Effects of dexamethasone on declarative memory function in posttraumatic stress disorder

Alterations in the hypothalamic-pituitary-adrenal (HPA) axis and hippocampal-based memory have been associated with posttraumatic stress disorder (PTSD), and the administration of exogenous glucocorticoids has been shown to result in a transient verbal declarative memory impairment in healthy human subjects. The purpose of this study was to assess the effects of the glucocorticoid dexamethasone on verbal declarative memory function in patients with PTSD. Forty-two men and women with (n=14) and without (n=28) PTSD received placebo or dexamethasone (1 and 2 mg on two successive days) in a double-blind, randomized fashion. Declarative memory was assessed with paragraph recall at baseline (day 1) and day 3. There was a significant interaction between diagnosis and drug (dexamethasone vs. placebo) on paragraph recall related to a relative detrimental effect of dexamethasone on memory function in healthy subjects, but not those with PTSD. These findings are consistent with an altered sensitivity of declarative memory function in PTSD to regulation by glucocorticoids, possibly explainable by alterations in glucocorticoid receptors in the hippocampus or other brain regions mediating declarative memory.     

Neuropsychological impairment in bipolar disorder: the relationship with glucocorticoid receptor function

OBJECTIVE: Basal levels of glucocorticoids, such as cortisol, are generally unaltered in bipolar disorder. However, neuroendocrine tests of glucocorticoid receptor (GR) function such as the dexamethasone suppression test (DST) are frequently abnormal. Neuropsychological impairment is well documented in healthy volunteers after administration of glucocorticoids and in patients with bipolar affective disorder. This suggests a potential link between neuropsychological and hypothalamic-pituitary-adrenal axis function. We examined the hypothesis that neuropsychological impairment in bipolar disorder is associated with abnormal GR function. METHODS: Seventeen euthymic bipolar patients and 16 controls completed tests of verbal declarative and working memory (WM) tests and the DST. The correlation between neuroendocrine and neuropsychological function was examined. RESULTS: Bipolar patients made significantly more errors of omission and commission on the WM paradigm and demonstrated impaired verbal recognition memory. Patients' post-dexamethasone cortisol correlated with WM commission errors (r(s) = 0.64, p = 0.0006). No such relationship was evident in controls. CONCLUSION: Deficits in declarative memory and WM are evident in patients with bipolar disorder. The deficit in retrieval accuracy from WM appears to be correlated with abnormal GR function.     

Dexamethasone blocks sleep induced improvement of declarative memory

To investigate the role of glucocorticoids for effects of early and late nocturnal sleep on declarative and procedural memory, 2 mg dexamethasone (versus placebo) were administered to healthy men 7 h prior to retention sleep. The retention sleep interval covered either the early or late half of nocturnal sleep. Following placebo, recall of a paired associate list (declarative memory) benefitted more from early than late sleep and recall of mirror tracing skills (procedural memory) benefitted more from late than early sleep. Dexamethasone did not affect slow wave sleep dominating early sleep, but blocked the beneficial effect of early sleep on recall of paired associates. Conversely, dexamethasone reduced rapid eye movement sleep dominating late sleep, but did not affect late sleeps beneficial effect on mirror tracing skills. The natural inhibition of endogenous glucocorticoid secretion during early sleep seems to be essential for a sleep-related facilitation of declarative memory.     

Decreased memory performance in healthy humans induced by stress-level cortisol treatment.

BACKGROUND: Glucocorticoids (GCs) can regulate hippocampal metabolism, physiologic functions, and memory. Despite evidence of memory decreases during pharmacological GC treatment, and correlations between memory and cortisol levels in certain disease conditions, it remains unclear whether exposure to the endogenous GC cortisol at levels seen during physical and psychological stress in humans can inhibit memory performance in otherwise healthy individuals. METHODS: Randomized, double-blind, placebo-controlled comparison of 2 fixed oral doses of cortisol (40 mg/d and 160 mg/d using split doses to approximate circadian rhythm) given for 4 days to matched groups of healthy subjects (n = 51). Lower-dose treatment approximated cortisol exposure during mild stress, whereas the higher dose approximated cortisol exposure during major stress. Cognitive testing and plasma sampling were done at baseline, after 1 and 4 days of treatment, and after a 6-day washout period, hypothesizing dose-dependent decreases in verbal declarative memory. RESULTS: Cortisol treatment at the higher dose produced reversible decreases in verbal declarative memory without effects on nonverbal memory, sustained or selective attention, or executive function. A significant interaction between time and treatment condition for paragraph recall was explained by treatment-induced differences in performance after 4 treatment days, with lower immediate and delayed recall performance during higher-dose cortisol treatment compared with lower-dose treatment and placebo. CONCLUSIONS: Several days of exposure to cortisol at doses and plasma concentrations associated with physical and psychological stress in humans can-similar to pharmacological GC treatment-reversibly decrease specific elements of memory performance in otherwise healthy individuals.     

Effects of acute cortisol administration on response inhibition in patients with major depression and healthy controls

Glucocorticoids (GCs) have repeatedly been shown to impair hippocampus-mediated, declarative memory retrieval and prefrontal cortex-based working memory in healthy subjects. However, recent experimental studies indicated that patients with major depressive disorder (MDD) lack these impairing effects. These missing effects have been suggested to result from dysfunctional brain GC receptors. The purpose of the present study was to investigate whether response inhibition, an executive function relying on the integrity of the prefrontal cortex, would be impaired after cortisol administration in patients with MDD. In a placebo-controlled, double blind crossover study, 50 inpatients with MDD and 54 healthy control participants conducted an emotional go/no-go task consisting of human face stimuli (fearful, happy, and neutral) after receiving a dose of 10 mg hydrocortisone and after placebo. GC administration had an enhancing effect on inhibitory performance in healthy control participants, indicated by faster responses, while no GC effect was revealed for the patients group. Moreover, patients showed an overall worse performance than healthy participants. In conclusion, this study further supports the hypothesis of impaired central glucocorticoid receptor function in MDD patients. Regarding the importance of inhibitory functioning for daily living, further studies are needed to examine the impact of glucocorticoids on response inhibition.     

Glucocorticoid-induced impairment of declarative memory retrieval is associated with reduced blood flow in the medial temporal lobe

Previous work indicates that stress levels of circulating glucocorticoids can impair retrieval of declarative memory in human subjects. Several studies have reported that declarative memory retrieval relies on the medial temporal lobe. The present study used H(2)(15)O-positron emission tomography to investigate whether acutely elevated glucocorticoid levels affect regional cerebral blood flow in the medial temporal lobe, as well as in other brain regions, during declarative memory retrieval in healthy male human subjects. When measured over four different declarative memory retrieval tasks, a single, stress-level dose of cortisone (25 mg) administered orally 1 h before retention testing, induced a large decrease in regional cerebral blood flow in the right posterior medial temporal lobe, the left visual cortex and the cerebellum. The decrease in the right posterior medial temporal lobe was maximal in the parahippocampal gyrus, a region associated with successful verbal memory retrieval. Cortisone administration also significantly impaired cued recall of word pairs learned 24 h earlier, while drug effects on performance in the other tasks (verbal recognition, semantic generation and categorization) were not significant. The present results provide further evidence that acutely elevated glucocorticoid levels can impair declarative memory retrieval processes and suggest that such impairments may be related to a disturbance of medial temporal lobe function.     

Deficits in hippocampal and anterior cingulate functioning during verbal declarative memory encoding in midlife major depression

OBJECTIVE: Prior studies showed that subjects with major depression have deficits in hippocampal-based verbal declarative memory (e.g., recall of a paragraph) and in hippocampal and prefrontal cortical functioning and structure. The purpose of the present study was to assess hippocampal and prefrontal functioning during performance of a verbal declarative memory task in subjects with midlife major depression. METHOD: Subjects with midlife major depression (N=18) and healthy subjects (N=9) underwent positron emission tomography imaging during a control task and verbal encoding of a paragraph. RESULTS: During the verbal memory encoding task the comparison subjects, but not the subjects with depression, activated the right hippocampus and prefrontal cortex (anterior cingulate), as well as the cuneus and cerebellum. CONCLUSIONS: These results are consistent with a failure of hippocampal and anterior cingulate activation in depression, and they support the hypothesis of deficits in hippocampal and anterior cingulate functioning in depression.     

Hippocampal volume, memory, and cortisol status in major depressive disorder: effects of treatment.

BACKGROUND: Depression has been linked to stress, memory deficits, and hypercortisolemia. However, the relationships between depression, hippocampal structure and function, and cortisol levels are unclear and the effects of antidepressant treatment on the measures are not well studied. METHODS: Whole hippocampal volume, performance on verbal and visual declarative memory function and cortisol status was evaluated in 38 subjects with major depressive disorder (MDD) and 33 healthy subjects. All measures were repeated in a subgroup (n = 22) of depressed patients after successful selective serotonin reuptake inhibitor (SSRI) treatment. RESULTS: Hippocampal volume was not significantly different between patients with untreated MMD and healthy subjects, after controlling for whole brain volume, age and gender. However, depressed subjects had significantly greater deficits in delayed memory and percent retention on the verbal portion of the Wechsler Memory Scale-Revised (WMS-R) compared with healthy subjects, without significant differences in visual memory, attention, vigilance, or distractibility. Baseline plasma or urinary free cortisol (UFC) was not related to either hippocampal volume or memory deficits. Successful treatment with antidepressants did not change hippocampal volume but did result in a significant improvement in memory function and a reduction in UFC excretion. CONCLUSIONS: Medication-free nonelderly depressed outpatients without alcohol dependence or adverse experiences in childhood had normal hippocampal volume. Focal declarative memory deficits in depression supported localized hippocampal dysfunction in depressed patients. Treatment with antidepressants significantly improved memory and depression but did not alter hippocampal volume, suggesting that antidepressants may improve hippocampal function in the absence of detectable structural changes.     

Lymphocyte glucocorticoid receptor resistance and depressive symptoms severity: a preliminary report

OBJECTIVE: Assessment of the temporal interrelationship of neuropsychiatric parameters requires technologies allowing frequent biological measurements. We propose glucocorticoid receptor (GR) function of lymphocytes to assess the temporal relationship between glucocorticoid resistance and the course of major depressive disorder. METHOD: Dexamethasone suppression of lymphocyte proliferation was in vitro assessed via 5-bromo-2' deoxyuridine (BrdU) incorporation in DNA. Optimal conditions were determined using blood of healthy volunteers. Thereafter the relation between depression severity (Hamilton Depression Rating Scale, HDRS, scores), lymphocyte proliferation and morning cortisol levels in blood was studied in thirteen depressed patients, mostly with a history of treatment resistance. RESULTS: Recovery from depression was not directly associated with changes in lymphocyte glucocorticoid resistance. However, a negative correlation was observed between HDRS and BrdU incorporation and a positive correlation between morning cortisol and BrdU incorporation. No significant correlation was found between cortisol and HDRS. Regression analyses showed that HDRS was related to both suppression of BrdU incorporation (beta -0.508, p<0.001) and cortisol levels (beta 0.364, p=0.001) in a highly significant model (F2,60=14,244, p<0.001) Except for one case, such relation could not be found within patients. CONCLUSION: Our preliminary results suggest a mutual relation between lymphocyte GR function, morning cortisol levels and MDD symptom severity. A direct relation between glucocorticoids resistance and recovery may not exist, but glucocorticoid resistance might attenuate or prevent recovery. It is clear that additional studies using larger and more homogenous groups of MDD patients are required to support our findings.     

Deficits in hippocampus-mediated Pavlovian conditioning in endogenous hypercortisolism.

BACKGROUND: Elevated endogenous levels of corticosteroids cause neural dysfunction and loss, especially within the hippocampus, as well as cognitive impairment in hippocampus-mediated tasks. Because Cushing's syndrome patients suffer from hypercortisolism, they represent a unique opportunity to study the impact of elevated glucocorticoids on cognitive functions. The aim of this study was to examine the performance of Cushing's syndrome patients on trace eyeblink conditioning, a cross-species, hippocampal-mediated test of learning and memory. METHODS: Eleven Cushing's syndrome patients and 11 healthy control subjects participated in an eyeblink trace conditioning test (1000-msec trace) and a task of declarative memory for words. Salivary cortisol was collected in both the patients and the control subjects, and urinary free cortisol was collected in the patients only. RESULTS: The patients exhibited fewer conditional responses and remembered fewer words, compared with the control subjects. Cortisol levels correlated with immediate and delayed declarative memory only. Conditional response correlated with delayed recall after controlling for the magnitude of unconditional response. CONCLUSIONS: The integrity of the hippocampus seems to be compromised in Cushing's syndrome patients. Trace eyeblink conditioning might be useful both as a clinical tool to examine changes in hippocampus function in Cushing's disease patients and as a translational tool of research on the impact of chronic exposure of glucocorticoids.     

Cortisol levels predict cognitive impairment induced by electroconvulsive therapy.

BACKGROUND: Elevated glucocorticoids may increase the vulnerability of the brain to the adverse effects of repeated seizures. This study tested the hypothesis that higher ambient cortisol levels would predict increased cognitive impairment in depressed patients subsequent to receiving electroconvulsive therapy (ECT) for major depression. METHODS: Sixteen subjects provided three samples of saliva the day before receiving unilateral nondominant ECT. Measures of mood, global cognitive functioning, attention, executive function, verbal and visuospatial memory, and visuospatial processing speed were obtained 1 day before the first ECT and 1 day after the sixth ECT treatment. The relationship between basal salivary cortisol obtained before the first ECT treatment and the change score of each cognitive measure after the sixth ECT treatment was examined and tested with Pearson correlation coefficients. RESULTS: Electroconvulsive therapy treatments delivered over 2 weeks resulted in a significant improvement in mood and a decline in most measures of cognitive performance. Elevated basal cortisol was associated with a greater decline in performance of executive function, visuospatial processing speed, and verbal memory. CONCLUSIONS: Although this study is limited by the small number of subjects and the high number of comparisons, all significant correlations were consistent with the hypothesis that elevated cortisol predicts a greater degree of ECT-induced cognitive impairment.     

Stress-induced cortisol elevations are associated with impaired delayed, but not immediate recall

Glucocorticoids are known to modulate memory functions, with elevated cortisol levels being associated with impaired declarative memory. This specific effect has been shown in several studies using pharmacological doses of cortisol. The present study was designed to assess the effects of stress-induced cortisol elevations on (1) the type of memory processing (encoding, consolidation and retrieval), and (2) on the emotional valence of the material under study. Sixteen healthy females were presented neutral and emotional material (words and paragraphs) before and after a stress challenge. Declarative memory was tested immediately after presentation and 24 h later (delayed recall). Delayed, but not immediate recall of the information presented after the stress challenge was significantly reduced compared with delayed recall of information presented before the stress challenge. In line with this, strong negative correlations were found for delayed recall of words and spatial memory presented after the challenge with post-stress cortisol levels, whereas no significant correlations were found between cortisol levels and delayed recall at day 1. These results suggest that stress-induced cortisol specifically affects long-term consolidation of declarative memories. These findings may have implications for understanding the effects of traumatic stress on memory functioning in patients with stress-related psychiatric disorders.     

Effects of adjunctive mifepristone (RU-486) administration on neurocognitive function and symptoms in schizophrenia.

BACKGROUND: It has been suggested that hypercortisolemia may cause or exacerbate both neurocognitive impairment and symptoms in schizophrenia. We hypothesized that antiglucocorticoid treatments, particularly glucocorticoid receptor (GR) antagonists, would improve neurocognitive functioning and clinical symptoms in this disorder. METHOD: Twenty patients with schizophrenia were treated with 600 mg/day of the GR-antagonist mifepristone (RU-486) or placebo for 1 week in a double-blind, crossover design. Neurocognitive function was evaluated at baseline and 2 weeks after each treatment. Neuroendocrine profiling was performed at these times and also immediately after each treatment. Symptoms were evaluated weekly. RESULTS: Mifepristone administration resulted in a temporary two- to threefold increase in plasma cortisol levels (p < .0001). No significant effects were observed on any measure of neurocognitive function, including the primary outcome measures of spatial working memory and declarative memory. Minor changes in symptoms occurred in both arms of the study and were indicative of a general improvement over time, irrespective of treatment. CONCLUSIONS: In contrast to our earlier report of positive effects in bipolar disorder, these data suggest that the GR-antagonist mifepristone has no effect on neurocognitive function or symptoms in this group of patients with schizophrenia. Future studies in schizophrenia should examine patients with demonstrable hypothalmic-pituitary-adrenal axis dysfunction.     

The modulatory effects of corticosteroids on cognition: studies in young human populations

In the present article, we report on two studies performed in young human populations which tested the cognitive impact of glucocorticoids (GC) in situations of decreased or increased ratio of mineralocorticoid (MR) and glucocorticoid (GR) receptor occupation. In the first study, we used a hormone replacement protocol in which we pharmacologically decreased cortisol levels by administration of metyrapone and then restored baseline cortisol levels by a subsequent hydrocortisone replacement treatment. Memory function was tested after each pharmacological manipulation. We observed that metyrapone treatment significantly impaired delayed recall, while hydrocortisone replacement restored performance at placebo level. In the second study, we took advantage of the circadian variation of circulating levels in cortisol and tested the impact of a bolus injection of 35 mg of hydrocortisone in the late afternoon, at a time of very low cortisol concentrations. In a previous study with young normal controls, we injected a similar dose of hydrocortisone in the morning, at the time of the circadian peak, and reported detrimental effects of GC on cognitive function. Here, when we injected a similar dose of hydrocortisone in the afternoon, at the time of the circadian trough, we observed positive effects of GC on memory function. The results of these two studies provide evidence that GC are necessary for learning and memory in human populations.     

Hippocampus, glucocorticoids and neurocognitive functions in patients with first-episode major depressive disorders

The aim of this study was to determine whether there was any relationship between hippocampal volume, and glucocorticoid regulation, and cognitive dysfunctions in drug-naïve major depressive disorder (MDD) patients during their first episode. Twenty drug-free female MDD patients in their first episode and 15 healthy females as control subjects were included in the study. All subjects underwent 3.0 Tesla (T) magnetic resonance imaging (MRI), comprehensive neuropsychological testing and dexamethasone suppression tests (DST). The volumes of the right and left hippocampus of the patients were found to be significantly smaller than those of the controls. Patients were found to have significantly lower scores on measures of attention, working memory, psychomotor speed, executive functions, and visual and verbal memory fields. The performance of the patients only in the recollection memory and memory of reward-associated rules were positively correlated with hippocampal volumes. The volumes of the left and right hippocampus did not correlate with basal or post-dexamethasone cortisol levels. Our findings indicate that depressed patients have smaller hippocampi even in the earlier phase of their illness. Further research efforts are needed to explain the mechanisms that are responsible for the small hippocampus in depressed patients.     

Saliva cortisol and response to dexamethasone in children of depressed parents.

BACKGROUND: Major depression (MDD) is heritable, and children of depressed parents are at higher risk for the development of depression. However, depression in a parent might also act as a stressor leading to increased activation of neuroendocrine stress circuits. To address this question we examined saliva cortisol in children whose parents have a history of MDD. METHODS: We recruited 15 families with one parent with MDD (26 prepubertal children) and 16 control families without history of parental MDD (32 prepubertal children). All parents and children underwent Structured Clinical Interview for DSM-IV and Kiddie Schedule For Affective Disorders And Schizophrenia interviews, respectively. Families were asked to collect morning, afternoon, and bedtime saliva samples for 4 days for 2 weeks. At bedtime of the 3rd day, dexamethasone was administered. Two doses, standard and low, were used in each family. RESULTS: The majority of children demonstrated no psychiatric diagnosis. Children with MDD parents showed higher cortisol basally and higher cortisol after both 25 mg and 5 mg dexamethasone. However, this effect occurred predominantly in children whose parents were currently depressed. There were strong correlations for cortisol between parents and children (r = 52 in depressed; r = 499 in control). CONCLUSIONS: Elevated cortisol and impaired feedback seemed to reflect an environmental effect of MDD in a parent.     

Plasma cortisol levels, brain volumes and cognition in healthy elderly men

PURPOSE: In ageing animals, exposure to chronic high levels of glucocorticoids is associated with cognitive impairment and hippocampal atrophy. However, there are few studies examining relationships among glucocorticoids, brain volumes and cognitive function in healthy older humans. This study examined the hypotheses that higher plasma cortisol levels and altered sensitivity to glucocorticoids are associated with worse cognition and more brain atrophy in elderly men. MATERIALS AND METHODS: Ninety-seven healthy men aged 65-70 had plasma cortisol measured at 09:00, 14:30 h, and post-dexamethasone (0.25mg, 09:00 h), and had dermal sensitivity to glucocorticoids measured. They also underwent cognitive testing, with scores adjusted for estimated prior mental ability, and had MRI measurements of intracranial area (a validated estimate of intracranial capacity), and hippocampus, temporal lobe and frontal lobe volumes. RESULTS: Plasma cortisol levels at 09:00 h were significantly and negatively correlated with a summary General Cognitive Factor accounting for 51% of the variance of cognitive function (rho=-0.22, p=0.035), and specific cognitive tests: delayed paragraph recall (rho=-0.28, p=0.036) and processing speed (rho=-0.23, p=0.026). Regional brain volumes adjusted for intracranial area generally did not correlate with cortisol levels. Tissue glucocorticoid sensitivity did not correlate with any measure of cognition or brain volume. CONCLUSIONS: In healthy older men, higher plasma cortisol levels are associated with worse ageing-related overall cognitive change but not ageing-related brain atrophy.     

Effects of cortisol on the memory bias for emotional words? A study in patients with depression and healthy participants using the Directed Forgetting task

Mood congruent alterations in information processing such as an impaired memory bias for emotional information and impaired inhibitory functions are prominent features of a major depressive disorder (MDD). Furthermore, in MDD patients hypothalamic–pituitary–adrenal axis dysfunctions are frequently found. Impairing effects of stress or cortisol administration on memory retrieval as well as impairing stress effects on cognitive inhibition are well documented in healthy participants. In MDD patients, no effect of acute cortisol administration on memory retrieval was found.The current study investigated the effect of acute cortisol administration on memory bias in MDD patients (N = 55) and healthy controls (N = 63) using the Directed Forgetting (DF) task with positive, negative and neutral words in a placebo controlled, double blind design. After oral administration of 10 mg hydrocortisone/placebo, the item method of the DF task was conducted. Memory performance was tested with a free recall test.Cortisol was not found to have an effect on the results of the DF task. Interestingly, there was significant impact of valence: both groups showed the highest DF score for positive words and remembered significantly more positive words that were supposed to be remembered and significantly more negative words that were supposed to be forgotten. In general, healthy participants remembered more words than the depressed patients. Still, the depressed patients were able to inhibit intentionally irrelevant information at a comparable level as the healthy controls. These results demonstrate the importance to distinguish in experimental designs between different cognitive domains such as inhibition and memory in our study.     

Cognitive effects of corticosteroids

In three independent studies with different designs and groups of subjects, the authors found that 1) depressed patients who did not suppress cortisol when given dexamethasone (compared to suppressors and normal control subjects), 2) healthy volunteers given a single 1-mg dose of dexamethasone (compared to those given placebo), and 3) healthy volunteers given 80 mg/day of prednisone for 5 days (compared to those given placebo) all made significantly more errors of commission in verbal memory tasks, with no significant change in their rates of errors of omission. These findings raise the possibility of specific corticosteroid-related cognitive impairments.