Chronic stress promotes palatable feeding, which reduces signs of stress: feedforward and feedback effects of chronic stress.

We suggested a new model of the effects of glucocorticoids (GCs) exerted during chronic stress, in which GCs directly stimulate activities in the brain while indirectly inhibiting activity in the hypothalamo-pituitary-adrenal (HPA) axis through their metabolic shifts in energy stores in the periphery. This study is an initial test of our model. In a 2 x 2 design, we provided ad lib access to calorically dense lard and sucrose (comfort food) + chow or chow alone, and repeatedly restrained half of the rats in each group for 5 d (3 h/d). We measured caloric intake, body weight, caloric efficiency, ACTH, corticosterone (B), and testosterone during the period of restraint and leptin, insulin, and fat depot weights, as well as hypothalamic corticotropin-releasing factor mRNA at the end of the period. We hypothesized that chronically restrained rats would exhibit a relative increase in comfort food ingestion and that these rats would have reduced HPA responses to repeated restraint. Although total caloric intake was reduced in both groups of restrained rats, compared with controls, the proportion of comfort food ingested increased in the restrained rats compared with their nonrestrained controls. Moreover, caloric efficiency was rescued in the stressed, comfort food group. Furthermore, ACTH and B responses to the repeated restraint bouts were reduced in the rats with access to comfort food. Corticotropin-releasing factor mRNA was reduced in control rats eating comfort food compared with those eating chow, but there were no differences between the stressed groups. The results of this experiment tend to support our model of chronic effects of stress and GCs, showing a stressor-induced preference for comfort food, and a comfort-food reduction in activity of the HPA axis.