, 2003, Wilson et al., 2005 and Wilson et al., 2006). Consistent with those findings, behavioral data obtained in three-choice recognition memory tasks from elderly humans show a shift in performance toward pattern completion, as reflected during lure trials by more incorrect “old” responses and fewer correct “similar” responses when compared Lumacaftor solubility dmso to young adults (Toner et al., 2009 and Stark et al., 2010). This error

profile on lure trials is further worsened in aMCI patients compared to normal aging (Yassa et al., 2010). Functional neuroimaging data was first subjected to a one-way analysis of variance (ANOVA) of trial type to select voxels that showed task related activity (see Experimental Procedures). To avoid selection bias while acknowledging the dependence of observations in the two treatment conditions selleck inhibitor for aMCI subjects, this analysis included data from all control subjects and aMCI patients randomly selected from either their placebo or their drug condition such that both treatment conditions were equally represented (approximately half from placebo and half from the drug condition). This analysis maintains independence of observations by including each aMCI patient only once. Unrelated foil items correctly identified as “new” were used as an implicit baseline against

which all other conditions were compared. This analysis resulted in an area of task-related activity localized within the left DG/CA3 subregion of the hippocampus (Figure 2C). To assess whether increased hippocampal activation was observed in the current study of patients with aMCI, we first compared functional activity during fMRI memory task performance between healthy control subjects and patients with aMCI on placebo within the DG/CA3 region as shown in Figure 2C.

The aMCI patients on placebo showed significantly increased blood oxygenation level-dependent (BOLD) activation during lure trials correctly identified as similar when compared to control subjects Cell press (t = 2.056, p = 0.048) (Figure 2E; see also Figure S1). This finding replicates earlier findings in Yassa et al. (2010). Behavioral performance of the aMCI subjects on placebo compared to healthy controls during the scanning task was assessed by the rates of each response option (old, similar, or new) on the critical lure trials. A between-groups ANOVA using only the response categories “old” and “similar” to maintain response independence revealed a significant effect of response type (F(1,32) = 5.357, p = 0.027) and, importantly, a significant group (control versus aMCI placebo) × response interaction (F(1,32) = 7.687, p = 0.009) showing that aMCI patients on placebo incorrectly identified lure items as “old” more often and gave relatively fewer correct responses of “similar” compared to control subjects (Figure 3A). That profile is consistent with reduced pattern separation and a shift to pattern completion in aMCI.