Effect of anticoagulation with citrate versus heparin on the adsorption of coagulation factors to blood purification resins with different charge.

In liver failure, hydrophobic toxins accumulate in the blood circulation. To support hepatic function, extracorporeal blood purification systems have been developed, in which both cationic and neutral adsorbents are used to remove albumin-bound metabolites from blood. An issue of these systems is the additional removal of coagulation factors containing negatively charged γ-carboxyglutamate (Gla) domains, which, in physiological conditions, are shielded by calcium ions. We hypothesized that complexation of calcium ions by citrate leads to exposure of negative Gla domains, resulting in their binding to the positively charged adsorbents. The data presented here confirm that the binding of coagulation factors containing Gla domains to positively charged polymers is enhanced in the presence of citrate as compared to heparin. This effect increased with increasing charge density of the polymer and has important implications for the clinical application of positively charged polymers.

Response of serum carboxylated and undercarboxylated osteocalcin to alendronate monotherapy and combined therapy with vitamin K2 in postmenopausal women.

Alendronate decreases the risk of femoral neck fracture by suppressing bone turnover, and also decreases the serum total osteocalcin level. A low serum carboxylated osteocalcin level or high undercarboxylated osteocalcin level could be risk factors for femoral neck fracture. Vitamin K mediates the carboxylation of osteocalcin, but the effect of alendronate therapy with or without vitamin K(2) supplementation remains unknown. Forty-eight postmenopausal women were enrolled in a 1-year prospective randomized trial and assigned to alendronate monotherapy (5 mg/day) (group A, n = 26) or vitamin K(2) (45 mg/day) plus alendronate (5 mg/day) (group AK, n = 22). Bone mineral density was measured by dual-energy X-ray absorptiometry at 0 and 12 months; bone turnover parameters were measured at 0, 3, and 12 months. Four patients discontinued alendronate therapy, and we analyzed the remaining 44 patients (23 in group A and 21 in group AK) who completed 1 year of treatment. Alendronate decreased undercarboxylated osteocalcin; carboxylated osteocalcin was not affected. Addition of vitamin K(2) enhanced the decrease of undercarboxylated osteocalcin levels and led to a greater increase of femoral neck bone mineral density. Alendronate monotherapy does not decrease carboxylation of osteocalcin, and combination of vitamin K(2) and alendronate brings further benefits on both osteocalcin carboxylation and BMD of femoral neck in postmenopausal women with osteoporosis.

Serum levels of calcification inhibition proteins and coronary artery calcium score: comparison between transplantation and dialysis.

Vascular calcifications in CKD are now linked to serum alterations of both divalent ions and calcification inhibitory proteins. Due to possible biochemical differences between dialysis (D) and transplantation (Tx), we examined the entity and severity of these biochemical modifications and of coronary artery calcium score separately in these two populations. We assayed, besides standard markers of inflammation, divalent ions and serum levels of fetuin, matrix Gla protein (MGP) and osteoprotegerin (OPG), in 51 Tx patients (age 45 +/- 12 years; 30 males, 21 females; previous D duration 4.8 +/- 4.2 years; Tx since 6.6 +/- 5.5 years; Cr 1.8 +/- 0.6 mg/dl) and in 49 D patients (age 49 +/- 14 years; 30 males,19 females; D duration 5.6 +/- 4.8 years). Additionally, coronary calcium score (AS) was evaluated by cardiac multi-slice CT. Compared with D patients, Tx patients had better values of divalent ions and inflammation markers, and lower prevalence (65 vs. 86%; p < 0.02) and severity (AS = 570 +/- 1,637 vs. 1,311 +/- 3,128; p < 0.008) of coronary calcification. In addition, a tendency toward normalization for all of the three calcification inhibitory proteins was evident. In both Tx and D, AS correlated with age and OPG (Tx: r(s) = 0.439, p < 0.001, and r(s) = 0.510, p < 0.0001; D: r(s) = 0.471, p < 0.001, and r(s) = 0.403, p < 0.005, respectively); in D patients, a correlation was present also with D duration (r(s) = 0.435; p < 0.002), other markers of inflammation and, notably, fetuin (r(s) = -0.442; p < 0.002). Regression analysis selected previous time on D in Tx patients (r(m) = 0.400; p < 0.004), and C-reactive protein and OPG in D patients (r(m) = 0.518; p < 0.004) as the most predictive parameters of AS. Discriminant analysis confirmed the major role of age and D duration in the appearance of AS and evidenced male gender as a distinct risk condition. At variance, Tx duration was never associated with AS. In conclusion, as compared to D, renal Tx patients show serum levels of calcification inhibition proteins and of divalent ions closer to normal. As this is associated with a lower prevalence and severity of AS, it is suggested that Tx antagonize the accelerating role of D in the progression of vascular calcification. Assessment of both coronary calcifications and serum levels of calcification inhibitory proteins may be of value to identify those subjects at higher risk of development and progression of vascular lesions, among whom males have the highest rate.

Comparative analysis of serum proteomes: discovery of proteins associated with osteonecrosis of the femoral head.

No means exist to evaluate the activity status, turnover, and prognosis of idiopathic osteonecrosis of the femoral head (IONFH) except for X-ray evidence of segmental collapse as a very good marker for prognosis. Moreover, the only current method for diagnosis of this disease is through physical examination and diagnostic imaging results, and no serum biochemical markers exist. A comparative analysis of serum proteomes was performed to discover proteins associated with osteonecrosis of the femoral head. Two-dimensional electrophoresis (2-DE) patterns of human sera from 10 patients with IONFH and 10 normal subjects were analyzed. The differentially expressed spots were identified by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), and 7 proteins were found. The expression levels of tissue-type plasminogen activator (t-PA), bone-carboxyglutamate protein (BGP), c-sis, and an unknown protein were downregulated in the sera of patients with IONFH, whereas the other 3 proteins, including plasminogen activator inhibitor type 1 (PAI-1), crosslaps, and anti-p53 antibody, were upregulated. To examine their applicability as diagnostic markers, levels of the 6 identified proteins in serum were validated from patients with IONFH, osteoarthritis (OA), rheumatoid arthritis (RA), and fracture using the enzyme-linked immunosorbent assay (ELISA) method. It was found that only serum levels of t-PA, PAI-1, crosslaps, and anti-p53 antibody in patients with IONFH were always significantly different from those in patients with OA, RA, and fracture. These results suggest that serum levels of t-PA, PAI-1, crosslaps, and anti-p53 antibody could be used as noninvasive diagnostic biomarkers for IONFH.

[Serum osteoprotegrin and serum bone gamma-carboxyglutamine acid-containing protein are correlated with bone mineral density in normal women].

OBJECTIVE: To study the relationships of serum osteoprotegrin (sOPG), serum bone gamma-carboxyglutamine acid-containing protein (sBGP), and urine deoxypyridinoline (uDPD)/creatinine (Cr) with age and bone mineral density (BMD) in women. METHODS: ELISA was used to examine the sOPG, sBGP, and uDPD/Cr of 672 female volunteers aged 20-80. The BMD (QDR4500A) value of the anteroposterior lumbar spine and femoral neck were measured by DXA. RESULTS: (1) The levels of sOPG, sBGP, and uDPD/Cr in the age group of 30-39 were 2.8 pmol/L +/- 1.4 pmol/L, 5 microg/L +/- 3 microg/L, and 4.9 nmol/mmol +/- 2.5 nmol/mmol respectively, all significantly lower than those in the age groups 40-49, 50-59, and 60-69 (all P < 0.05). (2) In the age group 40-49, the values of sOPG, sBGP, and uDPD in menopausal subjects were significantly higher than those of the non-menopausal subjects (5.7 pmol/L +/- 3.1 pmol/L vs 3.4 pmol/L +/- 2.0 pmol/L, 11 microg/L +/- 5 microg/L vs 6 microg/L +/- 3 microg/L, and 6.9 nmol/mmol +/- 3.3 nmol/mmol vs 5.2 nmol/mmol +/- 3.9 nmol/mmol, all P < 0.001). (3) Age was positively correlated with sOPG, sBGP, uDPD/Cr, and BMD of anteroposterior lumbar spine and femoral neck (r = 0.130, 0.355, 0.106, -0.600, -0.545; P < 0.01). sOPG and sBGP were negatively correlated to anteroposterior lumbar spine BMD (r = -0.183, -0.108, P < 0.01; and r = -0.541, -0.441, P < 0.001). sOPG was positively correlated with sBGP and uDPD/Cr (r = 0.216 and 0.083; both P < 0.05). CONCLUSION: sOPG, sBGP, and uDPD/Cr can be used as sensitive markers to determine the bone turnover status, which is changeable with age, and menopausal status in women, and predict the bone lose prior to BMD determination by DXA.

[Effect of raloxifene hydrochloride on bone mineral density, bone metabolism and serum lipids in Chinese postmenopausal women with osteoporosis].

OBJECTIVE: To determine the effect of raloxifene hydrochloride (RLX) on the lumbar spine and total hip bone mineral density (BMD), bone metabolism and serum lipids in Chinese postmenopausal women with osteoporosis. METHODS: 204 Chinese postmenopausal women with osteoporosis from 3 hospitals in Beijing and Shanghai were randomly divided into 2 groups of 102 women: RLX group (RLX of the dosage of 60 mg/day was given for 12 months) and placebo group. In addition, 500 mg of elemental calcium and 200 units of vitamin D were given daily to all women. BMD, serum bone markers and lipids were measured before and after drug administration. The BMD of lumber spine and hip was measured by dual-energy X-ray absorptiometry (DEXA). Serum bone gamma-carboxyglutamic acid-containing protein (BGP) and C-teloppeptide were analyzed by one-step ELISA. Serum lipids were measured by enzymatic method. RESULTS: By the end of the 12-month study period, the lumbar spine BMD was increased by 3.3% +/- 4.8% in the RLX group and 1.0% +/- 4.9% in the placebo group (P < 0.001); the hip BMD was increased by 1.4% +/- 4.8%in the RLX group and decreased by 0.9% +/- 5.0% in the placebo group (P < 0.01). New vertebral fracture occurred in none of the subjects in the RLX group and in 5 subjects of the placebo group (P = 0.059). The serum BGP and CTX decreased by 41.7% and 61.5% respectively in the RLX group, both significantly more than those in the placebo group (10.6% and 35.6% respectively, both P < 0.001). Both the total cholesterol and low-density lipoprotein cholesterol were significantly lower in the RLX group than in the placebo group (both P < 0.001), however, there were no significant differences in high-density lipoprotein cholesterol and triglycerides between these two groups. One subject in the RLX group and 5 subjects in the placebo group discontinued the study due to adverse events. There were no differences in the number of subjects with hot flushes (3 in the RLX group and 1 in the placebo group) and the number of subjects with leg cramps (9 in the RLX group and 4 in the placebo group). No venous thromboembolic event was reported. CONCLUSION: RLX of the dosage of 60 mg/day for 12 months significantly increases the lumbar spine and total hip bone BMD, significantly decreases bone turnover and has favorable effects on serum lipids in Chinese postmenopausal women with osteoporosis.

Analysis of gamma-carboxyglutamic acid content of protein, urine, and plasma by capillary electrophoresis and laser-induced fluorescence.

When the properties of an analyte are known, the separation system can be designed to make the analyte of interest migrate at either a much faster or a much slower velocity compared to other molecules in the sample matrix. A simple and sensitive method to analyze the gamma-carboxyglutamic acid (Gla) content of protein, urine, and plasma was developed using capillary electrophoresis with laser-induced fluorescence detection (CE-LIF). The separation method is designed according to the specific properties of three amino acids of interest. The number of Gla residues from three vitamin K-dependent proteins were estimated by quantifying the amount of fluorescein thiocarbamyl derivative of Gla after alkaline hydrolysis and fluorescein isothiocyanate labeling. Human prothrombin, blood coagulation factor X, and bovine osteocalcin were calculated to have 10.0 +/- 0.7, 11.0 +/- 0.6, and 2.1 +/- 0.1 Gla residues per mole of protein, respectively, which agreed well with amino acid sequencing data. The analysis of free Gla content in urine and plasma was also demonstrated by this method. It was demonstrated that submicrograms of protein can be characterized by CE-LIF.

Effects of nutritional supplementation on bone mineral status of children with rheumatic diseases receiving corticosteroid therapy.

OBJECTIVE: Because children with rheumatic disease receiving longterm corticosteroids are at high risk for developing osteoporosis, we attempted to determine whether nutritional supplementation would improve bone status in this group of children. METHODS: In a crossover design study, 10 corticosteroid treated children with rheumatic disease and osteoporosis received calcium and vitamin D supplementation for 6 months to determine their effect on bone density. They were then studied for 6 months without added nutrition supplements. The mean age was 13.1 years with a mean duration of disease of 4.2 years. Six patients had juvenile rheumatoid arthritis, 2 had systemic lupus erythematosus and 2 had mixed connective tissue disease. These children obtained a minimum of 1 g of calcium and 400 IU of vitamin D daily from diet and added supplements. Dual photon absorptiometry, laboratory and dietary data were obtained at baseline, 6 months, and one year. RESULTS: Spinal bone density significantly improved with supplementation. Osteocalcin values remained low throughout the study. CONCLUSION: Our results suggest some children with rheumatic disease receiving corticosteroids would benefit from calcium and vitamin D supplementation.

Increased plasma free gamma carboxyglutamic acid levels during deep vein thrombosis and intravascular disseminated coagulation.

Gammacarboxyglutamic acid (gla) is a non essential amino acid synthesized in presence of vitamin K, predominantly found in coagulation and bone proteins. In 14 cases of deep vein thrombosis and in 11 cases of disseminated intravascular coagulation, compared to 19 normal subjects and 9 patients hospitalized for leg pain, free plasma gla levels were found significantly elevated (respectively 372 +/- 244 and 559 +/- 361 versus 146 +/- 34 and 120 +/- 40 pmol/mL). In six paired plasma and serum, gla levels were similar. These results suggest an involvement of blood coagulation in gla generation with need of a catabolism of the activated factors. A significant decrease was noticed during vitamin K antagonist therapy and liver disease, both instances in which the synthesis of gla containing coagulation factors is affected. During hepatocellular carcinoma with elevated desgamma carboxyprothrombin, gla was found normal, denying an global impairement of the vitamin K metabolism.

Phylloquinone transport and its influence on gamma-carboxyglutamate residues of osteocalcin in patients on maintenance hemodialysis.

Plasma concentrations of phylloquinone in 42 fasting hemodialysis patients showed a much wider range than in healthy adults. Phylloquinone concentrations were best predicted by the concentration ratio of beta-very-low-density-lipoprotein to low-density-lipoprotein cholesterol (r = 0.71), which is closely related to chylomicron-remnant clearance. Phylloquinone concentrations in plasma were related to apolipoprotein E genotype in the order E2 > E3 > E4. The percentage of carboxylated osteocalcin (HBC) was related to the plasma concentration of phylloquinone in patients with the apolipoprotein E genotype E3/3 (r = 0.52, P < 0.05), and in patients with the genotypes E2/3 and E2/2 (r = 0.23, P < 0.1). Overall, plasma triglyceride concentration was a better predictor for HBC than was the plasma concentration of phylloquinone. These results point to the overriding importance of chylomicrons for the transport of phylloquinone to liver and bone. Delivery to osteocalcin-producing osteoblasts seemed impaired in patients with the low receptor-affinity apolipoprotein variant E2, suggesting a major role of receptor-mediated chylomicron-remnant uptake in the transport of phylloquinone to bone.

Role of gamma-carboxyglutamic acid residues in the binding of factor IXa to platelets and in factor-X activation.

To study the requirements for factor-IXa binding to platelets and factor-X activation, we examined the consequences of chemical modification (factor IXMOD) or enzymatic removal (factor IXDES) of gamma-carboxyglutamic acid (Gla) residues. In the presence of factor VIIIa and factor X, there were 344 (+/- 52) binding sites/platelet for factor IXaMOD (apparent dissociation constant [kdapp] = 4.5 +/- 0.9 nmol/L) and 275 (+/- 35) sites/platelet for factor IXaDES (kdapp = 5.0 +/- 0.8 nmol/L) compared with 580 (+/-65) sites/platelet for normal factor IXa (factor IXaN) (kdapp = 0.61 +/- 0.1 nmol/L) and 300 (+/-62) sites/platelet for factor IX (kdapp = 2.9 +/- 0.29 nmol/L). The concentrations of factor IXaN, factor IXaMOD and factor IXaDES required for half-maximal rates of factor-Xa formation were 0.67 nmol/L, 3.5 nmol/L, and 6.7 nmol/L. Whereas maximal velocities (Vmax) of factor Xa formation by factor IXaMOD (approximately 0.8 nmol/L.min-1) and factor IXaN (approximately 10.5 nmol/L.min-1), turnover numbers (kcat expressed as moles of factor Xa formed per minute per mole of factor IXa bound), and values of catalytic efficiency (kcat/Km) were normal, indicating that the decreased rates of factor X activation observed with factor IXaMOD and factor IXaDES are solely a consequence of the abnormal binding of these proteins to thrombin-activated platelets in the presence of factor VIIIa and factor X. Thus, factor IXa binding to platelets is mediated in part, but not exclusively, by high-affinity Ca2+ binding sites in the Gla domain of factor IX.

[Diagnostic evaluation of plasma free gamma-carboxyglutamic acid in maintenance hemodialytic patients with renal osteodystrophy].

The levels of serum Bone Gla Protein (BGP) and plasma free gamma-carboxyglutamic acid (free gamma-Gla) were measured in 36 patients on maintenance hemodialysis and compared with conventional parameters of renal osteodystrophy (ROD), such as C-PTH, intact PTH, Al-PIII and hydroxyproline levels in serum, and sigma GS/D. Even though the values of BGP were significantly elevated in patients group, 64.5 +/- 1.74 ng/ml, but the low values within normal range could be found in 3 of 36 patients, also. On the other hand, plasma levels of free gamma-Gla were markedly elevated in all cases as high as 10.40 +/- 3.10 nmol/ml in average compared to 1.29 +/- 0.37 nmol/ml (n = 15) in healthy controls. Comparative analysis of BGP with other parameters showed exclusively significant correlation as follows, C-PTH: r = 0.833 (p less than 0.01), intact PTH: r = 0.702 (p less than 0.01), Al-PIII: r = 0.771 (p less than 0.01), hydroxyproline: r = 0.462 (p less than 0.01), sigma GS/D: r = -0.407 (p less than 0.05). Respective for its higher levels, comparative analysis of plasma levels of free gamma-Gla with other parameters failed to show a significant correlation except for correlation with C-PTH (r = 0.459, p less than 0.01). Based upon these results, we could confirm a marked elevation in plasma levels of free gamma-Gla in all patients, but couldn't point out its clinical value in ROD in this study.

Serum bone-gla protein after fracture.

Serum bone Gamma-carboxyglutamic acid (bone-gla) protein (BGP), a marker of bone formation, was measured in serial blood samples drawn from 14 patients who had fractured at least one of their tibial or femoral diaphyses and from two other patients who had sustained major trauma without fracture but who had been immobilized. A total of 85 samples were taken and analyzed during the first three months after the fractures occurred. Serum BGP significantly increased and positively correlated with the time that had elapsed after the fracture, with an average twofold increase after two months. The fracture site and the duration of immobilization had no influence on the serum BGP levels. Serum BGP levels from the two non-fractured cases increased in the first two weeks with no subsequent consistent trend. These data suggest that serum BGP increases one to two months after a major fracture, possibly as a manifestation of bone repair. Further studies are required to determine the potential clinical value of serum BGP in the management of such patients.

Serum markers of bone formation in parenteral nutrition patients.

Bone gamma-carboxyglutamic acid containing protein (BGP) has been utilized effectively as a serum marker of bone turnover in healthy normals and in individuals with a variety of metabolic bone disorders including postmenopausal osteoporosis and Paget's disease. The utility of this serum marker in other bone disorders, including that associated with the maintenance of patients on long-term parenteral nutrition, still requires definition. Because of our interest in this clinical syndrome and the availability of serum and of bone formation rates (BFR) measured directly from double tetracycline labeling in 11 long-term parenteral nutrition patients, we measured BGP levels in these patients and attempted to correlate this measure with BFR. Serum vitamin D metabolites, immunoreactive parathyroid hormone (PTH), and alkaline phosphatase (alk phos) were also measured. Serum BGP was only weakly and not significantly correlated (r = 0.24, p = NS) with bone formation rate for the group as a whole. However, in a subgroup of 10 patients without hyperparathyroidism, there was strong and significant correlation (r = 0.81, P less than 0.01) between BGP and BFR. There was also a strong correlation between bone formation rate and serum 1,25 dihydroxyvitamin D [1,25(OH)2D] levels (r = 0.89, P less than 0.01, n = 11). The mechanism of this association could not be established. A correlation of borderline significance was observed between bone formation rate and serum alk phos (r = 0.60, P = 0.05, n = 11). The current data suggest that additional studies may help to more fully define the utility of serum measurements in quantifying bone dynamics in parenteral nutrition patients, and that measures of vitamin D metabolites, BGP, and alk phos may prove useful.

Reduced axial bone mineral content in patients taking an oral anticoagulant.

Recent literature indicates that the vitamin K cycle plays a role in the calcification process, presumably via its intervention on gamma-carboxylation of the noncollagenous proteins of bone osteocalcin and matrix gamma-carboxyglutamic acid protein. The major clinical evidence of this interference is fetal bone defect caused by oral anticoagulants given to the mother during the first trimester of pregnancy. No bone abnormalities have been reported so far in adults receiving oral anticoagulants. We studied 56 women who had had cardiac valve replacement and who were given acenocoumarol as anticoagulant, and 61 age-matched women who were in the same New York Heart Association functional class but who were not taking anticoagulants. Osteocalcin serum levels were comparable between the two groups; bone density values measured at appendicular bone were significantly lower in patients taking acenocoumarol. No significant correlation was found between duration of treatment and bone density. Significant osteopenia was present in the women being treated with oral anticoagulants.

[Calcium regulating hormone in the pathological changes of bone in rheumatoid arthritis].

To further elucidate the pathogenesis of calcium regulating hormones and bone Gla-protein (BGP) on the deteriorated bone changes in rheumatoid arthritis (RA), the degree of the appendicular bone atrophy, using microdensitometry method in patients with RA, and the concentrations of midchain mainly recognized parathyroid hormone (p-PTH1-84, intact parathyroid hormone (in-PTH1-84) and BGP in sera of patients with RA were measured. With concomitant progression in X-ray stage in RA, the levels of p-PTH1-84, in-PTH1-84, BGP, 25 (OH) D, and 24, 25 (OH)2D were significantly increased in sera of RA patients. The levels of 1,25 (OH)2D were not shown significant differences among RA patients classified by bone stage. These results might be concluded that the pathogenesis of the deteriorated bone change in RA seemed to be implicated in hyper-turnover bone presumably caused by the actions of parathyroid hormone and abnormal vitamin D metabolism.

Evidence that free gamma carboxyglutamic acid circulates in serum.

We report a rapid, sensitive and reproducible method to measure free gamma-carboxyglutamic acid (GLA) in serum using precolumn derivatization with O-phthalaldehyde, reversed-phase chromatography and deproteinization of serum by ultrafiltration. Serum free GLA level in 62 healthy adults was 167 +/- 46 pmol/ml and was increased (302 +/- 195 pmol/ml) in 14 patients with primary hyper-parathyroidism, a disease characterized by an increased bone turnover. Peptide bound GLA averaged 413 pmol/ml. In rabbits receiving massive doses of warfarin during 6 days there was a time- and dose-dependent decrease of serum free GLA but a significant fraction was still measurable. These data indicate that free GLA circulates and originates both from the metabolism of the vitamin K-dependent clotting factors and from bone metabolism.

Increased circulating levels of gamma-carboxyglutamic acid-containing protein and decreased bone mass in children on anticonvulsant therapy.

In order to investigate the pathophysiology of anticonvulsant-induced osteopenia, circulating levels of bone gamma-carboxyglutamic acid-containing protein (Bone Gla Protein: BGP) and urinary excretion of BGP were measured in 16 children on chronic anticonvulsant therapy and in 12 control children. Using microdensitometry analysis, osteopenia was found in 25% of the anticonvulsant therapy group, but it was not observed in the control group. Serum BGP and A1-P levels were significantly increased in the anticonvulsant group compared with the control group (P less than 0.05 and P less than 0.01, respectively), and a positive correlation was found between serum BGP and A1-P levels (P less than 0.05). Urinary excretion of BGP and hydroxyproline showed an increase in the anticonvulsant group, but it was not statistically significant. On the other hand, there was no significant difference between the two groups in serum levels of vitamin D metabolites, PTH, calcitonin, Ca, or P or in urinary excretion of Ca or P. It is suggested, therefore, that the increased BGP level in children receiving anticonvulsant therapy is a reflection of high bone turnover due to anticonvulsant drug complications.