Brain mechanisms supporting the modulation of pain by mindfulness meditation

Abstract

The subjective experience of one's environment is constructed by interactions among sensory, cognitive, and affective processes. For centuries, meditation has been thought to influence such processes by enabling a nonevaluative representation of sensory events. To better understand how meditation influences the sensory experience, we used arterial spin labeling functional magnetic resonance imaging to assess the neural mechanisms by which mindfulness meditation influences pain in healthy human participants. After 4 d of mindfulness meditation training, meditating in the presence of noxious stimulation significantly reduced pain unpleasantness by 57% and pain intensity ratings by 40% when compared to rest. A two-factor repeated-measures ANOVA was used to identify interactions between meditation and pain-related brain activation. Meditation reduced pain-related activation of the contralateral primary somatosensory cortex. Multiple regression analysis was used to identify brain regions associated with individual differences in the magnitude of meditation-related pain reductions. Meditation-induced reductions in pain intensity ratings were associated with increased activity in the anterior cingulate cortex and anterior insula, areas involved in the cognitive regulation of nociceptive processing. Reductions in pain unpleasantness ratings were associated with orbitofrontal cortex activation, an area implicated in reframing the contextual evaluation of sensory events. Moreover, reductions in pain unpleasantness also were associated with thalamic deactivation, which may reflect a limbic gating mechanism involved in modifying interactions between afferent input and executive-order brain areas. Together, these data indicate that meditation engages multiple brain mechanisms that alter the construction of the subjectively available pain experience from afferent information.

Figure 1.

Experimental procedures across time. First column, Psychophysical training session: Subjects first came in for psychophysical training. In this session, subjects were familiarized with visual analog scales, the range of thermal stimulation, and thermal stimulation paradigm used in the subsequent MRI sessions. Second column, MRI session 1: In the first two blocks, subjects were asked to reduce movement and keep eyes closed (rest). A heat (49°C) and neutral (35°) series were randomly presented in one of two blocks. Before anatomical acquisition, subjects were instructed to “begin meditating by focusing on the sensations of the breath.” Subjects continued to attend to the breath during a block of noxious stimulation (49°C) or neutral (35°). Pain ratings were assessed after each block. Third column: The 4 d (20 min/d) meditation intervention. Subjects were taught to focus on the changing sensations of the breath. They were taught that discursive thoughts were to be acknowledged without affective reaction and “let go” by redirecting their focus back on breath sensations. In sessions 3 and 4, sounds of the MRI scanner were introduced to familiarize subjects with the MRI environment. The fourth column describes the MRI session 2 (after meditation training). In the first four blocks, subjects were instructed to reduce movement and close their eyes (rest). Two heat (49°C) and two neutral (35°C) blocks were randomly administered. Before anatomical acquisition, subjects were instructed to “begin meditating by focusing on the sensations of the breath.” Subjects continued to meditate across two blocks of noxious stimulation (49°C) and neutral (35°C). Pain ratings were assessed after each block.

Figure 2.

Mean (SEM) psychophysical pain ratings across each session. Meditation, after training, significantly reduced pain intensity ratings and pain unpleasantness ratings when compared with rest *p < 0.05.

Figure 3.

Brain activations and deactivations illustrate the main effects of ATB and pain in the MRI session before training. In the main effect of pain, there was greater activation in SI corresponding to the stimulation site, ACC, SII, left putamen, and bilateral insula. There was no ATB-related brain activity, but the deactivations for the main effect of meditation were found in the medial PFC, posterior cingulate cortex, thalamus, and paracingulate gyrus. Slice locations correspond to standard stereotaxic space.

Figure 4.

Brain activations and deactivations illustrate the main effects of pain and meditation, as well as the overlap between pain and meditation in MRI session 2 (after training). Noxious stimulation activated the ACC, bilateral anterior insula, and posterior insula/SII. Meditation activated bilateral ACC, OFC, ventral striatum, anterior insula, SI, and SII. Moreover, meditation was associated with deactivations in the medial PFC and posterior cingulate cortex, consistent with default mode network activation. There was significant overlap between meditation and pain in the ACC and anterior insula, suggesting that these areas serve as a possible substrate for pain modulation. Slice locations correspond to standard stereotaxic space.

Figure 5.

Interaction between meditation and pain-related brain activation in MRI session 2. General linear modeling analyses detected a significant interaction in SI (z = 68) between meditation and rest in the presence of noxious stimulation. There was a significant activation of the contralateral SI during heat stimulation while subjects were not meditating. Meditation significantly reduced activation in SI during noxious heat stimulation. Slice locations correspond to standard stereotaxic space.

Figure 6.

Relationship between meditation-induced decreases in psychophysical pain ratings and pain-related brain activation. Subjects reporting the greatest decrease in pain intensity ratings also exhibited the largest increase in the ACC and right anterior insula activation. Top, Similarly, subjects reporting the greatest activation in the OFC exhibited the greatest decreases in pain unpleasantness. In contrast, greater deactivation in the thalamus was related to larger decreases in pain unpleasantness ratings. Bottom, Slice locations correspond to standard stereotaxic space.

Figure 7.

Paired t test illustrating differences in brain activation between the main effects of pain and meditation across MRI sessions. Top, Noxious stimulation activated significantly greater SI corresponding to the stimulation site, bilateral SII, and bilateral insula before training when compared with after training. After training, noxious stimulation activated greater medial PFC, frontal pole, thalamus, and ACC when compared with before training. Bottom, There was greater superior temporal gyrus activation during the ATB condition before training compared with meditation after training. However, there was greater OFC, ACC, and right anterior insula meditation-related activation after training when compared with ATB before training. Slice locations correspond to standard stereotaxic space.