Tolerability of facial electrostimulation in healthy adults and patients with facial synkinesis

This prospective interventional study was performed in 2017 at the Department of Otorhinolaryngology, authors blinded. Approval for the study was obtained through the local institutional review board and informed consent was obtained from all study participants. The investigation had two parts: (1) In the first study, two different waveforms of the pulses were compared only in healthy controls. The effects of surface facial electrostimulation using pulses with rectangular waveform were compared to pulses with triangular waveform to decide if a rectangular or triangular waveform should be used for the second study. (2) Based on the results, only pulses with rectangular waveform were used for the subsequent second study. Here, the effects of electrostimulation in healthy participants were compared to patients with aberrant reinnervation after facial palsy and subsequent postparectic facial synkinesis.

The first study investigated 48 healthy controls (20 females, median age: 45 years). Inclusion criteria were: no history of facial palsy or any other neurological disease, no other chronic disease. The stimulation was randomly performed on the left side or the right side in half of the controls, respectively. In the second study, another group of 30 healthy controls (8 females, median age: 49 years) with the same inclusion criteria were recruited. The additional group of patients consisted of 30 participants (22 females, median age: 51 years) with postparetic facial synkinesis as the result of unilateral peripheral facial palsy of different origin (idiopathic, herpes zoster, posttraumatic, postsurgical). The time interval between onset of the palsy and inclusion into the study was at least 12 months. The patients had postparetic synkinesis for at least 4 months. Postparetic synkinesis was confirmed by clinical and electromyographical examination. The patients were stimulated on the affected side.

Electrostimulation was performed using a standard electrostimulation device (Galva 4a, Zimmer Medizin Systeme GmbH, Neu-Ulm, Germany). For stimulation, round stimulation electrodes (diameter: 32 mm; Krauth + Timmermann GmbH, Hamburg, Germany) were used. A ground electrode (56 × 56 mm, Zimmer Medizin Systeme GmbH, Neu-Ulm, Germany) was fixed on the skin in the neck. All participants were examined in a sitting position. The stimulation took place at three defined positions (Supplemental Fig. 1): (P1) Midline between lateral corner of the eye and base of the helix for contraction of the orbicularis oculi muscle; (P2) below the zygomatic arch for zygomatic muscle contraction; and (P3) on the level of the angle of the mouth anterior to the masseter muscle for orbicularis oris muscle contraction. The sequence of the positions for electrostimulation was counterbalanced across participants. Single monophasic pulses were used. A sequence of fixed pulsation durations (0.1, 0.5, 1, 10, 50, 100, 500 ms) was tested. As a result of the first study, we refined the gradation of the pulsation durations (0.1, 0.5, 1, 2, 5, 10, 50, 100, 500, 1000 ms) for the second study. The stimulation at each stimulation point started with 0.1 mA and was stepwise increased in 0.1 mA steps. The following parameters were evaluated: (1) motor twitch threshold: the point at which a facial muscle contraction was observed, (2) tolerability threshold: the point at which the participant wanted to stop the stimulation because of discomfort or when the maximal amplitude (40 mA) was reached. The tolerability threshold was reached in some cases instantly at the first stimulation at the lowest threshold, i.e. the tolerability threshold was lower than the motor threshold. Additionally, the participant rated the felt discomfort after electrostimulation on a numeric rating scale (NRS) from 0 (no discomfort) to 10 (maximal discomfort). This was used to record (3) discomfort at motor threshold, and (4) discomfort at tolerability threshold. Finally, all side effects were noted. The participants were actively asked for the following side effects after each electrostimulation: light flashes, toothaches, metallic taste. If occurred, the participants were asked to qualify the side effect as mild, tolerable, intolerable. Furthermore, the participants were observed for unintended masseter muscle or shoulder muscle contractions. If occurred, the participants were also asked to qualify the side effect as mild, tolerable, intolerable.

All outcome variables were analyzed with IBM SPSS statistics software (Version 25; IBM, New York) for medical statistics. Data are presented as frequencies or mean ± standard deviation (SD) if not otherwise indicated. Amplitude and discomfort threshold as primary outcome parameters were analyzed using repeated measures analysis of variance (rmANOVA). Bonferroni corrected pairwise t-tests were used for post hoc comparisons. Gender and side (stimulation left or right side) were between-subject factors. Pulse duration up to 50 ms (not 100 ms and 500 ms, because the tolerability threshold was not reached at these pulse durations), stimulation site, and waveform of the stimulus were within-subject factors in each rmANOVA in the first study. In the second study, gender and health status (healthy controls, patients) were between-subject factors. Pulse duration up to 50 ms (not 100 ms, 500 ms and 1000 ms, because the tolerability threshold was not reached at these pulse durations), and stimulationsite were within-subject factors in each rmANOVA. The significance level was set at p < 0.05.

Average amplitude levels at motor threshold and discomfort at motor threshold, as well as discomfort at tolerability threshold and discomfort tolerability threshold averaged over the three stimulus positions P1, P2 and P3 are shown in Fig. 1. The complete dataset is summarized in Supplemental Table 1. Up to a pulse duration of 10 ms, the tolerability threshold was always higher than the motor threshold for rectangular and also for triangular pulses. At 50 ms, non-tolerable discomfort sensation was reported by one subject before the motor threshold was reached. At 100 ms and 500 ms, this was the case for 73–98% and 22–98% of the participants.

The visual impression of the results was confirmed statistically: a rmANOVA was performed for each of the four primary outcome factors and revealed the following: The amplitude at motor threshold was significantly lower for P1 than for P2 and P3 (main effect stimulation site F[2, 88] = 27.149, p < 0.001, η² = 0.382), decreased with longer pulse duration (main effect pulse duration F[4, 176] = 1054.567, p < 0.001, η² = 0.960), and was higher for triangular than for rectangular pulses (main effect waveform F[1, 44] = 434.00, p < 0.001, η² = 0.908). Gender and side had no influence on the motor threshold. The average threshold for rectangular pulses was 4.4 mA (95% confidence interval [CI] = 4.1–4.6) and for triangular pulses 6.4 mA (95% CI 6.0–6.9; p < 0.0001). The same effects were seen in the ANOVA for the amplitude at tolerability threshold. The motor threshold was significantly lower for P1 than for P2 and P3 (main effect stimulation site F[2, 88] = 12.696, p < 0.001, η² = 0.224), decreased with longer pulse duration (main effect pulse duration F[4, 176] = 832.581, p < 0.001, η² = 0.950), and was higher for triangular than for rectangular pulses (main effect waveform F[1, 44] = 357.578, p < 0.001, η² = 0.890). Gender, and side had no influence on threshold (all p > 0.05). The average threshold for maximal tolerance for rectangular pulses was 10.7 mA (95% CI 9.6–11.7) and for triangular pulses 14.7 mA (95% CI 13.5–15.9; p < 0.0001).

The rmANOVA for the discomfort at motor threshold revealed that discomfort perception increased with longer pulse duration (main effect pulse duration F[4, 176] = 109.074, p < 0.001, η² = 0.586), and discomfort was higher for triangular than for rectangular pulses (main effect waveform F[1, 44] = 17.650, p < 0.001, η² = 0.286). Average discomfort at motor threshold (NRS) for rectangular pulses was 1.6 (95% CI 1.3–1.9) and for triangular pulses 1.9 (95% CI 1.6–2.3; p < 0.0001). Stimulation site, gender, and side had no influence. The rmANOVA for discomfort at the tolerability threshold revealed that none of the factors or combinations of the factors influenced discomfort at the tolerability threshold (all p > 0.05). Longer pulse duration was associated to more discomfort per mA stimulation amplitude (Supplemental Fig. 2). The effect seemed to be more pronounced for rectangular pulses.

The side effects of electrostimulation are shown in Supplemental Table 2. Overall, side effects were more often seen for triangular than for rectangular pulses. The visual impression of light flashes during electrostimulation was by far the most frequent, but tolerable side effect. Tolerable weak light flashes were felt during 369 out of 1008 rectangular stimulations (36.6%) and during 393 of 1008 triangular stimulations (38.9%). Other relevant side effects in descending order of frequency were masseter muscle contraction, toothaches, a metallic taste, and shoulder muscle contraction. Light flashed occurred more often during electrostimulation at position P1 (eye region). Toothaches occurred most frequently at position P2 (cheek region). Masseter muscle contraction, shoulder muscle contraction, or metallic taste were seen most frequent at position P3 (oral region).

Average amplitude levels at motor threshold and pain at motor threshold, as well as at tolerability threshold and discomfort tolerability threshold averaged over the three stimulus positions P1, P2 and P3 re shown in Fig. 2. The related data are summarized in Supplemental Table 3. Up to pulse duration of 50 ms, the tolerability threshold was always higher than the amplitude at motor threshold for healthy participants and patients. Beginning at 100 ms and at higher pulse duration, a non-tolerable discomfort perception directly occurred in 73–98% of the participants.

A rmANOVA was also performed for each of the four primary outcome factors in the second study. The motor threshold was significantly lower for P1 than for P2 and P3 (main effectstimulationsiteF[2, 102] = 4.785, p < 0.010, η² = 0.086) and decreased with longer pulse duration (main effect pulse duration F[6, 306] = 231.815, p < 0.001, η² = 0.820). Gender and side had no influence. There was a significant lower motor threshold for healthy controls than for patients. The average threshold for controls was 3.4 mA (95% CI 3.0–3.7) and for patients 5.2 mA (95% CI 4.3–6.0; p < 0.0001). The same effects were seen in the rmANOVA for the tolerability threshold. The motor threshold was significantly lower for P1 than for P2 and P3 (main effect stimulation site F[2, 96] = 4.330, p < 0.016, η² = 0.083) and decreased with longer pulse duration (main effect pulse duration F[6, 288] = 296.281, p < 0.001, η² = 0.861). Gender and side had no influence. There was also a significant lower motor threshold for healthy controls than for patients (main effect health status; p = 0.025). The average tolerability threshold for healthy controls was 11.3 mA (95% CI 9.9–12.8) and for patients 13.8 mA (95% CI 12.8–15.7; p = 0.044).

The rmANOVA for the discomfort at motor threshold in the second study revealed that discomfort perception increased with longer pulse duration (main effect pulse duration F[6, 306] = 4.569, p < 0.001, η² = 0.082). Average discomfort at motor threshold (NRS) for healthy controls was 0.8 (95% CI 0.5–1.1) and for patients 1.6 (95% CI 1.0–2.2; p = 0.015). Stimulation site, gender, and side had no influence. The rmANOVA for discomfort at the tolerability threshold revealed that none of the factors or combinations of the factors had influence on discomfort at the tolerability threshold (all p > 0.05). Discomfort at average tolerability threshold (NRS) for healthy controls was 3.1 (95% CI 2.2–4.0) and for patients 4.1 (95% CI 3.1–5.0; p = 0.144), i.e. was not significantly different. As in the first study, longer pulse duration was associated to more discomfort per mA stimulation amplitude (Supplement Fig. 2). The effect was not different between healthy controls and patients, i.e. there was no interaction between the two factors.

The side effects of electrostimulation of the second study are shown in Supplemental Table 4. Again, tolerable light flashes were by far the most frequent side effect felt in 30% of the stimulation in controls as well as in patients. Overall, the frequency of felt side effects was not different for healthy controls and patients except for the sensation of metallic taste. Sensation of metallic taste was more frequently reported by the patients. Like in the first study, light flashes also occurred more often during electrostimulation at position P1 (eye region) in the second study. Toothaches occurred most frequently at position P2 (cheek region). Masseter muscle contraction and shoulder muscle contraction were seen most frequent at position P3 (oral region).