Similar States, Different Paths: Neurodynamics of diverse meditation techniques
Prakash Shrimali, Arun Sasidharan, Saketh Malipeddi, Bianca Ventura, Rahul Venugopal, Ajay Kumar Nair, Ravindra P. Nagendra, Bindu M. Kutty, Georg Northoff
bioRxiv (Cold Spring Harbor Laboratory) June 26, 2025 preprint DOI: 10.1101/2025.06.20.660652 via OpenAlex
Summary
AI-generated from the abstractMeditation involves training attention inward, but the brain activity that distinguishes meditative from non-meditative states across different traditions is not well understood. Analyzing high-density EEG data from 170 participants—121 advanced meditators and 49 controls—across Vipassana, Brahma Kumaris Raja Yoga, Heartfulness, and Isha Yoga traditions, researchers used random forest classifiers to distinguish meditative from non-meditative states with 91% accuracy. Nonlinear features contributed most, indicating a core neurodynamic profile. Classification was higher in advanced meditators (92%) than controls (85%), with different feature importance: nonlinear and aperiodic features dominated in meditators, while oscillatory and timescale features dominated in controls. Each tradition showed distinct neurodynamic profiles, suggesting multiple pathways lead to meditative states.
Study at a glance
| Characteristics | Observational study |
|---|---|
| Sample size | 170 |
| Population | Advanced meditators and controls across four traditions |
| Topics | Meditation |
| Keywords | Computer science Cognitive psychology Geography Archaeology |
| Citations | 1 |
| Key finding | Meditative and non-meditative states can be distinguished with 91% accuracy using EEG features, with nonlinear features contributing most, and each tradition showing distinct neurodynamic profiles. |
Abstract
Abstract Meditation encompasses diverse practices that train attention inward, in contrast to externally oriented task states. However, the neurodynamic features distinguishing meditative states from non-meditative states across traditions remain unclear. We analyzed high-density EEG data (N=170; 121 advanced meditators, 49 controls) across four traditions: Vipassana, Brahma Kumaris Raja Yoga, Heartfulness, and Isha Yoga. EEG features spanned oscillatory, aperiodic, nonlinear, and timescale components. Using random forest classifiers, we distinguished meditative from non-meditative states with robust classification performance (91%). Nonlinear features contributed the most, suggesting a core neurodynamic profile. Classification performance was higher in advanced meditators (92%) than in controls (85%), with distinct feature importance: nonlinear and aperiodic features dominated in meditators, and oscillatory and timescale features in controls. Each tradition showed distinct neurodynamic profiles, indicating technique-specific constellations. Our findings revealed shared yet distinct neurodynamic signatures across meditation techniques, suggesting that multiple neurodynamic pathways lead to meditative states.