FormalPara Key Points

Lysergic acid diethylamide (LSD) dose-dependently enhances mood both short- and long-term, according to the data extracted from the self-reports in the included studies.

LSD-induced positive mood enhancement in humans and reward learning in animals seems to be mediated by the serotonin 2A (5-HT2A) receptor.

LSD may have the potential to reduce reward responsiveness and effort expenditure for other reinforcers.

1 Introduction

In 2008, the National Institute of Mental Health (NIMH) introduced a domain-based framework of research domain criteria (RDoC) to facilitate collecting and integrating data from various classes of measurements or “units of analysis” as termed by the RDoC work group, namely molecules, cells, physiology, behavior, self-reports, and paradigms [1, 2]; to ensure a smooth flow of the text, we have used the terms “measurement”, “units of analysis”, “outcomes”, and “outcome measure” interchangeably. Among the six suggested domains, the positive valence system (PVS) is of particular clinical relevance [3, 4]. The PVS is a broadly defined domain, which consists of conceptually disparate yet functionally coherent processes such as reward responsiveness, reward learning, and reward valuation that govern goal-directed behavior [1]. Reward responsiveness refers to the processes that govern hedonic responses to present or anticipated rewarding stimuli [5]. Reward learning involves an organism experiencing and gathering information about rewarding stimuli and adapting its behavior accordingly [6]. Finally, reward valuation involves estimating possible outcomes based on past and new experiences [7].

Anhedonia and other deficits in the PVS are common characteristics occurring across diagnostic categories (e.g., major depressive disorder) and are known to be resistant to common pharmacological treatments, such as selective serotonergic reuptake inhibitors (SSRIs) [8, 9].

An increasing number of clinical studies support the therapeutic potential of classical psychedelics as a novel class of rapid-acting serotonergic drugs with a potentially unique clinical efficacy across diagnostic categories, and especially in treatment-resistant conditions [10, 11]. Classical psychedelics include agents such as psilocybin, mescaline, N,N-dimethyltryptamine (DMT), and lysergic acid diethylamide (LSD) that primarily, though not exclusively, activate serotonin 2A (5-HT2A) receptors [12]. These compounds are widely recognized for eliciting psychological, cognitive, and emotional alterations upon use [13]. Unlike SSRIs, which require several weeks to produce therapeutic effects, psilocybin-facilitated therapies have shown an ability to lower depressive and anxiety symptoms within a day, and this effect can persist for weeks and even months after a single administration of psilocybin [14,15,16]. The foregoing impressive findings have rekindled scientific interest in the mechanism of action of psychedelics and inspired a number of ongoing clinical trials exploring the therapeutic potential of these drugs [17]. The resulting abundance of data merits analysis within a multidomain framework such as RDoC, capable of integrating multidisciplinary outcomes into a more holistic appraisal of the efficacy of these pharmacological agents.

In a recent review, we reported that psilocybin affects several clinically relevant domains, particularly positive valence systems, as assessed by short- and long-term outcome measures across multiple units of analysis [18]. Drawing from these findings, the current review synthesizes existing data from the literature for another potent psychedelic agent, LSD, with a focus on the PVS to assess its clinical relevance. As such, this is the first review article aiming to evaluate the effect of LSD on reward processing across multiple units of analysis.

2 Methods

This study complies with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Two LSD-related (lysergic acid diethylamide and LSD) and 13 PVS-related (reward, happiness, bliss, motivation, reinforcement learning, satisfaction, operant, conditioning, decision making, habit, valence, affect, mood) search terms were used to create 26 distinct searches (e.g., “LSD”, “reinforcement learning”) across four databases, namely PubMed, Scopus, PsycINFO, and Web of Science, in July 2023. The search terms were identical in all databases; however, the syntax was modified to meet the requirements of each database. A manual search of the reference lists of the included literature, and a Google Scholar search was also conducted to identify additional potentially relevant papers. Articles were initially filtered for English language and year (1990–2022) as this period corresponds to the contemporary era of psychedelic research and adheres more closely to the current standards of experimental research [19].

Studies on animals and humans were included in the review if they contained RDoC-compatible outcome measures related to the PVS domain. In human studies, both healthy and patient populations with any psychiatric diagnosis, were included. The relevance of the selected articles in the previous step was assessed in two phases. During the initial identification phase of screening, studies were excluded if they fell into the following categories: (i) an irrelevant type of research report, such as review papers, commentaries, seminar abstracts, book chapters etc., (ii) on an irrelevant topic, such as chemistry, botany, anthropology, etc., or (iii) not available in full text. The identification phase yielded 487 articles. In the second phase and after exclusion of duplicate records, three reviewers (NP, AK, FYS) independently screened the titles and abstracts to identify only those studies with an LSD-elicited outcome measure related to the PVS domain. Accordingly, studies were excluded if they (i) did not incorporate comparators or reference groups such as cross-sectional studies, case studies, descriptive studies, observational studies, qualitative research, exploratory research, ecological studies, survey research, and content analysis studies (ii) only contained qualitative assessments such as qualitative interviews, (iii) only reported correlational analysis, (iv) contained duplicate data from an included study, e.g., in cases of multiple studies from the same study population, or (v) only contained outcome measures that were related to other RDoC domain (e.g., cognitive processes), were non-domain-specific (e.g., diagnostic measurements like Beck depression inventory), or informed more than one domain (e.g., whole brain analysis). The inclusion and exclusion criteria were also applied to the extracted findings from the included articles. In other words, data were considered only if they met the criteria mentioned earlier. Therefore, not all outcome measures from the included studies are presented in the results. The reviewers resolved disagreements through discussion. Forty-two studies were included in the review after the screening phase (Fig. 1). Data from the included studies were indexed by: the name of the author(s), year of publication, study characteristics (type of trial, study design), intervention features (i.e., dose, route of administration, number of doses), characteristics of the participants (gender, age, mental health condition, sample size), response criteria, units of analysis, and the durability of the effect (e.g., long- or short-term). Further details about the review’s methodology were elaborated and explained elsewhere [18].

Fig. 1
figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) study selection flow diagram. Non-PVS: non-positive valence system

Full size image

Relevant data were extracted from each study, with a focus on aforementioned key study characteristics. Thematic analysis was then conducted to identify common themes, patterns, and trends across studies by systematically coding and categorizing data related to study findings. Then risk of bias assessment was conducted (see Risk of bias assessment section). Accounting for contextual factors such as study settings and participant characteristics, etc., the key outcomes, proxy measures relevant to the RDoC-positive valence system (denoted with an asterisk), trends, and variations among studies were highlighted, and finally the overall conclusions regarding the effects of intervention on the PVS were drawn. The certainty of evidence for each outcome was determined, considering the collected data on study design, risk of bias, consistency, precision, and potential bias. This structured approach ensured a transparent evaluation of evidence reliability in our systematic review.

2.1 Risk of Bias Assessment

The quality of included studies was assessed by two reviewers, FYS and AK, using the National Toxicology Program’s Office of Health Assessment and Translation (NTP OHAT) risk of bias (RoB) tool [20]. The NTP OHAT tool employs 11 questions to assess potential bias sources within six domains, namely selection bias, confounding bias, performance bias, attrition/exclusion bias, detection bias, and selective reporting bias, all evaluated at the outcome level. For each bias domain, pre-specified rating criteria were used to answer as definitely low (++); probably low (+); probably high/not reported (−/NR); or definitely high (−−) RoB (Table 1). Rating discrepancies were addressed through a consensus reached between the reviewers.

Table 1 Risk of bias assessment of included studies. NA: not applicable, NR: not reported
Full size table

3 Results

3.1 Included Assessments

Following the inclusion criteria, 42 studies were included, consisting of four units of analysis, namely self-reports, paradigms, molecules (biochemical components such as neurotransmitters and receptors), and circuits (neural circuitry) (Tables 2 and 3).

Table 2 The effects of LSD on positive valence systems (human studies)
Full size table
Table 3 The effects of LSD on positive valence systems (animal studies)
Full size table

Twenty of the included studies measured LSD-elicited effects with the “self-reports” units of analysis (47.62 %), 17 with the “paradigms” units of analysis (40.48 %), ten with the “molecules” units of analysis (23.81 %) and five with the “circuits” units of analysis (11.90 %).

Thirteen unique self-reports were extracted. The positive affect subscale of the Positive and Negative Affect Scale (PANAS) was the only extracted outcome measure that was directly proposed by the RDoC task force [21]. This questionnaire evaluates the extent to which an individual experiences a positive mood such as joy, interest, or enthusiasm and is sensitive to signals of reward [22]. The positive affect (PA) subscale can reflect reward reactivity and was hence categorized in the reward responsiveness construct in the RDoC matrix [22, 23]. Five studies included PANAS as a self-report measure [24,25,26,27,28]. Encouraged by the flexible approach taken by the RDoC task force aimed at using the principles of the framework to integrate and define elements of the RDoC matrix, we have identified an additional 12 types of proxy measures that could similarly gauge altered mood consequent to a perceived reward [29]. Taking these newly identified proxy measures into consideration, 45 self-report measures were extracted from the included studies that investigated the effect of LSD on the PVS.

The most widely used proxy self-report was the Altered States of Consciousness questionnaire (ASC) and its variations [30, 31]. The ASC evaluates the subjective effects of psychedelic drugs in a retrospective manner and constitutes 19 (42.22 %) of the extracted outcome measures. The subscale “blissful state” assesses the positive mood induced by rewarding experiences during the peak drug effect [31]. We also included subjects’ and investigators’ rating of mood expressed on different scales such as the visual analogue scale (VAS) (eight outcome measures, 17.78%). Other commonly used proxy measures obtained from self-reports were the “vigor” subscale of the Profile of Mood State (POMS) questionnaire (four outcome measures, 8.89 %), the “emotional excitation” and “well-being” items of the Adjective Mood Rating Scale (AMRS) (four outcome measures, 8.89%), subscales “deeply felt positive mood” and “positive mood” from variations of the Mystical Type Experience Questionnaire (MEQ) (five outcome measures, 11.11%), subscales “ positive mood changes”, “positive attitude toward life and self”, “positive behavior changes” and “well-being and life satisfaction” from the Persisting Effect Questionnaire (PEQ) (one outcome measure, 2.22%), item “want more of you received” from drug effect questionnaire (DEQ) (one outcome measures, 2.22%), and subscale “optimism” from the revised life orientation test (one outcome measures, 2.22%) [32,33,34,35,36,37].

Paradigm units of analysis provided insights for all three constructs of the PVS, namely, reward responsiveness, reward learning, reward valuation. Included under the reward learning construct in the RDoC framework, conditioning paradigms evaluate learning behavior as modified by a present or prospective rewarding stimuli [23, 38]. Moreover, we have included nine paradigms as proxy measures. Two paradigms suggested as proxy measures for reward responsiveness are the electrophysiological monetary incentive delay task derived from human studies, and sucrose preference test from the animal studies. Two of the proxy paradigms, eyeblink conditioning task and probabilistic reward learning task, could potentially help determine whether LSD impacts the reward learning construct. Five proxy measures included under the reward valuation construct were the forced swim test, the reinstatement of self-administration of alcohol, the intracranial self-stimulation (ICSS) test, animal model of the Iowa gambling test, derived from preclinical studies, and the Cambridge gambling task, derived from clinical studies. The distribution of the included human and animal studies based on the PVS constructs (reward responsiveness, learning and valuation) across units of analysis (self-report, paradigm, circuit, and molecule) is summarized in Fig. 2.

Fig. 2
figure 2

Distribution of outcome measures extracted from human and animal studies by Research Domain Criteria (RDoC) units of analysis (self-report, paradigm, circuit, and molecule) and constructs of the positive valence system (PVS)

Full size image

The isolated involvement of the remaining units of analysis, molecules, and circuits, in the PVS is not feasible, given their potential relevance to multiple domains and constructs. For instance, the serotonergic system is involved in all six of the RDoC domains [39]. Consequently, reporting their involvement with a certain domain and construct is context-dependent and relies on supplementary higher-level analyses. For example, examining the impact of the serotonergic system on the positive affect subscale of the PANAS narrows the focus to its role within the context of the PVS. Accordingly, we have discussed circuits and molecules in their pertinent contexts.

3.2 Human Data

Twenty-eight unique clinical studies involving 477 participants were identified. As for the reported units of analysis, 23 studies assessed LSD-induced alterations of the PVS using self-reports, while five studies employed the molecules units of analysis, four examined circuits, and three utilized paradigms (Fig. 2). Twenty-seven studies were conducted with healthy participants primarily in biomedically oriented lab settings encompassing a wide range of psychometric, neurobehavioral, and brain imaging assessments. Only one study consisted of results from a clinical trial, which investigated the effects of LSD on patients suffering from anxiety disorders [40]. It is important to highlight that in all the included studies, the dosage range for LSD administration varied from 5 to 200 mcg, which was administered via and different routes, including oral, sublingual, and intravenous.

Besides self-reports, five of the studies reported molecular units of analysis, examining the effects of a 5-HT2A receptor antagonist, ketanserin, on the subjective effects of LSD. Additionally, we extracted results from three paradigms. These paradigms assessed the reward processing systems and included the Cambridge gambling task, probabilistic reversal learning task, and electrophysiological monetary incentive task. We will explore their potential relevance to distinct constructs within the PVS in the corresponding sections.

All the included studies utilized self-report questionnaires (psychometric) to assess participants' hedonic responses to the LSD-induced experience. Our results have indicated a dose-response relationship, particularly when comparing very low (“micro-doses”) to high doses of LSD, as reported in self-reports, which constituted the majority of our findings from the human studies. In light of this observed dose-response relationship and to streamline the organization of our findings, we have classified the human data into two distinct categories based on dosage: full-dose and micro-dose. In the literature, a micro-dose is typically defined as 1/10th or 1/20th of recreational dosing, which, according to the studies included in this review, amounts to ≤ 26 mcg. Furthermore, data for each of these categories, whether full-dose or micro-dose, were grouped by their measurement class (units of analysis), including self-report, paradigm, molecule, or circuits, where applicable.

3.2.1 Full-Dose Studies

3.2.1.1 Self Reports

Forty-five self-reports were extracted from the included studies that investigated the effects of full-dose LSD on the reward responsiveness. This construct involves an individual's hedonic response to rewarding stimuli, which is reflected in various measures, including subjective reports [5].

The self-reports from the full-dose studies indicated that LSD improved mood in both the short (15 studies) and the long term (three studies). Considering the plasma half-life of LSD, which typically falls within the 3- to 4-hour range and utilizing the standard calculation for drug clearance based on five half-lives, we have classified outcomes into two categories based on timeframes. Evaluations conducted within 24 hours of drug administration were classified as short-term outcomes, while those conducted thereafter were categorized as long-term outcomes (Table 2) [41, 42]; it is crucial to note that the long-term assessments included in this review extend well beyond this initial 24-hour timeframe. Specifically, these long-term assessments span a period ranging from two weeks to 12 months post-administration.

Six studies evaluated the dose-dependent effect of LSD. Among these studies, two were conducted in biomedical laboratory settings and involved the same group of participants. These two studies specifically examined the effects of two LSD doses, namely 100 and 200 mcg, on various factors, including “well-being” as measured by the Altered States of Consciousness Rating Scale (AMRS) “deeply felt positive mood” assessed through the MEQ (Mystical Experience Questionnaire) scale, and subjective ratings of the item “happy” [43, 44]. Their findings suggested a significant (p < 0.05) difference only between the higher dose of LSD and placebo for related items of AMRS and MEQ scales. The scores of the latter scale also showed that 200 mcg of LSD was associated with higher scores of “deeply felt positive mood” when compared to the effects of a 75 mg single dose of 3,4-methylenedioxymethamphetamine (MDMA). Conversely, the effects of both doses of LSD were similar but significantly higher than placebo (p < 0.05) for the item “happy” in the subjective ratings. In another study, Schmid et al. pooled the dose-effect findings of various doses of LSD, specifically 100 and 200 mcg, on the different mood assessments in healthy and patient populations [40]. Intriguingly, in the patient population, the effects of various doses of LSD on items “blissful state” on 5D-ASC and “deeply felt positive mood” and “positive mood” on different variations of the MEQ scale were not significantly different from one another but were nevertheless greater than placebo. The effects of LSD on healthy participants, however, were dose dependent with the 200 mcg eliciting higher scores than the 100 mcg on the related items of 5D-ASC and MEQ scales. Moreover, the effects of both doses on the aforementioned mood-related scales were significantly higher than placebo (p < 0.05). In another double-blind placebo-controlled study, Holze et al. studied the acute effects of two doses of LSD (100 and 200 mcg) as well as two doses of psilocybin (15 and 30 mg) on measures of mood [45]. In contrast to previous studies, they found no significant differences between the effects of the two LSD doses (100 and 200 mcg) on the relevant items of AMRS (“emotional excitation”), 5D-ASC (“blissful state”) and MEQ scales (“deeply felt positive mood” and “positive mood”). Furthermore, there were no significant differences between either the dose of LSD and the psilocybin doses (15 and 30 mg). However, the aforementioned items for acute changes in mood were significantly higher than placebo (p < 0.05).

3.2.1.2 Molecules

The five studies that reported LSD effect at the molecular level used ketanserin, a 5-HT2A receptor antagonist, to investigate whether the 5-HT2A receptor subtype is involved in the subjective and neural effects of LSD [24, 25, 27, 46,47,48]. All five studies showed that a pretreatment of 40 mg ketanserin can reverse LSD-elicited increases in subjective mood as assessed by items “blissful state” in 5D-ASC and “positive mood” in PANAS. Preller et al. further studied how ketanserin can affect the changes in connectivity associated with LSD [48]. Their main goal was to identify LSD-induced effects on neural and behavioral measures. Preller et al. showed that LSD concurrently increased somatosensory and thalamic connectivity while reducing associative network connectivity, and that ketanserin prevented these LSD-induced changes in neural connectivity.

3.2.1.3 Paradigms

Two studies investigated the effects of full-dose LSD with paradigms [47, 49]. Pokorny et al. conducted the only study that assessed reward valuation using the Cambridge Gambling Task (CGT) paradigm [47]. This task mainly evaluates risk-taking behavior and reward-based decision making outside of a learning context [50]. The CGT shares similarities with the probabilistic selection task, an RDoC-recommended task for the reward valuation construct. Both tasks involve decision making under conditions of uncertainty and requiring participants to learn through trial and error as to which trial is associated with positive versus negative outcomes. Accordingly, we suggest CGT as a proxy measure within the reward valuation construct [7]. The authors found no effect of LSD on the task parameters, namely the quality of decision making or risk taking [47].

Kanen et al. lead another study that used paradigms to evaluate the effects of LSD on reward processing [49]. The authors investigated the effects of 75 mcg intravenous LSD (same study as Carhart-Harris et al. 2016 [51]) in a probabilistic reversal learning task (PRL) [52]. In a PRL task, participants are asked to identify, through trial and error, the stimulus most frequently associated with reward during the acquisition phase. During the reversal phase, the contingencies change, necessitating participants to adapt their choices based on the new conditions. We propose that this paradigm aligns with the reward learning construct. Reward learning is a form of reinforcement learning that involves organisms gathering information about stimuli, actions, and situations that forecast favorable results. It also encompasses the adjustment of behavior when a new reward arises or when outcomes surpass initial expectations [6]. The PRL task effectively aligns with the construct of reward learning as it replicates key elements of this process. In the acquisition phase, participants must identify stimuli associated with rewards, mirroring how organisms gather information about positive outcomes. The subsequent reversal phase, where stimulus-reward contingencies change, necessitates participants to adapt their choices, reflecting the behavior modification component of reward learning when outcomes deviate from expectations or novel rewards emerge.

In their analysis, the authors explored the effects of LSD on different reinforcement learning parameters based on either raw data measures of behavior or the parameters obtained by fitting reinforcement learning computational models to the behavioral data. The results on the raw data measures showed that LSD did not affect the sensitivity to the immediate feedback but increased perseveration, which quantified the number of preservative errors involving stimuli that were formerly correct but are incorrect in the reversal trial. In terms of computational model parameters related to reinforcement learning, LSD increased both reward learning and punishment learning parameters, demonstrating its influence on learning from positive and negative outcomes. Additionally, LSD reduced “stimulus stickiness,” indicating a decreased propensity for choice repetition, irrespective of the outcome. The impact on reinforcement sensitivity, which measures the influence of past rewards and punishments on decision making, was different in acquisition and reversal trials. LSD did not significantly affect reinforcement sensitivity during acquisition trials but increased it during reversal trials, suggesting varying sensitivity to past reinforcement in different decision-making contexts.

3.2.1.4 Circuits

In two studies that also performed circuit-level analysis, notable associations were observed between self-reported positive mood states and functional connectivity as assessed by functional magnetic resonance imaging (fMRI) [53, 54]. Luppi et al. found that LSD-induced increases in “blissful state” are associated with increased “small-world” neural network organization, which is a pattern of brain activity that typically decreases in states of impaired or lowered consciousness [54]. In another study, Preller et al. found that LSD-induced increases in “blissful states” is correlated with elevated somatosensory connectivity [48].

3.2.2 Micro-dosing

3.2.2.1 Self Reports

Five of the included studies sought to determine the effects of micro-doses of LSD on mood, among other assessments [26, 28, 53, 55, 56]. The results of self-reports from all five studies shown non-significant effects with doses ≤ 20 mcg, suggesting that the subjective mood-enhancing properties of LSD are dose dependent.

Bershad et al. investigated the acute effects of single, low doses of 6.5, 13, and 26 mcg [26]. Following the administration of 13 and 26 mcg of LSD, the item “vigor” on POMS scored significantly higher than placebo (p < 0.05) only at the highest dose, and the item “blissful state” on 5D-ASC scored significantly higher than placebo (p < 0.05) after the administration of 13 and 26 mcg of LSD. None of the doses affected item “elation” on the POMS. The effects of the 13-mcg dose on measures of mood were not replicated in their later study as there was no significant difference between LSD and placebo on items “positive mood” on PANAS and “blissful state” on 5D-ASC scales [53]. In another study, Hutten et al. examined the acute effects of 5, 10, and 20 mcg of LSD on mood in a double-blind, within-subject, placebo-controlled study. They observed an enhancement of mood following the administration of 20 mcg, but not with the lower dosage of LSD, as measured by items “positive mood” on POMS and “happy” on subjective rating scales. None of the administered doses had an impact on the item “blissful state” on 5D-ASC [57].

In a study conducted by de Wit et al., the effects of four repeated doses of LSD, either 13- or 26 mcg, or placebo, were investigated on different subjective assessment of mood, namely, POMS, PANAS, DEQ, and 5D-ASC [28]. Their results suggest that only 26 mcg of LSD increased the scores on items “vigor” on the POMS and “blissful state” on the 5D-ASC. No doses affected the relevant subscales of mood of PANAS or DEQ. Moreover, they found no residual effect on the drug-free follow-up session.

Murray et al. investigated the effects of 13- and 26-mcg doses compared to a placebo on subjective measures of mood indexed by POMS, DEQ, and 5D-ASC in healthy volunteers [56]. Murray et al. found that only 26 mcg of LSD produced perceptible subjective changes in mood, namely by increasing the score of items “vigor” and “elated” on the POMS questionnaire, “want more of what you received” on the DEQ, and “blissful state” on the 5D-ASC questionnaires [56].

3.2.2.2 Paradigms and Circuits

Two studies that investigated the effects of micro-doses of LSD also reported results from concurrent assessments using paradigms alongside neuroimaging (circuits).

In a complementary study to that of Murray et al., elaborated above, Glazer et al included neural measures of reward processing as assessed by task-based EEG [58]. While the results of Murray et al. indicated no subjective effects for doses lower than 26 mcg, Glazer et al. showed that 13 mcg of LSD led to neural changes in a reward-processing paradigm called the electrophysiological monetary incentive delay (eMID) task [58]. During this task, the event-related potentials (ERPs) of the EEG signals were recorded while processing positive and negative feedback under both reward and non-reward conditions. We interpreted the MID task as a proxy for measuring reward responsiveness because this task directly aligns with the RDoC guidelines, which emphasize assessing “neural activity in response to the receipt of rewards and reward cues [5]. The results of Glazer et al. suggested that at 13 mcg, LSD increased both the affective values of reward feedback (vs neutral feedback) as measured by late positive potential (LPP) amplitudes and hedonic impact of the positive feedback in reward conditions, as indexed by Reward Positivity (ReWP). They also showed that both doses of LSD, 13 and 26 mcg, increased the motivational salience of the stimuli by elevating the feedback P3 (FB-P3) amplitudes. Intriguingly, none of these neural measurements were associated with the subjective measures of mood in their respective studies. In a similar vein, in another study involving the “micro-dose” group, led by Bershad et al, no subjective effects were reported with the 13-mcg dose of LSD, but there was an observed increase in the correlation between amygdala-prefrontal connectivity and measures of positive mood [53].

3.3 Animal Data

All 14 included animal studies reported LSD effects using paradigm units of analysis. Five studies additionally included molecular outcomes, while one included outcome at the circuit level to assess the effects of LSD on the PVS. The studies encompass both short-term assessments (conducted on the administration day) and long-term assessments (conducted two days or more after administration). In these studies, all three RDoC-defined PVS constructs were addressed, although not all subconstructs were tested. Three of the identified studies employed measures of reward responsiveness [59,60,61], and the rest of the studies focused on the constructs of reward valuation [62,63,64,65,66,67], reward learning [68,69,70], or a combination of the two [71, 72].

3.3.1 Effects of LSD on Reward Responsiveness

Three of the included studies reported outcome measures related to the reward responsiveness construct using paradigms [59,60,61]. One study further reported LSD-induced modifications at the receptor level (molecule units of analysis) [59]. Two studies reported the results of the effects of LSD on place-preference conditioning and taste reactivity paradigms [60, 61].

Various designs of place-preference conditioning paradigms are employed to evaluate whether a specific drug can influence an animal’s preference for a particular environment [73,74,75,76]. If animals exhibit a preference for the drug-paired environment, it suggests that they find the drug’s effects rewarding. Such paradigms align more closely with the concept of reward responsiveness, specifically the immediate response to a rewarding stimulus, which, in this case, is the drug itself. While the primary focus of the place preference paradigm is on the immediate response to the drug’s rewarding effects (reward responsiveness), it also encompasses the aspect of learning and assigning value to the drug-paired context (reward valuation). Consequently, whether the primary emphasis aligns with reward responsiveness hinges upon the task design and research focus. When the initial chamber has a neutral value, and animals develop a preference or lack of preference based on exposure to a rewarding or aversive stimulus, the primary assessment centers on reward responsiveness. Such a paradigm primarily captures the immediate responses of animals to a new rewarding or aversive stimulus. However, in scenarios where animals already possess a pre-existing preference due to another rewarding stimulus, and the research intervention aims to alter that preference, the emphasis leans more toward measuring reward valuation. Here, the focus shifts from immediate responses to the long-term value assigned to the chamber based on the impact of the intervention on the animals’ preference.

Taste reactivity (TR) was the other paradigm used by the above-mentioned papers [60, 61]. Taste reactivity assesses an animal’s immediate, unconditioned facial and oromotor responses, such as tongue protrusion, lip licking, and facial grimacing, when exposed to various taste stimuli, such as sweet or bitter solutions [77]. Taste reactivity is closely related to the construct of reward responsiveness within the PVS, because it allows researchers to directly observe and quantify an animal’s real-time hedonic reactions to taste stimuli.

Using both place-preference conditioning and TR paradigms, Parker et al. tested the effects of different doses of LSD ranging from 0.025 mg/kg to 0.2 mg/kg on Sprague-Dawley rats. At 0.2 mg/kg, LSD produced a conditioned place preference, but this preference was prevented by a single pre-exposure to the conditioning chamber [61]. As for the TR paradigm, LSD doses of 0.05 mg/kg and 0.2 mg/kg produced less positive hedonic taste reactions than the saline control solution, but none of the doses induced aversive taste reactions.

In the other study that included place preference paradigm, Meehan et al. evaluated the effect of 0.2 mg/kg of LSD on male and female Fawn hooded rats [60]. They showed that LSD induced a preference for place only in male rats and not in female rats. These results suggest that LSD-induced rewarding experiences using the place conditioning and TR paradigm may vary based on sex. Moreover, the rewarding effects are weak relative to other drugs of abuse, such as cocaine and amphetamines [78]. These results encourage further investigations of LSD using TR and conditioned place preference paradigms.

In the third reward responsiveness study, Marona-Lewicka et al. tested chronic administration of 0.16 mg/kg of LSD on 16 male Sprague Dawley rats as part of their efforts to develop an animal model for psychosis [59]. Related to the current review, they specifically investigated the effects of LSD on a sucrose preference test as an indicator of anhedonia. Sucrose preference test is commonly used to test responsiveness to sugar, a natural rewarding stimuli for rodents [79, 80]. Compared to controls, their results showed significantly (p < 0.05) lower sugar consumption in the LSD group, tested 34 hours after the cease of chronic administration. Interestingly, there was a concomitant increase in dopamine receptor (DRD2) mRNA gene expression and a decrease in 5-HT2C receptor mRNA in the medial prefrontal cortex one month after LSD administration cessation.

3.3.2 Effect of LSD on Reward Learning

Overall, five animal studies reported outcome measures related to the reward learning construct. Three of the studies assessed the effect of LSD with eyeblink conditioning paradigms [68,69,70]. Eyeblink Classical Conditioning (EBC) is a well-established paradigm for the study of associative, or Pavlovian, learning. In EBC, an unconditioned stimulus (US), in this case a puff of air delivered directly to the eye, is paired with a conditioned stimulus (CS) such as light and tone [81, 82]. Eyeblink Classical Conditioning is widely recognized as motor associative learning, in which cerebellum plays a pivotal role [83, 84]. One must not overlook the fact that this paradigm, like many others, has undeniable contributions from other domains and constructs, such as the negative valence systems and sensorimotor systems domains [84,85,86]. However, in the context of this review, our primary focus centers on the associative learning aspect of EBC as a straightforward Pavlovian conditioning paradigm [87]. We have therefore classified it within the reward learning construct in the RDoC framework [88].

In the first study using the EBC, Welsh et al. examined the effects of 30 nmol/kg (9.7 mcg/kg) of intravenous (IV) LSD administration on EBC in 51 New Zealand white albino rabbits [70]. The researchers found that LSD administration improved acquisition of the eyeblink CR to both tone and light, while two other 5-HT1A receptor agonists (8-OH-DPAT; 50 and 200 nmol/kg (12.35 mcg/kg and 49.4 mcg/kg) and lisuride; 0.3–30 nmol/kg (0.1 mcg/kg–10 mcg/kg) employed in the study had no effect on CR acquisition. Furthermore, these effects were reversible by administering ritanserin, a selective 5-HT2A/2C antagonist [70].

In another eyeblink response paradigm, Harvey et al. gauged the effect of chronic administration (8 consecutive days) of 0.013 mg/kg of IV LSD on conditioned eyeblink responses in 16 male New Zealand rabbits [68]. In contrast to the findings of Welsh et al., the researchers found no significant differences in CR acquisition in LSD-treated rabbits compared to the control condition.

Romano et al. also studied the effects of eight sessions of intrahippocampal injection of LSD (1, 3, 10 and 30 nmol per hippocamp side) on conditioned eyeblink responses [69]. The results of their work suggest that LSD increases CR acquisition. While the 1-nmol dose of LSD per hippocampus side had no significant effect on CR acquisition compared to the control group, higher doses elicited a small but significant (p < 0.05) increase in CR acquisition on different days of the experiment. Additionally, higher doses of LSD resulted in slower CR acquisition. For example, the highest dose resulted in an increase in CR acquisition on Day 5, while the lowest dose had no effect on Day 2 (Table 3). Notably, the LSD-elicited increases in CR acquisition at the highest dose only reached significance on one day. Their findings also showed that LSD dose-dependently lowered the head bob counts elicited by a serotonin 5-HT2A receptor agonist, (±)-1(2, 5-dimethoxy-4-iodophenyl)-2-aminopropane hydrochloride (DOI). This reduction was interpreted as 5-HT2A receptor desensitization in the hippocampus resulting from chronic LSD administration. In the remaining two studies that utilized operant test batteries, there were outcome measures that pertained to both reward learning and reward valuation constructs, which will be addressed in Sect. 3.2.2.3 [71, 72].

3.3.3 Effect of LSD on Reward Valuation

Six animal studies employing paradigms to assess the reward valuation construct were identified [63,64,65,66,67, 89]. The paradigms used in these studies were forced swim test, intracranial self-stimulation, rodent model of Iowa gambling task, and alcohol relapse paradigms. In this section, we also incorporated the findings of two additional animal studies that employed an operant test battery with measures that had implications for both reward valuation and reward learning constructs [71, 72]. In the following paragraphs, we elaborate on each paradigm as a possible proxy to gauge reward valuation.

In a rodent version of the Iowa gambling task (IGT), Elsila et al. tested the effects of 0.025, 0.1, 0.2, and 0.4 mg/kg of LSD on previously learned reward-driven decision-making tasks [63]. The IGT is a widely used psychological task that assesses decision-making and risk-taking behavior under uncertain conditions [90]. A similar paradigm to the probabilistic selection task, which falls under the “reward (probability)” subconstruct in the reward valuation construct, the critical aspect of the IGT is that participants must learn through trial and error which decks are advantageous and which are disadvantageous [91, 92]. Therefore, the study by Elsila et al. was classified under the reward valuation construct.

The group showed that the lower doses of LSD had no effect on task parameters such as option selection and general functioning. At the highest dose (0.4 mg/kg), LSD increased omission rates (failure to respond within a determined interval) and decreased premature responding but had no effect on option selection [63]. In contrast, amphetamine decreased correct responses and premature responses while increasing omission rates.

Another study in this construct, Sakloth et al., explored the abuse potential of LSD using an intracranial self-stimulation (ICSS) paradigm [67]. In the ICSS paradigm, operant responding is reinforced by electrical pulses delivered directly into the brain [93]. This behavioral paradigm is able to avoid some inherent limitations associated with the use of food reinforcement paradigms, such as the possible anorectic effects of serotonergic drugs.

Like any other associative learning paradigm, ICSS also employs principles of reinforcement learning to establish a specific behavior, such as lever pressing to obtain a reward. The investigators then assess whether the animal assigns more value to the rewarding stimuli (in this case, brain stimulation) and increases its effort to receive the reward. In other words, they investigate whether the animal assigns more value to the rewarding stimuli (in this case, brain stimulation) and increases its effort to receive the reward. This effortful reward seeking is the primary reason for classifying the results based on principles of reward valuation rather than reward learning.

In this study, Sakloth et al. used short- and long-term LSD administration on ICSS facilitation in three dose schedules: acute-effect studies, repeated-dose studies, and time-course studies (Table 3). In the acute dose-effect and time-course studies, LSD administration was associated with inconsistent and weak ICSS facilitation, especially when compared to the ICSS facilitation by other drugs of abuse, such as methamphetamine. Intriguingly, the repeated-dose schedule of LSD caused no ICSS facilitation compared to baseline, while methamphetamine dose-dependently facilitated ICSS. However, the repeated-dose schedule of LSD attenuated the ICSS depression resulting from administration of a pure kappa opioid agonist test drug (U69,593).

Hibicke et al. investigated the long-term effects of a single LSD administration on the forced swim test (FST), a behavioral despair paradigm, in 12 male Wistar-Kyoto (WKY) rats. The FST assesses animals' active coping strategies, such as swimming or climbing, and passive coping strategies, namely immobility, in response to stress-inducing conditions [65]. Passive coping has been recognized as a significant contributor to major depression risk [94]; a decrease in immobility during this test is commonly interpreted as an antidepressant effect [95]. Notably, both of the outcome measures in this paradigm, reflecting effort invested in achieving a desired result, align with reward valuation and thus making them potential proxy outcome measures for this construct.

Hibicke et al. reported that LSD significantly (p < 0.05) increased “swimming” and “climbing” and decreased “immobility” when compared to placebo 35 days post-administration, and this was interpreted as evidence of long-term and persistent anti-depressant–like effects of LSD. To rule out nonspecific motor effects, researchers also compared locomotor activity and found no significant differences between the treatment and control animals.

The following three studies investigated the effects of LSD on alcohol-seeking behaviors using two different main alcohol access schedules. In two of the models, intermittent access to alcohol in the formats of the alcohol deprivation effect (ADE) model or binge-like intermittent access was assessed with LSD [64, 66]. On the other hand, the third model assessed the effects of LSD on a stable-access paradigm [62]. As we will see in the next section, the results of these two models are radically different.

Models of alcohol-seeking behavior can be used to probe reward valuation and its related mechanisms, including the assessment of reward probability, incentive salience, and the perceived effort essential for achieving rewards [7]. Specifically, when an individual confronts a situation where the likelihood of obtaining a reward they previously had access to is altered, the significance or importance assigned to that reward tends to change. For instance, in a sudden state of deprivation when the chance of obtaining a reward diminishes, the perceived importance of that rewarding stimulus is likely to intensify. Such heightened salience may compel the individual to invest more effort to secure the reward, despite the reduced likelihood. This phenomenon is evident in various paradigms, such as those assessing food-seeking behavior [96].

In an alcohol deprivation effect model, Meinhardt et al. investigated the effect of two doses of LSD (either 0.08 or 0.32 mg/kg) on alcohol consumption in 27 male and female Wistar rats, administered one week apart [66]. The ADE is a rat model of alcohol relapse resembling human relapse behaviors with elements of compulsive drinking, craving and increased demand after deprivation [97]. The results of this study revealed no differences in alcohol consumption between placebo and either of the LSD dosing groups.

In the other intermittent schedule study by Elsila et al., the short- and long-term effects of different doses of LSD (0.05 and 0.1 mg/kg) were tested on binge-like ethanol consumption [64]. Their results showed that only at 0.1 mg/kg, LSD decreased intermittent ethanol consumption. This effect was observed solely on the test day (short-term effect) and on no follow-up days (long-term effect). Additionally, they investigated the effect of different LSD doses on a range of reward-related behaviors, including a natural reward (sucrose), intracranial self-stimulation (ICSS) with and without a well-known positive reinforcer (3 mg/kg d-amphetamine), as well as homeostatic feeding behavior. LSD had no impact on either sucrose consumption or ICSS administration, with or without d-amphetamine intake; however, it reduced food and water intake, notably at the 0.1 mg/kg dose. These results imply that the LSD-elicited decrease in ethanol consumption is likely due to reduced consummatory behavior rather than attenuating reward-related behavior.

However, the third mode of alcohol-seeking behavior showed contrasting results. Alper et al. studied the effect of a single LSD administration, either at a dose of 0.025 mg/kg or 0.05 mg/kg, on alcohol consumption using a two-bottle choice alcohol-drinking paradigm in male C57BL/6J mice [62]. Their results showed that only the 0.05 mg/kg dose of LSD reduced alcohol consumption by 17.9 % over 46 days following administration, which was measured at 24-hour intervals post-administration, along with fluid intake, until the study conclusion on Day 46. Fluid intake and locomotor activity were shown to be unaffected by LSD. Additionally, alcohol preference, defined as the proportion of alcohol intake relative to the total fluid intake (20 % v/v ethanol solution + water), was also reduced by 0.05 mg/kg of LSD.

Lastly, two studies that employed operant test batteries contained outcome measures that could be classified under both the reward learning and reward valuation constructs [71, 72]. Operant conditioning is a form of associative learning that focuses on the consequences of actions and how they impact the likelihood of future behavior [98, 99]. In operant conditioning, two main measures are often described: accuracy and response rate. Accuracy reflects precision of actions and the extent to which learned behaviors have been associated with specific outcomes, ergo falling under the reward learning construct [6]. The response rate measures speed and frequency at which individuals execute these actions, representing the effort individuals are willing to invest in obtaining a known reward. Accordingly, this measurement is more related to the “effort” subconstruct of the reward valuation construct [100].

Frederick et al. explored the effects of different doses of LSD (0.0003, 0.001, 0.003, 0.01, and 0.03 mg/kg) on learning and motivational behaviors using operant test battery (OTB) performance on six male Rhesus monkeys (Frederick et al., 1997). The test battery included both accuracy and response rate, which, as mentioned, can inform both reward learning and valuation constructs. Their result showed that at doses higher than 0.01 mg and 0.003 mg/kg, LSD significantly (p < 0.05) decreased the response rates in the motivation and learning tasks, respectively. Despite the decreased response rates, the monkeys’ accuracy in the learning task remained intact, suggesting that LSD did not affect their learning ability. During the progressive ratio motivation task, LSD only induced pauses in response at the highest dose (0.3 mg/kg), suggesting that this high dose reduced motivation for food, perhaps reflecting the anorectic effects of serotonin agonists. Further, as the pausing effect did not occur at doses lower than 0.3 mg/kg, it appears that this pausing effect did not play a role in the lowered motivation at 0.003 mg/kg. The fact that the monkeys continued the OTB despite lowered motivation might imply that OTB performance is not exclusively driven by food reinforcement. The percent task completed (PTC) in both tasks was decreased only at doses higher than 0.01 mg/kg for the motivation task and 0.03 mg/kg for the learning task.

Similarly, Kinney et al. investigated the effects of different doses of LSD (0.03, 0.1, 0.3 and 1 mg/kg) on a response duration differentiation (RDD) task in Long-Evans rats [72]. The RDD is an operant-response paradigm in which pressing and holding a lever for between 1.0 and 1.3 seconds is reinforced by food. In their study, Kinney et al. reported outcome measures that fall within both the reward learning (response accuracy) and the reward valuation (response rate) constructs. They found that LSD significantly (p < 0.05) reduced response rates, but it did not affect accuracy or mean response duration, and responding was completely suppressed at 1 mg/kg, suggesting a non-specific effect.

4 Discussion

Our review found that LSD affects positive valence systems, as evidenced by multi-level data collected from preclinical and clinical studies. Self-reports were the most commonly used measurements, followed by molecular analyses, paradigms, and circuit-level analyses. The self-reports obtained from the human studies provided valuable insights into the impact of LSD on the reward responsiveness construct. These reports revealed that LSD has both short-term and long-term mood-enhancing properties. These mood-enhancing effects seem to be mediated by the 5-HT2A receptor, a key serotonin receptor subtype. The subjective effects of LSD exhibited a clear dose-dependent pattern when comparing the outcomes of micro-dose studies. Notably, no noticeable subjective effects were reported for doses below 20 mcg of oral LSD. The results of the animal studies investigating the reward responsiveness construct suggest that LSD may serve as a weak reinforcer, potentially contributing to a decrease in responsiveness to sucrose, a natural reinforcer for rodents.

The reported impact of LSD on reward learning was inconsistent. Additionally, there is evidence to suggest that LSD may decrease previously learned reinforcement with other reinforcers, as indicated by data from both human and animal studies focusing on the reward valuation construct. In the upcoming sections, we will delve into the results of combined studies encompassing both animals and human subjects, examining each construct and analyzing data across reported units of analysis.

4.1 Effects of LSD on Reward Responsiveness

Reward responsiveness consists of processes that regulate individuals’ reactions to existing and impending positive reinforcers [23, 101]. Deficits in reward responsiveness have long been suggested as a significant contributor to functional impairment across various diagnostic categories, and current treatment options appear to yield only limited benefits [102].

Consequently, it is believed that aberrant reward responsiveness is associated with low response rates to both pharmacological and non-pharmacological treatment [103,104,105]. Twenty-three human studies (n = 477) and three animal studies (n = 249; rats) have documented outcome measures that could be pertinent to the reward responsiveness construct.

The short- and long-term assessments presented in the current review imply that LSD enhances hedonic states, as indicated by self-reported increases in mood from the human studies. The results from the microdosing studies (n = 142) that investigated the effects of low doses of oral LSD ranging from 5 to 26 mcg suggested a dose-dependent effect on the subjective measures of mood. These results generally showed that the effect of LSD is not subjectively perceivable at doses below 20 mcg. Interestingly, the results of two complementary assessments in two of the microdosing studies conducted by Bershad et al. and Glazer et al. (n = 42) suggest that the effects of can be detectable with neuroimaging despite no perceptible subjective effect on mood [53, 58]. Considering the potential use of sub-perceptual doses in clinical settings, whether as a placebo or as an actual treatment for sensitive patient populations, there is a clear need for future studies that incorporate complementary assessments across various units of analysis, especially neuroimaging.

The limited available data in this review also indicates that the administration of LSD is linked to rapid yet enduring changes in behavior, mood, and attitude. These changes persist well beyond the pharmacological half-life of the substance, as evidenced by measures such as increases in the item “optimism” from the revised life orientation test observed two weeks after administration [106]. These lasting effects align with the observations of other classical hallucinogens [107]. For instance, Agin-Liebes et al. have shown that reductions in demoralization and hopelessness, along with increases in positive mood and behavior, can endure up to four and a half years after psilocybin-assisted psychotherapy in cancer patients with anxiety or depression [108]. Similarly, initial studies of LSD treatment for depressive and anxiety symptoms in patients with life-threatening illnesses have indicated rapid and enduring improvements [109, 110]. Based on the foregoing data it would be worthwhile to study how LSD might benefit reward responsiveness in the clinical population, especially compared to available pharmacological therapies that require chronic administration for demonstrable clinical efficacy. It is crucial to recognize the significant contributions of psychotherapy in the effectiveness of psychedelics [111, 112].

The studies included in this review further emphasized the critical role of the 5-HT2A receptor in mood regulation. This was evidenced by the blocking of LSD-induced subjective mood enhancement through ketanserin, a selective 5-HT2A receptor antagonist [113]. These findings align with prior research on classical hallucinogens. For instance, Kometer et al., found that psilocybin-induced alterations in mood, cognition, perception, and understanding of self and others were inhibited by pretreatment with ketanserin [114]. In addition to the subjective effects, one of the included studies conducted by Preller et al. (n = 24) suggested that ketanserin blocks LSD-associated alterations in somatosensory, thalamic and associative network connectivity [48]. Further research is imperative to comprehensively understand the role of various neurotransmitters in hedonic responses to rewards, including dopamine and glutamate [115, 116].

The included studies suggest an association between alterations in mood and other RDoC domains, namely the social, cognitive, and sensorimotor systems. Notably, several studies (n = 89) have shown that LSD-elicited changes in the scores of “unity”, “oceanic boundlessness”, “auditory alterations”, “primary process thinking”, along with changes in “somatosensory network connectivity”, among others, were associated with alterations in acute and long-term changes in mood [44, 46, 48]. Despite the fact that such associations cannot be inferred as causal, they nonetheless appear to indicate alterations in normal patterns of waking consciousness. Such alterations, especially changes in perception of self and surroundings, can result in acute and persisting beneficial effects on the PVS [117]. Some authors have argued that this effect is attributable to increased psychological flexibility or mental plasticity, and neural plasticity during acute psychedelic experience [55, 118, 119]. Such psychological flexibility can, in turn, facilitate change and learning mechanisms during psychotherapy. Overall, mechanistic studies that encompass assessments across multiple units of analysis to investigate how LSD affects the reward responsiveness construct represent an exciting area for future research, although direct evidence for these explanatory models remains limited at the present time [10].

It is important to note that in human psychedelic trials, various non-pharmacological and contextual factors, such as “set” and “setting”, play a significant role in shaping the overall quality of the psychedelic experience [120,121,122]. Thus, it is plausible that some of the reported effects of LSD on positive valence systems may be attributed to these contextual factors. However, it is beyond the scope of this review to draw any firm conclusions regarding the influence of extra-pharmacological variables, as most of the reviewed trials were conducted within standardized biomedical research settings with prescreened healthy volunteers without therapeutic interventions specifically aimed at inducing positive experiences. It is essential to recognize that next to the pharmacological effects, the extent of concomitant psychological support may also contribute to the reported positive outcomes of psychedelic drug trials.

It is also worth mentioning that LSD can elicit adverse reactions in certain individuals, and the ability to predict who may experience positive or adverse drug effects remains challenging. While the consideration of factors like the setting, dosage, and individual risk profiles has generally contributed to a high level of safety, the identification of reliable predictors for distinguishing beneficial from adverse reactions to psychedelics remains an ongoing research priority [120]. Further investigation into extra-pharmacological factors is needed to broaden our understanding of how psychedelics, acting through 5-HT2A receptors, heighten sensitivity to contextual cues. For example, there is evidence that LSD enhances music-evoked emotions and that curated music during a psychedelic session may offer optimal supportive for peak experiences [123, 124].

Animal studies examining reward responsiveness have demonstrated that LSD can function as a weak reinforcer, supporting conditioned place preference in specific circumstances and not inducing conditioned place aversion [60, 61]. This observation aligns with the idea that LSD possesses rewarding properties, although the strength of this effect may vary. Furthermore, these studies have revealed that LSD, particularly under chronic administration conditions as examined by Marona-Lewicka et al. (n = 16; male Sprague Dawley rats), can lead to a reduced responsiveness to other rewards, as evidenced by the sucrose preference test [59]. It is worth noting that the studies generally exhibited a degree of consistency in the observed effects of LSD across different research contexts. However, as is often the case with complex behavioral phenomena, some behavioral effects were found to be relatively weak and inconsistent from one test to the next. This suggests that the outcomes are influenced by various factors, including the specific experimental conditions, dosage levels, and potentially individual differences among the animals. Overall, these results underscore the multifaceted nature of LSD’s impact on reward responsiveness. While it can act as a weak reinforcer in certain contexts, it may also influence anhedonia-related behavior by diminishing responsiveness to natural rewards (sucrose). These findings open up avenues for further investigation, including investigating the underlying neurobiological mechanisms and assessing how individual differences and experimental conditions modulate LSD’s effects on reward processing.

4.2 Effects of LSD on Reward Learning

Reward learning is a multifaceted process in which organisms acquire information about stimuli, actions, and contexts that predict positive outcomes, leading to adaptive behavioral modifications when novel rewards are encountered, or outcomes exceed expectations. This process encompasses subconstructs like probabilistic and reinforcement learning, reward prediction error, and the development of habitual behaviors within the PVS domain, offering insights into the complex interplay of cognitive and neural mechanisms in response to rewarding experiences [6]. Five animal studies (n = 177) and one human study (n = 15) reported outcomes from paradigms assessing the effect of LSD on the reward learning construct.

Three animal studies were conducted on 156 New Zealand white albino rabbits and focused on assessing the effect of LSD on the eyeblink conditioning response, which is a classical Pavlovian paradigm [87]. Two studies by Welsh et al. and Harvey et al., comparable in terms of administration (both systemic), yield contrasting results. While Welsh and colleague (n = 51) showed a significant (p < 0.05) effect of LSD on the eyeblink conditioning paradigm, the study of Harvey and colleague (n = 16) demonstrated no effect. In the third study conducted by Romano et al. (n = 89), LSD was administered through bilateral intrahippocampal injections, employing a paradigm similar to that used by Harvey et al. [69]. This study evaluated the impact of LSD on conditioned responses (CRs) over eight consecutive conditioning sessions. The findings indicated that varying doses of LSD had distinct effects on CRs, with lower doses influencing CRs during the earlier sessions and higher doses during the later sessions. On the molecular level, Welsh et al. further showed that ritanserin blocked the CR learning, implying a role for 5-HT2A in this form of associative learning, which is in agreement with other studies suggesting a possible role of this receptor in associative learning [125, 126]. The remaining two animal studies (six monkeys and 15 rats) that utilized an operant test battery to assess the effect of LSD on reward learning yielded inconsistent results [71, 72]. In one study, it was observed that LSD, administered at a dose of 0.03 mg/kg IV, resulted in a decrease in the accuracy of the learning task [71]. However, in the other study on 15 rats, even with higher doses of LSD, there was no discernible effect on task accuracy [72].

In summary, animal studies examining the effects of LSD on reward learning yield inconclusive and weak evidence, only hinting at a potential influence of LSD on associative learning. It is important to consider that the paradigm employed in all three studies includes substantial contributions from other domains, e.g., the sensorimotor and negative valence systems [127, 128]. The inconsistent results in studies assessing the effects of LSD on reward learning underscore the necessity for more selective and precise research paradigms that can directly gauge this crucial construct.

The result of the only human study for this construct revealed interesting results [49]. In their study, Kanen et al. investigated the effects of 75 mcg intravenous LSD on 15 participants through a probabilistic reversal learning task. Their findings indicated that LSD increased perseveration, characterized by a heightened tendency for persistent choices even when they became incorrect during reversals, without affecting the sensitivity to reward. Additionally, LSD amplified reward and punishment learning while reducing choice repetition, suggesting alterations in decision-making strategies. The study's findings, particularly the increased perseveration toward incorrect stimuli induced by LSD, emphasize the significance of future research dedicated to explore whether LSD alters the brain's ability to adapt to new information and adjust strategies in response to changing circumstances. Such investigations could provide valuable insights into the drug's impact on decision making and cognitive flexibility. Understanding these mechanisms may have implications for both therapeutic applications and safety considerations associated with LSD use.

Overall, the results from the included studies in the reward learning construct are inconsistent but suggest a possible enhancement of associative learning following LSD administration. Consequently, the reward learning construct remains relatively understudied, highlighting the need for future research to bridge this gap. Such studies are essential, given the crucial implications of reward learning across various diagnostic categories, particularly in conditions such as depression [129, 130]. This research is vital for gaining a comprehensive understanding of how psychedelics influence both healthy and aberrant reward learning processes.

4.3 Effects of LSD on Reward Valuation

Reward valuation encompasses the cognitive and neural processes involved in assessing the value of various facets of rewards, such as reward probability, incentive salience, and notably, the perceived effort necessary to attain those rewards [7]. Understanding the intricate interplay of these mechanisms helps elucidate individual differences in motivation, decision making, and susceptibility to mental health disorders, shedding light on the multifaceted nature of reward processing [131]. Seven distinct studies were identified: six involving animal subjects (n = 232) and one involving human subjects (n = 25). Each of these studies offer possible proxies to gauge the effects of LSD on the reward valuation construct.

Two of the included studies investigated the effects of LSD on the reward valuation construct using gambling tasks, IGT in 15 mice and CGT in 25 humans [47, 63]. The foregoing tasks are commonly used to simulate and assess risk-based decision-making processes, which have implications in several psychiatric disorders, including different forms of addiction, either substance use disorders or gambling addiction, where individuals exhibit aberrant decision making concerning the weighing of negative and positive outcomes in decision processes [132]. Both studies suggest no effect of LSD on risk-based decision making, as measured by different task parameters. These include no effect on option selection, decreased premature responding in the IGT, increased deliberation time, and no effect on the quality of decision making and risk taking in the CGT. In the study by Elsila et al., which also included a positive control group using amphetamine, there was a significant (p < 0.05) decrease in correct responses, an increase in premature responding, and a higher omission rate [63]. The aforementioned findings align with the notion that the serotonin system has a complex role in decision making, and its dysfunctions might lead to aberrant impulsive behavior [133]. It is important to note that these decision-making processes also significantly involve executive functions, which, although relevant, fall outside the scope of the current review [90].

In the remaining four studies, the paradigms primarily assessed LSD-elicited alterations in effort expenditure to obtain a previously learnt reward and were therefore proposed as proxies for the reward valuation construct. Hibicke et al. evaluated the long-term effect of LSD (n = 12 rats) on measures of active and passive coping, namely climbing and immobility [65]. In another study conducted by the same research group, they discovered findings consistent with these results, further supporting their observations. In this study, they examined the effects of high- and low-dose psilocybin on a Drosophila model for the FST [134]. The results revealed that both doses of psilocybin led to a reduction in immobility, aligning with their previous findings of decreased immobility as an outcome indicative of antidepressant effects. These findings raise the possibility that the FST could serve as a sensitive and face-valid tool for assessing the effects of psychedelic agents, although further replication is warranted to establish the robustness of these results.

Unlike other drugs of abuse, such as methamphetamine, Sakloth et al. (n = 39; rats) found only a weak and inconsistent effect of LSD on ICSS facilitation. Moreover, the repeated high-dose schedule of LSD attenuated the anhedonia-related ICSS depression elicited by opioids. None of the schedules of LSD administration had any effect on amphetamine-elicited ICSS facilitation [67]. These results align with previous literature, demonstrating that 5-HT2A and 5-HT2C receptor agonists (TCB-2 and WAY-161503) have a distinct effect profile, increasing rather than decreasing the ICSS thresholds, as observed with psychostimulants [135]. Given the entanglement of the current legal status of psychedelics with a potential for shared abuse profiles with other addictive drugs, such as psychostimulants, future studies are encouraged to investigate the possible distinct effects of LSD via such paradigms [136].

The following three studies reported the effects of LSD on variations of alcohol relapse paradigms and showed inconsistent outcomes. While Alper et al. demonstrated both short- and long-term (46 days) decreases in ethanol consumption and preference with 0.05 mg/kg intraperitoneal (IP) injections of LSD (n = 24 to 30 rats), Elsila et al. (n = 109 rats) found only a short-term decrease (specifically at 0.1 mg/kg IP), and Meinhardt et al. showed no significant change with either of the doses, 0.08 and 0.32 mg/kg (n = 27 rats). According to the authors, this discrepancy in results might stem from the different alcohol consumption schedules employed in the studies [64].

Additionally, two animal studies using an operant test battery with outcome measures to assess the effect of LSD on reward valuation yielded results supporting LSD-induced reductions in effort expenditure for another reinforcer [71, 72]. In both studies (on 6 monkeys and 15 rats), there was a reduction in response rates, which may serve as a proxy for the increased effort required to attain a reward following LSD administration.

Integrating findings from the aforementioned studies that assessed LSD's impact on reward-seeking behavior in response to different types of reinforcers suggests a possible unique profile for LSD in substance use disorders (SUD). In support of this concept, recent human clinical trials have indicated that the classical hallucinogen psilocybin shows promise as a treatment for tobacco and alcohol use disorders when administered alongside concurrent psychotherapy [15, 137, 138]. Future animal studies, adopting an approach similar to that of Elsila et al., are needed to account for the possible anorectic and inhibitory effects of LSD as potential sources of confounding in task short-term outcomes. This emerging mechanistic research field provides an exciting frontier for investigating how LSD, or other classical hallucinogens, influence reward-seeking behavior in substance-use disorders. However, it is essential to approach this research with a comprehensive perspective, considering not only the pharmacological aspects of LSD but also the potential contributions of non-pharmacological factors (“set” and “setting”) in shaping the outcomes [139].

5 Limitations

Using the RDoC framework to synthesize the literature on psychedelic research has several limitations, which were elaborated in our recent review [18]. Given the novelty of both RDoC and psychedelic research, the studies included in this review were expectedly diverse in terms of their research question, and methodology, and were not designed according to the RDoC framework. Subsequently, two limitations ensue from synthesizing the extant LSD literature based on RDoC framework. First, each of the included outcome measures was extracted according to the best judgment and understanding of the authors about the relevance of each outcome measure to the RDoC constructs and domains, and ergo is tentative. Second, due to compiling the outcome measures from studies with heterogenous methodologies and study designs, including different study populations, the resulting synthesis is prone to biased outcomes, samples and analyses [140].

While we have made every effort to conduct a comprehensive search of the available literature, it is possible that some relevant papers may have been inadvertently missed during our review process. This could occur because of various factors, including but not limited to variations in database coverage, differences in indexing practices, and the availability of publications in languages other than English. Despite our rigorous search strategy, it is important to acknowledge the possibility that some relevant studies may exist but were not included in this review. We have attempted to mitigate this limitation by conducting searches in multiple databases and consulting with subject experts to ensure a thorough search, but the potential for missing papers remains a limitation.

Despite these limitations, this review has shown that exploiting a multifaceted framework to integrate multi-units of analysis has the potential to provide a mechanistic understanding of how psychedelic drugs affect reward processing systems. Hence, there is a clear need for future studies with intentional designs aimed at assessing the PVS using specific tasks, as demonstrated in the study by Pokorny et al. [47]. Accordingly, and considering the limited number of specialized measurements to assess the PVS, developing novel tasks and paradigms and designing studies with a special focus on the interaction of psychedelics and PVS is imperative. An example of such specialized tasks is the monetary incentive delay task (MIDT), a paradigm suggested by RDoC to assess the reward responsiveness construct of the PVS [141]. In order to maximize benefits of LSD on positive valence systems and minimize potential harms, forthcoming studies should also explore the critical importance of extra-pharmacological and contextual variables (e.g., set and setting) to optimize treatment models and inform good practice guidelines. Finally, in future studies, it is important to consider the reduced task engagement observed following psychedelic administration, a phenomenon akin to that reported in human studies [142].

6 Conclusion

In conclusion, our findings align with previous literature showing that classical psychedelics (5-HT2A agonists) enhance measures related to PVS in both healthy and patient populations. While limited studies on reward learning suggest that LSD may act as a weak reinforcer, its effect on reward valuation suggests a potential decrease in effort for attaining other reinforcers. Using an integrated multidisciplinary approach, we have highlighted a potential unique profile of the influence of LSD on reward processing systems. We have further identified underdeveloped areas of future research that can improve our understanding of how LSD affects reward processing. The study of psychedelics holds the potential to address substantial clinical needs, so researchers in mental health should continue this work as a priority.