INTER NA TIO NA L JO U R NAL O F BE H A VIO R A L D EVELO PME NT, 1998, 22 (4), 779±797

Touch Therapy Effects on D evelopment

Tiffany M. Field
University of M iam i School of M edicine, USA

In this review empirical data are presented on the use of touch therapy,
speci® cally massage therapy for improving the clinical course of several
conditions including growth and development of pre-term infants, reducing
pain, increasing attentiveness, diminishing depression, and enhancing
immune function. Potential unde rlying mechanisms for the massage therapy
effects are propose d for each of these conditions. The general effect appears
to derive from the stimulation of pressure receptors and the ensuing increase
in vagal activity and slowing physiology which in turn facilitates a more
relaxed behavioural state, effects a decrease in stress hormones, most
particularly cortisol, and an increase in immune function, particularly natural
killer cells.

Touch therapies, such as massage therapy, are proba bly older than
recorded time, dating back at least to 1800 BC . They were the primary form
of medicine until the pharmaceutical revolution of the 1940s. Popularised
again as part of the alternative medicine movement, touch therapy,
speci® cally massage therapy, has recently received empirical support for
facilitating growt h, pain reduction, increasing alertness, diminishing
depression, and enhancing immune function. A s such, touch therapy can
affect the course of cognitive, emotional, and physical development. The
massage therapy used in these studies involve s rubbing and stroking
different parts of the body with some pressure and usually with oil. The
purpose of this report is to review studies that docum ent the massage
therapy effects, and models for potential unde rlying mechanisms. First
studies on the effects of massage therapy on the growt h and development

R equests for reprints should be sent to D r Tiffany Field, Touch R esearch Institute,
U niversity of Miami School of Medicine, P.O. Box 016820, Miami, FL 33101, U SA .
We would like to thank the parents, infants and children who participat ed in these studies
and our colleagues who collaborat ed on the research. This research was supporte d by an
NIMH R esearch Scientist A ward (No. MH 00331) and an NIMH R esearch G rant (No.
MH 46586) to Tiffany Field and funding from Johnson and Johnson.

c 1998 The International Society for the Study of Behavioura l D evelopment

780 FIELD

of pre-term and full-term infa nts are reviewed. This section is followed by
a review of pain reduction studies. The subsequent discussion includes
massage therapy effects on attentiveness in autism, attention de® cit
hypera ctivity disorder and in the norm al workpla ce. R educing depression
and stress horm ones is the topic of the subsequent section and ® nally,
studies are reviewed on the role of massage therapy in autoimmune
disorde rs including asthma, diabetes, and immune function of H IV
patients. The paper conclude s with a potential explanatory mecha nism
for the similar effects noted for massage therapy in many clinical
conditions.

GROWTH AND DEVELOPMENT

D ata from research on rats and monke y models support the use of touch as
therapy. In a recent model developed by Saul Schanberg and colleagues,
rat pups were ® rst removed from their mother to investigate touch
deprivation. Then, the mother’ s behaviour was simulated to restore the
physiology and biochemistry of the rat pups to nor mal. In several studies, a
decrease was noted in growt h horm one (ornit hine decarboxylase) when
the pups were removed from their mother (Schanberg & Field, 1987). This
decrease was observed in all body orga ns including heart and liver, and in
all parts of the brain including the cerebru m, cerebellum, and brainstem.
These values returne d to norma l when the pups were stimulated using
technique s approxima ting the mother’ s behaviour. A graduate student/
animal caretaker observed rat mothers’ nocturna l behaviour and noted
that they frequently tongue lick, pinch, and carry around the rat pups.
When the researchers tried each of these manoeuvre s, only the tongue
licking (simulated by a paint brush dipped in water and briskly stroked all
over the body of the rat pup) restored the growt h horm one values to their
norma l level. More recently, Schanberg and his colleagues discovered a
near immediate gene unde rlying protein synthesis. This is a gene that
triggers growth when the orga nism is touche d that responds to tactile
stimulation, suggest ing genetic origins of this touch/gro wth relationship
(Schanberg, 1995).
R elated studies by Meaney and his colleagues (Meaney et al., 1990)
sugge st a long-t erm impact of handling on the modulation of cortisol
(stress hormon e) product ion. R ats who were handled more as pups showe d
less corticosteroid product ion, more elaborate dendritic arborisa tion in the
hippoca mpal region, and better maze performa nce (memory function) in
the ageing rat. Parallels have already been noted in human infants at the
H ammersmith H ospital, London. R esearch report ed by Modi and G lover
(1996) provides magnetic resonance imaging (MR I) docum entation of
 TOUCH THERAPY EFFECTS 781
faster development of the hippoca mpal region in pre-term infants who
were being massaged.
We have report ed greater weight gain in several studies on pre-term
infants receiving massage therapy, including those who are cocaine-
expose d (Wheeden et al., 1993) and H IV-exposed (Sca® di et al., 1996). In
those studies, Swedish massage with some pressure was used. A lthough
infants, particularly premature infants, may seem to be fragile, some
pressure is needed for the massage to be effective. In our review of the
infant massage litera ture, we found that those who used light stroking did
not report weight gain, for example, whereas those who used stroking with
pressure reported weight gain (Sca® di et al., 1986). The pressure element
appeared to be necessary for touch therapy to positively affect weight gain.

Pre-term Infants
In our series of studies pre-term infants in the neonatal intensive care unit
were given 15-minut e massages three times a day for 10 days while they
were still in an incubator (massaged through the incubator porthole s).
H owever, when they were no longer in medical danger and were simply in
the ``growe r’ ’ nurse ry to gain weight (Field et al., 1986; Sca® di et al., 1990),
the treated infants, compared to control infants gained 47% more weight
and were hospitalised for 6 days less at a hospital cost saving of U S$10,000
per infant. These ® ndings sugge st that at least 4.7 billion dollars in hospital
costs could be saved per year by massaging the 470,000 pre-term infants
born each year. E levated norepinephrine and epinephrine levels in these
massaged infants sugge sted that massage therapy was facilitating the
norma l developm ental increase in these catecholamines at this stage in life
(Kuhn et al., 1991). Finally, the treated infants performe d better on the
Brazelton Neonatal Behavior A ssessment Scales including the orientation
and motor scales.
A t one year, the treated infants were still showing a weight advantage
and they also performe d better on the Bayley Scales of Infant
D evelopment (Field, Sca® di, & Schanberg, 1987). O n the mental scale
their scores averaged 12 points higher (101 vs. 89) and on the motor scale
15 points higher (105 vs. 90). The infants’ more responsive behaviour on
the Brazelton Neonat al Behavior A ssessment apparently elicited more
stimulation from their parents which led to the later gains in growt h and
development. The weight gain data has been replicated by at least two
other groups of researchers (G oldst ein-Ferber, 1997; Jinon, 1996).

Cocaine-exposed Pre-term Infants

Similar weight gains were noted in our study on cocaine-exposed infants
(Wheeden et al., 1993). In addition, these infants showe d superior motor
782
TABLE 1
Effects of Touch Therapy on Growth and Development
Studies G rowth D evelopment Stress Other
H ormones
1±8 1 1
Pre-term infants 47% . wt. gain . Brazelton ¯ Cortisol6 ­ Vagal activity7
2 5
31% . wt. gain . Bayley scores 1 year ­ Insulin 7
3
. wt. gain ­ Catecholamines8
4
. wt. gain
9 9 9
Cocaine-exposed pre-term infants . wt. gain . Brazelton motor scores
10 10 10
H IV-exposed full-term infants . wt. gain , Stress behaviour
11 11 11
Normal full-term infants . wt. gain , Irritability ¯ Cortisol11 ­ Serotonin 11
11
. Temperament
11
. Interactions
­ Sleep 11
1 2 3 4 5 6 7 8
Field et al., 1986; Sca® di et al., 1990; Jinon, 1996; G oldstein-Ferber, 1997; Field et al., 1987; Acolet et al., 1993; Sca® di et al., 1993; Kuhn
9 10 11
et al., 1991; Wheeden et al., 1993; Sca® di & Field, 1996; Field et al., 1996.
 TOUCH THERAPY EFFECTS 783
behaviour on the Brazelton scale. In attempting to unde rstand the
mechanism unde rlying the massage therapy/weight gain relationship we
assessed vagal tone using Por ges’ methods. The vagus (one of the 12
cranial nerves) slows the heart during tasks that require attention, and it
enhances gastric motility and facilitates the release of food absorpt ion
hormon es like insulin. Vagal activity and insulin levels increased following
the massage therapy (Sca® di et al., 1996). The massaged infants ate less
and slept less, so they were not simply conserving calories. R ather, the
weight gain seems to have been mediated by an increase in vagal activity,
which in turn facilitated the release of food absorpt ion hormon es (at least
insulin).

HIV-exposed Full-term Infants

In a study on H IV-expose d infants, the mothers of the infants were used as
the massage therapists. The mothers’ treatment compliance rates were
very high, perhaps because of the guilt they expressed for having possibly
transmitted H IV to their infants and their own high anxiety levels (Sca® di
& Field, 1996). Teaching parents to massage their infants often lowers
their anxiety levels related to their feeling helpless about their infants’ or
their child’ s condition. H elping with their children’ s treatment might be
expected to decrease their anxiety levels and make them feel that they
were contributing positively to the treatment. In addition, daily massages
are economically feasible when the parents are used as therapists. In the
study on H IV-expose d infants, the massaged infants’ weight gain was
signi® cantly greater and they showe d signi® cantly fewer stress behaviours.

Normal Full-term Infants

A lthough research on infant massage is limited, therapists sugge st that
massage: (1) reduces stress response s to painful procedures, such as
inoculations; (2) reduces pain associated with teething, constipation, and
colic; (3) helps indu ce sleep; (4) makes parents (who are typically the
massage therapists) ``feel good ’’ while they are massaging their infants; and
(5) facilitates parent-infant bonding and the development of warm,
positive parent-infant relationships (A uckett, 1981).
In a study of 40 full-term 1- to 3-month-old infants born to adolescent
mothers, given 15 minutes of either massage or rocking for 12 days over a
6-week period, the massaged versus the rocked infants: (1) spent more
time in active alert and active awake states; (2) cried less; (3) had lower
salivary cortisol levels, sugge sting lower stress levels; and (4) spent less
time in an active awake state and more time in sleep states after the
massage session (as oppose d to the rocking session), sugge sting that
784 FIELD

massage may be more effective than rocking for inducing sleep (Field,
G rizzle, Sca® di, A brams, & R ichardson, 1996a ). A nalyses of the longer-
term effects (Field et al., 1996) sugge sted that by the end of the 6-week
treatment period, the massage group infants; (1) gained more weight; (2)
improve d on emot ionality, sociability, and sootha bility temperament
dimensions; (3) showe d better face-to-face interaction behaviours; (4)
had decreased urinary stress hormon es (cortisol) and catecholamines
(nore pinephrine , epinephrine ); and (5) had increased serotonin levels.
The research by Schanbe rg and colleagues (Schanberg & Field, 1987)
presented earlier plus the work of U vnas-Moberg (U vna s-Moberg,
Widstrom, Marchini, & Windberg, 1987) in Sweden sugge sts a potential
unde rlying mechanism. Schanberg (1995) docume nted an increase in
protein synthesis. U vnas-Moberg and her colleagues (1987) report ed that
stimulating the inside of the mouth of the newborn (and the breast of the
mother) led to the increased release of gastrointestinal food absorpt ion
hormon es, including gastrin and insulin. Stimulating the entire body, as in
massage therapy would presumably lead to even higher levels of food
absorpt ion hormon es potentially mediated by vagal activity. Vagal
activity increases during massage therapy (Field, 1995), and the
vegetative branch of the vagus is known to stimulate the release of food
absorpt ion hormon es including insulin and gastrin (U vnas-Moberg, et al.,
1987). A ssays of (heelstick) blood samples of pre-term cocaine-expose d
infants yielded signi® cantly elevated insulin levels in those infants who
were massaged (Sca® di et al., 1996). E ven though the branch of the vagus
that slows down the heart (the typical measure of vagal activity) and the
branch that stimulates the release of food absorption horm ones are
different branches (the nucleus ambiguou s or the ``sma rt vagus’ ’ to the
heart and the dorsa l motor nucleus or the vegetative vagus to the
gastrointestinal tract), vagal activity and the release of food absorpt ion
hormon es appear to be related.
The superior habituation perform ance (an index of newbor n memory)
noted at the neonatal period and superior perfor mance on the Bayley
mental scale at one year (also related to infant memory skills) may derive
from enhanced hipp ocampa l development. A s noted earlier, Meaney and
his colleagues (1990) tracked a relationship between increased glucoco rti-
coids, decreased dendritic arborisa tion in the hippocampal region, and
inferior maze performa nce, sugge sting impaired memory function in the
ageing rats that had been deprive d of tactile stimulation as pups. Similarly,
the more elabora te dendritic arborisa tion noted in MR Is of the
hippoca mpal region in massaged pre-term neonates by Modi and
colleagues (Modi & G lover, 1996) may be related to the superior memory
perform ance not ed in the massaged newbor ns and their perform ance again
at one year.
 TOUCH THERAPY EFFECTS 785

PAIN REDUCTION DURING PAINFUL

PROCEDURES

Childbirth Labour
In many countries, pregnant women are massaged several times daily for
relaxation and to reduce their anxiety levels (O lder, 1982). In a study in
which we taught signi® cant others of pregnan t women to massage the
women during labour (Field et al., 1998c), we found reduced anxiety and
depression in the mothers and less need for medication and for Caesarean
section.

Therapy Prior to Debridement for Burn Patients

Massage therapy has also been used to reduce anticipatory anxiety prior to
debridement (skin bru shing for debris removal in severe burns) and to
indirectly alleviate pain during that procedure (Field, 1998d). A fter a
5-day course of 30-m inute massages prior to debridement patients
receiving massage therapy on the burn unit had lower anxiety levels and
associated decreases in stress horm ones (cortisol). Pain and depression
also decreased by day 5, probab ly because of the decrease in pain.

Juvenile Rheumatoid Arthritis

Chronic pain is a problem for children with juvenile rheumatoid arthritis.
A nti-in¯ ammatory agents used for their pain have ceiling effects, and
other drugs, such as narcotics, cannot be used due to their potentially
addictive effects. Thus, massage therapy is being assessed for its usefulne ss
for pain relief. In a study in which parents provide d these children daily
massages, several positive effects were noted (Field, Sunshine, H ernandez-
R eif, Krasnegor, & H enteleff, 1997b). The massaged children (vs. the
control children who received progressive muscle relaxation) had: (1)
decreased anxiety and cortisol after the ® rst and last sessions; and (2)
decreased pain over the 1-month period.

Fibromyalgia
In a study on ® brom yalgia syndrome (pain all over the body for no known
aetiology), subjects were random ly assigned to one of three conditions:
massage therapy, transcutaneous electrical stimulation (TE NS), a steel
roller the size of a pen that transmits a small barely discernible current as it
is rolled across the body; or transcutaneous electrical stimulation without
current (SH A M TE NS), for 30-m inute treatment sessions twice a week for
5 weeks (Sunshine et al., 1996). The massage therapy subjects reported
786
TABLE 2
Effects of Touch Therapy on Pain Reduction
Studies Pain A nxiety Depression Stress Other
H orm ones
1 1 1
Childbirth labour 1 ¯ ¯ ¯ ¯ Need for medication 1
¯ Need for a Caesarian 1
2 2 2 2
Debridement for burns ¯ ¯ ¯ ¯ Cortisol2
3 3 3
Juvenile rheumatoid arthritis ¯ ¯ ¯ Cortisol3
4 4 4 4
Fibromyalgia ¯ ¯ ¯ ¯ Cortisol4 ¯ Stiffness4
¯ Fatigue 4
¯ Sleep problems4
1
Field et al., 1998c; 2 Field et al., 1998d; 3 Field et al., 1997b; 4 Sunshine et al., 1996.
 TOUCH THERAPY EFFECTS 787
lower anxiety and depression, and their cortisol levels were lower
immediately after the therapy sessions on the ® rst and last days of the
study. The massage therapy group impro ved on a dolorimeter measure of
pain, report ed less pain, less stiffness and fatigue, and fewer nights of
dif® cult sleeping. Thus, massage therapy was the most effective therapy
with these ® brom yalgia patients.

Models for Mechanisms of Touch and Pain Relief

G ate T heory. Pain alleviation has most frequent ly been attributed to
the G ate Theory (Melzack & Wall, 1965). This theory sugge sts that pain
can be alleviated by pressure or cold temperature because pain ® bres are
short er and less myelinated than pressur e and cold temperature receptors.
The pressure or cold temperature stimuli are received before the pain
stimulus, the gate is closed and thus the pain stimulus is not processed.

Serotonin. A possible explanation for the alleviation of pain is

increased serotonin (i.e. moderate increases have been report ed in
seroton in levels after massage therapy for other conditions, Ironso n et
al., 1996). A n increase in the body’ s natural product ion of serotonin may
be superior to using the synthetic serotonergic pain relief drugs.

Sleep D e® cits. A nother potential theory for pain alleviation from

massage therapy relates to quiet or restorative sleep deprivation. D uring
deep sleep, somatostatin is norma lly released. Without this substance pain
is experienced. Substance P is also released in the absence of deep sleep,
and substance P is notable for causing pain. Thus, if a person is deprive d of
deep sleep they may have less somatostatin and increased substance P,
either or both causing greater pain. O ne of the leading theories for the pain
associated with ® brom yalgia syndrom e, for example, is the production of
substance P due to deep sleep deprivation (Sunshin e et al., 1996). In that
light, it is not surprising that the subjects with ® brom yalgia syndrom e in the
Sunshine et al. study experienced not only more deep sleep but less pain
following the massage therapy treatment period.

ENHANCING ATTENTIVENESS
Autism
O ne of the most salient proble ms of autistic children appears to be their
attentiveness to meaningless stimuli and their inattentiveness to stimuli
their teachers present . A lthough they have also been anecdotally described
as being extremely sensitive to touch and typically dislike being touched,
they showe d surprisingly little resistance to their being massaged. Massage
788 FIELD

TABLE 3
Effects of Touch Therapy on Enhancing Attentiveness
Studies A ttentiveness O ther
1 1 1
A utism ¯ Off-task behaviour ¯ A ttention to irrelevant stimuli
­ Social relatedn ess1
A ttention de® cit ­ On-task behaviour 2 ¯ Fidgeting 2
2
hyperactivit y disorder ¯ H yperactivity 2
3 3
E E G patterns of alertness ­ Alertness ­ E E G pattern alert ness
3
( ¯ alpha, ¯ beta, ­ delta waves)
­ Performance math computat ions 3
( ­ accuracy, ¯ time to complete)
¯ Job stress3
1 2 3
Field et al., 1996c; Field et al., 1998e; Field et al., 1996b.

may not have been aversive to them because it is pred ictable, unlike the
social stimulation they freque ntly resist. A lso, surprisingly, in our study on
autistic children, children who were massaged showe d a decrease in their
off-t ask behaviour in the classroom following a 10-da y period of massage,
their ``social relatedness’ ’ to their teachers improve d and they showe d
fewer autistic behaviours (orienting to sounds and stereotypic behaviours)
(Field et al., 1996c).

Attention Deficit Hyperactivity Disorder (ADHD)

Children with A D H D have similar proble ms staying on-task in the
classroom . In a recent study, A D H D adolescents were provide d massage
therapy or relaxation therapy for 10 consecutive school days (Field et al.,
1998e ). The massage therapy group, compared to the relaxation therapy
group, showe d less ® dgeting behaviour following the sessions. In addition,
after the 2-week period, their teachers (who were blind to the group
assignments) report ed that the children spent more time on-task and were
less hype ractive in the classroom .

EEG Patterns of Alertness

In a job stress study, medical schoo l faculty and staff received 15-m inute
chair massages during their lunch breaks (Field et al., 1996c). These
sessions involve d deep pressure in the head, neck, shoulde rs and back
regions. Surprisingly, instead of becoming sleepier after their midday
massage, the participants report ed experiencing heightened alertness,
much like a runne r’ s high. E lectroencephalogram (E EG ) recordings
before, during and after the massage sessions con® rmed the subjects’
impressions. A s compared to a group of relaxation therapy subjects, their
 TOUCH THERAPY EFFECTS 789
levels of alpha wave activity signi® cantly decreased during massage (in
contrast to alpha levels typically increasing during sleep). This decrease,
combined with increased delta and decreased beta waves sugge sted a
pattern of heightened alertness. A math compu tation task was added to
determine whether this E E G pattern of heightened alertness translated
into perform ance. Following the massage sessions the comput ation time
was signi® cantly reduced and comput ation accuracy increased, suggest ing
that 15-m inute massages during the lunch hour not only enhanced
alertness but also improve d cognitive perform ance.

ALLEVIATING STRESS, DEPRESSION, AND

ANXIETY
Post-traumatic Stress Disorder
This disorde r is characterised by depression and behaviour proble ms in
children. Many children showe d post-traumatic stress symptoms following
H urricane A ndre w, and several of the children’ s acting out proble ms in the
classroom were exacerbated by the hurricane (Field, Seligman, Sca® di, &
Schanberg, 1996d). A fter a month of massage therapy (twice a week) their
symptoms and their depression decreased in contrast to a control group
(who watched a relaxing video) . A nxiety also decreased and self-image
improve d as re¯ ected in their self-drawings. A girl’ s self-drawing, for
example, on the ® rst day was very small, had dark colour s, and no facial
features. By the last day, she drew a birthday party with balloons, sunshine,
birds, and friends attending the party.

Child and Adolescent Psychiatric Patients

Similarly, in a study of hospitalised depressed children and adolescents
(Field et al., 1992), those who received back massage s for a week versus
those who viewed relaxing videotapes (control group ), were less depressed
and anxious, had lower stress horm ones (lower saliva cortisol levels as well
as lower urinary cortisol and nore pinephrine levels) and their time-lapse
videotapes of their sleep/wake behaviour revealed less time awake and
more time asleep. In addition, the nurse s on the unit rated the massaged
adolescents as being less anxious and more co-ope rative by the last day of
the study.

Bulimia in Adolescent Girls

A dolescents with bulimia (ove reating and vomiting) experience severe
depression. Following one month of massage, bulimia adolescents had (1)
fewer symptoms of depression; (2) lower anxiety levels; and (3) lower
790 FIELD

stress horm ones (urina ry cortisol levels) (Field, 1998f) . Their eating habits
also improve d and their body image was less distorted.

Chronic Fatigue
D epression is also a proble m for adults with chronic fatigue syndrom e. In a
recent study with this popula tion, subjects were random ly assigned to a
massage therapy or a SH A M TE NS (see earlier) control grou p (Sunsh ine
et al., 1996). Immediately after the massage therapy and the SH A M TENS
sessions on the ® rst and last days of the study, the massage therapy subjects
had lower depression and anxiety scores and lower stress horm one
(salivary cortisol) levels. Longe r-term effects (last day vs. ® rst day)
indicated that the massage therapy group versus SH A M TE NS group were
experiencing fewer symptoms of depression as well as fewer somatic
symptoms, more hour s of sleep, lower urinary cortisol levels, and elevated
urinary dopa mine levels.

Grandparent Volunteers Massaging Infants

E lderly people can also be depressed, often due to social isolation, feelings
of less self-wort h, and touch deprivation. In a coun terbalanced design, we
compared grandpa rent volunt eers giving massage with their receiving
massage themselves (Field, 1998b). They alternatively gave infant
massages for a month and then received massages for a month. We were
surprise d to ® nd that the grandpare nts also bene® ted from giving the
massages. Their depression scores decreased followed a 1-mont h period of
massaging the infants and they experienced increased self-esteem and
decreased cortisol levels. The grandpa rent volunt eers bene® ted more from
giving than from receiving the massage. Their affect and self-esteem
improve d more, as did their lifestyle habits (e.g. they reported drinking
fewer cups of coffee per day, they made more social phone calls, and they
made fewer trips to the doctor’ s of® ce).

Models Underlying Touch-alleviating Depression

D epressed mood was decreased and anxiety levels and stress horm ones
(nore pinephrine , epinephrine , cortisol) were reduced in all of the
foregoing studies. O ne potential mechanism is sugge sted by a recent study
measuring front al E E G activation following massage in depressed
adolescents (Field & Jones, 1998). Shifts to a more posit ive mood were
notably accompanied by shifts from right front al E E G activation (norm ally
associated with sad affect) to left front al EE G activation (norm ally
associated with happy affect) or at least to symmetry (midway between sad
and happy affect). R ight front al E E G activation (noted in chro nically
 TABLE 4
Effects of Touch Therapy on Alleviating Stress, Depression, and Anxiety
Studies A nxiety Depression Stress O ther
Horm ones
1 1 1
Post-traumatic stress disorder ¯ ¯ ­ Self-esteem 1
1
. D rawings
1
, Behaviour problems classroom
2 2 2
Child and adolescent psychiatric patients ¯ ¯ ¯ Cortisol2 ­ Time asleep 2
2
¯ Norepinephrine 2 . Behaviour on unit
3 3 3 3
Bulimia in adolescent girls ¯ ¯ ¯ Cortisol3 . Body image
4 4 4 4 4
Chronic fatigue ¯ ¯ ¯ Cortisol . Sleep
­ D opamine 4
5 5 5
E lderly depressed volunteers massaging infants ¯ ¯ ¯ Cortisol5 ­ Self-esteem 5
5
, Coffee
5
. Social phone calls
5
, Trips to doctor
1 2 3 4 5
Field et al., 1996d; Field et al., 1992; Field et al., 1998f; Field et al., 1997c; Field et al., 1998b.

791
792 FIELD

depressed adults and also observed in the depressed mothers and infants in
our study) was shifted towards symmetry following a 20-m inute massage
(Field & Jones, 1998). Chemical and electrophysiological changes from a
negative to a positive balance may underlie the decrease in depression
noted following massage therapy.
A related potential mechanism may be the increase noted in vagal
activity following massage therapy. The nucleus ambiguou s branch of the
vagus (the ``smart vagus’ ’) stimulates facial expre ssions and vocalisations,
which would contribute to less depressed affect which in turn could
feedback to effect less depressed feelings.

AUTOIMMUNE DISORDERS
Children with Diabetes
In a study on the vagal activity and insulin levels following massage
therapy, we had reported (as described earlier) that both vagal activity and
insulin levels increased (Sca® di et al., 1996). This led us to investigate
massage therapy effects on diabetic children’ s clinical course . In this study
we used parents as therapists because we knew the cost of daily massages
was prohibitively expensive and because we knew that massaging children
had helped the therapists (the volunt eer grandpa rents) as well. Involve -
ment in the treatment of their children can be a particularly negative
experience for parents of diabetic children (e.g. monitoring dietary
compliance, taking blood samples, and giving insulin shots). In this study
parents were given a more positive role in their children’ s treatment by
massaging their children daily before bedtime (Field, H ernande z-R eif,
LaG reca, Shaw, Schanberg, & Kuhn, 1997a ). Immediately after the
massage therapy sessions, parents’ anxiety and depressed mood levels were
lower, and their children’ s anxiety levels and depressed mood levels were
lower. A t the end of the 1-month period, the parents’ assessment of their
children’ s insulin and food regulation improve d, and the children’ s blood
glucose levels decreased from 158 to 118 (ng/ml).

Children with Asthma

A ssuming that autoimmune disorde rs might similarly bene® t asthmatic
children, we had parents give daily 20-m inute bedtime massages to their
asthmatic children (Field et al., 1998a). Immediately after the massage: (1)
the parents’ anxiety decreased; (2) the children’ s self-reported anxiety
levels decreased; (3) their mood improve d; and (4) their cortisol levels
decreased. Most importantly, over the 1-mo nth period, they had
signi® cantly improve d pulmonary fun ctions including: peak air ¯ ow,
 TABLE 5
Effects of Touch Therapy on Autoimmune and Immune Disorders
Studies A nxiety D epression Stress Other
H ormones
A utoimm une disorders
1 1
Children with diabetes1 ¯ ¯ ¯ Parent anxiety and depression 1
¯ Blood glucose levels1

2 2 2
Children with asthma ¯ ¯ ¯ Cortisol2 ¯ Parent anxiety2
, A sthma attacks
­ Peak air ¯ ow, forced vital capacity, forced
expiratory volume, average ¯ ow rate, peak
expiratory ¯ ow rate

Im m une disorders
3 3 3 3 3
HIV positive adults ¯ ¯ ¯ Cortisol ­ NK cells and activity
¯ Catecholamines3
1 2 3
Field et al., 1997a; Field et al., 1998a; Ironson et al., 1996.

793
794 FIELD

forced vital capacity, forced expiratory volum e, average ¯ ow rate, and

peak expiratory ¯ ow rate.

IMMUNE DISORDERS
HIV positive adults
Immune disorde rs might be expected to bene® t from massage therapy
because of the decrease in cortisol levels noted in several previous studies.
E levated cortisol is known to dampen immune function. In a study we
conduct ed on H IV positive adults, natural killer (NK) cells and NK cell
cytotoxicity increased following 20 days of massage therapy (Ironso n et al.,
1996). Twenty-nine gay men (20 H IV positive, 9 H IV negative) were
massaged for 1 month. A subset of 11 of the H IV positive subjects served
as a within-subje cts control group (1 month with, and 1 month without
massages). Major immune ® ndings included a signi® cant increase in NK
cell num ber, NK cell cytotoxicity, and subsets of CD 8 cells. There was no
change in H IV disease progression markers (CD 4, CD 4/CD 8 ratios),
possibly because the H IV men were already severely immune compro-
mised. A signi® cant decrease was also noted in urinary cortisol and
nonsigni® cant trends showing decreased catecholamines. D ecreased
anxiety was signi® cantly correlated with increased NK cell num ber.
E levated stress hormon es (catecholamines and cortisol) are noted to
negatively affect immune function. The increase in cytotoxic capacity
associated with massage therapy proba bly derives from the decrease in
these stress horm ones following massage therapy. Because NK cells are
the front -line of defence in the immune system, combating the growt h and
proliferation of viral cells, the H IV patients who received the massage
therapy might experience fewer opport unistic infections such as pneum o-
nia and other viruses that often kill them. The increased NK cells also
sugge st that cancer patients may bene® t from massage therapy inasmuch as
NK cells are also noted to combat cancer cells.

CONCLUSION
The foregoing are the improve d functions noted following massage
therapy. In addition to each clinical condition being marked by unique
changes, such as the increased peak air ¯ ow noted in the asthma study and
the decreased glucose levels noted in the children with diabetes, there was
also a set of common ® ndings. A cross studies, decreases were noted in
anxiety, depression, stress horm ones (cortisol), and catecholamines.
Increased parasympathetic activity may be the unde rlying mechanism for
these changes. The pressur e stimulation associated with touch increases
vagal activity (Sca® di et al., 1996) which in turn lowers physiologica l
 TOUCH THERAPY EFFECTS 795
arousal and stress hormon es (cortisol levels). The pressure is critical
because light stroking is generally aversive (much like a tickle stimulus).
D ecreased cortisol in turn leads to enhanced immune function. Para-
sympathetic activity is also associated with increased alertness and better
perform ance on cognitive tasks. G iven that most diseases are exacerbated
by stress, and massage therapy alleviates stress, receiving massages should
proba bly be high on the health priority list along with diet and exercise.
Somewhat surprising was that the relatively short -term interventions
(nurt ure) in these studies apparently had positive effects on physiologica l
processes that are assumed to be dif® cult to change (nature). This
apparently occurs because tactile and pressure stimulation has a major
impact on the central and autonomic nervous systems. For that reason, any
long-t erm touch therapy would be expected to affect the course of
physical, cognitive, and emotional development.

Manuscript received July 1998

REFERENCES
A colet, D., Modi, N., G iannakoulo poulos, X., Bond, C., Weg, W., Clow, A . et al.
(1993). Changes in plasma cortisol and catecholam ine concentrat ions in response to
massage in pret erm infants. A rchives of Disease in Childhood , 68, 29±31.
A uckett, A.D. (1981). B aby m assage. New York: Newmarke t Press.
Field, T. (1995). T ouch in early developm ent. Hillsdale, NJ: E rlbaum.
Field, T., G rizzle, N., Sca® di, F., A brams, S., & R ichardson, S. (1996a). Massage therapy for
infants of depressed mothers. Infant B ehavior and D evelopm ent, 19, 109±114.
Field, T., Henteleff, T., H ernandez-R eif, M., Martinez, E ., Mavunda, K., Kuhn, C. et
al. (1998a). Asthmatic children have improved pulmonary functions after massage
therapy. Journal of Pediatrics, 132, 854±858.
Field, T., Hernande z-R eif, M., LaG reca, A ., Shaw, K., Schanberg, S., & Kuhn, C. (1997a).
Massage therapy lowers blood glucose levels in children with diabetes mellitus. Diabetes
Spectrum , 10, 237±239.
Field, T., H ernandez-R eif, M., Q uintino, O ., Wheeden, C., Schanberg, S., & Kuhn,
C. (1998b). E lder retired volunteers bene® t from giving massage therapy to infants.
Journal of A pplied Gerontology , 17, 229±239.
Field, T., H ernandez-R eif, M., Taylor, S., Q uintino, O ., Burman, I., Kuhn, C. et al.
(1997b). Labor pain is reduced by massage therapy. Journal of Psychosom atic O bstetrics
and G ynecology , 18, 286±291
Field, T., Ironson, G ., Pickens, J., Nawrocki, T., Fox, N., Sca® di, F. et al. (1996b). Massage
therapy reduces anxiety and enhances E E G pattern of alertness and math computat ions.
International Journal of N euroscience , 86, 197±205.
Field, T., & Jones, N. (1998). R ight frontal E E G asymmet ry is attenuated by massage and
music therapy. A dolescence. Submitted.
Field, T., Lasko, D ., Mundy, P., Henteleff, T., Talpins, S., & D owling, M. (1996c). Autistic
children’ s attentiveness and responsivity improved after touch therapy. Journal of A utism
and D evelopm ental Disorders, 27, 329±334.
796 FIELD

Field, T., Morrow, C., Valdeon, C., Larson, S., Kuhn, C., & Schanberg, S. (1992). Massage
reduces anxiety in child and adolescent psychiatric patient s. Journal of the A merican
A cadem y of A dolescent and Child Psychiatry , 31, 124±131.
Field, T., Peck, M., Krugman, S., Tuchel, T., Schanberg, S., Kuhn, C. et al. (1998c). Massage
therapy effects on burn patients. B urn Care and Rehabilitation .
Field, T., Q uintino, O ., & H ernandez-R eif, M. (1998d). A ttention D e® cit H yperactivity
D isorder adolescent s bene® t from massage therapy. A dolescence, 33, 103±108.
Field, T., Sca® di, F., & Schanberg, S. (1987). Massage of preterm newborns to improve
growth and developmen t. Pediatric N ursing, 13, 385±387.
Field, T., Schanberg, S., Kuhn, C., Fierro, K., H enteleff, T., Mueller, C. et al (1998e). Buli-
mic adolescent s bene® t from massage therapy. A dolescence.
Field, T., Schanberg, S.M., Sca® di, F., Bauer, C.R ., Vega-L ahr, N., G arcia, R . et
al. (1986). Tactile/kinesthetic stimulation effects on preform neonate s. Pediatrics, 77,
654±658.
Field, T., Seligman, S., Sca® di, F., & Schanberg, S. (1996d). A lleviating post traumat ic stress
in children following H urrican A ndrew. Journal of A pplied D evelopm ental Psychology , 17,
37±50.
Field, T., H ernandez-R eif, M., Seligman, S., Krasnegor, J., & Sunshine, W., (1997b). Ju-
venile rheumato id arthritis: bene® ts from massage therapy. Journal of Pediatric
Psychology , 22, 607±617.
Field, T., Sunshine, W., H ernandez-R eif, M., Q uintino, O ., Schanberg, S., Kuhn. C. et
al. (1997c) Chronic fatigue syndrome: Massage therapy effects on depression and
somatic symptoms in chronic fatigue syndrome. Journal of Chronic Fatigue Syndrom e, 3,
43±51.
G oldst ein-F erber, S. (1997). M assaging prem ature infants. Paper presente d at Child
D evelopment Conference , Bar-Ilan U niversity, Israel.
Ironson, G ., Field, T., Sca® di, F., Kumar, M., Patarca, R., Price, A. et al. (1996). Massage
therapy is associated with enhanceme nt of the immune systems cytotoxic capacity.
International Journal of N euroscience , 84, 205±218.
Jinon, S. (1996). The effect of infant massage on growth of preterm infant. In C. Yabes-
A lmirante & M.G . deLuna (E ds.), Increasingly safe and successful pregnancies , (pp. 265±
269). Amsterdam : Elsevier.
Kuhn, C., Schanberg, S., Field, T., Symanski, R ., Z immerman, E ., Sca® di, F. et al. (1991).
Tactile kinesthet ic stimulation effects on sympathet ic and adrenocor tical function in
preterm infants. Journal of Pediatrics, 119, 434±440. R ehabilitation M edicine, 19, 153±
158.
Meaney, M.J., A itken, D.H., Bhatnaga r, M., Bodnoff, S.R ., Mitchell, J.B., & Sarrieau, A .
(1990). Neonata l handling and the developme nt of the adrenocor tical response to stress.
In N. G unzenhauser , T.B. Brazelton, & T. Field (E ds.) A dvances in touch. Skillman, NJ:
Johnson & Johnson.
Melzack, R ., & Wall, P.D. (1965). Pain mechanisms: a new theory. Science, 150, 971±978.
Modi, N., & Glover, J. (1996). M assage therapy for preterm infants. Paper presente d at
Touch R esearch Symposium, Providence , R hode Island.
O lder, J. (1982). T ouching is healing. New York: Stein & D ay.
Sca® di, F., & Field, T. (1996). Massage therapy improves behavior in neonate s born to HIV
positive mothers. Journal of Pediatric Psychology , 21, 889±898.
Sca® di, F., Field, T., Wheeden, A ., Schanberg, S., Kuhn, C., Symanski, R ., Zimmerma n, E ., &
Bandstra, E .S. (1996). Cocaine-expo sed preterm neonate s show behaviora l and hormo-
nal differences. Pediatrics, 97, 851±855.
Sca® di, F., Field, T., & Schanberg, S. (1993). Factors that predict which preterm infants
bene® t most from massage therapy. D evelopm ental and B ehavioral Pediatrics, 14, 176±180.
 TOUCH THERAPY EFFECTS 797
Sca® di, F., Field, T., Schanberg, S., Bauer, C., Tucci, K., R oberts, J. et al. (1990). Massage
stimulates growth in preterm infants: A replicatio n. Infant B ehavior and D evelopm ent, 13,
167±188.
Sca® di, F., Field, T., Schanberg, S., Bauer, C., Vega-Lahr , N., G arcia, R . et al. (1986). E ffect
of tactile/kinesthetic simulation on the clinical course and sleep/wake behavior of preterm
neonate s. Infant B ehavior and D evelopm ent, 9, 91±105.
Schanberg, S. (1995). G enetic basis for touch effects. In T. Field (Ed.), T ouch in early
developm ent. H illsdale, NJ: E rlbaum.
Schanberg, S., & Field, T. (1987). Sensory deprivatio n stress and supplement al stimulation in
the rat pup and preterm human neonate . Child Developm ent, 58, 1431±1447.
Sunshine, W., Field, T., Schanberg, S., Quint ino, O ., Kilmer, T., Fierro, K. et al. (1996).
Massage therapy and transcuta neous electrical stimulation effects on ® bromyalgia. Journal
of Clinical R heum atology , 2, 18±22.
U vnas-M oberg, K., Widstrom, A.M., Marchini, G ., & Windberg, J. (1987). R elease of G l
hormone in mothers and infants by sensory stimulation. A cta Pediatrica Scandinavica , 76,
851±860.
Wheeden , A., Sca® di, F.A., Field, T., Ironson, G., Bandstra , E ., Schanberg, S. et al. (1993).
Massage effects on cocaine-expose d preterm neonate s. Journal of D evelopm ental and
B ehavioral Pediatrics, 14, 318±322.
Copyright of International Journal of Behavioral Development is the property of Sage
Publications Inc. and its content may not be copied or emailed to multiple sites or posted to a
listserv without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.