Clinical investigation of set-shifting subtypes in anorexia nervosa

• Clinical correlates of set-shifting in anorexia nervosa (AN) are unclear.
• According to patients׳ set-shifting we split them into poor and intact subtypes.
• Eating psychopathology and depression differed between AN set-shifting subtypes.
• AN set-shifting subtypes scored differently on personality dimensions.
• Multiple clinical variables may be correlated with set-shifting performances in AN.

While evidence continues to accumulate on the relevance of cognitive inflexibility in anorexia nervosa (AN), its clinical correlates remain unclear. We aimed at examining the relationship between set-shifting and clinical variables (i.e., eating psychopathology, depression, and personality) in AN. Ninety-four individuals affected by AN and 59 healthy controls (HC) were recruited. All participants were assessed using: Eating Disorders Inventory-2 (EDI-2), Temperament and Character Inventory (TCI), Beck Depression Inventory (BDI), and Wisconsin Card Sorting Test (WCST). The AN group scored worse than HCs on set-shifting. According to their neuropsychological performances, AN patients were split into two groups corresponding to poor (N=30) and intact (N=64) set-shifting subtypes. Interoceptive awareness, impulse regulation, and maturity fears on the EDI-2 and depression on the BDI differed across all groups (HC, intact, and poor set-shifting subtype). Self-directedness on the TCI differed significantly among all groups. Cooperativeness and reward dependence differed instead only between HC and AN poor set-shifting subtype. After controlling for depression, only interoceptive awareness remained significant with reward dependence showing a trend towards statistical significance. These findings suggest that multiple clinical variables may be correlated with set-shifting performances in AN. The factors contributing to impaired cognitive inflexibility could be more complex than heretofore generally considered.