رابطه معکوس بین کورتیزول و خشم و ارتباط آنها با حافظه آشکار و عملکرد

Cortisol has been found to increase in response to social evaluative threat. However, little is known about the cortisol response to induced anger. Thus, in the present study, we investigated the cortisol response to anger induction and its effects on performance and explicit memory. A variant of the Montreal Stress Imaging Task (MIST; Dedovic et al., 2005) was used to induce anger in 17 male and 17 female students. Consistent with previous observations, a significant decrease in cortisol was found from pre to post manipulation which was inversely related to increases in subjective anger. Moreover, whereas anger increase was related to impairments in performance, cortisol reduction was inversely related to cognitive performance and explicit memory (recall and recognition of persons’ features in a social memory task). The adaptive value of an increase in cortisol in response to fear or uncontrollability and of a decrease in cortisol in response to anger will be discussed.

Stress occurs in a situation in which a person needs to adapt to an increased situational demand so as to cope appropriately. In the case of a social evaluative threat, the hypothalamic-pituitary-adrenal (HPA) axis is activated and stress hormones, including cortisol, are released to supply the organism with additional energy to respond to the stressor (Dickerson and Kemeny, 2004, Sapolsky, 1992 and Sapolsky, 1996). While social evaluation seems to be primarily associated with acute cortisol release, fear of physical injury or death seems to be predominantly associated with an increase in activity of the sympathetic nervous system (SNS) (Herman et al., 2003 and Herman et al., 2005). However, not all negative emotions need necessarily be associated with an increase in activity of the stress systems. In fact, there is recent evidence that anger is related to reductions in the cortisol response rather than to increases in it (Herrero et al., 2010 and Matheson and Anisman, 2009; but see Moons et al., 2010). One possible reason for divergent effects of threat vs. anger on the cortisol response is that whereas threat elicits higher perceptions of risk and feelings of lack of personal control, anger may elicit optimistic perceptions of risk (Lerner and Keltner, 2001). Specifically, anger may trigger cognitive appraisals of certainty and control, which in turn lead to lower risk estimates. According to Ortony et al. (1989), anger is based on a cognitive interpretation that an undesirable event has occurred whose cause is considered controllable and external. Results of EEG studies (Harmon-Jones, 2003 and Harmon-Jones, 2007; see Harmon-Jones et al., 2010, for a recent review) showing that anger activates frontal areas of the left cerebral hemisphere (a pattern of Alpha wave reduction in the EEG), which are associated with approach but not with avoidance behavior (Harmon-Jones and Allen, 1998), are compatible with this view. Likewise, Carver (2004) and Carver and Harmon-Jones (2009) consider anger to be a special type of negative emotion associated with the behavioral approach system (BAS), but not with the behavioral inhibition system (BIS). The long-term effects of high cortisol levels on emotion regulation and cognitive performance (including declarative memory) are negative, especially when the level of stress is too high and not only mineralocorticoid but also glucocorticoid receptors in the hippocampus are engaged (Dickerson and Kemeny, 2004 and Lupien et al., 2005). These deficits have been attributed to lower efficiency of a chronically activated hippocampus, which is involved in endocrine stress regulation as well as in spatial and episodic memory (Sapolsky, 1992). The effects of cortisol on emotion, cognition, and behavior, however, are complex, especially those related to momentary cortisol changes. For example, it has been proposed that chronically increased cortisol levels are maladaptive (Dickerson and Kemeny, 2004 and Sapolsky, 1992), given its adverse effects on cognitive performance and declarative memory (Kirschbaum et al., 1995, Kirschbaum et al., 1996 and Wolf et al., 2004). However, other research suggests that while extreme levels of cortisol (too much or too little) may be maladaptive, moderate levels may be considered adaptive (e.g., de Kloet et al., 1999). This notion reminds of the Yerkes–Dodson law of arousal-performance relationships and is also compatible with the hormetic theory of glucocorticoids on memory (Lupien et al., 2005). Specifically, the latter theory implies that increases or reductions in cortisol as a response to an acute stressor, within a given range, may be finely tuned to the requirements of the task at hand, be it in terms of physical or mental energy (i.e., effort). If this is the case, it is understandable that the short-term effects of high levels of cortisol as a result of stress may sometimes facilitate certain types of cognitive performance (de Kloet et al., 1999), such as remembering emotional information (Buchanan and Lovallo, 2001, Cornelisse et al., 2011 and Nater et al., 2007). If anger is related to cortisol reduction, it could be speculated that it may have a negative short-term effect on performance of complex cognitive tasks and declarative memory (see Nater et al., 2007). Accordingly, lower than average cortisol levels might be inversely related to cognitive performance and declarative memory and in turn, average levels of cortisol might improve this type of performance ( de Kloet et al., 1999). Although anger is considered a major stressor by some (see Spielberger and Sarason, 2005), very few studies have specifically investigated its effects on HPA axis activation and the cortisol response (e.g., Herrero et al., 2010 and Matheson and Anisman, 2009). Matheson and Anisman (2009) investigated the effects of gender discrimination and mood priming of anger or shame on the cortisol response in two experiments with women. They found that whereas shame was associated with an increase of cortisol, anger was associated with a reduction of cortisol. More recently, Herrero et al. (2010) induced anger through a mood manipulation in a group of 30 men, and investigated its effects on cardiovascular, hormonal, and asymmetric brain activity. The anger manipulation consisted of reading descriptors of anger experiences (50 self-referent statements ranging from relatively neutral to those connoting irritability, hostility, rage and anger), recalling relevant personal memories, and evoking the mood suggested by the sentence/memory. They found that anger was associated with increased cardiovascular reactivity and with a reduction of cortisol levels. In addition, they also found that anger facilitated left-hemispheric processing as assessed by a dichotic listening test, which is consistent with the results of Harmon-Jones and associates (see Harmon-Jones et al., 2010). In contrast to the results of the previous studies, Moons et al. (2010), using the Trier Social Stress Test (Kirschbaum et al., 1993), found that whereas post-stress anger predicted higher cortisol levels, post-stress fear predicted lower levels of cortisol. One objection to this study is that these authors did not directly manipulate anger or fear during the task, but only measured them retrospectively, using a post-stressor mood questionnaire. In addition, it is not clear whether the anger reported was “anger-in” (towards one-self for deficient performance) or “anger-out” (towards the judges or the experimenter). Finally, the absolute levels of anger were not reported which makes it difficult to evaluate the direct effect of anger on the cortisol response. In summary, with one exception, the previously performed studies indicate that the effects of fear and anger on the cortisol response might be diametrically opposed, with fear leading to increases in cortisol levels, and anger leading to decreases. The aim of the current research is to investigate the relationship between anger and cortisol by first hypothesizing that anger is associated with reductions in the cortisol response. To induce anger, we used a modified version of the Montreal Imaging Stress Task (MIST; see Dedovic et al., 2005), which is described in Section 1. The MIST is normally applied to induce helplessness and evaluation anxiety, but it has the potential of inducing anger through bogus performance feedback. We also wanted to explore the effects of anger and its corresponding cortisol effects on cognitive performance and declarative memory (recall and recognition), hypothesizing that it would be associated with lower performance.

The effect of anger, a negative emotion, on the HPA-axis and cortisol regulation seems to run in the opposite direction to inductions of uncontrollability, helplessness, or negative social evaluation. Whereas manipulations of uncontrollability and negative social evaluation are associated with increments in cortisol, anger is associated with decrements in the cortisol level. Furthermore, anger appears to impair cognitive performance and declarative memory whereas moderate levels of cortisol were found to be beneficial to performance in these types of tasks. The effects of the negative emotions of uncontrollability and anger on the HPA axis thus appear not to be homogeneous and should be jointly and thoroughly investigated in future studies.