According to the World Health Organization, approximately 55 million people worldwide are living with dementia, with over 10 million new diagnoses each year [1]. The rapid increase in older adults is leading to a significant rise in the number of older individuals diagnosed with degenerative diseases, such as cognitive impairments [2,3]. This trend is causing major health concerns [4] and placing a greater economic burden on many countries tasked with their care.

Since age-related hearing loss has a significantly stronger correlation with reduced cognitive function due to dementia than other related factors in older adults, many recent studies have concentrated on the adverse interaction between these two conditions [5-7]. These studies also underscore the growing importance of addressing residual hearing ability in older adults proactively and as early as possible. Furthermore, clinicians have proposed that aural rehabilitation is necessary not only for older adults with hearing loss, but also for all patients with impaired cognitive functions, in the hope of potentially mitigating their accelerated hearing and cognitive issues [8,9]. Despite this recognized need, a specific protocol for developing and implementing an aural rehabilitation program for these populations has not yet been established for straightforward application in a clinical setting.

Several previous studies, however, have evaluated the performance of patients with cognitive decline using tests such as digit span, word recall, and story retelling, following the application of music-based auditory training [10-17]. For instance, Bottino et al. [10] found that patients with mild Alzheimer’s disease significantly increased the length of their backwards digit span after undergoing training. Similarly, Chu et al. [13] reported higher scores in the word recall test among dementia patients who received aural rehabilitation. Studies that utilized story retelling also demonstrated superior performance in the training group compared to the control group [12,14]. However, the research conducted by Ceccato et al. [12] and Giovagnoli et al. [15] did not reveal any significant changes in the digit span test results before and after rehabilitation. This lack of training benefit was also corroborated by Brotons and Kogor [11] and a study by Lyu et al. [16].

Although it is true that neuroplasticity resulting from auditory training can have a positive impact on cognitive function [17,18], previous studies have been hindered by certain limitations. First, there was an absence of standardized criteria for evaluating auditory and cognitive functions, and for identifying the effects of such training on older adults. Second, only a few studies included a control group without any intervention, leading to confusion about the existence of any rehabilitation effects [12-15]. Third, achieving comparable or consistent results across these studies was challenging due to the varied levels of cognitive function among participants [19,20]. Despite these limitations, an interesting observation is that the rehabilitation effect was greater when cognitive function impairment was less severe [10-16]. Additionally, active interventions, such as singing or playing an instrument, were found to yield greater improvements than passive methods such as simply listening to music [19,20].

Using music as stimuli in auditory training can stimulate a broader range of sensory cells than traditional methods, which typically use only a simple sound tone and/or speech utterances within their frequency range [21]. This results in more extensive brain activation [22]. Patel introduced the Overlap, Precision, Emotion, Repetition, Attention (OPERA) hypothesis [23], arguing that speech and music share networks. Therefore, incorporating music into training could potentially enhance speech processing [24]. Such an approach would necessitate trainees to utilize both analytical and comprehensive processing abilities for pitch, rhythm, harmony, and timbre [25]. Corrigall and Trainor [26] further suggested that an effective program should incorporate various task levels to maximize training effectiveness. However, previous studies have not successfully adapted these specialized rehabilitation methods for older adults with hearing loss and cognitive impairment, nor have they applied these diverse methods across various levels of training. Consequently, this study aims to examine the impact of music-based auditory training on older adults with hearing loss and mild cognitive impairment (MCI), which are both prevalent conditions in the older population. Simultaneously, we aimed to develop a training program that these patients can easily access and use effectively for rehabilitation.

The results of the pre-, post-, and FU tests are displayed in Fig. 2, along with a comparison of the mean and standard deviation between the ATG and CG groups. The statistical outcomes in terms of the main effects for each group, the testing times, and their interactions are then summarized in Table 2.

Fig. 2 shows that the ATG group outperformed the CG group in vowel identification. Following the training, the ATG group’s vowel identification significantly improved by approximately 9% (Fig. 2A). This improvement was also significantly sustained in the FU test conducted 2 weeks later (P<0.05). In contrast to vowel identification, no significant difference was observed between the two groups in terms of consonant identification (F(1,28)=4.378, P=0.051) (Fig. 2B). Despite the approximately 18% difference in scores between the two groups, the standard deviation for the CG group was notably high at 19% (Figs. 2 and 3). While no significant differences were detected between the groups, an improvement in scores over time was noted in the ATG group. In the FU test, the ATG group also maintained a significantly high score of 87.6% (P<0.05).

In the sentence recognition tests, both the completed sentences and target words demonstrated statistically significant differences between the groups and across the testing times, as detailed in Table 2. For the completed sentence test, the pre-test results for the ATG and CG were comparable, at 77% and 78% respectively (Fig. 2C). Following the training, the ATG’s performance improved to 95%, and impressively, it remained at 94% even after a 2-week period (P<0.05). Conversely, the target task required the identification of specific words within the sentence, rather than the correction of the entire sentence. As such, both groups achieved high scores of 90% or more, even in the pretest (Fig. 2D). After the training, the ATG demonstrated a significant improvement of 8% and maintained similar scores after a 2-week period (P<0.05).

In the sentence memory (Fig. 2E) and sentence sequence tests (Fig. 2F), a statistically significant difference was observed between the groups and across the testing times (Table 2). For instance, the sentence memory score of the ATG group significantly increased from 77.78% to 96.51% following the training, maintaining an improved score of 95.87% in the FU test (P<0.05). In contrast, the scores of the CG group remained approximately 78% throughout the testing times.

Interestingly, the score for the sentence sequence saw a significant increase of 35% or more following the training in ATG. This score was maintained, demonstrating a substantial difference with a P-value of less than 0.001 between the groups and testing time (Table 2). Fig. 2G and H present the results of the connected speech test and matrix sentence test at 6 dB SNR, respectively. Both tests exhibited an increase in scores by 18% and 16%, respectively, in the post-test compared to the pre-test results for ATG. These improved scores were still significantly maintained after a period of 2 weeks (P<0.05).

In Fig. 3, line graphs are presented for five cognitive tests and one questionnaire, comparing groups over various testing times. In Fig. 3A, a significant difference is evident in the MMSE test results between the groups (F(1,18)=19.995, P<0.001) and across the testing times (F(2,36)=27.036, P<0.001). Although the ATG’s improvement score was approximately 3–4 points higher than the CG’s, this difference was not as substantial as the results of other tests. Nevertheless, the improvement score was significant (P<0.05). In comparison to the CG, where the average across three testing times was five recall words, the ATG demonstrated a significant difference in the immediate recall on the K-CVLT test. This shifted from approximately 4.6 words in the pre-test to 7.2 words in the post-test (Fig. 3B). This result was sustained at 6.9 words in the FU test (P<0.05). For the delayed recall test, the ATG exhibited superior memory performance, recalling about four more words after the training (Fig. 3C). This demonstrated a statistically significant difference between the groups (F(1,18)=33.657, P<0.001) and across the testing times (F(2,36)=35.561, P<0.001).

In contrast, for the digit span, there was a statistically significant difference in the forward digit between the ATG and CG (F(1,18)=6.607, P=0.019), but no difference was observed across the three different testing times (F(2,36)=1.785, P=0.182). In other words, although the ATG showed an improvement of 1 digit length with the training, this finding was complicated by the fact that the CG had a mean and standard deviation of 5 digits and 1.25, respectively (Fig. 3D). Moreover, the backward digits did not demonstrate any significant difference between the groups (F(1,18)=3.101, P=0.095) or across the testing times (F(2,36)=3.219, P=0.052). Both groups started with 3.4 digits on the pretest, but only the ATG showed a slight increase to 4.2 after the training (Fig. 3E), which did not constitute a statistically significant change.

Finally, the results of the WHOQOL test did not show a significant difference between the groups (Fig. 3F). Upon applying a Bonferroni correction to account for the difference in testing time, a significant difference of approximately 4 points was observed between the pre-test and post-test results (P=0.024). However, this significance was not maintained in the FU test (P>0.05).

The present study investigated the effect of music-based auditory training on older adults (over 65 years) who had hearing loss and MCI. The ATG group, which consistently received the 40-minute training for 8 weeks, showed a meaningful change in overall auditory function including phoneme ability and word-level identification, sentence recognition ability, and cognitive function in sentence-level working/short-term memory and word-level long-term memory. When composed and delivered systematically, this training can provide benefits for auditory and cognitive function. This implies the need to clinically emphasize the importance of music-based auditory training in the future for older adults who have hearing loss and reduced cognitive function.

Auditory training is primarily based on the concept of auditory neuroplasticity [17]. Among the various methods, auditory training that incorporates music is recognized for its ability to stimulate a broader range of cortical and subcortical areas compared to traditional auditory training. This approach not only improves verbal and auditory skills, but also improves emotional sensitivity and recognition [34]. Previous research utilizing a neuroscientific approach to study the effects of music training on auditory treatment in older adults has confirmed a reduction in neural timing delays through the processing of rapidly changing auditory information [35]. These studies also report a positive enhancement in the perception of word sounds, directionality, discrimination, and even cognitive ability among hearing-impaired subjects [25,36].

In the present study, we conducted eight different tests to assess changes in auditory function. After 8 weeks of training, we observed an improvement of approximately 18%–20% (Fig. 2). Despite the statistically significant differences, vowel identification (Fig. 2A) and target words (Fig. 2D), which initially had high baseline scores, demonstrated an increase of less than 10%. Most notably, the assessment that exhibited the most significant training effect revealed a substantial difference of about 38% in the post-test between the two groups in the sentence sequence. This outcome is supported by Patel’s OPERA hypothesis, which posits that musical training enhances the neural encoding of speech due to improved attention through music activity [23,24]. In other words, it could be possible to compare the auditory encoding demands imposed by musical training and speech perception [23].

In this study, we trained participants in instrumental play and auditory attention using word-level referents, thereby minimizing the use of contextual clues. This type of training elicited auditory feedback through the activity of singing along while simultaneously performing tasks. Among the various tasks, some were designed to utilize the subjects’ substantial cognitive resources by establishing immersive listening for interference. The comprehension stage inherently involves auditory ability, which is a component of cognitive function [37]. Consequently, the participants in this study demonstrated a general improvement in both auditory and cognitive functions after undergoing the training. It appears that the training also increased their overall comprehension ability.

However, although there was a statistically significant difference, the effect on vowel identification (Fig. 2A) and target words (Fig. 2D), which initially had high baseline scores, increased by less than 10%.

Compared to the immediate recall test results, the delayed recall test results demonstrated an increase of approximately four words in the ATG after 8 weeks of training. In other words, the training had a greater impact on tests of slightly lower difficulty. This pattern was also observed in digit tests. Previous studies investigating the relationship between music training and working memory have found that subjects who underwent music training did indeed improve their working memory [38,39]. Moreover, a study by Janata et al. [22] used functional magnetic resonance imaging to confirm the activation of the subjects’ cortical and subcortical areas. They observed activation in the area associated with working memory management, thereby demonstrating a direct relationship between music training and working memory [22].

Generally, older adults with hearing loss struggle to accept and process auditory information efficiently [40]. As a result, they often rely on inferences for various meanings or nonverbal communication to recover the lost auditory information [41]. This process consumes a significant portion of their cognitive resources, creating a larger cognitive load in their already complex auditory environment [42]. The cognitive retrieval of auditory information delays the reliance on sensory memory, leading to further loss of auditory information in a complex environment [42,43]. This was corroborated by the results of the backward digit test, which did not yield significant results in the current study. In other words, regardless of whether training was conducted or not, long-term memory, a higher-order system of the auditory system, had a detrimental effect [44].

The current study demonstrated that a diverse range of music and task execution positively influenced the subjects’ cortical activity and their working memory capabilities. Training that involved playing an instrument and focusing on its sound amidst complex musical situations, while simultaneously concentrating on the researcher’s voice embedded in the music, positively impacted the subjects’ working memory management. This finding aligns with previous research conducted by Janata et al. [22]. Moreover, the training course, which encompassed tasks from three different types of training, appears to have enhanced the participants’ efficient use of their working memory’s cognitive resources and improved their executive function. From the perspective of music rehabilitation combined with auditory and cognitive training, the overall cognitive function of the subjects in this study did indeed improve. This improvement likely boosted performance in their higher cognitive systems, while simultaneously reducing cognitive load.

Unfortunately, the music-based auditory rehabilitation did not conclusively yield an effect on the participants’ overall quality of life. The WHOQOL-BREF questionnaire, which was used to evaluate participants’ overall quality of life, contains five scales: general health, physical health, psychological, social relations, and environmental areas [33]. This questionnaire focuses on the subjects’ perceptions of their lives over the past 2 weeks. However, several of the specific questions could not feasibly be altered by an 8-week music-based auditory training. Examples include questions such as “How healthy is the residential environment you are living in?,” “How do you rate your current means of transportation?,” or “Do you have sufficient funds to meet your needs?” [33].

Meta-analysis studies examining the impact of music rehabilitation on older adults with dementia have primarily focused on psychological factors, such as anxiety or depression. However, this study, which uniquely utilized a variety of testing instruments to assess diverse scales, yielded a different outcome. Furthermore, given the high correlation between depression and cognitive impairment in older adults, some studies have included subjects without diagnosed depression, as well as subjects with only MCI [45]. In our study, we required self-reporting on the presence or absence of depression during the screening test, and none of the subjects reported depression. Instead, this meta-analysis focused on older adults with dementia [46,47], which inevitably limited the scope of psychological changes resulting from rehabilitation when compared to subjects with only MCI. The fact that the WHOQOL-BREF questionnaire, which assess the overall quality of life, measured diverse scales—in addition to the limited changes in psychological factors—may explain the insignificant interaction effect between these groups. Therefore, it is believed that the results of this study were influenced not only by the effect of music-based auditory training, but also by differences arising from various external and internal variables [48].

The present study has several limitations. First, although the number of participants included in each group exceeded the threshold for statistical verification in G*Power, the sample size of only 10 individuals is too small to be generalized to an entire population. Secondly, the training period of 8 weeks may not have been sufficient to fully understand the impact of long-term rehabilitation performance. For instance, in the case of digit span tests that did not yield a statistically significant difference, it is plausible that a longer training period could reveal a difference. To discourage participants from withdrawing from the study, the period from rehabilitation to FU was kept to a brief 2 weeks. Fortunately, most of the test results indicated that performance was significantly maintained. However, it will be necessary to confirm whether performance can be sustained in some of the various tests evaluated in this study through future research. Lastly, the music-based rehabilitation used in this study appears to be quite broad. Given that our rehabilitation approach incorporated a variety of activities, it is challenging to identify the specific impact of each component on auditory and cognitive functions. Future studies could investigate the effects of these individual factors, identify which ones were ineffective, or determine whether significant effects were observed when these factors were combined. Such an investigation could offer deeper insights into the underlying mechanisms of music rehabilitation.

The implementation of an 8-week auditory training program, based on engaging and easy-to-follow music, was conducted for 10 older adults with MCI and hearing loss. It was confirmed that this approach had a significant impact on hearing and cognitive function when compared to the control group. Future studies are anticipated to further refine and standardize this music-based method, with the expectation that it will become widely utilized for auditory training in elderly patients within a clinical setting.