Cognitive functioning associated with acute and subacute effects of classic psychedelics and MDMA - a systematic review and meta-analysis

Classic psychedelics and MDMA have a colorful history of recreational use, and both have recently been re-evaluated as tools for the treatment of psychiatric disorders. Several studies have been carried out to assess potential long-term effects of a regular use on cognition, delivering distinct results for psychedelics and MDMA. However, to date knowledge is scarce on cognitive performance during acute effects of those substances. In this systematic review and meta-analysis, we investigate how cognitive functioning is affected by psychedelics and MDMA during the acute drug effects and the sub-acute (“afterglow”) window. Our quantitative analyses suggest that acute cognitive performance is differentially affected by psychedelics when compared to MDMA: psychedelics impair attention and executive function, whereas MDMA primarily affects memory, leaving executive functions and attention unaffected. Our qualitative analyses reveal that executive functioning and creativity may be increased during a window of at least 24 h after the acute effects of psychedelics have subsided, whereas no such results have been observed for MDMA. Our findings may contribute to inform recommendations on harm reduction for recreational settings and to help fostering differential approaches for the use of psychedelics and MDMA within a therapeutic framework.


Study quality
We assessed study quality and risk of bias using the Cochrane risk of bias (RoB) rating scales 58 for parallel-arm trials and cross-over trials.For sequentially designed trials, we used the rating tool from the National Heart, Lung, and Blood Institute (NHLBI) 59,60 .LAB and TGR independently rated study quality, and a consensus was reached through discussion for each parameter.

Meta-analysis
We included studies that reported results as mean with standard deviations into the meta-analyses and performed analyses if at least three viable studies per domain were included.As many studies reported multiple tests per domain we performed a three-level meta-analysis 61 on the standardized mean difference, incorporating between-study as well as within-study heterogeneity into the model.We modeled the effect sizes as nested within their respective studies by assigning a nested random effect to the grouping variable "study".Thus, the random intercept was allowed to vary across different studies.Test statistics and confidence intervals for the fixed effects used a t-distribution.The test of moderators checks the assumption that µ mdma = µ psychedelics = 0 , which is a test for an overall effect of the variable "drugType" (MDMA vs. psychedelics).To compare both drug types we defined a contrast, testing if µ mdma − µ psychedelics = 0 (Wald-Type Test).
Separate analyses were conducted for studies with sub-acute assessments (testing done up to 24h after drug administration) and microdosing approaches.Heterogeneity was measured using the I 2 statistic.For studies reporting multiple dosages and/or times of administration, data for the highest dosage and closest time to peak drug concentration (t max ; based on comparison values from [62][63][64] ) were selected.We planned six sensitivity analyses: (1) restricting included studies to accuracy-based test or (2) speed-based test), (3) excluding the study with the highest weight, (4) using measurements closest to drug administration instead of closest to t max in cases of multiple measurements, (5) restricting to the same psychedelic, (6) excluding studies with a high risk of bias rating.Statistical analyses were carried out using R version 4.3.2 65and the following R packages: dmetar v. 0.1.0 66, metafor v. 4.4.0 67 .All used packages are cited in the Supplementary Materials.

Study selection
Excluding duplicates, our search yielded 15,224 records, out of which 691 were selected for full-text screening.Following full-text screening, 122 records were included in the systematic review, and 31 records were viable for quantitative analysis.There was high inter-rater agreement (Fleiss' kappa k = 0.98; 95%CI [0.97-0.99]),with only ten cases of 691 screened full-texts involving disagreements.Two of these disagreements were related to the type of investigated substances, in which cases the third reviewer (TM) provided expertise.The other eight disagreements were due to the unclear nature of the included tests and were concluded based on literature research and expertise by reviewer TGM.Supplementary Tables S2-9 provide an overview of the included studies.Specifically, 68 studies investigated the cognitive effects of classic full-dose psychedelics (Supplementary Tables S2 and S6), 34 studies focused on MDMA (Supplementary Tables S3 and S7), nine studies explored microdosing (Supplementary Tables S4 and S8), and nine studies assessed cognition within 24 h after acute dosing (Supplementary Tables S5 and S9). Figure 1 (and Supplementary Table S11) displays details of the different phases of the search.

Quality rating
All 83 studies that were assessed using the Cochrane RoB tool, including both parallel-arm as well as cross-over studies, were rated as having a high risk of bias.Out of the 37 studies assessed with the NHLBI instrument, 34 were considered to be of poor quality, and three were rated as fair quality (see Supplementary Tables S2-5).The high risk of bias in the included studies is mainly due to the unblinding effects of psychedelics and MDMA 68 .If proper blinding would have been accomplished, five sequential studies would have been rated fair, while 41 placebo-controlled studies would have been rated as having some concern and 5 studies as having a low risk of bias.For the purpose of the respective sensitivity analyses, we excluded the blinding criterion from the quality rating, allowing to control for differences in other sources of bias risk between studies.

Qualitative analysis
Here, we present the results regarding the influences of psychedelics and MDMA on memory performance, executive functioning, and attention.Additional findings on other cognitive abilities, such as visuospatial skills, intelligence, motor skills, and language functions, are provided in the supplement.
Memory.Four of these studies (n = 124) 157,158,161,162 examined the effects of microdosed LSD on memory using five tests, and none of them found any significant effect.
Executive functioning.Five studies (n = 161) evaluated the effects of microdoses of LSD 157,161,163 and psilocybin 156,160 on executive functioning using nine different tasks.Among them, two studies 160,163 reported impaired performance in two tasks, one study 156 reported improved performance in two tasks, while the majority of studies 157,160,161,163 found no significant difference across eleven tasks.
Attention.Four of the reviewed studies (n = 106) assessed attentional performance under the influence of a LSD 158,159,163 or psilocybin 160 microdoses using six different tasks.Three of these studies 158,160,163 found no significant effect on attentional performance across five tasks, while one study 159 reports reduced attention in one task.

Full-dose, sub-acute assessment of psychedelics
Eight studies (n = 282) 4,99,164-169 had their participants perform cognitive tasks of memory, executive functioning, or attention on the day after consumption of a psychedelic substance (Supplementary Table S5).
Memory.Three studies (n = 46) 166,167,170 reported sub-acute effects of LSD on five different memory tasks.Two of these studies 166,170 showed increased memory performance across three tasks, and two studies 167,170 showed no effect in two task.
Executive functioning.Six studies (n = 288) evaluated sub-acute executive functioning using five different tasks in psilocybin 165 , ayahuasca 4,164,168 , 5-Meo-DMT 167,169 , and LSD 170 .Out of these studies, two 164,170 showed impaired performance in two tasks, while four studies 4,165,168,169 reported improvement across four measures, and three studies 4,167,170 found no effect in four tasks.
Attention.Two studies (n = 46) 167,170 investigated sub-acute effects of LSD using two tasks and observed no significant difference in performance compared to placebo.
Full-dose, sub-acute assessment of MDMA Two studies (n = 27) 144,171 assessed the subacute cognitive effects of MDMA and found no effect on three memory tasks, three executive functioning tasks and one attention task (see Supplementary Table S5).

Microdosing
Only three microdosing could be included for quantitative analysis and these studies were exclusively related to the domain of creativity.Thus, no quantitative analysis of microdosing effects on other cognitive domains could be performed.Three studies 156,157,160 , involving n = 81 participants, were included in a quantitative analysis of creativity tasks under microdose conditions.The overall effect estimate revealed no significant effect of psychedelic microdoses on creativity tasks (Z = 0.37 [95%CI − 2.51; 3.24], p = 0.64), see Figure 5.The estimated variance components were τ2 Level3 = 0.76 and τ2 Level2 = 0.96 Therefore

Sensitivity analyses
An overview of all completed sensitivity analyses can be found in Supplementary Table S10.

Discussion
In our analysis of the acute and subacute effects of classic psychedelics and MDMA on cognition, we report the following findings: acute effects of classic psychedelics are consistently associated with reduced cognitive performance across multiple domains.Our qualitative and quantitative results indicate that the most pronounced effects are observed in the domain of attention, whereas memory and executive functioning remain less intensively affected.In contrast, MDMA primarily impacts acute memory performance.Notably, the included studies were highly heterogeneous in terms of the tests applied, reported outcomes of each test, drug dosages, and time between drug and test administration, which hinders accurate comparisons across studies.Our results provide important information in the context of psychedelic harm reduction, for instance, when it comes to choosing a safe setting for recreational use, considering to drive a car or to elsewise participate in road traffic.In addition, impaired memory functions under MDMA might potentially reduce certainty in users about the amount of consumed MDMA which could lead to dangerous redosing and overdosing 172 .
Our results also provide some insight when it comes to fostering therapeutic interventions for the framework of psychedelic-assisted therapy (PAT).Specifically, during the acute experience, applying cognitive interventions  www.nature.com/scientificreports/might be challenging, if patients are unable to properly follow the therapist's guidance.In contrast, with psychedelics specifically, there might be a place for increased therapeutic support during the days after psychedelic dosing sessions, since cognitive performance is potentially increased during that subacute ('afterglow') window.
For MDMA, on the other hand, the effects on memory performance might actually be beneficial in the context of Posttraumatic Stress Disorder (PTSD)-specific therapy that involves reconsolidation of traumatic memories under the influence of MDMA 173 .The reduced ability to encode, especially emotional 149 , information might reduce reconsolidation of traumatic memories and promote extinction, thus promoting a reduction of PTSD symptoms 174 .In fact, there is already a line of research focused on producing psychopharmacological agents that can modify memory to support PTSD therapy 175 .Thus, our results support the notion that psychedelic and MDMA-assisted therapy are two distinct forms of therapies that are applicable to different disorders and different treatment approaches, as reflected in their distinct cognitive effects facilitating distinct therapeutic interventions.In addition to the consistent evidence that altered reconsolidation of memories is a main mechanism of MDMA-assisted therapy 173,176 , the re-experiencing of memories might be relevant in therapy with classic psychedelics as well 177,178 .
The disparity between psychedelics and MDMA with regard to acute cognitive effects may be linked to differences in acute subjective effects.In short, while psychedelic experiences might be characterized by strong perceptual distortions 26,76,179 and potentially distracting experiences such as reduced self-other boundaries 180 , MDMA experiences are often marked by an pronounced emotional pattern of effects 26,181 .The emotional effects of MDMA may impair the ability to retrieve state-incongruent memories (e.g., retrieving neutral stimuli such as word lists, in a highly aroused positive emotional state) 149,182,183 and as such lead to the observed memory deficits, while the effects of psychedelics might provide sensory distraction that reduces acute attention.Notably, under microdosing conditions which lack the perceptual alterations of psychedelics 184 , those effects on cognition disappear 156,159,160,163 .While this may suggest that the nature of the psychedelic alterations might be inherently distracting, this could also be the result of dose-dependent effects, as studies with multiple dosages of psilocybin showed dose-dependent effects on cognitive performance 71,73,73,76 .
While it is unclear if the observed difference is due to these factors, there is strong evidence indicating a reduction in attention under psychedelics which aligns with the recently proposed model of cognitive functioning under psychedelics by Sayalı and Barrett 185 .This model posits that psychedelics induce a transient increase of cognitive flexibility while simultaneously impairing attentional capacities.As we found preliminary evidence that attention was significantly reduced when focusing on speed-based tasks but not accuracy-based tasks, our  findings support the notion that psychedelics specifically reduce sustained attention or vigilance but not cognitive control (as is needed for accuracy-based tasks).Additional confirmatory evidence is provided by research regarding creative performance under the influence of psychedelics 99,111,186 and animal research 187,188 showing an increased ability to switch action patterns under psilocybin.Furthermore, this model of increased cognitive flexibility aligns with results regarding neuropsychological consequences of long-term psychedelic use.For instance, there is no evidence for lasting cognitive deficits 8,9 , but some preliminary evidence indicating potential sustained cognitive benefits beyond the acute psychedelic effects.Our qualitative review indicated that executive functioning and creativity may be increased within a window of 24 h after a using a psychedelic 4,165,168,169 .Furthermore, studies have shown improvements in executive functioning one week after psilocybin administration 189 , and there is as some evidence of improved executive performance in individuals who regularly use ayahuasca 190,191 .Finally alterations in neuronal networks of executive control 192 , salience and default mode networks 193 have been reported, as well as increased neuroplasticity for up to one week after administration 194 .As for MDMA, only few studies have investigated the sub-acute effects, and the evidence so far indicates no impairments or improvements on cognition 144,171 .Even though, animal research indicates that MDMA might enhance critical periods of social learning in the days after administration 195,196 , there is also evidence linking MDMA-induced memory deficits to alterations in serotonergic functioning [197][198][199] , which is in line with findings indicating deteriorations in the serotonergic system after repeated MDMA use 200 .Neurobiological long-term consequences of repeated use might also account for the proposed long-term cognitive deficits associated with MDMA use 27 .However, while neuronal adaptive processes that occur as a result of MDMA use are consistently reported, the significance and relevance of these changes for cognitive alterations remain a topic of ongoing debate 34 .

Limitations
First, the included studies reported results from various cognitive tests and employed heterogeneous research methods, including different dosages or times of drug administration.This heterogeneity limits the robustness of the presented results and reduces the certainty of our conclusion.However, we addressed this limitation by selecting the highest dosage and dosage closest to suspected peak effects for the meta-analysis and by combining multiple cognitive tests to load on the same factor, facilitating quantitative analysis (as is commonly done in other fields 201 ).This enables us to draw conclusions despite methodological variety, although these conclusions should be considered as preliminary.Future studies should take care to report common protocols to allow for analysis of different phases of cognitive processing under psychedelics (e.g. 183).
Secondly, the chosen methods did not allow for more fine-grained analyses of dose-dependent or timedependent effects of psychedelics or MDMA on cognition.Future studies should aim to assess the effects of a broader range of psychedelic dosages (from micro-to macro-dosing) and across multiple timepoints.
Thirdly, we did not analyze the contextual setting of the psychedelic/MDMA session or the test administration.Since psychedelic effects are known to be highly influenced by context 202 , variations in the settings of the experience itself may have impacted the effects on cognition.
Fourth, blinding participants and researchers in studies involving the acute administration of psychoactive substances is a well-known challenge 68 .As indicated in the risk of bias assessment, most of the included studies did not achieve full blinding, which may have introduced strong expectancy effects, reducing the validity of the reported results.However, unlike in clinical studies, it remains unclear what participants' expectations are in cognitive assessments under the influence of psychoactive substances and how these expectations might influence actual test performance.
Fifth, we were not able to incorporate analyses taking into account previous drug exposure of participants.While most studies reported if their participants had ever used the investigated substance, only a fraction reported use frequency in a fashion that would allow for co-variate analysis.This lack of data could be an additional factor contributing to the high heterogeneity found in our results, especially considering the chronic tolerance reported by MDMA users 27 .Sixth, while we reviewed a large number of studies in a qualitative manner only few studies could be included for quantitative analysis.Thus, our quantitative results are not representative of the literature at large but are based on a few select studies.Future studies in this field should take care to report as much of the raw data as possible to support more overarching analysis efforts like ours.
Finally, all included studies were conducted with healthy volunteers and did not include patient populations.As the effects of psychotropic medications sometimes can vary between the healthy and patient populations 203 , the same may be true for psychedelics and MDMA.Further studies are needed to investigate the cognitive effects of psychedelics and MDMA within the framework of psychedelic-assisted therapy.

Conclusion
This study is the first to contain a meta-analytic assessment of the acute cognitive effects of classic psychedelics, including a comparison with the effects of MDMA.Additionally, this is the largest synthesis of neuropsychological data for psychedelics as well as MDMA so far conducted.We report robust evidence that psychedelics acutely reduce attention, and preliminary evidence that psychedelics may also have acute detrimental effects on memory and executive functioning.In contrast, MDMA appears to have the strongest impact on memory performance, whereas attention and executive functions are less intensively affected in the acute phases.Conversely, during the sub-acute ('afterglow') period when acute effects have worn off, psychedelics potentially exert beneficial effects on executive functioning, while such effects have not been found for MDMA.Our findings add neuropsychological evidence for fostering distinct therapeutic approaches for psychedelics when compared to MDMA-assisted therapy.Given deteriorations of attention under psychedelics, psychotherapeutic interventions might be less fruitful during acute effects of the substance when compared to the subacute "afterglow" period.In contrast, given the specifics of acute effects of MDMA on cognition, psychotherapeutic techniques may be more fruitful during the acute experience when compared to psychedelics, whereas therapy during the subacute window might be somewhat less effective.In sum, there is a need for more research regarding cognitive underpinnings of classic psychedelics and MDMA with respect to prevention and for their use as tools for psychotherapy.