Report from a Tibetan Monastery: EEG neural correlates of concentrative and analytical meditation

Abstract

The positive effects of meditation on human wellbeing are indisputable, ranging from emotion regulation improvement to stress reduction and present-moment awareness enhancement. Changes in brain activity regulate and support these phenomena. However, the heterogeneity of meditation practices and their cultural background, as well as their poor categorization limit the generalization of results to all types of meditation. Here, we took advantage of a collaboration with the very singular and precious community of the Monks and Geshes of the Tibetan University of Sera-Jey in India to study the neural correlates of the two main types of meditation recognized in Tibetan Buddhism, namely concentrative and analytical meditation. Twenty-three meditators with different levels of expertise underwent to an ecological (i.e., within the monastery) EEG acquisition consisting of an analytical and/or concentrative meditation session at "their best," and with the only constraint of performing a 5-min-long baseline at the beginning of the session. Time-varying power-spectral-density estimates of each session were compared against the baseline (i.e., within session) and between conditions (i.e., analytical vs. concentrative). Our results showed that concentrative meditation elicited more numerous and marked changes in the EEG power compared to analytical meditation, and mainly in the form of an increase in the theta, alpha and beta frequency ranges. Moreover, the full immersion in the Monastery life allowed to share the results and discuss their interpretation with the best scholars of the Monastic University, ensuring the identification of the most expert meditators, as well as to highlight better the differences between the different types of meditation practiced by each of them.

Keywords: Buddhist monks; EEG; analytical meditation; concentrative meditation; neural correlates.

Figure 1

Exemplary of an analytical meditation session (left) and concentrative meditation session (right). Top: (A,B) tvPSD: each line in the spectrum corresponds to the PSD estimated in a 5-min-long window. Time is encoded in line color, from the beginning of the session (blue) to the end of the session (red). Bottom: (C,D) changes in tvPSD compared to baseline. The tvPSD is integrated in each considered frequency band.

Figure 2

Average scalp map distribution for analytical (A) and concentrative (C) meditation. For each frequency band, the maximum difference (dB) (averaged across subjects) between the PSD during the session and the baseline is reported. Top: maximal positive deviations from baseline. Bottom: maximal negative deviations from baseline.

Figure 3

Analytical vs. concentrative. Violin plots of frequency-band deviation in dynamic PSD. Statistically significant differences (p < 0.05, FDR- corrected) are marked with an asterisk (*).

Figure 4

Analytical vs. concentrative. Violin plots of Δ_α-peak. This feature was the feature with the most marked differences between analytical and concentrative sessions. Statistically significant differences (p < 0.05, FDR- corrected) are marked with an asterisk (*).

Figure 5

Exemplary of bump. A marked peak is observed at 21.5 Hz.

Figure 6

(A) In this subject the α peak is absent on the baseline. Then it emerges at frame 55–60 (green line), to turn off abruptly at frame 65–70 and remain off until the end of the session. (B) In this subject the α peak, is reduced in the baseline and then there is a first switch in frame 10–15 in which the whole spectrum grows, and the alpha peak is absorbed, then the spectrum returns to the state prior to the switch, and we assist to an increase of the alpha peak for approximately 95 min. In windows ranging from 110 to 115 min the phenomenon happens again and the alpha peak disappears.

Figure 7

Changes in the PSD compared to baseline for 3 subjects. Inverse behavior between θ/α₁/α₂ and β₂/γ AFB PSD. (A) After the first part of the session the power in α₁ and α₂ increase with respect to the baseline, whereas β₂ and γ decrease: the phenomenon is evident in the first 7 points (35 min). (B) The inverted trend starts after 10 min of meditation and continues for the whole session: the phenomenon is evident in the first 8 points (40 min). (C) The inverted trend is clear after at the very beginning of the session (10 to 20 min) and during the las 20 min of meditation: the inverse behavior is evident in the initial and final peaks situated, respectively, between 10 and 20 min and 55 and 70 min.

Figure 8

Changes in tvPSD compared to baseline. We report changes during an analytical session for the same meditator of Figure 7C.