Schizophrenia Research 147 (2013) 132–139

Contents lists available at SciVerse ScienceDirect

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

Comorbid substance abuse in first-episode schizophrenia: Effects on cognition and

psychopathology in the EUFEST study
T. Wobrock a, b,⁎, P. Falkai c, T. Schneider-Axmann c, A. Hasan c, S. Galderisi d, M. Davidson e, R.S. Kahn f,
E.M. Derks f, H. Boter g, J.K. Rybakowski h, J. Libiger i, S. Dollfus j, J.J. López-Ibor k, J. Peuskens l, L.G. Hranov m,
W. Gaebel n, W.W. Fleischhacker o, EUFEST study group
a
Department of Psychiatry and Psychotherapy, University Medical Centre, Georg-August-University, Göttingen, Germany
b
Centre of Mental Health, Darmstadt-Dieburg Clinics, Groß-Umstadt, Germany
c
Department of Psychiatry and Psychotherapy, Ludwig-Maximilians-University, Munich, Germany
d
Department of Psychiatry, University of Naples SUN, Naples, Italy
e
Department of Psychiatry, Sheba Medical Center, Tel-Hashomer, Israel
f
Rudolf Magnus Institute of Neuroscience, Department of Psychiatry, University Medical Center Utrecht, The Netherlands
g
Department of Epidemiology (unit HTA), University of Groningen, University Medical Centre Groningen, Groningen, The Netherlands
h
Poznan University of Medical Sciences, Poznan, Poland
i
Department of Psychiatry, University Hospital Hradec Kralove, Czech Republic
j
Department of Adult Psychiatry, Centre Hospitalier Universitaire, Centre Esquirol, Caen, France
k
Institute of Psychiatry and Mental Health, San Carlos Clinical University Hospital of Madrid, Spain
l
University Psychiatric Centre, Katholieke Universiteit Leuven, Campus St. Jozef Kortenberg, Leuven, Belgium
m
Department of Psychiatry, University Hospital of Neurology and Psychiatry, St. Naum, Sofia, Bulgaria
n
Department of Psychiatry and Psychotherapy, Heinrich-Heine-University, Düsseldorf, Germany
o
Department of Psychiatry, Department of Biological Psychiatry, Medical University of Innsbruck, Austria

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

Article history: Studies and meta-analyses investigating the influence of substance use disorder (SUD) (substance abuse or
Received 24 August 2012 dependence) on psychopathology and neurocognitive function in schizophrenia patients have revealed con-
Received in revised form 28 February 2013 troversial results. Most studies did only have small samples and did not focus exclusively on first-episode
Accepted 1 March 2013
schizophrenia patients.
Available online 26 March 2013
Method: In a post-hoc analysis of the European First Episode Schizophrenia Trial (EUFEST) psychopathology and
Keywords:
cognitive performances of patients with (FE-SUD, N = 119, consisting of N = 88 patients with persisting SUD at
Substance abuse baseline and N = 31 patients with previous SUD) and without SUD (FE-non-SUD, N = 204) were compared at
First-episode schizophrenia baseline and 6 months follow-up. Neurocognitive assessment included the Rey Auditory Verbal Learning Test
Cognitive function (RAVLT); Trail Making Tests A and B (TMT), Purdue Pegboard and Digit-Symbol Coding.
Dual diagnosis Results: In total 31.1% of patients reported SUD, and 22.2% of patients used cannabis. There were no significant
differences between patients with and without SUD concerning PANSS scores, extrapyramidal motor symptoms
or neurocognitive measures except better performance in psychomotor speed (TMT-A, p = 0.033, Cohen's
d = 0.26) in patients with SUD at 6 months follow-up. Interestingly, SUD patients with ongoing substance
use at follow-up showed elevated positive symptoms (PANSS positive score, p = 0.008, Cohen's d = 0.84)
compared to those who abstained. PANSS scores at baseline were increased in patients with an onset of SUD
before the age of 16 years. In addition we found a correlation between longer duration of cannabis use and
higher cognitive performance as well as reduced symptom improvement and more extrapyramidal motor
symptoms in patients with higher frequency of cannabis consumption.
Conclusions: FE-SUD and FE-non-SUD show similar psychopathology and neuropsychological performances at
baseline and during the first 6 months of antipsychotic treatment.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction

⁎ Corresponding author at: Centre of Mental Health, Darmstadt-Dieburg Clinics Cognitive deficits can be considered as a core deficit of schizophrenia
and Department of Psychiatry and Psychotherapy, University Medical Centre, Georg-
August-University Göttingen, Krankenhausstrasse 7, 64823 Groß-Umstadt, Germany.
and have been found even in untreated, first-episode patients (Heinrichs
Tel.: + 49 6078 79 2901; fax: + 49 6078 79 1836. and Zakzanis, 1998; Torrey, 2002). These impairments account for sig-
E-mail address: t.wobrock@kreiskliniken-dadi.de (T. Wobrock). nificant variance in measures of functional status (Green, 1996) and

0920-9964/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.schres.2013.03.001
 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139 133

cognitive functioning is a correlate of global and specific functional out- previous year or six weeks lifetime, iii) having a known intolerance
come in schizophrenia and contributes to poor judgment and lack of in- to one of the study drugs; and iv) meeting any of the contraindica-
sight (Jensen et al., 2004). tions for any of the study drugs.
Substance abuse in individuals with schizophrenia is the most prev- Sociodemographic and clinical parameters were obtained including
alent comorbid psychiatric condition with life-time prevalence between standardized assessment of psychopathology (Positive and Negative
10 and 65% (Kovasznay et al., 1997; Buckley et al., 2009) and has been Syndrome [PANSS] (Kay et al., 1987), Calgary Depression Rating Scale
associated with elevated positive symptoms (Soyka et al., 2001), lower for Schizophrenia [CDSS] (Addington et al., 1993)) disease severity
negative symptoms (Soyka et al., 2001), and reduced positive and nega- (Clinical Global Impression) [CGI] (Guy W, 1976), psychosocial func-
tive symptoms (Dixon et al., 1991). tioning (Global Assessment of Functioning [GAF] (Guy W, 1976)), and
Substance abuse and dependence (substance use disorder, SUD) extrapyramidal symptoms (St. Hans Rating Scale [SHRS] (Gerlach
are associated with deleterious consequences on cognitive function- et al., 1993)). The trial was conducted in accordance with the Decla-
ing. This is frequently reported for patients abusing alcohol and co- ration of Helsinki, Good Clinical Practice, and national regulatory
caine (Beatty et al., 1995), but is also described in patients with requirements.
long-term cannabis abuse (Pope et al., 1995). 498 patients were randomly assigned via a centralized, computer-
Despite the high prevalence of comorbid SUD in schizophrenia, ized online randomization system to open-label treatment with halo-
knowledge about the influence of substance abuse on neurocognitive peridol, 1 to 4 mg/day (N = 103); amisulpride, 200 to 800 mg/day
function is still limited and various studies revealed inconclusive results (N = 104); olanzapine, 5 to 20 mg/day (N = 105); quetiapine, 200
(Coulston et al., 2007). In some studies patients with comorbid SUD to 750 mg/day (N = 104); or ziprasidone, 40 to 160 mg/day (N = 82).
(mainly cocaine and alcohol) presented greater cognitive deficits than The treating physician determined the antipsychotic dose based on
patients without substance abuse, in other studies no differences be- his clinical experience, providing that it remained within the range
tween users and nonusers or even better performance of patients with above. The diagnosis of substance abuse and dependence was based
concomitant SUD have been demonstrated (Coulston et al., 2007). In a on DSM-IV criteria.
meta-analysis, subjects with schizophrenia and comorbid substance
abuse displayed similar cognitive performance as schizophrenia patients 2.2. Neuropsychological investigation
without substance abuse (Yucel et al., 2012). In a further meta-analysis
focussing on the effects of cannabis on neurocognition in schizophrenia Verbal memory was measured by the Rey Auditory Verbal Learning
cannabis-using patients performed better than patients with no history Test (RAVLT, (Rey, 1964)), the ability of complex visual scanning and
of cannabis use with small to medium effect sizes (Rabin et al., 2011). psychomotor speed by the Trail Making Test (TMT), Part A and addi-
A recent study including nearly thousand patients with non-affective tionally cognitive flexibility by Part B (Reitan and Wolfson, 1993), visual
psychosis found that lifetime cannabis use was associated with better motor skill/dexterity, processing speed, attention/concentration, visual
performance on acquired knowledge, facial affect recognition and face perception of abstract stimuli and short-term visual memory by Digit
identity recognition while current cannabis use was associated with Symbol Coding as part of the WAIS-III (Wechsler, 1997), and motor
poorer performance on immediate verbal learning, processing speed speed and motor coordination by the Purdue Pegboard Test (Costa
and working memory (Meijer et al., 2012). et al., 1963). Tests are described in more detail in a previous publication
One should note that most studies included only patients with a (Galderisi et al., 2009).
chronic disease course and a long duration of illness, but first-episode The tests were administered at baseline (visit 1) and after 6 months
and recent onset patients haven't been investigated extensively (Yucel of treatment (visit 7) (±3 weeks). For patients in whom the severity of
et al., 2012). In a comparison of first-episode patients with and without psychosis at baseline was such that cognitive testing was not feasible,
previous cannabis abuse, comorbid patients demonstrated slightly bet- testing could be postponed until the end of the first month of the trial.
ter (Wobrock et al., 2007), showed significant superior cognitive perfor- The tests were administered by psychologists and psychiatrists trained
mance in visual memory, working memory and executive functioning at investigators meetings to administer the specified tests.
(Yucel et al., 2012), had better attention and executive functions
(Rodriguez-Sanchez et al., 2010) or demonstrated better cognitive 2.3. Statistical analyses
task performance on all measures except working memory manipu-
lation (Leeson et al., 2012). All statistical analyses were performed with SPSS for Windows,
The aim of this post-hoc analysis of the EUFEST study was to ex- version 16. The significance level was set at α = 0.05. All tests were
amine the impact of comorbid substance abuse on neurocognition two-tailed.
in this large sample of first-episode patients. Additionally, the sub- The independent factor was substance use disorder (SUD), dividing
groups with and without comorbid SUD were compared in terms of the sample into two groups of patients: patients with lifetime SUD (cur-
positive, negative and depressive psychopathology and extrapyrami- rent abuse/dependence or history of abuse/dependence) and patients
dal motor symptoms. The mentioned variables were compared at without lifetime substance use disorder (non-SUD), excluding patients
baseline and 6-months follow-up to evaluate the stability or variabil- with alcohol dependence during the last 6 months before study entry.
ity of cognitive and clinical patterns in both subgroups over time. Drop-out was calculated after six months of antipsychotic treatment
and compared between SUD and non-SUD with a Chi² test. In the large
2. Methods group of patients from Romania (n = 112) there was only one patient
with substance abuse. As there were various sociodemographic differ-
2.1. Study design ences between patients from other countries and Romania, the latter
group was excluded from further analysis to avoid a bias.
The EUFEST study design is described in detail elsewhere (Kahn et al., Descriptive statistics were reported as mean ± standard devia-
2008). In brief, a total of 50 centers participated in 13 European countries tion. ANOVA was used to compare age and years of education be-
and Israel. To be eligible for inclusion, patients (18–40 years) had to meet tween the patients with SUD and patients without SUD. Chi² tests
DSM-IV criteria for schizophrenia, schizophreniform, or schizoaffective were used to compare the distribution of several sociodemographic
disorder as confirmed by the Mini International Neuropsychiatric Inter- variables between the groups. For the comparison of psychopatho-
view Plus (MINI +) (Sheehan et al., 1998). Exclusion criteria were logical items and neurocognitive variables at baseline and 6 months
i) more than two years had passed since the onset of positive symp- follow-up between SUD (fulfilling criteria for lifetime substance
toms, ii) the use of any antipsychotic exceeded two weeks in the abuse/dependence with still persisting or already stopped substance
134 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139

consumption at study entry) and non-SUD, analysis of covariance was inpatients (n = 282, 36.2% reporting SUD) and of Caucasian ethnic-
(ANCOVA) adjusted for age, education and gender was performed ity (92.6%, n = 299; 36.5% of them reporting SUD).
as main analysis. The same design was used comparing patients Substance abuse persisted in 26 patients after 6 months, 62 patients
with current SUD and without lifetime SUD (excluding the patients were able to stop substance consumption during the study, and 31
with previous SUD). These results were checked by analyzing a sam- patients showed lifetime abuse but had no persisting abuse at the
ple consisting of 88 patients with current substance abuse at base- beginning of the study.
line and 88 patients without lifetime SUD matched for gender, age
and duration of education. 3.1. Sociodemographic variables
As additional analyses, the ANCOVA design described above was
used to compare neurocognitive test results and psychopathology be- Sociodemographic and clinical characteristics at baseline divided
tween patients with cannabis use solely and patients consuming canna- by the subgroups of FE schizophrenia patients with comorbid lifetime
bis and other illegal drugs. In a next analysis, dependent variables were SUD and without substance abuse are presented in Table 2. Compared
compared between patients with persisting substance consumption at to nonabusers the sample of substance abusers was younger (23.9 ±
follow-up and patients that stopped substance consumption during 4.4 years vs. 26.4 ± 5.6 years; df = 1, 319, F = 12.1, p = 0.001),
the course of the study. Patients with lifetime SUD were grouped into was predominantly male (81.5% vs. 51.5%; df = 1, Chi 2 = 29.0;
patients with early (age ≤ 16 years) and late (age > 16 years) begin- p b 0.001), had a lower level of education (11.4 ± 2.3 years vs.
ning of regular cannabis abuse. Neurocognitive variables and psychopa- 13.0 ± 2.9 years; df = 1, 319, F = 18.4, p b 0.001), was less married
thology were compared between early and late onset and they were (0.8% vs. 15.2%, Chi2 = 17.4; p b 0.001) and less employed (30.3% vs.
correlated with the age at regular cannabis onset and duration of regu- 52.5%, df = 1, Chi2 = 15.0; p b 0.001) and had a higher income (633
lar cannabis use calculating partial correlations adjusted for age, gender Euro vs. 399 Euro, df = 1, 115, F = 15.27, p = b 0.001).
and education.
Data of patients with and without lifetime alcohol abuse or depen- 3.2. Drop-out rates
dence were compared using the described ANCOVA design, while pa-
tients with lifetime SUD diagnosis were excluded from this analysis. The percentage of drop-outs between both subgroups at follow-up
Due to the explorative character of this study the results are was similar (SUD: 29.4%, non-SUD: 27.4%; df = 1, Chi² = 0.15, p =
presented without error probability correction for multiple testing. 0.70). However, the percentage of patients taking a dose of the study
drug lower than advised to was significantly higher in the SUD group
3. Results (25.0% vs. 9.5%; df = 1, Chi² = 6.66, p = 0.01).

In the total sample of 498 patients 31.1% reported SUD or AUD 3.3. Psychopathology
according to DSM-IV criteria, 22.2% of patients used cannabis, 7.6%
solely cannabis, 5.4% in combination with alcohol, 5.2% in combina- At baseline and after 6 months there were no significant differences
tion with other illegal substances (mainly amphetamines, XTC and between patients with and without SUD concerning PANSS scores or ex-
cocaine), and 4.0% in combination with both. Two patients (0.4%) trapyramidal motor symptoms (see Table 2). At 6 month follow-up
reported no consumption of cannabis while using illegal substances, there was a trend towards a lower PANSS negative score in patients
2 patients (0.4%) used illegal substances in combination with alcohol with SUD (15.7 ± 5.9 vs. 16.2 ± 6.0, F = 3.2, df = 1, 222, p = 0.074).
and 8.0% of patients had only alcohol use disorder (AUD). In 4% of pa- Comparing only patients with and without current SUD at baseline,
tients the used substance was not clearly documented. In Table 1 there were still no significant differences in psychopathology. However,
more details about the number of patients using each substance, there was a trend for higher positive symptoms at baseline in patients
age of onset, frequency and total time of substance consumption are with current SUD (F = 3.2, df = 1, 283, p = 0.075).
presented. The analysis of the sample of 88 patients with current substance
As described in the Methods section, for the extremely reduced fre- abuse at baseline and 88 patients without lifetime SUD matched
quency of SUD in Romania (0.9%, n = 1) compared to the other partic- for gender, age and education showed no significant differences in
ipating countries these patients were excluded from further analysis to psychopathology.
avoid a bias. In addition, patients with SUD fulfilling criteria for alcohol In addition we compared the group of patients with persisting sub-
dependence during the last 6 months before study entry and patients stance consumption at follow-up (n = 26) with those who abstained
with only alcohol abuse/dependence (alcohol use disorder, AUD) but (n = 62). Patients with previous SUD (n = 31) were not included in
no SUD was excluded from main analysis because current or previous this analysis. We found no difference between the groups regarding
alcohol dependence may be a substantial confound influencing cogni-
tive performance in schizophrenia patients. Table 1
In our study sample for further analysis lifetime substance abuse/ Substance abuse pattern at baseline.
dependence (substance use disorder, SUD) according to DSM-IV
Substance Subjects with Subjects with Age at Years of Frequency
criteria was reported in 36.8% (n = 119) of the included first-episode persisting use previous use regular regular of use
schizophrenia patients (average rate without Romania). The rates of onset use (times per
SUD in the study samples of the participating countries were: 85.7% month)
(n = 12) in Germany, 85.7% (n = 6) in Belgium, 73.7% (n = 14) in m sd m sd m sd
France, 56.5% (n = 13) in Austria, 76.5% (n = 13) in The Netherlands, Cannabis 85 30 17.5 2.8 4.6 3.9 73.9 120.3
44.4% (n = 8) in Spain, 35.7% (n = 20) in Israel, 37.1% (n = 13) in Cocaine 16 5 20.8 4.2 2.2 2.3 13.8 17.9
Italy, 40.0% (n = 2) in Switzerland, 15.4% (n = 2) in Bulgaria, 18.5% Amphetamine 13 5 19.9 4.5 1.9 2.1 4.8 4.7
(n = 5) in Czech Republic and 10.2% (n = 10) in Poland. or other
stimulants
The frequencies of SUD in the diagnostic subcategories were: 33.3%
Heroin 4 1 18.0 3.5 5.4 5.3 10.0 7.2
for 295.1, 0% for 295.2, 37.1% for 295.3, 41.2% for 295.4, 25.0% for 295.7, XTC 17 2 17.9 2.7 1.5 0.9 3.9 5.5
and 26.3% for 295.9, revealing no significant differences concerning the LSD or other 3 0 19.0 2.6 5.7 5.7 1.1 0.6
frequency of SUD between, patients with first-episode schizophrenia hallucinogens
(35.6%), schizoaffective disorder (25.0%), and schizophreniform disorder Legend: m = mean, sd = standard deviation, XTC = ecstasy, LSD = lysergic acid
(41.2%). At baseline the vast majority of the study population (87.3%) diethylamide.
 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139 135

disease severity (CGI), depressive symptoms and extrapyramidal motor Table 2

symptoms, but those SUD patients with ongoing substance use at Comparison of sociodemographic and clinical parameters between subgroups at base-
line and 6 months follow-up.
6 months follow-up showed elevated positive symptoms (PANSS positive
score, 13.67 ± 4.4 vs. 9.94 ± 4.5, F = 7.6, df = 1, 54, p = 0.008, Variable FE-SZ FE-SZ Chi-square test
Cohen's d = 0.84). Patients using multiple substances demonstrated in- without with SUD
SUD
creased psychopathology and extrapyramidal motor symptoms (Table 3).
After exclusion of patients with lifetime SUD diagnosis, patients Sociodemographics at baseline N = 204 N = 119 df Chi² p
Gender (male/female) 105/99 97/22 1 29.0 b0.0001
with only alcohol use disorder showed no significant difference com-
pared to patients without alcohol use disorder. ANOVA
m sd m sd df F p
3.4. Cognition Age (years) 26.4 5.6 23.9 4.4 1, 12.1 0.001
319
Education (in years) 13.0 2.9 11.4 2.3 1, 18.4 b0.0001
First-episode schizophrenia patients with and without SUD showed a
319
similar performance in the assessed cognitive domains at baseline
(Table 4). There was only a trend for an increased forgetting of the initial Baseline N = 204 N = 119 ANCOVA
learned words in the RAVLT (difference of trials 5 and 7, RAVLT 1.3 ± PANSS positive subscale 22.3 6.2 23.3 6.4 1, 0.08 0.78
314
2.1, vs. 0.8 ± 2.3, df = 1, 269, F = 3.28, p = 0.071). At 6 months
PANSS negative subscale 22.0 7.5 20.5 8.1 1, 2.05 0.15
follow-up (Table 5) patients with SUD showed a faster performance in 314
TMT-A (time 34.4 ± 11.7 s. vs. 38.8 ± 21.0 s, F = 4.59, df = 1, 195, PANSS general psychopathology 44.9 11.2 43.5 11.2 1, 0.71 0.40
p = 0.033, Cohen's d = 0.26), while there was a trend towards a subscale 314
slower performance in TMT-B (time 88.9 ± 51.1 s. vs. 81.9 ± 40.8 s, PANSS total scale 89.1 21.7 87.3 21.5 1, 0.76 0.38
314
F = 2.83, df = 1, 195, p = 0.094). Comparing only patients with and
GAF 40.8 13.1 38.9 13.9 1, 0.04 0.84
without current SUD at baseline, for all neuropsychological test results 317
there were only some trends but no significant differences. CGI severity 4.8 0.8 4.8 0.9 1, 0.74 0.39
From the analysis of the sample matched for gender, age and du- 317
CDSS sum score 5.2 4.7 5.0 4.8 1, 0.01 0.93
ration of education it resulted a better performance of SUD patients
316
at the beginning of the study in the TMT-A (time 33.9 ± 11.0 s. vs. SHRS akathisia 0.34 0.88 0.35 0.90 1, 3.17 0.076
40.2 ± 14.7 s, F = 6.40, df = 1, 107, p = 0.013). 317
To detect the influence of the substance abuse pattern on SHRS dystonia 0.06 0.37 0.08 0.45 1, 0.06 0.80
neurocognition we compared the patients using solely cannabis 317
SHRS parkinsonism 0.27 0.67 0.21 0.65 1, 0.15 0.70
with the patients consuming cannabis and other illegal substances. The
317
only significant difference was a slightly better performance of the SHRS dyskinesia 0.02 0.17 0.00 0.00 1, 0.96 0.33
patients with solely cannabis abuse in verbal memory resulting in 317
a higher sum of correct recalled words at 6-months follow-up (RAVLT tri-
6 months N = 149 N = 84 ANCOVA
als 1–5, 49.9 ± 8.71 vs. 41.2 ± 10.08, F = 7.4, df = 1, 45, p = 0.009).
PANSS positive subscale 11.0 4.7 12.2 5.4 1, 0.06 0.81
In addition we compared the group of patients with persisting 222
substance consumption at follow-up (n = 26) with patients stopping PANSS negative subscale 16.2 6.0 15.7 5.9 1, 3.22 0.074
it (n = 62). We found no difference between the groups regarding 222
cognitive performance. PANSS general psychopathology 28.5 8.8 28.4 8.9 1, 0.52 0.47
subscale 222
After exclusion of patients with lifetime SUD diagnosis, patients
PANSS total scale 55.7 17.0 56.4 17.5 1, 0.86 0.35
with only alcohol use disorder (n = 22) demonstrated a significantly 222
reduced performance at baseline in the TMT-B error rate (1.6 ± 3.5 GAF 68.1 14.8 64.3 17.3 1, 0.01 0.91
vs. 0.9 ± 1.7; F = 7.29; df = 1, 239; p = 0.007, Cohen's d = 0.24) 227
and in the error difference between TMT-B and TMT-A (0.9 ± 3.1 CGI severity 2.7 1.1 3.0 1.3 1, 0.33 0.57
227
vs. 0.7 ± 1.6; F = 4.35; df = 1, 239; p = 0.038; Cohen's d = 0.07) CDSS sum score 2.4 3.5 2.4 2.9 1, 0.09 0.76
compared to patients without AUD (n = 223). 223
SHRS akathisia 0.26 0.80 0.15 0.53 1, 1.50 0.22
3.5. Influence of onset, duration and frequency of cannabis abuse 222
SHRS dystonia 0.01 0.08 0.03 0.16 1, 0.15 0.69
222
Since time of onset of cannabis use (before or after puberty/ SHRS parkinsonism 0.25 0.62 0.18 0.50 1, 0.99 0.32
adolescence) may have an impact on neurocognition and psycho- 222
pathology we divided the SUD group in those with starting cannabis SHRS dyskinesia 0.02 0.14 0.01 0.11 1, 0.49 0.48
use older than the age of 16 years and those beginning with or until 222

16 years. We could not detect differences in psychopathology between Legend: FE-SZ = first episode schizophrenia patients, SUD = substance use disorder,
these groups. However, patients with an onset of SUD up to 16 years ANOVA = analysis of variance, ANCOVA = analysis of covariance with independent
factors SUD and gender and covariates age and duration of education, m = mean,
performed slightly worse at 6-months follow-up in RAVLT (difference
sd = standard deviation, F = F statistic, Chi² = Chi square statistic, df = degrees of
trial 5 minus 7 increased, 2.25 ± 2.97 vs. 0.61 ± 2.11, F = 5.1, df = 1, freedom, p = error probability of falsely rejecting the null hypothesis, that there are
52, p = 0.029, Cohen's d = 0.64). PANSS scores at baseline were in- no mean differences between study groups. PANSS = Positive and Negative Syndrome
creased in patients with an onset of SUD up to 16 years (positive Scale, GAF = Global Assessment of Functioning, CGI = Clinical Global Impression,
CDSS = Calgary Depression Scale for Schizophrenia, SHRS = St. Hans Rating Scale.
score: 23.6 ± 6.7 vs. 22.7 ± 6.2, F = 2.8, df = 1, 98, p = 0.095, trend
Data are presented as mean ± standard deviation unless otherwise indicated.
level, general score: 44.2 ± 13.0 vs. 42.4 ± 10.6, F = 4.8, df = 1, 98,
p = 0.030, Cohen's d = 0.15, total score: 87.4 ± 23.9 vs. 85.7 ± 20.9,
F = 3.1, df = 1, 98, p = 0.080, trend level), however they also showed score: 17.2 ± 14.3 (≤16 years) vs. 13.5 ± 12.9 (>16 years), F = 8.2,
a stronger decrease of PANSS scores as compared follow-up with df = 1, 63, p = 0.006, Cohen's d = 0.28, total score: 34.9 ± 26.3
baseline data (positive score: 12.8 ± 7.0 (≤16 years) vs. 10.0 ± 7.8 (≤16 years) vs. 28.5 ± 24.2 (>16 years), F = 7.5, df = 1, 63, p =
(>16 years), F = 4.6, df = 1, 63, p = 0.036, Cohen's d = 0.39, general 0.008, Cohen's d = 0.25).
136 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139

Table 3 Partial correlations adjusted for age, gender and education resulted
Comparison of clinical parameters between patients with only cannabis abuse and pa- in significant correlations between age at onset of regular cannabis
tients using also other substances at baseline and 6 months follow-up.
use and RAVLT difference trial 5 minus 7 at follow-up (r(x,y)|z =
Variable FE-SZ with FE-SZ using ANCOVA −0.33, df = 53, p = 0.015) and CDSS score at baseline (r(x,y)|z =
only (also) −0.22, df = 100, p = 0.024), suggesting higher forgetting and depres-
cannabis other
sion in first-episode patients starting their cannabis consumption earli-
use substances
er. Duration of cannabis use correlated significantly with difference of
m sd m sd F df p
time required for TMT-B and TMT-A at baseline (r(x,y)|z = −0.21,
Baseline N = 71 N = 48 df = 92, p = 0.042), and with the number of digits, raw score
PANSS positive subscale 22.6 6.4 24.2 6.3 0.7 1, 111 0.41 skipped at follow-up (r(x,y)|z = − 0.44, df = 54, p = 0.001), dem-
PANSS negative subscale 20.1 8.1 21.0 8.1 2.3 1, 111 0.14
PANSS general 41.9 10.0 45.9 12.7 5.4 1, 111 0.021
onstrating a correlation between longer duration of cannabis use and
psychopathology subscale higher cognitive performance. The frequency of regular cannabis
PANSS total scale 84.7 20.5 91.1 22.7 4.1 1, 111 0.046 use correlated significantly with change in follow-up compared to
GAF 39.6 14.5 38.0 13.0 1.5 1, 113 0.22 baseline of PANSS general score (r(x,y)|z = −0.30, df = 53, p =
CGI severity 4.75 0.9 4.94 0.8 3.5 1, 113 0.065
0.028) and PANSS total score (r(x,y)|z = −0.29, df = 92, p = 0.035)
CDSS sum score 4.63 4.4 5.58 5.3 2.4 1, 112 0.13
SHRS akathisia 0.20 0.6 0.58 1.2 8.3 1, 113 0.004 and with SHRS akathisia symptoms at baseline (r(x,y)|z = 0.26, df =
SHRS dystonia 0.10 0.5 0.04 0.3 0.0 1, 113 1.00 82, p = 0.015), pointing to reduced symptom improvement and more
SHRS parkinsonism 0.14 0.4 0.31 0.9 3.8 1, 113 0.055 extrapyramidal motor symptoms in patients with higher frequency of
SHRS dyskinesia 0.00 0.0 0.00 0.0 0.0 1, 113 1.00 cannabis consumption.
6 months N = 51 N = 33
PANSS positive subscale 11.2 5.2 13.7 5.5 8.5 1, 74 0.005 3.6. Chlorpromazine equivalents
PANSS negative subscale 15.0 5.3 16.9 6.7 2.5 1, 74 0.15
PANSS general 26.6 7.8 31.4 9.9 10.1 1, 74 0.002
Since it has been demonstrated that acute or chronic administration
psychopathology subscale
PANSS total scale 52.8 16.2 61.9 18.3 9.7 1, 74 0.003
of antipsychotic medication, benzodiazepines and anticholinergic drugs
GAF 67.9 16.3 58.7 17.6 9.9 1, 78 0.002 impairs performance on neurocognitive tasks requiring vigilance, atten-
CGI severity 2.80 1.2 3.21 1.2 5.2 1, 78 0.025 tion, and especially motor speed in schizophrenic patients the pattern
CGI2 (CGI change) 2.20 1.3 2.61 1.3 3.9 1, 78 0.051 (percentage of patients receiving haloperidol, olanzapine, amisulprid,
CDSS sum score 2.35 2.9 2.53 3.0 0.6 1, 75 0.45
quetiapine, and ziprasidone) and dosage of administered antipsychotic
SHRS akathisia 0.16 0.5 0.13 0.6 0.3 1, 74 0.58
SHRS dystonia 0.04 0.2 0.00 0.0 0.5 1, 74 0.50 (in chlorpromazine equivalents (Woods, 2003)) were compared be-
SHRS parkinsonism 0.27 0.6 0.03 0.2 5.3 1, 74 0.024 tween abusers (SUD) and nonabusers (non-SUD). There were no signif-
SHRS dyskinesia 0.02 0.1 0.00 0.0 0.2 1, 74 0.64 icant differences between these subgroups in regard to these variables.
Legend: FE-SZ = first episode schizophrenia patients, SUD = substance use disorder,
ANCOVA = analysis of covariance with independent factor kind of SUD and gender and 4. Discussion
covariates age and duration of education, m = mean, sd = standard deviation, F = F sta-
tistic, df = degrees of freedom, p = error probability of falsely rejecting the null hypoth-
esis, that there are no mean differences between study groups. PANSS = Positive and In a previous analysis neurocognitive measures at baseline were
Negative Syndrome Scale, GAF = Global Assessment of Functioning, CGI = Clinical Glob- compared between the EUFEST patients and two-hundred-twenty
al Impression, CDSS = Calgary Depression Scale for Schizophrenia, SHRS = St. Hans Rat- healthy control subjects showing reduced performance of patients
ing Scale. Data are presented as mean ± standard deviation.
on all tests (Galderisi et al., 2009). The aims of the presented analysis

Table 4
Comparison of neuropsychological performance between subgroups at baseline.

FE-SZ without SUD FE-SZ with SUD ANCOVA

Test m sd m sd df F p

Trail Making Test (TMT) N = 186 N = 109

TMT A (time to complete, in s) 44.0 21.3 45.9 25.9 1, 289 1.98 0.16
TMT B (time to complete, in s) 104.0 85.5 115.0 72.8 1, 289 0.48 0.49
Difference TMT B − TMT A (in s) 60.0 77.2 69.1 61.9 1, 289 0.10 0.75
TMT A (total no. of errors) 0.24 0.84 0.34 0.95 1, 285 0.74 0.39
TMT B (total no. of errors) 1.01 1.83 1.22 1.63 1, 285 0.74 0.39
Difference TMT B − TMT A (errors) 0.77 1.65 0.89 1.57 1, 285 0.20 0.66

Digit Symbol Test (DST) N = 186 N = 103

Digit Symbol Test (no. of digits, raw score correct) 54.3 21.4 50.4 16.2 1, 283 0.26 0.61
Digit Symbol Test (no. of digits, raw score skipped) 2.8 17.8 1.2 10.5 1, 283 0.12 0.73
Digit Symbol Test (no. of digits, raw score wrong) 0.6 1.9 0.4 1.3 1, 283 0.96 0.33

Purdue Pegboard Test N = 183 N = 105

Purdue Pegboard (no. of pegs placed corr., dominant hand) 12.7 2.6 12.7 2.6 1, 282 2.12 0.15
Purdue Pegboard (no. of pegs placed corr., non-dominant hand) 12.0 2.6 11.9 2.4 1, 282 0.16 0.69
Purdue Pegboard (no. pegs placed corr., both hands simult.) 10.5 3.2 10.4 2.5 1, 282 0.22 0.64

Rey Auditory Verbal Learning Test (RAVLT) N = 174 N = 101

RAVLT, sum of trials 1–5 (recall, no. of correct responses) 43.3 12.3 41.5 10.8 1, 269 0.88 0.35
RAVLT, trial 1 (recall, no. of correct responses) 6.0 2.1 5.5 1.8 1, 269 0.00 0.95
RAVLT, trial 7 (delayed recall, no. of correct responses) 9.2 3.6 8.4 3.2 1, 269 0.01 0.94
RAVLT, difference of trials 5–7 (forgetting) 0.8 2.3 1.3 2.1 1, 269 3.28 0.071

Legend: FE-SZ = first episode schizophrenia patients, SUD = substance use disorder, no. = number, s = seconds, ANCOVA = analysis of covariance with independent factors
SUD and gender and covariates age and duration of education, m = mean, sd = standard deviation, F = F statistic, df = degrees of freedom, p = error probability of falsely
rejecting the null hypothesis, that there are no mean differences between study groups. Data are presented as mean ± standard deviation.
 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139 137

Table 5
Comparison of neuropsychological performance between subgroups at 6 months follow-up.

FE-SZ without SUD FE-SZ with SUD ANCOVA

Test m sd m sd df F p

Trail Making Test (TMT) N = 131 N = 70

TMT A (time to complete, in s) 38.8 21.0 34.4 11.7 1, 195 4.59 0.033
TMT B (time to complete, in s) 81.9 40.8 88.9 51.1 1, 195 2.83 0.094
Difference TMT B − TMT A (in s) 43.1 33.1 54.5 44.7 1, 195 0.72 0.40
TMT A (total no. of errors) 0.24 0.77 0.26 0.53 1, 195 0.51 0.48
TMT B (total no. of errors) 0.61 1.13 0.60 0.94 1, 195 2.08 0.15
Difference TMT B − TMT A (errors) 0.37 1.06 0.34 1.02 1, 195 0.89 0.35

Digit Symbol Test (DST) N = 132 N = 68

Digit Symbol Test (no. of digits, raw score correct) 61.1 19.6 60.3 16.5 1, 194 1.37 0.24
Digit Symbol Test (no. of digits, raw score skipped) 0.98 9.0 1.47 11.9 1, 194 0.05 0.82
Digit Symbol Test (no. of digits, raw score wrong) 0.77 3.8 0.31 1.1 1, 194 0.96 0.33

Purdue Pegboard Test N = 131 N = 71

Purdue Pegboard (no. of pegs placed corr., dominant hand) 13.2 2.6 13.4 2.4 1, 196 3.31 0.070
Purdue Pegboard (no. of pegs placed corr., non-dominant hand) 12.2 2.8 12.5 2.5 1, 196 1.83 0.18
Purdue Pegboard (no. pegs placed corr., both hands simult.) 10.5 2.8 10.9 2.2 1, 196 0.47 0.49

Rey Auditory Verbal Learning Test (RAVLT) N = 120 N = 68

RAVLT, sum of trials 1–5 (recall, no. of correct responses) 48.3 12.0 46.8 10.1 1, 182 2.31 0.13
RAVLT, trial 1 (recall, no. of correct responses) 7.0 2.3 6.6 1.9 1, 182 0.52 0.47
RAVLT, trial 7 (delayed recall, no. of correct responses) 10.2 3.4 9.4 3.0 1, 182 0.08 0.78
RAVLT, difference of trials 5–7 (forgetting) 0.6 2.6 1.1 2.6 1, 182 1.82 0.18

Legend: FE-SZ = first episode schizophrenia patients, SUD = substance use disorder, ANCOVA = analysis of covariance with independent factors SUD and gender and covariates
age and duration of education, m = mean, sd = standard deviation, F = F statistic, df = degrees of freedom, p = error probability of falsely rejecting the null hypothesis, that
there are no mean differences between study groups. Data are presented as mean ± standard deviation.

were to investigate the impact of concomitant substance abuse on psy- Therefore, the results of the above mentioned meta-analyses could
chopathology and neuropsychological functions in schizophrenia pa- not be supported by our analysis of the EUFEST sample. A recent
tients in the early stages of the disorder. Studies to date addressing study in 104 patients with a first-episode of non-affective psychosis
this question and including only small patient samples have revealed revealed a better attention and executive functions in the 47
inconsistent results. In our EUFEST sample we found nearly similar cog- patients using cannabis compared to non-users at baseline and one
nitive performance and psychopathology in patients with and without year follow-up, but the improvement in cognition during the
comorbid SUD. Patients with SUD only showed slightly reduced verbal follow-up was similar in the both groups (Rodriguez-Sanchez et al.,
memory performance at baseline and follow-up as well as slightly 2010). In addition, the amount of cannabis and length of time of con-
increased positive symptoms at baseline. Most previous studies did sumption did not relate to cognitive performance. In another study
not find pronounced cognitive impairments in individuals suffering with 99 first-episode patients (34 never using cannabis and 65 with can-
from schizophrenia who use substances, particularly focusing on al- nabis use, but no other illegal substances) cannabis users showed better
cohol and cannabis abuse (Cleghorn et al., 1991; Nixon et al., 1996). cognitive performance in all assessed tasks except working memory ma-
Some studies found even better neuropsychological performance of nipulation and cognitive measures remained stable during the fifteen
patients with lifetime history of cannabis abuse, e.g. in verbal and months follow-up (Leeson et al., 2012). Nevertheless, when premorbid
spatial recognition or recall (Liraud and Verdoux, 2002). A meta- IQ was entered as a covariate, all between group differences in cognition
analysis comprising 572 subjects out of 10 studies on cognitive function became nonsignificant except planning. Unfortunately we had no
in cannabis abusing schizophrenia patients found, contrary to the initial premorbid IQ to use as covariate in the EUFEST trial for our
hypothesis, superior cognitive functioning in comorbid patients com- post-hoc analysis. Interestingly in the above mentioned study can-
pared to those who never abused cannabis prior to examination (Yucel nabis use declined during the fifteen months with only 31% of pa-
et al., 2012). In their own study they examined 85 first-episode patients. tients with persisting use. This is similar to our cessation rate in
When comparing the patients without (n = 26) and with a history of the EUFEST analysis.
cannabis abuse (n = 59), the comorbid patients performed better on Heterogeneous samples, different cognitive assessments, and the in-
tests of visual memory, working memory, and executive functioning fluence of the variety of antipsychotic medication applied in the includ-
but not verbal memory (Yucel et al., 2012). A further more recent ed studies may contribute to the different findings of the meta-analyses
meta-analysis investigating the effect of cannabis on neurocognition and other first-episode studies when compared to our results. Regard-
in schizophrenia includes 8 studies with a total sample of 942 ing the influence of antipsychotics it is of note that in the EUFEST
patients (Rabin et al., 2011) and confirms the results of the former study most patients were drug-naïve at baseline and there was no
meta-analysis with superior performance of the cannabis abusing difference in cognitive performance between groups receiving the five
group. However, the meta-analyses included studies with chronic antipsychotics used in the study (Galderisi et al., 2009). Examining
schizophrenia as well as first-episode patients. the relative effect of cognitive impairment and substance abuse on
Our analysis of the largest sample of first-episode patients to date rehospitalization, a study found that substance abuse predicted
comprising 323 subjects revealed only slight differences not pointing prior hospitalizations. Furthermore, cognitive impairment revealed
to increased cognitive impairment in patients with persisting or by the Wisconsin Card Sorting Test (WCST) was a significant predictor
previous substance abuse. No significant differences in the exam- of duration of hospitalization in this study. However, in this trial
ined neurocognitive domains visual motor skill, psychomotor speed, substance abuse was not significantly associated with cognitive
attention/concentration, short-term visual memory, verbal memory measures (Jackson et al., 2001).
and cognitive flexibility between patients with and without previ- Cognitive performance may be linked to age, amount of consump-
ous substance abuse remained after correction for multiple testing. tion and further comorbidities. One study demonstrated pronounced
138 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139

impairments in memory function in older schizophrenia patients Contributors

T.W. wrote the manuscript. WW.F., RS.K., and EM.D. were strongly involved in the
with comorbid alcohol abuse and some studies indicate that cogni-
design of the study. All the authors, except A.H. and TS.-A. performed the measure-
tive impairments in drug abusers occur after prolonged and exces- ments and the ratings, and generated the source data. T.S.-A. undertook the statistical
sive use (Eckardt et al., 1995). In a study, including 203 schizophrenia analysis. A.H., and EM.D. contributed to the literature searches, and all the authors con-
patients, for patients with SUD history age and number of episodes tributed to the interpretation of the data. All authors contributed to and have approved
were predictive variables for the executive cognitive score, for pa- the final manuscript.

tients without SUD age and PANSS negative score predicted execu-
tive performance (Rodriguez-Jimenez et al., 2010). Conflict of interest
T. Wobrock was honorary speaker for Alpine Biomed, AstraZeneca, Bristol Myers
An explanation of our findings could be that substance abuse may
Squibb, Eli Lilly, I3G, Janssen Cilag, Novartis, Lundbeck, Sanofi-Aventis and Pfizer, and
contribute to an earlier onset of schizophrenia and induce psychotic has accepted travel or hospitality not related to a speaking engagement from AstraZeneca,
disorder before neurocognitive impairments become more promi- Bristol-Myers-Squibb, Eli Lilly, Janssen Cilag, and Sanofi-Synthelabo; and has received a
nent assuming an interaction between neurodevelopmental and neu- research grant from AstraZeneca, I3G and AOK (health insurance company).
rodegenerative processes in the etiology of schizophrenia. Patients P. Falkai was honorary speaker for Janssen-Cilag, Astra-Zeneca, Lilly, BMS, Lundbeck,
Pfizer, Bayer Vital, SKB, Wyeth, and Essex and during the last two years, but not presently,
without SUD in our sample may have higher neurodevelopmental vul-
he was member of the advisory boards of Janssen-Cillag, Astra-Zeneca, Lilly and Lundbeck.
nerability, and patients with SUD were prone to psychosis because of A. Hasan was honorary speaker for Desitin and Eli Lilly and was investigated to sci-
the effects of the abused substances on the dopaminergic system. entific conferences by Janssen Cilag, Lundberck and Pfizer.
Our result of significant enhanced positive symptoms in substance R.S. Kahn has received grants, honoraria for education programs, or served as consul-
tant for Astellas, AstraZeneca, BMS, Eli Lilly, Janssen-Cilag, Pfizer, Roche, and Sanofi-Aventis.
abusing patients at baseline is in accordance with most previous studies
W.W. Fleischhacker has received research grants from BMS/Otsuka, Eli Lilly,
(Caspari, 1999; Soyka et al., 2001; Buhler et al., 2002; Yucel et al., 2012). Janssen-Cilag, and Servier; honoraria for educational programs from AstraZeneca
Nevertheless, overall differences in psychopathology between sub- and Pfizer; speaking fees from AstraZeneca, BMS/Otsuka, Janssen-Cilag, and Pfizer;
stance using and not using schizophrenia patients were rather small and advisory board honoraria from AstraZeneca, BMS/Otsuka, Janssen-Cilag, Servier,
in some studies (Soyka et al., 2001). and Wyeth.
M. Davidson has received research grant support, travel support, speaker fees, or
The described sociodemographic characteristics in the substance
consultancy fees from JNJ, Pfizer, Lundbeck, Teva, BioLineRx, Eli Lilly, Sanofi-Aventis,
abuser group of the EUFEST sample are consistent with the literature, re- Roche, Servier, and Tangent Data.
vealing younger age, minor level of education, and higher rates of unem- J.K. Rybakowski has acted as a consultant or as a speaker for Adamed-Poland,
ployment and marriage (Mueser et al., 1990; Dixon et al., 1991; Linszen Bristol-Myers Squibb, Eli Lilly, Janssen-Cilag, Lundbeck, Organon, Pfizer, Sanofi-Aventis,
and Servier.
et al., 1994; Wobrock et al., 2007). In our sample some differences were
J. Libiger has received speaker's honoraria, travel grants, or consultancy fees from
not as striking as in previous studies, for instance one study found that AstraZeneca, Eli Lilly, Bristol-Myers Squibb, Lundbeck, and Servier. J. Libiger was a
cannabis abusing first-episode patients were about 7 years younger member of the advisory boards of AstraZeneca, Eli Lilly and Bristol-Myers Squibb and
than nonabusers (Veen et al., 2004) compared to a mean difference of he is a faculty member of the Lundbeck Institute (Lundbeck Neuroscience Foundation).
two years in our trial. The overall prevalence of substance abuse (27.2% J.J. López-Ibor is member of the scientific committee of Fundacion Lilly (Spain),
Wyeth member of the global strategy consultant board, and board member of the
with SUD) in our study sample is also in line with the reported rates in
Lundbeck Neuroscience Foundation (LINF). J.J. López-Ibor has attended several meet-
first-episode schizophrenia spectrum disorder (Linszen et al., 1994; ings financed by Eli Lilly, Bristol-Myers Squibb, GlaxoSmithKline, and Lundbeck.
Hambrecht and Hafner, 1996; Rabinowitz et al., 1998; Cantwell et al., L.G. Hranov has received speaker's honoraria, sponsorship for congresses, or
1999; Wobrock et al., 2007). investigator's fees from Bristol-Myers Squibb, Corcept Therapeutics, GlaxoSmithKline,
The pattern of abused substances reported in EUFEST is similar to Eli Lilly, Sanofi-Aventis, and Lundbeck.
W. Gaebel is or has been a member of the speaker bureau for AstraZeneca GmbH,
other European studies (Duke et al., 2001; Veen et al., 2004), pre- Bristol-Myers Squibb GmbH and Co KG, GlaxoSmithKline, Janssen-Cilag GmbH, Lilly
senting high prevalence of cannabis and low prevalence of other Deutschland GmbH, Lundbeck GmbH, and Sanofi-Synthelabo GmbH/Aventis. W. Gaebel
substances in contrast to US surveys showing high rates of cocaine is a member of the advisory board for Lilly Deutschland GmbH, Lundbeck GmbH, Novartis
consumers (Graham et al., 2001). That cognitive performance of per- Pharma GmbH, and Wyeth Pharma GmbH. W. Gaebel has received or is currently receiv-
ing research grants from Bristol-Myers Squibb GmbH and Co KG, Lilly Deutschland GmbH,
sistent users was similar to those who stopped substance use in our
Wyeth Pharma GmbH, Janssen-Cilag GmbH, and Lundbeck GmbH.
study was somehow unexpected, however, one study found that the J. Peuskens has received consultancy fees and research grants from and participat-
cognitive performance was relatively stable between a group of re- ed in clinical trials sponsored by AstraZeneca, Sanofi-Aventis, Eli Lilly, Pfizer, and
mittent comorbid cocaine abusers and current cocaine consumers Janssen-Cilag.
with schizophrenia (Peer et al., 2009). T. Schneider-Axmann, S. Galderisi, S. Dollfus, E.M. Derks and H. Boter declare that
they have no conflict of interest.
There are several limitations of our study. First, in the main analy-
sis we did not match the group of abusers with nonabusers for age,
Acknowledgments
gender or education. However, after performing the matching with
None.
88 patients in each group we found similar results compared to the
main analysis. Secondly we had no comparison with a healthy control
group. However, in an add-on study it was already demonstrated that References
the total EUFEST sample showed a reduced cognitive performance in Addington, D., Addington, J., Maticka-Tyndale, E., 1993. Assessing depression in
all tested measures compared to healthy controls. Third, we had no schizophrenia: the Calgary Depression Scale. Br. J. Psychiatry Suppl. (22), 39–44.
premorbid IQ to adjust our results. At least, the follow-up period of Beatty, W.W., Katzung, V.M., Moreland, V.J., Nixon, S.J., 1995. Neuropsychological per-
formance of recently abstinent alcoholics and cocaine abusers. Drug Alcohol De-
6 months may have been too short to detect differences between pend. 37 (3), 247–253.
the groups during the course of the disease. Buckley, P.F., Miller, B.J., Lehrer, D.S., Castle, D.J., 2009. Psychiatric comorbidities and
In conclusion, despite all limitations, in our study of first-episode schizophrenia. Schizophr. Bull. 35 (2), 383–402.
Buhler, B., Hambrecht, M., Loffler, W., an der Heiden, W., Hafner, H., 2002. Precipitation
patients substance abuse seems not to effect cognitive functioning
and determination of the onset and course of schizophrenia by substance abuse —
in schizophrenia. a retrospective and prospective study of 232 population-based first illness epi-
sodes. Schizophr. Res. 54 (3), 243–251.
Cantwell, R., Brewin, J., Glazebrook, C., Dalkin, T., Fox, R., Medley, I., Harrison, G., 1999.
Prevalence of substance misuse in first-episode psychosis. Br. J. Psychiatry 174,
Role of funding source 150–153.
This study was supported by the European Group for Research in Schizophrenia (EGRIS; Caspari, D., 1999. Cannabis and schizophrenia: results of a follow-up study. Eur. Arch.
Utrecht, the Netherlands) with grants from 3 pharmaceutical companies: AstraZeneca Psychiatry Clin. Neurosci. 249 (1), 45–49.
(Södertälje, Sweden), Pfizer (New York, NY), and Sanofi-Aventis (Paris, France). Cleghorn, J.M., Kaplan, R.D., Szechtman, B., Szechtman, H., Brown, G.M., Franco, S., 1991.
The EUFEST has been registered by Current Controlled Trials Ltd and assigned the Substance abuse and schizophrenia: effect on symptoms but not on neurocognitive
number ISRCTN68736636 (http://www.controlled-trials.com/ ISRCTN68736636). function. J. Clin. Psychiatry 52 (1), 26–30.
 T. Wobrock et al. / Schizophrenia Research 147 (2013) 132–139 139

Costa, L.D., Vaughan Jr., H.G., Levita, E., Farber, N., 1963. Purdue Pegboard as a predictor Meijer, J.H., Dekker, N., Koeter, M.W., Quee, P.J., van Beveren, N.J., Meijer, C.J., 2012. Can-
of the presence and laterality of cerebral lesions. J. Consult. Psychol. 27, 133–137. nabis and cognitive performance in psychosis: a cross-sectional study in patients
Coulston, C.M., Perdices, M., Tennant, C.C., 2007. The neuropsychology of cannabis and with non-affective psychotic illness and their unaffected siblings. Psychol. Med. 42
other substance use in schizophrenia: review of the literature and critical evalua- (4), 705–716.
tion of methodological issues. Aust. N. Z. J. Psychiatry 41 (11), 869–884. Mueser, K.T., Yarnold, P.R., Levinson, D.F., Singh, H., Bellack, A.S., Kee, K., Morrison, R.L.,
Dixon, L., Haas, G., Weiden, P.J., Sweeney, J., Frances, A.J., 1991. Drug abuse in schizo- Yadalam, K.G., 1990. Prevalence of substance abuse in schizophrenia: demographic
phrenic patients: clinical correlates and reasons for use. Am. J. Psychiatry 148 and clinical correlates. Schizophr. Bull. 16 (1), 31–56.
(2), 224–230. Nixon, S.J., Hallford, H.G., Tivis, R.D., 1996. Neurocognitive function in alcoholic, schizo-
Duke, P.J., Pantelis, C., McPhillips, M.A., Barnes, T.R., 2001. Comorbid non-alcohol sub- phrenic, and dually diagnosed patients. Psychiatry Res. 64 (1), 35–45.
stance misuse among people with schizophrenia: epidemiological study in central Peer, J., Bennett, M.E., Bellack, A.S., 2009. Neurocognitive characteristics of individuals
London. Br. J. Psychiatry 179, 509–513. with schizophrenia and cocaine dependence: comparison of currently dependent
Eckardt, M.J., Stapleton, J.M., Rawlings, R.R., Davis, E.Z., Grodin, D.M., 1995. Neuropsy- and remitted groups. J. Nerv. Ment. Dis. 197 (8), 631–634.
chological functioning in detoxified alcoholics between 18 and 35 years of age. Pope Jr., H.G., Gruber, A.J., Yurgelun-Todd, D., 1995. The residual neuropsychological
Am. J. Psychiatry 152 (1), 53–59. effects of cannabis: the current status of research. Drug Alcohol Depend. 38 (1),
Galderisi, S., Davidson, M., Kahn, R.S., Mucci, A., Boter, H., Gheorghe, M.D., Rybakowski, 25–34.
J.K., Libiger, J., Dollfus, S., Lopez-Ibor, J.J., Peuskens, J., Hranov, L.G., Fleischhacker, Rabin, R.A., Zakzanis, K.K., George, T.P., 2011. The effects of cannabis use on
W.W., 2009. Correlates of cognitive impairment in first episode schizophrenia: neurocognition in schizophrenia: a meta-analysis. Schizophr. Res. 128 (1–3),
the EUFEST study. Schizophr. Res. 115 (2–3), 104–114. 111–116.
Gerlach, J., Korsgaard, S., Clemmesen, P., Lauersen, A.M., Magelund, G., Noring, U., Rabinowitz, J., Bromet, E.J., Lavelle, J., Carlson, G., Kovasznay, B., Schwartz, J.E., 1998.
Povlsen, U.J., Bech, P., Casey, D.E., 1993. The St. Hans Rating Scale for extrapyrami- Prevalence and severity of substance use disorders and onset of psychosis in
dal syndromes: reliability and validity. Acta Psychiatr. Scand. 87 (4), 244–252. first-admission psychotic patients. Psychol. Med. 28 (6), 1411–1419.
Graham, H.L., Maslin, J., Copello, A., Birchwood, M., Mueser, K., McGovern, D., Georgiou, G., Reitan, R.M., Wolfson, D., 1993. The Halstead – Reitan Neuropsychological Test Battery:
2001. Drug and alcohol problems amongst individuals with severe mental health Theory and Clinical Interpretation, 2nd ed. Neuropsychology Press, Tucson.
problems in an inner city area of the UK. Soc. Psychiatry Psychiatr. Epidemiol. 36 Rey, A., 1964. L'Examen clinique en psychologie. Presses Universitaires de France, Paris.
(9), 448–455. Rodriguez-Jimenez, R., Bagney, A., Martinez-Gras, I., Ponce, G., Sanchez-Morla, E.M.,
Green, M.F., 1996. What are the functional consequences of neurocognitive deficits in Aragues, M., Rubio, G., Jimenez-Arriero, M.A., Santos, J.L., Palomo, T., 2010. Execu-
schizophrenia? Am. J. Psychiatry 153 (3), 321–330. tive function in schizophrenia: influence of substance use disorder history.
Guy W, B.R., 1976. CGI: Clinical Global Impressions. National Institute of Mental Health. Schizophr. Res. 118 (1–3), 34–40.
Hambrecht, M., Hafner, H., 1996. Substance abuse and the onset of schizophrenia. Biol. Rodriguez-Sanchez, J.M., Ayesa-Arriola, R., Mata, I., Moreno-Calle, T., Perez-Iglesias, R.,
Psychiatry 40 (11), 1155–1163. Gonzalez-Blanch, C., Perianez, J.A., Vazquez-Barquero, J.L., Crespo-Facorro, B., 2010.
Heinrichs, R.W., Zakzanis, K.K., 1998. Neurocognitive deficit in schizophrenia: a quan- Cannabis use and cognitive functioning in first-episode schizophrenia patients.
titative review of the evidence. Neuropsychology 12 (3), 426–445. Schizophr. Res. 124 (1–3), 142–151.
Jackson, C.T., Fein, D., Essock, S.M., Mueser, K.T., 2001. The effects of cognitive impair- Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Hergueta,
ment and substance abuse on psychiatric hospitalizations. Community Ment. T., Baker, R., Dunbar, G.C., 1998. The Mini-International Neuropsychiatric Interview
Health J. 37 (4), 303–312. (M.I.N.I.): the development and validation of a structured diagnostic psychiatric
Jensen, J., Nilsson, L.L., Levander, S., 2004. Neurocognitive and psychopathological corre- interview for DSM-IV and ICD-10. J. Clin. Psychiatry 59 (Suppl. 20), 22–33 (quiz
lates of self-monitoring ability in schizophrenia. Eur. Arch. Psychiatry Clin. Neurosci. 34-57).
254 (5), 312–317. Soyka, M., Albus, M., Immler, B., Kathmann, N., Hippius, H., 2001. Psychopathology in
Kahn, R.S., Fleischhacker, W.W., Boter, H., Davidson, M., Vergouwe, Y., Keet, I.P., dual diagnosis and non-addicted schizophrenics — are there differences? Eur.
Gheorghe, M.D., Rybakowski, J.K., Galderisi, S., Libiger, J., Hummer, M., Dollfus, Arch. Psychiatry Clin. Neurosci. 251 (5), 232–238.
S., Lopez-Ibor, J.J., Hranov, L.G., Gaebel, W., Peuskens, J., Lindefors, N., Riecher- Torrey, E.F., 2002. Studies of individuals with schizophrenia never treated with anti-
Rossler, A., Grobbee, D.E., 2008. Effectiveness of antipsychotic drugs in first-episode psychotic medications: a review. Schizophr. Res. 58 (2–3), 101–115.
schizophrenia and schizophreniform disorder: an open randomised clinical trial. Lancet Veen, N.D., Selten, J.P., van der Tweel, I., Feller, W.G., Hoek, H.W., Kahn, R.S., 2004. Can-
371 (9618), 1085–1097. nabis use and age at onset of schizophrenia. Am. J. Psychiatry 161 (3), 501–506.
Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale Wechsler, D., 1997. Wechsler Adult Intelligence Scale, 3rd ed. Psychological Corporation,
(PANSS) for schizophrenia. Schizophr. Bull. 13 (2), 261–276. San Antonio.
Kovasznay, B., Fleischer, J., Tanenberg-Karant, M., Jandorf, L., Miller, A.D., Bromet, E., Wobrock, T., Sittinger, H., Behrendt, B., D'Amelio, R., Falkai, P., Caspari, D., 2007. Comor-
1997. Substance use disorder and the early course of illness in schizophrenia and bid substance abuse and neurocognitive function in recent-onset schizophrenia.
affective psychosis. Schizophr. Bull. 23 (2), 195–201. Eur. Arch. Psychiatry Clin. Neurosci. 257 (4), 203–210.
Leeson, V.C., Harrison, I., Ron, M.A., Barnes, T.R., Joyce, E.M., 2012. The effect of cannabis Woods, S.W., 2003. Chlorpromazine equivalent doses for the newer atypical antipsy-
use and cognitive reserve on age at onset and psychosis outcomes in first-episode chotics. J. Clin. Psychiatry 64 (6), 663–667.
schizophrenia. Schizophr. Bull. 38 (4), 873–880. Yucel, M., Bora, E., Lubman, D.I., Solowij, N., Brewer, W.J., Cotton, S.M., Conus, P., Takagi,
Linszen, D.H., Dingemans, P.M., Lenior, M.E., 1994. Cannabis abuse and the course of M.J., Fornito, A., Wood, S.J., McGorry, P.D., Pantelis, C., 2012. The impact of cannabis
recent-onset schizophrenic disorders. Arch. Gen. Psychiatry 51 (4), 273–279. use on cognitive functioning in patients with schizophrenia: a meta-analysis of
Liraud, F., Verdoux, H., 2002. Effect of comorbid substance use on neuropsychological existing findings and new data in a first-episode sample. Schizophr. Bull. 38 (2),
performance in subjects with psychotic or mood disorders. Encephale 28 (2), 316–330.
160–168.