Understanding Decision-Making Across the Lifespan Needs Theoretical Neuroscience

A team of neuroscientists led by Michael B. Ryan, Letizia Ye, and Anne K. Churchland published a provocative paper arguing that research on decision-making across the lifespan has failed to incorporate two decades of computational neuroscience advances. While theoretical frameworks have transformed how we study young, healthy brains—providing tools to model neural dynamics, population codes, and interareal communication—aging research remains stuck with single-metric behavioral readouts and descriptive neural analyses. The authors contend this represents a major missed opportunity, as aging offers a unique platform for testing theories of neural computation under changing biological constraints.

Specifically, the paper outlines how recent advances in behavioral quantification, latent state modeling, dynamical systems, encoding models, representational geometry, and recurrent neural networks (RNNs) offer a rich theoretical toolkit. By applying these computational approaches, researchers could move beyond cross-sectional comparisons toward mechanistic explanations of why individuals age differently. This integration could fundamentally shift aging research from descriptive accounts to predictive models that explain variations in cognitive trajectories, potentially leading to more personalized interventions and a deeper understanding of neural stability and flexibility throughout life.