Effect of neural mobilisation in Bell’s palsy: A randomised controlled trial

Introduction

Bell’s palsy is a dysfunction of the facial nerve causing unilateral facial paralysis.¹ The annual incidence of Bell’s palsy is 15–30 cases per 100,000 people.² Bell’s palsy has shown to recover rapidly, with 85% of cases recuperate within 3 weeks.³ The residual effects of Bell’s palsy are seen in 29% of cases that range from minor facial asymmetry to facial contractures or disfigurement.⁴ A systematic review and meta-analysis reported that the combination of prednisone and antiviral drugs for 10 days provided a better prognosis but is not always fully successful.⁵

The facial nerve is a 7th cranial nerve emerging from the facial nerve nucleus of the brain stem.⁶ It begins as two roots (motor root and sensory root) and is accompanied by the 8th cranial nerve (vestibule-cochlear nerve) to enter the internal auditory meatus. The two roots of facial nerves leave the internal auditory meatus and enter the facial canal.⁷ Within the facial canal, the two roots fuse to form facial nerve and the geniculate ganglion (collection of nerve cell bodies). The facial nerve exits the facial canal via the stylomastoid foramen. Finally, the facial nerve divides into five motor branches (temporal, zygomatic, buccal, marginal mandibular, and cervical) to innervate muscles of facial expression.⁸

Neural tension was the term used earlier to test neural dysfunction in the peripheral nerves. In the past decade, the paradigm of neural mobilisation has shifted from a mere testing tool to a more dynamic system of evaluation and treatment of neural tissues.⁹ Neurodynamics is a new preferred term used by clinicians and researchers to describe the dynamic and adaptive nature of neural structures. The nervous system adapts to the movement performed in the limbs. The nerves adapt to tension, compression, and torsional stress during various functional activities. If these neural adaptive mechanisms fail, the nerve becomes vulnerable to inflammation, vascular hypoxia, or ischemia, which may alter neural function.¹⁰ In Bell’s palsy, facial nerve dysfunction is due to an inflammation of a facial nerve in a canal before emerging from the stylomastoid foramen. While the exact aetiology of Bell’s palsy is still largely unknown, it is thought that inflammation of the facial nerve could play a role.¹¹ The facial nerve can suffer from intrinsic pressure due to inflammation and the constricted S-shaped canal before it emerges through the stylomastoid foramen.¹²

The use of complementary therapies combined with drug therapy is a promising addition to the treatment of patients with Bell’s palsy. These complementary therapies are massage,¹³ electrical stimulation,¹⁴ Kinesiotaping,¹⁵ pulsed shortwave diathermy,¹⁶ and neural mobilisation. However, scientific evidence supporting these complementary therapies is limited. Neural mobilisation of the facial nerve is reported as a safe complementary therapy but lacks robust scientific evidence.¹⁷ The rationale of facial nerve mobilisation was derived from a study conducted by Kashoo et al.¹⁷ The proximity of the facial nerve to the outer ear, according to the authors of the case study, can be used to mobilise the facial nerve.

It is possible but not proven that a facial nerve mobilisation technique that moves the nerve as it emerges from the skull might aid reduction of swelling and tissue damage. Therefore, the purpose of our study was to evaluate the effect of addition of facial nerve mobilisation technique to the conservative treatment of acute Bell’s palsy. We hypothesise that the addition of facial nerve mobilisation to conservative treatment would lead to greater improvements in facial movement and symmetry than conservative treatment alone in patients with acute Bell’s palsy.

Materials and Methods

Study design: Single-centre, parallel, placebo-controlled randomised controlled trial with 1:1 randomisation. There was no change in the study method or procedure. The trial was conducted as planned in the protocol.

Settings: Our study started recruiting participants from 1 March 2020 and ended on 20 May 2021. This study was conducted in accordance with the ethical guidelines of the Declaration of Helsinki. The study was ethically approved by Central Institutional Review Board, Ministry of Health, Saudi Arabia (H-01-R-009). The study was conducted in the Physical Therapy Department of King Khalid Hospital, Majmaah, Saudi Arabia. The study protocol was registered in the US (United States) clinical trial registry (NCT04280120).

Participants’ recruitment: After the diagnosis by the neurologist, all patients were prescribed with corticosteroids and antiviral drugs for 10 days. The patients were referred to the out-patient physical therapy department within 1–2 days of Bell’s palsy onset and aged between 18–65 years. Patients were invited to take part and signed the written consent form. Patients who agreed and signed the consent form were assessed for the inclusion criteria that included no history of earlier Bell’s palsy or diabetes. Patients with moderate to severe acute Bell’s palsy evaluated by Modified House-Brackman Scale were included in the study. Patients with tumours, active skin or autoimmune disorders, or any neurodegenerative diseases were excluded from the study.

Sampling method: A Microsoft excel program was used to generate a list of 62 random numbers before the commencement of the study by a researcher who was not involved in the assessment or treatment of the prospective participants. The random number representing an intervention was confidentially communicated (sealed envelope) to the researcher involved in assessment and treatment of participants (third party concealed randomisation method).¹⁸ The unique number assigned to the participant was removed from the computer (Fig. 1).

Blinding: Randomisation of the participants was performed by a researcher who was neither part of the treatment nor the assessment. A researcher blinded to randomisation was involved in the baseline and post-treatment assessments. The participants were blinded to the type of treatment by providing placebo neural mobilisation.

Outcome variables: An independent researcher (blinded to participants allocation) who were certified manual therapists with master’s degree and had more than 10 years of clinical experience were tasked to assess the severity of the Bells palsy using the Sunnybrook Scale¹⁹ and Kinovea Movement Analysis Software²⁰ at baseline and after 3 weeks of intervention.

Primary outcome: Sunnybrook Scale is a reliable and valid tool to assess the severity of facial muscle paralysis based on the three subscales (resting symmetry, the symmetry of voluntary movements, and synkinesis).²¹ The scores obtained from the scale yields a numerical value ranging from 0 to 100, higher scores imply better facial recovery. At rest, each patient was evaluated based on the symmetry of the eye (0–1), cheek (0–2), and mouth (0–1). The second phase of the assessment involved the rating of five different facial expressions: eyebrow lift, eye closure, and smile with open mouth, snarl, and lip pucker, with a rating of 1–5 depending on the symmetry of movement on the affected side in comparison to the unaffected side. The scores obtained were added together and multiplied by 4. The last phase of the testing involved an assessment of the severity of synkinesis (associated movements) while performing five expressions, which were on a 3-point scale (0 for normal and 3 for severe). The scores obtained from the symmetry of the voluntary movements were subtracted from the added scores obtained from the resting symmetry and the synkinesis.

Secondary outcome: The same therapist evaluated the patients on Kinovea Movement Analysis Software (KMAS) through a videotape and analysed via free software. The software is an open and free tool for movement analysis. A plain video was recorded using a DSLR high-resolution camera (Nikon, USA model D-40) placed on a tripod. The patient faced the camera while sitting on the chair with the feet firmly planted on the ground. During recording, patients were asked to perform three movements as follows: first, raise the eyebrows (frontalis); second, smile with pursed lips (zygomatic); and third, pout the lips (orbicularis oris). The symmetry ratio was calculated while comparing each side.

Intervention: All the patients in the control and experimental group received 15 treatment sessions (5 times per week) by three independent therapists who were not involved at the baseline and post 3-week assessment. The therapists were trained by experts in facial nerve mobilisation techniques. Although the facial nerve mobilisation technique is simple to replicate, the therapists were supervised by an expert for at least five initial treatment sessions. Experimental group received facial massage, facial exercises electrical stimulation, and facial nerve mobilisation technique while as, control group received facial massage, facial exercises electrical stimulation, and placebo facial nerve mobilisation technique. The placebo facial nerve mobilisation followed all the procedures of real facial nerve mobilisation without applying horizontal and circular auricular traction.

The facial massage was applied while the patient was in the supine position. The massage was applied for 10 min on both sides of the face and neck. The massage included stroking using the tips of fingers and circular massage with three middle fingers for 2 min. Effleurage, kneading, picking up, and wringing were used to increase circulation and induce relaxation for 6 min. The massage ended with a hacking movement for 2 min to evenly distribute erythema.²²

Ten facial expression exercises including raising of eyebrows, flaring of nostrils, closing the eyes tightly, smiling with lips closed, sucking or sipping, joining lips or puffing, crying expression, laughing expression, and pouting were performed once at the clinic and 3–4 times at home per day in front of a mirror. Each exercise was performed 10 times, recommended by previous studies to avoid fatigue.²³ A standard exercise handout for home programs in Arabic was given to each subject upon initial consultation with the physiotherapist.

A standard calibrated machine called Electrotherapy Ultrasound Equipment; Combo-IFT-US-TENS (Bharat Medical Systems, Chennai, Tamil Nadu) was used for electrical stimulation of the facial muscles. The following setting parameters were chosen to mimic the natural action of the facial muscles: frequency 10$$Hz, pulse duration of 10$$$\mu$s, and intensity ranging from 1 to 1.5$$mA.¹⁴

Independent therapists who were not part of pre-assessment or post-assessment performed facial nerve mobilisation. Facial nerve mobilisation technique was applied when the patient was in supine position on the therapeutic table with head flat on the table in neural rotation. The therapist used sterile gloves for protection and talcum powder to reduce friction. The therapist sat down at the head-end of the patient. The table height was adjusted to the level slightly above the waist of the therapist while sitting on the chair. The therapist used one hand to perform facial nerve mobilisation, while the other hand was used to stabilise the patient’s head by placing the palm on the unaffected side of the face. The hand placement on the affected side is such that the therapist’s index finger circle behind the auricle of the ear and the thumb is gently placed at the opening of the external auditory meatus (Fig. 2). The intensity of the manipulation is decided by the patient reporting the level of discomfort. The force used in the traction is produced by gently moving the wrist in extension and circular manipulation from the elbow (mid-prone position) and shoulder joint. The circular movement and horizontal traction were applied 25 times in three sets with a 5-second rest period in between a single manipulation and a 30-second rest period in between the set.

Sample size calculation: The sample size was calculated by obtaining the mean and standard deviation scores from the primary assessment tool used in a similar previous study.²⁴ The calculation involved the insertion of data into the G*Power 3.1.9.2 software.²⁵ The mean difference and SD scores (Group 1-mean difference $=$ 26.3, SD $=$ 7.8, Group 2-mean difference $=$ 16.25, SD $=$ 3.99) of the Sunnybrook Facial Grading System from the previous similar²⁴ study were entered into G*Power 3.1.9.2 software to get the effect size ($d=1.16$). The statistical test used to calculate sample size was the t-test, and a priori power analysis was used to generate output parameters in G*power software. The results of the imputation yielded an ideal sample size of 54, with each group containing 27 participants. The power from the above sample size was 0.99. Expecting a drop-out rate of 20–30%, 62 was the anticipated sample size.

Statistical analysis

The analyses were performed using IBM SPSS Statistics for Windows, Version 20.0 (Armonk, NY: IBM Corp). The non-parametric data was analysed by the chi-square test and the parametric data by the Analysis of covariance ANCOVA with baseline scores as covariate and independent sample t-test. The Shapiro–Wilk test ($p>0.05$) and a visual inspection of their histograms, normal Q–Q plots, and box plots showed that the scores obtained from the Sunnybrook Scale and the KMAS were approximately normally distributed for both the experimental and control group. Levene’s test was conducted to satisfy the homogeneity of variance. Linearity of data was investigated through the scatter plot covariate and dependent variable. Sensitivity analysis was conducted for means scores of Sunnybrook scale whilst controlling for age, gender, body mass index, duration of Bell’s palsy. The secondary assessment was performed by KMAS for three major facial muscles (zygomatic, orbicularis oris, and frontalis muscle). The independent sample t-test was used for the data analysis of post-intervention scores of KMAS between control and experimental group because Leven’s test for homogeneity of variance was violated for ANCOVA.

Results

Of 189 people screened for eligibility, 62 persons were included (Fig. 1). No participants were lost in follow-up. Primary assessment was performed by Sunnybrook Scale. Among the participants belonging to the control group, the mean Sunnybrook Scale score at baseline was mean $=$ 5.74, SD $=$ 4.6 and the experimental group was mean $=$ 5.77, SD $=$ 4.3. At baseline, there was a non-statistically significant difference between the experimental and control group in terms of mean scores of Sunnybrook Scale (Table 1). The scores after 3 weeks of treatment changed to mean $=$ 72.2, SD $=$ 7.9 for control and mean $=$ 81.5, SD $=$ 8.0 for the experimental group.

For primary outcome, analysis of covariance with baseline data as covariate showed a significant difference between groups for the post-test mean scores of SBS after 3 weeks (between-group difference, 9.2 [95% CI, 5.1–13.3], $p=0.001$). Importantly, the effect size calculated by partial ${\eta }^2$ for neural mobilisation was 0.258.

For the secondary outcome, the independent sample t-test showed a significant difference between groups scores on KMAS for the zygomatic muscle after 3 weeks (between-group difference, $-27.2$ [95% CI, $-31$ to $-22.6$], $p=0.001$). Similarly, for the frontalis muscle, $-16.7$ [95% CI, $-9.9$ to $-23.4$], $p=0.001$ and the orbicularis oris muscle, $-15.0$ [95% CI, $-11.1$ to $-18.8$], $p=0.001$ (Table 2).

Sensitivity analysis for robustness of significant results was conducted through ANCOVA with gender, age, Body Mass Index (BMI), and duration of Bell’s palsy as covariates. The results revealed a non-significant difference in the post-test scores of SBS while controlling for gender, age, BMI, and duration of Bell’s palsy. Details about the sensitivity statistics are provided in Supplementary material.

Discussion

The results of this randomised controlled trial showed that 3-weeks of facial neural mobilisation, in addition to usual pharmacological and physical therapy treatment, improved facial movement and symmetry following acute idiopathic facial muscle paralysis (Bell’s palsy) to a greater extent than usual pharmacological and physical therapy treatment.

There was a post-test mean difference of approximately 9 units with small effect size for SBS scores between control and experimental group. Similarly, research conducted among 51 patients with acute Bell’s palsy reported 11 units of mean rank difference between groups receiving high level Laser therapy than low level Laser therapy.²⁶ A systematic review reported significant standardised mean difference of 13 units between experimental group who performing facial exercise therapy than the control group who did not perform facial exercises.²⁷ Similarly, a research study conducted among 60 patients with acute Bell’s palsy reported significant improvement in the facial profile with facial acupressure for 20$$min for two weeks.²⁸ The study also reported a decrease in depression symptoms.²⁸ Data from our study seems to confirm that neural mobilisation can be safely practised by experienced physical therapists in clinics since no deterioration in patient facial function occurred. Moreover, the neural mobilisation technique relatively easy technique to learn than peripheral nerve mobilisation because it involves simple traction and circular mobilisation facial nerve opening of the external auditory meatus (outer ear).

The results of our study are supported by many studies reporting beneficial effects of neural mobilisation of peripheral nerves. A study reported a reduction in the upper limb radiculopathic pain after 3 weeks of neural mobilisation.²⁹ Another study revealed a hypoalgesic effect caused by neural tension and gliding in normal healthy adults.³⁰ A reduction in pain and improvement in function after sciatic nerve mobilisation in patients with radiating low back pain were also described in another study.³¹ Similarly, a study reported improvement in the H-reflex latency of the flexor carpi radialis muscle after neural mobilisation of the median nerve among subjects with unilateral cervical radiculopathy.³² A cadaveric study reported significant dispersion of intra-neural edoema after passive neural mobilisation of a tibial nerve due to compression syndrome.³³ A recent study reported the positive effect of neural gliding on sciatic neural function and nerve sensitivity in healthy asymptomatic adults.³⁴ However, there is no randomised controlled trial published about neural mobilisation of facial nerve following Bell’s palsy.

We found better facial symmetrical voluntary movement performed by the zygomatic, frontalis, and orbicularis oris after 3 weeks of intervention (Fig. 3). Our study showed a relatively better improvement in action of frontalis as compared to zygomatic and orbicularis oris. A study using meridian facial massage³⁵ for 20$$min in Bell’s palsy reported the similar improvement in the action of frontalis muscle. Other complementary therapies such as short-wave diathermy¹⁶ and neuromuscular re-education.³⁶ (Resistance training of facial muscles) also reported improvement in resting and movement symmetry in Bell’s palsy. The possible beneficial effects of these massage techniques and neural mobilisation could be due to the facilitation of nerve gliding in the canal, reduced nerve adherence, dispersion of noxious inflammatory agents, increased neural blood supply, and improved exoplasmic flow. The sedative effect of the facial nerve mobilisation technique on nociceptive nerve fibres is likely to reduce inflammation and edoema surrounding the nerve. However, these effects due to neural mobilisation require validation studies. Future studies can consider subjects with chronic Bell’s palsy and can include Bell’s palsy with associated disorders.

Limitation

There are limitations in this study. The study lacks long term follow-up. The neural mobilisation technique used in the treatment is not fully developed and validated and it’s the first randomised controlled trial to report the effect of facial mobilisation in Bell’s palsy. The study population constituted only healthy participants without any history of diabetes or hypertension. Therefore, the results of the study cannot be generalised to all Bell’s palsy populations. The study population included in our study was relatively small and the magnitude of effect may be overestimated. The actual mobilisation of facial nerve could not be determined. Therefore, future research is recommended to include ultrasonography to measure the mobilisation of target facial nerve.

Conclusions

Facial nerve mobilisation added with conventional intervention was more effective in improving the facial symmetry among patients with acute Bell’s palsy. Our findings suggest that addition of neural mobilisation to acute rehabilitation of Bell’s palsy may facilitate early recovery and minimise facial muscle paralysis. Future studies can be conducted to evaluate the effect of neural mobilisation in chronic Bell’s palsy with associated disorders such as diabetes.

Acknowledgements

The authors would like to thank all the participants who volunteered to participate in the study.

Conflict of Interest

The authors declare no potential conflicts of interest with respect to the research, authorship and/or publication of this paper.

Funding/Support

The authors receive no funding for this research project.

Author Contributions

Raed Alharbi, Faizan Zaffar Kashoo, Mehrunnisha Ahmed, Mazen Alqahtani, and Saleh Aloyuni contributed to the concept and the design of the research, the supervision of the research work, the analysis of data and interpretation of the results, writing the whole paper and revising it critically for important intellectual content, ensuring the scientific accuracy and integrity of the paper, the final approval of the paper before submission, and the second revision of the paper as per the reviewers’ instructions. Msaad Alzhrani, Ahmad Dhahawi Alanazi, Mohammad Sidiq, Bander Hamud Alharbi, and Gopal Nambi contributed to the concept and design of the research, application of physiotherapy sessions, evaluation of the outcome measures, data acquisition, as well as the revision and final approval of the paper. Ahmad Dhahawi Alanazi and Mohammad Sidiq contributed to the concept of the research, the medical supervision of the research work, as well as the revision and final approval of the paper.