The study was approved by the South Yorkshire Ethics Committee and funded by a grant from the British Heart Foundation. All patients gave written informed consent
Patients were included if they had stable symptomatic chronic heart failure for at least 3 months, with evidence on echocardiography of left ventricular systolic dysfunction. Patients were excluded if they had symptomatic postural hypotension, sustained ventricular arrhythmias, hypertension (defined by persistent systolic blood pressure of 180 mm Hg or diastolic blood pressure >110 mm Hg), severe obstructive valvular disease or other respiratory of locomotor condition that would otherwise limit their exercise capacity.
Sixty‐nine of 274 outpatients who were initially approached agreed to participate. Of these 69 patients, four dropped out before randomisation, so 65 were randomised in the study. All were in New York Heart Association (NYHA) symptom class II–III. Patients were recruited in three cohorts of 20, 24, and 21, respectively, of whom 13 were lost to follow up (fig 1). Fifty‐two patients (25 experimental, 27 control) completed the trial. Patients were matched in pairs for baseline measures, then each pair randomly allocated to treatment or control. Given the need to train individuals in Tai Chi, it was necessary to ensure that the experimental (exp) group size did not exceed 10; the study was therefore performed in three consecutive groups each lasting 16 weeks. Group 1 included 15 patients (exp 7, control 8), group 2, 19 patients (exp 10, control 9), and group 3, 18 patients (8 exp, 10 control), respectively. Overall there were 42 men (mean age 68.9 years, range 46–90 years) and 10 women (mean age 70.0 years, range 58–82 years).
Figure 1Flow diagram of study recruitment.
Exercise capacity was assessed using the incremental shuttle walk test (ISWT). Two tests were performed at the start and at the end of the 16 week period. The average of the two tests was used in the analysis. The ISWT is a symptom limited exercise test with a progressive increase in workload designed to allow subjects to achieve maximum effort tolerance. Subjects walk back and forth along a horizontal 10 m course, marked out by two cones, and must complete the shuttle before a pre‐recorded signal from a cassette player, which shortens incrementally after each shuttle. The end point (distance walked in metres) is reached when the subject fails to complete the shuttle before the signal. The ISWT has been evaluated in patients with chronic heart failure, as an alternative to cardiopulmonary exercise testing and the traditional 6 min walk test. It is highly reproducible, preferred by patients, and correlates strongly with peak oxygen consumption (o2
0.84, p<0.0001). In multivariate analysis the ISWT was found to be the most significant independent predictor of peak o217,18
and, after 17 months follow up, predicted event‐free survival, whereas the 6 min walk test did not.19
Subjects completed the Minnesota Living with Heart Failure questionnaire (MLHF) to assess disease specific symptoms and quality of life, and the SCL‐R questionnaire to give a standardised assessment of mood at the beginning and end of the treatment period. Systolic and diastolic blood pressure was measured at the beginning and end of the treatment period.
The experimental group received two 55 min sessions of Tai Chi exercise per week for 16 weeks. The sessions were carried out every Monday and Friday of each week and the patients were encouraged to practice what they had learnt at home and to record the length of personal practice time. Each session began with 20 min of Chi Kung exercises (including an explanation of the exercise principles involved), followed by a 5 min rest period. The session recommenced with Tai Chi Chuan practice for 20 min followed by 5 min of cooling down stretching exercises. The Tai Chi exercises were designed around the Tai Chi and Chi Kung exercises to achieve the following:
- work the lower limbs, positively affecting muscular tone
- work rate was set at a moderate level as assessed by reference to the Borg Scale for patient self assessment
- relaxation/meditation methods were used to promote relaxation effects to encourage positive autonomic responses
- upper body (arm, shoulder and chest) exercises to improve tone in the muscles used in respiration.
Chi Kung is a series of exercises that complement Tai Chi practice and are commonly used while learning Tai Chi as an introductory warm up. These exercises purport to produce relaxation, stillness of mind and of mood. The Tai Chi Chuan employed in this study was the Wu Chian Chuan style as practised by the Ma family in Shanghai China. The Chi Kung exercises were derived from the Orchid Hand 21 Style and Wu's Chi Kung exercises.
Exercises were conducted by a trained Tai Chi trainer with a trained cardiac rehabilitation nurse also in attendance. The exercise programme was graded and gradually increased, reaching full potential on the eighth week. This was because Tai Chi is initially difficult to learn so it was necessary to gradually increase the forms and to allow the patients to become accustomed to the exercises, in order to minimise adverse reactions.
The control group received standard medical supervision and drug treatment. This comprised regular contact with specialist heart failure nurses and outpatient visits with medical staff every 3 months. Patients were not told to restrict activities but were not told to do extra exercise. In addition, they were told that they could contact a member of the research staff if any problems or queries arose during the trial.
A pilot study of four patients was carried out to test the acceptability of the intervention in patients with heart failure and to provide information on statistical power. A mean improvement of 18% in shuttle walk distance was observed from baseline, and the data showed that 25 patients in each group would have a power of 75% to identify a similar change in trial subjects.
In order to avoid large differences in baseline measures between the groups by chance, a Monte Carlo procedure was used to maximise similarities between pairs of subjects, selecting the best fit after the baseline measures were completed. The patients were matched in pairs for age, gender, systolic blood pressure (SBP) and diastolic blood pressure (DBP), NYHA rating, brain natriuretic peptide (BNP) concentration, shuttle walk test, SCL‐90‐R depression and anxiety scores, and Minnesota LWHF symptom scores; each member was then randomly allocated to either treatment or control. In the third cohort, where there was an odd number of recruits, the individual closest to the overall group mean was identified and separately randomised to treatment or control (table 1).
Table 1Baseline matched data for each group (includes those lost to follow up)
No change was seen in the systolic or diastolic blood pressures.
There was one objective and two subjective end points. The primary end point for the study was the change in the average (of two) shuttle walk distances after Tai Chi compared with the change in the control group. The secondary end points were changes in symptom scores of heart failure and mood. We had intended also to measure changes in BNP between the groups, but because of technical problems we were unable to complete the post‐intervention assays.
The primary outcome measure was the change in shuttle walk test distance with Tai Chi compared with the change after no intervention (that is, a comparison of the deltas). This change was compared using paired Student's t tests.