13Mechanism: Cochlear Damage and Deafferentation

FThe initiating event: losing the messengers

The cochlea turns sound into nerve impulses. Outer hair cells (OHCs) amplify and sharpen the travelling wave, while inner hair cells (IHCs) hand the signal to the auditory nerve. Noise, ageing and ototoxic drugs damage this machinery, and the consequence that matters for tinnitus is deafferentation — a fall in the steady stream of nerve impulses reaching the brain [2017].

Crucially, the brain does not simply go quiet. Deprived of its normal input, central auditory neurons begin to fire on their own. This is the cochlear-deafferentation hypothesis: a peripheral injury lights the fuse, but the noise the patient hears is generated centrally [2010].

THidden hearing loss: synaptopathy with a normal audiogram

Damage need not show on a pure-tone audiogram. Kujawa and Liberman demonstrated that a noise dose causing only a temporary threshold shift can permanently destroy up to roughly half of the ribbon synapses between IHCs and auditory-nerve fibres, with delayed loss of the spiral-ganglion neurons themselves — yet thresholds recover to normal [2009]. The high-threshold, low-spontaneous-rate fibres that encode sound in noise are preferentially lost.

This cochlear synaptopathy, or “hidden hearing loss,” is a compelling substrate for tinnitus in patients whose audiogram looks pristine: afferent input is reduced even when the threshold test cannot see it [2017].

TCentral takeover: hyperactivity and synchrony in the DCN

After acoustic trauma, recordings in the dorsal cochlear nucleus (DCN) and inferior colliculus show two changes: elevated spontaneous firing rates and increased neural synchrony, neurons firing together rather than independently [2005]. Noreña and Eggermont showed that this hyperactivity can appear within hours of trauma, implicating a rapid loss of inhibition rather than slow structural rewiring [2003].

A brain that reads firing rate as “sound present” will interpret this aberrant baseline as a tone or hiss — the percept of tinnitus [2010].

CWhy loudness does not track the audiogram

A central paradox: tinnitus severity correlates poorly with the depth of measured hearing loss. Patients with mild loss can be tormented; patients with profound loss may be untroubled. If the cochlea were the sole generator, loudness should track threshold loss — it does not [2010].

The resolution is that the cochlea initiates while central plasticity maintains and amplifies. Once central circuits reorganise, the phantom outlives, and is decoupled from, the peripheral lesion that started it [2017]. This is the clinical rationale for not promising that a hearing test predicts how loud a patient’s tinnitus will be.

CPredicting tinnitus pitch from the audiogram edge

The deafferentation idea makes a testable prediction. Schaette and Kempter modelled how homeostatic compensation for reduced input produces a peak of central hyperactivity at a particular frequency — and showed that the predicted frequency, and hence the dominant tinnitus pitch, falls near the edge of the audiogram, where hearing transitions from near-normal to impaired [2009].

Clinically this explains why high-frequency sloping losses so often produce a high-pitched tinnitus matched to the steep part of the curve, and it links the peripheral lesion quantitatively to the perceived sound [2004].