6The Central Gain Mechanism (Homeostatic Plasticity)

FHomeostatic plasticity — the thermostat that overshoots

Every neuron tries to keep its average firing within a working range. When its input drops, it compensates by becoming more excitable — strengthening its remaining synapses and dialling down inhibition — so that a smaller signal still produces a normal-sized output. This is homeostatic plasticity, and across the auditory pathway it manifests as central gain: an amplification applied to whatever drive survives the damaged cochlea. [2011]

The logic is sound. The problem is what gets amplified. With the gain turned up, the system magnifies not only residual sound but also its own spontaneous activity — the random and not-so-random neural noise that is always present. Amplify the noise enough and it crosses the threshold of perception. The thermostat, trying to restore normal activity, overshoots into a phantom percept. [2014]

TA model that predicts tinnitus from the audiogram

The power of the central-gain idea is that it can be made quantitative. Schaette and Kempter built a computational model in which auditory-nerve activity is reduced (as in deafferentation) and a homeostatic gain term restores the mean firing of central neurons to its pre-injury set-point. The model predicts that spontaneous activity will rise specifically in the frequency region of greatest input loss — and that this is where tinnitus pitch should fall. [2006]

Tested against patients, the model did real predictive work: it could forecast the perceived pitch of an individual’s tinnitus from the shape of their audiogram, predicting it from the edge of hearing loss where the recovered gain is highest. [2009] The framework also explains tinnitus with a normal audiogram: if hidden synaptopathy reduces auditory-nerve drive without raising thresholds, homeostatic gain still rises and can still generate tinnitus — a prediction the model made explicit. [2011] [2012]

TThe phantom-limb analogy and why quiet unmasks it

Central gain reframes tinnitus as a member of a wider family of phantom percepts. After amputation, the somatosensory system — deprived of input from the missing limb — reorganises and amplifies, and the result can be a vivid sensation referred to a hand that is no longer there. [1995] Tinnitus is the auditory counterpart: a sensory system, robbed of its normal afferent drive, raises its gain and generates a percept with no external source.

This view neatly explains a complaint every clinician hears — that tinnitus is worse in quiet. In a quiet room there is little external sound to occupy the high-gain system, so the amplified internal noise stands out in stark relief. Add ambient sound and the same gain now has real input to work on; the phantom recedes into the background. This is the mechanistic basis of sound therapy and of the simple advice to avoid silence. [2011]

CHyperacusis — the over-gain cousin

If tinnitus is gain applied to silence, hyperacusis is gain applied to real sound. The same homeostatic up-regulation that makes the system hear its own noise also makes ordinary, moderate sounds feel uncomfortably or even painfully loud, because the recovered gain over-amplifies suprathreshold input. [2014] This is why the two conditions so often travel together: they are two faces of a single dysregulation of central gain.

Clinically, the central-gain model carries a clear and somewhat counter-intuitive corollary: starving the system further — with earplugs and avoidance of all sound — tends to drive gain up and can worsen both tinnitus and hyperacusis over time. The rational strategy is the opposite: graded, enriched, comfortable sound exposure to give the gain something to act on and to coax it back down. [2016] The same homeostatic machinery that caused the problem is, in principle, the lever for unwinding it.