Neurogenic Obesity and Skeletal Pathology in Spinal Cord Injury

Abstract

Spinal cord injury (SCI) results in dramatic changes in body composition, with lean mass decreasing and fat mass increasing in specific regions that have important cardiometabolic implications. Accordingly, the recent Consortium for Spinal Cord Medicine (CSCM) released clinical practice guidelines for cardiometabolic disease (CMD) in SCI recommending the use of compartmental modeling of body composition to determine obesity in adults with SCI. This recommendation is guided by the fact that fat depots impact metabolic health differently, and in SCI adiposity increases around the viscera, skeletal muscle, and bone marrow. The contribution of skeletal muscle atrophy to decreased lean mass is self-evident, but the profound loss of bone is often less appreciated due to methodological considerations. General-population protocols for dual-energy x-ray absorptiometry (DXA) disregard assessment of the sites of greatest bone loss in SCI, but the International Society for Clinical Densitometry (ISCD) recently released an official position on the use of DXA to diagnose skeletal pathology in SCI. In this review, we discuss the recent guidelines regarding the evaluation and monitoring of obesity and bone loss in SCI. Then we consider the possible interactions of obesity and bone, including emerging evidence suggesting the possible influence of metabolic, autonomic, and endocrine function on bone health in SCI.

Introduction

In the United States, there are 294,000 persons living with spinal cord injury (SCI),¹ and the World Health Organization estimates a global incidence rate of 250,000 to 500,000 new cases yearly.² Somatic motor and sensory impairments are only some of the health impacts of SCI, and neurological deficits result in changes in bodily functions that accelerate risk of cardiometabolic disease (CMD). Accordingly, the Consortium for Spinal Cord Medicine (CSCM) recently released clinical practice guidelines for CMD in SCI,³ with obesity being the most prevalent CMD risk factor in SCI.⁴ The CSCM guidelines emphasize the importance of measurements of body composition in SCI due to regionally specific changes in adiposity that occur after SCI. The impact of SCI on skeletal muscle function can be determined by neurological exam⁵; muscle atrophy is readily observable, but changes in bone can be more difficult to detect. Recently, the International Society for Clinical Densitometry (ISCD) released an official position on skeletal pathology in SCI.⁶ The ISCD guidelines emphasize the importance of measuring specific skeletal regions that are at highest risk in SCI and endorse a new protocol for doing so. This review will begin by discussing clinical considerations for diagnosing obesity and skeletal pathology in SCI in light of these new guidelines.^3,6 We then consider the possible interaction of these metabolic consequences of SCI. This issue of Topics in Spinal Cord Injury Rehabilitation covers obesity in SCI; thus this review will emphasize skeletal pathology in SCI and the potential role of disordered fat metabolism in contributing to changes in bone after SCI.

Clinical Considerations for Obesity in Spinal Cord Injury

SCI results in dysregulation of fat metabolism^7–10 that increases the risk of morbidity and mortality¹¹ from CMD.^12–26 Central obesity in SCI^27–32 is the most serious risk component for this disease state.^33–35 SCI also increases CMD risk factors of dyslipidemia,^24,36–41 hypertension (depending on extent and level of injury), and insulin resistance.^14,38,42 Importantly, this population displays a unique clustering of CMD component risks with a higher prevalence of risks related to fat metabolism.^4,31 The most common risk factor for CMD in both paraplegia and tetraplegia is increased body mass index (BMI).⁴ At the established BMI cutoff of >25 kg/m², 53% to 60% of persons with SCI are considered overweight,^{15,31,43–48} which would appear to be a lower incidence of “overweight” than in the general population.⁴⁹ However, changes in body composition following SCI result in greater adiposity for a given BMI,^50–55 and therefore population-specific BMI cutoffs have been developed. Studies using an SCI-specific BMI cutoff of ≥22 kg/m² show an obesity prevalence ~75%.^3,4,17 Despite these adjustments, BMI does not account for the remarkable compositional changes that occur below the level of injury. Below the level of injury, lean tissue mass decreases due to skeletal muscle atrophy^56–58 and bone loss (discussed later). Adipose tissue does not atrophy concomitant to lean tissue, resulting in higher whole-body fat percentage.^50–55 Furthermore, fatty deposits encroach into the visceral compartment,^29,55,59,60 skeletal muscle,^55,56 and bone marrow.^61,62

Due to the complexities of changes in body composition in SCI, the CSCM clinical practice guidelines for CMD recommend using three- or four-compartment models of body composition to determine obesity in adults with SCI.³ A whole-body fat (BF) percentage of >22% for adult men and >35% for adult women classifies as obese in SCI.³ Secondarily, the BMI cutoff of ≥22 kg/m² can be used as a surrogate marker of obesity when compartmental assessment of body composition is not available.³ Multiple methods are available for determining body composition in SCI⁶³ including dual-energy x-ray absorptiometry (DXA), which is also the preferred method for monitoring skeletal pathology in SCI. DXA allows for the estimation of visceral adipose tissue (VAT)^27–32 and other regionally specific changes in body composition that are important in SCI, and it does not have the technical and/or monetary constraints of other methods. The clinical utility of DXA for assessing body composition in SCI has been considered elsewhere.⁶⁴ Furthermore, in the context of the topic of this review, DXA offers the ability to assess obesity and skeletal pathology in the same clinical session, which limits the duration and number of clinic visits, transfers, and radiation exposure.

Clinical Considerations for Skeletal Pathology in Spinal Cord Injury

Bone loss after SCI may result in part from lean muscle losses due to denervation, and the majority of persons living with SCI have osteoporosis.^65–68 Similar to skeletal muscle and adipose tissue, there is an important regional specificity to changes in bone as skeletal pathology disproportionately affects lower extremity long bones in this population. Bone is lost precipitously starting very early post injury,⁶⁹ with periarticular hip and knee bone mineral density (BMD) decreasing by 2% to 4% per month^69,70 and declining up to approximately 20%^71,72 within the first year of SCI. CT-based finite element analysis (FEA) modeling has estimated that within the first 5 months post injury, the femoral neck loses an estimated 40% of fracture resilience at the femoral neck.⁶⁹ Loss of BMD and bone strength of the proximal tibia by FEA appear to stabilize by the second year post injury.^70,73 The 2019 ISCD guidelines are the first to standardize knee region scanning in SCI, and therefore there is a paucity of data about the plateau in knee region BMD. It should be noted that BMD losses in other bone regions have been reported for up to 5 years post injury.⁷⁴ A new steady state in BMD is reached after as much as a 20% to 50% decrease in BMD at the periarticular hip and knee region,^{65,66,74–80} with generally no detectable loss at the spine.^{68,75–77,79,81} Preserved BMD at the lumbar spine is a consistent finding in SCI and is possibly due to maintenance of mechanical loading in this region during wheelchair use and activities of daily living. Differences in trabecular and cortical bone composition in vertebral and long bones may also influence this phenomenon. Further research is required to better understand the mechanisms of preserved lumbar spine BMD in SCI.

Fracture incidence rates that approximate two to three fractures per 100 patient-years in SCI^82,83 can be predicted by the rate of bone loss^68,84–88 and contribute to excess morbidity⁸⁹ and mortality.⁸² The distal femur and proximal tibia constitute the most common site for fracture in persons with SCI.^{85,87,90–92} Fracture risk increases with duration of injury,^86,93 with fracture being relatively rare in the first 1 to 3 years and then increasing to peak risk at 20 to 29 years post injury.⁸⁶ Persons with paraplegia are more likely to experience a fracture than persons with tetraplegia.^93,94 The reason for increased lower extremity fractures in paraplegia is incompletely understood, but they commonly occur during transfers to and from wheelchairs.⁹⁵ Lower extremity fracture is also more common than upper; a recent study of 3,365 fractures in SCI shows that upper extremity fractures constitute only 19.8% of all fractures in SCI.⁹⁴ Epidemiological studies of fracture in SCI have come primarily from retrospective observation studies carried out in the Veterans Affairs health care system.^{82,89,93,94,96–98} These studies suggest that opioid⁹⁶ and anticonvulsant⁹⁷ use are associated with increased fracture risk, while thiazide use is associated with decreased fracture risk.⁹⁸ These pharmacological effects are also observed in other populations with osteoporosis. However, large-scale data such as these are rare in SCI research,⁷⁹ and most interventional (e.g., randomized controlled) trials examining fracture risk in SCI are limited by small sample size. The lack of large-scale interventional studies required to understand fracture risk places an importance on quantifying bone loss via changes in BMD.

The early and precipitous loss of bone in specific skeletal regions following SCI emphasizes the importance of early detection and clinical intervention. Recently, the ISCD released an official position on the use of DXA testing to diagnose skeletal pathology in SCI.⁶ The ISCD recommends that, as soon as medically stable, all adults with new SCI have BMD assessed via DXA at the total hip, proximal tibia, and distal femur. The assessment of bone quality versus density has only recently been explored in SCI. Initial reports suggest some discordance between quality as assessed by DXA-derived trabecular bone score (TBS) and BMD.^80,99 The paucity of data in this area highlights the importance of focusing on population-specific BMD regions. The ISCD specifies that results for the total hip, distal femur, and proximal tibia subregions should not come from a segmental analysis of a whole-body scan. Furthermore, an important distinction in the SCI-specific recommendation is the lack of lumbar spine assessment, due to preserved BMD in this region, and the addition of the high-risk knee region. Compared to quantitative computed tomography (QCT), DXA has been established to be sufficiently precise and reliable for use as routine monitoring of knee BMD in SCI.¹⁰⁰ Prior to the ISCD statement, there were no standardized protocols developed by DXA machine manufacturers to measure BMD at the distal femur and proximal tibia. In the absence of manufacturer-provided software for clinical assessment of knee BMD, the ISCD has endorsed the implementation of the Toronto Rehab Protocol⁸⁷ for determination of BMD in the knee region of persons with SCI. Normative data at this region from the SCI population is available through the institution responsible for the development of this protocol.¹⁰¹ The adaptation and rapid implementation of this standardized methodology for quantifying knee BMD in early SCI will be imperative to therapies targeting preservation of bone during stabilization of skeletal and metabolic function. Despite the unique considerations for skeletal pathology in SCI, only one screening paradigm has been proposed,¹⁰² and there currently are no established guidelines for management of skeletal pathology in this population.

Pathophysiology and Potential Mechanisms

The early and precipitous loss of BMD in SCI (2% to 4% per month) exceeds that observed during bed rest¹⁰³ and space flight (1% to 1.5% per month).^104,105 The bone mechanostat hypothesis¹⁰⁶ helps explain why muscle size¹⁰⁷ and strength¹⁰⁸ are partial predictors of bone geometry and density in SCI. However, since the rate and magnitude of change in SCI exceeds that of even the weightlessness of spaceflight, contributions beyond the mechanical unloading must be considered.

Interaction of fat and bone metabolism in SCI

In the general population, obesity can be protective to bone as long as the increased body mass is transmitted as mechanical loading through physical activity.¹⁰⁹ In nonambulatory SCI, this mechanical effect of obesity on the lower extremities is largely absent. Furthermore, the proinflammatory nature of obesity is recognized as a factor contributing to bone pathology in this setting.¹¹⁰ Adipocytes and osteoblasts share common mesenchymal stem cell precursors. Increases in bone marrow fat seen in SCI may mark an increased adipocyte fate decision of mesenchymal stem cells and concomitantly decreased osteoblast precursors.^111,112 The exact mechanisms remain unknown, but recent evidence shows increased cytokine expression in bone after experimental SCI.¹¹² Elevated levels of circulating cytokines, such as tumor necrosis factor alpha (TNF-α), interleukin-1 (IL-1), and interleukin-6 (IL-6), are reported in SCI,^60,113 and these molecules are key mediators of bone resorption.¹¹⁰ Adipokines such as leptin¹¹⁴ and adiponectin¹¹⁵ are increased and decreased, respectively, in SCI and may influence bone metabolism.^113,116 Indeed adiponectin and BMD appear to be associated in SCI.^88,117 It is important to note that VAT, which is increased in SCI,^29,55,59,60 has a unique physiological profile and likely contributes to systemic inflammation.¹¹⁸ Because only one study has directly demonstrated a link between proinflammation and aberrant bone stem cell differentiation, it remains to be determined if inflammatory signaling in SCI is driving changes in bone adiposity.

The skeleton is not only a scaffolding system but it also performs endocrine functions,¹¹⁹ so bone-fat signaling may be occurring in both directions. One highly recognized signaling molecule originating from bone is osteocalcin, a peptide hormone secreted by osteoblasts and noted for its role in regulating energy metabolism.¹²⁰ It is well established that glutaminergic signaling mediates osteocalcin secretion,^121–123 but recent evidence suggests that sympathetic neurons mediate the release of osteocalcin from osteoblasts.¹²² This regulation of the acute stress response by the skeleton occurs in both humans and rodents, with the osteocalcin response peaking rapidly within 5 minutes post stressor and remaining steady for at least 3 hours.¹²² After release by osteoblasts, osteocalcin can act on skeletal muscle to stimulate nutrient uptake and secretion of IL-6.¹²⁴ Muscle-derived IL-6 then serves as a positive feedback for osteoblast secretion of bioactive osteocalcin.¹²⁴ It is possible that aberrant bursts of sympathetic activity (discussed later) is driving this process in SCI, with the combination of these signaling molecules contributing to interactions of obesity and bone in SCI.

Increased adiposity is seen in many tissues in persons with SCI, including bone. Bone marrow adipose volume increases 36% at 12 weeks post SCI,⁶¹ and femoral marrow adiposity is two- to threefold higher in persons with chronic SCI compared to persons without paralysis.⁶² In rodents, SCI leads to chronic bone marrow failure¹¹² due to impaired control of hematopoietic stem and progenitor cell (HSPC) proliferation and sequestration. HSCP and osteoblast precursors interact at different stages of development,^125–127 and there are multiple clinical examples of altered bone phenotypes in primarily hematopoietic disease.¹²⁸ It is possible that bone marrow failure in SCI as demonstrated in this study¹¹² creates a less favorable microenvironment for osteoblast precursors. Bone marrow failure could therefore drive increased differentiation of mesenchymal stem cells into adipocytes and lead to clinically recognizable increases in bone marrow adiposity. Marrow adiposity is related to lower trabecular BMD and prevalent vertebral fractures¹²⁹ and is a predictor of hip fractures in persons with obesity.¹³⁰ Accordingly, it has been proposed that marrow fat contributes to bone pathology in SCI.¹³¹ One potential mechanism of the negative impact of adiposity on bone could be that adipose depots in marrow change fluid shear via changes in flow area and/or the viscosity of marrow itself. Fluid shear has been proposed as a substantial stimulant for osteoprogenitor recruitment in bone marrow,¹⁰¹ and marrow viscosity is thought to be an important modulator of shear stress.¹³² It is possible that increased bone marrow adiposity in SCI changes the biomechanical properties of bone marrow and thereby the ability of bone cells to respond to mechanical stimuli. Furthermore, as previously mentioned, local fat depots could be providing paracrine and/or juxtracrine signals to bone cells concomitant to the systemic endocrine fat sources. However, the biomechanical and biochemical mechanisms by which excess bone marrow fat relates to poor bone health remain incompletely understood at this time.

Skeletal autonomic and neuroendocrine considerations in SCI

Multisynaptic retrograde tract-tracing has revealed that the rodent femur is innervated by a variety of neurons.^133–137 The neural regulation of bone and bone marrow has been recently and comprehensively reviewed.¹³⁷ A number of points are pertinent to SCI. Long bones are more densely innervated by sympathetic fibers than axial bone.¹¹⁹ Experimental elevation of sympathetic tone in rodents results in bone loss in long bones but not axial bone.^138,139 This preclinical evidence suggests that it is possible that the central nervous system contributes to skeletal physiology in humans by regulation of bone mass¹⁴⁰ and secretion of endocrine molecules from bone.¹²² In humans, indirect evidence suggests a role for the sympathetic nervous system in regulating bone mass, likely via β-2 adrenergic receptors expressed on both osteoblasts and osteocytes. In the general population, β-blocker agents have been shown to increase BMD and reduce fracture risk.^141–150 Furthermore, high resting heart rate has been suggested as a predictor of osteoporotic fracture risk.¹⁵¹ Direct microneurography measures of sympathetic nervous system (SNS) activity in premenopausal women has shown that sympathetic activity at rest is inversely correlated with trabecular bone volume and compressive strength of the tibia.¹⁵² The regionally specific lower extremity long bone pathology in SCI mirrors that seen with pathologically high sympathetic activity. The possible sympathetic contribution to bone loss should be considered in the context of the multiple episodes of autonomic dysreflexia (AD)¹⁵³ occurring through a day^154–156 in persons with SCI. These transient episodes of uncoordinated sympathetic outflow, occurring 10¹⁵⁶ to 40^154,155 times per day usually in persons with injuries above T6, might directly stimulate bone via the SNS. AD is commonly ascribed to injuries above T6, but it has been reported in individuals with injuries as low as T10.¹⁵⁷ Furthermore, sublesional SNS output in response to sublesional stimuli is similar for both high- and low-level SCI¹⁵⁸; in low injury, it occurs without the cardiovascular signs and symptoms of clinical AD.¹⁵⁸ It is important to note that SNS efferent output, such as that seen in AD, has not been shown to increase SNS activity within bone or bone marrow. The relationship between sympathetic dysfunction and skeletal pathology in SCI is still an area of considerable uncertainty. It remains to be determined if modifying autonomic function could be of benefit to bone pathology.

Beyond the potential effects of autonomic stimulation, the hormonal impact of recurring episodes of AD is difficult to evaluate due to the logistical complications of continuously monitoring blood analytes throughout an entire day. Other endocrine impairments of SCI are well established and likely impact bone. Hypogonadism is prevalent in SCI, with 46% to 60% of the adult male SCI population having low levels of testosterone.^159,160 Testosterone maintains bone density in men through its aromatization to estradiol,¹⁶¹ and low testosterone is likely to be involved in bone pathology in SCI.¹³¹ Secondary hyperparathyroidism^162–164 might interact with the ~35%^163,164 to 93%¹⁶⁵ prevalence of vitamin D deficiency in SCI.

Conclusion

SCI results in profound changes in body composition, including increases in adiposity and decreases in lean mass including early and precipitous bone loss. Recent CSCM clinical practice guidelines for CMD and the ISCD official position on DXA should be used by clinicians for the early detection and continued monitoring of obesity and bone health in persons with SCI. The interaction of adipose and skeletal pathologies in SCI is an important consideration for treating these conditions. In light of the relationship between secondary health complications and quality of life in persons with SCI,^166–168 knowledge about the physiological interaction of fat and bone should be explored by scientists to guide clinical strategies for treating disorders of obesity and bone in SCI.