Overall, the experimental and control groups did not differ significantly at baseline in cortisol levels and because of this lack of difference any variance between groups cannot be attributed to an inequality in cortisol level that existed prior to testing.  However, no significant differences between the groups were found at any of the cortisol measurements, suggesting the participants in the experimental group did not respond to the stressor with an increase in cortisol or that any increases in cortisol could be attributed to the stressor because the control and experimental groups did not differ significantly.  There was a significant difference across the different cortisol measurements for subjects.  The observed difference means that cortisol levels decreased significantly as a function of time.  In examining where the differences occur, the significant differences occurred between the baseline and measurements 5 and 4 as well as between measurements 4 and 3.  Differences between cortisol levels at various points in the manipulation further demonstrate how cortisol decreased over time and more so at the end of the manipulation.  Because there was no interaction between group and cortisol level, this decrease cannot be attributed to one group or the other. These results do not support the hypotheses set forth in this experiment.
         Over time object-location memory test scores for trial 1 and trial 2 significantly increased for both the control and experimental groups.  An increase from trial 1 to trial 2 in scores is expected because with practice participants should become more proficient at completing the task.  There were no significant differences between the control and experimental group, demonstrating further that the stressor had no effect on memory performance through increasing cortisol.  Although no differences between groups were significant for test scores on either trial, the experimental group, in examining means for test scores, had lower scores than the control group for the first trial while the experimental group had marginally higher means for the second trial.  Examination of individual cases were performed to explore why differences in score were observed and also because of the small sample size.  The small sample size imposes the disadvantage of not being able to generalize. Exploratory analyses of relationships between variables revealed that changes in cortisol level were related to certain object-location memory scores for the experimental group and the control group.
        For the control group the number of right hits from trial 1 of the object-location memory test was significantly correlated with the change in cortisol from baseline to cortisol 2 as well as with cortisol 4 to baseline.  From the examination in differences of cortisol trial by trial, it is know that cortisol decreases over time.  Therefore, because the correlation between right hits and various changes was negative, decreases in cortisol can be associated with increases in right hits.  In the experimental group the number of wrong hits from trial 1 and 2 of the memory task was negatively correlated with changes from baseline and cortisol 4.  Wrong hits in the first trial was negatively correlated as well with changes overall in cortisol and changes from cortisol 4 to 3.  Therefore, as cortisol decreased, the number of wrong hits increased.
         Examination of individual cases found that three subjects whose cortisol level was moderate had at least one object-location memory score that was above the mean.  In the case of subject #8, decreases in cortisol level were observed, however, cortisol levels remained in a moderate range.  As the cortisol level decreased, object-location memory scores fell below the mean.  Three individual cases in the control group had increases in cortisol level following the object-location memory test, while only one experimental subject showed the same trend.  Rather, four out of the six experimental subjects showed a decrease in cortisol after performing the object-location memory task.  Perception of the object-location memory test seems to be altered for the experimental subjects after receiving the stressor.  The experimental subjects are not as aroused by the object-location memory task as some of the control subjects.  For the control participants, the object-location memory task may have induced stress since this is a performance type task.
         A possible explanation for this observed dissociation and for participants’ failure to produce high cortisol under the stressor is the distinction between the sympathetic-adrenal medullary system (SAM) and hypothalamic-pituitary-adreno cortical axis (HPA) activation.  The SAM system is activated during effortful coping and is accompanied by an increase in heart rate, blood pressure, and secretion of epinephrine and norepinephrine (Cohen, Kessler, & Gordon, 1995; Peters et al., 1998).  Contrarily, the HPA system is associated with perceived uncontrollability, inability to cope and helplessness, and the secretion of cortisol.  Therefore, effort without distress would activate the SAM system, while distress without effort would activate the HPA system.  Based on qualitative, self-report data collected in this study, all participants that received the stressor reported that they put effort into performing the arithmetic task.  One plausible explanation for lack of cortisol activation in response to the stressor is that the SAM system was activated rather than the HPA system (Cohen et al., 1995).
        Another possible reason for the increase in cortisol for three of the control subjects and a decrease for four of the experimental subjects after the first trial of the object-location memory task could be stress inoculation.  In a clinical setting (Meichenbaum, 1977), stress inoculation is described as dealing with major stressors and gaining resistance so when a smaller stressor comes along, it is not as stressful.  Similarly, stress inoculation seems to be the reason for the difference between the control and experimental group in cortisol secretion.  The stressor desensitizes the experimental group because it is very stressful.  Performing the object-location memory task is viewed as relaxing because it is less stressful than the stressor, and in turn cortisol levels drop off.  This decrease in cortisol level could explain why the experimental groups’ means for object-location memory on the first trial were lower than the control groups’ mean.  Having the test preceded by a stressor alters perception of the object-location memory task.  The control group perceives the object-location memory task as a stressor and henceforth, cortisol levels increase.  This introduces a confound to the study.  The object-location memory task is acting as a stressor for the control group.  The control group is supposed to be “stress-free” in the sense that they do not receive the stressor, but the object-location memory test is acting as a stressor.
        Unstructured interviews of participants after testing revealed strategies participants used to diminish the amount of stress experienced in relation to the stressor.  The majority of experimental participants reported pausing when they felt overwhelmed while performing the visual PASAT.  Participants reported that this was not “giving-up,” but a means of reducing stress and allowing themselves to “catch-up.”  They also reported that the stressor was moderately stress inducing.  Given that participants found the stressor moderately stressful, but felt they had the ability to control the stressor and cope appropriately suggests SAM system activation.
        In research involving aversive emotional memory, the amygdala  is accepted as being important in the acquisition of this type of memory (for a review see Davis, 1992).  One view that has been offered is that neuroendocrine systems promote memory formation by adding “neural” significance to the event through particular hormones and regulation of memory storage so that the strength of the memory is susceptible to the importance of the experience  (Gold, 1995).  Memory improving events seem to have an inverted-U curve (White, 1995).  Certain events at an optimal level helps potentiate memory while those above or below the optimal level fail to enhance memory, which could explain object-location memory performance.  With extremely high or low levels of emotional arousal, the potentiation of memory is dampened (White, 1995).  The same type of inverted-U relationship is found to be related to levels of cortisol, catecholamines, and glucose (McEwen & Sapolosky, 1995).
        For the experimental group that had decreases in cortisol may have had decreases in catecholamines as well.  After the stressor that may have created SAM activation, the object-location memory, through stress-inoculation may have acted as a “destressor.”  Sympathetic arousal would drop during the encoding period for the object-location memory task, leading to less salient emotional coding.  For the control group, the object-location memory test may be viewed as a stressor and increases emotional arousal.  Sympathetic arousal through catecholamines would mediate more in-depth processing of the stimuli.  The results would support that object-location memory is hippocampal dependent even though cortisol levels were did not significantly effect performance on the memory test.  However, for the control group, because the stressor is emotionally arousing, mediation by the amygdala also occurs.
        These results do not suggest that object-location memory is necessarily limited to being declarative in nature.  As mentioned previously, attentional processes such as vigilance may have been implicated as important in experimental group performance.  Attentional processes are more important for short-term memory (STM) and working memory.  However, the involvement of STM in the object-location memory task is unlikely.  On average, participants were able to recognize ten or more objects that had exchanged position.  Ten exceeds the typical rule of 7+/-2 for STM capacity.  There could be a working memory component to object-location memory task as spatial memory is multicomponential (Postma & De Haan, 1996).  Further research needs to be done to examine the exact nature of object-location memory.
        In demonstrating object-location memory as declarative memory, a test of that has been shown to be declarative memory based such as the Weschler Memory Scale (WMS) Logical component should be compared with the object-location memory test.  Comparisons of performance between a group receiving the WMS and a group receiving the object-location memory experiment could demonstrate if the two are related.  Measuring performance at various delays could also address what type of memory is involved.  If after a couple of days, without seeing the initial array of objects, one could be fairly confident in assuming object-location memory has a declarative, long-term component.
        In summary, this study demonstrates that individual differences in stress appraisal, coping strategies, and biological reactivity to stressors exists in female college students and these variables may affect subsequent performance on other tasks such as those involving memory.  The possibility of arousal of not only the HPA system, but the SAM system in stressful situations is demonstrated.  Furthermore, the effects of stress and memory extend beyond the reactivity to cortisol secretion.  Stress can still impair memory without increases in cortisol.  Some of the individual results suggest that cortisol in low levels also has detrimental effects on memory.  The results were not significant and would need to be explored in greater depth to be conclusive.
 
 

click here to return to main page

(to References)