PT on the Net Research

Transversus Abdominis Revisited

Editor's Note: Although Mel Siff passed away in 2003, his work continues to represent some of the most intelligent thinking about stability and core strength, which is why we have chosen to re-release his article for use in this author symposium.

There are many therapists and fitness professionals who maintain that transversus abdominis (TvA) needs special isolated training because they consider that most people, even elite athletes, have some TvA deficiency. It would be interesting to find out how this whole mania started about attributing back problems more to some dysfunction of TvA or other abdominal muscles than to any other muscles in the chain of soft tissues involved with stabilizing and moving the trunk.

Trunk stabilization is a complex process involving many different muscles and other soft tissues, any of which could be blamed for leading to spinal inefficiency. This immediately raises the question as to why certain muscles of the abdominal muscle complex are being singled out as being crucial to spinal stabilization. Can anyone justifiably state that any single muscle involved in the trunk stabilization and mobilization process is more or less important in 'causing' spinal pain and dysfunction?

Recently, Dr Stuart McGill, one of the most prolific international scientists researching TvA actions and trunk stability, stayed with me for a few days and we both agreed that a great deal of what has been written by various popularists and therapists on testing and training does not accurately reflect what science has discovered about spinal stability and the role played by the different trunk muscles. This article is the outcome of the very productive time that we spent together.

More Recent Research Findings

An interesting point is that brain scans being taken during human movement are now showing that the same regions of the brain are not necessarily involved in controlling the identical movement (there was an article on this in a recent 'Scientific American', among other journals). It appears that dynamic and static stabilisation involves various opportunistic and nonlinear dynamic (NLD) processes of nervous and muscle action, so that it is questionable how anyone can justifiably single out individual muscles in trying to explain the aetiology of spinal pain and dysfunction.

Moreover, there is a huge body of complicated and contradictory research into the nature of motor control and it is very premature for anyone to popularize and simplify the situation with fairly trite recommendations for manipulating motor control mechanisms to "solve" back problems (if anyone thinks that this remark is exaggerated, try to sort out the high powered, very perplexing discussions concerning the "equilibrium point hypothesis" in motor control). The degree of variability, individuality and specificity involved in any action involving the human spine decrees that the problem is far form being satisfactorily solved. There are so many strategies for controlling back pain and dysfunction to different degrees of success (most of them involving a degree of placebo influence) that one can fairly safely state that nobody has yet solved the problem of spinal management. This is further compounded by the clinical observation that a fairly large percentage of non traumatic spinal disorder resolve themselves without treatment. In the light of all of this, the current zeal with which TvA or other abdominal theories are being propounded as the panacea for many spinal ills should be seen to be misplaced. One could quite happily suggest that any muscle "weakness" (or is it rather inefficient muscle activation due to neural activation issues?) could be proposed as the villain of the piece, and nobody could definitely prove such a premise to be wrong.

Some Fundamental Questions

So much has been written recently about how essential TvA is to trunk stabilization and mobility, but has anyone ever undertaken studies to show that people who may not be able to activate it very effectively may compensate very effectively by using other patterns of muscle activation and use? Is there really any convincing proof that "less than optimal" TvA activation definitely results in significantly less efficient and more dangerous trunk action in all daily activities?

Is there any genuine proof, other than the current conjecture, that TvA contributions to trunk action are not as predictable and deterministic as is being claimed by some therapists? Does it really matter all that much if it sometimes does not contribute as it "should" during human movement and its decrease in function is compensated for by other muscle actions? Those who work clinically often will notice how many patients with physical differences or dysfunctions "compensate" by using different muscles or patterns to carry out a given task without any apparent ill-effects.

Note that the combined force directions of the external and internal obliques can produce the force direction of the TvA, thereby carrying out roles like those of TvA. In other words, if active internal oblique muscles pull along one diagonal across the trunk and active external obliques pulling along the inverse diagonal, the resultant force will exerted roughly parallel to the floor (if you are standing), in other words in the same direction as tension in the TvA. This vector resultant alone may be more than adequate to compensate for any apparently undesirable TvA malingering. Moreover, activation of a muscle still does not prove its function, for example, as a Oprime mover¹ or stabilizer.

We do not even know if deficiencies in TvA action or timing are cause or effect of back pain or dysfunction. TvA activity may really be a secondary effect produced by appropriate control of intra-abdominal pressure via involuntary or voluntary breath management. Most of the time the TvA appears to activate quite automatically in the healthy person when the body is forced to stabilize or react to any quick positional or movement changes. I have often quoted EMG work by Basmajian to remind people that TvA activation is intimately linked to any actions which elicit changes in breathing status, especially if forced breath holding or exhalation takes place.

Remember Beevor's Axiom, namely that the body generally knows of actions not muscles? So, just learn some appropriate methods of controlling one's breath in different situations or carry out dynamic activities which compel the body to produce such breathing actions, and the rest will follow quite naturally! It is interesting that more and more scientific evidence is emerging to show that deliberate attempts to activate TvA, to "draw in the abs" or related strategies to enhance "core stability", balance and strength are misguided in healthy subjects, so it will be interesting to note how proponents of the TvA training hypothesis will now respond to this body of contradicting research.

More Scientific Clues Emerge 

Increasing numbers of folk are beginning to hear or read the evidence of scientists such as Dr Stuart McGill that these ideas about enhancing core performance and abdominal control are incorrect or misleading, because almost all of the research being quoted in defense of isolated TvA training has been done in controlled clinical situations and does not necessarily apply to competitive sport.

There is absolutely no evidence that TvA becomes dysfunctional on its own, because in any trunk stabilization task it functions in conjunction with the other abdominal and back muscles. In no real life movement is TvA is tensioned in total isolation. The individual degree of contribution by the abdominal and back muscles is under reflexive, proprioceptive control, not conscious, isolated attempts at activation in dynamic action.

The trunk, abdominals, lower extremities and so forth are not closed systems and any action involving these subsystems influences what is happening in all parts of the body and the body as a whole. It is vital that the body be regarded in terms of a systems theoretical approach, rather than one which makes very tenuous assumptions about the closed-ness of conveniently isolated subsystems whose apparent isolation from other systems invariably is based entirely on convenience or convenience.

Even if one attempts to apply a systems theoretical approach, it may still be inadequate to regard the entire body as the superordinate closed system, as is implied, for instance, by the current somewhat simplistic emphasis on so-called "core training". The limitations of the latter concept may readily be noticed if one observes that it is very rare in land-based sport for core stability to be manifested in the absence of contact with the ground or external objects.

Peripheral stability, which usually is reliant on solid contact between the extremities of the body with some surface, is essential before core stability becomes implicated in a given sporting action on land. Without adequate peripheral stabilization, the functional capabilities of the "core" are meaningless. The entire body or the body-surface constitutes the appropriate closed system for our attention. Thus, if terms such as "core stabilization" are to be used, then they need to be carefully applied within the appropriate context.

This is not to negate the value of approaches that use isolationist approaches for valid therapeutic or analytical reasons, such as those involving EMG mediated biofeedback, "Kegel" exercises, and post surgical respiratory exercises, but it is to stress that the unqualified application of isolationist approaches to sports conditioning needs to be viewed with careful circumspection.

Analysis of Studies on TvA Action

Before we go any further, let us examine some of the studies that have led to this invalidly extrapolated application of isolated TvA training to the world of sport and daily life.

  1. Hodges PW, Richardson CA Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis Spine 1996 Nov 15; 21(22): 2640-50. The delayed onset of contraction of transversus abdominis indicates a deficit of motor control and is hypothesized to result in inefficient muscular stabilization of the spine.
  2. Hodges PW, Richardson CA Delayed postural contraction of transversus abdominis in low back pain associated with movement of the lower limb J Spinal Disord 1998 Feb; 11(1): 46-56. In the control subjects the onset of electromyographic activity of all trunk muscles preceded that of the muscle responsible for limb movement, thus contributing to the feedforward postural response. The EMG onset of TvA was delayed in the low back pain (LBP) subjects with movement in each direction, while the EMG onsets of rectus abdominis, erector spinae, and oblique abdominal muscles were delayed with specific movement directions. This result provides evidence of a change in the postural control of the trunk in people with LBP.
  3. Hodges P, Richardson C, Jull G Evaluation of the relationship between laboratory and clinical tests of transversus abdominis function Physiother Res Int 1996; 1(1): 30-40. This study indicates that a reduction in the ability to draw in the abdominal wall is related to changes in the coordination of TvA, although the magnitude of the changes were not correlated. The degree of causality between these co-varying but independent manifestations of the function of TvA is uncertain. Comment: Many muscles in the body are activated in anticipation of a given movement, because this optimizes qualities such as the mechanical stiffness and damping ability and enhances movement efficiency. For example, if this did not happen with certain leg muscles, the acts of running and jumping would be impossible. Thus, a great deal more work has to be done to ascertain how the changes in TvA timing influence body movement and stability. Moreover, it has not yet been established if changes in timing of TvA activation are a cause or effect of low back pain or dysfunction.

Note, however, what McGill has to say:

"Some confusion exists in the interpretation of the literature regarding the issue of abdominal hollowing and abdominal bracing. Richardson's group observed that the hollowing of the abdominal wall recruits the transverse abdominis. On the other hand, an isometric abdominal brace coactivates the transverse abdominis with the external and internal obliques to ensure stability in virtually all modes of possible instability (Juker et al.,1998). Note that in bracing the wall is neither hollowed in nor pushed out. In this way, abdominal bracing is superior to abdominal hollowing in ensuring stability.

With this background, Richardson and colleagues (1999) noted that the recruitment of transverse abdominis is impaired following injury. The group developed a therapy program designed to reeducate the motor system to activate transverse abdominis in a normal way in LBD patients. Thus, while hollowing may act as a motor reeducation exercise, the act of hollowing does not ensure stability. Some clinical practitioners may have misinterpreted Richardson's work to suggest that abdominal hollowing should be recommended to patients who require enhanced stability in order to perform daily activities. This is misguided, Abdominal bracing, which activates the three layers of the abdominal wall (external oblique, internal oblique, transverse abdominis) with no "drawing in," is much more effective than abdominal hollowing at enhancing spine stability (McGill, 2001; Grenier and McGill, in press)."

This comment by Dr. David D. Wise in response to an Internet letter of mine on this topic is most relevant:

"There are basically two different types of stabilization. One is what is called global stabilization. This is the use of muscles to balance the trunk on the pelvis much like guy wires would support a flag pole. The guy wires, like the global stabilizers (large abdominal and back extensor muscles spanning from pelvis to trunk) match or overcome the torque of the pole leaning or torque of wind against it. Several studies have shown that transversus abdominis is able to produce little torque and has little effect on this type of stabilization.

Transversus abdominis (according to theories advanced by proponents of TvA training) has more to do with segmental stabilization. This type of stabilization is akin to the friction between ground and flag pole (that which a hole in the ground would make more effective). This would be called segmental stabilization by TvA training proponents. If the flag pole one were trying to erect were segmented, not only would it need stabilization at its base but also at each segment. The guy wires wouldn't offer much stabilization at each segment (other than compression) but intersegmental structures would. It is intersegmental stabilization that therapists are excited about.

However, I recently conducted a five week clinical study of 20 subjects with supposed lumbar instability (10 had transversus training and 10 had the usual lumbar stabilization training). At the end of 5 weeks there was no difference found between the groups. This proves little, but it is surprising to me how specific transversus abdominis training has become so popular based (to my knowledge) on very few clinical efficacy studies with young spondylolithesis patients as subjects, for example."


EMG studies reported in Basmajian show that TvA is automatically recruited if one suddenly or intensively has to stabilize the trunk or if one rapidly coughs or hold breath during such efforts. Now, when one is indulging in any fairly demanding activities which require stabilization or mobilization under exposure to any perturbation, that is exactly what happens, anyway, so why does the average person with healthy musculature have to learn TvA activation when neuromuscular processes reflexively take very adequate care of the action. Let the proponents of TvA training use invasive EMGs during actual sporting or exercise activities (and not simple static or slow, non-ballistic postures) to prove the hypothesis about the necessity for voluntary TvA activation.

So, while one may voluntarily or involuntarily recruit TvA more strongly during the beginning stage of a squat or deadlift, this does not mean that this type of muscle recruitment is either wise or correct at later stages of the movement. In fact, deliberate attempts to consciously try to change the patterns of muscle involvement in a dynamic movement may well slow down or disrupt important aspects of neural processing and distort the lifting pattern, thereby leading to inefficiency and possible injury.

In this respect many of us have stated that the simple act of learning the correct way of executing an exercise will automatically ensure that the relevant muscles will become involved quite reflexively in that exercise. As was stated before, "the body (generally) knows only of movements, not muscles," which stresses that the primary task is not to focus on a given single muscle, but on the pattern of movement required.

How many lifters can we definitely identify who have failed during the squat, deadlift or the Olympic lifts because of some deficiency in TvA strength or timing? Most unsuccessful or injurious lifts are the result of imperfect technique or lack of adequate leg and back strength or power rather than of any weak or lazy abdominal muscles.

As increasing numbers of scientists are now beginning to stress, we cannot necessarily extrapolate physical therapy and medical studies of posture based on abnormal or fairly sedentary subjects to the world of elite sport and daily living - we need similar proof obtained from competitive and everyday situations before we can accept any claims about what certain components of the abdominal muscles do or don't do.

Meanwhile, those who advocate special isolated TvA or other abdominal exercises need to state clearly that theirs is one of several other theories and that it may help in some clinical situations, but that theirs is not the only universally correct approach.

Maybe we should state it even more simply so that many more will assimilate its truth more readily: "Mary had a little lamb and everywhere that Mary went, the lamb was sure to follow" .... or, "My body had a little muscle and everywhere that body went, that muscle was sure to follow."


  1. Basmajian, J. Muscles Alive. 1978.
  2. McGill, S. Low Back Disorders. 2002.
  3. Siff, M. Facts and Fallacies of Fitness. 2003.
  4. Siff, M. Supertraining. 2003.