Abdominal muscle feedforward activation in patients with chronic low back pain is largely unaffected by 8 weeks of core stability training

Blikk på forskning i Fysioterapeuten 8/2012


Blikk på forskning utarbeides i samarbeid med Journal of Physiotherapy, som trykker forskningspresentasjonene under betegnelsen Critically Appraised Papers, CAPs.


Summary of: Vasseljen O et al (2012) Effect of core stability exercises on feedforward activation of deep abdominal muscles in chronic low back pain: a randomized controlled trial Spine 37: 1101–1108. [Prepared by Margreth Grotle and Kare B Hagen, CAP Editors.]

Question: Does timing of abdominal muscle activation in response to rapid shoulder flexion change after 8 weeks with low-load core stability exercises (CSE), high-load sling exercises (SE), or general exercises (GE) in chronic nonspecific low back pain (LBP) patients?Design: A randomised, controlled trial with concealed allocation.

Setting: Patients were recruited from general practitioners, physiotherapists, or by advertising at a regional hospital in Norway.

Participants: Men and women, aged 18–60 years, with chronic nonspecific LBP for 3 months or more, and pain score of 2 or more on a 0–10 numeric rating scale were included. Key exclusion criteria included radiating pain below the knee or neurological signs from nerve root compression, and former back surgery. Randomisation of 109 participants allocated 36 to CSE, 36 to SE, and 37 to GE.

Interventions: Patients in the three groups attended treatment once a week for 8 weeks, supervised by a physiotherapist. All were encouraged to stay active and received an information booklet with general information on LBP. The CSE were individualised according to protocols focusing on isolated activation of transversus abdominis during an abdominal drawing-in manoeuver in supine hook-lying position with ultrasound feedback. Written instructions to carry out the drawing-in exercise (10 × 10 seconds 2–3 times per day) at home were also provided. The SE maintained the lumbar spine stable in neutral position throughout a range of leg/arm positions and movements, using elastic bands attached to the pelvis to help the patient maintain a neutral spine position. The SE was performed for 40 minutes in a physiotherapy clinic. The GE group received generalised trunk strengthening and stretching exercises supervised by a physiotherapist at a fitness centre.

Outcome measures: Primary outcome was change in onset of the deep abdominal muscles in response to rapid shoulder flexion.

Results: 102 participants completed the study. No or small changes were found in onset after treatment. Baseline adjusted between-group differences showed a 15 milliseconds (95% CI 1 to 28) and a 19 millisecond (95% CI 5 to 33) improvement with SE relative to CSE and GE, respectively, but on one side only. There was no association between changes in pain and onset over the intervention period (R2 ≤ 0.02).

Conclusion: Abdominal muscle onset was largely unaffected by 8 weeks of exercises in chronic LBP patients with changes in onset of less than 20 milliseconds between groups. 


This RCT utilises a large cohort to examine mechanical onsets of the deep abdominal muscles and response to different exercises. The findings show limited changes in the timing of the core onsets and no association with pain or disability. Interestingly 99% of the 109 cohort subjects had feedforward (FF) onsets of the contralateral abdominal muscles. The current dogma is that a small percentage of the LBP cohort should have had FF responses. Therefore, this may question how any exercise regimen may “improve” the onset of the LBP cohort if they already have what could be within a normal range. This could be the basis of the continued discussion on the significance and validity of the FF corset hypothesis and the method of detecting onsets (Masse-Alarie H et al 2012). Another observation is that the assessment of mechanical movement ‘onsets’ may not correlate with activation (EMG) onsets because movement can be achieve via relaxation. We have previously shown that the FF response of (ipsilateral) transversus abdominus can be inhibitory; this is also highly directional specific and controlled by planned rotational torques (Morris et al 2012, Allison et al 2008a,b). Therefore these underlying rotation mechanisms may in part explain the observed side to side differences in change of the mechanical onsets as well as the greater improvements with the sling exercises.

This RCT contributes to evidence that the association between the deep muscle FF responses and spinal pain is not simple. It adds to the growing diversity of opinion of the hypothesised mechanisms of motor control in LBP. This is an important reminder that there should be a separation between the research question askingifthe treatment works, andhoworwhythe treatment works. Too many therapists and researchers rely on one to justify the other. 

Garry T Allison, Faculty of Health Sciences, Curtin Health Innovation Research Institute, Australia 


Allison GT et al (2008b)J Orthop Sports Phys Ther38: 228–237.

Allison GT et al (2008a)Br J Sports Med42: 930–931.

Massé-Alarie H et al (2012)Exp Brain Res218: 99–109.

Morris SL et al (2012)Clin Biomech27: 249–254.