Default Mode Network Modulation by Psychedelics: A Systematic Review

This work was conducted as a systematic review reported according to PRISMA guidance, although no protocol was registered. Electronic searches were run in PubMed and Scopus from database inception up to April 2022. The search strategy was framed using PICO principles, although the extracted text does not include the precise search strings or exact date ranges beyond the April 2022 cut-off. Eligibility criteria restricted inclusion to human clinical studies examining DMN modulation by classical psychedelics; animal studies were excluded. Two reviewers screened articles with disagreements resolved by discussion and arbitration when needed. Only English-language papers were eligible; there were no restrictions on psychedelic dose, treatment duration or statistical approach. The authors included studies that performed novel analyses on previously published datasets. The screening and selection process yielded the set of studies synthesised in the review, and a flowchart of selection was reported (figure referenced but not reproduced in the extracted text).

The combined searches returned 161 records initially, reduced to 119 after duplicate removal; after screening 24 articles met inclusion criteria. The included evidence set comprised 14 psilocybin studies (twelve using fMRI, two using EEG/MEG), four ayahuasca fMRI studies, and seven LSD studies (the extracted text contains an internal inconsistency when listing modality counts for LSD studies; it reports six 3T fMRI studies plus one EEG and one MEG, which numerically exceeds the stated total). Across the psilocybin studies the mean subject age was reported as 34 years (SD 19.3) and the mean sample size 32 (SD 9.3); 10 of the 14 psilocybin studies included a placebo control. For ayahuasca the mean age was 35 years (SD 17.8) with mean sample size 28 (SD 5.9) and one placebo-controlled trial. The LSD studies had a mean reported age of 29 years (SD 2.7) and mean sample size 19 (SD 3.0); the authors state all LSD studies had placebo controls. No published human studies assessing mescaline or DMT effects on the DMN were identified in the search. Synthesis of the included studies revealed a consistent pattern: acute reductions in intra-DMN functional connectivity alongside increases in functional connectivity between the DMN and other resting-state networks (internetwork integration). The review notes specific neurophysiological correlates in electrophysiological studies, including desynchronisation and reduced spectral power in posterior DMN hubs such as the posterior cingulate cortex (PCC). One quantitative figure reported was that, in an analysis of whole-brain connections, 695 of 35,778 possible connections were significantly different from baseline after psychedelic administration, indicating widespread changes in network architecture. Multiple studies linked DMN perturbation to subjective experiences and clinical outcomes. Reduced DMN coactivation correlated with measures of ego dissolution and mystical-type experiences in several reports, and increased measures of brain entropy (greater signal complexity or unpredictability) were associated with those subjective effects. Clinical findings summarised in the extracted text included reductions in depressive symptom measures following psilocybin (reported at 1 week and longer timepoints in some cohorts), with one reanalysis finding decreased brain modularity after psilocybin was associated with improved clinical outcomes at 6 months. Electrophysiological and magnetic imaging studies showed decreased low-frequency power in posterior DMN regions and reductions in blood flow and BOLD signal in DMN hubs during acute drug effects. The authors emphasise that direction of causality remains unresolved: while DMN modulation co-occurs with subjective and clinical effects, it is unclear whether network changes mediate therapeutic benefit or are an epiphenomenon. The review also summarised mechanistic modelling from included studies: 18 of the 24 papers were described as aligning most closely with the REBUS (Relaxed Beliefs Under Psychedelics) model, which links increased brain entropy and relaxed hierarchical priors to therapeutic plasticity. Other models discussed in the results included the Cortico-Striatal-Thalamo-Cortical (CSTC) model and a Cortico-Claustro-Cortical (CCC) model, each with some empirical support reported in individual studies (for example, altered thalamic-to-PCC effective connectivity in support of CSTC; decreased claustrum BOLD signal in support of CCC).

Gattuso and colleagues interpret the body of evidence as showing that classical psychedelics reliably perturb the DMN acutely, typically reducing within-network functional coupling while increasing internetwork connectivity and global integration. They link these network changes to subjective phenomena such as ego dissolution and mystical experiences, and to increased brain entropy, proposing that greater neural flexibility may permit revision of pathological cognitive priors and thereby underlie therapeutic benefit — a formulation consistent with the REBUS model. The authors observe that other mechanistic frameworks (CSTC and CCC) are not mutually exclusive and may jointly contribute to the complex phenomenology and clinical effects of psychedelics. However, the investigators acknowledge several important limitations and uncertainties in the literature. Many studies had small sample sizes, used participants with prior psychedelic experience, or re-analysed previously published small datasets, which limits generalisability and independence of findings. Methodological heterogeneity across studies — differing imaging modalities and field strengths, variable statistical thresholds, variable dosing regimens, retrospective symptom rating in some cases, and occasional lack of placebo controls — complicates synthesis. The authors highlight neuroimaging-specific confounds such as neurovascular coupling (BOLD signal limitations), potential artefacts in source localisation for EEG/MEG, and the problem of reverse inference when mapping brain activation patterns to psychological processes. They also note potential publication bias and encourage the field to treat current models as parsimonious hypotheses rather than established facts. For future research the authors suggest several priorities that they and others have discussed: larger, well-controlled studies (including placebo arms and psychedelic-naïve participants), longitudinal designs with repeated imaging to track acute and enduring changes, exploration of dose–response including micro-dosing and non-hallucinogenic analogues, and combined interventions (for example, psychotherapy or mindfulness) to probe synergistic effects. The review also proposes further comparative testing of REBUS, CSTC and CCC frameworks within the same datasets to develop a more integrated mechanistic account. Clinically, the authors point to the potential of DMN measures as biomarkers and to the prospect of translating DMN-modulation strategies into treatment paradigms, coupled with considerations of regulatory and policy pathways for scheduled substances.

The review concludes that classical psychedelics modulate the Default Mode Network in humans, with consistent acute decreases in intra-DMN connectivity and increases in connectivity between canonical resting-state networks. These neural changes are correlated with ego-dissolution, greater brain entropy and, in several studies, with improvements in mental health outcomes. While the findings support the utility of psychedelics as tools for probing brain–mind relationships and suggest therapeutic potential, the authors emphasise that causality remains to be established and that methodological limitations warrant cautious interpretation. They advocate for further rigorous research to clarify mechanisms, dose effects and clinical applications.