Figure 1: Experimental design and results from physiology and behaviour.

(a,b) The experiment consisted of three phases, including acquisition, Think/NoThink and post-scan memory test. During the acquisition phase outside the scanner, participants performed two training sessions on day 1 and day 2, which occurred 24 h and 30 min before the Think/NoThink phase, respectively. Participants were trained to memorize 26 pairs in each of the two acquisition sessions. (c) During the Think/NoThink phase, participants underwent fMRI with concurrent recording of skin conductance responses (SCR) while they were performing a ‘Think/NoThink’ task. (d) During the test phase, participants were given faces as cues and asked to recall their associated target pictures. Behavioural suppression scores were calculated separately for newly acquired (that is, 30-min) and overnight (that is, 24-h) aversive memories based on the difference in memory accuracy between their respective baseline items and NoThink items accordingly. A total number of 18 participants were included in the final analysis. (e) Bar graphs depict SCR associated with ‘Think’ (green) and ‘NoThink’ (red) conditions for aversive memories acquired either 24 h or 30 min ago (referred to as remote or recent memories when appropriate). (f) Bar graphs depict cued-recall accuracy for newly acquired and consolidated aversive memories as a function of ‘Think’, ‘NoThink’ and ‘Baseline’ (gray) during the test phase. (g) Bar graphs depict suppression scores for both newly acquired and overnight consolidated aversive memories. Error bars represent standard error of mean. *P<0.05; **P<0.01.