Summary of Journal Article
Mason, L., O’sullivan, N., Montaldi, D., Bentall, R. P., & El-Deredy, W. (2014). Decision-making and trait impulsivity in bipolar disorder are associated with reduced prefrontal regulation of striatal reward valuation. Brain: A Journal of Neurology, 137, 2346-2355. doi:10.1093/brain/awu152
Summary of Introduction
Past research indicates that risky behavior and impulsivity are hallmark traits of bipolar disorders and many other, often comorbid, disorders. Impulsivity in this context is defined by the tendency to make decisions based on hedonic, pleasure-driven behaviors even when this behavior contradicts long-term optimal outcome and the lack of the self-control associated with delayed-gratification. Past research has shown specific brain regions to be associated with impulsive vs. optimal decision making. Impulsivity can be associated with suicidality, substance use disorders, and addiction (whether it be drugs, food, or gambling). Some of the brain regions indicated include but are not limited to: (1) the ventral striatum, (2) the dorsolateral prefrontal cortex and (3) the ventromedial prefrontal cortex. (1) The striate nucleus, striatum, or corpus striatum is what we are referencing when we speak of the ventral striatum. The striate nucleus is often subdivided between the ventral and dorsal regions. The ventral striatum contains the nucleus accumbens and olfactory tubercle. The nucleus accumbens connects to the ventral tegmental area via the mesolimbic pathway; addictive and/or appetitive behaviors are associated with this circuitry, specifically the ventral tegmental region and the nucleus accumbens. The ventral tegmental area, located in the midbrain, projects to the ventral striatum of the basal ganglia in the forebrain. Dopamine activity is primarily responsible for regulation of these appetitive behaviors. (2) The orientation of the dorsolateral prefrontal cortex is somewhat self-explanatory; referring to the dorsal (toward the back, like a dorsal fin), lateral (away from the midline), prefrontal cortex (the area most associated with executive function, integrative reasoning, and human personality). The dorsolateral PFC in particular, is associated with responsible, sound, higher-order, decision-making in healthy individuals. (3) The ventromedial prefrontal cortex is also somewhat self-explanatory and refers to orientation of the area of the prefrontal cortex we are referring to. This area is the part of the PFC that is ventral (toward the belly button) and medial (toward the midline). The ventromedial PFC is indicated in providing a weighted valuation signal based on integrated signals/projections from the ventral striatum and the dorsolateral PFC. Essentially, the dorsolateral PFC should prefer responsible decisions, the ventral striatum wants to have fun and reap the short-term rewards and the ventromedial PFC integrates these signals and helps us make a decision by providing a weighted valuation signal. It is worth mentioning that there are many functions related to these regions but for purposes of this summary I provided simplistic definitions. According to a study cited in this research, the dorsolateral PFC is also activated during re-evaluation of the possibility of positive reward-based outcome. There is a body of research related to the neural basis of reward evaluation, but often results are contradictory or unclear. A 2012 study on bipolar individuals in remission indicated hyperactivity of reward-circuitry and ventromedial PFC during anticipation, yet no differentiation based on outcome. In abnormally, or clinically, impulsive individuals reward and reward size are shown to trump loss in subjective decision making which suggests a failure to deactivate reward-sensitive structures in the face of loss. One study that was mentioned, showed continued activation of the nucleus accumbens or ventral striatal region even when rewards were omitted, further identifying failure to deactivate reward-sensitive structures. The “healthy” control group exhibited the opposite, showing an appreciation for magnitude of loss and probability in relation to higher-order goals, independent of magnitude potential gains. This conglomeration of research shows clear difficulty in regulating goal-driven, higher-order optimized behavior but non-explicit examination of the neural circuitry at play.
The aim behind this study was to find the neural basis for impulsivity in bipolar disorder. Impulsivity is a key trait in the disorder and many other disorders, and I would speculate that impulsive behavior—and maybe even the desire to behave impulsively—is highly connected to diminished quality of life. The second goal of the study was to shed light on the ambiguity related to information on affective state. Individuals with bipolar disorder are known for having fluctuating affective states. Past research has shown that impulsive behavior is heightened during the manic phase of the disorder, which is marked by heightened energy and restlessness. Participants in the study are asked to play a roulette type game in order to test the theory that the ventromedial PFC is indicated in valuating reward-based options, based on signals it receives from the ventral striatum and dorsolateral PFC. The nucleus accumbens codes probability and delay unsystematically, preferring risky and well-paying rewards—especially if the payout is immediate. A reward that is objectively superior with a long delay will be less preferable if we only consider the dopaminergic mesolimbic pathway without input from the dorsolateral prefrontal cortex. According to the study, past research indicated the dorsolateral PFC is not just limited to mediating risky vs. safe behavior, it is also indicated when suppressing a desire based on higher-order thought processes telling us the end result will be unfavorable —like when we use self-control to choose an apple over a candy bar. The prefrontal region is associated with choosing delayed gratification if that option is objectively superior. Further, the current research suggests the hallmark trait impulsivity and risky behavior of bipolar disorder is associated with a subjective valuation and tendency to bias decision toward lower-order preference without consideration of optimal long-term decision—or failure to integrate input for optimal outcome. Understanding the neural basis behind this trait feature can potentially aid in the process of evaluating interventions and therapies associated with the exhibited hedonically driven behavior.
Summary of Methods
As mentioned above, a gambling task was used during fMRI in order to evaluate neural activity during selection, anticipation, and outcome. Participants included 40 individuals between the ages of 18 and 45. The control group consisted of 20 “healthy”, case-matched individuals. Age, gender and level of education were factors considered in avoidance of confounding variables. In the bipolar group, individuals were euthymic (the state of remission between manic and depressive stages), had no current substance abuse issues—alcohol problems in particular were excluded. The bipolar individuals were not in receipt of antipsychotic medication and had not taken antipsychotic medication for at least 6 months prior. A number of tests were given using the appropriate and trusted scales, in order to evaluate participants; some of which included structured clinical interview based on DSM-IV, 17-item Hamilton Depression Rating Scale, and the 12-item Bech-Ralfaelsen Mania Scale. Selection, anticipation, and outcome stages were timed, and participants were informed that if they did not make a decision in the allotted amount of time a random decision would be made for them. Four choices were available, all choices accompanied by equal knowledge of reward, and probability (risk). Manipulated variables included level of risk and magnitude of earnings; these decisions were separated by trial instead of offering opposing choices. A risky, low-probability, option was defined as 25% chance for reward and a safe bet was 75% chance at reward. Participants were informed that they could keep the profits accrued after completion; However, magnitude of reward was offset by magnitude of loss. In other words, the size of the bet was fixed so participants would lose the amount of the bet if they lost, giving them incentive to make wise decisions. There were 8 runs, which were 6 minutes per run and a total of 272 trials.
Summary of Results
The study does not suggest that the dorsolateral prefrontal cortex does not activate in bipolar individuals, it simply suggests that it shows abnormality in activation pattern, with a negative correlation to sound decision-making practices. The controls prejudicially activate dorsolateral PFC for high-probability gamble results, while the bipolar disorder group show the opposite pattern, using this region more for the low-probability rewards. Increased risk-taking in the bipolar disorder group can related to diminished response to safe, responsible, reward outcomes in the right dorsolateral PFC. Intensified pleasure in response to reward is indicated in the bipolar disorder group, they show greater enticement for reward relative to loss than does the control group.
During the aforementioned tasks, fMRI was used to examine correlations and activity among ventral striatal, ventromedial prefrontal cortical, and dorsolateral prefrontal cortical regions in euthymic bipolar individuals. It was expected that preferential activation of dorsolateral PFC would correlate with predilection for maximization of overall winnings and coordination between ventral striatal region and ventromedial PFC would be associated with risky decision-making tendencies. In healthy participants, it was posited that optimal integration of projections to the ventromedial PFC would correlate with dorsolateral PFC activation, more than it would correlate with activity in the ventral striatum. Furthermore, dorsolateral PFC activity can be associated with response to sound decision-making and negatively correlated to hyper-hedonic reward-driven behavior which implies that perhaps trait impulsivity (not dependent on particular disorder) is a result of highly diminished dorsolateral PFC-moderated upregulation and subsequent sensitivity to signals that affect goal-driven behavior. The idea that impulsivity is heightened during mania and hypomanic phases of the disorder is relatively well known, though not as much from a neural vantage point as a behavioral vantage point. This study shows that even in remission from symptoms, individuals with bipolar disorder exhibit abnormalities in mechanisms meant to regulate hyper-hedonic tendencies. This highlights the idea that bipolar individuals may be more likely to chase rewards because they take more pleasure in them, yet they are less affected by losses relative to controls further confounding the problem. In other words, the bipolar group does not get excited about high probability/low yield gambles, but they focus their attention and drive on the low probability/high yield gambles and the dorsolateral prefrontal area fails to devalue the projections that cause the heavily weighted pleasure-seeking trait. The study’s findings both agreed with and elaborated on past research. As previously mentioned, past research indicates the ventral striatum, ventromedial PFC and the dorsolateral PFC contribute to decision-making; This study confirmed this and examined related neural structures. Delineation based on affective state is ambiguous in past research, which is a key motivating factor behind this study. There are many valuable insights gained in examining these conditions, such as evaluation of current therapeutic interventions and future efforts to cater interventions to the core trait of the disorder, focusing on impulse-control, recognizing the far reach of the problem and subsequent diminished quality of life.