. 46
( 68 .)


(2001a), for example, looked at 392 bipolar affective patients and found no
difference between mixed and pure manic patients with regard to substance-
abuse disorders. Similar results were obtained from our own group: in a prospec-
tive study of 149 bipolar patients, there were no significant differences in the
frequency of substance-abuse disorders in bipolar affective patients with or with-
out mixed affective episodes. Similarly, as shown in the study of Cassidy et al.
(2001a), on a descriptive and non-significant level, the ˜˜mixed™™ bipolar group had
even lower rates of substance-abuse disorders than the pure mania group (Brieger
et al., in preparation).
Rapid-cycling bipolar affective disorders may concur with higher rates of
substance-abuse disorders (Calabrese et al., 2000), but such an idea is mainly
based on clinical observation or preliminary data (Keller et al., 1986; Sonne et al.,
1994). One family study reported (without statistical significance) higher rates of
substance abuse in families of subjects with rapid-cycling than in families of
subjects with non-rapid-cycling bipolar affective disorder (Lish et al., 1993). In
our own data set, we have also observed (without significance) that patients with
rapid cycling exhibited higher rates of substance-abuse disorders (Brieger et al., in
preparation), although our rapid-cycling group is very small (n ¼ 12). There is some
consideration that patients with substance-abuse disorder and bipolar affective
267 Comorbidity in mixed states and rapid-cycling disorders

disorder may have more illness episodes than patients with bipolar affective disorder
without substance-abuse disorder (Sonne and Brady, 1999; Calabrese et al., 2000),
but not all data support such a hypothesis. Yet, patients with substance abuse and
bipolar affective disorders differ from patients with bipolar affective disorders without
substance abuse in other respects: they show more hospitalization (Brady et al., 1991),
have an earlier age at onset of bipolar disorder (Sonne et al., 1994; Brieger et al., in
preparation) and have an overall more unfavorable course of bipolar illness (Sonne
and Brady, 1999; Marneros and Brieger, 2002; Brieger et al., in preparation).

Anxiety disorders
There is a fundamental relationship between bipolar affective disorder and anxiety
disorders (including obsessive-compulsive disorder: OCD) (Perugi et al., 1999;
Zarate and Tohen, 1999; Brieger, 2000), which has been shown by epidemiological
(e.g., Kessler, 1999) and clinical (e.g., McElroy et al., 2001; Perugi et al., 2001)
studies. Panic disorder, social phobia, and OCD seem to have a specific relationship
with bipolar affective disorder (Chen and Dilsaver, 1995a, b; Perugi et al., 1999).
One series of factor analyses (Cassidy et al., 1998a, b, 2001a, 2001b) delineated
five independent factors in mania: (1) dysphoric mood; (2) psychomotor pressure;
(3) psychosis; (4) increased hedonic function; and (5) irritable aggression. The
factor ˜dysphoria™ strongly correlated with anxiety. Based on these studies, an
alternative set of six diagnostic criteria for mixed episodes was proposed (Cassidy
et al., 2000), which (with a threshold of two symptoms) consisted of (1) anxiety;
(2) depressed mood; (3) anhedonia; (4) guilt; (5) suicide; and (6) fatigue. In this
concept, anxiety is an integral aspect of mixed bipolar affective phenomenology, as
was earlier observed by Post et al. (1989). Furthermore, a comparison between
patients with mixed mania and agitated depression (Swann et al., 1993) proved that
these two groups were (at least partially) similar in respect to observed anxiety.
These observations support the idea that dysphoria and bipolar mixed states are
characterized by elevated levels of dimensional anxiety. This is in agreement with
the theories of Akiskal, Koukopoulos, and others, who have postulated that mixed
bipolar affective disorders are the product of an admixture of anxiety or depressive
symptoms (or such temperaments) with manic episodes (Koukopoulos and
Koukopoulos, 1999; Akiskal and Pinto, 2000; Akiskal et al., 2002). In a comparison
of bipolar patients with pure and mixed manic episodes, we found that patients
with mixed mania exhibited higher rates of anxious temperament than patients
with pure mania (Brieger et al., 2003b). Here, anxious temperament may be
interpreted as a trait marker with longitudinal stability. Nevertheless, there are
studies that do not support such an idea and have found no significant difference
between patients with mixed and pure manic episodes concerning dimensional
268 P. Brieger

anxiety. For example, one Italian study found no differences concerning self-rated
or observer-rated anxiety (Dell™Osso et al., 2000) in 90 patients with mixed and
pure mania with psychotic features.
Patients with mixed bipolar affective disorder may also exhibit higher frequencies
of full-blown anxiety disorders than patients with pure mania. This was already
observed in the Iowa study (Winokur et al., 1969). A later study (Feske et al., 2000)
showed that, amongst bipolar affective patients, those with a depressive or mixed
episode fulfilled criteria of an anxiety disorder in roughly half of cases, while this was
only the case in every fifth patient with a pure manic episode. Yet, partially negative
studies can also be found: in a comparison of patients with mixed mania, pure
mania, and bipolar depression with psychotic features, only two differences con-
cerning anxiety disorders were reported: depressive patients had higher rates of
OCD and simple phobia (Dell™Osso et al., 2000).
The temporal relationship between anxiety disorder and mixed bipolar affective
disorder remains complex. A certain proportion of panic disorders in bipolar
affective patients presents during hypomanic episodes, while social phobia nearly
always precedes (hypo)mania (Perugi et al., 2001). Therefore, Perugi et al. conclude
that in such patients panic disorder may be a reflection of mixed (hypo)manic
symptomatology (Perugi et al., 2001).
Altogether there is some indication that mixed bipolar affective episodes corres-
pond with an elevated risk for co-occurring anxiety disorders and higher rates of
trait anxiety (or anxious temperament). Concerning OCD, it was once observed
(Strakowski et al., 1998) that OCD occurred more often in mixed mania than in
pure mania, although full-blown OCD had a relatively lower prevalence than
many other psychiatric disorders “ a result stemming from a rather small sample
size (12 patients in both groups).
We are not aware of substantial data that support the idea that rapid cycling has
a special link to anxiety disorders, including OCD.

Personality disorders
Personality disorders in affective disorders occur at higher rates (Brieger et al.,
2003a). Nevertheless, interactions between personality and affective disorder are
complex (Akiskal et al., 1983): is the assessed personality a predisposing or
complicating factor, or is it a result or an epiphenomenon of the affective disorder?
The present diagnostic system of personality disorders has undergone fundamen-
tal critique (Livesley, 1998). The relevance of comorbid personality disorders in
affective disorder may be more complex than generally assumed; for example,
there is an indication that only specific patterns of personality disorders have a
prognostic value (Brieger et al., 2002a). With regard to mixed episodes, there is
269 Comorbidity in mixed states and rapid-cycling disorders

some indication that they may (at least partially) result from an admixture of a
certain personality type or a certain temperament to an affective episode “ a
theory that has been advocated most of all by Hagop Akiskal (Akiskal, 1996;
Akiskal et al., 1998). There are now several studies supporting such a hypothesis
(Dell™Osso et al., 1991; Akiskal et al., 1998; Brieger et al., 2003b), although the
observed differences between pure and mixed manic patients in respect to tem-
perament are not significant enough to explain the difference fully. Furthermore,
methodologically, the distinction between mood ˜˜state™™ and temperament ˜˜trait™™
cannot always be made satisfactorily. Concerning personality as assessed by the
five-factor model (Costa and Widiger, 1994), we found no difference between
patients with pure and mixed manic episode in a relatively small sample (Brieger
et al., 2002b).
We are not aware of any studies that have linked rapid cycling to the presence of
a personality disorder.

Other psychiatric disorders
There is a large diagnostic overlap between attention-deficit hyperactivity disorder
(ADHD) and bipolar disorder in children and adolescents (Geller et al., 2002b).
Geller et al. (2002a, b, c) have outlined a ˜˜prepubertal and early adolescent bipolar
disorder phenotype,™™ which may be superior to DSM-IV criteria in recognizing
bipolar children and adolescents. Still, in their sample of 93 subjects with a
˜˜prepubertal and early adolescent bipolar disorder phenotype,™™ 86.5% suffered
from comorbid ADHD, 87.1% showed rapid cycling (77.4% ultradian!), and
54.8% presented with mixed mania. As only 16 of 93 subjects with bipolar disorder
had no ADHD, it seems futile to analyze the effect of comorbid ADHD on bipolar
disorder. Rather, one has to acknowledge that (at least with the present diagnostic
criteria of ADHD), the great majority of childhood and adolescent cases of bipolar
disorders present with rapid (and even ultradian) cycling and with co-occurring
ADHD, and that more than half of the subjects with bipolar affective disorder
exhibit mixed episodes.
We are not aware of any studies that have looked specifically at the relationship
between mixed states or rapid-cycling forms of bipolar affective disorders and
eating disorders, sexual and gender-identity disorders, somatoform disorders, or
dissociative disorders.
Furthermore, we have not discussed the relationship between psychotic and
schizophrenic disorders and ˜˜atypical™™ forms of bipolar affective disorders, as this
lies outside the realm of this chapter. Where to draw the line between bipolar
affective disorders, schizophrenic disorders, and schizoaffective disorders is a
270 P. Brieger

fundamental question of concept and phenomenology, and not a question of

The general effect of comorbidity
The majority of patients with a bipolar affective disorder and a co-occurring
psychiatric disorder suffer from more than one additional disorder (McElroy
et al., 2001; Vieta et al., 2001). In one analysis from the Stanley Foundation
Bipolar Treatment Outcome Network, 35% of all bipolar patients (n ¼ 288) had
no lifetime comorbid Axis I disorder, 23% had one such comorbid disorder, 18%
had two, and 24% had three or more. With regard to current comorbid Axis I
disorders, the numbers were as follows: 67% had none, 20% had one disorder, 7%
had two disorders, and 6% had three or more disorders (McElroy et al., 2001).
In another comparison of bipolar affective patients (n ¼ 129) with or without a
current comorbid psychiatric disorder (Vieta et al., 2001), patients with a current
comorbid disorder had a history of significantly more mixed episodes than
patients without such current comorbid disorders. In our follow-up study
(n ¼ 149), we came to the same result (Brieger et al., in preparation). Both studies
have included Axis I and II disorders and, in our study, this difference is primarily
the effect of an excess of cluster C personality disorders in the group of patients
with previous mixed episodes. Interestingly, the Stanley Foundation Bipolar
Treatment Outcome Network study (McElroy et al., 2001), which looked at only
Axis I disorders and not at personality disorders, found no significant difference
between patients with or without such comorbid diagnoses in regard to dysphoric
mania, whether one looked at lifetime or current comorbidity.
The Stanley Foundation Bipolar Treatment Outcome Network study (McElroy
et al., 2001) indicated that patients with comorbid Axis I disorders show more
cycle acceleration and possibly more rapid cycling. The latter result barely missed
statistical significance, but the study used strict correction for multiple compari-
sons; therefore, this may have been a false-negative result. Vieta et al. (2001) did
not come to the same result, but here the smaller sample size has to be considered.

Comorbidity research illustrates the shortcomings of the present diagnostic
systems and must therefore go further than merely describing present rates of
co-occurring disorders. Instead, it has to uncover trends, relationships, and links
between different disorders. Even if one were to suppose that the different defin-
itions of mixed episodes and of rapid cycling described two widely homogeneous
groups, the present review does not give much support for the idea that such forms
271 Comorbidity in mixed states and rapid-cycling disorders

of bipolar disorder constitute distinct diagnostic categories. Rather, this review
supports the idea that rapid cycling and mixed episodes are manifestations of
bipolar illnesses, which occupy certain regions of the multidimensional bipolar
affective spectrum. Still, this does not contradict the idea that these forms of
bipolar affective disorders have a prognostic validity. In all areas of our review,
there is the overall tendency that mixed states and rapid-cycling forms of bipolar
affective disorder constitute more unfavorable forms of the underlying illness.
Mixed affective episodes have a link with anxiety disorders and anxious-dependent
personality disorders (cluster C). Rapid cycling may have a link with substance
abuse and with certain neuropsychiatric disorders, or, perhaps, these neuro-
psychiatric disorders may mimic rapid cycling. Undoubtedly, children and ado-
lescents with bipolar affective disorders usually suffer from ultradian cycling and
ADHD. All of these co-occurring factors are reasons for the deterioration of the
course of a bipolar affective disorder.
It seems simplistic to distinguish between ˜˜typical™™ and ˜˜atypical,™™ or between
˜˜mixed™™ and ˜˜non-mixed,™™ or between ˜˜rapid-cycling™™ and ˜˜non-rapid-cycling™™
forms of bipolar illness. The reality of bipolar illness is much more complex.
Several “ mostly dimensional “ factors affect the course and prognosis. Amongst
these, we count phenomenology of episodes, cycle acceleration, cycle frequency,
occurrence of full cycles, early age at onset, comorbid substance abuse, comorbid
personality disorders, comorbid anxiety disorders, neuropsychiatric abnormal-
ities, time in depression, and other genetic, biological, psychosocial, and person-
ality factors. We think it is necessary to assess bipolar affective illness in the light of
these factors in order to understand it better.


Akiskal, H. S. (1996). The temperamental foundations of affective disorders. In Interpersonal
Factors in the Origin and Course of Affective Disorders, ed. C. Mundt, K. Hahlweg, and
P. Fiedler, pp. 3“30. London: Gaskell.
Akiskal, H. S. and Pinto, O. (2000). Soft bipolar spectrum: footnotes to Kraepelin on the
interface of hypomania, temperament and depression. In Bipolar Disorders: 100 Years after
Manic-Depressive Insanity, ed. A. Marneros and J. Angst, pp. 37“62. Dordrecht: Kluwer
Academic Publishers.
Akiskal, H. S., Hirschfeld, R. M., and Yerevanian, B. I. (1983). The relationship of personality to
affective disorders. Arch. Gen. Psychiatry, 40, 801“10.
Akiskal, H. S., Hantouche, E. G., Bourgeois, M. L., et al. (1998). Gender, temperament, and the
clinical picture in dysphoric mixed mania: findings from a French national study (EPIMAN).
J. Affect. Disord., 50, 175“86.
272 P. Brieger

Akiskal, H. S., Brieger, P., Mundt, C., Angst, J., and Marneros, A. (2002). Temperament und
¨ ¨
affektive Storungen. Die TEMPS-A Skala als Konvergenz europaischer und US-amerika-
nischer Konzepte. [Temperament and affective disorders. The TEMPS-A scale as a conver-
gence of European and US-American concepts.] Nervenarzt, 73, 262“71.
American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders,
4th edn. Washington, DC: American Psychiatric Association.
Bamrah, J. S. and Johnson, J. (1991). Bipolar affective disorder following head injury.
Br. J. Psychiatry, 158, 117“19.
Bartalena, L., Pellegrini, L., Meschi, M., et al. (1990). Evaluation of thyroid function in patients
with rapid-cycling and non-rapid-cycling bipolar disorder. Psychiatry Res., 34, 13“17.
Basso, M. R., Lowery, N., Neel, J., Purdie, R., and Bornstein, R. A. (2002). Neuropsychological
impairment among manic, depressed, and mixed-episode inpatients with bipolar disorder.
Neuropsychology, 16, 84“91.
Bauer, M. S., Whybrow, P. C., and Winokur, A. (1990). Rapid cycling bipolar affective disorder.
I. Association with grade I hypothyroidism. Arch. Gen. Psychiatry, 47, 427“32.
Berthier, M. L. (1992). Post-stroke rapid cycling bipolar affective disorder. Br. J. Psychiatry, 160, 283.
Brady, K., Casto, S., Lydiard, R. B., Malcolm, R., and Arana, G. (1991). Substance abuse in an
inpatient psychiatric sample. Am. J. Drug Alcohol Abuse, 17, 389“97.
Brieger, P. (2000). Comorbidity in bipolar disorder. In Bipolar Disorders: 100 Years after Manic-
Depressive Insanity, ed. A. Marneros and J. Angst, pp. 215“29. Dordrecht: Kluwer Academic
¨ ¨
Brieger, P. and Marneros, A. (2000). Komorbiditat bei psychiatrischen Storungen. Einige
¨ berlegungen. [Comorbidity in psychiatric diseases. Some theoretical consid-
theoretische U
erations.] Nervenarzt, 71, 525“34.
Brieger, P., Ehrt, U., Bloeink, R., and Marneros, A. (2002a). Consequences of comorbid
personality disorders in major depression. J. Nerv. Ment. Dis, 190, 304“9.
Brieger, P., Ehrt, U., Roettig, S., and Marneros, A. (2002b). Personality features of patients with
mixed and pure manic episodes. Acta Psychiatr. Scand, 106, 179“82.
Brieger, P., Ehrt, U., and Marneros, A. (2003a). Frequency of comorbid personality disorders in
bipolar and unipolar affective disorders. Compr. Psychiatry, 44, 28“34.


. 46
( 68 .)