The population from which the study sample was drawn included 69 elderly people who attended a day service center. We included participants aged 65 years and older, and those with Japanese long-term care insurance (LTCI) services [25
]. Elderly people using LTCI services had physical or mental impairment and needed assistance with activities of daily living [26
]. We excluded participants with Mini-Mental State Examination (MMSE) scores below 24, and those with orthopedic, neurological, cardiovascular, or psychiatric diseases that might influence the results. Finally, 12 elderly people (female, n
= 9; male, n
= 3; mean age ± standard deviation: 81.9 ± 3.9 years) were recruited to the study. Recruitment of participants was conducted by a third party, and all those who were approached to participate consented to taking part. The power calculation was performed using G*Power [27
The study was conducted according to the principles of the Declaration of Helsinki and was approved by the local institutional ethics committee of Kyoto Tachibana University. All participants provided informed written consent and were free to withdraw from the study at any time.
2.2. Study Protocol
This study was a randomized cross-over study. Initially, participants were randomly divided into two groups (A and B) using random numbers generated by Microsoft Excel 2010 (Microsoft, Redmond, WA, USA). Group A received hand massage and group B received foot massage, both for 15 min each. After 1 week, the body part for massage was changed such that group A received foot massage and group B received hand massage, both for 15 min each.
Participants received hand massage while sitting on a chair with a back and armrests. Foot massage was received while in a supine position on a bed. Each massage was performed by two therapists who had worked as instructors at the Japan Wellness Therapist Association for five years. One therapist (female, 65 years old) performed only hand massage and the other (female, 60 years old) performed only foot massage. Fragrance-free massage oil (2E32000, Naris Cosmetics Co., Ltd., Osaka, Japan) was used during each massage to prevent friction and discomfort.
A standard massage technique [13
], without pressure on points indicated by reflexology [29
], was used in this study. Hand massage was performed according to the following procedure. First, the therapist stroked the entire lower arm, from the wrist to the elbow, using her entire hand (5 min). Next, she stroked the entire hand from the fingers to the wrist (5 min). Finally, she stroked the fingers one by one from the base of the finger toward the tip (5 min).
Foot massage was performed according to the following procedure. First, the therapist stroked the entire lower leg, from the ankle to the knee, using her entire hand (5 min). Next, she stroked the entire foot, from the toes to the ankle (5 min). Finally, she stroked the toes, one by one, from the base of the toe toward the tip (5 min). The therapists were requested to keep the pressure intensity and rate of massage as consistent as possible for all participants. Each massage was conducted for 15 min. Emotions and mood states were assessed after each massage (Post1 and Post2), and resting-state EEG activity was measured before (Pre1 and Pre2) and after (Post1 and Post2) each massage. The measurement of emotions and mood states and resting-state EEG activity was conducted by the study investigators.
To assess emotions and mood states, a Likert scale [30
] was used to measure the degree to which participants felt pleasant, relaxed, and refreshed after each massage [13
]. The Likert scale consisted of a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher scores indicating positive emotions and mood states.
Resting-state EEG activity before and after each massage was measured for 90 s in the eyes open condition and while the patients were seated in a chair with a backrest. The EEG was obtained with an electroencephalograph (EEG-9100, Nihon Kohden Co., Ltd., Tokyo, Japan) and an active dry electrode system (Miyuki Giken Co., Ltd., Tokyo, Japan). The EEG was recorded with 19 channels (Fp1, Fp2, F7, F3, Fz, F4, F8, T3, C3, Cz, C4, T4, T5, P3, Pz, P4, T6, O1, and O2) based on the international 10–20 system and at a sampling rate of 1000 Hz. Reference electrodes were attached to both earlobes. Participants in both groups were seated for several minutes after massage before EEG recording.
Recorded EEG data were processed in EEGLAB as follows: down sampling to 256 Hz, band pass filter set to 1–40 Hz, independent component analysis to remove artifacts, and re-referencing to the average reference [31
]. Next, exact low-resolution brain electromagnetic tomography (eLORETA) was used to analyze the cortical distribution of current source density [32
]. In eLORETA, the solution space consists of 6239 cortical gray matter voxels at 5 mm spatial resolution, in a realistic head model [33
], using the Montreal Neurological Institute (MNI) 152 template [34
]. The eLORETA image was calculated in the alpha frequency band (8–13 Hz) because the resting alpha activity reflects the emotions and mood states [35
2.4. Statistical Analysis
The baseline characteristics of the groups A and B were compared to check if the two groups were comparable. The Kolmogorov–Smirnov Test was used to test the normality of distributions, and differences between groups were analyzed using Student’s t-tests for normally distributed variables and the Mann–Whitney U test for variables that were not normally distributed. The degree to which participants felt pleasant, relaxed, and refreshed after each massage was analyzed with the Wilcoxon signed-rank test. Statistical analyses were performed with SPSS ver. 24.0 (IBM, Chicago, IL, USA). The pre- and post-massage cortical distributions of current source density were compared in the alpha frequency band with voxel-by-voxel dependent sample F-ratio tests, based on the eLORETA log-transformed current density power. In the resulting statistical three-dimensional images, cortical voxels exhibiting significant differences were identified with a non-parametric approach (statistical non-parametric mapping; SnPM). The level of significance was set at p < 0.05.