Background: In this meta-analysis we review the findings from neuropsychological studies on set-shifting in people with eating disorders (EDs) or overweight/obesity.
Method: Four databases (PubMed, PsycINFO, PSYNDEX and Web of Science) were searched for eligible studies. Effect sizes (ESs) were pooled using random-effects models. Moderator analyses were conducted for ED and overweight/obese subgroups, adult/adolescent samples and measures of set-shifting.
Results: Sixty-four studies with a total of 1825 ED patients [1394 anorexia nervosa (AN), 376 bulimia nervosa (BN) and 55 binge eating disorder (BED)] and 10 studies with a total of 449 overweight/obese individuals were included. The meta-analysis revealed a small to medium ES for inefficient set-shifting across all three ED diagnoses (Hedges' g = -0.45). Subgroup analyses yielded small to medium ESs for each ED subtype (g = -0.44 for AN, -0.53 for BED, -0.50 for BN), which did not differ significantly. There was a medium ES for restricting type AN (ANR; g = -0.51) but no significant ES for binge/purge type AN (AN/BP; g = -0.18). A medium ES was found across obesity studies (g = -0.61). The ES across overweight studies was not significant (g = -0.07). Adult samples did not differ from adolescent samples in either ED or overweight/obesity studies. The different set-shifting measures were associated with largely varying ESs.
Conclusions: The meta-analysis provides strong support that inefficient set-shifting is a salient neuropsychological phenomenon across ED subtypes and obesity, but is less prominent in AN/BP and overweight. Compulsivity seems to be a common underlying factor supporting a dimensional and transdiagnostic conceptualization of EDs and obesity.