| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Health Research and Policy, Stanford University School of Medicine (A.S.), Stanford, California 54305-5405; Department of Microbiology and Immunology, University of Arizona and Cord Blood Registry’s Stem Cell Bank (D.T.H.), Tucson, Arizona 85721; and Department of Anthropology (P.R.B.), University of California, Berkeley, California 94720; and GeneSage, Inc. (P.R.B.), San Francisco, California 94105
Address all correspondence and requests for reprints to: Atsuko Shibata, M.D., Ph.D., Department of Health Research and Policy, Stanford University School of Medicine, HRP-Redwood Building, Room T211, Stanford, California 94305-5405. E-mail: ashibata{at}stanford.edu
Abstract
It has been hypothesized that exposure to elevated levels of estrogens and IGFs before birth may increase breast cancer risk in female offspring. We examined whether the concentrations of estrone, E2, IGF-I, IGF-II, and IGF-binding protein-1 (IGFBP-1), -2, and -3 in umbilical cord blood plasma differed in female neonates of three racial/ethnic groups with contrasting breast cancer risk. The study included 57 Caucasian, 22 Hispanic, and 22 Asian-American subjects. Relative contribution of race/ethnicity to the analyte level variability was the largest for IGFBP-1 (P = 0.06). The only statistically significant (P < 0.05) mean difference was the lower IGFBP-3 levels in Asian than in Caucasian subjects. Adjusted mean levels of estrone and E2 for Asian subjects were 128% and 109% of the Caucasian means, respectively, whereas the Hispanic group showed lower means (85% and 84% of the Caucasian means). IGF-I, IGFBP-1, and IGFBP-3 showed lower adjusted means for both Hispanics and Asians compared with Caucasians. However, these differences were not statistically significant. In summary, we have shown that concentrations of estrogens, IGF-I, IGF-II, and IGFBPs are not different in cord blood samples from Caucasian, Hispanic, and Asian-American subjects. These data do not support a link between antenatal exposure to elevated levels of estrogens and IGFs and breast cancer.
INCIDENCE RATES OF breast cancer vary significantly across countries and among racial/ethnic groups within the United States (1). For instance, among women living in California, age-standardized incidence rates of breast cancer are higher in non-Hispanic Caucasians than in Hispanics and Asians (2).
Findings from experimental and epidemiological studies (3) suggest that the intrauterine environment may influence the risk of breast cancer in female offspring (4, 5). Some investigators who examined the associations of pre- and perinatal characteristics, such as parental age at birth and birth weight, with subsequent risk of breast cancer speculated that exposure to high concentrations of sex hormones during the fetal period might explain the associations (6, 7, 8, 9, 10). The breast depends on estrogens for its development, growth, and physiological functions. Thus, estrogens may be reasonable candidate mediators for the effects of the perinatal environment on breast cancer risk. Other hormones and growth factors, including IGFs, may be as reasonable candidates as estrogens, as they also play an important role in the regulation of fetal and neonatal growth (11). This hypothesized association would have public health implications if the mediators (i.e. hormones and growth factors) could be tied to practical prenatal interventions.
A corollary of the prenatal hypothesis is that female fetuses in a population with higher breast cancer risk are exposed to, on the average, elevated levels of endogenous estrogens or IGFs than their counterparts in a population with lower breast cancer risk. The goal of this study was to examine whether concentrations of estrogens and IGFs in umbilical cord blood plasma were higher in female neonates of Caucasian descent than in their counterparts of Hispanic and Asian descent. All of the neonates included in this study were born in the United States. We also measured cord blood concentrations of IGF-binding proteins (IGFBPs).
Subjects and Methods
Study subjects and data collection
Subjects were identified and recruited from among the clients of the Cord Blood Registry (CBR), a company based in San Bruno, CA, that harvests and stores stem cells from umbilical cord blood for potential uses in medical treatment. The study protocol was approved by the panel on medical human subjects at Stanford University.
Between May and September 1999, invitation letters calling for participation in the study were mailed to prospective CBR clients along with the cord blood collection kit (prospective recruitment). Additional invitation letters were mailed in September 1999 to the CBR clients, probably of Hispanic or Asian descent based on their surnames, whose babies had been born by August 1999 (retrospective recruitment). The eligibility criteria were 1) both the mother and father of the child whose cord blood would be or had been banked fell into the same broad category of Caucasian, Hispanic, or Asian by self-identification; 2) the sex of the child was female; 3) the child was born in the United States; 4) the mother completed a questionnaire by mail or telephone (see below); and 5) the parents provided written consent for participation in the study at the time of completing the questionnaire. Of the 103 subjects who met all of the above criteria, cord blood plasma was available for 101 subjects, consisting of 57 Caucasian, 22 Hispanic, and 22 Asian subjects. The neonates were born between February 1996 and December 1999.
A self-administered questionnaire was mailed to each eligible family when they were enrolled in the study. The questionnaire included questions regarding demographic and socioeconomic variables for the parents of the child, reproductive history of the mother, and complications during the pregnancy with the child under study. For the subjects who had not returned the questionnaire after a second mailing and several reminder telephone calls, we attempted a telephone interview using the questionnaire, slightly modified for queries over the telephone. Of the 101 study subjects, 86 (85%) returned self-administered questionnaires by mail, and 15 (15%) completed telephone interviews. For the purpose of this study, the following conditions were considered complications during pregnancy that might possibly affect circulating levels of hormones and growth factors: toxemia, hypertension, diabetes, lupus, hyperemesis gravidarum, intrahepatic cholestasis, placental abruption, and urinary tract infections.
Information on the mother’s age at delivery, birth weight of the child, and gestational age at birth was obtained from the CBR client database with consent of the study participants (i.e. parents of the neonates).
Cord blood specimens and laboratory assays
Cord blood was drawn from the umbilical vein using heparinized syringes within 10 min of birth, shipped at room temperature to the CBR laboratory, and processed within 24 h according to the previously described protocol (12). Briefly, cord blood was diluted 1:1 with PBS, underlain with Histopaque, and centrifuged at room temperature for 30 min at 1800 rpm. Plasma was drawn, aliquoted in cryovials, and stored at -20 C. In May 2000, one plasma aliquot (
5 ml) was retrieved for each study subject and shipped to Stanford University, where the specimen was thawed, subaliquoted into three plastic vials, and refrozen at -70 C. Each subaliquot was labeled with a randomly chosen vial identification number to blind the personnel at the endocrine laboratories to the racial/ethnic identify of specimens and the identity of replicate samples (see below).
In September 2000, 1 subaliquot for each subject was shipped on dry ice to the Endocrine Sciences division of the ESOTERIX Center for Clinical Trials (Calabasas Hills, CA) for estrogen assays. Plasma estrone and E2 concentrations were measured by extraction with hexane/ethyl acetate and chromatography on Sephadex LH-20 columns with benzene/methanol, followed by RIA with specific antiserum, using a modification of the method of Wu and Lundy (13). The coefficients of variation (CVs) for interassay variations from routine clinical assays at the laboratory were 8.2% for estrone (based on 45 assays with a mean of 20.8 ng/dl and an SD of 1.7 ng/dl) and 6.9% for E2 (based on 44 assays with a mean of 30.5 ng/dl and an SD of 2.1 ng/dl).
In December 2000, a second subaliquot for each subject was shipped on dry ice to Diagnostics Systems Laboratories, Inc. (Webster, TX) for IGF and IGFBP assays. Plasma concentrations of IGF-I, IGF-II, IGFBP-1, and IGFBP-3 were measured by immunoradiometric assay using the DSL-5600, -9100, -7800, and -6600 kits (Diagnostics Systems Laboratories, Inc.), respectively. IGFBP-2 levels were measured by double antibody RIA using the DSL-7100 kit (Diagnostics Systems Laboratories, Inc.). All specimens were analyzed in duplicate, and mean values were calculated. Those specimens with more than 15% CV between duplicates were subjected to repeat analysis. The mean value for each analyte in each specimen was reported. Assay results reported by the Endocrine Sciences and Diagnostics Systems Laboratories, Inc., laboratories were multiplied by 2 before statistical analysis to account for the 1:1 dilution of cord blood with PBS in the initial processing.
To assess the reliability of assays, a set of replicate samples from nine randomly selected subjects (three each from the Caucasian, Hispanic, and Asian groups; different sets for the two laboratories) was included in the shipment of plasma specimens to each laboratory without revealing the identity of replicate samples to the laboratory personnel. The CV was calculated for each of the nine paired replicates as a ratio of SD to mean multiplied by 100. The intraclass correlation coefficient (ICC) of reliability was estimated for logarithmically transformed values of assay results according to the method described by Fleiss (14). CVs varied widely for estrone (4.5–77.9%) and E2 (9.2–63.5%). ICCs were 0.22 for estrone and 0.35 for E2. When two pairs of replicates with CV larger than 50% (same pairs for estrone and E2) were excluded, ICCs were 0.46 for estrone and 0.80 for E2. The ranges of CVs were 0.5–27.1% for IGF-I, 1.4–21.2% for IGF-II, 1.7–26.0% for IGFBP-1, 2.4–34.7% for IGFBP-2, and 0.2–48.2% for IGFBP-3. The same pair of replicates had the largest CVs for all IGF and IGFBP measurements, except for IGFBP-2. ICCs were 0.90 for IGF-I, 0.80 for IGF-II, 0.99 for IGFBP-1, 0.67 for IGFBP-2, and 0.73 for IGFBP-3.
Statistical analysis
Subsequent statistical analyses were performed using the SAS statistical package, release 8.1 (SAS Institute, Inc., Cary, NC) (15). Assay results of replicate specimens were not included in these analyses. All P values reported here are nominal (i.e. not adjusted for multiple comparisons), except for pairwise comparisons of three group means by Dunnett’s t test (see below).
Means of continuous variables were compared among three racial/ethnic groups by ANOVA using the generalized linear model procedure. Proportions for categorical variables were compared among the three groups by the 
2 test or Fisher’s exact test using the FREQ procedure.
Contribution of variation due to race/ethnicity to the overall variability in plasma concentrations of analytes (hormones, growth factors, and proteins), while adjusting for covariates, was examined by analysis of covariance (ANCOVA) (16), using the general linear models procedure. As the distributions of all of the analytes measured in this study were skewed to the right, logarithmically transformed values of measurement were used in the following ANCOVA model: Y = µ + 
1 x A + 2 x H + ß1 x [birth weight (in grams)] + ß2 x [gestational age (in weeks)] + ß3 x [mother’s age (in years)] + 
x F + 
x C + e (Eq I), where Y denotes a logarithmically transformed value of an analyte; µ denotes an overall mean; 1 and 2 denote fixed effects of race/ethnicity (for Asians and Hispanics, respectively, compared with Caucasians); ß1, ß2, and ß3 denote regression coefficients for birth weight, gestational age, and mother’s age, respectively; denotes a fixed effect of birth rank (first born vs. all others); denotes a fixed effect of complications during pregnancy (yes vs. no); and e denotes residual errors. In model 1, A, H, F, and C are indicator variables defined such that they equal 1 if a subject is Asian, Hispanic, firstborn, and with complications during pregnancy, respectively, and 0 otherwise.
Least squares means (adjusted means) for each analyte, i.e. mean values that would be expected if the three racial/ethnic groups were balanced with respect to the distributions of birth rank, birth weight, gestational age, mother’s age, and complications during pregnancy, were estimated as part of the ANCOVA procedure (17) by holding all continuous variables to their mean value in the entire study sample and assuming that the subjects were evenly distributed across strata for categorical variables (e.g. 50% firstborn and 50% later born for birth rank). To determine whether adjusted means of either Asians or Hispanics were different from those of Caucasians, Dunnett’s two-tailed t tests were performed (17).
Results
Characteristics of the 101 neonates and their parents are summarized in Table 1
. Mean birth weight (±SEM) was similar between Caucasians (3493 ± 63 g) and Hispanics (3446 ± 120 g) and was slightly lower in Asians (3271 ± 83 g). Mean gestational age at birth was similar among the three racial/ethnic groups. Hispanic mothers were about 5 yr younger (30.6 ± 1.2 yr) than Caucasian (34.9 ± 0.6 yr) and Asian (35.5 ± 0.8 yr) mothers. The majority (59.1%) of Hispanic neonates were firstborn, whereas Caucasian and Asian neonates were mostly second or later born. Complications during pregnancy were reported most frequently by Asian mothers (27.3%), followed by Caucasian (14.3%) and Hispanic (4.6%) mothers. The birthplace of the neonate’s parents showed distinct patterns among the three groups: Caucasian parents were mostly born in the U.S., with less than 10% being foreign born; more than 65% of Asian parents were born outside the U.S., and about a third of Hispanic parents were foreign born.
|
shows the geometric mean and its 95% confidence interval for estrogens, IGFs, and IGFBPs by race/ethnicity. These estimates are not adjusted for other covariates. Pairwise comparisons of means by Dunnett’s t test showed that Asian subjects had a statistically significantly lower mean IGFBP-3 level than Caucasian subjects at = 0.05 (adjusted for the fact that multiple pairwise comparisons were made).
|
shows F values and corresponding P values from the ANCOVA. The larger the F value for a given number of degree(s) of freedom, the larger the relative contribution of a covariate to the overall variability of analyte levels. The relative contribution of race/ethnicity to the analyte level variability was the largest for IGFBP-1 (P = 0.06). The statistically significant difference in mean IGFBP-3 levels between Asian and Caucasian subjects by Dunnett’s t test persisted after adjustment for the covariates in the ANCOVA. Additional ANCOVA, in which a single fixed effect of race/ethnicity for Asians and Hispanics combined vs. Caucasians was assessed, indicated no statistically significant effect of race/ethnicity on estrogen, IGF, or IGFBP levels, except for a lower mean IGFBP-3 level for Asians and Hispanics combined than that for Caucasians (data not shown).
|
). Estimates of parameters for the covariates in model 1 (i.e. ß1, ß2, ß3, , and ) indicate the direction of correlation between analyte concentrations and covariates. In Table 3, positive and inverse associations, regardless of whether they were statistically significant, are indicated by bold type and italics, respectively. The following associations showed P values smaller than 0.05. Birth weight was correlated positively with IGF-I, IGF-II, and IGFBP-3 levels and inversely with IGFBP-2 levels. Gestational age at birth was correlated inversely with IGF-I and IGF-II levels and positively with IGFBP-2 levels. Mother’s age was inversely correlated with estrone and E2 levels. IGFBP-1 levels were higher in firstborn than in later born neonates.
To translate the ANCOVA results into a more interpretable form, expressed as concentrations of analytes, we estimated mean levels of estrogens, IGFs, and IGFBPs for the Caucasian, Hispanic, and Asian groups, respectively, adjusted for the covariates included in the ANCOVA. These adjusted means (Table 4) can be interpreted as group-specific means that would be expected if all subjects had the identical values of birth weight, gestational age, and mother’s age (set to mean values in the entire study sample) and the subjects were evenly distributed across strata for birth rank and complications during pregnancy. Adjusted estrogen levels for Asian subjects were no lower than those for Caucasians, while Hispanics had the lowest adjusted means for both estrone and E2. Only IGF-I, IGFBP-1, and IGFBP-3 showed lower adjusted means for both Hispanics and Asians in comparison with Caucasians. Mother’s birth place (the United States vs. other), when added to the ANCOVA model, accounted for little variability in analyte concentrations (data not shown) except that the adjusted mean of estrone levels was somewhat higher in neonates of foreign-born mothers (1174 ng/dl; 95% confidence interval 833-1654) than in those of US-born mothers (867 ng/dl; 95% confidence interval 686-1096).
|
Experimental data in rodents support the hypothesis that exposure to high estrogen levels during the perinatal period increases breast cancer risk in the female offspring (3). One may speculate that the same association exists in humans. Epidemiological approaches have limited ability to investigate associations of perinatal exposure to elevated levels of estrogens or other endogenous compounds with breast cancer risk. Major difficulties arise from the long putative induction period spanning several decades between exposure and outcome as well as potential confounding by risk factors during childhood, adolescence, and adulthood (18). One practical way of providing proof of principle is to examine cross-sectional variation in perinatal levels of candidate compounds in relation to established risk factors for breast cancer or in populations with contrasting breast cancer risks. The present study aimed to compare cord blood levels of estrogens, IGFs, and IGFBPs among three racial/ethnic groups. Our hypothesis was that cord blood of Caucasian neonates (high risk group) had higher plasma levels of estrogens and IGFs than that of Hispanic and Asian neonates (intermediate and low risk groups). This study also provides previously unavailable data on cord blood concentrations of estrogens, IGFs, and IGFBPs from a study sample diverse in race/ethnicity.
Overall, the ranges of analyte concentrations in cord blood plasma observed in this study were consistent with those reported by other investigators (11, 19, 20, 21, 22, 23, 24, 25, 26, 27). Our results of higher estrogen levels in Asians than in Caucasians are similar to the observations by Lipworth et al. (28) that maternal plasma levels of E2 at wk 16 of gestation were significantly higher in Chinese women than in Caucasian women, although the inference on maternal plasma levels during pregnancy cannot be directly extended to that on cord blood levels. To our knowledge, no data are available on racial/ethnic differences in maternal plasma levels of IGFs or IGFBPs.
In the comparison of estrogen and IGF levels among racial/ethnic groups, we adjusted means for selected characteristics of the neonate, mother, and pregnancy that have been associated with pre- and perinatal hormone levels in previous studies. In the present study, estrone and E2 levels were most strongly correlated with mother’s age, followed by birth rank in multivariate analysis. Higher estrogen levels in the first pregnancy than in later pregnancies have also been found in maternal plasma (29, 30, 31). We found no association between pregnancy complications and cord blood estrogen concentrations. Lower estrogen levels observed in pregnant women with preeclampsia (32) may not persist until the time of delivery. We found that IGF-I and IGFBP-3 levels were positively correlated with birth weight, consistent with other studies on IGF-I (20, 21, 22, 23, 24, 33, 34, 35) and IGFBP-3 (23, 24). IGFBP-2 levels were inversely correlated with birth weight, also consistent with a previous report (23). Gestational age was inversely correlated with IGF-I and IGF-II levels after adjustment for birth weight and other covariates, contradicting some previous reports on IGF-I (20, 21, 33, 34) and IGF-II (20, 21, 22, 25), which may be due in part to the relatively narrow range of gestational age in the present study. Unlike estrone and E2 levels, IGF and IGFBP levels showed no statistically significant correlation with mother’s age.
Interpretation of the present findings, particularly in the context of the hypothesized effects of intrauterine exposure to elevated levels of estrogens and IGFs on breast cancer risk, requires caution because of some limitations of the study. First, observed differences in means of hormone and growth factor concentrations, even though not statistically significant, may be large enough to explain part of the racial/ethnic variation in breast cancer risk. Less than perfect reproducibility of hormone and growth factor measurements, reflected by a wide range of CVs for some assays, also must be considered when interpreting subtle differences in analyte levels. Nonetheless, the present results do not seem to support the prediction that Asian neonates are exposed to lower estrogen or IGF levels than Caucasian counterparts. Second, measurements in cord blood plasma captures only a snapshot of analyte levels at a single point in time. It remains to be shown whether the time around birth is relevant to the potential effects of hormones and growth factors on mammary carcinogenesis, or whether perinatal analyte levels are correlated with the levels at a critical time period. Third, a possible birth cohort effect needs to be considered. This study examined female babies born in the last few years. By the time these babies reach their fifties and sixties, when breast cancer incidence rates start to rise, the racial/ethnic variation in breast cancer rates may be different from what it is now. Finally, most of the Hispanic and Asian subjects were recruited retrospectively, whereas all but two Caucasian subjects were enrolled prospectively before the babies were born. Although differential misclassification of subjects with regard to the covariates included in the present analysis is likely to be minimal, we cannot rule out a potential effect of variable length of frozen plasma specimen storage on the validity of hormone and growth factor measurements. In terms of the implications for breast cancer risk in populations, marginal distributions of estrogens and IGFs may be more relevant than means adjusted for covariates related to pregnancy and birth. However, comparisons of unadjusted means among the three racial/ethnic groups in this study would not be very informative, because the study subjects were not a population-based sample of newborns. The distributions of relevant covariates in target populations must be taken into account when addressing the hypothesis regarding potential effects of perinatal hormone levels on breast carcinogenesis in humans.
In summary, we have shown that concentrations of estrogens, IGF-I, IGF-II, and IGFBPs are not different in cord blood samples from Caucasian, Hispanic, and Asian-American subjects. Although this study provided previously unavailable or sparse data on racial/ethnic variation in perinatal levels of estrogens, IGFs, and IGFBPs, the data do not provide support for the hypothesized link between antenatal exposure to elevated levels of estrogens and IGFs and breast cancer.
Acknowledgments
We greatly appreciate the contributions of the following people: Thomas Moore of the CBR and Gloria Ochoa of the Cord Blood Donor Foundation (CBDF) for their support and invaluable advice, without which this study would have been impossible; Chris Goodman and Beth Mapother of the CBR laboratory for plasma specimen retrieval; the staff of the CBR and CBDF for their assistance with subject recruitment, data management, and telephone interviews; Ann Yuriko Minn, Maria DeJoseph, Iona Cheng, Maria Isabel Garcia, and Sarah Larson for their technical assistance with various aspects of the study; Dr. Linda Giudice for her advice on IGF assays and comments on the manuscript; and Drs. Byron Wm. Brown, Jr., and Philip Lavori for their advice on statistical analysis.
Footnotes
This work was supported by Grant Contract 9730 from the Susan G. Komen Breast Cancer Foundation.
Abbreviations: ANCOVA, Analysis of covariance; CBR, Cord Blood Registry; CV, coefficient of variation; ICC, intraclass correlation coefficient; IGFBP, IGF-binding protein.
Received July 25, 2001.
Accepted October 30, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  R. Troisi, P. Lagiou, D. Trichopoulos, B. Xu, L. Chie, F. Z. Stanczyk, N. Potischman, H.-O. Adami, R. N. Hoover, and C.-C. Hsieh Cord Serum Estrogens, Androgens, Insulin-Like Growth Factor-I, and Insulin-Like Growth Factor Binding Protein-3 in Chinese and U.S. Caucasian Neonates Cancer Epidemiol. Biomarkers Prev., January 1, 2008; 17(1): 224 - 231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Troisi, N. Potischman, and R. N. Hoover Exploring the Underlying Hormonal Mechanisms of Prenatal Risk Factors for Breast Cancer: A Review and Commentary Cancer Epidemiol. Biomarkers Prev., September 1, 2007; 16(9): 1700 - 1712. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A. Arslan, A. Zeleniuch-Jacquotte, A. Lukanova, Y. Afanasyeva, J. Katz, M. Levitz, G. Del Priore, and P. Toniolo Effects of Parity on Pregnancy Hormonal Profiles Across Ethnic Groups with a Diverse Incidence of Breast Cancer. Cancer Epidemiol. Biomarkers Prev., November 1, 2006; 15(11): 2123 - 2130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sanderson, J. R. Daling, D. R. Doody, and K. E. Malone Perinatal factors and mortality from breast cancer. Cancer Epidemiol. Biomarkers Prev., October 1, 2006; 15(10): 1984 - 1987. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. E. Murphy, R. Smith, W. B. Giles, and V. L. Clifton Endocrine Regulation of Human Fetal Growth: The Role of the Mother, Placenta, and Fetus Endocr. Rev., April 1, 2006; 27(2): 141 - 169. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. Pinheiro, M. D. Holmes, M. N. Pollak, R. L. Barbieri, and S. E. Hankinson Racial Differences in Premenopausal Endogenous Hormones Cancer Epidemiol. Biomarkers Prev., September 1, 2005; 14(9): 2147 - 2153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Cerhan, T. A. Sellers, C. A. Janney, V. S. Pankratz, K. R. Brandt, and C. M. Vachon Prenatal and Perinatal Correlates of Adult Mammographic Breast Density Cancer Epidemiol. Biomarkers Prev., June 1, 2005; 14(6): 1502 - 1508. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Potischman, R. Troisi, R. Thadhani, R. N. Hoover, K. Dodd, W. W. Davis, P. M. Sluss, C.-C. Hsieh, and R. Ballard-Barbash Pregnancy Hormone Concentrations Across Ethnic Groups: Implications for Later Cancer Risk Cancer Epidemiol. Biomarkers Prev., June 1, 2005; 14(6): 1514 - 1520. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Baik, W. J. DeVito, K. Ballen, P. S. Becker, W. Okulicz, Q. Liu, E. Delpapa, P. Lagiou, S. Sturgeon, D. Trichopoulos, et al. Association of Fetal Hormone Levels with Stem Cell Potential: Evidence for Early Life Roots of Human Cancer Cancer Res., January 1, 2005; 65(1): 358 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. DeLellis, S. Rinaldi, R. J. Kaaks, L. N. Kolonel, B. Henderson, and L. Le Marchand Dietary and Lifestyle Correlates of Plasma Insulin-Like Growth Factor-I (IGF-I) and IGF Binding Protein-3 (IGFBP-3): The Multiethnic Cohort Cancer Epidemiol. Biomarkers Prev., September 1, 2004; 13(9): 1444 - 1451. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hilakivi-Clarke, E. Cho, A. Cabanes, S. DeAssis, S. Olivo, W. Helferich, M. E. Lippman, and R. Clarke Dietary Modulation of Pregnancy Estrogen Levels and Breast Cancer Risk among Female Rat Offspring Clin. Cancer Res., November 1, 2002; 8(11): 3601 - 3610. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
