PDF
Citation
Print
INTRODUCTION
Hearing loss associated with advancing age is a common condition that is challenging to manage. Typically, older individuals with hearing impairments first begin to notice hearing loss in their forties or fifties [1]. The risk of experiencing hearing loss doubles with each passing decade [2]. If not addressed, this condition can impose considerable burdens on both individuals and society. Affected patients may struggle with communication, leading to social isolation and psychological issues such as depression [3]. Other serious consequences include a decline in cognitive ability and increased risks of hospitalization and mortality [4]. Some studies suggest a strong link between hearing loss, cognitive impairment, and dementia [5]. With global life expectancy on the rise, the prevalence of hearing loss is increasing. In the Republic of Korea, 12.0% of the general population and 44.7% of those over 70 years of age report subjective hearing loss [6]. Hearing loss is the fourth leading cause of disability worldwide, affecting approximately 6% to 8% of the global population [7].
Hearing aids (HAs) are the primary method for managing age-related sensorineural hearing loss, given the lack of effective medical or surgical treatments for mild to moderate cases [8]. These devices have been demonstrated to improve the quality of life and auditory capacity of adults with mild to moderate hearing loss over time [9].
Nevertheless, in developing nations, few people—approximately 2% of those considered to require hearing rehabilitation— have adopted HAs [10]. The rate of consistent HA usage among individuals with moderate to severe hearing loss in the Republic of Korea is 12.6%, which is lower than rates documented elsewhere [11]. Multiple factors contribute to the low HA adoption rate, including high costs, societal stigma, suboptimal efficacy, and inconvenience [1]. One common perception about HAs that deters patients from using them is that these devices may cause additional deterioration of the remaining hearing ability. Although this is often regarded as a misconception, insufficient scientific research is available to conclusively determine whether the belief that HAs cause further hearing loss is indeed unfounded.
Although previous research has suggested a typical progression of hearing loss over 5 or more years of HA use, few studies have investigated the long-term effects of HA on objective audiometric outcomes [12]. Therefore, it is essential to explore the lasting impact of HAs on hearing, especially in light of the emergence of non-prescription devices such as over-the-counter HAs and personal sound amplification products. These devices are designed to meet unmet needs for traditional HAs, but they may cause users to neglect regular hearing monitoring [13].
Thus, we aimed to determine the long-term effects of HAs on auditory function as measured by objective audiometry, including pure-tone average (PTA) and word recognition score (WRS), and to explore potential factors associated with changes in hearing.
MATERIALS AND METHODS
Study population
Among patients who purchased a new unilateral HA between 2015 to 2017 at a single tertiary center, those who showed consistent use for five years were enrolled in this study, and their medical records were retrospectively analyzed. Patients were excluded from the study if they had any of the following conditions: asymmetric hearing loss greater than 15 dB, middle-ear pathology, inner ear malformation, previous otologic surgery, previous radiation therapy, comorbidities that could affect hearing (such as probable or definite Meniere’s disease or brain tumor), or bilateral hearing amplification.
This study was approved by the Institutional Review Board of Samsung Medical Center (No. 2023-04-002) and performed in accordance with the tenets of the Declaration of Helsinki. The Board waived the requirement for informed consent considering low risk associated with a retrospective review of data.
Clinical data collected included diabetes mellitus, hypertension, anemia, major depressive disorder, stroke, cardiovascular disease, history of noise exposure, and previous HA use. Self-reported questionnaires of Hearing Handicap Inventory for the Elderly (HHIE) and International Outcome Inventory for Hearing Aids (IOI-HA) were completed at each visit. Audiometric data, HA type, time of use and the follow up period were also noted.
Hearing evaluation
During each appointment, pure-tone air conduction (AC) audiometry was conducted in a sound-proof booth using a GSI-61 Clinical Audiometer (Grason-Stadler) and Telephonics TDH-50P headphones (Teledyne Avionics). The audiometers were calibrated every 6 months to meet the American National Standards Institute guidelines (ANSI, 1996, 2004). The test measured the pure-tone AC thresholds of each ear at six frequencies of 250, 500, 1,000, 2,000, 4,000, and 8,000 Hz. If the threshold at any frequency was greater than 25 dB, bone-conduction thresholds were also tested. A PTA of 500, 1,000, 2,000, and 4,000 Hz was calculated for each patient [14]. Simultaneously, the WRS of each ear was obtained by presenting 50 single-syllable, single words at a sound level 30 dB louder than the speech reception threshold [15]. Additionally, a thorough otological evaluation, which included an endoscopic examination of the tympanic membrane, was performed for each participant in the outpatient clinic.
Fitting process
Every HA used in this study had a directional microphone and multiple channel processor. Participants who purchased a new HA were directed to our HA clinic for adjustment of programs. Prior to using the HA, their hearing-related difficulties were assessed using the HHIE questionnaire during their first visit. Initially, a proprietary formula was applied to set the gain at each frequency, and follow-up visits were conducted at 1, 3, and 12 months after the initial fitting to gradually increase the gain and achieve the target provided by the fitting formula. At each visit, participants completed self-reporting questionnaires, including the HHIE and IOI-HA, with the assistance of clinicians. The HA settings were modified based on the user’s needs, such as adding programs to increase gain in noisy conditions. At the three-month visit, the real-ear aided response was obtained to ensure the proper gain was achieved.
Questionnaires
The effectiveness of HA was assessed by analyzing the overall scores obtained from two questionnaires: HHIE and IOI-HA. To use questionnaires to evaluate outcomes in different languages, a series of steps including forward translation, reconciliation, reverse translation, and cognitive debriefing is required. The Korean versions of the HHIE and IOI-HA questionnaires have been accurately translated and validated, and they are dependable tools for evaluating self-reported outcomes related to hearing impairment in Republic of Korea [16,17].
The HHIE is composed of 25 questions that gauge the social and emotional impact of hearing loss [18]. Although it was originally developed for individuals with hearing impairment, it is a reliable measure to evaluate the benefits of HA usage. A higher score on the HHIE implies greater difficulties related to hearing impairment [19,20]. The IOI-HA questionnaire evaluates the effectiveness of HA based on seven key aspects of fitting outcomes: HA usage, benefit, activity limitations, satisfaction, participation restrictions, impact on others, and quality of life. A higher score on the IOI-HA indicates a better outcome with the use of HA [21].
Statistical analysis
Paired t-tests were performed to investigate statistically significant differences of audiometry results in the aided ear and unaided ear of each patient. A two-sided P-value <0.05 was considered statistically significant. Furthermore, Fisher’s exact test, the chi-square and Kruskal-Wallis tests were conducted for a subgroup analysis to find possible factors that attribute to the difference between the group which showed a significant hearing deterioration (AC_d) and the group which did not (AC_nd). Significant hearing deterioration was defined as 5 dB or more deterioration in the aided side compared to the unaided side. Pure tone AC was further analyzed by frequencies and divided into low (500 Hz), middle ([1 kHz+2 kHz]/2) and high (4 kHz) frequencies. Patients who showed deterioration of WRS on the aided side and patients who did not were also analyzed using Fisher’s exact test and the chi-square test. Statistical analyses were performed using IBM SPSS version 25.0 (IBM Corp.).
RESULTS
The study included 55 patients who consistently used a unilateral HA for 5 years (Table 1). The mean age of the participants was 73.49±12.83 years (range, 46–85 years), with 18 men (32.7%) and 37 women (67.3%). Of these patients, 31 were fitted with the HA on their right side (56.4%) and 24 on their left side (43.6%). Eighteen patients (32.7%) had a diagnosis of diabetes mellitus, 31 (56.4%) had hypertension, one (1.8%) had anemia, and six were receiving treatment for major depressive disorder. Additionally, four patients (7.3%) had a history of stroke, 26 (47.3%) had cardiovascular disease, and one patient (1.8%) reported a history of frequent noise exposure. Regarding device type, receiver-in-the-canal HAs were the most common, accounting for 54.5% of the devices. Twenty-five patients (45.5%) used their HA for more than 8 hours daily. The mean follow-up period for this study was 67.25±19.56 months.
Aided vs. unaided ears
The paired t-test was performed to assess differences between the aided and unaided ears at various frequencies (Table 2). No significant difference in AC PTA was found between the aided and unaided sides in pre-fit audiometry (53.47±9.25 dB vs. 52.86±13.84 dB, respectively; P=0.73), final follow-up audiometry (60.65±10.77 dB vs. 60.60±14.30 dB; P=0.98), or the change between the two (7.18±7.18 dB vs. 7.74±6.00 dB; P=0.58). Frequency-based analysis of the change in AC PTA from the pre-fit assessment to the final evaluation revealed no significant differences between the aided and unaided sides at low frequency (7.09±8.80 dB vs. 8.36±8.93 dB, respectively; P=0.30), middle frequency (7.45±8.07 dB vs. 7.95±6.75 dB; P=0.69), or high frequency (6.18±7.70 dB vs. 6.27±6.89 dB; P=0.93). Similarly, WRS showed no significant differences between aided and unaided ears in the pre-fit evaluation (82.38%±14.43% vs. 78.58%±20.93%, respectively; P=0.11), the final follow-up assessment (72.74%±19.94% vs. 66.79%±16.73%; P=0.07), or the difference between the two (9.33%±14.69% vs. 10.33%±17.31%; P=0.70
Subgroup analysis by AC threshold
Subgroup analysis was conducted to evaluate potential factors influencing hearing deterioration attributable to HA (Table 3). Of the participants, 11 patients (20.0%) exhibited an AC deterioration of 5 dB or greater on the side aided by the HA compared to the unaided side, with this group termed AC_d. In contrast, 44 patients displayed less deterioration, represented as AC_nd. Demographic data, questionnaire scores, and WRS results were compared between the AC_d and AC_nd groups.
Age, sex, laterality of amplification, medical comorbidities, type of HA, regular usage time of the HA, and fitting gain of the aided ear demonstrated no significant difference between the AC_d and AC_nd groups. Similarly, no significant difference was found between the groups regarding the difference between final and pre-fit HHIE or between final and first-fit IOI-HA. To determine if a decline in WRS corresponded with deterioration of AC, the ratio of the aided to the unaided WRS was used. Specifically, the AC_d and AC_nd groups were compared regarding the change in this ratio from pre-fit to final fit. This comparison between groups did not show statistical significance. Therefore, deterioration in AC thresholds was not associated with deterioration in WRS on the aided side.
The change in AC PTA from the pre-fit to the final assessment on the aided side was analyzed by frequency. Significant differences were observed between the AC_d and AC_nd groups for the lower and middle frequencies (P<0.001 and P=0.001, respectively). However, no significant differences were found at the high frequency (P=0.10).
Subgroup analysis by WRS
Regarding WRS, 23 patients displayed greater deterioration on the aided side (53.5%), while 20 patients experienced a greater decline on the unaided side (46.5%) (Table 4). No significant differences were found between these two groups in terms of age, sex, laterality of amplification, medical comorbidities, type of HA, regular usage time of the HA, or fitting gain of the aided ear. Similarly, the differences in HHIE scores between the final and pre-fit assessments, as well as those between the final and first-fit IOI-HA scores, did not differ significantly between groups. The change in PTA from pre-fit to final fit was analyzed between the WRS_deteriorated and WRS_non-deteriorated groups. Neither the absolute nor the proportional changes in AC from the pre-fit evaluation to the final follow-up were significantly different.
DISCUSSION
In the present study, 20.0% of participants exhibited a deterioration in AC of 5 dB or greater in the aided compared to the unaided side after long-term HA use. WRS showed a greater decline on the aided side in 53.5% of cases. However, we found no significant differences in hearing deterioration between the aided and unaided sides, in terms of either AC or WRS.
The National Institute on Deafness and Other Communication Disorders and the Department of Veterans Affairs conducted a longitudinal HA clinical trial in the 1990s. The findings revealed that patients who did not use HAs experienced a milder progression of hearing impairment in pure-tone thresholds compared to those who did use the devices [22,23]. These findings led to the hypothesis that the difference in threshold shift could stem from long-term acoustic trauma from the HAs. Another study from the same decade reported that despite audiologists recommending caution, the use of HAs could lead to temporary and even permanent changes in hearing thresholds due to exposure to unsafe sound levels [24]. This finding aligns with research suggesting that the impact of HAs is greater at higher frequencies than at lower frequencies [25].
In contrast, recent studies have shown results consistent with our own. Song et al. [26] found that the degree of change in pure-tone thresholds did not differ according to the use of HAs for amplification. Similarly, Johnson [27] reported a minimal threshold shift in children with HAs, except in cases of severe to profound hearing loss. As HA technology has advanced, allowing for more precise adjustments, the risk of acoustic trauma from excessive gain may have diminished. This could explain why recent studies, including ours, have found no significant difference in hearing impairment between aided and unaided ears. This also aligns with our finding that the change in AC threshold on the aided side did not differ significantly between the AC_d and AC_nd groups at a high frequency (P=0.10), suggesting that acoustic trauma was not a contributing factor.
In contrast, the changes in low and middle frequencies did exhibit significant differences between the AC_d and AC_nd groups (P<0.001 and P=0.001, respectively). Considering that the mean participant age was 73.49±12.83 years and that structural pathologies of the middle or inner ear, as well as low-frequency hearing-related diseases such as Ménière disease, were excluded, age-related hearing loss (ARHL) is likely the primary etiology of hearing loss in the enrolled participants. These individuals exhibited relatively poorer pre-fit high-frequency thresholds compared to the low or middle frequencies (Table 2). Since the quantitative study by Zwaardemaker [28] in 1891, ARHL has been recognized as sensorineural hearing loss that begins at high frequencies and progresses to low and middle frequencies. Thus, the significant differences observed solely at low and middle frequencies may not result from noise exposure due to HA use but rather the natural progression of ARHL.
Several prospective studies from the 1990s have suggested that HAs may have a protective effect against age-related decline in word recognition [29-32]. However, subsequent analyses examining the long-term use of HAs indicated that the use of a unilateral HA may not sustain or improve speech recognition abilities [26,33,34]. Our data indicate that the rate of decline in WRS was comparable on the unaided and the aided sides. To address potential selection bias—in which the side with poorer hearing might be fitted with an aid and subsequently exhibit greater long-term deterioration—patients with asymmetric hearing loss of 15 dB or more were excluded from the study. Accordingly, we noted no significant difference in pre-fit AC between the aided and unaided ears (P=0.73). Although we did find a noticeable difference in pre-fit WRS between the aided and unaided ears, it was not statistically significant (P=0.11). This is likely because, in cases of symmetrical hearing loss, unilateral HAs are typically prescribed for the ear with better speech recognition [35].
To identify variables associated with hearing progression in individuals using HAs, we compared various characteristics between the AC_d and AC_nd groups. The factors examined included comorbid medical conditions, type of HA, regular usage time of the HA, and the fitting gain of the aided ear. However, our analysis did not reveal any significant factors distinguishing the AC_d group from the AC_nd group. This suggests that the use of HAs does not affect the residual hearing of the user, regardless of individual characteristics. Additionally, a deterioration in AC was not significantly associated with a decline in WRS.
The primary limitation of this study was its retrospective design. To address the associated shortcomings, a prospective study of the potential hazardous impacts of HA use would be valuable. Furthermore, the sample size of our study may have been insufficient to detect factors associated with hearing deterioration. Additionally, the threshold value used to differentiate between AC_d and AC_nd was established at 5 dB, which is less than the commonly accepted testing error for PTA (10 dB) [36,37]. This decision was based on the observation that the mean hearing deterioration for both the aided and the unaided side did not exceed the testing errors (7.18±7.18 dB and 7.74±6.00 dB, respectively). Only one patient exhibited a hearing change that differed by more than 10 dB between aided and unaided sides.
Despite these limitations, the rigorous follow-up protocol at the HA clinic and the minimal amount of missing data lead us to believe that this study has effectively demonstrated the safety of HA use. After at least 5 years of HA use, 20% of participants exhibited greater deterioration in hearing on the aided side, as measured by AC PTA, and 53.5% showed more decline in WRS compared to the unaided side. Nevertheless, no significant factors were identified as contributing to this hearing deterioration, and overall, the use of HAs did not adversely impact residual hearing.
HIGHLIGHTS
▪ Hearing was compared between aided and unaided sides to assess long-term hearing deterioration attributable to the use of hearing aids (HAs).
▪ Of the patients, 20% exhibited a deterioration of air conduction of 5 dB or greater in the aided ear compared to the unaided side.
▪ Furthermore, 53.5% of patients displayed greater deterioration of word recognition score for the aided ear compared to the unaided side.
▪ Overall, the use of HAs did not negatively impact residual hearing.
▪ No significant factors were found to contribute to hearing deterioration.
XML Download