Dr Masayuki Sawada MD, Department of Psychiatry, Nara Medical University School of Medicine, 840 Shijocho Kashihara, Nara 634-8522, Japan. Email: email@example.com
Background: Cerebral vascular disorder (CVD) might result in a quantifiable decrease in quality of life, which is determined not only by the neurological deficits but also by impairment of cognitive functions. There are few studies that report on the cognitive effect of Tai Chi exercise (Tai Chi) on the elderly with CVD. The purpose of the present study was to examine the cognitive effect of Tai Chi on the elderly with CVD using P300 measurement, in addition to the General Health Questionnaire (GHQ) and Pittsburgh Sleep Quality Index (PSQI).
Methods: A total of 34 patients with CVD were recruited from outpatient Akistu-Kounoike Hospital and randomly assigned to receive Tai Chi (n= 17) or rehabilitation (n= 17) in group sessions once a week for 12 weeks. To examine the time courses of each score (P300 amplitude, P300 latency, GHQ score and PSQI score), repeated-measures analysis of variance was carried out with groups and time as factors.
Results: For the time courses of P300 amplitudes and latencies, there were no significant effects of interaction between group and time. However, significant time-by-group interactions were found for Sleep Quality (P= 0.006), GHQ total score (P= 0.005), anxiety/insomnia score (P= 0.034), and severe depression score (P= 0.020).
Conclusions: Tai Chi might therefore be considered a useful non-pharmacological approach, along with rehabilitation, for the maintenance of cognitive function in the elderly with CVD and might be a more useful non-pharmacological approach for the improvement of sleep quality and depressive symptoms in the elderly with CVD than rehabilitation.
Cerebral vascular disorder (CVD), often called stroke, has attracted increased concern in terms of its clinical importance in Japan, where it is now the third most common cause of death. According to the National Medical Care Expenditure, the annual medical expenses for CVD in Japan reached ¥1.8 trillion and accounted for 7.3% of all medical expenses in 2001; thus, this condition is clearly of major medical economic importance.1 CVD might result in a quantifiable decrease in quality of life, which is determined not only by neurological deficits, but also by impairment of cognitive functions.2 Snaphaan et al. reported that the prevalence of post-stroke memory dysfunction varies from 23% to 55% 3 months after stroke and declines to 11–31% one year after stroke.3
Event-related potentials (ERP) are commonly used as physiological measures of cognitive function, because they are easily measured and are non-invasive. Of the major waves observed in the ERP, the P300 component corresponds to mental processes, such as recognition, categorization of stimuli, expectancy and short-term memory, and there are many brain regions, especially in the temporal lobe, the parietal lobe and the hippocampus, that are thought to be responsible for its generation.4 With respect to the relationship between ERP and stroke, Khedr et al. reported that stroke patients showed significant prolongation of P300 latencies.5 Golob et al. examined auditory cortical activity in amnestic mild cognitive impairment (MCI) and reported that P300 latencies were significantly longer in MCI and Alzheimer's disease compared with controls.6 Numerous other clinical P300 studies strongly suggest that this ERP component might be clinically useful as an index of cognitive function.7–9
Tai Chi exercise, often called Tai Chi, has been regarded as both the highest form of martial arts10 and an important regimen,11 and thus Tai Chi has obtained growing popularity not only in the Far East but also in many Western countries. Tai Chi is viewed as a way of life that maintains or restores health, and it is a powerful centering activity in which an individual learns control over some bodily functions and calms the mind.12,13 Few studies have reported on the physical effect of Tai Chi for patients with CVD. Au-Yeung et al. reported that regular practice of a short-form of Tai Chi for 6–12 weeks improves standing balance in people with chronic stroke.14
To our knowledge, there are few studies that have reported the cognitive effect of Tai Chi on the elderly with CVD, and no study has examined the effect of Tai Chi on the elderly with CVD using P300. We therefore designed a single-blinded randomized controlled trial (Tai Chi versus Rehabilitation for the elderly with CVD) using measured P300, as well as the General Health Questionnaire (GHQ)15 and Pittsburgh Sleep Quality Index (PSQI).16
This pilot study was a prospective, single-blind, randomized, controlled clinical trial to evaluate the psychological effects of Tai Chi for patients with CVD. The study was approved by Nara Medical University Investigation Review committee and was carried out in Akitsu-Kounoike Hospital.
Study setting and population
Elderly patients, aged 50 years and older, with CVD were eligible to participate. The subjects, who had been diagnosed with intracerebral haemorrhage, subarachnoid haemorrhage or cerebral infarction with computed tomography (CT) or magnetic resonance imaging, were included in the present study as CVD sufferers. Subjects were excluded if one or more of the following applied: (i) prior experience with Tai Chi or other similar types of complementary and alternative medicine, such as Qigong, yoga and acupuncture, and/or if they were unwilling to participate in Tai Chi; (ii) a Mini-Mental State Examination (MMSE)17 score under 20; or (iii) participation in any other clinical trial within the last 30 days.
Patients continued routine medications and maintained treatment visits with their primary care physician throughout the study. The investigators recorded any changes made to treatment, but did not change or recommend changes in medical therapy.
A total of 34 patients with CVD were recruited from Akistu-Kounoike Hospital outpatient clinic and randomly assigned to receive Tai Chi (n= 17, the Tai Chi group) or rehabilitation (n= 17, the control group; Fig. 1). Written informed consent was obtained from all subjects before the study.
The Tai Chi program was based on classical Yang style.18 Participants of the Tai Chi group participated in 50-min Tai Chi sessions conducted once a week for 12 weeks. Each session included the following: (i) 10 min of warm-up and a review of Tai Chi principles; (ii) 30 min of Tai Chi practice; and (iii) 10 min of cool down.
The rehabilitation program was composed of two parts. The first part was non-resistance training (such as walking and/or standing). The second part was resistance training using the exercise machines and Thera-Band Tubing. The two parts required approximately 20 min and 60 min, respectively. Therefore, participants of the control group participated in approximately 80 min of rehabilitation sessions carried out once a week for 12 weeks.
Based on the guidelines for evoked potential measurement, P300 was obtained by an auditory odd-ball task. A NEC Multi Stim II (NEC, Tokyo, Japan) was used as the auditory stimulus system.
Measurement. Infrequent target stimuli were presented as tone bursts at 2000 Hz (P= 0.2), and frequent non-target stimuli were presented as bursts at 1000 Hz (P= 0.8), with each stimulus lasting 50 ms. Both types of stimuli were given at intervals of 1.5 s and an intensity of 80 dB. The infrequent and frequent stimuli were given in random order through headphones. The subjects were instructed to pay attention to the target stimuli with their eyes open and to press the button as quickly as possible when each target stimulus was delivered.
Recording. P300 was recorded with an MEB 2200 (Nihon Kohden, Tokyo, Japan). Electroencephalograms (EEG) were obtained at Fz, Cz, Pz, C3 and C4 positions on the scalp using disk electrodes. The bilateral ear lobes were used as the reference electrode sites. The resistance of the electrodes was set at ≤5 kΩ. P300 was analyzed during the period between the 50 ms pre-stimulus and the 750 ms post-stimulus. Artifact-free responses to the stimuli were added and averaged after EEG amplitude data ≥100 µV and eye movements were removed. To prevent the subjects from getting tired of, or used to, carrying out the tasks, each trial was carried out only once under the specified conditions.
Analysis. Thirty responses to infrequent target stimuli were averaged. P300 was identified as a positive wave with a peak latency from 250–550 ms. P300 latency and amplitude were also measured. Examiners who were blind to diagnoses measured the P300 of the participants at both start-point and end-point.
The GHQ is a widely used self-report measure of general health developed by Goldberg.15 In the present study, we used the previously validated Japanese version19 of the 60-item GHQ15 to estimate the participants' health-related quality of life. The GHQ-60 used in this study consists of four subscales: (i) somatic symptoms; (ii) anxiety/insomnia; (iii) social dysfunction; and (iv) severe depression. Lower scores indicate better mental health. Examiners who were blind to diagnoses measured the GHQ scores of the participants at both the start-point and end-point.
The PSQI consists of 19 self-rated questions and five questions intended to be completed by the subject's roommate or bed partner (if applicable).16 The 19 self-rated questions are grouped into seven component scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction. Each component score ranges from 0 (no difficulty) to 3 (severe difficulty) and contributes equally to a PSQI global score that ranges from 0 to 21. Negative change scores indicate improvement on the PSQI. This validated instrument has been shown to be a sensitive and specific measure of sleep quality over a 1-month interval.16 The PSQI underwent field testing, which demonstrated that the questionnaire is considered easy to use by subjects and patients. In addition, field testing showed that the seven major components, as well as the 19 individual questions, are internally consistent. Responses for global scores, component scores and individual question responses are stable across time, and its validity is supported by its ability to discriminate patients from controls.16 Examiners who were blind to diagnoses measured PSQI scores of the participants at both the start-point and end-point.
To examine the time courses of each score (P300 amplitude, P300 latency, GHQ score, and PSQI score), repeated-measures analysis of variance (anova) was carried out with groups (the Tai Chi group or the Control group) and time (start-point or end-point) as factors. Group differences at both start-point and end-point were tested with t-tests. All statistical analyses were carried out with spss 17.0 J for Windows (SPSS, Tokyo, Japan).
There were no significant differences between the Tai Chi group (30.8% male subjects; mean age 76.53 years, SD 9.74 years; mean MMSE 25.71, SD 2.71) and the control group (23.5% male subjects; mean age 77.59 years, SD 12.33 years; mean MMSE 26.06, SD 2.77) for gender (P= 0.467), age (P= 0.783), and MMSE score (P= 0.710) at the start-point.
P300 data between the Tai Chi group and the control group
There were no significant differences between the Tai Chi group and the control group in P300 amplitudes and latencies at both the start-point and end-point.
As shown in Table 1, in the time courses of P300 amplitudes (Fz, Cz, Pz, C3, and C4), there were no significant effects of interaction between group and time.
Table 1. P300 across two assessment points
Tai Chi group
Time × group
P300 amplitude (Fz)
P300 amplitude (Cz)
P300 amplitude (Pz)
P300 amplitude (C3)
P300 amplitude (C4)
P300 latency (Fz)
P300 latency (Cz)
P300 latency (Pz)
P300 latency (C3)
P300 latency (C4)
The grand-averaged P300 at Pz for both the Tai Chi group and the control group at start-point and end-point are shown in Figure 2. The grand-averaged P300 at the end-point of the control group seemed later than the grand-averaged P300 at the start-point of the control group, and the grand-averaged P300 at the end-point of the Tai Chi group seemed earlier than the grand-averaged P300 at the start-point of the Tai Chi group.
However, in the time courses of P300 latencies, there were no significant effects of interaction between group and time (Table 1).
PSQI data between the Tai Chi group and the control group
Daytime dysfunction from the control group at the end-point was significantly less than the Tai Chi group at the end-point (P= 0.041). There was no significant difference between the Tai Chi group and the control group in PSQI scores at both the start-point and the end-point among other factors.
As shown in Table 2, significant time-by-group interactions were found for sleep quality (P= 0.006), and significant time effects were found for global sleep quality (P= 0.011), habitual sleep efficiency (P= 0.022), sleep duration (P= 0.009), sleep disturbance (P= 0.022) and daytime dysfunction (P= 0.002).
Table 2. Pittsburgh Sleep Quality Index across two assessment points
GHQ data between the Tai Chi group and the control group
There was no significant difference between the Tai Chi group and the control group in GHQ scores at both the start-point and the end-point.
As shown in Table 3, significant time-by-group interactions were found for GHQ total score (P= 0.005), anxiety/insomnia score (P= 0.034) and severe depression score (P= 0.020), and significant time effects were found for GHQ total score (P= 0.001), social dysfunction score (P= 0.017) and severe depression score (P= 0.004).
Table 3. General Health Questionnaire across two assessment points
P300 latency has been found to increase as dementia symptoms increase and is considered a consequence of attention process, reaction speed and immediate memory. Shorter P300 latencies indicate superior mental performance relative to longer latencies.20 However, to our knowledge, no study has examined the effect of Tai Chi on the elderly with CVD using P300. Therefore, the present report, examining the potential benefits of Tai Chi in this patient population, is a useful study. We regret that there were no significant differences between the Tai Chi group and the control group in P300 amplitudes and latencies at the end-point, and there were no significant effects of interaction between group and time. However, P300 latencies of the Tai Chi group did show more of a trend toward decreasing than those of the control group (P= 0.185 (Fz), P= 0.179 (Cz), P= 0.165 (Pz), P= 0.131 (C3), P= 0.103 (C4), data not shown). Furthermore, the grand-averaged P300 at the end-point of the control group seemed later than the grand-averaged P300 at its start-point, and the grand-averaged P300 at the end-point of the Tai Chi group seemed earlier than the grand-averaged P300 at its start-point (as shown in Fig. 2). Therefore, Tai Chi might promote cognitive function in the elderly with CVD.
Exercise, leisure and diet are known to be important factors in the primary prevention of dementia and related disorders.21 Additionally, reminiscence therapy in particular has been found to be a useful non-pharmacological approach for elderly with dementia in past years, the recent studies have reported that reminiscence therapy is a useful non-pharmacological approach for elderly both with and without dementia,22 and is more effective than everyday conversation in the treatment of elderly with dementia.23 The present study suggests that Tai Chi is also at least a useful non-pharmacological approach for maintaining cognitive function in the elderly with CVD. Therefore, Tai Chi might be an effective cognitive function in the elderly as reminiscence therapy.
With respect to the relationship between Tai Chi and sleep quality, Irwin et al. examined the efficacy of Tai Chi in promoting sleep quality in the elderly.24 They designed a randomized controlled trial (Tai Chi versus health education for the elderly), measuring PSQI, and showed that Tai Chi can be considered a useful non-pharmacological approach for improving sleep quality in the elderly with moderate complaints. Our findings about sleep quality (as shown in Table 2) show that both Tai Chi and rehabilitation have a positive effect on global sleep quality, habitual sleep efficiency, sleep duration, sleep disturbance and daytime dysfunction, and furthermore show that Tai Chi can improve sleep quality better than rehabilitation. Thus, these results support the Irwin et al. study24 and, furthermore, show that Tai Chi might be a more useful non-pharmacological approach to promote sleep quality in the elderly with CVD than rehabilitation.
Our findings about general health, including somatic symptoms, anxiety/insomnia, social dysfunction, and depression (as shown in Table 3), show that both Tai Chi and rehabilitation promote general health. With respect to the relationship between Tai Chi and GHQ score, we previously examined the effects of Tai Chi on healthy elderly volunteers using GHQ and reported that significant time-by-group interactions were found for GHQ total score and severe depression score; thus, Tai Chi is related to improvement in GHQ among the elderly.25 Our findings showed that Tai Chi is more effective than rehabilitation for depressive symptoms, which is supported by our preliminary report. Furthermore, our findings showed that Tai Chi is more effective than rehabilitation for anxiety/insomnia score. Thus, Tai Chi might be a more useful non-pharmacological approach to promote not only depressive symptoms, but also anxiety/insomnia in the elderly with CVD than rehabilitation. Tsang et al. suggested the hypothesis that Qigong increases the levels of monoamine neurotransmitters in the brain. Increased tryptophan levels, as a consequence of exercise, might result in increased brain serotonin synthesis and release.26 Our findings might be based on their hypothesis. Biological studies would be needed to support their hypothesis.
There were two main limitations to the present study. First, the sample size was small. However, we did enter 34 participants in the present study, which is a large sample according to ERP study standards. Second, the study term was short. There might have been no significance between the Tai Chi group and the Control group in P300 amplitudes and latencies because of the short-term nature of the intervention. Therefore, future ERP studies with large samples and long-term intervention are needed to determine whether Tai Chi is effective for improving cognitive function. In the present study, the Tai Chi participants completed the study with the exception of one participant who dropped out and was unable to be contacted. Furthermore, no mental and/or physical adverse events were reported. Tai Chi might therefore be considered a useful non-pharmacological approach, along with rehabilitation, for the maintenance of cognitive function in the elderly with CVD and might be a more suitable exercise for the elderly to improve their health-related quality of life and sleep quality than rehabilitation.