Age-related hearing loss, presbycusis, isn’t a single condition. It’s at least four distinct patterns of degeneration inside the cochlea, each with a different signature on your audiogram and a different impact on your ability to follow conversation. Understanding which type you have (or more often, which combination) helps explain why two seventy-year-olds with identical pure-tone thresholds can have wildly different experiences in a noisy restaurant.
The classification system we still use comes from Harold Schuknecht, a temporal bone pathologist who spent decades examining temporal bones donated to research. What he found changed how we think about ageing ears: the problem isn’t always the cells that detect sound, and it isn’t always fixable with louder volume.
The Cochlea’s Invisible Power Station
Before we get to the four types, you need to know about the stria vascularis. It’s a three-cell-thick stripe of tissue lining the outer wall of the cochlea, and it does something remarkable: it pumps potassium ions uphill to create an electrical potential of roughly +80 millivolts in the endolymph, the fluid that bathes the tops of your hair cells. That doesn’t sound like much until you realise it’s one of the largest standing electrical potentials in the body.
When a sound wave bends a hair cell’s stereocilia, ion channels open and that +80 mV battery drives current through the cell. No battery, no signal. When the stria declines, hearing sensitivity drops across the board.
Schuknecht’s Four Types of Presbycusis
1. Metabolic (Strial) Presbycusis
This is stria vascularis failure. The battery slowly goes flat. What’s characteristic about metabolic presbycusis is that the audiogram is flat, similar loss across all frequencies, and often fairly mild, maybe 30 to 40 dB. But the word recognition scores are usually quite good, because the hair cells and nerve fibres themselves are intact. You just need more volume.
The stria is one of the most metabolically active tissues in the body and relies on a good blood supply, which is why cardiovascular risk factors such as smoking, diabetes and high blood pressure are thought to contribute to its decline. Patients often describe everyone mumbling, because soft consonants (the s, f and th sounds that carry meaning) fall below their threshold.
This type usually responds well to hearing aids, provided they’re programmed and verified with real-ear measurement. When the problem is mainly a lack of gain, well-fitted amplification goes a long way.
2. Sensory Presbycusis
This is classic high-frequency hair cell loss, starting in the basal turn of the cochlea where high-pitched sounds are encoded. The audiogram slopes steeply: normal or near-normal thresholds at 250 and 500 Hz, then dropping to 50, 60, 70 dB or worse by 4000 Hz.
Sensory presbycusis often begins earlier, sometimes in your forties if you’ve had significant noise exposure, and it’s cumulative. Every rock concert, every power tool, every year of commuting on the M2 with the windows down adds up. The hair cells don’t regenerate. Once they’re gone, they’re gone.
Word recognition is generally proportional to the pure-tone average. If we amplify the high frequencies appropriately (and avoid over-amplifying the lows, which causes that “talking in a barrel” sensation), most people do reasonably well. The challenge is that many people wait years before seeking help, and by then the brain has had a long stretch of reduced input to adapt to.
3. Neural Presbycusis
This is loss of the spiral ganglion cells: the actual nerve fibres that carry the signal from hair cells to the brainstem. The audiogram might look relatively benign, but word recognition scores are disproportionately poor. Someone with a 40 dB pure-tone average might score 60% on phonemically balanced word lists in quiet, where you’d expect closer to 90%.
Neural presbycusis is insidious. Hearing aids can make things louder, but they can’t restore clarity when the problem is the wire, not the microphone. This is the patient who says “I can hear you, I just can’t understand you,” and they’re exactly right. Speech testing, particularly speech-in-noise testing, often reveals the deficit more clearly than pure-tone audiometry.
There’s no strong fix for neural presbycusis yet. We can improve the signal-to-noise ratio with directional microphones and noise reduction algorithms, and some patients benefit from cochlear implants if the loss is severe enough. But fundamentally, if you’ve lost 50% of your spiral ganglion cells, no amount of digital processing will recover the information they once carried.
4. Mechanical (Cochlear Conductive) Presbycusis
The rarest type. This is stiffening of the basilar membrane, the structure that vibrates in response to sound and determines which hair cells fire. The audiogram shows a gradually sloping loss, but unlike sensory presbycusis, it’s not restricted to the high frequencies: it’s more linear.
Schuknecht hypothesised this was due to changes in the physical properties of the membrane itself, possibly calcification or changes in the collagen matrix. It has always been the most speculative of the four categories, and it can’t be confirmed from an audiogram alone.
The Reality: Mixed Patterns
Real human ears don’t read Schuknecht’s textbook. By the time you’re seventy, you’ve probably got some combination of strial atrophy, high-frequency hair cell loss, and a bit of neural dropout. The audiogram is a snapshot of the cumulative damage.
The relative importance of each type is also still debated. A 2020 study of human temporal bones by Wu and colleagues at Massachusetts Eye and Ear, published in the Journal of Neuroscience, found that hair cell loss explained age-related threshold shifts better than strial damage did, which challenges the older view that metabolic presbycusis is the most common form.
What matters clinically is whether the loss is symmetrical (usually metabolic or genetic) or asymmetrical (raise suspicion for acoustic neuroma, Ménière’s, or other pathology requiring imaging), and whether word recognition matches the pure-tone thresholds. A 50-year-old with a 40 dB loss in one ear and 50% word scores is referred for an MRI. A 75-year-old with symmetrical high-frequency loss and proportional word scores gets hearing aids and cardiovascular counselling.
Contributing Factors Beyond Candles on the Cake
A 2017 Deloitte Access Economics report for the Hearing Care Industry Association estimated that 3.6 million Australians had some degree of hearing loss, with prevalence rising sharply with age. But age is only one variable. Others include:
- Noise exposure: Occupational (construction, manufacturing, military) and recreational (motorcycles, firearms, live music).
- Genetics: Some families lose hearing at fifty, others at ninety. DFNA mutations can present as progressive adult-onset loss.
- Ototoxic medications: Gentamicin, cisplatin, high-dose aspirin, loop diuretics. Always relevant in oncology and ICU survivors.
- Cardiovascular disease: The cochlea’s blood supply comes through a single small artery with no backup route, so interruptions to blood flow can cause permanent damage.
The same report put the cost of hearing loss in Australia at $33.3 billion a year once productivity losses and lost wellbeing are counted alongside direct financial costs.
What This Means for You
If you’re over sixty and noticing you need the TV louder, or you’re struggling in group settings, the first step is a diagnostic hearing assessment with pure-tone and speech audiometry. That tells us what you’ve lost. Tympanometry and bone-conduction testing check for middle-ear problems such as fluid or a stiffened ossicular chain, and speech-in-noise testing tells us how much functional difficulty you’re facing.
If it’s metabolic or sensory presbycusis, well-fitted hearing aids programmed with real-ear measurement usually help significantly. If there’s a neural component, we set realistic expectations and optimise what we can. And if it’s asymmetric or progressing rapidly, we refer for imaging and ENT review.
If you’re wondering whether what you’re experiencing is “normal ageing” or something else, the answer is: both can be true, and both deserve proper assessment.
