From putting weight on premature infants to improving peak air flow in children with asthma, massage therapy has demonstrated impressive results. This pioneer touch researcher reviews the data and offers some intriguing hypotheses for how massage produces its effects.
By Tiffany M. Field, PhD
From putting weight on premature infants to improving peakair flow in children with asthma, massage therapy has demonstrated impressiveresults. This pioneer touch researcher reviews the data and offers someintriguing hypotheses for how massage produces its effects.
Massage therapy is one of the oldest forms of treatment in the world.It was first described in China during the second century BC, and Hippocratesthought it so central that he defined medicine as "the art of rubbing."Massage continued to be used widely in the US and abroad well into thiscentury, but it disappeared from the American medical scene in the 1940sat approximately the same time as the pharmaceutical revolution broughtabout by the introduction of antibiotics.
Today, massage is once again becoming popular, as part of the alternativemedicine movement. Recent studies suggest that massage, as complementarytherapy, can facilitate growth and development, enhance attentiveness andlearning, lessen pain, reduce stress, and alleviate symptoms of autoimmuneand immune disorders. Some of these studies will be reviewed here so thatpediatricians will be able to offer information to parents considering massagetherapy for their children.
Several studies have reported that preterm infants who were massagedgained more weight than those who were not. It is important to note thatthe Swedish massage used in these studies involved a fair degree of pressure;a review of the infant massage literature suggests that light stroking doesnot produce similar results. In one study, 20 preterm infants in the neonatalintensive care unit of the University of Miami Medical School were comparedwith 20 controls matchedfor birth weight and medical/ surgical condition.Infants in the treatment group were given 15-minute massages through theincubator portholes three times a day for 10 days by students trained inmassage technique.1The infants who were massaged gained 47% moreweight than controls (Figure 1) and had higher scores on the Brazelton NeonatalBehavior Assessment Scale. The hospital stay for the massaged infants was,on average, six days shorter than for controls, with cost savings from thereduced stay averaging $10,000 per infant.
At 1 year of age, the treated infants still weighed more than the controls.They also performed better on the Bayley Scales of Infant Development, withscores averaging 12 points higher on the mental scale and 13 points higheron the motor scale.2 One possible explanation for the higherscores is that the infants' more responsive behavior, as demonstrated bytheir higher Brazelton scores, elicited more stimulation from their parents,which led to the later gains in growth and development.
These results have been replicated in studies conducted in Israel3and the Philippines.4 In the Philippines study, which reproducedour University of Miami study exactly, the preterm infants who were massagedgained 45% more weight than infants who were not massaged. In the Israelistudy, massaged infants gained 31% more weight than controls. In addition,the mothers who provided the massage experienced a decrease in depression,as measured by their scores on a measurement scale for depression. Finally,a recent study by our group suggests that a 47% greater weight gain canbe achieved in preterm infants after only one week of massage therapy.
Massage is also beneficial for full-term infants. In one study, 40 full-terminfants 1 to 3 months of age were given 15 minutes of massage or rockingfor 12 days over a six-week period.5 The infants who were massagedspent more time in active alert and awake statesthan the infants who wererocked and had better habituation scores on the Brazelton Neonatal BehaviorScales. The massaged infants also cried less and had lower salivary cortisollevels. After the sessions, the massaged infants spent less time activeand awake than the babies who had been rocked, suggesting that massage mayinduce sleep more effectively than rocking does. By the end of the six-weektreatment period, infants in the massage group had gained more weight; scoredhigher on measures of emotionality, sociability, and soothability; sustainedface-to-face interactions for longer periods; and had lower levels of urinarystress hormones (cortisol) and catecholamines (norepinephrine, epinephrine)and higher levels of serotonin than infants who were rocked. Lower levelsof cortisol, norepinephrine, and epinephrine suggest lower levels of stressfollowing massage. Higher levels of serotonin are noted when infants' activitylevels and affect are less depressed.
Parents who want to know how to massage infants can learn how from Touch...AParent's Guide to Infant Massage, an illustrated brochure published by theJohnson & Johnson Pediatric Institute. For information on how to obtainfree copies, call the Institute at 877-JNJ-LINK. Some of the techniquesare illustrated in Figure 2.
Converging data suggest a possible mechanism for the relationship betweenmassage therapy and weight gain. Uvnas-Moberg and colleagues have reportedthat when newborns are stimulated on the inside of the mouth, increasedamounts of gastro-intestinal food-absorption hormones, including gastrinand insulin, are released.6My colleagues and I found that preterminfants who sucked a pacifier during gavage feedings gained significantlymore weight than controls.7 Stimulating the entire body, as massagetherapy does, leads to increased vagal activity,1 which facilitatesthe release of food-absorption hormones such as insulin.6 Themassaged infants in the studies described here did not take in more caloriesthan controls, nor did they sleep more and thus expend less energy. Rather,their weight gain seems to have been mediated by this increase in vagalactivity.
The quicker habituation (an index of newborn memory) noted in the massagedbabies at the neonatal period and their superior performance on the Bayleymental scale at 1 year of age (also related to memory) may derive from enhancedhippocampal development, according to findings from animal studies.8A neonatologist at Hammersmith Hospital in London is currently studyingMRIs of massaged preterm infants to determine whether hippocampal changesalso occur in humans.9
The potential of massage to relieve chronic pain has been studied inchildren with juvenile rheumatoid arthritis (JRA). Parents were taught toadminister massage therapy and asked to provide it for 15 minutes everynight as part of their children's bedtime ritual. Having parents providethe massage kept costs down and, researchers hypothesized, might benefitparents as well as children. For more information on parent-administeredmassage, see "Tips for Parents" box.
In the JRA study, children who received bedtime massage for a month werecompared with controls who received progressive muscle relaxation trainingfrom their parents.10 The children who were massaged reportedless anxiety and had lower cortisol levels than the control group in measurementstaken after the first and last sessions. In addition, according to reportsfrom parents and the children's physicians, the massaged children had lesspain and fewer limitations on activity than the controls.
A variety of mechanisms may account for these results. According to gatetheory,11 pressure alleviates pain because pain fibers are shorterand less myelinated than pressure receptors. Pressure stimuli are receivedbefore the pain stimulus, the gate is closed, and thus the pain stimulusis not processed.
Another possible explanation points to an increase in serotonin levels(5HIAA in urine) detected after massage therapy in both infants5and adults.12 Serotonergic drugs alleviate pain, so it wouldnot be surprising if the body's naturally produced serotonin did also.
Finally, massage may alleviate pain because it allows the patient tohave longer periods of restorative deep sleep. When my colleagues and Istudied the effects of massage therapy on JRA10 and fibromyalgia,13we noted improved sleep. We speculated that low levels of deep sleep mayresult in pain because they are associated with low levels of somatostatin,which is an anti-pain substance. We also noted that patients who are deprivedof deep sleep release substance P, which causes pain, and hypothesized thatthis combination of less somatostatin and more substance P in sleep-deprivedpatients may be factors that contribute to the pain associated with fibromyalgiasyndrome.
Inattention is a salient problem for children with autism. Although suchchildren have been described as extremely sensitive to touch and as typicallydisliking being touched, in a recent study they showed surprisingly littleresistance to being massaged by massage therapists who were strangers tothem.14 The children in the study may have been able to toleratemassage because it is predictable, unlike the social touch they often resist,and because it uses pressure rather than light touch.
The children in the study were massaged with firm, moderate pressurefor 30 minutes twice a week for five weeks. The control group of childrenwith autism were held by their teachers and shown objects for the same amountof time. By the end of the study, the children who were massaged showedless "off-task" behavior in the classroom than the controls, accordingto observers who were blinded to the experimental condition of the subjects.In addition, they were more able to relate to their teachers and showedfewer stereotypic behaviors.
Children with attention deficit hyperactivity disorder (ADHD) also haveproblems staying on task in the classroom. In a recent study, adolescentswith ADHD were given firm, moderate-pressure massage therapy or relaxationtherapy for 30 minutes twice a week for five weeks.15 When thesessions were over, teens in the massage group were less fidgety than teensin the relaxation group. In addition, teachers' ratings on the Conners Scaleshowed that the massage therapy subjects spent significantly more time payingattention to their tasks and were less hyperactive than the controls. Theteachers were not aware of the students' status as experimental or controlsubjects.
A recent study on massage for medical school faculty and staff providesa possible explanation for enhanced alertness and attentiveness after massage.The subjects in this study were given 15-minute chair massages during lunchbreaks, including deep pressure applied to the head, neck, shoulders, andback.16 Surprisingly, instead of becoming sleepy, participantssaid they felt more alert. Some said it felt like a "runner's high."EEG recordings before, during, and after the massage sessions confirmedthese impressions. Levels of alpha wave activity were significantly lowerduring massage (an increase in alpha levels typically occurs during sleep),suggesting heightened alertness. A math computation task was added to thestudy to determine whether the EEG pattern of heightened alertness translatedinto performance. Computation time was significantly reduced and accuracyincreased after massage. This result may have been produced by increasedparasympathetic (vagal) activity, which follows massage therapy and is typicallycorrelated with attentiveness.17
Massage therapy has been studied as a way to reduce symptoms of stress-relatedpsychiatric conditions in children. Posttraumatic stress disorder (PTSD)is an example. In one study, a group of children whose neighborhoods hadbeen devastated by Hurricane Andrew showed posttraumatic stress symptomsof depression, and several had exacerbations of acting out behavior in theclassroom.18 In that study, one group of children exposed toAndrew was given 20-minute sessions every week by massage therapists, whilechildren in a control group watched a relaxing video while sitting on atherapist's lap. After two months of therapy, the massage group had fewerPTSD symptoms and less depression than the controls. Anxiety also decreasedin the massage group, and self-image as reflected in the children's drawingsimproved. For example, one girl drew herself on the first day of the studyas a very small figure in dark colors and without facial features. By thelast day of the study, the same child drew herself at a birthday party withballoons, sunshine and birds, and friends attending the party.
Another study explored the effects of massage on children and adolescentshospitalized for depression.19 One group of subjects receivedback massages for a week, while a control group viewed relaxing videotapes.At the end of the one-week study, the massaged patients were less depressedand anxious than the controls, and their stress hormone levels (saliva cortisol,urinary cortisol, and norepinephrine) were lower. Time-lapse videotapesof sleep/wake behavior revealed more organized sleep patterns in the massagegroup. In addition, nurses on the unit rated the patients who were gettingmassage as less anxious and more cooperative by the end of the week of massagethan they had been at the beginning of the week.
Adolescents with eating disorders including bulimia and anorexia alsoexperience severe depression. One study compared adolescents with bulimiawho received one month of twice weekly massages plus standard daily grouptherapy treatment with bulimic adolescents who received group therapy only.20The results showed less depression and anxiety, and lower stress hormonelevels (urinary cortisol), in the massaged patients than in the controls.Eating habits improved and body image was less distorted. In a similar study,adolescents with anorexia reported less anxiety and had lower cortisol levelsthan control patients who were not massaged. Over the one-month treatmentperiod, patients in the massage group reported less body dissatisfactionon the eating disorder inventory and had higher dopamine levels than controls.
In all of these studies, depressed mood, anxiety, and stress hormonelevels (norepinephrine, epinephrine, cortisol) declined. A number of mechanismsmight account for these effects. A recent study measuring frontal EEG activationwhen depressed adolescent mothers were massaged suggests one possibility.21In both the mothers and their infants, shifts to a more positive mood aftermassage accompanied shifts from right frontal EEG activation (normally associatedwith sad affect) to left frontal EEG activation (normally associated withhappy affect), or at least to symmetry (midway between sad and happy affect).Right frontal EEG activation, which has been noted in depressed adults andadolescents, shifted toward symmetry following a 20-minute massage. Chemicaland electro-physiologic changes from a negative to a positive balance mayunderlie the decrease in depression noted following massage therapy.
A related mechanism may be the increase in vagal activity following massagetherapy.22 The nucleus ambiguous branch of the vagus nerve stimulatesfacial expressions and vocalizations. More animated facial expressions andvocalizations provide feedback that reduces feelings of depression.17
The effects of massage therapy on a number of severe illnesses have beenstudied, including research on children with diabetes, asthma, HIV infection,and leukemia. The diabetes study grew out of earlier work that demonstratedincreased insulin levels after massage.23 We used parents astherapists for the same reasons as in the JRA study: to keep costs low andto assess the effects on the parents. Many parents have negative feelingsabout having to monitor their children's restrictive diets, take blood samples,and give insulin shots. We hoped that giving children a bedtime massagewould be a positive experience. Right after the massage therapy sessions,we found, both parents and children reported feeling less anxious and depressed,and our own observations confirmed this. At the end of a month, parentssaid they felt less stressed about managing their child's illness, and thechildren's blood glucose levels decreased from very high (158) to normal(118). We do not know what mechanism produced this change.
To see whether massage therapy might have similar benefits for asthmaticchildren, we studied 32 children with asthma randomly assigned to receiveeither massage therapy or relaxation therapy from their parents for 20 minutesbefore bedtime over a one-month period.24 Immediately after themassage, parents and children in the massage group felt less anxious andhappier and had lower cortisol levels than the relaxation group. Most important,peak air flow and other measures of pulmonary function showed significantlygreater improvement over the course of the study for children in the massagegroup than for those in the relaxation group.
Preliminary unpublished data on HIVpositive adolescents also suggestsome important effects of massage therapy. We assessed 20 HIVpositiveadolescents for immune changes, stress hormone levels, depression, mood,and anxiety after 12 weeks of massage or relaxation therapy. Subjects whoreceived 20 minutes of chair massage twice a week for 12 weeks felt lessanxious, less depressed, and in a better mood than the control group. Theircortisol levels were lower, and the cytotoxicity and number of natural killer(NK) cells had increased. In addition, the CD4 count of the massage groupincreased. These findings are consistent with our published data on HIV-positiveadults25 and not surprising, since cortisol is reputed to destroyNK cells and massage therapy has decreased cortisol levels in all our studies.
Because NK cells are the front line of defense in the immune system,combating the growth and proliferation of viral cells, HIV patients whoreceive massage therapy may get fewer opportunistic infections such as pneumonia.The increased NK cells also suggest that cancer patients may benefit frommassage therapy, inasmuch as NK cells also combat cancer cells.
The research reviewed here is summarized in the table. The results indicatethat massage therapy can produce clinical improvements in a number of specificconditions: better peak flow in asthma patients, lower glucose levels indiabetic patients, and faster weight gain in premature infants, for example.The research also identifies certain effects that are common across allthe studies: decreases in anxiety, depression, stress hormones (cortisol),and catecholamines. This stress-reduction effect may be important in alleviatingsymptoms in stress-related disease.
The underlying mechanism for the effects of massage may be increasedparasympathetic activity. The pressure associated with massage appears toincrease vagal activity, which in turn lowers physiologic arousal and stresshormone levels. Pressure is critical, because light stroking is generallyaversive (like being tickled) and has not been associated with beneficialeffects. Decreased cortisol leads to enhanced immune function. Parasympatheticactivity is also associated with increased alertness and better performanceon cognitive tasks.17
Where does this review leave the pediatric clinician? Clearly, some questionsarise that can only be answered by further research: Are treatment benefitslasting, or do they cease when therapy ends? What are the contraindications,if any? Exactly which techniques are most effective?
Regarding the duration of effects, there is no reason to suppose thatthe effects of massage would persist after termination of treatment anymore than the effects of diet, exercise, and many drugs do. Contraindications,if any, remain a matter of controversy.
Nevertheless, the findings accumulated thus far point to a valid rolefor massage therapy for children. While American attitudes toward caretakerswho touch children are generally hostile, therapeutic touch is beginningto gain favor here as part of the general pursuit of health and wellnessand a more open attitude toward alternative medicine.
I would like to thank the parents, infants and children who participatedin these studies and my colleagues who collaborated on the research. Thisresearch was supported by the National Institute of Mental Health, the NationalInstitute of Child Health and Human Developent, the National Institute onDrug Abuse, and by Johnson & Johnson.
THE AUTHOR is Director, Touch Research Institutes, University of MiamiSchool of Medicine and Nova Southeastern University, Miami, FL. Establishedin 1992, the Touch Research Institutes conduct research on touch and movementtherapies.
1. Field T, Schanberg SM, Scafidi F, et al: Tactile/ kinesthetic stimulationeffects on preterm neonates. Pediatrics 1986;77:654
2. Field T, Scafidi F, Schanberg S: Massage of preterm newborns to improvegrowth and development. Pediatric Nursing 1987;13:385
3. Goldstein-Ferber S: Massaging premature infants, paper presented atChild Development Conference, Bar-Ilan University, Israel, 1998
4. Jinon S: The effect of infant massage on growth of preterm infant,in Yabes-Almirante C, deLuna MG (eds): Increasingly Safe and SuccessfulPregnancy. Netherlands, Elsevier Science, 1996, pp 265269
5. Field T, Grizzle N, Scafidi, F, et al: Massage therapy for infantsof depressed mothers. Infant Behav Devel 1996;19:109
6.Uvnas-Moberg K, Widstrom AM, Marchini G: Release of GI hormone in mothersand infants by sensory stimulation. Acta Paediatr Scand 1987;76:851
7. Field T, Ignatoff E, Stringer S, et al: Nonnutritive sucking duringtube feedings: Effects on preterm neonates in an ICU. Pediatrics 1982;70(3):381
8. Meaney MJ, Aitken DH, Bhatnagar M, et al: Neonatal handling and thedevelopment of the adrenocortical response to stress, in Gunzenhauser N,Brazelton TB, Field T (eds): Advances in Touch. Skillman, NJ, Johnson &Johnson, 1990
9. Modi N, Glover J: Massage therapy for preterm infants. Paper presentedat Touch Research Symposium, Providence, Rhode Island, 1996.
10. Field T, Sunshine W, Hernandez-Reif M, et al: Juvenile rheumatoidarthritis: Benefits from massage therapy. J Ped Psychol 1997;22:607
11. Melzack R, Wall PD: Pain mechanisms: A new theory. Science 1965;150:971
12. Hernandez-Reif M, Field T, Theakston H: Multiple Sclerosis patientsbenefit from massage therapy. J Bodywork Movement Therapies 1998;2:168
13. Sunshine W, Field T, Schanberg S, et al: Massage therapy and transcutaneouselectrical stimulation effects on fibromyalgia. J Clin Rheumatol 1997;2:18
14. Field T, Lasko D, Mundy P, et al: Autistic children's attentivenessand responsivity improved after touch therapy.J Autism Dev Disorders 1997;27:329
15. Field T, Quintino O, Hernandez-Reif M: Adolescents with attentiondeficit hyperactivity disorder benefit from massage therapy. Adolescence1998;33:103
16. Field T, Ironson G, Pickens J, et al: Massage therapy reduces anxietyand enhances EEG pattern of alertness and math computations. Int J Neurosci1996;86:197
17. Porges SW: Emotion: An evolutionary by-product of the neural regulationof the autonomic nervous system. Ann NY Acad Sci 1997;807:62
18. Field T, Seligman S, Scafidi F, et al: Alleviating posttraumaticstress in children following Hurricane Andrew. J Applied Dev Psychol 1996;17:37
19. Field T, Morrow C, Valdeon C, et al: Massage reduces anxiety in childand adolescent psychiatric patients. J Amer Acad Child Adolesc Psychiatry1992;31:124
20. Field T, Schanberg S, Kuhn C, et al: Bulimic adolescents benefitfrom massage therapy. Adolescence 1998;33:555
21. Jones NA, Field T, Davalos M: Massage attenuates right frontal EEGasymmetry in one-month-old infants of depressed mothers. Infant Behav Dev1998;21:527
22. Field T, Schanberg S: Massage alters growth and catecholamine productionin preterm newborns. Neonatal/ Infant Stimulation 1995;96:105
23. Field T, Hernandez-Reif M, LaGreca A, et al: Massage therapy lowersblood glucose levels in children with diabetes mellitus. Diabetes Spectrum1997;10:237
24. Field T, Henteleff T, Hernandez-Reif M, et al: Children with asthmahave improved pulmonary function after massage therapy. J Pediatrics 1998;132:854
25. Ironson G, Field T, Scafidi F, et al: Massage therapy is associatedwith enhancement of the immune systems cytotoxic capacity. Int J Neuroscience1996;84:205