The current study is the first audiological assessment of the middle and inner ear functions of centenarian subjects. We show that all centenarian subjects have hearing loss with severity varying from moderate to severe in the low and mid frequencies and from severe to profound in the mid to high frequencies ( and ). Otoacoustic emissions, which reflect the status of outer hair cells, were undetectable in the majority of the subjects. The elevation of hearing threshold and absence of otoacoustic emissions suggest that sensorineural cause is largely responsible for age-related hearing loss seen in centenarian subjects.
Several large-scale epidemiologic studies were carried out in the United States and Europe to examine the prevalence and degree of hearing loss in the elderly population 
. The prevalence of hearing loss shown in these studies varies significantly from study to study. Comparisons of prevalence and degree of hearing loss among different studies are difficult because of the lack of agreement on a standard definition of hearing loss for use in epidemiologic studies, as well as differences in age and sex in the populations tested 
. Although there are many studies of the prevalence of age-related hearing loss in the United States and Europe, auditory function in the centenarian population has never been examined. The only study that examined hearing from more advanced age was done in an urban Swedish population aged between 85 and 90 
. The participants in that longitudinal study were followed audiometrically over a 20-year period from 70 to 90 years of age. The study focused on hearing and its decline during the later time span in which the participants were tested at the age of 85, 88 and 90 years. The results reveal that hearing loss in advanced age progresses only slightly in both men and women. The annual hearing threshold decline is about twice as large in the eighth decade of life as compared with the ninth 
. Although we did not follow the progression of age-related hearing loss of centenarian subjects in our study, the hearing loss appeared to be much worse than the population aged 85 and 90. We would like to point out that the current study is not an epidemiological study whose goal is to examine prevalence of hearing loss in the general population. Neither was our goal to compare and monitor hearing loss progression over time among different age groups. Our goal is to determine how well centenarian listeners can hear. It is likely that the extent and severity of hearing loss in the general centenarian population are slightly greater than what we reported here. This is because the current study excluded those who already had hearing loss due to genetic deficits (family history), histories of ototoxic drug usage and exposure to impulsive noise, and middle ear diseases. In addition, most subjects in this study were generally in reasonably good health (). It has been reported that changes in the blood supply to the ear because of heart disease, high blood pressure, and other circulatory problems can cause and/or aggravate presbycusis 
It has been demonstrated in several well-controlled studies that age-related hearing loss is more prevalent and severe in men than in women, especially at high frequencies 
. Cruickshanks et al. reported that the mean threshold difference was as large as 20 dB at 4,000 and 8,000 Hz between men and women aged between 60 and 64 years 
. Such difference is likely due to the fact that men have a greater risk of noise exposure in occupational settings. Interestingly, we did not observe a significant difference of hearing thresholds between men and women at any tested frequencies. The result is not entirely surprising since both men and women in this group already had severe to profound hearing loss. It is conceivable that the difference was minimized when majority of the subjects already suffered from a profound loss of hearing ().
Age-related hearing loss is a natural part of the aging process. Although it is generally accepted that morphological and physiological changes in the middle ear, cochlea, and central auditory system contribute to hearing loss, degeneration of the cochlear hair cells and/or atrophy of stria vascularis play the most important role 
. A recent study suggests that microRNAs, a class of short non-coding RNAs that regulate the expression of mRNA targets, are important regulators of age-related hearing loss 
. Animal studies have shown that outer hair cell loss, particularly in the basal turn of the cochlea, is associated with age-related hearing loss 
. Otoacoustic emissions decrease with age, likely signifying outer hair cell damage 
. A number of previous studies in humans have shown that as audiometric thresholds become poorer, the magnitude of the DPOAE response decreases and is ultimately eliminated 
. Some studies indicate that there is spiral ganglion cell loss during aging 
. Studies of the human temporal bone from patients with age-related hearing loss have shown a loss of capillaries within the spiral ligament and degeneration of the stria vascularis 
. It is important to point out that morphological and mechanical changes in the middle ear can also contribute to presbycusis. Such changes can result in reduced function of the tympanic membrane and the acicular chain. We show that the peak compliance of the middle ear was significantly reduced in centenarian subjects (). Furthermore, the bone-conduction thresholds are 10 to 20 dB better than the air-conduction thresholds (), suggesting that conductive hearing loss is partially responsible for age-related hearing loss in centenarian subjects. A number of previous studies have demonstrated the impact of reduced conductive hearing loss on otoacoustic emissions 
. Therefore, reduced middle ear function can partially be responsible for the absence/reduction of otoacoustic emissions seen in the centenarian listeners.
Finally, it is worth noting that the audiogram is not a particularly good measure of how one hears in real life. Hearing pure tones under headphones is quite different from listening to complex and dynamically changing sounds coming from different directions. Listening to pure tones in a quiet environment is also different from listening to conversation in background noise. Presbycusis is characterized by decreased hearing sensitivity and reduced speech recognition in a noisy environment. Although we did not perform speech audiometry, many studies have shown that speech discrimination is significantly reduced in the elderly 
. The reduced speech recognition is generally believed to be caused by degeneration of the central auditory pathway. Loss of function of the cochlear nerve has been shown in aged animals with reduced synchronous neural activity 
. This asynchrony may contribute to the decline in temporal resolution during aging. Other animal studies have shown decreased function in the cochlear nucleus 
. Thus, it is conceivable that morphological and physiological changes in the periphery (middle ear and cochlea) and central auditory system contribute to age-related hearing loss and difficulty understanding spoken language.
In conclusion, we show that although centenarian subjects still retain some residual hearing, more than 95% of them have server to profound hearing loss. It appears that both conductive and sensorineural causes contribute to the age-related hearing loss seen in the centenarian listeners.