Inter-individual variability in neural response to low doses of LSD

Self-reported ‘‘microdosing’’ with psychedelics such as LSD and psilocybin to boost mood, cognition and mental health has become increasingly common, yet placebo-controlled studies have produced inconsistent results. The introduction frames several contributors to this inconsistency, including variability in dose, pharmacokinetics, genetic differences, personality and context, and highlights that baseline arousal may be an important moderator. Resting-state EEG power in low-frequency bands (delta, theta, alpha) and event-related potentials (ERPs) such as mismatch negativity (MMN) are presented as objective measures of arousal, attention and pre-attentive processing that can reveal subtle drug effects that behavioural measures might miss. Hutten and colleagues set out to test whether neurophysiological responses to repeated low doses of LSD (15 mcg) depend on individuals’ baseline cognitive state. The study targeted three domains assessed with EEG: resting-state oscillatory power (arousal), pre-attentive processing using a roving auditory oddball task (MMN, P3a), and perceptual learning/memory using a visual long-term potentiation (LTP) paradigm. Based on earlier low-dose work, the authors expected reductions in low-frequency EEG power reflecting increased arousal and hypothesised that stimulatory effects would be largest in participants with high baseline low-frequency power (interpreted as low arousal). No firm directional predictions were made for ERP measures, but baseline-dependent effects were anticipated.

This double-blind, placebo-controlled, between-subjects trial randomised healthy volunteers to receive either four oral doses of LSD base (15 mcg) or four matched placebo administrations over two weeks. Baseline and follow-up visits were scheduled one week before and one week after the two-week dosing period. EEG assessments were collected at baseline, after dosing session 1, after dosing session 4, and at the one-week follow-up. Participants completed a 4-hour training session before study entry and underwent medical screening; the extracted text states that inclusion/exclusion criteria are reported in the Supplementary Materials. EEG protocols comprised resting-state recordings with eyes open (EO) and eyes closed (EC) for 2.5 minutes each, a roving auditory oddball task to probe MMN and P3a (tones presented in trains with deviant vs standard positions), and a visual LTP paradigm using vertical and horizontal sine gratings to index stimulus-locked N1 and P200 components across pre-potentiation, early-post and late-post (≈30 min) conditions. Primary EEG outcomes were oscillatory power in five bands (delta 1–4 Hz, theta 4–8 Hz, alpha 8–13 Hz, beta 13–30 Hz, gamma 30–45 Hz) across a set of frontal, central and parietal electrodes, MMN/P3a amplitude and latency across the same electrodes, and N1/P200 amplitudes and latencies at posterior electrodes for the LTP task. Blood samples for LSD and O-H-LSD concentrations were taken 2 h after dose 1 and dose 4. Participants were also asked to guess treatment allocation at the end of dosing days. Data were analysed using linear mixed models (LMMs) with restricted maximum likelihood estimation and an unstructured covariance matrix, implemented in SPSS. Models included fixed effects for Treatment, Test day (where baseline differences were absent models used Test day levels dose 1, dose 4, follow-up), Electrode, and relevant interactions; Time (pre, early-post, late-post) was added for LTP analyses. Pairwise comparisons were Bonferroni corrected where interactions were significant. Correlational analyses (Bonferroni corrected) assessed associations between baseline EEG/ERP measures and treatment-induced changes; Pearson correlations were used for normally distributed data and Kendall’s Tau-b for non-normal data. Missing data were handled via listwise deletion. Sample size planning used a G*Power calculation targeting small effects (f = 0.25), 80% power and alpha = 0.05.

Of 53 randomised participants, six were removed from analyses (two dropped out due to COVID restrictions, one for non-adherence, three excluded for technical or noisy EEG), leaving 47 participants for the reported EEG analyses. Resting-state EEG showed no baseline differences between treatment groups. LMMs revealed significant main effects of Treatment and Treatment by Test day interactions across low-frequency bands: compared to placebo, LSD reduced delta and theta power in both EO and EC conditions during dosing sessions 1 and 4 (all p < 0.001 for many contrasts) and reduced alpha power in EO on dosing days (p < 0.001). Delta power remained lower in the LSD group at the one-week follow-up (EO p = 0.002; EC p = 0.008). Some alpha differences in EC appeared as a group offset rather than an acute treatment effect. Gamma power (EC) was higher on dosing session 1 under LSD (p = 0.015), while EO gamma did not differ by treatment. Within the LSD group, higher baseline delta, theta and alpha power correlated with larger acute decreases in those bands (direction and significance reported in the extracted text), consistent with stronger stimulatory effects in participants with higher baseline low-frequency power. In the roving auditory oddball task, the paradigm produced robust deviant–standard differences for MMN and P3a across conditions. At baseline there were minimal treatment differences, although one electrode showed earlier MMN in the LSD group. Across dosing days, LMMs indicated a Treatment by Test day interaction for MMN amplitude (F 2,381.67 = 7.05; p = 0.001) and a main effect of Treatment for P3a amplitude (F 1,273.75 = 21.63; p < 0.001). Latencies for both MMN and P3a were shorter (earlier) in the LSD condition (MMN F 1,407.58 = 35.9; p < 0.001; P3a F 1,401.91 = 26.81; p < 0.001). Pairwise tests showed that MMN amplitude was significantly less negative (i.e. larger) in the LSD condition at follow-up (p = 0.005), and P3a amplitude was more positive under LSD across test days. Correlational analyses linked larger LSD-induced changes in MMN/P3a to poorer baseline pre-attentive performance; some similar baseline–change associations were also observed in the placebo group, which the authors note may reflect practice or other non-treatment factors. Habituation/repetition-suppression profiles differed subtly between groups across sessions (details reported in Supplementary Materials). Visual LTP analyses found significant potentiation across electrodes at baseline in both groups. Input specificity (tetanized vs non-tetanized stimulus) did not reliably influence N1 or P200, so analyses used averaged stimuli. A Treatment × Test day × Time interaction for N1 amplitude indicated that at dosing session 4 the placebo group showed significant early and late potentiation (more negative N1) relative to pre-potentiation (p < 0.013) whereas the LSD group did not. For P200, a Treatment by Test day interaction was present (F 1,1084.31 = 27.36; p < 0.001) with lower P200 amplitude on dosing session 4 in the LSD group compared to placebo (p = 0.001); follow-up values converged. Importantly, baseline P200 potentiation strongly predicted LSD-induced reductions in P200 at dose 4 within the LSD group (r(21) = -0.75, p < 0.001), indicating larger inhibitory effects in participants with greater baseline LTP. No significant correlations were reported for N1. Treatment blinding appeared adequate: the distribution of correct versus incorrect/inconsistent treatment guesses did not differ from chance in either placebo or LSD groups. Mean (SE) plasma LSD concentrations 2 h after doses 1 and 4 were 302 (105) pg/mL and 326 (117) pg/mL, respectively; O‑H‑LSD concentrations averaged 17 (7.2) pg/mL and 17 (4.9) pg/mL.

Hutten and colleagues interpret the pattern of results as showing that repeated low doses of LSD produce acute stimulatory neural effects, namely reductions in resting-state low-frequency EEG power and earlier, larger indices of pre-attentive processing, and that these effects are moderated by baseline neurophysiological state. Specifically, individuals with higher baseline low-frequency power (interpreted as lower arousal) showed the largest reductions in delta/theta/alpha power after LSD, consistent with a baseline-dependent stimulant-like action. The authors note that reductions in low-frequency EEG power are associated with increased wakefulness and have been observed after other stimulants and after higher and lower doses of psychedelics. Potential mechanisms advanced by the authors include dopaminergic receptor modulation and stress hormone (cortisol) release for acute stimulatory effects, although they acknowledge these mechanisms are speculative and unlikely to fully explain the sustained decrease in delta power observed at one-week follow-up. For such longer-lasting changes the authors propose persisting neuroplastic or immune-related adaptations that have been reported after single psychedelic doses. Regarding pre-attentive processing, the study found accelerated MMN and P3a latencies and increased P3a amplitude under LSD, with MMN amplitude increases at follow-up; the authors interpret this as improved novelty detection and pre-attentive processing under low-dose LSD in healthy volunteers. They caution that some baseline–change correlations also occurred in the placebo group, so practice or other non-specific factors may contribute; nevertheless, the baseline-dependent association for P3a amplitude was specific to LSD, supporting an interaction with drug state. These results contrast with prior studies using higher doses, where MMN is often blunted; the authors suggest dose and task demands (sensory versus affective-cognitive oddball paradigms) may explain divergent findings. On perceptual learning, LSD reduced visual LTP (notably P200 potentiation) at the fourth dosing session, and individuals with stronger baseline LTP showed the largest impairments. The authors link this to previously reported acute memory impairments at moderate-to-high psychedelic doses and discuss potential shifts in glutamatergic/GABAergic balance as a mechanism; they also observe that evidence for post-acute increases in neuroplasticity (for example BDNF rises) does not appear to translate into increased LTP in their sensory paradigm at this low dose. The paper highlights two principal contributions: first, neural effects of low-dose LSD vary between individuals and depend on baseline cognitive/neurophysiological state; second, some neural changes (reduced delta power and increased oddball ERPs) persisted up to one week after dosing, suggesting possible lasting neuroadaptations. The authors discuss clinical implications cautiously, proposing that disorders characterised by elevated low-frequency EEG power such as ADHD or OCD could, in theory, benefit from interventions that reduce low-frequency power, but they emphasise that randomised clinical trials in patient populations are needed. Key limitations acknowledged include the two-week treatment window (not addressing long-term microdosing schedules), the healthy volunteer sample (limiting generalisability to patients), and the general concern about treatment unblinding—although in this study treatment guesses did not exceed chance. Overall, the authors conclude that low doses of LSD can modulate arousal, pre-attentive processing and perceptual learning in a baseline-dependent manner, with some effects persisting beyond the dosing period.