Verbal and visual dissociation in retrieval practice paradigm in neurodegenerative diseases

As is well known, prefrontal functions assert the control of inhibition for the retrieval of semantically related elements that lead to forgetfulness. The aim of this study is to document any verbal and visual dissociation in the Retrieval- Induced Forgetting (RIF). We want to test the hypothesis that representations in long-term visual memory are sufficiently rooted to be immune to impairment based on recognition, unlike oblivion induced by the retrieval of verbal material. The aim of this project is also to investigate which brain areas are most involved in the processes of inhibition and facilitation in patients with different types of cognitive impairment. To do this, 21 patients with Alzheimer's disease (AD), 21 mild cognitive impairment (MCI), 16 patients with subjective cognitive decline (SCD) and 23 healthy subjects (HS) were enrolled. All participants underwent an extensive neuropsychological evaluation by Addenbrooke's Cognitive Examination-Revised (ACE-R; Mioshi et al., 2006) and the Retrieval Practice Paradigm (RPP), with an experimental recognition task, where the same items were presented both in visual and verbal form to calculate, at a group level, the RIF and FAC (facilitation effect) effects. Only subjects who did not have contraindications to perform MRI, (14 AD, 16 MCI, 14 SCD and 18 HS) underwent 3 T-MRI scanning including a T1-w volume. Voxel-based morphometry (VBM) was used to assess associations between grey matter (GM) volumetrics and RPP indices: items practiced by practiced categories (Rp+), items which were not practiced, but were members of the same category as the Rp+ items (Rp-), items which received no additional retrieval practice and were not members of a practiced category (Nrp) in each group separately. ANOVA models were used to assess cross-sectional differences in all neuropsychological measures and experimental conditions and effects. In each of the 4 groups, separately, Pearson's correlations were used to assess potential association between each domain of ACE-R and the three RPP indices (Nrp, RP +, Rp-). Regarding the results about the RPP indices (Nrp, RP +, Rp-), for the verbal task, we observed a significant between-group difference in Nrp items. Post-hoc showed that the proportion of items in the Nrp condition is statistically higher in HS group than AD group. Moreover, we found a significant effect of group in Rp+ items: post-hoc revealed a significant difference between AD and SCD and between AD and HS. Interestingly, the tendency of significance was present on the comparison between AD and MCI. No significant differences were observed between MCI and SCD and between MCI and HS. Similarly, HS and SCD groups showed comparable performance on Rp+ items. Regarding the RPP effects (RIF and FAC), we observed a significant main effect of group in the FAC effect. Post-hoc revealed significant a main effect of group since AD recognized less items than SCD and HS, respetively. We found also a significant main effect of Condition, because the proportion of items retrieved in the Rp+ condition is statistically higher than that retrieved in the Nrp condition in all groups. Finally, we found no significant Group by Condition interaction indicating that the FAC effect was present in all considered groups. Finally, regarding the RIF effect, we observed a significant main effect of group, due to a significant difference exclusively between AD and HS groups; we revealed no significant effect of condition: indeed, in this case the means of item retrieved in two conditions Nrp and Rp- were almost the same. Finally, we found a significant Group by Condition interaction. The planned comparisons showed that the RIF effect was present in the SCD group only, but no RIF effect was observed in AD, in MCI and HS groups. For the visual task, no significant differences were observed between groups in the accuracy of the Nrp items, RP+ items, RP- items. As regards the FAC effect, we observed a significant main effect of group, due to a significant difference exclusively in the recognition of the global accuracy between AD and HS groups. We found also a significant main effect of Condition: in this case, the Rp+ items were better recalled than Nrp items in all groups. Finally, we observed no significant Group by Condition interaction, indicating that the FAC effect was present in all groups. Furthermore, about the RIF effect, in the verbal task, we observed no significant main effect of group, but significant main effect of Condition, since the Nrp items were better remembered than Rp- items. Moreover, we found no significant Group by Condition interaction: in this case, the planned comparisons revealed that the RIF effect was present in the SCD only. Comparing the groups individually, no significant dissociation emerged between two tasks, verbal and visual. There is only a tendency in SCD, due to greater accuracy of subjects who performed the verbal task, compared to those who performed the visual task, in the Rp- condition. For the MRI analysis results, in the Verbal task, VBM results showed in AD patients, revealed a positive association between Rp- items and GM volumes in the Right Putamen, Cingulus Gyrus and Left Putamen. In the Visual task the VBM analyses, revealed in AD patients a positive association between Nrp items and GM volumes in the Middle Frontal Gyrus, Inferior Frontal Gyrus, pars triangularis and Insular Cortex. Moreover, in SCD subjects, a significant positive association were found between Rp- items and GM volumes in the Postcentral Gyrus, Supramarginal Gyrus, Superior Frontal Gyrus, and Precuneous Cortex, bilaterally. In conclusion, in our study, in line with other experiments (Hogge et al., 2008; Saunders and Summers., 2011; Traykov et al., 2011; Ortega et al., 2012; Serra et al., 2022), shows the presence of an inhibitory effect: the RIF is not evident in the pathological group of AD and MCI, but not even in healthy subjects, which generally show good inhibition capacity; this result could be because the recognition task was too simple for healthy subjects, who were able to recall multiple items without distinction. Although there is a trend, in both tasks, in favor of a RIF effect as the pathology increases. We can also hypothesize that the worse performance observed in the RIF obtained from patients is not a direct expression of an inhibitory deficit but could depend on a general deficit of episodic long-term memory processes.