Specialising in musculoskeletal, orthopaedic, spinal and sports rehabilitation

There has been a significant amount of work published in the past 15 years investigating the essence of spinal stability (Brown et al., 2003; Chiang and Potvin, 2001; Cholewicki et al., 1997 and 1999; Essendrop et al., 2002; Hodges et al., 2001; Quint et al., 1998; Solomonow et al, 1998 and Wilke et al., 1995). Much of this work is applicable to the investigation behind the concept of core stability. This chapter reviews the evidence behind the anatomical, biomechanical and physiological principles of core stability. Table 8 displays the evidence table for intervention studies that was used when considering guideline statements for this section.


It has been widely reported that spinal stability is primarily the result of interaction between different components of the neuromuscular system (Gibson and McCarron, 2004; Granata and Orishimo, 2001; Kaigle et al., 1995 and Panjabi, 1992a). This can provide both dynamic, controlled movement and variable degrees of stiffness dependent on the functional requirements of the task being performed by an individual or the external forces acting on the individual (McGill et al., 2003).


There is a considerable variation in the literature from both scientific sources down to the popular press that influences much of the fitness industry on what anatomical structures provide the foundation for core stability. The fitness industry is responsible to a large extent for the practical application of core stability training methods to the general population. Unfortunately, much of the material on which they base their programmes are derived from poorly researched publications that are highly biased towards anecdotal evidence and even influenced by companies that produce ‘fitness products’. The aim of the following section is to provide a comprehensive review of what structures have been shown to contribute to core stability and to highlight areas that warrant further investigation.


Simply stated core stability is a term that would appear to encompass structures that provide stability to the trunk. Soft tissue structures that connect the pelvis, spine, ribs and shoulders that have varying levels of responsibility in terms of movement, stability and combined function would make up some of the essential components needed to provide adequate core stability. Panjabi (1992a) described three subsystems that provided a conceptual model for spinal stability: passive, active and neural. The passive subsystem was described as being made up of the vertebrae, discs, joint capsules and ligaments (Panjabi, 1992a). Notably he did not specifically include the pelvis, ribs or scapulae in his model. The active system was described as consisting of the muscles and tendons surrounding the spinal column. Thirdly the neural and feedback subsystem constituted the various force and motion transducers, located in ligaments, tendons, and muscles, and in the neural control centres. Core stability could be considered as the interplay between these three components .


To provide structure to the following the above subsystems will be used as a format to describe the components of core stability. Whilst the concept behind these terms has been proposed by Panjabi (1992a) the following discussion is not necessarily an expression of his views rather a representation of the available evidence from data searches performed.


Passive Subsystem


The components of the passive subsystem do not provide significant stability to the spine whilst in the neutral zone (Panjabi 1992a and 1992b). They provide soft tissue attachment points for muscles and fascia of the active subsystem and attempt to provide a safe anatomical passage for some of the structures that make up the neural and feedback subsystem. The osseous structures that fit within the core stability framework, starting inferiorly, include the pelvic girdle made up of the coccyx (4 fused vertebrae), sacrum (five fused vertebrae) and 2 ilia (Williams et al., 1989). The lower limb connects with the pelvic girdle directly through the connection with the hip joint (Calais-Germain, 1993). The spine interacts with the pelvic girdle through the lumbosacral junction at L5/S1(Bogduk, 1997). The sacroiliac joints within the pelvic girdle play an important role in controlling torsional movements through the lumbopelvic region (Vleeming et al., 1990a and 1990b; Lee, 1999). The lumbar spine is made up of 5 lumbar vertebrae (Bogduk, 1997). The thoracic cage is made up of 12 ribs on each side of the thoracic spine (Edmondston and Singer, 1997). The shoulder girdle articulates with the thoracic cage at the scapulothoracic joint (Motttram, 1997). This effectively provides the superior border for core stability structures and has a similar function as the pelvic girdle but for the upper limb that is, creating a stable platform for the upper limbs to move from and to control the initial transference of upper limb forces to the trunk (Mottram,1997).


Two adjacent vertebrae constitute a spinal segment (Bogduk, 1997; Norris, 2000). They are attached (or articulated) by joints, ligaments, capsules and overlying muscle attachments (Maitland, 1997). There are 3 joints at each spinal segment, the ‘articulating triad’(Norris, 2000). The intervertebral disc forms the joint between the vertebral bodies and the two zygapophyseal facet joints are made up of the inferior articular processes on either side of the upper vertebra and the superior articular processes on either side of the lower vertebra (Norris, 2000; Williams, 1989). The zygapophyseal facet joints have a function in controlling movement and transmitting loads (Goel et al, 1993). They are not designed to be weight bearing joints and only become so in extremes of spinal movement and pathological states such as degenerative disc disease that results in increased facet joint contact pressures and subsequently degenerative changes of the articular surface (Norris, 2000).


Bogduk (1997) described 12 possible degrees of movement at each spinal segment. These possible movements included distraction, compression, translation (lateral, medial, posteroanterior and anteroposterior), side flexion (left and right), axial rotation (left and right), flexion and extension. A coordinated approach by all three subsystems is necessary to control these potential movements (Panjabi, 1992a, 1992b; McGill et al., 2003).


From a biomechanical perspective sacroiliac joint stability has been described as being derived from a combination of form and force closure (Liebenson, 2004, Vleeming et al., 1990 a and b). Form closure is related to the inherent anatomical properties of the sacroiliac joint (Vleeming et al, 1990a and 1990b). Force closure is related to how the muscles, ligaments and the thoracolumbar fascia aid in stabilising the pelvis (Pool-Goudzwaard et al., 1998). During activities such as walking a unilateral force is transmitted up the leg and introduces a shear force through the sacroiliac joint. The muscle-ligament-fascia system works to compress the sacroiliac joints together to ensure stability and hence force closure (Liebenson, 2004a, Vleeming et al., 1990a and 1990b). The study by Avery et al. (2000) demonstrated a possible role for the pelvic floor muscles in the force closure system. The combined evidence for this concept is moderate. No studies were identified that disproved or opposed the above principles.


The ligamentous structures of the spinal segments and the pelvic girdle have definite biomechanical roles in controlling end of range spinal movement and may also factor in the neural feedback system (Panjabi, 1992a). Solomonow et al. (1998) demonstrated how stimulation of the supraspinous ligament influenced activity in the multifidus muscle in both cats and humans, thus providing some evidence to this concept. In terms of the passive subsystem at the pelvis the sacrotuberous and long dorsal sacroiliac ligaments are responsible for limiting nutation and counter-nutation, respectively (Liebenson, 2004a). Nutation is the forward motion of the sacral promontory into the pelvis about a coronal axis within the interosseous ligament (Lee, 1999). Conversely, counter nutation is the backward motion of the sacral promontory about a coronal axis within the interosseous ligament (Lee, 1999). These movements can be bilateral or unilateral.


Ligaments of the spinal segment have been classified as neural arch, capsular and ventral (Willard, 1997). The neural arch ligaments constitute the ligamentum flavum, interspinous ligament, supraspinous ligament and intertransverse ligament (Williams et al., 1989). The capsular ligament surrounds the zygapophyseal joint capsule and is reinforced by the ligamentum flavum (Bogduk, 1997). The fundamental biomechanical difference between fibrous joint capsules and ligaments is the arrangement of collagen fibres (Lee, 1999). The functional role of ligaments are to resist tensile stresses between bones therefore the collagen fibres are aligned in a longitudinal arrangement (Bartlett, 1999; Lee, 1999). The collagen fibres within the fibrous joint capsule are arranged in an irregular, random pattern to allow a degree of extensibility to permit joint movement (Lee, 1999). The ventral ligaments are described as being the anterior longitudinal ligament and the posterior longitudinal ligament (Willard, 1997; Norris, 2000).


The ligaments of the neural arch have been considered to act as a single functional structure (Norris, 2000). Dissection studies have described how the deep abdominal muscles may assist spinal stability by influencing the transmission of forces through the ligaments and fascia (Willard, 1997). The following is an example of this proposed feature. The posterior border of the interspinous ligament is thickened into the supraspinous ligament. The supraspinous ligament combines with the thoracolumbar fascia which further merges with the deep abdominals (Norris, 2000). It has been proposed that this chain of interconnecting structures can prevent the ligamentum flavum from buckling towards the spinal cord due to the force generated by the deep abdominal muscles (Willard, 1997; Norris, 2000). This interspinous-supraspinous-thoracolumbar ligamentous complex not only attaches the fascia to the back of the lumbar spine but also creates a pathway for tension in the extremities to be transmitted to the vertebral column (Norris, 2000). This would become an important consideration during rehabilitation and performance training.


The facet joint capsule has been described as a ‘bridge’ of connective tissue connecting the ligaments of the neural arch with the vertebral body (Wilard, 1997). The posterior longitudinal ligament is not the strongest ligament limiting flexion of the spine (Bogduk, 1997). Bogduk (1997) reported that 52% of the resistance to flexion in the lumbar spine is provided by the combined ligamentum flavum and the facet joint capsule. The ligamentum flavum has the highest amount of elastic fibres of any ligament in the body and is held under tension when the lumbar spine is in a neutral position (Butler et al., 1978). The anterior longitudinal ligament is very strong and extends from the occiput to the sacroiliac joint capsule (Norris, 2000). Rapid loading of the longitudinal ligaments results in a ‘stiffening’ response due to their viscoelastic properties. Due to the hysteresis effect these ligaments become stiffer with repeated loading and therefore become prone to fatigue failure (Hukins, 1987). If the local stabilising muscles were working optimally then it would be assumed that their functional role would be to limit the passive structures from being pushed to their failure point. In this evidence based review no studies were found to prove this point.


In reference to anatomy and biomechanics the thoracic spine has been largely neglected compared with the the cervical and lumbar spines (Edmondston and Singer, 1997). The rib cage and its articulations enhance the stability of the thoracic spine (Shea et al., 1996). Berg (1993) hypothesised that the rib cage-sternum complex may comprise an additional ‘column’ for load transfer and stability. No studies were identified that provided evidence for this. A number of investigative studies demonstrated the importance of stability provided by the costovertebral joints (Oda et al., 1996), thoracic ligament complex (Panjabi et al., 1981; Shea et al., 1996) and intercostal fascia (Jiang et al., 1994). The reduced disc height to vertebral body height ratio compared with the lumbar spine reduces motion of the thoracic functional spinal unit (Kapandji, 1978). Zygapophysal joint alignment in the thoracic spine also limits flexion and anterior translation (White, 1969) however enhances torsional stiffness (Singer, 1989).


While there has been considerable research and opinion on the benefits of a neutral pelvic position and the importance of a natural lumbar curve (Chek, 2004; Cholewicki et al., 1997; Norris, 2000; Panjabi, 1992a) the natural thoracic kyphotic curve appears to be primarily related to the osseous asymmetry of the vertebral bodies and tension in the ligament complex (Singer, 11989). There appears to be little supporting evidence to suggest that increasing muscle activity in the thoracic spine will alter the degree of thoracic spine kyphosis. In a load-carrying task, Klausen (1965) reportedly found greater electromyographic activity in the thoracic spine extensors compared with the lumbar extensors however the thoracic curvature remained unchanged while the lumbar curvature was altered. This finding suggested that osseous and non-contractile tissue have a greater influence over thoracic spine curvature than muscular control (Edmondston and Singer, 1997).


Edmondston and Singer (1997) concluded from their review that, “posture correction is achieved predominately through changes in the cervical and lumbar regions while the less mobile thoracic spine remains relatively unaltered.” They acknowledged that the thoracic curvature would influence patterns of load-bearing movement and the degree of stiffness may produce compensatory changes in the cervical and lumbar regions that naturally have more movement. It would appear the more anatomically static thoracic region provides stable attachment points for muscles and fascia that have the potential to influence intra-abdominal pressure and dynamic control of the lumbar spine. There is moderately good evidence of synergistic muscle activity in the trunk producing a stabilising effect (Brown et al., 2003; Cholewicki et al., 1997 and 1999; Essendrop et al., 2002 and 2004; Hodges et al., 2001; Sapsford and Hodges, 2001; Stokes et al., 2000; Radebold et al., 2000; Wilke et al. (1995). In contrast to these studies which were generally from experiments measuring spine stiffness from rapid or informed loading, Kumar (1997) found that with prolonged load the IAP readings were inconsistent when compared with the magnitude of the load applied. The Kumar (1997) study did not record EMG activity of transverse abdominus or multifidus which would have provided a more indepth analysis of this finding.


Core stability rehabilitation in the thoracic region could potentially require manual therapy intervention as part of the programme to correct biomechanical anomalies. An example would be during active movement assessment of unilateral arm elevation that produced an asymmetrical spinal movement response. This response could be due to an upper thoracic movement dysfunction and may be best managed with manual therapy techniques. Joint mobilisation techniques such as those described by Maitland (1997), Mulligan (1999) and Kaltenborn (1993) cannot be excluded from a core stability programme as abnormal mechanics will affect the neural and active subsystems. Normalisation of physiological and accessory joint movements would appear to be critical in  maintaining ideal timing of muscle activation patterns for muscles acting on the trunk. Such statements are currently largely based on observation and anecdote, direct evidence has not been forthcoming to date however some evidence for this can be found in the literature for manipulative therapy. Keller and Colloca (2000) reported on a clinical trial where they investigated the effect of mechanical force, manually-assisted spinal manipulative therapy on trunk muscle strength pre and post manpulation. They found that muscles produced stronger muscle activation patterns immediately following manipulation of the underlying joints. This study could be improved by extending the length of follow up time to determine the lasting effects of manipulation. Research is necessary to develop a greater understanding for the role of the thoracic spine in the core stability concept. The influence that manual therapy techniques can have on improving muscle activation patterns and thereby core stability also warrants further research.


Active Subsystem

A number of models have been proposed to describe the functional role of the trunk muscles as stabilisers and mobilisers (Bergmark, 1989; Comerford and Mottram, 2000; Goff, 1972; Sahrmann, 2002) The classification schema by Bergmark (1989) divided muscles into two functional categories, the local stabilising and global stabilising systems. The local stabilising muscles were considered to have a primary role in maintaining segmental stability. Many of these muscles have their origin or insertion on the lumbar vertebrae. Panjabi’s (1992a and 1992b) hypothesis of clinical instability highlighted the role of these muscles in maintaining the ‘neutral zone’. The neutral zone has been described as the area of physiological intervertebral motion within which the spinal motion is produced with minimal internal resistance (Gibbons and Comerford, 2001a).


Muscles that make up the local stabilising system are generally considered to be continuously active throughout movement, their activation is not dependent on direction of movement and due to their anatomical close proximity to the centre of rotation of the spinal segments they may also have a proprioceptive role (Gibbons and Comerford, 2001a; Richardson et al., 1999). Panjabi (1992a) suggested they have a role in maintaining the lumbar lordotic curvature.


The muscles that constitute the global stabilising system are multi-segmental, more superficial, generate force to control movement, contractions tend to be eccentric to control motion segments throughout range of movement, activity is direction dependent and muscle activation patterns are phasic (Gibbons and Comerford, 2001b; Richardson et al., 1999).


A third functional classification has been proposed by Comerford and Mottram (2000) which they have labelled the ‘global mobilisers’. The function and characteristics of these muscles is to generate torque to produce range of movement, shock absorption of load, activity is direction dependent and phasic in nature. Muscle contractions are generally concentric in nature therefore produce movement through concentric activity rather than the eccentric control displayed by the global stabiliser muscles. Gibbons and Comerford (2001a) suggested that the global mobility muscles work eccentrically to decelerate high loads and would adopt a stabilising function when under load or when subjected to high-speed movements. Table 7 displays the classification of the local and global muscles of the trunk according to their proposed stability function.


Table 7. Muscle Functional Classification Table. Based on work of Bergmark, 1989; Richardson et al, 1999; Gibbons and Comerford, 2001b and Norris, 2000).

Local Stabiliser Global Stabiliser Global Mobiliser
Transversus Abdominus Oblique Externus Rectus Abdominus
Deep Lumbar Multifidus Spinalis Iliocostalis
Psoas Major (Posterior Fascicles) Gluteus Medius Piriformis
Intertransversarii Quadratus lumborum lateral fibres Hamstrings
Interspinales Oliquus internus abdominis  
Iliocostalis Lumborum pars lumborum    
Quadratus lumborum, medial fibres    
Obliquus internus abdominis (thoracolumbar fascia insertion)    


Further terms that are used to describe the muscular function of the trunk include the ‘inner’ and ‘outer unit’ (Lawrence, 2003; Richardson et al., 1999). The corset or girdle effect created by the deep stabilising muscles that provides a stiffening effect on the spine is also known as the inner unit (Richardson,et al., 1999 and Chek, 2004). The inner unit of musculofascial support is independent  of the outer unit both anatomically and functionally (Richardson et al., 1999). Muscles that are part of the inner unit include the transverse abdominus, multifidus, pelvic floor, diaphragm and posterior fibres of the internal oblique (Richardson et al., 1999; Chek, 2004).


The outer unit has a combined role of stability and movement initiation (Richardson et al., 1999). Movement is the primary function and this takes place in unison with the stabilising role of the inner unit (Richardson et al., 1999 and Lawrence, 2003). The outer unit has been further described as being made up of four subsystems (Lee, 1999). These systems have been labelled the the posterior oblique, the deep longitudinal, the anterior oblique and the lateral system (Vleeming, et al., 1990a and 1990b). The four subsystems are effectively another term to describe the specific roles of the global muscles.


The posterior oblique system of the outer unit is made up of the latissimus dorsi, gluteus maximus and the thoracolumbar fascia. The anterior oblique system includes the external and internal oblique, the contralateral adductors of the thigh and the anterior abdominal fascia.The lateral system includes the gluteus medius and minimis and the contralateral adductors of the thigh. The deep longitudinal system includes the erector spinae, deep lamina of the thoracolumbar fascia, sacrotuberous ligament and the biceps femoris muscle (Vleeming, et al., 1990a and b and 1995; Lee, 1999). Alteration of normal muscle activation patterns or weakness of the inner or outer units would reduce the force closure mechanism of the sacroiliac joint (Vleeming et al., 1990b; Lee, 1999). Lee (1999) proposed that this could result in compensatory movement strategies to accommodate the dysfunction. She suggested that this could culminate in pathology in the lower back, hip and knee. This pathology would most likely be in the form of tissue trauma as a result of the cumulative effect of undesirable forces. This indicates that by maintaining adequate control of these muscle groups and their associated synergies and agonist versus antagonistic relationship that core stability training programmes could either prevent or rehabilitate injuries. The amount that each system contributes to spinal stability is currently still under investigation (Richardson et al., 1999).


The functional unit of local stabilisation (Richardson et al., 1999) or cylinder of stability (Chek, 2004; Hodges et al., 2001) has been used to describe the synergistic actions of the pelvic floor, transversus abdominis, multifidus, and the diaphragm. This conceptual cylinder is made up of the transverse abdominis and multifidus as forming the walls, pelvic floor as the base and diaphragm as the lid. Evidence suggests that with activation of the transverse abdominis muscle there is coactivation of the multifidus, diaphragm and pelvic floor muscles (Avery et al., 2000; Hodges and Richardson, 1996; Hodges et al., 1997; Hodges and Gandevia, 2000; Richardson et al., 1999).


The coactivation of the pelvic floor and diaphragm with the transversus abdominus would act to increase the intra-abdominal pressure and potentially stabilise the spine (Hodges and Gandevia, 2000a; Hodges et al. 2001). The hypothesis here is that this synergistic activity would provide an additional extensor torque to the spine. The pressure in the ‘cylinder’ is optimised if as the abdominal wall is engaged and drawn in causing the pelvic floor to be drawn up a deep breath is taken to force the diaphragm down (Hodges et al., 2001). If the breath is held the intra-abdominal pressure is sustained (Valsalva maneuver). This may be appropriate for heavy loads to prevent bending stress through the lumbar spine (cylinder) (Hodges et al., 1997; Hodges, 1999; Hodges et al., 2001; Norris, 2000). Hodges et al. (2001) found a causal relationship between intra-abdominal pressure and lumbar extensor torque in an in vivo study. Notably this study was performed on a small sample size of 5 males.


The role of intrathoracic pressure in trunkal stability has been much less lauded. Intrathoracic pressure is created during inspiration by expanding the lungs against the rib cage. The role of intrathoracic pressure within the core stability concept appears to have been given very little attention. Norris (2000) considered that the coordination of combined inspiration and abdominal hollowing to increase intra-abdominal and intrathoracic  pressure was unsuitable for most rehabilitation programmes. The effect of timing inspiration with effort could result in the use of the Valsalva maneuver which could potentially raise blood pressure to dangerous levels (Linsenbardt et al., 1992).


The thoracolumbar fascia gain is an anatomical principle that has been compared with erecting a tent where the guy ropes (latissimus dorsi, gluteals, thoracolumbar fascia and internal obliques) act together as a team to support the main structure (lumbar spine) (Lawrence, 2003; Norris, 2002). When the transverse abdominus muscle contracts and draws in the abdominal wall a synergistic relationship with the internal obliques is invoked that causes them to contract and increase tension through the thoraco-lumbar fascia (Hodges, 1999). Evidence for this concept has been provided in studies by Gibson and McCarron (2004) and Vleeming et al. (1995). The spinal extensors are encased in a layer of fascia therefore when they contract they act to stretch the fascia and tension it (Norris, 2002). The combined actions of the latissimus dorsi (pulling fascia upwards), gluteals (pulling fascia downwards) and the transversus abdominus and internal oblique (pulling fascia sideways) is to resist bending forces through the spine especially when lifting (Liebenson, 2004c; Norris, 2002; Vleeming et al., 1995). This coordinated approach to spinal stability requires all muscles to be activated in an appropriate sequence and to have adequate endurance capacity (Elfving et al., 2003; Thomson, 2002). This highlights the need for core stability training programmes to consider coordinated muscle function and not just isolated activity.


The transversus abdominis is the deepest abdominal muscle and has been the subject of intense interest in spinal stability models and research (Hodges, 1999; Richardson et al., 1999; Norris, 2000). The transversus abdominis muscle has its origin on the iliac crest, lower six ribs, and the lateral raphe of the thoracolumbar fascia and inserts medially at the linea alba (Williams et al., 1989). The horizontal fibre orientation results in causing a reduction of the abdominal circumference, increased intra-abdominal pressure and increased tension in the thoracolumbar fascia (Hodges, 1999).


The contributing role of transversus abdominis to spinal stability has been investigated from two view points. Firstly, a mechanical evaluation of the transversus abdominis and its ability to contribute to spinal control, and secondly, the influence of the central nervous system on transversus abdominus during specific tasks (Hodges, 1999). Hodges (1999) critically reviewed the evidence on the recruitment pattern of transversus abdominis. Much of the earliest research on transversus abdominis as a potential contributor to spinal control was performed by Cresswell et al. (1992). In this study they considered the surface electromyographical activity of rectus abdominus, internal oblique and external oblique muscles during isometric trunk extension that caused an increase in intra-abdominal pressure. As there was minimal electromyographic activity in these muscle groups it was proposed that the increase in intra-abdominal pressure was due to transversus abdominis involvement. Transversus abdominis was also found to be the only abdominal muscle continuously active during trunk flexion and extension in both dynamic and isometric situations (Cresswell et al., 1992).


The erector spinae muscle group has been found to behave in a phasic manner during trunk movements (Hodges, 1999; Cresswell et al., 1992). This provokes the question of what contributes to the extensor torque of the lumbar spine or provides the stablity to keep it in the neutral zone whilst the erector spinae is being latent. The increased intra-abdominal pressure as discussed above provides some of the extensor torque to the lumbar spine however studies have demonstrated the role of multifidus and iliopsoas (psoas major and iliacus) as also having major contributory roles in spinal stability (Gibbons et al., 2002; Wilke et al., 1995). Neuromuscular processes appear to link transversus abdominis and multifidus coactivation (Richardson et al., 1999). A number of studies have highlighted the combined action of transversus abdominis and multifidus in maintaining spinal stability (Essendrop, 2002; Hodges et al., 1999).


Physiological evidence of the lumbar multifidus and the lumbar longissimus that is suggestive of their capacity as integral spinal stabilisers has come from studies of their histochemical composition, capillarisation and muscle enzyme activities (Richardson et al., 1999). Research indicates that these muscles have a greater capillary network, concentration of oxidative enzymes and type I fibres which would allow them to perform with a much greater endurance capacity compared with normal skeletal muscle which have a relatively even supply of type I and type II fibres (Jorgensen et al., 1993; Astrand and Rodahl, 1986). This supports the proposed role of these muscles in contributing to a tonic holding and therefore stabilising function of the lumbar spine.


The role of the internal oblique in the local stability mechanism hasn’t been fully explored (Richardson et al., 1999).  There appears to be a stabilising role due to its posterior attachment to the lower lumbar section of the thoracolumbar fascia. It is therefore possible that it may contribute to stability when the transversus abdominis causes tensioning of the thoracolumbar fascia by assisting  stabilising forces through the lower lumbar levels (Richardson et al., 1999).


In this review of the evidence behind the concept of core stability the scapulothoracic joint has been considered within the physical boundaries of the core stability model due primarily to the role that muscle and fascia that attach onto the scapula have in stabilising the trunk. Stability at the scapulothoracic joint is hypothesised to be due to a co-contraction between serratus anterior and the trapezius muscles (Mottram, 1997). It is thought that they have a similar stabilising role as transverse abdominus as muscles that are preactivated before upper limb movement (Mottram, 1997). This review found no direct evidence to prove this point however new research is currently investigating this feature.


The latissimus dorsi originates from the sacral and iliac crests, thoracolumbar fascia, spinous processes of T7-T12 and the posterior surface of the lower ribs (Calais-Germain, 1993; Williams et al., 1999) . The muscle inserts on the bicipital groove of the humerus however also has a slip to the inferior scapula angle so therefore may influence scapula position (Calais-Germain, 1993; Mottram, 1997). Conversely, it can therefore be seen from the above attachment points that scapula position could influence lines of tension along the thoracolumbar fascia and therefore trunkal stability. Other muscles that should also be considered when assessing scapula alignment and function in relation to core stability are the rhomboids, subclavius, pectoralis minor and levator scapulae (Calais-Germain, 1993; Kibler, 1995). While not having attachments to the spinal column or rib cage the rotator cuff muscle group could also have some influence on scapula alignment and therefore secondarily on core stability (Horsley, 2005). The kinetic chain between the shoulder girdle and thoracic spine has been considered by a number of authors who acknowledged the influence of muscle imbalance and thoracic spine mobility on scapula control (Kebaetse et al., 1999; Lee, 1995; Janda, 1991). No direct evidence was found to support the influence of scapula position or control on core stability function.


Another possible consideration would be to consider the scapulothoracic joint as part of a second cell of peripheral core stability units. Muscle stabilising principles similar to the lumbar spine could be applied to the shoulder, elbow, hip, knee and ankle joints so that potentially these areas could be viewed as having their own unique core stability features. This potentially controversial view could provide geater continuity in describing the cause and effect of movement dysfunction throughout the body.


Control and Feedback Subsystem


McGill et al. (2003) acknowledged the importance of the sensory organs when considering the motor control response to sudden spine loading. They reported that the proprioceptive feedback gained from a variety of proprioceptors including muscle spindles, Golgi tendon organs, joint receptors and cutaneous receptors was an important drive to directing motor responses to attempted perturbations of the spine. Conversely, it has been well established from back pain research that subjects with lumbopelvic dysfunction have impaired muscle activation patterns, poor postural contol and a diminished concept of a neutal pelvic position (Radebold et al., 2000).


Panjabi (1992a) expressed his opinion that while there was no published evidence that spinal stability could be enhanced by improving neural control alone he felt the possibility existed. This current review found a number of studies that intimated that neural adaptations had occurred through low level training of the deep stabilising muscles (Davidson and Hubley-Kozey, 2005; Richardson et al., 1990; Hagins et al., 1999; O’Sullivan et al., 1997a). Muscles were trained at levels below that required to cause hypertrophy and strength gains however were able to be activated at higher intensities (measured by % maximum voluntary contraction) following training programmes.  It has been suggested that this is the direct effect of enhancing neural pathways (Davidson and Hubley-Kozey, 2005; McGill et al., 2003).


A decade on from Panjabi’s (1992a) hypothesis came the work by Holm et al. (2002) who investigated the role of viscoelastic structures with special focus on their sensory motor functions. The intervertebral disc, fibrous joint capsule and spinal ligaments are included in the passive subsystem for their physical properties however their ability to monitor sensory information warrants their inclusion in the neural and feedback system. The background hypothesis for Holm et al. (2002) was that lesions in avascular supporting structures could cause perturbations in the proprioceptive function of various receptors and result in increased and prolonged muscle activation that may cause pain. They further suggested that subsequent irritation of low threshold nerve endings in the sacroiliac joint, intervertebral disc and zygapophyseal joint tissue may trigger a reflex activation of the gluteal and paraspinal muscles over time that may provoke further pain.


Another consideration here would be the impact of a ‘double crush’ like syndrome where a peripheral nerve pathway could be compressed twice e.g. centrally due to derangement of an intervertebral disc and peripherally due to tonic muscle spasm. In this scenario the nerve root irritation due to the disc compression would inflame the neural tissue and provoke muscle spasm along the course of the nerve root and therefore further compress the peripheral nerves potentially irritating them locally. This could then setup a chronic pain cycle. While there is a potential degree of overactivity due to the neural irritation there could also be associated inhibition of some muscle groups (Arvidsson et al., 1986). As the concept of core stability appears to rely on a coordinated pattern of muscle activation it would appear that aberations in the passive and neural subsystems would undermine core stability function.


Holm et al. (2002) used 80 domestic pigs during their investigation of the neuromuscular interaction between the spinal structures. They found that reflexive contraction of the multifidus and also longissimus muscles resulted following electrical stimulation of the lumbar afferents in the discs, capsules and ligaments. The muscular activity was greatest at the level of excitation and found to have weaker excitation at 1-2 levels above and below (Holm et al., 2002). This finding is suggestive that elements of the passsive subsystem that help to maintan the neutral zone as described by Panjabi (1992b) also have a role in monitoring sensory information and controlling spinal muscles especially the multifidus which has been reported to have a local stabilising role in a number of studies (Hides et al., 1996; Jemmett et al., 2004 and Moseley et al., 2002). They also found that the multifidus and longissimus muscles had increased levels of activation due to mechanical stimulation of the viscoelastic structures. Electromyographic recordings were greater if more than one viscoelastic structure was stimulated (Holm et al., 2002). Holm et al. (2002) also suggested that the viscoelastic structures provided kinesthetic perception to the sensory cortex.


Panjabi (1992b) hypothesised that if any one or more of the subsystems weren’t functioning optimally then overall spinal stability could be compromised. This could equate to influencing the core stabilising system to the same extent. Dysfunction of the passive subsystem could be the result of mechanical injury, development of microfractures in the endplates, disc protrusion or degeneration (Panjabi, 1992a). These disturbances in the passive subsystem could result in compensatory changes in the active subsystem due to the decreased load bearing and stability capacity of the passive subsystem. The alteration in muscle activation patterns would influence the core stabilising capacity. Any disruption to the neural components serving the active musculoskeletal subsystem could affect the muscle activation patterns and influence feedback of muscle tension to the neural system (Arvidsson et al., 1986; Panjabi, 1992a). It has been suggested by Panjabi (1992a) that perturbation of the active subsystem could influence the ability to provide compensatory muscle activation for stresses that challenge the passive subsystem and also erode the capacity to maintain trunk stability when subjected to unexpected multidirectional forces or heavy loading (Panjabi, 1992a). In this hypothesis it can be seen that trauma, disuse, degeneration and disease processes that affect the active subsystem would influence core stability directly. While there appears to be little direct evidence for this, there is some evidence to suggest that these noxious factors can influence muscle function and therefore potentiate instability of the trunk (Hides et al., 1994 and 1996; Silfies et al., 2005).


Dysfunction to the subsystems as described above would have far reaching effects on core stability rehabilitation programmes as these variables would need to be optimised or reduced below a threshold that could cause significant influence to the coordinated stabilising function of the subsystems. The majority of core stabilising rehabilitation programmes have an emphasis on physical training however the above discussion on reasons for dysfunction to the subsystems highlights the need to consider treatment of pathological states such as disc prolapse or ligament disruption with surgical and/or physiotherapy intervention. There is evidence to support both these modalities in the treatment of musculoskeletal disorders (Gibson, et al., 2005; Moseley et al., 2002).


In terms of the anatomical evidence for core stability much of the traditional anatomical work has been purely observational. Based on the observations and descriptions of structures researchers have hypothesised how they function in combination with each other. In this field there are very few randomised controlled trials so when developing guidelines based on anatomical evidence it would be very unlikely to ever be able to award the highest grading using the current criteria of the SIGN 50 (2004) protocol. There is a large proportion of observational and case series studies and from these expert opinions have been proposed (see Table 8). There has been some work on comparing biomechanical models with in vivo subjects and this has created a framework for predicting perturbations in the living model (Gardner-Morse and Stokes, 1998; Granata and Orishimo, 2001; Quint et al., 1998; Vleeming et al., 1995).


When comparing and contrasting observational studies in functional anatomy with the findings from biomechanical studies and biomechanical modelling, concepts and understandings begin to crystalise as findings from different sources assimilate. The continuity of findings allows concept development to gain momentum and credibility. There is a need for further studies in functional anatomy to allow the understanding of the core stability concept to progress.


The anatomical, physiological and biomechanical considerations behind the core stability concept are clearly complex. Future research will need to be well structured to provide further evidence especially in the areas of neural systems and synergistic muscle function and coactivation sequences. To see comprehensive evidence tables for the studies considered in this research question see Table 14 in Appendix C.


Table 8. Intervention studies for Question 1: What anatomical, biomechanical and physiological evidence exists that supports the concept of core stability?

Bibliographic citation Study type Ev Lev General comments
AVERY, A. F., O'SULLIVAN, P. B. & MCCALLUM, M. J. (2000) Evidence of pelvic floor muscle dysfunction in subjects with chronic sacro-iliac joint pain syndrome. Proceedings of the 7th Scientific Conference of the IFOMT in conjunction with the MPAA, 35-38.



Case-control 2- Investigation of the role of the pelvic floor muscles in providing SIJ stability as part of the theoretical ‘force closure’ system. Stability of the SIJ is hypothesised to be a result of its bony configuration (form closure) and the local muscle forces acting on the ligaments and fascia to compress the joints (force closure). The authors suggested that their findings support the notion that dysfunction of the pelvic floor muscles may disrupt the force closure mechanism of the SIJ in individuals with chronic SIJ pain syndrome. This study provided some anatomical, biomechanical and physiological evidence to support the concept of core stability. The pelvic floor is hypothesised to provide the base of the ‘cylinder’ or box in many core stability models. Disruption or deficiencies with this component would theoretically compromise the system. Limitations of this study must be considered including: small sample size, weak statistical methods, testing method of pelvic floor muscle activity/validity/specificity).
CHOLEWICKI, J., JULURU, K., RADEBOLD, A., PANJABI, M. M. & MCGILL, S. M. (1999) Lumbar spine stability can be augmented with an abdominal belt and/or increased intra-abdominal pressure. Eur Spine J, 8, 388-95. Cohort study 2- Both wearing an abdominal belt and increasing IAP can increase spinal stability. Of note was that the belt resulted in a decrease in activation of the thoracic and lumbar erector spinae muscles. The authors suggested that concomitant increase in IAP was due to muscle coactivation. Limitations of this study included a small sample size.
CHOLEWICKI, J., PANJABI, M. M. & KHAKHATRYAN, A. (1997) Stabilizing function of trunk flexor-extensor muscles around a neutral spine posture. Spine, 22, 2207-2212. Cohort 2+ This study demonstrated that antagonistic muscle coactivation patterns were present around the neutral spine posture. Coactivation increased with added mass to the torso. The importance of neuromuscular control was supported by the inverted pendulum biomechanical model. Researchers noticed flexor and extensor activity was very similar for all subjects however EMG tracings revealed some individual differences between subjects most notably with multifidus and the internal oblique muscles. These muscles appeared to exhibit a constant level of activation regardless of trunk angle. The implication here is that individuals used different muscle recruitment patterns to stabilise the lumbar spine.
GARDNER-MORSE, M. G. & STOKES, I. A. (1998) The effects of abdominal muscle coactivation on lumbar spine stability. Spine, 23, 86-91; discussion 91-2.


Exp. Biomech. Model design. 3 The biomechanical model suggested that the probable role of antagonistic abdominal muscle coactivation was to stabilise the spine. The authors suggested that in vivo the neuromuscular system must balance spinal stability vs trunk muscle fatigue vs spine compression. Results from this study need to be considered carefully as the simulated magnitudes of the abdominal coactivations may represent a suboptimal strategy for spinal stability invivo.
GIBSON, J. & McCARRON, T. (2004) Feedforward muscle activity: an investigation into the onset and activity of Internal oblique during two functional reaching tasks. Journal of Bodywork and Movement Therapies, 8, 104-113. Case series 3 This study demonstrated that the internal oblique muscle activity feeds-forward in anticipation of functional task performance. This suggests that prior to prime mover activity (deltoid) active stabilisation of intersegmental joints is thought to occur prior to arm displacement. This supports core stability theories that purport that by creating greater stiffness of the lumbar motion segment trunk movement may occur from a stable base. Supports Hodges (1999) study for transversus abdominus. This study indicated that a peak of activity occurred close to the onset of initial movement and was followed by a consistent lower level of activity until the arm came to rest. The authors suggested that this implies that a certain threshold of activity needed to be achieved and maintained in order to maintain a critical level of stiffness at the neutral zones of the spine to prevent strain of spinal structures during upper limb movement. From a training perspective this study implies that endurance training for the internal oblique muscle (type II fibres) is necessary so it can maintain spinal (stiffness) integrity during movement tasks of the upper limb. This study also supported the concept of local stabilising muscles and the need for them to provide spinal stability. Inhibition of these muscles in conditions such as low back pain that cause changes in the neuromuscular system would potentially expose the spine to excessive forces of translation and inefficiency along the kinetic chain. This study protocol needs to be repeated with a controlled matched subject trial with normals and a pathology group (e.g. chronic low back pain).
HIDES, J. A., STOKES, M. J., SAIDE, M., JULL, G. A. & COOPER, D. H. (1994) Evidence of multifidus muscle wasting ipsilateral to symptoms in patients with acuter/subacute low back pain. Spine, 19, 165-172.


Case -control 2+ Unilateral wasting of the multifidus muscle on the symptomatic side could be detected by real-time ultrasound. At the symptomatic vertebral level the degree of multifidus muscle cross sectional area difference was greatest. The reason for this was hypothesised to be due to inhibition from perceived pain via a long-loop reflex pathway. There was also a reported lack of correlation between severity of wasting and clinical findings. Some subjects with relatively minor clinical symptoms had significant wasting. Notably, some subjects in the ‘normal’ population that demonstrated asymmetry were symptom free and this may relate to sports that involve unilateral use of muscles. It could also potentially provide a window into the future as a measurable precursor to symptoms. This would need further investigation. This study provided some evidence on the intersegmental role of multifidus and an apparent link between atrophy and spinal dysfunction. Possibly the lack of stability at a specific spinal segment contributes to the pain pathology. The lack of spinal stiffness that multifidus could provide at that level would create difficulty for the spine to maintain motion in the neutral zone. This concept requires further consideration and research.
HIDES, J. A., RICHARDSON, C. A. & JULL, G. A. (1996) Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine, 21, 2763-2769. RCT 1+ Multifidus muscle recovery didn’t correlate with remission of painful symptoms. The authors proposed that reflex inhibition was the likely cause rather than via a long, loop reflex. However the patients in the exercise group made a more rapid recovery from a muscle physiology perspective. At the 10 week follow up group 1 patients still had decreased multifidus muscle size even though they had returned to normal activities of daily living. It would appear that early local stabilisation exercises can accelerate the return of more optimal spinal mechanics and therefore decrease the frequency of recurrent episodes of low back pain. Further research is required to determine optimal exercise programmes including exercise type, frequency and duration of programme.
HODGES, P., GANDEVIA, S. C. & RICHARDSON, C. A. (1997) Contractions of specific abdominal muscles in postural tasks are affected by respiratory maneuvers. Journal of Applied Physiology, 83, 753-760.




Case series 3 Transversus abdominus has an important role in respiration as well as spinal stability. Transversus abdominus normally only contributes to respiration during forced expiration or involuntarily by breathing in against an inspiratory load. This study suggested that transverses abdominus can contribute equally well to spinal stability in any respiratory phase. The authors suggested neural involvment in determining degree of abdominal muscle activity during postural perturbations to increase IAP.
KAIGLE, A. M., HOLM, S. & HANSSON, T. H. (1995) Experimental instability of the lumbar spine. Spine, 20, 421-430.


In vivo animal model of lumbar spine.



3 The purpose of this study was to evaluate the contribution from various spinal components, including muscular involvement of the lumbar motion segment, to the overall stability of the spine. They found that in the neutral region, where the muscles are under reduced tension, the motion segment is highly susceptible to instability, resting musculature influences spinal mechanics and increased muscular activity in the injured motion segment has a stabilising effect by reducing the abrupt patterns of motion in the neutral zone. This study although performed on pigs provided considerable evidence as to the stabilising role of some key muscle groups: erector spinae, multifidus, quadratus lumborum and psoas. While great care must be taken with extrapolating these results to the human spine it did provide a valuable insight into the effects that degenerative changes and surgical techniques can have on spinal mechanics. It would be unlikely for this study to be replicated on a human population in its current format however a revised version may gain ethical approval and could potentially provide conclusive evidence on muscle function.
LEINONEN, V., KANKAANPAA, M., AIRAKSINEN, O. & HANNINEN, O. (2000) Back and hip extensor activities during trunk flexion/extension: Effects of low back pain and rehabilitation. Archives of Physical Medicine and Rehabilitation, 81, 32-37.


Cross-sectional design.



2- The rehabilitation focused on improving back muscle strength, mobility, coordination and muscle elasticity. Exercises began at a low intensity and were gradually increased. Chronic back pain group demonstrated reduced gluteus maximus activity during the flexion-extension cycle. This study found no significant difference between groups for paraspinal muscle activity. Recommendation is to repeat study design but include investigation of hip and spinal extensor activity during spinal flexion and extension.
GRANATA, K. P. & ORISHIMO, K. F. (2001) Response of trunk muscle coactivation to changes in spinal stability. J Biomech, 34, 1117-23.





Two dimensional biomechanical model 3 Experimental data supported the test hypothesis that muscle activation increased when external loads were held at greater heights. Supporting the concept that the neuromuscular system responds to changes in stability.
KUMAR, S. (1997) The effect of sustained spinal load on intra-abdominal pressure and EMG characteristics of trunk muscles. Ergonomics, 40, 1312-34.



Case series 3 This study hypothesised that intra-abdominal pressure would correspond to the magnitude of the load applied. However this study showed the IAP to be inconsistent.The study concluded that intra-abdominal pressure is not an active spine load-relieving mechanism. This study should be compared with Hodges work. The results from this study are contrary to most others however this study investigated prolonged loading whereas most others considered transient loading.
NG, J. K., PARNIANPOUR, M., RICHARDSON, C. A. & KIPPERS, V. (2001) Functional roles of abdominal and back muscles during isometric axial rotation of the trunk. J Orthop Res, 19, 463-71. Case series 3 Different functional roles of trunk muscles during axial rotation were demonstrated in this study.  The more lateral and oblique the muscle fibres, the more these muscles contribute to movement force, but in the main part, only on the side ipsilateral to the movement.  The multifidus and rectus abdominis muscles act bilaterally during unilateral rotation movements indicating their role is more of a stabilising one.  The activity of multifidus was greater than its synergist, rectus abdominis, adding further weight to the body of evidence that points to its role as a local stabiliser of the lumbar spine, not just during sagittal plane exertions but in axial rotation exertions as well.
ESSENDROP, M., ANDERSEN, T. B. & SCHIBYE, B. (2002) Increase in spinal stability obtained at levels of intra-abdominal pressure and back muscle activity realistic to work situations. Appl Ergon, 33, 471-6.


Case series 3 Results of this study indicate that both abdominal and back muscles may have a role to play in stabilising the spine. Indication of involvement of trunk muscles and increased IAP levels to sudden spinal loading.
BROWN, S. H., HAUMANN, M. L. & POTVIN, J. R. (2003) The responses of leg and trunk muscles to sudden unloading of the hands: implications for balance and spine stability. Clin Biomech (Bristol, Avon), 18, 812-20.


Case-series 3 (Also evidence for question 6) This study demonstrated that in situations of sudden unloading , knowledge of the timing of the unloading may lead to anticipatory actions of postural muscles which actually decreases spinal stability, thereby increasing the risk of injury should an unexpected perturbation occur. The authors suggested that both loading and unloading of the spine can create relative instability. The magnitudes of the decreases were observed to become less pronounced as knowledge of the perturbation timing decreased. The random unloading condition demonstrated the importance of cocontractions about the spine as the posterior muscles showed a significant increase in activation and anterior muscles a slight increase. It would appear that anticipatory reactions attempt to keep compressive loads through the spine to a minimum. This study needs to be extended to look at the effects of pathological states and gender on anticipatory reactions. It would appear that unloading exercises should be performed as part of a core stability programme.
NG, J. K., RICHARDSON, C. A., PARNIANPOUR, M. & KIPPERS, V. (2002) EMG activity of trunk muscles and torque output during isometric axial rotation exertion: a comparison between back pain patients and matched controls. J Orthop Res, 20, 112-21. Case-control 2+ Authors expressed that training coordination and multiplanar movements as well as strength during rehabilitation is vital.  The authors hypothesised that the reduced activity of multifidus in the back pain group may indicate that spinal stability was compromised.
CHIANG, J. & POTVIN, J. R. (2001) The in vivo dynamic response of the human spine to rapid lateral bend perturbation: effects of preload and step input magnitude. Spine, 26, 1457-64. Repeated measures design 3 The authors reported that higher levels of trunk muscle preactivation increase spine stability. Higher loads increase activation of agonistic and antagonistic muscle activation. Larger multisegmental muscles are important in the maintenance of spine stability. Supports the concept that cocontraction increases stability of the spine. Supportive of the work by Cholewicki.
ESSENDROP, M. & SCHIBYE, B. (2004) Intra-abdominal pressure and activation of abdominal muscles in highly trained participants during sudden heavy trunk loadings. Spine, 29, 2445-51.






Case series 3 IAP rose quickly with the loading effect. This reinforced the notion that physiological components that result in increased IAP have an important role in maintaining and returning the trunk to neutral. Men tended to develop higher IAP and higher torques than women when the trunk was suddenly heavily loaded. The authors described an ‘internal air bag’ effect with increasing IAP. Rectus abdominus was found to have minimal activity in IAP production therefore confirming its role more as a mobilising rather than stabilising muscle. In terms of core stability this paper described how the ability to generate higher IAP is crucial to stabilising the spine from perturbations. This study needs to be repeated with non trained individuals to determine the effect of loading an untrained person with sudden heavy loads.
STOKES, I. A., GARDNER-MORSE, M., HENRY, S. M. & BADGER, G. J. (2000) Decrease in trunk muscular response to perturbation with preactivation of lumbar spinal musculature. Spine, 25, 1957-64. Case series 3 This study supported the theory that preactivated muscles can stabilise the loaded spine. The authors concluded that preactivation of the muscles increased spinal stiffness therefore helped to maintain the spinal system in equilibrium. Evidence for anatomical and biomechanical concept of trunk core stability.
Van DIEEN, J. H. (1996) Asymmetry of erector spinae muscle activity in twisted postures and consistency of muscle activation patterns across subjects. Spine, 21, 2651-2661.


Repeated measures design 3 Notably 2 subjects demonstrated significantly different muscle activity from the general pattern. Possibly this could represent an underlying pathology and a precursor for back pain. This is worthy of further investigation. Some evidence for the effects of fatigue and spinal stability. Twisting was found to cause asymmetry in back muscle activation resulting in stress concentrations in spinal motion segments. The distribution model was able to demonstrate the effects of asymmetrical loading on the spine.
HODGES, P. W. & RICHARDSON, C. A. (1997) Relationship between limb movement speed and associated contraction of the trunk muscles. Ergonomics, 40, 1220-30.


Case series 3 This study was one of the earlier works that suggested the transverse abdominus muscle had an important role in stabilising the trunk in anticipation of upper limb movement. This study provided initial normative data for preprogrammed anticipatory muscle activity. This knowledge can be incorporated into the assessment of and treatment of stabilising muscles. Future studies should investigate the responses of subjects who have pathology especially low back pain. The EMG activity was recorded via fine wire as well as surface electrodes thus providing a more reliable picture of the true functioning of the deeper abdominal muscles.
HODGES, P., CRESSWELL, A. & THORSTENSSON, A. (1999) Preparatory trunk motion accompanies rapid upper limb movement. Exp Brain Res, 124, 69-79.


Case series 3 Transverse abdominus and internal oblique EMG timing did not vary between directions of bilateral shoulder movement. This non-specific activation pattern supports the non-direction-specific response of the deep abdominal muscles.
JEMMETT, R. S., MACDONALD, D. A. & AGUR, A. M. (2004) Anatomical relationships between selected segmental muscles of the lumbar spine in the context of multi-planar segmental motion: a preliminary investigation. Man Ther, 9, 203-10. In-vitro study 3 Dissection of the lumbar spine demonstrated segmental attachment patterns for transversus abdominis, psoas, quadratus lumborum and multifidus. Confirmed that muscle group surrounds the motion segments from the anterolateral aspect of a vertebral body to the spinous process. This study can only receive a low evidence rating due to small sample size and often significant variations in individual anatomy.
Van DIEEN, J. H., KINGMA, I. & VAN DER BUG, P. (2003) Evidence for a role of antagonistic cocontraction in controlling trunk stiffness during lifting. J Biomech, 36, 1829-36.


Case series 3 Study demonstrated greater coactivation with unstable load. This study demonstrated that antagonistic coactivation increases trunk stiffness during lifting tasks. Increasing spinal stiffnessmay also improve kinematic control of the spine. Repeating this study with back pain patients may provide some further information on compenation strategies.
STOKES, I. A. & GARDNER-MORSE, M. (2001) Lumbar spinal muscle activation synergies predicted by multi-criteria cost function. J Biomech, 34, 733-40.


Comparison of experimental data with a predictive model.



3 This study described the challenges of predictive analytical models. The authors found the role of the central nervous system to be difficult to account for in this model. Authors concluded that muscle activation strategies efficiently limit intervertebral forces and displacements.
HODGES, P. W., CRESSWELL, A. G., DAGGFELDT, K. & THORSTENSSON, A. (2001) In vivo measurement of the effect of intra-abdominal pressure on the human spine. J Biomech, 34, 347-53.


Case series

In vivo study

3 This study provided evidence that IAP contributed to spinal stability. The in vivo results correlated with the biomechanical model. The role of the diaphragm in contributing to the roof of the cylinder in the core stability model can be deduced from this study.
MARRAS, W. S., DAVIS, K. G. & JORGENSEN, M. (2002) Spine loading as a function of gender. Spine, 27, 2514-2520. Cohort 2+ During whole-body free-dynamic condition women experienced greater relative loads. Men experienced greater absolute spine loading due to body mass and women had greater relative spine loading due to kinematic compensations. Authors concluded that women were at greater risk of injury when considering spine tolerances. This is a relevant finding for rehabilitation programmes.
MOSELEY, G. L., HODGES, P. W. & GANDEVIA, S. C. (2002) Deep and superficial fibres of the lumbar multifidus muscle are diferentially active during voluntary arm movements. Spine, 15, E29-E36.


Cross-sectional 3 The authors concluded that the deep and superficial fibres of the multifidus are differentially active during single and repetitive movements of the arm. They also concluded that the superficial fibres of multifidus contributed to the control of spine orientation and that the deep multifidus has a role in controlling intersegmental motion.
QUINT, U., WILKE, H., SHIRAZI-ADL, A., PARNIANPOUR, M., LOER, F. & CLAES, L. E. (1998) Importance of the intersegmental trunk muscles for the stability of the lumbar spine. Spine, 23, 1937-1945. In vitro biomechanical design 3 The results of this study extend those of an earlier study by Wilke et al (1995).  They demonstrate the importance of the neural control strategy in increasing L4-L5 motion segment stiffness when subjected to axial torque and lateral bending moments.  Whilst the method of torque application in this study was more sophisticated than in the Wilke et al (1995) study, there are still several limitations when extrapolating results from in vitro to in vivo conditions.
RADEBOLD, A., CHOLEWICKI, J., PANJABI, M. M. & PATEL, T. C. (2000) Muscle response pattern to sudden trunk loading in healthy individuals and in patients with chronic low back pain. Spine, 25, 947-954.


Case control study 2+ Subjects with low back pain demonstrated a very different muscle response to a sudden trunk load in a range of spinal postures.  Patients had longer reaction times compared to their matched controls and, unlike the control group, demonstrated a pattern of agonist-antagonist co-contraction.  This may predispose to low back pain and / or be activated to enhance spinal stiffness.  The endurance capacity of the mucles involved in  this co-contraction pattern was not examined so this study does not provide clues as to whether this pattern is a successful compensation.
SAPSFORD, R. R. & HODGES, P. (2001) Contraction of the pelvic floor muscles during abdominal maneuvers. Archives of Physical Medicine and Rehabilitation, 82, 1081-1087.


Case series 3 This study was conducted in healthy subjects with the aim of determining whether there is a link between voluntary abdominal muscle contraction and pelvic floor muscle activity.  The data shows a positive link between the two and the increase in pelvic floor muscle activity before any increase in abdominal muscle activity suggests that the pelvic floor has a preprogrammed anticipatory function.  Dysfunction of the pelvic floor muscles can lead to urinary and fecal incontinence.  Whilst the effect of a combined abdominal / pelvic floor exercise programme was not studied, the results suggest that abdominal muscle training may be beneficial in restoring pelvic floor function.  This is of interest as the pelvic floor provides the inferior sling aspect of lumbo-pelvic, or “core” stability.


SILFIES, S. P., SQUILLANTE, D., MAURER, P., WESTCOTT, S. & KARDUNA, A. R. (2005) Trunk muscle recruitment patterns in specific chronic low back pain populations. Clinical Biomechanics, 20, 465-473. Non-randomised case-control 2- The chronic low back pain population in this study demonstrated altered muscle activation strategy when trying to enhance trunk stiffness.  This increase in overall trunk muscle activity was driven by increases in multisegmental abdominal muscles rather than the unisegmental multifidis muscles.  Data from this study suggested that it is not damage to the passive subsystem that drives this alteration in muscle activity and that the number of segments degeneratively damaged does not influence the aberrant muscle activation pattern.
SOLOMONOW, M., ZHOU, B., HARRIS, M., LU, Y. & BARATTA, R. V. (1998) The ligamento-muscular stabilizing system of the spine. Spine, 23, 2552-2562.




Experimental case series combined human and animal experiment. 3 This study demonstrated the direct link between multifidis and the supraspinous ligament in both human patients and cats.  The authors postulate that damage to the supraspinous ligament can lead to multifidis overactivity and possibly pain.  No pain outcome measures were included in this study and the human population size was limited to two, making it only marginally better than a case study.  More work needs to be done in this area with a greater population sample and pain assessment tools in order for these claims to be substantiated.
VLEEMING, A., POOL-GOUDZWAARD, A. L., STOECKART, R., VAN WINGERDEN, J. & SNIJDERS, C. A. (1995) The posterior layer of the thorocolumbar fascia. Its function in load transfer from spine to legs. Spine, 20, 753-758. Experimental in vitro case series 3 The thoracolumbar fascia provides the mechanism for an integrated force transfer system between the spine, pelvis, legs and arms.  Some muscles that attach to the thoracolumbar fascia (notably gluteus maximus and latissimus dorsi) can exert their influence contralaterally.  Hence, these muscles can provide a track for force transfer between pelvis and trunk.  In this manner, the thoracolumbar fascia provides plays a vital role in the integration of tunk rotation and in load transfer, thus augmenting the stiffness of the lumbopelvic region and theoretically in enhancing the force closure component of the sacroiliac joint.  Potentially therefore, it can be assumed that strengthening those muscles with direct attachments to the thoracolumbar fascia may increase the force closure forces.
WILKE, H., WOLF, S., CLAES, L. E., ARAND, M. & WIESEND, A. (1995) Stability increase of the lumbar spine with different muscle groups. Spine, 20, 192-198. In vitro study 3 This experiment showed the strong influence of muscle forces, particularly multifidis, on stiffening intervertebral motion segments (that is, reducing the neutral zone and ROM).  In contrast to Panjabi et al (1989) who reported increased flexion angles during similar muscle stimulation, this experiment showed a significant reduction in flexion angle (65%).  The major limitation of the study was that the muscle forces simulated were transferred through a small, discreet site of attachment onto the vertebrae, rather than the broad attachments seen in vivo.  It does however provide further weight to the concept of deep muscle activation reducing neutral zone movement with associated spinal movement.