Many regions of the brain are activated when listening to music or playing a musical instrument, as demonstrated by studies using functional magnetic resonance imaging (fMRI) and positron-emission topography (PET) [1–4]. This activation is identified by increases in blood volume relative to other brain regions. However, neither fMRI nor PET can provide information regarding absolute changes in blood volume or blood flow. Global cerebral blood volume and cerebral blood flow (CBF) can be stable if increases in blood volume and flow in activated regions of the brain are offset by slightly decreased in blood volume and flow in other regions that are not activated. The effect of playing a musical instrument on absolute changes in global CBF has yet to be determined.

Playing a musical instrument can increase global CBF. Middle cerebral artery (MCA) blood velocity (MCAv), which reflects a wide range of CBF, increases by up to 20% with physical activities such as cycling exercise by [5]. Playing a musical instrument has a much lower metabolic burden [6, 7] but is associated with more complex muscle activity than such physical activity. Complex movements require an interaction between local activation and brain interregional sensorimotor communication. Moreover, complex motor tasks involve larger regions of the brain with increased regional CBF than do simple motor tasks [8]. Thus, playing a musical instrument can increase CBF as much as physical activity.

Transcranial Doppler (TCD) flowmetry is a non-invasive method of assessing absolute changes in blood flow velocity in large arteries, such as the MCA in humans, in various situations and at a high sampling rate [9, 10]. TCD is inexpensive, safe, and useful for intracranial pressure monitoring [11]. It is generally accepted that the CBF measured by TCD reflects the change in global CBF. It has been demonstrated that during exercise, the change in MCAv measured by TCD is comparable to the change in mean blood flow in its upstream artery, the internal carotid artery [5]. TCD can record cerebral blood flow while musicians play music in normal circumstances. Also, PET and fMRI produce noise, whereas TCD generates no sound during the examination. Subjects can hear normal performance feedback of the sound they make.

The present study used TCD flowmetry to determine the absolute change in MCAv in participants while they were playing a musical instrument. The contributions of candidate factors underlying changes in CBF were explored, including blood pressure (BP) and arterial carbon dioxide partial pressure (PaCO₂), which are known to alter CBF [12]. We hypothesized that CBF increases when playing a musical instrument, as it does during exercise.

The 13 musicians who participated in this study were 3 males and 10 females aged 30±11 years, among whom 3 were violinists and 10 were pianists.

Resting MBP values were very similar when playing, irrespective of the difficulty of the score that was about to be played, being 86±12, 88±10 and 84±11 mmHg for the FS, MS and PR trials, respectively (Fig 1). Significant increases in MBP were observed in the FS (90±13 mmHg) and PR (89±13 mmHg) trials, but not in the MS trial.

There was also no significant difference in resting MCAv in subjects before playing the FS, MS and PR scores (0.54±0.09, 0.53±0.1 and 0.52±0.1 m/s, respectively; Fig 1). MCAv significantly increased when playing the MS (0.55±0.09 m/s, 8.6%) and PR (0.56±0.14 m/s, 5.2%) scores, but not when playing the FS score.

The only significant change in CI was observed when playing the MS score, when there was a significant decrease. P_ETCO₂ and oxygen consumption (VO₂) showed no significant changes with the score difficulty.

Analysis of relative changes in MBP versus MCAv revealed no association between the two parameters for the FS, PS and MS scores (Fig 2).

The main finding of this study was a 5–9% increase in MCAv in subjects when playing an instrument, except when playing a score for the first time (i.e., FS). The increase was observed in most subjects. There were no significant concomitant increases in metabolic demand or ventilation. These results suggest that playing an instrument increases CBF with no observable increase in metabolic demand.

The extent of the increase in MCAv was less than that previously been reported to occur during physical exercise. The degree of increase in MCAv during exercise ranges from 10% to 20%, with MCAv increasing by 20% when VO₂ increases to four times the resting level [5]. During exercise, MCAv continues to increase until VO₂ reaches 70% of its maximum [14–16]. In the present study, the extent of increase in CBF associated with playing a musical instrument was much lower than that observed during exercise. CBF increased during instrumental playing, even though there was no increase in VO₂. Therefore, the increase in MCAv during musical instrument playing was not due to metabolic processes.

There are two possible explanations for the lack of increase in MCAv in the FS trial. First, because of the difficulty inherent in playing music for the first time, the subjects chose music with a much lower level of difficulty than would be appropriate for their skills. For example, a subject who played Frederic Chopin’s Heroic Polonaise (Op. 53 in A-flat major) on the MS test played Gustav Lange’s Heather Rose (Op. 78, No. 3 in C major) on the FS test. According to the Japanese music publisher ZEN-ON MUSIC, Heroic Polonaise is ranked F (advanced upper) and Heather Rose is ranked A (beginner). Since they were professional musicians, the subjects in the present study would have found the FS score easy to perform. When a professional pianist requires bimanual coordination, the recruitment of motor networks in the brain is lower than that for non-musicians [1]. The subjects could perform complex bimanual movements more easily, which may be related to a reduced activity in the brain, possibly resulting in a subtle increase in MCAv. Second, the frequency of finger tapping was low because the subjects chose an easier FS music score. The frequency of fingering of both hands increases with increasing difficulty of the music. A relative increase in blood volume in the primary motor cortex observed by near-infrared spectroscopy was much larger in a maximum-effort finger-tapping task than in a low-frequency finger-tapping task [17]. The low frequency of finger tapping that occurs when playing an FS score may result in only a small degree of regional brain activation.

The MCAv measurements made during playing a musical instrument in the present study have two advantages over previous studies that have demonstrated activation in various brain regions. First, we targeted a large blood vessel and measured absolute changes in blood flow velocity. Studies using fMRI and PET have examined the occurrence of increases in blood volume in target regions of the brain relative to other brain regions while playing a musical instrument [1–4]. We have confirmed an increase in blood flow velocity in absolute value. Second, in the present study the subjects played their instruments under conditions close to those of a normal performance environment, whereas in previous fMRI and PET studies the subjects played fake instruments in a restricted environment. For example, a keyboard compatible with performing fMRI has been applied, but the subjects used only one hand to play it and/or only music scales were played rather than a full music score, subjects played their instrument while in a supine position, and the performance was done in loud extraneous noise conditions from fMRI and PET devices [18, 19]. These factors are likely to have affected the subjects’ performance and brain neural activity. The findings of the present study are more likely to reflect the normal CBF response to playing music, since the subjects were able to play their instrument in a normal playing environment.

The physiological factors underlying the increased MCAv observed during playing are controversial. Arterial pressure, PaCO₂, autoregulation, neural activity and metabolic relate all impact MCAv [20]. The increase in MBP may contribute to the increase in MCAv observed in the present study since there was no increase in CI, which reflects vasodilatation. However, its contribution to MCAv was not a clear and at least non-linear, since no relationship was shown in Fig 2. P_ETCO₂ did not change and thus did not contribute to the increase in MCAv. The contribution of MBP and vasodilatation to the increase in MCAv was not simple nor clear.

MCAv involves blood flow to regions relating to instrumental performance. Instrumental performance involves sensory, motor, and auditory senses, which are distributed throughout the brain as integrated regions [4, 7]. This network consists of the dorsolateral, inferior frontal cortex (including Broca’s area), superior temporal gyrus (Wernicke’s area), marginal superior gyrus, supplementary motor area, and premotor area [18]. All of these sites receive blood via the MCA. Thus the MCAv in the present study reflects blood flow in these regions.

The present study did not provide consistent results for increases in MCAv and MBP. This is partly due to the fact that the music performed were not strictly controlled in nature. The subjects themselves selected the music according to the difficulty level, since the subjects’ performance abilities varied. Replacing it with exercise, the intensity of the exercise was not steady and uniform. The degree of increase in MCAv and MBP varies with exercise intensity [21]. When the increase in MCAv and MBP accompanied with playing musical instruments is examined in terms of hand and arm exercises, it is necessary to match the music for the trials. In turn, we successfully provided the responses in all tasks, implying roles of visual processing of reading music score, and the act of listening to the sound, the motor activity of playing. Some kind of technique is needed to clearly describe effects of playing a musical instrument on MCAv and MBP.

MCAv as measured by TCD has been evaluated previously as an index of CBF. MCAv can be treated as CBF when the diameter of the artery does not change. It has been reported that the MCA diameter measured during craniotomy or by nuclear magnetic resonance imaging does not change when physiological stimuli (e.g. arterial BP and PaCO₂) are applied [22–24]. Even during exercise, changes in MCAv measured by TCD are comparable to changes in mean blood flow in the upstream artery, the internal carotid artery [5]. It is therefore reasonable to assume that MCAv indicates CBF in the present study.

Control data were not collected in this study so as to reduce the time burden on the subjects. However, given the early recovery of most variables back to the resting recorded before playing, it is reasonable to conclude that the observed changes were true responses to playing a musical instrument.

A limitation for this study is that Playing violin and piano are different in motor commitment. Further investigation using a larger sample size solely in violin is needed for a more definitive clarification.

Playing a musical instrument increases MCAv, at least when complex music was played, with the degree of increase being partially dependent on the difficulty of the piece being played. Previous studies using fMRI and PET have revealed regions of increased blood flow in the brain when playing a musical instrument; however, these studies did not assess CBF in the whole brain. To the best of our knowledge, this is the first study to find a clear impact of playing a musical instrument on CBF.