See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/328301956

Interventions and Social Functioning in Youth at Risk of Psychosis: A

Systematic Review and Meta‐analysis

Article in Early Intervention in Psychiatry · June 2018

DOI: 10.1111/eip.12689

CITATIONS READS

42 405

4 authors, including:

Daniel J Devoe Megan Farris

The University of Calgary Medlior Health Outcomes Research Ltd
72 PUBLICATIONS 531 CITATIONS 63 PUBLICATIONS 713 CITATIONS

SEE PROFILE SEE PROFILE

Parker Townes
University of Ottawa
12 PUBLICATIONS 111 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Risk Factors for Psychosis: Paradigms, Mechanisms, and Prevention View project

Persistent Negative Symptoms in Youth at Clinical High Risk for Psychosis: A Longitudinal Study View project

All content following this page was uploaded by Daniel J Devoe on 27 June 2018.

The user has requested enhancement of the downloaded file.

Received: 17 January 2018 Revised: 28 March 2018 Accepted: 16 May 2018
DOI: 10.1111/eip.12689

REVIEW ARTICLE

Interventions and social functioning in youth at risk of

psychosis: A systematic review and meta-analysis
Daniel J. Devoe | Megan S. Farris | Parker Townes | Jean Addington

Department of Psychiatry, Hotchkiss Brain

Institute, University of Calgary, Alberta, Aim: Youth at clinical high risk (CHR) for psychosis often exhibit difficulties in social functioning
Canada and poorer social functioning may be predictive of transition to a psychotic disorder. Therefore,
Correspondence the primary objective of this systematic review was to summarize the impact of all interventions
Dr Jean Addington, Mathison Centre for
on social functioning in CHR samples.
Mental Health Research and Education, 3280
Hospital Drive NW|Calgary, Alberta T2N 4Z6, Method: Electronic databases PsycINFO, CINAHL, Embase, EBM, and MEDLINE were searched
Canada. from 1951 to June 2017. Studies were selected if they included any intervention that reported
Email: jmadding@ucalgary.ca changes in social functioning in youth at CHR. Data were evaluated using random effects pair-
Funding information wise meta-analyses, stratified by time, and reported as the standardized mean difference (SMD).
National Institute of Mental Health, Grant/
Results: Nineteen studies met our inclusion criteria, including a total of 1513 CHR participants.
Award Number: RO1MH105178; Alberta
Innovates Graduate Studentship The mean age was 20.5 years and 47% were male. Cognitive behavioural therapy (4 studies) did
not significantly improve social functioning at 6 months (SMD = 0.06; 95% confidence interval
[CI] = −0.35, 0.46), 12 months (SMD = −0.15; 95% CI = –0.38, 0.08) and 18 months
(SMD = 0.20; 95% CI = −0.10, 0.50). Omega-3 (2 studies) did not significantly improve social
functioning at 6 months (SMD = 0.01; 95% CI = −0.21, 0.24) and 12 months (SMD = −0.08;
95% CI = −0.33, 0.17). Lastly, cognitive remediation (3 studies) did not significantly improve
social functioning at 2- to 3-month follow-up (SMD = 0.13, 95% CI = –0.18, 0.43).
Conclusions: This systematic review and meta-analysis demonstrated that no treatment signifi-
cantly improved social functioning in youth at CHR. Future randomized control trials are
required that are designed to target and improve social functioning in youth at CHR for
psychosis.

KEYWORDS

clinical high risk, psychosis, social functioning, treatment

1 | I N T RO D UC T I O N 2007). In schizophrenia research, social functioning has recently

become a primary outcome of interest for determining treatment effi-
Deficits in social functioning are a hallmark of patients with schizo- cacy (Green, Hellemann, Horan, Lee, & Wynn, 2012; Kane et al.,
phrenia (Addington, Penn, Woods, Addington, & Perkins, 2008; 2016; Velthorst et al., 2017). Nevertheless, attenuated psychotic
Velthorst et al., 2017) and are often present before the onset of illness symptoms and transition to psychosis remain the primary outcomes
(Carrion et al., 2013; Cornblatt et al., 2012). Likewise, those at clinical of interest in most CHR treatment trials to date, even though poor
high risk (CHR) for psychosis often present with difficulties in social social functioning may be more incapacitating than attenuated psy-
functioning such as social isolation (Cornblatt et al., 2003) that may be chotic symptoms (Addington et al., 2008; Cornblatt et al., 2007).
observed in those experiencing a first episode of psychosis or even Social functioning in those at CHR for psychosis has been
chronic course of schizophrenia (Addington et al., 2008). Definitions reported to be associated with transition and poor social functioning
of social functioning remain inconsistent in CHR studies due to the is predictive of transition to a frank psychotic disorder (Addington
differences in social functioning scales, however, for the purpose of et al., 2017; Cannon et al., 2008; Cornblatt et al., 2007; Cornblatt
this review social functioning was defined as the quantity and quality et al., 2012; Jang et al., 2011). Even in those who do not transition to
of peer relationships, level of peer conflict, age-appropriate intimate a psychotic disorder social functioning deficits remain prevalent
relationships and involvement with family members (Cornblatt et al., (Addington et al., 2011) and are associated with a range of symptoms

Early Intervention in Psychiatry. 2018;1–12. wileyonlinelibrary.com/journal/eip © 2018 John Wiley & Sons Australia, Ltd 1
2 DEVOE ET AL.

such as negative symptoms (Corcoran et al., 2011; Kim et al., 2013; P.T.) independently performed title and abstract screening using the
Lee, Kim, Lee, & An, 2017; Meyer et al., 2014; Schlosser et al., 2015), Covidence systematic review software (Mavergames, 2013). Next, full
poor emotional awareness (Kimhy et al., 2016), dyskinesia (Mittal text articles were screened by the same 2 reviewers to determine
et al., 2011) and cognitive deficits (Carrion et al., 2012). Thus, it is per- inclusion in this systematic review. Additionally, abstracts were
tinent to develop treatments that are designed to target and improve reviewed between the years 2010 and 2017 from both the Interna-
social functioning in youth at CHR for psychosis and to investigate tional Congress on Schizophrenia Research and the International
the impact of treatment strategies on social functioning up to the Association for Early Psychosis Conference. Further, the reference
present time. lists of included articles were hand searched for relevant studies not
To our knowledge, no meta-analysis has been conducted specifi- found through online database searching.
cally examining the impact of treatment on social functioning in CHR
samples. However, one previous meta-analysis looked at global func- 2.3 | Selection criteria
tioning scores (ie, scores including both social and role functioning
Studies that met the followed criteria were deemed eligible for inclu-
and severity of psychiatric symptoms) pooling a variety of treatments
sion in this systematic review by 2 reviewers (M.F. and P.T.): (1) original
(eg, cognitive behavioural therapy [CBT], risperidone plus CBT, inte-
research including participants at risk for psychosis meeting criteria for
grated psychological treatments and long-chain ω-3 (omega-3) polyun-
either CHR, at-risk mental state (ARMS), attenuated psychosis syn-
saturated fatty acids) and found no impact of treatment on overall
drome (APS), ultra-high-risk (UHR); (2) included an experimental or
functioning at 12 months (Van Der Gaag et al., 2013). Moreover, in a
observational treatment(s); (3) reported follow-up social functioning
segment of one systematic review, a qualitative approach to the social
scores on the following scales: global functioning social (GFS)
and role functioning literature and treatment in CHR samples was
(Cornblatt et al., 2007), social and occupational functioning assessment
described and concluded that treatments designed at improving func-
scale (SOFAS) (Morosini, Magliano, Brambilla, Ugolini, & Pioli, 2000),
tioning should be considered equally as important as decreasing
social functioning scale (SFS) (Birchwood, Smith, Cochrane, Wetton, &
transition (Cotter et al., 2014).
Copestake, 1990), social adjustment scale (SAS) (Weissman, Prusoff,
This systematic review complements the earlier reviews, by pro-
Thompson, Harding, & Myers, 1978); and (4) reported mean age
viding an in-depth examination of treatment strategies employed to
between 12 and 30 years at study inception. Studies were excluded
date and pooling the impact that these treatments have had on social
based on the following criteria: (1) only reporting baseline social func-
functioning using pairwise meta-analyses. By investigating the impact
tioning data; (2) ineligible study design (case reports, review articles,
of treatment strategies (eg, CBT vs Omega-3) on social functioning
opinion pieces, conference abstracts without sufficient description of
the evidence based on social functioning in CHR samples will be
the study/results, and editorials without original data); (3) studies with-
enhanced which in turn should facilitate researchers in the develop-
out an intervention; and (4) any study reporting social outcomes using
ment of clinical trials that target social functioning. Therefore, the pri-
either the global assessment of functioning scale (GAF) or the subjec-
mary objective of this systematic review and meta-analysis was to
tive well-being under neuroleptics scale (SWNS) because both scales
summarize the impact of all interventions on social functioning in
are heavily influenced by the severity of psychiatric symptoms (eg, sui-
CHR samples.
cidal ideation reduces a participant's score on the GAF significantly). Ini-
tial agreement on title/abstract screening was assessed using the kappa
statistic for interrater reliability between reviewers. Disagreements
2 | METHOD
were reconciled by a third reviewer (D.D.).

2.1 | Protocol
2.4 | Data extraction
This systematic review and meta-analysis was conducted in
Data abstraction was completed in duplication (M.F. and P.T.) includ-
accordance with the preferred reporting for systematic reviews and
ing the following study characteristics: first author, year of publication,
meta-analyses (PRISMA) and meta-analysis of observational studies in
study design, sample size and social functioning scales; participant
epidemiology (MOOSE) guidelines (Liberati et al., 2009; Moher et al.,
characteristics: number of CHR participants, transition percent, mean
2015). A PRISMA checklist is available in Appendix S1, Supporting
 SD age and number of males/percent male; and treatment charac-
Information (Moher, Liberati, Tetzlaff, & Altman, 2009). This review
teristics: intervention, primary outcome and impact of intervention on
was prospectively registered with the PROSPERO database of sys-
social functioning. For the meta-analysis, the following social function-
tematic reviews, number: CRD42017074560.
ing data were extracted: (1) mean  SD social scores at baseline and
follow-up and (2) CHR sample size per treatment group. Confidence
2.2 | Search strategy
intervals (CIs) and SEs were converted to SDs using methods
A comprehensive search of the literature was undertaken in the fol- described in the Cochrane Handbook (Higgins & Green, 2011). Corre-
lowing online databases: Embase, CINAHL, PsycINFO, MEDLINE and sponding authors were contacted to retrieve additional data missing
EBM from 1951 to June 2017. No geographical or language restric- in the data extraction phase. Articles and abstracts not published in
tions were applied. Examples of database searches are reported in English were translated using the Google Translator Toolkit (Google,
Appendix S2. After duplicates were removed, 2 reviewers (M.F. and Mountain View, California, USA).
DEVOE ET AL. 3

2.5 | Risk-of-bias assessment 3.2 | Study and participant characteristics

The studies were independently evaluated for quality by 2 reviewers As outlined in Table 1, most studies were performed in North America
(M.F. and P.T.) and reconciliation for any conflicts was resolved by a (n = 11), followed by Europe (n = 4), Asia (n = 2), Australia (n = 1) and
third reviewer (D.D.). The randomized-controlled trials (RCTs) were 1 multinational study. There were 11 RCTs, 5 open label studies, 2 natu-
assessed for quality using the Cochrane risk of bias assessment tool ralistic studies and 1 randomized discontinuity study design. Further,
(Higgins & Green, 2011), while the nonrandomized studies (ie, open- there was a total of 1513 CHR patients in total across all included stud-
label or naturalistic design) were examined using the Newcastle- ies, ranging from sample sizes of 6 to 304 participants in individual stud-
Ottawa Scale (NOS) (Wells et al., 2000). The Cochrane risk of bias ies. The mean age was 20.5 years and percentage of males was 47%.
assessment tool addresses biases present in randomized studies
focusing on: selection bias, performance bias, detection bias, attrition 3.3 | Features of treatment interventions and
bias, reporting bias and other biases. The NOS assesses the quality of controls
nonrandomized studies including: the selection of the study groups,
The most common interventions to report on social functioning were
comparability between groups and ascertainment of all study expo-
cognitive remediation studies (n =5) followed by antipsychotics (n = 4)
sures and outcomes. Studies were not excluded from this systematic
and CBT (n = 4), family-based therapy (n = 3), omega-3 (n = 2) and
review based on the quality assessments.
heart rate variability biofeedback (n = 1). The average intervention
duration was 24.9 weeks. The most common measure of social func-
2.6 | Data synthesis and analysis tioning was the GFS (n = 8) followed by the SOFAS (n = 5), SFS (n = 4)
Due to the differences in social functioning scales, effect sizes were and the SAS (n = 2). Table 2 outlines all social functioning scales with-
calculated as Hedges g and reported as the standardized mean details describing the scales and how baseline scores compare
difference (SMD) (Higgins, Green, & Scholten, 2008). Due to expected between healthy controls and CHR study populations. From a visual
heterogeneity between studies, DerSimonian and Laird (1986) assessment, it appeared that CHR samples experience noticeably
random-effects meta-analyses were performed on eligible RCT studies poorer functioning than healthy controls at the baseline (prior to any
to estimate pooled SMDs and 95% CIs for each treatment type sepa- intervention).

rately (ie, CBT, Omega-3, Cognitive Remediation). Meta-analyses were

stratified by time for each treatment type when available (eg, 6-, 12- 3.4 | Risk-of-bias assessment
or 18-month follow-up). If applicable, scales were inverted to match
Out of the 11 RCT studies assessed with the Cochrane risk-of-bias
the direction of effect of the majority of scales. Statistical heterogene-
assessment (Figure 2) random sequence generation bias was low.
ity was examined using the I2 statistic, with I2 ≥50% deemed moder-
However, 7 studies were at high risk-of-bias for performance bias due
ate and I2 ≥ 75% high heterogeneity. RevMan version 5 was used for
to the lack of blinding of participants and personnel and 3 studies had
all analyses with an α < .05 for statistical significance. All pooled
a high risk-of-bias for attrition bias and reporting bias based on the
SMDs effect sizes were assessed as a small effect (0.2), medium effect
authors’ risk-of-bias ratings. For observational studies, the NOS
(0.5) and large effect (0.8) (Cohen, 1988). described risk-of-bias highlighting a low risk-of-bias for outcomes not
being present and outcome follow-up length (Appendix S3).

3 | RESULTS
3.5 | Impact of treatment on social functioning
3.1 | Search yield Meta-analyses were only feasible to conduct for CBT, omega-3 and
cognitive remediation (CR) intervention studies due to similar follow-
After duplicate references were removed a total of 5608 abstracts
up time-points (eg, 6 months) and having at least 2 studies available
and titles were screened, of these, 190 were retrieved and reviewed
for comparison (Figure 3). Descriptions of observational studies not
in full text. The level of agreement between the 2 reviewers was high
included in meta-analyses are provided below in Table 1.
(κ = 0.90). Altogether, 19 studies (Addington, Epstein, et al., 2011;
Bechdolf et al., 2007; Cadenhead et al., 2017; Choi et al., 2017; Holzer
et al., 2014; Hooker et al., 2014; Ising et al., 2016; Landa et al., 2016; 3.6 | Psychosocial therapies
Loewy et al., 2016; McAusland & Addington, 2016; McFarlane et al., In pairwise meta-analyses, CBT interventions were not associated with
2015; McGorry et al., 2017; Miklowitz et al., 2014; Morita et al., a significant improvement in social functioning compared to controls at
2014; Piskulic, Barbato, Liu, & Addington, 2015; Rybakowski, 6 months (SMD, 0.06; 95% CI = −0.35, 0.46; I2 = 44%; P = .78, 3 stud-
Drozdz, & Borkowska, 2003; Shim et al., 2008; Stain et al., 2016; ies, N = 239), 12 months (SMD, –0.15; 95% CI = –0.38, 0.08; I2 = 6%;
Woods et al., 2007) met the inclusion criteria for this systematic P = .20, 4 studies, N = 321) and 18 months (SMD, 0.20; 95%
review, as noted in Figure 1. One study was translated using the Goo- CI = −0.10, 0.50; I2 = 0%; P = .20, 2 studies, N = 168). Nor did the
gle Translator Kit from Hungarian to English but was not eligible for results indicate a consistent trend across time-points. Heterogeneity
inclusion in this review due to having no measure of social functioning within the CBT-pooled estimates remained relatively low (6 months
(Keri, Kelemen, & Janka, 2006). I2 = 44%; 12 months I2 = 6% and 18 months I2 = 0%).
4 DEVOE ET AL.

Identification
Records identified through Additional records identified
database searching through other sources
(n = 7269) (n = 7)

Records after duplicates removed

Screening (n = 5608)

Records screened Records excluded

(n = 5608) (n = 5418)

Full-text articles assessed

Eligibility

Full-text articles excluded

for eligibility (n = 171)
(n = 190) -75 no social outcomes
-32 no intervention
-24 incorrect study design
-18 protocol/methods
-14 duplicate study population
-8 different patient population
Studies included in
qualitative synthesis
(n = 19)
Included

Studies included in meta-

analysis
(n = 9)

FIGURE 1 PRISMA flow diagram of social functioning systematic search and included studies

Family therapy could not be combined in a meta-analysis due to 2 demonstrated statistically significant improvements in social func-
having only 1 RCT and 2 open label observational studies. In the RCT tioning, specially risperidone (n = 8) and varied antipsychotics
(N = 129) both the control and the family therapy groups demon- (n = 27) (Rybakowski et al., 2003; Shim et al., 2008) from baseline to
strated a significant improvement in social functioning but there was follow-up and two studies, aripiprazole (n = 15) and varied antipsy-
no differences between groups (Miklowitz et al., 2014). In the obser- chotics study (n = 46) did not (Morita et al., 2014; Woods et
vational studies, one small study (N = 6) showed that family therapy al., 2007).
improved social functioning significantly (Landa et al., 2016) while the
other family study (N = 205) found changes in social functioning in
3.9 | Other interventions
the CHR group at 24-months (McFarlane et al., 2015).
In pairwise meta-analyses, omega-3 intervention studies were not
associated with a significant improvement in social functioning com-
3.7 | Cognitive remediation
pared to controls at 6 months (SMD, 0.01; 95% CI = –0.21, 0.24;
Three CR intervention studies (Choi et al., 2017; Loewy et al., 2016; I2 = 0%; P = .91, 2 studies, N = 309) and 12 months (SMD, –0.08;
Piskulic et al., 2015) were pooled together at 2- to 3-month (post-CR 95% CI = –0.33, 0.17; I2 = 0%; P = .51, 2 studies, N = 252). In one
scores) follow-up. In pairwise meta-analyses, CR interventions open label heartrate variability study (N = 20) no significant
was not associated with a significant improvement in social function- changes in social functioning were demonstrated (McAusland &
ing compared to controls at 2 to 3 months (SMD, 0.13 and 95% Addington, 2016).
CI = –0.18, 0.43; I2 = 0%; P = .41, 3 studies, N = 170). In addition,
one CR observational study (N = 14) found no improvement in social
functioning (Hooker et al., 2014) and one RCT with a mixed sample 4 | DI SCU SSION
(ie, CHR and FEP) found a significant improvement in both the control
and the CR groups from baseline to follow-up (Holzer et al., 2014). In summary, this systematic review and meta-analysis compared the
impact of various interventions on improving social functioning in
CHR samples. Pairwise meta-analyses demonstrated that CBT, CR and
3.8 | Antipsychotics omega-3 were not associated with a significant improvement in social
Antipsychotics could not be combined in a meta-analysis due to hav- functioning compared to controls at any time during the course of
ing no RCTs. However, in 4 open label noncontrolled studies treatment. Out of the 19 studies that met the inclusion criteria for this
TABLE 1 Intervention studies including social functioning outcomes (n = 19)

CHR patients Effect of

DEVOE ET AL.

Treatment Number treatment Included

Study Sample duration of CHR N/% Male Main study on social Social in
Author, year Country design Intervention Control size (wk) centres criterion converted Age  SD (N, %) outcome(s) functioning measure analysis

Cognitive behavioural
therapy (n = 4)
Addington, Epstein, Canada RCT CBT Supportive 51 24 2 SIPS 51/5.9% CBT: 20.8  4.5 36 (71) Conversion to No changes SFS PW
et al. (2011) therapy Supportive: psychosis
21.1  3.7
Bechdolf et al. (2007) Germany RCT CBT Supportive 128 52 4 EIPS 128/NR CBT: 25.2  5.3 75 (NR) Social Both groups SAS PW
therapy Supportive: adjustment significantly
26.4  5.7 improved
Ising et al. (2016) Netherlands RCT CBT TAU 196 24 6 CAARMS 196/17.3% CBT: 22.7  5.6 97 (49) Conversion to No changes SOFAS PW
Van Der Gaag et al. TAU: 22.6  5.4 psychosis
(2012)
Stain et al. (2016) Australia, RCT CBT NDRL 57 24 2 CAARMS 57/5.3% CBT: 16.2  2.7 23 (40) Conversion to No changes SOFAS PW
New Zealand TAU: 16.5  3.2 psychosis
Cognitive remediation
(n = 5)
Choi et al. (2017) USA RCT CR Active 62 8 2 SIPS 62/8.1% 18.4  3.7 30 (48) Social Significantly SAS-SR PW
control functioning improved
compared to
control
Holzer et al. (2014) Switzerland RCT CR Computer 32 8 1 SIPS 12/NR NR NR Cognitive Both groups SOFAS —
games functioning significantly
improved
Hooker et al. (2014) USA Open label CR Computer 28 8 4 SIPS 14/NR 21.9  4.2 7 (50) Cognitive, global, No changes GFS —
games role and social
functioning
Loewy et al. (2016) USA RCT CR Computer 83 8 1 SIPS 83/NR CRT: 17.8  3.1 42 (51) Cognitive No changes GFS PW
games Control: functioning
18.7  4.6
Piskulic et al. (2015) Canada RCT CR Computer 32 12 1 SIPS 32/NR CRT: 19.7  5.7 21 (66) Cognitive No changes GFS PW
games Games: functioning
17.5  3.5
Family-based therapy
(n = 3)
Landa et al. (2016) USA Open label Group-and-family- None 21 15 NR CAARMS 6/NR 19.5  1.5 2 (33) Feasibility Significantly SOFAS —
based CBT improved
McFarlane et al. (2015) USA RDD FACT None 337 104 6 SIPS 205/6.3% 16.4  3.3 116 (57) Conversion to Significantly GFS —
psychosis improved at
24 months
5
 6

TABLE 1 (Continued)

CHR patients Effect of

Treatment Number treatment Included
Study Sample duration of CHR N/% Male Main study on social Social in
Author, year Country design Intervention Control size (wk) centres criterion converted Age  SD (N, %) outcome(s) functioning measure analysis

Miklowitz et al. (2014) USA + Canada RCT Family-focused Enhanced 129 24 8 SIPS 129/4.7% 17.4  4.1 74 (57) Positive and Both groups GFS —
therapy care negative significantly
symptoms improved
Antipsychotics (n = 4)
Morita et al. (2014) Japan Naturalistic Supportive None 46 52 1 SIPS 46/6.5% 23.5  6.6 13 (28) Quality of life No changes SFS —
therapy and/or
psychotropic
medication
Rybakowski et al. (2003) Poland Naturalistic Risperidone: None 8 Varied 1 NR 8/12.5% 27.1  10.6 4 (50) Social Significantly SOFAS —
0.5-2 mg/d functioning improved
Shim et al. (2008) Korea Open label AP: varied None 27 Varied 1 CAARMS 27/7.4% 21.5  4.8 16 (59) Prodromal Significantly SFS —
symptoms improved
Woods et al. (2007) USA Open label Aripiprazole: None 15 8 1 SIPS 15/0% 17.1  5.5 8 (53) Total symptoms No changes SFS —
5-30 mg/d
Other interventions (n = 3)
Cadenhead et al. (2017) USA, Canada RCT Omega-3:740 Placebo 127 24 8 SIPS 118/5.9% 18.8  NR 71 (56) Conversion to No changes GFS PW
mg EPA, psychosis
400 mg DHA/d
McGorry et al. (2017) Multi-national RCT Omega-3 w-3 Placebo + 304 24 10 CAARMS 304/10.5% 19.1  4.6 139 (46) Conversion to Both groups GFS PW
PUFA: CBCM psychosis significantly
1.4 g/d + improved
CBCM
McAusland and Canada Open label HRV biofeedback None 20 4 1 SIPS 20/NR 16.7  2.3 6 (30) Anxiety, No changes GFS —
Addington (2016) training depression
and
functioning

Abbreviations: AP, anti-psychotics; CAARMS, comprehensive assessment of at-risk mental states; CBCM,cognitive-behavioural case management; CBT, cognitive behavioural therapy; CR, cognitive remediation; DHA,
docosahexaenoic acid; EIPS, early initial prodromal state; EPA, eicosapentaenoic acid; FACT, family-aided community therapy; GFS, global functioning social; HRV, heart rate variability; NDRL, non directive reflective lis-
tening; NR, not reported; PUFA, polyunsaturated fatty acid; PW, pairwise; RCT, randomized controlled trial; RDD, regression discontinuity design; SAS, social adjustment scale; SFS, social functioning scale; SIPS, struc-
tured interview for prodromal symptoms; SOFAS, social and occupational functioning assessment scale; TAU, treatment as usual.
DEVOE ET AL.
DEVOE ET AL. 7

TABLE 2 Description of social functioning scales

Baseline scores (meanSD)

Social functioning scales Description Healthy controls CHR
Global functioning: Social (GFS) • Range: 1-10 8.63  0.97 5.94  1.53
• Lower scores indicate poorer functioning (Carrión et al., 2011) (Carrión et al., 2011)
• Assesses quantity and quality of peer relationships,
level of peer conflict, age-appropriate intimate
relationships and involvement with family
members.
Social adjustment scale: self report • Range: 1-5 1.59  0.33 CBT: 3.7  0.85;
(SAS-SR) • Higher scores indicate poorer functioning (Weissman et al., SC: 3.5  1.0
• Global measure derived from 3 to 6 role area 1978) (Bechdolf et al., 2007)
scores: Occupational performance, social and
leisure functioning, and physical well-being,
relationship with family/marital/ children, domestic
life, and financial situation
Social functioning scale (SFS) • Normalized to mean = 100 SD = 15 161.65  20.23 125.29  22.77
• Lower scores indicate poorer functioning (Addington et al., (Addington et al.,
• Global measure derived from 7 sub-scores: 2008) 2008)
withdrawal/social engagement, interpersonal
communication, independence-performance,
independence competence, recreation, prosocial
and employment/occupation
Social and occupational functioning • Range: 0-100 88.83  4.09 56.25  10.02
assessments scale (SOFAS) • Lower scores indicate poorer functioning (Glenthøj et al., 2016) (Glenthøj et al., 2016)
• Derived from the global assessment scale and
specifically focuses on social and occupational
functioning not directly related to the individuals
psychological functioning

Abbreviations: CBT, cognitive behavioural therapy; SC, supportive counselling.

review only one CR study reported a significant improvement in social (Miklowitz et al., 2014). This is in contrast to a Cochrane review and
functioning compared to controls. meta-analysis that demonstrated that family therapy had a significant
As previously stated, psychosocial therapies (ie, CBT, family ther- impact on social functioning compared to controls in patients with
apy) had no impact on social functioning compared to controls in both schizophrenia (Pharoah, Mari, Rathbone, & Wong, 2010). A limitation
pairwise meta-analyses and individual studies. In contrast, one meta- of the Cochrane review is that it does not specify whether or not the
analysis in schizophrenia patients demonstrated a strong effect for treatments utilized individual or multiple family groups making it diffi-
several psychosocial therapies (ie, psychoeducation, multicomponent cult to know what elements of family therapy impact social function-
structured psychotherapies, art therapy, community care) compared ing (Pharoah et al., 2010).
to controls in improving social functioning (De De Silva, Cooper, Li, Next, CR studies had no impact on social functioning compared
Lund, & Patel, 2013). Interestingly, in 3 of the 4 CBT studies included to controls in our meta-analyses. In contrast, a meta-analysis in early
in this current review, the CBT interventions comprised of several fac- psychosis samples looking at CR studies showed a significant effect
tors related to social functioning such as social isolation (Addington, on functioning and a trend for impacting social cognition favouring
Epstein, et al., 2011; Stain et al., 2016), social perceptions and social the CR groups (Revell, Neill, Harte, Khan, & Drake, 2015). To our
skills training (Bechdolf et al., 2007) but demonstrated little to no knowledge the CR studies included in this current review did not
impact on social functioning. However, what remains unclear is how employ any social skills training in addition to the computer based
much the CBT treatments focused on ameliorating social functioning training, which in combination has been demonstrated to increase the
relative to other primary outcomes such as decreasing attenuated psy- impact of CR on social outcomes in schizophrenia trials (Hogarty et al.,
chotic symptoms and circumventing transition to psychosis. 2004; Spaulding, Reed, Sullivan, Richardson, & Weiler, 1999). Lastly, it
In regard to family therapy in the current review, the one RCT has been suggested that processing speed may be an important cogni-
asserted that their family treatment was designed to increase social tive competent to target impaired social functioning (Choi et al., 2017;
engagement (Miklowitz et al., 2014), while the 2 observational family Kelleher et al., 2013). All CR studies in this review reported that the
studies had elements in the family therapy group designed to improve CR intervention groups employed some element that targeted proces-
social environment with family, peers, work, in addition to reducing sing speeds specifically, but only one CR study improved social func-
social isolation (Landa et al., 2016; McFarlane et al., 2015). Nonethe- tioning compared to controls at 4 month follow-up (Choi et al., 2017).
less, even though social interactions are inherent in the design of fam- Antipsychotic studies included in this review utilized poor study
ily therapy, two family studies demonstrated an impact on social designs (ie, open label noncontrolled studies). Risperidone (Rybakowski
functioning from baseline to follow-up in the absence of a control et al., 2003) and varied antipsychotics (Shim et al., 2008) improved social
group (Landa et al., 2016; McFarlane et al., 2015), while there was no functioning in very small samples (n = 8 vs n = 27) from baseline to
difference between the family therapy and control group in the RCT follow-up, while one naturalistic study with varied antipsychotics usage
8 DEVOE ET AL.

FIGURE 2 A, Risk-of-bias graph for RCTs: review authors' judgement about each risk-of-bias item as percentages across all studies. B, Risk-of-
bias summary: review authors' judgements about each risk-of-bias item for each included study

amongst participants (ie, aripiprazole, quetiapine, perospirone, risperi- without treatment. Table 2 shows that CHR samples are certainly
done) and one aripiprazole study demonstrated no improvement in more socially impaired than healthy controls; however, since con-
social functioning from baseline to follow-up (Morita et al., 2014; trol and treatment groups rarely differ at follow-up in CHR studies
Woods et al., 2007). A more recent meta-analysis demonstrated that on social functioning it may be that there are not enough elements
antipsychotics had a significant impact in improving social functioning in in the treatment designs to target poor social functioning. It may
schizophrenia patients (Leucht et al., 2017). Certainly, high-quality evi- also be the case that CHR studies need to examine the impact of
dence is needed in the form of RCTs in CHR to determine if antipsy- treatment by isolating those with persistent deficits in social func-
chotics impact social functioning at all. tioning, which may in turn demonstrate similar results to more
To the best of our knowledge there is no cumulative evidence in chronic groups.
the literature to suggest that omega-3 or biofeedback has any impact
on social functioning in CHR, first episode patients or schizophrenia
4.1 | Strengths and limitations
patients, which is in agreement with our current review.
In this review there were no changes in social functioning in This systematic review included 19 studies that examined several types
CHR samples following the intervention compared to controls, of interventions on social functioning with more than 1500 participants
with the exception of one CR study. This is in contrast to first epi- at CHR for psychosis. We searched numerous databases, reviewed stud-
sode and schizophrenia studies that have demonstrated significant ies in duplicate and followed PRISMA guidelines. To our knowledge this
changes in social functioning relative to controls. This discrepancy is the first systematic review and meta-analysis that explicitly examines
may be due the heterogeneous nature of the CHR population such social functioning and the impact of all interventions in CHR to date.
as differences in attenuated psychotic symptom severity, age and However, this review has several significant limitations.
duration of impairment. In this review transition remained rela- First, the sum of a systematic review is only as good as the quality
tively low across most studies as can be seen in Table 1 and thus it of individual studies included in a review. Many studies included in this
may be that a significant number of individuals present with only review had poor study designs such as open labelled noncontrolled stud-
transient deficits in social functioning that ameliorate overtime ies; however, the majority of studies were RCTs and all studies included
DEVOE ET AL. 9

1) Cognitive behavioral therapy

a. 6-months

b. 12-months

c. 18-months

2) Omega-3
a. 6-months

b. 12-months

3) Cognitive remediation
a. 2-3 months

FIGURE 3 (1) Cognitive behavioural therapy: (a) 6 months; (b) 12 months; (c) 18 months. (2) Omega-3: (a) 6 months; (b) 12 months. (3) Cognitive
remediation: (a) 2-3 months (post-CR scores)

in the meta-analyses were RCTs. The RCTs included in this review suf- difficult to blind participants and impossible to blind therapist of the ther-
fered from several important biases such as blinding of participants and apy provided). Only one study had a high risk of bias for blinding of out-
personnel. This was the case in all the CBT RTCs included in this review, come assessments, which may be a more important and relevant
which is characteristic of psychosocial therapies study designs (eg, it is outcome to assess risk of bias in psychosocial studies.
10 DEVOE ET AL.

Second, we amalgamated several social functioning scales, which to target and improve social functioning in youth at CHR for
may have important implications when considering the results from the psychosis.
current systematic review. Most studies in the meta-analysis utilized
the GFS scale to measure social functioning followed by either the SAS
ACKNOWLEDGEMENTS
or the SOFAS. While the GFS and SAS both specifically measure social
functioning, the SOFAS incorporates a measurement of occupational This work was supported by NIH grant RO1MH105178 awarded to
functioning, which may influence the scores due to deficits in occupa- Dr. Jean Addington and by the Alberta Innovates Graduate Student-

tional functioning and not due to poor social functioning. However, ship awarded to Dan Devoe.

both studies included in the current meta-analyses that utilized the

ORCID
SOFAS (Ising et al., 2016; Stain et al., 2016) specifically stated the
SOFAS was utilized as a measurement of social functioning and not Daniel J. Devoe http://orcid.org/0000-0003-4931-0205
necessarily occupational functioning. Investigators may wish to utilize Jean Addington http://orcid.org/0000-0002-8298-0756
scales specifically geared towards measuring social functioning (ie, SAS,
GFS, and SFS) vs the SOFAS and the GAF which have elements of RE FE RE NC ES
occupational functioning to further elucidate the impact of treatment Addington, J., Cornblatt, B. A., Cadenhead, K. S., Cannon, T. D.,
on social functioning in future trials. Another issue that arises from the Mcglashan, T. H., Perkins, D. O., … Heinssen, R. (2011). At clinical high
risk for psychosis: Outcome for nonconverters. American Journal of
different scales is their sensitivity and ability of these scales to detect a
Psychiatry, 168, 800–805.
difference. For example, the GFS measures peer relationships; however, Addington, J., Epstein, I., Liu, L., French, P., Boydell, K. M., & Zipursky, R. B.
at the end of treatment the question remains whether or not partici- (2011). A randomized controlled trial of cognitive behavioral therapy
for individuals at clinical high risk of psychosis. Schizophrenia Research,
pants have had enough time to actually establish close friends. The SFS
125, 54–61.
was designed for patients with schizophrenia and the SAS has primarily Addington, J., Liu, L., Perkins, D. O., Carrion, R. E., Keefe, R. S. E., &
been tested in patients with schizophrenia, however, patients with Woods, S. W. (2017). The role of cognition and social functioning as
predictors in the transition to psychosis for youth with attenuated psy-
schizophrenia are often older than CHR participants and have elements
chotic symptoms. Schizophrenia Bulletin, 43, 57–63.
in their designs that may not be as applicable to CHR participants such Addington, J., Penn, D., Woods, S. W., Addington, D., & Perkins, D. O.
as independence-performance and financial situation. Indeed, (2008). Social functioning in individuals at clinical high risk for psycho-
sis. Schizophrenia Research, 99, 119–124.
researchers should consider if the scale they choose to employ is sensi-
Bechdolf, A., Wagner, M., Veith, V., Ruhrmann, S., Pukrop, R.,
tive, detailed, and relevant (eg, the advent of social media) to detect a Brockhaus-Dumke, A., … Klosterkotter, J. (2007). Randomized con-
change in social functioning in CHR youth. trolled multicentre trial of cognitive behaviour therapy in the early ini-
tial prodromal state: Effects on social adjustment post treatment. Early
Fourth, unfortunately due to the limit amount of studies, discrep-
Intervention in Psychiatry, 1, 71–78.
ancies in study design, and the lack of control groups many studies Birchwood, M., Smith, J., Cochrane, R., Wetton, S., & Copestake, S. (1990).
could not be pooled in a meta-analysis such as family therapy, antipsy- The social functioning scale. The development and validation of a new
scale of social adjustment for use in family intervention programmes
chotics and biofeedback. More high-quality studies are required to
with schizophrenic patients. The British Journal of Psychiatry, 157,
adequately understand the impact of these interventions on social 853–859.
functioning. Cadenhead, K., Addington, J., Cannon, T., Cornblatt, B., Mathalon, D.,
Mcglashan, T., … Woods, S. (2017). 23. Omega-3 fatty acid versus pla-
cebo in a clinical high-risk sample from the North American Prodrome
4.2 | Directions for future research Longitudinal Studies (NAPLS) consortium. Schizophrenia Bulletin, 43,
S16–S16.
Although the results of the systematic review and meta-analysis are Cannon, T. D., Cadenhead, K., Cornblatt, B., Woods, S. W., Addington, J.,
Walker, E., … Heinssen, R. (2008). Prediction of psychosis in youth at
null they may benefit researchers in several areas with future
high clinical risk: A multisite longitudinal study in North America.
research. First, few RCTs to date in CHR have examined the impact of Archives of General Psychiatry, 65, 28–37.
any intervention on social functioning as the primary outcome. Thus, Carrión, R. E., Goldberg, T. E., McLaughlin, D., Auther, A. M.,
Correll, C. U., & Cornblatt, B. A. (2011). Impact of neurocognition on
future trials may want to address social functioning by including ele-
social and role functioning in individuals at clinical high risk for psycho-
ments in the design that are primarily geared towards impacting social sis. The American Journal of Psychiatry, 168(8), 806–813.
functioning (ie, social skills training). Second, future studies may wish Carrion, R. E., Mclaughlin, D., Goldberg, T., Olvet, D., Auther, A., &
Cornblatt, B. (2012). Prediction of functional outcome in individuals at
to investigate what treatments impact social functioning in CHR par-
clinical highrisk for psychosis. Early Intervention in Psychiatry, 6, 4.
ticipants that have persistent impairments in social functioning, which Carrion, R. E., Mclaughlin, D., Goldberg, T. E., Auther, A. M., Olsen, R. H.,
was not addressed in any of the studies included in this review. Olvet, D. M., … Cornblatt, B. A. (2013). Prediction of functional out-
come in individuals at clinical high risk for psychosis. JAMA Psychiatry,
70, 1133–1142.
Choi, J., Corcoran, C. M., Fiszdon, J. M., Stevens, M., Javitt, D. C.,
5 | CO NC LUSIO N Deasy, M., … Pearlson, G. D. (2017). Pupillometer-based neurofeed-
back cognitive training to improve processing speed and social func-
This systematic review and meta-analysis demonstrated that no treat- tioning in individuals at clinical high risk for psychosis. Psychiatric
Rehabilitation Journal, 40, 33–42.
ment significantly improved social functioning in youth at CHR. Future Cohen, J. (1988). Statistical power analysis for the behavioral sciences
high-quality randomized control trials are required that are designed (pp. 20–26). Hillsdale, NJ: Lawrence Earlbaum Associates.
DEVOE ET AL. 11

Corcoran, C., Kimhy, D., Parrilla-Escobar, M., Cressman, V., Stanford, A., Keri, S., Kelemen, O., & Janka, Z. (2006). Therapy of mental states at high
Thompson, J., … Malaspina, D. (2011). The relationship of social func- risk for psychosis: Preliminary results from Hungary. Orvosi Hetilap,
tion to depressive and negative symptoms in individuals at clinical high 147, 201–204.
risk for psychosis. Psychological Medicine, 41, 251–261. Kim, K. R., Song, Y. Y., Park, J. Y., Lee, E. H., Lee, M., Lee, S. Y., …
Cornblatt, B. A., Auther, A. M., Niendam, T., Smith, C. W., Zinberg, J., Kwon, J. S. (2013). The relationship between psychosocial functioning
Bearden, C. E., & Cannon, T. D. (2007). Preliminary findings for two and resilience and negative symptoms in individuals at ultra-high risk
new measures of social and role functioning in the prodromal phase of for psychosis. The Australian and New Zealand Journal of Psychiatry, 47,
schizophrenia. Schizophrenia Bulletin, 33, 688–702. 762–771.
Cornblatt, B. A., Carrion, R. E., Addington, J., Seidman, L., Walker, E. F., Kimhy, D., Gill, K. E., Brucato, G., Vakhrusheva, J., Arndt, L., Gross, J. J., &
Cannon, T. D., … Lencz, T. (2012). Risk factors for psychosis: Impaired Girgis, R. R. (2016). The impact of emotion awareness and regulation
social and role functioning. Schizophrenia Bulletin, 38, 1247–1257. on social functioning in individuals at clinical high risk for psychosis.
Cornblatt, B. A., Lencz, T., Smith, C. W., Correll, C. U., Auther, A. M., & Psychological Medicine, 46, 2907–2918.
Nakayama, E. (2003). The schizophrenia prodrome revisited: A neuro- Landa, Y., Mueser, K. T., Wyka, K. E., Shreck, E., Jespersen, R.,
developmental perspective. Schizophrenia Bulletin, 29, 633–651. Jacobs, M. A., … Walkup, J. T. (2016). Development of a group and
Cotter, J., Drake, R. J., Bucci, S., Firth, J., Edge, D., & Yung, A. R. (2014). family-based cognitive behavioural therapy program for youth at risk
What drives poor functioning in the at-risk mental state? A systematic for psychosis. Early Intervention in Psychiatry, 10, 511–521.
review. Schizophrenia Research, 159, 267–277. Lee, S. J., Kim, K. R., Lee, S. Y., & An, S. K. (2017). Impaired social and role
De Silva, M. J., Cooper, S., Li, H. L., Lund, C., & Patel, V. (2013). Effect of function in ultra-high risk for psychosis and first-episode schizophre-
psychosocial interventions on social functioning in depression and nia: Its relations with negative symptoms. Psychiatry Investigation, 14,
schizophrenia: Meta-analysis. The British Journal of Psychiatry, 202, 186–192.
253–260. Leucht, S., Leucht, C., Huhn, M., Chaimani, A., Mavridis, D., Helfer, B., …
DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Con- Davis, J. M. (2017). Sixty years of placebo-controlled antipsychotic
trolled Clinical Trials, 7, 177–188. drug trials in acute schizophrenia: Systematic review, Bayesian
Glenthøj, L. B., Fagerlund, B., Hjorthøj, C., Jepsen, J. R. M., Bak, N., meta-analysis, and meta-regression of efficacy predictors. The Ameri-
Kristensen, T. D., … Nordentoft, M. (2016). Social cognition in patients can Journal of Psychiatry, 174, 927–942.
at ultra-high risk for psychosis: What is the relation to social skills and Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P. C.,
functioning? Schizophrenia Research: Cognition., 5, 21–27. Ioannidis, J. P., … Moher, D. (2009). The PRISMA statement for report-
Green, M. F., Hellemann, G., Horan, W. P., Lee, J., & Wynn, J. K. (2012). ing systematic reviews and meta-analyses of studies that evaluate
healthcare interventions: Explanation and elaboration. BMJ, 339,
From perception to functional outcome in schizophrenia: Modeling the
b2700.
role of ability and motivation. Archives of General Psychiatry, 69,
Loewy, R., Fisher, M., Schlosser, D. A., Biagianti, B., Stuart, B.,
1216–1224.
Mathalon, D. H., & Vinogradov, S. (2016). Intensive auditory cognitive
Higgins, J. P. T., & Green, S. (2011). Cochrane handbook for systematic
training improves verbal memory in adolescents and young adults at
reviews of interventions. Southern Gate, Chichester, West Sussex,
clinical high risk for psychosis. Schizophrenia Bulletin, 42(Suppl. 1),
England: Wiley.
S118–S126.
Higgins, J. P. T., Green, S., & Scholten, R. J. P. M. (2008). Maintaining
Maat, A., Cahn, W., Gijsman, H. J., Hovens, J. E., Kahn, R. S., & Aleman, A.
reviews: Updates, amendments and feedback. In Cochrane handbook
(2014). Open, randomized trial of the effects of aripiprazole versus ris-
for systematic reviews of interventions. Southern Gate, Chichester, West
peridone on social cognition in schizophrenia. European Neuropsycho-
Sussex, England: John Wiley & Sons, Ltd.
pharmacology, 24, 575–584.
Hogarty, G. E., Flesher, S., Ulrich, R., Carter, M., Greenwald, D.,
Mavergames, C. 2013. Covidence (Systematic Review Software). Freiburg,
Pogue-Geile, M., … Zoretich, R. (2004). Cognitive enhancement ther-
Germany.
apy for schizophrenia: Effects of a 2-year randomized trial on cognition
McAusland, L., & Addington, J. (2016). Biofeedback to treat anxiety in
and behavior. Archives of General Psychiatry, 61, 866–876.
young people at clinical high risk for developing psychosis. Early Inter-
Holzer, L., Urben, S., Passini, C. M., Jaugey, L., Herzog, M. H., Halfon, O., &
vention in Psychiatry, 49, 43–49.
Pihet, S. (2014). A randomized controlled trial of the effectiveness of
McFarlane, W. R., Levin, B., Travis, L., Lucas, F., Lynch, S., Verdi, M., …
computer-assisted cognitive remediation (CACR) in adolescents with
Spring, E. (2015). Clinical and functional outcomes after 2 years in the
psychosis or at high risk of psychosis. Behavioural and Cognitive Psycho- early detection and intervention for the prevention of psychosis multi-
therapy, 42, 421–434. site effectiveness trial. Schizophrenia Bulletin, 41, 30–43.
Hooker, C. I., Carol, E. E., Eisenstein, T., Yin, H., Lincoln, S. H., Tully, L. M., McGorry, P. D., Nelson, B., Markulev, C., Yuen, H. P., Schafer, M. R.,
… Seidman, L. J. (2014). A pilot study of cognitive training in clinical Mossaheb, N., … Amminger, G. (2017). Effect of omega-3 polyunsatu-
high risk for psychosis: Initial evidence of cognitive benefit. Schizophre- rated fatty acids in young people at ultrahigh risk for psychotic disor-
nia Research, 157, 314–316. ders: The NEURAPRO randomized clinical trial. JAMA Psychiatry, 74,
Ising, H. K., Kraan, T. C., Rietdijk, J., Dragt, S., Klaassen, R. M., Boonstra, N., 19–27.
… Van Der Gaag, M. (2016). Four-year follow-up of cognitive behav- Meyer, E. C., Carrion, R. E., Cornblatt, B. A., Addington, J.,
ioral therapy in persons at ultra-high risk for developing psychosis: The Cadenhead, K. S., Cannon, T. D., … Seidman, L. J. (2014). The relation-
Dutch Early Detection Intervention Evaluation (EDIE-NL) trial. Schizo- ship of neurocognition and negative symptoms to social and role func-
phrenia Bulletin, 42, 1243–1252. tioning over time in individuals at clinical high risk in the first phase of
Jang, J. H., Shin, N. Y., Shim, G., Park, H. Y., Kim, E., Jang, G. E., … the north American Prodrome longitudinal study. Schizophrenia Bulletin,
Kwon, J. S. (2011). Longitudinal patterns of social functioning and con- 40, 1452–1461.
version to psychosis in subjects at ultra-high risk. The Australian and Miklowitz, D. J., O'brien, M. P., Schlosser, D. A., Addington, J.,
New Zealand Journal of Psychiatry, 45, 763–770. Candan, K. A., Marshall, C., … Cannon, T. D. (2014). Family-focused
Kane, J. M., Robinson, D. G., Schooler, N. R., Mueser, K. T., Penn, D. L., treatment for adolescents and young adults at high risk for psychosis:
Rosenheck, R. A., … Heinssen, R. K. (2016). Comprehensive versus Results of a randomized trial. Journal of the American Academy of Child &
usual Community Care for First-Episode Psychosis: 2-year outcomes Adolescent Psychiatry, 53, 848–858.
from the NIMH RAISE early treatment program. The American Journal Mittal, V. A., Jalbrzikowski, M., Daley, M., Roman, C., Bearden, C. E., &
of Psychiatry, 173, 362–372. Cannon, T. D. (2011). Abnormal movements are associated with poor
Kelleher, I., Murtagh, A., Clarke, M. C., Murphy, J., Rawdon, C., & psychosocial functioning in adolescents at high risk for psychosis.
Cannon, M. (2013). Neurocognitive performance of a Schizophrenia Research, 130, 164–169.
community-based sample of young people at putative ultra high risk Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred
for psychosis: Support for the processing speed hypothesis. Cognitive reporting items for systematic reviews and meta-analyses: The
Neuropsychiatry, 18, 9–25. PRISMA statement. BMJ, 339, b2535.
12 DEVOE ET AL.

Moher, D., Shamseer, L., Clarke, M., Ghersi, D., Liberati, A., Petticrew, M., Van Der Gaag, M., Nieman, D. H., Rietdijk, J., Dragt, S., Ising, H. K.,
… Stewart, L. A. (2015). Preferred reporting items for systematic Klaassen, R. M., … Linszen, D. H. (2012). Cognitive behavioral therapy
review and meta-analysis protocols (PRISMA-P) 2015 statement. Sys- for subjects at ultrahigh risk for developing psychosis: a randomized
tematic Reviews, 4(1), 1–9. controlled clinical trial. Schizophrenia Bulletin, 38(6), 1180–1188.
Morita, K., Kobayashi, H., Takeshi, K., Tsujino, N., Nemoto, T., & Van Der Gaag, M., Smit, F., Bechdolf, A., French, P., Linszen, D. H.,
Mizuno, M. (2014). Poor outcome associated with symptomatic deteri- Yung, A. R., … Cuijpers, P. (2013). Preventing a first episode of psycho-
oration among help-seeking individuals at risk for psychosis: A natural- sis: Meta-analysis of randomized controlled prevention trials of
istic follow-up study. Early Intervention in Psychiatry, 8, 24–31. 12month and longer-term follow-ups. Schizophrenia Research, 149,
Morosini, P. L., Magliano, L., Brambilla, L., Ugolini, S., & Pioli, R. (2000). 56–62.
Development, reliability and acceptability of a new version of the Velthorst, E., Fett, A. J., Reichenberg, A., Perlman, G., Van Os, J.,
DSM-IV Social and Occupational Functioning Assessment Scale Bromet, E. J., & Kotov, R. (2017). The 20-year longitudinal trajectories
(SOFAS) to assess routine social functioning. Acta Psychiatrica Scandi- of social functioning in individuals with psychotic disorders. The Ameri-
navica, 101, 323–329. can Journal of Psychiatry, 174, 1075–1085.
Pharoah, F., Mari, J., Rathbone, J., & Wong, W. (2010). Family intervention Weissman, M. M., Prusoff, B. A., Thompson, W. D., Harding, P. S., &
for schizophrenia. The Cochrane Database of Systematic Reviews, 12, Myers, J. K. (1978). Social adjustment by self-report in a community
CD000088. sample and in psychiatric outpatients. The Journal of Nervous and Men-
Piskulic, D., Barbato, M., Liu, L., & Addington, J. (2015). Pilot study of cog- tal Disease, 166, 317–326.
nitive remediation therapy on cognition in young people at clinical high Wells, G. A., Shea, B., O'connell, D., Peterson, J., Welch, V., & Tugwell, P.
risk of psychosis. Psychiatry Research, 225, 93–98. (2000). The Newcastle-Ottawa Scale (NOS) for assessing the quality of
Revell, E. R., Neill, J. C., Harte, M., Khan, Z., & Drake, R. J. (2015). A sys- nonrandomised studies in meta-analyses. Ottawa, Canada: Ottawa Hos-
tematic review and meta-analysis of cognitive remediation in early pital Research Institute.
schizophrenia. Schizophrenia Research, 168, 213–222. Woods, S. W., Tully, E. M., Walsh, B. C., Hawkins, K. A., Callahan, J. L.,
Rybakowski, J. K., Drozdz, W., & Borkowska, A. (2003). Low dose risperi- Cohen, S. J., … Mcglashan, T. H. (2007). Aripiprazole in the treatment
done in the treatment of schizophrenia-like symptoms in high-risk sub- of the psychosis prodrome: An open-label pilot study. British Journal of
jects. Journal of Clinical Psychopharmacology, 23, 674–675. Psychiatry, 191, s96–s101.
Schlosser, D. A., Campellone, T. R., Biagianti, B., Delucchi, K. L., Gard, D. E.,
Fulford, D., … Vinogradov, S. (2015). Modeling the role of negative SUP P O R T I NG I NFO R M A T I O N
symptoms in determining social functioning in individuals at clinical
high risk of psychosis. Schizophrenia Research, 169, 204–208.
Shim, G., Kang, D.-H., Choi, J.-S., Jung, M. H., Kwon, S. J., Jang, G. E., & Additional supporting information may be found online in the Sup-
Kwon, J. S. (2008). Prospective outcome of early intervention for indi- porting Information section at the end of the article.
viduals at ultra-high-risk for psychosis. Early Intervention in Psychiatry,
2, 277–284.
Spaulding, W. D., Reed, D., Sullivan, M., Richardson, C., & Weiler, M. How to cite this article: Devoe DJ, Farris MS, Townes P,
(1999). Effects of cognitive treatment in psychiatric rehabilitation.
Addington J. Interventions and social functioning in youth at
Schizophrenia Bulletin, 25, 657–676.
Stain, H. J., Bucci, S., Baker, A. L., Carr, V., Emsley, R., Halpin, S., … risk of psychosis: A systematic review and meta-analysis. Early
Startup, M. (2016). A randomised controlled trial of cognitive behaviour Intervention in Psychiatry. 2018;1–12. https://doi.org/10.
therapy versus non-directive reflective listening for young people at ultra
1111/eip.12689
high risk of developing psychosis: The detection and evaluation of psy-
chological therapy (DEPTh) trial. Schizophrenia Research, 176, 212–219.

View publication stats