J. Psychosom. Obstet. Gynecol.

18 (1997) 286-291

Labor pain is reduced by massage

therapy

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

?: Field', M . Hernandez-Rkf,

S.Taylor', 0.Quintino' and I. Bumran2

'Touch Research Institute, University of Miami School of Medicine and *Educating Hands
School of Massage Therapy, Miami, Florida, USA
Key words: LABOR PAIN. ASSAG AGE

ABSTRACT
Twenty-eight women were recruited from prenatal classes
and randomly assigned to receive massage in addition to
coaching in breathingfrom theirpartners during labor, or to
receive coaching in breathing alone (a technique learned
during prenatal classes). 'The massaged mothers reported a
decrease in depressed mood, anxiety and pain, and showed
less agitated activity and anxiety and more positive affect
following thejrst massage during labor: In addition, the
massaged mothers had sign$cantly shorter labors, a shorter
hospital stay and less postpartum depression.

INTRODUCTION
Touch and massage have been used during labor in
nearly every culture for hundreds of years'. Only
recently has physical support been available to
Western women during delivery2. In the past,
massage and support during labor were used to
improve or correct the position of the fetus, to
stimulate uterine contractions, to prevent the fetus
from rising back up in the abdomen and to emrt
mechanical pressure to aid in the expulsion of the
child3.However, today the focus tends to center more
on relaxation to reduce anxiety and alleviate pain'.
A strong association between m a t e d anxiety
(typically measured by self-report questionnaires)
and labor discomfort has been reported. Labor
discomfort is thought to arise fiom fear of the

unknown, which leads to sympathetic arousal producing tension in the circular fibers of the uterus
and rigidity at the opening of the cervix'. This force
acts against the expulsive muscle fibers in labor,
producing tension within the uterine cavity which
is interpreted by the laboring mother as pain. Prolonged uterine muscle tension can produce ischemia
(local and temporary anemia due to poor blood
flow), resulting in pain. Maternal anxiety can cause
increased catecholamines, resulting in a decrease in
uterine contractility and blood flow, and therefore
pain and maternal complications during delivery2.
henatal classes often include instructions on
visualization and imagery, with the expectation that
women will be more relaxed and in control of pain
during labor. A recent study on the effectiveness of
imagery, however, failed to find differences
between women who participated in visual imagery
training sessions and a control group with regard to
self-reports on anxiety and pain levels during labo9.
However, when intense visual imagery training
sessionswere used in the management of labor pain
and other chronic pain conditionsa, the results
were more favorable. Thus, visual imagery
techniques may be effective for reducing pain
associated with labor, although at the cost of
extensive training.
Being .touched in general during labor and
delivery is perceived by mothers as a positive
experience9.For example, abdominal massage used

Correspondence to:T. Field, Touch h r c h I n ~ t i t u University

~.
of Miami School of Mdicine. PO Box 016820. Mimi. FL 33101.

USA

286

Field et al.

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

Massage and labor

in some Lamaze childbirth educational classes has

been suggested for easing discomfort of tired or
cramping muscles and facilitating control and
relaxation during labor'". The presence and
involvement of partners who touch the women
during labor have been significantly correlated with
less need for drugs, shorter labors, fewer perinatal
problems and more optimal maternal interactions".
Studies involving doulas (labor support people),
who have already experienced a labor of their own
and remain at the patient's bedside from time of
admission to the end of birth, show a positive
impact on labor'. Studies involving male partners,
such as the husband, have also reported positive
effects'*. In a study comparing doulas and male
partners, doulas touched the laboring women more
and were present for more time during labor than
male partnersI2. A pilot study by Kennel1 and
colleague^'^ reported inconclusive results when
examining male partner behavior during labor and
delivery, with some partners providing excellent
support while others did not.
These reports combined suggest that persons
providing more touch have been more favorably
viewed by women in labor. The touching seems to
reduce maternal anxiety and facilitate labor and
delivery. However, it is still unclear whether the
reduced anxiety and pain derive from the
supportive presence of another person, such as the
doula, or whether active touch such as massage
therapy could further reduce those problems. The
present study compared partners massaging women
in labor versus the partners being present and
simply doing what came naturally during labor
(typically the coaching in breathing exercises they
had learned in prenatal classes). We expected that
massage coupled with the breathing exercises
would have more positive effects than the breathing
exercises alone on anxiety and pain, as well as
length of labor.

METHODS
Sample
The sample comprised 28 middle socioeconomic
status women (mean age=29.7) who were
recruited from Lamaze classes during their last
trimester (mean = 37 weeks gestation). The sample
size was predetermined by a power analysis based
on a medium size effect, an alpha of 0.05 and power
> 0.80 in previous massage study data. The women

J. Psychosom. Obstet. Gynecol.

were most frequently married (91%) and

distributed 34% White, 9% Black and 57%
Hispanic. Other sample characteristics were (1)
74% had worked through the seventh month of
pregnancy; (2) 75% had more than 12 prenatal
visits; (3) 60% had attended from one to six
childbirth classes, 63% had learned the breathing
technique and 60% reported using the breathing
technique during delivery; and (4) 67% reported
having had a massage previously The women were
randomly assigned (based on a table of numbers) to
massage therapy or a control group. The groups did
not differ on any of the above baseline variables.
The groups also did not differ on the relationship
with labor partner (88% were husbands, 9% parents
and 3% relatives).

Procedures
Massage condition
Following the admissions interview, the massage
was taught to the partner for a mean of.10 min by a
massage therapist. At approximately 3-5cm cervical
dilation, the subjects then received 20 min of head,
shoulder/back, hand and foot massage, respectively
The massage entailed moderate pressure and
smooth movements specifically adapted to relax the
strained and stressed areas of the laboring body The
20-min sequence consisted of smooth timed
clockwise circular stroking movements for 5-min
consecutive periods in each of the four regions
while the mother was laying on her side: (1) around
the head, down the temple to the (2) neck and
shoulder, across and down the back; (3) to the hand
and (4) then down to the foot. The same 20-min
massage was repeated by the partner every hour for
5 h. The massage therapist was present only until
the partner felt comfortable giving the massage on
hidher own. None of the partners refused to give
the massage and none of them reported being
uncomfortable delivering the massage. Immediately
after the first massage the research associate, who
was blind to group assignment, was allowed into
the labor room to record measures on the
immediate effects of the massage.
Attention control condition

The control subjects and their partners were simply

asked to engage in whatever activities they had been
taught, for example the breathing coaching, or

287

Field ef al.

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

Massage and labor

whatever came naturally during labor. Their

activitieswere observed by a research associate,who
served not only as a control for the subject receiving
extra attention but also to record whether anyone in
the control group was receiving massage. None of
the control group mothers requested or received
massage from their partners, even thwgh 60% had
been previously exposed to massage. The same
baseline and follow-up assessments were conducted at the same time as those of the massage
group (a pre-/post-2O-min interval).

Partners mpott mearum

The partners rated the mothers stress level and
labor progression on 5-point Likert scales before
and after the massage.

The Behavior Observation Scak (BOS)i5

Before and after the massage session, the mothers
behavior was rated by an observer blind to the
mothers group status on a 3-point continuum on
four scales including activity, anxiety, and positive
facial expressions.

Self-report measures
Labor w a s considered that time period from the
onset of hospitalization until the patient was taken
to the delivery room at full cervical dilation. As
soon as the subject was admitted to the hospital, a
demographic interview was conducted to ensure
comparabilityof the two groups.

Post-labor measures

The Centerfor Epidemiological Studies Scalefor

Depression (CES-D)16

Pre -post-massage measures

This is a 20-item scale containing questions relating

to depressed mood and psychophysiological
indicators of depression. Respondents were asked
to rate how frequently each symptom was
experienced during the past week on a 4-point
scale. The ratings form a summary score ranging
from 0 to 60. A score greater than 16 indicates a
high level of depression symptoms. This cutpoint
corresponded to the 80th percentile of scores in
community samples. Both the reliability and
validity of the CES-D have been supported across
demographically diverse subsarnples of the general
population16.

The Profile 0fMood States D e p m i o n Scale

(POMS)~~

The Touch Switivity Scale

Demographic information
This included maternal age, ethnicity, marital
socioeconomic status, prenatal care, attendance at childbirth classes, self-report measures on
importance of touch and previous experience being
massaged.
status,

This 14-item, Likert-type self-report questionnaire

of adjectives describing current depressed mood
w a s completed by the mothers before and after the
massage session.

This was u x d to document the mothers sensitivity

to tactile stimulation. This scale includes 22 items
on reactions to different types of touch, including
whether the individual finds being accidentally
touched.aversive.

Feeling Good Thermometer

This is a visual analog scale that was used to assess
in a less cognitive way the mothers feeling of wellbeing.
Stress level and labor pains

The mothers rated their stress levels and labor pains

before and afier the massage on a 5-point Likert
scale.

288

Labor and neonatal measures

The hospital records were then examined by a
research assistant who was blind to the mothers
group assignment. The following data were
recorded hours of labor, days hospitalization, the
infants gestational age, birthweight, length, head
circumference, Ponderal Index, and perinatal data
on the Obstetric and Postnatal Complications
Scales*.

J. Psychosom. Obstet. Gynecol.

Field ei al.

Massage and labor

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

RESULTS
A repeated measures MANOVA on self-report and
partner report measures with pre- and post-sessions
as the repeated measure revealed a group by
repeated measures interaction effect. Post hoc
comparisons suggested the following for the selfreport measures (see Table 1): (1) the massaged
group mothers versus the control group mothers
reported less depressed mood (on the POMS),
feeling better (on the visual analog thermometer),
lower stress levels and decreased labor pains; and
(2) the control group mothers reported increased
labor pains across the same period of time. The
massage versus control group partners reported
lower maternal stress levels and greater labor
progression.
A repeated measures by group interaction effect
and post hoc comparisons for the behavior
observation measures suggested the following (see
Table 2): (1) the massage group showed lower
activity and anxiety levels and more positive facial
expressions after the massage; and (2) the control
group showed more positive facial expressions
following a similar time period.
A MANOVA on the post-labor variables yielded
a significant interaction effect. Post hoc comparisons
suggested the following effects favoring the
massage group (see Table 3): (1) less touch
sensitivity; (2) lower levels of perinatal depression;
(3) fewer hours in labor; and (4) a shorter hospital
stay.
A MANOVA yielded no group differences on the
following neonatal measures: (1) gestational age
(mean = 39.0 weeks); (2) birthweight (mean =
3304g); (3) length (mean = 48.4cm); (4) head
Table 1 Means for self-report and partner's measures for
pre-post-labor massage/control sessions (control means in
parentheses)
Ptr
SeFreport measures
Depressed mood (POMS)

Feeling Good Thermometer'

Stress level'
Labor pains
Partners' report measura
Mother's stress level'

Labor progression'

Post

14.0 (14.4)
5.6 (6.5)
3.3 (3.4)
5.0 (4.3)

6.9(14.9)
6.8' (6.6)
5.2- (3.5)
3.5' (5.0)'

3.4 (3.3)
3.7 (4.1)

5.4' (3.6)
4.1' (3.7)'

'High values arc optimal. Superscripts denote group diffcrcnm

c p < 0.05."p < 0.001)

J. Psychosom. Obstet Gynecol.

Table 2 Means for observer measures pre-post-labor

musagelcontrol sessions (control means in parentheses)
PR

Bahavior observation measures

Activity'
Anxiety'
Positive facial expressions'
~~

Post

1.7 (1.9) 2.5- (2.0)

1.9 (1.8) 2.4" (1.8)
1.9 (2.0) 2.3' (2.6)~

~~

'High values arc optimal. Superscripts denote group differences

(p < 0.05. "p < 0.01, "p < 0.001)

Table 3 Means for post-labor measures

Post-labor measures

Postpartum depression (CES-D)

Touch sensitivity
No. of hours in labor
No. days hospital stay

Group
Massage Control

15.4
27.9
8.5
1.3

19.8'
11.1'
11.3'
2.2'

Superscripts denotes group differences ('p < 0.05)

circumference (mean = 33.5cm); (5) Ponderal

Index (mean = 1.6); (6) Obstetric Complications
Scale Scores (mean = 116); and (7) Postnatal
Complications Scale Scores (mean = 128).

DISCUSSION
Data from three different sources (the mother,
partner and observer) converged to suggest that
massage therapy reduces stress and pain during
labor. The pregnant women themselves reported
less depressed mood state, feeling better, having less
stress and fewer labor pains following massage. In
contrast, labor pains increased in those women who
were not massaged. The mothers' partners (most
frequently their husbands) also evaluated the
massaged women as being less stressed and labor
progressing better following the massage sessions.
Behavioral observations by an observer who w a s
blind to the women's g o u p assignment rated the
women as having lower activity and anxiety levels.
The only similarity between the groups was an
increase in positive facial expressions.
Lower levels of self-reported stress and less
depressed mood state and decreased behavioral
anxiety have been reported following massage
therapy in several st~dies'~J0.These lower
anxiety/stress levels have typically been asseated
with lower stress hormone (cortisol) levels. Lower
pain levels have also resulted fiom massage therapy
given to adults with pain syndromes such as
fibromyalgia*'. The underlying mechanism for

289

Field ct al.

Massage and tabor

reduced stress hormones are unclear, although we

have speculated elsewhere that reduced anxiety and
stress hormones may be mediated by increasing
parasympathetic activity accompanying massage.
The mechanism underlying reduced pain is also
unclear. Some investigators speculate that pressure.
stimulation from massage pre-empts the processing
of pain stimulation because pressure fibers arc
longer and more myelinated and thus can relay the
signal to the brain faster than pain fibersP. That
these immediate effects of massage therapy
translated into longer term effects including shorter
labor, shorter hospital stay less touch sensitivityand
less postpartum depression was more surprising.
These findings are tempered by the fact that the
women's self-report might have been influenced by
their wish to please their partners following the
massage, and by the partners who provided the
massage also providing their assessment on at least

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

'

REFERENCES
1. Hedstrom LW, Newton N. Touch in labor: a
comparison of cultura and eras. Birth Issues Perinat

Care Educ 1986;13:181-6.

2. Kennell J. Klaus M, McGrath

S. Robertson S.
Hickley C. Continuous emotional support during
labor in a US Hospital. J Am Med Assoc
1991;265:2197-201.
3. Engdman G. Labour Among Primitive Peoples. St
Louis, M O JH Chvnber~1982,216-229.
4. Read D. Childbirth Without Feu: The Principles
and Practices of Natural Childbirth. NewYork New
York Press, 1972.
5. Ibrol C, Von Baeycr C. Effects of brief instruction in
imagery and birth visualization in prenatal education.
J Ment Imagery 1992.16167-72.
6. Brown J. Imagery coping strategies in the treatment
of migraine. Pain 1984,18157-67.
7. Lindberg C, Lawlii GF. The effectiveness of imagery
as a childbirth preparatory technique. J Ment
Imagery 1988.12103-14.
8. Rlft D, Smith M. Wurrn LD. Selection of imagery
in the relief of chronic and acute clinical pain. J
+horn Res 198630481-8.
9. Stolte KM. An Exploratory Study of Patient
Perceptions of tfie Touch They Receive During Labor.
Lawrence, Kansas 1976; unpublished disscrption.
10. Wideman Ml( Singer JEThe role of psychologiul
mechanisms in preparation for childbirth. Am
Psych01 1984,39:1357-71.

two measures. In addition, the sample was small,

suggesting the need for studies on larger samples
and more comprehensive labor outcome measures.
Having a larger sample size might reveal effects on
neonatal outcome. Nonetheless, these data
highlight the cost-effectiveness of significant others
providing massage during labor. Further research is
needed on the underlying mechanisms, and
whether neonatal outcome can be enhanced by
providing massage therapy during pregnancy long
before the onset of labor.

ACKNOWLEDGEMENTS
This research was supported by an NIMH
Research Scientist Award (#MH00331) and an
NIMH Research Grant (#MH46586) to Tiffany
Field and a grant from Johnson &Johnson to the
Touch Research Institute.

11. Sosa R, Kennell J, Klaus M. Urrutia J. The effect of

a supportive companion on perinatal problems,
length of labor. and mother infant interaction. N
En61J Med 1980,303:597-600.
12. Bertsch TD, Nagashima WL, Dykeman E, Kennell
JH. Labor support by first-time fathers: direct observations.J Psychosom Obstet Gynecol1990,11:251-60.
13. Nagashima L., Berxhi T, Dykeman S, McGrath S.
Delay T, K e ~ e l J.
l Fathers during labor: do we
expect too much? Pedktr Res 1987,21:281.
14. McNair DM. LOrr M. Droppleman LF. POMSProfile of Mood States. San Diego, CA: Educational
and Industrid Testing Service 1971.
15. Platania-Soluzo A, Field T, Blank J. Seligman F.
Kuhn C, Schanberg S, S u b I! Rekcation therapy
reduces anxiety in child adolescent psychiatry
patients. Acta Padopsychiatr 199235 115-20.
16. Radloff L The CES-D scale: a self-report depression
scale for research in the general population. Appl
Psycho1 Meu 1977;1:385-401.
17. Royeen CB. Test-retest reliability of a touch scale for
tactile defensiveness. Occup Phys Ther Pediatr 1987;
745-52.
18. Littman D, Parmelee A Medical correlates of infant
development Ftdiatrics 1978;61:470- 82.
19. Field T, Morrow C, Wdeon C, Larson S, Kuhn C.
Schanberg S. Mysage therapy reduces anxiety in
child and adolescent psychiatric patients. J Am Acad
Child ad ole^^ psychi?try 199231:125-31.
20. Field TM, Hernandez-Reif M, Quintino 0.
Schanberg S, Kuhn C. Elder retired volunteers

J. Psychosom. Obstet. Gynecol.

Massage and labor

benefit from giving massage therapy to infants. J

Appl Gerontol; 1998, in press.
21. Sunshine W, Field T, Schanberg S. Quintino 0,
Kilmer T, Fierro K, Burman I, Hashimoto M,
McBride C, Henteloff T. Massage therapy and

Field et al.

transcutaneous electrical stimulation effects on

fibromyalgia.J Clin Rheumatol 19962: 18-22.
22. Melzack R. Wall PD. Pain mechanisms: a new
theory. Science 1965;150:971-8.

J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Louisville on 12/30/14

For personal use only.

Received 21 October 1996; accepted 14 April 1997

J. Psychosom. Obstet. Gynecol.

291