نوسانات وابسته به چرخه قاعدگی در هورمونهای تخمدان موثر بر حافظه هیجانی

The hormones progesterone and estradiol modulate neural plasticity in the hippocampus, the amygdala and the prefrontal cortex. These structures are involved in the superior memory for emotionally arousing information (EEM effects). Therefore, fluctuations in hormonal levels across the menstrual cycle are expected to influence activity in these areas as well as behavioral memory performance for emotionally arousing events. To test this hypothesis, naturally cycling women underwent functional magnetic resonance imaging during the encoding of emotional and neutral stimuli in the low-hormone early follicular and the high-hormone luteal phase. Their memory was tested after an interval of 48 h, because emotional arousal primarily enhances the consolidation of new memories. Whereas overall recognition accuracy remained stable across cycle phases, recognition quality varied with menstrual cycle phases. Particularly recollection-based recognition memory for negative items tended to decrease from early follicular to luteal phase. EEM effects for both valences were associated with higher activity in the right anterior hippocampus during early follicular compared to luteal phase. Valence-specific modulations were found in the anterior cingulate, the amygdala and the posterior hippocampus. Current findings connect to anxiolytic actions of estradiol and progesterone as well as to studies on fear conditioning. Moreover, they are in line with differential networks involved in EEM effects for positive and negative items.

Estradiol (E2) and progesterone (P4) modulate neuronal plasticity in a variety of brain areas, where most evidence accumulated so far for hormonal effects in the amygdala, the hippocampus and the prefrontal cortex (Foy et al., 1999, Hao et al., 2006, McEwen and Woolley, 1994, Murphy and Segal, 1996 and Tang et al., 2004). These cellular effects of E2 and P4 stimulated research about hormonal influences on various cognitive and emotional functions. Consistent with their opponent effects on hippocampal neurotransmission, E2 exerts beneficial effects on memory that are reversed by P4 (Bimonte-Nelson, Francis, Umphlet, & Granholm, 2006; Gibbs et al., 2004 and Harburger et al., 2007). However, hormonal influences differ across brain regions, with E2 stimulating neural transmission for instance in the hippocampus while decreasing neuronal excitability in the amygdala (Foy et al., 1999 and Womble et al., 2002). Moreover, E2 and P4 act opponent on some cellular processes in the hippocampus but exert conjoint neuroprotective effects in the prefrontal cortex (Djebaili et al., 2004, Foy et al., 2008, Hill et al., 2009 and Kritzer and Kohama, 1998). The superior memory formation for emotionally arousing information, i.e. the emotional enhancement of memory (EEM), depends upon the interplay between all these structures. Therefore, predictions concerning hormonal influences on this memory function are less straightforward (for a review, see Dolcos, Denkova, & Dolcos, 2012). Prefrontal areas, the hippocampus and the amygdala contribute differentially to the EEM, i.e. to distinct aspects and at different stages of emotional memory formation. When an emotional stimulus is encountered, a network between the amygdala and cortical areas rapidly initiates emotional processing and the allocation of attentional resources (Davis and Whalen, 2001 and Pessoa, 2008). A prefrontal-parietal network then mediates the elaborate cognitive processing of stimulus valence (Canli et al., 1999, Dolcos et al., 2004, Kensinger and Schacter, 2006 and Mickley and Kensinger, 2008). After the initial processing, emotional arousal leads to enhanced hippocampal consolidation of arousing stimuli via the amygdala and increased noradrenergic neurotransmission (Huff et al., 2013, McGaugh, 2004, Roozendaal et al., 1999 and Strange and Dolan, 2004). Consequently, the effect of emotional arousal on memory requires a consolidation-delay between encoding and retrieval (Schwarze et al., 2012 and Sharot et al., 2007). Because E2 and P4 affect cellular mechanisms in brain areas mediating initial processing but also encoding and consolidation of emotionally arousing information, it is conceivable that hormonal fluctuations modulate the EEM effect at various stages. In animal models, both hormones exert anxiolytic and antidepressant actions (Frye et al., 2000, Frye and Walf, 2004, Lund et al., 2005, Reddy et al., 2005, Shirayama et al., 2011 and Walf and Frye, 2006). Functional imaging in humans reveals that E2 and P4 change reactivity to emotional stimuli in the amygdala, the hippocampus and the medial prefrontal cortex (Derntl et al., 2008, Goldstein et al., 2005, Guapo et al., 2009, Ossewaarde et al., 2010, Sripada et al., 2013 and Van Wingen et al., 2007). Likewise, E2 enhances extinction of contextual and cued fear conditioning in both animal models and humans (Graham and Milad, 2013, Gupta et al., 2001, Markus and Zecevic, 1997, Milad et al., 2009, Milad et al., 2006, Milad et al., 2010 and Zeidan et al., 2011). Hormonal influences on declarative emotional memory have been investigated only in studies employing between-subject designs, in which women were Post Hoc grouped according to their hormonal states (Ertman et al., 2011, Felmingham et al., 2012 and Nielsen et al., 2013). In detail, Ertman et al. (2011) found better memory for negative images in the second half relative to the first half of the menstrual cycle and a positive relationship between P4 levels and recall of negative images. In contrast, Felmingham et al. (2012) did not detect significant differences in memory for negative images between a high- and a low-P4 group. Nielsen et al. (2013) showed that only women in the second, but not the first, half of the menstrual cycle recalled more details from an emotional than a neutral story. Memory performance in this study did not show significant relationships to P4 or E2 levels. The current study aims to further explore how natural hormonal variations affect the superior consolidation of emotional arousing information within the same women. Therefore, functional neuroimaging was performed in healthy young women while encoding positive, negative and neutral pictures once during menstruation in the early follicular phase, where both hormones are low, and once in the middle of the luteal phase where the concentration of P4 is at its maximum and E2 reaches a second peak. These phases of relatively stable hormone levels were chosen to reduce the variance between women. Moreover, the chosen cycle phases avoid the effects of sharp increase and decrease in E2 and P4 on cognition but also of the luteinizing hormone that rises shortly before ovulation ( Berry et al., 2008, Casadesus et al., 2007 and Hausmann et al., 2000). In order to assess the effects of hormones on arousal-stimulated consolidation rather than the initial processing of emotional stimuli, memory was tested two days after encoding. Memory accuracy was assessed by means of d-prime. Additionally, memory quality was examined by estimating the process parameters recollection and familiarity of a prominent dual-process recognition memory model using confidence ratings ( Yonelinas, 2002). In this model, recollection represents the process of remembering an item together with contextual information, such as thoughts, feelings or other details of the learning context. As such, memory based on recollection is always accompanied by high confidence. Familiarity is the sense of having something encountered before, but without recollecting any further details. Familiarity can be accompanied by the full range of confidence levels. Arousal ratings were acquired from each subject on a 9-point Likert scale for target images and stimuli were grouped according to arousal ratings into either high (HA) or low arousing (LA). According to the goal of the study, analyses were primarily conducted on HA emotional and LA neutral stimuli. Menstrual cycle phases were verified by assessments of E2 and P4 in saliva.