Controversies in Osteoporosis Treatment of Nursing Home Residents

Abstract

Osteoporotic fractures are a common and serious health problem for older adults living in nursing homes (NHs). Risk of fracture increases with age and dementia status, yet gaps in evidence result in controversies around when to start and stop treatment for osteoporosis in NH residents, particularly those who have high fracture risk but have limited life expectancy. In this paper, we discuss these areas of controversy. We provide an overview of current guidelines that explicitly address osteoporosis treatment strategies for NH residents; review the evidence for osteoporosis medications in NH residents; and use these sources to suggest practical recommendations for clinical practice and for research. Three published guidelines (U.S., Canada, Australia) and several studies provide the current basis for clinical decisions about osteoporosis treatment for NH residents. Practical approaches may include broad use of Vitamin D and selective use of osteoporosis medication based on risks, benefits, and goals of care. Clinicians still lack strong evidence to guide treatment of NH residents with advanced dementia, multimorbidity, or severe mobility impairment. Future priorities for research include identifying optimal approaches to risk stratification and prevention strategies for NH residents and evaluating the risk-benefit profile of pharmacologic treatments for osteoporosis NH residents across key clinical strata. In the absence of such evidence, decisions for initiating and continuing treatment should reflect a patient-centered approach that incorporates life expectancy, goals of care, and the potential burden of treatment.

INTRODUCTION

Clinical Scenarios

Consider two NH patients. Ms. A is an 80-year-old woman with Parkinson’s disease complicated by moderate dementia and orthostatic hypotension who is admitted to a long-term care facility after a hip fracture. She uses a wheelchair, but due to forgetfulness and impulsivity, she attempts to walk independently and falls frequently. Our second patient, Ms. B, is a 90-year-old ambulatory woman with mild dementia and three falls in the past year. She has no prior fractures but is taking prednisone for temporal arteritis. For these patients, who have both high fracture risk and limited life expectancy, is osteoporosis treatment an important component of a comprehensive care plan, or a potentially burdensome regimen that is more likely to harm than benefit them? In what situations would you not offer treatment or recommend deprescribing existing treatment?

Significance of Fractures in Nursing Homes

Osteoporotic hip fractures are a common and serious health problem for the 1.3 million older adults living in American nursing homes (NHs).¹ The rate of hip fracture is twice the rate of those living in the community. Fractures are also prevalent in the assisted living (AL) setting, where more than half of adults over age 75 fall each year.² In the six months after a hip fracture, more than 1 in 3 NH residents die.³ Nursing home residents with recent fractures are also at higher risk of reduced mobility, infections, pressure ulcers, rehospitalization, and other complications that adversely affect quality of life.^4,5

Current Practice Patterns

Observational studies of NH residents with osteoporosis demonstrate inconsistent use of pharmacologic therapies for fracture prevention.^6–10 Treatment rates range from as high as 40% to as low as 1.5%, suggesting potential undertreatment.¹⁰ At the same time, one study found that more than 10% of residents with severe mobility dependence and more that 10% of residents with <6 months life expectancy continued to receive pharmacologic therapy for fracture prevention.⁸ Variable patterns of treatment in real-world practice may be due to the lack of clear recommendations available to guide osteoporosis treatment in NHs.

The Controversy

Despite the significance of osteoporotic fractures, older adults living in NHs are rarely included in clinical trials of osteoporosis screening and treatment. Thus, clinicians face challenging decisions – estimating the likelihood of benefit from treatment in the context of life expectancy and fall risk and weighing this against the potential harms and burden of treatment, considering patient preferences around these issues.

Clinical trials enrolling community-dwelling older adults show that several medications are effective for fracture prevention in older people with osteoporosis, including bisphosphonates and others (denosumab, romosozumab, parathyroid hormone analogues, and selective estrogen receptor modulators). NH residents, however, differ from their community-dwelling counterparts in ways that may affect the balance of benefits and risks, including higher rates of polypharmacy¹¹, pill dysphagia¹², and chronic kidney disease¹³. Many of these issues are also prevalent in the AL population, albeit to a lesser degree.¹⁴ Given that the median life expectancy of an older adult entering a skilled nursing facility is about 2 years¹⁵, care must also be taken to select patients with sufficient time to have a chance of benefitting from treatment.¹⁶

Clinicians are challenged by an additional gap in research as to whether osteoporosis medications achieve the same degree of benefit for key subpopulations, such as those with limited ambulation, dementia, and multimorbidity resulting in reduced life expectancy. Most NH residents have a high burden of multimorbidity and impairments of functional status: nearly half (48%) have dementia, and 92% need assistance with walking.¹ Observational studies show greater dementia severity and comorbidity burden are associated with lower likelihood of osteoporosis treatment⁸, indicating concerns about the likelihood of benefit, or that fracture prevention is incompatible with more palliative goals of care. Nursing home clinicians must make choices about strategies for fracture prevention in the face of this uncertainty – should they initiate, continue, or deprescribe treatment based on their assessment of risks and benefits, life expectancy, and goals of care? In the sections that follow, we will: 1) provide an overview of guidelines that explicitly address osteoporosis treatment strategies for NH residents; 2) present a scoping review of evidence for osteoporosis medications in NH residents; 3) provide practical recommendations for clinical decision-making regarding osteoporosis medications, and for research.

WHAT DO NATIONAL GUIDELINES RECOMMEND FOR OSTEOPOROSIS MANAGEMENT IN THE LONG-TERM CARE SETTING?

We conducted targeted searches and engaged clinician experts to identify national osteoporosis guidelines and reviewed these for content specific to NH residents.

In the United States, several national expert groups publish practice guidelines that address osteoporosis broadly, but do not explicitly provide guidance on the NH setting. The Bone Health and Osteoporosis Foundation (formerly National Osteoporosis Foundation) Clinician’s Guide to Prevention and Treatment of Osteoporosis (2022)¹⁷ briefly mentions that alendronate and zoledronic acid have been shown to improve bone mineral density (BMD) in frail NH residents. Otherwise, clinical practice guidelines from the American Academy of Clinical Endocrinologists¹⁸, American College of Physicians¹⁹, and the American Society for Bone and Mineral Research²⁰ do not reference NH populations or individuals with mobility impairment or limited life expectancy.

Three guidelines from the United States, Canada, and Australia focus on osteoporosis treatment as well as fall injury prevention in the NH setting. In the United States, the American Medical Directors Association (now Society for Post-Acute and Long-Term Care Medicine) publishes a set of clinical practice guidelines that include osteoporosis management, last updated in 2009²¹. The Scientific Advisory Council of Osteoporosis Canada²² developed guidelines graded by level of evidence and updated in 2015 for osteoporosis management in long-term care residents, stratifying recommendations by high fracture-risk and low fracture-risk residents. The third Consensus Conference on Treatment of Osteoporosis in Residential Aged Care Facilities in Australia was held in October 2020²³, resulting in evidence- and consensus-based recommendations.

In Table 1, we have summarized key recommendations presented in the U.S., Canadian, and Australian guidelines on strategies for risk assessment and pharmacologic treatment for osteoporosis. For each recommendation, we determined whether the rationale was based on evidence collected in the NH setting, in the community and extrapolated to the NH setting, or expert consensus in the absence of published evidence.

Table 1.

Summary of Recommendations and Strategies for Risk Assessment and Pharmacologic Treatment for Osteoporosis

Recommendation	Source			Rationale
	US (AMDA 2009) [21]	Australia (CCPOFR 2021) [23]	Canada (Osteoporosis Canada 2015) [22]	NH Evidence	Community Evidence	Expert Consensus
Interventions for all NH residents
Employ fall injury prevention strategies (medication review, environmental assessment, etc.)	x	x	x	x
Assess fracture and/or fall risk on NH admission	x	x	x			x
Consider cholecalciferol (vitamin D3) in ambulatory residents	800-1000IU/d or 50,000IU monthly	1000IU/d (avoid periodic high doses)	800-2000IU/d (avoid periodic high doses)	x
Consider 1200-1300mg calcium intake daily from diet and/or supplements	Max supplement 1500mg/d	Max supplement 600mg/d	Max supplement 500mg/d	x
Medications for NH Residents with Osteoporosis
Treat patients with high fracture risk with oral bisphosphonates	x		x	x
Treat patients with high fracture risk with zoledronic acid	x	x	x	x
Treat patients with high fracture risk with denosumab		x	x		x
Avoid oral bisphosphonates in patients with dysphagia or disordered swallowing	x	x	x		x
For patients with low GFR, use denosumab first line		If GFR < 35	If GFR < 30		x
Consider using anabolic therapy for some high-risk patients	If intolerant to other drugs	If fracture after ≥1 year of antiresorptive use AND T<−3 or 2+ fractures	“High risk” patients only		x
Consider life expectancy and goals of care in treatment decisions for antifracture medications	x	LE >1 year	LE > 1 year			x

GFR = Glomerular filtration rate; T = T-score; LE = Life expectancy

The referenced guidelines differ in their overall scope and focus but have consensus on key components. All three guidelines address individualized fall and/or fracture risk assessment and recommend broad prescription of vitamin D and calcium supplements, while differing in specifics on dose and administration. All recommend selective pharmacologic treatment of osteoporosis after considering life expectancy and goals of care, with a few notable differences. The Australian guideline²³ does not recommend oral bisphosphonates as first-line treatment, citing the complexities of administration for frail NH residents (e.g., sitting upright, swallowing difficulty), differing from the other two NH guidelines and U.S. national guidelines. The U.S. AMDA guideline was last updated in 2009²¹, and therefore does not include more recently approved treatment options (e.g. denosumab), and includes treatments no longer widely utilized (e.g. raloxifene, calcitonin).

Guidelines are based on available evidence and on expert opinion. In the next section, we present a scoping narrative review and evidence synthesis of pharmacologic treatments for osteoporosis and fracture prevention for the NH population, guided by the evidence referenced in the guidelines above. We do not address non-pharmacologic strategies, as falls prevention interventions are heterogeneous in design and scope and several meta-analyses have been published that adequately address this expansive body of literature.^24,25

WHAT EVIDENCE SUPPORTS OSTEOPOROSIS TREATMENT DECISIONS FOR NURSING HOME RESIDENTS?

We conducted a scoping review of studies evaluating the effectiveness of osteoporosis treatments in NH residents. Our literature review highlights key studies referenced in the guidelines presented above. We also conducted a targeted literature search in Medline using terms related to aging, nursing homes, osteoporosis, and medications to identify additional observational or other studies conducted specifically among NH residents. Two authors reviewed all studies with input from clinician scientist experts in the field regarding strengths, limitations, and generalizability to sub-populations. The following sections present a narrative synthesis of evidence for supplementation with vitamin D and calcium as well as prescription medications for treating osteoporosis.

Vitamin D and Calcium Supplementation:

Evidence for supplementation with vitamin D alone or in combination with calcium requires careful interpretation. Although most trials evaluating supplementation have been conducted among community-dwelling older adults, some randomized studies either include a large proportion of NH residents²⁶ or focus on this population specifically.^27,28 A Cochrane Review that included NH residents²⁶ noted that vitamin D alone is unlikely to reduce fractures, including hip fractures. However, the combination of vitamin D and calcium was associated with reduced fractures in older adults, noting that benefits were most likely attributable to frail older adults residing in NHs. A systematic umbrella review of meta-analyses²⁹ supports this assertion, showing no significant fracture risk reduction among studies conducted in community-dwelling participants, versus modest risk reduction when limited to participants in NHs (range of absolute risk reduction: 0.67-0.85). Just two studies^27,28 have been conducted specifically in the NH setting and were limited to relatively healthy, ambulatory older adults with no severe medical conditions and excluded those who received medications that alter bone metabolism. Thus, a limited but consistent body of evidence supports broad use of Vitamin D and calcium supplements for NH residents, particularly for those who remain ambulatory.

Prescription Osteoporosis Medications:

Several randomized studies have evaluated the effectiveness of antiresorptive osteoporosis medications in the NH setting. The first was a randomized placebo-controlled trial of alendronate conducted among older women living in NHs or an AL setting.³⁰ Participants (n=327) were required to be ambulatory with a BMD T-score < −2.0 and were randomized to receive either alendronate 10 mg/day or placebo over 24 months. Compared to studies in community dwelling women, this study found greater increases in BMD at the spine and femoral neck, and greater decreases in bone turnover markers. The second study³¹ was a randomized placebo-controlled trial of a one-time zoledronic acid infusion among 181 older women living in NHs or AL with either low BMD or a history of vertebral or hip fracture. Those with a life expectancy <2 years and those with impaired renal function were excluded. As in the prior study, a statistically significant increase in bone mineral density was observed compared to placebo at 12- and 24-months. Neither study was powered to detect a difference in the rate of fractures, falls, or deaths.

Two observational studies have examined the comparative effectiveness of pharmacologic treatments among long-stay NH residents using Medicare data. A retrospective cohort study of approximately 10,000 residents³² evaluated the effectiveness of bisphosphonates among new initiators against an active comparator, calcitonin. Bisphosphonate initiation was associated with a modest reduction in hip fractures over 2.5 years of follow-up with a time to benefit as early as 6 months. Results were consistent across subgroup analyses stratified by age, sex, and baseline fracture risk. A second observational study³³ examined the comparative effectiveness of denosumab, teriparatide, and zoledronic acid for prevention of hip fractures in a sample of approximately 2,000 residents. Denosumab and zoledronic acid were found to have comparable effectiveness to teriparatide in preventing hip fractures.

RECOMMENDATIONS FOR PRACTICE AND RESEARCH

We present a suggested approach to osteoporosis treatment decisions in NH residents in Figure 1.

Figure 1.

Flow Diagram of Practical Considerations for Pharmacologic Fracture Prevention Treatment for Older Nursing Home Residents

What screening tools should be used to identify NH residents who are most likely to benefit from osteoporosis treatment?

Recommendations:

• We recommend use of a clinical screening tool, such as history of fracture +/− FRAiL or FRS model (described below), when feasible, to identify candidates for osteoporosis treatment.

• We do not recommend routinely incorporating BMD into decision-making, due to barriers to testing for most NH residents, although it may be useful for considering fracture risk when available.

• Candidates for osteoporosis treatment should undergo individualized decision-making incorporating life expectancy, goals of care, and other personal factors.

Rationale:

There is a lack of consensus as to what constitutes the target population for osteoporosis medications in the NH setting. Guidelines recommend fracture risk assessment, but there is no optimal assessment for the NH setting. Community-based fracture prevention models such as the Fracture Risk Assessment Tool (FRAX)³⁴ that incorporate BMD suggest that over 90% of NH women would be eligible for treatment.³⁵ However, these tools may not be appropriate for this setting, as they do not incorporate fall risk or functional characteristics and bone densitometry is not often accessible in NHs. At least two models specific to NH patients utilize the interRAI-Minimum Data Set (MDS) but have not been widely studied. The FRAiL model³⁸ has been validated to predict 2-year risk of hip fracture in US NHs, while the FRS model³⁹ has been validated to predict 1-year hip fracture risk in Canadian NHs. However, studies are needed to determine the optimal risk cut-point for treatment. Whether these or other risk models are feasible for incorporation in an electronic medical record (EMR), or at the bedside is ripe for study in future research. Estimation of life expectancy and communication of the risks, benefits, costs, and hassle of starting osteoporosis medications in the context of a patient’s health trajectory and goals of care is also remains unaddressed for fracture prevention.

Which pharmacologic therapies are most appropriate for NH residents?

Recommendations:

• Available therapies, particularly bisphosphonates, are likely to be at least as effective for NH residents with sufficient life expectancy as they are in community-dwelling older adults.

• Although supplementation with vitamin D and calcium is a low-risk intervention for which even a modest benefit may outweigh the risk, there is no evidence for reduced fracture risk in immobile residents or those approaching end-of-life.

• Decision-making should reflect a patient-centered approach that considers pill burden, dysphagia, and risk for adverse effects relative to risk for fracture.

• The choice to use bisphosphonates should be positively influenced by a resident’s ambulatory status and risk for falls, preserved renal function, and life expectancy of least 1 year.

• Consider whether the magnitude of fracture risk reduction is clinically meaningful to the patient (0.4-1% absolute risk reduction for nonvertebral fractures at 1 year, increasing with longer duration of treatment).

• Consider deprescribing bisphosphonates if life expectancy is less than 2 years, unless fracture risk is particularly high and patient goals continue to prefer treatment.

• Evidence for treatments other than bisphosphonates is limited and cannot be recommended with certainty. If denosumab is used, attention should be given to duration of use and risk for rebound fractures after deprescribing.

Rationale:

The benefits of supplementation with vitamin D plus calcium likely outweighs the low risk and low burden associated with these preventive measures for many patients, except for those particularly affected by pill burden or constipation. However, the populations included in randomized studies evaluating these benefits may not be representative of the NH population at large.

For bisphosphonates, observational studies have identified sub-populations of NH residents for which the likelihood for benefit may be reduced, e.g., those who are no longer ambulatory and those with short life expectancy. One study of NH residents newly starting bisphosphonates reports a number needed to treat (NNT) to prevent one non-vertebral fracture compared to calcitonin after 1 year of 270 (individual absolute risk reduction [ARR] 0.4%).³² This estimate is higher than that of a recent meta-analysis of randomized trials of bisphosphonates in community-dwelling women, which reported a NNT of 100 (ARR 1%) to prevent one non-vertebral fracture at about 1 year.¹⁶ The difference in these estimates is likely due to research methodologies, though differences in NH vs community-dwelling adults may also play a role.

Data on the optimal duration of treatment is also limited, which may hamper efforts to reduce low-value prescribing. Although there is some data to suggest that the benefits of bisphosphonates may persist for 2 or more years after stopping treatment^39–42, data on fractures is too limited to make a strong recommendation on optimal time frame for deprescribing and should also consider fracture risk and goals of care. Considering the barriers to conducing randomized studies in this medically frail population, there is opportunity for well-designed, large observational studies to address questions related to treatment allocation, time-to-benefit, optimal treatment duration, and deprescribing for this population. However, observational studies may be limited by data availability across care transitions, duration of follow up, and residual confounding due to the high prevalence of multimorbidity.

Finally, there is limited data on the relative risks and benefits of non-bisphosphonate therapies for osteoporosis such as denosumab, PTH analogues, and sclerostin inhibitors in NH residents. Denosumab presents an opportunity to treat patients with stage 4 chronic kidney disease and/or dysphagia and has lower administration burden (subcutaneous, twice yearly). However, there is a risk of rapid bone loss and “rebound fractures” when deprescribing denosumab in patients who remain at fall risk, complicating treatment toward the end of life.⁴³ Higher cost and insurance coverage may also be barriers. Additional studies are needed to evaluate the comparative effectiveness of different treatments considering feasibility of administration, medication costs, and risk for adverse effects.

What other approaches are recommended in the NH setting?

Recommendations:

• Multi-faceted fracture-prevention interventions that combine screening and pharmacologic treatment options with non-pharmacologic strategies are likely to have a synergistic effect on health outcomes.

• Recommendations should not be strictly applied based on care setting, but rather based on individual clinical considerations as relevant to patient-centered decision-making.

Rationale:

Clinicians should also consider the value of falls prevention as part of osteoporosis management. Fall prevention in NH is beyond the scope of this paper and has been described elsewhere.^25,44 Hip protectors provide modest benefit in preventing fractures⁴⁵ for very high-risk patients who are willing to wear them, but adherence may be limited by patient comfort and staffing. Targeted deprescribing of fall-risk increasing drugs is a logical strategy to reduce fractures, though interventions often fail to show significant reductions in negative outcomes.^46,47 This may be attributable to low adherence to deprescribing recommendations or the narrowed focus on CNS-active medications that fails to consider other medications that contribute to falls, such as antihypertensives, antidiabetics, and anticholinergics.

Finally, we acknowledge that much of the evidence presented does not explicitly address the AL population or other frail older adults. Many individuals living in AL are similarly affected by multiple chronic conditions (66%), dementia (34%), and mobility limitations (69%).⁴⁸ Yet, the AL setting lacks extensive regulatory oversight to encourage initiatives for fracture prevention.

CONCLUSION

Several guidelines address considerations specific to the NH population, but there is a general lack of strong evidence regarding optimal management of osteoporosis for NH residents across key clinical strata (e.g., advanced dementia, multimorbidity, and severe mobility impairment). Decisions for initiating and continuing treatment should reflect a patient-centered approach that incorporates life expectancy, goals of care, and the potential burden of treatment. Future priorities for research include identifying optimal risk stratification and further evaluation of the risk-benefit profile of pharmacologic treatments for osteoporosis NH residents.