Review
Is meditation associated with altered brain structure? A systematic review and meta-analysis of morphometric neuroimaging in meditation practitioners

https://doi.org/10.1016/j.neubiorev.2014.03.016Get rights and content

Highlights

  • Many studies have examined ‘morphometric’ neuroimaging of meditation practitioners.

  • We conduct a meta-analysis of these gray and white matter differences in meditators.

  • We find consistent differences in prefrontal cortex and body awareness regions.

  • Global mean effect size is ‘moderate’ (Cohen's d = 0.46; r = .19).

  • Results suggest consistent and medium-sized brain structure differences.

Abstract

Numerous studies have begun to address how the brain's gray and white matter may be shaped by meditation. This research is yet to be integrated, however, and two fundamental questions remain: Is meditation associated with altered brain structure? If so, what is the magnitude of these differences? To address these questions, we reviewed and meta-analyzed 123 brain morphology differences from 21 neuroimaging studies examining ∼300 meditation practitioners. Anatomical likelihood estimation (ALE) meta-analysis found eight brain regions consistently altered in meditators, including areas key to meta-awareness (frontopolar cortex/BA 10), exteroceptive and interoceptive body awareness (sensory cortices and insula), memory consolidation and reconsolidation (hippocampus), self and emotion regulation (anterior and mid cingulate; orbitofrontal cortex), and intra- and interhemispheric communication (superior longitudinal fasciculus; corpus callosum). Effect size meta-analysis (calculating 132 effect sizes from 16 studies) suggests a global ‘medium’ effect size (Cohen's d¯=0.46; r¯=.19). Publication bias and methodological limitations are strong concerns, however. Further research using rigorous methods is required to definitively link meditation practice to altered brain morphology.

Introduction

A range of effects have been associated with long- and short-term training in the mental practices broadly referred to as ‘meditation.’ A few striking examples include enhancement of executive functions, such as attention (Jha et al., 2007), working memory (Jha et al., 2010), and introspection (Fox et al., 2012, Sze et al., 2010); improved immune function (Davidson et al., 2003, Jacobs et al., 2011); better perceptual discrimination (MacLean et al., 2010); increased prosocial (compassionate) behavior (Condon et al., 2013); and symptom improvements in clinical disorders, such as anxiety and depression (Vollestad et al., 2012). Skepticism is certainly warranted, however, when a relatively straightforward intervention demonstrates such a wide variety of benefits. With the aim of evaluating the consistency and practical significance of this body of results, a recent comprehensive meta-analysis found robust evidence that meditation practice is associated with an array of cognitive and emotional benefits that often achieve medium to large effect sizes (Sedlmeier et al., 2012). As evidence mounts that meditation may have wide-ranging and measurable effects on many aspects of brain, body, and behavior, understanding the biological mechanisms that underlie these effects is of paramount scientific and public health importance.

The study of the functional neuroanatomical bases that drive meditation's apparently salutary effects remains in its infancy, however. This is all the more true of research examining putative differences in the anatomical structure of the brains of meditation practitioners. Although many studies have examined meditation with functional methods such as electroencephalography (EEG), event-related potentials (ERPs), positron emission tomography (PET), and functional magnetic resonance imaging (fMRI) (reviewed in Cahn and Polich, 2006, Hölzel et al., 2011b, Vago and Silbersweig, 2012), an understanding of potential structural differences via ‘morphometric’ neuroimaging remains limited (Table 1). Such an understanding is important because evidence is mounting that experience-dependent structural differences in both gray (Draganski and May, 2008, Lövdén et al., 2013) and white matter (Johansen-Berg, 2010) are fundamental to many aspects of learning and behavior in humans (though for a counterpoint, see Thomas and Baker, 2012).

Since the first morphometric study of meditation less than a decade ago (Lazar et al., 2005), numerous studies have addressed the potential effects of meditation on brain morphology (Table 2), with over 120 results already reported (Table S1). However, the wide variety of results, sample sizes, and meditation styles makes it very difficult to readily draw a cohesive picture of meditation's relationship to brain morphology. This difficulty is compounded by the diversity of morphometric neuroimaging methods employed (Table 1).

. Summary of regions in which morphometric heterogeneities have been found in either experienced meditation practitioners, or in normal subjects after short-term meditation training.

In the absence of an overall view of what has been achieved so far, two central questions remain: (1) Is meditation associated with altered brain morphology in any consistent, replicable ways? And if so, (2) what is the magnitude (size) of these differences – i.e., are the differences potentially of any practical importance or clinical relevance? Another open question concerns whether meditation is in fact the causative factor in these brain morphology changes, because the majority of studies to date have employed cross-sectional methods, comparing long-term meditation practitioners to meditation-naïve controls. Such cross-sectional studies cannot address the critical question of whether meditation is in fact causing the observed structure differences, or whether pre-existing differences of brain morphology make certain people more likely to engage in intensive meditation practice (see more details in Section 6.2). A few recent studies, however, have used pre–post interventions that can begin to address the causal role of meditation in altering brain morphology. Another key goal, then, was to examine which regions are consistently altered after short-term meditation training, therefore suggesting a causal role for meditation – and to examine the extent to which such meditation-induced changes resemble the cross-sectional differences observed in long-term practitioners. A third, ancillary question, then, was (3) what evidence there might be for mediation as the causative factor in altered brain morphology.

These are all complex questions. A detailed review and meta-analysis of both neuroimaging results and effect sizes therefore seemed necessary for understanding how morphometric neuroimaging has been applied to the study of meditation, and what conclusions, if any, can be drawn from this body of work. Here, we attempt to synthesize the literature to date by performing both a systematic review and quantitative meta-analyses of all extant morphometric neuroimaging studies of meditation. Our central aim is to provide answers to the two fundamental questions posed above, and to address, so far as is possible given the limited evidence to date, the third question regarding causation.

First, of the many findings on brain morphology differences related to meditation practice, are there regions that have been repeatedly implicated in subsequent studies? To answer this question, we used anatomical likelihood estimation (ALE), a quantitative meta-analytic technique (Eickhoff et al., 2009, Eickhoff et al., 2012, Laird et al., 2005, Turkeltaub et al., 2002), to identify brain regions showing consistent heterogeneities in meditation practitioners. Where use of ALE was not possible, we also ‘qualitatively’ reviewed all results to see which regions were repeatedly (in ≥3 studies) implicated in meditation. Neuroimaging studies typically report a ‘peak’ focus in the brain, where differences between groups (meditators vs. controls) is most statistically significant. Each study reports multiple such peaks of greatest anatomical difference; compiling all these peaks together, the ALE method seeks for statistically significant overlaps in the peaks from independent studies. In this way, we were able to compile a list of regions that appear to be consistently altered in meditation practitioners, across many independent studies and samples (see Section 2 for detailed information).

Knowing that certain regions are consistently different in meditation practitioners, however, does not necessarily imply that the differences are of any practical significance. Even if consistent and statistically significant, such brain structure differences might be too small to be considered relevant in a practical, everyday sense. Calculation of effect sizes, however, which indicate not the significance but the magnitude of results, can begin to address these questions of practical significance (Cumming, 2013). Simply testing whether a result is significant or not (null-hypothesis significance testing) is limited by the fact that attaining significance is very much dependent on sample size (Cumming, 2013) – and sample size is generally quite small in most neuroimaging studies, due to the high costs involved. Effect sizes, however, estimate the magnitude of differences between groups, regardless of whether the result was statistically significant (where non-significance, e.g., might be due simply to small sample size). Our effect size meta-analysis therefore allowed an overview of the apparent magnitude of brain structure differences reported in meditators. Although effect sizes are still rarely reported in neuroimaging studies, and their interpretation with respect to brain structure differences remains problematic and poorly developed at the theoretical level (Poldrack et al., 2008), we nonetheless aimed to present all quantitative effect size data and offer some preliminary interpretations of their significance.

Brain ‘morphology’ refers to the structure, shape, and composition of the brain; the measurement and analysis of brain morphology via various neuroimaging techniques is generally known as ‘morphometry’ or ‘morphometric neuroimaging’ (Draganski and May, 2008, May and Gaser, 2006, Zatorre et al., 2012). Broadly speaking, morphometric neuroimaging techniques aim to characterize anatomical differences based on a variety of morphological characteristics. Some relate solely to the brain's shape or size (e.g., cortical gyrification), others take into account the relative concentration or organization of gray and white matter (e.g., gray matter concentration), and yet others combine both aspects (e.g., volumetry of predefined gray matter structures). Morphometric neuroimaging stands in contrast, then, to ‘functional’ neuroimaging techniques such as fMRI, EEG, and PET, which aim to characterize not brain structure, but brain activity, such as changes in electrical potentials or blood flow.

A brief overview of measures used to date in morphometric studies of meditation is presented in Table 1 (for in-depth reviews outside the field of meditation, see Draganski and May, 2008, May and Gaser, 2006, Zatorre et al., 2012; for specific methods, see Ashburner and Friston, 2000, Beaulieu, 2002, Fischl and Dale, 2000).

A sometimes-tacit assumption underlying morphometric neuroimaging is that greater values (structural ‘increase’) on a given morphometric measure entail a corresponding enhancement of function. The structural increases in question could be, e.g., an increased concentration of gray matter in a given region; an increased thickness of cerebral cortex; increased integrity of white matter fibers; or any number of other measures (see Table 1). In support of this view, there are well-established connections between brain maturation and cognitive development, as well as a complementary link between neurodegenerative disease, or atrophy, and cognitive decline.

More specifically, there exists fairly robust evidence in favor of the brain structure–function connection in both animal models and human neuroimaging. Several important studies have established relationships between structural ‘increases’ of both gray and white matter (for recent reviews, see Taubert et al., 2012, Zatorre et al., 2012) and beneficial outcomes, including achievement in a variety of fine motor skills, such as juggling (Draganski et al., 2004, Scholz et al., 2009) and musical instrument playing (Hyde et al., 2009). Even gross physical activities, such as aerobic exercise, show an ‘enhancing’ effect on brain morphology (Colcombe et al., 2006).

Importantly, such differences are observed not only in response to physical or motor skill training: some studies have recently found morphometric differences after mental training in reasoning (Mackey et al., 2012) and working memory (Takeuchi et al., 2011). Conversely, structural deterioration or deficiencies measured via morphometric neuroimaging have been linked to various forms of cognitive decline, including normal age-related cognitive decline (Good et al., 2001) and Alzheimer's disease (Frisoni et al., 2007).

The possibility remains, of course, that ‘less is more’ in at least some cases: the phenomenon of synaptic pruning provides a forceful example (Low and Cheng, 2006). Structural increases might also indicate functional impairments in at least some cases: several brain regions related to stimulus-response learning and habit formation show structural increases in obsessive compulsive disorder, for instance (Pujol et al., 2004).

Morphometric neuroimaging in meditation practitioners has generally aimed to explore whether meditation, too, is analogous to a form of (mental) skill learning, and can produce such anatomical changes. If so, brain structure increases related to meditative practice might provide at least a partial neural explanation of the numerous cognitive and emotional benefits associated with meditation (Sedlmeier et al., 2012). It should be acknowledged, however, that both in the field of morphometric neuroimaging as a whole, as well as within the smaller realm involving meditation practitioners in particular, the meaning of these brain structure differences is still very poorly understood. Very few studies have been directly replicated, and very few have correlated behavioral changes with brain structure differences. Enthusiasm about altered brain structure in meditation practitioners should therefore be tempered by the fact that the significance of these changes remains controversial (cf. Thomas and Baker, 2012); indeed, this is one of the main reasons for the present meta-analysis.

In collating data from multiple morphometric neuroimaging modalities, our interest is in the regions where differences have consistently been reported, irrespective of imaging method. The assumption is not that morphometric methods are necessarily directly comparable, but rather that particular brain regions are reliably involved in particular cognitive and emotional processes. Accordingly, alteration of a region's structure (regardless of imaging method) is presumed to entail a corresponding alteration in its function(s).

Whether a morphological difference in a single region will yield consistent results across morphometric methods is poorly understood. Since very few studies employ multiple methods simultaneously, direct comparisons are rare. However, there is preliminary evidence that results from disparate methods are comparable. For instance, Hutton et al. (2009) found broadly similar results when comparing two different-aged populations, using both gray matter concentration and cortical thickness analysis, and Testa et al. (2004) found that volumetry methods showed results consistent with gray matter concentration analysis. Nevertheless, different methods should not be expected to produce entirely consistent results, since they likely rely on different underlying cellular changes for their outcomes (see Section 5.5). Ultimately, the differing sensitivity of various methods may prove to be a source of additional information, rather than a shortcoming (for a critical discussion, see Lemaitre et al., 2012).

Meditation techniques vary enormously in aims, scope, difficulty, and tentatively, recruitment of brain regions (Brewer et al., 2011, Lee et al., 2012, Lou et al., 1999, Manna et al., 2010). Zen practice, for instance, tends to involve an open, undirected awareness of the present moment (Austin, 1999). Some traditions of Vipassana (‘Insight’) meditation, on the other hand, focus very explicitly on body sensations in a directed, systematic fashion (Goenka, 2000). Yet other practices involve detailed visualizations, simple awareness of the breath, or audible repetition of a particular phrase (a ‘mantra’) (Singh, 1979).

There are several influential attempts to find commonalities among techniques, however. The most well-known scheme categorizes practices into either ‘focused attention’ or ‘open monitoring’ meditations (Lutz et al., 2008), alternatively referred to as ‘concentrative’ and ‘mindfulness’ techniques, respectively (Cahn and Polich, 2006). Focused attention practices involve concentration of attention on a single object of meditation (e.g., the sensations of the breath, the recitation of a phrase, or the mental visualization of an image). Open monitoring practices, sometimes referred to as ‘choiceless awareness,’ instead involve an open, receptive, non-judgmental attitude toward any and all experience, regardless of origin (external/sensory or internal/mental) and affective tone (positive, negative, or neutral).

With respect to morphometric neuroimaging, however, it is difficult to study the neural basis of each category (much less each particular technique) independently of the others, for several reasons. Most practitioners examined to date have substantial experience with multiple categories, and more specifically, there is a dearth of studies examining only focused attention meditation practitioners (since focused attention meditation is almost always combined with, or followed by, open monitoring and compassion types of meditation). Moreover, numerous studies mix practitioners from multiple traditions in their analyses (Table 2). Therefore, despite the potential value of various classification schemes, comparative analyses based on meditation type were not undertaken here (although, where possible, a tentative discussion is offered). Whether distinct patterns of structural differences are related to particular forms of meditation practice therefore remains a question for future research.

Why the need for a new review and meta-analysis? Although a number of major efforts toward theoretical integration have been published in recent years (Hölzel et al., 2011b, Vago and Silbersweig, 2012, Farb et al., 2012), only a few thorough reviews of functional and morphometric neuroimaging in meditation practitioners have been undertaken (Cahn and Polich, 2006, Chiesa and Serretti, 2010, Ivanovski and Mahli, 2007, Rubia, 2009). Though generally comprehensive, several include only ‘mindfulness’ meditation, and only two recent studies (Sperduti et al., 2012, Tomasino et al., 2013) have conducted quantitative meta-analyses (ALE) of the burgeoning neuroimaging literature. These meta-analyses (Sperduti et al., 2012, Tomasino et al., 2013) suffer from certain limitations, such as no calculation or discussion of effect sizes, and no basic checks to ensure the robustness of the meta-analytic data (e.g., funnel plots or fail-safe N calculations; see Egger et al., 1997). These limitations are common to many earlier meta-analyses of meditation's cognitive and emotional effects as well (see Sedlmeier et al., 2012). Moreover, no synthesis or quantitative meta-analysis whatsoever of morphometric (i.e., structural) neuroimaging of meditation practitioners has yet been undertaken, despite the fact that the 21 studies examined here have already been cited more than 2200 times. Prior reviews and meta-analyses have instead tended to focus on functional neuroimaging results. In the present work we aim to fill this gap in the literature by providing a systematic review and quantitative meta-analysis of all morphometric neuroimaging studies of meditation.

Section snippets

Search strategy

Two of us (KCRF and SN) searched MEDLINE (http://www.pubmed.com), Google Scholar (http://scholar.google.com), and PsycINFO (http://www.apa.org/pub/databases/psycinfo/index.aspx) for all papers containing the word ‘meditation’ since the first morphometric study of contemplative practices was published (Lazar et al., 2005). These extensive lists of articles were then refined by searching within results for studies that contained any of the words or phrases ‘magnetic resonance imaging’, ‘MRI,’

Qualitative review of group differences in long-term practitioners and novices

Among all group differences (Table S1), we found 9 regions to be consistently (in ≥3 studies) reported (Table 3 and Fig. 1): rostrolateral prefrontal cortex (RLPFC)/BA 10, anterior/mid-cingulate cortex, insular cortex, somatomotor cortices, inferior temporal gyrus, fusiform gyrus, hippocampus, corpus callosum, and superior longitudinal fasciculus. We pool all results together in Table 3 to obtain an overview of consistent brain differences associated with meditation generally, not only

Convergent findings

In this section we discuss brain regions that have shown structural differences in multiple (≥3) studies (Table 3), and/or show significant clusters in the ALE meta-analysis (Table 4). We relate morphometric findings to relevant functional neuroimaging research in meditators and non-meditators, as well as anatomical investigations in non-human primates. We offer overviews of the putative functionality of each region along with hypotheses regarding how structural increases in a region might

Is meditation associated with altered brain structure?

Evidence for meditation practice as the causative factor in structural brain change remains tenuous, and much further work is needed before such a relationship is either established or disconfirmed. Several regions show consistent differences in advanced practitioners vs. meditation-naïve controls (Section 4), but the possibility remains that pre-existing brain structure heterogeneities explain the observed group differences. Other findings bear on the question of causality, however,

Meta-analytic methods, reliability, and limitations

In the present section we discuss in more detail the validity of the present meta-analysis, the checks of its robustness, and the limitations imposed both by our meta-analytic techniques and the methodology of the primary research reviewed.

Conclusions and directions for future research

At the outset of this review, we asked two fundamental questions about the morphometric neuroimaging of meditation practitioners: (1) Is meditation associated with altered brain morphology in any consistent, replicable way? And if so, (2) what is the magnitude of these differences? In this final section we present what we consider the best available answer to each question, and also provide some suggestions for future work along these lines.

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgments

The authors thank Dr. Tal Yarkoni, Dr. David Fresco, and Dr. Clifford D. Saron for extremely helpful discussions of meta-analytic methods and findings. We also thank Jelena Markovic, Zachary Stansfield, Dr. Lawrence M. Ward, and Dr. Todd C. Handy for helpful comments on earlier drafts of the manuscript, as well as four anonymous reviewers for detailed, considerate, and constructive comments. Finally, we thank Dr. Norman A.S. Farb and Dr. Sara Lazar for graciously providing previously

References1 (220)

  • J. Debiec et al.

    Cellular and systems reconsolidation in the hippocampus

    Neuron

    (2002)
  • R.C. deCharms et al.

    Learned regulation of spatially localized brain activation using real-time fMRI

    Neuroimage

    (2004)
  • B. Draganski et al.

    Training-induced structural changes in the adult human brain

    Behav. Brain Res.

    (2008)
  • M. Dresler et al.

    Dreamed movement elicits activation in the sensorimotor cortex

    Curr. Biol.

    (2011)
  • S.B. Eickhoff et al.

    Activation likelihood estimation meta-analysis revisited

    Neuroimage

    (2012)
  • M. Ellamil et al.

    Evaluative and generative modes of thought during the creative process

    Neuroimage

    (2012)
  • C.R. Genovese et al.

    Thresholding of statistical maps in functional neuroimaging using the false discovery rate

    Neuroimage

    (2002)
  • P.R. Goldin et al.

    The neural bases of emotion regulation: reappraisal and suppression of negative emotion

    Biol. Psychiatry

    (2008)
  • C.D. Good et al.

    A voxel-based morphometric study of ageing in 465 normal adult human brains

    Neuroimage

    (2001)
  • J.A. Grant et al.

    A non-elaborative mental stance and decoupling of executive and pain-related cortices predicts low pain sensitivity in Zen meditators

    Pain

    (2011)
  • J.A. Grant* et al.

    Cortical thickness, mental absorption and meditative practice: possible implications for disorders of attention

    Biol. Psychol.

    (2013)
  • T.V. Gurvits et al.

    Magnetic resonance imaging study of hippocampal volume in chronic, combat-related posttraumatic stress disorder

    Biol. Psychiatry

    (1996)
  • C.M. Hamamé et al.

    Reading the mind's eye: online detection of visuo-spatial working memory and visual imagery in the inferior temporal lobe

    Neuroimage

    (2012)
  • H.S. Hodgins et al.

    Attentional processes and meditation

    Conscious. Cogn.

    (2010)
  • S. Hofer et al.

    Topography of the human corpus callosum revisited – comprehensive fiber tractography using diffusion tensor magnetic resonance imaging

    Neuroimage

    (2006)
  • B.K. Hölzel et al.

    Differential engagement of anterior cingulate and adjacent medial frontal cortex in adept meditators and non-meditators

    Neurosci. Lett.

    (2007)
  • B.K. Hölzel* et al.

    Mindfulness practice leads to increase in regional brain grey matter density

    Psych. Res. Neuroimag.

    (2011)
  • C. Hutton et al.

    A comparison between voxel-based cortical thickness and voxel-based morphometry in normal aging

    Neuroimage

    (2009)
  • J.M. Allman et al.

    The anterior cingulate cortex

    Ann. N. Y. Acad. Sci.

    (2001)
  • Analayo

    Satipatthana: The Direct Path to Realization

    (2003)
  • B.J. Anderson

    Plasticity of grey matter volume: the cellular and synaptic plasticity that underlies volumetric change

    Dev. Psychobiol.

    (2011)
  • J.R. Andrews-Hanna

    The brain's default network and its adaptive role in internal mentation

    Neuroscientist

    (2012)
  • J.H. Austin

    Zen and the Brain

    (1999)
  • R.A. Baer et al.

    Construct validity of the five facet mindfulness questionnaire in meditating and nonmeditating samples

    Assessment

    (2008)
  • K.B. Baerentsen et al.

    Onset of meditation explored with fMRI

    Neuroimage

    (2001)
  • K.B. Baerentsen et al.

    An investigation of brain processes supporting meditation

    Cogn. Process.

    (2010)
  • C. Beaulieu

    The basis of anisotropic water diffusion in the nervous system – a technical review

    NMR Biomed.

    (2002)
  • M. Beckmann et al.

    Connectivity-based parcellation of human cingulate cortex and its relation to functional specialization

    J. Neurosci.

    (2009)
  • J.S. Beer et al.

    Orbitofrontal cortex and social behavior: integrating self-monitoring and emotion-cognition interactions

    J. Cogn. Neurosci.

    (2006)
  • T.E. Behrens et al.

    Learning the value of information in an uncertain world

    Nat. Neurosci.

    (2007)
  • A.M. Brausch et al.

    The role of body image and disordered eating as risk factors for depression and suicidal ideation in adolescents

    Suicide Life-Threat. Behav.

    (2009)
  • J.A. Brefczynski-Lewis et al.

    Neural correlates of attentional expertise in long-term meditation practitioners

    Proc. Natl. Acad. Sci. U. S. A.

    (2007)
  • J.A. Brewer et al.

    Meditation experience is associated with differences in default mode network activity and connectivity

    Proc. Natl. Acad. Sci. U. S. A.

    (2011)
  • D. Brown et al.

    Differences in visual sensitivity among mindfulness meditators and non-meditators

    Percept. Mot. Skills

    (1984)
  • R.L. Buckner et al.

    The brain's default network: anatomy, function, and relevance to disease

    Ann. N. Y. Acad. Sci.

    (2008)
  • B.R. Cahn et al.

    Meditation states and traits: EEG, ERP, and neuroimaging studies

    Psychol. Bull.

    (2006)
  • A.E. Cavanna et al.

    The precuneus: a review of its functional anatomy and behavioural correlates

    Brain

    (2006)
  • A. Chiesa et al.

    Mindfulness-based stress reduction for stress management in healthy people: a review and meta-analysis

    J. Altern. Complement. Med.

    (2009)
  • A. Chiesa et al.

    A systematic review of neurobiological and clinical features of mindfulness meditations

    Psychol. Med.

    (2010)
  • K. Christoff et al.

    The frontopolar cortex and human cognition: evidence for a rostrocaudal hierarchical organization within the human prefrontal cortex

    Psychobiology

    (2000)
  • Cited by (505)

    • Mindfulness and Management of Low Desire and Vulvovaginal Pain

      2024, Obstetrics and Gynecology Clinics of North America
    View all citing articles on Scopus
    1

    References marked with an asterisk indicate studies included in the meta-analysis.

    View full text