The Neuroanatomical Correlates of Mindfulness Meditation: A Literature Review

(1)

The

Neuroanatomical

Correlates of

Mindfulness

Meditation: A

Literature Review

Literature Thesis

L.K.M. Han (6175937) Supervisor: dr. O. Colizoli Co-assessor: dr. H.A. Slagter

MSc in Brain and Cognitive Sciences, University of Amsterdam Cognitive Science track

Date: 28-07-2014 Words: ± 9100

(2)

The Neuroanatomical Correlates of Mindfulness Meditation: A Literature Review

2

ABSTRACT

Over the last three decades, mindfulness meditation (MM) practices have gained an ever-increasing popularity in the Western world. This growing popularity is a result of accumulating scientific evidence suggesting that practicing MM has beneficial effects on general, psychological, and physical well being, in addition to improving cognitive abilities. The neurobiological mechanisms underlying these beneficial effects remain to be elucidated. However, a growing body of literature has demonstrated that there are regional neuroanatomical differences in mindfulness meditators with respect to non-meditators. It is corroborated that MM is able to actively alter neuroanatomical structure, specifically in brain regions that are involved in bodily attention and visceral awareness, emotion and cognition, attention, respiratory and cardiac control, and pain regulation. Nevertheless, studies showed large differences in conceptualization and operationalization of MM. In addition, previous studies largely identified different neuroanatomical regions that showed higher values of brain structure measurements in meditators, with respect to non-meditators. The current issues of theory and methodology make it difficult to directly compare findings across studies. Generally, it is suggested that mental practice in the context of MM, rather than components only specific to MM, can actively alter neuroanatomical structure. Mental practice includes bodily awareness, self-perception, emotional-, attentional-, and cognitive regulation.

(3)

3

1. Introduction

Over the last three decades, meditation practices have gained an ever-increasing popularity in the Western world, specifically the subgroup of meditative practices known as Mindfulness Meditation (MM). Besides MM, there are Mindfulness-Based Cognitive Therapies (MBCT), Mindfulness-Based Interventions (MBI) such as Mindfulness-Based Stress Reduction (MBSR), centers for mindful eating, and increasingly more companies that encourage their employees to become more “mindful” to improve work productivity (Chiesa & Malinowski, 2011; Brantley, 2005; Kabat-Zinn, 2003; Kelly, 2014; Wolever et al., 2012). While significant differences exist between them, all of the above meditative techniques include MM as an active component of their practice (Chiesa & Serretti, 2010).

The growing popularity of MM in Western culture is a result of accumulating scientific evidence suggesting that practicing MM has beneficial effects on general and psychological well being, in addition to improving cognitive abilities (Brown & Ryan, 2003; Chiesa et al., 2011; Zeidan et al., 2006). MM practices are therefore increasingly used as clinically relevant tools in treating medical illnesses and psychiatric disorders (Arias et al., 2006; Chiesa & Serretti, 2010; Pickut et al., 2013). Generally, it seems that scientific studies are confirming certain benefits that MM practitioners have been advocating for years.

MM has its roots in ancient Buddhist meditation traditions, and its core characteristics include attentional, emotional, and postural self-regulation, as well as the change in perspective on the self and awareness of the sensations of breathing (Hölzel et al., 2011; Nielsen & Kaszniak, 2006; Pagnoni & Cekic, 2007). Subsequently, the goal of MM is to achieve a state that can be described by nonjudgmental and metacognitive monitoring of momentary thoughts, resulting in an open and receptive attitude (Deikman, 1982; Froeliger et al., 2012; Martin, 1997). It is speculated that the above-mentioned characteristics of MM are the mechanisms through which it is able to exert its beneficial effects (Hölzel et al., 2011).

It follows that scientific studies are not only trying to reveal the mechanisms through which MM is able to produce positive effects on well-being from a conceptual perspective, but also from a biological perspective. From a clinical viewpoint, MM based training and practices have shown efficacy in treating psychiatric and physical disorders by reducing relapses of depression as well as reducing blood pressure and alcohol and substance abuse (Chiesa & Serretti, 2010). In addition, MM has also been shown to be beneficial to healthy subjects, by reducing their perceived stress (Hölzel et al., 2010). Likewise, MM practices

(5)

2

produce favorable results in terms of several types of attention, working memory capacity, and other executive functions (Chambers et al., 2008; Jha et al., 2010). It is thus implied that practicing MM helps to overcome several health-related issues, as well as enhance mental capabilities. With regard to the confirmed beneficial clinical outcomes associated with MM, neuroimaging studies have explored the neurobiological processes underlying these mechanisms.

The brain determines behavior, yet it is also altered by the very behaviors it produces. To date, it is well-established that brain structure can be modified by experience, as this lies at the essence of the neural basis of cognition, learning and neuroplasticity (Zatorre et al., 2012). Consequently, changes in brain structure can be directly linked to changes in the way the brain functions (Johansen-Berg, 2010; Kanai & Rees, 2011). Therefore, it is of profound importance to establish correlations between brain structure and MM, in order to understand the neurobiological mechanisms that conceivably precede brain function and behavior. To this end, this current review specifically focuses on the existent literature of MM in relation to brain structure.

At present, there are several available techniques for structural brain imaging. One of the most common technique to experimentally investigate brain anatomy is magnetic resonance imaging (MRI) (Haacke et al., 1999). Magnetic resonance images are composed of several volume elements, so called voxels. The main advantages of MRI are that it is non-invasive and does not involve exposure to radiation. In short, MRI techniques allow us to map brain structures with relatively good sensitivity and resolution.

Furthermore, the structure of the brain can be divided into two major classes, gray matter and white matter. In general, gray matter mostly consists of neuronal cell bodies, and can be measured through T1-weighted MRI images (Purves et al., 2008). There are different morphometric techniques to determine measures of gray matter volume, density or cortical thickness. However, note that multiple gray matter tissue properties such as cell density, cell size, and myelination can influence these measures (Zatorre et al., 2012). A drawback of MRI is thus that it cannot distinguish between different gray matter tissue characteristics.

In contrast, white matter primarily consists of myelinated axon tracts (Purves et al., 2008) and can be measured with diffusion-weighted MRI. With diffusion tensor imaging (DTI) techniques it is possible to estimate parameters such as fractional anisotropy (FA) to quantify the degree of directional sensitivity of water diffusion within a voxel (Basser & Pierpaoli, 1996). In general, FA values are interpreted as an indicator for white matter fiber connectivity and coherence (Luders et al., 2012). As mentioned before, it is of importance to note that individual differences in gray and white matter structure have been linked to brain function and behavior.

(6)

3

A growing body of literature has demonstrated that there are regional neuroanatomical differences in mindfulness meditators with respect to non-meditators (Hölzel et al., 2008; Kang et al., 2013; Luders et al., 2009; Luders et al., 2012; Pagnoni & Cekic, 2007; Vestergaard-poulsen et al., 2009). Some studies even showed a positive correlation between the amount of MM experience and cortical thickness, demonstrating that the most experienced meditators were associated with the thickest cortex (Lazar et al., 2005; Grant et al., 2010; Grant et al., 2013). In addition, a longitudinal study showed significant induced gray matter alterations in subjects participating in a mindfulness-based intervention program (Hölzel et al., 2011). Furthermore, another study showed that the Five Facet Mindfulness Questionnaire (FFMQ), a questionnaire developed to assess individual differences in mindfulness states, was positively associated with gray matter volume in brain regions involved in interoception and cognitive control of emotional responses (Murakami et al., 2012). However, to date it remains to be elucidated if, and how exactly, MM is able to alter neuroanatomical structure.

While the findings of gray matter alterations in participants of MM practices are starting to converge, different brain regions across studies have been linked to MM. Specifically, different brain regions have been found to be greater in mindfulness meditators than controls and altered after a mindfulness based intervention program. The discrepancies between findings are most likely due to both theoretical as well as methodological issues. This literature review therefore aims to compare and contrast several studies investigating the neuroanatomical correlates of MM by evaluating the issues of theory and methodology and discussing the limitations of correlational research adopted by MRI-based morphological studies.

It will be argued that studies conducted so far on MM lack interpretations of what neuroanatomical alterations in mindfulness meditators actually signify. Other studies on expert taxi drivers, musicians and jugglers supported an experience-dependent explanation of gray matter alterations (Draganski et al., 2004; Maguire et al., 2000; Bermudez et al., 2009). However, these studies revealed anatomical group differences that reflected motor skill, training or expertise, rather than mental practice in the context of MM. Nevertheless, mindfulness meditators dedicate their life to regular, frequent and ongoing cognitive efforts (Luders et al., 2011). It seems conceivable that expert mindfulness meditators support an experience-dependent explanation of neuroanatomical alterations as well. The second aim of this review is therefore to discuss the potential physiological nature of the different underlying correlates and consider which cognitive phenomena are thought to drive these changes.

In the end it will be argued that the current issues in theory and methodology make it difficult to directly compare findings across studies. In addition, well designed longitudinal

(7)

4

studies are lacking at present. Therefore, it remains unclear whether components only specific to MM are able to actively alter neuroanatomical structure, or that alterations are an effect of non-specific MM components. Instead, it is suggested that mental practice in the context of MM demonstrates experience-dependent neuroanatomical alterations, and that mindfulness meditators are good human models for investigating structural plasticity. In conclusion, current evidence suggests that more general cognitive processes such as bodily awareness, self-perception, emotional-, cognitive-, and attentional regulation, rather than MM in specific is associated with neuroanatomical changes.

(8)

5

2. The concept and operationalization of Mindfulness Meditation

The concept of mindfulness originates from many ancient Buddhist and Hindu meditation traditions, but its current practice does not necessarily imply any specific involvement of philosophical or religious tradition (Kabat-Zinn et al., 2000). As mentioned before, there are multiple meditative techniques that include MM as an active component. “Mindfulness” is the skill or capacity to increase awareness of the present moment, but allows us to be less reactive (Germer, 2004). When engaged in MM, one focuses attention and is not entangled in the past or future, and, can therefore experience the present moment with non-judgmental openness. The world authority on mindfulness, Jon Kabat-Zinn (1994), describes mindfulness in his book “wherever you go, there you are: mindfulness meditation in everyday life” as: “Paying attention in a particular way: on purpose, in the present moment, and non-judgmentally.”

The first MRI-based study that investigated MM in relation to gray matter structure, conceptualized MM as the cultivation of attention, without cognitive elaboration, in Buddhist Insight meditation practitioners (Lazar et al., 2005). Other studies conceptualized MM in a similar fashion, investigating comparable Buddhist MM practices (Pagnoni & Cekic, 2007; Hölzel et al., 2008; Vestergaard-Poulsen et al., 2009; Luders et al., 2009; Luders et al., 2012; Grant et al., 2013). In contrast, for example, Grant et al. (2010) conceptualized MM of Zen practitioners particularly in the context of the acceptance of pain, suggesting MM can alter sensitivity on both the affective and sensory dimensions of pain. While conceptualization of MM in scientific research remains to be diverse and inconsistent, there is general consensus concerning the involvement of sustained attention to the present moment (Chiesa & Malinowsky, 2011).

Furthermore, there are several studies that have investigated MM so as to reveal its neuroanatomical correlates. However, researchers investigating the neural substrates of MM largely examine different types of MM practices, as can be seen in Table 1. Most studies include more traditional ancient meditations, such as Zen and Zazen (Kapleau, 1965), Vipassana (Gunaratana, 1993), Samatha (Lamrimpa, 1992), and other Buddhist meditations, whereas one longitudinal study investigated a modern standardized 8-week mindfulness-based intervention, MBSR (Kabat-Zinn, 1990). In addition, Murakami and colleagues (2012) related FFMQ questionnaire scores (an instrument developed to measure mindfulness) to brain structure. The FFMQ questionnaire is composed of five facets: non-reactivity to inner experience, non-judging, acting with awareness, describing, and observing, and shows adequate to good consistency (Baer et al., 2008). Furthermore, the most recent study by Kang et al. (2013) investigated MM by means of Brain Wave Vibration

(9)

6

(BWV) training, an eclectic form of yoga and meditation (Bowden et al., 2011). It becomes apparent that there is substantial heterogeneity of practices under study in the attempt to uncover the neuroanatomical correlates of MM.

Although studies are inconsistent in their operationalization of MM, all studies claimed that their meditative technique had the capacity to evoke mindfulness, and therefore included the active beneficial component of MM. Consequently, it is speculated that this component potentially underlies the structural differences between meditators vs. non-meditators. The most common factor between the operationalization of the discussed studies is thus the capacity to evoke mindfulness, with attention playing a key role.

Table 1

Overview of operationalization of Mindfulness Meditation

MBSR, Mindfulness-Based Stress Reduction; FFMQ, Five Facet Mindfulness Questionnaire; BWV, Brain Wave Vibration. * Indicates longitudinal study.

Study Operationalization of Mindfulness Meditation

Lazar et al. (2005) Insight

Pagnoni & Cekic (2007) Zen

Hölzel et al. (2008) Insight

Vestergaard-Poulsen et al. (2009) Tibetan Buddhist

Luders et al. (2009) Zazen, Samatha, Vipassana and others

Grant et al. (2010) Zen

Luders et al. (2011) Samatha, Vipassana, Zazen

Hölzel et al. (2011)* 8-week MBSR intervention

Luders et al. (2012) Chenrezig, Kriya, Samatha, Vajrayana, Vipasanna, Zazen

Murakami et al. (2012) FFMQ

Grant et al. (2013) Zen

(10)

7

3. The neuroanatomical correlates of Mindfulness Meditation

In general, MM is studied quite extensively, however, the existent literature on the associations of MM and neuroanatomical structure is relatively sparse. More specifically, while there are several studies that have investigated MM in relation to gray matter structure, there are only a few studies that have done so in relation to white matter structure. To date, the structure of the mindful brain remains to be elucidated. An overview of the current studies and associated neuroanatomical structures can be seen in Table 2.

Existent findings demonstrated that the associations of MM and brain anatomy are widely distributed across the entire brain, encompassing both cortical, as well as subcortical regions. In addition, neuroanatomical structures within the brain stem and cerebellum have also been associated with MM. Findings are somewhat starting to converge, however, studies also largely identified different brain regions showing higher values of brain structure measurements in mindfulness meditators with respect to non-meditators, especially in the frontal brain regions. Interestingly, findings were mostly unidirectional, with positive correlations of MM and brain structure. However, there was one study that identified gray and white matter structures to be greater, and more enhanced, in controls than in meditators (Kang et al., 2013). When trying to isolate consensual findings across studies, the frontal cortex, insula, and cingulate cortex seem to be relatively common structures associated with MM.

(11)

8 Table 2

Overview of brain structures associated with Mindfulness Meditation

Study Gray matter associations Related function or

involvement Lazar et al. (2005) Right middle and superior frontal

sulci, right anterior insula

Integration of emotion and cognition, bodily attention and visceral awareness

Pagnoni & Cekic (2007) No age related decline in left putamen

Cognitive flexibility, attentional processing

Hölzel et al. (2008) Left interior temporal lobe, right hippocampus, right insula

Experience of insight, attentional and emotional processes, interoception and visceral awareness

Luders et al. (2009) Right orbito-frontal cortex, left inferior temporal lobe, right hippocampus, right thalamus,

Emotional self-regulation and behavioral flexibility,

experience of insight, attentional and emotional processes, regulator of sensory information

Vestergaard-Poulsen et al.

(2009) Left superior and inferior frontal gyri, left fusiform gyrus, medulla oblongata, anterior lobe of the cerebellum

Self-relation and active memory retrieval, face and body

recognition, respiratory and cardiac control, unconscious proprioception

Grant et al. (2010) Dorsal anterior cingulate cortex, primary and secondary

somatosensory cortex

Control of emotional experience, pain regulation Hölzel et al. (2011) Posterior cingulate cortex,

temporo-parietal junction, cerebellum, left hippocampus

Integration of self-referential stimuli, conscious experience of self and social cognition, regulation of emotion and cognition, modulation cortical arousal and responsiveness Murakami et al. (2012) Right amygdala, right anterior

insula Cognitive control of emotional responses, interoception Grant et al. (2013) Left superior frontal gyrus, middle

frontal gyrus, anterior cingulate cortex, supramarginal gyrus, superior parietal lobe

Self-awareness, attention and executive control

(12)

9 * Indicates gray matter increases/white matter enhancement in controls vs. mindfulness meditators.

3.1. Gray matter associations

The first study to associate MM with gray matter structure was conducted by Lazar and colleagues in 2005. A decade later, about ten other studies have investigated this relationship further. As you can see in Table 2, gray matter associations can be roughly divided into a few different functional categories, namely, bodily attention and visceral awareness, emotion and cognition, attention, respiratory and cardiac control, and pain regulation. From a behavioral perspective, the involvement of these brain regions seems to correspond with the goals and outcomes of MM training, that being, “Paying attention in certain way: on purpose, in the present moment, and non judgmentally” (Kabat-Zinn, 1994). In addition, MM is sometimes performed while sitting in an uncomfortable position

Kang et al. (2013) Medial prefrontal cortex, superior frontal cortex, temporal pole, middle and inferior temporal cortices, bilateral postcentral cortex*, inferior parietal cortices*, left posterior cingulate cortex* middle occipital cortex*

Emotional regulation, regulation and monitoring of attention, emotional processing, self-referential processing, egocentric processing

White matter associations Luders et al. (2011) Corticospinal tract, temporal

component of superior longitudinal fasciculus, uncinate fasciculus

Major projection pathways: travels through medulla involved in respiratory and cardiac control, commissural pathways: travels through caudal areas of temporal lobe perhaps involved in language, association pathways: connecting orbital cortex to amygdala and hippocampal gyrus

Luders et al. (2012) Callosal regions Interhemispheric connectivity,

integration of information Kang et al. (2013) Medial prefrontal cortex, medial

prefrontal cortex*, posterior cingulate cortex*, occipital cortex*

Emotional processing, self-referential processing, egocentric processing,

attentional regulation, executive attention

(13)

10

and/or while actively trying to remain in a certain posture, potentially explaining the involvement of brain regions associated with pain regulation (Grant et al., 2010).

Most strikingly, almost no gray matter structures were identified to be significantly increased in controls than in meditators, except in one recent study by Kang and colleagues (2013). They demonstrated significantly thinner cortical thickness in the posterior regions of the brain, including the postcentral cortex, inferior parietal cortex, middle occipital cortex and posterior cingulate cortex (PCC). This is specifically inconsistent with the finding of Hölzel et al. (2011) that found gray matter concentration in the PCC to be greater after an 8-week MBSR intervention. Kang et al. (2013) argued that thinner cortical thickness in the PCC may have been a result of non-linear training-induced changes, in which an initial increase in gray matter volumes can be followed by a later decrease, as previously demonstrated by both Driemeyer and colleagues (2008), and Boyke and colleagues (2008). The findings of both studies may potentially reflect this exact mechanism.

Another interesting finding is that mindfulness meditators, compared to non-meditators, showed less age-related cerebral volume loss, as demonstrated by Lazar and colleagues (2005) and Pagnoni & Cekic (2007). It is suggested that commitment to, and regular practice of MM, may counteract the rate of age-related neural degeneration in cortical areas, specifically in the prefrontal cortex. This is of particular interest, because age-related gray matter volume decline is normally most substantial in the prefrontal cortex, contributing to age-related cognitive decline (Raz et al., 1997). Engaging in MM practices can therefore potentially offset cognitive decline.

Furthermore, in an attempt to isolate similar findings, three studies have demonstrated a positive relationship between MM and gray matter within the right anterior insula (Lazar et al., 2005; Hölzel et al., 2008; Murakami et al., 2012). The right anterior insula has consistently shown to be involved in interoception and visceral awareness (Critchley et al., 2004). The associations of MM with the right anterior insula seem to be in line with the concept of MM. That being, focusing attention to the self, in order to experience the present moment with non-judgmental openness. Moreover, MM constitutes training in interoception and conscious awareness (Hölzel et al., 2011).

Furthermore, the cingulate cortex seemed to be linked to MM across different studies. However, the studies by Grant and colleagues (2010, 2013) found positive associations with the anterior cingulate cortex, whereas Kang and colleagues (2013) identified the posterior part of the cingulate cortex to be negatively associated with MM. Under the hypothesis that repeated activation, or deactivation, of particular brain areas could lead to durable neuroanatomical alterations, the latter finding fits well with a recent finding of Garrison and colleagues (2013) that showed deactivation of the PCC in experienced meditators during meditation. The anterior cingulate cortex is involved in self

(14)

11

and emotion regulation, and these functions also seem to be in line with the concept and goals of MM.

Findings remain inconsistent concerning the specific subareas of the frontal regions of the brain that have been associated with MM, yet associations with the frontal cortex seem to be evident. It can be speculated that involvement of frontal brain regions associated with MM can be explained from a more metacognitive and attentional perspective. Metacognition is the knowledge and regulation of cognition, and MM can be considered a metacognitive skill given that its main goal is attentional self-regulation (Bishop et a., 2004). Previous studies have indicated frontal and prefrontal contributions during metacognitive processes (Fleming et al., 2012). In addition, both the cingulate cortex and insula have been involved in metacognition as well (Fleming & Dolan, 2012). Moreover, prefrontal regions share reciprocal anatomical connections with the cingulate cortex and insula (Medalla & Barbas, 2010). Fleming & Dolan (2012) suggested that the prefrontal regions receive input from the cingulate and insula in order to monitor and control cognition. It can be seen in Table 2 that (pre)frontal brain regions and the insula and cingulate cortex have also been linked to MM. Therefore, it can be speculated that metacognitive processes during MM, rather than MM itself, are associated with these brain structures.

Although elaborate discussion of functional MRI (fMRI) is beyond the scope of this review, it is of importance to note some of the findings of these studies. First, there are only a few studies that have investigated brain function in response to MM with the use of fMRI. A study by Baerentsen and colleagues (2001) showed an increase of brain activation in the dorsolateral prefrontal cortex and anterior cingulate cortex during meditation vs. rest in mindfulness meditators, whereas a signal decrease in the orbitofrontal cortex was apparent. Similarly, Ritskes et al. (2003) also demonstrated an increase in signal in the dorsolateral prefrontal cortex during meditation vs. rest in Zen meditators. On the other hand, Ritskes and colleagues showed a signal increase in the basal ganglia and decreased activation of the anterior cingulate cortex and the right anterior superior gyrus. Another fMRI study by Lazar et al. (2003) only showed an increase in activation in the dorsolateral anterior cingulate cortex and right temporal lobe during meditation vs. a control task in mindfulness meditators and Kundalini yoga practitioners. These discrepant findings are most likely a consequence of the heterogeneity of types of MM practices under study and the different experimental designs and control tasks. While an attempt to isolate consensual findings showed involvement of the dorsolateral prefrontal cortex and anterior cingulate cortex, until now, there is limited insight into which neural network is recruited by the active MM component (Ives-Deliperi et al., 2011). Therefore, brain regions found to be activated during MM, do not consistently overlap with brain regions found to be associated with MM in neuroanatomical studies.

(15)

12

3.2. White matter associations

As mentioned before, there are only a handful of studies that have investigated MM in relation to white matter structure. Luders and colleagues (2011) estimated the FA values for twenty different fiber tracts in mindfulness meditators and well-matched controls. Their results showed enhanced structural connectivity throughout the whole brain in meditators, with the largest group differences observed within the corticospinal tract, temporal component of the superior longitudinal fasciculus, and the uncinate fasciculus. The uncinate fasciculus projects from the orbital cortex to the amygdala and hippocampal gyrus, brain regions that indeed have been previously associated with MM (Hölzel et al., 2008; Hölzel et al., 2011; Luders et al., 2009; Murakami et al., 2012).

Another study by Luders et al. (2012) demonstrated enhanced white matter integrity and thicker callosal regions in mindfulness meditators. These findings indicated increased connectivity between the hemispheres, potentially reflecting an enhanced integration of information between them during cerebral processes that involve (pre)frontal regions. However, to date there is no behavioral data to support the assumption that mindfulness meditators actually benefit from enhanced white matter integrity and thicker callosal regions.

The third publication on white matter associations is a study by Kang et al. (2013). In contrast to the previous studies, their results not only showed higher FA values in meditators, but also reduced FA values, specifically in the medial prefrontal cortex (MPFC), PCC and occipital cortex. According to the authors, the number of crossing fibers in these regions may explain reduced FA values. That is to say, when white matter fibers cross, or when an additional connection in another direction is present, connectivity in the principal direction is less coherent (Wedeen et al., 2008). This is reflected in reduced FA values. A drawback of the DTI technique is therefore that the source of the FA value differences is not specified. Nevertheless, the explanation of reduced FA values due to crossing fibers is in line with the view that the MPFC and PCC perform the role of connector hubs in the brain (Hagmann et al., 2008).

A technique that makes it possible to investigate the connectivity between regions in gray matter is fiber tracking. Fiber tracking is necessary in order to further interpret whether FA changes are caused by crossing fibers, or because of intrinsic changes to the primary fiber. There is a need for future studies to include fiber tracking in their methods. Due to the rare amount of DTI studies, the inconsistent findings, and the lack of longitudinal studies, it remains unclear if and how MM is able to influence white matter structure. Obviously, further studies are needed in order to establish any specific relationships.

(16)

13

4. From correlation to causation

As with all studies investigating brain anatomy with the use of MRI, findings are intrinsically correlational. Yet most of the existent literature suggests that MM is able to act as a plausible causal agent, increasing neuroanatomical structure in meditators, and potentially modulating behavioral outcomes as well (Hölzel et al., 2008; Grant et al., 2010; Grant et al., 2013; Lazar et al., 2005). Additionally, the study by Pagnoni and Cekic (2007) suggested that MM played a protective role against natural age-related cortical thinning in the left putamen. The aforementioned studies corroborate a causal relationship, because experienced long-term mindfulness meditators that have had more hours of meditation practice showed the greatest increase in gray matter, or least age-related cortical thinning, respectively. However, regarding the cross-sectional nature of the studies, findings are not sufficient to infer causality. The interpretation of these findings is limited by the fact that individuals might have pre-existent differences in their propensity or willingness to practice MM.

Besides, Luders and colleagues (2009) were unable to show any significant effects of number of years in meditation experience. Similarly, Vestergaard-Poulsen and colleagues (2009) were also not able to detect a correlation between structural differences and amount of practice in mindfulness meditators. However, the authors of the latter study suggest that this could be due to a ceiling effect in their sample, as their participants were highly experienced meditators with a minimum of 8000 hours of accumulated practice. In other words, potential relationships between gray matter structure and meditation experience might have been overseen in these groups of highly experienced meditators. Nevertheless, this remains speculative, and in order to overcome the major limitation of cross sectional research, longitudinal studies are required to verify the findings of gray matter alterations in mindfulness meditators.

To date, there are no studies that follow mindfulness meditators from the very start of the trajectory of becoming an experienced meditator. Nonetheless, in order to verify the assumptions that MM is able to actively alter neuroanatomical structure, a longitudinal study was conducted that investigated participants before and after an 8-week mindfulness-based intervention (Hölzel et al., 2011). The authors hypothesized that the beneficial effects of mindfulness-based interventions on psychological well-being and amelioration of symptoms of disorders would also be visible on a brain level, resulting in morphological changes after such intervention. The findings of their study showed that taking part in an 8-week MBSR intervention was indeed associated with increases in gray matter concentration in the left hippocampus, posterior cingulate cortex, temporo-parietal junction (TPJ), and the cerebellum, as can be seen in Table 2. According to the authors, these brain regions are

(17)

14

especially involved in emotion regulation, self-referential processing, embodiment, and emotional and cognitive processes, respectively.

It is worth mentioning that other studies have also showed gray matter alterations in response to an 8-week MBSR intervention program. For example, Pickut and colleagues (2013) showed significant gray matter density increases in the caudate nucleus, occipital lobe, thalamus and TPJ, while Hölzel and colleagues (2010) showed that perceived stress was positively correlated with decreases in right basolateral amygdala gray matter density. However, it is of importance to note that the study of Pickut and colleagues (2013) was performed in patients suffering from Parkinson’s disease. Due to abnormalities of gray matter atrophy in Parkinson patients (Beyer et al., 2007), it is difficult to compare these findings to the findings of Hölzel and colleagues (2011). Similarly, the longitudinal study of Hölzel et al. (2010) is beyond direct comparison, because participating subjects reported high levels of stress during the experiment. Perceived stress has previously been associated with gray and white matter differences in specific brain areas, inconsistent with the areas identified by Hölzel et al. (2011) (Li et al., 2014). In other words, the different participants across studies would already show abnormal brain morphometry at baseline and are therefore not directly comparable. Nevertheless, although the limited amount of longitudinal studies does not yet provide strong evidence for a causal relationship of MM and neuroanatomical alterations, they complement cross-sectional outcomes and demonstrated neurobiological changes in response to a MBSR intervention program.

Previous longitudinal studies that were not specific to MM, have shown that various types of training such as aerobic exercise, juggling, and mirror reading can induce task-specific neuroanatomical alterations (Colcombe et al., 2006; Draganski et al., 2004; Ilg et al., 2008). In addition, cross-sectional studies have substantiated that gray matter differences were positively associated with behavioral performance. This suggests that increased gray matter structures correspond to increased functioning in the relevant area. It seems plausible that mental training in the context of MM is also able to produce training-induced neuroanatomical changes.

At the same time, it remains to be substantiated whether training or skill should be associated with increased or reduced neuroanatomical structure because of the complex relationship between structural changes and underlying functionality (Zatorre et al., 2012). For example, it can be speculated that decreases in relevant brain areas show more efficient connectivity to be able to achieve the same behavioral performance in the context of gray matter. Regarding white matter, it might be possible to find reduced FA values, while it potentially reflects increases in axon diameters or secondary fiber populations maturing in an area of fiber crossing. This is speculated to be the case in the study of Kang et al. (2013).

(18)

15

with MM can consequently increase its neuroanatomical structure. However, this assumption is challenged by the fact that some brain regions that are activated during MM do not lead to modifications in cortical structure. Hölzel and colleagues (2011) argued that it is credible that not all brain regions are just as amenable to neuroanatomical alterations than others. To date, it remains unclear why some structural changes are transient, why some brain regions are more readily altered over others, and why neuroanatomical changes in different brain regions require different amounts of practice or time.

Furthermore, it remains to be elucidated what the physiological explanation is behind neuroanatomical changes in response to mental training in the context of MM. The nature of the observed changes can be categorized into three types of neuronal changes, namely, gray matter, white matter, and extra-neuronal changes. Concerning gray matter, explanations of larger volumetric measures are threefold: a) neurogenesis, b) synaptogenesis, and c) changes in neuronal morphology. Concerning white matter, different FA values can reflect the number of axons, axon diameter, packing density of fibers, axon branching, axon trajectories, and myelination. However, FA is an indicator for the coherence of the aforementioned white matter microstructure characteristics, making it difficult to specify the source of FA values. Regarding extra-neuronal changes, it may be possible that increases in glial cell size and/or angiogenesis were observed (Zatorre et al., 2013). The MRI methods employed by the studies under review cannot distinguish between either of these possibilities, making it difficult to reveal the physiological nature of the underlying correlates.

All studies suggested that practicing MM promotes morphological alterations in mindfulness meditators vs. non-meditators. However, the associated cognitive and physical phenomena that are suggested to drive these alterations are quite divergent. For example, Hölzel et al. (2011) argued that a perceptual shift modulates internal representations of the self in mindfulness meditators, as evident by the anatomical changes in the brain network involved in the projection of oneself into another perspective. This was accompanied by changes in gray matter concentration associations in brain regions involved in learning and memory processes, emotion regulation, and self-referential processing.

Conversely, Lazar and colleagues (2005) found structural cortical thickness differences in brain regions involved in attention, and argued that repeatedly directing attention towards sensory stimuli during the practice of MM is the driving force for the experience dependent alterations. Similarly, Pagnoni & Cekic (2007) argued for a driving force of the capacity for sustained attention, and attentional processing, by showing thicker cerebral and subcortical structures involved in attention in mindfulness meditators. Likewise, Grant et al. (2013) also argued for the mechanism of attentional absorption in

(19)

16

order to promote neuroanatomical alterations. It is suggested that attention and attentional processes are cognitive phenomena driving brain anatomy changes.

In contrast, Vestergaard-Poulsen (2009) spoke of a driving force of regulation of breathing, showing lasting effects on respiration control as a trait of meditative practice. In addition, they argued that the repeated breathing exercises drive lower brain stem structural alterations that further project to the entire brain, also influencing cognitive and emotional processes. Grant and colleagues (2010) specifically hypothesized that MM is associated with structural changes in brain areas related to pain regulation, given the observation that the practice of MM is often associated with pain in the knees and ankles, due to the cross-legged posture. However, this hypothesis was not entirely confirmed, since an increase of cortical thickness was also demonstrated in areas of the primary somatosensory cortex that had associations with the hands, while the hand is not subjected to any discomfort or pain during MM.

If we broadly categorize the phenomena that are thought to drive neuroanatomical structure changes, we can divide MM components in the context of mental phenomena, and physical phenomena. The mental phenomena that are thought to play a key role in MM induced alterations are perception of the self, attention, and cognitive and emotional regulation. The physical phenomena that are thought to drive structural changes are regulation of breathing and pain. It is noteworthy that not all phenomena are present in every MM practice. For example, not every MM practice requires its meditators to sit cross-legged, or sit down at all for that matter. These differences should be taken into consideration when comparing findings across studies.

(20)

17

5. Methodology

There are different morphological measures and analysis methods to determine neuroanatomical structure. While most studies have associated MM with gray matter, many of them used different morphological measures, as can be seen in Table 3. Several studies related MM to cortical thickness (Grant et al., 2010; Grant et al., 2013; Kang et al., 2013; Lazar et al., 2005), while others used measures of gray matter volume (Luders et al., 2009; Murakami et al., 2012; Pagnoni & Cekic, 2007; Vestergaard-Poulsen et al., 2009), gray matter density (Hölzel et al., 2008; Vestergaard-Poulsen et al., 2009), and gray matter concentration (Hölzel et al., 2011). Only three studies investigated white matter structure in relation to MM (Kang et al., 2013; Luders et al., 2011; Luders et al., 2012). It should be noted that distinct differences exist between the different morphological measures (Hölzel et al., 2010; Hutton et al., 2009; Winkler et al., 2010).

(21)

18 Table 3

Overview of the morphological measures and analysis method

Study Morphological measures Analysis method

Lazar et al. (2005) Cortical thickness Manual analyses

Pagnoni & Cekic (2007) GMV VBM in SPM5

Hölzel et al. (2008) GMD VBM in SPM2

Vestergaard-Poulsen et al. (2009) GMD & GMV VBM in SPM5

Luders et al. (2009) GMV VBM in SPM5

Grant et al. (2010) Cortical thickness Automated analyses

Hölzel et al. (2011) GMC VBM in SPM5

Luders et al. (2011) WM integrity In-house software

Luders et al. (2012) WM thickness & integrity Manual analyses

Murakami et al. (2012) GMV VBM in SPM5

Grant et al. (2013) Cortical thickness Automated analyses

Kang et al. (2013) Cortical thickness, WM integrity Freesurfer & AFNI GMV, Gray Matter Volume; GMD, Gray matter Density; GMC, Gray Matter Concentration; WM, White matter; VBM, Voxel-Based Morphometry; SPM, Statistical Parametric Mapping; MNI, Montreal Neurological Institute; AFNI, Analysis of Functional NeuroImages

Cortical thickness, gray matter volume, gray matter density and gray matter concentration were all assessed using T1-weighted images. All VBM methods were voxel-based, but measured different features of gray matter. All cortical thickness methods were surface based. Gray matter density and concentration both refer to the gray matter values for unmodulated images, and can therefore be used interchangeably. Gray matter volume refers to modulated images that reflect the influence of regional volumes on the size of gray matter values, but also cortical thickness and/or volume averaging (Shukla, 2004). The exact interpretation of gray matter volume, density and cortical thickness is complicated and can also depend on the preprocessing steps used (Ashburner & Ridgway, 2010). In addition, gray matter density or concentration is not the same as neuronal packing density or other cytoarchitectonic tissue properties. Although available evidence is limited

(22)

19

so far, it seems that the morphological measures of cortical thickness, gray matter volume, density, and concentration do have reasonably high correspondence (Hutton et al., 2009). Nevertheless, some biological differences of the distinct neuroimaging techniques remain unclear.

5.1. Neuroimaging parameters

With structural brain imaging using MRI, local imaging metrics over time are compared across individuals. The resulting brain images can potentially be subject to error, bias or variation, due to the different neuroimaging parameters and analysis steps that can be applied (Zatorre et al., 2012). The first step in generating a structural brain image is to apply a magnetic field. The strength of this magnetic field affects the quality of the images that are being formed (Goldstein & Price, 2004). That is to say, by increasing field strength, sensitivity and resolution will also increase. In order to increase statistical power, spatial smoothing is applied. The goal of smoothing is to cope with anatomical variability between individuals, by improving the signal to noise ratio (Reimold et al., 2006). Generally, the size of a smoothing kernel is selected by matching the width of the kernel to the size of the signal to detect (Rosenfeld & Kak, 1982). Furthermore, because MRI-based analyses rely upon performing thousands of statistical tests, the probability of obtaining false positives is high. Therefore, it is extremely important to apply multiple comparison corrections to MRI data. As can be seen in Table 4, the studies mentioned in this review showed variations regarding the above mentioned neuroimaging parameters.

(23)

20 Table 4

Overview of neuroimaging parameters

T, Tesla; mm, millimeter; -, not mentioned; FDR, False Discovery Rate; FWE, Family Wise Error; RFT, Random Field Theory.

Seven out of the 12 discussed studies used an MRI scanner with field strength of 1.5T, whereas the other five studies were able to use higher field strength of 3T, and therefore obtained a higher image resolution. The study of Lazar et al. (2005) failed to mention the specifics of their used scanner resolution and smoothing kernel. While measurement variability between different field strengths is fairly small, a study by Han et al. (2006) showed that cortical thickness measurements between different field strengths were slightly biased, with thickness values being marginally higher at 3T, than 1.5T.

As for spatial smoothing, the approaches of how to apply smoothing have been an area of discussion. It is known that the width of the smoothing kernel can affect the analyzed data, and therefore the results. When one expects to associate MM with brain

Study Scanner

resolution

Smoothing Kernel Multiple comparison correction

Lazar et al. (2005) - - FDR

Pagnoni & Cekic (2007) 3 T 12 mm -

Hölzel et al. (2008) 1.5 T 12mm FWE

Vestergaard-Poulsen et al.

(2009) 3 T 4 mm & 12 mm FDR

Luders et al. (2009) 1.5 T 14 mm Small volume & FWE

Grant et al. (2010) 3 T 20 mm Small volume & RFT

Hölzel et al. (2011) 1.5 T 8 mm Small volume

Luders et al. (2011) 1.5 T - Bonferroni

Luders et al. (2012) 1.5 T - FDR & Bonferroni

Murakami et al. (2012) 1.5 T 12 mm -

Grant et al. (2013) 3 T 20 mm RFT

Kang et al. (2013) 1.5 T 25 mm Monte-Carlo, FDR,

(24)

21

regions of a small extent, it is more suitable to use a small smoothing kernel, however, when one is looking for larger volumes, smoothing may be used more generously (Weibull et al., 2008). For example, Grant and colleagues (2013) measured cortical thickness by using large regions of interest, such as the medial prefrontal cortex. In cortical thickness analyses, a smoothing kernel of 20 mm seems fit concerning the required sensitivity and specificity. In contrast, Hölzel et al. (2011) were specifically interested in studying small anatomical structures such as the hippocampus and insula; therefore a smoothing kernel of 8 mm seems fit. Overall, previous studies have shown that smoothing schemes can have great impact on the reproducibility of cortical thickness measurements (Han et al., 2006).

The adopted strategies for applying multiple comparison correction are displayed in Table 4. The authors of Pagnoni & Cekic (2007) and Murakami and colleagues (2012) have failed to mention the specifics of the correction they applied, if even done so. Corrections applied by other studies involve false discovery rate, small volume correction, Bonferroni and random field theory. It is beyond the scope of this review to elaborately discuss the specifics of the different strategies, but depending on the data, corrections can be either too liberal or too conservative (Bennett et al., 2009). This should therefore be taken into consideration when adjusting the parameters of the different multiple comparison correction strategies.

5.2. Participants and covariates

When comparing the different studies it becomes apparent that there is great heterogeneity regarding the degree of expertise of the mindfulness meditators, as can be seen in Table 5. Some studies included highly experienced meditators with as much as 24 years of experience (Luders et al., 2009), whilst others also included participants that had a little over three years of meditation experience (Pagnoni & Cekic, 2007). Previous studies investigating training-induced cortical plasticity with juggling, showed that some of the structural changes revert to baseline levels after induction (Driemeyer et al., 2008). This potential transient changes should be taken into consideration when comparing the findings of the different studies. Besides, not only is there great heterogeneity of meditation practices under study between studies, it was also large within some of the studies. For example, Luders et al. (2009) used a sample size of 22 meditators, and those 22 people reported to practice more than four different types of MM.

In addition, no studies reported using any covariates that would be possible confounding variables such as exercise, diet, alcohol and drugs consumption, or smoking behavior. Previous studies have shown that these lifestyle factors can influence brain structure (Colcombe et al., 2006; Gallinat et al., 2006; Kramer et al., 2006; Paul et al., 2008). Besides, it is conceivable that MM practitioners commit to an overall better lifestyle,

(25)

22

including a healthy diet and regular exercise. It is suggested that these possible confounding variables are measured and used as covariates in future analyses in an attempt to isolate the specific effects of MM.

Table 5

Overview of participants and degree of expertise

The degree of expertise is displayed in years, except stated otherwise; -, not mentioned.

Study N (meditators/controls) Degree of expertise (years)

Lazar et al. (2005) 20/15 9.1 ± 7.1

Pagnoni & Cekic (2007) 12/13 > 3

Hölzel et al. (2008) 20/20 8.6 ± 5.0 Vestergaard-Poulsen et al. (2009) 10/10 16.5 ± 5.1 Luders et al. (2009) 22/22 24.18 ± 12.36 Grant et al. (2010) 17/18 14.4 ± 8.39 Hölzel et al. (2011) 16/17 - Luders et al. (2011) 27/27 23.3 ± 12.2 Luders et al. (2012) 30/30 20.2 ± 12.2 Murakami et al. (2012) 19 -

Grant et al. (2013) 18/18 6406 ± 2955 hours

(26)

23

6. Discussion

The concept of MM is to “pay attention in a particular way: on purpose, in the present moment, and non-judgmentally” (Kabat-Zinn, 1990). Over the years, scientific evidence has accumulated supporting beneficial effects of MM on general well being, physical and mental health, and behavioral outcomes. For this reason, researchers have focused on studying the brains of mindfulness meditators, so as to uncover the neurobiological mechanisms associated with these beneficial effects. Moreover, since brain structure conceivably precedes brain function and behavior, scientists have explored the associations of MM and brain structure.

Several studies have linked specific brain structures to MM, and while it is tempting to credit neuroanatomical alterations to specific components of MM, findings may also point to components that are non-specific to MM. The assumption that MM by itself is able to actively alter brain structure must be considered with caution. To date, it remains to be elucidated what the specific effects of MM are on the structure of the human brain.

At the moment, research into the neuroanatomical correlates of MM suffers from both theoretical, as well as methodological inconsistencies. Therefore, findings of relatively sparse studies concerning MM and brain structure cannot be directly compared and should be considered with caution. It is suggested that more general cognitive processes are associated with neuroanatomical alterations in mindfulness meditators, including bodily awareness, self-perception, emotional-, attentional-, and cognitive regulation in the context of MM. There is crucial need for future research to elucidate which components truly specific to MM are able to actively alter brain structure in relevant brain areas.

6.1. The structure of the mindful brain

The findings of the discussed articles showed that MM as conceptualized by the authors is associated with widely distributed parts of the brain. These widely distributed brain regions can be broadly categorized in being involved with functions regarding bodily attention and visceral awareness, emotion and cognition, attention, respiratory and cardiac control, and pain regulation. While these functions correspond fairly high to the goals and outcomes of MM, in my opinion, they do not necessarily reflect MM specific characteristics only. Rather, they represent mental practice in the context of MM.

When trying to isolate convergent findings, it becomes apparent that the right anterior insula is a brain region that is somewhat consistently associated with MM. Given the fact that Lazar and colleagues (2005), as well as Hölzel and colleagues (2008) found this brain region in Insight meditators, it is suggested that the right anterior insula is specifically associated with this type of MM only. Additionally, only the facet of “describing” of the

(27)

24

FFMQ questionnaire was positively correlated to the right anterior insula. It is suggested that the core characteristic of Insight meditation is the describing of one’s own feelings, and the nonjudging of experience, and can therefore show its associations with the right anterior insula that has the functionality of interoception.

Furthermore, two of the current studies have incorporated the FFMQ in their study (Grant et al., 2013; Hölzel et al., 2011). However, in the study of Hölzel et al. (2011), FFMQ scores were not correlated to any neuroanatomical changes in response to the 8-week MBSR intervention program. That is to say, components that have shown to be inherent to MM states were not able to actively alter brain structure. This suggests that MBSR intervention as a whole, rather than the true MM components, were responsible for neuroanatomical alterations. In my opinion, the findings should therefore be placed in a much broader framework, with both MM specific as well as non-specific components affecting differential brain networks. For example, brain regions other than the right anterior insula that were linked to MM are more likely linked to more general mental processes such as self-perception, bodily awareness, and attentional and emotional regulation. It is tempting to assign these mental practices under the umbrella term of MM, yet findings have demonstrated obvious differences of associated brain structures with mental practice in the context of MM. In addition, Grant et al. (2013) showed an indirect relationship of FFMQ scores with gray matter in the context of attentional absorption, supporting the misuse of the umbrella term of MM. In other words, attentional absorption and mindfulness may have shared experiential dimensions, but FFMQ scores were not directly related to gray matter structures. Generally, a validated measure of MM does not seem to have strong correlations with neuroanatomical structure.

To my knowledge, there are only a few studies that investigate MM in relation to white matter structure. This is very surprising, given the fact that white matter associations of MM could reveal potential enhanced brain connectivity patterns. Given the increased gray matter volumes and alterations in mindfulness meditators, it is not unlikely that mindfulness meditators also show increased brain connectivity. It is of particular interest to investigate the white matter tracts between the gray matter structures that are found to be associated with MM. Inconsistent findings regarding gray matter associations clearly need to be complemented by DTI studies in order to study the issue of structural associations of MM in more detail.

Furthermore, the same researchers conducted most of the studies that have investigated the neuroanatomical correlates of MM. While I do not impeach any of the researchers in any way, I do think it is possible that ideas regarding, for example, the conceptualization and operationalization of MM might be somewhat biased. For example, it can be argued that modern MM practices such as MBSR are an admixture of different

(28)

25

types of meditation, stretching exercises, group therapies, and social interactions. Nevertheless, the authors of Hölzel et al. (2011) still promote in their journal article title that it is “Mindfulness practice” that leads to increases in regional brain gray matter density. In addition, it is worth mentioning that the meditation population under study by Luders and colleagues in 2011 is partly the same sample as in their 2009 gray matter study and 2011 DTI study. Therefore, I would put some question marks as to how generalizable the findings of the discussed articles are.

6.2. Potential mechanisms underlying gray matter alterations in mindfulness meditators By using MRI-based volumetric analyses, a general limitation is that we cannot distinguish the nature of the underlying correlates. While it remains highly speculative, it can be hypothesized that associations of MM and telomere length might shed led on the underlying biology of volumetric increases in mindfulness meditators with respect to non-meditators. Telomeres are specialized DNA structures that cap off the end of chromosomes, protecting them from degradation (Wicky et al., 1996). Consequently, a loss of telomeres assists the progress of increased genetic recombination and end-to-end chromosomal fusions, leading to genome instability (Masutomi et al., 2010). Furthermore, telomere length is considered a measure of cellular aging and cell proliferation, for the reason that telomeres become progressively shorter with each cell division (Wikgren et al., 2012). Accordingly, telomeric DNA components can become critically short such that it leads to cellular senescence or apoptosis (Harris et al., 2006). However, telomerase is an enzyme that counteracts telomere shortening by adding telomeric repeats to DNA, and thus has the potential to elongate telomeres (Olovnikov, 1996). Most human cells express low quantities of telomerase and therefore have limited cell growth.

Strikingly, a recent meta-analytic review showed that MM leads to increased telomerase activity in peripheral blood mononuclear cells (Schutte & Malouff, 2013). It is suggested that central telomerase activity in the brain is also increased by MM, yet it remains difficult to confirm, as it is impossible to directly assess telomerase in living individuals. Nevertheless, telomerase assessed from the peripheral blood mononuclear cells can be used as a proxy. Other studies have also demonstrated a requirement for telomerase in the cell survival-promoting actions of Brain Derived Neurotrophic Factor (BDNF) in hippocampal neurons (Fu et al., 2002). In other words, telomerase activity is suggested to elongate telomere length, as well as promote cell growth and survival, both in the periphery and the brain. This is of particular interest in the context of natural age-related gray matter atrophy. It can be argued that increased telomerase activity associated with MM may be able to counteract natural aging and gray matter atrophy, thus maintaining and potentially increasing gray matter volume.

(29)

26

In addition, gray matter volume in the hippocampus seems to decline by chronic stress (Gianaros et al., 2007). Recent studies have shown a dose-response relationship between chronic stress and telomere shortening in patients suffering from major depression disorder (Verhoeven et al., 2013). While it remains highly speculative, it is suggested that increased telomerase activity is a potential underlying neurobiological mechanism that is associated with increased gray matter volumes, specifically in the hippocampus. This is in line with the findings of Hölzel et al. (2008) and Luders et al. (2009) that found the right hippocampus to be larger in mindfulness meditators vs. non-meditators. Due to limited cell proliferation outside the hippocampus, it remains unlikely that telomerase activity can account for other increased neuroanatomical structures associated with MM. Telomerase activity can be assayed with different commercially available kits, for example, one of them being TRAPeze (Chemicon, USA). The methods of how to exactly assay telomerase activity is described in detail elsewhere (Wolkowitz et al., 2012). Simply by including telomerase activity as a predictor of gray matter volume in a structural MRI analysis (Voxel-Based Morphometry, FSL-FIRST and/or Freesurfer) can easily further test this potential mechanism.

6.3. Issues of theory & methodology

It is of profound importance that there is clarity on the meaning of MM; however, to date this is not yet the case in scientific research. As a consequence, inconsistent findings of the discussed studies in this review can potentially be explained by the variations in conceptualization of MM. A major limitation of the inconsistent conceptual and operational definition of MM is represented by the fact that different types of meditation may recruit distinct neural networks. The handful of fMRI studies previously mentioned showed activation of different brain mechanisms and regions, potentially reflecting this theoretical problem. In sum, the issues of theory make it increasingly difficult to directly compare findings across studies.

Furthermore, another limitation is represented by the multifaceted nature of MM practices. While the active beneficial component of MM might be present in all the adopted MM practices in previously mentioned studies, it remains unclear whether positive effects were produced by the same component, or even a component specific to MM for that matter (Chiesa & Serreti, 2010). For example, the mindfulness-based therapeutic interventions occur mostly in the form of groups, rather than individually. It therefore remains debatable whether the beneficial effects produced by the intervention program are a result of the active MM component, or merely, for example, the social group interactions. In addition, other non-specific MM components such as stress education and/or gentle stretching exercises may very well contribute to the positive effects of mindfulness-based

(30)

27

therapies (Hölzel et al., 2011). To date, it remains to be substantiated to which extent non-specific aspects of MM, and/or bias due to other theoretical shortcomings, are able to contribute to the positive effects that MM produces on a neuroanatomical level. Obviously, there is need for a validated way of investigating MM measures.

In neuroimaging techniques such as (f)MRI, brain images are acquired on different points over time and are then statistically compared across individuals. The different preprocessing and analysis steps can introduce a susceptibility to error, bias or variation. Therefore, one must always carefully consider neuroimaging parameters such as spatial smoothing and multiple comparison correction (Zatorre et al., 2012). Moreover, field strength appears to be a measuring bias that should also be taken into account when we compare the findings of the different studies, given that thickness values are reported to be higher in studies with a field strength of 3T, compared to 1.5T. In addition, it remains unclear what the exact correspondence is between the different morphological measures of cortical thickness, gray matter density/concentration, and gray matter volume. Therefore, the direct comparing of findings should be considered with caution.

Furthermore, there currently is a need for longitudinal studies that follow mindfulness meditators from the start of the trajectory of becoming an experienced meditator. That way, methodological issues such as the heterogeneity of meditation practices under study, and the degree of expertise can be controlled to a greater extent. However, it remains extremely difficult to find naïve participants that are willing to commit and dedicate their life to such intense mental training over a time span of many years. Nevertheless, with gray matter alterations already detectable after an 8-week intervention program, it would be of major interest to pursue such a study.

The only relevant MM specific longitudinal study by Hölzel et al. (2011) showed that the number of minutes of formal MM homework exercise was not positively correlated with changes in gray matter concentration. This lends confidence to the idea that the MBSR program as a whole influences morphological change, instead of MM per se. It is therefore obvious that more controlled, and eloquently designed longitudinal studies are needed in order to verify that MM is able to actively alter neuroanatomical structure.

The Neuroanatomical Correlates of Mindfulness Meditation: A Literature Review

The

Neuroanatomical

Correlates of

Mindfulness

Meditation: A

Literature Review

Literature Thesis

Table of Contents

1. Introduction

2. The concept and operationalization of Mindfulness Meditation

3. The neuroanatomical correlates of Mindfulness Meditation

4. From correlation to causation

5. Methodology

6. Discussion