Core Stability for Low Back Pain: Separating the Research from the Marketing

In 1996, researchers Hodges and Richardson published a study suggesting that transversus abdominis (TVA) activation was delayed in people with low back pain. The implication was that targeting this muscle might address a root cause of spinal instability.

This was a single study. What followed was an industry.

Pilates programmes, specific stabilisation protocols, clinical assessments of "core weakness," and treatment rationales built around TVA activation proliferated across physiotherapy, exercise physiology, and personal training. The original finding — one study, a specific population, a specific measure — became clinical gospel.

What happened when the research caught up?

Seven or more systematic reviews have now examined specific stabilisation exercise for low back pain. Six of them found no benefit over general graded exercise.

A 2014 systematic review with meta-analysis was direct:

"There is strong evidence that stabilisation exercises are not more effective than any other form of active exercise in the long term… further research is unlikely to considerably alter this conclusion." (Smith et al., 2014)

A companion review addressed the proposed mechanism:

"Changes in morphometry or activation of transversus abdominis following conservative treatments tend not to be associated with the corresponding changes in clinical outcomes." (Wong et al., 2014)

Not only does specific stabilisation training not outperform general exercise — the TVA changes it produces don't predict whether someone gets better. The mechanism the intervention is built on doesn't appear to drive the outcome.

What about the one exception?

There is a subgroup of patients who appear to respond better to stabilisation training than general exercise: those who score highly on the Lumbar Spine Instability Questionnaire (LSIQ). The complication is that LSIQ items include questions like "I have had this pain for a long time" and "I am sometimes fearful to move because of my pain" — neither of which has anything to do with biomechanical instability.

This raises an important question: does the exercise outperform general exercise because it targets a specific instability, or because it aligns with that patient's expectations and internal model of their condition?

What this means in practice

General exercise works for low back pain. The evidence is reasonably consistent that supervised exercise of most kinds is effective — the specific modality matters less than the engagement, progressive loading, and behavioural factors around it.

Specific stabilisation training as a superior intervention for a universal muscle deficit is not supported by current evidence. That doesn't mean Pilates or core work can't be part of a programme. It means the clinical rationale for it needs to be honest.

Telling a patient their back pain is caused by a weak core and that the fix is specific TVA activation creates a structural narrative that may not serve them — particularly if the symptom returns when they stop the exercises.

If you want to understand the evidence behind common low back pain interventions — including what the research actually supports and what has been extrapolated beyond its evidence base — our Low Back Pain and BPS Practice course provides a comprehensive, referenced review.

→ Explore the course at tkex.org/courses