. 6
( 68 .)


(1) Four or more episodes of depression, mania, or hypomania in the previous 12 months.
(2) Patients need not have an intervening euthymic interval for a mania and a depression to be
counted as two episodes.
(3) Numbers of episodes were tabulated, rather than numbers of cycles; for example, two cycles
in which manic episodes are biphasically coupled with depressions followed by euthymic
intervals would count as four episodes and satisfy criteria for rapid cycling.
(4) Episodes are demarcated by a switch to a mood state of opposite polarity or by a period of
relative remission lasting 2 months (DSM-IV, American Psychiatric Association, 1994).
Therefore, consecutive episodes with the same polarity must be separated by a period of
relative remission lasting 2 months.

DSM-IV included rapid cycling as a specifier of longitudinal course, but not as a
specific mood disorder subtype (American Psychiatric Association, 1994). ICD-10
(World Health Organization, 1991) did not include any specifier or subgroup
˜˜rapid cycling.™™ According to DSM-IV, the specifier ˜˜with rapid cycling™™ can be
applied to bipolar I disorder or bipolar II disorder.
The essential feature of a rapid-cycling bipolar disorder is the occurrence of four
or more mood episodes during the previous 12 months. These episodes can occur
in any combination and order. The episodes must meet both the duration and
symptom criteria for a major depressive, manic, mixed, or hypomanic episode and
must be demarcated by either a period of full remission or by a switch to an episode
of the opposite polarity. Manic, hypomanic, and mixed episodes are counted as
being on the same pole (e.g., a manic episode immediately followed by a mixed
episode counts as only one episode = when considering the specifier ˜˜with rapid
cycling™™). Except for the fact that they occur more frequently, the episodes that occur
in a rapid-cycling pattern are no different from those that occur in a non-rapid-cycling
pattern. Mood episodes that count toward defining a rapid-cycling pattern
exclude those episodes directly caused by a substance (e.g., cocaine, corticoster-
oids) or a general medical condition (American Psychiatric Association, 1994).
27 Beyond major depression and euphoric mania

Prolonged single episodes accompanied by intermittent fluctuations within the
mood state (i.e., cycling above or below baseline due to changes in medication doses or
levels) are counted as one episode. For example, patients who have one long period of
mania followed by a short period of hypomania due to the transient use of neuroleptics
or benzodiazepines, followed by a return to mania, are counted as having only one
episode (Calabrese et al., 2000). DSM-IV applies ˜˜rapid cycling™™ only to bipolar I and
bipolar II disorders in recognition of the reality that rapid cycling of unipolar depres-
sion is extremely rare, and when it does occur, the family history is usually positive for
bipolar disorder (Tay and Dunner, 1992; Kilzieh and Akiskal, 1999).
As noted above, the cut-off of four episodes per year has been criticized for being
arbitrary. The key question, still unanswered, is whether cycle length distributes
more or less evenly across a spectrum, or if there is a true bimodal distribution
into two concrete (as opposed to arbitrary) subgroups “ namely, rapid-cycling and
non-rapid-cycling (Goodwin and Jamison, 1990; Coryell et al., 1992). Another
problem regarding the definition of rapid cycling relates to the nature and duration
of the interepisodic period, which varies widely among studies. Studies differ regarding
its duration and level of symptoms (Kilzieh and Akiskal, 1999): some require partial
or full remission for at least 2 months or a switch to an episode of opposite polarity,
other researchers have set the duration of remission as low as 4 weeks, while still
others require a period of euthymia as long as the proximate episodes. The
prevalence of rapid cycling in bipolar populations is also indeterminate because most
of the data come from studies mostly done at tertiary centers with a high proportion of
difficult-to-treat patients (Kilzieh and Akiskal, 1999). No community-based studies
have been conducted, and the true prevalence of rapid cycling in an unselected bipolar
patient population remains unknown (Kilzieh and Akiskal, 1999). Another problem
related to the estimation of the prevalence of rapid cycling is its longitudinal instability;
that is, it often occurs intermittently during the course of illness (Coryell et al., 1992).
Studies on the prevalence of rapid cycling in a clinical bipolar population range
from 24.2% (Tondo et al., 1998) down to 13.6% (Maj et al., 1994), with others in
between: Coryell et al. (1992) 18.5%, Dunner and Fieve (1974) 20%, and
Koukopoulos et al. (1980) 19%. A prevalence of rapid cycling with ˜˜approximately
5“15% of people with bipolar disorders seen in mood disorder clinics™™ is cited.
Higher prevalence rates than those noted above were reported by Cowdry et al.
(1983) from the US National Institute of Mental Health (56%), but this probably
reflects the specialization of this major research center.
The gender difference “ on average, more than 70% of rapid cyclers are females “
is the most extensively replicated finding in rapid cycling (Table 1.3).
The meta-analysis of Tondo et al. (1998) showed that, although the majority of
rapid-cycling cases (72%) are women, rapid cycling occurred in less than 30% of
the total female bipolar population.
28 A. Marneros and F. K. Goodwin

Another explanation delivered by the finding of Coryell et al. (1992) may derive
from the assumption that a greater cyclicity seems to be frequently associated with

Age at onset
According to the findings of Fujiwara et al. (1998), age at onset (for rapid-cycling
patients) can be divided into early (onset at an age of 25 years and younger) and later
onset (onset at an age of 26 years or older). These data suggest that early- and late-onset
bipolar disorders are distinct illness subtypes with different courses and responses
to treatment (Calabrese et al., 2000). The Japanese authors concluded that patients
with an earlier onset tend to have rapid cycling at an early stage and a good response
to carbamazepine. Those with later onsets tended to have relatively long latency until
the appearance of rapid cycling and a good response to lithium.
Rapid cycling, as well as mixed states in childhood and adolescence, has not
been investigated systematically, but some reports have shown that the prevalence
of these states in childhood and adolescence is not rare (Calabrese et al., 2000; see
Chapter 10).

Family studies and genetics
Most of the studies of the families of patients with rapid-cycling bipolar disorder
show no difference between rapid- and non-rapid-cycling patients. That is, rapid
cycling is not more frequent in families of patients with rapid cyclers (Nurnberger
et al., 1988; Coryell et al., 1992; Lish et al., 1993; Maj et al., 1994). Although the
studies mentioned appear to argue convincingly against any inheritance of rapid
cycling in general, the less common form of early-onset rapid cycling may be
family-related (Calabrese et al., 2000). Studies reporting on genetic abnormalities
(more or less anecdotal) are rare and should be replicated (Kilzieh and Akiskal,
1999; Calabrese et al., 2000).

Biological data
Also uncertain are other biological correlations of rapid cycling, as described in the
present book by Grunze and Walden (Chapter 14).

The issue of comorbidity as it relates to rapid cycling is complex. Calabrese et al.
(2000) reviewed the extensive literature on thyroid dysfunction in patients with
bipolar rapid cycling, noting that while many studies do report an association
between rapid cycling and reduced thyroid function, not every study confirmed it.
Alcohol and drug abuse is another comorbid disorder that is frequently associated
with the acceleration of remanifestations and rehospitalizations, but there is no
29 Beyond major depression and euphoric mania

systematic research regarding association with rapid cycling (Brieger and
Marneros, 1999; Calabrese et al., 2000; see Chapter 12). Finally, a few reports,
essentially anecdotal, suggest an association between rapid cycling and neuropsy-
chological deficits (Calabrese et al., 2000; see Chapter 12).

Longitudinal prognosis
The impact of rapid cycling on longitudinal prognosis is also uncertain. While
Coryell et al. (1992) and Wu and Dunner (1993) do not find an association between
rapid cycling and a worsening of long-term prognosis, Okuma (1993) did. These
differences may reflect sampling. That is, among patients for whom rapid cycling is
transient and intermittent, one might not expect a negative effect or prognosis.

The treatment of rapid cycling is discussed extensively by Calabrese et al. (2000), as
well as by Elhaj and Calabrese (see Chapter 3).

Bipolar schizoaffective mixed states
Schizoaffective disorders present as unipolar or bipolar forms in a way similar to
mood disorders (Marneros et al., 1989a“c; 1990a“c; 2000), as is reflected in both
DSM-IV and ICD-10 (American Psychiatric Association, 1994; World Health
Organization, 1993). However, there are differences between DSM-IV and ICD-10.
While DSM-IV defines two subtypes based on longitudinal course, namely bipolar
and depressive, ICD-10 defines three types (manic, depressive, and mixed) based
on the most recent episode, rather than longitudinal course (World Health
Organization, 1993). These differences present a difficulty for cross-national
Tables 1.4 and 1.5 illustrate how differently ICD-10 and DSM-IV handle the
definition of schizoaffective disorder.
While the main problem with the definition of ICD-10 concerns the long-
itudinal issue, the problem with DSM-IV concerns both “ cross-sectional and
longitudinal issues. The problem with the cross-sectional definition of DSM-IV
concerns time criteria for criterion B (during the period of illness there have been
delusions or hallucinations for at least 2 weeks in the absence of prominent mood
symptoms). Obviously, that is an attempt of the DSM-IV to separate schizoaffec-
tive disorders from psychotic mood disorders. The DSM-IV definition of mood
disorders is broad, including even those with mood-incongruent symptoms (even
first-rank schizophrenic symptoms) as the mood disorders. But the chronological
criterion is an arbitrary one (2 weeks of psychotic symptoms without mood
disorders is schizoaffective; less than 2 weeks is a psychotic mood disorder). One
30 A. Marneros and F. K. Goodwin

Table 1.5 Schizoaffective disorders (295.70) according to DSM-IV

A. An uninterrupted period of illness during which, at some time, there is a major depressive
episode, a manic episode, or a mixed episode concurrent with symptoms that meet criterion
A for schizophrenia. Note: The major depressive episode must include criterion
A1: depressed mood
B. During the same period of illness, there have been delusions or hallucinations for at least 2
weeks in the absence of prominent mood symptoms
C. Symptoms that meet criteria for a mood episode are present for a substantial portion of the
total duration of the active and residual periods of the illness
D. The disturbance is not due to the direct physiological effects of a substance (e.g., a drug of
abuse, a medication) or a general medical condition.

Specify type:
Bipolar If the disturbance includes a manic episode (or a mixed episode or a manic or a
type: mixed episode and major depressive episodes)
Depressive If the disturbance only includes major depressive episodes

problem with this is that the beginning of a psychotic episode can rarely be assessed
exactly. Every clinician knows that there is usually a gap of many days, weeks, or
months between the beginning of a psychotic episode and admission to a hospital.
Reconstruction of the psychopathological picture, retrospectively, is fraught
with difficulty. Given the likelihood that the psychotic period would be under-
estimated, many patients who are really schizoaffective could be diagnosed as
schizophrenic or as having psychotic mood disorder.
Furthermore, the intensity of both concurrent syndromes can vary enormously
during an episode “ it seems arbitrary to give chronological priority to the psychotic
symptoms over the mood component. It is curious that DSM-IV rejected Jasper™s
hierarchical diagnostic principle, which suggested a diagnostic superiority of schizo-
phrenic symptoms over affective symptoms but, regarding the chronological criter-
ion of the schizoaffective definition, obviously made an exception!
Considering what is now known about schizoaffective disorders (see overviews
in Marneros and Tsuang, 1986, 1990; Marneros et al., 1995), we suggest that the
definition of schizoaffective disorders should contain two components: cross-
sectional and longitudinal. The cross-sectional definition should be the definition
of an episode, while the longitudinal definition should be that of a disease or
disorder. The cross-sectional definition of a schizoaffective episode should be
based on the simultaneous occurrence of symptoms of a schizophrenic and a
mood episode, independent of the chronological manifestation. Thus, we agree
31 Beyond major depression and euphoric mania

Manic Schizomanic Mixed schizoaffective
Depressive Schizodepressive Schizophrenic Others

Fig. 1.12 Longitudinal course of patients with mixed schizo-manic-depressive episodes in the Halle
Bipolarity Longitudinal Study (HABILOS).

with the definition of ICD-10, which yields three types of schizoaffective episodes:
schizodepressive, schizomanic, and mixed.
The longitudinal definition of the schizoaffective disorder should consider the
sequential occurrence of mood and schizophrenic episodes during the course. The
longitudinal research demonstrates that the course of schizoaffective disorders can
be very unstable because schizoaffective episodes, pure mood episodes, and pure
schizophrenic episodes can each occur at different points in the patient™s long-
itudinal course (Fig. 1.12). What are such disorders when viewed longitudinally?
Are they to be considered mood disorders because of the pure mood episodes, or
schizophrenic disorders because of some pure schizophrenic episodes, or schizo-
affective disorders because of some schizoaffective episodes? Relevant to this
question is the finding that there are no differences between patients who have
only had schizoaffective episodes, and those in whom schizoaffective episodes
occur along with pure mood and schizophrenic episodes. Thus, there are no
differences between the ˜˜concurrent™™ and the ˜˜sequential™™ type of schizoaffective
disorder (Marneros et al., 1986, 1991a, b). Patients who change from pure mood
episodes to pure schizophrenic episodes and vice versa do not differ from patients
having schizoaffective episodes. In this sense, Marneros et al. suggest a longitudinal
definition of schizoaffective disorders, including a concurrent and a sequential
32 A. Marneros and F. K. Goodwin

Bipolar patients
(n = 277)

Bipolar affective
Bipolar schizoaffective
(n = 100)
patients 8.3%
(n = 23)

55.6% Bipolar unclassifiable
(n = 154)

Fig. 1.13 Tenth Revision of the International Classification of Diseases (ICD-10) longitudinal
classification of bipolar patients in the Halle Bipolarity Longitudinal Study (HABILOS).

type (Marneros et al., 1986“2004): the concurrent type is characterized by the
coincidence of schizophrenic and affective episodes, while the sequential type is
characterized by the longitudinal change from schizophrenic to affective episodes
and vice versa (Marneros et al., 1986, 1988c, 1989a, 1991b, 2000; Marneros and
Angst, 2000).
How essential it is to make a longitudinal diagnosis is illustrated by the


. 6
( 68 .)