years were enrolled, including 1030 (95%) with spot urine measurements of six DAP metabolites measured within two weeks of parental completion of the Strengths and Difficulties Questionnaire to assess mental and behavioral outcomes (Table 1) (Oulhote and Bouchard 2013). The five-dimension scales of this question naire evaluate emotional symptoms, conduct problems, hyperactivity/inattention, peer problems, and prosocial behavior (not analyzed due to insufficient variability), each scored on a 10-point scale; a global total difficulties scale is computed based on the sum of all scales except prosocial behavior. .Scores were dichotomized between high and low/borderline using cutoffs recommended by the author of the instrument. Of the eligible children, 779 (72%) had complete covariate data and were included in the analysis. The median urinary level of DAPs was 99.2 nmol/L (l'QR = 34.3-273.3), that of DMPs was 62.0 (IQR --18.7-192.8), and that of DEPs w'as 25.0 (IQR = 10.5-51.3). When analyzed on the lo g U) scale and adjusted for multiple covariates, with or without creatinine standardization, urinary DAPs, DMPs, and DEPs were all statistically unassociated with elevated scores for total difficulties (OR for DAPs = 0.6, 95% C l - 0 .3 , 1.3; OR for D M Ps- 0 .8 , 95% C l - 0 .4 , 1.6; OR for DEPs --0.3, 95% Cl = 0.1, 1.8), conduct problems, emotional symptoms, hyperaetivity/inattention, and peer problems (Table 2). No significant heterogeneity was observed by child sex. The methodological strengths and limitations of this studyare essentially the same as those of the NHANES studydescribed above (Bouchard et al. 2010). Advantages include the population-based setting and extensive information on potential confounders (but not diet), w'hereas major drawbacks include the cross-sectional design and the one-time spot urine measurement of DAP metabolites. Parent-reported outcome measures were subject to misclassification that might have been differential. Selection bias could have influenced the results in unpredictable ways if participation in the Canadian Health Measures Survey or provision o f complete covariate data were related to both the exposure and the outcome. In light of these limitations and the statistically null findings, this study offers no evidence to support a causal effect of OP insecticide exposure on behavioral problems in children. E arly life E xposed in M exico to E nvironm ental Toxicants Study The Early Life Exposed in Mexico to Environmental Toxi cants (ELEMENT) study sequentially enrolled 827 healthy pregnant women from a general hospital and affiliated clinics in Mexico City (Table 1) (Fortenberry et al. 2014). Of the original cohort participants, 187 (23%) mother-child pairs had third-trimester urine specimens and completed child psy chometric assessments to screen for ADHD-related symptoms at ages 6-11 years in 2007-2011; these assessments included the Conners' Parental Rating Scales-Revised, the Behavior Assessment System for Children-Parental Rating Scales, and Conners' Continuous Performance Test. The Conners' Paren tal Rating Scales-Revised, a parent-completed assessment tool for children and adolescents aged 3-17 years, included scales for an ADHD index, global restlessness/impulsivity, hyperactivity/impulsivity ADHD, inattention ADHD, and combined-type ADHD, mostly based on guidelines from the DSM-IV. The Behavioral Assessment System for ChildrenParental Rating Scales, a parent-completed assessment tool for children aged 6-11 years, were used to assess attention problems and hyperactivity. The geometric mean concen tration of TCPy in maternal prenatal urine was 1.76 ng/mL (IQR --0.91-3.57). In a subset of 21 subjects who provided prenatal urine specimens in all three trimesters, the geometric mean concentration did not vary significantly across trimes ters, but significant within-subject variability w7as detected (intraclass correlation --0.41 without correction for specific gravity and 0.29 with correction). No significant association, including after stratification by sex, was found between maternal prenatal urinary TCPy level and any of the outcome measures studied, including all ADHD and restlessness/impulsivity scales based on the Conners' Parental Rating Scales-Revised; the two scales for attention problems and hyperactivity based on the Behavior Assessment System for Children-Parental Rating Scales; and the clinical index for ADHD and the hit reaction time block change measure (used to assess vigilance or sustained atten tion) based on the Conners' Continuous Performance Test (Table 2). The authors highlighted "suggestive trends'' (with P values >0.05 but < 0 .1 0 ) between maternal prenatal urinary TCPy and increasing hit reaction time block change and the Conners' Parental ADHD index among boys, but no appar ent trends were detected (P --0.18-0.99) for any other ADHD screening measures. The ELEMENT study benefits from prospective collection of prenatal urine specimens, its measurement of a specificmetabolite of chlorpyrifos, and its adjustment for numerous potential confounders (excluding diet). The scope of the study is confined by the measurement of only one OP insecticide metabolite. Other limitations, which are shared by other studies discussed in this review, include the lack of serial biomonitoring, potential selection bias, possible outcome mis classification due to the use of parent-reported data, a modest number of subjects, and multiple comparisons, with no a p r i o r i hypothesis regarding why TCPy should be associated with some measures of ADHD-related symptoms but not others. Thus, chance must he considered as a reasonable explanation for the two marginally significant trends detected among at least 27 tested. In general, the results of this study suggest no consistent or convincing associations between prenatal TCPy levels and ADHD-related symptoms. Shenyang birth cohort In another prospective birth cohort study, 249 healthy pregnant women were enrolled from a hospital in Shenyang, China, between 2011 and 2012 and followed through delivery of a healthy neonate (Table 1) (Zhang et al. 2014). The Neonatal SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00101 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 632 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 Behavioral Neurological Assessment, developed lor Chinese ceding section, outcomes that were uniquely evaluated in only newborns, was administered at 3 days of age to measure one study (e.g., brain morphology (Rauh et al. 2012) and spe functional abilities, reflexes and responses, and behavioral sta cific autonomic nervous system functions (Quiros-Alcala et al. tus based on five scales: behavior, passive tone, active tone, 2011)) are not included in the weight-of-evidence evaluation primary reflexes, and general assessment. Concentrations of because of the absence of independent results for comparison. five DAP metabolites were measured in prenatal maternal urine (timing of collection not specified), with the following geomet Strength. As in the case o f associations with birth outcomes, ric means: 18.03 itg/L (IQR ----7.83-39.43) for DMP, 8.53 pg/L the strength of observed associations between OP metabolites (IQR= 3.4-15.67) for DMTP, 7.14 jig/L (IQR = 3.54-17.17) and neurodevelopmental outcomes cannot be compared read for DEP, 5.64 |ig/L (IQR = 2.34-13.55) lor DETP, and < 1 ily across studies, due to variations in the unit of exposure pg/L (LOD; IQR = LOD-LOD) for DEDTP. measurement, the inconsistent use of logarithmic transforma In adjusted linear regression models, logU)-unit increases tion or creatinine standardization of metabolite levels, out in maternal prenatal urinary levels of DAPs, DMPs, and come measurement and classification methods, and the format DEPs were all significantly associated with lower summary of reported results. "Strong" versus "weak" associations also scores on the Neonatal Behavioral Neurological Assess are not objectively defined, especially for continuous expo ment (beta for DAPs = - 1.78, 95% Cl = - 2 . 1 2 , -1 .4 5 ; sures and outcomes. Even so, most observed associations beta for DMPs = - 0.96, 95% C I= - 1.35, -0 .5 7 ; beta for entail relatively modest changes in outcomes---for example, DEPs = - 0.88, 95% Cl = - 1.30, - 0.47) (Table 2). Signifi ORs and RRs between 0.5 and 2.0, and increases or decre cant inverse associations were also observed between maternal ments of a tew points on a scale standardized to a mean of prenatal DAPs and DEPs and behavior, between DAPs and 100 and SD of 15. Confounding and bias cannot confidently DMPs and passive tone, between DAPs and DMPs and active be ruled out as explanations for associations of such a magni tone, and between DAPs and DMPs and primary reflexes. tude. Several ORs around or above 5.0 were reported in the These associations were detected in both boys and girls, and CHAMACOS study (Eskenazi et al. 2010, Marks et al. 2010, with and without creatinine standardization. Analyses with Rauh et al. 2006, Young et al. 2005), but most of these were maternal prenatal urinary DAP concentrations categorized statistically unstable, with lower 95% confidence limits near or into quintiles were consistent with linear inverse exposure- below 1.0. Although these associations with large ORs merit a response associations with all five outcomes examined. When closer look, most of these and other reported associations are regression coefficients were standardized, the associations statistically non-significant, making them consistent with no between maternal prenatal urinary DAP levels and all out association between OP metabolites and neurodevel opmental comes w'ere stronger than those with gestational age, cord outcomes. blood lead levels, and maternal prenatal BMI. Like other birth cohort studies, the Shenyang study is Consistency. To evaluate the consistency of findings across strengthened by the measurement of urinary DAP metabolite study settings, we assume that neurodevelopmental outcomes levels prior to the measurement of neurological outcomes, evaluated using different assessment tools are reasonably com which rules out reverse causality. However, urine specimens parable. In four studies conducted in four different settings, appear to have been collected from various subjects at dif neonatal behavior was evaluated using the Brazelton Neonatal ferent times throughout pregnancy, and it may or may not be Behavioral Assessment Scale, the NICU Network Neurobe- plausible that exposures at different stages of neurodevelop havioral Scale, and the Neonatal Behavioral Neurological ment would have the same effect on behavioral outcomes. The Assessment (Engel et al. 2007, Yolton et al. 2013, Young et al. participation rate (81%) among eligible women was relatively 2005, Zhang et al. 2014). Three of these four studies found an high, thereby reducing concerns about selection bias, and association between prenatal OP metabolite levels and poorer information was collected on numerous potential confound- reflexes at or shortly after birth (Engel et al. 2007, Young et al. ers, thereby lessening the probability of strong confounding. 2005, Zhang et al. 2014). Three studies also showed no asso However, due to the reliance on a single biospecimen and the ciation with any other adverse neonatal behavioral outcomes measurement of non-specific DAP metabolites, the observed (Engel et al. 2007, Yolton et al. 2013, Young et al. 2005). inverse associations between prenatal DAP metabolite levels The statistically null results for poorer reflexes in the HOME and neonatal behavioral outcomes cannot reliably be inter Study (Yolton et al. 2013), in which newborns were older at preted as causal. In addition, the applicability of the outcome the time of assessment than those in the other three studies, assessment instrument outside of China, where it was devel may suggest that the association is no longer detectable by oped and tested, is unknown. age 5 weeks. Alternatively, the heterogeneity might be due to chance, confounding or bias, or true differences in study Bradford Hill evaluation of weight o f evidence populations or assessment tools. Among infants and toddlers evaluated in four different study Some measures of neurodevelopmental outcomes, including settings, behavioral outcomes were measured using the Bayley the Brazelton Neonatal Behavioral Assessment Scale, the Scales of Infant Development and the Gesell Developmental Bayley Scales of Infant Development, the Wechsler Intelli Schedules (Engel et al. 2011, Eskenazi et al. 2010, Eskenazi gence Scales, the Conners' Parent Rating Scales, and the Child et al. 2007, Guodong et al. 2012, Lovasi et al. 2011, Rauh Behavior Checklist, were used in more than one study, but sev et al. 2006). Although all three studies that measured pre- or eral were not. Although all relevant studies of OP metabolites perinatal OP metabolites and used the Bayley Scales of infant and neurodevelopmental outcomes were described in the pre Development found a significant inverse association between SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00102 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 633 OP metabolite levels and scores on the Mental Development at ages 24 months and 3.5 years (Eskenazi et al. 2007, Marks Index (Engel et al. 2011, Eskenazi et al. 2007, Ranh et al. et al. 2010), the Canadian Health Measures Survey at ages 6-11 2006), this apparent consistency is no longer evident after a years (Oulhote and Bouchard 2013), or the ELEMENT study closer examination of results. Specifically, the CCCEH study at ages 6-11 years (Fortenberry et al. 2014). The consistency detected an association at 36 months among African American of results across studies is difficult to judge, due to differences children but not at 12 or 24 months or in Dominican children in measurement instruments, analytic approaches, the timing (Rauh et al. 2006); the Mount Sinai CECS detected an asso of metaboli te measurement, and the timing of neurodeveiop ciation at 12 months but not at 24 months among black and mentai assessment; the internal inconsistency of the findings Hispanic children, and an association in the opposite direction in the CHAMACOS cohort also complicate interpretation. at 12 months among white children (Engel et al. 2011); and Overall, the findings for ADHD and attention problems were the CHAMACOS cohort found an association at 24 months approximately equally balanced between positive and null. but not at 6 or 12 months (Eskenazi et al. 2007). Thus, none Other behavioral problems in preschool- and school-aged of these studies detected persistent decrements in the Mental children were measured in three study settings using the Child Development Index related to OP insecticide exposure across Behavior Checklist and the Strengths and Difficulties Ques infancy and early childhood age groups. No adverse cross- tionnaire (Eskenazi et al. 2010, Eskenazi et al. 2007, Oulhote sectional associations between child urinary OP metabolite and Bouchard 2013, Rauh et al, 2006). The only specific levels and mental development at 24 months were reported in behavioral outcome measured in more than one study was per CHAMACOS (Eskenazi et al. 2007) and the Shanghai study vasive developmental disorder based on the Child Behavior (Guodong et al. 2012), and most (three out of four) studies Checklist, which was positively associated with pre- or peri did not detect any significant associations with infant psycho natal OP metabolite levels in the CCCEH study (Rauh et al. motor development (Engel et al. 2011, Eskenazi et al. 2007, 2006) and the CHAMACOS study (Eskenazi et al. 2007). Guodong et al. 2012). Although scores from the Strengths and Difficulties Question Four studies in four separate settings assessed cognitive naire have been shown to be highly correlated with those from outcomes in preschool- and school-aged children using the the Child Behavior Checklist (Goodman and Scott 1999), it is Wechsler Intelligence Scales, the Children's Memory Scale, unclear whether the global total difficulties scale-- which was the Wisconsin Card Sorting Test, and the Trail Making Tests, not significantly associated with child urinary DAP, DMP, or although only the Wechsler Scales were used in more than one DEP metabolite levels in the Canadian Health Measures Sur study (Bouchard et al. 2011, Engel et al. 2011, Lizardi et al. vey (Oulhote and Bouchard 2013)--is comparable to that for 2008, Rauh et al. 2011). Two studies found an inverse associa pervasive developmental disorder based on the Child Behavior tion between prenatal OP metabolite levels and the Wechsler Checkhst. Working Memory Index at 7 years (Bouchard et al. 2011, In summary, multiple studies reported a variety of associa Rauh et al. 2011), but one study did not (Engel et al. 2011), tions of OP metabolites with poorer reflexes in neonates and and another found no association based on child DAP levels working memory, perceptual reasoning, measures of ADHD (Lizardi et al, 2008). In addition, one study found an inverse and attention problems, and pervasive developmental disorder association with the Wechsler Perceptual Reasoning Index at in school-aged children. However, this apparent consistency age 7 years (Bouchard et al. 2011) and another detected that was detected among only three studies for neonatal reflexes association among P O N 1 ,92 QQ carriers (Engel et al, 2011), and selected measures of ADHD and attention problems, and whereas no significant associations were detected in the other only tw'o studies for working memory, perceptual reason tw'o studies (Lizardi et al. 2008, Rauh et al. 2011). Three of four ing, and pervasive developmental disorder. In addition, there studies detected no significant associations between prenatal were no associations across three studies for adverse neonatal or child OP metabolite levels and the Wechsler Full-Scale IQ, behavioral outcomes, infant psychomotor development, other Processing Speed, and Verbal Comprehension Scales (Engel measures of ADHD and attention problems in school-aged et al, 2011, Lizardi et al. 2008, Rauh et al. 2011). children, and full-scale IQ, processing speed, and verbal com Six studies in five settings evaluated ADHD and other prehension in school-aged children. attention problems in preschool- and school-aged children When studies were closely compared according to design, using the Child Behavior Checklist, the NEPSY visual exposure metric, timing of exposure measurement, age group attention subtest, the Conners' Parental Rating Scales and of subjects, and neurodeveiopmentai test, at most only two Continuous Performance Test, the Hillside Behavior Rating studies were directly comparable. That is, the Mount Sinai Scale, composite ADHD indices, the Diagnostic Interview CECS (Engel et aL 2007 2011) and CHAMACOS (Young Schedule for Children IV, the Strengths and Difficulties Ques et al. 2005, Eskenazi et al. 2007, Bouchard et al. 2011) both tionnaire, and the Behavior Assessment System for Children used a prospective cohort study design to evaluate prenatal (Bouchard et al, 2010, Eskenazi et al. 2007, Fortenberry et al. maternal DAP, DMP, and DEP levels in association with 2014, Marks et al. 2010, Oulhote and Bouchard 2013, Rauh neurodeveiopmentai outcomes measured using the Brazelton et al. 2006). Significant positive associations between pre Neonatal Behavioral Assessment Scale, the Bayley Scales natal or child OP metabolite levels and some (but not all, in of Infant Development, and the Wechsler Intelligence Scale. the case of CHAMACOS) measures of ADHD or attention Other prospective birth cohort studies used some of these problems were detected in the CCCEH study at age 36 months neurodeveiopmentai tests but not the same exposure metrics (Rauh et al. 2006), in the CHAMACOS cohort at age 5 years (Rauh et al, 2006 2011, Lovasi et al. 2011), the same exposure (Marks et al. 2010), and in NHANES at ages 8-15 years metrics but different neurodevelopmental tests (Yoiton et al. (Bouchard et al. 2010), but not in the CHAMACOS cohort 2013), or different exposure and outcome measures (Forten- SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00103 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 634 R. Reiss et al. berry et al, 2014). Thus, using our a.priori requirement of three independent studies to evaluate the weight of epidemiologic evidence (stated in the "Scope of review" section), the avail able data are insufficient to establish consistent associations between specific OP metabolites and specific neurodevelopmental outcomes. Temporality, issues related to the temporality of measured OP metabolites and neurodevelopmental outcomes are simi lar to those discussed above in the evaluation of associations with birth outcomes, except that perinatal or early childhood exposures could plausibly be related to subsequent neurode velopment. However, because little is known about the tim ing of various neurodevelopmental impairments, it is unclear whether environmental exposures in early gestation, late gestation, infancy, early childhood, or later childhood--or perhaps a combination of these---are most etiologically rel evant. The paucity of knowledge about possible biological mechanisms and latency periods in neurodevelopment may justify the practice of multiple comparisons, with exposures and outcomes measured at multiple time periods being tested for associations. However, the pitfalls of this approach should be acknowledged; it is scientifically invalid to test numerous associations and choose the statistically significant ones as being the etiologically correct ones while dismissing the sta tistically non-significant associations. Crit Rev Toxicol. 2015; 45(7): 531-641 In NHANES data, ordinally increasing categories--unde tectable, below median, or above median--o f urinary DMTP (the only metabolite evaluated as a categorical variable) were associated with progressively higher ORs for parent-reported prevalent ADHD, suggesting a monotonic gradient (Bouchard et al. 2010). In school-aged Mexican children, no statistically significant monotonic trends were detected across increasing tertiles of maternal prenatal urinary TCPy levels and vari ous screening measures of ADHD (Fortenberry et al. 2014). Borderline significant trends were detected with increasing scores on the Conners' Parental ADHD index among boys (Pjrerd = 0.06) and increasing hit reaction time block change on the Conners' Continuous Performance Test among boys and girls combined, but evidence of non-monotonicity was detected for both outcomes among girls. Finally, analyses based on maternal prenatal urinary DAP levels categorized into quintiles clearly illustrated monotonic inverse associations with the behavior, passive tone, active tone, primary reflexes, and summary scores on the Neonatal Behavioral Neurological Assessment among Chinese neonates, supporting the results of linear regression models (Zhang et al. 2014). Overall, then, most of this subset of studies detected biological gradients that strengthen the evidence in support of exposure -response relationships between OP insecticide exposure and adverse neurodevelopmental outcomes, although some results were not consistent with monotonic trends. Biological gradient. Although most studies implicitly Plausibility and coherence. The biological plausibility and assumed a log-linear exposure-outcome relationship between coherence of the epidemiologic and toxicological evidence on OP metabolites and neurodevelopmental outcomes, several OP insecticides in relation to neurodevelopmental outcomes explicitly tested for a biological gradient by categorizing was discussed above. The OP insecticide levels measured exposures into at least three ordinal categories and assessing in the epidemiologic studies are far lower than would cause trends across those categories, with mixed results. Specifically, meaningful AChE inhibition based on animal and (limited) in the Mount Sinai CECS, although positive associations were human toxicology data, and lower than has been established detected between a continuous increase in maternal prenatal as clinically significant in animal studies. There are no known urinary DAP and DEP levels and number of abnormal reflexes biological pathways for OP insecticides to cause the neurode- in neonates, the estimated association with DAP concentra velopmental effects examined in the epidemiologic studies. tions was stronger for die second-lowest quartile than the high Although the lack of established pathways does not mean that est (compared with the lowest), and the association with DEP they do not exist, the existing evidence does not support a concentrations w'as stronger for the second-highest quartile causal interpretation. than the highest, suggesting a non-monotonic relationship A consideration in the evaluation of coherence of evidence (Engel et al. 2007). In the same study, the inverse association is w'hether observed interactions between OP metabolite observed between maternal prenatal DAP and DMP levels levels and PON1 activity levels or genotypes suggest greater and the Bayley Mental Development index at 12 months susceptibility to adverse neurodevelopmental effects of OP among black or Hispanic infants appeared to be monotonic insecticides in individuals with low'er PON1 activity levels. when evaluated across tertiles of metabolite levels (Engel Three studies evaluated such interactions. In the Mount Sinai et al. 2011). In the CHAMACOS cohort, exposure-response CECS, significantly stronger positive associations between gradients were tested between quintiles of maternal prenatal maternal prenatal DAP and DMP metabolite levels and hav urinary DAP, DMP, and DEP levels and > 3 versus < 3 abnor ing > 2 versus < 2 abnormal neonatal reflexes were detected mal reflexes in infants aged > 3 days to < 2 months (Young among those with lower levels of maternal plasma PON l et al. 2005), as well as between quintiles of maternal prenatal enzymatic activity, with an increasing exposure-response urinary DAP levels and all five Wechsler Intelligence Scale pattern in the RR across tertiles of decreasing PON1 activity measures at age 7 years (Bouchard et al. 2 0 1 1). Al though (Engel et al. 2007). By contrast, in the same cohort, mixed Wechsler scores were consistently lower in the second than the results were obtained in analyses o f interactions o f maternal third quartile of prenatal DAP levels, significant monotonic prenatal urinary DAP, DMP, and DEP levels with PON! ,g2, trends were detected in all of these analyses, supporting their PO N!L5SM, and PON!_wsc > T polymorphisms and PON 1 validity. No significant trends were detected between ordinal enzymatic activity (Engel et al. 2011). In particular, PON17P2 categories of maternal prenatal urinary MDA and TCPy levels genotype interacted with prenatal DAP and DMP levels in and the Bayley Motor and Psychosocial Development Indices the expected direction (i.e., with stronger adverse associa at 6, 12, and 24 months (Eskenazi et al. 2007). tions in R allele carriers) with respect to the Bayley Mental SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00104 DOI Low-level OP insectide exposures (A) Birth weight (g) Low-level OP insectide exposures 635 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. (B) Reference WoW207(Mt Sinai) Birth length (cm) * i Eskenazi 204 (CHAMACOS) (C) Reference Woff 2007 m . Sinai) { {i 1.1 - : . 0 -0 .9 ~ U .S -0 ? -0 .3 -0 .5 -0 .4 -C .3 - 0 .2 -0.1 0 0 0.1 0.2 0.3 0 4 0.5 0.6 0.7 0.8 0.9 1.0 ' .1 Beta estimate Head circumference (cm) rIiiii Eskenass 2004 (CHAiVACOS) : I ;i::j__________________________________Iiii_________________________________ iI _ .0 - 3 .9 - 0 .8 - 0 .7 - 0 .6 - 0 . 5 - 0 .4 - 0 .3 -0 .2 ~ 0 1 0 .0 0.1 0.2 0.3 0.4 0.5 0.0 0.7 0 .0 0 .9 1.6 Beta estimate Figure 3. Estimated associations between maternal prenatal urinary DAP levels and birth outcomes. Circles indicate estimated regression coefficients (betas), with 95% CIs indicated by whiskers. Exposures are iogl0 DAP concentrations in nmol/g creatinine (Wolff et ai. 2007), nmol/L (Eskenazi et ai. 2004), and nmol/L, creatinine-standardized (Rauch et ai. 2012). A. Associations with birth weight in grants. B. Associations with birth length in centimeters. C. Associations with head circumference in centimeters. D. Associations with ponders] index in grams per cubic centimeter. E. Associations with gestational age in weeks. Development Index at 12 months in blacks and Hispanics, genotype and enzyme measures or neurodevelopmental out hut opposite to the hypothesized direction with respect to comes tested. In the CHAMACOS study, interactions were the Wechsler Perceptual Reasoning Index at 6 -9 years, and examined between maternal prenatal urinary DAP, DMP, no significant interactions were found for the other PON1 and DEP levels and maternal and child PON1 enzyme mea- SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00105 636 R. Reiss et al. m Pondera! index (g/cm3) Crit Rev Toxicol. 2015; 45(7): 531-641 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. (E) Gestational age {weeks} Beta estimate Figure 3. (Continued). surements and genotypes with respect to the Bayley Mental Discussion and Psychomotor Development Indices and Child Behavior Checklist pervasive developmental disorder score (Eskenazi et al. 2010). No apparent interactions or patterns suggest ing higher susceptibility with lower PON1 activity were detected. Altogether, these limited findings do not provide consistent, coherent evidence to support the hypothesis that low PON1 activity levels augment individual susceptibility to impaired neurodevelopment from OP insecticide expo sure. One possible reason for the lack o f consistent evidence for higher susceptibility wi th lower PON 1 activi ty is that not all OP insecticides are detoxified by PON 1 (Coombes et al. 2014). Additionally, Coombes et al. (2014) found that PON1 may not affect metabolism of chlorpyrifos at environmen tally relevant exposures. This paper reviews a large body of epidemiologic literature and weighs the overall evidence using the framework of the Bradford Hill guidelines. In this section, we focus on three prospective cohort studies (CECS, CHAMACOS, and HOME) that we judged to have the most informative design, and that measured maternal prenatal urinary DAP levels prior to birth or neurodevelopmental outcomes. In addition, we focused on associations with total urinary DAPs in all study subjects combined, to facilitate comparisons across studies, because DAPs were the common exposure metric. Where available, we used associations with creatinine-standardized urinary DAP levels and those that were fully adjusted for potential confounders. As showm in Figure 3, these most informative cohort Specificity, experiment, and analogy. As discussed in the evalu ation of the weight of evidence on OP insecticides and birth out comes, no specificity is evident in the relationships between any particular OP insecticide and any particular neurodevelopmental studies on balance reported no consistent significant asso ciations between maternal prenatal urinary DAP levels and birth w'eight, birth length, head circumference, ponderal index, or gestational age. Figure 4 show's that, in general, outcome. Relevant experimental or quasi-experimentai evidence there was also a lack of a significant association between pertaining to low-dose OP insecticide exposure and adverse neu maternal prenatal urinary DAP levels and most measures rodevelopmental outcomes in humans is lacking, and analogies o f neonatal neurodevelopment. Urinary DAP levels were to other neurotoxic or non-neurotoxic prenatal exposures do not significantly associated with increased risk of abnormal convincingly confirm or negate a causal hypothesis. reflexes at birth in tw'o studies (Engel et al. 2007, Young 8 IQ H I S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00106 DOI Low-level OP insectide exposures (A) Reference Outcome Neonatal neurodevelopment 200 Enge; ? {Mi Sinai) EramMon i-iabituahon before disea; ge 200 Young c:(CHAMACO)S; rszsi'on heb:ftaton before ? po/mhe 2013 3 Voften vncinns; MICH Rewom.habituation at wee*.? 2007 Bige; (Mt Sinai) Brazelton one;i:a;ion before oleenergo 2005 2 Voting (CHAMACOiS) Erst/sTon orte:itation before months 2013 voiton {Cincinnati) MICH Network.attention end arousal at weeks 2007 Zim {Mt Sinai) Brazekon mohy performance before di- iierge 2005 7 2 Voting (CHAM.AC S) E re ile n mol;x penerrnarme before mihe 2013 Yoijcn {Cincinnati) N(CU Network.quality of movement at ( reeks 2007 Engel (M: Sinai) EVazeimn rane;e of state before Jst'hertit: 2005 2 .5 Voting CHAMAC(>S) Eraesimn range of state before mor 2013 6 Yolton {Cincinnati) HICU fceiwom axHtabiit-y a: weeks 2007 Engel (Mt Sinai) Brazskon rep::iishori of state before ci?<merge 2005 5 Voting (CHAM.ACC)S) Era/eilen regt;liation o state before ?ronhs 2013 6 (Mon {Cincinnati) fclCUNiSVW.:ii-tewSem<n H woo; 2007 Engel {Ml Snai) Ei'sizeiton soteinorrer; stability before tShcharge 2005 2 Young (CHAMACO)S) Brazelton sutenomlc stability before rnenths 2013 7 YMon {Cincinnati) HICUHmworr ainenomic ekes;: at w 2007 Engel (Mi. Sinei; Era/efron # atnorma: lefie-ms bMeie oischatoe oung 2GG5 } Braeeto ob-om ('R) y oiton 2Ui.i fcICU N* tefie: weeks Low-level OP insectide exposures 637 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. (B) Rsffcfence Enge; 2011(Mi GY'y) O utcom e Infant neurodevelopment 8 % - M y ITiftl'iM: a t 1 2 m o n t h ? 2007 Ek&ni (CHAMACOSiBoyloy at 12 montt Engel 2011(Ml Sinai) EayHy a: 24merits 7 20 03 54 :f:Y/;i C? :GHAMAC ;Eay;ey mema; a: mentf:$ Engs! 2011(Ml Sir:ai) 2 Buymy pay'Murcio? s i months 2007 12 Eskenazi CHAMACOS; Bxwev peychomotor s: months Engel 2011(M. Sy y ) 2-1 Bayley peyehomoior a; months 2007 5 24 Eskenazi :CHAMACOS S*y:ev peyehomotor a: - 7 - 6 --5 - 4 - 3 i -1 0 ! Beta estimate (C) Childhood neurodevelopment Engel 2011(Mr Sinai; 0-5 VVpcYY-y fulixcaie ;Q a? yearn 2011 -7 Bouchard -CHAMACOSi VVsYhsm. in-scaie IQ a yeas 20 3 EhQei : i (Mi. S:i': i! 3 6-0 hVftohsiS:' p*roept:J?i ? ve*rs 20 1 03 7 SoiKhard ! (CHAMAG ; VVechsler peMebiaei M year? Enge; 2011(Mt. Sinai.: 6-3 VVechsier verba; at years 211 Boncham; ! -OHAMACCSI Vvechsiemmrba at ? vea?. Enge: 2011(Mr. Sinai; 7-0 VVocPH-y pnx easing speed at yearn 2011 7 Bouchard (CHAMACCS.; VV&chsier processing speed at years 21 3 EY:Qw! ) : ! (Mr. S:h :) 7 tACmhijarviCMOna memory O? - yper? 20 1 7 & BoaMiAid (CHAMACOS? WW;h<;lerwork:ng memo;y years Figure 4. Estimated associations between maternal prenatal urinary DAP levels and neurodevelopmental outcomes. Circles indicate estimated regression coefficients (betas), except for associations with abnormal reflexes in neonates, where circles indicate estimated relative risks. Whiskers indicate 95% Cls. Exposures are logl0 DAP concentrations in nmol/L (Engel et al. 2007, Engel et al. 2011, Young et ai. 2005, and Bouchard et al. 2011) and in nmol/g creatinine (converted from log2) (Yolton et al. 2013). A. Associations with neurodevelopmental outcomes in neonates. Most quantitative estimates were not reported by Yolton et ai. (2013). who stated that associations not shown were statistically non-significant. B. Associations with neurodevelopmental outcomes in infants. C. Associations with neurodevelopmental outcomes in children. et al. 2005), though not in the third cohort (Yolton et al. 2013). No consistent significant associations were detected between maternal prenatal urinary DAPs and Bayley measures of neurodevelopment in infancy. Several sig nificant associations between prenatal urinary DAPs and Wechsler measures of cognitive development in childhood SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00107 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 638 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7): 531-641 were detected in one study (Bouchard et al. 2011), but not current literature is subject to publication and reporting bias. the other (Engel et al. 2011), although point estimates in It is has been shown empirically that null results in general the latter study w;ere below the null value. are less likely to be reported (Dickersin and Min 1993, Dwan Overall, these three most informative and comparable studies et al. 2008), or if reported, presented in the conclusions (Kyzas did not establish any consistent associations between maternal et al. 2007). prenatal urinary DAP levels and birth or neurodevelopmental outcomes. Although results for abnormal neonatal reflexes and Conclusions poorer childhood cognitive development suggested a possible association, these w'ere not entirely consistent across studies and require independent confirmation. in summary, associations observed between OP metabolites and birth outcomes in epidemiologic studies have been mostly weak or imprecise, inconsistent, temporally ambiguous, not clearly monotonic, not biologically plausible or coherent with toxicological evidence given the estimated degree of AChE inhibition at observed DAP concentrations, and not specific to any OP insecticide or health outcome. Associations between OP metabolites and neurodevelopmental outcomes in observed in these epidemiologic studies likewise do not unequivocally meet any of the Bradford Hill guidelines. Sir Austin Bradford Hill stated that none of these guidelines must necessarily be met to establish a causal relationship: "None of my nine viewpoints can bring indisputable evidence for or against the cause-and-effect hypothesis and none can be required as a sine qua non" (Hill 1965). Some might consider the standards for causality used in this analysis as restrictive (e.g., use of Bradford Hill, criteria for agreement across three independent studies). It is possible that the use of other standards could yield different conclusions. The inconsistencies across the studies also have to be con sidered in light of the lack of a biologically plausible mecha nism for the adverse birth outcomes or neurodevelopmental effects evaluated in the studies. Even far less severe effects, such as mild AChE inhibition, occur at dosages that are sub Recent epidemiologic studies on balance have found weak and inconsistent associations of maternal exposures to OP insec ticides with birth outcome and neurodevelopmental testing results in the offspring. Perhaps the most important limitation o f the extant literature is the exposure classification, which is subject to significant uncertainties due to limited sampling during pregnancy, despite the high temporal variability in exposure. In addition, the available studies cannot differentiate metabolites that form directly on and in food items and are not the result of OP insecticide exposure. Given the heterogeneity across studies in terms of overall design, types of exposure biomarkers assessed, timing of exposure measurement, birth outcomes, neurodevelopmental tests, statistical modeling approaches, and reporting of results, inter-study comparisons are challenging, and consistency of findings has not been established. The available toxicology data show that the dosages required to cause AChE inhibition are far higher than the levels observed in the epidemiologic studies, a finding that raises further uncertainties about the biological plausibility of the epidemiologic findings. Nonetheless, the studies evalu ate potential effects of major public health importance, and some of the findings, particularly poorer reflexes in neonates, ADHD/attention problems, lower cogni Live scores in preschool or school-aged children, and changes in brain morphology, warrant additional study. stantially higher than the OP insecticide levels measured in the epidemiologic studies. A common limitation of existing studies is their reliance on non-specific DAP metabolites measured in one or two urine specimens. Given the high variability of DAP metabolite lev els in time-series analysis, more frequent sampling is needed to more accurately estimate exposure during pregnancy. In addition, studies with repeated serum or plasma measure Acknowledgements The authors gratefully acknowledge the review and com ments from six anonymous reviewers selected by the editor. The manuscript was significantly strengthened as the result of these reviews. The authors also gratefully acknowledge Mr. Rick Nelson at Exponent for providing a careful editing of the manuscript. ments could provide further insight into the relationship of OP exposure with birth outcomes, childhood growth, and neu Declaration of interests rodevelopment. Standardization of exposures and neurodevel The research reported in this paper w;as sponsored by opmental measures would also aid comparisons across studies. CropLife America, which represents agrochemical compa In general, studies must be larger to enable statistically robust nies that manufacture OP insecticides, under a contract with analyses of gene/environment interactions; to this end, pooling Exponent. RJR and MG were subcontractors to Exponent. of study populations might be useful. However, efforts should CropLife America reviewed the paper, but the authors had be made to recognize and adjust for the expected frequency of ultimate authority to determine its content. The author's false-positive results that arise due to multiple comparisons, affiliations are listed on the cover page. RR and ETC w'ork especially in exploratory analyses (Wacholder et al. 2004, for Exponent, which performs risk assessment consulting for Glickman et al. 2014). In light of the existing limitations and companies that produce OP insecticides. RR has served as inconsistency of studies, the body of epidemiologic data avail an expert witness on behalf of defendants for cases involv able at this time does not convincingly demonstrate an effect ing OP insecticides, though not related to the topics in this of low-level OP insecticide exposure on any adverse health paper. RJR has been the Principal Investigator on research outcomes in humans. grants and research donations to the University of Michigan Although or goal was to include all relevant data on this from Dow Chemical Company and Dow AgroSciences. He topic, it is possible that some studies published in non-English has also been appointed by the University of Michigan as the journals were missed in our review. It is also possible that the Dow Professor o f Toxicology, a professorship endow'ed by SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00108 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Low-level OP insectide exposures 639 the Dow Foundation. In addition, he has served as a consul tant and expert witness on behalf of Dow' Chemical Company and Dow AgroSciences. References Abu-Saad K, Fraser D. (2010). Maternal nutrition and birth outcomes. Epidemiol Rev, 32, 5-25. Andres RL, Day MC. (2000). Perinatal complications associated with maternal tobacco use. Semin Neonatol, 5, 231-41. Anjos T, Altmae S, Emmett P, Tierneier H, Closa-Monasterolo R, Luque V, et al. (2013). Nutrition and neurodevelopment in children: focus on NUTR1MENTHE project. Eur I Nutr, 52, 1825-12. Attfield KR, Hughes MD, SpenglerJD.LuC. (2014). Within- and betweenchild variation in repeated urinary pesticide metabolite measurements over a 1-year period. Environ Health Perspect, 122, 201-6. Aviram M, Vaya I. (2013). Paraoxonase 1 activities, regulation, and interactions with atherosclerotic lesion. Cun: Opin Lipidol. 24, 339-44. Barr DB, Ananth CV. Yan X, Lashley S, Smulian JC, Ledoux TA, et al.. (2010). Pesticide concentrations in maternal and umbilical cord sera and their relation to birth outcomes in a population of pregnant women and newborns in New Jersey. Sci Total Environ, 408, 790-5. Berkowitz GS, Wetmur JG, Birman-Deych E, Obi J, Lapinski RH, Godbold JH, et ai. (2004). In utero pesticide exposure, maternal paraoxonase activity, and head circumference. Environ Health Perspect, 112, 388-9!. Berman T, Goldsmith R, Goen T, Spungen J, Novack L, Levine H, et al. (2013). Urinary concentrations of organophosphate pesticide metabolites in adults in Israel: demographic and dietary predictors. Environ Int, 60, 183-9. Bliddal M, Olsen J, Stovring H, Eriksen HE, Kesmodel US, Sorensen Tl, Nohr EA. (2014). Maternal pre-pregnancy BM1 and intelligence quotient (IQ) in 5-year-old children: a cohort based study. PLoS One, 9, e94498. Boon PE, Van der Voet H, Van Raaij MT, Van Klaveren JD. (2008). Cumulative risk assessment of the exposure to organopbospborus and carbamate insecticides in the Dutch diet Food Chem Toxicol, 46, 3090-8. Bouchard M, Gosselin NH. Brunet RC. Samuel O. Dumoulin MJ, Carrier Ci. (2003). A toxicokinetic model of malathion and its metabolites as a tool to assess human exposure and risk through measurements of urinary biomarkers. Toxicol Sci, 73, 182-94. Bouchard ME, Bellinger DC, Wright RO, Weisskopf MG. (2010). Attention-deficit/hyperactivity disorder and urinary metabolites of organophosphate pesticides. Pediatrics, 125, el270-7. Bouchard MF, Chevrier J, Harley KG, Kogut K, Vedar M, Calderon N, et al. (2011). Prenatal exposure to organophosphate pesticides and IQ in 7-year-old children. Environ Health Perspect, 119, 1189-95. Bradman A, EskenaziB, BarrDB, Bravo R. Castorina R, Chevrier J, et ai. (2005). Organophosphate urinary metabolite levels during pregnancy and after delivery in women living in an agricultural community. Environ Health Perspect, 113, 1802-7. Bradman A, Kogut K, Eisen EA, Jewell NP, Quiros-Alcala L, Castorina R, et al. (2013). Variability of organophosphorous pesticide metabolite levels in spot and 24-hr urine samples collected from young children during 1 week. Environ Health Perspect, 121, 118-24. Burkhalter TM, Hillman CH. (2011). A narrative review of physical activity, nutrition, and obesity to cognition and scholastic performance across the human lifespan. Adv Nutr, 2, 201S-6S. Bums Cl, McIntosh LJ, Mink PJ, Jurek AM, Li AA. (2013). Pesticide exposure and neurodevelopmental outcomes: review of the epidemiologic and animal studies. J Toxicol Environ Health B Crit Rev, 16, 127-283. Cantor KP, Blair A, Everett G, Gibson R, Burmeister LF, Brown LM, et al. (1992). Pesticides and other agricultural risk factors for nonHodgkin's lymphoma among men in Iowa and Minnesota. Cancer Res, 52,2447-55. Carr RL, Adams AL, Kepler DR, Ward AB, Ross MK, (2013). Induction of endocannabinoid levels in juvenile rat brain following developmental chlorpyrifos exposure. Toxicol Sci, 135, 193-201. Centers for Disease Control and Prevention (CDC). (2014). National Health and Nutrition Examination Survey. CDC Website. http://www. edc.gov/nehs/nhanes.htm. Chen L, Zhao T, Pan C, Ross J, Ginevan M, Vega H, Krieger R. (2013). Absorption and excretion of organophosphorous insecticide biomarkers of malathion in the rat: implications for overestimation bias and exposure misclassification from environmental biomonitoring. Regul Toxicol Pharmacol, 65, 287-93. Chen L, Zhao T, Pan C, Ross JH, Krieger RL (2012). Preformed biomarkers including dialkylphosphates (DAPs) in produce may confound biomonitoring in pesticide exposure and risk assessment. J Agric Food Chem, 60, 9342-51. Ciaeys WL. De Voghel S, Schmit JF. Vromman V. PussemierL. (2008). Exposure assessment of the Belgian population to pesticide residues through fruit and vegetable consumption. Food Addit Contam Part A Chem Anal Control Expo Risk Assess, 25, 851-63. Coombes RH, Meek EC, Dail MB, Chambers HW, Chambers JE. (2014). Human paraoxanase 1 hydrolysis of nanomolar chlorpyrifos-oxon concentrations is unaffected by phenotype or Q192R genotype. Toxicol Lett, 230, 57-61. Curl CL, Fenske RA, Elgethun K. (2003). Organophosphorous pesticide exposure of urban and suburban preschool children with organic and conventional diets. Environ Health Perspect, 111, 377-82. Dickersin K, Min YI. (1993). Publication bias: the problem that won't go away. Ann NY Acad Sci, 703, 135-46; discussion 146-138. Duggan A, Chamley G, Chen W, Chukwudebe A, Hawk R, Krieger RL, et al. 2003. Dialkyiphosphate biomonitoring data: assessing cumulative exposure to organophosphate pesticides. Regul Toxicol Pharmacol, 37, 382-95. Dwan K, Altman DG, Arnaiz JA, Bloom J, Chan AW, Cronin E, et al. (2008). Systematic review' of the empirical evidence of study publication bias and outcome reporting bias. PLoS One. 3, e.3081. Gonzales ET, Bauer SB. (1999). Pediatric urology practice. Philadelphia, PA: Lippincott Williams , & Wilkins. Eaton DL, Daroff RB, Autrup FI, Bridges J, Bffler P, Costa LG, et al. (2008). Review of the toxicology o f chlorpyrifos with an emphasis on human exposure and neurodevelopment. Crit Rev Toxicol, 38, 1-125. Eddieston M, Buckley NA, Eyer P, Dawson AH. (2008). Management of acute organopbosphorus pesticide poisoning. Lancet, 371,597-607. Engel SM, Berkowitz GS, Barr DB, Teitelbaum SL, Siskind J, Meisel SJ, et al. (2007). Prenatal organophosphate metabolite and organochlorine levels and performance on the Brazelton Neonatal Behavioral Assessment Scale in a multiethnic pregnancy cohort. Am J Epidemiol. 165, 1397-404. Engel SM, Wetmur J, Chen J, Zhu C, Barr DB, Canfield RL, Wolff MS. (2011). Prenatal exposure to organophosphates, paraoxonase 1, and cognitive development in childhood. Environ Health Perspect, 119, 1182-8. Eskenazi B, Harley K. Bradman A, Weitzien E, Jewell NP, Barr DB. et al. (2004). Association of in utero organophosphate pesticide exposure and fetal growth and length of gestation in an agricultural population. Environ Health Perspect. 112, 1116-24. Eskenazi B. Huen K, Marks A. Harley KG, Bradrnan A, Barr DB. Holland N. (2010). PON1 and neurodevelopment in children from the CHAMACOS study exposed to organophosphate pesticides in utero. Environ Health Perspect, 118, 1775--81. Eskenazi B, Marks AR, Bradman A, Harley K, Barr DB, Johnson C, et al. (2007). Organophosphate pesticide exposure and neurodevelopment in young Mexican-American children. Environ Health Perspect, 115, 792-8. Ferioli A, Maroni M. (2011). Biological monitoring. In: La Ferla F, Lauwerys RR, eds. Encyclopedia of Occupational Health and Safety. Geneva, Switzerland: International Labor Organization. Fortenberry GZ, Meeker JD, Sanchez BN, Barr DB, Panuwet P, Bellinger D, et al. (2014). Urinary 3,5,6-trichioro-2-pyridinol (TCPY) in pregnant women from Mexico City: distribution, temporal variability, and relationship with child attention and hyperactivity. Int J Hyg Environ Health, 217, 405-12. Garfitt SJ, Jones K, Mason HJ, Cocker J. (2002). Exposure to the organophosphate diazinon: data from a human volunteer study with ora! and derma! doses. Toxicol Lett, 134, 105-13. Giles D, Dickson J. (2000). A randomized double blind ascending single oral dose study with malathion to determine the no effect level on plasma and RBC cholinesterase activity. Tranent, Scotland: Inveresk Research, Glickman ME, Rao SR, Schultz MR. (2014). False discovery rate control is a recommended alternative to Bonferroni-lype adjustments in health studies. J Clin Epidemiol, 67, 850-7. 8 IO H T S V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00109 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. 640 R. Reiss et al. Crit Rev Toxicol. 2015; 45(7); 531-641 Gonzales ET, Bauer SB. (1999). Pediatric urology practice. Philadelphia, Medina JL,Reinicke K, Simpfendorfer R, Roa A, Oliveros H, Bardisa L, PA: Lippincott Williams & Wilkins. Rudolph MI. (1993). Characterization and distribution of cholinesterase Goodman R, Scott S. (1999). Comparing the strengths and difficulties activity in mouse uterine horns: changes in estrous cycle. Comp questionnaire and the child behavior checklist: is small beautiful? Biochem Phys C, 106, 473-78, J Abnorm Child Psychol, 27. 17-24. Marks AR, Harley K, Bradman A, Kogut K, Barr DB. Johnson C, et al. Gordis L. (2013). Epidemiology. 5th ed. Philadelphia: Elsevier. (2010). Organophosphate pesticide exposure and attention in young Griffith W, Curl CL, Eenske RA, Lu CA, Vigoien EM, Faustman EM. Mexican-American children: the CHAMACOS study. Environ Health (2011). Organophosphate pesticide metabolite levels in pre-school Perspect, 118, 1768-74. children in an agricultural community: within- and between-child Marshall NE, Spong CA? (2012). Obesity, pregnancy complications, and variability in a longitudinal study. Environ Res, 111, 751-6. birth outcomes. Semin Reprod Med, 30, 465-71. Guodong D, Pei W, Ying T, Jun Z, Yu G, Xiaojin W, et al. (2012). Mason JB, Saldanha LS. Ramakrishnan U, Lowe A, Noznesky EA, Organophosphate pesticide exposure and neurodevelopment in young Girard AW, et al. (2012). Opportunities for improving maternal Shanghai children. Environ Sci Technol. 46, 2911-7. nutrition and birth outcomes: synthesis of country experiences. Food Harley KG. Huen K, Aguilar Schell R, Holland NT, Bradman A, Nutr Bull, 33, SI 04-37. Barr DB. Eskenazi B. (2011). Association of organophosphate McDaniel CY, Dail MB, Wills RW, Chambers HW, Chambers JE. (2014) pesticide exposure and paraoxonase with birth outcome in Mexican- Paraoxonase 1 polymorphisms within a Mississippi USA population American women. PLoS One, 6, e23923. as possible biomarkers of enzyme activities associated with disease Hill AB. (1965). The environment and disease: association or causation?. susceptibility. Biochem Genet, 52, 509-23. Proc R Soc Med, 58, 295-300. Mileson BE, Chambers JE, Chen WL, Dettbarn W, Ehrich M, Horton MK, Kahn LG, Perera F, Barr DB, Rauh V. (2012). Does the Eidefrawfi AT, et al. (1998) Common mechanism of toxicity: a case home environment and the sex of the child modify the adverse effects study of organophosphorus pesticides. Toxicol Sci, 41. 8-20. of prenatal exposure to ehlorpyrifos on child working memory? Millichap JG, Yee MM. (2012). The diet factor in attention- Neurotoxicol Teratol, 34, 534-41. deficit/hyperactivity disorder. Pediatrics, 129, 330-7. Jensen AF, Petersen A. Granby K. (2003). Cumulative risk assessment Mink PJ, Kimmei CA, Li AA. (2012). Potential effects of ehlorpyrifos of the intake of organophosphorus and carbamate pesticides in the on fetal growth outcomes: implications for risk assessment. J Toxicol Danish diet. Food Addit Contain, 20, 776-85. E nvH ealB , 15, 281-316. Jensen BH, Petersen A, Christensen T. (2009). Probabilistic assessment Morgan MK, Sheldon LS, Jones PA, Croghan CW, Chuang JC, of the cumulative dietary acute exposure of the population of Denmark Wilson NK. (2011 ). The reliability of using urinary biomarkers to to organophosphorus and carbamate pesticides. Food Addit Contain estimate children's exposures to ehlorpyrifos and diazinon. J Expo Sci Part A Chem Anal Control Expo Risk Assess, 26. 1038-48. Environ Epidemiol, 21,280-90, Jurek AM, Greenland S, Maldonado G. (2008). How far from non Neggers YH, Goldenberg RL, Ramey SL, Cliver SP. (2003). Maternal differential does exposure or disease misciassiftcation have to be to prepregnancy body mass index and psychomotor development in bias measures of association away from the null? Int J Epidemiol, 37, children. Acta Obstet Gynecol Seand, 82, 235-40. 382-5. Nougadere , Sirot V, Kadar A, Pastier A, Trachot E, Vergnet C, et al. (2012). Jurek AM. Greenland S, Maldonado G, Church TR. (2005). Proper Total diet study on pesticide residues in France: levels in food as consumed interpretation of non-differential misclassification effects: expectations and chronic dietary risk to consumers. Environ Int, 45, 135-50. vs observations. Tnt J Epidemiol, 34, 680-7. Oulhote Y, Bouchard ME'. (2013). Urinary metabolites of organophosphate Krasowski MD, Mcgehee DS, Moss J. (1997). Natural inhibitors of and pyretbroid pesticides and behavioral problems in Canadian cholinesterases: implications for adverse drug reactions. Can J Anaesth, children. Environ Health Perspect, 121, 1.378-84. 44,525-34. Ozarowski M, Mikolajczak PL, Bogacz A, Gryszezynska A, Kujawska Krieger RI, Chen L. Ginevan M, Watkins D, Cochran RC, M, Jodynis-Liebert J, et al. (2013). Rosmarinus officinalis L. leaf Driver JH, Ross JH. 2012. Implications of estimates of residential extract improves memory impairment and affects acetylcholinesterase organophosphate exposure from dialkylphosphates (DAPs) and their and butyryieholinesterase activities in rat brain. Fitoterapia, 91, relevance to risk. Regul Toxicol Pharmacol, 64, 263-66. 261-71." Kyzas PA, Denaxa-Kyza D, Ioannidis JP. (2007). Almost all articles on Perera FP. Rauh V, Tsai WY. Kinney P, Camann D, Barr D, et al. (2003). cancer prognostic markers report statistically significant results. Europ Effects of transplacental exposure to environmental pollutants on birth J Cancer, 43, 2559-2579. outcomes in a multiethnic population. Environ Health Perspect, 111, Lefkowitz LJ, Kupina JM, Hirth NL, Henry RM, Noland GY, 201-5. Barbee j YJ, et al, (2007). Intraindividual stability of human erythrocyte Quitos-Alcala L, Alkon AD, Boyce WT, Lippert S, Davis NV, cholinesterase activity. Clin Chem, 53, 1358-1363. Bradman A, et al. (2011). Maternal prenatal and child Lester BM, Tronick EZ, Brazelton TB. (2004). The Neonatal Intensive organophosphate pesticide exposures and children's autonomic Care Unit Network Neurobehavioral Scale procedures. Pediatrics, 113. function. Neurotoxicology, 32, 646-55. 641-67. Rauch SA, Braun JM, Barr DB, Calafat AM, Khoury J, Montesano AM, Li B, Sedlacek M, Manoharan I, Boopathy R, Duysen EG, Masson P, et al. (2012). Associations of prenatal exposure to organophosphate Lockridge O (2005). Butyryieholinesterase, paraoxonase, and albumin pesticide metabolites with gestational age and birth weight. Environ esterase, but not carboxylesterase, are present in human plasma. Health Perspect, 120, 1055-60. Biochem Pharmacol, 70, 1673-1684. Rauh V, Arunajadai S. Horton M, Perera F, Hoepner L, Barr DB, Li AA, Lowe KA, McIntosh LJ, Mink PJ. (2012). Evaluation of Whyatt R. (2011 ). Seven-year neurodevelopmental scores and prenatal epidemiology and animal data for risk assessment: ehlorpyrifos exposure to ehlorpyrifos, a common agricultural pesticide. Environ developmental neurobehavioral outcomes. J Toxicol Environ Health B Health Perspect, 119, 1196-201. Crit Rev, 15, 109-84. Rauh VA, Garfinkel R, Perera FP, Andrews HE, Hoepner L, Barr DB. et al. Lizardi PS, O'Rourke MK, Morris RJ. (2008). The effects of (2006). Impact of prenatal ehlorpyrifos exposure on neurodevelopment organophosphate pesticide exposure on Hispanic children's cognitive in the first 3 years of life among inner-city children. Pediatrics, 118, and behavioral functioning. J Pediatr Psychol, 33, 91--101. el 845-59. Lovasi GS, Quinn JW, Rauh VA, Perera FP, Andrews HE, Garfinkel R, Rauh VA, Perera FP, Horton MK, Whyatt RM, Bansal R, Hao X, et al. et al. (2011). ehlorpyrifos exposure and urban residential environment (2012). Brain anomalies in children exposed prenatally to a common characteristics as determinants of early childhood neurodevelopment. organophosphate pesticide. Proc Natl Acad Sci U S A, 109, 7871-6. Am J Public Health, 101, 63-70. Reiss R. (2012). Benchmark dose modeling for an omethoate acute Lu C, Barr DB, Pearson MA, Waller LA. (2008). Dietary intake comparative cholinesterase study in neonatal and adult rats. Prepared and its contribution to longitudinal organophosphorus pesticide exp for Gheminova, Inc. Exponent report number VA10532.000 BT0 osure in urban/suburhan children. Environ Health Perspect, 116, 0212 KIEL:. 537-42. Rice D, Barone S Jr. (2000). Critical periods of vulnerability for the Macharia M, Kengne AP, Blackhurst DM, Erasmus RT, Matsha TE. developing nervous system: evidence from humans and animal models. (2014). The impact of chronic untreated hypergiycaemia on the long Environ Health Perspect, 108, 511-33. term stability of paraoxonase 1 (PON1) and antioxidant status in Rothman KJ, Greenland S, Lash TL. (2012).Modern epidemiology. human sera. .1Clin Pathol, 67, 55-9. 3rd ed. (revised). Philadelphia, PA: Lippincott Williams & Wilkins SIQHIS V Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 ED 002061 00046431-00110 Critical Reviews in Toxicology Downloaded from informahealthcare.com by Kathi Lohmann on 07/20/15 For personal use only. DOI Low-level OP insectide exposures Sandjaja, Poh BK, Rojroonwasinkul N, Le Nyugen BK, Budiman B, Ng LO, et al. (2013). Relationship between anthropometric indicators and cognitive performance in Southeast Asian school-aged children. Br I Nutr. 110, S57-64. Schrader C, Rimbach G. (2011). Determinants of paraoxonase 1 status: genes, drugs and nutrition. Curr Med Chetn, 18, 5624-43. Sexton K, Ryan AD. (2012) Using exposure biomarkers in children to compare between-chiid and within-child variance and calculate correlations among siblings for multiple environmental chemicals. J Expo Sci Environ Epidemiol .22, 16-23. Slotkin TA, Seidler FJ. (2007). Comparative developmental neuro toxicity of organophosphates in vivo: transcriptional responses of pathways for brain cell development, cell signaling, cytotoxicity and neurotransmitter systems. Brain Res Bull, 72, 232-74. Smithers LG. Lynch JW. Yang S, Dahhou M, Kramer MS. (2013). Impact of neonatal growth on IQ and behavior at early school age. Pediatrics, 132, e53-6oZ Soleeki R. (2002). Pesticide residues in food - 2002 - joint FAO/WHO meeting on pesticide residues - acephate. IPCS International Program on Chemical Safety [Online] Available At: http://www.Inchenr.Org/ Doeuments/Jmpr/Jmpmono/2002prQ2.Htm. Accessed on 14 July 2014. Sorahan T, Gilthorpe MS. (1994). Non-differential misclassification of exposure always leads to an underestimate of risk: an incorrect conclusion. Occup Environ Med, 51, 839-40. Stromberg U, Bjork J, Broberg K, Mertens F, Vincis P. (2008). Selection of influential genetic markers among a large number of candidates based on effect estimation rather than hypothesis testing: an approach for genome-wide association studies. Epidemiology, 19, 302-8. Sudakin DL, Stone DL. (2011). Dialkyl phosphates as biomarkers of organophosphates: the current divide between epidemiology and clinical toxicology. Clin Toxicol, 49, 771-81. Timchalk C, Nolan RJ, Mendrala AL, Dittenber DA, Brzak KA, Mattsson JL. (2002). A Physiologically based pharmacokinetic and pharmacodynamic (PBPK/PD) model for the organophosphate insecticide chlorpyrifos in rats and humans. Toxicol Sci, 66, 34-53. U.S. Department of Agriculture (USDA). (2014). Pesticide data program, annual summary, calendar year 2012. In: Agriculture LTSDO (ed.). Washington, DC: U.S. Department of Agriculture. IJ.S. Environmental Protection Agency (ERA), (2009a). Dimethoate: human health assessment scoping document in support of registration review. USEPA Office of Pesticide Program. U.S Environmental Protection Agency (EPA). 2009b. Registration eligibility decision (RED) for malathion. USEPA Office of Pesticide Program, U.S. Environmental Protection Agency (EPA). (2005). Dimethoate and omethoate: comparative toxicity and determination of toxicity adjustment factors, (addendum to PLED nos. 0050651 and 0050901). USEPA Ofice of Pesticide Programs, U.S. Environmental Protection Agency (EPA). (2011). Chlorpyrifos: preliminary human health risk assessment for registration review. USEPA Ofice of Pesticide Programs (ed.), US Environmental Protection Agency (EPA). (2002). Guidance on cumulative risk assessment of pesticide chemicals that have a common mechanism of toxicity. USEPA Ofice of Pesticide Programs, US Environmental Protection Agency (EPA). (2000). Office of pesticide programs science policy on the use of data on cholinesterase inhibition for risk assessments of organophosphorous and carbamate pesticides. USEPA Office of Pesticide Programs. Wacholder S, Chanock S, Garcia-Closas M, El Ghormli L, Rothman N. (2001). Assessing the probability that a positive report is false: an Low-level OP insectide exposures 641 approach for molecular epidemiology studies. J Natl Cancer Inst, 96, 434-42. Wacholder S, Hartge P. Lubin JH. Dosemeci M. (1995). Non-differential misclassification and bias towards the null: a clarification. Occup Environ Med, 52, 557-8. Wakefield J. (2007). A Bayesian measure of the probability of false discovery in genetic epidemiology studies. Am I Hum Genet, 81, 208-27. Wang P. Tian Y, Wang XJ, Gao Y, Shi R. Wang GQ, et al. (2012). Organophosphate pesticide exposure and perinatal outcomes in Shanghai, China. Environ Int. 42, 100-4. Weitkunat R, Kaelin E, Vuillaume G, Kallischnigg G. (2010). Effectiveness of strategies to increase the validity of findings from association studies: size vs. replication. BMC Med Res Methodol, 10, 47. World Health Organization (WHO). (1996). Biological monitoring of chemical exposure in the workplace, guidelines, volume 1. WHO/ HPR/OCH 96.1. Geneva. Whyatt RM, Camann D, Perera FP, Rauh VA, Tang D. Kinney PL, et al. (2005). Biomarkers in assessing residential insecticide exposures during pregnancy and effects on fetal growth. Toxicol Appi Pharmacol, 206, 246-54. Whyatt RM, Camann DE, Kinney PL, Reyes A, Ramirez I, Dietrich J, et al. (2002). Residential pesticide use during pregnancy among a cohort of urban minority women. Environ Health Perspect, 110, 507-14. Whyatt RM, Rauh V, Barr DB, Camann DE, Andrews FIE, Garfinke! R, et al. (2004). Prenatal insecticide exposures and birth weight and length among an urban minority cohort. Environ Health Perspect, 112, 1125-32. Wickerharn EL, Lozoff B, Shao J, Kaciroti N, Xia Y, Meeker JD. (2012). Reduced birth weight in relation to pesticide mixtures detected in cord blood of full-term infants. Environ Int, 47, 80-5. Wolff MS, Engel S, Berkowitz G, Teitelbaum S, Siskind J, Barr DB, Wetmur J. (2007). Prenatal pesticide and PCB exposures and birth outcomes. Pediatr Res, 61, 243-50. Yang D, Lauridsen H, Buels K, Chi LH, La Du J, Brunn DA, et al. (2011). Chlorpyrifos-oxon disrupts zebrafish axonal growth and motor behavior. Toxicol Sci, 121, 146-59. Ye X, Pierik EH, Hauser R, Duty S, Angerer J, Park MM. et al. (2008). Urinary metabolite concentrations of organophosphorous pesticides, bisphenol A, and phthalates among pregnant women in Rotterdam, the Netherlands: the Generation R study. Environ Res, 108, 260-7. Yolton K, Xu Y. Sucharew H, Succop P, Altaye M, Popeiar A, et al. (2013). Impact of low-level gestational exposure to organophosphate pesticides on neurobehavior in early infancy: a prospective study. Environ Health, 12, 79. Young JG, EskenaziB, Gladstone EA, Bradman A, Pedersen L, Johnson C, et al. (2005). Association between in utero organophosphate pesticide exposure and abnormal reflexes in neonates. Neurotoxicology, 26, 199-209. Zhang X, Drive JH, LI Y, Ross JH, Krieger RT. (2008). Dialkylphosphates (DAPs) in fruits and vegetables may confound biomonitoring in organophosphorus insecticide exposure and risk assessment. .1 Agric FoodChem, 56, 10638-45. Zhang Y, Han S, Liang D, Shi X, Wang F, Liu W, et al. (2014). Prenatal exposure to organophosphate pesticides and neurobehavioral development of neonates: a birth cohort study in Shenyang, China. PLoS One, 9, e88491. Sierra Club v. EPA 18cv3472 NDCA Tier 3/4 BlflHTS V ED 002061 00046431-00111