Reproductive Effects in Men and Women

The Volume 11 committee identified several new studies on reproductive effects and exposure to the OPs of interest. The new studies assessing OP exposure and reproductive effects in men and women are listed in Table 5-2. The effects of OPs on male fertility have been studied by several researchers, including a recent systematic review by Martenies and Perry (2013). In general, human studies have focused on sperm quality associated with either environmental or occupational exposures.

TABLE 5-2

Summary of Reproductive Effects of Organophosphate Pesticides.

Meeker et al. (2004a) assessed TCPy, a urinary marker of chlorpyrifos exposure, in 272 men who were not occupationally exposed to pesticides and who were attending an infertility clinic in Massachusetts. They reported that increasing urinary TCPy levels were associated, although not significantly, with increased risk of having sperm concentrations below 20 million/mL, less than 50% motile sperm, and less than 4% of sperm with normal morphology. Meeker et al. (2004b) further studied sperm DNA integrity, which they assessed by neutral comet assay and reported as comet extent, percentage DNA in comet tail (Tail%), and tail distributed moment (TDM). Statistically significant increases in Tail% for an interquartile range (IQR) increase in TCPy (β=2.8, 95% confidence interval [CI] 0.9–4.6) were reported, while a decrease in TDM was associated with IQR changes in TCPy (β= –2.5, 95% CI –4.7– –0.2).

Using the same population as Meeker et al. (2004b), Figueroa et al. (2015) examined whether exposure to OPs was associated with sperm chromosomal abnormalities—specifically XX18, YY18, XY18, and total disomy in sperm nuclei. Increasing levels of urinary metabolites, both individual and total sum of DAPs, were significantly associated with increased levels of disomy, particularly XX18 and XY18, although the increases were not monotonic. The highest significant association was observed between the third exposure quartile of DMTP (2.21–6.47 ng/mL) and XX18, with a 52% increase in the incidence rate ratio (95% CI 1.36–1.69).

Sperm integrity, specifically the sperm chromatin structure, was also evaluated in 33 men occupationally exposed to OPs in Mexico (Sánchez-Peña et al., 2004). About 75% of the men had semen with altered sperm chromatin structure (>30% of DNA fragmentation index [DFI]) that would classify them as having poor fertility potential, compared with only 4% of the unexposed men. Only the DAP concentrations for DETP were significantly associated with mean DFI (p=0.026) and standard deviation-DFI (p=0.022) and marginally associated with DFI% (p=0.079) after adjustment for potential confounders, such as the presence of immature germ cells (IGCs). No other DAPs had any associations with measured outcomes. No significant associations were found between semen quality parameters, including IGCs, and DETP or any other DAP, although IGC values were above the reference value for 82% of the samples versus only 9% of unexposed men. There was a negative correlation between mean DFI and sperm viability (r=0.254; p<0.027) and also seminal volume (r=0.333; p<0.004).

A further assessment of the effect of exposure to environmental OPs on sperm parameters and male reproductive hormones was conducted by Melgarejo et al. (2015). This was a study of 116 men attending an infertility clinic in southern Spain, all of whom had detectable concentrations of at least one urinary OP metabolite (the possible sources of exposure were not discussed). The serum levels of reproductive hormones were within the normal range, and only DEDTP showed a positive association with lutenizing hormone (LH) levels (β=11.4, 95% CI 0.81–22.1) and follicle-stimulating hormone (FSH) levels (β=3.2, 95% CI 0.08–6.2) and a negative association with the testosterone/LH ratio (β= –9.6, 95% CI –18.7− –0.50). However, the semen quality parameters were inversely correlated with the urinary OP metabolites. The sperm concentration and total sperm count were both significantly inversely associated with urinary concentrations of the DMP, DMDP, and DMDTP as well as ΣDAP. There was a significant inverse association between the percentage of motile sperm and DMTP, DMDTP, and DEP, and the total motile count was significantly inversely associated with urinary DMP and DMDTP concentrations (all significant at the ≤0.05 level). The seminal volume and percentage of sperm with normal morphology were not affected by DAP levels.

Although their study was not specific for any particular OP or carbamate pesticide (the latter are discussed in the next section), Miranda-Contreras et al. (2013) measured erythrocyte AChE and plasma butyrylcholinesterase (BuChE) levels in 64 male agricultural workers in Venezuela with exposure to these and other pesticides and also in 35 unexposed men. A total of 87.5% of the exposed workers showed exposure to OP and carbamate pesticides, and 82.8% had mild inhibition of BuChE activity. The levels of reproductive hormones and sperm quality were assessed in the workers. In the exposed group, there were significant adverse effects observed in semen pH, sperm motility, abnormal sperm, the number of live sperm, and the DNA fragmentation index versus those values in the unexposed group. Reproductive and thyroid hormones (testosterone, prolactin, free T4, and thyroid stimulating hormone) were normal in most of the exposed workers, although 44% had levels of LH and 21% had levels of FSH above reference values.

In a systematic review that assessed exposure to OPs and semen parameters, Martenies and Perry (2013) found that all six studies published between 2007 and 2012 reported significant adverse associations between OP exposure and at least one semen parameter (Hossain et al., 2010; Perez-Herrera et al., 2008; Perry et al., 2007, 2011; Recio-Vega et al., 2008; Yucra et al., 2008). The Volume 11 committee notes that exposures in most of the studies cited in this review were generally to more than one class of pesticide, e.g., OPs and carbamates, and to more than one pesticide within a class.

In a small biomonitoring study in China, Perry et al. (2007) examined sperm concentration in 18 men who did not work directly with pesticides but who had visited an agricultural area. Urine samples were analyzed for 13 parent OPs or their metabolites. Only the crude difference in sperm concentration for low versus high concentrations of DETP in urine was statistically significant (absolute sperm concentration difference= –1.0, 95% CI –1.8− –0.2). In a larger study of 94 exposed men and 95 controls in eastern China, Perry et al. (2011) confirmed that men with lower sperm concentration and motility were more likely to have higher urinary levels of DMP (odds ratio [OR]=1.30, 95% CI 1.02–1.65) than men with higher sperm concentrations; there was no association with five other urinary OP metabolites.

Recio-Vega et al. (2008) found similar effects on sperm quality in a study of 52 male agricultural workers in Mexico who were exposed to a number of pesticides, including diazinon. OP metabolites were found in the urine of 83% of the occupationally exposed participants. Of the seven semen parameters evaluated, only semen volume and total sperm count were significantly lower (p=0.004 and p=0.01, respectively) in the most exposed subjects compared with unexposed controls, and in the highly exposed group there was a negative correlation between semen volume and sperm count; there were no other significant associations. There were seasonal variations in the application of pesticides, with a corresponding variation in the levels of urinary metabolites and semen quality. Yucra et al. (2008) also found that exposure to OPs as reflected in the levels of urinary metabolites, particularly DETP and DEDTP, was associated with semen quality among 31 pesticide workers in Peru who were compared with 31 unexposed workers. Men with the highest levels of these two metabolites had lower volumes of semen and higher seminal pH than men with lower levels of DEDTP and DETP; there were no other significant associations. The reported use of the pesticides paraquat and/or malathion was correlated with semen quality in a group of 62 farmers in Malaysia who were compared with nonexposed controls (Hossain et al., 2010). The risk of having impaired semen parameters (lower semen volume, lower sperm concentration, a smaller percentage of motile sperm, and more abnormal sperm) was significantly increased (ORs 4.5–8.7) in exposed workers; seminal pH was not significantly changed. The committee notes that in the Hossain et al. study, exposure to the pesticides was not confirmed nor were the various pesticides distinguished by metabolite analysis.

Finally, Pérez-Herrera et al. (2008) evaluated the role of the PON1_Q192R² polymorphism on the susceptibility to OP toxicity in 54 male agricultural workers in Mexico. All participants had abnormal sperm morphology, 46% had low ejaculate volume, 30% had low sperm motility, and 13% had low sperm viability. OP exposure, either at the month of sample collection or during the preceding 3 months before sampling, was not associated with any parameters of semen quality or with DNA integrity. However, for men with the 192RR but not QQ genotype, there was a significant negative association between OP exposure during the 3 months before sampling (representing one spermatogenic cycle) and decreased sperm viability (p=0.086).

The Volume 11 committee identified two studies on female reproductive effects from exposure to OPs (Al-Hussaini et al., 2018; Hu et al., 2018). In the first study, Al-Hussaini and colleagues reported on 94 couples at an infertility clinic in Egypt. Follicular fluid from each woman was collected during in vitro fertilization/embryo transfer and analyzed for the presence of chlorpyrifos (mean=146.9±43 [s.d.] μg/L), diazinon (mean=284.7±108.2 μg/L), and malathion (mean=38.5±9.5 μg/L), among other pesticides and polychlorinated biphenyls. There were negative correlations between the follicular fluid concentrations of all three pesticides and endometrial thickness (p<0.05 for diazinon and malathion and p<0.01 for chlorpyrifos). The concentrations of diazinon and chlorpyrifos, but not of malathion, were negatively associated with the number of oocytes retrieved (p<0.01 for both pesticides) and with the implantation rate (that is, the number of sacs; p<0.01 for both pesticides). None of the OPs were associated with reduced fertilization or early embryo cleavage rates.

Hu et al. (2018) conducted a TTP study with 569 women recruited from two preconception care clinics in Shanghai, China, from August 2013 and April 2015. The subjects were enrolled in the study before conceiving a child and were followed for approximately 1 year. Exposure to organophosphates (and pyrethroids) was determined through the measurement of urinary metabolites. After adjustments were made for age, body mass index (BMI) before pregnancy, smoking, and education, women in the highest quartile for urinary levels of DETP were found to have significantly longer TTP (fecundability OR=0.68, 95% CI 0.51–0.92) and increased infertility (OR=2.17, 95% CI 1.19–3.93) than women in the lowest quartile.

Adverse Pregnancy Outcomes

Numerous researchers have examined the effects of exposure to OPs on a variety of pregnancy outcomes, such as infant body weight and length and head circumference, preterm birth, and duration of gestation. Some studies have included large cohorts of mother–child pairs and have followed the children for several years after birth. The postnatal developmental effects observed in these children are discussed in the section on developmental effects. In this section the committee focuses on fetal growth and birth outcomes, including spontaneous abortion, as important reproductive effects.

The committee notes that none of the studies considered in this section provide information on maternal or paternal exposures during the preconception period or the first trimester of pregnancy. Rather, the studies usually assess maternal and fetal exposure by measuring OP metabolites in maternal or cord blood or maternal urine samples during pregnancy. In addition, for the Columbia Center for Children's Environmental Health (CCCEH) study the mothers wore personal air monitors during pregnancy to measure ambient exposures, and the researchers measured the parent compounds chlorpyrifos and diazinon in maternal blood within one day of delivery.

There are four major longitudinal U.S. birth cohort studies of pesticide exposure and child health that have examined the effects of prenatal exposure to OPs and other pesticides on fetal growth: the CCCEH study in New York City; the Center for the Health Assessment of Mothers and Children of Salinas (CHAMACOS) study in California; the Health Outcomes and Measures of the Environment (HOME) study in Cincinnati, Ohio; and the Children's Environmental Cohort Study at Mount Sinai Hospital in New York City. Each of these studies has combined an assessment of maternal pesticide exposure with some measure of birth outcomes. Harley et al. (2016) evaluated birth outcomes across all four U.S. cohorts.

The CCCEH study followed a cohort of 263 inner-city African American or Dominican women enrolled between 1998 and 2002 (Perera et al., 2003). Women wore personal air monitors for 48 hours in the second or third trimester to determine their exposure to airborne toxicants. Blood and urine were also sampled. Chlorpyrifos was detected in 98% of maternal blood samples and 94% of cord blood samples at birth. Prenatal chlorpyrifos exposure (pg/g blood) was associated with a significant decrease in birth weight (grams) overall (β= –0.04, p=0.01) and among African Americans (β= –0.05, p=0.04), but not among Dominican infants (β= –0.03, p=0.11). It was also associated with reduced birth length overall (β = –0.01, p=0.003) and in Dominicans (β= –0.02, p<0.001); chlorpyrifos exposure was not significantly associated with reduced head circumference in any group.

The CHAMACOS study investigated a group of 488 low-income Latina women and their infants in an agricultural area of Salinas Valley, California, in 1999−2000 (Eskenazi et al., 2004). DAPs were measured in maternal urine at about 13 weeks of gestation and again at 26 weeks of gestation, and cholinesterase (ChE) and BuChE were measured in maternal and umbilical cord blood at delivery. Increased DAP concentrations were associated with increased infant body length (p=0.06) and were significantly associated with an increased head circumference (p=0.03). A 10-fold increase in DMP, but not DEP, metabolites was associated with a 3-day decrease in gestational duration (β= –0.41 weeks/log10 unit increase; 95% CI –0.75– –0.02; p=0.02), but only after 22 weeks of gestation. DAP metabolites were not associated with birth weight, infant ponderal index (a weight–height index used to measure fetal growth), the risk of preterm delivery, low birth weight, or small-for-gestational-age births. Lower levels of ChE in umbilical cord blood were associated with a significantly shorter length of gestation (β=0.34 weeks/unit increase; 95% CI, 0.13–0.55; p=0.001), increased risk of preterm delivery (OR=2.3, 95% CI 1.1–4.8; p=0.02), and lower birth weight (OR=4.3, 95% CI 1.1–17.5; p=0.04). However, 6 of the 11 low-birth-weight infants were also preterm. BuChE levels in maternal and umbilical cord blood were not associated with any birth outcome.

A further analysis of the CHAMACOS cohort by Harley et al. (2011) examined whether PON1 would modify the effect of OPs on adverse fetal outcomes. Maternal and cord blood samples were analyzed for the PON1 genotype and enzyme activity. Associations of PON1 with decreased gestational age were found only in those infants who had susceptible PON1 genotypes (PON1_-108TT); the presence of susceptible PON1 genotypes in mothers was not associated with decreased gestational age. No associations were found between any marker of PON1 genotype or activity as measured in maternal blood and the gestational age or infant birth weight, length, or head circumference. There was some evidence of effect modification with DAPs: maternal DAP concentrations were associated with shorter gestational age only among infants with the susceptible PON1_-108TT genotype (p-value interaction=0.09). However, maternal DAP concentrations were associated with larger birth weight (p-value interaction=0.06) and head circumference (p-value interaction 0.01) in infants with a nonsusceptible genotype (PON1_192QQ).

In the Cincinnati HOME study, 344 women less than 19 weeks pregnant were assessed for residential exposure to toxicants between 2003 and 2006 (Rauch et al., 2012). Urine samples collected at 16 and 26 weeks of gestation were analyzed for six DAPs; all mothers had detectable levels of at least one DAP. A 10-fold increase in total DAP concentration was significantly associated with a decrease in gestational age (β= –0.5 weeks, 95% CI –0.8– –0.1) and with a lower birth weight (β= –151 g, 95% CI –287– –16), and the association with decreases in gestation age was stronger for white mothers than for black mothers, but the reverse was true for birth weight. Decrements in birth weight and gestational age associated with ΣDAP concentrations were greatest among infants with the PON1_192QR and PON_–108CT genotypes.

Berkowitz et al. (2004) reported on the Children's Environmental Cohort Study at Mount Sinai Hospital in New York City, a cohort of 404 multiethnic mother–infant pairs with babies born between 1998 and 2002. TCPy (a metabolite of chlorpyrifos) and 3-phenoxybenzoic acid (PBA; a pyrethroid metabolite) were each detected in more than 40% of prenatal maternal urines. There were no associations between maternal pesticide exposures or urinary levels of TCPy, PBA, or pentachlorophenol and adjusted birth weight, adjusted length, adjusted head circumference, or adjusted gestational age. A positive trend was found between maternal blood PON1 activity and head circumference, but the trend was only significant for mothers whose TCPy levels were above the level of detection. No trends were seen for birth weight or birth length for TCPy or for the other metabolite levels when maternal PON1 genotype level was taken into account. Infant PON1 activity had no association with any of the fetal growth measures.

Harley et al. (2016) pooled data from the four U.S. cohort studies and found that there were no significant associations between DAP metabolites and birth weight, length, or head circumference. However, when the mothers were stratified by ethnicity, there was an association between increasing urinary DAP concentrations and decreased birth weight for non-Hispanic black women. Infants with the PON1_192RR genotype were at greater risk for decreased birth length with increasing prenatal exposure to DAP metabolites. Gestational duration was not examined in the pooled analysis.

The committee identified one study in Shanghai, China, that also sought to determine whether prenatal exposure to pesticides resulted in any adverse fetal growth outcomes (Wang et al., 2012). Women (n=187) were recruited into the study between 2006 and 2007, and they provided urine samples at the onset of labor for DAP analysis. There were no significant associations between prenatal OP exposure and infant body weight or length or length of gestation, with the exception of DEP, which showed a significant inverse relationship with duration of gestation for infant girls but not boys (β= –1.79; p=0.001 versus β=0.17; p=0.164, respectively).

In an in vitro study with human villous cytotrophoblast cells isolated from human placentas, exposure to chlorpyrifos was found to cause marked changes in placental tissues including chorionic villi, suggesting the placenta may be a target of chlorpyrifos toxicity (Ridano et al., 2017).

Animal Studies

Although the Volume 2 committee (IOM, 2003) did not report on human or animal studies of reproductive effects and OP exposure, information from animal studies was available at the time. In 1997, ATSDR had published toxicological profiles of dichlorvos and of chlorpyrifos that included the results of animal studies on the reproductive effects of those pesticides. Dermal exposure to chlorpyrifos had resulted in transient adverse effects on semen volume and quality in bulls. Oral exposure to chlorpyrifos had no effect on female reproduction in mice, rats or dogs; however, the doses of chlorpyrifos needed to cause death in pregnant mice were approximately six times lower than those needed to cause death in nonpregnant mice. Chlorpyrifos administered via inhalation had no effect on rat testes (ATSDR, 1997a). ATSDR concluded that in several animal species oral and inhalation exposures to dichlorvos did not have reproductive effects in either males or females (ATSDR, 1997b). The ATSDR Toxicological Profile for Malathion (2003a) indicated that malathion was not a reproductive toxicant in animals; however, several animal studies had found that the oral administration of malathion did produce transient testicular effects. Finally, ATSDR concluded that animal studies had provided no evidence that diazinon was a reproductive toxicant. The studies on which that conclusion was based had involved oral exposures in mice, rats, and dogs (ATSDR, 2008).

In the past 10 years, rodent research has focused on characterizing the testicular toxicity of OPs. As discussed below, these studies have treated rodents with OPs, usually over at least one sperm cycle, and found effects on sperm quality, serum hormones, testes and epididymis histopathology, biochemistry, and gene expression.

DNA damage in sperm, which has been associated with OP exposure in humans (Meeker et al., 2004b), has also been observed in animal studies after in vivo treatment with diazinon (Pina Guzman et al., 2005; Sarabia et al., 2009) and malathion (Giri et al., 2002) as well as in vitro with human sperm after exposure to chlorpyrifos (Salazar-Arredondo et al., 2008). These findings in sperm agree with the in vitro genotoxicity that has been observed relating to chromosomal disruption in lymphocytes (Ohja and Gupta, 2015).

As in human studies such as Meeker et al. (2008a), rodent experiments in male animals have found decreased levels of serum testosterone as well as of the gonadotropins FSH and LH after short-term exposure to chlorpyrifos (Alaa-Eldin et al., 2017; Mandal and Das, 2012; Sai et al., 2014), dichlorvos (Dirican and Kalender, 2012), diazinon (Leong et al., 2013), and malathion (Bustos-Obregón and González-Hormazabal, 2003; Geng et al., 2015; Uzun et al., 2009). These hormone effects were seen in connection with sperm abnormalities and testicular histopathology. Hormone effects were observed consistently, as was testicular toxicity, when measured, with the exception of the two dichlorvos studies (Okamura et al., 2005, 2009).

Two mechanisms for OP reproductive toxicity have been studied. The suggestion that OPs generate reactive oxygen species (ROS) is supported by the presence of tissue oxidative damage markers in male reproductive organs after treatment with chlorpyrifos (Mandal and Das, 2011, 2012; Mosbah et al., 2016; Sai et al., 2014; Umosen et al., 2012). ROS-mediated genotoxicity has been established for in vitro chlorpyrifos exposures (Chauhan et al., 2016). Other studies have supported this hypothesis by showing that antioxidants provide protection against testicular toxicity (Alaa-Eldin et al., 2017; Dirican and Kalender, 2012; Dutta and Sahu, 2013; Elmazoudy et al., 2011; Elsharkawy et al., 2014; Mosbah et al., 2016; Umosen et al., 2012; Uzun et al., 2009). A second line of mechanistic work demonstrated a failure in the final steps of sperm maturation in the epidydimis after treatment with malathion and dichlorvos (Akbarsha et al., 2000). ROS are known to be involved in sperm capacitation (Aitken, 2017).

OP animal studies that included a mating trial after the males had been treated found that preconception exposure to chlorpyrifos reduced fertility (85% negative fertility using mating exposure test; Joshi et al., 2007) and decreased the number of live fetuses, increased the number of dead fetuses, and increased the number of early resorptions (Farag et al., 2010). The treatment of male rats with malathion prior to mating caused an 80% decrease in fertility (Choudhary et al., 2008).

As with humans, there is little information on female animal reproductive toxicity resulting from exposures prior to pregnancy. Chebab et al. (2017) found that 6.75 mg/kg chlorpyrifos administered to female rats for 30 days significantly decreased the serum concentrations of sex hormones and thyroid hormones.

A study with dichlorvos did not find any effects on pregnancy when female mice were treated before mating (Dean and Blair, 1976). Gomes and Lloyd (2009) and Gomes et al. (2008) found effects on both pregnancy outcome and the incidence of malformation when a mixture of OPs was given to female mice before mating. The mixture included chlorpyrifos and dichlorvos (both OPs used in the Gulf War) as well as other pesticides at LD₅₀ levels. Such high doses coupled with the use of commercial formulations and the large number of agents make these findings difficult to interpret. Malathion has also been shown to affect the maturation of porcine cumulus–oocyte complexes in vitro by inducing oxidative stress (Flores et al., 2017).

The Volume 11 committee found a lack of relevant information in the animal literature on adverse birth outcomes associated with preconception or prenatal exposure to any of the relevant OPs considered in this chapter.

Neurodevelopmental Effects

The Gulf War and Health, Volume 2 committee did not discuss any studies on developmental effects that may result in children whose parents were exposed to OPs. On the other hand, that committee did consider a number of animal studies, which are discussed later in this section.

In November 2016, EPA released Chlorpyrifos: Revised Human Health Risk Assessment for Registration Review as part of the Federal Insecticide, Fungicide, and Rodenticide Act Section 3(g) Registration Review program. The review gave special consideration to chlorpyrifos' neurodevelopmental effects on children. The risk assessment document (EPA, 2016) stated:

In the 2015 updated literature review (USEPA, D331251, 09/15/2015), the agency conducted a systematic review expanding the 2012/2014 review which was focused only on US cohort studies with particular emphasis on chlorpyrifos. The expanded 2015 review includes consideration of the epidemiological data on any OP pesticide, study designs beyond prospective cohort studies, and non-U.S. based studies. The updated literature review identified seven studies which were relevant (Bouchard et al. 2010; Fortenberry et al. 2014; Furlong et al. 2014; Guodong et al. 2012; Oulhote and Bouchard, 2013; Zhang et al. 2014; Shelton et al. 2014). These seven studies have been evaluated in context with studies from the 2012/2014 review (D. Drew et al. D424485, 12/29/2014). In addition, the agency has also reviewed more recent studies from CCCEH (Rauh et al. 2015) and a pooled analysis of U.S. cohort studies (Engel et al. 2015) (E. Holman, D432184, 03/25/2016). As discussed below, Rauh et al. (2015) provides further evidence of neurodevelopmental outcomes in the CCCEH study. The Engel et al. (2015) study shows relatively consistent results compared to previous studies conducted at 24 months (Engel et al. 2011; Rauh et al. 2006). . . . The agency continues to conclude that the 3 U.S. cohort studies (CCCEH, CHAMACOS, and Mt. Sinai) provide the most robust available epidemiological evidence.

The agency acknowledges the lack of established MOA/AOP pathway [mode of action/adverse outcome pathway], the inability to make strong causal linkages, and the unknown window(s) of susceptibility. These uncertainties do not undermine or reduce the confidence in the findings of the epidemiology studies. The epidemiology studies reviewed in the 2012/2014 and 2015 literature reviews represent different investigators, locations, points in time, exposure assessment procedures, and outcome measurements. Despite differences in study design, with the exception of two negative studies in the 2015 literature review (Guodong et al. 2012; Oulhote and Bouchard, 2013) and the results from the more recent Engel et al. (2015) study, all other study authors have identified associations with neurodevelopmental outcomes associated with OP exposure; these conclusions were across four cohorts and twelve study citations. Specifically, there is evidence of delays in mental development in infants (24-36 months), attention problems and autism spectrum disorder in early childhood, and intelligence decrements in school age children who were exposed to OPs during gestation. Investigators reported strong measures of statistical association across several of these evaluations (odds ratios 2-4 fold increased in some instances), and observed evidence of exposures-response trends in some instances, e.g., intelligence measures. . . .

In summary, the EPA's assessment is that the CCCEH study, with supporting results from the other 2 U.S. cohort studies and the seven additional epidemiological studies reviewed in 2015, provides sufficient evidence that there are neurodevelopmental effects occurring at chlorpyrifos exposure levels below that required for AChE inhibition.

On December 15, 2017, the California Environmental Protection Agency, Office of Environmental Health Hazard Assessment, Developmental and Reproductive Toxicant Identification Committee determined that chlorpyrifos was “clearly shown through scientifically valid testing according to generally accepted principles to cause reproductive toxicity,” based on the developmental endpoint (OEHHA, 2017).

The Volume 11 committee considered a number of studies on the developmental effects, particularly neurological outcomes, in children whose mothers or fathers had been exposed to organophosphate pesticides, primarily chlorpyrifos. In addition to studies conducted in Spain, Mexico, and China, researchers have also used four U.S. cohort studies to look at the effects of prenatal OP exposure on a variety of neurodevelopmental outcomes in children. Only studies not cited in Chlorpyrifos: Revised Human Health Risk Assessment for Registration Review (EPA, 2016) and considered by the committee to be relevant to its assessment are included in Table 5-3.

The CCCEH study assessed children born between February 1998 and May 2002 for psychomotor and cognitive development at 12, 24, and 36 months (Rauh et al., 2006), and at 7 years of age (Rauh et al., 2012). Younger children were administered the Bayley Scales of Infant Development II (BSID-II), and their mothers completed the Child Behavior Checklist (CBCL); the 7-year-olds were administered the Wechsler Intelligence Scale for Children, 4th edition (WISC-IV). Prenatal exposure of chlorpyrifos levels was measured in maternal blood or cord blood samples obtained at delivery, and the results were categorized as high exposure (>6.17 pg chlorpyrifos/g plasma) or low exposure (≤6.17 pg/g). At 12 and 24 months of age, there were no differences in mental or motor function between high- and low-exposure children. However, at 36 months of age, children with high prenatal exposure to chlorpyrifos scored significantly lower on the psychomotor development index (PDI; p=0.01) than children with low exposure; were in the clinical problems range for all five CBCL scales, particularly the attention problems and attention deficit–hyperactivity disorder (ADHD) scales (p=0.03 each); and were more likely than the low-exposure children to have mental delays (OR=2.4, 95% CI 1.12–5.08) and motor delays (OR=4.5, 95% CI 1.61–12.70) (Rauh et al., 2006). At 7 years of age, there were significant inverse associations between chlorpyrifos exposure and two of the WISC-IV scales: working memory (r= –0.21, p <0.0001) and full-scale intelligence quotient (r= –0.13, p=0.02) (Rauh et al., 2011). Horton et al. (2012) found a significant interaction between exposure and the child's sex, suggesting that boys had a greater decrement in working memory than girls (β= –1.714, 95% CI –3.753–0.326). At 11 years of age, there was a significant association of prenatal chlorpyrifos exposure with tremor (Rauh et al., 2015). A subgroup of 40 CCCEH 7-year-old children (20 from the upper tertile and 20 from the lowest tertile) also underwent magnetic resonance imaging of the brain. Children in the high-exposure group had a significant enlargement of several brain areas (e.g., superior temporal, posterior middle temporal, and inferior postcentral gyri bilaterally; gyrus rectus) as well as frontal and parietal cortical thinning and an inverse dose–response relationship between chlorpyrifos and cortical thickness (Rauh et al., 2012).

A second New York City study, the Mount Sinai Children's Environmental Health Cohort, recruited 404 mother–infant pairs during pregnancy in 1998–2001 (Engel et al., 2007, 2011). Prenatal exposure to OPs (including malathion) was determined by an analysis of maternal blood and urine samples taken at about pregnancy week 31. The Brazelton Neonatal Behavioral Assessment Scale (BNBAS) was administered to 311 neonates before they left the hospital (Engel et al., 2007). BNBAS scores showed that neonates born to women with elevated urinary malathion dicarboxylic acid (MDA) levels had a greater number of abnormal reflexes (OR=2.24, 95% CI 1.55–3.24) than children born with elevated total DEP metabolites (OR=1.49, 95% CI 1.12–1.98). Increasing levels of total DAP and DMP metabolites were also associated with an increase in abnormal reflexes. In a further study of this cohort (n=169 children or more, depending on which assessment), neurodevelopment was assessed within 5 days of birth, at 12 and 24 months, and at 6–9 years of age (Engel et al., 2011). At 12 and 24 months, in black and Hispanic infants decreased mental development (as measured with the BSID-II) was associated with higher exposure to prenatal total DAP metabolites, whereas among whites, increasing exposure was associated with improved mental development index (MDI) scores. Furthermore, prenatal maternal DAP levels were inversely associated with children's 24-month MDI but not their PDI. At 7–9 years of age, there were only modest effects of any prenatal OP metabolites on any psychometric measure in the children (Engel et al., 2011). A further evaluation of the 7- to 9-year-old children was done using the Social Responsiveness Scale (SRS) to assess for characteristics of autism spectrum and related disorders. No associations overall were found between any of the total DAP, DEP, or DMP metabolites and total SRS scores. However, increasing total DEP levels were associated with poorer SRS scores among blacks (β=5.1 points, 95% CI 0.8–9.4) and boys (β=3.5 points, 95% CI 0.2–6.8), although not among whites and Hispanics (β=0.2 points, 95% CI –2.8–3.2), or among girls (β= –0.4 points, 95% CI –4.1–3.3) (Furlong et al., 2014). Further analysis of the children found that DMP metabolites were negatively associated with internalizing factor scores (β= –0.13, 95% CI –0.26–0.00), but positively associated with executive functioning factor scores (β=0.18, 95% CI 0.04–0.31). However, DEP metabolites were negatively associated with the working memory index (β= –0.17, 95% CI –0.33– –0.03) (Furlong et al., 2017a).

The longitudinal CHAMACOS cohort study has tracked children born between October 1999 and October 2000 to mothers residing in an agricultural area of the Salinas Valley in California. The mothers were generally young, of Mexican origin, and from farm-working families, and most had not completed high school. Urine samples were collected from the mothers (n=381) during the first and the second half of pregnancy, and each sample was analyzed for DAP metabolites of OP pesticides. Children (n=327) were assessed for developmental parameters at 6, 12, and 24 months, 3.5 years, 5 years, and 7 years of age (Eskenazi et al., 2014). Newborn infants showed a significant increase in the number of abnormal reflexes on the BNBAS which was correlated with increasing levels of maternal total DAP, DMP, and DEP levels measured in urine (β=0.23, 95% CI 0.05–0.41; β=0.18, 95% CI 0.02–0.34; β=0.22, 95% CI 0.04–0.40, respectively); however, none of the maternal metabolite levels were correlated with any of the other BNBAS scores (Young et al., 2005). Eskenazi et al. (2007) found that maternal prenatal DAP levels were negatively associated with the child's MDI on the BSID-II and that this association reached statistical significance at 24 months of age (per 10-fold increase in prenatal DAPs: β= –3.5 points; 95% CI –6.6– –0.5). Prenatal DAPs were also associated with pervasive developmental disorder (per 10-fold increase in prenatal DAPs: OR=2.3, p=0.05). Prenatal DAPs were not significantly associated with attention problems or ADHD when the children were 3.5 years of age, but they were associated with both outcomes by the time the children were 5 years of age (CBCL attention problems: β=0.7 points; 95% CI 0.2–1.2; ADHD: β=1.3, 95% CI 0.4–2.1), and some associations were found only among boys (Marks et al., 2010). Finally, at 7 years of age, a 10-fold increase in maternal DAP concentrations was associated with lower cognitive scores on all four cognitive subtests of the WISC-IV. The most significant associations were for verbal comprehension (β= –5.3, 95% CI –8.6– –2.0) and a decrease of 5.6 full-scale IQ points (95% CI –9.0– –2.2). Higher DAP levels later in pregnancy were associated with greater deficits in the children than were levels in the first half of pregnancy (Bouchard et al., 2011). The association between DAPs and full-scale IQ was strongest for the children of mothers with the lowest-tertile levels of arylesterase, a measure of the enzyme activity of PON1. This relationship held for both diethyl and dimethyl DAPs and for all subscales of the WISC (Eskenazi et al., 2014). Further assessment of neurobehavior in 238 of the 7-year-old children in the CHAMACOS cohort found consistent but not significant negative associations between DNA methylation at several CpG sites close to the PON1 transcription start site in the children's cord blood and full-scale IQ and other composite measures of cognition as measured with the WISC-IV. Nonsignificant associations were also found between cord blood methylation at cg15887283 and the children's working memory scores (Huen et al., 2018). Studies using Pesticide Use Reporting data from California reported an association between the use of OPs in fields near the maternal residence during pregnancy and WISC Full Scale and Verbal Comprehension IQ scores in children at age 7 (Gunier et al., 2017).

A further study by Sagiv et al. (2018) of the CHAMACOS cohort examined whether there is an association between prenatal exposure to OP pesticides and autism spectrum disorder (ASD) traits in children and adolescents. Women were interviewed twice during pregnancy, once after delivery and then when the children reached the ages of 6 months, 1, 2, 3.5, 5, 7, 9, 10.5, 12, and 14 years. Exposure was assessed by measuring DAPs in urine and by collecting information on the mother's residential proximity to OP application while pregnant, using California's Pesticide Use Reporting data. The authors measured traits in 247 children based on the parents' and teachers' reporting on the child's performance at ages 7, 10.5, 12, and 14 years on tests specifically aimed at evaluating a child's ability to use facial expressions and to recognize the mental state of others. Results showed that prenatal exposure to DAPs was associated with unfavorable social behavior reported by parents and teachers. DAP levels were also associated with a 2.7-fold increase in social responsiveness among children at 14 years of age (β=2.7, 95% CI 0.9–4.5). The SRS subscales are social awareness, social cognition, social communication, social motivation, and restricted interests and repetitive behavior. Normal scores are below 59, and scores above 75 are interpreted as indicating severe deficiencies in reciprocal social behaviors. No association was found between proximity to an area where OPs were used during pregnancy and ASD traits in children. In short, this study presented inconsistent results on the association between OP pesticides and ASD-related traits.

The committee considered two studies that examined pregnant women recruited into the HOME study (Donauer et al., 2016; Millenson et al., 2017). Urine from 327 women was collected at about gestation week 16 and again around gestation week 26 and analyzed for metabolites of OPs. Children were administered the BSID-II at 1, 2, and 3 years of age; the Clinical Evaluation of Language Fundamentals at age 4; and the Weschsler Preschool and Primary Scale of Intelligence at age 5 (Donauer et al., 2016). No association was found between prenatal exposure and cognition at any age from 1 through 5 after adjustment for relevant covariates. Millenson et al. (2017) further examined the children for PON1 polymorphisms from cord blood. Children were assessed with the SRS. Total DAP levels were not associated with changes in SRS scores (β= –1.2, 95% CI –4.0–1.6), nor were the associations modified by the child's PON1 genotype.

Engel et al. (2016) attempted to pool data from the CHAMACOS study, the HOME study in Cincinnati, and the CCCEH and Mount Sinai cohort studies in New York City. Those researchers found significant heterogeneity for estimates of the association between DAPs and MDI (p=0.09) and also found that PON1 genotypes, ethnicity, and the center performing the study all influenced the results. Despite the limitations, the authors concluded that some population subgroups were at risk of neurodevelopment problems as shown by changes in the MDI.

The Childhood Autism Risk from Genetics and Environment (CHARGE) study in California also attempted to link pesticide use near the mother's residence to children having ASD (n=486) or developmental disorder (n=168) (Shelton et al., 2014). Each pregnancy was given an exposure profile based on the pesticide applications made near the mother's home and the days of her pregnancy on which those applications occurred. Children were eligible for the study if they were 2–5 years of age (mean 36–38 months). Any prenatal proximity (within 1.25 km) to OPs was associated with an increased risk for ASD (OR=1.6, 95% CI 1.02–2.51); the risks were higher the later in the pregnancy the exposure occurred, but except for third-trimester exposures, those risks were not significant (OR=1.37, 95% CI 0.76–2.50 for preconception exposure; OR=1.53, 95% CI 0.87–2.68 for first trimester; OR=1.57, 95% CI 0.87–2.83 for second trimester; and OR=1.99, 95% CI 1.11–3.56 for third trimester). For developmental disorder and residential proximity to agricultural pesticide application the ORs were 1.23 (95% CI 0.65–2.31) for all pregnancies, 1.20 (95% CI 0.54–2.65) for preconception exposure, 1.29 (95% CI 0.60–2.79) for the first trimester, 1.62 (95% CI 0.75–3.48) for the second trimester, and 1.10 (95% CI 0.46–2.67) for the third trimester. The authors concluded that prenatal exposure to OPs could be linked with neurodevelopmental disorders.

Several studies in China assessed the effect of prenatal exposure to OPs on child neurodevelopment. Zhang et al. (2014a) administered the Neonatal Behavioral Neurological Assessment (NBNA) to 3-day-old infants in Shenyang, China. Urine samples of 249 pregnant women were analyzed for five nonspecific OP metabolites. Higher prenatal DAP concentrations were associated with lower scores in all NBNA scales, especially the Behavior Scale (β for a 10-fold increase in concentration= –0.65, 95% CI –0.84– –0.45, p=0.01). In a further assessment of the effects of prenatal exposure to over 30 OPs on child neurodevelopment, Silver et al. (2017) analyzed cord blood from 237 Chinese infants born in Fuyang Maternal and Children's Hospital between 2008 and 2011. Exposure to chlorpyrifos, as determined by gas chromatography, was associated with significant motor function deficits in 199 infants assessed at 6 weeks and 9 months of age using the Peabody Developmental Motor Scales, 2nd edition. In another study, Wang et al. (2017) measured six DAP metabolites in urine specimens from 436 women taken at delivery. Women were part of the Laizhou Wan Birth Cohort study in Shandong, China. Children were assessed for neurodevelopment at 12 months (n=297) and 24 months (n=262). Prenatal OP concentrations were not correlated with any neurodevelopmental deficits, as measured with the Gesell Developmental Schedules, at 12 months. At 24 months, however, prenatal, but not postnatal, OP exposure was associated with a reduced developmental quotient, especially among boys.

González-Alzaga et al. (2015) used the WISC-IV to study the effects of high and low exposure to OP applications on cognitive functioning in 305 6- to 11-year-old children in Spain. Maternal prenatal residential exposure to OPs was associated with a poorer score on the processing speed subtest for boys (β= –5.7; 95% CI –9.6– –1.8), but not for girls (β= –1.6, 95% CI –6.4– –3.3). No other associations were significant for either high or low exposures or any of the WISC-IV subtests. The authors concluded that there was a weak association between prenatal OP exposure and neurological deficits.

The Early Life Exposures in Mexico to Environmental Toxicants (ELEMENT) study measured metabolites in maternal urine to assess prenatal exposure to chlorpyrifos and ADHD in children ages 6–11 years old in Mexico City (Fortenberry et al., 2014). There were no significant associations between chlorpyrifos metabolite concentrations in maternal urine and ADHD outcomes in the children.

A recent integrative review by Burke et al. (2017) examined the developmental neurotoxicity of OPs with an emphasis on chlorpyrifos and summarized the extensive epidemiologic studies of chlorpyrifos (most of which are discussed earlier in this section). The authors concluded that a “disruption of the structural integrity of the brain can be an important determinant of the cognitive deficits” associated with prenatal exposure to the pesticide. The authors also considered eight studies of cognitive and locomotor phenotypes following prenatal exposure to chlorpyrifos in rats and mice. The authors argued that “it is unclear to what extent, if any, AChE inhibition contributes to the developmental neurotoxicity of chlorpyrifos” and concluded that this made it difficult to extrapolate from animal models to humans.

Respiratory Effects

Respiratory effects in children following prenatal exposures were not reported in Gulf War and Health, Volume 2, nor were they discussed in EPA's 2016 Chlorpyrifos: Revised Human Health Risk Assessment for Registration Review, ATSDR's 2008 Toxicological Profile for Diazinon, or ATSDR's 1997 Toxicological Profile for Dichlorvos.

The Volume 11 committee identified two recent publications by Raanan et al. (2015, 2016) that assessed respiratory effects in young children after prenatal exposure to OPs and other pesticides in the CHAMACOS cohort. The assessments of maternal exposures to OPs were based on an analysis of DAPs in maternal urine at gestation weeks 13 and 26. Mothers reported their children's respiratory symptoms at 5 and 7 years of age, with 359 children assessed. Children's respiratory symptoms were not associated with any maternal DAP concentrations in the first half of pregnancy. However, there were significantly increased risks of respiratory symptoms in children whose mothers had elevated prenatal total DAPs and diethyl metabolite levels during the second half of pregnancy (ΣDAPs OR for a 10-fold increase in concentration=1.77, 95% CI 1.06–2.95; DEPs OR=1.61, 95% CI 1.08–2.39); there was no increased risk observed for elevated dimethyl metabolites (Raanan et al., 2015). A spirometric assessment of lung function in the 7-year-old children found that the mothers' DAP urinary concentrations, whether in the first half or second half of pregnancy, were not significantly associated with lung function measurements (FEV₁, FVC, FEV₁/FVC ratio, and FEF_25–75) in their children (Raanan et al., 2016).

Childhood Cancer

Genotoxicity studies of OPs reveal primarily chromosomal abnormalities and DNA fragmentation. IARC has classified two of the Gulf War OPs—diazinon and malathion—as probable human carcinogens (IARC, 2016a,b). IARC has also classified dichlorvos as a possible human carcinogen (IARC, 1991). No transplacental cancer studies were reviewed for any of the three OPs.

The committee found only one new study that assessed the risk of childhood cancers and exposure to OPs before birth. Monge et al. (2007) examined the effects of maternal exposures (n=876) and paternal exposures (n=762) to pesticides on the risk of childhood leukemia (n=334) in Costa Rica in 1995–2000. Parents were interviewed about risk factors for childhood leukemia, and their exposures were rated as to hazard value at the time of occurrence. The mothers' exposure to OPs in the first trimester of pregnancy was significantly associated with total leukemia (OR=3.5, 95% CI 1.0–12.2) and acute lymphoblastic leukemia (ALL) (OR=3.7, 95% CI 1.0–13.1). Although the risk of any leukemia and ALL was elevated in children whose mothers were exposed in the year before conception (leukemia OR=2.3, 95% CI 0.9–6.0; ALL OR=2.5, 95% CI 0.9–6.7) and during the second trimester (leukemia OR=2.5, 95% CI 0.7–9.5; ALL OR=3.0, 95% CI 0.8–11.5), the increases were not significant. In the fathers, exposure to OPs was associated with a increased risk of leukemia at all time points: the year before conception (OR=1.5, 95% CI 1.0–2.2), the first trimester (OR=1.6, 95% CI 1.0–2.6), and the second trimester (OR=1.4, 95% CI 0.8–2.3). Further analysis of the fathers' exposures to specific pesticides showed that exposure to malathion in the year before conception was associated with an increased risk of leukemia and of ALL for boys (OR=8.5, 95% CI 1.1–74.1 and OR=10.4, 95% CI 1.2–91.1, respectively), but not for girls (OR=0.9, 95% CI 0.2–4.9 and OR=0.5, 95% CI 0.1–4.8, respectively).

Other Developmental Effects

Carmichael et al. (2014) examined the association between exposure to pesticides during pregnancy and the prevalence of eight types of congenital heart defects in a population-based case-control study. Pesticide exposure within a 500-meter radius of each mother's residence was estimated for the 3 months prior to conception to delivery using geospatial coding for eight San Joaquin counties. The California Pesticide Use Reporting records were used to identify any pesticides applied at more than 100 lbs per year in any of the eight counties. Using data from the California Birth Defects Monitoring Program from 1997 to 2006, the exposures of the mothers of 569 case infants was compared with 785 mothers who had an infant without a birth defect. Prenatal exposure to chlorpyrifos was significantly associated with pulmonary valve stenosis (OR=2.4, 95% CI 1.0–6.3) and atrial septal defects, secundum (OR=1.9, 95% CI 1.1–3.4). A further assessment for five types of other birth defects was conducted (Carmichael et al., 2016), based on the exposure of 367 mothers of case infants compared with the same 785 control mothers. Prenatal exposure to OPs as a group (based on mechanism of action) was associated with a reduced risk of anorectal atresia/stenosis (OR=0.4, 95% CI 0.2–0.9) but was significantly associated with an increased risk of craniosynostosis (OR=1.9, 95% CI 1.1–3.2). Chlorpyrifos specifically was significantly associated with anotia (OR=2.0, 95% CI 1.1–3.8) and craniosynostosis (OR=2.4, 95% CI 1.2–4.6). The ORs for all other OPs or birth defects were not significant and ranged from 0.5 to 2.0.

Buscail et al. (2015) used the French PELAGIE mother–child cohort to examine the prevalence of otitis media in children at 2 years of age. They found that the children of 246 mothers who had exposure to OPs, determined by an analysis of DAP, DMP, and DEP metabolites in the mother's urine taken before the 19th week of pregnancy, were not at an increased risk of having one or more occurrences of otitis media before age 2 years.

The Volume 11 committee identified one small study on epigenetic effects associated with prenatal exposure to pesticides. Declerck et al. (2017) examined 25 children born in Denmark between 1996 and 2001 to mothers who had occupational exposures to a variety of pesticides during the first trimester of pregnancy and 23 unexposed control children. Blood samples taken from the children at 6–11 years of age were analyzed for the PON1 Q192R genotype and genome-wide DNA methylation patterns related to prenatal pesticide exposure was determined. The majority of mothers (91%) were exposed to organophosphate pesticides, including chlorpyrifos, dichlorvos, and dimethoate, as well as pyrethroid pesticides (e.g., deltamethrin) and carbamate pesticides (e.g., methomyl). Prenatal pesticide exposure was found to be associated with changes in DNA methylation patterns in children with the PON1 192R-allele compared with the 192Q-genotype and the control children, and the hypermethylation changes were significantly associated with higher body fat content (p<0.001) and serum leptin concentrations (p=0.02).

Animal Studies

In studies of pregnant women exposed to toxicants, including OPs, there are three confounding issues: exposures are seldom limited to one insecticide (i.e., exposures are typically to mixtures of toxicants); exposures are rarely limited to one period of time (i.e., preconception, pregnancy or trimester of pregnancy, or postnatal); and outcome assessments are typically made in early childhood and not at later stages of development, such as adolescence. These issues can be addressed in animal models. Socioeconomic status can also be a confounder in human studies, but it is not in animal studies.

The ATSDR Toxicological Profile for Malathion (2003a) indicated that malathion was not a developmental toxicant in animals at doses that did not induce maternal toxicity.

Animal models have been used extensively to study sensitive periods for chlorpyrifos exposure in pregnancy. This is important for female veterans whose deployments are terminated shortly after pregnancy is identified. Researchers using rodent models to compare the effects of chlorpyrifos exposure in early and late pregnancy found that early pregnancy exposure had various long-lasting effects on brain development in rats, including altered behavior (Icenogle et al., 2004), brain morphology (Chen et al., 2017; Qiao et al., 2004), modified brain chemistry (Aldridge et al., 2003, 2004, 2005; Meyer et al., 2004a,b; Qiao et al., 2002), and altered brain development (Qiao et al., 2003). For example, adult rats that had been exposed to chlorpyrifos on gestational days (GDs) 9–12 or 17–20 showed significant changes in the adenylyl cyclase signaling cascade, which mediates the cellular responses to numerous neurotransmitters, in a wide variety of brain regions in adulthood (Meyer et al., 2004a). However, exposure during GD 9–12 was less sensitive and required doses that resulted in decreased maternal weight gain. In another study, early pregnancy exposure resulted in cardiac and hepatic effects in offspring (Meyer et al., 2004b).

Exposure to chlorpyrifos later in pregnancy (GDs 14–20) also produced anxiety-like symptoms in rats assessed on postnatal day (PND) 21, but the anxious behavior was reversed by PND 70 (Silva et al., 2017). There were no depressive symptoms.

Studies of mid- to late-pregnancy exposure (GDs 14–17) of CD-1 mice to chlorpyrifos at a dose of 6 mg/kg resulted in early effects on pup vocalizations (Venerosi et al., 2009) and, later, alterations in social vocalizations and social investigation in female offspring (Venerosi et al., 2006, 2015). There was “specific and sex-dependent vulnerability of affective/emotional domains” and a disruption of serotoninergic transmission in both male and female animals (Venerosi et al., 2010). Lan et al. (2017) also treated pregnant mice in mid-gestation and found a delayed development of neonatal reflexes; by PND 90 the mice at the high dose had reduced social behavior, while the low-dose mice showed increased restricted interest behaviors. Thus, these animal studies demonstrate that chlorpyrifos exposure during sensitive brain development periods may affect various behavioral domains, including cognition, motor behavior, anxiety, and social behavior, and these detrimental neurodevelopmental effects may persist to adulthood.

While chlorpyrifos has been the most extensively studied OP with regard to neurodevelopmental effects, one study did demonstrate long-term effects on behavior in rodents following early pregnancy exposure to dichlorvos (Lazarini et al., 2004).

The associations between early pregnancy exposure to OPs and malformations and transplacental cancer have also been studied. All registered pesticides are assessed for teratogenicity before EPA approval. For malathion, diazinon, and dichlorvos there are published studies that reported finding no teratogenesis (ATSDR 1997b, 2003a, 2008). Recently, studies using chlorpyrifos doses with high systemic toxicity during embryogenesis did produce malformations (Akhtar et al., 2006), and studies testing mixtures of OPs, or mixtures including OPs with other pesticides, were also positive (Gomes and Lloyd, 2009; Yu et al., 2017). At doses that produce high systemic toxicity, pregnancy outcome measures (resorption, fetal loss, neonatal death, growth retardation) were also affected by OPs. However, it is important to note that the studies that identified neurodevelopmental toxicity of OPs used doses below the threshold for systemic maternal toxicity.

Oxidative damage has been implicated as a possible mechanism of action for the developmental effects of OPs (Qiao et al., 2004). While it has been demonstrated that antioxidants protect against the reproductive effects of OPs in males (Mosbah et al., 2016), this has not been studied for neurodevelopmental effects in offspring.

Reproductive Effects

Chlorpyrifos exposure has been shown to result in significant effects on sperm quality and lead to significantly increased levels of disomy in sperm chromosomes (Figueroa et al., 2015; Martenies and Perry, 2013; Meeker et al., 2004a,b; Melgarejo et al., 2015; Miranda-Contreras et al., 2013; Sánchez-Peña et al., 2004). Occupational exposure to OPs has been associated with altered sperm integrity, specifically the sperm chromatin structure, although the association was only with the presence of the OP metabolite DETP (Sánchez-Peña et al., 2004). Adverse effects on sperm parameters and some reproductive hormones were also seen in agricultural workers with OP exposure, although the men were also exposed to other pesticides (Miranda-Contreras et al., 2013). Similar adverse effects on sperm parameters were seen in men with environmental exposures to OPs, although reproductive hormones were largely unaffected (Melgarejo et al., 2015). A systematic review of six studies assessing the effects of exposure to OPs on male reproduction concluded that all six studies reported significant negative associations between OP exposure, whether occupational or otherwise, and at least one semen parameter. However, the Volume 11 committee noted that in many of these studies exposure was assessed using a single nonspecific urinary measure of the metabolites to which most OPs devolve, and, as indicated above, these biomarkers have some inherent limitations (e.g., they reflect only recent exposure, they may reflect exposure to preformed metabolites, etc.). Meeker et al. (2004a,b) examined six sperm parameters and reported no significant associations between urinary levels of TCPy, a metabolite of chlorpyrifos, and sperm concentration, motility, morphology, or percentage DNA tail or comet extent in sperm; only a decrease in tail distributed moment reached significance.

Recent rodent studies also focused on the testicular toxicity of OPs. These studies treated rodents with a pesticide, usually over at least one sperm cycle, and found effects on sperm quality, serum hormones, testes and epididymis histopathology, biochemistry, and gene expression. Such rodent studies have been conducted for all the OPs recognized for Gulf War toxicants: chlorpyrifos (Mandal and Das, 2012; Sai et al., 2014), dichlorvos (Dirican and Kalender, 2012), diazinon (Leong et al., 2013; Zidan, 2009), and malathion (Bustos-Obregón and González-Hormazabal, 2003; Geng et al., 2015; Uzun et al., 2009). The Volume 11 committee found that these animal studies described findings that were supportive of male reproductive effects being associated with OPs. These studies used measures similar to those used in human studies (serum hormones, sperm quality, sperm DNA damage) and were able to add testes histopathologic and biochemical measures. Although sperm production is affected by chlorpyrifos exposure in both animal and human studies, the effects on spermatogenesis and male fertility after the discontinuation of exposure have not been elucidated in humans. The Volume 11 committee concludes that these animal studies are supportive of male reproductive effects associated with OPs in human studies, as both human and animal studies assess the same sperm parameters. Additionally, animal studies of ROS generation as the mechanism of testicular toxicity provide biological plausibility for the associations reported in human studies.

The Volume 11 committee identified two studies that assessed the effects of OP exposure on female reproduction. Although the study in China (Hu et al., 2018) showed increased TTP and reduced fertility, the second study did not show decreased fertility. OP effects on ovarian function either during or after exposure have not been examined in any human studies and have been examined in only a few animal studies.

The Volume 11 committee considered several major longitudinal U.S. birth cohorts—the CCCEH study in New York City, the CHAMACOS study in California, the Children's Environmental Cohort Study at Mt. Sinai Hospital in New York City, and the HOME study in Cincinnati, Ohio—that examined the association between prenatal exposure to one or more pesticides, including OPs, and birth outcomes.

The effects of prenatal exposure to chlorpyrifos specifically, or OPs in general, on birth outcomes were not consistent across the studies. In particular, the Volume 11 committee notes that there were ethnic and sex differences for several of the outcomes related to specific DAPs. The Harley et al. (2016) pooled analysis found no significant associations between increases in maternal DAP concentrations and decreases in birth weight, birth length, or infant head circumference for white and Hispanic women, but it did find among black women that increased prenatal OP exposure was associated with decreased infant birth weight, length, and head circumference. However, gestational duration could not be examined in this pooled analysis, and an association between OP exposure and shortened gestational duration was noted in two of the studies (CHAMACOS and HOME) that examined length of gestation (Eskenazi et al., 2004; Rauch et al., 2012). The Volume 11 committee also emphasizes that it was not possible in any of these cohort studies to say which particular OP the subjects were exposed to. Furthermore, maternal OP exposure may not be restricted to preconception, to the first trimester of pregnancy, or to pregnancy.

The Volume 11 committee found a lack of relevant information in the animal literature on adverse birth outcomes associated with preconception or prenatal exposure to any of the relevant OPs considered in this chapter.

The Volume 11 committee concludes that there is limited/suggestive evidence of an association between exposure to OPs and reproductive effects in men.

The Volume 11 committee also concludes that there is inadequate/insufficient evidence to determine whether an association exists between exposure to OPs and reproductive effects in women, or with adverse pregnancy outcomes.

Developmental Effects

Research into the developmental effects in infants or children following prenatal exposure to OPs has focused primarily on neurodevelopmental effects, although effects on the respiratory system and the presence of childhood cancers have also been assessed (see Table 5-1).

TABLE 5-1

Summary of Neurodevelopmental Effects of Organophosphate Pesticides.

The potential neurodevelopmental effects associated with prenatal exposure to OPs—and to chlorpyrifos in particular—have been the subject of several comprehensive reviews. In 2016 EPA concluded that the CCCEH study in New York City, along with other epidemiologic studies, provided “sufficient evidence that there are neurodevelopmental effects occurring at chlorpyrifos exposure levels below those required for AChE inhibition.” The Volume 11 committee reviewed a number of studies of child neurodevelopment and prenatal OP exposure conducted in Spain, Mexico, and China, as well as the CCCEH, CHAMACOS, and Mount Sinai, and HOME cohort studies. Results of many, but not all, of these studies indicate that prenatal exposure to OPs can result in lower scores on developmental indices such as IQ and other assessment scales and that these deficits may persist up to at least the 9th year of life. In addition, associations have been found with attention deficits and tremor. The CHARGE study in California found that close prenatal proximity to OP application was associated with an increased risk for ASD (Shelton et al., 2014), and Burke et al. (2017) stated that “the association between cord blood levels of chlorpyrifos and neurodevelopmental outcomes cannot be ignored.”

Assessment of the CHAMACOS cohort found increased respiratory symptoms, but no impairment of lung function, in children at 5 and 7 years of age as reported by mothers, although the increase was not associated with maternal DAP measures in the first half of pregnancy. Studies on the effects of prenatal OP exposure and the risks of childhood leukemia were inconsistent, although the one positive study examined both the father's and the mother's exposure prior to conception and in the first trimester. There was little information on the effects of prenatal OP exposure on the risk of birth defects or on other adverse childhood health effects.

In addition to the numerous epidemiologic studies discussed above, the Volume 11 committee considered more than 50 new animal studies, many of which discussed a variety of reproductive and developmental endpoints. Because of the large number of studies that assessed developmental effects from prenatal exposure to OPs, the committee described only those that specifically showed effects in early pregnancy compared with effects seen later in pregnancy. Animal studies indicate that early pregnancy exposures produce developmental toxicity, but it is unclear whether these effects would be expected for human pregnancies initiated after deployment ends since organ systems in animals and humans develop at different times during gestation.

The Volume 11 committee concludes that there is sufficient evidence of an association between prenatal exposure to OPs and neurodevelopmental effects.

The Volume 11 committee also concludes that there is inadequate/insufficient evidence to determine whether an association exists between prenatal exposure to OPs and other developmental effects.

Reproductive Effects in Men and Women

The Volume 2 committee found little relevant information on the reproductive and developmental effects of human or animal exposure to propoxur and methomyl. The committee did review several epidemiologic studies that assessed reproductive effects associated with exposure to either carbamates in general or to carbaryl specifically. The Volume 11 committee found a similar dearth of epidemiologic information on propoxur and methomyl and only one set of studies on carbaryl published since Volume 2.

In Volume 2, two studies on sperm effects in the same cohort of male carbaryl production workers were assessed (Whorton et al., 1979; Wyrobek et al., 1981); they reported that compared with unexposed controls there was an exposure-dependent increase in the number of oligospermic men among the male carbaryl production workers (p=0.07). There was also an increase in teratospermic men (having more than 60% abnormal sperm), but the increase was not dose-dependent. The latter study also found that men who were younger or who had more years of exposure had a greater proportion of abnormal sperm. The Volume 2 committee concluded that both of these cross-sectional studies had flaws in their methodology, including the choice of chemical workers as controls and crude measures of exposure.

The Volume 11 committee identified two new studies that examined reproductive effects in men (see Table 5-4). Meeker and colleagues (2004a,b, 2006, 2008a) completed a series of reports on the effects of low-level environmental exposure on human sperm and male reproductive hormones. Men (n=330) who were recruited from an infertility clinic provided semen and urine samples; the latter were analyzed for 1-naphthol, a urinary metabolite of carbaryl. None of the subjects reported being occupationally exposed to carbaryl or other pesticides. Compared with the lowest-exposure tertile, ORs for the medium and high tertiles of 1-naphthol and below-reference sperm concentration were 4.2 (95% CI 1.4–13.0), and 4.2 (95% CI 1.4–12.6; p-value for trend=0.01); for 1-naphthol exposures and sperm motility, the ORs were 2.5 (95% CI 1.3–4.7) and 2.4 (95% CI 1.2–4.5; p-value for trend=0.01). There was no association between 1-naphthol and abnormal sperm morphology. However, increasing carbaryl exposure was associated with increasing DNA damage as determined by the neutral comet assay. Reduced testosterone and estradiol levels were associated with carbaryl exposures but were not dose-dependent.

In a study of 16 production workers in China who had occupational exposure to carbaryl for at least 1 year, no significant differences in semen volume, sperm concentration, sperm number per ejaculum, or sperm motility were seen between the carbaryl-exposed workers and members of two control groups (12 unexposed workers in the same plant and 18 workers outside the plant) (Xia et al., 2005). However, morphological defects such as sperm abnormalities were significantly higher in the exposed workers than in the external control group (p=0.008), as were the frequencies of disomic sex-chromosome-bearing (highest the XY-bearing) and disomic chromosome 18 spermatozoa (p<0.05 and/or p<0.01, respectively).

The Volume 2 committee also considered several studies on fertility and carbamate exposure. Several researchers reported on the large Ontario Farm Family Health Study, which assessed both male and female exposures to pesticides during the month of attempted conception or at any time during the previous 2 months. Neither maternal-only nor paternal-only exposure to carbaryl was associated with significant effects on TTP; the close correlation of exposures between members of a couple made it difficult to determine whether any effects were specifically maternally or paternally mediated (Curtis et al., 1999).

Adverse Pregnancy Outcomes

In the Ontario Farm Family Health Study reviewed by the Volume 2 committee, both male and female exposures to pesticides were assessed during the month of attempted conception or at any time during the previous 2 months (Curtis et al., 1999). For men who reported working with carbaryl, there was a significant increase in spontaneous abortions, although there was no association with altered sex ratios, small for gestational-age deliveries, or preterm births (Savitz et al., 1997). A further analysis of preconception exposure for the couple as a unit showed no increased risk for overall, early, or late spontaneous abortions (Arbuckle et al., 2001).

Bell et al. (2001) studied the relationship between fetal death and maternal residential proximity to agricultural areas in California with pesticide applications. They found that exposure to carbamates from the third to eighth week of pregnancy was not significantly associated with an increased risk of fetal death due to congenital abnormalities. Wives of pesticide applicators were at increased risk of spontaneous abortion, although the Volume 2 committee noted that the study had few details on its methodology.

The Volume 11 committee reviewed one study (see Table 5-4) and one EPA summary on the effects of exposure to carbamate pesticides on fetal growth and birth outcomes.

Whyatt et al. (2004, 2005) assessed birth outcomes in women enrolled in the CCCEH study between 1998 and 2002. All women had detectable levels of chlorpyrifos, diazinon, and propoxur in personal air samples collected at their homes over a 48-hour period during their third trimester. The insecticides were detected in 45–74% of blood samples collected from the mothers and newborns at delivery. Among the birth outcomes measured in newborns (birth weight, birth length, and head circumference), for those born before 2001 only the association between 2-isopropoxyphenol (a propoxur metabolite) and decreased birth length was statistically significant (β= –0.73 cm/unit, p=0.01); for those born after January 1, 2001, the association remained inverse, but the magnitude of the effect was less and no longer significant (β= –0.30 cm/unit, p=0.56).

Animal Studies

Volume 2 reported on two animal studies on the effects of carbaryl on sperm. When 50 mg or 100 mg carbaryl/kg was fed to rats 5 days per week for 60 or 90 days, dose- and age-dependent decreases in sperm count and motility and increased abnormal sperm structure were observed, particularly in young animals (Pant et al., 1995, 1996). In a sperm-abnormality assay, rats fed carbaryl for 3, 6, 9, 12, 15, and 18 months at 12.5, 25, and 250 mg/kg per day had genotoxic effects (Luca and Balan, 1987).

The Volume 2 committee further reported that the reproductive toxicity of carbaryl has been examined in rats, gerbils, dogs, and primates following a variety of protocols, including multigeneration studies and studies to determine whether carbamate pesticides act as endocrine modulators. The oral administration of carbaryl to rats for 1 year resulted in a decreased sperm count, decreased motility, and abnormalities (Kitagawa et al., 1977). In a three-generation reproductive study in rats, even the lowest dose (100 mg/kg per day) decreased weaning weights (Collins et al., 1971). In a three-generation study of gerbils, a dose of 150 mg/kg per day was found to be the no-observed-effect level; higher doses were associated with decreases in fertility, litter size, viability, and survival (Collins et al., 1971). In two dog studies, teratogenesis was seen at doses similar to those that did not result in fetotoxicity (Imming et al., 1969; Smalley et al., 1968).

The EPA Integrated Risk Information System (IRIS) Chemical Assessment Summary (EPA, 1987) reported on one dog study in which doses of methomyl up to 1,000 ppm had no effect on reproductive organs. A 2016 EPA fact sheet for propoxur reported that no information was available on the reproductive or developmental effects of propoxur in humans and that no adverse reproductive or developmental effects had been observed in an oral study of rabbits; however, a few studies of rats orally exposed to propoxur found fetotoxic effects, decreased numbers of pups, and depressed fetal weight (EPA, 2000). No studies on adverse effects on maternal reproduction or fetal outcomes were identified.

The Volume 11 committee identified no additional studies of carbaryl exposure in animals, but it did identify one on propoxur and one on methomyl. Male rats exposed to propoxur at a dose of 5.2 mg/kg body weight had significant (p<0.05) decreases in body weight gain, sperm density, serum and intratesticular total cholesterol concentration, and intratesticular total protein. However, there were no significant effects on gestation, fertility and parturition indices, average birth weight, litter size, and pups' sex ratio for untreated females mated with treated males (Ngoula et al., 2007).

Male rats that were administered methomyl orally (17 mg/kg in saline) daily for 2 months had a significant decrease in their levels of testosterone, while their levels of FSH, LH, and prolactin were significantly increased. The testes had degenerative changes in the seminiferous tubules up to total cellular destruction (Mahgoub and El-Medany, 2001).

The 1992 EPA IRIS Chemical Assessment for Baygon (a brand name for propoxur) briefly summarized three animal studies conducted by the chemical industry: a three-generation reproduction study in rats that found decreased pup numbers at 37.5 mg/kg/day (the generation in which the effects were seen was not stated); a no-observed-effect level of 500 mg/kg/day in a teratogenicity study in rabbits; and a fetotoxicity study in rats that showed a significantly lower average fetal weight, at 150 mg/kg/day (EPA, 1992).

Animal Studies

No animal data on developmental effects from exposure to carbamate pesticides was reported in Volume 2, nor did the Volume 11 committee identify any new animal studies on the relevant carbamates.

Reproductive Effects

Information on the reproductive and developmental toxicity of the carbamate pesticides is limited almost exclusively to carbaryl, with little information on propoxur or methomyl. With the exception of occupational studies in carbaryl workers—the CCCEH study and the studies by Meeker et al.—virtually all other epidemiologic studies on male or female reproductive effects relied on self-reports of exposure to carbamates, often in conjunction with other potential pesticide exposures. This poses considerable challenges for the interpretation of the findings. The Volume 2 committee found adverse effects on sperm in the two occupational studies it reviewed, but it concluded that the studies were flawed. The Volume 11 committee identified one new occupational study in China that found an increase in sperm abnormalities (Xia et al., 2005) and a second study of men with no occupational exposure to carbaryl who attended an infertility clinic in the United States (Meeker et al., 2004a,b). The latter study also found sperm damage and decreased sperm concentrations to be associated with carbaryl exposure. Animal studies with carbaryl, propoxur, or methomyl supported the adverse effects on sperm seen in humans; other studies had inconsistent results for other reproductive effects in male animals, and there were no studies on female reproduction. Additional animal evidence indicates that carbamates, specifically carbaryl, act as an endocrine-disrupting compound, providing biological plausibility and mechanistic support for the reproductive effects in humans. There were no relevant animal studies on female reproductive effects.

Studies on adverse pregnancy outcomes associated with carbamate exposure reviewed by the Volume 2 committee showed that paternal occupational exposure to carbaryl increased the risk of spontaneous abortion, but when couples were analyzed as a unit, there was no increased risk (Curtis et al., 1999). Maternal exposure had no significant effects on birth outcomes. The Volume 11 committee identified only one new study that examined birth outcomes for women with exposure to chlorpyrifos, diazinon, and propoxur during pregnancy. Only propoxur was significantly associated with birth length, and it was not associated with birth length or head circumference at birth. The animal data provide inconsistent evidence of an association between exposure to carbamates and adverse reproductive effects in male and female animals. Adverse pregnancy outcomes were reported in two multigeneration studies of carbaryl.

The Volume 11 committee concludes that there is sufficient evidence of an association between exposure to carbamates and reproductive effects in men.

The Volume 11 committee also concludes that there is inadequate/insufficient evidence to determine whether an association exists between exposure to carbamates and reproductive effects in women, or with adverse pregnancy outcomes.

Developmental Effects

The epidemiologic studies on developmental effects in children exposed prenatally to carbamates, particularly the prevalence of childhood cancers, are of concern. While the CHAMACOS study (Gunier et al., 2017) indicated that there are some early neurocognitive deficits associated with maternal residential proximity to carbamate application, these deficits resolved with age (Rowe et al., 2016). The small study by Lafiura (2007) also showed neurodevelopmental effects with maternal carbamate exposure. Other studies were not restricted to carbamates alone. There is a lack of animal data on developmental effects in offspring following parental exposure to carbamates reported in Volume 2, and the Volume 11 committee did not identify any new animal studies.

The Volume 11 committee concludes that there is inadequate/insufficient evidence to determine whether an association exists between prenatal exposure to carbamates and developmental effects.

Reproductive Effects in Men and Women

The Volume 11 committee considered several papers that explored the association between environmental exposure of men to pyrethroids and the impacts on their reproductive systems, particularly sperm quality (see Table 5-5). The majority of the studies assessed exposure to pyrethroids by analyzing for urinary metabolites, including PBA, CDCCA, TDCCA, and DBCA. In general, the men in the studies were recruited from infertility clinics.

TABLE 5-5

Summary of Reproductive and Developmental Effects of Pyrethroid Pesticides.

Xia et al. (2008) reported a dose–response relationship between increasing urinary PBA quartiles and decreased sperm concentration in 376 Chinese men. ORs for the first, second, third, and fourth concentration quartiles were 1.00, 1.31 (95% CI 0.65–2.64), 1.73 (95% CI 0.87–3.45), and 2.04 (95% CI 1.02–4.09), respectively (p-value for trend=0.027), whereas sperm volume, sperm number per ejaculum, and sperm motility showed no significant associations with PBA quartiles. In another study of Chinese men, Ji et al. (2011) found a significant inverse correlation between urinary PBA levels and sperm concentration (β= –0.27, 95% CI –0.41– –0.12, p<0.001) and a significant positive correlation between PBA levels and sperm DNA fragmentation (β=0.27, 95% CI 0.15–0.39, p<0.001).

Imai et al. (2014) found in a sample of 323 Japanese men (university students) that environmental exposure to pyrethroid insecticides as determined by urinary PBA did not affect semen volume, sperm concentration, motility, the total number of sperm, or the total number of motile sperm. Furthermore, the researchers did not find any association between PBA and the levels of FSH, LH, testosterone, sex hormone-binding globulin, inhibin B, or calculated free testosterone. More than 91% of the men had detectable levels of PBA in their urine (Yoshinaga et al., 2014).

In a study of 195 men in Poland conducted to evaluate sperm DNA damage (Jurewicz et al., 2015) and aneuploidy (Radwan et al., 2015), there was a positive association between CDCCA greater than the 50th percentage concentration distribution in urine and the percentage of medium DFI, the percentage of immature sperm, and disomy of chromosome 18 (p=0.04, p=0.04, and p=0.05, respectively). A level of PBA was positively related to the percentage of high DFI (p=0.03). The TDCCA, DBCA levels, and the sum of pyrethroid metabolites were not associated with any sperm DNA damage measures. PBA concentrations in the urine that were both below and above the 50th percentile were associated with disomy of sex chromosomes, although to different degrees—XY disomy (p=0.05 and p=0.02, respectively), Y disomy (p=0.04 and 0.02, respectively), disomy of chromosome 21 (p=0.04 and p=0.04, respectively), and total disomy (p=0.03 and p=0.04, respectively); while PBA levels that were above the 50th percentile were associated with disomy of chromosome 18 (p=0.03). Urinary levels of TDCCA above the 50th percentile were related to XY disomy (p=0.04) and disomy of chromosome 21 (p=0.05) but not with any sperm DNA damage. The levels of DBCA and the sum of pyrethroid metabolites also were not associated with any sperm DNA damage measures. The three metabolites were also significantly associated with an increase in the number of sperm having abnormal morphology and a decrease in sperm concentration and levels of testosterone (Radwan et al., 2014).

One study of U.S. men also found environmental exposure to pyrethroids to be associated with semen quality. In a study of 207 men attending an infertility clinic in Massachusetts, Meeker et al. (2008b) found that increasing levels of urinary PBA and CDCCA were associated with increased sperm DNA damage. TDCCA levels showed a dose-dependent increase in the risk of having sperm concentration, motility and morphology below the reference values. Further assessment of these men showed that all three metabolites were associated with an increased incidence of disomy of chromosomes X, Y, and 18. Specifically, disomy was increased in YY18 for PBA, CDCCA, and TDCCA (incidence ratio rates [IRRs]=1.28, 95% CI 1.15–1.42; 1.18, 95% CI 1.05–1.31; and 1.19, 95% CI 1.06–1.34, respectively). Both CDCCA and TDCCA were significantly associated with increased disomy for XY18 and total disomy, but PBA was not. None of the metabolites were associated with disomy for XX18 or 1818 (Young et al., 2013).

Meeker et al. (2009) assessed the effects of pyrethroids on serum hormone levels in men and found that PBA, TDCCA, and CDCCA were all positively associated with FSH levels (all p-values for trend <0.05). CDCCA was also significantly associated with luteinizing hormone; however, CDCCA and TDCCA were inversely associated with inhibin B (p-values for trend=0.03 and 0.02, respectively).

In a systematic review of three papers—Ji et al. (2011); Meeker et al. (2008b); Xia et al. (2008)—published between 2007 and 2012 on the effects of exposure to pyrethroid insecticides on sperm parameters, Martenies and Perry (2013) concluded that although all three studies reported inverse associations between pyrethroid metabolites in urine and decreased sperm concentrations, the results were suggestive but not definitive.

The Volume 11 committee identified only one study on the effects of pyrethroids on female reproduction. Hu et al. (2018) conducted a TTP study of 615 women in Shanghai, China, who were recruited from two preconception care clinics between August 2013 and April 2015. Women were enrolled in the study before conceiving a child and were followed for approximately 1 year after they had stopped using a contraceptive. Exposures to pyrethroids were determined by measuring PBA, TDCCA, and CDCCA in urine. After adjustments were made for age, BMI before pregnancy, smoking, and education, women who had never been pregnant and were in the highest quartile for PBA concentrations had significantly longer TTP (fecundability OR=0.72, 95% CI 0.53–0.98) and increased infertility (OR=2.03, 95% CI 1.10–3.74) than women in the lowest quartile.

Adverse Pregnancy Outcomes

The ATSDR review Toxicological Profile for Pyrethrins and Pyrethroids (2003b) reported that no studies had been identified on the reproductive effects, including birth outcomes, of pyrethroid pesticides following inhalation, oral, or dermal exposures in humans. Several studies in Poland, China, and Japan have been published since the ATSDR profile. The studies examine the impact of prenatal exposure to pyrethroid pesticides on fetal outcomes.

In a 2003 study of the effect of first- and second-trimester exposures to pesticides on birth outcomes in Poland, women (n=95) who reported residential or occupational exposure to pesticides, including pyrethroids, were found to be at an increased risk (p=0.01) of delivering a low-birth-weight infant, but the risk was not significant for pyrethoid exposure specifically (p=0.286), and there was no effect on pregnancy duration (Dabrowski et al., 2003). The study is limited in that exposures were to mixtures of pesticides and based on the mother's recall of pesticide use at the time of delivery. As such, pesticide-specific effects cannot be evaluated.

The Volume 11 committee considered one study of birth outcomes in babies born to women in rural China who had prenatal exposure to pyrethroids (Ding et al., 2015). Maternal urine samples were analyzed for CDCCA, TDCCA, and PBA. The authors found that log₁₀ unit increases in the total concentration of the metabolites, but not individual metabolite levels, were associated with a decrease in birth weight (β= –96.76, 95% CI= –173.15– –20.37). No associations were found between individual or total metabolite levels and birth length, head circumference, or gestational duration.

Zhang et al. (2014b) analyzed the urine of 147 Japanese women in their first trimester of pregnancy for PBA metabolites. Thyroid hormone levels—free thyroxine (FT4) and thyroid-stimulating hormone (TSH)—were determined in neonatal and maternal blood at birth. The levels of both thyroid hormones were within the normal range for all infants except one. There was no significant correlation between PBA concentrations in maternal urine and neonatal thyroid hormone concentration or with body size at birth (p=0.05), nor was there a significant correlation between the levels of maternal PBA and other neonatal outcomes of body weight, body length, head circumference, or chest circumference. However, in a multiple-regression analysis the authors found that PBA concentrations in maternal urine sampled in the first trimester were significantly associated with reduced birth weight (β=0.183, p=0.017) and head circumference (β=0.243, p=0.010).

The Baltimore THREE (Tracking Health Related to Environmental Exposures) study collected cord blood from 300 singleton live births delivered between November 2004 and March 2005 at Johns Hopkins Hospital. The cord blood was analyzed for the cis- and trans-isomers of permethrin, for the permethrin synergist piperonyl butoxide, and for nine cytokines. Infants were evaluated for birth weight, length, head circumference, ponderal index, and gestational age. Permethrin was negatively associated with the anti-inflammatory cytokine IL-10 (β= –0.14, 95% CI –0.22– –0.05). The permethrin concentrations measured in the cord blood had no significant associations with gestational age, birth weight, length, head circumference, or ponderal index (Neta et al., 2011).

Animal Studies

The Volume 2 committee considered several studies on the reproductive effects of pyrethroids in animal models. Pyrethroids at 100–10,000 nM were found not to be estrogenic or antiestrogenic, nor did they alter human estrogen-receptor α-mediated mechanisms in vitro (Saito et al., 2000; Sumida et al., 2001). The Volume 2 committee cited numerous studies that indicated that pyrethroids are not gonadotoxic, embryotoxic, or teratogenic (EPA, 1983; Hallenbeck and Cunningham-Burns, 1985; Kaloianova and El Batawi, 1991; Kidd and James, 1991; Litchfield, 1985; Miyamoto, 1976; Polakova and Vargova, 1983; Ray, 1991), although high oral doses (250 mg/kg per day) of permethrin during GDs 6–15 were found to reduce the number of offspring (Wauchope et al., 1992). A three-generation study of resmethrin reported slight increases in premature stillbirths and a decrease in pup weight (Ray, 1991). The Volume 2 committee concluded that pyrethroids are unlikely to cause reproductive and teratogenic effects (IOM, 2003).

The ATSDR review Toxicological Profile on Pyrethrins and Pyrethroids (2003b) reported that a few studies have indicated that relatively low-level, intermediate-duration oral exposure of adult male rats to some pyrethroids may result in damage to reproductive organs, abnormal sperm characteristics, reduced plasma testosterone levels, and reduced fertility. Relative to controls, male rats administered oral doses of deltamethrin had significantly reduced sperm cell concentrations, live cell percentage, and motility index; had a significantly higher percentage of total sperm abnormalities; and had significantly reduced plasma testosterone levels. Male fertility, determined by successful matings to untreated female rats, was 50% that of controls (Abd El-Aziz et al., 1994).

ATSDR also reported on an EPA study that showed decreased pup survival in rats following parental oral exposure to resmethrin at 70.8 mg/kg/day prior to mating and throughout gestation and lactation; rats receiving 80 mg/kg/day during GDs 6–15 had slight increases in skeletal variations and delayed ossification (EPA, 1994).

The Volume 11 committee identified several animal studies that assessed the reproductive toxicology of all three pyrethroids of interest. One study examined the effect of 20 or 40 mg/kg/day permethrin on rat ovaries (Kotil and Yon, 2015). After treatment for 14 days via gavage, dose-dependent degenerative changes were seen in the follicular and corpus luteum cell morphology, with evidence of apoptotic cell death. All the other studies assessed the reproductive toxicity of pyrethroids in male animals.

In studies of male reproductive effects, Zhang et al. (2007, 2008) dosed male ICR mice with cispermethrin at 0, 35, or 70 mg/kg/day for 6 weeks. Caudal epididymal sperm count and sperm motility, testicular testosterone production, and plasma testosterone concentrations were all significantly and dose-dependently decreased, with an increase in LH. Testicular mitochondrial mRNA expression was also reduced, with mitochondrial membrane damage in the Leydig cells. Cis-permethrin administered at 90 μmol/kg/day for 6 weeks did not affect plasma testosterone levels, although the co-exposure of animals with 10 μmol/kg/day diazinon reduced plasma testosterone levels to 45% of those in the control animals. Cis-permethrin alone also reduced translocator protein mRNA levels by 46% compared with controls but did not reduce P450scc mRNA levels. Administration of permethrin to adult male rats at 35 mg/kg/day for 60 days resulted in reductions in reproductive organ weights, the number of live sperm and Leydig cells, sperm motility, and serum testosterone levels and in increased abnormal sperm morphology (Khaki et al., 2017).

Deltamethrin was also found to have adverse effects on male reproduction in animal studies, including reduced sperm quality, sex hormones, and libido and increased testicular damage (Ben Slima et al., 2017). Issam et al. (2009) treated male rats with deltamethrin at 2 ppm for 30 days, 20 ppm for 45 days, and 200 ppm for 60 days and reported a significant decrease in FSH, LH, and testosterone at the highest dose. The deltamethrin treatment of rats also arrested spermatogenesis and produced a significant increase in malondialdehyde levels that was related to dose, the length of treatment, and lipid peroxidation. Abdallah et al. (2010) studied the effect of deltamethrin at doses of 0, 10, 50, 100 and 200 μmol on the oxidative stress response in rat spermatozoa after 3 hours of incubation. This in vitro exposure caused a significant decline in sperm motility and viability and increased abnormal sperm morphology as well as levels of malondialdehyde, superoxide dismutase, and catalase. The administration of deltamethrin at 5 mg/kg/day to male rats for 4 weeks resulted in decreased serum testosterone, LH, and FSH levels and significantly increased testicular total oxidant capacity, poly(ADP-ribose) polymerase, lactate dehydrogenase, and DNA damage (Ismail and Mohamed, 2012). A dose of 1 mg/kg deltramethrin administered intraperitoneally for 5 days/week for 1 month increased the weight of rat testes and produced adverse histopathologic changes. There was also a significant reduction in the number of seminiferous tubules per unit area and adverse morphologic changes in the tubules (Kumar and Nagar, 2014).

Cypermethrin was found to have similar effects to deltamethrin on male rodents. Elbetieha et al. (2001) studied the effects of cypermethrin on reproductive and fertility parameters in male mice receiving 0, 13.15, 18.93, and 39.66 mg cypermethrin in drinking water for 12 weeks. All doses resulted in a significant reduction in the number of viable fetuses in females mated to the treated males, lower body weight gain in the treated males, a significant increase in preputial gland weights, and a significant decrease in epididymal and testicular sperm counts and daily sperm production. At the two higher doses, there was a significant increase in the weight of the testes and seminal vesicles and a significant reduction in the perimeter and number of cell layers of the seminiferous tubules, along with hemorrhage in the testes and a large number of immature spermatids in the tubules. The two lower doses (13.15 and 18.93 mg) produced significantly reduced fertility, as indicated by fewer females being impregnated by the treated males, while 39.66 mg cypermethrin caused a significant reduction in the number of implantation sites. On the other hand, Al-Hamdani and Yajurvedi (2010) did not see a decrease in fertility in mice administered cypermethrin by gavage at 1.38, 2.76, and 5.52 mg/kg/day for 6 or 12 weeks. All doses and both durations caused a significant reduction in epididymal spermatozoa count and an increase in abnormal spermatozoa, but 6 weeks after dosing ceased, the counts returned to normal in mice administered the two lower doses but not the highest dose. All mice had normal fertility, although the weight of the litters was significantly lower than for controls; the gestation period and the litter size were unaffected. Dahamna et al. (2010) also found that cypermethrin administered to male mice at 20% and 5% of the LD₅₀ (485 mg/kg) for 2 weeks and 4 weeks, respectively, and at 20% of the LD₅₀ for 12 weeks, resulted in a 20% decrease in sperm count and an increase in abnormal sperm and in testicular and epididymal morphology. Similar results were seen by Hu et al. (2013) who treated rats with cypermethrin at 0, 6.25, 12.5, 25 and 50 mg/kg/day by gavage for 15 days. At 25 and 50 mg/kg there was a significant reduction in sperm production and changes in the seminiferous tubules, including reduced and deformed spermatogonia. Furthermore, at 50 mg/kg cypermethrin, androgen receptor levels were significantly reduced, as were serum testosterone levels. Rodriguez et al. (2017) also found adverse effects on Sertoli cells in mice administered 20% of the LD₅₀ intraperitoneally.

Sharma et al. (2018) found that the administration of cypermethrin, deltamethrin, or a combination of the two pesticides significantly reduced sperm count and motility and concentrations of some steroidogenic enzymes and increased the number of abnormal sperm; the combination of the two pesticides had a greater effect on all sperm parameters and enzyme levels. Both pesticides and their combination also significantly reduced testosterone and LH concentrations, but only the combination of the two pesticides significantly reduced concentrations of FSH.

The mechanism by which pyrethroids act on male fertility was explored by Taylor et al. (2010) using an in vitro mouse Sertoli cell assay. Pyrethroids, at both low and high serum concentrations, affected the expression of the two estrogen receptors; however, the influence on estrogen receptor gene expression was different from the effect of exposure to 17beta-estradiol.

Neurodevelopmental Effects

The CCCEH study assessed 348 children born between February 1998 and May 2002 for psychomotor and cognitive development at several times during their childhoods (Horton et al., 2011). In this study, the children were administered the BSID-II to generate MDI and PDI scores. Prenatal exposure to permethrin was measured in maternal and umbilical cord plasma collected on delivery, and permethrin and piperonyl butoxide levels were measured in personal air collected during pregnancy. When the children were assessed at 36 months of age, the authors found no significant associations between their MDI or PDI scores and permethrin concentrations in either the personal air samples or the maternal or umbilical cord plasma samples. However, the highest prenatal piperonyl butoxide exposure was associated with a 3.9-point mean decrement in MDI scores (95% CI –0.25– –7.49), and the most highly exposed children had the greatest risk (OR=3.11, 95% CI 1.38–6.98) of delayed mental development.

In a further follow-up of the CHAMACOS study in California described in the section on OPs, Gunier et al. (2017) used the Pesticide Use Reporting data from California to determine the use of pyrethroids in fields near the maternal residence during pregnancy. Neurodevelopment in children at age 7 was assessed using WISC-IV. Pyrethroid use was associated with lower scores on the full-scale IQ (β= –2, 95% CI –3.7– –0.3), verbal comprehension (β= –1.8, 95% CI –3.4– –0.3), and perceptual reasoning (β= –2.1, 95% CI –4.0– –0.2) subscales, but not on the working memory or processing speed scores.

The Mt. Sinai Children's Environmental Health Cohort, described in the section on OPs, assessed 162 mother–infant pairs during pregnancy in 1998–2001. Prenatal exposure to pyrethroids was determined by an analysis of maternal urine samples taken in the third trimester of pregnancy; the samples were analyzed for PBA, TDCCA, and CDCCA. Children were assessed for neurodevelopment at 4, 6, and 7–9 years of age. Parents reported on their children's development using the Behavioral Assessment System for Children and the Behavior Rating Inventory of Executive Function. Detectable levels of PBA during pregnancy were associated with lower scores for the indices of internalizing (β= –4.50, 95% CI –8.05– –0.95), depression (β= –3.21, 95% CI –6.38– –0.05), somatization (β= –3.22, 95% CI –6.38– –0.06), behavioral regulation (β= –3.59, 95% CI –6.97– –0.21), emotional control (β= –3.35, 95% CI –6.58– –0.12), shifting (β= –3.42, 95% CI –6.73– –0.11), and monitoring (β= –4.08, 95% CI –7.07– –1.08). Detectable levels of CDCCA were associated with worse externalizing (β= –4.74, 95% CI –9.37– –0.10), conduct problems (β= –5.35, 95% CI –9.90– –0.81), behavioral regulation (β= –6.42, 95% CI –11.39– –1.45), and inhibitory control (β= –7.20, 95% CI –12.00– –2.39) (Furlong et al., 2017b).

The impact of maternal exposure to pyrethroids on the development of 102 18-month old Japanese infants was examined by Hisada et al. (2017) using the Kinder Infant Development Scale, which is based on a questionnaire completed by the caretaker. Infants whose mothers had higher levels of PBA in their urine at 10–12 weeks of gestation had higher development quotient scores than infants whose mothers had lower PBA concentrations (p=0.017), which the authors considered to be unexpected.

Other studies have found in utero pyrethroid exposure to be associated with adverse effects on neurodevelopment. Xue et al. (2013) found that among 497 mother–infant pairs in China, increasing levels of pyrethroid metabolites in maternal urine during pregnancy were inversely correlated with the development quotient of the infants at 1 year of age (β= –0.1527, p=0.05), although there was no such correlation for the MDI. The Development Screen Test scale was used to assess the intellectual development of infants. Children born to mothers residing in rural areas had greater impairment than children born to mothers living in cities. The impairment was correlated with higher levels of pyrethroid metabolites in the urine of mothers residing in rural areas.

The impact of maternal exposure to pyrethroids on child neurodevelopment was also assessed by Watkins et al. (2016) in the ELEMENT longitudinal study of a mother–child cohort in Mexico City recruited from 1997 to 2001. PBA was measured in the urine of 187 mothers during their third trimester, and children were administered the BSID-II Spanish version (BSID-IIS) to assess the developmental functioning of infants and children at 24 and 36 months of age using the MDI and PDI. Children whose mothers had medium and high urinary PBA levels had nonsignificantly lower MDI scores (β= –3.54, 95% CI –7.86−0.78 and β= −3.80, 95% CI –8.44−0.84, respectively) at 24 months than those with mothers in the low (reference) PBA category (p-value for trend=0.07), and the effect was lower at 36 months (p-value for trend=0.14). At 24 months the associations were significantly stronger for female children (β= –6.20, 95% CI –12.3− –0.14) than for males (β= –2.58, 95% CI –9.10−3.94) of mothers with medium PBA levels, but the differences were not significant at 36 months. The third-trimester prenatal level of maternal PBA was not associated with PDI scores at either time point.

Eskenazi et al. (2018) assessed the effects of prenatal exposure to multiple pesticides on neurodevelopment, as measured by the BSID-II, in 752 children in South Africa participating in the Venda Health Examination of Mothers, Babies and Their Environment (VHEMBE) birth cohort study. Maternal exposure to pyrethroids during pregnancy was determined through the identification of four urinary pyrethroid metabolites—DBCA, CDCCA, TDCCA, and PBA. Maternal urine samples were collected around the time of delivery. When the children were 1 year of age, each 10-fold increase in CDCCA, TDCCA, and PBA was significantly associated with a decrement in social–emotional scores (β= –0.63, 95% CI –1.14– –0.12; β= –0.48, 95% CI –0.92– –0.05; and β= –0.58, 95% CI –1.11– –0.06, respectively). At 2 years of age, each 10-fold increase in maternal CDBCA levels was associated with significant decrements in language composite scores (β= –1.74, 95% CI –3.34– –0.13) and expressive communication scores (β= –0.40, 95% CI –0.77– –0.04). The authors noted that “significant differences by sex were estimated for pyrethroid metabolites and motor function scores at 2 years of age, with higher scores for boys and lower scores for girls.”

The PELAGIE study in Brittany, France, is a large mother–child cohort that has assessed child development since the children were born between 2002 and 2006 (Viel et al., 2015). Maternal exposure to pyrethroids was determined on the basis of urine samples collected between 6 and 19 weeks of pregnancy and adjusted for exposure to OPs. At a child's sixth birthday, the mother and child participated in a neuropsychological follow-up using the WISC-IV; a total of 287 mother–child pairs took part. None of the WISC scores were associated with maternal pyrethroid metabolite concentrations (PBA, TDCCA, or CDCCA). The children's behavior was further evaluated for three subscales of the Strengths and Difficulties Questionnaire: prosocial behavior, internalizing disorders, and externalizing disorders (Viel et al., 2017). Maternal prenatal CDCCA was associated with internalizing difficulties, while maternal prenatal PBA was associated with externalizing difficulties and higher odds of abnormal or borderline social behavior per the maternal report on the Strengths and Difficulties Questionnaire (Viel et al., 2017).

The CHARGE study in California attempted to link pesticide use near the mother's residence with children having ASD (n=486) or developmental disorder (n=168) (Shelton et al., 2014). Prenatal residence proximity (within 1.5 km) to areas sprayed with pyrethroids was associated with an increased risk for ASD if the exposure was prior to conception (OR=1.82, 95% CI 1.00–3.31) or in the third trimester (OR=1.87, 95% CI 1.02–3.43). Exposure at 1.75 km was associated with an increased risk for ASD only in the third trimester (OR=1.83, 95% CI 1.04–3.23). For developmental disorders and residential proximity to agricultural pesticide application, there was a significant risk only for a distance of 1.75 km during the third trimester (OR=2.34, 95% CI 1.18–4.67), indicating that the timing of exposure during pregnancy had a significant effect on the risk of ASD and developmental disorder.

Other Developmental Effects

Researchers have linked other developmental endpoints to prenatal exposure to pyrethroids. Carmichael et al. (2014), discussed in the section on OPs, compared the exposure of mothers of 569 infants with any of eight congenital heart defects with 785 mothers who had an infant without a birth defect. Prenatal exposure to permethrin was nonsignificantly associated with ventricular septal defect, perimembranous (OR=2.2, 95% CI 0.8–5.7), but not with any of the heart defects. Monge et al. (2007) examined the effect of maternal (n=876) and paternal (n=762) exposures to pesticides on the risk of childhood leukemia (n=334) in Costa Rica in 1995–2000. Parents were interviewed about risk factors for childhood leukemia, and their exposures were rated by hazard value and by time of occurrence. Neither maternal nor paternal exposure to pyrethroid insecticides, primarily deltamethrin, at any time prior to birth was significantly associated with childhood leukemia for either boys or girls.

The association between prenatal exposure to permethrins and piperonyl butoxide, a synergist for residential pyrethroid pesticides, and the presence of childhood cough at 5–6 years of age was examined by Liu et al. (2012) as part of the CCCEH birth cohort. Prenatal exposure to cis- and trans-permethrin was not significantly associated with a noninfectious cough at age 5–6 years, but there was a positive association between a noninfectious cough and prenatal piperonyl butoxide in children ages 5–6 years (OR=1.27 for unit change in log piperonyl butoxide, 95% CI 1.09–1.48). Exposure to piperonyl butoxide or permethrins at ages 5–6 years was not associated with an increase in childhood cough.

Coker et al. (2018) assessed the effects of prenatal exposure to multiple pesticides on child body weight and body composition in 708 children in South Africa's VHEMBE birth cohort. Maternal exposure was determined by the identification of four urinary pyrethroid metabolites—DBCA, CDCCA, TDCCA, and PBA. Maternal urine samples were collected around the time of delivery. The length or height and the weight of the children were measured at 1 and 2 years of age; maternal exposure to dichlorodiphenyltrichloroethane (i.e., DDT) was also assessed. Using a single-pollutant linear mixed-effects model, BMI-for-age and weight-for-height were inversely related to DBCA and TDCCA levels; the effects were more pronounced for boys than for girls for both measures and metabolites. Weight-forage was also inversely related to TDCCA for boys only, but not overall. None of the four metabolites significantly affected any of the measures for girls.

Animal Studies

The Volume 2 committee reported that pyrethroids administered to female rats during pregnancy did not affect the development of their offspring (Extoxnet, 1994). No other animal studies on the developmental effects of pyrethroids were cited in Volume 2.

The ATSDR review Toxicological Profile on Pyrethrins and Pyrethroids (2003b) did not find the typical signs of developmental toxicity in animals exposed to pyrethrins or pyrethroids, particularly deltamethrin. There were no studies of reproductive or developmental effects following inhalation. ATSDR reported that numerous studies of pyrethorids administered orally in rats and mice did not indicate any developmental toxicity at doses below those which cause maternal toxicity. In utero exposure of rats resulted in cellular changes indicative of compromised immunological function. Santoni et al. (1997, 1998, 1999) reported treatment-related increases in natural killer (NK) cell and antibody-dependent cytotoxic activity, impaired thymocyte function, increased levels of T cells in peripheral blood, and decreased levels of T cells in the spleen in rats after their mothers were given 50 mg/kg/day cypermethrin orally during gestation—a dose schedule that did not result in clinical signs of maternal toxicity. Malaviya et al. (1993) observed significant increases in the levels of dopamine and muscarinic receptors of striatal membrane in rat pups that had been exposed to 15 mg cypermethrin/kg on GDs 5–21.

The Volume 11 committee identified three new studies and a systematic review that assessed the effects of preconception and prenatal exposure to permethrin on the development of offspring. In a two-generation study with mice, males and females were given 0, 4.9, 9.8, and 19.6 mg/kg/d permethrin by gavage for 4 weeks before mating (Farag et al., 2006). The highest dose produced signs of toxicity in the F0 mice, with body weight loss in the females during gestation and lactation. There was a significant (p<0.05) reduction in fetal weight and in the number of live pups born to females receiving the two highest doses. Postnatal assessment of the F1 offspring showed that at 9.8 and 19.6 mg/kg there was a significant reduction in righting reflexes, swimming ability, and open field activity and a slowed development of cliff avoidance; these effects were not seen in pups whose dams received 4.9 mg/kg permethrin. Imanishi et al. (2013) found other effects of prenatal permethrin exposure on neurodevelopment. Pregnant mice received a single dose of 0, 2, 10, 50, or 75 mg/kg permethrin on GD 10.5. On GD 17.5, exposure to 10 mg/kg resulted in significant increases in abnormalities of the anterior communicating arteries in the brain of male fetuses but not of female fetuses. Fetal brains obtained on GD 17.5 had anatomical abnormalities, including altered vascular formation, and on PND 7 rat brains had alterations in the thickness of the neocortex and hippocampus and significantly increased norepinephrine and dopamine levels. Even the lowest dose of 2 mg/kg showed significant influences on vascular development in fetuses and on adult offspring behaviors. At 8 and 12 weeks of age, male mice had a significant reduction in standing ability and locomotor activity at the 2 mg/kg dose; these effects were not observed in female offspring.

One new study on the effects of deltamethrin on adult male offspring of treated dams was identified (Ben Slima et al., 2012). Pregnant mice received 5 mg/kg/day deltamethrin by gavage from GD 3 to GD 21. Male offspring were assessed for fertility when the animals achieved sexual maturity. Although the body weights of the dams were significantly reduced, there were no significant differences in absolute and relative epididymides weights in offspring. However, there were significant differences in absolute and relative testes weights (reduced to approximately half of the control value, p≤0.05); reduced spermatozoa counts and percent of motile and viable sperm; increased abnormal sperm; and increases in testicular histopathology.

Saillenfait et al. (2017) assessed the effects of prenatal exposure to cypermethrin on fetal testicular steroidogenesis. Cypermethrin (0.1, 1, 5, or 10 mg/kg/day) administered on GDs 13–19 had no effect on anogenital distance of the male fetuses assessed on GD 19, but the two highest doses reduced testosterone production. Testicular mRNA expressions of HMG-CoA synthase and reductase, and the genes SRB1, StAR, P450scc, 3βHSD, P450 17A1, and 17βHSD, were not affected at any dose.

A systematic review of inherited transgenerational effects was conducted by the National Institute of Environmental Health and Sciences (Walker et al., 2018). The researchers found that there was no evidence of transgenerational effects of pesticide exposure in humans. However, there were three rodent studies—all from the same group of investigators—that examined whether transgenerational effects might occur following gestational exposure to a mixture of permethrin and DEET. The studies assessed the effects of the pesticide mixture on growth and development, reproductive effects in offspring, and renal effects in offspring. Two studies evaluated the body weight of F3 offspring and reported no impact of exposure on post-weaning weight at PND 21, but an increased weight at 12 months of age (Manikkam et al., 2012a,b). Two studies reported on male reproductive effects with no associations for most endpoints (Manikkam et al., 2012a,b). Three studies (Manikkam et al., 2012a,b; Nilsson et al., 2012) reported decreased ovarian follicle counts and increased ovarian cysts in female rats of the F3 generation of F0 gestating females exposed to 190 mg/kg/day permethrin and DEET, but inconsistent effects on puberty and no effects on fertility and ovarian weight were reported. Confidence in these findings is limited by inconsistencies across the studies (heterogeneity of endpoints, differences in the measurement of endpoints), a lack of studies from other research groups, and a lack of analysis using the litter as the unit of analysis.

Reproductive Effects

Several studies have assessed the effects of exposure to pyrethroids on human male reproductive parameters. Most of the studies assessed pyrethroid exposure by measuring the levels of the metabolite PBA in urine. With the exception of one study in Japanese men (Imai et al., 2014), which found no association between PBA levels and semen parameters or male reproductive hormones, the other five studies considered by the committee found that increasing levels of pyrethroid metabolites in urine were associated with decreased sperm concentrations and motility and increased sperm abnormal morphology and DNA damage and disomy. Numerous animal studies in rats and mice also indicate that exposure to permethrin, cypermethrin, and deltamethrin has adverse effects on male reproduction, especially sperm parameters and reproductive hormones. One limitation of these studies is that the PBA reflects only recent exposure, given the short half-life of pyrethroids, and in most cases only a single measurement of pyrethroid metabolites was made.

The Volume 11 committee identified one study of the effects of pyrethroid exposure on female reproduction and it showed increased TTP and reduced fertility (Hu et al., 2018).

Several epidemiologic studies on the effects of prenatal exposure to pyrethroids on birth outcomes were considered. When the assessed exposure to pyrethroids was based on the concentration of metabolites in urine, there was a negative association between exposure and birth weight, although this was not seen in a study where permethrin exposure was based on its concentration in cord blood. There were no animal studies that added to the evidence base for effects on female reproduction or birth outcomes.

The Volume 11 committee concludes that there is limited/suggestive evidence of an association between exposure to pyrethroid pesticides and reproductive effects in men.

The Volume 11 committee also concludes that there is inadequate/insufficient evidence to determine whether an association exists between exposure to pyrethroid pesticides and reproductive effects in women, or with adverse pregnancy outcomes.

Developmental Effects

The effects of prenatal pyrethroid exposure on the neurodevelopment of children were assessed in several studies with mixed results. Hisada et al. (2017) found that mothers' exposure to pyrethroids during the first trimester of pregnancy was positively associated with their children's development scores at 18 months of age—that is, greater exposure in the mothers was associated with better performance in the children. This conclusion was contradicted by several other studies, including that of Watkins et al. (2016), who found that children born to mothers with higher PBA concentrations during any trimester of pregnancy had lower scores on developmental tests at 24 months of age; a study by Xue et al. (2013) in China that reported an association between pyrethroid metabolite concentrations in the mother's urine and the developmental quotient in their children at 1 year of age; a study by Gunier et al. (2017) that used geospatial data to estimate prenatal exposure and found negative associations with child IQ at age 7 years; a study by Furlong et al. (2017b) that found increased PBA and CDCCA were both associated with decrements in children's neurodevelopment at 4, 6, and 7–9 years of age; and a study by Shelton et al. (2014) that found that preconception or third-trimester exposure (but not first- or second-trimester exposure) to pyrethroids (as determined by proximity to areas of agricultural pesticide application) was associated with an increased risk of ASD. However, the PELAGIE study in France found no significant association between a mother's urinary pyrethroid metabolite concentrations during early pregnancy and cognitive outcomes in their children at age 6 years, with the exception of one developmental subscale (Viel et al., 2015, 2017), and a study from Japan (Hisada et al., 2017) found better developmental quotients associated with prenatal exposure.

Prenatal exposure to pyrethroids was not associated with childhood leukemia. Studies of the effects of prenatal exposure to permethrin and piperonyl butoxide on noninfectious cough at ages 5–6 years were inconclusive. Although exposure to pyrethroids was associated with reduced child growth at 1 and 2 years of age, the effect was limited to boys only and for only two of the four metabolites measured (Coker et al., 2018).

Animal studies reviewed by the Volume 2 committee and ATSDR did not indicate that prenatal exposure of animals to pyrethroids causes postnatal developmental effects, although one research group did find effects on immune function that deserve further study. The Volume 11 committee considered two additional animal studies of pyrethroids that indicated that prenatal exposure at levels that do not cause maternal toxicity may result in neurodevelopmental effects (structural brain abnormalities, poor performance on neurodevelopmental tests) in offspring. Although there is evidence in rodent models that transgenerational effects are possible (Walker et al., 2018), their relevance to human health is not yet clear.

The Volume 11 committee concludes that there is limited/suggestive evidence of an association between prenatal exposure to pyrethroid pesticides and developmental effects.

Reproductive Effects in Men and Women

The Volume 2 committee (IOM, 2003) and ATSDR (2005) both identified only one study that assessed lindane's reproductive effects in men—an examination of workers at a lindane production facility for 8 years (Tomczak et al., 1981). Blood samples from the exposed workers showed elevated serum LH concentrations, somewhat elevated levels of FSH, and somewhat depressed testosterone levels.

The Volume 11 committee identified three studies on female reproductive effects from exposure to lindane. In a study of 94 couples at a fertility clinic in Egypt, follicular fluid from each woman was collected during in vitro fertilization–embryo transfer and analyzed for the presence of lindane (mean=418.6±171.4 [sd] μg/L) among other pesticides and polychlorinated biphenyls (Al-Hussaini et al., 2018). There were negative correlations between the follicular fluid concentration of lindane and endometrial thickness (p=0.0001), and a high concentration of lindane was associated with a lower implantation rate (p=0.006). Lindane did not reduce fertilization or early embryo cleavage rates.

In an earlier study of the association between exposure to lindane and endometriosis, Buck Louis et al. (2012) assessed 473 women 18–44 years of age scheduled for a laparoscopy or laparotomy (operative cohort) and an age and residence matched population cohort of 127 women in San Francisco, California, between 2007 and 2009. Serum and fat samples were analyzed for HCH and its β and γ isomers along with other persistent organic pollutants. Lindane in the fat of the operative cohort was significantly associated with endometriosis (OR=1.27, 95% CI 1.01–1.59). Serum β-HCH was significantly associated with endometriosis in the population cohort (OR=1.72, 95% CI 1.09–2.72), but not the operative cohort; fat samples were not analyzed for the population cohort, only sera. In another study in western Washington State, conducted from 1996 to 2001, Upson et al. (2013) compared levels of several endocrine-disrupting chemicals in the sera of 248 women with endometriosis and 538 population-based women. β-HCH—but not γ-HCH (lindane)—was associated with endometriosis (third versus lowest quartile OR=1.7, 95% CI 1.0–2.8; highest versus lowest quartile OR=1.3, 95% CI 0.8–2.4), and the risk was even greater for ovarian endometriosis specifically (highest versus lowest quartile OR=2.5, 95% CI 1.1–5.3). A third study of 55 women in France with endometriosis (24 with deep infiltrating endometriosis only and 25 with deep infiltrating endometriosis and ovarian endometriosis) and 44 controls was conducted between 2013 and 2015 (Ploteau et al., 2017). Concentrations of β-HCH in adipose tissue collected during surgery were not associated with an increased risk of deep infiltrating endometriosis (OR=1.58, 95% CI 0.94–2.80), but there was a significant association with deep infiltrating endometriosis combined with ovarian endometriosis (OR=3.64, 95% CI 1.52–11.15). The researchers analyzed for γ-HCH in the fat tissue, but the detection rate was so low that further analysis was not performed (Ploteau et al., 2016).

Adverse Pregnancy Outcomes

Volume 2 discussed two studies that examined the effects of lindane on fetal outcomes; however, lindane was not a unique exposure in either study. Willis et al. (1993) found that among 535 pregnant women who resided in an agricultural area in San Diego, California, only those who were not exposed to pesticides—which included carbaryl and lindane—had spontaneous abortion or preterm delivery; furthermore, the unexposed women also had a greater incidence of low-birth-weight infants. The Volume 2 committee found this study hard to interpret because critical information such as the number of exposed women in the cohort was not presented.

The second study assessed paternal insecticide exposure and pregnancy outcomes among 32 Italian pesticide applicators and their wives (Petrelli et al., 2000). Although there was an increased risk of spontaneous abortion among the exposed wives (OR=3.8, 95% CI 1.2–12.0), there was little information on the selection and recruitment of study subjects, and a crude exposure measurement was used.

The 2005 ATSDR review Toxicological Profile for Alpha-, Beta-, Gamma-, and Delta-Hexachlorocyclohexane had no reports on fetal outcomes following inhalation exposure in humans. One study found that babies born to pregnant Indian women with higher blood levels of HCH isomers, including the gamma isomer, also had higher levels of lindane and were more likely to have intrauterine fetal growth retardation (Siddiqui et al., 2003); however, mothers also had exposure to other organochlorine pesticides.

The Volume 11 committee identified one new study that examined the effects of prenatal exposure to organochlorine pesticides in infants born in an urban area of Northern India (Tyagi et al., 2015). The levels of γ-HCH in maternal blood and placental tissue were measured in 50 women who were at less than 37 weeks of gestation and 50 women who were at 37 weeks or more; γ-HCH residues were detected in the blood of all the women. There was a negative correlation between the blood levels of γ-HCH and placental weight (r= –0.401, p=0.006) but no correlation with the weight of the infant or length of gestation.

Animal Studies

The Volume 2 committee found mixed results from studies of the reproductive effects of lindane in animals. Lindane was found to produce reversible decreases in the total number of sperm in mice (Smith, 1991), testicular degeneration in rats (Chowdhury et al., 1987), and reduced “reproductive behavior” in young rams (Beard et al., 1999). Lindane also had adverse effects on reproduction in female animals. Low doses (10 mg/kg/day for 138 days) of lindane reduced fecundity and litter size in rats (Trifonova et al., 1970); increased the number of resorbed fetuses in hamsters, rabbits, and rats (Dzierzawski, 1977); and increased the duration of diestrus, shortened the duration of estrus, lengthened the gestation period, decreased the number of fetuses, increased the number of dead fetuses, and decreased the growth of the young in rats (Nayshteyn and Leybovich, 1971). However, another study found lindane to have no effect on the numbers of pregnancies, abortions, and dead or resorbed fetuses when given to rats on GDs 6-15 (Khera et al., 1979).

The 2005 ATSDR review Toxicological Profile for Alpha-, Beta-, Gamma-, and Delta-Hexachlorocyclohexane found no studies on reproductive effects in animals following inhalation exposure to any HCH isomer. Anti-estrogenic properties were found in female rats given γ-HCH by the oral route (Chadwick et al., 1988), and female rabbits treated orally with γ-HCH had a reduced ovulation rate (Lindenau et al., 1994). The results of single and multigeneration reproduction studies in rats and mink indicate that exposure to γ-HCH reduces receptivity to mating and reduces whelping rate, decreases numbers of offspring at birth, reduces neonatal viability, and delays the maturation of pups. Those effects were primarily the result of prenatal or postnatal developmental toxicity (Beard and Rawlings, 1998; Beard et al., 1997; King, 1991).

The Volume 11 committee identified five new animal studies on the reproductive effects of lindane. In a two-generation study in rats, lindane administered in the diet to male and females for 10 weeks through mating was found to have no reproductive effects in males; in females, a lack of maternal behavior was seen in the F0 and F1 generations, but no effects were observed on mating, fertility, pregnancy, or parturition in F1 females (Matsuura et al., 2005).

Oral administration of lindane to male rats at 5 mg/kg for 30 days resulted in significantly reduced weights of the reproductive organs and in reductions of testicular DNA, RNA, protein, and enzymes (Sujatha et al., 2001). Damage to Leydig cells was seen in rats treated dermally with a single 1% dose of lindane (the medical dose for scabies) (Suwalsky et al., 2000).

Two in vitro animal studies were considered by the committee. Walsh and Stocco (2000) found that alpha-, delta-, and gamma-HCH inhibited steroidogenesis by reducing StAR protein expression, an action that may contribute to the pathogenesis of lindane-induced reproductive dysfunction. Ronco et al. (2001) also found that lindane produced a dose-dependent inhibition of testosterone production in hCG-stimulated Leydig cells.

Animal Studies

In Volume 2, a number of studies indicated that lindane is not teratogenic. A three-generation study of rats (Palmer et al., 1978) and experiments with mice (Chernoff and Kavlock, 1983; Gray and Kavlock, 1984) did not demonstrate any teratogenic effects.

The 2005 ATSDR review Toxicological Profile for Alpha-, Beta-, Gamma-, and Delta-Hexachlorocyclohexane reported that there were no inhalation data on the developmental effects of lindane in animals. Adverse effects on testicular histology and sperm numbers occurred in the adult male offspring of mice that were exposed to γ-HCH during gestation (Traina et al., 2003).

No adverse prenatal developmental effects of γ-HCH from oral exposure have been found in rats or rabbits (Khera et al., 1979; Palmer et al., 1978; Seiler et al., 1994). Decreases in fetal weight, fetal thymic weight, and placental weight were reported in mice exposed to a single oral dose of γ-HCH on day 12 of gestation (Hassoun and Stohs, 1996). No effects on embryonic development were seen in rabbits treated orally with γ-HCH (Seiler et al., 1994). Due to the lack of developmental toxicity studies in humans as well as the lack of inhalation and dermal data in animals, insufficient information is available to indicate whether HCH affects development via all three routes of exposure. Pharmacokinetic data suggest that HCH isomers might have the potential to affect development across all routes of exposure.

The Volume 11 committee identified three studies of oral exposure to lindane in animals. In a series of experiments, pregnant mice were treated on GDs 9 to 16 with 15 or 25 mg/kg lindane, and male and female offspring were assessed for developmental effects. There were no significant effects on gross malformations in pups, the average number of live pups per litter, the incidence of stillbirths or runts, sex ratio, survival to weaning, or developmental landmarks. Adult F1 males had changes in testicular enzyme activity, germ cell distribution, and chromatic abnormalities of the sperm at PND 60 (i.e., sexual maturity) that were not significant by PND 100 (Di Consiglio et al., 2009; Traina et al., 2003). Female F1 mice showed early increases in uterus weight that were not evident at PND 100; only the diameter of follicular oocytes was significantly reduced in treated females (Maranghi et al., 2007). Preimplantation embryotoxicity with degenerating two-cell embryos was induced by exposure of preovulatory oocytes in mice at the highest lindane dose tested (25 mg/kg); no other signs of embryotoxicity were observed (e.g., apoptosis or micronucleus induction). In one of the few studies to assess preconception exposure only, female mice were given up to 64.4 mg/kg/day lindane in drinking water for 2 weeks prior to mating. There were no cytopathological effects on the male offspring germ cells (Lopez-Casas et al., 2012).

Reproductive Effects

There is little epidemiologic information available on lindane. Both the Volume 2 committee and ATSDR found little information on the effects of lindane exposure on male reproduction in humans and none for lindane's effects on female reproduction. Three studies of endometriosis found that serum levels of β-HCH were associated with an increased risk of endometriosis, but only one of the studies found an increased risk specifically with lindane exposure (Buck Louis et al., 2012). Lindane was also associated with reduced endometrial thickness but not embryo implantation in one study of in vitro fertilization (Al-Hussaini et al., 2018). Thus, all three studies suggest that hexachlorocyclohexane may affect endometrial tissue but the association is stronger for the β-isomer and not the γ-isomer (lindane). Thus, the committee finds that more research on the association between the γ-isomer and female reproductive effects is warranted as lindane is still available to the general public.

The studies of the effects of lindane on birth outcomes are difficult to interpret because lindane was not the only pesticide to which the mothers were exposed. The single new study identified by the Volume 11 committee, Tyagi et al. (2015), found only a negative correlation with lindane and placental weight and no other adverse effects.

In general, the animal studies examined by ATSDR and the Volume 2 and Volume 11 committees found that oral exposure to lindane was associated with numerous reproductive outcomes in both male and female animals, including reduced mating behavior in both sexes and poor fetal outcomes. This body of animal evidence suggests that additional studies of the effects of lindane in humans is warranted to ascertain whether effects seen in animals may be occurring in humans. Furthermore, animals studies with gestational exposures that are equivalent to human organogenesis would be helpful.

The Volume 11 committee concludes that there is inadequate/insufficient evidence to determine whether an association exists between exposure to lindane and reproductive effects in men, or with adverse pregnancy outcomes.

The Volume 11 committee also concludes that there is limited/suggestive evidence of an association between exposure to lindane and reproductive effects in women.

Developmental Effects

There is a lack of information on developmental effects in humans following prenatal exposures to lindane; only one study of women in China found an association—in this case, between the women's lindane exposure and neural tube defects in their children (Ren et al., 2011).

Lindane has not been shown to be teratogenic in animals, and the results of studies to ascertain adverse developmental outcomes following oral exposure prenatally are mixed. Developmental effects were seen in two-generation studies of mice; adult offspring of both sexes displayed altered reproductive parameters at sexual maturity, although the effects appeared to resolve with age. ATSDR (2005) indicated that there was insufficient information to determine if lindane affects development in animals or humans. Although the animal studies indicate that lindane may be both a reproductive and a developmental toxicant, confirmation of the effects in humans is lacking. The available human studies are generally small and may be confounded by exposure to other pesticides or organochlorines; and exposure may have occurred at any point prior to or during pregnancy.

The Volume 11 committee concludes that there is inadequate/insufficient evidence to determine whether an association exists between prenatal exposure to lindane and developmental effects.