Effects of Psychedelic Microdosing versus Conventional ADHD Medication Use on Emotion Regulation, Empathy, and ADHD Symptoms in Adults with severe ADHD symptoms: A Naturalistic Prospective Comparison Study

Attention deficit hyperactivity disorder (ADHD) is marked by inattention, hyperactivity and impulsivity, and many adults with ADHD also experience difficulties with emotion regulation (ER). ER includes cognitive reappraisal (reinterpreting emotional situations) and expressive suppression (inhibiting emotional expressive behaviour). Previous work indicates that standard ADHD medications improve core ADHD symptoms but have more limited effects on ER and social functioning; research on empathy in adults with ADHD is sparse. Meanwhile, naturalistic reports and some laboratory work on low doses of psychedelics (microdosing, MD) suggest possible benefits for emotional and social functioning, but no prior study had prospectively examined ER and empathy in adults with severe ADHD symptoms using MD. Haijen and colleagues designed two linked prospective naturalistic studies to address this gap. Study 1 followed adults with a diagnosis of ADHD or with severe ADHD symptoms who intended to microdose, measuring ER and empathy at baseline, two weeks and four weeks after starting MD. Because Study 1 lacked a comparator, Study 2 compared a subsample of participants who used MD only with a treatment-as-usual (TAU) control group of people already using conventional ADHD medication, adding measurement of ADHD symptom severity. The overarching aim was to explore whether self-initiated MD is associated with changes in ER, empathy, and ADHD symptoms over four weeks and to provide initial data to inform potential randomised trials.

Both studies used a prospective, naturalistic online design in which adults who intended to begin MD on their own completed surveys at baseline (1–3 days before MD initiation), and at two and four weeks thereafter. Participants were recruited via an online advertisement on microdosinginstitute.com and provided informed consent. The sample included adults with a clinical ADHD diagnosis and adults without a diagnosis but with severe ADHD symptoms; the Conners' Adult ADHD Rating Scale (CAARS-S:SV) T-scores were used at baseline to determine symptom severity, and individuals without an ADHD diagnosis who scored below a T-score of 65 on all CAARS subscales were excluded. Outcome measures comprised ADHD symptom severity (CAARS-S:SV), emotion regulation (Emotion Regulation Questionnaire, ERQ; cognitive reappraisal and expressive suppression subscales) and empathy (Interpersonal Reactivity Index, IRI; perspective-taking, fantasy, empathic concern, personal distress). Daily reports tracked MD intake, substance and dose, although substance/dose information was only reported by about half the sample. Psychiatric and medical comorbidity, demographic information and prior psychedelic use history were collected at baseline. Statistical analyses used linear mixed models (LMMs) with Time (0W, 2W, 4W) as a within-subjects factor. In Study 1 covariates included concurrent medication use and comorbidity; in Study 2 Group (MD-only versus TAU) was a between-subjects factor and comorbidity was included if significant. Separate LMMs were run for each ERQ and IRI subscale; the best-fitting covariance structure was chosen by lowest AIC. Restricted maximum-likelihood estimation handled missing data. Pairwise comparisons were Bonferroni-corrected and the significance level was adjusted to .025 to account for multiple subscales. Effect sizes were reported as partial eta squared (ηp2). Study 2 excluded participants who reported psychedelic use during the study in the TAU group to preserve group contrasts.

Study 1 analytic sample comprised 233 participants at baseline, 64 at two weeks and 44 at four weeks after exclusions for unrealistic responses and insufficient ADHD symptom scores. Linear mixed models indicated changes over time on several outcomes. Cognitive reappraisal (CR) showed a Time effect (F(2,132.0)=4.29, p=.016, ηp2=0.06) with Bonferroni-corrected increases at 2W (Δ=+2.08, p=.013) and 4W (Δ=+2.47, p=.011) versus baseline. Expressive suppression (ES) decreased over time (F(2,123.8)=4.20, p=.017, ηp2=0.06), with a significant reduction at 4W versus baseline (Δ=-2.01, p=.002). For empathy, perspective-taking (PT) increased (F(2,65.9)=5.01, p=.009, ηp2=0.13), with higher scores at 4W compared to baseline (Δ=+1.92, p<.001) and 2W (Δ=+1.12, p=.011). Personal distress (PD) decreased at 2W versus baseline (F(2,122.8)=5.21, p=.007, ηp2=0.08; Δ=-1.09, p=.046). Empathic concern and fantasy showed no significant Time effects. Interactions between Time and concurrent medication use or comorbidity were nonsignificant. Study 2 compared the MD-only subsample (n=180, 50, 38 at 0W/2W/4W) with a TAU group already using conventional ADHD medication (n=37, 27, 28 at 0W/2W/4W). A significant Time by Group interaction for overall ADHD symptom T-scores (F(2,85.8)=10.26, p<.001, ηp2=0.19) indicated the MD group had higher baseline scores but lower scores at 4W compared with TAU. Subscale analyses showed significant Time by Group interactions for hyperactivity/impulsivity (F(2,150.5)=4.81, p≈.000, ηp2=0.06), DSM-IV total symptoms (F(2,173.5)=9.55, p<.001, ηp2=0.10), and ADHD index (F(2,70.6)=17.03, p<.001, ηp2=0.33), with lower T-scores for the MD group at 4W. For ER, no Time by Group interaction was found for CR (F(2,152.9)=1.24, p=.294, ηp2=0.02); a main Time effect was present but pairwise comparisons were non-significant. For ES, a Time by Group interaction was observed (F(2,146.6)=4.58, p=.012, ηp2=0.06), with mean ES scores lower in the MD group compared to TAU at baseline and at 4W. Regarding empathy, no Time by Group interactions were found for fantasy, empathic concern or personal distress. Perspective-taking showed no interaction (F(2,69.7)=1.14, p=.325, ηp2=0.03) but did show a Time effect (F(2,69.7)=3.88, p=.025, ηp2=0.10) with higher scores at 4W versus 0W and a main Group effect (F(1,147.5)=6.97, p=.009, ηp2=0.09) indicating higher PT overall in the MD group. The authors note that only the ES improvement persisted when comparing MD-only participants to the TAU control; the CR and empathy changes seen in Study 1 largely disappeared in Study 2 comparisons. The extraction indicates that MD substance and dose data were incompletely reported by participants and that one participant in the TAU group who used psychedelics during the study was excluded.

Haijen and colleagues interpret the findings as preliminary evidence that self-initiated microdosing over four weeks is associated with reductions in ADHD symptom severity and changes in emotion regulation among adults with severe ADHD symptoms. The most robust ER effect that survived comparison with a treatment-as-usual control was a reduction in expressive suppression, which the authors suggest could reflect improved capacity to manage emotional responses without continual suppression. Improvements in cognitive reappraisal and some empathy subscales were observed in the uncontrolled study (Study 1) but largely did not remain when compared with the TAU group in Study 2, indicating weaker or non-specific effects for those domains. The investigators propose a plausible neurobiological mechanism, noting that psychedelics act primarily at the serotonin 5-HT2A receptor, which is densely expressed in prefrontal regions that regulate amygdala activity; modulation of prefrontal–amygdala circuitry could plausibly underlie improved ER. They also report that, by four weeks, the MD group’s ADHD subscale scores fell below clinical cut-offs on all subscales while the medication group did so on only one subscale, a pattern the authors read as consistent with potential therapeutic effects of MD on ADHD symptoms. Important limitations are emphasised. The naturalistic design leaves the study vulnerable to self-selection bias, high attrition, and expectancy/placebo effects. Lack of experimental control over MD substance, dose, route and contextual factors (intentions, setting, expectations) limits causal inference. Outcome assessment relied solely on self-report measures and participants’ diagnostic status was self-declared rather than clinically verified. Substance/dose reporting was incomplete for many participants, and the study did not follow longer-term psychedelic or other substance use. Given these constraints, the authors state that placebo-controlled randomised trials are required to determine whether observed improvements are specific pharmacological effects of MD or reflect non-specific factors. Finally, they suggest that if MD were to be considered clinically, it should be integrated into a therapeutic framework with trained clinicians, and they note potential transdiagnostic interest given ER and empathy deficits across several psychiatric conditions.