The effect of exogenous cortisol during sleep on the behavioral and neural correlates of emotional memory consolidation in humans
Introduction
A striking and adaptive feature of human memory is its propensity to prioritize biologically salient over emotionally neutral information (Hamann, 2001, LaBar and Cabeza, 2006). An established animal literature posits that this emotional enhancement effect is at least partly related to neuromodulatory action during memory consolidation (McGaugh and Roozendaal, 2002). Stress hormones like cortisol (corticosterone in rats) are, together with stress-related neurotransmitters like norepinephrine, thought to exert their modulatory effect through potentiation of the amygdala. The amygdala subsequently modulates mnemonic processing in the hippocampus and other brain regions (McGaugh, 2004). Although well established in animals, these effects remain relatively uninvestigated in humans, despite recent reports on a possible role for glucocorticoids in the prevention of traumatic memory (de Quervain et al., 2009).
During memory consolidation, initially fragile memory traces are reorganized and integrated into long-term storage (Müller and Pilzecker, 1900, Marr, 1970, McGaugh, 2000). This process occurs when awake, but is particularly facilitated during sleep (Stickgold, 2005, Diekelmann and Born, 2010). In case of emotional memory, current theories suggest that this consolidation process additionally serves to reduce the affective tone of the memory trace by integrating it into conceptualized, semantic networks (Walker and van der Helm, 2009). Within neuroimaging, memory consolidation is often probed by measuring the outcome of consolidation during retrieval, for instance using a remote versus recent memory paradigm (Takashima et al., 2006).
Previous studies using similar designs have shown that the emotional enhancement effect of memory develops progressively over consolidation (Nishida et al., 2009), and that the retrieval of consolidated emotional memories engages the amygdala and hippocampus (Dolcos et al., 2005, Sterpenich et al., 2009). However, none of these studies consider the role of cortisol. On a behavioral level, several studies have suggested a facilitatory role of cortisol in (emotional) memory consolidation (Buchanan and Lovallo, 2001, Cahill et al., 2003, Maheu et al., 2004, Abercrombie et al., 2006, Kuhlmann and Wolf, 2006, Payne et al., 2007, but see: Wagner et al., 2005b, and see Lupien et al., 2007, de Quervain et al., 2009 for reviews). However, these findings are either complicated by concomitant drug effects on encoding, or related to other stress-induced neuromodulatory changes in addition to cortisol, most importantly norepinephrine. A recent study by Wilhelm and colleagues did study the isolated effect of cortisol on consolidation, but only found an effect on relational (and not item) memory of neutral text material (Wilhelm et al., 2011). Thus, it remains unknown to what extent the reorganization of emotional memory traces as a result of prioritized memory consolidation directly depends on cortisol.
We assessed the effect of hydrocortisone administration on the consolidation of negative and neutral pictures studied just prior to polysomnographically monitored sleep (remote set), in a double-blind, placebo-controlled, between-subjects design. Memory consolidation was probed the following evening during a recognition memory test in the MR scanner by assessing the difference in memory performance and brain activity for the ‘remote’ items and new, ‘recent’ items studied just prior to test. We administered hydrocortisone after learning and just before sleep, because processes underlying memory consolidation are thought to take place most prominently, although not exclusively, during sleep (Diekelmann and Born, 2010). Remote items thus transitioned through a phase of memory consolidation during sleep, which in the cortisol group was accompanied by elevated cortisol levels. In contrast, recent items were neither consolidated, nor affected by the cortisol manipulation. Note that although this experimental setup targets memory consolidation during sleep, it cannot establish sleep dependence. We did not include a non-sleep condition at a different time of day or a sleep deprivation condition because such controls would be confounded by cortisol differences resulting respectively from the diurnal cycle of cortisol or sleep deprivation itself (Balbo et al., 2010).
Based on extensive animal literature concerning the facilitatory role of glucocorticoids in the memory consolidation of emotionally arousing events (McGaugh and Roozendaal, 2002), we hypothesized that hydrocortisone administration during post-encoding sleep would enhance the consolidation benefit for emotional (vs. neutral) memory, which would be reflected in altered reactivation patterns of limbic circuitry during retrieval.
Section snippets
Participants
Forty-two healthy male volunteers participated in this study, which was approved by the local ethics committee (CMO region Arnhem-Nijmegen, The Netherlands) in accordance with the declaration of Helsinki. Only men were included to minimize heterogeneity related to gender differences in HPA-axis activity (Wang et al., 2007). Volunteers were screened before entering the study. Prior to screening, they were informed about all procedures and risks and asked to sign informed consent. At screening,
Cortisol measurements
Hydrocortisone administration resulted in a significant increase in total cortisol concentration during the post-encoding night as compared to placebo (measured as AUCg, T(37) = 3.2, p < 0.005), while not affecting cortisol levels during the recent encoding and retrieval on day 2 (T(37) = 0.1, p = 0.9) (Table 1). Baseline cortisol levels were not different between groups (T(37) = 0.9, p = 0.4).
Behavioral results
As expected, we observed main effects of picture valence (negative > neutral, F(1,37) = 31.8, p < 0.0001, Pη2 = 0.46), and
Discussion
We investigated the effect of cortisol on emotional memory consolidation using a remote/recent memory paradigm. Administration of 20 mg of hydrocortisone resulted in a substantial increase of cortisol levels during the night, with no group difference at recent encoding or test. In the CORT group we observed an increased relative retention benefit for emotional (vs. neutral) items. This prioritization of emotional memory consolidation by cortisol was accompanied by reduced activation in
Role of funding source
This work was supported by Grant 918.66.613 and 451.07.019 from the Netherlands Organization for Scientific Research.
Conflict of interests
S.O. has received speaker fees from Boehringer, Ingelheim, UCB Pharma and Novartis. The other authors report no biomedical financial interests or potential conflicts of interest.
Acknowledgement
We would like to thank Maarten van Hal for scoring the sleep data.
References (56)
- et al.
Endogenous cortisol elevations are related to memory facilitation only in individuals who are emotionally aroused
Psychoneuroendocrinology
(2006) - et al.
Measuring emotion: the self-assessment manikin and the semantic differential
J. Behav. Ther. Exp. Psychiatr.
(1994) - et al.
Response and habituation of the human amygdala during visual processing of facial expression
Neuron
(1996) - et al.
Enhanced memory for emotional material following stress-level cortisol treatment in humans
Psychoneuroendocrinology
(2001) - et al.
The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research
Psychiatr. Res.
(1989) - et al.
Glucocorticoids and the regulation of memory in health and disease
Front. Neuroendocrinol.
(2009) - et al.
Initial posttraumatic urinary cortisol levels predict subsequent PTSD symptoms in motor vehicle accident victims
Biol. Psychiatr.
(2000) Cognitive and neural mechanisms of emotional memory
Trends Cogn. Sci.
(2001)- et al.
The effects of stress and stress hormones on human cognition: implications for the field of brain and cognition
Brain Cogn.
(2007) - et al.
Role of adrenal stress hormones in forming lasting memories in the brain
Curr. Opin. Neurobiol.
(2002)