Determinants of serum sclerostin in healthy pre- and postmenopausal women

Authors

  • Mohammed-Salleh M Ardawi,

    Corresponding author
    1. Center of Excellence for Osteoporosis Research, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    2. Department of Clinical Biochemistry, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    • PO Box No. 20724, Jeddah 21465, Saudi Arabia.
    Search for more papers by this author
  • Hanan A Al-Kadi,

    1. Center of Excellence for Osteoporosis Research, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    2. Department of Medical Physiology, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    Search for more papers by this author
  • Abdulrahim A Rouzi,

    1. Center of Excellence for Osteoporosis Research, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    2. Department of Obstetrics & Gynecology, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    Search for more papers by this author
  • Mohammed H Qari

    1. Center of Excellence for Osteoporosis Research, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    2. Department of Hematology, Faculty of Medicine and King Abdulaziz University Hospital, King Abdulaziz University, Jeddah, Saudi Arabia
    Search for more papers by this author

Abstract

Sclerostin is a secreted Wnt antagonist produced almost exclusively by osteocytes that regulates bone mass. However, there is currently limited information on the determinants of sclerostin in a large population-based study. The main objectives of the present study were to: (1) establish reference normative interval values for serum sclerostin in randomly selected healthy premenopausal women; (2) study the changes in serum sclerostin in relation to age in premenopausal and postmenopausal women and the factors that may influence bone turnover; and (3) determine the effect of menopausal status on serum sclerostin. A total of 1803 women were studied (including [n = 1235] premenopausal, and [n = 568] postmenopausal women, respectively, aged 20 to 79 years). A total of 443 healthy premenopausal women (aged 35 to 45 years) were used to establish reference normative intervals for serum sclerostin. All women studied were medically examined and had their bone mineral density values obtained for the lumbar spine (L1–L4) and femoral neck according to a detailed inclusion criteria. In all women, values of serum sclerostin increased with increasing age up to the age of 45 years, and remained increased in postmenopausal women. Significant increases were evident in serum sclerostin in postmenopausal women with increasing years since menopause. Using stepwise multiple linear regression analysis, several variables were identified as determinants of serum sclerostin, including age, parathyroid hormone, estradiol (E2), and follicle-stimulating hormone (FSH) for premenopausal women; age, FSH, and E2 for postmenopausal women; and age, serum osteocalcin, FSH, and E2 in the entire sample studied. Further studies are needed to establish the potential role of this increase in mediating the known age-related impairment in bone formation. © 2011 American Society for Bone and Mineral Research

Introduction

Based on the work of the past decade, it is now well established that the Wnt/β-catenin signaling pathway is a major regulator of bone mass.1, 2 Activation of this pathway results in increased proliferation and differentiation of osteoprogenitor cells and reduced apoptosis of mature osteoblasts, favoring higher bone formation and increased bone density.3–5 In addition, the Wnt signaling may inhibit osteoclastogenesis.6 Activation of the Wnt/β-catenin canonical signaling pathway in osteoblasts is mediated via binding of any of multiple Wnt ligands to a seven-transmembrane domain-spanning frizzled receptor and either of two coreceptors, namely, low-density lipoprotein receptor-related proteins (LRPs) 5 or 6.7

Sclerostin is a glycoprotein secreted almost exclusively by osteocytes and to a lesser extent other cell types (kidney, vascular),1 which travels through osteocytic canaliculi to the bone surface and binds to LRP-5 and LRP-6 to inhibit the Wnt/β-catenin canonical signaling pathway,8–10 thereby decreasing osteoblastogenesis and bone formation. In humans, the importance of sclerostin is highlighted by two genetic disorders associated with significant progressive increases in bone mass,11–14 namely: sclerosteosis (caused by loss-of-function mutation in the gene encoding sclerostin, resulting in an improperly spliced SOST-mRNA)11–12; and Van Buchem's disease (caused by a deletion of an enhancer element that is normally downstream of the SOST-gene),13, 14 respectively. Furthermore, SOST-null mice have a high-bone-mass phenotype,15 and over-expression of normal human SOST-alleles in mice causes osteopenia.16 These observations suggested that inhibition of sclerostin may have therapeutic potential for the treatment of low-bone-mass disorders. Indeed, recent studies showed that using anti-sclerostin neutralizing antibodies in rats17 and primates18 increased bone density and bone strength. Similarly, in rodent models of fracture healing, sclerostin antibodies treatment resulted in increased callus density and bone strength at fracture sites and accelerated bone repair.19 This increased bone formation and bone mass by sclerostin antibody was not blunted in ovariectomized rats pretreated with alendronate.20 More recently, Padhi et al. showed that treatment of postmenopausal women with anti-sclerostin antibody resulted in dose-dependent increased bone formation markers.21

There are very limited reports in the literature on the serum levels of sclerostin in humans: Mirza et al. showed elevated sclerostin levels in 20 postmenopausal women compared with an equal number of premenopausal women.22 Modder et al. reported similar results in 152 postmenopausal women compared with 123 premenopausal women and showed a sex difference whereby men exhibited higher sclerostin levels than corresponding women.23 Moreover, Gaudio et al. showed increased circulating sclerostin levels in 40 long-term immobilized patients.24 Given the availability of a commercially validated immunoassay for sclerostin and no available information on a large-population studies, the main objectives of the present study are to: (1) establish age-specific normative interval values of serum sclerostin in randomly selected premenopausal women; (2) study the changes in sclerostin in relation to age, bone mineral density (BMD), and bone turnover markers (BTMs) (namely: serum osteocalcin [s-OC]; serum procollagen type 1 N-terminal propeptide [s-PINP]; serum cross-linked C-terminal telopeptide of type 1 collagen [s-CTX]; and urinary N-telopeptide of type 1 collagen [u-NTX)]) in pre- and postmenopausal women; and factors reported to influence bone turnover; and (3) determine the effect of menopausal status on serum sclerostin.

Subjects and Methods

Study Design

Over a period of 50 months (October 2005 to November 2009), a total of 5850 Saudi women were prospectively recruited at random during a health survey from 40 primary health care centers (PHCCs) scattered around the city of Jeddah (divided into 7 geographical areas) to ensure that the average health status of the studied group reflected a randomly selected adult population. The sample size was calculated using the sample-size determination option in Epi-Info Statistical Package (version 6) (USD, Stone Mountain, GA, USA). A description of the study design and rationale has been reported previously.25 Women who agreed to participate in the study were asked to visit a special clinic at the CEOR, King Abdulaziz University, Jeddah, Saudi Arabia, to be enrolled in the study. Age, body weight, height, body mass index (BMI) calculated as the weight (kg) divided by height (m) squared (kg/m2), and waist-to-hip ratio (WHR) were recorded. All women were medically examined and interviewed using a locally developed and validated standardized questionnaire to collect information on lifestyle, smoking habits, and level of physical activity in leisure time; coffee and tea consumption and the use of vitamins and medications. Women with chronic diseases, including osteoarthritis or established osteoporosis; with evident endocrine disorders; on any form of drug therapy with possible effect on bone metabolism (eg selective estrogen receptor modulators, calcitonin, phytoestrogens, glucocorticoids, anticonvulsants thyroid hormones and/or estrogen therapy); or with cancer were excluded from the final analysis. Women who were pregnant or lactating, taking oral contraceptives, vitamin-D deficient with serum 25-hydroxy-vitamin D [25(OH)D] levels being < 50 nmol/L,26 smokers, or who reported a recent fracture (within two years) were also excluded. In addition, all participants in the present study showed: (1) normal blood counts; (2) normal values for renal creatinine (serum creatinine in women < 105 µmol/L); and (3) normal hepatic function tests (serum aspartate aminotransferase < 30 U/L; alanine aminotransferase < 30 U/L; alkaline phosphatase between 80 U/L and 280 U/L; and gamma-glutamyl transferase < 60 U/L). Premenopausal women were included if they were between 20 and 45 years of age, and had serum FSH levels < 15 mIU/L, and were cycling regularly. Postmenopausal women were included if they had experienced their last menstrual cycle at least one year before and were not taking any medications that are known to affect bone metabolism. The levels of FSH in postmenopausal women were > 15 mIU/L. Accordingly, a total of 1803 healthy Saudi women (age range 20 to 79 years) (premenopausal, n = 1235; postmenopausal, n = 568) living in the Jeddah area, participated in the present study and were included in the final analysis (see Fig. 1). Demographic characteristics, hormones, and BTMs together with BMD values for these women are presented in Table 1.

Figure 1.

Subjects screening and recruitment according to inclusion and exclusion criteria. A total of 5850 women originally recruited, of which only 1803 were included in the final analysis and 4047 women were excluded accordingly.

Table 1. Demographic, Anthropometric Characteristics, Various Hormones, 25(OH)D, BTMs, Minerals, and BMD Values of the Studied Healthy Pre- and Postmenopausal Women
VariablesMenopausal StatusANOVA (p-value)
Premenopausal (n = 1235)Postmenopausal (n = 568)
  1. BMI = body mass index; WHR = waist-to-hip ratio; FSH = follicle-stimulating hormone; LH = luteinizing hormone; E2 = estradiol; intact-PTH = intact parathyroid hormone; 25(OH)D = 25-hydroxyvitamin D; s-OC = serum osteocalcin; s-PINP = serum procollagen type-1 N-terminal propeptide; s-CTX = serum cross-linked C-terminal telopeptide of type 1 collagen; u-NTX = urinary N-telopeptide of type 1 collagen; Creat = creatinine; Ca = calcium; PO4 = phosphate; Mg = magnesium; BMD = bone mineral density.

  2. Mean ± SD, or as percentage.

Age (yrs)33.83 ± 8.4162.38 ± 8.34< 0.0001
Age at menarche (yrs)12.46 ± 3.8214.20 ± 4.16< 0.051
Age at menopause (yrs)47.51 ± 6.18
Years since menopause (yrs)15.62 ± 8.75
Parity3.28 ± 2.614.18 ± 3.23< 0.0001
BMI (kg/m2)27.97 ± 5.3031.63 ± 6.07< 0.0001
 Lean (BMI < 25)30.4%12.9%< 0.001
 Overweight (BMI ≥ 25 – < 30)37.6%29.0%< 0.001
 Obese (BMI ≥ 30)32.0%58.1%< 0.0001
WHR0.817 ± 0.0820.897 ± 0.082< 0.001
FSH (IU/L)4.57 ± 2.4828.72 ± 7.47< 0.0001
LH (IU/L)8.07 ± 2.1025.59 ± 7.34< 0.0001
E2 (pmol/L)369.2 ± 87.3135.1 ± 56.6< 0.0001
Intact-PTH (pmol/L)2.72 ± 0.764.55 ± 0.96< 0.0001
25(OH)D (nmol/L)66.47 ± 13.0255.74 ± 5.32< 0.0001
s-OC (ng/mL)13.30 ± 4.5117.18 ± 3.61< 0.0001
s-PINP (µg/L)51.33 ± 22.7347.38 ± 5.52< 0.0001
s-CTX (pg/ml)279.5 ± 67.8338.9 ± 43.7< 0.0001
u-NTX (nmol/mmol creat)34.49 ± 7.8157.69 ± 8.36< 0.0001
Serum Ca (mmol/L)2.38 ± 0.132.38 ± 0.090.145
Serum PO4 (mmol/L)1.25 ± 0.161.27 ± 0.16< 0.015
Serum Mg (mmol/L)0.891 ± 0.090.791 ± 0.16< 0.031
BMD (g/cm2)
 Lumbar spine (L1–L4)1.079 ± 0.2790.991 ± 0.074< 0.001
 Neck femur1.107 ± 0.1440.928 ± 0.075< 0.001

Bone Mineral Densitometry Measurements

BMD (g/cm2) was determined for the anteroposterior lumbar spine (L1–L4) and mean of proximal right and left femur (total and subregions) by dual-energy X-ray absorptiometry (DXA), using LUNAR Prodigy Model (Lunar Corp., Madison, WI, USA) according to standard protocol. Quality-control procedures were carried out in accordance with the manufacturer's recommendations as described previously.25 BMD values were classified according to WHO criteria: a T-score between −1 and −2.5 is indicative of osteopenia, whereas a T-score of −2.5 and below reflects osteoporosis, and a T-score of −1.0 and above is considered normal.27

Specimen Collection

Venous blood samples were collected in the morning under standardized conditions. Second-void morning urine samples were collected on the same day of blood sampling. Serum and urine samples were stored at −85°C within 30 minutes after centrifugation at 2500 g for 10 minutes. All samples were collected between 9:00 and 11:00 a.m., after an overnight fast. The samples were stored until analyzed for the determinations of serum sclerostin, s-OC, s-PINP, s-CTX, and u-NTX. Serum sclerostin and all biochemical BTMs together with other hormones, and other analytes were carried out at the same time-point, according to the manufacturer's instructions.

Measurements of Sclerostin and Biochemical BTMs

Serum sclerostin levels were measured on coded specimen by enzyme-linked immunosorbent assay (ELISA) supplied by Biomedica (Biomedica Gruppe, Biomedica Medizinprodukte GmbH & Co KG, Wien, Austria). This assay uses a polyclonal goat anti-human sclerostin antibody as a capture antibody and a biotin-labeled mouse monoclonal anti-sclerostin antibody for detection. The intra-assay and interassay coefficients of variations (CVs) were 4.5% and 5.6%, respectively. Further validation studies were performed for this assay. Linearity was assessed serially diluting serum samples with sample diluents (dilutions 1:2, 1:4, and 1:8) and comparing observed values with expected values (observed recoveries ranged from 102% to 108%) (data not shown). Also, recovery of spiked standards was tested by adding different concentrations of human recombinant sclerostin (three different concentrations) into eight different human serum samples presenting with various levels of endogenous sclerostin. Spiked recovery ranged from 92% to 108% (data not shown). Serum OC was measured using electrochemiluminescence immunoassay (ECLIA) Elecsys autoanalyzer (Roche Diagnostics GmbH, D-68298 Mannheim, Germany). The intra-assay and interassay CVs were 1.8% and 1.2%, respectively. Serum PINP was measured using ECLIA Elecsys autoanalyzer [Roche Diagnostics GmbH, D-68298 Mannheim, Germany]. The intra-assay and interassay CVs were 2.3% and 2.1%, respectively with a sensitivity of < 5 µg/L. Serum CTX was measured by Elecsys β-CrossLaps assay using ECLIA Elecsys autoanalyzer. The intra-assay and interassay CVs were 4%. Urinary NTX was determined by utilizing a competitive-inhibition ELISA using Osteomark kits (Ostex International, Seattle, USA). The intra-assay and interassay CVs were 7.8% and 4.5%, respectively. In all manual assays, a standard curve was run simultaneously with the samples, and the curve was fitted for serum sclerostin and u-NTX; the results were calculated using MultCalc (Wallac, Turku, Finland). The results of all u-NTX were corrected for creatinine (creat) concentration (mmol creat/l).

Measurements of Hormones and Other Analytes

Serum FSH, luteinizing hormone (LH), E2, and intact-PTH were measured by commercially available immunoassays using Elecsys autoanalyzer (Roche Diagnostics GmbH, D-68298 Mannheim, Germany). The intra-assay and interassay CVs were less than 4.0%. Serum 25(OH)D was measured by direct competitive chemiluminescence immunoassay using LIASON autoanalyzer (DiaSorin Inc, Stillwater, MN, USA). The intra-assay and interassay CVs were 7.8% and 3.8%, respectively. Serum creatinine (creat), calcium (Ca), phosphate (PO4), magnesium (Mg), and other biochemical analytes were measured by kits and reagents supplied by Ortho-Clinical Diagnostics, USA using Vitros 250 Chemistry System Autoanalyzer (Ortho-Clinical Diagnostics–Johnson & Johnson Co., USA).

Statistical Analysis

Results are presented as means or Geometric means (± SD) and categorical variables are expressed as frequencies as appropriate. Data were analyzed using SPSS-Statistical Package (version 15.0 for Windows Smart Viewer) supplied by SPSS Inc. 2000, Mapinfo Corp. (Troy, NY, USA). Results that were not normally distributed were log-transformed before analysis. A 95% reference intervals for serum sclerostin together with that for s-OC, s-PINP, s-CTX, and u-NTX measured were calculated for each as mean ± 1.96 SD, and the confidence intervals (CI) for the lower and upper bounds of the reference intervals were computed as boundary ± 1.96 standard error (SE). Initially locally weighted scatterplot smoothing (LOWESS) curve analysis was used to establish the trend of serum sclerostin and measured biochemical BTMs data among the various age groups (five-year bands). The LOWESS curve analysis is used to fit a curve to data without selecting a model and cannot be used to obtain the best-fit values; however, it is very useful in providing a simple approach to produce smoothing of the data.28 Thus, in the present analysis of serum sclerostin and the biochemical BTMs measured, the LOWESS curve analysis was applied to establish the general trend and distribution of the scatter of BTMs data among the various age groups. This approach was followed by segmental linear regression analysis to fit two regression lines to serum sclerostin and BTMs data by using suitable critical cut-off point. Accordingly, the gradient of each regression line generated was used to compare differences between the lines. Further, ANOVA testing (independent two-samples t-test) was applied to examine whether serum sclerostin and measured BTMs values differ among women aged 30 to 34 or younger versus women who were 35- to 45-years-old. Based on the above two approaches, women aged 35- to 45-years-old were used for further analysis to examine other determinants of serum sclerostin and measured BTMs values to remove the confounding effects of age. All women were stratified by 10-years age and also by quartiles of serum FSH, LH, PTH, and E2 to assess variations in serum sclerostin among these variables using one-way ANOVA testing. Associations between continuous variables were examined by Pearson's correlation coefficient. ANOVA was used to examine differences among the groups for different variables, and the Bonferroni criterion was used when significance tests were made. Independent relationships between serum sclerostin and that of measured biochemical BTMs and other variables were assessed by multiple regression and partial correlation analysis.

Results

Table 1 shows the basic demographic, anthropometric characteristics, various hormones, BTMs, and BMD values of the studied premenopausal (n = 1235) and postmenopausal (n = 568) women with a mean age of 42.82 ± 15.69 years. Of the women studied, 24.9%, 34.9%, and 40.2% were lean (BMI < 25.0 kg/m2); overweight (BMI ≥ 25.0 to < 30.0 kg/m2); and obese (BMI ≥ 30 kg/m2), respectively. Postmenopausal women exhibited higher parity (p < 0.0001), BMI (p < 0.0001), and WHR (p < 0.0001) values compared with premenopausal women (Table 1). Postmenopausal women exhibited higher levels of serum FSH (p < 0.0001), LH (p < 0.0001), and intact-PTH (p < 0.0001), but lower E2 (p < 0.0001) compared with premenopausal women. All premenopausal women exhibited normal BMD values for lumbar spine (L1–L4) and femoral neck with lower values for postmenopausal women. However, postmenopausal women showed significantly higher s-OC, s-PINP, s-CTX, and u-NTX values and significantly lower s-PINP and serum 25(OH)D levels compared with premenopausal women (Table 1).

Using LOWESS curve analysis, the sclerostin values exhibited a general trend showing that women under the age of 35 years showed decreased sclerostin values: accordingly, segmental linear regression analysis was performed using the critical values of 35 (Fig. 2). Supplemental Table 1 shows two types of analysis: (1) calculation of gradients of segmental linear regression for 95% CIs for women in the age group 30 to 34 years compared with that of 35 to 45 years; and (2) independent t-testing between the age group 30 to 34 years compared with that of 35 to 39 years, 40 to 45 years, and 35 to 45 years. Using the gradient of segmental regression line for women aged 30 to 34, it was significantly different from zero compared with the same values obtained for women in the age group 35 to 45 years (which was not significantly different from zero), as indicated by the CI values (Supplemental Table 1). The ANOVA testing showed significantly lower values for serum sclerostin for the age group 30 to 34 years compared with that observed in the age group 35 to 45 years (see Supplemental Table 1). Similar trends with higher values among the age group 30 to 34 years compared with that among the age group 35 to 45 years were obtained for the BTMs: s-OC, s-PINP, s-CTX, and u-NTX. Accordingly, a total of 443 premenopausal women (age 35 to 45 years) were used to establish sclerostin reference normative values, and their demographic characteristics and other variables are given in Table 2.

Figure 2.

Correlation between age (years) and serum sclerostin levels among all women (n = 1803) studied. Significant positive correlation was observed (R2 = 0.828; p < 0.0001).

Table 2. Anthropometric Characteristics and Other Variables in Premenopausal Women Used for Reference Intervals of Biochemical Bone Turnover Markers
VariablesnResults
  1. n = number; BMI = body mass index; WHR = waist-to-hip ratio; 25(OH)D = 25-hydroxyvitamin D.

  2. Mean ± SD, or as percentage.

Age (years)44339.83 ± 3.25
BMI (kg/m2)44329.10 ± 4.99
WHR4430.830 ± 0.087
Parity
 0 (nullipara)8819.9%
 1–27316.5%
 3–419243.3%
 ≥ 59020.3%
Physical activity
 Exercise: Yes25256.9%
 No19143.1%
 Regular: Yes7429.4%
 No17870.6%
Total dietary: Ca intake (mg/day)
 < 60029065.5%
 > 600–100010122.8%
 > 10005211.7%
Total caffeine intake (mg/day)
 < 20028564.3%
 200–4009120.5%
 > 4006715.2%
Vitamin D intake (IU/day)
 None22550.8%
 < 40017840.2%
 ≥ 400409.0%
25(OH)D (nmol/L)44366.08 ± 13.55

In the present study, the levels of serum sclerostin were stable among the age group 35 to 45 years, and the reference normative intervals of serum sclerostin are given in Supplemental Table 2, including values for geometric mean; log10 mean ± SD; 95% reference interval; and lower and upper limits of 95% CIs. The geometric mean and SD values for serum sclerostin for the 10-year age grouping of women studied are presented in Table 3, together with relevant BTMs, compared with that of the reference normative intervals group. Serum sclerostin showed a general linear increase with age (Fig. 3), with serum sclerostin showing a steep increase until the age of 35, and remaining stable in those between 35 and 45 years of age; then, serum sclerostin exhibited an increase with increasing age. Stratifying women according to BMI (kg/m2) showed that obese women (≥ 30) had significantly higher serum sclerostin values by 26.4% over that for lean (< 25) women (p < 0.049) (data not shown). However, stratifying women by quartiles of WHR showed no significant differences among the groups (data not shown). Supplemental Table 3 shows the mean values of serum sclerostin in relation to years since menopause (YSM) in postmenopausal women studied compared with the premenopausal values. Significant increases were observed for serum sclerostin among postmenopausal women according to YSM periods. Similar trends were evident for the BTMs examined (Supplemental Table 3).

Table 3. Geometric Mean and 95% Confidence Interval (95% CI) Age-Specific Values of Sclerostin and Bone Turnover Markers Among Women Studied
Age groupnSclerostin (pmol/L)s-OC (ng/mL)s-PINP (µg/L)s-CTX (pg/mL)u-NTX (nmol/mmol creat)
  • CI = confidence interval; n = number of women; s-OC = serum osteocalcin; s-PINP = serum procollagen type-1 N-terminal propeptide; s-CTX = serum cross-linked C-terminal telopeptide of type 1 collagen; u-NTX = urinary N-telopeptide of type 1 collagen; Creat = creatinine.

  • All values are presented as geometric means with (95% CI).

  • a

    Statistically significant from reference values.

  • b

    Reference values were calculated for the age group 35 to 45 years as geometric mean with (95% CI).

20–2943516.87a (16.54–17.20)17.27a (17.11–17.43)78.79a (77.49–80.09)360.9a (358.3–363.5)29.05 (28.62–29.48)
30–3943525.04 (24.68–25.40)11.18a (10.81–11.55)33.90 (33.28–34.52)200.9a (197.4– 204.4)18.74a (18.31–19.17)
40–4936530.80a (30.15–31.45)9.66 (9.35–9.97)35.30 (34.63–35.97)198.0 (195.1–200.9)25.09 (23.75–26.43)
50–5924640.29a (39.41–41.17)14.45a (14.06–14.84)46.63a (45.98–47.27)334.6a (330.4–338.8)51.78a (51.23–52.34)
60–6919651.01a (49.48–52.54)19.34a (19.08–19.60)49.37a (48.67–50.07)373.2a (369.5–376.9)63.86a (62.74–64.98)
70–7912661.95a (59.98–63.91)19.16a (18.64–19.68)45.75a (44.66–46.84)294.4a (287.4–301.4)59.65a (58.45–60.85)
Total (20–79)180331.72 (31.03–32.41)14.24 (14.02–14.46)49.31 (48.40–50.22)282.4 (278.6–286.2)34.79 (33.95–35.63)
Reference valueb (35–45)44326.68 (26.39–26.97)9.09 (2.54–15.20)31.8 (21.90–32.31)181.7 (179.3–230.2)30.84 (19.35–46.32)
Figure 3.

Scatterplots showing segmental regression analysis for age and serum sclerostin (A), together with that for serum osteocalcin (B); serum procollagen type 1 N-terminal propeptide (s-PINP) (C); serum cross-linked C-terminal telopeptide of type 1 collagen (s-CTX) (D); urinary N-telopeptide of type 1 collagen (u-NTX) (E), respectively, among 668 women aged 30 to 45 years.

In premenopausal women, serum sclerostin showed significant positive correlations with the following variables: age (r = 0.879; p < 0.0001); BMI (r = 0.223; p < 0.0001); parity (r = 0.368; p < 0.0001); serum FSH (r = 0.595; p < 0.0001); and serum LH (r = 0.538; p < 0.001), respectively. Negative correlations were evident between serum sclerostin and the following variables: s-OC (r = −0.671; p < 0.0001); s-PINP (r = −0.713; p < 0.0001); s-CTX (r = −0.767; p < 0.0001); u-NTX (r = −0.432; p < 0.0001); serum E2 (r = −0.519; p < 0.0001); serum intact-PTH (r = −0.123; p < 0.001); serum 25(OH)D (r = −0.072; p < 0.011); BMD lumbar spine (L1–L4) (r = −0.168; p < 0.0001); and BMD neck femur (r = −0.545; P < 0.0001), respectively. In premenopausal women age- and BMI-adjusted correlations persisted for s-OC (r = −0.095; p < 0.014); s-CTX (r = −0.289; p < 0.0001); u-NTX (r = −0.359; p < 0.0001); intact-PTH (r = −0.111; p < 0.01); FSH (r = −0.122; p < 0.002); and E2 (r = 0.091; p < 0.020), respectively. In postmenopausal women, serum sclerostin showed significantly positive correlations with age (r = 0.696; p < 0.0001); BMI (r = 0.120; p < 0.039); WHR (r =0.131; p < 0.042); parity (r = 0.153; p < 0.0001); YSM (r = 0.672; p < 0.0001) and s-OC (r = 0.479; p < 0.0001), respectively. However, significant negative correlations were observed between serum sclerostin and various variables including: serum FSH (r = −0.622; p < 0.0001); LH (r = −0.646; p < 0.0001); serum E2 (r = −0.256; p < 0.001); serum intact-PTH (r = −0.172; p < 0.004); serum 25(OH)D (r = − 0.303; p < 0.0001); BMD lumbar spine (L1–L4) (r = −0.354; p < 0.0001); and BMD neck femur (r = −0.372; p < 0.0001), respectively (data not shown). In postmenopausal women, age-adjusted and BMI-adjusted correlations persisted for FSH (r = −0.101; p < 0.018); and E2 (r = −0.122; p < 0.002), respectively. Moreover, when combining both pre- and postmenopausal women (n = 1803), serum sclerostin showed significant correlations with various variables examined, including: age (r = 0.893; p < 0.0001); BMI (r = 0.271; p < 0.0001); WHR (r = 0.146; p < 0.012); parity (r = 0.299; p < 0.0001); YSM (r = 0.731; p < 0.0001); FSH (r = 0.587; p < 0.0001); LH (r = 0.520; p < 0.0001); intact-PTH (r = 0.679; p < 0.0001); and s-OC (r = 0.076; P <0.001), respectively. Significant negative correlations were evident between serum sclerostin and other variables: s-PINP (r = −0.441; p < 0.0001); s-CTX (r = −0.080; p < 0.001); u-NTX (r = −0.121; p < 0.001); serum E2 (r = −0.751; p < 0.0001); serum 25(OH)D (r = −0.350; p < 0.0001); serum calcium (r = −0.055; p < 0.018); BMD lumbar spine (L1–L4) (r = −0.670; p < 0.0001) and BMD neck femur (r = −0.312; p < 0.0001), respectively (data not shown). In all women, age-adjusted and BMI-adjusted correlations persisted for s-OC (r = 0.165; p < 0.0001); s-PINP (r = 0.115; p < 0.0001); s-CTX (r = 0.087; p < 0.0001); u-NTX (r = 0.123; p < 0.0001); FSH (r = 0.187; p < 0.001); LH (r 0.112; p < 0.001); and E2 (r = 0.109; p < 0.0001), respectively.

Serum sclerostin is identified as an independent variable in the multiple linear stepwise regression analysis, whereas variables showing a significant correlations (age, BMI, parity, FSH, LH, E2, intact-PTH, 25[OH]D, s-OC, s-PINP, s-CTX, u-NTX, BMD lumbar spine [L1–L4] and BMD neck femur) are identified as dependent variables; the best predictors in all women were age, s-OC, FSH, and E2, accounting for about 84% of the variance in serum sclerostin (Supplemental Table 4). We also analyzed predictors of serum sclerostin separately for pre- and postmenopausal women. In premenopausal women, age, intact-PTH, E2, and FSH were the best predictors of serum sclerostin, accounting for about 73% of the variance. Whereas, in postmenopausal women only age, FSH, and E2 were found to be predictors of serum sclerostin, accounting for about 62% of the variance in serum sclerostin (Supplemental Table 4).

Discussion

This study provides the first and the largest population-based study of pre- and postmenopausal women using a commercially available validated immunoassay for serum sclerostin. We have shown that serum sclerostin levels are stable between the ages of 35 and 45 in this group of premenopausal women. In the present study the reference normative intervals for serum sclerostin were identified to be in the age group of 35 to 45 years. Women under the age of 30 years were excluded from the reference normative intervals group: this lowered the possibility of including women with lowered serum sclerostin as a result of skeletal immaturity, lowered osteocytes activity, or both. Further, women over 45 years were also excluded, because detailed analysis showed that serum sclerostin increased beyond this cut-off age. Similar observations were evident for the analysis of the measured BTMs confirming previous studies.25, 29 The study shows that serum sclerostin levels increased significantly with age in women. Thus, serum sclerostin levels increased over life by an average of 3.7-fold (p < 0.0001): this suggests that increased sclerostin production by osteocytes may be involved in the age-related impairment of bone formation. Moreover, given that sclerostin is produced almost exclusively by osteocytes,1, 30, 31 this observation is concurrent with the expected changes in bone mass associated with aging. However, it is possible that the increase with age of serum sclerostin was related, at least partly, to decreased clearance of the protein, but the increased serum sclerostin with age was minimally affected after adjustment for the expected age-related decreases in creatinine clearance. The large number of women studied ensured a small uncertainty around the upper- and lower-reference limits and thus, providing reliable reference ranges for serum sclerostin in studied women. Women using hormonal contraceptives were excluded from the study, because several previous studies showed that bone remodeling, bone turnover, or both are markedly decreased in women receiving combined oral contraception, particularly women over 35 years of age.25, 29

The mean reference normative interval values obtained for serum sclerostin in the present study were similar to that given by the manufacturer and the ranges for the values reported here are higher than that reported for premenopausal women (n = 123) by Modder et al. (15.1 ± 1.2 pmol/L) using the same immunoassay system.23 Such differences are most likely because of differences in the selection criteria of women studied as well as the mean age of the group studied by Modder et al., although no information was given on the mean of the age of women studied.23 It is possible, however, that environmental or lifestyle and/or genetic influences on bone mass, and consequently serum sclerostin, not accounted for in the present study should be taken into consideration to account for any differences in serum sclerostin levels among the women studied or to be studied in the future.

In the present study, the pattern of age-related changes in serum sclerostin among women is shown by the increase in bone turnover at menopause. Postmenopausal women showed significantly higher serum sclerostin levels (48.79 ± 12.68 pmol/L; n = 568) than premenopausal women (26.90 ± 9.71 pmol/L; n = 1235). The higher serum sclerostin levels in postmenopausal women maybe a cause, an effect, or both of the increased bone turnover occurring in the postmenopausal state. These observations are consistent with previously reported studies by Modder et al.23 and by Polyzos et al.32 Similar changes were obtained for the BTMs studied confirming previous studies in different populations.25, 29 However, it is uncertain whether the onset of the changes in serum sclerostin takes place early in the menopausal transition or before the cessation of menstruation. In addition, the results of the present study showed that serum sclerostin continued to increase with increasing YSM and remained elevated in the elderly (Supplemental Table 3). This is possibly related to the expected increased bone turnover (and possibly osteocytic activities) associated with estrogen imbalance,33 with the rate of bone turnover slowing down after 4 to 8 years for most women,34 whereas serum sclerostin continued to increase with increasing YSM in the present study. To our knowledge, there is no information in the literature on the changes of serum sclerostin in relation to YSM.

Previous studies have shown that BMI is an important determinant of bone turnover and BMI plays a major role in bone mass in women regardless of gonadal status and is a key risk factor for osteoporotic fractures.35, 36 In the present study, women with higher BMI exhibited significantly higher serum sclerostin: thus serum sclerostin was 26.4% higher in obese versus lean women (p < 0.0001). Conversely, women with higher BMI exhibited significantly lower s-CTX and u-NTX values and a trend toward lower s-OC and s-PINP values (data not shown). Based on the previously known effects of sclerostin in inhibiting bone formation,1, 30 bone-formation markers would be expected to be inversely related to serum sclerostin levels. Indeed, in the present study, premenopausal women exhibited significantly negative correlations between serum sclerostin and BTMs examined (namely, s-OC, s-PINP, s-CTX, and u-NTX); however, such correlations disappeared after adjustment for age and BMI. Conversely, in postmenopausal women, significant positive correlations with s-OC, but negative weak correlations with bone resorption markers (namely, s-CTX and u-NTX) with serum sclerostin were observed, respectively; however, none of these correlations persisted after adjustment for age and BMI. These findings are consistent with that obtained for both pre- and postmenopausal women reported previously22 and contrast with that reported by Modder et al. for postmenopausal women studied.23 However, these results indicate that in healthy women serum sclerostin may provide information on bone metabolism, ie, osteocyte functional activity, which is not reflected and/or captured by measured BTMs.

In the present study, significant negative correlations were evident between serum sclerostin and BMD for both lumbar spine (L1–L4) and neck femur for both pre- and postmenopausal women, which persisted when the entire population was examined. However, these correlations disappeared after adjustment for age and BMI, in contrast to what was reported previously for elderly women, but concurrent with what was reported in premenopausal women.23

In the present study, serum FSH and LH levels in postmenopausal women are 6.3 and 3.2 times greater, respectively, than those in premenopausal women (Table 1). Several studies showed that serum FSH levels are not only related to age and menopause but also to ethnic backgrounds.37–39 Our results showed that a decline in ovarian function, as indicated indirectly by menopausal rise in serum FSH levels, is associated with higher serum sclerostin and bone loss. Furthermore, serum FSH levels in the studied women showed significantly positive correlations with serum sclerostin. In addition, serum FSH exhibited positive correlations with the BTMs examined (data not shown), suggesting that bone turnover speed becomes more rapid in women with menopausal rise in FSH; such finding for BTMs is similar to those of other studies.40 In contrast, however, Vural et al.41 found no correlation between serum FSH and that of s-OC. Our results also show that serum sclerostin in quartile groups with high serum FSH (Q3 and Q4) are higher than those in the quartile groups with lower FSH levels (Q1 and Q2) (data not shown). Moreover, we have examined and compared the effects of serum FSH and LH on serum sclerostin and BTMs using multiple linear regression analysis, and we have found that the association between serum FSH and serum sclerostin and BTMs is greater than that with serum LH. These phenomena indicate that FSH in circulation is more significantly associated with osteoblasts and possibly osteocytes than with osteoclasts, results that contrast from those of previous in vitro cellular and experimental animal studies.42, 43 However, the recent work by Drake et al. showed that FSH did not directly regulate bone resorption in postmenopausal women after suppression of their FSH secretion.44 The significant association between serum sclerostin, FSH, and LH are considered to be the first reported in the literature, and further studies in other populations should be conducted to further our understanding of this association, particularly at the cellular levels (ie, osteocytes).

Estradiol deficiency is considered to be partly responsible for the rapid bone loss after menopause.43 In the present study, serum E2 showed significant negative correlations with serum sclerostin in both pre- and postmenopausal women, which persisted after adjustment for age and BMI. Further, serum E2 was significantly associated with serum sclerostin in both pre- and postmenopausal women. Modder et al., observed among postmenopausal women that serum sclerostin levels were significantly lower in women on estrogen therapy (ET) compared with women not on ET.45 These observations were concurrent with the recent observations from the same group showing that ET of postmenopausal women for 4 weeks resulted in a 27% decrease in serum sclerostin levels.47 Such observations are consistent with the correlation analysis of the present study and consistent with previous findings by both Mirza et al.22 and Modder et al.,45 but contrast with that reported more recently by Modder et al. for both pre- and postmenopausal women.23

Parathyroid hormone has been shown to decrease sclerostin transcription in vitro,46, 47 and continuous or intermittent chronic administration of PTH to rodents is associated with decreased SOST-mRNA and sclerostin expression in osteocytes.48, 49 Moreover, van Lierop et al. showed significantly lowered serum sclerostin in patients with primary hyperparathyroidism compared with euparathyroid controls.48 The present study showed an inverse relationship between serum sclerostin and intact-PTH levels: such a relationship was stronger for postmenopausal (r = −0.172; p < 0.004) compared with that of premenopausal (r = −0.111; p < 0.01) women, respectively. Also, stratifying women into quartiles of serum intact-PTH showed that serum sclerostin was significantly lower in women with the highest quartile (Q4) of intact-PTH compared with the lowest quartile (Q1) (p < 0.001) (data not shown). Furthermore, PTH contributed significantly to serum sclerostin variation among the premenopausal women. Thus, our results are consistent with the observations of Mirza et al.22 for postmenopausal women but contrast with that observed by Modder et al. for both pre- and postmenopausal women.23 Also, our results are indirectly consistent with the recent report by Drake and colleague, on the observed significant reduction in circulating sclerostin levels after intermittent PTH 1-34 treatment of postmenopausal women.49 Interestingly, in this latter study, bone marrow plasma and peripheral serum sclerostin levels were significantly correlated (p < 0.0001), suggesting that the circulating levels may be a good index of local bone production. Taken all together, these observations including the results of the present study, lend further support to the hypothesis that, in humans and experimental animals, at least part of the anabolic effect of PTH on bone may be mediated via an inhibition of sclerostin production and/or stimulation of its clearance.

The present study has several strengths and limitations. The strengths of our study include its large sample size and acceptable sampling errors; its random selection of women from the local population and covering a wide range of ages, thus, avoiding sampling bias; and its very strict detailed inclusion criteria and comprehensive details of lifestyle characteristics and BMD measurements. In addition, another strength of our study is that minimizing pre-analytical variations as a result of possible circadian rhythm and/or food intake was ensured by standardized sampling time and that all women studied were fasting overnight. However, currently there is no published information on the effects of various pre-analytical factors (eg, variations related to food intake or fasting, seasonal changes, or circadian rhythm) on serum sclerostin; such information requires further study. The principal limitation of the present study was its cross-sectional design, and thus the causative nature of the associations between sclerostin and other variables cannot be established. Moreover, another limitation of our study is the single fasting measurement of serum sclerostin and other analytes (eg, PTH, E2). The accuracy of self-reported data concerning lifestyle practices may have been subject to report bias, as can be the case with such type of studies; we acknowledge that there may be unrecognized confounding. Women who participated in the present study may have chosen to take part because they are more aware of health issues and healthier than average. Also, women in the present study exhibited low dietary calcium intake (about 65.5% of women had daily calcium intake estimated at < 600 mg/day); that this will affect the observations of the present study cannot be ruled out, and further studies are needed in this regard. Furthermore, our results were based on premenopausal women living in Jeddah, Saudi Arabia, with quite rigorous inclusion criteria; thus, whether the findings reported here will be completely applicable to other populations remains to be determined in other populations. Finally, we recognize that there is currently no information on the metabolism or renal clearance of sclerostin and/or the stability of sclerostin or its degradation in circulation and/or upon storage at −85°C; further studies are needed in this regard.

In summary, our study represents the first randomly selected, large population-based assessment of circulating sclerostin levels and provides reference normative interval values among pre- and postmenopausal women over a wide age range using a well-validated immunoassay. The study protocol specified detailed inclusion criteria, which combined BMD measurement and through medical examination, allowed a well-defined and characterized study population to be examined. The results of the present study point to the need for future further studies examining the mechanisms for the age-related changes in serum sclerostin levels described in pre- and postmenopausal women, and also the contribution of sclerostin in mediating the well-established age-related decrease in bone formation in humans.1, 30 Serum FSH together with that of E2 showed significant associations with serum sclerostin in both pre- and postmenopausal women. The results have confirmed that the levels of serum sclerostin increase in women with increasing serum FSH and LH levels, and that FSH showed stronger association with serum sclerostin than LH, particularly in postmenopausal women. Serum sclerostin was poorly associated with BTMs (except for s-OC), and based on its relationship with measured BTMs in the present study indicate that serum sclerostin measurements provide additional information on bone metabolism that are not captured by currently available BTMs, including the activity of osteocytes in vivo and the modulation of their functions by mechanical loading and therapeutic intervention in osteoporosis and other bone diseases. However, measuring serum sclerostin and BTMs simultaneously may prove helpful in evaluating and/or monitoring the changes in the rate of bone turnover and ostecytic function caused by aging or menopause in women or in bone-diseased states in humans.

Disclosures

All the authors state that they have no conflicts of interest.

Acknowledgements

We are grateful to the Ministry of Higher Education for financial support to the Center of Excellence for Osteoporosis Research (CEOR) at King Abdulaziz University, Jeddah, Saudi Arabia. This study was supported by grants from the Ministry of Higher Education to the Center of Excellence for Osteoporosis Research (CEOR) at King Abdulaziz University (grants # CEOR/001-08 and CEOR/004-08), Jeddah, Saudi Arabia. It was approved by the Human Research Ethical Committee of the Center of Excellence for Osteoporosis Research, and the study protocol was in agreement with King Abdulaziz University Hospital (KAUH) ethical standards and the Helsinki Declaration of 1975, as revised in 1989. We thank all the subjects who participated in the study, and we thank all the staff and colleagues at CEOR, King Abdulaziz University Hospital, and the Primary Care Health Centers for their invaluable assistance during the execution of the study. Special thanks are due to Ms Veronica Orbacedo for her excellent help in preparing the manuscript. Professor MSM Ardawi was awarded grants CEOR/001-08 and CEOR/004-08 from Center of Excellence for Osteoporosis Research, Supported by MOHE, Saudi Arabia.

Authors' roles: All authors have contributed to the design of the study, analysis and interpretation of data, the writing of the article, and approval of the version to be published.

Ancillary