Nineteen healthy volunteers (mean age, 43.6 ± 10.1 years; 15 women and 4 men) were enrolled. None of the participants had a history of medical illness. Participants with a history of health problems, taking chronic medication or supplemental vitamins, and those who weighed < 40 kg [18–22] were excluded. Good health was assessed by physical examination, blood pressure, and heart rate. The protocol was approved by the Ethics Committee of Osaka City University, and all participants provided written informed consent.
The day before the experiment, participants finished dinner by 9:00 pm and then fasted overnight. The following morning, they had breakfast before the visit. At 10:00 a.m., after the visit, they started the experiment. Before the start of the experiment, a paper-and-pencil questionnaire was distributed to participants to evaluate their daily level of fatigue. As a fatigue-inducing mental task session, participants performed 2-back test  trials for 30 min , and as a fatigue-evaluating mental task, they performed the advanced trail making test (ATMT; ) for 30 min  before and after the fatigue-inducing task session. Just before and after the fatigue-inducing session, they recorded their subjective sensation of fatigue on a visual analogue scale (VAS) from 0 (no fatigue) to 100 (complete exhaustion)  and underwent electrocardiography (ECG) with their eyes closed for 1 min while sitting on a chair. VAS and ECG recordings were performed before the ATMT trials. This study was conducted in a quiet, temperature- and humidity-controlled environment. For 1 day before the visit, participants refrained from intense mental and physical activities and caffeinated beverages, consumed a normal diet, and maintained normal sleeping hours.
A paper-and-pencil questionnaire was distributed to participants. The severity of daily level of fatigue was measured using Chalder's fatigue scale (Chalder et al. 1993), which has been previously used in Japanese participants . The reliability and validity of the Japanese version of Chalder's fatigue scale to evaluate the severity of daily fatigue have been previously confirmed . The fatigue scale consists of 11 questions using a 4-point (0-3) Likert scale that allows the following responses: 0 = less than usual; 1 = no more than usual; 2 = more than usual; 3 = much more than usual during the past several weeks. The total score for the 11-item fatigue scale ranges from 0 to 33, with higher scores indicating a greater degree of daily fatigue.
Fatigue-inducing mental task
As a fatigue-inducing mental task, participants performed the 2-back test for 30 min . During this task, one of four letters was presented on a display of a personal computer every 3 sec, and they had to judge whether the target letter presented at the center of the screen was the same as the one that had appeared 2 presentations before. If it was, they were to press the right mouse button with their right middle finger; if it was not, they were to press the left mouse button with their right index finger. They were instructed to perform the task trials as quickly and as correctly as possible. The results of each 2-back trial, that is, a correct response or error, were continuously presented on the display of the personal computer.
Fatigue-evaluating mental task
As a fatigue-evaluating mental task, participants performed the ATMT for 30 min . In this test, circles numbered from 1 to 25 were randomly placed on the display of a personal computer, and participants were required to use a computer mouse to touch these circles in sequence, starting with number 1. Tasks A, B, and C all ended when they touched the 25th target. They continued directly with the next Tasks B, C, and A, in that order, on and on for 30 min. The number of hits counted and the time were counted. In task A of the ATMT, when they touched a target circle, it remained in the same position, but the color changed from black to yellow. The positions of the other circles remained the same. In task B of the ATMT, when they touched the first target circle, it disappeared, and circle number 26 appeared in a different position on the screen. The positions of the other circles remained the same. For example, touching circles 2, 3, and 4 resulted in their disappearance and the addition of circles 27, 28, and 29 on the screen, so that there were always 25 circles on the screen. In task C of the ATMT, when they touched the first target circle, it disappeared and circle number 26 appeared in a different position on the screen and the position of all other circles changed at random. As in task B, there were always 25 circles on the screen. Participants performed tasks A, B, and C consecutively. They were instructed to perform all task trials as quickly and as correctly as possible.
ECG was recorded using active tracer AC301 (Global Medical Solution Inc., Tokyo, Japan), and the ECG was analyzed using MemCalc for Windows (Global Medical Solution Inc.). Data were analyzed offline after analogue-to-digital conversion at 250 Hz. R-R wave. Irregularities in the ECG recordings were excluded from the analyses. Variability was measured as an indicator of autonomic nerve activity. For frequency domain analyses of the R-R wave intervals, the percent of low-frequency power (LF) was calculated as the power within the frequency range of 0.04 to 0.15 Hz, and the percent of high-frequency power (HF) was calculated as that within the frequency range of 0.15 to 0.4 Hz. LF and HF were measured in absolute and normalized units; normalization was performed by dividing the absolute power by the total variance and then multiplying by 100. The %HF is vagally mediated [28–30], but the %LF originates from a variety of sympathetic and vagal mechanisms [28, 31]. The LF/HF ratio represents the sympathetic to parasympathetic balance .
Values are shown as mean ± SD unless otherwise noted. Paired t-tests were used to evaluate the differences before and after the mental fatigue-inducing task as for the ATMT performances and VAS scores and Wilcoxon's signed rank tests as for the indices of the heart rate variability. Pearson's correlation analyses were conducted to evaluate relationships between two variables. All P values were 2-tailed, and P values less than 0.05 were considered statistically significant. Statistical analyses were performed using SPSS 17.0 software package (SPSS Inc., Chicago, IL).