Effects of LSD and Psilocybin on Heart Rate in Patients Receiving Psychedelic Treatment for Depressive and Anxiety Disorders: A Retrospective Observational Study

The authors situate the study within a renewed clinical interest in classic psychedelics (LSD and psilocybin) as adjuncts to psychotherapy for conditions such as treatment‑resistant depression and anxiety. Prior experimental work in healthy volunteers shows that both substances produce mild, transient autonomic activation (including raised heart rate) that is dose dependent and typically resolves within hours. The introduction emphasises that clinical patients differ from healthy research participants (comorbidities, concomitant medications, emotionally intense therapeutic sessions), and that subjective anxiety during sessions may itself influence cardiovascular measures; therefore, real‑world data on heart rate dynamics in treated patients are needed to inform safety and clinical practice. M. and colleagues state that the present work aims to characterise heart rate trajectories during supervised LSD‑ or psilocybin‑assisted psychotherapy in a clinical, naturalistic sample. The primary focus is on temporal patterns of heart rate across seven measurements from 30 to 300 minutes after dosing and on whether these trajectories differ between LSD and psilocybin, accounting for age and session anxiety.

This was a retrospective, single‑centre observational study using routinely collected data from the Psychedelic Program at Geneva University Hospital. All patients had received compassionate‑use authorisation and gave written consent for treatment and data use. The eligible clinical population comprised adults with treatment‑resistant major depressive disorder or an anxiety disorder; exclusion criteria included active psychosis or bipolar disorder, imminent suicide risk, severe cardiovascular disease, pregnancy and other contraindications. Patients could continue many concomitant psychotropic medications, although some agents (e.g. antipsychotics, stimulants) were stopped before dosing. The extracted text reports 30 participants (mean age 51.56 ± 12.19 years; range 22–80), 15 female and 15 male; 17 received psilocybin and 13 received LSD. Interventions were administered in an outpatient therapeutic setting with psychiatric supervision. Reported dose ranges were LSD 100–200 µg and psilocybin 15–30 mg in the procedures section, although the Discussion elsewhere refers to psilocybin doses as 15–25 mg (the extraction contains this inconsistency). Sessions began at 09:00 with pre‑dose vital signs taken while seated; the first in‑session physiological measurement was at 30 minutes post‑dose. Trained psychiatric nurses continuously supervised sessions, and standard safety protocols were in place. Outcome measurement focused on heart rate recorded at seven time points between 30 and 300 minutes post‑administration. Subjective anxiety was measured at the same timepoints using a single‑item 0–100 visual analogue scale (VAS), chosen for feasibility during altered states. Baseline (pre‑dose) anxiety and a systematic pre‑dose heart‑rate baseline were not available in the dataset; the 30‑minute measurement was therefore treated as an early quasi‑baseline given evidence that immediate physiological drug effects are minimal within the first 30 minutes. For statistical analysis the researchers fitted linear mixed models (LMMs) with heart rate as the dependent variable. Time was entered as a categorical factor with seven levels (30 min reference) to allow non‑linear trajectories; substance (psilocybin reference) and age were fixed effects, with interactions tested. A random‑intercept model with an AR(1) covariance structure (first‑order autoregression, modelling higher correlation between measurements closer in time) was used to account for repeated measures within participants. A second model added the time‑varying raw VAS anxiety score as a covariate; the authors note this reflects between‑person differences rather than within‑person changes. Analyses were performed in SPSS v28 and statistical significance was defined as p < 0.05.

The analysis included data from the first treatment session of each of 30 patients (17 psilocybin, 13 LSD). The primary LMM found no significant main effect of time (F(6, 77.25) = 0.76, p = 0.60) and no significant main effect of substance (F(1, 30.82) = 0.66, p = 0.42), indicating there was no overall change in heart rate across the sampled timepoints nor an overall difference between the LSD and psilocybin groups. However, there was a significant time × substance interaction (F(6, 77.25) = 3.03, p = 0.01), indicating that heart rate trajectories over time differed between the two drug groups. The three‑way interaction including age (time × substance × age) was not significant (F(14, 62.55) = 1.61, p = 0.10). When anxiety (raw VAS) was added as a time‑varying covariate, the main effects of time (F(6, 79.75) = 1.21, p = 0.31) and substance (F(1, 30.89) = 0.38, p = 0.54) remained non‑significant, while the time × substance interaction remained significant (F(6, 79.75) = 3.29, p = 0.006). The three‑way interaction of time × substance × anxiety was not significant (F(14, 91.88) = 1.35, p = 0.19), leading the authors to conclude that reported momentary anxiety did not materially modify the differential time courses between substances in these models. The estimated participant‑level random intercept variance was 102.73 (SE = 32.31), significantly greater than zero, indicating meaningful individual differences in baseline heart rate. Residual variance was 47.53 (SE = 8.84). The AR(1) autocorrelation parameter was 0.36 (SE = 0.12), also significant, consistent with greater correlation among measurements closer in time. The descriptive pattern reported by the authors (and depicted in a figure referenced in the text) was that LSD was associated with a delayed but sustained increase in heart rate peaking around 3–4 hours post‑dose, whereas psilocybin showed an earlier decline after the initial period. No participant met the study's criteria for tachycardia (>120 bpm) and no clinically significant cardiovascular adverse events were reported. Minor transient symptoms (nausea, dizziness, blurred vision) occurred in some sessions. Heart rate variability (HRV) was not recorded in this dataset.

The authors interpret their findings as preliminary evidence that LSD and psilocybin may produce different temporal patterns of cardiovascular activation in a clinical, naturalistic setting. Although there were no overall main effects of time or substance, the significant time × substance interaction suggests LSD produced a delayed and more sustained elevation of heart rate while psilocybin exhibited a quicker decline. These temporal differences remained after adjustment for age and for between‑person differences in momentary anxiety as measured by the VAS. M. and colleagues relate these observations to known pharmacokinetic and pharmacodynamic differences: LSD has a longer half‑life and a broader receptor profile (the Discussion highlights affinity at receptors including 5‑HT4, D2 and H2) that could plausibly prolong sympathomimetic effects, whereas psilocybin is described as more 5‑HT2A selective with a shorter duration of action and faster recovery of parasympathetic tone. The authors note that prior laboratory studies in healthy volunteers typically show peak cardiovascular effects within 1–2 hours and return to baseline within 6–12 hours; their clinical sample showed a later and more sustained LSD effect compared with psilocybin in routine care. The authors acknowledge several limitations that constrain interpretation and generalisability. Key limitations include the retrospective and observational design, small sample size, and an imbalance in dose ranges between substances (the extraction contains inconsistent reporting of psilocybin dose ranges: 15–30 mg in Procedures but 15–25 mg mentioned elsewhere). Baseline pre‑dose heart rate and anxiety ratings were not systematically collected, so the 30‑minute measurement served as an early quasi‑baseline. Blood pressure was recorded at baseline and not repeatedly in most participants, preventing trajectory modelling for pressure. Anxiety was assessed with a single‑item VAS and was modelled in a way that captures between‑person differences rather than within‑person fluctuations during sessions; therefore anxiety cannot be interpreted as a causal mediator. Objective autonomic markers such as HRV or electrodermal activity were not captured, limiting inferences about sympathetic‑parasympathetic balance. Concomitant medications (four patients on SSRIs, three on benzodiazepines) could have influenced physiological responses but subgroup analyses were not feasible given small numbers. The open‑label, naturalistic context increases external validity but reduces experimental control and prevents causal inference. Given these caveats, the authors present the results as hypothesis‑generating and recommend prospective, dose‑controlled studies with more comprehensive physiological monitoring (including HRV), standardised baseline assessments, validated multi‑item measures of state anxiety (e.g. STAI‑State or dedicated instruments for challenging psychedelic experiences), and larger samples to better separate dose effects from substance effects. They also note the reassuring safety observation that no serious cardiovascular events occurred under supervised clinical care in this small sample, while emphasising the need for more detailed monitoring in future work.