Research Review by Dr. Shawn Thistle©


Jan. 2008

Study Title:

Effectiveness of an extension-oriented treatment approach in a subgroup of subjects with low back pain: A randomized clinical trial


Browder DA, Childs JD, Cleland JA, Fritz JM
Author Affiliations: Physical Therapists from various military (US Army-Baylor University) and non-military research and clinical centers in the United States

Publication Information:

Physical Therapy 2007; 87: 1608-1618.


Recently, the importance of sub-grouping low back pain (LBP) patients to guide clinical decision making has been revealed, and supported by quality research. Past efforts to study treatment options for LBP have produced largely equivocal results, perhaps in part due to a lack of patient sub-grouping. A substantial body of recent research has led to the development of a Clinical Prediction Rule (CPR) for LBP to assist clinicians when determining treatment options for these patients. The CPR uses historical and physical examination findings to subgroup patients into one of 4 main groups.

One of the patient/treatment subgroups identified in the CPR is “specific exercise”, which includes patients that are likely to respond to direction-specific exercise in the form of flexion, extension, or lateral shift. Please note, the LBP CPR has been previously summarized by the RRS – refer to the review titled “Low Back Pain – Clinical Prediction Rule” written in July 2007 for a complete summary.

This study focused on a group of patients who, by utilizing the CPR, were identified a priorias those likely to respond to an extension-oriented exercise program. Extension exercise programs typically include a combination of active and passive, sustained and repeated movements to promote extension in the lumbar spine. Further inclusion criteria for this study included:
  • age between 18-65
  • LBP symptoms of any duration extending distal to the buttocks in at least one extremity
  • centralization (see below) with active extension movement of the lumbar spine
  • OSWESTRY score of at least 30% (subjects also completed a Numeric Pain Rating Scale and the Fear Avoidance Beliefs Questionnaire)
Basically, centralization means that distal symptoms in the buttock or leg centralize toward the spine (or midline), or abolish, with a certain movement (repeated or sustained) of the spine. Subjects were excluded for the following reasons:
  • presence of “red flags” for serious pathology (ex. tumour, fracture, infection)
  • current pregnancy
  • recent surgery to the lumbar spine within 6 months
48 subjects (mean age 39, 31% female) were randomized into one of two exercise groups after undergoing a standardized history and physical examination. Subjects in both groups attended physical therapy sessions twice weekly for the first two weeks, then once per week for the next two weeks, for a total of 6 sessions. They were also asked to perform exercises on days in between treatments. Further, all subjects were provided with a copy of an exercise instruction booklet (specific to their intervention) and were required to record their exercises in a log to track adherence. The study groups differed as follows:
  1. Extension-Oriented Treatment Approach [EOTA] (n=26): Subjects performed a series of exercises to promote lumbar extension during treatment sessions which included active movements in standing and prone positions (10 reps x 3 sets for standing and prone positions, held at end-range for 2-3 seconds). The second component was posterior-to-anterior mobilization of lumbar segments selected by the treating therapists (10-20 Grade I-IV mobilizations according to Maitland). At home, patients were instructed to perform 1 set of 10 repetitions of prone extensions every 2-3 hours for the 4 week treatment period. Therapists also advised patients on how to maintain neutral lordosis during daily activities, and to avoid sitting for greater than 30 minutes.
  2. Strengthening (n=22): Subjects performed core stability exercises targeting the Transverse Abdominus (TrA) and other abdominal musculature (from work by Richardson, Jull, Hicks, Hides et al.). Therapists provided hands on feedback as required, and instructed subjects to perform the exercises on non-treatment days. (Note: with the mode of exercise chosen here, it is odd that they chose “strengthening” as the title for this group – perhaps “stabilization” or “motor control” would have been more appropriate)
Sample size calculations were based on detecting a 10-point difference on the OSWESTRY – 24 subjects per group were required to provide 80% power to detect this difference (p < 0.05).

Pertinent Results:

  • a significant difference existed between groups at baseline – more patients (n=5) in the EOTA group had previous lumbar surgery compared to zero in the strengthening group
  • within-group comparison of those 5 subjects to the others in the EOTA group revealed that the 5 surgical subjects had significantly less improvement in disability at 1 week, 4 weeks, and 6 months (approaching significance even with only 5 subjects: p = .07)
  • a significant group x time interaction was found for disability (p = 0.02) – the EOTA group had less disability at each follow-up interval (1 week, 4 weeks, 6 months*)
  • EOTA subjects also had lower pain levels, but only at 1 week, after which both groups experienced similar pain reduction
  • 17 subjects in each treatment group sought additional treatment, but no differences were noted in this factor at 6 month follow-up (this includes 2 subjects in each group undergoing surgery during the study period)

Conclusions & Practical Application:

In this study, the EOTA was more effective than a trunk strengthening program in patients hypothesized a priorito be appropriate for extension exercise intervention. This provides preliminary evidence that utilizing the LBP CPR to identify patients likely to benefit from an extension-oriented exercise program can have a favorable impact on disability outcomes and pain in the short term. Overall, this supports the notion that outcomes can be improved when patients are matched to appropriate interventions.

Previous studies using more heterogenous patient samples have not shown positive results for EOTA, so this study is a step in a positive direction, as many clinicians utilize extension exercise with great success. Further studies are required however, to address and rectify some of the shortcomings of this study which include:
  • *this study was slightly underpowered, and lost a significant number of patients at the 6 month follow-up interval (11 in the EOTA group and 6 in the strengthening group) – therefore the 6 month results should be interpreted with caution*
  • a significant difference in pain between groups was not found beyond 1 week – suggesting that further specification of patient selection may be necessary to maximize the effectiveness of EOTA
  • further work is required to determine the exact interaction between previous surgery and response to EOTA – as those in this study with previous surgery did not respond as well to EOTA as those without (this factor alone may have weakened the results of this study