**3. Results**

tion). The resulting image can record the condition of the iris and eyeball movement on the monitor. For example, when normal subjects are exposed to continuous light stimulation for 1 sec, we can obtain a pupillogram from the IRIS apparatus; constricted pupils indicate para‐ sympathetic activity, and pupil dilation indicates sympathetic activity. Pupil diameter in normal subjects is generally 2-5 mm, which changes under various adaptation conditions. We examined the transverse diameter and velocity of pupil constriction or dilatation after

**Figure 2.** IRIS apparatus (A and B) and typical data acquired following light stimulation (C and D). IRIS records the bilateral pupillary reflex simultaneously. IRIS records the constricted and dilated pupil reflex for one second after light stimula‐ tion. Pupillary reflex parameters (e.g., most constricted pupil velocity) can be quantified. Experimental schedule (E).

Figure 2 shows the IRIS experimental apparatus used in Figures 2A and 2B. The pupillary light-reflex test was executed using an infrared pupillometer (Iriscorder C10641, Hamamat‐ su Photonics Co.), and pupil diameter in both the right and left eyes was measured after 3 min of rest or stimulation. Figure 2E shows the timeline of this experiment. Five particular experimental conditions were explored over 5 days; we recorded HRV modulation during the pupillary reflex adaptation test. However, only the right pupil was exposed to light stimulation, as shown in Figure 4. The pupillary test is non-invasive and enables real-time diagnosis. We examined the initial diameter (D1), minimum diameter (D2), constriction ra‐ tio (CR), time to total construction (T3), maximum velocity of constriction (VC), and maxi‐ mum acceleration of constriction (AC) among the parameters obtained from the IRIS. Pupil diameter decreased with parasympathetic activity and increased with sympathetic activity. The IRIS records pupil parameters of the right and left eyes simultaneously. However, we adopted parameters from the right pupil because data from both sides were similar, as shown in Figure 4. We conducted these examinations with eight normal subjects (six males, two

vibrotactile stimulation to explore changes in autonomic activity.

358 Advances in Vibration Engineering and Structural Dynamics

## **3.1. Vibrotactile stimulation of the parotid and submandibular glands**

We examined differences between vibrotactile stimulation of the parotid and submandibu‐ lar glands. We found that the most effective frequency to induce salivation was 89 Hz-S regardless of whether vibrotactile stimulation was delivered to the parotid or submandibu‐ lar glands, as shown in previous paper (Hiraba et al. 2011).

Because patients with hyposalivation often have psychiatric disorders, we conducted an experiment to realistically approximate natural conditions. We examined whether effective salivation occurred continuously when vibrotactile stimulation was performed daily. Specif‐ ically, we used the 89 Hz-S frequency with a single motor from the previous experiment. None of the glands (i.e., right and left parotid glands or right submandibular and sublingual glands) showed a reduced response. Regression curves for each gland showed non-adaptation to continuous stimulation; instead they showed parallel or increasing curves, indicating that continuous use of this apparatus should not be problematic, as shown in a previous paper (Hiraba et al. 2011).

## **3.2. Relationship between stimulation and fNIRS activity**

The OEG16 spectroscope was used to record BBF haemoglobin concentration from areas in the frontal cortex using 16 channels. We determined the oxyHb concentration schema evoked by 89 Hz-S vibrotactile stimulation by analysing 16 channels. The results showed very weak oxyHb concentrations (i.e., near zero) during 89 Hz-S vibrotactile stimulation. Changes in oxyHb, deoxyhaemoglobin (deoxyHb), and total haemoglobin (totalHb) concentrations dur‐ ing salivation measurements at rest and for each vibrotactile stimulation frequency were measured. As shown in a previous paper (Hiraba et al. 2011), changes during the following six conditions were measured: (1) resting; (2) 89 Hz-S vibrotactile stimulation; (3) 89 Hz-D (89 Hz-D, 89 Hz frequency with double motors, 3.5-μm amplitude); (4) 114 Hz-S; (5) 114 Hz-D (114 Hz-D, 114 Hz frequency with double motors, 3.5-μm amplitude); and (6) "A-" phona‐ tion. Each wave was recorded for 3 min, and each 2-min vibrotactile stimulus is shown between the vertical lines (Fig. 6B and Fig. 4 in the previous paper, 2011). Although each wave meas‐ ured during resting salivation, at 114 Hz-D, and during "A-" phonation showed increased activity, the 89 Hz-D and 114 Hz-S vibrotactile stimuli decreased activity. However, vibrotac‐ tile stimulation at 89 Hz-S showed a value of almost zero. Particularly, when we focused on oxyHb changes based on these results, increased oxyHb occurred during "A-" phonation, the resting condition, and at 114 Hz-D vibrotactile stimulation, whereas a decrease in oxyHb was observed during vibrotactile stimulation at 114 Hz-S and at 89 Hz-D. However, oxyHb con‐ centration during vibrotactile stimulation at 89 Hz-D was almost zero, as were all other data (oxyHb, deoxyHb, totalHb). From these results, we computed oxyHb integral rates over 2 min, as shown by the area between the longitudinal bars (Fig. 6B and Fig. 4 in the previous paper, 2011).

**3.4. Pupillary reflex after vibrotactile stimulation**

right pupillary reflex for data analysis.

vibrotactile stimulation.

Among the parameters obtained from the IRIS, we examined D1, D2, CR, T3, VC, and AC, as shown in Figures 4 A and 4B. The pupillary light reflex showed significantly decreased D1, D2, and T3 compared with the resting state. Furthermore, the pupillary light reflex showed increased AC, as shown in Figure 4B. Data from the right and left pupils were simi‐ lar following light stimulation, as shown in Figure 4B. Thus, we employed data from the

Optimal Vibrotactile Stimulation Activates the Parasympathetic Nervous System

http://dx.doi.org/10.5772/51156

361

**Figure 4.** Typical example of data from the pupillary reflex. Right pupillary reflex data after light stimulation in the

**Figure 5.** Effect of the pupillary reflex following right-side light stimulation between the resting condition and 89 Hz-S

right pupil (A-a and B-a) and left pupillary reflex data after light stimulation in the left pupil (A-b and B-b).

Furthermore, we examined integral rates while subjects listened to classical music for 2 min. We divided the subjects into two groups: (1) one group that disliked listening to classical music and (2) one group that enjoyed listening to classical music. Although the subjects who enjoyed the music did not show a larger spread of values, the former did. Specifically, vibro‐ tactile stimulation at 89 Hz-S led to a small, similar value spread. All integral rates during the vibrotactile stimulation at 89 Hz-S and listening to classical music showed similar averages and standard deviations (SDs), as shown in Figure 6B.

## **3.3. Total salivation during vibrotactile stimulation**

During facial 89 Hz-S stimulation, values of total salivation and salivation in the parotid or submandibular and sublingual glands were examined in comparison with resting salivation values. Submandibular and sublingual gland total salivation values increased; however, parotid gland salivation values were similar, as shown in Figure 3. Parotid gland salivation values were 0.15 ± 0.12 m*l* on both sides during rest and 0.14 ± 0.12 m*l* under the 89 Hz-S stimula‐ tion condition. Salivation values of submandibular and sublingual glands on both sides was 0.79 ± 0.44 m*l* during rest and 1.00 ± 0.58 m*l* under the 89 Hz-S condition (*p*< 0.01; Wilcoxon singed-rank test, two-tailed). Total gland salivation values on both sides were 0.91 ± 0.55 m*l* during rest and 1.16 ± 0.60 m*l* under the 89 Hz-S condition (p < 0.01). Particularly, although parotid gland salivation was similar between the resting and 89 Hz-S conditions, that in the submandibular and sublingual glands showed an absolute increase between the resting and 89 Hz-S conditions, as shown in Figure 3.

**Figure 3.** Salivation during the resting condition and 89 Hz-S vibrotactile stimulation over 3 min. Wilcoxon signedrank test (two-tailed), *P*< 0.01.

## **3.4. Pupillary reflex after vibrotactile stimulation**

centration during vibrotactile stimulation at 89 Hz-D was almost zero, as were all other data (oxyHb, deoxyHb, totalHb). From these results, we computed oxyHb integral rates over 2 min, as shown by the area between the longitudinal bars (Fig. 6B and Fig. 4 in the previous

Furthermore, we examined integral rates while subjects listened to classical music for 2 min. We divided the subjects into two groups: (1) one group that disliked listening to classical music and (2) one group that enjoyed listening to classical music. Although the subjects who enjoyed the music did not show a larger spread of values, the former did. Specifically, vibro‐ tactile stimulation at 89 Hz-S led to a small, similar value spread. All integral rates during the vibrotactile stimulation at 89 Hz-S and listening to classical music showed similar averages

During facial 89 Hz-S stimulation, values of total salivation and salivation in the parotid or submandibular and sublingual glands were examined in comparison with resting salivation values. Submandibular and sublingual gland total salivation values increased; however, parotid gland salivation values were similar, as shown in Figure 3. Parotid gland salivation values were 0.15 ± 0.12 m*l* on both sides during rest and 0.14 ± 0.12 m*l* under the 89 Hz-S stimula‐ tion condition. Salivation values of submandibular and sublingual glands on both sides was 0.79 ± 0.44 m*l* during rest and 1.00 ± 0.58 m*l* under the 89 Hz-S condition (*p*< 0.01; Wilcoxon singed-rank test, two-tailed). Total gland salivation values on both sides were 0.91 ± 0.55 m*l* during rest and 1.16 ± 0.60 m*l* under the 89 Hz-S condition (p < 0.01). Particularly, although parotid gland salivation was similar between the resting and 89 Hz-S conditions, that in the submandibular and sublingual glands showed an absolute increase between the resting and

**Figure 3.** Salivation during the resting condition and 89 Hz-S vibrotactile stimulation over 3 min. Wilcoxon signed-

and standard deviations (SDs), as shown in Figure 6B.

360 Advances in Vibration Engineering and Structural Dynamics

**3.3. Total salivation during vibrotactile stimulation**

89 Hz-S conditions, as shown in Figure 3.

rank test (two-tailed), *P*< 0.01.

paper, 2011).

Among the parameters obtained from the IRIS, we examined D1, D2, CR, T3, VC, and AC, as shown in Figures 4 A and 4B. The pupillary light reflex showed significantly decreased D1, D2, and T3 compared with the resting state. Furthermore, the pupillary light reflex showed increased AC, as shown in Figure 4B. Data from the right and left pupils were simi‐ lar following light stimulation, as shown in Figure 4B. Thus, we employed data from the right pupillary reflex for data analysis.

**Figure 4.** Typical example of data from the pupillary reflex. Right pupillary reflex data after light stimulation in the right pupil (A-a and B-a) and left pupillary reflex data after light stimulation in the left pupil (A-b and B-b).

**Figure 5.** Effect of the pupillary reflex following right-side light stimulation between the resting condition and 89 Hz-S vibrotactile stimulation.

We analysed right pupillary reflex data from eight normal subjects, as shown in Figure 5. D1 was 6.15 ± 0.64 mm under the resting condition and 5.20 ± 1.12 mm under the 89 Hz-S condi‐ tion (*p*< 0.01: Wilcoxon signed-rank test, two-tailed). D2 was 4.03 ± 0.79 mm under the rest‐ ing condition and 3.47 ± 0.84 mm under the 89 Hz-S condition (*p*< 0.01). CR was 0.37 ± 0.10 ms under rest and 0.35 ± 0.09 under the 89 Hz-S condition. T3 was 1.089 ± 0.094 ms under rest and 0.973 ± 0.175 ms under the 89 Hz-S condition (*p*< 0.05). VC was 4.90 ± 0.95 mm/s under rest and 5.03 ± 0.89 mm/s under the 89 Hz-S condition. AC was 52.4 ± 16.8 mm/s2 under rest and 56.7 ± 17.0 mm/s2 under the 89 Hz-S condition (*p*< 0.05). Of particular note, D1, D2, and T3 showed an absolute decrease between the resting state and 89 Hz-S stimula‐ tion, and AC increased between the resting and 89 Hz-S conditions. CR and VC did not change between the resting and 89 Hz-S conditions, as shown in Figure 5.

### **3.5. Analysis of HRV during vibrotactile stimulation**

We recorded typical heart rate changes and performed a power-spectral analysis (HRV module, AD Instruments, Japan) under the following six conditions: (1) resting state; (2) 89 Hz-S face; (3) listening to Mozart (Mozart); (4) Mozart + 89 Hz-S face; (5) 89 Hz-S neck; and (6) listening to noise (Noise), as shown in Figure 6. For example, Mozart + 89 Hz-S face "pulse (+)" indicates that participants were listening to Mozart classical music while receiv‐ ing 89 Hz-S vibrotactile stimulation on the face. When comparing heart rates between the rest state and under various stimuli, we used the RR interval peak value from the powerspectral analysis (Fig. 6A and 6C). RR interval peak values (ms) from the power-spectral analysis were compared. The values were as follows: resting state, 757.5 ± 57.0 ms; 89 Hz-S face, 905.1 ± 189.5 ms, Mozart, 771.7 ± 86.7 ms; Mozart + 89 Hz-S face, 875.3 ± 188.3 ms; 89 Hz-S neck, 901.7 ± 188.4 ms; and Noise, 831.7 ± 114.6 ms (Fig. 6C). Significant differen‐ ces were observed between resting state and 89 Hz-S face (paired *t*-test, *P*< 0.01) and be‐ tween resting state and Mozart + 89 Hz-S face, 89 Hz-S neck, and Noise (paired *t*-test, *P*< 0.05; Fig. 6C).

**Figure 6.** Changes in power spectrums (A) and HRV modulation (C) during various stimuli. B. fNIRS OxyHb concentra‐ tion during various stimuli (this graph was described in a previous article, Hiraba et al. 2011). RS, 89 Hz-S face, Mozart, Mozart + 89 Hz-S face, 89 Hz-S neck, and Noise indicate 89 Hz-S on the face, listening to Mozart, both listening to Mozart and 89 Hz-S vibrotactile stimulation on the face, 89 Hz-S vibrotactile stimulation on the nape of the neck, and listening to noise, respectively. There were significant differences between RS and 89 Hz-S face (paired *t*-test, *P*< 0.01),

Optimal Vibrotactile Stimulation Activates the Parasympathetic Nervous System

http://dx.doi.org/10.5772/51156

363

We reported that 89 Hz-S vibrotactile stimulation evoked rest and increased salivation, as shown in previous papers (Hiraba et al. 2008, 2011). We further investigated increased sali‐ vation during 89 Hz-S. We were the first to show that increased salivation during 89 Hz-S stimulation was due to increased salivation from the submandibular and sublingual glands but not from the parotid glands, as shown in Figure 3. We knew that the amylase-rich paro‐ tid glands were principally responsible for the increased salivation. Salivation also occurs during mechanical stimulation during mastication when eating (Matuo, 2003). In addition to hunger- and mastication-induced salivation, salivation was also increased through 89 Hz-S vibrotactile stimulation of the facial and intraoral structures. This increased salivation may be different from salivation produced by hunger, as increased salivation during 89 Hz-S caused salivation in the submandibular and sublingual glands. In particular, increased sali‐ vation evoked by 89 Hz-S vibrotactile stimulation may be due to somatosensory input from the facial skin and intraoral cavity. Vibrotactile stimulation at 89 Hz-S may evoke a different

between RS and Mozart + 89 Hz-S face, and between 89 Hz-S neck and Noise (paired *t*-test, *P*< 0.05).

**4.1. Relaxation produced by 89 Hz-S vibrotactile stimulation**

perception from masticatory mechanical stimuli.

**4. Discussion**

The resting-state peak value had the lowest frequency. The Mozart-listening peak value was closest to the resting-state value, which might be because the majority of subjects disliked listening to classical music (three subjects favourite music was classical, and seven people reported classical music was not their favourite), as shown in Figure 6C. The 89 Hz-S stimu‐ lation led to the highest heart-beat frequency in comparison with the resting condition, as shown in Figures 6A and 6C. However, 89 Hz-S face stimulation was effective in many sub‐ jects, as 89 Hz-S face had the smallest SD, as shown in Figure 6C. On the other hand, heart rates during the resting condition and while listening to noise were similar. We generated noise with fractioned foam polystyrene. Many subjects may have felt discomfort due to the noise; however, we believe that discomfort induced by this noise was unlikely.

**Figure 6.** Changes in power spectrums (A) and HRV modulation (C) during various stimuli. B. fNIRS OxyHb concentra‐ tion during various stimuli (this graph was described in a previous article, Hiraba et al. 2011). RS, 89 Hz-S face, Mozart, Mozart + 89 Hz-S face, 89 Hz-S neck, and Noise indicate 89 Hz-S on the face, listening to Mozart, both listening to Mozart and 89 Hz-S vibrotactile stimulation on the face, 89 Hz-S vibrotactile stimulation on the nape of the neck, and listening to noise, respectively. There were significant differences between RS and 89 Hz-S face (paired *t*-test, *P*< 0.01), between RS and Mozart + 89 Hz-S face, and between 89 Hz-S neck and Noise (paired *t*-test, *P*< 0.05).
