Article Text


Gut, inflammation and osteoporosis: basic and clinical concepts
  1. H Tilg1,
  2. A R Moschen1,
  3. A Kaser1,
  4. A Pines2,
  5. I Dotan3
  1. 1 Christian Doppler Research Laboratory for Gut Inflammation and Department of Medicine, Division of Gastroenterology and Hepatology, Medical University Innsbruck, Austria
  2. 2 Department of Medicine “T”, Tel Aviv Sourasky Medical Center, Tel Aviv University, Tel Aviv, Israel
  3. 3 IBD Center, Department of Gastroenterology and Liver Diseases, Tel Aviv Sourasky Medical Center, Tel Aviv University, Tel Aviv, Israel
  1. Dr Herbert Tilg, Christian Doppler Research Laboratory for Gut Inflammation and Division of Gastroenterology and Hepatology, Medical University Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria; herbert.tilg{at}


Chronic inflammatory disorders such as inflammatory bowel diseases (IBD) affect bone metabolism and are frequently associated with the presence of osteoporosis. Bone loss is regulated by various mediators of the immune system such as the pro-inflammatory cytokines tumour necrosis factor-alpha (TNF-α), interleukin-1beta (IL-1β), IL-6, or interferon-gamma. TNF-α, a master cytokine in human IBD, causes bone erosions in experimental models and these effects are exerted by osteoclasts. Other TNF-related cytokines such as receptor activator of nuclear factor kappa B (RANK), its ligand, RANKL, and osteoprotegerin are important mediators in inflammatory processes in the gut and are critically involved in the pathophysiology of bone loss. The awareness and early diagnosis of osteoporosis in states of chronic inflammation, together with applied therapies such as bisphosphonates, may be beneficial in inflammation-associated osteoporosis. Although several mechanisms may contribute to osteoporosis in patients with IBD and coeliac disease, inflammation as an important factor has so far been neglected. As key inflammatory mediators in IBD such as TNF-α are involved in the disease process both in gut and bone, we hypothesise that neutralisation of TNF-α could prove an efficient strategy in the treatment of inflammation-related osteoporosis in the future.

View Full Text

Statistics from

Bone remodelling is an active, progressive process, involving several cell populations, soluble mediators, ligands and receptors (see below). During adulthood, bone mass, represented by bone mineral density (BMD), the major determinant of bone strength, increases, reaches a peak around the second to third decades and then gradually declines. Women suffer more bone loss compared to men, and this is true especially after the menopause when 5–15% of bone loss occurs within the first 5 years. According to WHO, osteoporosis is defined as “a systemic skeletal disease characterised by low bone mass and microarchitectural deterioration of bone tissue, with a consequent increase in bone fragility and susceptibility to fracture”.1 ,2 It occurs if bone resorption exceeds bone formation. In “low-turnover” osteoporosis, bone resorption is normal whereas synthesis of bone matrix is reduced. In contrast, in “high-turnover” osteoporosis, activity of osteoclasts is increased resulting in increased bone resorption. Osteoporosis associated with chronic inflammation usually follows the high-turnover pattern whereas corticosteroid-induced osteoporosis is usually of the low-turnover pattern.3 ,4

Practically, BMD measurements are used in order to quantitatively assess osteoporosis. Thus, the T-score is the number of SDs above or below the average BMD value for young healthy adults (the current recommended reference database is derived from NHANES) (adults aged 20–29 years) and the Z-score is the number of SDs above or below the average BMD for age-matched controls. According to WHO definitions, osteoporosis is present when the T-score is below −2.5 SDs, and osteopenia is present when it is between −1.0 and −2.5.1 ,2 ,5

Disorders of the liver and the gastrointestinal tract, in particular chronic inflammatory processes like inflammatory bowel disease (IBD) and coeliac disease, are commonly associated with osteopenia and osteoporosis.6 ,7 The pathogenesis of osteoporosis in IBD patients is incompletely understood and most likely multifactorial. Whereas there is increasing evidence that inflammation per se may contribute to osteoporosis,8 ,9 several other factors have an important effect.10 These factors include age, corticosteroid use, malnutrition, vitamin D and calcium deficiency, and immobilisation. Of particular relevance in these diseases is the role of corticosteroids on bone remodelling.11 Corticosteroids negatively affect bone mass by various mechanisms including impairment of osteoblast function, reduction of intestinal calcium absorption, and increase of renal calcium excretion.12 Together, these mechanisms result in a pattern of low turnover bone loss. Patients treated with corticosteroids therefore have an increased risk for osteoporosis and associated fractures.13 ,14 More importantly, risk for osteoporosis increases early on, hence the worst effect on bone mineral density is within the first few months of treatment.15 ,16 This increased bone resorption observed in the early stages of corticosteroid therapy is mainly due to increased production of receptor activator of nuclear factor kappa B ligand (RANKL) and decreased synthesis of osteoprotegerin.12 Interestingly, low doses of corticosteroids (<10 mg prednisone per day) have only modest effects on bone formation possibly due to its accompanied beneficial effect on inflammatory processes;17 however, it should be stressed that even these low doses were associated with osteoporosis.18 Several key factors are involved in the pathogenesis of osteoporosis in chronic inflammatory disorders. This article will focus on the role of inflammation in the pathogenesis of osteoporosis in chronic inflammatory diseases such as IBD and coeliac disease and suggest clinical algorithms on how to prevent and treat patients with bone loss.


Introduction to bone cells

Osteoblasts are derived from a multipotential mesenchymal cell that can alternatively differentiate also into marrow stromal cells or adipocytes.19 ,20 The signals that are involved for the development of osteoblasts from mesenchymal progenitor cells are not fully understood.8 ,21 Wnt signalling through Wnt10b is involved in the fate decision in differentiating mesenchymal progenitor cells between adipocytes and osteoblasts.22 Osteoclasts are multinucleated macrophage-like giant cells that resorb bone.23 They are haematopoietic in origin and derive from myeloid precursors that also give rise to macrophages and dendritic cells. Signals that control osteoclasts to form and resorb bone involve several transcription factors and cytokines which are discussed later in detail (fig 1A).3 ,8 The critical role of nuclear factor kappa B (NFκB) in osteoclast formation has been demonstrated in NFκB1 and NFκB2 knockout mice that showed osteopetrosis caused by a defect in osteoclast development.24 To remove organic components of bone, osteoclasts produce several enzymes such as cathepsin. Osteoclasts are highly motile, move across the bone surface, and resorb large areas of bone.23 They die by apoptosis regulated by several paracrine acting cytokines. Inflammation-related osteoporosis is driven by activation of osteoclasts and therefore this cell type is of critical importance.

Figure 1 (A) Bone remodelling under physiological conditions (homeostasis). Bone remodelling involves several cell types including osteoblasts (OBs) that build up non-mineralised bone matrix, bone marrow stromal cells and bone resorbing osteoclasts (OCs). Osteoblasts are derived from a mesenchymal stem cell, and Wnt10b signalling via frizzled and lipoprotein receptor-related protein (LRP)-5 plays a critical role for the induction of the osteoblastogenic transcription factors Runx2, Dlx5 and osterix. Osteoclasts on the other hand are derived from myeloid precursors under the permissive influence of receptor activator of nuclear factor kappa B (RANK) signalling by activating several transcription factors such as NFκB and AP-1 (c-fos). RANK is activated by RANK ligand (RANKL), a member of the TNF superfamily, which is the key osteoclastogenic cytokine and is mainly produced by bone marrow stromal cells and osteoblasts. RANKL activity is inhibited by osteoprotegerin (OPG) which is a soluble decoy receptor for RANKL. Therefore, bone homeostasis is regulated by the balance of bone resorbing RANKL and bone “protecting” osteoprotegerin. (B) Bone remodelling in chronic inflammation. Activated T cells produce RANKL and its soluble form and therefore directly trigger bone loss by inducing and activating OCs via RANK. The amount of RANKL exceeds the protective capacity of OPG. Moreover, several pro-inflammatory cytokines such IL-1β, IL-6 and TNF-α promote osteoclastogenesis and bone resorption in synergy with RANKL. Interferon gamma (IFN-γ) promotes bone loss indirectly through its capacity to upregulate antigen presentation and T cell activation by antigen-presenting cells.

The osteoprotegerin–RANK axis

The concept that stimulation of bone resorption requires an interaction between cells of the osteoclastic lineages was put forward many years ago, but its molecular mechanisms have been identified only recently.25 Various soluble mediators are involved in inflammation-associated bone loss (see table 1). Osteoclastogenesis (and therefore bone loss) is stimulated by a member of the TNF family called RANKL by its attachment to the RANK receptor on the surface of progenitor cells. RANKL and its soluble form, sRANKL, are produced by activated T cells and also osteoblastic cells. Opposing this is a natural decoy receptor called osteoprotegerin which combines with RANKL (fig 1A and B). Bone homeostasis is thus achieved by the balance of the bone resorbing effect of RANKL and the blocking of RANKL by osteoprotegerin.26 Osteoprotegerin is mainly derived from monocytes/macrophages and dendritic cells and is released under the control of various pro-inflammatory cytokines.27 Ashcroft and colleagues28 have shown that in IL-2-deficient mice which spontaneously develop colitis, bone loss is mediated via activated T cells secreting RANKL. Although transcription as well as plasma levels of the decoy receptor of RANKL, osteoprotegerin, is substantially increased in IL-2-deficient mice, such upregulation of osteoprotegerin does not suffice to ameliorate bone loss which is mediated by activated T cell-derived RANKL. More importantly, exogenous administration of osteoprotegerin to IL-2-deficient mice reverses bone loss, and notably also ameliorates intestinal inflammation. Colitic IL-2−/− mice develop substantial osteopenia with increasing age, with bone loss detectable after the onset of colitis at the age of 4 weeks. Notably, elevated sRANKL serum levels are only observed at an early stage at the onset of colitis and bone loss, and decline rapidly to normal levels by week 9–10. In contrast, osteoprotegerin levels start to substantially increase later, without ever reaching a plateau until the end of the observation period. This model therefore might quite reliably reflect the “human situation” of gut inflammation observed bone loss.

Table 1 Soluble mediators involved in the pathophysiology of inflammation-associated bone loss

As activated T cells are central in human IBD and coeliac disease, it is likely that inflammation, for example via the RANK/RANKL/OPG system, could contribute to bone loss in these diseases. We have recently demonstrated that IBD and chronic liver diseases are associated with alterations in these systems.27 ,29 In IBD patients we observed an increased osteoprotegerin production in colonic explant cultures whereas sRANKL levels from these cultures were low. As in the above-mentioned mouse model, increased osteoprotegerin levels may represent a continuing homeostatic response attempting to reverse established osteopenia and RANKL-driven osteoclastogenesis, thus maintaining normal bone mass. This view is supported by the gradual increase in osteoprotegerin levels in osteopenic and osteoporotic IBD patients in our study. There was a negative correlation between osteoprotegerin plasma levels and femoral neck/lumbar spine bone mineral density in our IBD study suggesting that despite these increased osteoprotegerin levels endogenous RANKL production might not be antagonised sufficiently.

Osteoclasts and the role of TNF-α in bone erosions

Osteoclast precursors and mature osteoclasts are abundant at sites of bone erosions.30 Redlich and colleagues31 have recently shown that osteoclasts are the essential link between synovial inflammation and bone destruction. c-fos, a component of the dimeric transcription factor AP-1, is an important regulator of bone cell proliferation and differentiation.32 In their elegant experiments they crossed transgenic mice that over-expressed human TNF-α and that develop a severe and destructive arthritis with osteoporosis, with c-fos deficient mice, which completely lack osteoclasts.31 These mice developed severe arthritis but were fully protected against bone destruction. The clinical and histological extent of inflammation and matrix metalloproteinase expression were similar in TNF-α transgenic mice versus c-fos−/− TNF-α transgenic mice, indicating that c-fos is not critically involved in TNF-α-mediated joint inflammation. Their data suggested that TNF-α-dependent bone erosion is mediated by osteoclasts and that the absence of osteoclasts alters TNF-α-mediated arthritis from a destructive to a non-destructive arthritis.

Indeed, the whole range of signals essential for osteoclastogenesis is found over-expressed in human rheumatoid arthritis.8 ,33 Apart from the well known presence of IL-1 and TNF-α, osteoclastogenic cytokines M-CSF and RANKL are found in synovial inflammatory tissue.34 Furthermore, synovial T cells and cells of the monocyte/macrophage lineage, possibly serving as osteoclast progenitors, accumulate in the inflammatory lesions of rheumatoid arthritis.34 Several other experimental animal models have revealed an important role for osteoclasts in local bone erosion. Treatment with osteoprotegerin prevents local bone erosion in adjuvant arthritis by blocking the osteoclastogenic properties of activated RANKL-expressing T cells.35 Osteoprotegerin significantly reduces local bone erosion in a TNF-α-driven model of arthritis, which is T cell and B cell independent, suggesting that RANKL expression by immunocompetent cells is not needed for osteoclast-mediated local bone resorption. Further evidence comes from the study of osteoclast-free RANKL−/− mice, which are almost completely protected from bone erosion but not from arthritis.36 As TNF-α is a critical mediator in human IBD, it may lead to bone loss by the mechanisms discussed (fig 1B).

Besides TNF-α, IL-6, a central cytokine of the acute phase response, might be involved in inflammation-related bone loss. Neutralising antibodies to IL-6 decrease the formation of osteoclast-like cells.37 Furthermore, increased circulating IL-6 levels in paediatric IBD patients are negatively correlated with BMD38 and IL-6 gene polymorphisms influence BMD in patients with Crohn’s disease (fig 1B).39

Box 1: Risk factors for osteoporosis in IBD

  • Increasing age1621

  • Female gender121

  • Previous fragility fracture122

  • Hypogonadism121

  • Type of IBD2 ,11 ,45 ,54 ,92 ,121

  • Low BMI121

  • Chronic inflammatory activity10

  • Chronic/recurrent corticosteroid use44

  • Smoking54

  • Family history54 ,123

  • Vitamin D deficiency/insufficiency54 ,123

The central role of T cells and IFN-γ in bone resorption

T cell-produced cytokines play an important role in the bone loss caused by inflammation, infection and oestrogen deficiency (fig 1B).40 Osteoclast renewal is regulated by the key osteoclastogenic cytokines M-CSF and RANKL.3 Some controversy existed about the role of IFN-γ on bone homeostasis. IFN-γ has been shown to directly inhibit RANKL-mediated osteoclastogenesis via induction of TRAF6 degradation.41 However, Gao et al42 have recently demonstrated that the net effect of IFN-γ on bone remodelling is bone loss. It potently stimulates osteoclastogenesis in vivo via antigen-driven T cell activation as a consequence of its capacity to upregulate antigen presentation. Consequently, these authors have shown that this IFN-γ effect is only observed in T cell-replete but not T cell-deficient mice. This is corroborated by experiments showing that T cell-deficient nude mice fail to undergo bone loss following rIFN-γ administration. IFN-γ-induced bone loss was reversed by T cell reconstitution. T cell activation and proliferation leads to subsequent upregulation in the production of the key osteoclastogenic cytokines RANKL and TNF-α.42 In vivo, this indirect pro-osteoclastogenic activity overcomes the direct suppressive activities of IFN-γ on osteoclast precursors, leading to a net bone loss. Furthermore, as the dominant source of IFN-γ is activated T cells as observed in IBD and coeliac disease, a potent amplificatory loop is likely established which further drives up and sustains antigen presentation and T cell activation, thus perpetuating an inflammatory and pro-resorptive environment. Inhibition of IFN-γ thus might represent a novel strategy to reduce inflammation and bone loss in inflammation-associated osteoporosis.43


Patients with IBD have increased bone loss that may result in osteopenia and osteoporosis.4446 In Crohn’s disease the reported prevalence of osteopenia and osteoporosis is 22–55% and 3–57.6%, respectively, for the lumbar spine and femoral neck;4447 and in ulcerative colitis the reported prevalence of osteopenia and osteoporosis is 32–67% and 4–50%.4446 ,48 ,49 More importantly, this results in an increased risk of fracture – symptomatic or asymptomatic – that may be up to 40–60% higher than in the control population.50 ,51 Alternatively described, the relative fracture risk in IBD has been shown to be about 1.3–1.4.52 The risk is greater in Crohn’s disease than in ulcerative colitis5355 (as summarised in Lewis and Scott: Guidelines for osteoporosis in inflammatory bowel disease and coeliac disease).56 Factors considered important in the pathogenesis of osteoporosis and fractures in IBD patients are the inflammatory process itself, age, gender and menopausal status in women, and the use of corticosteroids.50 Other potentially contributing factors are malnutrition and malabsorption, immobilisation, low body mass index (BMI),48 smoking and hypogonadism. Controversy exists regarding the effect of bowel resections on osteoporosis in Crohn’s disease patients, while in ulcerative colitis improvement after restorative proctocolectomy and ileal pouch anal anastomosis has been reported.20 ,21 ,57

In coeliac disease, villous atrophy and associated calcium and vitamin D malabsorption, reduced calcium intake, and secondary hyperparathyroidism are major reasons for the increased rates of osteoporosis and risk for fractures previously reported.5860 Several studies did not show a significant increase in the risk for fractures in patients with coeliac disease.61 ,62 Individuals with coeliac disease may, however, be at increased risk of hip fracture and any fracture type (hazard ratio  =  1.4) as recently described in a large, general, population-based Danish cohort.5 Weight loss, resulting in low BMI and changes in insulin-like growth factor-1 (IGF-1) and leptin may be contributing factors for reduced BMD in coeliac disease,58 which, in contrast to BMD in IBD patients, tends to be low at diagnosis and to improve on a gluten free diet.6371

Several questions exist regarding screening for osteoporosis in inflammatory gastrointestinal diseases. As data and guidelines differ between IBD and coeliac disease, they will be discussed separately. It is also important to notice that inflammation, osteoporosis and its treatment is a multidisciplinary area, where opinions of different authors and societies may slightly differ. This review represents our suggested approach. The recommended literature includes references to summaries and views of European and North American societies.12 ,56 ,63 ,7274

Should all patients with inflammatory gastrointestinal diseases be screened for osteoporosis? When?

IBD patients are at an increased risk for the development of osteoporosis because of the inflammatory disease process itself, use of corticosteroids, as well as other known risk factors for osteoporosis such as age, low BMI, malnutrition, hypogonadism/amenorrhoea, and smoking (box 1). Indeed, high rates of osteopenia and osteoporosis were reported by several groups in IBD patients worldwide, with an increased risk of both vertebral and hip fractures.21 ,4446 On the other hand, when pooling together the major reports on BMD and IBD, the conclusion is that “IBD had only a modest effect on BMD, with a pooled Z-score of −0.5”.72 Who, then, should be screened? Screening, usually with dual-energy x ray absorptiometry (DEXA), should be selectively ordered after a thorough risk assessment as recommended in the AGA Technical Review on Osteoporosis.6 The important risk factors are shown in box 1. It has been recommended6 that if one or more risk factors exist, patients should undergo initial screening with DEXA which should be repeated after 2–3 years or, in the case of corticosteroid treatment, after 1 year of treatment. An alternative is to screen all patients, but to justify that would require further evaluation of cost-effectiveness.

Surprisingly, despite osteoporosis being a well-recognised side effect of corticosteroid treatment and the reported high prevalence of osteopenia, osteoporosis and fractures in IBD patients, and despite the guidelines published in 2003,6 several studies demonstrate that suboptimal attempts are made in general to diagnose or treat this problem.75

For coeliac disease the situation is different. Previously, it was assumed that high rates of osteoporosis in this patient population require assessment of bone mineral density upon diagnosis.76 However, recent data demonstrating lower osteoporosis rates put these guidelines in question.64 There have been reports suggesting high prevalence of positive serologies for coeliac disease in postmenopausal women with “idiopathic” osteoporosis;77 however, screening patients with low BMD for coeliac disease has low yields, according to more recent studies.78

What is the best way to screen for osteoporosis?

Amongst diagnostic procedures [including radiographs, dual-photon absorptiometry (DPA), quantitative computed tomography (QCT) and ultrasound], the U.S. Preventive Services Task Force Recommendations state that bone density measurements using DEXA are the best way to predict fractures.73 However, as bone density measurements account for only about 70% of fracture risk, it is important to consider additional risk factors for osteoporosis (box 1) as well. It should be noted that using DEXA in the paediatric population necessitates special adjustments as this measurement is stature dependent.79


Prophylaxis and general aspects

Randomised controlled studies on physical activity in IBD patients are rare. Robinson et al80 had shown that in fully compliant Crohn’s disease patients BMD increased when adhering to a low impact physical activity programme. As the beneficial effects of exercise programmes designed to stimulate and maintain bone mass in healthy subjects are known, it would be reasonable to further investigate their effects in IBD patients (fig 2)

Figure 2 Bone disease in chronic inflammatory bowel diseases: prevention and treatment strategies

Should all patients starting corticosteroid treatment be given prophylactic therapy against osteoporosis? Bone loss in patients treated with corticosteroids may occur rapidly, especially in corticosteroid naive patients, and even at low corticosteroid doses. The detrimental effect on bone correlates with corticosteroid doses. Thus, all IBD patients on corticosteroid treatment should receive calcium and vitamin D supplementation. The reported suboptimal intake of calcium and vitamin D of IBD patients4 strengthens this recommendation. In IBD patients on current corticosteroid treatment planned for more than 3 months, who are corticosteroid-dependent or have T scores lower than –1, bisphosphonates may be used for primary or secondary prevention, respectively.12

An important preventive measure is to avoid or decrease the use of corticosteroids. As has been mentioned, there is more than one mechanism by which corticosteroids negatively affect bone remodelling, and one potential mechanism where they may positively affect it (i.e. by decreasing inflammatory activity). Bernstein et al81 had indeed demonstrated that Crohn’s disease (but not ulcerative colitis) patients who used corticosteroids had a greater risk of fracture. When active disease necessitates the use of anti-inflammatory agents more effective than 5-aminosalicylates, alternatives to conventional prednisolone, such as azathioprine and anti-TNF-α agents which may preserve or even increase BMD, should be considered. Budesonide had a beneficial effect especially on the prevention of early bone loss in corticosteroid naive Crohn’s disease patients, as shown by Schoon et al.82 Its use may, however, be limited by its moderate efficacy, and the selective sites of action in the gut. Whether its use is significantly better for bone health than the short-term use of systemic corticosteroids, with their stronger effect on inflammation and the potential consequent improvement of BMD, is uncertain.

Other therapeutic options

Several therapeutic options for the treatment of osteoporosis that are applicable to the general postmenopausal population may also be applicable to IBD patients, especially if they are post-menopausal (fig 2). These were summarised by von Tirpitz and Reinshagen.83 An important point should be stressed that the number of studies and population size evaluated in studies on postmenopausal osteoporosis is, as expected, significantly greater than that for IBD. Thus, the recommendations given for IBD are sometimes extrapolations from postmenopausal osteoporosis; and we assume, without good evidence, that they should apply also to IBD.

Calcium and vitamin D

Calcium and vitamin D are effective for the prevention of postmenopausal osteoporosis, but are insufficient for its treatment. The recommended doses for calcium are 600–800 mg/day for premenopausal women, 800–1200 mg/day postmenopausal;84 ,85 and for vitamin D, 700–800 IU/day.86 In Crohn’s disease, vitamin D with or without calcium was effective in preventing bone loss.87 ,88


Bisphosphonates are pyrophosphate analogues that are effective for the prevention and treatment of postmenopausal osteoporosis. Some of them such as alendronate were also effective for the prevention and treatment of corticosteroid-induced osteoporosis and thus may be particularly important for IBD patients.89 Indeed, a significantly increased BMD after 1 year of oral alendronate and 2 years of intravenous ibandronate every 3 months combined with calcium and vitamin D was demonstrated in randomised, controlled trials in Crohn’s disease patients.90 ,91 Oral risedronate was significantly better than placebo (and when combined with calcium) in increasing BMD of osteopenic Crohn’s disease as well as ulcerative colitis patients in a randomised controlled trial.9 Supportive uncontrolled trials had shown that after 1 year of intravenous pamidronate every 3 months combined with calcium and vitamin D, BMD in the lumbar spine of Crohn’s disease patients significantly increased.92 Other studies, however, had less positive results: Bartram et al93 also used intravenous pamidronate every 3 months in addition to calcium and vitamin D for the treatment of Crohn’s disease patients with T scores ⩽−1.5. Patients increased their BMDs whether or not they received pamidronate.93 Siffledeen et al94 reported that there was no additional effect of intermittent 14/90-day cycles of oral etidronate (compared to no treatment), followed by calcium and vitamin D repeatedly for 2 years in Crohn’s disease patients with low BMDs. Thus, while bisphosphonates are potent “bone-saving and maintaining” preparations, the above-mentioned studies demonstrate that they may not be needed in many cases, and that the addition of calcium and vitamin D to young IBD patients, with BMDs in the osteopenic range may suffice. Potential exceptions are corticosteroid-induced osteoporosis where bisphoshonates are effective therapeutic modalities as shown by several groups using different preparations such as risedronate and etidronate (summarised by Bernstein et al6). They are also effective in case of patients with low impact fractures, and patients with T scores less than –2.5. Potential problems with bisphosphonate therapy exist, which may be specifically bothersome in the patient population with intestinal inflammation. For instance, gastrointestinal side effects (such as oesophagitis), their long half-life combined with their ability to cross the placenta, which should make their use in women of childbearing age carefully considered as effects on fetal skeletal development may occur, and the insufficient intestinal absorption of the oral preparations. For the reasons of intolerance and suboptimal intestinal absorption of some of the oral bisphosphonates, preparations which may be administered once weekly, monthly or intravenous 3-monthly, had been designed.2 ,93

Hormone-replacement therapy and selective oestrogen receptor modulators

The use of oestrogens and combined oestrogen and progestin preparations was associated with a significant decrease in fractures rates (24–34% reduction, depending on fracture location).95 Selective oestrogen receptor modulators (SERMs), such as tamoxifen and raloxifene (60 mg/day), increase BMD and decrease risk of fractures in postmenopausal women.96 Direct evidence for the beneficial effects of HRT on BMD in postmenopausal women with IBD were reported earlier by Clements et al.97 HRT was a major treatment modality for postmenopausal osteoporosis until the publication of the Women’s Health Initiative study in 200293 which reported an increased risk of breast cancer and cardiovascular disease associated with HRT. As a result US and EU health authorities recommended severe restriction on the use of HRT as a first-line treatment for postmenopausal osteoporosis. However, further analysis of the data suggests that this may have been an over-reaction.9699 Further advice from health authorities is awaited. Until then, HRT may be considered as a second-line treatment for postmenopausal osteoporosis. This should apply to postmenopausal IBD patients as well.

Other drugs


Calcitonin decreases bone resorption by directly interacting with receptors on osteoclasts. This preparation, whichwas introduced intranasally at 200 IU/day, was associated with BMD improvement and, more importantly prevented vertebral fractures in postmenopausal women.98 However, as controversy exists regarding its effects on the prevention of fractures, as well as due to the side effects of calcitonin and high cost, this modality is approved only for established postmenopausal osteoporosis.


Recombinant human parathyroid hormone (rh PTH 1–34; teriparatide) has recently shown promise in the treatment of postmenopausal and corticosteroid-induced osteoporosis, increasing BMD and reducing the occurrence of fractures.99 ,100 In a 1-year randomised study comparing BMD in women taking 5–20 mg/day prednisone equivalent concomitantly with oestrogen alone or oestrogen and teriparatide, those on oestrogen and teriparatide had an increase in lumbar spine BMD compared with no increase in those on oestrogen alone.101 More interestingly, Saag et al102 recently demonstrated in an 18-month randomised, double-blind, controlled trial that among patients with osteoporosis at high risk for fracture, BMD increased more in patients receiving teriparatide than those treated with alendronate.

Strontium ranelate

Strontium ranelate is a novel agent composed of ranelic acid and two atoms of non-radioactive strontium. Its mechanism of action is possibly related to concomitantly increasing bone formation and reducing bone resorption. Its beneficial effect in increasing BMD and decreasing the incidence of vertebral and non-vertebral fractures in postmenopausal osteoporosis had recently been reported.103 ,104

Decreasing osteoporosis by controlling inflammation

As in IBD the inflammatory process itself may contribute to disturbed bone remodelling21 (summarised by Bernstein and Leslie10) as discussed in detail earlier, it is reasonable to assume that treatment of inflammation with agents other than corticosteroids may improve BMD derangement. Reffitt et al105 studied IBD patients in remission compared to active disease and reported that those in remission had significantly increased Z-scores, and that this correlated with increasing duration of remission. Importantly, patients treated with azathioprine had significantly increased Z-scores, possibly due to the drug’s effect on disease activity. Thus, while acting indirectly, azathioprine therapy was associated with an additional benefit of increasing BMD in IBD patients.

Is anti-TNF-α therapy the logical treatment in inflammation-associated osteoporosis? As mentioned in the previous paragraphs, TNF-α is involved at various steps in inflammation-associated osteoporosis and contributes to this process in various ways such as activating osteoclasts.31 Therefore, it seems logical and attractive to assess and study the effects of anti-TNF-α agents on osteoporosis in IBD. So far, only few, mainly retrospective, reports have been published studying this aspect. Pazianas and colleagues106 retrospectively studied 61 patients with Crohn’s disease and low BMD by serial DEXA scans. After controlling for corticosteroid use, patients with concurrent infliximab and bisphosphonate treatment exhibited a greater increase in BMD compared to those on bisphosphonates alone. However, infliximab alone had no effect on BMD suggesting that concurrent anti-TNF-α therapy may confer no additional benefit to that already observed for bisphosphonates alone. Another study, however, suggested a decrease in osteoporosis in Crohn’s disease patients treated with infliximab only, pointing to a potential effect of this treatment per se on bone remodelling.107 This study, although not reporting on endpoints such as BMD or fractures, did demonstrate that there was improvement in markers of bone formation in 59% of the patients. Interestingly, the increase in markers of bone formation did not correlate with clinical activity or response. Abreu and colleagues108 also studied the effects of treatment with infliximab on parameters of bone formation. They investigated prospectively sera from 38 patients for bone alkaline phosphatase, N-telopeptide of type I collagen, immunreactive parathyroid hormone and calcium at baseline and 4 weeks following infliximab infusion. Serum levels of bone alkaline phosphatase were increased whereas serum levels of N-telopeptide of type I collagen, a marker of bone resoption, were not affected suggesting that infliximab positively affects markers of bone formation. Ryan and colleagues109 prospectively studied active Crohn’s disease patients treated with infliximab and demonstrated similar results. They investigated serum levels of bone alkaline phosphatase and N-telopeptide of type I collagen demonstrating that infliximab again increased levels of bone alkaline phosphatase and did not affect levels of N-telopeptide of type I collagen further supporting that an anti-TNF-α therapy may lead to bone formation. As in the previous study, no correlation with disease activity or response was demonstrated, and there were no clinically relevant endpoints such as fracture rates or BMD. Thus, the potential role of infliximab as a bone sparing agent remains to be demonstrated.

Similar findings have been observed in studies assessing anti-TNF-α therapies in patients with rheumatoid arthritis. Seriolo and colleagues110 studied markers of bone formation and resorption in 30 rheumatoid arthritis patients treated with infliximab or etanercept. Bone formation markers again increased whereas bone resorption as assessed by deoxypyridinoline decreased. Vis and colleagues111 prospectively studied 102 patients with active rheumatoid arthritis treated with infliximab over 1 year in an open cohort study. In this study, spine and hip bone loss could be arrested, whereas metacarpal cortical hand bone loss could not be stopped. In a prospective open-label pilot study, 26 patients with rheumatoid arthritis received infliximab therapy for 1 year.112 A significant increase in BMD in the spine could be observed after 12 months of therapy, with a significant increase in BMD at the left femoral neck but only a trend towards improvement in the spine during the study period suggesting that anti-TNF-α therapy might beneficially affect bone formation. It can be concluded that although so far only limited data on anti-TNF-α therapies are available, such a treatment approach might be beneficial supporting the “inflammatory concept” of osteoporosis in patients with IBD. However, larger prospective studies focussing on this aspect are eagerly awaited.


Several groups had demonstrated that in coeliac disease, a gluten-free diet had a beneficial effect on BMD.68 ,113 Whether the beneficial effect on BMD is mediated via mucosal healing is still controversial. Thus, a gluten-free diet, which is recommended for all coeliac disease patients, may also improve BMD. However, it should be noted that Ludvigsson et al5 recently reported that coeliac patients years after diagnosis still had an increased risk of hip fracture. As fracture risk, and not the surrogate marker of an increase in BMD is the clinically relevant end point, these results suggest that control of intestinal inflammation may be just one factor (and an insufficient one) in decreasing the risk of osteoporotic fractures. Calcium and vitamin D supplementation is mainly recommended for the prevention and treatment of the secondary hyperparathyroidism that may accompany untreated coeliac disease, although a gluten-free diet itself may correct this metabolic abnormality.


Much insight has come into the pathophysiology of bone remodelling in inflammatory states in the last few years. It has become evident that several mediators involved in chronic inflammation (e.g. in IBD and rheumatic disorders) are of crucial importance not only at the primary site of inflammation, such as the gut, but also in bone, regulating its activity. Thereby, this complex group of mediators such as TNF-α, IL-1, IL-6, IFN-γ, RANK, RANKL or osteoprotegerin regulates bone remodelling, activates osteoclasts and finally contributes to the development of osteoporosis. Although several aspects are involved in the development of osteoporosis, inflammation as a key factor has only been partially addressed so far. For clinicians, the most important task is to identify risk factors for osteoporosis that are associated with gastrointestinal-associated inflammatory diseases, to screen for osteoporosis in this population, to take the respective prophylactic “bone-saving” decisions and to treat this condition properly.


View Abstract


  • Competing interests: None.

  • Funding: This work was supported by a grant from the Austrian Science Foundation (P17447) and the Christian–Doppler Research Society.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.