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 Polychlorinated Biphenyls and Risk of Hypospadias and Cryptorchidism in Male Offspring

Maternal Pregnancy Levels of Polychlorinated Biphenyls and Risk of Hypospadias and Cryptorchidism... Research Children’s Health Maternal Pregnancy Levels of Polychlorinated Biphenyls and Risk of Hypospadias and Cryptorchidism in Male Offspring 1 2 1 3 4 Katherine A. McGlynn, Xuguang Guo, Barry I. Graubard, John W. Brock, Mark A. Klebanoff, and Matthew P. Longnecker Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and 2 3 Human Services, Bethesda, Maryland, USA; Westat, Inc., Durham, North Carolina, USA; National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; Division of Epidemiology, Statistics, and Prevention Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, 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 and 1965 at 12 U.S. medical centers located in Background : The etiologies of the male urogenital anomalies cryptorchidism and hypospadias are Baltimore, Maryland; Boston, Massachusetts; poorly understood. It has been suggested, however, that in utero hormone levels may be related to Buffalo, New York; Memphis, Tennessee; risk. Endocrine-disrupting chemicals, including polychlorinated biphenyl (PCB) compounds, may Minneapolis, Minnesota; New Orleans, alter hormone levels and thereby affect the fetus. Louisiana; New York City (two centers); o Bjectives : To examine whether in utero PCB exposure is related to cryptorchidism and hypo- Philadelphia, Pennsylvania; Portland, Oregon; spadias, we examined PCB levels among pregnant women enrolled in the Collaborative Perinatal Providence, Rhode Island; and Richmond, Project (CPP). Virginia. Eleven centers recruited participants Methods : The CPP enrolled pregnant women at 12 U.S. medical centers between 1959 and 1965. from the prenatal clinics of a university hos- For the present research, we analyzed third-trimester serum samples from the mothers of 230 sons pital, whereas one (Buffalo) recruited from with cryptorchidism, 201 sons with hypospadias, and 593 sons with neither condition. We esti- 13 private obstetric practices. Participant mated adjusted odds ratios (ORs) and 95% confidence intervals (CIs) using logistic regression and selection varied among the medical centers. examined the associations of each anomaly with individual PCB congener levels, sum of PCBs, and Women were ineligible if they were incarcer- several functional groupings of PCBs. ated, planned to leave the area on delivery, or results : In general, the ORs for cryptorchidism or hypospadias showed no notable associations planned to place their child for adoption or if with individual PCB congener levels or functional groupings of PCBs. However, the ORs and 95% they delivered on the day they were recruited CIs for the sum of PCBs associated with hypospadias were as follows: 0–1.9 µg/L, reference group; for the study. The characteristics of the women 2–2.9 µg/L, OR = 1.57, 95% CI, 1.05–2.34; 3–3.9 µg/L, OR = 1.45, 95% CI, 0.90–2.34; and in the sample were, at registration, essen- ≥ 4.0 µg/L, OR = 1.69, 95% CI, 1.06–2.68; p-value for trend = 0.08. tially the same as those in the sampling frame conclusions : Given the large number of associations examined, these findings do not strongly (Niswander and Gordon 1972). Four per- support the hypothesis that PCBs are associated with cryptorchidism or hypospadias. Because cent of the participants were lost to follow-up population serum PCB levels at the time of sample collection were considerably higher than levels before delivery. Once enrolled, the women’s at present, it is unlikely that current PCB exposure is related to the development of either anomaly. nonfasting blood was drawn approximately k e y w o r d s : cryptorchidism, hypospadias, polychlorinated biphenyls, testicular dysgenesis syn- every 8 weeks during pregnancy, at delivery, drome. Environ Health Perspect 117:1472–1476 (2009). doi:10.1289/ehp.0800389 available via and at 6 weeks postpartum. Sera were stored http://dx.doi.org/ [Online 20 April 2009] in glass vials at –20°C with no recorded thaws. Approximately 42,000 women were enrolled, The hormone hypothesis of male reproductive since the 1960s but remain detectable in most and 55,000 children were born in the study. disorders suggests that both endogenous and segments of the population (Ross 2004). To The children were systematically assessed exogenous hormones might be risk factors date, only four published studies have exam- for the presence of birth defects and other for cryptorchidism, hypospadias, impaired ined the association between PCBs and either spermatogenesis, and testicular cancer (Sharpe cryptorchidism or hypospadias. Two case– Address correspondence to K.A. McGlynn, Hormonal 2003). Exogenous hormonal exposures involve control studies of PCBs and cryptorchidism and Reproductive Epidemiology Branch, Division of a variety of endocrine-disrupting chemicals reported no relationship with risk (Hosie Cancer Epidemiology and Genetics, National Cancer Institute, NIH, EPS Suite 550, 6120 Executive Blvd., whose effects range from estrogenic to anti - et al. 2000; Mol et al. 2002), whereas a third Rockville, MD 20852-7234 USA. Telephone: (301) estrogenic, androgenic, and antiandrogenic. offered measured support (Brucker-Davis 435-4918. Fax: (301) 402-0416. E-mail: mcglynnk@ Evidence in support of the hypothesis has et al. 2008). One ecologic study of PCBs and mail.nih.gov consisted largely of animal data (Toppari et al. hypospadias reported an inverse association Support for this research was provided by the 1996) and temporal trends (Paulozzi 1999) in (Giwercman et al. 2006). Because the prior Intramural Research Programs of the National Cancer the prevalence of the so-called testicular dys- studies may have been too small to detect Institute, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, genesis syndrome (TDS) conditions. Human statistically significant differences, the present and the National Institute of Environmental studies of the relationship between endocrine- study was conducted among a large, well- Health Sciences of the National Institutes of Health disrupting chemicals and the TDS disorders described population in which the serum (NIH) and by the Centers for Disease Control and have been relatively few. samples were collected at a time when PCB Prevention (CDC). Polychlorinated biphenyl (PCB) com- levels in the United States were higher. The findings in this study are those of the authors pounds are ubiquitous, stable, environmental and do not necessarily reflect the official opinion of Materials and Methods the NIH or the CDC. pollutants that were widely used in devel- X.G. is employed by Social and Scientific Systems, oped countries between 1929 and 1977. Even Study population. The Collaborative Perinatal Inc., which has no PCB-related contracts. The other after they were banned, PCBs continued Project (CPP) was a prospective study of authors declare they have no competing financial to be used in “closed” source applications neurologic disorders and other conditions interests. such as electrical capacitors and transform- in children (Niswander and Gordon 1972). Received 11 November 2008; accepted 20 April ers. Levels of PCBs in humans have declined Pregnant women were enrolled between 1959 | | 1472 volume 117 number 9 September 2009 • Environmental Health Perspectives PCBs and risk of hypospadias and cryptorchidism outcomes at birth and through 7 years of age. Laboratory assays. Serum levels of 11 PCBs as continuous variables and serum p,p´-dichloro- Follow-up to 7 years of age was completed for were measured at the Centers for Disease diphenyldichloroethylene (p,p´-DDE) as a five- about 75% of children born into the study. Control and Prevention after solid-phrase stratum categorical variable. For the present study, we employed a extraction cleanup and dual-column gas chro- We evaluated confounding by comparing nested case–control design to examine the matography using electron capture detection the hazard ratio of the baseline model (includ- association between maternal serum PCB (Brock et al. 1996). The congeners, designated ing sum of PCBs, triglycerides, cholesterol, and levels and risk of cryptorchidism and hypo- by International Union of Pure and Applied p,p´-DDE) to the hazard ratio of the model spadias among sons. To be eligible for inclu- Chemistry code, were PCB-28, PCB-52, that also included the possible confounder. We sion in the present study, the mother must PCB-74, PCB-105, PCB-118, PCB-138, modeled sum of PCBs both as a continuous have given birth to a singleton, live-born male PCB-153, PCB-170, PCB-180, PCB-194, and variable and as a categorical variable and used infant and must have had a 3-mL aliquot of PCB-203. The proportion of PCBs in each results of both analyses to determine whether third-trimester maternal serum available. We sample recovered by extraction was approxi- the factor was a confounder. If the hazard ratio defined cryptorchidism as having had a diag - mately 60%. The results shown are unadjusted per microgram per liter of PCB or the haz- nosis of undescended testis(es) at any time for recovery. The between-assay coefficient of ard ratio for the contrast of highest-to-lowest during the first year of life. The diagnosis was variation was 19% at 3.49 µg PCB. A labora- PCB strata changed by 15% or more, the fac- made by pediatricians based on serial exami- tory result was not obtained for 3% of samples tor was considered a confounder. The factors nations that included inspection and palpa- because the measured value did not meet the considered as possible confounders were race, tion of the genitalia. Boys first noted in the quality-control standards for acceptance. Thus, season of birth, maternal age, maternal parity, medical record to have undescended testis(es) the PCB results were not available for 30 sam- socioeconomic index, prepregnancy body mass after the first year of life were not considered ples. Because PCBs are lipid soluble and because index, weight gain during pregnancy, smoking cryptorchid because they may have had retrac- serum contains a variable amount of lipid, we during pregnancy, hyperemesis gravidarum, tile testes. Hypospadias was defined as having measured serum cholesterol and triglycerides gestational hypertension, age at menarche, his- a diagnosis any time during the first 7 years of using standing enzymatic assays. tory of infertility, menstrual cycle irregularity, life. Degree of hypospadias was not noted in In addition to analyzing each congener estrogen use during pregnancy, and progester- the medical records. sepa rately, we examined four groupings of con- one use during pregnancy. Similarly, we con- Among the 28,444 boys in the CPP, 267 geners. The Wolff groupings (Wolff et al. 1997) sidered the effects of adjustment for preterm were not live born and 441 were not singletons. were as follows: Wolff group 1A, potentially birth, birth weight, placental weight, and small- No blood sample was available from the moth- estrogenic, weak phenobarbital inducers, not for-gestational-age even though these were ers of 5,389 boys. Among the eligible 22,347 persistent congeners (PCB-52); Wolff group potentially intermediate variables. None of the boys, there were 241 cases of cryptorchidism 2A, potentially antiestrogenic and immuno- factors considered, however, changed the hazard and 214 cases of hypospadias. The five boys toxic, dioxin-like, non-ortho– and mono-ortho– ratios by ≥ 15%. Using a similar approach, but with both cryptorchidism and hypospadias substituted, moderately persistent congeners with cross-product terms, we evaluated effect were included in each group for the analysis. (PCB-74, PCB-105, PCB-118); Wolff group modification by maternal age, race, smoking, From the pool of eligible mothers, 599 were 2B, limited dioxin activity, di-ortho substituted, prepregnancy body mass index, triglycerides, selected at random so that the control:case ratio persistent congeners (PCB-138, PCB-170) and cholesterol, serum p,p´-DDE, gestational hyper- would be > 2:1 for each condition. Wolff group 3, phenobarbital inducers, CYP1A tension, socioeconomic index, and study center. Among the boys diagnosed as cryptorchid and CYP2B inducers, and persistent congeners We supplemented evaluation of effect modifica - during the first year of life, the study records (PCB-153, PCB-180, PCB-203). The toxic tion by categorical variables with more than two indicated that the testicles were descended at equivalency factor (TEF) groupings (Ahlborg categories by comparing the model fit statistics birth in 103, suggesting that these boys may et al. 1994) were TEF mono-ortho (PCB- for models with and without the cross-product have had acquired, rather than congenital, 105, PCB-118), and TEF di-ortho (PCB-170, terms. If the p-value associated with the inter- undescended testis (Barthold and Gonzalez PCB-180). The uridine diphosphate-glucurono - action term based on the likelihood ratio test 2003). To evaluate whether a relationship syl transferase (UDP-GT) inducer, cytochrome had a value ≤ 0.10, the degree of potential effect with PCBs varied by type of cryptorchidism, P450 1A (CYP1A), and CYP2B inducer group- modification was further considered by examin - we considered the boys with testis descended ing (Chevrier et al. 2007) was PCB-118 and ing tables stratie fi d by the potentially modifying at birth separately in a sensitivity analysis. PCB-180. The molecular-weight groups were factor(s). We further evaluated sum of PCBs Among the boys with undescended testicle(s) low (PCB-28, PCB-52, PCB-74, PCB-105, by forming ordinal categories and including at birth (n = 138), all but one also had a sub- PCB-118) and high (PCB-138, PCB-153, corresponding dummy variables in the logistic sequent observation of cryptorchidism in at PCB-170, PCB-180, PCB-194, PCB-203). regression analysis to estimate adjusted odds least one of the three subsequent examination Statistical methods. We converted PCBs ratios (ORs). We performed trend tests across (ages 4 months, 1 year, 7 years) or documen- to molar concentrations (PCB in micrograms categories by including a single independent tation of orchidopexy. per liter divided by the congener’s molecu- variable with integer scores corresponding to The socioeconomic index calculated lar weight) and then analyzed on a continu- the categories testing the significance of the for participants in the CPP was the mean ous scale. Five congeners had missing values regression coefficient. All statistical analyses of the three percentile scores: education of (PCB-28, n = 2; PCB-74, n = 26; PCB-118, were conducted using the SAS statistical soft- head of the household, occupation of head n = 42; PCB-138, n = 22; PCB-180, n = 2), ware package, version 9.1 (SAS Institute Inc., of the household or chief wage earner, and mainly because the measured value did not meet Cary, NC). All p-values were two-sided. family income. The score used to calculate the quality control standards for acceptance Results the percentile for an occupation was based (Ballschmiter et al. 1992). We estimated the on the percentiles of education and income change in log odds of having cryptorchidism or Based on the distribution of the controls, the among persons with the same occupation hypospadias per millimole increase in PCB level study population was 46% white, 48% black, (Myrianthopoulos and French 1968). Verbal using conditional logistic regression, condi- and 5.6% other racial/ethnic groups (primarily consent to participate was elicited from all tioned on study center (12 strata). Models were Hispanic and Asian) (Table 1). Birth weight mothers (Hardy 2003). adjusted for serum triglycerides and cholesterol and gestational age medians were typical of | | Environmental Health Perspectives • volume 117 number 9 September 2009 1473 McGlynn et al. healthy births. The mothers were a median age previous report has described the study popula- cryptorchidism, the ORs ranged from 1.03 of 22 years and had a median socioeconomic tion in detail (Longnecker et al. 2002). (PCB-118) to 1.79 (PCB-203), and none index (mean of three percentile scores: educa- Table 2 shows median maternal PCB lev- were statistically significant. For hypospadias, tion of head of household, occupation of head els, lipid levels, and DDE levels for each of the ORs ranged from 0.80 (PCB-52) to 1.99 of household or chief wage earner, and family the study groups. The median level of sum (PCB-203), and as with cryptorchidism, none income) of 4.5, which was just under of the of PCBs was slightly lower in the controls were statistically significant. median of 5.0 for the United States population (2.7 µg/L), whereas the median levels of Table 3 also shows the adjusted ORs for in the 1960s. About 31% of the mothers were p,p´-DDE (24.5 µg/L) and total cholesterol groupings of PCBs. We found no significant primiparas. Compared with the control boys, (234 µg/L) were slightly higher. The median increase in risk of either cryptorchidism or the boys with cryptorchidism and hypospadias level of triglycerides in the controls (204 µg/L) hypospadias with increasing level of sum of were more likely to be white, to be born pre- was intermediate to the two case groups. PCBs [cryptorchidism: OR = 1.01, 95% con- maturely, and to be born small for gestational Table 3 shows the adjusted ORs for the fidence interval (CI), 0.99–1.04; hypospadias: age (birth weight below 10th percentile). The PCB congener-specific analyses, conditioned OR = 1.01, 95% CI, 0.98–1.04]. Similarly, we boys with hypospadias also had a lower median on study center and adjusted for triglyc- found no significant increases in risk of either birth weight than did the control boys. A erides, cholesterol, and p,p´-DDE level. For outcome with any of the Wolff groupings (Wolff et al. 1997), TEF groupings (Ahlborg Table 1. Characteristics of mothers and sons according to the son’s case–control status, CPP, 1959–1965. et al. 1994), or enzyme inducer (Chevrier Characteristic Cryptorchidism (n = 230) Hypospadias (n = 201) Control (n = 593) et al. 2007) groupings. Because the ORs in the congener-specific analysis tended to be higher Race (%) among the higher-molecular-weight congeners, White 57.0 49.3 46.0 Black 41.3 44.8 48.4 we also analyzed PCBs stratified by molecular Other 1.7 6.0 5.6 weight. The analyses did not detect statistically Gestation (week) significant associations with either cryptorchid - Median (Q1, Q3) 39 (38, 41) 39 (38, 41) 39 (38, 40) ism (low: OR = 1.02, 95% CI, 0.97–1.06; Preterm birth (%) 16.6 19.5 14.0 high: OR = 1.04, 95% CI, 0.98–1.09) or hypo- Birth weight (g) spadias (low: OR = 1.00, 95% CI, 0.95–1.06; Median (Q1, Q3) 3,260 (2,835, 3,629) 3,147 (2,665, 3,487) 3,260 (2,948, 3,600) high: OR = 1.04, 95% CI, 0.98–1.10). Small for gestational age (%) 9.1 17.8 4.9 Maternal age (years) Because the examinations of cryptorchid- Median (Q1, Q3) 24 (21, 30) 24 (20, 29) 22 (20, 28) ism and hypospadias with the sum of PCBs Previous live births (%) approached statistical significance, we also 0 27.0 30.8 30.9 analyzed risk by PCB category (Table 4). 1 22.2 21.4 22.5 Cryptorchidism was not significantly related ≥ 2 50.9 47.8 46.6 to PCB level in any category, and the test for Socioeconomic index trend was not statistically significant (p for Median (Q1, Q3) 4.7 (3.3, 6.3) 4.3 (3.0, 6.2) 4.5 (3.3, 6.0) trend = 0.19). The results of a similar analysis Prepregnancy body mass index Median (Q1, Q3) 22.2 (20.3, 25.0) 21.8 (19.3, 24.1) 22.2 (20.0, 24.9) that modeled sum of PCBs on a lipid basis Gestational hypertension (%) 5.9 6.5 6.4 in quintiles produced the same conclusions Study Center (%) (data not shown). Although the trend in risk Boston, MA 31.3 25.4 23.6 of hypospadias approached significance ( p for Buffalo, NY 7.0 3.5 3.5 trend = 0.08) and the ORs for all categories New Orleans, LA 6.5 3.5 4.7 a were greater than unity, we found, overall, no New York City, NY 3.0 4.5 3.2 linear relationship. The examination of effect Baltimore, MD 6.1 6.5 7.6 Richmond, VA 7.0 5.5 5.7 modification showed that no factor signifi- Minneapolis, MN 3.5 4.0 5.2 cantly altered the results of the baseline models New York City, NY 1.7 6.0 7.8 (data not shown). A final examination of cryp - Portland, OR 4.3 5.5 7.1 torchidism excluded the boys (n = 103) whose Philadelphia, PA 14.8 23.9 18.5 testicles were initially descended at birth. Providence, RI 11.3 10.0 5.9 The results of that analysis (data not shown), Memphis, TN 3.5 2.0 7.1 however, did not differ from the analysis that a b Columbia-Presbyterian Medical Center. New York Medical College included all boys with cryptorchidism. Table 2. Maternal serum values by son’s case–control status, CCP, 1959–1965. Discussion Cryptorchidism (n = 230) Hypospadias (n = 201) Control (n =593) The results of the present study do not support Total cholesterol (ug/L) the hypothesis that in utero PCB exposure Median 231 229 234 increases the risk of hypospadias or cryp- (Q1, Q3) (190, 272) (190, 275) (195, 279) torchidism. Because the present study samples Triglycerides (ug/L) were collected in the 1960s, when PCB expo- Median 207.5 187 204 sure was substantially greater than at present, (Q1, Q3) (163, 258) (152, 253) (159, 259) the results also suggest that current low-level DDE (ug/L) PCB exposure is unlikely to be related to the Median 23.6 23.9 24.5 (Q1, Q3) (15.9, 35.3) (16.2, 34.4) (16.7, 37.2) development of either condition. Sum of PCBs with imputed congener (ug/L) With the inclusion of > 1,000 participants, Median 2.8 2.9 2.7 the present study is by far the largest to have (Q1, Q3) (2.0, 3.9) (2.1, 4.2) (1.8, 3.8) examined the relationship between PCBs and | | 1474 volume 117 number 9 September 2009 • Environmental Health Perspectives PCBs and risk of hypospadias and cryptorchidism cryptorchidism or hypospadias. Although the would also be related to the TDS disorders to affect fertility or to have direct hormone- prior literature is not extensive, three studies (Skakkebaek et al. 2001) that become evident like activity (Bonde et al. 2008). To date, the of cryptorchidism and PCBs and one study of in adulthood: impaired spermatogenesis and sole published study of testicular cancer and hypospadias and PCBs have been published. testicular cancer. More studies have examined PCBs found no association with sum of PCBs, Mol et al. (2002) examined cord blood PCB the relationship of PCBs with the former than estrogenic PCBs, or enzyme-inducing PCBs levels in a population known to have high the latter. In general, the results of the PCB– (Hardell et al. 2003). Taken as a group, then, at PCB exposure via s fi h consumption. Studying fertility studies are somewhat equivocal. Several the present time, there is little evidence that the 196 Faroe Island boys born between 1986 studies have found statistically significant associ - TDS disorders are related to PCB exposure. and 1987, the investigators found no relation- ations with impaired sperm parameters (Hauser Several potential weaknesses of the present ship between PCB levels and cryptorchidism. et al. 2003; Richthoff et al. 2003; Rozati et al. study merit consideration. A first consideration A similar conclusion was reached by Hosie 2002), whereas others have found no associa- is that the assays were run on stored, rather et al. (2000). Studying adipose PCB levels in tion (Weiss et al. 2006) or associations only in than fresh, serum samples. PCB levels, how- 18 German boys with cryptorchidism and subsets of their populations (Bush et al. 1986; ever, are quite stable over time, as has been 30 control boys, the investigators found no dif- Dallinga et al. 2002; Rignell-Hydbom et al. demonstrated in samples stored for 15 years ference in PCB levels. A third study examined 2004). In contrast, several studies have reported (Lunden and Noren 1998; Noren 1988). In PCB levels in both cord blood and breast milk direct associations between PCB levels and fer- addition, the cholesterol and triglyceride levels, in a French population (Brucker-Davis et al. tility (Cok et al. 2008; Ensslen et al. 1990). as reported in Table 2, were in the expected 2008). The comparison of 78 boys diagnosed The summary of the international INUENDO range, suggesting that substantial degradation at birth with cryptorchidism and 86 control study of PCBs and fertility in four populations, had not occurred. A second potential weakness boys found a significant association with breast however, concluded that PCBs did not appear is that determination of one of the outcomes, milk PCB levels but not with cord blood PCB Table 3. Adjusted ORs (95% CIs) for cryptorchidism and hypospadias in relation to maternal PCB levels levels. When they restricted the analysis to the (mmol/L), CPP, 1959–1965. boys who remained cryptorchid at 3 months, Percentiles among controls Cryptorchidism Hypospadias however, the relationship with breast milk PCB a a levels was no longer significant. The only prior 25th 50th 75th 95th Adjusted OR (95% CI) Adjusted OR (95% CI) study of hypospadias and PCB levels reported PCB congener to date was an ecologic study in Greenland. PCB-28 0.39 0.70 0.97 1.79 1.07 (0.86–1.34) 1.07 (0.86–1.32) The investigators found a low prevalence rate of PCB-52 0.00 0.00 0.00 0.51 1.11 (0.81–1.52) 0.80 (0.42–1.52) hypospadias despite high levels of PCBs in the PCB-74 0.51 0.79 1.13 2.16 1.04 (0.87–1.25) 1.03 (0.84–1.26) population (Giwercman et al. 2006), suggest- PCB-105 0.00 0.34 0.52 1.19 1.11 (0.82–1.50) 1.03 (0.69–1.53) PCB-118 1.13 1.65 2.39 5.21 1.03 (0.95–1.12) 0.99 (0.88–1.11) ing that PCBs might be inversely, rather than directly, associated with risk of hypospadias. PCB-138 1.16 1.61 2.36 3.85 1.09 (0.94–1.25) 1.08 (0.93–1.26) PCB-153 1.22 1.69 2.52 4.02 1.11 (0.97–1.27) 1.12 (0.98–1.30) PCBs have been suspected of being related to male urogenital anomalies princi- PCB-170 0.00 0.23 0.38 0.66 1.07 (0.59–1.94) 1.34 (0.73–2.46) PCB-180 0.35 0.56 0.83 1.39 1.19 (0.84–1.69) 1.28 (0.89–1.84) pally because some congeners and metabo- PCB-194 0.00 0.00 0.21 0.37 1.42 (0.54–3.72) 1.41 (0.48–4.19) lites display weak estrogenic activity. There is PCB-203 0.00 0.00 0.21 0.42 1.79 (0.71–4.52) 1.99 (0.75–5.28) no evidence in the present results, however, Sum of PCBs 5.29 7.86 10.92 19.48 1.01 (0.99–1.04) 1.01 (0.98–1.04) that estrogenic activity is a risk factor. The UPD-GT inducers 1.44 2.15 3.12 6.19 1.03 (0.96–1.11) 1.01 (0.92–1.11) lower-molecular-weight congeners have greater Wolff groups estrogenic activity than the higher-molecular- 1A 0.00 0.00 0.00 0.51 1.11 (0.81–1.52) 0.80 (0.42–1.52) weight congeners (Wolff et al. 1997), yet it 2A 1.78 2.61 3.75 8.06 1.02 (0.97–1.07) 1.00 (0.94–1.07) was the higher-molecular-weight congeners 2B 1.22 1.75 2.66 4.41 1.07 (0.94–1.21) 1.07 (0.94–1.22) that had the greatest, although nonsignificant, 3 1.59 2.30 3.49 5.59 1.08 (0.98–1.18) 1.09 (0.98–1.20) ORs. In addition, it has been noted that the TEF Groups estrogenic potency of PCBs is extremely low Mono-ortho 1.19 1.93 2.85 6.13 1.02 (0.96–1.10) 0.99 (0.91–1.09) compared with naturally occurring estrogens, Di-ortho 0.38 0.73 1.21 1.97 1.09 (0.87–1.38) 1.16 (0.92–1.48) so exposure may have little ability to affect the Molecular weight groups in utero estrogenic milieu (Safe 1995). In addi- Low 2.31 3.36 4.68 9.98 1.02 (0.97–1.06) 1.00 (0.95–1.06) tion to estrogenic effect, there is evidence that High 2.8 4.21 6.32 10.27 1.04 (0.98–1.09) 1.04 (0.98–1.10) PCBs also may have antian drogenic effects, Results are from conditional logistic regression, conditioned on study center and adjusted for serum DDE level (5 strata), although the data are few and inconsistent triglycerides, and cholesterol. OR is change in lnOR per mmol/L increase in PCB. (Ulbrich and Stahlmann 2004). Reduced anogenital distance has been reported, as Table 4. Adjusted ORs (95% CIs) for cryptorchidism and hypospadias in relation to total PCB level in mater- has interference by PCB-138 with androgen nal serum, CPP, 1959–1965. receptor–mediated effects (Bonefeld-Jorgensen Control Cryptorchism Hypospadias et al. 2001; Portigal et al. 2002; Schrader and a a Total PCBs (ug/L) (no.) No. Adjusted OR (95% CI) No. Adjusted OR (95% CI) Cooke 2003). Coplanar PCBs may also have 0–1.9 180 55 1.00 41 1.00 antiandrogenic effects (Mocarelli et al. 2008). 2–2.9 171 72 1.27 (0.88–1.83) 67 1.57 (1.05–2.34) Although no coplanar PCBs were measured in 3–3.9 111 48 1.32 (0.86–2.02) 38 1.45 (0.90–2.34) the present study, their levels generally corre- ≥ 4.0 131 55 1.41 (0.90–2.20) 55 1.69 (1.06–2.68) late well with sum of PCB levels (Gladen et al. p-Trend 0.19 0.08 1999; Longnecker et al. 2000). If PCBs are able to alter the maternal a b Adjusted for serum DDE concentration (five categories), triglycerides, and cholesterol. Ordinal test across four categories, hormonal milieu, it is possible that they using the median value within each group. | | Environmental Health Perspectives • volume 117 number 9 September 2009 1475 McGlynn et al. cryptorchidism, can be problematic. A sin- Brock JW, Burse VW, Ashley DL, Najam AR, Green VE, Mol NM, Sorensen N, Weihe P, Andersson AM, Jorgensen N, Korver MP, et al. 1996. An improved analysis for chlori- Skakkebaek NE, et al. 2002. Spermaturia and serum hor- gle determination of testicular descent in the nated pesticides and polychlorinated biphenyls (PCBs) mone concentrations at the age of puberty in boys prena- delivery room is not always accurate. Testes in human and bovine sera using solid-phase extraction. tally exposed to polychlorinated biphenyls. Eur J Endocrinol descend late in the third trimester and are not J Anal Toxicol 20(7):528–536. 146(3):357–363. Brucker-Davis F, Wagner-Mahler K, Delattre I, Ducot B, Myrianthopoulos NC, Chung CS. 1974. Congenital malforma- uncommonly undescended at birth. In addi- Ferrari P, Bongain A, et al. 2008. Cryptorchidism at birth tions in singletons: epidemiologic survey. Report from the tion, it is now generally acknowledged that in Nice area (France) is associated with higher prenatal Collaborative Perinatal Project. Birth Defects Orig Artic “acquired undescended testes” does occur, exposure to PCBs and DDE, as assessed by colostrum Ser 10(11):1–58. concentrations. Hum Reprod 23(8):1708–1718. Myrianthopoulos NC, French KS. 1968. An application of the although the age at occurrence remains a mat- Bush B, Bennett AH, Snow JT. 1986. Polychlorobiphenyl con- U.S. Bureau of the Census socioeconomic index to a large, ter of some debate (Barthold 2008). However, geners, p,p’-DDE, and sperm function in humans. Arch diversified patient population. Soc Sci Med 2(3):283–299. the CPP examined the children systematically Environ Contam Toxicol 15(4):333–341. Niswander KR, Gordon M. 1972. The Women and Their Chevrier J, Eskenazi B, Bradman A, Fenster L, Barr DB. 2007. Pregnancies: The Collaborative Perinatal Study of the over time to ascertain congenital anomalies Associations between prenatal exposure to polychlori- National Institute of Neurological Diseases and Stroke. (Myrianthopoulos and Chung 1974). As a nated biphenyls and neonatal thyroid-stimulating hormone Philadelphia:Saunders. result, the rates of both cryptorchidism and levels in a Mexican-American population, Salinas Valley, Noren K. 1988. Changes in the levels of organochlorine pesti- hypospadias reported in the CPP were higher California. Environ Health Perspect 115:1490–1496. cides, polychlorinated-biphenyls, dibenzo-para-dioxins and Cok I, Donmez MK, Satiroglu MH, Aydinuraz B, Henkelmann B, dibenzofurans in human-milk from Stockholm, 1972–1985. (Myrianthopoulos and Chung 1974) than Shen H, et al. 2008. Concentrations of polychlorinated Chemosphere 17(1):39–49. they were other U.S. studies (Paulozzi 1999). dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans Paulozzi LJ. 1999. International trends in rates of hypospadias For the present study, the ratio of hypospadias (PCDFs), and dioxin-like PCBs in adipose tissue of infertile and cryptorchidism. Environ Health Perspect 107:297–302. men. Arch Environ Contam Toxicol 55(1):143–152. Portigal CL, Cowell SP, Fedoruk MN, Butler CM, Rennie PS, to cryptorchidism was somewhat higher Dallinga JW, Moonen EJ, Dumoulin JC, Evers JL, Geraedts JP, Nelson CC. 2002. Polychlorinated biphenyls interfere with than might be expected because we included Kleinjans JC. 2002. Decreased human semen quality androgen-induced transcriptional activation and hormone hypospadias diagnosed up until 7 years of age, and organochlorine compounds in blood. Hum Reprod binding. Toxicol Appl Pharmacol 179(3):185–194. 17(8):1973–1979. Richthoff J, Rylander L, Jonsson BA, Akesson H, Hagmar L, whereas we excluded cryptorchid cases diag- Ensslen SC, Riedel HH, Bluthgen H, Heeschen W. 1990. Nilsson-Ehle P, et al. 2003. Serum levels of 2,2’,4,4’,5,5’- nosed after 1 year. A final limitation of the Chlorinated hydrocarbons in seminal plasma and male hexachlorobiphenyl (CB-153) in relation to markers of study was that the medical records did not note fertility [in German]. Zentralbl Gynakol 112(13):817–821. reproductive function in young males from the general Giwercman YL, Kleist KE, Giwercman A, Giwercman C, Toft G, Swedish population. Environ Health Perspect 111:409–413. degree of hypospadias among the boys affected. Bonde JP, et al. 2006. Remarkably low incidence of hypos- Rignell-Hydbom A, Rylander L, Giwercman A, Jonsson BA, In conclusion, the results of the present padias in Greenland despite high exposure to endocrine Nilsson-Ehle P, Hagmar L. 2004. Exposure to CB-153 and study do not support an association between disrupters: possible protective effect of androgen receptor p,p’-DDE and male reproductive function. Hum Reprod genotype. Pharmacogenet Genomics 16(5):375–377. 19(9):2066–2075. in utero PCB exposure and either cryptorchid- Gladen BC, Longnecker MP, Schecter AJ. 1999. Correlations Ross G. 2004. The public health implications of polychlorinated ism or hypospadias. PCBs are just one type among polychlorinated biphenyls, dioxins, and furans in biphenyls (PCBs) in the environment. Ecotoxicol Environ of environmental endocrine modulator, how- humans. Am J Ind Med 35(1):15–20. Saf 59(3):275–291. ever. Whether other endocrine modulators Hardell L, van Bavel B, Lindstrom G, Carlberg M, Dreifaldt AC, Rozati R, Reddy PP, Reddanna P, Mujtaba R. 2002. Role of envi- Wijkstrom H, et al. 2003. Increased concentrations of poly- ronmental estrogens in the deterioration of male factor are related to risk is unclear and should be chlorinated biphenyls, hexachlorobenzene, and chlor- fertility. Fertil Steril 78(6):1187–1194. examined in other studies. danes in mothers of men with testicular cancer. Environ Safe SH. 1995. Environmental and dietary estrogens and human Health Perspect 111:930–934. health: is there a problem? Environ Health Perspect Hardy JB. 2003. The Collaborative Perinatal Project: lessons 103:346–351. Refe Rences and legacy. Ann Epidemiol 13(5):303–311. Schrader TJ, Cooke GM. 2003. Effects of Aroclors and individ- Hauser R, Chen Z, Pothier L, Ryan L, Altshul L. 2003. The relation- ual PCB congeners on activation of the human androgen Ahlborg UG, Becking GC, Birnbaum LS, Brouwer A, Derks HJGM, ship between human semen parameters and environmen- receptor in vitro. Reprod Toxicol 17(1):15–23. Feeley M, et al. 1994. Toxic equivalency factors for dioxin- tal exposure to polychlorinated biphenyls and p,p’-DDE. Sharpe RM. 2003. The “oestrogen hypothesis”—where do we like PCBs: Report on a WHO-ECEH and IPCS consultation, Environ Health Perspect 111:1505–1511. stand now? Int J Androl 26(1):2–15. December 1993. Chemosphere 28(6):1049–1067. Hosie S, Loff S, Witt K, Niessen K, Waag KL. 2000. Is there Skakkebaek NE, Rajpert-De Meyts E, Main KM. 2001. Testicular Ballschmiter K, Bacher R, Mennel A, Fischer R, Riehle U, Swerev a correlation between organochlorine compounds and dysgenesis syndrome: an increasingly common develop- M. 1992. The determination of chlorinated biphenyls, chlo- undescended testes? Eur J Pediatr Surg 10(5):304–309. mental disorder with environmental aspects. Hum Reprod rinated dibenzodioxins, and chlorinated dibenzofurans by Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, 16(5):972–978. GC-MS. HRC CC J High Resolut Chromatogr Chromatogr Needham LL, et al. 2002. Maternal serum level of 1,1-dichloro- Toppari J, Larsen JC, Christiansen P, Giwercman A, Grandjean P, 15(4):260–270. 2,2-bis(p-chlorophenyl)ethylene and risk of cryptorchidism, Guillette LJ Jr, et al. 1996. Male reproductive health and Barthold JS. 2008. Undescended testis: current theories of hypospadias, and polythelia among male offspring. Am J environmental xenoestrogens. Environ Health Perspect etiology. Curr Opin Urol 18(4):395–400. Epidemiol 155(4):313–322. 104(suppl 4):741–803. Barthold JS, Gonzalez R. 2003. The epidemiology of congenital Longnecker MP, Ryan JJ, Gladen BC, Schecter AJ. 2000. Ulbrich B, Stahlmann R. 2004. Developmental toxicity of poly- cryptorchidism, testicular ascent and orchiopexy. J Urol Correlations among human plasma levels of dioxin-like chlorinated biphenyls (PCBs): a systematic review of 170(6 pt 1):2396–2401. compounds and polychlorinated biphenyls (PCBs) and experimental data. Arch Toxicol 78(5):252–268. Bonde JP, Toft G, Rylander L, Rignell-Hydbom A, Giwercman A, implications for epidemiologic studies. Arch Environ Health Weiss JM, Bauer O, Bluthgen A, Ludwig AK, Vollersen E, Spano M, et al. 2008. Fertility and markers of male repro- 55(3):195–200. Kaisi M, et al. 2006. Distribution of persistent organochlo- ductive function in Inuit and European populations spanning Lunden A, Noren K. 1998. Polychlorinated naphthalenes and rine contaminants in infertile patients from Tanzania and large contrasts in blood levels of persistent organo chlorines. other organochlorine contaminants in Swedish human milk, Germany. J Assist Reprod Genet 23(9–10):393–399. Environ Health Perspect 116:269–277. 1972–1992. Arch Environ Contam Toxicol 34(4):414–423. Wolff MS, Camann D, Gammon M, Stellman SD. 1997. Proposed B o n e f e l d - J o r g e n s e n E C , A n d e r s e n H R , R a s m u s s e n T H , Mocarelli P, Gerthoux PM, Patterson DG Jr, Milani S, Limonta G, PCB congener groupings for epidemiological studies. Vinggaard AM. 2001. Effect of highly bioaccumulated poly- Bertona M, et al. 2008. Dioxin exposure, from infancy Environ Health Perspect 105:13–14. chlorinated biphenyl congeners on estrogen and androgen through puberty, produces endocrine disruption and affects receptor activity. Toxicology 158(3):141–153. human semen quality. Environ Health Perspect 116:70–77. | | 1476 volume 117 number 9 September 2009 • Environmental Health Perspectives http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Environmental Health Perspectives Pubmed Central

Maternal Pregnancy Levels of Polychlorinated Biphenyls and Risk of Hypospadias and Cryptorchidism in Male Offspring

Environmental Health Perspectives , Volume 117 (9) – Apr 20, 2009

Loading next page...
 
/lp/pubmed-central/maternal-pregnancy-levels-of-polychlorinated-biphenyls-and-risk-of-D4bxiW6PJn

References (44)

Publisher
Pubmed Central
ISSN
0091-6765
eISSN
1552-9924
DOI
10.1289/ehp.0800389
Publisher site
See Article on Publisher Site

Abstract

Research Children’s Health Maternal Pregnancy Levels of Polychlorinated Biphenyls and Risk of Hypospadias and Cryptorchidism in Male Offspring 1 2 1 3 4 Katherine A. McGlynn, Xuguang Guo, Barry I. Graubard, John W. Brock, Mark A. Klebanoff, and Matthew P. Longnecker Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and 2 3 Human Services, Bethesda, Maryland, USA; Westat, Inc., Durham, North Carolina, USA; National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; Division of Epidemiology, Statistics, and Prevention Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, 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 and 1965 at 12 U.S. medical centers located in Background : The etiologies of the male urogenital anomalies cryptorchidism and hypospadias are Baltimore, Maryland; Boston, Massachusetts; poorly understood. It has been suggested, however, that in utero hormone levels may be related to Buffalo, New York; Memphis, Tennessee; risk. Endocrine-disrupting chemicals, including polychlorinated biphenyl (PCB) compounds, may Minneapolis, Minnesota; New Orleans, alter hormone levels and thereby affect the fetus. Louisiana; New York City (two centers); o Bjectives : To examine whether in utero PCB exposure is related to cryptorchidism and hypo- Philadelphia, Pennsylvania; Portland, Oregon; spadias, we examined PCB levels among pregnant women enrolled in the Collaborative Perinatal Providence, Rhode Island; and Richmond, Project (CPP). Virginia. Eleven centers recruited participants Methods : The CPP enrolled pregnant women at 12 U.S. medical centers between 1959 and 1965. from the prenatal clinics of a university hos- For the present research, we analyzed third-trimester serum samples from the mothers of 230 sons pital, whereas one (Buffalo) recruited from with cryptorchidism, 201 sons with hypospadias, and 593 sons with neither condition. We esti- 13 private obstetric practices. Participant mated adjusted odds ratios (ORs) and 95% confidence intervals (CIs) using logistic regression and selection varied among the medical centers. examined the associations of each anomaly with individual PCB congener levels, sum of PCBs, and Women were ineligible if they were incarcer- several functional groupings of PCBs. ated, planned to leave the area on delivery, or results : In general, the ORs for cryptorchidism or hypospadias showed no notable associations planned to place their child for adoption or if with individual PCB congener levels or functional groupings of PCBs. However, the ORs and 95% they delivered on the day they were recruited CIs for the sum of PCBs associated with hypospadias were as follows: 0–1.9 µg/L, reference group; for the study. The characteristics of the women 2–2.9 µg/L, OR = 1.57, 95% CI, 1.05–2.34; 3–3.9 µg/L, OR = 1.45, 95% CI, 0.90–2.34; and in the sample were, at registration, essen- ≥ 4.0 µg/L, OR = 1.69, 95% CI, 1.06–2.68; p-value for trend = 0.08. tially the same as those in the sampling frame conclusions : Given the large number of associations examined, these findings do not strongly (Niswander and Gordon 1972). Four per- support the hypothesis that PCBs are associated with cryptorchidism or hypospadias. Because cent of the participants were lost to follow-up population serum PCB levels at the time of sample collection were considerably higher than levels before delivery. Once enrolled, the women’s at present, it is unlikely that current PCB exposure is related to the development of either anomaly. nonfasting blood was drawn approximately k e y w o r d s : cryptorchidism, hypospadias, polychlorinated biphenyls, testicular dysgenesis syn- every 8 weeks during pregnancy, at delivery, drome. Environ Health Perspect 117:1472–1476 (2009). doi:10.1289/ehp.0800389 available via and at 6 weeks postpartum. Sera were stored http://dx.doi.org/ [Online 20 April 2009] in glass vials at –20°C with no recorded thaws. Approximately 42,000 women were enrolled, The hormone hypothesis of male reproductive since the 1960s but remain detectable in most and 55,000 children were born in the study. disorders suggests that both endogenous and segments of the population (Ross 2004). To The children were systematically assessed exogenous hormones might be risk factors date, only four published studies have exam- for the presence of birth defects and other for cryptorchidism, hypospadias, impaired ined the association between PCBs and either spermatogenesis, and testicular cancer (Sharpe cryptorchidism or hypospadias. Two case– Address correspondence to K.A. McGlynn, Hormonal 2003). Exogenous hormonal exposures involve control studies of PCBs and cryptorchidism and Reproductive Epidemiology Branch, Division of a variety of endocrine-disrupting chemicals reported no relationship with risk (Hosie Cancer Epidemiology and Genetics, National Cancer Institute, NIH, EPS Suite 550, 6120 Executive Blvd., whose effects range from estrogenic to anti - et al. 2000; Mol et al. 2002), whereas a third Rockville, MD 20852-7234 USA. Telephone: (301) estrogenic, androgenic, and antiandrogenic. offered measured support (Brucker-Davis 435-4918. Fax: (301) 402-0416. E-mail: mcglynnk@ Evidence in support of the hypothesis has et al. 2008). One ecologic study of PCBs and mail.nih.gov consisted largely of animal data (Toppari et al. hypospadias reported an inverse association Support for this research was provided by the 1996) and temporal trends (Paulozzi 1999) in (Giwercman et al. 2006). Because the prior Intramural Research Programs of the National Cancer the prevalence of the so-called testicular dys- studies may have been too small to detect Institute, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, genesis syndrome (TDS) conditions. Human statistically significant differences, the present and the National Institute of Environmental studies of the relationship between endocrine- study was conducted among a large, well- Health Sciences of the National Institutes of Health disrupting chemicals and the TDS disorders described population in which the serum (NIH) and by the Centers for Disease Control and have been relatively few. samples were collected at a time when PCB Prevention (CDC). Polychlorinated biphenyl (PCB) com- levels in the United States were higher. The findings in this study are those of the authors pounds are ubiquitous, stable, environmental and do not necessarily reflect the official opinion of Materials and Methods the NIH or the CDC. pollutants that were widely used in devel- X.G. is employed by Social and Scientific Systems, oped countries between 1929 and 1977. Even Study population. The Collaborative Perinatal Inc., which has no PCB-related contracts. The other after they were banned, PCBs continued Project (CPP) was a prospective study of authors declare they have no competing financial to be used in “closed” source applications neurologic disorders and other conditions interests. such as electrical capacitors and transform- in children (Niswander and Gordon 1972). Received 11 November 2008; accepted 20 April ers. Levels of PCBs in humans have declined Pregnant women were enrolled between 1959 | | 1472 volume 117 number 9 September 2009 • Environmental Health Perspectives PCBs and risk of hypospadias and cryptorchidism outcomes at birth and through 7 years of age. Laboratory assays. Serum levels of 11 PCBs as continuous variables and serum p,p´-dichloro- Follow-up to 7 years of age was completed for were measured at the Centers for Disease diphenyldichloroethylene (p,p´-DDE) as a five- about 75% of children born into the study. Control and Prevention after solid-phrase stratum categorical variable. For the present study, we employed a extraction cleanup and dual-column gas chro- We evaluated confounding by comparing nested case–control design to examine the matography using electron capture detection the hazard ratio of the baseline model (includ- association between maternal serum PCB (Brock et al. 1996). The congeners, designated ing sum of PCBs, triglycerides, cholesterol, and levels and risk of cryptorchidism and hypo- by International Union of Pure and Applied p,p´-DDE) to the hazard ratio of the model spadias among sons. To be eligible for inclu- Chemistry code, were PCB-28, PCB-52, that also included the possible confounder. We sion in the present study, the mother must PCB-74, PCB-105, PCB-118, PCB-138, modeled sum of PCBs both as a continuous have given birth to a singleton, live-born male PCB-153, PCB-170, PCB-180, PCB-194, and variable and as a categorical variable and used infant and must have had a 3-mL aliquot of PCB-203. The proportion of PCBs in each results of both analyses to determine whether third-trimester maternal serum available. We sample recovered by extraction was approxi- the factor was a confounder. If the hazard ratio defined cryptorchidism as having had a diag - mately 60%. The results shown are unadjusted per microgram per liter of PCB or the haz- nosis of undescended testis(es) at any time for recovery. The between-assay coefficient of ard ratio for the contrast of highest-to-lowest during the first year of life. The diagnosis was variation was 19% at 3.49 µg PCB. A labora- PCB strata changed by 15% or more, the fac- made by pediatricians based on serial exami- tory result was not obtained for 3% of samples tor was considered a confounder. The factors nations that included inspection and palpa- because the measured value did not meet the considered as possible confounders were race, tion of the genitalia. Boys first noted in the quality-control standards for acceptance. Thus, season of birth, maternal age, maternal parity, medical record to have undescended testis(es) the PCB results were not available for 30 sam- socioeconomic index, prepregnancy body mass after the first year of life were not considered ples. Because PCBs are lipid soluble and because index, weight gain during pregnancy, smoking cryptorchid because they may have had retrac- serum contains a variable amount of lipid, we during pregnancy, hyperemesis gravidarum, tile testes. Hypospadias was defined as having measured serum cholesterol and triglycerides gestational hypertension, age at menarche, his- a diagnosis any time during the first 7 years of using standing enzymatic assays. tory of infertility, menstrual cycle irregularity, life. Degree of hypospadias was not noted in In addition to analyzing each congener estrogen use during pregnancy, and progester- the medical records. sepa rately, we examined four groupings of con- one use during pregnancy. Similarly, we con- Among the 28,444 boys in the CPP, 267 geners. The Wolff groupings (Wolff et al. 1997) sidered the effects of adjustment for preterm were not live born and 441 were not singletons. were as follows: Wolff group 1A, potentially birth, birth weight, placental weight, and small- No blood sample was available from the moth- estrogenic, weak phenobarbital inducers, not for-gestational-age even though these were ers of 5,389 boys. Among the eligible 22,347 persistent congeners (PCB-52); Wolff group potentially intermediate variables. None of the boys, there were 241 cases of cryptorchidism 2A, potentially antiestrogenic and immuno- factors considered, however, changed the hazard and 214 cases of hypospadias. The five boys toxic, dioxin-like, non-ortho– and mono-ortho– ratios by ≥ 15%. Using a similar approach, but with both cryptorchidism and hypospadias substituted, moderately persistent congeners with cross-product terms, we evaluated effect were included in each group for the analysis. (PCB-74, PCB-105, PCB-118); Wolff group modification by maternal age, race, smoking, From the pool of eligible mothers, 599 were 2B, limited dioxin activity, di-ortho substituted, prepregnancy body mass index, triglycerides, selected at random so that the control:case ratio persistent congeners (PCB-138, PCB-170) and cholesterol, serum p,p´-DDE, gestational hyper- would be > 2:1 for each condition. Wolff group 3, phenobarbital inducers, CYP1A tension, socioeconomic index, and study center. Among the boys diagnosed as cryptorchid and CYP2B inducers, and persistent congeners We supplemented evaluation of effect modifica - during the first year of life, the study records (PCB-153, PCB-180, PCB-203). The toxic tion by categorical variables with more than two indicated that the testicles were descended at equivalency factor (TEF) groupings (Ahlborg categories by comparing the model fit statistics birth in 103, suggesting that these boys may et al. 1994) were TEF mono-ortho (PCB- for models with and without the cross-product have had acquired, rather than congenital, 105, PCB-118), and TEF di-ortho (PCB-170, terms. If the p-value associated with the inter- undescended testis (Barthold and Gonzalez PCB-180). The uridine diphosphate-glucurono - action term based on the likelihood ratio test 2003). To evaluate whether a relationship syl transferase (UDP-GT) inducer, cytochrome had a value ≤ 0.10, the degree of potential effect with PCBs varied by type of cryptorchidism, P450 1A (CYP1A), and CYP2B inducer group- modification was further considered by examin - we considered the boys with testis descended ing (Chevrier et al. 2007) was PCB-118 and ing tables stratie fi d by the potentially modifying at birth separately in a sensitivity analysis. PCB-180. The molecular-weight groups were factor(s). We further evaluated sum of PCBs Among the boys with undescended testicle(s) low (PCB-28, PCB-52, PCB-74, PCB-105, by forming ordinal categories and including at birth (n = 138), all but one also had a sub- PCB-118) and high (PCB-138, PCB-153, corresponding dummy variables in the logistic sequent observation of cryptorchidism in at PCB-170, PCB-180, PCB-194, PCB-203). regression analysis to estimate adjusted odds least one of the three subsequent examination Statistical methods. We converted PCBs ratios (ORs). We performed trend tests across (ages 4 months, 1 year, 7 years) or documen- to molar concentrations (PCB in micrograms categories by including a single independent tation of orchidopexy. per liter divided by the congener’s molecu- variable with integer scores corresponding to The socioeconomic index calculated lar weight) and then analyzed on a continu- the categories testing the significance of the for participants in the CPP was the mean ous scale. Five congeners had missing values regression coefficient. All statistical analyses of the three percentile scores: education of (PCB-28, n = 2; PCB-74, n = 26; PCB-118, were conducted using the SAS statistical soft- head of the household, occupation of head n = 42; PCB-138, n = 22; PCB-180, n = 2), ware package, version 9.1 (SAS Institute Inc., of the household or chief wage earner, and mainly because the measured value did not meet Cary, NC). All p-values were two-sided. family income. The score used to calculate the quality control standards for acceptance Results the percentile for an occupation was based (Ballschmiter et al. 1992). We estimated the on the percentiles of education and income change in log odds of having cryptorchidism or Based on the distribution of the controls, the among persons with the same occupation hypospadias per millimole increase in PCB level study population was 46% white, 48% black, (Myrianthopoulos and French 1968). Verbal using conditional logistic regression, condi- and 5.6% other racial/ethnic groups (primarily consent to participate was elicited from all tioned on study center (12 strata). Models were Hispanic and Asian) (Table 1). Birth weight mothers (Hardy 2003). adjusted for serum triglycerides and cholesterol and gestational age medians were typical of | | Environmental Health Perspectives • volume 117 number 9 September 2009 1473 McGlynn et al. healthy births. The mothers were a median age previous report has described the study popula- cryptorchidism, the ORs ranged from 1.03 of 22 years and had a median socioeconomic tion in detail (Longnecker et al. 2002). (PCB-118) to 1.79 (PCB-203), and none index (mean of three percentile scores: educa- Table 2 shows median maternal PCB lev- were statistically significant. For hypospadias, tion of head of household, occupation of head els, lipid levels, and DDE levels for each of the ORs ranged from 0.80 (PCB-52) to 1.99 of household or chief wage earner, and family the study groups. The median level of sum (PCB-203), and as with cryptorchidism, none income) of 4.5, which was just under of the of PCBs was slightly lower in the controls were statistically significant. median of 5.0 for the United States population (2.7 µg/L), whereas the median levels of Table 3 also shows the adjusted ORs for in the 1960s. About 31% of the mothers were p,p´-DDE (24.5 µg/L) and total cholesterol groupings of PCBs. We found no significant primiparas. Compared with the control boys, (234 µg/L) were slightly higher. The median increase in risk of either cryptorchidism or the boys with cryptorchidism and hypospadias level of triglycerides in the controls (204 µg/L) hypospadias with increasing level of sum of were more likely to be white, to be born pre- was intermediate to the two case groups. PCBs [cryptorchidism: OR = 1.01, 95% con- maturely, and to be born small for gestational Table 3 shows the adjusted ORs for the fidence interval (CI), 0.99–1.04; hypospadias: age (birth weight below 10th percentile). The PCB congener-specific analyses, conditioned OR = 1.01, 95% CI, 0.98–1.04]. Similarly, we boys with hypospadias also had a lower median on study center and adjusted for triglyc- found no significant increases in risk of either birth weight than did the control boys. A erides, cholesterol, and p,p´-DDE level. For outcome with any of the Wolff groupings (Wolff et al. 1997), TEF groupings (Ahlborg Table 1. Characteristics of mothers and sons according to the son’s case–control status, CPP, 1959–1965. et al. 1994), or enzyme inducer (Chevrier Characteristic Cryptorchidism (n = 230) Hypospadias (n = 201) Control (n = 593) et al. 2007) groupings. Because the ORs in the congener-specific analysis tended to be higher Race (%) among the higher-molecular-weight congeners, White 57.0 49.3 46.0 Black 41.3 44.8 48.4 we also analyzed PCBs stratified by molecular Other 1.7 6.0 5.6 weight. The analyses did not detect statistically Gestation (week) significant associations with either cryptorchid - Median (Q1, Q3) 39 (38, 41) 39 (38, 41) 39 (38, 40) ism (low: OR = 1.02, 95% CI, 0.97–1.06; Preterm birth (%) 16.6 19.5 14.0 high: OR = 1.04, 95% CI, 0.98–1.09) or hypo- Birth weight (g) spadias (low: OR = 1.00, 95% CI, 0.95–1.06; Median (Q1, Q3) 3,260 (2,835, 3,629) 3,147 (2,665, 3,487) 3,260 (2,948, 3,600) high: OR = 1.04, 95% CI, 0.98–1.10). Small for gestational age (%) 9.1 17.8 4.9 Maternal age (years) Because the examinations of cryptorchid- Median (Q1, Q3) 24 (21, 30) 24 (20, 29) 22 (20, 28) ism and hypospadias with the sum of PCBs Previous live births (%) approached statistical significance, we also 0 27.0 30.8 30.9 analyzed risk by PCB category (Table 4). 1 22.2 21.4 22.5 Cryptorchidism was not significantly related ≥ 2 50.9 47.8 46.6 to PCB level in any category, and the test for Socioeconomic index trend was not statistically significant (p for Median (Q1, Q3) 4.7 (3.3, 6.3) 4.3 (3.0, 6.2) 4.5 (3.3, 6.0) trend = 0.19). The results of a similar analysis Prepregnancy body mass index Median (Q1, Q3) 22.2 (20.3, 25.0) 21.8 (19.3, 24.1) 22.2 (20.0, 24.9) that modeled sum of PCBs on a lipid basis Gestational hypertension (%) 5.9 6.5 6.4 in quintiles produced the same conclusions Study Center (%) (data not shown). Although the trend in risk Boston, MA 31.3 25.4 23.6 of hypospadias approached significance ( p for Buffalo, NY 7.0 3.5 3.5 trend = 0.08) and the ORs for all categories New Orleans, LA 6.5 3.5 4.7 a were greater than unity, we found, overall, no New York City, NY 3.0 4.5 3.2 linear relationship. The examination of effect Baltimore, MD 6.1 6.5 7.6 Richmond, VA 7.0 5.5 5.7 modification showed that no factor signifi- Minneapolis, MN 3.5 4.0 5.2 cantly altered the results of the baseline models New York City, NY 1.7 6.0 7.8 (data not shown). A final examination of cryp - Portland, OR 4.3 5.5 7.1 torchidism excluded the boys (n = 103) whose Philadelphia, PA 14.8 23.9 18.5 testicles were initially descended at birth. Providence, RI 11.3 10.0 5.9 The results of that analysis (data not shown), Memphis, TN 3.5 2.0 7.1 however, did not differ from the analysis that a b Columbia-Presbyterian Medical Center. New York Medical College included all boys with cryptorchidism. Table 2. Maternal serum values by son’s case–control status, CCP, 1959–1965. Discussion Cryptorchidism (n = 230) Hypospadias (n = 201) Control (n =593) The results of the present study do not support Total cholesterol (ug/L) the hypothesis that in utero PCB exposure Median 231 229 234 increases the risk of hypospadias or cryp- (Q1, Q3) (190, 272) (190, 275) (195, 279) torchidism. Because the present study samples Triglycerides (ug/L) were collected in the 1960s, when PCB expo- Median 207.5 187 204 sure was substantially greater than at present, (Q1, Q3) (163, 258) (152, 253) (159, 259) the results also suggest that current low-level DDE (ug/L) PCB exposure is unlikely to be related to the Median 23.6 23.9 24.5 (Q1, Q3) (15.9, 35.3) (16.2, 34.4) (16.7, 37.2) development of either condition. Sum of PCBs with imputed congener (ug/L) With the inclusion of > 1,000 participants, Median 2.8 2.9 2.7 the present study is by far the largest to have (Q1, Q3) (2.0, 3.9) (2.1, 4.2) (1.8, 3.8) examined the relationship between PCBs and | | 1474 volume 117 number 9 September 2009 • Environmental Health Perspectives PCBs and risk of hypospadias and cryptorchidism cryptorchidism or hypospadias. Although the would also be related to the TDS disorders to affect fertility or to have direct hormone- prior literature is not extensive, three studies (Skakkebaek et al. 2001) that become evident like activity (Bonde et al. 2008). To date, the of cryptorchidism and PCBs and one study of in adulthood: impaired spermatogenesis and sole published study of testicular cancer and hypospadias and PCBs have been published. testicular cancer. More studies have examined PCBs found no association with sum of PCBs, Mol et al. (2002) examined cord blood PCB the relationship of PCBs with the former than estrogenic PCBs, or enzyme-inducing PCBs levels in a population known to have high the latter. In general, the results of the PCB– (Hardell et al. 2003). Taken as a group, then, at PCB exposure via s fi h consumption. Studying fertility studies are somewhat equivocal. Several the present time, there is little evidence that the 196 Faroe Island boys born between 1986 studies have found statistically significant associ - TDS disorders are related to PCB exposure. and 1987, the investigators found no relation- ations with impaired sperm parameters (Hauser Several potential weaknesses of the present ship between PCB levels and cryptorchidism. et al. 2003; Richthoff et al. 2003; Rozati et al. study merit consideration. A first consideration A similar conclusion was reached by Hosie 2002), whereas others have found no associa- is that the assays were run on stored, rather et al. (2000). Studying adipose PCB levels in tion (Weiss et al. 2006) or associations only in than fresh, serum samples. PCB levels, how- 18 German boys with cryptorchidism and subsets of their populations (Bush et al. 1986; ever, are quite stable over time, as has been 30 control boys, the investigators found no dif- Dallinga et al. 2002; Rignell-Hydbom et al. demonstrated in samples stored for 15 years ference in PCB levels. A third study examined 2004). In contrast, several studies have reported (Lunden and Noren 1998; Noren 1988). In PCB levels in both cord blood and breast milk direct associations between PCB levels and fer- addition, the cholesterol and triglyceride levels, in a French population (Brucker-Davis et al. tility (Cok et al. 2008; Ensslen et al. 1990). as reported in Table 2, were in the expected 2008). The comparison of 78 boys diagnosed The summary of the international INUENDO range, suggesting that substantial degradation at birth with cryptorchidism and 86 control study of PCBs and fertility in four populations, had not occurred. A second potential weakness boys found a significant association with breast however, concluded that PCBs did not appear is that determination of one of the outcomes, milk PCB levels but not with cord blood PCB Table 3. Adjusted ORs (95% CIs) for cryptorchidism and hypospadias in relation to maternal PCB levels levels. When they restricted the analysis to the (mmol/L), CPP, 1959–1965. boys who remained cryptorchid at 3 months, Percentiles among controls Cryptorchidism Hypospadias however, the relationship with breast milk PCB a a levels was no longer significant. The only prior 25th 50th 75th 95th Adjusted OR (95% CI) Adjusted OR (95% CI) study of hypospadias and PCB levels reported PCB congener to date was an ecologic study in Greenland. PCB-28 0.39 0.70 0.97 1.79 1.07 (0.86–1.34) 1.07 (0.86–1.32) The investigators found a low prevalence rate of PCB-52 0.00 0.00 0.00 0.51 1.11 (0.81–1.52) 0.80 (0.42–1.52) hypospadias despite high levels of PCBs in the PCB-74 0.51 0.79 1.13 2.16 1.04 (0.87–1.25) 1.03 (0.84–1.26) population (Giwercman et al. 2006), suggest- PCB-105 0.00 0.34 0.52 1.19 1.11 (0.82–1.50) 1.03 (0.69–1.53) PCB-118 1.13 1.65 2.39 5.21 1.03 (0.95–1.12) 0.99 (0.88–1.11) ing that PCBs might be inversely, rather than directly, associated with risk of hypospadias. PCB-138 1.16 1.61 2.36 3.85 1.09 (0.94–1.25) 1.08 (0.93–1.26) PCB-153 1.22 1.69 2.52 4.02 1.11 (0.97–1.27) 1.12 (0.98–1.30) PCBs have been suspected of being related to male urogenital anomalies princi- PCB-170 0.00 0.23 0.38 0.66 1.07 (0.59–1.94) 1.34 (0.73–2.46) PCB-180 0.35 0.56 0.83 1.39 1.19 (0.84–1.69) 1.28 (0.89–1.84) pally because some congeners and metabo- PCB-194 0.00 0.00 0.21 0.37 1.42 (0.54–3.72) 1.41 (0.48–4.19) lites display weak estrogenic activity. There is PCB-203 0.00 0.00 0.21 0.42 1.79 (0.71–4.52) 1.99 (0.75–5.28) no evidence in the present results, however, Sum of PCBs 5.29 7.86 10.92 19.48 1.01 (0.99–1.04) 1.01 (0.98–1.04) that estrogenic activity is a risk factor. The UPD-GT inducers 1.44 2.15 3.12 6.19 1.03 (0.96–1.11) 1.01 (0.92–1.11) lower-molecular-weight congeners have greater Wolff groups estrogenic activity than the higher-molecular- 1A 0.00 0.00 0.00 0.51 1.11 (0.81–1.52) 0.80 (0.42–1.52) weight congeners (Wolff et al. 1997), yet it 2A 1.78 2.61 3.75 8.06 1.02 (0.97–1.07) 1.00 (0.94–1.07) was the higher-molecular-weight congeners 2B 1.22 1.75 2.66 4.41 1.07 (0.94–1.21) 1.07 (0.94–1.22) that had the greatest, although nonsignificant, 3 1.59 2.30 3.49 5.59 1.08 (0.98–1.18) 1.09 (0.98–1.20) ORs. In addition, it has been noted that the TEF Groups estrogenic potency of PCBs is extremely low Mono-ortho 1.19 1.93 2.85 6.13 1.02 (0.96–1.10) 0.99 (0.91–1.09) compared with naturally occurring estrogens, Di-ortho 0.38 0.73 1.21 1.97 1.09 (0.87–1.38) 1.16 (0.92–1.48) so exposure may have little ability to affect the Molecular weight groups in utero estrogenic milieu (Safe 1995). In addi- Low 2.31 3.36 4.68 9.98 1.02 (0.97–1.06) 1.00 (0.95–1.06) tion to estrogenic effect, there is evidence that High 2.8 4.21 6.32 10.27 1.04 (0.98–1.09) 1.04 (0.98–1.10) PCBs also may have antian drogenic effects, Results are from conditional logistic regression, conditioned on study center and adjusted for serum DDE level (5 strata), although the data are few and inconsistent triglycerides, and cholesterol. OR is change in lnOR per mmol/L increase in PCB. (Ulbrich and Stahlmann 2004). Reduced anogenital distance has been reported, as Table 4. Adjusted ORs (95% CIs) for cryptorchidism and hypospadias in relation to total PCB level in mater- has interference by PCB-138 with androgen nal serum, CPP, 1959–1965. receptor–mediated effects (Bonefeld-Jorgensen Control Cryptorchism Hypospadias et al. 2001; Portigal et al. 2002; Schrader and a a Total PCBs (ug/L) (no.) No. Adjusted OR (95% CI) No. Adjusted OR (95% CI) Cooke 2003). Coplanar PCBs may also have 0–1.9 180 55 1.00 41 1.00 antiandrogenic effects (Mocarelli et al. 2008). 2–2.9 171 72 1.27 (0.88–1.83) 67 1.57 (1.05–2.34) Although no coplanar PCBs were measured in 3–3.9 111 48 1.32 (0.86–2.02) 38 1.45 (0.90–2.34) the present study, their levels generally corre- ≥ 4.0 131 55 1.41 (0.90–2.20) 55 1.69 (1.06–2.68) late well with sum of PCB levels (Gladen et al. p-Trend 0.19 0.08 1999; Longnecker et al. 2000). If PCBs are able to alter the maternal a b Adjusted for serum DDE concentration (five categories), triglycerides, and cholesterol. Ordinal test across four categories, hormonal milieu, it is possible that they using the median value within each group. | | Environmental Health Perspectives • volume 117 number 9 September 2009 1475 McGlynn et al. cryptorchidism, can be problematic. A sin- Brock JW, Burse VW, Ashley DL, Najam AR, Green VE, Mol NM, Sorensen N, Weihe P, Andersson AM, Jorgensen N, Korver MP, et al. 1996. An improved analysis for chlori- Skakkebaek NE, et al. 2002. Spermaturia and serum hor- gle determination of testicular descent in the nated pesticides and polychlorinated biphenyls (PCBs) mone concentrations at the age of puberty in boys prena- delivery room is not always accurate. Testes in human and bovine sera using solid-phase extraction. tally exposed to polychlorinated biphenyls. Eur J Endocrinol descend late in the third trimester and are not J Anal Toxicol 20(7):528–536. 146(3):357–363. Brucker-Davis F, Wagner-Mahler K, Delattre I, Ducot B, Myrianthopoulos NC, Chung CS. 1974. Congenital malforma- uncommonly undescended at birth. In addi- Ferrari P, Bongain A, et al. 2008. Cryptorchidism at birth tions in singletons: epidemiologic survey. Report from the tion, it is now generally acknowledged that in Nice area (France) is associated with higher prenatal Collaborative Perinatal Project. Birth Defects Orig Artic “acquired undescended testes” does occur, exposure to PCBs and DDE, as assessed by colostrum Ser 10(11):1–58. concentrations. Hum Reprod 23(8):1708–1718. Myrianthopoulos NC, French KS. 1968. An application of the although the age at occurrence remains a mat- Bush B, Bennett AH, Snow JT. 1986. Polychlorobiphenyl con- U.S. Bureau of the Census socioeconomic index to a large, ter of some debate (Barthold 2008). However, geners, p,p’-DDE, and sperm function in humans. Arch diversified patient population. Soc Sci Med 2(3):283–299. the CPP examined the children systematically Environ Contam Toxicol 15(4):333–341. Niswander KR, Gordon M. 1972. The Women and Their Chevrier J, Eskenazi B, Bradman A, Fenster L, Barr DB. 2007. Pregnancies: The Collaborative Perinatal Study of the over time to ascertain congenital anomalies Associations between prenatal exposure to polychlori- National Institute of Neurological Diseases and Stroke. (Myrianthopoulos and Chung 1974). As a nated biphenyls and neonatal thyroid-stimulating hormone Philadelphia:Saunders. result, the rates of both cryptorchidism and levels in a Mexican-American population, Salinas Valley, Noren K. 1988. Changes in the levels of organochlorine pesti- hypospadias reported in the CPP were higher California. Environ Health Perspect 115:1490–1496. cides, polychlorinated-biphenyls, dibenzo-para-dioxins and Cok I, Donmez MK, Satiroglu MH, Aydinuraz B, Henkelmann B, dibenzofurans in human-milk from Stockholm, 1972–1985. (Myrianthopoulos and Chung 1974) than Shen H, et al. 2008. Concentrations of polychlorinated Chemosphere 17(1):39–49. they were other U.S. studies (Paulozzi 1999). dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans Paulozzi LJ. 1999. International trends in rates of hypospadias For the present study, the ratio of hypospadias (PCDFs), and dioxin-like PCBs in adipose tissue of infertile and cryptorchidism. Environ Health Perspect 107:297–302. men. Arch Environ Contam Toxicol 55(1):143–152. Portigal CL, Cowell SP, Fedoruk MN, Butler CM, Rennie PS, to cryptorchidism was somewhat higher Dallinga JW, Moonen EJ, Dumoulin JC, Evers JL, Geraedts JP, Nelson CC. 2002. Polychlorinated biphenyls interfere with than might be expected because we included Kleinjans JC. 2002. Decreased human semen quality androgen-induced transcriptional activation and hormone hypospadias diagnosed up until 7 years of age, and organochlorine compounds in blood. Hum Reprod binding. Toxicol Appl Pharmacol 179(3):185–194. 17(8):1973–1979. Richthoff J, Rylander L, Jonsson BA, Akesson H, Hagmar L, whereas we excluded cryptorchid cases diag- Ensslen SC, Riedel HH, Bluthgen H, Heeschen W. 1990. Nilsson-Ehle P, et al. 2003. Serum levels of 2,2’,4,4’,5,5’- nosed after 1 year. A final limitation of the Chlorinated hydrocarbons in seminal plasma and male hexachlorobiphenyl (CB-153) in relation to markers of study was that the medical records did not note fertility [in German]. Zentralbl Gynakol 112(13):817–821. reproductive function in young males from the general Giwercman YL, Kleist KE, Giwercman A, Giwercman C, Toft G, Swedish population. Environ Health Perspect 111:409–413. degree of hypospadias among the boys affected. Bonde JP, et al. 2006. Remarkably low incidence of hypos- Rignell-Hydbom A, Rylander L, Giwercman A, Jonsson BA, In conclusion, the results of the present padias in Greenland despite high exposure to endocrine Nilsson-Ehle P, Hagmar L. 2004. Exposure to CB-153 and study do not support an association between disrupters: possible protective effect of androgen receptor p,p’-DDE and male reproductive function. Hum Reprod genotype. Pharmacogenet Genomics 16(5):375–377. 19(9):2066–2075. in utero PCB exposure and either cryptorchid- Gladen BC, Longnecker MP, Schecter AJ. 1999. Correlations Ross G. 2004. The public health implications of polychlorinated ism or hypospadias. PCBs are just one type among polychlorinated biphenyls, dioxins, and furans in biphenyls (PCBs) in the environment. Ecotoxicol Environ of environmental endocrine modulator, how- humans. Am J Ind Med 35(1):15–20. Saf 59(3):275–291. ever. Whether other endocrine modulators Hardell L, van Bavel B, Lindstrom G, Carlberg M, Dreifaldt AC, Rozati R, Reddy PP, Reddanna P, Mujtaba R. 2002. Role of envi- Wijkstrom H, et al. 2003. Increased concentrations of poly- ronmental estrogens in the deterioration of male factor are related to risk is unclear and should be chlorinated biphenyls, hexachlorobenzene, and chlor- fertility. Fertil Steril 78(6):1187–1194. examined in other studies. danes in mothers of men with testicular cancer. Environ Safe SH. 1995. Environmental and dietary estrogens and human Health Perspect 111:930–934. health: is there a problem? Environ Health Perspect Hardy JB. 2003. The Collaborative Perinatal Project: lessons 103:346–351. Refe Rences and legacy. Ann Epidemiol 13(5):303–311. Schrader TJ, Cooke GM. 2003. Effects of Aroclors and individ- Hauser R, Chen Z, Pothier L, Ryan L, Altshul L. 2003. The relation- ual PCB congeners on activation of the human androgen Ahlborg UG, Becking GC, Birnbaum LS, Brouwer A, Derks HJGM, ship between human semen parameters and environmen- receptor in vitro. Reprod Toxicol 17(1):15–23. Feeley M, et al. 1994. Toxic equivalency factors for dioxin- tal exposure to polychlorinated biphenyls and p,p’-DDE. Sharpe RM. 2003. The “oestrogen hypothesis”—where do we like PCBs: Report on a WHO-ECEH and IPCS consultation, Environ Health Perspect 111:1505–1511. stand now? Int J Androl 26(1):2–15. December 1993. Chemosphere 28(6):1049–1067. Hosie S, Loff S, Witt K, Niessen K, Waag KL. 2000. Is there Skakkebaek NE, Rajpert-De Meyts E, Main KM. 2001. Testicular Ballschmiter K, Bacher R, Mennel A, Fischer R, Riehle U, Swerev a correlation between organochlorine compounds and dysgenesis syndrome: an increasingly common develop- M. 1992. The determination of chlorinated biphenyls, chlo- undescended testes? Eur J Pediatr Surg 10(5):304–309. mental disorder with environmental aspects. Hum Reprod rinated dibenzodioxins, and chlorinated dibenzofurans by Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, 16(5):972–978. GC-MS. HRC CC J High Resolut Chromatogr Chromatogr Needham LL, et al. 2002. Maternal serum level of 1,1-dichloro- Toppari J, Larsen JC, Christiansen P, Giwercman A, Grandjean P, 15(4):260–270. 2,2-bis(p-chlorophenyl)ethylene and risk of cryptorchidism, Guillette LJ Jr, et al. 1996. Male reproductive health and Barthold JS. 2008. Undescended testis: current theories of hypospadias, and polythelia among male offspring. Am J environmental xenoestrogens. Environ Health Perspect etiology. Curr Opin Urol 18(4):395–400. Epidemiol 155(4):313–322. 104(suppl 4):741–803. Barthold JS, Gonzalez R. 2003. The epidemiology of congenital Longnecker MP, Ryan JJ, Gladen BC, Schecter AJ. 2000. Ulbrich B, Stahlmann R. 2004. Developmental toxicity of poly- cryptorchidism, testicular ascent and orchiopexy. J Urol Correlations among human plasma levels of dioxin-like chlorinated biphenyls (PCBs): a systematic review of 170(6 pt 1):2396–2401. compounds and polychlorinated biphenyls (PCBs) and experimental data. Arch Toxicol 78(5):252–268. Bonde JP, Toft G, Rylander L, Rignell-Hydbom A, Giwercman A, implications for epidemiologic studies. Arch Environ Health Weiss JM, Bauer O, Bluthgen A, Ludwig AK, Vollersen E, Spano M, et al. 2008. Fertility and markers of male repro- 55(3):195–200. Kaisi M, et al. 2006. Distribution of persistent organochlo- ductive function in Inuit and European populations spanning Lunden A, Noren K. 1998. Polychlorinated naphthalenes and rine contaminants in infertile patients from Tanzania and large contrasts in blood levels of persistent organo chlorines. other organochlorine contaminants in Swedish human milk, Germany. J Assist Reprod Genet 23(9–10):393–399. Environ Health Perspect 116:269–277. 1972–1992. Arch Environ Contam Toxicol 34(4):414–423. Wolff MS, Camann D, Gammon M, Stellman SD. 1997. Proposed B o n e f e l d - J o r g e n s e n E C , A n d e r s e n H R , R a s m u s s e n T H , Mocarelli P, Gerthoux PM, Patterson DG Jr, Milani S, Limonta G, PCB congener groupings for epidemiological studies. Vinggaard AM. 2001. Effect of highly bioaccumulated poly- Bertona M, et al. 2008. Dioxin exposure, from infancy Environ Health Perspect 105:13–14. chlorinated biphenyl congeners on estrogen and androgen through puberty, produces endocrine disruption and affects receptor activity. Toxicology 158(3):141–153. human semen quality. Environ Health Perspect 116:70–77. | | 1476 volume 117 number 9 September 2009 • Environmental Health Perspectives

Journal

Environmental Health PerspectivesPubmed Central

Published: Apr 20, 2009

There are no references for this article.