Impaired adaptation of learning to contingency volatility in internalizing psychopathology

Abstract

Using a contingency volatility manipulation, we tested the hypothesis that difficulty adapting probabilistic decision-making to second-order uncertainty might reflect a core deficit that cuts across anxiety and depression and holds regardless of whether outcomes are aversive or involve reward gain or loss. We used bifactor modeling of internalizing symptoms to separate symptom variance common to both anxiety and depression from that unique to each. Across two experiments, we modeled performance on a probabilistic decision-making under volatility task using a hierarchical Bayesian framework. Elevated scores on the common internalizing factor, with high loadings across anxiety and depression items, were linked to impoverished adjustment of learning to volatility regardless of whether outcomes involved reward gain, electrical stimulation, or reward loss. In particular, high common factor scores were linked to dampened learning following better-than-expected outcomes in volatile environments. No such relationships were observed for anxiety- or depression-specific symptom factors.

A neural and behavioral tradeoff underlies exploratory decisions in normative anxiety

“…Combined with the present results, one possible hypothesis is that anxious individuals perceive the world as being more volatile and therefore needs to be explored more frequently. Previous studies reported that trait anxiety and internalizing psychopathology decreased the ability to adjust learning rates as a function of volatility (Browning et al 2015; Gagne et al 2020), and it has been suggested that anxious individuals misestimates uncertainty (Pulcu and Browning 2019). Indeed, one simple heuristic which can be used to maintain a low level of uncertainty in a decision-making environment is to explore frequently, independently of actual volatility levels.…”

“…This provides two benefits: first, observing how exploration varies across the distribution, rather than comparing two populations which are somewhat arbitrarily divided (patients vs. controls); and second, examining the normal range, allowing us to examine how maladaptive decisions are mediated by exploration even in healthy individuals. Because such maladaptive decision have a huge impact on daily-life in all individuals, and definitely on societies and industry, we argue that more studies should use such gradual scales over non-clinical populations (Browning et al 2015; Fung et al 2019a; Gagne et al 2020). It should be noted that Spielberger’s Trait-Anxiety Inventory has been debated for its lack of convergent and discriminant validity, suggesting that it estimates ‘negative affectivity’ rather than proneness to anxiety per-se (Balsamo et al 2013).…”

Anxiety and Depression Related Abnormalities in Socio-affective Learning

“…Seminal work by Browning and colleagues (2015) suggests that anxiety is associated with impaired learning in aversive environments, potentially explaining maladaptive behaviour as the consequence of abnormal learning. Learning abnormalities have also been shown in depression (Robinson & Chase, 2017), and recent work shows associations with general affective distress (Gagne et al, 2020), raising questions about the specificity of learning deficits to anxiety (Pulcu & Browning, 2019). Concerns about the overlap in symptom features of anxiety and depression (Regier et al, 2013;Watson, 2009), as well as a focus on threat learning and reward learning in anxiety and depression respectively (see Bishop & Gagne, 2018), raise questions as to whether findings such as those identified by Browning and colleagues (2015) may relate more generally to negative affect rather than anxiety specifically, as suggested by more recent work (Gagne et al, 2020).…”

“…Human behaviour is guided by the simultaneous goals of maximising reward while minimising punishment. However, individuals with depression and anxiety often exhibit behaviours at odds with their best interests (Browning et al, 2015;Gagne et al, 2020). Seminal work by Browning and colleagues (2015) suggests that anxiety is associated with impaired learning in aversive environments, potentially explaining maladaptive behaviour as the consequence of abnormal learning.…”

“…Recent findings have shown that a general, internalizing factor (rather than anxiety or depression specifically) was associated with poorer overall learning and decreased learning in better-than-expected outcomes under volatile environments (Gagne et al, 2020). However, Gagne and colleagues (2020), as is consistent with most work in the field, compared nonequivalent outcomes across both versions of their tasks (i.e., positive reinforcement -monetary gain vs. negative punishment -monetary loss from previously provided funds; positive punishment -electrical shock vs. positive reinforcement -monetary gain).…”