Page 184

5


Tissue Mechanobiology

Basic biology studies, such as those that examine microscopic or biochemical changes in tissues, are a source of our understanding of details of injury mechanisms at the tissue or cellular level. These studies are primarily performed using cadaver samples, animal models, and tissues or cells grown in culture. Direct study of the tissues of concern in live humans (e.g., nerve, tendon, disc, muscle) is limited due to ethical and methodological considerations. The purpose of this chapter is to systematically review basic biology studies in order to determine to what degree they support an association between loading and tissue injury, especially at load levels well below those that cause tissue disruption. Biological plausibility, one of the Bradford Hill criteria for causality (see Chapter 3), refers in this case to the likelihood that associations between loading and tissue damage are compatible with existing knowledge of basic biological mechanisms. These basic biology studies may also address other Bradford Hill criteria, such as specificity, temporality (tissue injury occurs after loading is initiated), and dose-response associations.

In reference to our overall model of the person in the workplace (Figure 1.2), this chapter explores the “Internal Tolerances” box, that is, the tolerance of tissues to loading. The focus is primarily on the mechanical and biological responses of tissues to repeated or continuous loading, not on damage due to a single, sudden load. However, cyclical loads are frequently compared to the single load that causes tissue to rupture or grossly fail. It is well recognized that loading is required to maintain tissue integrity. Lack of loads or disuse leads to tissue atrophy and impaired function (e.g., osteoporosis, muscle atrophy). The implications are that there may be an optimal range of loading below which atrophy occurs and above which tissue injury may occur. This chapter does not



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 184
Page 184 5 Tissue Mechanobiology Basic biology studies, such as those that examine microscopic or biochemical changes in tissues, are a source of our understanding of details of injury mechanisms at the tissue or cellular level. These studies are primarily performed using cadaver samples, animal models, and tissues or cells grown in culture. Direct study of the tissues of concern in live humans (e.g., nerve, tendon, disc, muscle) is limited due to ethical and methodological considerations. The purpose of this chapter is to systematically review basic biology studies in order to determine to what degree they support an association between loading and tissue injury, especially at load levels well below those that cause tissue disruption. Biological plausibility, one of the Bradford Hill criteria for causality (see Chapter 3), refers in this case to the likelihood that associations between loading and tissue damage are compatible with existing knowledge of basic biological mechanisms. These basic biology studies may also address other Bradford Hill criteria, such as specificity, temporality (tissue injury occurs after loading is initiated), and dose-response associations. In reference to our overall model of the person in the workplace (Figure 1.2), this chapter explores the “Internal Tolerances” box, that is, the tolerance of tissues to loading. The focus is primarily on the mechanical and biological responses of tissues to repeated or continuous loading, not on damage due to a single, sudden load. However, cyclical loads are frequently compared to the single load that causes tissue to rupture or grossly fail. It is well recognized that loading is required to maintain tissue integrity. Lack of loads or disuse leads to tissue atrophy and impaired function (e.g., osteoporosis, muscle atrophy). The implications are that there may be an optimal range of loading below which atrophy occurs and above which tissue injury may occur. This chapter does not

OCR for page 184
Page 185focus on the effects of disuse. We do, however, investigate a number of questions: Is there evidence that tissue damage occurs at levels below the strength of the tissue? Is there evidence for microtrauma or damage accumulation? If so, what is the dose-response relationship? What are the mechanisms of injury and repair? How are the responses of tissues to load modified by intrinsic factors (e.g., age, gender)? Do existing studies support or refute an association between repeated loading and injury? For the purposes of this chapter, several terms are defined. Elastic materials are those that regain their original shape after a load is removed. For such materials, the change in shape (e.g., strain) is proportional to the applied load (within certain limits). The constant of proportionality is called the stiffness. The force (e.g., stress) necessary to cause rupture or fracture is called the strength. In some cases, cyclically applied forces that are below the tissue strength may cause rupture or fracture via damage accumulation. This is called fatigue. Tissue fatigue can also lead to changes in other mechanical properties, such as reduced tissue stiffness. Since pain is a common and important endpoint for humans with musculoskeletal disorders, this chapter begins with a review of the pain pathways from musculoskeletal tissues to the brain. This is followed by reviews of the biological responses of six tissues—vertebral bone, spinal disc, tendon and ligament, muscle, peripheral nerve, and spinal nerve root—to loading. These reviews are based on systematic evaluations of the scientific literature. The chapter summary integrates the findings across all tissues and draws conclusions about current knowledge of injury mechanisms. We conclude with suggestions for future research directions. PAIN PATHWAYS FROM PERIPHERAL TISSUES Evolution has provided our bodies with many senses by which to interact with the environment. Each sense (smell, vision, hearing, taste, and somatic sensibilities) has a highly specialized neural pathway. Pain is one of the somatic sensibilities (others are touch, temperature sensation, and proprioception) and itself has its own highly specialized set of neural pathways. This specialization begins in the peripheral tissues. “Nociceptor” is the term given to the specialized receptors that serve as injury (or noxious stimuli) detectors. Activation of nociceptors evokes pain. Pain arouses us to protect the injured or threatened body part and hence plays a crucial role in survival. Nociceptors innervate a variety of tissues in ways that are appropriate from a teleological perspective. Lightly touching the cornea can injure the eye, and so the nociceptors that serve the cornea are quite sensitive to mechanical stimuli. The skin is a more resilient tissue, and

OCR for page 184
Page 186nociceptors that serve the skin are sensitive to higher intensities of stimuli. Not all tissues have nociceptors (e.g., fat tissues are relatively insensitive to noxious stimuli). However, muscle, periosteum, and especially the interface between ligaments or tendons and bone are richly innervated by nociceptors. Correspondingly, surgical manipulation of fat is relatively painless, whereas manipulation of muscle or bone at tendon insertion sites is painful. Nociceptors can also assist in healing and may even be involved in neuroimmune mechanisms. When nociceptor innervation to the skin is blocked, there is delay in wound healing, and the thickness of the epidermis is reduced. Neurogenic inflammation is another function of nociceptors; activation of nociceptors prompts a release of potent vasoactive peptides that leads to redness and increased permeability of the vessels. Signals from nociceptors are transmitted by the peripheral nerve to cells in the spinal cord. Damage to the peripheral nerve (e.g., carpal tunnel syndrome, spinal root compression) may lead to unusual sensations, the sensation of pain, or the loss of sensation (e.g., numbness) in the part of the extremity served by the nerve. The spinal cord is an important processing center for noxious information. Nociceptive inputs have connections to motor neurons in the dorsal horn; this accounts for pain-induced muscle contractions (muscle spasms). Specialized cells in the spinal cord also transmit information from nociceptors to higher brain centers. These inputs to higher centers arouse descending pathways back down the spinal column, which in turn regulate the sensitivity of the nociceptive neurons. Other inputs from peripheral pathways (e.g., touch systems) may interact with the nociceptive inputs to regulate the sensitivity of the cells in the spinal cord. Thus, the sensitivity of the pain-signaling pathways is highly plastic. Nociceptors that serve different deep tissues have convergent inputs to the spinal cord; these lead to the phenomenon of referred pain. Thus, a person with a heart attack may feel pain in the left arm; a person with a herniated cervical disc feels muscle tenderness in the trapezius muscle, and a person with carpal tunnel syndrome may feel pain in the elbow and upper arm. Injury may induce changes in pain sensibility. Tissues may become hyperalgesic; that is, the same stimulus produces a greater sensation of pain. Lightly touching the skin may be associated with pain (allodynia). Hyperalgesia results from two forms of sensitization: peripheral and central. Nociceptors (peripheral) themselves become more sensitive to heat and mechanical stimuli, and the spinal cord cells (central) become sensitized as well. As part of this central sensitization, the nerve fibers concerned with touch sensation acquire the capacity to activate the spinal cord cells that serve pain. This accounts for the phenomenon of allodynia,

OCR for page 184
Page 187in which touch stimuli evoke pain in patients with inflammation of the skin and in patients with nerve injury. LITERATURE REVIEW The panel reviewed the scientific literature to evaluate the state of knowledge of the effects of loading on vertebral bone, spinal disc, tendon and ligament, muscle, peripheral nerve, and spinal nerve root. Online databases (e.g., Embase, MEDLINE, Pre-Medline) were searched at least back to 1980 for articles with relevant keywords (e.g., tissue type, damage, pathology, fatigue, tension, compression, repetitive, loading). Appropriate articles were considered for review only if they were published in English-language, peer-reviewed scientific journals. For each tissue type, this process identified between 28 and 190 articles for consideration. The reviews that follow summarize, for each tissue, the function and structure of the tissue; the effects of loads on microstructure, mechanical characteristics, and biological function; and the influence of heterogeneity, aging, and other factors on the response of the tissue to load. The types of load considered, along with the biological and mechanical responses, are appropriate to the tissue. VERTEBRAL BONE AND SPINAL DISC Structural and Functional Properties The intervertebral disc is a complex structure consisting of four distinct tissues: the nucleus pulposus, the annulus fibrosus, the cartilaginous endplates, and the adjacent vertebral bodies (Figure 5.1). The nucleus ~ enlarge ~ FIGURE 5.1 Schematic representation of the intervertebral disc (from Bass, 1999:2). Reprinted with permission from the author.

OCR for page 184
Page 188 pulposus is a viscous, mucoprotein gel that is approximately centrally located within the disc. It consists of abundant sulfated glycosaminoglycans in a loose network of type II collagen, with a water content that is highest at birth (approximately 80 percent) and decreases with age. The annulus fibrosus is a ligamentous tissue that becomes differentiated from the periphery of the nucleus and forms the outer boundary of the disc. The transition between the nucleus and the annulus is progressively more indefinite with age. The annulus is made up of coarse type I collagen fibers arranged in layers, running obliquely between the adjacent vertebral bodies. The fibers run in the same direction within a given layer, but opposite to those in an adjacent layer. The cartilaginous endplates cover the end surfaces of the opposed vertebral bodies and serve as the upper and lower surfaces of the intervertebral disc; they are composed predominantly of hyaline cartilage. The vertebral bodies consist of a trabecular (porous) bone core (centrum) surrounded by a thin shell of cortical (dense) bone. The facet joints are part of the posterior vertebral arch and serve as additional points of articulation between adjacent vertebra. They guide vertebral motion by constraining rotation and supporting some axial load. In the adult, the cells residing within the endplate, nucleus, and inner annulus resemble chondrocytes (cartilage cells), while the cells populating the middle and outer annulus are fibroblastic (fibrous tissue cells). Because the disc is avascular, these cells receive nutrition via diffusion from adjacent vascularized tissues and convective fluid flow (Maroudas, 1988). Normal Disc Mechanics The disc derives its structural properties largely through its ability to attract and retain water. The proteoglycans (biochemicals that help resist compressive loading) of the nucleus osmotically pull in water, exerting a “swelling pressure” that enables it to support spinal compressive loads. The pressurized nucleus also creates tensile stress within the collagen fibers of the annulus and ligamentous structures surrounding the disc. In other words, although the disc principally supports compression, the fibers of the annulus experience significant tension. This annular and ligamentous prestress, in turn, functions synergistically with the facet joints to guide normal spinal motion (Adams et al., 1987). Under long duration loading, in which the spinal stress exceeds the nuclear swelling pressure, water is slowly forced from the disc, principally through the semipermeable cartilaginous endplates, resulting in a creep response (continual change in height from a constant applied force). As a result of this mechanism, a significant disc water loss can occur over the course of hours due to activities of daily living (Tyrell et al., 1985).

OCR for page 184
Page 189Diurnal loss of disc height can approach 2 mm, leading to increased spinal instability via decreased tissue prestress (Adams et al., 1987). Water loss and stability can be recovered during periods of bed rest (LeBlanc et al., 1994). Effects of Age and Degeneration After skeletal maturity, the intervertebral disc undergoes numerous alterations with age. These include a progressive loss of cellularity, disorganization of the extracellular matrix, and, as a result, morphological changes and alterations in biomechanical properties (Buckwalter, 1995). These age-related changes represent a form of degeneration, which may be accelerated by a number of factors and have been implicated in increasing the risk of discogenic back pain. Discogenic pain refers to pain originating from the intervertebral disc and is distinguished from back pain of other origins, such as facet joints, spinal ligaments, and muscles. The most consistent chemical modification observed with aging is loss of proteoglycans and concomitant loss of water (Pearce et al., 1987). Secondary changes in the annulus include fibrocartilage production with disorganization of the annular architecture and increases in type II collagen (Rufai et al., 1995). These alterations precede the morphological reorganization usually attributed to degeneration: loss of disc height, disc bulge, sometimes called protrusion, and disc herniation, sometimes referred to as prolapse (Pearce et al., 1987). Disc bulge can occur when loss of water causes the disc to flatten, bulge beyond its normal margins, and may place pressure on a nerve exiting from or traversing along the spinal column. Disc bulge can also occur when fibrocartilage proliferates within the substance of the annulus fibrosus (Yasuma, 1990). Disc herniation occurs when disc material escapes through a fissure in the annulus fibrosus, which, like a bulge, can place pressure on nerve roots or the spinal cord. Nociceptive nerve fibers are sparsely present in the outer annulus and vertebral body and extensively present in the facet joint capsule and posterior longitudinal ligament (Cavanaugh et al., 1997; Antonacci et al., 1998; Palmgren et al., 1999). With increasing degeneration, nerves can penetrate to deeper layers within the tissue (Coppes et al., 1997; Freemont et al., 1997), including the vertebral endplate (Brown et al., 1997). Innervation is thought to advance deeper into the disc in concert with vascular granulation tissue (Yoshizawa et al., 1980). These nerves can be stimulated both mechanically and chemically (Yamashita et al., 1993). There are several mechanisms that purportedly link disc degeneration and low back pain. First, degeneration leads to tissue dehydration (Pearce et al., 1987). Breakdown of the nuclear polymeric structure results

OCR for page 184
Page 190in reduction of its osmotic properties— the disc loses its ability to attract and retain water (Urban and McMullin, 1985). Tissue volume loss from dehydration, in turn, leads to a decrease in disc height and an increase in disc bulging (Adams et al., 1987). Both of these geometric changes can adversely affect patients by accelerating facet joint arthritis and by causing mechanical impingement on the adjacent spinal cord or nerve roots. In these cases, alterations in nerve function secondary to chronic compression are thought to be the primary mediators of back pain (Devor, 1995). Dehydration is also correlated with decreases in disc cellularity, disorganization of the annular layers, and alterations in the density and architecture of adjacent vertebra (Vernon-Roberts, 1988). These changes can begin early in life and have significant consequences for the disc's biomechanical behavior. Disc degeneration may cause pain indirectly via chemicals secreted by disc cells. These inflammatory factors can diffuse to and sensitize surrounding innervated tissues (McCarron et al., 1987; Kawakami et al., 1996, 1997; Kayama et al., 1998). As mentioned previously, degenerated discs may be considered a normal consequence of aging. Indeed, a large percentage of the adult population has degenerated discs, with a significant percentage of these being asymptomatic (Wiesel et al., 1984; Powell et al., 1986). For instance, in an MRI study of symptomless adults, greater than 50 percent had disc bulges, protrusions, or vertebral endplate abnormalities (M. C. Jensen et al., 1994). While these data suggest that the presence of a degenerated disc is not diagnostic of back pain, the severity of spinal degeneration (extent and number of levels affected) does correlate with increased risk for symptoms (Luoma et al., 2000). Influences of Loading on the Disc Via Mechanical and Biologic Pathways The disc behaves as a composite structure when loaded: forces exerted on it are distributed among the tissues from which it is constructed (the annulus, nucleus, cartilage endplate, and adjacent vertebra). This tissue stress distribution is dependent on the type of loading (e.g., compression, flexion, lateral bending, or torsion) and duration of loading (creep response). Tissue stress induced by spinal loading affects the disc through both mechanical and biological pathways. These pathways are usually coupled: that is, the mechanical response influences the biology, and the biological response influences the mechanics. This load-induced response can be either beneficial or detrimental.

OCR for page 184
Page 191 An important detrimental mechanical response is overload injury (material failure). This occurs when the tissue stress exceeds the tissue strength. The human tolerance to overload injury has been investigated largely in cadaveric models. These in vitro experiments demonstrate that failure will occur within the tissue that is stressed most severely. The tissue at risk, and therefore the mode of injury, is in part dependent on the type of loading (compression, flexion, lateral bending, or torsion). For instance, under pure compression the disc fails by vertebral body fracture, whereas excessive bending injures the ligaments of the neural arch (Table 5.1). Vertebral body compressive strength is strongly correlated with its cross-sectional size and bone density (Brinckmann, Biggeman, and Hilweg, 1989a, 1989b), making it feasible to predict noninvasively in humans. Disc tissues can also be injured through a process of fatigue, where subfailure loads are applied repetitively for sufficient cycles to ultimately cause tissue failure via damage accumulation (such as during exposures TABLE 5.1 Summary of Static Strength for Intact Spinal Segments Loading Mode Injury Mode Average Strength Notes Compression Vertebral endplate fracture 5.2 (± 1.8) kNa 6.1 (± 1.8) kN (male, 20-50 yrs)a 10.2 (± 1.7) kN* (male, 22-46 yrs)b Dependent on vertebral cross-sectional area and bone density Shear Neural arch, facet joint fracture 1.0 kNc Uncertain Flexion Posterior ligaments 73 (± 18) Nm measured with 0.5 - 1.0 kN compressive preload Extension Neural arch 26 (± 9) Nmd Anterior annulus may be damaged Torsion Neural arch /facets 25 - 88 Nme Compression plus flexion Posterior annulus, vertebral body 5.4 (± 2.4) kNf Disc can prolapse under hyperflexion a Brinckmann, Biggemann, and Hilweg, 1989a, 1989b b Hutton and Adams, 1982 c Miller et al., 1986; Adams et al., 1994 d Adams et al., 1988 e Farfan et al., 1970; Adams and Hutton, 1981 f Adams and Hutton, 1982

OCR for page 184
Page 192 to extremes of whole-body vibration). Cadaveric experiments demonstrate that cyclic loading in compression, bending, torsion, shear, or combinations thereof damage vertebra (including facet joints) prior to damaging the annulus fibrosus. These studies suggest that vertebral fatigue may result from physiological loading regimens that include between 1,000 and 10,000 compressive cycles (5,000 cycles may easily accumulate in vivo during 2 weeks of industrial exposure). Cyclic compressive stress as low as 50 percent of the vertebral failure strength may result in fracture after 1,000 cycles (Hansson, Keller, and Spengler, 1987). Brinckmann, Biggemann, and Hilweg (1988) developed a probability model to predict the fatigue strength given vertebral size, density, and load magnitude. Based on this model, a “fatigue limit” of 30 percent of ultimate compressive strength has been hypothesized for living vertebrae; in other words, cyclic loading of less than 30 percent of vertebral compressive strength would never cause fatigue failure (Table 5.2). Vertebral microdamage cannot be identified utilizing clinical radiographs, bone scans, or MRIs (Hansson et al., 1980; Mosekilde and Mosekilde, 1986). The clinical significance of these pathological changes remains uncertain. Some evidence suggests that the annulus may also be injured via fatigue and damage accumulation (Gordon et al., 1991; Buckwalter, 1995; Walsh et al., 2000) Tissue stress developed during spinal loading can influence disc biology. Within vertebra, stress can stimulate cells to produce more bone in areas of high stress or remove bone in areas of low stress. This process, TABLE 5.2 Cyclic Loading Reduces the Compressive Strength of Lumbar Motion Segments Relative Load Number of loading cycles % 10 100 500 1000 5000 60-70 10%a 55% 80% 95% 100% 50-60 0% 40% 65% 80% 90% 40-50 0% 25% 45% 60% 70% 30-40 0% 0% 10% 20% 25% 20-30 0% 0% 0% 0% 10% a Values indicate the probability of compressive failure if a motion segment is loaded for the specified number of cycles at the specified relative load. Relative load is the actual compressive load expressed as a percentage of the load required for compressive failure from single loading cycle. Data from Brinckmann, Biggemann, and Hilweg (1988).

OCR for page 184
Page 193 called remodeling, is the body's mechanism to optimize the density and shape of bones for a particular mechanical exposure (e.g., tennis players have denser bone in their dominant arms) (Cowin et al., 1985). These bone cells are also responsible for healing fractures, including microdamage resulting from fatigue (as described above). It is inferred from known bone healing times, amounting to several weeks or months (Martin et al., 1998), that minimal repair of bone microfractures would be expected in a time interval of approximately 2 weeks. This observation suggests that vertebral fatigue damage may accumulate in vivo, not be offset by healing, and lead to fractures. However, while the presence of endplate microfractures appears to increase with age (Roberts et al., 1997), whether these are responsible for patient symptoms is uncertain (Braithwaite et al., 1998). Within the nucleus and annulus, spinal loading can alter tissue water content (via creep, as discussed above) and tissue shape, leading to altered cell metabolism. In particular, changes in water content, in addition to concomitant modifications of tissue permeability, fixed charge density, oxygen tension, and cell shape, can have adverse biologic consequences (Ohshima et al., 1989; Ohshima and Urban, 1992; Ishihara et al., 1996; Handa et al., 1997; Ishihara and Urban, 1999). For instance, Urban and coworkers utilized an in vitro model to demonstrate that disc cell function is harmed by extremes of water content (either too high or too low) induced by fluctuations of disc compression (Ohshima et al., 1995). The detrimental effect was thought to be due to alterations in the disc cells' pericellular environment. In vivo loading in animals demonstrates that altered disc cell metabolism and death may be related to spinal loading via a quantifiable dose-response relationship (Hutton et al., 1998; Lotz et al., 1998; Lotz and Chin, 2000). These studies and others demonstrate that certain regimens of spinal loading can be harmful to the disc. Implied, though not demonstrated directly, is that other regimens, involving lower compressive loads, may be beneficial. Summary and Conclusions The intervertebral disc manifests a complex, time-dependent response to spinal loading. Loading, in turn, alters the joints' biomechanical behavior and the tissues' biological activity. Overload injury and fatigue may cause vertebral body failure, while coupling between tissue stress and cell activity may accelerate annular and nuclear degeneration through more subtle, biological pathways. Spinal discs degenerate with age. The independent contribution of physical force to degeneration is currently unknown due to inherent physiological variability among individuals, and because aging, by definition, signifies lengthened exposure to cumulative trauma. Furthermore,

OCR for page 184
Page 194due to a lack of specificity of disc degeneration for back pain, the patho-physiological mechanisms linking spinal load and pain in humans are still uncertain. However, significant data exist by which to quantify the failure and fatigue strength of vertebral bodies in humans (related to bone density and bone size). The biological response to spinal stress and its contribution to damage accumulation have been demonstrated in animal and laboratory models, yet the extent by which this pathway affects humans still needs to be established. TENDONS AND LIGAMENTS Properties of Tendon and Ligament and Injury Endpoints Tendon and ligament are composed of dense connective tissue. The collagen fibrils visible with the electron microscope are grouped into fibers and fascicles that are visible with a light microscope. The fibers and fascicles are enclosed in a thin film of loose connective tissue called endotendon or endoligament. The whole tendon is wrapped in a connective tissue called the epitenon, which in turn is surrounded by the paratenon, a loose, areolar connective tissue (Figure 5.2). In some areas of the body, for example at the wrist, the paratenon forms a double layer sheath lined with synovial cells. This tendon sheath or tenosynovium facilitates smooth gliding of the tendon. Tendons connect muscle to bone, while ligaments connect bone to bone. Tendons and ligaments primarily transmit tension forces but can also experience shear and compressive loads (Luo et al., 1998). Compression occurs when the tendon path is altered, for example around a bony structure or pulley system, or if there is impingement between the bony structures. Cellular remodeling and adaptation of these tissues occurs in response to different types of loading. When a tendon experiences compressive loading in addition to tension, the tendon in this region is gradually transformed from linear bands of collagen fascicles into irregular patterned fibrocartilage. The transformation is accompanied by changes in proteoglycans (Malaviya et al., 2000). During tendon gliding, the amount of friction against the surrounding sheath and tissue depends on the amount of tension in the tendon, the friction coefficient, and the arc of contact (Uchiyama et al., 1997). Friction force can generate heat and cause thermal effects indirectly and can stimulate cellular reaction directly (Birch, Wilson, and Goodship, 1997). Joint movement determines the amount of tendon excursion; therefore, the specific joint posture, as well as tendon tension, are important determinants of compressive and shear load.

OCR for page 184
Page 208explanation for this is that the cycles of passive stretch damage the cross-bridges that attach and detach in relaxed muscle at a rate of 32 s−1 at 37° C (Eisenberg, Hill, and Chen, 1980). Such damage to cross-bridges would not be visible using a conventional microscope. The mechanism of muscle damage during passive stretch has not yet been fully elucidated. Similarly, the effects of cyclical, passive loading and the role of duty cycle and total duration have not been systematically studied. Vibration Injury of Skeletal Muscle Working with hand-held vibrating tools has been linked to neurologic, vascular, and musculoskeletal disorders (Pelmear and Taylor, 1992; Armstrong et al., 1987; Stromberg et al., 1997; Färkkilä et al., 1979). In terms of skeletal motor unit function, vibration exposure impairment during intermittent and sustained maximal voluntary contractions has been shown in humans to include reduced electromyogram (EMG) firing rate, decreased motor unit firing rate, and decreased skeletal muscle force generation (Bongiovanni, Hagbarth, and Stjernberg, 1990). The effects on skeletal muscle per se are not so well documented, and the decrease in skeletal muscle maximal force generation is due at least in part to reduced firing of α motor neurons (Färkkilä, 1978; Färkkilä et al., 1980). Relocation of nuclei to the center of muscle fibers, although not structural damage per se, is used as a marker of injury in vibration-exposed muscles because this nuclear change is observed to be a common feature of neuromuscular disorders. In studies, using centralized nuclei as a marker for injury, only the skeletal muscles most directly exposed to the vibration were found to be affected by it (Necking et al., 1992, 1996b; Dubowitz, 1985). The results of studies of rats by Necking and coworkers (Necking et al., 1996a, 1996b) show that frequency displacement and duration of the vibration interacted as determinants of changes in fiber nuclei location in the contracting skeletal muscle. The most direct indication of skeletal muscle damage from exposure to vibration during active contraction is that plasma levels of intracellular muscle enzymes increase, suggesting disruption of the skeletal muscle fiber cell membrane (Miyashita et al., 1983; Okada, 1986). Muscular weakness is a common complaint among vibration-exposed workers (Färkkilä, 1978; Färkkilä et al., 1980; Pyykko et al., 1986), and reduced hand grip strength, corrected for aging, may persist for years after exposure (Färkkilä et al., 1986). However, the evidence of skeletal muscle damage per se has not been thoroughly studied for conditions leading to vibration-induced loss of force generation.

OCR for page 184
Page 209 Age-Related Skeletal Muscle Injury and Risk Aging is associated with skeletal muscle decrements in force, power, endurance, and recovery from injury. There are also age-related changes in motor unit innervation and in muscle morphology and metabolism (Kirkendall and Garrett, 1998; Bemben, 1998; Faulkner, Brooks, and Zerba, 1990; Shephard, 1999). In humans, the decrease in muscle strength begins around age 40 and is more dramatic in humans after age 65; most of this decline is associated with inactivity (Faulkner, Brooks, and Zerba, 1990; Kirkendall and Garrett, 1998; Brooks and Faulkner, 1990, 1994; Shephard, 1999). However, approximately 20 percent of the age-related skeletal muscle weakness cannot be explained by the decrease in muscle mass or cross-sectional area associated with inactivity (Brooks and Faulkner, 1988; Bruce, Newton, and Woledge, 1989; Phillips et al., 1992; Brooks and Faulkner, 1994; Degens, Hoofd, and Binkhorst, 1995; Brown and Hasser, 1996; Jubrias et al., 1997). As evidence for this, training does not completely protect against the changes due to aging (Faulkner and Brooks, 1995). Aged skeletal muscle is also more vulnerable to injury. In studies of rats subjected to eccentric contractions, researchers have demonstrated that aged skeletal muscle fibers are more easily injured by single and multiple eccentric contractions, muscle fibers regenerate less, and structural and functional recovery is not complete (Zerba, Komorowski, and Faulkner, 1990; Brooks and Faulkner, 1990, 1996; Carlson and Faulkner, 1989). However, extensive studies of the effects of eccentric contraction on aged human muscle have not been published. Summary The scientific studies reviewed support the conclusion that repetitive mechanical strain exceeding tolerance limits, imposed in a variety of ways, results in chronic skeletal muscle injury. This conclusion must be tempered by the limitations of the animal studies, which examined only a limited number of independent variables for short time periods that do not match the time frame for chronic work-related exposures. A major void in this area is concrete animal data that links repetitive use to injury after chronic exposure at levels of use that do not cause short-term injury. The conclusions related to repetitive mechanical strain and chronic skeletal muscle injury are dependent on extrapolation of data from short-term animal experiments. However, human studies support the same conclusions, even though the measures of dependent and independent variables are less definitive and the experimental conditions are less controlled. More importantly, the earliest molecular contractile changes in skeletal

OCR for page 184
Page 210muscle structural injury have not been identified, and measures to detect them are not available for animals or humans. Standardizing work will not necessarily guarantee safety or similar risk of injury for all workers. Constant, or standardized, external loads and strains, encountered in the performance of work, will have a different impact on each person, because of individual variations in skeletal muscle (mass, type, condition, structure) and for a given person over time, because of effects of aging and adaptation. For example, younger conditioned persons with the largest contractile mass, or skeletal muscle cross-sectional area, will have the greatest contraction force generation to oppose external load. Similarly, people with longer skeletal muscles have the capacity to withstand the larger length changes, and a person's relative proportion of slow- versus fast-twitch fibers will determine their tolerance for low-intensity endurance versus high-intensity burst-type work. This creates person-based variations in the risk and degree of injury for fixed work and external loads. Better noninvasive measures of skeletal muscle injury threshold are needed, since the match of work to task will be imprecise. PERIPHERAL NERVE Structure and Function Peripheral nerves carry electrical impulses from peripheral tissues (e.g., skin, tendon, muscle) to the spinal columns and from the spinal column to the periphery (e.g., vessels, muscle). A nerve is composed of hundreds or thousands of axons, which are each an extension of a nerve cell body located in the spinal cord. The axon is surrounded by Schwann cells to form myelinated nerve fibers ( Figure 5.4). Myelinated and non-myelinated nerve fibers are grouped together in bundles, called fascicles, and surrounded by a perineurial membrane. The amount of connective tissue in and surrounding the nerve varies by level. For example, nerves located superficially in the limb or parts of the peripheral nerve that cross a joint contain an increased quantity of connective tissue, possibly as a response to repeated loading (Sunderland, 1978). The energy needs of impulse propagation and nutritional transport (axonal transport) are provided by a unique microvascular system. The small vessels supplying the nerve from the surrounding tissue have a coiled appearance that permits the normal gliding of the nerve during movement. When the vessels reach the nerve, they divide into branches running longitudinally in various layers of the nerve. In the endoneurium the environment is protected by a blood-nerve barrier. There are no lymphatic vessels to drain the endoneurial space; therefore, when edema

OCR for page 184
Page 211 ~ enlarge ~ FIGURE 5.4 Drawing of a peripheral nerve with bundles of nerve fibers surrounded by perineurium (p) forming 12 fascicles. The fascicles are embedded in a loose connective tissue called the epineurium (epi). Myelinated (c) and nonmyelinated (b) fibers are shown with Schwann cells (Schw), myelin sheath(my), axons (ax), and nodes of Ranvier (nR). (Reproduced, with modification, from Lundborg, 1988:186. Reprinted with permission.) forms in this space, the pressure in the fascicle may increase and rapidly interfere with the endoneurial microcirculation (Lundborg and Dahlin, 1996). Short-Term Effects of Compression The effects of loading on the peripheral nerve have recently been reviewed (Rempel et al., 1999). The primary mechanism of mechanical injury to the nerve is by regional compression or nerve stretching. Extraneural compression pressures as low as 20 mmHg can decrease intraneural microvascular flow, and pressures of 30 mmHg can impair axonal transport. By increasing vascular permeability, a brief low-pressure (30 mmHg) compression of the nerve can lead to endoneurial edema formation, which persists for at least 24 hours after the compression is

OCR for page 184
Page 212removed. In turn, the resultant edema reduces blood flow in the nerve. Both increasing duration of compression and higher pressure lead to greater edema formation. The effect of fluctuating extraneural pressure on nerve function was investigated in a rat tibial nerve model, wherein a sinusoidal pressure pattern was applied at 1 Hz for 20,000 cycles (Szabo and Sharkey, 1993). The study indicated that when extraneural pressure fluctuates rapidly, the effect on nerve function is associated with the mean value of the pressure waveform, rather than the minimal or peak value. Long-Term Effects of Nerve Compression The long-term biological effects of brief, graded nerve compression have been studied in several animal models using small inflatable cuffs (Powell et al., 1986; Dyck et al., 1990). Pressures of 0, 10, 30, and 80 mmHg were applied for 2 hours to a nerve; then at intervals up to 28 days the nerves were examined for evidence of injury. Within 4 hours endoneurial edema formed within all compressed nerves and persisted for the entire time of the study. Inflammation and fibrin deposits occurred within hours of compression, followed by a proliferation of endoneurial fibroblasts and capillary endothelial cells. Within days, vigorous proliferation of fibrous tissue was noted, with marked fibrosis at day 28 and sheets of fibrous tissue extending to adjacent structures. Demyelination and axonal degeneration were first observed a week after compression. The degree of axonal degeneration and demyelination were correlated with the initial pressure. To model chronic nerve compression, other investigators have placed short silicon tubes of varying internal diameters or loose ligatures around the rat sciatic or sural nerve (Mackinnon et al., 1994; Sommer et al., 1993). These are very effective models for studying pain-related behavior (Mosconi and Kruger, 1996). The biological response of the nerve is similar to that found in the cuff experiments, with early perineural edema followed by a short-term inflammatory response, fibrosis, demyelination, and, finally, nerve fiber degeneration. It is not possible to precisely control the compression level with these chronic models. Vibration Exposure Work with handheld vibrating tools can lead to a complex of symptoms known as the hand-arm vibration syndrome, in which sensorineural disturbances are prominent (Strömberg et al., 1996). Biopsies of the posterior interosseus nerve 5 cm proximal to the wrist, from men exposed to hand vibration at work, revealed such pathological changes as break-

OCR for page 184
Page 213down of myelin and the presence of interstitial and perineurial fibrosis in comparison to controls (Strömberg, 1997). The histology results suggest that demyelination may be a primary lesion in the neuropathy, which is followed by fibrosis associated with incomplete regeneration or with organization of an edema. Similar pathological changes are seen in the small nerves at the fingertips from patients exposed to vibrating handheld tools (Takeuchi et al., 1986). Animal models exposing peripheral nerves to vibration demonstrate an initial edema formation followed by demyelination and later a loss of axons (Lundborg et al., 1990, 1987; Ho and Yu, 1989; Chang, Ho, and Yu, 1994). Summary Several animal models demonstrate that low magnitude, short- or long-term compression of a peripheral nerve leads to a biological response of endoneurial edema, demyelination, inflammation, axon degeneration, and fibrosis. The degree of axonal degeneration is dependent on the applied pressure in a dose-response pattern. The critical pressure or threshold causing acute changes in nerve function is known, but the critical pressure-duration threshold for chronic nerve compression is unknown. Exposure to vibrating hand tools at work can lead to permanent peripheral nerve injury. Animal models of vibration exposure confirm a pathophysiological process of edema formation followed by demyelination and axonal degradation. No animal model has been developed to evaluate the effects of repetitive hand-finger loading on nerve structure and function. SPINAL NERVE ROOTS Structure and Function of Spinal Nerve Roots The nerve roots are located along the axis of the spine and serve as routes of communication between the central and peripheral nervous systems. Enclosed by the vertebral bones, the spinal nerve roots are relatively well protected from external trauma. However, spinal canal pathology that compromises the neural space, such as disc herniation or protrusion, spinal stenosis, and degenerative disorders, can create high risk of injury, even under what might be considered moderate physical exposures. Furthermore, nerve roots do not possess so much protective connective tissue as do the peripheral nerves, which makes them particularly sensitive to mechanical and chemical irritation. Structurally, the axons of the nerve root are located in the endoneural space, which is similar to that of the peripheral nerve but with five times less collagen. The root sheath separates the nerve root from the cerbro-

OCR for page 184
Page 214spinal fluid, which is surrounded by the spinal dura mater. The vascular supply is complex and may be involved in the pathophysiology of injury. The vessels from the periphery and from the spinal cord meet in the proximal one-third of the nerve root; it has been suggested that this region is particularly vulnerable to injury from ischemia. The blood-nerve barrier in the nerve root is not so well developed as in peripheral nerves; this creates a higher risk of edema. Acute Nerve Root Compression The most common mode of nerve root injury is mechanical compression. Recent experiments that precisely control nerve root compression reveal that capillary blood flow can be disrupted by venular occlusion with pressures as low as 5 to 10 mmHg. Such a compression-induced impairment of the vasculature will impede nerve root nutrition and lead to nerve root dysfunction. There is no significant secondary route of nutrition via diffusion from the cerebrospinal fluid. Low-pressure compression of a nerve root will lead to an increase in the vascular permeability and intraneural edema formation (Olmarker, Rydevik, and Holm, 1989), a response well documented for peripheral nerves (Rydevik and Lundborg, 1977). In peripheral nerves, such edema may increase the endoneurial fluid pressure (Low and Dyck, 1977; Lundborg, Myers, and Powell, 1983; Rydevik, Myers, and Powell, 1989), which in turn may impair the endoneurial capillary blood flow and jeopardize the nerve root nutrition (Myers et al., 1982; Low, Dyck, and Schmelzer, 1982; Low et al., 1985). Edema may negatively affect the nerve root for a longer period than the compression itself, since the edema usually persists for some time after the removal of a compressive agent. The presence of an intraneural edema is also related to subsequent formation of intraneural fibrosis (Rydevik, Lundborg, and Nordborg, 1976), which may delay recovery in some patients with nerve compression disorders. Experimental compression studies have demonstrated that the sensory fibers are more susceptible to compression than the motor fibers (Pedowitz et al., 1992; Rydevik et al., 1991). Chronic Experimental Nerve Root Compression Compression that evolves gradually may allow time for the remodeling and adaptation of axons and vasculature. In this case, the clinical consequences of compression may be less severe than if the compression was applied acutely. Despite this, a very gradual increase in compression, over two weeks, still results in structural and functional changes consistent with constriction (Delamarter et al., 1990; Cornefjord et al., 1997).

OCR for page 184
Page 215Following compression of nerve root and dorsal root ganglion, there can be an increase in a chemical factor called substance P, which is a neurotransmitter related to pain transmission (Cornefjord et al., 1995). This finding suggests that compression may lead to pain via both mechanical and chemical pathways. Mechanical Deformation, Biochemical Factors, and Pain Mechanical deformation of nerve roots may induce impulses that are perceived by the individual as pain. For instance, mechanical stimulation of nerve roots or peripheral nerves results in nerve impulses of short duration; these impulses are prolonged if the nerve tissue had been exposed to mechanical irritation by a chronic gut ligature (Howe, Loeser, and Calvin, 1977; Cavanaugh, Özaktay, and Vaidyanathan, 1994). Severe mechanical deformation, such as ligation of the nerve root, is generally not painful (Chatani et al., 1995; Kawakami et al., 1994a, 1994b). An increase in the level of neurotransmitters related to pain transmission has been found in the dorsal root ganglion in response to whole-body vibration of rabbits (Weinstein, 1986; Weinstein et al., 1987). A similar increase has also been seen in the dorsal root ganglion and nerve root after local constriction of the same nerve root (Cornefjord et al., 1995). Recent data suggest that chemical factors in nucleus pulposus can sensitize a nerve root, making it more susceptible to mechanical perturbations. Applied individually, nucleus pulposus or slight mechanical movement will not produce pain behavior in a rat model, whereas the combination of the two factors produces pain (Olmarker and Myers, 1998; Olmarker et al., 1998). This sensitization is thought to be orchestrated by the cytokine TNF-alpha (Olmarker and Larsson, 1998). Summary Although experimental models of low-grade compression have not been so extensively applied to the spinal root as they have to the peripheral nerve, they reveal similar initial damage mechanisms. Compression or direct mechanical stimulation may release cytokines and neurotransmitters (e.g., substance P, TNF-alpha) that stimulate pain transmission. However, due to experimental difficulties associated with chronic pain models, the entire pathway leading to chronic pain remains uncertain. The pathway is likely to involve a combination of compression of the dorsal root combined with the release of biochemical mediators of pain.

OCR for page 184
Page 216 SUMMARY The structural tissues—bone, disc, tendon, and muscle—demonstrate fatigue failure well below their strength. For example, tendons subjected to repeated stretching at 10 percent of strength will ultimately fail. For tissues studied in detail, the relationship between load and number of cycles to failure follows a log-linear dose-response model. The demonstration of this relationship provides evidence for damage accumulation. The ultrastructural correlates to damage accumulation have been observed for some tissues. For example, vertebral bone subjected to cyclical loading develops microfractures. In other tissues, such as the disc, the ultrastructural correlates have yet to be documented. Although a threshold for fatigue damage below which no damage accumulates has been postulated, such thresholds have yet to be proven. For example, 30 percent compression strength has been postulated as the fatigue threshold for the disc. The duration of fatigue testing at low stress levels (e.g., 10 percent) has been short relative to the time frame for repair and remodeling. Tissue repair and remodeling takes place over weeks and months, but the fatigue tests are conducted for hours or days. The documentation of thresholds for fatigue damage would be a valuable adjunct to the current data. In vivo animal models are necessary to investigate the whole organism's response to tissue loading. To some degree, the ultrastructural damage due to cyclical loading may be repaired as long as the time frame for the repair and remodeling is not long relative to the rate of damage and as long as the remodeling mechanism is not overwhelmed. With a repair system in place, one would expect a load-duration or a load-repetition threshold below which there is no damage accumulation and a disorder would never manifest. In addition, some injury mechanisms, for example, those mediated by inflammation or ischemia, can only be thoroughly investigated with in vivo models. In vivo animal models have been developed for the disc, tendon, muscle, and nerve that can support the investigation of the effects of cyclical loading on cellular, biochemical, and mechanical endpoints of tissues in the intact organism. Generally, these studies demonstrate specific damage endpoints, with sustained or repeated loading, that are similar to the pathology observed in humans. For example, the rabbit tendinitis model developed by Backman et al. (1990) demonstrated edema, increased capillary network, and inflammatory cells in the paratenon and degenerative changes in the tendon, findings also observed in histopathology studies of human tenosynovitis and epicondylitis. Most of the in vivo findings have been observed in more than one laboratory. These studies, in addition to demonstrating the specificity of endpoints, also

OCR for page 184
Page 217document the temporal relationship between exposure and effect. Due to the different levels of development of these models and the methods used to evaluate tissue damage, the depth of knowledge of the mechanism of injury varies from tissue to tissue. In vertebral bone, strength diminishes and damage is irreversible with each trabecular fracture. However, the role of the smaller microcracks in the pathophysiology of bone damage is uncertain; they may be a harbinger of fractures, or they may be repaired. In humans, vertebral microfractures are not visible using current imaging methods (e.g., X-ray, CT, MRI), which contributes to the known poor correlation between images of the spine and low back pain. Both disc and vertebra demonstrate fatigue failure, but vertebral endplates fail before the disc in response to repeated loading. Cumulative spinal loading affects disc pressure and water content, and these factors, in turn, influence cell viability and function. In vivo loading of sufficient magnitude can cause altered cell metabolism and cell death following a dose-response relationship. In vivo animal models that expose tendons to repeated loading demonstrate an inflammatory response with fibrosis in the peritenon. The process may be mediated by an initial release of inflammatory mediators and microtrauma. In the central tendon, degenerative changes are observed with edema, collagen disorganization, and fibrosis. Although the initial steps may include microtrauma, they are not well characterized. In muscle, prolonged fatigue damage can occur with single or cyclical loads. Eccentric loading is more damaging than concentric loading. Repeated eccentric loading can produce a persistent force decrement with structural damage to the sarcomeres (Z-line streaming, fiber necrosis, and inflammation). A threshold or endurance limit for injury has not been identified. The mechanical fatigue injury mechanism may be complemented by physiological mechanisms (e.g., intramuscular pressure, Cinderella hypothesis). In the peripheral nerve, compression causes edema accumulation, elevated endoneural pressure, vascular disruption, fibrosis, demyelination, and axon injury. The steps linking the initial effects to demyelination and permanent nerve damage are uncertain. Compression of the nerve for a sufficient duration also leads to chronic pain. Exposure to vibration causes a similar process of edema formation followed by demyelination and axon degradation. Although the relationship between nerve injury and compression follows a dose-response model, the critical pressure and duration relationships for chronic nerve compression have not been determined. For the spinal nerve roots, adjacent tissue compression may release cytokines that stimulate pain transmission. Age can influence the mechanical and biological properties of bone, disc, muscle, and nerve. Increasing age leads to increased degenerative

OCR for page 184
Page 218changes in vertebrae and discs, increased accumulation of extracellular matrix in the peripheral nerve, and reduced muscle strength. However, the independent role of age in modifying the negative effects of cyclical loading on these tissues is not determined. The role of gender as a covariate in the response of tissues to cyclical loading has not been investigated.