A growing body of research shows that lactate is not a “dead-end metabolite,” and it is not the “mediator” of fatigue and inefficiency as widely published in the 1960s through the 1980s.² On the contrary, current research implicates the hydrogen ion excess as the primary agent of diminished contractile power. The lactate ion may serve multiple roles in maintaining constant energy; recruiting new energy sources (gluconeogenesis); recruiting new vascularity (angiogenesis); and promoting a local cascade of healing, plasticity, and hyperplasia.²^–⁴ A growing body of research based on the experimental work of Brooks,⁴ termed lactate shuttle theory, suggests that higher concentrations of lactic acid produced in the skeletal muscles stressed by exercise have significant increased benefit in remote tissue such as brain, liver, heart, peripheral nerves, and peripheral vasculature over baseline metabolism.²

The ATPase work of Engel ⁵ in 1962 established a body of research showing the presence of distinctly different motor units within skeletal muscle. There are many myotype classification schemes based on histology, morphology, or function. In brief, the interaction between the type of myosin heavy chain (ATP binding site) and actin within individual sarcomeres is probably the greatest contributor to functional differences within myofibrils. The functional difference is related to the rate that the myosin heavy chains can repetitively bind ATP and release ADP under conditions of physiologic stress.⁶

Roughly divided, sarcomeres fall into one of three broad functional categories. Type I fibers have a slower “twitch” response with good fatigue resistance and lower tension development.⁷ Structurally, these groups of sarcomeres (known collectively as fibers) have rich capillary beds and high concentrations of mitochondrial enzymes with relatively low concentrations of glycogen and myosin ATPase. They seem ideally suited for aerobic activity with good fatigue resistance. These type I muscles predominate in areas that require aerobic or endurance demands. Type II muscle displays a fast twitch with good strength but relatively poor endurance compared with type I muscle fibers. Type II fibers can be subdivided further into type IIA, which still show a fast twitch response but a fatigue threshold between type I and type IIB, and type IIB, which show the fastest contractions, the highest tension development, and the most rapid onset of fatigue.^7,⁸ Although other fiber subtypes continue to be identified, type I, type IIA, and type IIB show the major functional categories of voluntary skeletal muscles.⁷ Whether individual fibers have biochemical characteristics for high-intensity, short-duration contractile bursts or more sustained activity, each muscle is a heterogeneous, woven tapestry of all of the above-mentioned fiber subtypes. Any relative predominance of one particular fiber type is based on genetics, anatomic location of the muscle, demand on the muscle, age of the individual, nutrition, and multiple other external factors.⁹

Although the contribution to parental lineage to phenotypic expression may remain paramount in predicting functional potential, the actual fiber composition of muscle groups has great capacity for plasticity in response to environmental stress. Multiple studies show fiber conversion within major groups from type IIB to type IIA to be common.¹⁰ Other conversions, such as type I to either type IIA or type IIB, are less common and seen most often with denervation, immobilization, and profound deconditioning.⁸ There is emerging evidence of conversion of type II to type I in the electrical stimulation literature, but data remain sparse and confounded by the fact that type II muscle tends to atrophy with age. The relative absence of type II muscle may confound data trying to analyze conversion.¹¹

A human infant is normally born with a full complement of muscle fibers that may continue to differentiate in childhood. Muscle fiber genesis seems to be biphasic. Myoblasts fuse to form fibrils at 8 to 10 weeks of gestation—primarily resulting in the formation of type I (slow-twitch) fibers. The second wave of fiber formation occurs at 15 to 18 weeks of gestation creating fast-twitch, type II fibers, which differentiate and change with early postpartum use.⁷ The growth and change in muscle size after adolescence seems to be due to increase in size of the fibers rather than increase in numbers (although controversial studies still contend that extreme muscle exertion may result in some new fiber development) that accompanies increased neuronal innervation and finer control of contractile force.

Although animal data suggest that hyperplasia and genesis of new muscle fibers may be possible with training and endogenous, hormonal feedback, these data have not been replicated in humans. Instead, increasing strength of contraction with training seems to be related not only to muscular factors, such as muscle size, fiber type, and fiber number, but also (and perhaps to a greater extent) to neural factors.¹² The specific neural characteristics that change with training include frequency, extent, order, and synchrony of motor unit firing, with feedback to the spinal pyramidal tracts to control system efficiency—maximal force at minimal energy expenditure. Training, age, and certain anabolic hormones (either endogenous or exogenous) seem to exert the greatest effect on muscle plasticity—the characteristic rapid increase in strength and muscle diameter under the hormonal influence of puberty is a specific example.

Muscle hypertrophy seems to occur through two processes that may proceed simultaneously: myofibrillar hypertrophy and splitting. The degree and sequence to maximize muscle size and function remain under study, but it seems clear that isometric contraction (in which the contracting muscle is not permitted to shorten) is far more effective in increasing muscle bulk than concentric (isotonic or isokinetic) contractions. Nevertheless, this increased bulk may come at a price of increased injury risk and lower dynamic function with very poor correlation to strength gains (isotonic and isokinetic training correlate better than isometric with dynamic strength). By contrast, various pathologic factors, such as denervation, starvation, and immobilization and disuse, produce muscle atrophy. Multiple factors, including a functional nerve supply, good nutrition, hormonal input, and periodic muscle activity, are necessary to maintain, or increase, muscle function and fibrillar “bulk.”

Anatomy

The spine consists of a series of bilaterally symmetrical joints phylogenetically adapted for protection of the neural communications network linking brain to periphery.¹³ The critical role of the spine musculature in dynamically protecting and vitalizing these articulations with their passive ligamentous supports and accompanying neural transmission lines is infrequently acknowledged in biomechanical modeling. For this reason, a perspective based in evolutionary theory aids in understanding the complexity of lumbar spine musculature.¹⁴ From an evolutionary standpoint, it seems that all land-dwelling vertebrates (including mammals) evolved from ocean-dwelling cousins; in the ocean, gravitational force acted differently when contributing to function and form. Large paravertebral spine muscles likely developed to provide lateral (coronal plane) flexion-extension to propel bodies through water, as shown by the lateral tail motion of the fish. This form of locomotion was initially preserved in amphibians and land reptiles (even after the development of extremities), as seen in the Crocodylia species including the modern alligator and crocodile. Reptiles with legs (currently living) all propel themselves through lateral spinal motion in which propulsion is achieved by contracting spine muscles to create alternating coronal convexities on one side of the spine that allow the ipsilateral foreleg to move forward while the contralateral hindfoot (on the concave side) is brought closer to the contralateral foreleg in preparation for the next reciprocal lateral movement that repeats the motion-event.

Mammalian adaptations resulted in a 90-degree evolutionary shift to the sagittal spinal motion characteristic of all four-legged mammals (the platypus and echidna being partial exceptions), which presumably provides an advantage during land locomotion, allowing for explosive growth of the class Mammalia. Land mammals that subsequently retreated to the sea (e.g., whales, porpoises, seals) maintain the sagittal orientation of spinal locomotion resulting in the 90-degree rotation of the tail fluke (when compared with fish), even after adaptive pressure resulted in other changes to their extremities. The later adaptation of humans to full bipedal stance and locomotion presumably necessitated a lordotic lumbar spine, kyphotic thoracic spine, and lordotic cervical spine for balance and ambulation (Fig. 5–2).

FIGURE 5–2 A, Muscle activation pattern needed to maintain lumbar spine model under compressive follower loads. The resultant force acting on the spine approximates the tangents to the deformed shape of the spine. B, Response of spine model to a compressive vertical load applied at L1 and to compressive follower load. Lumbar spine model could support substantially larger compressive loads when load path approximated the tangent to curve of lumbar spine.

(Borrowed with permission from Patwardhan AG, Meade KP, Lee B: A frontal plane model of the lumbar spine subjected to a follower load: implications for the role of muscles. J Biomech Eng 123:212-217, 2001.)

It is further theorized that the lordotic curve converts lateral bending to produce torsion at the hips, adding propulsion efficiency to the balanced human gait that would be impossible without lumbar lordosis. Ambulation without lumbar lordosis leads to the shuffling strides of the upright apes, whose gait is clearly dissimilar to that of humans. Laboratory modeling strongly suggests that cocontraction of spinal and abdominal muscles is the primary generator of the curvilinear structure of the spine that sustains logarithmically more force than “straight” models of spine motion (>1200 N vs. 100 N). The theory holds that instantaneous, axial-rotational forces between segments in “straight-spine” models lead to rapid failure when the spine is progressively loaded.¹⁵

This model focuses on the dynamic contribution of spine muscles to create a compressive-stabilizing force through bony and ligamentous structures that withstands physiologic forces modeled in sagittal motion. Similar to taut “guy wires” or tensioned cables adding structural stability to allow gauzy tent material to withstand 90 mph winds, muscle tension is hypothesized by the “follower-load” theory to provide a stabilizing force (in at least one plane of motion—sagittal) that directs the force vector to pure compression of the motion segment, minimizing shear (shear force being implicated in degenerative cascade).¹⁶

Another unique human evolutionary adaptation for manipulation of objects near the ground from a bipedal stance involves the ability to bend efficiently from the waist in combination with squatting and hunkering. A stable “biomechanical chain,” transferring force efficiently from hands through arms, shoulder girdle, spine, pelvis, legs, and feet to make ground contact, is necessary (Fig. 5–3). In this functional concept, the muscle acts as a dynamic stabilizer of the biomechanical chain in several ways. During flexion-extension of the lumbar spine, the co-tensioning abdominal muscles at a distance (force plus lever arm) allows maintenance of a “balance point” at each individual motion segment. Next, coupled anterior and posterior forces have a net compressive (downward) force, which balances motion at the instantaneous axis of rotation for each motion segment—maintaining compression at the disc and minimizing angular change. The net muscle tension force downward serves to offset any other forces, maintaining the force vector perpendicular to the disc’s plane similar to the tent “guy-wires” explained previously.^16,¹⁷ Muscles supply dynamic and static downward force to create a form that allows maximal load bearing (for bipedal carrying and lifting) capacity, while maintaining function that allows maximal efficiency of muscle energy output—maintaining a plumb line (not working against gravity to maintain posture), sharing tension bands to distribute loads, coupling forces when motion is required, and distributing load/work among the other osteoligamentous static structures.

FIGURE 5–3 Biomechanical chain for permitting manual handling of objects while maintaining balance over bipedal base, transmitting forces through four functional units: (1) upper extremities, (2) shoulder girdle and thoracic spine, (3) lumbar spine and pelvic unit, and (4) lower extremities.

(From Mayer TG, Gatchel RJ: Functional Restoration for Spinal Disorders: The Sports Medicine Approach. Philadelphia, Lea & Febiger, 1988.)

As stated previously, controlling forward flexion with muscles alone is inefficient, and the space required for the abdominal and thoracic contents imposes size constraints on spine muscles. The evolutionary solution is twofold: (1) strong, elastic posterior spinal ligaments (midline ligaments, joint capsules, and lumbodorsal fascia) that (a) produce passive constraint, particularly to lumbar spine flexion, and (b) allow static “hanging on the ligaments” subject only to slow, plastic “creep” but without muscular effort; (2) a manipulation of the lever arm advantage from quadrupeds through use of posterior pelvic muscles as “motors” and “stabilizers” of lumbar extension and abduction motion. This combination of a posterior ligamentous complex and powerful muscles of the buttocks and posterior thighs (along with the psoas muscle controlling degree of lordosis) permits the spine to function in a way not generally recognized: as a crane, whose boom is the ligament-stabilized flexed spine, whose fulcrum is the hips, and whose engine is the pelvic extensor musculature.

These observations lead to multiple functional inferences, one of which is the tendency of the spine to “hang on its ligaments” in an efficient, muscle-sparing manner frequently observed in “stooped laborers.” Efficiency of motion is documented by the way most normal subjects make the universal unconscious choice of allowing spine flexion to precede hip flexion when bending forward at the waist, eventually resulting in an EMG “silent period” when the spine is fully flexed and stabilized by posterior ligaments.^14,^18,¹⁹

It is beyond the scope of this chapter to discuss these concepts in greater depth; however, one quickly develops an appreciation of the spinal musculature as an efficiently evolved functional unit, improved on from earlier iterations, that encompasses paravertebral, abdominal, buttock, pelvic floor, and hamstring muscles to exert specific force vectors that combine with gravity and the constraint of the passive structures to allow bipedal stance, locomotion, and balance. This spine and pelvis functional module links the carrying and manipulation capacity of the cervicothoracic, shoulder girdle, and upper extremity functional modules with the propulsion capability of the lower extremity functional module. This relationship provides a complete biomechanical chain that allows manual handling tasks by the upper extremities to be performed while maintaining stable foot-to-ground contact for maximal evolutionary advantage in virtually any environmental context.

The small interconnecting vertebrae motion segments with the multiplanar motions of the three-joint complex produce difficulties when assigning specific uniaxial functions to individual groups of muscles. The erector spinae group of muscles are generally thought of as extensors of the spine. Functioning unilaterally, they may also be powerful abductors or lateral stabilizers (assisting with locomotion) and have been shown to function to some extent in spine derotation.¹⁸ Similarly, lateral abdominal musculature (internal and external oblique and transversus abdominis) may act as spine flexors and extensors (working through the lumbosacral fascia).¹⁴ The “girdling” abdominal muscles are also powerful spine rotators and assist in abduction and lateral stabilization.²⁰ As noted previously in this chapter, the force-coupling of spinal, pelvic, and abdominal muscles acting in concert increases compressive load through the disc and vertebral bodies to minimize shear force and imbalance axially across the three-joint complex.

In describing the gross anatomy of the spinal muscular functional unit, the extreme importance of the posterior stabilizing structures cannot be overemphasized in spinal functional integrity. The intrinsic lumbar musculature is only a part of the functional unit. The lumbodorsal fascia, interspinous ligaments, and facet joint capsules are crucial structures providing constraint and fulcrum points.²¹ There is growing appreciation among some surgeons to maintain the integrity of these collagenous and elastic structures, which may decrease adjacent segment shear, dysfunction, or instability.

Mammals as a class share many characteristics, including those of the spine musculature, that may account for the evolutionary success of the class. The most superficial layers of spine musculature, extensions of the functional unit of the shoulder girdle, include muscles such as the serratus posterior and latissimus dorsi. In contrast to other spinal musculature, the modified proximal function of these muscles is matched by innervation from the proximal spinal cord. The true lumbar spinal muscles, by contrast, have segmental innervations that arise from the posterior rami of the contiguous spinal nerves (the same nerves that perceive proprioceptive input from facet capsules, posterior ligaments, and peripheral anulus). Although true spinal muscles function together, their most characteristic differentiating factor is span length. The deepest muscles, such as the interspinalis, span only a single segment. The most superficial muscles may traverse a large portion of the entire spinal column. Controlled, coordinated action of individual vertebrae is a critical part of spine function, whereas loss of appropriate musculoligamentous control may contribute to various pathologic syndromes, such as segmental instability, segmental rigidity, facet syndromes, and perhaps even discogenic pain.²²

Musculature of the Lumbar Spinal Functional Unit

Intrinsic Muscles

Erector Spinae

The erector spinae is a large and superficial muscle that lies just deep to the lumbodorsal fascia and arises from an aponeurosis on the sacrum, iliac crest, and thoracolumbar spinous processes.¹³ The muscle mass is poorly differentiated, but divides into three sections in the upper lumbar area: (1) The iliocostalis is most lateral and inserts into the angles of the rib; (2) the longissimus, the intermediate column, inserts into the tips of the spinous processes of thoracic and cervical vertebrae; (3) the spinalis is most medial and inserts into spinous processes of the cervical and thoracic vertebrae (Fig. 5–4). The innervation of all the lumbar paraspinal muscles is from the dorsal rami of the nerve root as it exits at the most approximate level and overlaps proximally and distally along the muscle length.

FIGURE 5–4 Cross section of body musculature and fascia through L3 showing intrinsic spinal musculature. Abdominal muscles also function in containing viscera and respiration.

(From Finneson B: Low Back Pain, 2nd ed. Philadelphia, JB Lippincott, 1977.)

The importance of the lumbar musculature is inferred by the intricate redundancy of the paraspinal innervation. Each area of these long muscles has overlapping innervation that may include up to two segments craniad and two segments caudal of coinnervation. This redundancy allows maintenance of function even if a particular level is affected by injury to its respective neural structure.

Multifidi

Multifidi are a series of small muscles, best developed in the lumbar spine, that originate on the mammillary processes of the superior facets and run upward and medially for two to four segments, inserting on the spinous processes (Fig. 5–5). This orientation produces greater capacity for rotation and abduction, in addition to extension. They also share a multilevel innervation (similar to the erector spinae bundle or group) where function is maintained even with injury to the proximate dorsal rami.

FIGURE 5–5 Multifidi consist of numerous small muscle slips that arise from small bony prominences on the articular facet.

(From Finneson B: Low Back Pain, 2nd ed. Philadelphia, JB Lippincott, 1977.)

Quadratus Lumborum

The quadratus lumborum is the most lateral of the lumbar muscles (see Fig. 5–4); it originates on the iliac crest and iliolumbar ligament and runs obliquely to insert into the lowest rib and transverse processes of the upper four lumbar vertebrae. Its orientation provides strong abduction motor and stabilizer properties. The innervation is from the ventral rami of T12-L1-L2-L3 roots.

Deep Muscles

The interspinalis muscles are pairs of deep muscles spanning one segment on either side of the strong and elastic interspinous ligaments. In the lumbar spine, the intertransversarii consist of a pair of muscles on each side, spanning the transverse processes of adjacent vertebrae. Each side has dorsal and ventral slips.

Psoas and Iliacus Muscles

The psoas major, although usually thought of primarily as a hip flexor, has a direct effect on the vertebral column because it originates bilaterally from the vertebral bodies and posterior aspects of the transverse processes, providing the only intrinsic spinal muscle acting anterior to the sagittal axis. Yet, paradoxically, the psoas is usually an intersegmental extensor in the mid-lumbar spine, even as it flexes at the lumbosacral junction in the process of increasing the lumbar lordosis. It is an important spine stabilizer in sitting and standing.²³ Acting asymmetrically, the psoas may produce abduction or abduction-resistance to maintain coronal balance. Pathophysiologically, low back or pelvic pain may result from contracture or spasm of the iliopsoas producing combined hip and lumbosacral junction flexion. As in other situations, appropriately targeted stretches, followed by strengthening, may produce dramatic improvement.

The psoas major and iliacus muscles are innervated by the femoral nerve and lie in close proximity to the lumbosacral plexus. The primary innervation of this group is from the L2 and L3 root segments with minor contribution of L4 in some individuals. Proximal weakness and pain is occasionally a consequence of surgical approaches with susceptibility to denervation or devascularization injury from aggressive retraction. Care should be taken by surgeons because injury to this muscle can be a harbinger of occult lumbosacral plexus injury. The iliopsoas complex as the floor of Scarpa fascia serves as an important surgical landmark.

Extrinsic Muscles

Abdominal Musculature

There are four important abdominal girdling muscles in spine function.²⁴ The rectus abdominis is primarily a flexor, spanning the anterior abdomen from its origin on the pubic crest to its insertion on the anterior rib cage between the fifth and seventh ribs. The obliquely oriented abdominal muscles are, from superficial to deep, the external oblique, internal oblique, and transversalis abdominis. They all may act to produce rotation or abduction and assist flexion and extension under different circumstances.¹⁴

The fibers of the external oblique run in an anteroinferior direction from attachments on the lower eight ribs to insert along the anterior rectus sheath and anterior wall of the iliac crest (Fig. 5–6). The external oblique fibers are almost perpendicular in direction to those of the internal oblique fibers. This muscle courses transversely only in its lowermost portion, with most of the muscle running anteriorly and proximally from its origins on the lumbodorsal fascia and anterior two thirds of the iliac crest. It inserts on the lower three ribs and rectus sheath anteriorly. The transversalis abdominis, the deepest muscle of the group, runs transversely like a horizontal girdle from the lumbodorsal fascia, anterior iliac crest, and inner surface of the lower six ribs. The main mass of the muscle inserts into the linea alba in the midline. In the act of flexion, it is probable that the abdominal muscles act not only to create a ventral moment, but also to stabilize the spine posteriorly through their action on the lumbodorsal fascia. The innervation of the abdominal muscles is shared via intercostal nerves from root levels T7-T12. The thoracic nature of the innervation means that these muscles are spared from radicular-type injuries.

FIGURE 5–6 External oblique muscle.

(From Finneson B: Low Back Pain, 2nd ed. Philadelphia, JB Lippincott, 1977.)

Gluteal Muscles

The large muscles of the buttocks, chiefly the gluteus maximus, gluteus medius, and gluteus minimus, act variously as hip extensors and abductors. As such, they act as motors to the spinal “crane” in forward bending and twisting movements. They also provide the “spinal engine” for locomotion (Fig. 5–7).^13,¹⁴ The gluteus maximus receives innervation via the inferior gluteal nerve and is primarily an S1 muscle, although it receives contributions from L5 and S2 root levels. The superior gluteal nerve innervates the gluteus medius primarily with L5 contribution. Although it receives contributions from L4 and S1 root segments, it is often an important “internal verifier” of a true L5 radiculopathy when combined with clinical or electrodiagnostic abnormalities in distal L5 muscles.

FIGURE 5–7 Musculature of buttocks and proximal thigh.

(From Hollinshead WH: Anatomy for Surgeons, 3rd ed. Hagerstown, MD, Harper & Row, 1982.)

Posterior Thigh Musculature

Muscles attached to the ischial tuberosity, such as the hamstrings, are also strong pelvic extensors acting around the hip fulcrum. They provide powerful assistance to the buttocks musculature in raising or resisting lowering of the pelvis. Additionally, the hamstrings provide efficient passive restraint on pelvic flexion when the knees are locked in extension. The hamstrings are the inferior restraint providing the most efficient forward flexion by controlled forward rotation of the pelvis around the hips. These muscles, which include the two heads (long and short) of the biceps femoris and semimembranosus and semitendinosus, are innervated by the tibial portion of the sciatic nerve (except the short head of the biceps from the peroneal division of the sciatic nerve) and confirm lower root dysfunction (L5, S1, and S2).

Rectus Femoris

The rectus femoris serves a similar but weaker role than the iliopsoas complex in transmitting force from the spine to hip and pelvis motion segments. The rectus femoris crosses the hip and knee joint and has a primary role in creating a more efficient and synchronized gait. Intrinsic spinal muscle limit the sagittal motion transmitting forces to the lower limbs. Coordination of these muscle groups is fundamental to the spine and pelvis functional unit to supporting bipedal function and efficiency. Patients with cerebral palsy are exemplars of the importance of spine and pelvis coordination insofar as hip flexor spasticity and contracture create a kyphotic imbalance that prevents or extinguishes gait. The rectus femoris is innervated by the femoral nerve and L2 and L3 root segments with minor L4 contribution.

Electrophysiology

A great deal of energy has been devoted clinically to the use of the EMG signal for detecting radiculopathy in the lower extremity musculature associated with lumbar disc derangements. EMG needle electrodes have been used in the extremities to detect characteristic denervation, reinnervation, and muscle changes to muscle fibers within a discrete motor unit that is seen with nerve injury or muscle injury. These tests are often accompanied by nerve conduction velocity studies to assess peripheral nerve function, serving as a check to rule out other pathologic conditions, such as peripheral nerve disease, motor neuron disease, or myopathies that may mimic intracanal spinal lesions.

EMG employed to analyze normal function, as opposed to pathologic conditions, of the spine musculature is less well understood. There has not been a concerted research effort to document the natural history of adaptive electrodiagnostic findings as patients evolve from symptomatic to asymptomatic. This section discusses the clinical utility and limitations of electrodiagnostics further. In assessing low back pain without corresponding neurologic sequelae in the extremities, EMG and nerve conduction velocity remain investigational. Biomechanical studies have often relied on use of the raw, integrated EMG or root mean square EMG signals to estimate muscle loads in studies involving lifting performance in normal subjects. Trunk stabilization and initiation of motion are two particular patterns of EMG activity that have been identified in trunk movements.²⁵^–²⁷ Different movements recruit muscles in varying patterns of activity. Intrinsic spine musculature supplies a net sum of movement using various lever arms to initiate and arrest motion by combining concentric and eccentric contraction in either movement or maintenance of posture.

Longissimus and other paravertebral muscles are frequently silent in the “flexion-relaxation” position, in which the maximally flexed spine is “hanging on its posterior ligaments.” It is also relatively quiet in gentle extension of the spine, but with full extension, lateral bend, or torsion, the role of the intrinsic musculature is as a “balancer” of spinal motion—shown by its prominent activity.²⁶ Subsequently, it was found that loading of the flexed spine in the posture normally producing EMG silence leads to an increase in myoelectric activity in the intrinsic spine musculature proportional to the load applied. This proportion of increased EMG and muscle activity is similar to the changes seen in loading the spine in the upright position.²⁸ Presumably, these increasing loads stress and stretch the posterior ligamentous structures sufficiently to require compensatory muscle firing. The multifidus and rotatores muscles have similar activity in sagittal plane movements. They are active in rotation to the contralateral side, however, in conjunction with bilateral abdominal muscular contraction.²⁰

These muscles also achieve flexion-relaxation silence when the spine is in the ligamentous support phase. Early investigators discovered that the slouching or full flexion seated posture (often considered “bad posture”) is, in reality, quite comfortable for prolonged periods and that EMG silence is generally maintained in the erector spinae.²⁹ Issues of posture, movement, load, and speed have usually been studied through surface EMG measurements.³⁰ This “silent period” is also known as the flexion-relaxation phenomenon.³¹ The phenomenon represents an EMG pattern seen in most normal subjects but frequently absent in subjects with chronic low back pain. These patients show elevated muscle activity during full voluntary trunk flexion and fail to achieve flexion-relaxation.³²^–³⁶ The phenomenon is noted with needle and surface EMG measurements, which has led many researchers to use surface EMG in studying the phenomenon.³⁷

More recent research suggests that the hypothesis of a dichotomous pattern (muscles either “on” or “off”) is likely misleading. Physiologically, lumbar muscles show electrical activity because they are contracting, and the degree of electrical activity is roughly proportional to the rate and amplitude of muscle fiber firing. Lumbar muscle activity is almost never “completely silent” during trunk flexion maneuvers. More recent descriptions of the phenomenon suggest a significant contractile force associated with high surface EMG activity during flexion, followed by a relaxation phase associated with low surface EMG activity.^38,³⁹

Several researchers have also proposed quantitative formulas for defining the presence or absence of flexion-relaxation.^36,^40,⁴¹ More recent work has compared inclinometric range of motion and surface EMG measures, with several interesting findings.¹⁹ Using this quantified approach, almost all normal subjects can achieve flexion-relaxation, even if they lack completely normal spine motion. Even in patients with chronic low back pain who are symptomatic and completely disabled, approximately 30% can achieve flexion-relaxation before rehabilitation. After a functional restoration program that stressed improved lumbar mobility with pain management techniques aimed at the lumbar musculature, however, 94% of the patients completing treatment achieved flexion-relaxation. Larger studies performed subsequently show that quantified lumbar flexion-relaxation phenomenon can be an objective tool for measuring improvement in a functional restoration program, correlated to improvements in lumbar range of motion measurements.⁴² Almost all the improvement in achieving flexion-relaxation is directly attributable to a component of the interdisciplinary program that involves surface EMG biofeedback involving surface EMG–assisted stretching training.⁴³

Contractions of the abdominal musculature, particularly muscles attaching to the lumbosacral fascia, are also important in maintaining flexed postures and initiating extension. The lateral pull on the lumbodorsal fascia has two effects: creating a tightening of the craniocaudal dimension in the lumbar spine (“guy-wire model” discussed earlier) and “encapsulating” the intrinsic spine musculature to provide greater efficiency. In so doing, the abdominal mechanism also serves as a force-coupling mechanism: eccentrically controlling extension of the flexed spine while resisting flexion loads.¹⁴

Some groups have tried to correlate functional tasks and training, EMG data, and computed tomography (CT) and magnetic resonance imaging (MRI).⁴⁴ Small studies have shown correlations between postoperative strength deficits, imaging of muscle atrophy (decreased cross-sectional area or fatty infiltration), and changes in EMG characteristics.^18,^45,⁴⁶ CT seems to provide reliable measurements of paraspinal muscle cross-sectional area and density in normal subjects and patients with chronic low back pain.⁴⁷^–⁴⁹ Although some early pilot work at several institutions is intriguing, serial imaging remains cost prohibitive and is of little value for altering treatment outcomes. More recently, needle EMG sampling of the multifidus has been described in the diagnosis of lumbar spinal stenosis.

Trunk Muscle Strength

The obvious relationship between extremity joints and strength of the contiguous musculature in athletic and pathologic (traumatic, arthritic, or deconditioned) situations has stimulated many investigators to study similar relationships in the spine.^29,⁵⁰ Early investigators were limited to the use of cable tensiometers and isometric and isotonic machines. Work began 30 years ago using individually modified isokinetic dynamometers in various positions.⁵¹^–⁵³ Later technology provided computerized isokinetic devices for separately measuring isolated lumbar trunk strength in the sagittal and axial planes concentrically and eccentrically that also allowed measurements isometrically and isotonically.^20,^41,⁵⁴^–⁵⁶

Isokinetics has been deemed a safe way to measure muscular output.⁵⁷^–⁵⁹ Studies have shown a diversity of normative data owing to factors such as different protocols, positioning (reclined, sitting, or standing), instrument, study samples, gravity correction, and axis of rotation. The normative data are specific to instrumentation, protocol, and positioning.⁶⁰^–⁶³ The development and availability of isokinetic dynamometers or attachments specifically for back testing increased beginning around 1980. The number of manufacturers of isokinetic dynamometers available to test the back has decreased since then as a result of a combination of managed care reimbursement issues and lack of training for physical or occupational therapists likely to use such devices. At the present time, only Biodex (Shirley, NY), CSMi (formerly Cybex; Stoughton, MA), and Technogym (Gambettola, Italy) remain, limiting the development in the field of quantitative functional measurement of the spine (Fig. 5–8).

FIGURE 5–8 Modern-day test isokinetic equipment for back extension and flexion in semistanding position and lift analysis.

Cady and colleagues ⁶⁴ showed a relationship between physical fitness and back injury rates that has been confirmed through isometric lifting testing in several industrial environments. Blay and colleagues⁶⁵ supported the use of isokinetic measurements for the assessment of trunk muscle strength in scoliotic patients. Results indicate that the reliability of trunk testing was satisfactory (Fig. 5–9). These results are in accordance with other studies of trunk testing relating to healthy subjects and patients with low back pain.⁶⁶ Although joint range of motion seems to be an independent variable in trunk function, muscular factors such as trunk strength, endurance, and neuromuscular coordination seem to be critical factors in maintaining the integrity of the lumbar spine. Tests of isokinetic trunk strength in the sagittal plane have revealed a typical gaussian distribution of strength in the normal population. Normalizing strength by body weight narrows the width of the distribution curve. Using other normalizing factors, men seem to be 10% to 20% stronger than women and have greater ability to sustain strength at high speeds. The variety of test methods used to examine trunk muscles makes it difficult to compare results. The assessment requires an understanding of the force relationship throughout the defined range of motion (work).⁶⁷

FIGURE 5–9 Isokinetic torque measurement of isolated thoracolumbar motion segment during dynamic flexion-extension in a normal subject.

Another challenge is the absence of a contralateral measurement within the same person to provide a “gold standard.” Comparison with a normative database stratified by age (past the 5th decade), weight, and gender has emerged as the standard for quantifying strength. Patient inhibition by pain or reduced voluntary effort is often a major factor in low torque output (Fig. 5–10), a pattern that had been identified in sports medicine testing of limbs but not as well documented in patients with chronic low back pain. Age through the 5th decade does not seem to be a critical variable.

FIGURE 5–10 Torque production in dynamic thoraco-lumbar motion between 90 degrees of extension and 10 degrees of flexion in three different populations. A, Normal force production. Note rapid increase in torque and ability to maintain strength throughout motion. B, Postrehabilitation force production. Note similarity to normal force production in shape, although the amount of work (area under the curve) is less than normal. C, Before rehabilitation, force production exhibits a slow increase in force and inability to sustain force through the motion.

Similar findings are noted when measuring isolated axial strength.^20,^68,⁶⁹ Ability to generate trunk rotational torques within the same individual is generally symmetrical, although there is a trend toward slightly greater strength rotating to the dominant side in men, which may be due to training in pulling activities. There seems to be a greater tolerance for applying loads at high speeds in the axial plane. This tolerance may be due to the greater fast-twitch fiber type composition of the lateral abdominal musculature, which is primarily responsible for axial plane lumbar spine movements. Rotational strength depends on the direction; for men and women the force produced from the rotated position toward the neutral position is stronger than the neutral position rotating outward.⁷⁰

In the pathologic state, significant decrements of muscle strength are frequently noted. Patients with chronic back pain show a selective loss of extensor strength compared with flexors and an inability to maintain strength at high speeds.^59,⁷¹^–⁷⁴ In rotation, there is also a substantial decrement in strength, but it seems to be relatively symmetrical in the chronic state and less subject to high-speed variation.⁵⁴ Although most early work was based on peak torque measurements, advances in computerization made measurements of work performed, power consumed, and curve analysis possible, while allowing variability determination to assess “effort.” Because only maximal muscular effort is truly reproducible, variability of curve shape and height becomes a potential measure of effort. In the absence of visual feedback of trunk muscle function to the clinician, these measures become exceedingly important in documenting optimal functional capacity and effort.^75,⁷⁶ Some controversy exists regarding clinical utility and discriminating power of trunk strength testing, partly because of normal human variability and partly because of unrecognized sources of error related to testing procedures.^59,^73,⁷⁷ The discriminating power of a test depends not only on the ability to distinguish the “normal” from the pathologic state, but also on distinguishing prerehabilitation and postrehabilitation performance. Many such longitudinal studies have been performed, including some that correlate spinal strength performance to imaging (e.g., CT and MRI) findings.^18,^45,⁷⁸^–⁸¹

By contrast, supernormal or athletic individuals who have been studied (Fig. 5–11), such as female gymnasts, male soccer players, male tennis players, and male wrestlers, seem to exceed mean torque-to-body weight strength ratios for the normal population by 15% to 40%.⁸² In contrast to pathologic and normal populations, supernormal or athletic individuals show almost no “high-speed drop-off”—that is, decreased torque output at high speeds. They also maintain a very stable ratio of extensor to flexor strength (well-balanced, efficient use of coupled forces).

FIGURE 5–11 Normal force production at two different speeds for an athlete. A, 60 degrees/sec. B, 120 degrees/sec.

The factors that contribute to decreased strength in the pathologic state are not entirely understood. Although muscle atrophy undoubtedly occurs with prolonged disuse and deconditioning, pain may inhibit neuromuscular function through nociceptive reflex feedback mechanisms. Similarly, various psychosocially induced phenomena, such as anxiety, fear of reinjury, or depression, may unconsciously attenuate effort, producing submaximal measurements. At this time, techniques for assessing subject motivation do not seem to be available, although curve variability and ratio of work to peak torque seem to be promising tools for identifying attempts to produce submaximal output consciously.^76,⁸³

Myofascial Pain

In recent years, there has been an increase in the diagnosis of myofascial pain syndromes. Often, there is overlap, with lumbar axial “mechanical” pain (without concurrent extremity pain) having a strong or occasionally predominant myofascial component. Frequently, patients are referred to spine specialists with simultaneous diagnoses of back pain and myofascial pain that do not correlate with radiologic or other diagnostic information. Aggressive use of invasive options often yields poor outcomes. Marginal benefit from pills often leaves the spine specialist at a loss for how to treat this poorly characterized cluster of symptoms that seem to center on primary muscle “dysfunction” generating pain with concomitant patient fear-inhibition creating disability.

A highly used but poorly understood treatment for lumbar myofascial or “back muscle pain” is “trigger point” management. Trigger points themselves are fairly well described, hyperirritable, isolated, focal areas in a “taut band” of multiple fibers of skeletal muscle that produce sustained contraction that fails to relax or release fully, which creates local and referred pain in discrete, predetermined patterns.⁸⁴ Although numerous authors have described the diagnosis and treatment of trigger points, many others have questioned the interphysician reliability of judging trigger points and the more slippery generalized diagnosis of myofascial pain.⁸⁵^–⁸⁸ Although by no means definitive, a growing number of physicians diagnose the widespread, painful summation of multiple areas of hyperirritability nonspecifically as myofascial pain or fibromyalgia—which is treated partially with trigger point injection and pharmacologic intervention.

Although trigger points are a clinical diagnosis, some academicians believe that the presence of a local twitch response (identified most reliably with EMG) is a prerequisite to treatment.^84,⁸⁵ Nevertheless, many patients find relief from interventions aimed at modifying trigger points, even if present in multiple body regions. Treatments directed toward trigger points are widespread among many different physician fields and alternative medicine specialists. Although many practitioners have a favorite “needling” technique, multiple studies have found no statistical difference between “dry needling,” saline injection, or medication.⁸⁹ The preeminence of needle injection as the most effective way to inactivate a trigger point has been questioned in several published articles.^90,⁹¹ Other modalities for refractory myofascial or trigger point pain include ultrasound, manipulation, massage, acupressure, acupuncture, and “spray and stretch” techniques.

Within the last decade, use of botulinum toxin (Botox) has become popular, first to inactivate refractory trigger points and then more generally for use in muscle spasticity associated with back pain. Botulinum toxin, one of the most potent biologic poisons known, is created by a gram-positive, rod-shaped bacillus, Clostridium botulinum, first identified in the 1820s. The toxin is a heat-labile, zinc-dependent, metalloprotease polypeptide composed of a heavy chain and a light chain about 150 kDa in size (too large to cross the blood-brain barrier) that denatures at 80° C.⁹²^–⁹⁴ There are currently eight classified toxin types: A, B, C1, C2, D, E, F, and G. The presynaptic, cholinergic neuromuscular junctions are the target of botulinum toxin. The toxin inhibits the release of acetylcholine, which results in blockage of muscle activation and paralysis.^95,⁹⁶ The toxin targets zinc-dependent endoproteases of the SNARE complex—SNAP-25, synaptobrevin II, and syntaxin I—each involved in the packaging and release of acetylcholine at the postsynaptic nerve terminal.⁹⁷ Each toxin type (A to G) acts slightly differently on the three SNARE proteins, leading to speculation, without clinical evidence, that different toxins may have different clinical uses and safety profiles.

A Medline search reveals 230 publications in the last 10 years that describe direct or indirect reduction of pain as one of the sequelae of Botox/Myobloc use. Despite the sheer number of publications, most of these are pilot or open-label studies and fail to show convincingly a discrete mechanism of action for additional analgesia beyond that of muscle fiber paralysis. At the time of this writing, two types of toxins, A and B, enjoy widespread, off-label use. Their safety profiles seem to be equal. Both inactivate neuromuscular junctions at roughly equal concentrations, “weakening” the muscle, for 70 to 120 days, with a mean peak occurring around 90 days. Over time, the body builds an immunologic response to the toxin, which may eventually render some individuals nonresponders after multiple exposures. This tolerance seems to have a heterogeneous distribution with no clear prognostic data as to who will fail to respond over time. Some reported side effects include respiratory distress, dysphagia, dysphonia, autonomic instability, and temporary impotence. The temporary nature of the muscular dysfunction and the emerging evidence of additional analgesic effects has made this medication extremely popular not only for muscular back pain, but also in a wide assortment of medical fields and for cosmetic indications.

Muscle-Sparing Surgery

In the last decade, improvements in technology, visualization, technique, material innovation, and device performance have led to greater use of lumbar surgical approaches that claim to be “muscle sparing.” These techniques are discussed in greater detail in the rest of the book, but it is worth examining the claims of muscle sparing. As noted earlier, the lumbar spine is a finely balanced biomechanical mechanism that relies on the integration of intervertebral height, joint mobility and proprioception, muscle balance, and osseoligamentous constraint to allow people to function without pain. Although ample redundancy is undoubtedly built into the system, it is hoped that minimizing disruption of biomechanical integrity will lead to better functional outcomes for all spine specialists.

Lumbar surgery has classically involved extensive dissection of the posterior muscle, fascia, ligamentous structures, and occasionally joints that is even more extensive when fusion (with or without instrumentation) is attempted. Additionally, unintended consequences of denervation, compression, retraction, and vascular injury have led to other biomechanical sequelae.^18,⁹⁸ Decreased morbidity by sparing vulnerable structures may serve to maintain mobility and function, improving outcomes.

The posterior lateral approach first described by Cloward ⁹⁹ in 1953 has been controversial.¹⁰⁰ The improved tubular retractor combined with fluoroscopy and improved implantable devices have allowed for a repopularization of this approach, although denervation and devascularization of the deep muscles is still a likely possibility. The more lateral approach first described in 1968 by Wiltse and colleagues¹⁰¹ for fusion allows paraspinal muscles to be bluntly divided along their aponeurosis sparing the muscle, fascia, and some of the ligamentous structures. Adequate decompression of far lateral or foraminal disc herniations can be achieved, but more centrally located pathology is difficult to access from this approach. Good fusion rates are reportedly achieved with marginally increased operative time and conflicting length of stay comparisons as a tradeoff for “sparing” devascularization and denervation injury to muscle.¹⁰² Multiple sources advocate percutaneous or open supplementary instrumentation to the contralateral side to achieve a consistently high fusion rate.

The anterior transperitoneal approach has been described since the 1930s in various iterations.¹⁰³ A spine surgeon can achieve (indirect) decompression, stabilization, fusion (at a variable rate), and now motion preservation through this approach. Often these surgeries lose the benefit of posterior muscle preservation when surgeons later elect for additional posterior stabilization. Additionally, retraction can damage abdominal and anterior lumbar musculature that play a key role in maintenance of proper spine balance. The anterior approach has been modified to go retroperitoneally with laparoscopic devices that allow perceived (unconfirmed) benefit of splitting of the abdominal musculature to allow for more rapid postoperative healing and less perturbation of the abdominal viscera but often at the cost of crossing and possibly injuring the psoas muscle, which may have a very important role in lumbar pelvic coordination of functional tasks. Keeping in mind the advantages of protecting lumbar musculature and maintaining mobility, there is hope that the functional outcomes of spine specialists will continue to improve.