Learning from sensory and reward prediction errors during motor adaptation

Voluntary motor commands produce two kinds of consequences. Initially, a sensory consequence is observed in terms of activity in our primary sensory organs (e.g., vision, proprioception). Subsequently, the brain evaluates the sensory feedback and produces a subjective measure of utility or usefulness of the motor commands (e.g., reward). As a result, comparisons between predicted and observed consequences of motor commands produce two forms of prediction error. How do these errors contribute to changes in motor commands? Here, we considered a reach adaptation protocol and found that when high quality sensory feedback was available, adaptation of motor commands was driven almost exclusively by sensory prediction errors. This form of learning had a distinct signature: as motor commands adapted, the subjects altered their predictions regarding sensory consequences of motor commands, and generalized this learning broadly to neighboring motor commands. In contrast, as the quality of the sensory feedback degraded, adaptation of motor commands became more dependent on reward prediction errors. Reward prediction errors produced comparable changes in the motor commands, but produced no change in the predicted sensory consequences of motor commands, and generalized only locally. Because we found that there was a within subject correlation between generalization patterns and sensory remapping, it is plausible that during adaptation an individual''s relative reliance on sensory vs. reward prediction errors could be inferred. We suggest that while motor commands change because of sensory and reward prediction errors, only sensory prediction errors produce a change in the neural system that predicts sensory consequences of motor commands.     

Dual Reward Prediction Components Yield Pavlovian Sign- and Goal-Tracking

Reinforcement learning (RL) has become a dominant paradigm for understanding animal behaviors and neural correlates of decision-making, in part because of its ability to explain Pavlovian conditioned behaviors and the role of midbrain dopamine activity as reward prediction error (RPE). However, recent experimental findings indicate that dopamine activity, contrary to the RL hypothesis, may not signal RPE and differs based on the type of Pavlovian response (e.g. sign- and goal-tracking responses). In this study, we address this discrepancy by introducing a new neural correlate for learning reward predictions; the correlate is called “cue-evoked reward”. It refers to a recall of reward evoked by the cue that is learned through simple cue-reward associations. We introduce a temporal difference learning model, in which neural correlates of the cue itself and cue-evoked reward underlie learning of reward predictions. The animal''s reward prediction supported by these two correlates is divided into sign and goal components respectively. We relate the sign and goal components to approach responses towards the cue (i.e. sign-tracking) and the food-tray (i.e. goal-tracking) respectively. We found a number of correspondences between simulated models and the experimental findings (i.e. behavior and neural responses). First, the development of modeled responses is consistent with those observed in the experimental task. Second, the model''s RPEs were similar to dopamine activity in respective response groups. Finally, goal-tracking, but not sign-tracking, responses rapidly emerged when RPE was restored in the simulated models, similar to experiments with recovery from dopamine-antagonist. These results suggest two complementary neural correlates, corresponding to the cue and its evoked reward, form the basis for learning reward predictions in the sign- and goal-tracking rats.     

When Does Reward Maximization Lead to Matching Law?

What kind of strategies subjects follow in various behavioral circumstances has been a central issue in decision making. In particular, which behavioral strategy, maximizing or matching, is more fundamental to animal''s decision behavior has been a matter of debate. Here, we prove that any algorithm to achieve the stationary condition for maximizing the average reward should lead to matching when it ignores the dependence of the expected outcome on subject''s past choices. We may term this strategy of partial reward maximization “matching strategy”. Then, this strategy is applied to the case where the subject''s decision system updates the information for making a decision. Such information includes subject''s past actions or sensory stimuli, and the internal storage of this information is often called “state variables”. We demonstrate that the matching strategy provides an easy way to maximize reward when combined with the exploration of the state variables that correctly represent the crucial information for reward maximization. Our results reveal for the first time how a strategy to achieve matching behavior is beneficial to reward maximization, achieving a novel insight into the relationship between maximizing and matching.     

Risky Decisions and Their Consequences: Neural Processing by Boys with Antisocial Substance Disorder

Background

Adolescents with conduct and substance problems (“Antisocial Substance Disorder” (ASD)) repeatedly engage in risky antisocial and drug-using behaviors. We hypothesized that, during processing of risky decisions and resulting rewards and punishments, brain activation would differ between abstinent ASD boys and comparison boys.

Methodology/Principal Findings

We compared 20 abstinent adolescent male patients in treatment for ASD with 20 community controls, examining rapid event-related blood-oxygen-level-dependent (BOLD) responses during functional magnetic resonance imaging. In 90 decision trials participants chose to make either a cautious response that earned one cent, or a risky response that would either gain 5 cents or lose 10 cents; odds of losing increased as the game progressed. We also examined those times when subjects experienced wins, or separately losses, from their risky choices. We contrasted decision trials against very similar comparison trials requiring no decisions, using whole-brain BOLD-response analyses of group differences, corrected for multiple comparisons. During decision-making ASD boys showed hypoactivation in numerous brain regions robustly activated by controls, including orbitofrontal and dorsolateral prefrontal cortices, anterior cingulate, basal ganglia, insula, amygdala, hippocampus, and cerebellum. While experiencing wins, ASD boys had significantly less activity than controls in anterior cingulate, temporal regions, and cerebellum, with more activity nowhere. During losses ASD boys had significantly more activity than controls in orbitofrontal cortex, dorsolateral prefrontal cortex, brain stem, and cerebellum, with less activity nowhere.

Conclusions/Significance

Adolescent boys with ASD had extensive neural hypoactivity during risky decision-making, coupled with decreased activity during reward and increased activity during loss. These neural patterns may underlie the dangerous, excessive, sustained risk-taking of such boys. The findings suggest that the dysphoria, reward insensitivity, and suppressed neural activity observed among older addicted persons also characterize youths early in the development of substance use disorders.     

Does General Motivation Energize Financial Reward-Seeking Behavior? Evidence from an Effort Task

We aimed to predict how hard subjects work for financial rewards from their general trait and state reward-motivation. We specifically asked 1) whether individuals high in general trait “reward responsiveness” work harder 2) whether task-irrelevant cues can make people work harder, by increasing general motivation. Each trial of our task contained a 1 second earning interval in which male subjects earned money for each button press. This was preceded by one of three predictive cues: an erotic picture of a woman, a man, or a geometric figure. We found that individuals high in trait “reward responsiveness” worked harder and earned more, irrespective of the predictive cue. Because female predictive cues are more rewarding, we expected them to increase general motivation in our male subjects and invigorate work, but found a more complex pattern.     

Intention Understanding in Autism

When we observe a motor act (e.g. grasping a cup) done by another individual, we extract, according to how the motor act is performed and its context, two types of information: the goal (grasping) and the intention underlying it (e.g. grasping for drinking). Here we examined whether children with autistic spectrum disorder (ASD) are able to understand these two aspects of motor acts. Two experiments were carried out. In the first, one group of high-functioning children with ASD and one of typically developing (TD) children were presented with pictures showing hand-object interactions and asked what the individual was doing and why. In half of the “why” trials the observed grip was congruent with the function of the object (“why-use” trials), in the other half it corresponded to the grip typically used to move that object (“why-place” trials). The results showed that children with ASD have no difficulties in reporting the goals of individual motor acts. In contrast they made several errors in the why task with all errors occurring in the “why-place” trials. In the second experiment the same two groups of children saw pictures showing a hand-grip congruent with the object use, but within a context suggesting either the use of the object or its placement into a container. Here children with ASD performed as TD children, correctly indicating the agent''s intention. In conclusion, our data show that understanding others'' intentions can occur in two ways: by relying on motor information derived from the hand-object interaction, and by using functional information derived from the object''s standard use. Children with ASD have no deficit in the second type of understanding, while they have difficulties in understanding others'' intentions when they have to rely exclusively on motor cues.     

Neural Sensitivity to Absolute and Relative Anticipated Reward in Adolescents

Adolescence is associated with a dramatic increase in risky and impulsive behaviors that have been attributed to developmental differences in neural processing of rewards. In the present study, we sought to identify age differences in anticipation of absolute and relative rewards. To do so, we modified a commonly used monetary incentive delay (MID) task in order to examine brain activity to relative anticipated reward value (neural sensitivity to the value of a reward as a function of other available rewards). This design also made it possible to examine developmental differences in brain activation to absolute anticipated reward magnitude (the degree to which neural activity increases with increasing reward magnitude). While undergoing fMRI, 18 adolescents and 18 adult participants were presented with cues associated with different reward magnitudes. After the cue, participants responded to a target to win money on that trial. Presentation of cues was blocked such that two reward cues associated with $.20, $1.00, or $5.00 were in play on a given block. Thus, the relative value of the $1.00 reward varied depending on whether it was paired with a smaller or larger reward. Reflecting age differences in neural responses to relative anticipated reward (i.e., reference dependent processing), adults, but not adolescents, demonstrated greater activity to a $1 reward when it was the larger of the two available rewards. Adults also demonstrated a more linear increase in ventral striatal activity as a function of increasing absolute reward magnitude compared to adolescents. Additionally, reduced ventral striatal sensitivity to absolute anticipated reward (i.e., the difference in activity to medium versus small rewards) correlated with higher levels of trait Impulsivity. Thus, ventral striatal activity in anticipation of absolute and relative rewards develops with age. Absolute reward processing is also linked to individual differences in Impulsivity.     

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants

Understanding typical and atypical development remains one of the fundamental questions in developmental human neuroscience. Traditionally, experimental paradigms and analysis tools have been limited to constrained laboratory tasks and contexts due to technical limitations imposed by the available set of measuring and analysis techniques and the age of the subjects. These limitations severely limit the study of developmental neural dynamics and associated neural networks engaged in cognition, perception and action in infants performing “in action and in context”. This protocol presents a novel approach to study infants and young children as they freely organize their own behavior, and its consequences in a complex, partly unpredictable and highly dynamic environment. The proposed methodology integrates synchronized high-density active scalp electroencephalography (EEG), inertial measurement units (IMUs), video recording and behavioral analysis to capture brain activity and movement non-invasively in freely-behaving infants. This setup allows for the study of neural network dynamics in the developing brain, in action and context, as these networks are recruited during goal-oriented, exploration and social interaction tasks.     

Motor Inhibition during Overt and Covert Actions: An Electrical Neuroimaging Study

Given ample evidence for shared cortical structures involved in encoding actions, whether or not subsequently executed, a still unsolved problem is the identification of neural mechanisms of motor inhibition, preventing “covert actions” as motor imagery from being performed, in spite of the activation of the motor system. The principal aims of the present study were the evaluation of: 1) the presence in covert actions as motor imagery of putative motor inhibitory mechanisms; 2) their underlying cerebral sources; 3) their differences or similarities with respect to cerebral networks underpinning the inhibition of overt actions during a Go/NoGo task. For these purposes, we performed a high density EEG study evaluating the cerebral microstates and their related sources elicited during two types of Go/NoGo tasks, requiring the execution or withholding of an overt or a covert imagined action, respectively. Our results show for the first time the engagement during motor imagery of key nodes of a putative inhibitory network (including pre-supplementary motor area and right inferior frontal gyrus) partially overlapping with those activated for the inhibition of an overt action during the overt NoGo condition. At the same time, different patterns of temporal recruitment in these shared neural inhibitory substrates are shown, in accord with the intended overt or covert modality of action performance. The evidence that apparently divergent mechanisms such as controlled inhibition of overt actions and contingent automatic inhibition of covert actions do indeed share partially overlapping neural substrates, further challenges the rigid dichotomy between conscious, explicit, flexible and unconscious, implicit, inflexible forms of motor behavioral control.     

Excitability of the Primary Motor Cortex Increases More Strongly with Slow- than with Normal-Speed Presentation of Actions

Introduction

The aim of the present study was to investigate how the speed of observed action affects the excitability of the primary motor cortex (M1), as assessed by the size of motor evoked potentials (MEPs) induced by transcranial magnetic stimulation (TMS).

Methods

Eighteen healthy subjects watched a video clip of a person catching a ball, played at three different speeds (normal-, half-, and quarter-speed). MEPs were induced by TMS when the model''s hand had opened to the widest extent just before catching the ball (“open”) and when the model had just caught the ball (“catch”). These two events were locked to specific frames of the video clip (“phases”), rather than occurring at specific absolute times, so that they could easily be compared across different speeds. MEPs were recorded from the thenar (TH) and abductor digiti minimi (ADM) muscles of the right hand.

Results

The MEP amplitudes were higher when the subjects watched the video clip at low speed than when they watched the clip at normal speed. A repeated-measures ANOVA, with the factor VIDEO-SPEED, showed significant main effects. Bonferroni''s post hoc test showed that the following MEP amplitude differences were significant: TH, normal vs. quarter; ADM, normal vs. half; and ADM, normal vs. quarter. Paired t-tests showed that the significant MEP amplitude differences between TMS phases under each speed condition were TH, “catch” higher than “open” at quarter speed; ADM, “catch” higher than “open” at half speed.

Conclusions

These results indicate that the excitability of M1 was higher when the observed action was played at low speed. Our findings suggest that the action observation system became more active when the subjects observed the video clip at low speed, because the subjects could then recognize the elements of action and intention in others.     

Research of a holiday kind: A clinical trial gone awry: the Chocolate Happiness Undergoing More Pleasantness (CHUMP) study

The randomized controlled trial is the “gold standard” for evaluating the benefits and harms of interventions. The Chocolate Happiness Undergoing More Pleasantness (CHUMP) study was designed to compare the effects of dark chocolate, milk chocolate and normal chocolate consumption on happiness. Although the intention-to-treat analysis showed that participants who received either dark or milk chocolate were happier than those who received no additional chocolate, the actual-consumption analysis showed that there were no differences between any of the groups. The reason for this result is that many participants switched groups mid-study because of their personal chocolate preferences. Although the CHUMP study was pleasurable, it demonstrated the difficulties associated with performing a truly blinded clinical trial.The randomized controlled trial is the “gold standard” for testing the beneficial and harmful effects of interventions. There has been a growing concern in the literature about the potential for bias during randomization of participants in clinical trials.^1–5 I designed the Chocolate Happiness Undergoing More Pleasantness (CHUMP) study to compare the effects of dark chocolate, milk chocolate and “normal” chocolate consumption on happiness. The CHUMP study was a double-blinded clinical trial, and it demonstrated the difficulties associated with performing a truly blinded clinical trial.     

Relationship between Oscillatory Neuronal Activity during Reward Processing and Trait Impulsivity and Sensation Seeking

Background

The processing of reward and punishment stimuli in humans appears to involve brain oscillatory activity of several frequencies, probably each with a distinct function. The exact nature of associations of these electrophysiological measures with impulsive or risk-seeking personality traits is not completely clear. Thus, the aim of the present study was to investigate event-related oscillatory activity during reward processing across a wide spectrum of frequencies, and its associations with impulsivity and sensation seeking in healthy subjects.

Methods

During recording of a 32-channel EEG 22 healthy volunteers were characterized with the Barratt Impulsiveness and the Sensation Seeking Scale and performed a computerized two-choice gambling task comprising different feedback options with positive vs. negative valence (gain or loss) and high or low magnitude (5 vs. 25 points).

Results

We observed greater increases of amplitudes of the feedback-related negativity and of activity in the theta, alpha and low-beta frequency range following loss feedback and, in contrast, greater increase of activity in the high-beta frequency range following gain feedback. Significant magnitude effects were observed for theta and delta oscillations, indicating greater amplitudes upon feedback concerning large stakes. The theta amplitude changes during loss were negatively correlated with motor impulsivity scores, whereas alpha and low-beta increase upon loss and high-beta increase upon gain were positively correlated with various dimensions of sensation seeking.

Conclusions

The findings suggest that the processing of feedback information involves several distinct processes, which are subserved by oscillations of different frequencies and are associated with different personality traits.     

Miliary Tuberculosis in Adults

Of 40 adults with miliary tuberculosis 24 had “overt” disease; in them miliary mottling was usually present on the chest radiograph, and tubercle bacilli were readily isolated from sputum, urine, or cerebrospinal fluid. In the remaining 16 patients the disease was termed “cryptic” because its usual clinical and radiographic features were absent. This cryptic type is as common as the overt type in patients over 60 years. In this series the peak age incidence was in the eighth decade, and possibly this increase in the incidence age is due to the breakdown of old tuberculous foci in patients with diminished immunological mechanisms.Cryptic miliary tuberculosis is a difficult diagnostic problem and should be suspected in any elderly patient, particularly a woman, who has an unexplained pyrexia, pancytopenia, or leukaemoid reaction. In 10 cases it was diagnosed by a therapeutic trial with para-aminosalicylic acid and isoniazid, a fall of temperature to normal (usually within a week), weight gain, a rise in haemoglobin, and increased well-being being the criteria of improvement The use of such a trial is strongly advocated as a specific method of diagnosing cryptic miliary tuberculosis.     

A Higher Number of TMS-Elicited MEP from a Combined Hotspot Improves Intra- and Inter-Session Reliability of the Upper Limb Muscles in Healthy Individuals

We aimed to determine, using transcranial magnetic stimulation (TMS), the number of elicited motor evoked potentials (MEPs) that induces the highest intra- and inter-sessions reliability for the extensor carpi radialis (ECR) and first dorsal interosseus (FDI) muscles. Twelve healthy subjects participated in this study on two separate days. Single pulse magnetic stimuli were triggered with Magstim 200² to obtain MEPs from the muscles of interest, with the subjects in a relaxed position. Reliability of MEP responses was investigated in three blocks of 5, 10 and 15 trials. The intra- and inter-session reliability of the MEPs'' amplitudes and latencies were assessed using intraclass correlation coefficients (ICCs). Repeated measures ANOVA and paired t-tests revealed no significant time effect in the MEP amplitude and latency measurements (P>0.05). The ICCs indicated high intra-session reliability in the MEPs'' amplitudes for the ECR and FDI muscles (0.77 to 0.99, 0.90 to 0.99, respectively) and latency (0.80 to 1.00, 0.75 to 0.97, respectively). The MEPs'' amplitudes also had high inter-session reliability (0.84 to 0.97, 0.88 to 0.93, respectively), as did their latency (0.80 to 0.90, 0.75 to 0.97, respectively). Highest intra- and inter-session reliability was achieved for blocks of 10 and 15 trials. Our data suggest that intra- and inter-session comparisons should be performed using at least 10 MEPs in “combined hotspot” stimulation technique to ensure highest reliability.     

Estimating the dimensionality of the manifold underlying multi-electrode neural recordings

It is generally accepted that the number of neurons in a given brain area far exceeds the number of neurons needed to carry any specific function controlled by that area. For example, motor areas of the human brain contain tens of millions of neurons that control the activation of tens or at most hundreds of muscles. This massive redundancy implies the covariation of many neurons, which constrains the population activity to a low-dimensional manifold within the space of all possible patterns of neural activity. To gain a conceptual understanding of the complexity of the neural activity within a manifold, it is useful to estimate its dimensionality, which quantifies the number of degrees of freedom required to describe the observed population activity without significant information loss. While there are many algorithms for dimensionality estimation, we do not know which are well suited for analyzing neural activity. The objective of this study was to evaluate the efficacy of several representative algorithms for estimating the dimensionality of linearly and nonlinearly embedded data. We generated synthetic neural recordings with known intrinsic dimensionality and used them to test the algorithms’ accuracy and robustness. We emulated some of the important challenges associated with experimental data by adding noise, altering the nature of the embedding of the low-dimensional manifold within the high-dimensional recordings, varying the dimensionality of the manifold, and limiting the amount of available data. We demonstrated that linear algorithms overestimate the dimensionality of nonlinear, noise-free data. In cases of high noise, most algorithms overestimated the dimensionality. We thus developed a denoising algorithm based on deep learning, the “Joint Autoencoder”, which significantly improved subsequent dimensionality estimation. Critically, we found that all algorithms failed when the intrinsic dimensionality was high (above 20) or when the amount of data used for estimation was low. Based on the challenges we observed, we formulated a pipeline for estimating the dimensionality of experimental neural data.     

Synchronized drumming enhances activity in the caudate and facilitates prosocial commitment--if the rhythm comes easily

Why does chanting, drumming or dancing together make people feel united? Here we investigate the neural mechanisms underlying interpersonal synchrony and its subsequent effects on prosocial behavior among synchronized individuals. We hypothesized that areas of the brain associated with the processing of reward would be active when individuals experience synchrony during drumming, and that these reward signals would increase prosocial behavior toward this synchronous drum partner. 18 female non-musicians were scanned with functional magnetic resonance imaging while they drummed a rhythm, in alternating blocks, with two different experimenters: one drumming in-synchrony and the other out-of-synchrony relative to the participant. In the last scanning part, which served as the experimental manipulation for the following prosocial behavioral test, one of the experimenters drummed with one half of the participants in-synchrony and with the other out-of-synchrony. After scanning, this experimenter “accidentally” dropped eight pencils, and the number of pencils collected by the participants was used as a measure of prosocial commitment. Results revealed that participants who mastered the novel rhythm easily before scanning showed increased activity in the caudate during synchronous drumming. The same area also responded to monetary reward in a localizer task with the same participants. The activity in the caudate during experiencing synchronous drumming also predicted the number of pencils the participants later collected to help the synchronous experimenter of the manipulation run. In addition, participants collected more pencils to help the experimenter when she had drummed in-synchrony than out-of-synchrony during the manipulation run. By showing an overlap in activated areas during synchronized drumming and monetary reward, our findings suggest that interpersonal synchrony is related to the brain''s reward system.     

Palatable Meal Anticipation in Mice

The ability to sense time and anticipate events is a critical skill in nature. Most efforts to understand the neural and molecular mechanisms of anticipatory behavior in rodents rely on daily restricted food access, which induces a robust increase of locomotor activity in anticipation of daily meal time. Interestingly, rats also show increased activity in anticipation of a daily palatable meal even when they have an ample food supply, suggesting a role for brain reward systems in anticipatory behavior, and providing an alternate model by which to study the neurobiology of anticipation in species, such as mice, that are less well adapted to “stuff and starve” feeding schedules. To extend this model to mice, and exploit molecular genetic resources available for that species, we tested the ability of wild-type mice to anticipate a daily palatable meal. We observed that mice with free access to regular chow and limited access to highly palatable snacks of chocolate or “Fruit Crunchies” avidly consumed the snack but did not show anticipatory locomotor activity as measured by running wheels or video-based behavioral analysis. However, male mice receiving a snack of high fat chow did show increased food bin entry prior to access time and a modest increase in activity in the two hours preceding the scheduled meal. Interestingly, female mice did not show anticipation of a daily high fat meal but did show increased activity at scheduled mealtime when that meal was withdrawn. These results indicate that anticipation of a scheduled food reward in mice is behavior, diet, and gender specific.     

Altered Brain Activity during Reward Anticipation in Pathological Gambling and Obsessive-Compulsive Disorder

Background

Pathological gambling (PG) and obsessive-compulsive disorder (OCD) are conceptualized as a behavioral addiction, with a dependency on repetitive gambling behavior and rewarding effects following compulsive behavior, respectively. However, no neuroimaging studies to date have examined reward circuitry during the anticipation phase of reward in PG compared with in OCD while considering repetitive gambling and compulsion as addictive behaviors.

Methods/Principal Findings

To elucidate the neural activities specific to the anticipation phase of reward, we performed event-related functional magnetic resonance imaging (fMRI) in young adults with PG and compared them with those in patients with OCD and healthy controls. Fifteen male patients with PG, 13 patients with OCD, and 15 healthy controls, group-matched for age, gender, and IQ, participated in a monetary incentive delay task during fMRI scanning. Neural activation in the ventromedial caudate nucleus during anticipation of both gain and loss decreased in patients with PG compared with that in patients with OCD and healthy controls. Additionally, reduced activation in the anterior insula during anticipation of loss was observed in patients with PG compared with that in patients with OCD which was intermediate between that in OCD and healthy controls (healthy controls < PG < OCD), and a significant positive correlation between activity in the anterior insula and South Oaks Gambling Screen score was found in patients with PG.

Conclusions

Decreased neural activity in the ventromedial caudate nucleus during anticipation may be a specific neurobiological feature for the pathophysiology of PG, distinguishing it from OCD and healthy controls. Correlation of anterior insular activity during loss anticipation with PG symptoms suggests that patients with PG fit the features of OCD associated with harm avoidance as PG symptoms deteriorate. Our findings have identified functional disparities and similarities between patients with PG and OCD related to the neural responses associated with reward anticipation.     

Attenuated Neural Processing of Risk in Young Adults at Risk for Stimulant Dependence

Objective

Approximately 10% of young adults report non-medical use of stimulants (cocaine, amphetamine, methylphenidate), which puts them at risk for the development of dependence. This fMRI study investigates whether subjects at early stages of stimulant use show altered decision making processing.

Methods

158 occasional stimulants users (OSU) and 50 comparison subjects (CS) performed a “risky gains” decision making task during which they could select safe options (cash in 20 cents) or gamble them for double or nothing in two consecutive gambles (win or lose 40 or 80 cents, “risky decisions”). The primary analysis focused on risky versus safe decisions. Three secondary analyses were conducted: First, a robust regression examined the effect of lifetime exposure to stimulants and marijuana; second, subgroups of OSU with >1000 (n = 42), or <50 lifetime marijuana uses (n = 32), were compared to CS with <50 lifetime uses (n = 46) to examine potential marijuana effects; third, brain activation associated with behavioral adjustment following monetary losses was probed.

Results

There were no behavioral differences between groups. OSU showed attenuated activation across risky and safe decisions in prefrontal cortex, insula, and dorsal striatum, exhibited lower anterior cingulate cortex (ACC) and dorsal striatum activation for risky decisions and greater inferior frontal gyrus activation for safe decisions. Those OSU with relatively more stimulant use showed greater dorsal ACC and posterior insula attenuation. In comparison, greater lifetime marijuana use was associated with less neural differentiation between risky and safe decisions. OSU who chose more safe responses after losses exhibited similarities with CS relative to those preferring risky options.

Discussion

Individuals at risk for the development of stimulant use disorders presented less differentiated neural processing of risky and safe options. Specifically, OSU show attenuated brain response in regions critical for performance monitoring, reward processing and interoceptive awareness. Marijuana had additive effects by diminishing neural risk differentiation.