Altered levels of biochemical indices of bone turnover and bone-related vitamins in patients with Crohn's disease and ulcerative colitis

Authors


Professor K. D. Cashman, Department of Food and Nutritional Sciences, Department of Medicine and Alimentary Pharmabiotic Centre, University College, Cork, Ireland.
E-mail: k.cashman@ucc.ie

Summary

Background  The pathogenesis of inflammatory bowel disease-associated osteopenia may be related to pathological rates of bone turnover; however, the literature shows mixed results.

Aim  To compare bone biomarkers in inflammatory bowel disease patients (Crohn's disease: n = 68, and ulcerative colitis: n = 32, separately) with age- and sex-matched healthy controls.

Subjects  Patients and controls were recruited from Cork University Hospital and Cork City area, respectively.

Results  Relative to that in their respective controls, Crohn's disease (n = 47) and ulcerative colitis (n = 26) patients (i.e. excluding supplement users) had significantly (P < 0.05–0.001) higher serum undercarboxylated osteocalcin (by 27% and 63%, respectively) and bone-specific alkaline phosphatase (by 15% and 21%, respectively) and urinary Type I collagen cross-linked N-telopeptides concentrations (by 87% and 112%, respectively). Relative to that in their respective controls, Crohn's disease and ulcerative colitis patients had significantly (P < 0.01) lower serum total osteocalcin (by 20% and 42%, respectively) and 25-hydroxyvitamin D (by 37% and 42%, respectively), while serum parathyroid hormone levels were similar. In the combined patient group (n = 100), undercarboxylated osteocalcin was positively associated with bone markers.

Conclusions  Both Crohn's disease and ulcerative colitis patients have altered bone turnover relative to that in healthy controls.

Introduction

Crohn's disease (CD) and ulcerative colitis (UC), collectively referred to as inflammatory bowel disease (IBD), are chronic aggressive disorders, which have a combined incidence of about 15–25/100 000 in western Europe.1 CD and UC are important diseases, increasing in frequency, disabling for many patients, and generating a significant burden in the health-care system.2 Osteopenia and osteoporosis are recognized complications of IBD, and while there is some uncertainty about their exact prevalence, it is generally accepted that the prevalence of reduced bone mineral density (BMD) is increased in IBD patients.3, 4 This bone loss is a major risk factor for osteoporotic fractures of the spine, wrist and hip,5 which can have a deleterious effect on the quality of life of IBD patients, especially in young patients who have a normal life expectancy.

The pathogenesis of osteopenia and osteoporosis in CD has been suggested to result from pathological rates of bone remodelling and turnover arising from multifactorial, but incompletely characterized, mechanisms that include sex hormone deficiency, reduced physical activity, prolonged corticosteroid therapy, bowel surgery, smoking, and the potential deleterious effects of circulating cytokines and other mediators [e.g. interleukin-1 and -6 and tumour necrosis factor (TNF)] released by the inflamed intestines.6 Another pathogenetic mechanism which has been implicated in the low BMD in CD is the existence of one or more nutritional inadequacies, not only calcium7–9 but also vitamins D10, 11 and K,12, 13 in these patients. While the factors underpinning bone loss in UC are also incompletely understood, it is assumed that many of the above mechanisms are also involved. Thus, there is a need for an improved understanding of the pathophysiological processes underlying bone loss and altered rates of bone turnover in IBD.

Biochemical markers that reflect the processes of bone resorption and bone formation, and thus bone turnover, can be measured in blood and urine.14 Utilization of such markers in studies of IBD patients may provide some insight into how altered rates of bone metabolism may underpin IBD-related osteopenia and osteoporosis. However, studies of bone marker levels in IBD patients have produced conflicting results. While some studies of IBD patients have reported increased levels of bone resorptive markers without a compensatory increase in formation markers,12, 15–19 other studies have reported either reduced levels of markers of bone formation and no difference in resorptive markers,20 elevated levels of both types of markers,21, 22 or indeed no difference in markers23 between IBD patients and control subjects. However, in addition to differences in the biochemical markers used, some caution is also warranted in comparing the results of some of these studies because of major differences in the various IBD patient populations, which were studied, especially with regard to the type of disease (CD vs. UC), corticosteroid usage and disease activity (active disease vs. disease remission).13 For example, only two studies have compared the levels of bone turnover markers in UC patients alone compared with controls,21, 22 and these have produced conflicting results. Similarly, only two studies have compared the levels of bone turnover markers in patients with quiescent CD, and not taking steroids, with those of controls,20, 23 and these also have produced conflicting results.

Thus, the aim of the present study was to compare the rate of bone formation and bone resorption in quiescent IBD patients (CD and UC, separately) with age- and sex-matched healthy controls, using sensitive and specific biochemical markers of bone turnover. In addition, indices of vitamin D and K status were compared between IBD (CD and UC, separately) patients and their respective healthy controls, so as to get a better understanding of the prevalence of selected nutritional aetiological factors for IBD-associated bone loss.

Methods

Subjects

Patients attending the Cork University Hospital IBD Clinic in whom there was a confirmed diagnosis of CD and UC and who were in clinical remission were randomly asked to participate in the study. There was no selection bias other than that those selected to participate in the study had to be in close geographical proximity to the research centre in consideration of patient inconvenience. CD and UC were previously diagnosed on the basis of consistent clinical findings, barium radiology and histology. Remission was defined at the time of study as the absence of gastrointestinal symptoms and not requiring therapeutic doses of corticosteroids. A total of 68 CD patients (27 male, 41 female) and 32 UC patients (14 male, 18 female) participated in the study. Age- and sex-matched healthy control subjects (within 2 years of age maximum), without disorders of the gastrointestinal tract or bone, were recruited by leaflet or direct contact from the Cork City area. Mean age, weight, height, body mass index (BMI) and the ratio of males to females for the CD patient population and their age- and sex-matched healthy controls, and for the UC patient population and their age- and sex-matched healthy controls are provided in Tables 1 and 2 respectively.

Table 1.  Characteristics of patients with CD (n = 68) compared with age- and sex-matched healthy control subjects (n = 68)
 CD patientsControl subjectsP-value*
Means.d.Means.d.
  1. * Comparison of mean values between CD patients and age- and sex-matched control subjects, where appropriate, using unpaired Student's t-test.

  2. CD, Crohn's disease; BMI, body mass index.

Age (years)36.011.635.911.50.96
Height (m)1.70.091.70.090.72
Weight (kg)69.817.368.311.00.57
BMI (kg/m2)24.35.224.13.40.78
Males/females27/4127/41
Table 2.  Characteristics of patients with UC (n = 32) compared with age- and sex-matched healthy control subjects (n = 32)
 UC patientsControl subjectsP-value*
Means.d.Means.d.
  1. * Comparison of mean values between UC patients and age- and sex-matched control subjects, where appropriate, using unpaired Student's t-test.

  2. UC, ulcerative colitis; BMI, body mass index.

Age (years)40.511.040.311.20.93
Height (m)1.70.091.70.090.64
Weight (kg)74.514.669.910.20.15
BMI (kg/m2)26.04.024.63.20.16
Males/females14/1814/18

All CD patients were on maintenance 5-aminosalicylates (2–3 g/day), eight were taking mercaptopurine (6-mercaptopurine; 50 mg/day) or azathioprine (100 mg/day). No patients were taking warfarin or methotrexate. None had received infliximab or other anti-TNF therapy. Only one patient was taking a low dose of steroid (prednisolone, 5 mg/day) during the study and one was taking metronidazole at the time of the study (200 mg b.d.). The mean erythrocyte sedimentation rate was 10 (range: 2–28/h) and only four patients had modestly elevated C-reactive protein concentration of 25–28 (normal: <10) mg/L. Fifteen of the patients had small bowel involvement only, 11 patients had inflammation of the colon and 42 had inflammation in both sites. Thirty-one patients had a previous terminal ileal resection (<50 cm).

Eight female CD patients and eight female control subjects were postmenopausal; of those three patients and two control subjects were receiving hormone replacement therapy (HRT). Seven female CD patients and six control females were taking oral contraceptive medications. At the time of inclusion, none of the patients or controls was taking activated forms of vitamin D (25-hydroxyvitamin D [25 (OH) D] or 1,25 (OH)2D), calcitonin, or bisphosphonates, while 12 patients and 10 controls were regular cigarette smokers. None of the subjects was taking vitamin K supplements at the time of the study. Twenty-one patients (31%) and 20 controls (29%) were taking vitamin D- (range: 2.5–20 μg/day) and/or calcium-containing supplements (range: 250–1000 mg/day).

All UC patients were on maintenance 5-aminosalicylates (2–3 g/day). No patients were taking warfarin, methotrexate, mercaptopurine or azathioprine. None had received infliximab or other anti-TNF therapy. In addition, none of the patients was taking steroids or metronidazole during the study. The mean erythrocyte sedimentation rate was 9 (2–25/h) only two patients had modestly elevated C-reactive protein concentration of 28–30 (normal: <10) mg/L. Fourteen of the patients had left-sided colitis, 10 patients had pan (total colitis) and eight had limited colitis.

Six female UC patients and five female control subjects were postmenopausal; of those one patient and one control subject were receiving HRT. Four female UC patients and three control females were taking oral contraceptive medications. At the time of inclusion, none of the patients or controls was taking activated forms of vitamin D [25 (OH) D or 1,25 (OH)2D], calcitonin, or bisphosphonates, while six patients and four controls were regular cigarette smokers. None of the subjects was taking vitamin K supplements at the time of the study. Six patients (19%) and six controls (19%) were taking vitamin D- (range: 2.5–20 μg/day) and/or calcium-containing supplements (range: 250–1000 mg/day).

Ethical considerations

Before participation in this study, all subjects signed an informed consent document approved by the Clinical Research Ethics Committee of the Cork Teaching Hospitals.

Collection and preparation of samples

Each participant was invited to provide a fasting blood and urine sample at the University. After an overnight fast, a blood sample (20 mL) was taken between 08:30 and 10:30 hours by a trained phlebotomist. Blood was collected by venepuncture into a vacutainer tube with no additive and processed to serum, which was immediately stored at −80 °C until required for analysis. Subjects were supplied with suitable collection containers for urine samples and asked to collect first morning void urine samples. Portions of urine were stored at −20 °C from the morning of collection until required for analysis.

Experimental techniques

Urinary Type I collagen cross-linked N-telopeptides

Type I collagen cross-linked N-telopeptides (NTx) was measured in urine samples by an enzyme-linked immunosorbent assay (ELISA; Osteomark, Ostex International, Inc., Seattle, WA, USA). The intra-assay and interassay coefficient of variation (CV) was 6% and 5% respectively.

Urinary creatinine

Creatinine was determined in urine samples using a diagnostic kit (Metra Creatinine Assay Kit, catalogue no. 8009, Quidel Corporation, San Diego, CA, USA). The intra-assay and interassay CV was 2% and 3% respectively.

Serum intact parathyroid hormone

Serum intact parathyroid hormone (iPTH) concentrations were measured in serum using an ELISA (OCTEIA Intact Parathyroid Hormone, Immuno Diagnostic Systems, Ltd, Boldon, UK). The intra-assay and interassay CV were 3% and 4% respectively. Based on the manufacturer's instruction the suggested normal range for PTH is 0.8–3.9 pm, while values between 4.1 and 29.0 pm are suggestive of primary hyperparathyroidism.

Serum 25-hydroxyvitamin D

Serum 25 (OH) D levels were measured in serum samples using a recently developed ELISA (OCTEIA 25-Hydroxy Vitamin D, Immuno Diagnostic Systems, Ltd). The intra-assay and interassay CV were 6% and 7% respectively. The quality and accuracy of our serum 25 (OH) D analysis is assured on an ongoing basis by participation in the Vitamin D External Quality Assessment Scheme (DEQAS, Charing Cross Hospital, London, UK). There is no international consensus on cut-off levels for vitamin D deficiency/insufficiency.24, 25 In the present study, we adopted the definitions proposed by Vieth26 (>40 nm, adequate; 25–40 nm, marginally deficient; <25 nm, severely deficient).

Serum total and undercarboxylated osteocalcin

Total osteocalcin concentrations were measured in serum samples using an ELISA (Metra Osteocalcin EIA Kit, Quidel Corporation). The intra-assay and inter-assay CV was 6% and 8% respectively. Undercarboxylated osteocalcin (Glu) concentrations were measured in serum samples using a recently developed ELISA [undercarboxylated (Glu-OC) EIA kit, Takara Biomedical Group, Otsu, Shia, Japan]. The intra-assay and interassay CV for Glu was 4% and 8% respectively.

Serum bone-specific alkaline phosphatase

Bone-specific alkaline phosphatase activity was measured in serum samples using an ELISA (Metra Osteocalcin EIA Kit, Quidel Corporation). The intra-assay and interassay CV was 5% and 6% respectively.

Statistical analysis

Data are presented as mean values and standard deviations. Student's unpaired t-test was used when appropriate. When the variances were unequal or the distribution was not normal, the Mann–Whitney test was used. Two-tailed tests of significance were used in all statistical analysis. A P-value of <0.05 was considered statistically significant. Differences in age, height, weight and BMI between the patients (CD and UC, separately) and healthy control subjects were examined by unpaired Student's t-tests. Differences in markers of vitamin K and vitamin D status and bone turnover between the patients (CD and UC, separately) and healthy control subjects were examined by unpaired Student's t-tests or Mann–Whitney tests. These comparisons were performed in all patients (as the approach used in some studies);20 as well as in only those patients not taking vitamin D and/or calcium supplements (an approach used in other studies).6 Differences in markers of vitamin K status and bone turnover between the CD and UC patients were examined by unpaired Student's t-tests or Mann–Whitney tests. Differences in proportion of CD and UC patients, separately, and control subjects that were vitamin D-deficient were assessed by Fisher's exact tests. Correlations between markers of bone turnover and C-reactive protein (and haemoglobin in UC patients only) were assessed using Pearson's correlation coefficients. The Statistical Package for the Social Sciences (spss for Windows Version 10.0, SPSS Inc., Chicago, IL, USA) and Data Desk 6.0 for Mac OS (Data Description Inc., New York, NY, USA) were used for statistical analysis.

Results

There were no significant differences between CD patients and age- and sex-matched healthy control subjects for age, weight, height or BMI (Table 1). Similarly, there were no significant differences between UC patients and age- and sex-matched healthy control subjects for age, weight, height or BMI (Table 2). Of the group of CD patients (and of the healthy control subjects, as these were matched to the same sampling month as the age- and sex-matched CD patients), 69%, 12%, 12% and 7% were sampled during autumn, winter, spring and summer respectively. Of the group of UC patients (and of the healthy control subjects, as these were matched to the same sampling month as the age- and sex-matched UC patients), 19%, 25%, 25% and 31% were sampled during autumn, winter, spring and summer respectively.

On exclusion of supplement users, there were no significant differences between CD (n = 47) and UC patients (n = 26) and their respective age- and sex-matched healthy control subjects for age, weight, height or BMI (data not shown). Serum PTH concentration was similar in CD patients and age- and sex-matched healthy control subjects (Table 3). Patients with CD had significantly (P < 0.05) higher serum Glu and bone-specific alkaline phosphatase concentrations compared with those of age- and sex-matched healthy control subjects (Table 3). Patients with CD had significantly (P < 0.0002) higher urinary NTx concentrations compared with those of healthy control subjects. Serum total osteocalcin (s-OC) concentration was significantly (P < 0.01) lower in CD patients compared with that of healthy control subjects.

Table 3.  Serum 25 (OH) D, Glu and parathyroid hormone and markers of bone turnover in CD patients and age- and sex-matched control subjects‡
 CD patients (n = 47)Control subjects (n = 47)
Means.d.Means.d.
  1. 25 (OH) D, 25-hydroxyvitamin D; PTH, parathyroid hormone; Glu, undercarboxylated osteocalcin; NTx, Type I collagen cross-linked N-telopeptides; BCE, bone collagen equivalents; Cr, creatinine; CD, Crohn's disease.

  2. Mean values significantly different from that of control subjects using unpaired Student's t-test: * P < 0.05, ** P < 0.01, *** P < 0.0002.

  3. † n = 32 for serum PTH in CD patients.

  4. ‡ Including only those subjects not taking calcium- and/or vitamin D-containing supplements.

Serum
 25 (OH) D (nm)71.6***33.011369.2
 PTH (pm)†2.11.31.80.80
 Glu (ng/mL)5.1*3.44.02.0
 Osteocalcin (ng/mL)9.5**5.011.84.0
 Bone-specific alkaline phosphatase (U/L)24.6*7.221.36.1
Urine
 NTx (nmol BCE/mmol Cr)49.4***26.526.420.2

Serum 25 (OH) D concentration was significantly (P < 0.0002) lower in CD patients compared with that of healthy control subjects (Table 3). Six CD patients and two controls were marginally vitamin D-deficient [serum 25 (OH) D: >25 but <40 nm], whereas three of the CD patients and none of the control subjects had severe vitamin D deficiency [serum 25 (OH) D: <25 nm]. Therefore, nine and two CD patients and matched controls, respectively, had inadequate vitamin D status [serum 25 (OH) D: <40 nm], a difference that was significant (P ≤ 0.05).

Patients with UC had significantly (P < 0.05) higher serum Glu and bone-specific alkaline phosphatase concentrations compared with those of age- and sex-matched healthy control subjects (Table 4). Patients with UC had significantly (P < 0.001) higher urinary NTx concentrations compared with those of healthy control subjects (Table 4). Serum total osteocalcin concentration was significantly (P < 0.0002) lower in UC patients compared with that of healthy control subjects (Table 4).

Table 4.  Serum 25 (OH) D, Glu and markers of bone turnover in UC patients and age- and sex-matched control subjects†
 UC patients (n = 26)Control subjects (n = 26)
Means.d.Means.d.
  1. 25 (OH) D, 25-hydroxyvitamin D; Glu, undercarboxylated osteocalcin; NTx, Type I collagen cross-linked N-telopeptides; BCE, bone collagen equivalents; Cr, creatinine; UC, ulcerative colitis.

  2. Mean values significantly different from that of control subjects using unpaired Student's t-test, * P < 0.05, ** P < 0.01, *** P < 0.001; **** P < 0.0002.

  3. † Including only those subjects not taking calcium- and/or vitamin D-containing supplements.

Serum
 25 (OH) D (nm)63.9****20.510950.8
 Glu (ng/mL)5.7*3.93.52.2
 Osteocalcin (ng/mL)6.3****2.310.83.4
 Bone-specific alkaline phosphatase (U/L)26.7*7.122.18.5
Urine
 NTx (nmol BCE/mmol Cr)51.2***31.624.219.1

Serum 25 (OH) D concentration was significantly (P < 0.0002) lower in UC patients compared with that of healthy control subjects (Table 4). Only one UC patient was marginally vitamin D-deficient and another UC patient had severe vitamin D deficiency, whereas none of the control subjects had vitamin D deficiency (severe or marginal). Therefore, while two UC patients had inadequate vitamin D status [serum 25 (OH) D: <40 nm], none of the matched controls had inadequate vitamin D status; however, this difference was not significant (P = 0.49).

These significant differences in bone marker levels and indices of vitamin D and K status between CD and UC patients and their respective controls were also evident when the statistical analysis was re-performed after inclusion of data from those patients taking vitamin D and/or calcium supplements (i.e. using data from all 68 CD and 32 UC patients; data not shown).

While the two groups of patients were not matched by age or gender, serum Glu and bone-specific alkaline phosphatase as well as urinary NTx concentrations were similar (P > 0.05) in CD and UC patients. Patients with UC had significantly (P < 0.01) lower total osteocalcin concentrations compared with that of CD patients.

Gender, age, serum Glu and serum 25 (OH) D were included in a multiple linear regression analysis for the combined group of IBD patients (n = 100). Serum Glu correlated significantly with urinary NTx (P = 0.007) and serum bone-specific alkaline phosphatase (P = 0.041; Table 5). No correlations were found between serum 25 (OH) D and urinary NTx or serum bone-specific alkaline phosphatase (Table 5).

Table 5.  Results of multiple linear regression analysis in patients with IBD (n = 100), with urinary Type I collagen cross-linked N-telopeptides and serum bone-specific alkaline phosphatase as dependent, and serum Glu, serum vitamin D, age and gender as independent variables
 Urinary NTx levelsSerum bone-specific alkaline phosphatase levels
Coefficient95% CIP-valueCoefficient95% CIP-value
  1. CI, confidence interval for coefficient; NTx, Type I collagen cross-linked N-telopeptides; Glu, undercarboxylated osteocalcin; 25 (OH) D, 25 hydroxyvitamin D.

Serum Glu5.41.6–9.30.0070.490.02–0.950.041
Serum 25 (OH) D−0.09−0.47 to 0.300.650−0.02−0.07 to 0.030.430
Gender−0.39−24.4 to 23.60.974−2.6−5.5 to 0.330.108
Age−0.46−1.6 to 0.640.400−0.05−0.19 to 0.080.413

Bone biomarker levels were not correlated with C-reactive protein levels in CD or UC patients, or between bone biomarker levels and haemoglobin levels in UC patients (data not shown).

Discussion

Inadequate vitamin D status has been suggested as an aetiological factor for IBD-associated osteopenia.10, 11 In the present study, both CD and UC patients, separately, had significantly lower serum 25 (OH) D levels compared with respective healthy-matched control subjects. These findings are in line with those of several, but not all,21, 27–29 studies which show vitamin D status to be lower in CD and UC patients than control subjects or a healthy population reference range.11, 16, 19, 29–32 Several reasons have been suggested for the lower vitamin D status of IBD patients, including a reduced efficiency of intestinal absorption of vitamin D, as a consequence of ileopathy,12, 33 a disrupted enterohepatic circulation of vitamin D,10 renal insufficiency,34 reduced dietary intake and/or exposure to sunshine,35 although these were not determined in the present study.

Even though the patients had lower vitamin D status than sex- and age-matched controls, their mean serum 25 (OH) D levels were, on average, relatively good. For example, much lower mean serum 25 (OH) D levels and/or higher prevalence of low vitamin D status among adult IBD patients have been reported compared with that observed in the present study.11–13, 29–31, 36 It must be emphasized; however, that vitamin D status is strongly influenced by geographical differences,37 making between-country comparisons difficult to interrupt. While our laboratory has recently reported lower vitamin D status in CD patients compared with age- and sex-matched controls during summer and winter,38 this is the first study, to our knowledge, which has investigated the vitamin D status of UC patients in Ireland, despite our relatively northerly latitude (51–55°N). In northern latitudes of 40–60°N, sunlight is not of sufficient intensity during the winter months (October–March) to stimulate cutaneous synthesis of vitamin D.24

The prevalence of vitamin K deficiency in chronic gastrointestional disorders has been known for some time.39 In the present study, both CD and UC patients had significantly higher serum Glu levels, a marker of vitamin K nutritive status,40 compared with age- and sex-matched control subjects. The reasons for the lower vitamin K status in IBD patients are unclear and warrant further investigation. While IBD patients are often treated for their disease with antibiotics that could kill vitamin K-producing flora, none of the patients in the present study was receiving broad-spectrum antibiotics. It is possible that patients with predominately colonic CD and UC may be more likely to have altered bacterial flora and produce less vitamin K (vitamin K2). However, vitamin K2 produced by colonic microflora is probably poorly absorbed, and therefore, bacterial synthesis of vitamin K2 only represents a minor source of vitamin K in human nutrition (see review by Institute of Medicine41). There may have been malabsorption of this fat-soluble vitamin in some patients, as a consequence of ileopathy. While it could also be argued that the lower vitamin K status in IBD patients merely reflects overall poor nutritional status, which itself could have implications for bone turnover; none of the patients in the present study had a BMI <19 kg/m2, suggesting none of the patients was malnourished.6

In recent years, there has been mounting evidence for a role of suboptimal vitamin K intake and/or status in the pathogenesis of low BMD and osteoporotic fracture in healthy adult populations, especially the elderly (see reviews by Szulc and Meunier42 and Weber43). Low vitamin K status may also be a causative factor in CD-associated osteopenia. For example, Schoon et al.13 provided evidence of low serum vitamin K1 concentrations in patients with long-standing CD (who were in remission and receiving no or very low doses of steroids at the time of the study), together with increased concentrations of Glu, which is also a predictor of hip fracture risk.44 Similarly, our laboratory has recently reported that Irish patients with long-standing CD (who were in remission and receiving no or very low doses of steroids at the time of the study) had increased concentrations of serum Glu.12 Moreover, the increased concentrations of Glu in CD patients in these studies appeared to be positively associated with the rate of bone turnover12 and inversely correlated with BMD at the lumbar spine.13 There has been no previous study which investigated the role of vitamin K status on bone health in UC. In the present study, serum Glu was independently correlated with the rates of bone resorption and bone formation, and thus the rate of bone turnover, as assessed using sensitive and specific biochemical markers, in the entire group of IBD patients, as well as in the UC patients on their own.

In the present study, mean urinary NTx concentrations (a specific and sensitive marker of bone resorption45) in both CD and UC patients, separately, were almost double those in age- and sex-matched control subjects, whereas serum concentrations of markers of bone formation were either higher (bone-specific alkaline phosphatase) or lower (total osteocalcin) than those of controls. While some studies of IBD patients have reported increased levels of bone resorptive markers,12, 15–19, 21, 22 other studies have reported no difference in resorptive markers17, 20, 23 between IBD and control subjects. However, in addition to major differences in the various IBD patient populations, as mentioned already, the type of resorptive markers differed in many of these studies, and even within a study, resorptive markers may show differential responses.17

The only other study that used urinary NTX as a marker of bone resorption in IBD also found elevated levels compared with those in healthy controls.19 Dresner-Pollak et al.46 recently showed, using quartile analysis that IBD patients with the highest urinary NTx concentrations experienced the greatest decrease in spinal BMD over a 2-year period compared to patients with the lowest NTx concentrations.

The literature is mixed on whether serum osteocalcin and/or bone-specific alkaline phosphatase levels are elevated, reduced or similar in IBD patients and healthy controls.12, 15–23 Even in the two studies which examined the levels of bone formation markers in patients with quiescent CD, and not taking steroids, the results have been conflicting.20, 23 The reasons for the discordant findings between the two markers of bone formation in the CD and UC patient groups compared with respective control groups are unclear. Ardizzone et al. also found differential responses of serum osteocalcin and bone-specific alkaline phosphatase in UC patients.21 It is possible that one marker is more sensitive than the other. Should serum bone-specific alkaline phosphatase be the more sensitive marker of bone formation, then the present findings would suggest that the processes of bone resorption and bone formation are still coupled in IBD, and thus, the patients have an increased rate of bone turnover. An increased rate of bone turnover in older adults contributes to faster bone loss and is recognized as a risk factor for fracture.47 In addition, high rates of bone turnover are associated with a disruption of the trabecular network leading to loss of connectivity that is not necessarily reflected in a decrease of bone mass.48 On the other hand, should serum osteocalcin be the more sensitive marker of bone formation, then the present findings would suggest that the processes of bone resorption and bone formation are uncoupled in IBD, and that bone loss associated with IBD may be occurring by unbalanced bone metabolism. This uncoupled bone metabolism can be a risk factor for progressive loss of bone in IBD patients, even during the quiescent phase of disease.20While the impact of alterations in bone turnover markers on fracture risk has not been studied in IBD patients, as mentioned already there is evidence that increased levels of markers of bone turnover in IBD patients are associated with bone loss,46 which is recognized as a risk factor for fracture, at least in postmenopausal women.47

In conclusion, the altered rates of bone metabolism observed in IBD patients in the present study may explain, at least in part, the osteopenia and osteoporosis which are common among IBD patients.3, 4 While there may be several factors underpinning the altered rates of bone metabolism in these patients, two of the principal reasons that have been put forth as causative factors for the osteopenia in IBD, namely, high-dosage steroid use9, 49 and an active inflammation process,6, 46 were not factors in the patients in the present study. Further research is needed to explore whether bone metabolism in IBD patients, currently in remission, is coupled or uncoupled as this may be of importance for future preventive and therapeutic strategies. In addition, the present study provides further evidence that IBD patients had significantly lower status of vitamin D and K compared with age- and sex-matched control subjects. Moreover, the rates of bone resorption and bone formation (and thus bone turnover) were inversely correlated with vitamin K status, but not vitamin D status, in these patients. Thus, further research is needed to investigate whether improving vitamin K status could beneficially influence bone turnover and bone mass in IBD patients.

Acknowledgements

This study was supported by funding made available under the National Development Plan 2000–2006 with assistance from the European Regional Development Fund.

Ancillary