Simulating mood disorders in systems of interacting agents: the role of intrinsic and extrinsic factors in mood regulation

The incidence of neuropsychiatric disorders, such as major depressive disorder, is on the rise and represents a significant burden on modern societies. While previous research has focused on their neuronal origins and the development of pharmacological therapies, there has been a recent trend towards viewing neuropsychiatric disorders through the lens of complex systems. In this context, top-down causation from higher levels of organization, such as large-scale brain networks or the self, is seen as a determining factor in both their symptomatology and the efficacy of their treatment, as evidenced by the widespread use of behavioral therapy. That said, we consider that this perspective should be taken further, incorporating the putative top-down effects of social interactions and dynamics in understanding the etiology of major depressive disorder.

With that in mind, we will study the behavior, collective dynamics, and internal states of simulated interacting agents in collaborative foraging tasks. Such agents will be equipped with internal systems to track their actions and associated rewards, as in classical reinforcement learning paradigms, but also a proxy system for tracking their mood, following empirically-validated mathematical models. Such internal states will, i turn, influence how agents act in the world and interact with their peers.

Our first goal is to evaluate how disruptions in the mood systems of individuals impact their internal dynamics and behavior, leading them to exhibit symptoms characteristic of MDD, such as chronically low mood and social withdrawal. Then, and most importantly, we aim to investigate how different social norms, shaping the level of collaboration between agents, affect the behavior of individuals with disruptions to their mood systems. Finally, we will investigate the multi-level interactions between these variables and the possibility of existing negative feedback loops and societal-level dynamics that might increase the propensity of agents to develop "mood disorders".

With this, we aim to demonstrate how the symptomatology and severity of MDD depend on the interaction between internal deficiencies in mood-controlling systems and societal norms affecting cooperation and interdependence between agents.