Abstract
Objective
To assess whether tai chi is non-inferior to cognitive behavioural therapy for insomnia (CBT-I), the first line treatment, for managing chronic insomnia in middle aged and older adults.
Design
Randomised, assessor masked, non-inferiority trial.
Setting
A single research site in Hong Kong with participants recruited from the local community between 18 May 2020 and 14 July 2022.
Participants
Chinese participants aged ≥50 years with chronic insomnia diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, fifth edition.
Interventions
Tai chi and CBT-I interventions were delivered in group format over three months, consisting of one hour sessions twice a week for a total of 24 sessions.
Main outcome measures
The primary outcome was the change in perceived insomnia severity measured by the Insomnia Severity Index after the intervention (month 3) and at 12 month follow-up (month 15). To assess whether tai chi is non-inferior to CBT-I, a threshold of four points was used as the margin of non-inferiority.
Results
200 participants were randomised (1:1) to receive tai chi (n=100) or CBT-I (n=100). The per protocol principle was adopted. At month 3, the tai chi group showed a reduction of 6.67 (95% confidence interval 5.61 to 7.73) in Insomnia Severity Index scores, while the CBT-I group had a reduction of 11.19 (10.06 to 12.32), resulting in a between group difference of 4.52 (−∞ to 5.81). Tai chi was deemed inferior to CBT-I at month 3 because the upper confidence limit exceeded the non-inferiority margin. At month 15, the reductions for tai chi and CBT-I were 9.51 (8.47 to 10.54) and 10.18 (8.97 to 11.40), respectively, with a between group difference of 0.68 (−∞ to 2.00). At this point, tai chi was considered non-inferior to CBT-I because the upper confidence limit fell within the non-inferiority margin. Results from the intention-to-treat analysis were consistent with the per protocol findings. No adverse events occurred during the intervention period.
Conclusion
Tai chi was inferior to CBT-I at month 3 but non-inferior at month 15. This finding supports the use of tai chi as an alternative approach for the long term management of chronic insomnia in middle aged and older adults.
Trial registration
ClinicalTrials.gov NCT04384822
Introduction
Chronic insomnia is one of the most common sleep disorders in middle aged and older adults, with global prevalence rates ranging from 4% to 22%, depending on the diagnostic criteria used.1 In Hong Kong, 30-50% of middle aged and older adults report having insomnia problems, with old age considered to be a main predictor.2 3 Chronic insomnia has a median duration of three years, with up to 74% of patients experiencing persistent symptoms over the past year.4 Chronic insomnia in middle aged and older adults has been associated with increased risk of cardiovascular diseases, mental disorders, and cognitive impairment, which prospectively increases hospital admission and mortality rates.5 Therefore, a large financial burden is placed on healthcare and socioeconomic systems, with the direct and indirect costs of insomnia treatments estimated to be $150bn (£114bn; €130bn) each year in the United States alone.6
Cognitive behavioural therapy for insomnia (CBT-I) is the first line treatment for chronic insomnia owing to its excellent treatment efficacy and minimal side effects.7 However, accessibility to CBT-I is often limited because of the associated high costs and the low availability of trained therapists, preventing its wider implementation in the community.8 In the US, it was reported that only 4% of primary care providers working in Veterans Affairs healthcare used CBT-I in their practice,9 while less than 10% of patients with insomnia at university medical centres were referred for CBT-I.10 Data from 12 European countries indicate that only 4-10% of patients are prescribed non-pharmacological insomnia treatments such as CBT-I.11 Limited access in primary care settings and insufficient education on the therapeutic potential of CBT-I have been recognised as key factors hindering the use of CBT-I in the community.9 10 Therefore, a pressing need exists to explore alternative treatment approaches that are effective, accessible, and inexpensive to meet the demands of the increasing number of patients with chronic insomnia. Considering the high persistence of insomnia symptoms,4 treatment approaches with sustainable efficacy should also be explored to improve the long term management of insomnia.5
Tai chi, a form of mind-body exercise that is widely practiced in Chinese communities, is becoming increasingly popular in “western” countries.12 Each year, more than 100 cities in over 80 countries recognise and celebrate the World Tai Chi and Qigong Day, reflecting the global popularity of tai chi and mind-body exercises.13 14 One recent survey in the US reported that approximately 4.05 million people reported practicing tai chi or other mind-body exercises over the past year for health purposes.15 Tai chi is perceived as a suitable exercise modality for middle aged and older adults, even among those who are inactive or unfit.16 17 Tai chi is inexpensive, easily accessible, and amenable to middle aged and older adults, which can enable its large scale implementation in the community. Previous clinical trials and meta-analyses support the beneficial effects of tai chi in middle aged and older adults with insomnia, with improvements that can be sustained for up to 24 months.16 18 19 However, most studies used passive control designs (eg, stretching) or compared it with other exercise modalities (eg, aerobic exercise), while direct comparisons with first line insomnia treatments (eg, CBT-I) in middle aged and older adults with primary chronic insomnia are currently lacking. Therefore, we need studies assessing the clinical efficacy of tai chi as an alternative insomnia treatment to manage chronic insomnia in middle aged and older adults.
To validate the use of tai chi as an alternative approach to manage chronic insomnia in middle aged and older adults, this study aimed to examine whether a three month tai chi intervention (experimental treatment) is non-inferior to a three month CBT-I intervention (first line treatment for chronic insomnia) for treating insomnia; and whether tai chi is non-inferior to CBT-I in sustaining beneficial improvements 12 months after the intervention. A non-inferiority evaluation was used to determine the acute and long term treatment efficacy of tai chi in improving sleep to establish its potential as an alternative approach for managing chronic insomnia in middle aged and older adults. We hypothesised that tai chi is non-inferior to CBT-I for treating insomnia in middle aged and older adults, and sustaining the sleep improvements 12 months after the intervention.
Methods
Study design
This randomised, assessor blinded, non-inferiority trial adhered to the Consolidated Standards of Reporting Trials (CONSORT) extension for non-inferiority and equivalence trials.20 Participants were enrolled between 18 May 2020 and 14 July 2022, with the last follow-up ending on 2 November 2023. The study was approved by the University of Hong Kong/Hospital Authority Hong Kong West institutional review board (IRB approval No UW 18-621). Written informed consent was obtained on a voluntary basis before baseline assessments. The study was conducted in accordance with the Declaration of Helsinki. The study was prospectively registered at ClinicalTrials.gov (NCT04384822) on 12 May 2020. Details about the protocol, including outcomes and intervention content, are further described in the trial protocol paper published on 4 November 2022 (supplementary material I).21
Participants and settings
This study was conducted at a single research site in Hong Kong. Participants were recruited from the community using study leaflets and posters. To be included, participants needed to be middle aged and older adults aged ≥50 years who were ethnic Chinese and had a diagnosis of chronic insomnia (including difficulty in getting to sleep or maintaining sleep, early morning awakening, complaints of impaired daytime functioning, and sleep disturbances occurring at least three nights each week and lasting for at least three months) according to the Diagnostic and Statistical Manual of Mental Disorders, fifth edition. Exclusion criteria were the following: inability to walk without an assistive device (eg, a cane); somatic condition that limits tai chi participation (eg, limb loss); regular participation in aerobic exercise or mind-body exercise such as tai chi, yoga, qigong, or meditation (more than three times a week for >60 minutes each session); severe chronic disease that may profoundly affect sleep (eg, active cancer, end stage renal disease); receiving treatments or drugs that can affect sleep (eg, chemotherapy for cancer); dementia or receiving dementia treatments; untreated sleep disorder such as obstructive sleep apnea; current or past CBT-I treatment; and shift worker. Participants with common physical or mental chronic conditions, or both, including central obesity, hypertension, diabetes, depression and anxiety, were still included in the study. However, participants who might be disruptive during the group sessions, such as those showing uncontrolled violence or psychotic symptoms, were excluded.
Randomisation and masking
Eligible participants were randomly assigned to the two intervention groups in a 1:1 ratio. Randomisation was conducted using an online random generator (https://www.sealedenvelope.com/) with block sizes of four to six. The computer generated randomised allocation sequence was securely placed in sealed opaque envelopes and kept by an independent researcher who did not interact with the participants during recruitment or during the interventions. A second independent researcher who was responsible for conducting the group allocation needed to contact the first independent researcher to retrieve the allocation sequence after the baseline assessments and before the start of the interventions. These two independent researchers were not involved in participant recruitment or outcome assessment. Participants and intervention instructors were not blinded to the group allocation because of the nature of the tai chi and CBT-I interventions. Outcome assessors were blinded to the group allocation and participants were instructed not to disclose their group allocation to the outcome assessors during the outcome assessments.
Interventions
Tai chi
Participants in the tai chi group attended a three month, instructor led training programme, with 8-12 participants in each group, delivered by two certified tai chi instructors. The programme adopted the 24 form Yang style of tai chi, the most common style adopted in the literature.22 We examined the feasibility and efficacy of this intervention in improving sleep in older adults with chronic insomnia in our previously published randomised controlled trial.16 The intervention was prescribed as a three month programme consisting of a one hour session twice a week for a total of 24 sessions. The tai chi intervention was delivered in the first two months followed by one month of consolidation. The instructors discussed safety issues, proper training principles, and introduced the skills in the first class to minimise any avoidable adverse events caused by improper skills or practice. Each tai chi session consisted of 10 minutes of breathing and relaxation as warm-up exercises, followed by 45 minutes of tai chi practice, and then five minutes of cool-down exercises. Supplementary material II gives details of the 24 sessions of the tai chi intervention.
Cognitive behavioural therapy for insomnia
Participants in the CBT-I group attended a three month, instructor led training programme, with 8-12 participants in each group, delivered by two certified CBT-I therapists. We also evaluated the feasibility and efficacy of this therapy in our previous study.23 The intervention covers typical CBT-I components, including sleep education, stimulus control, sleep restriction, relaxation training, and cognitive therapy. The intervention was prescribed as a three month programme with a one hour session twice a week for a total of 24 sessions. The intervention was delivered in the first two months followed by one month of consolidation. Supplementary material II gives details of the 24 sessions of the CBT-I intervention.
Both interventions took place at the School of Public Health, Li Ka Shing Faculty of Medicine, the University of Hong Kong. Attendance was recorded by the instructors as a measure of treatment adherence.
Outcomes
All outcomes were assessed at baseline, after the intervention (month 3), and at 12 month follow-up (month 15).
Primary outcome
The primary outcome was perceived insomnia severity measured using the Insomnia Severity Index (ISI) at month 3 and month 15. The ISI includes seven items assessing the following: sleep onset, sleep maintenance difficulties, satisfaction with the current sleep pattern, interference with daily functioning, noticeable impairment owing to sleep problems, degree of distress, and concerns caused by sleep problems. Each item is rated on a five point Likert scale (eg, 0=no problem; 4=very severe problem), yielding a total score ranging from 0 to 28. The total score is interpreted as follows: absence of insomnia (0-7); subthreshold insomnia (8-14); moderate insomnia (15-21); and severe insomnia (22-28).24 A score change ≥8 has been established as the minimally important difference with 60% sensitivity and 70% specificity to detect moderate improvement.25 The Chinese version of the ISI has been shown to have a satisfactory content validity index of 0.94 and high internal consistency, with a Cronbach’s alpha of 0.81.26
Secondary outcomes
Secondary outcomes included insomnia remission rate (defined as “no longer meeting the criteria for insomnia” according to the Diagnostic and Statistical Manual of Mental Disorders, fifth edition); insomnia treatment response rate (defined as a decrease in ISI of ≥8 points25); subjective sleep quantity and quality measured by the Pittsburgh Sleep Quality Index; seven day actigraphy; seven day sleep diary; sleep drug use; health related quality of life measured by the Short Form (SF) 12 item version 2 (SF-12v2); body balance and lower extremity function measured by the Short Physical Performance Battery; physical activity level measured by the International Physical Activity Questionnaire Short Form; severity of symptoms of depression and anxiety measured by the Hospital Anxiety and Depression Scale; and total energy intakes measured by a three day food diary.
Statistical analysis
Sample size
We conducted sample size estimation for a two sample t test between tai chi and CBT-I interventions to detect non-inferiority using POWER in SAS OnDemand for Academics (SAS Institute). The sample size estimation was based on a primary comparison of perceived insomnia severity measured by ISI between the two groups using a 95% power and 0.05/2=0.025 maximum chance of committing false type I errors to account for multiplicity. To assess whether tai chi is non-inferior to CBT-I, we adopted an ISI threshold of 4 points as the margin of non-inferiority, which is 50% of the minimally important difference in the ISI (reported as 8 points).25 The margin of 4 points in the ISI has been adopted in several non-inferiority trials conducted in different populations with chronic insomnia.23 27 Our pilot study indicated that the tai chi intervention resulted in a mean reduction of 5.05 points in the ISI.16 Assuming that the CBT-I group would also induce a similar improvement in the ISI with standard deviation conservatively taken as 6.5,16 we needed a total of 70 participants in each group. Expecting a 20% study dropout rate, we aimed to recruit 90 participants (rounded up from 88) for each of the tai chi and CBT-I interventions, giving an overall study sample of 180 participants. Eventually, a total of 200 participants were enrolled in this study.
Data analysis
A non-inferiority analysis was performed on the primary outcome by comparing differences in the changes in ISI scores between the two groups at month 3 and month 15. Per protocol analyses were used to provide a more conservative and stringent non-inferiority comparison.28 A margin of 4 points in the ISI (50% of minimally important difference) was adopted to assess whether tai chi was non-inferior to CBT-I.23 27 The per protocol approach included all the participants who completed the assessments at the specific time point (ie, month 3 or month 15). For instance, for the non-inferiority analysis at month 3, only participants who completed month 3 assessments were included. Additionally, we adopted another per protocol approach that only included those who completed both month 3 and month 15 assessments.
We used generalised estimating equations analyses to examine the treatment effects on the quantitative secondary outcomes, adjusting for baseline values.16 17 Pairwise comparisons were performed using linear contrasts. We conducted logistic regression to analyse insomnia remission rate and treatment response rate. The intention-to-treat principle was adopted in the secondary outcomes analyses for all randomised participants, except for insomnia remission and treatment response rates because logistic regression cannot account for missing data. All statistical analyses were performed using SAS OnDemand for Academics (SAS Institute). Statistical significance was set at a P value <0.05.
Sensitivity analysis
We performed sensitivity analysis using three approaches to assess the robustness of the results of the primary outcome. The first approach was a more stringent per protocol analysis of participants who completed the assessments at all three time points. The other two approaches followed the intention-to-treat principle by replacing missing values using the last observation carried forward and multiple imputation methods, respectively. In the last observation carried forward approach, missing values were replaced by the last observed value for the same participant. In the multiple imputation approach, a total of 20 imputed datasets were generated using the fully conditional specification method. We included the following variables in the multiple imputation model: subject code, time, group, demographic factors (eg, sex), primary outcome (ie, ISI), and other secondary outcomes (eg, Pittsburgh Sleep Quality Index). Non-inferiority analysis was performed on each generated dataset and the results were then pooled according to Rubin’s rules. Additionally, we conducted sensitivity analyses on the primary outcome adjusting for baseline variables that significantly differed between those who completed and those who did not complete the outcome assessments.
Subgroup analysis
We conducted generalised estimating equations analysis to examine whether the changes in primary outcome were associated with age, sex, and body mass index, all of which are strongly associated with insomnia. The differences in mean changes and 95% confidence intervals were presented for each subgroup based on age (50-59 years, 60-69 years, and 70 years and older), sex (male and female), and body mass index categories (normal weight, overweight, and obese). We applied the per protocol principle.
Patient and public involvement
Patients were not involved in formulating the research question, selecting outcome measures, or creating recruitment and implementation strategies. They were also not consulted on interpreting or reporting the results. Nevertheless, patient engagement during the trial played a crucial role in its successful completion.
Results
Baseline characteristics of participants
A total of 272 people registered for the study and underwent eligibility screening. Of these, 200 eligible participants provided written informed consent and were included in the study. They were randomly allocated in a 1:1 ratio to the tai chi (n=100) or CBT-I (n=100) group (fig 1). Table 1 shows baseline characteristics of all participants. Participants in both groups presented with moderate levels of insomnia severity, as indicated by their ISI scores. Among the 200 participants, 166 (83%; 82 in the tai chi group and 84 in the CBT-I group) completed the assessments at month 3, and 167 (83.5%; 85 in the tai chi group and 82 in the CBT-I group) completed the assessments at month 15. Baseline characteristics were compared among participants who completed the assessments at month 3 and month 15, and those who did not. We observed a statistically significant difference in the baseline physical component score in SF-12v2 between those who completed and those who did not complete the assessment at month 3. Additionally, there were statistically significant differences in the baseline sleep parameters recorded in sleep diaries—including total time in bed, wake after sleep onset, and sleep efficiency—between those who completed and those who did not complete the assessment at month 15 (supplementary material IV). The adherence rates were similar for the tai chi and CBT-I groups (78.2% and 80.7%, respectively), as measured by class attendance.
Fig 1.
Flow chart showing screening, randomisation, interventions, and follow-up of participants. CBT-I=cognitive behavioural therapy for insomnia
Table 1.
Baseline characteristics of participants
| Characteristics | Tai chi (n=100) | CBT-I (n=100) |
|---|---|---|
| Female, n (%) | 77 (77) | 84 (84) |
| Age (years) | 64.83 (6.31) | 63.76 (6.15) |
| Education level, n (%) | ||
| Below tertiary education | 73 (73) | 70 (70) |
| Tertiary education or above | 27 (27) | 30 (30) |
| Marital status, n (%) | ||
| Single | 15 (15) | 21 (21) |
| Married | 66 (66) | 64 (64) |
| Divorced | 13 (13) | 8 (8) |
| Widowed | 6 (6) | 7 (7) |
| Monthly income ($), n (%) | ||
| <20 000 | 12 (12) | 10 (10) |
| 20 000-50 000 | 5 (5) | 5 (5) |
| >50 000 | 3 (3) | 2 (2) |
| Retired | 79 (79) | 82 (82) |
| Undisclosed | 1 (1) | 1 (1) |
| Presented with physical comorbidities, n (%) | 62 (62) | 59 (59) |
| Presented with mental comorbidities, n (%) | 13 (13) | 7 (7) |
| Insomnia Severity Index | 16.45 (3.69) | 17.41 (4.42) |
| Pittsburgh Sleep Quality Index | 13.13 (2.68) | 13.79 (2.61) |
| Sleep diary | ||
| Sleep onset latency (min) | 45.69 (43.49) | 52.49 (47.11) |
| Wake after sleep onset (min) | 89.91 (71.76) | 103.77 (92.70) |
| No of awakenings | 2.33 (1.12) | 2.46 (1.40) |
| Average wake-up time (min) | 43.40 (40.64) | 49.00 (44.47) |
| Total sleep time (min) | 332.40 (73.37) | 323.52 (94.46) |
| Total time in bed (min) | 468.00 (79.51) | 479.79 (76.32) |
| Sleep efficiency (%) | 72.00 (16.05) | 68.22 (19.00) |
| Actigraph | ||
| Sleep onset latency (min) | 7.55 (4.13) | 7.96 (4.17) |
| Wake after sleep onset (min) | 70.82 (34.92) | 81.23 (37.28) |
| No of awakenings | 19.92 (7.09) | 20.63 (6.86) |
| Average wake-up time (min) | 3.58 (1.04) | 4.03 (2.08) |
| Total sleep time (min) | 386.88 (63.52) | 390.25 (72.19) |
| Sleep efficiency (%) | 83.07 (7.59) | 81.46 (7.53) |
| Hypnotic drug use* | 7.38 (12.35) | 8.57 (10.17) |
| SF-12v2 | ||
| Physical component score | 44.09 (8.33) | 43.47 (8.39) |
| Mental component score | 45.19 (8.56) | 43.49 (10.98) |
| IPAQ-SF, activity† | 1974.47 (1447.02) | 1603.50 (1230.41) |
| IPAQ-SF, sedentary behaviour† | 383.85 (161.33) | 390.00 (170.08) |
| Short Physical Performance Battery | 11.02 (1.36) | 11.09 (1.37) |
| Hospital Anxiety and Depression Scale | ||
| Anxiety subscale | 7.08 (3.55) | 7.89 (4.31) |
| Depression subscale | 6.91 (3.66) | 7.45 (4.19) |
| Total energy intake (kcal/day) | 2072.03 (856.70) | 1869.30 (555.38) |
£1.00=$1.32, €1.14. Data are presented as mean (standard deviation) unless specified otherwise.
Physical comorbidities mainly included hyperlipidaemia, hyperuricemia, hyperthyroidism, osteoporosis, prediabetes and diabetes, and hypertension. Mental comorbidities included depression and anxiety.
CBT-I=cognitive behavioural therapy for insomnia; IPAQ-SF=International Physical Activity Questionnaire—Short Form; SF-12v2=Short Form 12 item, version 2.
Thirty participants in tai chi group and 26 participants in CBT-I group were taking hypnotic drugs (lowest recommended dosage, units/week).
Metabolic equivalent minutes per week.
Primary outcome
Figure 2 and supplementary material III present the results of the primary outcome. The per protocol analysis showed that, at month 3, the tai chi group showed a reduction of 6.67 points (95% confidence interval 5.61 to 7.73) in ISI scores, while the CBT-I group had a reduction of 11.19 (10.06 to 12.32), resulting in a between group difference of 4.52 (−∞ to 5.81). Tai chi was deemed inferior to CBT-I at month 3 because the upper confidence limit exceeded the non-inferiority margin. At month 15, the reductions for the tai chi and CBT-I groups were 9.51 (8.47 to 10.54) and 10.18 (8.97 to 11.40), respectively, with a between group difference of 0.68 (−∞ to 2.00). At this point, tai chi was considered non-inferior to CBT-I because the upper confidence limit fell within the non-inferiority margin.
Fig 2.
Primary outcome (per protocol analysis): non-inferiority analysis for Insomnia Severity Index (ISI) at months 3 and 15 in tai chi and cognitive behavioural therapy for insomnia (CBT-I) groups. Sensitivity analyses: completed assessments, intention-to-treat last observation carried forward (LOCF), and intention-to-treat pooled multiple imputation. First approach (completed assessments) was more stringent per protocol analysis of participants who completed assessments at all time points. For LOCF, missing values replaced by last observed value for same participant. For multiple imputation, 20 imputed datasets generated using fully conditional specification method (variables were subject code, time, group, personal factors (eg, sex), primary outcome (ie, ISI), and other secondary outcomes (eg, Pittsburgh Sleep Quality Index)). Non-inferiority analysis performed on each generated dataset and results pooled according to Rubin’s rules. Dots indicate differences in changes in mean ISI scores between two groups; dotted line represents non-inferiority margin of four points of ISI. Upper confidence limit not exceeding margin represents non-inferiority. Supplementary material III includes table presenting primary outcome analysis
Results of the intention-to-treat analysis were consistent with the per protocol findings. At month 3, the intention-to-treat analysis using the multiple imputation method showed that there was a 6.88 (95% confidence interval 5.84 to 7.90) and 10.73 (9.53 to 11.93) reduction in the ISI scores for the tai chi and CBT-I groups, respectively, with a between group difference of 3.85 (−∞ to 5.46). Tai chi was considered inferior to CBT-I at month 3 because the upper confidence limit exceeded the predefined non-inferiority margin. At month 15, there was a 9.43 (8.37 to 10.49) and 10.15 (8.94 to 11.35) reduction in the ISI scores for the tai chi and CBT-I groups, respectively, with a between group difference of 0.71 (−∞ to 2.28). Tai chi was considered non-inferior to CBT-I at month 15 because the upper confidence limit was within the predefined non-inferiority margin.
Secondary outcomes
A substantial proportion of participants in the tai chi and CBT-I groups achieved remission from chronic insomnia or showed a positive treatment response. The insomnia remission rates differed between the CBT-I and tai chi groups at month 3 (83.3% v 56.1%, P<0.001) but were similar at month 15 (63.4% v 76.5%, P=0.067). The treatment response rates also differed between the CBT-I and tai chi groups at month 3 (77.4% v 43.9%, P<0.001) but were similar at month 15 (73.2% v 62.4%, P=0.137; fig 3).
Fig 3.
Insomnia remission rate (definition of chronic insomnia from Diagnostic and Statistical Manual of Mental Disorders, fifth edition) and treatment response rate (defined as decrease of ≥8 points on Insomnia Severity Index score) shown as percentage of participants. Error bars indicate range of remission or treatment response assuming that missing data all relate to participants who had remission/response or did not have remission/response. Data analysed by logistic regression. CBT-I=cognitive behavioural therapy for insomnia
No statistically significant group by time interaction effects were observed in all secondary outcomes, indicating that there were no between group differences in the changes in these outcomes over time between the tai chi and CBT-I groups (table 2, table 3).
Table 2.
Primary and secondary outcomes assessed at baseline, after the intervention (month 3), and at 12 month follow-up (month 15)
| Outcome | Assessment | Cohen’s d† | Group effect | Time effect | Group by time interaction effect | ||||
|---|---|---|---|---|---|---|---|---|---|
| Baseline | Month 3 | Month 15 | Month 3 | Month 15 | |||||
| Primary outcome | |||||||||
| Insomnia Severity Index | — | — | — | — | — | <0.001* | <0.001* | 0.74 | |
| Tai chi | 16.45 (3.69) | 9.66 (5.24) | 7.09 (4.31) | −1.50 | −2.33 | — | — | — | |
| CBT-I | 17.41 (4.42) | 6.14 (4.35) | 7.54 (4.53) | −2.57 | −2.21 | — | — | — | |
| Secondary outcomes | |||||||||
| Pittsburgh Sleep Quality Index | — | — | — | — | — | <0.001* | <0.001* | 0.40 | |
| Tai chi | 13.13 (2.68) | 8.38 (3.62) | 6.48 (3.54) | −1.49 | −2.12 | — | — | — | |
| CBT-I | 13.79 (2.61) | 6.23 (2.89) | 7.04 (3.46) | −2.74 | −2.20 | — | — | — | |
| Sleep diary | |||||||||
| Sleep onset latency (min) | — | — | — | — | — | 0.01* | <0.001* | 0.67 | |
| Tai chi | 45.69 (43.49) | 35.59 (34.25) | 31.64 (24.22) | −0.26 | −0.40 | — | — | — | |
| CBT-I | 52.49 (47.11) | 22.33 (19.18) | 31.16 (31.70) | −0.84 | −0.53 | — | — | — | |
| Wake after sleep onset (min) | — | — | — | — | — | 0.08 | <0.001* | 0.75 | |
| Tai chi | 89.91 (71.76) | 71.58 (61.50) | 45.27 (39.34) | −0.27 | −0.77 | — | — | — | |
| CBT-I | 103.77 (92.70) | 49.44 (48.40) | 49.84 (53.95) | −0.73 | −0.71 | — | — | — | |
| No of awakenings | — | — | — | — | — | 0.64 | <0.001* | 0.88 | |
| Tai chi | 2.33 (1.12) | 1.83 (0.90) | 1.70 (0.93) | −0.49 | −0.61 | — | — | — | |
| CBT-I | 2.46 (1.40) | 1.77 (1.32) | 1.73 (1.15) | −0.51 | −0.57 | — | — | — | |
| Average wake-up time (min) | — | — | — | — | — | 0.17 | 0.001* | 0.91 | |
| Tai chi | 43.40 (40.64) | 45.88 (49.48) | 30.90 (32.41) | 0.05 | −0.34 | — | — | — | |
| CBT-I | 49.00 (44.47) | 36.03 (39.73) | 31.63 (35.14) | −0.31 | −0.43 | — | — | — | |
| Total sleep time (min) | — | — | — | — | — | 0.94 | <0.001* | 0.13 | |
| Tai chi | 332.40 (73.37) | 368.97 (68.88) | 395.31 (59.35) | 0.51 | 0.94 | — | — | — | |
| CBT-I | 323.52 (94.46) | 366.45 (58.79) | 373.74 (65.44) | 0.55 | 0.62 | — | — | — | |
| Total time in bed (min) | — | — | — | — | — | <0.001* | 0.21 | 0.19 | |
| Tai chi | 468.00 (79.51) | 476.14 (64.97) | 472.21 (54.78) | 0.11 | 0.06 | — | — | — | |
| CBT-I | 479.79 (76.32) | 438.23 (62.39) | 454.74 (60.22) | −0.60 | −0.36 | — | — | — | |
| Sleep efficiency (%) | — | — | — | — | — | 0.009* | <0.001* | 0.80 | |
| Tai chi | 72.00 (16.05) | 77.97 (13.39) | 84.02 (10.26) | 0.40 | 0.89 | — | — | — | |
| CBT-I | 68.22 (19.00) | 84.15 (10.81) | 82.73 (12.78) | 1.03 | 0.90 | — | — | — | |
| Actigraph | |||||||||
| Sleep onset latency (min) | — | — | — | — | — | 0.006* | 0.03* | 0.56 | |
| Tai chi | 7.55 (4.13) | 8.47 (4.20) | 6.96 (2.68) | 0.22 | −0.17 | — | — | — | |
| CBT-I | 7.96 (4.17) | 6.59 (2.09) | 6.86 (1.35) | −0.42 | −0.35 | — | — | — | |
| Wake after sleep onset (min) | — | — | — | — | — | 0.03* | 0.36 | 0.36 | |
| Tai chi | 70.82 (34.92) | 76.29 (35.09) | 74.19 (33.74) | 0.16 | 0.10 | — | — | — | |
| CBT-I | 81.23 (37.28) | 61.71 (27.48) | 70.12 (29.02) | −0.60 | −0.33 | — | — | — | |
| No of awakenings | — | — | — | — | — | 0.06 | 0.12 | 0.93 | |
| Tai chi | 19.92 (7.09) | 20.11 (7.25) | 21.13 (7.78) | 0.03 | 0.16 | — | — | — | |
| CBT-I | 20.63 (6.86) | 18.88 (6.95) | 20.59 (7.18) | −0.25 | −0.01 | — | — | — | |
| Average wake-up time (min) | — | — | — | — | — | 0.04* | 0.05 | 0.60 | |
| Tai chi | 3.58 (1.04) | 3.89 (1.82) | 3.52 (0.89) | 0.21 | −0.06 | — | — | — | |
| CBT-I | 4.03 (2.08) | 3.29 (0.93) | 3.47 (1.13) | −0.46 | −0.33 | — | — | — | |
| Total sleep time (min) | — | — | — | — | — | 0.19 | 0.18 | 0.84 | |
| Tai chi | 386.88 (63.52) | 386.69 (59.67) | 379.22 (58.63) | −0.003 | −0.13 | — | — | — | |
| CBT-I | 390.25 (72.19) | 372.41 (63.14) | 372.21 (51.67) | −0.26 | −0.29 | — | — | — | |
| Sleep efficiency (%) | — | — | — | — | — | 0.07 | 0.61 | 0.68 | |
| Tai chi | 83.07 (7.59) | 82.16 (7.22) | 82.59 (6.17) | −0.12 | −0.07 | — | — | — | |
| CBT-I | 81.46 (7.54) | 84.51 (5.77) | 82.97 (6.00) | 0.45 | 0.22 | — | — | — | |
Data are presented as estimated mean (standard deviation), baseline covaried. CBT-I=cognitive behavioural therapy for insomnia.
Statistical significance.
Effect size: change from baseline to 3 months and 15 months.
Table 3.
Further secondary outcomes assessed at baseline, after the intervention (month 3), and at 12 month follow-up (month 15)
| Secondary outcomes | Assessment | Cohen’s d† | Group effect | Time effect | Group by time interaction effect | ||||
|---|---|---|---|---|---|---|---|---|---|
| Baseline | Month 3 | Month 15 | Month 3 | Month 15 | |||||
| Hypnotic drug use‡ | — | — | — | — | — | 0.009 | 0.05 | 0.65 | |
| Tai chi (n=30) | 7.38 (12.35) | 1.80 (2.42) | 5.79 (13.86) | −0.63 | −0.12 | — | — | — | |
| CBT-I (n=26) | 8.57 (10.17) | 7.37 (11.20) | 4.93 (7.40) | −0.11 | −0.41 | — | — | — | |
| SF-12v2 | |||||||||
| Physical component score | — | — | — | — | — | 0.72 | 0.05 | 0.98 | |
| Tai chi | 44.09 (8.33) | 47.80 (6.70) | 47.00 (9.19) | 0.49 | 0.33 | — | — | — | |
| CBT-I | 43.47 (8.39) | 47.13 (8.22) | 46.65 (8.80) | 0.44 | 0.37 | — | — | — | |
| Mental component score | — | — | — | — | — | 0.30 | <0.001* | 0.99 | |
| Tai chi | 45.19 (8.56) | 50.41 (7.20) | 52.83 (7.71) | 0.66 | 0.94 | — | — | — | |
| CBT-I | 43.49 (10.98) | 51.52 (7.84) | 52.05 (8.67) | 0.84 | 0.87 | — | — | — | |
| IPAQ-SF, activity§ | — | — | — | — | — | 0.06 | <0.001* | 0.76 | |
| Tai chi | 1974.47 (1447.02) | 3236.22 (2165.24) | 3215.42 (2411.85) | 0.69 | 0.62 | — | — | — | |
| CBT-I | 1603.50 (1230.41) | 2584.69 (2040.69) | 2739.68 (1873.17) | 0.58 | 0.72 | — | — | — | |
| IPAQ-SF, sedentary behaviour§ | — | — | — | — | — | 0.46 | 0.02* | 0.09 | |
| Tai chi | 383.85 (161.33) | 324.09 (148.29) | 367.47 (163.96) | −0.39 | −0.10 | — | — | — | |
| CBT-I | 390.00 (170.08) | 355.36 (179.77) | 418.96 (192.55) | −0.20 | 0.16 | — | — | — | |
| Short Physical Performance Battery | — | — | — | — | — | 0.44 | 0.13 | 0.45 | |
| Tai chi | 11.02 (1.36) | 11.35 (1.05) | 11.20 (1.42) | 0.27 | 0.13 | — | — | — | |
| CBT-I | 11.09 (1.37) | 11.54 (0.77) | 11.48 (0.92) | 0.40 | 0.33 | — | — | — | |
| Hospital Anxiety and Depression Scale | |||||||||
| Anxiety subscale | — | — | — | — | — | 0.21 | <0.001* | 0.48 | |
| Tai chi | 7.08 (3.55) | 4.09 (3.00) | 2.92 (3.09) | −0.91 | −1.25 | — | — | — | |
| CBT-I | 7.89 (4.31) | 3.39 (2.86) | 3.04 (3.56) | −1.23 | −1.23 | — | — | — | |
| Depression subscale | — | — | — | — | — | 0.13 | <0.001* | 0.91 | |
| Tai chi | 6.91 (3.66) | 4.24 (3.20) | 2.96 (3.06) | −0.78 | −1.17 | — | — | — | |
| CBT-I | 7.45 (4.19) | 3.57 (2.78) | 3.30 (3.62) | −1.09 | −1.06 | — | — | — | |
| Total energy intake (kcal/day) | — | — | — | — | — | 0.72 | 0.88 | 0.19 | |
| Tai chi | 2072.03 (856.70) | 1933.06 (628.23) | 1894.03 (717.04) | −0.18 | −0.22 | — | — | — | |
| CBT-I | 1869.30 (555.38) | 2002.39 (587.03) | 1903.59 (745.69) | 0.23 | 0.05 | — | — | — | |
Data are presented as estimated mean (standard deviation), baseline covaried. CBT-I=cognitive behavioural therapy for insomnia; IPAQ-SF=International Physical Activity Questionnaire—Short Form; SF-12v2=Short Form 12 item, version 2.
Statistical significance.
Effect size: change from baseline to 3 months and 15 months.
The sample size for all secondary outcomes was n=100 participants in tai chi group and n=100 participants in CBT-I group; 30 participants in tai chi group and 26 participants in CBT-I group were taking hypnotic drugs (lowest recommended dosage, units/week).
Metabolic equivalent minutes per week.
Sensitivity analysis, subgroup analysis, and adverse events
The results of the sensitivity analyses on the primary outcome were consistent with the primary analysis (fig 2 and supplementary materials III, IV, V). The generalised estimating equations analyses revealed the changes in ISI were not associated with age and age groups, body mass index, or sex. Supplementary materials VI and VII present the differences in mean changes in ISI and 95% confidence intervals for each subgroup. No adverse events were observed in either group during the intervention period.
Discussion
Principal findings
This randomised, non-inferiority trial compared the acute and long term efficacy of tai chi and CBT-I for treating chronic insomnia in middle aged and older adults. Our data showed that tai chi was inferior to CBT-I after the three month intervention, but became non-inferior at the 12 month follow-up (month 15).
Short term efficacy of tai chi
Most studies exploring tai chi as a treatment for insomnia used passive control designs or other exercise modalities as comparators, while direct comparison with first line clinical treatments is scarce.16 19 In this study, we provide scientific evidence about the comparative efficacy of tai chi and CBT-I using a non-inferiority design with a stringent comparison margin. The results showed that tai chi was inferior to CBT-I after the three month intervention, but the lower efficacy of tai chi was mitigated during the 12 month post-intervention follow-up. Another study comparing tai chi and CBT-I found that tai chi was non-inferior to CBT-I after the three month intervention and at the 12 month post-intervention follow-up in patients with insomnia who had undergone successful treatment for breast cancer.23 The discrepancy between the results could be attributed to differences in clinical characteristics and insomnia profiles between patients with a history of breast cancer and the participants in our study. For instance, patients with insomnia and a cancer history differ from those without cancer in terms of risk factors, such as cancer stage and history of lumpectomy and chemotherapy, and clinical characteristics, such as the number of comorbidities and increased levels of depression, anxiety, and fatigue.29
Our findings suggest that tai chi can lead to substantial improvements in insomnia severity after a three month intervention, while the long term efficacy of tai chi is non-inferior to that of CBT-I, the gold standard treatment for chronic insomnia. The observed effect size of the tai chi intervention on ISI was large (Cohen’s d 1.50) and comparable to the effect sizes reported in previous studies investigating the efficacy of tai chi and conventional treatments (eg, hypnotic drugs and CBT-I, with Cohen’s d between 0.36 and 1.13).16 19 30 Overall, our findings support the use of tai chi as a potential treatment approach for the management of chronic insomnia.
Long term efficacy of tai chi
Current theories suggest that physiological and cognitive hyperarousal contribute to the evolution and chronicity of insomnia.31 A longitudinal study reported that 38% of patients with insomnia presented with persistent symptoms after five years of follow-up.32 Additionally, chronic insomnia is associated with high relapse rates because studies have shown that 27% of treated patients have a relapse after remission.33 This finding highlights the need for treatments with long term efficacy and acceptability. Our study showed that the full benefits of tai chi do not emerge 12 months after the intervention. These findings align with previous research showing that the beneficial effects of insomnia remission on systemic and cellular inflammation are delayed and non-significant until 12 months after the intervention.34 The effect size on the ISI in the tai chi group at 12 month follow-up increased by 55% from levels measured immediately after the intervention. Importantly, the long term improvements after tai chi were non-inferior to those of CBT-I. This result is particularly important considering the high costs of CBT-I and its limited availability in healthcare systems worldwide.11
Our data revealed that 31 (36.5%) of 85 participants in the tai chi group who completed the outcome assessments at month 15 maintained their engagement in tai chi practice. The average practicing duration was 60±107 minutes, although the overall training volume decreased compared with the intervention period. However, 13 (15.9%) of 82 participants in the CBT-I group who completed the outcome assessments at month 15 reported self-practicing of the knowledge and skills learnt during the three month intervention. Therefore, the positive effects of tai chi being non-inferior to CBT-I at follow-up may be partially attributed to participants’ continued practice after the end of the intervention. Taken together, the large acute and long term efficacy, and the observed adherence to the intervention even after the end of the study, support the sustained benefits of tai chi, suggesting that tai chi could be used as an alternative treatment approach for the long term management of chronic insomnia.
Underlying physiological and psychological mechanisms of tai chi and CBT-I
A previous randomised controlled trial by Irwin and colleagues compared the physiological responses of tai chi and CBT-I on systemic, cellular, and genomic markers of inflammation in 90 patients with chronic insomnia and a history of breast cancer over a three month intervention and a 12 month post-intervention follow-up.34 Both interventions statistically significantly reduced toll-like receptor 4 stimulated monocyte production of interleukin 6 and tumour necrosis factor α, and their coexpression. Additionally, they were associated with reductions in antiviral gene expression and inflammatory gene transcripts, and with increases in antiviral gene transcripts over 15 months (all P values <0.05). Interestingly, tai chi showed greater and more sustained reductions in systemic and cellular inflammation at the 12 month post-intervention follow-up, while CBT-I produced stronger increases in antiviral gene transcripts (all P values <0.05). These findings suggest distinct physiological pathways by which each intervention modulates inflammatory processes and offers therapeutic benefits, and aligns with the results of our study relating to the long term benefits of tai chi in treating insomnia.
In addition to the inflammatory profile, previous research has also examined the potential neurological mechanisms underlying the effects of these two interventions. Studies have shown that a 12 week tai chi intervention modulated resting state functional connectivity in cognitive control networks,35 the default mode network,36 and the hippocampus,37 and improved fractional amplitude of low frequency fluctuations,38 while the CBT-I intervention enhanced functional connectivity in the frontoparietal network and reduced hyperarousal by decreasing thalamic activity.39 40 Although studies directly comparing the psychological mechanisms underlying tai chi and CBT-I are needed, existing evidence highlights notable differences. Cognitive therapy, a core component of CBT-I, targets dysfunctional beliefs, attitudes towards sleep and insomnia, attentional biases, and cognitive arousal, addressing specific maladaptive patterns that sustain insomnia.41 In contrast, tai chi is a holistic mind-body practice integrating physical, psychological, emotional, spiritual, and behavioural components,14 which has been shown to reduce stress, depressive symptoms, and cognitive hyperarousal.42 43 44 Importantly, only a limited number of well controlled studies have explored these underlying mechanisms, particularly in older adults with chronic insomnia. Further mechanistic research is warranted to clarify how these distinct pathways contributed to the observed sustained therapeutic effects, such as those we observed with tai chi at the 12 month post-intervention follow-up.
Study limitations
This study has several potential limitations. A large proportion (77.5%, n=155) of the study participants were older adults aged ≥60 years, which might limit the generalisability of our results when considering younger populations. However, older adults have a higher prevalence of chronic insomnia and higher risk of developing comorbid conditions such as depression and cognitive impairment.5 Therefore, it is crucial to determine the efficacy of treatment approaches that target this vulnerable population. Additionally, this study was conducted at a single research centre, which might further limit the generalisability of the results. However, our study participants were recruited from various districts of Hong Kong, representing a diverse study population. Further studies are needed to replicate our findings and determine whether the benefits of tai chi can be applied to other countries or regions with different demographic characteristics.
Tai chi was chosen as the exercise modality in this study, which might restrict the implementation of similar interventions in “western” countries owing to cultural differences. Nevertheless, national trend surveys in the US showed that the practice of tai chi and other mind-body exercises increased from 6% to 15% from 2002 to 2012, highlighting the growing popularity of mind-body practices.45 Additionally, the tai chi intervention was standardised and manualised to enable replication. Finally, future studies are warranted to regularly assess (eg, every three months) adherence to tai chi during the post-intervention follow-up period to capture the full tai chi practice behaviour.
Conclusion
After the three month intervention, tai chi and CBT-I improved insomnia symptoms in middle aged and older adults with chronic insomnia, although the improvements were more marked in the CBT-I group and tai chi was inferior to CBT-I in treating insomnia. At the 12 month follow-up (month 15), tai chi was non-inferior to CBT-I in improving insomnia symptoms. Additionally, tai chi and CBT-I had comparable benefits on subjective sleep parameters, quality of life, mental health, and physical activity level. Our study supports tai chi as an alternative treatment approach for the long term management of chronic insomnia in middle aged and older adults.
What is already known on this topic
Tai chi has been found to be non-inferior to cognitive behavioural therapy for insomnia (CBT-I), which is the first line treatment for chronic insomnia, after a three month intervention period and a 12 month post-intervention follow-up in patients with insomnia and a history of breast cancer
The long term clinical potential of tai chi for managing chronic insomnia in middle aged and older adults needs further investigation because patients with insomnia and a history of cancer have different characteristics
What this study adds
This randomised non-inferiority trial compared the effects of tai chi and CBT-I on chronic insomnia in middle aged and older adults
The efficacy of the tai chi intervention was found to be non-inferior to CBT-I at the 12 month post-intervention follow-up, but not after the 3 month intervention
These results support the use of tai chi for the long term management of chronic insomnia in middle aged and older adults
Acknowledgments
The authors thank all the participants for their contribution to this study. They also thank the tai chi instructors (Ms Wong Hung Wah and Ms Lo Lai Wah) and cognitive behavioural therapy for insomnia (CBT-I) instructors (Dr Chan Ngan Yin Rachel and Dr Li Xin Shirley) for delivering the interventions. The authors thank Mr Tsang Fan Pong, Dr Cheung Tin Wai Forrest, and Ms Hao Fong Sit Millet for supporting the interventions. The Tai Chi and CBT-I instructors were compensated for their contributions.
Web extra.
Extra material supplied by authors
Web appendix: Supplementary material
Contributors: PMS, SXL, DYF, DKCC, SSCH, KFC, JW, CW, and MRI conceived the study, formulated the research question, and sought funding support for this study. DJY and APY coordinated the intervention arrangement and outcome assessments. SXL, RNYC, DJY, and APY supported the interventions. DJY, APY, and DYF supported the statistical analyses. PMS, DJY, APY, and FR drafted the initial version of the manuscript. All authors revised the manuscript for important intellectual content and approved the final version of the manuscript. PMS, DJY, and APY contributed equally to this work as co-first authors. PMS is the guarantor for the paper and accepts full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish. The corresponding author (PMS) attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. During the preparation of this work the authors used GPT-4o in order to improve the readability and language of this work. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Funding: This study was supported by General Research Fund of Research Grants Council, Hong Kong University Grants Committee (project No 17112819), and Seed Fund for Basic Research of the University of Hong Kong. CW is supported by the US National Institutes of Health under awards K24AT007323 and UG3/UH3AT012413. The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; or the decision to submit the manuscript for publication.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/disclosure-of-interest/ and declare: support from General Research Fund of Research Grants Council, Hong Kong University Grants Committee, Seed Fund for Basic Research of the University of Hong Kong, and the US National Institutes of Health for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Transparency: The lead author (PMS) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Dissemination to participants and related patient and public communities: The authors will circulate all study results to participants and patient organisations immediately after publication and will showcase the trial findings at pertinent national and international conferences.
Provenance and peer review: Not commissioned; externally peer reviewed
Ethics statements
Ethical approval
The study was approved by the University of Hong Kong/Hospital Authority Hong Kong West Institutional Review Board (IRB approval No UW 18-621). Interested individuals were invited to a preliminary visit to confirm eligibility. Verbal and written information on the study was provided and written informed consent was obtained before baseline outcome assessments.
Data availability statement
Statistical code and dataset are available from the corresponding author at pmsiu@hku.edu.hk. The code used to analyse the data in the paper can be found in the supplementary files. The data underlying the findings in this paper are openly and publicly available and can be found here: https://doi.org/10.6084/m9.figshare.29203967.v3. If you encounter problems accessing the data, please contact the corresponding author.
References
- 1. Morin CM, Drake CL, Harvey AG, et al. Insomnia disorder. Nat Rev Dis Primers 2015;1:15026. 10.1038/nrdp.2015.26 [DOI] [PubMed] [Google Scholar]
- 2. Wong WS, Fielding R. Prevalence of insomnia among Chinese adults in Hong Kong: a population-based study. J Sleep Res 2011;20(1pt1):117-26. 10.1111/j.1365-2869.2010.00822.x [DOI] [PubMed] [Google Scholar]
- 3. Bedford LE, Tang EHM, Dong W, et al. Who reports insufficient and disturbed sleep? Results from a representative population-based health survey in Hong Kong. BMJ Open 2022;12:e058169. 10.1136/bmjopen-2021-058169 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Pillai V, Roth T, Drake CL. The nature of stable insomnia phenotypes. Sleep 2015;38:127-38. 10.5665/sleep.4338 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Morin CM, Benca R. Chronic insomnia. Lancet 2012;379:1129-41. 10.1016/S0140-6736(11)60750-2 [DOI] [PubMed] [Google Scholar]
- 6. Léger D, Bayon V. Societal costs of insomnia. Sleep Med Rev 2010;14:379-89. 10.1016/j.smrv.2010.01.003 [DOI] [PubMed] [Google Scholar]
- 7. Edinger JD, Arnedt JT, Bertisch SM, et al. Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med 2021;17:255-62. 10.5664/jcsm.8986 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Manber R, Alcántara C, Bei B, Morin CM, van Straten AA. Integrating technology to increase the reach of CBT-I: state of the science and challenges ahead. Sleep 2023;46:zsac252. 10.1093/sleep/zsac252 [DOI] [PubMed] [Google Scholar]
- 9. Ulmer CS, Bosworth HB, Beckham JC, et al. Veterans affairs primary care provider perceptions of insomnia treatment. J Clin Sleep Med 2017;13:991-9. 10.5664/jcsm.6702 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Conroy DA, Ebben MR. Referral practices for cognitive behavioral therapy for insomnia: a survey study. Behav Neurol 2015;2015:819402. 10.1155/2015/819402 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Baglioni C, Altena E, Bjorvatn B, et al. The European Academy for Cognitive Behavioural Therapy for Insomnia: an initiative of the European Insomnia Network to promote implementation and dissemination of treatment. J Sleep Res 2020;29:e12967. 10.1111/jsr.12967 [DOI] [PubMed] [Google Scholar]
- 12. Yang G-Y, Sabag A, Hao W-L, et al. Tai Chi for health and well-being: a bibliometric analysis of published clinical studies between 2010 and 2020. Complement Ther Med 2021;60:102748. 10.1016/j.ctim.2021.102748 [DOI] [PubMed] [Google Scholar]
- 13.W. D. World Tai Chi & Qi Gong Day. Available from: https://www.worldtaichiday.org/.
- 14. Wayne PM, Ahn A, Clark J, et al. The science of Tai Chi and Qigong as whole person health-part I: rationale and state of the science. J Integr Complement Med 2025;31:499-520. 10.1089/jicm.2024.0957 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Wang CC, Li K, Gaylord A S. Trends and characteristics of Tai Chi and Qi Gong use among U.S. adults: results from the 2007-2017 National Health Interview Survey. Complement Ther Med 2022;71:102890. 10.1016/j.ctim.2022.102890 [DOI] [PubMed] [Google Scholar]
- 16. Siu PM, Yu AP, Tam BT, et al. Effects of Tai Chi or exercise on sleep in older adults with insomnia: a randomized clinical trial. JAMA Netw Open 2021;4:e2037199. 10.1001/jamanetworkopen.2020.37199 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Siu PM, Yu AP, Chin EC, et al. Effects of Tai Chi or conventional exercise on central obesity in middle-aged and older adults: a three-group randomized controlled trial. Ann Intern Med 2021;174:1050-7. 10.7326/M20-7014 [DOI] [PubMed] [Google Scholar]
- 18. Hasan F, Tu Y-K, Lin C-M, et al. Comparative efficacy of exercise regimens on sleep quality in older adults: a systematic review and network meta-analysis. Sleep Med Rev 2022;65:101673. 10.1016/j.smrv.2022.101673 [DOI] [PubMed] [Google Scholar]
- 19. Yu DJ, Recchia F, Bernal JD, et al. Effectiveness of exercise, cognitive behavioral therapy, and pharmacotherapy on improving sleep in adults with chronic insomnia: a systematic review and network meta-analysis of randomized controlled trials. Healthcare 2023;11:2207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Piaggio G, Elbourne DR, Pocock SJ, Evans SJ, Altman DG, CONSORT Group . Reporting of noninferiority and equivalence randomized trials: extension of the CONSORT 2010 statement. JAMA 2012;308:2594-604. 10.1001/jama.2012.87802 [DOI] [PubMed] [Google Scholar]
- 21. Yu DJ, Yu AP, Li SX, et al. Effects of Tai Chi and cognitive behavioral therapy for insomnia on improving sleep in older adults: Study protocol for a non-inferiority trial. J Exerc Sci Fit 2023;21:67-73. 10.1016/j.jesf.2022.10.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Yang G-Y, Wang L-Q, Ren J, et al. Evidence base of clinical studies on Tai Chi: a bibliometric analysis. PLoS One 2015;10:e0120655. 10.1371/journal.pone.0120655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Irwin MR, Olmstead R, Carrillo C, et al. Tai Chi Chih compared with cognitive behavioral therapy for the treatment of insomnia in survivors of breast cancer: a randomized, partially blinded, noninferiority trial. J Clin Oncol 2017;35:2656-65. 10.1200/JCO.2016.71.0285 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Bastien CH, Vallières A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med 2001;2:297-307. 10.1016/S1389-9457(00)00065-4 [DOI] [PubMed] [Google Scholar]
- 25. Morin CM, Belleville G, Bélanger L, Ivers H. The Insomnia Severity Index: psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep 2011;34:601-8. 10.1093/sleep/34.5.601 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Yu DS. Insomnia Severity Index: psychometric properties with Chinese community-dwelling older people. J Adv Nurs 2010;66:2350-9. 10.1111/j.1365-2648.2010.05394.x [DOI] [PubMed] [Google Scholar]
- 27. Garland SN, Carlson LE, Stephens AJ, Antle MC, Samuels C, Campbell TS. Mindfulness-based stress reduction compared with cognitive behavioral therapy for the treatment of insomnia comorbid with cancer: a randomized, partially blinded, noninferiority trial. J Clin Oncol 2014;32:449-57. 10.1200/JCO.2012.47.7265 [DOI] [PubMed] [Google Scholar]
- 28. D’Agostino RB, Sr, Massaro JM, Sullivan LM. Non-inferiority trials: design concepts and issues—the encounters of academic consultants in statistics. Stat Med 2003;22:169-86. 10.1002/sim.1425 [DOI] [PubMed] [Google Scholar]
- 29. Fiorentino L, Ancoli-Israel S. Insomnia and its treatment in women with breast cancer. Sleep Med Rev 2006;10:419-29. 10.1016/j.smrv.2006.03.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. De Crescenzo F, D’Alò GL, Ostinelli EG, et al. Comparative effects of pharmacological interventions for the acute and long-term management of insomnia disorder in adults: a systematic review and network meta-analysis. Lancet 2022;400:170-84. 10.1016/S0140-6736(22)00878-9 [DOI] [PubMed] [Google Scholar]
- 31. Riemann D, Nissen C, Palagini L, Otte A, Perlis ML, Spiegelhalder K. The neurobiology, investigation, and treatment of chronic insomnia. Lancet Neurol 2015;14:547-58. 10.1016/S1474-4422(15)00021-6 [DOI] [PubMed] [Google Scholar]
- 32. Morin CM, Jarrin DC, Ivers H, Mérette C, LeBlanc M, Savard J. Incidence, persistence, and remission rates of insomnia over 5 years. JAMA Netw Open 2020;3:e2018782. 10.1001/jamanetworkopen.2020.18782 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Morin CM, Bélanger L, LeBlanc M, et al. The natural history of insomnia: a population-based 3-year longitudinal study. Arch Intern Med 2009;169:447-53. 10.1001/archinternmed.2008.610 [DOI] [PubMed] [Google Scholar]
- 34. Irwin MR, Hoang D, Olmstead R, et al. Tai Chi compared with cognitive behavioral therapy and the reversal of systemic, cellular and genomic markers of inflammation in breast cancer survivors with insomnia: a randomized clinical trial. Brain Behav Immun 2024;120:159-66. 10.1016/j.bbi.2024.05.022 [DOI] [PubMed] [Google Scholar]
- 35. Tao J, Chen X, Egorova N, et al. Tai Chi Chuan and Baduanjin practice modulates functional connectivity of the cognitive control network in older adults. Sci Rep 2017;7:41581. 10.1038/srep41581 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Liu J, Tao J, Liu W, et al. Different modulation effects of Tai Chi Chuan and Baduanjin on resting-state functional connectivity of the default mode network in older adults. Soc Cogn Affect Neurosci 2019;14:217-24. 10.1093/scan/nsz001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Tao J, Liu J, Egorova N, et al. Increased hippocampus–medial prefrontal cortex resting-state functional connectivity and memory function after Tai Chi Chuan practice in elder adults. Front Aging Neurosci 2016;8:25. 10.3389/fnagi.2016.00025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Tao J, Chen X, Liu J, et al. Tai Chi Chuan and Baduanjin mind-body training changes resting-state low-frequency fluctuations in the frontal lobe of older adults: a resting-state fMRI study. Front Hum Neurosci 2017;11:514. 10.3389/fnhum.2017.00514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Sabot D, Baumann O. Neuroimaging correlates of cognitive behavioral therapy for insomnia (CBT-I): a systematic literature review. J Cogn Psychother 2023;37:82-101. 10.1891/JCPSY-D-21-00006 [DOI] [PubMed] [Google Scholar]
- 40. Lee YG, Kim S, Kim N, et al. Changes in subcortical resting-state functional connectivity in patients with psychophysiological insomnia after cognitive-behavioral therapy: changes in resting-state FC after CBT for insomnia patients. Neuroimage Clin 2017;17:115-23. 10.1016/j.nicl.2017.10.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Altena E, Ellis J, Camart N, Guichard K, Bastien C. Mechanisms of cognitive behavioural therapy for insomnia. J Sleep Res 2023;32:e13860. 10.1111/jsr.13860 [DOI] [PubMed] [Google Scholar]
- 42. Zou L, Sasaki JE, Wei G-X, et al. Effects of mind–body exercises (Tai Chi/Yoga) on heart rate variability parameters and perceived stress: a systematic review with meta-analysis of randomized controlled trials. J Clin Med 2018;7:404. 10.3390/jcm7110404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Yin J, Yue C, Song Z, Sun X, Wen X. The comparative effects of Tai chi versus non-mindful exercise on measures of anxiety, depression and general mental health: a systematic review and meta-analysis. J Affect Disord 2023;337:202-14. 10.1016/j.jad.2023.05.037 [DOI] [PubMed] [Google Scholar]
- 44. Liu Z, Zhang L, Bai L, et al. Repetitive transcranial magnetic stimulation and tai chi chuan for older adults with sleep disorders and mild cognitive impairment: a randomized clinical trial. JAMA Netw Open 2025;8:e2454307. 10.1001/jamanetworkopen.2024.54307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Wang CC, Li K, Choudhury A, Gaylord S. Trends in yoga, Tai Chi, and Qigong use among US adults, 2002-2017. Am J Public Health 2019;109:755-61. 10.2105/AJPH.2019.304998 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Web appendix: Supplementary material
Data Availability Statement
Statistical code and dataset are available from the corresponding author at pmsiu@hku.edu.hk. The code used to analyse the data in the paper can be found in the supplementary files. The data underlying the findings in this paper are openly and publicly available and can be found here: https://doi.org/10.6084/m9.figshare.29203967.v3. If you encounter problems accessing the data, please contact the corresponding author.
