By Nancy Selfridge, MD
Dr. Selfridge is Professor, Clinical Foundations Department, Ross University School of Medicine, Barbados, West Indies
Dr. Selfridge reports no financial relationships relevant to this field of study.
• Music therapy — represented primarily by daily listening to Mozart’s “K.448” — has been shown in several studies to reduce seizure frequency in both adult and pediatric patients with drug-resistant epilepsy, although studies have been hampered by methodologic flaws.
• In this well-constructed randomized controlled trial with a crossover design, patients with drug -resistant epilepsy listened daily to either Mozart “K.448” or a cleverly constructed, non-rhythmic audio control, recording seizure frequency in a diary.
• Seizures were reduced significantly in the Mozart treatment group compared to both baseline, pre-study seizure experience and during the control exposure period in both intention-to-treat and per-protocol analyses.
SYNOPSIS: A randomized, controlled, crossover trial in adult patients with drug-resistant epilepsy compared listening to a Mozart piano sonata daily to an active control. It showed reduction in seizures during the Mozart treatment compared to both baseline and the control treatment.
SOURCE: Rafiee M, Patel K, Groppe DM, et al. Daily listening to Mozart reduces seizures in individuals with epilepsy: A randomized control study. Epilepsia Open 2020;5:285-294.
The “Mozart Effect” was coined after a 1993 publication in Nature reported temporary improvement in spatial task performance by college subjects after exposure to brief periods of listening to a Mozart sonata for two pianos (“K.448”).1 Since then, many studies have investigated the “Mozart Effect” as a potential intervention for cognitive disorders, mood disorders, and other psychiatric diagnoses, as well as neurologic disorders.2
A 1998 publication by Hughes et al documented significant seizure and electroencephalogram epileptiform activity reduction in patients with epilepsy listening to Mozart “K.448,” even in status epilepticus and coma. A similar therapeutic effect was not associated with listening to a pop piano piece.3
This work subsequently inspired significant research activity evaluating this Mozart piece therapeutically for refractory epilepsy. In 2014 and 2018, systematic reviews of studies over the previous 15 to 20 years investigated the efficacy of Mozart as a treatment tool for epilepsy. They concluded that a therapeutic effect appeared likely, but studies to date had been hampered by methodologic flaws, the most common of which was lack of an active control.4,5 It makes sense that any control condition other than an alternative style or piece of music would be challenging.
In this well-designed study, investigators implemented a randomized, controlled, crossover trial that addressed some of the methodologic flaws of the previous work. Based on its consistent application in extant research, Mozart's “K.448” was chosen as the active treatment. Researchers then created an active control from the same musical piece, subjected to phase-scrambling that preserved the auditory frequency and amplitudes of “K.448” but without any rhythmicity.
The auditory interventions were limited to the first “allegro con spirito” movement (six minutes) of “K.448” (24 total minutes) for practical purposes. They hypothesized that there would be no difference in seizure frequency between active treatment and control periods. Participants were selected from adult patients with epilepsy previously admitted to an epilepsy monitoring unit at Toronto Western Hospital. Inclusion criteria consisted of:
- dissatisfaction with seizure control despite drug therapy;
- at least three seizures in the three-month “baseline” period preceding the study and at least one seizure within the past two months.
Exclusion criteria were extensive and included:
- any history of brain resection surgery;
- any changes in seizure medication during the one-year study period;
- use of vagus nerve or deep brain stimulation or a ketogenic diet at the time of enrollment or during the study period;
- inability to recall or document seizure episodes in a diary;
- inability to understand or speak English;
- a score of less than 55% on a pitch perception and hearing impairment online test administered prior to starting the intervention.
Of 1,107 subjects assessed for eligibility, 13 met study inclusion criteria and were randomized to either Group A (7 participants) or Groups B (6 participants). After a baseline period of three months, allowing documentation of seizure occurrence, Group A and B began treatment. Both groups experienced three months of once-daily Mozart listening and three months of “scrambled” Mozart listening; Group A began with the Mozart piece and Group B began with the “scrambled” control. After three months, the groups switched interventions.
A three-month follow-up period allowed subjects to assess seizure control after the six months of interventions. Subjects documented seizures in a diary throughout the one-year study period. Auditory files for the interventions could be accessed via a password-protected website that also would prompt subjects to complete a survey question after listening, allowing some assessment of adherence.
Subjects who did not choose to use the website for access to audio files were provided an electronic copy of the sound files and then requested to indicate any missed days of listening in their diaries. Primary outcome was seizure frequency as indicated by diary entries.
Two subjects from Group A dropped out of the study, one because of a change of medication and one because of dissatisfaction with seizure control during the control intervention. Therefore, results included both per-protocol and intention-to-treat analyses of data. Baseline period seizure numbers for each subject were used to normalize paired observations of recorded seizures during the treatment and control periods. Effects sizes were calculated from mean differences between these normalized, paired observations.
Results are summarized in Table 1. Seizure count was reduced for subjects in both groups during the treatment period compared to control period for both intention-to-treat analysis (-44% for Group A, -25% for Group B) and per-protocol analysis (-59% for Group A, -25% for Group B). Paired t tests confirmed a significant treatment effect and Cohen’s d values indicated a large effect size (> 0.8). An average 35% reduction in seizure counts resulted from daily listening to Mozart's “K.448” compared to listening to a scrambled version, devoid of any rhythmicity, of the same piece.
Table 1. Summary of Results*
|
| |
Intention-to-Treat Analysis
|
Per-Protocol Analysis
|
| |
Group A (n = 7)
|
Group B (n = 6)
|
Group A (n = 5)
|
Group B (n = 6)
|
|
Mean seizure count reduction
(treatment period compared to control period)
|
-44%
|
-25%
|
-59%
|
-25%
|
|
Treatment effect
(paired t test results)
|
P value = 0.0005
t value = 4.75
|
P value = 0.0009
t value = 4.61
|
|
Paired Cohen’s d (95% confidence interval)
|
1.5 (0.7 to 2.1)
|
1.6 (0.7 to 2.5)
|
|
*Compliance rates: Group A 83 ± 11%; Group B 72 ± 16%
|
COMMENTARY
A significant strength of this study is that it addresses the absence of an active control present in previously published, quasi-experimental study designs. The sample size was ultimately small because of the numerous exclusion criteria, but the crossover design allowed valid statistical inferences that could not be made in a parallel study design with so few subjects. The intervention certainly was reasonable and manageable, supported by high compliance rates (83 ± 11% and 72 ± 16% for treatment and control periods, respectively).
The authors suggested that the lack of a washout period between treatment and control periods ignores a potential carryover effect and represents a weakness of the study that should be addressed in future research, along with optimum treatment time and “dosing” for maximum and/or sustained effect.
Experts continue to debate what element of the musical intervention is responsible for its positive therapeutic effect. Hypotheses include a parasympathetic effect from perceived pleasurable experience, brain wave entrainment caused by specific sound frequencies and amplitudes, or physiologic changes induced by auditory rhythms.2 Interestingly, despite the proposed hypothesis that the therapeutic effect attributed to this Mozart piece may be the result of physiologic effects of listening to pleasurable music in general, similar therapeutic effects for other musical pieces has not been found, except for one other Mozart piano sonata (“K.545”).6,7
Some experts have continued to argue that the nearly universal human experience of Mozart compositions as pleasing may yet explain the therapeutic effect of this music. However, a study of rats demonstrated improved maze running performance when exposed to Mozart “K.448,” whereas no improvement in maze running occurred when these rats were similarly exposed to Beethoven’s “Für Elise,” suggesting that Mozart’s works are unique in this regard.8
Is brain activity then “entrained” by the specific sound frequencies in these Mozart works? One study noted a seizure-reducing effect of the “K.448” piano piece that was not noted for a string version of the same piece, suggesting a therapeutic effect because of sound frequencies.9
However, the cleverly scrambled “K.448” control employed by Rafiee et al suggests that the therapeutic effect of listening to Mozart's “K.448” is not exclusively the result of sound frequencies and amplitudes, which were identical in this study’s treatment and control interventions — a noteworthy contribution to this debate.
In summary, based on the positive results reported in this study and while awaiting further research, practicing clinicians can consider recommending this no-cost/no-risk intervention to patients with epilepsy who are not content with their seizure control on prescribed medication.
REFERENCES
- Rauscher FH, Shaw GL, Ky KN. Music and spatial task performance. Nature 1993;365:611.
- Pauwels EKJ, Volterrani D, Mariani G, Kostkiewics M. Mozart, music and medicine. Med Princ Pract 2014;23:403-412.
- Hughes JR, Daaboul Y, Fino JJ, Shaw GL. The “Mozart effect” on epileptiform activity. Clin Electroencephalogr 1998;29:109-119.
- Dastgheib SS, Lavegh P, Sadeghi R, et al. The effects of Mozart’s music on interictal activity in epileptic patients: Systematic review and meta-analysis of the literature. Curr Neurol Neurosci Rep 2014;14:420.
- Brackney DE, Brooks JL. Complementary and alternative medicine: The Mozart effect on childhood epilepsy-a systematic review. J Sch Nurs 2018;34:24-37.
- Lin LC, Lee MW, Wei RC, et al. Mozart K.545 mimics Mozart “K.448” in reducing epileptiform discharges in epileptic children. Evid Based Complement Alternat Med 2012;2012:607517.
- Lin LC, Chiang CT, Lee MW, et al. Parasympathetic activation is involved in reducing epileptiform discharges when listening to Mozart music. Clin Neurophysiol 2013;124:1528-1535.
- Aoun P, Jones T, Shaw GL, Bodner M. Long-term enhancement of maze learning in mice via a generalized Mozart effect. Neurol Res 2005;27:791-796.
- Lin LC, Lee WT, Wu HC, et al. Mozart “K.448” and epileptiform discharges: Effect of ratio of lower to higher harmonics. Epilepsy Res 2010;89:238-245.
