Specialising in musculoskeletal, orthopaedic, spinal and sports rehabilitation

To highlight the problem with the popular use of the term core stability one just has to look at its most popular application, training. Core stability training is often used to describe strengthening (overload or high threshold training) of the proximal trunk muscles (Gibbons and Comerford, 2001). The net effect here is potentially a cocontraction of all regional muscles (local and global muscles). Gibbons and Comerford (2001a) suggested that it may not be appropriate to extrapolate the research on low threshold movement dysfunction and training of the local stability muscle system with strength training of other muscle groups (global muscles). The indiscriminate and haphazard use of the terms strength and stability is a consistent problem with the core stability concept Gibbons and Comerford, 2001a and 2001b).


On review of the evidence concerning specific exercise programmes focused on trunk musculature the majority followed guidelines of segmental stabilisation as described by Richardson et al. (1990 and 1999) (Davidson and Hubley-Kozey, 2005; Hubley-Kozey and Vezina, 2002; Marshall et al., 2005; Souza et al., 2001). Some programmes were more general in approach to exercising the trunk musculature and applied only some of the principles such as a neutral pelvic position (Arokoski et al, 2004, Haynes, 2004; Koumantakis et al., 2005). The majority of the related research is derived from studies in exercise and back pain (Arokoski et al., 2004; Nelson et al., 1999; O’Sullivan et al., 1997; Richardson et al., 2002). The literature consistently described some desirable traits that reportedly constituted the foundations of good core stability such as low threshold spinal stabilisation exercises performed with the pelvis in a neutral position (Bardin, 2003;  Hagins et al, 1999).


The established elements of what constitutes a physical training programme include training intensity, duration, frequency and exercise type (Astrand and Rodahl, 1986). One could assume that core stability training could fit into these parameters however on review of the available evidence it would appear that the traditional strength training guidelines do not apply directly to core stability training. Arguably the physiological adaptations required to enhance core stability muscle function are as specific as their functional roles (Boyle, 2004; Liebenson, 2003). This evidence based review has highlighted that training protocols described for strength and power training should not be directly correlated with core stability training. Overload training programmes tend to recommend a training intensity of between 60-80% of 1 repetition maximum (RM) (Astrand and Rodahl, 1986; Campos et al., 2002) A number of studies demonstrated improved function of the core stabilising muscles at training intensities less than 40% of a maximal voluntary contraction (Davidson and Hubley-Kozey, 2005; Hubley-Kozey and Vezina, 2002; Sung, 2003).  This would indicate that factors other than muscle fibre hypertrophy have contributed to enhanced muscle control such as neuromuscular adaptations (Sung, 2003). This lends support to the programmes that encourage low intensity exercises of the deep stabilising muscles (Richardson et al., 1999 and Stott, 2002).


Campos et al. (2002) evaluated the effect of three resistance-training programmes (low repetition 3-5 RM, intermediate repetition 9-11 RM and high repetition 20-28 RM) on the physiological responses in men. The high repetition group (2 sets of 20-28 RM) were the only group to show improvements in endurance capacity, aerobic power and time to exhaustion. The low repetition group (4 sets of 3-5RM) demonstrated maximal strength gains. The target muscle group in this investigation was the quadriceps muscle. The methodology of this study should be repeated specifically on muscles that make up the core stabilisers. In its current form the results of this study provide an indication that endurance training requires high repetition rates.


While the benefits of low intensity stabilising exercises have been heralded in core stability exercise programmes for the past decade there appears to have been a lack of investigation into the effects of higher intensity exercise after the initial muscle activation patterns have been enhanced through low grade exercise. The natural progression following achieving enhanced muscle activation patterns during low intensity stabilising exercises has tended to be functional training. Functional training has been described as, “a continuum of exercises that teach athletes to handle their own body weight in all planes of movement” (Boyle, 2004). Essentially functional training involves maintaning the principles of core stability training and reproducing these concepts into task specific exercises. The specificity of training principle is well established and this progression makes perfect empirical sense however this review failed to find intervention studies on this topic. Interestingly no studies were found that considered the effects of progressing to higher intensity core stability training following achieving adequate standards of basic, low intensity exercise. Perhaps this progression is unnecessary for everyday life and normal activities of daily living though this level of conditioning could potentially influence performance at elite exercise levels. There may also be a role for this progression in high velocity , high impact sports such as rugby where perturbations to spinal alignment are caused by large forces. This is an area of potential future research.


The effect of traditional strength training on spinal stability has been poorly reported (Linton and van Tulder, 2001; van Tulder et al., 2000). Studies that have considered the effects of strength training about the spine have tended to use pain as the outcome measure rather than stability. The holistic benefit of an integrated strength and stability exercise programme on an individual’s pathology and performance could prove to be a very beneficial combination that warrants further scientific investigation (Boyle, 2004).


Maintaining the neutral lumbar spine position with low intensity exercises appears to be  important from a specificity of training point of view (McGill et al., 2003; Richardson et al., 1990 and 1999; Stott, 2002). Classic abdominal curl exercises were originally performed with the pelvis placed in a posteriorly tilted position (Stott, 2002). This not only resulted in increased intradiscal pressure but also recruited the global muscles such as rectus abdominus greater than the deeper stabilising muscles like transversus abdominus (Norris, 2000). Of note many of the early Contrology exercises described by Joseph Pilates (1945) were originally performed with the pelvis tilted in a posterior direction. The influence of recent researchers like Essendrop et al. (2002), Hides et al. (2001), Hodges and Gandevia (2000), Panjabi (1992a) and Richardson et al. (1999),  has resulted in a change in the method of Pilates teaching to incorporate the neutral spine. This has bourne a whole new genre of exercise known as Clinical Pilates. The proponents of this branch of exercise have been proactive in implementing new research and findings in the area of segmental spinal stabilisation exercise into a formal training programme (Stott, 2002). By having accesible exercise programmes like Clinical Pilates it allows easy access for the public to exercise programmes based on scientific evidence. What the public is not aware of is that many studios teach the traditional Pilates techniques that do not apply the findings from recent research. It is very important to consider the fundamental principles of an exercise programme that is being considered primarily for its core stability benefits. As there is no universal registration body that controls Pilates standards there is an unfortunate inconsistency in how this exercise format is taught.


There is much variation in the types of core stability training programmes described in the literature however there appears to be some consistent parameters that exist between the different programmes. These common attributes may provide the essence to what parameters make up the essential components of core stability training. The following discussion will attempt to define the essential components of a core stability training programme based on a review of the available evidence. Recommendation for further research will be made as appropriate.

Core Stability Training (CST) Protocol

Stage 1 of CST: General Assessment and Biomechanical Analysis


A comprehensive subjective examination is important as information about variables that can influence core stability can be identified. Of prime significance are mechanisms of injury, presence of pain, previous operations, congenital disorders and physical training techniques (McRrae, 1997; Sahrmann, 2002).


A biomechanical analysis commencing with a static observation then a dynamic overview of the lumbopelvic, thoracolumbar, scapulothoracic, upper and lower limbs provides an initial reference point for assessment of core stability function (Norris, 2000; Richardson et al., 1999; Horsley, 2002; Sahrmann, 2002) Consideration must be given to all factors acting on this region that may cause a perturbation from the position of ideal equilibrium (neutral pelvis) in both a static and dynamic environment such as muscle length (Toppenberg and Bullock, 1986), joint stiffness (Cyriax, 1996; Kaltenborn, 1993;  Maitland, 1997; Mulligan, 1999), neural mobility (Butler, 1991) and muscle function (Richardson et al., 1999; Wisbey-Roth, 2000; Sahrmann, 2002)


A valid objective measure for core stability should be taken to assist with assessment and diagnosis of components that need to be optimised and to allow for comparative monitoring in the future (see Chapter VI). The measures taken should depend on the presenting condition and the ultimate goals of the individual. Components that have objective measures include spinal alignment (Liemohn et al, 2002; Toppenberg and Bullock, 1986), muscle control (Richardson et al., 1999; Sahrmann, 2002; Wisbey-Roth, 2000), muscle activation patterns (Hodges, 2005, Oddson and De Luca, 2003) and endurance capacity (De Luca, 1984; Koumantakis et al., 2001; Roy et al. 1989 and 1994).


Stage 2 of CST: Treat pathology and implement biomechanical corrections


It may be appropriate to provide treatment to specific musculoskeletal injuries that may directly cause movement dysfunction or pain. (Commerford and Mottram, 2001). Pain can cause muscle inhibition which results in compensatory strategies and dysfunctional muscle activation patterns (Arvidsson et al., 1986; Hides et al, 1994 and 1996; McConnell, 2002). This can be very important with spinal rehabilitation as pain can limit exercise capacity which prevents efficient treatment of musculoskeletal dysfunction (Richardson and Jull, 1995). It may be appropriate in these cases to create a ‘pain free window’ to allow the performance of appropriate exercises (Hackley and Wiesel, 1993). The pain free window can be provided by oral medications or interventions such as nerve root, facet joint or epidural injections. Physiotherapy treatment has been shown to be effective in decreasing pain and improving function and could be used in conjunction or independently of other therapeutic interventions. Efficacy of physiotherapy treatment can be found in the Physiotherapy Evidence Database (PEDro) (Moseley et al., 2002).


This phase involves optimising the previously identified biomechanical deficiencies to achieve a more desirable spinal alignment and to achieve a state of relatively low tonic activity of the local stabilizing muscles while the spine is in the neutral zone (McGill et al., 2003; Panjabi, 1992b). This includes both active and passive structures. Implementation of strategies to decrease excessive compressive forces through the spine that are caused by tight or over active soft tissue structures (Norris, 2000). For example the iliopsoas muscle can cause a significant increase in extensor loading of the spine as well as decrease flexion range if shortened or overactive (Gibbons et al., 2002).


It is desirable to improve spinal alignment as a precursor to improving muscle activation patterns, strength and endurance components (Richardson et al., 1999). Acknowledgement of the mobility of neural structures is important as the restriction of neural mobility may influence muscle activation patterns (Butler, 1991). It is reported that neural mobility can be restricted by adhesions around nerve roots, vertebral joint dysfunction and muscle spasm (Butler, 1991). The area of neuro-dynamics is one that warrants further research.


There is no clear evidence in the literature concerning optimum muscle stretching techniques. The role of stretching is contested in the literature with its position being challenged in rehabilitation, prevention and performance enhancement (Shrier, 2002). Stretching soft tissue structures and mobilising neural structures anecdotally appear to be very important during rehabilitation (Butler, 1991). Scientific evidence has not delivered a concise answer with regard to stretching and CST. More specific investigations are necessary based on the above techniques. Sustained stretching has been reported to recruit slow motor units (Burke, 1968). This may benefit the local stabilising muscles and possibly the global stabilising muscles as they are purported to respond to low loads by encouraging slow motor unit recruitment. This physiological response to sustained stretch may be problematical with the stretching of global mobilising muscles. Implementing contract-relax stretching techniques may be more appropriate for global mobilizing muscles (Norris, 2000). This has not been vindicated by evidence to date. Gibbons and Comerford (2000) reported that the global mobility system is dysfunctional when it responds to low loads e.g. like postural sway where gluteal muscle insufficiency results in overactive hamstrings in single leg stance. Clark (1999) reported that stretching a mobiliser may encourage them to respond to low loads. This is contrary to their proposed function as global mobilisers that have a dominate role in tasks requiring speed and loading with fast motor unit recruitment (Comerford and Mottram, 2000). This is an important component that requires further investigation. It may be possible to create a biomechanical alteration by stretching a muscle group and then immediately performing specific exercises that ensure the optimal muscle recruitment pattern is maintained. This hypothesis warrants investigation.


 Stage 3 of CST: Implementation of neuromuscular phase


Commence low intensity exercise focusing on the ‘local muscle’ stabilising system performed in a neutral spine posture (Davidson and Hubley-Kozey, 2005; Hubley-Kozey and Vezina, 2002; Marshall et al., 2005; Souza et al., 2001). These muscles include: multifidus, transversus abdominis and iliopsoas (Erb, 2004; Gibson et al., 2002; Richardson et al., 1999).


Intensity levels up to 30-40% of maximal voluntary contraction (Davidson and Hubley-Kozey, 2005; Hubley-Kozey and Vezina, 2002; Sung, 2003). Tonic, continuous contractions. No phasic or ballistic movements (Richardson et al., 1999).


Frequecy of exercise has been poorly considered. Davidson and Hubley-Kozey (2005) had subjects train 4 times per week at levels < 40% of a maximum voluntary isometric contraction. They found that muscle activation patterns improved significantly following this programme. Sung (2003) reported improved muscle activation patterns due to neural adaptations with a training frequency of 3 times per week. Hagins et al. (1999) demonstrated improvements in subjects who performed exercises 3-4 times per week for 4 weeks. They did not report methodology further to include repetition rate and number of sets. This was a shortcoming in all the above mentioned studies.  DeMichele et al. (1997) found that two sessions per week produced the same strength gains as three. It would be incorrect to directly extrapolate this finding to exercises primarily working on stabilising muscles. The frequency of exercise required to produce neuromuscular adaptations is poorly reported in the literature and hence the optimal frequency for training is not established. Current studies such as DeMichele et al. (1997) can only act as indicators for optimum training frequencies. Repetitions of individual exercise and number of sets is poorly reported. Studies have ranged from 1 set of 15 repetitions (Stott, 2002)  to neuromuscular fatigue (Nelson et al., 1999). More research is required in this area.


Richardson et al. (1999) recommended isometric type contractions as this would meet the functional characteristics of the local stabilizing muscles as they demonstrated minimal length changes in different spinal positions. Encourage co-contractions of the inner unit of muscles i.e. transverses abdominus (Hodges and Richardson, 1996, Hodges, 1999), multifidus (Hides et al., 1996 and 2001), diaphragm (Hodges and Gandevia, 2000) and pelvic floor (Sapsford et al., 2001; Sapsford and Hodges, 2001). Observe for patterns of breathing and be sure the individual is breathing normally whilst drawing the abdomen in. (Stott, 2002). Progress to movement patterns incorporating cocontractions with the outer unit (global muscles) ensuring they do not mask function of the inner unit (Comerford, 2000; Comerford and Mottram, 2001)


Stage 4 of CST: Advanced physiological adaptations


Improving the endurance capacity of stabilising muscles (McGill, 2001) and functional training (Boyle, 2004; Haynes, 2004; O’Sullivan, 2000). Good evidence available concerning benefits of endurance training (O’Sullivan, 2000; Roy et a., 1999) however functional training guidelines currently lack credible foundation studies.


This simplified 4 stage approach to CST has been created from evidence derived from intervention studies where possible. Further research is required to fill the many gaps to optimise the efficiency and efficacy of this protocol. While the stages have been set in a logical sequence of progression accounting for timing of desirable adaptations it is acknowledged that the maximum potential for optimisation of anatomical and physiological structures will not be achieved prior to each stage progression. Some elements of each stage would be performed concurrently. The challenge here for the clinician is that the current research base has very limited guidelines on determining firstly, the timing of exercise progression and secondly, what exercise type to progress to. Of the many established exercise systems ‘purported’ to enhance core stability such as Pilates, Swiss Ball training, yoga, martial arts and the Alexander technique they all describe progressions of exercise however the decision on when and how to progress an individual is often based on subjectivity or intuition on the part of the clinician or trainer. It would appear that limited work has been performed on exercise progression criteria. This creates inconsistencies in the management of individuals undertaking CST.


Most core stability exercise programmes are based on the hypothesis of lumbar stabilisation proposed by Bergmark (1989), in which he described the concept of local and global muscle function. McGill et al (2003) stated that any exercise can be classified as a stabilisation exercise depending on how it is performed. Essentially by ensuring that sufficient joint stiffness is achieved during the exercise, repetitive practice will develop the appropriate motor patterns for the lumbar spine (McGill et al., 2003). McGill et al. (2003) did not report on thoracic spine or scapulothoracic function as part of lumbopelvic control however the assumption would be that maintaining a neutral thoracic kyphosis and maintaining scapulothoracic control during stabilising exercises is advantageous to overall trunk stability. These theories warrant further investigation.


Intervention studies considered in the development of the guideline from Key Question 6 can be seen in Table 13. More detail for individual studies can be found in Table 19, Appendix H which contains the evidence table for Key Question 6.


Table 13. Intervention studies for Question 6: Is there any evidence to support specific exercise programmes to enhance core stability?

Bibliographic citation Study type Ev lev General comments

AROKOSKI, J. P., VALTA, T., KANKAANPAA, M. & AIRAKSINEN, O. (2004) Activation of lumbar paraspinal and abdominal muscles during therepeutic exercises in chronic low back pain patients. Archives of Physical Medicine and Rehabilitation, 85, 823-832.

Cross-sectional 3 The validity of the 18 exercises chosen for the treatment of back pain were questionable. The frequency and intensity of the exercises were an uncontrolled variable as the majority of exercises were performed independently. Participants had only limited supervision, 4-6 times over 3 months. Exercises were based on a general protocol rather than being customised to an individual’s specific physiological deficiency. Validity of surface EMG measurements as a measure of muscle function. It would appear that this form of exercise training/therapy was not appropriate for the group. Interestingly, the exercises were not along the lines of segmental stabilisation theories. There was a significant emphasis on global stabilisers and movers.
AROKOS          AROKOSKI, J. P., VALTA, T., AIRAKSINEN, O. & KANKAANPAA, M. (2001) Back and abdominal muscle function during stabilization exercises. Archives of Physical Medicine and Rehabilitation, 82, 1089-1098.


Cross-sectional 3 All exercises caused EMG activity in the abdominal and paraspinal muscles. This study demonstrated that different exercises caused different muscle activation patterns and generated varying %MVC. Higher loads created by a change of position or unbalanced limb movements produced higher %MVC. Women generally achieved higher activation levels than men however it was hypothesised that men only needed to activate a smaller amount of their MVC to perform a similar activity.
BARDIN, L. D. (2003) Physiotherapy management of accelerated spinal rehabilitation in an elite level athlete following L4-S1 instrumented spinal fusion. Physical Therapy in Sport, 4, 40-45. Case study 3 This case study applied current principles of spinal stabilisation exercises to a specific athlete post surgery for L4-S1 spinal fusion. The rehabilitation followed a programme to rehabilitate postural muscles for segmental and global stabilization, proprioception, balance, trunk muscle endurance, coordination and leg strengthening. Spinal stabilization exercises were performed in unloaded positions of supine and 4 point kneeling. Subject returned to preinjury running performance level.

This study demonstrated the successful application of some key principles of core stability training i.e. transversus abdominus strengthening and neutral lumbo-pelvic position training.


DEMICHELE, P. L., POLLOCK, M. L., GRAVES, J. E., FOSTER, D. N., CARPENTER, D., GARZARELLA, L., BRECHUE, W. & FULTON, M. (1997) Isometric torso rotation strength: effect of training frequency on its development. Archives of Physical Medicine and Rehabilitation, 78, 64-69. Pre-test-post test randomised group design.



1+ The researchers concluded that a training frequency of 2 x week obtained better adherence to the programme with equal strength gains as the 3 x week group. They concluded that 2 x week is optimal for torso rotation training when training over a 12 week period.

Considerations for future research would be a comparison of different muscle groups and speeds of muscle contractions. Of the 11 other studies cited by the researchers it appeared to be a genuine consistent trend that significant strength gains could be achieved with a training programme of 2-3 x per week of 7-15 repetitions over 6-20 weeks. It would have been beneficial if a comparative statistical analysis was performed comparing the other studies. How useful is this finding clinically as most core stability programmes advocate dynamic functional stability as the method of training? There is a need to compare the isometric results with more dynamic movement patterns. This study didn’t directly measure the impact this form of isolationist strength training of the muscles involved in rotation of the torso would have on core stability as a whole.


HAGINS, M., ADLER, K., CASH, M., DAUGHERTY, J. & MITRANI, G. (1999) Effects of practice on the ability to perform lumbar stabilization exercises. Journal of Orthopaedic and Sports Physical Therapy, 29, 546-555.

Randomised pretest-posttest control group design 1+ (Also evidence for question 2) Study demonstrated that a modified pressure biofeedback isometric stability test was reliable and that a 4 week lumbar stabilisation exercise programme with weekly intervals of reinstruction and testing, improved the subject’s ability to perform progressively difficult lumbar stabilisation exercises. This programme of exercise was performed in a supine position so further investigation would be required to determine if this has carryover to standing positions. May be a useful first stage in activating the intersegmental local stabilising muscles prior to more dynamic stabilisation exercises.
HERRINGTON, L. & DAVIES, R. (2005) The influence of Pilates training on the ability to contract the Transversus Abdominis muscle in asymptomatic individuals. Journal of Bodywork and Movement Therapies, 9, 52-57.


Case-control 2+ This study demonstrated that the Pilates trained subjects’ maintained better lumbo-pelvic control compared with those who performed regular abdominal curl exercises or those who performed no exercise. There is potential bias between the exercise groups as the Pilates group exercised TrA for 45 min and the abdominal curl group only performed 15 min of exercise. It is also possible that Pilates style exercises recruit the deeper abdominal stabilising muscles better however more research would be necessary to confirm this.
HUBLEY-KOZEY, C. L. & VEZINA, M. J. (2002) Muscle activation during exercises to improve trunk stability in men with chronic low back pain. Archives of Physical Medicine and Rehabilitation, 83, 1100-1108.


Case series.

Prospective, comparative, repeated measures design.

3 The authors concluded that the 3 exercises were appropriate for low level muscle activation which may be appropriate for the initial phases of dynamic stabilisation programmes however due to their low activation levels were deemed inappropriate as part of a strengthening programme. This has significant relevance for core stability programmes as many exercise systems only concentrate on low level activation exercises. Further research is necessary to indicate which exercises provide higher muscle activation levels to progress patients/athletes to higher levels. If low level activation exercises are an important first stage of a core stability programme then the exercises reviewed may be appropriate. There is a degree of bias regarding gender in this study. No female population.
KOUMANTAKIS, G. A., WATSON, P. J. & OLDHAM, J. A. (2005) Supplementation of general endurance exercises with stabilisation training versus general exercise only. Physiological and functional outcomes of a randomised controlled trial of patients with recurrent low back pain. Clinical Biomechanics, 20, 474-482. RCT 1+ This study concluded that general exercise alone was just as effective as a combined approach of graded specific stabilisation exercises over an 8 week period for subjects suffering non-specific low back pain. A further consideration would be a comparison of a normal healthy population and a more specific pathology group such as lumbar spine ‘instability’. This study considered the important interaction between the local stabilising and the global mobilising systems and suggested that these systems may be able to work in parallel. Employing this strategy for the first few weeks of a core stability programme would potentially accelerate patients/athletes to a greater functional capacity faster than just performing isolated, low level activation exercises for the first few weeks of a programme. This requires further investigation. The integration of both types of exercise may have significant cost saving implications for rehabilitation and also recurrence rates for low back pain. Consideration must also be given to the type of specific stabilisation exercises prescribed to group 1.
DAVIDSON, K. L. & HUBLEY-KOZEY, C. L. (2005) Trunk muscle responses to demands of an exercise progression to improve dynamic spinal stability. Arch Phys Med Rehabil, 86, 216-23. Case series.

Prospective, comparative, repeated measures design.


Different levels of exercise elicited differing amplitude patterns in stabilising muscles. This potentially highlighted how the muscular stabilising system responded to specific challenges.  Recommended loads for a strength training effect are 60-100% of 1 maximal voluntary isometric contraction (MVIC) (e.g. Astrand and Rodahl, 1986; American College of Sports Medicine, Position Paper.Progressive models in resistance training for healthy adults. Med Sci Sport and Ex 2002;34:364-80). Levels 2-4 of this programme were low load (<40% MVIC) so would not have a strengthening effect however would potentially influence endurance capacity if performed at higher repetition rates. This warrants further investigation. Endurance training is considered more important for stability training than strength therefore exercise Levels 1-4 could provide valuable input to the initial phases of a core stability programme (McGill et. al,2003). This study demonstrated the stabilising role of the global stabilising muscles such as rectus abdominus and the influence this has on spinal stability. In this example the more challenging Level 5 exercises elicited a higher NMRS.

MARSHALL, P. W. & MURPHY, B. A. (2005) Core stability exercises on and off a Swiss ball. Arch Phys Med Rehabil, 86, 242-9.



Case series.

Prospective comparison study.

3 Muscle activation patterns on the Swiss ball were greater than the ‘stable’ environment. Specific exercises appeared to involve different synergistic relationships between the muscles. The authors suggested that this point justifies why core stability programmes should have a variety of different exercises to challenge the neuromuscular system i.e. no clear muscle recruitment patterns associated with stabilisation exercises. For training the local stabilising system the abdominal drawing-in task and quadruped exercise demonstrated minimal rectus abdominus activity in both a stable and unstable environment. Increased rectus abdominus activity with the Swiss ball may be inappropriate during a lumbar stabilisation programme if the model supported by Richardson et.al (1990 and 1999) is correct. The role of rectus abdominis needs further investigation in the core stability concept. Supported theory that external oblique muscle activation is unaffected by exercise type or surface. Limitations of this study include a small sample size and narrow age range.
SAPSFORD, R. R., HODGES, P., RICHARDSON, C. A., COOPER, D. H., MARKWELL, S. J. & JULL, G. A. (2001) Co-activation of the abdominal and pelvic floor muscles during voluntary exercises. Neurology and Urodynamics, 20, 31-42. Case-series 3 Voluntary pelvic floor contractions correlated with abdominal muscle activity. An additional pilot experiment was carried out using the opposite protocol i.e. voluntary contraction of the abdominal muscles and recording pelvic floor activity. This demonstrated pubococcygeus muscle activity associated with submaximal abdominal exercises. The cylinder theory of core stability described by a number of authors suggests that the pelvic floor muscles make up the base of the cylinder (Chek, 2004). This study demonstrated some potential methods of incorporating pelvic floor and specific abdominal muscle activity in relation to lumbar spine position. This study needs to be broadened to include males and pathological groups and greater trial numbers.

SOUZA, G. M., BAKER, L. L. & POWERS, C. M. (2001) Electromyographic activity of selected trunk muscles during dynamic spine stabilization exercises. Archives of Physical Medicine and Rehabilitation, 82, 1551-1557.


Case series 3 Degree of EMG activity was suggestive that there was unlikely to be a strengthening effect in healthy subjects. Regardless of exercise type or level the maximal EMG value achieved was 41%MVIC. This study needs to be repeated with an injured population and a wider age group. Back pain patients may achieve a strengthening effect. Study demonstrated that dead bug and quadruped exercises complement each other as part of the same exercise programme due to their muscle activation patterns. This study made no attempt to differentiate between internal and external oblique muscle activity.
SUNG, P. S. (2003) Multifidi muscles median frequency before and after spinal stabilization exercises. Archives of Physical Medicine and Rehabilitation, 84, 1313-1318.


Cohort 2- Functional disability improved within a 4 week period suggestive that mechanisms behind the improvement are other than the benefits of strengthening. The inconsistency of the multifidus muscle fatigue responses between gender suggested the intensity, duration or type of exercise performed needed to be reviewed. It has been suggested that individual differences in lumbar curvature and mobility may influence fatigue levels. Other studies have indicated that back muscle fatigability was significantly more prevalent in men (e.g.Moffroid et al, (1993) in Physical Therapy)
VEZINA, M. J. & HUBLEY-KOZEY, C. L. (2000) Muscle activation in therapeutic exercises to improve trunk stability. Archives of Physical Medicine and Rehabilitation, 81, 1370-1379. Case series.

Prospective comparative study

3 Demonstrated EMG muscle activation patterns for 3 exercises in healthy males therefore providing a reference for injured populations. The 3 exercises demonstrated different muscle activation patterns thereby providing the option for the clinician to match any deficiencies to appropriate rehabilitation exercises. No muscle groups achieved an activation level significant for strengthening benefits. Level 1 trunk stability test demonstrated coactivation between abdominal and erector spinae muscles. Minimal activity for trunk extensors during pelvic tilt and abdominal hollowing exercises. Study should be extended to include a female population and injured population.
WEBBER, S. C. & KRIELLAARS, D. J. (2004) The effect of stabilisation instruction on lumbar acceleration. Clinical Biomechanics, 19, 777-783. Case series 3 (Also evidence for Question 2) This study demonstrated that stabilisation instruction had an immediate effect on decreasing lumbar acceleration and therefore institute a neuromuscular change to influence overall kinematics. Accelerometry was shown to be able to objectively quantify mechanical change in the lumbar spine. Further research is necessary to compare these results to an injured population and to more complex tasks. Emphasises the importance that verbal and practical instructions can have on performance of an exercise. There are many “how to” books written on how to improve core stability, an interesting study would be with a population that only had written instructions.
COSIO-LIMA, L. M., REYNOLDS, K. L., WINTER, C., PAOLONE, V. & JONES, M. T. (2003) Effects of physioball and conventional floor exercises on early phase adaptations in back and abdominal core stability and balance in women. J Strength Cond Res, 17, 721-5.



Case-control 2+ This study demonstrated that short-term core exercises using a physioball resulted in greater EMG neuronal activity and torso balance in previously untrained women compared with performing exercises on the floor. The validity of measuring core stability with the CYBEX system is questionable. This intensity of exercise may not be able to be extrapolated to other populations such as athletes, males or injured populations. These variables warrant further investigation.
RICHARDSON, C.A., TOPPENBERG, R. & JULL, G. A. (1990) An initial evaluation of eight abdominal exercises for their ability to provide stabilisation for the lumbar spine. Australian Journal of Physiotherapy, 36, 6-11.


Case series

Pilot trial

3 The authors proposed an equation to create a stability score. This concept warrants further exploration as it could assist in creating a score for core stability. The study concluded that exercises that involved trunk stabilisation during rotating resistance appeared to produce muscle activation patterns that could stabilise the vertebral column. Studies on the internal validity of the stability score need to be performed if this concept was to be used clinically.