Abstract
BACKGROUND: The aim of this study was to explore the construct validity and diagnostic properties of the Reading the Mind in the Eyes Test (RMET) in non-demented patients with amyotrophic lateral sclerosis (ALS). MATERIALS: A total of 61 consecutive patients and 50 healthy controls (HCs) were administered the 36-item RMET. Additionally, patients underwent a comprehensive assessment of social cognition via the Story-Based Empathy Task (SET), which encompasses three subtests targeting Causal Inference, Emotion Attribution (SET-EA), and Intention Attribution (SET-IA), as well as global cognitive [the Edinburgh Cognitive and Behavioral ALS Screen (ECAS)] and behavioral screening [the Frontal Behavioral Inventory (FBI); the Dimensional Apathy Scale (DAS); the Beck Depression Inventory (BDI); and the State and Trait Anxiety Inventory-Y]. The construct validity of the RMET was tested by regressing it within a stepwise model that encompassed as predictors the abovementioned cognitive and behavioral measures, covarying for demographic and motor confounders. Receiver-operating characteristics (ROC) analyses allowed exploring intrinsic and post-test properties of the RMET both in discriminating patients from HCs and in identifying patients with a defective SET-EA performance. RESULTS: The RMET was solely predicted by the SET-EA (p = 0.003) and SET-IA (p = 0.005). RMET scores showed high accuracy both in discriminating patients from HCs (AUC = 0.81) and in identifying patients with a defective SET-EA score (AUC = 0.82), with adequate-to-optimal both intrinsic and post-test properties. DISCUSSION: The RMET is a convergently and divergently valid measure of affective social cognition in non-demented ALS patients, also featuring optimal intrinsic and post-test diagnostic properties in both case-control and case-finding scenarios.