Tai Chi in Patients With Heart Failure With Preserved Ejection Fraction


  • Gloria Y. Yeh MD, MPH,

    1. From the Osher Center for Integrative Medicine, Harvard Medical School, Boston, MA
    2. Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Deaconess Medical Center, Brookline, MA
    Search for more papers by this author
  • Malissa J. Wood MD,

    1. Division of Cardiology, Massachusetts General Hospital, Boston, MA
    Search for more papers by this author
  • Peter M. Wayne PhD,

    1. From the Osher Center for Integrative Medicine, Harvard Medical School, Boston, MA
    2. Division of Preventive Medicine, Brigham and Women’s Hospital, Boston, MA
    Search for more papers by this author
  • Mary T. Quilty BS,

    1. From the Osher Center for Integrative Medicine, Harvard Medical School, Boston, MA
    2. Division of Preventive Medicine, Brigham and Women’s Hospital, Boston, MA
    Search for more papers by this author
  • Lynne W. Stevenson MD,

    1. Division of Cardiovascular, Brigham and Women’s Hospital, Boston, MA
    Search for more papers by this author
  • Roger B. Davis ScD,

    1. From the Osher Center for Integrative Medicine, Harvard Medical School, Boston, MA
    2. Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Deaconess Medical Center, Brookline, MA
    Search for more papers by this author
  • Russell S. Phillips MD,

    1. From the Osher Center for Integrative Medicine, Harvard Medical School, Boston, MA
    2. Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Deaconess Medical Center, Brookline, MA
    Search for more papers by this author
  • Daniel E. Forman MD

    1. Division of Cardiovascular, Brigham and Women’s Hospital, Boston, MA
    2. New England Geriatric Research, Education, and Clinical Center, Veterans Administration Boston Healthcare System, Boston, MA
    Search for more papers by this author

Daniel E. Forman, MD, Brigham and Women’s Hospital, Cardiovascular Division, 75 Francis Street, Boston, MA 02115


Although exercise is an important component of heart failure management, optimal regimens, particularly in heart failure with preserved ejection fraction (HFPEF), are uncertain. Tai chi (TC) is a mind-body exercise that may have potential benefits but has not been studied in this population. The authors randomized 16 patients with HFPEF to either 12 weeks TC or aerobic exercise. Assessments included peak oxygen uptake, 6-minute walk, quality of life, echocardiography, mood, and self-efficacy at baseline and at 12 weeks. Cardiorespiratory measures during exercise were obtained to characterize training intensities. Baseline characteristics were as follows: age 66±12 years, E/A ratio 1.3±0.7, and E/e′ ratio 15.9±4.8. Overall, adherence was excellent (89% attendance). Change in peak oxygen uptake was similar between groups after 12 weeks, but 6-minute walk distance increased more after TC (69±46 m vs 10±31 m, P=.02). While both groups had improved Minnesota Living With Heart Failure scores and self-efficacy, Profile of Mood States (POMS)-Depression scores improved more with TC (−1.7±2.8 vs 1.6±3, P=.05). Cardiorespiratory assessment during TC showed lower oxygen uptake (4.3 mL/kg/min vs 9.4 mL/kg/min, P<.01), respiratory rate, and heart rate. TC is feasible and safe in HFPEF. Therepeutic endpoints appear similar with TC relative to aerobic exercise despite a lower aerobic training workload.

Heart failure (HF) with preserved ejection fraction (HFPEF) is widely prevalent and growing in incidence, particularly since aging increases susceptibility to the disease and the elderly population is expanding.1 Definitive therapy for HFPEF, however, remains unclear and there is a critical need to expand and optimize therapeutic insights and options.

Exercise training is an important component of HF management; however, most studies to date have investigated HF with systolic dysfunction. More recently, exercise training in HFPEF has gained attention. One study by Kitzman and colleagues demonstrated that a 16-week aerobic training regimen provided significant benefits in HFPEF.2 Peak exercise oxygen consumption (VO2) and 6-minute walk (6MW) distance improved significantly as compared with an attention control group. Physical quality-of-life score also improved (P=.03). These data provide important rationale for exercise. Unfortunately, multiple trials have demonstrated problematic adherence to aerobic exercise over time, particularly with diverse populations and settings.3,4 Moreover, studies have suggested that peripheral effects play a particularly important role.5 Thus, one question that has been raised is whether exercise gains might be adequately and efficiently achieved with alternative training options that may directly address these behavioral and physiological issues.

Tai chi (tai chi chuan or taijiquan) is a gentle, meditative exercise that has its roots in ancient Chinese martial arts. It employs detailed regimens of flowing circular arm movements, balance and weight shifting, breathing techniques, and cognitive tools (such as visualization and focused internal awareness).6,7 Also inherent in tai chi is a component of core strengthening and stretching. This may facilitate a beneficial training effect, particularly in patients who are deconditioned, older, or have chronic disease. In addition, compared with aerobic training, the relatively lower intensity and meditative qualities of tai chi may help accessibility and promote longer-term exercise adherence over time. Prior studies have reported good long-term adherence with tai chi in various chronic disease populations.8–11

In past studies, we investigated the potential of tai chi to improve exercise capacity and quality of life in patients with systolic HF.12,13 In this pilot investigation, we expand our inquiry to explore the potential benefits of tai chi for HFPEF. Our aims were to preliminarily assess changes in functional and physiological outcome measures after 12 weeks of tai chi as compared with low-intensity aerobic exercise, explore relative differences in training intensities, and determine feasibility of a larger, more definitive trial.


Study Design

Sixteen patients were recruited from the ambulatory cardiology practice at Beth Israel Deaconess Medical Center in Boston, MA. Inclusion criteria were based on the following: (1) physician diagnosis of HFPEF; (2) New York Heart Association functional class I, II, or III; (3) left ventricular ejection fraction ≥50% (by echocardiography, radionuclide angiography, or contrast angiography) within 2 years of screening. Exclusion criteria included: (1) unstable angina, myocardial infarction or major cardiac surgery in the past 3 months; (2) cardiac arrest in the past 6 months; (3) significant valvular or pericardial disease accounting for signs and symptoms of HF; (4) severe chronic obstructive pulmonary disease on bronchodilators or chronic lung disease with cor pulmonale; (5) moderate or severe pulmonary hypertension; (6) atrial fibrillation as the dominant rhythm; (7) severe peripheral vascular disease, claudication, or other physical condition that would preclude a walk test; (8) inability to perform bicycle ergometry; (9) technically inadequate echocardiographic windows or valvular conditions precluding assessment of diastolic function, ie, patients with severe mitral annular calcification; (10) cognitive dysfunction (Mini-Mental State Examination score ≤24); (11) non–English-speaking; or (12) current participation in conventional cardiac rehabilitation or regular practice of tai chi.

Patients were assigned randomly to receive 12 weeks of either tai chi or aerobic exercise. All participants continued to receive their usual care, which included pharmacologic therapy and general exercise advice as per American College of Cardiology/American Heart Association (ACC/AHA) HF guidelines. Permuted block randomization with variable block size was used to generate treatment assignments. All patients provided written informed consent. The Beth Israel Deaconess Medical Center’s human subjects review board approved the protocol.

Tai Chi Intervention

The tai chi intervention consisted of 1-hour group classes held twice weekly for 12 weeks with an experienced instructor. A standard training regimen was utilized that we have used in prior trials for patients with systolic HF (Table I). The protocol included traditional warm-up exercises followed by 5 simplified tai chi movements. Classes incorporated weight shifting, arm swinging, gentle stretches (of the neck, shoulders, spine, arms, legs), visualization techniques, and traditional breathing. These exercises focus on releasing tension in the physical body, incorporating mindfulness and imagery into movement, increasing awareness of breathing, and promoting overall relaxation of body and mind. The core tai chi movements were adapted from Master Cheng Man-Ch’ing’s Yang-style short form6 and performed repetitively. Chairs were provided for resting, and patients were allowed to progress at their own comfort and pace.

Table I. Description of the Tai Chi Intervention
WeekActivitiesApproximate Duration, min
1 Introductory session: overview of program  
1. Tai chi principles, philosophies15
2. Demonstration of tai chi form10
3. Expectations of participants10
4. Description of class format5
5. Participation in warm-up exercises30
2–5 Warm-up exercises (Repeated during all sessions)  
1. Standing: 
 (a) “Drumming the body”6
 (b) “Swinging to connect kidney and lungs”3
 (c) “Washing the body with qi”3
 (d) Standing meditation and breathing3
2. Sitting: 
 (a) Neck/shoulder stretches6
 (b) Arm/leg stretches3
 (c) Sitting meditation and breathing6
Total warm-up time 30
Tai chi movements  
1. “Raising the power”5–10
2. “Withdraw and push”5 per side
6–9 (Warm-up and Movements 1–2)  
3. “Grasp sparrows tail”5 per side
4. “Brush knee Twist step”5 per side
10–12 (Warm-up and Movements 1–4)  
5. “Wave hands like clouds”5–10

Each patient in the tai chi group was also provided with a 35-minute instructional videotape that reviewed the exercises presented in class and was encouraged to practice at home at least 3 times per week.

Aerobic Exercise Control

The conventional aerobic exercise group also met for 1 hour twice a week under the guidance of an experienced instructor. The aerobic training was modeled on a typical community-based conventional low-impact aerobic exercise class offered at a local community center and designed for a senior population with low to moderate training intensity. Warm-up, peak exercise, and cool-down stages were routine. Exercises included stretching of the torso/waist, neck, upper and lower extremity; step in place and side to side with rhythmic leg and arm movements; and seated strengthening exercises. One- or two-pound hand weights and resistance bands were optional. Chairs were available for seated cool-down exercises and for support when performing basic stretches and strengthening exercises. As this class replicated what was readily available in the community, patients were allowed to progress at their own comfort and pace.

Each patient in the aerobic exercise group was also provided with a 35-minute instructional videotape that reviewed the exercises and instructions presented in class, and patients were encouraged to practice at home at least 3 times a week.

Outcome Measures

Comprehensive assessments were completed at baseline and 12 weeks.

Exercise Capacity and Functional Status.  Symptom-limited cardiopulmonary exercise testing was completed using an electronically calibrated upright bicycle and continuous electrocardiographic monitor. A standard 10-watt graded bicycle ramp protocol was utilized.14 Patients were encouraged to exercise to exhaustion. Gas exchange was assessed using a SensorMedics (Yorba Linda, CA) metabolic cart. Peak VO2 values were averaged from the final 20 seconds of the test. The exercise testing staff was blinded to patients’ exercise group assignments.

The 6MW test per the American Thoracic Society protocol and timed up-and-go (TUG) assessments were also completed.15–17 The 6MW is often cited as providing a more meaningful assessment of functional performance with respect to daily activity. TUG serves as an integrative assessment of functional and strength capacities and is often used to gauge susceptibility to falls. The personnel assessing 6MW and TUG was blinded to the patients’ exercise group assignments.

Physical activity was assessed using the Community Healthy Activities Model Program for Seniors (CHAMPS) Physical Activity Questionnaire for Older Adults,18 an assessment that captures weekly frequency and total time spent in different activities and allows estimation of caloric expenditure.

Health-Related Quality-of-Life, Symptoms, Mood, Psychosocial Functioning.  The Minnesota Living With Heart Failure Questionnaire (MLHF) was used to assess health-related quality of life. MLHF consists of 21 validated items covering physical, psychological, and socioeconomic dimensions (eg, swelling in the ankles, difficulty in climbing stairs, fatigue, feeling depressed, spending money for HF). The score ranges from 0 to 105, with a lower number denoting better quality of life.19

The Profile of Mood States (POMS) is a well-validated instrument for assessing emotional states that are transient and expected to respond to clinical intervention.20 It has been widely used to assess the effects of exercise interventions. We used the POMS-Brief version, which consists of 30 single-word items rated on a 5-point scale to indicate recent mood in 6 dimensions: tension/anxiety, depression/dejection, anger/hostility, vigor/activity, fatigue/inertia, and confusion/bewilderment. A decreased total mood disturbance score denotes an improved emotional state (range 0–200). Prior studies of tai chi have reported improvement in mood, reduction in anxiety, and enhancement in vigor as measured by the POMS scale.21

The Self-Efficacy-Barriers to Exercise Scale is a reliable and validated measure that assesses one’s confidence to perform exercise 3 times a week for 20 minutes each in the face of different barriers (range 0–100). Items are scored from 0 to 10 (not confident to very confident).22

Biomarkers.  B-type natriuretic peptide (BNP) was analyzed on whole blood collected in EDTA using a commercially available BNP test point-of-service meter (fluorescence immunoassay). Serum BNP >100 pg/mL supports a diagnosis of symptomatic HF.

Echocardiography.  Serial echocardiographic studies were performed at baseline and at 12 weeks by a blinded assessor using a commercially available echocardiographic machine. Assessments included mitral annular tissue Doppler measures, strain, strain rate, and flow propagation velocity. In particular, E/A (the ratio of mitral peak velocity of early filling to mitral peak velocity of late filling) and E/e′ (the ratio of mitral peak velocity of early filling to early diastolic mitral annular velocity) were calculated to estimate degree of diastolic dysfunction. Image analysis was performed offline using a digital system. Tapes were read by a single experienced cardiologist who was blinded to patient treatment assignment.

Cardiorespiratory Assessments During Exercise

At weeks 2 and 10 of training, all patients underwent cardiorespiratory assessments for 5 minutes during each of 3 separate training phases. Mean VO2, respiratory rate, and heart rate were assessed to provide a composite gauge of exercise intensity. In the tai chi group, VO2 was assessed during warm-up, meditation, and tai chi movement phases. In the aerobics group, VO2 was assessed during warm-up, cool-down, and the aerobic training phases. Gas exchange was assessed in-class at the community health facility using a portable metabolic cart and heart rate measurement was captured using a Holter monitor, digitized at 128 Hz, and annotated using a Holter scanner.

Statistical Analysis

All statistical analyses were performed on an intention-to-treat basis. Two-sample Wilcoxon rank-sum tests were used to compare the distribution of change from baseline to 12 weeks between treatment and control groups for exercise and functional parameters, questionnaire scores, echocardiographic indices, and BNP. Continuous data are summarized as mean (standard deviation). Cardiorespiratory assessments during exercise are reported using descriptive statistics. Comparisons were also made between groups examining the pooled mean (not separated by phase or timepoint) for heart rate, percent maximal heart rate, respirations, VO2, and Borg rating using Wilcoxon rank-sum tests.


Baseline Characteristics

Characteristics of the 16 patients enrolled in the study are described in Table II. Mean age (±standard deviation) of the study population was 66(±12) years, and the distribution of New York Heart Association functional class I, II, and III was 3, 9, and 4 patients, respectively. Based on echocardiographic indices, 7 patients had mild diastolic dysfunction (abnormal relaxation), 6 patients had moderate dysfunction (pseudo-normalization), and 3 patients had severe dysfunction (restrictive filling); 7 patients exhibited left ventricular hypertophy. All patients were clinically euvolemic at baseline. The groups were similar with respect to demographics and comorbidities except for physical activity and POMS score. Members of the tai chi group reported more baseline activity and relatively worse moods. There were no differences in baseline exercise performance or quality-of-life scores.

Table II. Baseline Characteristics of the Study Sample
 Tai Chi (n=8)Aerobic Exercise (n=8)
  1. Abbreviations: ACE, angiotensin-converting enzyme; ARBs, angiotensin receptor blockers; BP, blood pressure; RER, respiratory exchange ratio; 6MW, 6-minute walk test; VO2, peak oxygen consumption. Data are presented as mean±standard deviation (SD) or number (percentage). aAs measured by the Community Healthy Activities Model Program for Seniors questionnaire.

Demographic factors
 Age, mean±SD, y 68±11 63±11
 Male sex4 (50)4 (50)
  Black0 (0)3 (37)
  White 8 (100)5 (62)
Clinical classification
 Body weight, kg 87±32 94±35
 Body mass index, kg/m2 32±10 34±14
 Systolic BP, mm Hg135±10134±19
 Diastolic BP, mm Hg 74±10 68±11
 Heart rate, bpm 64±10 68±11
 New York Heart Association
  Class I1 (12)2 (25)
  Class II4 (50)5 (63)
  Class III3 (38)1 (12)
 Weekly caloric expenditure of reported activities, kcal/wka 2.4±1.91.3±2.3
Physical capacity
 Peak VO2, mL/kg/min14.5±713.1±5
 Peak RER 1.15±0.08 1.15±0.14
 Exercise time, min 7.2±6.3 6.6±2.8
 6MW, m335.4±174349.7±216
 ACE inhibitors/ARBs6 (75)4 (50)
 Digoxin0 (0)0 (0)
 Diuretics6 (75)6 (75)
 β-Blockers6 (75)5 (62)
 Calcium antagonists2 (25)5 (62)
 Nitrates1 (12)1 (12)
Self-reported comorbidities
 Coronary artery disease3 (37)3 (37)
 Arrhythmia2 (25)3 (37)
 Hypertension5 (62)7 (87)
 Diabetes2 (25)1 (12)
 High cholesterol4 (50)4 (50)
 Asthma1 (12)3 (37)
 Anxiety4 (50)1 (12)
 Depression3 (37)3 (37)
 Arthritis2 (25)3 (37)
 Cancer3 (37)1 (12)
 Renal disease3 (37)3 (37)

Feasibility, Adherence, and Safety

We demonstrated feasibility of tai chi as an intervention for HFPEF. Patients were willing to be randomized. Mean class attendance rate was 89% for tai chi and 88% for aerobic exercise training. Data are unavailable with respect to specific frequency and duration of home practice. However, 7 of 8 in the tai chi group and 4 of 8 in the aerobics group self-reported compliance with home practice 3 times a week. No adverse events or hospitalizations occurred during class sessions or the study period. During a follow-up phone call conducted 3 months after the intervention ended, 6 of 8 in the tai chi group reported continued practice of tai chi; 2 of these individuals reported starting aerobic training in addition to tai chi. Four of 8 in the aerobics group reported continued aerobic training.

Change in Clinical Outcomes After 12 Weeks

Table III, Table IV, and Table V present outcome measures at baseline and 12 weeks. Compared with patients randomized to conventional aerobic exercise, patients who performed tai chi showed no significant differences in peak VO2 but relatively greater improvements in 6MW.

Table III. Exercise and Functional Performance
 Tai ChiAerobic Exercise P Valuea
Baseline12 WeeksBaseline12 Weeks
  1. Abbreviations: BP, blood pressure; bpm, beats per minute; VO2, peak oxygen consumption. Data are presented as mean±standard deviation. aComparing the change scores between groups.

Peak exercise (bike)
 Peak VO2, mL/kg/min14.5±715.2±613.1±513.0±4.73
 Exercise duration, min7.2±68.3±66.6±27.2±3.81
 Heart rate, bpm122±29116±25119±22118±24.95
 Systolic BP, mm Hg174±15172±30198±30171±28.18
 Diastolic BP, mm Hg81±679±1185±1375±16.22
 Respiratory exchange ratio 1.15±0.081.14±0.041.15±0.141.14±0.10.67
6-Min walk, m335.4±174404.2±190349.7±216360.1±205.02
Time up and go, s15.9±1412.7±917.8±1512.8±81
Table IV. Quality of Life, Mood, Self-Efficacy
 Tai ChiAerobic Exercise P Valuea
Baseline12 WeeksBaseline12 Weeks
  1. Data are presented as mean±standard deviation. aComparing the change scores between groups. bLower score denotes improvement.

Minnesota Living With Heart Failure questionnaire totalb32.8±1828.7±1642.0±3028.6±25.09
Profile of Mood States: Total Mood Disturbanceb20.0±217.8±8 5.2±15 5.2±15.13
Profile of Mood States: Depressionb4.0±22.3±21.3±23.0±3.05
Exercise self-efficacy59.6±3066.6±2950.3±24  53±28.18
Table V. Echocardiography and B-Type Natriuretic Peptide
 Tai ChiAerobic Exercise P Valuea
Baseline12 WeeksBaseline12 Weeks
  1. Data are presented as mean±standard deviation. aComparing the change scores between groups. bRatio of mitral peak velocity of early filling (E) to mitral peak velocity of late filling (A). cRatio of mitral peak velocity of early filling (E) to early diastolic mitral annular velocity (E’).

Left ventricular ejection fraction, %62±962±965±864±7.48
Left atrial dimension, mm 3.8±0.4 3.7±0.3 3.7±0.3 3.8±0.3.04
Left atrial volume, mL 77±22 78±1970±7 68±24.94
E/A ratiob 1.2±0.6 1.1±0.4 1.3±0.8 1.1±0.7.12
E/e′ ratioc14±413±517±414±4.01
B-type natriuretic peptide, ng/mL98±85 98±116 72±85107±99.11

While both groups had improved quality-of-life scores, there was no differential change between groups detected based on MLHF score. However, the tai chi group showed improvements in the POMS-Total Mood Disturbance and Depression subscale.

There were increases in physical activity outside of classes as measured by the CHAMPS questionnaire in both groups, but no statistically significant difference between the groups. There were no significant changes in BNP in relation to exercise training. Echocardiographic indices of diastolic filling showed that E/e′ improved after exercise in the aerobics group more so than in the tai chi group, while left atrial size decreased in the tai chi group.

Characterization of Tai Chi and Aerobic Training Intensity

Table VI shows the mean values of cardiorespiratory data collected during 3 phases of both tai chi and aerobic exercise sessions. Mean heart rate and VO2 were lower in tai chi. Respiratory rates were also slower in tai chi, particularly in the phase of training corresponding to meditation, which deliberately incorporates slow, deep breathing.

Table VI. Cardiorespiratory Measurements During Tai Chi and Aerobic Exercise Traininga
  1. Abbreviations: bpm, beats per minute; SD, standard deviation. aP<.01 for comparison between groups of each variable mean (heart rate, % maximal heart rate, respiratory rate, peak oxygen uptake, rating of perceived exertion). Mean values (combining each phase) not shown. bMeasuring overall perceived exertion (scale 0–10). ct1=week 2, t2=week 10.

 Heart Rate, bpm±SD%Heart Ratemax, %±SDRespiratory Rate, bpm±SDPeak Oxygen Uptake, mL/kg/min±SDRate of Perceived Exertion (Borg)b
t1c t2c t1 t2 t1 t2 t1 t2 t1 t2
Tai chi
 Warm up72.8±1174.4±1262±1866±1425.2±626.8±86.4±1.86.0±1.32.6±21.5±1
Aerobic exercise
 Warm up88.9±1696.2±1476±1483±1724.9±832.1±108.1±3.89.4±2.92.3±12.1±1
 Cool down89.1±1291.3±1176±1279±1625.4±625.3±67.1±2.97.0±2.23.1±12.1±1


While HFPEF is common and associated with high morbidity and mortality rates, definitive treatment remains unclear. Exercise options that promote adherence are needed. Overall, this pilot study provides preliminary information regarding feasibility and potential safety of tai chi. Despite its relatively lower aerobic training intensity, tai chi may provide benefits that are similar to a typical conventional, community-based low-impact aerobics regimen in this population.

Within the context of this small study, we found that 6MW distance improved more significantly than with aerobic training. Likewise, depression scores decreased to a greater extent with tai chi. While preliminary, these findings are provocative. One possible physiological rationale is that as a training stimulus, tai chi places relatively more emphasis on skeletal muscle strengthening, stretching, and neuroregulatory enhancements that can translate into significant functional gains and might be achieved rather efficiently using tai chi rather than conventional aerobic exercise techniques. While strength was not measured here, there are trials in the elderly that suggest that tai chi can enhance peripheral muscle strength (legs and hand grip), peripheral blood flow, and neuromuscular coordination (gross postural control and fine motor skills).9,11,23–25

Whereas Kitzman and colleagues5 showed benefit of 16 weeks of aerobic training for patients with HFPEF, related analyses attributed peripheral differences in skeletal muscle to be a primary determinant of these training benefits. Aerobic training is associated with important peripheral advantages, particularly in regard to increased endothelial health and improved vasodilatory and perfusion capacities. However, strength and neuroautonomic physiologic effects also play important roles. These may be adequately achieved by other nonconventional exercise training modes. Consistently, guidelines for systolic HF call for strength training as part of a composite exercise regimen.26–28 In addition, intrinsic respiratory limitations have been recognized as an important component of HF pathophysiology with bearing on exercise intolerance, well-being, and prognosis. Inspiratory muscle training may provide key benefits, including improved inspiratory muscle and peripheral strength, functional capacity, ventilatory efficiency, perception of dyspnea, and quality of life.29,30 Tai chi constitutes a unique blend of these exercise training attributes that may potentially address aerobic, strength, and even respiratory aspects of HFPEF. Importantly, studies of tai chi in varied populations have suggested good adherence and significant patient well-being.11

A related aspect of this study is that training intensity of tai chi appeared considerably lower than aerobic training while benefits were similar. Reduced VO2 during tai chi training is likely related to the lower heart rate, since VO2 is based, in part, on heart rate responses. However, reduced breathing rate during tai chi suggests fundamentally different training effects compared with aerobic exercise.

Implications bear not only on the physiological gains facilitated by training, but the behavioral advantages of a lower cardiorespiratory training burden. Prior literature shows that exercise adherence and efficacy is more likely to be achieved with a regimen that can provide effective low-intensity options, particularly as exercise is initiated, but that can also progress in intensity as tolerated or facilitate intermittent intervals of higher intensity.31–34 In this respect, tai chi, which is inherently less strenuous but easily modified to tailor intensity, may directly address tolerability and accessibility, and thus, facilitate exercise adoption and maintenance long-term.

Safety implications may also be relevant. The slower heart rate and cardiovascular workload of tai chi as compared with aerobic exercise constitutes a more modest cardiovascular risk. The likelihood of significant adverse arrhythmia, ischemia, and/or other cardiovascular instability is proportionally reduced. Apparent safety with tai chi practice has been reported in prior trials in high-risk cardiovascular patients,35–37 as well as other populations, including the transitionally frail,38 patients with arthritis,39,40 and balance-impaired patients with vestibular disease.23 Tai chi has also been proposed to be particularly suitable in older age, and thus among patients particularly predisposed to HFPEF. Optimal HFPEF exercise therapy must therefore address the fundamental pathophysiology of the disease itself, as well as the context of frailty and age-related limitations that are inherently associated with the disease.


Key limitations of the study are the small sample size and single recruitment site potentially affecting overall generalizability. This study, however, was designed to be a pilot study to assess feasibility and gather preliminary estimates for a future trial. Nonetheless, because of the small sample, the likelihood of spurious findings may increase. In particular, we found that E/e′ improved after aerobics, suggesting that aerobic training facilitated improved lucitropic function. While prior studies suggest that training benefits in both HFPEF and with left ventricular dysfunction are mediated primarily by peripheral physiological adaptations rather than intrinsic cardiac effects,27 a recent trial by Edelmann41 reported improvements in diastolic function. It is also noted that the aerobic exercise group did not have significantly improved peak VO2 after 12 weeks, despite our cardiorespiratory data suggesting that adequate training should have been achieved. Limitations in aerobic training efficacy in HF have been described, and it is possible that these are amplified in older adults, specifically with HFPEF with a typical community-based intervention.


Despite the prevalence of HFPEF, clinical studies in this population have been notoriously challenging as many confounding issues (eg, chronic pulmonary disease and normal age-related ventricular diastolic filling changes) complicate identification of an ideal study population. Our study, albeit with a small sample, included clinical, echocardiographic, and serological parameters to assess patients with HFPEF. To our knowledge, this is the first study to demonstrate the applicability of tai chi to an HFPEF population and to characterize training intensity as compared with aerobic exercise in this population. In summary, our data suggest that tai chi may achieve exercise training benefits with improved 6MW test that are similar to a typical community-based low-impact aerobic exercise in patients with HFPEF. Benefits were seen with tai chi at a lower training intensity that may potentially facilitate safety and longer-term exercise adherence. As HF is a pervasive and rapidly growing medical problem, our data provide intriguing rationale for further study of tai chi exercise in this cardiac population.


Acknowledgements:  This study was supported by an award from the National Center for Complementary and Alternative Medicine (Yeh, K23 AT00002624) and in part by the Beth Israel Deaconess Medical Center General Clinical Research Center grant (RR 01032) from the National Institutes of Health.