The trusted source for
healthcare information and
September 2001; Volume 3; 69-71
By Edzard Ernst, MD, PhD, FRCP
Low back pain (LBP) is a burden for the patient, the family, and the economy.1 At the same time it represents one of the most significant unresolved therapeutic challenges to date. Although exercise and spinal manipulation often are recommended, their clinical effectiveness has not been documented beyond reasonable doubt.2,3 Other treatment options may be well worth considering.
Classical (Swedish) muscle massage has a long history and is associated with a range of effects that are potentially beneficial in the symptomatic treatment of LBP.4 For instance, massage relaxes the mind as well as the musculature and increases the pain threshold, possibly through endorphin release.5 It also can enhance local blood flow, which could increase the clearance of local pain mediators.5 However, physiological effects should be clearly differentiated from clinical effectiveness.
In Europe, massage has been routine therapy for acute and chronic LBP for many decades.4 A recent survey suggests that no fewer than 87% of Austrian back pain patients received massage.6 The evidence for or against its effectiveness in LBP recently has been summarized systematically.7 This article is an update of this evidence in the light of recently available data.
The following databases were searched from their beginnings to July 2000: Medline, Embase, and Cochrane Library. I also searched my own (extensive) files and asked colleagues with an interest in massage therapy for additional references. The bibliographies of all trials and reviews thus retrieved were scanned for further relevant publications. Randomized clinical trials (RCTs) were considered if they included subjects with LBP and if one patient group was treated with massage therapy as the sole treatment. Massage therapy included manual or apparative (using a device or apparatus) massage of muscular and soft-tissue structures of the back. Trials in which massage was combined with other treatments (other than mild heat, a routine adjunct to massage) were excluded. When massage was used in both the reference and the experimental group, trials also were excluded. No language restrictions were applied. Data were extracted and validated in a standardized, pre-specified way by the present author. In cases of dual publication, only one report was used.
Six trials meeting the above criteria were found.8-13 Hsieh et al randomized 63 patients into four groups.8 The patient group vaguely was defined as having "non-specific LBP" for more than three weeks and less than six months. Patients with neurological deficits, fractures, tumors, infections, or spondyloarthropathy were excluded, as were those with sciatica or pain radiating below the knee and positive nerve root tension signs. All groups were treated for three weeks with chiropractic manipulation, corset, transcutaneous electrical nerve stimulation, or massage. The latter intervention consisted of "gentle stroking massage to the whole back area without any deep soft-tissue manipulation" administered by massage therapists three times per week. The evaluator was blinded to group assignment. In the massage group, scores on the Oswestry scale decreased from 41.9 ± 13.0 to 32.7 ± 18.7 and scores on the Roland-Morris scale (a standard, validated outcome measure for LBP) decreased from 45.7 ± 27.3 to 37.5 ± 28.9. Chiropractic manipulation was significantly superior to massage; there were no other significant inter-group differences.
Godrey and coworkers randomized 81 patients with acute LBP (otherwise undefined) into three groups.9 Patients received chiropractic manipulation, electrostimulation, or massage. Massage was "administered by a kinesiologist from the sciatic notch to the thoracolumbar junction with light effleurage (stroking) for 10 minutes." Treatment was assessed using a custom-made, non- validated index. The statistical evaluation was complex and provided no qualitative data for the massage group. All groups improved significantly with no difference among groups.
Hoehler et al randomly assigned 95 patients with acute or chronic LBP to two groups.10 Patients were included if they had "palpatory cues indicating hyperalgesia or a restricted or painful range of vertebral motion." No further description was given. Patients received either spinal manipulations or "soft-tissue massage." No further details were provided as to the massage treatment. Endpoints included subjective pain, straight leg raising to pain, and the fingertip-to-floor distance. Both groups improved. Immediately after a therapy session, spinal manipulation seemed to yield superior results compared to massage, but there were no significant differences between groups at the end of the treatment period.
Konrad and colleagues divided 158 outpatients into four groups.11 Patients had "back pain localized to the lumbosacral region, with or without radiation to the thigh" for at least three months. They received either balneotherapy, traction, massage, or no such treatments. Massage consisted of underwater massage with a jet of hot water (37° C, 1 atm, 10 cm distance) over the affected area. Pain was assessed by visual analogue scales and analgesic consumption was monitored. Each patient was assessed on admission, at the end of the four-week treatment period and one year later. Analgesic consumption decreased from a baseline value of 5.1 to 2.2 and 2.1, respectively. Pain decreased from 56.7 to 24.6 and 45.8, respectively. These changes were statistically significant compared to those seen in the no treatment group.
Preyde randomized 98 patients with subacute non-specific LBP into four groups: One received soft-tissue massage only, while the other groups received "comprehensive massage therapy" (massage, remedial exercise, and posture education); exercise and posture education; or treatment with a detuned laser (placebo).12 Pain was quantified with the McGill Pain Questionnaire, which includes a rating of pain intensity, the Present Pain Index (PPI), and a rating of pain quality, the Pain Rating Index (PRI). Ninety-two percent of subjects completed treatment and 85% completed follow-up evaluations.
After six treatments (within about one month), the massage/exercise/posture education group had significantly lower scores on the PPI compared to the other three groups. The massage-only and the massage/exercise/posture groups both had significantly lower scores on the Roland Disability Questionnaire (RDQ) than the remedial exercise and sham laser group, and both were significantly better than sham laser on the PPI.
One month post treatment, 63% of the massage/exercise/posture group reported no pain compared to 27% of the soft-tissue manipulation group, 14% of the remedial exercise group, and 0% of the sham therapy group. One month after treatment, the massage-only group was statistically better than the placebo group but was not superior to the exercise/posture education group.
Cherkin et al randomized 262 patients (ages 20-70 years) with persistent back pain into three groups: massage, acupuncture, or self-care educational materials.13 Up to 10 sessions of massage or acupuncture were administered during 10 weeks. Symptoms and function were assessed via telephone interviews after four, 10, and 52 weeks. At 10 weeks, massage therapy yielded significantly better results on symptom and disability scales than the other two approaches. At one-year follow-up, massage therapy still was significantly better than acupuncture, but not significantly different from self-care. The massage group sustained the lowest cost of subsequent care and used the least amount of medication for treating their back pain.
Considering the popularity of massage as a therapy for LBP, the fact that only six RCTs have been published to date (and that most of these have serious methodological limitations) is disappointing. Three of the studies are not, strictly speaking, trials testing the effectiveness of massage; massage was used in these studies as a control rather than as an experimental intervention. RCTs designed to test the effectiveness of massage therapy for LBP have become available only recently.
Methodological limitations include failure to describe the patient populations in sufficient detail; failure to account for dropouts; inadequate description of the randomization procedure; failure to base sample size on proper power calculations; and other statistical flaws. Moreover, several trials provide inadequate descriptions of the massage therapy applied.8-10 However, the study with the most rigorous design13 yields a clearly positive result and should further encourage high-quality work in this area.
Considering these limitations, conclusions from this review must be tentative. Massage therapy appears to be an effective treatment for LBP, although its efficacy has not been established beyond doubt. Studies certainly do suggest that massage is helpful for LBP. Unquestionably, more research on massage therapy is warranted.
Dr. Ernst is Professor, Department of Complementary Medicine, School of Postgraduate Medicine and Health Sciences, University of Exeter, United Kingdom.
1. Ernst E. Physical stress in low back pain. Eur J Phys Med Rehab 1994;4:112-113.
2. Koes BW, et al. Spinal manipulation and mobilisation for back and neck pain: A blinded review. BMJ 1991; 303:1298-1303.
3. Ernst E, Assendelft WJJ. Chiropractic for low back pain. BMJ 1998;317:160.
4. Westhof E, Ernst E. Geschichte der Massage. Dtsch med Wschr 1992;117:150-153.
5. Ernst E, Fialka V. The clinical effectiveness of massage therapy—a critical review. Forsch Komplementärmed 1994;1:226-232.
6. Wiesinger GF, et al. Benefit and costs of passive modalities in back pain outpatients: A descriptive study. Eur J Phys Med Rehab 1997;7:182-186.
7. Ernst E. Massage therapy for low back pain: A systematic review. J Pain Symptom Manage 1999;17:65-69.
8. Hsieh CY, et al. Functional outcomes of low back pain: Comparison of four treatment groups in a randomized controlled trial. J Manipulative Physiol Ther 1992; 15:4-9.
9. Godrey CM, et al. A randomized trial of manipulation for low-back pain in a medical setting. SPINE 1984;9: 301-304.
10. Hoehler FK, et al. Spinal manipulation for low back pain. JAMA 1981;245:1835-1838.
11. Konrad K, et al. Controlled trial of balneotherapy in treatment of low back pain. Ann Rheum Dis 1992;51: 820-822.
12. Preyde M. Effectiveness of massage therapy for sub-acute low-back pain: A randomized controlled trial. CMAJ 2000;162:1815-1820.
13. Cherkin DC, et al. Randomized trial comparing traditional Chinese medical acupuncture, therapeutic massage, and self-care education for chronic low back pain. Arch Intern Med 2001;161:1081-1088.