Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Maternal Pregnancy Levels of trans-Nonachlor and Oxychlordane and Prevalence of Cryptorchidism and Hypospadias in Boys

Maternal Pregnancy Levels of trans-Nonachlor and Oxychlordane and Prevalence of Cryptorchidism... Research Children’s Health Maternal Pregnancy Levels of trans-Nonachlor and Oxychlordane and Prevalence of Cryptorchidism and Hypospadias in Boys 1 2 3,4,5 6 1 Britton Trabert, Matthew P. Longnecker, John W. Brock, Mark A. Klebanoff, and Katherine A. McGlynn Hormonal and Reproductive Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Rockville, Maryland, USA; Epidemiology Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, North Carolina, USA; National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; 4 5 6 Department of Environmental Studies, and Department of Chemistry, Warren Wilson College, Asheville, North Carolina, USA; Center for Perinatal Research, The Research Institute, Nationwide Children’s Hospital, Columbus, Ohio, USA concomitant genital abnormalities (Toppari Background : The etiologies of the male urogenital anomalies—cryptorchidism and hypospadias— et al. 2010). The etiologic role of lifestyle and are poorly understood. Given positive associations between chlordane isomers and testicular germ environmental exposures such as smoking and cell tumors, it is reasonable to assume that chlordanes might also be associated with other testicular organochlorines has been evaluated in some dysgenesis syndrome disorders, namely cryptorchidism and hypospadias. studies, but few have evaluated the association o B j e c t i v e : To examine whether exposure to in utero chlordane is related to cryptorchidism and between maternal chlordane levels and either hypospadias, we evaluated levels of chlordane derivatives, trans-nonachlor and oxychlordane, among male congenital anomaly. In the only pub- pregnant women enrolled in the Collaborative Perinatal Project (CPP). lished report to date, Damgaaard et al. (2006) Methods : From 1959 to 1965, the CPP enrolled pregnant women at 12 U.S. medical centers. found a positive association between breast We analyzed serum trans-nonachlor and oxychlordane levels measured in third-trimester serum milk chlordane levels and cryptorchidism in from the mothers of 217 sons with cryptorchidism, 197 sons with hypospadias, and 557 sons with sons. The paucity of data on the relation of neither condition. Adjusted odds ratios (ORs) and 95% cond fi ence intervals were calculated using chlordane isomers with male birth defects conditional logistic regression. indicates the need for additional studies. rsultse : The quartile-specific ORs for cryptorchidism or hypospadias show no notable associations In a case–control study nested within a with trans-nonachlor or oxychlordane. Further, there were no significant trends with increasing large prospective cohort of pregnant women, quartile of maternal trans-nonachlor or oxychlordane level in either cryptorchidism or hypospadias we evaluated whether levels of trans-nonachlor (p-trend all > 0.45). and oxychlordane in maternal third-trimester c onclusions : The results do not support an association between chlordane levels and cryptorchid- serum were associated with cryptorchidism or ism or hypospadias. It is unlikely that current chlordane exposure is related to the development of hypospadias among male children. either anomaly, given that serum chlordane levels at the time of sample collection, the early 1960s, were considerably higher than levels at present. Materials and Methods k ey words : chlordanes, cryptorchidism, hypospadias, oxychlordane, pregnant women, prospective Study population. The Collaborative Perinatal cohort, trans-nonachlor. Environ Health Perspect 120:478–482 (2012). http://dx.doi.org/10.1289/ Project (CPP) was a prospective study of neu- ehp.1103936 [Online 7 September 2011] rologic disorders and other conditions in chil- dren (Niswander and Gordon 1972). From Chlordane, a human-made chemical mix- retrospective evaluation in Sweden (Hardell 1959 to 1965, the study enrolled 48,197 ture of structurally similar organochlorines, et al. 2003), and two were prospective evalu- women upon presenting for prenatal care was widely used on agricultural crops and ations, one in the United States and one in at 12 medical centers in the United States. as a termiticide in the United States until Norway (McGlynn et al. 2009b; Purdue et al. Centers were located in Baltimore, Maryland; 1988 (Abadin et al. 1994). Chlordane con- 2009). In animal experiments, chlordane has Boston, Massachusetts; Buffalo, New sists of > 140 isomers; the most abundant been shown to interfere with steroid hormone York; Memphis, Tennessee; Minneapolis, include trans-chlordane, cis-chlordane, trans- metabolism, perhaps because of the induction Minnesota; New Orleans, Louisiana; New nonachlor, beta-chlordane, and heptachlor. In of cytochrome oxidase enzymes (Welch et al. York, New York (two centers); Philadelphia, humans the predominant chlordane-related 1971). Given the associations between chlor- Pennsylvania; Portland, Oregon; Providence, contaminants detected are trans-nonachlor dane isomers and TGCT, it is reasonable to Rhode Island; and Richmond, Virginia. and oxychlordane, a major metabolite of the assume that chlordane levels might also be Eleven centers recruited participants from chlordanes and nonachlors (Dearth and Hites associated with other testicular dysgenesis the prenatal clinics of a university hospital, 1991; Ivie 1973). Chlordanes have a 10- to syndrome disorders, namely cryptorchidism Address correspondence to B. Trabert, Hormonal 20-year half-life in soil, and exposure as a (failure of one or both testicles to descend and Reproductive Epidemiology Branch, Division of result of termite treatment around dwellings into the scrotum) and hypospadias (urethral Cancer Epidemiology and Genetics, National Cancer may have continued for decades after applica- opening on the ventral side of the penis or on Institute, NIH, DHHS, 6120 Executive Blvd., Suite 550, Rockville, MD 20852 USA. Telephone: (301) tion (Bennett et al. 1974). the perineum) (Skakkebaek 2003). 451-4435. Fax: (301) 402-0916. E-mail: britton. Chlordanes, specifically nonachlor deriv - Cryptorchidism and hypospadias are com- trabert@nih.gov atives (trans-nonachlor and cis-nonachlor), mon genital birth defects, affecting approxi - We gratefully acknowledge the contributions to have been among the organochlorines most mately 3–4% and 0.2–1% of male newborns, the project by X. Guo. consistently associated with testicular germ respectively. Increases in both congenital Support for this research was provided by the cell cancer (Cook et al. 2011). Of the four malformations have been reported in several Intramural Research Programs of the National Cancer Institute, the Eunice Kennedy Shriver National published studies evaluating serum chlor- but not all countries (Paulozzi 1999; Toppari Institute of Child Health and Human Development, dane levels and testicular germ cell tumors et al. 2010). Both conditions can be outcomes and the National Institute of Environmental Health (TGCT) (Biggs et al. 2008; Hardell et al. of androgen insufficiency (Toppari et al. Sciences of the National Institutes of Health. 2003; McGlynn et al. 2009b; Purdue 2010). Shared risk factors for cryptorchid- e Th authors declare they have no actual or potential et al. 2009), three reported positive asso- ism and hypospadias include premature competing financial interests. ciations with serum nonachlor. One was a delivery, intrauterine growth restriction, and Received 4 March 2011; accepted 7 September 2011. | | 478 volume 120 number 3 March 2012 • Environmental Health Perspectives Chlordanes and male congenital anomalies and one (Buffalo) recruited from 13 pri- captured in the medical records. The diagnosis the instrument’s LOD yield better estimates vate obstetric practices. The study was not of cryptorchidism (failure of one or both tes- of true concentration than imputed values intended to be representative of the United ticles to descend into the scrotum) was made (Chevrier et al. 2010). Undetected values States, and each clinical site had its own sam- by pediatricians based on serial examinations were set to missing. pling approach (varying from 10 to 100% of that included inspection and palpation of the Statistical analysis. Of the 1,054 total eligible women, either by enrolling a random genitalia. We defined cryptorchidism as hav - subjects (241 cryptorchid, 214 hypospadias, or systematic sample or all women). ing had a diagnosis of undescended testis(es) at 599 controls), 3% were missing data on Women were ineligible if they were incar- any time during the first year of life. Boys with trans-nonachlor. Of those with data on trans- cerated, planned to leave the area after deliv- undescended testis(es) after the first year of life nonachlor, 6% were missing data on other ery, planned to place their child for adoption, were not considered cryptorchid because they covariates, and 5% were also missing data on or if they delivered on the day they were may have had retractile testes. Among the boys oxychlordane. A total of 971 (217 cryptorchid, recruited for the study. Four percent of the with undescended testicle(s) at birth (n = 138), 197 hypospadias, 557 controls) subjects participants were lost to follow-up before all but one had documented orchiopexy or a were included in the analysis of trans-non- delivery. As part of data collection, the moth- subsequent observation of cryptorchidism in achlor, and a total of 919 (206 cryptorchid, ers were asked to donate non fasting blood at least one of the three subsequent examina- 181 hypospadias, 532 controls) were included samples at approximately 8-week intervals tions (ages 4 months, 1 year, 7 years). For the in the analysis of oxychlordane. throughout their pregnancies. Serum samples remaining boys diagnosed as cryptorchid dur- We categorized trans-nonachlor and oxy- were stored in glass vials at –20°C with no ing the first year of life ( n = 103), study records chlordane concentrations according to the recorded thaws. indicated that the testicles were descended at quartile distributions in controls, with the There were 142,130 pregnancies among birth, suggesting that these boys may have lowest quartile serving as the reference cat- the 48,197 women, including 54,390 preg- had acquired, rather than congenital, unde- egory. Odds ratios (ORs) and 95% confi- nancies prospectively (observed) captured by scended testis (Barthold and Gonzalez 2003). dence intervals (CIs) for the association the CPP. The children were systematically To evaluate whether a relationship with oxy- between trans-nonachlor or oxychlordane assessed for the presence of birth defects and chlordane or trans-nonachlor varied by type and cryptorchidism or hypospadias were esti- other outcomes at birth and through 7 years of cryptorchidism, we considered boys with mated using conditional logistic regression of age, with follow-up completed for approxi- testis(es) descended at birth separately in an conditioned on study center (12 strata). We mately 75% of the children born into the additional analysis. assessed a linear trend across quartile catego- study. All mothers provided verbal consent to For analyses in CPP, the socioeconomic ries by including a single independent vari- participate (Hardy 2003). The present study index was calculated as the mean of three per- able taking the value of the corresponding was approved by an institutional review board centile scores: education of head of the house- median of the category. Models were adjusted at the National Institutes of Health. hold, occupation of head of the household or for serum p,p´-DDE as a five-stratum cate- To evaluate our hypothesis, we used data chief wage earner, and family income. The gorical variable, total PCBs as a four-stratum from a nested case–control study of cryp- score used to calculate the percentile for an categorical variable, and serum triglycerides torchidism and hypospadias among sons occupation was based on the percentiles of and cholesterol as continuous variables. We within the CPP. Details of the nested case– education and income among persons with included serum triglycerides and cholesterol control study have been described previously the same occupation (Myrianthopoulos and as independent variables in all statistical analy- (Longnecker et al. 2002). Inclusion criteria French 1968). ses to account for interindividual variations were based on the characteristics of the mother Laboratory assays. Serum levels of chlor- in serum lipid concentration. The results of and infant. The sole maternal inclusion crite - dane (trans-nonachlor and oxychlordane) were analyses that modeled lipid-adjusted trans- rion was the availability of a 3-mL aliquot of measured at the Centers for Disease Control nonachlor and oxychlordane produced results third-trimester serum. Inclusion criteria based and Prevention (CDC) after solid-phase extrac- similar to those presented. We included serum on the characteristics of the infant included tion cleanup and dual-column gas chroma- lipids as a covariate in our model rather than alive at birth, male sex, and singleton birth. tography using electron capture (Brock et al. using lipid-standardized chlordane concen- Of the 28,444 boys born to mothers enrolled 1996). Serum levels of p,p´-dichlorodiphenyl - trations because the latter may be prone to prospectively in the CPP cohort, 6,097 were dichloroethylene (p,p´-DDE) and 11 poly- bias, depending on the underlying mecha- not eligible for inclusion in the current analy- chlorinated biphenyls (PCBs) were measured in nism of the chlordane–lipid disease associa- sis because there was no maternal 3-mL ali- the same laboratory, and the laboratory meth- tion (Gaskins and Schisterman 2009). quot of third-trimester serum available (n = ods have been described previously (Longnecker All models included trans-nonachlor 5,389), the son was not a singleton (n = 441), et al. 2002; McGlynn et al. 2009a). Serum cho- or oxychlordane as the main exposure and or the son was not live born (n = 267). Among lesterol and triglycerides were measured using were adjusted for total PCBs, p,p´-DDE, the eligible 22,347 boys, there were 241 cases standard enzymatic assays. triglycerides, and cholesterol as a priori of cryptorchidism and 214 cases of hypospa- The between-assay coefficient of varia- selected variables. Additional variables were dias. Five boys had both cryptorchidism and tion was 25% for trans-nonachlor and assessed as potential confounders using the hypospadias and were included in each group 20% for oxychlordane. These were deter- change in estimate method (Maldonado and for analysis. For comparison, we randomly mined at concentrations of 0.52 and 0.57 Greenland 1993), starting with all variables selected a group of boys (n = 599) from the μg/L (301 batches), respectively. Limits of in the models with deletion of one by one remaining eligible boys without a diagnosis of detection (LODs) were 0.28 μg/L for trans- in a stepwise manner. If, on deletion, the cryptorchidism and/or hypospadias. Controls nonachlor and 0.20 μg/L for oxychlordane ; OR for the contrast of the highest-to-lowest were selected so that the control:case ratio 29% of the values for trans-nonachlor were chlordane strata or the OR from the trend would be > 2:1 for each case group. below that value, whereas 31% of the values test changed by ≥ 15%, the factor was con- Hypospadias (urethral opening on the for oxychlordane were below the LOD. For sidered a confounder and was included in ventral side of the penis) was defined as having values below the LOD, we used the signal the adjusted analyses. Potential confounding a diagnosis any time within the first 7 years recorded by the instrument, when available, factors included race, maternal age, mater- of life. The degree of hypospadias was not because it is thought that the signals below nal history of previous live birth, season of | | Environmental Health Perspectives • volume 120 number 3 March 2012 479 Trabert et al. birth, socioeconomic index, smoking dur- index was the only variable that changed the triglycerides, cholesterol, serum p,p´-DDE, ing pregnancy, and gestational hypertension, OR by ≥ 15%. We also considered the effects total PCBs, gestational hypertension, and as defined in Table 1. Additional potential of adjustment for the continuous variables socioeconomic index, using the cross-prod- confounding factors included the categorical birth weight and placental weight and the uct terms. Variables were coded as defined in variables (yes/no) hyperemesis gravidarum, categorical variables preterm birth and small- Table 1, with the exception of prepregnancy history of infertility, menstrual cycle irregu- for-gestational-age, even though these were BMI, which was coded as < 25, 25–29.9, larity, estrogen use during pregnancy, and potentially intermediate variables. ≥ 30 kg/m . We supplemented evaluation of progesterone use during pregnancy, as well as We evaluated effect modification by effect modification by categorical variables the continuous variables age at menarche and maternal age, race, smoking, prepregnancy with more than two categories by compar- weight gain during pregnancy. Socioeconomic body mass index (BMI), previous live births, ing the model fit statistics for models with and without the cross-product terms. If the Table 1. Characteristics of mothers and sons according to case–control status of the son, CPP, 1959–1965. p-value associated with the interaction term based on the likelihood ratio test had a value Characteristic Cryptorchidism (n = 217) Hypospadias (n = 197) Control (n = 557) < 0.10, the degree of potential effect modi - Race (%) fication was further evaluated by examining White 57.6 50.8 46.5 tables stratified by the potentially modify- Black 41.5 43.7 47.8 Other 0.9 5.6 5.7 ing factor(s). Statistical significance was set at Gestation (week) p < 0.05 for main effects based on two-sided Median (IQR) 39 (37–40) 39 (37–41) 39 (38–40) tests. Statistical analyses were conducted using Preterm birth (%) 19.3 20.5 13.8 SAS statistical software package, version 9.2 Birth weight (g) (SAS Institute Inc., Cary, NC, USA). Median (IQR) 3,232 (2,863–3,629) 3,147 (2,722–3,487) 3,260 (2,948–3,600) Small for gestational age (%) 10.7 19.9 5.0 Results Maternal age (years) The distribution of selected demographic Median (IQR) 24 (21–30) 23 (20–29) 22 (20–28) Previous live births (%) and health characteristics of the mothers and 0 26.3 31.0 30.3 sons stratified by study group are provided 1 23.0 21.8 22.8 in Table 1. For both cryptorchid and hypos- ≥ 2 50.7 47.2 46.9 padias case groups, a higher percentage of Season of birth (%) case boys than control boys were white, born January–March 24.4 17.3 24.6 preterm, and small-for-gestational-age (birth April–June 28.1 25.4 24.2 weight < 10th percentile). The distribution of July–September 24.9 29.9 25.9 October–December 22.6 27.4 25.3 other potential confounding factors according Socioeconomic index to case–control status has been published pre- Median (IQR) 4.7 (3.3–6.3) 4.7 (3.0–6.3) 4.7 (3.3–6.0) viously (Longnecker et al. 2002). Prepregnancy BMI The median maternal serum concentra- Median (IQR) 22.2 (20.3–25.0) 21.8 (19.6–24.2) 22.2 (20.0–24.9) tions of trans-nonachlor, oxychlordane, DDE, Gestational hypertension (%) 5.9 6.3 6.2 total PCBs, cholesterol, and triglycerides Maternal smoking status (%) for each study group are shown in Table 2. None 50.7 52.0 55.0 The median level of maternal serum trig- 1–10 cigarettes per day 32.3 27.0 26.7 > 10 cigarettes per day 17.0 20.9 18.4 lycerides was slightly lower for hypospadias Study center (%) cases (188 μg/L) compared with cryptorchid Boston, MA 32.7 27.4 24.6 cases (204 μg/L) or controls (204 μg/L). The Buffalo, NY 7.4 4.1 3.8 Spearman correlation coefficients of trans- New Orleans, LA 6.5 4.1 4.7 nonachlor, oxychlordane, DDE, total PCBs, New York City, NY 2.3 3.6 3.1 cholesterol, and triglycerides are provided in Baltimore, MD 5.5 6.1 7.9 Table 3. All of the exposures evaluated were Richmond, VA 7.4 6.1 6.1 Minneapolis, MN 3.7 4.1 5.4 positively correlated, with the strongest corre- New York City, NY 1.4 5.1 7.5 lation between trans-nonachlor and oxychlor- Portland, OR 4.1 5.6 7.2 dane (r = 0.78), as expected. trans-Nonachlor Philadelphia, PA 15.2 23.9 17.6 was moderately associated with DDE Providence, RI 10.6 8.1 5.2 (r = 0.52), while the remaining compounds Memphis, TN 3.2 2.0 7.0 were less correlated (r = 0.08–0.46). IQR, interquartile range (quartiles 1–3). Table 4 provides the adjusted ORs for a b Columbia-Presbyterian Medical Center. New York Medical College. trans-nonachlor and oxychlordane quartiles conditioned on study center and adjusted Table 2. Maternal serum values by son’s case–control status, CPP, 1959–1965 [median (IQR)]. for serum DDE concentration, total PCBs, Characteristic Cryptorchidism (n = 217) Hypospadias (n = 197) Control (n = 557) triglycerides, cholesterol, and socioeconomic trans-Nonachlor (μg/L) 0.36 (0.26–0.52) 0.40 (0.24–0.58) 0.38 (0.25–0.57) index. For cryptorchidism, the quartile- Oxychlordane (μg/L) 0.29 (0.14–0.47) 0.34 (0.18–0.58) 0.31 (0.16–0.55) specific ORs for oxychlordane or trans-non- DDE (μg/L) 23.6 (15.9–35.3) 23.8 (16.6–34.4) 24.5 (16.7–37.5) achlor were close to the null as well as not Total PCBs with imputed congener (μg/L) 2.8 (2.0–3.9) 2.9 (2.1–4.2) 2.7 (1.8–3.9) statistically significant. Similarly for hypos - Total cholesterol (μg/L) 232 (190–273) 232 (192–277) 234 (197–280) padias, the quartile-specific ORs for trans- Triglycerides (μg/L) 204 (161–252) 188 (156–256) 204 (159–259) nonachlor and oxychlordane were close to IQR, interquartile range (quartiles 1–3). Eleven cases of cryptorchidism, 16 cases of hypospadias, and 25 controls were missing data for oxychlordane, respectively. the null as well as not statistically significant. | | 480 volume 120 number 3 March 2012 • Environmental Health Perspectives Chlordanes and male congenital anomalies We found no significant increase or decrease from the analysis that included all boys with lipid in controls) (Damgaard et al. 2006). The in risk of either cryptorchidism or hypos- cryptorchidism (results not shown). investigators reported no relationship between padias with increasing quartile of maternal oxychlordane and cryptorchidism and a Discussion trans- nonachlor or oxychlordane level (trend border line significant association between p-values all > 0.40). In the present study, a prospective evalua- cis-chlordane and cryptorchidism. However, Potential effect modification at a p-value tion of chlordane levels during pregnancy and the percent detection for cis-chlordane was of < 0.10 was present for the cryptorchid- cryptorchidism or hypospadias, the results do only 30.8%, so median concentrations were ism–oxychlordane association by smoking, not support an association between chlordane not reported (Damgaard et al. 2006). Our the hypospadias–oxychlordane association by levels and cryptorchidism or hypospadias. finding of no association between oxychlor - BMI, and the hypospadias–trans-nonachlor Further, the lack of association was consis- dane level and cryptorchidism is consistent association by smoking. After further evalu- tent in the subgroup of primiparous women. with that of Damgaard and colleagues. We ation in stratified models, there were some The median serum concentration of trans- did not evaluate trans- or cis-chlordane; how- differences in ORs by smoking or BMI ; how- nonachlor in the United States was 14.8 ng/g ever, we report no association between trans- ever, all estimates were consistent with no lipid in 2003–2004, the most recent years for nonachlor, another chlordane derivative, and association. Given that the main effects were which data are available; the median of oxy- cryptorchidism. not indicative of an association, the results chlordane was 10.3 ng/g lipid (CDC 2011). Teratogenic effects of chlordane expo- from the stratified models are not presented. The median serum concentrations of trans- sure have not been observed in animal studies In subanalyses, we evaluated the associa- nonachlor and oxychlordane were 48.0 ng/g (Abadin et al. 1994). However, chlordane tions between the chlordanes and cryptorchid- lipid and 37.5 ng/g lipid, respectively, in the exposures has been shown to affect repro- ism and hypospadias among women with no present study samples that were collected in duction in test animals, delaying puberty, history of a live birth. We hypothesized that the 1960s. This suggests that current low-level disrupting estrous cycling in females, and the trans-nonachlor and oxychlordane expo- chlordane exposure is unlikely to be related to reducing fertility by as much as 50% (Welch sure to the male fetus would be highest in the development of either condition. et al. 1971). Furthermore, chlordane is clas- this group of women. Although the quartile- To our knowledge, this is the first epi- sified as potentially carcinogenic to humans specific ORs were increased slightly compared demiologic evaluation of the association (Group 2B) by the International Agency for with the corresponding ORs for all women between chlordane levels and hypospadias Research on Cancer (IARC 2001) and has (data not shown), there were no significant and only the second such study of the asso- been associated with risk of TGCT (Hardell trends with increasing quartile of mater- ciation between chlordane levels and cryp- et al. 2003; McGlynn et al. 2009b; Purdue nal trans-nonachlor or oxychlordane level torchidism. Using breast milk as a proxy for et al. 2009). Although chlordane may be a for either cryptorchidism (trans-nonachlor: maternal exposure, Damgaard et al. (2006) carcinogen, both animal data and the present p-trend = 0.75; oxychlordane: p-trend = 0.97) evaluated persistent pesticide concentrations study suggest it is not a teratogen. or hypospadias (trans-nonachlor: p-trend = in milk samples collected from mothers of The present study has several strengths. 0.68; oxychlordane: p-trend = 0.90). Finally, cryptorchid boys (n = 62) and healthy boys (n The major advantages were that the study was after exclusion of the boys (n = 103) whose = 68). Maternal milk levels of trans-chlordane a large and prospective evaluation of chlor- testicles were initially descended at birth, were slightly higher in mothers of cryptorchid dane exposure and congenital malformations. the association between chlordane level and boys than mothers of healthy boys; however, This study has several potential weaknesses cryptorchidism was not substantially different the overall exposure was very low (0.04 ng/g that merit consideration. Chlordane levels were measured on samples stored for approxi- Table 3. Spearman correlation coefficients of selected covariates, CPP, 1959–1965. mately 40 years. However, prior studies of freeze–thaw cycles have demonstrated that trans-Nonachlor Oxychlordane DDE Total PCBs Total cholesterol Triglycerides chlordane levels are quite stable over time trans-Nonachlor (μg/L) 1.00 (Becker et al. 1997). In addition, cholesterol Oxychlordane (μg/L) 0.78 1.00 and triglyceride levels were in the expected DDE (μg/L) 0.52 0.38 1.00 Total PCBs (μg/L) 0.38 0.46 0.37 1.00 range, suggesting that substantial degrada- Total cholesterol (μg/L) 0.10 0.12 0.15 0.25 1.00 tion had not occurred. The chlordane con- Triglycerides (μg/L) 0.11 0.12 0.08 0.15 0.36 1.00 centrations for this study were measured in Table 4. Adjusted ORs and 95% CIs for cryptorchidism and hypospadias in relation to maternal oxychlordane and trans-nonachlor levels (μg/L), CPP, 1959–1965. Cryptorchidism Hypospadias a b b Chlordane (quartiles) Control (n) Case (n) Unadjusted OR Adjusted OR (95% CI) Case (n) Unadjusted OR Adjusted OR (95% CI) trans-Nonachlor (µg/L) 0.00 to < 0.25 134 47 1.00 (reference) 1.00 (reference) 50 1.00 (reference) 1.00 (reference) 0.25 to < 0.38 139 67 1.25 1.27 (0.85, 1.90) 44 0.83 0.84 (0.54, 1.33) 0.38 to < 0.57 142 62 1.33 1.46 (0.93, 2.30) 52 1.00 1.04 (0.64, 1.70) ≥ 0.57 142 41 1.10 1.22 (0.70, 2.12) 51 0.99 1.08 (0.62, 1.89) Trend test 557 217 0.55 197 0.60 Oxychlordane (µg/L) 0.00 to < 0.16 131 54 1.00 (reference) 1.00 (reference) 38 1.00 (reference) 1.00 (reference) 0.16 to < 0.31 131 56 0.95 0.90 (0.60, 1.35) 47 1.12 1.06 (0.66, 1.69) 0.31 to < 0.55 137 51 0.90 0.91 (0.59, 1.42) 44 1.03 1.07 (0.64, 1.80) ≥ 0.55 133 45 0.94 0.95 (0.55, 1.64) 52 1.15 1.24 (0.69, 2.22) Trend test 532 206 0.90 181 0.46 a b Quartile cut points were defined based on 557 and 532 controls for trans-nonachlor and oxychlordane, respectively. Adjusted ORs and 95% CIs are from conditional logistic regression models adjusted for serum DDE concentration (five categories), total PCBs (four categories), triglycerides, cholesterol, and socioeconomic index. Ordinal test across four categories using the median value within each group. | | Environmental Health Perspectives • volume 120 number 3 March 2012 481 Trabert et al. third-trimester samples; the critical window Refe Rences chlordanes in mothers of men with testicular cancer. Environ Health Perspect 111:930–934. of exposure might be earlier for hypospadias, Hardy JB. 2003. The Collaborative Perinatal Project: lessons Abadin HG, Baynes R, Goetchius PF. 1994. Toxicological Profile likely mid-to-late first trimester (Husmann and legacy. Ann Epidemiol 13:303–311. for Chlordane. Atlanta, GA:U.S. Department of Health 2002). However, third-trimester assays should Husmann DA. 2002. Micropenis: an animal model and its human and Human Services, Agency for Toxic Substances and correlates. Adv Exp Med Biol 511: 41–54. reflect first-trimester exposure, given the Disease Registry (ATSDR). IARC (International Agency for Research on Cancer). 2001. Barthold JS, Gonzalez R. 2003. The epidemiology of congenital long biological half-life of chlordane (Abadin Some Thyrotropic Agents. IARC Monogr Eval Carcinog cryptorchidism, testicular ascent and orchiopexy. J Urol et al. 1994). In the CPP, the prevalence of Risk Hum 79. 170:2396–2401. Ivie GW. 1973. Nature and toxicity of two oxychlordane photoi- cryptorchidism and hypospadias among Becker PR, Mackey EA, Demiralp R, Schantz MM, Koster BJ, somers. J Agric Food Chem 21:1113–1115. Wise SA. 1997. Concentrations of chlorinated hydro- males within the first year of life was 56 Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, carbons and trace elements in marine mammal tissues per 10,000 and 41 per 10,000, respectively Needham LL, et al. 2002. Maternal serum level of 1,1-di- archived in the U.S. National Biomonitoring Specimen chloro-2,2-bis(p-chlorophenyl)ethylene and risk of cryp- (Myrianthopoulos and Chung 1974). These Bank. Chemosphere 34:2067–2098. torchidism, hypospadias, and polythelia among male Bennett GW, Ballee DL, Hall RC, Fahey JE, Butts WL, Osmun JV. prevalences were higher than in other U.S. offspring. Am J Epidemiol 155:313–322. 1974. Persistence and distribution of chlordane and diel- populations (cryptorchidism: 19 per 10,000; Maldonado G, Greenland S. 1993. Simulation study of con- drin applied as termiticides. Bull Environ Contam Toxicol hypospadias: 21 per 10,000) (Paulozzi 1999). founder-selection strategies. Am J Epidemiol 138:923–936. 11:64–69. McGlynn KA, Guo X, Graubard BI, Brock JW, Klebanoff MA, Biggs ML, Davis MD, Eaton DL, Weiss NS, Barr DB, Doody DR, This difference was likely the result of reduced Longnecker MP. 2009a. Maternal pregnancy levels of et al. 2008. Serum organochlorine pesticide residues and misclassification in CPP, given that congeni - polychlorinated biphenyls and risk of hypospadias and risk of testicular germ cell carcinoma: a population-based tal anomalies were identified in a series of cryptorchidism in male offspring. Environ Health Perspect case-control study. Cancer Epidemiol Biomarkers Prev 117:1472–1476. systematic examinations, whereas birth defect 17:2012–2018. M c G l y n n K A , Q u r a i s h i S M , G r a u b a r d B I , W e b e r J P , Brock JW, Burse VW, Ashley DL, Najam AR, Green VE, registries are typically based on routine reports Rubertone MV, Erickson RL. 2009b. Polychlorinated biphe- Korver MP, et al. 1996. An improved analysis for chlori- and records only. The ratio of hypospadias to nyls and risk of testicular germ cell tumors. Cancer Res nated pesticides and polychlorinated biphenyls (PCBs) 69:1901–1909. cryptorchidism was somewhat higher than in human and bovine sera using solid-phase extraction. Myrianthopoulos NC, Chung CS. 1974. Congenital malforma- J Anal Toxicol 20:528–536. what might be expected. We included hypo- tions in singletons: epidemiologic survey. Report from the CDC (Centers for Disease Control and Prevention). 2011. Fourth spadias diagnosed up until 7 years of age, Collaborative Perinatal project. Birth Defects Orig Artic National Report on Human Exposure to Environmental Ser 10:1–58. whereas we included cryptorchidism cases Chemicals. NCEH Pub No. 01-0164:1-6. Atlanta, GA:Centers Myrianthopoulos NC, French KS. 1968. An application of the for Disease Control and Prevention. diagnosed through the first year of life only. U.S. Bureau of the Census socioeconomic index to a large, Chevrier J, Harley KG, Bradman A, Gharbi M, Sjodin A, Further, the medical records did not denote diversified patient population. Soc Sci Med 2:283–299. Eskenazi B. 2010. Polybrominated diphenyl ether (PBDE) Niswander KR, Gordon M. 1972. The Women and Their the degree of hypospadias, which limited our flame retardants and thyroid hormone during pregnancy. Pregnancies. Philadelphia:WB Saunders. Environ Health Perspect 118:1444–1449. ability to evaluate the relationship of exposure Paulozzi LJ. 1999. International trends in rates of hypospadias Cook MB, Trabert B, McGlynn KA. 2011. Organochlorine com- and disease severity. and cryptorchidism. Environ Health Perspect 107:297–302. pounds and testicular dysgenesis syndrome: human data. In conclusion, these results provide no Purdue MP, Engel LS, Langseth H, Needham LL, Andersen A, Int J Androl 34(4 pt 2):e68–e84. Barr DB, et al. 2009. Prediagnostic serum concentrations Damgaard IN, Skakkebaek NE, Toppari J, Virtanen HE, Shen H, clear evidence of an effect of in utero chlor- of organochlorine compounds and risk of testicular germ Schramm KW, et al. 2006. Persistent pesticides in human dane (trans-nonachlor or oxychlordane) levels cell tumors. Environ Health Perspect 117:1514–1519. breast milk and cryptorchidism. Environ Health Perspect on cryptorchidism or hypospadias, and the Skakkebaek NE. 2003. Testicular dysgenesis syndrome. Horm 114:1133–1138. Res 60(suppl 3):49. findings are in line with results from animal Dearth MA, Hites RA. 1991. Chlordane accumulation in people. Toppari J, Virtanen HE, Main KM, Skakkebaek NE. 2010. Environ Sci Technol 25:1279–1285. studies. Further, it is unlikely that current Cryptorchidism and hypospadias as a sign of testicular Gaskins AJ, Schisterman EF. 2009. The effect of lipid adjust- chlordane exposure is related to the develop- dysgenesis syndrome (TDS): environmental connection. ment on the analysis of environmental contaminants and Birth Defects Res A Clin Mol Teratol 88:910–919. ment of either anomaly, because serum chlor- the outcome of human health risks. Methods Mol Biol Welch RM, Levin W, Kuntzman R, Jacobson M, Conney AH. 580:371–381. dane levels at the time of sample collection, 1971. Effect of halogenated hydrocarbon insecticides on Hardell L, Van BB, Lindstrom G, Carlberg M, Dreifaldt AC, the early 1960s, were considerably higher than the metabolism and uterotropic action of estrogens in rats Wijkstrom H, et al. 2003. Increased concentrations levels at present. and mice. Toxicol Appl Pharmacol 19:234–246. of polychlorinated biphenyls, hexachlorobenzene, and | | 482 volume 120 number 3 March 2012 • Environmental Health Perspectives http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Environmental Health Perspectives Pubmed Central

Maternal Pregnancy Levels of trans-Nonachlor and Oxychlordane and Prevalence of Cryptorchidism and Hypospadias in Boys

Environmental Health Perspectives , Volume 120 (3) – Sep 7, 2011

Loading next page...
 
/lp/pubmed-central/maternal-pregnancy-levels-of-trans-nonachlor-and-oxychlordane-and-78fr4uo8Yh

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
ISSN
0091-6765
eISSN
1552-9924
DOI
10.1289/ehp.1103936
Publisher site
See Article on Publisher Site

Abstract

Research Children’s Health Maternal Pregnancy Levels of trans-Nonachlor and Oxychlordane and Prevalence of Cryptorchidism and Hypospadias in Boys 1 2 3,4,5 6 1 Britton Trabert, Matthew P. Longnecker, John W. Brock, Mark A. Klebanoff, and Katherine A. McGlynn Hormonal and Reproductive Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Rockville, Maryland, USA; Epidemiology Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, North Carolina, USA; National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; 4 5 6 Department of Environmental Studies, and Department of Chemistry, Warren Wilson College, Asheville, North Carolina, USA; Center for Perinatal Research, The Research Institute, Nationwide Children’s Hospital, Columbus, Ohio, USA concomitant genital abnormalities (Toppari Background : The etiologies of the male urogenital anomalies—cryptorchidism and hypospadias— et al. 2010). The etiologic role of lifestyle and are poorly understood. Given positive associations between chlordane isomers and testicular germ environmental exposures such as smoking and cell tumors, it is reasonable to assume that chlordanes might also be associated with other testicular organochlorines has been evaluated in some dysgenesis syndrome disorders, namely cryptorchidism and hypospadias. studies, but few have evaluated the association o B j e c t i v e : To examine whether exposure to in utero chlordane is related to cryptorchidism and between maternal chlordane levels and either hypospadias, we evaluated levels of chlordane derivatives, trans-nonachlor and oxychlordane, among male congenital anomaly. In the only pub- pregnant women enrolled in the Collaborative Perinatal Project (CPP). lished report to date, Damgaaard et al. (2006) Methods : From 1959 to 1965, the CPP enrolled pregnant women at 12 U.S. medical centers. found a positive association between breast We analyzed serum trans-nonachlor and oxychlordane levels measured in third-trimester serum milk chlordane levels and cryptorchidism in from the mothers of 217 sons with cryptorchidism, 197 sons with hypospadias, and 557 sons with sons. The paucity of data on the relation of neither condition. Adjusted odds ratios (ORs) and 95% cond fi ence intervals were calculated using chlordane isomers with male birth defects conditional logistic regression. indicates the need for additional studies. rsultse : The quartile-specific ORs for cryptorchidism or hypospadias show no notable associations In a case–control study nested within a with trans-nonachlor or oxychlordane. Further, there were no significant trends with increasing large prospective cohort of pregnant women, quartile of maternal trans-nonachlor or oxychlordane level in either cryptorchidism or hypospadias we evaluated whether levels of trans-nonachlor (p-trend all > 0.45). and oxychlordane in maternal third-trimester c onclusions : The results do not support an association between chlordane levels and cryptorchid- serum were associated with cryptorchidism or ism or hypospadias. It is unlikely that current chlordane exposure is related to the development of hypospadias among male children. either anomaly, given that serum chlordane levels at the time of sample collection, the early 1960s, were considerably higher than levels at present. Materials and Methods k ey words : chlordanes, cryptorchidism, hypospadias, oxychlordane, pregnant women, prospective Study population. The Collaborative Perinatal cohort, trans-nonachlor. Environ Health Perspect 120:478–482 (2012). http://dx.doi.org/10.1289/ Project (CPP) was a prospective study of neu- ehp.1103936 [Online 7 September 2011] rologic disorders and other conditions in chil- dren (Niswander and Gordon 1972). From Chlordane, a human-made chemical mix- retrospective evaluation in Sweden (Hardell 1959 to 1965, the study enrolled 48,197 ture of structurally similar organochlorines, et al. 2003), and two were prospective evalu- women upon presenting for prenatal care was widely used on agricultural crops and ations, one in the United States and one in at 12 medical centers in the United States. as a termiticide in the United States until Norway (McGlynn et al. 2009b; Purdue et al. Centers were located in Baltimore, Maryland; 1988 (Abadin et al. 1994). Chlordane con- 2009). In animal experiments, chlordane has Boston, Massachusetts; Buffalo, New sists of > 140 isomers; the most abundant been shown to interfere with steroid hormone York; Memphis, Tennessee; Minneapolis, include trans-chlordane, cis-chlordane, trans- metabolism, perhaps because of the induction Minnesota; New Orleans, Louisiana; New nonachlor, beta-chlordane, and heptachlor. In of cytochrome oxidase enzymes (Welch et al. York, New York (two centers); Philadelphia, humans the predominant chlordane-related 1971). Given the associations between chlor- Pennsylvania; Portland, Oregon; Providence, contaminants detected are trans-nonachlor dane isomers and TGCT, it is reasonable to Rhode Island; and Richmond, Virginia. and oxychlordane, a major metabolite of the assume that chlordane levels might also be Eleven centers recruited participants from chlordanes and nonachlors (Dearth and Hites associated with other testicular dysgenesis the prenatal clinics of a university hospital, 1991; Ivie 1973). Chlordanes have a 10- to syndrome disorders, namely cryptorchidism Address correspondence to B. Trabert, Hormonal 20-year half-life in soil, and exposure as a (failure of one or both testicles to descend and Reproductive Epidemiology Branch, Division of result of termite treatment around dwellings into the scrotum) and hypospadias (urethral Cancer Epidemiology and Genetics, National Cancer may have continued for decades after applica- opening on the ventral side of the penis or on Institute, NIH, DHHS, 6120 Executive Blvd., Suite 550, Rockville, MD 20852 USA. Telephone: (301) tion (Bennett et al. 1974). the perineum) (Skakkebaek 2003). 451-4435. Fax: (301) 402-0916. E-mail: britton. Chlordanes, specifically nonachlor deriv - Cryptorchidism and hypospadias are com- trabert@nih.gov atives (trans-nonachlor and cis-nonachlor), mon genital birth defects, affecting approxi - We gratefully acknowledge the contributions to have been among the organochlorines most mately 3–4% and 0.2–1% of male newborns, the project by X. Guo. consistently associated with testicular germ respectively. Increases in both congenital Support for this research was provided by the cell cancer (Cook et al. 2011). Of the four malformations have been reported in several Intramural Research Programs of the National Cancer Institute, the Eunice Kennedy Shriver National published studies evaluating serum chlor- but not all countries (Paulozzi 1999; Toppari Institute of Child Health and Human Development, dane levels and testicular germ cell tumors et al. 2010). Both conditions can be outcomes and the National Institute of Environmental Health (TGCT) (Biggs et al. 2008; Hardell et al. of androgen insufficiency (Toppari et al. Sciences of the National Institutes of Health. 2003; McGlynn et al. 2009b; Purdue 2010). Shared risk factors for cryptorchid- e Th authors declare they have no actual or potential et al. 2009), three reported positive asso- ism and hypospadias include premature competing financial interests. ciations with serum nonachlor. One was a delivery, intrauterine growth restriction, and Received 4 March 2011; accepted 7 September 2011. | | 478 volume 120 number 3 March 2012 • Environmental Health Perspectives Chlordanes and male congenital anomalies and one (Buffalo) recruited from 13 pri- captured in the medical records. The diagnosis the instrument’s LOD yield better estimates vate obstetric practices. The study was not of cryptorchidism (failure of one or both tes- of true concentration than imputed values intended to be representative of the United ticles to descend into the scrotum) was made (Chevrier et al. 2010). Undetected values States, and each clinical site had its own sam- by pediatricians based on serial examinations were set to missing. pling approach (varying from 10 to 100% of that included inspection and palpation of the Statistical analysis. Of the 1,054 total eligible women, either by enrolling a random genitalia. We defined cryptorchidism as hav - subjects (241 cryptorchid, 214 hypospadias, or systematic sample or all women). ing had a diagnosis of undescended testis(es) at 599 controls), 3% were missing data on Women were ineligible if they were incar- any time during the first year of life. Boys with trans-nonachlor. Of those with data on trans- cerated, planned to leave the area after deliv- undescended testis(es) after the first year of life nonachlor, 6% were missing data on other ery, planned to place their child for adoption, were not considered cryptorchid because they covariates, and 5% were also missing data on or if they delivered on the day they were may have had retractile testes. Among the boys oxychlordane. A total of 971 (217 cryptorchid, recruited for the study. Four percent of the with undescended testicle(s) at birth (n = 138), 197 hypospadias, 557 controls) subjects participants were lost to follow-up before all but one had documented orchiopexy or a were included in the analysis of trans-non- delivery. As part of data collection, the moth- subsequent observation of cryptorchidism in achlor, and a total of 919 (206 cryptorchid, ers were asked to donate non fasting blood at least one of the three subsequent examina- 181 hypospadias, 532 controls) were included samples at approximately 8-week intervals tions (ages 4 months, 1 year, 7 years). For the in the analysis of oxychlordane. throughout their pregnancies. Serum samples remaining boys diagnosed as cryptorchid dur- We categorized trans-nonachlor and oxy- were stored in glass vials at –20°C with no ing the first year of life ( n = 103), study records chlordane concentrations according to the recorded thaws. indicated that the testicles were descended at quartile distributions in controls, with the There were 142,130 pregnancies among birth, suggesting that these boys may have lowest quartile serving as the reference cat- the 48,197 women, including 54,390 preg- had acquired, rather than congenital, unde- egory. Odds ratios (ORs) and 95% confi- nancies prospectively (observed) captured by scended testis (Barthold and Gonzalez 2003). dence intervals (CIs) for the association the CPP. The children were systematically To evaluate whether a relationship with oxy- between trans-nonachlor or oxychlordane assessed for the presence of birth defects and chlordane or trans-nonachlor varied by type and cryptorchidism or hypospadias were esti- other outcomes at birth and through 7 years of cryptorchidism, we considered boys with mated using conditional logistic regression of age, with follow-up completed for approxi- testis(es) descended at birth separately in an conditioned on study center (12 strata). We mately 75% of the children born into the additional analysis. assessed a linear trend across quartile catego- study. All mothers provided verbal consent to For analyses in CPP, the socioeconomic ries by including a single independent vari- participate (Hardy 2003). The present study index was calculated as the mean of three per- able taking the value of the corresponding was approved by an institutional review board centile scores: education of head of the house- median of the category. Models were adjusted at the National Institutes of Health. hold, occupation of head of the household or for serum p,p´-DDE as a five-stratum cate- To evaluate our hypothesis, we used data chief wage earner, and family income. The gorical variable, total PCBs as a four-stratum from a nested case–control study of cryp- score used to calculate the percentile for an categorical variable, and serum triglycerides torchidism and hypospadias among sons occupation was based on the percentiles of and cholesterol as continuous variables. We within the CPP. Details of the nested case– education and income among persons with included serum triglycerides and cholesterol control study have been described previously the same occupation (Myrianthopoulos and as independent variables in all statistical analy- (Longnecker et al. 2002). Inclusion criteria French 1968). ses to account for interindividual variations were based on the characteristics of the mother Laboratory assays. Serum levels of chlor- in serum lipid concentration. The results of and infant. The sole maternal inclusion crite - dane (trans-nonachlor and oxychlordane) were analyses that modeled lipid-adjusted trans- rion was the availability of a 3-mL aliquot of measured at the Centers for Disease Control nonachlor and oxychlordane produced results third-trimester serum. Inclusion criteria based and Prevention (CDC) after solid-phase extrac- similar to those presented. We included serum on the characteristics of the infant included tion cleanup and dual-column gas chroma- lipids as a covariate in our model rather than alive at birth, male sex, and singleton birth. tography using electron capture (Brock et al. using lipid-standardized chlordane concen- Of the 28,444 boys born to mothers enrolled 1996). Serum levels of p,p´-dichlorodiphenyl - trations because the latter may be prone to prospectively in the CPP cohort, 6,097 were dichloroethylene (p,p´-DDE) and 11 poly- bias, depending on the underlying mecha- not eligible for inclusion in the current analy- chlorinated biphenyls (PCBs) were measured in nism of the chlordane–lipid disease associa- sis because there was no maternal 3-mL ali- the same laboratory, and the laboratory meth- tion (Gaskins and Schisterman 2009). quot of third-trimester serum available (n = ods have been described previously (Longnecker All models included trans-nonachlor 5,389), the son was not a singleton (n = 441), et al. 2002; McGlynn et al. 2009a). Serum cho- or oxychlordane as the main exposure and or the son was not live born (n = 267). Among lesterol and triglycerides were measured using were adjusted for total PCBs, p,p´-DDE, the eligible 22,347 boys, there were 241 cases standard enzymatic assays. triglycerides, and cholesterol as a priori of cryptorchidism and 214 cases of hypospa- The between-assay coefficient of varia- selected variables. Additional variables were dias. Five boys had both cryptorchidism and tion was 25% for trans-nonachlor and assessed as potential confounders using the hypospadias and were included in each group 20% for oxychlordane. These were deter- change in estimate method (Maldonado and for analysis. For comparison, we randomly mined at concentrations of 0.52 and 0.57 Greenland 1993), starting with all variables selected a group of boys (n = 599) from the μg/L (301 batches), respectively. Limits of in the models with deletion of one by one remaining eligible boys without a diagnosis of detection (LODs) were 0.28 μg/L for trans- in a stepwise manner. If, on deletion, the cryptorchidism and/or hypospadias. Controls nonachlor and 0.20 μg/L for oxychlordane ; OR for the contrast of the highest-to-lowest were selected so that the control:case ratio 29% of the values for trans-nonachlor were chlordane strata or the OR from the trend would be > 2:1 for each case group. below that value, whereas 31% of the values test changed by ≥ 15%, the factor was con- Hypospadias (urethral opening on the for oxychlordane were below the LOD. For sidered a confounder and was included in ventral side of the penis) was defined as having values below the LOD, we used the signal the adjusted analyses. Potential confounding a diagnosis any time within the first 7 years recorded by the instrument, when available, factors included race, maternal age, mater- of life. The degree of hypospadias was not because it is thought that the signals below nal history of previous live birth, season of | | Environmental Health Perspectives • volume 120 number 3 March 2012 479 Trabert et al. birth, socioeconomic index, smoking dur- index was the only variable that changed the triglycerides, cholesterol, serum p,p´-DDE, ing pregnancy, and gestational hypertension, OR by ≥ 15%. We also considered the effects total PCBs, gestational hypertension, and as defined in Table 1. Additional potential of adjustment for the continuous variables socioeconomic index, using the cross-prod- confounding factors included the categorical birth weight and placental weight and the uct terms. Variables were coded as defined in variables (yes/no) hyperemesis gravidarum, categorical variables preterm birth and small- Table 1, with the exception of prepregnancy history of infertility, menstrual cycle irregu- for-gestational-age, even though these were BMI, which was coded as < 25, 25–29.9, larity, estrogen use during pregnancy, and potentially intermediate variables. ≥ 30 kg/m . We supplemented evaluation of progesterone use during pregnancy, as well as We evaluated effect modification by effect modification by categorical variables the continuous variables age at menarche and maternal age, race, smoking, prepregnancy with more than two categories by compar- weight gain during pregnancy. Socioeconomic body mass index (BMI), previous live births, ing the model fit statistics for models with and without the cross-product terms. If the Table 1. Characteristics of mothers and sons according to case–control status of the son, CPP, 1959–1965. p-value associated with the interaction term based on the likelihood ratio test had a value Characteristic Cryptorchidism (n = 217) Hypospadias (n = 197) Control (n = 557) < 0.10, the degree of potential effect modi - Race (%) fication was further evaluated by examining White 57.6 50.8 46.5 tables stratified by the potentially modify- Black 41.5 43.7 47.8 Other 0.9 5.6 5.7 ing factor(s). Statistical significance was set at Gestation (week) p < 0.05 for main effects based on two-sided Median (IQR) 39 (37–40) 39 (37–41) 39 (38–40) tests. Statistical analyses were conducted using Preterm birth (%) 19.3 20.5 13.8 SAS statistical software package, version 9.2 Birth weight (g) (SAS Institute Inc., Cary, NC, USA). Median (IQR) 3,232 (2,863–3,629) 3,147 (2,722–3,487) 3,260 (2,948–3,600) Small for gestational age (%) 10.7 19.9 5.0 Results Maternal age (years) The distribution of selected demographic Median (IQR) 24 (21–30) 23 (20–29) 22 (20–28) Previous live births (%) and health characteristics of the mothers and 0 26.3 31.0 30.3 sons stratified by study group are provided 1 23.0 21.8 22.8 in Table 1. For both cryptorchid and hypos- ≥ 2 50.7 47.2 46.9 padias case groups, a higher percentage of Season of birth (%) case boys than control boys were white, born January–March 24.4 17.3 24.6 preterm, and small-for-gestational-age (birth April–June 28.1 25.4 24.2 weight < 10th percentile). The distribution of July–September 24.9 29.9 25.9 October–December 22.6 27.4 25.3 other potential confounding factors according Socioeconomic index to case–control status has been published pre- Median (IQR) 4.7 (3.3–6.3) 4.7 (3.0–6.3) 4.7 (3.3–6.0) viously (Longnecker et al. 2002). Prepregnancy BMI The median maternal serum concentra- Median (IQR) 22.2 (20.3–25.0) 21.8 (19.6–24.2) 22.2 (20.0–24.9) tions of trans-nonachlor, oxychlordane, DDE, Gestational hypertension (%) 5.9 6.3 6.2 total PCBs, cholesterol, and triglycerides Maternal smoking status (%) for each study group are shown in Table 2. None 50.7 52.0 55.0 The median level of maternal serum trig- 1–10 cigarettes per day 32.3 27.0 26.7 > 10 cigarettes per day 17.0 20.9 18.4 lycerides was slightly lower for hypospadias Study center (%) cases (188 μg/L) compared with cryptorchid Boston, MA 32.7 27.4 24.6 cases (204 μg/L) or controls (204 μg/L). The Buffalo, NY 7.4 4.1 3.8 Spearman correlation coefficients of trans- New Orleans, LA 6.5 4.1 4.7 nonachlor, oxychlordane, DDE, total PCBs, New York City, NY 2.3 3.6 3.1 cholesterol, and triglycerides are provided in Baltimore, MD 5.5 6.1 7.9 Table 3. All of the exposures evaluated were Richmond, VA 7.4 6.1 6.1 Minneapolis, MN 3.7 4.1 5.4 positively correlated, with the strongest corre- New York City, NY 1.4 5.1 7.5 lation between trans-nonachlor and oxychlor- Portland, OR 4.1 5.6 7.2 dane (r = 0.78), as expected. trans-Nonachlor Philadelphia, PA 15.2 23.9 17.6 was moderately associated with DDE Providence, RI 10.6 8.1 5.2 (r = 0.52), while the remaining compounds Memphis, TN 3.2 2.0 7.0 were less correlated (r = 0.08–0.46). IQR, interquartile range (quartiles 1–3). Table 4 provides the adjusted ORs for a b Columbia-Presbyterian Medical Center. New York Medical College. trans-nonachlor and oxychlordane quartiles conditioned on study center and adjusted Table 2. Maternal serum values by son’s case–control status, CPP, 1959–1965 [median (IQR)]. for serum DDE concentration, total PCBs, Characteristic Cryptorchidism (n = 217) Hypospadias (n = 197) Control (n = 557) triglycerides, cholesterol, and socioeconomic trans-Nonachlor (μg/L) 0.36 (0.26–0.52) 0.40 (0.24–0.58) 0.38 (0.25–0.57) index. For cryptorchidism, the quartile- Oxychlordane (μg/L) 0.29 (0.14–0.47) 0.34 (0.18–0.58) 0.31 (0.16–0.55) specific ORs for oxychlordane or trans-non- DDE (μg/L) 23.6 (15.9–35.3) 23.8 (16.6–34.4) 24.5 (16.7–37.5) achlor were close to the null as well as not Total PCBs with imputed congener (μg/L) 2.8 (2.0–3.9) 2.9 (2.1–4.2) 2.7 (1.8–3.9) statistically significant. Similarly for hypos - Total cholesterol (μg/L) 232 (190–273) 232 (192–277) 234 (197–280) padias, the quartile-specific ORs for trans- Triglycerides (μg/L) 204 (161–252) 188 (156–256) 204 (159–259) nonachlor and oxychlordane were close to IQR, interquartile range (quartiles 1–3). Eleven cases of cryptorchidism, 16 cases of hypospadias, and 25 controls were missing data for oxychlordane, respectively. the null as well as not statistically significant. | | 480 volume 120 number 3 March 2012 • Environmental Health Perspectives Chlordanes and male congenital anomalies We found no significant increase or decrease from the analysis that included all boys with lipid in controls) (Damgaard et al. 2006). The in risk of either cryptorchidism or hypos- cryptorchidism (results not shown). investigators reported no relationship between padias with increasing quartile of maternal oxychlordane and cryptorchidism and a Discussion trans- nonachlor or oxychlordane level (trend border line significant association between p-values all > 0.40). In the present study, a prospective evalua- cis-chlordane and cryptorchidism. However, Potential effect modification at a p-value tion of chlordane levels during pregnancy and the percent detection for cis-chlordane was of < 0.10 was present for the cryptorchid- cryptorchidism or hypospadias, the results do only 30.8%, so median concentrations were ism–oxychlordane association by smoking, not support an association between chlordane not reported (Damgaard et al. 2006). Our the hypospadias–oxychlordane association by levels and cryptorchidism or hypospadias. finding of no association between oxychlor - BMI, and the hypospadias–trans-nonachlor Further, the lack of association was consis- dane level and cryptorchidism is consistent association by smoking. After further evalu- tent in the subgroup of primiparous women. with that of Damgaard and colleagues. We ation in stratified models, there were some The median serum concentration of trans- did not evaluate trans- or cis-chlordane; how- differences in ORs by smoking or BMI ; how- nonachlor in the United States was 14.8 ng/g ever, we report no association between trans- ever, all estimates were consistent with no lipid in 2003–2004, the most recent years for nonachlor, another chlordane derivative, and association. Given that the main effects were which data are available; the median of oxy- cryptorchidism. not indicative of an association, the results chlordane was 10.3 ng/g lipid (CDC 2011). Teratogenic effects of chlordane expo- from the stratified models are not presented. The median serum concentrations of trans- sure have not been observed in animal studies In subanalyses, we evaluated the associa- nonachlor and oxychlordane were 48.0 ng/g (Abadin et al. 1994). However, chlordane tions between the chlordanes and cryptorchid- lipid and 37.5 ng/g lipid, respectively, in the exposures has been shown to affect repro- ism and hypospadias among women with no present study samples that were collected in duction in test animals, delaying puberty, history of a live birth. We hypothesized that the 1960s. This suggests that current low-level disrupting estrous cycling in females, and the trans-nonachlor and oxychlordane expo- chlordane exposure is unlikely to be related to reducing fertility by as much as 50% (Welch sure to the male fetus would be highest in the development of either condition. et al. 1971). Furthermore, chlordane is clas- this group of women. Although the quartile- To our knowledge, this is the first epi- sified as potentially carcinogenic to humans specific ORs were increased slightly compared demiologic evaluation of the association (Group 2B) by the International Agency for with the corresponding ORs for all women between chlordane levels and hypospadias Research on Cancer (IARC 2001) and has (data not shown), there were no significant and only the second such study of the asso- been associated with risk of TGCT (Hardell trends with increasing quartile of mater- ciation between chlordane levels and cryp- et al. 2003; McGlynn et al. 2009b; Purdue nal trans-nonachlor or oxychlordane level torchidism. Using breast milk as a proxy for et al. 2009). Although chlordane may be a for either cryptorchidism (trans-nonachlor: maternal exposure, Damgaard et al. (2006) carcinogen, both animal data and the present p-trend = 0.75; oxychlordane: p-trend = 0.97) evaluated persistent pesticide concentrations study suggest it is not a teratogen. or hypospadias (trans-nonachlor: p-trend = in milk samples collected from mothers of The present study has several strengths. 0.68; oxychlordane: p-trend = 0.90). Finally, cryptorchid boys (n = 62) and healthy boys (n The major advantages were that the study was after exclusion of the boys (n = 103) whose = 68). Maternal milk levels of trans-chlordane a large and prospective evaluation of chlor- testicles were initially descended at birth, were slightly higher in mothers of cryptorchid dane exposure and congenital malformations. the association between chlordane level and boys than mothers of healthy boys; however, This study has several potential weaknesses cryptorchidism was not substantially different the overall exposure was very low (0.04 ng/g that merit consideration. Chlordane levels were measured on samples stored for approxi- Table 3. Spearman correlation coefficients of selected covariates, CPP, 1959–1965. mately 40 years. However, prior studies of freeze–thaw cycles have demonstrated that trans-Nonachlor Oxychlordane DDE Total PCBs Total cholesterol Triglycerides chlordane levels are quite stable over time trans-Nonachlor (μg/L) 1.00 (Becker et al. 1997). In addition, cholesterol Oxychlordane (μg/L) 0.78 1.00 and triglyceride levels were in the expected DDE (μg/L) 0.52 0.38 1.00 Total PCBs (μg/L) 0.38 0.46 0.37 1.00 range, suggesting that substantial degrada- Total cholesterol (μg/L) 0.10 0.12 0.15 0.25 1.00 tion had not occurred. The chlordane con- Triglycerides (μg/L) 0.11 0.12 0.08 0.15 0.36 1.00 centrations for this study were measured in Table 4. Adjusted ORs and 95% CIs for cryptorchidism and hypospadias in relation to maternal oxychlordane and trans-nonachlor levels (μg/L), CPP, 1959–1965. Cryptorchidism Hypospadias a b b Chlordane (quartiles) Control (n) Case (n) Unadjusted OR Adjusted OR (95% CI) Case (n) Unadjusted OR Adjusted OR (95% CI) trans-Nonachlor (µg/L) 0.00 to < 0.25 134 47 1.00 (reference) 1.00 (reference) 50 1.00 (reference) 1.00 (reference) 0.25 to < 0.38 139 67 1.25 1.27 (0.85, 1.90) 44 0.83 0.84 (0.54, 1.33) 0.38 to < 0.57 142 62 1.33 1.46 (0.93, 2.30) 52 1.00 1.04 (0.64, 1.70) ≥ 0.57 142 41 1.10 1.22 (0.70, 2.12) 51 0.99 1.08 (0.62, 1.89) Trend test 557 217 0.55 197 0.60 Oxychlordane (µg/L) 0.00 to < 0.16 131 54 1.00 (reference) 1.00 (reference) 38 1.00 (reference) 1.00 (reference) 0.16 to < 0.31 131 56 0.95 0.90 (0.60, 1.35) 47 1.12 1.06 (0.66, 1.69) 0.31 to < 0.55 137 51 0.90 0.91 (0.59, 1.42) 44 1.03 1.07 (0.64, 1.80) ≥ 0.55 133 45 0.94 0.95 (0.55, 1.64) 52 1.15 1.24 (0.69, 2.22) Trend test 532 206 0.90 181 0.46 a b Quartile cut points were defined based on 557 and 532 controls for trans-nonachlor and oxychlordane, respectively. Adjusted ORs and 95% CIs are from conditional logistic regression models adjusted for serum DDE concentration (five categories), total PCBs (four categories), triglycerides, cholesterol, and socioeconomic index. Ordinal test across four categories using the median value within each group. | | Environmental Health Perspectives • volume 120 number 3 March 2012 481 Trabert et al. third-trimester samples; the critical window Refe Rences chlordanes in mothers of men with testicular cancer. Environ Health Perspect 111:930–934. of exposure might be earlier for hypospadias, Hardy JB. 2003. The Collaborative Perinatal Project: lessons Abadin HG, Baynes R, Goetchius PF. 1994. Toxicological Profile likely mid-to-late first trimester (Husmann and legacy. Ann Epidemiol 13:303–311. for Chlordane. Atlanta, GA:U.S. Department of Health 2002). However, third-trimester assays should Husmann DA. 2002. Micropenis: an animal model and its human and Human Services, Agency for Toxic Substances and correlates. Adv Exp Med Biol 511: 41–54. reflect first-trimester exposure, given the Disease Registry (ATSDR). IARC (International Agency for Research on Cancer). 2001. Barthold JS, Gonzalez R. 2003. The epidemiology of congenital long biological half-life of chlordane (Abadin Some Thyrotropic Agents. IARC Monogr Eval Carcinog cryptorchidism, testicular ascent and orchiopexy. J Urol et al. 1994). In the CPP, the prevalence of Risk Hum 79. 170:2396–2401. Ivie GW. 1973. Nature and toxicity of two oxychlordane photoi- cryptorchidism and hypospadias among Becker PR, Mackey EA, Demiralp R, Schantz MM, Koster BJ, somers. J Agric Food Chem 21:1113–1115. Wise SA. 1997. Concentrations of chlorinated hydro- males within the first year of life was 56 Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, carbons and trace elements in marine mammal tissues per 10,000 and 41 per 10,000, respectively Needham LL, et al. 2002. Maternal serum level of 1,1-di- archived in the U.S. National Biomonitoring Specimen chloro-2,2-bis(p-chlorophenyl)ethylene and risk of cryp- (Myrianthopoulos and Chung 1974). These Bank. Chemosphere 34:2067–2098. torchidism, hypospadias, and polythelia among male Bennett GW, Ballee DL, Hall RC, Fahey JE, Butts WL, Osmun JV. prevalences were higher than in other U.S. offspring. Am J Epidemiol 155:313–322. 1974. Persistence and distribution of chlordane and diel- populations (cryptorchidism: 19 per 10,000; Maldonado G, Greenland S. 1993. Simulation study of con- drin applied as termiticides. Bull Environ Contam Toxicol hypospadias: 21 per 10,000) (Paulozzi 1999). founder-selection strategies. Am J Epidemiol 138:923–936. 11:64–69. McGlynn KA, Guo X, Graubard BI, Brock JW, Klebanoff MA, Biggs ML, Davis MD, Eaton DL, Weiss NS, Barr DB, Doody DR, This difference was likely the result of reduced Longnecker MP. 2009a. Maternal pregnancy levels of et al. 2008. Serum organochlorine pesticide residues and misclassification in CPP, given that congeni - polychlorinated biphenyls and risk of hypospadias and risk of testicular germ cell carcinoma: a population-based tal anomalies were identified in a series of cryptorchidism in male offspring. Environ Health Perspect case-control study. Cancer Epidemiol Biomarkers Prev 117:1472–1476. systematic examinations, whereas birth defect 17:2012–2018. M c G l y n n K A , Q u r a i s h i S M , G r a u b a r d B I , W e b e r J P , Brock JW, Burse VW, Ashley DL, Najam AR, Green VE, registries are typically based on routine reports Rubertone MV, Erickson RL. 2009b. Polychlorinated biphe- Korver MP, et al. 1996. An improved analysis for chlori- and records only. The ratio of hypospadias to nyls and risk of testicular germ cell tumors. Cancer Res nated pesticides and polychlorinated biphenyls (PCBs) 69:1901–1909. cryptorchidism was somewhat higher than in human and bovine sera using solid-phase extraction. Myrianthopoulos NC, Chung CS. 1974. Congenital malforma- J Anal Toxicol 20:528–536. what might be expected. We included hypo- tions in singletons: epidemiologic survey. Report from the CDC (Centers for Disease Control and Prevention). 2011. Fourth spadias diagnosed up until 7 years of age, Collaborative Perinatal project. Birth Defects Orig Artic National Report on Human Exposure to Environmental Ser 10:1–58. whereas we included cryptorchidism cases Chemicals. NCEH Pub No. 01-0164:1-6. Atlanta, GA:Centers Myrianthopoulos NC, French KS. 1968. An application of the for Disease Control and Prevention. diagnosed through the first year of life only. U.S. Bureau of the Census socioeconomic index to a large, Chevrier J, Harley KG, Bradman A, Gharbi M, Sjodin A, Further, the medical records did not denote diversified patient population. Soc Sci Med 2:283–299. Eskenazi B. 2010. Polybrominated diphenyl ether (PBDE) Niswander KR, Gordon M. 1972. The Women and Their the degree of hypospadias, which limited our flame retardants and thyroid hormone during pregnancy. Pregnancies. Philadelphia:WB Saunders. Environ Health Perspect 118:1444–1449. ability to evaluate the relationship of exposure Paulozzi LJ. 1999. International trends in rates of hypospadias Cook MB, Trabert B, McGlynn KA. 2011. Organochlorine com- and disease severity. and cryptorchidism. Environ Health Perspect 107:297–302. pounds and testicular dysgenesis syndrome: human data. In conclusion, these results provide no Purdue MP, Engel LS, Langseth H, Needham LL, Andersen A, Int J Androl 34(4 pt 2):e68–e84. Barr DB, et al. 2009. Prediagnostic serum concentrations Damgaard IN, Skakkebaek NE, Toppari J, Virtanen HE, Shen H, clear evidence of an effect of in utero chlor- of organochlorine compounds and risk of testicular germ Schramm KW, et al. 2006. Persistent pesticides in human dane (trans-nonachlor or oxychlordane) levels cell tumors. Environ Health Perspect 117:1514–1519. breast milk and cryptorchidism. Environ Health Perspect on cryptorchidism or hypospadias, and the Skakkebaek NE. 2003. Testicular dysgenesis syndrome. Horm 114:1133–1138. Res 60(suppl 3):49. findings are in line with results from animal Dearth MA, Hites RA. 1991. Chlordane accumulation in people. Toppari J, Virtanen HE, Main KM, Skakkebaek NE. 2010. Environ Sci Technol 25:1279–1285. studies. Further, it is unlikely that current Cryptorchidism and hypospadias as a sign of testicular Gaskins AJ, Schisterman EF. 2009. The effect of lipid adjust- chlordane exposure is related to the develop- dysgenesis syndrome (TDS): environmental connection. ment on the analysis of environmental contaminants and Birth Defects Res A Clin Mol Teratol 88:910–919. ment of either anomaly, because serum chlor- the outcome of human health risks. Methods Mol Biol Welch RM, Levin W, Kuntzman R, Jacobson M, Conney AH. 580:371–381. dane levels at the time of sample collection, 1971. Effect of halogenated hydrocarbon insecticides on Hardell L, Van BB, Lindstrom G, Carlberg M, Dreifaldt AC, the early 1960s, were considerably higher than the metabolism and uterotropic action of estrogens in rats Wijkstrom H, et al. 2003. Increased concentrations levels at present. and mice. Toxicol Appl Pharmacol 19:234–246. of polychlorinated biphenyls, hexachlorobenzene, and | | 482 volume 120 number 3 March 2012 • Environmental Health Perspectives

Journal

Environmental Health PerspectivesPubmed Central

Published: Sep 7, 2011

References