Does the MRI/fMRI Procedure Itself Confound the Results of Meditation Research? An Evaluation of Subjective and Neurophysiological Measures of TM Practitioners in a Simulated MRI Environment.

Abstract

Early research into meditation, including Transcendental Meditation (TM), relied exclusively on EEG to measure brain activity during meditation practice. Since the advent of neural imaging, MRI, and later fMRI, have dominated this field. Unfortunately, the use of this technology rests on the questionable assumption that lying down in a confining tube while exposed to very loud sounds would not interfere with the meditation practice. The present study was designed to assess the effects of the fMRI procedure on both the subjective and neurophysiological responses of short and long-term TM practitioners. Twenty-three TM practitioners volunteered to participate in this study: 11 short-term meditators, averaging 2.2 years practice, and 12 long-term meditators, averaging 34.8 years. The repeated-measures design included two activities for each participant, eyes-closed rest, and TM practice, in each of three conditions: sitting quietly in an upright position (normal TM practice); lying quietly in a supine position; and lying, with earplugs, inside a simulated fMRI tube (simMRI), while exposed to 110 dB recordings of an actual fMRI machine. Subjective experiences were collected after each activity in each condition. Physiological arousal was recorded using skin conductance levels. Scalp EEG was averaged into eight frequency bands within frontal and parietal leads; eLORETA software was used to explore the 3-D cortical distribution of EEG sources. During the simMRI condition, participants reported having more shallow meditation experiences, and greater agitation/distraction. Skin conductance levels paralleled self-reports, decreasing least during the simMRI condition. Frontal and parietal power decreased from sitting to simMRI in the alpha2 through gamma bands. Parietal power was higher during rest compared to TM in the alpha1 through beta2 bands. Frontal and parietal alpha1 coherence were highest during the simMRI condition. The eLORETA analysis revealed that the default mode network was more active during TM when sitting compared to the simMRI condition. The responses to the supine condition were generally between sitting and simMRI, with some significant exceptions. In conclusion, these data indicate that the fMRI procedure itself (high dB noise; lying down) strongly influences subjective and neurophysiological responses during meditation practice, and may therefore confound the interpretation of results from fMRI studies.