Abstract
Anorexia nervosa (AN) is a chronic, life-threatening disease with mental and physical components that include excessive weight loss, persistent food restriction, and altered body image. It is sometimes accompanied by hyperactivity, day-night reversal, and amenorrhea. No medications have been approved specific to the treatment of AN, partially due to its unclear etiopathogenesis. Because adiponectin is an appetite-regulating cytokine released by adipose tissue, we hypothesized that it could be useful as a specific biomarker that reflects the disease state of AN, so we developed a modified AN mouse model to test this hypothesis. Twenty-eight 3-week-old female C57BL/6J mice were randomly assigned to the following groups: 1) no intervention; 2) running wheel access; 3) food restriction (FR); and 4) activity-based anorexia (ABA) that included running wheel access plus FR. After a 10-day cage adaptation period, the mice of the FR and ABA groups were given 40% of their baseline food intake until 30% weight reduction (acute FR), then the body weight was maintained for 2.5 weeks (chronic FR). Running wheel activity and the incidence of the estrous cycle were assessed. Spontaneous food restriction and the plasma adiponectin level were evaluated at the end of the acute and chronic FR phases. An increase in running wheel activity was found in the light phase, and amenorrhea was found solely in the ABA group, which indicates that this is a good model of AN. This group showed a slight decrease in spontaneous food intake accompanied with an attenuated level of normally induced plasma adiponectin at the end of the chronic FR phase. These results indicate that the plasma adiponectin level may be a useful candidate biomarker for the status or stage of AN.