| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
University of Jyväskylä (Q.W., P.H.F.N., M.S., A.L., E.H., H.S., S.C.) and LIKES-Foundation for Sport and Health Sciences (Q.W., M.S., A.L.), Jyväskylä Fin-40014; and Peurunka-Medical Rehabilitation Center (M.A.), Jyväskylä Fin-41340, Finland
Address all correspondence and requests for reprints to: Sulin Cheng, Department of Health Sciences, P.O. Box 35 (LL), Fin-40014 University of Jyväskylä, Finland. E-mail: Cheng{at}sport.jyu.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Testosterone (T) has profound effects on longitudinal bone growth, but its effects on bone calcium accrual or cross-sectional development have been investigated even less extensively than the effect of E2. It is believed that T also exerts its effect in adult men by suppressing bone turnover and that the effects may be mediated by the estrogen pathway (6). Some of the studies in adult females also indirectly confirmed that the effect of T on bone mass and its change are mediated by E2 (7, 8).
Sex hormones are mainly stored and transported in serum by binding to their sex hormone-binding globulin (SHBG), which is an important regulator of the bioavailability of sex hormones. Studies in old women have shown that SHBG levels are negatively associated with bone mass and its change and that the association is independent of serum E2 level (9, 10, 11), which indicates that an extra pathway exists for SHBG to exert its effect on bone mass. Such a phenomenon has not been investigated in growing girls.
In the present study, we measured the midshaft of the tibia using peripheral quantitative computed tomography (pQCT), a technique capable of separately analyzing cortical bone, trabecular bone, and marrow cavity compartments. Our aim was to investigate whether levels of E2, T, or SHBG were associated with the geometric and densitometric properties of the tibia.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The subjects consisted of 248 Finnish girls aged 10–13 yr. To be eligible for the study, the participants had to have no history of disease or medication known to affect bone metabolism. None of the participants had menarche. The physical maturational stage was determined by the developmental patterns of breast and pubic hair assessed by a public nurse according to the Tanner grade system (12). If there was a disagreement between breast and pubic hair development, the final decision was made based on the development of breast. The numbers of girls at each maturational stage were 126, 109, and 13 at Tanner stages I, II, and III, respectively. All participants provided written consents by themselves and their guardians also provided consent in accordance with the ethical committees of University of Jyväskylä, the Central Hospital of Central Finland, and the Finnish National Agency of Medicines.
Anthropometric assessments
Body weight was measured using an electronic scale and recorded to the 0.1 kg with subjects wearing light clothes and no shoes. Height was determined with stadiometer to the nearest 0.5 cm.
Sex hormone assessments
Serum T, E2, and serum SHBG were assessed using time-resolved fluoroimmunoassays (Delfia; Wallac Oy, Turku, Finland). Blood was drawn in the morning between 0700 and 0900 after an overnight fast. Inter- and intraassay coefficients of variation were 5.2% and 5.1% for E2, 9.2% and 9.4% for T, and 1.1% and 1.1% for SHBG, respectively. The free E2 was calculated as: free E2 (pmol/liter) = E2 (pmol/liter)/[K * SHBG (nmol/liter) +1], and the free T was calculated as: free T (pmol/liter) = T (nmol/liter) * 1000/[K * SHBG (nmol/liter) + 1], according to Ekins (13), where K for E2 and T binding to SHBG are the equilibrium constants (0.68 * 109 liter/mol and 1.6 * 109 liter/mol, respectively).
There were 22 girls with the serum T concentration under detectable level, and therefore, they were excluded from the analysis of the association between T and other variables.
Bone densitometric and geometric measurements
A pQCT device (XCT-2000; Stratec Medizintechnik, GmbH, Pforzheim, Germany) was used to scan the left tibial shaft of the participants. A 2-mm thick single tomographic slice with pixel size of 0.59 mm was taken from the transverse plane 60% of lower leg length up from the lateral malleolus. The lower leg length was defined as the distance between the lateral condyle of tibia and lateral malleolus and measured when the participant was in sitting position with the knee at a 90° angle. The pQCT device was calibrated once a week using a standard phantom and once a month using a cone phantom provided by the manufacture.
The results were analyzed using Bonalyse 1.3 software (Bonalyse Oy, Jyväskylä, Finland). The outer-bone border was determined using a specified threshold of 280 mg/cm3. The total cross-sectional area (CSA) of the tibial shaft was then defined as the area enclosed within the outer-bone border. The total CSA was separated into three parts, the cortex, subcortex, and medullary cavity, based on the following two thresholds: 711 mg/cm3 to distinguish cortex and subcortex, and 100 mg/cm3 to separate subcortex and medullary cavity. The areas of these three compartments were then expressed as percentages relative to the total bone CSA. The volumetric bone mineral density (vBMD, mg/cm3) of the total and cortical bone was also obtained. Cortical thickness was obtained from the Stratec software version 5.4 (Stratec Medizintechnik) using the circular ring model in which the cortical thickness was calculated as follows: (total bone CSA/
)0.5 – [(total bone CSA – cortical bone CSA)/]0.5. The coefficients of variation were 1% for CSA and less than 1% for vBMD and cortical thickness.
Statistical analysis
Descriptive statistics were used to present the background and bone information. All the hormone values were first transformed into normal or approximately normal distribution by applying a logarithm or square root transform before further analysis. Body weight and height were highly correlated with each other, and therefore, they were reduced to one body size index using principal components factor analysis, and this index was used for adjustment instead of body weight and height. The correlations between age, body weight and height, sex hormones, and bone geometric and densitometric properties were assessed using the Pearson product moment correlation. Partial correlation analysis was used to evaluate the relationship of bone geometric properties and vBMD to hormone variables, controlling for age and body size index. To explore the E2-independent effect of SHBG on bone density and geometric properties, partial correlation analysis was used, adjusting for total E2, age, and body size. A P < 0.05 was considered statistically significant. Statistical analysis was carried out using SPSS version 11.0 for Windows (SPSS Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
.
|
). Age did not correlate with cortical, subcortical, or marrow proportions, or with total vBMD of tibial shaft, but it did correlate positively with cortical vBMD (P = 0.001). Body weight was positively associated with cortical proportion and negatively associated with subcortical proportion (P = 0.009 and P < 0.001, respectively), but it was not associated with either tibial total or cortical vBMD. Body height was negatively correlated with subcortical proportion (P < 0.001), but it was not associated with tibial total or cortical vBMD or cortical proportion.
|
), and negatively correlated with subcortical and marrow proportions (P = 0.007 and P = 0.013, respectively). Free E2 had similar results to E2.
|
After controlling for age and body size, the results showed that E2 correlated positively with total vBMD, as well as with cortical thickness and proportion (P < 0.001, Table 3), and negatively with marrow proportion (P = 0.003, Table 3), but E2 did not correlate with either total CSA or cortical vBMD of tibia shaft. Free E2 showed similar associations with these bone geometric and densitometric properties as total E2. However, T and free T did not show any relationship with these bone variables. SHBG correlated negatively with total vBMD, as well as with cortical thickness and proportion, and positively with marrow proportion (P = 0.005, P = 0.010, P = 0.029, and P = 0.007, respectively; Table 3). When further controlled for the total E2 level, the negative correlation between SHBG and total vBMD, cortical thickness, and cortical proportion, and the positive correlation between SHBG and marrow proportion remained (P = 0.003, P = 0.021, P = 0.038, and P = 0.014, respectively).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
During early puberty, body height increases by 0.5–1.0 cm per month; therefore, even a half year’s difference in age can elicit large differences in bone size. To evaluate the association of E2 and bone geometric properties, the confounding effect of age and body height and weight must be controlled for.
In the present study, we divided the total CSA of tibial shaft into three compartments (cortex, subcortex, and marrow cavity) and expressed them as proportions to the total CSA and found that no correlation existed between any of the three proportions and chronological age. Because the total CSA was not correlated with E2 after adjustment for age and body size, the larger cortical proportion and the thickening of the cortex relating to higher E2 level cannot be attributed to periosteal expansion but can only be due to reduction, or relatively less expansion, of the marrow cavity. Our results indicate that E2 may not influence the total CSA, but rather, E2 acts to constrict the marrow cavity or suppress bone resorption at the endosteal surface. Libanati et al. (14) reported a decrease in marrow diameter of the second metacarpal bone in girls with more advanced puberty, which is consistent with our finding. On the other hand, they did not find any correlation between E2 and second metacarpal indices, such as total bone thickness, cortical thickness, or marrow diameter. However, the authors reported that both the E2 level and total metacarpal thickness increased with the increase of the maturational stages, which indicated that the sample size might not be big enough to show the slight correlation between E2 and marrow diameter. Many studies have demonstrated that serum E2 levels are negatively associated with both bone formation and resorption markers during the growth or postmenopausal periods (15, 16, 17, 18, 19, 20, 21), which indicates that E2 at physiological levels might exert its suppressing effect on both bone formation and resorption, which are coupled. A study by Plato and Purifoy (22) showed that postmenopausal women who were currently using estrogen had significantly smaller medullary cavity and larger cortical area of the second metacarpal bone compared with estrogen nonusers measured using radiographic morphometry. Using the same method, Horsman et al. (23) also demonstrated that a high dose of E2 up to 25 mg/d could completely inhibit bone resorption at the endosteal surface. Although the physiological conditions are different between postmenopausal women and growing girls, the effect of E2 on bone resorption at the endosteal surface is evident in the two different populations. Therefore, we could propose that the negative relationship of E2 with the marrow cavity proportion found in our study is mainly due to the suppressing effect of E2 on the bone resorption and not from new bone formation.
Total vBMD measured by pQCT is defined as the mass of mineral contained in the bone slice per volume of the slice. It is mainly determined by the following three elements: mean mineral density, the porosity of solid components, and the structural compartmentalization of total CSA. The pQCT provides us with a possibility to separately analyze the trabecular, subcortical, and cortical bone compartments. No correlation was found between serum E2 level and cortical vBMD of tibial shaft in our study. This implies that physiological E2 levels do not influence the mineralization or the porosity of the cortical bone. This view is supported by ovariectomized young animal studies in which the cortical vBMD at long bone shaft did not differ over experimental period (24, 25). The positive association between cortical vBMD and age found in this study, which is consistent with a cross-sectional study by Schoenau et al. (26) showing that cortical vBMD increases until adulthood, may not be related to changes in E2 level over the growing and fertile period but, instead, may be related to other factors. Naturally, it is conceivable that the positive correlation between E2 and total vBMD, with or without adjustment for age and body size, is due to the smaller proportion of marrow cavity of the total CSA of tibial shaft associated with high E2 levels.
Bone size and mass are determined by many factors such as heredity, body size, aging, disease, diet, physical activity, medication, and alcohol and tobacco use. Evidence from twin and family studies suggests that genetic factors account for up to 85% of the variance in peak BMD and other bone properties. Although epidemiological and experimental studies have identified estrogen deficiency as an important risk factor for osteoporosis and E2 as a vital determinant of sexual skeletal dimorphism, estrogen exposure is not necessarily required for the normal occurrence of the major functions of the skeleton, which has been exhibited in sexually immature and hypogonadal individuals. Few studies have reported the amount of the variance explained by E2 in bone mass during puberty. In the present study, E2 was found to account for only 5.2% of the variance in total vBMD, 6.7% in cortical proportion of total tibial CSA, and 2.5% in marrow proportion. We have not found any comparable studies with which to confirm our results.
Compared with E2, the effects of T on bone mineralization and cross-sectional development appear insignificant after controlling for body size. Very few studies have investigated the effects of T on development of bone CSA in pubertal girls to date. One study from Libanati et al. (14) reported that free serum T level positively related to cortical thickness and total bone thickness of the second metacarpal bone in girls at Tanner stage II–IV. These findings are in agreement with our results that T level positively correlated with cortical thickness and total bone CSA of tibia shaft when body size was not controlled (see Table 2
). However, the significant correlations between T and cortical thickness and bone CSA of the tibia shaft disappeared when the difference of body size between individuals was adjusted. This indicates that the effect of growth plays a critical role in the relationship between T and bone size in growing girls, which suggests that it is improper to leave the body size ignored when assessing the relationship between sex hormones and bone size at the growing period. It is believed that the effects of androgens on bone mineralization are mediated by estrogens. This is supported by results in men, which showed that destructive mutation in the ER gene (estrogen resistance) or in the CYP19 gene (aromatase deficiency) is associated with low BMD despite adequate or even well above normal androgen levels (27, 28, 29, 30). Men with aromatase deficiency respond to estrogen therapy with marked increases in bone mass, whereas T treatment has no effect on bone mass (27, 28, 29, 30). In the present study, no significant correlation was found between T and either vBMD or cortical and marrow proportion. This is in agreement with some previous studies in which T showed less association with bone mineral density (31, 32).
SHBG modulates the bioavailability of sex steroids by decreasing their serum free levels. We found that SHBG was negatively associated with bone variables, which was consistent with previous studies in older women (9, 10). One study even showed that SHBG was superior to sex hormone levels for predicting the occurrence of postmenopausal fracture (11). Another explanation for our finding could be that SHBG actually has a more complex role than simply acting as a transporter and bioavailability modulator of sex steroids in the circulatory system (33, 34). An inhibiting effect of SHBG on the E2 induction of cell proliferation through SHBG receptor and cAMP was also found in estrogen-dependent breast cancer (35). It is conceivable that the SHBG/SHBG-receptor system, if it exists in bone tissue, may also exert an antiestrogen effect on the skeletal system through this mechanism. Our results support this view.
Compared with chronological age, biological age is a more relevant and reliable measurement for monitoring the stage of bone growth during puberty. In this study, we were not able to determine the biological age of the bone because of the study protocol. Another shortcoming is that we only measured the serum E2 level, which fluctuates diurnally, once for every girl. It is possible that a single measurement did not give us a representative indicator of E2 exposure. Nevertheless, in clinical settings, a single measurement from serum sample drawn in the early morning is commonly used and acceptable.
In summary, we conclude that, in early pubertal girls, E2 levels are associated positively with vBMD, cortical thickness, and proportion of cortical bone relative to total CSA and negatively with medullary area proportion, and are not associated with total bone CSA. SHBG showed precisely the opposite pattern of associations with bone properties as those seen with E2, and T was not associated at all with bone properties. Our results indicate that E2 influences bone cross-sectional development mainly by suppressing bone resorption at endocortical surface and provide new insights into the effects of sex hormones on bone in early pubertal girls.
|
Footnotes |
|---|
Abbreviations: CSA, Cross-sectional area; E2, 17ß-estradiol; pQCT, peripheral quantitative computed tomography; SHBG, sex hormone-binding globulin; T, testosterone; vBMD, volumetric bone mineral density.
Received June 30, 2003.
Accepted December 30, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Utriainen, J. Jaaskelainen, A. Saarinen, E. Vanninen, O. Makitie, and R. Voutilainen Body Composition and Bone Mineral Density in Children with Premature Adrenarche and the Association of LRP5 Gene Polymorphisms with Bone Mineral Density J. Clin. Endocrinol. Metab., November 1, 2009; 94(11): 4144 - 4151. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Travison, A. B. Araujo, T. J. Beck, R. E. Williams, R. V. Clark, B. Z. Leder, and J. B. McKinlay Relation between Serum Testosterone, Serum Estradiol, Sex Hormone-Binding Globulin, and Geometrical Measures of Adult Male Proximal Femur Strength J. Clin. Endocrinol. Metab., March 1, 2009; 94(3): 853 - 860. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D M L Cooper, Y Ahamed, H M Macdonald, and H A McKay Characterising cortical density in the mid-tibia: intra-individual variation in adolescent girls and boys Br. J. Sports Med., August 1, 2008; 42(8): 690 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kaptoge, N. Dalzell, E. Folkerd, D. Doody, K.-T. Khaw, T. J. Beck, N. Loveridge, E. B. Mawer, J. L. Berry, M. J. Shearer, et al. Sex Hormone Status May Modulate Rate of Expansion of Proximal Femur Diameter in Older Women alongside Other Skeletal Regulators J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 304 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Wang, M. Alen, P. H. F. Nicholson, J. M. Halleen, S. L. Alatalo, C. Ohlsson, H. Suominen, and S. Cheng Differential Effects of Sex Hormones on Peri- and Endocortical Bone Surfaces in Pubertal Girls J. Clin. Endocrinol. Metab., January 1, 2006; 91(1): 277 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J H Davies, B A J Evans, and J W Gregory Bone mass acquisition in healthy children Arch. Dis. Child., April 1, 2005; 90(4): 373 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mora Relationship of Sex Hormones to Bone Geometric Properties and Mineral Density in Early Pubertal Girls: Use of Correlation Analyses J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5866 - 5866. [Full Text] [PDF]
|
|
|
|
|
|
Q. Wang, P. Nicholson, M. Alen, and S. Cheng Authors' Response: Relationship of Sex Hormones to Bone Geometric Properties and Mineral Density in Early Pubertal Girls: Use of Correlation Analyses J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5866 - 5866. [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 |
