Effects of Auditory Training on Cognition in Hearing Loss: A Systematic Review and Meta-analyses

Abstract

Purpose:

The aim of this study was to examine the efficacy of auditory training to improve cognitive function in patients with age-related hearing loss (ARHL).

Research Design:

This is a systematic review and meta-analysis.

Study Sample:

Seven studies involving 443 participants met the inclusion criteria. Participants were typically older adults (mean age = 67.23 years, standard deviation = 7.14) with mild to severe hearing loss.

Intervention:

Auditory training includes speech perception training, phoneme discrimination training, and so on.

Data Collection and Analysis:

A literature search of academic databases (Cochrane Library, PubMed, Web of Science, Embase, Wanfang, Weipu, and China National Knowledge Infrastructure) identified relevant articles published up to December 2023. This review includes only randomized controlled trials. The primary outcome is cognition function, measured by Montreal Cognitive Assessment, Mini-Mental State Examination, and other cognition-related subtest indicators.

Results:

The overall effect of auditory training on overall cognition and executive function in ARHL is statistically significant (overall cognition: g = 0.79, 95 percent confidence interval [CI]: 0.57, 1.01; executive function: g = 3.84, 95 percent CI: 1.49, 6.19), but executive function domain has high heterogeneity (I² = 100 percent). The effect of auditory training on attention/processing speed and working memory is small and not significant (attention/processing speed: g = 1.47, 95 percent CI: −0.48, 3.42; working memory: g = 0.68, 95 percent CI: −2.22, 3.58), but both attention/processing speed (I² = 96 percent) and working memory domain (I² = 98 percent) have high heterogeneity.

Conclusions:

The overall impact of auditory training on overall cognition and executive function seems to be significant, but because of the low quality of the literature and certain biases, it is impossible to conclude that auditory training can improve the cognitive function of ARHL; therefore, more high-quality evidence is needed.

With aging, most older adults have age-related hearing loss (ARHL), which adversely affects both the audibility and clarity of speech, particularly in noise (Gates and Mills, 2005; Anderson et al, 2013b). With the growing trend of population aging, ARHL has become a global health problem. According to the World Health Organization (WHO), approximately 320 million older adults worldwide suffer from ARHL. Because of difficulties in listening, older adults are unwilling to communicate (Rönnberg et al, 2013), resulting in limited social interaction and triggering a sense of loneliness and frustration, which has a negative impact on the psychology of the elderly (Mick et al, 2014). More importantly, ARHL is strongly associated with cognitive decline, and hearing loss (HL) is considered the greatest modifiable risk factor for dementia (Livingston et al, 2017). During the follow-up period of up to 10 years, the risk of dementia in patients with mild, moderate, and severe HL increased by two, three, and five times, respectively. Every 10 dB of HL will increase the risk of dementia by about 1.27 times, and HL is related to an accelerated rate of cognitive decline (Lin, 2011). There have been several mechanism hypotheses to explain the strong link between ARHL and cognitive dysfunction, including the information degradation hypothesis, which suggests that because ARHL results in a reduced ability to encode and use speech signals, it requires the mobilization of more cognitive resources for processing and dealing with them, leaving insufficient cognitive resources for other tasks, which leads to overall cognitive decline (Humes and Young, 2016).

There have been a few interventions to alleviate cognitive impairment caused by ARHL (Livingston et al, 2020), one of which is auditory training. Auditory training can be defined as a process that involves actively engaging with sounds to teach the brain how to listen. During such training, individuals learn to make distinctions between different sounds presented systematically. Perceptual training focuses on enhancing listening and communication skills by refining the sensory perception of auditory stimuli (Fu et al, 2004; Stecker et al, 2006), such as tones, phonemes, and words. Although the literature primarily emphasizes the bottom-up approach to training, which involves actively listening to auditory stimuli, it also acknowledges the implicit role of top-down cognitive processes in auditory training and subsequent learning. This is evident in studies demonstrating learning in the auditory domain as a result of training on nonauditory tasks, such as visual discrimination or visuospatial tasks, using identical stimuli (Amitay et al, 2006). These findings suggest that top-down cognitive processes facilitate learning. Thus, auditory training may offer a potential method for improving both auditory and cognitive processes in adults with HL (Pichora-Fuller and Levitt, 2012).

Previous systematic reviews have focused on whether auditory training can improve speech recognition in ARHL instead of cognitive functions. Evidence suggests that phonemic training can enhance the identification of targeted phonemes, while training with nonsense words can improve sentence recognition in noise for recipients of cochlear implant (Cambridge et al, 2022).

There is still debate about whether auditory training can improve cognitive function in ARHL. It has been found that auditory training may improve cognitive functioning in older adults with ARHL (Ferguson et al, 2014), and one study concluded that evidence from five randomized controlled trial (RCT) studies found that auditory training improved patients’ working memory and overall cognition (Lawrence et al, 2018). A total of 13 studies are included in the systematic review, but only 1 study is targeted at cognitive improvement. Although this one study had shown an enhancement in cognitive performance, the improvement is small and not robust (Henshaw and Ferguson, 2013). Therefore, more secondary research evidence is needed.

The first aim of this systematic review and meta-analysis is to determine whether auditory training can improve cognitive functioning in ARHL and, if so, which cognitive domains can be improved. The secondary aim is to determine whether participants adhered to the intervention with limited supervision.

METHODS

Search Strategy

This review was done according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. Potential eligible articles were identified by consulting seven databases: Cochrane Library, PubMed, Web of Science, Embase, Wanfang, Weipu, and China National Knowledge Infrastructure. The search time was from the establishment of the database to December 2023. The following search terms are used: “auditory training,” “hearing disorder,” “hearing loss,” “hearing impairment,” “working memory,” “processing speed,” “executive function,” “Presbycusis” [MeSH], and “Cognition” [MeSH] (see Supplemental Table S1). We retrieved related synonyms in MeSH according to the subject words and merged them with the subject words so that the retrieval scope can be expanded and the retrieval rate can be improved. In addition, we performed hand searching by looking at relevant studies cited in other studies.

Study Selection

Studies that met the following criteria were included in the study:

• 1.Population: Participants (mean age [M age] ≥ 60 years old) had HL, using the WHO definition of HL (individuals with hearing thresholds no greater than 20 dB [reference group]; mild HL, 20–34 dB; moderate HL, 35–49 dB; moderately severe HL, 50–64 dB; severe HL, 65–79 dB; profound HL, 80–94 dB; and complete or total HL, ≥95 dB) or a clinical definition of HL. The gender, race, and nationality of the subjects were not limited.
• 2.Intervention: Auditory training included speech perception training, phoneme discrimination training, and so on.
• 3.Comparator: The comparative training was sham training or no intervention.
• 4.Outcomes: Primary outcomes were cognition function, measured by Montreal Cognitive Assessment (MoCA), Mini-Mental State Examination (MMSE), and other cognition-related subtest indicators (e.g., working memory-spatial working memory [Guidetti et al, 2020], executive function-Stroop Color and Word test [Scarpina and Tagini, 2017], attention/processing speed-visual matching subtest [Jones et al, 2008]).
• 5.Study type: Only RCTs are included.

Trials that met any of the following criteria were excluded:

• 1.expert opinion, case report, unpublished meeting summary, review paper, and conference abstract;
• 2.lack of required outcome measures or failure to report data required for meta-analysis (such as mean, standard deviation [SD], etc.); or
• 3.only abstracts and unavailable full text through various channels.

Screening and Data Extraction

The titles and abstracts of identified articles were managed using a commercial citation manager (Endnote X6 Thomson, London, UK). Two authors (J.W. and H.R.) conducted a literature search independently, and the title, abstract, and key words of all studies were preliminarily screened with the same screening criteria. Then, they thoroughly read the initially included studies and eliminated the records that did not meet the requirements according to the exclusion criteria. Detailed information was extracted from each study, including first author, study design, HL and hearing device, sample size, M age, intervention, stimuli, training frequency and duration, cognitive outcome measures, compliance, and follow-up. When the data in the study were incomplete, there was an attempt to contact the author for the missing data. When the relevant data were not obtained, the study was excluded. Any disagreements were discussed and, if required, resolved by consultation with a third reviewer (Z.J.). All eligible studies were included in the end.

Quality of Evidence

For RCT studies, two investigators (H.Q. and C.G.) independently assessed the risk of bias (RoB) for each of the included studies using the Cochrane Risk of Bias 2.0 tool. Disagreements were discussed and resolved by consultation with a third investigator (J.W.) if necessary. Following the general principles outlined in chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions version 6.4 (updated August 2023), the following five domains were assessed: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported results. The RoB for each of the five domains and overall was described as low, with some concerns, or high. The overall RoB for the result is the least favorable assessment across the domains of bias (Higgins et al, 2024). Both the proposed domain-level and overall risk-of-bias judgments can be overridden by the review authors, with justification.

Statistical Analysis

Effect Sizes

Differences in mean and SD between the intervention and control groups were extracted for the effect size calculations after training, and if they were not provided, calculations were based on standard errors and standardized mean differences (SMDs). Considering that in calculating effect sizes, Hedge’s g considered sample size, it was chosen as the effect size analysis instead of Cohen’s d (Andrel et al, 2009). Because some studies had the same research objectives but the use of different tests for the outcomes could not be directly compared, SMD was used to transform these effect indicators before analysis. The domain-specific effect sizes and adjusted variances were then combined using a random-effects model, providing 95 percent confidence intervals (CIs; Bipat and Zwinderman, 2010). The analysis was performed using Revman software version 5.4.

Heterogeneity

Heterogeneity was assessed using the I² statistic, which estimates the percentage of variation across the studies that is due to heterogeneity. Heterogeneity levels were categorized as low (0–40 percent), medium (41–60 percent), and high (61–100 percent) (Moher et al, 2015). For pooled-effect estimates showing significant heterogeneity, sensitivity analyses were conducted to explore the impact of removing studies with anomalous characteristics, such as different intervention types or participant subgroups, to determine their influence on the pooled effects.

Quality Certainty

The certainty of quality was evaluated by two independent raters using the Grading of Recommendation Assessment, Development, and Evaluation (GRADE) (Balshem et al, 2011). Within GRADE, randomized trials are initially rated high, observational studies low, and other levels of evidence very low. However, high-quality evidence will be downgraded if methodological flaws exist, and low-quality evidence will be upgraded when high rigor and large effect sizes exist.

RESULTS

Search Results

The search of seven databases and other sources provided a total of 426 citations. After deduplication, 320 citations remained. After selection of the title and abstract based on the modified PICO (patient/population, intervention, comparison, and outcomes) criteria, 51 studies were withheld for full-text evaluation. Forty-three were excluded because they did not include standardized neuropsychological outcomes measuring cognition (19), auditory training (5), or participants with HL (8); were a review article (4); lacked statistics (7); or were of low quality (1). This led to a total of seven studies used for data extraction. Figure 1 shows the flow chart for the process of study selection and the main reasons for exclusion.

Figure 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) flow diagram of the study selection process. Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers). If automation tools were used, indicate how many records were excluded by a human and how many were excluded by automation tools. For more information, visit: http://www.prisma-statement.org/. (From Page MJ, McKenzie JE, Bossuyt PM, et al. [2021] The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71. 10.1136/bmj.n71.)

Study Characteristics

Participants

Seven studies involving 443 participants met the inclusion criteria (see Supplemental Table S2). Participants were typically older adults (M age = 67.23 years, SD = 7.14) with mild to severe HL. Intervention and control group sample sizes ranged from 15 to 76 participants (intervention group: M = 32.00, SD = 21.53; control group: M = 31.28, SD = 20.65). One study included cochlear implant recipients (Magits et al, 2023), three included hearing aid users (Lowe et al, 2023; Van Wilderode et al, 2023; Yanping et al, 2023), and three did not report this information (Anderson et al, 2013a, 2013c; Yusof et al, 2019).

Intervention

Two studies used personalized LUISTER training, which mainly used speech materials (Magits et al, 2023; Van Wilderode et al, 2023). Also, one study used conversations with selected communication partners for training with additional competing noise (Lowe et al, 2023). Other studies have integrated auditory-cognitive training panels (Anderson et al, 2013a, 2013c; Yusof et al, 2019; Yanping et al, 2023), including subtests such as sound source localization, tone tapping, etc.

Controls

All studies used sham training in the control group, four studies were conducted by watching DVD films (Anderson et al, 2013a, 2013c; Yusof et al, 2019; Van Wilderode et al, 2023), and the others changed some of the training settings (Lowe et al, 2023; Magits et al, 2023; Yanping et al, 2023), such as nonpersonalization, talking to a communication partner in a quiet environment, and so on.

Outcome Measures

Different studies used various neuropsychological tests to assess cognitive function. Refer to Table 1 for a concise overview of the cognitive outcome measures employed in the included studies.

Table 1.

Neuropsychological Tests Used to Measure Cognition Among Included Studies

Cognitive Domain	Neuropsychological Tests	Original References for Test Development	Studies Included in This Article
Executive function	Stroop Color-Word test	Golden CJ, Freshwater SM. (1978) Stroop Color and Word Test. Torrance, CA: Western Psychological Services.	Magits et al (2023) Van Wilderode et al (2023)
	Trail Making Test	Delis DC, Kaplan E, Kramer JH. (2007) Trail Making Test (TMT): D-KEFS. San Antonio, TX: Harcourt.	Magits et al (2023) Van Wilderode et al (2023)
	Letter Memory test	Friedman NP, Miyake A, Young SE, DeFries JC, Corley RP, Hewitt JK. (2008) Individual differences in executive functions are almost entirely genetic in origin. J Exp Psychol Gen 137:201–225.	Magits et al (2023) Van Wilderode et al (2023)
Attention/processing speed	Dual task of listening and memory	Howard CS, Munro KJ, Plack CJ. (2010) Listening effort at signal-to-noise ratios that are typical of the school classroom. Int J Audiol 49:928–932.	Lowe et al (2023)
	Integrated Visual and Auditory Continuous Performance Test	Woodcock RW, McGrew KS, Mather N. (2001) Woodcock-Johnson III Tests of Cognitive Abilities. Itasca, IL: Riverside Publishing.	Anderson et al (2013a, 2013c)
Overall cognition	Montreal Cognitive Assessment	Nasreddine ZS, Phillips NA, Bédirian V, et al (2005) The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53(4):695–699.	Yanping et al (2023) Yusof et al (2019)
	Mini-Mental State Examination	Folstein MF, Folstein SE, McHugh PR. (1975) “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12(3):189–198.	Yanping et al (2023)
Working memory	Dichotic Digits Test	Mukari SZ, Keith RW, Tharpe AM, Johnson CD. (2006) Development and standardization of single and double dichotic digit tests in the Malay language. Int J Audiol 45(6):344–352.	Yusof et al (2019)
	Numbers reversed	Woodcock RW, McGrew KS, Mather N. (2001) Woodcock-Johnson III Tests of Cognitive Abilities. Itasca, IL: Riverside Publishing.	Anderson et al (2013a)
	Memory for Words test		Anderson et al (2013c)

Study Type

All studies are RCTs.

Quality of Evidence

The RoBs in the eight included studies were assessed using the Cochrane Collaboration tool (see Figure 2). For the overall quality of each study, all have some concerns regarding RoB. All the studies were randomized but did not provide information on randomization methods and allocation concealment. This meta-analysis is a study on the impact of auditory training on cognitive function, in which adherence to intervention plays an important role. Therefore, all included studies adopted the per-protocol analysis method. One study (Anderson et al, 2013c) had a high noncompliance rate (>5 percent) and due to the large number of dropouts in the intervention groups and the lack of appropriate correction bias analysis, the outcome data were incomplete. Specifically, that study (Anderson et al, 2013c) did not explain the reasons for loss of follow-up. Five studies had low noncompliance rates, or all participants completed the trial without intervention deviation, so all or nearly all outcome data were collected. All methods for evaluating research results are appropriate, and there is no difference between intervention groups. Considering that the evaluation of cognitive function is mostly subjective outcome indicators, although the evaluation process is relatively rigorous, speculation may still be made during the data collection process. Two studies blinded outcome evaluators and generated all the specified outcomes in the protocol (Magits et al, 2023; Van Wilderode et al, 2023). The protocols for the other five studies were not available (Anderson et al, 2013a, 2013c; Yusof et al, 2019; Lowe et al, 2023; Yanping et al, 2023).

Figure 2.

Risk of bias summary. The judgments of the review authors regarding each “risk of bias” item for each of the included studies.

Adherence to Training

The adherence to auditory training was assessed by calculating the percentage of participants who completed the recommended intervention duration. Among the eight studies, five reported participants’ adherence. In the study by Lowe et al (2023), all participants in both the experimental and active-control groups showed a 100 percent adherence rate, completing the required number of training sessions. However, there was some variability in the time frame in which these sessions were completed. Approximately two-thirds of participants in both groups (61.1 percent in the experimental group and 66.6 percent in the active-control group) reported following the planned training schedule of five sessions per week for 4 weeks. The remaining participants completed all 20 sessions but over a longer duration. Only 1 participant of 42 dropped out of the study by Van Wilderode et al (2023) because he was not able to use the tablet training program proficiently. Yusof et al (2019) stated that 100 percent of the participants (n = 76) completed the full auditory training intervention, and Anderson et al (2013b) reported an 87.5 percent compliance rate. Three studies did not report adherence to training (Anderson et al, 2013a; Magits et al, 2023; Yanping et al, 2023), but one study (Magits et al, 2023) mentioned that one participant in the training group and three participants in the active-control group practiced less than the recommended duration.

Effect Size, Heterogeneity Outcomes, and Sensitivity Analyses

We categorized cognition into four domains: overall cognition, working memory, executive function, and attention/processing speed. The overall effect of auditory training on overall cognition and executive function on ARHL was statistically significant (overall cognition: g = 0.79, 95 percent CI: 0.57, 1.01; p < 0.001; executive function: g = 3.84, 95 percent CI: 1.49, 6.19; p = 0.001), but executive function domain with high heterogeneity (I² = 100 percent) (see Figures 3 and 4). The effect of auditory training on attention/processing speed and working memory was small and not significant (attention/processing speed: g = 1.47, 95 percent CI: −0.48, 3.42; p = 0.19; working memory: g = 0.68, 95 percent CI −2.22, 3.58; p = 0.64); both attention/processing speed (I² = 96 percent) and working memory domain (I² = 98 percent) are with high heterogeneity (see Figures 5 and 6).

Figure 3.

Forest plot comparing the overall cognition of age-related hearing loss with and without auditory training. (a) Mini-Mental State Examination. (b) Montreal Cognitive Assessment. CI = confidence interval.

Figure 4.

Forest plot comparing the executive function of age-related hearing loss with and without auditory training. (a) Stroop Color-Word Test. (b) Trail Making Test. (c) Letter Memory Test. CI = confidence interval.

Figure 5.

Forest plot comparing the attention/processing speed of age-related hearing loss with and without auditory training.

Figure 6.

Forest plot comparing the working memory of age-related hearing loss with and without auditory training.

Following the identification of the high I² score, the studies in outcome were reexamined for significant sources of heterogeneity. However, through running a sensitivity analysis, no potential source of heterogeneity was identified.

Certainty of Evidence

The GRADE standards indicated that the certainty of the evidence for improved overall cognition, executive function, attention/processing speed, and working memory after auditory training were “moderate,” “low,” “low,” and “low,” respectively (Table 2). “Moderate” means that the real effect is likely to be close to the effect estimate, but there is still a possibility they are quite different (Balshem et al, 2011). “Low,” in contrast, means that there is a limit to how much we can trust effect estimates; real effects may be very different from those estimated. The certainty of the evidence was most often downgraded because of the RoB, the heterogeneity within pooled effect estimates, and the small number of published studies available to include in this review.

Table 2.

GRADE Summary of Evidence for Auditory Training

Certainty Assessment							Effect
No. of Studies	Study Design	Risk of Bias	Inconsistency	Indirectness	Imprecision	Other Considerations	g [95 percent CI]	Certainty	Importance
Overall cognition
2	Randomized trials	Not serious	Not serious	Not serious	Serious*	None	0.79 [0.57, 1.01]	⨁⨁⨁◯ Moderate	Critical
Executive function
2	Randomized trials	Not serious	Serious†	Not serious	Serious*	None	3.84 [1.49, 6.19]	⨁⨁◯◯ Low	Important
Attention/processing speed
3	Randomized trials	Serious‡	Not serious	Not serious	Serious*	None	1.47 [−0.48, 3.42]	⨁⨁◯◯ Low	Critical
Working memory
3	Randomized trials	Not serious	Serious†	Not serious	Serious*	None	0.68 [−2.22, 3.58]	⨁⨁◯◯ Low	Important

^* Certainty of the evidence was downgraded by one level for imprecision because the CI around the absolute effects just included the threshold between a small benefit to trivial and no benefit.

^† Certainty of evidence was downgraded one level for inconsistency because there was a large degree of heterogeneity within the pooled effect.

^‡ Certainty of evidence was downgraded one level because a proportion of studies was graded as some concerns of bias across multiple criteria.

CI = confidence interval; GRADE = Grading of Recommendation Assessment, Development, and Evaluation; SMD = standardized mean difference.

DISCUSSION

This study summarized the existing evidence from RCTs on the effects of auditory training on cognitive function in ARHL. The main purpose of auditory training is to improve the patient’s auditory processing ability and speech perception capability to help the patient achieve better communication; however, recent evidence indicated potential enhancements in cognitive function as well (Yanping et al, 2023). In this study, we focused on whether auditory training improves cognitive function and which cognitive domains are most improved for the population. Understanding the extent to which auditory training can improve cognitive function can have important implications for designing effective intervention programs for individuals with ARHL, ultimately improving their overall communication abilities and quality of life.

Auditory training appears to improve overall cognitive function, based on results from comprehensive cognitive assessment scales such as MoCA and MMSE, which are in line with the results of a 2018 publication (Lawrence et al, 2018). This effect can be explained by the sensory deprivation hypothesis and neuroplasticity (Jafari et al, 2019). Individuals with ARHL often experience difficulties in speech perception because of a long-term lack of sound signals. Consequently, they need to engage more frontal lobe resources to process auditory information, resulting in a decline in cognitive functioning because of reduced cognitive resources available for complex cognitive processes (Bosmans et al, 2020). The hippocampus, a brain region crucial for cognitive function, undergoes neuropathological changes, including synaptic and cortical connection loss, leading to cognitive decline characterized by reduced synaptophysin content and number (Billig et al, 2022). Auditory training can stimulate neural cell activity, increase dendritic density, and establish new neural pathways. This process allows for enhanced reception of afferent impulses and central nervous system reorganization, ultimately improving cognitive function (Gómez-Soria et al, 2023).

When dividing cognition into different domains, the overall effect of auditory training on executive function in ARHL was statistically significant, but not on attention/processing speed or working memory. It is worth noting that a meta-analysis reported that auditory training can improve working memory in patients (Lawrence et al, 2018). Discrepancies in the analyses may be because of differences in the type and number of studies included. There are some reasons why auditory training does not work when smaller cognitive domains are involved: the training modules do not correlate with the cognitive assessment outcome indicators, making it difficult to transfer the results of the training; and the training period is too short to allow for the development of neuroplasticity. When comparing training durations, patients who underwent executive function training had longer durations than those trained in working memory and attention/processing speed, typically 12 or 16 weeks versus 8 weeks, potentially explaining the significant improvement in executive function results. In the future, for a precise comparison of different cognitive subdomains improvement in ARHL through auditory training, it is crucial to minimize other influences, such as ensuring patients undergo the same training program for the same training duration and using diverse outcome measures to draw more accurate conclusions.

The adherence of participants to auditory training with limited supervision is crucial in determining the effectiveness of a new treatment. Insufficient adherence to the training program is indeed considered a significant cause for the lack of improvement (Hnath Chisolm et al, 2013). When participants do not consistently adhere to the prescribed training regimen, it can hinder their ability to fully benefit from the program and impede their progress. More than half of the studies included in this review reported adherence, and the compliance was high (all >85 percent), but this may not be quite the same as the reality. In the study conducted by Sweetow and Sabes (2010), they found an adherence rate of 30 percent in their clinical population, which was significantly lower than the compliance rate of 73 percent reported in their earlier study (Sweetow and Sabes, 2006). This decrease in adherence is a concerning finding, so it is important to investigate what factors affect participants’ adherence and to explore potential strategies to increase participant engagement and compliance. Factors affecting participant adherence to auditory training include the engagement of communication partners (Lowe et al, 2023), difficulty of the training, participant proficiency with the device and their motivation (Sweetow and Sabes, 2010), etc. By improving adherence to the training program, the overall effectiveness of the treatment can be enhanced.

There are several limitations to this review. Only seven RCTs were found to meet the inclusion criteria, and several pooled effects within individual domains were supported by limited evidence from only one or two studies. In addition, heterogeneity was high on executive functions and working memory domains. It did not decrease after sensitivity analyses, suggesting that there is a high degree of variability across studies and that the results of the analyses may be questionable, but the limited number of included studies also influenced this. Although this review included only RCTs that were considered to have a high grade of original evidence (Glasziou, 2011), the evaluation of the included studies using the Cochrane Risk of Bias 2.0 tool revealed that none of the studies were at low RoB, and they all had some concerns regarding bias. This suggested that the quality of the included studies was not well controlled, and that future RCTs in this area need to be better designed.

CONCLUSIONS

This study focuses on the RCT, a type of study with a high level of original evidence, to investigate the effect of auditory training on the cognitive functions of ARHL. The findings from the seven included RCTs indicated that auditory training may improve overall cognition; this can probably be explained by the sensory deprivation hypothesis. The improvement of cognitive subdomains by auditory training remains controversial, and from the studies included in this article, it was found that auditory training mainly improves executive function rather than working memory and attention/processing. However, because of limitations and biases in the literature, it is not possible to determine this conclusively. In the future, to determine whether auditory training is effective in improving cognitive function in ARHL, more high-quality RCTs with long-term follow-up assessments are needed.

Abbreviations

ARHL

age-related hearing loss

CI

confidence interval

GRADE

Grading of Recommendation Assessment, Development, and Evaluation

HL

hearing loss

M age

mean age

MMSE

Mini-Mental State Examination

MoCA

Montreal Cognitive Assessment

RCT

randomized controlled trial

RoB

risk of bias

SD

standard deviation

SMD

standardized mean difference

WHO

World Health Organization