Food reward: brain substrates of wanting and liking.

What are the neural substrates of food reward? Are reward and pleasure identical? Can taste pleasure be assessed in animals? Is reward necessarily conscious? These questions have re-emerged in recent years, and there is now sufficient evidence to prompt re-examination of many preconceptions concerning reward and its relation to brain systems. This paper reviews evidence from many sources regarding both the psychological structure of food reward and the neural systems that mediate it. Special attention is paid to recent evidence from "tasty reactivity" studies of affective reactions to food. I argue that this evidence suggests the following surprising possibilities regarding the functional components and brain substrates of food reward. (1) Reward contains distinguishable psychological or functional components--"liking" (pleasure/palatability) and "wanting" (appetite/incentive motivation). These can be manipulated and measured separately. (2) Liking and wanting have separable neural substrates. Mediation of liking related to food reward involves neurotransmitter systems such as opioid and GABA/benzodiazepine systems, and anatomical structures such as ventral pallidum and brainstem primary gustatory relays. Mediation of wanting related to food reward involves mesotelencephalic dopamine systems, and divisions of nucleus accumbens and amygdala. Both liking and wanting arise from vastly distributed neural systems, but the two systems are separable. (3) Neural processing of food reward is not confined to the limbic forebrain. Aspects of food reward begin to be processed in the brainstem. A neural manipulation can enhance reward or produce aversion but no single lesion or transection is likely abolish all properties of food reward. (4) Both wanting and liking can exist without subjective awareness. Conscious experience can distort or blur the underlying reward process that gave rise to it. Subjective reports may contain false assessments of underlying processes, or even fail at all to register important reward processes. The core processes of liking and wanting that constitute reward are distinct from the subjective report or conscious awareness of those processes.