Hertz-Picciotto, Irva*; Charles, M Judith†; James, Rebecca A.*; Keller, Jean A.*; Willman, Eric*; Teplin, Stuart‡
Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
Polychlorinated biphenyls (PCBs) are a class of chemically stable compounds that were produced in the United States for decades, until 1977. These were widely used in industry as heat transfer fluids, in capacitors and transformers, and in some consumer products. In this class are a possible 209 individual PCB congeners, differing in the number and location of attached chlorine atoms. These compounds bioaccumulate through the food chain, with the most important sources for humans being meat, fish, and both animal milk and human breastmilk. Because of their lipophilic properties, organochlorines are not cleared readily from the body.1
PCBs, along with their metabolites and some contaminants, are endocrine-disrupting compounds. PCBs are structurally similar to thyroxine, which is critical to in utero development.2 Additionally, PCB congeners and their metabolites have been shown to have estrogenic, antiestrogenic, or antiandrogenic activity.3,4 Two accidents, one in 1968 in Japan5 and the other in 1979 in Taiwan,6,7 led to the conclusion that high in utero exposures to heat-degraded PCBs could cause severe deficits in growth and development. These findings raised new concerns about possible adverse effects on prenatal and early childhood development from lower-level exposures. At lower environmental levels, investigations have shown reductions in average birth weight or increased risks for low birth weight as consumption of fish from contaminated waters increased8–12 or as measured PCB levels increased13,14; however, null results also have been reported.15–17 One study found the association to be limited to male infants,9 although no direct measurement of prenatal exposure was made.
In this study, we measured 11 PCB congeners in archived serum samples from an historical cohort of pregnant women from the Child Health and Development Study (CHDS). The CHDS was conducted during the 1960s, a time of peak production and use of PCBs. Thus, body burdens were higher than they currently are in most populations in the United States and other developed countries,18 although lower than in some groups.19,20 The CHDS follow-up of children through their early years allowed us to examine both intrauterine and early childhood growth in relation to their prenatal PCB exposures.
The CHDS was a prospective cohort study that enrolled approximately 20,000 pregnant women either attending prenatal clinics or giving birth at Kaiser Foundation Health Plan Medical Centers in the San Francisco Bay Area during the 1960s.21 During this time, the Kaiser Health Plan, a prepaid health maintenance organization, had enrolled 90% of membership through employer or union groups (G. Friedman, personal communication). The population from which we sampled was thus made up largely of persons having stable employment, with wealthy and very poor persons under-represented. In other respects, members represented a broad cross-section of the San Francisco Bay Area population.
The 399 woman/child pairs in this study were selected from 3412 CHDS children born April 1964 through April 1967 who participated in an extensive examination at 5 years of age, as described previously.21 To be eligible for our study, women had to have completed the enrollment interview before delivery and donated a blood specimen in the second or third trimester, with sufficient volume still remaining. We then excluded nonsingleton deliveries, children with severe anomalies, and children who did not complete 2 cognitive examinations and a hearing screening. We also excluded mothers who were deaf, who contracted or had contact with rubella during pregnancy, who were taking thyroid medication in the 60 days before blood draw, or who took iodine-containing medications22 during pregnancy or in the 6 months before conception. Finally, we excluded infants born at gestational ages that were unknown, less than 35 completed weeks, or greater than 45 completed weeks. On the basis of regularly updated contact information provided by the CHDS investigators, we selected children residing in 8 counties surrounding the greater San Francisco Bay Area so that further follow-up examinations might be feasible.
After these exclusions (Table 1), 1291 children were eligible for our study. From these eligible children, we sampled from 3 strata: those with cognitive delay, children who failed a hearing screening examination, and a random sample of the remainder. The motivation for this design was to enhance power for the hypotheses related to neurocognitive and neurosensory deficits (which were of relatively low prevalence) while not adversely affecting efficiency for the other hypotheses, including those related to physical growth, the subject of the present paper. Several authors have published papers on “reusing” data from case–control studies— designs that are, in essence, a special case of stratified sampling.23,24 We selected (1) all children with low cognitive scores defined as less than the 10th percentile on the Peabody Picture Vocabulary Test or an abbreviated Raven's Progressive Matrices (n = 194); (2) all who failed the hearing screening (an additional n = 46); and (3) a 17% random sample of all others (n = 159) to achieve the sample size goal. Where mothers had more than one eligible child, no more than one was included in the pool for selection. Serum PCB and lipid concentrations were determined for 399 women in this final sample.
Measurements of Organochlorine Compounds
Serum specimens collected during pregnancy were aliquoted and kept at −20°C at National Institutes of Health facilities. Laboratory methods for measurement of PCBs and of lipids have been described previously.18,25 Samples were thawed and homogenized; surrogate standards and glacial acetic acid were added; and the solution was mixed. PCBs were extracted 3 times and the extract passed through a florisil column. PCBs were eluted, then concentrated by rotary evaporation, and an internal standard (PCB 204) was added. Serum extracts were analyzed by gas chromatography with electron capture detection (Hewlett-Packard 6890 Series, Agilent Technologies, Palo Alto, CA). All samples were analyzed between 14 January 1998 and 22 September 1999 in batches of 10 to 20 samples. PCB concentrations were adjusted for percent recovery for each sample. Specimens analyzed on batch dates with extremely high or low variability were reanalyzed, and the 2 values averaged. Total lipids were estimated according to the formula of Phillips.26
Before analyzing samples from our subjects, noncritical samples from the CHDS cohort (women lost to follow-up) were analyzed to assess integrity of the stored serum specimens. Organochlorine determinations demonstrated comparability with other historical samples, and lipids were in reference ranges.
We measured specific PCBs that (1) were present in mixtures previously reported to have associations with reproductive, developmental or thyroid outcomes in animals or humans; (2) did not coelute with other compounds; and (3) were present in sufficient concentrations in our samples to achieve reasonable precision.25 These included: PCB IUPAC #101, #105, #110, #118, #137, #138, #153, #156, #170, #180, and #187. Values below the limit of quantitation were imputed as described previously.18 “Total PCBs” was calculated as the sum of the 9 congeners with fewer than 30% below the limit of quantitation (PCB IUPAC #105, #110, #118, #137, #138, #153, #170, #180, and #187).
Data Collection and Outcome Variables
We focused on outcomes at birth and at 5 years of age. Newborn measures were birth weight, gestational age at delivery, birth length, head circumference recorded at birth, and birth weight adjusted for gestational age at delivery. Adjusted birth weight was expressed as a Z-score representing the deviation from the mean of a standard Normal distribution of birth weight specific to week of gestation, sex, race, and parity.27 The standards were based on all U.S. births in 1989. Gestational age at delivery was calculated by subtracting the date of the last menstrual period (LMP) reported by the mother from the date of delivery in the medical chart. If month of LMP was known but day was not, reports of the “beginning of the month” were coded as the 7th of the month, the “end of the month” as the 23rd, and no day information as the 15th.
A physical examination was conducted as close to the child's fifth birthday as possible (approximately 70% were within 1 month of the birth date). Precise anthropometric measurements were made by personnel trained in high-precision protocols; growth standards established from the CHDS28,29 were published and cited by others in the field. Weight was measured in completed pounds and ounces; height and sitting height were measured in completed inches and sixteenths of an inch. The biacromial distance, chest breadth and depth, and bi-iliac distance were measured in millimeters with the Harpenden digital anthropometer (Holtain Ltd, Crymych PEMBS SA41 3UF, UK).
Mothers were interviewed during pregnancy regarding sociodemographic variables, lifestyle information, and reproductive histories. Medical records were abstracted. From these databases, we extracted potential confounders. Maternal factors were age, race, education, occupation, height, prepregnancy body mass index (BMI), parity, adequacy of prenatal care,30 smoking during pregnancy, alcohol consumption during pregnancy, medications taken during pregnancy, chronic hypertension, pre-eclampsia, and diabetes. We also obtained paternal height, education and occupation; sex of the child and whether the child was breastfed; and specimen characteristics, namely duration of pregnancy at the time of the blood draw, batch-date of laboratory analysis, and sample storage history (see Appendix, available with the online version of this article). Self-reported race and ethnicity were used to create one variable. Because of their relatively low representation in our sample, Hispanics (2.4%), Asians (4.2%), and multiracial or other races (3.5%) were combined with African-Americans. Maternal and paternal occupations were initially evaluated, but did not predict growth and therefore were removed from all models.
Medications taken during pregnancy were reviewed for potential effects on intrauterine growth, using references pertinent for the time period of the study.31–33 Active ingredients in medications from earlier years were checked against current references.34,35 Mothers were coded (yes/no) for medications prescribed or taken during pregnancy that have been reported to have an association with intrauterine growth in humans.
Data were cleaned for plausibility and consistency; continuous variables were screened for extreme values. Because PCBs are highly lipid soluble and circulating lipids change rapidly after ingestion of fatty meals, the measurements of PCBs were expressed as μg/g total lipids.
A multiple linear regression model predicting each outcome was fit for each individual congener and for total PCBs. Models were built separately for the birth and 5-year outcomes. The full set of covariates was reduced by eliminating factors that were not predictive (p > 0.15) of any outcome measured at that time point and did not confound the associations between PCBs and outcomes (criterion was <15% change in estimate).36 We conducted weighted least squares regression and used design effects to adjust for potential correlations within sampling strata. Resulting coefficients provide unbiased estimates of associations in the population from which we sampled. Regressions for growth variables were conducted in the Sudaan (RTI International, Research Triangle Park, NC) statistical software package and collinearity diagnostics were conducted in SAS (SAS Institute, Cary, NC).
Three women with diabetes were excluded from all birth models. Paternal height was missing for 14% of the children. Because paternal height was not a confounder of the PCBs in models of birth outcomes, it was eliminated, thereby permitting an analysis on the larger sample. It was, however, a confounder of the 5-year growth outcome models, and therefore these analyses were conducted on the smaller subset with complete data. We examined interactions between PCBs and race and between PCBs and child's sex. Associations did not differ for whites compared with nonwhites. However, the interactions with child's sex proved to be important and were therefore examined in models of birth outcomes.
We initially used the PCB measurements without transformation, but to ensure robustness of our findings we have reported only the results using log-transformed values. We further evaluated confounding by adjusting for p,p′-DDE.
Table 2 provides demographic, lifestyle and medical information about the study population, as well as a comparison to the original CHOS cohort. (For more details, see Appendix, available with the online version of this article). Our sample did not differ from the original cohort with regard to age, parity, prepregnancy BMI, maternal education, diabetes prevalence, or smoking. The mothers of those examined at 5 years, eligible for our study, and selected for our study were more likely to be African-American, more likely to have been diagnosed with essential hypertension, and more likely to consume low levels of alcohol; fathers were less likely to be college graduates or professionals.
The mean, median, 5th and 95th percentile of total PCBs were: 696, 616, 378, and 1115 ng/g serum lipids. For IUPAC PCB congener #153, these statistics were 148, 133, 72, and 273 ng/g lipids. These concentrations parallel those in other reports from the same time period in the United States and are higher than most but not all populations under study today.37
Table 3 presents results for each birth outcome. Overall, total PCBs were associated with a tendency toward lower birth weight, smaller head circumference, and shorter gestational age. However, heterogeneity by child's sex was observed, with the reduction in birth weight only in boys and shorter gestational age only in girls. Reduced head circumference was at least suggestive in both sexes. These models were adjusted for maternal age, BMI, height, parity, smoking, race, and hypertension or preeclampsia, as well as the adequacy of prenatal care,30 medications associated with intrauterine growth restriction that were taken during pregnancy, and child's sex. We also adjusted for the 3 specimen characteristics: duration of pregnancy at time of blood draw, batch-date of laboratory analysis, and an indicator for storage history (only at an NIH facility or stored on-site and later shipped to the facility). Additional adjustment for p,p′-DDE did not substantially alter the findings (not shown: see Appendix, available with the online version of this article).
On average, male infants had lower birth weight as maternal PCBs increased. For instance, at the 90th percentile of total PCBs, birth weights in male infants were reduced −290 g (95% confidence interval = −504 to −76 g) as compared with the 10th percentile of total PCBs. Other factors strongly related to birth weight were parity, maternal BMI, maternal height, and smoking (not shown). PCB's were not associated with birth weight in female infants. Figure 1 (top row) plots the log of the PCB concentration against birth weight residuals from multivariate models omitting PCBs. For girls, there is only random scatter, whereas for boys, a trend can be seen across the range of values in these data. The trend remains after removal of the 3 low birth-weight outliers (not shown). Five of the 6 most abundant congeners (#118, #138, #153, #170, and #180) had similar associations with birth weight to those for total PCBs, although again, only in boys (data not shown).
Findings for the z-score of weight-for-gestational-age were similar to those for birth weight (Figure 1, row 4). Head circumference was also reduced in association with increasing PCBs, a finding that was stronger in boys but still suggestive in girls (Table 3). A trend towards shorter birth length was observed, but results were not definitive. Gestational age was associated with total PCBs overall and in girls, but not in boys (Figure 1, row 2). In girls, although gestational age was inversely associated with total PCBs, the findings for specific congeners with regard to gestational age (as well as the other birth outcomes), were inconsistent, ie, some coefficients were positive and others negative, and precision was low.
Total prenatal PCBs showed a suggestive positive association with sitting height, standardized weight, and bi-iliac distance at 5 years of age. In girls, sitting height was greater in those with higher exposures, and standing height and standardized weight were suggestively increased. There was less consistency across congeners as compared with the birth weight and weight-for-gestational-age results (not shown).
Our finding of reduced birth weight in male infants with higher in utero PCB exposures, even after adjustment for gestational age at delivery, suggests that some PCBs may differentially affect intrauterine growth of male fetuses. Low birth weight infants represent a mix of premature deliveries and neonates whose in utero growth has been slower than optimal. Because gestational age at delivery was not associated with PCBs in boys, but weight-for-gestational-age was, our results specifically indicate an association with intrauterine growth restriction, rather than premature deliveries inboys.
Despite many reports, most previous epidemiologic studies on intrauterine growth relied on surrogate indicators (eg, fish consumption, residence near waste sites or in areas with high consumption of contaminated fish, etc.) rather than direct measures of exposure. Few reported sex-specific results, although when such results were available, they frequently indicated a stronger deficit in boys.8,9,38–40 In a Dutch cohort, Patandin et al13 found reduced birth weight with higher cord and maternal plasma PCB concentrations, but did not report sex-specific results. Total dioxin-like “equivalents” (which included PCBs) in breast milk of Finnish mothers were negatively associated birth weight, particularly among male offspring.39 Others8,11,12 also observed smaller head circumference in populations that consumed contaminated fish from the Great Lakes or Baltic Sea. Dewailly40 found reduced birth length in boys in relation to the most abundant PCBs in our study population. These authors reported a positive association between coplanar PCBs and birth length in girls, but no association with noncoplanar congeners. We were unable to measure coplanar PCBs because of the small volume of serum available. Experimental data in animals suggest that an effect on birth weight could be a result of impaired skeletal growth, with male fetuses more susceptible to the antiestrogenic effect of some congeners.41
Some studies contradict our birth weight findings, though none conducted sex-specific analyses.15–17,42 In one study, the authors erroneously adjusted for maternal weight gain, which is itself partly a result of fetal growth and therefore an inappropriate control variable.42 Two other null studies used biologic measures of exposure; each was half the size of ours, and low birth weight babies were under-represented, either because of the design17 or because of the heavy consumption of sea mammals rich in omega fatty acids.16
In contrast to our birth weight association in boys, the gestational age result we observed in girls was not consistent across individual congeners. This pattern may suggest that chance played a role, or that contaminants or co-occurring compounds, rather than PCBs, were involved. Berkowitz et al43 saw no association between PCBs and preterm delivery, while Taylor and colleagues44 did find shorter gestations associated with model-based estimates of PCB exposures in women working in a capacitor manufacturing plant.
We found no evidence of growth deficits at 5 years for children with relatively higher prenatal PCB exposures. In fact, we observed increased growth in girls. The Yu Cheng children, who experienced much higher prenatal exposure to PCBs as well as to PCDFs,45 were shorter in height and had less total lean and soft tissue mass at ages 6–12 years.46 Shorter height at age 18 years,47 and lower weight at puberty48 and at 4 years49 have been reported for environmentally exposed populations; the findings for weight were in relation to prenatal exposures. However, Patandin et al,13 who found reduced growth rate to 3 months of age, observed no continued deficit between 3 and 42 months, consistent with our findings. The growth rate of school children exposed in the Yusho incident was initially stunted but subsequently returned to that of controls,50 also suggesting a catch-up phenomenon even though this cohort was exposed postnatally. Although the literature is mixed, our data are consistent with several studies suggesting an initial adverse effect on physical growth that is reversed over time. Increased growth in girls in association with PCBs has not been reported in other cohorts.
As compared with the original cohort, those who participated in the 5-year examination and those who met our eligibility criteria were more highly educated and more likely to be African-American. Greater residential mobility may explain the lower percentage of whites retained for the 5-year examination. The resulting near-balance of whites and African-Americans maximized our power to assess heterogeneity by race, and it was not observed. In other respects, the sample was representative of the original cohort, except where our specific selection criteria skewed the sample, namely, in the exclusion of a small group who delivered at less than 35 weeks of pregnancy. Our selection criteria appear unlikely to have introduced bias into the associations investigated here. The self-selection for the 5-year examination and our inclusion of only those with known paternal height for 5-year growth analyses may point to an effect of PCB's towards greater growth in girls from more favorable home environments. Alternatively, our finding may be a chance one.
Validity of our data is supported by the confirmation of known risk factors for intrauterine growth restriction, such as maternal smoking, short stature, and low BMI, as well as female sex of the infant. High-quality PCB determinations were ensured by use of within-batch replicates, standards across batches, and surrogate standards.26
In conclusion, we observe that increased maternal serum PCB concentrations are associated with reduced birth weight in newborn boys, shorter gestation in girls, and smaller head circumference in both sexes. Because PCBs after adjustment for lipids are very stable across trimesters of pregnancy (correlations greater than 0.8051), measurements taken in the second and third trimester provide a good indication of fetal exposures throughout pregnancy. Combined with a lower male:female ratio at birth in association with higher PCB exposures (Hertz-Picciotto et al, unpublished data), these results indicate greater susceptibility of male conceptuses either in the fetal or embryonic period. Greater growth in girls was observed but may have been a chance finding or a result that pertains to a select subset.
Given the long-term decline in PCB concentrations, how relevant are these findings? First, studies in historical populations such as the CHDS provide an upper limit on the magnitude of toxic effects expected from today's environmental PCB levels in most of the world. Second, these investigations are informative for assessing risk in populations currently exposed to higher levels, eg, those who rely on fish from contaminated lakes and seas or who reside near former manufacturing facilities. Third, if contaminants of PCB mixtures are responsible for these associations, trends in these exposures deserve to be monitored. Finally, other chemical classes with similar structure, such as polybrominated diphenyl ethers, are widely used in consumer products and share many of the biochemical and toxicologic properties of PCBs.52,53 As polybrominated diphenyl ethers are rapidly increasing in human populations,54 studies such as this one may provide a clue regarding potential human health effects from these newer compounds.
We thank Bea van den Berg and Barbara Cohn for making the CHDS specimens available for this study, Bobbie Christianson for sharing her wealth of knowledge about the CHDS database, and Teri Greenfield, who assisted with statistical programming and manuscript preparation in the final stages.
1.Matthews HB, Dedrick RL. Pharmacokinetics of PCBs. Annu Rev Pharmacol Toxicol
2.Porterfield SP. Vulnerability of the developing brain to thyroid abnormalities: environmental insults to the thyroid system. Environ Health Perspect
. 1994;102(Suppl 2):125–130.
3.Hany J, Lilienthal H, Sarasin A, et al. Developmental exposure of rats to a reconstituted PCB mixture or aroclor 1254: effects on organ weights, aromatase activity, sex hormone levels, and sweet preference behavior. Toxicol Appl Pharmacol
4.Letcher RJ, Lemmen JG, van der Burg B, et al. In vitro antiestrogenic effects of aryl methyl sulfone metabolites of polychlorinated biphenyls and 2,2-bis(4-chlorophenyl)-1,1-dichloroethene on 17beta-estradiol-induced gene expression in several bioassay systems. Toxicol Sci.
5.Kuratsune M, Yoshimura T, Matsuzaka J, Yamaguchi A. Yusho, a poisoning caused by rice oil contaminated with polychlorinated biphenyls. HSMHA Health Rep
6.Chen YC, Yu ML, Rogan WJ, Gladen BC, Hsu CC. A 6-year follow-up of behavior and activity disorders in the Taiwan Yu-cheng children. Am J Public Health
7.Guo YL, Lambert GH, Hsu CC. Growth abnormalities in the population exposed in utero and early postnatally to polychlorinated biphenyls and dibenzofurans. Environ Health Perspect
. 1995;103(Suppl 6):117–122.
8.Rylander L, Stromberg U, Hagmar L. Decreased birthweight among infants born to women with a high dietary intake of fish contaminated with persistent organochlorine compounds. Scand J Work Environ Health
9.Rylander L, Stromberg U, Hagmar L. Dietary intake of fish contaminated with persistent organochlorine compounds in relation to low birthweight. Scand J Work Environ Health
10.Rylander L, Stromberg U, Hagmar L. Agreement between reported fish consumption obtained by two interviews and its impact on the results in a reproduction study. Eur J Epidemiol
11.Rylander L, Stromberg U, Hagmar L. Lowered birth weight among infants born to women with a high intake of fish contaminated with persistent organochlorine compounds. Chemosphere
12.Fein GG, Jacobson JL, Jacobson SW, Schwartz PM, Dowler JK. Prenatal exposure to polychlorinated biphenyls: effects on birth size and gestational age. J Pediatr
13.Patandin S, Koopman-Esseboom C, de Ridder MA, Weisglas-Kuperus N, Sauer PJ. Effects of environmental exposure to polychlorinated biphenyls and dioxins on birth size and growth in Dutch children. Pediatr Res
14.Rylander L, Stromberg U, Dyremark E, Ostman C, Nilsson-Ehle P, Hagmar L. Polychlorinated biphenyls in blood plasma among Swedish female fish consumers in relation to low birth weight. Am J Epidemiol
15.Rogan WJ, Gladen BC, McKinney JD, et al. Neonatal effects of transplacental exposure to PCBs and DDE. J Pediatr
16.Grandjean P, Bjerve KS, Weihe P, Steuerwald U. Birthweight in a fishing community: significance of essential fatty acids and marine food contaminants. Int J Epidemiol
17.Gladen BC, Shkiryak-Nyzhnyk ZA, Chyslovska N, Zadorozhnaja TD, Little RE. Persistent organochlorine compounds and birth weight. Ann Epidemiol
18.James RA, Hertz-Picciotto I, Willman E, Keller JA, Charles MJ. Determinants of serum polychlorinated biphenyls and organochlorine pesticides measured in women from the child health and development study cohort, 1963–1967. Environ Health Perspect
19.Kocan A, Petrik J, Drobna B, Chovancova J. Levels of PCBs and some organochlorine pesticides in the human population of selected areas of the Slovak Republic. I. Blood. Chemosphere
20.Dewailly E, Ayotte P, Bruneau S, Laliberte C, Muir DC, Norstrom RJ. Inuit exposure to organochlorines through the aquatic food chain in arctic quebec. Environ Health Perspect
21.Child Health and Development Study. Data Archive and User's Manual of the Child Health and Development Studies
. Vol. 1 & 2. Berkeley, CA: School of Public Health, University of California at Berkeley and Western Consortium for Public Health; 1994.
22.Surks MI, Sievert R. Drugs and thyroid function. N Engl J Med
23.Nagelkerke NJ, Moses S, Plummer FA, Brunham RC, Fish D. Logistic regression in case-control studies: the effect of using independent as dependent variables. Stat Med
24.Lee AJ, McMurchy L, Scott AJ. Re-using data from case-control studies. Stat Med
25.Willman E, Hertz-Picciotto I, Keller JA, Martinez E, Charles MJ. A reproducible approach to the reporting of organochlorine compounds in epidemiologic studies. Chemosphere
26.Phillips DL, Pirkle JL, Burse VW, Bernert JT Jr., Henderson LO, Needham LL. Chlorinated hydrocarbon levels in human serum: effects of fasting and feeding. Arch Environ Contam Toxicol
27.Zhang J, Bowes WA Jr. Birth-weight-for-gestational-age patterns by race, sex, and parity in the United States population. Obstet Gynecol
28.Wingerd J, Schoen EJ, Solomon IL. Growth standards in the first two years of life based on measurements of white and black children in a prepaid health care program. Pediatrics
29.Wingerd J, Solomon IL, Schoen EJ. Parent-specific height standards for preadolescent children of three racial groups, with method for rapid determination. Pediatrics
30.Kotelchuck M. The Adequacy of Prenatal Care Utilization Index: its US distribution and association with low birthweight. Am J Public Health
31.Physician's Desk Reference to Pharmaceutical Specialties and Biologicals.
16th ed. Oradell, NJ: Medical Economics; 1962.
32.Physician's Desk Reference to Pharmaceutical Specialties and Biologicals
. 19th ed. Oradell, NJ: Medical Economics, 1965.
33.Physician's Desk Reference to Pharmaceutical Specialties and Biologicals
. 22nd ed. Oradell, NJ: Medical Economics, 1968.
34.Physician's Desk Reference to Pharmaceutical Specialties and Biologicals
. 54th ed. Oradell, NJ: Medical Economics, 2000.
35.Behrman RE, Kliegman R, Jenson HB. Nelson Textbook of Pediatrics
. 16th ed. Philadelphia: W.B. Saunders Co., 2000.
36.Greenland S. Modeling and variable selection in epidemiologic analysis. Am J Public Health
37.Longnecker MP, Wolff MS, et al. Comparison of polychlorinated biphenyl levels across studies of human neurodevelopment. Environ Health Perspect
38.Baibergenova A, Kudyakov R, Zdeb M, Carpenter DO. Low birth weight and residential proximity to PCB-contaminated waste sites. Environ Health Perspect
39.Vartiainen T, Jaakkola JJ, Saarikoski S, Tuomisto J. Birth weight and sex of children and the correlation to the body burden of PCDDs/PCDFs and PCBs of the mother. Environ Health Perspect
40.Dewailly E, Bruneau S, Ayotte P, et al. Health status at birth of Inuit newborn prenatally exposed to organochlorines. Chemosphere
41.Lind PM, Larsson S, Oxlund H, et al. Change of bone tissue composition and impaired bone strength in rats exposed to 3,3′,4,4′,5-pentachlorobiphenyl (PCB126). Toxicology
42.Dar EKM, Anderson HA, Sonzogni WC. Fish consumption and reproductive outcomes in Green Bay, Wisconsin. Environ Res
43.Berkowitz GS, Lapinski RH, Wolff MS. The role of DDE and polychlorinated biphenyl levels in preterm birth. Arch Environ Contam Toxicol
44.Taylor PR, Stelma JM, Lawrence CE. The relation of polychlorinated biphenyls to birth weight and gestational age in the offspring of occupationally exposed mothers. Am J Epidemiol
45.Guo YL, Ryan JJ, Lau BP, Yu ML, Hsu CC. Blood serum levels of PCBs and PCDFs in Yucheng women 14 years after exposure to a toxic rice oil. Arch Environ Contam Toxicol
46.Guo YL LC, Yao WJ, Ryan JJ, Hsu CC. Musculoskeletal changes in children prenatally exposed to polychlorinated biphenyls and related compounds (Yu-Cheng children). J Toxicol Environ Health
47.Rylander L, Hagmar L. Medical and psychometric examinations of conscripts born to mothers with a high intake of fish contaminated with persistent organochlorines. Scand J Work Environ Health
48.Blanck HM, Marcus M, Rubin C, et al. Growth in girls exposed in utero and postnatally to polybrominated biphenyls and polychlorinated biphenyls. Epidemiology
49.Jacobson JL, Jacobson SW, Humphrey HE. Effects of exposure to PCBs and related compounds on growth and activity in children. Neurotoxicol Teratol
50.Yoshimura T, Ikeda M. Growth of school children with polychlorinated biphenyl poisoning or yusho. Environ Res
51.Longnecker MP, Klebanoff MA, Gladen BC, Berendes HW. Serial levels of serum organochlorines during pregnancy and postpartum. Arch Environ Health
52.McDonald TA. A perspective on the potential health risks of PBDEs. Chemosphere
53.Hooper K, McDonald TA. The PBDEs: an emerging environmental challenge and another reason for breast-milk monitoring programs. Environ Health Perspect
54.Petreas M, She J, Brown FR, et al. High body burdens of 2,2′,4,4′-tetrabromodiphenyl ether (BDE-47) in California women. Environ Health Perspect.
© 2005 Lippincott Williams & Wilkins, Inc.