Blink Recovery in Patients with Bell’s Palsy: A Neurophysiological and Behavioral Longitudinal Study

Abstract

purpose. To examine the recovery process of blinking in a longitudinal study of nine patients severely affected by Bell’s palsy.

methods. Kinematics of bilateral eyelid and eye movements and concomitant orbicularis oculi activity during voluntary blinking and air-puff– and acoustic-click-induced reflex blinking were determined by using the magnetic search coil technique and electromyographic recording of the orbicularis oculi muscle (OO-EMG).

results. In the first 3 months of absence of OO-EMG activity, reduced eyelid and eye movement of the palsied eyelid were observed during all types of blinking. First OO-EMG activity was determined 3 months after onset of the affliction. After 1 year, OO-EMG activity was normalized and showed values similar to those on the nonpalsied side. Clinically, eyelid movements were normal after 1 year, although corresponding maximum amplitudes and corresponding velocities were two times smaller, expressed in reduced eyelid motility. Directions of eye movement during reflex blinking were normal after 1 year, although maximum amplitudes were smaller on the palsied side. Eye movements during voluntary blinking remained impaired. A simultaneous horizontal upward shift of both eyes in the same direction was recorded throughout the study.

conclusions. Although OO-EMG activity on the palsied side was normalized 1 year after onset of the affliction, the accompanying eyelid movements and their maximum amplitudes and velocities remained smaller throughout the study. The consistent impairment of eye movements in voluntary blinking during the study and reduced motility of eyelid movements indicates that higher brain structures, which modify eyelid and eye movement control during blinking, may be altered by the affliction.

Blinking is imperative to protect the eye against corneal drying and damage. ¹ The characteristic eye and eyelid movements during a blink obtain an optimal tear film distribution over the cornea. Disturbances in this blinking pattern can lead to inflammation of the eye and, if not treated, to functional blindness. 

Eyelid movement during blinking is mainly mediated by the levator palpebrae superioris and orbicularis oculi (OO) muscles. ^{2 3 4} Humans exhibit three types of blinks: spontaneous, voluntary, and reflex blinks. ^{3 5} During blinking stereotypical eye movements occur. ^{6 7} In healthy subjects, the eyeball rotates from a straight-ahead position nasally downward, directly followed by laterally upward movement during blinking. A slight displacement of the eyeball occurs in the orbit, ⁶ involving a horizontal and vertical rotation ^{8 9} and torsion. ^{10 11} The underlying neuronal structure initiating eye movement during blinking is not known, but recently evidence was obtained that showed that specific areas in the lateral medullar reticular formation are involved in eyelid and eye movements during blinking. ^{12 13}  

Patients with Bell’s palsy, a form of unilateral facial nerve palsy, characteristically have unilateral peripheral facial weakness and are unable to blink on the palsied side. ¹⁴ Those with complete clinical facial paralysis also have decreased tearing on the palsied side ^{15 16} and eye irritation. The worldwide incidence of this affliction is approximately 30 cases per 100,000 people per year. ¹⁶ Clinically, facial functions appear to recover completely in these patients, although a significant number have residual facial weakness and synkinesia. Bell’s palsy patients, classified as House-Brackmann grade VI, ¹⁷ show complete absence of excitability of the facial nerve within 1 week. A main feature of this grade of affliction is that blink restoration will not occur sufficiently. Blink restoration remains completely absent in some patients, and surgical interference is the only way for them to regain eyelid motility. 

Research on blinking of patients with Bell’s has mainly concentrated on reflex blinks elicited by electrical stimulation of the supraorbital (SO) nerve. ^{18 19 20 21 22 23} Analysis of eyelid kinematics during spontaneous and voluntary blinks has shown various degrees of recovery in patients with unilateral facial nerve palsy. ²⁰ In other studies the OO-EMG recording technique was used to investigate the recovery of OO-muscle activity at different stages of facial nerve palsy. ^{18 21 22 23} Adaptive changes of eyelid movements on the nonpalsied side were observed. ^{24 25} Whether adaptive changes and normal kinematical values of eyelid movement during blinking will be reached after recovery is not known. To examine these phenomena during blink recovery, eye and eyelid kinematics and OO-EMG features were monitored in a longitudinal study. 

Materials and Methods

Subjects

Nine patients (six men and three women) with Bell’s palsy, House-Brackmann grade VI, were included in the study (Table 1) . Inclusion criteria were: acute unilateral facial paralysis, onset and deterioration within 48 hours, and a facial nerve impervious to excitation within 10 days after onset of the affliction. All patients were given prednisolone, 60 mg orally for 5 days, decreasing to 10 mg at day 10, after which medication was stopped. Exclusion criteria were: viral Herpes simplex I, Varicella zoster, and diabetic cranial neuropathy (all established with laboratory tests) otological disorders, history of cerebral vascular accident, psychiatric history, and pregnancy. The mean age of the patients at the onset of palsy was 46 ± 12 years (mean ± SD; range, 29–67). The mean duration of disease was 16 ± 3.7 months (mean ± SD). All subjects gave informed consent according to the Declaration of Helsinki, and the Medical Ethics Committee of Erasmus MC University approved the study. 

Eyelid Movements, Eye Movements, and OO-EMG Recordings

Movements of both eyelids were measured with the magnetic search coil technique. ^{26 27} Detailed procedures have been described previously. ⁵ Although the subject looked straight ahead, a search coil (2.2 mm diameter, 20 turns, 20 mg) of copper wire (0.1 mm diameter) was placed 1 mm above the rim of the upper eyelid above the center of the pupil. Movements of both eyes were also measured with the magnetic search coil technique using standard scleral search coils (Skalar Medical BV, Delft, The Netherlands). Before insertion of the search coil, the eye was anesthetized with 2 drops of the topical ophthalmic anesthetic oxybuprocaine 0.4% (Novesin; CIBA Vision,. Ltd., Hettlingen, Switzerland). A previously described device was used to calibrate eye and eyelid rotations. ⁵  

For OO-EMG recordings, two 6-mm diameter Ag/AgCl surface electrodes were used. The active recording electrode was placed approximately 10 mm below the margin of the lower eyelid, and the reference-recording electrode was placed 10 mm lateral to the temporal margin of the eyelids. A self-attached circle electrode (10 × 10 mm) was positioned on the forehead and served as the ground electrode. 

Experimental Procedures

Blinking was recorded in the patients every 6 weeks. Three types of blinks were recorded: (1) voluntary; (2) acoustic-click–induced, and (3) corneal trigeminal reflex (air-puff–induced, ipsi- and contralateral stimulated). The mean number of recordings per subject was 13. Of the 52 recordings, 16 were simultaneous eye and eyelid recordings. The subject was seated comfortably in a chair, and the subject’s head was stabilized with a head holder by fixing his or her chin in a standard position. The measured eyelid/eye was placed near the center of the magnetic field. 

Voluntary Blinks

Subjects were asked to focus on a fixation point in the middle of a 2 × 2-m flat transparent screen at a distance of 82 cm and to respond with a gentle blink as short as possible every time the fixation point disappeared for 50 ms. We recorded at least 15 blinks with a 6-second interval between successive blinks. 

Trigeminal Reflex Blinks

Trigeminal blink reflexes were elicited by an air-puff stimulus (2 bar, 10 ms). For this purpose, the end of a silicon rubber tube (diameter 1 mm) was directed toward the lateral rim of the iris, 2 to 3 mm from the eyeball. Air-puff–induced blinks were recorded with random stimulus intervals between 25 and 30 seconds. At least 15 registrations were made both after stimulation of the nonpalsied and the palsied sides. 

Acoustic-Click–Induced Reflex Blinks

Blinks were elicited by an acoustic click of 90 dB SPL (sound pressure level), with a duration of 10 ms. The source box was placed 60 cm from the subject’s head, at the level of the external ear on the nonpalsied side. Ten registrations were made with random stimulus intervals between 30 and 40 seconds. 

Data Acquisition and Statistical Analysis

Data acquisition was achieved as described previously. ⁵ For each trial, the computer displayed bilateral OO-EMG and eyelid and eye movements in vertical and horizontal directions. The raw data were stored on disks for off-line analysis. 

For eyelid movements, we analyzed the start time, the down-phase duration, the up-phase duration, the maximum down-phase amplitude, the maximum down velocity, the time of maximum amplitude, and the time of maximum velocity. For OO-EMG onset latency and integrated OO-EMG until the time of maximum amplitude of the eyelid movement were analyzed for voluntary, air-puff–, and acoustic-click–induced blinks. We also determined the delay between the onset of OO-EMG and the onset of eyelid movement for all types of blinks. Eye movements during blinking were depicted by their angle and maximum amplitude according to a polar coordinate system. 

In the text and tables standard errors of the mean (SEM) are indicated unless otherwise mentioned. If the data of chosen parameters were not normally distributed, they were logarithmically transformed, after which statistical analysis was performed. 

Ratios of the integrated OO-EMG of the palsied and nonpalsied eyelid and kinematics were determined for the up phase, the maximum amplitude and velocity. Ratios were logarithmically averaged and values of P < 0.05 were considered significant. 

Results

There was a significant difference in eyelid kinematics between all types of blinks (Table 1) . At the onset of the affliction, the OO-EMG activity was absent in the palsied eyelid, and both eye and eyelid movements of both eyes were disturbed. Almost no recovery of eyelid movement on the palsied side was recorded within the first 18 weeks. At the end of the study, the blink seemed clinically recovered, though the maximum amplitude and maximum velocity remained different between both eyelids during voluntary and reflex blinking, this resulted in reduced eyelid motility (Figs. 1A 1B 2C 2F) . Eye movements remained impaired during voluntary blinking (Fig. 2C) . 

Eyelid Kinematics

The start times of the movement of both palsied and nonpalsied eyelids showed a clear time delay in the first 18 weeks (Fig. 3A ; Table 1 ). The nonpalsied eyelid exhibited a slight shortening of the start time until 1 year. 

The difference of start times of OO-EMG and eyelid movement shortened directly after onset of the affliction and reached normal values at 18 weeks. The mean start time difference for voluntary blinking, air-puff– (nonpalsied/palsied side induced), and acoustic-click–induced of the nonpalsied eyelid was 15.8 ± 2.8, 12.5 ± 1.4, 11.1 ± 1.3, and 10.6 ± 1.1 ms, respectively. For the palsied eyelid the mean values were: 13.8 ± 4.5, 13.7 ± 1.3, 12.5 ± 1.3 and 11.4 ± 0.9 ms, respectively. The start times in both eyes were not significantly different. 

In voluntary- and air-puff–induced reflex blinks, the down-phase duration of the palsied side was prolonged during the first 18 weeks. After 18 weeks’ active eyelid movements reverted and the down-phase duration became shorter, although measured values never reached values on the nonpalsied side (Table 1) . On the nonpalsied side, the down-phase duration slightly shortened from onset up to 1 year. 

The up-phase duration of the palsied eyelid was always shorter than that of the nonpalsied eyelid, except for acoustic-click–induced blinking (Fig. 3B , Table 1 ). 

The maximum amplitude and velocity increased significantly on the palsied side at least until 84 weeks (Figs. 3C 3D ; Table 1 ). The largest velocity increase was measured within 36 weeks, followed by a small continuous increase from 782 to 923 deg/s for the air-puff–induced blink stimulated on the palsied side at 84 weeks. However, these values never reached values measured on the nonpalsied side. On the nonpalsied side, no significant changes in the maximum amplitude and velocity were measured during the study. 

On the palsied side, the time of maximum amplitude and velocity shortened after 18 weeks until the end of the study, but they never reached the values measured on the nonpalsied side, except for acoustic-click–induced blinks (Fig. 3E , Table 1 ). 

Voluntary Blinks.

The difference between the down-phase duration of palsied and nonpalsied eyelids, measured at the end of the study, was the largest of all types of blinks (e.g., 28 ms; Table 2 ). The down- and up-phase duration of the nonpalsied eyelid ranged during the study between 65 and 84 ms and between 194 and 265 ms, respectively. 

The maximum amplitude and velocity of the palsied eyelid was increased at 18 weeks from 4.8° at onset to 19.7° and from 65 to 331 deg/s, respectively (Table 1) . 

Reflex Blinks.

Independent of the stimulation side, air-puff–induced blinks showed shortening of the start time of eyelid movement at 18 weeks. When the eye was stimulated on the nonpalsied side, shortening of start time of the palsied eyelid continued until the end of the study from 49 ms at 18 weeks to 41 ms. On the nonpalsied side the shortening was from 60 ms at 18 weeks to 40 ms at 84 weeks. 

The down- and up-phase durations of both eyelids slowly shortened during the study. 

When the palsied eye was stimulated, the maximum amplitude and velocity of the palsied eyelid was increased at 18 weeks from 7.3° at onset to 19.5° and from 123 to 395 deg/s, respectively (Table 1) . After stimulation of the nonpalsied side, the maximum amplitude was 1.28 ± 0.04 (P < 0.00002) times larger than after stimulation of the palsied side. The maximum velocity was 1.24 ± 0.05 (P < 0.0002) times larger. 

When the nonpalsied eye was stimulated, the maximum amplitude and velocity of the palsied eyelid was increased at 18 weeks from 5.1° at onset to 18.4° and from 109 to 450 deg/s, respectively (Table 1) . The maximum amplitude of the nonpalsied eyelid was 1.10 ± 0.02 (P < 0.0005) times larger than the amplitude measured after stimulation on the palsied side; the maximum velocity was 1.11 ± 0.03 (P < 0.005) times larger. 

In acoustic-click–induced reflex blinks, the start times of both eyelids were comparable after 18 weeks (Table 1) . 

The down-phase durations of nonpalsied and palsied eyelids ranged between 49 and 59 ms and between 53 and 64 ms, respectively, during the study. The up-phase duration of the palsied eyelid increased from 116 at onset to 179 ms until the end of the study. 

The maximum amplitude and velocity of the palsied eyelid was increased at 18 weeks from 3.0 to 5.6 deg/s and from 107 to 146 deg/s, respectively (Table 1) . 

Electromyography of the OO Muscle

At onset of the affliction, the start times of the OO-EMG were not measurable on the palsied side, because the integrated OO-EMG measured on the palsied side was at noise level in all types of blinking examined. Six weeks later, the start time of the OO-EMG of the palsied eyelid was measurable in reflex blinks and shortened until the end of the study (Fig. 3H , Table 1 ). On the nonpalsied side, the start time of the OO-EMG remained constant. At 36 weeks, start times of eyelid movement and OO-EMG and integrated OO-EMGs of both eyes were no longer significantly different (Table 2) . 

The integrated OO-EMG of the palsied eyelid increased in the first 18 weeks, followed by a continuous small increase until 84 weeks (Table 1) . In reflex blinking, OO-EMG of the palsied eyelid was reset after 1 year and showed values similar to the OO-EMG of the nonpalsied eyelid (Fig. 3F) . The integrated OO-EMG of the nonpalsied eyelid slightly decreased during the study. In individual cases, an overshoot of the integrated OO-EMG on the palsied side was recorded between 36 weeks and 1 year. After 1 year, values were comparable with those measured on the nonpalsied side. 

The sum of the integrated OO-EMG of palsied and nonpalsied eyelids in all measurements was almost constant throughout the study (Fig. 3G) . 

Voluntary Blinks.

On the nonpalsied side, the integrated OO-EMG decreased only from 4.8 to 2.8 μV during the study. On the palsied side the integrated OO-EMG increased from 0.7 at onset to 3.5 μV at 84 weeks (Table 1) . 

Reflex Blinks.

Although in all reflex blinks the start times of the OO-EMG of the palsied eyelid were not measurable at onset, the OO-EMG continuously decreased from around 45 at 18 weeks to 28 ms at the end of the study. The start time of the nonpalsied eyelid was measurable at onset, and the values were stable after 18 weeks. Mean values measured from 18 weeks until the end of the study are 28.9 ± 0.6, 33.1 ± 0.6, and 38.5 ± 0.8 ms for air puff on the nonpalsied side, air puff on the palsied side, and acoustic click of the nonpalsied eyelid, respectively. If the eye on the nonpalsied side was stimulated with an air puff, the start time of the OO-EMG on the nonpalsied side was 4.2 ± 0.7 ms (P < 0.0001) shorter than if the eye on the palsied side was stimulated. 

The integrated OO-EMG of the nonpalsied eyelid decreased independent of the stimulation side until 18 weeks and then ranged between 3.2 and 4.5 μV. The integrated OO-EMG of the palsied eyelid, measured after air-puff stimulation on the nonpalsied side, increased from 0.7 at onset to 3.0 μV at 84 weeks. If the eye was stimulated with an air puff on the palsied side, the integrated OO-EMG of the palsied eyelid was 1.19 ± 0.05 times (P < 0.001) larger than if the eye on the nonpalsied side was stimulated. 

In acoustic-click–induced blinks, the integrated OO-EMG of the palsied eyelid, increased until 36 weeks and then remained constant. The integrated OO-EMG of the nonpalsied eyelid was at onset 1.3 μV and was almost stable until the end of the study, 1.4 μV. 

Eye Movements

The maximum amplitude and time of maximum amplitude of eye movement were recorded during voluntary blinking and air-puff–induced and acoustic-click–induced reflex blinking 30 and 72 weeks after onset of the affliction. In general, directly after onset, impaired eye movements were observed during blinking. 

In all types of blinking, the maximum amplitude of eye movements measured on the palsied side always remained two times smaller than that on the nonpalsied side (Table 3) . Maximum amplitudes did not change significantly between 30 and 72 weeks on both the palsied and nonpalsied sides during reflex blinking. 

After 72 weeks, the direction of eye movement remained impaired in voluntary blinking, but was normal in reflex blinking. Of note, at 30 weeks, a shift in direction was observed in the rotation of both eyes during voluntary and reflex blinking (Figs. 2B 2E) . The vertical component was never more than 2°; the horizontal component ranged between 5° and 10°. The sweep of both eyes during voluntary blinking was much more pronounced than during reflex blinking (Figs. 2B 2C 2E 2F ; Table 3 ). No relation was found between the direction of the shift and the side of affliction. In five subjects, eye movements during blinking were directed toward the palsied side and in the remaining four toward the nonpalsied side. 

Voluntary Blinks.

Differences between the maximum amplitude of both eyes did not change significantly. 

The amplitude ratio slightly decreased from 30 until 72 weeks (Table 3) . Aberrant eye movements of both eyes remained until the end of the study. 

Reflex Blinks.

With air-puff stimulation, maximum amplitudes of eye movements on the palsied side remained between 2.3° and 2.9° and on the nonpalsied side between 5.5° and 8.7° during the study (Table 3) . 

The time of maximum amplitude was the same for both eyes at 72 weeks. On the nonpalsied side, values measured at 30 and 72 weeks did not differ independent of the side of stimulation. 

In acoustic-click–induced blinks the maximum amplitude increased slightly during the study. The time of maximum amplitude and time difference between maximum amplitudes remained constant. 

Discussion

The present study describes blink recovery in nine patients with a complete (House-Brackmann grade VI) Bell’s palsy, who were given prednisolone. The recovery of OO-EMG and eyelid kinematics occurred in phases. 

In the first phase, start times of OO-EMG and lid movement synchronized between onset of the affliction and 18 weeks. During the second phase, which lasted from 18 weeks until 36 weeks after onset of the affliction, the palsied eyelid of these patients showed the first signs of OO-EMG activity and active eyelid movement at the same time. At the end of the second phase, OO-EMG values of both eyes were no longer significantly different (Table 2) . 

The third phase was characterized by overshoot of OO-EMG activity of the palsied eyelid. The overshoot was no longer measurable approximately 1 year after onset of the affliction. Together with the increase of OO-EMG activity on the palsied side, a decrease in OO-EMG activity was observed on the nonpalsied side. The sum of OO-EMG of both eyelids remained almost constant throughout the study. In the fourth phase, subtle increases in maximum amplitude and velocity were found. However, these increases were not significant. Except for the start times of the eyelid movements, we found that recovery of eyelid movements on the palsied side during reflex blinking was incomplete at 84 weeks. 

Of note, the direction of the eye movements during voluntary blinking remained impaired, but the direction was normal in reflex blinking 1 year after onset of the affliction. The maximum amplitudes of eye movements remained aberrant throughout the study. 

Eyelid Kinematics

This longitudinal study suggests that the recovery of Bell’s palsy induced adaptations in eyelid kinematics. Probably, motoneuron excitability was changed, which resulted in eyelid asymmetry during blinking. ²⁸  

At onset of the paralysis, the palsied eyelid generated only a passive down phase of eyelid movement during all types of blinking. The down phase of the palsied eyelid was not in concert with the down phase of the nonpalsied eyelid. After 18 weeks, signs of active down-phase movement on the palsied side appeared for the first time, about 1 to 2 months later than is described in the literature. ^{21 22 24}  

The maximum amplitudes and velocities measured on the palsied side, never reached the levels of those in healthy volunteers. ⁵ This is another example of reorganization of motor units in the OO muscle. On the palsied side, this ongoing reorganization modifies motoneuron excitability. ²⁸ Maximum amplitude and velocity values on the nonpalsied side did not differ significantly from our previous study. ⁵ Huffman et al. ²⁰ had similar observations in a group of unrecovered patients during spontaneous and voluntary blinking. In contrast, they found normal amplitudes and peak velocities in a group of patients clinically recovered from Bell’s palsy. The grade of affliction of those patients was not indicated, and the appearance of synkinesia was not mentioned, whereas we had already observed synkinesia in all our patients after 6 months. 

Aberrant regeneration of peripheral facial nerve palsy was observed earlier in a primate study ²⁹ and in a study of 29 patients. ²¹ In the latter study of the OO muscle, activity was found at least 4 months after facial nerve degeneration. These investigators also observed involuntary facial movement disorders, which were recently confirmed in a patient with Bell’s palsy who had signs of blepharospasm on the nonpalsied side. ³⁰ We did not find any sign of blepharospasm in our patients. 

Voluntary Blinks.

From the onset of the paralysis, the down- and up-phase duration and maximum amplitude on the nonpalsied side showed clear fluctuations after the recovery. This means that the neural blinking circuits are activated on both the palsied and nonpalsied sides. ³¹ We found that these duration adaptations are transient and independent of incomplete recovery of kinematics of the palsied eyelid, although a reduced maximum amplitude and velocity was still present at the end of the study. 

Several brain regions are involved in voluntary and spontaneous blinking. ³² Specific regions in the motor cortex, including regions in control of eyelid and mouth performance project directly or indirectly toward the facial motor nucleus. ^{33 34 35} Kaneko et al. ³⁶ showed neural activity of the supplementary motor area during voluntary blinks. Transcranial magnetic stimulation showed that the cortical center, initiating upper facial movements, including blinking, is located in the mesofrontal region rather than in the facial region of the primary motor cortex. ³⁷ Subcortical dopaminergic pathways appear to play a role in the inhibition of the levator palpebrae superioris and/or inactivation of the OO muscle during voluntary blinks. ² These pathways may change during the paralysis, because the representation fields of facial muscles like the OO and orbicularis oris muscles in the primary motor cortex probably alter in composition and size during recovery from this affliction. 

Reflex Blinks.

Several studies have shown that movements of palsied eyelids during reflex blinking are reduced during and after recovery. ^{25 29 38 39} The characteristics of air-puff– and acoustic-click–induced reflex blinks found in the present study were consistent with each other and confirmed that reflex blinks have the shortest down and up phases of all types of blinks. ^{5 6 28} Of note, the present study showed that acoustic-click–induced reflex blinks recovered fastest. 

Schicatano et al. ²⁵ observed a difference in synchronism of eyelid closure during contralateral SO stimulation at the onset of unilateral facial nerve palsy. This asynchronism was accompanied by blink oscillations that decreased during recovery. In that study, a reduced amplitude and a nonsignificant increased down-phase duration of the eyelid movement was observed as well. Stimulation on the palsied side prolonged the blink duration, and “blink oscillations” were often seen. These oscillations may arise from oscillatory processes within trigeminal reflex blink circuits developed as a consequence of aging of the sensory trigeminal complex neurons or decreased lacrimation associated with aging. ³⁹ We had similar observations after air puff stimulation on the palsied side (e.g., reduced motility and blink oscillations). However, when the nonpalsied side was stimulated, hardly any blink oscillations were found. 

Electromyography of the OO Muscle

Testing the blink reflex using OO-EMG may predict the recovery trajectory of Bell’s palsy patients. ²² At onset of the affliction, OO-EMG activity during voluntary blinks and click and air-puff–induced blinks was at noise level. For all types of blinks, we registered both integrated OO-EMG, and OO-EMG start time on the palsied side did not show a significant difference from that on the nonpalsied side after 36 weeks. Of note, the sum of integrated OO-EMG of palsied and nonpalsied eyelids was almost constant throughout the study, indicating that both facial motoneuron pools possibly receive a common cortical input from representation fields M2 and M3 in the motorcortex. ³⁴  

Another negative effect, which was present in all patients, was the development of facial weakness during the disease. In individual patients, an overshoot of OO-EMG on the palsied side was found between 36 weeks and 1 year. Despite improvement in maximum amplitude and maximum velocity of the eyelid, facial weakness remained. This together with synkinesia creates less effective eyelid movement during blinking. ¹⁹ Our finding is not in agreement with observations in a human Bell’s palsy study in which larger OO-EMG responses on the palsied side after recovery of the affliction were found, even after 18 months. ⁴⁰  

Facial symmetry of patients with late or partially recovered Bell’s palsy can be improved nonsurgically, with facial physiotherapy or botulinum toxin injections, or surgically. ⁴¹ However, the result is often unsatisfactory and few alternatives are available. Electrical stimulation of facial muscles during recovery is one of these alternatives to decrease facial weakness. Although a 3-month therapeutic electrical stimulation improved the eyelid movement, no significant improvement was recorded in spontaneous maximum amplitudes and maximum velocities. ⁴² An electrophysiological study in patients with Bell’s palsy showed that patients with residual facial weakness had enhanced reflex blink recovery after electrical SO nerve stimulation on the palsied side. ¹⁹ Cossu et al. ²³ suggested starting “treatment” approximately 3 months after the onset of complete facial nerve palsy, the period of the very first connections between growing axons and denervated muscles. At this stage very little muscle activity was detected in our patients. Starting therapy at that stage risks an overstimulation (activation) of facial muscles on the nonpalsied side, which could result in pronounced facial asymmetry. To prevent the overstimulation, prudence should be used with the start of the training. We recommend starting directly after the first signs of innervation of both the OO and orbicularis oris muscles. 

Eye Movements during Blinking

It was Bell ⁴³ who first noticed that patients with hemifacial paralysis moved the eye upward during blinking on the palsied side. Using standard scleral search coils, we recorded eye movements during voluntary, acoustic-click– and air-puff–induced blinks. 

Until 30 weeks, eye movements during all types of blinking were disturbed; their direction was oblique upward, parallel to each other, but not straight up, as described previously. ⁴⁴ After 72 weeks, eye movements during reflex blinking were normal in direction, but had smaller amplitudes on the palsied side. On the nonpalsied side, amplitudes were normal. ⁶  

During voluntary blinking, eye movements of both eyes were enlarged and the direction remained disturbed. Specifically, the vertical eye component remained smaller than the values measured in healthy subjects. The exceptional increase in amplitude of eye movements during voluntary blinking suggests that, besides the facial nerve, other cranial nerves are involved in the etiology of Bell’s palsy. This implies that Bell’s palsy is a cranial neuropathy, as was suggested by Adour et al. ¹⁵  

Contrary to our results, a restraint eyelid study of Collewijn et al. ⁸ showed no alteration in eye movements during blinking. Voluntary blinking was recorded with one eyelid kept open by adhesive tape. Relatively large blink-related eye movements were observed. The relatively large eye movement during voluntary blinking in the lid restraint experiment may not be caused by a mechanical factor, as Bour et al. ⁶ showed that the extent of eye rotation depends on the initial eye position. In the present study, large eye movements during blinking were seen as well, although the direction of eye movements was disturbed. 

Adaptive Changes

Bell’s palsy can lead to changes in the blinks of the nonpalsied eyelids. ^{30 45} These changes may result in a Bell’s palsy-induced blepharospasm, which was observed in the nonpalsied eyelid. Fortunately, treatment with apomorphine or other dopamine receptor agonists’ can reduce blepharospasm and blink excitability and possibly change motoneuron recruitment. ³⁰ We observed blink oscillations on the palsied side; however, none of the nine patients with Bell’s palsy revealed signs of involuntary eyelid movement disorders on the nonpalsied side. 

It is stated in a lid-restraint study that unilateral facial nerve palsy is a long-term version of the adaptive process of eyelid movement after lid restraint ²⁵ and that hyperexcitability of the neural circuit of the blink reflex during an acute period of facial palsy is an adaptive response to compensate impaired facial motor function. ²⁸ We think that these functional changes are more complex and are also determined centrally, as voluntary and, to a lesser extent, reflex blinking remained disturbed in our group of patients. This persistent disturbed blinking may be the result of changes in representative cortical areas of eye and eyelid movements developed during recovery. ⁴⁶ Therefore, a proper facial function needs peripheral and central adaptations in facial motor control. Further studies are needed to elucidate the role of the sensory and motor cortex during the recovery process. 

 Submitted for publication May 2, 2006; revised September 14, 2006; accepted November 27, 2006.

 Disclosure: F. VanderWerf, None; D. Reits, None; A.E. Smit, None; M. Metselaar, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

OO-EMG and Eyelid Kinematics at Four Moments during Recovery of Voluntary and Reflex Blinking

Type of Blinking	OO-EMG and Eyelid Kinematics	Onset			18 Weeks			36 Weeks			84 Weeks
				Nonpalsied	Palsied	Nonpalsied	Palsied	Nonpalsied	Palsied	Nonpalsied	Palsied
Voluntary
	Integrated OO-EMG (μVs)	4.8 ± 0.7	0.7 ± 0.2	4.1 ± 0.9	2.1 ± 0.6	3.6 ± 0.5	3.1 ± 0.6	2.8 ± 0.3	3.5 ± 0.5
	Start lid - start OO-EMG (ms)	19 ± 5	Indefinite	9 ± 6	15 ± 2	12 ± 3	16 ± 2	29 ± 5	30 ± 3
	Down phase (ms)	77 ± 8	128 ± 6	69 ± 6	135 ± 12	84 ± 9	121 ± 8	65 ± 6	93 ± 4
	Up phase (ms)	265 ± 40	171 ± 40	212 ± 19	173 ± 22	194 ± 21	175 ± 22	220 ± 13	160 ± 23
	Maximum amplitude (°)	44.0 ± 6.7	4.8 ± 0.8	48.7 ± 3.7	19.7 ± 4.0	46.3 ± 2.9	31.2 ± 4.4	36.9 ± 2.0	27.1 ± 4.3
	Maximum velocity (deg/s)	1304 ± 145	65 ± 7	1359 ± 135	331 ± 87	1379 ± 78	650 ± 117	1204 ± 74	676 ± 120
Air puff-induced, nonpalsied eyelid
	Start OO-EMG (ms)	27 ± 2	Indefinite	32 ± 1	43 ± 5	31 ± 2	34 ± 2	27 ± 2	28 ± 2
	Integrated OO-EMG (μVs)	5.8 ± 1.2	0.7 ± 0.2	3.9 ± 0.7	2.4 ± 0.7	3.6 ± 0.6	2.9 ± 0.5	3.2 ± 0.4	3.0 ± 0.8
	Start time (ms)	56 ± 4	75 ± 6	49 ± 3	60 ± 5	40 ± 1	44 ± 1	41 ± 4	40 ± 2
	Start lid - start OO-EMG (ms)	29 ± 7	Indefinite	17 ± 2	17 ± 7	9 ± 1	10 ± 2	14 ± 2	11 ± 1
	Down phase (ms)	90 ± 23	97 ± 21	65 ± 6	93 ± 16	78 ± 8	93 ± 12	63 ± 5	81 ± 8
	Up phase (ms)	294 ± 20	213 ± 13	237 ± 17	184 ± 11	199 ± 18	176 ± 19	207 ± 28	171 ± 26
	Maximum amplitude (°)	47.9 ± 5.0	5.1 ± 1.0	46.0 ± 2.6	18.4 ± 4.8	44.7 ± 3.1	27.1 ± 4.9	39.8 ± 4.1	24.4 ± 5.3
	Maximum velocity (deg/s)	1675 ± 109	109 ± 9	1638 ± 92	460 ± 150	1566 ± 123	715 ± 133	1569 ± 136	748 ± 201
	Time maximum amplitude (ms)	146 ± 27	172 ± 21	114 ± 3	153 ± 18	118 ± 8	137 ± 12	104 ± 3	121 ± 9
	Time maximum velocity (ms)	70 ± 1	98 ± 9	72 ± 2	88 ± 4	69 ± 3	90 ± 4	71 ± 3	82 ± 3
Air puff-induced, palsied eyelid
	Start OO-EMG (ms)	31 ± 2	Indefinite	34 ± 2	41 ± 4	34 ± 2	31 ± 2	30 ± 3	26 ± 2
	Integrated OO-EMG (μVs)	7.7 ± 2.8	1.1 ± 0.2	4.5 ± 0.6	2.6 ± 0.6	3.0 ± 0.4	3.5 ± 0.9	3.7 ± 0.9	3.9 ± 1.3
	Start time (ms)	57 ± 3	82 ± 6	46 ± 2	54 ± 4	41 ± 2	42 ± 2	40 ± 5	38 ± 4
	Start lid - start OO-EMG (ms)	26 ± 3	Indefinite	12 ± 4	12 ± 7	7 ± 1	11 ± 2	10 ± 3	12 ± 4
	Down phase (ms)	101 ± 25	124 ± 28	80 ± 11	115 ± 18	73 ± 6	85 ± 7	74 ± 6	99 ± 18
	Up phase (ms)	285 ± 36	205 ± 30	273 ± 34	213 ± 30	224 ± 35	211 ± 36	205 ± 22	163 ± 13
	Maximum amplitude (°)	45.3 ± 4.6	7.3 ± 2.5	44.3 ± 3.6	19.5 ± 3.9	40.7 ± 3.3	28.7 ± 4.8	38.7 ± 3.1	32.1 ± 4.6
	Maximum velocity (deg/s)	1596 ± 125	123 ± 37	1569 ± 126	395 ± 100	1532 ± 144	782 ± 138	1561 ± 120	923 ± 144
	Time maximum amplitude (ms)	152 ± 17	206 ± 33	126 ± 9	169 ± 17	114 ± 6	127 ± 7	114 ± 7	136 ± 20
	Time maximum velocity (ms)	72 ± 2	102 ± 12	73 ± 3	98 ± 7	72 ± 2	83 ± 4	72 ± 5	82 ± 5
Acoustic click-induced
	Start OO-EMG (ms)	39 ± 3	Indefinite	39 ± 1	45 ± 1	41 ± 1	40 ± 2	37 ± 2	37 ± 1
	Integrated OO-EMG (μVs)	1.3 ± 0.1	0.2 ± 0.2	1.2 ± 0.2	1.0 ± 0.2	1.2 ± 0.2	1.2 ± 0.2	1.4 ± 0.1	1.3 ± 0.1
	Start time (ms)	56 ± 4	62 ± 4	49 ± 2	54 ± 3	48 ± 2	50 ± 2	49 ± 4	51 ± 3
	Start lid - start OO-EMG (ms)	16 ± 8	Indefinite	10 ± 2	11 ± 3	7 ± 2	10 ± 3	11 ± 6	14 ± 4
	Down phase (ms)	52 ± 8	54 ± 13	53 ± 2	57 ± 5	53 ± 3	57 ± 6	54 ± 4	54 ± 6
	Up phase (ms)	155 ± 28	116 ± 20	150 ± 17	140 ± 17	149 ± 15	151 ± 14	202 ± 28	179 ± 29
	Maximum amplitude (°)	22.1 ± 5.3	3.0 ± 0.5	22.0 ± 4.3	5.6 ± 1.1	21.5 ± 4.8	12.2 ± 2.5	17.5 ± 5.4	12.5 ± 3.1
	Maximum velocity (deg/s)	790 ± 180	107 ± 31	813 ± 168	146 ± 31	797 ± 184	373 ± 76	630 ± 189	439 ± 117
	Time maximum amplitude (ms)	108 ± 7	116 ± 11	102 ± 2	111 ± 7	99 ± 2	105 ± 4	103 ± 1	105 ± 4
	Time maximum velocity (ms)	79 ± 1	111 ± 26	80 ± 2	85 ± 5	80 ± 2	84 ± 2	84 ± 1	85 ± 1

 The differences between both eyelids and their significances are presented in Table 2 .

Figure 1.

 

Simultaneously recorded OO-EMG activities, upper eyelid and eye movements during blinking in one subject. (A) Superposition of six successive traces of voluntary blinks and their OO-EMG activities. (B) Superposition of six successive traces of palsied corneal air puff-induced blinks and their OO-EMG activities. Lines 1 and 2: OO-EMG activities in the palsied and nonpalsied eyelid, respectively; lines 3 and 5: vertical displacement; lines 4 and 6: the horizontal eyelid displacement at palsied and nonpalsied sides; lines 7 and 9: represent the vertical eye displacement at palsied and nonpalsied sides; and lines 8 and 10:represent the horizontal eye displacement at palsied and nonpalsied sides. Note the abnormal vertical displacement (lines 7 and 9) and the large horizontal displacement (lines 8 and 10) of the eyes during voluntary blinking. The bar in front of lines 1 and 2 corresponds with a 200 μV OO-EMG signal. Vertical bar in front of lines 3 to 10corresponds with a 50° rotation; duration bar: 100 ms.

Figure 2.

 

Profiles of six superimposed successive traces of simultaneous recorded eyelid and eye movement during voluntary and reflex blinking measured from the same patient as in Figure 1 . (A) Eyelid movement during voluntary blinking recorded at the onset of the affliction. (B, C) Eyelid and eye movement during voluntary blinking recorded at 30 and 72 weeks, respectively. (D) Eyelid movement after a corneal air puff on the palsied side recorded at the onset of the affliction. (E, F) Eyelid and eye movement after a corneal air puff on the palsied side recorded at 30 at 72 weeks, respectively. Eyelid movement motility remained impaired in both types of blinking throughout the study. Both eyes move in the same abnormal direction during voluntary blinking, whereas in reflex blinking the direction of eye movement is normal after 72 weeks; however, the amplitude on the palsied side remains smaller.

Figure 3.

 

Schematic representations of OO-EMG and eyelid kinematics during recovery. Shown are eyelid movement start time (A), duration of the up phase (B), maximum amplitude (C), maximum velocity (D), and time maximum velocity (E), along with the amplitude (F), the summed amplitudes (G), and the start time (H) of the OO-EMG over the recovery time. Light blue: voluntary blinking, red: corneal air-puff–induced blinking on the nonpalsied side; green: corneal air-puff–induced blinking on the palsied side; purple: acoustic-click–induced blinking. Thick lines: values on the palsied side; thin lines: values on the nonpalsied side. In (A), (E), and (H), values from voluntary blinking (light blue) are absent, as an exact start time of a “trigger” cannot be determined for this type of blinking.

Table 2.

 

Differences in OO-EMG and Eyelid Kinematics between Nonpalsied and Palsied Sides

Type of Blinking	OO-EMG and Eyelid Kinematics	Onset					18 Weeks					36 Weeks					84 Weeks
				Mean	SEM	P<	Mean	SEM	P<	Mean	SEM	P<	Mean	SEM	P<
Voluntary	Start OO-EMG (ms)	Indefinite			11	5	0.03	−1	4	NS	2	7	NS
	Log (integrated OO-EMG ratio)	0.897	0.130	0.0005	1.053	0.162	0.001	0.101	0.071	NS	−0.080	0.061	NS
	Start time (ms)	11	9	NS	6	3	0.04	3	2	NS	3	5	NS
	Start lid - start OO-EMG (ms)	Indefinite			−5	6	NS	4	5	NS	1	5	NS
	Down phase (ms)	51	9	0.002	66	13	0.005	37	10	0.005	28	7	0.02
	Log (up phase ratio)	0.228	0.090	0.03	0.137	0.014	0.00001	0.053	0.028	0.05	0.155	0.042	0.01
	Log (maximum amplitude ratio)	0.950	0.118	0.00005	1.021	0.112	0.00002	0.202	0.044	0.001	0.153	0.089	NS
	Log (maximum velocity ratio)	1.288	0.095	0.000002	1.378	0.070	0.0000002	0.381	0.059	0.0001	0.277	0.085	0.01
	Time maximum amplitude (ms)	58	23	0.02	27	5	0.0005	17	4	0.002	11	2	0.01
	Time maximum velocity (ms)	46	21	0.04	27	9	0.01	25	4	0.0002	14	7	NS
Air puff-induced, nonpalsied eyelid	Start OO-EMG (ms)	Indefinite			11	5	0.03	3	2	NS	1	3	NS
	Log (integrated OO-EMG ratio)	0.397	0.075	0.0005	0.436	0.112	0.005	0.145	0.052	0.02	0.082	0.081	NS
	Start time (ms)	19	5	0.005	11	3	0.01	4	1	0.005	−1	5	NS
	Start lid - start OO-EMG (ms)	Indefinite			0	6	NS	1	2	NS	−2	3	NS
	Down phase (ms)	7	6	NS	28	9	0.03	15	5	0.02	18	5	0.02
	Log (up phase ratio)	0.104	0.041	0.02	0.110	0.035	0.01	0.057	0.017	0.005	0.086	0.039	0.04
	Log (maximum amplitude ratio)	0.502	0.102	0.001	0.621	0.159	0.005	0.270	0.073	0.005	0.247	0.116	0.04
	Log (maximum velocity ratio)	0.769	0.123	0.0002	0.846	0.188	0.001	0.412	0.103	0.002	0.379	0.157	0.03
	Time maximum amplitude (ms)	26	9	0.02	39	17	0.02	19	6	0.01	17	6	0.04
	Time maximum velocity (ms)	28	9	0.01	16	5	0.01	21	5	0.001	11	5	NS
Air puff-induced at palsied eyelid	Start OO-EMG (ms)	Indefinite			7	4	NS	−3	3	NS	−4	2	NS
	Log (integrated OO-EMG ratio)	0.101	0.071	NS	0.145	0.052	0.02	0.031	0.103	NS	0.007	0.089	NS
	Start time (ms)	25	5	0.005	8	2	0.005	1	3	NS	−2	3	NS
	Start lid - start OO-EMG (ms)	Indefinite			0	5	NS	4	1	0.005	2	3	NS
	Down phase (ms)	23	16	NS	35	11	0.01	12	4	0.01	25	13	NS
	Log (up phase ratio)	0.053	0.028	0.05	0.057	0.017	0.005	0.036	0.017	0.04	0.094	0.016	0.001
	Log (maximum amplitude ratio)	0.202	0.044	0.001	0.270	0.073	0.005	0.201	0.063	0.01	0.091	0.070	NS
	Log (maximum velocity ratio)	0.381	0.059	0.0001	0.412	0.103	0.002	0.350	0.085	0.002	0.243	0.096	0.03
	Time maximum amplitude (ms)	54	15	0.02	43	12	0.005	13	2	0.0002	22	14	NS
	Time maximum velocity (ms)	30	13	0.03	25	9	0.02	11	3	0.002	10	5	NS
Acoustic click-induced	Start OO-EMG (ms)	Indefinite			6	2	0.005	−1	2	NS	0	1	NS
	Log (integrated OO-EMG ratio)	−0.081	0.061	NS	0.082	0.081	NS	0.021	0.031	NS	0.020	0.015	NS
	Start time (ms)	6	6	NS	5	2	0.04	2	2	NS	2	2	NS
	Start lid - start OO-EMG (ms)	Indefinite			−1	3	NS	3	3	NS	2	3	NS
	Down phase (ms)	2	7	NS	4	3	NS	4	3	NS	0	3	NS
	Log (up phase ratio)	0.155	0.042	0.02	0.086	0.039	NS	−0.007	0.021	NS	0.064	0.022	0.02
	Log (maximum amplitude ratio)	0.153	0.089	NS	0.247	0.116	NS	0.291	0.077	0.005	0.135	0.165	NS
	Log (maximum velocity ratio)	0.277	0.085	0.03	0.379	0.157	0.05	0.351	0.077	0.001	0.160	0.196	NS
	Time maximum amplitude (ms)	8	7	NS	9	6	NS	6	2	0.03	2	4	NS
	Time maximum velocity (ms)	32	26	NS	5	5	NS	4	2	0.04	1	1	NS

 The differences of the integrated OO-EMG, the up phase, the maximum amplitude, and velocity are expressed in logarithmic ratios to normalize their distributions. The remaining parameters are expressed as time differences (in milliseconds). Apart from the means and the SEMs probabilities are indicated to determine the significance of each parameter. P > 0.05 was considered nonsignificant (NS).

Table 3.

 

Eye Movement Parameters after 30 and 72 Weeks of Recovery

Type of Blinking	Eye Movement Kinematics	30 Weeks			72 Weeks
				Nonpalsied	Palsied	Nonpalsied	Palsied
Voluntary
	Maximal amplitude (°)	9.1 ± 1.7	3.4 ± 3.4	10.9 ± 3.1	5.6 ± 3.1
	Time difference maximum amplitudes (ms)	10 ± 19	—	1 ± 10	—
	Amplitude ratio	2.68 ± 0.18	—	1.97 ± 0.09	—
Air Puff-induced at nonpalsied eyelid
	Maximal amplitude (°)	7.5 ± 1.2	2.3 ± 0.3	8.7 ± 1.3	2.5 ± 0.4
	Time maximal amplitude (ms)	99 ± 6	134 ± 6	100 ± 2	103 ± 6
	Time difference maximum amplitudes (ms)	35 ± 7	—	3 ± 4	—
	Amplitude ratio	3.26 ± 0.10	—	3.52 ± 0.07	—
Air puff-induced at palsied eyelid
	Maximal amplitude (°)	5.5 ± 0.6	2.6 ± 0.2	6.9 ± 1.6	2.9 ± 0.6
	Time maximal amplitude (ms)	98 ± 4	137 ± 7	97 ± 2	98 ± 2
	Time difference maximum amplitudes (ms)	39 ± 5	—	1 ± 3	—
	Amplitude ratio	2.13 ± 0.02	—	2.23 ± 0.12	—
Acoustic Click-induced
	Maximal amplitude (°)	2.4 ± 0.2	1.1 ± 0.2	3.0 ± 1.2	1.5 ± 0.2
	Time maximal amplitude (ms)	96 ± 2	111 ± 5	96 ± 1	110 ± 4
	Time difference maximum amplitudes (ms)	15 ± 5	—	14 ± 4	—
	Amplitude ratio	2.18 ± 0.06	—	1.97 ± 0.11	—

 Time differences of voluntary blinking at 30 and 72 weeks and of air puff induced blinking at 72 weeks are not significant. Neither is the time of maximal amplitudes of air puff-induced blinking after 72 weeks.