A new focus in the field of emotional memory is the study of sex-related differences. Whether the sex-related lateralization of amygdala function (i.e., the female-left/male-right effect) in the emotional enhancement of memory (EEM) is time-dependent remains unclear. To evaluate this phenomenon, we conducted a two time-point study (20 min vs. 24 h) using fMRI and behavioral paradigms. We found that the right amygdala predicted 20-min EEM, while the left amygdala predicted 24-h EEM. The sex-related lateralization of amygdala function was not detected in either the 20-min or the 24-h EEM. Our results further confirm and extend the idea that the amygdala exhibits a lateralized and time-dependent dissociation, occurring even in the 24-h EEM relative to the 20-min EEM. The present and previous studies indicate that sex-related lateralization of amygdala function occurs in the 2- to 3-week EEM, but it does not occur in the 1-week, 24-h, or less than 30-min EEM, suggesting that this effect on emotional memory may also be time-dependent.
It is easier to remember emotionally arousing events than it is to recall those that are neutral (Bradley, Greenwald, Petry, & Lang, 1992). This phenomenon is known as the emotional enhancement of memory (EEM) (Sommer et al., 2008 and Talmi et al., 2008). Sex-related differences have become a new focus in the study of emotional memory (Cahill, 2006). An important finding from imaging studies is that successful encoding of emotional memory activates the left amygdala in women but the right amygdala in men (Cahill et al., 2004, Cahill et al., 2001, Canli et al., 2002 and Mackiewicz et al., 2006). This female-left/male-right effect of amygdala activity was coined the “sex-related lateralization of amygdala function” (Cahill, 2003), or the “sex-related difference” in the present study. Other recent studies, however, have failed to find this effect (Kensinger and Schacter, 2006, Kensinger and Schacter, 2008 and Talmi et al., 2008), although the reason for this inconsistency remains to be elucidated.
McGaugh (2000) defined second-to-hour delayed memory as short-term memory (STM) and hour-to-month delayed memory as long-term memory (LTM). The processes of STM and LTM are dissociated or time-dependent. Previous studies have shown that EEM occurs over both short (several minutes) (Kensinger and Corkin, 2004 and Sergerie et al., 2006) and long (one day to weeks) time periods (Adolphs et al., 2000, Cahill et al., 2004 and Canli et al., 2000), corresponding to short-term EEM and long-term EEM, respectively. Neuroimaging studies have demonstrated that the involvement of the amygdala is time-dependent and dissociated between short-term EEM and long-term EEM (Hamann et al., 1999, Mackiewicz et al., 2006 and Ritchey et al., 2008). Thus, a key factor underlying the inconsistent results observed in the sex-related differences could be the differences in delay time. To date, few studies have explored this discrepancy. It is therefore necessary to systematically investigate whether the sex-related differences are time-dependent.
Previous imaging studies on emotional memory, particularly those investigating sex-related differences, have often been designed using a single time-point model (Cahill et al., 2001, Cahill et al., 2004, Canli et al., 2002, Dolcos et al., 2004 and Kensinger et al., 2007) that only focused on one stage of EEM (Hamann, 2001 and LaBar and Cabeza, 2006). To further evaluate the time-dependence of sex-related differences in EEM, multiple time-point experiments investigating the neural correlates of the different stages of EEM are needed. To our knowledge, only a limited number of imaging studies have adopted the two time-point model for EEM (Hamann et al., 1999), and even fewer have accounted for sex-related differences (Mackiewicz et al., 2006 and Ritchey et al., 2008). These studies all adopted a relatively longer delay (1–4 weeks) for the long-term EEM, while behavioral experiments have indicated that EEM can also be observed in 24-h delayed memory (Sharot and Phelps, 2004 and Sharot and Yonelinas, 2008). Moreover, neurological studies have suggested that lesions of the amygdala can impair 24-h long-term EEM (Adolphs et al., 2000 and Buchanan et al., 2001). To date, however, few imaging studies have adopted a delay time of 24 h. The primary goal of this study was to confirm whether the involvement of the amygdala is dissociated between the 20-min and the 24-h EEM and further explore whether the sex-related lateralization of amygdala function is time-dependent in EEM.
The results of d′S, d′L, Δd′S, and Δd′L are listed in Supplementary Table 1. A MANOVA showed that the interactions of emotional categories (F, S, H) × memory session (Δd′S, Δd′L) were significant [F(2, 108) = 3.838, p < 0.05], and no main effect of sex was observed [F(2, 108) = 0.257, p > 0.50]. During short- and long-term recognition, a repeated measures ANOVA yielded a significant main effect of emotional category for Δd′S [F(2, 57) = 5.592, p < 0.01] and Δd′L [F(2, 57) = 20.005, p < 0.001], and the post hoc tests (LSD) showed that the Δd′ of fear pictures was higher than that of sad and happy pictures at both time points (p < 0.05). A paired samples t test revealed that Δd′L was significantly larger than Δd′S for fear pictures (p < 0.05), while there was no significant difference between Δd′S and Δd′L for happy and sad pictures ( Fig. 2).
