Dynamic Neural State Transitions Predict Psychedelic Phenomenal Richness: Magnitude, Not Direction, Drives Subjective Intensity

Background: Psychedelic substances produce profound alterations in conscious experience, yet the neural mechanisms underlying phenomenal richness remain poorly understood. Static measures of neural integration and complexity have shown limited success in predicting subjective intensity, suggesting alternative approaches may be needed. Methods: We analyzed EEG data from 33 participants (after excluding 2 for excessive artifacts) who inhaled N,N-Dimethyltryptamine (DMT) in naturalistic settings. We computed the Consciousness Gradient Index (CGI = √(φ × ρ)), where φ represents global cortical integration and ρ represents spectral entropy, across 30-second windows throughout the experience. We correlated the rate of CGI change with validated phenomenological measures (5D-Altered States of Consciousness scale). Results: Static measures of integration (φ), complexity (ρ), and their combination (CGI) showed no correlation with phenomenal richness (all r < 0.10, p > 0.65). In contrast, the maximum rate of CGI change significantly predicted overall phenomenal intensity (r = 0.378, p = 0.028, 95% CI [0.060, 0.632]) and composite phenomenal richness (r = 0.378, p = 0.028). This effect remained significant after controlling for baseline anxiety and mean CGI levels. Critically, both increases and decreases in CGI predicted richness equally (z = 0.08, p = 0.94), indicating that the magnitude of neural state transitions - not their direction - drives subjective intensity. Conclusions: Phenomenal richness in psychedelic states depends on dynamic exploration of neural state space rather than occupation of specific integration or complexity levels. These findings challenge static theories of consciousness (e.g., Integrated Information Theory) and support dynamic models emphasizing state-space traversal. The bidirectional effect suggests consciousness richness reflects the brain's capacity to explore diverse neural configurations, with implications for understanding altered states and developing consciousness theories