Abstract
Background
Little is known about relationships among reproductive factors, estrogens and estrogen metabolites (jointly referred to as EM), and estrogen metabolism patterns.
Methods
In a cross-sectional analysis, we examined the associations of age at menarche, menstrual cycle length and regularity, parity, age at first and last birth, breastfeeding, and duration of and time since use of oral contraceptives (OC) with mid-luteal phase urinary EM in a sample of 603 premenopausal women, ages 33–51, within the Nurses’ Health Study II (NHSII). Fifteen individual urinary EM were measured using liquid chromatography-tandem mass spectrometry (LC-MS/MS) and analyzed both individually and in metabolic pathways.
Results
Compared to women with extremely regular cycles, those with irregular cycles had lower levels of total EM (percent difference=24%; ptrend=0.01), estradiol (23%; ptrend=0.02), and 16-hydroxylation pathway EM (32%; ptrend<0.01). Longer menstrual cycles were associated with higher levels of estrone (percent difference ≥32 vs. <26 days: 25%; ptrend=0.03), estradiol (24%; ptrend=0.01), and 16-hydroxylation pathway EM (22%; ptrend=0.02). Among parous women, older age at first birth was associated with lower 16-hydroxylation pathway EM (percent difference age at first birth ≥35 vs. ≤25 years: 20%; ptrend=0.02). The other reproductive factors were not statistically significantly associated with individual urinary EM or EM pathways.
Conclusions and Impact
These data, based on a LC-MS/MS assay with high specificity and precision, provide an initial, comprehensive evaluation of the associations between reproductive factors and estrogen metabolism patterns.
Keywords: estrogen metabolism, urinary estrogens, premenopausal, reproductive factors
Introduction
Prior research suggests inverse associations between parity, breastfeeding, and age at menarche and breast cancer risk; and a positive association between age at first birth, current oral contraceptive use and risk (1, 2). Data also suggest a decrease in risk associated with an irregular menstrual cycle pattern and longer cycles (1), though these data are not consistent. The positive association between endogenous parent estrogens (estrone and estradiol) and breast cancer is established in postmenopausal women (3, 4), and suggested in premenopausal women in some (5), but not all (6, 7) studies.
Despite the hormonal nature of these reproductive factors and the observed associations between endogenous estrogens and breast cancer, lifestyle correlates of estrogen metabolism have been minimally explored in epidemiologic studies. The parent estrogens can be irreversibly hydroxylated at the 2-, 4-, or 16-positions of the steroid ring, and the reactive catechol estrogen metabolites formed by 2- and 4-hydroxylation can be stabilized by methylation of one of the two adjacent hydroxyl groups. Only three prior studies have presented reproductive correlates of a subset of urinary estrogen metabolites, with all three studies measuring only 2-hydroxyestrone and 16α-hydroxyestroneand the ratio of the two metabolites in urine (8–10). These studies found no association between these two estrogen metabolites and age at menarche, age at first birth, or parity in premenopausal and postmenopausal women.
The hypothesized effects of individual estrogens and estrogen metabolites (jointly referred to as EM) vary, given that EM differ in their relative capacity to stimulate cell proliferation and mutagenicity (11–13). A higher 2-hydroxyestrone to 16α-hydroxyestrone ratio has been hypothesized to be protective against breast cancer (14, 15), though that association has not consistently been observed when these two metabolites are measured in urine (8, 10, 16, 17), plasma (18) or serum (19). In the Nurses’ Health Study II (NHSII), we recently reported inverse relations of urinary parent estrogens and EM in the 2- and 4-pathways with breast cancer risk among premenopausal women (16).
Using the recently developed stable isotope dilution high-performance liquid chromatography-tandem mass spectrometry (LC-MS/MS) method we measured concurrently all 15 EM in urine with high sensitivity, specificity and reproducibility (20–23).Within the NHSII, among premenopausal women, we evaluated the associations between reproductive characteristics and urinary EM individually, by metabolic pathway (e.g., 2, 4-, and 16-hydroxylation; catechols and methylated catechols) and by metabolic pathway ratios.
MATERIALS AND METHODS
Study population and biospecimen collection
The NHSII was initiated in 1989 when 116,430 registered nurses enrolled in the cohort by completing and returning a questionnaire. Since that time, participants have completed biennial questionnaires to update exposures and collect health and disease data. Between 1996 and 1999, 29,611 participants provided blood and urine samples. These women were between 32 and 54 years of age, and were cancer-free at the time of collection. A total of 18,521 premenopausal women who had not used oral contraceptives, been pregnant, or breastfed during the past 6 months provided samples timed within the menstrual cycle. Women were asked to collect blood samples in the early follicular phase (days 3–5 of the menstrual cycle) and blood and urine samples in the mid-luteal phase (7–9 days prior to the expected onset of the next cycle). All participants who provided biospecimens completed a brief questionnaire at the time of specimen collection. Women recorded the first day of the menstrual cycle during which the samples were collected and returned a postcard recording the first day of their next cycle to allow determination of the timing of the luteal phase collection. Luteal urine samples were shipped, via overnight courier with an ice pack, to our laboratory with approximately 93% of samples received within 26 hours of collection. Samples have been stored in liquid nitrogen freezers since collection.
The current analysis is limited to premenopausal women with timed, luteal urine samples who were selected as controls for a nested-case control study of breast cancer (N=493) (16), as well as premenopausal women who participated in a biomarker reproducibility sub-study of the NHSII (N=110) (23). The study was approved by the Committee on the Use of Human Subjects in Research at Brigham and Women’s Hospital.
Exposure and covariate assessment
Women provided data on reproductive factors on the biennial questionnaires. Data collected included (year of collection): age at menarche (1989), parity (1997), age at first and last birth (1997), use of oral contraceptives (1997), time since oral contraceptive use (1997), and breastfeeding (1997; cumulative across pregnancies). For menstrual cycle length, women were asked “What is the current usual length of your menstrual cycle (interval from first day of period to first day of next period)?” and were provided categorical response options ranging from “<21 days” to “51+ days or too irregular to estimate” (1993). Similarly, for menstrual cycle regularity women were asked “What is the current usual pattern of your menstrual cycles (when not pregnant or lactating)?” with response options ranging from “extremely regular (no more than 1–2 days before or after expected)” to “no periods” (1993). Data on covariates, including alcohol use and physical activity, also were collected on the biennial questionnaires; height was reported in 1989. Data for reproductive factors and covariates was collected prior to urine collection (i.e. through the 1997 questionnaire), with the exception of alcohol and physical activity for which the average of the values on the 1997 and 2001 questionnaires were used for statistical adjustment. In addition, the specimen collection questionnaire provided data on current weight, age, blood and urine collection date and time, and whether the urine was a first morning urine sample.
Laboratory assays
For the urinary EM assay, 500µL of frozen urine was sent to the Laboratory of Proteomics and Analytical Chemistry, SAIC-Frederick, Inc., Frederick, MD. Endogenous estrogens and their metabolites are usually present in urine as glucuronide and sulfate conjugates; therefore, an initial enzymatic hydrolysis step was included. Each urine sample was thawed and mixed, and 400 µL was immediately aliquoted into a clean screw-cap glass tube and 20 µL of an internal standard solution containing 1.6 ng of each of five deuterated EM (17β-estradiol-d4, estriol-d3, 2-hydroxy-17β-estradiol-d5, 2-methoxy-17β-estradiol-d5, 16-epiestriol-d3) was added, followed by 0.5 mL of 0.15 M acetate buffer, pH 4.1, containing 2 mg of ascorbic acid and an enzymatic preparation from Helix pomatia with β-glucuronidase and sulfatase activity(Sigma-Aldrich, St. Louis, MO). The isotopically labelled EM are used to correct for loss of urinary EM during the hydrolysis, extraction, dansyl chloride derivatization, and liquid chromatography-tandem mass spectrometry (LC-MS/MS) steps of the assay procedure. Details of the assay have been published previously (21, 22). In brief, LC-MS/MS was performed with a TSQ Quantum-AM triple quadrupole mass spectrometer coupled with a Surveyor high performance liquid chromatography system (Thermo, San Jose, CA). Both the liquid chromatography system and mass spectrometer were controlled by Xcalibur software (Thermo). Quantification of each EM in urine was carried out using Xcalibur Quan Browser (Thermo). Calibration curves for the 15 EM were constructed by plotting EM/deuterium-labeled EM peak area ratios versus amounts of the EM. The amount of each EM in the urine sample was then interpolated using a linear function. Masked replicate quality control samples were placed in each batch to assess laboratory variability. The overall coefficients of variation (CVs) were <7% except for 4-methoxyestrone (17%) and 4-methoxyestradiol (15%), the two EM with the lowest concentrations. The lower level of quantitation for each EM is about 150 fmol/mL urine.
Creatinine was measured in three batches at the Endocrine Core Laboratory at Emory University (Atlanta, GA), Dr. Nader Rifai’s laboratory at Boston Children’s Hospital (Boston, MA), and Dr. Vincent Ricchiuti’s laboratory at Brigham and Women’s Hospital (Boston, MA). Overall CVs were ≤9.2% in all labs.
Statistical analysis
Urinary EM concentrations (pmol/mL) were adjusted for creatinine levels to account for urine volume, which resulted in units of picomoles EM per milligram of creatinine, and were further log transformed to achieve an approximately normal distribution. EM were evaluated individually, as well as grouped by metabolic pathways (e.g., methylated catechols, 16-hydroxylation pathway) and as pathway ratios (e.g., 2-hydroxylation pathway/16-hydroxylation pathway). We also examined the 2-hydroxyestrone/16α-hydroxyestroneratio given the prior hypothesized association between this ratio and breast cancer risk (14, 15). Statistical outliers were identified using the extreme studentized many deviate procedure (24). This resulted in the exclusion of up to ten values for individual EM, with the exception 2-Methoxyestradiol which had sixteen outliers, and up to nine values for ratio measures.
Reproductive factors evaluated include: duration of past OC use (never, 1–47 months, ≥48 months), time since last oral contraceptive use (never, 6–23 months, 24–47 months, 48–71 months, 48–71 months, 72–95 months, 96–119 months, ≥120 months), age at menarche (<12, 12, 13, ≥14 years), menstrual cycle pattern (extremely regular/no more than 1–2 days deviation from expected, very regular/within 3–4 days, regular/within 5–7 days, usually/always irregular), usual menstrual cycle length(<26, 26–31, ≥32 days), and parity (nulliparous; 1, 2, 3 or ≥4 children). Also evaluated among parous women were age at first and last birth (≤25, >25–30, >30–35, >35 years) and total duration of breastfeeding (0 to <1 month, 1 to 6 months, >6 to 12 months, >12 to 18 months, >18 to 24 months, >24 to 36 months, and >36 months).
We calculated geometric means using generalized linear models for each of the individual EM, metabolic pathway groups, and pathway ratios by categories of reproductive factors. Tests for trend were conducted by modeling continuous exposure measures (e.g. parity, age at first birth) or including the categorical variable in the model as an ordinal variable (e.g. menstrual cycle regularity). Multivariable models were adjusted for age at urine collection (years, continuous), luteal day at collection (<5, 6–7, 8–9, ≥10 days before next menstrual period), first morning urine (yes/no), BMI at collection (<25, 25–30, ≥30 kg/m2), total physical activity (<3, 3 to <9, 9 to<18, 18 to<27, 27 to<42, ≥42 MET-hrs/week), and alcohol consumption (non-drinker, up to 3 drinks/month, 3 drinks/month to 2 drinks/week, 2 drinks/week to 5 drinks/week, more than 5 drinks/weeks). Reproductive factors were mutually adjusted using combined age at first birth/parity (nulliparous, age at first birth <25 years/1–2 children, age at first birth 25 to 29 years/1–2 children, age at first birth ≥30 years/1–2 children, age at first birth <25 years/≥3 children, age at first birth ≥25 years/≥3 children) and, as categorized above, age at menarche, menstrual cycle regularity and menstrual cycle length. In the age at first birth analysis, parity was included as a continuous variable.
All p-values are two-sided and considered statistically significant if p<0.05. Analyses were conducted with SAS version 9 (SAS Institute, Cary, NC).
RESULTS
Participants had a mean age of 42.8 years at sample collection (Table 1). The majority of participants were parous (82%), and overall reported low consumption of alcohol (69% less than 2 drinks per week). Mean BMI was 25 kg/m2 (SD=5.3). Nearly half of the women (49%) experienced menarche at age 12 years or younger and the majority had used oral contraceptives (84%). Among parous women, 45% had an age at first birth of 25 years or younger and 87% had breastfed for at least 1 month.
Table 1.
Characteristics of Study Population at Urine Collection: Nurses’ Health Study II (n=603)
| Characteristic | mean (SD) | ||
|---|---|---|---|
| Age (years) | 42.8 (3.8) | ||
| BMI (kg/m2) | 25.1 (5.3) | ||
| n | % | ||
| Age at menarche | |||
| <12 | 126 | 20.9 | |
| 12 | 168 | 27.9 | |
| 13 | 181 | 30.0 | |
| ≥ 14 | 128 | 21.2 | |
| Menstrual Cycle Length | |||
| <26 days | 115 | 19.8 | |
| 26–31 days | 405 | 69.6 | |
| ≥32 days | 62 | 10.7 | |
| Menstrual Cycle Regularity* | |||
| extremely regular | 160 | 27.5 | |
| very regular | 248 | 42.7 | |
| regular | 146 | 25.1 | |
| usually/always irregular | 27 | 4.7 | |
| Past Oral Contraceptive Use | |||
| Never | 94 | 15.9 | |
| Ever, < 4 years | 266 | 44.9 | |
| Ever, ≥ 4 years | 232 | 39.2 | |
| Parity | |||
| Nulliparous | 109 | 18.1 | |
| 1 child | 76 | 12.6 | |
| 2 children | 247 | 41.0 | |
| 3 children | 119 | 19.7 | |
| ≥ 4 children | 52 | 8.6 | |
| Age at First Birth | |||
| ≤25 | 222 | 45.0 | |
| >25 – 30 | 177 | 35.9 | |
| >30–35 | 72 | 14.6 | |
| >35 | 22 | 4.5 | |
| Breastfeeding | |||
| 0 to <1 month | 65 | 13.2 | |
| 1 to 6 months | 47 | 9.6 | |
| >6 to 12 months | 87 | 17.7 | |
| >12 to 18 months | 92 | 18.7 | |
| 18 to 24 months | 51 | 10.4 | |
| >24 to 36 months | 84 | 17.1 | |
| >36 months | 65 | 13.2 | |
| Alcohol, ~10 g serving | |||
| Non-drinker | 183 | 30.4 | |
| up to 3 servings/month | 99 | 16.4 | |
| 3/month - 2/week | 135 | 22.4 | |
| 2/week – 5/week | 99 | 16.4 | |
| > 5/week | 87 | 14.4 | |
| Race, Caucasian | 571 | 94.7 | |
| First Morning Urine | 472 | 79.6 | |
| Not Perimenopausal† | 529 | 87.7 | |
| Sample taken within 4–10 days of next menstrual period | 516 | 85.6 | |
| Ovulatory Cycle | 542 | 89.9 | |
extremely regular: no more than 1–2 days deviation from expected; very regular/within 3–4 days; regular/within 5–7 days; usually/always irregular
women considered perimenopausal at urine collection if report being menopausal within 4 years of urine collection.
We observed several statistically significant associations between urinary EM levels and both regularity and length of menstrual cycles. In general, as compared to extremely regular menstrual cycles, irregular menstrual cycles were associated with lower urinary levels of total EM, estradiol, the 16-hydroxylation pathway, and four of the five 16-pathway EM; and the trends for these EM measures by cycle regularity were statistically significant (Table 2).
Table 2.
Adjusted Geometric Means (pmol/mg creatinine) of Estrogen/Estrogen Metabolite Levels by Menstrual Cycle Regularity: Nurses’ Health Study II (n=603)
| Estrogen/estrogen metabolite measures | Extremely regular |
Very regular |
Regular | Usually/ Always irregular |
p-trend |
|---|---|---|---|---|---|
| n | 160 | 248 | 146 | 27 | |
| Total estrogens and estrogen metabolites | 227.9 | 189.0 | 199.0 | 173.8 | 0.01 |
| Parent estrogens | 46.8 | 36.6 | 42.5 | 38.0 | 0.10 |
| Estrone | 31.1 | 24.9 | 28.3 | 26.3 | 0.14 |
| Estradiol | 15.9 | 13.0 | 14.0 | 12.2 | 0.02 |
| Catechols | 62.8 | 54.6 | 60.6 | 62.7 | 0.77 |
| 2-Catechols | 55.1 | 46.4 | 52.8 | 54.4 | 0.73 |
| 2-Hydroxyestrone | 48.3 | 40.7 | 46.9 | 48.2 | 0.82 |
| 2-Hydroxyestradiol | 6.1 | 4.9 | 5.6 | 5.3 | 0.29 |
| 4-Catechols | |||||
| 4-Hydroxyestrone | 5.9 | 5.2 | 6.2 | 5.8 | 0.72 |
| Methylated catechols | 11.6 | 8.8 | 10.8 | 10.0 | 0.32 |
| Methylated 2-catechols | 11.2 | 8.5 | 10.4 | 9.8 | 0.34 |
| 2-Methoxyestrone | 8.9 | 6.7 | 8.1 | 7.9 | 0.28 |
| 2-Methoxyestradiol | 0.73 | 0.64 | 0.71 | 0.77 | 0.93 |
| 2-Hydroxyestrone-3-methyl ether | 1.3 | 1.1 | 1.3 | 1.0 | 0.77 |
| Methylated 4-catechols | 0.21 | 0.19 | 0.21 | 0.15 | 0.51 |
| 4-Methoxyestrone | 0.14 | 0.12 | 0.13 | 0.09 | 0.30 |
| 4-Methoxyestradiol | 0.05 | 0.05 | 0.06 | 0.04 | 0.64 |
| 2-Hydroxylation pathway | 68.0 | 56.0 | 64.5 | 65.5 | 0.60 |
| 4-Hydroxylation pathway | 6.3 | 5.7 | 6.6 | 6.1 | 0.74 |
| 16-Hydroxylation pathway | 81.3 | 67.0 | 68.2 | 55.3 | <0.01 |
| 16a-Hydroxyestrone | 14.2 | 11.4 | 11.9 | 9.4 | 0.02 |
| Estriol | 37.0 | 31.3 | 30.7 | 24.8 | 0.01 |
| 17-Epiestriol | 1.7 | 1.5 | 1.9 | 1.1 | 0.76 |
| 16-Ketoestradiol | 16.3 | 13.4 | 14.2 | 11.3 | 0.01 |
| 16-Epiestriol | 7.3 | 6.3 | 6.3 | 5.4 | 0.01 |
| Ratios of metabolic pathways | |||||
| 4-Catechols/2-Catechols | 0.10 | 0.12 | 0.11 | 0.10 | 0.65 |
| 2-Catechols/16-Pathway | 0.68 | 0.67 | 0.78 | 0.98 | 0.09 |
| Catechols/16-Pathway | 0.78 | 0.77 | 0.89 | 1.1 | 0.08 |
| 4-Pathway/2-Pathway | 0.09 | 0.11 | 0.10 | 0.09 | 0.29 |
| 2-Pathway/16-Pathway | 0.84 | 0.79 | 0.95 | 1.2 | 0.12 |
| 4-Pathway/16-Pathway | 0.08 | 0.09 | 0.10 | 0.11 | 0.03 |
| 2,4-Pathway/16-Pathway | 0.94 | 0.91 | 1.1 | 1.3 | 0.09 |
| 2-Pathway/4,16-Pathway | 0.75 | 0.71 | 0.82 | 1.0 | 0.17 |
| 2-Catechols/methylated 2-catechols | 4.9 | 5.2 | 5.2 | 5.6 | 0.38 |
| 4-Catechols/methylated 4-catechols | 26.2 | 26.9 | 29.7 | 37.3 | 0.27 |
| Catechols/methylated catechols | 5.4 | 5.9 | 5.7 | 6.2 | 0.31 |
| Parent estrogens/estrogen metabolites | 0.27 | 0.26 | 0.28 | 0.26 | 0.67 |
| 2-Pathway/parent estrogens | 1.5 | 1.4 | 1.5 | 1.7 | 0.37 |
| 4-Pathway/parent estrogens | 0.13 | 0.15 | 0.16 | 0.16 | 0.11 |
| 16-Pathway/parent estrogens | 1.7 | 1.8 | 1.6 | 1.4 | 0.13 |
| 2-Hydroxyestrone/16α-hydroxyestrone | 3.4 | 3.5 | 3.9 | 5.1 | 0.09 |
Adjusted for age, luteal day, first morning urine, BMI, alcohol consumption, physical activity, age at menarche, cycle length, breastfeeding, age at first birth/parity
Specifically, relative to women with extremely regular menstrual cycles, women with irregular menstrual cycles had lower levels of total EM (24% difference; ptrend=0.01), estradiol (23%; ptrend= 0.02), and 16-pathway EM (32%; ptrend<0.01). Irregular menstrual cycles also were associated with a higher ratio of 4/16-hydroxylation pathways (irregular vs. extremely regular cycles: 0.08 vs. 0.11; 45% difference; ptrend=0.03). Menstrual cycle regularity was not associated with the 2- or 4- hydroxylation pathway EM, either individually or in groups.
Menstrual cycle length was statistically significantly positively associated with estrone, estradiol, the 16-hydroxylation pathway, and three of the 16-pathwayEM, and nonsignificantly positively associated with total EM and parent estrogens (Table 3). Longer menstrual cycles were associated with higher levels of estrone (≥32 vs. <26 days: 25%difference; ptrend=0.03), and estradiol (24%; ptrend=0.01) (Table 3). Longer cycles also were associated with higher levels of the 16-hydroxylation pathway (≥32 vs. <26 days: 22%difference; ptrend=0.02), and three of the 16-pathway EM (17-epiestriol, 16-ketoestradiol, and 16-epiestriol). The individual EM in the 2- and 4- hydroxylation pathways were not associated with menstrual cycle length when considered individually or as groups, however there was a significant inverse association between cycle length and the ratio of 2-pathway/parent EM (ptrend=0.04).
Table 3.
Adjusted Geometric Means (pmol/mg creatinine) of Estrogen/Estrogen Metabolite Levels by Menstrual Cycle Length: Nurses’ Health Study II (n=603)
| Menstrual Cycle Length, days |
||||
|---|---|---|---|---|
| Estrogen/estrogen metabolite measures | <26 | 26–31 | 32+ | p-trend |
| n | 115 | 405 | 62 | |
| Total estrogens and estrogen metabolites | 176.7 | 186.3 | 199.9 | 0.17 |
| Parent estrogens | 36.5 | 41.7 | 43.4 | 0.06 |
| Estrone | 23.8 | 27.6 | 29.8 | 0.03 |
| Estradiol | 12.3 | 14.0 | 15.2 | 0.01 |
| Catechols | 52.3 | 55.7 | 55.0 | 0.94 |
| 2-Catechols | 45.4 | 48.8 | 46.5 | 0.97 |
| 2-Hydroxyestrone | 40.1 | 43.3 | 40.6 | 0.93 |
| 2-Hydroxyestradiol | 4.7 | 4.9 | 5.3 | 0.67 |
| 4-Catechols | ||||
| 4-Hydroxyestrone | 5.0 | 5.2 | 5.9 | 0.69 |
| Methylated catechols | 9.3 | 9.7 | 9.2 | 0.92 |
| Methylated 2-catechols | 9.0 | 9.4 | 9.0 | 0.97 |
| 2-Methoxyestrone | 7.1 | 7.4 | 7.1 | 0.92 |
| 2-Methoxyestradiol | 0.64 | 0.71 | 0.66 | 0.67 |
| 2-Hydroxyestrone-3-methyl ether | 1.0 | 1.2 | 1.1 | 0.39 |
| Methylated 4-catechols | 0.17 | 0.18 | 0.18 | 0.93 |
| 4-Methoxyestrone | 0.10 | 0.11 | 0.10 | 0.76 |
| 4-Methoxyestradiol | 0.05 | 0.04 | 0.05 | 0.88 |
| 2-Hydroxylation pathway | 56.0 | 60.0 | 56.7 | 0.95 |
| 4-Hydroxylation pathway | 5.5 | 5.5 | 6.4 | 0.83 |
| 16-Hydroxylation pathway | 59.6 | 66.2 | 72.5 | 0.02 |
| 16a-Hydroxyestrone | 10.3 | 11.0 | 12.0 | 0.15 |
| Estriol | 30.8 | 30.1 | 32.4 | 0.54 |
| 17-Epiestriol | 1.2 | 1.5 | 1.9 | 0.01 |
| 16-Ketoestradiol | 11.5 | 13.0 | 14.8 | 0.01 |
| 16-Epiestriol | 5.6 | 6.3 | 7.3 | <0.01 |
| Ratios of metabolic pathways | ||||
| 4-Catechols/2-Catechols | 0.11 | 0.11 | 0.13 | 0.75 |
| 2-Catechols/16-Pathway | 0.76 | 0.72 | 0.64 | 0.10 |
| Catechols/16-Pathway | 0.86 | 0.80 | 0.75 | 0.13 |
| 4-Pathway/2-Pathway | 0.10 | 0.10 | 0.11 | 0.70 |
| 2-Pathway/16-Pathway | 0.92 | 0.87 | 0.78 | 0.10 |
| 4-Pathway/16-Pathway | 0.09 | 0.08 | 0.09 | 0.23 |
| 2,4-Pathway/16-Pathway | 1.0 | 1.0 | 0.90 | 0.10 |
| 2-Pathway/4,16-Pathway | 0.80 | 0.79 | 0.68 | 0.13 |
| 2-Catechols/methylated 2-catechols | 5.0 | 5.1 | 5.1 | 0.95 |
| 4-Catechols/methylated 4-catechols | 30.1 | 27.9 | 33.7 | 0.88 |
| Catechols/methylated catechols | 5.5 | 5.6 | 5.9 | 0.86 |
| Parent estrogens/estrogen metabolites | 0.27 | 0.29 | 0.29 | 0.19 |
| 2-Pathway/parent estrogens | 1.5 | 1.4 | 1.3 | 0.04 |
| 4-Pathway/parent estrogens | 0.15 | 0.13 | 0.15 | 0.26 |
| 16-Pathway/parent estrogens | 1.6 | 1.6 | 1.7 | 0.61 |
| 2-Hydroxyestrone/16α-hydroxyestrone | 3.9 | 3.8 | 3.4 | 0.19 |
Adjusted for age, luteal day, first morning urine, BMI, alcohol consumption, physical activity, age at menarche, cycle regularity, breastfeeding, age at first birth/parity
There were no statistically significant associations between parity and individual EM, metabolic pathway groups, or pathway ratios when comparing nulliparous to parous women (Table 4; all p>0.05) or across increasing number of children (data not shown). Among parous women, older age at first birth was associated with lower levels of the 16-hydroxylation pathway, both grouped (>35 vs. ≤25 years: 20%difference; ptrend=0.02), and individually, including 16-ketoestradiol (21%; ptrend<0.01), and 16-epiestriol (28%; ptrend<0.01) (Table 4). Older age at first birth was positively associated with higher ratios when the 16-hydroxylation pathway served as the denominator, including the following ratios: catechols/16-hydroxylation pathway (ptrend=0.02), 2-hydroxylation pathway/16-hydroxylation pathway (ptrend=0.01), 4-hyrdroxylation pathway/16-hydroxylation pathway (ptrend=0.01), 2-,4-hydroxylation pathway/16-hydroxylation pathway (ptrend=0.01). Age at first birth was not associated with 2- and 4-hydroxylation pathways, with the exception of the 2-pathway/parent EM ratio (ptrend=0.03). Age at last birth is significantly correlated with age at first birth (r=0.72, p<0.01), and results for age at last birth were similar to those for age at first birth (e.g. 16-hydroxlation pathway, >35 vs. ≤25 years: 18% difference, ptrend=0.05) (data not shown).
Table 4.
Adjusted Geometric Means (pmol/mg creatinine) of Estrogen/Estrogen Metabolite Levels by Parity and Age at First Birth: Nurses’ Health Study II (n=603)
| Parity† | Age at First Birth | ||||||
|---|---|---|---|---|---|---|---|
| Estrogen/estrogen metabolite measures |
Nulli- parous |
Parous | ≤ 25 | 26–30 | 31–35 | >35 | p- trend* |
| n | 109 | 494 | 222 | 177 | 72 | 22 | |
| Total estrogens and estrogen metabolites | 203.1 | 198.3 | 193.9 | 195.4 | 189.6 | 172.7 | 0.29 |
| Parent estrogens | 43.6 | 40.3 | 42.3 | 41.7 | 39.9 | 34.2 | 0.28 |
| Estrone | 29.4 | 27.0 | 28.1 | 28.2 | 26.0 | 21.8 | 0.31 |
| Estradiol | 14.2 | 13.5 | 13.8 | 14.2 | 13.3 | 12.4 | 0.58 |
| Catechols | 64.1 | 58.5 | 53.3 | 56.9 | 55.1 | 58.6 | 0.84 |
| 2-Catechols | 55.9 | 50.6 | 45.5 | 50.0 | 47.3 | 48.0 | 0.83 |
| 2-Hydroxyestrone | 49.6 | 44.6 | 40.1 | 44.6 | 41.7 | 41.8 | 0.83 |
| 2-Hydroxyestradiol | 5.6 | 5.3 | 4.8 | 5.0 | 4.9 | 5.7 | 0.73 |
| 4-Catechols | |||||||
| 4-Hydroxyestrone | 5.7 | 5.7 | 5.5 | 5.5 | 6.0 | 7.0 | 0.47 |
| Methylated catechols | 10.8 | 9.9 | 9.1 | 9.6 | 10.1 | 9.1 | 0.34 |
| Methylated 2-catechols | 10.4 | 9.5 | 8.8 | 9.3 | 9.6 | 8.8 | 0.42 |
| 2-Methoxyestrone | 8.0 | 7.5 | 6.9 | 7.4 | 7.5 | 6.9 | 0.42 |
| 2-Methoxyestradiol | 0.73 | 0.69 | 0.65 | 0.66 | 0.80 | 0.52 | 0.58 |
| 2-Hydroxyestrone-3-methyl ether | 1.3 | 1.2 | 1.1 | 1.1 | 1.2 | 0.95 | 0.59 |
| Methylated 4-catechols | 0.22 | 0.20 | 0.18 | 0.18 | 0.18 | 0.17 | 0.90 |
| 4-Methoxyestrone | 0.13 | 0.12 | 0.11 | 0.11 | 0.10 | 0.09 | 0.48 |
| 4-Methoxyestradiol | 0.06 | 0.05 | 0.04 | 0.05 | 0.05 | 0.06 | 0.11 |
| 2-Hydroxylation pathway | 67.5 | 61.4 | 55.9 | 60.7 | 58.0 | 59.0 | 0.73 |
| 4-Hydroxylation pathway | 6.1 | 6.1 | 5.8 | 5.8 | 6.3 | 7.7 | 0.36 |
| 16-Hydroxylation pathway | 67.5 | 69.2 | 71.8 | 67.7 | 62.6 | 57.2 | 0.02 |
| 16a-Hydroxyestrone | 11.3 | 12.0 | 12.1 | 11.8 | 10.4 | 11.4 | 0.09 |
| Estriol | 29.8 | 31.5 | 33.4 | 30.2 | 29.4 | 25.6 | 0.06 |
| 17-Epiestriol | 1.6 | 1.6 | 1.7 | 1.4 | 1.5 | 1.3 | 0.05 |
| 16-Ketoestradiol | 13.5 | 14.1 | 14.4 | 13.1 | 12.5 | 11.3 | <0.01 |
| 16-Epiestriol | 6.5 | 6.4 | 6.9 | 6.4 | 6.2 | 4.9 | <0.01 |
| Ratios of metabolic pathways | |||||||
| 4-Catechols/2-Catechols | 0.09 | 0.11 | 0.12 | 0.10 | 0.12 | 0.15 | 0.75 |
| 2-Catechols/16-Pathway | 0.83 | 0.73 | 0.64 | 0.75 | 0.79 | 0.83 | 0.05 |
| Catechols/16-Pathway | 0.95 | 0.83 | 0.73 | 0.85 | 0.92 | 1.0 | 0.02 |
| 4-Pathway/2-Pathway | 0.09 | 0.10 | 0.11 | 0.09 | 0.11 | 0.13 | 0.70 |
| 2-Pathway/16-Pathway | 1.0 | 0.87 | 0.76 | 0.91 | 0.97 | 1.0 | 0.01 |
| 4-Pathway/16-Pathway | 0.09 | 0.09 | 0.08 | 0.09 | 0.10 | 0.13 | 0.01 |
| 2,4-Pathway/16-Pathway | 1.1 | 1.0 | 0.87 | 1.0 | 1.1 | 1.2 | 0.01 |
| 2-Pathway/4,16-Pathway | 0.87 | 0.77 | 0.69 | 0.80 | 0.82 | 0.85 | 0.05 |
| 2-Catechols/methylated 2-catechols | 5.4 | 5.2 | 5.2 | 5.4 | 4.9 | 5.4 | 0.50 |
| 4-Catechols/methylated 4-catechols | 25.5 | 28.6 | 31.2 | 29.4 | 31.5 | 42.2 | 0.71 |
| Catechols/methylated catechols | 5.9 | 5.8 | 5.9 | 6.0 | 5.5 | 6.4 | 0.40 |
| Parent estrogens/estrogen metabolites | 0.28 | 0.27 | 0.28 | 0.28 | 0.28 | 0.24 | 0.81 |
| 2-Pathway/parent estrogens | 1.6 | 1.5 | 1.3 | 1.4 | 1.5 | 1.8 | 0.03 |
| 4-Pathway/parent estrogens | 0.14 | 0.15 | 0.14 | 0.14 | 0.16 | 0.23 | 0.05 |
| 16-Pathway/parent estrogens | 1.6 | 1.7 | 1.7 | 1.6 | 1.5 | 1.7 | 0.08 |
| 2-Hydroxyestrone/16α-hydroxyestrone | 4.4 | 3.7 | 3.4 | 3.8 | 4.2 | 3.6 | 0.09 |
Adjusted for age, luteal day, first morning urine, BMI, alcohol consumption, physical activity, age at menarche, cycle length and regularity, breastfeeding, parity
Test for trend across age at first birth; excludes nulliparous women
differences between nulliparous and parous all p>0.05 with the exception of 4-Catechols/2-Catechols where p=0.05
Breastfeeding, duration of oral contraceptive use, years since oral contraceptive use, and age at menarche were not associated with either total urinary concentration of EM or concentrations of individual EM or metabolic pathway groups, or pathway ratios (data not shown). For example, for total EM, the geometric mean for <1 month of breastfeeding was 151.2 pmol/mg creatinine and for >36 months of breastfeeding 171.9 pmol/mg creatinine, with no clear pattern across the intermediate categories (ptrend=0.28).
We further evaluated these associations in three subgroups of participants: those who gave their urine samples during an ovulatory cycle (defined by plasma mid-luteal progesterone levels >400 ng/dL) (n=542), those whose urine sample was collected within 4–10 days of their next menstrual period (n=516), and those who were not perimenopausal (i.e., women reporting they did not become menopausal within 4 years after sample collection) (n=529). Results among each of these subgroups were similar to the overall results (e.g., among women who were not perimenopausal, the association between ≥32 vs. <26 days for the 16-hydroxylation pathway was 69.2 vs. 59.2 pmol/mg creatinine, ptrend=0.06, as compared to 72.5 vs. 59.6 pmol/mg creatinine, ptrend=0.02 among all women).
DISCUSSION
In this first study of 15 urinary EM and reproductive factors among premenopausal women, we observed statistically significant associations of menstrual cycle regularity, menstrual cycle length, and age at first birth with parent estrogens and EM in the 16-hydroxylation pathway. We observed higher levels of estrone and estradiol among women with longer compared with shorter menstrual cycle length and lower levels of estradiol among women with irregular compared with regular menstrual cycles. Of the three major pathways (2-, 4-, and 16-hydroxylation) of estrogen metabolism, most of our statistically significant findings were for the 16-hydroxylation pathway, with longer menstrual cycle length associated with higher levels of these EM, and menstrual cycle irregularity and older age at first birth associated with lower levels. There were no statistically significant associations between these reproductive factors and 2- and 4-hydroxylation pathway EM. In addition, we observed no statistically significant associations of age at menarche, duration of or time since oral contraceptive use, parity, or breastfeeding with individual EM or estrogen metabolism pathways.
Three prior studies have evaluated the association between reproductive factors and urinary EM, but each of these included only 2-hydroxyestrone and 16α-hydroxyestrone (8–10) and only one study presented results separately for premenopausal women (8). In the Guernsey III cohort, Meilahn and colleagues found no statistically significant associations between age at menarche, parity, or age at first birth and the 2-hydroxyestrone/16α-hydroxyestroneratio in premenopausal (n=139), as well as postmenopausal (n=184), women (8). The remaining two studies presented data on pre- and postmenopausal women combined (9) (% premenopausal, cases=26%; controls=34.4%) or postmenopausal women (10). In these studies, no statistically significant associations were observed between parity (9, 10), age at menarche (10), or age at first birth (9) and 2-hydroxyestrone, 16α-hydroxyestrone, or the ratio of the two metabolites. These studies were limited by small sample size, and limited representation of premenopausal women. In the current analysis, we found no association between parity, age at menarche, or age at first birth and 2-hydroxyestrone, 16α-hydroxyestrone, or their ratio, in agreement with these prior findings. To our knowledge, there is no prior epidemiologic data on the association between oral contraceptive use, breastfeeding, or menstrual cycle regularity and length and urinary measures of estrogen metabolism.
Experimental data suggest that individual EM vary in their estrogenic and genotoxic potential. EM in the 2- and 4-hydroxylation pathways bind to the estrogen receptor, but have different rates of dissociation, which lead to potentially differential effects. EM in the 2-hydoxylation pathway demonstrate a faster rate of dissociation than EM in the 4-hydroxylation pathway (25–27), which suggests a higher estrogenic potential of 4-hydroxylation pathway EM compared with 2-hydroxylation pathway EM. Data also suggest higher genotoxicity for 4-hydroxylation pathway EM (11). In addition, neoplastic tissue, both benign and malignant, has higher levels of 4-hydroxylation pathway EM than normal breast tissue while 2-hydroxylation pathway EM are similar in neoplastic and normal breast tissue (11). The 16-hydroxylation pathway EM also exhibit estrogenic and genotoxic potential. Specifically, 16α-hydroxyestrone can bind covalently to the estrogen receptor, which leads to a constitutively activated receptor (28). In addition, animal data show unscheduled DNA synthesis in mouse mammary cells (29) associated with 16α-hydroxyestrone, suggesting genotoxicity. However, these hypotheses about the roles of specific EM in human carcinogenesis are all based on experimental model systems.
To date, our prior analysis of these 15 EM and breast cancer risk, the controls from which are included in the current analysis, is the first and only comprehensive epidemiologic study of these 15 EM in urine and breast cancer risk among premenopausal women (16). We observed a decreased risk with higher urinary levels of estrone and estradiol (RR estrone=0.52, 95% CI=0.30–0.88; estradiol=0.51, 95% CI=0.30–0.86), inverse, but not statistically significant patterns with 2- and 4- hydroxylation pathway EM, no clear relationship with urinary 16-pathway EM, and an increased risk associated with higher urinary 17-epiestriol levels (top vs. bottom quartile RR=1.74, 95% CI=1.08–2.81, ptrend=0.01).
The importance of excreted vs. circulating estrogens is unclear. Circulating estrogens may not reflect all of the activity in the breast tissue and urinary estrogens are an additional step removed. Considering plasma levels alone, data from the NHSII suggest an increased risk with high levels of estradiol in the follicular phase, but not in the luteal phase (5). In analyses in our EM case-control study including both urine and plasma measures, high urinary estrone and estradiol (i.e. high excretion) was consistently associated with decreased breast cancer risk, regardless of plasma levels. Therefore, from our data it appears that estrogen excretion may be an important factor.
Age at menarche, parity, and breastfeeding are inversely associated and age at first birth positively associated with risk of breast cancer (1, 2). The associations between irregular and/or long menstrual cycles and breast cancer risk have not been consistent, with either no association (30, 31) or an inverse association (32–34) reported. In the NHSII cohort, cycles >32 days long or too irregular to estimate at ages 18–22 (but not in later adulthood) were associated with decreased risk of breast cancer only in the subgroup of women diagnosed before age 40 (HR: 0.71; 95% CI: 0.53–0.97) (32). Given experimental data suggesting the 4- and 16-hydroxylation pathways may have the highest genotoxic and proliferative potential, we hypothesized that these EM would be positively associated with hormonally related breast cancer risk factors; and 2-hydroxylation pathway EM inversely associated with these risk factors (e.g. higher 4- and 16-pathway EM and lower 2-pathway EM associated with the higher risk group). Our findings for menstrual cycle regularity were in the hypothesized direction with more irregular menstrual cycles associated with lower 16-hydroxylation pathway EM. However, for menstrual cycle length, our results indicating a positive association with higher 16-pathway levels are contrary to the hypothesized direction based on experimental evidence. Similarly, our finding of older age at first birth being associated with lower levels of EM in the 16-hydroxylation pathway, including 17-epiestriol, is contrary to the hypothesized direction, given the consistent, positive association between older age at first birth and breast cancer risk (1, 2) and the positive association between 17-epiestriol and breast cancer in the NHSII case-control study (16). Despite our hypotheses, based on the experimental literature, that 4-pathway EM would be positively associated with hormonally related breast cancer risk factors and 2-pathway EM would be inversely related, we found no association between these pathways and menstrual cycle characteristics or age at first birth.
There are several limitations to the current study. This is a cross-sectional study, which limits our ability to infer causality. While the reproductive factors studied were all measured prior to the specimen collection, it is possible that reverse causality was in part responsible for these associations, given that a woman’s pattern of estrogen metabolism may affect reproductive events. We used data on usual menstrual cycle length and regularity, collected in 1993, and cycle length and regularity may change over time. This would result in some non-differential classification of the exposure, and bias our results toward the null. Second, data from a single urine collection were used for this analysis. However, although multiple measures would be preferable, a within-person stability analysis in a subset of the samples included in this analysis (n=110) revealed moderately high intraclass correlation coefficients (ICCs) over three years, with ICCs ≥ 0.50 observed for most of the individual EM, estrogen metabolic pathways and pathway ratios (23). Finally, while we made multiple comparisons, we interpret our results cautiously and in the context of biologically driven hypotheses and what is currently known about estrogen biochemistry and metabolism. Our study has important strengths including a sample size substantially larger than previous studies, a more comprehensive investigation of reproductive factors than earlier work, and a comprehensive examination of individual and grouped EM representing the three main pathways of estrogen metabolism. Prior analyses of urinary estrogen metabolites used enzyme immunoassays (8–10), while the present study used a LC-MS/MS assay, a method that concurrently measures all 15 urinary EM and provides improved reproducibility, specificity, and accuracy, compared to previous methods (35).
In conclusion, this first analysis in premenopausal women of the association between reproductive factors and 15 urinary luteal EM yielded some statistically significant associations between menstrual cycle length and regularity and age at first birth and patterns of estrogen metabolism; no associations were observed with age at menarche, oral contraceptive use, parity, and breastfeeding. Regular menstrual cycles were associated with higher levels of 16-hydroxylation pathway EM, which have estrogenic and genotoxic potential, but the opposite was observed for longer menstrual cycle length and older age at first birth. Similarly, estradiol was associated in opposite directions with menstrual cycle irregularity and length. Further work is needed to confirm our observed associations between reproductive factors and estrogen metabolism patterns and determine additional correlates of urinary EM in premenopausal women.
Acknowledgments
Funding/Support: This study was supported by Research Grants CA67262 and CA50385 from the National Cancer Institute and the Intramural Research Program of the Division of Cancer Epidemiology and Genetics of the National Cancer Institute, and with federal funds of the National Cancer Institute awarded under Contract HHSN261200800001E to SAIC-Frederick. RT Fortner is supported in part by T32 CA09001. The content of this publication does not necessarily reflect the views or policies of the U.S. Department of Health and Human Services; nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
References
- 1.Kelsey JL, Gammon MD, John EM. Reproductive factors and breast cancer. Epidemiol Rev. 1993;15(1):36–47. doi: 10.1093/oxfordjournals.epirev.a036115. 00/01; [DOI] [PubMed] [Google Scholar]
- 2.Key TJ, Verkasalo PK, Banks E. Epidemiology of breast cancer. The Lancet Oncol. 2001;2(3):133–140. doi: 10.1016/S1470-2045(00)00254-0. 00/01; [DOI] [PubMed] [Google Scholar]
- 3.Kaaks R, Rinaldi S, Key TJ, Berrino F, Peeters PH, Biessy C, et al. Postmenopausal serum androgens, oestrogens and breast cancer risk: The European Prospective Investigation into Cancer and Nutrition. Endocr Relat Cancer. 2005 Dec;12(4):1071–1082. doi: 10.1677/erc.1.01038. [DOI] [PubMed] [Google Scholar]
- 4.Key T, Appleby P, Barnes I, Reeves G Endogenous Hormones and Breast Cancer Collaborative Group. Endogenous sex hormones and breast cancer in postmenopausal women: Reanalysis of nine prospective studies. J Natl Cancer Inst. 2002 Apr 17;94(8):606–616. doi: 10.1093/jnci/94.8.606. [DOI] [PubMed] [Google Scholar]
- 5.Eliassen AH, Missmer SA, Tworoger SS, Spiegelman D, Barbieri RL, Dowsett M, et al. Endogenous steroid hormone concentrations and risk of breast cancer among premenopausal women. J Natl Cancer Inst. 2006 Oct 4;98(19):1406–1415. doi: 10.1093/jnci/djj376. [DOI] [PubMed] [Google Scholar]
- 6.Kaaks R, Berrino F, Key T, Rinaldi S, Dossus L, Biessy C, et al. Serum sex steroids in premenopausal women and breast cancer risk within the European Prospective Investigation into Cancer and Nutrition (EPIC) J Natl Cancer Inst. 2005 May 18;97(10):755–765. doi: 10.1093/jnci/dji132. [DOI] [PubMed] [Google Scholar]
- 7.Walker K, Bratton DJ, Frost C. Premenopausal endogenous oestrogen levels and breast cancer risk: A meta-analysis. Br J Cancer. 2011 doi: 10.1038/bjc.2011.358. 00/13; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Meilahn EN, De Stavola B, Allen DS, Fentiman I, Bradlow HL, Sepkovic DW, et al. Do urinary oestrogen metabolites predict breast cancer? Guernsey III cohort follow-up. Br J Cancer. 1998;78(9):1250–1255. doi: 10.1038/bjc.1998.663. 00/01; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kabat GC, Chang CJ, Sparano JA, Sepkovie DW, Hu XP, Khalil A, et al. Urinary estrogen metabolites and breast cancer: A case-control study. Cancer Epidemiol Biomarkers Prev. 1997;6(7):505–509. 00/01; [PubMed] [Google Scholar]
- 10.Ursin G, London S, Stanczyk FZ, Gentzschein E, Paganini-Hill A, Ross RK, et al. Urinary 2-hydroxyestrone/16alpha-hydroxyestrone ratio and risk of breast cancer in postmenopausal women. J Natl Cancer Inst. 1999;91(12):1067–1072. doi: 10.1093/jnci/91.12.1067. 00/16; [DOI] [PubMed] [Google Scholar]
- 11.Joubert A, Van Zyl H, Laurens J, Lottering M. C2- and C4-position 17beta-estradiol metabolites and their relation to breast cancer. Biocell. 2009;33(3):137–140. 00/01; [PubMed] [Google Scholar]
- 12.Yager JD, Davidson NE. Estrogen carcinogenesis in breast cancer. N Engl J Med. 2006;354(3):270–282. doi: 10.1056/NEJMra050776. 00/19; [DOI] [PubMed] [Google Scholar]
- 13.Seeger H, Wallwiener D, Kraemer E, Mueck A. Comparison of possible carcinogenic estradiol metabolites: Effects on proliferation, apoptosis and metastasis of human breast cancer cells. Maturitas. 2006;54(1):72–77. doi: 10.1016/j.maturitas.2005.08.010. 00/20; [DOI] [PubMed] [Google Scholar]
- 14.Bradlow HL, Hershcopf RJ, Martucci CP, Fishman J. Estradiol 16 alpha-hydroxylation in the mouse correlates with mammary tumor incidence and presence of murine mammary tumor virus: A possible model for the hormonal etiology of breast cancer in humans. Proc Natl Acad Sci U S A. 1985 Sep;82(18):6295–6299. doi: 10.1073/pnas.82.18.6295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bradlow HL, Hershcopf RE, Fishman JF. Oestradiol 16 alpha-hydroxylase: A risk marker for breast cancer. Cancer Surv. 1986;5(3):573–583. [PubMed] [Google Scholar]
- 16.Eliassen AH, Spiegelman D, Xu X, Keefer LK, Veenstra TD, Barbieri RL, et al. Urinary estrogens and estrogen metabolites and subsequent risk of breast cancer among premenopausal women. Cancer Res. 2012;72:696–706. doi: 10.1158/0008-5472.CAN-11-2507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Muti P, Bradlow HL, Micheli A, Krogh V, Freudenheim JL, Schunemann HJ, et al. Estrogen metabolism and risk of breast cancer: A prospective study of the 2:16alpha-hydroxyestrone ratio in premenopausal and postmenopausal women. Epidemiology. 2000;11(6):635–640. doi: 10.1097/00001648-200011000-00004. 00/01; [DOI] [PubMed] [Google Scholar]
- 18.Eliassen AH, Missmer SA, Tworoger SS, Hankinson SE. Circulating 2-hydroxy- and 16alpha-hydroxy estrone levels and risk of breast cancer among postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2008 Aug;17(8):2029–2035. doi: 10.1158/1055-9965.EPI-08-0262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Arslan AA, Shore RE, Afanasyeva Y, Koenig KL, Toniolo P, Zeleniuch-Jacquotte A. Circulating estrogen metabolites and risk for breast cancer in premenopausal women. Cancer Epidemiol Biomarkers Prev. 2009;18(8):2273–2279. doi: 10.1158/1055-9965.EPI-09-0312. 00/05; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ziegler RG, Faupel-Badger JM, Sue LY, Fuhrman BJ, Falk RT, Boyd-Morin J, et al. A new approach to measuring estrogen exposure and metabolism in epidemiologic studies. J Steroid Biochem Mol Biol. 2010 Aug;121(3–5):538–545. doi: 10.1016/j.jsbmb.2010.03.068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Xu X, Veenstra TD, Fox SD, Roman JM, Issaq HJ, Falk R, et al. Measuring fifteen endogenous estrogens simultaneously in human urine by high-performance liquid chromatography-mass spectrometry. Anal Chem. 2005;77(20):6646–6654. doi: 10.1021/ac050697c. 00/01; [DOI] [PubMed] [Google Scholar]
- 22.Falk RT, Xu X, Keefer L, Veenstra TD, Ziegler RG. A liquid chromatography-mass spectrometry method for the simultaneous measurement of 15 urinary estrogens and estrogen metabolites: Assay reproducibility and interindividual variability. Cancer Epidemiol Biomarkers Prev. 2008 Dec;17(12):3411–3418. doi: 10.1158/1055-9965.EPI-08-0355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Eliassen AH, Ziegler RG, Rosner B, Veenstra TD, Roman JM, Xu X, et al. Reproducibility of fifteen urinary estrogens and estrogen metabolites over a 2- to 3-year period in premenopausal women. Cancer Epidemiol Biomarkers Prev. 2009 Nov;18(11):2860–2868. doi: 10.1158/1055-9965.EPI-09-0591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rosner B. Percentage points for a generalized ESD many-outlier procedure. Technometrics. 1983 May;25(2):165–172. [Google Scholar]
- 25.Van Aswegen CH, Purdy RH, Wittliff JL. Binding of 2-hydroxyestradiol and 4-hydroxyestradiol to estrogen receptors from human breast cancers. J Steroid Biochem. 1989 Apr;32(4):485–492. doi: 10.1016/0022-4731(89)90380-4. [DOI] [PubMed] [Google Scholar]
- 26.Anstead GM, Carlson KE, Katzenellenbogen JA. The estradiol pharmacophore: Ligand structure-estrogen receptor binding affinity relationships and a model for the receptor binding site. Steroids. 1997 Mar;62(3):268–303. doi: 10.1016/s0039-128x(96)00242-5. [DOI] [PubMed] [Google Scholar]
- 27.Barnea ER, MacLusky NJ, Naftolin F. Kinetics of catechol estrogen-estrogen receptor dissociation: A possible factor underlying differences in catechol estrogen biological activity. Steroids. 1983 May;41(5):643–656. doi: 10.1016/0039-128x(83)90030-2. [DOI] [PubMed] [Google Scholar]
- 28.Lustig RH, Mobbs CV, Pfaff DW, Fishman J. Temporal actions of 16 alpha-hydroxyestrone in the rat: Comparisons of lordosis dynamics with other estrogen metabolites and between sexes. J Steroid Biochem. 1989 Sep;33(3):417–421. doi: 10.1016/0022-4731(89)90332-4. [DOI] [PubMed] [Google Scholar]
- 29.Telang NT, Suto A, Wong GY, Osborne MP, Bradlow HL. Induction by estrogen metabolite 16 alpha-hydroxyestrone of genotoxic damage and aberrant proliferation in mouse mammary epithelial cells. J Natl Cancer Inst. 1992 Apr 15;84(8):634–638. doi: 10.1093/jnci/84.8.634. [DOI] [PubMed] [Google Scholar]
- 30.Grabrick DM, Vierkant RA, Anderson KE, Cerhan JR, Anderson VE, Seller TA. Association of correlates of endogenous hormonal exposure with breast cancer risk in 426 families (United States) Cancer Causes Control. 2002 May;13(4):333–341. doi: 10.1023/a:1015292724384. [DOI] [PubMed] [Google Scholar]
- 31.Titus-Ernstoff L, Longnecker MP, Newcomb PA, Dain B, Greenberg ER, Mittendorf R, et al. Menstrual factors in relation to breast cancer risk. Cancer Epidemiol Biomarkers Prev. 1998 Sep;7(9):783–789. [PubMed] [Google Scholar]
- 32.Terry KL, Willet WC, Rich-Edwards JW, Hunter DJ, Michels KB. Menstrual cycle characteristics and incidence of premenopausal breast cancer. Cancer Epidemiol Biomarkers Prev. 2005;14(6):1509–1513. doi: 10.1158/1055-9965.EPI-05-0051. 00/01; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Clavel-Chapelon F E3N-EPIC Group. Differential effects of reproductive factors on the risk of pre- and postmenopausal breast cancer. Results from a large cohort of French women. Br J Cancer. 2002;86(5):723–727. doi: 10.1038/sj.bjc.6600124. 00/04; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wu AH, Ziegler RG, Pike MC, Nomura AM, West DW, Kolonel LN, et al. Menstrual and reproductive factors and risk of breast cancer in Asian-Americans. Br J Cancer. 1996 Mar;73(5):680–686. doi: 10.1038/bjc.1996.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Faupel-Badger JM, Fuhrman BJ, Xu X, Falk RT, Keefer LK, Veenstra TD, et al. Comparison of liquid chromatography-tandem mass spectrometry, RIA, and ELISA methods for measurement of urinary estrogens. Cancer Epidemiol Biomarkers Prev. 2010 Jan;19(1):292–300. doi: 10.1158/1055-9965.EPI-09-0643. [DOI] [PMC free article] [PubMed] [Google Scholar]