SEARCH

SEARCH BY CITATION

Keywords:

  • psychosis;
  • schizophrenia;
  • clinical high-risk;
  • ultra-high risk;
  • prodrome;
  • relatives;
  • familial;
  • cognition;
  • neuropsychology

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Objective

It is likely that cognitive deficits are vulnerability markers for developing schizophrenia, as these deficits are already well-established findings in first-episode psychosis. Studies at-risk adolescents and young adults are likely to provide information about cognitive deficits that predate the onset of the illness.

Method

We conducted meta-analyses of studies comparing familial-high risk (FHR) or ultra-high risk (UHR; n = 2113) and healthy controls (= 1748) in youth studies in which the mean age was between 15 and 29.

Results

Compared with controls, high risk subjects were impaired in each domain in both UHR (d = 0.34–0.71) and FHR (d = 0.24–0.81). Heterogeneity of effect sizes across studies was modest, increasing confidence to the findings of the current meta-analysis (I2 = 0–0.18%). In both risk paradigms, co-occurrence of genetic risk with attenuated symptoms was associated with more severe cognitive dysfunction. In UHR, later transition to psychosis was associated with more severe cognitive deficits in all domains (d = 0.31–0.49) except sustained attention. However, cognitive impairment has a limited capacity to predict the outcome of high-risk patients.

Conclusion

Cognitive deficits are already evident in adolescents and young adults who have familial or clinical risk for psychosis. Longitudinal developmental studies are important to reveal timing and trajectory of emergence of such deficits.

Summations
  • Both familial and clinical high-risk for psychosis is associated with cognitive deficits.
  • Co-occurrence of both risk factors is associated with more severe cognitive dysfunction.
  • Youth who develops psychotic disorders at follow-up have more severe cognitive dysfunction at baseline assessment.
Considerations
  • More studies are needed to directly compare cognitive functioning in genetic and clinical high-risk risk subjects.
  • Effect of the diagnosis on baseline cognitive deficits at individuals who developed psychotic disorders at follow-up needs further clarification.
  • There is a significant overlap of baseline cognitive performances of clinical high risk individuals who did or did not develop psychosis, despite the fact that first group is significantly more impaired. There is no sufficient data regarding the effect of baseline cognition on transition to psychosis in familial high risk individuals.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Cognitive dysfunction is one of the core features of schizophrenia and is an important predictor of occupational and functional impairments observed in people with this disorder [1-5]. Reduced cognitive performance is already evident at the first-episode of psychosis and is a stable characteristic of the illness [6-8]. This implies that cognitive dysfunction may be a likely neurobiological marker of psychosis before the onset of the illness.

The most important paradigms to investigate cognitive deficits before the onset of the illness are the genetic and clinical high-risk approaches. The genetic high risk paradigm has traditionally been used to study neurobiological alterations prior to onset of schizophrenia [9, 10]. There are a substantial number of studies that examined cognitive functions in the offspring and siblings of patients with schizophrenia in childhood, adolescence and adulthood [11-14]. Over the last 15 years, a clinical-high risk paradigm targeting identification in the prodrome of psychosis, known as ultra-high-risk (UHR), clinical high risk, at-risk mental state or psychosis risk syndrome, has successfully been used to study individuals who are at substantial risk for developing a psychotic disorder in the near future [15-17].

There is evidence of cognitive impairment of both individuals with UHR and young people with family history of schizophrenia (FHR); however, there are inconsistent findings regarding specific abnormalities associated with clinical and familial high risk to psychosis. Two recent meta-analyses examining cognitive deficits in UHR subjects in comparison to healthy controls found evidence of cognitive impairment in adolescents and youth with UHR [18, 19], but no meta-analysis has examined cognitive deficits in FHR at this age period, or compared this group to UHR populations.

One important reason for inconsistent results is the small sample size of most studies. Another is the fact that both of these risk paradigms will recruit a substantial proportion of ‘false positives’ that have no real susceptibility to develop psychotic disorders. These young people may never be destined to develop frank psychotic disorder. Combining familial and clinical risk could be important, but to date it is not clear whether such an approach is associated with more severe cognitive deficits. It is also unclear whether cognitive deficits can contribute to prediction of individuals who will develop frank psychosis among high-risk subjects. Meta-analytic methods can be beneficial to reveal the most consistent cognitive deficits in FHR and UHR subjects and potential predictors of transition to psychosis.

Aims of the study

Our goals were to conduct a meta-analysis of cognitive deficits in familial-high risk and ultra-high risk to i) examine consistent cognitive deficits in youth with familial-high risk and ultra-high risk; ii) examine effect of co-occurrence of genetic and clinical risk on cognitive deficits; iii) examine whether baseline cognitive deficits in both high risk groups are associated with transition to psychosis.

Material and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Study selection

We followed PRISMA 2009 guidelines in conducting this meta-analysis [20]. A literature search was conducted using the databases Pubmed, PsycINFO and Scopus to identify the relevant studies (January 1990 to April 2013). We used the combination of the following keywords: high risk, familial high risk, relatives, clinical high risk, ultra-high risk, prodrome, at-risk mental state, psychosis, schizophrenia, cog*, neurops*, memory, attention and executive function. We also reviewed the reference lists of published studies. Inclusion criteria were studies that i) reported neurocognitive data or statistics sufficient to calculate effect sizes (where necessary, authors were contacted to provide such data); ii) written in English; iii) compared performance of a help-seeking clinical-high risk group (UHR) and/or unaffected relatives of people with schizophrenia (FHR) with healthy controls; or compared transitioned and non-transitioned at-risk subjects, and each group included at least five subjects; iv) had a mean high-risk sample age between 15 and 29 to examine FHR and UHR subjects in the same developmental stage. All available UHR studies met this fourth criterion, and FHR studies included were selected from larger number of studies with a wider age range. FHR was defined as having a parent or sibling with schizophrenia, or at least two relatives with schizophrenia. UHR was defined as having one or more of three psychosis risk syndromes in help-seeking youth or young adults. Although there are various tools used to assess these, (e.g. CAARMS, SIPS), essentially these three risk syndromes included i) recent onset or worsening of attenuated positive symptoms (APS); ii) recent onset of frank level psychotic symptoms which were significant but not sufficiently sustained to meet the criteria for DSM-IV psychotic disorder (brief limited intermittent psychotic syndrome: BLIPS); iii) genetic/familial risk to psychosis plus deterioration in functioning (recent onset or worsening functional decline, or chronic poor functioning) syndrome (GRDS). Some studies used an alternative basic symptom approach by combining three syndrome criteria above (late prodromal state) with other clinical features that was hypothesized to predate them (early prodromal state). In these studies, only late prodromal state data was included if it was reported separately. Studies comparing high risk samples with help-seeking controls, recruited non-help-seeking clinical high risk groups based on screening, and those which defined risk syndrome based on psychometric risk or schizotypal personality were excluded. For each cognitive measure, only the study with largest sample size was included from each group. Studies that were included in reports of multisite studies were not included in the meta-analysis unless they reported a different cognitive measure. A flow chart of study selection is shown in Fig. 1. Forty-four Studies [21-64] including two multisite studies included in the study (Tables 1 and 2).

Table 1. Study characteristics of FHR studies included
StudiesSampleAgeCharacteristicsCognitive tasks
  1. HC, Healthy controls; FHR, Familial high risk to psychosis; WMS, Wechsler memory scale; CPT, continuous Performance test; CVLT, California verbal learning test; LNS, Letter number sequencing; TMT, trail making test; RCFT, Rey complex figure test; WCST, Wisconsin cart sorting test; HSCT, Hayling sentence completion test; RAVLT, Rey Auditory verbal learning test; IQ, intelligence coefficient; EHRS, Edinburgh High Risk Study; NAPLS, North American Prodrome Longitudinal Study.

Barcelona
Barrantes-Vidal et al. 2007 [20]

38 FHR

63 HC

FHR = 27.3

Siblings

No personality disorder

clusters A or B

WAIS-III, CPT, WMS-R, WCST, letter and category fluency
EHRS
Bryne et al. [27]

157 FHR

34 HC

FHR = 21.2With at least 2 family member schWAIS-R, NART, Stroop, letters/Category fluency, HSCT, digit symbol, Digit span, RAVLT, CPT, WMS-R visual
HHAHRS
Seidman et al. [54]

73 FHR

84 HC

FHR = 18.1 WRAT reading, vocabulary, BD, TMT, CPT, Digit symbol, Digit span, Story recall, letter fluency
Israel
Hans et al. [35]

39 FR

36 C

FHR = 16.9

Offspring

Schizotypy

correlated

TMT, CPT, Stroop, WCST, digit

Span

Jena-Germany
Klemm et al. [40]

32 FHR

32 HC

FHR = 16.0

Siblings

Offsprings

WCST, TMT, Stroop, d2

Concentration test, SPM

London
Hughes et al. [36]

25 FHR

25 HC

FHR = 28

Siblings

SPD excluded

IQ, CPT, LNS, WMS-R, Visual memory, RAVLT, WCST, TMT, Stroop, letter, category fluency
NAPLS
Seidman et al. [55]

49 FHR

109 HC

FHR = 18.7

Offsprings, siblings

No prodrome

 
Nottingham
Groom et al. [34]

36 FHR

72 HC

FHR = 17.5Siblings No prodromal symWASI, RAVLT,CPT, HSCT, Letter fluency
NYHRP
Wolf et al. [61]

73 FHR

120 HC

FHR = 25.6OffspringWSCT
Palau
Myles Worsley et al. [45]

99 FHR

99 HC

FHR = 16.9

Offspring or 2 siblings +

55/99 mild symptoms

WMS, digit symbol, digit span, spatial

Span, LNS, CPT

Seoul
Nam et al. [46]

44 FHR

100 HC

FHR = 28.8SiblingsDSST, letter/category fluency, Digit span, N-back, TMT, CPT, RAVLT, RCFT
Pitsburgh
Prasad et al. [48]

74 FHR

86 HC

 

Offsprings

Schizotypy

Correlated

IQ, WCST, CPT
Quebec
Maziade et al. [44]

22 FHR

45 HC

FHR = 17.3OffspringsIQ, CPT, CVLT
Seoul Youth Clinic
Choi et al. [29]

23 FHR

34 HC

FHR = 23.42 relativesIQ
St Louis
Delawalla et al. [30]

31 FHR

42 HC

FHR = 21.3SiblingsVocabulary, CPT
Swedish
Schubert and Mc Neil [52, 53]

38 FHR

88 HC

FHR = 22.4

Offsprings

8/74 has cluster A personality

WCST, Word-pair, verbal fluency, TMT, Digit span, reaction time, BD, Selective attention
Washington University
Bonner-Jackson et al. [24]

24 FHR

40 HC

FHR = 21.1SiblingsLogical memory, family pictures, CVLT
New York
Bertisch et al. [22]

16 FHR

31 HC

 One first-degree or Multiple second degree 6/16 prodromal symptomsVocabulary, CVLT, digit span, LNS, Stroop, Visual memory, word pairs, Matrix reasoning
Table 2. Study characteristics of UHR studies included
StudiesSampleAgeCharacteristicsCognitive tasks
  1. HC, Healthy controls; UHR, Ultra-high risk to psychosis; COPE, Center of prevention and evaluation; NAPLS, North American Prodrome Longitudinal Study; RAP, Recognition and prevention programme; PRIME, Prevention through risk management and education; OASIS, Outreach and Support in South London service; WMS, Wechsler memory scale; CPT, continous Performance test; CVLT, California verbal learning test; LNS, Letter number sequencing; TMT, trail making test; RCFT, Rey complex figure test; WCST, Wisconsin cart sorting test; HSCT, Hayling sentence completion test; RAVLT, Rey Auditory verbal learning test; HVLT, Hopkins verbal learning test; CANTAB, Cambridge Neuropsychological Test Automated Battery; WM, Working memory; SWM, Spatial working memory; SOPT, Self-ordered pointing test; ToH, Tower of London test; D-KEFS, Delis-Kaplan Executive Function System; DRT, Delayed response task; IQ, intelligence coefficient.

PACE

Melbourne

Lin et al. 2013 [42]

325 UHR

66 HC

UHR = 19.1

244 UHR-NP

81 UHR-P

IQ, premorbid IQ, WMS-R, RAVLT

Trails A &B, Stroop

Letter fluency

Additional data below

Francey et al. [31]

70 UHR

51 HC

UHR = 20.1

20 UHR-P

50 UHR-NP

CPT
Wood et al. 2003 [62]

16 UHR

17 HC

UHR = 18.8

1–2 year follow-up

7-UHR-P

9 UHR-NP

CANTAB

COPE

Columbia

Stanford et al. 2011 [58]

63 CHR

24 HC

UHR = 19.6Cross-sectionalWMS, WAIS

NAPLS

Multisite Emory, Harvard, UCSD UCL, Toronto, Yale, North Carolina, Zucker-Hillside

304 UHR

193 HC

UHR = 18.2 WCST, CPT, Premorbid IQ, Symbol coding, letter fluency, Verbal memory, symbol coding, IQ
Seidman et al. [55]

167 UHR

109 HC

UHR = 18.2

2.5 year follow-up

54 UHR-P

113 UHR-NP

WCST, CPT, Premorbid IQ, Symbol coding, letter fluency, Verbal memory, BD, IQ

RAP

Zucker-Hillside

NAPLS  

Data included in NAPLS

Not overlapping data (below)

Smith et al. 2006 [57]

8 UHR

10 HC

UHR = 16.3BaselineSWM
Carrion et al. 2011 [28]

127 UHR

80 HC

UHR = 16.0BaselineVerbal WM (Digit span, LNS)

PRIME

UNC, Toronto

NAPLS LongitudinalData included in NAPLS
UCL and Emory NAPLS  Data included in NAPLS
UCSD NAPLS  

Data included in NAPLS

Not overlapping data (below)

Jahshan et al. 2010 [37]

48 UHR

29 HC

UHR = 18.7LongitudinalSpatial Span, Stroop, HVLT, LNS

GRNS

Bonn, Cologne Dusseldorf, Munich

Frommann et al. [32]

89 LPS

87 HC

LPS = 25.3Cross-sectionalAVLT, SOPT, TMT A &B, CPT, symbol coding, LNS, Letter fluency

FETZ

Cologne

GRNS  

Data included in GRNS

Not overlapping data (below)

Pukrop et al. 2006 [49]

90 UHR

174

UHR = 24.6 ROCF, DRT, WCST
Pukrop et al. 2007 [50]

83 CHR

44 HC

UHR = 24

44 UHR-P

39 UHR-NP

UHR-P vs. UHR-NP

CPT, SOPT, ROCFT, AVLT,

Letter/Number, WCST, SC

Trail A B, Fluency, DRT

Munich GRNS  

Data included in GRNS

Not overlapping data (below)

Koutselaris et al. [41]

48 UHR

30 HC

UHR = 24.7

20 EPS/28 LPS

15 UHR-P/

20 UHR-NP

UHR-P vs. UHR-NP:

Digit span, SOPT, AVLT,

Letter/Number, symbol coding

Trail A B, Fluency, DRT

UHR vs. HC

Digit span

FEPSY
Pflueger et al. 2007 [47]

54 CHR

51 HC

UHR = 2754 HR, 6 low riskIQ, ToH, WCST, Go/No-Go, WM-visual 2 back, CPT
Amsterdam
Becker et al. [21]

47 UHR

42 UHR

UHR = 20.7

18 UHR-P, 2 year

69 FE

Letter and category fluency
Becker et al. [22]

41 UHR

17 HC

UHR = 19.2

17 UHR-P

24 UHR-NP

CVLT, Letter and category fluency, CPT, SWMT, ROCF,Premorbid IQ

PIER

Portland

Woodberry et al. 2010 [63]

73 HR

34 HC

UHR = 16.5

13 UHR-P

55 UHR-NP

WCST, LM, CVLT, WAIS, Fluency, Trail making test, WRAT, CPT-IP

Seoul 1

Seoul Youth Clinic

Kim et al. [38]

49 UHR

45 HC

UHR = 21.1

2.8 year mean follow-up

13 UHR-P

Trail A, Stroop, CVLT, ROCF

WCST, letter fluency, DS, Spatial location,

Seoul 2

Yonsei University

Kim et al. [39]

33 UHR

27 HC

 17% transit over 5–17 monthsVerbal and visual memory
Helsinki
Lingdren et al. [43]

62 UHR

72 HC

UHR = 16.6Cross-sectionalLetter, semantic fluency, TMT-A/B, list learning, Vis and digit span, reaction time, Dot cancellation

OASIS

London

Broome 2007 [25]

35 UHR

23 HR

UHR = 24.9Cross-sectionalPremorbid IQ, Beads task, IQ?
Broome 2012 [26]28 UHRUHR = 24.4

2 years follow-up

5 UHR-P

23 UHR-NP

WM, IQ, premorbid IQ
Buones Aires
Serrani 2011 [56]

27 UHR

38 HC

UHR = 17.4Cross-sectionalIQ, Premorbid IQ, category fluency, CPT, symbol coding, HVLT, LNS, TMT A, spatial span. Mazes, Visual memory
Utrecht
Van Rijn et al. [60]

36 UHR

21 HC

UHR = 15.2Cross-sectionalIQ, shifting set task
Budapest
Szily et al. 2009 [59]

26 UHR

50 HC

UHR = 22.0Cross-sectionalIQ
image

Figure 1. Flow diagram for meta-analysis of cognitive deficits in UHR and FHR. From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097.

Download figure to PowerPoint

FHR or UHR vs. Controls

The FHR and healthy control comparison included 18 main studies (20 samples) including one multi-site study. About 929 FHR (178–757 FHR for each variable) subjects (48.3% males) and 1140 (126–881 for each variable) controls (46.6%) males were included. There was no significant between-group difference for age (d = 0.01, CI = −0.11 to 0.14, Z = 0.23, = 0.82).

Sixteen main UHR studies including two multisite studies consisting of 1184 UHR (351–907 UHR for each variable) subjects (53.8% males) and 717 (311–557 for each variable) controls (50.1% males) were included. There was no significant difference for age (d = 0.05, CI = −0.12 to 0.21, Z = 0.54, = 0.59).

Co-occurrence of genetic and clinical risk

Meta-regression analyses were used to examine effect of the percentage of cases with genetic risk (11 studies) in UHR samples. 20.3% of UHR subjects met the genetic risk criteria. A vast majority of UHR subjects with genetic risk met criteria for GRDS+APS rather than pure GRDS. In FHR, a preliminary meta-analysis conducted for three studies which provided separate data for genetic at-risk subjects with (= 116) and without (= 159) symptoms. This analysis in FHR was only conducted for verbal memory and processing speed as they were the only cognitive domains reported in all three studies.

Baseline cognition and prediction of psychosis

Comparison of UHR subjects who developed psychotic disorder (UHR-P) and did not (UHR-NP) were based on data from nine samples (263 UHR-P and 586 UHR-NP, 11 reports) including up to 229 UHR-P and 518 UHR-NP for each of the variable. There was no significant difference for gender (60.9% vs. 52.1% male, RR = 1.10, CI = 0.98–1.24, = 0.12) and age (= 0.06, CI = −0.11 to 0.22, = 0.50) between UHR-P and UHR-NP. Only one study [the Edinburgh high risk study (EHRS)] compared FHR subjects who developed illness with others, and therefore meta-analysis was not possible for this group.

Cognitive measures

We combined individual tasks into broader cognitive domains of current and premorbid IQ, verbal memory, visual memory, executive functions, fluency, sustained attention, verbal and visuospatial working memory. This step was undertaken because there were not sufficient studies to perform meta-analyses for all individual tasks (See Appendix). In addition to cognitive domain analyses, task-specific analyses were conducted when at least three independent studies had employed a given task. Individual tasks that were analyzed separately included list learning and delayed recall, the Wisconsin cart sorting test (WCST) perseveration errors, Continuous performance test (CPT) sensitivity (d) score, Trail Making Task A and B, digit span, symbol coding, letter number sequencing and letter (phonemic) fluency.

Statistical analyses

Meta-analyses were performed using mix software version 1.7 on a Windows platform [65]. For each cognitive task, an effect size and standard error was estimated. Effect sizes were weighted using the inverse variance method, and a random effects model (DerSimonian–Laird estimate) was used because the distributions of effect sizes were heterogeneous for a number of variables. For studies that reported more than one cognitive task for each domain, a pooled effect size was calculated. The Q test was used to measure the ‘heterogeneity of the distribution of effect sizes’ (the dispersion of the effect sizes around their mean is greater than expected from sampling error alone). When the Q test was significant, ‘I2′ was used to quantify heterogeneity [66]. I2 estimates the percentage of total variation across studies due to heterogeneity rather than chance. I2 values between 0 and 0.25 indicate small magnitudes of heterogeneity, I2 values in the range 0.25–0.50 indicate medium magnitudes and I2 > 0.50 indicate large magnitudes. Publication bias was assessed by Egger's test that rely on the theory that small studies with significant rather than negative findings in studies with small sample size would be likely to be reported, while large-scale studies would be more likely to be published regardless of significance of findings. We also calculated homogeneity statistics using Qbet to test the differences between FHR and UHR.

Meta-regression analyses in UHR were conducted for age, gender, transition rate, duration of follow-up, the ratio of subjects with genetic risk and Global assessment of functioning (GAF) score. Meta-regression analyses in FHR were conducted for age and gender. Meta-regression analyses (weighted generalized least squares regressions) were conducted using SPSS version 11.0 (SPSS Inc., Chicago, IL, USA). Meta-regression analyses were performed with a random effects model and conducted using the restricted-information maximum likelihood method with a significance level set at < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Cognitive deficits in FHR and UHR

FHR vs. HC

Compared with controls, FHR subjects were impaired in every domain (= 0.24–0.81; Fig. 2) and on all individual tasks (= 0.25–0.51; Table 3). There was significant heterogeneity in some of the cognitive variables, but the magnitude of heterogeneity was small for each of these variables (range of I2 = 0–0.09). There was no evidence of publication bias in any of the analyses. In meta-regression analyses, age and gender had no effect on observed between-group differences.

Table 3. Mean weighted effect sizes for cognitive differences between UHR, FHR and healthy controls
TestSampleHRHC D 95% CI Z P Q-test P I 2 Bias
  1. D, Effect size of between-group difference; Bias, P value of the Egger's test; UHR, Ultra-high risk to psychosis; FHR, Familial high risk to psychossi; TMT, Trail making test; WM, Working memory; EF, executive functions; CPT, Continous performance test; LNS, Letter Number sequencing; CI, Confidence interval; HC, Healthy controls; HR, High risk; WCST per, Wisconsin cart sorting test perseverative.

Premorbid IQ
UHR9 (2)8734970.300.13–0.483.35<0.0010.020.040.13
FHR63704000.630.47–0.797.65<0.0010.600
Current IQ
UHR12 (2)9075330.400.25–0.545.48<0.0010.150.020.36
FHR84274730.810.61–1.018.06<0.0010.070.04
Processing speed
UHR85634110.470.27–0.664.72<0.0010.040.040.55
FHR13 (11)6868080.350.22–0.495.07<0.0010.130.02
TMT A
UHR5 (1)3442790.410.17–0.663.30.00090.080.040.88
FHR41873040.270.04–0.5023.40.020.230.02
Symbol coding
UHR4 (2)4172710.700.41–0.994.76<0.0010.050.050.41
FHR6 (5)4204260.370.23–0.515.20<0.0010.460
Stroop
UHR31951110.38–0.09–0.851.590.110.020.130.78
FHR42271260.280.04–0.522.330.020.640
TMT B
UHR6 (2)5563840.410.16–0.673.150.0020.0050.070.63
FHR8 (7)5665030.380.17–0.603.46<0.0010.030.05
Verbal memory
UHR107104950.500.32–0.685.39<0.0010.030.040.35
FHR12 (11)7577900.450.29–0.615.65<0.0010.060.03
Learning
UHR6 (1)3962420.680.34–1.013.97<0.0010.0030.120.21
FHR52562070.510.30–0.724.70<0.0010.970
Visual memory
UHR85104450.500.23–0.773.670.00020.0010.100.37
FHR8 (7)5824030.510.30–0.724.79<0.0010.080.04
EF
UHR87435570.380.19–0.563.98<0.0010.030.040.24
FHR11 (10)6056720.240.06–0.422.630.0090.020.04
WCST per
UHR54714180.310.08–0.522.630.0080.040.040.63
FHR83305590.250.06–0.452.530.010.050.04
Verbal WM
UHR96644720.410.20–0.613.89<0.0010.0070.060.31
FHR10 (9)6056010.320.12–0.513.230.0010.020.05
LNS
UHR42371880.540.10–0.992.400.020.0030.160.28
FHR5 (4)1782180.500.22–0.773.53<0.0010.100.05
Digit span
UHR42741830.210.01–0.412.070.040.8000.45
FHR7 (6)3023970.250.04–0.452.380.020.120.03
VisSpa WM
UHR94024000.710.39–1.044.28<0.001<0.0010.180.06
FHR42192070.350.01–0.711.970.040.020.09
Attention
UHR8 (2)5834590.370.25–0.505.78<0.0010.5900.35
FHR145708810.300.16–0.444.29<0.0010.080.03
CPT
UHR7 (2)5213870.400.26–0.535.60<0.0010.5300.28
FHR12 (11)5307610.300.15–0.463.84<0.0010.040.04
Fluency
UHR85284020.520.30–0.744.65<0.0010.010.060.53
FHR104986510.390.16–0.613.400.0010.0020.08
Letter fluency
UHR5 (2)4642950.450.20–0.703.550.00040.060.040.91
FHR94705630.450.31–0.596.34<0.0010.400
image

Figure 2. Cognitive deficits in UHR and FHR. PS, Processing speed; VerMem, Verbal memory; VisMem, Visual memory; Ver WM, Verbal working memory; EF, executive functions; FHR, Familial high risk; UHR, Ultra-high-risk to psychosis.

Download figure to PowerPoint

UHR vs. HC

Compared with controls, UHR subjects were impaired in every domain (= 0.34–0.71; Fig. 2) and on all individual tasks (= 0.21–0.70), except the Stroop interference score (= 0.38, = 0.11; Table 3). There was significant heterogeneity in some of the cognitive variables, but the magnitude of heterogeneity was small for each of these variables (range of I2 = 0–0.23). There was no evidence of publication bias in any of the analyses. The most robust impairments in the UHR subjects compared with controls were observed for symbol coding (= 0.70), visuospatial working memory (= 0.71) and smell identification (= 0.77).

In meta-regression analyses, there was a significant negative association between GAF score (level of functioning) and executive dysfunction (2.75, = 0.006), verbal working memory (= 3.95, = 0.0001) and premorbid IQ (= 2.87, = 0.004). There were no significant associations between any cognitive impairment and any other variables (with the exception of the association between genetic risk ratio and cognitive impairment in UHR, which is reported in the section 'Effect of combined genetic and clinical risk on cognition').

Comparison of cognitive deficits in FHR and UHR

In FHR subjects, impairment in IQ (= 0.81 vs. = 0.40, Qbet = 20.0, < 0.001) and premorbid IQ (= 0.64 vs. = 0.30, Qbet = 13.1, < 0.001) were more robust than deficits in UHR. In other cognitive domains, impairments in UHR and FHR groups were comparable, except more impaired visuospatial working memory in UHR than FHR (= 0.71 vs. = 0.35, Qbet = 4.6, = 0.03; Fig. 2). Examination of individual tasks showed more impairment in the UHR group on symbol coding (= 0.70 vs. = 0.37, Qbet = 6.40, = 0.01).

Effect of combined genetic and clinical risk on cognition

Meta-regression in the UHR sample showed that a higher ratio of GRDS was associated with more severe deficits in verbal memory (= 2.0, = 0.04), premorbid IQ (= 1.95, = 0.05) and CPT (= 2.06, = 0.04). There was also a trend association for verbal working memory (= 0.07).

In FHR studies, symptomatic subjects were significantly more impaired than asymptomatic subjects in the two domains examined: verbal memory (= 0.28, CI = 0.0–0.56, = 0.05, I2 = 0) and processing speed (= 0.51, CI = 0.0–1.02, = 0.05, I2 = 0.14).

Baseline cognitive functioning and transition to frank psychosis

UHR-P vs. UHR-NP

At baseline, UHR-P subjects were cognitively more impaired than UHR-NP in all domains (= 0.31–0.49), except sustained attention (= 0.11; Table 4). UHR-P were also more impaired in most individual task analyses (= 0.26–0.52), except Trail Making Task A and letter number sequencing. Non-overlap ratios for these effect sizes were only between 9 and 34%, suggesting that the vast majority of UHR-P and UHR-NP subjects have similar performances. Heterogeneity of distribution of effect sizes was small (range of I2 = 0–0.22) and significant only for verbal fluency tasks. There was no evidence of publication bias.

Table 4. Mean weighted effect sizes for cognitive differences between UHR-P and UHR-NP
TestSampleUHR-PUHR-NP D 95% CI Z P Q-test P I 2 Bias
  1. UHR, Ultra-high risk to psychosis; UHR-P, UHR who converted to psychotic disorder at follow-up; UHR-P, UHR who has not converted to psychotic disorder at follow-up; TMT, Trail making test; WM, Working memory; EF, executive functions; CPT, Continous performance test; LNS, Letter, Number sequencing.

IQ 61724720.390.13–0.652.960.0030.140.040.39
Premorbid72295180.380.21–0.544.54<0.0010.7300.66
Processing speed61913450.330.14–0.523.390.00070.470.060.94
Symbol coding41652540.390.06–0.722.300.020.080.060.81
TMT B61863320.370.18–0.563.810.00010.8200.37
Verbal memory72033520.430.24–0.624.43<0.0010.3900.60
Learning51372080.260.03–0.502.210.030.6200.72
Delayed51231810.480.13–0.842.660.0080.080.120.57
Visual memory 41251660.490.24–0.753.810.00010.8700.45
EF 61624410.340.15–0.533.530.00040.4800.37
WCST51452750.260.03–0.482.210.030.340.010.40
Verbal WM 61672760.490.20–0.783.340.00080.120.050.45
LNS3721140.32–0.01–0.641.950.050.6500.18
Visspa WM 71161940.300.06–0.542.450.010.8400.99
Attention 61693130.11–0.08–0.311.150.250.8500.95
CPT61693130.11–0.08–0.311.150.250.8500.95
Fluency 72043570.490.11–0.862.520.010.00080.180.37
Letter fluency61913020.520.09–0.952.350.020.00040.220.36

In meta-regression analyses, lower transition rate was significantly associated with larger IQ (= 3.9, SE = 1.8, = 0.03) differences between UHR-P and UHR-NP.

The effect of transition to psychosis in FHR

In the current review, only one study examined the relationship between baseline cognitive deficits and the development of psychosis by follow-up. Verbal memory was the only measure that was significantly impaired in FHR subjects who developed psychosis by follow-up compared with those who did not [67]. There is some additional evidence of for cognitive deficits in childhood as predictors of psychosis in the follow-up in FHR samples, but these studies are excluded based on inclusion criteria.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Findings of this meta-analysis suggest that cognitive functions are significantly impaired in both FHR and UHR. The severity of most of these deficits is modest, and cognitive deficits appear to be more severe in people who have attenuated psychotic symptoms together with genetic risk for psychosis. General intellectual deficits were more strongly associated with genetic risk and poor performance on visuospatial working memory and digit symbol coding was more strongly associated with clinical high risk. This meta-analysis also provided evidence that cognitive functions were significantly more impaired in UHR-P compared with UHR-NP, although non-overlap ratios of both groups were small. These differences were most pronounced for general intellectual capacity in samples with lower transition rate, suggesting that UHR subjects in enriched samples are more similar to each other, independent of outcome.

Cognitive deficits in FHR

Findings of this meta-analysis suggest that FHR is associated with modest sized cognitive deficits, and distribution of effect sizes were quite homogenous. Cognitive deficits in FHR are comparable to abnormalities reported in meta-analysis of older unaffected relatives of schizophrenia [11]. However, older relatives have less severe deficits than our findings in general intelligence [12]. This difference might be related to the fact that older relatives are past the age of peak risk for psychosis onset, while young relatives may still develop the disorder. In older samples, better intelligence might be a protective factor in this population, as intellectual impairment is associated with higher risk for psychosis [68]. The relative strength of IQ finding in FHR suggests that general rather than specific cognitive deficits might be the genetically transmitted neurocognitive correlate of schizophrenia. This is consistent with evidence suggesting that genetics of intellectual disability and schizophrenia might be related [69, 70], and the fact that psychosis is more common among people with intellectual disabilities than the general population [68]. It could be argued that deficits observed in range of cognitive domains in FHR might be secondary to general cognitive dysfunction, and some authors have already suggested that this might be also true in established schizophrenia [71]. However, it is not possible to exclude the likely contributions of more specific cognitive deficits to psychosis risks. Future FHR studies controlling IQ might likely reveal such deficits.

Another important question is whether neurocognitive deficits in FHR are specific to genetic risk to schizophrenia or they are shared risk factor for psychosis in general. Few studies have compared neurocognitive deficits in young relatives of schizophrenia and affective psychoses and bipolar disorder. Schubert et al. [53] reported that offspring of patients with schizophrenia have more severe cognitive deficits than those of parents with affective psychosis, but offspring of patients with affective psychosis also underperformed in some tasks including selective attention and grammatical reasoning. Maziade et al. [72] found shared neurocognitive dysfunction in young offspring of patients with schizophrenia and bipolar disorders. Available evidence does not support the specificity of cognitive deficits to genetic risk to schizophrenia. Current data suggest that cognitive deficits are evident in young relatives of patients with psychosis (affective and non-affective), but might be modestly more severe in relatives of schizophrenia.

Cognitive deficits in UHR in comparison to FHR

Similarly, UHR status was associated with modest sized cognitive deficits and distribution of effect sizes across studies were more homogenous than reported in Giuliano et al. [19] (I2 = 30–70% for most measures) and Fusar-Poli et al. [18] (I2 = 50–64% for 4/9 of measures) who reported high I2 scores. The differences in the studies included in the current, meta-analysis and previous studies (i.e. new PACE data) are likely to explain these differences. Also, Fusar-Poli et al. [18] combined the data of the studies including healthy controls and help-seeking patient controls, which is problematic and would contribute to heterogeneity. Our findings of consistent cognitive deficits in UHR suggest that cognitive dysfunction predates the onset of first-episode psychosis, and that these cognitive deficits contribute to functional impairment. However, magnitudes of these deficits are modest compared with FEP and established illness. Indeed, the size of effects is more comparable to findings in FHR. For example, the magnitude of impairment in UHR in digit symbol task is half of that observed in schizophrenia [73]. A number of factors are likely to contribute to the observed cognitive differences between UHR and first-episode psychosis. Most importantly, like any other construct defining milder severity cases within a category of a mental illness, UHR is expected to recruit higher percentage of patients with good prognosis and false-positives [74]. Second, UHR is a more heterogeneous concept than established schizophrenia and FEP: it is likely to include a mixture of true prodromal schizophrenia, affective psychosis and other psychotic disorders; subjects who are in psychotic disorders spectrum but have a favourable outcome or present predominantly with negative symptoms and subjects with non-psychotic conditions (false positives). Similarly, many relatives in FHR group might not carry susceptibility genes that their parents or siblings carry. Also, use of antipsychotics is relatively less common in at-risk subjects than FEP and chronic samples which might be related to poor performance in some cognitive tasks. Therefore, it is not unexpected that lesser percentage of UHR and FHR subjects would have cognitive deficits of modest effect sizes. These findings has also been interpreted as an indirect evidence of neuroprogression; however, most of the available longitudinal studies did not show progressive cognitive deficits in UHR and FE patients [23, 75, 76], and a recent meta-analysis of follow-up studies of neurocognition in UHR and FEP studies did not find any evidence of cognitive decline [77].

The preliminary comparison between FHR and UHR suggests that general intellectual deficit is more likely associated with FHR. One explanation is that genetic risk paradigms might be more likely to recruit subjects with vulnerability to certain subgroups of schizophrenia (chronic cases that remain in care and may be associated with lower IQ) than clinical risk. The UHR criteria rely on the presence of positive, but not negative, symptoms and referral system through the school system is often used in recruitment of UHR. In contrast, visuospatial working memory and digit symbol tasks were more impaired in UHR compared with FHR. However, more studies directly comparing FHR and UHR are necessary to be conclusive in the subject as statistical comparison of findings of independent studies has limitations.

Co-occurrence of genetic and clinical risk

Our meta-analysis suggests that co-occurring attenuated symptoms and genetic risk was associated with more severe cognitive deficits. Seidman and colleagues found similar results in a clinical high risk sample [52]. Our meta-analysis provides additional evidence for the effect of co-occurrence on FHR studies. One explanation might be cumulative effect of different risk factors on cognition. However, association between GRDS+APS and cognitive deficits might be also a severity and reliability issue: clinical high risk together with family history is associated with higher risk of transition in follow-up, and attenuated symptoms on the background of genetic risk are expected to be a better indicator of to true prodrome or vulnerability. Cognitive deficits in pure UHR or FHR are likely to be less than reported in studies included. These findings might have implications for the discussions and reservations about introducing attenuated psychosis syndrome in psychiatric classification systems due to concern regarding validity and false-positives. Defining a risk syndrome based on attenuated psychotic symptoms only in relatives of people with schizophrenia can significantly increase the validity of the syndrome and decrease false positives. However, such a risk syndrome would have low sensitivity as only one third of the patients with schizophrenia have a family history of the illness.

Prediction of psychosis in high-risk samples

One important goal of the high risk research is to identify markers which can be use to predict illness onset [78]. In our meta-analysis, cognitive deficits are significantly more severe in subjects who developed psychosis by follow-up, suggesting that some of cognitive impairments in UHR samples are associated with risk to transition to full-blown psychotic disorder. This may also be true for FHR, but more research examining the transition to psychosis in FHR samples is needed. Interestingly, unlike other domains, sustained attention performance was not different between UHR-P and UHR-NP. One explanation might be that sustained attention is a vulnerability marker for being at risk for psychosis but not related to onset of the illness. Alternatively, sustained attention deficits might be observed not only in conditions within psychotic spectrum but also in other conditions which might be overrepresented in UHR-NP group such as mood disorders in which sustained attention deficit suggested to be a trait-marker [79-82] and has been shown to be not related to history of psychosis [83].

While cognitive dysfunction was more severe in UHR-P compared with UHR-NP, effects sizes for between-group differences were modest, approximately Cohen = 0.5 at most (for verbal fluency, verbal and visual memory and working memory), suggesting that there is significant overlap even for these measures (67%) between cognitive performances of UHR-P and UHR-NP. Such a high overlap rate suggests that potential of neurocognition to predict illness prognosis is very limited. In addition, not all UHR-P subjects will be diagnosed with schizophrenia, and it is possible that cognitive deficits might have a better predictive value for schizophrenia than other psychotic illnesses. A recent study failed to show significant differences between UHR subjects who developed schizophrenia and affective psychosis [84]. Still, more research is needed in this subject. On the other hand, Lin et al. [85] showed larger and broader differences in cognitive performance when UHR individuals with a poor functional outcome compared with good outcome (regardless of transition status), than when UHR-P and UHR-NP were compared. In this meta-analysis, we also found that decreased GAF scores are related to more severe cognitive deficits. These findings suggest that reducing the heterogeneity of the UHR sample by examining functional outcome may be a useful approach. Of note, our confounding factor analyses suggested that these differences in general intellectual abilities might be more severe in samples with lower transition rates and that UHR-P and UHR-NP groups perform more similarly in samples with a high transition rate. This finding supports the possibility that low intellectual capacity might be a vulnerability factor for risk of developing psychosis (whether they develop psychosis or not) as samples with high transition rates are likely to be enriched UHR samples. In contrast, many UHR-NP subjects in samples with low transition rate would be false-positives who have preserved intellectual abilities and would drive the between-group differences. In accordance with these findings, studies examining non-help-seeking UHR samples reported less cognitive deficits [86].

The advantages of this meta-analysis include homogenous distribution of effect sizes, co-examination of both familial and genetic risk and large sample size. Limitations of this meta-analysis include differences in methodology, such as follow-up duration and the various risk criteria and exclusion criteria used in the included studies. Many studies did not report variables such as positive/negative symptoms, functioning levels, antipsychotic use, and cannabis use and other substance use.

In conclusion, findings of this meta-analysis confirm that UHR and FHR subjects have cognitive deficits that are modest compared with established schizophrenia. Cognitive impairment is larger in subjects who have both genetic and clinical risk and those who developed psychotic disorder at follow-up. However, it is not possible to know timing of the development of cognitive deficits with cross-sectional methods. There is a need for large-scale longitudinal studies that follow at-risk subjects many years before developing UHR to reveal timing and development of cognitive deficits in psychotic disorder.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Prof Christos Pantelis and Prof Patrick McGorry were supported by a National Health and Medical Research Council of Australia (NHMRC) Program Grant (ID: 566529). Prof Christos Pantelis was also supported by a NHMRC Senior Principal Research Fellowship (ID: 628386).

Declaration of interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Over the last two years, Christos Pantelis has participated on Advisory Boards for Janssen-Cilag and Lundbeck. He has received honoraria for talks presented at educational meetings organized by Janssen-Cilag and Lundbeck. Patrick McGorry received research grant support from Janssen Cilag and Astra Zeneca and honoraria for educational lectures from Janssen Cilag, Eli Lilly, Lundbeck, Pfizer and Roche. Other authors have no conflict of interests to be reported. None of the authors have financial or personal relationships, interests and affiliations relevant to the subject matter of the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix
  • 1
    Bora E, Yücel M, Pantelis C. Cognitive impairment in schizophrenia and affective psychoses: implications for DSM-V criteria and beyond. Schizophr Bull 2010;36:3642.
  • 2
    Heinrichs RW, Zakzanis KK. Neurocognitive deficit in schizophrenia: a quantitative review of the evidence. Neuropsychology 1998;12:426445.
  • 3
    Bora E, Yücel M, Pantelis C. Theory of mind impairment: a distinct trait-marker for schizophrenia spectrum disorders and bipolar disorder? Acta Psychiatr Scand 2009;120:253264.
  • 4
    Green MF, Kern RS, Braff DL, Mintz J. Neurocognitive deficits and functional outcome in schizophrenia: are we measuring the right stuff? Schizophr Bull 2000;26:119136.
  • 5
    Liddle PF. Cognitive impairment in schizophrenia: its impact on social functioning. Acta Psychiatr Scand Suppl 2000;400:1116.
  • 6
    Mesholam-Gately RI, Giuliano AJ, Goff KP, Faraone SV, Seidman LJ. Neurocognition in first-episode schizophrenia: a meta-analytic review. Neuropsychology 2009;23:315336.
  • 7
    Hofer A, Bodner T, Kaufmann A, et al. Symptomatic remission and neurocognitive functioning in patients with schizophrenia. Psychol Med 2011;41:21312139.
  • 8
    Szöke A, Trandafir A, Dupont ME, Méary A, Schürhoff F, Leboyer M. Longitudinal studies of cognition in schizophrenia: meta-analysis. Br J Psychiatry 2008;192:248257.
  • 9
    Lawrie SM, Byrne M, Miller P, et al. Neurodevelopmental indices and the development of psychotic symptoms in subjects at high risk of schizophrenia. Br J Psychiatry 2001;178:524530.
  • 10
    Tarrant CJ, Jones PB. Precursors to schizophrenia: do biological markers have specificity? Can J Psychiatry 1999;44:335349.
  • 11
    Sitskoorn MM, Aleman A, Ebisch SJ, Appels MC, Kahn RS. Cognitive deficits in relatives of patients with schizophrenia: a meta-analysis. Schizophr Res 2004;71:285295.
  • 12
    Snitz BE, Macdonald AW III, Carter CS. Cognitive deficits in unaffected first-degree relatives of schizophrenia patients: a meta-analytic review of putative endophenotypes. Schizophr Bull 2006;32:179194.
  • 13
    Meijer J, Simons CJ, Quee PJ, Verweij K, GROUP Investigators. Cognitive alterations in patients with non-affective psychotic disorder and their unaffected siblings and parents. Acta Psychiatr Scand 2012;125:6676.
  • 14
    Agnew-Blais J, Seidman LJ. Neurocognition in youth and young adults under age 30 at familial risk for schizophrenia: a quantitative and qualitative review. Cogn Neuropsychiatry 2013;18:4482.
  • 15
    Yung AR, McGorry PD. The prodromal phase of first-episode psychosis: past and current conceptualizations. Schizophr Bull 1996;22:353370.
  • 16
    Rietdijk J, Klaassen R, Ising H, et al. Detection of people at risk of developing a first psychosis: comparison of two recruitment strategies. Acta Psychiatr Scand 2012;126:2130.
  • 17
    Zimbrón J, Ruiz DE Azúa S, Khandaker GM, et al. Clinical and sociodemographic comparison of people at high-risk for psychosis and with first-episode psychosis. Acta Psychiatr Scand 2013;127:210216.
  • 18
    Fusar-Poli P, Deste G, Smieskova R, et al. Cognitive functioning in prodromal psychosis: a meta-analysis. Arch Gen Psychiatry 2012;69:562571.
  • 19
    Giuliano AJ, Li H, Mesholam-Gately RI, Sorenson SM, Woodberry KA, Seidman LJ. Neurocognition in the psychosis risk syndrome: a quantitative and qualitative review. Curr Pharm Des 2012;18:399415.
  • 20
    Moher D, Liberati A, Tetzlaff J, Altman DG, Group PRISMA. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535.
  • 21
    Barrantes-Vidal N, Aguilera M, Campanera S, et al. Working memory in siblings of schizophrenia patients. Schizophr Res 2007;95:7075.
  • 22
    Becker HE, Nieman DH, Dingemans PM, van de Fliert JR, de Haan L, Linszen DH. Verbal fluency as a possible predictor for psychosis. Eur Psychiatry 2010;25:105110.
  • 23
    Becker HE, Nieman DH, Wiltink S, et al. Neurocognitive functioning before and after the first psychotic episode: does psychosis result in cognitive deterioration? Psychol Med 2010;40:15991606.
  • 24
    Bertisch HC, Fava J, Kattan A, Delisi LE. Preliminary neuropsychological findings in individuals at high genetic risk for schizophrenia. Early Interv Psychiatry 2008;2:4549.
  • 25
    Bonner-Jackson A, Csernansky JG, Barch DM. Levels of processing effects in first-degree relatives of individuals with schizophrenia. Biol Psychiatry 2006;61:11411147.
  • 26
    Broome MR, Johns LC, Valli I, et al. Delusion formation and reasoning biases in those at clinical high risk for psychosis. Br J Psychiatry 2007;51:s38s42.
  • 27
    Broome MR, Day F, Valli I, et al. Delusional ideation, manic symptomatology and working memory in a cohort at clinical high-risk for psychosis: a longitudinal study. Eur Psychiatry 2012;27:258263.
  • 28
    Byrne M, Clafferty BA, Cosway R, et al. Neuropsychology, genetic liability, and psychotic symptoms in those at high risk of schizophrenia. J Abnorm Psychol 2003;112:3848.
  • 29
    Carrión RE, Goldberg TE, McLaughlin D, Auther AM, Correll CU, Cornblatt BA. Impact of neurocognition on social and role functioning in individuals at clinical high risk for psychosis. Am J Psychiatry 2011;168:806813.
  • 30
    Choi JS, Kang DH, Park JY, et al. Cavum septum pellucidum in subjects at ultra-high risk for psychosis: compared with first-degree relatives of patients with schizophrenia and healthy volunteers. Prog Neuropsychopharmacol Biol Psychiatry 2008;32:13261330.
  • 31
    Delawalla Z, Csernansky JG, Barch DM. Prefrontal cortex function in nonpsychotic siblings of individuals with schizophrenia. Biol Psychiatry 2008;63:490497.
  • 32
    Francey SM, Jackson HJ, Phillips LJ, Wood SJ, Yung AR, McGorry PD. Sustained attention in young people at high risk of psychosis does not predict transition to psychosis. Schizophr Res 2005;79:127136.
  • 33
    Frommann I, Pukrop R, Brinkmeyer J, et al. Neuropsychological profiles in different at-risk states of psychosis: executive control impairment in the early–and additional memory dysfunction in the late–prodromal state. Schizophr Bull 2011;37:861873.
  • 34
    Green MF, Bearden CE, Cannon TD, et al. Social cognition in schizophrenia, Part 1: performance across phase of illness. Schizophr Bull 2012;38:854864.
  • 35
    Groom MJ, Jackson GM, Calton TG, et al. Cognitive deficits in early-onset schizophrenia spectrum patients and their non-psychotic siblings: a comparison with ADHD. Schizophr Res 2008;99:8595.
  • 36
    Hans SL, Auerbach JG, Nuechterlein KH, et al. Neurodevelopmental factors associated with schizotypal symptoms among adolescents at risk for schizophrenia. Dev Psychopathol 2009;21:11951210.
  • 37
    Hughes C, Kumari V, Das M, et al. Cognitive functioning in siblings discordant for schizophrenia. Acta Psychiatr Scand 2005;111:185192.
  • 38
    Jahshan C, Heaton RK, Golshan S, Cadenhead KS. Course of neurocognitive deficits in the prodrome and first episode of schizophrenia. Neuropsychology 2010;24:109120.
  • 39
    Kim HS, Shin NY, Jang JH, et al. Social cognition and neurocognition as predictors of conversion to psychosis in individuals at ultra-high risk. Schizophr Res 2011;130:170175.
  • 40
    Kim KR, Park JY, Song DH, Koo HK, An SK. Neurocognitive performance in subjects at ultrahigh risk for schizophrenia: a comparison with first-episode schizophrenia. Compr Psychiatry 2011;52:3340.
  • 41
    Klemm S, Schmidt B, Knappe S, Blanz B. Impaired working speed and executive functions as frontal lobe dysfunctions in young first-degree relatives of schizophrenic patients. Eur Child Adolesc Psychiatry 2006;15:400408.
  • 42
    Koutsouleris N, Patschurek-Kliche K, Scheuerecker J, et al. Neuroanatomical correlates of executive dysfunction in the at-risk mental state for psychosis. Schizophr Res 2010;123:160174.
  • 43
    Lin A, Yung AR, Nelson B, et al. Neurocognitive transition to psychosis: medium to long-term findings from a sample at ultra-high risk for psychosis. Psychol Med 2013;43:23492360.
  • 44
    Lindgren M, Manninen M, Laajasalo T, et al. The relationship between psychotic-like symptoms and neurocognitive performance in a general adolescent psychiatric sample. Schizophr Res 2010;123:7785.
  • 45
    Maziade M, Rouleau N, Cellard C, et al. Young offspring at genetic risk of adult psychoses: the form of the trajectory of IQ or memory may orient to the right dysfunction at the right time. PLoS One 2011;6:e19153.
  • 46
    Myles-Worsley M, Ord LM, Ngiralmau H, Weaver S, Blailes F, Faraone SV. The Palau Early Psychosis Study: neurocognitive functioning in high-risk adolescents. Schizophr Res 2007;89:299307.
  • 47
    Nam HJ, Kim N, Park T, et al. Cognitive profiles of healthy siblings of schizophrenia patients:application of the cognitive domains of the MATRICS consensus battery. World J Biol Psychiatry 2009;10:452460.
  • 48
    Pflueger MO, Gschwandtner U, Stieglitz RD, Riecher-Rössler A. Neuropsychological deficits in individuals with an at risk mental state for psychosis - working memory as a potential trait marker. Schizophr Res 2007;97:1424.
  • 49
    Prasad KM, Sanders R, Sweeney J, et al. Neurological abnormalities among offspring of persons with schizophrenia: relation to premorbid psychopathology. Schizophr Res 2009;108:163169.
  • 50
    Pukrop R, Schultze-Lutter F, Ruhrmann S, et al. Neurocognitive functioning in subjects at risk for a first episode of psychosis compared with first- and multiple-episode schizophrenia. J Clin Exp Neuropsychol 2006;28:13881407.
  • 51
    Pukrop R, Ruhrmann S, Schultze-Lutter F, Bechdolf A, Brockhaus-Dumke A, Klosterkötter J. Neurocognitive indicators for a conversion to psychosis: comparison of patients in a potentially initial prodromal state who did or did not convert to a psychosis. Schizophr Res 2007;92:116125.
  • 52
    Riecher-Rössler A, Pflueger MO, Aston J, et al. Efficacy of using cognitive status in predicting psychosis: a 7-year follow-up. Biol Psychiatry 2009;66:10231030.
  • 53
    Schubert EW, McNeil TF. Neuropsychological impairment and its neurological correlates in adult offspring with heightened risk for schizophrenia and affective psychosis. Am J Psychiatry 2005;162:758766.
  • 54
    Schubert EW, McNeil TF. Neurobehavioral deficits in young adult offspring with heightened risk for psychosis who developed schizophrenia-spectrum disorder. Schizophr Res 2007;94:107113.
  • 55
    Seidman LJ, Giuliano AJ, Smith CW, et al. Neuropsychological functioning in adolescents and young adults at genetic risk for schizophrenia and affective psychoses: results from the Harvard and Hillside Adolescent High Risk Studies. Schizophr Bull 2006;32:507524.
  • 56
    Seidman LJ, Giuliano AJ, Meyer EC, et al. Neuropsychology of the prodrome to psychosis in the NAPLS consortium: relationship to family history and conversion to psychosis. Arch Gen Psychiatry 2010;67:578588.
  • 57
    Serrani D. Neurocognitive assessment of ultra high risk of psychosis states using the MATRICS battery (Measurement and Treatment Research to Improve Cognition in Schizophrenia). Rev Psiquiatr Clín 2011;38:4.
  • 58
    Smith CW, Park S, Cornblatt B. Spatial working memory deficits in adolescents at clinical high risk for schizophrenia. Schizophr Res 2006;81:211215.
  • 59
    Stanford AD, Messinger J, Malaspina D, Corcoran CM. Theory of Mind in patients at clinical high risk for psychosis. Schizophr Res 2011;131:1117.
  • 60
    Szily E, Kéri S. Anomalous subjective experience and psychosis risk in young depressed patients. Psychopathology 2009;42:229235.
  • 61
    van Rijn S, Schothorst P, Wout M, et al. Affective dysfunctions in adolescents at risk for psychosis: emotion awareness and social functioning. Psychiatry Res 2011;187:100105.
  • 62
    Wolf LE, Cornblatt BA, Roberts SA, Shapiro BM, Erlenmeyer-Kimling L. Wisconsin Card Sorting deficits in the offspring of schizophrenics in the New York High-Risk Project. Schizophr Res 2002;57:173.
  • 63
    Wood SJ, Brewer WJ, Koutsouradis P, et al. Cognitive decline following psychosis onset: data from the PACE clinic. Br J Psychiatry 2007;51:s52s57.
  • 64
    Woodberry KA, Seidman LJ, Giuliano AJ, Verdi MB, Cook WL, McFarlane WR. Neuropsychological profiles in individuals at clinical high risk for psychosis: relationship to psychosis and intelligence. Schizophr Res 2010;123:188198.
  • 65
    Bax L, Yu LM, Ikeda N, Tsuruta H, Moons KG. Development and validation of MIX: comprehensive free software for meta-analysis of causal research data. BMC Med Res Methodol 2006;6:50.
  • 66
    Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:15391558.
  • 67
    Whyte MC, Brett C, Harrison LK, et al. Neuropsychological performance over time in people at high risk of developing schizophrenia and controls. Biol Psychiatry 2006;59:730739.
  • 68
    Morgan VA, Leonard H, Bourke J, Jablensky A. Intellectual disability co-occurring with schizophrenia and other psychiatric illness: population-based study. Br J Psychiatry 2008;193:364372.
  • 69
    Owen MJ, O'donovan MC, Thapar A, Craddock N. Neurodevelopmental hypothesis of schizophrenia. Br J Psychiatry 2011;198:173175.
  • 70
    Toulopoulou T, Goldberg TE, Mesa IR, et al. Impaired intellect and memory: a missing link between genetic risk and schizophrenia? Arch Gen Psychiatry 2010;67:905913.
  • 71
    Dickinson D, Iannone VN, Wilk CM, Gold JM. General and specific cognitive deficits in schizophrenia. Biol Psychiatry 2004;55:826833.
  • 72
    Maziade M, Rouleau N, Gingras N, et al. Shared neurocognitive dysfunctions in young offspring at extreme risk for schizophrenia or bipolar disorder in eastern quebec multigenerational families. Schizophr Bull 2009;35:919930.
  • 73
    Dickinson D, Ramsey ME, Gold JM. Overlooking the obvious: a meta-analytic comparison of digit symbol coding tasks and other cognitive measures in schizophrenia. Arch Gen Psychiatry 2007;64:532542.
  • 74
    Velthorst E, Nieman DH, Klaassen RM, et al. Three-year course of clinical symptomatology in young people at ultra high risk for transition to psychosis. Acta Psychiatr Scand 2011;123:3642.
  • 75
    Bozikas VP, Andreou C. Longitudinal studies of cognition in first episode psychosis: a systematic review of the literature. Aust N Z J Psychiatry 2011;45:93108.
  • 76
    Hawkins KA, Keefe RS, Christensen BK, et al. Neuropsychological course in the prodrome and first episode of psychosis: findings from the PRIME North America Double Blind Treatment Study. Schizophr Res 2008;105:19.
  • 77
    Bora E, Murray RM. Meta-analysis of cognitive deficits in ultra-high risk to psychosis and first-episode psychosis: do the cognitive deficits progress over, or after, the onset of psychosis? Schizophr Bull 2013; doi: 10.1093/schbul/sbt085 [Epub ahead of print].
  • 78
    Nieman DH, Velthorst E, Becker HE, et al. The Strauss and Carpenter Prognostic Scale in subjects clinically at high risk of psychosis. Acta Psychiatr Scand 2013;127:5361.
  • 79
    Clark L, Goodwin GM. State- and trait-related deficits in sustained attention in bipolar disorder. Eur Arch Psychiatry Clin Neurosci 2004;254:6168.
  • 80
    Bora E, Vahip S, Akdeniz F. Sustained attention deficits in manic and euthymic patients with bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry 2006;30:10971102.
  • 81
    Ancín I, Santos JL, Teijeira C, et al. Sustained attention as a potential endophenotype for bipolar disorder. Acta Psychiatr Scand 2010;122:235245.
  • 82
    Bora E, Yücel M, Pantelis C. Cognitive endophenotypes of bipolar disorder: a meta-analysis of neuropsychological deficits in euthymic patients and their first-degree relatives. J Affect Disord 2009;113:120.
  • 83
    Bora E, Vahip S, Akdeniz F, et al. The effect of previous psychotic mood episodes on cognitive impairment in euthymic bipolar patients. Bipolar Disord 2007;9:468477.
  • 84
    Olvet DM, Stearns WH, McLaughlin D, et al. Comparing clinical and neurocognitive features of the schizophrenia prodrome to the bipolar prodrome. Schizophr Res 2010;123:5963.
  • 85
    Lin A, Wood SJ, Nelson B, et al. Neurocognitive predictors of functional outcome two to 13 years after identification as ultra-high risk for psychosis. Schizophr Res 2011;132:17.
  • 86
    Mukkala S, Ilonen T, Nordström T, et al. Different vulnerability indicators for psychosis and their neuropsychological characteristics in the Northern Finland 1986 Birth Cohort. J Clin Exp Neuropsychol 2011;33:385394.

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Declaration of interest
  9. References
  10. Appendix

Individual cognitive tests under each domain

  1. TMT, Trail making test; HSCT, Hayling sentence completion task; RAVLT, Rey auditory verbal learning test; HVLT, Hopkins verbal learning test; WMS, Wechsler memory scale; WMS, Wechsler memory scale; RCFT, Rey complex figure test; CVLT, California Hopkins verbal learning test; SOPT, Self-ordered pointing test.

Processing speedTMT-A, digit symbol, letter fluency, category fluency, reaction time, TMT-B, Stroop interference, HSCT, set shifting
Verbal memoryRAVLT, CVLT, HVLT, logical memory, word pair, paired associate learning
Visual memoryVisual reproduction/ WMS visual memory, RCFT
Verbal working memoryDigit span, letter number sequencing
Visual working memorySpatial span, SWM, DMTS, SOPT, DRT, visual 2 back, beads task
Executive functionsWCST, mazes, matrix reasoning
AttentionContinuous performance test, d2 concentration test, cancellation test, selective attention
FluencyLetter fluency, category fluency