Work

Make it worth it: Effort-reward modulations on reinforcement-learning and prediction-error signaling across adolescence

Year: 2025

Type: article

Abstract: Adolescence is characterized by significant shifts in effort allocation. A well-known neuro-economic framework suggests that rewards help overcome potential effort costs. However, few studies have examined the neurobiological mechanisms by which rewards and associated effort costs drive adolescent learning. This study utilized functional magnetic resonance imaging in a sample of adolescents ( N = 146, 13–25 years) and employed a reinforcement-learning paradigm that manipulated effort and reward levels, by varying task demands and varying potential rewards. The analysis of trial-by-trial learning signals (reward prediction errors) and behavioral learning performance demonstrated that greater reward levels enhanced adolescent learning, especially when faced with greater effort demands. Moreover, this effect was more pronounced in those experiencing greater effort demands: younger adolescents and adolescents who place less value on effort for demanding tasks. Neuroimaging results revealed that the dorsal anterior cingulate cortex (dACC) was a key region in signaling the interaction between reward and effort demands. That is, greater reward strengthened prediction error coding in the dACC, particularly under conditions of greater task demands, with these effects being more pronounced in younger adolescents and adolescents who place less value on effort for demanding tasks. These findings support a role for dACC in the engagement of cognitive control, especially in situations where more cognitive control would be beneficial despite its associated effort costs, such as in high-demanding learning situations. This comprehensive approach aims to inform strategies for supporting effort allocation in learning during this crucial developmental period. • We examined the neurobiological mechanisms underlying how rewards and effort costs influence learning across adolescence. • We observed that greater rewards enhanced learning performance across both low and high effort learning contexts. • Younger adolescents and those valuing effort less learned more from high rewards, but only in high-effort learning contexts. • Neuroimaging showed stronger prediction error coding for high vs. low rewards, but only in high-effort learning contexts. • Dorsal ACC signals the interaction between effort and reward, and its activation shows age-related differences across adolescence.

Source: Developmental Cognitive Neuroscience

DOI: 10.1016/j.dcn.2025.101559

Authors: Anne-Wil Kramer, Lydia Krabbendam, Jessica V. Schaaf, Hilde M. Huizenga, Anna C.K. Van Duijvenvoorde