To investigate how neural activation for correctly recalled memories varied across different time delays, we examined the contrast of remote >recent correct trials during object presentation at retrieval (Figure 4, ‘Retrieval fMRI’).
Mean signal differences between correct remote and recent memories.
The figure presents mean signal difference for remote > recent contrast across sessions and groups during the object presentation time window in (A) anterior and posterior hippocampus; (B) anterior and posterior parahippocampal gyrus; (C) cerebellum; (D) medial prefrontal cortex; (E) ventrolateral prefrontal cortex; (F) precuneus; (G) retrosplenial cortex; (H) lateral occipital cortex. Note: Bars indicate the group mean for each session (solid lines for day 1, dashed lines for day 14), plotted separately for children and young adults. Error bars represent ± 1 standard error of the mean. The color indicated the age groups: purple for children and khaki yellow for young adults. Across all panels, the mean of individual subject data is shown with transparent points. The connecting faint lines reflect within-subject differences across sessions. Orange asterisks denote significant difference of remote > recent contrast from zero. An upward orange arrow indicates that this difference is greater than zero, while a downward arrow indicates that this is less than zero. *p<0.05; **p<0.01; ***p<0.001 (significant difference); nonsignificant differences were not specifically highlighted. Significant main and interaction effects are highlighted by the corresponding asterisks. All main and interaction p-values were false discovery rate (FDR)-adjusted for multiple comparisons.
We first tested whether the remote > recent contrast significantly differed from zero in each age group and session (day 1 and day 14), as an indicator of differential engagement during memory retrieval. FDR-adjusted results showed no significant results in the anterior and posterior HC (Figure 4A), anterior PHG (Figure 4B), and RSC (Figure 4G) across sessions and age groups (all p>0.054; see Supplementary file 3 for details). To rule out the possibility that these nonsignificant differences reflect an overall absence of retrieval-related activation, we tested whether mean activation for recent and remote items – each relative to the implicit baseline – was significantly above zero. FDR-adjusted results revealed that activation in these ROIs was significantly greater than zero (all p<0.031), except in the recent day 1 condition in children for the posterior HC (p>0.141) and the precuneus (p>0.056, see Supplementary file 4 and Figure 4—figure supplement 1 for details). These findings indicate that the anterior and posterior HC, anterior PHG, and RSC are similarly engaged during successful retrieval of both recent and remote memories, regardless of delay or age group. (As a control analysis, we tested whether the anterior and posterior HC, anterior PHG, and RSC were similarly engaged during retrieval of recent and remote items over time using the LME models. These models included mean activation relative to the implicit baseline, a Session × Delay × Group interaction, and Subject as a random intercept. The results were consistent with the earlier findings, showing no significant main effect of Delay [all p>0.106], Group [all p>0.060], or Session × Delay interaction [all p>0.340], indicating comparable engagement of these ROIs across delays and age groups [see Supplementary file 6 for full statistical details].) Other ROIs showed more differentiated patterns, which are discussed below. (In contrast, the vlPFC, CE, posterior PHG and LOC, precuneus, and mPFC showed a significant main effect of Delay [all p<0.009, see Supplementary file 5 for details], indicating time-related changes in the remote > recent contrast. These effects are examined in more detail below. Notably, these findings are consistent with results from the whole-brain analyses; Supplementary file 7.)
To further explore the more differentiated patterns observed in other ROIs, we examined changes in the remote >recent contrast across age groups and sessions (day 1 and day 14) using LME models, controlling for sex, handedness, general intelligence, and mean reaction time. All main and interaction effects were FDR-adjusted, and all post hoc tests were Sidak-corrected (see Supplementary file 5 for details).
For the posterior PHG (Figure 4B), a significant Session × Group interaction, F(1,83) = 9.54, p=0.020, ω2=0.09, indicated a more pronounced increase in remote >recent mean signal difference over time in young adults compared to children, b=0.11, t(83) = 3.09, p=0.003.
Similarly, also for the cerebellum (Figure 4C), a significant Session × Group interaction, F(1,161) = 7.68, p=0.020, ω2=0.04, indicated a stronger increase in remote > recent mean signal difference over time in young adults compared to children, b=0.09, t(160) = 2.77, p=0.006.
For the mPFC (Figure 4D), a significant main effect of Group, F(1,86) = 7.61, p=0.023, ω2=0.07, denoted that the overall remote > recent mean signal difference in children was higher than in young adults, b=–0.10, t(86) = –2.76, p=0.007.
For the vlPFC (Figure 4E), a significant main effect of Group, F(1,82) = 31.35, p=<0.001, ω2=0.13, indicated an overall lower remote > recent mean signal difference in children compared to young adults, b=–0.125, t(108) = –3.91, p<0.001. In addition, a significant main effect of Session, F(1,99)=10.68, p=0.005, ω2=0.09, pointed out overall higher remote > recent mean signal difference on day 14 compared to day 1, b=0.08, t(99) = 3.27, p=0.001.
For the precuneus (Figure 4F), a significant main effect of Group, F(1,161) = 5.09, p=0.027, ω2=0.02, indicated an overall lower remote > recent mean signal difference in adults compared to children, b=–0.05, t(160) = –2.26, p=0.037. In addition, a significant main effect of Session, F(1,161) = 6.50, p=0.036, ω2=0.03, denoted an overall lower remote > recent contrast for day 14 compared to day 1, b=–0.05, t(160) = –2.55, p=0.012. Although the remote > recent contrasts were mostly negative, the mean activation for recent and remote items – each relative to the implicit baseline – was significantly greater than zero for all delays and group (all p<0.023), except for children’s recent items on day 1 (p=0.056).
For the LOC (Figure 4H), a significant main effect of Group, F(1,82) = 9.12, p=0.015, ω2=0.09, indicated a higher remote > recent mean signal difference in young adults compared to children, b=0.07, t(82) = 3.02, p=0.003. Additionally, a significant main effect of Session, F(1,97) = 16.76, p=<0.001, ω2=0.14, showed an overall increase in remote > recent mean signal difference on day 14 compared to day 1, b=0.07, t(97) = 4.10, p=<0.001. Furthermore, a significant Session × Group interaction, F(1,81) = 6.42, p=0.032, ω2=0.06, demonstrated higher increase in remote > recent mean signal difference over time in adults compared to children, b=0.09, t(81) = 2.53, p=0.013.
Of note, we conducted an additional univariate analysis using a subsample that included only participants who needed two learning cycles to reach the learning criteria (see Supplementary file 8 for details). The subsampled results fully replicated the findings from the full sample and demonstrated that the amount of re-exposure to stimuli during encoding did not affect consolidation-related changes in memory retrieval at the neural level.
In summary, our findings revealed distinct consolidation-related neural upregulation for remote memory between children and adults. From day 1 to day 14, adults showed a higher increase in remote > recent signal difference for remembered items in the posterior PHG, LOC, and cerebellum than children. Adults showed overall higher remote > recent difference in the vlPFC than children, while children showed overall higher remote > recent difference in the mPFC than adults. Furthermore, we observed a constant activation of anterior and posterior HC, anterior PHG, and RSC in memory retrieval across age groups irrespective of memory type or delay.