INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY

Int J Geriatr Psychiatry 2003; 18: 820–829.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/gps.929

Comparative effects of risperidone and olanzapine

on cognition in elderly patients with schizophrenia
or schizoaffective disorder
Philip D. Harvey1*, Judy A. Napolitano2, Lian Mao2, and Georges Gharabawi2
1
Department of Psychiatry, Mount Sinai School of Medicine, New York, USA
2
Janssen Pharmaceutica, Inc., Titusville, NJ, USA

SUMMARY
Objective To examine the effects of risperidone and olanzapine on cognitive functioning in elderly patients with schizo-
phrenia or schizoaffective disorder.
Method One hundred seventy-six elderly inpatients and outpatients with schizophrenia or schizoaffective disorder were
enrolled in this multicenter, double-blind trial. After their antipsychotic medications were tapered for 1 week, patients were
randomly assigned to receive either risperidone 1 to 3 mg/day or olanzapine 5 to 20 mg/day for 8 weeks. Performance on the
Continuous Performance Test (CPT), Serial Verbal Learning Test (SVLT), TMT (Trail Making Test) Parts A and B, Wis-
consin Card Sorting Test (WCST), and Verbal Fluency Examinations (VFE) was assessed at baseline and at end point.
Results Patients in the risperidone group had improved scores on at least one test of attention, memory, executive function,
and verbal fluency, and those in the olanzapine group had improved scores on at least one test of attention and memory
function. Scores on the TMT Part B, WCST total errors (executive function domain), and the VFE improved significantly
from baseline in the risperidone group but not in the olanzapine group. No significant differences in change scores between
the two groups were found. Higher baseline scores on each test predicted more improvement at endpoint.
Conclusions Low doses of risperidone and olanzapine improve cognitive function in elderly patients with schizophrenia
or schizoaffective disorder. Consistent with research in younger populations, these improvements occur in aspects of cog-
nitive functioning related to functional outcome. Copyright # 2003 John Wiley & Sons, Ltd.

key words — risperidone; olanzapine; cognition; schizophrenia; schizoaffective disorder; elderly

INTRODUCTION suggested that even patients with a history of chronic

Recent research has challenged many of the long-held
assumptions about schizophrenia in old age. Evidence
suggests that the overall course and prognosis of schi-
zophrenia changed little between the 1890s and the
1980s (Hegarty et al., 1994). Clinical observations
during the 20th century initially led to the widespread
belief that symptoms of schizophrenia were attenu-
ated in late life (Bridge et al., 1978). Some research
* Correspondence to: Dr P. D. Harvey, Department of Psychiatry,
Box 1229, Mount Sinai School of Medicine, New York, NY 10029,
USA. Tel: 212-659-8713. Fax: 212-860-3945.
E-mail: Philipdharvey1@cs.com
Contract/grant sponsor: Janssen Pharmaceutica Products, L.P.
Copyright # 2003 John Wiley & Sons, Ltd.
institutionalization subsequently were able to live
in the community with minimal adaptive deficits
(Harding et al., 1987).
Recent studies paint a markedly different picture
with respect to the severity of clinical symptoms
and functional capacity in many older patients with
schizophrenia. As these studies have shown, older
institutionalized patients in late life have positive
and negative symptoms that are consistent with the
severity of symptoms in younger institutionalized
patients (Davidson et al., 1995), as well as marked
deficits in cognitive functioning as severe as those
in patients with a confirmed diagnosis of degenerative
dementia (Davidson et al., 1996). Even older schizo-
phrenic patients with a lifetime history of better func-
tional outcomes have severe deficits in cognitive
Received 6 January 2003
Accepted 10 June 2003
 effects of risperidone and olanzapine on cognition
functioning (Eyler Zorrilla et al., 2000; Palmer et al.,
1997) and functional skills (Klapow et al., 1997) in
later life, compared with older healthy individuals.
Studies of younger schizophrenic patients suggest
that deficits in functional skills are best predicted by
impairment in cognitive domains such as memory,
attention, and executive function (Green, 1996; Green
et al., 2000), even when the severity of both negative
and positive symptoms is considered. Studies of older
patients with schizophrenia have demonstrated that
the severity of deficits in these specific cognitive
domains is similar to that of impairments seen in
younger patients (Heaton et al., 1994) and that these
impairments are also related to specific deficits in
functional skills (Harvey et al., 1997, 1998). This
relationship between cognitive and functional deficits
in older patients is consistent across a wide range of
lifetime functional outcomes, with similar relation-
ships occurring in patients with a lifetime history of
chronic institutional care and in patients who have
been ambulatory their entire lives (Harvey et al.,
1998). These correlations are found even in older
schizophrenic patients with severe cognitive impair-
ments (Bowie et al., 2002).
Because many older patients with schizophrenia
are now required to live in the community, deficits
in self-care, instrumental living skills, and social skills
are important targets for intervention (Auslander et al.,
2001; Patterson et al., 2001a,b). The atypical antipsy-
chotics risperidone and olanzapine have been shown
in open-label studies to reduce both positive and
negative symptoms in elderly patients to a greater
extent than do conventional antipsychotics and also
to improve cognitive deficits (Davidson et al., 2000;
Jeste et al., 1996; Madhusoodanan et al., 2000;
Berman et al., 1996). Further, both of these medica-
tions have been shown in a recent double-blind study
to enhance cognitive functioning in younger schizo-
phrenic patients to a relatively equivalent extent (Har-
vey et al., in press).
This report details the effects of risperidone and
olanzapine on cognitive function in elderly patients
with schizophrenia or schizoaffective disorder. This
is the first randomized, double-blind study of cogni-
tive effects of newer antipsychotics in elderly patients
with schizophrenia or schizoaffective disorder. We
addressed several questions in this study, based on
previous studies of younger patients.
1. We examined the extent that cognitive functioning
was enhanced compared to baseline assessments
by treatment with newer antipsychotic medica-
tions.
Copyright # 2003 John Wiley & Sons, Ltd.
821
2. We tested for differential effects of risperidone and
olanzapine on enhancing cognition.
3. We determined whether changes in cognitive
functioning were due to improvements in clinical
symptoms or side effects.
METHODS
Subjects
Patients older than age 60 with a DSM-IV diagnosis
of schizophrenia or schizoaffective disorder were
enrolled in this study. Inpatients, outpatients, nursing
home residents, and board and care patients with
Positive and Negative Syndrome Scale (PANSS)
baseline scores between 50 and 120 were eligible
(Kay et al., 1987a; Kay and Opler, 1987).
Exclusion criteria included a DSM-IV Axis I diag-
nosis other than schizophrenia or schizoaffective
disorder, a major depressive episode according to
DSM-IV criteria within 6 months, or DSM-IV diagno-
sis of substance abuse or dependence in the previous
3 months. Additional exclusion criteria included
Mini-Mental State Examination (MMSE) (Folstein
et al., 1975) score of less than 18, QTc interval greater
than 500 milliseconds, a history of clozapine therapy
for treatment-refractory schizophrenia for more than
4 consecutive weeks or within 28 days before screen-
ing, and treatment with a depot antipsychotic within
two treatment cycles of the beginning of the washout
period.
The study was approved by the appropriate institu-
tional review boards at every research site, and all
patients or their legal guardians provided written
informed consent before patients entered the study.
Study design
This was a double-blind, randomized, parallel,
8-week study. After a 1-week dosage-tapering period,
during which antipsychotic medications were gradu-
ally discontinued, 176 patients were randomly as-
signed to receive either risperidone (n ¼ 87) or
olanzapine (n ¼ 89) for 8 weeks on a flexible-dose
schedule. All other psychotropic medications were
discontinued as well. Baseline medications received
are presented in Table 1. Study medications were
administered once daily and the doses for each patient
were gradually increased by the investigator to bal-
ance efficacy and extrapyramidal symptoms. The
dosage schedule for risperidone called for 1 mg/day
on days 1 through 3, 1 to 2 mg/day on days 4 through
7, and 1 to 3 mg/day on days 8 through 56. The
Int J Geriatr Psychiatry 2003; 18: 820–829.
822 p. d. harvey et al.

Table 1. Baseline Characteristics of the Patients

Risperidone Olanzapine Test statistic
(n ¼ 74) (n ¼ 79)

Gender (% Male) 38 34 X2(1) ¼ 0.22, p ¼ 0.64

Ethnicity (%)
White 55 65 X2(3) ¼ 3.47, p ¼ 0.32
Black 12 12
Hispanic 4 1
Asian 3 1
Diagnosis (%)
Schizophrenia 82 85 X2(1) ¼ 0.16, p ¼ 0.69
Schizoaffective 18 15

M SD M SD

Age 71.22 6.03 71.39 5.85 t(151) ¼ 0.18, p ¼ 0.86

Age first admission 36.15 14.85 33.93 14.17 t(151) ¼
0.93, p ¼ 0.35
N Previous Admit 5.30 8.60 5.99 8.76 t(151) ¼ 0.47, p ¼ 0.64
PANSS
Total 77.32 14.16 77.75 15.66 t(151) ¼ 0.17, p ¼ 0.86
Positive 21.28 5.85 22.20 6.42 t(151) ¼ 0.42, p ¼ 0.67
Negative 21.70 7.60 21.49 7.45 t(151) ¼
0.17, p ¼ 0.87
MMSE 25.64 3.57 24.92 3.37 t(151) ¼
1.27, p ¼ 0.21
BQoL 4.64 1.55 4.68 1.34 t(151) ¼ 0.14, p ¼ 0.89
HAM-D 7.35 5.61 7.96 6.16 t(151) ¼ 0.64, p ¼ 0.53
ESRS 10.65 10.75 12.10 11.06 t(151) ¼ 0.82, p ¼ 0.41

PANSS ¼ Positive and Negative Syndrome Scale; MMSE ¼ Mini-Mental State Examination; BQoLI ¼ Brief Quality of Life Interview;
HAM-D ¼ Hamilton Depression Scale; ESRS ¼ Extrapyramidal Symptom Rating Scale.

schedule for olanzapine called for 5 mg/day on days 1
through 7, 5 to 10 mg/day on days 8 through 14, 5 to
15 mg/day on days 15 through 21, and 5 to 20 mg/day
on days 22 through 56. Both groups received open-
label risperidone during the tapering period from days
57 through 65.
response sensitivity for discrimination of target and
non-target stimuli.
Trail Making Test (TMT) Part A. In this test of
visuomotor speed (Spreen and Strauss, 1998), the
time required to complete the test is the dependent
variable.

Efficacy assessments Memory domain

Efficacy was assessed in all patients who entered the
randomization phase, had received at least one dose of
study medication, and had at least one post-baseline
cognitive assessment. Efficacy variables included
change from baseline in scores on cognitive tests of
attention, memory, executive function, and verbal flu-
ency. The following cognitive tests were performed at
baseline (randomization) and at 4 and 8 weeks of
treatment (or at early termination).
Attention domain
Continuous Performance Test (CPT). In this test of
vigilance (Cornblatt et al., 1989), patients are asked to
press a computer key whenever the same four-digit
target stimulus occurs twice in a row. The dependent
variable is the signal detection index, d’, which is the
Serial Verbal Learning Test (SVLT). This test was
developed to evaluate patients with Alzheimer’s
disease (Morris et al., 1989) and has been used
extensively in elderly patients with schizophrenia
(e.g. Harvey et al., 1998). A ten-item list of words is
presented to the patient in each of three separate
learning trials. A delayed free-recall test was
performed after the TMT. Dependent variables are
the total number of words recalled in the three
learning trials and at delayed recall.
Executive function domain
Wisconsin Card Sorting Test (WCST). This test
assesses executive functioning, cognitive flexibility,
maintenance of a cognitive set, and working memory
(Heaton et al., 1993). The critical dependent variables

Copyright # 2003 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2003; 18: 820–829.
 effects of risperidone and olanzapine on cognition
for this trial are the number of categories completed
and the total number of errors.
Trail Making Test (TMT) Part B. This second part of
the TMT evaluates both visuomotor speed and the
ability to alternate between sets (Spreen and Strauss,
1998). The dependent variable is the time required to
complete the task.
Verbal fluency examination
Two verbal fluency tests were administered: category
and phonologic fluency (Lezak, 1997). Total scores
for category and letter fluency were the dependent
variables.
Assessment methods
All tests were performed by testers who received
training at each site, at investigator meetings, or via
a videotaped demonstration. Case record forms were
inspected by the first author and all scores that seemed
questionable were queried and resolved before the
data were entered into the final database, which was
before any analyses, particularly nonblinded ones,
were performed.
Patients were enrolled at 19 sites in the United
States and 13 sites in Austria, Israel, Norway, Poland,
and the Netherlands. After following the procedures
we used in a previous multinational clinical trial
(Harvey et al., 2003), we translated all tests and
instructions into the local languages. A senior neurop-
sychologist was identified as the local consultant in
each country where translations of the English ver-
sions of the tests were required. This consultant pro-
vided information on several critical aspects of the
cognitive assessment, including the verbal fluency let-
ters typically used and any suggested variations in the
words used for the SVLT. The English-language clin-
ical trial manual was written by the first author and
specific instructions provided for the tests by their
publisher were translated by a professional translator
into the local language, edited and reviewed by the
local neuropsychologist for accuracy, and then trans-
lated back into English by a different translator.
Dr Harvey then reviewed these back-translations.
A similar process was applied to the translations
and back-translations of the test stimuli, after any sug-
gested adjustments by the local consultant. The local
experts provided the letters typically used for phono-
logical fluency examinations and altered the words
used in the SVLT when necessary. In our previous
research on a large sample of patients with first-epi-
sode schizophrenia, we did not observe any perfor-
Copyright # 2003 John Wiley & Sons, Ltd.
823
mance differences associated with translation of the
tests and stimulus materials (Harvey et al., 2003).
Assessments of noncognitive variables
Total PANSS and subscale scores, including positive
and negative symptoms (Kay and Opler, 1987b), were
rated at screening, at randomization, and at weekly
intervals over the 8-week treatment protocol. Other
evaluations included the Hamilton Depression Scale
(HAM-D) (Hamilton, 1960) and the Brief Quality of
Life Interview (BQoLI) (Lehman, 1988), obtained at
randomization and at end point, and MMSE scores,
evaluated at screening, randomization, and at end
point. Extrapyramidal Symptom Rating Scale (ESRS)
scores (Chouinard et al., 1980) were obtained at ran-
domization and week 8 as part of the safety analysis.
Results of these assessments are being reported else-
where, but these variables were used in some correla-
tional analyses described below. All data for
correlational analyses were collected on the day of
the endpoint assessment.
Data analysis
Two-tailed tests were used for all analyses, and the
results were interpreted at the 5% significance level.
T-tests were used to compare the demographic and
baseline characteristics data. Multivariate analysis
of Covariance (MANCOVA) was used to compare
the change sores from all cognitive tests, incorporat-
ing factors of baseline score and site. We examined
patients for their improvement in performance at their
final cognitive assessment (LOCF) compared with
baseline in order to answer our first question. The
Cochran-Mantel-Haenszel test was also used to
examine the differences in the distribution of patients
who had markedly improved (>1 SD), substantially
improved (>0.5 to 1 SD), slightly improved (0.01 to
0.5 SD), or not improved at their final assessment
(40 SD) relative to their baseline performance. We
compared the two treatment groups for their changes
in performance over time in order to answer our sec-
ond question. The Spearman rank-order correlation
coefficient was used to assess the relationship
between the changes in cognitive scores from baseline
to end point and the baseline and change scores in the
PANSS, ESRS, and HAM-D tests and sex, age at first
psychiatric admission, age at study entry, and the
number of previous admissions in order to answer
our third question.
Finally, follow-up analyses were performed to
compare the patients who did and did not complete
Int J Geriatr Psychiatry 2003; 18: 820–829.
824 p. d. harvey et al.

post baseline cognitive assessments on all of the base-
line variables. Analyses of time trends were per-
formed in order to examine whether the cognitive
test performance of subjects who received two post
baseline assessments differed at the endpoint assess-
ment from those who were examined only once.
RESULTS
Demographic and baseline treatment
characteristics
Of the 176 patients who entered the randomization
phase, 175 patients received at least one dose of the
study drug. Of the 175 patients, 153 completed one
or more post-baseline cognitive tests, of whom 74
were in the risperidone group and 79 were in the olan-
zapine group (Table 1). Of the 74 risperidone patients,
54 completed the protocol (and received two cogni-
tive assessments) and 20 received their final assess-
ment before 8 weeks. The numbers for olanzapine
were 55 completers and 24 early terminations respec-
tively. As can be seen in the table, there were no dif-
ferences in baseline symptomatic or demographic
factors. Baseline medications were similar between
the groups as well. Only one of the patients rando-
mized to risperidone was receiving an anticholinergic
medication, while none of the olanzapine patients
were receiving anticholinergics. Similarly, no patients
were receiving standing benzodiazepines at baselines.
In terms of antipsychotic medications, 60% of the
patients were receiving one or more conventional
antipsychotic medications, while 40% were already
receiving newer antipsychotics. Of the patients rando-
mized to receive risperidone, 25% were already
receiving risperidone at baseline, while 20% were
receiving olanzapine. For patients randomized to
olanzapine, 11% were receiving olanzapine at base-
line, while 20% were receiving risperidone. These
distributions were not significantly different from
each other (X2(1) ¼ 1.05, p ¼ 0.58).
Baseline scores on the cognitive assessments did
not differ significantly ( p > 0.05) between treatment
groups (Table 2) based on a MANOVA. The overall
effects of site were statistically significant ( p > 0.05),
so this covariate was preserved for later analyses.
Mean risperidone dose at the end point assessment
was 1.95, SD ¼ 0.85 (range ¼ 1–3), while the olanza-
pine patients received a mean dose of 11.46 mg,
SD ¼ 5.44 (range ¼ 5–20). We compared the baseline
cognitive scores of the patients who did and did not
complete any of the post-baseline cognitive assess-
ments with a multivariate analysis of variance. The
baseline cognitive scores of the patients did not differ
as a function of whether any follow-up analyses were
performed, F(11, 284) ¼ 0.23, p ¼ 0.88.
Treatment effects
Changes from baseline. In the attention domain,
scores on the TMT Part A improved significantly
from baseline in both groups; however, there was no
significant change from baseline in CPT d’ total
scores for either treatment group. In the memory
domain, scores on the SVLT improved significantly in
both groups. For executive function, only risperidone
was associated with a significant improvement in
TMT Part B scores and WCST total error scores. The

Table 2. Cognitive performance at baseline and cognitive changes from baseline

Variable Risperidone Olanzapine

Baseline Change from Baseline Change from

baseline baseline

N M SD M SD t p N M SD M SD t p

Attention domain
CPT d’ 54 0.21 0.54 0.0 0.51 0.00 0.99 50 0.24 0.52 0.14 0.72 1.11 0.22
Trail-making Part A 71 139.14 89.6
20.71 52.54 2.80 0.007 76 155.13 84.36
20.12 21.10 6.94 0.001
Memory domain
Serial verbal test, 72 13.32 5.24 2.73 4.70 4.82 0.001 75 12.35 4.68 2.39 4.01 5.11 0.001
Total learning
Delayed recall 72 4.02 2.23 1.13 1.89 5.00 0.001 75 3.62 2.69 1.02 1.58 5.26 0.001
Executive domain
Trail-making Part B 69 258.19 64.76
14.74 52.23 2.35 0.021 72 258.44 67.67
6.33 59.89 0.89 0.379
WCST categories 63 1.62 1.69 0.1 1.4 0.56 0.656 63 1.19 1.34 0.00 0.78 0.00 0.999
WCST total errors 63 68.1 21.27
4.94 16.49 2.80 0.023 63 71.45 18.21
1.01 18.85 0.93 0.673
Verbal fluency 71 38.08 17.29 2.92 9.94 3.31 0.002 75 36.02 17.48 2.14 9.68 1.86 0.070
Total production

Copyright # 2003 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2003; 18: 820–829.
 effects of risperidone and olanzapine on cognition
WCST categories score did not change significantly
in either group. The mean score on the VFE sum of
categories and letters also improved significantly
from baseline in the risperidone group but not in the
olanzapine group. No difference between treatment
groups in change scores from baseline was demon-
strated at end point for any of the cognitive variables
(Table 2).
For all tests but the TMT Part A, the overall distri-
bution of patients who markedly improved (>1 SD),
substantially improved (>0.5 to 1.0 SD), slightly
improved (0.01 to 0.5 SD), or not improved (40
SD) was similar in both treatment groups. In the
TMT Part A only, there was a significant difference
( p < 0.006) in the distribution of overall change
scores. Substantial or marked improvement occurred
in 25.0% of patients in the olanzapine group and
18.3% of those in the risperidone group, whereas no
improvement occurred in 52.6% of patients in the
olanzapine group and 33.8% of those in the risperi-
done group. Thus, in the olanzapine group, there were
not only slightly more patients who had greater
improvement but also more patients who did not
improve.
In the next analyses, a MANOVA approach was
applied to examine differences in change scores from
baseline as a function of medication treatment. This
analysis was performed for the entire sample at end-
point, as well as patients who completed the protocol.
The endpoint analysis revealed no statistically signif-
icant differences between the groups, Wilks
lambda ¼ 0.90, Pillais Approx. F (11, 84) ¼ 0.84,
p ¼ 0.60. Similar results were obtained when protocol
completers were examined, Wilks lambda ¼ 0.87,
Pillais approx. F (11, 52) ¼ 0.72, p ¼ 0.71. In order
to examine whether there was continued improvement
in patients who completed the protocol vs those who
did not, endpoint scores were compared in a
group (completer, noncompleter)  medication group
MANOVA. There were no statistically significant
differences in endpoint cognitive scores associated
with completer status F(11, 84) ¼ 0.49, p ¼ 0.84 and
no significant interaction of completer status  med-
ication group F(11, 84) ¼ 0.23, p ¼ 0.91.
In the next analyses, baseline values for all of the
demographic and other clinical and cognitive vari-
ables, as well as changes in the PANSS and ESRS
scores, were related to the cognitive change scores
with Spearman correlations, limiting the analysis to
those cognitive variables for which there was signifi-
cant improvement from baseline. For risperidone-
treated patients baseline scores predicted the extent
of improvement on each test for all variables other
Copyright # 2003 John Wiley & Sons, Ltd.
825
than WCST total errors. For olanzapine patients, there
were fewer predictors of change, while baseline
scores for SVLT delayed recall, WCST total errors,
and TMT Part B predicted changes on that variable
for olanzapine. In all cases where baseline scores
were significant predictors, higher baseline scores
predicted greater improvement.
In both treatment groups, the change from baseline
to end point in cognition scores on the tests in which
there was significant change from baseline demon-
strated only modest correlations with changes from
baseline in other variables. In the olanzapine group,
a positive correlation was demonstrated between the
change in the ESRS akathisia score from baseline
and the change in scores on TMT Part B. In addition,
there were positive correlations in both groups
between improvements in negative and total PANSS
scores and scores on the VFE. The improvement in
HAM-D scores related to improvement in the VFE
in the risperidone group. No correlations between
improvements in the WCST or the SVLT and changes
in clinical symptoms or side effects were significant
in either group.
DISCUSSION
This study demonstrated that performance on tests of
attention and memory improved in elderly patients
with long-term schizophrenia or schizoaffective dis-
order after 4 to 8 weeks of treatment with low doses
of risperidone or olanzapine. In addition, scores on
tests of executive function and verbal fluency also
improved in patients treated with risperidone. There
were no variables on which the groups differed signif-
icantly in the average improvements from baseline.
These patients, despite having years of illness, sub-
stantial baseline cognitive deficits, and an extensive
history of conventional antipsychotic therapy, demon-
strated significant improvements in multiple domains
of cognition previously shown to be outcome-related.
These changes were statistically significant even
when the diversity of patients across sites and nations
was considered. Finally, these changes in cognitive
functioning were not caused by improvements in clin-
ical symptoms or side effects (other than one variable
in one of the treatment groups) or the previous use of
anticholinergic medications.
Before evaluating the significance of these findings,
some limitations should be reviewed. There was no
conventional antipsychotic treatment group in this
study. While this may be viewed as a limitation,
both the larger literature and several recent clinical
trials indicate that conventional medications lead to
Int J Geriatr Psychiatry 2003; 18: 820–829.
826 p. d. harvey et al.
essentially no improvement in cognitive functioning.
For instance, a recent study comparing the effects of
conventional medications to novel treatments in treat-
ment refractory younger patients (Bilder et al., 2002)
found that both risperidone and olanzapine improved
cognitive functioning more than haloperidol treat-
ment, which had no beneficial effect. Harvey et al.
(2000) reported that even extensive practice on atten-
tional tests failed to produce significant improve-
ments in performance when patients were treated
with haloperidol, while improvements associated
with risperidone treatment were noted within one
week. Harvey et al. (2002) examined 45 older
patients with schizophrenia with an extensive cogni-
tive assessment at an eight-week retest interval. None
of 22 different cognitive tests changed with retesting
over this very similar follow-up period. Finally, Blyler
and Gold (2000) analyzed the overall research litera-
ture and concluded that treatment with conventional
medication may even interfere with the ability to
improve in cognitive functioning. Thus the risks asso-
ciated with conventional antipsychotic medications in
older patients may outweigh the scientific benefits of
including such a comparison sample in a contempor-
ary clinical trial. At the same time, the lack of this
comparator does not allow firm conclusions regarding
relative benefits of these newer medications to con-
ventional treatments in the elderly.
The findings of this study confirm previous studies
of risperidone and olanzapine involving relatively
younger patients with a markedly shorter history of
illness. Improvement on tests of attention, executive
function, memory, and verbal fluency after risperi-
done treatment has been demonstrated in studies of
both elderly (Davidson et al., 2000) and younger adult
schizophrenic populations (Green et al., 1997; Rossi
et al., 1997; Purdon et al., 2000; Harvey et al.,
in press). Previous studies of olanzapine have demon-
strated cognitive improvements in younger schizo-
phrenic populations (Purdon et al., 2000; Cuesta
et al., 2001) and improved MMSE scores in elderly
psychotic patients (Madhusoodanan et al., 2000).
Improvements in cognitive function have been
demonstrated in younger patients with schizophrenia
as early as 4 weeks with risperidone (Green et al.,
1997) and 6 weeks with olanzapine (Purdon et al.,
2000). Further, in our study, the proportion of patients
with schizophrenia who had improvements of 0.5 SD
or more in their memory, motor speed, and executive
functioning was actually greater than that in a pre-
vious and similar double-blind clinical trial compar-
ing the effects of risperidone and olanzapine
(Harvey et al., in press).
Copyright # 2003 John Wiley & Sons, Ltd.
There are some inconsistencies in our findings with
the results of previous studies. For instance, Meltzer
and McGurk (1999) reported that verbal fluency
improved with olanzapine treatment but not with ris-
peridone and that working memory improved with
risperidone treatment but not with olanzapine. We
did not find differential effects of the two medica-
tions, noting also that the Meltzer and McGurk paper
reported the results of different open-label switch stu-
dies to evaluate the effects of these two medications.
Similarly, Purdon et al. (2000) reported a substantial
benefit of olanzapine compared to both risperidone
and haloperidol. The results of the present study are
consistent with other studies of younger patients, both
refractory and nonrefractory, that no substantial dif-
ferences between the medications (Harvey et al., in
press; Bilder et al., 2002).
The average doses in this study were approximately
equal to those recommended by the Health Care
Financing Association (Gurvich and Cunningham,
2000) for elderly patients with psychosis (2 and
10 mg/day for risperidone and olanzapine, respec-
tively), suggesting that these results would be general-
izable to the typical older patient treated with these
medications.
Because the components of cognitive functioning
that improved after treatment in our study are all
aspects of cognitive functioning previously reported
to be associated with functional outcome in schizo-
phrenia (Green, 1996, Green et al., 2000), we
hypothesize that changes in functional status in this
population might be expected after extended treat-
ment with newer medications. Because of the dura-
tion of the study, this possibility remains to be
examined in later studies. Our results might be influ-
enced by the patients’ previous high doses of older
antipsychotics, which might affect the ability of
newer antipsychotic medications to enhance cognitive
function. Although patients completed a 7-day taper-
ing period, some residual effects of previous medica-
tions might have been present at baseline (Harvey and
Keefe, 2001). Also, because this was a multinational
trial in which patients were recruited in countries with
varying educational standards, educational data were
not considered for these patients, and it is unknown
whether this population was typical with respect to
educational levels in their native countries.
In summary, this study provides further evidence
that the atypical antipsychotics risperidone and olan-
zapine enhance cognition in elderly patients with
schizophrenia. This effect occurs as early as 8 weeks
and is particularly evident among patients with the
highest levels of cognitive functioning at initiation
Int J Geriatr Psychiatry 2003; 18: 820–829.
 effects of risperidone and olanzapine on cognition
of therapy. Changes are not due to improvements
in clinical symptoms or side effects and were not
influenced by anticholinergic treatments. Longer-
term studies are necessary to evaluate the beneficial
effects of these agents on functional status in this
population.
ACKNOWLEDGEMENT
This study was funded by Janssen Pharmaceutica
Products, L.P.
REFERENCES
Auslander LA, Lindamer LL, Delapena J, et al. 2001. A comparison
of community-dwelling older schizophrenia patients by residen-
tial status. Acta Psychiatr Scand 103: 380–386.
Berman I, Merson A, Rachov-Pavlov J, Allan E, Davidson M,
Losonczy MF. 1996. Risperidone in elderly schizophrenic
patients. Am J Geriatr Psychiatry 4: 173–179.
Bilder RM, Goldman RS, Volavka J, et al. 2002. Neurocognitive
effects of clozapine, olanzapine, risperidone, and haloperidol
in patients with chronic schizophrenia or schizoaffective disor-
der. Am J Psychiatry 159: 1018–1028.
Blyler CR, Gold JM. 2002. Cognitive effects of typical antipsycho-
tic medication treatment: another look. In Cognition in Schizo-
phrenia, Sharma T, Harvey PD (eds). Oxford University Press:
Oxford; 241–265.
Bowie CR, Harvey PD, Moriarty PJ, et al. 2002. Reliability and
validity of cognitive assessment in very low functioning patients
with schizophrenia. Arch Clin Neuropsychology 17: 611–623.
Bridge TP, Cannon HE, Wyatt RJ. 1978. Burned-out schizophrenia:
evidence for age effects on schizophrenic symptomatology. J
Gerontol 33: 835–839.
Chouinard G, Ross-Chouinard A, Annable L, Jones BD. 1980. The
Extrapyramidal Symptom Rating Scale. Can J Neurol Sci 7: 233.
Cornblatt BA, Lenzenweger MF, Erlenmeyer-Kimling L. 1989. The
Continuous Performance Test, Identical Pairs version: II. Con-
trasting attentional profiles in schizophrenic and depressed
patients. Psychiatry Res 29: 65–85.
Cuesta MJ, Peralta V, Zarzuela A. 2001. Effects of olanzapine and
other antipsychotics on cognitive function in chronic schizophre-
nia: a longitudinal study. Schizophr Res 48: 17–28.
Davidson M, Harvey P, Welsh KA, Powchik P, Putnam KM,
Mohs RC. 1996. Cognitive functioning in late-life schizophre-
nia: a comparison of elderly schizophrenic patients and patients
with Alzheimer’s disease. Am J Psychiatry 153: 1274–1279.
Davidson M, Harvey PD, Powchik P, et al. 1995. Severity of symp-
toms in chronically institutionalized geriatric schizophrenic
patients. Am J Psychiatry 152: 197–207.
Davidson M, Harvey PD, Vervarcke J, et al. 2000. A long-term mul-
ticenter, open-label study of risperidone in elderly patients with
psychosis. Int J Geriatr Psychiatry 15: 506–514.
Eyler Zorrilla LT, Heaton RK, McAdams LA, Zisook S, Harris MJ,
Jeste DV. 2000. Cross-sectional study of older outpatients
with schizophrenia and healthy comparison subjects: no differ-
ences in age-related cognitive decline. Am J Psychiatry 157:
1324–1326.
Folstein MF, Folstein SE, McHugh PR. 1975. ‘Mini-mental State.’
A practical method for grading the cognitive state of patients for
the clinician. J Psychiatr Res 12: 189–198.
Copyright # 2003 John Wiley & Sons, Ltd.
827
Green MF, Kern RS, Braff DL, Mintz J. 2000. Neurocognitive def-
icits and functional outcome in schizophrenia: are we measuring
the ‘right stuff’? Schizophr Bull 26: 119–136.
Green MF, Marshall BD, Jr, Wirshing WC, et al. 1997. Does risper-
idone improve verbal working memory in treatment-resistant
schizophrenia? Am J Psychiatry 154: 799–804.
Green MF. 1996. What are the functional consequences of neuro-
cognitive deficits in schizophrenia? Am J Psychiatry 153:
321–330.
Gurvich T, Cunningham JA. 2000. Appropriate use of psychotropic
drugs in nursing homes. Am Fam Physician 61: 1437–1446.
Hamilton M. 1960. A rating scale for depression. J Neurol Neuro-
surg Psychiatry 23: 56–62.
Harding CM, Brooks GW, Ashikaga T, Strauss JS, Breier A. 1987.
The Vermont longitudinal study of persons with severe mental
illness, II: long-term outcome of subjects who retrospectively
met DSM-III criteria for schizophrenia. Am J Psychiatry 144:
727–735.
Harvey PD, Ariola L, De Smedt G, Vester-Blockland E. 2003.
Cross-national cognitive assessment in schizophrenia clinical
trials: a feasibility study. Schizophr Res 59: 243–251.
Harvey PD, Heaton RK, Palmer B, Mohammed S, Kennedy J,
Brickman A. 2002. Stability of cognitive impairment in elderly
schizophrenic patients receiving conventional antipsychotic
treatment. J Clin Psychiatry 63: 1065.
Harvey PD, Howanitz E, Parrella M, et al. 1998. Symptoms, cog-
nitive functioning, and adaptive skills in geriatric patients with
lifelong schizophrenia: a comparison across treatment sites.
Am J Psychiatry 155: 1080–1086.
Harvey PD, Keefe RSE. 2001. Studies of cognitive change in
patients with schizophrenia following novel antipsychotic treat-
ment. Am J Psychiatry 158: 176–184.
Harvey PD, Meltzer HY, Green MF, McGurk SR. Cognitive effects
of risperidone and olanzapine in schizophrenia and schizoaffec-
tive disorder. Psychopharmacology (in press).
Harvey PD, Sukhodolsky D, Parrella M, White L, Davidson M.
1997. The association between adaptive and cognitive deficits
in geriatric chronic schizophrenic patients. Schizophr Res 27:
211–218.
Harvey PD, Moriarty PJ, Serper MR, Schnur E, Lieber D. 2000.
Practice-related improvement in information processing with
novel antipsychotic treatment. Schizophr Res 46: 139–148.
Heaton R, Paulsen JS, McAdams LA, et al. 1994. Neuropsycholo-
gical deficits in schizophrenics. Relationship to age, chronicity,
and dementia. Arch Gen Psychiatry 51: 469–476.
Heaton RK, Chelune CJ, Talley JL, Kay GG, Curtiss G. 1993. Wis-
consin Card Sorting Test Manual-Revised and Expanded. Psy-
chological Assessment Resources: Odessa, FL.
Hegarty JD, Baldessarini RJ, Tohen M. 1994. One hundred years of
schizophrenia: a meta-analysis of the outcome literature. Am J
Psychiatry 151: 1409–1416.
Jeste DV, Eastham JH, Lacro JP, Gierz M, Field MG, Harris MJ.
1996. Management of late-life psychosis. J Clin Psychiatry
57: 39–50.
Kay S, Fiszbein A, Opler LA. 1987. The Positive and Negative Syn-
drome Scale (PANSS) for schizophrenia. Schizophr Bull 13:
261–276.
Kay SR, Opler LA. 1987. The positive-negative dimension in schi-
zophrenia: its validity and significance. Psychiatr Dev 2: 79–103.
Klapow JC, Evans J, Patterson TL, Heaton RK, Koch WL, Jeste DV.
1997. Direct assessment of functional status in older patients
with schizophrenia. Am J Psychiatry 154: 1022–1024.
Lehman AF. 1988. A quality of life interview of the chronically
mentally ill. Eval Program Plan 11: 51–62.
Int J Geriatr Psychiatry 2003; 18: 820–829.
828 p. d. harvey et al.
Lezak MD. 1997. Neuropsychological Assessment, 3rd edn. Oxford
University Press: New York.
Madhusoodanan S, Brenner R, Suresh P, et al. 2000. Efficacy
and tolerability of olanzapine in elderly patients with psychotic
disorders: a prospective study. Ann Clin Psychiatry 12:
11–18.
Meltzer HY, McGurk SR. 1999. The effects of clozapine, risperi-
done, and olanzapine on cognitive function in schizophrenia.
Schizophr Bull 25: 233–255.
Morris JC, Heyman A, Mohs RC, et al. 1989. The Consortium to
Establish a Registry for Alzheimer’s Disease (CERAD). Part I.
Clinical and neuropsychological assessment of Alzheimer’s dis-
ease. Neurology 39: 1159–1165.
Palmer BW, Heaton RK, Paulsen JS, et al. 1997. Is it possible to be
schizophrenic yet neuropsychologically normal? Neuropsychol-
ogy 11: 437–446.
Copyright # 2003 John Wiley & Sons, Ltd.
Patterson TL, Goldman S, McKibbin CL, Hughs T, Jeste DV.
2001a. UCSD Performance-Based Skills Assessment: develop-
ment of a new measure of everyday functioning for severely
mentally ill adults. Schizophr Bull 27: 235–245.
Patterson TL, Moscona S, McKibbin CL, Davidson K, Jeste DV.
2001b. Social skills performance assessment among older
patients with schizophrenia. Schizophr Res 48: 351–360.
Purdon S, Jones BDW, Stip E, et al. 2000. Neuropsychological
change in early phase schizophrenia during 12 months of treat-
ment with olanzapine, risperidone, or haloperidol. Arch Gen Psy-
chiatry 57: 249–258.
Rossi A, Mancini F, Stratta P, et al. 1997. Risperidone, negative
symptoms and cognitive deficit in schizophrenia: an open study.
Acta Psychiatr Scand 95: 40–43.
Spreen O, Strauss JS. 1998. A Compendium of Neuropsychological
Tests and Norms, 2nd edn. Oxford University Press: New York.
Int J Geriatr Psychiatry 2003; 18: 820–829.