The prevalence of older people with cognitive decline is increasing since the aging population is growing substantially worldwide. Cognitive impairment places older adults at high risk for functional disability. Previous researches have provided strong evidence on the beneficial effects of physical exercise on maintaining cognitive functions in older adults. Tai Chi is considered as a low to moderate intensity exercise, which is performed in a slow and rhythmic movement. It is a popular and safe exercise suitable for older people who have weaker muscle strength. The aim of this study was to determine the effects of Tai Chi on cognition and instrumental activities of daily living (I-ADLs), and health-related quality of life (HRQOL) in older Chinese adults with mild cognitive impairment (MCI).

The research hypothesis of the study stated: There is significant improvement in the general cognitive performance, functional capabilities of instrumental ADLs,the physical component and mental component of HRQOL between the group of community-dwelling older people with MCI who have participated and the control group who have not participated in the Tai Chi program.

1. Study design & sample recruitment:

   This was a quasi-experimental study which adopted the multi-site non-randomized control group pre- and post-test design. By adopting the multi-site design, better representation of the population could be obtained. This intervention study involved two groups of participant, the experimental and control groups, whereas no randomization would be taken place.

   The study was carried out in District Elderly Community Centers (DECCs) in Hong Kong. The DECC has been implemented since 2001 under the support from the Social Warfare Department (SWD) to provide community support services at district level to enable older people to remain in the community and to lead a healthy, respectful and dignified life.The service users are community-dwelling older citizens aged 60 or above living in the locality. Invitation letters were sent to various DECCs for joining the study. Finally, there were four elderly centers willing to participate in the study. The participants were recruited from the four DECCs using convenience sampling.

   To avoid the possible diffusion effect that might interfere the research result, all eligible participants in a community center were allocated to either experimental group or control group, who were treated with single intervention, so there would be no discussion among participants who have received different treatment regimes.

   The present study had employed a control group to enhance methodological rigor and included a pre-test to assess the equivalence between two groups, so the problem on selection biases could be alleviated.

2. Sample size estimation A medium effect size of 0.5 was set for the study, based on previous research on the effect of Tai Chi on cognitive function. Based on the use of a sample size calculator (https://www2.ccrb.cuhk.edu.hk/stat/Other.htm), a sample of 64 per group was expected to provide 80% power to detect a medium effect size of 0.5, at a significance level of 0.05. Allowing for an attrition rate of 20% in the Tai Chi groups, 80 participants were recruited for each group.

3. Intervention The intervention group received a 16-week Tai Chi program, of 32 sessions (two sessions per week) each one hour long.The capacity was around 20 people per class. A certified Tai Chi master taught the Yang-style simple form of Tai Chi to the participants allocated to the experimental group.There were four sets of movements included in the simplified form; this required less time for the older practitioners, who usually had less exercise endurance, to complete the whole set of movements. Moreover, it was an appropriate form of Tai Chi for older people with MCI because it involved only a few postures, hence it was easier for them to learn and remember. The Tai Chi classes were conducted in the open hall of the community center or a public playground with sufficient space, a flat floor and good lighting.To promote safety, measures include appropriate footwear, appropriate warm-up and cool-down, correct exercise technique and gradual increases in exercise dose will be emphasized in Tai Chi class. Incident report will be done to record and analyses causes for any adverse effect on Tai Chi training.To ensure adherence, the participants were required to sign attendance records. Telephone follow-up was arranged to provide emotional support and reinforce compliance if they were absent from two consecutive lessons. As well, the participants were advised to do daily practice at home to foster better mastery of the Tai Chi movements and build up a regular exercise habit. A guide book was issued to each participant, providing guidance to reinforce correct postures.

   For participants in the control group, they were considered as usual care group which could attend for different kinds of recreational activities provided by the community center. However, they were advised not to practice Tai Chi and not to join any structured physical training program during the study period to avoid possible aerobic capacity changes with exercise practices causing interferon on experimental outcomes.

4. Data collection The outcome variables of the study comprised of general cognitive status, instrumental ADLs, and HRQOL of older participants. The outcome assessments were conducted at baseline, and at 4-month (at the end of study) for all participants.The Chinese version of the Mini-Mental State Examination (CMMSE) was used for global cognitive assessment. The Hong Kong Chinese version of Lawton's Instrumental Activities of Daily Living (IADL-CV) was used to assess the participants' IADLs levels and the Chinese version of the Short Form-12 Health Survey-Standard 1 (SF-12) was employed to assess the participants' perceived health status.

5. Data analysis The IBM SPSS Version 24.0 (IBM Crop. Armonk, N.Y.) was used for data entry and analysis. All statistical tests involved were 2-tailed and the level of significance was set at 5%. Descriptive statistics, including mean, standard deviation, frequency, and percentage, were used to summarize and present the sample characteristics and outcome measures.

   Chi-square, Fisher's exact, and independent t- tests were used, as appropriate, to compare the characteristics of intervention and control group at the baseline. If there were significant differences between two groups of these characteristics, they were adjusted in the subsequent outcome analysis.

   Generalized Estimating Equations (GEE) was applied for comparing each of the repeated measures study outcomes (i.e. CMMSE score, IADL-CV score, physical and mental components of SF-12 as well as subscales of SF-12) between the two groups with adjustment for those demographic characteristics showing statistical incomparability between groups. Two dummy variables, Group and Time, were included in each GEE model to represent the baseline group difference (Intervention - Control) and the time effect on the control group (Post 16 weeks - Baseline), respectively. Furthermore, the interaction term, Group by Time (Group*Time) was also included to assess the differential change in each outcome between the two groups. GEE model could account for intra-correlated repeated measures outcome and produce unbiased effect estimated under the data missingness mechanism of missing completely at random (MCAR).

6. Ethical consideration Approval for conducting the study was obtained from the Joint CUHK-NTEC Clinical Research Ethics Committee of the Chinese University of Hong Kong and the selected community elderly centers.The participants were required to sign a consent form. Information and the purpose of the study were explained. They could withdraw from the study at any time of their own will.

Inclusion Criteria:

• Chinese people aged 60 years or above;
• Obtained the CMMSE screening score ranging from 19 to 28, which was corrected based on educational level (≥ 18 for illiterate respondents and ≥ 22 for those having received more than two years of education);
• Have ability to perform self-care functions on their own;
• No confirmed diagnosis of dementia, depression or other psychiatric illnesses;
• Not engaged in any structured physical exercise program or Tai Chi practice in the preceding year.

Exclusion Criteria:

• Had a medical history of chronic alcoholism or brain trauma occurred in previous years; - Regular users of medications that could affect cognition;
• Contraindicated light to moderate physical exercise.

All IPD that underlie results in a publication

The supporting information (as listed above) will become available starting from 1 year after publication for 6 months

The datasets used and/or analyzed during the current study are available from the corresponding author ( Name: Siu Mei Yi at minniesiumy@gmail.com) on reasonable request. All data analyzed during this study are included in this published article and its supplementary information files.

All personals involved the study (participants, community center supporting staff, investigators and outcome assessors) had not masked for treatment assignment.