Estrogen and peptide YY are associated with bone mineral density in premenopausal exercising women

Bone 49 (2011) 194–201

Contents lists available at ScienceDirect

Bone j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b o n e

Estrogen and peptide YY are associated with bone mineral density in premenopausal exercising women J.L. Scheid, R.J. Toombs, G. Ducher, J.C. Gibbs, N.I. Williams, M.J. De Souza ⁎ Women's Health and Exercise Laboratory, Department of Kinesiology, Penn State University, University Park, PA, USA

a r t i c l e

i n f o

Article history: Received 8 October 2010 Revised 11 March 2011 Accepted 14 April 2011 Available online 28 April 2011 Edited by: Toshio Matsumoto Keywords: Estrogen Peptide YY Exercise Bone mineral density Women

a b s t r a c t Background: In women with anorexia nervosa, elevated fasting peptide YY (PYY) is associated with decreased bone mineral density (BMD). Prior research from our lab has demonstrated that fasting total PYY concentrations are elevated in exercising women with amenorrhea compared to ovulatory exercising women. Purpose: The purpose of this study was to assess the association between fasting total PYY, average monthly estrogen exposure and BMD in non-obese premenopausal exercising women. Methods: Daily urine samples were collected and assessed for metabolites of estrone 1-glucuronide (E1G) and pregnandiol glucuronide (PdG) for at least one menstrual cycle if ovulatory or a 28-day monitoring period if amenorrheic. Fasting serum samples were pooled over the measurement period and analyzed for total PYY and leptin. BMD and body composition were assessed by dual-energy X-ray absorptiometry. Multiple regression analyses were performed to determine whether measures of body composition, estrogen status, exercise minutes, leptin and PYY explained a significant amount of the variance in BMD at multiple sites. Results: Premenopausal exercising women aged 23.8 ± 0.9 years with a mean BMI of 21.2 ± 0.4 kg/m2 exercised 346 ± 48 min/week and had a peak oxygen uptake of 49.1 ± 1.8 mL/kg/min. Thirty-nine percent (17/44) of the women had amenorrhea. Fasting total PYY concentrations were negatively associated with total body BMD (p = 0.033) and total hip BMD (p = 0.043). Mean E1G concentrations were positively associated with total body BMD (p = 0.033) and lumbar spine (L2–L4) BMD (p = 0.047). The proportion of variance in lumbar spine (L2–L4) BMD explained by body weight and E1G cycle mean was 16.4% (R2 = 0.204, p = 0.012). The proportion of variance in hip BMD explained by PYY cycle mean was 8.6% (R2 = 0.109, p = 0.033). The proportion of variance in total body BMD explained by body weight and E1G cycle mean was 21.9% (R2 = 0.257, p = 0.003). Conclusion: PYY, mean E1G and body weight are associated with BMD in premenopausal exercising women. Thus, elevated PYY and suppressed estrogen concentrations are associated with, and could be directly contributing to, low BMD in exercising women with amenorrhea, despite regular physical activity. © 2011 Elsevier Inc. All rights reserved.

Introduction Exercise-related menstrual disturbances are often observed in women who participate in physical activity ranging from the recreational to the competitive level, and from exercise of moderate to strenuous intensity [1,2]. Amenorrhea represents the most extreme presentation of menstrual irregularity and is described as “functional hypothalamic amenorrhea” (FHA) since the disruption occurs at the level of the hypothalamus, concomitant with an energy deficiency, leading to chronically suppressed reproductive function and reduced circulating estrogen concentrations [3,4]. Alterations in metabolic and endocrine homeostasis indicative of an energy deficiency in exercising women with FHA include suppressed resting energy expenditure (REE) [5–7], reduced concentrations of total triiodothyronine (TT3) [5,6,8] ⁎ Corresponding author at: Department of Kinesiology, Penn State University, University Park, PA 16802, USA. Fax: + 1 814 865 4602. E-mail address: [email protected] (M.J. De Souza). 8756-3282/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2011.04.011

insulin-like growth factor-1 (IGF-1) [9], and insulin [7], and elevated concentrations of growth hormone (GH) [10], ghrelin [8,11,12], and cortisol [2,9,13]. The metabolic and endocrine alterations demonstrated in exercising women with FHA are likely contributing to low bone mineral density (BMD) and consequently, leading to musculoskeletal consequences such as stress fractures and pathological bone loss [14]. In our laboratory, we have reported elevated peptide YY (PYY) concentrations in exercising women with FHA secondary to an energy deficit [15]. Peptide YY is a gastrointestinal peptide secreted from the endocrine L cells of the ileum of the intestine and appears to be involved with appetite suppression, satiety, and energy homeostasis [19–22]. PYY acts centrally to inhibit neuropeptide Y (NPY) and activate the proopiomelanocortin (POMC) neurons in the arcuate nucleus resulting in a reduction in food intake [19]. Interestingly, while exercising women and adolescent athletes with FHA demonstrate elevations in fasting PYY concentrations [15,23], women with a more severe energy deficiency, i.e. women and adolescent girls with anorexia nervosa, also demonstrate elevated fasting PYY concentrations [16–18]. These findings

J.L. Scheid et al. / Bone 49 (2011) 194–201

indicate that an association exists between elevated PYY concentrations and energy deficiency and suggests the need to further explore the mechanism for alterations in PYY with energy deficiency and consequently, the suppression of reproductive function. Exercising women with FHA and women with anorexia nervosa also demonstrate energy deficiency-related decreases in BMD and estrogenrelated disruptions in bone metabolism [24–28]. A previous study from our lab [26] demonstrated that women with FHA experience increased bone resorption related to an estrogen deficiency and decreased bone formation related to an energy deficiency that translate to lower BMD in these women. It is of interest that concentrations of gut peptides, such as elevated PYY and ghrelin are also associated with suppressed BMD [25,29,30] and suppressed markers of bone formation [16]. In women with anorexia nervosa, an energy deficient population with elevated PYY concentrations, Utz et al. [25] observed a negative correlation between mean PYY concentrations (samples over 12 h) and BMD. Thus an elevation in PYY associated with an energy deficiency may be directly participating in a central mechanism promoting the suppression of bone formation and as a result, a decrease in BMD in energy deficient populations, such as women with anorexia nervosa and, potentially, exercising women with FHA. The purpose of this study was to assess the association between fasting total PYY, average monthly estrogen exposure and BMD in nonobese premenopausal exercising women. We hypothesized that elevated total PYY concentrations would be negatively associated with BMD, while estrogen exposure would be positively associated with BMD. Methods Experimental design

195

participation in the study before signing an informed consent approved by the University of Toronto and Penn State University Institutional Review Board. Screening procedures During an initial visit, study details and participation requirements were explained, and written informed consent was obtained. Once consent was obtained, height and weight were measured, and subjects completed questionnaires to assess demographics, medical history, exercise history, menstrual history, eating behaviors, bone health, and mental health. A physical exam was performed on all subjects by an onsite clinician to determine overall health and check for physical symptoms of PCOS such as acne or hirsutism. In addition, a fasting blood sample was analyzed for a complete blood count, basic chemistry panel, and an endocrine panel which included measures of luteinizing hormone (LH), follicle stimulating hormone (FSH), thyroid stimulating hormone (TSH), thyroxine (T4), prolactin (PRL), and dihydroepiandrosterone (DHEA) (Quest Diagnostics, Pittsburgh, PA), total testosterone and sex hormone binding globulin (SHBG) to rule out endocrine or metabolic disease or other illnesses. A clinical psychologist or licensed clinical social worker interviewed each subject to determine if she was suffering from major psychiatric disorders including depression or clinical eating disorders. Subjects met with a General Clinical Research Center (GCRC) registered dietitian after completing a 3-day diet log (2 weekdays and 1 weekend day) to discuss eating patterns and preferences. Additionally, dual-energy X-ray absorptiometry (DXA) scans of the total body, lumbar spine, and dual femur were performed to assess bone mineral density (BMD) and body composition (GE Lunar Prodigy and iDXA, Madison, WI). Group categorization

This investigation includes data from a prospective study designed to determine the impact of increased caloric intake on bone health and menstrual cyclicity in energy deficient, amenorrheic exercising women that took place at the University of Toronto and Penn State University. The current investigation includes cross-sectional data from 44 premenopausal exercising women who were monitored for at least one menstrual cycle if eumenorrheic or at least one 28-day monitoring period if not regularly cycling, and assessed for reproductive status, BMD, and metabolic hormones. Five of the 44 women were also included in an earlier publication [15] from our lab assessing PYY concentrations in exercising women. The reproductive profile evaluation included the evaluation of menstrual history, detection of the presence or absence of an luteinizing hormone (LH) peak, and quantification of daily urinary ovarian steroid metabolites, estrone 1glucuronide (E1G) and pregnanediol glucuronide (PdG). The metabolic hormones were measured after pooling two fasting samples and included PYY and leptin.

Exercising women were monitored for at least one menstrual cycle if eumenorrheic (cycle length 26–34 days) and for one 28-day monitoring period if amenorrheic. BMD and body composition was assessed once during the study period. Two fasting samples of blood were collected during each menstrual cycle or 28-day monitoring period. All data presented represent the mean of these repeated measurements.

Recruitment and medical screening

Menstrual characteristics

Volunteers were recruited by posters targeting physically active women for a study on women's health. Screening procedures included questionnaires on exercise history, eating behaviors, menstrual history, and self-reported medical health history. Inclusion criteria were: 1) no history of any serious medical conditions; 2) no current clinical diagnosis of an eating or psychiatric disorder; 3) age 18–35 years; 4) weight-stable (±2 kg) for the past six months; 4) non-smoking; 5) no medication use that would alter metabolic or reproductive hormone concentrations and 6) ≥150 min/wk self-reported purposeful exercise. Exclusion criteria were: 1) BMIN 25 kg/m2; 2) menstrual cycle length less than 26 days; 3) history of a clinical diagnosis of PCOS; 4) evidence of prolactin secreting tumors; and 5) evidence of premature failure. Sixty-six volunteers met the initial inclusion criteria and were monitored for one menstrual cycle or 28-day monitoring period. Each subject was informed of the purpose, procedures, and potential risks of

Menstrual status was assessed in all volunteers, defined in this study by classifying cycle length and presence or absence of menses, i.e., eumenorrheic, or amenorrheic, and by ovulatory status, i.e., ovulatory or anovulatory. Subject categorization to groups was confirmed by daily urinary hormone evaluations. Volunteers were considered amenorrheic if they failed to menstruate for a minimum of 3 consecutive months and eumenorrheic if menses occurred at regular intervals of 26–34 days [1,2,8]. Menstrual cycle length was defined as the number of days from day one of menses to the day before the first day of the next menses. The follicular phase length was defined as the number of days from the first day of menses up to and including the day of the LH surge [1,2]. The luteal phase was defined as the menstrual cycle length minus the length of the follicular phase [1,2]. Daily urine samples were assayed for LH, E1G and PdG to assess ovulatory status, and estrogen exposure and progesterone exposure,

After determination of menstrual status (eumenorrheic and ovulatory or amenorrheic) based on self reported menses and confirmation by daily urinary profiles of reproductive hormones, volunteers were grouped as follows: 1) exercising women with eumenorrhea, ovulatory menstrual cycles (ExOv; n =27) and 2) exercising women with hypothalamic amenorrhea (ExAmen; n= 17). Observational time periods

196

J.L. Scheid et al. / Bone 49 (2011) 194–201

assessed by E1G and PdG area under the curve (AUC). To determine estrogen exposure and progesterone exposure, E1G and PdG urinary metabolites were assessed using a modified trapezoidal integrated AUC method. Values for the menstrual cycle characteristics and reproductive hormones (E1G and PdG) were obtained from one baseline cycle. E1G and PdG cycle mean represent the mean E1G and PdG concentrations during the cycle or 28 day monitoring period. Ovulatory status was determined by urinary LH, identified as a LH peak after the midcycle E1G peak [1,2]. Specific hormonal criteria for detecting ovulation included an LH peak concentration above 25 mIU/mL, the E1G peak concentration above 35 ng/mL, and the peak PdG concentration above 5 μg/mL during the luteal phase [1,2,31,32]. An anovulatory cycle was defined as one of the following: 1) a cycle in which a minimal increase in E1G was observed concomitantly with a failure of LH to rise at midcycle, 2) when PdG concentration failed to increase during the luteal phase from a 5-day follicular phase baseline, or 3) when the peak PdG value was below 2.49 μg/mL [1,2,31]. Amenorrhea was confirmed by chronically suppressed E1G and PdG profiles. All cases of amenorrhea in this study were secondary amenorrhea and not previously associated with PCOS, as determined by medical history and androgen status to include a Free Androgen Index (FAI) b 3, which was calculated according to the following equation: FAI = 100 X [total testosterone (nmol/L) / serum hormone binding globulin (nmol/L)] [33]. An FAI of 3 was used as the cutoff because women diagnosed with PCOS often exhibit calculated FAIs ≥ 3 and women without PCOS exhibit calculated FAIs b 3 [34,35].

some biases in the total BMD, total BMC, total fat, and % fat relative to the magnitude of the variable. Equations were derived using simple linear regression to remove these biases and report the Prodigy values calibrated to the iDXA. Determination of exercise status and peak oxygen uptake All participants kept daily logs of purposeful physical activity to determine total exercise minutes per week. Reported physical activity represents the averaged physical activity during the monitoring period. Peak oxygen uptake (VO2 peak) was measured once during a progressive treadmill test to volitional exhaustion using open circuit spirometry as previously reported [5]. Blood sampling and storage Blood samples were collected twice during each menstrual cycle or 28-day monitoring period. All hormone measurements obtained from repeated blood samples on a given participant were pooled before analysis. Participants were instructed not to exercise or consume food within 12 h prior to blood sampling. Samples were allowed to clot for 30 min at room temperature (20–24 °C) and then centrifuged at 3225.6 g-force (3000 rpm) for 15 min at 4 °C. The serum was aliquoted into 2-mL polyethylene storage tubes and stored frozen at −80 °C until analysis. The serum samples were analyzed for leptin and PYY.

Urinary measurement of E1G and PdG

Serum hormone measurements

The secretion of E1G and PdG metabolites in the urine parallels serum concentrations of the parent hormones [36]. Microtiter plate competitive enzyme immunoassays were used to measure the urinary metabolites E1G and PdG. The E1G (R522-2) and PdG (R13904) assays use a polyclonal capture antibody supplied by Coralie Munro, University of California (Davis, CA). The inter-assay coefficients of variation for high and low internal controls for the E1G assay were 12.2% and 14.0% respectively. The PdG intra- and inter-assay variability was determined in-house as 13.6% and 18.7% respectively. All urine samples were corrected for specific gravity using a hand refractometer (NSG Precision Cells) to account for hydration status [37], which has been reported to perform as well as creatinine correction for adjusting urinary hormone concentrations [37]. Urinary LH was determined by coat-a-count immunoradiometric assay (Siemens Healthcare Diagnostics, Deerfield, IL). The sensitivity of the LH assay is 0.15 mIU/L. The intra-assay and inter-assay coefficients of variation were 1.6% and 7.1%, respectively.

Serum leptin concentration was measured using a solid-phase sandwich enzyme-linked immunoassay for total leptin (Millipore, St. Charles, MI). The inter-assay and intra-assay coefficients of variation for the low control were 6.2% and 4.6%, respectively. This assay is sensitive to leptin concentrations of 0.5 ng/mL. Total PYY was measured using an RIA for total PYY (Linco Research, St. Charles, MO). The total PYY assay recognizes both PYY1–36 and PYY3–36 and does not require the addition of inhibitors. The intra-assay and interassay coefficients of variation were 5.3% and 7.0%, respectively. The sensitivity of the assay was 10 pg/mL. Total testosterone was measured using a radioimmunoassay kit (Siemens, Los Angeles, CA) through competitive immunoassay. Analytical sensitivity for the testosterone assay was 0.14 nmol/L (4.0 ng/dL). The intra-assay and inter-assay coefficients of variation were 6.4% and 7.5%, respectively. Serum hormone binding globulin (SHBG) was analyzed using a chemiluminescence analyzer (Immulite, Euro Diagnostic Products Corporation, Lianberis, UK) through competitive immunoassay. Analytical sensitivity for the SHBG was (0.2 nmol/L) 5.76 ng/dL. The intra-assay and inter-assay coefficients of variation were 6.4% and 8.7%, respectively. TSH, T4, prolactin, DHEA, FSH and LH were sent out for analysis (Quest Diagnostics, Pittsburgh, PA).

Anthropometric testing Total body weight was measured weekly to the nearest 0.1 kg on a physician's scale (Seca, Model 770, Hamburg, Germany), and height was measured to the nearest 0.5 cm at the beginning of the study. Body mass index (BMI) was calculated as weight divided by height squared (kg/m2). Percentage body fat, fat mass, and lean body mass (LBM) were determined by dual-energy X-ray absorptiometry (DXA) once during the study protocol. Forty-four women were scanned on one of two scanners at different centers, either a GE Lunar Prodigy (n= 30, enCORE 2002 Software version 6.50.069) or a GE Lunar iDXA (n = 14, enCORE 2008 software version 12.10.113). Consistent with the International Society of Clinical Densitometry guidelines, a cross calibration study was performed to remove systematic bias between the systems. Fourteen participants were scanned in triplicate on both machines. The majority (n= 8) were scanned on both machines within 5 days; however, there was approximately one month between scans for some subjects (n= 6). The values were found to be highly correlated with no significant difference between the population mean values. However, we did find

Statistical analysis Data screening was conducted prior to statistical analysis in order to identify whether the data met the assumptions required by specific statistical techniques. Data screening involved outlier detection, and examination of variable distributions within each of the four groups for normality. Since all distributions were normal, parametric analyses were utilized for inferential analyses. Pearson's correlation analyses were used to detect associations between BMD and variables of interest. Running a separate analysis, forward multivariate stepwise regression using P = 0.05 for entry, and P = 0.10 to leave in the model was used to explore predictors of BMD. Variables included in the model were measurements of body composition, E1G metabolites, exercise status, and endocrine hormones of interest to the current study (leptin, PYY and estrogen). All data were analyzed

J.L. Scheid et al. / Bone 49 (2011) 194–201

197

by using SPSS for Windows (version 18.0; Chicago, IL). A significance level of 0.05 was used to detect the presence of significant differences. Data were expressed as mean ± SEM. Results Demographic characteristics Demographic characteristics of the women are presented in Table 1. The ExOv and ExAmen groups were similar with respect to age (p = 0.321), height (p = 0.847) and weight (p = 0.234). However, the ExAmen group had a lower (p = 0.047) BMI compared to the ExOv group. Body fat (%) (p = 0.011) and fat mass (kg) (p = 0.010) were lower in the ExAmen group compared to the ExOv group. There was no difference between the ExOv and ExAmen group with respect to lean body mass (p = 0.572), peak aerobic capacity (p = 0.087), and weekly physical activity minutes (p = 0.134). Bone mineral density Compared to the ExOv group, the ExAmen had lower BMD at the total body (p= 0.019), lumbar (L2–L4) spine (p= 0.001), and hip (p= 0.032) sites. Total body (p= 0.050) and L2–L4 (p= 0.002) Z-score was also lower in the ExAmen group vs. the ExOv group. The groups were similar with respect to total hip mean Z-score (p= 0.183) and neck of femur Z-score (p= 0.680) and BMD (p= 0.283) (Fig. 1). Metabolic and gastrointestinal hormones Metabolic and gastrointestinal hormone concentrations are presented in Fig. 2. PYY concentrations were higher in the ExAmen compared to the ExOv group (p b 0.001). Serum leptin (p = 0.042) concentrations were lower in the ExAmen compared to the ExOv group.

Fig. 1. Bone density Z-scores in exercising women with ovulatory or amenorrheic menstrual cycles. ExOv= exercising women with ovulatory cycles; ExAmen= exercising women with amenorrheic menstrual cycles. *Significant difference between exercising women with ovulatory cycles and exercising women with amenorrheic menstrual cycles (pb 0.05).

2–5 (p= 0.893) and days 2–12 (p= 0.356) were similar in the ExOv and ExAmen groups. ExAmen group had lower PdG cycle mean (pb 0.001) and AUC (p= 0.001) compared to the ExOv group. There were no differences between the two groups with respect to serum LH (p = 0.186) or FSH (p = 0.121) concentrations. Fig. 2 (Panel A) demonstrates suppressed E1G and PdG concentrations associated with amenorrhea. Fig. 2 (Panel B) displays the rise in E1G concentrations during the follicular phase, and E1G and PdG concentrations that are classic characteristics of an ovulatory cycle.

Reproductive characteristics Menstrual cycle parameters are presented in Table 2. The average menstrual cycle length in the ExOv group was 29.8 ± 0.7 days. The average duration of amenorrhea for the ExAmen group was 222.7 ± 35.7 days. E1G and PdG analyses are presented in Table 2 and composite menstrual cycle graphs of the ExOv and ExAmen groups are presented in Fig. 3. The E1G cycle mean (p= 0.001) and AUC (pb 0.001) were lower in the ExAmen group compared to the ExOv group. The E1G AUC on days

Table 1 Demographic characteristics and bone mineral density of exercising women with ovulatory and amenorrheic menstrual cycles.

Demographic characteristics Age (years) Height (cm) Body weight (kg) Body mass index (kg/m2) Body composition characteristics Body fat (%) Fat mass (kg) Lean body mass (kg) Training characteristics Peak aerobic capacity (mL/kg/min) Physical activity (min/week)

ExOv (n = 27)

ExAmen (n = 17)

p-value

24.3 ± 0.9 165.9 ± 1.2 59.1 ± 0.9 21.6 ± 0.3

22.9 ± 0.8 166.0 ± 1.7 56.9 ± 1.8 20.5 ± 0.4

0.321 0.847 0.234 0.047

26.1 ± 0.8 15.5 ± 0.6 41.3 ± 0.8

22.5 ± 1.2 12.8 ± 0.8 42.1 ± 1.3

0.011 0.010 0.572

47.2 ± 1.8 302.7 ± 38.4

52.0 ± 1.8 409.4 ± 63.1

0.087 0.134

Values are the mean ± SEM. ExOv = exercising/ovulatory, ExAmen = exercising/ amenorrheic.

Fig. 2. Bar graphs represent the mean (± SEM) concentrations of PYY (pg/mL) and leptin (ng/mL) of exercising women with ovulatory or amenorrheic menstrual cycles. ExOv = exercising women with ovulatory cycles; ExAmen = exercising women with amenorrheic menstrual cycles; PYY = Peptide YY. *Significant difference between exercising women with ovulatory cycles and exercising women with amenorrheic menstrual cycles (p b 0.05).

198

J.L. Scheid et al. / Bone 49 (2011) 194–201

Table 2 Reproductive profiles of exercising women with ovulatory or amenorrheic menstrual cycles.

Menstrual cycle characteristics Menstrual cycle length (days) Duration of amenorrhea (days)

ExOv (n = 27)

ExAmen (n = 17)

p-value

29.8 ± 0.7 n/a

n/a 222.7 ± 35.7

n/a n/a

18.80 ± 2.71 506.08 ± 72.48 0.78 ± 0.12 21.14 ± 3.32

0.001 b 0.001 b 0.001 0.001

Estrogen and progesterone analysis E1G cycle mean (ng/mL) 32.97 ± 2.51 E1G cycle AUC 958.87 ± 74.11 PdG cycle mean (μg/mL) 2.39 ± 0.14 PdG cycle AUC 69.44 ± 4.03 Gonadotropin data Luteinizing hormone (LH) (mIU/L) Follicular stimulating hormone (FSH) (mIU/L)

10.26 ± 2.96

4.96 ± 1.55

0.186

4.67 ± 0.71

6.33 ± 1.90

0.121

Values are the mean ± SEM. ExOv = Exercising/ovulatory, ExAmen = Exercising/ amenorrheic.

Associations with bone mineral density Pearson correlations between BMD variables, body composition, E1G metabolites, and endocrine hormones are presented in Table 3. There was a significant correlation between total body BMD and PYY (r = −0.310, p = 0.041), E1G cycle mean (r = 0.323, p = 0.033), and lean body mass (r = 0.318,p = 0.035), while there was a significant positive correlation between lumbar spine BMD (L2–L4) and E1G cycle mean (r = 0.301, p = 0.047) and a significant negative correlation between hip BMD and PYY (r = − 0.307, p = 0.043). PYY was not correlated with any of the other variables of interest including: E1G cycle mean (r = −0.224, p = 0.144), BMI (r = − 0.210, p = 0.176), body weight (r = −0.110, p = 0.479), fat mass (kg, r = − 0.261, p = 0.087), and lean body mass (kg, r = 0.090, p = 0.563).

Predictors of bone mineral density Using pooled data for forward stepwise linear regression, we entered all significant bivariate variables into the prediction model, including both variables that showed a significant association with BMD (i.e. lean body mass, PYY, E1G), and non-significant correlates that have previously been reported by others to be associated with BMD such as

Fig. 3. Composite graphs of daily urinary reproductive hormones of exercising women with ovulatory or amenorrheic menstrual cycles. Panel A demonstrates the mean daily E1G and PdG concentrations of exercising women amenorrhea. Panel B demonstrates the mean daily E1G and PdG concentrations of exercising women ovulatory cycles. E1G = estrone1-glucuronide; PdG = pregnanediol glucuronide. Data is expressed as mean ± SEM.

J.L. Scheid et al. / Bone 49 (2011) 194–201 Table 3 Pearson correlations between bone mineral density variables, body composition, estrogen status, and endocrine hormones. Total body bone mineral density (g/cm2)

Lumbar spine (L2–L4) bone mineral density (g/cm2)

Total hip bone mineral density (g/cm2)

R-value

R-value

R-value

Body composition Fat mass (kg) Lean body mass (kg) Metabolic hormones Peptide YY (pg/mL) Leptin (ng/mL)

p-value

p-value

p-value

0.200 0.318

0.194 0.035*

0.216 0.222

0.159 0.147

0.117 0.239

− 0.310 0.070

0.041* 0.653

− 0.254 0.029

0.096 0.850

− 0.307 − 0.059

0.323

0.033⁎

0.301

Estrogen E1G cycle mean

0.047⁎

0.279

0.450 0.119

0.043⁎ 0.706

0.067

⁎Significant correlations (p b 0.05).

fat mass, exercise status, and leptin. For the first time in exercising premenopausal women, we report herein that the strongest predictors of total body BMD were E1G cycle mean and body weight (adjusted R2 = 0.219, p = 0.003). While the strongest predictor of lumbar spine (L2–L4) BMD was also E1G cycle mean and body weight (adjusted R2 = 0.164, p = 0.012), the strongest predictor of hip BMD was PYY (adjusted R2 = 0.086, p = 0.033, Table 4).

Discussion In this cross-sectional analysis of 44 premenopausal exercising women, we demonstrate that fasting total PYY concentrations are negatively associated with total body BMD and total hip BMD and that mean E1G concentrations are positively associated with total body BMD and lumbar spine BMD. Additionally, we speculate that the elevated PYY and suppressed estrogen concentrations may contribute to decreased BMD in exercising women with amenorrhea, with a seemingly stronger effect of monthly estrogen exposure on lumbar spine BMD while circulating total PYY concentration mainly affects total hip BMD. The low bone mass of women with amenorrhea and estrogen deficiency, in comparison to women with eumenorrhea and an estrogen replete environment, is well-documented. However, an association between elevated PYY and low BMD has only recently been observed [23]. Russell et al. [23] demonstrated PYY to be negatively associated with apparent BMD and a marker of bone formation, PINP, in adolescent female athletes both with and without amenorrhea. In clinical populations of patients with anorexia nervosa and amenorrhea, concentrations of PYY have been demonstrated to be negatively

Table 4 Predictors of bone mineral density in exercising women. β Value

p Value

Variability explained by variable

Cumulative variability explained by model

0.011 0.003

13.1% 8.8%

13.1% 21.9%

Lumbar spine (L2–L4) bone mineral density Body weight (kg) 0.006 0.026 E1G cycle mean (ng/mL) 0.003 0.012

9.6% 6.8%

9.6% 16.4%

Total hip bone mineral density PYY (pg/mL) − 0.003

8.6%

8.6%

Total body mineral density Body weight (kg) E1G cycle mean (ng/mL)

0.005 0.002

0.033

Variables entered into a stepwise regression included: exercise minutes per week, body weight, fat mass, lean body mass, E1G cycle mean, leptin, and PYY.

199

associated with bone turnover [16] and BMD [25], indicating that PYY may play a role in disrupting bone turnover and result in detrimental bone pathology. Additionally, in adolescent girls with anorexia nervosa who have been followed prospectively, changes in PYY were demonstrated to predict changes in BMD [38]. In the current study, we confirm the association between PYY and BMD; however, this is the first study to demonstrate an association between PYY and BMD in premenopausal exercising women to include women with amenorrhea. Although the current study can only conclude an association between PYY and BMD, animal studies suggest that PYY may directly regulate bone formation [39] although the mechanisms are still unclear [40]. Baldock et al. [40] showed that in Y2 deficient mice, suppression of the Y2 receptor increased bone volume, suggesting that hypothalamic Y2 receptors modulate bone formation. Since PYY can cross the blood brain barrier and bind with receptors of the Y family, PYY may regulate BMD through its binding to the Y2 receptor in the hypothalamus. According to these findings, low levels of PYY would be beneficial for bone formation. However, Wortly et al. [41] reported that PYY deficient mice have a reduction in trabecular bone mass. This suggests that PYY may be acting in a region other than the hypothalamus to regulate bone turnover and elevated PYY may actually be beneficial for bone formation. If elevated PYY has positive effects on trabecular bone mass, exercising women with amenorrhea may be resistant to the positive effects of elevated PYY on bone, similar to women with anorexia nervosa who have also been suggested to be resistant to the positive effects that PYY may have on bone [42]. Both the animal models [40,41] and the human models [16,23,25,38] suggest that elevated PYY concentrations regulate bone turnover at the neuroendocrine level in women with amenorrhea and, as a result, contribute to bone loss leading to low BMD. Interestingly, Miller and colleagues [43] demonstrated that weight gain in women with anorexia nervosa resulted in increased hip BMD, and we speculate that nutritional factors associated with weight gain, for example, changes in IGF-1, leptin, insulin, adiponectin, and PYY, may be involved with the increase in hip BMD [30,44–46]. Misra et al. [16] demonstrated that weight gain in adolescent girls with anorexia nervosa resulted in a decrease in circulating PYY concentrations, further supporting the premise that decreases in circulating PYY associated with weight gain may contribute to improvements in total body BMD or site-specific BMD. However, the exact mechanism of site-specific regulation of BMD by hypothalamic factors is unknown. The current study supports the notion that both elevated PYY and suppressed monthly estrogen exposure contribute to low BMD in exercising women with amenorrhea. Hypoestrogenism has been demonstrated to be associated with low bone mass and perpetual bone loss in adults [45,46,47]. Drinkwater and colleagues [48] demonstrated that exercising women who resume menses increase spine BMD. Interestingly, in women with anorexia nervosa, Grinspoon et al. [47] demonstrated an association between increased duration of amenorrhea and decreased BMD at both the anterior–posterior spine and at the lateral spine. To this end, Miller and colleagues [43] also demonstrated that resumption of menses, independent of increases in weight, resulted in increased spinal BMD. All of these studies support the concept that estrogen concentrations, or mean EIG concentrations in the current study, are a strong predictor of lumbar spine BMD. We speculate that the lumbar spine may be more sensitive to changes in circulating estrogen concentrations than the hip because it contains about 70% of trabecular bone which has a more rapid turnover than cortical bone. The results from Miller and colleagues [43] support our findings that circulating estrogen is the strongest predictor of lumbar spine BMD and emphasize the importance of resuming menses and increased endogenous estrogen exposure to positively translate to an improvement in BMD. In contrast, exogenous estrogens in the form of oral contraceptives were shown to have no positive impact on BMD [43], presumably via suppression of IGF-1 levels [43], thereby attenuating the benefits of estrogen on BMD. Alternatively, this may

200

J.L. Scheid et al. / Bone 49 (2011) 194–201

suggest that the nutritional factors associated with the resumption of menses also impact BMD and that the resumption of menses is likely to be accompanied by significant changes in bone-trophic hormones, such as IGF-1. The major limitation to this study is inherent in the cross-sectional design and the correlational nature of the study, particularly because association does not imply causation. A second limitation is that only fasting PYY concentrations were measured. However, fasting PYY concentrations have a strong correlation with the PYY post-meal peaks [49], indicating that the post-meal PYY concentrations likely observed in our population of women would be relative to the observed fasting PYY concentrations. Additionally, we also want to acknowledge that the hormones included in this study are not the only hormones that have been identified in the literature as strong predictors of BMD, for example IGF-1 is known to be a very strong predictor of BMD [50]; however, IGF-1 was not measured in the current study. In conclusion, this study demonstrates that PYY, E1G metabolites and body weight are independently associated with BMD in premenopausal exercising women. Future prospective studies need to explore endocrine alterations during recovering of amenorrhea and the effects on weight gain on BMD in exercising women with amenorrhea.

Acknowledgments This study was supported by the United States Army Medical Research and Material Command Peer Reviewed Medical Research Program (Award Number PR054531).

References [1] De Souza MJ, Toombs RJ, Scheid JL, O'Donnell E, West SL, Williams NI. High prevalence of subtle and severe menstrual disturbances in exercising women: confirmation using daily hormone measures. Hum Reprod 2010;25(2):491–503. [2] De Souza MJ, Miller BE, Loucks AB, Luciano AA, Pescatello LS, Campbell CG, et al. High frequency of luteal phase deficiency and anovulation in recreational women runners: blunted elevation in follicle-stimulating hormone observed during luteal–follicular transition. J Clin Endocrinol Metab 1998;83(12):4220–32. [3] De Souza MJ, Williams NI. Physiological aspects and clinical sequelae of energy deficiency and hypoestrogenism in exercising women. Hum Reprod Update 2004;10(5):433–48. [4] Williams NI. Lessons from experimental disruptions of the menstrual cycle in humans and monkeys. Med Sci Sports Exerc 2003;35(9):1564. [5] De Souza MJ, Lee DK, VanHeest JL, Scheid JL, West SL, Williams NI. Severity of energy-related menstrual disturbances increases in proportion to indices of energy conservation in exercising women. Fertil Steril 2007;88(4):971–5. [6] De Souza MJ. Menstrual disturbances in athletes: a focus on luteal phase defects. Med Sci Sports Exerc 2003;35(9):1553. [7] Kaufman BA, Warren MP, Dominguez JE, Wang J, Heymsfield SB, Pierson RN. Bone density and amenorrhea in ballet dancers are related to a decreased resting metabolic rate and lower leptin levels. J Clin Endocrinol Metab 2002;87(6): 2777–83. [8] De Souza MJ, Leidy HJ, O'Donnell E, Lasley B, Williams NI. Fasting ghrelin levels in physically active women: relationship with menstrual disturbances and metabolic hormones. J Clin Endocrinol Metab 2004;89(7):3536–42. [9] Laughlin GA, Yen SS. Hypoleptinemia in women athletes: absence of a diurnal rhythm with amenorrhea. J Clin Endocrinol Metab 1997;82(1):318–21. [10] Waters DL, Qualls CR, Dorin R, Veldhuis JD, Baumgartner RN. Increased pulsatility, process irregularity, and nocturnal trough concentrations of growth hormone in amenorrheic compared to eumenorrheic athletes. J Clin Endocrinol Metab 2001;86(3):1013–9. [11] Christo K, Cord J, Mendes N, Miller KK, Goldstein MA, Klibanski A, et al. Acylated ghrelin and leptin in adolescent athletes with amenorrhea, eumenorrheic athletes and controls: a cross-sectional study. Clin Endocrinol (Oxf) 2008;69(4):628–33. [12] Schneider LF, Monaco SE, Warren MP. Elevated ghrelin level in women of normal weight with amenorrhea is related to disordered eating. Fertil Steril 2008;90(1): 121–8. [13] Suh BY, Liu JH, Berga SL, Quigley ME, Laughlin GA, Yen SS. Hypercortisolism in patients with functional hypothalamic-amenorrhea. J Clin Endocrinol Metab 1988;66(4):733–9. [14] Nattiv A, Loucks AB, Manore MM, Sanborn CF, Sundgot-Borgen J, Warren MP. American College of Sports Medicine position stand. The female athlete triad. Med Sci Sports Exerc 2007;39(10):1867–82.

[15] Scheid JL, Williams NI, West SL, VanHeest JL, De Souza MJ. Elevated PYY is associated with energy deficiency and indices of subclinical disordered eating in exercising women with hypothalamic amenorrhea. Appetite 2009;52(1):184–92. [16] Misra M, Miller KK, Tsai P, Gallagher K, Lin A, Lee N, et al. Elevated peptide YY levels in adolescent girls with anorexia nervosa. J Clin Endocrinol Metab 2006;91(3):1027–33. [17] Pfluger PT, Kampe J, Castaneda TR, Vahl T, D'Alessio DA, Kruthaupt T, et al. Effect of human body weight changes on circulating levels of peptide YY and peptide YY3-36. J Clin Endocrinol Metab 2007;92(2):583–8. [18] Nakahara T, Kojima S, Tanaka M, Yasuhara D, Harada T, Sagiyama K, et al. Incomplete restoration of the secretion of ghrelin and PYY compared to insulin after food ingestion following weight gain in anorexia nervosa. J Psychiatr Res 2007;41(10):814–20. [19] Batterham RL, Cohen MA, Ellis SM, Le Roux CW, Withers DJ, Frost GS, et al. Inhibition of food intake in obese subjects by peptide YY3-36. N Engl J Med 2003;349(10):941–8. [20] Huda MS, Wilding JP, Pinkney JH. Gut peptides and the regulation of appetite. Obes Rev 2006;7(2):163–82. [21] Adrian TE, Ferri GL, Bacarese-Hamilton AJ, Fuessl HS, Polak JM, Bloom SR. Human distribution and release of a putative new gut hormone, peptide YY. Gastroenterology 1985;89(5):1070–7. [22] Kim BJ, Carlson OD, Jang HJ, Elahi D, Berry C, Egan JM. Peptide YY is secreted after oral glucose administration in a gender-specific manner. J Clin Endocrinol Metab 2005;90(12):6665–71. [23] Russell M, Stark J, Nayak S, Miller KK, Herzog DB, Klibanski A, et al. Peptide YY in adolescent athletes with amenorrhea, eumenorrheic athletes and non-athletic controls. Bone 2009;45(1):104–9. [24] Cobb Kl, Bachrach LK, Greendale G, Marcus R, Neer R, Nieves J, et al. Disordered eating, menstrual irregularity, and bone mineral density in female runners. Med Sci Sports Exerc 2003;35(5):711–9. [25] Utz AL, Lawson EA, Misra M, Mickley D, Gleysteen S, Herzog DB, et al. Peptide YY (PYY) levels and bone mineral density (BMD) in women with anorexia nervosa. Bone 2008;43(1):135–9. [26] De Souza MJ, West Sl, Jamal SA, Hawker GA, Gundberg CM, Williams NI. The presence of both an energy deficiency and estrogen deficiency exacerbate alterations of bone metabolism in exercising women. Bone 2008;43(1):140–8. [27] Drinkwater BL, Nilson K, Chesnut III CH, Bremner WJ, Shainholtz S, Southworth MB. Bone mineral content of amenorrheic and eumenorrheic athletes. N Engl J Med 1984;311(5):277–81. [28] Salisbury JJ, Mitchell JE. Bone mineral density and anorexia nervosa in women. Am J Psychiatry 1991;148(6):768–74. [29] Misra M, Miller KK, Stewart V, Hunter E, Kuo K, Herzog DB, et al. Ghrelin and bone metabolism in adolescent girls with anorexia nervosa and healthy adolescents. J Clin Endocrinol Metab 2005;90(9):5082. [30] Misra M, Miller KK, Cord J, Prabhakaran R, Herzog DB, Goldstein M, et al. Relationships between serum adipokines, insulin levels, and bone density in girls with anorexia nervosa. J Clin Endocrinol Metab 2007;92(6):2046–52. [31] Santoro N, Crawford SL, Allsworth JE, Gold EB, Greendale GA, Korenman S, et al. Assessing menstrual cycles with urinary hormone assays. Am J Physiol Endocrinol Metab 2003;284(3):E521–30. [32] Kesner JS, Wright DM, Schrader SM, Chin NW, Krieg Jr EF. Methods of monitoring menstrual function in field studies: efficacy of methods. Reprod Toxicol (Elmsford, NY) 1992;6(5):385–400. [33] Rosner W, Auchus RJ, Azziz R, Sluss PM, Raff H. Position statement: utility, limitations, and pitfalls in measuring testosterone: an Endocrine Society position statement. J Clin Endocrinol Metab 2007;92(2):405–13. [34] Stener-Victorin E, Holm G, Labrie F, Nilsson L, Janson PO, Ohlsson C. Are there any sensitive and specific sex steroid markers for polycystic ovary syndrome? J Clin Endocrinol Metab 2010;95(2):810–9. [35] Barber TM, Wass JA, McCarthy MI, Franks S. Metabolic characteristics of women with polycystic ovaries and oligo-amenorrhoea but normal androgen levels: implications for the management of polycystic ovary syndrome. Clin Endocrinol (Oxf) 2007;66(4):513–7. [36] Munro CJ, Stabenfeldt GH, Cragun JR, Addiego LA, Overstreet JW, Lasley BL. Relationship of serum estradiol and progesterone concentrations to the excretion profiles of their major urinary metabolites as measured by enzyme immunoassay and radioimmunoassay. Clin Chem 1991;37(6):838–44. [37] Miller RC, Brindle E, Holman DJ, Shofer J, Klein NA, Soules MR, et al. Comparison of specific gravity and creatinine for normalizing urinary reproductive hormone concentrations. Clin Chem 2004;50(5):924–32. [38] Misra M, Prabhakaran R, Miller KK, Goldstein MA, Mickley D, Clauss L, et al. Prognostic indicators of changes in bone density measures in adolescent girls with anorexia nervosa—II. J Clin Endocrinol Metab 2008;93(4):1292–7. [39] Wong IP, Baldock PA, Herzog H. Gastrointestinal peptides and bone health. Curr Opin Endocrinol Diabetes Obes 2010;17(1):44–50. [40] Baldock PA, Sainsbury A, Couzens M, Enriquez RF, Thomas GP, Gardiner EM, et al. Hypothalamic Y2 receptors regulate bone formation. J Clin Invest 2002;109(7):915–21. [41] Wortley KE, Garcia K, Okamoto H, Thabet K, Anderson KD, Shen V, et al. Peptide YY regulates bone turnover in rodents. Gastroenterology 2007;133(5):1534–43. [42] Jacobson-Dickman E, Misra M. Skeletal abnormalities in anorexia nervosa. IBMS BoneKEy 2010;7(2):63–83. [43] Miller KK, Lee EE, Lawson EA, Misra M, Minihan J, Grinspoon SK, et al. Determinants of skeletal loss and recovery in anorexia nervosa. J Clin Endocrinol Metab 2006;91(8):2931–7. [44] Legroux-Gerot I, Vignau J, D'Herbomez M, Collier F, Marchandise X, Duquesnoy B, et al. Evaluation of bone loss and its mechanisms in anorexia nervosa. Calcif Tissue Int 2007;81(3):174–82.

J.L. Scheid et al. / Bone 49 (2011) 194–201 [45] Grinspoon S, Miller K, Coyle C, Krempin J, Armstrong C, Pitts S, et al. Severity of osteopenia in estrogen-deficient women with anorexia nervosa and hypothalamic amenorrhea. J Clin Endocrinol Metab 1999;84(6):2049–55. [46] Miller KK, Grinspoon S, Gleysteen S, Grieco KA, Ciampa J, Breu J, et al. Preservation of neuroendocrine control of reproductive function despite severe undernutrition. J Clin Endocrinol Metab 2004;89(9):4434. [47] Grinspoon S, Thomas E, Pitts S, Gross E, Mickley D, Miller K, et al. Prevalence and predictive factors for regional osteopenia in women with anorexia nervosa. Ann Intern Med 2000;133(10):790–4.

201

[48] Drinkwater BL, Nilson K, Ott S, Chesnut III CH. Bone mineral density after resumption of menses in amenorrheic athletes. JAMA 1986;256(3):380–2. [49] Stock S, Leichner P, Wong AC, Ghatei MA, Kieffer TJ, Bloom SR, et al. Ghrelin, peptide YY, glucose-dependent insulinotropic polypeptide, and hunger responses to a mixed meal in anorexic, obese, and control female adolescents. J Clin Endocrinol Metab 2005;90(4):2161–8. [50] Yakar S, Rosen CJ, Beamer WG, Ackert-Bicknell CL, Wu Y, Liu JL, et al. Circulating levels of IGF-1 directly regulate bone growth and density. J Clin Invest 2002;110 (6):771–81.