Cognitive Functioning, Cortisol Release, and Symptom
Severity in Patients with Schizophrenia
Deborah J. Walder, Elaine F. Walker, and Richard J. Lewine
Background: There is substantial evidence of dysregula-
tion of cortisol secretion, hippocampal abnormalities, and
memory deficits in schizophrenia and other psychotic
disorders. Research also suggests that cortisol secretion
augments dopaminergic activity, which may result in
increased symptom expression in this clinical population.
Methods: We examined the relations among cortisol
release, cognitive performance, and psychotic symptom-
atology. Subjects were 18 adults with schizophrenia or
schizoaffective disorder, seven with a nonpsychotic psy-
chiatric disorder, and 15 normal control subjects. Tests of
memory and executive function were administered. Corti-
sol was assayed from multiple saliva samples.
Results: Findings indicated the following: 1) patients with
psychotic disorders scored below the comparison groups
on the cognitive measures; 2) for the entire sample,
cortisol levels were inversely correlated with performance
on memory and frontal tasks; and 3) among patients,
cortisol levels were positively correlated with ratings of
positive, disorganized, and overall symptom severity, but
not with negative symptoms.
Conclusions: The present results suggest that abnormal-
ities in the hypothalamic–pituitary–adrenal axis and hip-
pocampal systems play a role in observed cognitive
deficits across populations. Among psychotic patients,
elevated cortisol secretion is linked with greater symptom
severity. Biol Psychiatry 2000;48:1121–1132 © 2000
Society of Biological Psychiatry
Key Words: Cortisol, hippocampus, memory, cognitive,
(Gold et al 1992a, 1992b; Goldberg et al 1990; Hoff et al
1992; Saykin et al 1994; Schmand et al 1992). Paralleling
these findings, research has shown that schizophrenia is
associated with hippocampal volumetric reductions (e.g.,
atients with psychotic disorders manifest deficits in
short-term verbal and visuospatial memory functions
Bogerts et al 1985; Breier et al 1992; Jeste and Lohr 1989;
Suddath et al 1990; Waldo et al 1994) and cellular
abnormalities (e.g., Akbarian et al 1993; Altshuler et al
1990; Benes et al 1991; Bogerts et al 1990; Jeste and Lohr
1989; Kovelman and Scheibel 1984). In recent studies,
reductions in hippocampal volume have been found in
first-episode schizophrenia patients, suggesting that the
abnormality is not a treatment artifact (Hirayasu et al
1998; Velakoulis et al 1999). This also extends to patients
with other psychotic disorders (Velakoulis et al 1999).
A link between hippocampal morphology and memory
has been established in studies of nonschizophrenic sam-
ples, such that the volume of the hippocampus is inversely
related with memory test performance (Lencz et al 1992;
Starkman et al 1992). This is consistent with the prevailing
theory that the hippocampal system plays a major role in
declarative (or explicit) memory (Eichenbaum et al 1996;
Squire 1987, 1992). A few studies show no significant
relationship between delayed memory and hippocampal
volume in schizophrenia spectrum patients (Colombo et al
1993; Delisi et al 1991; Nestor et al 1993; Torres et al
1997). In contrast, Goldberg et al (1994) found a strong
association between intrapair differences in a parameter of
verbal memory and intrapair differences in left hippocam-
pal volume in monozygotic twins discordant for schizo-
phrenia. Discrepancies in these findings may be due to
variability in the cognitive measures selected to index
memory. It is also noteworthy that Goldberg et al (1994)
focused on the hippocampal volume difference between
affected twins and their healthy monozygotic cotwins.
Thus, there was some control for genetically determined
differences in volume.
In addition to its role in memory functioning, the
hippocampus is assumed to play a role in the regulation of
the hypothalamic–pituitary–adrenal (HPA) axis. Specifi-
cally, it appears that the stimulation of glucocorticoid
receptors in the hippocampus contributes to a negative
feedback system that dampens HPA activity. Animal
studies suggest that prolonged stress and elevations in
corticosteroids can damage the hippocampus (Sapolsky
and McEwen 1986; Sapolsky et al 1990), which may lead
to dysregulation of the HPA axis via a decrease in
glucocorticoid negative feedback. Thus, studies of human
subjects revealing an inverse correlation between cortisol
From the Departments of Psychology (DJW, EFW) and Psychiatry (RJW),
Emory University, Atlanta, Georgia.
Address reprint requests to Deborah J. Walder, M.A., Emory University, Depart-
ment of Psychology, 532 North Kilgo Circle, Atlanta GA 30322.
Received March 28, 2000; revised August 28, 2000; accepted September 1, 2000.
© 2000 Society of Biological Psychiatry 0006-3223/00/$20.00
and hippocampal volume (Rao et al 1989; Starkman et al
1992) are not surprising.
Consistent with the above findings, there is also an
inverse relation between cortisol level and performance
on measures of hippocampal function, particularly declar-
ative memory, in healthy human subjects. This relation has
been shown in experimental studies in which glucocorti-
coids are manipulated and memory decrements ensue
(Kirschbaum et al 1996; Newcomer et al 1994, 1999),
studies in which stress-induced alterations in cortisol are
associated with memory deficits (Lupien et al 1997), and
longitudinal studies of the relation between age-related
declines in cortisol and memory performance (Seeman et
al 1997). It is noteworthy that Kirschbaum et al (1996)
showed that memory performance deficits follow acute,
relatively low-stress exposure (e.g., Trier Social Stress
Test), or what is presumed to be relatively low-dose
increases in cortisol. In contrast, one more recent study
(Newcomer et al 1999) found that similar cognitive
impairments following exogenous glucocorticoid treat-
ment are high-dose specific (e.g., approximating cortisol
exposure during moderate to maximum stress conditions).
Future research aimed at clarifying the specific effects of
low- versus high-dose and acute versus chronic glucocor-
ticoid exposure is warranted.
The precise nature of memory deficits following acute
elevations in cortisol is also unclear. Some studies suggest
impairments in immediate recall following stress-induced
cortisol secretion (Lupien et al 1997), whereas others
suggest a retrieval-specific impairment following exoge-
nous administration of corticosteroids (De Quervain et al
2000). Overall, this body of literature suggests that cortisol
is linked with memory function in two ways: 1) directly,
by acutely disrupting working memory and short-term
recall, and 2) indirectly, through the effects of persistent
cortisol elevations on hippocampal integrity.
To date, only two studies have explored the relation
between glucocorticoids and memory performance in
schizophrenia. One early report (Newcomer et al 1991)
showed an inverse relation between early morning (8:00
AM) postdexamethasone cortisol levels and auditory verbal
learning deficits in unmedicated schizophrenic patients. In
a more recent experimental study, Newcomer et al (1998)
examined the effects of dexamethasone versus a placebo
on verbal memory performance in schizophrenic patients
over 4 days. The finding of an association between higher
plasma cortisol concentrations (before dexamethasone
treatment) and reduced memory performance in patients
with schizophrenia extended their earlier findings. There
was no relation, however, between postdexamethasone
cortisol levels and recall performance. It is noteworthy that
although several studies evidence elevated basal levels of
cortisol (Walker and Diforio 1997) and postdexametha-
sone nonsuppression (Arana et al 1983; Sharma et al 1988)
in schizophrenia, these findings are not consistently ob-
served in this population.
Previous studies indicate a relation between HPA activ-
ity and symptomatology in schizophrenia. In some, corti-
sol secretion was primarily associated with more severe
positive symptoms (Kaneko et al 1992; Keshavan et al
1989; Rybakowski et al 1991), whereas in others it was
associated with higher ratings of negative symptoms
(Newcomer et al 1991; Tandon et al 1991). It has been
suggested that the relation between cortisol levels and
symptom severity is due to the augmenting effects of
cortisol on dopamine activity (Walker and Diforio 1997).
Findings from several studies revealing a relation of
pre- or postdexamethasone suppression test cortisol levels
with psychotic symptoms suggest that this association is
not affected by level of depression (Schatzberg and Roth-
schild 1988; Sharma et al 1988). Thus, the association of
cortisol with symptom severity is not attributable to the
presence of mood disorder in some psychotic patients.
This study is based on the assumption that heightened
cortisol release is associated with memory deficits in
normal healthy control subjects. Thus, it was hypothesized
that cortisol level would be inversely correlated with
memory performance in patients with psychotic disorders.
In addition, we examined the association of symptom
severity with both cortisol secretion and memory perfor-
mance. Given the evidence that HPA activation augments
dopamine activity (Schatzberg et al 1985), it was predicted
that elevated cortisol levels would be associated with
increased symptom severity. This study is novel in that it
examines the relations among cognitive performance,
symptom expression, and neuroendocrine measures in a
psychiatric population, in the absence of the stress typi-
cally associated with blood sampling procedures or exog-
enously administered pharmacologic agents.
To better understand the association of the HPA axis
with memory and symptom expression, cortisol levels
were assessed at multiple time points. It is important to
note that participation in a research study may, in itself,
constitute a psychological stressor for many individuals
(Weinstein et al 1999). At the very least, the research
laboratory is a novel context for most people. This, in turn,
may influence tonic measures of cortisol. Thus, what is
often conceived of as “baseline” cortisol (e.g., the initial
cortisol sample upon entry into a study) may not be a true
index of stable cortisol levels for the individual. Instead,
this initial measure of cortisol may be an index of
sensitivity to novelty and reflective of HPA stress respon-
sivity. In contrast, cortisol levels sampled after the indi-
vidual is acclimated to the research setting may better
approximate baseline levels of cortisol and serve as an
index of hippocampal integrity, as the novelty of the
1122 D.J. Walder et al
situation diminishes. These factors were considered when
interpreting the findings in this study.
Methods and Materials
The study samples were 15 normal control (NC) subjects, 18
patients with current or past (in remission) diagnosis of a
psychotic disorder (PD), and seven individuals with other psy-
chiatric disorders (ODs). Subjects ranged from 18 to 58 years of
age, with an overall mean age of 30.8 years (see Table 1 for
demographic characteristics). The PD group included patients
who met DSM-IV criteria for schizophrenia, including paranoid,
disorganized, residual, and undifferentiated type (n ? 12) or
schizoaffective disorder (n ? 6). Subjects in the OD group met
criteria for affective disorders without psychotic features (e.g.,
major depressive disorder, history of a major depressive epi-
sode), and one subject met criteria for pervasive developmental
disorder. Normal control subjects were recruited from the com-
munity; all other subjects were recruited from inpatient and
outpatient psychiatric services. Fifteen of the 18 subjects with
psychotic disorders were being treated with antipsychotics. Some
patients refused medication.
Informed consent was obtained from all subjects before their
participation in this research study. The study protocol was
reviewed and approved by the Human Investigations Committee
at Emory University before data collection.
view for DSM-IV, research version (SCID-IV; First et al 1997),
the Schedule for Assessment of Positive Symptoms (SAPS;
Andreasen 1983), and the Schedule for Assessment of Negative
Symptoms (SANS; Andreasen 1981) were administered by
trained examiners to all participants at the initial assessment.
Diagnostic interviews were videotaped so that consensus diag-
noses could be established.
The Structured Clinical Inter-
SALIVA SAMPLING FOR THE ASSAY OF CORTISOL.
measurement of salivary cortisol has been purported to be a
reliable tool for investigating HPA activity (Kiess et al 1995;
Kirschbaum and Hellhammer 1989; Laudat et al 1988). More-
over, as pointed out by Weinstein et al (1999), salivary, urinary,
and plasma measures of cortisol are highly interrelated, as well as
comparable in sensitivity as measures of stress reactivity (Bassett
et al 1987; Shipley et al 1992). Finally, the noninvasive nature of
saliva sampling yields it a more preferable method for repeated
assessments (Baum and Grunberg 1995).
Saliva samples (about 1 mL each) for the assay of cortisol
were obtained from subjects three to five times during the
assessment. These samples were obtained by asking the subjects
to spew saliva into plastic specimen tubes, in which the samples
were stored. To maintain uniformity in the portion of the diurnal
variation in cortisol excretion represented in the samples, all
assessments were conducted at the same time of day. The first
sample was obtained at approximately 8:00 AM. Subsequent
samples were obtained at approximately 9:00 AM, 10:00 AM,
11:00 AM, and 12:00 PM. By obtaining multiple measures, it was
possible to examine group and individual differences in change,
as well as mean levels.
assay according to standard procedures. They were stored at
?20°C in a 13–cu ft S/P (Asheville, NC) cryofreezer. In
preparation for assay, the samples were rapidly thawed and
centrifuged at 300 g for 10 min to remove coagulated protein and
other insoluble material. Cortisol was assayed in duplicated
200-?L aliquots of the clear supernatant using materials and
procedures provided by Incstar (Stillwater, MN). The assay was
performed in tubes coated with an antiserum that shows signif-
icant cross-reactivity only with prednisone (83%), 11-deoxycor-
tisol (6.4%), cortisone (3.6%), and corticosterone (2.3%). Stan-
dards in the range 1–30 ng/mL consisted of the serum standards
provided with the kit materials diluted with 200 ?L of phos-
phate-buffered saline. Protein concentrations were equalized in
standards and samples by adding cortisol-free serum to the
The saliva samples were prepared for
Table 1. Characteristics of Current Sample
(3) Grand total
Number of subjects
36.72 (11.66) 26.71 (3.95) 25.67 (8.18)30.83 (10.72)2 ? 1,a3 ? 1b
ap ? .05, one-tailed test.
bp ? .01, one-tailed test.
Cognition, Cortisol, and Schizophrenia 1123
samples. For the most recent assays conducted by the lab, the
mean coefficients of variation between duplicates and between
assays were less than 5%. Relative to the serum standards, the
mean recovery of cortisol from saliva has been indistinguishable
The neuropsychological measures were selected to yield a
comprehensive profile of hippocampal functioning and atten-
ing Test (CVLT; Delis et al 1983) The CVLT is typically used to
assess memory deficits associated with hippocampal functioning.
It measures recall and recognition of lists containing related
words, across several trials.
The composite score derived for this test was an average of the
standardized scores on the seven subtests of this measure,
including two measures of immediate recall, two measures of
short-delay recall (free and cued), two measures of long-delay
recall (free and cued), and one measure of long-delay recogni-
tion. Higher scores indicate better performance.
California Verbal Learn-
Wechsler Memory Scale—Revised (WMS-R; Wechsler 1987)
Subtests of the WMS-R are also frequently used to assess
memory deficits associated with hippocampal functioning. The
WMS-R scales require recall of stories and paired word associ-
ates. Patients with schizophrenia have significant memory im-
pairments on the original Wechsler Memory Scale (Goldberg et
al 1990) and the WMS-R (Gold et al 1992b). Research findings
suggest that memory impairment based on this measure is not
solely attributable to general cognitive deterioration in schizo-
phrenia (Gold et al 1992b). The four subtests administered from
this measure (Logical Memory I, Logical Memory II, Verbal
Paired Associates I, and Verbal Paired Associates II) focus on
short-term learning, recall, and delayed recall of verbal material.
The composite score derived for the WMS-R was an average
of the standardized scores on the four subtests, immediate and
delayed recall of a story, and immediate and delayed recall of
verbalpaired associates. Higher
Modified Wisconsin Card Sorting Test (MCST; Nelson 1976)
This test is a simplified version of the Wisconsin Card Sorting
Task, revised to reduce ambiguity concerning the sorting princi-
ple active at any point in the test. Poor performance on the MCST
has been demonstrated in schizophrenic patients (Nelson 1976).
The MCST is a measure traditionally used to assess frontal lobe
(or executive) functioning. There is evidence, however, suggest-
ing that volume reductions in the anterior hippocampal formation
may predict neuropsychological deficits in executive functions
(Bilder et al 1995). Therefore, given the hypothesized connec-
tivity between the frontal and hippocampal brain regions, the
MCST was selected as an additional potential indicator of
In this test the subject is required to sort a deck of stimulus
cards into groups corresponding to the color, form, or number
printed on four key cards before him or her. The subject is not
told the correct sorting principle but is informed by the examiner
whether or not each card was sorted correctly. The subject must
deduce the correct sorting rule based upon the information
Indices derived from this test are total number of correct
responses, total number of errors, and number of perseverative
errors. The composite score was an average of the standardized
scores for the total number of errors and number of perseverative
errors. Thus, for total number of correct responses, higher scores
indicate better performance. For the composite score, however,
higher scores indicate poorer performance.
Vigil) (ForThought, Nashua NH) CPT-Vigil is a computerized
attention measure typically used to assess vigilance, or mainte-
nance of attention, over time. It has been shown to be a valid test
of individual differences in vigilance performance (Buchsbaum
and Sostek 1980). Because memory involves aspects of attention,
this measure was administered to a subgroup of 33 subjects to
examine the relation between memory performance and atten-
tional functioning. Data were not collected on the remaining
seven subjects of the total sample because of technical problems
with computer equipment.
This computerized task measures sustained concentration and
attention by presenting target stimuli (e.g., letters) on the
computer screen one at a time. The “AK” version of the task,
which requires the subject to press the space bar on the keyboard
in response to a complex target stimulus (e.g., the letter K only
when preceded by the letter A), was used. This stimulus is
presented among other randomly presented nontarget stimuli
(e.g., other letters of the alphabet presented in an order not
ascribing to the rule outlined above).
The test apparatus consisted of a desktop computer monitor
and keyboard. The signal stimuli were presented on the computer
screen for a duration of 85.3 msec and at an interstimulus interval
of 910.3 msec. A total of 100 targets out of 480 total stimuli were
presented. The stimulus sequence was randomized. A response to
a nontarget stimulus was considered an “error of commission,”
and a failure to respond in the presence of a target stimulus was
considered an “error of omission.” From these measures, an
index of sensitivity (d?) was derived and used in the current
analyses. Sensivity (d?) has been shown to have moderate
reliability (Buchsbaum and Sostek 1980) and is a measure of
attentiveness or discriminability.
Continuous Performance Task—Vigil (CPT-
Procedures and Time Frame
The order of administration of the tests was the same for all
subjects: CVLT (immediate recall), MCST, CVLT (20-min
delayed recall), subtests of the WMS-R (immediate recall),
subtests of the WMS-R (delayed recall), and CPT-Vigil.
Assessments began at approximately 8:00 AM, with a descrip-
tion of the study goals and measures, and took roughly 3.5–4.5
hours to complete. The consent form described the procedures
for maintaining confidentiality, the study protocol, and the
subject’s right to withdraw from participation at any time.
The first saliva sampling was at 8:00 AM, followed by one
1124D.J. Walder et al
every hour. The diagnostic interview (SCID-IV), from which
SANS and SAPS ratings were determined, followed the initial
saliva sampling and was followed by the neuropsychological
assessment. All interviewers were trained in conducting the
SCID-IV before commencement of this study. The neuropsycho-
logical assessment was administered by another trained research
assistant, who was blind to the subjects’ interview results.
Subjects were asked to refrain from alcoholic beverages for
one day before the assessment. They were also instructed to
avoid consumption of caffeinated food and beverages beginning
the night before the assessment and to eat a light breakfast before
their appointment. Any deviations from the instructions were
3 for all subjects. Cortisol levels at times 4 and 5 were only
collected for those subjects for whom the duration of the
assessment continued for 3–4 hours. This was more often the
case for subjects belonging to the inpatient and outpatient clinical
samples, as compared with NC subjects. Because 42.5% of the
total sample was missing data for cortisol at time 4 and 82.5% of
the total sample was missing data for cortisol at time 5, these
measures were excluded from all analyses.
Two additional indices of HPA activity were computed for
each subject: 1) the average of cortisol at times 1, 2, and 3 and
2) the slope of cortisol across times 1, 2, and 3 (i.e., the linear
slope of the regression of cortisol on time). The linear slope
provides an index of the magnitude of the change in cortisol
levels during the assessment period for each subject. The slope of
cortisol has previously been used as a sensitive measure to index
cortisol responsivity to the initial novelty of similar assessment
procedures (Weinstein et al 1999). One-sample t tests were
conducted comparing the slope for each of the four groups to a
test value of zero. Results indicated that the slope value for each
of the four groups was significantly different from zero (total
sample, t(39) ? 14.143, p ? .000; NC group, t(14) ? 7.421, p ?
.000; PD group, t(17) ? 9.296, p ? .000; OD group, t(6) ?
9.906, p ? .000).
Cortisol levels were measured at times 1, 2, and
Table 2 lists correlation coefficients among measures of
cortisol. The high intercorrelations indicate reliable assay of
It has been shown that exposure to novelty can heighten
cortisol levels (Al’Absi and Lovallo 1993; Hennessy et al 1995).
Thus, it is possible that in this study cortisol at time 1 served as
an index of sensitivity to the initial novelty of the assessment. In
contrast, saliva samples at times 2 and 3 may reflect baseline
levels of cortisol. If this is the case, then the slope of cortisol may
be an alternative measure of stress sensitivity as well, given that
it indexes the magnitude of the linear change in cortisol from the
initial level over time.
WMS-R, CVLT, and MCST standardized scores were examined
and internal consistency (?) was computed. Results indicated
high internal consistency among the four WMS-R subtests (? ?
.8804), the seven CVLT subtests (? ? .9445), and the two
MCST indices (? ? .9424). Therefore, z scores for the subtests
of each of these measures were averaged together to derive
CVLT, WMS-R, and MCST composite scores, respectively.
Sensitivity (d?) was calculated according to standard procedures
(McNicol and Willson 1971).
The CVLT and WMS-R composite scores were highly corre-
lated (? ? .8516). Therefore, they were averaged to compute an
index of “explicit memory.” There were high intercorrelations
among the CVLT, WMS-R, and MCST total number of correct
responses scores (? ? .7947). Therefore, the three scores were
averaged to compute an aggregate index of “hippocampal func-
tioning.” Scaling was the same for all variables in the composite
Interitem correlations among the
a three-symptom dimension model may best fit the data on
clinical symptoms in patients with schizophrenia (Andreasen et
al 1995). Accordingly, positive symptoms are divided into two
dimensions, a psychosis dimension (i.e., delusions and halluci-
nations) and a disorganized dimension (i.e., disorganized speech,
disorganized behavior and inappropriate affect). The third di-
mension is composed of negative symptoms (i.e., affective
flattening, physical anergia). Using Andreasen et al’s (1995)
three-factor solution, we derived Positive, Negative, and Disor-
ganized Symptoms scale scores to measure symptomatology.
Internal consistency (Chronbach’s ?) was examined for each
of the hypothesized scales, excluding items with missing data
and/or that had restricted variance across ratings. The reliabilities
for the hypothesized Positive Symptoms scale (? ? .8631),
Negative Symptoms scale (? ? .6986), and Disorganized Symp-
toms scale (? ? .7100) were high. Therefore, ratings on the items
within each scale were averaged to obtain a composite index of
severity of positive, negative, and disorganized symptoms, re-
spectively. The items comprising each scale and their associated
criteria are contained in Appendix 1.
Neuropsychological dysfunction in schizophrenia may be
more closely associated with overall symptom severity than
specific symptom dimensions (Goldberg and Weinberger 1995).
Therefore, ratings across the three symptom scales were aver-
aged to create a global index of symptom severity. The reliability
Recent research findings suggest that
Table 2. Correlation Coefficients for Cortisol Measures for
at time 1
at time 2
at time 3
ap ? .01, one-tailed test, n ? 40.
Cognition, Cortisol, and Schizophrenia 1125
for this scale was low (? ? .3078). This was expected, given that
the three scales were presumed to measure different dimensions
DIAGNOSTIC GROUP DIFFERENCES.
ples t tests were employed to test for group differences in age,
cognitive performance, cortisol levels, and symptom scale rat-
ings. One-tailed tests were used because the hypothesized group
differences in the examined variables were directional in nature.
than the OD group [t(23) ? 2.198, p ? .019] and NC group
[t(31) ? ?0.3090, p ? .002] (Table 1). ?2tests comparing the
PD, OD, and NC groups on gender and race indicated no group
differences in these demographic characteristics.
The PD group was significantly older
group on d? [t(9) ? ?1.885, p ? .038], the global index of
hippocampal functioning [t(23) ? ?5.000, p ? .000], and the
global index of explicit memory [t(23) ? ?5.018, p ? .000]. The
PD group also scored below the NC group on d? [t(28) ? 3.241,
p ? .002], the global index of hippocampal functioning [t(31) ?
6.233, p ? .000], and the global index of explicit memory
[t(31) ? 5.941, p ? .000]. The PD group scored higher (poorer
performance) on the MCST composite score than the OD group
[t(23) ? 2.416, p ? .012] and the NC group [t(31) ? ?3.373,
p ? .001].
These analyses were also conducted with age as a covariate
because there were group differences in age. There was no longer
a significant difference between the PD and OD groups on d?.
This was the only change in the pattern of results.
The PD group scored below the OD
Symptom Scale Ratings
higher ratings than the OD group on the Positive Symptoms scale
[t(23) ? 1.854, p ? .039], the Disorganized Symptoms scale
[t(23) ? 2.526, p ? .010], and the Overall Symptoms scale
[t(23) ? 3.160, p ? .002]. The diagnostic groups did not differ
on negative symptoms.
The PD group had significantly
measures of cortisol. However, it is possible that psychotropic
There were no diagnostic group differences on the
medication was masking group differences, given evidence that
antipsychotics reduce cortisol levels (Walker and Diforio 1997).
In addition, some studies have evidenced hypercortisolemia in
depressed patients (Carpenter and Bunney 1971; Deuschle et al
1998). More specifically, Galard et al (1991) and Gotthardt et al
(1995) showed significantly higher basal levels of cortisol in
depressed patients relative to NC subjects. The inclusion of a
large majority of subjects with a history of depression in the OD
group might therefore account for the absence of a significant
difference in cortisol secretion between the PD and OD groups.
Some psychiatric populations (viz., depressed patients) have
also been shown to exhibit hypersecretion of cortisol during the
afternoon (Christie et al 1986; Seckl et al 1991). Given the
diagnostic nature of the OD group subjects, restriction of the
cortisol samplings to morning hours in this study may account
for the lack of significant group differences in cortisol.
The Relation between Cortisol and Cognitive
All correlational analyses were conducted using one-tailed
tests because of the directional nature of the hypothesized
relations among the examined variables. Pearson correla-
tion analyses conducted across the total sample (PD group,
OD group, and NC group) indicated that explicit memory
and hippocampal functioning were negatively correlated
with time 3 cortisol (Table 3). Hippocampal functioning
was also negatively correlated with time 2 cortisol. The
pattern of results indicates that higher cortisol was asso-
ciated with poorer performance. Performance on the
MCST, as indexed by the composite score, was not
correlated with any of the cortisol measures.
Correlational analyses examining discrete aspects of
memory performance—namely, immediate, short-term,
and long-term memory—were also conducted (Table 4).
Results indicated that time 1 cortisol was not associated
with any of the three derived subcomponents of memory
in either the total sample or the NC group. Time 2 cortisol
Table 3. Correlation Coefficients and Partial Correlation Coefficients Relating Cortisol with Explicit Memory and Hippocampal
Functioning, for Total Sample
at time 1
at time 2
at time 3
Explicit memorya(no covariate)
Explicit memoryc(with CPT sensitivity as covariate)
Hippocampal functioninga(no covariate)
Hippocampal functioningc(with CPT sensitivity as covariate)
MCSTc(with CPT sensitivity as covariate)
CPT, Continuous Performance Test; MCST, Modified Wisconsin Card Sorting Test.
an ? 40.
bp ? .05, one-tailed test.
cn ? 33.
dp ? .01, one-tailed test.
1126D.J. Walder et al
was significantly correlated with all three memory vari-
ables in the NC group, but with only immediate recall in
the total sample. Time 3 cortisol was significantly corre-
lated with all three memory variables in the total sample,
but with only immediate and long-term memory in the NC
group. Mean cortisol and cortisol slope were significantly
correlated with immediate and long-term memory in the
NC group. In the total sample, cortisol slope was signifi-
cantly correlated with immediate memory; no other cor-
relations were significant within this group. There were no
significant findings in the PD and OD groups alone.
Overall, these results suggests that cortisol levels were
associated with immediate, short-term, and long-term
memory performance .
It is possible that attentional problems contribute to the
observed memory deficits in schizophrenia. The correla-
tional analyses were therefore repeated controlling for
attention, as indexed by sensitivity performance on the
CPT-Vigil. All of the significant correlations held after
covarying for attention. In addition, there were significant
negative correlations for cortisol slope with explicit mem-
ory and hippocampal functioning, as well as time 2
cortisol with explicit memory, and significant positive
correlations for time 2 cortisol, mean cortisol, and cortisol
slope with MCST (Table 3). These findings suggest that
cognitive impairment attributable to deficits in attention
did not account for the relation between cortisol and
memory performance. Instead, individual differences in
attention were obscuring the relations of cortisol with
memory and executive functioning.
The above analyses were repeated with the PD group
alone (Table 5). Results indicated no significant correla-
tions for cortisol with explicit memory, hippocampal
functioning, or MCST performance. After covarying for
attention, there was a trend toward a negative correlation
for time 2 cortisol with hippocampal functioning, as well
as a positive correlation for time 2 cortisol and cortisol
slope with MCST performance.
The majority of the findings among the PD group were
consistent with findings among the total sample in that,
although they were of lesser magnitude, they were in a
similar direction. Findings in the total sample, and to a
lesser degree in the PD group, suggest that increased
cortisol secretion, particularly those measures presumed to
reflect baseline cortisol, is associated with impaired per-
formance on cognitive measures indexing hippocampal
The Relation between Cortisol and Symptoms
Age was negatively correlated with time 1 cortisol (r ?
?.451, p ? .030) and mean cortisol (r ? ?.410, p ? .045)
within the PD group. Therefore, correlational analyses of
Table 4. Correlation Coefficients Relating Cortisol with Immediate, Short-Term, and Long-Term
Memory for Total Sample, Psychotic Disorder Group, Other Disorders Group, and Normal
at time 1
at time 2
at time 3
an ? 40.
bp ? .05, one-tailed test.
cn ? 15.
dp ? .01, one-tailed test.
en ? 18.
fn ? 7.
Cognition, Cortisol, and Schizophrenia1127
cortisol with symptom severity in the PD group controlled
for age for these two measures of cortisol (Table 6).
The pattern of results indicated that higher cortisol
levels were associated with more severe symptoms (Table
6). Time 1 cortisol was positively correlated with disor-
ganized and overall symptoms, and mean cortisol was
correlated with positive and overall symptoms. After
covarying for age, these findings held and mean cortisol
was additionally correlated with disorganized symptoms.
In addition, time 2 cortisol was correlated with positive
symptoms and time 3 cortisol was correlated with positive,
disorganized, and overall symptoms. Higher cortisol slope
was linked with higher ratings of positive and overall
The Relation between Cognitive Performance and
First-order correlations revealed that none of the symptom
dimensions was associated with the three cognitive indi-
ces. After covarying for attention, however, positive
symptoms were negatively correlated with hippocampal
functioning (r ? ?.729, p ? .002) and positively corre-
lated with MCST performance (r ? .841, p ? .000). Thus,
higher symptom ratings were associated with poorer
The Relation of Cortisol and Task Performance
Stepwise regression analyses were conducted across all
subjects with psychotic disorders. Memory performance,
hippocampal functioning, MCST performance, and all
measures of cortisol were entered as predictor variables.
The only significant predictors of symptom severity were
time 1 and time 3 cortisol. Time 1 cortisol accounted for
26.9% of the variance in Disorganized Symptom scale
ratings. Time 3 cortisol accounted for 53.7% of the
variance in Positive Symptom scale ratings and 52.7% of
the variance in Overall Symptom scale ratings.
This study was designed to examine the relations among
memory, cortisol release, and symptom expression in
patients with psychotic disorders. In general, the results
provide modest support for the hypothesis that elevations
in cortisol secretion are associated with deficits in explicit
memory performance. Among patients with psychotic
Table 5. Correlation Coefficients and Partial Correlation Coefficients Relating Cortisol with Explicit Memory and Hippocampal
Functioning for Psychotic Disorder Group
at time 1
at time 2
at time 3
Explicit memorya(no covariate)
Explicit memoryb(with CPT sensitivity as covariate)
Hippocampal functioninga(no covariate)
Hippocampal functioningb(with CPT sensitivity as covariate)
MCSTb(with CPT sensitivity as covariate)
CPT, Continuous Performance Test; MCST, Modified Wisconsin Card Sorting Test.
an ? 18.
bn ? 12.
cp ? .01, one-tailed test.
dp ? .05, one-tailed test.
Table 6. Correlation Coefficients and Partial Correlation
Coefficients Relating Cortisol with Symptom Scale Ratings,
for Psychotic Disorder Group
Symptoms Positive Negative Disorganized
Cortisol at time 1
Cortisol at time 1
(with age as a
Cortisol at time 2
Cortisol at time 3
Mean cortisol (no
(with age as a
Cortisol slope (no
.156 .335 .559b
n ? 18.
ap ? .05, one-tailed test.
bp ? .01, one-tailed test.
1128 D.J. Walder et al
disorders, cortisol shows a strong positive relation with the
severity of symptoms, especially positive symptoms.
The Relation between Cortisol and Cognitive Task
Previous research has demonstrated an inverse association
between cortisol secretion and memory (e.g., Lupien et al
1994; Seeman et al 1997). Current findings of an inverse
relationship of cortisol levels at times 2 and 3 (presumably
reflective of basal levels of cortisol) with task performance
in the total sample are consistent with these results. Thus,
it appears that cortisol levels measured after acclimation to
the research setting are most strongly linked with cogni-
tive function. It is possible that this reflects the relation of
persistent elevations in circulating corticosteroids with the
integrity of various regions of the brain, particularly the
hippocampus, which has been implicated in memory
function. Some previous research has shown that cognitive
testing successfully induces cortisol release (Lupien et al
1997). Thus, it is important to consider that the current
finding of an inverse relationship of cortisol with cognitive
performance may have been due to heightened cortisol
secretion subsequent to cognitive testing.
Although the direction of the relation between cortisol
and task performance was similar for the psychotic pa-
tients, only two of the coefficients reached statistical
significance. It is likely that this is due to insufficient
statistical power in this smaller group (n ? 18) of patients.
The Relation between Cortisol and Symptom
Correlational analyses revealed that heightened cortisol
release is associated with increased symptom expression
in psychotic disorders. Most of the subjects in the PD
group were medicated with or previously medicated with
antipsychotics. Thus, findings of an association of symp-
toms with cortisol secretion, despite medication effects,
are all the more striking.
The findings are not supportive of a stronger association
between heightened HPA activity and the deficit symp-
toms of schizophrenia, as suggested by Tandon et al
(1991). Rather, these results suggest that heightened HPA
activity is generally associated with overall symptom
expression in psychotic disorders. In terms of specific
symptom dimensions, cortisol indexing sensitivity to nov-
elty (time 1) and mean cortisol were associated with
disorganized symptom severity, whereas “basal” (times 2
and 3) measures of cortisol were associated with positive
symptom severity. The association between cortisol and
positive symptoms is consistent with several previous
findings (Kaneko et al 1992; Keshavan et al 1989; Ryba-
kowski et al 1991). No measures of cortisol were associ-
ated with negative symptoms.
This suggests a possible dissociation between positive
and disorganized symptoms on the one hand and negative
symptoms on the other. More specifically, the pattern of
association of positive, negative, and disorganized symp-
toms with the various measures of cortisol suggests that
the symptom dimensions are differentially influenced by
1) basal HPA activity linked with the integrity of the
hippocampus versus 2) acute changes in HPA activity in
response to novelty or stress, respectively.
The Relation between Cognitive Performance and
Previous studies of the relation between memory perfor-
mance and symptom severity have yielded mixed findings.
Some showed an inverse relation of memory with negative
symptoms (Basso et al 1998; Sullivan et al 1994), disor-
ganized symptoms (Basso et al 1998), or total symptom
severity (Sullivan et al 1994). Sullivan et al (1994) found
a trend in the relationship between memory and positive
symptoms. Other studies found no association between
memory and symptom severity (Hoff et al 1992; Schmand
et al 1992).
In this study, an index of MCST errors and an index of
hippocampal functioning, which incorporates the MCST,
were linked with positive symptoms but not with the other
three symptom scales. This suggests that positive symp-
toms are associated with more pronounced impairment on
measures of frontal function, rather than measures of
The Relation of Cognitive Performance and
Cortisol with Symptoms
Regression analyses indicated that time 1 cortisol pre-
dicted disorganized symptoms and time 3 cortisol pre-
dicted positive and overall symptom severity. No other
measures of cortisol or indices of cognitive functioning
(e.g., memory, executive, or hippocampal functioning)
predicted symptom severity.
In conclusion, analyses of data from the total sample
provided further support for a relationship of cortisol with
cognitive functions subserved by the hippocampus. Insuf-
ficient statistical power may account for the failure of
these relations to reach statistical significance within the
group of psychotic patients. However, highly significant
relations were found between cortisol levels and the
severity of patients’ symptoms. It is speculated that this
may reflect the ability of cortisol to increase dopaminergic
activity. Overall, these findings suggest that abnormalities
in the HPA axis and hippocampal system may play a role
Cognition, Cortisol, and Schizophrenia 1129
in the clinical manifestation of schizophrenia. Additional
research on larger samples of patients is warranted.
Akbarian S, Vinuela A, Kim JJ, Potkin SG, Bunney WE (1993):
Distorted distribution of nicotinamide-adenine dinucleotide
phosphate-diaphorase neurons in temporal lobe of schizo-
phrenics implies anomalous cortical development. Arch Gen
Al’Absi M, Lovallo WR (1993): Cortisol concentrations in
serum of borderline hypertensive men exposed to a novel
experimental setting. Psychoneuroendocrinology 18:355–363.
Altshuler LL, Casanova MF, Goldberg TE, Kleinman JE (1990):
The hippocampus and parahippocampus in schizophrenic,
suicide, and control brains. Arch Gen Psychiatry 47:1029–
Andreasen NC (1981): Scale for the Assessment of Positive
Symptoms (SAPS). Iowa City: University of Iowa Press.
Andreasen NC (1983): Scale for the Assessment of Negative
Symptoms (SANS). Iowa City: University of Iowa Press.
Andreasen NC, Arndt S, Alliger R, Miller D, Flaum M (1995):
Symptoms of schizophrenia: Methods, meanings, and mech-
anisms. Arch Gen Psychiatry 52:341–351.
Arana GW, Barreira PJ, Cohen BM, Lipinski JF, Fogelson D
(1983): The dexamethasone suppression test in psychotic
disorders. Am J Psychiatry 140:1521–1523.
Bassett JR, Marshall PM, Spillane R (1987): The physiological
measurement of acute stress in bank employees. Int J Psy-
Basso MR, Nasrallah HA, Olson SC, Bornstein RA (1998):
Neuropsychological correlates of negative, disorganized and
psychotic symptoms in schizophrenia. Schizophr Res 31:99–
Baum A, Grunberg N (1995): Measurement of stress hormones.
In: Cohen S, Kessler RC, Gordon LU, editors. Measuring
Stress: A Guide for Health and Social Scientists. New York:
Oxford University Press, 175–192.
Benes FM, Sorensen I, Bird ED (1991): Morphometric analyses
of the hippocampal formation in schizophrenic brain. Schizo-
phr Bull 17:597–608.
Bilder RM, Bogerts B, Ashtari M, Wu H, Alvir JM, Jody D, et
al (1995): Anterior hippocampal volume reductions predict
“frontal lobe” dysfunction in first episode schizophrenia.
Schizophr Res 17:47–58.
Bogerts B, Falkai P, Haupts M, Greve B, Ernst S, Tapernon-
Franz U, et al (1990): Post-mortem volume measurements of
limbic system and basal ganglia structures in chronic schizo-
phrenics. Initial results from a new brain collection. Schizo-
phr Res 3:295–301.
Bogerts B, Meertz E, Schonfeldt-Bausch R (1985): Basal ganglia
and limbic system pathology in schizophrenia. Arch Gen
Breier A, Buchanan R, Elkashef A, Munson RC, Kirkpatrick B,
Gellad F (1992): Brain morphology and schizophrenia: A
magnet resonance imaging study of limbic prefrontal cortex
and caudate structures. Arch Gen Psychiatry 49:921–926.
Buchsbaum MS, Sostek AJ (1980): An adaptive-rate continuous
performance test: Vigilance characteristics and reliability for
400 male students. Percept Mot Skills 51:707–713.
Carpenter W, Bunney W (1971): Adrenal cortical activity in
depressive illness. Am J Psychiatry 128:31.
Christie JE, Whalley LJ, Blackwood DHR, Blackburn IM, Fink
G (1986): Raised plasma cortisol concentrations are a feature
of drug-free psychotics and are not specific for depression.
Br J Psychiatry 148:58–65.
Colombo C, Abbruzzese M, Livian S, Scotti G, Locatelli M,
Bonfanti A, et al (1993): Memory functions and temporal-
limbic morphology in schizophrenia. Psychiatry Res 50:45–
Delis DC, Kramer JH, Kaplan E, Ober BA (1983): CVLT:
California Verbal Learning Test Manual. San Antonio: Psy-
Delisi LE, Hoff AL, Schwartz JE, Shields GW, Halthore SN,
Gupta SM, et al (1991): Brain morphology in first-episode
schizophrenic-like psychotic patients: A quantitative mag-
netic resonance imaging study. Biol Psychiatry 29:159–175.
De Quervain DJF, Roozendaal B, Nitsch RM, McGaugh JL,
Hock C (2000): Acute cortisone administration impairs re-
trieval of long-term declarative memory in humans. Nat
Deuschle M, Weber B, Colla M, Depner M, Heuser I (1998):
Effects of major depression, aging and gender upon calcu-
lated diurnal free plasma cortisol concentrations: A re-
evaluation study. Stress 2:281–287.
Eichenbaum H, Schoenbaum G, Young B, Bunsey M (1996):
Functional organization of the hippocampal memory system.
Proc Natl Acad Sci U S A 93:13500–13507.
First MB, Spitzer RL, Gibbon M, Williams JBW (1997):
Structured Clinical Interview for DSM-IV Axis I Disorders—
Patient Edition (SCID-I/P Version 2.0, 4/97 Revision). Wash-
ington, DC: American Psychiatric Press, Biometrics Research
Galard R, Gallart JM, Catalan R, Schwartz S, Arguello JM,
Castellanos JM (1991): Salivary cortisol levels and their
correlation with plasma ACTH levels in depressed patients
before and after the DST. Am J Psychiatry 148:505–508.
Gold JM, Randolph C, Carpenter CJ, Goldberg TE, Weinberger
DR (1992a): Forms of memory failure in schizophrenia. J
Abnorm Psychol 101:487–494.
Gold JM, Randolph C, Carpenter CJ, Goldberg TE, Weinberger
DR (1992b): The performance of patients with schizophrenia
on the Wechsler Memory Scale-Revised. Clin Neuropsychol
Goldberg TE, Ragland JD, Torrey EF, Gold JM, Bigelow LB,
Weinberger DR (1990): Neuropsychological assessment of
monozygotic twins discordant for schizophrenia. Arch Gen
Goldberg TE, Torrey EF, Berman KF, Weinberger DR (1994):
Relations between neuropsychological performance and brain
morphological and physiological measures in monozygotic
twins discordant for schizophrenia. Psychiatry Res 55:51–61.
Goldberg TE, Weinberger DR (1995): A case against subtyping
in schizophrenia. Schizophr Res 17:147–152.
Gotthardt U, Schweiger U, Fahrenberg J, Lauer CJ, Holsboer F,
Heuser I (1995): Cortisol, ACTH, and cardiovascular re-
sponse to a cognitive challenge paradigm in aging and
depression. Am J Physiol 268:R865–R873.
1130 D.J. Walder et al
Hennessy MB, Mendoza SP, Mason WA, Moberg GP (1995):
Endocrine sensitivity to novelty in squirrel monkeys and titi
monkeys: Species differences in characteristic modes of
responding to the environment. Physiol Behav 57:331–338.
Hirayasu Y, Shenton ME, Salisbury DF, Dickey CC, Fischer IA,
Mazzoni P, et al (1998): Lower left temporal lobe MRI
volumes in patients with first-episode schizophrenia com-
pared with psychotic patients with first-episode affective
disorder and normal subjects. Am J Psychiatry 155:1384–
Hoff AL, Riordan H, O’Donnell DW, Morris L, DeLisi LE
(1992): Neuropsychological functioning of first-episode
schizophreniform patients. Am J Psychiatry 149:898–903.
Jeste DV, Lohr JB (1989): Hippocampal pathologic findings in
schizophrenia. A morphometric study. Arch Gen Psychiatry
Kaneko M, Yokoyama F, Hoshino Y, Takahagu K, Murata S,
Watanabe M, et al (1992): Hypothalamic pituitary adrenal
axis function in chronic schizophrenia: Association with
clinical features. Neuropsychobiology 25:1–7.
Keshavan MS, Brar J, Ganguli R, Jarrett D (1989): DST and
schizophrenic symptomatology. Biol Psychiatry 26:847–858.
Kiess W, Meidert A, Dressendorfer RA, Schriever K, Kessler U,
Konig A, et al (1995): Salivary cortisol levels throughout
childhood and adolescence: Relation with age, pubertal stage
and weight. Pediatr Res 37:502–506.
Kirschbaum C, Hellhammer DH (1989): Salivary cortisol in
psychobiological research: An overview. Neuropsychobiol-
Kirschbaum C, Wolf OT, May M, Wippich W, Hellhammer DH
(1996): Stress- and treatment-induced elevations of cortisol
levels associated with impaired declarative memory in
healthy adults. Life Sci 58:1475–1483.
Kovelman JA, Scheibel AB (1984): A neurohistological correlate
of schizophrenia. Biol Psychiatry 19:1601–1621.
Laudat MH, Cerdas S, Fournier C, Guiban D, Guilhaume B,
Luton JP (1988): Salivary cortisol measurement: A practical
approach to assess pituitary-adrenal function. J Clin Endocri-
nol Metab 66:343–348.
Lencz T, McCarthy G, Bronen RA, Scott TM, Inserni JA, Sass
KJ, et al (1992): Quantitative magnetic resonance imaging in
temporal lobe epilepsy: Relationship to neuropathology and
neuropsychological function. Ann Neurol 31:629–637.
Lupien S, Lecours AR, Lussier I, Schwartz G, Nair NPV,
Meaney MJ (1994): Basal cortisol levels and cognitive
deficits in human aging. J Neurosci 14:2893–2903.
Lupien SJ, Gaudreau S, Tchiteya BM, Maheu F, Sharma S, Nair
NP, et al (1997): Stress-induced declarative memory impair-
ment in healthy elderly subjects: Relationship to cortisol
reactivity. J Clin Endocrinol Metab 82:2070–2075.
McNicol D, Willson RJ (1971): The application of signal
detection theory to letter recognition. Aust J Psychol 23:311–
Nelson HE (1976): A modified card sort test sensitive to frontal
lobe defects. Cortex 12:313–324.
Nestor PG, Shenton ME, McCarley RW, Haimson J, Smith RS,
O’Donnell B, et al (1993): Neuropsychological correlates of
MRI temporal lobe abnormalities in schizophrenia. Am J
Newcomer JW, Craft S, Askins K, Hershey T, Bardgett ME,
Csernansky JG, et al (1998): Glucocorticoid interactions with
memory function in schizophrenia. Psychoneuroendocrinol-
Newcomer JW, Craft S, Hershey T, Askins K, Bardgett J (1994):
Glucocorticoid-induced impairment in declarative memory
performance in adult humans. J Neurosci 14:2047–2053.
Newcomer JW, Faustman WO, Whiteford HA, Moses JA Jr,
Csernansky JG (1991): Symptomatology and cognitive im-
pairment associate independently with post-dexamethasone
cortisol concentrations in unmedicated schizophrenic pa-
tients. Biol Psychiatry 29:855–864.
Newcomer JW, Selke G, Melson AK, Hershey T, Craft S,
Richards K, Alderson AL (1999): Decreased memory perfor-
mance in healthy humans induced by stress-level cortisol
treatment. Arch Gen Psychiatry 56:527–533.
Rao VP, Krishnan RR, Goli V, Saunders WB, Ellinwood EH,
Blazer DG, et al (1989): Neuroanatomical changes and
hypothalamic-pituitary-adrenal axis abnormalities. Biol Psy-
Rybakowski J, Linka M, Matkowski K, Kanarkowski R (1991):
Dexamethasone suppression test and the positive and nega-
tive symptoms of schizophrenia. Psychiatr Pol 25:9–15.
Sapolsky RM, McEwen BS (1986): Stress, glucocorticoids, and
their role in degenerative changes in the aging hippocampus.
In: Crook T, Bartus R, Ferris S, Gershon S, editors. Treatment
Development Strategies for Alzheimer’s Disease. Madison,
CT: Mark Powley, 151–171.
Sapolsky RM, Uno H, Rebert CS, Finch CE (1990): Hippocam-
pal damage associated with prolonged glucocorticoid expo-
sure in primates. J Neurosci 10:2897–2902.
Saykin AJ, Shtasel DL, Gur RE, Kester DB, Mozley LH,
Stafiniak P, et al (1994): Neuropsychological deficits in
neuroleptic naı ¨ve patients with first-episode schizophrenia.
Arch Gen Psychiatry 51:124–131.
Schatzberg AF, Rothschild AJ (1988): The roles of glucocorti-
coid and dopaminergic systems in delusional (psychotic)
depression. Ann N Y Acad Sci 537:462–471.
Schatzberg AF, Rothschild AJ, Langlais PJ, Bird ED, Cole JO
(1985): A corticosteroid/dopamine hypothesis for psychotic
depression and related states. J Psychiatry Res 19:57–64.
Schmand B, Brand N, Kuipers T (1992): Procedural learning of
cognitive and motor skills in psychotic patients. Schizophr
Seckl JR, Campbell JC, Edwards CRW, Christie JE, Whalley LJ,
Goodwin GM, Fink G (1991): Diurnal variation of plasma
corticosterone in depression. Psychoneuroendocrinology 15:
Seeman TE, McEwen BS, Singer BH, Albert MS, Rowe JW
(1997): Increase in urinary cortisol excretion and memory
declines: MacArthur studies of successful aging. J Clin
Endocrinol Metab 82:2458–2465.
Sharma RP, Pandey GN, Janicak PG, Peterson J, Comaty JE,
Davis JM (1988): The effect of diagnosis and age on the DST:
A meta-analytic approach. Biol Psychiatry 24:555–568.
Shipley JE, Alessi NE, Wade SE, Haegle AD, Helmbold B
Cognition, Cortisol, and Schizophrenia 1131
(1992): Utility of an oral diffusion sink (ODS) device for
quantification of saliva corticosteroids in human subjects.
J Clin Endocrinol Metab 74:698–700.
Squire LR (1987): Memory and Brain. New York: Oxford
Squire LR (1992): Memory and the hippocampus. A synthesis
from findings with rats, monkeys, and humans. Psychol Rev
Starkman MN, Gebarski SS, Berent S, Schteingart DE (1992):
Hippocampal formation volume, memory dysfunction, and
cortisol levels in patients with Cushing’s syndrome. Biol
Suddath RL, Christison GW, Torrey EF, Casanova MF, Wein-
berger DR (1990): Anatomical abnormalities in the brains of
monozygotic twins discordant for schizophrenia. N Engl
J Med 322:789–794.
Sullivan EV, Shear PK, Zipursky RB, Sagar HJ, Pfefferbaum A
(1994): A deficit profile of executive, memory and motor
functions in schizophrenia. Biol Psychiatry 36:641–653.
Tandon R, Mazzara C, DeQuardo J, Craig KA, Meador-Woo-
druff JH, Goldman R (1991): Dexamethasone suppression
test in schizophrenia: Relationship to symptomatology, ven-
tricular enlargement, and outcome. Biol Psychiatry 29:953–
Torres IJ, Flashman LA, O’Leary DS, Swayze V II, Andreasen
NC (1997): Lack of an association between delayed memory
and hippocampal and temporal lobe size in patients with
schizophrenia and healthy controls. Biol Psychiatry 42:1087–
Velakoulis D, Pantelis C, McGorry PD, Dudgeon P, Brewer W,
Cook M, et al (1999): Hippocampal volume in first-episode
psychoses and chronic schizophrenia: A high-resolution mag-
netic resonance imaging study. Arch Gen Psychiatry 56:133–
Waldo MC, Cawthra E, Adler LE, Dubester S, Staunton M,
Nagamoto H, et al (1994): Auditory sensory gating, hip-
pocampal volume, and catecholamine metaboism in schizo-
phrenics and their siblings. Schizophr Res 12:93–106.
Walker EF, Diforio D (1997): Schizophrenia: A neural diathesis-
stress model. Psychol Rev 104:667–685.
Wechsler D (1987): Wechsler Memory Scale—Revised. New
York: Psychological Corp.
Weinstein DD, Diforio D, Schiffman J, Walker E, Bonsall R
(1999): Minor physical anomalies, dermatoglyphic asymme-
tries, and cortisol levels in adolescents with schizotypal
personality disorder. Am J Psychiatry 156:617–623.
Appendix 1. Items in the SANS-SAPS–Derived Symptom Scales
Item no. Criteria
7 Appropriateness of emotional expressions
39 Illogical speech
42 Distractible speech
44Clothing and appearance
Thoughts of being persecuted by other people
Thoughts of sin or guilt
Thoughts concerning the body
Ideas and thoughts of self-reference
Thoughts of being controlled
Thoughts of mind reading
Hallucinatorylike experiences involving noises or voices
Hallucinatorylike experiences concerning the body
Hallucinatorylike experiences involving odors/sense of smell
Visual hallucinatorylike experiences
Facial expressions of emotion
Spontaneous movements of arms and legs
Emotional responsivity to other people or events
Grooming and hygiene
Repetitive or stereotyped behavior
SANS, Schedule for Assessment of Negative Symptoms; SAPS, Schedule for
Assessment of Positive Symptoms.
1132 D.J. Walder et al