Twenty studies assessing 34 comparisons were included (n=1,341 participants). Study sample sizes ranged from nine to 262. Quality was limited, quality scores out of 10 ranged from 0 to 3 for studies using healthy participants, and from 3 to 7 for clinical studies. Studies using healthy participants also had smaller sample sizes. Most studies were underpowered. Details and rationale of the massage interventions were rarely reported. There was no difference in methodological scores between studies with a positive outcome regarding pain/soreness and those with a negative outcome.
Therapeutic massage was judged to be effective at reducing pain/soreness compared to control in half of the 20 studies (four of nine studies in healthy patients with post-exercise pain, six of 11 studies in patients with musculoskeletal pain, three of seven studies in patients with low back pain).
Therapeutic massage was better than no treatment in five of 10 studies (four of nine studies in healthy patients with post-exercise pain, one study in patients with shoulder pain receiving six 20 minute massage sessions over two weeks).
Therapeutic massage was better than active treatment in seven of 22 comparisons (active interventions included self care educational material, traditional Chinese acupuncture, exercise and education, muscle relaxation, mental relaxation, and standard care).
There was no difference in the total amount of massage given or in the amount of massage administered per study day when comparing studies with positive or negative outcomes with respect to pain relief. However, median dosing per study day was eight minutes and lack of effect may have been due to insufficient dosing despite the inability to find a relationship between dosing and outcome.