Research Review By Dr. Daniel Avrahami©


Download MP3

Date Posted:

January 2012

Study Title:

The McKenzie Method compared with manipulation when used adjunctive to information and advice in low back pain patients presenting with centralization or peripheralization: A randomized controlled trial


Peterson T, Larsen K, Nordsteen J et al.

Author's Affiliations:

Back Center Copenhagen, Denmark; Holstebro University Hospital, Denmark; and Department of Rheumatology, Copenhagen University Hospital, Rigshospitalet, Denmark.

Publication Information:

Spine 2011; 36(24): 1999-2010.

Background Information:

As health care providers we all understand the prevalence of low back pain. Commonly used treatments for nonspecific low back pain (NSLBP) include (but are not limited to) self-management through education, tailored exercise programs (i.e. McKenzie method) and manual therapy (i.e. spinal manipulation, mobilization, soft tissue therapy etc.).

A good clinician understands that there is no one therapy solution for low back pain sufferers. Furthering this concept, researchers are stressing the importance of subgrouping NSLBP patients to aid in decision-making and effective management strategies. Recently, there has been increasing literature to support subgrouping LBP patients. For example, Fritz et al. (1) demonstrated that patients with LBP and hypomobility experienced greater benefit from an intervention of manipulation and those with LBP and hypermobility were more likely to benefit from a stabilization exercise program. [EDITOR’S NOTE: we have reviewed much of this literature on RRS – please refer to the Clinical Prediction Rules section of the database.]

Several studies have examined McKenzie method versus spinal manipulation in patient subgroups with acute or subacute NSLBP. However, these studies have not come to consensus and the verdict is still out. Therefore, Petersen et al. set out to compare the effects of the McKenzie method with spinal manipulation in addition to education and advice for a patient population with persistent LBP (greater than 6 weeks).

Pertinent Results:

Approximately half of both treatment groups reported success and a reduction in mean disability above 50% at the end of the treatment program along with the follow-up at 2 months and 1 year. A significantly larger proportion of patients withdrew or was excluded during treatment in the manipulation group, likely due to lack of treatment effect according to the authors (43 patients in the manipulation group vs. 28 in the McKenzie group).

At the end of the treatment there was a statistically significant difference of 13% in favor of the McKenzie group. For every 8 patients, one would have success with the McKenzie method who would not have achieved this response with manipulation.

Two months post-treatment there was a statistically significant difference of 12% in favor of the McKenzie group. For every 7 patients, one would have success with the McKenzie method who would not have achieved this response with manipulation. In addition, The McKenzie group demonstrated a 1.5 point greater decrease (95% CI 0.2 to 2.8) in level of disability compared to the manipulation group.

One year post-treatment the McKenzie group demonstrated a greater reduction in disability, by 1.5 points (95% CI 0.2 to 2.9), compared to the manipulation group.

Clinical Application & Conclusions:

After the treatment period the McKenzie method was shown to be more effective in treating patients with LBP showing symptoms of centralization or peripheralization compared with manipulation. However, the relative improvement between the two groups was fairly small and its clinical relevance is questionable.

The largest difference between the two groups was found at 2 and 12 months after the completion of treatment. Once again the differences between the two groups were not very big and clinical relevance is once again somewhat questionable.

Since the majority of patients were classified as reducible disc syndromes based on centralization of symptoms one would presume that patients would profit the most from the McKenzie method. However, practically speaking many patients with possible lumbar disc disease with or without nerve involvement receive spinal manipulative treatment in practice. Due to the inherent risks associated with rotational forces induced on a patient with a presumed disc deficit, and considering more patients in this study from the manipulation group were referred to surgical evaluation (nine patients in the manipulation group vs. five in the McKenzie group), the authors recommended the McKenzie method as the first choice of treatment for these patients.

The reason there might not have been much difference between the two groups is that, in the end, both treatment methods intended to mobilize the lumbar spine, and both were monitored by the patient’s pain response during the course of treatment. Thus, both treatments are likely to have similar effects on the outcomes, namely the pain mechanism.

The question remains: are centralization and peripheralization prognostic factors regardless of treatment – are these factors influencing the prediction of the outcome of their disease? Or are they treatment effect modifiers related to a specific treatment – do they change the outcomes related to certain types of treatment?

This study continues to support the value of a classification approach based on clinical examination findings in the management of patients with LBP. In the end our goal continues to be to thoroughly assess a patient, providing an accurate diagnosis and apply a treatment program that is specific to the needs of that patient and their pathology.

Study Methods:

LBP patients, between 18-60 years of age, with more than 6 weeks of pain and disc-related symptoms were recruited. Disc-related symptoms were determined via centralization (the abolition of symptoms in the most distal body area) or peripheralization (the production of symptoms in a more distal body area) of symptoms with or without signs of nerve root involvement.

1969 patients were screened for eligibility. Only 350 of the screened patients were selected for the study. Baseline measures were obtained and 175 patients were randomized into the McKenzie group or manipulation group. Comparison of 17 baseline characteristics and treatment information for the patients in the two treatment groups showed no significant differences between groups.

In both treatment groups, patients were educated about their physical assessment, the benign course of back pain, back care guidance and given a Danish version of “The Back Book.” A maximum of 15 treatments for a period of 12 weeks was provided. Each patient received an individual program of self-administered mobilizing, stretching, stabilizing, and/or strengthening exercises chosen by their physical therapist or chiropractor dependent on the treatment goals. Patients continued the exercises for a minimum of 2 months after completion of the treatment at the back center.

Treatment for the McKenzie group included specific directional preference exercises that centralized the patient’s symptoms. Manual vertebral mobilization techniques including high velocity thrust were not allowed. Treatment for the spinal manipulation group included all types of manual techniques including vertebral mobilization, high velocity thrust and myofascial trigger-point therapy. It was at the discretion of the chiropractor as to which technique, or combination of techniques was given to each patient. General mobilizing exercises were allowed but not specific exercises in the directional preference of a particular patient, if one existed.

The primary outcome measures in this study were changes in the 23-item modified Roland Morris Disability Questionnaire (RMDQ) 2 months after the completion of the treatment program. Secondary outcome measures included RMDQ changes throughout the study, changes in pain, global perceived effect, quality of life, days with reduced activity, return-to-work, satisfaction with treatment, and use of health care after the completion of treatment.

Treatment success was defined as an absolute reduction of at least 5 points on RMDQ. The “true baseline” measure was taken using serial measurements of the RMDQ and calculating the average value over 16 days. This study used between-group analyses on the dichotomous variables. Multivariate analyses were performed to examine the influence of various baseline covariates on main and secondary outcomes. A logistic regression analysis was used to examine treatment success, and for the main continuous outcome, RMDQ, multiple regression analysis was used. A secondary analysis combining the three follow-up points was performed by means of repeated measures analysis of variance (ANOVA).

Study Strengths / Weaknesses:

Unfortunately this study did have several notable weaknesses. Some of the McKenzie group patients were given an educational booklet describing a “lumbar roll” for correction of the seated position and some of the manipulation group was given manual techniques including vertebral mobilization as well as myofascial trigger-point massage, alternating lumbar flexion/extension movements, and stretching exercises. These additional therapies were provided for the patients based on the discretion of the therapist. Therefore, the treatment successes may have been highly dependent on the skill of the clinician. Now this may seem realistic to everyday practice, however, the generalizability of the treatment results are likely hampered by the fact that clinical decision making was performed without standardization by the clinicians.

In addition to the weakness previously mentioned, stabilizing and strengthening home exercises were only given to those patients that the clinicians felt that it was indicated in order to achieve their treatment goals. Once again this brings up the clinician’s skill and bias into the treatment results. It is difficult to pin-point the treatments results to the McKenzie or manipulation treatment as there were various confounding treatments provided to different patients.

A final limitation worth noting was that the study had a relatively high withdrawal rate during the intervention. The withdrawal rate covers patients who decided to discontinue treatment during the course as well as patients that were excluded by decision of the practitioner. The majority withdrew or was excluded for reasons likely, in the authors’ opinion, to be related to lack of treatment effect (43 patients in the manipulation group vs. 28 in the McKenzie group). The authors suggested that the difference in withdrawal rate between groups supports the conclusion that the McKenzie treatment was the most suitable for our patient sample.

Even with these flaws, this study is a step in the right direction towards more accurately determining patient subgroupings and their outcomes based on specific treatment techniques regularly used in daily practice. Since there were some interesting and noteworthy findings from this study, future research should examine other clinical findings that would support the use of McKenzie method, manipulation or both therapies for patients with acute, subacute or chronic low back pain.

Additional References:

  1. Fritz JM, Whitman JM, Childs JD. Lumbar spine segmental mobility assessment: an examination of validity for determining intervention strategies in patients with low back pain. Archives of Physical Medicine and Rehabilitation 2005; 86(9): 1745-52.
  2. Petersen T, Larsen K, Nordsteen J, Olsen S. The McKenzie method compared with manipulation when used adjunctive to information and advice in low back pain patients presenting with centralization or peripheralization. Spine 2011; 36(24): 1999-2010.