Age-related hearing loss and its consequences

Hearing loss is common and currently affects 1.6 billion people worldwide [1]. The consequences of hearing loss have a significant impact on individual quality of life and place a considerable burden on global health and economic systems [2], with a conservative cost-analysis estimated at $981 billion [3]. This is exacerbated due to a rapidly ageing population, with hearing loss projected to extend globally to 2.1 billion people by 2050 [4]. Age-related hearing loss (ARHL, or presbycusis) is a progressive and bilateral sensorineural hearing loss that impacts both central and peripheral auditory functions, and disproportionately affects higher frequencies [5, 6]. ARHL is highly prevalent amongst older adults, with the estimated incidence doubling for every ten-year increase in age [7]. For hearing losses greater than 25 dB (equivalent to a mild hearing loss or worse), the estimated incidence varies from 42.3% to 63.1% for adults aged 70 years and older, depending on test context and population factors [7, 8].

ARHL is associated with communication challenges, which is typically exemplified by difficulties perceiving speech in the presence of noise [9–11]. This is significant given the congregation of humans at social activities and experiences such as dining-out at eateries are typified by noisy sound propagation. Real world noise levels indicate that older adults with mild to moderate hearing loss spend approximately 7.5% of their time in listening situations where noise meets or exceeds speech levels [12]. Importantly, while speech-in-noise (SIN) tasks used in clinical and research contexts are typically receptive in nature (i.e., focussed on individual ability to perceive SIN); the real-world outcomes are far broader, with significant social and emotional impacts that extend to communication partners and beyond [13, 14]. The ability to communicate effectively is fundamental to the development of strong interpersonal relationships and helps counteract loneliness and isolation, which is becoming a growing public health challenge [15]. Mechanistically, it has been posited that ARHL communication challenges reduce active social participation; this withdrawal has a cascading effect leading to social isolation and feelings of loneliness; finally, there are also increased risks to health and mortality [16–19]. While these risk factors are associated with general ageing [20, 21], these risks are significantly exacerbated by hearing loss.

There is currently no way to reverse ARHL and management is generally focused on the provision of assistive listening devices such as hearing aids to amplify sounds. However, despite the noted benefits of hearing aids for communication and broad quality of life outcomes [22], hearing aid adoption rates are low, with global estimates indicating 17% of persons that could benefit from a hearing aid do so [23]. Hence, most older adults with hearing loss have unaddressed hearing loss [24], acknowledging this has been referred to synonymously as untreated [17], unperceived [25], or uncorrected [26]. To the best of our knowledge, there is no gold-standard terminology and there may be subtle variations in definition. For the purposes of our study, a participant with unaddressed hearing loss refers to an individual with a mild-to-moderate sensorineural hearing loss that does not currently use an assistive listening device such as a hearing aid.

Benefits of group singing

The importance of social support, social networks, and social participation has been highlighted as a target for audiological rehabilitation, specifically for adults with ARHL [19]. In response to this, music and arts-based interventions for health and wellbeing have come to prominence in recent years with the benefits of being effective for preventing and promoting better health outcomes; helping to manage and treat health; and being engaging and cost-effective [27]. One potential non-clinical intervention showing promise is group singing—a universally practiced music activity that is more common than individual singing in approximately 70% of all societies [28]. There is an intimate and important cultural connection between music making and social context, with group singing being strongly associated with religious practice, dance, games, ceremony, and work [28].

Community choirs are popular, with 3.5 million Canadians having sung in 28,000 choirs in 2016 [29]. Typically consisting of amateur singers, the focus is to provide a supportive and inclusive environment where people can enjoy singing. These activities promote community engagement, cultural enrichment, and social connections among participants. Moreover, a diverse group of older adults that participated in a methodologically robust community choir intervention experienced significant benefits to loneliness and interest in life, compared to controls [30].

A potential benefit of group singing is the enhancement of speech perception, which is particularly relevant as ARHL is primarily associated with difficulties perceiving SIN [9–11]. Here, we provide a brief overview of two key theoretical frameworks that support speech perception: the OPERA hypothesis [31–33], and the Processing Rhythm in Speech and Music (PRISM) framework [34].

The OPERA hypothesis posits that music may engage speech processing networks if certain conditions are met. 1) Overlap—sensory and cognitive processing of music and speech is encoded by brain networks that overlap. More commonly referred to as neural sharing, there is ample neurological evidence to support the overlap between music and speech processing [35, 36]. 2) Precision—music must place a higher sensory and cognitive demand than speech. For example, pitch perception thresholds in music tend to operate at the semitone interval, whereas the speech intonation curve of a question utterance is much larger, typically extending beyond an octave (or 12 semitones) [37, 38]. 3) Music training that is enjoyable, regularly practiced, and engaging will satisfy the final requirement related to Emotion, Repetition, and Attention.

The PRISM framework suggests that three rhythm processes in both speech and music may lead to improved speech perception. 1) Precise auditory processing—both speech and music perception require accurate processing of small deviations. However, the regularity and precision required in music production (particularly when performing with others) may fine-tune the auditory system, leading to enhancements in perceiving speech processes such as temporal envelope [39], or formant tracking (such as in plosives such as /ba/ and /da/) [40]. 2) Synchronization/entrainment of neural oscillations to external rhythmic stimuli—because musical rhythm typically features temporal regularities, music is a useful stimulus to entrain neural oscillations. Entrainment optimizes processes associated with attention and prediction [41], and hierarchical processing (i.e., phonemic, syllabic, phrase level speech perception) [42, 43], playing a critical role in speech and music perception. 3) Sensorimotor coupling—auditory and motor networks are linked and implicated in effective processing of both speech and music domains [44, 45]. Several studies have shown that music training has the capacity to strengthen sensorimotor coupling [46, 47], supporting a potential mechanism for the musician advantage.

Group singing also confers a unique opportunity to practice sensorimotor coupling—the dynamic integration of auditory, somatosensory, and motor systems [48, 49], which likely involves a dorsal pathway and its constituent fibre tracts integrating auditory cortices to the inferior frontal gyrus through the posterior parietal cortex [50]. This same dorsal pathway has been implicated in speech perception in noise, particularly as listening conditions become more challenging [51]. However, because a choir will experience tempo drift and micro-perturbations of intonation, a very high demand is placed on brain areas and fibre tracts subserving sensorimotor coupling. Singers in a choir are understandably motivated to coordinate sensory input and motor output to ensure accurate and harmonious vocal performance. Thus, motivated practice over a period of weeks and months may lead to neuroplastic changes (e.g., enhanced motor areas of the brain), which may ultimately support SIN perception [52, 53].

The musician advantage was brought to prominence in 2009 and is supported by empirical evidence that musicians outperformed non-musicians in SIN tasks—with years of consistent musical practice being positively associated with better SIN, working memory, and pitch perception [54]. More recently, a systematic review [55], and meta-analyses [56, 57], provided further support for this claim. Musicians routinely outperformed non-musicians on various SIN tasks, with more pronounced benefits across lower signal-to-noise ratios (SNRs), as well as across different distractor types and target properties, such as sentences, single words, phonemes, and tones. Other studies have shown stronger positive effects of musicianship on SIN performance across more challenging listening situations, suggesting greater benefits when the auditory system is under greater strain. One possible mechanism involves the type of training musicians receive (i.e., focusing on specific sounds in compositions that contain multiple simultaneous sounds), which may help them to better segregate auditory information in noisy environments. Neuroimaging studies have also revealed better auditory discrimination for musicians at the neural level (e.g. stronger spectral amplitudes in the Frequency-Following Response (FFR), more precise timing of transient response peaks, and improved Event-Related Potentials (ERP) waves such as P2 and N400). Studies have also shown a musician pattern of neural recruitment via shared pathways (both ventral and dorsal auditory streams in both hemispheres). The musician advantage in SIN may also be influenced by cognitive factors such as auditory working memory and selective attention, as well as age-related differences, with older musicians showing a greater SIN advantage compared to non-musicians of the same age. Additionally, an early age of onset and more extensive duration of musical training is correlated with stronger neural encoding and better performance on SIN tasks. Overall, these studies revealed a complex interaction between lower-level auditory processing and higher-level cognitive functions, indicating that musical training may enhance SIN perception through neuroplasticity and improved auditory processing. Despite these positive findings, some studies show variability in results, with either no significant advantage observed, or an advantage only under specific experimental conditions. A recent review raises questions regarding the causality of music training and its transfer to non-musical benefits, such as SIN perception, are often weak or nonexistent [58].

Hence, the role of music training for individuals with hearing loss has been explored with cautious optimism, with the possibility that music training may enhance SIN perception as well as psychosocial outcomes. However, a systematic review of studies that tested the hypothesis that music training can enhance speech understanding in people with hearing loss showed that most of the studies had flaws in their methodologies and design such as no (or an inappropriate) control group, lack of randomisation, failure to account for multiple comparisons, or inadequate limiting of participant or tester bias [59]. Additionally, the four studies with appropriate designs did not show any benefit to speech understanding [60–63].

The strongest evidence that group singing enhances SIN for older adults with hearing loss, was a study involving a choir-singing group and an age- and audiometrically-matched control group, which did not participate in any activities, over a 10-week period [64]. Both groups were carefully matched based on previous musical experience, quantified by the years of formal musical training, thereby enabling causal inferences that are typically unachievable in correlational or cross-sectional studies of musicians’ auditory capabilities. The choir participants exhibited significant improvements in SIN perception, pitch discrimination, and the robustness of the neural encoding of speech stimuli (frequency following response, FFR). The enhancement in SIN perception among choir participants was mediated by improved pitch discrimination, which was, in turn, predicted by the strength of the FFR. Although these results support the hypothesis that short-term participation in choir singing serves as an effective intervention for mitigating ARHL, reasonable concerns about participant bias may be raised due to the use of a self-selected choir group and a do-nothing control group.

Several systematic reviews have been conducted exploring the psychosocial benefits of singing (which is predominantly focused on group singing) in a range of populations and health conditions [65–68]. Overall, these findings indicate that group singing is an enriching activity that shows promise, but more research needs to be conducted to understand the therapeutic benefits for psychosocial wellbeing. Group singing was associated with better health-related quality of life, reduced mental distress, reduced symptoms of anxiety and depression, enhanced mood and emotional states, self-efficacy, a sense of purpose and achievement, social connectedness, and a sense of community and belonging. Physiological benefits have also been shown, with coordinated group singing leading to enhanced synchronization of respiration and heart rate variability (HRV) [69, 70].

There are several frameworks that suggest how and why music provides psychosocial benefits. Firstly, the musical and social bonding hypothesis posits that music-making is a coevolved system for social bonding [71]. The musical and social bonding hypothesis suggests that music can be likened to a toolkit, consisting of the following features: rhythm and dance (which supports synchronization and coordination); melody, harmony, and vocal learning (which supports unison and harmony); repetitive structure (which supports prediction); and music and social identity (which supports group identity). Importantly, these are all features common to group singing—or choir singing—which is a popular form of active music participation in the Western world. Notably, the musical and social bonding hypothesis argues that music making is more effective for large-scale bonding than other socially-facilitated contexts such as grooming or language [71].

Music has also been conceptualized in relation to its therapeutic benefits, consolidated into seven capacities in the Therapeutic Music Capacities Model [72]: 1) Music is engaging, utilizing a wide range of cognitive functions such as attention and working memory. There is evidence that long-term music training induces neuroplastic changes [73, 74]. Furthermore, while still inconclusive, there is some evidence that suggests music may enhance cognitive reserve [75], and limit age-related cognitive decline [76, 77]; 2) Music is emotional, and can induce a wide variety of emotional states, and heighten arousal and reward [78]; 3) Music is physical, with movement and dance a typical component to music engagement. Significantly, physical activity (such as dance) is a known contributor for supporting healthy ageing and cognition [79, 80]; 4) Music permits synchronization. The ability of individuals to synchronize to an external rhythm (such as a musical beat) is a fairly unique feature of musical activity. Group synchronization is associated with cooperation, group cohesion, collective identity, enhanced non-verbal communication, and better learning outcomes [81–84]; 5) Music is personal, allowing for personal connection, such as the reinforcement or re-evaluation of self. This is particularly important in the context of ARHL, as the construction and negotiation of identity and self may face pressures due to age and hearing loss stigma [85, 86]; 6) Music is social, and helps create bonds and connections with greater efficacy compared to other group-based interventions, such as crafting [87]; and 7) Music is persuasive. The authors of the Therapeutic Music Capacities Model argue that music holds a unique cultural position in cultural contexts such as religion, advertising, and health, that may reinforce and support positive treatment outcomes. Thus, music has the potential to amplify therapeutic benefit, or enhance adherence to a program.

To summarise, in the context of a rapidly ageing population, untreated hearing loss has significant consequences that affect communication and psychosocial wellbeing at the individual and community level. On the other hand, there is promising but limited evidence that group singing may support both communication and psychosocial outcomes. However, numerous systematic reviews have consistently highlighted limitations that limit generalizability as well as the ability to formally recommend group singing as an intervention with robust psychosocial benefits [65–68]. These include heterogenous populations; small sample sizes; single site studies lacking control groups (or appropriate control groups); a lack of theoretical frameworks; a lack of large and well-designed experimental randomized controlled trials (RCTs); and a level of bias that was typically moderate to high [65–68]. Hence, to address these concerns, this SingWell Project [88] study utilizes the theoretical frameworks of the musical and social bonding hypothesis, Therapeutic Music Capacities Model, OPERA, and PRISM that will guide hypothesis testing; adopted a multisite design to recruit a sufficiently large sample size based on a statistical power analysis; will limit of risk of bias through our recruitment strategy and experimenter blinding; and will use an RCT design with planned statistical analyses.

Aims and hypotheses

The aim of this study is to explore if group singing is beneficial for older adults with unaddressed hearing loss, hypothesizing that group singing may be effective at improving speech perception and psychosocial outcomes. At the macro-level (i.e., across a 12-week period), the outcomes of interest include: 1) speech-in-noise perception, and 2) psychosocial health. At the micro-level (i.e., pre- and post-session effects at week 2, 7, and 11), the outcomes of interest are 1) psychosocial health, and 2) heart rate variability. Note that additional detail for macro and micro-level tests, and the active control group is described in the Materials and Methods section.

Macro-level hypotheses

1. H1. Group singing may be associated with better music perception such as pitch, rhythm, timbre, and higher-level music skills—more so than participants in the control group.
2. H2. Group singing may be associated with better speech perception such as speech-in-noise and emotional prosody—more so than participants in the control group.
   1. The mechanism for speech perception enhancement may be due to improvements in music perception (i.e., pitch, rhythm, timbre, or higher-level music skills.)
3. H3. Group singing may be associated with better psychosocial health—more so than participants in the control group.
   1. The primary outcome of interest is the enhancement of quality of life and social connectedness.
   2. The secondary outcomes of interest are the reduced symptoms of depression and anxiety, and the enhancement of self-esteem.

Micro-level hypotheses

1. H1. Group singing may lead to increases in positive mood—more so than the participants in the control group.
2. H2. Group singing may foster greater feelings of social closeness—more so than the participants in the control group.

Ethics

Ethical approval to conduct this study was obtained from the Toronto Metropolitan Research Ethics Board (ID: REB 2024–103). A copy of the ethics protocol is available in S1 Appendix. All participants will be required to provide informed written consent.

Study design

A schematic of the study design can be seen in Fig 1. The study is a longitudinal RCT. Initially, participants will undergo an eligibility screening process and complete an information and consent form (described in detail under ‘Testing Procedure’). Participants will be randomly assigned to either group singing or audiobook club (control group) intervention for a training period of 12-weeks and will be unaware if they are in the experimental or control group (i.e., participant-blinded). After this, the study has multiple timepoints for testing, that are broadly categorized as macro, or micro timepoints.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Overview of the study design.

https://doi.org/10.1371/journal.pone.0314473.g001

The rationale for the use of an audiobook club as a suitable control group is due to several factors. Firstly, listening to audiobooks is a common clinical recommendation for hearing loss and have been used in other RCTs as a control [89, 90]. Secondly, discussion from the audiobook club will generate asynchronous speech, which is a suitable compliment to investigate the targeted benefits of synchronized group singing. Finally, we anticipate that the credibility and expectation between both interventions will be similar given they are both group activities in the auditory domain.

This study was registered as a clinical trial (clinicaltrials.gov, U.S. National Library of Medicine, ID: NCT06580847); and adheres to a Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) Checklist with 2022 extension [91–93], which is provided in S2 Appendix. Any modifications to the protocol will require approval from the Toronto Metropolitan Research Ethics Board and will be noted in the subsequent registered report article.

Participants

Participants will be recruited from seven sites around the world: Toronto, Montreal, and St John’s (Canada); Los Angeles (United States of America); Groningen (Netherlands); Oldenburg (Germany); and Adelaide (Australia). Each site will aim to recruit a total of 30 participants with a simple randomisation using a 1:1 computer generated allocation (15 randomly assigned to group singing, and 15 randomly assigned to the audiobook club), for a total of 210 participants among all the sites. Participants will not be able to change their allocation.

Recruitment will be variable as each site will have access to different databases, networks, and communities. Hence, recruitment flyers will be sent to a range of organisations such as: deaf and hard-of-hearing organisations, aged-care homes/retirement centres, senior centres, veterans organisations; and to databases such as the Toronto Metropolitan University Auditory Participant Pool and the Toronto Metro Senior Participant Pool. To avoid potential participant bias, flyers will advertise that the study is looking for participants to engage in a creative group activity, rather than group singing or an audiobook club, specifically.

Test battery

Test procedure

Below, we list the three facets to the test procedure, noting that some occur over several sessions. For consistency and logistical pragmatics, we will aim to test all participants within two-week windows, relative to key training timepoints.

1. Consent and Screening—initial session.
2. Macro Timepoint Testing—two sessions in a test booth:
   1. Baseline (up to 2-weeks before training begins).
   2. Completion (up to 2-weeks after training ends).
3. Micro Timepoint Testing—six sessions at the site of training:
   1. Pre- and Post-session testing at Week 2.
   2. Pre- and Post-session testing at Week 7.
   3. Pre- and Post-session testing at Week 11.

Training procedure

Participants will be randomly assigned to either the group singing or audiobook club intervention. Training will occur once-a-week over a total duration of 12-weeks, with each training session lasting approximately 1.5-hours, including informal networking and discussion. Adherence to the intervention will be reported.

Group singing will be facilitated by a choir master with at least one-year of experience leading a choir, a piano accompanist, and support from up to two research assistants who will sing with the participants. The broad goal of the group singing is to provide synchronous vocal pitch training in a supportive and casual environment. The beginning of each session will involve approximately 15-minutes of warm-up exercises, and the repertoire will be decided through discussion between the choir master and participants. Given the choir master will have some flexibility and need to adapt the sessions depending on group ability, a full outline of each lesson will be provided as supplementary materials in the final report. Participants will also be assigned an individual online ‘homework’ task each week—one-hour of using pitch-based games in Theta Music Trainer (https://trainer.thetamusic.com), an application designed for ear training and music theory [118]. The goal of the homework task is to maximize pitch-based perception. Theta Music Trainer provides logging capability that will allow us to monitor participants’ weekly homework tasks. Email reminders will be sent to encourage participant adherence.

Participants in the audiobook club will listen to a 1-hour length audiobook segment prior to each audiobook club session. Participants will be provided with a pool of 10 audiobooks with a brief description on genre and plot. Each audiobook club will collectively select which audiobook/s they would like to listen to during their 12-week intervention. Given they are expected to listen to ~1-hour per week; the audiobook/s must total approximately 12-hours in total duration. The pool has been selected to cover a wide range of genres and interests, and are available in English, German, and Dutch. These sessions will feature a facilitator that will constructively guide the participants through a series of open-ended questions developed by the research team that has been designed to elicit open and constructive discussion. Up to two research assistants will also be present and contribute to the discussions. Email reminders will be sent to encourage participant adherence to the weekly audiobook segment.

Dissemination plan

The main findings will be published in the subsequent registered report article. Given the multisite nature of this study, individual sites may choose to develop their own specific hypotheses that extend beyond this protocol, and may choose to develop additional outputs beyond the associated registered report article. The SingWell Project is also committed to knowledge mobilization activities involving key stakeholders such as choirs, physicians, hearing health professionals, and community centres. Full details can be found on their website (https://www.singwell.ca/) and from the SingWell Project protocol [88].

Statistical power analysis

An a priori power analysis was conducted to determine the sample size needed to detect the smallest effect size of interest. We determined the number of participants needed to detect the cross-level interaction of group (i.e., group singing vs. audiobook club) and time (i.e., baseline vs. completion) on SIN perception. The power analysis was conducted using traditional approaches where the resulting sample size refers to the level 2 clusters–in this case, the number of participants.

G*Power version 3.1.9.4 was used to conduct the analysis for an ANOVA within-between factor interaction [119]. For a conservative estimate, we assumed a small effect size of, η² = .01, corresponding to an effect size of f = 0.1. We estimated the correlation among the repeated measures to be r = .62, based on a previous group-singing controlled trial that reported an intraclass correlation (ICC) of .62 [64]. Accounting for a 20% attrition rate [64], the power analysis (f = .10, α = .05, r = .62) suggests a total sample size of N = 206 participants, approximately n = 30 per site for 85% power.

In addition to the power analysis, we also calculated the reliability of our repeated measurements to assess for sufficient reliability for each participant based on [120] (1): (1) where n min is the number of time points and ρ is the population correlation across all the timepoints. Given our two timepoints and estimated ICC of .62, we have adequate-to-good reliability of our timepoints, ρ = .77.

Planned statistical analyses