Psychological Medicine Prevalence of cognitive impairments and

cambridge.org/psm
strengths in the early course of psychosis
and depression
Alexandra Stainton1,2 , Katharine Chisholm3, Siân Lowri Griffiths4,
Original Article
Lana Kambeitz-Ilankovic5,6, Julian Wenzel5, Carolina Bonivento7,
*Joint senior authorship.
Paolo Brambilla8,9, Mariam Iqbal10, Theresa K. Lichtenstein5, Marlene Rosen5,
†The notes appear after the main text.
Linda A. Antonucci11, Eleonora Maggioni12, Joseph Kambeitz5,
Cite this article: Stainton A et al (2023). Stefan Borgwardt13,14, Anita Riecher-Rössler15, Christina Andreou14,
Prevalence of cognitive impairments and
strengths in the early course of psychosis and André Schmidt14, Frauke Schultze-Lutter16,17,18, Eva Meisenzahl16,
depression. Psychological Medicine 53,
5945–5957. https://doi.org/10.1017/ Stephan Ruhrmann5, Raimo K. R. Salokangas19, Christos Pantelis20,
S0033291723001770
Rebekka Lencer13,21, Georg Romer22, Alessandro Bertolino11,
Received: 28 February 2023
Revised: 12 May 2023
Rachel Upthegrove4,23, Nikolaos Koutsouleris24,25,26, Kelly Allott1,2, *,
Accepted: 1 June 2023 Stephen J. Wood1,2,27,* and on behalf of the PRONIA Consortium†*
First published online: 6 July 2023

Keywords:
Mental health; psychosis; working memory; Abstract
processing speed; verbal learning
Background. Studies investigating cognitive impairments in psychosis and depression have
Corresponding author: typically compared the average performance of the clinical group against healthy controls
Alexandra Stainton; (HC), and do not report on the actual prevalence of cognitive impairments or strengths within
Email: [email protected] these clinical groups. This information is essential so that clinical services can provide
adequate resources to supporting cognitive functioning. Thus, we investigated this prevalence
in individuals in the early course of psychosis or depression.
Methods. A comprehensive cognitive test battery comprising 12 tests was completed by 1286 indi-
viduals aged 15–41 (mean age 25.07, S.D. 5.88) from the PRONIA study at baseline: HC (N = 454),
clinical high risk for psychosis (CHR; N = 270), recent-onset depression (ROD; N = 267), and recent-
onset psychosis (ROP; N = 295). Z-scores were calculated to estimate the prevalence of moderate or
severe deficits or strengths (>2 S.D. or 1–2 S.D. below or above HC, respectively) for each cognitive test.
Results. Impairment in at least two cognitive tests was as follows: ROP (88.3% moderately,
45.1% severely impaired), CHR (71.2% moderately, 22.4% severely impaired), ROD (61.6%
moderately, 16.2% severely impaired). Across clinical groups, impairments were most preva-
lent in tests of working memory, processing speed, and verbal learning. Above average per-
formance (>1 S.D.) in at least two tests was present for 40.5% ROD, 36.1% CHR, 16.1%
ROP, and was >2 SDs in 1.8% ROD, 1.4% CHR, and 0% ROP.
Conclusions. These findings suggest that interventions should be tailored to the individual, with
working memory, processing speed, and verbal learning likely to be important transdiagnostic
targets.

Introduction
Cognitive impairments are a prominent feature of early-stage mental illnesses, particularly in
full-threshold psychotic disorder (Fioravanti, Bianchi, & Cinti, 2012; Mesholam-Gately,
Giuliano, Goff, Faraone, & Seidman, 2009), clinical high-risk (CHR) for psychosis (Catalan
et al., 2021; Fusar-Poli et al., 2012; Pukrop et al., 2006), and major depressive disorder
(MDD; Ahern & Semkovska, 2017; Goodall et al., 2018). Such impairments can significantly
impact an individual’s everyday functioning and long-term outcomes. Cognition has been
© The Author(s), 2023. Published by found to be predictive of progression through the psychosis spectrum (Koutsouleris et al.,
Cambridge University Press. This is an Open 2011). Further, in a transdiagnostic sample of young people with mental illness, baseline cog-
Access article, distributed under the terms of
nition was the strongest predictor of two-year functional outcomes (Lee et al., 2013). While the
the Creative Commons Attribution licence
(http://creativecommons.org/licenses/by/4.0/), literature to date has clearly demonstrated a meaningful link between mental health difficulties
which permits unrestricted re-use, distribution and cognitive impairments, studies have largely taken a group-level approach to analysis. Such
and reproduction, provided the original article studies have shown that individuals with mental illnesses, on average, perform more poorly
is properly cited. than healthy controls (HC) in multiple cognitive domains (East-Richard, R-Mercier,
Nadeau, & Cellard, 2020). However, statistically significant differences at a group level cannot
necessarily equal a clinically meaningful impairment (Abramovitch & Schweiger, 2015;
Abramovitch, Short, & Schweiger, 2021; Michel, Ruhrmann, Schimmelmann, Klosterkötter,
& Schultze-Lutter, 2014). At present there is no universally agreed definition of a clinically
5946 Alexandra Stainton et al.

meaningful cognitive impairment (Abramovitch et al., 2021). Holshausen, and Bowie (2021) also investigated the prevalence
However, by examining the amount to which individual perform- of impairment relative to the individuals’ estimated premorbid
ance deviates from the average (standard deviation difference), we performance in a community sample of individuals with MDD.
can obtain a clearer picture of how common moderate or severe They identified that 62.2% of the sample were performing at
cognitive difficulties may be within early intervention services, least 1 S.D. below their estimated premorbid performance.
and who may require further cognitive assessment or treatment. Finally, some studies have calculated the prevalence of a compos-
Clinical guidelines for early psychosis (e.g. Early Psychosis ite cognitive impairment (the average of cognitive test scores).
Guidelines Working Group, 2016; NICE, 2014) and depression The prevalence of composite impairment in outpatient MDD
(e.g. Malhi et al., 2021; National Institute for Clinical samples has ranged from 11.8% (Douglas et al., 2018) to 25.2%
Excellence, 2009) recommend an assessment of cognition and (Tran et al., 2021) at a level of >1 S.D. relative to HC, and has
remediation therapies where cognition may be impairing func- been observed in 1.5% of an outpatient MDD sample at >2 SDs
tional recovery. Thus, it is important to estimate the prevalence relative to HC (Douglas et al., 2018). Thus, while rates of impair-
of people presenting to services with impaired cognition who ment appear high in adults with established MDD, less is known
may need to access those additional services. This would allow about the prevalence of cognitive impairment in younger
for more adequate resourcing of clinical services and could inform recent-onset samples.
relevant training for clinicians. Furthermore, because there may Beside understanding the prevalence of cognitive impairments
be a variation in the cognitive strengths or impairments in people during the early stages of serious mental illnesses, it is equally
presenting to services, it is important to screen cognition at ser- important to consider the prevalence of cognitive strengths.
vice entry (Bryce & Allott, 2019; Bryce, Bowden, Wood, & Current approaches to treatment of cognition often focus on
Allott, 2021), and consider the cognitive profile of individuals the remediation of deficits. However, reinforcing and building
who are in the early course of mental illness. Cognitive impair- on strengths could be useful as an adjunct to such remediation
ments are often present long before the onset of clinical symp- (Allott et al., 2020), to prevent further deterioration (Pantelis
toms (e.g. Mollon, David, Zammit, Lewis, & Reichenberg, et al., 2015), and to leverage important psychological factors for
2018), and may also show further decline following the first- the treatment process such as self-esteem and motivation (Allott
episode of psychosis (Fett et al., 2020; Flaaten et al., 2022), or et al., 2020). By understanding the prevalence of cognitive
with multiple depressive episodes (Allott, Fisher, Amminger, strengths, we can provide clinicians and services with a more
Goodall, & Hetrick, 2016; Semkovska et al., 2019). Thus, under- comprehensive picture of cognition during the first-episode,
standing the profile of cognitive performance in the early course which could be used to inform service delivery and training. To
of illness will allow clinicians to intervene early to preserve intact our knowledge, no study has yet examined the prevalence of cog-
cognitive skills or prevent further decline (Pantelis, Wannan, nitive strengths in mental illness.
Bartholomeusz, Allott, & McGorry, 2015), providing the best pos- In summary, the studies conducted to date have demonstrated
sible chance of functional recovery. that there is a significant association between serious mental dis-
The idea that a subgroup of individuals with schizophrenia orders and cognitive impairment at the group level. Further, indi-
may be ‘neuropsychologically normal’ has been previously dis- viduals with psychosis or depression may exhibit poorer
cussed (Keefe, 2008; Wilk et al., 2005), and some more recent performance on cognitive tasks than they would have if they
studies have used data-driven approaches, which identified clus- had never developed the illness (Keefe, Eesley, & Poe, 2005;
ters of relatively spared v. impaired cognitive performance in Tran et al., 2021). However, few studies have reported the preva-
early psychosis (Gould et al., 2014; Uren, Cotton, Killackey, lence of cognitive impairments in the earliest stages of mental ill-
Saling, & Allott, 2017; Wenzel et al., 2021). Few studies have ness, and no studies have reported the prevalence of cognitive
examined the prevalence of cognitive impairments in this popula- strengths. We aimed to report the prevalence of clinically mean-
tion. Two previous reports estimated that 70–80% of individuals ingful levels of cognitive strengths or impairments in a multi-
with chronic schizophrenia demonstrate cognitive impairments diagnostic sample of individuals diagnosed with recent-onset
(Allen, Goldstein, & Warnick, 2003; Palmer et al., 1997). Less is psychosis (ROP) or depression, or identified at clinical high-risk
known about the prevalence of cognitive impairments in the for psychosis. We categorised cognitive test performance at vari-
early stages of the illness. In a sample of Ugandan inpatients ous levels ranging from >2 SDs above average to >2 SDs below
with first-episode psychosis, 62% presented with a cognitive average. We defined impairments and strengths as both moderate
impairment, defined as mean scores of >2 SDs below HC in (1–2 SDs) and extreme (>2 SDs) in two or more tests, as well as
one domain, or of >1 SDs below HC in two or more domains report a composite cognitive score across these levels. This defin-
(Mwesiga et al., 2022). Furthermore, to the best of our knowledge ition allows for comparison with previous literature (e.g. Douglas
there are no previous studies which have investigated the preva- et al., 2018; Tran et al., 2021), and, in the case of impairment,
lence of cognitive impairments in CHR. highlights that which might require clinical attention.
The prevalence of cognitive impairments in MDD has been
investigated in a few studies. Gualtieri and Morgan (2008)
observed that up to 32% of their treated MDD sample had an Methods
impairment of 1–2 SDs relative to HC in at least one cognitive
Participants
domain, which was >2 SDs in up to 18%. When cognitive impair-
ment is defined as being >1 S.D. below HC in at least two cognitive This sample comprised 1286 individuals aged 15–41 who were
tests, Douglas et al. (2018) observed a prevalence of 78.6% in their recruited into the multi-site ‘Personalised Prognostic Tools for
outpatient MDD sample, which was >2 SDs below HC in 14.3% of Early Psychosis Management’ (PRONIA; https://www.pronia.eu)
the sample. They also observed that their inpatient depression study and completed a cognitive test battery at baseline.
group had higher rates of impairment, at 91.4% (>1 S.D. in two Although the maximum age for inclusion was 40, one participant
tests) and 32.8% (>2 SDs in two tests). Tran, Milanovic, in the HC group was assessed on the date of their 41st birthday
Psychological Medicine 5947

and retained in analysis. Full details on the methods and recruit- outcome scores for each cognitive test were chosen, with any
ment of the PRONIA study are provided in Koutsouleris et al. ‘reaction time’ or ‘error’ scores reversed by subtracting the indi-
(2018). In summary, participants were recruited from ten inter- vidual raw score from the highest raw score on that test, so that
national sites (Germany: Munich, Cologne, Münster, and higher scores always indicated better performance. Following
Düsseldorf; UK: Birmingham; Italy: Udine, Bari, and Milan; this, z-scores based on the mean and standard deviation (S.D.)
Finland: Turku; Switzerland: Basel) if they met criteria for ROP, of the HC study group were calculated using the formula: (raw
recent-onset depression (ROD), or clinical high risk for psychosis score – HC mean) / HC S.D.. A ‘composite’ cognitive score was
(CHR), or for HC with no personal or family history of mental computed for each individual as the average of all twelve z-scores.
illness. Inclusion and exclusion criteria for each study group are Missing data values were small in this dataset, particularly in ele-
summarised in online Supplementary Table S1. All participants ven of the twelve cognitive tasks (0.7–5.4%). However, more data
provided their written consent (or assent for participants aged was missing on the RAVLT (12.8%), as an alternative measure was
<18) after having received all study information. All sites gained used by the Finnish site, and those Finnish data were not used in
approval from their respective ethics committees. the current study. The prevalence of various levels of cognitive
impairment or strength was then investigated by allocating the
proportion of z-scores for each cognitive test into the following
Procedure
categories: >2 SDs below HC (severely impaired); 1–2 SDs
At baseline, participants were screened for inclusion and then below HC (moderately impaired); ‘average performance’, meaning
completed the cognitive test battery and demographic measures that performance was within 1 S.D. relative HC; 1–2 SDs above HC
including age, sex, education, and ethnicity. Cognitive tests were (above average), and >2 SDs above HC (extremely high).
completed in a standardised order and in the native language of Chi-square analysis examined whether there was a significantly
each site. different proportion of the study groups meeting each level of
cognitive performance. Post-hoc comparisons (with Bonferroni
correction for multiple comparisons) were conducted to examine
Measures
any group differences in each level of cognitive performance for
The cognitive test battery comprised the following tests and pri- each individual test. For each test and the composite score, we
mary outcome scores (also see online Supplementary Material, also calculated the odds ratio of each clinical group demonstrating
‘Description of Cognitive Tasks’ and online Supplementary a cognitive impairment (anything >1 S.D.) relative to HC and the
Table S2): Trail Making Test (TMT) parts A & B, total reaction 95% confidence intervals of the odds ratio, using the formula pro-
time (Army Individual Test Battery, 1944); phonetic (PVF) and posed by Altman (1990). Next, we examined the proportion of
semantic (SVF) verbal fluency, total number of correct words each study group showing moderate or severe cognitive impair-
(Benton, Hamsher, & Sivan, 1994); Continuous Performance ment, or above average performance on a given number of tests.
Task (CPT), d-Prime Sensitivity Index (Cornblatt, Risch, Faris, Cognitive impairment (or strength) was reported at the level of
Friedman, & Erlenmeyer-Kimling, 1988); Rey Auditory Verbal both 1–2 and >2 SDs relative to HC in at least two tests, or in
Learning Test (RAVLT), sum of trials 1–5 (Rey, 1964); the composite score. This definition was chosen to facilitate com-
Rey-Osterrieth Complex Figure (ROCF), total score at immediate parison with previous research (e.g. Douglas et al., 2018; Tran
recall (Osterrieth, 1944; Rey, 1941); Self-Ordered Pointing Task et al., 2021). For each participant group, the mean number of
(SOPT), total errors (Petrides & Milner, 1982); Auditory Digit tests impaired or above average at both levels was also calculated.
Span task, number of correct trials forwards (FDS) and backwards
(BDS) (Wechsler, 1939); Digit-Symbol Substitution Test (DSST),
Results
total score of number correct subtracted by number of errors
(Copyright free version, component of the Wechsler Adult The characteristics of the sample are reported in Table 1. The four
Intelligence Scale-Revised; Wechsler, 1981); Diagnostic Analysis study groups significantly differed on age, sex, site, IQ estimate,
of Non-Verbal Accuracy (DANVA), total number of correct years of education, and ethnicity. Post-hoc examinations of
faces identified (Nowicki & Duke, 1994); and the vocabulary these group differences are shown in Table 1. Mean raw cognitive
and matrix reasoning subscales of the Wechsler Abbreviated scores for each study group are presented in online
Scale of Intelligence (Wechsler, 2011). These tests were chosen Supplementary Table S3.
to represent a comprehensive evaluation of the participant’s cog-
nitive functioning. Results were reported according to individual
Prevalence of cognitive strengths and impairment
test scores to facilitate use of the findings for clinicians.
Clinicians may use different cognitive test batteries in practice The percentages of each study group demonstrating each level of
and thus, by reporting individual tests they will be able to directly cognitive performance are presented in Table 2. Figures 1a and b
check the expected prevalence of strength or impairment in a clin- show the prevalence of any level of impairment (e.g. anything >1
ical population, relative to controls, on a particular test. S.D. below HC) and strengths (e.g. anything >1 S.D. above HC).
Though the focus of this paper was only to report the prevalence
of cognitive impairments and strengths, we also examined
Statistical analysis
between-group differences. Chi-square analysis showed that
Statistical analysis was completed using IBM SPSS Statistics 25 there was a significant difference in the proportion of participants
software (IBM Corp, 2017). Group differences on demographic in each study group demonstrating each level of cognitive impair-
variables were examined using one-way ANOVA for continuous ment in all cognitive tests. Post-hoc analysis of significant between
variables, or chi-squared for categorical variables. Post-hoc ana- group differences is presented in online Supplementary Table S6.
lysis then determined the direction of any significant group Table 2 shows the odds ratio and 95% confidence interval for that
effects, using Tamhane or Bonferroni adjustment. The primary odds ratio, showing how much higher the odds of a cognitive
 5948
Table 1. Characteristics of the sample

Whole sample
(N = 1286) HC (N = 454) ROD (N = 267) CHR (N = 270) ROP (N = 295) Statistic p value Post-hoc

Age CHR<HC, p = 0.003*

CHR<ROD, p = 0.001*
Mean (S.D.) 25.07 (5.88) 25.17 (5.99) 26.64 (6.13) 23.67 (5.26) 25.67 (5.84) F = 6.939 <0.001* CHR<ROP, p < 0.001*
a a
Range 15–41 15–41 15–40 15–40 15–40
Gender ROP>HC (Male), p < 0.001*
ROP>CHR (Male), p = 0.032*
Male % 47.5 40.6 49.2 46.1 57.8 χ = 21.771
2
<0.001*
Estimated HC>CHR, p < 0.001*
Current IQ HC>ROD, p < 0.001*
HC>ROP, p < 0.001*
Mean (S.D.) 105.58 (12.09) 110.00 (10.03) 105.54 (11.83) 105.46 (10.99) 98.92 (13.10) F = 56.655 <0.001* CHR>ROP, p < 0.001*
ROD>ROP, p < 0.001*
Education Years HC>CHR, p < 0.001*
HC>ROD, p < 0.001*
Mean (S.D.) 14.60 (3.12) 15.71 (3.09) 14.35 (3.06) 13.60 (2.55) 14.00 (3.20) F = 34.339 <0.001* HC>ROP, p < 0.001*
ROD>CHR, p = 0.014*
Ethnicity
N (%) χ2 = 43.591 <0.001* ROD>ROP (White), p = 0.004*
White 1067 (83.0) 381 (83.9) 235 (88.0) 226 (83.7) 225 (76.3) HC>ROD (Asian), p = 0.001*
Asian 72 (5.6) 39 (8.6) 4 (1.5) 11 (4.1) 18 (6.1) ROP>ROD (Asian), p = 0.026*
Black 26 (2.0) 6 (1.3) 3 (1.1) 4 (1.5) 13 (4.4) ROP>HC (Black), p = 0.041*
Mixed race 28 (2.2) 14 (3.1) 3 (1.1) 4 (1.5) 7 (2.4) CHR>HC (Other), p = 0.045*
Other 62 (4.8) 10 (2.2) 16 (6.0) 16 (5.9) 20 (6.8) ROD>HC (Other), p = 0.045*
Unknown 31 (2.4) 4 (0.9) 6 (2.2) 9 (3.3) 12 (4.1) ROP>HC(Other), p = 0.008*
N, Number of Participants; S.D., Standard Deviation; F, one-way ANOVA; χ2, Chi-squared; *, Statistically significant at a level of p⩽0.05.

Alexandra Stainton et al.

a
Note: One participant in the HC group was assessed on the date of their 41st birthday and retained in analysis.
Psychological Medicine 5949

Table 2. Percentage of participants in each group meeting the criteria for each level of cognitive impairment or strengths per cognitive test

Cognitive test (N ) HC ROD CHR ROP χ2 (df) p value Odds Ratio 95% CI for Odds Ratio

TMT A (433) (248) (256) (281) χ2 = 49.943 <0.001* ROD- 1.50 ROD-1.01–2.23
>2 SDs Above 0 0.4 0 0 (12) CHR-1.91 CHR-1.31–2.79
1–2 SDs Above 14.1 10.9 9 6.4 ROP- 2.78 ROP-1.95–3.97
No impairment 70 66.5 64.5 59.1
1–2 SDs Below 11.1 12.1 12.9 17.4
>2 SDs Below 4.8 10.1 13.7 17.1
TMT B (435) (247) (255) (280) χ2 = 101.057 <0.001* ROD- 1.67 ROD-1.09–2.55
>2 SDs Above 0 0 0.4 0 (12) CHR- 2.22 CHR-1.48–3.32
1–2 SDs Above 10.6 6.1 3.9 1.8 ROP- 4.68 ROP-3.23–6.77
No impairment 76.8 74.5 71.4 57.9
1–2 SDs Below 8.5 8.5 11.8 19.6
>2 SDs Below 4.1 10.9 12.5 20.7
Phonetic verbal (450) (264) (266) (288) χ2 = 70.502 <0.001* ROD- 2 ROD-1.35–2.96
fluency
>2 SDs Above 4.2 1.5 2.3 1 (12) CHR- 1.81 CHR-1.22–2.70
1–2 SDs Above 12.9 8.3 10.5 4.9 ROP- 3.35 ROP-2.33–4.83
No impairment 69.6 66.7 65.4 60.1
1–2 SDs Below 12 21.6 21.1 27.8
>2 SDs Below 1.3 1.9 0.8 6.3
Semantic verbal (451) (264) (266) (287) χ2 = 127.906 <0.001* ROD- 2.14 ROD-1.47–3.13
fluency
>2 SDs Above 2.7 2.3 1.5 0.3 (12) CHR- 2.51 CHR-1.73–3.64
1–2 SDs Above 12.2 9.8 5.6 3.1 ROP- 5.66 ROP-3.98–8.03
No impairment 70.7 61.4 63.2 47.7
1–2 SDs Below 12 20.1 24.8 33.1
>2 SDs Below 2.4 6.4 4.9 15.7
CPT d’ (451) (263) (266) (288) χ2 = 96.778 <0.001* ROD- 1.74 ROD-1.17–2.59
>2 SDs Above 2.2 4.6 3.4 0 CHR- 2.12 CHR-1.44–3.10
1–2 SDs Above 13.7 9.5 7.9 5.2 (12) ROP- 4.04 ROP-2.83–5.76
No impairment 70.1 63.9 63.2 55.2
1–2 SDs Below 12.4 17.9 24.4 30.9
>2 SDs Below 1.6 4.2 1.1 8.7
RAVLT Trials 1–5 (389) (245) (238) (249) χ2 = 123.299 <0.001* ROD- 1.69 ROD-1.14–2.50
>2 SDs Above 0 0 0 0 (9) CHR- 2.21 CHR-1.50–3.24
1–2 SDs Above 14.4 10.6 8 6 ROP- 5.72 ROP-3.97–8.23
No impairment 68.9 64.1 61.3 40.6
1–2 SDs Below 12.3 15.9 20.6 25.7
>2 SDs Below 4.4 9.4 10.1 27.7
ROCF Immediate (452) (266) (265) (288) χ2 = 88.896 <0.001* ROD- 1.55 ROD-1.05–2.28
memory
>2 SDs Above 0 0 0 0 (9) CHR- 1.80 CHR-1.23–2.64
1–2 SDs Above 17 15.4 12.1 7.3 ROP- 3.91 ROP-2.76–5.53
No impairment 67.7 62.8 63.4 51.4
1–2 SDs Below 11.7 14.7 12.1 20.1
(Continued )
5950 Alexandra Stainton et al.

Table 2. (Continued.)

Cognitive test (N ) HC ROD CHR ROP χ2 (df) p value Odds Ratio 95% CI for Odds Ratio

>2 SDs Below 3.5 7.1 12.5 21.2

SOPT total errors (449) (263) (267) (286) χ2 = 100.049 <0.001* ROD-1.75 ROD-1.21–2.53
>2 SDs Above 0 0 0 0 (9) CHR- 1.58 CHR-1.09–2.29
1–2 SDs Above 17.8 13.3 12 6.6 ROP- 3.98 ROP-2.83–5.58
No impairment 65.5 6.5 63.7 48.6
1–2 SDs Below 12.5 19.4 14.2 21.7
>2 SDs Below 4.5 6.8 10.1 23.1
Forward digit span (452) (263) (268) (292) χ2 = 62.584 <0.001* ROD- 2.02 ROD-1.43–2.85
>2 SDs Above 3.3 1.9 3.7 1.4 (12) CHR- 1.41 CHR-0.99–2.02
1–2 SDs Above 17.7 11.8 12.7 4.8 ROP- 2.43 ROP-1.74–3.38
No impairment 59.3 53.2 57.8 56.5
1–2 SDs Below 17.5 25.9 19 26
>2 SDs Below 2.2 7.2 6.7 11.3
Backward digit span (451) (263) (269) (292) χ2 = 126.958 <0.001* ROD- 1.85 ROD-1.32–2.58
>2 SDs Above 0 0 0 0 (9) CHR- 2.03 CHR-1.46–2.83
1–2 SDs Above 25.9 18.6 14.9 5.8 ROP- 4.71 ROP-3.42–6.48
No impairment 51.2 46 47.6 36
1–2 SDs Below 22.4 35 37.2 54.1
>2 SDs Below 0.4 0.4 0.4 4.1
DSST Total score (453) (264) (265) (291) χ2 = 271.929 <0.001* ROD- 2.32 ROD-1.59–3.39
>2 SDs Above 2 1.9 1.5 0 (12) CHR- 3.29 CHR-2.28–4.75
1–2 SDs Above 12.1 9.5 4.9 1 ROP- ROP-7.24–14.77
10.34
No impairment 72 61.4 58.9 36.4
1–2 SDs Below 13 23.5 28.7 36.1
>2 SDs Below 0.9 3.8 6 26.5
DANVA Total correct (453) (264) (269) (291) χ2 = 64.349 <0.001* ROD- 0.94 ROD-0.63–1.40
>2 SDs Above 0.9 1.9 0.4 0 (12) CHR- 1.36 CHR-0.94–1.97
1–2 SDs Above 17 9.8 11.9 9.6 ROP- 2.19 ROP-1.56–3.09
No impairment 63.8 70.8 64.3 57.4
1–2 SDs Below 13.5 12.9 16 15.5
>2 SDs Below 4.9 4.5 7.4 17.5
Composite scorea (363) (227) (219) (230) χ2 = 138.974 <0.001* ROD- 2.98 ROD-1.6–5.56
>2 SDs Above 0 0 0 0 (9) CHR- 3.74 CHR-2.03–6.88
1–2 SDs Above 1.1 0.9 0.5 0 ROP- ROP-7.24–21.97
12.61
No impairment 94.2 86.3 84 61.7
1–2 SDs Below 4.7 11 14.6 27.8
>2 SDs Below 0 1.8 0.9 10.4
χ , Chi-squared examining differences in the proportion of each study group at each level of performance; *, Statistically significant at a level of p ⩽ 0.05; Odds Ratio, The odds of the clinical
2

group having any impairment >1 S.D., relative to HC; CI, Confidence Interval.
a
Calculated only for participants who completed all tests.
Psychological Medicine 5951

Figure 1. Prevalence of impairments (a); >1 S.D. below average and strengths (b); >1 S.D. above average per group in the individual cognitive tests.
Note: Sample sizes provided are of the whole sample. Individual sample sizes for each test can be found in Table 2. Tests are presented in order from highest to
lowest prevalence in the ROP group.

impairment (>1 S.D. below HC) are for the clinical group com- Regarding above average performance, 16.1% of the ROP
pared to HC. group were above average (1–2 SDs above HC) on at least two
Next, we examined the proportion of each study group show- tests, but no participant in the ROP group performed at the
ing moderate or severe cognitive impairment on a given number ‘extremely high’ level (>2 SDs above HC) on at least two tests.
of tests (Table 3). We observed that 88.3% of the ROP group were In the CHR group, 36.1% were above average on at least two
moderately impaired (1–2 SDs below HC) on at least two tests, tests, and 1.4% were extremely high on at least two tests. In the
and 45.1% showed severe impairment (>2 SDs below HC) on at ROD group, 40.5% were above average on at least two cognitive
least two tests i In the CHR group, 71.2% were moderately tests, and 1.8% were extremely high on at least two cognitive tests.
impaired on at least two tests, and 22.4% were severely impaired Finally, the average number of impaired test performances at a
on at least two tests. In the ROD group, 61.6% were moderately moderate (1–2 SDs below) or severe (>2 SDs below) level was cal-
impaired on at least two cognitive tests, and 16.2% were severely culated per group (Table 4). One-way ANOVA demonstrated a
impaired on at least two tests. significant difference between the groups in the number of tests
5952 Alexandra Stainton et al.

Table 3. The percentage of each group demonstrating moderate or severe impairment, above average, or extremely high performance on a given number of
cognitive tests

HC ROD CHR ROP χ2 (df) p value

Number of tests moderately impaired (1–2 SDs Below) χ2 = 192.472 (15) <0.001*
0 25.6 19.8 9.1 4.3
1 26.2 18.5 19.6 7.4
2–3 27.5 31.3 31.5 22.2
4–5 12.5 12.7 20.5 21.8
6–8 6.9 9.7 15.1 25.2
9–12 1.7 7.9 4.1 19.1
% Moderately Impaired on Two or More Tests 48.6 61.6 71.2 88.3
Number of Tests Severely Impaired (>2 SDs Below) χ2 = 175.494 (15) <0.001*
0 75.8 66.1 55.7 34.3
1 17.6 17.6 21.9 20.4
2–3 5.8 11 16.4 23.5
4–5 0.9 3.5 5 11.3
6–8 0 1.7 0.5 9.1
9–12 0 0 0.5 1.2
% Severely Impaired on Two or More Tests 6.7 16.2 22.4 45.1
Number of Tests Above Average (1–2 SDs Above) (N = 363) (N = 227) (N = 219) (N = 230) χ2 = 108.104 (15) <0.001*
0 27.5 36.1 42.9 60.0
1 20.1 23.3 21 23.9
2–3 31.4 29.1 26.1 13.9
4–5 15.7 8.3 7.3 1.3
6–8 5.3 2.2 2.8 0.9
9–12 0 0.8 0 0
% Above Average on Two or More Tests 52.3 40.5 36.1 16.1
Number of Tests Extremely High (>2 SDs Above) (N = 363) (N = 227) (N = 219) (N = 230) χ2 = 19.083 (6) 0.004*
0 87.3 86.3 88.6 97.0
1 10.5 11.9 10 3
2–3 2.2 1.8 1.4 0
4–5 0 0 0 0
6–8 0 0 0 0
9–12 0 0 0 0
% Extremely High on Two or More Tests 2.2 1.8 1.4 0
χ2, Chi-squared; *, Statistically significant at a level of p ⩽ 0.05.

moderately (F(3, 1035) = 72.810, p < 0.001), and severely (F(3, group who were demonstrating various levels of performance
1035) = 61.705, p < 0.001) impaired, as well as above average (F from extremely high to severely impaired. Overall, the findings
(3, 1035) = 31.708, p < 0.001) and extremely high (F(3, 1035) = of this study are in line with the previous literature, in that a larger
6.113, p < 0.001). Post-hoc examinations are presented in Table 4. proportion of the clinical groups demonstrated both moderate
and severe cognitive impairment than HC. However, even within
the clinical groups, a subset of participants demonstrated unim-
Discussion
paired cognitive performance at the individual test level, and a
This study investigated the prevalence of cognitive impairments small proportion demonstrated above average performance.
and strengths in a sample of individuals in the early stages of In this sample, cognitive impairment was most prevalent in the
psychosis or depression. While previous studies have largely ROP group. Previous reports have estimated that 70–80% of indi-
examined cognitive impairment in mental illness using group- viduals with chronic schizophrenia (Allen et al., 2003; Palmer
level means, we examined the proportion of individuals in each et al., 1997), and 62% of individuals with first-episode psychosis
Psychological Medicine 5953

Table 4. Average number of tests that each study group is impaired on

HC ROD CHR ROP F p Tamhane Post Hoc CI

Number of tests moderately impaired HC<ROD, p < 0.001* −1.50 to −0.34

HC<CHR, p < 0.001* −1.79 to −0.72
Mean (S.D.) 2.04 (2.11) 2.96 (2.84) 3.29 (2.51) 5.3 (3.22) 72.810 <0.001* HC<ROP, p < 0.001* −3.90 to −2.63
Min-Max 0–10 0–12 0–12 0–12 CHR<ROP, p < 0.001* −2.73 to −1.29
ROD<ROP, p < 0.001* −3.09 to −1.59
Number of tests severely impaired HC<ROD, p = 0.001* −0.65 to −0.11
HC<CHR, p < 0.001* −0.83 to −0.29
Mean (S.D.) 0.35 (0.74) 0.73 (1.4) 0.91 (1.4) 2.04 (2.34) 61.705 <0.001* HC<ROP, p < 0.001* −2.11 to −1.27
Min-Max 0–5 0–8 0–9 0–11 CHR<ROP, p < 0.001* −1.62 to −0.65
ROD<ROP, p < 0.001* −1.79 to −0.84
Number of tests above average HC>ROD, p = 0.022* 0.04–0.85
HC>CHR, p < 0.001* 0.29–1.06
Mean (S.D.) 2.00 (1.90) 1.55 (1.75) 1.32 (1.55) 0.67 (1.06) 31.708 <0.001* HC>ROP, p < 0.001* 1.01–1.65
Min-Max 0–8 0–10 0–7 0–6 CHR>ROP, p < 0.001* 0.32–0.99
ROD>ROP, p < 0.001* 0.53–1.24
Number of tests extremely high HC>ROP, p < 0.001* 0.05–0.18
CHR>ROP, p = 0.003* 0.02–0.17
Mean (S.D.) 0.15 (0.41) 0.15 (0.41) 0.13 (0.37) 0.03 (0.17) 6.113 <0.001* ROD>ROP, p < 0.001* 0.05–0.20
Min-Max 0–2 0–2 0–2 0–1
*, Statistically significant at a level of p ⩽ 0.05; CI, 95% Confidence Intervals for Mean Difference.

(Mwesiga et al., 2022), will demonstrate cognitive impairment. at least two tests, and 22.4% were severely impaired in at least
We observed that 88.3% of the ROP group were moderately two tests. In the composite cognitive score, 14.6% of the CHR
impaired (1–2 SDs below HC) in at least two tests, and 45.1% group were demonstrating an impairment of 1–2 SDs below
were severely impaired (>2 SDs below HC) in at least two tests, HC, whereas 0.9% of the sample were severe (>2 SDs below
which is a slightly higher prevalence than those previous esti- HC). To the best of our knowledge, this is the first study to inves-
mates. When an average composite cognitive score was calculated, tigate the prevalence of cognitive impairment in individuals at
27.8% of the ROP group were demonstrating a global cognitive CHR for psychosis. Nevertheless, these observations align with
impairment of 1–2 SDs below HC, and severe impairment of the previous literature demonstrating that cognitive impairments
>2 SDs below HC in 10.4% of the sample. We also saw a much can begin well before the first-episode of psychosis (Catalan
higher proportion of the present ROP group showing significant, et al., 2021; Fusar-Poli et al., 2012). These findings also suggest
and widespread impairment on multiple tests when compared to that almost three quarters of CHR populations may be experien-
the other study groups. Further, our ROP group demonstrated a cing moderate deficits, and almost one quarter may already be
significantly higher average number of tests impaired at both a experiencing severe deficits and require cognitive interventions
moderate (5.3) and severely (2.04) impaired level than all other by the time they seek help.
clinical groups. Current findings can also be compared with data-driven
We observed that 61.6% of the ROD group were moderately approaches, which have broadly identified either two (Amoretti
(1–2 SDs below HC) impaired in at least two cognitive tests, et al., 2021; Wenzel et al., 2021) or three (Uren et al., 2017) dis-
and 16.2% were severely (>2 SDs below HC) impaired in at tinct clusters of cognitive performance in individuals with ROP
least two cognitive tests. When an average composite cognitive spanning from relatively spared to severely impaired. The rela-
score was calculated, 12.8% of the ROD group were demonstrating tively preserved profile of cognitive performance is often asso-
a moderate cognitive impairment (1–2 SDs below HC), and severe ciated with milder clinical symptoms and better functioning
impairment (>2 SDs below HC) was seen in 1.8% of the sample. (e.g. Crouse, Moustafa, Bogaty, Hickie, & Hermens, 2018;
The prevalence of impairment in at least two tests in this sample Oomen et al., 2021; Uren et al., 2017). Similar findings have
is broadly comparable to the outpatient sample of Douglas et al. emerged for samples with mood disorders (Cotrena, Branco,
(2018). They observed that 78% were moderately impaired, and Ponsoni, Shansis, & Fonseca, 2017), but there are also suggestions
14.3% were severely impaired, according to these criteria. that in depression samples there may be separable clusters based
Furthermore, the composite impairments of 11.8% <1 S.D. and on the type of domains impaired, as opposed to the severity
1.5% <2 SDs in that study are comparable to the present findings. (Hermens et al., 2011; Tickell et al., 2019). Together, previous
Tran et al. (2021) observed a slightly higher prevalence of global and current findings suggest that while cognitive impairments
impairment <1 S.D. in 25.2% of their sample of outpatients with are a core feature of mental illnesses such as psychosis and depres-
MDD. Their sample was slightly older than the current sample sion, there will be a subsample of individuals who are not present-
(mean age 48.6 compared to 26.6 in this ROD group), and ing with such impairments at illness onset. This paper further
there was no restriction to first-episode MDD. Thus, it is possible extends that notion by reporting the prevalence of each group
that their sample may have experienced some further cognitive with above average cognitive performance at baseline.
decline which can be associated with repeated or prolonged When we examined the individual tests showing the largest
depressive episodes (Allott et al., 2016; Semkovska et al., 2019). prevalence of impairment (Table 2, Fig. 1a), we saw largely the
The current CHR group demonstrated a prevalence of cogni- same six tests in all three clinical groups; Backwards Digit
tive impairment which was comparable, but less severe, than Span, Forward Digit Span, Digit-Symbol Substitution Test, Rey
that seen in the ROP group; 71.2% were moderately impaired in Auditory Verbal Learning Test Trials 1–5, Self-Ordered
5954 Alexandra Stainton et al.

Pointing Test Total Errors, and Semantic Verbal Fluency. These performance between individuals. These findings should help us
tests assess working memory, processing speed, verbal learning to refine our approach to cognitive interventions for individuals
and memory and verbal fluency. This observation aligns with pre- with early psychosis or depression. By introducing, or promoting,
vious systematic reviews and meta-analyses, in which deficits in routine cognitive screening at service entry (Bryce & Allott, 2019;
these domains are often seen with the largest effect sizes in psych- Bryce et al., 2021), those individuals who are already demonstrat-
osis (Fioravanti et al., 2012; Schaefer, Giangrande, Weinberger, & ing severe impairment could be efficiently referred to cognitive
Dickinson, 2013), CHR (Fusar-Poli et al., 2012; Pukrop & interventions as required. Such treatment could also be tailored
Klosterkötter, 2010; Zheng et al., 2018), and MDD (Ahern & towards the individual’s specific profile of cognitive performance.
Semkovska, 2017; Goodall et al., 2018). This would suggest that In addition to interventions which take a deficit-reduction
these are core cognitive functions that are impacted across diagno- approach, patients may also benefit from an approach which
ses (East-Richard et al., 2020), and that these domains are particu- aims to highlight their cognitive strengths (Allott et al., 2020)
larly important when screening and providing treatment for and to work on preserving cognitive domains which are intact
cognitive impairment at the first-episode of mental illness. at the first episode to prevent further decline (Pantelis et al.,
To the best of our knowledge, no previous studies have 2015). Here, we used the terms ‘strengths’ and ‘impairments’,
reported the prevalence of cognitive strengths in the early course referring to the degree to which cognitive performance is above
of psychosis or depression. Here, we observed that up to 40.5% of or below average, taking a normative, as opposed to individual,
the ROD group, 36.1% of the CHR group, and 16.1% of the ROP approach (Lezak, Howieson, & Loring, 2004). We cannot report
group were performing at least one S.D. higher than average in at the extent to which an individual’s performance may have
least two cognitive tests. This would indicate that, although cog- declined from their own previous levels (Lezak et al., 2004; Tran
nitive impairments may be pervasive in the early course of mental et al., 2021), which can be associated with poorer outcomes
illness, there may also be areas of relative strength or above aver- even when performance is still within an ‘average’ range
age performance which clinicians could harness to enhance treat- (Raucher-Chéné et al., 2022). Assessment of an individual’s cog-
ment, with a focus on bolstering motivation and functioning nitive performance, including their relative strengths and weak-
(Allott et al., 2020). nesses, would involve a full neuropsychological assessment in
One limitation of this study is that the CHR sample was sig- which the clinician brings together the full picture of their current
nificantly younger than the other study groups, which may have performance, as well as estimated premorbid performance or
had slight impacts on the standardised z-scores calculated based intraindividual assessment, and demographic variables (Lezak
on the HC group. It is possible that the prevalence of cognitive et al., 2004). Further, the terms ‘strength’ and ‘impairment’ may
strengths and impairments in the CHR group may need to be not necessarily represent how the individual perceives their own
interpreted with more caution than the other clinical groups. current cognition relative to previous levels, and thus, targeted
The study groups did also significantly differ on the other demo- treatments for cognition should be planned in consultation with
graphic variables including estimated current IQ and years of the individual to assist with both their objective and subjective
education. However, differences on these factors reflect what is areas of strength and impairment. In addition, this work reports
commonly seen in clinical practice. The purpose of this paper on cognitive performance at baseline only. Cognition may take
was not to determine the cause of cognitive impairment (e.g. a heterogeneous trajectory following the first-episode, with
potentially due to reduced education level in clinical groups), some individuals potentially experiencing further declines in
but rather to provide clinicians with a picture of cognitive per- their cognitive performance (e.g. Fett et al., 2020; Flaaten et al.,
formance which they might expect to see in individuals with 2022). Further work is required to improve our understanding
recent-onset mental illness. Although the individual cognitive of the nature and trajectory of cognition both before and follow-
tests may each have their own norms which are matched on ing the first episode of mental illness. Such work will allow clin-
these demographic variables, these are often different normative icians to maximise the use of time and resources, ensuring that
samples for each test. By creating z-scores based on the present cognition is routinely assessed and treatment tailored to the indi-
HC sample (who were recruited from the same areas as the clin- vidual’s particular needs based on their presentation and stage of
ical groups), we were able to provide a standardised picture of illness.
cognitive performance across all twelve tests. It must also be
acknowledged that at present, there is no universally agreed defin-
ition of a clinically meaningful cognitive impairment. Here, we Conclusion
have provided the exact prevalence of three clinical groups per- By understanding the prevalence of cognitive impairment and
forming at various levels of impairment to compare to previous strength in early course mental illness, we can advocate for cogni-
literature and provide clinicians with a full picture of the patterns tive screening at intake, more accurately resource clinical services,
of performance that they are likely to observe in individuals with and advise clinicians as to the proportion of young people who
recent-onset mental illness. We have also provided the prevalence may need to access additional cognitive treatments or therapies.
of impairments and strengths in a HC group recruited from the
same locations for comparison. This is intended as a starting Supplementary material. The supplementary material for this article can
point for clinical services to estimate the need for cognitive treat- be found at https://doi.org/10.1017/S0033291723001770.
ments. However, cognitive impairment should always be inter-
Acknowledgements. The authors would also like to acknowledge the signifi-
preted within the individual’s wider clinical picture, including
cant contributions of the PRONIA consortium members who performed the
their premorbid functioning, and their own treatment preferences. screening, recruitment, rating, examination, and follow-up of the study parti-
Given the variability in the prevalence of impairment across cipants. They were involved in implementing the examination protocols of the
individual tests, this suggests that while cognitive impairment study, setting up its IT infrastructure, and organising the flow and quality con-
may be common even at the early stages of serious mental illness, trol of the data analysed in this manuscript between the local study sites and
there is likely to be high variability in the profile of cognitive the central study database.
Psychological Medicine 5955

Financial support. PRONIA is a Collaborative Project funded by the University of Bari “Aldo Moro”, Bari, Italy; 12Department of Electronics,
European Union under the 7th Framework Programme under grant agreement Information and Bioengineering, Politecnico di Milano, Milan, Italy;
13
n° 602152 Department of Psychiatry and Psychotherapy, University of Lübeck, Lübeck,
Germany; 14Department of Psychiatry, Psychiatric University Hospital,
Authors’ contributions. CP: Participated on Advisory Boards for University of Basel, Basel, Switzerland; 15Medical Faculty, University of Basel,
Janssen-Cilag, AstraZeneca, Lundbeck, and Servier. Received honoraria for Basel, Switzerland; 16Department of Psychiatry and Psychotherapy, Medical
talks presented at educational meetings organised by AstraZeneca, Faculty, Heinrich-Heine University, Düsseldorf, Germany; 17Department of
Janssen-Cilag, Eli Lilly, Pfizer, Lundbeck, and Shire. NK and RS: Received hon- Psychology, Faculty of Psychology, Airlangga University, Surabaya, Indonesia;
oraria for talks presented at education meetings organised by Otsuka/ 18
University Hospital of Child and Adolescent Psychiatry and Psychotherapy,
Lundbeck. RU: Received grants from the Medical Research Council, University of Bern, Bern, Switzerland; 19Department of Psychiatry, University of
National Institute for Health Research: Health Technology Assessment, Turku, Turku, Finland; 20Melbourne Neuropsychiatry Centre, University of
European Commission – Research: The Seventh Framework Programme, Melbourne and Melbourne Health, Melbourne, VIC, Australia; 21Institute for
and personal speaker fees from Sunovion, outside the submitted work. TL: Translational Psychiatry, University of Münster, Münster¸ Germany;
Supported by the Koeln Fortune Program/Faculty of Medicine, University of 22
Department of Child Adolescent Psychiatry and Psychotherapy, University of
Cologne (Grant No. 370/2020). JW: Partly supported by the NARSAD Münster, Münster, Germany; 23Birmingham Early Intervention Service,
Young Investigator Award of LK through the Brain & Behavior Research Birmingham Women’s and Children NHS Foundation Trust, Birmingham, UK;
Foundation (grant number 28474). KA: Supported by a Dame Kate 24
Department of Psychiatry and Psychotherapy, Ludwig-Maximilian-University,
Campbell Fellowship from the University of Melbourne. Munich, Germany; 25Max-Planck Institute of Psychiatry, Munich, Germany;
26
Institute of Psychiatry, Psychology and Neuroscience, King’s College London,
Ethical standards. The authors assert that all procedures contributing to London, UK and 27School of Psychology, University of Birmingham, Edgbaston,
this work comply with the ethical standards of the relevant national and insti- UK
tutional committees on human experimentation and with the Helsinki
Declaration of 1975, as revised in 2008.

References
*PRONIA Consortium members:
Nikolaos Koutsouleris, Dominic B. Dwyer, Lana Kambeitz-Ilankovic, Anne Abramovitch, A., & Schweiger, A. (2015). Misuse of cognitive neuropsych-
Ruef, Alkomiet Hasan, Claudius Hoff, Ifrah Khanyaree, Aylin Melo, Susanna ology in psychiatry research: The intoxicating appeal of neo-reductionism.
Muckenhuber-Sternbauer, Yanis Köhler, Ömer Öztürk, Nora Penzel, David Behavior Therapist, 38(7), 187–191.
Popovic, Adrian Rangnick, Sebastian von Saldern, Rachele Sanfelici, Moritz Abramovitch, A., Short, T., & Schweiger, A. (2021). The C factor: Cognitive
Spangemacher, Ana Tupac, Maria Fernanda Urquijo, Johanna Weiske, dysfunction as a transdiagnostic dimension in psychopathology. Clinical
Antonia Wosgien, Camilla Krämer, Shalaila S. Haas, Rebekka Lencer, Inga Psychology Review, 86, 102007. doi: https://doi.org/10.1016/j.cpr.2021.
Meyhoefer, Marian Surmann, Udo Dannlowski, Olga Bienek, Georg Romer, 102007
Marlene Rosen, Theresa Lichtenstein, Stephan Ruhrmann, Joseph Kambeitz, Ahern, E., & Semkovska, M. (2017). Cognitive functioning in the first-episode
Karsten Blume, Dominika Julkowski, Nathalie Kaden, Ruth Milz, Alexandra of major depressive disorder: A systematic review and meta-analysis.
Nikolaides, Mauro Silke Vent, Martina Wassen, Christos Pantelis, Stefan Neuropsychology, 31(1), 52–72. doi: https://doi.org/10.1037/neu0000319
Borgwardt, Christina Andreou, André Schmidt, Anita Riecher-Rössler, Laura Allen, D. N., Goldstein, G., & Warnick, E. (2003). A consideration of neurop-
Egloff, Fabienne Harrisberger, Ulrike Heitz, Claudia Lenz, Letizia Leanza, sychologically normal schizophrenia. Journal of the International
Amatya Mackintosh, Renata Smieskova, Erich Studerus, Anna Walter, Sonja Neuropsychological Society, 9(1), 56–63. doi: https://doi.org/10.1017/
Widmayer, Alexandra Korda, Rachel Upthegrove, Chris Day, Sian Lowri S135561770391006X
Griffiths, Mariam Iqbal, Mirabel Pelton, Pavan Mallikarjun, Alexandra Allott, K., Fisher, C. A., Amminger, G. P., Goodall, J., & Hetrick, S. (2016).
Stainton, Ashleigh Lin, Paris Lalousis, Raimo K. R. Salokangas, Alexander Characterizing neurocognitive impairment in young people with major
Denissoff, Anu Ellilä, Tiina From, Markus Heinimaa, Tuula Ilonen, Päivi depression: State, trait, or scar? Brain and Behavior, 6(10), e00527. doi:
Jalo, Heikki Laurikainen, Antti Luutonen, Akseli Mäkela, Janina Paju, Henri https://doi.org/10.1002/brb3.527
Pesonen, Reetta-Liina Säilä, Anna Toivonen, Otto Turtonen, Frauke Allott, K., Steele, P., Boyer, F., de Winter, A., Bryce, S., Alvarez-Jimenez, M., &
Schultze-Lutter, Eva Meisenzahl, Sonja Botterweck, Norman Kluthausen, Phillips, L. (2020). Cognitive strengths-based assessment and intervention
Gerald Antoch, Julian Caspers, Hans-Jörg Wittsack, Ana Beatriz Solana, in first-episode psychosis: A complementary approach to addressing func-
Manuela Abraham, Timo Schirmer, Alessandro Bertolino, Linda tional recovery? Clinical Psychology Review, 79, 101871. doi: https://doi.
A. Antonucci, Giulio Pergola, Ileana Andriola, Barbara Gelao, Paolo org/10.1016/j.cpr.2020.101871
Brambilla, Carlo Altamura, Marika Belleri, Francesca Bottinelli, Adele Ferro, Altman, D. G. (1990). Practical statistics for medical research. London: CRC
Marta Re, Emiliano Monzani, Maurizio Sberna, Armando D’Agostino, press.
Lorenzo Del Fabro, Giampaolo Perna, Maria Nobile, Alessandra Alciati, Amoretti, S., Rabelo-da-Ponte, F. D., Rosa, A. R., Mezquida, G.,
Matteo Balestrieri, Carolina Bonivento, Giuseppe Cabras, Franco Fabbro, Sánchez-Torres, A. M., Fraguas, D., … Pina-Camacho, L. (2021).
Marco Garzitto & Sara Piccin. Cognitive clusters in first-episode psychosis. Schizophrenia Research, 237,
31–39. doi: https://doi.org/10.1016/j.schres.2021.08.021
1 Army Individual Test Battery. (1944). Army Individual Test Battery: Manual of
Orygen, Melbourne, Victoria, Australia; 2Centre for Youth Mental Health,
directions and scoring. In. Washington, DC: War Department, Adjutant
University of Melbourne, Melbourne, Victoria, Australia; 3School of Psychology,
General’s Office.
Aston University, Birmingham, UK; 4Institute for Mental Health and Centre for
Benton, A. L., Hamsher, K. S., & Sivan, A. B. (1994). Multilingual aphasia
Human Brain Health, University of Birmingham, Birmingham, UK; 5Department
examination: Token test. Iowa City: AJA associates.
of Psychiatry and Psychotherapy, Faculty of Medicine and University Hospital of
Bryce, S., & Allott, K. (2019). Cognitive screening: A significant unmet need in
Cologne, Cologne, Germany; 6Faculty of Psychology and Educational Sciences,
youth mental health. Australian & New Zealand Journal of Psychiatry, 53
Department of Psychology, Ludwig-Maximilian University, Munich, Germany;
7 (8), 813–813. doi: https://doi.org/10.1177/0004867419834356
Scientific Institute, IRCCS E. Medea, Pasian di Prato, Udine, Italy; 8Department
Bryce, S., Bowden, S. C., Wood, S. J., & Allott, K. (2021). Brief, performance-
of Neurosciences and Mental Health, Fondazione IRCCS Ca’ Granda Ospedale
based cognitive screening in youth aged 12–25: A systematic review. Journal
Maggiore Policlinico, University of Milan, Milan, Italy; 9Department of
of the International Neuropsychological Society, 27(8), 835–854.
Pathophysiology and Transplantation, University of Milan, Milan, Italy;
10 Catalan, A., De Pablo, G. S., Aymerich, C., Damiani, S., Sordi, V., Radua, J., …
Department of Psychology, Woodbourne Priory Hospital, Birmingham, UK;
11 Stone, W. S. (2021). Neurocognitive functioning in individuals at clinical
Department of Translational Biomedicine and Neuroscience (DiBraiN),
high risk for psychosis: A systematic review and meta-analysis. JAMA
5956 Alexandra Stainton et al.

Psychiatry, 78(8), 859–867. doi: https://doi.org/10.1001/jamapsychiatry. Koutsouleris, N., Davatzikos, C., Bottlender, R., Patschurek-Kliche, K.,
2021.1290 Scheuerecker, J., Decker, P., … Meisenzahl, E. M. (2011). Early recognition
Cornblatt, B. A., Risch, N. J., Faris, G., Friedman, D., & Erlenmeyer-Kimling, and disease prediction in the at-risk mental states for psychosis using neu-
L. (1988). The continuous performance test, identical pairs version rocognitive pattern classification. Schizophrenia Bulletin, 38(6), 1200–1215.
(CPT-IP): I. New findings about sustained attention in normal families. doi: https://doi.org/10.1093/schbul/sbr037
Psychiatry research, 26(2), 223–238. doi: https://doi.org/10.1016/0165-1781 Koutsouleris, N., Kambeitz-Ilankovic, L., Ruhrmann, S., Rosen, M., Ruef, A.,
(88)90076-5 Dwyer, D. B., … the PRONIA Consortium. (2018). Prediction models of
Cotrena, C., Branco, L. D., Ponsoni, A., Shansis, F. M., & Fonseca, R. P. (2017). functional outcomes for individuals in the clinical high-risk state for psych-
Neuropsychological clustering in bipolar and major depressive disorder. osis or with recent-onset depression: A multimodal, multisite machine
Journal of the International Neuropsychological Society, 23(7), 584–593. learning analysis. JAMA Psychiatry, 75(11), 1156–1172. doi: https://doi.
doi: https://doi.org/10.1017/S1355617717000418 org/10.1001/jamapsychiatry.2018.2165
Crouse, J. J., Moustafa, A. A., Bogaty, S. E., Hickie, I. B., & Hermens, D. F. Lee, R. S. C., Hermens, D. F., Redoblado-Hodge, M. A., Naismith, S. L., Porter,
(2018). Parcellating cognitive heterogeneity in early psychosis-spectrum ill- M. A., Kaur, M., … Hickie, I. B. (2013). Neuropsychological and socio-
nesses: A cluster analysis. Schizophrenia Research, 202, 91–98. doi: https:// occupational functioning in young psychiatric outpatients: A longitudinal
doi.org/10.1016/j.schres.2018.06.060 investigation. PLoS ONE, 8(3), e58176. doi: https://doi.org/10.1371/
Douglas, K. M., Gallagher, P., Robinson, L. J., Carter, J. D., McIntosh, V. V., journal.pone.0058176
Frampton, C. M., … Porter, R. J. (2018). Prevalence of cognitive impair- Lezak, M., Howieson, D., & Loring, D. (2004). The rationale of deficit meas-
ment in major depression and bipolar disorder. Bipolar Disorders, 20(3), urement. Neuropsychological assessment, 4, 86–99.
260–274. doi: https://doi.org/10.1111/bdi.12602 Malhi, G. S., Bell, E., Bassett, D., Boyce, P., Bryant, R., Hazell, P., … Porter, R.
Early Psychosis Guidelines Working Group. (2016). Australian clinical guide- (2021). The 2020 royal Australian and New Zealand college of psychiatrists clin-
lines for early psychosis. Orygen The National Centre of Excellence in Youth ical practice guidelines for mood disorders. Australian & New Zealand Journal
Mental Health: Melbourne, Australia. of Psychiatry, 55(1), 7–117. doi: https://doi.org/10.1177/0004867420979353
East-Richard, C., R-Mercier, A., Nadeau, D., & Cellard, C. (2020). Mesholam-Gately, R. I., Giuliano, A. J., Goff, K. P., Faraone, S. V., & Seidman,
Transdiagnostic neurocognitive deficits in psychiatry: A review of L. J. (2009). Neurocognition in first-episode schizophrenia: A meta-analytic
meta-analyses. Canadian Psychology/Psychologie canadienne, 61(3), 190– review. Neuropsychology, 23(3), 315. doi: https://doi.org/10.1037/a0014708
214. doi: https://doi.org/10.1037/cap0000196 Michel, C., Ruhrmann, S., Schimmelmann, B. G., Klosterkötter, J., &
Fett, A.-K. J., Velthorst, E., Reichenberg, A., Ruggero, C. J., Callahan, J. L., Schultze-Lutter, F. (2014). A stratified model for psychosis prediction in
Fochtmann, L. J., … Kotov, R. (2020). Long-term changes in cognitive func- clinical practice. Schizophrenia Bulletin, 40(6), 1533–1542. doi: https://doi.
tioning in individuals with psychotic disorders: Findings from the Suffolk org/10.1093/schbul/sbu025
County mental health project. JAMA Psychiatry, 77(4), 387–396. doi: Mollon, J., David, A. S., Zammit, S., Lewis, G., & Reichenberg, A. (2018).
https://doi.org/10.1001/jamapsychiatry.2019.3993 Course of cognitive development from infancy to early adulthood in the
Fioravanti, M., Bianchi, V., & Cinti, M. E. (2012). Cognitive deficits in schizo- psychosis spectrum. JAMA Psychiatry, 75(3), 270–279. doi: https://doi.org/
phrenia: An updated metanalysis of the scientific evidence. BMC Psychiatry, 10.1001/jamapsychiatry.2017.4327
12(1), 64. Mwesiga, E. K., Robbins, R., Akena, D., Koen, N., Nakku, J., Nakasujja, N., &
Flaaten, C. B., Melle, I., Bjella, T., Engen, M. J., Asbo, G., Wold, K. F., … Stein, D. J. (2022). Prevalence, profile and associations of cognitive impair-
Ueland, T. (2022). Domain-specific cognitive course in schizophrenia: ment in Ugandan first-episode psychosis patients. Schizophrenia Research:
Group- and individual-level changes over 10 years. Schizophrenia Cognition, 28, 100234. doi: https://doi.org/10.1016/j.scog.2021.100234
Research. Cognition, 30, 100263. doi: https://doi.org/10.1016/j.scog.2022. National Institute for Clinical Excellence. (2009). Depression in adults:
100263 Recognition and management. Clinical guideline [CG90]. National
Fusar-Poli, P., Deste, G., Smieskova, R., Barlati, S., Yung, A. R., Howes, O., … Institute for Health and Clinical Excellence. Published October, 28.
Borgwardt, S. (2012). Cognitive functioning in prodromal psychosis: A NICE. (2014). Psychosis and schizophrenia in adults: Prevention and manage-
meta-analysis. Archives of General Psychiatry, 69(6), 10. doi: https://doi. ment. NICE. https://www.nice.org.uk/guidance/cg178.
org/10.1001/archgenpsychiatry.2011.1592 Nowicki, S., & Duke, M. P. (1994). Individual differences in the nonverbal
Goodall, J., Fisher, C., Hetrick, S., Phillips, L. J., Parrish, E. M., & Allott, K. communication of affect: The diagnostic analysis of nonverbal accuracy
(2018). Neurocognitive functioning in depressed young people: A system- scale. Journal of Nonverbal Behavior, 18(1), 27. doi: https://doi.org/10.
atic review and meta-analysis. Neuropsychology Review, 28, 216–231. doi: 1007/BF02169077
https://doi.org/10.1007/s11065-018-9373-9 Oomen, P. P., Begemann, M. J., Brand, B. A., de Haan, L., Veling, W., Koops,
Gould, I. C., Shepherd, A. M., Laurens, K. R., Cairns, M. J., Carr, V. J., & S., … van Beveren, N. (2021). Longitudinal clinical and functional outcome
Green, M. J. (2014). Multivariate neuroanatomical classification of cognitive in distinct cognitive subgroups of first-episode psychosis: A cluster analysis.
subtypes in schizophrenia: A support vector machine learning approach. Psychological Medicine, 53(6), 2317–2327. doi: https://doi.org/10.1017/
NeuroImage: Clinical, 6, 229–236. doi: https://doi.org/10.1016/j.nicl.2014. S0033291721004153.
09.009 Osterrieth, P. A. (1944). Le test du copie d’une figure complexe. Archives de
Gualtieri, C. T., & Morgan, D. W. (2008). The frequency of cognitive impair- psychologie, 28, 206–356.
ment in patients with anxiety, depression, and bipolar disorder: An Palmer, B. W., Heaton, R. K., Paulsen, J. S., Kuck, J., Braff, D., Harris, M. J., …
unaccounted source of variance in clinical trials. Journal of Clinical Jeste, D. V. (1997). Is it possible to be schizophrenic yet neuropsychologi-
Psychiatry, 69(7), 1122–1130. cally normal? Neuropsychology, 11(3), 437. doi: https://doi.org/10.1037/
Hermens, D. F., Redoblado Hodge, M. A., Naismith, S. L., Kaur, M., Scott, E., 0894-4105.11.3.437
& Hickie, I. B. (2011). Neuropsychological clustering highlights cognitive Pantelis, C., Wannan, C., Bartholomeusz, C. F., Allott, K., & McGorry, P. D.
differences in young people presenting with depressive symptoms. Journal (2015). Cognitive intervention in early psychosis – preserving abilities ver-
of the International Neuropsychological Society, 17(2), 267–276. doi: sus remediating deficits. Current Opinion in Behavioral Sciences, 4, 63–72.
https://doi.org/10.1017/S1355617710001566 doi: https://doi.org/10.1016/j.cobeha.2015.02.008
IBM Corp. (2017). IBM SPSS Statistics for Windows, Version 25.0. Armonk, Petrides, M., & Milner, B. (1982). Deficits on subject-ordered tasks after
NY: IBM Corp,. frontal-and temporal-lobe lesions in man. Neuropsychologia, 20(3), 249–
Keefe, R. S. E. (2008). Should cognitive impairment be included in the diag- 262. doi: https://doi.org/10.1016/0028-3932(82)90100-2
nostic criteria for schizophrenia? World Psychiatry, 7(1), 22–28. Pukrop, R., & Klosterkötter, J. (2010). Neurocognitive indicators of clinical
Keefe, R. S. E., Eesley, C. E., & Poe, M. P. (2005). Defining a cognitive function high-risk states for psychosis: A critical review of the evidence.
decrement in schizophrenia. Biological Psychiatry, 57(6), 688–691. doi: Neurotoxicity Research, 18(3-4), 272–286. doi: https://doi.org/10.1007/
https://doi.org/10.1016/j.biopsych.2005.01.003 s12640-010-9191-1
Psychological Medicine 5957

Pukrop, R., Schultze-Lutter, F., Ruhrmann, S., Brockhaus-Dumke, A., Tran, T., Milanovic, M., Holshausen, K., & Bowie, C. R. (2021). What is nor-
Tendolkar, I., Bechdolf, A., … Klosterkötter, J. (2006). Neurocognitive func- mal cognition in depression? Prevalence and functional correlates of nor-
tioning in subjects at risk for a first episode of psychosis compared with mative versus idiographic cognitive impairment. Neuropsychology, 35(1),
first-and multiple-episode schizophrenia. Journal of Clinical and 33. doi: https://doi.org/10.1037/neu0000717
Experimental Neuropsychology, 28(8), 1388–1407. Uren, J., Cotton, S. M., Killackey, E., Saling, M. M., & Allott, K. (2017).
Raucher-Chéné, D., Pochiet, G., Lavigne, K. M., Heinrichs, R. W., Malla, A., Cognitive clusters in first-episode psychosis: Overlap with healthy controls
Joober, R., & Lepage, M. (2022). Normal-range verbal memory in the and relationship to concurrent and prospective symptoms and functioning.
first-episode of psychosis: Clinical and functional outcomes across 24 months Neuropsychology, 31(7), 787. doi: https://doi.org/10.1037/neu0000367
and impact of estimated verbal memory decrement. Schizophrenia Research, Wechsler, D. (1939). The measurement and appraisal of adult intelligence (1st
246, 75–84. doi: https://doi.org/10.1016/j.schres.2022.06.007 ed.). Baltimore: Williams and Wilkins Corporation.
Rey, A. (1941). L’examen psychologique dans les cas d’encéphalopathie Wechsler, D. (1981). Wechsler Adult Intelligence Scale-Revised (WAIS-R) In.
traumatique.(Les problems.). Archives de psychologie, 28, 215–285. San Antonio, TX: Psychological Cooperation.
Rey, A. (1964). L’examen clinique en psychologie [The clinical psychological Wechsler, D. (2011). Wechsler abbreviated scale of intelligence (2nd ed.)
examination]. Presses Universitaires de France: Paris. (WASI-II). NCS Pearson.
Schaefer, J. D., Giangrande, E., Weinberger, D. R., & Dickinson, D. (2013). The Wenzel, J., Haas, S. S., Dwyer, D. B., Ruef, A., Oeztuerk, O. F., Antonucci, L.
global cognitive impairment in schizophrenia: Consistent over decades and A., … the, P. c. (2021). Cognitive subtypes in recent onset psychosis:
around the world. Schizophrenia Research, 150(1), 42–50. doi: https://doi. Distinct neurobiological fingerprints? Neuropsychopharmacology, 46(8),
org/10.1016/j.schres.2013.07.009 1475–1483. doi: https://doi.org/10.1038/s41386-021-00963-1
Semkovska, M., Quinlivan, L., O’Grady, T., Johnson, R., Collins, A., O’Connor, Wilk, C. M., Gold, J. M., McMahon, R. P., Humber, K., Iannone, V. N., &
J., … Gload, T. (2019). Cognitive function following a major depressive epi- Buchanan, R. W. (2005). No, it is not possible to be schizophrenic yet neu-
sode: A systematic review and meta-analysis. The Lancet Psychiatry, 6(10), ropsychologically normal. Neuropsychology, 19(6), 778. doi: https://doi.org/
851–861. doi: https://doi.org/10.1016/S2215-0366(19)30291-3 10.1037/0894-4105.19.6.778
Tickell, A. M., Scott, E. M., Davenport, T., Iorfino, F., Ospina-Pinillos, L., Harel, Zheng, W., Zhang, Q.-E., Cai, D.-B., Ng, C. H., Ungvari, G. S., Ning, Y.-P., &
K., … Hermens, D. F. (2019). Neurocognitive clusters: A pilot study of young Xiang, Y.-T. (2018). Neurocognitive dysfunction in subjects at clinical high
people with affective disorders in an inpatient facility. Journal of Affective risk for psychosis: A meta-analysis. Journal of Psychiatric Research, 103,
Disorders, 242, 80–86. doi: https://doi.org/10.1016/j.jad.2018.08.045 38–45. doi: https://doi.org/10.1016/j.jpsychires.2018.05.001
Copyright © The Author(s), 2023. Published by Cambridge University Press.
This work is licensed under the Creative Commons Attribution License This
is an Open Access article, distributed under the terms of the Creative
Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/),
which permits unrestricted re-use, distribution and reproduction, provided the
original article is properly cited. (the “License”). Notwithstanding the
ProQuest Terms and Conditions, you may use this content in accordance
with the terms of the License.