Studies suggest that psilocybin, a classic psychedelic drug (partial agonist of the serotonin 2A or 5-HT receptor), _2A ), may be effective in treating depression and anxiety. ^{1 , 2 , 3} , tobacco use disorder ^4 , 5 and alcohol use disorder ^6 , 7 . It has been shown that the reduction in clinical symptoms lasts up to 3 ³ , 6 ^1 , 2 and 12 ⁸ months after 1 to 3 administrations of psilocybin. Despite these promising advances, the neural and psychological mechanisms underlying the lasting therapeutic effects of psychedelic drugs are not well understood. Two potentially interactive transdiagnostic targets that may be affected by psilocybin are negative affect and brain network plasticity.

Increased negative affect, reduced positive affect, and hypersensitivity to negatively biased information are characteristics of mood disorders. ^{9 , 10 , 11} . Negative affect is also a central component of the addiction cycle, in which the craving and withdrawal symptoms experienced after intoxication lead to preoccupation, anticipation, and readmission of drugs of abuse. ¹² . The amygdala has been shown in clinical and preclinical models to track the importance of stimuli in the environment. ^13 , 14 and is highly responsive to negative emotional stimuli. ^{15 , 16 , 17} . The abnormally high reactivity of the amygdala to negative affective stimuli has been implicated in the pathophysiology of depression. ¹⁸ . Areas within the anterior cingulate cortex (ACC) are known to monitor cognitive conflicts. ^{19 , 20 , 21 , 22} , are involved in the evaluation and expression of negative emotions ²² , respond to levels of suffering associated with pain ²³ and negative social affect ²⁴ , and have been implicated in negative rumination and depression. ²⁵ . Both amygdala and ACC dysfunction have been implicated in the pathophysiology of substance use disorders. ¹² and were specifically implicated in supporting aberrant negative affect in these disorders.

Psychedelic drugs have been shown to acutely reduce the processing of negative affective stimuli. ²⁶ while increasing positive mood in humans ^27 , 28 . In behavioral paradigms, psychedelics have been shown to reduce sensitivity during the encoding of fear faces. ²⁹ Recognition of negative facial expressions. ³⁰ and response to negative stimuli in an emotional inhibition task ²⁷ . Functional magnetic resonance imaging (fMRI) studies have found that psilocybin acutely reduces amygdala activity and connectivity when viewing negative emotional facial expressions. ^{28 , 31 , 32} . Psilocybin has also been shown to acutely decrease activity in the ACC during the resting state. ³³ and during the recollection of autobiographical memory ³⁴ . If the acute effects of psychedelic drugs on affect and associated neurobiology are maintained after the resolution of other acute effects of these drugs, these sustained effects may reveal a transdiagnostic mechanism of the long-lasting therapeutic effects of psychedelics.

The available neuroimaging evidence can be interpreted to suggest that the acute effects of psychedelic drugs on emotion perception (e.g., ²⁷ These factors are associated with altering emotional reactivity from the bottom up by modulating the amygdala's response to negative affective stimuli. ^{28 , 31 , 32} . However, the changes in emotion perception and in positive and negative affect that are observed with psychedelic drugs may also result from changes in top-down emotion control that can lead to observed effects as downstream outcomes, and recent qualitative and self-report evidence supports this. Recent research has shown that, while psychedelics were not believed to reduce the physiological components of craving and withdrawal, they may have reduced the affective components of craving and withdrawal. ³⁵ . Other reports identify an increase in "connection with life," an increase in the creation of meaning. ³⁶ and changes in other higher-level psychological factors ³⁷ as well as involvement with music ³⁸ , as potential mechanisms underlying the effectiveness of the treatment.

Further evidence abounds for a possible role of psychedelics in the acute decrease in resting-state connectivity within the default mode network (DMN). ^{33 , 39 , 40} , and between and within various sensory and cognitive brain networks. ^41 , 42 . Two reports also provided evidence of a post-acute change in DMN connectivity following psilocybin administration, which was shown to be decreased two days after psilocybin in a cohort of long-term meditators.⁴³ ^and Paradoxically, increased levels one day after psilocybin treatment in patients with treatment-resistant depression. ⁴⁴ . These findings support a potential neuroplastic effect of psilocybin on brain network function, loosely consistent with evidence in vitro and in vivo for increased neuritogenesis and spinogenesis in cortical neurons in response to a wide range of classic psychedelic 5-HT receptor agonists. _2A ⁴⁵ . Plasticity within higher-order cortical brain networks may allow for greater modulation of affect by top-down cognitive circuits.

The current open-label, in-person pilot study was conducted to examine whether a single administration of a high dose (25 mg/70 kg) of psilocybin could lead to a lasting increase in positive affect, a lasting reduction in negative affect, a lasting change in neural response to emotional stimuli, and lasting changes in resting-state functional connectivity. A battery of self-report measures of state and trait affect, including the Profile of Mood States (POMS) ⁴⁶ , the State-Trait Anxiety Inventory (STAI) ⁴⁷ , the Positive and Negative Affect Schema – Form X (PANAS-X) ⁴⁸ , the Depression, Anxiety and Stress Scale (DASS) ⁴⁹ , and the Dispositional Scale of Positive Emotion (DPES) ⁵⁰ , was completed one day before, one week after, and one month after psilocybin administration, and responses were compared between time points to investigate the lasting effect of psilocybin on state and trait affect. The participants completed the Big Five Inventory (BFI). ⁵¹ and the Tellegen Absorption Scale (TAS) ⁵² One day before and one month after taking psilocybin, responses were compared between the time points to investigate the lasting effect of psilocybin on personality. One day before, one week after, and one month after psilocybin, participants also underwent fMRI measurements at rest and during the completion of three separate emotion processing tasks (the emotion discrimination task 15, the recognition task). ^of emotions ⁵³ , and a Stroop task of emotional conflict. ⁵⁴ ). fMRI data collected during emotional tasks were compared across time points to determine the lasting effects of psilocybin in response to emotional stimuli in the amygdala and ACC, and the analyses were repeated with whole-brain data to determine if the effects could be detected outside of a priori regions of interest (ROIs). Functional connectomes calculated from resting-state scans were compared across time points to determine the lasting effects of psilocybin on functional network connectivity.

Psilocybin reduced negative affect and increased positive affect.

The main time point effects were observed in the DASS stress (F[2,20] = 4.45, p = 0.025 n ² _p = 0.284), negative affect PANAS (F[2,20] = 9.28, p = 0.0014 n ² _p = 0.466), STAI state (F[2,20] = 3.91, p = 0.037 n ² _p = 0.27) and trace (F[2,20] = 3.96, p = 0.036 n ² _p = 0.277) anxiety, and tension POMS (F[2,20] = 6.37, p = 0.007 n ² _p = 0.376), depression (F[2,20] = 5.46, p = 0.013 n ² _p = 0.316) and total score on the mood disturbance scale (F[2,20] = 5.66, p = 0.011 n ² _p = 0.352). Post-hoc tests demonstrated that DASS stress, PANAS negative affect, STAI state anxiety and POMS tension, depression, and total mood disturbance scale scores were significantly lower 1 week after psilocybin compared to baseline and returned to baseline assessments 1 month after psilocybin (Table). 1 ). POMS depression was significantly greater 1 month after psilocybin compared to 1 week after psilocybin. Trait anxiety scores were reduced 1 month after psilocybin treatment compared to baseline.

The main time point effects were also observed in the DPES joy (F[2,20] = 6.03, p = 0.009 n ² _p = 0.36), content (F[2,20] = 5.11, p = 0.016 n ² _p = 0.314), pride (F[2,20] = 5.85, p = 0.011 n ² _p = 0.343), compassion (F[2,20] = 7.69, p = 0.004 n ² _p = 0.44) and fun (F[2,20] = 7.66, p = 0.004 n ² _p = 0.435) scales. Post-hoc tests (Table) 1 ) demonstrated that DPES scores were significantly higher 1 week and 1 month after psilocybin compared to baseline. The only significant changes observed in personality between baseline and 1 month post-psilocybin (Table) 1 ) were conscientious ( t = 2.33 p = 0.042 d = 0.738) and absorption ( t = 3.55 p = 0.005 d = 1,122) . Descriptive statistics for all self-reported measures are presented in Supplementary Information (Table). S1 ).

Psilocybin led to changes in the neural response to affective stimuli.

The accuracy of the response in the emotion recognition task was nearly maximum for all emotional facial categories at all times (average accuracy of 96.7%, SEM = 0.47%). No effect of the condition (F[4,1421] = 1.171, p = 0.263), time point (F[2,1421] = 1.338, p = 0.1954) or interaction between time point and condition (F[8,1421] = 0.820, p = 0.5847) was observed on the precision.

The ROI analysis produced a main effect of the time point on the BOLD response to stimuli in the emotion recognition task in the left amygdala ( F [2.165] = 6.38 p < 0.0005 n ² _p = 0.098), right tonsil ( F [2,165] = 7,54, p < 0.005 n ² _p = 0.068), and left ACC ( F [2,165] = 6,66, p < 0.05 n ² _p = 0.053), but not right ACC ( F [2,165] = 3,34, p = 0.108 n ² _p = 0.026). No effect of emotional state or interaction between time point and emotional state on any ROI was observed. Post-hoc comparisons (Fig. 1 ) demonstrated a significant reduction in the BOLD response to all facial stimuli in the left amygdala ( t [118] = 4,303, p < 0.00005 d = 0.79) and in the right amygdala ( t [118] = 4,199, p = 0.00005 d = 0.77) in 1 week compared to the baseline. Interestingly, both on the left ( t [118] = 4,557, p < 0.00005 d = 0.83) as to the right ( t [118] = 3,043, p < 0.005 (d = 0.56) the amygdala's response to all stimuli returned to baseline levels in 1 month, compared to 1 week, with no significant difference in the left amygdala ( t [118] = 0.909 p = 0.365 d = 0.17) or right tonsil ( t [ 118] = 0.841, p = 0.402 d = 0.15) in 1 month compared to the post-psilocybin baseline. Individual data for the amygdala response in the emotion recognition task are presented in Supplementary Information (Fig. S1 ). Exploratory associations between changes in self-reported affect over time and changes in amygdala response over time in the emotion recognition task are also presented in Supplementary Information (Figs. S2 - S4 ). No effect was observed in the whole-brain voxel analysis of the emotion recognition task.

The accuracy of performance during the emotion discrimination task was near the ceiling at baseline (92.04%, SEM = 1.7%) and increased from baseline to 1 week (94.7%, SEM = 1.4%). t = 3,089, p = 0.006) and 1 month (94.1%, SEM = 1.5%, t = 2,669, p = 0.014) after psilocybin. No differences were observed in the amygdala or ACC response to the emotional discrimination task between baseline, 1 week, or 1 month, and no significant effect was observed in whole-brain voxel analyses of the emotional discrimination task.

Psilocybin increased the neural response to conflicting emotional information in decision-making circuits.

Participants performed at their peak at all time points on the Stroop Emotional Conflict task (98.9%, SEM = 0.21%). The classic Stroop interference was observed, with a main effect of the task condition (emotionally congruent vs. incongruent trials) on response accuracy (F[3,2575] = 5.704, p = 0.00069) and response time (F[3,2575] = 7.019, p = 0.00011) for the Stroop task, with fewer correct responses and longer response time for incongruent trials compared to congruent trials. No primary time point effect or interaction between time point and condition was observed in behavioral data.

No effect of task condition or time point on amygdala or ACC response was observed in ROI analyses for the Stroop emotional conflict task. However, voxel analysis of the whole brain from the Stroop emotional conflict task identified significant findings. It has been shown that trial-to-trial changes in task conditions can alter cognitive control processes, with the greatest interference effects in Stroop-like paradigms being found in incongruent trials following congruent trials. ^19 , 55 . In a high-demand incongruent testing contrast (incongruent tests following congruent tests or CI tests) compared with low-demand congruent testing (congruent tests following congruent tests or CC tests; contrast [CI > CC]), the BOLD response increased from baseline to 1 week after psilocybin in the dorsal lateral prefrontal cortex (DLPFC) and medial orbitofrontal cortex (MOFC) (Fig. 2A ; Table 2 ), and from baseline to 1 month after psilocybin in the somatosensory cortex and fusiform gyrus (Fig. 2B ; Table 2 ), but no decreases were observed in the BOLD responses (baseline >1 week, baseline >1 month). A greater BOLD response was also observed at 1 week compared to 1 month (1 week > 1 month) in a distributed network of left hemisphere brain regions, including the inferior frontal gyrus (IFG), anterior insula, parietal lobule, and fusiphorous gyrus (Fig. . 2C ; Table 2 ), without significant effects in the opposite contrast (1 month > 1 week).

Psilocybin increases resting-state functional connectivity in brain networks.

Of the 35,778 possible functional connections in Shen's atlas ⁵⁶ , 695 were significantly different from zero (after the Bonferroni correction) at at least one point in time. Functional connectivity increased throughout the brain from baseline to 1 week after psilocybin (greatest connectivity strength for 38 edges and least for 10 edges), and this pattern persisted at one month (greatest connectivity strength for 29 edges and least for 18 edges at 1 month; Fig. 3). _ ). Of the 29 edges that showed increased connectivity 1 month after psilocybin, 7 of them were the same edges that increased from baseline to 1 week after psilocybin, and these edges were evenly distributed across different lobes and brain networks. Changes in static functional connectivity did not follow any discernible network pattern (Fig. 4A However, there were more numerical increases than numerical reductions in functional connectivity within and between networks one week and one month after psilocybin compared to baseline (Fig. 4B ).

The measures of dispersion of connectivity forces within and between networks were not affected at the time points (Figs. S – S8 ).

The current open-label pilot study identified four main sustained effects of a single high dose of psilocybin on affect and the neural correlates of affective processing. First, negative affect decreased 1 week after psilocybin and returned to baseline levels 1 month after psilocybin. Secondly, there was a decrease in amygdala responses to emotional stimuli 1 week after psilocybin, which recovered 1 month after psilocybin. Thirdly, there was an increase in responses in the reward, attention, and decision-making learning circuits 1 week after psilocybin, and an increase in responses in the somatosensory and fusiform gyri 1 month after psilocybin, during high-demand incongruent testing in the Stroop emotional conflict task. Finally, there were overall increases in functional connectivity at 1 week and 1 month post-psilocybin.

A notable feature of the current report is that the reported effects of psilocybin were observed well after the psilocybin would have been eliminated from the body and in addition to the expected transient effects of receptor trafficking that may occur after psilocybin administration. The half-life of psilocybin and psilocin (the active metabolite of psilocybin) is approximately 3 hours. ^57 , 58 This indicates that more than 50 half-lives of the drug have passed before 1 week, ensuring the elimination of the drug from each participant. Furthermore, while the 5-HT receiver _2A It is known to internalize rapidly with both agonism and antagonism, and is believed to be re-expressed approximately 24 to 48 hours after internalization (in the absence of chronic involvement). ⁵⁹ , and thus any transient changes in receptor dynamics related to psilocybin administration would be resolved by the 1-week time point. Instead of receptor trafficking or other residual pharmacological effects, the reported findings may be better explained by a neuroplastic period during which the neural processing of affective stimuli is altered.

The sustained decreases in negative affective states and traits, increases in positive affective states and traits, and decreases in amygdala responses to emotional stimuli observed in this study resemble the reported acute effects of psilocybin. ^27 , 28 . The changes observed in MOFC, DLPFC, IFG, insula, parietal lobe, and fusiform response to conflicting tests, however, are unexpected findings that may reveal a potential top-down mechanism underlying the sustained effects of psilocybin on affect and brain function.

Psilocybin may increase top-down control of emotional processes.

The DLPFC is widely implicated in a range of tasks spanning the domains of working memory. ⁶⁰ decision making ⁶¹ and emotional regulation ⁶² . The hypoactive response of the DLPFC to emotional interference has been demonstrated in major depressive disorder, suggesting a deficit in the neural circuit underlying emotional regulation and top-down control of emotionally conflicting information 63, ^and This hypoactive response was shown to recover with antidepressant treatment. ⁶⁴ . DLPFC has also been shown to exert a top-down influence on the amygdala's response during emotional regulation. ⁶² . Reduced DLPFC recruitment and increased amygdala recruitment during negative emotion downregulation have been identified in a variety of disorders, including mood disorders and substance use disorders. ⁶⁵ .

The MOFC response is observed in a wide range of decision-making tasks. ^66 , 67 , and can encode the reward value of reinforcers, with a forward-to-back gradient within the OFC suggesting that more abstract reinforcers elicit a more forward OFC response. ⁶⁸ . The increase observed in the previous MOFC 1 week post-psilocybin is consistent with increased sensitivity to the abstract reinforcer of positive emotional stimuli. The amygdala and MOFC also have dense bidirectional structural connections that facilitate top-down modulation of saliency detection and reward learning. ⁶⁸ .

Taken together, the increased DLPFC and MOFC response to high-conflict Stroop emotional conflict tests 1 week after the psilocybin task may reflect greater top-down control of emotionally conflicting information and suppression of the amygdala response to negative affective stimuli, which may lead to a shift in the relative relevance of positive and negative affective information in the environment and an overall shift in affect. The lack of change observed in behavior during the Stroop emotional conflict may indicate that behavioral performance was already at its ceiling during the baseline. However, this leaves open the possibility that executive control over emotionally conflicting information was less efficient 1 week after psilocybin, leading to greater recruitment of DLPFC and MOFC to maintain the same level of behavior.

IFG has been implicated in supporting a general domain interference resolution process. ⁶⁹ . Activity in the anterior insula is understood to contribute to interoceptive mapping. ⁷⁰ , and both the IFG and the insula may be involved in the assessment of socio-emotional stimuli. ⁷¹ . The fusiform gyrus contains a series of functionally defined subregions dedicated to the recognition of specific stimulus objects. ^{72 , 73 , 74} , with specialized regions that respond to facial stimuli ^75 , 76 . Greater recruitment of these brain regions in response to conflicting stimuli at 1 week compared to 1 month post-psilocybin may reflect increased attentional load and sharper visceral representation of emotionally conflicting information.

Comparison of current findings with recently reported literature.

Publications from only three other studies reported effects of a psychedelic drug that lasted beyond the drug's acute effect. One of these studies showed an increased amygdala response to negative facial affect stimuli measured using fMRI one day after psilocybin administration in a cohort of patients with treatment-resistant depression. ⁷⁷ . This finding is somewhat intriguing because the amygdala's response to negative stimuli is abnormally increased in patients with depression. ^18 , 78 And one would expect that an antidepressant response would be accompanied by normalization (e.g., decrease) in the amygdala's response to negative stimuli. ^79 , 80 . One possible explanation for this short-term rebound effect in the amygdala response may be that the increased amygdala response one day after psilocybin administration reflects a transient alteration in serotonergic signaling, as 5-HT receptors _2A They are well known for readily internalizing agonistic and antagonistic sentiments. ⁵⁹ including after administration of classic psychedelics ⁸¹ , with re-expression occurring up to 48 hours later ⁵⁹ . It is important to note, however, that increased amygdala reactivity one day after psilocybin was associated with therapeutic outcomes in this sample, and it may be that a subacute rebound process one day after psilocybin is followed by a subsequent drop in amygdala responsiveness within a week (as in the current sample). It is also important to note that any apparent differences between this and previous studies in the direction of the findings regarding the amygdala response may be a function of the different populations studied, where the current study is investigating the effects in healthy volunteers and the previous study was conducted in patients with treatment-resistant depression.

Although no single brain network stood out as uniquely (or significantly) altered in relation to post-psilocybin connectivity in the current report, the pattern of increased connectivity in brain networks that was sustained 1 week and 1 month after psilocybin is generally consistent with the reported acute effects of psilocybin. ⁴¹ , where connectivity between various canonical brain networks is increased. The current findings are also consistent with recent reports that demonstrated increased connectivity within the DMN network the day after the second of two psilocybin administrations in patients with treatment-resistant depression. ⁴⁴ , two days after a single dose of psilocybin was administered to individuals at a 5-day meditation retreat ⁴³ , and the day after administering a closely related substance, ayahuasca, to healthy participants. ⁸² . It should be noted, however, that the aforementioned reports of changes in DMN connectivity one or two days after the administration of psychedelic drugs were restricted to an a priori analysis of the DMN, and these studies did not report connectivity between or within the network of other canonical networks. Furthermore, these studies differed from the current report because all previous studies performed resting-state imaging with eyes closed, while the current study conducted resting-state imaging with eyes open. Furthermore, while the current study collected 16 minutes of resting-state data, previous studies measured resting-state scans of 7 or 8 minutes, which may produce less reliable estimates of resting-state connectivity than scans of 12 minutes or more. ⁸³ .

Changes in personality traits after psilocybin administration are also noteworthy. Although previous studies have observed that trace absorption is a predictor of response to psychedelic drugs ^84 , 85 This is the first demonstration that psilocybin administration can lead to a change in absorption. Numerical increases in openness and extraversion and numerical reductions in neuroticism are consistent with previous effects of psilocybin on the Big Five personality traits. ^86 , 87 But surprisingly, the strongest effect of psilocybin on the Big Five traits was an increase in awareness ( d = 0.738). It has not yet been determined whether these effects are idiosyncratic to the sample in question, generalize to other samples of healthy participants, or have relevance to therapeutic outcomes in patient populations.

Limitations

All conclusions in the current report are limited by the small sample size. However, concerns regarding sample size can be mitigated by the moderate to strong effect sizes that were observed in both self-report and neurobiological outcomes. The lack of a placebo or positive control comparison, the open-label nature of the study, and the lack of multiple time points before psilocybin administration leave open the possibility that some of the reported effects are due to expectation, experimental demand characteristics, and learning or habituation effects. These concerns are somewhat mitigated by the fact that negative affect and task-based fMRI effects at 1 week return to baseline levels at 1 month post-psilocybin. However, replication of this study in a larger sample with convincing control conditions is warranted.

The current report provides preliminary evidence that psilocybin administration can lead to changes in affect and neural correlates of affective processing that persist beyond the acute effects of the drug. Within a dimensional or domain-based taxonomy of brain function and pathology, the reported findings are consistent with a transdiagnostic process that may underlie mood and substance use disorders. Reducing negative affect can impair ruminative processes that contribute to the development and maintenance of mood disorders, and these effects are consistent with the psychological and neural changes that can accompany the antidepressant effects of psilocybin. Disrupting the negative components of craving and withdrawal can impair the development and maintenance of substance use disorders. consistent with psychological and clinical changes observed in patients with tobacco and alcohol use disorders. The reported findings may also account for long-term positive changes in mood, attitude and well-being that have been reported in healthy individuals ^{88 , 89} .

Although negative affect and the brain's response to affective stimuli were reduced 1 week after psilocybin, they recovered by the 1-month time point, suggesting that psilocybin may have initiated a dynamic and neuroplastic process that was sustained for at least a few weeks. It is possible that this neuroplastic period allows for a more lasting shift toward positive affect. The observed increase in the strength of functional connectivity indiscriminately between networks may reflect a general-domain cortical plasticity process that supports the observed changes in affective processing, consistent with preclinical evidence of the psychoplastogenic properties of psychedelic drugs. ^45 , 90 . Overall, current findings identify negative affect as a potential therapeutic target for psilocybin.

participants

Twelve volunteers (7 women, mean age 32.1 ± 7.5 years) participated in this open-label, longitudinal pilot study within individuals. Volunteers were included if they were right-handed, aged between 18 and 45 years, clinically healthy (as determined by medical history, physical examination, electrocardiogram, blood analysis, and urine test for common drugs of abuse), and psychiatrically healthy (as determined by the Structured Clinical Interview for DSM-IV). Individuals were excluded due to contraindications for magnetic resonance imaging (including previous head trauma, claustrophobia, presence of certain implants and/or non-removable ferrous metals), as well as potential contraindications to psilocybin (personal or family history of psychotic or bipolar disorder, history in the last 5 years of moderate or severe substance use disorder, and taking medications with psychoactive effects or affecting the CNS). A urine pregnancy test (for women) and a urine test for common drugs of abuse (for all participants) should be negative during screening and on the morning of drug administration.

The sample was racially homogeneous (100% Caucasian), more than half (58.3%) were married at the time of their participation, 83.3% had a bachelor's degree or higher, and all reported limited lifetime use of hallucinogens (median of 1, range of 1 to 4 uses), with the most recent use occurring an average of 8.3 years ago. This study was registered at ClinicalTrials.gov (NCT02971605, registered on November 23, 2016). All participants provided informed consent in accordance with the Common Rule and the Declaration of Helsinki. All procedures were approved by the Institutional Review Board of the Johns Hopkins University School of Medicine, and participants received a total of $240 upon completion of the study.

Study procedures

After registration, participants underwent preparation, intensive care, and post-treatment care for psilocybin administration sessions, following published safety guidelines. ⁹¹ . Participants were assigned two session monitors whom they met during two preparatory meetings prior to medication administration, for a total of approximately eight hours of preparation time. During the preparatory meetings, participants recounted their life stories and important life events, received training, and practiced each of the three emotional tasks that would be performed during the MRI evaluations. (See “Affective Tasks” below ) and monitors instructed participants regarding the range of possible experiences that may be encountered during the acute effects of drugs. Practical sessions involving emotional tasks were included to ensure that participants were familiar with all tasks before the start of the MRI procedures and to minimize the initial learning effects on these tasks. The participants then completed a single psilocybin administration session lasting approximately 7 hours, using established procedures. ⁹¹ based on several previous and ongoing studies with healthy participants ^{89 , 92 , 93 , 94 , 95} and clinical populations ^1 , 4 . Participants returned one day after the psilocybin session to meet with the study team and review the previous day's psilocybin session.

psilocybin session

Participants consumed a small, low-fat breakfast > 1 hour before arriving at the Behavioral Pharmacology Research Unit at Johns Hopkins Bayview Medical Center. Participants remained lying on a couch under the supportive supervision of two study staff members after ingesting a capsule containing a high dose of psilocybin (25 mg/70 kg) prepared by our research pharmacy. Blood pressure, heart rate, and participant behavior assessments by the team were evaluated as safety measures at 0, 30, 60, 120, 180, 240, 300, and 360 minutes after capsule administration.

Questionnaires

A battery of questionnaires was completed one day before, one week after, and one month after psilocybin administration to assess emotional function. Positive and Negative Affect Scale - X (PANAS-X) ⁴⁸ It is a 60-item adjective rating scale with a 5-point response format (0 – very slightly or not at all, 1 – a little, 2 – moderately, 3 – somewhat, 4 – extremely) that is scored on general subscales of positive and negative affect, as well as a range of facets of positive and negative affect. Participants were asked to indicate the degree to which they generally feel (“i.e., how you feel on average”) the different feelings and emotions described by each adjective. The Profile of Mood States (POMS) ⁴⁶ is a 65-item rating scale with a 5-point response format (0 – not at all, 1 – a little, 2 – moderately, 3 – a little, 4 – extremely) that is scored on seven subscales (tension, depression, anger, fatigue, confusion, vigor, and total mood disturbance). Participants were asked to indicate the degree to which each item described how they felt during the past week, including today. The Dispositional Scale of Positive Emotions (DPES) ⁵⁰ is a 38-item Likert scale with a 7-point response format (with response anchors at 1 “Strongly Disagree”, 4 “Neither Agree nor Disagree”, and 7 “Strongly Agree”) that is scored on seven subscales (joy, contentment, pride, love, compassion, amusement, and admiration). Participants were asked to think about each statement and decide how much they agreed or disagreed with it. The Depression, Anxiety and Stress Scale (DASS) ⁴⁹ is a 21-item rating scale with a 4-point response format (0 – does not apply to me at all, 1 – applies to me to some degree or some of the time, 2 – applies to me to a considerable degree or a good part of the time, 3 – applies to me very much, or most of the time) that is scored on three subscales (depression, anxiety, and stress). Participants were asked to indicate how much each statement in the DASS applied to them in the past week. The State-Trait Anxiety Inventory (STAI) ⁴⁷ is a 40-item rating scale with a 4-point response format (0 – almost never, 1 – sometimes, 2 – often, 3 – almost always) that is classified into two subscales (state anxiety and trait anxiety). For questions about "state" anxiety, participants were asked to select the response for each item that best described how they feel "now, that is, at this moment." For questions about “trait” anxiety, participants were asked to select the response that best describes how they “generally feel, i.e., most of the time”.

Participants also completed personality measures at screening and again one month after psilocybin treatment. The Big Five Inventory (BFI) ⁵¹ It is a 44-item Likert scale with a 5-point response format (1 – Strongly disagree, 2 – Somewhat disagree, 3 – Neither agree nor disagree, 4 – Somewhat agree, 5 – Strongly agree) that is scored on five subscales (extroversion, neuroticism, agreeableness, conscientiousness, openness). The Tellegen Absorption Scale (TAS) ⁵² It is a 34-item rating scale with a 4-point response format (with response anchors at 0 “Never” and 3 “Always”) that is scored on a single total score for absorption.

Magnetic resonance imaging (MRI) assessments

One day before, one week after, and one month after psilocybin administration, participants completed the Stroop tasks of emotional discrimination, emotional recognition, and emotional conflict in that order, with an 8-minute open-eyes resting-state scan between each pair of tasks (total of 16 minutes of resting scans for each visit), during measurement of the blood oxygenation level-dependent signal (BOLD) using echoplanar imaging (EPI; TR/TE = 2200/30 ms, tilt angle = 75°, voxel size = 3 mm). ^3,37 axial slices collected in an interleaved manner with a slice interval of 1 mm, with an acceleration factor SENSE = 2). All scanning procedures were performed on a Philips 3T MRI scanner equipped with a 32-channel head coil at the FM Kirby Research Center for Functional Brain Imaging at the Kennedy Krieger Institute in Baltimore, MD. Each scanning session lasted 60 minutes.

Task performance during MRI sessions began with a short practice task on the scanner before the MRI measurement, followed by full task performance during the MRI measurement. All facial emotional stimuli were selected from the NimStim Emotional Facial Expression database. ⁹⁶ , and balanced within the task and between conditions to the degree possible based on sex, race, and frequency of open versus closed mouth in response to each stimulus. Visual stimuli were projected onto a frosted Plexiglas shield at the open end of the scanner orifice, which was viewed through a mirror placed on the head coil. Participants provided responses using a fiber optic magnetic resonance imaging (fMRI) safe response device. Stimuli and responses were presented and recorded using presentation software (Neurobehavioral Systems, Inc. Berkeley, CA).

Emotional discrimination. During this task, participants viewed a series of three images (one at the top of the screen and two at the bottom of the screen) containing three emotional facial expressions (fear or anger) or three geometric shapes (vertically or horizontally oriented ellipsoids). ^{16 , 17 , 28} . Participants were instructed to press a button (on their right or left hand) to indicate the image at the bottom of the screen (right or left) that corresponded to the image at the top of the screen. Participants completed four 30-second blocks of facial matching attempts interspersed with five 30-second blocks of shape matching attempts. Each block began with a 3-second cue (“match faces” or “match shapes”) followed by 6 trials (4.5 seconds per trial) and a 500-second interval between stimuli (total task time: 4 minutes 57 seconds).

Recognizing emotions. In this task, participants are presented with a series of facial expressions of joy, sadness, fear, anger, and neutrality and are instructed to press a button to identify the emotion expressed in each face. ^{53 , 97 , 98} . Sixty stimuli (12 stimuli for each emotion) are presented one at a time for 4 seconds each, with an ISI jittered average of 3 seconds and 15 seconds of rest at the beginning and end of the task, for a total task time of 7 minutes and 30 seconds. An equal number of male and female faces were presented for each emotion. The order of emotions was pseudorandomized according to a genetic algorithm to maximize the statistical separation of each condition. ⁹⁹ , but within each emotional state, the order of the actual stimuli is random.

Stroop emotional conflict. This task requires participants to identify the valence of emotional facial expressions (targets) with superimposed emotional words (distractors). ^22 , 54 . The emotional facial expressions consist of 18 happy and 18 sad emotional faces (9 male and 9 female each), matched between emotional conditions based on the strength of the emotional valence and presented in pseudo-random order. The emotional words consisted of 18 emotional words with positive valence and 18 with negative valence from the Affective Norms for English Words (ANEW). ¹⁰⁰ that are combined between valence conditions in valence intensity (degree of pleasure versus displeasure), arousal, dominance, and word length (in characters) and paired in pseudo-random order with facial emotional stimuli. A given target-distractor pair may have congruent or incongruent emotional valence. The stimuli are pseudorandomized to control the order of congruent (C) and incongruent (I) stimuli, balancing the order effects for the following sequences in gender and emotional valence of the target stimulus: congruent trials following a previous congruent trial (CC), congruent trials following a previously incongruent trial (IC), incongruent trials following a previous incongruent trial (II), and incongruent trials following a previous congruent trial (CI).

Analysis

Analysis of a self-report questionnaire

Mixed-effects, repeated measures ANOVAs were used to determine the persistent effects of psilocybin on self-reported affect measures, comparing each measure between each time point (baseline, 1 week, and 1 month post-psilocybin). Where a significant main effect was observed, we followed up with post-hoc comparisons between each time point, corrected for multiple comparisons using Tukey's method for multiple comparisons of all paired means. ¹⁰¹ . Tests t pairs They were used to test changes in personality measures between screening and 1 month post-psilocybin.

BOLD task-based data preprocessing and analysis

All task-based BOLD data underwent preprocessing, region of interest (ROI) extraction, and ROI analysis to determine the response of the left and right amygdala and left and right ACC to task conditions in each fMRI task. The preprocessing steps consisted of slice time correction, realignment/motion correction, and normalization to an EPI model registered in MNI space. ¹⁰² and smoothing using a 6mm FWHM kernel. The first eigenvariant of all voxels within four ROIs (left and right amygdala and left and right ACC) was extracted for each subject and each scan and subjected to separate general linear model (GLM) analyses at the subject level for each affective task at each time point (baseline, 1 week post-psilocybin, and 1 month post-psilocybin).

The subject-level GLM design matrices consisted of six realignment motion parameters, a motion sensor or "debug" regressor generated using outlier detection, and intermediate settings (global signal z-value limit = 5, subject motion mm limit = 0.9) in the ART toolbox. ¹⁰³ , the average signal within each execution, a linear term to model the signal deviation, a regressor to model all button presses made by the participant, and regressors of interest for each task. The design matrix for the emotion discrimination task included regressors of interest for face blocks and shape blocks, and a contrast [face > shapes] was set as the contrast of interest for each subject and point in time. The design matrix for the emotion recognition task included a regressor indicating the onset of each stimulus and separate regressors of interest for each emotional state of the face (happy, angry, sad, fearful, and neutral). An emotion greater than the total contrast of stimuli was appropriate for each emotional state ([happy > all stimuli], [angry > all stimuli], etc.). The design matrix for the Stroop task of emotional conflict included regressors of interest for each of the four types of first-order sequence (congruent trials following a congruent trial, or CC, incongruent trials following a congruent trial, or CI, incongruent trials following an incongruent trial, or II, and congruent trials following an incongruent trial, or IC). Two contrasts of interest were set: one for all incongruent trials greater than all congruent trials ([CI & II > CC & IC]) and one for high-demand incongruent trials greater than low-demand congruent trials ([CI < CC]). and congruent trials that follow an incongruent trial, or IC). Two contrasts of interest were set: one for all incongruent trials greater than all congruent trials ([CI & II > CC & IC]) and one for high-demand incongruent trials greater than low-demand congruent trials ([CI < CC]). and congruent trials that follow an incongruent trial, or IC). Two contrasts of interest were set: one for all incongruent trials greater than all congruent trials ([CI & II > CC & IC]) and one for high-demand incongruent trials greater than low-demand congruent trials ([CI < CC]).

SPM12 ( http://www.fil.ion.ucl.ac.uk/spm/software/spm12/ ) was used to preprocess all the data, and SPM12, MaRSBaR ( http://marsbar.sourceforge.net ) and MATLAB (R2017a, version 9.2.0.556344) were used to conduct GLM analyses. A one-way ANOVA was adjusted for subject-level ROI contrasts to determine a main effect of time point on the BOLD response on each ROI for each task. Post-hoc comparisons were performed using tests. t , corrected for multiple comparisons using the Holm-Bonferroni method ¹⁰⁴ . The analyses were repeated as general exploratory linear models of the whole brain voxel to investigate potential effects outside the hypothetical areas. Whole-brain analyses were thresholded at p < 0.0005 (uncorrected), with a clustering threshold of p < 0.05 (uncorrected).

Resting-state fMRI analysis

The resting-state data were pre-processed as task-based data and then subjected to band-pass filtering. ^105 , 106 simultaneous (0.009–0.08 Hz) and regression of annoying parameters. The troublesome parameters consisted of linear trends, the first 5 eigenvectors of cerebrospinal fluid and white matter signal (identified using masks derived from segmented and normalized T1-weighted structural images), 6 realignment motion parameters, and a motion sensor or “scrubbing” regressor generated using outlier detection and intermediate settings (global signal z-value threshold = 5, subject motion mm threshold = 0.9) in the ART toolbox. ¹⁰³ . The pre-processed and annoyance-regressed data were then parceled using the Shen 268-node functional brain atlas. ⁵⁶ . Voxels within each node were calculated in each acquisition to produce 268 time series (one for each node) for each participant. One subject was excluded from the resting-state analysis due to a lack of resting-state data from the 1-week time point.

The static functional connectivity between each edge (each pair of nodes in Shen's atlas) was calculated using Pearson correlations. These values ​​and all other correlations were transformed using Fisher's z-transform for all statistics. To explore the differences in static whole-brain connectivity, significant (negative and positive) boundaries were identified using tests. t from a sample separated between participants for each edge and time point, limited using the Bonferroni correction for all 35,778 edges. Although statistically conservative, this procedure produces the most reliable edges in our relatively small sample. All edges that survived this boundary for at least one time point were then contrasted between the time points (baseline x one week and baseline x one month) using Paired t -tests (α = 0.05).

Two resting-state scans were collected at each MRI visit, and all resting-state dependent variables were averaged within the subject at each time point and each edge before analysis. Shen's atlas nodes are grouped into eight canonical functional networks: medial frontal, frontoparietal, default mode, subcortical-cerebellum (including salience), motor, visual I (medial), visual II (occipital pole), and visual association (lateral), yielding 8 additional observations within the network and 28 observations across networks for each outcome measure (static functional connectivity, DCC, and entropy). In order to explore within and between network differences, all edges within each network, or all edges between each pair of networks, were averaged and compared at points in time (via (t test). Visual analysis of the value matrix- t It was used to identify obvious patterns in changing connectivity, but it should be interpreted with caution.