Enhancing the regulatory function of the autonomic nervous system using sounds with inaudible high-frequency components

Abstract

The dysregulation of the autonomic nervous system (ANS) activity notably contributes to the onset and progression of numerous diseases, including lifestyle-related and psychiatric disorders. This necessitates the development of effective nonpharmacological methods for regulating ANS function for therapeutic purposes and disease prevention. This study examined how the presence or absence of the inaudible high-frequency component (HFC) of sounds—which activates deep-brain structures—affects the ANS regulatory function. Under the N-back task condition, which requires concentration, exposure to sounds with HFC resulted in significantly higher sympathetic and parasympathetic nervous activities compared to sounds without HFC. Conversely, under the relaxation condition, the sounds with HFC significantly suppressed sympathetic nervous activity relative to sounds without HFC. Therefore, sounds with HFC may flexibly adjust the sympathetic and parasympathetic nervous activities based on situational demands.

Introduction

The autonomic nervous system (ANS), comprising the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS), involuntarily regulates physiological functions to maintain homeostasis in response to environmental and situational demands. The dysregulation of ANS activity attributed to different factors—such as chronic SNS overactivation caused by aging, genetic predisposition, and physical and psychological stress—has been implicated in various physiological dysfunctions, particularly those related to the metabolic and circulatory systems¹. Moreover, ANS imbalance significantly contributes to the onset and progression of numerous diseases, including lifestyle-related and psychiatric disorders¹. Therefore, the development of effective nonpharmacological methods for regulating ANS function is promising for therapeutic purposes and disease prevention.

The current study focused on the use of the hypersonic effect, a phenomenon previously reported by our group^2–11, as a potential nonpharmacological intervention. The hypersonic effect refers to a set of physiological and psychological responses elicited by a hypersonic sound, containing complex high-frequency components (HFC) that are inaudible to humans, which exceeds the upper limit of human auditory perception (20 kHz). Our earlier findings showed that compared with sounds within the audible frequency range, exposure to hypersonic sounds significantly enhanced the alpha-band power in spontaneous electroencephalogram activity^3,4,6–8,10 and increases the regional cerebral blood flow of the deep-lying brain structures, such as the midbrain and hypothalamus^3,4,9. In addition, the anterior cingulate cortex and the medial prefrontal cortex, which are the regions receiving input from monoaminergic reward-related circuits, are activated⁹. These findings explain that hypersonic sound induces psychological effects that promote pleasant sensations and emotional responses^3,4,8 and behavioral effects such as approach behavior^4,7,10,11.

Notably, hypersonic sound significantly suppresses the rise in blood glucose levels during an oral glucose tolerance test—a key marker of glucose metabolism and diabetes risk⁵. Blood glucose homeostasis is regulated by complex endocrine pathways involving hormones such as insulin, glucagon, and adrenaline, which are further modulated by higher-order systems, including the ANS. Chronic overactivation of the SNS, often resulting from stress, can impair glucose regulation. Therefore, the observed glucose-lowering effect can be mediated by suppressing the SNS activity¹². Notably, this effect was more pronounced in older adults than in younger individuals⁵, indicating a potential age-related difference in ANS responsiveness to hypersonic stimulation.

Meanwhile, another study¹³ reported findings that contradict the previously mentioned⁵ suppressive effect of hypersonic sound on SNS activity. Specifically, exposure to hypersonic sound notably enhanced performance on the N-back task¹³, a cognitively demanding task requiring sustained attention and working memory. Therefore, SNS activation potentially contributes to successful performance on tasks involving high level of concentration and tension^14,15. For example, Vincent et al.¹⁴ administered a lexical recognition task to participants while simultaneously measuring physiological indices such as heart rate and T-wave amplitude. Indicators of SNS activity—specifically, increased heart rate and decreased T-wave amplitude—were positively associated with task performance. Therefore, SNS activation contributed to the execution of lexical recognition tasks. Similarly, Peccinenda and Smith¹⁵ performed a lexical memory task on participants and measured skin conductance (SC) as an index of SNS activity. Their results showed a positive correlation between SC and task performance. Altogether, SNS activation is involved in tasks requiring tension and sustained attention.

Taken together, these seemingly contradictory findings raise the possibility that hypersonic sound modulates SNS activity in a context-dependent manner. In particular, under conditions requiring relaxation, it may promote restfulness by reducing SNS activity. Meanwhile, in cognitively demanding contexts, it may facilitate arousal and task engagement via SNS activation. Therefore, hypersonic sound may enhance the adaptive regulatory capacity of SNS and PNS in accordance with situational demands, thereby representing a promising tool for ANS modulation.

The current study aimed to examine the hypothesis that hypersonic sound enhances the ANS regulatory function. Thus, two distinct task conditions (one requiring mental tension and concentration and another promoting relaxation) were designed. During each task, the participants were exposed to either hypersonic sound or placebo sound, which was acoustically identical except for the absence of inaudible HFC. The indices of the SNS and PNS activities were measured and compared to assess whether the effects of hypersonic sound varied based on the task context. Further, to explore potential age-related differences in responsiveness, the data were analyzed separately for the younger and older groups.

Methods

Participants

Forty healthy adults (22 women, 18 men) aged 21–66 (mean age: 44.0 ± 18.5) years participated in this study. Participants aged 18–70 years when the informed consent was obtained were included. However, individuals deemed incapable of providing informed consent, diagnosed with lifestyle-related diseases (e.g., diabetes, hypertension, and/or dyslipidemia) and undergoing pharmacological treatment, with disorders affecting the ANS, regardless of treatment status, and considered unsuitable for participation by the research director were excluded; additionally, we excluded all participants whose medications or health conditions could potentially influence ANS function. To examine age-related differences in ANS function, participants were categorized into younger and older groups, with equal numbers assigned to each group. This classification was based on a previous study examining the effects of hypersonic sound on glucose tolerance⁵. Consequently, the older group consisted of 12 women and 8 men (age range: 49–66 years; mean ± SD: 61.3 ± 4.72), while the younger group included 10 women and 10 men (age range: 21–48 years; mean ± SD: 26.8 ± 7.69).

Auditory stimuli and presentation system

Natural environmental sounds recorded in the primary forests of the Bornean tropical rainforest were used as auditory stimuli, consistent with a previous study⁵. These soundscapes comprised various natural sounds, including insect chirping, bird calling, and leaf rustling. Further, they were characterized by abundant HFC inaudible to humans, exceeding 20 kHz, with complex temporal fluctuations on the millisecond scale. The average frequency bandwidth extended up to 150 kHz, with peaks reaching 200 kHz. A 20-min segment was used for the experimental playback.

The following three auditory conditions were administered:

• Full-range sound (FRS): This was an unfiltered soundscape containing all frequency components. Based on a previous study, the hypersonic effect requires coexistence of the audible frequency range below 20 kHz and the inaudible HFC above 40 kHz that reach the body surface^3,6. The FRS satisfied this requirement.

• High-cut sound (HCS): This was a low-pass filtered version of the original sound with a cutoff at 20 kHz (attenuation:− 200 dB/octave). Hence, all components below this threshold were preserved, and inaudible HFC was eliminated, serving as a placebo sound condition for the FRS condition.

• No-sound (NS): This was background laboratory noise (e.g., air conditioning) that was presented without any added auditory stimulus to clarify any nonspecific effects to the presence of audible components.

Playback in the FRS and HCS conditions used the TASCAM DA3000 recorder (TEAC Corporation, Japan) with the 5.6448-MHz DSD format. Four OOHASHI Monitor Op.8 speakers (GAIA HSS-801, Action Research Co., Ltd., Japan), which is capable of an ultra-wideband output (20–120 kHz, − 10 dB), were used and supplemented by four super tweeters (HSST-01P, Action Research Co., Ltd., Japan; frequency response of 20–200 kHz). The setup aimed to replicate a natural rainforest acoustic environment, with speakers positioned at four points surrounding the participant (front-right, front-left, rear-right, and rear-left). The Supplementary Fig. S1 online shows the spectral power measurements at the ear level (height: 110 cm).

Task conditions

A cognitively demanding condition (N-back task) and a relaxation condition were the two task conditions applied to regulate ANS activity. Both tasks were 5 min in length.

N-back task condition: Participants were presented with a randomized sequence of four Japanese speech stimuli—ue (up), shita (down), migi (right), and hidari (left). The task was to decide for each stimulus whether it matches the one presented N items before¹⁶, specifically two items before in this study. Responses were recorded using “correct” or “incorrect” buttons operated with the dominant hand. Response time and accuracy were emphasized. Before the experiment, a training session was conducted to determine the optimal stimulus interval, aiming for an accuracy rate between 75 and 85%.

Relaxation condition: The participants were instructed to release physical tension, breathe slowly and deeply, and focus on the sensation of relaxation. Eye closure was not restricted. However, continuous closing was discouraged.

ANS assessment

To decrease measurement-induced stress, a low-intrusiveness protocol was adopted. Three indices were employed to assess the ANS activities. Details of the recording and analysis methods are provided in the Supplementary Information.

Heart rate variability: The high-frequency (HF; 0.15–0.4 Hz) component of heart rate variability (HRV) was used as a marker of PNS function^17,18. During the 5-min N-back task and relaxation conditions, the R-R interval (RRI) data were recorded. The HF components were obtained by calculating the power spectral density within a range of 0.15–0.4-Hz using the Fast Fourier Transform of the 5-min RRI time-series data.

Skin conductance: The tonic skin conductance level (SCL), derived from SC time-series data, was used as an index of the SNS function because it reflects general SNS activity¹⁹. The mean SCL over the 5-min period was calculated and used as the quantitative measure of SNS activity.

Skin temperature: The differential skin temperature (DST) between the nasal tip and the forehead, was used as a marker of the SNS function, as this helps account for environmental influences such as ambient temperature^20–23. SNS activation induces vasoconstriction, further reducing blood flow and consequently lowering nasal skin temperature^{20–22,24–26}. Conversely, forehead skin temperature is relatively unresponsive to SNS activation^{20–22,25–28}. The mean DST—calculated as nasal tip skin temperature minus forehead skin temperature—over the 5-min period served as the index of SNS activity.

Experimental procedure

The Ethics Committee of the National Center of Neurology and Psychiatry approved this study (approval no. A2023-061). All participants provided a written informed consent. The current study was conducted in accordance with the Declaration of Helsinki.

Each participant completed two experimental sessions on separate days, one for each task condition (N-back or relaxation). The order of conditions was counterbalanced across the participants. All sessions commenced at 10:00 AM to control for circadian influences. The participants were instructed to refrain from consuming caffeine and have adequate sleep prior to participation. Each experimental day comprised four sessions, with 10-min breaks between sessions. During the first three sessions, the FRS, HCS, and NS conditions were assigned in a counterbalanced manner. In the fourth session, the NS condition was consistently assigned. However, data from this session were excluded from the analysis. The inclusion of an additional NS session at the end, despite its exclusion from the analysis, aimed to decrease the potential confounding effects on ANS indices due to the psychological influences associated with the final session. In the N-back and relaxation conditions, a 4-min baseline rest period was implemented at the beginning to stabilize physiological signals and eliminate potential carryover effects from the preintervention state. No-sound stimuli were presented during this period. In the N-back condition, participants subsequently performed a 5-min 2-back task, during which sound stimuli were presented (in the FRS or HCS conditions) or not presented (in the NS condition). Conversely, for the relaxation condition, the initial 4-min baseline rest period alone could have induced a relaxation state, making it difficult to detect physiological changes attributable to further relaxation. Therefore, we first induced a mild arousal state by showing participants a 2-min emotionally arousing video clip (AKIRA, BANDAI VISUAL CO., LTD., Japan), followed by a 5-min relaxation period with sound presentation (in FRS or HCS) or without sound presentation (in NS).

To reduce environmental stress, plants were placed throughout the laboratory, and the equipment was arranged unobtrusively. The experimenter remained outside the room, observing the participants using audiovisual monitoring to maintain a nonintrusive presence. A separate glass-walled area served as a break room.

Statistical analysis

All analyses were performed using EZR (Jichi Medical University, Japan)²⁹, a graphical interface for R designed for biostatistics. Data from participants with missing values or frequent arrhythmias were excluded from the analysis. Normality was assessed using the Shapiro–Wilk test, and outliers were identified using the Smirnov–Grubbs test. The normality was not rejected in any analysis.

First, to examine whether differences in the task conditions set in this study influenced ANS indices and whether age-related changes in autonomic function were reflected in the measured indices, the mean values across the three sound conditions were calculated for each task condition and each participant. The participants (n = 40) were then evenly divided into two groups: the older group (n = 20) and the younger group (n = 20). For each ANS index (HF, SCL, and DST), a two-way repeated measures analysis of variance was conducted, with the task condition considered as a within-subject factor and the age group as a between-subject factor. The HF data were log-transformed to ensure normality.

Next, the effects of differences in sound conditions on each physiological index within each task condition were examined. The HF and SCL data were normalized by dividing the value for each sound condition by the mean value across the three sound conditions. Regarding DST, nasal tip skin temperature can be either higher or lower than forehead temperature, resulting in positive or negative DST values. Consequently, the mean DST value across the three sound conditions, which were used for normalization, could be negative or very close to zero. Unlike HF and SCL data, which are always positive and normalized by dividing by the mean value across the three sound conditions, DST required a different approach. Therefore, the mean forehead ST across all participants, task conditions, and sound conditions was first added to each DST value. Then, normalization was conducted by dividing the adjusted DST value of each sound condition by the mean of the three sound conditions. Repeated measures analysis of variance with the sound condition as a within-subject factor was conducted. Holm-corrected multiple comparisons were carried out for post hoc analysis between the sound conditions. In addition, to investigate the effects of age, similar analyses were performed individually for the older and younger groups. A p value less than 0.05 was considered statistically significant.

Results

Data validity and characteristics of the participants

The HF data of HRV obtained from 37 participants (20 women and 17 men) aged 21–66 (mean age: 43.6 ± 18.4) years were valid. Among them, 18 were assigned to the older group (9 women, 9 men; 49–66 [mean: 61.2 ± 4.79] years) and 19 to the younger group (11 women, 8 men; 21–48 [mean: 27.1 ± 7.81] years).

The SCL data of 39 participants (21 women and 18 men), who were aged 21–66 (mean: 44.2 ± 18.8) years, were valid. Of them, 20 were included in the older group (10 women, 10 men; mean age: 61.3 ± 4.72 years) and 19 in the younger group (11 women, 8 men; mean age: 26.2 ± 7.42 years).

The DST data of 35 participants (20 women, 15 men), who were aged 21–66 years (mean age: 46.5 ± 18.4 years), were analyzed. The older group included 20 participants (10 women, 10 men; mean age: 61.3 ± 4.72 years), and the younger group comprised 15 participants (10 women, 5 men; mean age: 26.8 ± 8.04 years).

Statistical analyses were conducted using the participant population described above. Outlier tests were performed for each analysis, and the number of participants included in the final analyses varied depending on the results of the outlier screenings. The number of participants who passed the outlier test and were included in each analysis is indicated in the corresponding figures and Supplementary Table S1 online.

Effect of the task condition and age group

Figure 1 shows the mean values and standard errors (SEs) of the HF, SCL, and DST under the N-back and relaxation task conditions stratified by age group. These data were analyzed to assess the effects of cognitive load and age on ANS function. In this context, higher HF values reflected a greater PNS activity. Meanwhile, higher SCL and lower DST values indicated increased SNS activity.

Fig. 1.

Mean values of the HF, SCL, and DST for each task condition (N-back and relaxation) across the older and younger groups. The error bars represent the standard error of the mean. A p value less than 0.05 was considered statistically significant. The blue and pink bars correspond to the N-back and relaxation conditions, respectively. The task condition exhibited significant main effects in all indices. In addition, the age group had significant main effects on the HF and DST. No significant interaction effects were observed between the task condition and age group for any of the indices.

The task condition had a significant main effect on all three physiological indices. In particular, the HF in the N-back task condition was significantly lower than that in the relaxation condition (F (1, 32) = 8.76, p = 0.00576), thereby indicating a reduced PNS activity under the cognitive load. In contrast, the SCL was significantly higher in the N-back task condition compared with the relaxation condition (F (1, 37) = 16.5, p = 0.000240), which was consistent with increased SNS activity. Similarly, a significant decrease in the DST was observed during the N-back task condition compared with the relaxation condition (F (1, 26) = 6.47, p = 0.0173), thereby further supporting heightened sympathetic arousal during the cognitive task.

The age group also exhibited main effects. The older group had significantly lower HF values than the younger group (F (1, 32) = 19.03, p = 0.000125), thereby indicating a reduction in baseline PNS activity with aging. A significant age-related difference was found for DST (F (1, 26) = 4.32, p = 0.0477), with the older group having lower values than the younger group, which suggests a higher baseline SNS activity. However, the age group did not have a significant effect on SCL (F (1, 37) = 0.103, p = 0.751), which indicated that the tonic SC did not differ significantly between the age groups.

In contrast, the interaction effects between the task condition and age group were not statistically significant for any of the indices examined. In particular, the interaction terms for HF (F (1, 32) = 0.920, p = 0.345), SCL (F (1, 37) = 2.20, p = 0.146), and DST (F (1, 26) = 2.13, p = 0.157) did not reach statistical significance. Based on these results, although task condition and age independently influenced autonomic responses, the patterns of task-related change were consistent across age groups.

Effect of the sound condition

Figures 2 and 3 present the mean values and SEs of the HF, SCL, and DST under each sound condition (FRS, HCS, and NS) for the N-back and relaxation conditions, respectively. The data of the whole group and the older and younger groups were presented, thereby allowing the examination of the influence of sound condition on SNS and PNS activity.

Fig. 2.

Mean values of the HF, SCL, and DST during the N-back task condition across the whole group, older group, and younger group. The error bars represent the standard error of the mean. A p value less than 0.05 was considered statistically significant. The blue, pink, and gray bars correspond to the high-cut sound (HCS), full-range sound (FRS), and no-sound (NS) conditions, respectively. For the HF, the significant main effects of the sound condition were detected in the whole and older groups. Post hoc comparisons with Holm’s correction indicated that the HF values were significantly higher in the FRS condition than in the HCS condition (whole group: p = 0.0425; older group: p = 0.0436). For SCL, the significant main effect of the sound condition was observed only in the older group, with higher values observed in the FRS condition than in the HCS condition (p = 0.0410). The sound condition did not have significant main effects on the DST in any of the groups.

Fig. 3.

Mean values of the HF, SCL, and DST during the relaxation condition across the whole group, older group, and younger group. The error bars represent the standard error of the mean. A p value less than 0.05 was considered statistically significant. The blue, pink, and gray bars correspond to the high-cut sound (HCS), full-range sound (FRS), and no-sound (NS) conditions, respectively. For DST, the significant main effect of the sound condition was observed in the whole group. Post hoc comparisons with Holm’s correction showed that the DST values were significantly higher in the FRS condition than in the HCS condition (p = 0.0100). In the older group, although the main effect of the sound condition did not reach statistical significance, the DST values were more likely to be higher in the FRS condition than in the HCS condition (p = 0.100). The sound condition did not have significant main effects on the HF or SCL in any of the groups during the relaxation condition.

N-Back Task Condition: As shown in Fig. 2, the sound condition had a significant main effect on HF in the whole group (F (2, 62) = 5.53, p = 0.00619) and in the older group (F (2, 26) = 6.56, p = 0.00494). Post hoc comparisons revealed that the FRS and NS conditions had significantly higher HF values than the HCS condition. In particular, in the whole group, the HF in the FRS (p = 0.0425) and NS (p = 0.00260) conditions was significantly higher than those in the HCS condition. However, the HF did not significantly differ between the FRS and NS conditions (p = 0.711). A similar pattern was observed in the older group, where the HF was significantly greater in the FRS (p = 0.0436) and NS (p = 0.00190) conditions than in the HCS condition. Nevertheless, there was no significant difference between the FRS and NS conditions (p = 0.512). The sound condition did not have a significant main effect on HF in the younger group (F (2, 32) = 2.79, p = 0.0763). Thus, the influence of HF sound components may be more evident in older adults than in younger ones.

The sound condition had a significant main effect on the SCL only in the older group (F (2, 34) = 3.36, p = 0.0467). Post hoc analysis showed that SCL in the FRS (p = 0.0410) and NS (p = 0.0230) conditions were significantly higher relative to those in the HCS condition. However, no significant difference was observed between the FRS and NS conditions in terms of the SCL (p = 0.753). In contrast, the sound condition did not have a significant effect on the SCL in the whole group (F (2, 74) = 1.62, p = 0.206) and the younger group (F (2, 36) = 0.423, p = 0.658).

The sound condition did not have significant main effects on the DST in any of the groups. In whole group, the DST did not differ significantly across the conditions (F (2, 60) = 0.0371, p = 0.964). Further, the DST did not have significant effects in the older (F (2, 34) = 0.266, p = 0.768) and younger (F (2, 26) = 1.33, p = 0.281) groups.

Relaxation Condition: As shown in Fig. 3, the sound condition had a significant main effect on the DST in the whole group (F (2, 62) = 6.11, p = 0.00379). Post hoc comparisons showed that the DST was significantly higher in the FRS (p = 0.0100) and NS (p = 0.0200) conditions than in the HCS condition. Meanwhile, there was no significant difference in the DST between the FRS and NS conditions (p = 0.900). In the older group, although the main effect did not reach statistical significance (F (2, 36) = 3.16, p = 0.0542), the DST was more likely to be higher in the FRS condition than in the HCS condition (p = 0.100), thereby suggesting a possible age-related sensitivity. Nonetheless, there were no significant differences in the DST in the younger group (F (2, 22) = 2.79, p = 0.0834).

The sound condition did not have significant effects on the HF during the relaxation condition in any of the groups. The F-values and p values were as follows: whole group, F (2, 70) = 0.0894, p = 0.915; older group, F (2, 32) = 1.33, p = 0.280; and younger group, F (2, 36) = 2.95, p = 0.0649. Similarly, the sound condition did not significantly affect the SCL in any of the groups, as evidenced by the following values: whole group, F (2, 70) = 2.84, p = 0.0652; older group, F (2, 36) = 1.12, p = 0.337; and younger group, F (2, 36) = 0.361, p = 0.699.

Discussion

Effect of hypersonic sound on the regulatory function of the ANS

To investigate the primary objective of this study—the effects of inaudible HFC—the FRS and HCS conditions were compared.

Regarding SNS activity, under the N-back condition, the older group showed significantly higher SCL, indicating SNS activation, in the FRS condition. Conversely, under the relaxation condition, the whole group showed notably higher DST, indicating SNS suppression, in the FRS condition. The older group exhibited a similar, although nonsignificant, trend. Regarding PNS activity, under the N-back task condition (Fig. 2), the HF values, an index of PNS activity, were substantially higher in the FRS condition than in the HCS condition in the whole group and the older group. In contrast, under the relaxation condition (Fig. 3), no significant difference in HF values between the FRS and HCS conditions was observed. These findings suggest that the effects of inaudible HFC (FRS condition) versus their absence (HCS condition) on ANS activity vary based on task demands. In particular, under conditions requiring tension and sustained attention (N-back task), the FRS condition enhanced the SNS and PNS activities. Meanwhile, under the relaxation condition, the FRS condition was more likely to suppress the SNS activity. These results support the study hypothesis that hypersonic sound regulates the SNS and PNS activities in a condition-dependent manner.

The underlying mechanisms should still be elucidated. However, previous studies have revealed that hypersonic sound increases cerebral blood flow to the hypothalamus, a key brain region responsible for regulating ANS output⁹. The hypothalamus integrates diverse physiological signals to maintain homeostasis and orchestrates the activities of the SNS and PNS pathways. Therefore, the context-dependent modulation of ANS activity observed in the current study may be attributed to the enhanced hypothalamic function caused by increased cerebral blood flow induced by the hypersonic sound.

Further, the absence of significant differences between the FRS and HCS conditions in the younger group and the evident effects observed in the older group are consistent with the findings of previous study showing that the glucose-suppressive effects of hypersonic sound were more evident in older individuals or those with elevated Hemoglobin A1c levels⁵. Considering that the regulatory function of the ANS declines with age³⁰, these results suggest that hypersonic sound may be particularly effective in individuals with impaired glucose tolerance or reduced ANS function. Moreover, the current findings imply that hypersonic sound may enhance the regulatory capacity of the ANS in populations with age-related or pathological decline and that it has minimal effects in individuals with preserved autonomic function.

In addition, under the N-back task condition, SNS and PNS indices were remarkably increased in the FRS condition than in the HCS condition. Hansen et al.³¹ reported that successful performance on tasks requiring tension and sustained attention involves coordinated activation of the SNS and PNS at the cardiac level. Furthermore, as noted in the Introduction, Peccinenda and Smith¹⁵ reported a positive correlation between performance on such tasks and SC, an index of SNS activity. Altogether, showing coactivation of PNS activity at the cardiac level and SNS activity as reflected in SC, our findings suggest that hypersonic sound may promote the ANS activity necessary for task performance and support cognitive engagement. The SNS and PNS are traditionally viewed as antagonistic systems. However, their simultaneous coactivation under certain physiological or psychological conditions indicates a more nuanced and dynamic regulatory model³². Under the relaxation condition, SNS activity substantially reduced under the FRS condition, as indicated by an increase in the DST. However, the HF, a PNS index, did not remarkably increase. One possible explanation is that the PNS activity may have already been sufficiently elevated during the relaxation condition, resulting in a ceiling effect that masked further enhancements.

Limitations in the interpretation of the results

The finding that the SCL and DST—both considered indices of the SNS activity—did not respond uniformly to the sound conditions should be further investigated. One plausible explanation lies in the distinct response characteristics of the SC and ST. As noted in a previous study, the SCL, derived from the SC, reflects the tonic component and is highly sensitive to transient sweating responses triggered by psychological stress, thereby serving as a rapid indicator of increased SNS activity¹⁹. In contrast, the sympathetic vasoconstrictor nerve activity, which plays a key role in ST regulation, maintains a steady level of vascular tone at rest³³. When SNS activity is suppressed, these nerves induce near-maximal vasodilation, making ST particularly sensitive to reductions in the SNS activity. These differences in response mechanisms may account for the finding that SNS activation during the N-back task condition was more sensitively captured by SCL, meanwhile, SNS suppression during the relaxation condition was more clearly reflected by ST changes. Another plausible explanation is that ST and SC are mediated through distinct neurophysiological pathways. Typically, postganglionic sympathetic neurons release norepinephrine that acts on target organs³⁴. The sympathetic vasoconstrictor nerves that regulate cutaneous blood flow—as reflected by ST—are primarily adrenergic and release norepinephrine³⁴. Conversely, sympathetic innervation of the sweat glands, contributing to SCL, is mediated by cholinergic neurons that release acetylcholine³⁴. This neurochemical difference may explain why SCL and DST did not respond uniformly to the sound conditions.

This study primarily aimed to investigate the effects of inaudible HFC by comparing the FRS and HCS conditions. However, under the N-back task condition, HF (a PNS index) and SCL (an SNS index) were significantly higher in the NS condition than in the HCS condition. Under the relaxation condition, the DST was significantly higher in the NS condition than in the HCS condition. Further, there were no significant differences between the FRS and NS conditions in any of these comparisons. These findings suggest that the differences observed between the FRS and HCS conditions may not be solely attributable to the presence of inaudible HFC in the FRS condition. Rather, they raise the possibility that the HCS condition—characterized by the presentation of only audible-range components—may suppress the ANS regulatory function. This finding is supported by previous studies that investigated the impact of HFC on brain activity, which similarly reported that the presentation of only audible-range sound (HCS condition) leads to reductions in alpha-band electroencephalogram power and decreased regional cerebral blood flow in the brainstem and thalamus compared with the FRS and NS conditions^3,4. The biological mechanisms underlying these effects remain unclear. However, we have previously reported that natural acoustic environments, such as tropical rainforests, are characterized by a rich presence of inaudible HFC, whereas modern urban environments lack such features³⁵. Further, the HCS condition, in which HFC were removed from the original sound source, is acoustically comparable to the audio quality of CDs and digital broadcasts, both of which are now widespread in modern urban environments as a part of conventional digital media. Considering these observations, it is essential to cautiously monitor and further investigate the potential physiological effects associated with inaudible HFC deficiency in contemporary acoustic environments.

Appropriateness of the task conditions and age group classification

As shown in Fig. 1, the task condition had a significant main effect across all measured indices. In particular, the HF, an index of PNS activity, was significantly lower during the N-back task condition than during the relaxation condition. Conversely, the SNS indices exhibited opposing patterns. Specifically, the SCL was significantly higher during the N-back task condition, and the DST was significantly lower. A higher HF value indicates a greater PNS activity. Meanwhile, elevated SCL and reduced DST values reflect an enhanced SNS activity. These results indicate that the N-back task induced a physiological state characterized by SNS dominance. Meanwhile, the relaxation condition elicited a PNS-dominant state, thereby showing that the two task conditions successfully elicited distinct patterns of the ANS activity.

Regarding age-related differences, the HF and DST exhibited significant main effects in the age groups, with both indices significantly lower in the older group than in the younger group (Fig. 1). This pattern indicates that older participants had a reduced PNS activity and an increased SNS activity relative to their younger counterparts. These findings are consistent with previous evidence showing that aging is associated with diminished parasympathetic cardiovascular regulation, increased sympathetic tone, and an overall decline in ANS regulatory capacity³⁰. Thus, the age group classifications and physiological indices used in the current study may be appropriate and effective for capturing age-related differences in the ANS function. In contrast, the SCL, which is another SNS-related index, did not exhibit a significant main effect in the age groups. Although DST and SCL are indicators of the SNS activity, as mentioned above, the distinct response characteristics and the distinct neurochemical pathways may account for the divergent age-related patterns observed between the two SNS indices.

Future directions

The current study showed the effects of hypersonic sound on ANS regulation. This finding can provide important insights into the mechanisms underlying the hypersonic effect previously reported in the literature. The activation of brain regions such as the brainstem and hypothalamus regulates a broad range of physiological systems via the ANS, including the visceral organs, body surface, and vascular system. In this context, our results suggest that hypersonic sound may be a foundation for future medical applications aiming to treat and prevent disorders associated with ANS dysregulation or dysfunction. Future research should investigate the clinical potential of hypersonic sound in different conditions, such as lifestyle-related and stress-induced disorders, which are characterized by impaired ANS regulation.

This research direction is in accordance with the emerging field of information medicine, which seeks to optimize environmental information input to regulate brain function^2,36. Considering that the brain operates not only as a chemically driven organ but also as an information-processing system responsive to internal and external stimuli, disruptions in the environmental information flow may impair neural processing and contribute to the development and progression of various psychiatric and neurological disorders. Information medicine represents a complementary approach to traditional pharmacological and surgical interventions, focusing on the information-processing aspects of brain function.

By elucidating how environmental information influences neural activity, this field aims to develop nonpharmacological strategies for the treatment and prevention of psychiatric and neurological conditions. As discussed in the previous text, natural environments such as tropical rainforests—where human genetic and neural systems are believed to have evolved—are characterized by an abundance of inaudible HFC. If human evolution shaped the genetic and neural architectures to process such environmental information and enhance survival under these conditions, then an acoustic environment rich in inaudible HFC may represent an optimal state for the brain’s information-processing mechanisms. In contrast, the modern urban soundscape, which is significantly devoid of these HFC, may constitute a significant deviation from this evolutionary ideal. In light of the findings related to the hypersonic effect, advancing the development of information medicine approaches that aim to supplement contemporary urban environments with hypersonic sound is essential^2,36,37.

Author contributions

K.J., N.K., and M.H. conceived and designed the experiment. N.K. and E.N. created the sound sources. K.J. acquired the data. K.J., N.K., and M.H. analyzed the data. K.J. and M.H. wrote the paper. All authors revised the paper.

Data availability

The datasets generated in this study will be provided upon reasonable request to the corresponding authors.

Declarations

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-11190-9.