International Psychogeriatrics, Vol. 7, No.

4, 1995
0 1995 Springer Publishing Company

Effects of Early Treatment of Poststroke

Depression on Neuropsychological
Rehabilitation
Josk L. Gonzalez-Torrecillas, Julien Mendlewicz,
and Antonio Lob0

ABSTRACT. This study was made in an attempt todocument the effects of early
treatment of poststroke depression (including fluoxetine treatment) on neuro-
psychological rehabilitation (including cognitive function). Assessment mea-
sures used included the Schedule for Affective Disorders and Schizophrenia
( S A D S )and Research Diagnostic Criteria (RDC),as well as standard measures
of seventy of depression, functional ability, cognitive function, and neurologi-
cal function. Thirty-seven patients with poststroke depression, treated with
fluoxetine (n = 26) or nortriptyline (n = ll), were compared with 11 poststroke
depressed patients who received no depression treatment and 82 poststroke
nondepressed patients who received no depression treatment.Our findings about
the prevalenceof depression (37%), more frequent with anterior lesion (p = .009)
and left hemispherelesion (not statistically significant),tend to c o n f i i previous
reports. Early treatment (4th week poststroke) with either fluoxetine or
nortriptyline significantly improved the depressed patients’ mood neurological
function, functional ability, and cognitive ability. A close relationship between
appropriate early treatment (including fluoxetine treatment) of poststroke de-
pression and improved neuropsychological rehabilitation (including cognitive
improvement) is suggested by our findings. This is the first report, to our
knowledge, of the beneficial effects of early antidepressant treatment on the
cognitive function of poststroke depressed patients.

For many stroke patients, depression s e e m to be an important negative factor,

hindering their participation and progress in neurological (Robinson et al., 1984a;
Starkstein et al., 1987), functional (Feibel & Springer, 1982; Labi et al., 1980),
andcognitive (Ebrahim et al., 1985; Starksteinet al., 1988) rehabilitationprocesses.
The prevalence of poststroke depression is estimated to be between 25% and
50% (Dam et al., 1989; Eastwood et al., 1989; Finklestein et al., 1982; Folstein

From the Department of Psychiatry, Erasme Hospital, Free University of Brussels, Brussels, Belgium
(J. L. Gonzalez Torrecillas, MD; and J. Mendlewicz, MD), the Department of Psychiatry, Hospital
Clinic0 Universitario. Universidad de Zaragoza, Zaragoza, Spain ( A . Lobo, MD), and the
Department of Psychiatry, Centro Neuropsiquiatrico “Ntra. Sra. del Carmen,” Zaragoza, Spain (J. L.
Gonzalez-Torrecillas, MD).
547
548 J. L Gonzdez-Torrecillas et al.

et al., 1977; Kikumoto, 1990; Kotila et al., 1984; Robinson et al., 1982; Ross et
al., 1986; Wade et al., 1987).
Despite the known therapeuticeffect of pharmacological treatment of poststroke
depression (House, 1987), systematic studies at an early stage in its evolution
are very rare. Lipsey and colleagues (1984), in a double-blind study with
nortriptyline, and Reding and colleagues (1986), in a controlled study with
trazodone, reported a better outcome on standardized measures of depres-
sion. In both of these studies, patients with cardiovascular disease were
excluded. Fedoroff and colleagues (1989) reported a common passive therapeu-
tic attitude in poststroke depression, partly explained by the side effects of
traditional tricyclic antidepressant drugs in patients with severediseases, such as the
cardiovascular type.
In our investigation, we studied (a) the frequency of poststroke depression in the
4th week after stroke, (b) the relationship between lesion localization and depres-
sion, and (c) the responseto pharmacologicaltreatment (fluoxetineand nortriptyline) at
this early stage and the treatment’s effects on the level of neurological, functional,
and cognitive recovery.

METHOD
Sample

Our study was carried out with an initial sample of 165 patients consecutively
admitted to the neuropsychological rehabilitation service of a Brussels hospital.
All patients were in the 4th week after a stroke. Informed consent was obtained
from the patients or their representatives after the procedure had been fully explained.
Only patients who had a unilateral lesion documented by computed tomographic
(CT) scan and who were considered capable of compliance were included. In the
case of aphasia, only patients with a level of 2b/3 or lower according to the Goodglass
criteria (Goodglass & Kaplan, 1972) were included. Other reasons for exclusion
were alcoholism, drug abuse, any pathological condition capable of resembling
a depressive condition, or any antidepressant treatment in the 6 months before
the cerebrovascular accident. None of the patients had had a major or minor
depressive condition according to Research Diagnostic Criteria (RDC) in the 6
months before the study.
Thirty-five patients did not fulfill the inclusion criteria and were eliminated.
Table 1 summarizes some characteristics of the excluded patients that poten-
tially might have influenced the results. The excluded and included patients did
not differ significantly in distribution by race, socioeconomic status, marital
status, or family history of psychiatric disorder.
It was previously decided that for depression treatment purposes, four groups
would be formed:
1. Nondepressed patients not treated for depression comprised one group
(n = 82); 74 of them were considered to be noncases and 8 patients had
diagnoses other than depression.
2. Among the depressed patients, three groups would be formed. Approxi-
mately one fourth of them, selected with a table of random numbers among
Early Trearmenf of Poststroke Depression 549
TABLE 1. Characteristics of Patients Eliminated From Study (n = 35)

Age(Mean t SD) 70+11

Sex
Male 20 (57.1)
Female 15 (42.9)
History of psychiatric disorder
RDC-Major depression 3 (8.6)
RDC-Minor depression 2 (5.8)
Demented 3 (8.6)
Other RDC diagnosis 5 (14.3)
Family history of psychiatric disorders 7 (20)
Previous CVA 15 (42.9)
Lesion location
Right hemisphere 7 (20)
Left hemisphere 19 (54.3)
Nonhemispheric 5 (14.3)
Undetermined 4 (11.4)
Lesion location
Anterior 20 (57.1)
Posterior 11 (3 1.4)
Undetermined 4 (11.4)
Type of CVA
Ischemic 22 (62.9)
Hemorrhagic 9 (25.7)
Undetermined 4 (11.4)
Aphasic 20 (37.1)
Note. CVA = cerebrovascular accident; RDC = Research Diagnostic Criteria

consecutive depressed patients, would not be treated with antidepressants

and would serve as the control group ( n = 11).
3. The rest of the depressed patients would be allocated to the treatment groups.
To be included in the fluoxetine treatment group (n = 26), receiving a single
dose of 20 mg in the morning, for the 6 weeks of the study, they had to have
at least one of the three following characteristics: a left ventricular ejection
fraction (LVEF) of 40% or less, a prolonged QT interval on an electrocar-
diogram, andor orthostatic hypotension, without severe kidney failure.
4. Finally, depressed patients without any of the described abnormalities in
cardiac function were allocated to the nortriptyline treatment group ( T I =
11). These patients had an LVEF of 40% or greater, and no cardiac rhythm
disorder or orthostatism. The initial dose of nortriptyline was 25 rng per
day, increased every 3 days until it reached 75 mg and subsequently adjusted
according to plasma concentration.

Instruments

The patients were classified as having major depressive disorder, minor depres-
sive disorder, other psychiatric disorder, or no psychiatric disorder, according
to RDC (Spitzer et al., 1978). Depression was evaluated using the Schedule for
Affective Disorder and Schizophrenia (SADS; Spitzer & Endicot, 1975). The
degree of severity of depression was measured with the Hamilton Depression
550 J. L. Gonzdez-Torrecillas et al.
Rating Scale (HAMD; Hamilton, 1960), the Montgomery and Asberg Depres-
sion Rating Scale (MADRS; Montgomery & Asberg, 1979), and the 13-item
Beck Depression Inventory (BDI; Beck & Beck, 1972). The patients’ functional
state was assessed with Barthel’s Index (BI; Mahoney & Barthel, 1965) and
Karnofsky’s Performance Status Scale (KPS; Karnofsky & Burchenal, 1949).
Both functional measures are scored on a 100-point scale, with a low score indicating
greater disability. The cognitive evaluation was done by means of the Mini-Mental
State Examination (MMSE; Folstein et al., 1975).
The neurological evaluation was performed by the attending neurologist, who
was blind to the treatment group. He used the diagnostic criteria set by the Pilot
Stroke Data Bank (Kunitz et al., 1984) and Orgogozo’s scale (Orgogozo et al.,
1983). This scale ranges from 0 to 100, with a low score indicating more severe
neurological deficiency. Patients’ CT scans were evaluated by a neuroradiologist
who was blind to the clinical findings. Two simple classifications of lesion
location were considered: (a) rightlleftlnonhemisphere; (b) anterior/posterior
(calculated as the mean distance of the anterior lesion border from the frontal
pole for all CT slices in which the lesion was visible, following Robinson’s
criteria [Robinson et al., 1985aI).

Statistical Analyses

The statistical methods used to comparegroups included a chi-square test for the
categorical measures and two-tailed Student’s t test for the continuous mea-
sures. A one-way analysis of variance (ANOVA) was used to compare quantitative
variables in more than two groups of patients and Kruskal Wallis’ test was used
when the variance was not homogeneous. Bartlett’s and Cochran’s tests were
used for the homogeneity of variance. For measurement of the same variable
repeated over time, a multivariate analysis of variance (MANOVA) was per-
formed that analyzed the occasional effect due to inclusion in the group, the
effect of time, and the effect of interaction between group and time. The contrast
used was difference, each measure being compared to the mean of the
preceding measures.
When the same control group is compared with two groups of depressed
patients treated with different methods, an alpha type error might be increased.
A conservative way to deal with this was to divide by 2 the significance levels
(Bonferroni’s correction). A level of real significance was achieved only when
the rate was ,025. Similarly, when using scales measuring the same variable
(severity of depression: HAMD, MADRS, BDI), a conservative decision was to
divide the levels of significance by 3 (Bonferroni’s correction).
Dropouts were excluded for the statistical analysis of results of treatment.
Computations were performed using Statistical Package for the Social Sciences
statistical software (Nie et al., 1975).

RESULTS
On the basis of the SADS interview in the 4th week after a stroke, out of the 130
patients who made up our sample, 34 (26%) fulfilled the RDC for major depression,
Eariy Treatment of Posrstroke Depression 55 1

14 (1 1%) for minor depression, and 8 (6%) for other psychiatric disorders (anxiety,
personality disorder, etc.); 74 (57%) had no psychiatric disorders and were
considered to be “noncases” (Table 2).
The main characteristics of the 130 patients, grouped by RDC diagnosis, are
shown in Table 2. No significant differences in sociodemographic characteris-
tics (criteria of Hollingshead & Redlich, 1958) or personal psychiatric back-
ground were observed between groups. Previous strokes were more common in
both groups of depressed patients when compared with both groups of non-
depressed patients, but the differences were not statistically significant. A
subanalysis was performed to compare the collapsed groups of depressed patients
(18 patients out of 48 depressed patients or 37.5% had had a previous stroke)
with the collapsed group of nondepressed patients (21 patients out of 82
nondepressed patients or 25.6% had had a previous stroke), but the differences
still did not reach statistical significance (x’
= 1.5 1, df= 3 , p = 214). Therefore,
the four groups of stroke patients formed for treatment purposes were compa-
rable.
With regard to the location of the cerebral lesion (Table 3), major depression
was more frequent with anterior lesions (73%) and minor depression with
posterior lesions (69%), the differences being statistically significant (xz= 1 1.39,

TABLE 2. Characteristics of Study Population by Diagnostic Group (N = 13)

Major Minor Other RDC
Depression Depression Diagnosis Noncases
(n = 34) (n = 14) ( n = 8) ( n = 14)
Patient n vo n 70 n 070 n 070

Sex
Male 18 53% 7 50% 2 25% 33 45%
Female 16 47% 7 50% 6 15% 41 55%
Race
White 32 94% 12 86% I 87% 69 93%
Other 2 6 ‘70 2 14% 1 13% 5 I%
Married 11 50% 8 57% 5 62% 38 51%
Socioeconomic status
(070 IV or V). I 21% 2 14% 1 12% 18 24%
Personal history of
psychiatric disorder
RDC-Major.
depression 3 9 vo 2 14% 1 12% 5 7v o
RDC-Minoi-
depression 2 6 070 6 36% 1 12% 6 8%
Other RDC di-
agnosis 6 18% 3 21% 1 12% 14 19%
Family history of
psychiatric
disorders 7 21% 3 21% 2 25% 14 19%
Previous CVA 13 38% 5 36% 2 25% 19 26%
Age
(Mean ? SD) 61 t 1 3 66 ? 12 6 9 t 11 68 t 12
Note. There were no statistically significant differences for any of the variables among the four groups.
CVA = cerebrovascular accident; RDC = Research Diagnostic Criteria.
aIV or V is socioeconomic status category, by Hollingshead and Redlich criteria.
552 J. L. Gonzdlez-Torrecillas ef af.

TABLE 3. Lesion Location (CT Scan Findings) Among Depressed and Nondepressed
Patients
Major Minor Other RDC No
Depression Depression Diagnosis Depression
(n = 34) ( n = 14) ( n = 8) ( n = 74)
n oio n vn n oio n %

Lesion location*
Right hemisphere 13 38% 7 50% 2 2540 33 45%
Left hemisphere 19 56% 4 29% 6 75yo 30 40%
N o n hemispheric 2 6% 3 21% 0 0To 11 16%
Lesion location**
Anterior 24 7370 4 31% 6 75Vo 31 44%
Posterior 9 27070 9 69% I 25% 39 56%
Note. CT = computed tomographic; RDC = Research Diagnostic Criteria.
*ns. **pi.Ol.

df= 3,p = .009).

Major depression was also more frequent in left hemisphere lesions
(56%) compared to right hemisphere ones (38%), but the differences were not
statistically significant.
Figure 1 shows the results of antidepressive treatment on HAMD scores. On
Day 0, the three subgroups of depressed patients (treated with either fluoxetine
or nortriptyline or nothing) did not differ on HAMD scores (homogeneity on the
ANOVA: F = .649, df= 2 , p = S 2 7 ) . In the 6-week follow-up period, both groups
of treated patients improved, the course of the illness being similar, with no
statistical differences being demonstrated between them in the HAMD scores
(MANOVA analysis with Bonferroni’s corrections: F = .49, df = 6, p = .807).
Therefore, a simple group of treated poststroke depressed patients can be formed.
This group of depressed patients, treated with either fluoxetine or nortriptyline,
improved when compared with the nontreated depressed group (Figure l), and
the differences were statistically significant (HAMD: F = 9.12, df= 6 , p < .OOl).
All these results were similar when the severity of the patients’ depression
was assessed witheither MADRS ( F = 14.42, d’= 6 , p < .oOl) or BDI ( F = 7.23,
df= 6 , p < .OOl). The statistically significant improvement appeared in Week 3
on the HAMD and in Week 4 on both the MADRS and the BDI. At the end of
the 6 weeks, the group of treated depressed patients reached overall mean values
i n the depression scales similar to those in the group of nondepressed,
nontreated patients.
Figure 2 shows the results of antidepressant treatment on neurological function.
On Day 0, in the 4th week after stroke, the ANOVA gave homogeneity for the
Orgogozo’s scale mean values for the three depressed patient groups ( F = .157,
df = 2,p = .850). The MANOVA showed no differences in the mean values for
the groups treated with either fluoxetine or nortriptyline ( F = 1.56, df = 6,
p = .197). Therefore, a single group of treated depressed patients can be formed,
whose evolution of neurological recovery over time was significantly better
Early Trearnient 01Postsrroke Depression 553

Q) 30
-
m
U
1
V) 25
rn
.-S -
cr *O
iu
nc
c ’5
.-0
tn -
v) 10
2
a
Q ) 5
n p = .001
c
0 0 I I I I I I I

-
c,

-
.-
E
m
I Fluoxetine ( n = 26) Nontreated ( n = 11)
_t_ Nortriptyline ( n = 1 1) __o_Nondepressed ( n = 82)

Figure 1. Effects of antidepressive treatment with fluoxetine and nortriptyline on the

Hamilton’s (HAMD) scores on poststroke depressed patients.

than that of untreated patients ( F = 6.58, df = 6, p = .005). The differences were

significant from the 5th week of treatment on.
Similarly, antidepressant treatment improved the patients’ functional abili-
ties (Figure 3). On Day 0, in the examination of functional ability made with
the BI, the ANOVA showed homogeneity in the overall mean values for the
three depressed patient groups ( F = .006, df = 2 , p = .994). The MANOVA also
showed a similar evolution over time of the mean values on this scale for the two
depressed patient groups given fluoxetine or nortriptyline ( F = .34, df = 6,
p = .908). Consequently, a single group of treated depressed patients could be
formed whose functional ability evolution was significantly better compared to
the nondepressed untreated patient group ( F = 3.92, df = 6, p = .004). The
differences were statistically significant from the 3rd week of treatment on for
the BI. The results of antidepressant treatment were also demonstrated
when the patients’ functional abilities were assessed with the KPS. Statistically
significant differences were documented ( F = 8.69, df = 6 , p < .00l) from the 5th
week of treatment on for the KPS.
In terms of cognitive impairment, the three groups of depressed patients were
comparable on Day 0, before the treatment was started (MMSE, F =.2 19, df =
2, p = 304). However, the nondepressed group had a significantly higher score
on the MMSE (Figure 4). In a pattern similar to those for neurological and
functional ability, both subgroups of treated patients had a similar course of illness
in the 6-week follow-up period 0, = S 3 2 ) . The MANOVA permitted, therefore,
formation of a single treated group whose evolution over time was significantly
ss4

80 -
-
Q)
rn
u
v)
v) 70-
0
N
0
60-
z)
0
50 ‘ I I I I I 1 I

Day 0 Day 7 Day 14 Day 21 Day 28 Day 35 Day 42

--u- Fluoxetine ( n = 26) Y - Nontreated ( n = 1 1)

_5_ Nortriptyline ( n = 1 1 ) Nondepressed ( n = 82)

Figure 2. Effects of antidepressive treatment with fluoxetine and nortriptyline on the

neurological function (Orgogozo’s Scale) of poststroke depressed patients.

80 -

70 -

60 -

50 -

40 -
p = .004 -
30 I I I I I I I

- Fluoxetine ( n = 2 6 )
Nortriptyline ( n = 1 1)
-
- Nontreated ( n = 1 11
42

Nondepressed ( n = 8 2 )

Figure 3. Effects of antidepressive treatment with fluoxetine and nortriptyline on the

functional abilities (Barthel’s Index) of poststroke depressed patients.
Early Treatment of Poststroke Depression 555

28

c 27
.-*
0
m
.E 26
E
m
X
W 25
a,
c,
m
*
vl
24
- p = .036
m
c1
c 23
a Day 0 Day 7 Day 14 Day 21 Day 28 Day 35 Day 42
z
.-

-
I

-
.-t
I Fluoxetine ( n = 26) Nontreated ( n = 1 1)
Nortriptyline ( n = 1 1) ~ Nondepressed ( n = 82)

Figure 4. Effects of antidepressive treatment with fluoxetine and nortriptyline on the

cognitive function (MMSE scores) of poststroke depressed patients.

better compared to that in the nontreateddepressedgroup (F=2.58, df= 6,p = .036).

These differences were significant from the 5th w e k of treatment on, with the
treated group reaching values similar to those of the nondepressed patient group
after 6 weeks of treatment (Figure 4).
Since MMSE scores may be disproportionately influenced by left hemi-
sphere cognitive activities and because there was an enriched population of
depressed patients with left hemisphere lesions, we conducted a subanalysis of
treated and untreated depressed patients compared within hemisphere group.
Specifically, within the left hemisphere group, we compared the MMSE scores
ofthe 17 patients treated with either fluoxetine or nortriptyline with the scores
of the 6 nontreated patients. Similarly, within the right hemisphere group, we
compared thc MMSE scores of the 15 treated patients with the scores of the 3
nontreated patients. In both left and right hemisphere patients the treated group
did better, the differences i n MMSE scores being significant from the 5th week
of treatment on ( F = 8.604, df = 1, p = .009 in the left hemisphere patients;
F = 15.60, d j = 1, p = .01S in the right hemisphere patients).
I a t , in relation to dropouts and side effects, only 5 patients out of the 130
included could not complete the study. Two patients were transferred to another
hospital before the assessment was completed: One of them was nondepressed
and nontreatedand the other one was depressed and nontreated. One patient had
556 J. L. Gonzhlez-Torrecillas et al.
a new cerebrovascular accident (CV A) that prevented his participation. Only
two other patients abandoned the study because of side effects of fluoxetine (one
patient) or nortriptyline (one patient).

DISCUSSION

Our results confirm that depression is a common disorder in stroke evolution.

In spite of great sociocultural differences between study samples, our results are
similar to other findings in intrahospital studies, using criteria of the Diagnostic
and Statistical Manual of Mental Disorders (3rd ed., American Psychiatric
Association, 1980) (Sinyor et al., 1986b), and in extrahospital studies, using
RDC (Wade et al., 1987). We do not believe the excluded patients have seriously
influenced the results, because their characteristics were similar to those of the
included patients. However, since the proportion of both left hemisphere lesions
and anterior lesions was slightly higher in the excluded group, the prevalence of
depression might be even higher i n this sample.
Our findings tend to support the importance of lesion location in the develop-
ment of poststroke depression. Previous researchers have reported the associa-
tion between anterior lesions and depression (Ebrahim et al., 1987; Sinyoret al.,
1986a). Robinson and colleagues (1985b) have argued in favor of the associa-
tion with left hemisphere lesions, although Ross and colleagues (1986) and
Ebrahim and colleagues (1987) could not confirm those findings. We did not
find a significant relationship between left hemisphere lesions and depression,
although there was a trend toward a positive association that would favor
Robinson’s arguments.
Additional features of the CT lesions, such as lesion volume and lesion location
in relation to limbic structure and cortex, would be of interest and might be studied
in future investigations. We have studied functional dominance, indicated by
patient handedness. Only two patients in this sample were left-handed: One of
them was depressed and the other one was in the nondepressed nontreated group.
Therefore, we believe functional dominance has not biased the results in this study.
Furthermore, Robinson and colleagues (198Sa) suggested that the association
between depression and left-side stroke lesions was also valid in left-handed
patients.
Like other researchers (Robinson et al., 1984b), we recognize the important
negative effect of poststroke depression. However, there remain very few systeni-
atic studiesonearly antidepressive treatment. The current availability of nontricyclic
antidepressants (fluoxetine), whichother studies (Feighner & Cohn, 1985; Steinberg
et al., 1986) have shown not to have anticholinergic effects and very slight
cardiovascular side effects, enabled us to address the treatment of depression
after stroke associated with heart disease, as opposed to previous studies
(Lipsey et al., 1984; Reding et al., 1986) that excluded heart disease patients.
Our study tends to confirm the results reported by others (Reding et al., 1986)
Early Treafment of Poststroke Depression 557

in later treatment of post-CVA depression. After 6 weeks, the early post-CVA

treated group’s scores on the three measuring scales for depression descended
to values siniilar to those of the group initially nondepressed. This suggests that
it is easy and relatively simple and quick to improve depression in stroke
patients, including cases of severe intensity and major depression.
The results for neurological and functional evolution also suggest the nega-
tive effect of post-CVA depression on those aspects of neuropsychological
rehabilitation. However, the effects of early antidepressive treatment were positive,
because both aspects improved significantly during the follow-up period in the
treated patients, when compared to the depressed untreated. It has been suggested
that in some patients, depression may appear as a direct effect of the brain
lesion, in some as a functional response to the physical disability, and in some
as a combination of both (Caplan & Ahmed, 1992). Basedon our results, we tend
to agree with the first interpretation (Parikh et al., 1987; Starkstein et al., 1987)
and now emphasize the importance of antidepressant medication.
The beneficial effects of early antidepressive treatment on the cognitive
function of poststroke depressed patients should be emphasized since, to our
knowledge, it has not been reported previously. On Day 0 , 4 weeks after stroke,
the depressed patients scored significantly lower on the MMSE. Other authors
have previously reported similar findings (Fogel & Sparadeo, 1985; Kellner et
al., 1986). Both fluoxetine and nortriptyline significantly improved the MMSE
scores on the 5th week of follow-up treatment. It is remarkable that no other
treatment of poststroke patients has been documented to improve neurological
rehabilitation in this way. It might be argued that the MMSE is primarily a screening
battery for dementia and not a marker for change in poststroke cognitive function.
However, it has been used for this purpose (House et al., 1990). Future studies
trying to corroborate the suggestive results in our study could probably benefit
from more detailed neuropsychometric testing.
The number of dropouts from the study was limited and we believe has not
modified the main results of this research. Only two patients abandoned the
study because of the side effects of fluoxetine (one patient) or nortriptyline (one
patient), although the depression i n these two patients was rather severe.
Particularly severe was the condition of the patients treated with fluoxetine,
because their LVEF (less than 40%) was worse than that of the nortriptyline-
treated patients (LVEF greater than 40%). Therefore, this new, nontricyclic
antidepressant appears to have excellent potential i n poststroke depression.
Double-blind studies of fluoxetine efficacy are probably justified. If pos-
sible, such future studies should take into consideration some of the shortcom-
ings of our investigation, such as the limited follow-up period and the limited
number of patients studied. They should also analyze whether improvements seen
are disproportionately due to changes in particular items on the scales used,
because this might indicate the particular neurological and cognitive features on
which depression has a disproportionate impact.
558 J. L. Gonzdlez-Torrecillas et al.
REFERENCES
American Psychiatric Association. (1980). Diagnostic and statistical tizanual of nlental
disorders (3rd ed.). Washington, DC:Author.
Beck, A. T., & Beck, R. W. (1972). Screening depressed patients in family practice: A rapid
technique. Postgraduate Medicine, 52, 8 1-85.
Caplan, L. R., & Ahmed, I. (1992). Depression and neurological disease: Their distinction
and association. General Hospital Psychiatry, 14, 177-185.
Dam, H., Pedersen, H. E., & Ahlgren, P. (1989). Depression among patients with stroke.
Acta Psycliiatrica Scandinavica, 80, 118- 124.
Eastwood, M. R., Rifat, S. L., Nobbs, H., & Ruderman, J. (1989). Mood disorder following
cerebrovascular accident. British Journal of Psychiatry, 154, 195-200.
Ebrahim, S., Barer, D., & Nouri, F. (1987). Affective illness after stroke. British Journal of
Psychiatry, 1S1, 52-56.
Ebrahim, S., Nouri, F., & Barer, D. (1985). Cognitive impaiment after stroke. Age and
Ageing, 14, 345-350.
Fedoroff, J. P., & Robinson, R. G. (1989). Tricyclic antidepressant in the treatment of post-
stroke depression. Journal of Clinical Psychiatry, 50(7, Suppl.), 18-23.
Feibel, J. H., & Springer, C. J. (1982). Depression and failure to resume social activities after
stroke. Archives of PILysician and Medical Rehabilitalion, 63, 276-28 1.
Feighner, J. P., & Cohn, J. (1985). Double-blind comparative trials of fluoxetine and doxepin
in geriatric patients with major depressive disorder. Journal of Clinical Psychiatry, 46,
20-25.
Finklestein, S., Benowitz, L., Baldessarini, R. J., k a n a , G. W., Levine, D., et al. (1982).
Mood, vegetative disturbance and dexamethasone suppression test after stroke. Annals
of Neurology, 12,463-468.
Fogel, B. S., & Sparadeo, F. R. (1985). Focal cognitive deficits accentuated by depression.
The Journal of Nenioiis and Mental Disease, 173, 120-124.
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-Mental State”: A practical
method for grading the cognitive stale of patients for the clinician. Journal of
Psychiatric Research, 12, 189-198.
Folstein, M. F., Maiberger, R., & McHugh, P. R. (1977). Mood disorder as a specific
complication of stroke. Journal of Neurology, Neurosurgery and Psychiatry, 40,1018-
1020.
Goodglass, H., & Kaplan, E. (1972). The assesstiwnt of aphasia and related disorders.
Philadelphia: Lea & Febiger.
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurger-y
and Psychiatry, 23, 56-62.
Hollingshead, A. B., & Redlich, F. C. (1958). Social class and riwnlal illness: A comtnunizy
study (pp. 387-397). New York: John Wiley.
House, A. (1987). Mood disorders after stroke: A review of the evidence. International
Journal of Geriatric Psyckiutry, 2, 21 1-221.
House, A., Denis, M., Warlow, C., Hawton, K., & Molyneux, A. (19W). The relationship
between intellectual impairment and mood disorder in the first year after stroke. Psycho-
logical Medicine, 20, 805-814.
Kamofsky, D. A,, & Burchenal, J. H. (1949). The clinical evaluation of chemotherapeutic
agents in cancer. In C. M. Macleod (Ed.), Ewluation of cheiiwtherr~peziticagents (pp.
191-205). New York: Columbia University Press.
Emly Trratrnenr of Poststroke Depression s59

Kellner, C. H., Rubinow, D. R., & Post, R. M. (1986). Cerebral venuicularsizeandcognitivc

impairment in depression. Journal of Affective Disorders, 10, 215-219.
Kikumoto,0. (1990).Clinical study on depressive state following stroke.Seishin-ShinkeiXakif-
Zusshi, 92,411-434.
Kotila, M., Waltimo, O., Niemi, M. L., Laaksinen, K., & Lempinen, M. (1984). The profile
of recovery from stroke and factors influencing outcome. Stroke, 15, 1039-1044.
Kunitz, S. C., Gross,C. R., Heyman, A,, Kase, C. S., Mohr, J. P.,etal. (1984). The Pilot Stroke
Data Bank: Definition, design and data. Stroke, 15, 740-746.
Labi, M. L., Phillips, T. F., & Gresham, G. E. (1980). Psychosocial disability in physically
restored long-term stroke survivor. Archives of Physician and Medical Rehabilitation,
61, 561-565.
Lipsey, J. R., Robinson, R. G., Pearlson, G. D., Rao, K., & Price, T. (1984). Nortriptyline
treatment for post-stroke depression: A double-blind study. Lancet, 1, 297-300.
Mahoney, F. I., & Barthel, D. W. (1965). Functional evaluation: Barthel Index. Mayland
Srate Medical Journal, 14,61-65.
Montgomery, S . A,, & Asberg, M. (1979). A new depression scale designed to be sensitive
to change. British Journal of Psychiatry, 134, 382-389.
Nie, N., Hull, (1.H., &Jenkins, J. G. (1975). StatisticalpacX-ageSorthe social sciences (2nd
ed.). New York: McGraw-Hill.
Orgogozo, J. M., Capildeo, R., Anagnostou, C. N., Juge, O., Dartigues, J. F., et al.
(1983). Mise au point d’un score neurologique pour I’evaluation clinique des infarctus
sylviens [Standardization of a neurological score for the clinical assessment of infarcts
in the sylvian territory]. La Presse Mkdicale, 12, 3039-3044.
Parikh, R. M., Lipsey, J. R., Robinson, R. G., & Price, T. R. (1987). A two-year longitudinal
study of post-stroke mood disorders: Dynamic changes in correlates of depression at one
and two years. Stroke, 18, 579-584.
Keding, M. J., Orto, L. A., Winter, S. W., Fortuna, I. M., Di Polite, P., et al. (1986). Anti-
depressan1 therapy after stroke: A double-blind trial. Archives of Neurology, 43, 763-
765.
Robinson, R. G., Kubos, K. L., Starr, L. B., Kao, K., & Price, T. R. (1984a). Mood disorder
in stroke patients: Importance of location of lesion. Bruin, 107, 81-93.
Robinson, R . G., Lipsey, J. R., Bolla-Wilson, K., Bolduc, P. L., Pearlson, G. D., e t a l .
(1985a). Mood disorders in left-handed stroke patients. American Journal of Psyckia-
try, 142, 1424-1429.
Robinson, R. G., Lipsey, J. R., & Price, T. R. (1985b). Diagnosis and clinical management
of post-stroke depression. Psychosovuztics, 6,769-778.
Robinson, R. G., & Price, T. R. (1982). Post-suoke depressive disorders: A follow-up study
of 103 patients. Stroke, 13, 635-641.
Robinson, R. G., Starr, L. B., Lipsey, J. R., Kao, K., & Price, T. R. (1984b). A two-year
longitudinal study ofpost-stroke mooddisorders: Dynamicchanges inassociafedvariables
over the first six months of follow-up. Stroke, 15, 510-517.
Ross, E. D., Gordon, W. A., Hibbard, M., & Egelco, S. (1986).The dexamethasone suppression
test post-stroke depression, and the validity of DSM-III-based diagnostic criteria Ketter].
Arnericun .Journal of Psychiatry, 143, 1200-1201.
Sinyor, D., &to, P., Kaloupek, D. G., Becker, R., Goldenberg, M., et al. (1986a). Post-
stroke depression: Relationships to functional impairment, coping strategies, and
rehabilitation outcome. Stroke, 17, 1102-1107.
Sinyor, D., Jacques, P., Kaloupek, D. G., Becker, R., Goldenberg, M., et al. (1986b). Post-
stroke depression and lesion location: An attempted replication. Brain, 109, 537-546.
560 J . L. Gonzdez-Torrecillas et al.

Spitzer, R. L., & Endicot, J. (1975). Schedule for Affective Disorders and Schizophrenia
(SADS) (2nd ed.). New York: New York State Psychiatric Institute.
Spitzer, R. L., Endicot, J., & Robins, E. (1978). Research Diagnostic Criteria (RDC)fir a
selectedgroup offinctiunal disorders (2nd ed.). New York: New York State Psychiatric
Institute.
Starkstein, S. E., Robinson, R. G., & Price, T. R. (1987). Comparison of cortical and
subcortical lesions in the production of poststroke mood disorders. Brain, 110, 1045-
1059.
Starkstein, S. E., Robinson, R. G., & Price, T. R. (1988). Comparison of patients with and
without post-stroke major depression matched for size and location of lesion. Archives
of General Psychiatry, 45, 241-252.
Steinberg, M. I., Smallwood, J. K., & Holland, D. R. (1986). Hemodynamic and electro-
cardiographic effects of fluoxetine and its major metabolite, norfluoxetine, in
anesthetised dogs. Toxicology and Applied Pkarmacology, 82, 70-79.
Wade, D. T., Legh-Smith, J. E., & Hewer, R. A. (1987). Depressed mood after stroke: A
community study of its frequency. British Journal of Psychiatry, 151, 200-205.

Ackmowledgrnmt This work was presented at the 146th Annual Meeting of the
A m e r i c a n Psychiatric Association, S a n Francisco, California, M a y 22-27,
1993.

Offprints. Requests for offprints should be directed to Antonio Lobo, MD, De-
partamento de Psiquiatria (planta 1 I ) , Hospital Clinico Universitario, 50.009
Zaragoza, Spain.