Effects of familiar voices on brain activity


  • Yuji L Tanaka RN PhD,

    Corresponding author
    1. Associate Professor, Department of Physiology and Biochemistry, Chiba University, Graduate School of Nursing, Chiba, Japan
    Search for more papers by this author
  • Yumi Kudo RN MA

    1. Doctor Candidate, Course on Studies of Public Affairs, Chiba University, Graduate School of Humanities and Social Science, Chiba, Japan
    Search for more papers by this author

Yuji L. Tanaka, Department of Physiology and Biochemistry, Chiba University Graduate School of Nursing, 1-8-1, Inohana, Chuo-ku, Chiba-city, Chiba 260-8672, Japan. Email: yuji@faculty.chiba-u.jp


This study aimed to examine the extent to which a familiar voice influences brain activity. Participants were nine healthy female volunteers aged 21–34 years old (with a mean age of 25.78 ± 4.04 years). Brain activity was recorded during periods of silence, familiar and unfamiliar voices. Electroencephalographic data were collected and analyzed using a frequency rate set at 5 min. To account for emotional influences imbedded into the contents of the voice stimuli, both the voice of a familiar family member and the voice of a stranger were used to record a well-known Japanese fairy tale, ‘Momotaro’. Results revealed that listening to familiar voices increased the rate of the β band (13–30 Hz) in all four brain areas (F3, F4, C3 and C4). In particular, increased activity was observed at F4 and C4. Findings revealed that in study, participants' familiar voices activated cerebral functioning more than unfamiliar voices.


Contact and communication with others is a vital part of human awareness and development. Patients who are hospitalized, especially those in intensive care units (ICUs) with critical conditions, are for the most part in an isolated environment deprived of the necessary human contact. In order to facilitate patients' physical and psychological recovery, some ICUs have employed various methods including playing music, playing recordings and messages from loved ones and extending visiting hours for friends and family.

Accordingly, numerous researchers have examined the acoustical environments in hospitals and have demonstrated that several aspects are related to the investigation on noise and countermeasures aimed at reducing noise.1–7 However, these studies suggest that the theoretical support for examining hospital environments is seldom utilized with nursing interventions in mind. A recent study on patients in vegetative states from severe traffic accidents or stroke showed residual cognitive functioning and conscious awareness.8 These results suggest that it is possible for unconscious patients to recover and restore their levels of consciousness with longer periods of treatments. In contrast, previous studies show that there are particular sounds that may make patients feel more comfortable and enhance the quality of life.9–15 There are limited research effects of familiar voice on the level of consciousness.

Thus, the present study investigated the influence that familiar voices and sounds exert on brain activity. More specifically, we examined differences between baseline electroencephalography and brain activity during the administration of audio recordings using both familiar and unfamiliar voices. Brain activity was analyzed from the following four frontal and central brain sites: left frontal area (F3), right frontal area (F4), left central area (C3) and right central area (C4).



Nine healthy adult women with a mean age of 25.78 ± 4.04 years (age range, 21–34 years) volunteered to participate in the present study. The participants were recruited via a poster advertisement that was placed in one university in Chiba, Japan. The participants had no known neurological or otherwise underlying disorder that may influence data collection or outcomes. Additionally, the women were instructed (i) to refrain from drug use, excessive drinking and vigorous exercise 1 day before the experiment and (ii) to get a good night's rest, so that they would be adequately prepared for the experiment.

Ethical consideration

Following our university's research protocols for graduate student research, all of participants were briefed on the details of the experiment and written consent was obtained from each prior to the start of the study. They were informed that they could withdraw from the study at any time during the experiment. Also they were full apprised that all data and information would be kept confidential and anonymous. The volunteers were paid for their participation.

Experimental environment

Experiments were performed in a quiet, dark and shielded room where electric waves and external sounds were sufficiently blocked and removed. In addition, the room temperature and humidity were controlled and set at approximately 24°C and 50%, respectively.

Voice stimuli

Subjects were instructed to listen to voice recordings of a well-known Japanese fairy tale, ‘Momotaro’,16 delivered by either their own mother or an unfamiliar voice (a stranger). Participants' mothers were recruited to tape record the stories in order to avoid any additional emotional influence from stories being read aloud. Additionally, individuals not acquainted with the women were also recruited to tape record the stories and instructed to read slowly, include only a few intonations and reduce their emotions to the story. The acoustic pressure of the voice hearings was set at 55 dB.

Experimental procedure

Equipment for the present study included the electrocap (E1-M; Electrocap International Inc., Eaton, OH, USA) for electroencephalographic (EEG) data collection and Walkman earphones (SONY Co. Ltd, Tokyo, Japan) for the presentation of the voice stimuli. Participants were instructed to sit on a chair and close their eyes while sitting in a dark experimental room. EEG was recorded for a total of 30 min with the following breakdown: (i) a period of silence (6 min; SP1); (ii) a period including either a familiar or unfamiliar voice (6 min; F or U), followed by an additional period of silence (6 min; SP2); (iii) a period of unfamiliar or familiar voice (6 min; U or F); and (iv) a final period of silence (6 min; SP3) (Fig. 1). In order to avoid order effects, the presentation order for the voice periods was counterbalanced. Thus, five participants listened to a family member's voice first (hereafter, family early group) and four participants listened to the voice of a stranger first (hereafter, others early groups).

Figure 1.

Experimental procedure. Electroencephalographic (EEG) recording. Shaded areas indicate the period for EEG analyses (5 min, respectively). Recorded sessions: SP1, silent period 1; SP2, silent period 2; SP3, silent period 3; F, familiar voice; U, unfamiliar voice.

EEG recording and analysis

Administration of EEG recording electrodes and recording method

EEG activity was recorded from a total of 16 sites using an electroencephalograph (EEG-4518; Nihon Koden Co. Ltd, Tokyo, Japan) via an electrode cap based on the International 10–20 System for Electrode Method and Placement (Fig. 2). The Hi-cut and Lo-cut filters were set at 30 Hz and 0.3 s, respectively.

Figure 2.

The International 10–20 system electrode application. For the present study, electroencephalographic data from four recording sites (F3, F4, C3 and C4) were analyzed. F3: left frontal area, F4: right frontal area, C3: left central area and C4: right central area. NASION: craniometric point at the bridge of the nose where the frontal and nasal bones of the skull meet. INION: craniometric point that is the most prominent point at the back of the head (at the occipital protuberance).

EEG analysis

EEG data were analyzed using A/D converted output signal obtained from the electroencephalograph and collected directly onto a personal computer (PC-9821; NEC Corporation (old English company name; Nippon Electric Company, Ltd.), Tokyo, Japan). The analysis program Application Program for Topographic Mapping (ATAMAP; Kissei Comtec Co. Ltd, Matsumoto, Japan) was used for mapping EEG signals using a band pass filter set at 2.0–30 Hz. After the data were collected, fast Fourier transformation (FFT) analyses were conducted.

The data were analyzed using 5 min intervals after the first minute from each of the 6 min periods were excluded because the first minute of each sections may be influenced by previous conditions (see shaded areas of Fig. 1). EEG power spectral analyses were conducted using a total power value. EEG bandwidth was divided into the following six bands: δ (2–4 Hz), θ (4–8 Hz), α1 (8–10 Hz), α2 (10–13 Hz), β1 (13–20 Hz) and β2 (20–30 Hz). The power values for each of the bandwidths were calculated in proportion to the total power value and presented using a content ratio percentage (%). For the present study, EEG data from four recording sites (F3, F4, C3 and C4) were assessed and analyzed with regard to the projecting pathway hypotheses where auditory stimuli are said to influence the frontal areas of the brain (see Fig. 2).17–19 Thus, an increase in the content ratio within the beta band (13–30 Hz) would be indicative of brain activity.

Statistical analysis

The statistical analysis program SPSS version 12.0J (SPSS Co., Ltd, Tokyo, Japan), was used to conduct all analyses. All data were presented as mean ± standard error of the mean. To assess the association in presentation order between the two groups (mother's voice and unfamiliar voice), the data were analyzed using a two-way analysis of variance (ANOVA) and tests for mean differences in values were also conducted. Moreover, significant differences between both groups were analyzed using paired t-tests. Significance levels were set at *P < 0.05.


Change in baseline activities in EEG

Analyses of the EEG data revealed that subjects' brain activity exhibited lower amplitudes and faster waves while listening to a voice period vs. listening to a period of silence. Additionally, there were clear differences in EEG activity when subjects listened to their mother's voice vs. an unfamiliar voice.

Differences in presentation order

We examined whether an increase in the content ratio within the β band (13–30 Hz) was related to a difference in presentation order. The total power value while listening to each voice period in both the familiar and unfamiliar early groups was assumed to be 100%. Additionally, tests for mean differences in values for the content ratio within the β band were analyzed. Results from the nine subjects revealed that the content ratio within the β band was higher in all sites, except in C4, when subjects in the unfamiliar early group listened to both voice periods. No significant differences were observed in presentation order; therefore the data from the nine subjects were combined.

Comparison between familiar and unfamiliar voices

Comparison of content ratios within the β band

Comparisons in β band (13–30 Hz) frequency content ratios between the two voice periods (i.e. mother's voice vs. the unfamiliar voice) within all four brain sites (F3, F4, C3 and C4) revealed that the content ratios were statistically higher while listening to a familiar voice vs. an unfamiliar one. More specifically, the content ratios for the mother's voice condition were significantly greater in the F4 and C4 sites (with F4: mother's voice condition = 16.19% and unfamiliar voice = 12.60%; and in C4: mother's voice = 17.55% and unfamiliar voice = 13.86%) (Fig. 3).

Figure 3.

Comparisons of β wave content ratios in four brain sites. F3, F4, C3 and C4 indicate the recording sites. F3: left frontal area, F4: right frontal area, C3: left central area and C4: right central area. The white portion of the columns represents β1 wave (13–20 Hz) contents; the black portion of the columns represents β2 wave (20–30 Hz) contents. Asterisks indicate statistically significant differences (P < 0.05). F: familiar voice, U: unfamiliar voice. □, β1; inline image, β2.

Additionally, for subsequent analyses, the β band (13–30 Hz) was divided into its two respective portions: β1 band (13–20 Hz) and β2 band (20–30 Hz). Subsequent comparisons examining the content ratios within the four sites (F3, F4, C3 and C4) revealed a significant increase in brain activity while listening to a familiar voice (a mother's voice) specifically in the β1 band at the F4 site (Fig. 3) and in the β2 band at the F4 and C4 sites (Fig. 4).

Figure 4.

Comparisons of β2 band (20–30 Hz) content ratios in the four brain sites. F: familiar voice, U: unfamiliar voice. inline image: F3; left frontal area, inline image: F4; right frontal area, inline image: C3; left central area and inline image: C4; right central area. Asterisks indicate statistically significant differences (P < 0.05).

Comparison of presentation order

For comparisons of presentation order in β2 band frequencies in the four sites (F3, F4, C3 and C4), the content rate during SP1 for each subject was set and standardized at 100%, then the content ratio for the other voice periods was accordingly normalized (Fig. 5). In the F4 and C4 sites, there was a tendency for the content ratios to increase while the subjects were listening to the familiar voice (during the mother's voice condition), even when the presentation order was altered.

Figure 5.

Temporal differences in β2 band (20–30 Hz) content ratios illustrated by presentation order. SP1, silent period 1; SP2, silent period 2; SP3, silent period 3; F, familiar voice; U, unfamiliar voice. The presentation order in (a) is SP1–familiar voice–SP2–unfamiliar voice–SP3 and in (b) SP1–unfamiliar voice–SP2–familiar voice–SP3. inline image: F4; right frontal area, inline image: C4; right central area.


Presentation order effects

We thought that the influence from the first voice period (see Fig. 1; between SP1 and SP2) was stronger than the influence from the presentation of the second voice period (see Fig. 1; between SP2 and SP3). Additionally, the degree of interest in the content was higher during the first voice period. This may be due to habituation effects from the repetition of the content during the second voice period. Therefore, if a familiar voice such as the voice of a mother and an unfamiliar one show the same effect on brain activity, then the influences on brain may be larger from the first voice period rather than the second voice period (within an experimental environment).

However, listening to the sound of a mother's voice regardless of presentation order indicated tendency that the content ratio of the β2 band was increased when listening to a familiar voice such as one's mother (Fig. 5). There were no differences in the content ratio of the β2 band for either the first or second familiar voice period in two brain sites (F4 and C4) where EEG data were collected and analyzed. Thus, it was acknowledged that the content ratio of the β2 band was smaller when listening to an unfamiliar voice whether it was presented first or second. Results suggested that listening to a familiar voice such as a mother's voice may be an effective way of activating brain activity. Encouraging the use of familiar voice may in turn help the recovery of comatose patients. For example, a study conducted on an 8-year-old boy who nearly drowned and was left in a vegetative state for 4 years prior to the study demonstrated that there were significant differential activations in EEG frequencies at 14–58 Hz, with specific brain activity at 33.2 Hz (gamma band) to the sound of his mother's voice when compared with the voices of unfamiliar women.20 An additional study analyzed electrophysiological data on how newborns process their mother's voice compared with the voice of a stranger and found that exposure to one's maternal voice elicited early language-relevant processing, whereas the stranger's voice elicited more voice-specific responses.21 In an experiment examining event-related potentials to audio of a subject's own name (SON) recorded by a familiar voice vs. an unknown voice that was recorded by an unfamiliar person, the SON elicited large response amplitudes during the later phase of a novel P3 (after 300 ms).19 A large parietal component for familiar voice was observed, which suggested a higher level of information processing, even when the subjects were not required to explicitly differentiate between the two voices.22

Comparison between listening to a familiar voice vs. listening to an unfamiliar voice

Increase in the content ratios of the β band

Research has shown that the frequency of β waves increases while engaging in thought or when psychologically excited.23,24 Moreover, β waves appear during tasks requiring mental activity as well as when sensory stimuli are administered and received. Then, why is it that the content ratio of the β band increases when listening to one's mother's voice?

Even when participants heard the same fairy tale, there were two reasons that may have influenced the content rate of the β band by listening to their mother's voice vs. a stranger's voice. As previously suggested by others, firstly, it may be easier to focus on the contents of a story while listening to a familiar person's voice vs. a stranger's voice.20 Secondly, it seems that the brain may be activated by things that are not directly related to the contents of the fairy tale, for example, thoughts about their family, and events that may have occurred among their family members, such as recalling a memory connecting their family to the fairy tale.

The effects on the right hemisphere

Why do the F4 and the C4 sites in the right hemisphere activate to a mother's voice? In general, the left hemisphere is related to language processes and functions involving theoretical thoughts, whereas the right hemisphere is more involved in creative thinking and artistic creation and memory. Additionally, the right hemisphere can be involved in spoken language, but its ability to process spoken language is not as strong as the left hemisphere.

These brain processes within the left and right hemispheres may provide explanations why the right hemisphere is activated when listening to a familiar voice. Firstly, after processing the content of the fairy tale within the left hemisphere, the function to create the scenes within the fairy tale and the appearance of characters' speech and behaviour in detail might be further promoted by the right hemisphere. Secondly, it is thought that the association between the subjects and their families may be processed within the right hemisphere.

Additionally, the right superior temporal sulcus (STS) is said to be involved in the processing of human voices.25 By using functional magnetic resonance imaging (fMRI), experiments have found that the right anterior and posterior STS was activated more during voice recognition when compared with the recognition of verbal content. While the right anterior STS responded equally to both voice categories, the right posterior STS displayed stronger responses to unfamiliar voices when compared with familiar voices. During speaker recognition tasks, analyses examining the connectivity in psychophysiological interactions revealed that both the anterior and posterior right STSs interacted with the mid/anterior portion of the right STS, which is a region that has been previously implicated in the processing of the acoustic properties of voices. Moreover, the anterior and posterior STS regions that process different properties from voices interact in a specific manner depending on the familiarity with the speaker.

Significance of listening to a familiar voice

Most intensive care nurses report that verbal communication with unconscious patients is critical to their recovery; however, some ambiguity still remains regarding the level of awareness exhibited by unconscious patients.26 A previous study examined the influence of music vs. voice messages on 30 unconscious patients and showed that voice messages had a stronger effect on the patients' vital signs and facial expression when compared with the effects from music.27

The present study's results demonstrated that the stronger effects observed to the mother's voice may be due to the fact that it is easier to focus on and understand the content of a story when it is told by a familiar voice (such as a mother's voice). Alternatively, as previous studies have suggested, it is also possible that the subjects' brains were activated by hearing a familiar voice due to an association or recollection of a fond memory between the subject and their families.28 Furthermore, it is assumed that one would be more relaxed and comfortable when listening to a familiar voice. Perhaps, familiar voices convey something above and beyond the content of a story and ignite an emotional or creative response in the brain, thus activating the brain. Globally, the layout and organization of ICUs vary by hospital. Thus, future studies should closely examine the mutual trust, closeness, environment and other attributes between unconscious patients and their families.

Limitations of the study

The limitations of this study are the following: (i) the sample was small (n = 9); (ii) participants were conscious; (iii) the setting was not an ICU, but experimental room; and (iv) all were female. Furthermore, we used the mother's voice for familiar in this study. Therefore, a limitation is that the study did not describe the differences between gender, culture and age. Although, these findings cannot be generalized to the ICU or other cultures, they illuminate the potential for research on the effect of familiar voice on the consciousness of patient.


In conclusion, results from this study suggest that familiar voices may have different influence when compared with unfamiliar voices, which may serve as an additional recovery tool for critically ill patients particularly those in comas. Nurses in clinical practice should explore evidence-based interventions that promote recovery of unconscious patients. When culturally appropriate, fostering verbal and tactile communication between patients and close family members may be a useful strategy to hasten recovery or provide inner peace for the comatose.


We wish to express our deepest gratitude to the volunteers and their families for participating in the present study.


The authors declared no conflict of interest.