What’s in a name? Semantic, as well as episodic, memory plays an important role in the diagnostic sensitivity of the free and cued selective reminding test in MCI

Abstract

AbstractBackgroundThe Free and Cued Selective Reminding Test (FCSRT) is highly sensitive in the detection of AD, even at early stages (Grober, et al., 2008; Papp et al., 2015; Slachevsky et al., 2018). By controlling encoding and retrieval, the task is especially sensitive to memory consolidation deficits that characterize AD. During the learning phase, patients name items that are paired with semantic cues aimed to boost encoding and retrieval. However, benefit from the semantic cues requires intact semantic memory and access. As the “cognitive signature” of AD includes reduced semantic processing (Henry and Crawford, 2004.), we asked whether this reduction impacts FCRST performance.Method170 patients (M age=74.1±6.2, M edu=16±3.0; 57% male) clinically referred for neuropsychological evaluation to address cognitive decline completed the FCSRT and semantic access measures, the Boston Naming Test (BNT) and semantic (animal) fluency. Neuropsychologists independently and blindly assigned diagnosis specifying NCD severity (Mild v. Major) and type (amnestic v. non‐amnestic). To determine the impact of semantic processing on FCRST performance we examined whether naming accuracy (NA): (1) differed between patient groups on first FCSRT learning trial; (2) was impacted by item prototypicality, 3) was predicted by BNT and semantic fluency performances, and (4) predicted subsequent FCSRT performance (total free, and total cued recall accuracy).ResultSs included 50 Mild NCD, amnestic; 59 Mild NCD, non‐amnestic; 23 Major NCD, amnestic; and 38 Major NCD, non‐amnestic. NA on the first FCSRT learning trial was higher for Mild than Major NCD and non‐amnestic than amnestic patients without severity x type interaction. Across all groups: NA was predicted by BNT (p < .05) and semantic fluency (p < .05), and was better for high than for low prototypicality items. Also across all groups, NA predicted subsequent memory for the items, including total free recall accuracy (p < .001) and total cued recall accuracy (p < .001).ConclusionThe diagnostic sensitivity of the FCSRT in AD is likely impacted not only by reduced episodic memory but also by reduced semantic memory and access. Understanding what underlies the sensitivity of the FCSRT may ultimately help improve the detection of early cognitive changes in AD.