Anxious individuals are more sensitive to changes in outcome variability and value differences in dynamic environments

Anxiety is known to alter learning in uncertain environments. Standard experimental paradigms and computational models addressing these differences have mainly assessed the impact of volatility, and anxious individuals have been shown to have a reduced learning rate when moving from a stable to volatile environment. Previous research has not, however, independently assessed the impact of both changes in volatility, i.e., reversals in reward contingency, and changes in outcome variability (noise) in the same individuals. Here we use a simple probabilistic reversal learning paradigm to independently manipulate the level of volatility and noise at the experimental level in a fully orthogonal design. We replicate general increases, irrespective of anxiety levels, in both positive and negative learning rates when moving from low to high volatility, but only in the context of low noise. When low volatility is combined with high noise, more anxious individuals display negative learning rates similar to high volatility with high noise, whereas those lower in anxiety show the usual negative learning rate increase from low to high volatility. Within-individual increases in lose-shift responses from low to high noise conditions scale with levels of anxious traits, but this occurs under low volatility only. We furthermore find that people with higher anxious traits are more accurate overall and utilize a more exploitative decision-making strategy in this dynamic environment. Our findings suggest that changes in both sources of uncertainty, volatility and noise, should be carefully considered when assessing learning, particularly in relation to anxiety and other neuropsychiatric conditions, and implicate anxiety-related differences in dopaminergic and noradrenergic neurotransmitter signalling when learning in highly changeable environments.