Seated T'ai Chi in Older Taiwanese People Using Wheelchairs: A Randomized Controlled Trial Investigating Mood States and Self-Efficacy

Abstract

Objective:

There is growing interest in t'ai chi, but little research has addressed whether t'ai chi is effective in older people using wheelchairs for mobilization. The aim of this study was to compare the effects of seated t'ai chi exercise and usual standard activities on mood states and self-efficacy in older people living in a long-term care facility and using wheelchairs for mobilization.

Design:

Randomized controlled trial (trial registration no. ACTRN12613000029796).

Setting:

One long-term-care facility in Taiwan.

Participants:

Sixty participants were randomly assigned by a computer-generated random sequence to a t'ai chi group (n = 30) or a usual exercise and entertainment activities group (n = 30).

Intervention:

Seated t'ai chi exercise for 40 minutes three times a week for 26 weeks was provided.

Main outcome measures:

Mood states (Profile of Mood States Short Form [POMS-SF]) and self-efficacy (Self-Efficacy for Exercise [SEE]).

Results:

At week 26, participants in the t'ai chi group reported significantly lower mood states on the fatigue-inertia dimension of the POMS-SF (mean score ± standard deviation, 3.56 ± 3.71) than did the control group (mean score, 7.16 ± 6.36) (F [1, 58] = 7.15; p < 0.05). The t'ai chi group recorded significantly higher SEE levels (mean, 35.66 ± 36.83) than did those in the control group (mean, 15.30 ± 26.43) (F [1, 58] = 6.05; p < 0.05).

Conclusion:

The findings highlight the importance of t'ai chi for a reduction in the fatigue-inertia mood state and an increase in self-efficacy for older people using wheelchairs.

Introduction

Growing old is a multidimensional process that influences the physiologic, biomedical, psychosocial, social, and economic perspectives of older life. As a result, older people may have to adapt to age-related changes, such as cognitive impairment, frailty, and disability. Such changes increase the risk for falls and may require older people to mobilize in a wheelchair and live in a long-term care (LTC) facility. In some countries, such as Canada¹ and Taiwan,² it is common for frail older people to use wheelchairs for mobilization. In Taiwan, an estimated 67% of people age 65 years or older in LTC facilities use wheelchairs for their daily activities.² The use of a wheelchair reduces opportunities to undertake exercise and as a result may reduce physical health.³

Older people who experience poor physical health are at risk for depression.⁴ Poor physical health can result in older people presenting with multiple mood states, such as being anxious or easily upset and having feelings of guilt, low self-esteem, helplessness, and/or hopelessness.⁵ Furthermore, this population is at risk for limited involvement in social and physical activities; have suicidal tendencies with or without open expression; and as a result may be unable to achieve social function.⁶ Many studies have found a positive effect of “perceived self-efficacy” for older people, in terms of an improvement in their sense of capacity to perform a task or meet a challenge, in relation to participating in an intervention such as a physical activity.^7,8 It is therefore argued that if mood states and self-efficacy in older people can be successfully improved, there will be a potential increase in health benefits for older people,⁵ including those using wheelchairs for mobility.

Although participation in an activity for older people in wheelchairs can be meaningful, not all activities are feasible for them. T'ai chi has been reported as being beneficial for enhancing psychological health, for example by lessening mood disturbances and improving positive mood states⁹ and by increasing social functioning.¹⁰

Despite increasing interest in t'ai chi, little research has addressed the efficacy of whether a t'ai chi intervention can improve individual perception of self-efficacy or mood states in older people using wheelchairs for mobility and living in an LTC facility. Few randomized controlled trials have evaluated the effectiveness of a seated t'ai chi program for these people. The aims of the current study were to examine the effect of a 26-week seated simplified t'ai chi exercise program (STEP) developed by Chen et al.¹¹ especially for frail older people. The seated movements used in this study replicated the movements developed by Chen et al.,¹² which involve few leg movements suited to persons undertaking t'ai chi in a wheelchair.

Materials and Methods

Study design

This study was a single site, repeated-measures randomized controlled trial. A convenience sample of participants was recruited from the Jen-Tao LTC facility in the Changhua region of Taiwan. Residents who met the following inclusion criteria were invited to participate: (1) use of a wheelchair and age 65 years or older, (2) ability to speak and understand Chinese or Taiwanese and no severe hearing impairment, (3) ability to use and raise both arms while sitting in a wheelchair, and (4) cognitive function with Mini-Mental State Examination (MMSE) score greater than 24/30. Residents were excluded if they had (1) symptoms of acute pain or infection, (2) any lower-extremity infection, (3) current palliative care, (4) blindness, or (5) documented or observable psychiatric or neurologic disorders that might interfere with participation. On the basis of G-Power to detect an effect size of 0.8 with 80% probability at 0.05 and accounting for attrition, 60 participants (i.e., 30 residents in each group) were recruited for this study.

Ethics approval was obtained from the Griffith University Human Research Ethics Committee (ethics approval registration number NRS/17/10/HREC). Participants were given written and verbal information about the study, and they provided written consent to participate.

Intervention

Participants were randomly assigned by using a computer-generated random sequence to one of two parallel groups in a 1:1 ratio: an intervention group (t'ai chi) or a control group (usual care). Persons in the intervention group participated in three 40-minute sessions of seated t'ai chi exercise (STEP^11,12) each week for 26 weeks. Each 40-minute session consisted of a 10-minute warm-up, 25-minute STEP session, and 5-minute cool-down. All classes took place mid-afternoon, according to the American Council of Exercise¹³ recommendation. A trained and qualified t'ai chi instructor led two classes of 15 participants each. The intervention sessions had three parts, as described below.

1. Warm-up for 10 minutes

This part consisted of the following: (1) nodding the head, (2) shrugging the shoulders, (3) waving the hands, (4) swinging the arms, (5) twisting the waist, (6) circling the legs, (7) jiggling the knees, (8) raising the feet, and (9) closing movement. The sequence of movements from steps 1–9 was repeated twice, and the warm-up then finished with movement 9.

2. STEP sequence for 25 minutes

This part involved the following: (1) commencing form, (2) curving the back arms, (3) pushing the palms forward, (4) wind pushing the wall, (5) holding the palms as if on horseback, (6) palming the energy of the tiger, (7) performing energy tower to the sky, (8) performing needle at the sea bottom, (9) embracing the moon, (10) lifting the moon to the sky, (11) punching down, and (12) closing movement. The sequence of movements 1–11 was done 12 times, and the STEP section then finished with closing movement (step 12).

3. Cool-down for 5 minutes

This part consisted of just two movements: (1) heaven and earth and (2) rubbing through circulation. These movements were done eight times.¹¹

Control group

The participants allocated to the control group received their usual standard exercise and entertainment activities with no seated t'ai chi exercise. Standard activities provided by the facility, included use of machines for rehabilitation, such as a balancer hand trainer; watching TV and DVDs; karaoke singing; playing card games; chess and croquet; throwing activities; listening to music; drawing; arranging flowers; origami; and doing stretching exercises by following an instructor on a DVD.

Outcome measures

This study used the Chinese versions of the Profile of Mood States Short Form (POMS-SF)¹⁴ and the Self-Efficacy for Exercise (SEE)¹⁵ instruments to collect data at baseline, week 13, and week 26. The POMS-SF presents 30 adjectives describing feelings and moods that the respondent is asked whether they have experienced during the past week. Participants respond on a 5-point scale ranging from 0 (not at all) to 4 (extremely). The POMS-SF includes six dimensions of mood states: tension-anxiety, depression-dejection, anger-hostility, vigor-activity, fatigue-inertia, and confusion-bewilderment. Each of these six dimensions is defined by the responses to five adjectives. Chen et al.^9,16 developed and used the Chinese version of the POMS-SF with 80 older community-dwelling people in Taiwan with good reliability for the six dimensions scales, with Cronbach α ranging from 0.98 to 0.99.

Resnick and Jenkins¹⁵ developed the SEE scale. This consists of nine items that measure the participants' perception of their confidence to continue doing exercise in the face of various circumstances, such as bad weather, tiredness, and feeling depressed. The items are scored from 0 (not confident) to 10 (very confident): The higher scores represent greater exercise self-efficacy. Reliability of the SEE scale was tested with 187 older people from a North American population, and the Cronbach's α was 0.92. Lee et al.⁸ introduced the Chinese version of the SEE scale using it to assess 192 older Chinese adults in Taiwan, with good reliability: The Cronbach α was 0.75.

Data analysis

The data were analyzed by using the Statistical Package for Social Sciences (SPSS) software, version 18.0 (SPSS, Inc., College Station, TX), and an intention-to-treat approach. Data missing completely at random were handled by using the last observation carried forward method.¹⁷ For descriptive statistics, frequency, percentages, means, and standard deviations (SDs) were used to examine demographic characteristics and POMS-SF and SEE scores. Demographic characteristics and baseline data of control and intervention groups were compared by using the chi-square test for categorical variables and t-test for continuous variables. Reliability of instruments was determined via Cronbach α.

The effect of t'ai chi on mood states and self-efficacy were compared between the control and intervention groups at baseline, week 13, and week 26 via analysis of covariance (ANCOVA) tests. Effects of the MMSE scores, length of time living in the LTC facility, and the length of time using a wheelchair on the outcomes of the t'ai chi intervention were examined as covariates. To determine the likelihood of outcome measure improvements for the t'ai chi group compared with the control group, odds ratios were computed by using the Pearson chi-square test with risk estimates because there were generally 10 or more in any cell of the data table; in contrast, the Fisher exact testi is a more conservative test that is better suited for data tables with small cell counts. A significant α level of 0.5 was set for all analyses.

Results

Figure 1 outlines the flow of participants throughout the randomized controlled trial. As shown in Table 1, and as might be expected given the randomized allocation, at baseline there were no significant differences between the demographic characteristics of participants in the intervention group and those in the control group. The Cronbach α was 0.70 for the POMS-SF and 0.66 for the SEE. Participants' MMSE scores, length of stay in a long-term care facility, and period of wheelchair use did not influence the results.

FIG. 1.

The flow of participants throughout the randomized controlled trial.

Table 1.

Demographic Characteristics of Participants (n = 60)

Characteristic	Intervention group (n = 30)	Control group (n = 30)	p-Value
Mean age ± SD (yr)	80.73 ± 9.68	81.77 ± 6.32	0.62^a
Sex
Male	11 (36.7)	11 (36.7)	1.00^b
Female	19 (63.3)	19 (63.3)
Ethnicity
Taiwanese	30 (100.0)	28 (93.3)	0.15^b
Chinese	0 (0.0)	2 (6.7)
Religion
Buddhist	12 (40.0)	7 (23.3)	0.34^b
Taoist	16 (53.4)	21 (70.0)
Christian	1 (3.3)	0 (0.0)
None	1 (3.3)	2 (6.7)
Highest education level
None	18 (60.0)	21 (70.0)	0.62^b
Primary school	6 (20.0)	7 (23.4)
Junior high school	3 (10.0)	1 (3.3)
High school	2 (6.7)	1 (3.3)
College/university	1 (3.3)	0 (0.0)
Current marital status
Married	13 (43.3)	13 (43.3)	0.17^b
Single	3 (10.0)	1 (3.3)
Widowed	11 (36.7)	16 (53.4)
Separated	3 (10.0)	0 (0.0)
Length of residency in long-term care facility
≤1 yr	13 (43.3)	16 (53.3)	0.64^b
1–5 yr	12 (40.0)	9 (30.0)
>5 yr	5 (16.7)	5 (16.7)
Duration of wheelchair use
≤1 yr	16 (53.3)	15 (50.0)	0.77^b
1–5 yr	10 (33.3)	12 (40.0)
>5 yr	4 (13.4)	3 (10.0)
Mean MMSE score ± SD	20.50 ± 3.19	20.50 ± 2.92	1.00^b

Unless otherwise noted, values are the number (percentage) of participants.

t-test.

Chi-square test.

SD, standard deviation; MMSE, Mini-Mental State Examination.

No significant differences (ANCOVA) were found between the overall mood state levels of participants in the intervention group and those in the control group at baseline, week 13, and week 26 (Table 2). However, the fatigue-inertia dimension of the POMS-SF did achieve statistical significance (F [1, 58] = 7.15; p < 0.05) with a large effect size (Cohen d) of 0.69 at week 26 (Table 2). Comparison of the mean scores at week 26 indicates that the intervention group (mean, 3.56 ± 3.71) had a lower mood state on the fatigue-inertia dimension than the control group (mean, 7.16 ± 6.36). In comparison of the change in mean scores from baseline to week 13 and week 26, no significant difference was found for overall mood states or its dimensions (Table 3).

Table 2.

Responses on Profile of Mood States Short Form

Variable	Intervention group (n = 30)	Control group (n = 30)	p-Value	Effect size (Cohen d)
Overall mood states
Baseline	13.20 ± 22.08	12.80 ± 18.21	0.93	0.02
Week 13	12.73 ± 18.90	18.93 ± 23.28	0.26	0.29
Week 26	12.43 ± 19.52	20.93 ± 24.99	0.14	0.38
Tension-anxiety
Baseline	4.36 ± 5.26	3.90 ± 4.29	0.70	0.09
Week 13	3.86 ± 4.09	4.73 ± 4.71	0.44	0.19
Week 26	4.36 ± 4.90	5.20 ± 5.37	0.53	0.16
Depression-dejection
Baseline	4.90 ± 4.77	4.36 ± 4.61	0.66	0.11
Week 13	4.96 ± 5.79	5.96 ± 5.53	0.49	0.17
Week 26	4.43 ± 4.93	6.33 ± 5.60	0.16	0.36
Anger-hostility
Baseline	3.40 ± 4.43	2.46 ± 3.84	0.38	0.22
Week 13	3.00 ± 3.71	2.93 ± 4.01	0.94	0.01
Week 26	3.53 ± 3.74	3.53 ± 4.17	1.00	0.00
Vigor-activity
Baseline	8.96 ± 5.21	6.53 ± 5.44	0.08	0.45
Week 13	7.93 ± 4.72	6.26 ± 5.39	0.20	0.32
Week 26	7.76 ± 6.17	6.06 ± 6.19	0.29	0.27
Fatigue-inertia
Baseline	3.93 ± 4.36	4.70 ± 5.47	0.55	0.15
Week 13	4.03 ± 4.45	6.80 ± 6.31	0.05	0.50
Week 26	3.56 ± 3.71	7.16 ± 6.36	0.01^a	0.69
Confusion-bewilderment
Baseline	5.56 ± 4.42	3.90 ± 3.40	0.10	0.42
Week 13	4.80 ± 3.32	4.76 ± 3.18	0.96	0.01
Week 26	4.30 ± 3.16	4.76 ± 2.87	0.55	0.15

Unless otherwise noted, values are mean ± SD.

Significant at p < 0.05.

Table 3.

Mean Change in Profile of Mood States Short Form from Baseline to Week 13 and Week 16 (n = 60)

Time period	Intervention group	Control group	F-value	p-Value
Week 13 to baseline (overall POMS-SF)	−0.46 ± 20.61	6.13 ± 25.72	1.20	0.28
Week 26 to baseline (overall POMS-SF)	−0.76 ± 22.09	8.13 ± 31.29	1.61	0.20
Week 13 to baseline (tension-anxiety)	−0.50 ± 4.41	0.83 ± 6.00	0.96	0.33
Week 26 to baseline (tension-anxiety)	0.00 ± 5.42	1.30 ± 6.82	0.66	0.41
Week 13 to baseline (depression-dejection)	0.06 ± 5.55	1.60 ± 6.56	0.95	0.33
Week 26 to baseline (depression-dejection)	−0.46 ± 4.68	1.96 ± 7.96	2.08	0.15
Week 13 to baseline (anger-hostility)	−0.40 ± 4.06	0.46 ± 4.54	0.60	0.44
Week 26 to baseline (anger-hostility)	0.13 ± 3.85	1.06 ± 4.95	0.66	0.41
Week 13 to baseline (vigor-activity)	−1.03 ± 4.58	−0.26 ± 6.44	0.28	0.59
Week 26 to baseline (vigor-activity)	−1.20 ± 5.20	−0.46 ± 8.21	0.17	0.68
Week 13 to baseline (fatigue-inertia)	0.10 ± 5.78	2.10 ± 6.89	1.48	0.22
Week 26 to baseline (fatigue-inertia)	−0.36 ± 4.70	2.46 ± 8.41	2.59	0.11
Week 13 to baseline (confusion-bewilderment)	−0.76 ± 4.21	0.86 ± 4.35	2.18	0.14
Week 26 to baseline (confusion-bewilderment)	−1.26 ± 5.13	0.86 ± 4.42	2.97	0.09

Unless otherwise noted, values are mean ± SD.

POMS-SF, Profile of Mood States Short Form.

Interestingly, odds ratio estimation (Pearson chi-square) indicated that participants in the intervention group were 2.9 times more likely to show an improvement in their overall mood states than those in the control group. Furthermore, participants in the intervention group were 2.5 times more likely to improve in their fatigue-inertia dimension (p = 0.1) for exercise levels than those in the control group (Table 4). Although these results did not achieve statistical significance, they may suggest a potential clinical effect of the intervention for overall mood states and the fatigue-inertia dimension.

Table 4.

Odds Ratios for Profile of Mood States Short Form for the Intervention and Control Groups

Variable	Intervention group (n = 30), n (%)	Control group (n = 30), n (%)	Odds ratio (95% confidence interval)	p-Value
Overall mood states			2.93 (0.94–8.71)	0.05
Yes	14 (46.7)	7 (23.3)
No	16 (53.3)	23 (76.7)
Tension-anxiety			1.83 (0.61–5.45)	0.27
Yes	12 (40.0)	8 (26.7)
No	18 (60.0)	22 (73.3)
Depression-dejection			2.10 (0.71–6.22)	0.17
Yes	13 (43.3)	8 (26.7)
No	17 (56.7)	22 (73.3)
Anger-hostility			1.00 (0.33–3.01)	1.00
Yes	9 (30.0)	9 (30.0)
No	21 (70.0)	21 (70.0)
Vigor-activity			0.49 (0.17–1.41)	0.18
Yes	9 (30.0)	14 (46.7)
No	21 (70.0)	16 (53.3)
Fatigue-inertia			2.51 (0.82–7.64)	0.10
Yes	13 (43.3)	7 (23.3)
No	17 (56.7)	23 (76.7)
Confusion-bewilderment			1.72 (0.61–4.84)	0.29
Yes	15 (50)	11 (36.7)
No	15 (50)	19 (63.3)

No statistically significant difference (ANCOVA) was found between SES scores of those in the intervention group and those in the control group at baseline and week 13 (Table 5). A significant difference (F [1, 58] = 6.05; p < 0.05) was, however, found between the intervention group and the control group at week 26. Comparison of the mean scores at week 26 indicates that those in the intervention group (mean, 35.66 ± 36.83) had higher levels of self-efficacy for exercise than those in the control group (mean, 15.30 ± 26.43). Comparing the change in mean scores from baseline to week 13 and week 26, no significant difference was found for self-efficacy (Table 6).

Table 5.

Self-Efficacy for Exercise

Time period	Intervention group (n = 30)	Control group (n = 30)	p-Value	Effect size (Cohen d)
Baseline	33.26 ± 32.13	27.16 ± 29.06	0.44	0.19
Week 13	32.73 ± 31.54	23.30 ± 27.84	0.22	0.31
Week 26	35.66 ± 36.83	15.30 ± 26.43	0.01^a	0.63

Unless otherwise noted, values are mean ± SD.

Significant at p < 0.05.

Table 6.

Mean Change in Self-Efficacy for Exercise from Baseline to Week 13 and Week 16 (n = 60)

Time period	Intervention group	Control group	F-value	p-Value
Week 13 to baseline	−0.53 ± 34.23	−3.86 ± 26.79	0.17	0.67
Week 26 to baseline	2.4 ± 40.93	−11.86 ± 39.51	1.88	0.17

Unless otherwise noted, values are mean ± SD.

Odds ratio estimation (Pearson chi-square) indicated that participants in the intervention group were 2.9 times more likely to show an improvement in their SEE scores than those in the control group (Table 7). A close to significant difference (95% confidence interval, 0.94–8.71; p = 0.05) was found between the improvement in the SEE levels of the intervention (t'ai chi) group (14 of 30 [46.7%]) and the control group (7 of 30 [23.3%]) when the two groups were compared (Table 7). Promising clinical significance of the t'ai chi intervention on self-efficacy is evidenced by the large effect size (Cohen d) of 0.63 at week 26 (Table 5) and the positive odds ratio for self-efficacy scores on the SEE.

Table 7.

Odds Ratio for Self-Efficacy for the Intervention and Control Groups

Improvement in self-efficacy	Intervention group (n = 30), n (%)	Control group (n = 30), n (%)	Odds Ratio (95% confidence interval)	p-Value
			2.87 (0.94–8.71)	0.05^a
Yes	14 (46.7)	7 (23.3)
No	16 (53.3)	23 (76.7)

Significant at p ≤ 0.05.

Discussion

This study examined the effect of a 26-week seated STEP intervention¹¹ on mood states and self-efficacy and found the STEP intervention had several significant effects in older people in wheelchairs living in an LTC facility. The STEP intervention was associated with a reduction of the fatigue-inertia dimension of mood states and an increase in self-efficacy.

The STEP exercise intervention significantly reduced the fatigue-inertia dimension of participants' mood state, as measured by the POMS-SF. Previous studies offer some possible explanation for this finding. Chen et al.⁹ and Chen et al.¹² found that practicing t'ai chi positively enhanced health status, especially that of older people with chronic illness. According to Adler and Roberts,¹⁸ t'ai chi provides the benefits of flexibility and muscle strengthening because it is a slow and gentle form of exercise. Such a view may partially explain this study finding; for example, participants may have felt improvement in their muscle strength as a result of the t'ai chi exercise and then reported this as a lower level of fatigue-inertia. Several studies indicate that t'ai chi affects the body through physical and psychological relaxation.^19
–21 Consequently, the participants in this current study may have experienced physical and psychological relaxation, flexibility, and muscle strengthening through the slow and gentle t'ai chi exercise. Indeed, the positive effects of the interrelationship between the autonomic nervous system and musculoskeletal functions can result in the person achieving a positive health status and therefore a reduction in the sense of fatigue-inertia.

It is difficult to explain the finding of no significant difference between the two groups regarding their perceptions of the other negative mood states: the tension-anxiety, depression-dejection, anger-hostility, and confusion-bewilderment dimensions on the POMS-SF. This may be related to whether participants were willing to express negative mood states openly, especially because these were older Chinese participants and traditional Chinese culture and social norms would be expected to strongly influence their expression of negative mood states.²²

Despite the lack of significant differences between the two groups for the other negative mood states, a trend was noticed: the findings in the STEP group indicate lower negative mood states than in the control group. This suggests that the other negative mood states in the STEP group may have been positively influenced by the intervention.

Another possible reason for the lack of significant difference in vigor-activity is that the intensity of the exercise experience and/or the length of the intervention might not have been sufficient to result in a noticeable improvement in strength, balance, and/or flexibility. Those with cardiovascular disease, for example, may need further practice and participation to result in their cardiovascular circulation and respiratory system function being noticeably and measurably better; that, in turn, would give them a more positive mood state, which would then be reflected in their responses to questions measuring the vigor-activity dimension of the POMS-SF.

The STEP intervention was associated with significantly increased self-efficacy. This finding is consistent with those of previous studies, which measured a significant improvement in the self-efficacy of older people who did some form of t'ai chi.^7,22,23

The results, however, need to be interpreted with some caution because the current study used a convenience sample recruited from one LTC facility in the Changhua region of Taiwan; this limits the generalizability of the findings. Further limitations of the design include no monitoring of the level or extent of the participants' learning of seated t'ai chi. In addition, the lack of a follow-up assessment at the end of the program impeded an understanding of the sustainability of the program. Moreover, participants could not be blinded to the treatment they received, and one outcome assessor was not blinded to study treatment.

Conclusions

These results contribute to the growing area of knowledge about seated t'ai chi and offer staff working in LTC facilities a practical way to help older people in wheelchairs living in these facilities to engage in exercise. The use of a seated t'ai chi exercise program as a psychological health maintenance activity and/or social activity for older people in wheelchairs in LTC facilities should be supported.

Footnotes

Acknowledgments

The authors acknowledge participation of the residents as well as staff cooperation.

Author Disclosure Statement

No competing financial interests exist.

References

Mortenson

, Miller

, Backman

, Oliffe

. Association between mobility, participation, and wheel-chair-related factors in long-term care residents who use wheelchairs as their primary means of mobility. J Am Geriatr Soc, 2012; 60:1310–1315.

. Usage survey of locomotive assistive devices for long-term care clients. Formosan J Physic Ther, 2004; 29:405–420.

Penedo

, Dahn

. Exercise and well-being: a review of mental and physical benefits associated with physical activity. Curr Opin Psychiatry, 2005; 18:189–193.

Tabloski

. Gerontological Nursing, 2nd ed. Upper Saddle River, NJ: Pearson, 2010.

Kerse

, Falloon

, Moyes

, et al. DeLLITE depression in late life: an intervention trial of exercise. Design and recruitment of a randomised controlled trial. BMC Geriatr, 2008; 8:12–12.

Waughfield

ed. Mental Health Concepts, 5th ed. Scarborough, Ontario, Canada: Delmar, 2002.

Dechamps

, Onifade

, Decamps

, et al. Health-related quality of life in frail institutionalized elderly: effects of a cognition-action intervention and Tai Chi. J Aging Phys Act, 2009; 17:236–248.

Lee

, Perng

, Ho

, et al. A preliminary reliability and validity study of the Chinese version of the self-efficacy for exercise scale for older adults. Int J Nurs Stud, 2009; 46:230–238.

Chen

, Snyder

, Krichbaum

. Tai chi and well-being of Taiwanese community-dwelling elders. Clin Gerontol, 2002; 24:137–156.

10.

Chen

, Hsu

, Chen

. et al. Well-being of institutionalized elders after Yang-style Tai Chi practice. J Clin Nurs, 2007; 16:845–852.

11.

Chen

, Chen

, Huang

. Development of the simplified Tai Chi exercise program (STEP) for frail older adults. Complement Ther Med, 2006; 14:200–206.

12.

Chen

, Lin

, et al. The effects of a Simplified Tai-Chi Exercise Program (STEP) on the physical health of older adults living in long-term care facilities: a single group design with multiple time points. Int J Nurs Stud, 2008; 45:501–507.

13.

American Council on Exercise. The best time to exercise. Online document at: http://www.bellinghamfitness.com/downloads/itemid_53.pdf. Accessed April 12, 2015 .

14.

McNair

, Lorr

, Droppleman

. Profile of Mood States (POMS) Manual. San Diego, CA: Educational and Industrial Testing Service, 1992.

15.

Resnick

, Jenkins

. Testing the reliability and validity of the self-efficacy for exercise scale. Nurs Res, 2000; 49:154–159.

16.

Chen

, Snyder

, Krichbaum

. Translation and equivalence: the Profile of Mood States Short Form in English and Chinese. Int J Nurs Stud, 2002; 39:619–624.

17.

Tsikriktsis

. A review of techniques for treating missing data in OM survey research. J Oper Manag, 2005; 24:53–62.

18.

Adler

, Roberts

. The use of Tai Chi to improve health in older adults. Orthop Nurs, 2006; 25:122–126.

19.

Wang

. Tai-chi chuan and health promotion. Taiwan Geriatr Gerontol, 2009; 4:116–129.

20.

Saravanakumar

, Higgins

, van der Riet

, Marquez

, Sibbritt

. The influence of tai chi and yoga on balance and falls in a residential care setting: a randomised controlled trial. Contemp Nurse, 2014; 48:76–87.

21.

Yang

, Hao

, Tian

, et al. The effectiveness of Tai Chi for patients with Parkinson's disease: study protocol for a randomized controlled trial. Trials, 2015; 16:111.

22.

Taylor-Piliae

, Haskell

, Waters

, et al. Change in perceived psychosocial status following a 12-week Tai Chi exercise program. J Adv Nurs, 2006; 54:313–329.

23.

Yeh

, McCarthy

, Wayne

, et al. Tai chi exercise in patients with chronic heart failure: a randomized clinical trial. Arch Intern Med, 2011; 171:750–757.