A growing body of evidence suggests that being physically active reduces a woman's risk for developing breast cancer (18). To date, the mechanism for this reduction in risk remains unknown. One biologically plausible hypothesis that has been suggested (18) is that exercise results in lower estrogen levels, and estrogens are strongly implicated in the etiology of breast cancer (14,29). In general, however, the levels of physical activity that have been observed to result in breast cancer risk reduction are unlikely to result in decreases in circulating estrogens (17,19). Moreover, occupations that require more physical activity also have been associated with decreased breast cancer risk (19) and are unlikely to result in a decrease in circulating estrogen concentrations. It is also now generally accepted that exercise interventions that lead to changes in circulating estrogens in premenopausal women must be accompanied by an energy deficit such that the energy expenditure of exercise is not matched by increased food intake (8,13). Thus, exercise itself is not viewed as a mechanism to modulate the central control of the reproductive axis (25) and the ovarian output of estrogen.
Given that the levels of physical activity reported to decrease breast cancer risk are unlikely to incur an energy deficit and are, therefore, unlikely to result in decreased circulating estrogens, it may be speculated that exercise, with or without energy deficiency, exerts more subtle changes in estrogen metabolism-that is, an increase in the proportion of estrogen metabolized through the 2-hydroxylation pathway and a reduction through the 16α-hydroxylation pathway-and perhaps it is this change that results in lowered breast cancer risk. Estrogen is primarily metabolized in humans through these two major competing pathways to yield two hydroxylated metabolites (41). These two metabolites, 2-hydroxyestrone (2-OHE1) and 16α-hydroxyestrone (16α-OHE1), exhibit very different biological activities. 16α-OHE1 has been shown to be estrogenic (15,37), is able to form covalent bonds with the estrogen receptor (34), and has been associated with breast cancer in humans (35) and mammary tumor development in animal models (6). 2-OHE1, in contrast, seems to display no, or antiestrogenic, activity (32,37), and increased levels have been associated with decreased breast cancer incidence in animal models (7). Further, lifestyle interventions that have been implicated in reducing breast cancer risk, such as altering dietary consumption of cruciferous vegetables (16) and soy (24,39), seem to increase 2-hydroxylation and the ratio of 2-OHE1/16α-OHE1 (2/16).
Because of their competitive production and their different biological properties, it has been proposed that the ratio of 2/16 may be a marker for breast cancer risk (5), with a higher 2/16 being associated with a decreased risk. This relationship has been found in many (22,23,26,27,40) but not all studies (10,36). We propose that exercise and/or energy deficiency increases the ratio of 2/16, and this, in turn, is a factor in the physical activity-mediated reduction in breast cancer risk.
Early studies (31,33) have suggested that athletes have higher 2-OHE1 levels than sedentary women. Further, a case study that involved brief bouts of exercise training reported increases in estrogen 2-hydroxylation (12). More recently, Bentz et al. (4) have reported that premenopausal women with higher self-reported physical activity levels had higher 2/16 levels than women who reported less physical activity. Campbell et al. (9), however, found that maximal oxygen consumption, a measure of aerobic fitness, was unrelated to 2-OHE1, 16α-OHE1, or the ratio of 2/16. To date, only one study has been published (1) in which estrogen metabolites were measured in response to an exercise intervention. In that study, Atkinson et al. (1) report that estrogen metabolism was generally unaltered in postmenopausal women who exercised 5 d·wk−1 for 1 yr, although changes in intraabdominal fat tended to be associated with changes in 2-OHE1 and 2/16.
To our knowledge, there are no data on the effects of exercise training, with or without weight loss, on estrogen metabolites in premenopausal women. Although the majority of breast cancers develop in women after menopause, the potential to impact risk factors earlier in life may have important implications for lifelong risk. Therefore, the purpose of the present study was to determine the effect of moderate exercise training combined with calorie restriction on 2-OHE1, 16α-OHE1, and the ratio of 2/16. The exercise training and calorie-restriction program was designed so that the subjects would lose body weight and body fat and would increase aerobic capacity (V˙O2max).
MATERIALS AND METHODS
Experimental design overview.
This study was part of a prospective study designed to test the effects on the menstrual cycle of a 4-month intervention of moderate aerobic exercise combined with calorie restriction. The primary study was aimed at understanding the impact of energy deficiency on circulating estrogens and biomarkers of breast cancer. The study was conducted through the Penn State University general clinical research center (GCRC) during 3 yr. All subjects were studied for a baseline period of one menstrual cycle, followed by four menstrual cycles during which a combination diet and exercise intervention occurred. Each subject participated in supervised exercise sessions throughout the 4-month intervention. The exercise intervention was designed to produce an increase in calorie expenditure of 20% of a woman's eucaloric intake. Similarly, the subjects were prescribed a diet equivalent to a 20-35% reduction in daily caloric intake. Repeated measurements of body weight, body composition, fitness, caloric intake, and physical activity were performed; menstrual status was documented via daily urine collections and menstrual records. Assays of 2-OHE1 and 16α-OHE1 were performed on urine samples from each menstrual cycle during the midfollicular and midluteal phases.
Screening and recruitment.
Recruitment was from the local area and the Penn State University community by newspaper, television, radio, and bulletin board an-nouncements. Recruitment occurred on a rolling basis, and subjects began the study at all times of the year. Inclusion in the study was based on the following criteria: 1) no history of serious medical conditions; 2) no current history of depression, disordered eating, or other affective disorders; 3) age 25-40, 4) weight 50-90 kg; 4) body fat 15-45%; 5) body mass index (BMI) 18-35 kg·m−2; 6) nonsmoking; 7) no medication use that would alter metabolic or reproductive hormone levels; 8) no significant weight loss/gain (± 2.3 kg) in the last year; 9) less than 1 h of purposeful aerobic exercise per week; 10) gynecological age ≥ 13 yr, 11) not taking hormonal contraceptives for the past 6 months; 12) eumenorrheic with ovulatory menstrual cycles; and 13) suitable candidate for an exercise study. Each subject was informed of the purpose, procedures, and potential risks of participation in the study before signing an informed consent approved by the university biomedical institutional review board.
After subjects met initial eligibility criteria, and during the baseline menstrual cycle, they completed surveys to assess demographics, medical history, and current and past menstrual history. To ensure that they were good candidates for a study that involved control over food intake and exercise, all subjects completed the Beck Depression Inventory (2), the Eating Disorder Inventory (20), a questionnaire regarding food preferences and food allergies, and a survey of their habitual physical activity.
During the baseline menstrual cycle in the early follicular phase, each subject visited the GCRC after an overnight fast to provide a blood sample via venipuncture, to be analyzed for a complete blood count (CBC), basic chemistry panel, and endocrine screen. Candidates were eliminated if the results of the endocrine screening, chemistry panel, or CBC were out of the normal range. During a subsequent visit to the GCRC, each subject had a physical examination by a GCRC clinician. Each subject also underwent a clinical interview under the supervision of a clinical psychologist, to determine mental health status. On a separate visit, subjects were also screened by a GCRC registered dietitian, who examined a 3-d record of each subject's diet and went over the dietary expectations of the study. Overall, 18 subjects who had signed informed consents were eliminated for the following reasons: medical, cycle length during baseline was outside criteria, progesterone levels were not indicative of ovulation, elevated prolactin levels were detected with the endocrine screen blood sample, and values for body composition and body weight parameters were outside the inclusion criteria.
Thirty-one subjects began the 4-month intervention. Subjects were mostly Caucasian (24/31), but there was some representation from minority groups: three were Asian, one was Hispanic, and three were from other minority groups. Of these 31 subjects, seven (21%) were eliminated or dropped out after the intervention began because of medical reasons, time constraints, pregnancy, visa restrictions, or noncompliance. Twenty-four women completed the study. A post hoc analysis of those who completed the study versus those who began but did not complete the study revealed that those who did not complete the study had higher baseline BMI (nonfinishers = 26.6 ± 1.4 kg·m−2 vs finishers 23.6 ± 0.5 kg·m−2; P = 0.04), percent body fat (nonfinishers = 36.6 ± 2.0% vs finishers 31.6 ± 1.01%; P = 0.03), and fat mass (nonfinishers = 26.1 ± 2.9 kg vs finishers 20.2 ± 1.1 kg; P = 0.03). Finishers and nonfinishers did not differ in initial body weight, age, height, V˙O2max, or ethnicity.
Baseline aerobic capacity (V˙O2max) was determined with a maximal graded exercise test on a treadmill, using indirect calorimetry. Body composition was determined by hydrostatic weighing after correction for residual lung volume. Body weight was measured using a digital scale to the nearest 0.01 kg (Seca scale, Hamburg, Germany) on the same day that body composition was determined, and twice per week throughout the intervention. All body weight measurements were obtained the subject wearing shorts and a T-shirt, without shoes.
Menstrual status and reproductive hormones.
Menstrual status was determined by self-reported dates for menses and other menstrual characteristics, as recorded by subjects on menstrual calendars. Daily urine samples (first morning void) were collected every day during the baseline cycle and the subsequent four intervention cycles. During the baseline period, ovulatory status was determined by documentation of a midcycle LH surge (First Response, Church and Dwight, 2004) and confirmation of a luteal rise in serum progesterone as determined from three midluteal blood samples, typically on days 21, 23, and 25. During the baseline cycle and during the last cycle of the intervention, blood samples were collected via venipuncture after an overnight fast every 2-3 d across each subject's menstrual cycles. The average of these serum samples was computed for each subject for each of these two cycles. Serum estradiol was measured in the laboratory, using radioimmunoassay procedures (Diagnostic Products Inc, Los Angeles, CA; Coat ACount Estradiol6). Intraassay coefficient of variation was 7.6%, and interassay coefficient of variation was 15.3%. Assay sensitivity was 7.4 pg·mL−1.
Dietary intake during the intervention.
The target macronutrient composition for all subjects during the intervention was 55% carbohydrate, 30% fat, and 15% protein. Using the average total daily caloric intake from each subject's 3-d diet log, in combination with the Harris-Benedict equation as adjusted for daily physical activity (21), an estimate of the eucaloric (weight maintenance) energy needs of each subject was calculated. To promote weight loss, this level of dietary intake was reduced by 20-35%, and this new level became the target daily caloric intake throughout the intervention. Before the intervention, the GCRC dietitian taught all subjects how to use the food exchange system (American Diabetes Association, 2003 Edition, Chicago, IL) to achieve their target caloric intake and prescribed macronutrient composition. Subjects kept track of their daily food exchanges by recording them on monitoring forms for 7 d at a time, every 2 wk, and they met with the GCRC dietitian every other week to review these records and discuss their progress. Estimates of daily caloric intake from food exchange monitoring sheets were determined by the dietitian using Dietitian's Assistant v.2.99 (Compu-Cal Inc., Olympia, WA). Once per menstrual cycle during the intervention, subjects also completed a 3-d diet log. Total calories and macronutrient content were determined using Nutritionist Pro (First Data Bank, Indianapolis, IN). In addition, servings of soy-containing foods and cruciferous vegetables were quantified at baseline and during the last month of the intervention. If difficulties arose with subjects meeting their targeted caloric intake, the dietitian counseled them on how to overcome these challenges. If subjects were not losing weight, appropriate changes to the target caloric intake were made. Dietary counseling sessions included discussion of food education modules, including shopping tips, low fat/low calorie food, food preparation, dining out, iron, calcium, fiber, and vitamins in food. Subjects attended 98% of the nutrition counseling sessions.
Exercise training during the intervention.
Subjects attended supervised exercise sessions monitored by a head trainer and several personal trainers who had experience in fitness assessment and personal training. Workouts were conducted four times per week and consisted of a 5-min warm-up and then approximately 40-90 min of aerobic activity at a heart rate 60-90% (average 79 ± 1%) of the maximal heart rate obtained from the V˙O2max test, then by a 10-min cool-down. Modes of aerobic activity included treadmill walking, stationary cycling, and stair stepping. Average attendance was 96 ± 1% of the total workouts; they attended an average of 3.6 workouts per week. The duration of exercise for each individual was equal in minutes to that required to expend a target amount of calories, determined to be 20% of the subjects' baseline eucaloric intake. For example, if a subject's predetermined eucaloric intake was 2000 calories, then the subject's exercise calorie-expenditure target was 400 calories. The total amount of calories expended during each exercise session was measured using the OwnCal feature on the Polar S610 heart rate monitor (Polar Electro Oy, Kempele, Finland). Throughout the study, the heart rate monitors were continually reinitialized with the most recent values of weight, maximum heart rate, maximal aerobic capacity, and age.
Urinary estrogen metabolite assays.
First void of the morning urine samples were collected and brought to the lab on ice. First-void samples have been reported to not differ from 24-h collections and are used to facilitate subject compliance (38). Boric acid (0.17 mg·mL−1) was added to the urine samples to preserve the estrogen metabolites, which were then refrigerated for no more than 4 h. Individual aliquots of 1-2 mL were frozen at −80°C within 4 h and were stored until analysis. Urinary estrogen metabolites in the midfollicular and midluteal phases were measured for each subject during the baseline cycle and during each cycle in the intervention. Midfollicular samples represented the midpoint day between day 1 of menses and the actual day of ovulation. Midluteal samples represented the midpoint day between the day after ovulation and the day before the subsequent menses. Samples were shipped on dry ice from Penn State to the AMC Cancer Research Center, where they were stored at −80°C until all samples from a given subject were available to be assayed.
Urinary estrogen metabolite concentrations were measured using a commercially available, competitive, solid-phase, enzyme-linked immunoassay (ESTRAMET, ImmunaCare, Corp., Bethlehem, PA), and the ratio of 2/16 was computed. The frozen urine samples were brought to room temperature before testing. Samples were diluted 1:4 before testing, with manufacturer-supplied diluent. In urine, 2-OHE1 and 16α-OHE1 are found in the glucoronide conjugate form and require removal of the sugar moiety before recognition by the monoclonal antibodies. Samples were deconjugated with β-glucoronidase and arylsulphatase and were then neutralized. Samples were incubated for 3 h at room temperature and then kinetically read every 2 min for 20 min, using a Thermomax Microplate Reader (Molecular Devices, Sunnyvale, CA). Estrogen metabolite values were determined from a calibration curve derived from six standards supplied with the kit (0.625-15.0 ng·mL−1). All samples, controls, and standards were assayed in triplicate, and all of the samples from a given subject were batch assayed to minimize interassay variability. In-house and manufacturer-supplied controls were included in each of the assays performed. Any sample outside of the range of the standard curve or with a coefficient of variation greater than 10% was reassayed. Twelve percent of the samples needed to be reassayed. Intraassay and interassay CV for 2-OHE1 were 4.1 and 9.4%, respectively; for 16α-OHE1, they were 4.1 and 7.3%, respectively.
Urinary 2-OHE1 and 16α-OHE1 were normalized to urinary creatinine concentration and expressed in nanograms per milligram of creatinine. Urinary creatinine was measured in duplicate using a Diagnostic Chemicals Limited Assay (Diagnostic Chemical Limited, PEI, Canada). The ratio of 2/16 was computed by dividing the absolute concentration of 2-OHE1 by 16α-OHE1. No units are used to express the ratio.
Data analysis and statistics.
Results were analyzed using SPSS statistical software, version 13.0 (SPSS, Inc., Chicago, IL). Data were analyzed with paired Student's t-tests, one-way ANOVA, and univariate repeated-measures ANOVA. When necessary, Greenhouse-Geisser correction factors were used when the sphericity assumption was violated. Associations between fitness and body composition parameters and estrogen metabolites were determined using Pearson product-moment correlation coefficients. A significance level of P ≤ 0.05 was preset to indicate statistical significance. All data are presented as means ± standard error (SE) unless otherwise specified.
Baseline descriptive characteristics are presented in Table 1. Age was unrelated to any estrogen metabolite measure and was not used as a covariate in any subsequent analyses. Similarly, baseline V˙O2max was not associated with 2-OHE1, 16α-OHE1, or 2/16 (data not shown) and was not further incorporated into the analyses. The absolute concentrations of 2OHE1 and 16α-OHE1 were not found to be associated with any baseline body composition parameter (data not shown), with the exception of 2-OHE1 in the follicular phase being negatively correlated with baseline BMI values (r = −0.49; P = 0.02). Similarly, no relationship was found between the ratio of 2/16 in either phase with baseline percent body fat, weight, or BMI (data not shown).
The effects of the intervention on body composition, body weight, and fitness are also shown in Table 1. Significant improvements and changes were observed in all variables during the course of the intervention, with the exception of no change in lean body mass.
Daily caloric intake from the average of 3-d diet logs recorded during the baseline period and during the fourth menstrual cycle was significantly decreased (−22%) over time (P < 0.001) Overall macronutrient distribution during the intervention was 55% carbohydrate, 19% protein, and 29% fat. There was no change in the number of servings of soy products (0.85 ± 0.2 servings/3 d (baseline) vs 0.50 ± 0.2 servings/3 d (month 4 of intervention)) or cruciferous vegetables (0.04 ± 0.2 servings/3 d (baseline) vs 0.12 ± 0.12 servings/3 d (month 4)).
Estrogen metabolite values at baseline and during the 4-month intervention period are presented in Table 2. Data from the follicular and luteal phases are presented for 2-OHE1 and 16α-OHE1. In addition to the reported 2/16 ratio for each phase, we also present an average of the two ratios per cycle. No significant difference in the ratio was observed between the follicular and luteal phases during any menstrual cycle (P = 0.20-0.99).
2-OHE1 and 16α-OHE1 concentrations were significantly higher during the luteal phase than during the follicular phase (Table 2). This would be expected, because the concentrations of the metabolites parallel the changes in estradiol during the menstrual cycle. The ratio of 2/16 did not vary as a function of the phase of the menstrual cycle. Baseline menstrual cycle length was 29.7 ± 0.5 d (range 24-36 d) and did not change significantly across the intervention (P > 0.05). Average serum estradiol did not change significantly during the intervention (87.2 ± 5.1 pg·mL−1 (baseline) vs 80.1 ± 5.7 pg·mL−1 (postintervention); P = 0.20).
No effect of exercise on estrogen metabolism was observed over time, with the exception of a significant polynomial (cubic) effect on 16α-OHE1 in the luteal phase and a similar effect that approached significance (P = 0.06) for 2-OHE1 in the luteal phase. This time effect was a function of an elevation in both metabolites during the third exercise month in the luteal phase. We have no explanation for this finding. This was not seen in other variables that were measured.
Interestingly, there was a significant association between a woman's baseline ratio of 2/16 and her overall percent change in 2/16 during the intervention (r = −0.68; P < 0.001). This finding, coupled with evidence in the literature (3,11,24,28,30) and from our lab (unpublished data) suggesting that some women respond to interventions with an increase in 2/16 (responders) whereas others do not (nonresponders), led us to analyze the data further. We divided the subjects into tertiles according to their baseline urinary 2/16 value. Three distinct groups were identified, with average 2/16 ratios of 0.91 ± 0.13 (tertile 1), 1.97 ± 0.13 (tertile 2), and 2.95 ± 0.07 (tertile 3). The subjects in each tertile did not differ according to initial body weight, body fat, or fitness (Table 3). Further, the amount of weight or percent fat lost with the intervention did not differ among the three groups (Table 3). Similarly, there was no difference among the three tertiles in calories consumed or energy deficit experienced. Average exercise intensity during the intervention, expressed as percentage of maximal heart rate, was significantly lower in the women grouped in tertile 2 compared with tertiles 1 and 3 (Table 3). There were no differences between tertile groups in baseline menstrual cycle lengths or cycle length changes across the intervention (data not shown).
Table 4 provides information on the percent and absolute change in 2/16 for the three tertiles. Women with the lowest baseline 2/16 ratio had significantly greater percent increases in 2/16 during the exercise intervention than did women in the second and third tertiles, who did not differ from each other. Given that the percent change in 2/16 is a function of the initial values, we analyzed the data to determine whether the change in 2/16 among tertiles persisted when we looked at absolute changes in 2/16. With the exception of the follicular phase (P = 0.11), we again saw that women in the lowest tertile had significantly larger absolute increases in 2/16 than women in the second and third tertiles.
Evidence continues to accumulate that being physically active reduces a woman's risk for developing breast cancer; yet, to date, the mechanism remains unknown. We hypothesized that exercise training coupled with calorie restriction and subsequent weight loss would alter estrogen metabolism in a direction that would be associated with proportionately more estrogen 2-hydroxylation and less 16α-hydroxylation.
On the surface, we found that an intervention program of moderate-intensity exercise, coupled with calorie restriction and weight loss, did not result in alterations in baseline estrogen metabolism that would suggest decreased breast cancer risk. This is consistent with the data of Atkinson et al. (1), who examined the effect of 12 months of moderate exercise on estrogen metabolites in postmenopausal women. Pasagian-McCauley et al. (28) also have reported that women counseled to decrease weight and dietary fat content and to increase physical activity increased their 2/16 ratio during a 20-week intervention, but the control group did as well, and thus there was no difference between the groups.
When we examined the present data more closely, however, we found that only women with low baseline 2/16 ratios had significant increases in their 2/16 ratio in response to the intervention. This was true when expressed as a relative percent change or as an absolute change in 2/16. To our knowledge, none of the other physical activity intervention studies have determined whether a woman's baseline ratio was predictive of her change in 2/16 with the intervention (1,28). Pasagian-McCauley et al. (28) did group women into tertiles according to baseline urinary 2/16 ratios, but they used those groupings only to compare baseline measures of weight, physical activity, and body composition. Consistent with our findings, they observed no differences among the tertiles in body weight or body fat as functions of the baseline 2/16 ratio. Interestingly, we did find that tertiles 1 and 3 had higher average exercise intensities compared with tertile 2. It is unclear whether this statistically significant difference in exercise intensity is physiologically relevant, because exercise intensity in all groups was sufficient to promote increases in fitness. Additionally, a higher exercise intensity was not exclusive to the lowest tertile group; thus, it seems unlikely that changes in the 2/16 ratio were associated with exercise intensity.
To determine whether low baseline 2/16 ratios have been associated with greater changes in 2/16 with other interventions, we evaluated the data of Bell et al. (3), who conducted a placebo-controlled trial of indole-3-carbinol to treat cervical cancer. When we compared their subjects who would have been included in our first tertile (2/16 ≤ 1.3; N = 12) with the remainder of the subjects (2/16 > 1.3; N = 5), the changes in 2/16 tended to be greater in women with low 2/16 ratios (percent change = 56%; absolute change = 0.42) compared with women with higher 2/16 ratios (−2% and +0.01, respectively). Others have reported a variability in response to interventions, with subjects showing increases, no change, and decreases in 2/16. Whether these were a function of baseline 2/16, however, was not discussed (11,24,28,30).
It is impossible to determine whether the changes in 2/16 observed in the present study were a function of exercise, calorie restriction, weight/fat loss, or some combination of these factors. The subjects lost body weight, body fat, and increased their fitness levels. This effect was very consistent among subjects; only one subject exhibited an increase in body weight (caused solely by an increase in fat-free mass), only one subject exhibited a slight increase in body fat, and only one subject experienced a decline in V˙O2max. However, the women in the lowest tertile of 2/16 who exhibited the greatest change in 2/16 did not lose more weight or fat or become more fit than women in the other tertiles. Moreover, women in the lowest tertile of 2/16 were no fatter, did not weigh more, and were no less fit at baseline than women in any other tertile. This would suggest that it is not the initial body weight/fat level or the magnitude of the change in body composition that determines the amount of change in 2/16. When women were grouped in tertiles according to the amount of weight or fat that was lost, no differences in the change in the 2/16 ratio were found among the groups. Similarly, when grouped by tertiles of energy deficit, no difference in changes in 2/16 was observed. In previous work, loss of body weight of 10 lb resulted in an increase in 2/16, but an increase in 2/16 also occurred in the control group that did not lose weight (28). Subjects in the study by Atkinson et al. (1) lost, on average, 1.3 kg, but this was not associated with any change in estrogen metabolism.
Taken together, these data suggest that a loss of body weight or fat is not necessarily associated with alterations in estrogen metabolism. Rather, it seems to be a function of a woman's initial estrogen-metabolism profile that determines whether she responds to a particular intervention. The current data may suggest a threshold effect. Women in the second and third tertiles of baseline 2/16 (range 1.45-3.22) did not differ from each other in terms of their changes in 2/16 in response to the intervention. Our data suggest that women with higher 2/16 ratios may not exhibit increases in 2/16 in response to an exercise/calorie-restriction intervention. It may be speculated that these women are already at lower risk for developing breast cancer because of their preexisting estrogen-metabolism patterns.
On the basis of the current data, it may be speculated that by measuring urinary estrogen metabolites and computing the 2/16 ratio, we may be able to identify women at higher risk for developing breast cancer who would most likely benefit from targeted interventions. If women with low 2/16 ratios are at higher risk for breast cancer, as has been proposed (5), and if increasing 2/16 results in decreased risk (22,26,27,40), then using a lifestyle program that is recognized to promote general health may also favorably impact a woman's risk for developing breast cancer by altering how she metabolizes estrogen.
To our knowledge, this is the first prospective trial designed to assess the effects of exercise and calorie restriction on estrogen metabolism in premenopausal women. Limitations to the study include the small sample sizes and the lack of a control group. Although initial efforts were made to recruit a nonexercising control group, this effort was abandoned because potential subjects were reluctant to risk being put in a control group that would not exercise or lose weight and yet would be subjected to frequent biological and physiological measurements during the 5-month period. On the other hand, the major strengths of this study include the rigorous control of each individual's exercise and calorie-restriction program and the high compliance rate.
In conclusion, 4 months of exercise and calorie restriction did not, at the surface, seem to significantly alter estrogen metabolism. However, when women were grouped according to their initial estrogen metabolite profiles, we observed that women with low urinary 2/16 ratios had large increases in 2/16 in response to a moderate exercise and calorie-restriction intervention.
This work was possible through funding by the Cancer Research and Prevention Foundation (K. C. Westerlind) and the Department of Defense DAMD 17-01-1-0361 (N. I. Williams). The authors wish to gratefully acknowledge all the subjects who participated in this study and technical staff Regan Story, Susan Bell, and Ann Albert for their contributions.
1. Atkinson, C., J. W. Lampe, S. S. Tworoger, et al. Effects of a moderate intensity exercise
intervention on estrogen metabolism in postmenopausal women. Cancer Epidemiol. Biomarkers Prev
. 13:868-874, 2004.
2. Beck, A. T., C. H. Ward, M. Mendelson, J. Mock, and J. Erbaugh. An inventory for measuring depression. Arch. Gen. Psychiatry
3. Bell, M. C., P. Crowley-Nowick, H. L. Bradlow, et al. Placebo-controlled trial of indole-3-carbinol in the treatment of CIN. Gynecol. Oncol
. 78:123-129, 2000.
4. Bentz, A. T., C. M. Schneider, and K. C. Westerlind. The relationship between physical activity and 2-hydroxyestrone, 16α-hydroxyestrone and the 2/16 ratio in premenopausal women. Cancer Causes Control
5. Bradlow, H. L., D. L. Davis, G. Lin, D. Sepkovic, and R. Tiwari. Effects of pesticides on the ratio of 16α-/2-hydroxyestrone: a biological marker of breast cancer
risk. Environ. Health Perspect
. 103(Suppl 7):147-150, 1995.
6. Bradlow, H. L., R. J. Hershcopf, C. P. Martucci, and J. Fishman. Estradiol
16α-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
. 82:6295-6299, 1985.
7. Bradlow, H. L., N. T. Telang, D. W. Sepkovic, and M. P. Osborne. 2-hydroxyestrone: the "good" estrogen. J. Endocrinol
. 150:5259-5265, 1996.
8. Bullen, B. A., G. S. Skrinar, I. Z. Beitung, G. Von Mering, B. A. Turnbull, and J. W. McArthur. Induction of menstrual disorders by strenuous exercise
in untrained women. N. Engl. J. Med
. 312:1349-1353, 1985.
9. Campbell, K. L., K. C. Westerlind, V. J. Harber, C. M. Friedenreich, and K. S. Courneya. Association between aerobic fitness and estrogen metabolites in premenopausal women. Med. Sci. Sports Exerc
. 37:585-592, 2005.
10. Cauley, J. A., J. M. Zmuda, M. E. Danielson, et al. Estrogen metabolites and the risk of breast cancer
in older women. Epidemiol
. 14:635-640, 2003.
11. Dalessandri, K. M., G. L. Firestone, M. D. Fitch, H. L. Bradlow, and L. F. Bjeldanes. Pilot study: effect of 3.3′-diindolylmethane supplements on urinary hormone metabolites in postmenopausal women with a history of early-stage breast cancer
. Nutr. Cancer
12. DeCree, C., P. Ball, B. Seidlitz, G. Van Kranenburg, P. Geurten, and H. A. Keizer. Effects of a training program on resting plasma 2-hydroxycatecholestrogen levels in eumenorrheic women. J. Appl. Physiol
. 83:1551-1556, 1997.
13. DeSouza, M. J., and N. I. Williams. Physiological aspects and clinical sequelae of energy deficiency and hypoestrogenism in exercising women. Hum. Reprod. Update
14. Eliassen, A. H., S. A. Missmer, S. S. Tworoger, and S. E. Hankinson. Endogenous steroid hormone concentrations and risk of breast cancer
: does the association vary by a woman′s predicted breast cancer
risk? J. Clin. Oncol
. 24:1823-1830, 2006.
15. Fishman, J., and C. Martucci. Biological properties of 16α-hydroxyestrone: implications in estrogen physiology and pathophysiology. J. Clin. Endocrinol. Metab
. 51:611-615, 1980.
16. Fowke, J. H., C. Longcope, and J. R. Herbert. Brassica
vegetable consumption shifts estrogen metabolism in healthy postmenopausal women. Cancer Epidemiol. Biomarkers Prev
. 9:773-779, 2000.
17. Friedenrech, C. M., K. S. Courneya, and H. E. Bryant. Relation between intensity of physical activity and breast cancer
risk reduction. Med. Sci. Sports Exercise
18. Friedenreich, C. M., and M. R. Orenstein. Physical activity and cancer prevention: etiologic evidence and biological mechanisms. J. Nutr
. 132:3456S-3464S, 2002.
19. Gammon, M. D., E. M. John, and J. A. Britton. Recreational and occupational physical activities and risk of breast cancer
. J. Natl. Cancer Inst
. 90:100-117, 1998.
20. Garner, D. M., and M. P. Olmstead. Manual for Eating Disorder Inventory
. Odessa, FL: Psychological Assessment Resources, Inc., 1984.
21. Harris, J. A., and F. G. Benedict. A Biometric Study of the Basal Metabolism in Man
. Washington, DC: Carnegie Institution of Washington, DC, 1919. Publication no. 279.
22. Kabat, G. C., C. J. Chang, J. A. Sparano, et al. Urinary estrogen metabolites and breast cancer
: a case-control study. Cancer Epidemiol. Biomarkers Prev
. 6:505-509, 1997.
23. Kabat, G. C., E. S. O'Leary, M. D. Gammon, et al. Estrogen metabolism and breast cancer
24. Lee-Jane, W. L., M. Cree, S. Josyula, M. Nagamani, J. J. Grady, and K. E. Anderson. Increased urinary excretion of 2-hydroxyestrone but not 16α-hydroxyestrone in premenopausal women during a soya diet containing isoflavones. Cancer Res
. 60:1299-1305, 2000.
25. Loucks, A. B., and J. R. Thuma. Lutenizing hormone pulsatility is disrupted at a threshold of energy availability in regularly menstruating women. J. Clin. Endocrinol. Metab
. 88:297-311, 2003.
26. Meilahn, E. N., B. De Stavola, D. S. Allen, et al. Do urinary oestrogen metabolites predict breast cancer
? Guernsey III cohort followup. Br. J. Cancer
27. Muti, P., H. L. Bradlow, A. Micheli, et al. Estrogen metabolism and risk of breast cancer
: a prospective study of the 2:16α-hydroxyestrone rat in premenopausal and postmenopausal women. Epidemiology
28. Pasagian-Macaulay, A., E. N. Meilahn, H. L. Bradlow, et al. Urinary markers of estrogen metabolism 2- and 16α-hydroxylation in premenopausal women. Steroids
29. Pike, M. C., D. V. Spencer, L. Dahmoush, and M. F. Press. Estrogens, progestogens, normal breast cell proliferation, and breast cancer
risk. Epidemiol. Rev
. 15:17-35, 1993.
30. Reed, G. A., K. S. Peterson, H. J. Smith, et al. A phase I study of indole-3-carbinol in women: tolerability and effects. Cancer Epidemiol. Biomarkers Prev
. 14:1953-1960, 2005.
31. Russell, J. B., D. Mitchell, P. I. Musey, and D. C. Collins. The relationship of exercise
to anovulatory cycles in female athletes: hormonal and physical characteristics. Obstet. Gynecol
. 63: 452-456, 1984.
32. Schneider, J., M. M. Huh, H. L. Bradlow, and J. Fishman. Anti-estrogen activation of 2-hydroxyestrone on MCF-7 human breast cancer
cells. J. Biol. Chem
. 259:4840-4845, 1984.
33. Snow, R. C., R. L. Barbieri, and R. E. Frisch. Estrogen 2-hydroxylase oxidation and menstrual function among elite oarswomen. J. Clin. Endocrinol. Metab
. 69:369-376, 1989.
34. Swaneck, G. E., and J. Fishman. Covalent binding of the endogenous estrogen 16α-hydoxyestrone to estradiol
receptor in human breast cancer
cells: characterization and intranuclear localization. Proc. Natl. Acad. Sci
. 85:7831-7835, 1988.
35. Telang, N. T., D. M. Avelrod, G. Y. Wong, H. L. Bradlow, and M. P. Osborne. Biotransformation of estradiol
by explant culture of human mammary tissue. Steroids
36. Ursin, G., S. London, F. Z. Stanczyk, et al. Urinary 2-hydroxyestrone/16α-hydroxyestrone ratio and risk of breast cancer
in postmenopausal women. J. Natl. Cancer Inst
. 91:1067-1072, 1999.
37. Westerlind, K. C., K. J. Gibson, P. Malone, G. L. Evans, and R. T. Turner. Differential effects of estrogen metabolites on bone and reproductive tissues of ovariectomized rats. J. Bone Miner. Res
. 13:1023-1031, 1998.
38. Westerlind, K. C., K. J. Gibson, and P. Wolfe. The effect of diurnal and menstrual cyclicity and menopausal status on estrogen metabolites: implications for disease-risk assessment. Steroids
39. Xu, X., A. M. Duncan, K. E. Wangen, and M. S. Kurzer. Soy consumption alters endogenous estrogen metabolism in postmenopausal women. Cancer Epidemiol. Biomarkers Prev
. 9: 781-786, 2000.
40. Zheng, W., L. Dunning, F. Jin, and J. Holtzman. Correspondence re: G. C. Kabat et al, Urinary estrogen metabolites and breast cancer
: a case control study. Cancer Epidemiol. Biomarkers Prev
. 78:85-86, 1998.
41. Zhu, B. T., and A. H. Conney. Functional role of estrogen metabolism in target cells: review and perspectives. Carcinogenesis