Heterogeneous appetite patterns in depression: computational modeling of nutritional interoception, reward...

Area of Science:

• Computational psychiatry
• Neuroscience
• Decision-making

Background:

• Accurate interoception is vital for homeostasis and health; disruptions link to depression.
• Nutrient homeostatic dysregulation is a focus for depression classification and treatment.
• Appetite changes in depression correlate with altered mesolimbic dopamine and insular cortex activity.

Purpose of the Study:

• To explore computational psychiatry's potential in understanding interoception's role in depression.
• To investigate how altered interoception affects homeostatic behavior and reward activity using a homeostatic reinforcement learning model.

Main Methods:

• Developed a homeostatic reinforcement learning model.
• Conducted simulation experiments to model altered interoception.
• Compared interoception manipulation effects with traditional reinforcement learning parameters (e.g., inverse temperature β, delay discount γ).

Main Results:

• Simulations showed altered interoception produced behaviors resembling depression subtypes, including appetite dysregulation.
• Reduced interoception decreased reward activity and increased punishment, mirroring decreased appetite in depression.
• Increased interoception led to heightened reward activity and impaired goal-directed behavior, reflecting increased appetite.

Conclusions:

• Disruptions in interoception and related reinforcement learning parameters contribute to depressive symptoms by affecting homeostatic regulation.
• Integrating interoception and homeostasis into decision-making models is crucial for depression subtype classification and treatment development.