Schizophrenia Research Forum - A Catalyst for Creative Thinking

News Brief: Methamphetamine Points to Reward-Learning Deficits Underlying Psychosis

July 10, 2013. Disruptions in how the brain assigns value to a stimulus or memory may underlie features of psychosis, according to a study published online June 4 in the American Journal of Psychiatry. Led by Graham Murray of the University of Cambridge, U.K., the study examined the effects of a low dose of methamphetamine, which boosts dopamine as well as norepinephrine and serotonin, on reward-learning tasks in 18 healthy controls. Using functional magnetic resonance imaging (fMRI), the researchers found that the drug attenuated signals in the ventromedial prefrontal cortex (also called the orbitofrontal cortex) and that these reductions correlated with the degree of psychotic symptoms experienced.

The results support the idea that symptoms of schizophrenia can stem from reward learning, in which associations are made between a stimulus or action and an outcome. Faulty reward learning could lead people to misjudge the relevance of something or to mistakenly associate unconnected ideas or events—processes that could build delusional beliefs (Kapur et al., 2003) or the profound apathy characteristic of negative symptoms (see SRF related news story). Typically, reward learning involves connections between the ventral striatum, which encodes prediction error signals, and the ventromedial prefrontal cortex, which represents value. In the new study, the researchers used methamphetamine to tweak how the circuitry worked and to see if anything resembling psychosis resulted.

When subjects were under the influence of methamphetamine, first author Javier Bernacer and colleagues write, the learning paradigm produced attenuated signals in both the ventral striatum and the cortex relative to placebo, but only the cortical value signals varied according to degree of psychosis. The researchers also found that reduced signals in the posterior cingulate cortex, a region also implicated in encoding value, correlated with greater psychosis symptoms. Methamphetamine also slowed the rate of learning, though it did not make a difference in what was ultimately learned. Amisulpride, a selective dopamine D2 receptor blocker, did not alter the effects of methamphetamine. This suggests that other neuromodulatory pathways impinge on reward-learning circuitry and parsing these may offer a fuller picture of the mechanisms of psychosis.—Michele Solis.

Bernacer J, Corlett PR, Ramachandra P, McFarlane B, Turner DC, Clark L, Robbins TW, Fletcher PC, Murray GK. Methamphetamine-Induced Disruption of Frontostriatal Reward Learning Signals: Relation to Psychotic Symptoms. Am J Psychiatry. 2013 Jun 4. Abstract

Comments on News and Primary Papers

Primary Papers: Methamphetamine-Induced Disruption of Frontostriatal Reward Learning Signals: Relation to Psychotic Symptoms.

Comment by:  Lynn Selemon
Submitted 23 June 2013
Posted 23 June 2013

Bernacer et al. make an important contribution to the field by linking the basic neurobiologic mechanism underlying motivation and learning to psychosis in mental illness. It is particularly interesting that the potent D2 receptor antagonist amisulpride did not prevent the methamphetamine-induced disruption of incentive value signaling in the ventral prefrontal cortex. The authors discuss other possible pharmacologic sites, among them the D1 receptor, as the active site for methamphetamine in the cortex. I look forward to future studies by this group that elucidate the neurotransmitter(s) that may be involved.

View all comments by Lynn Selemon

Primary Papers: Methamphetamine-Induced Disruption of Frontostriatal Reward Learning Signals: Relation to Psychotic Symptoms.

Comment by:  J. Daniel Ragland
Submitted 25 June 2013
Posted 25 June 2013
  I recommend this paper

Phenomenologically, there can be striking similarities between acute psychotic symptoms in individuals with schizophrenia and individuals with acute methamphetamine abuse. There is also growing evidence that both conditions disrupt dopamine-rich frontostriatal brain networks, leading to similar changes in cognition, such as impaired performance on cognitive control tasks such as the Stroop (e.g., Salo et al., 2011). This current study by Bernacer and colleagues is one of the first to investigate the impact of methamphetamine on reward learning in healthy participants and uses fMRI, a methamphetamine challenge and a pharmacological intervention (the antipsychotic amisulpride) to get at putative monoaminergic mechanisms. fMRI results were convincing in linking reward prediction error to the ventral striatum and incentive value to the ventromedial prefrontal cortex. The disruptive effects of methamphetamine challenge on task performance and fMRI activation in candidate brain regions were also consistent with the notion that methamphetamine disrupts frontostriatal network function.

The links to psychosis were intriguing, but not universal. Increases in psychotic symptoms with methamphetamine challenge were associated with disruption in ventromedial prefrontal and posterior cingulate cortex, and related processing of incentive value. However, pretreatment with amisulpride did not rescue task performance or fMRI activation following drug challenge, and psychotic symptoms were not associated with reward prediction errors or related activation in the ventral striatum. The lack of an antipsychotic effect may not be surprising in retrospect, as antipsychotic efficacy depends upon sustained treatment. However, the lack of association with reward prediction errors and related brain function is more difficult to reconcile, as reward prediction errors are a central component of reward learning deficits in psychotic disorders such as schizophrenia (e.g., Morris et al., 2012). Nevertheless, this research design is an impressive way to obtain convergent data linking reinforcement learning with clinical phenomenology, and functional and pharmacological brain mechanisms. Future studies employing this approach with a patient group of first-episode psychosis patients before and after treatment would be most informative.


Salo R, Ravizza S, Fassbender C (2011). Overlapping cognitive patterns in schizophrenia and methamphetamine dependence. Cogn Behav Neurol, 24(4): 187-193. Abstract

Morris RW, Vercammen A, Lenroot R, Moore L, Langton JM, Short B, Kulkarni J, Curtis J, O'Donnell M, Weickert CS, Weickert TW (2012). Disambiguating ventral striatum fMRI-related BOLD signal during reward prediction in schizophrenia. Mol Psychiatry, 17(3):235, 280-289. Abstract

View all comments by J. Daniel Ragland

Comments on Related News

Related News: Deconstructing Negative Symptoms in Schizophrenia

Comment by:  Laurie Kimmel
Submitted 25 October 2012
Posted 26 October 2012

As a clinician, I find this research encouraging.

View all comments by Laurie Kimmel