help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sowers, M.
Right arrow Articles by Willing, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sowers, M.
Right arrow Articles by Willing, M.
The Journal of Clinical Endocrinology & Metabolism Vol. 89, No. 2 733-739
Copyright © 2004 by The Endocrine Society

Estrogen Receptor Genotypes and Their Association with the 10-Year Changes in Bone Mineral Density and Osteocalcin Concentrations

MaryFran Sowers, Mary L. Jannausch, Wei Liang and Marcia Willing

Departments of Epidemiology (M.S., M.L.J.) and Biostatistics (W.L.), School of Public Health, University of Michigan, Ann Arbor, Michigan 48104; and Department of Pediatrics (M.W.), University of Iowa, Iowa City, Iowa 52242

Address all correspondence and requests for reprints to: MaryFran Sowers, Ph.D., 339 East Liberty, Suite 310, Ann Arbor, Michigan 48104. E-mail: mfsowers{at}umich.edu.


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
We conducted a 10-yr prospective study of peak bone mass and its change in 604 women, aged 24–44 yr at study initiation, and related changes in bone mineral density (BMD) and osteocalcin (OCN) concentrations to estrogen receptor (ER) {alpha} gene polymorphisms in 442 of these women.

We examined the association of ER{alpha} PvuII and XbaI polymorphisms with the 10-yr change in lumbar spine (LS) and femoral neck (FN) BMD, measured by densitometry, as well as serum OCN levels, after accounting for weight and menstrual status change. The women were members of the Michigan Bone Health Study, a population-based longitudinal study of BMD.

There was a linear loss of LS BMD and curvilinear loss of FN BMD from peak bone mass over a 10-yr period. Women homozygous for the ER{alpha} gene variant without an XbaI restriction site (XbaI -/- genotype) had higher FN BMD and less change in LS over time. Women homozygous for the ER{alpha} gene variant without a PvuII restriction site (PvuII -/- genotype) had less LS BMD change over time as well as higher FN BMD. However, this higher FN BMD was dependent upon the rate of bone turnover as estimated from serum OCN change over time.

The ER{alpha} genotype associations were statistically significant in explaining the rate of perimenopausal bone loss and its turnover; however, BMI or becoming postmenopausal contributed more to the magnitude of the difference in bone change.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
FAILURE TO ACCRUE adequate peak bone mass and loss of peak bone mass, particularly during the later perimenopause, can lead to lower bone mineral density (BMD) immediately after menopause and into subsequent older age (1). Although evidence suggests that the bone loss process can occur before menopause, few risk factors for low premenopausal BMD and for greater pre- and perimenopausal bone loss have been delineated (2, 3, 4, 5). Genetic, anthropometric, reproductive, and lifestyle-related factors may contribute to the variation in either peak BMD or its rate of loss. For example, among premenopausal women, body weight has been consistently and positively associated with BMD (5, 6, 7). Selective reproductive characteristics, such as irregular menses and amenorrhea, have been consistently associated with lower BMD, lactation has been associated with transient bone loss, and age at menarche has shown inconsistent association with BMD or long-term bone change (8, 9, 10, 11).

BMD is highly heritable, as demonstrated by comparisons of intrapair differences in BMD between monozygotic and dizygotic adult twin pairs, where heritability estimates of BMD range from 45–84%, depending on the skeletal site being examined (12, 13, 14). Sowers et al. (14) reported that approximately 50% of variation in femoral neck (FN) peak BMD was potentially genetic, based on data from a modified family study. The extensive list of candidate genes for study as determinants of BMD includes the nuclear estrogen receptor (ERß or ER{alpha}), including two polymorphisms in the first intron that can be defined by the restriction enzymes PvuII or XbaI (15, 16). This candidate gene has been widely studied based on the recognition of the importance of estrogen on bone as demonstrated by change in bone mass with the menopausal transition, treatments involving estrogen replacement (17), and the decreased mineralization observed in the ER{alpha} knockout mouse model (18). Despite numerous studies, the role of ER polymorphisms remains unclear, although a recent metaanalysis suggests a potential role for groups defined by the XbaI genotypes compared with groups defined by the PvuII genotypes (15).

To date, most human studies of genetic associations with BMD have been cross-sectional, in which genotypes have been related to level of BMD or to fractures, but few studies have related genotypes to BMD change over time. Likewise, there are few studies that have included bone turnover markers, such as osteocalcin (OCN) or N-telopeptides, which would give a more proximal measure of the association of genotype with BMD and its relationship to bone remodeling activity.

We completed a large, community-based, 10-yr prospective cohort study of bone and its change in pre- to early postmenopausal women, and we related BMD change and OCN serum levels to selected polymorphisms of the ER{alpha} gene. We determined the association between ER{alpha} genotypes and the 10-yr change in BMD and 10-yr OCN patterns in women approaching and transitioning to menopause. We also determined whether the association of change in serum OCN concentrations with FN and spine BMD change over time was different according to genotype.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
This study is a component of the longitudinal Michigan Bone Health Study, a study that has been described previously (4). This population-based study of women, aged 24–44 yr in 1992 when BMD measurement with dual energy x-ray densitometry was initiated, recruited the women from two lists. The first list identified the female children in families from the Tecumseh Community Health Study, a prospective cohort study of families that had been established in 1959. There were 539 adult daughters (>80% participation) of these families still living in the community who were successfully recruited to the Michigan Bone Health Study using letters, telephone calls, and in-person visits. A second list was used to identify and recruit those women who were current Tecumseh community residents but whose families had not lived in Tecumseh when the Tecumseh Community Health Study was undertaken in 1959–1969. There were 135 age-eligible female Tecumseh residents who were identified from Kohl’s Directory, a list that included age, name, address, and telephone number, and of these women, 121 (87%) were enrolled. The total cohort, identified from the combined lists, included 660 age-eligible Caucasian women.

Data for these analyses were collected nine times in the 10-yr period (1992/1993 to 2002 and excluding 1996/1997). Women may not have been evaluated in any given year because they declined (<14% in any given year), moved from the area (1–5% in any given year), reported significant health impairments (1–2% in any given year), or died (n = 6). An annual measurement was deferred to the following year if the enrollee was pregnant (1–3% in any given year). More than 64% of the study population was assessed at all nine evaluations, and almost 80% of the population had at least seven of the nine evaluations. Genotyping was completed in 442 of these enrollees. Written informed consent was obtained annually from all participants, and approval for the conduct of the study was obtained from the University of Michigan Institutional Review Board.

BMD and OCN measures

BMD was measured at the lumbar spine (LS; L2–4) and the FN with dual-energy x-ray absorptiometry (DPX-L; Lunar Corp., Madison, WI) and the same software (version 1.3y) at all time points. Coefficients of variation for the dual-energy x-ray absorptiometry measurements of the LS and FN were less than 1.0% in a study of repeated measures conducted at the study inception in 20 women. The densitometer was calibrated daily and a spine phantom was measured weekly to assure that drift did not occur in BMD measures or that periodic maintenance did not result in change that was artifact.

At each annual examination, blood was drawn on d 3–7 of the follicular phase of the menstrual cycle and after women had been fasting for 8 h. For those women without menstrual bleeding, blood was drawn fasting on the anniversary date of their first annual examination. Serum was stored in an ultralow freezer at -80 C and had no thaws and refreezes before assay for OCN. OCN, a bone matrix protein released into circulation by osteoblasts during bone formation and during the enzymatic degradation of the bone matrix, was selected as a measure of bone turnover (19). OCN concentrations were measured in batch using a RIA (Incstar, Stillwater, MN) with a combined inter- and intraassay variation of less than 10% and modeled as a time-varying covariate. We used logistic regression to generate a model for imputation and address OCN measurements that were below assay detection limits (0.05 nmol/liter). Based on the property of homogeneity, we replaced a random number for the data censored due to underdetection. The number of OCN values that were below detection [42 (7.5%) in 1992; 55 (9.7%) in 1993; 32 (7.4%) in 1994; 23 (4.6%) in 1995; and three or fewer (<0.6%) per year in 1998 and later] declined over time as the population increased with age.

Anthropometry

The height (cm) and weight (kg) of participants, wearing light clothing and stockings without shoes, were measured annually with a stadiometer and a calibrated balance-beam scale, respectively. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters (kg/m2) and was modeled as a time-varying covariate.

Reproductive history

Interviews were used to assess menstrual status and exogenous hormone use at each examination. A multilevel variable for menstrual status was based on the frequency of menstrual bleeds per year, hormone use, and having gynecologic surgery or treatment with chemotherapy. There was medical record confirmation of 75% of reported unilateral oophorectomies, 79% of reported hysterectomies, and 90% of reported bilateral oophorectomies. Failure to confirm surgeries and chemotherapy was, in large measure, due to the consolidation of health care providers and systems in the area and misplacement of archival records. Women were classified as postmenopausal if they reported the absence of a menstrual bleed for a 12-month period or a history of bilateral oophorectomy.

Lifestyle

At each interview, current physical activity level was measured with the self-administered Stanford Five-City instrument (20), where participants recalled the number of minutes of physical activity during the previous January and the previous July. Intensity of physical activity (summed to encompass the winter and summer data) was expressed in METS per week, with one MET defined as the energy consumed per minute of sitting at rest, and was modeled as a time-varying covariate. Dietary information was collected in the first four annual examinations with the Block-National Cancer Institute Health, History, and Habits Questionnaire (21) and DIETSYS software (HHHQ-DIETSYS Analysis software, version 3.0; National Cancer Institute, Bethesda, MD). Depending on the year, data from 55–73 women were excluded from the dietary analysis because their reports included too few foods or too few calories to meet the data checking criteria. Women reported their annual alcohol beverage consumption, and the data were expressed as ounces per month. Nutrients, dietary energy, and alcohol consumption were modeled as time-varying covariates. At each interview, women classified themselves as a current smoker, ex-smoker, or never smoker, and a summary variable was updated when a woman reported a change in smoking status.

Family history and genotyping

Enrollees reported whether any maternal or paternal first-degree relative had an osteoporosis diagnosis, dowager’s hump, or weak bones (a term used to describe osteoporosis among health care providers around the community). Family history of osteoporosis was defined as any positive report through 1998/1999 and was modeled as a summary variable.

DNA was extracted and amplified from 2–3 ml of whole blood, and genotyping was completed for 442 women. Red blood cells were lysed at 4 C in cell lysis buffer (0.32 M sucrose; 10 mM Tris-HCl, pH 8.0; 5 mM MgCl2; and 1% Triton X-100). After three cycles of lysis, the final lymphocyte pellet was resuspended in 10 mM Tris-HCl (pH 8.0), 10 mM NaCl, 10 mM EDTA, and proteinase K (final concentration, 1 mg/ml). Samples were incubated at 55 C overnight with periodic mixing to disperse the pellet; an aliquot of 1–3 µl (50–300 ng of DNA) was used for each PCR.

Primers designed to amplify intragenic polymorphic PvuII and XbaI sites were derived from Kobayashi et al. (22). Amplification conditions were as follows: 94 C (3 min), followed by 35 cycles at 94 C (1 min), 58 C or 60 C (45 sec), and 72 C (1.5 min). Amplifications were carried out in a 480 Perkin-Elmer thermocycler (Wellesley, MA). After each PCR, an aliquot of amplified material was cleaved with the appropriate restriction endonuclease, according to manufacturers’ specifications (New England Biolabs, Beverly, MA). The presence (+) or absence (-) of the enzyme recognition site was identified by ethidium bromide staining of fragments separated in 6% polyacrylamide. Genotypes were independently verified by two individuals.

Statistical analysis

The relationship between time and 10-yr change was assessed for FN BMD, LS BMD, and OCN. Univariate statistics were computed for continuous variables, and frequencies were determined for categorical variables. Variables with highly skewed distributions were transformed using a logarithmic function or categorized.

The linear and quadratic relationships with time were evaluated for each outcome variable alone and after inclusion of each genotype. LS BMD and OCN had a constant (linear) rate of change. The rate of change in FN BMD was identified as having a curvilinear (quadratic) component. Accordingly, all models for FN BMD included a quadratic time component, whereas the time component for OCN and LS was modeled as a linear function.

Repeated measures models were used to test for an association between phenotypic characteristics and genotypes at each genetic locus. Genotype variables were tested for their associations with bone loss or OCN change and were modeled either as main effects or as interactions with time, with the SAS Procedure MIXED (SAS Institute, Inc., Cary, NC) in the following form:

where BMDij is the value of BMD for the ith person at time j, time'ij is the jth time point for person i [adjusted for mean time, i.e. time'ij = (timeij - mean time)], xi is the value of the genotype for the ith person, bi is a random intercept term to account for correlated errors among repeated measures on the same woman, and {epsilon}ij is an error term. The ß1 (time') coefficient provides an estimate of the average annual rate of bone loss for premenopausal (the referent group) women, and the ß2 (time'2) demonstrates how the annual rate of bone loss changes over time. The ß3 genotype coefficient provides an estimate for the difference in BMD or OCN associated with a unit change in the specific genotype. A significant ß4 (risk factor) coefficient is interpreted as a difference in baseline BMD according to level of the risk factor only if there is a corresponding time' by risk factor interaction. The ß4 coefficient can be interpreted as the association between the genotype and BMD at any time over the 10-yr study. The ß4 coefficient (time' by genotype) and the ß5 coefficients (time'2 by genotype) estimate the BMD or OCN change according to genotype level over time (23, 24).

Covariates available for incorporation into the models included baseline smoking, alcohol consumption, usual daily dietary intake of calories and calcium, thiazide medication use, and physical activity, but these were not included in final models because they were not statistically significant. Covariates retained in the model included baseline BMD (or baseline OCN, as appropriate), BMI, and a status variable to describe menstrual status and hormone use. Gene-by-environment interactions were not estimated. Family history of osteoporosis or osteoporotic fracture was evaluated for its potential contribution as a main effect and in interaction with OCN or genotype. Analyses were undertaken using SAS 8.0 (SAS Proprietary Software, Release 8.0; SAS Institute).


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
Baseline characteristics of the study population are shown in Table 1Go. The average age of the study population at the inception of the 10-yr study period was 37 yr, with a span from 24–44 yr. At baseline, women were premenopausal, with the exception of 34 women who had a bilateral oophorectomy and three women with premature menopause, and 21 of these women were using hormone replacement therapy (HRT). In 2002, when the age range was 34–54 yr, 69% of women were still classified as premenopausal. The mean (± SD) usual daily dietary calcium intake was 758 (± 455) mg, whereas the usual median dietary calcium intake was 667 mg.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Characteristics of population at the baseline examination (1992/1993): Michigan Bone Health Study (n = 604)

 
The amount and direction of BMD change over the 10-yr period is shown in Table 2Go as a ß coefficient ± SE. Over the 10-yr period, there was a modest and linear 1.5% BMD loss at the LS (P < 0.0001) and a 2.2% loss of FN BMD that was mildly curvilinear. As shown in Fig. 1Go, as OCN concentrations increased over time, there was BMD decline at both the FN and LS. The OCN concentrations increased significantly and linearly by approximately 10% over the 10-yr period, as estimated with a regression coefficient (P < 0.003).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Annual rate of change in BMD at the FN and LS and in OCN with adjustment for baseline BMI and time-varying reproductive status

 


View larger version (15K):
[in this window]
[in a new window]
  		FIG. 1. The fitted line for the amount of 10-yr BMD change (change values shown on the left vertical axis) is significantly and inversely related to the fitted line for the amount of 10-yr OCN change (change values shown on the right vertical axis).

 
A family history of osteoporosis, reported by 38% of women, had slightly greater OCN turnover over time and lower BMD, but the difference in BMD change was not significant when compared with women with no family history of osteoporosis.

The frequency of each of the genotypes is shown in Table 1Go. There is no evidence of deviation from Hardy-Weinberg equilibrium for the selected polymorphisms (P > 0.05).

LS loss

As shown with regression coefficients in Table 3Go, there was linear decline in LS BMD over time and a greater change in postmenopausal women compared with premenopausal women. A statistically significant interaction term was used to identify that the rate of bone loss varied according to genotype. There was less change in LS BMD in those women who were PvuII (-/-) homozygotes compared with women who were PvuII (+/+) homozygotes (P = 0.0007). Likewise, there was less change in LS BMD in those women who were XbaI (-/-) homozygotes compared with women who were XbaI (+/+) homozygotes (P < 0.02).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Mean annual LS BMD (g/cm2) change over time, as predicted by PvuII and XbaI genotypes and other covariates, after adjustment for baseline BMD

 
FN

As shown with the regression coefficients in Table 4Go, the change in FN over the 10-yr period was mildly curvilinear downward and more evident in those women who because postmenopausal compared with those women who remained premenopausal. Lower OCN concentrations and their change over time were linearly associated with change in FN BMD.


View this table:
[in this window]
[in a new window]
  		TABLE 4. FN BMD (g/cm2), OCN (nmol/liter),1 and their change over time, as predicted by PvuII and XbaI genotypes and other covariates, after adjustment for baseline BMD

 
Both ER genotypes were associated with level of BMD, with greater levels being observed for women who were PvuII and XbaI (-/-) homozygotes. As shown in Fig. 2Go, there was a significant OCN by genotype interaction, suggesting that the amount of FN BMD change per unit of OCN change depended upon the genotype group, particularly the XbaI genotype groups. As an example, there is a significant difference (-0.029 g/cm2; P = 0.05) between the unit change of FN BMD of XbaI (-/-) homozygotes. However, when OCN levels increased over time, the FN BMD decreased more rapidly in the -/- group and somewhat faster in the +/- group than in the +/+ group (P < 0.05).



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 2. A, The slopes for the fitted line of 10-yr FN BMD per unit change in OCN (change values shown on the left vertical axis) are significantly different among women homozygous for the presence (+/+) or absence (-/-) of the PvuII genotypes. B, The slopes for the fitted line of 10-yr FN BMD per unit change in OCN (change values shown on the left vertical axis) are different among the three genotypes of the XbaI polymorphism.

 

   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
Examining the potential genetic contribution to peak bone mass accretion and maintenance is important for two reasons. First, there are relatively few factors that explain peak BMD variation. Second, there is little information about whether peak bone mass is similarly retained among all women with diverse genotypes until the menopausal transition. Further, a genetic contribution to BMD and its change may be more readily identified in premenopausal women before menopause and aging confound the effect of genotype on BMD. We found that the linear 10-yr rate of change in LS was different among groups of women characterized by the presence or absence of variants of the ER as defined by the restriction enzymes PvuII and XbaI. The FN BMD loss was also different among the genotype groups, but the rate of change was more strongly tied to the change in OCN over time. To our knowledge, this is the first time that the change in BMD or OCN according to ER genotype has been reported.

The ER{alpha} gene maps to chromosome 6q25, and the polymorphisms defined by the two restriction enzymes PvuII and XbaI are in close proximity (~45 bp apart) in the first intron (25). We found that women with the XbaI (-/-) genotype had higher FN BMD and less change in LS over time. These findings are consistent with and extend the summary findings from the metaanalysis of the associations of ER{alpha} with BMD and fracture risk (16).

We found that women with the PvuII (-/-) genotype had less LS BMD change over time as well as higher FN BMD. This is inconsistent with the metaanalysis that reported no association of the PvuII genotypes with BMD or fractures (16). However, we observed that the higher FN BMD was dependent upon the rate of bone turnover as estimated by OCN concentrations and their change over time. We found that increasingly higher OCN concentrations, a bone matrix protein released into circulation by osteoblasts during bone formation and during the enzymatic degradation of the bone matrix (19), were greater in women with greater bone loss, which was consistent with other studies (26). However, the women that were PvuII (-/-) homozygotes did not have the same BMD turnover for the increase in OCN units. The metaanalysis did not address such interactions, probably because they are infrequently reported in the literature. Further, our study does not identify the mechanism of this interaction because, as identified by Lindberg et al. (27), this could include gene by gene interactions with growth factors and inflammatory cytokines.

Our study reports statistically significant associations of the ER genotypes with BMD and its change, but it is important to consider the relative impact of these genotypes with respect to other covariates. As seen in each of the tables, the relative importance of these genotypes was minimal, which is not surprising considering the polygenic and multifactorial nature of BMD (28). Two factors, baseline BMI and change in menstrual status, were stronger predictors of both level and 10-yr rate of change in BMD than the ER genotypes, and these predictors have consistently and repeatedly been reported (3, 5, 6, 28, 29, 30).

We observed progressively greater bone change in postmenopausal women compared with premenopausal women, even when these women were using HRT. The effect of estrogen is believed to be mediated by the nuclear ER{alpha} and/or ERß, although the direct mechanisms of estrogen on bone are not completely understood (17). However, because most women had not actually experienced an observable transition to postmenopause, we did not test whether the rate of BMD loss among postmenopausal women was different in groups defined by the ER genotypes.

Women with greater BMI at baseline had a positive, not negative, change in BMD. The contribution of BMI could reflect at least two mechanisms. First, women with greater BMI can have the experience of increased bone loading, generating an environment of greater BMD (31). Likewise, adipose tissue is increasingly acknowledged as an endocrine organ, and with greater BMI, a greater number of adipocytes increases with the potential for aromatization of androstenedione to estrone (32). Although estrone is less biologically active than circulating estradiol, the most biologically active estrogen in premenopausal women, adipose tissue could become an increasingly important estrogen hormone source in the postmenopausal period. Although we identified no ER{alpha} genotype interaction with BMI, the size of the adipose pool in relation to ER markers may become more important as more study enrollees become postmenopausal.

Although this study reported an association that appears related to the differential response of BMD with OCN, it does not establish an actual functional role for the genotypes. If the ER genotypes are important in this relationship, it is possible that the reported genotypes are linked to other causative sequence variants. For that reason, one recent investigation reported the association of eight ER{alpha} variants, including the polymorphisms reported here, to HRT therapy and high-density lipoprotein cholesterol in women with coronary disease (33), yet a functional role was not resolved in that report and remains uncertain.

This study has a number of strengths and limitations. The longitudinal methods used in this study included the ability to reclassify an exposure that may change with time (such as OCN concentration) and to accumulate events (such as reproductive events) over time (23, 24). The use of these methods may have diminished any misclassification bias and increased the power necessary to detect associations with BMD change over time. This study has sustained high participation over time, suggesting that findings are not likely to be biased because of loss to follow-up. Indeed the lowest participation occurred in mid-study, and participation actually increased over time. However, the size of the study population coupled with the distribution of the genotypes limited the power to detect some associations. Because almost two thirds of the women were still premenopausal at the end of the study period, potentially important interactions between the genotypes and the experience of the menopausal transition could not be evaluated.

In summary, there was a linear loss of LS BMD and curvilinear loss of FN BMD over a 10-yr period in a population-based sample of women aged 24–44 yr at study inception. Those women homozygous for the ER{alpha} gene variant without an XbaI restriction site had higher FN BMD and less change in LS over time. Further, women homozygous for the ER{alpha} gene variant without a PvuII restriction site had less LS BMD change over time as well as higher FN BMD. However, this higher FN BMD was dependent upon the rate of bone turnover as approximated by OCN concentrations and their change over time. Although these genotype associations were statistically significant, their magnitude was overshadowed by two other factors. Women with higher baseline BMI had less bone loss, and women who became postmenopausal within the 10-yr period had greater loss even when using HRT.


   Footnotes
 
This work was supported by Grants AR R01-40888 [to M.S., Principal Investigator (PI)], AR P60-20557 (to M.S., PI), and NIAMS R55-43507 (to M.W., PI).

Abbreviations: BMD, Bone mineral density; BMI, body mass index; ER, estrogen receptor; FN, femoral neck; HRT, hormone replacement therapy; LS, lumbar spine; OCN, osteocalcin.

Received April 18, 2003.

Accepted October 22, 2003.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Sowers MF 2000 Lower peak bone mass and its decline. Baillieres Best Pract Res Clin Endocrinol Metab 14:317–329[CrossRef][Medline]
  2. Bainbridge KE, Sowers MF, Crutchfield M, Lin X, Jannausch M, Harlow SD 2002 Natural history of bone loss over 6 years among premenopausal and early postmenopausal women. Am J Epidemiol 156:410–417[Abstract/Free Full Text]
  3. Melton LJ, Atkinson EJ, O’Connor MK, O’Fallon WM, Riggs BL 2000 Determinants of bone loss from the femoral neck in women of different ages. J Bone Miner Res 15:24–31[CrossRef][Medline]
  4. Sowers MF, Crutchfield M, Bandekar R, Randolph JF, Shapiro B, Schork MA, Jannausch M 1998 Bone mineral density and its change in pre- and perimenopausal white women: the Michigan Bone Health Study. J Bone Miner Res 13:1124–1140[CrossRef][Medline]
  5. Reid IR, Plank LD, Evans MC 1992 Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J Clin Endocrinol Metab 75:779–782[Abstract]
  6. Revilla M, Villa LF, Hernandez ER, Sanchez-Atrio J, Rico H 1997 Influence of weight and gonadal status on total and regional bone mineral content and on weight-bearing and non-weight bearing bones, measured by dual-energy X-ray absorptiometry. Maturitas 28:69–74[CrossRef][Medline]
  7. Sowers MFR, Kshirsagar A, Crutchfield MM, Updike S 1992 Joint influence of fat and lean body composition compartments on femoral bone mineral density in premenopausal women. Am J Epidemiol 136:257–265[Abstract/Free Full Text]
  8. Sowers MF 1996 Pregnancy and lactation as risk factors for subsequent bone loss and osteoporosis. J Bone Miner Res 11:1052–1060[Medline]
  9. Snow-Harter CM 1994 Bone health and prevention of osteoporosis in active and athletic women. Clin Sports Med 13:389–404[Medline]
  10. Sowers MFR, Galuska DA 1993 Epidemiology of bone mass in premenopausal women. Epidemiol Rev 15:374–398[Free Full Text]
  11. Galuska DA, Sowers MF 1999 Menstrual history and bone density in young women. J Womens Health Gend Based Med 8:647–656[Medline]
  12. Hunter DJ, de Lange M, Andrew T, Snieder H, MacGregor AJ, Spector TD 2001 Genetic variation in bone mineral density and calcaneal ultrasound: a study of the influence of menopause using female twins. Osteoporos Int 12:406–411[CrossRef][Medline]
  13. Arden NK, Baker J, Hogg C, Baan K, Spector TD 1996 The heritability of bone mineral density, ultrasound of the calcaneus and hip axis length: a study of postmenopausal twins. J Bone Miner Res 11:530–534[Medline]
  14. Sowers MF, Boehnke M, Jannausch ML, Crutchfield M, Corton G, Burns TL 1992 Familiality and partitioning of variability of femoral bone mineral density of women of childbearing age. Calcif Tissue Int 50:110–114[CrossRef][Medline]
  15. Willing M, Sowers MF, Aron D, Jannausch M, Clark MK, Burns T, Bunten C, Crutchfield M, D’Agostino D 1998 Bone mineral density and it change in white women: estrogen and vitamin D receptor genotypes and their interaction. J Bone Miner Res 13:695–705[CrossRef][Medline]
  16. Ioannidis JPA, Stavrou I, Trikalinos TA, Zois C, Brandi ML, Gennari L, Albagha O, Ralston SH, Tsatsoulis A 2002 Association of polymorphisms of the estrogen receptor {alpha} gene with bone mineral density and fracture risk in women: a meta-analysis. J Bone Miner Res 17:2048–2060[CrossRef][Medline]
  17. Turner RT, Riggs BL, Spelsberg TC 1994 Skeletal effects of estrogen. Endocr Rev 15:275–300[Abstract/Free Full Text]
  18. Korach KS 1994 Insights from the study of animals lacking a functional estrogen receptor. Science 266:1524–1527[Abstract/Free Full Text]
  19. Brown JP, Delmas PD, Malaval L, Edouard C, Chapuy MC, Meunier PJ 1984 Serum bone GLA-protein: a specific marker for bone formation in postmenopausal osteoporosis. Lancet 1:1091–1093[Medline]
  20. Sallis JF, Haskell WL, Wood PD, Fortmann SP, Rogers T, Blair SN, Paffenbarger 1985 Physical activity assessment methodology in the five-city project. Am J Epidemiol 121:91–106[Abstract/Free Full Text]
  21. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L 1986 A data-based approach to diet questionnaire design and testing. Am J Epidemiol 124:453–469[Abstract/Free Full Text]
  22. Kobayashi S, Inoue S, Hosoi T, Ouchi Y, Shiraki M, Orimo H 1996 Association of bone mineral density with polymorphisms of the estrogen receptor gene. J Bone Miner Res 11:306–311[Medline]
  23. Diggle PJ, Liang K-Y, Zeger SL 1994 Analysis of longitudinal data. Oxford, UK: Clarendon Press
  24. Verbeke G 1997 Linear mixed models for longitudinal data. In: Verbeke G, Molenberghs G, eds. Linear mixed models in practice: a SAS oriented approach. New York: Springer Verlag; 63–153
  25. Albagha OM, McGuigan FE, Reid DM, Ralston SH 2001 Estrogen receptor {alpha} gene polymorphism and bone mineral density: haplotype analysis in women from the United Kingdom. J Bone Miner Res 16:128–134[CrossRef][Medline]
  26. Ravn P, Fledelius C, Rosenquist C, Overgaard K, Christiansen C 1996 High bone turnover is associated with low bone mass in both pre- and postmenopausal women. Bone 19:291–298[Medline]
  27. Lindberg MK, Moverare S, Eriksson AL, Skrtic S, Gao H, Sahlman-Wright K, Gustafsson JA, Ohlsson C 2002 Identification of estrogen-regulated genes of potential importance for the regulation of trabecular bone mineral density. J Bone Miner Res 17:2183–2195[CrossRef][Medline]
  28. Pollitzer WS, Anderson JJB 1989 Ethnic and genetic differences in bone mass: a review with a hereditary vs. environmental perspective. Am J Clin Nutr 50:1244–1259[Abstract/Free Full Text]
  29. Harris S, Dallal GE, Dawson-Hughes B 1992 Influence of body weight on rates of change in bone density of the spine, hip, and radius in postmenopausal women. Calcif Tissue Int 50:19–23[CrossRef][Medline]
  30. Brot C, Jensen LB, Sorensen OH 1997 Bone mass and risk factors for bone loss in perimenopausal Danish women. J Int Med 242:505–511[Medline]
  31. Rubin CT, Lanyon LE 1985 Regulation of bone mass by mechanical strain magnitude. Calcif Tissue Int 37:411–417[Medline]
  32. Rebuffe-Scrive M 1991 Neuroregulation of adipose tissue: molecular and hormonal mechanisms. Int J Obes 15:83–86
  33. Herrington DM, Howard TD, Hawkins GA, Reboussin DM, Xu J, Zheng SL, Brosnihan KB, Meyers DA, Bleecker ER 2002 Estrogen-receptor polymorphisms and effects of estrogen replacement on high-density lipoprotein cholesterol in women with coronary disease. N Engl J Med 346:967–974[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am J EpidemiolHome page
L. Gennari, D. Merlotti, V. De Paola, A. Calabro, L. Becherini, G. Martini, and R. Nuti
Estrogen Receptor Gene Polymorphisms and the Genetics of Osteoporosis: A HuGE Review
Am. J. Epidemiol., February 15, 2005; 161(4): 307 - 320.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sowers, M.
Right arrow Articles by Willing, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sowers, M.
Right arrow Articles by Willing, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals