Otology & Neurotology

29:557Y560 2008, Otology & Neurotology, Inc.

Effects of Exercises on Bell_s Palsy:

Systematic Review of Randomized Controlled Trials
*Jefferson Rosa Cardoso, Elsie Cobra Teixeira, *Michelle Damasceno Moreira,
Francis Meire Favero, Sissy Veloso Fontes, and Acary Souza Bulle de Oliveira
*Universidade Estadual de Londrina; Private Physiotherapy Practice, Jacarei;
and Universidade Federal de Sao Paulo, Sao Paulo, Brazil

Objective: This study examined the effects of facial exercises

associated either with mirror or electromyogram (EMG) biofeedback with respect to complications of delayed recovery in
Bell_s palsy.
Patients: Patients with unilateral idiopathic facial palsy were
included in this review.
Interventions: Facial exercises associated with mirror and/or
EMG biofeedback as treatment.
Main Outcome Measure: Report of facial symmetry, synkinesis, lip mobility, and physical and social aspects.
Results: Four studies of 132 met the eligibility criteria. The
studies described mime therapy versus control (n = 50), mirror
biofeedback exercise versus control (n = 27), Bsmall[ mirror

movements versus conventional neuromuscular retraining

(n = 10), and EMG biofeedback + mirror training versus mirror
training alone. The treatment length varied from 1 to 12
months.
Conclusion: Because of the small number of randomized controlled trials, it was not possible to analyze if the exercises, associated either with mirror or EMG biofeedback, were effective. In
summary, the available evidence from randomized controlled
trials is not yet strong enough to become integrated into clinical practice. Key Words: Facial paralysis/rehabilitationV
Meta-analysisVPhysical therapy techniquesVRandomized
controlled trialVReview (publication type).
Otol Neurotol 29:557Y560, 2008.

Bell_s palsy (BP) can cause unilateral paralysis of the

facial muscles and synkinesis. Because of the persistence
of synkinesis and use of exercises or electric stimulation
in an inadequate way, total recovery of the function of
the facial muscles can be jeopardized (1). A systematic
review of the effects of electrotherapy and thermotherapy
was accomplished in 2003, and some modalities were
investigated separately: electric stimulation, short-wave,
ultrasound, laser and mirror, or electromyogram (EMG)
biofeedback. As a synthesis, the author pointed out the
usefulness of these modalities, but no single form of
treatment seemed to be superior (2). Because of the
uncertainties, a systematic review is proposed for evaluating the effects of facial exercises associated either with
mirror or EMG biofeedback with respect to the complications of delayed recovery in BP.

MATERIALS AND METHODS

Location of the randomized controlled trials (RCTs) was
achieved through a computerized database search performed
at the Cochrane Controlled Trials Register Library (Issue 3,
2007), MEDLINE (1966Y2007), Excerpta Medica Database
(1980Y2007), Latin American and Caribbean Health Sciences
Literature (1982Y2007), Physiotherapy Evidence Database,
and Database of Reviews of Effects. The search strategy followed the Neuromuscular Disease criteria of the Cochrane
Collaboration.
The studies selected for this review were those that used
facial exercises associated either with mirror or EMG biofeedback as a treatment for patients with BP. These exercises use
facial mimicry or gesture making. The purpose of these exercises is to promote facial symmetry, to control and to reduce
synkinesis, and to maintain active musculature and movement
perception (1). Biofeedback is a technique that uses equipment
that demonstrates to an individual both his normal and abnormal physiologic events, either visually or auditorially (3).
The variables considered were facial symmetry, presence or
absence of synkinesis, muscular stiffness, labial mobility,
and the patients_ sociophysical aspects resulting from BP.
The scales used for evaluating these outcomes were the following: the House-Brackmann Scale, the Linear Measurement
Index, the Facial Disability Index, Lip-length and Snout
Indices, and the 4-Level Scale.

Address correspondence and reprint requests to Jefferson Rosa

Cardoso, P.T., Ph.D., Physical Therapy Department, University Hospital, Universidade Estadual de Londrina, Av. Robert Kock, 60,
Londrina-PR, Brazil 86038-440; E-mail: jeffcar@uel.br

557

Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.

558

J. R. CARDOSO ET AL.
TABLE 1.

Study

Participants

Characteristics of included studies

Interventions

Beurskens and 25 patients with

Mime therapy
Heymans (5)
unilateral PFP for
(combination of
at least 9 mo  25
stimulation of facial
patients on a
emotional expression
waiting list for 3 mo and functional
movements) for
45 min

Outcomes
Stiffness, lip mobility,
or immobility
and Facial
Disability Index

Main results

Conclusions/
limitations

These outcome
measures improved
substantially because
of the therapy

Mime therapy was

a good treatment
choice for patients
with sequelae of PFP.
Although the author
affirmed that waiting
lists are normal in
their health care
system, they did not
cite how the concealed
allocation occurred
and did not perform
the intention-to-treat
analysis
Nakamura et al. 12 patients with
Method of biofeedback
Percent asymmetry of Not only the percent
The method of biofeedback
(6)
complete FNP  15 rehabilitation (patients
eye opening width
asymmetry of eye
rehabilitation was very
patients served
tried to keep their eyes
(still images of the
opening width was
effective for preventing
as controls
open symmetrically
3 designated mouth
significantly greater in the development of
during 3 designated
movements were
the training group than synkinesis. The authors
mouth movements
chosen and measured in the control, but also, mentioned neither
using a mirror) for
by using
the degree of
measures of central
30 min
Adobe Photoshop)
synkinesis was less
tendency (and variability)
of the asymmetry nor the
intention-to-treat analysis
Ross et al. (7) 13 patients with
G1 included specific
Linear measurement of Improvements were
Feedback training in
unilateral peripheral retraining strategies
facial movement,
noted in both treatment combination with a
facial nerve injuries using mirror exercises
blinded visual
groups with respect to
structured home
(G1)  11 patients
alone. G2 included
assessment of
symmetry of voluntary rehabilitation program
with same diagnosis mirror exercises in
voluntary movement
movement and linear
is a clinically
(G2)  7 patients
combination with 0.5 h
using a standardized
measurement of
efficacious treatment
served as controls
of EMG biofeedback.
videotape, and facial
facial expression
for patients with
All patients were
nerve response
facial nerve paresis.
prescribed a minimum
to electroneurography
The third group was
of 0.5 to 1 h of
not formed by
daily home practice
random assignment
Segal et al. (8) 5 patients with Bell_s Conventional neuromuscular Symmetry and objective Facial movements in
The small-movement
palsy  5 patients
retraining program
(method that counted
both groups were
program was at least
with the same
(patient education,
the number of
significantly more
as good as the standard
diagnosis
relaxation, exercises,
synkinetic muscles
symmetric after
one. The study used
and biofeedback training)
contracting during a
treatment. The
a very small number
versus small-movement
given specific action) reduction of synkinesis of patients with a
program (identical to
and subjective
was not significant
short history of
program above, but patients (subjective synkinesis
facial paralysis
were first taught to do the
rating) synkinesis
exercises, stopping when
estimates
synkinesis occurred) in
1-h sessions
EMG indicates electromyogram; FNP, facial nerve palsy; G1, Group 1; G2, Group 2; PFP, peripheral facial paralysis.

The data were extracted from relevant studies presented by 1

review author. The methodologic quality of potentially
included studies was assessed independently by 2 review
authors using the Delphi List (4). The data are presented in
descriptive form without meta-analysis because the studies
had an incompatible combination of treatment characteristics,
duration, and outcome measures.

RESULTS
One hundred thirty-two related abstracts regarding BP
were found in the databases previously mentioned. Only
4 studies contemplated the goals of this review and were

acquired and analyzed: Beurskens and Heymans (5),

Nakamura et al. (6), Ross et al. (7), and Segal et al. (8)
(Table 1). The number of patients from the 4 studies
was 118 (66 in the treatment group and 52 in the control
group). The treatment duration was between 1 and
12 months. The description of included studies is in
Table 1.
Beurskens and Heymans (5) studied 50 patients during
a period of 3 months: 25 in the treatment group and 25
in the control group (without treatment). The authors_
conclusion was that mime therapy is a good treatment
choice for patients with sequelae of BP. Nakamura
et al. (6) refer to the efficacy of exercises using mirror

Otology & Neurotology, Vol. 29, No. 4, 2008

Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.

EXERCISE THERAPY FOR BELLS PALSY

TABLE 2.
Score
Study
Beurskens and
Heymans (5)
Nakamura et al.
(6)
Ross et al. (7)
Segal et al. (8)

Quality assessment score of the studies using

the Delphi List
1

Total

Yes No Yes Yes No No No Yes No

Yes No Yes Yes Yes No No No No

Yes No Yes Yes Yes No No Yes No

Yes No Yes Yes No No No Yes No

5
4

Delphi List items: 1, random allocation; 2, concealed allocation;

3, groups similar at baseline; 4, description of the eligibility criteria;
5, outcome assessor blinded; 6, care provider blinded; 7, patient
blinded; 8, point estimates and measure of variability 9, intention-totreat analysis.

biofeedback. Twelve patients were allocated for the treatment group and 15 in the control group. The authors
concluded that this treatment was effective in the prevention of the synkinesis.
In the study of Ross et al. (7), the treatment group
comprised 11 patients that received EMG plus mirror
feedback. The comparison group consisted of 13 patients
treated with exercises performed in front of the mirror
alone, and 7 patients were used as controls. Both experimental groups obtained significant improvement in symmetry of voluntary movement and linear measurement.
Segal et al. (8) performed a study with 2 treatment groups
composed of 5 patients each. The treatment offered to the
main group was a protocol of Bsmall movement[ exercises, whereas standard exercises were used with the
control group. The authors found significant improvement for symmetry in both groups, but a nonsignificant
reduction of synkinesis. The results of the quality assessment studies are presented in Table 2.
DISCUSSION
This study is relevant for the lack of an effective treatment protocol for patients with BP. Few studies were
found in the several databases used, which serves to
reinforce the importance of this work. The 4 RCTs
were found in a period from 1958 to 2007. This should
serve as an alert to health professionals because primary
studies that respond to the treatment question are only
being pursued in an unfocused and sporadic way.
Beurskens and Heymans (5) used the House-Brackmann
Scale, recommended as an important instrument of evaluation, but only to include and classify participants in the
study and not to verify their improvement. Nakamura
et al. (6) used a group of biofeedback exercises that
used mirrors and an assessment procedure not mentioned
by any of the other authors; thus, its comparison with the
others became impossible. Nevertheless, they also found
significant improvement in percent of asymmetry of
eye opening width in the training group and a reduction
in synkinesis.

559

Segal et al. (8) published an RCT with few participants

in each group (n = 5), some patients having had BP for
only 5 months. This factor could have biased the study,
seeing that it falls within the period in which spontaneous
improvement can occur. Nevertheless, it stresses the
diagnosis of BP in almost all patients (9 of 10). However,
in the study of Ross et al. (7) comprising only 3 groups,
the 31 participants had various causes of facial paralysis, and only 4 were diagnosed with BP. All of the studies used facial exercises (mimicry or facial expression) as
the treatment; however, each one had its own emphasis.
Two studies (5,8) used facial expression exercises demonstrating sensations such as happiness (smiling), sadness (crying), and singing (whistling). Another study
(6) imposed a form of lip movement without eye closure,
and, finally, the last study (7) did not describe the type of
exercise used. The use of mime therapy, whether associated or not with mirror biofeedback, seems to have
beneficial effects on the patients with BP. Each author
used different outcome measures. None used the scale of
House-Brackmann as the main outcome and, when they
used the scale, it was just to either include or classify
patients. Another outcome used by 2 authors (7,8) was
the Linear Measurement Index, which is similar to the
House-Brackmann Scale.
Of the 118 included patients, 57 were diagnosed with
BP. The remaining patients had other types of facial paralysis caused by acoustic neuromas, herpes zoster, traumas,
neuromas of the meninges, Lyme disease, jugular artery
thrombosis, facial neuromas, and surgical trauma.
Regarding the quality assessment of the studies, these
were considered moderate due to nonallocation concealed description (through stamped and opaque envelopes, after the generation of the numbers) and the
nondescription of the intention-to-treat analysis. Another
reason is the impossibility of blinding (therapist and
patient) when the exercises are performed. It is important
to emphasize that 2 of the included studies were written
before the Consort Statement in 2001 (9).

IMPLICATIONS FOR PRACTICE

The research question of this study still remains open
because we did not detect in the included studies
decreases of degree II or less in the House-Brackmann
Scale or in the 4-Level Scale, which involves a reduction
of synkinesis and stiffness, together with the improvement of symmetry and global mobility of the face. From
a practical point of view, the following should be taken
into account: the facility of the use of the mirror (low
cost and easy access) when compared with electrostimulation devices or EMG biofeedback, patient educationV
the patient_s understanding and participation in the
accomplishment of the facial exercises, and the therapist_s
expertise in differentiating several degrees of both scales
and the relative experience in supervising those exercises.
Furthermore, the improvements observed in the outcome
measures can be temporary.
Otology & Neurotology, Vol. 29, No. 4, 2008

Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.

560

J. R. CARDOSO ET AL.
IMPLICATIONS FOR RESEARCH

One of the main resultant implications of this systematic review is the need for new studies to investigate
the effects of facial exercises associated either with
mirror or EMG biofeedback in patients with BP.
These exercises can prevent or reduce synkinesis,
improve orofacial functionality, and therefore avoid
negative social repercussions. Furthermore, it suggests
the development of an RCT according to the guidelines
of Consort Statement (9).
CONCLUSION
Because of the small number of RCTs, it was not
possible to analyze if the exercises, associated either
with mirror or EMG biofeedback, were effective.
Although all the studies reported clinical and statistically
significant differences among the different treatment
groups, the nonstandardized outcome measures used do
not allow definitive conclusions. The available evidence
from RCT is not yet strong enough to be integrated into
clinical practice.
Acknowledgment: The authors thank Gloria Thomas,
formerly of the Robert Jones and Agnes Hunt Orthopaedic

Hospital, Oswestry, UK, for suggestions and correction of

the article.

REFERENCES
1. Diels HJ. Facial paralysis: is there a role for therapist? Facial Plas
Surg 2000;16:361Y4.
2. Quinn R, Cramp F. The efficacy of electrotherapy for Bell_s palsy:
the systematic review. Phys Ther Rev 2003;8:151Y64.
3. Basmajian JV. Research foundations of EMG biofeedback in rehabilitation. Biofeedback Self-Regul 1988;113:275Y98.
4. Verhagen AP, Vet HC, Bie RA, et al. The Delphi List: the criteria
list for quality assessment of randomised clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51:1235Y41.
5. Beurskens CH, Heymans PG. Positive effect of spoils therapy on
sequelae of facial paralysis: stiffness, lip mobility, and social and
physical aspects of facial disability. Otol Neurotol 2003;24:677Y81.
6. Nakamura K, Toda N, Sakamari K, Kasimha K, Takeda N. Biofeedback rehabilitation for prevention of synkinesis after facial palsy.
Otolaryngol Head Neck Surg 2003;128:539Y43.
7. Ross B, Nedzelski JM, McLean A. Efficacy of feedback training in
long-standing facial nerve paresis. Laryngoscope 1991;101:744Y50.
8. Segal B, Hunter T, Danys I, Freedman C, Martin B. Minimizing
synkinesis during rehabilitation of the paralyzed face: preliminary
assessment of the new small-movement therapy. J Otolaryngol
1995;24:149Y53.
9. Moher D, Schulz KF, Altman DG. The Consort-Statement: revised
recommendations for improving the quality of reports of parallelgroup randomised trials. Ann Intern Med 2001;134:657Y62.

Otology & Neurotology, Vol. 29, No. 4, 2008

Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.