Schizophrenia Research 83 (2006) 65 75

www.elsevier.com/locate/schres

A 12-month prospective follow-up study of patients with

schizophrenia-spectrum disorders and substance abuse:
Changes in psychiatric symptoms and substance use
Howard C. Margolese a,b,d, Juan Carlos Negrete a,d, Raymond Tempier c,d,
Kathryn Gill a,d,*
a

Addictions Unit, McGill University Health Centre, Canada

Clinical Psychopharmacology Unit, McGill University Health Centre, Canada
c
Continuing Care Service, McGill University Health Centre, Canada
d
Department of Psychiatry, McGill University, Montreal, Quebec, Canada

Received 19 July 2005; received in revised form 17 November 2005; accepted 18 November 2005
Available online 7 February 2006

Abstract
While it is widely known that patients with schizophrenia-spectrum psychoses and co-occurring substance use disorders are
more difficult to manage, there is limited data on the course of their psychiatric symptoms when they remain in treatment over
time. This prospective 12-month study evaluated changes in psychiatric symptoms and substance use to ascertain if the coexistence of substance use disorders influences ratings of psychiatric symptoms at follow-up.
147 outpatients in a continuing care program were assessed at intake and followed prospectively for 12 months. Psychiatric
symptoms were measured at baseline and 12-month follow-up using the Positive and Negative Syndrome Scale (PANSS) and
Hamilton Depression Rating Scale (HAM-D). Subjective psychological distress was rated with the Brief Symptom Inventory
(BSI) and quality of life by the Satisfaction with Life Domains Scale (SDLS). Drug and alcohol use was measured with the
Addiction Severity Index (ASI).
50.3% of patients were diagnosed with dual disorders (DD) (current and lifetime). The most common primary substances of
abuse were alcohol (35.6%) and cannabis (35.1%). DD subjects had higher baseline PANSS positive scores but experienced a
greater reduction at 12 months compared to single diagnosis (SD) patients. Severity of substance abuse as measured by ASI
composite scores did not decrease significantly between baseline and 12 months.
DD patients with schizophrenia and related psychoses treated for their psychiatric illness showed a reduction in PANSS
scores over 12 months, even when their substance use remained largely unchanged. However, co-morbidity cases continued to

* Corresponding author. Addictions Unit, Montreal General Hospital, 1604 Pine Avenue West, Montreal, Quebec, Canada H3G 1B4. Tel.: +1
514 934 8311; fax: +1 514 934 8262.
E-mail address: kathryn.gill@mcgill.ca (K. Gill).
0920-9964/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.schres.2005.11.019

66

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

show higher depression and anxiety ratings. Ongoing substance abuse appears to be related to levels of depression as 62.5% of
DD-current versus 34.7% of SD patients had HAM-D scores in the depressed range at 12-month follow-up. Implications for
treatment are discussed.
D 2005 Elsevier B.V. All rights reserved.
Keywords: Alcohol; Cannabis; Dual-diagnosis; Psychosis; Schizophrenia; Substance use disorders

1. Introduction
Dual diagnosis (DD) patients are known to present
more difficulties from a clinical management perspective than single diagnosis (SD) patients. Substance use
has been found to exacerbate the positive symptoms
of schizophrenia (Drake et al., 1989; Margolese et al.,
2004; Negrete et al., 1986; Pulver et al., 1989) and to
increase aggression, violence (Angermeyer, 2000;
Soyka, 2000), as well as medication noncompliance
(Bhanji et al., 2004; Coldham et al., 2002; Kamali et
al., 2001; Margolese et al., 2004; Olfson et al., 2000;
Swartz et al., 1998) in the chronic mentally ill.
Furthermore, time to readmission or community
survival is significantly reduced among DD compared
to SD subjects, even when controlling for noncompliance (Hunt et al., 2002).
Some follow-up studies have found that only a
small percentage of patients with severe mental illness
(SMI) achieve stable substance use remission over
time (Drake et al., 1996). The prevalence of active
substance use disorders changed little during a 7-year
naturalistic follow-up (Bartels et al., 1995) of a SMI
sample, suggesting that remission and new cases were
approximately equal. Another study, which found
high remission rates among dually diagnosed patients,
reported significantly higher dropout rates among
patients with schizophrenia and other psychoses
compared to patients with other diagnoses (Dixon
et al., 1998).
In a retrospective 18-month study of 100 schizophrenic outpatients, between 30% and 40% were
found to be using substances during each 3-month
interval (Chouljian et al., 1995). Analysis of 59 of the
subjects with complete data demonstrated that cocaine
and multiple substances increased, while problem use
of alcohol, marijuana and other drugs remained stable
(Chouljian et al., 1995).
Caspari (1999) examined the impact of cannabis
abuse on schizophrenia in a sample of 27 DD patients

compared to 26 SD controls. He found higher

rehospitalization rates and worse psychosocial functioning in the cannabis abuse group at follow-up after
68.7 F 28.3 months. Of note, 48% of the DD patients
had ceased all substance abuse, while only 1 SD
patient started excessive alcohol use, and none
showed significant drug use. The cannabis group also
had additional thought disturbance measured on the
Brief Psychiatric Rating Scale (BPRS) and hostility
measured on the Arbeitsgememeinscaft fur Methodik
und Dokumetation in der Psychiatrie (AMDP) (German scale for measurement of psychotic symptoms)
compared to the control group at the follow-up
assessment (Caspari, 1999). The study design produced group comparisons for psychiatric symptoms
that were cross-sectional, rather than within-subjects
as the BPRS and AMPD were conducted only at
follow-up and not at baseline.
It seems possible that increased symptoms expressed by DD patients in some studies are accounted
for by medication non-compliance as well as the
direct effect of substances of abuse on the production
of psychiatric symptomatology. Buhler et al. (2002)
followed 29 DD and 29 SD first episode patients for 5
years. Among the patients available for follow-up
assessments, those with a DD had more positive
symptoms and less affective flattening (Buhler et al.,
2002). Once again, changes in drug use prevalence
or severity during the follow-up period were not
explored.
The impact of 12 months of treatment in a general
hospital outpatient psychiatry service for patients
with severe and persistent mental illness including
schizophrenia, schizoaffective disorder and related
psychoses was explored in the present study.
Changes in the expression of psychiatric symptoms
and use of substances of abuse were prospectively
evaluated to ascertain if continuing substance abuse
and dependence were associated with poorer therapeutic response.

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

2. Methods
2.1. Sample selection
The sample consisted of all patients with schizophrenia-spectrum disorders (schizophrenia, schizoaffective disorder, delusional disorder, psychosis NOS),
who were referred to the Continuing Care Service
(CCS) of the McGill University Health Centre
Montreal General Hospital (MGH) site. The MGH is a
tertiary care hospital serving a diverse urban population and the CCS cares for all those patients with
severe and persistent mental illness (Margolese et al.,
2004). There are no specific selection criteria for
referral to the CCS other than presence of a chronic or
recurring psychotic disorder (schizophrenia and related psychoses) and home address in a defined
geographical catchment area. The presence or absence
of substance use disorders plays no role in this
referral. The catchment area (defined by home postal
code) requires patients to obtain all psychiatric
services at the MGH. The MGH catchment area is
wide, representing a multicultural population from
both inner city and suburban regions. CCS admissions
who presented with a pure substance-induced psychosis or bipolar disorder and those whose clinical
and/or hospital chart diagnosis of schizophreniaspectrum psychosis was not confirmed by an SCIDP interview were excluded from the present analyses.
Characteristics of the sample (n = 207) at baseline
were presented in a prior report (Margolese et al.,
2004). Briefly, there was no statistically significant
difference between DD and SD groups on age,
diagnosis, education level, living arrangements, percent on social assistance (welfare), employment
status, lifetime or recent (past 2 years) number of
hospitalizations, or type of antipsychotic prescribed at
baseline (typical (oral or IM), atypical or both).
Significantly more DD subjects were male and they
first received help at a younger age.
While patients were evaluated at 4-month intervals, the complete set of evaluations was performed
only at study entry and study completion. Furthermore, some patients available at 12 months were not
available at some of the intervening time points, and
vice versa. Comparison of baseline to study end
point was used in order to maintain the maximum
possible sample size.

67

The follow-up study was designed to test the

hypothesis that DD patients with a current diagnosis
of substance abuse/dependence would fare significantly worse than non-using SD patients on psychotic
and affective symptoms, subjective psychological
distress, quality of life and medication adherence.
Power analyses were conducted in order to determine
whether the follow-up sample was adequate to test the
primary hypothesis. There are no standard methods
for power/sample size calculations in the context of
multivariate models for unbalanced repeated measures
MANOVA. Approximate power was determined by
generalizing conventional multiple linear analysis to
the case of repeated measurements of an interval-scale
dependent variable. Calculations indicated that a total
sample of 140 patients would be required to ensure
80% power in a 2-tailed test ( p = 0.05) of the
significance of the difference between current drug
abusers and other patients, after having adjusted for
potential confounders (e.g. age, gender).
At 12 months, 147 of the initial sample of 207
patients (71.0%) were still receiving services at the
MGH and agreed to the follow-up interview. Followup rates were 66.1% for SD, 73.5% for DD-lifetime,
and 82.1% for DD-current subjects (v 2 = 3.3, df = 2,
p = 0.190). Compared to those lost to follow-up,
subjects remaining in the study were significantly
older (mean age 39.6 F 10.4 vs. 36.2 F 10.9; p = 0.03),
were more likely to live alone (37.0% vs. 19.7%;
p = 0.05) and had significantly lower PANSS negative
scores at intake (15.8 F 5.2 vs. 17.6 F 5.6; p = 0.03).
There were no differences between subjects remaining
in the study and those lost to follow-up on any other
demographic, psychiatric, or drug use variable,
including compliance status. Regarding compliance
status, there was in fact a trend for retention to be
greater for those not compliant at intake (87.0% vs.
68.2%). All analyses were conducted on the 12-month
follow-up sample (n = 147).
2.2. Instruments
Diagnoses were confirmed through the administration of the mood, psychotic and psychoactive substance abuse sections of the Structured Clinical
Interview for DSM-IV (SCID-P). Baseline demographics including age, ethnicity, education, marital
status, and personal and family psychiatric history

68

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

were recorded. Date of onset of psychiatric symptoms,

number and length of previous hospitalizations, and
medications prescribed (dose, form, compliance status) were obtained from patient interviews and review
of previous hospital charts.
Psychiatric symptomatology was assessed at baseline and 12-month follow-up using the Positive and
Negative Syndrome Scale (PANSS) (Kay et al., 1988,
1989) and the Hamilton Depression Rating Scale
(HAM-D) (Hamilton, 1960); subjective psychological
distress was measured with the Brief Symptom
Inventory (BSI) (Derogatis and Melisaratos, 1983);
while quality of life (QOL) was measured by the
Satisfaction with Life Domains Scale (SDLS) (Baker
and Intagliata, 1982). The PANSS is a 30-item
standardized instrument that measures positive symptoms (e.g. hallucination, delusions), negative symptoms (e.g. affective blunting, emotional withdrawal),
and general symptoms (e.g. motor retardation, anxiety, disorientation) using a semi-structured interview
and chart review (Kay et al., 1988, 1989). The HAMD is a 23-item clinician administered scale that rates
cognitive, affective, somatic, and vegetative symptoms of depression (Hamilton, 1960). The BSI is a 53item self-rating questionnaire that evaluates psychological distress in nine areas (e.g. hostility, depression,
somatization, anxiety) over the past week (Derogatis
and Melisaratos, 1983). A global severity index (GSI)
score is also obtained from the nine dimensions
providing an indication of overall distress. The SDLS
is a 7-point scale wherein stylized faces are used to
rate patients feelings about relationships, autonomy,
leisure activities, health and housing (Baker and
Intagliata, 1982). It has been widely used with
mentally ill patients (Baker and Intagliata, 1982;
Mercier et al., 1992) and it has been demonstrated
to have good psychometric properties in terms of
reliability, internal consistency, and convergent and
content validity (Horley, 1985; Kamman et al., 1983;
Larsen et al., 1985).
The Addiction Severity Index (ASI) was used to
determine current and lifetime drug and alcohol use
levels (McLellan et al., 1980, 1985). The ASI was
found reliable and valid for assessing drug-related
behaviours and consequences in mentally ill patients
(Carise et al., 2001; el Guebaly and Hodgins, 1992). It
assesses the number of days and routes of administration of specific drug (e.g. cannabis, cocaine,

amphetamines, etc.) and alcohol use during the past

30 days, as well as the number of days of drug
abstinence and extent of substance abuse treatment.
Composite scores range from 0 (abstinent) to 1 (daily
use with significant impairment and intoxication).
2.3. Treatment during the follow-up period
The CCS provides long-term care with a focus on
improving quality of life, encouraging community
living, prevention of hospitalizations and increased
autonomy. Clients are treated by a multidisciplinary
team that includes a psychiatrist, social worker, nurse
and occupational therapist. A case manager who
becomes the primary contact person for both the
client and family follows each client. A number of
services were offered during treatment including
medication and side-effect management, supportive
psychotherapy, crisis intervention, social skills training, as well as linkage with community services
related to general health care, housing, finances,
vocational training and recreation. Case managers
were not specifically trained to address substance use
but used the standard case management approach to
help each patient. Thus, general encouragement to
reduce substances of abuse and psychoeducation on
the harmful effects of substance use was discussed
with those patients whose case manager was aware of
their ongoing substance use. This program contained
no specific component focussed on co-morbid substance use disorders; it was not an integrated treatment
program.
2.4. Statistical analysis
Data for each patient across all variables including
demographic and diagnostic information was coded
and entered into a database using Microsoft ExcelR.
Statistical analysis was conducted using the microcomputer version 10.0 of SPSSR. Fishers exact tests
and chi-square tests of association were used to assess
differences in categorical variables between groups.
Comparisons between groups for continuous variables
were conducted using independent t-tests and Analysis of Variance (ANOVA) techniques, including
multivariate tests (MANOVA). Comparisons between
baseline and follow-up time points were performed
using repeated measures analyses. All post-hoc tests

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

69

Table 1
Demographics (n = 147)
Variable

SD (n = 73)

DD-lifetime (n = 50)

DD-current (n = 24)

Age (mean F S.D.)

% Male
Diagnosis at intake
% Schizophrenia
% Schizoaffective
% Other (delusional disorder, psychosis NOS)
% Married
% Caucasian
Living arrangements
% Alone
% Institutiona
% With other people
Education: % above high school
% Employed
% On welfare
Age first received help (mean F S.D.)
Age first hospitalizedb (mean F S.D.)
Number of hospitalizations at baseline (mean F S.D.)
% With history of medical problems
% With current medical problems
% Non-compliant at baseline
% Non-compliant at follow-up

40.4 F 10.4
47.9

40.1 F 9.7
74.0*

38.6 F 10.8
79.2*

63.0
26.0
11.0
9.6
89.0

66.0
24.0
10.0
12.0
90.0

58.3
16.7
25.0
12.5
95.8

32.9
24.7
49.7
65.8
5.5
65.8
24.2 F 7.9
28.2 F 9.0
4.6 F 4.4
54.8
42.5
5.6
10.9

42.9
22.0
36.0
52.0
2.0
68.0
21.2 F 9.4
25.6 F 8.6
4.8 F 5.4
52.0
28.0
20.8*
12.0

37.5
25.0
37.5
58.3
0
70.8
22.0 F 9.2
29.4 F 9.4
3.9 F 3.0
58.3
37.5
25.0*
12.5

S.D. = standard deviation.

a
Institution includes any supervised living setting.
b
For age first hospitalized, n = 67 for SD, n = 45 for DD-lifetime, and n = 20 for DD-current as some patients were never hospitalized.
* DD group significantly different than SD group, v 2 or Fishers Exact Test, p b 0.05 corrected for multiple comparisons.

were conducted using t-tests with a Bonferroni

correction.
Linear hierarchical multiple regression analysis
was performed to examine the relationships among
drug use and psychiatric symptomatology (e.g.
PANSS scores) at follow-up. Variables with significant individual correlations, as well as key variables
of interest (PANSS scores) were entered into hierarchical multiple linear regression using the following
model: Step 1: Baseline ASI alcohol or drug severity
composite scores; Step 2: demographics (age and
sex); Step 3: mood variables at follow-up (HAM-D,
BSI, antidepressant status); Step 4: variables related to
psychosis at follow-up (PANSS, antipsychotic type).

when they first received psychiatric treatment (Table

1). There were no significant differences between SD
and DD patients on any other demographic variable.
Medication compliance rates differed markedly at
baseline, but were similar among the 3 groups at
follow-up, with a higher percentage of DD subjects
compliant at follow-up versus baseline and a lower
percentage of SD subjects compliant at follow-up

3. Results
The patients in this cohort were chronically ill with
approximately 16 years since first diagnosis. Subjects
with a DD were more likely than SD to be male, noncompliant with medications at intake and younger

Fig. 1. Non-compliance at baseline and follow-up.

70

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

Table 2
Primary substance of abuse for current and lifetime DD patients at
12 months
Substance of
abuse

DD-lifetime
(n = 50)

DD-current
(n = 24)

Total sample
(n = 74)

Alcohol (%)
Cannabis (%)
Cocaine (%)
Other (%)

29.2
38.5
12.5
16.7

40.0
32.0
20.0
8.0

36.5
35.1
17.6
10.8

versus baseline (Fig. 1). At baseline, 41.5% of the

sample were prescribed typical antipsychotics, 37.4%
atypicals, 17.0% both and 4.1% none. The percentage
of patients prescribed different antipsychotic types
(typical, atypical, both) did not differ significantly
between SD and DD patients. Furthermore, there was
little shift over time so that similar percentages of
patients were receiving typicals, atypicals or both
antipsychotic types at follow-up.
Only 15% (22/147) of patients required hospitalization during the follow-up period and most needed
either 1 (N = 14) or 2 (N = 6) hospitalizations. There
were no significant differences between groups (SD,
DD-current, DD-lifetime) with respect to hospitalization rates.
Primary substance of abuse for DD-current and
DD-lifetime patients is shown in Table 2. A higher
percentage of DD-lifetime subjects used cannabis
while a lower percent used both alcohol and cocaine
compared to DD-current subjects. These differences
were not statistically significant. Of the 74 patients
with current or lifetime diagnoses of DD, 58 were
actively abusing substances at baseline. At follow-up,
36.2% (21/58) were abstinent from their primary drug
during the month prior to evaluation. However, 7 of

Fig. 2. ASI alcohol composite severity scores.

Fig. 3. ASI drug composite severity scores.

these 21 patients were using substances of abuse other

than their identified primary drug of abuse. Thus,
severity of substance use as measured by the
composite scores on the ASI did not differ significantly when comparing baseline to 12-month followup measures. There were no significant time
[ F(1,135) = 1.42, p = 0.24] or group  time interactions [ F(2,135) = 0.34, p = 0.72]. This lack of change
in overall substance use was also observed for alcohol
(all p values N 0.3) (Fig. 2) and drug (all p values N 0.2) (Fig. 3) use measures separately.
During the course of psychiatric outpatient treatment, DD subjects experienced a greater reduction in
PANSS positive symptom measures compared to SD
patients at 12 months (Fig. 4). Repeated measures
ANOVA with group (3 levels) and time (2 levels)
yielded significant effect of time [ F(1,140) = 47.50,
p = 0.0001] and a significant group  time interaction

Fig. 4. PANSS positive scores.

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

[ F(2,140) = 3.97, p = 0.021]. Post-hoc analysis failed

to reveal a statistically significant between group
difference at 12 months. In other words, 12 months of
routine outpatient treatment caused all groups to look
more similar in terms of positive symptom expression.
HAM-D total scores were analyzed using repeated
measures ANOVA with group (3 levels) and time (2
levels) yielding a significant main effect for group
[ F(2,138) = 5.41, p = 0.005] and time
[ F(1,138) = 4.92, p = 0.028] but no significant group  time interaction [ F(2,138) = 0.379, p = 0.685]. Posthoc analysis indicated that DD-current subjects
remained significantly ( p b 0.05) more depressed than
SD at both baseline and 12 months. Substance abuse
appeared to impact on rates of clinically significant
depression as 62.5% of DD-current and 50% of DDlifetime versus 34.7% of SD patients had HAM-D
scores in the depressed range of 12 or more at followup (v 2 = 0.039, df = 2). Most of the patients who did
receive concomitant antidepressants received an SSRI
(19/31, 61.3%).
While there was an absolute significant difference
between SD versus DD-current and DD-lifetime
subjects on changes in total subjective QOL scores
(Table 3), the result was no longer significant after
individual items were considered (i.e. after correction
for multiple comparisons).

71

Self-rating of symptoms, as measured by the BSI,

was different between SD, DD-lifetime and DDcurrent groups (Table 3). While all groups had some
reduction in their levels of distress (Global Severity
Index scores) these remained the most elevated for
DD-current subjects [ F(2,139) = 9.39, p b 0.001]. Furthermore, while all three groups had some reduction
in their Positive Symptom Total (PST) scores, the DDcurrent group remained the most symptomatic, followed by the DD-lifetime group [ F(2,142) = 10.30,
p b 0.001].
The relationship between psychiatric symptomatology and alcohol/drug abuse severity at follow-up was
examined using hierarchical multiple regression. ASI
alcohol and drug severity composite scores at followup were predicted using a model that included
variables related to baseline ASI composite scores,
demographics, psychological distress and psychotic
symptoms. Results of the regression demonstrated that
alcohol and drug severity at intake largely predicted
severity at follow-up (Table 4). ASI alcohol composite
severity scores at follow-up were predicted by severity
at intake (R 2change, F 1,135 = 75.73, p b 0.001). Similarly,
drug severity at follow-up was predicted by drug
severity at baseline (R 2change, F 1,130 = 84.53, p b 0.001).
Younger age also predicted alcohol, but not drug ASI
severity scores. No additional variance in alcohol or

Table 3
Psychiatric and substance use measures at baseline and follow-up
Variable

PANSS
Positive
Negative
Total
HAM-D Total score
ASI
Alcohol Comp score
Drug Comp Score
SDLS Total score
BSI
GST
PST
PSDI

SD (n = 73)

DD-lifetime (n = 50)

DD-current (n = 24)

Baseline

Follow-up

Baseline

Follow-up

Baseline

Follow-up

14.01 F 5.36
16.10 F 5.26
59.90 F 13.82
11.7 F 6.90

12.16 F 5.45
14.92 F 6.63
54.52 F 17.52
9.5 F 7.7

15.22 F 5.16
15.22 F 5.05
61.70 F 13.15
14.4 F 8.4

12.72 F 5.16
15.11 F 5.92
55.96 F 15.66
11.7 F 8.0

18.17 F 5.44
15.29 F 4.93
66.83 F 13.53
15.4 F 7.5

13.04 F 5.43
15.63 F 5.26
59.54 F 16.74
14.7 F 7.1

0.02 F 0.06
0.006 F 0.02
91.9 F 18.5

0.04 F 0.08
0.003 F 0.008
98.6 F 20.7

0.06 F 0.08
0.02 F 0.05
88.1 F17.7

0.09 F 0.15
0.02 F 0.07
89.7 F 20.3

0.20 F 0.20
0.11 F 0.11
85.3 F 19.6

0.19 F 0.20
0.13 F 0.13
86.6 F 22.3

0.80 F 0.57
22.7 F 12.4
1.72 F 0.68

0.55 F 0.55
17.4 F 13.5
1.45 F 0.54

1.05 F 0.64
27.8 F 12.8
1.88 F 0.55

0.84 F 0.56
25.6 F 14.4
1.53 F 0.43

1.30 F 0.72
33.2 F 12.6
2.06 F 0.60

1.05 F 0.56
29.7 F 14.7
1.74 F 0.44

PANSS = Positive and Negative Syndrome Scale; HAM-D = Hamilton Depression Rating Scale; ASI = Addiction Severity Index; SDLS = Satisfaction with Life Domains Scale; BSI = Brief Symptom Inventory, GST = global symptom total, PST = positive symptom total, PSDI = positive
symptom distress index.
For none of the above variables was DD-current or DD-lifetime group significantly different from SD group, or DD-current group significantly
different from DD-lifetime group, p b 0.05 corrected for multiple comparisons.

72

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

Table 4
Hierarchical multiple regression predicting ASI composite severity ratings for alcohol and drug abuse at follow-up
Predictors

Step 1
Baseline ASI alcohol or drug
severity composite scores
Step 2: demographics
Age
Sex
Step 3: mood variables
HAM-D total
BSI depression
BSI anxiety
BSI GSI
Any antidepressant use
Step 4: psychosis variables
PANSS positive
PANSS negative
PANSS total
Antipsychotic type

ASI alcohol severity at 12 months

2

0.60

0.36

0.62

0.64

0.65

0.39

0.41

0.42

R change
0.36,
p b 0.0001*
0.03,
p = 0.06
0.24,
p = 0.39

0.10,
p = 0.73

ASI drug severity at 12 months

b

R2

R 2 change

0.63

0.39

0.39,
p b 0.0001*

0.60*
0.63

0.40

0.16*
0.02
0.64

0.42

0.03
0.03
0.16
0.04
0.03
0.66
0.11
0.01
0.12
0.09

0.44

0.002,
p = 0.77
0.02,
p = 0.58

0.23,
p = 0.30

ba
0.63*

0.02
0.04
0.03
0.07
0.12
0.01
0.07
0.27
0.28
0.54
0.10

b weights are standardized.

* p b 0.05.

drug severity at follow-up was accounted for by the

severity of psychosis or depression. The presence or
absence of antidepressant medications and the type of
antipsychotic medication prescribed during treatment
did not add to understanding the variance in alcohol/
drug abuse severity at follow-up.

4. Discussion
The present study demonstrates that when DD
patients are adequately treated for their psychiatric
illness there is considerable amelioration in their
psychotic symptoms. Moreover, the reduction in
psychosis does not appear to be related to the type
of antipsychotic prescribed (typical vs. atypical). The
12-month follow-up rate of 71% was considered to be
reasonable for a chronic psychosis population in an
ambulatory care setting. While the numbers did not
reach statistical significance, there was a trend for
higher retention among the current (82.1%) and
lifetime DD (73.5%) groups compared to the SD
group (66.1%). This finding was unexpected, as it is
often assumed that DD patients tend to drop out of
treatment more readily. BootsMiller et al. (1998), for
instance, reported that dual diagnosis subjects have a

higher attrition rate and are more difficult to track in

follow-up studies.
This study replicates the often-reported finding of
poor treatment compliance among DD subjects; they
were more likely to be non-compliant with prescribed
medications at baseline. However, 12 months later all
groups showed similar compliance rates; largely due
to considerable improvement in the dual diagnosis
groups. This change for the better can be seen as a
measure of success for the case management approach
adopted by the CCS program, and the significant
impact of this approach on treatment compliance
among DD subjects confirms current best-practices
recommendations (Drake et al., 2004).
Alcohol and cannabis were the most commonly
abused substances in the present cohort, and overall
severity of substance use, as measured by ASI
composite scores, did not change significantly over
the 12-month follow-up period. While 36.2% of DD
patients abstained from using their primary drug over
the month preceding follow-up evaluation, 1/3 of
them substituted another substance of abuse. There
was little change in drug/alcohol dependence over the
follow-up period.
On average, the severity of psychotic symptoms as
measured by the PANSS, did diminish over the 12-

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

month treatment period. All groups benefited from

treatment, but the gains were most marked in the DD
despite the fact that they continued to use substances of
abuse during the follow-up period.
It would appear that DD patients in this cohort
did no worse than SD patients with respect to
positive and negative psychiatric symptoms, living
arrangements, hospitalizations, and employment
status over the 12-month follow-up. This observation is somewhat at odds with previous reports,
mostly from the United States. The relatively low
rates of cocaine usea more potent psychotomimeticin this Montreal sample; may help explain
the difference. There are, nonetheless, other followup studies that found that substance disorders do
not impact negatively on psychiatric status. Blow et
al. (1998) explored inpatient service use and general
functioning over 2 years in 632 veterans with an
Axis 1 psychotic disorder, 90% of whom had
schizophrenia, 7.7% an affective psychosis and
198 (29%) a co-morbid substance use disorder.
They found that Brief Psychiatric Rating Scale
(BPRS) (Overall and Gorman, 1962) scores were
consistently lower (i.e. better) among DD compared
to SD subjects. Clinical rated Global Assessment of
Functioning scores were higher (better) for DD
compared to SD groups and all mean group GAF
scores increased between baseline and follow-up.
Similar to our findings, DD patients reported less
satisfaction with life compared to SD patients on a
subjective measure of life satisfaction (Blow et al.,
1998). However, the sample differs from the present
study in that they were all veterans and 96.8% of them
were male. Moreover, Blow et al. did not evaluate
changes in substance use measures over time, nor
specify the types of substances that were abused. Bell
et al. examined patients with schizophrenia in a work
rehabilitation program in the United States (Bell et al.,
2002). 80% of the group was classified as DDlifetime, as opposed to DD-current, on the ASI. At 1year follow-up, most patients were classified as
remaining in stable remission, and substance use
was found not to influence work participation rates
(Bell et al., 2002). Thus the impact of substance abuse
on social functioning as measured by work, was
minimal. Unfortunately, this paper did not report
changes in PANSS or ASI scores over time although
these data were collected.

73

Evidence suggests that several different patient

profiles exist among the DD population. For example,
some have higher symptom levels and poor social
skills; others lower symptom levels, especially negative symptoms, and even better social skills compared
to non-abusing schizophrenia patients (Arndt et al.,
1992; Dixon et al., 1991; Salyers and Mueser, 2001).
Joyal et al. (2003) compared 16 DD to 14 SD men
discharged from a forensic inpatient unit. The DD
group had fewer negative symptoms and less frontal
impairment as measured by soft signs and performance on the WCST (Wisconsin Card Sorting Test)
and COWAT (Controlled Oral Word Association
Test). They suggest that planning and organization
skills are required to obtain substances of abuse and
maintain substance use. Therefore, these patients are
able to continue their addiction in part because of their
better cognitive function and fewer negative symptoms. Thus these different patient profile types may
help to explain some of the differential findings
between studies as this variance is not yet accounted
for in most published studies to date.
The only major between group-differences at
follow-up were the greater percentage of current and
lifetime DD patients who had clinically significant
depression. However, there was no statistically
significant difference in mean group HAM-D scores.
Subjective positive and global symptoms (i.e., BSI
results) were higher among DD patients. Thus, it is
quite evident that DD patients experience more severe
internal distress than their SD counterparts. Similar to
the general population without severe mental illness
(Kendler et al., 1996; Kessler et al., 1994; Regier et
al., 1990), substance use disorders are associated with
higher levels of stress, depression and anxiety among
those with dual disorders. Among this cohort, these
higher rates of clinically significant depression were
seen despite improved antipsychotic and concomitant
medication compliance, suggesting that medication
compliance does not have a strong impact on
depression and anxiety symptoms when patients
continue to use substances of abuse. This is likely
because substance abuse and dependence can increase
or even induce states of depression and anxiety
(Kendler et al., 1996; Kessler et al., 1994; Regier et
al., 1990).
The data from this cohort imply that severity of
psychosis and drug and alcohol use at follow-up

74

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

varied independently, and that 12 months of routine

outpatient treatment for schizophrenia and related
psychotic disorders did not significantly impact
levels of alcohol and drug use. All groups experienced a reduction in objective positive symptoms of
psychosis. In fact, objective positive symptoms
decreased more for DD-current than DD-lifetime
than SD patients. This may be related to the fact that
those patients who remained in treatment wanted
help and thus made an effort to be compliant with
treatment. Thus, antipsychotic medication compliance despite ongoing substance use may result in a
reduction of positive symptoms. Patients that were
lost to follow-up may have continued to escalate
their drug use and/or were more likely to be
medication noncompliant.
Limitations of this study include the design whereby
psychiatric symptoms were only measured at two time
points. Since psychiatric symptoms vary and are
expected to vary with the pattern of substance use,
the intimate relationship between use and expression of
psychiatric symptoms was not adequately explored in
this study. As well, clinical samples that use a
naturalistic follow-up are by their very nature biased.
However, this sample was from a large catchment area
and both the initial baseline refusal rate (14%) and the
portion lost to follow-up (29%) were relatively low,
thus reducing the impact of the inherent bias in the
study design. This study relied on self-report data,
some key measures such as medication compliance and
current substance abuse were not independently
confirmed; thus some patients may have been incorrectly classified as SD instead of DD.
Although this study suggests that standard psychiatric treatment is sufficient to reduce psychosis,
specific substance-related interventions may be required to reduce alcohol and drug consumption and
improve quality of life in this population.

Acknowledgements
The authors would like to thank Drs. Allan
Fielding, Richard Montoro and Warren Steiner for
their assistance in patient recruitment. This research
was supported by a grant awarded to K.G. from the
National Health Research and Development Program
(Canada).

References
Angermeyer, M.C., 2000. Schizophrenia and violence. Acta
Psychiatr. Scand., Suppl. 102, 63 67.
Arndt, S., Tyrrell, G., Flaum, M., Andreasen, N.C., 1992.
Comorbidity of substance abuse and schizophrenia: the role of
pre-morbid adjustment. Psychol. Med. 22, 379 388.
Baker, F., Intagliata, J., 1982. Quality of life in the evaluation of
community support systems. Eval. Program Plann. 5, 69 79.
Bartels, S.J., Drake, R.E., Wallach, M.A., 1995. Long-term course
of substance use disorders among patients with severe mental
illness. Psychiatr. Serv. 46, 248 251.
Bell, M., Greig, T., Gill, P., Whelahan, H., Bryson, G., 2002. Work
rehabilitation and patterns of substance use among persons with
schizophrenia. Psychiatr. Serv. 53, 63 69.
Bhanji, N.H., Chouinard, G., Margolese, H.C., 2004. A review of
compliance, depot intramuscular antipsychotics and the new
long-acting injectable atypical antipsychotic risperidone in
schizophrenia. Eur. Neuropsychopharmacol. 14, 87 92.
Blow, F.C., Barry, K.L., BootsMiller, B.J., Copeland, L.A., McCormick, R., Visnic, S., 1998. Longitudinal assessment of inpatient
use and functioning of seriously mentally ill veterans with and
without co-occurring substance use disorders. J. Psychiatr. Res.
32, 311 319.
BootsMiller, B.J., Ribisl, K.M., Mowbray, C.T., Davidson, W.S.,
Walton, M.A., Herman, S.E., 1998. Methods of ensuring high
follow-up rates: lessons from a longitudinal study of dual
diagnosed participants. Subst. Use Misuse 33, 2665 2685.
Buhler, B., Hambrecht, M., Loffler, W., an der, H.W., Hafner, H.,
2002. Precipitation and determination of the onset and course of
schizophrenia by substance abusea retrospective and prospective study of 232 population-based first illness episodes.
Schizophr. Res. 54, 243 251.
Carise, D., McLellan, A.T., Cacciola, J., Love, M., Cook, T.,
Bovasso, G., Lam, V., 2001. Suggested specifications for a
standardized Addiction Severity Index database. J. Subst. Abuse
Treat. 20, 239 244.
Caspari, D., 1999. Cannabis and schizophrenia: results of a followup study. Eur. Arch. Psychiatry Clin. Neurosci. 249, 45 49.
Chouljian, T.L., Shumway, M., Balancio, E., Dwyer, E.V., Surber,
R., Jacobs, M., 1995. Substance use among schizophrenic
outpatients: prevalence, course, and relation to functional status.
Ann. Clin. Psychiatry 7, 19 24.
Coldham, E.L., Addington, J., Addington, D., 2002. Medication
adherence of individuals with a first episode of psychosis. Acta
Psychiatr. Scand. 106, 286 290.
Derogatis, L.R., Melisaratos, N., 1983. The Brief Symptom
Inventory: an introductory report. Psychol. Med. 13, 595 605.
Dixon, L., Haas, G., Weiden, P.J., Sweeney, J., Frances, A.J., 1991.
Drug abuse in schizophrenic patients: clinical correlates and
reasons for use. Am. J. Psychiatry 148, 224 230.
Dixon, L., McNary, S., Lehman, A.F., 1998. Remission of substance
use disorder among psychiatric inpatients with mental illness.
Am. J. Psychiatry 155, 239 243.
Drake, R.E., Osher, F.C., Wallach, M.A., 1989. Alcohol use and
abuse in schizophrenia. A prospective community study. J. Nerv.
Ment. Dis. 177, 408 414.

H.C. Margolese et al. / Schizophrenia Research 83 (2006) 6575

Drake, R.E., Mueser, K.T., Clark, R.E., Wallach, M.A., 1996. The
course, treatment, and outcome of substance disorder in
persons with severe mental illness. Am. J. Orthopsychiatry
66, 42 51.
Drake, R.E., Mueser, K.T., Brunette, M.F., McHugo, G.J., 2004. A
review of treatments for people with severe mental illnesses and
co-occurring substance use disorders. Psychiatr. Rehabil. J. 27,
360 374.
el Guebaly, N., Hodgins, D.C., 1992. Schizophrenia and substance
abuse: prevalence issues. Can. J. Psychiatry 37, 704 710.
Hamilton, M., 1960. A rating scale for depression. J. Neurol.
Neurosurg. Psychiatry 23, 56 62.
Horley, J.L.B.R., 1985. Affective and cognitive components of
global subjective well-being measures. Soc. Indic. Res. 17,
189 197.
Hunt, G.E., Bergen, J., Bashir, M., 2002. Medication compliance
and comorbid substance abuse in schizophrenia: impact on
community survival 4 years after a relapse. Schizophr. Res. 54,
253 264.
Joyal, C.C., Halle, P., Lapierre, D., Hodgins, S., 2003. Drug abuse
and/or dependence and better neuropsychological performance
in patients with schizophrenia. Schizophr. Res. 63, 297 299.
Kamali, M., Kelly, L., Gervin, M., Browne, S., Larkin, C.,
OCallaghan, E., 2001. Psychopharmacology: insight and
comorbid substance misuse and medication compliance among
patients with schizophrenia. Psychiatr. Serv. 52, 1613, 166.
Kamman, R., Farry, M., Harbison, P., 1983. The analysis and
measurement of happiness as a sense of well-being. Soc. Indic.
Res. 15, 91 115.
Kay, S.R., Opler, L.A., Lindenmayer, J.P., 1988. Reliability and
validity of the Positive and Negative Syndrome Scale for
schizophrenics. Psychiatry Res. 23, 99 110.
Kay, S.R., Opler, L.A., Lindenmayer, J.P., 1989. The Positive and
Negative Syndrome Scale (PANSS): rationale and standardisation. Br. J. Psychiatry Suppl., 59 67.
Kendler, K.S., Gallagher, T.J., Abelson, J.M., Kessler, R.C., 1996.
Lifetime prevalence, demographic risk factors, and diagnostic
validity of nonaffective psychosis as assessed in a US
community sample. The National Comorbidity Survey. Arch.
Gen. Psychiatry 53, 1022 1031.
Kessler, R.C., McGonagle, K.A., Zhao, S., Nelson, C.B., Hughes,
M., Eshleman, S., Wittchen, H.U., Kendler, K.S., 1994. Lifetime
and 12-month prevalence of DSM-III-R psychiatric disorders in
the United States. Results from the National Comorbidity
Survey. Arch. Gen. Psychiatry 51, 8 19.

75

Larsen, R.J., Diener, D., Emmons, R.A., 1985. An evaluation of

subjective well-being measures. Soc. Indic. Res. 17, 1 17.
Margolese, H.C., Malchy, L., Negrete, J.C., Tempier, R., Gill, K.,
2004. Drug and alcohol use among patients with schizophrenia
and related psychoses: levels and consequences. Schizophr. Res.
67, 157 166.
McLellan, A.T., Luborsky, L., Woody, G.E., OBrien, C.P., 1980.
An improved diagnostic evaluation instrument for substance
abuse patients. The Addiction Severity Index. J. Nerv. Ment.
Dis. 168, 26 33.
McLellan, A.T., Luborsky, L., Cacciola, J., Griffith, J., Evans, F.,
Barr, H.L., OBrien, C.P., 1985. New data from the Addiction
Severity Index. Reliability and validity in three centers. J. Nerv.
Ment. Dis. 173, 412 423.
Mercier, C., Tempier, R., Renaud, C., 1992. Community services
and quality of life: a study of the impact in a remote region. Can.
J. Psychiatry 37, 553 563.
Negrete, J.C., Knapp, W.P., Douglas, D.E., Smith, W.B., 1986.
Cannabis affects the severity of schizophrenic symptoms: results
of a clinical survey. Psychol. Med. 16, 515 520.
Olfson, M., Mechanic, D., Hansell, S., Boyer, C.A., Walkup, J.,
Weiden, P.J., 2000. Predicting medication noncompliance after
hospital discharge among patients with schizophrenia. Psychiatr.
Serv. 51, 216 222.
Overall, J.E., Gorman, D.R., 1962. The brief psychiatric rating
scale. Psychol. Rep. 10, 799 802.
Pulver, A.E., Wolyniec, P.S., Wagner, M.G., Moorman, C.C.,
McGrath, J.A., 1989. An epidemiologic investigation of
alcohol-dependent schizophrenics. Acta Psychiatr. Scand. 79,
603 612.
Regier, D.A., Farmer, M.E., Rae, D.S., Locke, B.Z., Keith, S.J.,
Judd, L.L., Goodwin, F.K., 1990. Comorbidity of mental
disorders with alcohol and other drug abuse. Results from the
Epidemiologic Catchment Area (ECA) Study. JAMA 264,
2511 2518.
Salyers, M.P., Mueser, K.T., 2001. Social functioning, psychopathology, and medication side effects in relation to substance use
and abuse in schizophrenia. Schizophr. Res. 48, 109 123.
Soyka, M., 2000. Substance misuse, psychiatric disorder and violent
and disturbed behaviour. Br. J. Psychiatry 176, 345 350.
Swartz, M.S., Swanson, J.W., Hiday, V.A., Borum, R., Wagner,
H.R., Burns, B.J., 1998. Violence and severe mental illness: the
effects of substance abuse and nonadherence to medication. Am.
J. Psychiatry 155, 226 231.