Mind–body practices for patients with cardiac disease: a systematic review and meta-analysis

Abstract

Background

Due to new treatment modalities in the last decades, a decline in cardiovascular deaths has been observed. There is an emerging field of secondary prevention and behavioural programmes with increased interest in the use of mind–body practices. Until now, these have not been established in cardiovascular disease treatment programmes.

Design

We performed a systematic review and meta-analysis of the available evidence on the effectiveness of mind–body practices for patients with diagnosed cardiac disease.

Methods

We included randomized controlled trials (RCTs), published in English, reporting mind–body practices for patients with diagnosed cardiac disease. EMBASE, MEDLINE, Pubmed, Web of Science, The Cochrane Central Register of Controlled Trials and PsycINFO were searched up to July 2013. Two reviewers independently identified studies for inclusion and extracted data on study characteristics, outcomes (Quality of Life, anxiety, depression, physical parameters and exercise tolerance) and quality assessment. Standardized effect sizes (Cohen’s d) were calculated comparing the outcomes between the intervention and control group and random effects meta-analysis was conducted.

Results

We identified 11 unique RCTs with an overall low quality. The studies evaluated mindfulness-based stress reduction, transcendental meditation, progressive muscle relaxation and stress management. Pooled analyses revealed effect sizes of 0.45 (95%CI 0.20–0.72) for physical quality of life, 0.68 (95%CI 0.10–1.26) for mental quality of life, 0.61 (95%CI 0.23–0.99) for depression, 0.52 (95%CI 0.26–0.78) for anxiety, 0.48 (95%CI 0.27–0.69) for systolic blood pressure and 0.36 (95%CI 0.15–0.57) for diastolic blood pressure.

Conclusions

Mind–body practices have encouraging results for patients with cardiac disease. Our review demonstrates the need for high-quality studies in this field.

Meditation, mind–body, cardiovascular disease, psychosocial risk factors, secondary prevention, systematic review

Introduction

Advances in medical treatment have resulted in a decline in mortality from cardiovascular disease (CVD).¹ CVD is, however, still the leading cause of death globally.² Furthermore, the decline in CVD mortality has led to an increase in prevalent CVD, requiring treatment and secondary prevention of more and more patients with documented disease. Traditionally, secondary prevention has focused on well-established and modifiable risk factors, such as smoking, hypertension, diabetes, dyslipidaemia and physical inactivity. An emerging field is that of psychosocial risk factors (anxiety, depression and stress) in the aetiology and prevention of CVD.^3–5

Clinical trials studying modification of psychosocial risk factors in CVD have focused on the use of behavioural and psychological interventions in the setting of cardiac rehabilitation. Multiple studies show promising results of stress management on psychological outcomes and mind–body practices have become popular for stress reduction.⁶ Commonly used in Eastern cultures for centuries, mind–body practices have more recently been making their way into Western lifestyle and clinical practice with currently almost 20% of the population routinely doing some form of mind–body practice.^7,8 Furthermore, evidence is accumulating that mind–body practices can be used as safe adjuncts to existing medical treatment and are effective in several conditions, such as insomnia, chronic pain, depression, post-traumatic stress, irritable bowel syndrome (IBS), hypertension and cardiovascular disease (CVD).^9–13

Mind–body practices predate modern biomedicine but there is accumulating evidence on the connection between the mind and body.¹⁴ The encouraging results on improved psychological well-being and measurable changes in physical parameters suggest that mind–body practices can be of added value in patients with CVD. The aim of this review was to evaluate the effectiveness of meditation-based mind–body practices on quality of life, anxiety, depression, physical parameters and exercise tolerance in patients with diagnosed cardiac disease.

Methods

Search strategy

Together with a professional librarian we conducted electronic searches through Embase, Medline, Pubmed, Web of Science (WoS), The Cochrane Central Register of Controlled Trials (CENTRAL) and PsycINFO of the literature published up to 31 July 2013. We reviewed mind–body practices with the focus on meditation and relaxation in patients with diagnosed cardiac disease¹⁵ that could be taught without external biofeedback techniques. Practices considered included: (transcendental) meditation, mindfulness meditation, autogenic training and relaxation methods. The diagnosed cardiac diseases included were: heart failure, ischaemic heart disease, hypertensive heart disease, inflammatory heart disease, valvular heart disease, congenital heart disease and cardiomyopathy. The search included keywords corresponding to the mind–body techniques and cardiac diseases. To optimize the search strategy we used synonyms including ‘text words’ and MeSH (Medical Subject Headings) (Appendix A, see online Supplementary Material). We screened for other potentially relevant randomized controlled trials by searching through the references and citation lists from identified papers meeting our inclusion criteria.

Study selection

Two reviewers (JY and CB) independently selected all potential studies, which were identified through the results of our search strategy. Disagreements were resolved by consensus. We included relevant studies if they met the following criteria: (1) written in English; (2) study design was a randomized controlled trial (RCT); (3) study population with diagnosed cardiac disease; (4) the mind–body practice was compared to standard care, pharmacological intervention, psychological intervention or no treatment at all; (5) at least one of the following clinically relevant outcomes was assessed – any form of Quality of life (QoL) measurement as described by Ski et al.¹⁶ and Wenger et al.¹⁷ including subjective health status and health-related QoL, mental health scales on anxiety or depression, physical parameters (blood pressure and heart rate) or exercise tolerance (measured by 6-minute walk test or cardiopulmonary exercise testing).

We selected studies reporting both objective and physiological outcomes because in cardiac disease especially, physiological parameters are good indicators of overall physical well-being. Several studies show the hypothesis and rational of mind–body interaction. We know that hypertension is one of the major risk factors of cardiovascular morbidity and mortality.¹⁸ The prevalence of hypertension is projected to increase to about 60% in 2025.¹⁹ Conventional medical treatment may be associated with various adverse effects,²⁰ especially in cases of multiple drug use and treatment-resistant hypertension, which has led to investigations of the supporting role of non-pharmacological interventions.

We chose to limit our search to English written studies based on two reports, which showed no systematic bias with the use of a language restriction.^21,22

We excluded studies if they met one of the following criteria: (1) the article was a review or meta-analysis; (2) data were from abstracts or letters; (3) studies that evaluated psychological or psychosocial interventions focusing on interactions between people; (4) studies in which the mind–body practice was not the main area of interest; (5) studies that evaluated stress management with a major cognitive component; (6) studies evaluating practices using an external biofeedback technique; and (7) studies that evaluated practices performed during the peri-operative phase.

Data extraction

Data from studies were extracted independently by two authors (JY and RG) using a data extraction form in accordance with the Cochrane handbook of systematic reviews.²³ The following information was obtained using this form: author, journal, year of publication, country, setting, funding, number of patients, cardiac diagnosis, mean age, gender, co-morbidity, use of medication, type of mind–body practices, details of the practices, comparators, follow-up duration, outcome definition, unit of measurement, pre- and post-outcome measurements on exercise tolerance, QoL, anxiety, depression, blood pressure and heart rate. If more than one practice or control arm was present, we included only the mind–body practices versus the control group. Disagreements were resolved by consensus. If data were unclear or unavailable, the authors were contacted by email. In case of multiple publications from one source, we only included data from different papers if different outcomes were reported. Resting heart rate and blood pressure results were used and not values measured during the relaxation practice.

Quality assessment

Two reviewers (JY and RG) independently scored the methodological quality of each study by using the Cochrane Collaboration ‘risk of bias’ tool²³ in which the following domains are considered: (1) random sequence generation; (2) allocation concealment; (3) blinding of participants and personnel; (4) blinding of outcome assessment; (5) incomplete outcome data; (6) selective reporting; (7) other sources of bias. The judgment is categorized as follows: A. ‘Low risk’ of bias. B. ‘High risk’ of bias. C. ‘Unclear risk’ of bias. Additionally, the heterogeneity between studies was reported and the PRISMA statement for reporting systematic reviews and meta-analysis was used for reporting of results.²⁴ Disagreements were resolved by consensus. Small sample size can lead to bias but no formal agreement exists on how large the sample size of a study should be to limit the risk of bias. Consensus between JY and RG led to the cut-off point of 30 participants per study to score the study at high risk of ‘other bias’ due to their small sample size.

Data synthesis and analysis

To analyse the maximum treatment effect we used data of post-treatment values on the pre-defined outcome measurements. If multiple follow-up data were available we used the outcome assessment at the end of the intervention period in order to determine the maximum treatment effect. If multiple follow-up data were unavailable we used the data as reported. We calculated the Cohen’s d effect size, comparing the intervention and control groups, of the means and confidence intervals according to the equations as shown in Appendix B (see online Supplementary Material). Data for each outcome were summarized in forest plots, and random effects models were used to calculate summary estimates with 95%CI. Funnel plots were constructed to check for the presence of publication bias. Of the eight domains of the Short-Form Health Survey-36, we present the summary measures of the Physical and Mental Component Score (PCS and MCS respectively). QoL measures were grouped according to their subscale (i.e. physical, mental/emotional or other). If more than one QoL measure was reported (i.e. The Short-Form Health Survey 36 (SF-36) and the Minnesota Living With Heart Failure Questionnaire (MLWHFQ)), the SF-36 was used over the other measures in the random effects meta-analysis.

Results

Study selection

Our literature search resulted in 1487 studies, 1485 from the original search and two from reference lists (Figure 1). Of these, 551 studies were excluded because they were duplicates and 902 studies were excluded when title and abstract were reviewed. Of the 34 studies selected for full text review, 21 were excluded for not meeting the predefined inclusion criteria (Figure 1). Finally, 13 RCTs, consisting of 11 unique studies, met our inclusion criteria and were included in this review. They were published between 1996 and 2012 and were carried out in six different countries (UK, USA, China, Brazil, India and Portugal).

Figure 1.

Flow diagram of identified studies.

General characteristics

The characteristics of interventions and patients are presented in Table 1. The studies included 793 unique patients (46% female) with a mean age of 66 (±11 years). The studies ranged in sample size from 18 to 201 patients. There were two cases of duplicate reports on the same patient population.^25,26 One study did not fully report its baseline characteristics.²⁷

Table 1.

Study characteristics of included randomized controlled trials.

Author	Year	Study location	Cardiac diagnosis	Age (years) (mean ± SD)	Female %	N
Trzcieniecka-Green	1996	UK	AMI/CABG	60 ± 10	13	100
Wilk	2002	USA	CABG/MI/Angioplasty	63 ± NR	14	14
Tacon	2003	USA	Cardiaca	61 ± 9	100	18
Robert-Mccomb	2004	USA	Cardiaca	61 ± 8	100	18b
Curiati	2005	Brazil	HF	75 ± 7	74	19
Chang	2005	USA	HF	70 ± 14	0	63
Paul-Labrador	2006	USA	CAD	67 ± 14	18	103
Jayadevappa	2007	USA	HF	64 ± 11	61	23
Yu	2007	China (HK)	HF	76 ± 8	51	121
Yu	2009	China (HK)	HF	76 ± 8	51	121b
Neves	2009	Portugal	CAD	60 ± 15	85	81
Schneider	2012	USA	CAD	59 ± 15	43	201
Nehra	2012	India	CHD	n/a	n/a	50
All				66 ± 11	51	743b

AMI, acute myocardial infarction; CABG, coronary bypass artery grafting; CVD, cardiovascular disease

CAD, coronary artery disease; HF, heart failure; CHD, coronary heart disease; n/a, not available.

a

Patients diagnosed with angina pectoris, valvular disease, hypertension or CAD

b

743 unique patients.

Patients

Most studies included patients with heart failure (n = 5),^26,28–31 or coronary artery disease (n = 4).^27,32–34 One study only recruited heart failure patients after optimal medical treatment and at least two months of carvedilol therapy.²⁹ Two studies included patients following AMI, angioplasty or CABG, when considered in a stable condition (2–3 months post surgery or percutaneous intervention).^35,36 Two studies included various diagnoses including angina, hypertension, valve disorders and coronary artery disease.^25,37

Interventions

The included studies compared the effects of different mind–body practices with various control interventions. The mind–body practices studied were: mindfulness-based stress reduction,^25,27,37 transcendental meditation,^30,33,34 meditation (which consisted of three components: breathing, mental repetition of a word and a guided image),²⁹ progressive muscle relaxation training,^26,31,36 relaxation response (which consisted of eight different components: breathing awareness; mental repetition of a word, sound, phrase or prayer; mindfulness meditation; guided body scan; progressive muscle relaxation; guided countdown; autogenic and guided imagery),²⁸ relaxation,³² and stress management (based on autogenic training).³⁵ Further details and the content of the interventions are shown in Table 2. Two studies evaluated a mind–body practice in addition to regular cardiac rehabilitation compared to cardiac rehabilitation alone.^32,36 The duration of the interventions ranged from 4 to 26 weeks. One study did not provide the duration of the intervention.³⁴ All but one study³² gave daily home exercises by audiotape or instructions. In one study the trial was conducted in two parts due to a hiatus in funding.³⁴

Table 2.

Details of interventions in included studies.

Intervention
Author	Year	Name	Details	Home assignments	Duration (weeks)	Control goup
Trzcieniecka- Green	1996	Stress management	Weekly sessions	Audiotape on relaxation, twice daily	10	Waiting list
Wilk	2002	Progressive muscle relaxation (+Cardiac rehabilitation)	50 minute weekly sessions	Audiotape on PMR	4	Cardiac rehabilitation
Tacon	2003	Mindfulness based stress reduction	Two-hour weekly sessions	Audiotape, daily	8	Waiting list
Robert-Mccomb	2004	Mindfulness based stress reduction	Two-hour weekly sessions	Audiotape, daily	8	Waiting list
Curiati	2005	Meditation	Two classes 1hour introduction	Audiotape	12	Weekly talks
Chang	2005	Relaxation response	Weekly 90 min group sessions	Audiotape	15	Usual care
Paul-Labrador	2006	Transcendental meditation	Introduction week, one month of biweekly meetings, weekly thereafter	Daily home assignments	16	Health education
Jayadevappa	2007	Transcendental meditation	Seven days introduction, 3 months biweekly meetings, 3 months monthly meetings	Twice daily, 20 minutes of TM	26	Health education
Yu	2007	Progressive muscle relaxation (grade 2)	One-hour weekly sessions, biweekly phone calls, skill skill revision at 4 weeks	Audiotape	14	Attention control (phone calls)
Yu	2009	Progressive muscle relaxation	One-hour weekly sessions, biweekly phone calls, skill skill revision at 4 weeks	Audiotape	14	Attention control (grade 2) (phone calls)
Neves	2009	Relaxation (+ Cardiac rehabilitation)	3 one-hour weekly sessions after cardiac rehabilitation session	None	10	Cardiac rehabilitation
Schneider	2012	Transcendental meditation	Introduction week, 1 month weekly meetings, 2 months biweekly meetings, monthly meetings thereafter	Twice daily, 20 minutes of TM	n/a	Health education
Nehra	2012	Mindfulness based stress reduction	Eight weekly individual 150 minutes sessions and one 7 hour weekend session and extended MBCT	Yes	10–17	Usual care

PMR, progressive muscle relaxation; TM, transcendental meditation; MBCT, mindfulness-based cognitive therapy.

Control intervention

The control interventions used were: waiting list,^25,35,37 usual care (in which the patients were expected to solely attend the two moments of study outcome assessment and were not invited to group sessions),^27,28 attention control (extra attention by phone calls to balance the effect of the extra attention that patients received in the PMRT training).^13,26,31 Some studies made comparisons with an active intervention, such as health education^30,33,34 or cardiac rehabilitation.^32,36 One study had weekly talks on stress management.²⁹ The study by Chang et al.²⁸ used two control groups (education and usual care) of which we present only the usual care.

Risk of bias

The risk of bias of included studies is shown in Table 3. Due to the nature of the intervention, none of the studies was able to blind the participants and personnel teaching the intervention. Blinding of the outcome assessor is possible and was included in Table 3. Furthermore, the presence of small groups often made the risk of other bias inevitable.

Table 3.

Risk of bias in included studies.

Random sequence generation

Five studies reported the randomization procedure used.^{26–28,30,31,33} The other studies did not provide enough information to judge which randomization procedure was used and were classified as ‘Unclear risk’.

Allocation concealment

Only three studies reported allocation concealment.^30,33,34 Most studies failed to state clearly how randomization had been achieved and were judged as ‘Unclear risk’.

Blinding of participants, personnel and outcome assessment

The outcome assessor was blinded in five studies.^{26,30,31,33,34} In one study the outcome assessor was also the instructor of the intervention making blinding to the outcome impossible.³⁶ In the other studies the assessment procedure of the outcome was unclear.

Addressing incomplete outcome data

Almost all studies reported missing data whereas three studies did not.^26,31,35 Eight studies reported adequate information about how many participants had withdrawn after having consented to participate.

Selective reporting

Most studies reported outcomes that were predefined in the methods section and were thus judged as low risk of selective reporting bias. In three studies the risk of selective reporting was considered high since some of the outcomes mentioned in the methods section were not reported.^25,27,37 Furthermore, one study failed to provide enough statistical data (i.e. standard deviations were missing).³⁶

Other bias

We considered other sources of bias to be present if the study included a small sample size (n < 30), was reported to be underpowered, or minimal or no group comparison at baseline was reported. Almost all studies had a small sample size. Only four studies included 100 or more patients.^26,31,33,35

Heterogeneity across studies

Overall, based on the clinical diversity of the included patients in the studies (Table 1), the methodological diversity in study design and risk of bias (Table 3) and the statistical heterogeneity of outcomes reported, we conclude there is significant heterogeneity across studies.

Subjective outcomes

Quality of Life

Six RCTs reported quality of life outcomes assessed with five different questionnaires. Results measurements used are presented in Table 4. Pooled effect sizes of the physical QoL measures revealed an overall medium statistically significant effect size of d = 0.45 (95% CI 0.18–0.72) and an overall medium statistically significant effect size of d = 0.68 (95%CI 0.10–1.26) for mental scores (Figure 2). The other subscales showed a small non-significant effect of d = 0.25 (95%CI -0.10–0.61). The results were mainly influenced by the large study of Yu et al.²⁶

Figure 2.

Forest plot of QoL results.

CI, confidence interval; n(I), number of patients in intervention group; n(C), number of patients in control group.

^aWHOQOL-BREF-HK, World health Organization Quality of Life Hong Kong version. ^bSF-36, Short-Form Health Survey 36. ^cMLWHQ, Minnesota Living With Heart Failure Questionnaire. ^dQWB, Quality of Well-Being. ^eGWB, General Well-Being.

Table 4.

Type of outcome in included studies (methods).

Author	Year	Outcome	Type of scale	Moment of assessment
Trzcieniecka-Green	1996	QoL, anxiety, depression	GWB, HADS	Baseline, week 10, week 26
Wilk	2002	Anxiety, DBP, SBP, HR	STAI	Baseline, week 4
Tacon	2003	Anxiety	STAI	Baseline, week 8
Robert-Mccomb	2004	QoL	SF-36	Baseline, week 8
Curiati	2005	QoL, exercise tolerance, DBP, SBP, HR	MLWHFQ, VO2max	Baseline, week 14
Chang	2005	QoL, exercise tolerance	MLWHFQ, VO2max	Baseline, week 15–19
Paul-Labrador	2006	Anxiety, depression, SBP, DBP	STAI, CESD	Baseline, week 16
Jayadevappa	2007	QoL, depression, exercise tolerance	SF-36, MLHFQ, QWB,	Baseline, week 12, week 26
			CESD, 6MWT
Yu	2007	Anxiety, depression	HADS	Baseline, week 8, week 14
Yu	2009	QoL	WHOQOL-BREF-HK	Baseline, week 8, week 14
Neves	2009	SBP, DBP, HR		Baseline, week 12
Schneider	2012	Depression, SBP, DBP, HR	CESD	Baseline, week 12, every 26 weeks
Nehra	2012	Anxiety, depression	HADS	Baseline, week 10–17

QoL, quality of life; GWB, general well-being; HADS, Hospital Anxiety and Depression Scale; DBP, diastolic blood pressure; SBP, systolic blood pressure; HR, heart rate; STAI, State-Trait Anxiety Inventory; SF-36, 36-item Short Form Health Survey; MLWHFQ, Minnesota Living with Heart Failure Questionnaire; HRV, heart rate variability; CESD, Center for Epidemiologic Studies Depression Scale; QWB, Quality of Well Being; 6MWT, six-minute walk test; WHOQOL-BREF-HK, World Health Organization Quality of Life-BREF Hong Kong Chinese version.

Depression

Depression was assessed with only two instruments: the Center for Epidemiologic Studies Depression Scale (CESD) and the Hospital Anxiety and Depression scale (HADS). Depression was reported in six studies. The study by Schneider et al.³⁴ consisted of repeated measurements over a period of 5.4 years. The depression outcomes revealed an overall medium statistically significant effect size of d = 0.61 (95% CI 0.23–0.99) (Figure 3). Results were heterogeneous with small effects of d = 0.14 in the study by Schneider et al.³⁴ and large effects of d = 1.25 in the study of Yu et al.³¹ (Figure 3).

Figure 3.

Forest plots of depression and anxiety.

n(I), number of patients in intervention group; n(C), number of patients in control group; CI, confidence interval.

Anxiety

Anxiety was reported in five studies. An overall medium statistically significant effect of 0.52 (95% CI 0.26–0.78) was found (Figure 3). Again, results were heterogeneous with small effect sizes ranging from d = 0.31 in the study of Paul-Labrador et al.³³ to large effects of d = 1.34 in the study of Tacon et al.³⁷ (Figure 3).

Physiological outcomes

Resting blood pressure

Blood pressure was reported in five RCTs. Due to missing SDs no effect size with 95% CI could be calculated for Wilk et al.³⁶ Effect sizes for SBP ranged from small effects of d = 0.42 in the study of Schneider et al.³⁴ to a large effect of d = 1.10 in the study of Curiati et al.²⁹ (Figure 4). The meta-analysis resulted in an overall medium statistically significant effect of d = 0.48 (95% CI 0.27–0.69). Effect sizes for DBP showed a similar heterogeneous pattern with results of d ranging from 0.25 in the study by Schneider et al.³⁴ to 0.80 in the study by Curiati et al.²⁹ Results showed an overall lower, but statistically significant, effect than the SBP results (0.36, 95% CI 0.15–0.57) (Figure 4).

Figure 4.

Forest plots of physiological parameters.

SBP, systolic blood pressure; n(I), number of patients in intervention group; n(C), number of patients in control group; CI, confidence interval; DBP, diastolic blood pressure; HR, heart rate.

Resting heart rate

Four RCTs reported on resting heart rates (Figure 4). Wilk et al.³⁶ reported that a significant difference was found but no effect size with 95% CI could be calculated due to missing SDs.

An overall small effect of d = 0.15 (95%CI -0.08–0.39) was found. All studies showed small effects of d ranging from 0.13 to 0.20 (Figure 4).

Exercise tolerance

Only three studies reported exercise tolerance as outcome (Table 5).^28–30 No effect size could be calculated for Wilk et al.³⁶ due to missing SDs. A medium effect of d = 0.51 was seen for VO2 max testing and a large effect of d = 1.04 was seen on the 6 minute walk test in the study of Jayadevappa et al.³⁰ (Table 5).

Table 5.

Results of studies that reported exercise tolerance.

		Patients analysed (n)		Effect size (95% CI)b
Author	Year	Intervention	Control	Exercise tolerance	Exercise test
Curiati	2005	8 (5)a	7 (5)a	0.51 (−0.75–1.77)	VO2max
Chang	2005	31	28	n/a	VO2max
Jayadevappa	2007	13	10	1.04 (0.15–1.91)	6MWT

CI, confidence interval; n/a, not available; 6MWT, six-minute walk test.

a

Five patients did VO2max testing

b

effect size is Cohen’s d.

Publication bias

Funnel plots of the subjective and physiological outcomes were constructed and are shown in Appendix C (see online Supplementary Material). Overall, the funnel plots show no clear evidence of publication bias.

Discussion

The eleven unique studies in this review had an overall low quality and used a variety of outcome measurements. There was some evidence for the effectiveness of mind–body practices for patients with diagnosed cardiac disease. Promising but heterogeneous results were seen on overall effect sizes of mental and physical QoL, anxiety, depression and blood pressure.

There are several reasons why patients engage in mind–body practices. Firstly, these therapies are easy to learn and they allow patients to take a more active role in their treatment. Secondly, most exercises can be done at home without the help of external means. Thirdly, low emotional and physical risk is involved. Finally, the costs are relatively low.^38–40 Although mind–body practices require commitment in both adherence and time, they are becoming more and more popular.^{7,  8}

Studies on the effectiveness of mind–body practices in CVD have shown some promising results. They improved psychosocial risk factors, reduced blood pressure and even showed survival benefits.^41–44 It is hypothesized that patients with CVD are more likely to seek additional treatment since psychological stress, whether cause or consequence, often accompanies their clinical condition.^45,46 Stress can cause an imbalance between the mind and body, and several studies suggest stress is related to CVD at several stages of the disease from the development of arteriosclerosis to acute cardiac events to chronic disease.^5,47 Furthermore, other studies have shown associations between psychosocial variables and vascular function, inflammation and increased blood clotting.^48–50 Thus, although the precise pathophysiological mechanism still needs to be unravelled, it is clear that a relationship exists between stress and CVD.

Many biological pathways have been studied that could explain the working mechanism of mind–body practices. Several studies show positive physiological effects in blood pressure, heart rate, respiration rate and oxygen consumption with mind–body practices.^10,51 Three studies showed that the autonomic nervous system releases endorphin and serotonin, which leads to counteraction of norephinephrine and activates the parasympathetic response.^10,12,51 In a small study on the effect of yoga and meditation on endothelial function, favourable changes were observed in endothelial-dependent vasodilatation in CAD patients.⁴⁴ A mechanism illustrating how mind–body practice can have influence on health is provided by research on the interaction between the central nervous system (CNS) and the endocrine, immune and peripheral autonomic nervous systems.⁵² There is also evidence for a positive effect on the immune system and endothelial functioning.^51,53,54 All together, these studies show the profound effects of stress reduction and meditation techniques on the body, suggesting that the mind may be able to influence the working of the heart.

Until now, a core component in the treatment of CAD patients is cardiac rehabilitation with the components exercise training, healthy nutrition and smoking cessation.⁵⁵ Lifestyle modification programmes have been shown to be of added value in treating patients with CAD.⁵⁶ These interventions have proven to be favourable for physical and psychosocial risk factors and also showed a survival advantage.^57–59 Additional stress management has been shown to have additional value in cardiac rehabilitation programmes.^60,61 Even though the American Heart Association (AHA) has recognized the importance of psychosocial interventions as a core component in cardiac rehabilitation programmes,⁶² these interventions have only been integrated in a limited number of settings. Furthermore, only 25–31% of eligible patients participate in these comprehensive programmes.⁶³

Some limitations of this systematic review must be addressed. Firstly, we limited our search strategy to English studies exclusively, with no consideration of studies conducted in the East – for example written in Chinese – from which mind–body practices originate. Secondly, we focused primarily on meditation-based mind–body practices that could be undertaken by patients at home without external tools. Therefore, no conclusions can be drawn on the effect of other mind–body practices in CVD.

The connection between the mind and heart is a complex phenomenon and the use of meditation-based mind–body practices is still limited in clinical practice. Clearly, the role of mind–body practices in the advanced treatment of cardiac patients still needs to be defined. Behavioural cardiology is an emerging field that is bringing about awareness of the mind–heart connection and the management of psychosocial risk factors.⁵⁷

However, current evidence on the efficacy and effectiveness of most mind–body practices has not been established due to a lack of well-performed randomized clinical trials. Mind–body research has several drawbacks that can hamper the internal validity and generalizability of published studies. A paper by Caspi et al.⁶⁴ showed several important features to consider in meditation research such as monitoring, assessment procedures, integration of qualitative methods and a pragmatic design. Most of the studies included in this review failed to provide detailed information on the randomization procedure and a study protocol was often not available. Furthermore, most studies lacked power and were too small to draw firm conclusions from. Until now, there is no clear understanding of the study design in mind–body research and future methodological studies should provide guidance.

Conclusions

In our review we showed that mind–body practices have encouraging results for patients with cardiac disease on selected QoL outcomes, anxiety, depression and blood pressure. Due to an overall low quality of studies, no firm conclusions can be drawn. Future clinical trials should focus on using a rigorous study design in order to minimize methodological flaws and enhance their validity and generalizability.

Acknowledgement

We would like to thank Wichor Bramer for his constructive help on the search strategy of this review.

Funding

JO Younge, RA Gotink, JW Roos-Hesselink and MGM Hunink were supported by an Internal Grant from the Erasmus Medical Centre.

Conflicts of interest

None declared by authors.

Author notes

*

These authors contributed equally to this article.