Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Massage for chronic low back pain

Three Part Question

[In patients presenting with chronic low back pain] is [massage better than simple advice] at [decreasing pain, increading mobility, and speeding a return to work]

Clinical Scenario

A 32 year old female patient presents to her general practitioner with a 3 month history of lower back pain. She is previously fit and well, has no neurological symptoms or signs or features to suggest a serious underlying cause. You diagnose a simple low back strain and advise her to take analgesics as required and to mobilise. She informs you that her friend is a massage therapist and wonders if you think it would help her get better quicker.

Search Strategy

Medline, OVID interface via ATHENS 1966-Feb 2005
back or exp Back Pain/ or exp Low Back Pain/ or lumbar or exp Back Pain/ and exp MASSAGE/ or limit to humans and english language and abstracts and 2001-2005
back-pain and massage

Search Outcome

Medline. 83 papers. 4 Systematic review since 2001 included in this BET.
1 relevant systematic review found.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Furlan AD
Not stated
Randomised or Quasi randomised studies investigating the use of any massage (hands or mechanical device) as a treatment for non-specific back pain. Extensive search along Cochrane guidelines.Systematic review and meta-analysisNumber of papers found9 publications included in studyOnly one of the trials had an effective non-intervention (sham) control group. All studies included in the paper had participants with back pain lasting more than 6 weeks, often much longer making the results difficult to extrapolate to the acute setting.
Quality of papers (according to Cochrane back pain study group criteria)5 had high quality scores. 3 had low quality scores.
Massage vs. sham LaserOne study. Massage better than sham therapy at long and short term function, short term pain, but not on quality of pain. Only 51 patients in trial.
Massage vs. spinal manipulation4 studies. 3 showed manipulation better than massage in early stages. Limited evidence to suggest that at 3 weeks little difference in outcomes.
Massage vs. electrical stimulation3 studies. No difference found.
Massage vs. Corset2 studies. No differences found
Massage vs. exercise,1 study. Massage better at short term function. No difference at long term. No differene in pain intensity, ROM, pain quality at short or long term.
Massage vs relaxation therapy1 study. No significant difference.
Massage vs. acupuncture1 study. Massage better in terms of function at 10 and function and symptoms at 52 weeks.
Massage vs. self care education1 study. Massage better at 10 weeks (SMD -2.79 CI -3.22 to -2.36) and 52 weeks (-1.49 (-1.84 to -1.15) for pain intensity and symptom botherness.Better for function at 10 weeks SMD -1.8 (CI -4.31 to -3.29) but not at 52 weeks SMD 0.57 (0.26 to 0.88)
Massage as component of combined therapy1 study. Massage more effective if combined with an exercise program.
Ernst E
Papers were sought if they were either chiropractic or massage therapy systematic reviews. Data sources from electronic databases, references and authors personal files.Review article of previously published systematic reviews.Number of papers found6 papers. Only 2 related to massage therapyBoth of the massage papers included in the review were written by the author of this review! No mention of the widely known Cochrane review published 2 years earlier. Inclusion criteria for studies not explicit. No explicit quality score for studies.
Quality of studiesNot explicit
Systematic reviews of low back pain and massage1 study that included 6 primary studies. Benefit of massage unproven. (this review was also written by Ernst)
Cherkin DC.
A review of systematic reviews of RCTs published since 1995 evaluating massage, acupuncture, spinal manipulationfor non-specific back pain. The authors also looked for any papers puiblished since the publication of any systematic reviews.Narrative review of systematic reviews and subsequent relevant controlled trials.Papers found2 systematic review. 3 subsequent controlled trials.Search strategy limited to time after 1995.
Quality of included studiesNot explicit.
Effectiveness of massage therapyMassage appears to have some beneficial effect that may last up to 52 weeks. Recent higher quality studies suggest greater benefit to massage.


The role of massage in low back pain has been assessed in a number of trials and in a number of systematic reviews as detailed above. Whilst the methodology of the reviews differs there does appear to be a consistent finding that massage may be of benefit in low back pain. Massage appears to be a worthwhile therapy for patient with back pain of more than 6 weeks duration. As with many other aspects of the treatment of back pain it is often difficult to isolate one treatment alone and there appears to be evidence to suggest that massage may be most effective when combined with exercise treatments.

Clinical Bottom Line

Massage appears to be a useful treatment for subacute and chronic back pain, particularly when combined with other treatments such as exercise.

Level of Evidence

Level 1 - Recent well-done systematic review was considered or a study of high quality is available.


  1. Furlan AD, Brossseau L, Imamura M, Irvin E. Massage for low back pain. Spine 2002;17:1896-1910
  2. Ernst E Manual therapies for pain control: Chiropractic and Massage Clin J Pain 2004;20:8-12
  3. Cherkin,D.C.; Sherman,K.J.; Deyo,R.A.; Shekelle,P.G. A review of the evidence for the effectiveness, safety, and cost of acupuncture, massage therapy, and spinal manipulation for back pain. Ann Intern Med 2003;138:898-906