SCHRES-07716; No of Pages 7

Schizophrenia Research xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Schizophrenia Research

journal homepage: www.elsevier.com/locate/schres

Clinical and cognitive correlates of unsheltered status in homeless

persons with psychotic disorders
Katiah Llerena a,b,c,⁎, Sonya Gabrielian a,b, Michael F. Green a,b
a
Department of Veterans Affairs VISN 22 Mental Illness Research, Education, and Clinical Center, Los Angeles, CA, United States
b
UCLA Semel Institute for Neuroscience & Human Behavior, David Geffen School of Medicine, United States
c
Department of Veterans Affairs VISN 21 Mental Illness Research, Education, and Clinical Center, San Francisco, CA, United States

a r t i c l e i n f o a b s t r a c t

Article history: Homeless persons with psychosis are particularly susceptible to unsheltered homelessness, which includes living
Received 12 October 2017 on the streets, in cars, and other places not meant for human habitation. Homeless persons with psychosis have
Received in revised form 22 December 2017 distinct barriers to accessing care and comprise a high-need and hard-to-serve homeless subpopulation. There-
Accepted 16 February 2018
fore, this study sought to understand unsheltered homelessness in persons with psychosis and its relationship to
Available online xxxx
cognitive impairment, clinical symptoms, and community functioning, examined both categorically and dimen-
Keywords:
sionally. This study included 76 homeless participants with a history of a psychotic diagnosis who were enrolled
Cognition in a supported housing program but had not yet received housing. This study used two different housing stability
Community functioning thresholds (literally homeless at any point vs. literally homeless N20% of days) for comparing homeless Veterans
Homelessness with psychosis living in sheltered versus unsheltered situations on cognition, clinical symptoms, and community
Neurocognition integration. Dimensional analyses also examined the relationship between percentage of days spent in
Psychosis unsheltered locations and cognition, clinical symptoms, and community integration. Sheltered and unsheltered
Schizophrenia Veterans with psychosis did not differ on clinical symptoms or community integration, but there was an incon-
sistent group difference on cognition depending on the threshold used for determining housing stability. In the
unsheltered group, cognitive deficits in overall cognition, visual learning, and social cognition were related to
more days spent in unsheltered locations. Rehabilitation efforts targeting specific cognitive deficits may be useful
to facilitate greater access to care and successful interventions in this population.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction
Homeless adults have higher rates of chronic medical and mental
health conditions, injury, and mortality rates compared to the general
population (Cheung and Hwang, 2004; Gozdzik et al., 2015; Hwang,
2000; Morrison, 2009). The homeless population is very heterogeneous,
which has led to attempts to identify meaningful subgroups of individ-
uals. One key distinction is whether homeless people live primarily in
unsheltered or sheltered situations. Unsheltered homelessness refers
to places not meant for human habitation, such as streets, abandoned
buildings, vehicles, or parks; whereas sheltered homelessness includes
emergency shelters, transitional housing programs, or safe havens
(Henry et al., 2016).
It is important to understand how sheltered and unsheltered living
conditions are associated with cognitive and clinical factors, because
persons living in unsheltered situations are particularly vulnerable and
⁎ Corresponding author at: University of California, Los Angeles & VA Greater Los
Angeles, Healthcare System, MIRECC, Bldg. 210, 11301 Wilshire Blvd., Los Angeles, CA
90073, United States.
E-mail address: kllerena@ucla.edu (K. Llerena).
may require greater outreach and more intensive health and housing
services than sheltered persons. For example, those in unsheltered situ-
ations have higher rates of medical and mental health needs (Shern et
al., 2000), are more likely to be chronically homeless (Cousineau,
1997; Montgomery et al., 2016b; O'Toole et al., 1999; Shern et al.,
2000; Tsai et al., 2014), and have higher rates of substance use disorders
and severe mental illness (Byrne et al., 2016) than their sheltered coun-
terparts. People living in unsheltered situations comprise one of the
highest-need and hardest-to-serve homeless subpopulations (Byrne et
al., 2016) and they receive more fragmented care (Lam and
Rosenheck, 1999; Levitt et al., 2009; Montgomery et al., 2016b).
Unfortunately, homeless persons with psychosis are particularly
susceptible to unsheltered situations (Foster et al., 2012) and we do
not know the specific reasons why. Psychosis may be associated with
unsheltered homelessness because people with psychosis often have
poor independent living skills, impaired cognition, and few interper-
sonal relationships (Foster et al., 2012; Kuno et al., 2000;
Stergiopoulos et al., 2010). Further, persons with psychotic disorders
often have difficulties managing even routine landlord-tenant conflicts
(Lamb and Bachrach, 2001). Moreover, they are less likely to engage
in rehabilitative services that facilitate sheltered housing (Foster et al.,

https://doi.org/10.1016/j.schres.2018.02.023
0920-9964/© 2018 Elsevier B.V. All rights reserved.

Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
2 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx
2012; Lincoln et al., 2009). Beyond these factors, the degree of psychotic
and negative symptoms might distinguish between those in sheltered
and unsheltered locations (Drake et al., 1991); for example, distrust
and paranoia may create obstacles to obtaining or maintaining housing.
Cognitive impairment is also an important factor for explaining poor
housing and community integration outcomes (Backer and Howard,
2007; Burra et al., 2009; Seidman et al., 1997; Spence et al., 2004). A re-
view of cognition in homeless adults found considerable cognitive im-
pairment in the homeless population (Depp et al., 2015). Cognitive
dysfunction may lead to difficulties navigating the social service system
and planning and prioritizing shelter and treatment. While cognitive
dysfunction is a core feature of psychosis (Green et al., 2000;
Heinrichs and Zakzanis, 1998), we know very little about cognition in
homeless unsheltered persons (Depp et al., 2015).
A key limitation in resolving these questions is that unsheltered
homelessness is usually treated as a categorical variable (Byrne et al.,
2016; Montgomery et al., 2016b; Tsemberis et al., 2007), but we do
not know if the relationships (e.g., with clinical symptoms or cognition)
are dimensional in nature. Persons experiencing homelessness often
vacillate between sheltered and unsheltered living arrangements. We
do not know if persons experiencing any days in unsheltered homeless-
ness comprise a meaningful subgroup, or if those with low percentages
of unsheltered days closely resemble those who had sheltered housing.
To address this question, it is necessary to examine the relationship be-
tween duration of unsheltered homelessness and specific cognitive,
clinical, and functional features in both a categorical and dimensional
manner. Understanding these relationships may help improve home-
less services and outcomes for homeless persons with psychosis.
The U.S. Department of Veterans Affairs (VA) has prioritized ending
Veteran homelessness (HUD, 2011; USICH, 2010) employing a Housing
First model in the U.S. Department of Housing and Urban Development
– Veterans Administration Supported Housing (HUD-VASH) program. A
Housing First approach provides homeless individuals with indepen-
dent housing and supportive services without traditional treatment-re-
lated preconditions for housing entry (Kertesz et al., 2017). The HUD-
VASH program combines a permanent housing subsidy provided by
HUD's Housing Choice Voucher Program and clinical services through
the VA. Although the program has been successful in housing homeless
Veterans (CRS, 2015; Rosenheck et al., 2003), it may have more diffi-
culty reaching homeless Veterans with psychosis who live in
unsheltered situations (Montgomery et al., 2016a).
Thus, the purpose of this study is to understand unsheltered home-
lessness in persons with psychosis and its relationship to cognitive im-
pairment, clinical symptoms, and community functioning, examined
both categorically and dimensionally. We hypothesized that
unsheltered homelessness would be associated with greater severity
of clinical symptoms and greater cognitive impairment. We examined
this in a sample of homeless Veterans with psychotic disorders. Al-
though unsheltered populations present substantial challenges to data
collection, this study had the advantage of examining homelessness
while participants were at the same point in the housing process; all
participants were Veterans newly enrolled in the HUD-VASH program
who had received a Housing Choice Voucher and were currently
searching for housing with assistance from the VA.
2. Methods
2.1. Participants
This study included 76 Veteran participants who: 1) were homeless;
2) had a history of psychosis; and 3) had recently enrolled in the HUD-
VASH program at the VA Greater Los Angeles Healthcare System (GLA)
but had not yet received housing. This study is based on the baseline as-
sessments of a three-year longitudinal study designed to identify the
determinants of successful community integration in a large sample of
homeless Veterans with psychotic disorders who have recently received
Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
HUD-VASH vouchers. A VA administrative database (VA Informatics
and Computing Infrastructure; VINCI) was used to identify all Veterans
enrolled in HUD-VASH at GLA in the preceding month who received in-
patient or outpatient mental health care for major depressive disorder
with psychotic features, any bipolar disorder, psychosis not otherwise
specified, schizoaffective disorder, and schizophrenia (identified by In-
ternational Classification of Diseases and Related Health Problems-9
(ICD-9) codes) in the preceding five years. Opt-in letters were mailed
to Veterans from this list, and Veterans who did not respond to the letter
were subsequently contacted by phone. Additionally, research assis-
tants attended patient orientation sessions for the HUD-VASH program
and distributed information about the study.
To obtain DSM-5 diagnoses, participants were assessed by trained
interviewers using the Structured Clinical Interview for DSM-5 Disor-
ders (SCID; First et al., 2016). All SCID interviewers were trained to a
minimum kappa of 0.75 for key psychotic and mood items through
the Treatment Unit of the Department of Veterans Affairs VISN 22 Men-
tal Illness Research, Education, and Clinical Center (MIRECC). Twenty-
two participants were excluded because they did not meet the DSM-5
diagnostic criteria for a psychotic disorder. The final sample consisted
of the following DSM-5 diagnoses: schizophrenia, schizoaffective disor-
der, other specified/unspecified schizophrenia spectrum disorder, or
mood disorder with psychotic features (see Table 1 for breakdown).
Participants were included if they were between 18 and 60 years of
age, had an estimated premorbid IQ N 70 (based on reading ability as
assessed with the Wide Range Achievement Test), understood spoken
English sufficiently to comprehend testing procedures, had no clinically
significant neurological disease, and no history of serious head injury
(such as loss of consciousness longer than 1 h based on self-report
and medical chart review). Seven participants were excluded due to
head injury (n = 4) and premorbid IQ (b70; n = 3). All participants pro-
vided written informed consent to participate in the study after all pro-
cedures were fully explained in accordance with procedures approved
by the IRB at the GLA.
2.1.1. Symptom assessment
Psychotic symptoms and depression/anxiety in the previous month
were evaluated using the Expanded Brief Psychiatric Rating Scale
(BPRS) (Ventura et al., 1993). The Clinical Assessment Interview for
Negative Symptoms (CAINS) (Kring et al., 2013) was used to assess neg-
ative symptoms. The CAINS is comprised of two subscales: 1) The Moti-
vation and Pleasure (MAP) subscale (9 items) assesses motivational
symptoms (e.g., internal experiences of motivation, drive, and interest);
2) the Expression subscale (4 items) assesses affective flattening and
alogia. For the BPRS and CAINS, higher scores indicate the presence of
more symptoms. Symptom raters were trained to a minimum intra-
class correlation coefficient of 0.80.
2.1.2. Cognitive assessment
The MATRICS Consensus Cognitive Battery (MCCB) (Green et al.,
2004; Nuechterlein and Green, 2006) was used to assess cognition.
The MCCB includes 10 measures within 7 separable cognitive domains,
including speed of processing, verbal learning, visual learning, working
memory, reasoning and problem solving, attention/vigilance, and social
cognition. T-scores for each cognitive domain and an overall composite
score served as dependent variables. T-scores were based on established
age and gender norms for the MCCB (Kern et al., 2008).
2.1.3. Community functioning
The Role Functioning Scale (RFS) (McPheeters, 1984) assesses de-
gree of work, independent living, social connections, and family interac-
tions. Higher scores indicate better functioning. All clinical interviewers
had a masters or doctoral-level degree and were trained through the
Treatment Unit of the VA VISN 22 MIRECC.
 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx 3

Table 1
Demographic and clinical data.

Patient characteristic Sheltered Unsheltered Statistic All participants

n = 46 n = 30 n = 76

Demographics M SD M SD t df p-Value M SD
Age 48.67 9.80 49.33 11.29 −0.31 74 .757 48.97 10.34
Education (years) 13.17 1.48 12.77 1.17 1.27 74 .208 13.01 1.37
Parent education (years) 12.33 3.17 12.10 2.41 0.30 63 .765 12.25 2.90

N % N % χ2 df p N %
Sex 2.75 1 .097
Male 42 91.3 30 100 72 94.70
Female 4 8.70 0 0 4 5.30
Marital status 0.94 1 .331
Unmarried 44 95.7 27 90.0 71 93.40
Married/cohabitating 2 4.30 3 10.0 5 6.60
Race (n, %) 9.38 5 .095
Caucasian 7 15.20 5 16.7 12 15.80
African-American 30 65.20 15 50.0 45 59.20
Asian 1 2.20 0 0 1 1.30
Hispanic 6 13.00 6 20.0 12 15.80
American Indian/Alaskan 2 4.30 0 0 2 2.60
More than one race 0 0 4 13.30 4 5.30

Clinical characteristics M SD M SD t df p M SD
Age of illness onset 22.47 6.97 25.36 10.63 −1.36 66 .180 23.53 8.54
Age of first hospitalization 30.28 11.71 27.46 14.15 0.88 63 .385 29.15 12.71
Total lifetime hospitalizations 4.32 6.34 2.86 2.95 1.14 67 .260 3.73 5.25

N % N % χ2 df p N %
Antipsychotic medication (n, %)
Typical 2 4.30 0 0 4.91 2 .086 2 2.60
Atypical 34 73.90 17 56.70 51 67.10
No medications 10 21.70 13 43.30 23 30.30
Diagnosis (n, %) 5.58 5 .349
Serious mental illness
Schizophrenia 15 32.60 13 43.30 28 36.80
Schizoaffective disorder 4 8.70 4 13.30 8 10.50
Delusional disorder 0 0 1 3.30 1 1.30
Other specified/unspecified schizophrenia spectrum and other psychotic disorder 16 34.80 9 30.00 25 32.90
Bipolar I disorder, with psychotic features 7 15.20 3 10.00 10 13.20
Major depressive disorder, with psychotic features 4 8.70 0 0 4 5.30
Substance use disorder 0.94 1 .332
At least one 38 82.60 22 73.30 60 78.90
None 8 17.40 8 26.70 16 21.10

Note. There was missing data for parental education (n = 11), first psychiatric hospitalization (n = 11), and first signs of illness (n = 8). The number of substance use diagnoses per subject
ranged from 0 to 5, with 19.5% of the sample having one diagnosis, 34.1% having two, 17.1% having three, 7.3% having four, and 4.9 having five diagnoses. Alcohol, stimulant, and cannabis
use disorders were the most common substance use disorders.

2.1.4. Residential outcomes
The Residential Time-Line Follow-Back (TLFB) Inventory (Tsemberis
et al., 2007) was used to gather retrospective participants' residencies
over the past 6 months. The percentage of days that participants spent
in sheltered (temporary, stable, institutional) vs. unsheltered (literal
homelessness) housing was calculated.
independent samples t-tests to determine whether demographic vari-
ables, cognition, clinical symptoms, and community functioning varied
by housing stability. To examine the data dimensionally, we used bivar-
iate Pearson correlations (data were normally distributed) to assess the
relationship between percent days unsheltered and cognition, clinical
symptoms, and community functioning within the entire unsheltered
group (i.e., literal homeless at any point in the past 6 months).

2.2. Statistical analyses 3. Results

We conducted both categorical and dimensional analyses using IBM
SPSS version 24. For categorical analyses, we divided groups based on
housing stability (sheltered vs. unsheltered). Inspection of the TLFB
data revealed no natural cut-point for categorizing participants; thus,
we examined two thresholds for determining housing stability. First,
participants who were literally homeless at any point in the past
6 months (in this sample the least number of days any participant was
unsheltered was 5 days) were categorized as unsheltered (n = 30),
and the remaining participants were considered to be sheltered (n =
46). To ensure that the results are not solely based on the selection of
a low threshold, we repeated the analyses using a higher threshold: par-
ticipants who were literally homeless N20% of days were classified as
unsheltered (n = 22) and the remaining participants were included in
the sheltered group (n = 54). We used chi-squared (χ2) and
3.1. Group differences on sample characteristics, clinical symptoms, com-
munity functioning, and cognition
We present the results in full for the analyses using the low thresh-
old in which unsheltered participants were literally homeless at any
point in the past 6 months. We describe in the text any difference be-
tween using the low versus high (N20% unsheltered) threshold. Demo-
graphic and clinical statistics are summarized in Table 1. There were no
significant differences between sheltered and unsheltered homeless
Veterans on demographic or diagnostic variables regardless, and this
was true regardless of the threshold used for categorizing the
unsheltered group, all p's N .05.
For cognition, there was a significant group difference on reasoning
and problem solving, but in a counter-intuitive direction (see Table 2).

Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
4 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx

Table 2
Descriptive statistics for non-social cognition, social cognition, clinical symptoms, and community functioning.

Sheltered Unsheltered Statistic All participants

(n = 46) (n = 30) (n = 76)

M SD M SD t df p M SD

Non-social cognition (MCCB)

Speed of processing 40.33 11.49 40.97 9.61 −0.25 74 .801 40.58 10.73
Attention/vigilance 37.65 12.73 38.07 12.95 −0.26 74 .891 37.82 12.73
Working memory 23.57 8.20 25.10 8.60 −0.782 74 .436 24.17 8.34
Verbal learning 38.93 7.58 39.13 9.09 −0.10 74 .918 39.01 8.15
Visual learning 38.41 12.64 39.17 11.84 −0.26 74 .795 38.71 12.26
Reasoning and problem solving 43.76 11.63 48.97 8.84 −2.09 74 .040 45.82 10.86
Social cognition 35.28 11.27 38.10 15.23 −0.93 74 .358 36.39 12.96
Overall composite 28.50 11.42 31.07 11.65 −0.95 74 .345 29.51 11.51

Clinical symptoms
BPRS
Positive 2.07 0.88 2.35 0.79 0.36 70 .721 2.04 0.77
Depression/anxiety 2.35 0.79 2.33 0.96 0.10 70 .918 2.34 0.86
CAINS-MAP total 18.20 8.64 17.41 8.98 0.37 70 .715 17.90 8.72
CAINS-EXP total 3.65 3.85 3.20 3.91 0.50 70 .622 3.47 3.85

Community functioning
RFS
Working productivity 2.22 1.38 2.37 1.33 −0.47 74 .641 2.28 1.35
Independent living 4.30 1.28 4.17 1.09 0.49 74 .629 4.25 1.20
Immediate social network 4.50 1.97 4.33 2.02 0.36 74 .724 4.43 1.99
Extended social network 4.15 1.80 3.93 1.76 0.53 74 .601 4.07 1.78

Note. BPRS = Brief Psychiatric Rating Scale; CAINS-MAP/EXP = Clinical Assessment Interview for Negative Symptoms (Motivation and Pleasure scale and Expression scale); RFS = Role
Functioning Scale. BPRS and CAINS data were missing for 4 participants. Numbers in bold represent p-values b .05.

That is, sheltered participants showed more impairment relative to the pattern of scores were in the opposite direction of what was ex-
unsheltered participants, d = −0.49. When using the higher threshold pected for reasoning and problem solving; that is, unsheltered Veterans
(N20%), this difference became a non-significant trend, F(74) = −1.90, scored significantly higher on this domain relative to the sheltered
p = .062, d = −0.44. There were no significant group differences on the group. Perhaps Veterans with more noticeable impairments in reason-
other MCCB domains or the composite score. ing and problem solving are more likely to receive services. However,
Participants showed low to moderate levels of clinical symptoms. the effect of reasoning and problem solving became nonsignificant
There were no significant group differences between sheltered and when using the higher threshold (N20% literally homeless) for
unsheltered homeless Veterans on symptoms or on community func- unsheltered homelessness. In dimensional analyses, there was no asso-
tioning, all p's N .05 (see Table 2). This absence of differences was true ciation between reasoning and problem solving and days of unsheltered
for both threshold levels. homelessness. This indicates that using a categorical approach, despite
using two thresholds, did not accurately reveal the aspects of cognition
3.2. Correlations between percent days unsheltered and cognition, clinical
symptoms, and community functioning
Table 3
Correlations between percent days unsheltered and cognition, clinical symptoms, and
Correlations for the unsheltered group (n = 30) are presented in functional outcome in the unsheltered group (n = 30).
Table 3. Regarding cognition, days unsheltered were correlated with
% days unsheltered
overall cognition such that poorer cognition was associated with more
r p
days unsheltered (see Fig. 1). Among the cognitive domains that con-
tribute to this correlation, poorer visual learning and social cognition MCCB cognition
were associated with more days spent unsheltered, p's b .01 (see Fig. 1 Speed of processing −0.095 .528
Attention/vigilance −0.125 .409
panels B and C). Regarding symptoms, higher levels of expressive nega- Working memory −0.212 .156
tive symptoms were associated with greater percent days unsheltered, Verbal learning −0.238 .111
p b .05, but there were no significant correlations with motivational Visual learning −0.490 .001
negative symptoms. There were no significant correlations between Reasoning and problem solving −0.226 .130
Social cognition −0.407 .005
percent days unsheltered and positive symptoms and community func-
Overall composite −0.388 .008
tioning, all p's N .05. Of note, there was no significant correlation be-
tween percent days unsheltered and substance use diagnosis in the BPRS symptoms
Positive 0.222 .143
full sample or in the unsheltered group only (all p's N .05). This suggests
Depression/anxiety 0.101 .507
that correlations between percent days unsheltered and cognition, clin-
ical symptoms, and community functioning within the unsheltered CAINS Negative Symptoms
Motivation and Pleasure 0.204 .194
group are not better explained by substance use disorders.
Expression 0.346 .018

4. Discussion Role Functioning Scale

Work −0.079 .607
Independent living −0.055 .719
Despite large differences in housing histories, there were no signifi- Family functioning −0.229 .125
cant differences in clinical symptoms or community functioning be- Social functioning −0.232 .120
tween sheltered and unsheltered homeless Veterans with psychotic Note. MCCB = MATRICS Consensus Cognitive Battery; BPRS = Brief Psychiatric Rating
disorders. However, there was one significant difference in cognition Scale; CAINS = Clinical Assessment Interview for Negative Symptoms; RFS = Role Func-
between sheltered and unsheltered homeless participants. Specifically, tioning Scale. Numbers in bold represent p-values b .05.

Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx 5

Fig. 1. Relationship of MCCB domains and percent days unsheltered. The dashed line represents the mean MCCB domain scores for the sheltered group.

that contribute to unsheltered homelessness. Further, this approach did
not capture the effect of overall cognition on unsheltered homelessness.
A dimensional approach examining the relationship between days
unsheltered and clinical and cognitive variables was more useful. Poorer
overall cognition was associated with more days of unsheltered home-
lessness, and visual learning and social cognition contributed to this re-
lationship. Visual learning requires several skills (i.e., psychomotor
skills, spatial approximation/estimation, and visual working memory)
that are related to important aspects of functional outcome, such as em-
ployment status (Kern et al., 2011); thus one could speculate that visual
learning impairments might predispose Veterans with psychosis to-
ward factors (e.g., unemployment) that may contribute to poorer hous-
ing stability.
Poorer social cognition was also related to a greater duration of
unsheltered homelessness, and this has important implications. Social
cognition shows consistent impairment in people with psychotic disor-
ders (Savla et al., 2013) and is associated with poor community func-
tioning (Fett et al., 2011; Green et al., 2012). Deficits in social
cognition may result in poorer ability to manage one's emotions,
which in turn could lead to difficulties in managing conflicts with ten-
ants and landlords. This difficulty could lead to the loss of sheltered
housing or reduced ability to engage in services that facilitate housing.
Therefore, addressing social cognitive deficits may be especially impor-
tant for engaging unsheltered homeless Veterans in rehabilitative hous-
ing efforts.
In general, the level of cognitive impairment in the full sample was
consistent with what is typically seen in schizophrenia and other psy-
chotic disorders (Kern et al., 2013; McCleery et al., 2014), as well as
the cognitive impairment seen in the homeless (Depp et al., 2015). In
the full sample, participants showed impairments that were 1–2 stan-
dard deviations on most domains below that of age and gender
corrected norms.
Higher levels of expressive, but not motivational, negative symp-
toms were associated with greater percent days unsheltered. This pat-
tern is not consistent with findings linking motivational deficits to
poor functional outcome in psychosis (Fervaha et al., 2014). Further,
positive symptoms did not contribute to unsheltered homelessness.
This indicates that factors related to cognitive impairment, rather than
disorder-specific positive symptoms or motivational deficits, may play
a key role in imparting risk of unsheltered homelessness in persons
with psychosis.
Community integration was not associated with percent days
unsheltered. However, the RFS was used to measure community inte-
gration during the baseline assessment in which most Veterans were
temporarily housed as they awaited their HUD-VASH voucher. Since
percent days unsheltered was measured retrospectively (i.e., the past
6 months), the RFS may not have captured community integration
while Veterans were actually experiencing unsheltered homelessness.
This study had several limitations. First, the cross-sectional design of
this study limits the generalizability of findings; however, our findings

Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
6 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx
are the initial report from a baseline assessment and will be followed by
longitudinal analyses to understand the cognitive and motivational pre-
dictors of community integration in homeless Veterans. Further, these
data are limited only to Veterans who received resources for housing
that are not available to other populations experiencing homelessness,
and the rate of women in this sample was not comparable to the rate
of non-Veteran women experiencing homelessness. Second, the current
study investigated only a limited number of social cognitive processes.
Future studies should investigate different social cognitive processes,
as well as measures of social skills competence, to provide opportunities
for targeted recovery and rehabilitation-focused interventions. Third,
this study only investigated a limited number of risk factors. Future
studies should examine the course of other risk factors that may be as-
sociated with cognitive impairment, such as history of substance use
and traumatic brain injury, in relationship to the onset of homelessness,
duration of homelessness, and response to rehabilitative interventions.
Others have also suggested adopting more clinically informative
methods for stratifying risk factors, such as subgrouping based on
neurodevelopmental risks (such as schizophrenia or intellectual disabil-
ity) versus acquired risks (such as traumatic brain injury or substance
use disorder) (Depp et al., 2015).
Despite these limitations, this is one of the few studies to concur-
rently examine unsheltered homelessness using categorical and dimen-
sional approaches. It is recommended that future studies use
dimensional approaches to understand the contribution of specific cog-
nitive deficits, and other possible risk factors, on unsheltered homeless-
ness. More research is also needed to understand the consequences of
these cognitive deficits on the effectiveness of interventions and ser-
vices; for example, cognitive impairment may reduce participation in
and compliance with treatment services (e.g., missed appointments,
misunderstanding instructions, poor planning). Further, assessment of
cognitive deficits has the potential to improve interventions that can in-
crease access to care and engagement in rehabilitative housing efforts.
For example, certain people may benefit from compensatory strategies,
structured skill learning, and/or social cognition training to increase
participation in interventions and maintain housing.
Conflict of interest
Dr. Green has been a consultant for DSP, Lundbeck, and Takeda, a member of the Sci-
entific Board of Cadent, and has received research funds from Forum. The other authors re-
port no disclosures.
Role of funding source
A Department of Veterans Affairs grant awarded to Dr. Green titled Homeless Vet-
erans with Mental Illness: Predicting and Enhancing Recovery (Project #:
1I01RX001116-01A2), and the VA Research Enhancement Award Program (REAP; Project
#: I50RX001875) on Enhancing Community Integration for Homeless Veterans funded
this project. The funding source had no direct role in study design; in the collection, anal-
ysis and interpretation of data; in the writing of the report; and in the decision to submit
the article for publication.
Contributors
All authors designed the study. K. Llerena managed the literature searches, conducted
the statistical analyses, and wrote the first draft of the manuscript. All authors contributed
to and have approved the final manuscript.
Acknowledgement
We gratefully acknowledge all of the people who participated in the present study.
Funding for this study was provided by the Department of Veterans Affairs grant awarded
to Dr. Green titled Homeless Veterans with Mental Illness: Predicting and Enhancing Re-
covery (Project #: 1I01RX001116-01A2), and the VA Research Enhancement Award Pro-
gram (REAP; Project #: I50RX001875) on Enhancing Community Integration for
Homeless Veterans. Dr. Llerena was supported by the Office of Academic Affiliations, Ad-
vanced Fellowship Program in Mental Illness Research and Treatment, Department of Vet-
erans Affairs. Dr. Sonya Gabrielian was supported by a Career Development Award (CDA
15-074) from the VA Health Services Research & Development Service.
References
Backer, T.E., Howard, E.A., 2007. Cognitive impairments and the prevention of homeless-
ness: research and practice review. J. Prim. Prev. 28 (3–4), 375–388.
Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
Burra, T.A., Stergiopoulos, V., Rourke, S.B., 2009. A systematic review of cognitive deficits
in homeless adults: implications for service delivery. Can. J. Psychiatr. 54 (2),
123–133.
Byrne, T., Montgomery, A.E., Fargo, J.D., 2016. Unsheltered homelessness among veterans:
correlates and profiles. Community Ment. Health J. 52 (2), 148–157.
Cheung, A.M., Hwang, S.W., 2004. Risk of death among homeless women: a cohort study
and review of the literature. CMAJ 170 (8), 1243–1247.
Cousineau, M.R., 1997. Health status of and access to health services by residents of urban
encampments in Los Angeles. J. Health Care Poor Underserved 8 (1), 70–82.
CRS, 2015. Congressional Research Service Report for Congress: Veterans and Homeless-
ness, Congress, U.D.O, (Washington, DC).
Depp, C.A., Vella, L., Orff, H.J., Twamley, E.W., 2015. A quantitative review of cognitive
functioning in homeless adults. J. Nerv. Ment. Dis. 203 (2), 126–131.
Drake, R.E., Wallach, M.A., Teague, G.B., Freeman, D.H., Paskus, T.S., Clark, T.A., 1991. Hous-
ing instability and homelessness among rural schizophrenic patients. Am.
J. Psychiatry 148 (3), 330–336.
Fervaha, G., Foussias, G., Agid, O., Remington, G., 2014. Impact of primary negative symp-
toms on functional outcomes in schizophrenia. European Psychiatry 29 (7), 449–455.
Fett, A.-K.J., Viechtbauer, W., Dominguez, M.-d.-G., Penn, D.L., van Os, J., Krabbendam, L., 2011.
The relationship between neurocognition and social cognition with functional outcomes in
schizophrenia: a meta-analysis. Neurosci. Biobehav. Rev. 35 (3), 573–588.
First, M.B., Williams, J.B.W., Karg, R.S., Spitzer, R.L., 2016. Structured Clinical Interview for
DSM-5 Disorders (SCID-5), Clinician Version.
Foster, A., Gable, J., Buckley, J., 2012. Homelessness in schizophrenia. Psychiatr. Clin. North
Am. 35 (3), 717–734.
Gozdzik, A., Salehi, R., O'Campo, P., Stergiopoulos, V., Hwang, S.W., 2015. Cardiovascular
risk factors and 30-year cardiovascular risk in homeless adults with mental illness.
BMC Public Health 15, 165.
Green, M.F., Kern, R.S., Braff, D.L., Mintz, J., 2000. Neurocognitive deficits and functional
outcome in schizophrenia: are we measuring the “right stuff”? Schizophr. Bull. 26,
119–136.
Green, M.F., Kern, R.S., Heaton, R.K., 2004. Longitudinal studies of cognition and functional
outcome in schizophrenia: implications for MATRICS. Schizophr. Res. 72, 41–51.
Green, M.F., Hellemann, G., Horan, W.P., Lee, J., Wynn, J.K., 2012. From perception to func-
tional outcome in schizophrenia: modeling the role of ability and motivation. Arch.
Gen. Psychiatry 69 (12), 1216–1624.
Heinrichs, R.W., Zakzanis, K.K., 1998. Neurocognitive deficit in schizophrenia: a quantita-
tive review of the evidence. Neuropsychology 12 (3), 426–445.
Henry, M., Watt, R., Rosenthal, L., Shivji, A., 2016. The 2016 annual homeless assessment report
to congress. Part 1: Point-in-Time Estimates of Homelessness (Washington, DC).
HUD, 2011. Supplemental Report to the 2009 Annual Homeless Assessment Report to
Congress. U.S. Department of Housing and Urban Development.
Hwang, S.W., 2000. Mortality among men using homeless shelters in Toronto, Ontario.
JAMA 283 (16), 2152–2157.
Kern, R.S., Nuechterlein, K.H., Green, M.F., Baade, L.E., Fenton, W.S., Gold, J.M., Keefe, R.S.E.,
Mesholam-Gately, R., Mintz, J., Seidman, L.J., Stover, E., Marder, S.R., 2008. The
MATRICS consensus cognitive battery: part 2. Co-norming and standardization. Am.
J. Psychiatr. 165, 214–220.
Kern, R.S., Gold, J.M., Dickinson, D., Green, M.F., Nuechterlein, K.H., Baade, L.E., Keefe, R.S.,
Mesholam-Gately, R.I., Seidman, L.J., Lee, C., Sugar, C.A., Marder, S.R., 2011. The MCCB
impairment profile for schizophrenia outpatients: results from the MATRICS psycho-
metric and standardization study. Schizophr. Res. 126 (1–3), 124–131.
Kern, R.S., Penn, D.L., Lee, J., Horan, W.P., Reise, S.P., Ochsner, K.N., Marder, S.R., Green, M.F.,
2013. Adapting social neuroscience measures for schizophrenia clinical trials, part 2:
trolling the depths of psychometric properties. Schizophr. Bull. 39 (6), 1201–1210.
Kertesz, S.G., Austin, E.L., Holmes, S.K., DeRussy, A.J., Van Deusen Lukas, C., Pollio, D.E.,
2017. Housing first on a large scale: fidelity strengths and challenges in the VA's
HUD-VASH program. Psychol. Serv. 14 (2), 118–128.
Kring, A.M., Gur, R.E., Blanchard, J.J., Horan, W.P., Reise, S.P., 2013. The Clinical Assessment
Interview for Negative Symptoms (CAINS): final development and validation. Am.
J. Psychiatr. 170 (2), 165–172.
Kuno, E., Rothbard, A.B., Averyt, J., Culhane, D., 2000. Homelessness among persons with
serious mental illness in an enhanced community-based mental health system.
Psychiatr. Serv. 51 (8), 1012–1016.
Lam, J.A., Rosenheck, R., 1999. Street outreach for homeless persons with serious mental
illness: is it effective? Med. Care 37 (9), 894–907.
Lamb, H.R., Bachrach, L.L., 2001. Some perspectives on deinstitutionalization. Psychiatr.
Serv. 52 (8), 1039–1045.
Levitt, A.J., Culhane, D.P., DeGenova, J., O'Quinn, P., Bainbridge, J., 2009. Health and social
characteristics of homeless adults in Manhattan who were chronically or not chron-
ically unsheltered. Psychiatr. Serv. (Washington, D.C.) 60 (7), 978–981.
Lincoln, A.K., Plachta-Elliott, S., Espejo, D., 2009. Coming in: an examination of people
with co-occurring substance use and serious mental illness exiting chronic homeless-
ness. Am. J. Orthop. 79 (2), 236–243.
McCleery, A., Ventura, J., Kern, R.S., Subotnik, K.L., Gretchen-Doorly, D., Green, M.F.,
Hellemann, G.S., Nuechterlein, K.H., 2014. Functioning in first-episode schizophrenia:
MATRICS consensus cognitive battery (MCCB) profile of impairment. Schizophr. Res.
157 (0), 33–39.
McPheeters, H.L., 1984. Statewide mental health outcome evaluation: a perspective of
two southern states. Community Ment. Health J. 20 (1), 44–55.
Montgomery, A.E., Cusack, M., Blonigen, D.M., Gabrielian, S., Marsh, L., Fargo, J., 2016a.
Factors associated with veterans' access to permanent supportive housing. Psychiatr.
Serv. (Washington, D.C.) 67 (8), 870–877.
Montgomery, A.E., Szymkowiak, D., Marcus, J., Howard, P., Culhane, D.P., 2016b. Home-
lessness, unsheltered status, and risk factors for mortality: findings from the
100,000 homes campaign. Public Health Rep. 131 (6), 765–772.
 K. Llerena et al. / Schizophrenia Research xxx (2018) xxx–xxx
Morrison, D.S., 2009. Homelessness as an independent risk factor for mortality: results
from a retrospective cohort study. Int. J. Epidemiol. 38 (3), 877–883.
Nuechterlein, K.H., Green, M.F., 2006. MATRICS Consensus Cognitive Battery. MATRICS As-
sessment, Inc., Los Angeles.
O'Toole, T.P., Gibbon, J.L., Hanusa, B.H., Fine, M.J., 1999. Utilization of health care services
among subgroups of urban homeless and housed poor. J. Health Polit. Policy Law 24
(1), 91–114.
Rosenheck, R., Kasprow, W., Frisman, L., Liu-Mares, W., 2003. Cost-effectiveness of sup-
ported housing for homeless persons with mental illness. Arch. Gen. Psychiatry 60
(9), 940–951.
Savla, G.N., Vella, L., Armstrong, C.C., Penn, D.L., Twamley, E.W., 2013. Deficits in domains
of social cognition in schizophrenia: a meta-analysis of the empirical evidence.
Schizophr. Bull. 39 (5), 979–992.
Seidman, L.J., Caplan, B.B., Tolomiczenko, G.S., Turner, W.M., Penk, W.E., Schutt, R.K.,
Goldfinger, S.M., 1997. Neuropsychological function in homeless mentally ill individ-
uals. J. Nerv. Ment. Dis. 185 (1), 3–12.
Shern, D.L., Tsemberis, S., Anthony, W., Lovell, A.M., Richmond, L., Felton, C.J., Winarski, J., Cohen,
M., 2000. Serving street-dwelling individuals with psychiatric disabilities: outcomes of a
psychiatric rehabilitation clinical trial. Am. J. Public Health 90 (12), 1873–1878.
Please cite this article as: Llerena, K., et al., Clinical and cognitive correlates of unsheltered status in homeless persons with psychotic disorders,
Schizophr. Res. (2018), https://doi.org/10.1016/j.schres.2018.02.023
7
Spence, S., Stevens, R., Parks, R., 2004. Cognitive dysfunction in homeless adults: a system-
atic review. J. R. Soc. Med. 97 (8), 375–379.
Stergiopoulos, V., Dewa, C., Durbin, J., Chau, N., Svoboda, T., 2010. Assessing the mental
health service needs of the homeless: a level-of-care approach. J. Health Care Poor
Underserved 21 (3), 1031–1045.
Tsai, J., Kasprow, W.J., Kane, V., Rosenheck, R.A., 2014. Street outreach and other forms of
engagement with literally homeless veterans. J. Health Care Poor Underserved 25 (2),
694–704.
Tsemberis, S., McHugo, G., Williams, V., Hanrahan, P., Stefancic, A., 2007. Measuring
homelessness and residential stability: the residential time-line follow-back inven-
tory. J. Community Psychol. 35 (1), 29–42.
USICH, 2010. Federal Strategic Plan to Prevent and End Homlessness. U.S. Interagency
Council on Homelessness.
Ventura, J., Lukoff, D., Nuechterlein, K.H., Liberman, R.P., Green, M.F., Shaner, A., 1993. Brief
Psychiatric Rating Scale (BPRS) expanded version: scales, anchor points, and admin-
istration manual. Int. J. Methods Psychiatr. Res. 3, 227–243.