Hearing Loss: What It Is, What Causes It, What Helps

Hearing loss is one of the most common long-term health conditions in the world, and one of the least well explained to the people who have it. Most people arrive at an audiologist with a story rather than a measurement: people keep mumbling, I cannot follow conversation in restaurants, the phone has got hard, I turn the television up. Every one of those complaints is real. None of them tell you what is actually happening in the ear.

This page is written to close that gap. It explains the three main types of hearing loss in plain terms, what the decibel bands on an audiogram actually mean, what tends to cause hearing loss at different ages, how it is measured, and what the established options are. It also says plainly which things do not work, because the supplement category around hearing has grown large and most of what it sells is not aimed at a mechanism that exists.

It is educational information, not medical advice, and it is not a substitute for being assessed. Nothing here is designed to tell you what is wrong with your ear. That is a job for a clinician with a booth, a scope and an audiogram.

Sound reaches the brain through a long chain of structures, and hearing loss is usually named after the point in that chain where the problem sits. The NIDCD description of how we hear is worth keeping in mind, because the vocabulary only makes sense once you can picture it.

Sound waves enter the ear canal and vibrate the eardrum. The eardrum passes those vibrations to three tiny bones in the middle ear, the malleus, incus and stapes, which amplify them. The amplified vibrations reach the cochlea, a fluid-filled snail-shaped organ in the inner ear. A travelling wave forms along a membrane inside it, and hair cells riding that wave bend tiny hair-like projections called stereocilia. That bending opens pores, creates an electrical signal, and the auditory nerve carries it to the brain. Hair cells near the wide end of the cochlea detect higher-pitched sounds; hair cells nearer the centre detect lower pitches.

Every one of those steps can fail, and the step that fails determines the type of loss, the likelihood of recovery, and what can be done about it.

Conductive hearing loss

Conductive loss means sound is not getting through efficiently on its way to the inner ear. The inner ear is working; the delivery route is impaired. Common causes include earwax blocking the canal, fluid or infection behind the eardrum (otitis media), a perforated eardrum, a nose blockage that stops the middle ear from equalising pressure, and otosclerosis, in which abnormal bone remodelling causes the stapes to become fixed in place and stop transmitting vibration. The NIDCD notes that otosclerosis is usually first noticed as an inability to hear low-pitched sounds or a whisper, starting in one ear and often moving to the other.

The clinical significance is straightforward. A conductive cause is something a clinician can often act on directly, sometimes with a procedure. A blocked canal can be cleared, fluid can be drained or allowed to resolve, and otosclerosis can sometimes be treated with a stapedectomy, in which a prosthetic device bypasses the abnormal bone. None of those outcomes is available for damaged hair cells.

Sensorineural hearing loss

Sensorineural loss means the problem is in the cochlea or the auditory nerve: the sensory cells themselves, or the fibres carrying the signal. This is the commonest type of hearing loss in adults, and it is what most people mean when they talk about going deaf. The NIDCD defines it as hearing loss caused by damage to the sensory cells and/or nerve fibres of the inner ear.

It matters that human cochlear hair cells do not regenerate. The NIDCD is explicit about this in the context of noise-induced hearing loss: unlike the hair cells of birds and amphibians, human hair cells do not grow back, and they are gone for good. That single fact is the reason this page spends a section on prevention rather than on products.

Mixed hearing loss

Mixed loss is simply a combination: there is a conductive problem and a sensorineural problem in the same ear. It is not a third mechanism, but it is recorded separately because it changes management. Treating only the conductive component will improve hearing but will not restore it to the level of someone without sensorineural damage, and the audiogram has to be read by someone trained to see the difference. In practice, mixed loss is common in people who have lived with a long-standing middle-ear problem and have also accumulated age-related or noise-related inner-ear damage.

Decibels are not like centimetres or ounces. The scale is logarithmic, which is why the NIDCD uses a building as its illustration: if a building is 80 feet tall, adding another 10 feet makes it 12.5 percent taller, which feels trivial. Adding 10 decibels to a sound makes it ten times more intense, and it sounds about twice as loud. That is why small numbers on a decibel scale represent large physical differences, and why 85 dBA is a threshold that matters more than 80 dBA.

For everyday noise, the A-weighted scale (dBA) is used, because it weights frequencies according to how the human ear responds. The NIDCD gives useful reference points: about 0 dBA is a sound you can barely hear, whispering is around 30 dBA, normal conversation sits around 60 to 70 dBA, and a jet taking off reaches roughly 140 dBA.

Two cautions about those bands. First, they are averages across the frequencies tested, and the shape of the curve matters as much as the average. Someone with a mild average but a steep high-frequency drop can struggle badly to understand speech in noise, because consonants carrying much of the intelligibility of speech sit in exactly that region. Second, the bands describe degree, not cause. A 50 dB average caused by removable fluid and a 50 dB average caused by decades of noise exposure are the same measurement and completely different problems.

If you are trying to interpret your own results, take them to a clinician. The number is the easy part; the pattern, the symmetry between ears and the history are what carry the meaning.

Noise exposure, at any age

The NIDCD states that sounds at or below 70 dBA are unlikely to cause hearing loss even after long exposure, while long or repeated exposure at or above 85 dBA can cause it, and that the louder the sound, the shorter the time needed. Repeated exposure causes damage and eventual death of hair cells. Sometimes exposure to impulse or continuous loud noise causes a temporary hearing loss that disappears 16 to 48 hours later, but the NIDCD notes that although the loss seems to disappear, there may be residual long-term damage.

That is the mechanism behind a lot of adult hearing loss that people do not identify as noise-related, because the damage was accumulated decades earlier: band rehearsals, club nights, power tools, hunting, and years of industrial work. The NIDCD also flags personal audio devices, noting that headphones can reach 110 dBA at maximum volume.

Children and young adults

In this group the dominant causes are genetic or developmental rather than acquired. A child born with a hearing impairment usually falls into one of two broad patterns: loss present from birth, which is very often genetic in origin, and loss acquired during a critical period of speech and language development, whether from an infection, a medical condition affecting the ear, or the use of certain medicines. The NIDCD treats genetic causes of hearing loss as an active research area, which reflects how large and heterogeneous this group is. It also notes that children are not simply small adults when it comes to the ear: their eustachian tubes are smaller and more level, their immune systems are still developing, and both of those differences make middle-ear problems more common in childhood than in adulthood.

In older children and teenagers, the picture shifts towards the noise-related pattern, because recreational exposure begins. The <a href="/hissing-noise-in-ears/">first thing many people notice is a high-pitched hiss</a>, which is exactly the pattern associated with damage in the high-frequency part of the cochlea.

Middle age

Two things dominate. The first is the beginning of age-related change, which the NIDCD describes as gradual and typically occurring in both ears, and which is covered in full in our guide to <a href="/presbycusis/">presbycusis</a>. The second is continuing noise exposure, occupational or recreational, which frequently adds to the age-related component rather than appearing on its own.

Middle-aged adults also commonly present with the conductive causes discussed above, particularly wax and middle-ear fluid, and with a sensorineural pattern that is asymmetric. A clearly one-sided loss in a middle-aged or older adult is taken seriously by clinicians, because a tumour on the auditory nerve has to be excluded even though it is an uncommon cause.

Older adults

The NIDCD lists the contributors to age-related hearing loss: changes in the inner ear, changes in the middle ear, and complex changes along the nerve pathways from ear to brain. It also notes that long-term noise exposure can play a role, that certain genes may make some people more susceptible, and that conditions more common in older people such as high blood pressure and diabetes are associated with hearing loss. It also points out that some medicines toxic to the sensory cells of the ear, some chemotherapy drugs among them, can cause hearing loss.

Note the wording there. Those are described as associations and as factors that can also play a role, not as proven causes. The distinction matters and we come back to it in the <a href="/presbycusis/">presbycusis guide</a>.

Pure-tone audiometry: the audiogram

Pure-tone audiometry is the standard test. In a sound-treated booth you wear headphones and are played tones at a range of pitches. For each pitch the audiologist searches for the quietest level at which you indicate you can hear a tone, and that threshold is recorded in decibels. The results for each ear are plotted separately on an audiogram, with frequency along one axis and hearing level in decibels along the other.

The NIDCD notes that the audiometer was invented by Alexander Graham Bell, and that a modern version of it is still the device used to diagnose hearing loss today. The shape of the resulting plot is where the clinical information sits: a flat line near normal, a downward slope affecting high pitches, a notch at a single frequency, or a gap between air and bone conduction that tells you the loss is conductive. That last part is not one test but a comparison, because clinicians test thresholds with sound delivered through air and, where a conductive component is suspected, through the skull bone as well.

Speech testing

Pure tones are a convenient laboratory stimulus, not a particularly good model of real speech. Speech testing therefore exists alongside audiometry. In a speech reception test, recorded word lists are presented at a set level and you are asked to repeat what you hear, and the result is scored as a percentage of words understood. Speech detection thresholds record the level at which you detect speech without understanding it. Comparing the two tells the clinician something important: a person whose detection is reasonable but whose understanding score is poor is having difficulty with intelligibility, not merely with volume. The classic illustration of this is the high-frequency pattern of age-related loss, where words are audible but not reliably understood, especially in background noise.

Other tests you may meet

• Tympanometry. A small probe in the ear canal varies air pressure and measures how the eardrum moves, which identifies fluid or pressure problems in the middle ear. The NIDCD describes it as a test using sound tones and air pressure, used when the diagnosis of a middle-ear problem is otherwise unclear.

• Otoscopy. A lighted instrument, or a video otoscope, so the clinician can see the eardrum directly. Wax, infection, a bulge or a perforation are visible this way.

• Acoustic reflex testing. Measures the involuntary contraction of a middle-ear muscle in response to sound, which helps distinguish between types and sites of lesion.

• Self-screening questionnaires. The NIDCD publishes a ten-item questionnaire adapted from the Hearing Handicap Inventory for Adults, for which the source research is the Hearing Handicap Inventory for the Elderly, designed for adults aged 65 and over. Answering yes to three or more questions is the threshold the NIDCD gives for considering an otolaryngologist or an audiologist. These tools are designed to decide who should be tested, not to measure anything.

The AAO-HNS clinical practice guideline on sudden hearing loss defines sudden sensorineural hearing loss as a loss of 30 dB or more over three consecutive frequencies within 72 hours. The AAO-HNS patient information on the same topic notes that sudden hearing loss usually affects one ear and may come with dizziness, ringing or a feeling of pressure, and advises seeing a healthcare provider as soon as possible.

The recommendation that makes this a time-critical problem is the one about treatment. The guideline recommends obtaining audiometry as soon as possible, notes that corticosteroids may be offered within two weeks of onset, and recommends follow-up audiometry within six months. Two weeks is the operative number. It is a short window for symptoms that people routinely explain away, because the NIDCD notes that people with sudden deafness often put off seeing a doctor by attributing the loss to allergies, a sinus infection or earwax.

That last point is the trap. The three most common self-diagnoses for sudden loss are all conductive causes, and all three are also capable of producing a blockage sensation rather than a true sensorineural loss, which is precisely the confusion that <a href="/muffled-hearing/">the muffled-hearing guide</a> is about. The NIDCD is clear that when sudden deafness is suspected, a doctor should rule out conductive hearing loss first, and that pure tone audiometry should be performed within a few days of symptom onset to identify any sensorineural component.

So: if your hearing dropped over hours or days, do not start a supplement, do not clean your ears, and do not wait to see whether it settles. Be seen.

Hearing aids

The NIDCD describes hearing aids as devices that amplify sound for people with hearing loss caused by damage to the inner ear, and they are the most established intervention in the guideline literature. The logic is mechanical: if the difficulty is that too little sound reaches the cochlea, then increasing the sound that reaches it addresses the problem directly. Modern aids are fitted to an individual audiogram by an audiologist, and over-the-counter options now exist for adults with mild to moderate loss. The NIDCD also notes the access gap plainly: of people who could benefit from hearing aids, only about one in four has ever used one.

Cochlear implants

A cochlear implant is a different mechanism, and the NIDCD is precise about it. An implant has an external portion behind the ear and an internal portion placed under the skin. A microphone picks up sound, a speech processor selects and arranges it, and an electrode array sends electrical impulses to different regions of the auditory nerve. Where a hearing aid amplifies, an implant bypasses the damaged portion of the ear and stimulates the nerve directly.

The NIDCD also states the limitations clearly, and they should be understood before anyone pursues this route. An implant does not restore normal hearing. It gives a useful representation of environmental sound and can help people understand speech, and it allows many people to recognise warning signals, but hearing through an implant is different from normal hearing and takes time to learn. Using one requires surgery and significant therapy afterwards, speech-language pathologists and audiologists are usually involved, and the decision should be discussed with an experienced implant surgeon.

Assistive listening devices

Assistive devices sit alongside hearing aids rather than instead of them, and they are consistently under-used. The NIDCD maintains its own resources on assistive devices, and the practical categories are worth knowing. A personal alerting device can be set to flash a light for a doorbell, a smoke alarm or a telephone. A telephone amplifier or a captioned telephone makes calls usable. A television hearing system streams the audio directly to headphones or a receiver. In meetings, restaurants and small rooms, a portable system that loops audio through the ear or sends it to a receiver can be the difference between joining a conversation and withdrawing from it. These devices do not treat the ear. They remove the acoustic obstacles that a hearing aid alone cannot solve, particularly distance and background noise.

Audiological rehabilitation

The fourth route is the one most often left out of consumer information, and it is the reason an audiologist appointment is not simply a measurement session. Audiological rehabilitation is the training that makes hearing a usable sense again: communication strategies, speech-reading, structured listening practice, counselling about hearing-aid use and care, and, where tinnitus is present, directed approaches to managing it.

It is worth being clear about what each of these routes is and is not. They do not repair a cochlea. They compensate for the ear, and in doing so they require work from the person using them, which is why the training component is not optional packaging but part of the mechanism.

The honest position is short and unambiguous. There is no supplement that has been shown to treat hearing loss. No clinical trial has demonstrated that a dietary supplement improves audiometric thresholds, restores hair cells, or reverses established sensorineural damage. The most-studied supplement in the hearing category, ginkgo biloba, failed the Cochrane reviews for tinnitus, and tinnitus is a different condition again.

There is legitimate laboratory science about why people find this plausible. Oxidative stress and microvascular insufficiency are real, well-characterised contributors to inner-ear injury in noise-trauma and ageing models, and ingredients with antioxidant or vascular activity have real pharmacology. That gives a mechanism a reason to be investigated. It does not give a product a result. Animal cochlear protection from green tea catechins is not a demonstration that swallowing green tea extract helps a person with hearing loss, and no human trial of a hearing supplement blend on audiometric thresholds exists.

The practical consequence is worth stating plainly, because it is the point of this page. If you have measurable hearing loss, the highest-value purchase available to you is an assessment, not a bottle. The assessment tells you what type of loss you have, what caused it, whether any of it is reversible, and whether aids or implants are appropriate. A supplement cannot return any of that information, and money spent on one is money not spent on the part of the pathway that has evidence behind it.

If you have already been assessed, been told there is nothing treatable, and simply want to try a supplement anyway, that is a legitimate choice of how to spend your own money, provided it is understood for what it is. What it is not is a treatment for hearing loss.

If you have read to the end of this page, the practical next step is not complicated. The NIDCD points to two kinds of professional: an otolaryngologist, which is an ear, nose and throat specialist, and an audiologist, who is trained to identify, measure and treat hearing disorders. Both are appropriate. An audiologist is where most people should start for measurement; an ENT is where you need a medical assessment, a decision about a procedure, or an urgent look at a sudden or one-sided change.

It helps to know the questions worth asking. What is the type of loss, conductive, sensorineural or mixed? Is it symmetrical? What does the pattern across frequencies suggest? Is any part of this reversible? Is amplification indicated, and is assistive listening worth adding? If the plan is a hearing aid, is there a trial period, and is there a fitting follow-up included rather than being treated as an extra? These are ordinary, reasonable questions, and the answers should be given without difficulty.

Published 25 Feb 2026Last updated 25 Sept 2026