Acta Oto-Laryngologica, 2007; 127: 888
891

ORIGINAL ARTICLE

Prognosis for Bells palsy: a comparison of diabetic and nondiabetic

patients

ATSUKO KANAZAWA, SHIN-ICHI HAGINOMORI, ATSUKO TAKAMAKI,

RYUZABURO NONAKA, MICHITOSHI ARAKI & HIROSHI TAKENAKA
Department of Otolaryngology, Osaka Medical College, Takatsuki, Japan

Abstract
Conclusion. The present study indicates that recovery from Bells palsy in a diabetic group (DG) is delayed, and the facial
movement score remains low in comparison with a nondiabetic group (NDG). More aggressive treatments, such as higherdose corticosteroid administration and/or facial nerve decompression surgery, might be considered in diabetic patients with
severe Bells palsy. Objectives. The purpose of this study was to reveal prognostic differences for Bells palsy in the DG and
NDG. Patients and methods. The grades of facial palsy in 19 diabetic and 57 nondiabetic patients with Bells palsy were
assessed using the House-Brackmann grading system (HB system). Recovery was defined as grade I. The average of HB
system grades and recovery rates were compared in the DG and NDG at the start of the treatment, and 1 month, 3 months,
and 6 months after onset. Results. There were no differences in the HB system between the DG and NDG at the start of
treatment and at 1 month after onset. However, facial movement in the DG was poorer than that in the NDG at 3 months
and 6 months after onset. In terms of the recovery rate, the rate in the DG (52.6%) was much lower than that in the NDG
(82.5%) at 6 months after onset.

Keywords: Bells palsy, diabetes mellitus, recovery, prognosis

Introduction
Bells palsy, an idiopathic peripheral facial nerve
paralysis of sudden onset, affects approximately 25
in every 100,000 people [1
3]. Many events, including herpes viral infection [4
6], ischemia [7],
and autoimmune response [8], have been proposed
as causes of Bells palsy. Molecular biological analyses have implicated herpes simplex virus (HSV)
infection or reactivation in geniculate ganglions as
one of the causative pathogens of Bells palsy [2].
Once HSV infection or reactivation occurs in the
facial nerve, edema and ischemia of the nerve

which become more severe in a vicious circle as
a result of swelling in the narrow fallopian bony
canal
may lead to facial palsy.
Recently, an association with diabetes mellitus
has been noted, since the number of diabetic
patients is increasing in proportion to the aging
society. Adour et al. have mentioned that diabetes is
more common among patients with Bells palsy

than among those who have never had this disease

[9]. In our hospital, over 20% of patients with Bells
palsy also suffer from diabetes mellitus. Neurologically, motor nerve conduction velocity [10,11] and
amplitude [11] are lower in diabetic patients
compared with nondiabetic patients. In diabetics,
the Schwann cells and myelin sheaths of nerves are
much more prone to damage than in nondiabetic
patients [9,12]. Moreover, some other complications are observed in patients with diabetes, including angiopathy and severe viral or bacterial
infections. These complications may influence the
patients ability to recover from facial nerve paralysis. In the present study, we analyzed prognostic
differences in Bells palsy between diabetic and
nondiabetic patients in terms of facial movement
scores and recovery rates. In diabetic patients,
correlations between glycosylated hemoglobin A1c
(HbA1c) or age and facial movement were also
examined.

Correspondence: Shin-Ichi Haginomori, MD, Department of Otolaryngology, Osaka Medical College, 2-7 Daigakumachi, Takatsuki, Osaka 569-8686, Japan.
E-mail: hagi@poh.osaka-med.ac.jp

(Received 19 August 2006; accepted 19 October 2006)

ISSN 0001-6489 print/ISSN 1651-2551 online # 2007 Taylor & Francis
DOI: 10.1080/00016480601075399

Prognosis for Bells palsy

889

Patients and methods

Results

Subjects

No differences were observed in the averaged HB

system grades between the DG and NDG at the start
of treatment (5.159/1.01 in DG and 4.889/0.76 in
NDG, p / 0.16) and 1 month after onset (3.379/
1.67 in DG and 2.609/1.36 in NDG, p/0.07).
However, facial movement in the DG was poorer
than that in the NDG at 3 months (2.269/1.37 vs
1.619/0.94, p/0.039) and at 6 months (1.799/1.03
vs 1.259/0.58, p /0.01) after onset (Figure 2).
In terms of recovery rate, at 1 month after onset, 3
of 19 patients (15.8%) in the DG showed recovery
from palsy; 16 of 57 (28.1%) in the NDG had
recovery. At 3 months, 9 of 19 patients (47.4%) in
the DG showed recovery; 35 of 57 (61.4%) in the
NDG demonstrated recovery. At 6 months, 10 of 19
(52.6%) in the DG showed recovery from palsy; 47
of 57 (82.5%) in the NDG achieved recovery.
Statistical analysis revealed that the DG recovery
rate was lower than that of the NDG at 6 months
(52.6% vs 82.5%, p/ 0.01) after onset. However, no
differences were seen in the recovery rates at 1
month (15.8% in DG vs 28.1% in NDG, p / 0.23)
and at 3 months (47.4% in DG and 61.4% in NDG,
p/ 0.21) after onset (Figure 3).
In the DG, no statistical correlations were observed between facial movement at 6 months after
onset and HbA1c (p / 0.29) and age (p / 0.21).
Through the treatment, there were no patients
observed in the DG with severe side effects, e.g.
uncontrollable hyperglycemia, ketoacidosis, coma or
critical infection, due to corticosteroid injection.

This study enrolled 76 patients with Bells palsy (19

type 2 diabetic patients and 57 nondiabetic patients)
divided into 2 groups: a diabetic group (DG) and a
nondiabetic group (NDG). There was no statistical
difference in age between the groups (mean /64.69/
8.5 years in the DG and 61.39/8.5 years in the
NDG) after selecting patients whose ages were 45 or
higher because the age of the youngest patient with
DM was 45. HbA1c in the DG ranged from 5.3% to
9.0% (mean 7.259/1.20%). All patients began
receiving intravenous prednisolone injection (starting at 120 mg per day) within 4 days after the onset
of palsy, and continued to receive this treatment for
8 days (Figure 1). All diabetic patients required
subcutaneous insulin injections during the treatment
for blood glucose control.
Analysis
The grade of each patients facial score was assessed
by two expert otolaryngologists without information
as to whether the patients had diabetes or not. The
House-Brackmann facial nerve grading system (HB
system) [13] was used. This system has six grades
(from grade I to grade VI, Table I) and is used
worldwide as a method for evaluating facial nerve
function. Recovery was defined as grade I. The mean
values of HB system points and recovery rates were
analyzed four times in each group: at the beginning
of treatment, and at 1, 3, and 6 months after the
onset of facial nerve paralysis. Data obtained from
the DG and NDG were compared statistically using
Mann-Whitneys U test (HB system points) and
Fishers exact probability method (recovery rate). In
DG, correlations between HbA1c or age and facial
score
HB system score at 6 months after the onset

were analyzed using Spearmans correlation coefficient by rank test.

120 mg
Prednisolone

80 mg
40 mg

2 days

6 days intravenous injection

20 mg
oral intake
2 days oral intake

Figure 1. Therapeutic schedule for Bells palsy in the present

study.

Discussion
In the present study, no differences in HB system
grades were observed between the DG and NDG at
the start of treatment, suggesting that being diabetic
does not influence the severity of facial palsy at the
time of onset. However, recovery from Bells palsy
was found to be delayed in patients with diabetes
compared with nondiabetic patients, indicating that
diabetes may influence recovery from facial palsy.
Nerve injury associated with Bells palsy is classified
into neurapraxia and axonotmesis [14]. Complete
recovery from neurapraxia, i.e. damage of the myelin
sheath, is most likely to occur within 8
12 weeks
[14]. The grade of the facial score is considered to
reflect the degree of neurapraxia. On the other hand,
axonotmesis, i.e. damage of the axon and Wallerian
degeneration in the distal nerve fibers, is liable to
develop sequelae such as persistent facial weakness,
synkinesis, or contracture [14]. The degree of nerve
degeneration and recovery from facial palsy depends
on the severity of axonotmesis. Our results suggest

890

A. Kanazawa et al.

Table I. House-Brackmann facial nerve grading system (HB system).

Grade

Description

Characteristics

Normal

Normal facial function in all areas

II

Mild dysfunction

Gross: slight weakness noticeable on close inspection;

may have very slight synkinesis
At rest: normal symmetry and tone
Motion
Forehead: moderate to good function
Eye: complete closure with minimum effort
Mouth: slight asymmetry

III

Moderate dysfunction

Gross: obvious but not disfiguring difference between two sides;

noticeable but not severe synkinesis, contracture, and/or hemi-facial
spasm
At rest: normal symmetry and tone
Motion
Forehead: slight to moderate movement
Eye: complete closure with effort
Mouth: slight weak with maximum effort

IV

Moderately severe dysfunction

Gross: obvious weakness and/or disfiguring asymmetry

At rest: normal symmetry and tone
Motion
Forehead: none
Eye: incomplete closure
Mouth: asymmetric with maximum effort

Severe dysfunction

Gross: only barely perceptible motion

At rest: asymmetry
Motion
Forehead: none
Eye: incomplete closure
Mouth: slight movement

VI

Total paralysis

No movement

that diabetes may have more influence on the

process of axonotmesis than on neurapraxia through
its microangiopathy and low immunity. However,
further experimental studies are needed to confirm
this speculation.
The present analysis showed that the HB system
grades in diabetic patients without complete recovery observed within 6 months of onset are statistically lower than those in nondiabetic patients. Adour
et al. have noted striking differences in the course
H-B
score

and outcome between the DG and NDG who were

untreated for Bells palsy [15]. On the other hand,
Sittel et al. [16] reported observing no significant
differences in complete recovery rate from Bells
palsy between DG (16 patients) and NDG. Their
recovery rates reached 87.5% (14 of 16 patients).
They chose 250 mg intravenous prednisolone administration per day as an initial dose, which is
almost twice as high as the dose used in the present
study (120 mg per day). When comparing our study
%

DG

100

DG

NDG

NDG

80

II
III

60

IV

40

V
*p<0.05

20
*p<0.05

VI
0M

1M

3M

6M

0
0M

Figure 2. Course of averaged HB system grades. Facial movement in the diabetic group was poorer than that in the nondiabetic
group at 3 and 6 months after onset. Collateral bars indicate one
standard deviation. DG, diabetic group; NDG, nondiabetic
group.

1M

3M

6M

Figure 3. Course of recovery rate. The recovery rate in the

diabetic group is lower than that of the nondiabetic group at
6 months after onset. DG, diabetic group; NDG, nondiabetic
group.

Prognosis for Bells palsy

with the study by Sittel et al., it is likely that the
lower dose of prednisolone administration had some
influence on the poor recovery from Bells palsy in
patients with diabetes mellitus in our study. Corticosteroid administration to these patients requires
great caution because it can easily lead to hyperglycemia. Nevertheless, with adequate subcutaneous
insulin injection, there is no reason to hesitate as
regards high-dose steroid administration in patients
with severe Bells palsy with diabetes mellitus for
avoidance of undesirable sequelae.
When electrodiagnostic tests including electroneurography and evoked electromyography (usually
performed 7
14 days after onset) indicate denervation and predict a poor prognosis, diabetic patients
with Bells palsy need additional treatment. For
instance, surgical decompression performed by otologists, which removes the bony facial canal wall to
expose the facial nerve to the middle ear cavity,
prevents degeneration and helps functional recovery
of the nerve, and may arise as a candidate for such
aggressive therapy for diabetic patients with severe
Bells palsy. Fisch [17] emphasized the surgical
indication for Bells palsy as 90% nerve degeneration
reached within 14 days after onset. Grantz et al. [18]
recommended a decompression from the internal
auditory meatal portion to the tympanic portion of
the facial nerve performed within 2 weeks of onset.
Yanagihara et al. [19] mentioned the efficacy of
decompression from the labyrinthine portion to the
mastoid portion of the facial nerve performed within
1 month of onset. The role of surgical decompression in treatments for patients with Bells palsy
remains controversial. Some papers have insisted
that the decompression surgery is unnecessary
[2,20]. Grogan and Gronseth [21] remarked that
there is insufficient evidence to make recommendations regarding facial nerve decompression surgery
for Bells palsy and further studies are necessary to
obtain evidence to present the benefits of surgical
decompression. However, for patients with diabetes
mellitus in whom diabetic neuropathy and angiopathy that may have influence on degeneration and
regeneration of the facial nerve are often observed,
surgical decompression is likely to avoid complete
degeneration and promote recovery from Bells
palsy. We need to reconsider the effectiveness of
decompression surgery for such cases
Bells palsy
with diabetes mellitus
and well-designed studies
are still needed to reveal it.

891

References
[1] Hadar T, Tovi F, Sidi J, Sarov B, Sarov I. Specific IgG and
IgA antibody to herpes simplex virus and varicella zoster
virus in acute peripheral facial palsy patients. J Med Virol
1983;12:237
45.
[2] Gilden DH. Bells palsy. N Engl J Med 2004;351:1323
31.
[3] Kamaratos A, Kokkoris S, Protopsaltis J, Agorgiantis D,
Koumpoulis H, Lentzas J, et al. Simultaneous bilateral facial
palsy in a diabetic patient. Diabetes Care 2004;27:623
4.
[4] Murakami S, Mizobuchi M, Nakashiro Y, Doi T, Hato N,
Yanagihara N. Bell palsy and herpes simplex virus: identification of viral DNA in endoneural fluid and muscle. Ann
Intern Med 1996;124:27
30.
[5] McCormick DP. Herpes-simplex virus as a cause of Bells
palsy. Lancet 1972;29:937
9.
[6] Adour KK, Bell DN, Hilsinger RL Jr. Herpes simplex virus
in idiopathic facial paralysis (Bell palsy). JAMA 1975;233:
527
30.
[7] Devriese PP. Compression and ischaemia of the facial nerve.
Acta Otolaryngol (Stockh) 1974;77:108
18.
[8] Abramsky O, Webb C, Teitelbaum D, Arnon R. Cellular
immune response to peripheral nerve basic protein in
idiopathic facial paralysis (Bells palsy). J Neurol Sci 1975;
26:13
20.
[9] Adour KK, Wingerd J, Doty HE. Prevalence of concurrent
diabetes mellitus and idiopathic facial paralysis (Bells palsy).
Diabetes 1975;24:449
51.
[10] Lawrence DG, Locke S. Motor nerve conduction velocity in
diabetes. Arch Neurol 1961;5:483
9.
[11] Bloomgarden ZT. Clinical diabetic neuropathy. Diabetes
Care 2005;28:2968
74.
[12] Winegrad AI. Diabetic neuropathy. N Engl J Med 1972;286:
1261
2.
[13] House JW, Brackmann DE. Facial nerve grading system.
Otolaryngol Head Neck Surg 1985;93:146
7.
[14] Sittel C, Stennert E. Prognostic value of electromyography
in acute peripheral facial nerve palsy. Otol Neurotol 2001;
22:100
4.
[15] Adour KK, Bell DN, Wingerd J. Bell palsy: dilemma of
diabetes mellitus. Arch Otolaryngol 1974;99:114
7.
[16] Sittel C, Sittel A, Guntinas-Lichius O, Eckel HE, Stennert
E. Bells palsy: a 10-year experience with antiphlogisticrheologic infusion therapy. Am J Otol 2000;21:425
32.
[17] Fisch U. Maximal nerve excitability testing vs electroneuronography. Arch Otolaryngol 1980;106:352
7.
[18] Grantz BJ, Rubinstein JT, Gidley P, Woodworth GG.
Surgical management of Bells palsy. Laryngoscope 1999;
109:1177
88.
[19] Yanagihara N, Hato N, Murakami S, Honda N. Transmastoid decompression as a treatment of Bell palsy. Otolaryngol
Head Neck Surg 2001;124:282
6.
[20] Adour KK. Decompression for Bells palsy: why I dont do
it. Eur Arch Otorhinolaryngol 2002;259:40
7.
[21] Grogan PM, Gronseth GS. Practice parameter: steroids,
acyclovir, and surgery for Bells palsy (an evidence-based
review): report of the quality standards subcommittee of
the American Academy of Neurology. Neurology 2001;56:
830
6.
/