Neurochemical properties of depression and anxiety and its effect on reinforcement learning

Abstract

Despite having more direct information than functional magnetic resonance imaging, little is known about how the brain’s neurochemical mechanisms influence mental health and decision-making. This study, with 1H-functional MRS scanning during reward learning and computational modeling, examined whether baseline glutamate/glutamine (Glx) and lactate concentrations were related to general anxiety and depression and how they changed according to gains and losses. The glutamate concentration in the anterior insular cortex (AIC), not the medial prefrontal cortex, was positively associated with the general psychopathological factor of anxiety and depression through the mediation of error sensitivity. The AIC Glx decreased mainly after learning from gains and as the entire task was performed. Conversely, lactate reduction in AIC occurred only during learning from loss and was associated with higher general psychopathology scores. It is the first time to demonstrated that both glutamate-related error sensitivity and loss-specific lactate reduction in AIC are associated with general psychopathology. This study’s results suggest that AIC is a potential target brain region for glutamate and/or lactate-mediated therapeutics for anxiety and depression.