The concepts of stability and stiffness are benchmarks for all human motion, including spine function. Without accounting for stability, almost all human activities could not be successfully performed. Three major tenets with regard to stiffness and stability exist for human function¹⁷:

In the context of the spine, Bergmark described the stability-stiffness continuum using a spring-mass system: the stability of a system (spine) is dependent on the stiffness of the springs that attach to it (the “core” muscles, i.e., rectus abdominis, quadratus lumborum, internal and external obliques).¹⁸ External perturbation (i.e., externally applied loads or movement) can potentially disrupt the equilibrium of the system. To maintain static equilibrium (ensuring the spine experiences no unwanted motion), the core muscles must generate sufficient stiffness. This is akin to how the rigging on a ship stabilizes its mast (Fig. 19.1). The stiffness of the “guy wires” provides stability to the mast to ensure it stays upright against external forces such as the wind or rocking of the ship. Similarly, stiffness generated by the core muscles helps to “anchor” the spine, enhancing its ability to resist external perturbation. Insufficient stiffness of the guy wires may cause the mast to fall over or buckle if excessive perturbations are applied. Similarly, insufficient core stiffness will cause the system to mechanically buckle, resulting in an involuntary bending of the spine when trying to resist external perturbation (i.e., yielding during a heavy deadlift).

However, unlike guy wires, the core muscles have the ability to “fine-tune” the stiffness they generate. This is mitigated through muscular activation: when a muscle contracts it creates both force and stiffness. Though muscular stiffness always adds to joint stability,¹⁹ excessive force can contribute to destabilizing the system. Applied spinal forces will increase with muscular activation, but high levels of activation may prove harmful: increased compressive loads may take away from the spine’s work capacity and ultimately lead to injury. Because peak stiffness is not achieved at full activation of the muscle, in order to achieve optimal stability (maximizing stiffness with minimal force applied), the goal is to not activate the muscle as hard as possible but to finely tune the activation. So, what level of activation is required to achieve optimal stability? It has been demonstrated that stiffness values level off at approximately 25% core musculature maximum voluntary contraction (MVC).^19,20 Based on this information, one goal of training is for the patient to learn how to tune core activation to create “sufficient stiffness”: maximizing spinal stability through maintaining just enough stiffness, while minimizing the amount of internal force generated from this contraction. According to Cholewicki and McGill, spine stability is greatly enhanced by cocontraction of antagonistic trunk muscles.²¹ Cocontractions increase spinal compressive load, as much as 12% to 18% or 440 N, but they increase spinal stability even more, by 36% to 64% or 2,925 N.²² Cocontractions have been shown to occur during most daily activities.²³ This mechanism is present to such an extent that, without cocontractions, the spinal column is unstable even in an upright posture!⁶

Cocontractions are most obvious during reactions to unexpected or sudden loading.^24,25 Stokes et al has described how there are basically two mechanisms by which this co-activation occurs²⁶: (a) a voluntary pre-contraction to stiffen and, thus, dampen the spinal column when faced with unexpected perturbations; and (b) an involuntary reflex contraction of the muscles quick enough to prevent excessive motion that would lead to buckling after either expected or unexpected perturbations.^{24,25,26,27,28,29}

One mechanism of tissue injury is repetitive, end-range loading. A stable spine can avoid injurious, repetitive, end-range loading via the buttressing effect of agonist-antagonist co-activation in maintaining the integrity of the “neutral zone.” The neutral zone is the inner region of a joint’s range of motion (ROM), where minimal resistance to motion is encountered.³⁰ This inner region’s mobility is restricted by passive ligamentous factors alone, and, when it is expanded (i.e., spinal flexion), creates joint instability, which places greater demands on the muscles to stabilize a joint. Thus, the most observable and measurable sign of instability is not joint hypermobility, but excessive agonist-antagonist muscular co-activation.³¹

Various studies have pointed out how important the motor control system is for preventing spinal injury. Ironically, when under load, the spine is best stabilized, but when “surprised” by trivial load at a vulnerable time, such as in the morning or after prolonged sitting, the spine stability system is most dysfunctional.^32,33 Inappropriate muscle activation during seemingly trivial tasks, such as bending over to pick up a pencil after sitting for a prolonged period of time, can compromise spine stability and increase the likelihood of buckling of the passive ligamentous restraints.³¹ This motor control skill has also been shown to be more compromised under challenging aerobic circumstances.³³ The basic science aspects of the spine stability system are presented in greater detail in Chapters 1 and 5.

Coordinated activation and control (motor control) of the core musculature is a major factor contributing to stabilizing the spine and preventing LBP. Stiffness created by activation of the core muscles arrests painful spinal micromovements, and buttresses the spine against unwanted motion when challenged by external perturbations and/or loads.¹⁷ However, dysfunctions in the timing, magnitude, and pattern of activation can contribute to pain and impaired spinal function.

Radebold et al have shown that there is a predictable muscle response pattern to sudden trunk loading in individuals with LBP.³⁴ This includes delayed initial activation, over-activation, and delayed subsequent relaxation of muscles. Researchers in Queensland, Australia, have found that delayed activation of the transverse abdominis during arm or leg movements distinguishes LBP patients from asymptomatic individuals.^35,36 O’Sullivan found that an abdominal hollowing maneuver helped retrain perturbed motor patterns in deep abdominal muscles.^7,8 This has motivated the use of hollowing exercises for use in rehabilitation programs for patients with segmental spinal instability. Unfortunately, others have misinterpreted these data to mean that hollowing maneuvers or minimizing rectus abdominis activity during core stability exercises directly enhances stability. In the context of the previous guy wire analogy, focusing on training a single muscle is like focusing on a single guy wire.³⁷ Research from the University of Waterloo in Canada has found
that while certain muscles, such as multifidus and transverse abdominis, may have special relevance in distinguishing LBP subjects from asymptomatic individuals, these muscles are part of a much bigger orchestra responsible for spinal stability.³⁷ This effect was explored further by Vera-Garcia et al, who showed that abdominal bracing performed better than abdominal hollowing for stabilizing the spine against rapid perturbations.³⁸ Specifically, bracing actively stabilized the trunk and reduced the lumbar spine displacement after loading, but hollowing was ineffective for buttressing the spine against external perturbation. Brown and McGill also showed that coordinated activation of the abdominal wall (external oblique, internal oblique, transverse abdominis) had a synergistic effect in force and stiffness transmission.³⁹ Though this study was performed in situ in rat specimens, the take-home message here is the coordinated activation of the core muscles (i.e., abdominal bracing) has a synergistic effect in creating core stiffness, rather than activating a single muscle (i.e., transverse abdominis through abdominal hollowing). Sufficient stability, according to Brown and McGill, is defined as the amount of muscle stiffness necessary for stability along with a safety margin.¹⁹ Cholewicki et al showed that modest levels of co-activation are necessary, but if a joint has lost its stiffness, greater amounts of co-activation are needed.^40,41

Marras et al have reported that there is a different pattern of antagonist muscle co-activation in LBP individuals than in asymptomatic individuals’ torso kinematics (motion) while performing a functional lifting assessment.¹⁶ Patients were found to have greater spine load and less ability to generate normal trunk movement patterns during lifting tasks. Altered trunk movements were strongly related to spine load, being able to predict 87% of the variability in compression, 61% in anteroposterior shear, and 65% in lateral shear. The kinematic picture for the LBP individual showed excessive levels of antagonistic muscle co-activation, which reduced trunk motion, but increased spine loading.

Hodges and Moseley presented a proposed model for the interaction between pain and motor control.⁴⁴ In this model, psychological factors related to fear, stress, and attention from pain/nociceptor stimulation interact with a complex pathway of various brain centers to affect motor control. The researchers stated it is unlikely that the simple inhibitory pathways (i.e., cortical inhibition, reflex inhibition, motoneuron inhibition) can mediate the complex changes in motor control of the trunk muscles, but the most likely candidates are changes in motor planning via a direct influence of pain on the motor centers, fear avoidance, or because of changes in the sensory system.

Although impaired motor control can lead to pain and dysfunction, it is possible to retrain these patterns using motor control exercise (MCE)—exercises that improve coordination and control of muscles that support the spine. A systematic review by Saragiotto et al revealed that MCE plays a large role in reducing pain and dysfunction, and improving quality of life.⁴⁵ A detailed summary of the efficacy of exercise for spinal stability can be found in the following section.

We can summarize that the ability to voluntarily control core musculature stiffness can provide relief and long-term benefit to an LBP patient via enhancing spinal stability. However, the “trainability” of core stiffness is an important question to answer: can we train to enhance core stiffness and if so, what is the most effective way possible?

Lee and McGill showed that it is possible to enhance core stiffness through a single session of isometric core exercise, and following a 6-week isometric
training protocol. A single bout of core exercise (5 sets of 10-second hold planks, side bridges, and bird dogs) showed significant increases in core stiffness, though how long this effect lasted is unknown.¹³ Both untrained individuals and individuals savvy to core exercise experienced increased measures of stiffness. In comparing long-term adaptations to isometric versus dynamic (repetition-based) core exercises, a 6-week isometric training protocol was also shown to be superior to a dynamic protocol for enhancing stiffness.¹² It should be noted that both of these studies examined asymptomatic individuals, and may need to be modified to suit a pained population’s needs.

Australian researchers have reported that multifidus atrophy occurs with acute LBP.⁴⁶ The atrophy does not spontaneously go away when the pain does.⁴⁷ However, motor control training does restore the multifidus and reduces long-term recurrence rates of LBP.⁴⁸ Tsao and Hodges showed that specific training of the transverse abdominis restored function of this muscle to resemble profiles of non-pained individuals in participants with LBP.⁴⁹ This effect was observed with just a single training session and the magnitude of change correlated with the quality of training. Long-term adaptations were measured over a 4-week training period, and this change was retained even after a 6-month follow-up.⁵⁰ O’Sullivan et al showed that specific spine stabilization exercises achieved superior outcomes to isotonic exercises in chronic patients with spondylolisthesis.^7,8 In a large, randomized, controlled clinical trial, Timm showed that exercise was superior to passive care in treating failed back surgery in patients.⁵¹ In this study, a further comparison of exercise types showed that low-technology exercise (i.e., McKenzie and stabilization) was superior to high-technology exercise (i.e., isotonics and Cybex).

Yilmaz et al administered an 8-week stabilization program to postoperative lumbar microdiscectomy patients,⁵² and compared this program to home exercise and no exercise. At week 12, superior results were achieved in pain, function, mobility, and lifting ability for the stabilization group. Supervised stabilization training was superior to home exercises, which was superior to no exercise.

Stuge et al found that stabilizing exercises were superior to traditional physical therapy for pelvic girdle pain after pregnancy.⁵³ The stabilization group had lower pain intensity and disability, higher quality of life and less impairments. The results persisted at the 1-year check, postpartum.

Safe exercises for acute and subacute low back patients should have favorable biomechanical load profiles. The goal of these exercises is to create adequate core musculature challenge for the patient in order to strengthen tissues and enhance their load tolerance, while minimizing imposed spine loads. It is known that without muscles, the spine buckles at 90 N of compression (under its own weight!), yet, during routine activities of daily living (ADLs), compressive loads 20 times are routinely encountered.³⁰ Thus, proper functioning muscles controlled by the central nervous system enable stability to be maintained. In fact, demanding ADLs involve compressive loads of approximately 6,000 N, and the National Institute for Occupational Safety and Health (NIOSH) work demand limit (a threshold value for elevated risk of back injury set by the NIOSH) is 6,400 N.^6,54,55 Elite weight lifters manage, through highly skilled motor control strategies, to safely lift loads of nearly 20,000 N.⁵⁶ McGill recommends that, for subacute exercise training, a safe limit is approximately 3,000 N.⁵⁷ Table 19.1 lists exercises with varying safety levels based on loads imposed to the spine.

McGill’s recommendation helped influence the development of his “non-negotiable Big 3” core exercises (curl-up, side bridge, bird dog—see Stability Training section for descriptions of these exercises). Traditional core exercises involving motion of the spine (i.e., sit-ups, back extensions) provide a high degree of challenge to the core musculature but impose excessive spinal loads (>3,000 N).⁵⁸ The “non-negotiable big 3” provide a much safer and effective alternative, imposing much lower spine loads while still creating a high degree of neuromuscular challenge.¹⁵ This is why they are considered non-negotiable—they provide a biomechanical onramp with low-load exercises for more sensitive patients. Selecting the wrong exercises will only worsen
their condition and adds to the myth that exercise is not well tolerated for individuals in pain. LBP patients can perform these exercises daily with minimal risk of overloading their spine, thus more effectively inducing stability and motor control adaptations to remove and prevent pain. Performed daily, these exercises allow patients build endurance of their core musculature allowing them to tolerate greater applied loads. Over time, these exercises can be progressed in terms of neuromuscular challenge to suit the patients’ functional needs.

Specificity of Training The specificity principle is often referred to as the SAID principle (Specific Adaptation to Imposed Demands). This means that the locomotor system will specifically adapt to the type of demand placed on it. Evidence shows that training leads to length-, task-, and speed-specific changes.^55,59,60 For example, long-distance running will improve cardiovascular endurance, but not speed. Also, if a person regularly weight trains with maximal resistance and few repetitions, this will produce greater strength or power gains, but little endurance gains.

During the first 1 to 2 months of training, rapid improvement, as much as 100%, in weight-lifting ability occurs. However, if unrelated tasks are attempted, gains will be less than 20%.

The more similar the exercise is to the actual activity (i.e., position, whole body coordination, speed, resistance), the greater the likelihood that improvements in function at home, sports or work will occur. This is known as the transfer-of-training effect. Therefore, if training programs do not address the specific functional needs of the individual, the goal cannot be achieved.

Endurance training of agonist and antagonist cocontraction ability about a joint has been shown to improve joint stability by enhancing muscle stiffness.³¹ This does not require a very strong muscular effort. Hoffer and Andreassen showed that efforts of just 25% MVC provided maximal joint stiffness.⁶¹ A prolonged tonic holding contraction at a low MVC is ideally suited to selectively train type 1 tonic muscle fiber function. According to McArdle et al, tonic fibers only operate at levels less than 30% to 40% MVC.⁶²

Isometric holds should be no more than 7 or 8 seconds based on recent infrared spectroscopy indicating rapid loss of available oxygen in muscles contracting at mild to moderate levels of intensity (<50% of MVC).⁵⁴ In athletic, nonpainful populations the isometric hold time can be increased dramatically.

The variables in stability training include intensity, sets, repetitions, hold times, form, frequency, and duration (Table 19.2).

Patient reactivation is a gradual process. Behavioral medicine or sports psychology tenets of “paced activity” and the relationship between hurt and harm should be discussed with the patient. Many LBP patients have excessive fear avoidance beliefs or catastrophizing behaviors that promote a passive, symptom-driven approach, excessive pathoanatomic diagnostic testing, and a poor prognosis.⁵⁹ At the other end of the spectrum are individuals who are overly aggressive, which can lead to a “boom or bust” mentality.

The middle path is best exhibited by the modern emphasis on quota-based graded exposures.⁶³ This operant conditioning model successively demonstrates to patients that hurt does not necessarily equal harm, and that activity—contrary to the patient’s pain expectancy or fear avoidance beliefs—is actually beneficial. In graded exposure training (GET), the patient’s activity levels are gradually increased in a stepwise manner limited by quota, not pain.^64,65 For individuals at the other end of the spectrum who ignore pain and continue with or complete activities that may be harmful, GET is equally important.⁶⁶ GET or pacing is important and ensures that either too little or too much activity is avoided.^67,68,69

It is important that the clinician prescribes only those exercises that have a large safety/stability margin. Such exercises should be mutually agreed on with the patient. They should be performed to a quota even if mildly uncomfortable. In chronic pain patients, the expectancy of reinjury is typically based on an activity avoidance belief or catastrophizing tendency and not an actual experience.^26,70 Ciccione and Just showed that in susceptible individuals, there is discordance between pain expectancies and actual pain intensity with activities.⁷¹ Vlaeyen et al have demonstrated that GET can change the individual’s fears and beliefs.⁷²

GET starts with baseline testing to identify feared activities (activity intolerances [AIs]). Initial quotas should be set to subthreshold levels to assure success. Then, patients are gradually exposed to their feared stimuli so they can experience that it is safe to do so.^73,74 Posttreatment auditing of previously provocative maneuvers is crucial to “prove” the effectiveness of the program.

Patients should be given self-care advice before passive interventions, so that they attribute their progress to self-care (see Chapter 14).^7,8 Such positive attribution to self-management is motivational.^68,69 The McKenzie system is designed to facilitate this self-attribution via its rigorous audit process (posttreatment reevaluation) (see Chapter 17).

Chronic patients should be educated to expect relapses. Flare-ups are not failures to manage the pain. The challenge is to learn how to better self-manage such “flare-ups.”^68,75 It is important for patients to learn that there are both first aid and preventive, conditioning selfcare programs. Chronic patients must master both.

Individualizing the exercise prescription necessitates that the clinician have a strategic plan. What follows is a step-by-step guide to customize the right exercise for each patient.

First, identify the patient’s AI. This is derived from a history of things in their daily life that they are having trouble performing. Restoring these functional abilities should be mutually agreed on as the goal of care.

Second, identify the patient’s capabilities or functional deficits. This is derived from an examination of capabilities, and is termed the patient’s functional range (FR). Dennis Morgan defines this as “the range of movement which is both painless and appropriate for the task at hand.” The FR emerges from a rigorous functional assessment of both a patient’s mechanical sensitivities (MS)—what the patient feels—and abnormal motor control (AMC)—what the clinician observes. Whenever possible, the patient’s functional capacity or deficits should be quantified with reliable tests that have normative databases. This is a key to establishing the patient’s initial baselines that will be used later to monitor progress.

Third, identify exercises and supportive treatments that can close the gap between the patient’s functional goals (AIs) and capabilities. The exercises prescribed should ideally be performed without MS or AMC and be relevant for the patient’s goals and deficits.

Fourth, regularly audit (recheck) the patient’s AIs and FR (AMC and MS) to determine if the exercises are achieving the twin goals of reducing AIs and enhancing capabilities. McGill calls this finding the positive slope with the exercise(s).¹⁷ The goal is to find this as quickly as possible. However, in severe, acute pain when inflammation is predominant, or in disabling, chronic pain cases in which central sensitization is present, it can take a few sessions of empirical trial to establish the unique exercise prescription that yields a positive slope. Follow this audit frequency:

Table 19.3 summarizes this prescriptive approach. Motivating patients to stay in their painless range is easy. When the patients are in acute pain, they should avoid what hurts them. In this phase, hurt and harm may go hand in hand. The art of the McKenzie approach is successfully examining the patient to find the painful (or pain peripheralizing) movement and the pain centralizing or reducing movements. Once this is done, it is easy to teach the patient what positions and movements to avoid and which to repetitively perform by recommending those that reduce or centralize symptoms.

Motivating patients to perform exercises appropriately (biomechanically correctly) is not so easy. Pain is not a sufficient guide. In fact, patients often use “trick” movement patterns, with excessive global muscle substitution, to increase repetitions, thereby reinforcing dysfunction (i.e., stooping excessively during squat exercises). Unfortunately, many dysfunctional movements don’t hurt in patients with chronic pain. Also, certain movements, such as stretches for tissues that have adaptively shortened, do hurt, but are not harmful.

How patients acquire the skill of “core” stability during functional activities generally follows certain established stages of motor learning (Table 19.4).⁷⁶ These may be unnoticed by the patient, but the astute clinician guides the patient effortlessly through these stages with the help of encouraging and facilitatory cues, contacts, resistance, commands, etc.

The first stage of motor learning is the cognitive-kinesthetic stage. Most patients have poor kinesthetic awareness of how to produce and/or control motion of their problem area. In this first stage, the patient learns to “discover” how to move an important region, such as the lumbo-pelvic, scapulo-thoracic, or cervico-cranial. They acquire the skill to perform the movement and then to limit it to a “painless” or pain centralizing range. Examples include the following:

Lumbo-pelvic control—cat-camel

Scapulo-thoracic—shoulder rolls or Brügger position

Cervico-cranial—nodding of the head as if saying “yes”

Movement control then progresses to the second stage of motor learning called the associative stage. This is entered when the patient has sufficiently developed the kinesthetic awareness to move within their FR, so they can safely and appropriately perform more complex exercises using a “key” region. For instance, a progression in level of motor control difficulty occurs when a movement progresses from simple, unresisted concentric motions (i.e., cat-camel) to one requiring isometric “core” stabilization during peripheral mobilization (i.e., quadruped single-leg reach).

The third stage of motor learning is the autonomous stage. This is accomplished when the patient does not have to think about the exercise to perform it properly. This is demonstrated when the patient can control lumbo-pelvic posture during an exercise despite unexpected perturbations from a labile surface (i.e., stability pad, mini-trampoline) or from the clinician (quick, gentle pushes). In this way, the unexpected nature of “real-life” situations involving sudden movements and jostles are trained.

Because many patients lack the motivation to concentrate during exercise training, it is crucial to reach the autonomous stage of an exercise as quickly as possible. Additionally, the most functional exercise that the patient can perform with good stability is always a wise choice because compliance will be better. Gary Gray terms this “attacking success.”⁷⁵ Vladimir Janda also emphasized this point by suggesting that exercises such as sensory-motor training on a rocker board were preferred to cortically demanding training, such as complex floor exercises because it automatically trained the stability system without requiring much voluntary control.⁷⁷

Part of the art of prescribing exercises is determining how long to stay with exercises that require the patient to hypervigilantly (voluntarily) control posture versus having the patient practice simple, functional movements that they automatically (involuntarily) perform with good “core” control. This is an important clinical issue because modern stabilization training may require 4 to 6 weeks of practice before abdominal hollowing can be learned.¹⁷ Karel Lewit⁷⁸ sums it up thusly, “remedial exercise is always time consuming, and time should not be wasted … We should not attempt to teach patients ideal locomotor patterns, but only correct the fault that is causing the trouble.”

There are innumerable different ways to progress patients. Table 19.5 outlines a few of the most important ones.