The trial, registered as NCT03866252, was designed as a parallel-group comparison. Patients were randomly assigned to one of two dosing regimens: either two low doses of 25 micrograms of oral LSD, or two high doses, the first at 100 micrograms and the second at 200 micrograms, administered four weeks apart. The researchers, led by Friederike Holze and Kristian Larsen, who contributed equally, with senior investigators including Gitte M. Knudsen, Matthias E. Liechti, Felix Müller, and Patrick M. Fisher, scanned participants during an n-back working memory task on two occasions: before any drug was given, and again approximately one week after the second treatment. This timing matters, because one week after dosing the acute psychedelic effects have long worn off, meaning any differences in brain activity reflect lingering changes in how the brain does its work rather than the drug acting in the moment.

The n-back task is a workhorse of cognitive neuroscience and deserves a brief technical explanation. During the scan, participants view a stream of stimuli and must indicate when the current item matches the one presented one step back, the 1-back condition, or two steps back, the 2-back condition, while a simpler 0-back control condition requires only a response to a predefined target. Increasing the load from 0-back to 2-back places progressively heavier demands on working memory, the mental scratchpad that allows information to be held and manipulated over seconds. By contrasting brain activity during the harder conditions against the easier one, researchers can isolate the neural signature of working memory itself, largely free of the visual and motor processes common to all conditions.

When the team analyzed these task-activated responses using linear mixed effects models, focusing on the treatment-by-time interaction, a clear pattern emerged. Compared with the low-dose group, patients who received the high-dose regimen showed significantly increased working memory-related brain responses, with the difference reaching statistical significance at p equal to 0.007. The effect was not scattered randomly across the brain. It was concentrated precisely within the regions that the task recruits in healthy brains: the dorsolateral prefrontal cortex, a hub for executive control and manipulation of held information; the parietal cortex, which supports attention and the storage side of working memory; the insula, which integrates internal bodily signals with cognition; and subcortical structures including the striatum and thalamus, which gate information flow and support reward and arousal systems.

The direction of this effect came as a surprise to the investigators themselves. The team had hypothesized that high-dose LSD would reduce working memory-related brain responses, a prediction grounded in the idea that psychedelic therapy works by relaxing top-down control and loosening the grip of rigid thought patterns, which might be expected to quiet the very prefrontal and parietal circuits that light up during demanding cognitive tasks. Instead, the opposite occurred. One week after a 200-microgram session, the depressed patients’ working memory networks were more active, not less, when challenged with the same task they had performed before treatment. Contrary to the hypothesis, high-dose LSD significantly increased brain responses within the task-relevant network.

What makes this increased activity potentially meaningful rather than merely noisy is its relationship to behavior. The magnitude of the working memory-related brain responses correlated positively with task performance, suggesting that the extra neural recruitment was not an inefficient compensation or a sign of cortical hyperexcitability, but was coupled to better cognitive output. The authors interpret this as a potential LSD-induced boost in neural function, a phrase that will surely fuel further research into whether psychedelics can enhance, rather than merely reorganize, cognitive processing in clinical populations. The responses also correlated with subjective perceptual changes reported by participants, linking the intensity of the psychedelic experience to the degree of subsequent neural change, although the study design cannot establish that the subjective experience itself caused the brain effects.

Just as telling is what did not correlate. The working memory-related brain responses showed no association with changes in depressive symptom severity. In other words, patients whose mood improved the most were not the same patients whose working memory circuits became most active. The authors conclude that task-related brain activity may underlie different mechanisms than changes in mood, a dissociation with real consequences for how psychedelic therapy is studied and deployed. If the antidepressant effects of LSD operate through one set of neural changes and the cognitive effects through another, then clinical trials that measure only depression scales may miss important dimensions of what these drugs do to the brain, and treatment protocols may eventually need to be tailored depending on whether mood, cognition, or both are the therapeutic target.

The clinical context makes these findings timely. Cognitive dysfunction, including impaired working memory, attention, and executive function, is common in major depressive disorder and often persists even when mood symptoms respond to standard antidepressants. These residual cognitive symptoms are strongly linked to functional impairment, difficulty returning to work, and risk of relapse, yet most conventional treatments offer limited benefit for them. Classic psychedelics such as LSD act primarily as agonists at the serotonin 2A receptor, triggering cascades of downstream changes in glutamate signaling, cortical excitability, and large-scale network dynamics that have been implicated in the plasticity underlying both their acute perceptual effects and their longer-lasting therapeutic potential. Whether those same mechanisms could repair or strengthen cognitive circuits in depression had, until now, been largely a matter of speculation, because imaging studies of psychedelic treatment effects on cognitive brain function in patient populations have been sparse.

The study’s design carries both strengths and caveats worth weighing. The randomized, double-blind, parallel-group structure with an active low-dose comparison group is a rigorous standard for psychedelic research, controlling for expectancy effects that can otherwise inflate apparent benefits. The use of task-based functional MRI provides a mechanistic window that self-report questionnaires cannot. Yet the sample of 43 patients is modest, the follow-up imaging occurred only about one week after the second dose, leaving the durability of the neural changes unknown, and the correlation between brain responses and perceptual changes raises the possibility that residual differences in the psychedelic experience between dose groups contributed to the imaging results. The absence of a placebo-only group also means the findings speak to dose-dependent differences rather than to effects against no treatment at all.

Even with those qualifications, the study marks a genuine milestone. It demonstrates that LSD treatment in patients with major depressive disorder measurably affects cognitive brain function, that the effect is dose-dependent, that it runs counter to the direction many researchers expected, and that it is behaviorally meaningful in the domain of working memory while remaining dissociable from mood improvement. As psychedelic-assisted therapy moves closer to regulatory review for depression, questions are shifting from whether these drugs can lift mood to what exactly they do to the mind’s machinery, and for whom. This trial suggests that the answer will be richer, and stranger, than a simple story of quieted default networks and dissolved egos. The same compound that loosens the boundaries of perception may, a week later, be asking the prefrontal cortex to work a little harder, and a little better, at one of the brain’s most demanding jobs.

Subject of Research: Effects of LSD treatment on working memory-related brain function in major depressive disorder

Article Title: LSD treatment affects working memory-related brain function in patients with major depressive disorder

Article References: Holze, F., Larsen, K., Ozenne, B., Knudsen, G. M., Becker, A. M., Avram, M., Borgwardt, S., Liechti, M. E., Müller, F., & Fisher, P. M. (2026). LSD treatment affects working memory-related brain function in patients with major depressive disorder. Translational Psychiatry. https://doi.org/10.1038/s41398-026-04522-4

Image Credits: AI Generated

DOI: 10.1038/s41398-026-04522-4

Keywords: LSD, major depressive disorder, working memory, functional MRI, psychedelic therapy, dorsolateral prefrontal cortex, n-back task, Translational Psychiatry, cognitive dysfunction, serotonin 2A receptor, clinical trial, neuroimaging