The main effect for region and region × measure interactions were not significant (χ (1) = 2.39, P = 0.1219 and χ (1) = 0.01, P = 0.9932, respectively).
While Aβ-42 levels were not different in anterior and posterior white matter regions (main effect of Region: F 1,15) = 1.001, p = 0.333; Figure 2B), there was a trend for them to be higher in anterior than in posterior regions (anterior: M = 45.84 pmol/L, SD = 46.06 pmol/L; posterior: M = 27.56 pmol/L, SD = 37.03 pmol/L).
Aβ-40 levels did not differ between anterior and posterior white matter regions in either the overall ANOVA (main effect of Region: F 1,15) = 0.718, p = 0.410; Figure 2A) or when the regional effect was further probed using a Wilcoxon signed-ranks test (Z = −0.675, p = 0.500).
There was still neither a main effect of region (F 1,14) = 1.513, p = 0.239) nor an interaction between group and region (Group x Region interaction: F 1,14) = 3.237, p = 0.094) after controlling for total cortical neuritic plaque severity.
After statistically controlling for total cortical neuritic plaque severity, there was still neither a main effect of region (F 1,14) = 0.940, p = 0.349) nor an interaction between group and region (Group x Region interaction: F 1,14) = 0.141, p = 0.713).
As shown in Figure 3 and Table 5, besides a main effect of Region, we did not find significant interaction effects with Region during encoding.
P3a latency was also analysed and a main effect of Region [F (1, 34) = 6.228, p =.018] showed shorter latency in the frontal region in comparison to the central region.
A main effect of region [F 1,30) = 7.9, p<0.01] suggested that most negative voltages occurred at anterior regions.
We did this by including interaction terms by region, adding: α r + ∑ r d r, q X i, q * M i separately to the model for each covariate q, where α r is a categorical variable for the main effect of region, and M i is a zero/one indicator for whether location i is in a given region or not.
This showed there to be a significant effect of model (F3.48 = 6.579, P = 0.001. η2 = 0.291), which was driven by stronger encoding of the CV structure model compared with all other models (all Ps < 0.007), and a marginal region by model interaction (F6,96 = 1.973, P = 0.077, η2 = 0.110), with no main effect of region (F2,32 = 1.906, P = 0.165, η2 = 0.106).
