A Scientific Review of Neuroscience Research on Meditation

Author: Ranganatha Sitaram

---

Introduction

Results from neuroscience studies of meditation are accumulating. However, these studies have not been well replicated. The results, however, have been used by clinicians, scientists and practitioners to inform their own work. Many investigators are themselves practitioners of meditation, and hence there may be a strong positive bias in the practitioners to the results of meditation (Forget et al 2014, metanalysis).

Current studies of meditation have many methodological problems: 1. They are mainly only cross-sectional studies with very few of them being longitudinal studies that track the changes in neurophysiology with increased practice, 2. Very few are well-controlled experiments, 3. The sample size is usually small, 4. Most studies are not beginning with proper theoretical bases and experimental hypotheses but produce post-hoc interpretations and tentative conclusions.

Cross sectional and longitudinal studies

Meditation studies in the early periods of research tended to be cross-sectional in design which compared differences in structure and function of the brain of the meditators, and their behavior between groups of meditators with hundreds and thousands of hours of practice (experts) with controls who could be beginners or other types of meditators, of a different sect or school, at a particular point in time. Such studies reveal differences in the brain, behavior and physiology due to long-term meditation practice. However, cross-sectional studies may fail to see the actual effects of meditation as the effects that are found could be due to pre-existing differences such as, interests, personalities and other traits of the meditators.

In other words, cross-sectional studies do not show the ‘causal effects’ of meditation practice and may only bring out the correlated changes, which may consist of both pre-existing and practice-related effects.

Longitudinal studies evaluate changes in the body and mind at differnet points in time spanning hours, days, weeks and months, and comparing between different groups. Such studies are rate, but the few that have been reported have shown differences in brain function and structure, behavior and physiological measures due to shorter (few days) and slightly longer (1-3 months) periods of mediation training especially when the experiments have been properly controlled.

Evidence for differences between expert and novice meditators are emerging due to a combination of cross-sectional and longitudinal studies. It would be highly ‘contributional’ and interesting to the field to conduct very long-term, longitudinal studies where beginners or experts are tracked for much longer periods of time stretching to years. However, practical and economical limitations may preclude studies and make them infeasible for scientific investigation.

Control conditions

To extract and tease out the specific effects of meditation, researchers have to control for confounding factors or effects that are associated with meditation instruction, training and practice, but perhaps not directly related to meditation. Examples of these confounding factors are: expectancy effects of meditation (beneficial or otherwise), intentional effects, influence of interaction with teachers and other practitioners, posture and physical exercise effects, diet and life style changes etc. If the study evaluates specific effects, then the other effects need to be maintained equal between the experimental and control groups so that statistical analysis can subtract out or eliminate those balanced effects leaving behind the specific effects of meditation practice under investigation.

To give an example, one study used ‘sham meditation practice’ by incorrectly instructing subjects (beginners) to perform meditation but still maintaining other factors such as expectancy, intention, practice length, teacher interaction, posture etc. the same. Mechanistic studies that aim to understand the specific brain and behavioral practice, especially, need to control for confounding conditions precisely to evaluate the changes in the target variable.

Changes in functional activation of the brain (brain states) during and due to meditation have been investigated. The challenge of conducting such studies lies in the difficulty in funding control tasks and conditions where the target meditative state or functional activation is not known. Some studies have used resting condition as the control condition. However, the rest condition is problematic in that expert meditators may slide into meditative states due to long practice or expertise in the rest conditions too making the control condition weak or even invalid.

Changes in brain structure due to meditation

Several studies have looked at structural changes in the gray and white matter of the brain in different traditions of meditation. Studies used varied measurement techniques and measures, including cortical thickness, gray matter volume and density, fractional anisotropy, and axial and radial diffusivity. Most studies have been cross-sectional in design, comparing between expert meditators and novices, while only a few studies have also incorporated longitudinal designs showing changes in beginners learning and practicing meditation. There have also been studies which tried to investigate the effect of meditation on the brain and behavior related to stress reduction, emotion regulation and/or wellbeing. There is a large diversity in the design of these studies, imaging modality and its specifics used, mediation tradition, and measures used. Given the large diversity in these studies, many brain areas and structural connections have been identified in the above studies. These are not surprising results.

The brain areas found altered in most studies so far are:

1. Sensory cortices and insula related to body awareness

2. The frontopolar cortex related to meta awareness

3. The hippocampus related to memory processes

4. Anterior cingulate cortex, mid cingulate cortex (orbitofrontal cortex) related to the self and emotion regulation.

5. Superior longitudinal fasciculus and corpus callosum involved in inter- and intra-hemispheric communication.

These results indicate the early attempts to integrate, consolidate and make sense of the disparate experimental findings. However, our understanding of how these brain regions and connections are related to behavioral and clinical changes that accompany meditation practice are still very tentative.

A theoretical framework to tie together experimental evidences and be able to explain and make predictions of the effects of different practices and traditions of meditation on brain function, structure and behavior is necessary.

Mindfulness and attention

Meditation practices require the engagement, monitoring and sustained control of attention. Also, long periods of meditation has also been known and reported to improve attention in practitioners.

Attention consists of 3 components: 1. Alerting component, and 2. Orienting component, and 3. Conflict monitoring.

Alterting component represents the readiness in preparation for any impending stimulus. This consists of tonic effects due to spending time on a task and phasic effects due to functional changes in the brain due to changes in the stimulus.

Orienting component indicates the selection of specific information from multiple sensory stimuli.

Conflict monitoring represents monitoring and resolution of conflict between processes related to different stored formation, also referred to as executive attention.

Other distinctions have also been described in literature between Sustained Attention and Selective Attention. Sustained Attention is the state of vigilance during long tasks, involving both tonic alerting alerting and orienting. Selective Attention involves either orienting to a stimulus or internally stored information and its processing.

Performance of an individual on any or all of the above different attentional functions is tested by the Attention Network Test (ANT). The ANT allows to measure the performance of alerting and orienting aspects of attention.

Alerting is based in the brain’s noradrenaline system, which originates in the Locus Coeruleus. Orienting involves frontal and areas, including the frontal eye fields and inferior and superior parietal lobes. The Executive Network which takes care of conflict resolution involves the anterior cingulate cortex (ACC), anterior insula and basal ganglia.

ANT and other experimental paradigms have been used to investigate effects of meditative attention. Following is a summary of the findings so far:

1. Improved conflict monitoring due to 5 days (20 minutes each) of Integrative Body-Mind Training (IBMT) based on a longitudinal study (Tang et al 2007).

2. Reduced attentional blink due to mindfulness meditation based on a cross-sectional study (Slagter et al 2007).

3. Improved performance in conflict monitoring in experienced meditators (Van den Hurk 2010).

4. Mindfulness based Stress Reduction (MBSR) technique has not found to improve conflict monitoring (Anderson 2007, Jha 2007).

5. It may be argued that the negative results are due to the short periods of meditation that is being studied, and alternatively, this type of meditation may not influence conflict monitoring. While this may be the case, studies in long-term meditators (months to years) did find effects in alerting aspects of attention (Tang 2012, Jha 2007, McLean 2010, Pagoni 2007).

6. Enhanced alerting and orienting to a visual target was observed in a cross-sectional study in meditators who trained for relatively longer periods (3 months) in Shamatha meditation (McLean 2010).

7. However, in another study 8 weeks of MBSR training did not help in sustained attention although some improvement in orienting was observed (McCoon 2014).

The above results indicate that in the current state of research we do not know whether the differences in findings of studies are due to different meditation types, duration of practice, expertise and the type of task being investigated. A systematic review of meditation (Chiesa 2011) shows that, early phase of mindfulness meditation may be influencing conflict monitoring and orienting of attention, and later phase in improved alerting.

Neural mechanisms of enhanced attention

Mindfulness meditation has been shown to enhance activations in the ACC. This brain region is known for its involvement in attention and its control in the past and other current neuroscience research. ACC and insula form part of a network of brain regions (through long-range connections) that switch and regulate attention to resolve conflicting information processing in various cognitive acts. Cross-sectional studies on meditation have reported increased levels of activation in the ACC in meditators in comparison to controls during focused attention meditation and mindfulness meditation during the application of painful stimuli (Cahn & Polich, 2006, Gard 2014). In another longitudinal study involving control subjects and conditions (say what they are) higher activation levels of ventral and rostral ACC were observed in the resting state after 5 days of meditation practice (Tang 2009). Another study reveals the subtle differences in early and later levels of expertise: shorter periods of meditation practice lead to greater activation in ACC that decreases with greater expertise (Brefczynski-Lewis 2007). Meditation practice has also been shown to increase greater thickness of the neocortex and white connectivity of the ACC (Grant 2010, Grant 2013, Tang 2012) from and to other regions in the attentional network that have been shown to be affected by meditation are: PFC, which increases during executive processing in meditators, and parietal cortex.

Despite these results showing changes in the brain regions related to attention processing, it is still not clear whether the changes are due to the improved attentional performance. Studies which measure structural and functional changes in the brain with frequent, longitudinal measures of attentional control need to be performed. If a close relationship between these variables are found, replicated and understood, then the application of mindfulness to psychiatric disorders would become highly promising.

 

Mindfulness and emotion regulation

Mindfulness meditation has been shown to some extent to improve emotion regulation. Emotion regulation involves the control of how different emotions arise, how often and for how long, how these emotions are experienced by the individual, and how they are expressed on an ongoing basis. A number of processes underlie emotion regulation, namely:

1. Attention deployment: attention to thoughts, 2. Cognitive change: changes in the habitual patterns of appraisal of one’s own emotions, 3. Response modulation: modifying and reducing the habitual suppression of response to emotions (Refer to Gross 2014, for a review on the concepts of Emotion Regulation).

Effects of mindfulness meditation on emotion regulation

Experimental investigation on the effects of mindfulness meditation have used self reports from meditators, measures of peripheral physiology and neuroimaging. The following effects have been found:

1. Reduction in emotional interference by unpleasant stimuli (Ortner 2007),

2. Decreased physiological reactivity and quicker return to the baseline condition compared to controls, after watching a stress movie (Goleman 1976).

3. Reduced difficulties in emotion regulation (Robins 2012).

4. Lower intensity and frequency of negative affect (Chambers 2008, Ding 2014).

5. Improved positive mood states (reference).

Neural mechanisms of improved emotion regulation after mindfulness meditation

Most neuroimaging studies of mindfulness meditation so far have presented participants words, sounds and pictures of emotional content to test how participants and their brains react to when confronted by these stimuli (Hoetzel 2011). The predominant hypotheses in the studies can be divided into “top-down” control and “bottom-up” processing. Top-down control refers to the influence of prefrontal cortical regions on the emotional and limbic regions, such as the amygdal and insula. Bottom-up processing indicates the influence of primary and higher sensory activations having a greater role to play in perceptual and cognitive outcomes.

A major sub-hypothesis is also that mindfulness meditation leads to strengthening of prefrontal cognitive control mechanisms that down-regulate or constrain activation in emotion processing area such as amygdala. This in turn is assumed to lead to lesser emotional reactivity or behavioral outcomes that are closer to the baseline state and not the overt emotional responses that one might observe otherwise.

Mindfulness meditation is hypothesized to result in the developed habit of being the present moment, in the sense that the individual is less preoccupied with thoughts that are not representative of events in the outside world, but actually very aware of these sensory and bodily occurrences. Further, mindfulness meditation is also hypothesized to lead to a cognitive trait of acceptance of the inside (bodily) and outside (environmental) events without categorizing, judging, stereotyping and evaluating those events in ordinary, daily value based manner.

The above two aspects, namely, present moment awareness (or the now) and the non-judgmental acceptance developed through the practice of mindfulness meditation are considered to be crucial in cognitive control, because they increase sensitivity to affective cues that in turn may signal or send message for activating control.

Current research suggests that the existence of top-down and bottom-up control their effects on behavior differ between experts and beginners. While studies have looked at the involvement of the functional activation of the fronto-limbic structures in mindfulness meditation, the relation to changes in the measures of behavior or well-being related to the above hypotheses are not investigated that much yet (Hoetzel 2010). Many studies have reported findings that mindfulness meditation is associated with reduced activation of amygdala (a brain region critically involved in emotion processing and fear conditioning) during viewing of emotionally charged images or during rest (Goldin 2010, Lutz 2014, Taylor 2011). However, these effects are more often found in beginners than in experienced practitioners (Taylor 2011). Prefrontal activations (in the Dorsolateral Prefrontal Cortex – DLPFC) were observed to increase in beginners in a stroop task, which relates to executive processing, after mindfulness meditation practice of 6 weeks than before the beginning of the practice (Allen 2012). This study was one of the first studies in meditation research that used active controls. In another example of this type of effect in beginners, enhanced activations in Dorsomedial and Dorsolateral Prefrontal Cortices (DMPFC and DLPFC, respectively) when participants were presented emotionally negative images while actively performing or being in a mindful state (Lutz 2014). In another interesting study on Mindfulness Based Stress Reduction (MBSR) type of meditation, patients of anxiety who completed the course showed increased activation of the ventral prefrontal cortex when they labeled the emotions on the images displayed to them (Hoetzel 2013).

In complete contradiction, experienced meditators have been seen to show reduced activation in the medial prefrontal cortex during mindfulness meditation (Taylor 2011). It has been proposed that the reduced prefrontal cortical activation could be interpreted as reduced cognitive control in terms of appraisal and elaborative thoughts during viewing emotionally laden stimuli and also greater acceptance of the present state of the mind, body and stimuli.

Similarly, neuroimaging studies of pain processing and reduction in pain after mindfulness meditation found contradicting patterns between beginners and experts of meditation. While beginners showed heightened activation of prefrontal and pain processing regions such as the anterior cingulate cortex and insula, the expert meditators showed reduced activations in the prefrontal regions. The beginners, on the other hand, showed relatively lower activations in the primary sensory processing areas of pain, while experts showed increased activation in these regions including in the insula, somatosensory cortex and the thalamus (Gard 2012, Pogoni 2007).

Based on the above results, the current theoretical stand is that mindfulness meditation affects beginners quite differently from experts. It is believed that inn the beginners, mindfulness facilitates active cognitive control or regulation when they confront emotional stimuli to overcome habitual ways of responding or reacting to one’s emotions. This might be the reason for a greater activation in the prefrontal cortex, a region involved in cognitive control and regulation. In contrast, experts do not actively regulate their emotions like the beginners, but rather take an accepting and observing stand, which is non-judgmental in nature. This might explain the lower prefrontal cortical activation and a greater activation in the primary sensory areas. Thus the beginners and expert meditators may represent the top-down and bottom-up ways of responding neutrally and behaviorally to emotional and other stimuli (Chiesa 2013).

Hence, it is now considered that in the early stages of training to meditate there might be a greater cognitive effort and need for attentional control. This view is supported by the evidence that in beginners lateral prefrontal and parietal cortices are more active than in the experts (Tang 2012, Pagoni & Cekic 2007, Farb 2007, Jensen 2012). In the advanced stages of meditation, attention control may become an acquired skill. In this stage, activations in the prefrontal and parietal parts of the brain is much reduced, while the anterior cingulate, striatum and insula activity remains. The role of cognitive effort and attentional control in meditation needs to be further looked at by comparing between begineers and experts.

Functional connectivity of the brain during meditation

A cross-sectional study of pain processing in meditators showed decreased functional connectivity between prefrontal cortex and pain processing regions of the brain, such as anterior cingulate cortex, primary somatosensory cortices (S1 and S2), and insula (Grant 2010). This result suggests a functional decoupling between the executive network and the pain network. In another study of a group of a cigarettes who performed mindfulness meditation as beginners, when mindfulness state was compared with a normal state of viewing of craving cues, reduced connectivity between craving related brain regions was observed (Westbrook 2013).

A randomized, controlled, longitudinal study in anxiety patients observed changes in a previously existing negative correlation between prefrontal regions and amygdala to a positive correlation after mindfulness meditation training (Hoetzel 2013). Negative correlation like the above occur when the prefrontal regions down-regulate or reduce activation of amygdala through a top-down modulation (Jensen 2012, Kirk 2014), which could have been the way the patients coped with their symptoms/problems before meditation training. However, after meditation training the functional connectivity changed to a positive correlation. This could be interpreted as a change in neural processing and communication due to monitoring of emotional arousal and without down-regulation or suppression of emotional responses. Tong and colleagues suggested that (2015) this change in functional connectivity could be a unique signature of mindfulness meditation. A clear improvement in anxiety symptoms has also been seen after mindfulness intervention (Hoetzel 2013).

 

(to be continued)