| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Studies |
Douglas Hospital Research Center, Department of Psychiatry, McGill University (S.J.L., S.S., N.P.V.N., M.J.M.), Montreal, Canada H4H 1R3; the Théophile-Alajouanine Laboratory, Research Center of the Hospital Côte-des-Neiges (S.J.L., S.G., B.M.T., F.M.), University of Montreal, Montreal, Canada H3W 1W5; the Laboratory of Neuroendocrinology, Rockefeller University (B.S.M.), New York, New York 10021; and the Department of Psychiatry, Veterans Administration Medical Center, University of California-San Diego (R.L.H.), La Jolla, California 92093-0603
Address all correspondence and requests for reprints to: Sonia Lupien, Ph.D., Douglas Hospital, 6875 boulevard Lasalle, Montreal, Quebec, Canada H4H 1R3. E-mail: Lupiens{at}ere.umontreal.ca
 |
Abstract |
|---|
 Top Abstract IntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
In the rat, sustained exposure to elevated glucocorticoid levels in later life is associated with an increased loss of hippocampal neurons, accompanied by severe memory impairments (5). Elevated glucocorticoid levels in this species appear to compromise both the survival (6) as well as the function (7, 8) of hippocampal neurons. A comparable relationship between cortisol levels and cognition is apparent in humans (3, 4). Among a group of healthy elderly human subjects, elevated basal cortisol levels have been shown to be associated with impaired declarative memory performance, a form of memory that appears to be mediated by the hippocampal formation. Declarative memory refers to the conscious or voluntary recollection of learned information, whereas nondeclarative memory refers to memory abilities that are retrieved without conscious or explicit access (e.g. priming) (9, 10). The observation of impaired declarative memory performance in aged human subjects showing increased cortisol levels suggests a close functional relationship between high levels of corticosteroids and hippocampal function.
Although studies in animals and humans have reported detrimental effects of chronic endogenous or stress-induced elevations of glucocorticoids on memory function, other studies performed both in animals (11, 12, 13) and humans (14, 15, 16) have shown that the acute administration of synthetic glucocorticoids also impairs declarative memory function, leaving nondeclarative memory unimpaired. These results suggest a relationship between the acute stress-induced increase in glucocorticoid levels and hippocampally relevant memory function.
The observation of acute and chronic effects of increased cortisol levels on cognitive function in humans leads to the question of whether the stress-induced increase in cortisol levels in elderly subjects may be related to the memory impairments frequently reported in these individuals. Indeed, aging has long been viewed in the geronto-geriatric literature as being either accelerated by stress factors or reflecting a decreased adaptation to stress (17, 18). To test whether stress responsiveness may play a role in memory impairments in the elderly individual, we examined cortisol reactivity and declarative/nondeclarative memory performance in response to a psychological stressor in elderly subjects.
|
Experimental Subjects |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
Stress-induced effects of cortisol elevation on declarative and nondeclarative memory function was studied in comparison to that during a nonstressful condition, and memory was measured before and after each condition (nonstressful vs. stressful). The nonstressful condition consisted of a computer-generated detection task for which the subjects had to search for a predetermined target. Subjects were first familiarized with the computer to avoid computer-based biases in performance (19). During the nonstressful condition, subjects were told that this task was for practice only and that they could adjust their performance at their own pace. This task lasted 10 min.
The stressful condition consisted of a public speaking task, a test that has proven to be a provocative laboratory stressor in young normal controls (20, 21). In the testing room, subjects were introduced to an individual who was presented as being an expert in neurolinguistics and who would evaluate their discourse. This colleague was, in fact, a confederate instructed to be polite but cold, not showing any emotions toward the subjects. Once in the room, subjects were instructed by the second experimenter to prepare a 5-min speech on a given topic that they were then to present in front of a video camera. They were also told that their discourse would be evaluated by five experts on neurolinguistics who were standing behind the false mirror placed behind the video camera. They were finally told that the filming of their performance would go on for the entire 5 min of the speech, even if they stopped talking after a few minutes. The stressful condition lasted 10 min.
Memory tests
Before each condition, subjects were presented with a declarative/nondeclarative memory task consisting of an acquisition phase of 12 pairs of words and a declarative and nondeclarative recall phase of these words. This test has been described in length in other reports (3, 4) and will only be summarized here. Two lists of word-pairs were constructed for the stressful and nonstressful conditions using the same parameters described below. Each list of words presented to the subjects was comprised of six moderately related word pairs (related pairs) and six unrelated word pairs (unrelated pairs). The unrelated pairs were constructed so as to be matched to the related pairs in terms of word frequency, word length, and grammatical category. Presentation of material and response recording were monitored by a Macintosh computer. For the declarative memory task, the subjects were presented with the list of word pairs that they had to read aloud. The list was presented twice in a different order. After each presentation, the subjects made a cued recall, where they had to recall a member of a pair when presented with the other. Recall performance on the second word pairs presentation was used as the precondition memory score and compared to the postcondition memory performance. After the cued recall test, subjects were presented with a distractory task of word fragment completion (e.g. R - - - E). The goal of this distractory task was to mask the connection between the declarative and nondeclarative conditions and, at the same time, to create an interference to prevent recency effects on the completion task to follow. Nondeclarative memory was thereafter tested by a word completion task. Subjects were instructed to complete each presented syllable as fast as possible and with the first word that came to mind. The subjects were told that this was a completely different test. No restriction was imposed as to the category of word that could be given as a completion. After being presented with an example, the subjects had to complete 72 such syllables. Among these, 24 corresponded to the first syllable of each word of the pairs learned previously, whereas 24 others corresponded to the first syllable of words from a second list that was given for declarative and nondeclarative memory to another group of 19 elderly subjects to have a baseline for nondeclarative memory. The 24 last syllables were randomly chosen. Thus, only one third of the cues corresponded to the material presented in the studied list. This was again done to prevent subjects from making a connection between the presented syllables and the words that were presented previously. The dependent measure for the declarative memory task was the number of words correctly recalled, whereas for the nondeclarative task, it corresponded to the number of words generated with prior exposure minus the number of words generated without prior exposure (baseline).
Cortisol measurement
We chose to monitor cortisol levels using saliva samplings to prevent the stress-inducing effects of blood sampling (22). Moreover, it has been shown that determination of cortisol in saliva provides a reliable measure of the free unbound fraction of cortisol, and correlation coefficients on the order of r = 0.96 between cortisol levels in saliva and serum have been obtained in the elderly population (23).
Filters (3.5 x 5 cm) were cut from Whatman no. 42 filter paper (Whatman, Clifton, NJ) for the collection of saliva. The top centimeter of 5-cm length was used for recording subject data and was demarcated with a line. The subjects were asked to place the filter paper in their mouth until the saliva front reached just beyond the 4 cm line. The filter was then air-dried and stored at -4 C. In previous work with this procedure, we have established that protein content, measured in 0.1-N NaOH extracts, in salivary samples collected in this manner varies by an average of less than 1% across a wide range of samples.
Cortisol was extracted from the filter in 2 mL ethanol for 1 h at room temperature. A 300-µL aliquot of the extract was assayed using [3H]cortisol as radiotracer and a highly specific cortisol antibody (B-63 antibody from Endocrine Sciences, Tarzana, CA). This antibody cross-reacts less than 4% with deoxycorticosterone or deoxycortisol, and less than 0.5% with any other adrenal steroid. Intra- and interassay variabilities are 3.5% and 5%, respectively.
Procedure
All subjects were tested at 1330 h to control possible
differential effects of the circadian pattern of cortisol on cognition.
The experimental session lasted for 140 min and was approved by the
ethics committee of the research center where the study was performed.
Before starting the protocol, subjects gave their informed consent to
the "speech evaluation" that would be performed in about 90 min.
The subjects were not aware of the goal of this speech evaluation until
the end of the entire testing session, at which time they were
debriefed as to the nature of this test. Saliva samples were taken
before and after memory testing as well as before and after the
nonstressful and stressful conditions. Figure 1
provides
an overview of the experimental design.
|
After a resting period of 15 min, a sixth saliva sample was taken, a second memory list was presented to the subjects, and acquisition and declarative/nondeclarative recall of these new words were performed. This was followed by a seventh saliva sample and the stressful condition. After the stressful condition, subjects were returned to the neuropsychological testing room, where an eighth saliva sample was taken, and subjects were tested for declarative and nondeclarative recall of the words learned before the stressful condition. The two memory lists were counterbalanced across subjects, but the two conditions were not. This was to prevent the interfering effect of the stressful condition on the memory performance measured before and after the nonstressful condition. Two final saliva samples were taken before the end of the experiment, at which point the subjects were debriefed as to the nature of the stress condition.
Statistics
First, the effects of the nonstressful and stressful conditions on declarative and nondeclarative memory performance were assessed separately for each memory process using an ANOVA with condition (nonstressful vs. stressful) and period (pre- vs. posttreatment condition) as the within-subject factors. The time course of cortisol secretion across the experiment was assessed using an ANOVA, with cortisol levels (samples 1–10) as the only within-subject factor. Second, and as the existence of individual differences in coping with stressful situations does not predict a uniform endocrine response pattern to a given stressor, we separated, on a post-hoc basis, the subjects between those who responded to the acute stressful condition with an increase in cortisol levels (increase in cortisol level from samples 7 to 8; responder group) and those who did not respond to the acute stressful situation (no change or a decrease in cortisol levels; nonresponder group). This method of separating subjects on the basis of their acute cortisol stress response after a public speaking task as been described by Fehm-Wolfsdorf and collaborators (24).
|
Results |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
, A and B, presents the pre- and
postcondition performances for declarative and nondeclarative memory
before and after the nonstressful and stressful conditions in the
entire group of subjects. The ANOVA performed on declarative memory
changes revealed that the stressful condition induced a significant
decrease in declarative memory performance, whereas the nonstressful
condition did not [F(1, 12) = 6.11; P < 0.03].
Although performance on the declarative memory test tended to increase
on the second occasion (practice effect), this effect did not reach
significance level (P > 0.06). The ANOVA performed on
nondeclarative memory performance revealed a main effect of period,
showing that for both the nonstressful and stressful conditions,
performance increased at the second occurrence of the test (possibly
due to a priming mechanism). However, performance before and after each
condition was not different [F(1, 12) = 0.01; P =
NS].
|
presents the cortisol levels obtained at each
time point of the experiment. There were no effects of sample on the
time course of cortisol secretion (P > 0.05). However,
note the important variability of the cortisol level, a finding
consistent with many studies of cortisol responses to mild forms of
stress in humans (25). We thus separated the subjects with regard to
their stress-induced cortisol response following the criterion
described earlier. We found that 5 of 14 subjects (35.7% of the
population) were responders, and the remaining 9 (64.3% of the
population) were nonresponders. The groups did not differ in terms of
age, ratio of men/women, or education level (P >
0.05). Cortisol levels in both groups were not different before the
stressful condition [t(1, 12) = 0.4; P >
0.1; responders, 13.4 \ 1.3; nonresponders, 11.4 \ 2.2],
whereas there was a significant difference between groups after the
stressful condition [t(1, 12) = 7.7; P <
0.02; responders, 19.5 \ 3 (increase of 6.1 nmol/L);
nonresponders, 7.6 \ 2.6; (decrease of 3.8 nmol/L)]. Figure 4 presents the time course of cortisol secretion in
these 2 groups of individuals. As shown, a very different pattern of
cortisol secretion occurred in these 2 groups. Although the only
significant group or time difference with regard to cortisol secretion
was between samples 5 and 6 for the nonresponders (P <
0.05) and between samples 7 and 8 for the responders (P
< 0.05), the responders started to secrete high levels of cortisol
much earlier than the nonresponders.
|
|
presents group differences between the
responders and the nonresponders with regard to declarative memory
performance at each time point of the experiment. As no significant
difference was ever observed for nondeclarative performance for total
or separated grouping, we only discuss declarative memory performance
results in the next section. The responders did not differ from the
nonresponders for declarative memory performance before and after the
nonstressful condition, whereas they presented a significantly lower
declarative memory performance before [t(1, 12) = 11.9;
P < 0.005] and after [t(1, 12) = 16.1;
P < 0.002] the stressful situation.
|
|
Discussion |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
However, in our population of elderly subjects, cortisol levels did not significantly increase immediately after the stressful condition. A sharp increase in cortisol levels was observed 25 min before the actual stressor (sample 6), suggesting that the anticipation of the stress, rather than the actual stressor per se, may have played a more significant role in the stress-induced declarative memory deficits observed in this population. If this is the case, then a chronic exposure to high cortisol levels may be more predictive of declarative memory deficits than an acute stress-induced rise in cortisol levels.
There was a high degree of variability in the cortisol levels of these individuals over the entire experiment as well as in their responses to the stressor, suggesting large interindividual differences in the pattern of anticipation and response to the stressor. For this reason, we separated the subjects into responders and nonresponders on the basis of their cortisol responses to the stressor (24). The a posteriori analyses of cortisol secretion and declarative memory performance in the responders and nonresponders revealed very different patterns of cortisol secretion and declarative memory deficits in the two populations. We showed that the responders began to secrete high cortisol levels at the beginning of the protocol, immediately after the first neuropsychological evaluation (and, thus, 62 min before the actual stressor), whereas the nonresponders had low cortisol levels until sample 6, 25 min before the actual stressor. This suggests that the anticipation of the incoming stressor, rather than the actual stressor per se, may have played a more significant role in cortisol response in these subjects, with the responders starting to worry about the stressful experience much earlier than the nonresponders.
However, studies have shown that mental tasks themselves can induce a significant increase in cortisol levels in young subjects (27, 28). It is thus possible that the early rise in cortisol levels observed in the responder group be due to the mental effort introduced by the first memory task. The analysis of declarative memory function in the responder and nonresponder groups before and after the nonstressful condition showed that although it did not reach significance level, the responders tended to start the experiment with lower declarative memory performance than the nonresponders. Stress-induced increases in cortisol levels due to higher mental task demands may thus explain the early rise in cortisol levels observed in the responder group. In our study, we did not counterbalance the treatment order, as we postulated that stress-induced memory deficits after the stressful condition could have delayed detrimental effects on memory performance measured subsequently during the nonstressful condition. However, the possibility that cortisol levels may increase after the mental effort induced during a nonstressful condition requires a better control of treatment order in subsequent studies. We still suggest that a nonstressful (baseline) condition is necessary in stress studies in humans, but we now assume that counterbalancing stressful and nonstressful conditions in further studies could yield valuable data about the unique and combined influence of acute stress and mental effort on cortisol secretion.
Although the two groups of elderly subjects did not differ in their declarative memory performance before and after the nonstressful condition, they significantly differed in their declarative memory performance before and after the stressful condition. The observation of a significantly lower declarative memory performance before the stressor in the responder group suggests that the anticipation-induced increase in cortisol levels in this group may have affected acquisition of information and prevented the subjects from using previously learned rules to acquire information about the second memory list. This result is strengthened by the observation that at the start of the experiment, when cortisol levels were still similar in both groups, declarative memory performance was not different.
There is a large number of studies that have documented the provocative nature of anticipation on cortisol levels (for a review, see 29 . For example, admission to a hospital has been noted to be generally very provocative of cortisol (29), and other studies have shown that individuals anticipating exhausting exercise show significant rises in cortisol, comparable to those occurring during the actual physical task (29). From these studies, it is apparent that individuals show a significant hypothalamo-pituitary-adrenal activation when they are exposed to important changes in their environment. The results of this study demonstrate that elderly subjects who present an early increase in cortisol levels in anticipation of a stressor show poor declarative memory capacities before and after being confronted with a stress, whereas elderly individuals who show no change in cortisol levels in anticipation of a stressor do not present stress-induced declarative memory deficits.
Although these results are particularly interesting with regard to study of the genesis of memory deficits in elderly populations, it has to be noted that they are preliminary and were not induced by a prospective study. For this reason, additional studies measuring the anticipation-induced vs. stress-induced increase in cortisol levels in elderly subjects and their isolated and/or combined influence on declarative/nondeclarative memory function are now being conducted in our laboratory and should yield valuable data on the cause and consequence of the effects of stress reactions on memory function in aged individuals.
|
Footnotes |
|---|
2 Supported by a postdoctoral fellowship from the Fonds de la
Recherche en Santé du Québec (S.J.L.).
3 Recipient of a Medical Research Council of Canada Career Scientist
award.
Received May 30, 1996.
Revised October 28, 1996.
Revised April 1, 1997.
Accepted April 15, 1997.
|
References |
|---|
|
TopAbstractIntroductionExperimental SubjectsMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. M. Lai, K. A. Hawkins, C. P. Gross, and J. H. Karlawish Self-Reported Distress After Cognitive Testing in Patients With Alzheimer's Disease J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2008; 63(8): 855 - 859. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. K. Lee, T. A. Glass, M. J. McAtee, G. S. Wand, K. Bandeen-Roche, K. I. Bolla, and B. S. Schwartz Associations of Salivary Cortisol With Cognitive Function in the Baltimore Memory Study Arch Gen Psychiatry, July 1, 2007; 64(7): 810 - 818. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Duncko, B. Cornwell, L. Cui, K. R. Merikangas, and C. Grillon Acute exposure to stress improves performance in trace eyeblink conditioning and spatial learning tasks in healthy men Learn. Mem., May 1, 2007; 14(5): 329 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Ling, C. Campbell, T. M. Heffernan, and C. G. Greenough Short-term Prospective Memory Deficits in Chronic Back Pain Patients Psychosom Med, February 1, 2007; 69(2): 144 - 148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M I Peralta-Ramirez, M A Coin-Mejias, J Jimenez-Alonso, N Ortego-Centeno, J L Callejas-Rubio, A Caracuel-Romero, and M Perez-Garcia Stress as a predictor of cognitive functioning in lupus Lupus, December 1, 2006; 15(12): 858 - 864. [Abstract] [PDF]
|
|
|
|
|
|
M. Zorawski, N. Q. Blanding, C. M. Kuhn, and K. S. LaBar Effects of stress and sex on acquisition and consolidation of human fear conditioning Learn. Mem., July 1, 2006; 13(4): 441 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. G. Smith, L. Betancourt, and Y. Sun Molecular Endocrinology and Physiology of the Aging Central Nervous System Endocr. Rev., April 1, 2005; 26(2): 203 - 250. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kuhlmann, M. Piel, and O. T. Wolf Impaired Memory Retrieval after Psychosocial Stress in Healthy Young Men J. Neurosci., March 16, 2005; 25(11): 2977 - 2982. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S. Maheu, R. Joober, and S. J. Lupien Declarative Memory after Stress in Humans: Differential Involvement of the {beta}-Adrenergic and Corticosteroid Systems J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1697 - 1704. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Wilson, C. F. Mendes de Leon, J. L. Bienias, D. A. Evans, and D. A. Bennett Personality and Mortality in Old Age J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2004; 59(3): P110 - P116. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R S Wilson, D A Fleischman, R A Myers, D A Bennett, J L Bienias, D W Gilley, and D A Evans Premorbid proneness to distress and episodic memory impairment in Alzheimer's disease J. Neurol. Neurosurg. Psychiatry, February 1, 2004; 75(2): 191 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. F.C. van Rossum and S. W.J. Lamberts Polymorphisms in the Glucocorticoid Receptor Gene and Their Associations with Metabolic Parameters and Body Composition Recent Prog. Horm. Res., January 1, 2004; 59(1): 333 - 357. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. S. Wilson, D. A. Evans, J. L. Bienias, C. F. Mendes de Leon, J. A. Schneider, and D. A. Bennett Proneness to psychological distress is associated with risk of Alzheimer's disease Neurology, December 9, 2003; 61(11): 1479 - 1485. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Woodson, D. Macintosh, M. Fleshner, and D. M. Diamond Emotion-Induced Amnesia in Rats: Working Memory-Specific Impairment, Corticosterone-Memory Correlation, and Fear Versus Arousal Effects on Memory Learn. Mem., September 1, 2003; 10(5): 326 - 336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Lucassen, M. B. Muller, F. Holsboer, J. Bauer, A. Holtrop, J. Wouda, W. J. G. Hoogendijk, E. R. De Kloet, and D. F. Swaab Hippocampal Apoptosis in Major Depression Is a Minor Event and Absent from Subareas at Risk for Glucocorticoid Overexposure Am. J. Pathol., February 1, 2001; 158(2): 453 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Lyons, J. M. Lopez, C. Yang, and A. F. Schatzberg Stress-Level Cortisol Treatment Impairs Inhibitory Control of Behavior in Monkeys J. Neurosci., October 15, 2000; 20(20): 7816 - 7821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Cho, A. Ennaceur, J. C. Cole, and C. K. Suh Chronic Jet Lag Produces Cognitive Deficits J. Neurosci., March 15, 2000; 20(6): RC66 - RC66. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Deuster, J. S. Petrides, A. Singh, G. P. Chrousos, and M. Poth Endocrine Response to High-Intensity Exercise: Dose-Dependent Effects of Dexamethasone J. Clin. Endocrinol. Metab., March 1, 2000; 85(3): 1066 - 1073. [Abstract] [Full Text]
|
|
|
|
|
|
S. Kalmijn, L. J. Launer, R. P. Stolk, F. H. de Jong, H. A. P. Pols, A. Hofman, M. M. B. Breteler, and S. W. J. Lamberts A Prospective Study on Cortisol, Dehydroepiandrosterone Sulfate, and Cognitive Function in the Elderly J. Clin. Endocrinol. Metab., October 1, 1998; 83(10): 3487 - 3492. [Abstract] [Full Text]
|
|
|
|
 |
|
 Stress-Induced Cortisol Secretion Affects Memory in the Elderly Journal Watch Psychiatry, November 1, 1997; 1997(1101): 2 - 2. [Full Text]
|
|
|
|
|
|
M. H. Mesches, M. Fleshner, K. L. Heman, G. M. Rose, and D. M. Diamond Exposing Rats to a Predator Blocks Primed Burst Potentiation in the Hippocampus In Vitro J. Neurosci., July 15, 1999; 19(14): RC18 - RC18. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
