Bone health in adults with non‐ambulatory neuromuscular disorders: scoping review of risk factors, diagnosis and management

Abstract

Background

Non‐ambulatory adults have an increased risk of osteoporosis and fractures due to reduced weight‐bearing and diminished neuromuscular stimulation, resulting in substantial morbidity and mortality.

Aims

This scoping review aimed to systematically evaluate risk factors, diagnostic indicators and management strategies for optimising bone health in adults with non‐ambulatory neuromuscular disorders.

Methods

A structured search was conducted in Scopus, MEDLINE, Cochrane Database of Systematic Reviews, JBI Evidence Synthesis, Embase and Web of Science. Inclusion criteria defined adults with non‐ambulatory neuromuscular conditions and reported outcomes on bone mineral density (BMD), osteoporosis or fracture incidence. Screening was performed in accordance with Preferred Reporting Items for Systematic reviews and Meta‐Analyses (PRISMA)‐Scoping Review guidelines.

Results

Twenty‐nine studies met the inclusion criteria. The most significant risk factor associated with bone disease was reduced weight‐bearing and immobility, followed by a history of past fractures and wheelchair dependence. Dual‐energy X‐ray absorptiometry was the primary diagnostic modality to assess BMD, while biochemical markers showed limited diagnostic utility. Management strategies such as neuromuscular electrical stimulation and passive bone loading demonstrated minimal efficacy. Bisphosphonate therapy was supported in specific subgroups, including Duchenne muscular dystrophy and spinal cord injury. No consensus on evidence‐based management was identified for the target population.

Conclusions

Current evidence to guide diagnosis and management in this population is limited. Adults with non‐ambulatory neuromuscular disorders represent a small, heterogeneous group with variability in disease aetiology, onset, distribution and progression. This is reflected within the sparse literature, making it challenging to stratify risk factors, establish standardised outcome measures and develop robust, evidence‐based strategies to improve bone health.

Abbreviations

BMD

Bone mineral density

CP

Cerebral palsy

CT

Computed tomography

DMD

Duchenne Muscular Dystrophy

DXA

Dual‐energy X‐ray absorptiometry

GMFCS

Gross Motor Function Classification System

pQCT

Peripheral Quantitative Computerised Tomography

SCI

Spinal cord injury

Introduction

Osteoporosis and fragility fractures are common complications of bone disease, leading to early mortality, functional disability, reduced quality of life and substantial economic burden. ¹ Although risk factors, diagnostics and management strategies are well studied in ambulatory adults and non‐ambulatory children, evidence for non‐ambulatory adults remains limited, despite their heightened vulnerability to bone disease.

In neuromuscular disease, ambulation is shaped by progressive disease pathophysiology and limited rehabilitation capacity, predisposing individuals to bone disease. Fragility fractures and low bone mineral density (BMD) are both indicators of bone disease. In complete spinal cord injury (SCI), the lifetime risk of osteoporosis is 80%, with a 50% chance of at least one low‐impact fragility fracture, ² compared with lifetime osteoporotic fracture risks of 33% in women and 20% in men over 50 years. Adults aged 18–64 years with cerebral palsy (CP) also have a two‐fold higher fracture risk than adults without CP. ¹

A targeted, evidence‐based approach is necessary to manage bone health in non‐ambulatory adults, who differ from ambulatory populations due to absent weight‐bearing and neural stimulation affecting BMD. ³

This scoping review aimed to synthesise the evidence for non‐ambulatory adults with neuromuscular disease by assessing (1) risk factors for bone disease, (2) diagnostic methods for assessment and (3) management strategies to improve outcomes.

Non‐ambulatory status – defined as the inability to walk or mobilise independently and reliance on mobility aids – varies by disease aetiology, pathophysiology and clinical presentation. ² , ³ The American Spinal Injury Association (ASIA) Impairment Scale classifies SCI as complete (Grade A) or incomplete (Grades B–D) based on sacral sensory and motor sparing at S4–5. In CP, the Gross Motor Function Classification System (GMFCS) ranges from Level I (walking without limitation) to Level V (manual wheelchair use).

For this review, non‐ambulation was defined as ASIA Grade A or GMFCS Levels IV–V for SCI and CP, respectively. Neuromuscular conditions included muscular dystrophy, SCI, CP, myopathy and neuropathy, among others. These disorders impair muscle function via disrupted neural innervation, affecting contractility locally or systemically, and may have genetic or acquired aetiologies. Bone health outcomes were evaluated using BMD T‐ and Z‐scores, radiographic assessment of vertebral fragility fractures, fracture rates and serum biomarkers.

Inclusion criteria encompassed peer‐reviewed studies with accessible abstracts and full texts, published in English between 1 January 1990 and 16 May 2022, involving human subjects. Exclusion criteria were populations that were ambulatory or had unspecified ambulatory status, lacked a neuromuscular condition or included paediatric subjects.

Studies not assessing bone health, BMD, fracture risk or osteoporosis or those describing surgical interventions were excluded. Conference proceedings, posters, short communications, letters, editorials, abstract‐only publications, grey literature and unpublished studies were also excluded.

No restrictions were applied to geography, genetics or gender due to limited literature, supporting a broad, multidisciplinary search. Eligible sources included systematic reviews, meta‐analyses, experimental trials, observational studies and ecological studies.

Materials and methods

This scoping review followed JBI methodology and the Preferred Reporting Items for Systematic reviews and Meta‐Analyses for Scoping Reviews (PRISMA‐ScR) guidelines. ⁴ , ⁵ Ethics approval was not required.

The search strategy was developed with a research librarian and specialists in endocrinology, cardiology and rehabilitation medicine and adapted for each database (Appendix I). Key terms included ‘neuromuscular’, ‘muscular dystrophy’, ‘cerebral palsy’, ‘ambulation’, ‘osteoporosis’, ‘bone mineral density’ and ‘fracture’. The databases searched were Scopus, MEDLINE, Cochrane Database of Systematic Reviews, JBI Evidence Synthesis, Embase and Web of Science.

Search results were managed using Covidence (Veritas Health Innovation, Melbourne, Australia) and EndNote (Clarivate, Philadelphia, PA, USA), with duplicates removed. A single reviewer (TB) applied inclusion and exclusion criteria to screen titles, abstracts and full texts. Studies not meeting criteria were excluded at each stage. Uncertainties were discussed with the co‐authors. A PRISMA‐ScR flow diagram outlines the search process and exclusions. ⁵

Data relevant to the review question were extracted into Microsoft Excel 2016 (Microsoft Corp., Redmond, CA, USA). Extracted variables included citation details, study design, population demographics, interventions, controls, bone health outcomes, risk factors, diagnostic methods and management strategies (Appendix II). Data charting was performed by a single reviewer (TB), with uncertainties resolved through discussion with the co‐authors.

Data were tabulated to map risk factors, diagnostics and management of bone health in non‐ambulatory neuromuscular populations. A formal statistical analysis was not performed; findings were summarised narratively. A structured assessment of study bias was not undertaken.

Results

A total of 29 studies met eligibility criteria and were included for data extraction. The PRISMA‐ScR diagram (Fig. 1) outlines study selection and reasons for exclusion, while study characteristics are summarised in Table 1.

Figure 1.

PRISMA flow diagram of results screening.

Table 1.

Characteristics of included studies

	n	%
Study design
Cross‐sectional	12	41.4
Review	9	31
Experimental trial	5	17.2
Retrospective longitudinal study	2	6.9
Qualitative interview study	1	3.4
	29	100
Population
Cerebral palsy (CP)	11	37.9
Spinal cord injury (SCI)	7	24.1
Duchenne's muscular dystrophy (DMD)	4	13.8
Mixed – CP predominant	5	17.2
Other – IIM, BMD and LGMD (inflammatory muscle myositis) (Becker's muscular dystrophy) (limb‐girdle muscular dystrophy)	2	6.9
	29	100
Sex
Male	5	17.2
Female and male	12	41.4
Not stated	12	41.4
	29	100
Sample size
Total, n = 20 studies (reviews excluded)
1 to 25	6	30
26 to 50	7	35
50 to 100	4	20
100 to 150	1	5
150 to 200	0	0
200–300	2	10
	20	100
Non‐ambulatory definition
Complete SCI (spinal cord injury)	6	20.7
GMFCS (Gross motor functional classification scale)	5	17.2
Multiple measures, including GMFCS	4	13.8
Unable to walk independently or with assistance/device	4	13.8
Time‐based early to late non‐ambulatory	3	10.3
Other published scale	3	10.3
Wheelchair‐bound >75% time	2	6.9
Not stated	2	6.9
	29	100
Outcome measure
BMD Z or T‐scores (Bone mineral density)	12	41.4
Two or more BMD Z or T‐scores (Bone mineral density), bone biomarkers, radiography, fracture rates	12	41.4
Fracture rates	4	13.8
Body vibration transmission	1	3.4
	29	100

Risk factors

Of the 29 studies, 26 (90%) evaluated risk factors for bone disease in adult, non‐ambulatory neuromuscular populations (Table 2). Key risk factors for bone disease identified included reduced weight‐bearing and immobility (n = 22; 62.9%), prior fractures (n = 4; 11.4%) and wheelchair use (n = 4; 11.4%).

Table 2.

Results specifying study design, population and relevance to bone health risk factors, diagnosis and/or management

Author, year	Study design	Neuromuscular diagnosis, sample size (n)	Risk factor	Diagnosis	Management
Asselin, 2011	Experimental trial	Complete SCI, n = 17		None statistically significant	Neuromuscular electrical stimulation
Birnkrant, 2018	Review	DMD	Fractures	None statistically significant	Vitamin D, calcium, bisphosphonates
Bushby, 2010	Review	DMD	Reduced weight‐bearing, immobility and muscle weakness; no glucocorticoid use; glucocorticoid use	None statistically significant	Vitamin D, calcium, bisphosphonates
Carvalho, 2006	Experimental trial	Complete SCI, n = 21	Reduced weight‐bearing and immobility	None statistically significant	Neuromuscular electrical stimulation gait training
Danckworth, 2014	Cross‐sectional	Mixed, n = 89	Reduced weight‐bearing and immobility; wheelchair use	None statistically significant
Fowler, 2015	Cross‐sectional prospective	Cerebral palsy, n = 46	Reduced weight‐bearing and immobility	DXA BMD LS, FN, TH
Gannotti, 2021	Review	CP	Reduced weight‐bearing and immobility; fractures		Dynamic bone loading
Graupe, 2002	Review	Complete SCI			Neuromuscular electrical stimulation gait training
Jacobs, 2004	Review	Complete SCI	Reduced weight‐bearing and immobility	None statistically significant	Neuromuscular electrical stimulation gait training
Kannikeswaran, 2021	Cross‐sectional	CP, n = 58	Wheelchair use, transfer or limb stuck	None statistically significant
Kim, 2015	Cross‐sectional retrospective	CP, n = 58	Reduced weight‐bearing and immobility; spastic CP	DXA BMD femoral trochanter, TH
King, 2003	Cross‐sectional	Mixed, n = 48	Reduced weight‐bearing and immobility; fractures	DXA BMD LS
Marciniak, 2016	Chart review retrospective	CP, n = 42	Reduced weight‐bearing and immobility	DXA BMD left total hip
Morgenroth, 2012	Review	DMD	Delayed puberty	None statistically significant
Nakano, 2003	Cross‐sectional	CP, n = 123	Reduced weight‐bearing and immobility	DXA BMD second meta‐carpal in women only	Dynamic bone loading
Needham‐Shropshire, 1997	Experimental trial	Complete SCI, n = 16	Reduced weight‐bearing and immobility	None statistically significant	Neuromuscular electrical stimulation gait training
Rice, 2018	Qualitative semi‐structured interviews	Mixed, predominantly CP, n = 19	Falls risk
Shields, 2006	Experimental trial	Complete SCI, n = 7	Reduced weight‐bearing and immobility	None statistically significant	Neuromuscular electrical stimulation
Shin, 2017	Cross‐sectional	CP, n = 28	Reduced weight‐bearing and immobility	DXA BMD LS, FN, intertrochanteric and total femur; PTH, creatinine and sclerostin
Sutor, 2022	Review	Complete and incomplete SCI	Reduced weight‐bearing and immobility	None statistically significant	Dynamic bone loading
Trinh, 2016	Longitudinal retrospective	CP, n = 45	Reduced weight‐bearing and immobility; Bone mineral content and areal BMD	DXA BMD FN, LS and total body
Trinh, 2018	Cross‐sectional	CP, n = 43	Reduced weight‐bearing and immobility	DXA BMD LS and total body score
Trinh, 2019	Longitudinal retrospective	CP, n = 45	Reduced weight‐bearing and immobility	DXA areal BMD LS, TH and total body
Vestergaard, 2001	Cross‐sectional	DMD, Becker's MD, SMA II and III, n = 229	Reduced weight‐bearing and immobility; wheelchair use		Weight‐bearing exercise, bisphosphonates and calcium
Vincze. 2021	Review	Idiopathic inflammatory myopathy	Reduced weight‐bearing and immobility; fractures	BMD	Dynamic bone loading
Won, 2021	Cross‐sectional	Cerebral palsy, n = 87	Reduced weight‐bearing and immobility	DXA BMD LS, FN and total femur
Zelikovich, 2022	Clinical trial	Muscular dystrophy, n = 20		None statistically significant	Glucocorticosteroids
Zylstra, 2008	Cross‐sectional	Mixed, n = 298	Reduced weight‐bearing and immobility	DXA BMD

Grey box indicates study was not relevant to that category. BMD, bone mineral density; CP, cerebral palsy; DMD, Duchenne's muscular dystrophy; DXA, dual‐energy X‐ray absorptiometry; FN, femoral neck; LS, lumbar spine; SCI, spinal cord injury; SMA, spinal muscle atrophy; TH, total hip.

Reduced weight‐bearing and immobility were the most significant risk factors for bone disease across neuromuscular conditions, including CP, SCI and muscular dystrophy subgroups. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ Non‐ambulatory CP populations had significantly lower BMD Z‐scores at the lumbar spine, femoral neck, second metacarpal and total hip compared with ambulatory individuals with CP. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁵ A graded relationship between GMFCS severity and decreasing BMD was observed, with GMFCS identified as a strong predictor of BMD. ⁶ , ⁷ , ¹¹ Lower BMD correlated with a greater than three‐fold increased fracture risk. ¹⁵

After complete SCI, reduced weight‐bearing and immobility led to rapid, significant BMD loss within months to years after injury. ¹⁶ , ¹⁷ , ¹⁹ , ²⁰ BMD reductions were observed at the proximal tibia and femoral bones (cortical and trabecular) relative to baseline at the time of injury. Bone loss decelerates over time, eventually reaching a steady state. ²⁰ This chronic, irreversible disuse osteoporosis increases fracture risk 20‐ to 100‐fold compared with the general population. ¹⁷

Disuse osteoporosis is defined by absent neural innervation, lack of osteogenic weight‐bearing stimulus and muscle atrophy, resulting in bone resorption and lower BMD Z‐scores. ¹⁷ , ²⁰

In muscular dystrophy and mixed neuromuscular populations, reduced weight‐bearing and immobility were the primary risk factors for osteoporosis, falls and fracture risk, particularly in the lower extremities, compared with ambulatory individuals. ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ Vincze et al. ²⁷ reported a two‐fold increase in fracture risk per standard deviation decrease in BMD, consistent with patterns seen in the general ambulatory population without neuromuscular conditions.

Prior fractures were significant predictors of bone fragility and lower BMD Z‐scores compared with no fracture history. ¹⁴ , ²² , ²⁷ , ²⁸ Vertebral fractures were early indicators of fragility and were associated with progressive deformity, pain and impaired physical health. ¹⁴ , ²⁷ , ²⁹

Wheelchair use was associated with increased falls and fractures compared with ambulatory individuals. ²⁸ , ²⁹ , ³⁰ Wheelchair users also had lower serum 25‐OH vitamin D, higher rates of osteoporosis and a 4.2‐fold increased fracture risk. ²³ , ²⁴

Diagnosis

A total of 12 statistically significant diagnostic measures were identified (Table 2): dual‐energy X‐ray absorptiometry (DXA) BMD scans (n = 11); serum biomarkers (n = 1).

DXA scanning was the most common statistically significant diagnostic tool for assessing bone health. Sites assessed were heterogeneous but included the lumbar spine (L1–L4), femoral neck, total hip and total body. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁵ , ²¹ , ²² , ²⁷ Within a single study, measurements at multiple sites were common, and BMD changes were often discordant across anatomical regions. ⁷ , ⁹ , ¹⁰ , ¹¹ , ¹³ , ¹⁵ Cross‐sectional studies strongly supported DXA use in CP populations, whereas two studies in muscular dystrophy found no significant associations with DXA use, highlighting gaps in the literature. ¹⁸ , ³¹

Evidence for plasma biomarkers as indicators of bone health remains inconclusive. In non‐ambulatory CP patients, lower parathyroid hormone and creatinine and higher sclerostin levels correlated with significantly lower BMD Z‐scores at the lumbar spine and femoral sites. ¹² Generally, these biomarkers reflect calcium homeostasis rather than BMD. While three reviews supported their use, four studies found no association. ⁹ , ¹³ , ¹⁷ , ²³ , ²⁶ , ³¹ , ³² Other biomarkers – including serum calcium, 25‐OH vitamin D and alkaline phosphatase – lacked statistical significance, and diagnostic methods such as radiography and clinical examination were not supported by evidence in the target populations. ⁸ , ⁹ , ¹² , ¹³ , ¹⁶ , ¹⁷ , ¹⁹ , ²⁰ , ²² , ²³ , ²⁵ , ²⁶ , ³²

Management

A total of 15 studies (51.7%) evaluated management interventions (Table 2), including neuromuscular electrical stimulation (n = 6; 40.0%), bone‐loading exercises (n = 4; 26.7%) and vitamin D, calcium and bisphosphonates (n = 2; 13.3%).

Studies of neuromuscular electrical stimulation in SCI showed limited evidence for significant BMD increase effects. ¹⁸ , ¹⁹ , ²⁸ , ³³ However, Carvalho et al. ¹⁶ reported increased BMD and reduced bone resorption markers in muscular dystrophy patients using an electrical stimulation gait system compared with controls. Similarly, Shields et al. ²⁰ observed distal tibial BMD gains post‐SCI compared with the untrained limb, though the levels remained less than those at the time of injury.

Evidence for weight‐bearing exercise was mixed: several studies supported its role in reducing bone loss and fracture risk, ⁸ , ²⁴ , ²⁷ while others found no significant effects across GMFCS levels. ¹⁴ , ¹⁷

Review articles on Duchenne Muscular Dystrophy (DMD) populations recommend vitamin D, calcium and bisphosphonates to stabilise BMD and reduce fracture risk, although high‐quality evidence is lacking. ²⁶ , ³² No studies to date have evaluated the role of potent anti‐resorptive agents, such as denosumab, in adult non‐ambulatory neuromuscular populations.

Discussion

Evidence‐based guidelines for bone health in neuromuscular disease in non‐ambulatory adults are scarce, limiting consensus on optimal risk assessment, diagnosis and management strategies. ³

Risk factors

Reduced weight‐bearing and immobility were the most significant risk factors for osteoporosis and fractures in a non‐ambulatory neuromuscular population. A graded, dose–response relationship was observed, whereby greater loss of weight‐bearing capacity correlated with higher rates of osteoporosis and fracture risk, both outcomes of bone disease. Neurogenic disuse osteoporosis was hypothesised as the underlying mechanism, whereby chronic loss of neural stimulation and mechanoreceptor unloading lead to increased osteoclast activity and bone resorption. This process is accompanied by markedly impaired osteogenesis, mediated in part by the inhibitory effects of sclerostin, a potent inhibitor of the Wnt signalling pathway. ¹⁷ Microstructural bone damage, muscle atrophy and chronic inflammation further elevate osteoporosis and fracture risk. ²⁷ Distinguishing non‐weight‐bearing osteoporosis from weight‐bearing osteoporosis in the literature is clinically important, as it necessitates distinct approaches to evaluation and management. ²

Depending on the underlying pathophysiology, disease progression to immobility may be unmodifiable, with lower motor neuron lesions conferring a higher risk of osteoporosis due to loss of neural innervation. Risk stratification by diagnosis and motor neuron lesion type may therefore assist prognosis and management. However, the sparse literature in this population limits development of targeted interventions addressing modifiable risk factors, including fall prevention during daily activities such as wheelchair transfers. ²⁶ , ³² In CP, the hemiplegic pattern often preserves ambulation more than the diplegic pattern, decreasing the risk profile for wheelchair use and falls during transfers. ³ , ²⁹ Standard osteoporosis risk factors, including glucocorticoid use, family history, menopause, smoking and alcohol, should also be considered. ²

The effects of weight‐bearing loss on BMD are paralleled in spaceflight due to the absence of gravity. Astronauts demonstrate significant tibial BMD loss when compared before and 12 months after spaceflight, with recovery remaining uncertain. Weight‐bearing bones decline at a rate of 1%–1.5% BMD per month in space, consistent with a dose–response relationship. Resistive exercise alone is insufficient to prevent hip trabecular BMD loss, likely due to persistent osteoclast activity; however, combining resistive exercise with bisphosphonates effectively prevents demineralisation. ³⁴ These findings emphasise the critical role of weight‐bearing stimuli in maintaining BMD and targeted intervention.

Long‐bone fractures of the distal lower extremities were the predominant fracture type across non‐ambulatory neuromuscular populations, consistent with loss of ambulation, reduced mechanical loading, and transfer‐related trauma. This distribution contrasts with the weight‐bearing proximal fracture patterns observed in postmenopausal osteoporosis.

In DMD, lumbar vertebral fractures and distal femoral fractures were frequently reported. Reduced BMD at the lateral distal femur has been associated with increased fracture risk in this group. Vertebral fractures were strongly associated with long‐term glucocorticoid exposure and progressive postural kyphosis resulting in compression of the upper lumbar vertebrae. In contrast, the relationship between glucocorticoid therapy and limb fractures, particularly femoral fractures, remains less clearly defined. ²⁵ , ³⁵

In paraplegic SCI populations, fractures were predominantly localised to the lower limbs, most commonly involving the epiphyseal regions of the femur and tibia. This distribution aligns with immobilisation‐related bone loss, with a reported mean time to first fracture of 8.9 ± 1.4 years and an annual fracture incidence of approximately 2.2%. ³⁶ , ³⁷ Similarly, cross‐sectional studies in limb‐girdle muscular dystrophy and sporadic inclusion body myositis have identified tibial fractures as a prominent pattern. ³⁸

Among individuals with CP classified as GMFCS IV/V, lower extremity fractures, particularly at the distal femur, were overwhelmingly the most common, followed by vertebral column and upper extremity or hip fractures. These injuries were frequently associated with wheelchair use or transfer‐related activities. ²⁹

The fracture phenotype in non‐ambulatory neuromuscular populations differs from postmenopausal osteoporosis, which commonly involves the proximal femur, distal forearm and lumbar vertebrae. The differences in distribution reflect age‐related bone loss combined with fall mechanics and axial loading patterns. ³⁹ , ⁴⁰ Vertebral fractures represent an area of overlap, as progressive reductions in BMD contribute to fracture risk across both groups. Collectively, these distinct fracture distributions highlight potential limitations in applying risk stratification models and treatment paradigms derived from postmenopausal osteoporosis to non‐ambulatory neuromuscular populations.

Poor nutrition resulting in low body mass, including reduced lean muscle mass, likely contributes as a risk factor to osteoporosis in this population. Specific nutrient deficiencies, such as inadequate intake of calcium, vitamin D and other essential micronutrients, may further exacerbate bone fragility and should be considered on an individual basis and suggested to be optimised with a dietician. ¹⁴ , ²⁶ , ³⁵ , ⁴¹ However, there is a paucity of literature examining the relationship between nutrition and bone health in adult, non‐ambulatory neuromuscular populations.

Diagnosis

DXA BMD scanning was the most frequently cited diagnostic tool, although the assessed anatomical sites varied considerably. Discordant patterns of BMD loss may reflect disease‐specific presentations. The only statistically significant BMD loss was observed in cross‐sectional studies of CP populations, particularly at the lumbar spine and femoral hip. Non‐ambulatory status introduces logistical challenges for DXA accuracy, including positioning difficulties. DXA remains the World Health Organisation gold standard for BMD assessment, with osteoporosis defined as a T‐score < −2.5 standard deviations at the lumbar spine, femoral neck or total hip, or the occurrence of fragility fractures following minimal trauma. ² Novel techniques such as peripheral quantitative CT (pQCT) provide detailed bone geometry and microstructure analysis but are limited by cost, accessibility and lack of normative data. ³⁴

Serum biomarkers were heterogeneous and inconsistently applied, showing no reliable correlation with DXA BMD. The lack of high‐quality evidence, small sample sizes and variable disease pathophysiology contribute to the absence of statistically significant biomarkers. Key mediators such as sclerostin, vitamin D and parathyroid hormone may be disrupted in bone resorption pathways. ² Serum testosterone, known to influence BMD, was not evaluated in the included studies. In SCI, testosterone levels are significantly lower than in non‐SCI individuals, with an acute decline immediately post‐injury that may persist for 1–2 years. ⁴²

Accurate assessment of bone health requires integrating multiple diagnostic indicators – clinical evaluation and investigations – since BMD is influenced by bone microarchitecture, geometry and homeostatic mineralisation within the context of underlying pathophysiology and environmental stimuli. ¹⁰

Management

This review highlights a critical gap in effective strategies for managing bone health in non‐ambulatory adults. Identified strategies include neuromuscular electrical stimulation and weight‐bearing stimulus. Neuromuscular electrical stimulation was largely ineffective in SCI, which may be due to the pathophysiology of acute onset and lower‐motor neuron injury limiting osteogenesis. Shields et al. ²⁰ reported it was superior to no intervention but insufficient to restore pre‐SCI BMD. No studies evaluated this approach in other neuromuscular conditions, limiting comparisons between upper‐ and lower‐motor neuron lesion disease types. Dynamic bone loading and weight‐bearing exercise may help preserve BMD in CP, SCI and inflammatory myositis populations, but further research is needed to confirm effectiveness and feasibility.

Exoskeletal robotics and robotic‐assisted mobility for weight‐bearing support in SCI show limited evidence of BMD benefit. DXA assessments demonstrated no significant post‐intervention BMD improvements; however, these interventions may attenuate ongoing BMD loss.

Therapeutic strategies involving vitamin D, calcium, and bisphosphonates demonstrated statistically significant efficacy in stabilising BMD and reducing fracture risk in DMD and SCI populations. ³² Bisphosphonates prevent BMD loss in SCI but do not significantly affect serum biomarkers at 12 months. Evidence for bisphosphonate use beyond DMD and SCI non‐ambulatory populations was lacking, despite their frequent prescription in clinical practice. ² Notably, no studies evaluated the efficacy or safety of other anti‐osteoporotic agents, such as the anabolic therapies teriparatide and romosozumab, underscoring a gap for future research. ³⁴

Based on the limited evidence available, clinicians should aim to maximise and prolong ambulation and weight‐bearing stimulus through disease‐specific management and bisphosphonate therapy consistent with standard osteoporosis management. ² Achieving peak BMD and maintaining it over time reduces osteoporosis and future fracture risk; however, once ambulation is lost, strategies to replicate weight‐bearing stimulus demonstrate limited effectiveness. A further barrier to optimal care is poor follow‐up rates for osteoporosis, driven by the absence of standardised care pathways.

Limitations

The literature on adult neuromuscular non‐ambulatory conditions is limited by low‐quality evidence, with most studies being observational or review‐based and constrained by small sample sizes. The rarity of these conditions necessitated broad inclusion criteria; however, heterogeneity in study designs, populations and outcome measures limited meaningful comparisons and the ability to establish meaningful comparisons and consensus. The absence of randomised controlled trials further restricts the translation of findings into safe, evidence‐based conclusions and interventions.

Classification of non‐ambulatory status was inconsistent across studies. Presentations of adult, neuromuscular, non‐ambulatory conditions vary widely by genetic, environmental and pathophysiological factors, as well as disease onset (e.g. progressive childhood‐onset DMD versus acute adult‐onset SCI), progression and distribution (e.g. upper motor neuron involvement in CP versus lower motor neuron involvement in SCI). Studies often fail to specify participants’ age at loss of ambulation, which is relevant as individuals who achieve adult peak BMD prior to non‐ambulation may retain a greater physiological bone reserve, mitigating fracture risk. Variability in disease severity, functional outcomes and non‐ambulatory assessment scales further compounded these limitations. Additionally, this review is limited by screening conducted by a single reviewer and the lack of a formal risk of bias assessment.

Finally, 17 of the included studies reported a male predominance, reflecting the higher incidence of CP, SCI and DMD in males. As males generally attain higher peak BMD, these findings may have lower generalisability to females. Despite these limitations, this review highlights key gaps in risk factors, diagnostic approaches and management strategies, underscoring the need for higher‐quality research in this high‐risk population.

Conclusion

In summary, there is limited evidence to optimise bone health in adult neuromuscular non‐ambulatory populations. The loss of weight‐bearing and immobility were the most significant risk factors for osteoporosis in this population, with modifiability determined by underlying disease pathophysiology.

DXA was the key diagnostic modality; however, further research is needed to define optimal anatomical sites and validate serum biomarkers. A standardised, validated bone health assessment tool that integrates key risk factors and investigations may improve diagnosis and management.

Evidence for effective management interventions was minimal. Electrical neuromuscular stimulation and robotic‐assisted weight‐bearing have limited impact on BMD, while therapeutics, particularly bisphosphonates, showed efficacy in DMD and SCI populations. Further research is needed to establish evidence‐based clinical guidelines to optimise bone health in this high‐risk population.

Appendix I. Scopus search strategy

I.1.

Search run on 16/5/2022

#1 (TITLE‐ABS‐KEY (‘muscular dystrophy’ OR neuromuscular OR ‘cerebral palsy’ OR myopathy OR neuropathy) AND TITLE‐ABS‐KEY (ambulat* OR ‘weight‐bearing’ OR ‘load bearing’ OR mobil*) AND TITLE‐ABS‐KEY (osteoporo* OR osteopenia OR ‘bone strength’ OR ‘bone fragility’ OR ‘bone loss’ OR ‘bone mineral density’ OR ‘bone mineral content’ OR ‘bone density’ OR ‘bone deformity’ OR fracture)) AND PUBYEAR >1989 AND PUBYEAR > 1989 (1030).

The search strategy was run for all databases on 16/5/2022. The strategy was duplicated between databases, adjusting for formatting.

Appendix II. Results data extraction—risk factors

II.1.

Author, year	Study design	Neuromuscular diagnosis, sample size	Risk factors associated with bone disease	Evidence	Risk factors associated with bone disease lacking evidence
Fowler, 2015	Cross sectional prospective	Cerebral palsy, n = 46	Reduced weight‐bearing and immobility	Significantly lower BMD Z‐score in lumbar spine, femoral neck and total hip regions compared to ambulators. Positive association between lumbar spine BMD and GMFCS non‐ambulatory level	scoliosis, joint contractures, abnormal biomechanics, anticonvulsant use, alkaline phosphatase elevated in men, BMI. NOT age
Won, 2021	Cross sectional	Cerebral palsy, n = 87	Reduced weight‐bearing and immobility	Significantly lower BMD Z‐score in lumbar spine and femoral neck compared to ambulators. Positive association between lumbar spine BMD and GMFCS non‐ambulatory level	Bilateral cerebral palsy compared unilateral cerebral palsy. Negative association age and BMD
Nakano, 2003	Cross sectional	CP, n = 123	Reduced weight‐bearing and immobility	12% significantly lower BMD in second metacarpal compared to ambulators in females. 28% significantly higher serum alkaline phosphatase associated with lower second metacarpal BMD in men	Anticonvulsant use associated with risk of osteomalacia
Trinh, 2018	Cross sectional	CP, n = 43	Reduced weight‐bearing and immobility	Significantly lower trabecular bone score and lumbar spine BMD compared with ambulators	Hypogonadism, anticonvulsant use, malnutrition, lean mass
Trinh, 2019	Longitudinal retrospective	CP, n = 45	Reduced weight‐bearing and immobility	Significant negative association with areal BMD at lumbar spine, total hip and total body and GMFCS compared to ambulators. Significant decrease in lumbar spine areal BMD compared to ambulators	Subtype of CP, low BMI, hypogonadism, decreased lean muscle mass and age
Marciniak, 2016	Chart review retrospective	CP, n = 42	Reduced weight‐bearing and immobility	Significant negative association with BMD Z‐score at the left hip and GMFCS/non‐ambulatory status compared with ambulators
Shin, 2017	Cross sectional	CP, n = 28	Reduced weight‐bearing and immobility	Significantly lower BMD Z‐scores at lumbar spine and femoral regions, lower parathyroid hormone and creatinine, and higher plasma sclerostin compared to ambulators	Dysregulated muscle tone and posture
Kim, 2015	Cross sectional retrospective	CP, n = 58	Reduced weight‐bearing and immobility	Significantly lower BMD Z‐score in femoral, trochanter and total hip regions compared to ambulators	Poor nutrition and anticonvulsant medication
			spastic CP	Lowest BMD Z‐score total hip region compared all CP subtypes and ambulation status
Gannotti, 2021	Review	CP	Reduced weight‐bearing and immobility	Decrease in BMD compared healthy, age matched controls. Increased fracture risk.	Preterm birth/low birth weight, BMI, nutrition intake, anticonvulsant use, physical activity, co‐morbidities
			Fractures	Progressive deformity, pain and spinal stenosis
Trinh, 2016	Longitudinal retrospective	CP, n = 45	Reduced weight‐bearing and immobility	Significant inverse association with one unit decrease of femoral neck Z‐score BMD, increases fracture risk by 3.2‐fold. GMFCS is best predictor of lumbar spine, femoral neck and total body Z‐score BMD	Weight, age, decreased lean tissue mass
			Bone mineral content and areal BMD	Significant association at lumbar spine, femoral neck and total body Z‐scores with fractures
Carvalho, 2006	Experimental trial	Complete SCI, n = 21	Reduced weight‐bearing and immobility	BMD decreases significantly 3–4 months post injury, increasing bone fragility and non‐traumatic fracture risk
Sutor, 2022	Review	Complete and incomplete SCI	Reduced weight‐bearing and immobility	50%–100% lower BMD at proximal tibia and 40%–80% lower cortical femoral BMD 2–3 years post injury. 20 to 100× increased fracture risk compared to general population
Needham‐Shropshire, 1997	Experimental trial	Complete SCI, n = 16	Reduced weight‐bearing and immobility	Resultant osteopenia and osteoporosis, and non‐traumatic fractures
Jacobs, 2004	Review	Complete SCI	Reduced weight‐bearing and immobility	50% BMD loss at one year post injury, greatest loss at supracondylar femoral notch. Rapid bone demineralisation in first year, then slower decay. Increased urinary excretion calcium and hydroxyproline
Shields, 2006	Experimental trial	Complete SCI, n = 7	Reduced weight‐bearing and immobility	Post injury, bone demineralisation 2%–4% per month until reaching steady state for fracture threshold. Destruction of trabecular lattice leading to irreversible, chronic disuse osteoporosis
Zylstra, 2008	Cross sectional	Mixed, n = 298	Reduced weight‐bearing and immobility	Significantly greater rate of osteoporosis compared to ambulators	Older age, Caucasian
King, 2003	Cross sectional	Mixed, n = 48	Reduced weight‐bearing and immobility	Significantly lower BMD Z‐scores at lumbar spine compared to age and sex matched controls	NOT age
			Fractures	Previous fracture significant predictor of lower lumbar spine BMD Z‐score compared to no facture history
Danckworth, 2014	Cross sectional	Mixed, n = 89	Reduced weight‐bearing and immobility	Increased risk of lower BMD, increased fractures and falls	Chronic neurological conditions, older age, low BMI, post‐menopausal women, current smoking, glucocorticosteriod use, previous fractures and falls
			Wheelchair use	Increased risk of osteoporosis and decreased 25, OH Vitamin D3 levels in patients with myopathy compared with age and gender matched non‐myopathic disorder patients
Vestergaard, 2001	Cross sectional	DMD, Becker's MD, SMA II and III, n = 229	Reduced weight‐bearing and immobility	Significant increase fracture risk ratio by 21.4 times to lower legs, especially femurs
			Wheelchair use	Significant increase fracture increase risk ratio by 4.2 times
Apkon, 2018	Review	DMD	Reduced weight‐bearing and immobility	Increased fracture risk with progressive muscle weakness, loss of function, joint contractures, scoliosis and osteoporosis	Glucocorticoid use
Bushby, 2010	Review	DMD	Reduced weight‐bearing, immobility and muscle weakness	Long bone and vertebral fractures, osteopenia, osteoporosis, kyphoscoliosis, bone pain and reduced quality of life
			No glucocorticoid use	Significant 90% chance of progressive scoliosis
			Glucocorticoid use	Increase risk long bone and vertebral fractures
Vincze. 2021	Review	Idiopathic inflammatory myopathy	Reduced weight‐bearing and immobility	Fracture risk increases two‐fold each standard deviation Z‐score decrease in BMD	Glucocorticoid use
			Fractures	Poorer physical health function and status assessed via questionnaire
Birnkrant, 2018	Review	DMD	Fractures	Vertebral fractures early manifestation of bone fragility. Predictive of future fractures
Kannikeswaran, 2021	Cross sectional	CP, n = 58	Wheelchair use, transfer or limb stuck	40% of fracture causes compared to 2.9% of ambulators	Fracture of unknown cause or discovered due to change in behaviour post time of fracture 33.3%. Playing, falls and sports also cause of 16.7% of fractures
Rice, 2018	Qualitative semi‐structured interviews	Mixed, predominantly CP, n = 19	Falls risk	60% of wheelchair users fall during their lifetime, greatest falls risk during transfers, leading to fractures and poor physical recovery
Morgenroth, 2012	Review	DMD	Delayed puberty	Lower BMD and increased fracture risk in men	Age

Appendix III. Result data extraction—diagnostic strategies

III.1.

Author, year	Study design	Neuromuscular diagnosis, sample size (n)	Outcome measures used	Outcome measure statistical significance	Evidence/Explanation
Zylstra, 2008	Cross sectional	Mixed, n = 298	DXA BMD	DXA BMD
Vincze, 2021	Review	IIM	BMD	BMD
Fowler, 2015	Cross sectional	CP, n = 46	DXA BMDLS, femoral neck, total hip	DXA BMD LS, FN, TH
Won, 2021	Cross sectional	CP, n = 87	DXA BMD LS, FN, total femur and total body composition	DXA BMD LS, FN and total femur
Marciniak, 2016	Chart review, retrospective	CP, n = 42	DXA BMD LS, total spine, hip, TN and femoral total	DXA BMD left total hip
Trinh, 2018	Cross sectional cohort retrospective	CP, n = 43	DXA areal LS L2‐4, FN and total body	DXA BMD LS and total body score	OP diagnosis based on BMD criteria or minimal trauma fracture. Bone strength is determined by multiple factors including bone density, geometry, microarchitecture, turnover and mineralisation
Trinh, 2016	Longitudinal study, retrospective	CP, n = 45	DXA BMD LS L2‐4, FN, total body and proximal hip	DXA BMD FN, LS and total body
Kim, 2015	Cross sectional	CP, n = 58	DXA BMD LS L1‐4, femoral trochanter, TH	DXA BMD femoral trochanter, TH
			C‐telopeptide of type I collagen, osteocalcin, calcium, phosphate, FBC, intact PTH	None
King, 2003	Cross sectional	CP, n = 48	DXA BMD LS	DXA BMD LS
			Serum 25‐OH, vitamin D, calcium, phosphate, magnesium and osteocalcin	None
Trinh, 2019	Cross sectional retrospective	CP, n = 45	DXA areal BMD LS L2‐4, FN, TH, total body and bone quality	DXA areal BMD LS, TH and total body
			Serum calcium, phosphate, PTH, alkaline phosphatase and 25‐OH vitamin D	None
Nakano, 2003	Cross sectional	CP, n = 123	DXA BMD second meta‐carpal	DXA BMD second meta‐carpal in women only
			Serum calcium, alkaline phosphatase, phosphate	None
			BMI	None
Shin, 2017	Cross sectional	CP, n = 28	DXA BMD LS L1‐4, FN, intertrochanteric and total femur	DXA BMD LS, FN, intertrochanteric and total femur
			Sclerostin, PTH, osteocalcin, T terminal telopeptide type I collagen (Ctx), 25‐OH vitamin D, creatinine, calcium, phosphate	Ctx, PTH, creatinine and sclerostin
Bushby, 2010	Review	DMD	AP and lateral full spine radiography	Not stated	At time of wheelchair dependency and 6–12 monthly depending on degree scoliosis, kyphoscoliosis or back pain indicated
			Serum calcium, phosphate, alkaline phosphatase, 25‐OH vitamin D, magnesium and PTH; Urine calcium, sodium, creatinine	Not stated	Magnesium and PTH as clinically indicated
			DXA scan	Not stated	Baseline and annually as clinically indicated
			Left wrist radiography	Not stated	Bone age growth
Birnkrant, 2019	Review	DMD	Indicators of back pain, fractures or bone fragility	Not stated	Each clinical visit
			Serum calcium, phosphate, magnesium, alkaline phosphatase, PTH	Not stated	Baseline and as appropriate
			Calcium, vitamin D intake, spine DXA BMD, serum 25‐OH D3	Not stated	Baseline and annually
			Lateral thoracolumbar spine radiograph	Not stated	Baseline and 1–3 years depending steroid use or BMD Z‐score decline
Carvalho, 2006	Experimental trial	Complete SCI	DXA BMD LS, FN, trochanteric area and total femur	Not stated
			Serum osteocalcin, alkaline phosphatase, pyrinoline and deoxypyridinoline	Not stated
Sutor, 2022	Review	Complete SCI	DXA BMD LS, hip, distal femur and proximal tibia	Insufficient evidence
			Serum osteocalcin and sclerostin	Not stated
Needham‐Shropshire, 1997	Experimental trial	Complete SCI	DXA BMD FN, femoral head, Ward's triangle	None	Measured using dual photon densitometer
			Urine calcium and hydroxyproline	Not stated
Jacobs, 2004	Review	Complete SCI	Bone radiograph	Not stated	Progressive rarifying radiographs
Kannikeswaran, 2021	Cross sectional	CP, n = 58	BMD	Not stated	Stratify fracture risk
Morgenroth, 2012	Review	DMD	DXA BMD	Not stated	DXA BMD valuable diagnostic tool
Shields, 2006	Experimental trial	Complete SCI	Peripheral qCT	Not stated
Zelikovich, 2022	Clinical trial	Muscular dystrophy, n = 20	DXA BMD LS L1‐4, left forearm and total body	None
			Serum creatinine kinase	None
			Vignos lower extremity scale	None	Functional outcome
Apkon, 2018	Review	DMD	Focussed musculoskeletal examination	Not stated	Baseline and biannual
Danckworth, 2014	Cross sectional	Mixed, n = 89	Serum calcium, 25‐OH vitamin D, PTH, creatinine, alkaline phosphate	None
			Clinical history of falls, fractures, osteoporosis and related factors	None
			Brief pain inventory	None

Appendix IV. Results data extraction—management strategies

IV.1.

Author, year	Study design	Neuromuscular diagnosis, sample size (n)	Intervention	Evidence/Explanation
Graupe, 2002	Review	Complete SCI	Neuromuscular electrical stimulation gait training	No evidence intervention effects BMD
Needham‐Shropshire, 1997	Experimental trial	Complete SCI, n = 16	Neuromuscular electrical stimulation gait training	No evidence intervention effects BMD post intervention
Jacobs, 2004	Review	Complete SCI	Neuromuscular electrical stimulation gait training	Limited evidence intervention effects BMD
Carvalho, 2006	Experimental trial	Complete SCI, n = 21	Neuromuscular electrical stimulation gait training	Significant increase in BMD and significant decrease in bone resorption markers within subjects compared to control group
Asselin, 2011	Experimental trial	Complete SCI, n = 17	Neuromuscular electrical stimulation	No evidence intervention prevents bone loss or decreases fractures compared to controls
Shields, 2006	Experimental trial	Complete SCI, n = 7	Neuromuscular electrical stimulation	Attenuated BMD loss as 31% increase in distal tibial BMD compared to untrained limb
Gannotti, 2021	Review	CP	Dynamic bone loading	Unclear evidence for improvement of BMD across different exercise programs at different GMFCS levels
Sutor, 2022	Review	Complete SCI	Dynamic bone loading	Limited evidence intervention attenuates BMD loss of decreases fracture risk
Vincze, 2021	Review	Idiopathic inflammatory myopathy	Weight‐bearing exercise	Increase osteoblast activity, improve muscle strength, physical performance, balance, posture and decrease falls risk
Nakano, 2003	Cross sectional	CP	Weight‐bearing exercise	Minimise bone loss and decrease fracture risk
Bushby, 2010	Review	DMD	Vitamin D, calcium, bisphosphonates	Not stated
Birnkrant, 2018	Review	DMD	Vitamin D, calcium, bisphosphonates	Stabilisation phase, progression to maintenance phase with titration of bisphosphonates when BMD Z‐score > −2 SD or absence new fractures
Vestergaard, 2001	Cross sectional	Mixed	Weight‐bearing exercise, bisphosphonates and calcium	Maintain BMD in the lower extremities
Zelikovich, 2022	Clinical trial	Muscular dystrophy, n = 20	Glucocorticosteriod	No evidence total body BMD is affected post intervention for combined ambulatory and non‐ambulatory subjects
Apkon, 2018	Review	DMD	Fracture reduction, stabilisation and pain control	Clast or splint for distal femoral metaphyseal fracture, operative stabilisation for proximal femoral fracture

Data availability statement

The data that support the findings of this study are available in the appendix sections of this article.