Introduction

Osteopenia and osteoporosis are considered to be extra-intestinal manifestations of inflammatory bowel disease (IBD) [1]. Accordingly, the relative risk of fracture in IBD patients has been estimated to be 40 % higher than in the general population [2]. However, there is still no widely accepted consensus on the best evaluation method for the determination of osteoporosis in patients with IBD. Similarly, there is no established strategy for preventing or treating osteoporosis in IBD patients.

Osteoporosis is usually diagnosed by dual-energy X-ray absorptiometry (DEXA) scanning to quantify bone mineral density (BMD) [3]. BMD results are expressed as the number of standard deviations (SD) above or below the mean for a young adult population (T-score) or an age-matched population (Z-score) [4, 5]. DEXA scanning is a convenient noninvasive method for the evaluation of osteoporosis, but woman during pregnancy, or those expected to become pregnant, should not receive DEXA examination for the fear of increasing the incidence of congenital defects. However, it is possible to assess the state of bone metabolism by using biochemical bone metabolism markers in addition to BMD. Bone-specific alkaline phosphatase (BAP) is a marker of bone formation [6], while N-terminal telopeptide of type I collagen (NTx) is a biochemical marker of bone resorption [7]. Further, undercarboxylated osteocalcin (ucOC) has been shown to be a sensitive biomarker of vitamin K deficiency [8], while blood 1,25-(OH)2D concentration, which is an active form of vitamin D, has been found to be negatively correlated with BMD in patients with Crohn’s disease (CD) [9]. However, the role of these bone metabolism markers in the maintenance of bone health in IBD patients is not fully understood.

Recent studies have reported contradictory findings on osteoporosis and fractures in patients with CD and ulcerative colitis (UC) [10–15]. It would appear that osteoporosis in patients with CD is a multifactorial process. A greater risk of osteoporosis in CD than in UC might be attributed to the presence of small intestinal CD lesions causing malnutrition and impaired vitamin D or calcium absorption in CD patients [16]. Additionally, anti-tumor necrosis factor (TNF)-α biologics are suspected to increase the risk of osteoporosis [17]. However, currently there are no biochemical data to explain the mechanism by which TNF blockade leads to low BMD. This study was undertaken to investigate the biochemical basis of infliximab (IFX)-induced low bone density and increased turnover in patients with IBD.

Methods

Objectives

In a prospective setting, the objective of this study was to investigate the frequency of bone loss or abnormality of bone mineral markers among patients with IBD and to identify possible contributing risk factors. However, our major interest was to see whether IFX which is widely used in the treatment of patients with IBD affects bone turnover or other relevant bone turnover markers.

Study Populations

Between 2007 and 2011, 80 patients with UC (52 men, 28 women) and 40 patients with CD (27 men, 13 women) with a diagnosis of IBD [18] were included in this study. All patients had come to our IBD units for the treatment of IBD-related complications. We included patients between 25 and 60 years of age and excluded women who were pregnant or were hoping to become pregnant due to the potential risk from exposure to ionizing radiation by bone densitometry. Sixty-five age- and gender-matched healthy controls (33 men, 32 women) were also included (Table 1).

Table 1 Demographic features of patients with inflammatory bowel disease (IBD), age- and gender-matched controls
Full size table

Measurement of Bone Mineral Density

BMD was measured at the lumbar spine (L2–L4) and femoral neck with dual-energy X-ray absorptiometry (Discovery A, Hologic, Bedford, MA, USA). BMD was expressed in absolute values (g/cm2), T-scores (one SD of the mean of a young adult gender-matched reference population), and Z-scores (one SD of the mean of an age- and gender-matched reference population).

Biochemical Measurements

Blood samples were centrifuged at 3000 rpm for 10 min at 4 °C. Serum samples were stored frozen at −20 °C until analysis. Serum undercarboxylated osteocalcin (ucOC) was measured by an electrochemiluminescence immunoassay (Sanko, Tokyo, Japan) for ucOC. Serum 1,25-(OH)2D was measured by a radioimmunoassay (Immunodiagnostic Systems, Boldon, UK). Serum N-terminal telopeptide of type I collagen (NTx), a marker of bone resorption, was measured by enzyme-linked immunosorbent assay (Ostex, Seattle, USA). Bone-specific alkaline phosphatase (BAP), a marker of bone formation, was measured by an enzyme immunoassay (Quidel, San Diego, CA, USA). All laboratory measurements were blindly carried out at an SRL (special research laboratory).

Ethical Considerations

Informed consent was obtained from all individual participants included in the study after explaining the nature of the procedures involved. Similarly, the study protocol was reviewed and approved by the Institutional Review Board at Hamamatsu University School of Medicine. The investigation was conducted in accordance with the Principle of Good Clinical Practice, and the ethical standards laid down in the 1964 Helsinki Declaration and its subsequent amendments.

Statistics

When appropriate, parametric data are presented as the mean ± SD values. Continuous data are compared by using a two-tailed Fisher’s exact test, unless stated otherwise. Correlations between two independent measurements were assessed by using the Pearson’s correlation coefficient. A P value of less than 0.05 was considered statistically significant.

Results

Patients’ Demographic Variables

The baseline demographic variables and laboratory data on the study populations are presented in Table 1. In patients who had received IFX (Tables 2 and 3), the median duration of IFX therapy was 15 months, range 6–24 months. Data presented in Table 1 show that when we analyzed BMD, the mean T-score of the femoral neck in CD and UC patients was significantly lower than in the controls (P < 0.05). Likewise, the mean Z-score of the femoral neck was significantly lower in patients with CD as compared to controls (P < 0.05) or UC patients (P < 0.05). Bone metabolism markers including 1,25-(OH)2D, bone-specific alkaline phosphatase (BAP), N-terminal telopeptide of type I collagen (NTx), and undercarboxylated osteocalcin (ucOC) were significantly higher in patients with CD, but not in patients with UC as compared to the controls (P < 0.05; Table 1). Similarly, the mean level of serum ucOC was significantly higher in patients with CD compared with patients with UC (P < 0.05).

Table 2 Baseline demographic features of patients with Crohn’s disease (CD) who received or did not receive infliximab (IFX) therapy
Full size table
Table 3 Measurement outcomes of bone mineral density (BMD), T-score, Z-score, and other biochemical markers of bone formation and resorption in Crohn’s disease (CD) patients who received and those who did not receive infliximab (IFX) therapy
Full size table

IFX Therapy Was Associated with Increased Serum Level of NTx in Patient with CD

We were particularly interested to determine whether or not IFX therapy was associated with changes in BMD or bone metabolism markers in patients with CD. There was no significant difference in baseline demographic variables including the length of disease, disease location, previous CD surgery, corticosteroid use, or osteoprotective medications between patients with CD who were receiving IFX and CD patients who were not (Table 2). Likewise, the mean levels of BMD, T-score, and Z-score were not significantly lower in patients who were receiving IFX and those who were not (Tables 3, 4). Unexpectedly, there was no significant difference in the mean levels of T-score and Z-score between healthy controls and patients with CD who were taking corticosteroid (Fig. 1). Further, Fig. 2 shows that serum level of ucOC was significantly higher in CD patients as compared to controls regardless of IFX therapy, but the significance level was stronger in CD patients on IFX (P < 0.01) compared with CD patients not on IFX (P < 0.05). However, in patients with CD, the mean serum level of NTx (index of bone resorption) was significantly higher in patients who were receiving IFX compared with patients who were not (P < 0.01).

Table 4 Measurement outcomes on bone mineral density (BMD), T-score, and Z-score in Crohn’s disease (CD) patients who received and those who did not receive infliximab (IFX) therapy
Full size table
Fig. 1
figure 1

Serum levels of T-score and Z-score in patients with CD and healthy controls. Levels of T-score (a) and Z-score (b) in patients with CD who were not receiving IFX (n = 26), patients with CD who were receiving IFX (n = 14), and healthy controls (n = 65). Levels of T-score (c) and Z-score (d) in patients with CD who were not receiving corticosteroid (n = 26), patients with CD who were receiving corticosteroid (n = 14), and healthy controls (n = 65). Results represent the mean ± SD values. *P < 0.05; **P < 0.01

Full size image
Fig. 2
figure 2

Serum levels of ucOC, BAP, 1,25-(OH)2D3, and NTx in patients with CD who were not (n = 26) or were (n = 14) receiving IFX, and healthy controls (n = 65). Results represent the mean ± SD values. *P < 0.05; **P < 0.01

Full size image

The Relation Between Serum NTx and T-Score in Patients with CD

We were interested to see whether or not the serum level of NTx was correlated with BMD in healthy controls and patients with CD. The regression lines using T-score as outcome variable (y) and serum NTx as predictor variable (x) were y = 0.048x − 0.771 and y = −0.037x − 0.195 for controls and CD patients, respectively. The regression coefficient for controls was significantly greater than that for CD patients (t = 3.55, P < 0.01). There was no statistically significant correlation between serum NTx and T-score in CD patients regardless of IFX therapy, except a greater tendency toward negative correlation between NTx and T-score in CD patients who were on IFX (Fig. 3). Other bone metabolism markers including 1,25-(OH)2D, BAP, and ucOC were correlated with BMD in controls and patients with CD. A statistically significant (P < 0.05) positive correlation (r = 0.275) was observed between serum ucOC level and BMD in healthy controls, but other bone metabolism markers were not associated with BMD in healthy controls (data not presented).

Fig. 3
figure 3

Correlation between T-score and serum NTx level in patients with CD and healthy controls. a Correlation between T-score and serum NTx in healthy controls (r = 0.17, P > 0.05) and patients with CD (r = 0.22, P > 0.05). The regression lines using T-score as the outcome variable (y) and NTx as the predictor variable (x) were y = 0.048x − 0.771 and y = −0.037x − 0.195 for the controls and CD patients, respectively. The regression coefficient for the controls was significantly greater than for CD patients (t = 3.55, P < 0.01). b The correlation between T-score and serum NTx level in patients with CD who were not receiving IFX (r = 0.17, P > 0.05) and patients with CD who were receiving IFX (r = 0.35, P > 0.05). The regression lines using T-score as the outcome variable (y) and NTx as the predictor variable (x) were y = 0.034x − 0.328 and y = −0.063x + 0.437 for patients with CD who were not receiving IFX and patients with CD who were receiving IFX, respectively. There was no significant difference in the regression coefficient between patients with CD who were not receiving IFX and patients with CD who were receiving IFX (t = −1.01, P > 0.05)

Full size image

Discussion

From the observations described in this report, we could make the following statements on bone density, bone formation, and the related biochemical markers of bone turnover in patients with IBD. Compared with the controls, a lower level of BMD and a higher level of bone metabolism markers were seen in CD patients as compared to UC patients. The serum level of NTx was significantly higher in CD patients who were receiving IFX as compared to CD patients who were not receiving IFX, and this was the single most important finding of the present study and indicated that CD patients who are required for IFX therapy have an increased risk of bone resorption. There was a tendency toward negative correlation between the serum level of NTx and T-scores in CD patients, in particular those who were receiving IFX.

In this study, the mean Z-score of the femoral neck was significantly lower in patients with CD not only when compared with the controls, but also when compared with UC patients. Other studies have also shown that patients with CD are at an increased risk of low BMD compared with UC patients [10–13]. The presence of small intestinal CD lesions, which is common in CD patients, but not in UC patients, potentially can interfere with the absorption of vitamins and minerals, and contribute to the differences in BMD between CD and UC patients [16, 18]. Additionally, patients with CD are at a higher risk of developing intestinal strictures and often have abdominal discomfort as a primary symptom, which may lead to food avoidance and malnutrition.

In addition to measuring BMD, we assessed bone turnover markers in patients with IBD and controls. We found that the mean serum levels of 1,25-(OH)2D, BAP, NTx, and ucOC were significantly higher in patients with CD compared with the controls. These observations are in line with several recent studies. Abreu et al. [19] found that patients with CD have inappropriately elevated levels of 1,25-(OH)2D which is the active hormonal form of vitamin D and that the level of 1,25-(OH)2D correlated inversely with bone mineral density presumably because 1,25-(OH)2D mobilizes skeletal calcium stores. Nakajima et al. [20] reported that relative to healthy controls, serum level of ucOC, which is a sensitive biomarker of vitamin K deficiency, was significantly higher in patients with CD, but not in patients with UC, indicating that vitamin K was deficient in patients with CD. Gilman et al. [21] reported that compared with healthy controls, patients with CD had significantly higher concentrations of serum BAP (a sensitive marker of bone formation), and significantly higher concentrations of urinary NTx, a sensitive marker of bone resorption. These findings suggest that both the rates of bone resorption and bone formation are significantly increased in CD, reflecting an increased rate of bone turnover. An increased rate of bone turnover in elderly contributes to faster bone loss and is recognized as a risk factor for bone fracture [22].

We also were interested to see whether or not IFX affects the levels of bone metabolic markers in patients with CD. We found a significant increase in the level of serum NTx in patients with CD who were receiving IFX compared with patients who were not receiving IFX, indicating increased bone turnover associated with IFX therapy. At the time of this writing, we could not find any report describing the mechanism(s) associated with elevation of NTx in patients with Crohn’s disease. Potentially, this could be due to an excess collagen production or altered MMP–TIMP balance. However, an increased level of NTx is widely thought to indicate bone resorption and greater bone loss [23]. In other words, patients who have less bone density, and a high NTx level, may be at an increased risk of bone fracture [24, 25]. This risk is likely to be significantly higher in patients with IBD due to multiple medications they receive including biologics [17, 21]. Therefore, we believe that it is clinically relevant to measure NTx in a potentially at-risk population.

In this study, the T-score showed a declining tendency during a 6-month observation period in patients with CD who were receiving IFX (data not presented). The reason might be that these patients had more severe disease, which requires more aggressive treatment. A recent study show that the osteoclastogenic function of the Th17 TNFα+ cells that contribute to bone loss in IBD animal model [26]. Contrary to our findings, a number of studies have reported beneficial effect for IFX on bone turnover markers in patients with CD in the short term [27–31]. We believe that controlled long-term studies in large cohorts of patients are warranted to better understand the effects of IFX on osteoporosis.

The effect of corticosteroids on osteoporosis seen in patients with IBD is complex. While some studies have shown a clear relationship between lifetime corticosteroid dose and vertebral fracture rate or low BMD [32], other studies have suggested that BMD is unrelated to corticosteroid therapy [33–36]. In our study, BMD was significantly lower in patients with CD as compared to the controls, but not in patients with CD who were receiving corticosteroids. The conflicting results are probably a reflection of the heterogeneity of patient groups studied and the complex relationship between disease severity, systemic inflammation, and corticosteroid therapy.

A recent meta-analysis shows that bisphosphonate is effective for the treatment of low BMD patients with IBD and reduces the risk of vertebral fractures [37]. Our study indicates that CD patients receiving IFX considered having an increased risk of bone fracture because of low BMD and an increased bone resorption. Therefore, administration of bisphosphonate to these patients to reduce a risk of bone fracture should be considered.

In conclusion, we found that both bone resorption and bone formation were increased in patients with CD, but not in patients with UC. Furthermore, the serum level of NTx, which is a sensitive marker of bone resorption, was significantly increased in CD patients who were receiving IFX, indicating increased bone turnover associated with IFX therapy. Additionally, since there was a tendency toward negative correlation between NTx and T-score in CD patients who were receiving IFX, the risk of osteoporosis in patients with CD during IFX therapy may increase.