Intensity and Visual Content Dissociate in DMT: Dynamic Integration-Complexity Predicts One, Spatial Configuration the Other

Abstract

Background. The intensity and content of conscious experience are theoretically separable, but rarely dissociated empirically within one dataset using EEG. Existing consciousness measures typically capture level or state without addressing content, or decode content without indexing intensity. No existing measure tracks the dynamic rate of integration-complexity reconfiguration.

Results. We introduce mac_I (mean absolute change of CGI, where CGI = √(φ × ρ); φ = global coherence, ρ = Lempel-Ziv complexity), a rate-of-change measure of the integration-complexity geometric mean. In inhaled-DMT EEG (N = 35), mac_I predicted subjective experience intensity (Spearman ρ = 0.41, p = 0.016, surrogate-validated). Five alternative rate-of-change measures (coherence alone, complexity alone, broadband power, alpha power, aperiodic slope) all failed. The formula is specific and non-replaceable. Separately, EEG microstate spatial configuration (PAT_PC1) predicted visual content type (complex vs elementary imagery; ρ = 0.56, p < 0.001), while being orthogonal to intensity. The two measures formed a pre-specified double dissociation: each predicted its target and failed to predict the other’s (Steiger’s test: p = 0.007 for the pattern arm). The dissociation did not generalise to psilocybin (N = 57, all tests null).

Conclusion. mac_I is a formula-specific, surrogate-validated predictor of DMT experience intensity that dissociates from spatial content encoding. The double dissociation provides within-dataset evidence that intensity and visual content type are separable neural dimensions under DMT. This finding requires independent confirmatory replication.

Keywords: consciousness, DMT, EEG, integration, complexity, microstates, psychedelic, double dissociation, mac_I