Unit 5 Learning Activity
• Using the articles gathered for the unit 4 learning activity, summarize the studies,
similar to the introductory section of a research paper. You can find five different
articles if you prefer
• Follow this with a research question you would want to study that follows from
these studies and a hypothesis to accompany the research question.
• Write the Methods Section of your paper- consider who would be in the study,
how they would be recruited, and the study measure you would use
• Add a cover page and reference page to your paper. Using APA format, also
include a Running Head
Let’s talk through an example…
• Let’s say your overarching topic was treatments for autism and you
found articles on ABA, Pivotal Response, and LEND. You would
summarize these articles and INTEGRATE them (what were
overarching similarities/differences in the methods? What were the
overall findings?). The next slides will show an example of how this
can be done.
One sample paragraph
• Different studies have assessed the efficacy of different treatments
for autism, finding some support for a range of treatments, such as
ABA, Pivotal Response, and LEND. These results were found in
samples ranging in age from 3-15, including boys and girls. The
predominant ethnicity of participants in the studies was Caucasian (A,
2015, B, 2014, C, 2000, D, 2005, E, 1999).
Sample next paragraph
• A (2015) and B (2014) found that LEND was associated with increased
social interaction and daily functioning in children and adolescents in
their studies. A (2015) used a sample of 100 adolescents and B
(2014) used a sample of 50 fifth graders. D (2005) found that pivotal
response increased verbal ability in a sample of 3rd graders. C (2000)
and E (1999) showed that ABA was effective at teaching adaptive
skills and reducing aggressive behavior. C (2000) used a sample of 3
year olds and E (1999) used a sample of 12 year olds.
Sample next paragraph
• Overall, from these studies, it seems that different therapies can be effective in
different ways. These studies did not compare these three treatment techniques
in terms of their effectiveness on a single behavior. As such, it could be
interesting to explore whether ABA, Pivotal Response, or LEND is most effective
with a specific behavior, such as with increasing the range of foods eaten by
autistic children who have a restricted range of foods that they will eat. Thus, for
the following study, I will explore the question: is ABA, Pivotal Response, or LEND
most effective at increasing the range of foods eaten in autistic children who have
a restricted range of foods that they will eat? My hypothesis is that all three
treatments will show some efficacy, with ABA showing the most effectiveness in
increasing the range of foods that these children will eat.
Sample methods section
• For this study, I will recruit 150 children, aged 5-7, who are diagnosed with autism
and who have a restricted range of foods that they will eat. I will send fliers
about this study to special education schools that focus on children with autism
within Syracuse, providing contact information for interested parents. Parents
will undergo a brief phone screening, assessing when and how their child was
diagnosed with autism, the current foods their child will eat, and any medical or
dietary restrictions. Children who eat fewer than 15 different foods and who have
a valid diagnosis of autism will be included in the study. Children whose parents
are implementing dietary restrictions or who have medical issues that restrict the
foods they can eat will be excluded from the study.
Sample methods section continued
• Once 150 children are enrolled in the study, they will be randomly assigned to receive
ABA, Pivotal response, or LEND treatment focused on increasing the range of foods they
will eat, with 50 children assigned to each group. Foods eaten will be tracked through a
food diary that parents and teachers will complete daily. Baseline data will be collected
for two weeks prior to the start of treatment, with parents asked to complete the food
diaries daily starting two weeks prior to receiving the treatment. The children will receive
treatment for 3 weeks and parents will be asked to continue to complete the food diaries
for 2 weeks after the end of treatment. To compensate parents for their time and effort,
and to encourage continued participation in the study, treatment will be provided at no
cost and participating families will receive $10/week. Teachers who participate in the
study for enrolled children will receive $50 toward new classroom supplies at the end of
the study.
Sample methods continued
• I will count the different types of food eaten per week and will run
analyses to see if the number of different foods eaten during the first
two weeks of data collection differed from the number of different
foods eaten during the last two weeks of data collection. Additional
analyses on different foods eaten while the treatment was ongoing
will also be conducted to explore the rate of change in the different
groups.
From your Unit 5 Learning Activity, it should
be clear:
• Overall, what was found from prior studies- give the reader the big picture.
DON’T list studies (study A did…. And found…. Study B did…. And found….Study C
did…. And found…)- that is not integration!
• What one might be left curious about- what more is there to know on the topic?
• Your question and hypothesis related to what more there is to know on the topic
• An overview of how you would structure a study of this question (Consider: who
would be in it? How recruited? What data would you collect and how? How long
would the study last? Would there be any incentives? What might you do with
the data?)
Neuropsychology
2017, Vol. 31, No. 2, 209 –219
© 2016 American Psychological Association
0894-4105/17/$12.00 http://dx.doi.org/10.1037/neu0000309
The Effect of Positive Symptoms on Social Cognition in First-Episode
Schizophrenia Is Modified by the Presence of Negative Symptoms
Vibeke Bliksted
Poul Videbech and Birgitte Fagerlund
Aarhus University Hospital Risskov, Denmark, and
Aarhus University
Mental Health Centre Glostrup, Denmark
Chris Frith
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Aarhus University and University College London
Objective: There is considerable evidence that patients with schizophrenia have neurocognitive and social–
cognitive deficits. It is unclear how such deficits in first-episode schizophrenia relate to current clinical
symptoms. Method: Fifty-nine patients with first-episode schizophrenia (FES) were tested using the Danish
version of NART (premorbid IQ), subtests from WAIS-III (current IQ), and global cognition using Brief
Assessment of Cognition in Schizophrena (BACS), a neurocognitive test battery. Social perception was tested
using film clips of everyday interactions (TASIT). Theory of mind (ToM) was tested using silent animations
(Animated Triangles Task). The FES subjects had been experiencing psychotic symptoms for several years
(mean duration 9.5 years 95% confidence interval (CI [7.6;11.3]). The FES patients were divided into clinical
subgroups based on their level of positive and negative symptoms (using SANS and SAPS). Healthy controls
were matched to the patients. Results: High levels of negative symptoms were associated with low estimated
functional IQ and poor neurocognition and social cognition. All SANS subscales, but Avolition-Apathy, had
significant negative impact on social cognition. The effects of positive symptoms were complex. High levels
of delusions were associated with higher premorbid IQ. In the presence of high levels of negative symptoms,
high levels of positive symptoms were associated with the most comprehensive deficits in social perception,
while, in the absence of negative symptoms, high levels of positive symptoms were not associated with such
deficits. Conclusion: The results suggest that social– cognitive training will need to take account of the above
mentioned effects of symptoms.
Keywords: social cognition, neurocognition, negative symptoms, positive symptoms, theory of mind
deficit in the patients’ theory of mind abilities (Frith, 1992). Frith
et al. proposed four subgroups of schizophrenia with different
theory of mind abilities (Corcoran, Cahill, & Frith, 1997; Frith &
Corcoran, 1996; Pickup & Frith, 2001). Most severely affected
were patients with behavioral signs of negative symptoms and/or
incoherence. Patients with paranoid symptoms such as delusion of
persecution, delusions of reference, and third-person hallucinations were thought to be only moderately affected. Patients with
passivity experiences such as delusions of control, thought insertion, and thought broadcasting and finally patients in remission
were thought to have normal theory of mind abilities.
Sprong et al. (Sprong, Schothorst, Vos, Hox, & van Engeland,
2007) looked at subgroups according to Frith et al.’s suggestion in
29 studies of theory of mind in schizophrenia. They found deficits
of theory of mind in all the subgroups, including remitted patients
(patients with no behavioral signs and no positive symptoms on the
day of testing), indicating that these deficits are trait related and
even more comprehensive than in the model from 1992. Patients
with symptoms of disorganization had more severe theory of mind
deficits than the other patient subgroups.
Besides Theory of Mind (ToM), the social– cognitive domains
most affected in schizophrenia are social perception, social knowledge, attributional bias, and emotional processing (Green et al.,
2008; Penn et al., 2008). A meta-analysis of the above-mentioned
social– cognitive domains concluded that ToM and social percep-
In recent years, there has been an increasing interest in the importance of social– cognitive deficits in schizophrenia and how aspects of
social cognition might be able to explain the heterogeneity of symptoms in people with this disorder (Bliksted, Fagerlund, Weed, Frith, &
Videbech, 2014; Fletcher & Frith, 2009; M. F. Green & Leitman,
2008; Green et al., 2008; Penn, Sanna, & Roberts, 2008).
In 1992, Chris Frith proposed a model suggesting that specific
psychotic symptoms in schizophrenia could be explained by a
This article was published Online First November 3, 2016.
Vibeke Bliksted, Psychosis Research Unit, Aarhus University Hospital
Risskov, Denmark, and Interactive Minds Centre, Aarhus University; Poul
Videbech, Mental Health Services, Capital Region of Denmark, Mental
Health Centre Glostrup, Denmark; Birgitte Fagerlund, Mental Health Centre, Mental Health Services, Capital Region of Denmark, Glostrup, and
Lundbeck Foundation Centre for Clinical Intervention and Neuropsychiatric Schizophrenia Research (CINS), Mental Health Centre Glostrup;
Chris Frith, Interactive Minds Centre, Aarhus University, and Leopold
Müller Functional Imaging Laboratory, Wellcome Trust Centre for Neuroimaging, University College London.
Correspondence concerning this article should be addressed to Vibeke
Bliksted, Psychosis Research Unit, Aarhus University Hospital Risskov,
Denmark, Skovagervej 2, 8240 Risskov, Denmark. E-mail: vibeke.bliksted@
ps.rm.dk
209
BLIKSTED, VIDEBECH, FAGERLUND, AND FRITH
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
210
tion were the social– cognitive domains most affected in schizophrenia (Savla, Vella, Armstrong, Penn, & Twamley, 2012).
Recent studies have suggested that social– cognitive deficits in
schizophrenia are trait related and independent of the phase and
duration of the illness (Bliksted, Ubukata, & Koelkebeck, 2016;
Green et al., 2012; Horan et al., 2012). Similar impairments of
social cognition have been found in first-episode and chronic
schizophrenia and their first degree relatives, and no changes in the
social– cognitive deficits were seen at 12 months follow-up in
first-episode schizophrenia (Bora & Pantelis, 2013; Bora, Yucel, &
Pantelis, 2009; Green et al., 2012; Horan et al., 2012).
This adds to the complexity of understanding the interaction between specific social– cognitive domains and the clinical symptoms of
schizophrenia such as psychotic and negative symptoms. It also raises
the question of social– cognitive subgroups in schizophrenia.
Bell et al. found that patients with schizophrenia could be divided
into three subgroups: (a) Patients with high levels of negative symptoms, (b) patients with low levels of negative symptoms but high
social– cognitive scores, and (c) patients with low levels of negative
symptoms and low social– cognitive scores (Bell, Corbera, Johannesen, Fiszdon, & Wexler, 2013). These subgroups were based on a
principal component analysis of the subscales of negative symptoms
and the social– cognitive tests that was followed by a cluster analysis
that specified the above subgroups.
Other researchers argue that there is no simple delineation of
deficit profiles, but instead a complex matrix of correlations between clinical symptoms, neurocognitive domains, social–
cognitive domains, and functional outcome (Bliksted et al., 2014;
Mancuso, Horan, Kern, & Green, 2011). It has furthermore been
argued that patients with schizophrenia show deficits in all aspects
of neurocognitive and social– cognitive domains compared with
healthy controls, but to varying degrees (Green, Horan, & Sugar,
2013). However, a recent review suggests that some reflexive
aspects of social cognition, such as emotion experience, might be
intact in patients with schizophrenia (Green, Horan, & Lee, 2015).
In this study, we examined 59 patients with first-episode schizophrenia to account for the correlations between aspects of negative
and positive symptoms as well as neurocognition and social cognition. For this purpose we divided the patients into four groups
with high versus low levels of negative symptoms, and high versus
low levels of positive symptoms. A healthy control group was
included.
It was hypothesized that patients in general would perform
worse than the healthy control subjects on all neurocognitive and
social– cognitive tasks. We expected the social– cognitive deficits
to be most pronounced in patients with many negative symptoms
(Bell et al., 2013; Frith, 1992), and in very paranoid patients (Frith,
1992).
Method
Subjects
The patients were recruited from OPUS which is an intensive
2-year assertive community treatment program for first-episode
schizophrenia (Bertelsen et al., 2008). The patients had recently
been diagnosed with first-episode schizophrenia (FES) by experienced psychiatrists according to ICD-10 criteria. The patients were
included during two time periods: from January 1, 2009 until
February 1, 2010, and from August 1, 2011 until August 1, 2012
as consecutive parts of the same study. Some of the data from time
period one have been used in statistical analyses that differ from
the analyses in this article (Bliksted et al., 2014). Healthy control
subjects were recruited via advertisements in four local newspapers and were tested during the same time period as their patient
match.
FES Subjects
The first-episode schizophrenia (FES) patients were from 18 to
34 years old. Thirty-six patients were included during 2009 and
2010. None of the patients had received any antipsychotic medication for more than 3 months before inclusion. Twenty-three
patients were included during 2011 and 2012. These patients were
either antipsychotic-naïve or had received no more than 6 weeks’
lifetime psychopharmacological treatment at inclusion.
Twenty-nine from the entire group of patients received atypical
antipsychotics; 13 atypical antipsychotics and antidepressants; 4
atypical antipsychotics and other medication; 1 typical antipsychotic and other medication; 1 atypical antipsychotic, antidepressant, and benzodiazepine; 1 antidepressant; 1 patient had stopped
taking his atypical antipsychotic two weeks ago; and 9 patients
were antipsychotic-naïve. Other medication was for example, insulin and gastric ulcer medication.
Exclusion criteria were a history of neurological disorder, severe
head trauma, or drug- and alcohol dependency according to ICD10. Patients were excluded if they did not understand spoken
Danish sufficiently to understand the testing procedures or if they
had an estimated premorbid IQ below 70 based on their history.
Healthy Control Subjects
The healthy controls and patients were individually matched to
the patients based on age, gender, handedness, educational level
(based on the last commenced education), community of residence, and parental socioeconomic status (based on the highest
parental education and expected parental income regarding wages
according to Statistics Denmark). Parental educational level (rather
than the patients’s educational level) was chosen as the matching
variable because the onset of a mental illness interferes with
education. Thus, we presumed that the parental educational level
would reflect the potential level of function of the FES patients
more presisely. The parental socioeconomic groups were divided
into high (N ⫽ 19; 32.20%), middle (N ⫽ 30; 50.85%), and low
(N ⫽ 10; 16.95%). In two cases the patients and healthy control
subjects did not match perfectly one-to-one. This meant that there
was one more female in the healthy subject population than in the
FES patient group.
The 59 healthy control subjects had no history of mental illness
and neither had their first-degree relatives. They did not have a
history of neurological illness, severe head injury or drug- or
alcohol dependence according to ICD-10 criteria. This was controlled for by blood and urine tests.
Ethics
The participants received written and oral information about the
project, and written informed consent was obtained before inclu-
SOCIAL COGNITIVE SUBGROUPS IN SCHIZOPHRENIA
sion. The study was approved by The Central Denmark Region
Committees on Biomedical Research Ethics (Ref: M-2009 – 0035)
and the Danish Data Protection Agency. The project complied with
the Helsinki-II-declaration.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Measures
Social– cognitive measures.
Social perception. We used five sincere and five simple sarcastic film clips from the Danish translation of TASIT (The
Awareness of Social Inference Test—Part A2 Social Inference
minimal; Bliksted et al., 2014). The test is composed of small
video clips with professional actors performing everyday interactions (McDonald et al., 2006; McDonald, Flanagan, Rollins, &
Kinch, 2003). The clips last for 16 –53 s. After seeing each clip one
has to figure out if the people were being sincere or sarcastic based
on paralinguistic cues such as tone of voice, facial expressions, and
verbal content. Each clip was followed by four questions concerning the communicative intentions of the persons (what they were
doing, saying, thinking, and feeling; max score of each clip was 4
points). The TASIT film clips can also be analyzed as a signal
detection task, that is, how good the subjects are at distinguishing
sarcastic from sincere behavior. For this purpose we developed a
sensitivity score. We also constructed a bias score that measured
whether subjects were neutral (⫽0) or biased toward interpreting
the clips sincerely (⬎0) or sarcastic (⬍0). The TASIT film clips
were translated and back-translated for the purpose of this project
in agreement with Pearson Assessment in London.
Theory of mind. The Animated Triangles (Abell, Happé, &
Frith, 2000; Castelli, Happe, Frith, & Frith, 2000) were used to
measure ToM. We used four random clips where the two triangles
had no intended interaction (e.g., bouncing about) and four ToM
clips where the triangles interacted intentionally (e.g., the large
triangle trying to persuade the small triangle to come outside). The
duration of the animations is 38 – 41 s. The subjects were asked to
interpret what was going on in each animation and their answers
were recorded, transcribed, and scored according to the level of
intentionality (degree of mental state attribution, range 0 –5 per
animation, where a score of 4 or 5 meant use of words describing
different degrees of mental interaction), and appropriateness (accuracy of the description, range 0 –3 per animation, where 3 meant
a perfect description of the interaction intended by the animator;
Castelli et al., 2000). Also target items were recorded referring to
whether the right type of description was used to tell the degree of
intentionality being used in the animations (e.g., score 4 or 5 in
intentionality of the ToM clips and score 0 or 1 in the random
clips), no matter if the subject’s answer was appropriate or not
(max score of 4 per type of animation; Russell, Reynaud, Herba,
Morris, & Corcoran, 2006).
VB and a research student (MLT) evaluated each answer from
the 59 FES patients and 59 controls separately, and their mean
scores were calculated. Interrater agreement was substantial to
almost perfect (Random animations: ⫽ 0.75; ToM animations:
⫽ 0.84; ToM Appropriateness: ⫽ 0.63; Random Appropriateness: ⫽ 0.72).
Neuropsychological measures.
Neurocognition. We used the Danish translation of the Brief
Assessment of Cognition in Schizophrenia (BACS-DK), to measure verbal memory, working memory, motor speed, verbal flu-
211
ency, executive functions, attention, and speed of processing
(Keefe et al., 2004, 2008; Keefe, Poe, Walker, & Harvey, 2006).
Intelligence. Current functional intelligence was estimated by
the four following subtests from WAIS-III (Wechsler Adult Intelligence Scale, Third version; Wechsler, 1997), chosen because of
their high correlation with the total WAIS-III IQ score: Matrix
Reasoning, Block Design, Vocabulary, Similarities (Wechsler,
1997).
Premorbid intelligence was estimated using DART (Danish
Adult Reading Test), which is the Danish version of the NART
(The Nelson Adult Reading Test; Nelson & O’Connell, 1978).
Psychopathology, Clinical Measures, and
Drugs Screening
During inclusion at the OPUS clinic, all FES patients were
interviewed with the PSE-interview (Present State Examination,
ICD-10) by experienced psychiatrists regarding schizophrenia and
drug dependence (WHO, 1994). All healthy control subjects were
interviewed by VB using the entire PSE interview. All subjects
were SANS and SAPS rated by VB (Andreasen, 1984a, 1984b).
SANS (Scale for the Assessment of Negative Symptoms) consists
of 25 items divided into 4 subscales: Affective flattening or blunting, Alogia, Avolition-Apathy, Anhedonia-Asociality. SAPS
(Scale for the Assessment of Positive Symptoms) consists of 34
items which measure 4 psychotic domains: Hallucinations, Delusions, Bizarre behavior, Positive formal thought disorder. The
severity of each symptom is rated from 0 –5 and an overall global
score of each subscale is found (range 0 –5). Urine samples were
collected on the day of psychological testing measuring use of
amphetamine, benzodiazepines, cannabis, codeine, morphine, and
cocaine. The subjects tested during 2009 and 2010 were also tested
for potential alcohol dependence by blood samples (P-Carbohydrat-Deficient-Transferrin).
Data Analysis
Statistical analyses were carried out with Stata IC 11.2
software. The data were examined for distribution and outliers.
Continuous variables were examined by Wilcoxon’s rank sum test
(Mann–Whitney) and effect size in terms of Harrell’s C. Harrell’s
C is a rank parameter measuring the ordinal predictive power of a
model. Categorical variable were examined by Fisher’s exact test
and reported with the counts and proportions of the group total.
Means and 95 or 99% confidential intervals plus effect sizes were
calculated in accordance with the data distribution of each variable.
Composite scores of the BACS-DK were calculated as the
weighted mean of z-scores, separately computed for each subtest
relative to the mean and standard deviation of the healthy control
subjects (M ⫽ 0, SD ⫽ 1) as described by Keefe et al. (Keefe et
al., 2004). The same procedure was followed regarding the calculation of the social– cognitive composite score.
Correlations between social cognition, neurocognition, IQ, and
symptoms in the FES patients were calculated by Spearman’s rank
correlation (because not all variables were normally distributed).
Social– cognitive patient subgroups were found by dividing the
59 patients into subgroups on the basis of SANS and SAPS. Data
were split at the mean (SANS: ⱕ10 vs. ⬎10; SAPS: ⱕ12 vs. ⬎12)
BLIKSTED, VIDEBECH, FAGERLUND, AND FRITH
212
resulting in four groups. Results were adjusted for multiple testing
using Bonferroni corrections.
Comparison between symptom subgroups within patients and
controls was done using Kruskal-Wallis rank test. These p values
were from 3 df 2 tests corrected. Subgroup analyses were performed by comparing patients with their matched controls on a
group basis. This procedure lead us to the exclusion of two control
subjects who could not be directly matched to a patient (Table 1).
Correlations Between SANS/SAPS Subscales and
Neurocognition, Social Cognition, and IQ
Results
Demographics, Psychopathology, and IQ
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
no correlation between positive symptoms (SAPS total score) and the
social– cognitive composite score ( ⫽ 0.14, p ⫽ .29).
The application of TASIT as a detection task showed that the
patients were poor at making the distinction between sincerity and
sarcasm (TASIT sensitivity), t(96.37) ⫽ 4.33, p ⬍ .0001; d ⫽ 0.80
95% CI (0.42;1.17), but also had a bias toward treating the sarcastic behavior as sincere (TASIT bias), t(116) ⫽ ⫺2.77, p ⫽
.0066; d ⫽ ⫺0.51 95% CI (⫺0.87; ⫺0.14; Figure 2).
Demographics and psychopathology are summarized in Table 2.
Patients and controls did not differ in age and sex. The age range in
the FES subjects was 18 –34 years with a mean age of 22.9 years.
Twenty-three of the patients were women (39%). As expected, the
educational level of the patients was lower, probably because the
illness had interfered with their educational plans. The patients reported having had psychotic symptoms on average 9.5 years before
inclusion. As expected, the FES patients had lower premorbid IQ and
lower estimated current functional IQ than the controls.
Neurocognition
The FES patients performed significantly worse in all 6 BACS
subtests compared with the healthy controls (p ⬍ .0003). The
BACS composite score was 1.80 SDs below the healthy controls
(p ⫽ 1.6e-9; Harrell’s C (0.83) 99% CI (0.73;0.92)). There was a
significant negative correlation between negative symptoms
(SANS total score) and neurocognition ( ⫽ ⫺0.36, p ⫽ .005), but
no correlation between positive symptoms (SAPS total score) and
neurocognition ( ⫽ 0.16, p ⫽ .22; Figure 1).
Social Cognition
A social– cognitive composite score was derived based on the
statistically significant subscores of the social– cognitive tests. This
composite score was 1.91 SDs below the mean of the healthy controls
(p ⫽ 2.2e-11; Harrell’s C 0.86 99% CI (0.77;0.95), which is remarkably similar to the BACS composite score. There was a significant
negative correlation between negative symptoms (SANS total score)
and the social– cognitive composite score ( ⫽ ⫺0.39; p ⫽ .002), but
Table 1
FES Subgroups Based on SANS and SAPS Symptom Scores
Measure
SANS
Measure
Low
High
Total
SAPS
Low
High
Total
17
13
30
12
17
29
29
30
59
Note. FES ⫽ first-episode schizophrenia; SANS ⫽ Scale for Assessment
of Negative Symptoms; SAPS ⫽ Scale for Assessment of Positive Symptoms.
We started our subgroup analyses by looking at correlations
(using Spearmann’s Rank Correlations) between the subscales of
SAPS and SANS and neurocognition, social cognition, and IQ (see
Table 3). In the SAPS subscales, the only significant correlation
was a positive correlation between the global score of delusion and
premorbid IQ (DART) ( ⫽ 0.32, p ⫽ .01), with the presence of
delusions being associated with higher IQ.
The Four SANS Subscores All Showed Numerous
Significant Correlations
Global rating of Affective flattening (e.g., unchanging facial
expression, poor eye contact, and lack of vocal inflections) correlated negatively with the Animated Triangles total Appropriateness scores ( ⫽ ⫺0.30, p ⫽ .02) and the Appropriateness score of
the random animations ( ⫽ ⫺0.33, p ⫽ .01), the Social– cognitive
composite score ( ⫽ ⫺0.29, p ⫽ .03), and estimated functional
IQ (WAIS-III; ⫽ ⫺0.33, p ⫽ .01). However, no correlations
were found between Affective flattening and social perception
(TASIT: Sincere ⫽ ⫺0.03, p ⫽ .83; Simple sarcasm ⫽ ⫺0.16,
p ⫽ .22).
Global rating of Alogia (e.g., poverty of content of speech,
blocking, and increased latency of response) correlated negatively
with several aspects of social perception (TASIT: Sincere
⫽ ⫺0.27, p ⫽ .04; Simple sarcasm ⫽ ⫺0.29, p ⫽ .03;
Sensitivity ⫽ ⫺0.32, p ⫽ .01) and the Animated Triangles total
Appropriateness score ( ⫽ ⫺0.32, p ⫽ .01) and the overall level
of social cognition (Social– cognitive composite score ⫽ ⫺0.28,
p ⫽ .03) and estimated functional IQ (WAIS-III; ⫽ ⫺0.44, p ⬍
.001). Global rating of Avolition-Apathy (e.g., impersistence at
work or school, poor personal hygiene, and physical anergia)
correlated negatively with the neurocognitive BACS composite
score ( ⫽ ⫺0.30, p ⫽ .02).
Finally, the global rating of Anhedonia-Asociality (e.g., problems regarding ability to feel intimacy and closeness and have
relationships with friends and peers) correlated negatively with
social perception (TASIT: Simple sarcasm ⫽ ⫺0.32, p ⫽ .01;
Sensitivity ⫽ ⫺0.31, p ⫽ .02). There was also a negative
correlation with neurocognition (BACS composite score
⫽ ⫺0.38, p ⫽ ⬍0.01), estimated functional IQ (WAIS-III;
⫽ ⫺0.36, p ⬍ .01) and the Social– cognitive composite score
( ⫽ ⫺0.39, p ⬍ .01).
We adjusted the data for multiple comparisons (nine tests) using
the Bonferroni methods after which only four correlations remained significant: Alogia and estimated functional IQ;
Anhedonia-Asociality (AA) and estimated functional IQ; AA and
BACS composite score and AA and the social– cognitive composite score.
SOCIAL COGNITIVE SUBGROUPS IN SCHIZOPHRENIA
213
Table 2
Comparison of Patients With First-Episode Schizophrenia (FES) and Controls on demographics, Psychopathology, and IQ
Age
Females
Years of education
Current occupation
Unemployed
Work
Student
Sick leave
Pension
Vocational training
Days of FES-diagnosis
Years of untreated illness
SANS
SAPS
DART (estimated premorbid IQ)
WAIS-III (estimated functional IQ)
First-episode schizophrenia
(N ⫽ 59)
Healthy controls
(N ⫽ 59)
22.9 [22.0;23.8]
23 (39.0%)
12.1 [11.4;12.7]
23.1 [22.2;23.9]
24 (40.7%)
14.1 [13.4;14.7]
31 (52.5%)
1 (1.7%)
12 (20.3%)
5 (8.5%)
1 (1.7%)
9 (15.3%)
129.1 [98.3; 159.9]
9.5 [7.6; 11.3]c
10.1 [8.8;11.3]
12.1 [10.9; 13.3]
29.2 [27.3; 31.2]d
87.7 [82.7; 92.7]
4 (6.8%)
22 (37.3%)
33 (55.9%)
0
0
0
0
0
0
0
32.9 [31.5; 34.3]
103.0 [99.3; 106.7]
Harrell’s C
p value
.51 (.41–.62)
.81a
1.00b
2.1e-5a
1.7e-15b
.73 (.63–.82)
.66 (.56–.76)
.74 (.65–.83)
2.4e-3a
8.9e-6a
Note. Continuous variables were examined by Wilcoxon Rank-Sum Test (Mann-Whitney) and reported with mean (95% CI, confidence interval) and
effect size by terms of Harrell’s C (95% CI). Categorical variables were examined by Fisher’s Exact Test and reported with the counts and proportions of
group total, N (Percentage). FES ⫽ first-episode schizophrenia; SANS ⫽ Scale for Assessment of Negative Symptoms; SAPS ⫽ Scale for Assessment of
Positive Symptoms; DART ⫽ Danish Adult Reading Test; WAIS-III ⫽ Wechsler Adult Intelligence Scale-III (Matrix Reasoning, Block Design,
Vocabulary, and Similarities).
a
Mann-Whitney test. b Fisher’s exact test. c N ⫽ 58. d N ⫽ 57.
Based on the above findings, we concluded that various aspects
of social cognition, neurocognition, and functional IQ had significant associations with different aspects of negative symptoms in
our FES sample.
FES Subgroups
We compared the patients from the four symptom subgroups
with their matched controls on a subgroup basis. Two patients
were not matched one to one to a control subject which made us
leave out two nonmatched controls from the analysis (see Table 4).
Patients with a combination of high levels of positive symptoms
and high levels of negative symptoms had the most severe deficits in
-4
-2
0
2
As mentioned in the data analysis section, the FES patients were
divided into four subgroups of high and low negative and positive
SANS and SAPS scores.
Combined High Levels of Negative and Positive
Symptoms
-6
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Measure
Z_VeMeTo
Z_Token
Z_SymCod
Z_comp
Z_Digit
Z_VFIuTo
Z_ToLTot
Figure 1. Brief Assessment of Cognition in Schizophrenia (BACS) box and whiskers plot. Performance of first
episode schizophrenia patients on the Danish BACS subtests and composite score standardized to healthy
controls. All differences between patients and controls were statistically significant (p ⬍ 0.0003). See the online
article for the color version of this figure.
BLIKSTED, VIDEBECH, FAGERLUND, AND FRITH
-10
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
-5
0
5
214
Z_TASSin
Z_InteToM
Z_NoejTot
Z_TarRan
Z_TASSim
Z_invInteRan
Z_TarToM
Z_socomp
Figure 2. Social cognitive composite score. Performance of first episode schizophrenia patients on the social
cognitive subtests and composite score standardized to healthy controls. All differences between patients and
controls were statistically significant (p ⬍ 0.01). The Awareness of Social Inference Test (TASIT) Bias scores
were inverted so that higher scores indicated better performances as for all other tests. See the online article for
the color version of this figure.
social perception and ToM, while those with high levels of positive
symptoms and low levels of negative symptoms had the least severe
deficits, and in fact did not differ from their matched controls. In
general, patients with many negative symptoms also had more severe
social– cognitive and neurocognitive deficits. Patients with low levels
of positive and negative symptoms (resembling remission) had more
severe deficits in composite scores of neurocognition and social
cognition compared to their matched controls. A majority of results
remained significant when data were adjusted for multiple comparisons via the Bonferroni methods (see Table 4).
Comparison of Clinical Subgroups
We compared the four clinical subgroups by Wilcoxon’s rank
sum test (see Table 5). Patients dominated by many negative
symptoms had more severe social– cognitive and neurocognitive deficits than the other subgroups. It did not make a difference whether the high level of negative symptoms were accompanied by high levels of positive symptoms or not. On the
contrary, patients dominated by many positive symptoms and
few negative symptoms had better social– cognitive and neuro-
Table 3
Spearman’s Rank Correlations Between Clinical Symptoms, IQ, Social Cognition, and Composite Scores in First-Episode
Schizophrenia Patients
Measure
Intelligence
DART (estimated premorbid IQ)
WAIS-III (estimated functional IQ)
Animated triangles
Appropriateness, total score
Appropriateness, random
TASIT
Sincere
Simple sarcasm
Sensitivity
Composite scores
BACS composite score
Social cognitive composite score
Delusion
(SAPS)
Affective flattening
(SANS)
Alogia
(SANS)
Avolition-Apathy
(SANS)
Anhedonia-Asociality
(SANS)
⫺.20
⫺.33ⴱⴱ
⫺.14
⫺.44ⴱⴱ†
.16
⫺.20
.04
⫺.36ⴱⴱ†
⫺.19
⫺.22
⫺.30ⴱ
⫺.33ⴱⴱ
⫺.32ⴱⴱ
⫺.25
⫺.07
⫺.07
⫺.24
⫺.23
⫺.04
⫺.09
⫺.08
⫺.03
⫺.16
⫺.09
⫺.27ⴱ
⫺.29ⴱ
⫺.32ⴱⴱ
.02
⫺.12
⫺.04
⫺.19
⫺.32ⴱⴱ
⫺.31ⴱ
⫺.25
⫺.29ⴱ
⫺.21
⫺.28ⴱ
⫺.30ⴱ
⫺.26
⫺.38ⴱ†
⫺.39ⴱⴱ†
.32ⴱ
.11
.09
.06
Note. N ⫽ 59. TASIT ⫽ The Awareness of Social Inference Test; SAPS ⫽ Scale for Assessment of Positive Symptoms; SANS ⫽ Scale for Assessment
of Negative Symptoms; DART ⫽ Danish Adult Reading Test; WAIS-III ⫽ Wechsler Adult Intelligence Scale-III (Matrix Reasoning, Block Design,
Vocabulary, and Similarities); BACS ⫽ Brief Assessment of Cognition in Schizophrenia.
†
Bonferroni corrected p-values (p ⬍ .0056) adjusted for nine tests. ⴱ p ⱕ .05. ⴱⴱ p ⱕ .01.
11.5 [10.7; 12.3]
11.4 [10.6; 12.2]
.46 [.16; .75]
18.2 [16.7; 19.6]
18.0 [16.5; 19.5]
.46 [.12; .81]
15.8 [14.0; 17.7]
15.7 [13.5; 17.8]
.47 [.13; .81]
⫺.8 [⫺1.8; .1]ⴱ
⫺.1 [⫺.9; .8]
.69 [.38; 1.0]
⫺.6 [⫺1.5; .4]ⴱ
.2 [⫺.6; 1.0]
.71 [.40; 1.0]
15.7 [13.4; 18.0]ⴱ
18.1 [17.0; 19.2]
.71 [.45; .96]
11.1 [7.7; 14.4]ⴱ
14.9 [11.9; 18.0]
.75 [.51; 1.0]
⫺1.7 [⫺2.7; ⫺.7]ⴱⴱⴱ
.1 [⫺.6; .9]
.86 [.69; 1.0]
⫺1.7 [⫺2.9; ⫺.5]ⴱⴱⴱ
.5 [⫺.3; 1.3]
.85 [.67; 1.0]
98.1 [86.9; 109.2]
104.4 [94.8; 114.1]
.67 [.35; .99]
(NFES
10.9 [9.8; 12.0]
11.6 [11.0; 12.3]
.65 [.42; .88]
89.8 [75.3; 104.4]ⴱ
104.1 [94.1; 114.0]
.71 [.46; .97]
(NFES
Low SANS
High SAPS
⫽ 13; NCON ⫽ 12)
⫺2.4 [⫺4.0; ⫺.9]ⴱⴱ
.2 [⫺1.0; 1.3]
.88 [.67; 1.0]
⫺2.4 [⫺3.9; ⫺.9]ⴱⴱ
⫺.3 [⫺1.5; 1.0]
.83 [.58; 1.0]
12.2 [7.8; 16.6]
14.7 [11.8; 17.6]
.66 [.32; 1.0]
15.8 [13.0; 18.7]ⴱⴱ
18.9 [17.3; 20.5]
.83 [.57; 1.0]
10.4 [8.4; 12.4]
11.3 [10.1; 12.4]
.61 [.30; .91]
80.1 [62.1; 98.1]
92.7 [81.1; 104.2]
.69 [.35; 1.0]
(NFES
High SANS
Low SAPS
⫽ 12; NCON ⫽ 12)
⫺2.6 [⫺4.1; ⫺1.1]ⴱⴱⴱ
.4 [⫺.1; 1.0]
.97 [.90; 1.0]
⫺2.2 [⫺3.5; ⫺1.0]ⴱⴱⴱ
.1 [⫺.5; .6]
.91 [.78; 1.0]
9.4 [4.4; 14.5]ⴱⴱ
16.8 [14.4; 19.1]
.82 [.62; 1.0]
13.2 [9.2; 17.1]ⴱⴱⴱ
18.7 [17.3; 20.1]
.85 [.67;1.0]
10.1 [9.5; 10.8]ⴱ
11.7 [11.3; 12.1]
.71 [.48; .95]
83.1 [70.1; 96.1]ⴱⴱⴱ
108.2 [97.1; 119.2]
.87 [.70; 1.0]
(NFES
High SANS
High SAPS
⫽ 17; NCON ⫽ 16)
.03 (.15)
.72 (1.0)
.041 (.21)
.57 (1.0)
.012 (.062)
.29 (1.0)
.024 (.12)
.20 (1.0)
.31 (1.0)
.69 (1.0)
.058 (.35)
.043 (.26)
Kruskal-Wallis
p value
(adj. p)
Note. Comparison within SANS/SAPS subgroups was done by Wilcoxon Rank-Sum Test and the effect size is given in terms of Harrell’s C (HC). To account for multiple testing we adjusted p values
for five tests with the Bonferroni Method and correspondingly 99% confidence intervals (CIs) are used. The two composite scores are highly correlated and adjusting for six tests would be too
conservative. comparison by Kruskal-Wallis Rank Test between subgroups within patients and controls, respectively. These p values are from 3 df 2 tests corrected for ties. SANS ⫽ Scale for
Assessment of Negative Symptoms; SAPS ⫽ Scale for Assessment of Positive Symptoms. Wechsler Adult Intelligence Scale-III (Matrix Reasoning; Block Design, Vocabulary, and Similarities).
TASIT ⫽ The Awareness of Social Inference Test; BACS ⫽ Brief Assessment of Cognition in Schizophrenia.
Wilcoxon rank-sum test: ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001; Bonferroni corrected p-values adjusted for five tests p ⬍ .01.
Animated triangles
Appropriateness, random
FES
CON
HC
TASIT
Simple sarcasm
FES
CON
HC
Sensitivity
FES
CON
HC
Composite scores
BACS composite score
FES
CON
HC
Social cognitive composite score
FES
CON
HC
Intelligence
WAIS-III (estimated functional IQ)
FES
CON
HC
Mean (99% CI)
Low SANS
Low SAPS
⫽ 17; NCON ⫽ 17)
Sysptom subgroups versus CON
Table 4
Clinical First-Episode Schizophrenia (FES) SANS/SAPS Subgroups Compared With Healthy Controls (CON)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SOCIAL COGNITIVE SUBGROUPS IN SCHIZOPHRENIA
215
.21
1.00
.36 [.05; .67]
.67
1.00
.46 [.15; .76]
.79
1.00
.53 [.21; .84]
.58
1.00
.56 [.24; .88]
.23
1.00
.37 [.06; .68]
.22
1.00
.36 [.06; .66]
.19
.97
.64 [.34; .94]
.37
1.00
.59 [.32; .85]
.042
.21
.72 [.45; .98]
.0025
.013
.82 [.59; 1.00]
.075
.38
.69 [.42; .97]
.057
.28
.71 [.43; .98]
Subgroups
LNLP vs HNLP
.46
1.00
.43 [.14; .71]
.40
1.00
.42 [.13; .70]
.52
1.00
.44 [.15; .72]
.23
1.00
.38 [.11; .65]
.30
1.00
.40 [.14; .66]
.22
1.00
.38 [.10; .65]
.0077
.038
.19 [.00; .45]
.011
.053
.20 [.00; .46]
.029
.14
.24 [.00; .56]
.038
.19
.26 [.00; .56]
.21
1.00
.37 [.07; .66]
.022
.11
.23 [.00; .52]
Subgroups
LNHP vs HNLP
Comparison of symptoms subgroups
Subgroups
LNLP vs HNHP
.018
.090
.24 [.00; .50]
.025
.13
.26 [.00; .51]
.0098
.049
.22 [.00; .48]
.0049
.025
.20 [.00; .43]
.062
.31
.31 [.07; .56]
.024
.12
.26 [.00; .53]
Subgroups
LNHP vs HNHP
.89
1.00
.51 [.18; .85]
.69
1.00
.54 [.22; .87]
.30
1.00
.38 [.08; .68]
.27
1.00
.38 [.08; .68]
.71
1.00
.46 [.23;.69]
.64
1.00
.55 [.21; .89]
Subgroups
HNLP vs HNHP
Note. Comparison within SANS/SAPS subgroups was done by Wilcoxon Rank-Sum Test and the effect size is given in terms of Harrell’s C (HC). To account for multiple testing we adjusted p values
for five tests with the Bonferroni method and correspondingly 99% confidence intervals (CIs) are used. The two composite scores are highly correlated and adjusting for six tests would be too
conservative. Clinical subgroups: LNLP ⫽ Low SANSa ⫹ Low SAPSb (N ⫽ 17); LNHP ⫽ Low SANS ⫹ High SAPS (N ⫽ 13); HNLP ⫽ High SANS ⫹ Low SAPS (N ⫽ 12); HNHP ⫽ High SANS ⫹
High SAPS (N ⫽ 17). SANS ⫽ Scale for Assessment of Negative Symptoms; SAPS ⫽ Scale for Assessment of Positive Symptoms; WAIS-III ⫽ Wechsler Adult Intelligence Scale-III (Matrix
Reasoning; Block Design, Vocabulary, and Similarities); BACS ⫽ Brief Assessment of Cognition in Schizophrenia; TASIT ⫽ The Awareness of Social Inference Test.
Intelligence
WAIS-III (estimated functional IQ)
p
Adj. p
HC
Animated triangles
Appropriateness, random
p
Adj. p
HC
TASIT
Simple sarcasm
p
Adj. p
HC
Sensitivity
p
Adj. p
HC
Composite scores
BACS composite score
p
Adj. p
HC
Social cognitive composite score
p
Adj. p
HC
Mean (99% CI)
Subgroups
LNLP vs LNHP
Table 5
Comparison of the Clinical First-Episode Schizophrenia (FES) SANS/SAPS Subgroups
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
216
BLIKSTED, VIDEBECH, FAGERLUND, AND FRITH
SOCIAL COGNITIVE SUBGROUPS IN SCHIZOPHRENIA
cognitive functioning even than the subgroup with few positive
and negative symptoms.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Discussion
In this study, we found significant differences in social cognition and neurocognition between clinical subgroups of patients
with FES based on the level of their negative and positive symptoms. The results showed that patients with high levels of negative
symptoms had significant cognitive and social– cognitive deficits
irrespective of their level of positive symptoms compared to the
other subgroups of FES patients. FES patients with high levels of
negative symptoms combined with high levels of positive symptoms had the most pronounced difficulties perceiving sarcasm and
discriminating between sincerity and sarcasm. In the presence of
low levels of negative symptoms, high levels of positive symptoms
were associated with the least impairment of all the subgroups. In
fact, the patients dominated by high levels of positive symptoms
and low levels of negative symptoms did not differ from their
matched healthy controls. These results could imply that premorbid social– cognitive and neurocognitive deficits could be a predictor of future development of schizophrenia dominated by negative symptoms but not necessarily accompanied by high levels of
positive symptoms.
Patients with high levels of alogia (e.g., poverty of content of
speech, blocking, and increased latency of response) and
anhedonia-asociality (e.g., problems regarding ability to feel intimacy and closeness and have relationships with friends and peers)
had greater deficits in social perception. Alogia and affective
flattening (e.g., unchanging facial expression, poor eye contact,
affective nonresponsivity, inappropriate affect, and lack of vocal
inflections) were associated with the most inaccurate descriptions
of the Animated triangles film clips. This could imply that expressive linguistic difficulties impair the ability to analyze input from
the surroundings correctly. However, we did not find any correlation between affective flattening and social perception which is
surprising as it is reasonable to presume that the items measured in
the affective flattening subscale are involved in expressing aspects
of social perception, for example, sarcasm.
Our results are in accordance with the findings of Bell et al.
(Bell et al., 2013) who made a principal component analysis based
on subscales of negative symptoms and five social– cognitive tests.
They found that patients could be divided into three subgroups:
One group dominated by high levels of overall negative symptoms
and two groups with low levels of negative symptoms combined
with high versus low social– cognitive functioning. Bell et al.
suggested that patients with high social– cognitive functioning
combined with few negative symptoms had the best community
functioning compared to the other subgroups.
The findings in our study are also consistent with Friths ToM
subgroups (Frith, 1992) where the most affected subgroup of
patients with schizophrenia was thought to be patients dominated
by negative symptoms. However, our results could not confirm the
hierarchical structure of the Frith model since the most remitted
patients (low SANS and low SAPS scores) in fact had more
social– cognitive deficits than the patients dominated by positive
symptoms (high SAPS and low SANS scores).
Our findings regarding the relationship between high levels of
negative symptoms and severe deficits in IQ imply that these
217
patients could be comparable with deficit schizophrenia, which is
a subgroup of patients with poor prognosis (Carpenter, Heinrichs,
& Wagman, 1988; Galderisi, Bucci, et al., 2013). Aspects of
negative symptoms such as affective flattening, alogia, and anhedonia might be connected to deficits in ToM and social perception
or perhaps even be a result of such social– cognitive deficits. In
line with this assumption a recent multicenter study found that
blunted affect was the most persistent negative symptom in firstepisode schizophrenia at 12 month follow-up (Galderisi, Mucci, et
al., 2013). Lin et al. (Lin et al., 2013) suggest that negative
symptoms mediate the influence of neurocognition and social
cognition on functional outcome. Moreover, they suggest that
negative symptoms impair neuro- and social cognition by lowering
the motivation to attend to psychological tasks. They also hypothesize that negative symptoms decrease the motivation to participate in social activities and thereby influence functional outcome.
A similar hypothesis regarding a mediating effect of negative
symptoms with respect to the correlation between second-order
ToM and social functioning was suggested by Mehta et al. (Mehta,
Thirthalli, Kumar, Kumar, & Gangadhar, 2014). Because we did
not measure social functioning in our study, we were not able to
test if the same mediating effect was found in our sample. However, research has shown that ultra-high risk (UHR) subjects who
later develop psychosis have more pronounced social– cognitive
deficits compared with the other UHR subjects (Cornblatt et al.,
2012; Kim et al., 2011). These results are in accordance with our
findings that FES subjects with the most pronounced social–
cognitive and neurocognitive had the most severe levels of negative and positive symptoms. Similar results were found in a recent
study by Ventura et al. (Ventura, Wood, & Hellemann, 2013).
These results could imply that prodromal social– cognitive and
neurocognitive deficits enhances the risk of later development of
psychosis and/or schizophrenia. It is of note that Galderisi et
al.(Galderisi, Mucci, et al., 2013) did not find any difference in
neurocognition between FES subgroups dominated by persistent
negative symptoms (deficit schizophrenia) and remitted negative
symptoms.
The effects of positive symptoms depended upon the context in
which they occurred. Regarding detection of sarcasm in TASIT
(sensitivity), the FES patients with high levels of positive symptoms had the best social perception skills (regarding perception of
sarcasm and discriminating sarcasm from sincerity) in the presence
of low levels of negative symptoms, but the worst social perception in the presence of high negative symptoms. In other words,
patients with many negative symptoms had poor social perception,
but the presence of positive symptoms made their performance
even worse. These results suggest that the understanding of social–
cognitive deficits in schizophrenia is closely linked to a detailed
account of symptomatology. A significant positive correlation was
found between the global SAPS Delusion subscale score and
functional IQ. Our results contradict the notion that patients become increasingly paranoid because of deficits in ToM and social
perception. Recent studies have suggested that instead of ToM
deficits (hypomentalizing) paranoid patients may in fact be hypermentalizing attributing too much idiosyncratic meaning to social
interactions (Abu-Akel & Bailey, 2000; Ciaramidaro et al., 2015;
Crespi & Badcock, 2008; Fretland et al., 2015; Frith, 2004; Montag et al., 2011). Patients dominated by negative symptoms have
been hypothesized to have ToM deficits (hypomentalizing) resem-
BLIKSTED, VIDEBECH, FAGERLUND, AND FRITH
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
218
bling patients with autism spectrum disorder (Abu-Akel & Bailey,
2000; Bliksted et al., 2016; Crespi & Badcock, 2008; Frith, 2004;
Kästner et al., 2015).
The results of the present study imply that patients with FES is
not a clinically homogeneous group in terms of neurocognitive and
social– cognitive deficits, and that subtypes are differentially related to clinical symptoms. If the present results are replicated in
a larger sample of FES subjects (with more statistical power), this
would imply that there is no such thing as an average FES patient
with regard to social– cognitive and neurocognitive deficits.
The results of the present study indicate that a careful evaluation
of social– cognitive and neurocognitive skills is needed in subjects
with UHR and first-episode psychosis because such deficits tend to
make the subject more vulnerable to a future development of
psychotic and negative symptoms.
Strengths and Limitations
All patients were tested by the same psychologist and patients
were recruited from the same mental health clinic on both occasions. The two patient populations did not differ regarding age
(t(57) ⫽ ⫺0.73), gender (2(1) ⫽ 2.64, p ⫽ .10) or years of
education, t(57) ⫽ 0.06, p ⫽ .95.
The patients from time period two had only received medication
for a few weeks or were antipsychotic-naïve. The patients from the
first time period had been medicated for a longer period of time.
However, given the limited effects of antipsychotics on neurocognition, and the relative stability of social– cognitive deficits across
illness phases in schizophrenia (Green et al., 2012; Horan et al.,
2012), we do not expect this difference between previous antipsychotic medication to have affected our results. The fact that sample
one had received antipsychotic medication for a longer period of
time than sample two probably explains why sample one had
significantly fewer positive symptoms, t(57) ⫽ 0.03, p ⫽ .0009
than study two. No significant difference in negative symptoms,
t(57) ⫽ 0.03, p ⫽ .98 was found between the two samples.
Clinical Implications of the Results
In this study, we found that FES patients by no means are alike
regarding cognitive and social– cognitive deficits. Our results imply that cognitive and social– cognitive training must be tailored to
different subgroups of patients with a particular focus on FES
patients with high levels of negative symptoms while also taking
into account the presence of positive symptoms.
References
Abell, F., Happé, F., & Frith, U. (2000). Do triangles play tricks? Attribution of mental states to animated shapes in normal and abnormal
development. Cognitive Development, 15, 1–16. http://dx.doi.org/10
.1016/S0885-2014(00)00014-9
Abu-Akel, A., & Bailey, A. L. (2000). The possibility of different forms of
theory of mind impairment in psychiatric and developmental disorders.
Psychological Medicine, 30, 735–738. http://dx.doi.org/10.1017/
S0033291799002123
Andreasen, N. C. (1984a). Scale for the assessment of negative symptoms.
Iowa City, IA: University of Iowa.
Andreasen, N. C. (1984b). Scale for the assessment of positive symptoms.
Iowa City, IA: University of Iowa.
Bell, M. D., Corbera, S., Johannesen, J. K., Fiszdon, J. M., & Wexler, B. E.
(2013). Social cognitive impairments and negative symptoms in schizophrenia: Are there subtypes with distinct functional correlates? Schizophrenia Bulletin, 39, 186 –196. http://dx.doi.org/10.1093/schbul/sbr125
Bertelsen, M., Jeppesen, P., Petersen, L., Thorup, A., Øhlenschlaeger, J.,
leq Uach, P., . . . Nordentoft, M. (2008). Five-year follow-up of a
randomized multicenter trial of intensive early intervention vs standard
treatment for patients with a first episode of psychotic illness: The OPUS
trial. Archives of General Psychiatry, 65, 762–771. http://dx.doi.org/10
.1001/archpsyc.65.7.762
Bliksted, V., Fagerlund, B., Weed, E., Frith, C., & Videbech, P. (2014).
Social cognition and neurocognitive deficits in first-episode schizophrenia. Schizophrenia Research, 153, 9 –17. http://dx.doi.org/10.1016/j
.schres.2014.01.010
Bliksted, V., Ubukata, S., & Koelkebeck, K. (2016). Discriminating autism
spectrum disorders from schizophrenia by investigation of mental state
attribution on an on-line mentalizing task: A review and meta-analysis.
Schizophrenia Research, 171, 16 –26. http://dx.doi.org/10.1016/j.schres
.2016.01.037
Bora, E., & Pantelis, C. (2013). Theory of mind impairments in firstepisode psychosis, individuals at ultra-high risk for psychosis and in
first-degree relatives of schizophrenia: Systematic review and metaanalysis. Schizophrenia Research, 144, 31–36. http://dx.doi.org/10
.1016/j.schres.2012.12.013
Bora, E., Yucel, M., & Pantelis, C. (2009). Theory of mind impairment in
schizophrenia: Meta-analysis. Schizophrenia Research, 109, 1–9. http://
dx.doi.org/10.1016/j.schres.2008.12.020
Carpenter, W. T., Jr., Heinrichs, D. W., & Wagman, A. M. (1988). Deficit
and nondeficit forms of schizophrenia: The concept. The American
Journal of Psychiatry, 145, 578 –583. http://dx.doi.org/10.1176/ajp.145
.5.578
Castelli, F., Happé, F., Frith, U., & Frith, C. (2000). Movement and mind:
A functional imaging study of perception and interpretation of complex
intentional movement patterns. NeuroImage, 12, 314 –325. http://dx.doi
.org/10.1006/nimg.2000.0612
Ciaramidaro, A., Bölte, S., Schlitt, S., Hainz, D., Poustka, F., Weber, B.,
. . . Walter, H. (2015). Schizophrenia and autism as contrasting minds:
Neural evidence for the hypo-hyper-intentionality hypothesis. Schizophrenia Bulletin, 41, 171–179.
Corcoran, R., Cahill, C., & Frith, C. D. (1997). The appreciation of visual
jokes in people with schizophrenia: A study of ‘mentalizing’ ability.
Schizophrenia Research, 24, 319 –327. Retrieved from. http://www.ncbi.
nlm.nih.gov/pubmed/9134592
Cornblatt, B. A., Carrión, R. E., Addington, J., Seidman, L., Walker, E. F., Cannon,
T. D., . . . Tsuang, M. T. (2012). Risk factors for psychosis: Impaired social and
role functioning. Schizophrenia Bulletin, 38, 1247–1257.
Crespi, B., & Badcock, C. (2008). Psychosis and autism as diametrical
disorders of the social brain. Behavioral and Brain Sciences, 31, 241–
261. http://dx.doi.org/10.1017/S0140525X08004214
Fletcher, P. C., & Frith, C. D. (2009). Perceiving is believing: A Bayesian
approach to explaining the positive symptoms of schizophrenia. Nature
Reviews Neuroscience, 10, 48 –58. http://dx.doi.org/10.1038/nrn2536
Fretland, R. A., Andersson, S., Sundet, K., Andreassen, O. A., Melle, I., &
Vaskinn, A. (2015). Theory of mind in schizophrenia: Error types and
associations with symptoms. Schizophrenia Research, 162, 42– 46.
http://dx.doi.org/10.1016/j.schres.2015.01.024
Frith, C. D. (1992). The cognitive neuropsychology of schizophrenia. New
York, NY: Routledge Taylor & Francis Group.
Frith, C. D. (2004). Schizophrenia and theory of mind. Psychological
Medicine, 34, 385–389. http://dx.doi.org/10.1017/S0033291703001326
Frith, C. D., & Corcoran, R. (1996). Exploring ‘theory of mind’ in people
with schizophrenia. Psychological Medicine, 26, 521–530. http://dx.doi
.org/10.1017/S0033291700035601
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SOCIAL COGNITIVE SUBGROUPS IN SCHIZOPHRENIA
Galderisi, S., Bucci, P., Mucci, A., Kirkpatrick, B., Pini, S., Rossi, A., . . . Maj, M.
(2013). Categorical and dimensional approaches to negative symptoms of
schizophrenia: Focus on long-term stability and functional outcome. Schizophrenia Research, 147, 157–162. http://dx.doi.org/10.1016/j.schres.2013.03.020
Galderisi, S., Mucci, A., Bitter, I., Libiger, J., Bucci, P., Fleischhacker,
W. W., . . . the Eufest Study Group. (2013). Persistent negative symptoms in first episode patients with schizophrenia: Results from the
European First Episode Schizophrenia Trial. European Neuropsychopharmacology, 23, 196 –204. http://dx.doi.org/10.1016/j.euroneuro.2012
.04.019
Green, M. F., Bearden, C. E., Cannon, T. D., Fiske, A. P., Hellemann,
G. S., Horan, W. P., . . . Nuechterlein, K. H. (2012). Social cognition in
schizophrenia, Pt. 1: Performance across phase of illness. Schizophrenia
Bulletin, 38, 854 – 864. http://dx.doi.org/10.1093/schbul/sbq171
Green, M. F., Horan, W. P., & Lee, J. (2015). Social cognition in schizophrenia. Nature Reviews Neuroscience, 16, 620 – 631. http://dx.doi.org/
10.1038/nrn4005
Green, M. F., Horan, W. P., & Sugar, C. A. (2013). Has the generalized deficit
become the generalized criticism?. Schizophrenia Bulletin, 39, 257–262. http://
dx.doi.org/10.1093/schbul/sbs146
Green, M. F., & Leitman, D. I. (2008). Social cognition in schizophrenia.
Schizophrenia Bulletin, 34, 670 – 672. http://dx.doi.org/10.1093/schbul/
sbn045
Green, M. F., Penn, D. L., Bentall, R., Carpenter, W. T., Gaebel, W., Gur,
R. C., . . . Heinssen, R. (2008). Social cognition in schizophrenia: An
NIMH workshop on definitions, assessment, and research opportunities.
Schizophrenia Bulletin, 34, 1211–1220. http://dx.doi.org/10.1093/
schbul/sbm145
Horan, W. P., Green, M. F., DeGroot, M., Fiske, A., Hellemann, G., Kee,
K., . . . Nuechterlein, K. H. (2012). Social cognition in schizophrenia, Pt.
2: 12-month stability and prediction of functional outcome in firstepisode patients. Schizophrenia Bulletin, 38, 865– 872. http://dx.doi.org/
10.1093/schbul/sbr001
Kästner, A., Begemann, M., Michel, T. M., Everts, S., Stepniak, B., Bach,
C., . . . Ehrenreich, H. (2015). Autism beyond diagnostic categories:
Characterization of autistic phenotypes in schizophrenia. BMC Psychiatry, 15, 115. http://dx.doi.org/10.1186/s12888-015-0494-x
Keefe, R. S., Goldberg, T. E., Harvey, P. D., Gold, J. M., Poe, M. P., &
Coughenour, L. (2004). The Brief Assessment of Cognition in Schizophrenia: Reliability, sensitivity, and comparison with a standard neurocognitive battery. Schizophrenia Research, 68, 283–297. http://dx.doi
.org/10.1016/j.schres.2003.09.011
Keefe, R. S., Harvey, P. D., Goldberg, T. E., Gold, J. M., Walker, T. M.,
Kennel, C., & Hawkins, K. (2008). Norms and standardization of the
Brief Assessment of Cognition in Schizophrenia (BACS). Schizophrenia
Research, 102, 108 –115. http://dx.doi.org/10.1016/j.schres.2008.03.024
Keefe, R. S., Poe, M., Walker, T. M., & Harvey, P. D. (2006). The
relationship of the Brief Assessment of Cognition in Schizophrenia
(BACS) to functional capacity and real-world functional outcome. Journal of Clinical and Experimental Neuropsychology, 28, 260 –269. http://
dx.doi.org/10.1080/13803390500360539
Kim, H. S., Shin, N. Y., Jang, J. H., Kim, E., Shim, G., Park, H. Y., . . .
Kwon, J. S. (2011). Social cognition and neurocognition as predictors of
conversion to psychosis in individuals at ultra-high risk. Schizophrenia
Research, 130, 170 –175. http://dx.doi.org/10.1016/j.schres.2011.04.023
Lin, C. H., Huang, C. L., Chang, Y. C., Chen, P. W., Lin, C. Y., Tsai, G. E.,
& Lane, H. Y. (2013). Clinical symptoms, mainly negative symptoms,
mediate the influence of neurocognition and social cognition on func-
219
tional outcome of schizophrenia. Schizophrenia Research, 146, 231–
237. http://dx.doi.org/10.1016/j.schres.2013.02.009
Mancuso, F., Horan, W. P., Kern, R. S., & Green, M. F. (2011). Social
cognition in psychosis: Multidimensional structure, clinical correlates,
and relationship with functional outcome. Schizophrenia Research, 125,
143–151. http://dx.doi.org/10.1016/j.schres.2010.11.007
McDonald, S., Bornhofen, C., Shum, D., Long, E., Saunders, C., &
Neulinger, K. (2006). Reliability and validity of The Awareness of
Social Inference Test (TASIT): A clinical test of social perception.
Disability and Rehabilitation, 28, 1529 –1542. http://dx.doi.org/10.1080/
09638280600646185
McDonald, S., Flanagan, S., Rollins, J., & Kinch, J. (2003). TASIT: A new
clinical tool for assessing social perception after traumatic brain injury.
The Journal of Head Trauma Rehabilitation, 18, 219 –238. http://dx.doi
.org/10.1097/00001199-200305000-00001
Mehta, U. M., Thirthalli, J., Kumar, C. N., Kumar, J. K., & Gangadhar,
B. N. (2014). Negative symptoms mediate the influence of theory of
mind on functional status in schizophrenia. Social Psychiatry and Psychiatric Epidemiology, 49, 1151–1156. http://dx.doi.org/10.1007/
s00127-013-0804-x
Montag, C., Dziobek, I., Richter, I. S., Neuhaus, K., Lehmann, A., Sylla,
R., . . . Gallinat, J. (2011). Different aspects of theory of mind in
paranoid schizophrenia: Evidence from a video-based assessment. Psychiatry Research, 186, 203–209. http://dx.doi.org/10.1016/j.psychres
.2010.09.006
Nelson, H. E., & O’Connell, A. (1978). Dementia: The estimation of
premorbid intelligence levels using the New Adult Reading Test. Cortex,
14, 234 –244. http://dx.doi.org/10.1016/S0010-9452(78)80049-5
Penn, D. L., Sanna, L. J., & Roberts, D. L. (2008). Social cognition in
schizophrenia: An overview. Schizophrenia Bulletin, 34, 408 – 411.
http://dx.doi.org/10.1093/schbul/sbn014
Pickup, G. J., & Frith, C. D. (2001). Theory of mind impairments in
schizophrenia: Symptomatology, severity and specificity. Psychological
Medicine, 31, 207–220. http://dx.doi.org/10.1017/S0033291701003385
Russell, T. A., Reynaud, E., Herba, C., Morris, R., & Corcoran, R. (2006).
Do you see what I see? Interpretations of intentional movement in
schizophrenia. Schizophrenia Research, 81, 101–111. http://dx.doi.org/
10.1016/j.schres.2005.10.002
Savla, G. N., Vella, L., Armstrong, C. C., Penn, D. L., & Twamley, E. W.
(2012). Deficits in domains of social cognition in schizophrenia: A
meta-analysis of the empirical evidence. Schizophrenia Bulletin, 39,
979 –992.
Sprong, M., Schothorst, P., Vos, E., Hox, J., & van Engeland, H. (2007).
Theory of mind in schizophrenia: Meta-analysis. The British Journal of
Psychiatry, 191, 5–13. http://dx.doi.org/10.1192/bjp.bp.107.035899
Ventura, J., Wood, R. C., & Hellemann, G. S. (2013). Symptom domains
and neurocognitive functioning can help differentiate social cognitive
processes in schizophrenia: A meta-analysis. Schizophrenia Bulletin, 39,
102–111. http://dx.doi.org/10.1093/schbul/sbr067
Wechsler, D. (1997). WAIS-III/WMS-III technical manual. San Antonio,
TX: The Psychological Corporation.
WHO. (1994). Schedules for clinical assessment in neuropsychiatry (version 2.1). Geneva, Switzerland: Author.
Received May 22, 2015
Revision received June 22, 2016
Accepted July 21, 2016 䡲
Psychiatric Rehabilitation Journal
2017, Vol. 40, No. 1, 4 –11
© 2017 American Psychological Association
1095-158X/17/$12.00 http://dx.doi.org/10.1037/prj0000232
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Efficacy of Cognitive Rehabilitation Using Computer Software With
Individuals Living With Schizophrenia: A Randomized Controlled
Trial in Japan
Kazuhiko Iwata
Yasuhiro Matsuda
Osaka Psychiatric Medical Center, Osaka, Japan
Nara Medical University
Sayaka Sato
Shunichi Furukawa
National Institute of Mental Health, National Center of
Neurology and Psychiatry, Tokyo, Japan
Tokyo Metropolitan Police Hospital, Tokyo, Japan
Yukako Watanabe, Norifumi Hatsuse, and Emi Ikebuchi
Teikyo University School of Medicine
Objective: Cognitive impairment is common in schizophrenia, and is associated with poor psychosocial
functioning. Previous studies had inconsistently shown improvement in cognitive functions with cognitive
remediation therapy. This study examined whether cognitive remediation is effective in improving both
cognitive and social functions in schizophrenia in outpatient settings that provide learning-based psychiatric
rehabilitation. This study is the first randomized controlled trial of cognitive remediation in Japan. Method:
Study participants were individuals with schizophrenia from 6 outpatient psychiatric medical facilities who
were randomly assigned either a cognitive remediation program or treatment as usual. The cognitive
remediation intervention includes Cognitive training using computer software (CogPack; Japanese version)
administered twice a week and a weekly group over 12 weeks and was based on the Thinking Skills for Work
program. Most study participants were attending day treatment services where social skills training, psychoeducation for knowledge about schizophrenia, group activities such as recreation and sport, and other
psychosocial treatment were offered. Cognitive and social functioning were assessed using the Brief Assessment of Cognition in Schizophrenia (BACS) and Life Assessment Scale for Mentally Ill (LASMI) at pre- and
postintervention. Results: Of the 60 people with schizophrenia enrolled, 29 were allocated to the cognitive
remediation group and 31 were allocated to the treatment as usual group. Processing speed, executive function,
and the composite score of the BACS showed significantly greater improvement for the cognitive remediation
group than the treatment as usual group. In addition, there was significant improvement in interpersonal
relationships and work skills on the LASMI for the cognitive remediation group compared with the treatment
as usual group. Changes from pretreatment to posttreatment in verbal fluency and interpersonal relationships
were significantly correlated, as well as changes in attention and work skills. Conclusions and Implications
for Practice: The present findings showed that providing cognitive remediation on addition to psychiatric
rehabilitation contributed to greater improvement in both cognitive and social functioning than psychiatric
rehabilitation alone. Cognitive remediation may enhance the efficacy of psychiatric rehabilitation improving
social functioning.
Keywords: schizophrenia, cognition, social functioning, cognitive remediation, psychosocial treatment
Most people with schizophrenia continue to have cognitive
deficits even after their psychotic symptoms improve. Previous
studies (Green & Nuechterlein, 1999) have demonstrated an asso-
ciation between poor cognitive functioning and poor functional
outcomes. Therefore, improving both social and cognitive functioning is a very important goal in the treatment of schizophrenia.
This article was published Online First February 9, 2017.
Kazuhiko Iwata, Osaka Psychiatric Medical Center, Osaka, Japan; Yasuhiro Matsuda, Department of Psychiatry, Nara Medical University;
Sayaka Sato, Department of Psychiatric Rehabilitation, National Institute
of Mental Health, National Center of Neurology and Psychiatry, Tokyo,
Japan; Shunichi Furukawa, Department of Psychiatry, Tokyo Metropolitan
Police Hospital, Tokyo, Japan; Yukako Watanabe, Norifumi Hatsuse, and
Emi Ikebuchi, Department of Psychiatry, Teikyo University School of
Medicine.
The institutional review boards at each site approved this study. This
study is supported by Health Labour Sciences Research Grant of the
Ministry of Health, Labor and Welfare, Japan. We appreciate the help
received from Shunichi Fukuhara and Takeo Nakayama at Kyoto University School of Public Health with advice about study design and statistical
analysis.
Correspondence concerning this article should be addressed to Emi
Ikebuchi, 2-11-1 Kaga, Itabashi-Ku, Tokyo 173-8605, Japan. E-mail:
PXM06766@nifty.com
4
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
RCT OF COGNITIVE REHABILITATION FOR SCHIZOPHRENIA
Many different methods for cognitive rehabilitation or cognitive
remediation have been implemented by different researchers.
Some researchers have used puzzles (e.g., Tower of Hanoi), while
others have used neuropsychological tests (e.g., the Wisconsin
Card Sorting Test; Bell, Bryson, Greig, Corcoran, & Wexler, 2001;
Young & Freyslinger, 1995). In recent years there has also been
the increasing use of computer software programs for cognitive
remediation (Garrido et al., 2013; Kurtz, Seltzer, Shagan, Thime,
& Wexler, 2007; Lee, 2013; Rass et al., 2012).
Some studies have already shown that cognitive remediation is
effective in improving cognitive functioning in schizophrenia.
However, it remains unclear whether cognitive remediation improves social functioning in schizophrenia. Social functioning
means real world competence to live effectively and adaptably,
which includes several domains: social and interpersonal relationships, work and school activities, independent community living
skills, enduring family activities, and so on. Bowie, McGurk,
Mausbach, Patterson, and Harvey (2012) presented the results that
cognitive remediation produces robust improvements in neurocognition and generalization to functional competence and real-world
behavior was more likely when supplemental skills training and
cognitive remediation was combined. Therefore our study was
planned so that cognitive remediation added to psychiatric rehabilitation. Kiwanuka et al. (2014) showed that neuropsychological
impairment was associated with vocational outcomes, whereas
most of the self-reported measures were related to social outcomes.
Green et al. (2015) discussed the determinants of daily functioning
in schizophrenia: nonsocial and social cognition, which were mediated with defeatist beliefs and experimental negative symptoms
influencing functional outcomes. Further studies are needed to
make clear influences of neurocognitive functioning in a variety of
domains of social functioning.
There have been some interventional studies of cognitive remediation recently in urban areas of East Asia. Au et al. (2015)
reported that supported employment and cognitive remediation
demonstrated improvement in vocational, clinical, psychological,
and neurocognitive outcomes. However, there was no evidence to
show that that cognitive improvement in these domains beyond
gains associated with supported employment alone. A metaanalysis on working memory training studies suggested activation
of the dorsolateral prefrontal cortex in patients with schizophrenia
(Li et al., 2015). Another meta-analysis of prospective controlled
trials conducted in Singapore, Japan, and other nations showed that
patients receiving cognitive remediation had better work outcomes
than those not receiving cognitive remediation (Chan, Hirai, &
Tsoi, 2015). There have been few interventional studies in Japan
that have used comparison groups in examining the efficacy of
cognitive remediation with computer software (Ikezawa et al.,
2012; Sato et al., 2014). Therefore we conducted a randomized
controlled trial to evaluate whether cognitive remediation using
cognitive training software improves both cognitive and social
functions in people with schizophrenia. In Japan, social skills
training, psychoeducation for patients, group meetings, individual
work therapy, and group activities such as recreation, sports, and
cooking were widely disseminated services in day treatment programs. We sought to evaluate whether cognitive remediation
would enhance the effects of rehabilitation on improving social
functioning in this enriched learning environment. We also sought
5
to evaluate the associations between changes in cognitive functioning and changes in social functioning.
Method
Participants
Inclusion criteria for participation in this study were: (a) diagnosis of schizophrenia, based on the ICD-10 diagnostic criteria for
research (World Health Organization, 1990); (b) age from 20 to
50; (c) outpatients of psychiatric medical facilities participating in
this study; (d) willingness and capability to give informed consent
to participate in this study; (e) clients not receiving old and
conventional antipsychotics as their primary medication; and (f)
clients who have hope to work in the real world setting, and have
not worked yet during participation of the study.
Antipsychotic medications can affect cognitive functioning, especially the older, conventional antipsychotics. For this reason, we
limited study participation to clients receiving second generation
antipsychotics as their primary medication.
Exclusion criteria for participation were: (a) other comorbid
mental illness, (b) history of organic brain disorder, (c) comorbid
substance use disorders, (d) complicating congenital mental retardation, and (e) existence of severe psychiatric symptoms that
would preclude regular attendance at sessions.
Participants were recruited through referrals from attending
psychiatrists or clinical staff at the psychiatric medical facilities
where the study was conducted. Almost all of the participants were
attending day treatment services in both groups, where social skills
training, psychoeducation for patients (teaching information about
schizophrenia in an interactive way), community meetings, individual work therapy, and group activities such as recreation,
sports, and cooking were provided for 6 hr/day, 5 days per week.
Clients can choose to attend these programs after discussion with
their care-manager, and the total number of hours they attended
depended on individual preferences and conditions. These
learning-based rehabilitation programs are popular in Japan. Efforts were made to avoid changing clients’ pharmacological treatment during the study unless it was clinically necessary.
Design and Setting
This study was designed as a multicenter randomized controlled
trial, conducted at six psychiatric facilities in Japan: two sites in
Osaka (one hospital and one clinic), and one hospital each in
Fukushima, Tokyo, Saitama, and Niigata prefecture.
After completion of the baseline evaluations, clients were randomly assigned to either the cognitive remediation group or control treatment as usual group after stratification by age and sex. To
minimize potential disappointment for people assigned to the
control group, they served as a “waiting list” control group, and
were provided the cognitive remediation after 12 weeks.
Interventions
Cognitive remediation group. Cognitive remediation was
provided using the computer software: CogPack (Marker software,
Germany). This program was developed for rehabilitation of
higher brain dysfunction, and it contains 64 cognitive tasks to train
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
6
IWATA ET AL.
verbal memory, working memory, attention/vigilance, psychomotor speed, and executive function. Marker software gave us permission to develop a Japanese language version of the program for
use in research settings.
Our cognitive remediation method is theoretically based on the
Thinking Skills for Work program (McGurk, Mueser, et al., 2007;
McGurk, Mueser, & Pascaris, 2005). We conducted 24 cognitive
training sessions using CogPack, twice a week, with each session
lasting about 45– 60 min. We also conducted a weekly group
session designed to promote the transfer of improved cognitive
functioning to real-world situations. The trained therapists described below also provided participants teaching compensation
strategy or prompting additional practice if needed.
We developed a cognitive remediation manual for practitioners
in order to standardize treatment across sites. Additionally, at least
one therapist at each site had to take two 1-day training courses to
learn the cognitive remediation program before the study was
started. Therapists involved in the intervention were psychologists,
nurses, social workers, and occupational therapists who were familiar with psychiatric rehabilitation for schizophrenia, and supervised during the study period by members of the research team
who had several years of experience with cognitive remediation.
Internet conferences between members of the research team were
also held during the study period. Using computer software and the
manual also minimizes the disparity of efficacy in cognitive remediation.
In the computer cognitive training, participants were directed to
practice a wide range of cognitive domains in both the early and
later phases of remediation with adherence to the Thinking Skills
for Work program, and each participant could choose either preferable tasks or unskilled tasks to enhance their interests or selfefficacy in the later phase. In the groups, participants talked about
their weak tasks, and discussed with each other strategies to
complete tasks using some cognitive functioning in the early
phase. In the middle and later phases, they also discussed social
goals and how to transfer gained cognitive skills to achieve their
goals.
Treatment as usual group. All participants received standard
(treatment as usual) outpatient treatment.
Outcomes
Both the cognitive remediation group and the treatment as usual
group were evaluated using the following assessments in the preand postintervention phases (after 4 months from baseline) within
1 month.
Primary outcome. We assessed cognitive functioning as the
primary outcome in this study. Cognitive functioning was assessed
using the Brief Assessment of Cognition in Schizophrenia—Japanese version (BACS-J) preintervention. The BACS-J includes six
measures of cognitive functioning in the following domains: verbal memory, attention, verbal fluency, working memory, executive
functioning, and psychomotor processing. Composite score of
overall cognitive performance is also provided.
The BACS-J has established reliability and validity and has
good sensitivity for the types of cognitive deficits associated with
schizophrenia. Normalized standard scores of the BACS-J for each
age category have already been reported in Japan (Kaneda et al.,
2007; Keefe, Poe, Walker, Kang, & Harvey, 2006).
Secondary outcome. Social functioning and psychopathology
were assessed as secondary outcomes. Social functioning was
assessed with the LASMI (Ikebuchi, Iwasaki, Miyauchi, Oshima,
& Sugimoto, 1995; Iwasaki et al., 1994), which was rated based on
reports from clients, information of caregivers, and therapist observations during this study in day treatment activities or other
outpatient programs over the previous month. The LASMI yields
rating functioning in five domains: daily living, interpersonal
relations, work, endurance and stability, and self-recognition. We
used the Interpersonal Relations and Work subscales of the
LASMI in this study (Kay, Opler, & Fiszbein, 1991), because the
aim of the study was improving vocational abilities through cognitive remediation. The Interpersonal Relations subscale consists
of 13 items which are derived from basic communication skills
such as facial expression to informal relations with family or
friends. The Work subscale consists of 10 items, which include
abilities of work skills and the employee’s role which might
expected in a company. These abilities were rated in a simulation
setting of outpatients’ programs. Each item of the LASMI is
evaluated on a 5-point Likert scale, and more score points mean
more need for support for each activity. The score of each subscale
is calculated by summing all of the items in the subscale.
Symptom severity was rated using the Positive and Negative
Syndrome Scale (PANSS), which was scored according to the
three-factor model (Positive scale, Negative scale, and General
Psychopathology scale). The PANSS is composed of 30 items, and
each item is rated on a 7-point Likert scale from 1 (Absent) to 7
(Extreme).
We conducted two training sessions for the raters at each site
who assessed the primary and secondary outcomes.
Statistical Analysis
Analyses were performed on an intention-to-treat basis. Participants who had at least one assessment made up the intention-totreat sample. First we compared baseline demographic and clinical
characteristics between the cognitive remediation group and the
treatment as usual group with t tests or chi-square tests. After
comparing the two groups on demographic and baseline clinical
characteristics, we performed following analyses as described below. For the statistical analysis, we used JMP ver.10 (SAS Institute
Japan, Tokyo Japan).
Comparing primary and secondary outcomes between the
two groups. At first, improvements in cognitive and social functioning and psychotic symptoms were compared between two
groups. The groups were compared using analysis of covariance
(ANCOVA) with group as the independent variable, changes in
cognitive or social functioning of posttreatment as the dependent
variables, and baseline functioning and age as covariates. The
linear regression model of the analysis is used. We entered five
baseline characteristics as covariates (age, duration of illness,
premorbid IQ, dose of antipsychotics, and baseline score of clinical scales). We found no significant group difference of demographic and clinical variables by random allocation. However,
“allocation bias” remains because four variables (age, duration of
illness, premorbid IQ, dose of antipsychotics) and baseline score of
clinical scales have some effect on primary and secondary outcomes. Thus, adjusting these factors could minimize the allocation
bias and boosted the statistical power for the efficacy of the
RCT OF COGNITIVE REHABILITATION FOR SCHIZOPHRENIA
cognitive remediation. Finally, we set up the parameter () of
“allocation group” and its 95% confidence interval (95% CI) to
determine the efficacy of cognitive rehabilitation.
Correlational analysis. The objective of the correlational
analyses was to explore whether changes in cognitive functioning
from pretreatment to posttreatment were correlated with changes
in social functioning, and, if so, which areas of cognitive functioning were most strongly associated with improvement in social
functioning. Therefore we calculated Spearmen correlations between the changes of each domain in the BACS-J and LASMI.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Study Ethics
We administered this trial in accordance with Declaration of
Helsinki and Ethical Guidelines for Clinical Studies of the Ministry of Health, Labor, and Welfare. This study was approved by the
institutional review board or ethics committee at each site, and oral
and written informed consent was obtained from all participants.
We registered this study in UMIN Clinical trial registration
(UMIN CTR ID: UMIN000002775).
7
functioning, and psychiatric symptoms did not differ between the
groups. Fifty-five clients (91.7% of participants) were using the
day treatment program or psychosocial treatment programs for
outpatients, which were open to attendance every day. The percentage of attendance did not differ between the groups.
The Efficacy of Cognitive Rehabilitation in Improving
Cognitive Functioning
Analyses were performed on an intention-to-treat basis. Participants who had at least one assessment made up the intention-totreat sample. We used the last observation carried forward method,
and we assigned the pretest data into missing posttest data of four
participants who declined to posttest. Table 2 shows the changes in
cognitive functioning on the BACS-J before and after the intervention with cognitive remediation. ANCOVA indicated that there
was a significantly greater improvement for the cognitive remediation group on three domains of the BACS-J: composite score
(F ⫽ 8.209, df ⫽ 59, p ⫽ .006), processing speed (F ⫽ 6.345, df ⫽
59, p ⫽ .015), and executive functioning (F ⫽ 4.203, df ⫽ 59, p ⫽
.045).
Results
Patient Flow and Baseline Characteristics
The Efficacy of Cognitive Remediation in Improving
Social Functioning and Psychotic Symptoms
Among clients of the six hospitals, 61 clients met the inclusion/
exclusion criteria and provided informed consent. However, one
patient declined to complete the pretest, therefore 60 clients were
randomized. Of these, 29 clients were assigned to receive cognitive remediation, and 31 were assigned to the treatment as usual
group. Figure 1 shows a flow diagram of the clients through the
study.
Table 1 shows the patient characteristics and baseline scores in
both groups before intervention. There was no significant difference in baseline characteristics between the groups. In addition,
the baseline evaluation scores for cognitive functioning, social
Table 3 shows the changes in social functioning and psychotic
symptoms evaluated with the LASMI and PANSS at baseline and
posttreatment for both groups. The evaluation of changes in social
functioning indicated significantly greater improvements in interpersonal relationships on the LASMI (F ⫽ 12.817, df ⫽ 59, p ⬍
.001) and work skills (F ⫽ 8.037, df ⫽ 59, p ⫽ .007) in the
cognitive remediation group compared with the treatment as usual
group. In addition, psychiatric symptoms including the total
PANSS score and all of these subscales also showed significantly
greater improvements in the cognitive remediation group compared with the treatment as usual group (see Table 3).
Assessed for eligibility and obtained participants' consent
(n=61)
Excluded (n=1)
・Declined to pre-test(n=1)
Randomized(n=60)
Cognitive Remediation Group(n=29)
Treatment as Usual Group(n=31)
Intervention(n=29)
・Received cognitive remediation(n=29)
・Did not receive cognitive remediation(n=0)
Intervention (n=31)
Follow Up(n=29)
Follow Up(n=31)
・Lost to follow-up(n=0)
・Discontinued intervention (relapse, readmission etc.) (n=0)
・Received treatment as usual in outpatient unit(n=31)
・Did not receive treatment as usual in outpatient unit(n=0)
・Lost to follow-up(n=0)
・Discontinued intervention (relapse, readmission etc.) ( n=0 )
Analysed (n=28)
Analysed (n=28)
・Excluded from analysis ( declined to post-test ) ( n=1 )
・Excluded from analysis ( declined to post-test ) ( n=3 )
Figure 1.
CONSORT diagram of patient flow through study.
IWATA ET AL.
8
Table 1
Baseline Characteristics: Demographic Data
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Cognitive
remediation
group (N ⫽ 29)
Baseline data
Avg.
Age (years)
Sex (male/female)
Duration of Illness (month)
IQ (JART Score)
Dose of antipsychoticsa (mg)
34.2
Treatment as
usual group
(N ⫽ 31)
SD
Avg.
7.11
34.5
7/22
140.9
98.7
672.6
SD
6.73
8/23
96.8
9.59
618.3
144.5
97.2
674
83.1
10.2
417.5
Note. No significant differences between groups on any variables with t
test or chi square test. JART ⫽ Japanese Adult Reading Test.
a
Chlorpromazine equivalent.
Correlations Between Improvements on Cognitive and
Social Functioning
All clients were included in the analysis of the correlation
between cognitive and social functioning improvement. Table 4
shows the correlations between LASMI interpersonal relationships
and work skills, respectively, with the BACS-J composite score
and each cognitive functioning domain. We found significant
correlations between changes in BACS-J verbal memory and
changes in LASMI interpersonal relationships (Spearman’s rank
correlation coefficient; ⫽ ⫺0.315, p ⫽ .018), between changes
in BACS-J verbal fluency and changes in LASMI interpersonal
relationships (Spearman’s ⫽ ⫺0.263, p ⫽ .049), between
changes in BACS-J attention/vigilance and changes in LASMI
work skills (Spearman’s ⫽ ⫺0.311, p ⫽ .02), and between
changes in BACS-J executive functioning and changes in LASMI
work skills (Spearman’s ⫽ ⫺0.265, p ⫽ .049).
Discussion
Providing cognitive remediation in addition to the usual treatment involving psychiatric rehabilitation approaches such as social
skills training and other learning-based rehabilitation programs
was associated with greater improvement in cognitive functioning,
social functioning, and psychiatric symptoms in people with
schizophrenia, compared with the usual services alone. Research
has previously established the effectiveness of traditional psychi-
atric rehabilitation methods such as social skills training and
psychoeducation programs (Dixon et al., 2010), which are widely
used today in psychiatric practice in Japan. However, many clients
who receive psychiatric rehabilitation have difficulty achieving
their treatment goals. Cognitive remediation may impose cognitive
functioning which could enhance the benefits of rehabilitation
programs. Our findings are consistent with previous research on
cognitive remediation showing significant improvements in cognitive functioning, as well as improvements in social functioning,
when cognitive remediation was added to psychiatric rehabilitation
compared with psychiatric rehabilitation alone (McGurk et al.,
2013; McGurk et al., 2015).
In our analysis, significant correlation between improvement in
cognitive functioning and improvement in social functioning were
observed for some outcomes. As previously reviewed (Green,
Kern, Braff, & Mintz, 2000; Green & Nuechterlein, 1999), neurocognition strongly influences psychosocial functioning. However, this interpretation does not take into account the fact that
cognitive remediation programs that are provided in the absence of
psychiatric rehabilitation (e.g., social skills training) tend to produce weaker (or nonsignificant) effects on psychosocial functioning, despite showing beneficial effects on cognitive functioning.
Although previous studies of cognitive remediation report efficacy
in improving cognitive functioning, the findings have been inconsistent that improvement of cognitive functioning did not translate
into improved social functioning directly (Lu et al., 2012; Medalia
& Saperstein, 2013). However, some studies that added cognitive
remediation to psychiatric rehabilitation showed the greatest impact on psychosocial functioning (Franck et al., 2013; Krabbendam & Aleman, 2003; Wykes, Huddy, Cellard, McGurk, & Czobor, 2011). It is possible that cognitive remediation improves
capacity to learn through increased verbal memory or executive
functioning, and in the absence of concerted learning opportunities, improved cognitive functioning does not automatically lead to
improved psychosocial functioning, as discussed in McGurk et al.
(2013, 2015). The present results suggest that improved cognitive
functioning could facilitate improvement in social functioning in
the context of a social learning environment through transferring
skills from laboratory to real world. Subramaniam et al. (2014)
showed that task performance of cognitive training of working
memory and brain activity within the bilateral middle frontal gyri
predicted better occupational functioning at 6-month follow-up.
Table 2
The Changes of Cognitive Functions After Intervention
Cognitive remediation
group (N ⫽ 29)
Treatment as usual group
(N ⫽ 31)
BACS
Pre (SD)
Post (SD)
Pre (SD)
Post (SD)
F

p value
Verbal memory
Digit sequencing
Verbal fluency
Token Motor Task
Symbol-Coding Task
Tower of London
Composite score
.18 (.82)
.10 (.99)
.11 (.74)
.19 (1.02)
.25 (1.09)
.08 (1.08)
.15 (.48)
.63 (.82)
.49 (1.01)
.64 (.99)
.53 (.94)
.54 (1.12)
.32 (.74)
.55 (.64)
⫺.17 (1.12)
⫺.09 (1.02)
⫺.10 (1.20)
⫺.18 (.96)
⫺.24 (.86)
⫺.08 (.93)
⫺.14 (.70)
.07 (1.09)
.11 (1.14)
.09 (1.28)
⫺.18 (.88)
⫺.10 (1.00)
⫺.05 (.88)
.01 (.69)
3.150
1.487
3.129
6.345
1.467
4.203
8.209
.314
.227
.353
.466
.190
.239
.300
.082
.228
.087
.015ⴱ
.231
.045ⴱ
.006†
ANCOVA
Note. Analysis by analysis of covariance (ANCOVA) and last observation carried forward (LOCF). df ⫽ 59.
BACS ⫽ Brief Assessment of Cognition in Schizophrenia (Japanese version).
†
p ⬍ .01. ⴱ p ⬍ .05.
RCT OF COGNITIVE REHABILITATION FOR SCHIZOPHRENIA
9
Table 3
The Changes of Social Functions and Psychotic Symptoms After Intervention
Cognitive remediation group
(N ⫽ 29)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ANCOVA
Pre (SD)
Post (SD)
Pre (SD)
Post (SD)
F

p value
14.14 (5.47)
12.93 (5.17)
9.69 (5.41)
9.41 (5.52)
14.51 (7.38)
14.87 (6.10)
14.32 (7.45)
13.87 (6.20)
12.817
8.037
⫺4.317
⫺3.120
⬍.001†
.007†
12.93 (4.95)
14.83 (4.80)
29.48 (8.83)
57.27 (16.53)
11.38 (4.78)
12.66 (4.52)
26.72 (9.10)
50.76 (17.03)
12.77 (4.40)
16.55 (5.03)
31.12 (8.44)
60.45 (14.8)
13.03 (4.43)
15.87 (5.66)
31.39 (9.02)
60.29 (17.42)
7.996
5.215
9.015
9.407
⫺1.788
⫺1.871
⫺3.231
⫺6.691
.007†
.026ⴱ
.004†
.003†
The changes after intervention
LASMI
Interpersonal relationship
Work
PANSS
Positive scale
Negative scale
General Psychopathology Scale
Total
Treatment as usual group
(N ⫽ 31)
Note. Analysis by analysis of covariance (ANCOVA) and last observation carried forward (LOCF). df ⫽ 59. LASMI ⫽ Life Assessment Scale for Mental
Illness; PANSS ⫽ Positive and Negative Syndrome Scale.
†
p ⬍ .01. ⴱ p ⬍ .05.
The program of this study also focused on enhancing executive
functioning and working memory, and transferring learned abilities to the real world with group sessions and daily activities of day
treatment programs. Therefore, not only cognitive remediation but
also traditional psychiatric rehabilitation focused on social functioning may be necessary for some clients with schizophrenia. To
achieve social goals such improvements in both cognitive and
social functioning may be critical to facilitate the reintegration of
persons with schizophrenia in the community.
Psychiatric symptoms assessed on the PANSS also improved
significantly more in the cognitive remediation group than the
treatment as usual group. Meta-analyses of controlled studies of
cognitive remediation have showed small effect sizes on the reduction of psychiatric symptoms (Kurtz, 2012; McGurk, Twamley,
et al., 2007; Medalia & Choi, 2009). It is possible that the reduction in psychiatric symptoms found in this study were influenced
by the improvement in social functioning.
In this study, a standardized cognitive remediation program
provided for about 24 sessions over 12 weeks was found to be
effective at improving cognitive and social functioning. This result
Table 4
Correlation Between Improvements of Cognitive and
Social Functions
Change of subscales in LASMI
Interpersonal
relationship
(N ⫽ 60)
Work (N ⫽ 60)
Change of subscales
in BACS (N ⫽ 60)
Spearman’s
p value
Spearman’s
p
value
Verbal memory
Digit sequencing
Verbal fluency
Token Motor Task
Symbol-Coding Task
Tower of London
⫺.315
⫺.158
⫺.263
⫺.138
⫺.106
⫺.045
.018ⴱ
.246
.049ⴱ
.309
.438
.740
⫺.133
⫺.169
⫺.151
.056
⫺.311
⫺.265
.330
.212
.267
.681
.02ⴱ
.049ⴱ
Note. Analysis by correlational analysis and last observation carried
forward (LOCF). LASMI ⫽ Life Assessment Scale for Mental Illness;
BACS ⫽ Brief Assessment of Cognition in Schizophrenia Japanese version.
ⴱ
p ⬍ .05.
is consistent with the findings from Bowie et al. (2012) that
real-world behavior was more likely when supplemental skills
training and cognitive remediation were combined. It is possible
that some clients in the cognitive remediation group who did not
improve in cognitive or social functioning would have improved
from a longer program.
Strengths and Limitations of This Study
Some strengths of this study should be noted. The computer
software program used to engage clients in cognitive exercises was
designed to improve cognitive functioning. Moreover, the cognitive remediation procedures, which were mainly strategy coaching
and compensating methods by trained therapists, wer...
Purchase answer to see full
attachment