|
|
||||||||
CLINICAL STUDY |
1 Unit for Osteoporosis and Metabolic Bone Diseases and 2 Department of Public Health, Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium
(Correspondence should be addressed to P Crabbe, Department of Endocrinology, 9K12 I.E., Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium; Email: Patricia.Crabbe{at}UGent.be)
| Abstract |
|---|
|
|
|---|
Design: We performed cross-sectional (n = 270) and longitudinal (mean follow-up 3.4 years, n = 214) evaluations in elderly men.
Methods: Serum leptin, LEPR genotype, baseline bone mineral density (BMD), longitudinal BMD changes at the hip and forearm, and biochemical markers of bone turnover were determined.
Results: In cross-sectional analyses absolute fat mass was the index of body composition most strongly associated with leptin (r = 0.74; P < 0.001). LEPR genotypes and serum leptin were not associated. Serum bone-specific alkaline phosphatase (S-BAP) was associated with LEPR genotypes (P = 0.05) and urinary C-terminal telopeptides of type I collagen (U-CTX) were associated with leptin levels (P = 0.03), independently from age, fat mass and free estradiol. Baseline BMD at the hip and forearm was neither associated with leptin nor with LEPR genotypes. Prospectively assessed BMD loss was not associated with serum leptin at the hip, whereas BMD loss was positively associated with leptin at the forearm (P = 0.01), independently from age, fat mass and free estradiol. Longitudinal changes in hip or forearm BMD were not associated with LEPR genotypes.
Conclusion: The findings suggest a possible role for leptin as determinant of bone homeostasis in elderly men, but with only modest impact independently from body composition and free estradiol.
| Introduction |
|---|
|
|
|---|
Animal studies have indicated an anti-osteogenic action of leptin, mediated by the sympathetic nervous system, with inhibition of bone formation (3, 4) and stimulation of bone resorption (5, 6). However, others (79) have ascribed direct osteogenic effects and bone mineralization promoting actions to leptin. The apparently conflicting results in rodents, might be explained in part by site-specific actions of altered leptin signalling with differential effects on axial and appendicular regions, that may be dependent in part on muscle mass (10).
Leptin exerts its action through the leptin receptor, a member of the class I cytokine receptor family (11). In humans, two isoforms have been described (12). The long biologically active receptor isoform (OB-Rb) with full signalling capacity is primarily expressed in the hypothalamus but also peripherally; the second isoform is the soluble form (OB-Rs) corresponding to the extracellular domain of the receptor, lacking the domains capable of signal transduction. The human brain leptin receptor gene (LEPR) is encoded in 20 exons and spans over 70 kb of DNA (13). Previous studies revealed several sequence variants in the LEPR gene (1323), but mostly without any obesity-causing association (15, 16, 18). Some studies described association with serum leptin levels (13, 19, 22, 23). Only limited information is available on the relation between LEPR polymorphisms and bone (20, 24). One of the polymorphisms previously reported in Caucasians (13), is an A to G transition in the second position of codon 223, at position 668 in exon 6 of the LEPR gene. This results in the substitution of an arginine for a glutamine in the extracellular domain of the receptor with potential functional consequences on leptin binding activity (14, 19, 21).
The possible contribution of leptin to the regulation of human bone metabolism still remains to be clarified as studies of its role in the skeleton have yielded conflicting results with reports of both positive (7, 2529) and negative (3033) associations of leptin with bone mineral density (BMD), while other studies revealed no associations (30, 34) between leptin and skeletal status.
In this study, we aimed to explore the role of leptin in bone homeostasis in community-dwelling men over the age of 70 years. This elderly age group was characterized by relative increase of fat mass and decrease of lean body mass and by continuous bone loss with exponential increase in fracture risk. The present cross-sectional and longitudinal evaluation which considered association of bone parameters with serum leptin as well as the Gln223Arg LEPR gene polymorphism only revealed a limited role for leptin as a determinant of bone homeostasis in elderly men, independently of adiposity and serum-free estradiol.
| Materials and methods |
|---|
|
|
|---|
The subjects were recruited from the population register of a semi-rural community. A sample of 352 ambulatory men, aged 7186 years, agreed to participate. The longitudinal study was designed to investigate the process of aging by focusing on hormonal changes and bone metabolism at yearly intervals over a period of 4 years. All participants signed an informed consent approved by the ethical committee of the Ghent University Hospital (Belgium) and they completed questionnaires pertaining to medical and (family) fracture history, current smoking habits, alcohol consumption, dietary intake of calcium and physical activity during the previous year. Alcohol consumption was scored by the frequency of 2 or more alcoholic beverages a day. Calcium intake was estimated by a food questionnaire on dairy products accounting for the number of standard portions per week. Physical activity was assessed by recording the weekly frequency of activities and was scored as low, moderate or high. Following exclusions because of past or current history of disorders or treatments potentially affecting androgen status and/or bone metabolism, a complete data set for 270 subjects was available. Detailed exclusion criteria (3537) comprised hyperthyroidism, Pagets disease, inflammatory rheumatic disease, adrenal insufficiency, malignancy, fasting glycemia > 8.33 mmol/l, calcemia > 2.65 mmol/l, serum creatinine > 177 µmol/l or treatment with systemic glucocorticoids, (anti)-androgens, phenytoin, insulin, biphosphonates or vitamin D. The cross-sectional evaluation in the present study is part of the baseline examinations. The study population was invited for examination at yearly intervals for four years. The longitudinal study cohort consists of 214 eligible subjects for whom at least 1 follow-up visit was available; 5 serial measurements were available in 140 subjects; for the remaining subjects 2 to 4 data points were available.
Bone mineral density
Bone mineral density (g/cm2) at the proximal femur (reported here for the total hip region of interest) and at the distal forearm (reported here for the total distal ulna and radius region of interest) were measured on the nondominant side. Measurements were obtained at baseline and at yearly intervals in a 4-year follow-up period (38) using dual-energy X-ray absorptiometry with a Hologic QDR-1000 + device (Hologic, Inc., Bedford, MA, USA). Z-scores for BMD were calculated using the age- and gender-matched controls provided by the NHANESIII study group for the hip (39) and by the manufacturer for the distal forearm. The coefficient of variation (CV%) was < 1% for phantom measurements and ranged in vivo between 1.02.4%, as calculated from duplicate measurements in all elderly study subjects.
Body composition
The body composition, including fat and lean mass percentage, was estimated at baseline using bioelectrical impedance analysis (Bodystat 1500, Bodystat, Ltd, Isle of Man, UK) (37). The coefficient of variation (CV%) was 1.3% and 0.5% for fat mass and lean mass percentage, respectively, as calculated from duplicate measurements in 15 study subjects.
Biochemical indices of bone metabolism
At baseline, serum and second-void urine samples were obtained between 08 00 h and 10 00 h after overnight fasting, and were stored at 80 °C until analysis. The following markers of bone turnover and calciotropic hormones were measured by immunoassays: serum bone-specific alkaline phosphatase (S-BAP; Tandem-R Ostase; Hybritech, Inc., San Diego, CA, USA), serum intact osteocalcin (S-OC; N-MID Osteocalcin; Osteometer BioTech A/S, Copenhagen, Denmark), serum intact parathyroid hormone levels (PTH; Nichols Institute Diagnostics, San Juan Capistrano, CA, USA), serum 25-hydroxyvitamin D (25-OHD; Incstar Corp., Stillwater, MN, USA), serum C-terminal telopeptides of collagen type I (S-CTX; Elecsys ß-CrossLaps; Roche Diagnostics, Penzberg, Germany), urinary free deoxy-pyridinoline (U-DPD; Immulite Pyrilinks-D; Diagnostic Products Corp., Los Angeles, CA, USA) and urinary C-terminal telopeptides of type I collagen (U-CTX;
-CrossLaps; Osteometer BioTech, Copenhagen, Denmark). Concentrations of U-DPD and U-CTX were normalized for urinary creatinine concentration. The intra- and interassay coefficients of variation for all assays were below 10% and 15%, respectively.
Hormonal assays
Commercial immunoassay kits were used to determine serum levels of total testosterone (T) and luteinizing hormone (LH; results were expressed as IU/l of IRP 68/40) (Medgenix, Fleurus, Belgium), estradiol (E2) (Clinical assay, DiaSorin s.r.l., Saluggia, Italy: according to a modified protocol that doubles the serum amount), sex hormone binding globulin (SHBG) (Orion Diagnostica, Espoo, Finland), insulin (RIA, Pharmacia & Upjohn Diagnostics AB, Uppsala, Sweden), cortisol (Clinical Assay GammaCoat Cortisol RIA kit, Diasorin, Stillwater, Minnesota, USA), IGF-I and IGF-BP3 (IRMA, Diagnostic System Laboratories, Inc., Webster, TX, USA) and leptin (Human Leptin RIA kit, Linco Research, Inc., MO, USA). Dehydroepiandrosterone sulfate was measured with an in-house RIA. Serum-free testosterone (FT) and free estradiol (FE2) were calculated from serum total T, E2, SHBG and albumin concentrations using a previously validated equation (40, 41)
Determination of leptin receptor genotype
Genomic DNA (gDNA) was extracted from ethylene-diamine tetraacetic acid-treated blood using a commercial kit (Qiagen Midi Kit, Qiagen). PCR was performed using the primers described by Matsuoka et al. (15) and the reaction profiles were as follows: denaturation at 94 °C for 45 s, annealing at 63.5 °C for 45 s and extension at 72 °C for 90 s, for 30 cycles. Following digestion of the amplified gDNA with MspI, fragments were analysed on 2% agarose gel.
Statistical analysis
Using the KruskalWallis test, clinical characteristics of the 270 participants were compared among the 3 genotype groups (AA-AG-GG). A MannWhitney test was used for pairwise comparisons of U-CTX between genotypes and results were corrected for multiple comparison. For comparison of smoking, alcohol, physical activity and fracture incidence among LEPR genotypes, a chi square test was applied. Spearman correlations were first determined between leptin and age or body composition variables and secondly between leptin and hormones, biochemical markers of bone turnover, baseline BMD and longitudinal BMD changes after correction for absolute fat mass. The KruskalWallis test was used to study the relationship between leptin levels and smoking, alcohol and physical activity. To assess the association of leptin levels and LEPR with baseline BMD, longitudinal BMD changes and biochemical markers of bone turnover, all in relation to body composition and serum free E2 as major confounders, we used a general linear model with age, fat mass and free E2 included as covariates, next to leptin and LEPR. Variance inflation factors were used to test multicollinearity in the model, from this approach no evidence of multicollinearity was noticed. All statistical analyses were performed using SAS software (SAS Institute Inc., Cary, North Carolina, USA). A P value of < 0.05 was considered significant in all analyses.
| Results |
|---|
|
|
|---|
|
Although the presence of the G-allele showed a trend towards lower serum leptin levels, with means of 8.10, 6.82 and 6.52 ng/ml for AA, AG and GG genotypes, respectively, this effect did not reach statistical significance (P = 0.15), even after correction for absolute fat mass (P = 0.17). As such, LEPR genotypes and serum leptin were not associated.
In the cross-sectional analyses (n = 270), absolute fat mass was the index of body composition most strongly correlated with serum leptin levels in the elderly men (Table 2
). There was no association between leptin and smoking, alcohol or physical activity (results not shown). After correction for absolute fat mass, leptin was significantly correlated with free estradiol, insulin, cortisol, 25-OHD, creatinine, IGF-BP3, U-CTX and S-CTX (Table 2
).
|
|
|
|
| Discussion |
|---|
|
|
|---|
The observed strong correlation between leptin and fat mass lies within the line of expectations, since a strong association between serum leptin and adiposity in these elderly men had already been assessed in a previous study (37).
Limited associations between serum leptin and bone parameters are independent from adiposity. This might be explained by the strong correlation between serum leptin and total mass of fat tissue, which is the source of circulating leptin. Moreover, evidence from animal studies indicating a role of leptin in the regulation of bone metabolism, pertain mostly to experimental paradigms with initially perturbed bone metabolism, e.g. under unloading or after gonadectomy or to animal models of leptin deficiency (3, 4, 6, 9, 10).
Results of human studies relating leptin with bone have been inconsistent. Studies in women showed either a lack of association (after adjustment for body mass index (BMI)) (30, 42) or positive associations (7, 2529) with BMD. As to the findings in men, the lack of association between leptin and baseline BMD at the arm and hip in our elderly population is in agreement with previous studies (27, 34, 42), whereas serum leptin concentrations did not add to the prediction of bone mineral density. Other studies (31, 32) reporting a negative association between leptin and BMD in men are not confirmed in our cross-sectional analysis of baseline BMD, however they seem compatible with our finding that leptin was associated with BMD loss at the forearm.
Prospective assessment of longitudinal change in BMD showed association of higher leptin levels with greater BMD decrease, independently of fat mass and free estradiol, at the forearm but not at the hip. Possibly, a rather weak effect of leptin on BMD change, as observed at the forearm, is being masked by weight-bearing effects in the hip. In this context, it can also be noted that site-specific effects of leptin on the skeleton have been described (10). We are aware of one other report on the association of serum leptin with BMD changes. In the latter study by Dennison et al. (42), serum leptin was not associated with BMD changes at the spine or femoral neck in men.
The fact that the association of serum leptin with a change of longitudinal BMD at the forearm, in our study, has not translated into a similar association with baseline BMD, might indicate that the limited effect of serum leptin is being obliterated by lifelong exposure to a variety of positive and negative influences on bone mass in the considered elderly population.
As for the biochemical markers of bone turnover, we initially found a significant negative association of leptin with U-CTX. The seemingly paradoxical observation that leptin levels were associated with higher BMD loss, but at the same time linked with lower levels of U-CTX, might be due to confounding effects of creatinine excretions. Higher leptin levels were correlated with higher serum creatinine levels, a determinant for urinary creatinine levels for which U-CTX has been normalized. When taking into account creatinine as a confounder, leptin was no longer associated with U-CTX. Other studies report inconsistent results for biochemical markers of bone turnover. Scariano et al. (43) found a modest, but significant positive association with S-BAP in men and women. Upon adjustment for BMI, Schett et al. (44) observed an inverse relationship between leptin levels and osteocalcin and ß-crosslap, while Dennison et al. (42) found no significant associations with biochemical indices of bone turnover. Other studies in women have reported no (29) or only weak (7, 25) correlations with biochemical markers.
Next we studied the Gln223Arg polymorphism, which leads to an amino acid change in the extracellular domain common to all isoforms of the leptin receptor. This substitution has been reported to cause a change in charge (neutral to positive) and is therefore most likely to have functional consequences affecting peripheral and central leptin binding to the leptin receptor in humans (14, 19, 21). Quinton et al. 2001 (19) analysed serum leptin binding activity to determine whether the Gln223Arg polymorphism was directly associated with changes in ligand binding and they observed that Gln-carriers had significantly lower serum leptin binding activity.
In first instance we looked for associations of the Gln223Arg polymorphism with serum leptin levels and indices of body composition. Some previous studies did not find associations between LEPR and BMI or fat mass (18) while others (19, 23) reported that subjects carrying the Gln allele had higher body fat percentage and leptin levels. Although similar trends were noticed for leptin and BMI in our study, statistical significance was not found.
In second instance we looked for associations of the Gln223Arg polymorphism with BMD and biochemical markers of bone turnover. Our results indicated that in elderly men, LEPR polymorphisms had no important role on either baseline BMD or longitudinal change of BMD at the hip or the forearm. After correction for fat mass and free estradiol, the LEPR genotype emerged as a significant determinant of S-BAP and a trend was noticed for U-CTX. Allele-dose effects were observed with the Gln-allele being associated with lower levels of S-BAP and U-CTX. In the literature, data on the relation between LEPR mutations and bone metabolism are scarce. Two studies in young male subjects suggest a positive effect of the Gln-allele on bone acquisition involving a possible interaction with other genotypes (20) or with lifestyle-related factors such as physical activity (24).
With this study, we tried to get a better insight into the contribution of leptin to the regulation of skeletal homeostasis in elderly men. The strength of the study is the careful clinical evaluation of a well-defined community-dwelling study population, including prospective longitudinal BMD data and allowing adjustment for major confounders. Of course these observations are limited to elderly men and should not be readily extrapolated to other age groups; moreover, they seem to differ from findings in women. In conclusion, our overall findings seem to indicate a possible role for leptin as a determinant of bone homeostasis in elderly men, but with only modest impact independently from body composition and non SHBG-bound E2.
| Acknowledgements |
|---|
| Funding |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L. Richert, T. Chevalley, D. Manen, J.-P. Bonjour, R. Rizzoli, and S. Ferrari Bone Mass in Prepubertal Boys Is Associated with a Gln223Arg Amino Acid Substitution in the Leptin Receptor J. Clin. Endocrinol. Metab., November 1, 2007; 92(11): 4380 - 4386. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |