Change in Transversus Abdominis and Obliquus Internus Activity Following Different Kinds of Physiotherapy

Purpose: To investigate if transversus abdominis (TrA) and obliquus internus (OI) muscle thickness (a proxy for muscle activity in
these muscles) changes following spinal stabilisation exercises or other forms of physiotherapy in people with chronic low back pain (cLBP).
To investigate if changes in muscle thickness are related to clinical outcome.
Relevance: Spinal stabilisation exercises are moderately effective in cLBP. Their mechanisms of effect are said include change in TrA and
OI muscle activity. There is little evidence to support this claim.
Participants: 212 people, 136 (64%) female, mean (range) 44 (18-76) years old, with 7.4 (0.25-52) years history of non-specific cLBP
referred to physiotherapy.
Methods: Participants were randomised to spinal stabilisation exercises (small-group supervised individualised exercises; aims
included increasing TrA activation), individual physiotherapy (manual therapy and exercise according to clinicians' reasoning, some were
taught spinal stabilisation exercises), or physiotherapist-led pain management (including small-group supervised general exercise).
Participants had left and right TrA and OI muscle thickness measured with real-time ultrasound in relaxed supine lying, supine lying with
low abdominal drawing-in, and relaxed standing. Pain was measured with numerical analogue scale, disability with Roland-Morris
questionnaire. Data collection occurred before and six months after randomisation.
Analysis: Comparisons of muscle thickness between different treatment groups and between 'improvers' (participants with >50% pain
reduction) and 'non-improvers' (=50% pain reduction) were made using analysis of covariance with baseline data as covariate. Results
reported are mean(SD) of sum of left and right muscle in mm.
Results: 169 participants were reassessed. 101 provided complete ultrasound data, 32 were not measured because they did not
attend reassessment in person, and in 36 (21%) ultrasound images were not sufficiently clear to be confidently measured. Participants
not measured reported greater disability and pain, weighed more, and had higher BMI. Following interventions, both TrA and OI thickness
was unchanged in all three test positions and there was no between-group differences in either muscle (all p>0.1). In relaxed supine lying, for
all groups combined, 'improvers' (n=60) TrA thickness was 6.0(1.6) at baseline and 5.9(1.8) at six months; 'non-improvers' (n=41) was 6.1
(1.7) and 6.0(1.6)(p=0.74). In supine drawing-in, all 'improvers' TrA thickness was 10.2(3.4) at baseline and 10.7(3.0) at six months;
'non-improvers' was 9.7(3.2) and 9.9(2.6)(p=0.28). In relaxed standing, all 'improvers' TrA thickness was 7.7(2.7) at baseline and 8.5(3.5) at
six months; 'non-improvers' was 7.9(2.6) and 8.5(2.9)(p=0.61). Spinal stabilisation participants alone had no differences in TrA thickness in
any positions between 'improvers' (n=18) and 'non-improvers' (n=16)(p>0.2). Clinical results have been reported previously (Critchley et al
2007), all groups improved equally in pain and disability.
Conclusions: There were no differences in TrA thickness (activity) following treatment between specific spinal stabilisation exercise
and general exercise groups or between participants with >50% pain reduction and those without. Future research could investigate other
spinal stabilisation treatment mechanisms, including non-physiological mechanisms, and identify subgroups for whom spinal
stabilisation exercises are effective.
Implications: Pain reduction following spinal stabilisation training appears to be unrelated to change in transversus abdominis activity
which does not support the use of specific spinal stabilisation exercises people with in non-specific cLBP.