Considering that we want to continue the previous studies [11,12,13,14, 16], a total of 27 female Japanese university students (mean age: 21.9 ± 1.9 years) were included in the current study. None of the participants reported any physiological or psychiatric disorders in their personal histories, and all were non-smokers. Further, none of the participants were menstruating on the day of the experiment. All participants provided written informed consent to participate after being informed about the study’s aims and procedures. This study was conducted in accordance with the Declaration of Helsinki and adhered to the regulations set by the Ethics Committee of the Center for Environment, Health, and Field Sciences, Chiba University, Japan (Project Identification Code Number 34). This study is registered in the University Hospital Medical Information Network of Japan (Unique ID issued by UMIN: UMIN000032808).
Sugi (Japanese cedar, Cryptomeria japonica) wood grown in the Miyazaki Prefecture, Japan, was used as the base material for this experiment. Five wooden laminae without vertical joints (600 × 120 × 20 mm) were bonded together along their width using a water-based polymer isocyanate adhesive. A second bonding was made using sugi plywood (600 × 600 × 24 mm) to prevent bending; the total thickness of the resultant finished product was 44 mm. The surface was uncoated and finished via either (1) brushing using a stainless-steel wire brush (hereafter referred to as “uzukuri wood” Fig. 1a), or (2) sanding using an abrasive-band machine (hereafter referred to as “sanded wood” Fig. 1b). A marble slab (600 × 600 × 20 mm) was used as the control material, because its properties are substantially different from those of wood. Moreover, marble was used as a comparative sample in previous studies [11,12,13, 16]. In this study, it was placed on a piece of cedar plywood of the same size as that of the plywood bonded to sugi. The surface was processed by buffing (Fig. 1c). This buffed marble sample is hereafter referred to as “marble”. All materials were stored at room temperature. The physical properties of the materials are listed in Table 1.
Physiological effects were measured in a chamber with an artificial climate which has a soundproof function maintained at 24 °C, 50% relative humidity, and 230-lux illumination.
Figure 2 depicts the experimental setup, whereas Fig. 3 depicts the experimental protocol. Sensors for the physiological measurements were attached to each participant’s forehead and chest. After receiving information about the overall measurement flow, the participants initially trialed the test procedure by placing their feet on a practice material (plastic board) utilizing the same procedure to be followed for the actual measurements. The participants rested with their eyes closed for 60 s (Fig. 2a). By raising the sample using a scissor lift (Fig. 2b), the participants passively touched the samples with their sole of the feet while keeping their eyes closed for 90 s (Fig. 2c). The contact was ended by lowering the sample with the scissor lift (Fig. 2a). The researcher then set up another sample for the next measurement, covered the sample with a cloth so that the subjects could not see the sample, and then told the participants to open their eyes. Next, the subjective evaluation tests of the participants were performed. Stimulation with the subsequent materials was performed after a rest period of approximately 5–7 min. The materials were presented in a counterbalance to eliminate any order effects, such as adaptation and fatigue, caused by the sequence of tactile stimulation.
Near-infrared time-resolved spectroscopy (TRS)
Near-infrared time-resolved spectroscopy (TRS), which is one of the methods of near-infrared spectroscopy (NIRS), was used as an index for the prefrontal cortex activity. The prefrontal cortex is a part of the brain that is responsible for higher-order judgments, such as problem-solving and decision-making. It has been reported that NIRS signals were highly correlated with functional magnetic resonance imaging measurements by a previous study using simultaneous NIRS (placing NIRS probes over the frontal brain region including the prefrontal region) and functional magnetic resonance imaging during a battery of cognitive tasks . Thus, NIRS enables the detection of brain activity in the prefrontal cortex region. Sensors were mounted on the participant’s forehead, and the oxy-Hb concentrations in the prefrontal cortex were measured using the TRS-20 system (Hamamatsu Photonics K.K., Shizuoka, Japan). Changes in blood flow (rCBF) associated with the local nerve activity in the brain were correlated with changes in the oxy-Hb and total Hb concentrations in the near-infrared spectroscopy measurements [22,23,24]. Changes to the oxy-Hb concentration measured by TRS reflected the prefrontal activity via a sensor attached to the forehead [25,26,27]. It has been reported that oxy-Hb concentrations in the prefrontal cortex are reduced by pleasant emotions and increased by unpleasant emotions . The oxy-Hb concentrations in the left and right prefrontal cortices were measured for 30 s before (pre-measurement) and during the 90-s duration in which the participants’ sole of the feet touched the materials (post-measurement). All data were transformed by linear interpolation to every 1 s because their sampling intervals were approximately 1.07–1.19 s. In addition, post-measurement values (at every second) were calculated as the differences between the pre-measurement values (mean 30 s). In addition, the total hemoglobin (total Hb) and deoxygenated hemoglobin (deoxy-Hb) concentrations were measured and calculated as described for the oxy-Hb concentration.
Heart rate variability (HRV) and heart rate
HRV and heart rate were employed as indicators of autonomic nervous activity. HRV was analyzed for the periods between consecutive R waves (R–R intervals, RRI), as measured by a portable electrocardiograph (Activtracer AC-301A; GMS, Tokyo, Japan) [29, 30]. This device uses a 3-lead electrocardiogram (Lead II) to perform necessary measurements. High frequency (HF; 0.15–0.40 Hz) and low frequency (LF; 0.04–0.15 Hz) power level components of HRV were calculated using the maximum entropy method (MemCalc/Win; GMS, Tokyo, Japan) [31, 32]. The HF power indicated parasympathetic nervous activity, whereas the LF/HF power ratio indicated sympathetic nervous activity [29, 33]. The natural logarithmic values for HF, ln(HF) power and the LE/HF, ln(LF/HF) power ratios were employed to normalize the HRV parameters across participants . The mean HRV and heart rate were calculated for 90 s during which the participants’ soles were in contact with each material. Average values for the pre- and post-measurement values for ln(HF) and ln(LF/HF) were calculated.
The HRV power spectra were also used to estimate the participant’s respiratory frequency . Respiratory changes influenced the HRV data, with heart rate generally increasing during inspiration and decreasing during expiration [36, 37]; thus, the respiratory rate could be estimated from the dominant frequency of the HF component, found by detecting the maximal power of the HF component. To detect the peak frequency of the HF component, the model order for the spectral analysis was chosen from 8 to 11; however, the ninth order was used in principle. Since respiratory changes affect HRV data, respiratory frequencies were statistically compared between the mean values over the 90-s contact period for each material and pre- and post-measurement by paired t-tests.
The modified semantic differential (SD) method  and the short version of the Profile of Mood States Second Edition (POMS2) [39,40,41,42] were used to evaluate the psychological effects produced by the tactile stimulation from the materials. In the modified SD methods, seven pairs of adjectives from 13 scales of adjectives were assessed “comfortable–uncomfortable”; “relaxed–awakening”; “natural–artificial”; “warm–cold”; “uneven–flat”; “dry–moist”; and “soft–hard”. POMS2 questionnaire was used to evaluate the mood states from the scores for Tension–Anxiety (T-A), Depression–Dejection (D-D), Anger–Hostility (A-H), Fatigue–Inertia (F-I), Confusion–Bewilderment (C-B), Vigor–Activity (VA), and Friendliness (F). The total mood disturbance (TMD) score was calculated using this formula: [(T-A) + (D-D) + (A-H) + (F-I) + (C-B)−(V-A)]. The TMD score reflects the general mood state, whereby a higher score denotes negative feelings of the participant and vice versa. The POMS2 is a well-established means sure of psychological distress derived from factor analysis that has been shown to possess high levels of reliability and validity. In order to reduce the burden on the participants, a shortened Japanese version of POMS2 with 35 questions was used for this study.
The Statistical Package for the Social Sciences software (v21.0; IBM Corp., Armonk, NY, USA) was used for all of the statistical analyses. In all cases, p < 0.05 was considered statistically significant.
Paired t-tests were applied to analyze (1) differences in the physiological indices between the three materials (uzukuri wood vs. marble, sanded wood vs. marble, uzukuri wood vs. sanded wood) and (2) comparison before and after contact (pre- vs. post-measurement) for each material. Wilcoxon signed-rank tests were applied to analyze differences in the psychological indices between the three materials.
Holm correction was applied thrice to adjust the family-wise error rate in the comparison between stimuli . The test statistics were calculated for all comparisons, and P-values were obtained. Hypotheses were arranged in descending order based upon P-values. Since the total number of tests was three, the minimum P value was compared with the significance level α = 0.05/3. If the result of the aforementioned test was significant, then the second smallest P-value was compared with α = 0.05/2. In the same manner, proceeding in descending order based on the P-values, reduced the denominator by one, for each test. If a non-significant comparison is found, then the test is terminated at that point. For subsequent hypotheses, even if there was a P-value lower than the significance level at that step, we will hold all of the conclusions, and conclude that there was no difference found.