Uncertainties in the prevention and treatment of glucocorticoid-induced osteoporosis


  • This document was prepared on behalf of the Professional Practice Committee of the American Society for Bone and Mineral Research.


Much knowledge has accrued since the 2001 American College of Rheumatology (ACR) guidelines were published to assist clinicians in the prevention and treatment of glucocorticoid-induced osteoporosis (GIO). Therefore, the ACR undertook a comprehensive effort to review the literature and update the GIO guidelines [Grossman JM, Gordon R, Ranganath VK, et al. American College of Rheumatology 2010 recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Care Res (Hoboken). 2010;62:1515–1526]. Herein, we review the new guidelines for JBMR readers, highlighting the changes introduced by the 2010 publication. We discuss several patient scenarios for which the new treatment guidelines do not apply, or for which our committee interprets existing literature differently and suggests an alternative approach. © 2011 American Society for Bone and Mineral Research


Several developments since publication of the 2001 American College of Rheumatology (ACR) guidelines prompted the ACR to update its guidelines for the prevention and treatment of glucocorticoid-induced osteoporosis (GIO).1 Two new medications were Food and Drug Administration–approved for the prevention and treatment of GIO. New risks of estrogen replacement therapy were identified.2 The WHO Fracture Risk Assessment Tool was developed to assist clinicians in determining a patient's 10-year probability of fracture.3 Finally, new methods to establish guidelines were developed.4

Several differences exist between the 2001 and 2010 ACR GIO guidelines. The former5 suggested osteoporosis therapy for any patient with a T-score below −1.0, based on epidemiological data indicating that bones of patients taking glucocorticoids fracture at higher T-scores compared with patients not taking glucocorticoids.6 In contrast, the 2010 ACR guidelines suggest an individual approach to each patient. For postmenopausal women and men 50 or older, recommendations for treatment differ depending on whether the patient is considered to be at low, moderate, or high risk of fracture. The risk of fracture is determined using clinical judgment or by employing the WHO Fracture Risk Assessment Tool to estimate the 10-year probability of fracture, with low, medium, and high risk representing <10%, 10% to 20%, and >20% chances of fracture, respectively. Such an approach, though more complicated, is congruent with current practice to estimate a patient's risk of fracture based on history, examination, bone density findings, and/or the probability of fracture based on the WHO Fracture Risk Assessment Tool.

On behalf of the ASBMR, the ASBMR Professional Practice Committee reviewed and, by majority vote, endorsed the 2010 ACR GIO guidelines.1 During the process of reviewing the guidelines, the Professional Practice Committee identified a number of common patient scenarios for which treatment was uncertain, due to either a lack of research or a different interpretation of existing data. The committee therefore identified the need to compose a commentary to discuss the care of patients in such common scenarios.

The entire ASBMR Professional Practice Committee was invited to participate in authorship of the commentary via e-mail. Interested individuals were then assigned topics the committee had previously identified as areas of uncertainty in the care of patients or areas in which the Professional Practice Committee and ACR Committee had differing interpretations of existing data. Subsequently, the contributions of each individual were united in a single document that was sent to all members of the ASBMR Professional Practice Committee for feedback. All feedback was incorporated into the final document, which was approved by the authors before its submission. Below is the final product of this collective effort.

Overview of GIO and Food and Drug Administration (FDA)–Approved Therapies

GIO is associated with an initial increase in bone resorption7, 8 and, more importantly, subsequent reduced bone formation9 leading to microarchitectural deterioration and a predisposition to fragility fractures. Such increased bone resorption and reduced bone formation represent important therapeutic targets of bisphosphonates and teriparatide, respectively.10 In this regard, clinical trials document the beneficial effects of bisphosphonates and teriparatide on BMD in patients taking glucocorticoids. In randomized placebo-controlled clinical trials, alendronate, risedronate, and zoledronate each prevented declines in spine and proximal femur BMD in adults receiving oral glucocorticoid therapy.11–13 In a 36-month randomized trial comparing alendronate and teriparatide for the prevention and treatment of GIO, teriparatide was statistically superior in preventing declines in spine and hip BMD.14

Data supporting the use of various medications to reduce fractures is less consistent and might partly relate to differing study inclusion criteria, including prednisone dose and duration. Compared with a placebo, risedronate reduced the risk of morphometric vertebral fracture by >50% among patients taking glucocorticoids.15, 16 In a 36-month randomized trial comparing alendronate and teriparatide for the prevention and treatment of GIO, teriparatide was statistically better than alendronate in reducing the risk of morphometric vertebral fracture.14 Data supporting the ability of alendronate to reduce fractures are less robust. Rates of morphometric vertebral fracture did not significantly differ in alendronate- and placebo-treated subjects in an initial 48-week trial,11 and 2-year data extracted from a subset of subjects in the initial trial were difficult to interpret, due to a high dropout rate and exclusion of most participants with incident fractures in the first year of the study.17 In a GIO study in which 833 subjects were randomized to 1 year of risedronate or zoledronate,13 there was no significant difference in incident morphometric vertebral fractures between treatment groups (three and five new fractures, respectively).

In preparation for updating the GIO guidelines, the ACR core executive panel graded the strength of evidence for drug treatment of GIO using a method developed by the American College of Cardiology.18 Medications were assigned an A rating if data supporting their use were derived from multiple randomized controlled trials or a meta-analysis, and a B rating if data supporting their use came from a single randomized controlled trial or a nonrandomized study. The weighting of outcome variables such as fracture reduction or BMD change was not specified. The ACR committee gave alendronate and risedronate an A rating, whereas both zoledronate and teriparatide received a B rating. Our committee concurs with the rating decision for risedronate and zoledronate. However, after careful scrutiny of existing clinical trials the committee concludes that available data do not support ranking the evidence on GIO treatment efficacy for alendronate higher than teriparatide for two reasons. First, there is a lack of convincing fracture data for alendronate, and second, compelling clinical trials data found that teriparatide was statistically superior to alendronate for effects on BMD and fractures among subjects recruited into the study based on their high risk of fracture.

The FDA approved alendronate for the treatment of GIO in adults taking a prednisone dose of ≥7.5 mg/d who have low BMD. The FDA also approved risedronate for the prevention and treatment of GIO in adults initiating or continuing systemic glucocorticoid therapy at a prednisone dose of ≥7.5 mg/d. Zoledronate (5 mg intravenously [iv] once yearly) is also FDAapproved for the prevention and treatment of GIO in adults taking ≥7.5 mg of prednisone daily with anticipated use for 1 year. Most recently, the FDA approved teriparatide (20 µg subcutaneously daily for up to 2 years) for adults who are at high risk of fracture and take sustained systemic glucocorticoid therapy.

The package inserts for the three bisphosphonates do not mention a limit to the duration of bisphosphonate therapy for the treatment or prevention of GIO. However, it should be noted that the duration of osteoporosis therapy was relatively brief in each of the aforementioned trials. Zoledronate and risedronate were administered for 1 year,12, 13 alendronate for 2 years11, 17 and teriparatide for 3 years14 in the studies leading to their FDA approval for GIO. Many of our patients require several years of glucocorticoid therapy, but to our knowledge, there is essentially no information on the safety or the efficacy of long-term bisphosphonate or teriparatide therapy among these patients. Additionally, the primary outcome of the aforementioned GIO trials was a surrogate marker of bone health (ie, the change in spine BMD) rather than the risk of fracture. Research studies are needed to evaluate the efficacy and safety of bisphosphonate and teriparatide therapy for longer than 2 or 3 years. Ideally, researchers should conduct large studies with fracture outcomes, but such trials are very unlikely on two accounts. First, long-term studies with fracture outcomes would be very costly, and second, the medications are now FDA approved for patients taking glucocorticoids.

Use of the RAND/University of California Los Angeles (UCLA) Method to Reach Consensus

The ACR 2010 Guidelines were developed using the RAND/UCLA method for determining the appropriateness of procedures or treatments.4 The method utilizes an iterative process in which a multidisciplinary panel with relevant clinical expertise is provided a literature review regarding what is and is not known about the benefits and risks of the intervention. Based on this evidence report and clinical judgment, the panel independently rates the appropriateness of the intervention in each of a comprehensive set of clinical scenarios. Subsequently, panelists meet with a moderator, who attempts to resolve differences. Ultimately, four outcomes are possible. Panelists may agree that an intervention is appropriate, agree that it is not appropriate, disagree on its appropriateness, or be uncertain of its appropriateness, often because of inadequate or conflicting evidence. Several prospective studies report that this methodology has external validity. Compared with patients who fit a scenario judged inappropriate for a procedure but underwent it anyway, patients who met panelist agreement for the intervention had better clinical outcomes after hip or knee replacement,19 percutaneous transluminal coronary angioplasty or coronary artery bypass grafting,20 or spine surgery for back pain.21

Nevertheless, there are potential limitations to the RAND/UCLA methodology. Even when restricting consideration to randomized controlled trials, authors of the literature review and panelist might disagree on which comparisons and outcomes are most relevant for rating the level of evidence to support the benefit and/or harms of an intervention. Not surprisingly, studies indicate that outcomes are sensitive to panel composition (with kappas of 0.5 to 0.8), and it seems likely that outcomes would also be sensitive to moderator effects.22, 23 Further, guidelines derived using the RAND/UCLA approach may be valuable only as far as the evaluated scenarios capture real-life clinical decision-making dilemmas.

In the following paragraphs, we discuss a number of common patient scenarios for which treatment remains uncertain, due to either a lack of research or differing interpretations of existing data.

The Critical Timing of Bone Protective Therapy

A 65-year-old man was admitted to the hospital with hemoptysis. He required intubation for respiratory failure with associated interstitial infiltrates. Urinalysis revealed hematuria, and a renal biopsy confirmed glomerulonephritis. He was diagnosed with Wegener's granulomatosis and received several days of high-dose iv glucocorticoids along with cyclophosphamide. His 4-week hospital stay was complicated by prolonged immobility during which he required ventilator support. He presented to rheumatology clinic 2 weeks after hospital discharge with acute back pain; subsequent imaging confirmed a compression fracture.

The 2010 ACR guidelines suggest onset of osteoporosis therapy in individuals for whom the anticipated duration of glucocorticoid therapy is 3 months or longer. As illustrated by this patient, clinicians might misinterpret the ACR guidelines by waiting 3 months before prescribing osteoporosis therapy to prevent a fragility fracture. In our experience, the opportunity to prevent GIO is often missed in the initial phases of long-term glucocorticoid therapy. GIO is characterized by rapid bone loss at both the spine and the hip, which is most marked within the first 6 to 12 months of therapy.12, 17 More importantly, an increased fracture risk is noted within 3 months of starting glucocorticoids.24–26 Therefore, our committee reinforces the 2010 ACR Guidelines regarding the need for osteoporosis therapy early in the course of glucocorticoid therapy. We emphasize that clinicians should initiate bisphosphonate or teriparatide therapy at the onset of glucocorticoid therapy. Therapy should be utilized in individuals for whom the duration of therapy is expected to exceed 3 months and also in individuals who would meet criteria for osteoporosis therapy regardless of glucocorticoid therapy.

The patient reported herein represents a missed opportunity, which is not uncommon in the setting of severe illnesses. It is important that in the inpatient care of a critically ill patient bone health not be forgotten. In this patient, high-dose iv glucocorticoids, along with prolonged immobility and cyclophosphamide in a person of his age, likely played a role in the incident compression fracture.

Duration of Bone Protective Therapy

A 72-year-old woman began prednisone and risedronate 4 years ago after the onset of temporal artery biopsy-confirmed giant cell arteritis. Baseline evaluation included normal laboratory studies, no clinical or morphometric spine fracture, and T-scores of −4.0 and −2.5 at the spine and femoral neck, respectively. Attempts to taper prednisone to <7.5 mg daily were unsuccessful. Upon reevaluation 4 years later, her BMD was stable, and repeat laboratory and imaging studies and clinical history were otherwise unchanged.

While taking risedronate, this high-risk individual did not sustain a decline in BMD or an incident fracture secondary to long-term glucocorticoid therapy. However, as noted above, clinical trials of FDA-approved bisphosphonate for the prevention or treatment of GIO were of relatively brief duration. Since the completion of bisphosphonate trials for GIO, new potential adverse effects of long-term bisphosphonate therapy have been recognized. Although rare, osteonecrosis of the jaw and atypical subtrochanteric fractures appear more likely among patients taking glucocorticoids.27, 28 Bisphosphonates slow bone remodeling, raising the theoretical concern that long-term bisphosphonate and glucocorticoid therapy may have additive and suppressive effects on osteoblast function. In contrast, teriparatide stimulates osteoblast function and prolongs osteocyte survival.

As clinical trials for the prevention and treatment of GIO lasted no more than 3 years, it is the opinion of the ASBMR Professional Practice Committee that patients requiring longer-term prednisone therapy be evaluated for substitution of bisphosphonate treatment by teriparatide, or be considered for treatment with teriparatide initially, followed by treatment with a bisphosphonate. We acknowledge that research studies are needed to evaluate this strategy and confirm its utility in clinical practice.

Teriparatide in Patients at High Risk of Fracture

A 68-year-old white woman presented to clinic for advice regarding treatment of her osteoporosis. She reported a wrist fracture at age 50, a humerus fracture at age 58, and a 3-inch height loss with associated chronic back pain. Spine X-rays revealed three thoracic compression fractures. She received occasional prednisone tapers for exacerbations of chronic obstructive pulmonary disease.

The 2010 ACR guidelines recommend bisphosphonates, but not teriparatide, for older adults at high risk of fracture who are taking <5 mg of prednisone daily for less than 1 month. However, an increased risk of fracture has been reported with prednisolone doses of <2.5 mg daily.24 Whether this risk is due to low-dose glucocorticoid therapy or the underlying disease for which such therapy is prescribed is uncertain. Regardless, our committee contends that postmenopausal women and men over the age of 50 and at high risk of fracture are candidates for teriparatide, independent of glucocorticoid dose or duration. A recent study recruited patients who were at higher risk of fracture by virtue of low T-scores or prevalent low-trauma fracture at study entry; subjects randomized to 36 months of teriparatide had greater increases in BMD and fewer morphometric vertebral fractures than subjects randomized to alendronate.14 Thus, our committee feels that teriparatide is one treatment option for the patient presented above, even though her prednisone dose is expected to be <5 mg daily for <1 month.

Individuals with a particularly high risk of hip fracture might require special consideration. If the patient noted above had a high risk of hip fracture and required high-dose glucocorticoid therapy, a clinician might wish to consider combination therapy with teriparatide and zoledronate. Although simultaneous prescription of zoledronate and teriparatide has not been specifically tested in patients taking glucocorticoids, the combination of both drugs for 1 year resulted in greater and more rapid increases in spine and hip BMD in postmenopausal women.29 Thus, this combination of drugs might be suggested for a patient with severe osteoporosis who requires long-term high-dose glucocorticoid therapy.

Treatment of Premenopausal Women

A 32-year-old white woman was referred to the osteoporosis clinic for management of GIO. She was diagnosed with lupus nephritis 4 years ago and has required high-dose oral glucocorticoid therapy since that time to control life-threatening complications of lupus, including cerebritis and nephritis. A recent bone density study revealed a lumbar spine Z-score of −2.8 and a lowest hip Z-score of −2.3. Additionally, the patient experienced a 20% decline in spine BMD compared with a bone density study performed 3 years ago. Spine X-rays revealed no prevalent compression fractures.

In the 2001 guidelines, decisions to prescribe osteoporosis therapy rested upon having a T-score below −1.0, regardless of the patient's age or menopausal status. In the new guidelines, decisions to treat younger adults rest upon the presence of a fragility fracture. For premenopausal women and men younger than 50, treatment is recommended only if the patient has sustained a fragility fracture. Thus, this woman with lupus nephritis does not qualify for osteoporosis therapy, even though her significant decline in BMD coupled with ongoing glucocorticoid therapy increases her risk of future fractures. Such an approach raises several important issues.

What is the absolute risk of fracture in young adults taking glucocorticoids? The 2010 ACR GIO guidelines recommend osteoporosis medication only when younger adults have sustained a fragility fracture, because these individuals are at greatest risk of future fractures. The recommendation may have been made because of the large number of young adults needed to treat, to prevent one fracture related to GIO. According to one source,30 only 1 premenopausal woman among 157 sustained a fracture during the conduct of three placebo-controlled trials using etidronate,31 alendronate,11 and risedronate12 for the prevention and treatment of GIO. Although such low rates of fracture in premenopausal women might be due to lower bone resorption relative to postmenopausal women, the aforementioned studies lasted no more than 2 years. Thus, the lifetime risk of fracture related to glucocorticoid exposure in younger life is largely unknown.

Prevention of bone loss seems an appropriate focus of therapy for young adults taking glucocorticoids. In premenopausal women, the risk of incident vertebral fracture appears to be low during the initial 2 years of systemic glucocorticoid therapy,30 but the lifetime risk of fracture related to glucocorticoid therapy is largely unknown. Our committee recommends that clinicians measure BMD and perform spine imaging at the onset of glucocorticoid therapy, and again 2 years later. We agree with the ACR guidelines to start therapy for young individuals with fragility fracture, but we also suggest osteoporosis therapy for premenopausal women who experience a significant decline in BMD or have a Z-score below −2.0 and continue to require long-term systemic glucocorticoid therapy. Although we acknowledge that this recommendation is based on expert opinion rather than data, such an approach is suggested in order to reduce the risk of osteoporotic fracture in later life. Epidemiological and prospective research studies are needed to determine the lifetime risk of osteoporotic fracture among young adults exposed to glucocorticoids.

The second issue raised by the 2010 ACR guidelines relates to how clinicians will evaluate and treat bone health in younger adults taking glucocorticoids. Under the 2010 ACR guidelines, treatment is recommended for young adults with a fragility fracture, but not for young adults with a Z-score below −2.0. Most fractures related to GIO involve the spine,25 and more than half of such fractures occur without symptoms.32 If clinicians decide to embrace the 2010 ACR approach to prevention and treatment of GIO in young adults, then detection of compression fractures by periodic spine X-rays or VFA images will naturally replace periodic bone density tests in young adults. However, the proper interval for sequential spine imaging, the cost-benefit ratio of this approach, and sensitivity and specificity of VFA for detecting vertebral fracture among young adults on glucocorticoids have not been adequately evaluated.33

The 2001 and 2010 ACR GIO guidelines both recommend caution when prescribing bisphosphonates to premenopausal women, due to concerns about their potential effects on the fetal skeleton. We agree with the 2010 ACR guidelines that caution is needed when prescribing bisphosphonates, teriparatide, and denosumab to fertile women, as all drugs are categorized as class C by the FDA. Although the bisphosphonates likely vary with respect to their terminal half-lives in bone,34 all bisphosphonates have long enough half-lives to raise concern about their recirculation during pregnancy, even if discontinued before pregnancy. Bisphosphonates cross the placenta in both animals35 and humans.36 Pregnant animals receiving bisphosphonates can display prolonged parturition and hypocalcemia, while their exposed fetuses can display decreased survival, retarded skeletal or tooth growth, and higher rates of congenital malformations.37 In humans, hypocalcemia might occur in the mother or the baby.38 Very little information has been published on the outcomes of women and their children after exposure to bisphosphonates during pregnancy.37 Registries and epidemiological studies are needed to determine the impact of osteoporosis medication on the human fetal skeleton.

If a premenopausal woman sustains a fragility fracture and takes prednisone for at least 3 months at a dose of ≥7.5 mg/d, then the 2010 ACR GIO guidelines recommend prescription of alendronate, risedronate, or zoledronate. Our committee suggests that clinicians consider teriparatide as an alternative to the bisphosphonates in women considering future pregnancy and who have closed epiphyses. Teriparatide, unlike bisphosphonates, will not remain in the skeleton when therapy is terminated and therefore is not expected to affect fetal bone growth. Although teriparatide has not been shown to reduce fractures in premenopausal women,39 a similar lack of data on fracture risk reduction also exists for bisphosphonates.30 Among 62 premenopausal women randomized to alendronate or teriparatide,40 women receiving teriparatide experienced greater increases in lumbar spine and total hip BMD at 18 months (7% versus 0.7%, p < .001).39 The BMD response to teriparatide was comparable in premenopausal and postmenopausal women, suggesting a beneficial effect of teriparatide that is independent of baseline BMD, bone turnover, or menopausal status. By contrast, in subgroup analyses of placebo-controlled trials, premenopausal women randomized to alendronate 11 or risedronate 12, 41 did not experience statistically significant increases in spine or hip BMD. Whether this relates to the small sample size of this subset of subjects or reflects little benefit of bisphosphonate therapy in this group of patients with low bone resorption is unknown. In summary, teriparatide increases BMD in premenopausal women and has a short half-life, making teriparatide a reasonable treatment option for premenopausal women who plan a future pregnancy.

Although not yet FDA approved for GIO, a recent placebo-controlled study in patients with rheumatoid arthritis revealed that denosumab increased spine and hip BMD regardless of whether subjects were taking glucocorticoids.42 Denosumab has a relatively short half-life compared with the bisphosphonates and might represent a possible treatment option for premenopausal women with GIO who plan a future pregnancy and have contraindications to teriparatide therapy.

The Use of Zoledronate versus Other Bisphosphonates

A 66-year-old woman with limited scleroderma required prednisone 5 mg daily to control symptoms and signs related to interstitial lung disease. Spine X-rays revealed no prevalent compression fractures, and her fracture risk was judged to be intermediate, based on her BMD measurements and the score derived using the Fracture Risk Assessment Tool.

The 2010 ACR GIO guidelines recommend alendronate or risedronate for patients such as this. Although zoledronate is also FDA approved to prevent and treat GIO, the ACR guidelines suggest zoledronate only for older adults at moderate risk of fracture when the daily prednisone dose is ≥7.5 mg. It is not clear why there is a difference in the recommendations for the various bisphosphonates. The zoledronate clinical trials had inclusion criteria similar to those of the trials of alendronate and risedronate, recruiting only those subjects taking ≥7.5 mg of prednisone per day. The duration of the zoledronate clinical trials was 1 year, as was the risedronate trial. Perhaps the decision to withhold zoledronate related to lack of fracture outcome data, though similar concerns would also apply to alendronate. Our committee acknowledges the greater cost of zoledronate relative to oral generic bisphosphonates. However, treatment with zoledronate ensures excellent adherence, represents a treatment option for patients with gastrointestinal conditions prohibiting oral bisphosphonates, and, though more expensive, might lead to similar or lower out-of-pocket expenses for patients with Medicare or insurance coverage. Thus, the committee feels that zoledronate should be considered in the GIO treatment algorithm whenever risedronate and alendronate are considered, and certainly when oral bisphosphonates are contraindicated, as might be the case if our scleroderma patient had esophageal dysmotility.

The Use of Hormone Therapy

A 69-year-old man on chronic glucocorticoid therapy for rheumatoid arthritis reported poor libido, fatigue, and muscle weakness. Although his BMD remained stable since starting alendronate 3 years ago, he reported two rib fractures in the past year. His morning testosterone level was below normal on three separate occasions.

The 2001 guidelines recommended the use of estrogen or testosterone for patients with hypogonadism. By contrast, the 2010 guidelines do not recommend hormone therapy, even though hypogonadism is a recognized complication of glucocorticoid therapy. There are several valid reasons to exclude hormone therapy from the GIO treatment algorithm. First, the Women's Health Initiative Study2 revealed that some groups of postmenopausal women taking hormone therapy had a greater risk of heart disease, breast cancer, stroke, and pulmonary embolus. Second, studies using estrogen or testosterone therapy for patients taking glucocorticoids are limited to small trials of brief duration, with BMD instead of fracture as the studied outcome. Third, controversy exists regarding the impact of glucocorticoid-induced hypogonadism in the pathogenesis of GIO,43 as the histomorphometric changes of hypogonadism are distinct from those of glucocorticoid-induced alterations of bone.44

Existing data suggest that glucocorticoid therapy can contribute to hypogonadism. In a study of healthy men, glucocorticoid therapy decreased serum testosterone, estradiol, androstenedione, and dehydroepiandrosterone sulfate levels.45 Researchers found a positive correlation between serum estradiol levels and BMD in a study of 32 women with rheumatoid arthritis taking glucocorticoids.46

Data, albeit limited, indicate that sex hormone therapy has a salutary impact on BMD in patients with hypogonadism. In 23 postmenopausal women taking concomitant estrogen therapy and oral glucocorticoids, BMD was maintained for the 1-year duration of the study.47 Among 15 men taking oral glucocorticoids for asthma,48 testosterone increased BMD. The beneficial effects of testosterone on BMD in men appear to be mediated through its aromatization to estradiol because the compound nandrolone, which cannot be converted to an estrogen, has no effect on BMD.49

Thus, for the patient on glucocorticoids for whom a bisphosphonate or teriparatide cannot be used, our committee suggests determination of sex steroid status and consideration of hormone therapy. For men, this is an appealing option because testosterone may have a positive effect on muscle mass, which may benefit the myopathy resulting from glucocorticoid therapy. In some cases, hormone therapy can be combined with other pharmacologic therapy for osteoporosis, particularly in patients who experience muscle weakness or low libido.


In summary, the 2010 ACR GIO guidelines represent a useful update for clinicians who care for patients taking systemic glucocorticoid therapy. However, several uncertainties remain in the care of such patients. We identify the need for further studies on the safety of long-term bisphosphonate and teriparatide therapy for individuals taking long-term glucocorticoids. We also recommend establishing a global registry to evaluate the safety of bisphosphonate therapy for pregnant women and their babies and identify the need for longitudinal studies to evaluate the long-term skeletal consequences of glucocorticoid therapy in young adults. In the meantime, we suggest some modifications to the approach of the 2010 ACR GIO Guidelines (Tables 1–3). We realize that our suggestions lack rigorous scientific validation and that our committee's approach is based on clinical experience and expert opinion with regard to bone health. Both the ACR guidelines and our suggestions are affected by the lack of long-term studies of GIO management and minimal inclusion of young adults in the extant studies.

Table 1. Treatment of Postmenopausal Women and Men Over Age 50
 Low riskMedium riskHigh risk
  • a

    The consensus of the ASBMR Professional Practice Committee.

  • b

    Treatment with alendronate, risedronate, or zoledronate, which are all Food and Drug Administration approved for the treatment of glucocorticoid-induced osteoporosis.

  • c

    American College of Rheumatology 2010 guidelines for the prevention and treatment of glucocorticoid-induced osteoporosis.

ASBMR PPCaPrednisone <7.5 mg/d: If no therapy given, monitor closely for prevalent fracture and decline in BMD Prednisone ≥7.5 mg/day: BisphosphonatebBisphosphonatebBisphosphonateb or teriparatide
Comparison to ACRcAgreementACR recommended zoledronate only in patients taking ≥7.5 mg of prednisone dailyACR recommended teriparatide only in patients taking glucocorticoids >1 month, or ≥5 mg daily for <1 month
Table 2. Treatment of Premenopausal Infertile Women and Men Under Age 50
 No prevalent fracturePrevalent fracture
Prednisone ≤3 monthsPrednisone >3 months
  • a

    The consensus of the ASBMR Professional Practice Committee.

  • b

    Treatment with alendronate, risedronate, or zoledronate, which are all Food and Drug Administration approved for the treatment of glucocorticoid-induced osteoporosis.

  • c

    American College of Rheumatology 2010 guidelines for the prevention and treatment of glucocorticoid-induced osteoporosis.

ASBMR PPCaConsider therapy if Z-score −2.0 or significant decline in BMD related to glucorticoid therapyBisphosphonatebBisphosphonateb or teriparatide
Comparison to ACR guidelinescACR committee found inadequate data for this subgroupACR committee recommended zoledronate only if prednisone dose ≥7.5 mg/dAgreement
Table 3. Treatment of Premenopausal Fertile Women
 No prevalent fracturePrevalent fracture
Prednisone ≤3 monthsPrednisone >3 months
  • Contraception is recommended for all fertile women, regardless of which therapy is chosen.

  • a

    The consensus of the American Society for Bone and Mineral Research Professional Practice Committee.

  • b

    Treatment with alendronate, risedronate, or zoledronate, which are all Food and Drug Administration approved for the treatment of glucocorticoid-induced osteoporosis.

  • c

    American College of Rheumatology 2010 guidelines for the prevention and treatment of glucocorticoid-induced osteoporosis.

ASBMR PPCaConsider therapy if Z-score −2.0 or lower, or significant decline in BMD related to ongoing glucocorticoid therapyLittle data to support therapyPreference for short-acting drugs like teriparatide or denosumab instead of bisphosphonatesb
Comparison to ACR guidelinescACR committee found inadequate data for this subgroupAgreementNo consensus if prednisone dose <7.5 mg/d; alendronate, risedronate, or zoledronate (not teriparatide) if prednisone dose ≥7.5 mg/d


The authors received no funding to support the preparation of the document. RAA served on the Task Force Panel for the American College of Rheumatology 2010 recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. RAA has received research support from Amgen, Eli Lilly, Genentech, Merck, and Novartis. He has also served on advisory boards for Eli Lilly, GTX, Inc., and Merck. SM has served on the speaker's bureau for Amgen, Eli Lilly, Merck Sharp & Dohme, Nycomed, Roche, and Warner Chilcott. He has also served on advisory boards for Amgen, Eli Lilly, Medtronic, and Novartis. HAW has served on the speaker's bureau for Eli Lilly and Warner Chilcott. CZ has served on the speaker's bureau for Amgen, Eli Lilly, Proctor and Gamble, Sanofi-Aventis, and Warner Chilcott. All other authors have no conflicts of interest.


We thank the other members of the Professional Practice Committee for their review of the manuscript, particularly Drs. Richard Bockman and Sunil Wimalawansa.