Nordic Journal of Psychiatry

ISSN: 0803-9488 (Print) 1502-4725 (Online) Journal homepage: http://www.tandfonline.com/loi/ipsc20

Do patients with high-functioning autism have

similar social cognitive deficits as patients with a
chronic cause of schizophrenia?

Lotte Veddum, Heine Lund Pedersen, Anna-Sofie Lose Landert & Vibeke
Bliksted

To cite this article: Lotte Veddum, Heine Lund Pedersen, Anna-Sofie Lose Landert &
Vibeke Bliksted (2019): Do patients with high-functioning autism have similar social cognitive
deficits as patients with a chronic cause of schizophrenia?, Nordic Journal of Psychiatry, DOI:
10.1080/08039488.2018.1554697

To link to this article: https://doi.org/10.1080/08039488.2018.1554697

Published online: 14 Jan 2019.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ipsc20
NORDIC JOURNAL OF PSYCHIATRY
https://doi.org/10.1080/08039488.2018.1554697

ORIGINAL ARTICLE

Do patients with high-functioning autism have similar social cognitive deficits

as patients with a chronic cause of schizophrenia?
Lotte Vedduma, Heine Lund Pedersenb, Anna-Sofie Lose Landertc and Vibeke Bliksteda,d,e
a
Psychosis Research Unit Aarhus University Hospital Risskov, Risskov, Denmark; bDepartment of Psychosis, Neuropsychiatric Clinic, Aarhus
University Hospital Risskov, Risskov, Denmark; cAmbulatorium for psykotiske lidelser, Regionspsykiatrien Vest, Herning, Denmark;
d
Department of Clinical Medicine, Aarhus University, Aarhus N, Denmark; eInteracting Minds Centre Aarhus University, Aarhus C, Denmark

ABSTRACT ARTICLE HISTORY

Objective: There is substantial evidence that both patients with schizophrenia and patients with aut- Received 26 June 2018
ism spectrum disorders (ASD) have impaired social cognition including theory of mind (ToM) deficits. Revised 2 November 2018
However, it remains unclear if both verbal (explicit) and non-verbal (implicit) ToM as well as social per- Accepted 27 November 2018
ception are similarly affected in both disorders.
KEYWORDS
Methods: Twenty-one patients diagnosed with schizophrenia and 11 patients diagnosed with ASD Social cognition; theory of
were matched one-to-one to healthy controls based on gender, age, and educational level. Social mind; social perception;
functioning was measured by Personal and Social Performance (PSP) scale. Neurocognition was meas- implicit; explicit
ured using Brief Assessment of Cognition in Schizophrenia (BACS-DK), and four subtests from Wechsler
Adult Intelligence (WAIS-IV) scale were applied to estimate IQ. The Animated Triangles Task was used
to measure implicit ToM, while explicit ToM and social perception were measured by The Awareness
and Social Inference Test (TASIT).
Results: Patients with schizophrenia had deficits in implicit ToM and complex social perception com-
pared to their matched controls, but no problems with explicit ToM. Surprisingly, patients with ASD
solely had deficits with regard to complex social perception compared to their matched controls. The
two patient groups were similar regarding estimated IQ, social functioning and years of education, but
differed in age and neurocognition. When adjusting the p-values for age and neurocognitive deficits,
both patients groups had similar social cognitive deficits.
Conclusions: Results imply that we compared schizophrenia patients with substantial neurocognitive
deficits to a group of high-functioning patients with ASD. However, these two subgroups may have
the same level of social cognitive deficits.

1. Introduction disorders has emerged in recent years [12–14]. Similarities in

the neural correlates of ToM in the two disorders have been
Alterations and impairments in social cognition is a common
identified [15]. However, this contemplation of social cogni-
feature of many psychiatric disorders and conditions [1–4].
tive concordance is not novel as schizophrenia and ASD
For instance, schizophrenia and autism spectrum disorders
have a long history of diagnostic conflation stemming from
(ASD) are neurodevelopmental disorders characterized by apparent overlaps in certain areas of social dysfunction [16].
marked impairments in social cognition, including theory of In spite of this, much research has focused on the ToM defi-
mind (ToM) deficits, i.e. impairment in the ability to infer and cits appearing in the two disorders separately, whereas lim-
attribute independent mental states to oneself and others in ited empirical work has targeted at directly comparing these
order to explain and predict behavior [5]. Whereas ToM defi- two clinical groups. Research regarding direct comparisons is
cits have long been recognized as a primary abnormality in sparse, and findings from existing studies comparing ToM in
ASD [6], interest in ToM deficits in schizophrenia has evolved ASD and schizophrenia show mixed results. Some report
due to an increasing focus social cognitive impairments in comparable ToM deficits in both disorders [17–19], whereas
this mental disorder [7]. results from other studies indicate that patients with ASD
Results from various meta-analyses have revealed that have significantly more severe ToM deficits compared to
there is now substantial evidence showing that patients with patients with schizophrenia [20,21], or that the opposite is
schizophrenia likewise are characterized by marked ToM defi- occurring [22,23].
cits [8–10]. This indicates some overlap between social cogni- A recent meta-analysis examining ToM abilities in ASD
tion in schizophrenia and ASD [11]. An increasing interest in and schizophrenia found statistically large effect sizes at a
examining and comparing possible brain mechanisms or similar level in both disorders [24]. Results from this meta-
abberations underlying social cognitive deficits in the two analysis further revealed that patients with schizophrenia

CONTACT Vibeke Bliksted vibeke.bliksted@ps.rm.dk Psychosis Research Unit Aarhus University Hospital Risskov, Skovagervej 2, Indgang 51, 4, Risskov
DK-8240, Denmark
ß 2019 The Nordic Psychiatric Association
2 L. VEDDUM ET AL.

tend to have larger impairments on verbal mentalizing tasks 2. Materials and methods
compared to nonverbal mentalizing tasks, whereas patients
2.1. Subjects
with ASD had similar deficits in both types of tasks. This indi-
cates that the two patient groups may perform differently on 2.1.1. Recruited patients
distinct ToM tasks. This proposition was supported by results The thirty-two included patients were recruited from the
from a recent meta-analysis examining schizophrenia and Neuropsychiatric unit at the Psychosis department, Aarhus
ASD patients’ performance on the Animated Triangles Task University Hospital Risskov. This unit is specialized in diagnos-
[25]. Results showed that patients with ASD seem to make ing patients with complex psychotic disorders (e.g. suspicion
more inaccurate descriptions of interactions involving ToM of an organic mental disorder). All patients referred to the
compared to patients with schizophrenia, although no differ- clinic were recruited and tested from March 2013 until
ence was found between the two patient groups regarding October 2015 by an experienced neuropsychologist (HLP),
detection of intentionality. and diagnoses were validated by an experienced psychiatrist.
In research, a distinction between implicit and explicit In the present study, we only included those patients who
ToM is often made as these abilities are suggested to work met the ICD-10 (International Classification of Disease 10th
in different ways [26]. The implicit mentalizing system edition) criteria for either schizophrenia or ASD and had no
involves automatic and cognitive efficient ToM abilities, diagnosis of drug- or alcohol dependency according to ICD-10
which is thus limited and inflexible. The explicit mentalizing [37]. Patients were excluded if they did not understand spo-
system refers to a culturally inherited skill involving mental ken Danish sufficiently to comprehend the testing procedures.
flexibility to reason about beliefs, which is a process that
highly rely on language and executive functions [26,27]. It
2.1.2. Healthy control subjects
has been suggested that patients with schizophrenia have
The healthy control subjects were matched one-to-one to
intact implicit ToM, but impaired explicit ToM [28]. Contrary
the patients based on gender and age. Notably, it was not
to this, patients with ASD are known to have impaired impli-
possible to match perfectly one-to-one regarding age but
cit ToM, whereas some patients with higher verbal ability
the patients and their matched controls did not differ more
can pass explicit ToM tasks [29,30], presumably due to com- than 2 years, and on group level no difference occurred. We
pensatory strategies [31]. also intended to match on total years of education and cur-
However, in both schizophrenia and ASD, other social rent occupation. However, many of the included patients
cognitive areas are also identified to involve pivotal deficits. were on pension or sick leave and had fewer years of educa-
In schizophrenia, five distinct social cognitive domains have tion as could be expected according to their age. This
been specified to be particular affected: (1) ToM, (2) social strongly suggests that a group of controls matched on years
perception, (3) social knowledge, (4) attributional bias, and of education would not constitute a representative sample.
(5) emotional processing [7]. Results from a recent meta-ana- Instead, we aimed at recruiting healthy controls with differ-
lysis revealed that ToM and social perception were the ent employments and educational levels. The 32 included
domains most severely affected in schizophrenia [9]. For this controls had no history of mental illness and neither had
reason, we added a complex social perceptual task in order their first-degree relatives. In addition, they did not report a
to broaden the comparison of social cognitive deficits in history of neurological illness or any drug- or alcohol
schizophrenia and ASD. To our knowledge, no other study dependency according to ICD-10 criteria. The healthy con-
has compared these two patients groups regarding social trols were recruited through an advertisement in a local
perception, which involves the ability to attend to and inter- newspaper and through flyers posted in the local commu-
pret the whole combination of complex information from dif- nity. All healthy controls received a voucher for participating.
ferent social phenomena including body language, verbal
messages, paralinguistic cues as well as non-verbal behaviors
[32]. As such, social perception is a very complex domain 2.2. Social cognitive measures
concerning the ability to ‘read between the lines’. 2.2.1. Theory of mind
The main aim of the present study was to compare The Animated Triangles Task (AT) was applied to assess
implicit and explicit ToM as well as social perception in implicit ToM. This task was chosen due to its validity as a
schizophrenia and ASD. Furthermore, we also aimed to culturally independent task assessing implicit and non-verbal
compare the two patient groups regarding IQ and neuro- aspects of ToM. AT consists of short movie clips with two
cognition because both mental disorders have been found animated triangles, which either move in a random or an
to be characterized by impairments in neurocognition intentional way [38,39]. There are four clips of each type of
[28,33,34], and because ASD often is associated with low IQ animation lasting 38–41 s, and the different types of anima-
level or even mental retardation [35,36]. We hypothesized tions were shown in a random order. After each animation
that patients with ASD would have more comprehensive the participants were asked to describe what they thought
deficits in both social cognition as well as neurocognition was happening. The answers were rated regarding intention-
compared to patients with schizophrenia, but that both ality, that is, the degree of appreciation of mental states
patient groups would be significantly impaired relative to (ranging 0–5, where a score of 4 or 5 refers to mental state
healthy controls. attribution) and appropriateness, that is how well the
 NORDIC JOURNAL OF PSYCHIATRY 3

underlying script was captured (ranging 0–3, where 3 refers subjects were interviewed with the Personal and Social
to a perfect description). Performance Scale (PSP), which is a tool to measure social
Reports were recorded and transcribed. Two raters (L. V. functioning in schizophrenia [46]. The healthy controls were
and H. L. P.), blind to diagnosis, independently scored each interviewed with the Present State Examination interview
transcribed description after having been trained to use the (PSE, ICD-10).
published set of scoring criteria by V. B. [39,40], and their
scores were averaged for data analysis. Inter-rater agreements
2.5. Ethics
were fair to substantial (Random animations: Intentionality
j ¼ 0.74, Z ¼ 8.14, p < .0001, Appropriateness j ¼ 0.58, Z ¼ 9.47, All participants received written and verbal information about
p < .0001; ToM animations: Intentionality j ¼ 0.54, Z ¼ 8.55, the project and a written informed consent was obtained and
p < .0001, Appropriateness j ¼ 0.35, Z ¼ 7.58, p < .0001). signed before participation. The study was approved by the
Danish Data Protection Agency (Ref: 2007-58-0010), and the
project complied with the Helsinki-II declaration.
2.2.2. Social perception
The Danish version of The Awareness of Social Inference
Test, Part 2A (TASIT), was used to assess explicit ToM and 2.6. Data-analysis
social perception [41]. This test was chosen due to its sub-
Statistical analyses were performed with Stata IC 14.2 soft-
stantial ecological validity as a real life measure of social per-
ware (StataCorp, College Station, TX). All data were examined
ception and ToM abilities requiring a combination of visual
for distribution and outliers. Continuous variables were
and verbal information. TASIT consists of small video clips
examined by Wilcoxon rank-sum test (Mann–Whitney) and
lasting 16–53 s, where professional actors perform small
effect sizes were reported in terms of Harrell’s C (HC).
everyday interactions [32,42]. TASIT contains three different
Categorical variables were examined by Fisher’s exact test (if
types of video clips: (1) sincere clips, where congruence
N < 5 in a cell) or Chi2 and reported with the counts and
exists between what the persons are literally saying and the
proportions of the group total. Composite scores for BACS
paralinguistic and facial cues, (2) simple sarcastic clips, where
were calculated as the weighted mean of z-scores, separately
one of the persons is being sarcastic and no congruence
computed for each subtest relative to the mean and stand-
exists between the spoken word and the paralinguistic and ard deviation of the healthy control sample (mean ¼ 0, SD
facial cues, (3) paradoxical sarcastic clips, which only make ¼ 1). Regression analyses were calculated for the social cog-
sense if the participant is able to detect the sarcasm being nitive scores and p-values were adjusted with age and com-
used in the dialog. After each video clip, the participants posite BACS scores as covariates.
were asked to answer four questions concerning different
aspects of the communicative intentions of the persons
(what they were doing, saying, thinking, and feeling). Scoring 3. Results
of each type of video clip range from 0 to 20, which corre-
3.1. Demographics and social functioning
sponds to a maximum score of 60.
Demographics and social functioning of the patients and their
matched controls are summarized in Tables 1 and 2. Both
2.3. Neurocognitive measures patient groups and their matched controls did not differ in
2.3.1. Neurocognition gender and age. However, the two patient groups differed
The Danish version of the Brief Assessment of Cognition in markedly in age with the schizophrenia sample being older
Schizophrenia (BACS) was used to measure verbal memory, than the ASD sample (Z ¼ 3.42, p ¼ .0006). As expected, edu-
working memory, motor speed, verbal fluency, executive cational level and current occupation of both patient groups
functions, attention, and speed of processing [43,44]. were significantly lower compared to their matched controls.
Likewise, the social functioning of both patient groups was
significantly lower compared to their matched controls. The
2.3.2. Intelligence patient groups did not differ regarding years of education
Intelligence was estimated using four subtests from WAIS-IV (Z ¼ 1.07, p ¼ .29) or social functioning (Z¼ 1.13, p ¼ .26).
(Wechsler Adult Intelligence Scale, Fourth version):
Vocabulary, Similarities, Block Design, and Matrix Reasoning.
These subtests were chosen due to their high correlation 3.2. Pairwise comparisons of patients and
with the total WAIS-IV IQ-score [45]. matched controls
Pairwise comparisons revealed significant differences in esti-
mated IQ, neurocognition, and complex aspects of social
2.4. Psychopathology and social functioning
cognition between patients with schizophrenia and their
All patients had attended an extensive interdisciplinary diag- matched controls (see Table 3). However, no differences
nostic program based on ICD-10 criteria. Diagnoses were were found regarding the sincere and the simple sarcastic
based on clinical psychiatric interviews and neuropsycho- TASIT film clips. Surprisingly, no differences were found in
logical tests combined with interviews of relatives. All estimated IQ, neurocognition and all social cognitive domains,
4 L. VEDDUM ET AL.

Table 1. Pairwise comparisons of patients with schizophrenia and their matched controls on demographics
and social functioning.
Measure Schizophrenia (N ¼ 21) Matched controls (N ¼ 21) Harrell’s C p Value
Agea 40.48 (34.84;46.11) 25.45 (21.30;29.61) 0.49 (0.31;0.68) .95
Females 11 (52.4) 11 (52.4) – 1.00
Years of education 12.57 (11.31;13.83) 15.98 (14.69;17.26) 0.80 (0.65;0.94) .001
PSPb scores 48.24 (42.57;53.91) 85.67 (82.66;88.68) 0.99 (0.98;1.01) <.0001
Current occupation – <.001
Work 1 (4.8) 14 (66.7)
Student 2 (9.5) 4 (19.0)
Sick leave 13 (61.9) 2 (9.5)
Pension 5 (23.8) 1 (4.8)
a
Matched on group level.
b
Personal and Social Performance Scale.
Continuous variables were examined by Wilcoxon rank-sum test (Mann–Whitney) and reported with mean
(95% CI) and effect size in terms of Harrell’s C (95% CI). Categorical variables were examined by Fisher’s
exact test (if N < 5 in a cell) or Chi2 and reported with the counts and proportions of group total
N (percentage).

Table 2. Pairwise comparisons of patients with autism spectrum disorders (ASD) and their matched
controls on demographics and social functioning.
Measure ASD (N ¼ 11) Matched controls (N ¼ 11) Harrell’s C p Value
Agea 25.09 (20.77;29.41) 25.45 (21.30;29.61) 0.52 (0.24;0.80) .89
Females 7 (63.6) 7(63.6) – 1.00
Years of education 11.73 (9.82;13.63) 15.45 (14.12;16.79) 0.90 (0.72;1.09) .001
PSPb scores 52.45 (44.41;60.50) 87.73 (84.48;90.98) 1.00 .0001
Current occupation <.001
Unemployed 2 (18.2) 0 –
Work 0 4 (36.4)
Student 1 (9.1) 7 (63.4)
Sick leave 1 (9.1) 0
Pension 7 (63.6) 0
a
Matched on group level.
b
Personal and Social Performance Scale.
Continuous variables were examined by Wilcoxon rank-sum test (Mann–Whitney) and reported with mean
(95% CI) and effect size in terms of Harrell’s C (95% CI). Categorical variables were examined by Fisher’s exact
test (if N < 5 in a cell) or Chi2 and reported with the counts and proportions of group total N (percentage).

Table 3. Pairwise comparisons of schizophrenia patients and their matched controls regarding IQ, neurocogni-
tion, and social cognition.
Schizophrenia Matched controls
Measure (N ¼ 21) (N ¼ 21) Harrell’s C p Value
Estimated IQa 98.14 (88.83;107.46) 110.52 (103.66;117.39) 0.68 (050;0.85) .048
BACS composite scoreb 2.31 (3.01; 1.61) 0.06 (0.41;0.54) 0.90 (0.81;1.00) <.0001
TASITc
Sincere 16.05 (14.70;17.39) 17.48 (16.32;18.63) 0.64 (0.47;0.82) .11
Simple sarcasm 14.67 (12.27;17.06) 16.95 (15.61;18.29) 0.63 (0.45;0.81) .14
Paradoxical sarcasm 15.95 (14.19;17.71) 19.52 (18.92;20.13) 0.87 (0.76;0.99) <.0001
Animated Triangles
Appropriateness ToM 6.37 (5.56;7.18)d 8.14 (7.34;8.94) 0.78 (0.63;0.93) .003
Appropriateness Random 11.47 (10.98;11.97)d 10.40 (9.63;11.18) 0.29 (0.13:0.45) .016
Intentionality Random 0.66 (0.02;1.30)d 1.48 (0.77;2.18) 0.69 (0.52;0.85) .033
Intentionality ToM 9.84 (8.15;11.54)d 14.64 (13.48;15.81) 0.86 (0.74;0.98) .0001
a
Wechsler Adult Intelligence Scale-IV (Vocabulary, Similarities, Block Design, and Matrix Reasoning).
b
BACS, Brief Assessment of Cognition in Schizophrenia.
c
The Awareness and Social Inference Test, Part 2A.
d
N ¼ 19.
Variables were examined by Wilcoxon rank-sum test (Mann–Whitney) and reported with mean (95% CI) and
effect size by terms of Harrell’s C (95% CI).

besides TASIT paradoxical sarcasm, when comparing patients p ¼ .005). No differences were found between the two patient
with ASD to their matched controls (see Table 4). groups regarding social cognition, when p values were adjusted
for age and neurocognition (p > .25).

3.3. Comparisons of patients with schizophrenia and

patients with ASD 4. Discussion
We found no difference in estimated IQ (Z¼ 0.56, p ¼ .58), but Surprisingly, we found no significant differences in social
patients with schizophrenia had significantly more comprehensive cognition between schizophrenia and ASD. However, patients
neurocognitive deficits compared to patients with ASD (Z¼ 2.84, with schizophrenia were significantly impaired in both
 NORDIC JOURNAL OF PSYCHIATRY 5

Table 4. Pairwise comparisons of patients with autism spectrum disorders (ASD) and their matched controls regard-
ing IQ, neurocognition, and social cognition.
Measure ASD (N ¼ 11) Matched controls (N ¼ 11) Harrell’s C p Value
Estimated IQa 104.55 (89.11;119.98) 108.09 (99.58;116.60) 0.56 (0.31;0.88) .45
BACS composite scoreb 0.92 (1.73; 0.12) 0.12 (0.75;0.52) 0.66 (0.40;0.92) .20
TASITc
Sincere 17 (15.30;18.70) 17.27 (15.93;18.62) 0.52 (0.24;0.80) .87
Simple sarcasm 15.27 (12.65;17.89) 17.45 (16.10;18.81) 0.66 (0.39;0.93) .20
Paradoxical sarcasm 16.91 (15.31;18.51) 19.36 (18.82;19.91) 0.84 (0.66;1.02) .0056
Animated Triangles
Appropriateness ToM 7.61 (6.20;9.02)d 8 (6.91;9.09) 0.57 (0.28;0.87) .59
Appropriateness Random 10.94 (9.92;11.97)d 10.41 (8.92;11.89) 0.45 (0.17;0.74) .72
Intentionality Random 0.94 (0.12;1.77)d 1.64 (0.13;3.15) 0.54 (0.25;0.83) .75
Intentionality ToM 12.78 (10.46;15.09)d 13.73 (11.73;15.72) 0.63 (0.34;0.92) .34
a
Wechsler Adult Intelligence Scale-IV (Vocabulary, Similarities, Block Design, and Matrix Reasoning).
b
BACS, Brief Assessment of Cognition in Schizophrenia.
c
The Awareness and Social Inference Test, Part 2A.
d
N ¼ 9.
Variables were examined by Wilcoxon rank-sum test (Mann–Whitney) and reported with mean (95% CI) and effect
size by terms of Harrell’s C (95% CI).

implicit and explicit ToM as well as in social perception com- and longer-lasting course of illness to high-functioning
pared to their matched controls, whereas patients with ASD patients with ASD with intact neurocognitive abilities and no
performed on a similar level as their matched controls on mental retardation. This may contribute to the incongruence
most social cognitive domains. As such, our results do not between our preliminary hypothesis and our findings.
support the assumption that implicit ToM is intact in schizo- However, the results regarding ToM are in accordance with
phrenia [28]. Moreover, only the patients with schizophrenia previous research findings suggesting that patients with lon-
were significantly impaired in IQ and neurocognition com- ger-lasting schizophrenia have more severe ToM deficits than
pared to the matched controls and further, patients with patients with first-episodes or remitted schizophrenia [8,25].
schizophrenia had significantly neurocognitive deficits com- Moreover, recent research findings indicate that high-func-
pared to patients with ASD. tioning patients with schizophrenia have preserved social
Although the patients with schizophrenia in the present perceptual abilities compared to low-functioning patients
study seem to have more impaired social cognition than the [47]. Likewise, high-functioning patients with ASD have been
patients with ASD, these differences were not significant found to have less severe ToM deficits compared to low-
when controlling for age and neurocognition. As such, our functioning patients [48,49]. In accordance to this, our results
findings seem to differ from other studies who found that are in congruence with the hypothesis that some high-func-
schizophrenia and ASD differ significantly with regard to tioning patients with ASD are able to pass explicit mentaliz-
ToM deficits [20–23]. On the other hand, our results support ing tasks [29,30]. However, the fact that we did not find any
the findings from studies implying that comparable social significant differences regarding implicit nor explicit ToM
cognitive ToM deficits exist in schizophrenia and ASD between ASD and matched controls seems to be in conflict
[17–19]. It should be noted that we found a trend towards with the existing literature stating that ToM deficits are well-
more social cognitive deficits in the schizophrenia patients known in ASD [6]. Again, it is important to notice that the
(see Table 4), but the difference was non-significant which patients with ASD neither differentiated from the matched
might be due to the small sample size. controls regarding IQ nor neurocognition, and so our ASD
Former studies in this field have used various measures to sample may not be completely representative.
assess social cognition, but none has used TASIT to assess Future research in this area should bear in mind that dif-
explicit ToM and social perception. As suggested in the exist- ferent subgroups exist in both disorders and that these sub-
ing literature, appliance of different social cognitive tasks groups seem to perform differently on social cognitive
might be pivotal to research findings, as the two patient measures. These subgroups for instance involve level of func-
groups may perform differently on distinct tasks [24,25]. tioning, symptom severity, duration of illness, as well as IQ
Furthermore, the different tasks may in essence involve vari- and neurocognitive abilities. As such, it seems fallible to con-
ous abilities or aspects of the complexity of social cognition sider both patient groups as one entire group when making
making it difficult to compare studies. However, in the pre- comparisons of social cognition in schizophrenia and ASD.
sent study, we tried to accede this issue by applying two dis-
tinct social cognitive measures, but still no differences were
4.1. Limitations
found between the two patients groups. Future studies
should aim at continuing to compare the two patient groups The sample size was small. For this reason, replication in
with different social cognitive measures to ensure a more larger cohorts (with more statistical power) is needed.
comprehensive insight into the potential convergence or Moreover, the sample sizes in the two patient groups were
divergence in social cognition between the two disorders. not the same, and the age difference between the two clin-
Notably, our results imply that we have compared low- ical groups was high. But, each patient was closely matched
functioning patients with schizophrenia who had a chronic one-to-one to a healthy control subject and all patients were
6 L. VEDDUM ET AL.

recruited from the same mental health clinic. We investi- [5] Frith CD. The cognitive neuropsychology of schizophrenia. Hove
gated two different patient groups and the patients who (UK): Lawrence Erlbaum Associates; 1992.
[6] Baron-Cohen S. Theory of mind and autism: a fifteen year review.
were referred to the Neuropsychiatric unit were not neces-
In: Tager-Flusberg HC, D. J., editor. Understanding other minds:
sarily optimal medicated. This made it impossible for us to perspectives from developmental cognitive neuroscience. Oxford:
control for medication in our analyses. We cannot rule out, Oxford University Press; 2000. p. 3–20.
that medication may have had an impact on cognition. No [7] Green MF, Penn DL, Bentall R, et al. Social cognition in
symptom scales were used. However, all patients were tested Schizophrenia: an NIMH workshop on definitions, assessment,
and research opportunities. Schizophr Bull. 2008;34:1211–1220.
by the same experienced neuropsychologist and the exact
[8] Bora E, Yucel M, Pantelis C. Theory of mind impairment in schizo-
diagnoses were carefully made on the basis of a comprehen- phrenia: Meta-analysis. Schizophr Res. 2009;109:1–9.
sive and profound psychiatric interview program involving [9] Savla GN, Vella L, Armstrong CC, et al. Deficits in domains of
both the patient and their relatives. social cognition in schizophrenia: a meta-analysis of the empirical
evidence. Schizophr Bull. 2013;39:979–992.
[10] Sprong M, Schothorst P, Vos E, et al. Theory of mind in schizo-
4.2. Conclusion phrenia: meta-analysis. Br J Psychiatry. 2007;191:5.
[11] Frith CD, Frith U. Elective affinities in schizophrenia and child-
The results from the present study imply that different sub- hood autism. In Bebbington PE, editor, Social psychiatry: theory,
groups in both schizophrenia and ASD are characterized by methodology, and practice. New Brunswick: Transaction
differences in severity of ToM deficits and social perceptual Publishers; 1991.
[12] Eack SM, Wojtalik JA, Keshavan MS, et al. Social-cognitive brain
abilities. This seems to be a critical issue in the research of function and connectivity during visual perspective-taking in aut-
possible overlaps and diversities in social cognition between ism and schizophrenia. Schizophr Res. 2017;183:102–109.
these two mental disorders. Moreover, an important meth- [13] Meyer U, Feldon J, Dammann O. Schizophrenia and autism: both
odological question regarding measurement of social cogni- shared and disorder-specific pathogenesis via perinatal inflamma-
tion is emphasized. More specific, the results indicate that tion?. Pediatr Res. 2011;69:26R–33R.
[14] Nakagawa Y, Chiba K. Involvement of neuroinflammation during
we might have compared low-functioning patients with brain development in social cognitive deficits in autism spectrum
schizophrenia to a group of high-functioning patients disorder and schizophrenia. J Pharmacol Exp Therapeut. 2016;
with ASD. 358:504–515.
As such, low-functioning patients with schizophrenia may [15] Sugranyes G, Kyriakopoulos M, Corrigall R, et al. Autism spectrum
be more severely impaired in both neurocognition and social disorders and schizophrenia: meta-analysis of the neural corre-
lates of social cognition. PloS One. 2011;6:e25322.
cognition than high-functioning patients with ASD.
[16] Sasson NJ, Pinkham AE, Carpenter KLH, et al. The benefit of dir-
Prospective research should aim at comparing different sub- ectly comparing autism and schizophrenia for revealing mecha-
groups in both disorders in order to clarify the exact conver- nisms of social cognitive impairment. J Neurodevelop Disord.
gence and divergence in social cognition between these two 2011;3:87–100.
patient groups. This should be done by using different social [17] Couture SM, Penn DL, Losh M, et al. Comparison of social cogni-
tive functioning in schizophrenia and high functioning autism:
cognitive measures, and with an awareness of the diversity
more convergence than divergence. Psychol Med. 2010;40:
of particular tasks. 569–579.
[18] Craig JS, Craig FB, Hatton C, et al. Persecutory beliefs, attributions
and theory of mind: comparison of patients with paranoid delu-
Disclosure statement sions, Asperger’s syndrome and healthy controls. Schizophrenia
Res. 2004;69:29–33.
No potential conflict of interest was reported by the authors.
[19] Murphy D. Theory of mind in Asperger’s syndrome, schizophrenia
and personality disordered forensic patients. Cognit
Neuropsychiatry. 2006;11:99–111.
Funding [20] Ozguven HD, Oner O, Baskak B, et al. Theory of mind in schizo-
This study was financially supported by the Psykiatriens forskningsfond, phrenia and Asperger’s syndrome: relationship with negative
Central Denmark Region. symptoms. Klin Psikofarmakol B€ ul – Bull Clin Psychopharmacol.
2010;20:5–13.
[21] Radeloff D, Ciaramidaro A, Siniatchkin M, et al. Structural altera-
References tions of the social brain: a comparison between schizophrenia
and autism. PloS One. 2014;9:e106539.
[1] Hinterbuchinger B, Kaltenboeck A, Baumgartner JS, et al. Do [22] Lugnegaard T, Unenge Hallerb€ack M, Hj€arthag F, et al. Social cog-
patients with different psychiatric disorders show altered social nition impairments in Asperger syndrome and schizophrenia.
decision-making? A systematic review of ultimatum game experi- Schizophr Res. 2012;143:277–284.
ments in clinical populations. Cognit Neuropsychiatry. 2018;23: [23] Martinez G, Alexandre C, Mam-Lam-Fook C, et al. Phenotypic con-
117–141. tinuum between autism and schizophrenia: evidence from the
[2] Cotter J, Granger K, Backx R, et al. Social cognitive dysfunction as Movie for the Assessment of Social Cognition (MASC). Schizophr
a clinical marker: a systematic review of meta-analyses across 30 Res. 2017;185:161–166.
clinical conditions. Neurosci Biobehav Rev. 2018;84:92–99. [24] Chung YS, Barch D, Strube M. A meta-analysis of mentalizing
[3] Tulacı RG, Cankurtaran EŞ, Ozdel€ K, et al. The relationship impairments in adults with schizophrenia and autism spectrum
between theory of mind and insight in obsessive-compulsive dis- disorder. Schizophr Bull. 2014;40:602–616.
order. Nord J Psychiatry. 2018;72:273–280. [25] Bliksted V, Ubukata S, Koelkebeck K. Discriminating autism spec-
[4] Erol A, Akyalcin Kirdok A, Zorlu N, et al. Empathy, and its relation- trum disorders from schizophrenia by investigation of mental
ship with cognitive and emotional functions in alcohol depend- state attribution on an on-line mentalizing task: a review and
ency. Nord J Psychiatry. 2017;71:205–209. meta-analysis. Schizophr Res. 2016;171:16–26.
 NORDIC JOURNAL OF PSYCHIATRY 7

[26] Apperly IA, Butterfill SA. Do humans have two systems to track [39] Castelli F, Happ e F, Frith C, et al. Movement and mind: a func-
beliefs and belief-like states?. Psychol Rev. 2009;116:953–970. tional imaging study of perception and interpretation of complex
[27] Heyes CM, Frith CD. The cultural evolution of mind reading. intentional movement patterns. NeuroImage. 2000;12:314–325.
Science (New York, NY). 2014;344:1243091–1243091. [40] Castelli F, Frith C, Happe F, et al. Autism, Asperger syndrome and
[28] Frith CD. Schizophrenia and theory of mind. Psychol Med. 2004; brain mechanisms for the attribution of mental states to ani-
34:385–389. mated shapes. Brain: J Neurol. 2002;125:1839–1849.
[29] Bowler DM. "Theory of mind" in Asperger’s syndrome. J Child [41] Bliksted V, Fagerlund B, Weed E, et al. Social cognition and neu-
Psychol Psychiatry. 1992;33:877. rocognitive deficits in first-episode schizophrenia. Schizophr Res.
[30] Happe FGE. The role of age and verbal ability in the theory of 2014;153:9–17.
[42] McDonald S, Flanagan S, Martin I, et al. The ecological validity of
mind task performance of subjects with autism. Child Dev. 1995;
TASIT: a test of social perception. Neuropsychol Rehabil. 2004;14:
66:843–855.
285–302.
[31] Frith U. Why we need cognitive explanations of autism. Q J Exp
[43] Keefe RSE, Goldberg TE, Harvey PD, et al. The Brief Assessment of
Psychol. 2012;65:2073. Cognition in Schizophrenia: reliability, sensitivity, and comparison
[32] McDonald S, Bornhofen C, Shum D, et al. Reliability and validity with a standard neurocognitive battery. Schizophr Res. 2004;68:
of The Awareness of Social Inference Test (TASIT): a clinical test 283–297.
of social perception. Disabil Rehabil. 2006;28:1529–1542. [44] Keefe RSE, Harvey PD, Goldberg TE, et al. Norms and standardiza-
[33] Frith U. The neurocognitive basis of autism. Trends Cognit Sci. tion of the Brief Assessment of Cognition in Schizophrenia
1997;1:73–77. (BACS). Schizophr Res. 2008;102:108–115.
[34] Green MF, Nuechterlein KH, Gold JM, et al. Approaching a con- [45] Wechsler D. WAIS-IV: technical and interpretive manual. San
sensus cognitive battery for clinical trials in schizophrenia: the Antonio, TX: Pearson; 2008.
NIMH-MATRICS conference to select cognitive domains and test [46] Morosini PL, Magliano L, Brambilla L, et al. Development, reliabil-
criteria. Biol Psychiatry. 2004;56:301–307. ity and acceptability of a new version of the DSM-IV Social and
[35] First MB, Tasman A. Childhood disorders: pervasive developmen- Occupational Functioning Assessment Scale (SOFAS) to assess
tal disorders. DSM-IV-TR mental disorders: diagnoses, etiology routine social functioning. Acta Psychiatr Scand. 2000;101:
323–329.
and treatment. New York: Chichester John Wiley & Sons; 2004. p.
[47] Karpouzian TM, Alden EC, Reilly JL, et al. High functioning indi-
127–157.
viduals with schizophrenia have preserved social perception but
[36] Wing L. Mental retardation and the autistic continuum. In:
not mentalizing abilities. Schizophr Res. 2016;171:137–139.
Bebbington PE, editor. Social Psychiatry. Vol.1. New Brunswick [48] Begeer S, Malle BF, Nieuwland MS, et al. Using Theory of Mind to
(USA) and London (UK): Transaction Publishers; 1991. p. 113–138. represent and take part in social interactions: comparing individ-
[37] W.H.O. The ICD-10 classification of mental and behavioral disor- uals with high-functioning autism and typically developing con-
ders. Geneva: World Health Organization; 1993. trols. Eur J Dev Psychol. 2010;7:104–122.
[38] Abell F, Happe F, Frith U. Do triangles play tricks? Attribution of [49] Roeyers H, Demurie E. How impaired is mind-reading in high-
mental states to animated shapes in normal and abnormal devel- functioning adolescents and adults with autism?. Eur J Dev
opment. Cognit Dev. 2000;15:1–16. Psychol. 2010;7:123–134.