Journal of Bodywork & Movement Therapies (2017) 21, 565e568

Available online at www.sciencedirect.com

ScienceDirect

journalhomepage:www.elsevier.com/jbmt

QUASI EXPERIMENTAL STUDY

Effect of therapeutic Swedish massage on

anxiety level and vital signs of Intensive Care
Unit patients
Tatiana Alves da Silva, PT, Debora Stripari Schujmann, PT,
Leda Tomiko Yamada da Silveira, PT,
Fa´tima Aparecida Caromano, PhD, Carolina Fu, PhD*

Department of Physiotherapy, Communication Sciences & Disorders and Occupational Therapy, Faculty
of Medicine, Sao Paulo University, Sao Paulo, SP, Brazil

Received 16 March 2016; received in revised form 15 July 2016; accepted 16 August 2016

KEYWORDS Summary Objective: To evaluate how Swedish massage affects the level of anxiety and vital
Anxiety; signs of Intensive Care Unit (ICU) patients.
Massage; Methods: Quasi-experimental study. Inclusion criteria: ICU patients, 18e50 years old, cooper-
Intensive Care Units ative, respiratory and hemodynamic stable, not under invasive mechanical ventilation. Exclu-
sion criteria: allergic to massage oil, vascular or orthopedic post-operative, skin lesions,
thrombosis, fractures. A 30-min Swedish massage was applied once. Variables: arterial pres-
sure, heart rate, respiratory rate, S-STAI questionnaire. Timing of evaluation: pre-massage,
immediately post-massage, 30 min post-massage. Comparison: T-test, corrected by Bonferroni
method, level of significance of 5%, confidence interval of 95%.
Results: 48 patients included, 30 (62.5%) female, mean age 55.46 (15.70) years old. Mean S-
STAI pre-massage: 42.51 (9.48); immediately post-massage: 29.34 (6.37); 30 min post-
massage: 32.62 (8.56), p < 0.001 for all comparison. Mean vital signs achieved statistical sig-
nificance between pre-massage and immediately post-massage.
Conclusion: Swedish massage reduced anxiety of ICU patients immediately and 30 min post-
massage. Vital signs were reduced immediately post-massage.
ª 2016 Elsevier Ltd. All rights reserved.

* Corresponding author. Department of Physiotherapy, Communication Sciences & Disorders and Occupational Therapy, Faculty of
Medicine, Sao Paulo University, Rua Cipotaˆnea 51, Cidade Universita´ria, Sao Paulo, CEP 05360-000, Brazil. Fax: þ55 11
26617969. E-mail address: carolfu@usp.br (C. Fu).

http://dx.doi.org/10.1016/j.jbmt.2016.08.009
1360-8592/ª 2016 Elsevier Ltd. All rights reserved.
566
Introduction
Intensive Care Unit (ICU) patients may experience emotional
disorders such as loneliness, anxiety (Dunn et al., 1995) stress,
pain, and fear. Such disorders may even lead to physical re-
percussions such as elevation of arterial blood pressure, res-
piratory rhythm and heart rate levels and deterioration of level
of consciousness (Vahedian-Azimi et al., 2014).
Therapeutic Swedish massage has positive physiologic
effects such as a decrease of sympathetic activity, vasodi-
latation, reduction of muscle spasm and tissue congestion,
improvement of metabolites removal and pain relief. It also
influences psychological status, provides relaxation and
welfare, diminishes agitation, tension and anxiety
(Vahedian-Azimi et al., 2014; Cassar, 2001).
Previous studies have suggested that massage can be
beneficial for critically ill patients (Hill, 1993). Relaxation
interventions like therapeutic touch may reduce panic,
provide comfort and improve sleep quality (Gosselink et al.,
2008). However, not much is known about the impact of
massage on the level of anxiety and vital signs in ICU pa-
tients. Existent studies have some limitations such as sub-
jective evaluation, utilization of more than one massage
technique and massage given by a family member
(Vahedian-Azimi et al., 2014; Cutshall et al., 2010).
The objective of this study is to evaluate the effect of
Swedish massage on the anxiety status and vital signs of
ICU patients.
Method
This was a quasi-experimental study performed at the Cen-
tral Institute of Clinics Hospital, Medical School, University
of Sao Paulo, a high complexity school hospital. It was
approved by the Ethics Committee for the Analysis of
Research Pro-jects of Clinics Hospital of the Faculty of
Medicine of the University of Sao Paulo (process number
673.919). Data were collected from June to November 2014
at a 15-bed clinical ICU and a 17-bed post-operative ICU.
Inclusion criteria were: admission to ICU, age between
18 and 50 years old, who presented a Glasgow Coma
Score of 15, hemodynamic and respiratory stability and
were not under invasive mechanical ventilation. After being
assessed for inclusion in the study, patients who agreed to
partici-pate were asked to sign a consent form. All
participating patients were awake, alert and oriented, since
Coma Glasgow Score of 15 points was an inclusion criteria,
so they were able to sign the consent form by themselves.
Hemodynamic stability was defined as heart rate (HR)
between 60 and 140 beats per minute, systolic arterial
pressure (SAP) between 90 and 140 mmHg, diastolic
arterial pressure (DAP) between 60 and 90 mmHg, without
vasoac-tive drugs or with low and reducing doses.
Respiratory stability was defined as respiratory rate (RR)
between 12 and 35 breaths per minute, peripheral oxygen
saturation above 90% with maximum inspired fraction of
oxygen of 50% delivered by oxygen therapy.
Exclusion criteria were vascular and orthopedic post-
operative, limb amputation, allergy to massage oil, burn
wound, open wound, skin lesions, vascular thrombosis,
allodynia, hyperalgesia.
T. Alves da Silva et al.
Initial evaluation was performed immediately before
receiving the massage. Then, each patient was submitted
to one Swedish massage session. Patients were evaluated
again immediately after receiving the massage and 30 min
after massage was completed. Data collection and
massage were performed by the same investigator.
Collected variables were: age, gender, cause of ICU
admission, length of ICU stay from admission day until
study inclusion, vital signs and anxiety status. Evaluated
vital signs were: HR, RR, SAP, DAP and MAP, which were
collected using DX2010 or DX2020 Dixtal multiparametric
monitor (Dixtal Biome´dica Indu´stria e Come´rcio LTDA,
Manaus-AM, Brasil). Patient’s anxiety status was evaluated
by the State-Trait Anxiety Inventory for Adults (STAI)
questionnaire (Spielberger et al., 1983) in the Portuguese
version (Gorenstein and Andrade, 1996).
The state form of STAI questionnaire (S-STAI) was cho-
sen, since we intended to evaluate a temporary condition of
anxiety, which may vary according to the environment
and/or situation. Higher score reflects greater anxiety level
in a certain moment. S-STAI scale is a 20-item
questionnaire that graduates the levels of anxiety. Each
item of the questionnaire may score from one point (‘not at
all’) to four points (‘very much’). Ten items refer to the
presence of anxiety symptoms and the other ten items
reveal the absence of it. For sum score calculating, the
latter items’ score are inverted and then added to the score
of the first ten items. The sum score ranges from 20 to 80
points and is directly related to the level of anxiety. It is
usually administered as a self-completion questionnaire
(Spielberger et al., 1983) but in our study, the questions
were read to the patients by the same investigators.
Swedish massage was applied to the lower limbs, upper
limbs and to the upper trapezius muscle, with the patient in
dorsal decubitus position with a 30 elevation of the head.
Swedish massage was the chosen technique because of its
well-known relaxation effects. Massage was applied by the
same physiotherapist and lasted 30 min. Massage sequence
(Cassar, 2001) was: (1) stroking: backward and forward
movement of the therapist’s hands with light pressure; (2)
effleurage: backward and forward movement of the ther-apist’s
hands with a medium pressure; (3) kneading: compression of
the soft tissue using one hand’s thumb against the other hand’s
fingers, alternately; (4) effleurage;
(5) stroking. Each body part received one complete
sequence of the Swedish massage. Hypoallergenic oil
Dersani (Saniplan, Rio de Janeiro-RJ, Brasil) was used
during the massage.
Statistical analysis was carried out using statistic soft-
ware R (Lucent Technologies, Murray HilleKY, USA).
Descriptive analysis was made for all collected data.
Comparison of S-STAI and vital signs between the three
times of evaluation was performed using paired t-test
corrected by Bonferroni method, with significance level of
0.05 and confidence interval of 95%.
Results
A total of 48 patients were included and their character-istics
are displayed in Table 1. Mean age was 55.46 15.70 years old
and cause of ICU admission show that both clinical
Therapeutic Swedish Massage and Intensive Care Unit patients 567

post-massage vs. 30 min post-massage, with p < 0.001 for HR,

Table 1 Characterization of the study population. RR, DAP and MAP and p Z 0.003 for SAP.
Demographic data (n Z 48)
Female, n (%) 30 (62.5)
Age, years 55.46 (15.70)
ICU length of stay from admission 5.38 (6.23)
day to study inclusion, days
Cause of ICU admission, n (%)
Respiratory 10 (20.83%)
Renal 9 (18.75%)
Post-operative of abdominal surgery 7 (14.58%)
Neurologic 4 (8.33%)
Cardiologic 2 (4.17%)
Other 16 (33.34%)
Data are expressed as n(%), when indicated, or mean (standard
deviation). ICU: Intensive Care Unit.

and post-operative patients were included. Post-operative

patients included in our study were mainly abdominal sur-
gery patients.
S-STAI score was: 42.51 (9.48) before massage, 29.34
(6.37) immediately after and 32.62 (8.56) 30 min after
massage. There was statistically significant difference be-
tween the S-STAI pre-massage versus immediately post-
massage, pre-massage versus 30 min post-massage, and
also between immediately post-massage versus 30 min
post-massage, with p < 0.001 for all comparison.
Table 2 presents vital signs data. Heart rate, RR, SAP,
DAP and MAP were smaller immediately post-massage
when compared to pre-massage; however, they were
greater 30 min post-massage than immediately post-
massage. Comparing values between 30 min post-
massage and pre-massage, it was observed smaller values
30 min post-massage only for HR.

Table 2 Vital signs pre-massage, immediately post-

massage and 30 min post-massage.
Pre-massage Immediately 30 min
post-massage post-massage
b,c
HR (bpm) 91.64 (15.86) 84.81 (14.78) 90.16 (15.44)
a,c
RR (rpm) 21.41 (4.51) 18.56 (3.56) 20.37 (3.93)
a,c
SAP (mmHg) 131.87 (25.53) 125.64 (22.71) 132.41 (25.08)
a,c
DAP (mmHg) 77.00 (17.32) 71.39 (14.59) 75.93 (17.30)
a,c
MAP (mmHg) 96.14 (20.18) 89.45 (16.33) 95.56 (18.93)
Data are expressed as mean (standard deviation). HR: heart
rate; RR: respiratory rate; SAP: systolic arterial pressure; DAP:
diastolic arterial pressure; MAP: mean arterial pressure; bpm:
beats per minute; rpm: respirations per minute; mmHg: milli-
meters of mercury.
a
Statistically significant difference between pre-massage vs.
immediately post-massage, with p < 0.001 for HR, RR, DAP e
MAP and p Z 0.006 for SAP.
b
Statistically significant difference between pre-massage vs.
30 min post-massage, with p < 0.001 for HR.
c
Statistically significant difference between immediately
Discussion

The main finding in our study was the reduction of anxiety

level status, both immediately post-massage and 30 min
post-massage. Anxiety was evaluated using a specific and
quantitative questionnaire, after a single massage tech-
nique, in a population that comprised both clinical and post-
operative patients, mainly abdominal surgery.
We also observed reduction in vital signs values but only
immediately post-massage; this result was not sustained 30
min post-massage since S-STAI score 30 min post-massage
was not different from pre-massage S-STAI score.
A previous study (Adib-Hajbaghery et al., 2014) used a
specific questionnaire for anxiety evaluation and found that
anxiety levels decreased; however, that study included only
a specific population: patients admitted due to coronary
disease or acute myocardial infarction. In a prospective
controlled randomized trial (Cutshall et al., 2010) it was
observed that massage reduced anxiety levels, measured
by an analog scale. However, the studied population
comprised cardiac post-operative patients and they sug-
gested that other surgical patients should be evaluated.
In both studies mentioned above, different massage
techniques were used, making it difficult to determine what
is the effect of each technique. In our study, we included
other post-operative patients and also non-surgical patients.
Also, a physiotherapist applied one spe-cific massage
technique.
Previous studies performed the massage in a private
room, sometimes associated with relaxing music. In our
study, massages were performed at the ICU bed, thus pa-
tients were exposed to the ICU environment. The presence
of other patients, ICU team, loud noise, movement and
bright light, for example, may be stressful. One could argue
that the disturbing environment might compromise the
relaxing effects of the massage. However, anxiety level did
reduce immediately and 30 min post-massage. This may
suggest that removing the patient from the ICU to receive
massage sessions is not mandatory. This makes the appli-
cation of massage in the ICU more feasible.
We used S-STAI to analyze anxiety levels. S-STAI score
ranges from 20 to 80 points. Total score is directly related to
the level of anxiety. We observed that massage reduced
anxiety levels immediately post-massage. Thirty minutes
post-massage, S-STAI score was still smaller than pre-
massage, although it was higher than the score immedi-
ately post-massage.
The fact that S-STAI score 30 min post-massage was
greater than immediately post-massage may indicate that
perhaps the effect of massage did not last for long in this
population. However, since we applied massage only once and
final evaluation was performed after 30 min, we cannot make a
statement about the duration of the massage effects.
A recent study (Vahedian-Azimi et al., 2014) claims that
complementary therapies such as therapeutic massage
have beneficial effects on vital signs in cardiac post-
operative patients. Authors observed reduction of SAP of
ICU patients submitted to massage. This may be a sign o
relaxation, since it induces endorphin secretion and pro-
motes vascular dilatation, which leads to blood flow in-
crease and arterial pressure reduction. They affirm that
568
emotional changes such as stress and anxiety may affect
physiological parameters including vital signs. They
observed decrease in DAP, HR and RR, supporting the hy-
pothesis that massage is beneficial to patients in a coronary
unit since it promotes muscle relaxation, decreases
norepinephrine production, thus reducing anxiety levels.
We demonstrated that Swedish massage could be
beneficial for clinical and post-operative ICU patients
because it reduced anxiety status. It also reduced arterial
pressure, HR and RR, as previous studies suggest (Hill,
1993). In our study, the reduction of arterial pressure and
HR was not associated with hemodynamic instability. Vital
sign values after massage remained within normal limits.
We considered that it was a positive result because it is in
accordance with the physiologic effects of massage, such
as decrease of sympathetic activity, vasodilatation,
relaxation and pain relief. It is noteworthy that our
population included only patients who were stable from
hemodynamic and respiratory point of view.
We also observed that after 30 min, arterial pressure and
RR returned to the initial condition. Although it was found
statistical difference for HR, we do not consider that there was
a clinical difference, since HR pre-massage was 91.64 (15.86)
bpm and 30 min post-massage it was 90.16 (15.44) bpm. We
concluded that vital signs reduced immediately post-massage,
but this effect was not sustained after 30 min.
Our study suggests that Swedish massage can be another
tool to be used by physiotherapists in the ICU. The main finding
of this study was that massage reduced anxiety levels, which
may contribute to treatment adherence. Besides, recent studies
have shown that stress may have deleterious effects on wound
healing and immunologic system, and massage may be
beneficial once it reduces anxiety levels (Hill, 1993). Although
massage has been pointed out as an uncommon technique in
the ICU environment, we encoun-tered good acceptance
among patients and ICU team. This was also perceived in
previous study (Martorella et al., 2014).
One of the limitations of the present study was the se-
lection of a restrict population, since pediatric patients and
adult patients with Glasgow Coma Score inferior to 15
points were not included. This proceeding was adopted
because such patients would not be able to express the will
to receive the massage or not and to answer the anxiety
questionnaire. Another limitation was the use of a single
center, although we included one clinical and one post-
operative ICU. Finally, we used a convenience sample,
without control group. Since the objective of this study was
to analyze a cause-effect relationship, we chose a quasi-
experimental design, which is characterized by the absence
of a control group and randomization. Comparison was
made between the three moments of each subject.
Conclusion
Swedish massage reduced anxiety status of clinical and
post-operative ICU patients immediately after application
T. Alves da Silva et al.
and this effect remained after 30 min. Arterial pressure, HR
and RR were reduced immediately after massage.
Furthermore, there was no instability issue, since values
remained within normal limits.
Conflict of interest
None.
Acknowledgments
None.
References
Adib-Hajbaghery, M., Abasi, A., Rajabi-Beheshtabad, R., 2014.
Whole body massage for reducing anxiety and stabilizing vital
signs of patients in cardiac care unit. Medical J. Islamic Repub.
Iran 28, 47.
Cassar, M.P., 2001. Manual de massagem terapeˆutica. Manole,
Barueri.
Cutshall, S.M., Wentworth, L.J., Engen, D., Sundt, T.M., Kelly,
R.F., Bauer, B.A., 2010. Effect of massage therapy on pain,
anxiety and tension in cardiac surgical patients: a pilot study.
Com-plement Ther. Clinical Pract. 16 (2), 92e95.
Dunn, C., Sleep, J., Collett, D., 1995. Sensing an improvement: an
experimental study to evaluate the use of aromatherapy,
massage and periods of rest in an intensive care unit. J. Adv.
Nurs. 21 (1), 34e40.
Gorenstein, C., Andrade, L., 1996. Validation of a Portuguese
version of the beck depression inventory and the state-trait
anxiety inventory in Brazilian subjects. Braz. J. Medical Biol.
Res. 29, 453e457.
Gosselink, R., Bott, J., Johnson, M., Dean, E., Nava, S.,
Norrenberg, M., Scho¨nhofer, Stiller K., van de Leur, H.,
Vincent, J.L., 2008. Physiotherapy for adult patients with crit-ical
illness: recommendations of the european respiratory so-ciety
and european society of intensive care medicine task force on
physiotherapy for critically ill patients. Intensive Care Med. 34,
1188e1199.
Hill, C.F., 1993. Is massage beneficial to critically ill patients in
intensive care units? A critical review. Intensive Critical Care
Nurs. 9 (2), 116e121.
Martorella, G., Boitor, M., Michaud, C., Ge´linas, C., 2014. Feasi-
bility and acceptability of hand massage therapy for pain
management of postoperative cardiac surgery patients in the
intensive care unit. Heart Lung 43 (5), 437e444.
Spielberger, C.D., Gorsuch, R.L., Lushene, P.R., Vagg, P.R.,
Jacobs, A.G., 1983. Manual for the State-trait Anxiety Inventory
(Form Y). Consulting Psychologists Press, Palo Alto.
Vahedian-Azimi, A., Ebadi, A., Asghari Jafarabadi, M., Saadat, S.,
Ahmadi, F., 2014. Effect of massage therapy on vital signs and
GCS scores of ICU patients: a randomized clinical trial. Trauma
Mon. 19 (3), e17031.