Mindfulness meditation, well-being, and heart rate variability: A preliminary investigation into the impact of intensive Vipassana meditation

Highlights

• Examines change in well-being after 10 days of mindfulness meditation (Vipassana).

• HRV compared pre- and post-Vipassana training at baseline and in meditation task.

• Most measures of well-being (positive and negative) improved after post-Vipassana.

• Ln HF increased pre-, normalised HF increased, LF (THM) decreased post-Vipassana.

• HRV results interpreted in terms of flow, a state of positive immersion in a task.

Abstract

Mindfulness meditation has beneficial effects on brain and body, yet the impact of Vipassana, a type of mindfulness meditation, on heart rate variability (HRV) – a psychophysiological marker of mental and physical health – is unknown. We hypothesised increases in measures of well-being and HRV, and decreases in ill-being after training in Vipassana compared to before (time effects), during the meditation task compared to resting baseline (task effects), and a time by task interaction with more pronounced differences between tasks after Vipassana training. HRV (5-minute resting baseline vs. 5-minute meditation) was collected from 36 participants before and after they completed a 10-day intensive Vipassana retreat. Changes in three frequency-domain measures of HRV were analysed using 2 (Time; pre- vs. post-Vipassana) × 2 (Task; resting baseline vs. meditation) within subjects ANOVA. These measures were: normalised high-frequency power (HF n.u.), a widely used biomarker of parasympathetic activity; log-transformed high frequency power (ln HF), a measure of RSA and required to interpret normalised HF; and Traube–Hering–Mayer waves (THM), a component of the low frequency spectrum linked to baroreflex outflow. As expected, participants showed significantly increased well-being, and decreased ill-being. ln HF increased overall during meditation compared to resting baseline, while there was a time ∗ task interaction for THM. Further testing revealed that pre-Vipassana only ln HF increased during meditation (vs. resting baseline), consistent with a change in respiration. Post-Vipassana, the meditation task increased HF n.u. and decreased THM compared to resting baseline, suggesting post-Vipassana task-related changes are characterised by a decrease in absolute LF power, not parasympathetic-mediated increases in HF power. Such baroreflex changes are classically associated with attentional load, and our results are interpreted in light of the concept of ‘flow’ — a state of positive and full immersion in an activity. These results are also consistent with changes in normalised HRV reported in other meditation studies.

Introduction

In the last two decades, psychological interventions derived from mindfulness meditation practices have been increasingly used to treat a variety of stress, pain and anxiety-related conditions (Hofmann et al., 2010). Mindfulness refers to the state of being attentive to and aware of what is taking place in the present (Brown and Ryan, 2003, Shapiro, 2009); mindfulness meditation comprises a variety of techniques that help focus attention in a non-analytical way and avoid discursive, persistent, or obsessive thoughts (Shapiro, 1980). These techniques – such as quieting the mind, and exercising self-control – can have a profound influence on mind and body, and show promise as an alternative tool to regulate emotions, mood, and stress. However, the acute and longer-term concomitants of mindfulness meditation training, and potential mechanisms of action are still not well understood. In particular, there is a need to further understand the effects of meditation on the autonomic nervous system, a major component of emotional experience. While limited research has examined the effects of Zen meditation, different styles may have distinctive effects. For instance, Zen meditators show distinctive respiration changes (Lehrer et al., 1999) that are not evident in other styles such as yoga (Sarang and Telles, 2006) or traditional Chinese practices (Tang et al., 2009). Here we examine the impact of a particularly intensive form of mindfulness meditation – Vipassana – on heart rate variability, an important psychophysiological marker of mental health and wellbeing.

One of the goals of mindfulness is to allow thoughts to arise, be examined dispassionately, and allowed to fade, without practitioners being emotionally influenced by their contents. This process is a fundamental part of Vipassana meditation (Gethin, 1998). This technique is also similar to the reappraisal strategy for emotion regulation, which can serve to decrease subjective, physiological and neural responses, rather than increasing them as is the case with emotion suppression (Gross and Levenson, 1997, Gross, 1998, Goldin et al., 2008).

Mindfulness techniques appear to be linked in a variety of ways to well-being. Important behavioural examples include reduction in distractive and ruminative thinking (Jain et al., 2007) and symptoms of anxiety and mood disorders (Goldin and Gross, 2010, Hofmann et al., 2010), and improved emotion regulation (Arch and Craske, 2006). Individual differences in the ability to regulate emotional responses are also related to differences in mindfulness, even in non-meditators (Modinos et al., 2010). These findings suggest the possibility that mindfulness meditation influences well-being via changes in emotion regulation. Furthermore, trait mindfulness was associated with wider prefrontal and diminished amygdale activity during an affect labelling task in non-meditators, suggesting a possible mechanism for the role of mindfulness in emotion regulation (Creswell et al., 2007).

Neural and physiological benefits found to be associated with meditation include: increases in immune system activity and left-sided anterior activation, a pattern associated with positive affect (Davidson et al., 2003); decreased amygdale response to emotional stimuli (Desbordes et al., 2012) and increased brain connectivity (Luders et al., 2011). Moreover, long-term meditators had greater grey matter density in regions previously found to be involved in meditation including right anterior insula (involved in interoceptive awareness), left inferior temporal gyrus and right hippocampus and right orbito-frontal cortex (Hölzel et al., 2008, Luders et al., 2009). These particular studies are relevant because of the overlap with regions of the central autonomic network related to heart rate variability, especially insular and orbitofrontal cortices (Thayer and Lane, 2000, Thayer et al., 2009). Experienced meditators demonstrated increased cortical thickness in insula and prefrontal cortices compared to matched controls (Lazar et al., 2005), as well as larger gyrification in regions including left precentral and right fusiform gyrii, and the insula (Luders et al., 2012). These findings, taken together, suggest that meditation has dramatic long-term structural effects on the brain.

Vipassana is a widespread technique of mindfulness meditation, derived from Buddhist practice, based on objective observation of physical sensations in the body. Awareness of the breath is also used as an aid to concentration. It is taught in a standardised manner throughout the world, and involves up to 100 h of intense meditative practice over a 10-day period. This intense standardised training is ideally suited to examining the effects of mindfulness meditation on well-being and related physiological changes.

Heart rate variability (HRV) is a measure of beat-to-beat variability in heart rate that is mediated by the autonomic nervous systems. Parasympathetic influence on HRV is primarily mediated by the vagus nerve, which can provoke rapid changes from cardiac cycle to cardiac cycle, and is primarily responsible for fluctuations in respiratory sinus arrhythmia (RSA) and high frequency HRV (HF) (Dexter et al., 1992, Berntson et al., 1993). Sympathetic influence is primarily controlled by release of norepinephrine and catecholamines, precluding direct manifestation in short term fluctuations (Berntson et al., 1993) Sympathetic neural activity can alter cardiac behaviour only slightly from beat to beat (Levy et al., 1993), and thus RSA measured through HF HRV is often used as a biomarker of pure PNS activity. That is, the level of vagal outflow will be reflected in the magnitude of RSA, which is typically measured at the speed of normal breathing, at cycles from approximately 3 to 7 s (i.e. 0.15–0.4 Hz). THM, on the other hand is component of low frequency (LF) HRV which reflects an oscillation of arterial pressure (Julien, 2006). Recent studies suggest that low frequency power more closely approximates baroreflex outflow, rather than sympathetic activation (Moak et al., 2009, Goldstein et al., 2011).

Amongst other things, HRV reflects the capacity of the central autonomic network (CAN) – including the prefrontal cortex, central nucleus of the amygdala, hypothalamus and brainstem – to meet and adapt to environmental demands (Thayer and Friedman, 2002). HRV underpins an individual's capacity to regulate their emotions (Geisler et al., 2010), and may be key to psychological flexibility (Kashdan and Rottenberg, 2010).

HRV is reduced in patients with cardiovascular disease (Nolan et al., 1996), and reduced HRV is an indicator of risk of cardiac and all-cause mortality (Dekker et al., 2000, Tsuji et al., 1996). A number of studies and reviews (e.g. Kemp et al., 2010, Kemp et al., 2012) have indicated that HRV is reduced in patients with depression and anxiety, even without cardiovascular disease. While studies have often focused on links between decreased HRV, negative emotions and poor physical health, increased HRV is related to well-being (Kemp and Quintana, 2013--in this issue) over and above reductions in negative affect (Boehm and Kubzansky, 2012). There is growing evidence that positive psychological attributes such as mindfulness are independently related to cardiac health and autonomic function (DuBois et al., 2012), including individual differences in resting respiratory sinus arrhythmia (RSA) (Oveis et al., 2009) and THM (Fuller, 1992).

The effects of Vipassana on HRV have not to our knowledge been systematically researched, although several other mindfulness based meditation techniques have been examined in more detail in novice and experienced meditators.

The acute task-related cardiovascular effects of Zen meditation compared to resting baseline in practitioners with varying levels of experience have been better studied than Vipassana. Lehrer et al. (1999) found that respiration rates fell dramatically during Zen breathing meditation in experienced meditation practitioners. High frequency (HF) HRV decreased as a percentage of total variance (although there were no significant changes for absolute HF power). Total heart rate (HR) oscillation amplitude increased, as did absolute low frequency (LF) power — reflecting a shift in RSA towards lower-frequency waves. These findings confirmed that Zen breathing meditation results in an increase in low and very low frequency HR oscillations, shifting the majority of HRV spectral power into the low frequency band. Within-subjects shifts in RSA during Zen meditation compared to resting baseline varied with experience (Peressutti et al., 2010). Strong HF oscillations were observed in novices; while for the most experienced practitioners, variance centred in the LF range and was linked to RSA, being associated with decreased breathing rate during meditation (Peressutti et al., 2012).

Much of this research focused on meditation tasks involves slow breathing, relative to a normal breathing condition during a resting baseline. Very different task effects were found for an inward-directed attention meditation task which did not involve controlled respiration. During this task experienced Zen meditators exhibited increases in normalised HF (HF n.u.) and corresponding decreases in normalised LF and LF/HF ratio, compared to resting non-meditators and to their own resting baseline (Wu and Lo, 2008). A number of studies have reported similar acute HF increases after other styles of meditation, e.g. yoga-based cyclic meditation, (Sarang and Telles, 2006, An et al., 2010).

Differences between studies in which meditation acutely increased HF and those in which LF increased, can be partially explained by the influence of respiration rate on HRV (Schipke et al., 1999), especially as respiration was particularly slow in the study by Lehrer et al. (1999). The acute effects of being in a state of meditation, however, cannot be entirely reduced to breathing differences; mindfulness meditation appears to have an effect over and above simple breathing interventions. A similar pattern to that found by Wu and Lo (2008) – increased normalised HF power in meditation – emerged when novices were trained to perform a mindfulness of breathing task at a breathing pace of .25 Hz, and compared to a resting control in which they performed the same paced breathing. This allowed the acute effects of the meditation task to be separated from changes in respiration rate, and demonstrated that mindfulness meditation even with deliberately paced breathing had effects over and above the effects of paced breathing alone (Murata et al., 2004, Takahashi et al., 2005).

While the above studies focused on acute changes from being in a state of meditation, rather than longer-term changes in resting baseline HRV, there is evidence of changes in resting baseline HRV as a result of mindfulness meditation training. Tang et al. (2009), compared meditation-naïve participants who were given a short course of meditation involving mindful awareness with a relaxation control. Meditation resulted in decreased stress, calmer breathing, and increased normalised HF within subjects (vs. resting baseline measures), and between subjects (vs. the relaxation group).

The majority of these results report normalised high frequency HRV (HF n.u.), which has the distinct advantage of being comparable as a measure of vagal outflow between studies using different methods of spectral analysis (Task Force, 1996). However, HF n.u. suffers from an unavoidable interpretative problem — without comparative measures of raw or adjusted power for individual frequencies, changes in those frequencies may be obscured by their being reported as proportions. For example, if LF power decreases enough, HF n.u. may go up dramatically while spectral power in the HF band decreases. A change in normalised HF could be the result of increased HF power, decreased LF power, or a combination. Normalised units “should always be quoted with absolute values of LF and HF power in order to describe in total the distribution of power in spectral components” (Task Force, 1996, p. 358). Thus a complete description also requires the reporting of absolute units — a recommendation of the Task Force often neglected in HRV research. No other proportional units such as normalised LF power or LF/HF ratio need be reported, as they are equivalent to HF n.u. (Burr, 2007).

The primary goals of the current study were to examine the effect of Vipassana meditation on psychological measures related positively and negatively to well-being, and to extend research on the cardiovascular effects of mindfulness meditation to the techniques of Vipassana meditation. The Vipassana retreat studied here is of particular interest because of the intensity of training — maximising the opportunity to observe resting baseline and acute meditation-related associations with HRV. Vipassana meditation does not focus on controlled breathing or chanting, avoiding certain complications of the relationship between HRV and respiration. Finally, the Vipassana retreat is taught in a standardised manner, enabling easier replication of findings. We also sought to examine the effects of Vipassana training on psychological measures of well-being.

Because of the difficulties associated with direct interpretation of normalised HRV frequency measures, this study focused on measures of absolute HF and LF powers, as well as the normalised measures more commonly reported. We included the high frequency bands corresponding with respiratory sinus arrhythmia (RSA; 0.15–0.40 Hz) and the low frequency Traube–Hering–Mayer wave (THM; 0.06 to 0.10 Hz). Both of these specifically capture the expected bands in which change during a meditation task might be provoked due to an alteration in breathing. The THM wave specifically should be sensitive to the slow breathing effects typically observed in meditation, as the RSA contribution interacts with the baroreflex (Berntson et al., 1997). THM was included as a low frequency measure of baroreflex function, which was considered important in order to clarify possible effects in normalised HF. However no a priori hypotheses were put forward for THM in terms of time, task or interaction effects of meditation.

While the neural correlates of Vipassana meditation have been examined both in terms of state effects (Hölzel et al., 2007) and correlates of long-term practice (Hölzel et al., 2008), to our knowledge, there is no research available on the cardiovascular effects of Vipassana meditation, either acute or longer-term changes in resting baseline HRV. It is as yet unclear whether intensive Vipassana meditation has similar effects on HRV as other mindfulness techniques.

This primary goal of this experiment was to determine the effects of 10 days of intensive Vipassana meditation training on measures of well-being including frequency measures of heart rate variability. It was expected that intensive Vipassana meditation training would result in distinct changes in HF measures of HRV, both at the acute level with changes from resting baseline to the meditation task, and longer-term changes after Vipassana training, compared to before.

• It was hypothesised that there would be differences over time, comparing participants after Vipassana Meditation training to their pre-Vipassana resting baseline measures.

  • Well-being was expected to improve after meditation training compared to before. Increases were expected in measures of well-being and decreases were predicted in ill-being.

  • Before meditation training, positive measures of well-being were expected to correlate positively with resting baseline frequency-domain (ln HF, HF n.u., THM) HRV measures. The reverse pattern was expected for ill-being measures.

  • High frequency (ln HF and HF n.u.) HRV measures were expected to increase post- compared to pre-Vipassana, representing main effects of time.

• Both before and after Vipassana training, HF HRV measures were expected to increase in the meditation task compared to the resting baseline, representing a main effect of task.

• These HRV main effects of time (pre vs. post meditation training) and task (resting baseline vs. meditation task) were further expected to interact. The differences between the meditation task and the resting baseline were expected to be more pronounced after Vipassana training, compared to before.