‘I’ve got an iPod in my head’

To the Editor

Musical hallucinations are a rare phenomenon with a prevalence of 0.16% in the psychiatric inpatient setting (Fukunishi et al., 1998). Hearing loss, psychiatric illness, structural brain lesions, and epileptic foci have been commonly associated with musical hallucinations, which tend to onset late in life (mean age 61.5±19.4 years) and have a female predominance of 2:1 irrespective of aetiology (Evers et al., 2004). Musical hallucinations have not been previously reported with childhood onset in the scientific literature.

Mr V is a 30-year-old man with childhood onset musical hallucinations. Mr V’s musical hallucinations began at age 12, 2 years after an intracranial haemorrhage secondary to pineal choriocarcinoma. The haemorrhage resulted in hydrocephalus, rendering him comatose for 4 weeks. Mr V underwent biventricular peritoneal shunt insertion, followed by chemotherapy and whole brain irradiation. He recovered with no recurrence, but was left with mild intellectual disability.

At age 21, Mr V experienced a series of focal motor seizures and was commenced on carbamazepine with incidental attenuation of his musical hallucinations. He experienced no further seizures and carbamazepine was ceased 5 years later. He remained anti-epileptic free for 4 years, with progressive worsening of his musical hallucinations.

Mr V was subsequently referred to psychiatrists at age 29 with increased distress and sleep disturbance secondary to musical hallucinations. On presentation, Mr V experienced near-continuous musical hallucinations with onset upon waking, persistent throughout the day. The hallucinations consisted of repeating loops of song fragments ranging from a few bars to whole songs. The songs were familiar to him, and ranged from childhood and popular contemporary songs to mobile phone ringtones and simple repetitive beats, such as the ticking of a clock.

The hallucinations occurred independently of his will and were heard as if through internal speakers, controlled by an autonomous source, which Mr V likened to having ‘an iPod in my head’. The onset of musical hallucinations could be spontaneous or triggered by a true auditory musical perception, a fragment of which would repeat and replace the existing song fragment loop. Auditory or visual cues with associations to music were also frequent triggers. The musical hallucinations were not distractible, persisted simultaneously with conversation and caused initial and middle insomnia, recurring with light arousal from sleep.

Mr V experienced no other perceptual disturbances and had no delusions. He was not clinically depressed. Mr V had no personal or family history of psychiatric illness. His MMSE was 28/30. Neurological examination was unremarkable, except for strabismus, and there was no evidence of hypacusis.

On investigation, magnetic resonance imaging showed mild generalized atrophic features greater than expected for age, but was otherwise normal. No mass lesions were seen and bilateral-frontal ventricular shunt tips were seen in the third ventricle without ventricular dilation. Technetium-99m HMPAO single-photon emission computed tomography brain scan and electroencephalogram were normal.

Mr V was commenced on sodium valproate (500 mg b.d. titrated to 1700 mg in divided doses daily). Treatment response was seen after 1 month, with maximal response at 2 months, reflected by decreased musical hallucination duration (to 3 hours/day), less distress, and improved sleep as well as new distractibility with speech. However, after 3 months, there was a deterioration and plateau of anti-epileptic response. Gabapentin was subsequently added to augment sodium valproate treatment without significant benefit. Mr V continues to have near-continuous musical hallucinations distractible with speech. Overall, there has been only a partial response to sodium valproate. However, Mr V has adapted to the presence of musical hallucinations, the content of which is ego-neutral and at times even enjoyable.

The mechanism of musical hallucination is still unknown. A major theory is neuronal stimulation, where stored perceptual musical experiences are re-experienced though stimulation of the relevant neuronal circuitry by organic brain disease such as focal lesions and epilepsy (Keshevan, et al., 1992). A second major theory is that of deafferation, in which decreased afferent auditory stimulation such as in sensorineural deafness, in combination with reduced inhibition from higher centres, allows spontaneous activity in auditory association cortices, resulting in complex musical hallucinations with the salience of a true percept (Cope et al., 2009).

Mr V’s partial response to anti-epileptics and history of seizures supports the theory of an epileptiform process, with underlying predisposition from multiple organic brain insults including hydrocephalus, whole brain irradiation, and frontal biventricular peritoneal shunt insertion. An epileptiform process is also supported by phenomenological features of musical hallucination, which are shared with epilepsy including excitation, kindling, and self-perpetuation (Sacks, 2008) Mr V’s familiarity with all musical hallucination content supports the theory of musical hallucinations as a release of stored perceptual experiences.

This case highlights that musical hallucinations are somewhat distinct from other auditory hallucinations in psychiatry, particularly in the degree of organic contribution. There are similarities with epilepsy both in phenomenology and response to treatment. Clinicians should recognize anti-epileptics as a potential first line option in the treatment of isolated musical hallucinations. However, clinicians need to be aware that prognosis is somewhat guarded and musical hallucinations may persist as chronic near-continuous experiences. Nevertheless, many patients have shown a degree of adaptation to the experience of musical hallucinations.