Massage Therapy Effects Tiffany M. Field Touch Research Institute, University of Miami School of Medicine, and Nova Southeastern University Massage therapy is older than recorded time, and rubbing was the primary form of medicine until the pharmaceutical revolution of the 1940s. Popularized again as part of the alternative medicine movement, massage therapy has recently received empirical support for facilitating growth, reducing pain, increasing alertness, diminishing depression, and enhancing immune function. In this article studies are reviewed that document these effects, and models are proposed for potential underlying mechanisms. Massage therapy is one of the oldest forms of treatment in the world, having first been described in China during the second century B.C. and soon after in India and Egypt. Hippocrates, in 400 B.C., defined medicine as "the art of rubbing." Massage therapy disappeared from the American medical scene at approximately the time of the pharmaceutical revolution of the 1940s. Now considered an "alternative" therapy, it is becoming popular again as part of the alternative medicine movement. At this time, it is commonly defined by massage therapists as the manipulation of soft tissue by trained therapists for therapeutic purposes. Despite its long history and popularity, a Medline search yielded only approximately 200 articles from the last 30 years. Much of this literature suffers from classic methodological problems. First, although the literature has focused on clinical conditions, very few studies are based on clinical trials. Typical sampling problems are the failure to include control groups and the lack of random assignment to treatment and control conditions. Often, the participants have served as their own controls, and the measures were simply collected at the beginning and the end of the treatment period. Although within subject controls are important in controlling for individual differences, treatment individuals need to be compared with nontreatment or comparison-treatment individuals. Using a within-subjects design alone could result in effects that might be explained otherwise by spontaneous recovery, a placebo effect, and statistical regression. The control group would optimally be an attention control or a comparison-treatment control to avoid the possibility that the therapist's attention alone explained the effects. A second problem is the very small sample sizes used in most of the studies and that the treatment group often received more than one type of treatment. Another problem is the potential for initial level effects where the treatment may have differential effects depending on the initial level of the participants. This, in itself, could explain many of the mixed findings. Still another problem is the use of inappropriate statistics. Only two metaanalyses appear in the literature. This has occurred because there are not enough studies with comparable designs and standards, and too many different massage therapy techniques have been used across the studies. Very few replications and virtually no follow-up studies have been conducted. Although one might not expect massage therapy to have sustained effects any more than temporary drugs, diet, or exercise, follow-ups are needed to assess that question. Finally, there would appear to be a publication bias where inevitably positive results are published, and negative results lay idle. For the above reasons, I have conducted, along with colleagues at the Touch Research Institute, a number of massage therapy studies, focusing on a variety of different conditions that might benefit from massage therapy. In each of these studies there was a theoretical reason to expect positive results. In addition, other studies from the literature were included in this review if they met the criteria of adequate sample size determined by power analysis and random assignment to a treatment group, a comparison treatment, or an attention control group. The massage therapy technique used throughout all of these studies, unless otherwise specified, involved deep tissue manipulation with presumed stimulation of pressure receptors. Unless otherwise specified, adult studies involved eight sessions (two per week for four consecutive weeks) of 30 minutes duration, and all child sessions were performed by parents on a nightly basis for 30 days for 15 minutes duration. The parameters for the adult sessions were based on practical considerations, namely that most adults could not afford more than two half-hour sessions per week by a professional or by a significant other. The parameters of the child sessions were based on the consideration that children with chronic illness may benefit from a "daily dose" and Editor's note. Melissa G. Warren served as action editor