Annual Research Review: Transdiagnostic neuroscience of child and adolescent mental disorders--differentiating decision making in attention-deficit/hyperactivity disorder, conduct disorder, depression, and anxiety

Abstract

Background: Ineffective decision making is a major source of everyday functional impairment and reduced quality of life for young people with mental disorders. However, very little is known about what distinguishes decision making by individuals with different disorders or the neuropsychological processes or brain systems underlying these. This is the focus of the current review.

Scope and methodology: We first propose a neuroeconomic model of the decision-making process with separate stages for the prechoice evaluation of expected utility of future options; choice execution and postchoice management; the appraisal of outcome against expectation; and the updating of value estimates to guide future decisions. According to the proposed model, decision making is mediated by neuropsychological processes operating within three domains: (a) self-referential processes involved in autobiographical reflection on past, and prospection about future, experiences; (b) executive functions, such as working memory, inhibition, and planning, that regulate the implementation of decisions; and (c) processes involved in value estimation and outcome appraisal and learning. These processes are underpinned by the interplay of multiple brain networks, especially medial and lateralized cortical components of the default mode network, dorsal corticostriatal circuits underpinning higher order cognitive and behavioral control, and ventral frontostriatal circuits, connecting to brain regions implicated in emotion processing, that control valuation and learning processes.

Findings and conclusion: Based on clinical insights and considering each of the decision-making stages in turn, we outline disorder-specific hypotheses about impaired decision making in four childhood disorders: attention-deficit/hyperactivity disorder (ADHD), conduct disorder (CD), depression, and anxiety. We hypothesize that decision making in ADHD is deficient (i.e. inefficient, insufficiently reflective, and inconsistent) and impulsive (biased toward immediate over delayed alternatives). In CD, it is reckless and insensitive to negative consequences. In depression, it is disengaged, perseverative, and pessimistic, while in anxiety, it is hesitant, risk-averse, and self-deprecating. A survey of current empirical indications related to these disorder-specific hypotheses highlights the limited and fragmentary nature of the evidence base and illustrates the need for a major research initiative in decision making in childhood disorders. The final section highlights a number of important additional general themes that need to be considered in future research.

Keywords: Transdiagnostic; amygdala: CD; anxiety; attention-deficit/hyperactivity disorder; decision making; delayed reinforcement; depression; executive functions; inhibition; prefrontal cortex; reinforcement learning; ventral striatum; working memory.

Figure 1

A schematic representation of an integrated neuroeconomic model highlighting the complex interplay between multiple psychological systems and neural circuits in the control of the decision‐making process. The decision‐making process itself is divided into three distinct stages: Evaluation – where an estimate of the subjective utility of available choice options is made taking into account memory and learning from prior experience as well as prospection about future value mediated by implicit reinforcement learning mechanisms (encoded in a utility matrix) and explicit self‐referential processes (autobiographical memory); decision and management – during which the subjective utility assigned to competing alternatives is compared in terms of overall costs and benefits and a decision plan is implemented – processes heavily influenced by higher order executive functions; appraisal and accommodation – utility actually derived from decision is estimated (again on the basis of explicit and implicit value systems) and compared with prior expectations to generate a prediction error signal which drives learning and updates implicit and explicit value estimates for particular experiences and choices as represented by the feedback loops in the figure. The model proposes that these decision‐making stages are primarily controlled by three distinct brain systems: the default mode network (MPFC, medial prefrontal cortex; PCC, posterior cingulate cortex; LPC, lateral parietal cortex; MTG, medial temporal gyrus) primarily implicated in self‐referential cognition but also in some aspects of self‐regulation; executive control system (DLPFC, dorsolateral prefrontal cortex; ACC, anterior cingulate cortex; PAR, parietal cortex) which mediates top‐down control during self‐regulation and planning; and reinforcement evaluation and learning circuits (OFC, orbitofrontal cortex; AMYG, amygdala; ACC, anterior cingulate cortex, REINF, reinforcement)

Figure 2

Disorder‐specific hypotheses mapped onto the decision‐making stages outlined in our neuroeconomic model. Key: ADHD: attention‐deficit/hyperactivity disorder; DM: default mode network; PFC: prefrontal cortex; VS: ventral striatum [Correction added on 7 January 2016, after first online publication: The previous incorrect table was published without content and it has now been replaced with the corrected version.]