Neuroplastic white matter changes in patients with major depression following lysergic acid diethylamide treatment

Major depressive disorder is common, disabling, and often insufficiently relieved by standard antidepressant treatment. The introduction argues that depression may involve maladaptive neuroplasticity and notes that several treatments, including antidepressants, electroconvulsive therapy, ketamine, and psychedelics, have been linked to neuroplastic changes. However, the authors emphasise that evidence for in vivo neuroplastic changes in humans remains limited, particularly for white matter microstructure, and that it is unclear whether LSD produces measurable structural changes that relate to clinical benefit. Against this background, Avram and colleagues set out to test whether LSD treatment in patients with major depression alters white matter microstructure as measured by diffusion tensor imaging. Using data from a recent clinical trial comparing two low doses with two moderate-to-high doses of LSD, they hypothesised that the higher-dose condition would be associated with increased fractional anisotropy, decreased mean diffusivity, and a relationship between white matter changes and symptom improvement. The paper aims to help clarify a possible biological mechanism underlying LSD’s antidepressant effects and to extend human evidence on psychedelic-related neuroplasticity.

This study used diffusion tensor imaging data from a clinical trial in patients with major depressive disorder who received two low doses of LSD or two moderate-to-high doses. The extracted text indicates that the final diffusion analysis sample included 35 participants after quality checks, with 17 in the high-dose group and 18 in the low-dose group. Three participants were excluded because of corrupted imaging data, including motion or distortion artefacts. The groups were similar in age and sex, but the high-dose group had more severe depressive symptoms at baseline. The intervention compared two low doses of LSD, reported in the Discussion as 25 μg, with two moderate-to-high doses, reported as 100/200 μg. The study assessed depression symptom severity with the Inventory of Depressive Symptomatology Clinician-Rated and Self-Report scales and the Beck Depression Inventory. The primary clinical endpoint was 2 weeks after the second intervention, with follow-up assessments at 6 and 12 weeks after intervention. The DTI scans were used to evaluate white matter microstructure before and after treatment, and the researchers also examined subjective psychedelic effects using the oceanic boundlessness dimension from the 5-Dimensional Altered States of Consciousness questionnaire and the total MEQ30 score. Imaging data were preprocessed in FSL, including susceptibility distortion correction, eddy current and motion correction, skull stripping, and tensor fitting to derive fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity maps. Voxel-wise analyses used tract-based spatial statistics, with a white matter skeleton thresholded at FA > 0.2. Group-by-time effects were tested with permutation-based two-sample t tests using 5000 permutations and threshold-free cluster enhancement with family-wise error correction at p < 0.05. The researchers also performed control analyses for baseline white matter differences, age, sex, and timing between treatment and scan, and used Pearson correlations, partial correlations, and a multivariate analysis of covariance to examine relationships between imaging changes and clinical outcomes.

After quality control, diffusion data were available for 35 patients with major depressive disorder. The groups did not differ significantly in age or sex, but the high-dose LSD group began with more severe depression on both the IDS and BDI. At the primary endpoint, 2 weeks after the second intervention, the high-dose group showed a greater baseline-adjusted reduction in depressive symptoms than the low-dose group on both IDS and BDI. For white matter outcomes, there were no significant group-by-time interactions for mean diffusivity, axial diffusivity, or radial diffusivity. In contrast, fractional anisotropy showed a significant group-by-time interaction after multiple-comparison correction, with a peak t value of 5.59. The regions showing larger FA increases in the high-dose group included the internal capsule, external capsule, sagittal stratum, and fornix/stria terminalis. The extracted FA change in this cluster differed substantially between groups, with a large effect size (t33 = -4.37, p < 0.001, Cohen’s d = 1.48). Individual trajectories generally showed increases in FA in the high-dose group, whereas the low-dose group showed more variable and often minimal change. Control analyses suggested that the low-dose group had slightly higher baseline FA in the affected regions, but there was no whole-brain baseline FA difference between groups. Adjusting for age and sex did not alter the FA findings. The two groups did not differ in time from the second intervention to the second scan, although there was a trend suggesting that longer time to scan might be associated with higher FA across groups. Clinical scores fell in both groups numerically, but adjusted analyses showed significantly greater symptom reduction in the high-dose group at the primary endpoint and at both follow-ups, with similar findings on IDS-C, IDS-SR, and BDI. Across participants, post-intervention FA in the significant white matter regions correlated with symptom improvement in the high-dose group only. These associations were seen at the primary endpoint and remained present at the final follow-up, particularly after adjusting for age and sex. In the high-dose group, post-intervention FA correlated with better IDS-C, IDS-SR, and BDI outcomes, whereas the low-dose group showed no significant relationship. Subjective experience measures, including oceanic boundlessness and MEQ30 total score, were not correlated with post-intervention FA. In regression analyses, oceanic boundlessness predicted symptom improvement in the full sample, whereas FA was not significant; however, in the high-dose group, post-intervention FA was a significant predictor of clinical outcome and subjective experience was only a trend. The multivariate analysis also indicated that LSD dose significantly influenced both symptom reduction and post-treatment FA.

The authors interpret the study as evidence that two moderate-to-high doses of LSD, but not two low doses, are associated with increased white matter fractional anisotropy in patients with major depression. They describe the affected tracts as subcortical pathways involved in emotion regulation, memory, and executive function, and suggest that the FA increase may reflect a normalisation or restoration of white matter integrity in regions often altered in depression. They also argue that the dose-dependent divergence between groups points to a threshold for measurable plasticity rather than a simple linear dose-response relationship. Avram and colleagues place their findings alongside earlier work showing that other antidepressant interventions, such as paroxetine, electroconvulsive therapy, and repetitive transcranial magnetic stimulation, can alter white matter microstructure, although results across those studies have been mixed. They also note that the human psychedelic imaging literature is still limited and cite prior reports of structural changes after ayahuasca and psilocybin, which they contrast with their own longitudinal findings in depressed patients. The authors connect the imaging results with animal work suggesting psychedelic-induced neuroplasticity, including changes in dendritic spines, synaptic density, and stress-related structural recovery. The paper emphasises that the relationship between post-treatment FA and symptom improvement was specific to the higher-dose LSD group and remained evident at follow-up, supporting the idea that structural white matter change may be linked to the durability of clinical benefit. At the same time, the authors caution that acute subjective psychedelic experience also predicted outcome in the full sample, so experiential and biological mechanisms may both contribute, with their relative importance varying by dose. Several limitations are highlighted. The sample was small, especially in the high-dose subgroup, which limits generalisability and may reduce power to detect subtler effects. Baseline symptom severity differed between groups, raising the possibility of regression to the mean and meaning that the findings should be treated as associations rather than causal proof. The authors also note the absence of a placebo arm, the variability in time between treatment and scanning, the multiple DTI metrics examined without correction across all metrics, and the inherent limitations of fractional anisotropy and tract-based spatial statistics in identifying specific cellular mechanisms. They conclude that future studies should use larger samples, standardised longitudinal imaging schedules, and complementary imaging methods to clarify whether the observed FA changes reflect direct pharmacological effects, symptom-related changes, or both. The discussion further suggests that the findings may have clinical relevance because they indicate a possible structural neuroplastic mechanism for LSD’s antidepressant effects and a potential biomarker of response.