Clinical and investigational tools for monitoring noise-induced hyperacusis

Abstract

Hyperacusis is a recognized perceptual consequence of acoustic overexposure that can lead to debilitating psychosocial effects. Despite the profound impact of hyperacusis on quality of life, clinicians and researchers lack objective biomarkers and standardized protocols for its assessment. Outcomes of conventional audiologic tests are highly variable in the hyperacusis population and do not adequately capture the multifaceted nature of the condition on an individual level. This presents challenges for the differential diagnosis of hyperacusis, its clinical surveillance, and evaluation of new treatment options. Multiple behavioral and objective assays are emerging as contenders for inclusion in hyperacusis assessment protocols but most still await rigorous validation. There remains a pressing need to develop tools to quantify common nonauditory symptoms, including annoyance, fear, and pain. This review describes the current literature on clinical and investigational tools that have been used to diagnose and monitor hyperacusis, as well as those that hold promise for inclusion in future trials.

I. INTRODUCTION

As many as 24% of adults and 17% of teenagers in the United States demonstrate hearing test results that are consistent with ear damage from exposure to noxious noise levels (Henderson et al., 2011; NIDCD, 2019). Noise-induced hearing loss (NIHL) can occur after a single exposure to an intense impulse sound or as a consequence of repeated exposure to loud sounds over time. Both types of noise exposure can lead not only to threshold elevation but also to hearing deficits that are undetectable by standard audiometric assessments. Even when hearing sensitivity is normal or near-normal, many individuals with substantial noise exposure history experience difficulties encoding complex suprathreshold stimuli, tinnitus (ringing or noises in the ears or head), hyperacusis (decreased sound tolerance), or some combination (Bramhall et al., 2019; Le Prell, 2019). While research and clinical interest in suprathreshold processing deficits and tinnitus have grown substantially over the past several years, noticeably less attention has been paid to identifying and quantifying the generation of hyperacusis or changes in sound tolerance following acoustic overexposure.

Hyperacusis is a multifaceted condition wherein everyday sounds of low-to-moderate intensity are described as unbearably loud, aversive, painful, or some combination thereof (Pienkowski et al., 2014; Tyler et al., 2014). Population-based estimates suggest that hyperacusis affects between 8.6% and 15.2% of adults (Andersson et al., 2002; Fabijanska et al., 1999; Paulin et al., 2016; Smit et al., 2021) and the condition often leads to debilitating psychosocial consequences, including employment challenges, sleep disturbances, social phobia, anxiety, and depression (Jüris et al., 2013; Paulin et al., 2019; Tyler et al., 2014). Despite the profound impact of hyperacusis on quality of life, there are no universally accepted or effective treatments for the condition (Baguley and Hoare, 2018; Potgieter et al., 2017). The dearth of successful interventions is due in large part to a lack of consensus on appropriate subjective and objective clinical protocols for the diagnosis, surveillance, and management of hyperacusis symptoms.

Although symptoms of decreased sound tolerance have been documented in a variety of syndromes (Tyler et al., 2014), several researchers have posited that noise exposure is likely the most common cause of hyperacusis (Anari et al., 1999; Axelsson and Hamernik, 1987; Pienkowski et al., 2014; Tyler et al., 2014). Relationships between hyperacusis-like symptoms and occupational noise exposure have been documented in teachers and childcare workers (Anari et al., 1999; Fredriksson et al., 2021; Jüris et al., 2013; Meuer and Hiller, 2015; Sjödin et al., 2012), military personnel (Muhr and Rosenhall, 2010), and call center operators (McFerran and Baguley, 2007; Parker et al., 2014). Complaints of hyperacusis are also common amongst musicians (Anari et al., 1999; Couth et al., 2020; Halevi-Katz et al., 2015; Kähäri et al., 2003; Laitinen and Poulsen, 2008; Liberman et al., 2016; Schmuziger et al., 2006; Toppila et al., 2011; Wartinger et al., 2019) and individuals with high levels of self-reported noise exposure (Smit et al., 2021; Smith et al., 2019; Williams et al., 2021).

Emerging evidence suggests that the risk of developing hyperacusis may begin at sound exposure levels that are below permissible occupational noise exposure limits (Fredriksson et al., 2021). In a large retrospective study of 8328 Swedish women, Fredriksson et al. (2021) reported a significantly increased risk of hyperacusis for individuals who worked in environments with noise exposure levels of 75–85 dB(A) as compared to those with occupational noise exposure of <75 dB(A). The risk of hyperacusis was tripled amongst preschool teachers, who are often exposed to brief but high intensity impulse sounds throughout the workday (Fredriksson et al., 2021). While these correlational studies point toward a likely relationship between noise exposure and sound intolerance symptoms, prospective studies aimed at quantifying dose-response relationships between noise exposure and hyperacusis in humans remain a priority.

More direct evidence linking noise exposure and hyperacusis-like physiological and behavioral responses comes from decades of work on acoustic overexposure in animal models. In rodents, peripheral noise injury leads to a cascade of downstream central changes, including increased spike synchrony and spontaneous neural activity, overrepresentation of neural edge frequencies, poor neural adaptation to distractor noises, and abnormally elevated sound-evoked activity (Auerbach et al., 2014; Eggermont, 2017; Komiya and Eggermont, 2000; Seki and Eggermont, 2003). It is widely theorized that noise-induced hyperacusis arises from excess central gain that occurs following acoustic trauma, wherein a paradoxical increase in sound-evoked activity is observed across all levels of the central auditory system from the cochlear nucleus (Brozoski et al., 2002; Schrode et al., 2018) to the midbrain (Salvi et al., 1990; Salvi et al., 2000; Shaheen and Liberman, 2018) to the cortex (Asokan et al., 2018; Noreña et al., 2010; Resnik and Polley, 2017; Sun et al., 2012). This excess gain is generally most pronounced at later stages of central processing, particularly in the cortex, and likely results in a maladaptive amplified perception of sound that is similar to what patients with hyperacusis experience (Auerbach et al., 2014).

In fact, elevated sound-evoked activity along the central auditory pathway is accompanied by hyperacusis-like behavior in noise-exposed rodents (Chen et al., 2013; Hickox and Liberman, 2014; Manohar et al., 2017; Schrode et al., 2018; Sun et al., 2012). Following acoustic overexposure, rodents exhibit exaggerated startle behavior alongside sound-evoked neural hyperactivity in the auditory brainstem and cortex (Hickox and Liberman, 2014; Schrode et al., 2018; Sun et al., 2012). Notably, several investigations demonstrate similar central and behavioral hyperactivity following peripheral damage induced by ototoxic drugs (Auerbach et al., 2014).

Rodent studies support a central origin of hyperacusis following subtle or overt damage to the auditory periphery. While this central gain model of hyperacusis is ubiquitous in the literature, a compelling link between excess central gain and perceptual hypersensitivity has yet to be forged in humans. However, many existing psychoacoustic and electrophysiological tools may be harnessed to potentially elucidate the underlying neural mechanisms and successfully monitor hyperacusis in clinical and research settings. This review summarizes the measurement strategies that have been used to characterize audiologic profiles of individuals with hyperacusis and highlights the myriad of opportunities for clinically accessible investigational tools that could be used to quantify and survey sound intolerance symptoms.

II. METHODS

Several PubMed literature searches were completed in January 2022 to identify peer-reviewed studies that assessed subjective or objective auditory function in individuals who have hyperacusis. The PubMed search engine is developed and maintained by the National Center for Biotechnology Information (NCBI) at the United States National Library of Medicine (NLM) and contains over 34 × 10⁶ citations, dating back to at least the year 1966. A variety of search terms were combined with the words “hyperacusis” and “decreased sound tolerance” to identify relevant studies, including, for example, noise exposure, NIHL, questionnaire, audiogram, loudness, evoked potentials, otoacoustic emissions (OAEs), speech perception, neuroimaging, etc. The author carefully reviewed the titles and abstracts of resulting citations to determine whether the articles evaluated subjective or objective aspects of auditory function in humans who had hyperacusis. The full-text articles were reviewed for references to other relevant published data that were not identified during the initial search. When information about a particular aspect of auditory processing in hyperacusis was unavailable or severely limited in scope, additional PubMed searches were conducted to identify human or animal studies that could shed light on the utility of an investigational tool in the assessment of hyperacusis. Only articles written in the English language were included except for one French (Khalfa et al., 2001) and two German (Nelting and Finlayson, 2004; Nelting et al., 2002) citations for widely used self-report questionnaires. All relevant full-text articles were downloaded from the University of Texas at Dallas (UTD) library.

III. CONVENTIONAL METHODS FOR ASSESSING HYPERACUSIS IN CLINICAL AND RESEARCH SETTINGS

Conventional recommendations for the clinical assessment of hyperacusis reflect the dearth of targeted diagnostic assays for the condition. In addition to the standard pure-tone audiogram, the American Speech-Language-Hearing Association (ASHA) recommends that audiologists assess loudness discomfort levels (LDLs) and administer subjective patient questionnaires to assist in the differential diagnosis and surveillance of hyperacusis symptoms (American Speech-Language-Hearing Association, 2018). In a similar vein, studies characterizing the audiologic profiles of individuals with hyperacusis have focused almost exclusively on conventional clinical measures, including standard audiometry (Anari et al., 1999; Bläsing et al., 2010; Brandy and Lynn, 1995; Sheldrake et al., 2015; Smit et al., 2021), tonal LDLs (Aazh and Moore, 2017a, 2018; Anari et al., 1999; Bläsing et al., 2010; Brandy and Lynn, 1995; Sheldrake et al., 2015; Zaugg et al., 2016), and self-report questionnaires (Aazh and Moore, 2017b; Anari et al., 1999; Bläsing et al., 2010; Brandy and Lynn, 1995; Fackrell et al., 2015; Nelting et al., 2002; Paulin et al., 2016; Urnau and Tochetto, 2011).

A. Behavioral hearing sensitivity

The pure-tone audiogram remains the gold standard test for evaluating hearing sensitivity within the standard 0.25–8 kHz frequency range, and it is an important component of the hyperacusis test battery. Although the relationship between hyperacusis and audiometric thresholds is complex (Tyler et al., 2014), many individuals with hyperacusis present with some form of sensorineural hearing loss that is bilateral and mild-to-moderate in degree, on average (Anari et al., 1999; Bläsing et al., 2010; Nelson and Chen, 2004; Sheldrake et al., 2015; Smit et al., 2021). While hearing loss tends to be most pronounced in the high frequencies (>2 kHz), Smit et al. (2021) demonstrated that adults with hyperacusis had significantly poorer hearing thresholds at all octave frequencies between 0.5 and 8 kHz as compared to age-matched individuals without hyperacusis. This difference was observed even in frequency regions where average hearing thresholds were within the normal range (Smit et al., 2021).

In fact, many patients with hyperacusis have clinically normal hearing sensitivity across all standard audiometric frequencies, suggesting that perceptual hypersensitivity can arise through peripheral damage that is undetectable by conventional hearing tests (Anari et al., 1999; Brandy and Lynn, 1995; Sheldrake et al., 2015; Urnau and Tochetto, 2011). In individuals with normal audiograms, it is possible that hyperacusis is triggered by cochlear synaptopathy, or a loss of synapses between the inner hair cells and type I auditory nerve fibers. Findings from experimental studies demonstrate that rodents with noise-induced primary neural afferent degeneration exhibit paradoxically strong growth of neural responses across a range of suprathreshold intensities (Asokan et al., 2018; Chambers et al., 2016; Hickox and Liberman, 2014; Resnik and Polley, 2017). Emerging data in humans suggests that individuals at high risk of ear damage due to significant noise exposure history show physiological evidence of cochlear synaptopathy, as well as heightened reactions to sounds that are consistent with hyperacusis (Liberman et al., 2016). Similarly, elevated loudness perception across a range of suprathreshold sound intensities has been documented in ears with evidence of possible cochlear synaptopathy (Jahn et al., 2022).

Although the audiogram remains an important tool for identifying and monitoring the progression of hearing loss, it is evident that hearing sensitivity at standard audiometric frequencies (0.25–8 kHz) is insufficient for diagnosing and monitoring hyperacusis. Additional information about cochlear damage may be obtained by measuring hearing sensitivity at extended high frequencies (>8 kHz). Several investigations suggest that frequencies above 8 kHz may be more sensitive to noise-induced damage than conventional audiometry (Sheppard et al., 2020). To date, there are no published data characterizing extended high frequency (EHF) threshold profiles in individuals with hyperacusis. However, Liberman et al. (2016) showed that adults with notable noise exposure history and heightened aversion to everyday sounds had significantly poorer EHF thresholds (10–16 kHz) than a group of age- and sex-matched controls. Notably, thresholds at standard audiometric frequencies (0.25–8 kHz) were within normal limits and did not differ as a function of noise exposure history in that study (Liberman et al., 2016). Further, Sanchez and Roberts (2021) found that adolescents with tinnitus and reduced sound level tolerance had significantly poorer EHF thresholds (14–16 kHz) than a control group (Sanchez and Roberts, 2021). While empirical investigations quantifying EHFs in individuals with hyperacusis are warranted, monitoring EHF thresholds when conventional audiometry is within normal limits may provide a useful metric of peripheral damage in this population.

B. LDLs

LDLs, or the lowest intensity levels judged by an individual to be uncomfortably loud for a given stimulus, are commonly used clinically and in research settings to characterize an individual's ability to tolerate everyday sounds. Beyond the conventional audiogram, LDLs are the only psychoacoustic test currently recommended for the differential diagnosis of hyperacusis in the United States (American Speech-Language-Hearing Association, 2018). However, it has been suggested that LDLs do not provide a sufficient degree of sensitivity nor specificity to serve as a sole diagnostic marker of heightened loudness perception (Sheldrake et al., 2015). Although LDLs are often abnormally low in individuals with hyperacusis, the range of reported LDLs varies widely within and across investigations. Many individuals with hyperacusis have LDLs comparable to listeners with normal hearing and vice versa (Aazh et al., 2018; Aazh and Moore, 2017a; Jahn et al., 2022; Sheldrake et al., 2015).

The use of LDLs as a diagnostic marker of hyperacusis is further complicated by the fact that there is no consensus on the measurement technique, diagnostic cut-off criteria, or the optimal stimuli with which to elicit the response (Punch et al., 2004). The criteria used to categorize pure-tone LDLs as normal or abnormal are widely inconsistent across studies, ranging from <95 dB hearing level (HL; Goldstein and Shulman, 1996) to <70 dB HL (Anari et al., 1999) to ≤77 dB HL in the worse ear (Aazh and Moore, 2017a), to name a few. Absent a clear diagnostic consensus, it is imperative that clinicians and researchers document their exact methodologies and remain consistent across repeated test sessions.

Due to the immense variability of LDLs and inconsistent protocols, the relationship between one's self-reported ability to tolerate sound in everyday life and their LDLs remains ambiguous (Anari et al., 1999; Filion and Margolis, 1992; Knudson and Melcher, 2016; Zaugg et al., 2016). Knudson and Melcher (2016) showed that reduced LDLs are associated with enhanced startle responses in individuals with tinnitus but not to their self-reported hyperacusis severity. Similarly, Zaugg et al. (2016) found only weak correlations between LDLs and subjective reports of sound tolerance. These findings suggest that when measured in isolation, LDLs fail to capture the multifaceted nature of the hyperacusis condition. While many individuals with hyperacusis do report loudness intolerance, others report symptoms of sound-related distress or physical ear pain that can occur independent of their psychoacoustic loudness percept (Tyler et al., 2014; Urnau and Tochetto, 2011). The exact mixture of loudness hypersensitivity, emotional reactivity, and pain is likely specific to the individual. Although LDLs may provide a clear indication of hyperacusis in extreme cases of reduced loudness tolerance (Aazh and Moore, 2017a), they will be most informative when used in combination with additional subjective and objective tools to quantify the status of the auditory system and lived experience of the patient.

C. Self-report questionnaires

Patient self-report is an important component of the hyperacusis test battery as the condition is predominantly characterized by subjective symptoms that remain difficult to quantify objectively. Several self-report questionnaires exist to assist clinicians and researchers in assessing the functional impact of hyperacusis on everyday life. Here, we highlight the strengths and limitations of the most common questionnaires used to evaluate hyperacusis in clinical and research settings. Table I provides a summary of pertinent details from the initial validation studies for each questionnaire.

TABLE I.

An overview of hyperacusis questionnaire development studies.

Questionnaire	Purpose	Participants	Item development	Scoring	Relation to other measuresa
Hyperacusis questionnaire (HQ;b Khalfa et al., 2002)	Quantify and characterize the clinical presentation of hyperacusis	201 adults (132 female; mean age = 28 years old) randomly selected from the general population	Based on literature review	14 items scored on a 4-point scale ranging from “no” (0 points) to “yes, a lot” (3 points)	Higher HQ scores were associated with self-reported decreased noise tolerance
Multiple Activities Scale of Hyperacusis (MASH; Dauman and Bouscau-Faure, 2005)	Characterize situations during which an individual experiences noise discomfort	249 adults (105 female; mean age ∼ 53 years old) who had tinnitus	Not specified	Structured interview quantifies the level of sound-related annoyance during everyday activities (rated from 0 to 10)	MASH scores positively correlated with the level of self-perceived “annoyance of hyperacusis” (rated from 0 to 10)
German Questionnaire on Hypersensitivity to Sound (GÜF; Nelting et al., 2002;c Nelting and Finlayson, 2004;c Bläsing et al., 2010)	Assess subjective distress related to hypersensitivity to sound	226 hospitalized adults who had tinnitus combined with hyperacusis (27.9% inpatients; Nelting et al., 2002)	Based on clinician statements defining hypersensitivity to sound	15 items scored on a 4-point scale ranging from “never correct” (0 points) to “always correct” (3 points)	GÜF total score positively correlated with hyperacusis impairment as judged by an ear specialist (Nelting et al., 2002)
Noise Avoidance Questionnaire (NAQ, Blaesing and Kroener-Herwig, 2012)	Assess avoidance of sound in everyday life	Sample 1: 74 adults (40 females; mean age = 52 years old); sample 2: 86 adults (45 females; mean age ∼ 49 years old) who had either tinnitus alone, tinnitus combined with hyperacusis, or neither condition	Guided by two existing questionnaires (Avoidance-Endurance Questionnaire and Mobility Inventory)	25 items scored on a 5-point scale ranging from “never” (0 points) to “very often/always” (4 points)	NAQ scores positively correlated with GÜF scores; NAQ scores negatively correlated with LDLs; participants with hyperacusis had significantly higher NAQ scores than those without hyperacusis
Inventory of Hyperacusis Symptoms (IHS; Greenberg and Carlos, 2018)	Quantify the impact of decreased sound tolerance on quality of life, mental health, and general functioning	450 adults (60% female; mean age = 35 years old) across 37 countries (53% United States); 80% had tinnitus	Based on literature review, analysis of existing hyperacusis questionnaires, and qualitative data collected from audiologists and individuals who had tinnitus combined with auditory hypersensitivity	25 items scored on a 4-point scale ranging from “not at all” (1 point) to “very much so” (4 points)	IHS scores significantly correlated with self-report measures of quality of life, anxiety, depression, tinnitus severity, and overall levels of auditory sensitivity
Hyperacusis Impact Questionnaire (HIQ; Aazh et al., 2022)	Assess the impact of hyperacusis on the patient	266 adults (146 female; mean age = 54 years old) who had tinnitus alone or tinnitus combined with hyperacusis; retrospective analysis	Based on common complaints of hyperacusis patients who were seen at a tinnitus and hyperacusis therapy clinic in the United Kingdom	Eight items scored on a four-point scale ranging from “0–1 days” (0 points) to “11–14 days” (3 points)	HIQ score negatively correlated with minimum LDL; HIQ score positively correlated with HQ score and with self-report measures of mental health and tinnitus handicap
Sound Sensitivity Symptoms Questionnaire (SSSQ; Aazh et al., 2022)	Assess the type and severity of sound intolerance symptoms	266 adults (146 female; mean age = 54 years) who had tinnitus alone or tinnitus combined with hyperacusis; Retrospective analysis	Based on categories of hyperacusis described by Tyler et al. (2014)	5 items scored on a 4-point scale ranging from “0-1 days” (0 points) to “11-14 days” (3 points)	SSSQ score negatively correlated with age and minimum LDL; SSSQ score positively correlated with HQ score and self-report measures of mental health and tinnitus handicap

^a The relation to other measures as reported in the initial validation studies.

^b A 25-item modified version of the HQ is widely used in research and clinical practice, but it has not been validated. Fackrell et al. (2015) have also proposed a ten-item version of the HQ.

^c Nelting et al. (2002) and Nelting and Finlayson (2004) are written in the German language. A follow-up study by Bläsing et al. (2010) provides an English translation of the GÜF and is referenced in Table I.

The Hyperacusis Questionnaire (HQ; Khalfa et al., 2002) is the most widely used scale in clinical and research settings to identify hyperacusis symptoms and monitor treatment-related changes. The HQ consists of 14 questions that aim to characterize hyperacusis symptoms over 3 major dimensions (attentional, social, and emotional). The questions were developed based on a review of the hyperacusis literature, which was notably limited in scope at the time of publication. Participants in the original validation study consisted of 201 adults from the general population with unknown hyperacusis status who were primarily university students.

Despite its popularity as a tool for quantifying hyperacusis symptoms, the validity and reliability of the HQ have been called into question by multiple investigations (Fackrell et al., 2015; Greenberg and Carlos, 2018; Meeus et al., 2010; Wallén et al., 2012). Fackrell et al. (2015) demonstrated that the original factor structure of the HQ did not accurately quantify symptoms of auditory hypersensitivity in a cohort of 265 adults with tinnitus; however, they did report moderate correlations between the total HQ score and LDLs. Based on their findings, the authors proposed a new, ten-item version of the HQ that awaits validation (Fackrell et al., 2015). Some clinicians and researchers have alternatively adopted a 25-item modified version of the HQ (American Speech-Language-Hearing Association, 2018; Khalfa et al., 2001); however, the validity of the modified HQ has not been evaluated empirically.

Another popular self-report measure, the German Questionnaire on Hypersensitivity to Sound (GÜF), was designed to assess subjective distress related to hyperacusis (Bläsing et al., 2010; Nelting et al., 2002). The GÜF was first developed and validated in a cohort of 226 patients that suffered from hypersensitivity to sound and chronic tinnitus. The questionnaire items were generated from statements made by physicians and psychologists, describing hypersensitivity to sound. In a separate cohort of 91 patients who had both hyperacusis and tinnitus, Bläsing et al. (2010) determined that GÜF scores correlated more strongly with self-reported tinnitus severity than with self-reported hyperacusis severity, and there was no significant relationship between GÜF scores and LDLs. Moreover, the data from Bläsing et al. (2010) did not support the original factor structure of the GÜF. It is important to note that the English translation of the GÜF has not been validated even though it is often recommended for clinical use (American Speech-Language-Hearing Association, 2018).

The Noise Avoidance Questionnaire (NAQ) assesses self-reported sound avoidance behavior (Blaesing and Kroener-Herwig, 2012). Its development was guided by two existing questionnaires: the Avoidance-Endurance-Questionnaire (Hasenbring et al., 2009), which assesses avoidance behavior in response to pain, and the Mobility Inventory (Chambless et al., 1985), which quantifies agoraphobic behavior. The NAQ was initially evaluated in 28 individuals with tinnitus, 28 individuals with tinnitus and hyperacusis, and 30 control participants (Blaesing and Kroener-Herwig, 2012). While little information has been documented about the reliability and internal consistency of the scale, Blaesing and Kroener-Herwig (2012) demonstrated a significant, positive correlation between scores on the NAQ and GÜF.

Of the existing hyperacusis questionnaires, the Inventory of Hyperacusis Symptoms (IHS; Greenberg and Carlos, 2018) has arguably withstood the most rigorous tests of validity and reliability. The original IHS items were developed through a combination of scientific literature review and qualitative data collected from audiologists and individuals who had tinnitus and symptoms of auditory hypersensitivity. The initial validation of the IHS included data from 450 adults across 37 countries and resulted in a factor structure that allowed differentiation between primary hyperacusis subtypes of loudness, annoyance, fear, and pain (Greenberg and Carlos, 2018; Tyler et al., 2014). Aazh et al. (2021) found moderately strong, positive correlations between IHS scores and scores on the HQ in 100 patients with tinnitus, hyperacusis, or both conditions. In that study, participants with hyperacusis received significantly higher scores on the IHS than did those without hyperacusis (Aazh et al., 2021). A recent publication describes two new short-form questionnaires (Table I) that focus on assessing the impact of hyperacusis on the patient and are largely based on relevant items from the IHS and HQ (Aazh et al., 2022).

Oftentimes, open-ended interviews allow for more flexibility and individualized approaches to patient care than a set of standardized multiple-choice questions. The Multiple Activities Scale of Hyperacusis (MASH) offers a structured interview-style assessment that quantifies the level of sound-related annoyance that an individual experiences during various everyday activities (Dauman and Bouscau-Faure, 2005). Like many of the closed-set questionnaires, the MASH was developed in patients with chronic tinnitus (n = 249). While the scale has not been evaluated for internal consistency, reliability, or construct validity, scores on the MASH are strongly correlated with the level of self-perceived “annoyance of hyperacusis,” rated as a single value between 0 and 10 (Dauman and Bouscau-Faure, 2005).

In summary, few of the existing HQs were developed and validated using a comprehensive array of best practices for the creation of health-related scales (Boateng et al., 2018). Moreover, the internal consistency, reliability, and validity of many popular hyperacusis scales have been challenged or remain unevaluated. One possible explanation for these shortcomings is that most existing scales were not initially validated on individuals with hyperacusis as a primary complaint. Additionally, patients with hyperacusis were generally not interviewed during the item development phase, which is considered a valuable component of qualitative data collection and valid scale development (Boateng et al., 2018). To ensure that the lived experience of hyperacusis is adequately captured by self-report measures, future work should aim to include individual stakeholders in the scale development and validation process.

IV. BEYOND THE CONVENTIONAL AUDIOLOGIC TEST BATTERY

It is evident that the conventional audiologic test battery is insufficient for quantifying the multifaceted symptoms associated with hyperacusis, monitoring treatment outcomes, and understanding the functional impact of hyperacusis on quality of life. A handful of investigations have ventured beyond the standard audiometric test battery to explore the utility of loudness growth measures, OAEs, middle ear muscle reflexes (MEMRs), auditory evoked potentials, and cortical sound processing in individuals with hyperacusis or decreased sound tolerance. Although they are pending additional validation, these measures hold promise as clinically feasible tools that could assist in the differential diagnosis of hyperacusis and be incorporated into future investigational trials.

A. Loudness growth

Given the immense variability in LDLs amongst individuals with and without hyperacusis, it is likely beneficial to characterize loudness perception across the full auditory dynamic range rather than solely at the endpoints. While few studies have reported loudness growth data in individuals with hyperacusis, Brandy and Lynn (1995) provided evidence that adults with hyperacusis have steeper loudness growth at 1000 Hz than adults without hyperacusis. There is also evidence that loudness perception at sound intensities across the dynamic range can be modified in individuals with hyperacusis through structured auditory listening experiences (Noreña and Chery-Croze, 2007). Further, Jahn et al. (2022) showed that ears with estimated cochlear neural degeneration have elevated loudness perception for stimuli presented at intensity levels in the middle of the dynamic range, but not necessarily at the LDL (Jahn et al., 2022).

Although published data on loudness growth and hyperacusis are limited, incorporating loudness growth measures into many clinical and research protocols should be relatively innocuous. As discussed previously, it is common to measure LDLs as a component of the hyperacusis test battery. Oftentimes, and especially when testing individuals with decreased sound tolerance, LDLs are measured by first presenting sounds at very low intensities and gradually increasing the level until the patient reports that it is uncomfortable. When testing LDLs in patients with hyperacusis, Aazh and Moore (2017c) recommend a starting level equal to the audiometric threshold at the test frequency, which implies that the range of sound levels presented during the test should encompass the individual's full dynamic range. Regularly documenting loudness perception across the full auditory dynamic range would improve our understanding of whether loudness perception for intensity levels in between threshold and LDL can serve as a sensitive psychoacoustic marker of hyperacusis. Clinicians and researchers can implement loudness growth testing by including clinically feasible loudness scaling procedures such as the Contour Test of Loudness Perception (Cox et al., 1997) in their LDL protocols.

B. Physiological measures

1. MEMRs

Relationships between loudness discomfort and the sound level that elicits a reflexive contraction of the middle ear muscles have been reported (Olsen, 1999), suggesting a possible role for the MEMR in the assessment of hyperacusis. The MEMR represents bilateral contractions of the stapedius muscles that are evoked by a moderate-to-high intensity elicitor stimulus presented either ipsilaterally or contralaterally. These contractions stiffen the ossicular chain, altering the ratio of absorbed and reflected sound measured in the ear canal in response to an ipsilateral probe stimulus. MEMR strength is defined as the change in sound pressure in the probe ear during versus before presentation of the elicitor stimulus.

Data from animal studies suggest that the strength of the MEMR may be sensitive to noise damage (Valero et al., 2015; Valero et al., 2018), wherein weaker MEMRs are elicited in synaptopathic cochlear regions than in non-synaptopathic areas. Similarly, humans with clinically normal hearing and evidence of possible cochlear synaptopathy (Mepani et al., 2020) or noise-induced tinnitus (Wojtczak et al., 2017) show weaker MEMRs than control subjects. There is also evidence that temporary auditory deprivation leads to adaptive brainstem plasticity in individuals with normal hearing that is reflected in changes to MEMR thresholds and loudness perception (Munro et al., 2014; Munro and Blount, 2009). Munro and Blount (2009) compared MEMR thresholds across the two ears of adult participants before and after unilateral earplug use. Following one week of earplug-induced auditory deprivation, MEMRs were elicited at lower sound pressure levels in the ears that had been plugged. Conversely, MEMRs were measured at higher levels in the unplugged (control) ears following auditory deprivation. In a follow-up investigation, the researchers replicated the MEMR findings and showed that loudness perception was also elevated bilaterally following unilateral auditory deprivation (Munro et al., 2014).

To our knowledge, two investigations have quantified MEMR thresholds in individuals with hyperacusis-like symptoms. Although Brandy and Lynn (1995) did not find a difference in MEMR thresholds between hyperacusic and non-hyperacusic groups, they did observe abnormally low MEMR thresholds in a subgroup of individuals with endocrine disorder. Similarly, Ohmura et al. (2019) found that MEMR thresholds were lower in individuals with autism spectrum disorder (ASD), a condition that often co-occurs with hyperacusis, relative to a neurotypical control group. In that study, hyperacusis severity was negatively correlated with MEMR threshold across participants, irrespective of ASD status (Ohmura et al., 2019). While data on the MEMR and hyperacusis are emergent, available studies suggest that the MEMR holds promise for the differential diagnosis of hyperacusis and possibly even for distinguishing between etiological subtypes.

2. OAEs

OAEs offer an indirect measure of cochlear health that may be used as an objective tool to supplement the audiogram when behavioral hearing sensitivity is within normal limits. In a healthy ear, primary tone pairs generate distortion products that reflect active nonlinearities associated with the motility of the outer hair cells (OHCs). In addition to these frequency-specific distortion product otoacoustic emissions (DPOAEs), cochlear nonlinearities can be assessed rapidly over a broad frequency range by measuring transient-evoked otoacoustic emissions (TEOAEs) following click stimulation. A large body of literature suggests that OAEs can provide an early marker of noise damage in animal models (e.g., Fraenkel et al., 2001; Franklin et al., 1991; Jimenez et al., 2001; Vázquez et al., 2001; Vázquez et al., 2004) and humans (e.g., Helleman et al., 2010; Helleman et al., 2018; Job et al., 2009; Marshall et al., 2001; Sisto et al., 2007; Sliwinska-Kowalska and Kotylo, 2001; Sutton et al., 1994), wherein OAE amplitudes are often reduced following acoustic overexposure. Likewise, OAE amplitudes could theoretically provide an indication of early noise-induced peripheral damage in individuals with hyperacusis.

While traditional DPOAEs have not been widely studied in humans with hyperacusis, emerging evidence points toward the possibility of a hyperresponsive medial olivocochlear (MOC) system in individuals with decreased sound tolerance (Knudson et al., 2014; Wilson et al., 2017). The MOC system is part of the efferent auditory pathway that controls the mechanical movement of the OHCs (Guinan, 2006). Auditory nerve fibers of the MOC system project to a subset of multipolar neurons in the posteroventral cochlear nucleus (PVCN) followed by the superior olivary complex (SOC). From the SOC, MOC neurons project bilaterally to the OHCs in the cochlea, where they play a role in modulating cochlear gain and auditory nerve activity.

As a gain-control mechanism, the MOC system has been hypothesized to play a role in the generation of hyperacusis (Sturm and Weisz, 2015). The strength of the MOC system can be estimated by comparing the magnitude of DPOAEs measured during the presence versus absence of sound in the contralateral ear. In humans with normal hearing, this effect is generally suppressive, wherein the magnitude of the OAE is reduced by sound presented to the contralateral ear. A hyperresponsive system, as may be expected in individuals with hyperacusis, would theoretically exhibit greater OAE suppression than an ear without hyperacusis. Using this method, Knudson et al. (2014) assessed the magnitude of contralateral DPOAE suppression in individuals with and without tinnitus who demonstrated varying degrees of sound level tolerance. They observed greater suppression of contralateral DPOAE amplitudes in the participants with tinnitus and those with relatively low sound level tolerance, suggesting increased responsiveness of the MOC system. In children with ASD and hyperacusis, there is some evidence of greater contralateral TEOAE suppression as compared to their non-hyperacusic counterparts (Wilson et al., 2017). However, others have found no association between hyperacusis-like symptoms and MOC effects in either adults with tinnitus and hyperacusis (Urnau and Tochetto, 2012) or children with ASD (Ohmura et al., 2019).

3. Auditory evoked potentials

Auditory evoked potentials allow for noninvasive evaluation of neural responses along the peripheral and central auditory pathways. As described previously, the Central Gain Model of hyperacusis posits that peripheral damage from acoustic overexposure leads to enhanced sound-evoked activity in the central auditory system. The comparison of auditory evoked responses across generators at multiple levels of the auditory system may serve as an objective biomarker for the peripheral deficits and central compensation that are widely hypothesized to underlie hyperacusis.

a. Auditory brainstem responses and electrocochleography.

The auditory brainstem response (ABR) occurs within 10 ms after auditory stimulation and provides an objective index of the health of the auditory system from the cochlea through the brainstem. In humans, the primary generators of the ABR waves are complex, but they are generally believed to represent responses from the distal portion of the eighth nerve (wave I), the medial portion of the eighth nerve (wave II), the cochlear nucleus (wave III), neurons from the SOC (wave IV), and the fibers terminating from the lateral lemniscus in the contralateral inferior colliculus (wave V; Møller and Jannetta, 1985; Parkkonen et al., 2009). Waves I, III, and V are most often used to diagnose lesions along the auditory pathway in humans.

It has been hypothesized that individuals with noise-induced hyperacusis may exhibit reduced wave I amplitudes or elevated amplitudes for waves III and V relative to individuals without hyperacusis. Mice with noise-induced cochlear neural degeneration and hyperacusis-like behavior indeed show significantly reduced ABR wave I amplitudes and either unchanged or enhanced amplitudes for later waves relative to unexposed controls (Hickox and Liberman, 2014; Schrode et al., 2018). These data imply the presence of a compensatory neural hyperactivity in the auditory brainstem that can be induced by acoustic overexposure and recorded noninvasively.

Indirect evidence from individuals with noise exposure history or tinnitus provides clues into the utility of ABR measures in the differential diagnosis of noise-induced hyperacusis in humans. Liberman et al. (2016) quantified cochlear and auditory nerve function using click-evoked electrocochleography recordings in adults with (high-risk) and without (low-risk) substantial noise exposure history (Liberman et al., 2016). Individuals in the high-risk group reported heightened levels of sound annoyance and avoidance and showed differences in the electrocochleographic waveform peaks that were consistent with signs of cochlear synaptopathy. Specifically, the summating potential (SP)/action potential (AP) ratio, which represents the ratio between the waveform peaks generated by the cochlear hair cells (i.e., the SP) versus the auditory nerve fibers (i.e., the AP), was significantly elevated in the high-risk group as compared to the low-risk group. In a follow-up study, Jahn et al. (2022) showed that elevated SP/AP ratios coincided with elevated loudness perception at suprathreshold sound intensities in adults with normal audiograms. There is also evidence that individuals with tinnitus, which often co-occurs with hyperacusis, have enhanced amplitude ratios for wave V/I (Gu et al., 2012; Schaette and McAlpine, 2011) and wave III/I (Gu et al., 2012) relative to non-tinnitus controls. Together, these findings suggest that noise-induced loss of cochlear nerve synapses may contribute to the generation of central hyperactivity and hyperacusis-like symptoms, and ABR recordings may serve as effective noninvasive biomarkers for the condition in humans.

b. Auditory cortical responses.

The effects of noise exposure on central auditory function in hyperacusis may also be elucidated through cortically derived evoked potentials, including the middle latency response (MLR) and P1-N1-P2 cortical evoked response. In humans, the MLR represents a series of negative and positive waves (i.e., Na, Pa, Nb, and Pb) that occur between 10 and 50 ms post-stimulus onset. The Na and Pa waves of the MLR likely represent activation near the primary auditory cortex or Heschl's gyrus (Picton, 2011a). The last of the waves (Pb or P50) is essentially equivalent to P1 of the P1-N1-P2 cortical evoked response. The P1-N1-P2 cortical evoked response occurs between 50 and 180 ms post-stimulus onset and is generated by dipole fields in multiple areas of the cortex, most prominently from the auditory regions of the superior surface of the temporal lobe (Picton, 2011b).

Consistent with central gain theories, noise-exposed rodents demonstrate increased cortical evoked response amplitudes alongside decreased ABR amplitudes (Popelar et al., 2008; Syka et al., 1994; Syka and Rybalko, 2000). A recent report in young noise-exposed military Veterans showed that small ABR wave I amplitudes were associated with reduced MLR amplitudes and increased late latency response (i.e., P1-N1-P2) amplitudes (Bramhall et al., 2020). In that study, Veterans who reported noise exposure and chronic tinnitus had the largest cortical responses (Bramhall et al., 2020). In a case report of three adults with hyperacusis, MLR latencies tended to decrease following the provision of treatments aimed at expanding the auditory dynamic range (Formby et al., 2017). These findings from animals and humans suggest that reduced peripheral afferent input may lead to compensatory gain in the central auditory system.

Cortical event-related potentials have also been extensively investigated to infer the capacity of the auditory system to detect and predict signal regularity. Predictive coding models suggest that auditory predictive processes may rely on the presence of inhibitory neural templates in the central auditory system that serve to enhance the salience of novel sounds while diminishing the perception of repetitive sounds (Durai et al., 2018). It has been theorized that noise-induced sensory deafferentation triggers central neuroplasticity mechanisms that can disrupt the relationships between auditory sensory-memory, predictive coding, and attention-switching processes in the cortex (Durai et al., 2018).

Mismatch negativity (MMN) paradigms allow one to infer how the brain responds to violation of a stimulus pattern. A deviant-tone stimulus is presented amongst frequent standard-tone stimuli, and the brain generates a negative response waveform approximately 140–210 ms after presentation of the deviant stimulus. Multiple studies have demonstrated MMN amplitude reductions for auditory deviants amongst individuals with chronic tinnitus relative to non-tinnitus controls (Asadpour et al., 2020; Attias et al., 1993; Durai et al., 2018; Holdefer et al., 2013; Mahmoudian et al., 2013; Mohebbi et al., 2019). These findings suggest that noise-induced central compensatory mechanisms may extend to supratemporal (Giard et al., 1990; Tse et al., 2013) and frontal (Fishman, 2014; Giard et al., 1990) brain regions, leading to disrupted cognitive processing. The observed relationships between cortical event-related potentials and enhanced central gain suggest that auditory evoked potentials hold promise for the study of hyperacusis generation, and this remains an open area of investigation.

c. Envelope following responses.

Noise exposure impacts hearing at multiple stages of the auditory pathway, from the ear to the brain, and is hypothesized to affect suprathreshold sound processing by degrading complex temporal cues. Envelope following responses (EFRs) are steady-state auditory evoked potentials that faithfully mirror the slow-varying temporal envelope of an amplitude-modulated stimulus and may provide a window into how the auditory system processes temporal information across multiple time scales in individuals with noise-induced hyperacusis. Multiple aspects of the stimulus, including the carrier frequency, amplitude modulation (AM) rate, and AM depth, can be manipulated to facilitate comprehensive investigation of responses along the auditory pathway. Most importantly, careful stimulus selection can allow the experimenter to reasonably infer the neural generator of the response. Specifically, the degree to which neurons adequately represent temporal regularities in the auditory signal decreases along the ascending auditory pathway, wherein neurons in the auditory nerve can phase lock up to a few thousand hertz while neurons in the auditory cortex can only phase lock up to about 100 Hz (Joris et al., 2004).

Insights from cochlear synaptopathy and aging studies suggest that EFRs may prove useful in elucidating peripheral damage and central hyperactivity. Rodent studies demonstrate that the amplitude of EFR responses to high modulation rates (700–1000 Hz) are reduced in mice with noise-induced (Shaheen et al., 2015) and age-related (Parthasarathy and Kujawa, 2018) cochlear synaptopathy. In aging humans, EFR responses are generally degraded for high-rate stimuli but not for low-rate stimuli (∼20–40 Hz), suggesting declines in peripheral processing with concomitant central compensation (Anderson and Karawani, 2020). Moreover, most investigators agree that loudness perception is driven by higher-order cortical processing (Röhl and Uppenkamp, 2012), and there is some evidence to suggest that low-rate EFR responses (∼40–105 Hz) may serve as neural correlates of loudness perception (Van Eeckhoutte et al., 2016; Ménard et al., 2008; Parker and O'dwyer, 1998; Zenker Castro et al., 2008). Thus, EFRs may provide a tool to garner insight into perceptual and neural changes that occur following noise exposure, as well as the sites of lesion and enhancement along the central auditory pathway in individuals with hyperacusis.

4. Advanced neuroimaging techniques

The most compelling evidence supporting an association between central hyperactivity and hyperacusis in humans comes from two functional magnetic resonance imaging (fMRI) studies (Gu et al., 2010; Koops and van Dijk, 2021). Koops and van Dijk (2021) investigated subcortical and cortical sound-evoked responses in 35 participants with hearing loss and tinnitus with and without hyperacusis. They demonstrated that sound-evoked activity was greatest in subcortical and cortical auditory areas for the participants who had hyperacusis, and this hyperactivity extended across frequency regions with normal and impaired hearing (Koops and van Dijk, 2021). Similarly, Gu et al. (2010) assessed sound-evoked central auditory activation in individuals with and without tinnitus who had varying levels of sound tolerance. In that study, participants with relatively low sound level tolerance showed elevated sound-evoked activity in the auditory midbrain, thalamus, and primary auditory cortex relative to participants with higher sound level tolerance. Similar findings were reported in a rodent model of hyperacusis, wherein fMRI revealed hyperactive sound-evoked responses in the auditory cortex and inferior colliculus of rats with abnormally elevated loudness perception (Wong et al., 2020). While advanced neuroimaging techniques, such as fMRI and magnetoencephalography (MEG), may offer a promising research avenue to further elucidate the neural generators of hyperacusis and empirically evaluate potential treatment-related effects, it is noted that such equipment is expensive and not readily available to clinicians.

V. SPECIAL CONSIDERATIONS FOR SOUND-EVOKED TESTING IN INDIVIDUALS WITH HYPERACUSIS

As clinicians and researchers delve into the myriad of opportunities to explore new diagnostic and treatment avenues, it is critical to remember that sound-evoked testing is stressful for many individuals with hyperacusis. If not performed with proper care, many of the tests described in this review could lead to discomfort, pain, and diminished trust in scientists and healthcare professionals. Before proceeding with any assessment, it is crucial to obtain the patient's informed consent and ensure them that the testing can be stopped at any time. Participants should be informed about the level and duration of sound that will be presented for each test and whether that will change throughout the session. All tests (including audiometry) should proceed by presenting the acoustic stimuli at very low intensity levels and increasing the level gradually.

Many assessments, such as auditory evoked potentials and MEMR responses, traditionally require presenting sounds at moderate-to-high intensity levels, sometimes for prolonged durations. As discussed in this review, many individuals with hyperacusis have LDLs comparable to individuals without hyperacusis (Sheldrake et al., 2015); therefore, these tests are not necessarily contraindicated in this population. However, care must be taken to ensure that the sound levels are tolerable prior to testing and the participant is well-informed of the planned procedure. Better yet, researchers should work toward developing novel methods for eliciting objective auditory responses that do not require prolonged use of moderate-to-high intensity stimuli and can be applied across patient populations (Picton et al., 2007).

VI. BEYOND HEARING: AFFECTIVE SOUND PROCESSING AND PAIN

Many existing candidate assays of hyperacusis focus on quantifying loudness perception or sound-evoked activity along the canonical auditory pathway. However, hyperacusis often manifests as aversive reactions to sound that are not purely auditory, including annoyance, fear, anxiety, stress, and physical pain (Tyler et al., 2014). To ultimately address the multifaceted nature of hyperacusis, it is imperative to approach diagnosis and management with assays that are not traditionally included in evaluations of auditory function.

Rodent models of hypersensitivity show amplified sound-evoked neural responses in regions of the central nervous system that are involved in processing aversive auditory stimuli, including the amygdala and auditory-limbic network (Chen et al., 2016; Chen et al., 2014; Chen et al., 2015; Manohar et al., 2017). Enhanced sound-evoked responses in limbic regions have been linked to the generation of sound-evoked fear and avoidance behaviors in the rodents (Chen et al., 2016; Chen et al., 2014; Manohar et al., 2017). Thus, data from animal models support a central gain mechanism that extends beyond the canonical auditory pathway and may reflect the negative emotional hyperreactivity to sound that is reported by patients with hyperacusis.

A recent report documented a novel psychoacoustic tool that may be used to quantify affective sound processing in individuals with hyperacusis (Enzler et al., 2021). In that study, participants (n = 81) with and without hyperacusis rated the pleasantness of natural sounds, tone pips, and noises presented at various intensity levels. Using a subset of sounds that could best discriminate between the individuals with and without hyperacusis, Enzler et al. (2021) developed a metric called the Core Discriminant Sounds (CDS) score. Such assessments hold promise for characterizing hyperacusis complaints beyond standard assessments of loudness perception and auditory function.

In addition to heightened emotional reactions to sound, most patients with hyperacusis report some form of sound-induced pain, with 52.6% indicating that their pain symptoms were triggered by exposure to loud noise (Williams et al., 2021). Despite the prevalence of pain hyperacusis, the mechanisms underlying sound-evoked pain remain elusive. In mice, Flores et al. (2015) described a noncanonical auditory pathway between the cochlea and brain (possibly mediated by type-II cochlear neurons) that may detect noise-induced tissue damage and act as a pain receptor (Flores et al., 2015). The authors hypothesized that this auditory nociceptive system may elicit a reflex that triggers unpleasantness or a protective efferent response to intense noise. Research to elucidate the underlying mechanisms of sound-induced pain and develop targeted diagnostic assays and treatments for pain hyperacusis remains a foremost priority in the field.

VII. CONCLUSION

This review describes the current state of the literature regarding measurement strategies that have been used clinically and in research studies or could potentially be used pending further investigation to quantify and monitor hyperacusis symptoms. Published reports demonstrate substantial variability in methodology and outcomes, a lack of clear consensus on protocols and diagnostic criteria, and limited data from individuals with primary complaints of hyperacusis. These factors currently confound our ability to define a set of best practice protocols. Conventional recommendations for the assessment of hyperacusis (e.g., pure-tone audiometry, LDLs, and subjective questionnaires) are informative but insufficient to comprehensively characterize the condition on an individual level. Several clinically accessible candidate assays show promise for the differential diagnosis of hyperacusis, but they require further validation before they can be incorporated into clinical practice and investigational trials. Investigators are encouraged to subject available tools to rigorous validation studies, develop novel methods for assessing the multifaceted nature of decreased sound tolerance, and design standardized protocols that are effective and safe for individuals with hyperacusis.