The orbitofrontal cortex and amygdala are thought to encode information related to the reward value of food (Baxter and Murray, 2002; Holland and Gallagher, 2004; Kringelbach et al., 2003; O’Doherty et al., 2002; Rolls, 2010). The insula processes information related to the taste of food and its hedonic valuation (Balleine and Dickinson, 2000; Small, 2010). The nucleus accumbens and dorsal striatum, which receive dopaminergic input from the ventral tegmental area and substantia nigra, regulate the motivational and incentive properties of food (Baicy et al., 2007; Berridge, 1996, 2009; Farooqi et al., 2007; Malik et al., 2008; Söderpalm and Berridge, 2000). The lateral hypothalamus may regulate rewarding responses to palatable food and drive food-seeking behaviors (Kelley et al., 1996). These brain structures act in a concerted manner to regulate learning about the hedonic properties of food, shifting attention and effort toward obtaining food rewards and regulating the incentive value of environmental stimuli that predict availability of food rewards (Dagher, 2009). For the sake of clarity, not all interconnections between these structures are shown.

To measure reward thresholds, a stimulating electrode is surgically implanted into the lateral hypothalamus of rats, a region in which electrical stimulation is powerfully rewarding and can trigger intense bouts of feeding behavior. After recovery, animals are allowed to self-stimulate this region by turning a wheel. After stable self-stimulation behavior is established, the minimum stimulation intensity that maintained self-stimulation behavior is determined (i.e., the reward threshold). This reward threshold provides an operational measure of the activity of the reward system. Reward thresholds remain stable and unaltered in control rats that have access to standard lab chow and that remain drug naïve. However, thresholds gradually elevate in rats with extended daily access to an energy-dense palatable diet consisting of tasty food items (e.g., cheesecake, bacon, chocolate, etc.). Similarly, reward thresholds progressively elevated in rats that have extended daily access to intravenous cocaine or heroin infusions. Elevated rewards threshold are interpreted to reflect decreased sensitivity of the brain reward system. These effects suggest that overconsumption of palatable food and associated weight gain can induce profound deficits in brain reward similar to those induced by excessive consumption of addictive drugs.

Knockdown of dopamine D2 receptors (D2R) knockdown in rat striatum accelerates the emergence of reward dysfunction and compulsive eating in rats with extended access to palatable food.

The shell region of the nucleus accumbens (NAc) receives innervation from cortical and limbic brain sites and projects to the lateral hypothalamus and ventral pallidum. In turn, the lateral hypothalamus also projects to the ventral pallidum and also the PAG, IFN, VTA, and dorsal raphe nucleus. The IFN and dorsal raphe extend dopaminergic and serotonergic projections, respectively, back to the NAc. The lateral hypothalamus also innervates thalamic (PVN and PON) and epithalamic (LHb) structures. Not shown are the minor projections from the lateral hypothalamus to septal brain areas. 5-HT, serotonin; IFN, interfascicular nucleus; LHb, lateral habenula; PON, preoptic nucleus; PVN, paraventricular nucleus of the thalamus; VTA, ventral tegmental area. Figure is adapted with permission from Thompson and Swanson (2010).

Weight gain is associated with decreased striatal activation in response to palatable food, as measured by fMRI, and lower levels of striatal dopamine D2 receptor (D2R) availability in humans (see text for details).