Musical hallucinations are a particular and uncommon type of auditory hallucinations, in which the patient perceives instrumental sounds, melodies or songs.
Auditory hallucinations are mostly restricted to patients suffering from schizophrenia.1
However, there is a growing body of evidence to suggest that musical hallucinations widely occur in older persons, with a female predominance, no history of psychiatric disease and a frequent association to acquired hearing loss. In the literature reviewed, it also seems that there is a high correlation between hearing loss and cognitive impairment, mainly described in the memory components and executive function.2
Depression is also present in almost one third of the cases.3
Musical hallucinations are generally described as stereotyped and often familiar motifs, including religious, childhood and popular songs, continually repeated. Despite their music content, these hallucinations can be very disabling.
The biological basis of this phenomenon remains largely unknown. Since a history of acquired hearing impairment is often present, one of the most convincing hypotheses is that musical hallucinations are due to sensory deprivation.4
Drug-induced musical hallucinations have been also analyzed. Antihypertensive treatments, metoprolol, misuse of ergot alkaloids, salicylate, tramadol hydrochloride, morphine and a combination of benzodiazepine have been supposed to be the trigger.5
Musical hallucinations have been observed also in focal brain disease, epilepsy and general brain atrophy, but those alone have never been described as sufficient causes.6
We present two cases of musical hallucinations in outpatients with visuospatial impairment, with a detailed neuropsychological assessment; a brain fludeoxyglucose-positron emission tomography (FDG-PET) showed the involvement of medial temporal regions and occipital associative areas.
T.D. 81-year old (patient #1) and B.I. 78-year old (patient #2) are right-handed women with no previous history of dementia or psychiatric pathology that were sent by their family doctor for auditory hallucinations symptoms.
Both report to hear instrumental music as played by an orchestra with a predominance of violins. Patient #1 reported that her hallucinations consisted of folk-song melodies that she could repeat, as they had been real; sometimes she heard choirs. Patient #2 complained that her hallucinations had presented as the same repetitive music at the onset; otherwise she had been hearing different songs since few weeks, sung by choirs. In both patients, hallucinations were constantly heard in both ears during the day but their intensity increased at night and when surrounding noise level was low. They could not been interrupted and interfered with their sleep causing disturb.
T.D. was retired before coming to our attention. She had elementary education and after she worked as chef.
In her medical history she had been treated for carotid artery disease (left carotid artery stenting in 2001) and breast fibro adenomas. She had heart bypass surgery and now she’s taking statins and antiplatelets. She also had hypertension and hypothyroidism, which were treated with diuretics and levotiroxin. There was no relevant family history for psychosis or dementia, but there was high familiar cardiovascular risk.
Before sending the patient to our attention, the family doctor tried to treat the musical hallucinations with neuroleptics (promazine, haloperidol and quetiapine) without any benefits.
B.I. was also retired before coming to our attention and she had elementary education. She worked her entire life as beautician in a beauty center.
She has diabetes mellitus treated with insulin from 17 years and minimal eyes impairment; she had been treated for carotid artery disease (right carotid endarterectomy in 2013). Apart from insulin, her routine therapy consists only in antiplatelets.
For the musical hallucinations she never received medications before coming to our center.
Remarkably similar cognitive and neuropsychological pattern was found in both patients, even if the degree of deficits was different: T.D. performed worse than B.I.
The mini mental state examination7 showed mild cognitive impairment for T.D. while B.I. seemed to have normal scoring.
The neuropsychological profile of the patient #1 is characterized by difficulties in attention, in long-term visuospatial memory, in visual-perceptual functions and praxis component. Specifically, she obtained a score at the inferior limits of normality in short-term numerical and verbal memory, but the visuospatial ability to repeat sequences was normal (Table 1).8-16
The long-term verbal memory was considered normal in the tests with textured material as the brief story presented and in delayed recall of material not related, while the immediate re-enactment could be considered at the lower limits of normality.
The attention was deficient in all its components; executive functions were normal.
The language was characterized by good phonemic fluency, instead a worst performance in the semantic component. Language comprehension was totally normal. Visuospatial and perceptual functions were impaired; clinically relevant were the performances at clock drawing test.15
The tasks of praxia indicated visuopercettive disorder (in addition to the previously described visuospatial disorder) and deficits in apraxia ideo-motoria. The ability for logical reasoning was normal.
The performances of the patient #2 were globally better than patient #1. The B.I. cognitive profile was characterized by disorder of the visual-spatial, perceptual and praxis component; it was reported spatial planning disorder.
In particular, the performances for the memory were altogether normal: good short- and long-term performance, whether with textured material or unrelated words.
The attention appears normal, executive functions were normal, but she had marked difficulty in spatial planning emerged in the evidence of Elithorn’s perceptual maze test.9
The language was characterized by good phonemic fluency, instead a worst performance in semantic component. Language understanding was globally normal.
The tasks of praxis showed disorder visuopercettive and visuospatial deficits in combination with the ideo-motor apraxia.
The ability of logical reasoning could be considered normal, with low score.
Since depression is often associated with musical hallucinations, neuropsychological assessment investigated also the presence of subclinical depressive symptoms with neuropsychiatric inventory (NPI).17
In both patients depression was detected and was characterized by continuous low mood and sadness, feeling tearful and hopeless about future.
The disorders of constructive praxis were similar in both patients. Deficit in copy tests traditionally are associated with parieto-occipital lesions and the impairment of ideomotor praxia is associated with parietal lesions, sometimes with involvement of the thalamus and basal ganglia.
Worst performances in semantic component of language tests were also comparable in the two patients. This difficulty is usually associated with temporal lesions. To evaluate these anatomical correlations the patients went to detailed radiological study.
Cerebral magnetic resonance findings showed generalized brain atrophy and multiple white matter lacunar lesions affecting above all the cerebellum; T.D. brain imaging in particular revealed bilateral enlargement of the temporal horn.
Furthermore cerebral FDG-PET showed heterogeneous cortical metabolism. In both patients hypometabolism was more pronounced in posterior regions. In particular the medialinferior temporal cortex and the occipital associative areas were affected. Interestingly in both patients occipital associative areas involved were homolateral to previous carotid surgery (Figures 1 and 2).
Electroencephalograms (EEGs) were unremarkable.
In addition the presence of acquired hearing impairment was evaluated, because it might be associated to musical hallucinations. Both patients reported a gradual and progressive hearing loss. Audiometric testing revealed bilateral mild-moderate sensorineural and conducting hearing impairment.
Cases presented by our patients conformed closely to almost stereotyped pattern musical hallucinations consisting of folk- song music played by an orchestra and sang by a choir.
There’s still disagreement about the possible anatomical substrate of the phenomenon. Within this context our reported cases might suggest that musical hallucinations have been arisen from sensory distortions due to a combination of peripheral and central dysfunction.18 A further explanation might be that musical hallucinations result from multiple white matter lacunar lesions due to small vascular events.19
Brain atrophy and lacunar lesions might worsen these sensory deprivation. In support, our patients suffer from neurosensorial-conductive hearing impairment. No EEG alterations, therefore, have been detected.
In contrast to most published cases our patients’ hallucinations were unfamiliar to them. This observation might support a model in which musical hallucinations are based on musical experience but not necessarily reproduced entirely from musical memory.20
The complex music and their definite tonality, large scale structure and timbre are consistent with an origin in higher-order auditory areas in the antero-medial temporal or frontal lobes.21 In both our patients medial-inferior temporal cortex and the occipital associative areas were affected.
It is also remarkable that occipital associative areas impairment was homolateral to previous carotid disease surgery, our patients have undergone to. To the best of our knowledge, no data on musical hallucinations associated to this brain imaging findings have been reported in scientific literature. Neither data are available on the effects of carotid artery surgery on occipital areas functioning. Moreover, in both patients neuropsychological pattern suggested a visuospatial-executive deficit, consistent to the occipital involvement.
The anatomical features and the neuropsychological assessment of the two patients could suggest a form of Levy body dementia, but the criteria to diagnose this pathology are not satisfied at the moment.22
Both patients were treated with citalopram in order to reduce the depressive symptoms. Recently completed randomized controlled trial23 on novel or repositioned drugs (mibampator, dextromethorphan/quinidine, cannabinoids, and citalopram) showed some promise in treating agitation in Alzheimer’s disease, but still with safety concerns.
In the subsequent visits we saw some benefits from the therapy, T.D. reported better mood and reduced sleeping disorder and B.I. describe herself as more hopeful for the future and less anxious, moreover we observed a minimal reduction of her complaining for the musical hallucinations, reported also by her family.
After few mouths, during the follow-up visits, T.D. increased the behavioral disturb, which were evaluated with NPI. In order to control these symptoms the patient is now treated with low dosage of haloperidol.
A crucial question is whether musical hallucinations might be primarily associated with occipital areas hypometabolism and visuospatial alterations or might be just an epiphenomenon of the occipital areas involvement typically associated with LBD. Understand the physiopathology connected to this uncommon clinical manifestation is important in order to find a therapeutic options for these patients. Further studies are needed to reach that goal.