Effect and neurophysiological mechanisms of transcutaneous auricular vagus nerve stimulation (taVNS) for misophonia (taVNS-MISO): a study protocol for a single-centre double-blind randomised sham-controlled trial

Abstract

Introduction

Misophonia is a newly recognised sound sensitivity disorder with clinically significant symptoms affecting up to 18% of the population. It is characterised by extreme negative reactions to specific sounds which are often repetitive and generated by the human oral-nasal tract (eg, sniffing and eating sounds). Although misophonia currently has no standard treatment, research suggests that transcutaneous auricular vagus nerve stimulation (taVNS) holds promise therapeutically. This study aims to investigate both the effects of 4 weeks taVNS (compared with sham) on misophonia and related symptoms as well as its underlying neurophysiological mechanisms. To our knowledge, this is the first trial on taVNS in misophonia.

Methods and analysis

This is a single-centre double-blind sham-controlled trial in which 60 participants with clinically significant misophonia are randomly allocated in a 1:1 ratio to receive taVNS or sham stimulation. The intervention will be self-administered over 4 weeks (two times per day for 30 mins each). The primary efficacy outcome is self-reported misophonia severity with secondary outcomes, including mental health and audiological symptoms. In addition, all participants will undergo preintervention and postintervention testing, including MRI and physiology to investigate neurophysiological mechanisms underlying taVNS effects.

Ethics and dissemination

The study has been approved by the Brighton and Sussex Medical School ethics board (ER/GLP28/4). Results will be submitted for publication in peer-reviewed journals. Data will be anonymised and made available for sharing after completion of the study.

Trial registration number

This trial is registered in ISRCTN; ISRCTN79500062.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a single-centre double-blind sham-controlled trial to examine the effect of transcutaneous auricular vagus nerve stimulation (taVNS) in clinically significant misophonia.

• We measure mechanisms of taVNS in misophonia, including neural, physiological and behavioural, which are tested preintervention and postintervention. Self-administration of taVNS requires high compliance, increases the risk of unblinding and makes it harder to verify correct anatomical sites for active and sham stimulation, all of which may influence outcomes.

• Our samples are restricted to right-handed 18–40-year olds with fluent English abilities to aid functional MRI interpretations, but this may limit generalisability.

Introduction

Misophonia is a sound sensitivity disorder characterised by reduced tolerance to certain sounds and associated stimuli referred to as ‘triggers’. Trigger stimuli are often repetitive (eg, tapping, clicking), generated by the human oral-nasal tract (eg, breathing, chewing, sniffing) and sensitive to context (eg, the perceived source or meaning of the sound).¹ Misophonic triggers evoke unusually strong physiological and negative emotional responses (eg, anger, irritation, disgust and anxiety), with subsequent difficulty in emotion regulation and impairment to daily life functioning (eg, in educational, occupational and social settings).^{2 3} Estimated population prevalence rates of clinically significant misophonia from representative samples range from 5% to 18%.^{4 5}

Although treatment approaches for misophonia have been explored, including psychotherapy,6,8 medication^{9 10} and most recently, transcranial stimulation,¹¹ no widely accepted treatment exists.¹² A recent systematic review¹³ highlighted a lack of strong empirical support for any existing treatment, calling for greater experimental rigour and mechanistically driven approaches. Crucially, there is a need for more specialised interventions rooted in misophonia’s underlying pathophysiology, which target underlying neural, physiological and/or behavioural processes known to be involved in the disorder.¹⁴

Multiple lines of evidence point to transcutaneous auricular vagal nerve stimulation (taVNS) as a promising treatment approach for misophonia given its potential to modulate processes underpinning how misophonia symptoms are generated and maintained. taVNS is a non-invasive method, which stimulates the auricular branches of the vagus nerve via electrical pulses applied to the ear (concha cavity and conchae).¹⁵ Previous studies have shown that taVNS improves symptoms in numerous disorders including migraine,¹⁶ tinnitus,¹⁷ epilepsy¹⁸ and depression¹⁹ and is well-tolerated with minimal side effects.²⁰ Although not yet tested in misophonia, vagal nerve stimulation has been shown to have modulatory effects on processes involved in the disorder. This includes effects of taVNS on autonomic function21,23 and emotion regulation24,27 both of which are impaired in misophonia28,30 as well as effects on cortical activity, connectivity and network plasticity in misophonic neural circuitry.1431,36 If proven successful, taVNS could be either an alternative or a complementary treatment alongside existing cognitive behavioural therapy (CBT) based approaches.

Given the potential application of taVNS in misophonia treatment, this study tests whether 4-week taVNS can improve misophonia symptoms in a double-blind randomised sham-controlled trial with four objectives:

1. Establish whether taVNS (applied for 30 min two times per day, at least 5 days/week, for 4 weeks) improves misophonia symptoms. Our primary outcome metric is the change in misophonia symptoms compared across active stimulation and sham groups preintervention and postintervention and at 1-month follow-up.

2. Explore whether taVNS is associated with improvements in secondary outcomes, including mental health comorbidities, especially those previously linked to misophonia (Obsessive Compulive Disorder, anxiety and depression).^{3 37 38} We will examine changes in secondary outcome measures compared across active stimulation and sham groups pre and postintervention and at 1-month follow-up.

3. Determine the safety, tolerability, feasibility and acceptability of taVNS in misophonia. To determine safety and tolerability, we will record side effects and any worsening of misophonia. To determine feasibility, we will collect data on retention/completion rates and protocol adherence. To determine acceptability, we will collect metrics of usability of the device and treatment protocol.

4. Understand neurophysiological mechanisms underlying symptom improvement. Here, we include preintervention and postintervention assessments of: (1) resting-state neural activity, (2) neural activity under symptom provocation and (3) physiological reactivity and performance during a task capturing emotion regulation to misophonic stimuli. For more information on our mechanistic hypotheses, please see our OSF project (https://osf.io/368v5/overview?view_only=ae14d7c9c44b48b6b6c0ab9aef3f3ef6).

This study was developed in response to the need for new therapeutic options for misophonia, particularly those that are well-controlled and designed to target underlying neurophysiological mechanisms.¹³ Although we did not directly involve patients or members of the public in the trial design, our approach reflects research and treatment priorities identified by individuals with lived experience of misophonia. Qualitative interviews highlight a strong desire for practical, accessible treatments and the need for more research into the neural mechanisms underpinning misophonia.³⁹ These interviews also highlight a desire to explore the relationship between misophonia and co-occurring conditions such as anxiety and other forms of decreased sound tolerance, something reflected in our choice of secondary outcomes.³⁹ Findings from a large online survey of 252 adults with misophonia further support the need for treatment development. Although respondents had a strong willingness to engage in treatment, many reported dissatisfaction with previously pursued treatment options.¹² Notably, while neuromodulation approaches (including vagus nerve stimulation) were used less frequently than more conventional treatments (eg, relaxation techniques, psychological therapies or audiological interventions), they received comparable acceptability ratings. On a scale from 1 (extremely inappropriate) to 5 (extremely appropriate), mean acceptability scores were 3.45 for neuromodulation overall and 3.56 for vagus nerve stimulation, compared with 2.86 for medications, 3.36 for psychological interventions, 3.55 for relaxation-based treatments and 4.06 for audiological approaches. These findings suggest that while neuromodulation is not yet widely used, it may be perceived by patients as a promising and acceptable therapeutic option. Going forwards, to ensure alignment with patient perspectives, we will involve individuals with lived experience in the dissemination of study findings. This will include development of accessible summaries of results to be shared with relevant misophonia associations.

Methods and analysis

Study design

This study is a single-site double-blind randomised sham-controlled trial in participants with moderate to severe misophonia (preregistered in ISRCTN, trial number ISRCTN79500062) conducted at the University of Sussex within the Psychology School and Brighton and Sussex Medical School (see table 1).

Table 1. Trial characteristics based on WHO trial registration dataset.

Data category	Trial information
Primary registry and trial identifying number	ISRCTN79500062
Date of registration in primary registry	9/12/2025
Source(s) of monetary or material support	Misophonia Research Fund (MRF)—the funder has no role in design, data collection, analysis or publication
Primary sponsor	University of Sussex—researchsponsorship@sussex.ac.uk
Contact for public queries	GLP
Contact for scientific queries	GLP
Public title	A trial investigating the effectiveness of transcutaneous auricular vagus nerve stimulation (taVNS) for misophonia
Scientific title	Effect and neurophysiological mechanisms of taVNS in misophonia: a single centre double-blinded randomised sham-controlled trial
Country of recruitment	England
Health condition(s) or problems studied	Misophonia
Interventions	4-week taVNS vs sham administered at least twice daily for 5 days/week
Key inclusion and exclusion criteria	Inclusion: moderate to severe misophonia symptoms; not currently receiving active treatment for misophonia or planning to during the study; native English speaker; aged between 18 and 40; right-handed; no history of brain injury or neurological conditions; corrected or corrected to normal vision; normal hearing (except for hyperacusis, tinnitus or misophonia); owning and being willing to use a personal smartphone for stimulation reminders and questionnaire to be sent via a study app. Exclusion: pregnant or trying to get pregnant; active implants; cerebral shunt; wounds or skin diseases near the ear; history or family history of heart problems; history of seizures/epilepsy; any other contraindications to an MRI scan
Study type	Single-centre double-blinded sham-controlled trial in which 60 participants with clinically significant misophonia are randomly allocated in a 1:1 ratio to receive taVNS or sham stimulation
Target sample size	60
Recruitment status	Recruiting
Primary outcome(s)	Change in misophonia severity
Key secondary outcome(s)	Mental health and audiological measures; assessment of effects of taVNS on neurophysiology and behaviour (via MRI scans, laboratory task and concurrent measurement of heart rate and skin conductance)

An overview of the study design and participant timeline is shown in figure 1. Eligible participants are initially screened via online questionnaire for suitability (see inclusion/exclusion criteria). Those invited to take part are randomly assigned to 4-week at-home administration of taVNS or sham stimulation in a 1:1 ratio following baseline measurements at their first site visit (T1). Baseline (T1) and postintervention (T2) measurements include (1) misophonia symptoms, (2) mental health comorbidities, (3) audiological tests, (4) MRI scans (at rest and during symptom provocation) and (5) physiological reactivity at rest and during a task capturing emotion regulation to misophonic stimuli. Primary and secondary outcome measures are also assessed at 4-week follow-up via questionnaire (T3). Questionnaire measures of potential side effects and worsening of misophonia symptoms are collected and monitored weekly during the treatment period, with metrics of acceptability of the device and treatment protocol collected at T2.

Figure 1. Overview of study design and participant timeline. fMRI, functional MRI; taVNS, transcutaneous auricular vagus nerve stimulation.

This study protocol follows the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines.⁴⁰ Table 2 shows the SPIRIT flow diagram summarising study procedures, outcome measures, randomisation and assessments.

Table 2. Schedule of enrolment, interventions and assessments.

Study period	Enrolment	Baseline and allocation (T1)	Intervention	Postintervention (T2)	Follow-up (T3)
Timepoint	-t1	0	Weeks 1–4	Week 4	Week 8
Enrolment:
Eligibility screening (online)	X
Informed consent	X
Baseline assessments (T1)		X
Randomisation		X
Interventions:
taVNS stimulation (active)
Sham stimulation
Device fitting and stimulation tolerance		X
Compliance monitoring			Continuous
Assessments:
Misophonia symptom severity (AMISO-R, SMS)		X		X	X
Mental health comorbidities (SCL-90-R)		X		X	X
Audiological tests (PTA, ULL, TFI, HQ, HIQ)		X		X
MRI (resting-state, task-based, structural)		X		X
Behavioural task (Emotional Go-No-Go)		X		X
Psychophysiology (ECG, GSR)		X		X
Weekly side effect monitoring			X
Acceptability measures (SUS, comfort, willingness to reuse)				X
Blinding efficacy assessment					X
Serious adverse event monitoring			Continuous
Retention/completion monitoring				X

AMISO-R, Amsterdam Misophonia Scale Revised; fMRI, functional MRI; GSR, galvanic skin response; HIQ, Hyperacusis Impact Questionnaire; HQ, Hyperacusis Questionnaire; PTA, Pure Tone Audiometry (note only tested at T1); SCL-90-R, Symptom Checklist-90-Revised; SMS, Sussex Misophonia Scale; SUS, systems usability scale; taVNS, transcutaneous auricular vagus; TFI, Tinnitus Functional Index.

Sample size and statistical power

Sixty participants will be recruited (30 in each condition). Sample size is based on an a-priori estimate using G*Power⁴¹ to give sufficient power (0.80) to detect a medium effect size (f=0.25, p<0.05) on misophonia symptom severity for two groups (taVNS vs sham) at two time points (pre vs post-treatment) considering 25% attrition. Since taVNS has never been trialled in misophonia, we have been conservative by using a medium effect size. This is based on studies with both similar designs (randomisation to active and sham stimulation) and taVNS protocols (>2 weeks stimulation), which have reported medium to large effect sizes (eg, d=0.44 in migraine;⁴² d=1.36 and d=0.62 in depression).^{32 43} This is also conservative based on misophonia-based CBT trial effect sizes (d=1.97 (7)).

Recruitment and screening

Participants will be recruited by advertising sent to our misophonic participant volunteer list. This is a list of around 1200 UK-based individuals with clinically significant misophonia who have agreed to be contacted about misophonia research. We will also advertise around our university campus, including online research participation pools. Interested parties will be sent our participant information sheet (see online supplemental materials) and invited to take a prescreening questionnaire online to determine eligibility. Potential participants will be identified according to inclusion and exclusion criteria based on their responses and invited to participate. Those agreeing to participate will attend their first in-person site visit for baseline assessment after written and signed informed consent has been obtained. Note that informed consent will also include the use of anonymised data for future research.

Inclusion criteria

All participants will meet the following criteria: (1) moderate to severe misophonia symptoms (scores >20 on the revised Amsterdam Misophonia Scale⁴⁴ and scores >75 on the Sussex Misophonia Scale (SMS),⁴⁵ (2) not currently receiving active treatment for misophonia or planning to during the study, (3) aged between 18 and 40 years*, (4) right-handed*, (5) fluent English speakers*, (6) no history of brain injury or neurological conditions, (7) corrected or corrected to normal vision, (8) normal hearing (except for hyperacusis, tinnitus or misophonia) (9) owning and being willing to use a personal smartphone for tolerability questionnaires and stimulation reminders to be sent via a study app.

Criteria 3–5 above are included to assist in our functional MRI analyses where left-handedness, language fluency and brain age make greater differences in brain organisation, which make data harder to interpret.

Exclusion criteria

Exclusion criteria are as follows: (1) pregnant or trying to get pregnant, (2) active implants (eg, cochlear implants, any metallic implants, vagal nerve stimulating implants), (3) cerebral shunt, (4) wounds or skin diseases near the ear, (5) history or family history of heart problems, (6) history of seizures/epilepsy and (7) any other contraindications to an MRI scan.

Intervention

At their first on-site visit, participants are randomly allocated to receive either active taVNS or sham stimulation with the transcutaneous vagus nerve stimulation (tVNS) R device (tVNS Technologies GmbH). The tVNS R device has a stimulation unit and ear electrode, which delivers a low-voltage electrical signal to the ear. Stimulation is applied to the left ear with active stimulation applied to the cymba concha (rich vagus nerve branch distributions) and sham stimulation to the superior scapha (no vagus nerve distribution). Electrical stimulation consists of a 25 hz wave delivered in 28s-on 32s-off bursts. The device is explained to the participant and then fitted by the study researcher to set stimulation intensity. Participants are instructed on how to use the device independently (both oral and written instructions for how to apply stimulation at home are provided).

For maximum efficacy, stimulation intensity is adjusted and set to individual thresholds. We will ensure participant tolerance of stimulation by setting individual stimulation thresholds during device training. Current will be introduced at the lowest intensity (0.1 mA) and gradually increased to a level at which the participant just begins to report discomfort (maximum intensity is 5 mA). The device will then be adjusted to an intensity below the reported discomfort level and will remain at this intensity for at least 2 min to confirm tolerance. This intensity will be considered the maximum tolerated stimulation intensity for the participant and stimulation throughout the intervention will be administered at this level, with the option to lower up to 0.3 mA if required. If an individual’s stimulation tolerance lowers during the study beyond 0.3 mA, then we will repeat the stimulation tolerance protocol described above and adjust accordingly. Rates of stimulation and any changes will be reported in a stimulation tolerance monitoring log.

In order to ensure appropriate stimulation application at home each participant has a 30–60 min session to set up their stimulation and teach them how to correctly place the electrode. Before leaving participants have demonstrated their ability to independently place the electrode on themselves correctly to our unblinded RA (EMWK), and our RA takes a picture of the correct placement location for each individual participant that they can refer to in order to ensure correct placement at home. Additionally, participants are given both a written guide and are sent a video guide both of which demonstrate how to correctly place the electrode at home, and participants are regularly checked in by our Unblind RA who clarifies any questions that arise after participants have left.

Stimulation is applied for at least 30 min two times per day, at least 5 days per week by participants for the treatment period (4 weeks). Protocol compliance is monitored via the tVNS R stimulation logs and reminders to simulate are sent via a study app. Other than the stimulation location, active and sham groups are identical.

Randomisation and allocation concealment

Following baseline assessments, participants are randomly allocated to the taVNS or sham group at a 1:1 ratio. Random allocation sequencing is generated within SPSS (RV.UNIFORM(0,1)) by an independent research assistant not involved in any other aspect of the study. The randomisation order will be kept in numbered sealed envelopes, such that participants get their assignment to condition according to the order they enter the study. Enrolment to the study and other testing or monitoring other than taVNS allocation and training will be conducted by LJR.

Blinding

Active and sham devices are identical with the only difference being the stimulation location. An unblinded researcher (EMWK) will be setting the stimulation parameters for each participant and providing device training for at-home administration according to the randomisation schedule. All other researchers involved in the data collection and analyses are blinded to allocation group until after analyses are performed. To prevent participants from unblinding by researching taVNS, participants are informed that we are interested in the efficacy of taVNS at two different locations (ie, we do not explicitly say there is an active and sham condition). We also use the term tVNS and not specifically auricular taVNS, so that there’s more search results with more varying locations used if participants do search online. Blinding efficacy is measured at time three with two questions: (1) ‘Which type of stimulation do you believe you received?’ Active versus sham (placebo) versus I don’t know/not sure and (2) ‘How confident are you in your answer?’ 1 (not at all confident) to 10 (extremely confident). All participants are unblinded at the end of the study. Unblinding during the study is only permissible if serious adverse events occur and the investigators deem it appropriate to know treatment allocation.

Harms and adverse events

Data from previous studies show that taVNS is safe and well tolerated, with the most common side effects being local skin irritation (reported in 18% of participants) and less commonly (<4%) headache, nasopharyngitis, dizziness, fatigue and nausea.^{20 46} We do not expect the intervention to cause harm or significant discomfort for participants.

Participants are asked to notify a member of the research team if they are experiencing any adverse events that may be related to the stimulation. As well as immediate reporting, participants complete weekly questionnaires to document any potential side effects of stimulation. Specifically, participants rate the severity of each known (although rare) symptom (pain, tingling, burning, headache, nausea, dizziness, fatigue, tinnitus, nasopharyngitis) on scales from 0 (absence of symptom) to 3 (presence of symptom and intolerable). They are also asked about potential worsening of their misophonia symptoms and to list any additional side effects. Weekly logs will be promptly reviewed and any side effects reported will be further investigated, with participants contacted via email or phone. Participants will be closely monitored until resolution of any adverse events with details documented and logged. To assess safety and tolerability, a series of independent t-tests (or equivalent robust tests) will be conducted to compare the average presence and severity of adverse events and potential worsening of misophonia during the active versus sham stimulation.

In the unlikely event that a study participant experiences an adverse event, we will determine its seriousness, whether it is expected or not, and whether it is likely due to the intervention. Assessment of these three aspects follows guidance on clinical safety data management including definitions and standards for expedited reporting set out by the European Medicines Agency. Documentation will include the following aspects of any adverse events: (1) severity, (2) relationship to intervention, (3) action taken regarding study intervention (none, stopped, reduced dose, delayed dose), (4) outcome of adverse event, (5) whether expected or not and (6) seriousness. If a participant experiences a serious adverse event during the study related to the stimulation, the PI will notify the ethics board and sponsor within 5 days. If it is determined that the stimulation caused or may have caused the adverse event, the participant may be removed from the study and the protocol will be either modified or stopped.

Feasibility and acceptability

Feasibility is reported by collecting data on: (1) recruitment and retention rates, (2) protocol adherence rates and (3) intervention completion rates. Acceptability is assessed at T2 with the 10-item System Usability Scale⁴⁷ and two additional questions asking about comfort with the device and whether the participant would use it again.

Primary outcome measures

Misophonia

Symptom severity (our primary outcome measure) is measured with the Amsterdam Misophonia Scale Revised (AMISO-R,⁴⁸ chosen to enable comparison of taVNS to 3-month CBT), which showed an average reduction of 9.7 points (7). The AMISO-R consists of a severity scale consisting of 10 items rated on a Likert scale from 0 (none) to 4 (extreme), giving a total score between 0 and 40 where higher scores indicate higher misophonia severity.

We will additionally use the SMS, a validated misophonia measure,⁴⁵ which contains two parts. In part 1, participants select which of 48 frequent misophonia triggers within eight categories they have. Part 2 contains 39-items (eg, ‘Certain sounds make me feel disgusted, even if those sounds don’t disgust other people’), which are rated for how often they happen from 0 (never) to 4 (always). Part 2 of the SMS yields a total score between 0 and 156 with participants scoring ≥51 considered to have misophonia. As well as a total score, the SMS also gives 5-subscale scores of Feelings and Isolation, Intersocial Reactivity, Life Consequences, Avoidance or Repulsion, and Pain. The SMS will allow us to determine taVNS-related changes in trigger categories and functioning across five different domains.

Secondary outcome measures

Mental Health Comorbidities

The Symptom Checklist-90-Revised (SCL-90-R)⁴⁹ is a 90-item questionnaire administered to capture psychological and psychiatric symptoms across nine primary symptom dimensions: (1) somatisation (distress related to one’s body/physiological experiences), (2) obsessive-compulsive (intrusive thoughts and compulsive actions), (3) interpersonal sensitivity (self-perceived inadequacy/inferiority in social relationships), (4) depression (low mood and decreased sense of meaning), (5) anxiety (anxious symptoms and experienced tensions), (6) hostility (aggressiveness towards others), (7) phobic anxiety (fears related to specific stimuli), (8) paranoid ideation (projections to others and persecutory cognitions) and (9) psychoticism (psychotic and schizophrenic behaviours). Items are rated on 5-point Likert scales from 0 (not at all) to 4 (extremely), yielding total scores between 0 and 360 as well as the 9 subscale scores listed above.

Audiological assessment

Misophonia while a distinct condition co-occurs with audiological conditions of hyperacusis and tinnitus at higher than otherwise expected rates (42). We will, therefore, conduct several audiological and self-report measures capturing both tinnitus and hyperacusis and we will account for any change misophonia outcomes while controlling for other audiological changes using the below assessments.

Pure-tone audiometry (T1 only)

Participants’ hearing sensitivity will be assessed using pure tone audiometry according to the British Society of Audiology Recommended Procedures.⁵⁰ The procedure will be performed in a quiet room with ambient noise level below 20 dB SPL (ISO 8253–1:2010) using a calibrated audiometer (GSI Audiostar Pro) and supra-aural headphones (Sennheiser HDA300). This Hughson-Westlake method will be used to determine hearing threshold levels for each ear across 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz and 8000 Hz frequencies in octave intervals. The lowest level at which the participant responds to at least two out of the three (or three out of five) presentations will be recorded as the hearing threshold level for that frequency. Hearing will be considered within normal limits if thresholds are ≤20 dB HL across all tested frequencies.

Uncomfortable loudness level testing

Uncomfortable loudness levels (ULLs) will be measured at 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz and 8000 Hz in each ear using the same room and audiometer. The procedure will be performed according to the British Society of Audiology Recommended Procedures⁵¹ with modifications to instructions proposed by Aazh and Moore⁵² to minimise discomfort. Pure tone stimuli will be presented manually (roughly 1 s with 1–3 s silence gaps) starting at a comfortable level (eg, 60 dB) at each test frequency and increased in 5 dB increments until the participant reports the sound as uncomfortably loud but not painful. This intensity will be recorded as the ULL for that frequency. If the participant shows behavioural signs of discomfort such as flinching, the presentations will cease, and the last level will be recorded as the ULL. If no discomfort is reported up to 90 dB HL, the ULL for that frequency will be recorded as 95 dB HL. Hyperacusis will be considered present when the ULLmin is ≤77 dB HL.⁵³

Tinnitus

Tinnitus is a condition sometimes comorbid with misophonia (42), so we will measure this preintervention and postintervention. Participants are initially given a description of tinnitus and asked to indicate whether they experience it. If they answer affirmatively, they will complete the Tinnitus Functional Index (TFI);⁵⁴ a 25-item questionnaire asking about tinnitus-related distress (eg, ‘Over the past week, how often did your tinnitus make it difficult to fall asleep or stay asleep?’). For each item, participants answer on a 11-point scale either from 0 to 10 or 0%–100%. Total summed TFI scores will be computed ranging from 0 to 250 with higher scores indicating greater perceived tinnitus-related impairment.

Hyperacusis

Hyperacusis is a sound sensitivity condition sometimes comorbid with misophonia,⁴⁴ so we will measure this preintervention and postintervention. Participants will complete the Hyperacusis Questionnaire (HQ);⁵⁵ a 14-item questionnaire evaluating attentional, social and emotional responses to sounds (eg, ‘Do you have difficulty listening to conversations in noisy places?’). Responses are made on 4-point Likert scales from 0 (no) to 3 (yes, a lot). Following a precedent in the misophonia literature to refine the scoring methodology to distinguish between general hypersensitivity to sound (hyperacusis) and specific aversions to certain sounds (misophonia), separate subscales will be calculated for the 5 hyperacusis-specific items (total scores between 0 and 20) and the 5 misophonia-specific items (total scores between 0–20).⁵⁶ Participants will also complete the Hyperacusis Impact Questionnaire (HIQ);⁵⁷ an 8-item questionnaire in which participants indicate how often over the past 2 weeks they have experienced problems related to sound loudness (eg, ‘unable to relax in noisy places’) rated on 4-point Likert scales from 0 (0–1 days) to 3 (6–7 days). Total summed HIQ scores will be computed ranging from 0 to 24 with higher scores indicating greater negative impact of hyperacusis.

Experimental outcome measures

Neuroimaging measures (at T1 and T2)

Participants undergo a 3T MRI scan session, which includes: (1) a 6 min task-free, resting-state functional MRI scan (rs-fMRI—eyes open look at fixation cross) to capture brain activity at rest, (2) a high-resolution T1-weighted anatomical scan and (3) a sound exposure fMRI paradigm designed to elicit misophonic responses. The sound exposure paradigm uses 18 videos from,⁵⁸ which includes misophonic, aversive and neutral stimuli. Each condition (misophonic, aversive, neutral) comprises six 12 s clips (with a fixation cross interstimulus interval of 0.8 s), randomised per participant, with order of condition counterbalanced. Six, twelve second periods of fixation cross will also be included as a baseline. At the end of each video, participants will rate how ‘pleasant’ they found it from 1 (very unpleasant) to 10 (very pleasant) as well as how ‘disgusting’ and ‘annoying’ they found it (both rated from 1 (not at all) to 10 (completely)).

Behavioural task measures (at T1 and T2)

Participants will complete a misophonic emotional Go/No-Go task to capture cognitive control and emotion regulation. Two conditions (order counterbalanced) are: (1) ‘NoGo-Misophonic’ where participants respond as quickly as possible to neutral sounds (GO stimuli) and ignore misophonic sounds (NoGo stimuli) (2) ‘NoGo-Neutral’ where participants respond as quickly as possible to misophonic sounds (GO stimuli) and ignore neutral sounds (NoGo stimuli). Sounds were obtained from Hansen et al⁵⁹ and Daniels et al (only published their misophonic and unpleasant sounds (the latter not used here) but had additional neutral sounds that we use here.)⁶⁰ and were precategorised as either misophonic or neutral as part of these studies. Each condition has 200 trials with 70% GO trials to invoke a prepotent tendency for a GO response. Stimulus duration is 500 ms with 1000 ms interstimulus interval. Each condition lasts at least 5 min to ensure sufficient continuous recording for high frequency heart rate variability (HF-HRV) metrics.

At the end of each block participants will rate how ‘pleasant’ they found the sounds in that block from 1 (very unpleasant) to 10 (very pleasant) as well as how ‘disgusting’ and ‘annoying’ they found the sounds in that block (both rated from 1 (not at all) to 10 (completely)). Core outcomes are false alarm rates in each condition (measured by failing to withhold a response on NoGo trials). In the ‘NoGo-Misophonic’ condition, false alarm rates index emotion regulation (a reduced ability to inhibit a prepotent behavioural response in the context of emotional information). In the ‘NoGo-Neutral’ condition, false alarm rates index cognitive control (a reduced ability to inhibit a prepotent behavioural response in the context of neutral information).

Psychophysiological measures (at T1 and T2)

Physiology is recorded using the Biopac MP160 system (250 samples/s) during the behavioural task described above and analysed with Biopac AcqKnowledge software. ECG activity is recorded with three electrodes placed underneath the ribs on both sides, and on the right-side collarbone. Electrodermal (skin conductance) activity is recorded with two Ag-AgCl electrodes wrapped around the index and ring finger of the left hand, acquired by applying a small direct current (0.5V) through electrodes. Resting heart rate and heart rate variability and electrodermal activity are measured for 5 min under task-free conditions (eyes open, look at fixation cross). These measures are also obtained during exposure to misophonic and neutral sounds during the emotional Go/No-Go task described above.

Data monitoring

Internal monitoring will be conducted weekly to ensure that the protocol and Good Clinical Practice standards are followed. The monitor (LJR) will review trial data ensuring its accuracy and completeness as well as monitoring all signed informed consent documents and adverse event logs. Quality assurance audits by relevant regulatory authorities may be performed.

Data analysis

Per-protocol analyses will be performed first to assess intervention efficacy under ideal conditions. For per-protocol analyses, we will define inclusion as an adherence of 75% of completed stimulation sessions (ie, 75% of 40 30 min stimulation sessions across the intervention period). This is in-line with existing taVNS literature^{32 61 62} while being conservative considering our relatively long 4-week self-administration design. Second, intention-to-treat analyses will be conducted including all randomised participants regardless of protocol completion. This approach provides a comprehensive evaluation of intervention effectiveness considering potential real-world adherence issues while still retaining benefits of randomisation. Missing data will be handled by applicable statistical methods such as multiple imputation. With regards to blinding assessments, we will complete analyses with and without participants who correctly guessed the allocation with high confidence to ensure that any results are not driven by participants who have correctly guessed their condition, but this approach still ensure that we can maintain statistical power if these individuals are not significantly impacting the results.

Primary analyses for our outcome measures will be a series of 2 (intervention: taVNS vs sham) × 2 (time: T1 vs T2) mixed analysis of variance (ANOVAs) for each outcome measure. Where appropriate, such as analyses for our primary outcome measure of misophonia severity, we will run linear mixed effects models instead of mixed ANOVA’s including random intercepts and slopes to model each participants unique trajectories. We expect to find significant interaction effects, which would indicate improvements in outcome measures over time for the taVNS but not control group. We will also explore whether any improvements are sustained 1-month later (T3). For misophonia severity, we will also compare change in AMISO-R with prior a prior 3-month misophonia trial (7). We will use false discovery rate to control for multiple comparisons.

We ask our participants to disclose any comorbidities and any medication they use during our initial eligibility screening assessment. We do not exclude individuals on the basis of this information, but we will analyse if presence of comorbidities or if taking medications has any independent impact on our results by looking at differences in change on misophonia severity based on each of these factors (presence or absence of comorbidities, whether individuals are taking medication). If we find significant differences in either case, we will include these factors as random effects in our linear mixed effects models.

Primary analyses for our behaviour/physiological task will be a series of 2 (intervention: taVNS vs sham) × 2 (time: T1 vs T2) × 2 (condition: NoGo-Misophonic vs NoGo-Neutral) mixed ANOVAs for each outcome measure (false alarm rates, HF-HRV, skin conductance level (SCL), skin conductance response (SCRs)). We predict the following changes over time in the taVNS compared with sham group: (1) a reduction in false alarm rates (especially for NoGo-Misophonic), (2) an increase in HF-HRV (especially for NoGo-Neutral), (3) a decrease in SCL and SCRs (especially for NoGo-Neutral).

Resting-state data will be analysed to estimate patterns of functional connectivity within and across brain networks (eg, default mode network, auditory network, salience network), as defined by standardised and commonly used anatomical and functional atlases (eg, AAL atlas, Schaefer parcellation) as well as any Regions of Interest identified from analyses of the task-based scan. Network time-averaged maps will then be compared between groups (taVNS and sham) and across time points (T1 and T2). We expect to identify changes involving the functional profiles of both unimodal sensory regions and networks (eg, auditory network) as well as multimodal association cortices (eg, default mode network) and subcortical regions implicated in emotion processing and regulation (eg, amygdala, insula).

Task-based analyses. Contrasts of interest will be defined to compare neural activity between stimulus categories (eg, Misophonic>Aversive, Misophonic>Neutral, Aversive>Neutral) at T1 and T2 for each participant, across the whole brain. This will identify brain regions with a selective response to contrasting conditions (Misophonic vs Aversive vs Neutral stimuli). Next, we will calculate differences in activity between T1 and T2 for each contrast. This will allow us to compare changes over time between taVNS and sham conditions in neural responses to stimuli.

Task-based analyses will be complemented with psychophysiological interactions analysis to explore task-based synchronisation of neural activity. This will document functional connectivity between neural regions in the context of misophonic stimuli exposure and any changes to these context-specific functional connectivity patterns due to taVNS.

Dissemination

Any amendments to the protocol will only be implemented after approval from the Brighton and Sussex Medical School ethics committee. Any important protocol modifications will be updated in the trial registry.

At the end of the study, data will be collated into a single spreadsheet, with a data key indicating the content of each variable. The behavioural and questionnaire data will be stored in open data file formats (e.g., as comma delimited text files (.csv)). MRI images will be stored in DICOM and NIFTI file formats. All formats will allow anonymised sharing and long-term validity of the data. All original raw data will be retained for future use and access requests. Study data will be deposited in the University of Sussex server via Figshare (www.sussex.figshare.com) or another online data repository (e.g. www.OpenfMRI.org) where appropriate. The data that will be deposited at the group level, allowing other users to undertake the same analyses as those reported in the published studies, as well as new ones. Anonymised individual participant datasets and supporting metadata will be made available upon request. For the datasets containing individual anonymised patient data, a data-sharing agreement will be required.

Trial results will be published in peer-reviewed journals regardless of the outcome of the study (i.e., if the results are inconclusive or demonstrate that taVNS is not effective for the treatment of misophonia). Results will also be shared at national or international conferences. All publication rights belong to the principal investigator and authorship determined according to the Contributor Roles Taxonomy (CRediT) system ⁶³