Combined DMT-harmine formulation reduces negative self-referential emotions during social self-evaluation: a randomized placebo-controlled trial in healthy volunteers

Negative self-referential emotions such as embarrassment and shame are important features of several psychiatric disorders, including depression, social anxiety and trauma-related conditions, but existing treatments do not specifically target these processes well. The paper frames psychedelics, especially ayahuasca-related compounds, as potentially useful because they may alter rigid self-focused processing, increase self-compassion and reduce maladaptive self-criticism. The authors note that although earlier work suggests psychedelic effects on social cognition, emotional empathy and self-referential processing, it remains unclear whether these compounds can acutely reduce embarrassment and shame in a controlled, socially evaluative setting. Aicher and colleagues therefore aimed to test whether an ayahuasca-inspired formulation containing harmine and DMT would reduce negative self-referential emotions during a karaoke-based social self-evaluation task, and whether harmine alone would have similar effects. They hypothesised that HAR/DMT, but not harmine alone, would reduce embarrassment and shame relative to placebo, particularly when participants heard recordings of their own singing. The study was presented as part of a broader investigation of self-referential processing in the same sample.

The researchers conducted a randomised, double-blind, placebo-controlled crossover trial in healthy adult men. Although 36 volunteers were recruited, five withdrew and three were excluded because of technical recording problems, leaving 28 participants in the analysed sample. Participants were 20-40 years old, right-handed, fluent in German, with normal hearing, BMI between 18.5 and 30, minimal psychedelic experience, and no current or past psychiatric or neurological disorders. Female participants were excluded because of concerns about hormonal influences on endocrine and emotional measures. Each participant completed three intervention days separated by at least two weeks. Using a Latin square design, they received HAR/DMT, harmine alone, and placebo in randomised order. Randomisation was computer-generated by an independent researcher, and blinding was maintained with identical opaque containers and matched placebos. Harmine was given as a 100 mg orodispersible tablet for buccal delivery. DMT was delivered intranasally in repeated 10 mg increments to reach 100 mg in total, starting 30 minutes after harmine premedication; two 10 mg doses were skipped in the HAR/DMT condition because of strong acute effects, resulting in a 90 mg cumulative DMT dose in those instances. Participants were monitored throughout the day, with drug and alcohol screening, breakfast, baseline ratings, safety checks and blood sampling. The main behavioural task was an adapted karaoke paradigm designed to induce social self-evaluation. Participants selected songs, recorded their own singing, and later listened to three stimulus types: their own singing (Self), a confederate version (Other Bad) and original artist recordings (Other Good). After each stimulus sequence, they rated pleasantness, unpleasantness, embarrassment, funniness and perceived objective quality on visual analogue scales. After completing the task, they also completed the German version of the Experiential Shame Scale, which assessed state shame across emotional, physical and social subscales. Data were analysed in R using mixed-model ANOVA with drug condition and karaoke category as factors for the visual analogue scale outcomes, and drug condition for the Experiential Shame Scale outcomes. Pairwise contrasts were adjusted using the Tukey method, degrees of freedom were estimated with the Kenward-Roger method, and the significance threshold was p=0.05. Planned simple-effects contrasts comparing HAR/DMT, harmine and placebo were reported regardless of the omnibus interaction.

The karaoke task reliably elicited different emotional responses across stimulus types. There was a strong main effect of karaoke condition on all visual analogue scale ratings, including pleasantness, embarrassment, funniness, unpleasantness and objective quality, with all p-values below .001. There was also a significant main effect of drug condition for unpleasantness and objective quality, but not for pleasantness, embarrassment or funniness. The interaction between drug condition and karaoke category was significant for pleasantness and unpleasantness, but not for embarrassment, funniness or objective quality. For the main shame measure, the Experiential Shame Scale, drug condition significantly affected the global score and the emotional subscale, but not the physical or social subscales. The paper reports that HAR/DMT, compared with placebo, significantly reduced embarrassment in the Self condition, with an estimated marginal mean difference of -16.83 and p=.001. This effect was not present for the Other Bad or Other Good conditions. HAR/DMT versus harmine, and harmine versus placebo, did not significantly differ for embarrassment in any condition. In addition, HAR/DMT reduced the Experiential Shame Scale global score (F=4.26, p=.019) and emotional subscale (F=5.11, p=.009), while the physical and social subscales were not significant. Harmine alone did not show significant effects on the shame scale. The authors also note that HAR/DMT increased pleasantness and lowered unpleasantness in some comparisons, and that perceived objective quality of singing was rated more favourably under the active condition across all stimulus types. They report that funniness ratings were unchanged, which they interpret as suggesting that the findings were not simply due to nonspecific mood elevation or emotional detachment.

The authors interpret the findings as evidence that the combined harmine and DMT formulation can acutely reduce negative self-referential emotions in a socially evaluative context. They argue that the karaoke task successfully induced embarrassment and shame in response to hearing one’s own singing, and that HAR/DMT specifically reduced self-related unpleasantness, embarrassment and the emotional aspects of shame compared with placebo, whereas harmine alone did not produce the same pattern. They also emphasise that unchanged funniness ratings argue against a simple explanation based on general mood elevation. Aicher and colleagues compare their findings with earlier research showing that psychedelics can alter social cognition, increase emotional empathy and prosociality, and reduce negative affective bias. They suggest that the observed effects may reflect a broader shift towards more positive social processing, possibly involving reduced threat perception, decentering, self-compassion and altered self-other differentiation. They also discuss the possibility that changes in perceived singing quality may indicate a general positivity bias or altered auditory/aesthetic perception, although they argue that this does not fully explain the selective reduction in embarrassment. The authors consider the pattern of results on the shame scale to indicate that HAR/DMT may mainly reduce the internal emotional distress of shame rather than its social or bodily components, although they acknowledge alternative explanations such as altered embodiment or interoception. They also describe harmine as showing an intermediate, subtle pattern in some measures, but not a robust independent effect. Neurobiologically, they propose that modulation of the default mode network and related self-referential circuitry could underlie the behavioural effects, though this was not directly measured in the study. The main limitations they acknowledge are the exclusively healthy male sample, which restricts generalisability; possible expectancy effects and incomplete blinding; the retrospective administration of the shame scale after the full task; the lack of a formal blinding check and objective biomarkers; the absence of a prospectively powered analysis plan specific to this outcome; and the fact that the karaoke paradigm has not been formally validated against standard social-stress tasks. They also note the possibility of residual carryover or habituation effects despite the crossover and Latin square design. Clinically, they suggest that acute modulation of negative self-referential affect may be relevant to depression, social anxiety and trauma-related conditions, and that the formulation could potentially complement therapies such as Compassion-Focused Therapy or Acceptance and Commitment Therapy, although they stress that structured support and careful ethical consideration would be necessary.

The authors conclude that HAR/DMT reduced negative self-referential emotion in a controlled social-evaluative setting. They suggest that the paradigm may be useful for probing broader self-referential emotions such as self-consciousness, pride and self-compassion, and they present HAR/DMT as a promising tool for studying and potentially addressing maladaptive self-referential processing. They state that further studies in clinical populations are needed.