How the brain computes value when deciding what to eat.
What the evidence says
- The brain’s decision circuitry computes subjective value of food options at moment of choice[1]
- Value = function of expected reward, effort cost, and time delay[1]
- Three competing controllers govern food decisions: Pavlovian (pre-programmed responses to food cues), habitual (learned stimulus-action associations), and goal-directed (flexible, forward-looking value computation)[1]
- The habitual controller is “model-free” — it averages previous reward history and cannot learn delayed health consequences[1]
- The decision system is biased toward immediate, high-reward, low-effort options[1]
- Healthy eaters’ vmPFC signals reflect both health and taste attributes; unhealthy eaters’ vmPFC reflects only taste — dlPFC modulates vmPFC to incorporate health considerations[1]
- Successful dieting requires five simultaneous processes: Pavlovian/habitual inhibition, correct goal-directed values, conflict detection, response inhibition, and sufficient attention to health attributes[1]
- Feeding uniquely combines all six properties no other decision domain does: necessary/frequent consumption, multiple-controller regulation, dedicated homeostatic mechanisms, dedicated Pavlovian controllers, constrained learning systems, and powerful hedonic circuitry activation[1]
- This explains why convenient junk food wins: the neural economics favor it[1]