Low-Dose LSD Alters Early and Late Event-Related Potentials to Emotional Faces

Haggarty and colleagues situate this study within renewed scientific interest in LSD and growing popular use of very low ‘‘micro’’ doses, which proponents claim improve mood and cognition. Prior controlled studies have shown that single low doses (up to the doses used here) produce only mild subjective effects, yet imaging and connectivity studies suggest low doses can alter prefrontal–limbic and reward circuitry. The authors note that face‑processing paradigms and event‑related potentials (ERPs) such as the N170 and P300 provide a well‑established window onto how drugs alter social and emotional information processing, and that previous work with full and subhallucinogenic doses of psychedelics (psilocybin, LSD) has produced inconsistent effects on these ERPs. Mismatch Negativity (MMN) to deviant stimuli is also of interest given prior reports that LSD can blunt MMN to auditory deviants. The present study tested whether a single low sublingual dose of LSD tartrate (26 mg, within‑participant, double‑blind versus placebo) alters the N170, P300 and MMN elicited by angry, happy and neutral human faces presented infrequently in an oddball design with a frequent cartoon face. Based on prior findings and the hypothesis that neutral faces can be perceived as ambiguous or threatening, the investigators predicted that LSD would reduce responses to neutral faces, and additionally that LSD would dampen responses to angry but not happy faces.

Design and participants: The study used a double‑blind, within‑participant design. Healthy adults (initial n = 39) aged 18–35 years completed two 5‑hour sessions (LSD 26 mg sublingual and placebo), separated by at least 7 days. Five participants were excluded from EEG analyses for excessive noise, leaving a final sample of n = 34. Inclusion criteria included at least high school education, BMI 18–26, fluency in English and good physical and mental health; exclusion criteria included use of prescription medications, cardiac disease or hypertension, psychotic disorder, and prior negative experiences with hallucinogens. To reduce risk, participants were required to report at least one previous use of a hallucinogen. Procedure and drug administration: Sessions ran 9:00–14:00. Participants completed breathalyser and urine toxicology and pregnancy tests on arrival, baseline questionnaires and cardiovascular measures, then received 0.4 mL sublingual solution containing LSD tartrate (26 mg) or 0.4 mL distilled water (placebo) under double blind procedures. EEG electrodes were applied and recording began 120 min after administration, with electrodes removed at 180 min. Subjective measures and cardiovascular assessments were collected at preadministration and at 60, 90, 180 and 240 min postadministration. Participants rated the faces for valence and arousal after the EEG task. EEG task and acquisition: A visual emotional oddball task comprised 300 stimuli per session: 80% frequent schematic (cartoon) faces and 20% infrequent human faces (angry, happy, neutral) drawn from NimStim; stimuli were presented in three randomized blocks of 100 to balance vigilance. Subjects indicated with button presses whether a stimulus was a human or cartoon face to maintain attention. EEG was recorded from a 64‑channel net (acquired at 1024 Hz, downsampled to 250 Hz). Offline preprocessing was performed blinded to drug condition: PREP pipeline for line noise removal, PICARD ICA to identify and remove ocular, muscle and other artefacts, filtering (0.1–30 Hz), average referencing and segmentation from −200 to 1000 ms. ERP measures and analysis: The N170 was measured at PO8 as peak amplitude and latency between 150–200 ms (baseline −200 to 0 ms). The P300 was measured at Pz as mean amplitude and peak latency between 300–500 ms. MMN was computed at Pz between 130–180 ms by subtracting frequent (cartoon) from infrequent (human) peaks to confirm differential processing of human faces. Primary statistical tests were repeated‑measures ANOVAs with within‑subject factors Emotion (angry, happy, neutral) and Drug (placebo, LSD) on ERP amplitude and latency. Subjective/physiological outcomes (time points × drug) were analysed with repeated‑measures ANOVA; face valence/arousal ratings were compared by paired t‑tests. The 5D‑ASC questionnaire was analysed with a 5 (dimensions) × 2 (drug) ANOVA. Bonferroni correction was applied for multiple comparisons when no a priori hypothesis existed.

Sample and subjective/physiological effects: The analysed sample comprised 18 men, 17 women and 1 participant of undisclosed gender, mean age 26.1 years. Participants reported a history of psychedelic use (mean 19 lifetime uses, median 8). On end‑of‑session measures participants identified placebo correctly 52% of the time and identified LSD as a hallucinogen 47% of the time. Participants reported greater liking of LSD than placebo (LSD mean = 68.53, SEM = 4.27; placebo mean = 46.54, SEM = 5.49). LSD increased scores on the 5D‑ASC, with a significant Dimension × Drug interaction (F(4,132) = 12.99, p < 0.001, g p2 = 0.28); simple effects showed significant increases under LSD for Oceanic Boundlessness (p < 0.001), Auditory Alterations (p = 0.02), Anxious Ego Dissociation (p = 0.02) and Reduction of Vigilance (p < 0.001), but not for Visual Restructuralization (p = 0.12). Blood pressure was increased under LSD. Face ratings: In the supplementary results, LSD increased positive valence ratings of happy faces but did not affect arousal ratings or ratings for neutral or angry faces. ERP findings — N170 and P300: Under LSD, the peak amplitude of the occipitotemporal N170 at PO8 was significantly reduced for neutral faces compared with placebo; no significant reduction was observed for angry or happy faces. LSD also slowed N170 latency across all face types. For the midline posterior P300 at Pz, LSD significantly reduced mean amplitude for neutral and happy faces but not for angry faces. The reported reductions in N170 and P300 amplitudes to neutral faces were consistent with one of the primary hypotheses, while the P300 reduction to happy faces was not anticipated. MMN and novelty sensitivity: MMN differed across emotions as expected, confirming that human faces elicited different responses than the frequent cartoon stimulus, but LSD had no significant effect on MMN amplitude, indicating no detectable drug‑related change in visual novelty detection in this paradigm. Other analyses: The investigators report that the ERP effects did not appear to be simply explained by ratings of ‘‘liking’’ the drug; no clear association between liking and ERP amplitude reductions is reported in the extracted text. The paper notes that the study was not powered to examine individual differences such as gender or other demographics, and no pharmacokinetic measures were collected.

Haggarty and colleagues interpret their findings as evidence that a single low sublingual dose of LSD (26 mg) alters both early and later stages of cortical processing of facial stimuli: early occipitotemporal responses (N170) were reduced selectively for neutral faces, whereas later posterior P300 responses were reduced for neutral and happy faces. The authors suggest these changes may reflect reduced subjective salience or altered valence assignment to some facial expressions under low‑dose LSD, and they link the N170 changes to putative modulation of fusiform face area activity and the P300 changes to altered allocation of attentional or working memory resources. The results are positioned relative to prior studies in which full and low doses of psychedelic drugs elicited reductions in face‑responsive ERPs; the present pattern—stronger effects on neutral than on angry faces and a P300 reduction to happy faces—partially aligns with earlier observations but also diverges in important ways. The absence of an LSD effect on MMN for visual deviants contrasts with a previous report of blunted auditory MMN under LSD, and the authors note methodological differences (visual versus auditory paradigms) that could account for this discrepancy. They also point out that slowed N170 latency under LSD may relate to altered visual processing complexity rather than emotion‑specific effects. The investigators acknowledge several limitations: the limited spatial resolution of EEG constrains inference about deep sources and network mechanisms; the sample size, although adequate for primary drug effects, was insufficient for analyses of individual differences; and no pharmacokinetic measures (for example plasma levels or CYP2D6 genotypes) were obtained, which could contribute to between‑subject variance. They also note that participants were partially able to detect drug effects (less than half labelled LSD correctly), meaning expectancy cannot be ruled out entirely despite steps taken to minimise it. Finally, the authors emphasise that the relationship between subjective experience and ERP changes was not straightforward in their data and warrants further study. Implications and future directions described by the authors include testing whether these ERP effects generalise to other 5‑HT2A agonists, whether they can be blocked by antagonists, and whether the neural changes predict clinical benefit. The investigators propose that EEG measures at low doses may serve as biomarkers in mechanistic or therapeutic studies of psychedelics and as tools to probe serotonin’s role in affective processing.

The authors conclude that low sublingual doses of LSD attenuate the amplitude of N170 and P300 ERPs evoked by emotional face stimuli, providing preliminary evidence that subhallucinogenic doses can alter early and later stages of social‑emotional processing. They recommend further research to confirm these effects across other 5‑HT2A agonists, to test receptor specificity with antagonists, and to determine whether such neural changes relate to potential clinical benefits. EEG responses to low‑dose LSD are presented as candidate biomarkers for future therapeutic and mechanistic investigations.