Expert compassion meditators show cortical thickness increases in socio-affective brain networks

Introduction: In humans, only few studies have addressed plasticity in networks mediating socio-affective skills that may require the generation of positive affect, such as loving kindness, compassion, and empathic concern. A recent short-term longitudinal functional MRI study provided evidence of functional plasticity after compassion and loving-kindness training in meditation-naïve participants in brain networks associated with positive affect including ventral striatum, mid insular and medial orbitofrontal cortex (mOFC) (Klimecki et al., 2012). So far studies are lacking focusing on structural changes based on such affective mental training. In the absence of costly longitudinal imaging studies, cross-sectional assessment of experts offers a promising approach to infer training-related brain changes on a structural level. Specifically, studying long-term meditation practitioners (LTPs) specialized on specific compassion-enhancing meditation practices provides a useful framework to reveal brain plasticity associated to compassion expertise. We therefore assessed structural changes in LTPs by assessing MRI-based cortical thickness through surface-based analysis and region-of-interest assessment, focusing on cortical structures such as mOFC and insula as these were cortical region known to undergo functional plasticity after compassion training. Methods: We studied 17 LTPs (5 Women; 45-62 years). All meditators were practicing the Nyingma tradition of Tibetan Buddhism, the latter being known for specifically focusing on the cultivation of loving-kindness, altruism, and compassion. Participants were included only if they had previously participated in a full-time meditation retreat of at least 3 years. LTPs reported having practiced meditation with an estimated cumulative total of 40k h (10k-62k h). LTPs were compared to 15 age- and sex-matched healthy controls without prior meditation experience (5 women, 46-63 years). At 3T, we acquired T1-weigthed MRI data together with functional MRI data during a loving-kindness meditation and a resting-state (each 7 min long) in all subjects. On T1-weighted MRI, FreeSurfer was used to generate cortical surface models and to measure cortical thickness. To define a region-of-interest related to compassion and loving-kindness, we intersected functional MRI increases during loving-kindness meditation relative to a non-meditative resting-state in our LTPs with a network previously shown to undergo functional plasticity in meditation-naïve subjects after compassion training (Klimecki et al., 2012). This approach yielded a specific overlap in the mOFC. Analysis was performed using SurfStat for Matlab (Worsley et al., 2009); surface-based findings were controlled at FWE4.0, p