Low Bone Mineral Density Is Associated with High-Frequency Hearing Impairment in Women Over 50: An Observational Study in Korea
Article information
Abstract
Background
Osteoporosis and hearing impairment are known to be associated, but specific data regarding gender, bone mineral density (BMD) measurement sites, and hearing frequency ranges remain unclear. This study aimed to clarify the relationship between hearing loss and BMD in adults over the age of 50. Additionally, the study sought to determine the frequency ranges of pure tone audiometry (PTA) related to osteoporosis, identify BMD measurement sites, and investigate gender differences.
Methods
A total of 1,523 adults (651 men and 872 women) over the age of 50, who participated in a medical health check-up at a university hospital, were included. PTA was conducted to assess hearing, and BMD was measured using dual-energy X-ray absorptiometry at the lumbar vertebrae (LV) and femur.
Results
In women over the age of 50, a significant association was observed between hearing impairment and osteoporosis (P<0.01), but no such association was found in men. Lumbar BMD (L1–4) in women was significantly associated with hearing loss at 4,000 and 8,000 Hz (both P<0.05), whereas femoral neck and total femur BMD showed no significant relationship. In multiple logistic regression analysis, the odds ratio (OR) between osteoporosis and hearing threshold at 4,000 Hz (OR, 2.078; 95% confidence interval [CI], 1.092–3.954) and 8,000 Hz (OR, 2.648; 95% CI, 1.543–4.544) remained statistically significant in women after adjusting for age and other risk factors.
Conclusion
In women over the age of 50, low BMD at the LV is significantly associated with hearing impairment, particularly at the high frequencies of 4,000 and 8,000 Hz.
INTRODUCTION
Age-related hearing impairment (ARHI), also known as presbycusis, is a common condition among the elderly population. Globally, the World Health Organization (WHO) estimates that hearing loss affects 538 million people [1]. In the United States, the incidence of hearing impairment in adults over the age of 65 is estimated to be over 30% [2]. In the Republic of Korea, the incidence of presbycusis ranges from 25% to 40%, with approximately 1.7 million individuals affected [3]. The aging population is expected to increase not only in Korea but globally, leading to a corresponding rise in the prevalence of hearing loss.
Various risk factors contribute to hearing loss in the elderly, including genetic susceptibility, noise exposure, ear diseases, trauma, aging, and ototoxic drugs [4,5]. Specifically, the hardening of the basilar membrane of the cochlea, atrophy of the spiral ligament, degeneration of the central nervous system, eardrum and ear ossicles, and circulation disturbances caused by atherosclerosis are possible mechanisms of ARHI [4-7]. Additionally, traditional cardiovascular risk factors, such as hypertension (HTN), diabetes mellitus (DM), and chronic kidney disease, are related to presbycusis [6,8]. This suggests that systemic conditions can affect hearing impairment.
Reduced bone mineral density (BMD) in the cochlea can also be associated with hearing impairment. For instance, patients with severe otosclerosis exhibit decreased BMD in the temporal bone, resulting in impaired hearing. Increased formation of immature bone due to otosclerosis leads to decreased BMD in the temporal bone, subsequently reducing mineral concentration in the cochlea and resulting in auditory cilia loss [7,9-11]. Another example is Paget’s disease, which causes increased bone resorption in the temporal bone, leading to decreased BMD in the cochlea and affecting the auditory cilia [12]. In addition, systemic BMD loss is associated with coronary artery calcium score, a risk factor for cardiovascular disease (CVD), or CVD itself [13]. Therefore, BMD and presbycusis may be interconnected through systemic conditions.
These findings suggest a potential relationship between systemic bone loss due to aging, including osteoporosis, and ARHI. However, studies investigating the association between osteoporosis and presbycusis have yielded inconsistent results. Initial research suggested a correlation between BMD and hearing [10,11,14], and a meta-analysis showed a significant association [15]. Data from the Korea National Health and Nutrition Examination Survey also indicated a significant association between hearing loss and femur neck BMD [16]. However, a cross-sectional study found no significant association between BMD and hearing in postmenopausal women, instead identifying a relationship with serum estradiol concentration [17]. A case-control study also showed no significant association [18]. Additionally, another study reported a relationship between BMD and hearing, but this association disappeared after adjusting for age and body mass index (BMI) [19]. Furthermore, the relevance between BMD and loss of hearing sensitivity differed by sex and race [20]. One possible reason for the inconsistent relationship between low BMD and hearing impairment may be the varying impact of bone loss on hearing depending on the frequency range.
This study aimed to elucidate the association between hearing impairment and BMD in adults over the age of 50. Additionally, it sought to determine the frequency ranges related to osteoporosis and BMD measurement sites, with a focus on investigating gender differences.
METHODS
1. Study Design and Population
This retrospective, single-center, cross-sectional observational study included a total of 2,920 adults aged 50 to 80 who visited the Health Promotion Center at Ajou University Hospital in Suwon, Gyeonggi Province, Republic of Korea, for medical checkups between January 1, 2006, and December 31, 2008. Of these, 1,329 participants who did not undergo pure tone audiometry (PTA) for hearing tests and spine and femur dual-energy X-ray absorptiometry (DXA) for BMD testing were excluded. Additionally, five participants who did not complete the self-report questionnaire, 53 participants with ongoing ear diseases, and 10 participants taking medications affecting hearing, such as loop diuretics, antibiotics, or chemotherapeutic agents, were also excluded. Ultimately, a total of 1,523 participants (651 males, 872 females) were included in the final analysis (Figure 1).
Study flowchart. DXA, dual-energy X-ray absorptiometry; PTA, pure tone audiometry.
This study was conducted after receiving approval from the institutional review board of Ajou University Hospital (AJOUIRB-DB-2023-285) and adhered to the ethical principles outlined in the 1975 Declaration of Helsinki and current ethical guidelines. The requirement for informed consent from individual patients was omitted because of the retrospective design of this study.
2. Data Collection
BMD was measured by trained technicians using DXA (Lunar Prodigy Advance; GE Lunar, Madison, WI, USA). Bone mineral content (BMC; g), area (cm2), and T-scores were measured at the first, second, third, and fourth lumbar vertebrae (LV; L1–4), femoral neck (FN), and total femur (TF). BMD (g/cm2) was calculated by dividing the area (cm2) by the BMC (g). Osteopenia and osteoporosis were diagnosed based on the T-scores of L1–4, FN, and TF according to the WHO criteria: T-score ≥−1 (normal), −2.5< T-score <−1 (osteopenia), and T-score ≤−2.5 (osteoporosis).
PTA was conducted at frequencies of 500, 1,000, 2,000, 4,000, and 8,000 Hz using the Madsen Itera II audiometer (Natus Medical, Taastrup, Denmark). The pure tone average threshold (PTAT) was calculated as the mean of the thresholds at 500, 1,000, and 2,000 Hz. Hearing impairment was defined as a PTAT above 25 dB, based on the 1964 ISO standard, and thresholds above 40 dB at 4,000 and 8,000 Hz were also classified as hearing impairment.
All participants underwent a complete physical examination and anthropometric measurements. These measurements were conducted while participants wore light gowns and no shoes, following an overnight fast. Height was measured to the nearest 0.1 cm, and weight was measured to the nearest 0.1 kg using an automatic height and weight measuring machine. BMI was calculated by dividing weight (kg) by height squared (m²). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured in the sitting position using an automatic blood pressure measuring machine. Fasting blood sugar (FBS), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) levels were assessed from venous blood samples obtained after a fasting period of at least 12 hours. Participants were also asked to provide information about their medical history, including ear and systemic diseases, current medications, and smoking history, via a self-report questionnaire.
3. Statistical Analysis
Baseline characteristics of gender differences were compared using Student t-test. The hearing thresholds for each frequency range across the normal, osteopenia, and osteoporosis groups were compared using analysis of variance (ANOVA) with Bonferroni correction. Age-adjusted comparisons were conducted using analysis of covariance (ANCOVA). The same analysis (ANOVA with Bonferroni correction and ANCOVA for age adjustment) was performed to compare the T-scores of the L1–4, FN, and TF based on PTAT and thresholds at 4,000 and 8,000 Hz. Finally, multiple logistic regression analysis was conducted to calculate the odds ratios (ORs) for hearing impairment associated with each frequency range and osteoporosis, adjusting for age, current smoking, HTN, DM, hormone replacement therapy (HRT), history of myocardial infarction (MI), history of cerebrovascular accident (CVA), TC, LDL-C, and HDL-C. All statistical analyses were two-tailed, and a P-value of less than 0.05 was considered statistically significant. Data were analyzed using SPSS ver. 17.0 software (SPSS Inc., Chicago, IL, USA).
RESULTS
1. Baseline Characteristics
Baseline characteristics of the study subjects are presented in Table 1. The mean age of participants was 57.61±6.2 years, with no significant difference between males and females. BMI was similar between the sexes, with males having an average BMI of 24.38±2.80 kg/m² and females having an average BMI of 24.03±2.88 kg/m². There were no significant differences in SBP, DBP, FBS, and cholesterol levels. The prevalence of osteoporosis was significantly higher in females, with 4.0% of males and 12.0% of females affected. The smoking rate was significantly higher in males, with 37.6% of males and 3.1% of females being smokers. Regarding hearing impairment, the prevalence of low-frequency range impairment (500, 1,000, and 2,000 Hz) was higher in females, at 24.8% compared to 18.4% in males. However, the prevalence of high-frequency range impairment at 4,000 and 8,000 Hz was higher in males, at 44.2% and 53.5%, compared to 13.9% and 39.0% in females, respectively.
Baseline characteristics of the study population
2. Osteoporosis, Osteopenia, and Hearing Threshold
In males, despite a decrease in BMD, hearing thresholds remained consistent across all frequency ranges. However, in females, lower BMD was significantly associated with elevated hearing thresholds. In females, the difference in hearing threshold between the normal group and the osteoporosis group was 5.5 dB at 500 Hz in the left ear and 12.7 dB at 8,000 Hz in the left ear. This pattern was observed in both the left and right ears (Table 2). After adjusting for age, males showed no significant association between BMD and hearing thresholds, whereas females with osteopenia or osteoporosis exhibited significantly higher hearing thresholds across all frequency ranges, except for 1,000 Hz (left P=0.06, right P=0.23) (Table 3).
Differences in hearing threshold (dB) according to sex, osteopenia, and osteoporosis
Differences in hearing threshold (dB) according to sex, osteopenia, and osteoporosis, adjusted for age
3. BMD Measurement Sites and Hearing Threshold
The association between BMD at different measurement sites (L1–4, FN, and TF) and hearing loss in each frequency range was assessed using ANCOVA with age adjustment. In males, no significant associations were found between lumbar BMD and hearing thresholds. However, in females, a statistically significant association was observed between L1–4 BMD and hearing thresholds at the high-frequency ranges of 4,000 and 8,000 Hz (Figure 2). No significant associations were found between the BMDs of the femur (both neck and total) and hearing thresholds in either gender (Figures 3, 4).
Differences in hearing threshold by frequency ranges in the left ear and average bone mineral density (BMD) in lumbar vertebrae 1–4: (A) men and (B) women. The P-values were significant only in women for the high frequency ranges. *P<0.05 for 4,000 and 8,000 Hz. P-values were measured using analysis of covariance with age adjustment. Left pure tone average threshold (Lt PTA) was calculated as the average of hearing thresholds at 500, 1,000, and 2,000 Hz.
Differences in hearing threshold by frequency ranges in the left ear and average bone mineral density (BMD) in the femoral neck: (A) men and (B) women. No significant correlations were observed. P-values were measured using analysis of covariance with age adjustment. Left pure tone average threshold (Lt PTA) was calculated as the average of hearing thresholds at 500, 1,000, and 2,000 Hz.
Differences in hearing threshold by frequency ranges in the left ear and average bone mineral density (BMD) in the total femur: (A) men and (B) women. No significant correlations were observed. P-values were measured using analysis of covariance with age adjustment. Left pure tone average threshold (Lt PTA) was calculated as the average of hearing thresholds at 500, 1,000, and 2,000 Hz.
Multiple logistic regression analysis revealed that the OR for hearing loss in males increased as BMD decreased from normal to osteopenia and osteoporosis, based on the PTAT, before adjusting for confounding factors. However, after adjusting for risk factors such as age, smoking, HTN, HRT, DM, MI, CVA, TC, LDL-C, and HDL-C, the statistical significance disappeared. There were also no significant changes in the risk of hearing loss in the 4,000 and 8,000 Hz frequency ranges. In females, the OR for hearing loss significantly increased based on the PTAT before adjustment, but after adjusting for risk factors, the significant difference disappeared. Nevertheless, the risk of hearing loss increased with decreasing BMD in the 4,000 Hz (OR, 2.078; 95% confidence interval [CI], 1.092–3.954 after adjustment) and 8,000 Hz (OR, 2.648; 95% CI, 1.543–4.544 after adjustment) frequency ranges, both before and after adjustment (Table 4).
Odds ratios for osteopenia and osteoporosis with respect to hearing threshold, based on multiple logistic regression analysis
DISCUSSION
In the current study, no significant association was found between hearing impairment and the decline of BMD in males. However, in females, a significant association was observed between hearing impairment and BMD decline, especially in the 4,000 and 8,000 Hz frequency ranges after adjusting for various risk factors, including age. Specifically, the BMD of LV showed a significant correlation with hearing threshold, but not the BMD of the femur. Therefore, when managing patients with osteoporosis or hearing impairments, especially in the elderly, it is important to be aware of their relationship and assess them in relation to each other. In patients with osteoporosis or osteopenia, physicians should consider screening for symptoms of hearing loss and conducting PTA or consulting an ear doctor. Conversely, in patients with hearing loss, the ear doctor should also consider performing a BMD test and treating osteoporosis and osteopenia. In some double-blinded studies, the bisphosphonate treatment group showed maintenance of hearing thresholds and stabilization of hearing loss [21,22]. Additionally, many clinical studies have elucidated that vitamin D deficiency is strongly associated with hearing impairment [23,24].
As mentioned before, previous studies have reported mixed results regarding the association between BMD and hearing ability. Some studies found a relationship [10,11,14-16], while others did not, particularly in the low-frequency range [17,18]. Moreover, these associations often disappeared after adjusting for confounding factors such as age or BMI [19]. Additionally, inconsistent results have been reported across different populations and research methods [20]. Therefore, it is evident that different conclusions can be drawn depending on the study population and research approach.
In the current study, only females showed an association between hearing ability and BMD after adjusting for age and risk factors. Generally, males tend to have a higher prevalence of high-frequency hearing loss, which can be attributed to greater exposure to noisy environments [25]. Although the present study observed more severe high-frequency hearing impairment in males compared to females, this impairment did not appear to be related to BMD. This may be attributed to differences in BMD distribution between the sexes and the reduction of female hormones after menopause [26]. Studies have shown that females with Turner syndrome, who experience estrogen deficiency, exhibit hearing loss associated with low BMD [27]. Estrogen receptors have also been associated with ARHI in a mouse study [28]. Furthermore, research on a small population of postmenopausal women reported that HRT reduced the degree of hearing impairment [29]. Additionally, BMD is generally higher in males than in females at the same age [30]. Therefore, even with the same osteoporosis classification based on T-scores, males have higher BMD overall compared to females. Consequently, the impact of BMD reduction on the auditory system may be smaller in males than in females.
What are the mechanisms underlying the higher prevalence of hearing loss in osteoporosis patients, particularly in the high-frequency range? The reduction in total body BMD is proportional to the reduction in temporal bone BMD [10,31,32], which affects the mineral content of the periosteum surrounding the cochlea. This, in turn, influences auditory hair cells and leads to hearing loss. In this process, high-frequency hearing ability is typically affected first. Previous studies have suggested that the high-frequency range is associated with the organ of Corti at the basal end of the cochlea, and the loss of auditory hair cells in this region is related to high-frequency hearing loss [6]. Evidence supporting the association between BMD reduction and high-frequency hearing loss is evident in female Turner syndrome patients, who experience decreased BMD and demonstrate reduced hearing ability in the high-frequency range [27,33]. Additionally, with aging, apoptosis of osteocytes in the ear ossicles and hypermineralization can occur, which may connect hearing impairment and osteoporosis [34].
The current study has several limitations. First, it has a retrospective design, which does not allow for the establishment of any causal relationships between parameters. Second, the examination of ear diseases relied on self-administered questionnaires, without examinations such as otoscopy or evaluation by an otolaryngologist. Additionally, information on noise exposure was lacking. Third, serum estrogen levels were not measured, thus the influence of estrogen on hearing ability and BMD could not be investigated. Lastly, we could not verify the medication history for osteoporosis, which could also be an important confounding factor for the study results.
However, the current study possesses strengths. It was conducted with a large sample size using health screening data from a university hospital. Furthermore, we assessed hearing thresholds across various frequency ranges in relation to the BMD of LV and the femur. Additionally, we employed precise measurement methods such as PTA and DXA.
In conclusion, in the population group above the age of 50, which is predominantly affected by hearing loss, no association was found between hearing impairment and osteoporosis in males. However, in females, a significant relationship was observed between the severity of hearing loss and the reduction of BMD. This association was statistically significant in the 4,000 Hz and 8,000 Hz frequency ranges, rather than in the lower frequency range, and was consistent in both the left and right ears. Notably, the BMD of the LV showed a stronger correlation with hearing ability than the BMD of the femur. To establish a causal relationship between BMD and hearing ability, further long-term prospective studies will be necessary. Additionally, animal experiments will be required to elucidate the precise mechanisms by which the reduction in BMD affects hearing ability.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.