A Meta-Analysis of Massage Therapy

Christopher A. Moyer, James Rounds, and James W. HuffmanUniversity of Illinois -Champaign

Viki Bedford, Radford University, 2004

Introduction Massage Therapy (MT) has long been

trying to gain empirical support as for too long it has been associated with illicit sex, erotica or pampering for the rich.

Empirical support of efficacy of massage therapy as a healing modality is required before it will be accepted by insurance companies.

In general interest in massage therapy has grown in both professional and consumer groups as a complementary medicine.

I chose this meta-analysis specifically because I am an insider--- a massage therapist since 1978.

Purpose of this meta-analysis Consumers visits to MT’s have increased 36%

between 1990 & 1997

Consumers are spending between $4 & $6 billion annually for MT (not normally paid for w/ insurance!)

Tooted benefits by MT’s are improved circulation, relaxation, feelings of well-being and reductions in anxiety and pain, with decreases in stress hormones (cortisol) and reduction of depression.

Purpose of this meta-analysis (con’t)

This meta-analysis seeks to determine how well these claims are justified.

It also looked for MT as a valid treatment for psychological problems that drugs are often used for treatment in combination w/ psychotherapy, even though it is more known as a more physical type of therapy.

The scope of this meta-analysis

Researchers were hired to look for massage, massotherapy, acupressure, accupressure, applied kinesiology, bodywork, musculoskeletal manipulation, reflexology, relaxation techniques, Rolfing, Touch Research Institute, and Trager. Those found were used to find other articles.

Mean effect sizes were calculated from 37 studies for 9 dependent variables, state anxiety, negative mood, pain, cortisol, blood pressure, heart rate,, trait anxiety, depression, delayed assessment of pain.

Variables and Measures

1. State anxiety2. Negative mood3. Immediate assessment of pain4. Cortisol5. Blood pressure6. Heart rate7. Trait anxiety8. Depression9. Delayed assessment of pain

The scope of this meta-analysis (con’t.)

The present study’s operational definition of massage therapy (MT) is defined “as the manual manipulation of soft tissue intended to promote health and well-being.”

The AMTA definition is broader: manual soft tissue manipulation [that] includes holding, causing movement, and/or applying pressure to the body” and massage therapy as “a profession in which the practitioner applies manual techniques, and may apply adjunctive therapies, with the intention of positively affecting the health and well-being of the client” (AMTA, 1999a)

Various places and ways for giving massages

More massage pix

The scope of this meta-analysis (con’t.)

Studies were limited to human, other than infants, in English; ---studies using mechanical devices, ice or hot stones, chiropractic, passive motion, etc, were excluded.

These criteria yielded 144 studies, 37 actually used as they had random assignment, w/control groups, which combined for a total of 1,802 participants, 795 received MT, 1,007 were either on wait lists or receiving placebo therapy.

Massage Theories Considered:

Gate Control Theory of Pain Reduction

Promotion of Parasympathetic Activity

Mechanical Effects

Promotion of Restorative Sleep

Interpersonal Attention, one-on-one style

Moderators

Whether the length effects the magnitude of results, minimal or optimal times have never been examined as a variable

Are results due to the characteristics of the therapist or recipient

What are “the laboratory effects” (most has been done at TRI-32% of studies)

How much does the therapists’ training or experience have on effects

Moderators (con’t.)

The six specific moderators examined were non-significant:

1. Minutes/sessions (was the only moderator that approached an alpha level for statistical significance)

2. Age3. Training4. Laboratory effects5. Comparison group types6. Percent of female participants

Predictions That MT will “promote significant and

desirable reductions” of their 9 dependent variables.

That greater reductions will be associated with higher doses.

That effects are not expected to vary by age or gender.

MT effects will be larger than those generated in the comparison groups

And no prediction was made concerning therapists’ training or laboratory effects.

Effects

Single-dose- The term used for one time sessions; only state-anxiety, blood pressure, heart rate were significant

Multiple-dose- repeated sessions over time; all outcome variables were significant within this category.

Statistical AnalysisOutcome variables k g (d) 95% CI

Overall 37 0.34** 0.21, 0.48

Single-dose effects

State anxiety 21 0.37** 0.14, 0.59

Negative Mood 8 0.34 -0.08, 0.76

Immediate pain 15 0.28 -0.01, 0.57

Cortisol 7 0.14 -0.10, 0.38

Blood pressure 5 0.25* 0.03, 0.48

Heart rate 6 0.41** 0.19, 0.62

Multiple-dose effects

Trait anxiety 7 0.75** 0.27, 1.22

Depression 10 0.62** 0.37, 0.88

Delayed pain 5 0.31** 0.10, 0.52

“Note: a positive g indicates a reduction for any outcome variable. *p<.05, **p<.01

Results

MT single-dose, didn’t exhibit effects on immediate assessment of pain, which contradicts the common belief MT may provide analgesia per the gate control theory.

64% experienced reduction of state anxiety, and 60% a reduction in BP, and 66% for HR reduction over comparison groups.

Results (con’t)

For multiple-doses: “despite the fact that MT did not demonstrate an effect on immediate assessment of pain, a significant effect was found for delayed assessment of pain”, lower weeks later by 62% over the comparison groups, which may be due to the theory of sleep improvement.

Reduction of trait anxiety and depression after a course of treatments were MT’s largest effects.

MT Theories Outcomes Mixed support for existing theories

Strangely the ones most supported by therapists were least supported in this meta-analysis.

Reduction of BP and HR do support that MT promotes a parasympathetic response.

Lasting pain relief may be due to mechanical theories, through promotion of circulation, improved sleep and breakdown of adhesions that cause pain.

MT Theories (con’t.) MT from a psychological perspective,

not previous considered as a theory for MT effectiveness:

MT’s benefits may be because they parallel the “common-factors model” of psychotherapy--- a client who expects therapy will help, a therapist who is warm and has positive regard for the client, and the development of a healing alliance are considered more important in psychotherapy than the specific concepts one follows.

Findings show MT has a magnitude of effects on trait anxiety and depression to what would be expected in psychotherapy.

Further comparisons of MT to psychotherapy

Average psychotherapy clients fare better than 79% of untreated clients.

Average MT client experienced reduction of 77% for anxiety and 73% reduction for depression than comparison groups.

Discussion

Facts were supportive that the amount of training was not a factor in the effects.

Both MT and psychotherapy rely on repeated private interpersonal contact between 2 people.

Most of these studies used an average 30 minute session, but these authors feel if that matched more closely to psychotherapy’s 50-mintwice weekly for 5 week span, that MT’s effects may have matched or exceeded those of psychotherapy.

References

A Meta-Analysis of Massage Therapy Research

Psychological Bulletin2004, Vol.130, No. 1, 3-18Copyright 2004 by the

American Psychological Association, Inc

Photos from random Google image search

for “massage”