Psychiatry Research 228 (2015) 475–481

Contents lists available at ScienceDirect

Psychiatry Research
journal homepage: www.elsevier.com/locate/psychres

Stability of executive functions in first episode psychosis: One year

follow up study
Beathe Haatveit a,n, Anja Vaskinn a,b, Kjetil S. Sundet a,b, Jimmy Jensen a,c,
Ole A. Andreassen a, Ingrid Melle a, Torill Ueland a,b
a
NORMENT, KG Jebsen Centre for Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital & Institute of Clinical Medicine,
University of Oslo, Kirkeveien 166, 0407 Oslo, Norway
b
Department of Psychology, University of Oslo, P.O. Box 1094 Blindern, 0317 Oslo, Norway
c
Centre for Psychology, Kristianstad University, Elmetorpsvägen 15, 291 39 Kristianstad, Sweden

art ic l e i nf o a b s t r a c t

Article history: Executive functioning is a multi-dimensional construct covering several sub-processes. The aim of this
Received 9 September 2014 study was to determine whether executive functions, indexed by a broad range of executive measures
Received in revised form remain stable in first episode psychosis (FEP) over time. Eighty-two patients and 107 age and gender
19 May 2015
matched healthy controls were assessed on five subdomains of executive functioning; working memory,
Accepted 24 May 2015
Available online 25 June 2015
fluency, flexibility, and inhibitory control at baseline and at 1 year follow-up. Results showed that pa-
tients performed significantly poorer than controls on all executive measures at both assessment points.
Keywords: In general executive functions remained stable from baseline to follow-up, although both groups im-
Schizophrenia spectrum disorders proved on measures of inhibitory control and flexibility. In phonemic fluency, controls showed a slight
Cognition
improvement while patients showed a slight decline. Investigation of individual trajectories revealed
Longitudinal study
some fluctuations in both groups over time, but mainly supports the group level findings. The im-
Reliable change
plications of these results are discussed.
& 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction assesses all components of executive function. Thus, some criti-

Impairments in executive functioning are evident in the ma-
jority of schizophrenia spectrum patients, and are observed
throughout all stages of the illness (Rund et al., 2007; Holmen et
al., 2012b; Barder et al., 2013a, 2013b; Sanchez-Torres et al., 2013).
Executive dysfunction is present already in the first year of the
illness, even before the first contact with the public health care
service (Hoff et al., 2005), and is a contributing factor to functional
loss and disability. Executive functioning predicts degree of self-
care, as well as social, interpersonal, community, and occupational
functioning (Mcgurk and Mueser, 2003; Bowie and Harvey, 2006),
and is associated with treatment success. Impairments in this
domain are coupled with less engagement in therapy, poorer
medication adherence, and longer hospital stays (McKee et al.,
1997; Jackson et al., 2001; Robinson et al., 2002; Bowie and Har-
vey, 2006).
There is no established consensus of which tests are best suited
to assess executive functions, nor is there any single test that
n
Correspondence to: Section for Psychosis Research, Building 49, Division of
Mental Health and Addiction, Oslo University Hospital, Ulleval, Kirkeveien 166,
0407 Oslo, Norway.
E-mail address: b.c.haatveit@medisin.uio.no (B. Haatveit).
cism has been directed towards an inconsistently defined execu-
tive domain (Bozikas and Andreou, 2011) and longitudinal studies
have shown ambiguous findings (Liu et al., 2011). Executive
functioning is a multi-dimensional construct covering several sub-
processes including, working memory, fluency, flexibility, in-
hibitory control and problem-solving. In general, longitudinal
studies of executive functioning have investigated a limited range
of cognitive sub-processes, and yet refer to executive function as a
whole (Frangou, 2010). This may lead to misinterpretations, since
results may be more related to the sensitivity and psychometric
properties of the specific test used rather than specificity to the
executive process being measured. Several reviews of neurocog-
nition in schizophrenia report a discrepancy in effect sizes across
different executive measures (Szoke et al., 2008; Aas et al., 2014).
This could be due to variability in the difficulty level of the specific
tests or to the degree of dysfunction in the different sub-functions
being measured in the patient group, underlining the importance
of using a variety of tests.
There is still an ongoing debate as to whether schizophrenia is
a neurodegenerative disorder with brain related changes after
illness onset or a neurodevelopmental disorder with debut early in
life (Rund, 2009). This is critical knowledge with implications for
illness recovery and future functioning. Still, executive functions

http://dx.doi.org/10.1016/j.psychres.2015.05.060
0165-1781/& 2015 Elsevier Ireland Ltd. All rights reserved.
476 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481

Table 1
Demographic and clinical characteristics for patients and controls. Means, standard deviations, and results from group comparisons are reported.

1a: Demographics Patients Controls Group comparison

Age (years) 26.7(7.6) 28.6(6.8)
1.8 0.073

Education (years) 12.1(2.2) 14.1(2.1)
6.6 o 0.001
Estimated IQ (WASI) 102.5(14.8) (n81) 114.8(16.0)
5.4 o 0.001
IQ matched subsample (n62) 109.0(9.9) 110.7(9.5)
1.0 0.318
Sex (m/f) 50/32 62/45 x2(1, N ¼189) ¼0.2 0.674
1b: Clinical characteristics Baseline Follow-up

Age at onset 23.5(7.4)

DUP(median) 45(1–1040a)
Diagnoses
Schizophrenia 35(43%)
Schizophreniform 9(11%)
Schizoaffective 4(5%)
Other psychosis 25(30%)
Major depressive disorder 9(11.0%)
On anti-psychotic medication Z 1 64(78.1%)
Months on antipsychotic 2.6(3.1)
medication
GAF function 45.3(13.9) 53.4(15.5)
GAF symptom 41.9(12.2) 50.0(15.9)
PANSS positive score 15.6(5.0) 13.2(5.0)
PANSS negative score 15.2(6.4) 13.8(5.8)

PANSS: Positive and Negative Syndrome Scale.

GAF: Global Assessment of Functioning Scale.
a
Range.

are reported as relatively stable from treatment start in patients 2. Methods

with first episode psychosis (FEP). Longitudinal studies without a
control sample generally report minor improvements over the first
years, but an overall stability over longer periods (Rund et al.,
2007; Cohen et al., 2012; Barder et al., 2013a, 2013b). The same
stability is observed when patient samples are compared to
healthy control subjects (Hoff et al., 2005; Bozikas and Andreou,
2011; Sanchez-Torres et al., 2013). In studies in which improve-
ments are observed in patients (Addington et al., 2005; Hoff et al.,
2005; Mayoral et al., 2008; Rodriguez-Sanchez et al., 2008; Szoke
et al., 2008), the same pattern is most often also seen in the
control subjects, indicating that changes may be due to practice
effects or the natural development and maturation within the
samples (Goldberg et al., 2007).
To ensure that possible findings are not the result of practice
effects of repeated measurements, it is thus important to have an
age and gender matched control group. This can also help to de-
termine whether stability among patients actually reflects lack of
development. Although practice effects which are often reported
in patient studies might mask such a lack of development, rela-
tively few studies include a control group (Szoke et al., 2008). An
additional step in detecting actual changes in longitudinal studies
is to calculate reliable change indices (RCI). This could provide
information as to whether changes seen on the group level exist
on the individual level, and whether these changes are reliable and
unlikely to be caused by measurement error or practice effects
(Heaton et al., 2001; Iverson, 2001; Parsons et al., 2009; Duff,
2012). Thus, the aim of this study was to determine whether ex-
ecutive functions remain stable over one year in first episode pa-
tients compared to a healthy age and gender matched control
group, using a broad range of executive tests covering the execu-
tive subdomains of working memory, fluency, flexibility, and in-
hibitory control. Our main objective was to investigate perfor-
mance in these functions over time and to define the magnitude of
reliable changes on the individual level.
2.1. Participants
This study included 82 patients with first episode psychosis
(FEP), defined as less than one year since starting their first ade-
quate treatment for a DSM-IV diagnosis of schizophrenia (42.7%),
schizophreniform disorder (11.0%), schizoaffective disorder (4.9%),
major depression with mood incongruent psychotic symptoms
(11.0%), and other psychosis (30.5%), as well as 107 healthy control
participants from the ongoing Thematic Organized Psychosis (TOP)
research in Oslo, Norway. All participants were recruited between
2005 and 2012. The average test-interval time between baseline
and 12 months follow-up was 406.3 days (SD 66.5).
Diagnostic assessment was based on the “Structured Clinical
Interview for DSM-IV, axis 1 (SCID-1; (First et al., 1995)) and
symptom assessment on the Positive and Negative Syndrome Scale
(PANSS; (Kay et al., 1987)). Psychosocial functioning was assessed
with the Global Assessment of Functioning Scale split version
(GAF-S; (Pedersen et al., 2007)). Age at first psychotic episode was
calculated based on age at first psychotic symptoms (23.5 years,
SD 7.4), DUP was measured as the time from onset of psychotic
symptoms (the first week with PANSS score of 4 or above on the
last one (item) of the Positive Scale items 1, 3, 5, 6 or general item
9) until start of first adequate treatment. The DUP median was 45
weeks (SD 198.4, range 1-1040). Medication use is reported in
current usage of one or more antipsychotic drugs, in addition
number of months the participants had been on their main anti-
psychotic medication.
The healthy control group was randomly selected from the
same catchment area as the patient group using statistical records.
Exclusion criteria for both groups were traumatic brain injury,
neurological disorders and other medical condition interfering
with brain functioning, or signs of mental retardation (IQ o70).
Also to ensure valid test performance all participants had a score
Z15, on the CVLT forced recognition task (CVLT-II (Delis et al.,
2004). The healthy controls were screened before participating,
 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481
and participants who presented a history of severe drug abuse the
last 12 months, participant with severe mental illness or those
who had mental disorders in their close family, were excluded.
As the groups were matched, there were no differences in age
and gender between them. There were however, significant dif-
ferences in IQ and education level with lower levels in the pa-
tients. The healthy control group had on average 2 years more
education, and obtained 12.3 estimated IQ points more than the
FEP group, see Table 1a. To minimize the effect of random variance
in intellectual abilities, a subsample consisting of 62 controls and
62 patients were matched on IQ. To accomplish this all cases with
IQ below 90 (20 patients, 1 control) and above 130 (4 controls)
were excluded. The remaining patients (62) were randomly mat-
ched on group level with equal number controls. Further demo-
graphic and clinical characteristics are shown in Table 1.
All participants gave written informed consent, and the study
was approved by the Regional Committee for Medical Research
Ethics and the Norwegian Data Inspectorate. For further details
regarding selection and assessment, see Tandberg et al. (2012).
2.2. Assessment of executive function
The neurocognitive assessment was administered by a psy-
chologist or assistant trained in standardized neuropsychological
assessment. The following subdomains of executive functions
were investigated.
2.2.1. Working memory
The ability to hold and manipulate information for briefer
periods refers to our working memory capacity. To measure
working memory we used the Digit Span backward test (Wechsler,
2003), where increasingly longer sequences of digits are read
aloud to the participant. In addition we used the Letter-Number
Sequencing test (LNS) (Wechsler, 2003), where combinations of
increasingly longer sequences of intermingled numbers and letters
are read aloud. The longest series of correct digits was reported for
the Digit Span backward task and the total sum of correctly re-
ported sequences for the LNS task.
2.2.2. Fluency
Tests of fluency involve associative exploration and retrieval of
verbal information such as sematic and phonetic material (Henry
and Crawford, 2005). To measure phonetic and semantic fluency
we used the Letter and Category Fluency both form the Verbal
Fluency Test, part of the Delis Kaplan Executive Functioning Sys-
tem (D-KEFS)(Delis et al. 2005a). In Letter Fluency, subjects are
asked to generate as many words as possible beginning with the
letters F, A, and S within a given time frame (60 s). In Category
Fluency participants are asked to generate as many words as
possible, from a given category, within a given timeframe (60 s).
Fig. 1. Z-score profile revealing significant differences between patients and controls at baseline and 12 months follow-up on measures of executive functioning.
477
The categories given were animals and boys names. The total
number of words generated in the two conditions is reported.
2.2.3. Flexibility
Flexible thinking and the ability to switch between different
concepts or strategies can be categorized as our flexibility skills. To
measure flexibility we used the condition 3b Category Switching
from the D-KEFS Verbal Fluency Test (Delis et al., 2005b), and
perseverative responses form the short version of the Wisconsin
Card Sorting Test-64 (WCST (Kongs et al. 2000)). In Category
Switching participants are asked to alternate between two given
categories, fruits and furniture, and generate as many words as
possible within 60 s. Total number of correct set shifts are re-
ported. The WCST is a computer-based task, for which participants
are instructed to sort cards based on rules that change throughout
the course of the test. The total number of perseverative responses
is reported.
2.2.4. Inhibitory control
Inhibitory control is our ability to control overt behavior such
as impulses and motor responses. Inhibitory control was measured
using the Inhibition condition from the Color-Word Interference
Test, D-KEFS (Delis et al. 2005). In this condition participants are
asked to name the color of the ink on written words in incon-
gruent colors. The total time taken to complete the task is
reported.
Raw scores for all tests are reported.
2.3. Statistical analyses
All statistical analyses were conducted using the Statistical
Package for the Social Science (SPSS) for Windows, version 20.0
(SPSS Inc., Chicago, IL, USA). Chi-square analysis compared group
differences on categorical variables. Group differences on con-
tinuous variables were investigated using analyses of variance
(ANOVA), Independent samples t-tests, and Paired samples t-tests.
We used 7 separate two-way repeated measures analyses
(ANOVAs) (Bonferroni corrected) to investigate change over time
between and within the patient and control group. The executive
measures were entered as the dependent variables with group
affiliation as the independent variable and assessment time the
repeated variable. Reported are means, standard deviations, and F
values with flagged significance level for group, time, and time-
 group interactions. Partial eta-squared (η2) is also reported for
group differences across testing intervals. The z-scores used in
Fig. 1 were based on means and standard deviations in the control
sample. We calculated how much patients differed from controls
at baseline and follow-up, and also how much controls differed
from themselves at follow up. For follow-up analyses we used
Independent samples t-tests and to-way repeated measures
478 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481

analyses (ANOVAs).
Further, to investigate individual changes we calculated RCI
with adjustment for practice effects (RCIPE). This index evaluates
whether individual changes in test scores are reliable. RCIPE was
calculated based on Iverson's modified version (2001) of (Chelune
et al., 1993; Duff, 2012). RCIPE is calculated using the formula:
((T2
T1)
(M2
M1))/SEDIverson, where T1 and T2 are the dis-
crepancy between test and retest scores, and measurable practice
effect are the discrepancy between the means (M1, M2) at the two
assessment time points in the control sample. This practice ad-
justed discrepancy score is further the numerator in this formula,
and the standard error of the difference (SEDIverson) is the de-
nominator, taking both variability of test-retest scores into con-
sideration: SEDIverson ¼√((S1√(1
r12)) 2 þ(S2√(1
r12))2). S1 is
the standard deviation at T1 and S2 are the standard deviation at
T2, and r12 are the correlation between T1 and T2 scores. The re-
sulting RCIPE's is eventually compared with a normal distribution,
and all scores 71.645 are classified as statistically significant
changes (decline or improvement), outside the 6th percentile with
an 90% confidence interval (CI) (Duff, 2012). We used the case
report function in SPSS to descriptively follow individual change
over time, also to compare with relevant demographic or clinical
variables. In addition, we calculated an executive global decline
score based on the 7 possible outcome variables, and all subjects
with a reliable decline on one or more subdomains received a
global decline score between 1 and 7. We also conducted chi-
square tests for independence to explore the relationship between
group affiliation and degree of stability. Degrees of stability and
group affiliation were categorical variables. We categorized stable
cases with improvements versus those cases with decline, and
compared across patient and control group. Reported from the chi-
square analyses is continuity correction value with an associated
significance level. We used the fisher's exact test for variables that
violated the assumptions of chi-square analyses.
baseline and follow-up are shown in Table 2. The FEP group had
significantly poorer performance on all measures compared to the
healthy controls at both time-points. Letter Fluency and Category
Fluency, Letter Number Sequencing and Inhibition displayed the
largest group effects across the testing intervals (η2 ranging from
0.21 to 0.32). These are medium to large effect sizes (by def. small
 0.02, medium 0.13, large  0.26). Digit Span backward and
WCST perseverative responses showed the smallest group-differ-
ences (η2 ranging from 0.11 to 0.13). When analyzing performance
change over time for each executive measure we found significant
time effects for Inhibition and WCST perseverative responses,
where both FEP patients and healthy controls improve. There was
also one significant time  group interaction on Letter Fluency
(η2 ¼0.04), with patients showing slight deterioration and controls
slight improvements. All results remained significant after ad-
justing with a conservative Bonferroni alfa level (0.05/7(number of
tests) Ep o0.01). Fig. 1 show the z-scores for the two groups on all
the executive subdomains.
To control for possible confounding effects of baseline differ-
ences in intellectual abilities, we performed follow-up analyses on
a IQ matched subsample (t(122 ) ¼
1.00, p¼ 0.32), consisting of 62
patients (mean 109.0, SD 9.9) and 62 healthy controls (mean 110.7,
SD 9.6). The results from these analyses indicated that there was
only one confounding effect of baseline IQ level regarding the
observed group differences in Digit Span backward. The significant
group difference on this test disappeared when the samples were
matched on IQ. All other results remained unaffected also sur-
viving Bonferroni correction (Table 3). Two additional time-
 group interactions on the WCST preservative responses
(η2 ¼0.04.), and on the LNS (η2 ¼0.03) occurred, but when cor-
rected for multiple testing these findings did not remain
significant.
3.2. Reliable change index

To estimate the RCIPE we calculated the standard deviation at

3. Results
3.1. Changes in executive functions from baseline to follow-up
Results for the 7 tests measuring the 4 executive subdomains at
test-retest for both groups, and the correlation between the two
assessment points for all the executive measures for the control
subjects. The standard deviations are presented in Table 2, and the
correlations were as follow: Letter Number Sequencing ¼0.64,
Digit Span backward ¼0.45, Letter Fluency ¼0.74, Category

Table 2
Scores on the executive functioning tests at baseline and 12 months follow-up for the two groups. Means, standard deviations, and results from repeated measures ANOVA
are reported.

Patients Controls F values

Baseline 1 year Baseline 1 year Group Time Time  group

Working memory
Digit Span backward 4.3(1.2) 4.3(1.1) 5.0(1.2) 5.1(1.1) 24.4nn 0.1 0.6
Letter Number Sequencing 9.1(2.5) 9.5(2.7) 11.8(2.6) 11.6(2.5) 48.5nn 0.5 2.8

Fluency
Letter Fluency 35.2(12.9) 33.6(13.3) 45.2(10.2) 46.8(10.3) 52.6nn 0 7.5n
Category Fluency 38.0(10.3) 37.3(10.7) 49.2(7.0) 49.2(8.7) 88.6nn 0.4 0.3

Flexibility
Category Switching 11.2(3.0) 10.7(3.2) 13.8(2.8) 14.1(2.9) 69.8nn 0.1 3.1
WCST perseverative responses 10.1(5.5) 7.7(4.9) 6.8(3.9) 5.9(2.7) 27.5nn 17.0nn 3.6

Inhibitory control
Inhibition 65.2(22.0) 61.6(21.5) 49.0(11.3) 46.4(10.0) 49.0nn 12.9nn 0.4

WCST: Wisconsin Card Sorting Test.

n
pr 0.01.
nn
p r0.001.
 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481 479

Table 3
Baseline and 12 months follow-up test scores for IQ matched groups. Means, standard deviations, and results from repeated measures ANOVA are reported

Patients Controls F values

Baseline 1 year Baseline 1 year Group Time Time x group

Working memory
Digit Span backward 4.6(1.3) 4.6(1.1) 4.7(1.2) 4.9(1.1) 1.6 0.4 0.6
Letter Number Sequencing 9.7(2.3) 10.2(2.5) 11.5(2.7) 11.2(2.6) 10.3nn 0.2 4.3n

Fluency
Letter Fluency 38.5(11.7) 37.8(9.3) 42.5(9.5) 45.4(10.5) 10.1nn 2.7 7.5nn
Category Fluency 40.9(9.2) 40.6(9.3) 48.5(7.2) 47.8(8.1) 28.8nnn 0.6 0.1

Flexibility
Category Switching 11.7(3.0) 11.0(3.3) 13.7(2.5) 13.7(2.8) 28.8nnn 1.7 1.6
WCST preservative responses 8.9(4.8) 6.9(3.8) 6.4(3.4) 6.2(3.0) 7.3nn 7.9nn 5.3n

Inhibitory control
Inhibition 59.6(17.5) 55.6(14.6) 48.7(12.5) 46.6(10.8) 17.7nnn 12.3nnn 1.0

WCST: Wisconsin Card Sorting Test.

n
pr 0.05.
nn
p r0.01.
nnn
p r 0.001.

Table 4
Reliable changes within 90% CI 7 1.645.

Decline n (%) Stable n (%) Improvement n (%)

Patients Controls Patients Controls Patients Controls

Working memory
Letter Number Sequencing 0 10 (9) 73 (89) 95 (89) 9 (11) 2 (2)
Digit Span backward 8 (10) 11 (10) 73 (89) 94 (88) 1 (1) 2 (2)
Fluency
Letter Fluency 13 (16) 3 (3) 68 (83) 101 (94) 1 (1) 3 (3)
Category Fluency 4 (5) 4 (4) 77 (94) 95 (89) 1 (1) 8 (8)
Flexibility
Category Switching 5 (6) 5 (5) 76 (93) 97 (91) 1 (1) 5 (5)
WCST perseverative responsens 4 (5) 2 (2) 68 (83) 97 (91) 10 (12) 8 (8)
Inhibitory control
Inhibtion 10 (12) 4 (4) 62 (76) 96 (90) 10 (12) 7 (7)

WCST: Wisconsin Card Sorting Test.

Fluency ¼0.55, Category Switching ¼ 0.24, WCST preservative re-
sponses ¼ 0.15, Inhibition ¼0.80. All measures except the WCST
preservative responses showed significant correlations from
baseline to follow-up indicating high reliability i.e. high con-
sistency of change within each group. The results from the RCIPE
calculations are presented in Table 4. The single RCI scores on the
different executive tests are counted and divided into those stable,
declining, or improving according to the 71.645 cutoff point
(Table 4). With a 90% confidence interval, 90% of the controls re-
main stable. The FEP group ranges in stability scores varied from
76–94%. There were no patterns of a global decline with numerous
patients scoring lower on several subtests. Two patients had a
global decline score of 3, 6 had a score of 2, and 26 had a score of 1.
In the control group, seven had a global decline score of 2, and 25
had a score of 1.
Chi square test of independence revealed a significant asso-
ciation between group affiliation and degree of stability on Letter
Number Sequencing (x2(1, 189) ¼6.33, p ¼0.01), where controls de-
cline, and on the Letter Fluency test (x2(1, 189) ¼8.59, p o0.01),
where patients decline. There is no association between group
affiliation and degree of stability on the Digit Span backward (x2(1,
2
189) ¼ 0.00, p¼ 1.00), neither on the Inhibition condition (x(1,
189) ¼ 3.69, p¼ 0.06). Category Fluency, Category Switching and
WCST perseverative responses did not meet the criteria of chi-
square analysis, and fisher’s exact test shows now association
between group affiliation and degree of stability on these vari-
ables: Category Fluency (p¼ 0.73), Category Switching (p ¼0.75)
and WCST perseverative responses (p ¼ 0.41).
4. Discussion
The current study investigated a broad range of executive sub-
domains in FEP at study inclusion and 1 year later. There are three
main findings: First, there was an overall stability on measures of
executive functioning in patients and controls, i.e. with continuous
impairments at baseline and follow-up in the patient group. Second,
we found improvements in inhibitory control and on one measure of
flexibility, and a change across groups in phonemic fluency. Finally,
with regard to the individual trajectories (RCI calculations) we found
that both groups show some fluctuations in both directions over time,
but an overall stability is observed, giving further support to our
findings on the group level. However, in one sub-function, phonemic
fluency, patients fluctuate more and show a pattern of deterioration.
Overall our results are in accordance with the notion that ex-
ecutive deficits are present at the start of first treatment and remain
480 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481
relatively stable over the first year in FEP patients (Rodriguez-San-
chez et al., 2008; Szoke et al., 2008). In line with previous research
(Rodriguez-Sanchez et al., 2008: Bozikas and Andreou, 2011), the
patients scored on average Z1 SD below the healthy controls across
the various tasks with the largest impairments in inhibitory control
and fluency. The Inhibition condition, measuring inhibitory control,
has previously been shown to best discriminate between patient-
control samples (Holmen et al., 2012a), and sematic fluency has been
suggested to be a possible endophenotype being the most impaired
executive sub-function in schizophrenia (Szoke et al., 2008). In
phonemic fluency we find that patients deteriorate compared to
controls, somewhat contrary to previous findings reporting stability
or minor improvements in this domain (Szoke et al., 2008). Based on
our finding this subtest might serve as a potential sensitivity marker
of morbidity in longitudinal studies of schizophrenia, and should be
replicated in a larger population. Further, we find small improve-
ments in two sub-functions, inhibitory control and on one measure
of flexibility, also in line with previous research (Rodriguez-Sanchez
et al., 2008). However the effect of short-term improvements in
healthy populations after repeated measurements needs to be taken
into consideration, since such short intervals are often thought to
reflect practice effects. The extent to which observed improvements
in patients can be attributed practice effects also needs to be in-
vestigated. Cognitive improvements in FEP in early years may be
more related to practice than to real cognitive improvements
(Goldberg et al., 2007). Rodriguez-Sanchez et al. (2008) attributed all
their observed changes in non-affective FEP during the first year to
practice effects, rather than real cognitive enhancement since con-
trols showed the same pattern. Regarding our findings on the WCST,
few studies have investigated longitudinal changes on this test, and
results have been mixed. In a recent paper Ekerholm and colleagues
(2012) found no change in patients or controls in a five year follow-
up study (Ekerholm et al., 2012). Contrary to this, Basso and collea-
gues report practice effects on several WCST indices in their 1-year
follow-up study of healthy controls. They argue that prior knowledge
of the test may lead to recollection of prior strategies rather than
more efficient problem-solving (Basso et al., 2001). Whether the
observed changes reported in the current study reflect real changes,
changes due to practice effects or psychometric properties of the test
remains unclear. Regarding the Inhibition condition, small improve-
ments have been observed in schizophrenia samples previously
(Szoke et al., 2008). However, one could argue that there is little
advantage of prior strategy learning on this test, it has good test-
retest reliability (consistency of change within the groups), and thus
is well suited for repeated measurements.
When controlling for baseline differences in IQ level, results
mainly remained unaffected. The only exception was for the Digit
Span backward test of working memory where the group differ-
ence turned non-significant. This could be due to the smaller
sample size, but more presumably has to do with the removal of
higher achieving controls and the lowest preforming patients.
Having a sample weighted towards average performing controls
and higher performing patients may be misleading for several
reasons (Hoff et al., 2005). Patients are in a critical period, early in
the course of their illness, and removing the poorest performing
patients could in effect mean removing those that would change
the most over the next years. Thus, valuable information would be
missed. Also, IQ is usually lower in patient samples and we may be
interfering with factors that are inherent to the illness. Never-
theless, it does not appear that the observed group differences
over time could be explained by differences in baseline IQ.
4.1. Individual changes in performance
The individual trajectories in the control group suggest an
overall stability in executive sub-functions. There was however a
significant association between group affiliation and stability on
one measure of working memory where controls showed a decline
in their performance at follow-up. Closer investigation of these
participants revealed that their initial baseline performance was
very high, probably reflecting regression towards the mean. In the
FEP group we see a slight decline on one test from each of the
domains working memory and fluency, as well as the inhibitory
control domain (Z10% decline), but there is no pattern of global
decline with the patients scoring lower on several of these tests.
There was one significant association between group affiliation
and stability in phonemic fluency. Sixteen percent of all patients
preformed reliably poorer at follow-up. Closer inspection reveals
that these patients initially performed quite well, although still
below controls. For flexibility (WCST, perseverative responses) we
see the opposite pattern. Those who improve (in both groups)
initially had very poor performance. In inhibitory control we found
the most consistent impairment in the patient group at both as-
sessments. However, on the individual level a large number of
patients either improve or deteriorate. The controls are more
stable in comparison. It appears that initial high and low scores are
correlated with less and more improvements at follow up, again,
indicating regression towards the mean. Overall, we see that the
individual trajectories mostly display stabile functioning over
1 year. Both groups show some fluctuations in both directions over
time in accordance with our findings on the group level. The ob-
served interaction effect in phonemic fluency is further supported
by a reliable decline in the patient group on the individual level.
4.2. Strengths and limitations
The main strength of this study is the inclusion of a broad range
of executive tests. Further, the large age and gender matched
control sample recruited from the same catchment area in the
same time period, is a strength. Lastly, the individual trajectories
calculated with RCI enable a more thorough investigation of ex-
ecutive sub-functions, revealing information that may otherwise
be overlooked. A limitation of the study is the short follow-up
period. Nevertheless, the first years appear to be particularly im-
portant with regard to the developmental course of psychotic ill-
ness (Rund et al., 2007; Barder et al., 2013a).
4.3. Conclusion
These findings suggest that FEP patients have poorer executive
functioning than controls at treatment initiation and over the first
year. Some executive sub-functions vary across time in both
groups, but an overall stability is observed. The individual trajec-
tories lend further support to our findings on the group level. The
observed change in phonemic fluency is supported by an actual
decline in performance in the patient group, suggesting that this
function may be a sensitive marker of morbidity in schizophrenia.
Patients are in a critical phase while recovering from a psy-
chotic episode. However the fact that executive functioning re-
mains stable and thus not deteriorate further is positive. Still,
these patients preform much lower than the healthy population
and restorative and compensative measures should be pursued.
For future research we recommend that executive functioning
should be considered a target for interventions aimed at improv-
ing cognition
Acknowledgements
In addition to the clinicians and psychologist who contributed
with patient recruitment and assessment, we would like to thank
 B. Haatveit et al. / Psychiatry Research 228 (2015) 475–481
all the patients and volunteers for their participation in the study.
Special thanks go to Thomas Bjella, June Lystad, Carmen Simonsen,
Francesco Bettella and Gro Strømnes Dybedal for advice and as-
sistance in preparation of the manuscript.
This study was Granted by the Norwegian Research Council
(#421716, #223273), the Regional Health Authority South-Eastern
Norway (#N1, #2011085, #2013123, #52026) and from the K.G.
Jebsen Foundation (#SKGJ-2011-36).
Appendix A. Supplementary material
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.psychres.2015.05.
060.
References
Aas, M., Dazzan, P., Mondelli, V., Melle, I., Murray, R.M., Pariante, C.M., 2014. A
systematic review of cognitive function in first-episode psychosis, including a
discussion on childhood trauma, stress, and inflammation. Frot. Psychiatry 4,
182. http://dx.doi.org/10.3389/fpsyt.2013.00182.
Addington, J., Saeedi, H., Addington, D., 2005. The course of cognitive functioning in
first episode psychosis: changes over time and impact on outcome. Schizo-
phrenia. Res. 78 (1), 35–43.
Barder, H.E., Sundet, K., Rund, B.R., Evensen, J., Haahr, U., Ten Velden Hegelstad, W.,
Joa, I., Johannessen, J.O., Langeveld, H., Larsen, T.K., Melle, I., Opjordsmoen, S.,
Rossberg, J.I., Simonsen, E., Vaglum, P., McGlashan, T., Friis, S., 2013a. Neuro-
cognitive development in first episode psychosis 5 years follow-up: associa-
tions between illness severity and cognitive course. Schizophr. Res. 149 (1–3),
63–69. http://dx.doi.org/10.1016/j.schres.2013.06.016.
Barder, H.E., Sundet, K., Rund, B.R., Evensen, J., Haahr, U., Ten Velden Hegelstad, W.,
Joa, I., Johannessen, J.O., Langeveld, H., Larsen, T.K., Melle, I., Opjordsmoen, S.,
Rossberg, J.I., Simonsen, E., Vaglum, P., McGlashan, T., Friis, S., 2013b. Ten year
neurocognitive trajectories in first-episode psychosis. Front. Hum. Neurosci. 7,
643, doi0.3389/fnhum.2013.00643.
Basso, M.R., Lowery, N., Ghormley, C., Bornstein, R.A., 2001. Practice effects on the
Wisconsin card sorting test-64 card version across 12 months. Clin. Neu-
ropsychol. 15 (4), 471–478. http://dx.doi.org/10.1076/clin.15.4.471.1883.
Bowie, C.R., Harvey, P.D., 2006. Cognitive deficits and functional outcome in schi-
zophrenia. Neuropsychiatr. Dis. Treat. 2 (4), 531–536.
Bozikas, V.P., Andreou, C., 2011. Longitudinal studies of cognition in first episode
psychosis: a systematic review of the literature. Aust. N. Z. J. Psychiatry 45 (2),
93–108. http://dx.doi.org/10.3109/00048674.2010.541418.
Chelune, G.J., Naugle, R.I., Luders, H., Sedlak, J., Awad, I.A., 1993. Individual change
after epilepsy surgery: practice effects and base-rate information. Neu-
ropsychology 7 (1), 41–52.
Cohen, A.S., Couture, S.M., Blanchard, J.J., 2012. Neuropsychological functioning and
social anhedonia: three-year follow-up data from a longitudinal community
high risk study. J. Psychiatr. Res. 46 (7), 898–904. http://dx.doi.org/10.1016/j.
jpsychires.2012.03.020.
Delis, D.C., Kramer, J.H., Kaplan, E., Ober, B.A., 2005. California Verbal Learning Test
(CVLT-II). Pearson Assessment, Stockholm.
Delis, D.C., Kaplan, E., Kramer, J.H., 2005. Delis-Kaplan Executive Function System
(D-KEFS) Norwegian Manual. Pearson Assessment, Stockholm.
Duff, K., 2012. Evidence-based indicators of neuropsychological change in the in-
dividual patient: relevant concepts and methods. Arch. Clin. Neuropsychol. 27
(3). http://dx.doi.org/10.1093/arclin/acr120 248-26.
Ekerholm, M., Firus Waltersson, S., Fagerberg, T., Soderman, E., Terenius, L., Agartz,
I., Jönsson, E.G., Nyman, H., 2012. Neurocognitive function in long-term treated
schizophrenia: a five-year follow-up study. Psychiatry Res. 200 (2–3), 144–152.
http://dx.doi.org/10.1016/j.psychres.2012.05.008.
First, M., Spitzer, R., Gibbon, M., Williams, J.B.W., 1995. Structured Clinical Interview
for DSM-IV Axis I Disorders: Patient Edition (SCID 1/P, Version 2.0). Biometrics
Research Department, New York State Psychiatric Institute, New York, NY.
Frangou, S., 2010. Cognitive function in early onset schizophrenia: a selective re-
view. Front. Hum. Neurosci. 3, 79. http://dx.doi.org/10.3389/neuro.09.079.2009.
Goldberg, T.E., Goldman, R.S., Burdick, K.E., Malhotra, A.K., Lencz, T., Patel, R.C.,
Woerner, D.G., Schooler, N.R., Kane, J.M., Robinson, D.G., 2007. Cognitive
481
improvement after treatment with second-generation antipsychotic medica-
tions in first-episode schizophrenia: is it a practice effect? Arch. Gen. Psychiatry
64 (10), 1115–1122. http://dx.doi.org/10.1001/archpsyc.64.10.1115.
Heaton, R.K., Gladsjo, J.A., Palmer, B.W., Kuck, J., Marcotte, T.D., Jeste, D.V, 2001.
Stability and course of neuropsychological deficits in schizophrenia. ARCH GEN
PSYCHIATRY 58 (1), 24–32.
Henry, J.D., Crawford, J.R., 2005. A meta-analytic review of verbal fluency deficits in
schizophrenia relative to other neurocognitive deficits. Cogn. Neuropsychiatry
10 (1), 1–33. http://dx.doi.org/10.1080/13546800344000309.
Hoff, A.L., Svetina, C., Shields, G., Stewart, J., Delisi, L.E., 2005. Ten year longitudinal
study of neuropsychological functioning subsequent to a first episode of schi-
zophrenia. Schizophr. Res. 78 (1), 27–34.
Holmen, A., Juuhl-Langseth, M., Thormodsen, R., Sundet, K., Melle, I., Rund, B.R.,
2012a. Executive function tests in early-onset psychosis: which one to choose?
Scand. J. Psychol. 53 (3), 200–205. http://dx.doi.org/10.1111/
j.1467-9450.2012.00940.x.
Holmen, A., Juuhl-Langseth, M., Thormodsen, R., Ueland, T., Agartz, I., Sundet, K.,
Andreassen, O.A., Rund, B.R., Melle, I., 2012b. Executive function in early- and
adult onset schizophrenia. Schizophrenia Research 142 (1-3), 177–182. http:
//dx.doi.org/10.1016/j.schres.2012.10.006.
Iverson, G.L., 2001. Interpreting change on the WAIS-III/WMS-III in clinical samples.
Arch. Clin. Neuropsychol. 16 (2), 183–191.
Jackson, C.T., Fein, D., Essock, S.M., 2001. The effects of cognitive impairment and
substance abuse on psychiatric hospitalizations. Community Ment. Heal. J. 37,
303–312.
Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale
(PANSS) for schizophrenia. Schizophr. Bull. 13 (2), 261–276.
Kongs, S.K., Thompson, L.L., Iverson, G.L., Heaton, R.K., 2000. Wisconsin Card
Sorting Test-64 Card Version (WCST-64). Professional manual. Psychological
Assessment Resources, Odessa, Florida.
Liu, K.C., Chan, R.C., Chan, K.K., Tang, J.Y., Chiu, C.P., Lam, M.M., Chan, S.K., Wong, G.
H., Hui, C.L., Chen, E.Y., 2011. Executive function in first-episode schizophrenia:
a three-year longitudinal study of an ecologically valid test. Schizophr. Res. 126
(1–3), 87–92. http://dx.doi.org/10.1016/j.schres.2010.11.023.
Mayoral, M., Zabala, A., Robles, O., Bombin, I., Andres, P., Parellada, M., Moreno, D.,
Graell, M., Medina, O., Arango, C., 2008. Neuropsychological functioning in
adolescents with first episode psychosis: a two-year follow-up study. Eur.
Psychiatry 23 (5), 375–383. http://dx.doi.org/10.1016/j.eurpsy.2008.01.1420.
Mcgurk, S.R., Mueser, K.T., 2003. Cognitive functioning and employment in severe
mental illness. J. Nerv. Ment. Dis. 191 (12), 789–798.
McKee, M., Hull, J.W., Smith, T.E., 1997. Cognitive and symptom correlates of par-
ticipation in social skills training groups. Schizophr. Res. 23, 223–229.
Parsons, T.D., Notebaert, A.J., Shields, E.W., Guskiewicz, K.M., 2009. Application of
reliable change indices to computerized neuropsychological measures of con-
cussion. Int J Neurosci 119 (4), 492–507 http://dx.doi.org/10.1080/
00207450802330876.
Pedersen, G., Hagtvet, K.A., Karterud, S., 2007. Generalizability studies of the global
assesment of functioning-split version. Compr. Psychiatry 48, 88–94.
Robinson, D.G., Woerner, M.G., Alvir, J.M., 2002. Predictors of medication dis-
continuation by patients with first-episode schizophrenia and schizoaffective
disorder. Schizophr. Res. 57, 209–219.
Rodriguez-Sanchez, J.M., Perez-Iglesias, R., Gonzalez-Blanch, C., Pelayo-Teran, J.M.,
Mata, I., Martinez, O., Sanchez-Cubillo, I., Vazquez-Barquero, J.L., Crespo-Fa-
corro, B., 2008. 1-year follow-up study of cognitive function in first-episode
non-affective psychosis. Schizophr. Res. 104 (1–3), 165–174. http://dx.doi.org/
10.1016/j.schres.2008.05.020.
Rund, B.R., Melle, I., Friis, S., Johannessen, J.O., Larsen, T.K., Midboe, L.J., Opjords-
moen, S., Simonsen, E., Vaglum, P., McGlashan, T., 2007. The course of neuro-
cognitive functioning in first-episode psychosis and its relation to premorbid
adjustment, duration of untreated psychosis, and relapse. Schizophr. Res. 91 (1–
3), 132–140. http://dx.doi.org/10.1016/j.schres.2006.11.030.
Rund, B.R., 2009. Is there a degenerative process going on in the brain of people
with schizophrenia. Front. Hum. Neurosci. 3, 36. http://dx.doi.org/10.3389/
neuro.09.036.2009.
Sanchez-Torres, A.M., Basterra, V., Moreno-Izco, L., Rosa, A., Fananas, L., Zarzuela, A.,
Peralta, V., Cuesta, M.J., 2013. Executive functioning in schizophrenia spectrum
disorder patients and their unaffected siblings: a ten-year follow-up study.
Schizophr. Res. 143 (2–3), 291–296. http://dx.doi.org/10.1016/j.
schres.2012.11.026.
Szoke, A., Trandafir, A., Dupont, M.E., Meary, A., Schurhoff, F., Leboyer, M., 2008.
Longitudinal studies of cognition in schizophrenia: meta-analysis. Br. J. Psy-
chiatry 192 (4), 248–257. http://dx.doi.org/10.1192/bjp.bp.106.029009.
Wechsler, D., 2003. Wechsler adult intelligence scale (WASI-III). Norwegian Man-
ual, 3rd ed. Pearson Assessment, Stockholm.