8Anxiolytics and Anticonvulsants

FThe GABAergic rationale and its limits

One influential model of tinnitus pictures it as the brain’s response to lost input: deprived auditory pathways become hyperexcitable, with a tilt away from inhibition. That logic invites drugs which boost inhibitory GABA signalling or dampen neuronal excitability — benzodiazepines and anticonvulsants.

The trouble is that a plausible mechanism does not guarantee a clinical effect, and the trials that exist are small and inconsistent [2013]. None of these drugs is licensed to treat tinnitus, and guidelines do not recommend them for routine use [2014].

TBenzodiazepines: short-term help, long-term harm

Benzodiazepines such as clonazepam and alprazolam potentiate GABA-A receptors and have, in some small trials, been associated with reduced tinnitus annoyance — an effect that is hard to separate from their anxiolytic and sedative action. Their real-world value is therefore narrow: brief use to break a crisis of distress, panic or sleeplessness when tinnitus has become overwhelming.

Against this sit serious downsides. Tolerance develops, dependence is common, cognitive blunting and falls are risks in older patients, and abrupt withdrawal can transiently worsen tinnitus and anxiety. For these reasons long-term benzodiazepine use for tinnitus is discouraged, and any course should be short, planned and tapered [2013].

TGabapentinoids: the negative trial that matters

Gabapentin and pregabalin are GABA analogues that modulate voltage-gated calcium channels and are widely used for neuropathic pain. Early case reports of benefit in noise-trauma tinnitus raised hopes. The decisive test was the randomised, placebo-controlled trial by Piccirillo and colleagues, which found gabapentin no better than placebo for idiopathic subjective tinnitus [2007].

That result, set against a strong placebo response in tinnitus trials, is why gabapentinoids are not recommended for routine tinnitus. A pragmatic exception is the patient whose tinnitus coexists with genuine neuropathic pain or anxiety, where the drug is prescribed for that indication, not for the percept.

CCarbamazepine and typewriter tinnitus

There is one striking niche where an anticonvulsant earns its place. ‘Typewriter’ tinnitus is a paroxysmal, staccato, clicking or Morse-code-like sound, usually unilateral, attributed to neurovascular compression or microvascular irritation of the auditory nerve — a cochlear analogue of trigeminal neuralgia. Unlike continuous tonal tinnitus, it often responds dramatically to carbamazepine [2023].

Recognising this pattern matters because it changes management entirely: a rapid response to low-dose carbamazepine is both treatment and a confirming diagnostic clue. Outside this specific paroxysmal syndrome, carbamazepine has no established role and carries notable risks (hyponatraemia, marrow suppression, serious skin reactions), so it should not be used for ordinary continuous tinnitus.