Skip navigation

Effectiveness of massage

Clinical bottom line: Lack of high quality evidence for determining whether massage is an effective treatment.

Systematic review:

Ernst E. Clinical effectiveness of massage - A critical review. Forsch Komplementärmed 1994; 1:226-232

Date review completed: 1993

Number of trials included: unclear

Number of patients: Not stated for most studies.

Control groups: Some studies were not controlled. A variety of control interventions were used in the controlled studies including placebo, analgesics, TENS etc.

Main outcomes: blood flow; epidermal function; cardiovascular function; pulmonary gas exchange; composition of circulating blood; pain threshold; muscular tone; psychological variables.

Inclusion criteria were: assessment of the effectiveness of massage treatment using defined endpoints; adequate sample size; description of massage technique used and concomitant medications.

Medline (1970-1993) was searched for clinical trials using the terms massage and clinical trial. The reviewer's databases , textbooks on massage and physcial therapy and reference lists of retrieved reports were searched for additional citations. Experts were contacted for additional information. No language restrictions were made. A descriptive assessment was conducted.


Studies assessed the effects of massage in the treatment of a number of different medical conditions. Some positive effect of vibrational or manual massage was reported as improvements in mobility, doppler flow, expiratory volume and reduced lymphoedema in controlled studies. Positive trends for effectiveness of massage were also shown for musculoskeletal and phantom limb pain, but not cancer pain in controlled studies. Uncontrolled studies were invariably positive.

Adverse effects

Thrombophlebitis and local inflammation or ulceration of the skin were the most frequently documented adverse effects in textbooks. No other information was provided.



As stated by the reviewer, the bulk of the literature on massage is methodologically flawed and open to bias, mainly because of a lack of randomisation, blinding and placebo controls. The studies were of poor quality and validity. In addition the studies were small and, therefore, unlikely to be capable of showing the true effects of treatment. At the time this review was published there was no sound scientific evidence to show whether massage therapy works. Over 15 randomised, blind trials and some more recent reviews addressing specific conditions (e.g. low back pain) have been published since this review and supercede some, but not all, of the information in this paper.

Further reading

Related topics

Massage for low back pain

Abdominal massage for chronic constipation

Massage for delayed onset muscle soreness

Effleurage backrub for relaxation .