Research Review By Dr. Robert Rodine©


Download MP3

Date Posted:

July 2012

Study Title:

Centralization in patients with sciatica: are pain responses to repeated movement and positioning associated with outcome or types of disc lesions?


Albert HB, Hauge E & Manniche C

Author's Affiliations:

Spine Centre of Southern Denmark; Institute of Regional Health Services, University of Southern Denmark

Publication Information:

European Spine Journal 2012; 21: 630-636.

Background Information:

This study presents data obtained from secondary analysis of a prospective randomized controlled trial, previously published as ‘The efficacy of systematic active conservative treatment for patients with severe sciatica: A single-blind randomized clinical controlled trial’ (published in Spine 2012; 37(7).

Low back pain is known to be a disabling and costly problem to society, with an estimated lifetime prevalence of 80%. Within the catchment of low back pain is a subgroup presenting with sciatica, involving nerve root compression, with a lifetime prevalence estimated at 4-5%.

Within sciatica patients, centralization is a term used to describe the proximal retreat of distal leg/hip pain towards its point of origin (the spine) during specific movements or sustained positions. A standardized assessment and treatment approach was based on this theory and presented originally by Robin McKenzie, which has been referred to as ‘McKenzie’, but more formally it is also known as Mechanical Diagnosis and Therapy (MDT).

This theory postulates that in the presence of nerve root compression, patients will respond immediately or within days to dynamic loading strategies which reduce or abolish distal symptoms. This theory is based on the dynamic disc model which assumes that, given intact hydrostatic pressure, repeated end range loading will return displaced nuclear material within an intervertebral disc, reducing nerve root compression. This theory however, is based on the concept of an intact annulus and therefore intact hydrostatic pressure. If the annulus is presumed ruptured, in the case of an extruded or sequestered nucleus, it is believed that centralization cannot occur.

The purpose of this study was to examine a group of subjects with sciatica in order to:
  1. Determine the prevalence of certain pain responses following repeated movement;
  2. examine associations between baseline pain responses; and
  3. examine prognosis given baseline characteristics of disc lesions.

Pertinent Results:

  • A total of 182 subjects agreed to participate. Five were excluded and 11 were lost to follow-up, leaving 165 subjects for complete analysis. The mean age was 45 (range 37-52), 48% were female.
  • The mean Roland Morris Disability (RMQ) score at baseline was 15.5 (range 11-18).
  • Radiculopathy-related exam findings were reported by the authors as follows: ‘Sum score of positive straight leg raise, sensory and motor deficits and asymmetric reflexes in lower extremities was: 0 = 1%, 1 = 4%, 2 = 30%, 3 = 43% and 4 = 22%.’ (Note: this appears to be stating that subjects were granted 1 point for a positive SLR, 1 point for a motor deficit, etc. However, this data was poorly described.)
  • Seeking confirmation from Albert & Manniche (1), it appears that 95% of all subjects showed at least 2 positive neurological signs, with a mean of 2.8.
  • FOR INTEREST ONLY: While data on the entire sample was not presented, grouped data was. The intervention group showed that 94.7% +ve SLR, 52.6% with motor deficits, 77.9% with disturbed sensation and 55.8% with asymmetric reflexes. The sham group showed 94.2% with a +ve SLR, 50% with motor deficits, 83.6% with disturbed sensation and 53.5% with asymmetric reflexes. The intervention group showed current leg pain as 4.3/10 with total leg pain at 18/30 (ranging from 15-21). At the end of treatment, current leg pain was 1.5/10 with total leg pain at 4/30 (ranging from 0-9). One-year follow-up showed current leg pain at 1.5/10 with total leg pain at 3/30 (ranging 0-10). The sham group showed current leg pain as 4.5/10 with total leg pain at 18/30 (ranging from 12-21). At the end of treatment, current leg pain was 2.3/10 with total leg pain at 4/30 (ranging from 0-12). One-year follow-up showed current leg pain at 1.4/10 with total leg pain at 2/30 (ranging 0-8).
  • Within this study group, 84.8% of subjects were classified as centralizers (25.5% reported abolition, 43.6% reported reduction and 15.8% were unstable).
  • The remaining 15.2% of the study group were deemed non-centralizers (7.3% were peripheralizers and 7.9% reported no effect).
  • With respect to disc lesion morphology, 6% of subjects showed sequestration, 31% showed extrusion, 9% showed broad-based disc protrusion, 38% showed focal disc protrusion, 18% showed disc bulging and 8% were classified as normal.
  • The mean improvement in RMQ scores for all three centralization groups was 9.5 points at 8 weeks and 12.0 points at 12 months. The peripheralization group showed 7.0 points at 8 weeks and 14.0 points at 12 months. The no effect group showed 3.0 points at both 8 weeks and 12 months. When comparing the sub-groups of centralizers, the mean reduction at eight weeks was 8.1 for the abolitioners, 7.1 for the reducers and 9.7 for the unstable subjects. No significant difference between these groups were reported.
  • The authors also attempted to correlate pain response with MRI findings. Results found that 83.7% of those subjects with an intact annulus were able to demonstrate centralization. Based on the assumptions of the dynamic disc model, this makes excellent sense.
  • HOWEVER, subjects with extruded/sequestered discs and therefore a ‘certain lack of intra-discal pressure’ showed an even higher level of centralization, at 93.5%. (Keep in mind that we are comparing 63 subjects without an intact annulus to 102 subjects with an intact annulus.)
  • Compared to baseline, all three centralization groups showed a 15 point decrease in total leg pain at eight weeks and 16 points at one-year. The peripheralization group showed 14 points at eight weeks and 18 points at one-year. The no-effect group showed 8 points at eight weeks and 10 points at one-year.

Clinical Application & Conclusions:

The major take-home points of this study are:
  1. Both centralizers and peripheralizers demonstrated good outcomes at all time points, irrespective of treatment.
  2. Morphology of disc lesions identified via MRI had no association to the subjects’ pain response to movement or positioning.
  3. The presence of an intact annulus identified via MRI had no correlation to the presence of centralization.
The most interesting observation from this study is regarding prognosis. It is commonly believed that patients who centralize have a good prognosis, whereas those who do not centralize have a poorer prognosis (2). The current study challenges this finding, as the presence or absence of centralization had little to do with prognosis in the given sample. This being said, the no-effect group (those where repeated movements or position had no effect on the location or intensity of distal pain) were found to have a poorer prognosis at short and long term follow-up. Assuming the dynamic disc model is correct, we would classify this no-effect group as having pain that stems from an origin other than discal, given that directional loading is not affective. Therefore, this group would have an altered prognosis given its heterogeneity.

However, assuming the dynamic disc model is correct would also lead us to the conclusion that those without an intact annulus would not centralize, which this study also contradicts (finding centralization in greater proportion of subjects without an intact annulus than in those with.)

In the end, clinicians may use this information when making prognostic decisions regarding sciatica patients. This study does however point out that more research is needed in order to more fully understand the big picture of disc-related low back and leg pain.

Study Methods:

Study recruits were referred to a specialist spine centre from private practitioners (medical doctors), rheumatologists and chiropractors.

Subjects were required to be 18-65 years of age, with current dermatomal/radicular pain in the lower limb which measured at least 3/10 on the numerical pain rating scale and a duration of pain greater than 2 weeks but no longer than 1 year. Subjects were excluded in the presence of serious pathology, previous surgery, pregnancy, language/communications difficulties, inability to complete the rehabilitation protocol or a pending workers compensation claim.

For the purposes of primary examination, subjects were either randomized to sham exercises or symptom-guided exercises. Both groups contained identical patient education and advice. Subjects were required to complete at least four treatments, to a maximum of eight, over a period of 12 weeks. Treating practitioners were blinded to exam results.

Outcome Measures:
The primary outcome measure referred to in this study is the Roland Morris Disability Questionnaire (RMQ), which measures activity limitation.

ASIDE: The authors of the presented paper also report on improvements in leg pain intensity. Collection of this data or its purpose is not described in the methods section. For clarification, details were sought from Albert & Manniche (1): The Low Back Pain Rating Scale was used to measure low back and leg pain on a 0-10 point numerical scale and includes components of current leg pain, worst leg pain in the past two weeks and average leg pain in the last two weeks (creating a possible 30 point total score).

Centralization Procedures:
Within this study, centralization was defined as ‘the phenomenon by which distal limb pain emanating from the spine is immediately or eventually abolished in response to the deliberate application of loading strategies. Such loading causes reduction, then abolition of peripheral pain that appears to progressively retreat in a proximal direction.’

Each subject was then questioned regarding how movements or positions affected symptoms. The assessor attempted to target the exact movement or position most likely to centralize pain.

While standing, patients were instructed to rate their pain intensity (for lumbar, gluteal, thigh, crus and foot regions) and indicate the exact location of pain on a diagram. Subjects were positioned into the centralizing position, to its end-range. This was repeated until symptoms abolished, or until the most distal pain had improved. The maneuver was discontinued if symptoms worsened. The subject then walked freely for 1 minute, prior to filling out another pain drawing. The examiner then compared the two pain drawings assigning one of the following labels to each subject:
  1. Group A – Abolition Centralization: The most distal pain was abolished and pain was recorded more proximally on the second drawing.
  2. Group B – Reduction Centralization: The pain was located at the same distal location but with reduced intensity.
  3. Group C – Unstable Centralization: The pain was reduced or abolished during the repeated movement testing or positioning but after resuming a weight-bearing position for one minute, the pain intensity level returned to the pre-testing intensity.
  4. Group D – Peripheralization:The most distal pain increased in intensity or area in all test movement directions or positions.
  5. Group E – No Change: Throughout testing, there was no change in the location and intensity of the distal symptoms.
Groups A, B and C were classified as centralizers while groups D and E were classified as non-centralizers.

MRI Examination:
Following baseline assessment all subjects underwent MRI examination in the supine position. MRI’s were assessed by a medical radiologist who classified them as normal, bulging, focal-protrusion, broad-based protrusion, extrusion or sequestration.

Discs classified as normal, bulging or protruded were deemed to have an intact annulus and therefore intact hydrostatic pressure, whereas discs classified as extruded or sequestered were classified as not having an intact annulus and a loss of hydrostatic pressure.

Study Strengths / Weaknesses:

The most concerning limitation within this analysis is that the compared subject groups (centralizers, peripheralizers and no-effect) contained subjects who were exposed to both the intervention and the sham treatment with data collection at eight weeks and one year. As referenced in the primary analysis (1) both groups improved, though the intervention group showed statistically significant improvement over the sham group. This being said, further sub-grouping of the sample and examining the effect of treatment group may have yielded improved information with respect to prognosis of centralizers versus non-centralizers.

Another limitation, which the authors address, concerns the use of pain diagrams to determine if centralization is occurring. Overall, this process was based on the subjective comparison of assessors. Previous work in this area has divided the low back and leg into six regions on the pain diagram, whereby the assessor would evaluate the diagram before and after to determine if the pain has moved into a different region, rather than simply moved (2). While this method has the potential to miss centralizing subjects if they do not reach a certain anatomical landmark, it is certainly an objective method by comparison. Perhaps future research will look at validating a more sensitive method of identifying centralization.

Additional References:

  1. Albert & Manniche. The efficacy of systematic active conservative treatment for patients with severe sciatica. Spine 2012; 37(7): 531-42.
  2. Donelson et al. Centralization phenomenon. Its usefulness in evaluating and treating referred pain. Spine 1990; 15: 211-13.