Neural Signatures of Impulsivity in Adolescence: A Multivariate Pattern Analysis of Reward and Loss Processing
Abstract
Adolescents experience mood disturbances and impulsivity that increase vulnerability to maladaptive risk-taking, yet the underlying neurobiological mechanisms of these phenotypes remain poorly understood. Thus, the current pre-registered study sought to optimize models predicting anhedonic, sad, and manic mood, as well as impulsivity, from multivariate frontostriatal neural responses during incentive (i.e., reward and loss) processing. Participants were adolescents (n=153, healthy 12–18-year-olds, 104 assigned female at birth) with moderate to severe depression, as well as some controls, to have a full range of symptoms. They underwent functional neuroimaging during a monetary incentive delay task and completed self-report questionnaires on mood and impulsivity. We optimized a pre-registered 3-fold nested cross-validation support vector regression model predicting impulsivity, which showed statistically significant but small predictive performance (R² = .02, MAE = .81). Feature importance analyses showed distributed signal across frontostriatal circuits, particularly cognitive and motor-related circuits. In addition, neural activity to loss anticipation followed by reward receipt and reward anticipation contributed most to model performance. As a sensitivity analysis, we repeated the model with depression-residualized impulsivity to account for shared variance with depressive symptoms. This model showed modestly improved predictive performance (R² = .04, MAE = .80) with a similar pattern of feature contributions. No models successfully predicted anhedonic, sad, or manic mood. These findings suggest that multivariate neural activity within cognitive- and motor-related frontostriatal circuits during incentive processing is associated with impulsivity, and not mood disturbances, in adolescents, including those with mild to moderate depression. Public health efforts aimed at reducing impulsivity (and potentially risk-taking) should consider the role of incentive processing.
Related
- MINT Block: multivariate integration across independent studies to identify multi-omics signatures
- Depth of Processing and Congruency Interact to Shape Electrophysiological Signatures of Recognition Memory
- Individual behavioral signatures shape processing of context-dependent social fear
- Increasing cerebellar beta rhythms shifts cerebral oscillatory signatures of context-driven language processing
- Reliability-based arbitration reveals heterogeneous model-based and model-free reinforcement learning strategies linked to compulsivity and impulsivity
BibTeX
@misc{goncalves2026neural,
title = {{Neural Signatures of Impulsivity in Adolescence: A Multivariate Pattern Analysis of Reward and Loss Processing}},
author = {Stefanie Fraga Gonçalves and Morgan Lindenmuth and Helmet Karim and Rasim Diler and Cecile D. Ladouceur},
howpublished = {PsyArXiv},
year = {2026},
month = oct,
doi = {10.31234/osf.io/7numx_v1},
url = {https://osf.io/preprints/psyarxiv/7numx_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Neural Signatures of Impulsivity in Adolescence: A Multivariate Pattern Analysis of Reward and Loss Processing
AU - Stefanie Fraga Gonçalves
AU - Morgan Lindenmuth
AU - Helmet Karim
AU - Rasim Diler
AU - Cecile D. Ladouceur
PY - 2026
DA - 2026/10/01
PB - PsyArXiv
DO - 10.31234/osf.io/7numx_v1
UR - https://osf.io/preprints/psyarxiv/7numx_v1/
AB - Adolescents experience mood disturbances and impulsivity that increase vulnerability to maladaptive risk-taking, yet the underlying neurobiological mechanisms of these phenotypes remain poorly understood. Thus, the current pre-registered study sought to optimize models predicting anhedonic, sad, and manic mood, as well as impulsivity, from multivariate frontostriatal neural responses during incentive (i.e., reward and loss) processing. Participants were adolescents (n=153, healthy 12–18-year-olds, 104 assigned female at birth) with moderate to severe depression, as well as some controls, to have a full range of symptoms. They underwent functional neuroimaging during a monetary incentive delay task and completed self-report questionnaires on mood and impulsivity. We optimized a pre-registered 3-fold nested cross-validation support vector regression model predicting impulsivity, which showed statistically significant but small predictive performance (R² = .02, MAE = .81). Feature importance analyses showed distributed signal across frontostriatal circuits, particularly cognitive and motor-related circuits. In addition, neural activity to loss anticipation followed by reward receipt and reward anticipation contributed most to model performance. As a sensitivity analysis, we repeated the model with depression-residualized impulsivity to account for shared variance with depressive symptoms. This model showed modestly improved predictive performance (R² = .04, MAE = .80) with a similar pattern of feature contributions. No models successfully predicted anhedonic, sad, or manic mood. These findings suggest that multivariate neural activity within cognitive- and motor-related frontostriatal circuits during incentive processing is associated with impulsivity, and not mood disturbances, in adolescents, including those with mild to moderate depression. Public health efforts aimed at reducing impulsivity (and potentially risk-taking) should consider the role of incentive processing.
ER -