Download Axis I comorbidity in bipolar disorder with psychotic features.

Axis I comorbidity in bipolar disorder with psychotic features.
S Pini, L Dell'Osso, C Mastrocinque, G Marcacci, A Papasogli, S Vignoli, S Pallanti and G Cassano
BJP 1999, 175:467-471.
Access the most recent version at DOI: 10.1192/bjp.175.5.467
References
Reprints/
permissions
You can respond
to this article at
Downloaded
from
This article cites 0 articles, 0 of which you can access for free at:
http://bjp.rcpsych.org/content/175/5/467#BIBL
To obtain reprints or permission to reproduce material from this paper, please write
to [email protected]
http://bjp.rcpsych.org/cgi/eletter-submit/175/5/467
http://bjp.rcpsych.org/ on January 9, 2012
Published by The Royal College of Psychiatrists
To subscribe to The British Journal of Psychiatry go to:
http://bjp.rcpsych.org/site/subscriptions/
Axis I comorbidity in bipolar disorder
with psychotic features
STEFANO PINI, LlLlANA DELL'OSSO. CONCETTINA MASTROCINQUE,
GlOVANNl MARCACCI, ALESSANDRA PAPASOGLI, SERENA VIGNOLI.
STEFANO PALLANTI and GlOVANNl CASSANO
Background Axis l comorbidities
are prevalent among patients with
severe bipolar disorder but the clinical
and psychopathologicalimplications
are not clear.
Aims To investigate characteristics
offour groups of patients categorisedas
follows: substance abuse only (group I),
substance abuse associated with other
Axis I disorders (group 2), non-substanceabuse Axis I comorbidity (group 3), no
psychiatric comorbidity (group 4).
Method Consecutive patients
with bipolar disorder with psychotic
features (n=125) were assessed using
the Structured Clinical Interview for
DSM- Ill- R - patient version, and
several psychopathological scales.
Results By comparisonwith
group 4, group I had a higher riskof having
mood-incongruent delusions, group 2 had
an earlier age at onset of mood disorder,
a more frequent onset with a mixed state
and a higher riskof suicide, and group 3
had more severe anxiety and a better
awareness of illness.
Conclusions Substance abuse, nonsubstance-abuseAxis I comorbidity and
their reciprocal association are associated
with different characteristicsof bipolar
disorder.
Declaration of interest This
study was supported by funds from
the Departmentof Psychiatry,
University of Pisa.
The prevalence of psychiatric comorbidity
in bipolar disorder with psychotic features
ranges from 13% to 73.4%, with substanceuse disorders being the most common condition, followed by anxiety and eating disorders (Black et al, 1988; Strakowsky et al,
1992; Kessler et al, 1997). Such comorbidities are associated with more severe psychotic features, longer stays in hospital,
low recovery rates and earlier age at onset
of mood disorder (Some et al, 1994; Brady
& Some, 1995; Cassano et al, 1998; Scott
et al, 1998; Strakowsky et al, 1998). However, the extent to which such clinical correlates were due to substance abuse andlor
non-abuse Axis I comorbidity was not
exhaustively clarified.
In this study we investigated the clinical
characteristics of psychotic bipolar disorder
in patients categorised into four groups
according to their patterns of comorbidity:
substance abuse only (group I), substance
abuse associated with other Axis I disorders (group 2), non-substance-abuse Axis
I comorbidity (group 3), no psychiatric
comorbidity (group 4).
METHOD
The method of this study and the characteristics of the Pisa Centre have been described
in detail elsewhere (Cassano et al, 1998).
Consecutively hospitalised patients with
psychotic bipolar disorder (n=125) were
recruited and included in this study on the
basis of the following criteria: age over 16
years; presentation with psychotic symptoms (i.e. formal thought disorders, delusions, hallucinations, grossly disorganised
behaviour); provision of informed written
consent and approval from the local ethical
committee. Patients were selected independently of previous stays in hospital and/or
prior antipsychotic or mood-stabiser treatments and independently of having had
single or multiple episodes of psychosis.
Patients were excluded from the study if
psychotic symptoms either were secondary
to acute intoxication or withdrawal from
alcohol or other substances or were presenting with concomitant severe medical conditions defined according to Black et a1 (1998)
as any serious or acute life-threatening illness such as cancer, myocardial infarction,
stroke, or hepatic insufficiency. The inclusion diagnosis of psychosis was made by
three senior psychiatrists who were not
directly involved in the study, on the basis
of their clinical judgement and reading of
patient records. Then, the Structured Clinical
Interview for DSM-111-R - patient version
(SCID-P; Spitzer et al, 1987) was administered in the week preceding the patient's
discharge by three residents in psychiatry
who had been trained in the use of the
SCID-P. As recommended by Spitzer et a1
(1987), SCID-P interviewers were skilled
clinical researchers with at least three
years of clinical experience and with substantial familiarity with DSM-III-R criteria
(American Psychiatric Association, 1987).
In previous studies, psychiatric comorbidity
has been defined as the presence of an
antecedent or concurrent DSM-111-R Axis
I diagnosis in addition to the principal diagnosis. Ten Axis I diagnoses were assessed in
the present study (panic disorder, social
phobia, simple phobia, obsessive-compulsive
disorder, somatoform disorder, undifferentiated somatoform disorder, chronic pain
disorder, hypochondria, anorexia and bulimia). Substance abuse (involvingstimulants,
sedatives, opiates, hallucinogens, cocaine,
cannabis, alcohol and multiple drugs) was
also assessed by the SCID-P. In completing
the SCID-P, information was obtained from
any source available in addition to the
patient interview, including medical records,
first-degree relatives and treating clinicians.
Age at onset was investigated by the
SCID-P. Comorbid diagnoses were defined
as antecedent if patients endorsed full syndrome criteria more than one year before
the onset of bipolar disorder.
Psychopathology was assessed using
the 18-item version of the Brief Psychiatric
Rating Scale (BPRS; Overall & Gorharn,
1962). Negative symptoms were assessed
using the Scale for the Assessment of Negative Symptoms (SANS; Andreasen, 1983).
The awareness of illness was evaluated by
means of the Scale for the Unawareness of
Mental Disorder (SUMD; Amador et al,
1993). In this scale, scores range from 1
to 5, with higher scores indicating poorer
awareness of illness. Interrater reliability
for the SCID-P, BPRS, SANS and SUMD
was assessed in a small sample study (n=8)
on the basis of joint records for both principal and comorbid diagnoses. Good reliability established from joint ratings was
obtained for both principal and comorbid
diagnoses (~=0.87 and ~=0.82, respectively).
Socio-demographic characteristics
As shown in Table 2, group 2 was significantly younger than group 4 (PcO.05).
Males were significantly more represented
among substance users, with or without
other Axis I comorbidity, than those without
any comorbidity (P c 0.05). Married patients
were significantly more frequent in group 4
than in the other three groups (P< 0.05).
Statistical analyses
Analysis of variance (ANOVA) was performed on key demographic variables in
order to determine whether there were any
significant difkrences across groups. The
ANOVA was then performed on psychopathological variables by group. Pairwise
comparisons were performed on the psychopathological and clinical variables to
determine which groups sigrhcantly
differed from each other. All results at
Pd0.05 were judged to be sigmficant.
The overall error rate was also controlled
for by making pairwise comparisons using
the Tukey test of significant differences.
Categorical variables were analysed by the
2 test. The age at onset of bipolar disorder
in the four groups was assessed by means
of survival analyses with the Wicoxon
(Gehan) test (Gehan, 1997) for overall
and painvise comparisons. Multiple logistic
regression analysis was performed in order
to estimate the strength of association between clinical variables and comorbidity
after controlling for the principal psychotic
diagnosis and socio-demographic variables.
All analyses were performed using SPSS
version 7.0 for Widows 95.
Age at onset of bipolar disorder,
substance abuse and Axis I
comorbidities
Mean age at onset of bipolar disorder was
22.8 Years
5.7) in group 1, 21.2
(s.d.=4.6) in group 2, 25.3 (s.d.=8.8) in
group 3 and 26.1 (s.d.=7.7) in group 4.
As shown in Fig. 1, pairwise comparisons
from survival analysis showed that the
mean age at onset of bipolar disorder was
sigmficantly lower in group 2 than in group
4 (F=7.140, d.f.=l, P<0.01). Mean age at
onset of substance abuse was 19.6 years
(s.d.=S.S), with stimulant abuse being
earliest (15.8 years, s.d.=2.9) and alcohol
abuse the latest (22.9 years, s.d.=8.6), while
mean age at onset of non-abuse Axis I
comorbidity was 22.1 years (s.d.=9.5), with
social phobia being earliest (8.5, s.d.=2.1)
and generalised anxiety disorder the latest
(31.4, s.d.=18.9) (see Table 1).The onset
of substance abuse preceded that of mood
disorder in 23 (74.9%) abusers. The onset
of non-abuse Axis I comorbidity preceded
the onset of psychotic disorder in 20
(44.4%) patients with Axis I comorbidity
without substance abuse.
Age at onset of mood disorder was significantly earlier in subjects with cannabis
abuse than in those without (20.4, s.d.=
4.7 v. 25.6, s.d.=7.9; t=3.037, d.f.=123,
P<0.01), but not in those with alcohol
abuse (22.4, s.d.=5.2 v. 25.1, s.d.=8.0;
t=1.408, d.f.=149, P<0.162). We also
found a significant
negative
correlation
between age at 0of bipolar disorder
andcannabisabuse (Kenddl'srbtest=0.247,
Pc0.01) and panic disorder (Kendall's rb
test=-0.187, PcO.OS), but not with any
of the other &s I diagnoses.
Psychopathology
As shown in Table 3, levels of patients'
awareness of illness, of effects of medications
and of social consequences of mental disorder showed differences between the four
groups. Overall, group 3 had a better insight
than both group 4 and group 1.
There were no significant differences
between the four groups on BPRS scores
with the exception of the anxietyldepression
T.ble I Frequency &Axis I comorbidity and age of onset in 125 consecutively hospitalid patients with
bipdv disorder with psychic features
DSM-III-R Axis I diagnosis
Patients with bipolar disorder with psychotic features
(n=lX)
n (%)
Mean age at onset (years)(s.d.)
3 (2.4)
17 (13.6)
4 (3.2)
4 (3.2)
7 (5.6)
3 (24)
5 (4.0)
4 (3.2)
23 (18.4)
5 (4.0)
4 (3.2)
16 (12.8)
3 (24)
19 (15.2)
125(100)
39 (31.2)
5 (4.0)
8.5 (2.1)
I I .9 (3.7)
16.0 (2.7)
16.0 (2.5)
18.4 (3.2)
16.3 (9.5)
18.1 (3.1)
18.3 (0.5)
18.6 (3.4)
20.3 (5.5)
18.8 (2.4)
23.4 (9.9)
23.0 (7.1)
23.0 (7. I)
24.7 (7.6)
25.2 (9.1)
31.4 (18.9)
RESULTS
The total cohort of 125 bipolar patients
was categorised into the four groups as
follows: group 1 (patients with substance
abuse disorder without any other axis I
comorbidity) consisted of nine patients
(7.2%), group 2 (substance abuse plus at
least one other Axis I comorbid disorder)
consisted of 22 patients (17.6%), group 3
(at least one non-abuse Axis I disorder)
consisted of 45 patients (36.0%), and
group 4 (neither substance abuse nor
other Axis I comorbidities)consisted of 49
patients (39.2%). Frequencies and mean
age of onset of comorbid diagnoses and of
bipolar disorder are shown in Table 1.
Overall, total rates of substance-abuse and
non-abuse Axis I comorbidity were 24.8%
(n=31) and 36.0% (n=45), respectively.
Simple phobia
Social phobia
Multiple drug abuse
Stimulants abuse
Cocaine use
Anorexia
Hallucinogens use
Opiate use
Cannabis use
Bulimia
Sedative abuse
Obreuive-compulsivedisorder
Somatofonn disorders
Alcohol abuse
Bipolar disorder
Panic disorder
Generalised anxiety disorder
AXIS I COMORBIDITY IN B I P O L A R
Tabla 2 Sdi-demographic a d clinical chuacterirtkr of 125
DISORDER
hospitalid patieno with bipdv disorder with psychotic features
-
-
--
-
-
- - --
-- -
-
Substance rburc/no Substance abuse plus
Axis l comorbidkyl N o Axis l or rubstance ANOVA (Tukey pairwise
Axis Icornorbidiiy
Axis Icomorbidky
no substance abuse
abuse comorbidky
(n=9)
(n=22)
(n=45)
(n=49)
34.4 (s.d. 14.0)
30.6 (s.d. 7.3)
35.4 (s.d. 1 1.7)
39.2(s.d. 12.3)
10.7 (s.d. 3.6)
10.6 (s.d. 3.8)
I1.6 (rd. 3.5)
9.9 (rd. 4.1)
NS
5 (55.6%)
I5 (68.2%)
20 (4.4%)
16 (32.7%)
xL8.21 I.P<0.05
0 (0%)
4 (4.4%)
2 (9.1 %)
12 (54.5%)
I0 (22.2%)
19 (38.8%)
~'=11.171. P<O.M
23 (51.1%)
14 (28.6%)
NS
DSM-III-R global functioning
33.7 (s.d. 15.8)
30.4 (s.d. 12.5)
28.9 (s.d. 1 1.4)
34.1 (s.d. 12.9)
NS
Length of last stay in hospiial (days)
47.1 (s.d. 25.8)
33.1 (s.d. 25.5)
29.2 (s.d. 19.0)
29.7 (rd. 23.0)
NS
2.8 (s.d. 1.6)
3.3 (s.d. 1.6)
2.4 (s.d. 1.5)
2.9 (s.d. 1.6)
NS
Characteristic
Mean age
Educational level (years of school)
Gendcc male (n(%))
Married
UflemploYed (n(%))
Total number d stays in hospital
subscale, which scored significantly higher
in group 3 than in group 4 (2.74,
s.d.=1.22 v. 2.02, s.d.=1.02; F=3.804,
P<0.02). Among the 18 BPRS items, only
'depressed mood' (F=1.795, d.f.=3,124,
P<O.OS) scored higher in group 3 than in
group 1. As to the SANS, only the sub-scale
anhedonialasociality was significantly more
severe in group 3 than in group 4 (10.11,
s.d.=5.99 v. 5.48, s.d.=7.59; F=4.178,
d.f.=3,124, P<0.01).
comparisons)
F=4.323, P<O.OI (2<4)
1.2 7
1.O
-
- 0.8
-
'5
- a
Q
.z
-
0.6
3
Results of logistic regression
analyses
The association of each of the four comorbidity groups with a series of clinical variables was analysed by composite logistic
regression models. Delusions and hallucinations, formal thought disturbances, affective
flattening, polarity of first affective episode,
suicide attempts, and good insight were used
as dependent variables. Group, gender, age
and principal diagnosis of bipolar disorder
were used as independent variables.
A. shown in
4' mood-incongruent delusions were significantly associated
with group 1, an onset with a mixed episode and a history of previous suicide
attempts with group 2. An onset with a
manic episode was negatively and good
awareness of illness positively associated
with group 3.
DISCUSSION
This study aimed to investigate clinical
characteristics associated with substance
abuse and non-abuse Axis I comorbidity in
a cohort of patients with bipolar disorder
with psychotic features. Overall, 60.8% of
our patients had at least one Axis I comorbid disorder. Our centre specialises in the
S
o
0.4
-
0.2
-
0.0
gmup 4 ( n = 49)
- - - group3(n=45)
-group 2 ( n = 22)
..... group 1 ( n = 9)
,
I
I
-10
0
10
I
I
20
30
i
40
50
60
Age at onset of bipolar disorder
Fig. I Comparison dwrvival experience using theWlc-n
2<group 4; Wikomn=7.I4OOd.f.=l. P<0.01).
treatment of mood and aruciety disorders;
addicted patients are usually referred to
other centres. This might have contributed
to an under-representation of substance
abuse; our figures probably do not reflect
a true prevalence among bipolar patients.
However, consistent with previous studies
(Strakowsky et al, 1992; Kessler et al,
1997), our results indicate that a substantial
propomon of subjects with psychotic bipolar
disorder were complicated by additional
Axis I conditions.
(Gehan) statistic (pairwise comparison: group
Substance abuse without other
Axis I comorbidity (group I)
In group 1, onset of substance abuse
preceded onset of affective illness in the
majority (64%)of cases. Logistic regression
analysis showed that this group had a
higher risk of having mood-incongruent delusions than the rest of the subjects. The
reason for this last association not being
found in group 2 is difficult to define. It is
possible that these abusers, who were also
W e 3
Results of analysis of variance' for the Sale for the Unawareness of Mental Disorder (awareness of illness): six general items scores in 125 consecutively
hosplralised patients with bipolar disorder with psychotic features
Substance abuselno
Substance abuse plus
Axis l comorbidity/
No Axis Ior substance
Axis Icomorbidity
Axis Icornorbidity
no substance abuse
abuse cornorbidity
ANOVA (Tukey
paimise
(n=9) (mean (s.d.))
(n=22) (mean (s.d.))
(n=45) (mean (s.d.))
(n=49) (mean (s.d.))
comparisons)
Current episode
2.67 (1.73)
2.95 (1.68)
2.80 (1.44)
3.38 (1.44)
NS
Past episcde
3.43 (1.51)
3.m (1.84)
3.13 (1.41)
3.82 (148)
F=2.360, P c 0.05 (3 <4)
Current episode
3.33 (1.41)
2.86 (1.55)
2.58 (1.47)
3.16(1.45)
NS
Past episode
3.43 (0.98)
3.39 (1.61)
2.78 (1.49)
3.69 (1.34)
F=2.791. P c 0.05 (3 c 4 )
Current episode
3.43 (1.81)
2.91 (1.69)
2.81 (1.56)
3.47 (1.47)
NS
-
3.40 (1.52)
3.39 (1.65)
3.05 (1.47)
3.92 (1.38)
F=2.226, P < 0.05 (3 <4)
Awareness of illness
Awarenest of treatment efficacy
Awarmes of social consequencesof
Past episode
I. Pon hocTukcyitests am significant a t the P ~ 0 . 0 5bvel.
1
.4 Results of logistic regressionanalyses for the association of groups ofcomorMdiy with fm
features
Dependent variable
Independent variables' (odds ratio (95%confidence interval))
Substance abuse/ Substance abuse Axis Icomorbidiiy/
no Axis I
plus Axis I
cornorbidity
comorbidity
-
2.41 (1.03-8.29)
-
-
0.34 (O.IC0.81)
No Axis l or
no substance abuse substance abuse
mmorbidity
delusions
Onset with a mixed episode
Onset with a manic episode
Suicide attempts
Goad insight
in group 1also suggests that non-abuse Axis
I comorbidity may play an important role in
anticipating the onset of psychosis.
6.75 (1.36-33.41)
-
-
2.98 (1.32-6.42)
-
I. Gender, age and dbgnorir of bipolar disorder were also induded in the m o d s as independentvariables.
found to have low levels of insight, represent a subgroup of subjects with a greater
likelihood of having a bipolar disorder with
schizophrenia-like features than other subgroups. However, larger samples are needed
to support such a hypothesis empirically.
Substance abuse plus other Axis I
comorbidity (group 2)
The co-occurrence of substance abuse and
other Axis I diagnoses was associated with
a younger patient, an earlier age at onset of
bipolar disorder, an onset of illness with a
mixed manic episode, and a greater risk of
suicide attempts. Although this group had
a lower risk of having mood-incongruent delusions than the group with substance abuse
only, it showed a greater morbidity in terms
of course of illness and life-threatening
risk. The role of cannabis use in anticipating
the onset of psychotic illness appeared to
be more pronounced than that of alcohol
abuse. This finding is consistent with a
recent study by Brook et a1 (1998), which
reports a sigtllficant relationship between
earlier adolescent drug use and later depressive and disruptive disorders in young
adulthood, controlling for earlier psychiatric
disorders. Strakowsky et a1 (1996) found
that patients with bipolar disorder and antecedent alcohol abuse had a later onset of
affective illness, arguing that perhaps this
represented a subgroup of patients in whom
previous alcohol abuse was only necessary
to precipitate an affective episode.
Furthermore, given the remarkably
earlier age at onset of cannabis abuse than
of bipolar illness or alcohol abuse, our data
give support to the hypotheses that soft-drug
abuse is likely to anticipate the onset of psychosis, while alcohol is more likely to complicate it. However, the fact that the onset
of mood disorder was earlier in group 2 than
Non-abuse Axis I comorbidity
(group 3)
By comparison with the group without
comorbidity, the group with non-abuse
Axis I comorbidity had higher levels of
anxiety and depression, more anhedonid
asociality and a greater likelihood of having
an onset of illness with a mixed or depressive affective episode. Overall, the onset
of Axis I comorbidity preceded that of
bipolar disorder in 43% of cases.
This group appeared to have fewer psychotic symptoms and a lower risk of serious
suicide attempts than the other groups.
These features, and the more negative selfexperience of psychopathology - reflected
by a better insight of illness and a higher
level of anxiety - that characterisesthese patients, suggest that the boundaries between
psychosis and neurosis are likely to be less
marked in this cluster than in the others.
These findings seem to support the hypothesis of Black et a1 (1988) that manic patients
with non-abuse comorbidity resemble in
many respects the 'neurotic' secondary depression described by Wiokur et al (1988).
Limitations
It is important to acknowledge several
limitations of this study. We investigated
only hospitalised patients, limiting the
applicability of these results in relation to
patients with bipolar disorder in general.
Rates of lifetime substance abuse and Axis
I comorbidity, as well as of psychopathology
AXIS I C O M O R B I D I T Y I N BIPOLAR D I S O R D E R
by patients' or relatives' recall bias. Analyses
were performed independently of whether
bipolar patients had prevalently manic or
depressive episodes. The small size of the
cohort examined did not allow us to investigate the effect of each single comorbid Axis
I disorder on psychotic phenomenology.
Implications and futura directions
Our results support the importance of
dacctlng specific patterns of comorbidity
in patients with severe bipolar disorder (Cassano et al, 1998, 1999; Schatzberg, 1998).
Either substance abuse or non-abuse Axis I
comorbidity may have an impact on phenomenology of bipolar disorder. These associations in their various forms should be
investigated further because of potential
implications on age of onset of affective illness and risk of suicide. C l d c a t i o n of the
relationships between bipolar disorder and
other Axis I conditions, which are traditionally situated at a lower hierarchical level,
may be of great value in bemr understanding
the course of bipolar illness and the phenomenology of acute and interepisodic phases,
and in desaibii subsets of patients whose
symptoms are not entirely accounted for by
mood disorder. Furthermore, comorbidity
may have clear implications for treatment.
Different forms of bipolar disorder, for
example, could have different treatment
responses, and it would be inmesting to
see whether cornorbidity prcdicts treatment
outcome.
REFERENCES
!STEFAN0 PINI. MD. UUANA DEUYKXX MD. CONCETTlNA MASTROUNQUE. MD, GIOVANNI
MARCACCI. MD. ALESSANDRA PAW\SOCU. MD, SWNAWGNOU. MD. !STEFAN0PAUANTI. MD.
GIOVANNI B. CASSANQ FRCF3ych. Department of Rychiiry, Neurobidogy, Pharmacology and
Biotechdogy, U n i i y of Pisa. Italy
Correspondence: D r Stefano Pini. DPNFB-Department of Psychiatry,University of Pisa, via Roma 67,
56100 Pisa. ltaly.Tel: +39 50 835-419; Fax: +39 50 21-581
(First received 18 August 1998. final revision II February 1999. accepted 9 March 1999)
-. -.-.
u d (1999) Multiple anxiety disorders
comorbidity in patients with mood spectrum d i r s
with psrchotic features. Americwr Jomdof R y d m q
w
t
r
)
c
In,345-350.
G.k.n,E. A. (lr)The scientifi~basis of clinical trials.
Uinicol
k r c h . 31 2587-2590.
sonm,tC..~K.T.&~.WA.
(1994) substance abuse and bipolar
d i r . w d of N e w and Mental Disease,
112 349-352.
A m r k u , ~ ~ ( I # 1 ) ~ n o s t i c703-707.
and StotisticoJ Manual of Mental Disuders (3rd edn.
&)
(MM- Ill-R).\IVashington. DC: APA.
KdqRE.,Mc~S.L.Sadwdd.S.n,aaI
(1995) Outcome and m m h i d i i in first- cMnparcd
An&msm, N.C. (I=)The Scde @ the Arcenmnt of
with multiple episode mania.lownd of Nerwus and
Nega* 4ymptomr krua Ciy. IA: University of Iowa.
M e d Chase. IU. 320-324.
k r d w . ~ C . ~ ~ a . . ~ c . . a d
(lr)The e p i d e m i i of MM-Ill- R bipolar I
disorder in a general population wrwy. psVChdogico1
Medicine. ll. 1079- 1089.
S. C. (1995) The relationship
between substanceabuse and bipolar d i . j m m dof
U i M Whtg 56 (suppl. 3). 19-24.
WJ.S,bh.rrRLWPW(lm)
l m g i i study dm-occurring psychiatric d
and substance use.ofthe Americm Acdemy
ChiM ord no^ Psydrrov)c W, 322-330.
i
of
C.rmqGb,rW.S,Suaoni,M..ad(lm)
Occurrence and clinical amhOesof psychiatric
amwbb3iy in patients with psychotic dkorden.@d
of Uinical
B,60-68.
m
156,474-476.
K.14K.L&Kloh,D.N.(I9M)Therebtiwhip
between age at onset and cxxnorbidiy in psychiatric
d i . W $Nervous and Mend Di-,
184,
K.T. & Son-
htc.Y)dnraqS,~R.ad(lm)
Substance misuse and risk of asgnnion and dfending
among the severely mentally ill.British
of
McwS.L.~S.n,Wlrt,sA..ad
(I- PhaKumdogydaddexent and adult mania in
hospitalid patientswith bipolar d i . American
of Rychioq 154,4449.
~,J.€&Garhnn,RR(I~TheBrief
F'sychii RatingScale. RychdogicolReports. 14799-812.
Wmt&ug, A. F. (Im)Bipolar d
i recent
issws in diagnosis and dassifiiati0n.m ofoinicol
Psydrrov)c 59 (suppl. 6). 5- 10.
*
Spitsr,RL.~J.B.W.Qbbon.H..ad
(1987) hctured Uinicd Interview @ DSM-Ill-R btk4 kmion (KID-P). New York. New York State
R y c h i i Institute.
Jblbwdd.Ln,~n,Jsa.A.L..t.r(Im)
Comwbidiy in mania a first h o a p i t d i . America,
) o u d oW
f mc)rt
149,554556.
-.McBr~SL,KdqREJr..ad(l9M)
The dfeN d anbmdent substance abuse on the
devdopment of first-episode psychotic mania.
Jnmd ofk r c h , 30,594.
.
- kdqR Jr., Mew L.u d(1998)
khn-month outcome afber a first hospital&
for afTeck pqdmis. A r k of Generd fspzharty
55.49-55.
m n d a u , ~ r n ~ ~ ~ N V c * . h . ~ ( ~
D e p m seconQry *Dother p s y d r i disordas
a n d m e d i c d i N n s x s . ~ ~ o f P s y c h i o t r 145.
)c
233-237.