Animal and human studies have consistently shown differential effects of acute stress on memory depending on the memory phases (Goldfarb et al., 2019, Roozendaal, 2002). The enhancing or impairing effects of acute stress on episodic memory have been attributed to stress-induced physiological responses (Quaedflieg & Schwabe, 2018), particularly to the increase of glucocorticoid levels (cortisol in humans, corticosterone in rodents) (Lupien et al., 2007, Shields et al., 2017, Vogel and Schwabe, 2016). Findings revealed a consolidation enhancement, especially for emotionally arousing material (for a recent review Sazma et al., 2019; for a recent meta-analysis Shields et al., 2017). Conversely, studies consistently indicate detrimental effects of acute stress on memory retrieval (for a recent review Gagnon & Wagner, 2016; for a recent meta-analysis Shields et al., 2017). Studies employing a variety of stress-induction paradigms, such as pain or psychosocial stress, reliably demonstrated that acute stress impairs retrieval (Gagnon et al., 2019, Goldfarb et al., 2019; Kuhlmann, Piel, et al., 2005; Schwabe & Wolf, 2009). These detrimental effects have been directly attributed to endogenous elevation of cortisol after acute stress (Lupien et al., 2007). Findings suggest that participants who exhibit a stronger cortisol response to a stressor recall fewer words during a free recall task compared to participants who do not display a similar cortisol response or control participants (Buchanan et al., 2006). It is noteworthy to observe that the majority of the studies explored a delay of 20 to 30 min between the stressor and the retrieval tasks (Quaedflieg and Schwabe, 2018, Shields et al., 2017). This delay allows to reach salivary cortisol peak levels following a stressor (Kirschbaum and Hellhammer, 1989, Kirschbaum and Hellhammer, 1994, 2000; Kumar et al., 2005, Vining et al., 1983).

Consistent with acute stress-induction studies, pharmacological studies of cortisol administration, inducing a rise in cortisol levels exogenously, have consistently demonstrated a memory retrieval impairment (de Quervain et al., 1998, de Quervain et al., 2000; Kuhlmann, Kirschbaum, et al., 2005; Tollenaar et al., 2009). This memory retrieval impairment has typically been observed in studies where memory retrieval was tested within a time window up to 60 min following pharmacological cortisol administration (Buss et al., 2004, de Quervain et al., 2000, de Quervain et al., 2003, de Quervain et al., 2007; Kuhlmann, Kirschbaum, et al., 2005; Kuhlmann and Wolf, 2005, Kuhlmann et al., 2005, Kuhlmann et al., 2005, Kuhlmann and Wolf, 2006; but see Antypa et al., 2022, Tollenaar et al., 2009 for variations in delays). While there is a significant increase in cortisol levels compared to placebo administration with this delay (Buss et al., 2004, de Quervain et al., 2000; Kuhlmann, Kirschbaum, et al., 2005), it appears that salivary cortisol levels peak later, i.e. after 60 min, e.g. around 75 min for a 10 mg dose in Antypa et al., 2022; around 100 min for a 10 mg dose in Tops et al., 2003. To the best of our knowledge, only one study has investigated the effects of cortisol administration after a delay longer than 60 min and expanded the observed retrieval impairment (Tops et al., 2003). Here, we decided to extend the literature on the effect of cortisol administration on memory retrieval at a longer delay than 60 min. This decision was made to ensure that salivary cortisol levels were around peak levels (for a 10 mg dose).

Considering human studies with pharmacological cortisol administration, the majority have focused on investigating the effects of elevated cortisol levels on episodic memory using controlled laboratory material such as words (i.e. list of words, pairs of words) or pictures (Antypa et al., 2022, de Quervain et al., 2000; Kuhlmann, Kirschbaum, et al., 2005; Kuhlmann and Wolf, 2006, Roozendaal et al., 2003, Schilling et al., 2013, Tollenaar et al., 2009, Tops et al., 2003). However, the retrieval of real-world-like events is notably more intricate and complex than retrieval of simpler material used in the laboratory. Therefore, it is of utmost importance to investigate the impact of cortisol administration on the retrieval of ecologically-relevant material.

In addition, the nature of retrieval tasks (e.g. free recall, cued recall or recognition) appears to influence the findings regarding memory retrieval following cortisol administration (Gagnon & Wagner, 2016). After cortisol administration, free recall is consistently impaired (de Quervain et al., 2000, Kuhlmann and Wolf, 2005, Kuhlmann et al., 2005, Kuhlmann et al., 2005, Tops et al., 2003), while findings on cued recall and recognition are inconsistent with either impairing effects or no impact (Antypa et al., 2022, Buss et al., 2004, de Quervain et al., 2003, Tops et al., 2003). These inconsistencies may be directly associated with the complexity of the tasks, where recognition, being less demanding, is less susceptible to the influence of cortisol (Buchanan et al., 2006, Gagnon and Wagner, 2016). Concerning cued recall, the administration of different doses of cortisol has highlighted an inverted U-shape between cued recall and cortisol levels, indicating a performance-dose relationship (Schilling et al., 2013). Moreover, a concurrent noradrenergic activation might be necessary to induce memory retrieval impairment across all type of tasks (Gagnon and Wagner, 2016, Roozendaal et al., 2004). Given these discrepancies observed depending on the memory task type, it is vital to incorporate various memory tasks within the same study to explore whether cortisol administration impacts them in similar ways.

Additionally, this retrieval impairment upon increased cortisol levels appears to be more pronounced when the material (i.e. list of words) is emotional compared to neutral (studies on cortisol administration: Antypa et al., 2022, de Quervain et al., 2007; Kuhlmann, Kirschbaum, et al., 2005, but see Buss et al., 2004; studies on acute stress induction: Buchanan et al., 2006; Kuhlmann, Piel, et al., 2005; Smeets, 2011a; Smeets et al., 2008). This retrieval impairment has been attributed to the time-dependent neuro-modulatory response exerted by cortisol (de Quervain et al., 1998, Roozendaal et al., 2003), in conjunction with the activation of the noradrenergic system (Roozendaal, 2002, Roozendaal et al., 2004).

At the neural level, brain regions like the hippocampus, amygdala and prefrontal cortex (PFC) are particularly rich in cortisol receptors (Groeneweg et al., 2011, Reul and de Kloet, 1985). Altogether, these brain regions also support encoding, consolidation and retrieval processes (Cahill et al., 1995, Eichenbaum et al., 2007, Rubin et al., 2017), making cortisol level elevation influential in modulating episodic memory. Noteworthy, the amygdala exerts influences upon the hippocampus and the PFC, eliciting significant modulatory effects for emotionally laden content (Arnsten, 2009, McGaugh, 2004). Recent findings suggest that increased cortisol levels heighten neural excitability in the hippocampus and the amygdala (de Kloet et al., 1993, Groeneweg et al., 2011, Karst et al., 2005, Karst et al., 2010), enhancing consolidation (Cahill et al., 2003, Goldfarb et al., 2019, Roozendaal, 2002, van Marle et al., 2013), but impairing retrieval (de Quervain et al., 2000, Goldfarb et al., 2019, Roozendaal, 2002).

On this background, for this study, we aimed to examine the effects of cortisol administration on the retrieval of ecologically-relevant material. Participants learned neutral and emotional texts and twenty-four hours after, were orally administered 10 mg of hydrocortisone. Around 100 min after administration (in order to reach peak salivary cortisol levels, based on previous studies of 10 mg cortisol administration), participants engaged in a free recall, recognition and temporal sequential order tasks.

In light of the existing body of research demonstrating memory retrieval impairment after cortisol administration, we expected to see lower memory performance after cortisol administration compared to placebo administration in the free recall, recognition and sequential order tasks. Moreover, for emotional material, we hypothesized that memory retrieval of the emotional text will be more impaired than the retrieval of neutral text following cortisol administration compared to placebo administration.