DMT/Harmine Retreat Study Reveals Network-Specific Brain Effects

Subacute Effects of DMT/Harmine Are Network- and Timescale-Specific

A randomized, double-blind, placebo-controlled trial published on November 11, 2026 (OpenAlex W7225800875) provides new evidence that the subacute, or "afterglow," effects of N,N-dimethyltryptamine (DMT) combined with harmine are not global but instead are small, heterogeneous, and specific to certain brain networks and timescales. The study involved 39 participants (19 receiving DMT/harmine, 20 placebo) at a three-day meditation retreat, with resting-state functional MRI (rs-fMRI) scans conducted before and after the intervention. Researchers used connectome-based modeling to estimate changes in brain network control energy (CE)—a measure of how much effort is required for the brain to transition between different functional states.

Mechanisms: Control Energy, Timescales, and Network Dynamics

The brain's control energy landscape, as modeled in this study, reflects the accessibility of different brain states, where lower CE indicates more easily reachable states. Acute psychedelic effects are theorized to flatten this landscape, broadening the range of accessible states. However, this trial found no global shift in CE distribution post-retreat. Instead, effects were subtle and emerged only when examining specific networks under timescale-sensitive approaches.

• At longer timescales, the central executive network (CEN) showed a statistically significant increase in CE in the DMT/harmine group, suggesting increased effort to stabilize this network after the retreat.
• The default mode network (DMN) also showed a CE increase in both groups, but these changes did not survive correction for multiple comparisons.
• At shorter timescales, the visual network (VN) exhibited a significant decrease in CE within the DMT/harmine group, potentially reflecting reduced dynamical demand in a region acutely activated by DMT.

Importantly, these network-specific effects were only detectable when the analysis matched the network's intrinsic timescale with the modeling approach, highlighting a methodological insight: evaluating psychedelic effects through a single dynamical lens may obscure subtle, network-specific changes.

Implications for Psychedelic-Assisted Therapy and Research

The findings have direct relevance for the design and interpretation of psychedelic-assisted therapy protocols. The modest, network- and timescale-specific afterglow effects suggest that the therapeutic window following DMT/harmine administration may not be characterized by a broad, persistent increase in brain flexibility as sometimes hypothesized. Instead, lingering changes may be confined to particular networks and depend on the dynamical context. This nuance could inform the timing of integration sessions, safety monitoring, and the development of biomarkers for treatment durability.

From a research perspective, the study underscores the importance of using connectome-based and timescale-sensitive analyses to capture the full complexity of psychedelic effects. The authors note that previous reports of reduced within-network connectivity after psilocybin may be related to the increased CE observed here in the CEN, suggesting a possible cross-compound mechanism that warrants further investigation. Notably, the trial’s modest sample size and exploratory design mean these findings should be interpreted as hypothesis-generating rather than definitive.

Risks, Limitations, and Unknowns

While the study provides valuable mechanistic insights, several limitations temper its conclusions. The sample size (N = 39) limits statistical power, and the use of a meditation retreat as the setting introduces potential confounders related to environmental and psychological factors. The observed effects, though statistically significant in some cases, were small and did not always survive correction for multiple comparisons. Moreover, the relationship between control energy changes and clinical outcomes remains unclear; increased CE in the CEN, for example, could reflect either adaptive or maladaptive processes depending on context.

A non-obvious risk highlighted by the study is the potential for subtle, network-specific changes to be overlooked in larger clinical trials if analyses are not tailored to the relevant timescales and networks. This could lead to underestimation of both beneficial and adverse subacute effects, with implications for safety monitoring and regulatory evaluation.

Looking Forward: Toward Larger and More Nuanced Trials

The current findings point to the need for larger, multi-site trials with refined neuroimaging protocols that account for network-specific and timescale-dependent effects. As the field moves toward clinical translation, regulatory agencies and trial designers may need to consider more sophisticated endpoints beyond global brain metrics. The study also suggests that future work should examine how these network-level changes relate to subjective experience and long-term outcomes, and whether similar patterns are observed across different psychedelic compounds and retreat formats.

Reviewed by Dr. Alex M. Kessler, PhD (Neuroscience), November 2026. How we research: All findings are sourced from the original peer-reviewed article and trial registry; analysis is independently verified by credentialed experts.