| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Bone Metabolism Unit (P.B.R., J.C.G.) and Osteoporosis Research Center (G.H.), Creighton University School of Medicine, Omaha, Nebraska 68131
Address all correspondence and requests for reprints to: Prema Rapuri, Bone Metabolism Unit, Creighton University, School of Medicine, 601 North 30th Street, Room 6718, Omaha, Nebraska 68131. E-mail: thiyyari{at}creighton.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the longitudinal study, the rate of bone loss in the placebo group was significantly higher in total body (P < 0.05) and spine (P < 0.05) in women in the lowest tertile, compared with the highest tertile of serum bioavailable estradiol.
After treatment with conjugated equine estrogens 0.625 mg/d, the increase in BMD was 4–6% higher at the femoral sites (P < 0.05), total body (P < 0.05), and spine (not significant), in the lowest tertile, compared with the highest tertile of serum bioavailable estradiol or highest tertile, compared with the lowest tertile of serum SHBG.
In summary, small variations in endogenous serum estradiol and high serum SHBG determine differences in BMD and rate of bone loss in elderly women and also affect the response to treatment with estrogen. Women with a serum estradiol level of less than 9 pg/ml (33 pmol/liter) are optimal candidates for estrogen therapy for osteoporosis prevention.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In this study, which has both cross-sectional and longitudinal data, we examined the following hypotheses: lower serum estradiol (both total and bioavailable) and higher serum SHBG levels would be associated with lower bone mineral density (BMD), increased bone resorption, and an increased rate of bone loss, and women with lower serum estradiol and higher bone resorption would show a larger increase in BMD after treatment with estrogen.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
For cross-sectional analyses, we used the baseline data of 489 women enrolled into the study. One subject with suspected Paget’s disease was excluded from the analyses. After excluding women with no available measurements of serum total and bioavailable estradiol and serum SHBG, BMD at different skeletal sites and biochemical indices were compared between tertiles of total estradiol (n = 401), bioavailable estradiol (n = 400), and SHBG (n = 401).
For the longitudinal analyses, there were available measurements on 107 of a total of 112 women in the placebo group for serum total estradiol, bioavailable estradiol, and SHBG. In the HT/ET group, there were measurements available on 76 of a total of 101 women for serum total and bioavailable estradiol and on 77 women for serum SHBG. In the HT/ET + calcitriol group, the measurements were available on 73 of 102 women for serum total and bioavailable estradiol and SHBG. The percent change in BMD and various biochemical measures were compared between the tertiles of serum total estradiol, bioavailable estradiol, and SHBG.
Dietary intake, smoking, and alcohol history
Dietary intake at baseline and at the end of the study (36 months) was assessed using 7-d food diaries. Subjects completed a 7-d food diary and nutrient supplement record. Plastic food models (NASCO, Fort Atkinson, WI) were used to help participants to better estimate the quantities consumed. The average daily intakes of calcium and vitamin D intakes were calculated using the Food Processor II plus nutrition and diet analysis system (version 5.1; Esha Research, Salem, OR). At baseline recruitment and 36 months, participants were provided with a questionnaire to report their smoking and alcohol history; reproductive history; and present use of medications, vitamins, and mineral supplements. Current smokers were considered as smokers, whereas past smokers and women who never smoked were classified as nonsmokers. Alcohol use was stratified into drinkers and nondrinkers.
Biochemical analysis
Fasting blood samples and 24-h urine were obtained from the subjects at baseline and the end of the study (36 months). Serum was separated from the blood samples after allowing them to clot and centrifuging them at 2056 x g for 15 min at 4 C. The serum samples were stored at –70 C until analysis. All serum and urine chemistries were measured by automated procedures (chemistry analyzer, Nova Nucleus, Waltham, MA).
Endogenous hormones
Serum total estradiol and SHBG were measured in fasting baseline serum samples by RIA kits (Diagnostic Systems Laboratories, Webster, TX). Serum total estradiol was measured using an ultrasensitive assay. The minimum detection limit for these assays was 2.2 pg/ml for serum estradiol and 3 nmol/liter for serum SHBG. The intraassay coefficient was less than 5% for these assays. The interassay coefficient for the last 10 assays for serum total estradiol for the lowest standard used (5 pg/ml) was 8.4%, and the interassay coefficient for the in-house serum control (mean ± SD, 11.8 ± 1.35) used was 11.4%. The interassay coefficient for the lowest kit control tested for SHBG was 8.5%. The bioavailable (non-SHBG bound) serum estradiol was measured as described by Khosla et al. (37). Briefly, tracer amounts of 3H-estradiol were added to aliquots of serum (200 µl), made to 500 µl with saline. To this, equal volume of saturated solution of ammonium sulfate (final concentration, 50%) was added, which precipitates the SHBG with its bound steroid. The SHBG bound and unbound steroid were separated by centrifugation at 1100 x g for 30 min at 4 C. The percentage of labeled estradiol remaining in the supernatant (free and albumin bound fractions) was then calculated. The bioavailable estradiol concentration was obtained by multiplying the total estradiol concentration, as determined by RIA, by the fraction that was non-SHBG bound. All women except one had measurable levels of serum estradiol.
Calcium absorption test
Calcium absorption was measured in a fasting state by oral administration of 5 µCi (18.5 x 104 Bq) of 45Ca (Amersham, Arlington Heights, IL) in a 100-mg calcium chloride (CaCl2) carrier given in a total of 250 ml distilled water (13). A blood sample was collected at 2 and 3 h after the oral dose. 45Ca activity was counted in 2 ml serum using the 1900 CA tricarb liquid scintillation analyzer (Packard Instruments, Meriden, CT). A parallel standard taken from the patient’s dose before ingestion was counted at the same time. Calcium absorption was expressed as percent of actual dose per liter of blood (% AD/liter) and corrected for body mass index (BMI).
Calcitropic hormones and bone markers
Serum intact PTH was measured with Allegro immunoradiometric assay (Nichols Institute, San Juan Capistrano, CA). The interassay variation was 5% and the limit of detection for the assay was 1 ng/liter (1 pg/ml). Serum 25-hydroxy vitamin D [calcidiol or 25(OH)D] was assayed with a competitive protein binding assay (38) after prepurification of serum on Sep-Pak cartridges (39). Briefly, after precipitating plasma proteins with acetonitrile, the supernate was backwashed with potassium phosphate, 0.4 M (pH 10.5), to enhance the solubilization of lipids. The samples were then extracted on a reverse-phase Sep Pak C-18 columns (Waters Associates, Milford, MA). The acetonitrile fraction containing the vitamin D metabolites was taken through a normal phase extraction with a silica Sep Pak cartridge (Waters Associates, Milford, MA), in which the 25(OH)D was eluted with 96:4 hexane to isopropanol. 25(OH)D was quantitated by competitive protein binding assay employing normal rat serum as the source of binding protein. The competitive protein binding assay method was cross-calibrated with HPLC (40). The limit of detection for the assay was 5 µg/liter (12.5 mmol/liter) and the interassay variation was 5%. Serum concentrations of osteocalcin were determined by RIA (Diasorin Inc., Stillwater, MN). The limit of detection was 0.78 ng/ml (0.78 µg/liter) and the interassay variation was 5%. Urine collagen cross-links were measured by ELISA (Osteomark International, Seattle, WA) as N-telopeptides, a marker for bone type I collagen. The lower limit of detection was 20 nmol bone collagen equivalents (BCE), and the interassay variation was 6%. The data are expressed as nanomoles BCE per micromoles of creatinine.
BMD
BMD (grams per square centimeter) was measured using a dual-energy x-ray absorptiometry (model DPX-L, Lunar Corp., Madison, WI). The lumbar spine (L1-L4), total hip, two sites in proximal femur (femoral neck and trochanter), and whole-body BMD was determined by using standardized protocols for uniform subjects positioning, scan mode, and scan analysis. The hip and spine scans were performed in duplicate, and the mean was used for the analysis. The percent change in BMD was calculated as the difference between baseline and follow-up BMD (36-month value), divided by baseline BMD, and multiplied by 100.
Statistical analysis
Data were analyzed with the SAS statistical package (SAS.SAS/Stat User’s Guide, version 8.8.2; 2000, SAS Institute, Cary, NC). For the cross-sectional study, baseline characteristics were compared between the tertiles of endogenous estradiol and SHBG using a one-way ANOVA. For the cross-sectional analyses, biochemical indices and BMD measurements between tertiles of endogenous estradiol and SHBG were compared using analysis of covariance (ANCOVA) with adjustments for various relevant confounders (age, BMI, calcium intake, caffeine intake, fiber intake, vitamin D intake, smoking and alcohol drinking status), selected by the stepwise selection method. For the longitudinal analyses, similar ANCOVA models were considered with outcome variables being the percent change in BMD and biochemical variables and the corresponding baseline values added to the list of the independent variables in the models. The residuals of the final models were tested by graphic methods for deviation from normality. The BMD and biochemical parameters of the tertiles of endogenous estradiol and SHBG in both the cross-sectional and longitudinal analyses are summarized by the unadjusted least squares means and their respective SEs. The significance was derived from adjusted data with Tukey’s post hoc multiple comparison test to determine post hoc significance between the tertile groups.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Baseline study (Table 1
).
At baseline, mean serum total estradiol for the tertiles was 7.1 ± 0.1 (26.1 ± 0.51), 11.0 ± 0.1 (40.4 ± 0.37), and 17.7 ± 0.3 (65.0 ± 1.1) pg/ml (picomoles per liter) (Table 1). The mean serum bioavailable estradiol for the tertiles was 1.8 ± 0.04 (6.6 ± 0.15), 3.1 ± 0.04 (11.4 ± 0.15), and 5.8 ± 0.17 (21.3 ± 0.62) pg/ml (picomoles per liter) (Table 1). There was a high correlation between bioavailable and bound forms of serum estradiol (r = 0.88, P < 0.001) (Fig. 1A). Women in the highest tertile of serum total estradiol and bioavailable estradiol had significantly (P < 0.05) higher body weight and total body fat, compared with the lowest tertile (Table 1). For each 10-kg increase in body fat, serum bioavailable estradiol increased by 2.1 pg/ml (7.7 pmol/liter) (Fig. 1C). Mean serum SHBG level for the tertiles was 80.7 ± 1.8, 138 ± 1.21, and 222 ± 4.6 nmol/liter (Table 1). Serum SHBG was inversely correlated with total estradiol (r = –0.21, P < 0.001) and bioavailable estradiol (r = –0.49, P < 0.001) (Fig. 1B). Women in the lowest tertile of serum SHBG had significantly (P < 0.05) higher body weight and total body fat, compared with the highest tertile (Table 1). Age, age at menopause, and dietary calcium intake were not different across the tertiles of serum total estradiol, bioavailable estradiol, and SHBG at baseline (Table 1).
|
|
Baseline study and BMD (Table 1
and Fig. 2).
A comparison of BMD across the tertiles of serum bioavailable estradiol showed that in the lowest tertile, compared with the highest tertile, BMD was significantly (P < 0.05) lower at the femoral sites neck (10%) and trochanter (19%), total body (8%), and total femur (14%) and spine (9% not significant) (Fig. 2A and Table 1). The findings were similar for tertiles of serum total estradiol (data not given). In the highest tertile of serum SHBG, BMD was significantly (P < 0.05) lower at femoral sites neck (7%) and trochanter (12%) and total femur (10%) and total body (6%) and spine (5% not significant), compared with the lowest tertile (Fig. 2B and Table 1). Similar observations were made with respect to the T scores at multiple skeletal sites (data given for spine and femoral neck in Table 1). The association between BMD and serum bioavailable estradiol was much stronger than that seen with total estradiol or SHBG.
|
).
In the placebo group, the rate of bone loss over 3 yr in women in the lowest tertile of bioavailable estradiol, compared with the highest tertile, was significantly higher at the spine (–3.3% vs. –0.66%, P < 0. 05) and total body (–3.1% vs. –1.6%, P < 0.05) (Fig. 3A). There were no significant differences at the hip sites. The rate of bone loss was not significantly different across the tertiles of serum total estradiol and SHBG. However, a trend of increased bone loss at spine (–2.75 ± 0.73 vs. –0.85 ± 0.60) and total body (–2.8 ± 0.43 vs. –1.72 ± 0.50) in the lowest tertile, compared with the highest tertile of serum total estradiol, was observed. Similarly, there was a trend of higher bone loss in the highest tertile, compared with lowest tertile of serum SHBG at total body (–2.55 ± 0.51 vs. –1.49 ± 0.50) but not spine (–1.45 ± 0.82 vs. –1.33 ± 0.74).
|
), or in the highest vs. lower tertile of SHBG (data not given). Similar nonsignificant trends were seen for spine BMD. Results were similar for tertiles of serum total estradiol.
In the group given a combination of HT/ET + calcitriol, the response to treatment at all skeletal sites was higher in the lowest tertile of serum bioavailable estradiol, compared with that in the highest tertile, statistical significance being observed only for the femoral sites (Fig. 3C). Similar observations were made across the tertiles of serum total estradiol (data not given). The response to combination treatment with HT/ET + calcitriol was higher in the highest tertile of SHBG, compared with that of the lowest tertile, although statistical significance was seen only at the total femur site (data not given).
Baseline study and biochemical variables
Table 1 summarizes the relation at baseline between serum total estradiol, bio-available estradiol, and SHBG divided into tertiles and the biochemical measurements, after adjusting for the covariates mentioned in the statistical section. At baseline, the lowest tertile of serum bioavailable estradiol, compared with the highest tertile, had significantly (P < 0.05) higher serum ionized calcium, higher serum 25(OH)D, lower serum PTH, and higher serum osteocalcin; urine N-telopeptide of collagen cross-links (NTx) to creatine (Cr) ratio was higher but not significant (Table 1). Serum 1,25 dihydroxyvitamin D3 and calcium absorption were not different across the tertiles of serum bioavailable estradiol. Similar observations were made across the tertiles of serum total estradiol; however, statistical significant differences were noted between the lowest and highest tertiles for serum ionized calcium and serum 25(OH)D levels only. In contrast, the lowest tertile of serum SHBG had significantly lower serum 25(OH)D, serum osteocalcin, urine NTx to Cr ratio, and higher PTH, compared with the highest tertile (Table 1). Serum ionized calcium, serum 1,25 dihydroxyvitamin D3, and calcium absorption were not different across the tertiles of SHBG.
Longitudinal study and biochemical variables (Table 2).
In the placebo group, women in the lowest tertile of serum total and bioavailable estradiol had significantly (P < 0.05) higher decrease in serum ionized calcium, compared with the respective highest tertile (Table 2). The percent change over time in serum osteocalcin was higher in the lowest tertile of both serum total and bioavailable estradiol, compared with the highest tertile, although not significant (Table 2). Similarly, the percent change in urine NTx/Cr ratio was higher in the lowest tertile, compared with the highest tertile of serum total estradiol (but not bioavailable estradiol). The percent change in serum PTH (Table 2), 25(OH)D (Table 2) and calcium absorption (data not given) were not different across the tertiles of serum total and bioavailable estradiol. The percent change in serum ionized calcium was significantly (P < 0.05) higher in the highest tertile of serum SHBG, compared with the lowest tertile, whereas no other significant differences were noted across the tertiles of serum SHBG (data not given).
|
), serum PTH (Table 2), serum 25(OH)D levels (Table 2), and calcium absorption (data not given) across the tertiles of serum total and bioavailable estradiol. The percent change in serum osteocalcin was significantly higher in the highest tertile of serum total and bioavailable estradiol, compared with their respective lowest tertiles (Table 2). The percent change in urine NTx to Cr ratio was not different between the highest and lowest tertile of serum total and bioavailable estradiol. No significant differences were noted with regard to any of the biochemical measurements across the tertiles of SHBG (data not given).
In women receiving the combination treatment of HT/ET + calcitriol, there were no differences in percent change in any of the biochemical measurements across the tertiles of serum total and bioavailable estradiol (Table 2) and SHBG (data not given).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Similar associations between endogenous levels of serum estradiol and BMD and rate of bone loss were demonstrated earlier. In the Study of Osteoporosis Fractures, Ettinger et al. (27) reported that women with a serum total estradiol of 10–25 pg/ml (36.7–91.7 pmol/liter) (or SHBG 1.4 µg/dl) have higher BMD than those with levels below 5 pg/ml (18.4 pmol/liter) (SHBG 1.7 µg/dl). Furthermore, in the same population, it was shown that serum estradiol and SHBG predicts subsequent bone loss (31) and the risk of subsequent vertebral and hip fractures. Slemenda et al. (30) reported that SHBG and sex steroids account for a substantial proportion of the variance in BMD and a smaller proportion of the variance in bone loss. Ooms et al. (41) showed that in the elderly, higher serum SHBG is associated with low BMD at femoral neck, trochanter, and distal radius, suggesting that lower estrogen levels influence bone turnover and trabecular bone loss in the very old.
In the present study, women with serum total estradiol greater than 13 pg/ml (49 pmol/liter) had 6–14% higher baseline BMD at the spine, femoral sites, and total body, compared with women with serum estradiol less than 9 pg/ml (33 pmol/liter). The associations were stronger for serum bioavailable estradiol with women with serum bioavailable estradiol levels of greater than 4 pg/ml (14 pmol/liter) having BMD 8–19% higher at the various skeletal sites, compared with that of women with serum bioavailable estradiol less than 2.4 pg/ml (8.8 pmol/liter). Serum SHBG levels showed a significant negative association with BMD. Although significantly different for the hip sites and total-body BMD, the lack of statistically significant differences in spine BMD between the highest and lowest tertiles of serum total and bioavailable estradiol and serum SHBG could possibly be due to artifacts such as arthritis, which independently increase BMD measurements during aging. Although serum estradiol was associated with higher body weight and total body fat, the associations with BMD remained highly significant after adjusting for body weight and total body fat, suggesting that endogenous secretion of estradiol from the ovary and adrenal gland is important. The findings in the cross-sectional study were further confirmed in the prospective study in which the rate of bone loss over 3 yr in women on a placebo was significantly higher in women with very low levels of endogenous baseline serum bioavailable estradiol.
The importance of the role of endogenous hormones in determining BMD in the elderly is further substantiated by the fact that at baseline women in the lowest tertile of serum bioavailable estradiol and highest tertile of SHBG had significantly higher indices of bone remodeling. The magnitude of differences in the mean levels of the bone formation marker osteocalcin between the highest and lowest tertile of bioavailable estradiol and SHBG were 17 and 22%, respectively. In the longitudinal study, similar nonsignificant observations were seen in women receiving the placebo. At baseline, the difference in bone resorption marker, urinary N-telopeptides, between the highest and the lowest tertiles of bioavailable estradiol and SHBG were 20 and 31%, respectively. However, in the longitudinal study, urine N-telopeptides were not different across the tertiles. This could be explained by high biological variability in urine NTx levels and small numbers in the three tertiles. Consistent with our results, Heshmati et al. (35) recently demonstrated that even in late postmenopausal women, a reduction in the serum estradiol levels with an aromatase inhibitor increased bone remodeling markers. In the study of osteoporotic fractures, Chapurlat et al. (34) reported that in older women with low endogenous hormones, the bone turnover is marginally higher. A positive association between SHBG and bone turnover markers has also been shown by Ooms et al. (41).
In the present study, small changes in serum bioavailable estradiol at baseline are associated with increased bone resorption. Women in the lowest tertile of serum bioavailable estradiol have higher serum ionized calcium, decreased PTH, elevated serum 25(OH)D and marginally lower calcium absorption, a classical response seen in maintenance of calcium homeostasis. Similar trends were found with serum total estradiol and SHBG. Increased bone resorption due to estrogen deficiency results in a loss of skeletal calcium, which in turn increases serum calcium. The elevated serum calcium inhibits the production of PTH and also decreases the calcium absorption.
In the treatment phase of the study, we report for the first time that the increases in BMD in response to estrogen treatment (both in HT/ET and HT/ET + calcitriol groups) were much higher in women with the lower baseline levels of endogenous serum estradiol (total and bioavailable), compared with higher levels. The changes in BMD in response to HT/ET treatment, however, are not corroborated by the changes in biochemical markers of bone turnover, possibly due to high biological variability (42) and the low numbers.
Overall, these findings suggest that women with the greatest degree of relative estrogen deficiency benefit the most with estrogen treatment, compared with those who are better able to maintain their endogenous estradiol levels. Considering the highly variable endogenous estrogen status in elderly postmenopausal women, measurement of endogenous levels of serum estradiol before initiation of estrogen treatment would allow individualized selection of the appropriate women to be treated with estrogen for treatment of bone loss. Because of the large response in BMD to the dose of estrogen (conjugated equine estrogens 0.625 mg) used in this study, it is likely that much smaller doses of estrogen could be used for bone loss with fewer side effects. The use of ultrasensitive assays of serum estradiol should allow for individual titration of the estrogen dose similar to that done for patients with hypothyroidism on thyroxine, using serum T4, and TSH.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
This work in part was presented as an abstract at 24th and 25th Annual Meetings of the American Society for Bone and Mineral Research, San Antonio, Texas, September 20–24, 2002, and Minneapolis, Minnesota, September 19–23, 2003.
Abbreviations: % AD/liter, Percentage of actual dose per liter of blood; ANCOVA, analysis of covariance; BCE, bone collagen equivalent; BMD, bone mineral density; BMI, body mass index; Cr, creatine; ET, estrogen therapy; HT, hormone therapy; NTx, N-telopeptide of collagen cross-links; 25(OH)D, 25-hydroxy vitamin D.
Received March 17, 2004.
Accepted June 25, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-carboxyglutamic acid protein with aging in women. Implications for the mechanism of age-related bone loss. J Clin Invest 71:1316–1321
This article has been cited by other articles:
 |
|
 |
|
  F. Z. Stanczyk, J. S. Lee, and R. J. Santen Standardization of Steroid Hormone Assays: Why, How, and When? Cancer Epidemiol. Biomarkers Prev., September 1, 2007; 16(9): 1713 - 1719. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Riancho, C. Valero, A. Naranjo, D. J. Morales, C. Sanudo, and M. T. Zarrabeitia Identification of an Aromatase Haplotype That Is Associated with Gene Expression and Postmenopausal Osteoporosis J. Clin. Endocrinol. Metab., February 1, 2007; 92(2): 660 - 665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. T Zarrabeitia, J. L Hernandez, C. Valero, A. Zarrabeitia, J. A Amado, J. Gonzalez-Macias, and J. A Riancho Adiposity, estradiol, and genetic variants of steroid-metabolizing enzymes as determinants of bone mineral density Eur. J. Endocrinol., January 1, 2007; 156(1): 117 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Missmer, D. Spiegelman, E. R. Bertone-Johnson, R. L. Barbieri, M. N. Pollak, and S. E. Hankinson Reproducibility of Plasma Steroid Hormones, Prolactin, and Insulin-like Growth Factor Levels among Premenopausal Women over a 2- to 3-Year Period. Cancer Epidemiol. Biomarkers Prev., May 1, 2006; 15(5): 972 - 978. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, P < 0.05, compared with tertile 2.