Meditation attenuates activity of a trans-diagnostic brain network: 3T and 7T fMRI evidence

1331 

Abstract Submission 

Saampras Ganesan¹, Bradford Moffat², Nicholas Van Dam³, Valentina Lorenzetti⁴, Andrew Zalesky^1,5

¹Department of Biomedical Engineering, The University of Melbourne, Melbourne, Victoria, ²Melbourne Brain Centre Imaging Unit, The University of Melbourne, Melbourne, Victoria, ³Melbourne School of Psychological Sciences, The University of Melbourne, Melbourne, Victoria, ⁴Faculty of Health, Australian Catholic University, Melbourne, Victoria, ⁵Department of Psychiatry, The University of Melbourne, Melbourne, Victoria, Australia

Saampras Ganesan, M. Eng.  
Department of Biomedical Engineering, The University of Melbourne
Melbourne, Victoria

Bradford Moffat, PhD  
Melbourne Brain Centre Imaging Unit, The University of Melbourne
Melbourne, Victoria

Nicholas Van Dam, PhD  
Melbourne School of Psychological Sciences, The University of Melbourne
Melbourne, Victoria

Valentina Lorenzetti, PhD  
Faculty of Health, Australian Catholic University
Melbourne, Victoria

Andrew ZALESKY, PhD  
Department of Biomedical Engineering, The University of Melbourne|Department of Psychiatry, The University of Melbourne
Melbourne, Victoria|Melbourne, Victoria, Australia

Burgeoning fMRI research in meditation alongside advancements in ultra-high field fMRI and real-time neuroimaging uniquely enable multi-faceted investigations into the neurobiology of meditation. Meditation encompasses mind-body attentional training [7], that can alleviate core psychopathological symptoms, such as dysfunctional self-referential thinking and mood [1, 5, 6], prevalent across multiple internalizing psychiatric disorders including depression, anxiety, schizophrenia, etc. Elucidating the most robust neurobiological underpinnings of meditation can help objectively inform and develop the therapeutic value of meditation in treating prevalent psychiatric disorders [6].

Here, we present and distil fMRI-measured brain activations reported across 3 independent studies of task-fMRI meditation research (Figure 1), comprising N=210 healthy individuals that either: i) meditated by paying attention to breathing sensations (meditation condition) or; ii) engaged in a non-meditative control condition (e.g., let their minds wander freely) inside the MRI scanner. The first study (N=174) involves whole-brain random-effects activation likelihood estimation (ALE) meta-analysis of significant brain activation coordinates reported in PRISMA-eligible 3T meditation task-fMRI studies [3]. The second study (N=10) involves group-level whole-brain general linear modelling (GLM) inferences from first-of-a-kind ultra-high field 7T meditation task-fMRI following stringent artefact modelling [4]. The final study (N=26) involves group-level inferences from a real-time 7T fMRI neurofeedback paradigm (2-day fMRI sessions) that aimed to train participant self-regulation of BOLD activation within a target brain region via meditation.

Key nodes of the default-mode network (DMN) were implicated commonly across all 3 independent studies, demonstrating significantly reduced activation during meditation compared to control condition (Figure 2). Specifically, the coordinate-based meta-analysis revealed significantly reduced activation (FWE-corrected p<0.05, 1000 permutations) during meditation vs. control condition in the medial prefrontal cortex (mPFC). Group-level GLM of the 7T meditation task-fMRI, adjusted for head-motion, physiological (RETROICOR) and arousal (self-report) variables, found significantly reduced activation (FWE-corrected p<0.05, z>3.1, 1024 permutations) during meditation vs. rest in precuneus, posterior cingulate cortex (PCC) and mPFC. Finally, mean BOLD percent-signal change within PCC (neurofeedback target region) estimated in real-time during neurofeedback self-regulation, after real-time adjustment for physiological confounds, was found to be significantly negative (T(25) = -4.0, p = 0.0005) at the group-level for meditation vs. rest.

The distillation of multifaceted brain activation findings from 3T task-fMRI meta-analysis, 7T task-fMRI and 7T fMRI neurofeedback conclusively demonstrates that meditation attenuates activation within the DMN, a brain network predominantly associated with self-referential processing. Notably, given the trans-diagnostic implication of dysfunctional self-referential processing and DMN in psychiatric conditions [2], our findings highlight the potential utility of meditation-based therapeutics in alleviating excessive and aberrant self-referential processing characteristic of several internalizing psychiatric disorders. This work also highlights the value of synthesising multifaceted findings across 3T and 7T MRI to illuminate putative neurobiological mechanisms underlying complex subjective states of awareness (i.e., meditation).

Self Processes

Higher Cognitive Functions Other

Activation (eg. BOLD task-fMRI) ¹

Imaging Methods Other

Consciousness and Awareness ²

Consciousness

FUNCTIONAL MRI

HIGH FIELD MR

Meta- Analysis

NORMAL HUMAN

Psychiatric Disorders

Statistical Methods

Other - Neurofeedback; Default-mode network; trans-diagnostic