Journal of Affective Disorders 75 (2003) 1–10

www.elsevier.com / locate / jad

Cyclothymic OCD: a distinct form?

a, b c d e f
E.G. Hantouche *, J. Angst , C. Demonfaucon , G. Perugi , S. Lancrenon , H.S. Akiskal
a
´ ˆ ` Hospital, 47 Bd de l’ Hopital
Mood Center, Department of Psychiatry, Pitie-Salpetriere ˆ , 75013 Paris, France
b
Zurich University Psychiatric Hospital, Zurich, Switzerland
c
AFTOC, Chateaufort, France
d
Psychiatric Institute of Pisa, Pisa, Italy
e
Syliastat, Antony, France
f
International Mood Center, VA Hospital and University of California, San Diego, CA, USA

Received 30 January 2002; accepted 2 April 2002

Abstract

Background: Clinical research on the comorbidity of obsessive compulsive disorder (OCD) and other anxiety disorders
has largely focused on depression. However in practice, resistant or severe OCD patients not infrequently suffer from a
masked or hidden comorbid bipolar disorder. Method: The rate of bipolar comorbidity in OCD was systematically explored
among 453 members of the French Association of patients suffering from OCD (AFTOC) as well as a psychiatric sample of
OCD out-patients (n 5 175). As previous research by us has shown the epidemiologic and clinical sample to be similar, we
combined them in the present analyses (n 5 628). To assess mood disorder comorbidity, we used structured self-rated
questionnaires for major depression, hypomania and mania (DSM-IV criteria), self-rated Angst’s checklist of Hypomania and
that for the Cyclothymic Temperament (French version developed by Akiskal and Hantouche). Results: According to
DSM-IV definitions of hypomania / mania, 11% of the total combined sample was classified as bipolar (3% BP-I and 8%
BP-II). When dimensionally rated, 30% obtained a cut-off score $ 10 on the Hypomania checklist and 50% were classified
as cyclothymic. Comparative analyses were conducted between OCD with (n 5 302) versus without cyclothymia (n 5 272).
In contrast to non-cyclothymics, the cyclothymic OCD patients were characterized by more severe OCD syndromes (higher
frequencies of aggressive, impulsive, religious and sexual obsessions, compulsions of control, hoarding, repetition); more
episodic course; greater rates of manic / hypomanic and major depressive episodes (with higher intensity and recurrence)
associated with higher rates of suicide attempts and psychiatric admissions; and finally, a less favorable response to
anti-OCD antidepressants and elevated rate of mood switching with aggressive behavior. Limitation: Hypomania and
cyclothymia were not confirmed by diagnostic interview by a clinician. Conclusion: Our data extend previous research on
‘‘OCD-bipolar comorbidity’’ as a highly prevalent and largely under-recognized and untreated class of OCD patients.
Furthermore, our data suggest that ‘‘cyclothymic OCD’’ could represent a distinct form of OCD. More attention should be
paid to it in research and clinical practice.
 2003 Elsevier Science B.V. All rights reserved.

Keywords: Obsessive compulsive disorder; Hypomania; Cyclothymia; Comorbidity; OCD French Association

*Corresponding author.
E-mail address: [email protected] (E.G. Hantouche).

0165-0327 / 03 / $ – see front matter  2003 Elsevier Science B.V. All rights reserved.
doi:10.1016 / S0165-0327(02)00461-5
2 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10

1. Introduction (n 5 175), OCD out-patients of psychiatric services

were assessed. Patients had to present an OCD
The close relationship between obsessive-compul- (according to DSM-IV criteria) and be non-members
sive disorder (OCD) and mood disorders, especially of the AFTOC. The two samples were then com-
bipolar disorders, was already described by the pared in order to confirm prior data obtained in the
classical authors in French psychiatry during the XIX AFTOC sample, and to find if some bias existed by
century (Demonfaucon and Hantouche, 2001). De- the fact of belonging to a patient advocacy associa-
spite this historical antecedent, the recent literature tion. These analyses (AFTOC members versus non-
has been mainly focused on comorbidity with major members) showed comparable characteristics in both
depression giving less attention to bipolar comorbidi- populations (Hantouche et al., 2002b).
ty. Some epidemiological studies in the general A total number of 628 patients was obtained by
population (e.g., Chen and Dilsaver, 1995) demon- adding the two samples to conduct the present
strated significant association between OCD and analyses. Our aim was to delineate of a distinct
bipolar disorder, which is estimated to be at least as cyclothymic OCD type.
important as that with unipolar depressive comor-
bidity. Recently, in community subjects presenting 2.2. Clinical and psychometric assessment
hypomanic episodes (especially those with brief
recurrent episodes), Angst (1998) found elevated A questionnaire was prepared for self-assessment;
rates of OCD comorbidity. In two consecutive for patients under the age of 15 years (n 5 36), the
studies, Perugi et al. (1997, 2003) reported the parents were allowed to be informants. The ques-
significant presence of bipolarity in a cohort of tionnaire included 250 items related to socio-demo-
patients treated for OCD. The Pisa team showed graphic characteristics; the history of OCD (circum-
initially that the rate of bipolar comorbidity within stances of illness recognition, age of onset, clinical
OCD was 15.6%, mostly represented by BP-II aspects, course of illness, history and conditions of
disorder; in depressive OCD patients followed in a treatment). To explore mood comorbidity, we used
day-care unit, the rate was $ 50%. structured self-rated scales for major depression,
The foregoing studies led the French Association hypomania and mania (based on DSM-IV criteria) to
of patients suffering from OCD (AFTOC) to consi- generate categorical diagnoses. In addition, we added
der the question of bipolar comorbidity within OCD, Angst’s checklist of hypomania (Hantouche et al.,
taking into account the entire clinical spectrum of 2003), and the cyclothymic temperament ques-
mania, hypomania and cyclothymia. Earlier reports tionnaire (Akiskal et al., 2003) for dimensional
from the AFTOC survey confirmed the high rate of assessment of soft bipolarity. Cases obtaining a score
soft-bipolar comorbidity (Hantouche et al., 2002a,b). $ 10 on each questionnaire were classified as ‘‘hypo-
The present article will present the characteristics of manic’’ or ‘‘cyclothymic.’’ The threshold ( $ 10) for
cyclothymic OCD by a systematic comparison to both questionnaires had been previously validated by
non-cyclothymic OCD. Hantouche et al. (1998).
According to DSM-IV definitions of hypomania /
mania, 11% of the total sample were classified as
2. Methods bipolar (3% as BP-I and 8% as BP-II). When
dimensionally rated, 30% obtained a cut-off score
2.1. Creating the survey sample $ 10 on Hypomania checklist and 50% classified as
cyclothymic, ‘‘CYC-OCD’’ (score $ 10 on
The survey was first carried out among all mem- cyclothymic temperament questionnaire).
bers of AFTOC. At this writing, 780 subjects have
received by mail the survey questionnaire. To be 2.3. Statistical methods
included in the survey, subjects were to be recog-
nized and treated for OCD by a clinician. From this Statistical analyses were conducted by Syliastat. A
sample (n 5 780), 453 files (58%) were returned and mask for data management was developed with
retained for statistical analyses. In another sample GENMASK  software. The scientific committee
 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10 3

controlled all the files in order to check for illogisms. and psychiatric admissions in the cyclothymic group
Two independent operators of the project double (CYC-OCD).
seized the files in an interactive way. Statistical
analyses included comparative tests between two
3.2. OCD history and clinical picture
sub-populations: OCD with cyclothymic tempera-
ment versus those without cyclothymia. For com-
The past history of OCD is summarized in Table
parative analyses, the total number of files in which
2. History of alarming symptoms before the age of
data on hypomania and cylothymia questionnaires
15 was more common in CYC-OCD. Depression and
were obtained is 574 (91.4% of the responders). Two
impairment at work or school as the motive for
separate groups were created: 302 patients in the
seeking medical help were more prevalent in CYC-
cyclothymic OCD group and 272 in the non-
OCD. A progressive course of illness was more
cyclothymic group.
frequently reported in NC-OCD and precipitating
For categorical variables, a chi-square test (with
factors at the onset were more frequently reported in
Yates correction if needed) was used. The Student
the OCD-CYC group. In contrast to NC-OCD,
test for independent series or Mann–Whitney test (if
patients with CYC-OCD (Table 3) reported sig-
conditions for parametric tests were not satisfied)
nificantly higher rates of doubting, aggressive / im-
was used for continuous variables. All analyses were
pulsive, religious and sexual obsessions and of
two-tailed with a significance level set at P 5 0.05.
compulsions such as non-specific repetitive be-
Because of the high number of variables included in
haviors, hoarding, collecting and reassurance be-
the comparative tests, the significance level of P 5
havior (asking questions, need to tell or confess).
0.01 was also considered.

3.3. Coexisting neuro-psychiatric disorders and

family history
3. Results
Table 4 lists the rates of coexisting neuro-psychi-
3.1. Socio-demographic data atric disorders, which seemed in general of greater
relevance in the CYC-OCD group (with the excep-
Table 1 compares eight socio-demographic param- tion of tics, trichotillomania and onychophagia,
eters between the two groups. They differed only by which were comparable in the two groups). The
a higher divorce rate in the non-cyclothymic group family history showed higher rates of psychiatric
(NC-OCD) and elevated rates of suicide attempts problems (psychiatric admissions, presence of

Table 1
Socio-demographic characteristics
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Questionnaire filled by the patient (%) 88 79 0.015
Male (%) 43 45 ns
Age (years) 35 (612) 36 (614) ns
Divorce rate (%) 7 15 0.004
Alcohol consumption (%)
2–3 drinks / day 11 8
$ 4 drinks / day 4 2 ns
Suicide history (%)
Never thought about suicide 18 36
Attempts 20 12 , 0.0001
Psychiatric admissions (%) 46 35% 0.012
Number (mean) 3.0 3.6 ns
Age at first admission (years) 26 30 0.016
4 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10

Table 2
History of OCD
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Age of onset
Mean (years) 20.9 22.5 ns
Before 16 years (%) 38 31 ns
Alarming symptoms before age of 15 (mean, %) 2.7 1.8 , 0.0001
School difficulties 31 20 0.002
Conduct problems 26 18 0.019
Agitation / inattention 24 14 0.005
Somatic complaints 22 11 0.0007
‘‘Original character’’ / irritability 33 18 , 0.0001
Sleep disturbances 27 15 0.0006
Bizarre thinking / delusions / hallucinations 20 11 0.004
Panic crisis / phobias 72 63 0.017
Age at first seeking help (years) 24,6 26,2 ns
Motive for seeking medical help (%)
Persistence of symptoms 72 70 ns
Depression 53 36 0.0001
Complications at work / study 48 36 0.005
Number of doctors before correct diagnosis of OCD (%)
5 doctors or more 21 13 ns
Delay between OC onset and diagnosis
Mean (years) 12 11 ns
Delay . 10 years (%) 38 31 ns
Acute onset (%) 26 23 ns
Current OCD considered as (%)
First episode 2 4
Progressive slow disorder 24 38
Acute recurrent disorder with free interval 42 33
Acute recurrent disorder without free interval 33 25 0.002
Precipitating factors (%) 85 80 ns
Conjugal conflict 14 9 0.05
Death of loved person 25 14 0.002
Social difficulties 24 16 0.002
Trauma 25 13 0.0007

chronic symptoms, suicide attempts) in the CYC- CYC-OCD presented with greater number of hypo-
OCD group. manic symptoms (score 10 or more on the Angst’s
checklist). Moreover, the individual scoring on hypo-
3.4. Coexisting mood disorders manic items showed higher rate of all the items in
the CYC-OCD group.
The CYC-OCD group suffered more from co-
existing major depression (DSM-IV criteria), greater 3.5. Treatment history and medications
number of depressive symptoms within the episodes,
and higher rates of recurrence (Table 5). The analy- The CYC-OCD received anti-OCD medications
sis of chronological sequences showed no difference (serotonergic drugs) at an earlier age. This group
between the two groups. reported more frequently a global negative efficacy
Mania and hypomania according to DSM-IV were of medication. It is noteworthy that the CYC-OCD
present significantly more often in the CYC-OCD group reported more mood switching under anti-
group (Table 6). When dimensionally evaluated, the OCD medications associated with higher rates of
 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10 5

Table 3
Clinical picture of OCD
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Obsession inventory (%)
Contamination / fear of disease 46 46 ns
Order / symmetry 52 44 ns
Precision / perfectionism 61 57 ns
Doubting thoughts 74 63 0.004
Fear of catastrophes 34 29 ns
Somatic 20 15 ns
Superstitions / magic thoughts 33 26 ns
Aggressive / impulsive 32 20 0.002
Sexual thoughts 19 11 0.01
Religious 18 10 0.01
Parasites (neutral words, songs . . . ) 23 17 ns
Compulsion inventory (%)
Washing / cleaning 55 53 ns
Ordering / symmetry 47 42 ns
Checking 72 65 ns
Mental acts other than checking / counting 51 44 ns
Touching 36 29 ns
Repeating / in-out / up-down 65 54 0.009
Hoarding / fear of throwing 34 24 0.006
Collecting 15 7 0.001
Need to confess or to tell 18 11 0.03
Asking questions 36 27 0.03
Reassurance seeking 63 42 , 0.0001

aggressive behavior. Only 27% of the CYC-OCD 1909), Ballet responded to Seglas by insisting on the
group had been clinically recognized and treated as following:
bipolar (Table 7). More than half of the two groups
had received cognitive behavior therapy.
Obviously, obsessions are evolving in a parox-
ysmal mode. For intermittent obsessions I am
convinced that they are in almost all cases a part
4. Discussion
of periodic psychosis. However in psychasthenia,
the subject remains doubtful . . .
4.1. History

The classical authors in French psychiatry have Thereafter other authors such as Soukhanoff, Khan,
emphasized during the second half of the 19th Hartenberg and Bedel (Demonfaucon and Han-
century and the beginning of 20th century, the touche, 2001), considered obsessions as part of
special relationships between OCD termed ‘‘Folie de bipolar disorder, which was called at that time
´
Doute et Delire ´
du Toucher’’, ‘‘Psychasthenie’’ and ‘‘Cyclothymia’’ or ‘‘Periodic Psychosis with epi-
bipolar conditions termed ‘‘Folie a` Double Forme’’, sodes of mania and melancholia.’’ It is curious that
‘‘Folie Circulaire’’, ‘‘Cyclothymie’’, ‘‘Psychose with few exceptions (e.g., Chen and Dilsaver, 1995;
´
Periodique’’ (Demonfaucon and Hantouche, 2001). Perugi et al., 1998), contemporaneous research has
To the best of our knowledge, this question was first focused almost exclusively on comorbidity of OCD
considered by Ballet (1902), in a paper dedicated to with depression, and neglected the far-ranging as-
‘‘Intermittent melancholia.’’ During a famous debate sociation with the bipolar spectrum described long
in the French Society of Psychiatry (December 10, ago.
6 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10

Table 4
Associated neuropsychiatric disorders and family history
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Total comorbidity rate (%) 99.7 95.6 0.001
Motor tics 18 13 ns
Vocal tics 8 7 ns
Phobias 46 35 0.007
Anxiety—chronic worrying 78 60 , 0.0001
Anxiety—panic crisis 53 37 0.0002
Body dysmorphic concerns 31 17 0.0001
Excessive buying 26 14 0.001
Depression 75 54 , 0.0001
Sleep disturbances 50 36 0.0004
Eating disorders 38 21 , 0.0001
Pathologic gambling 1 1 ns
Trichotilllomania 12 9 ns
Onychophagia / body scratching 19 15 ns
Anger attacks 60 35 , 0.0001
Fava criteria 35 14 , 0.0001
Family history (%)
At least one psychiatric disorder 58 49 0.02
Psychiatric admissions 41 32 0.053
Chronic mental disorder 39 26 0.005
OCD 34 31 ns
Suicide attempts 26 18 0.05
Completed suicide 19 18 ns

Table 5
Major depression (MD) associated with OCD
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Major depression (DSM-IV criteria) (%) 89 64 , 0.0001
Number of depressive symptoms (mean) 7.2 5.3 , 0.0001
Rate of individual item scoring* (%)
Sadness 85 63
Loss of interest 88 69
Appetite disturbances 67 53
Sleep disturbances 88 71
Retardation or agitation 88 62
Fatigue 59 36
Guilt feelings 78 55
Concentration difficulties 92 73
Death or suicide thoughts 70 47
Subjects with more than 2 major episodes (%) 87 71
Chronological sequence (%)
MD before OCD 30 33
Simultaneous onset 46 35
MD after OCD 24 32 ns
* All comparisons are significant at P , 0.0001.
 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10 7

Table 6
Hypomanic (manic) episodes and symptoms associated with OCD
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
DSM-IV hypomania / mania (%) 18.8 7.4 0.0001
Hypomania 13.2 5.9
Mania 5.6 1.6
Hypomania check-list (Angst)
Global score (mean6S.D.) 8.9 (65.3) 4.2 (64.6) , 0.0001
Patients with score $ 10 (%) 49 17 , 0.0001
Rate of individual item scoring* (%)
Less sleep 50 27
More drive and energy 63 35
More self-confident 61 31
Increased work motivation 60 29
Increased social activity 61 31
More travelling, imprudent driving 29 10
Excessive shopping and spending 37 15
Foolish behaviors in business 12 3
Increased physical activity 43 18
More plans and ideas 58 33
Less shy, less inhibited 58 29
More talkative than usual 61 33
More irritable, impatient 47 23
Attention easily distractible 42 20
Increased sex drive and interest in sex 26 9
Increased consumption of coffee, cigarettes 24 10
Increased consumption of alcohol 15 5
Extremely happy mood, overeuphoric 34 11
More laughing 50 21
Faster thinking, more puns and jokes 55 23
* All comparisons are significant at P , 0.0001.

Table 7
Treatment history
CYC-OCD NC-OCD P
(n 5 302) (n 5 272)
Age at first anti-OCD treatment (years) 28,4 30,6 0.03
Delay from OCD onset more than 10 years (%) 40 36 ns
Efficacy of anti-OCD treatment (%)
Without any change / aggravation 26 19 0.04
Prior cognitive behavior therapy (%) 61 50 0.007
Prior diagnosis of bipolarity (%) 27 13 0.0001
Prior mood stabilizers treatment (%) 26 15 0.002
Mood switching under treatment (%) 48 24 , 0.0001
Severe switching 40 31 ns
Symptoms during mood switching (%) (n 5 126) (n 5 53)
Aggressive behavior 49 32 0.04
Bizarre behavior 24 23 ns
Delusions 17 11 ns
Hallucinations 6 4 ns
Severe insomnia 25 30 ns
Suicidal thoughts 27 26 ns
Homicide thoughts 5 4 ns
8 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10

4.2. Rate of concurrent bipolarity in OCD fluenced by concurrent cyclothymic traits such as
course of illness, which appeared more episodic
The AFTOC survey confirmed the high rate of (acute recurrent episodes with or without free inter-
bipolar comorbidity, especially when soft bipolarity vals) and the presence of precipitating factors like
was taken into account (hypomania and cyclothymia) conjugal conflicts, death of loved persons or social
in dimensional terms. This method of exploring problems. In this respect our data are reminiscent of
bipolarity could explain the high rate by comparison the Perugi et al. (1998) study.
to other work (e.g., Perugi et al., 1997), which relied The OCD clinical picture appears pathoplastically
on the more restrictive DSM-IV criteria used by influenced by cyclothymia. Globally all the OCD
clinicians. Recent research has validated the utility of symptoms are more significantly concentrated in
self-ratings in assessing hypomania and cyclothymic OCD-cyclothymic patients. Specifically, doubtful
temperament (Hantouche et al., 1998; Allilaire et al., thoughts, sexual, religious and aggressive obsessions,
2001; Angst and Hantouche, 2002; Akiskal et al., and repeating behaviors, hoarding and reassurance
2003) and even in acute mania (Altman, 1998; compulsive behaviors are more frequent in this
Hantouche et al., 2001; Akiskal et al., 2001). group. These findings could have clinical relevance.
On the basis of DSM-IV criteria of mania or When the total number of symptoms is high or
hypomania, our data are similar to the Pisa study, in obsessions are dominating (with themes related to
which a rate of 15,6% for mania and hypomania was sex, religion, aggression, doubt) or reassurance be-
found (Perugi et al., 1997). However the present data havior is omnipresent (need to confess or ask ques-
are best compared with the recent data from the tions), the OCD clinical picture could be more
Zurich survey (Angst, 1998), which showed a special striking. This should alert the clinician to suspect the
association between OCD and recurrent brief hypo- presence of soft bipolar comorbidity.
mania. We submit that OCD most frequently emerges More cues are also presumptive of soft bipolarity
from rapid, brief and recurrent mood swings, which such as suicide attempt, episodic course, recurrent
represent a separate sub-type of soft bipolarity and severe depression, and greater concurrent dis-
(Akiskal and Pinto, 1999). orders (panic, phobias, anger attacks, impulsive
disorders, eating disorders) and / or family history for
4.3. Cyclothymic OCD? suicide attempts, chronic symptoms or psychiatric
admissions. Some of these presumptive cues were
In contrast to their non-cyclothymic counterparts, suggested by the Pisa study in which concurrent
CYC-OCD patients suffered from more severe disor- substance abuse and panic attacks appeared as
der as suggested by greater number of psychiatric indicators of bipolar OCD (Perugi et al., 1998,
admissions (with earlier age at first admission), 1999).
elevated risk of suicide and greater number and Even if our data are insufficient to isolate
intensity of concurrent mania or hypomania, and ‘‘cyclothymic OCD’’ as a distinct form, there are
major depressive episodes. Furthermore CYC-OCD many important facets, which should be considered:
group had more behavioral problems, and disturbing Firstly, the suicide risk (as suggested by a prior
symptoms before the age of 15: It would appear that report by Lester and Abdel-Khalek, 1999) and the
the cyclothymic temperament can lead at an earlier elevated recurrence rate of severe depression; sec-
age to alarming emotional problems, school difficul- ondly, the less favorable response of OCD to antide-
ties, as well as to sleep and cognitive disturbances. pressant therapy associated with higher rate of mood
However, in some aspects the two groups did not switching (48%), and ‘‘paradoxical’’ worsening
differ: mean age at onset of OCD, the rate of early under drug therapy (emerging aggressive behaviors,
onset cases before the age of 15, the age at onset of suicidal thoughts, psychotic features, severe insom-
full symptomatic OCD, the age at first medical help- nia). These phenomena could look like ‘‘refractory
seeking for OCD, and the delay to OCD recognition OCD’’ in which experts might recommend the use of
or the number of consulted doctors before the correct neuroleptics. In the Pisa study (Perugi et al., 2003),
diagnosis of OCD. Other aspects of OCD are in- previous administration of clomipramine resulted in
 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10 9

induced (hypo)mania in 39% of the cases. Other ment in psychiatric services or those from the rank of
reports, too, indicate the emergence of antidepres- membership in patient advocacy organizations). The
sant-induced mania in OCD (Berk et al., 1996). French survey has the merit to extend the comorbid
Moreover, despite combination of different mood bipolarity to a systematically evaluated cyclothymic
stabilizers (lithium plus antiepileptics), many bipolar temperament. The research from these three sources
OCD patients present significant residual affective supports the pioneering contributions of now forgot-
and OCD symptomatology. Emerging data from both ten ‘‘classical’’ French psychiatrists who emphasized
clinical and systematic trials suggest that atypical the major role of the ‘‘emotional constitution’’ in
antipsychotics might be clinically useful in such understanding both OCD and periodic affective
patients in terms of avoiding depression and protect- psychosis (Demonfaucon and Hantouche, 2001). We
ing against switching, though most of the papers in submit that ‘‘cyclothymic OCD’’ represents a distinct
the literature deal with OCD refractory to form of OCD in terms of different clinical picture,
serotoninergic antidepressants (Jacobsen, 1995; course, and outcome of OCD, and of higher recur-
Ravizza et al., 1996; Saxena et al., 1996; Stein et al., rence of depression and suicide risk. Our findings are
1997; Fitzgerald et al., 1999; Weiss et al., 1999; of major clinical and public health significance, and
Koran et al., 2000; McDougle et al., 2000; Pfanner et should be replicated by other research teams.
al., 2000; Francobandiera, 2001; Khullar et al., 2001;
De et al., 2002; Mohr et al., 2002). The foregoing
literature is cumulatively and clinically persuasive,
though except for one double blind study (McDougle References
et al., 2000), is based on systematic open case series.
The clinician must note that sporadic worsening of Akiskal, H.S., Pinto, O., 1999. The evolving bipolar spectrum:
OCD with atypical antipsychotics use has also been prototypes I, II, III and IV. PCNA 22, 517–534.
Akiskal, H.S., Hantouche, E.G., Bourgeois, M.L., Azorin, J.M.,
reported in the literature cited above, though it is Sechter, D., Allilaire, J.F., Chatenet-Duchene, L., Lancrenon,
impossible to tell in the absence of a placebo arm S., 2001. Toward a refined phenomenology of DSM-IV mania:
whether the worsening was due to natural course of Combining clinician-assessment and self-report in the French
severe comorbid OCD. EPIMAN study. J. Affect. Disord. 67, 89–96.
The lessons from the AFTOC survey would be to Akiskal, H.S., Hantouche, E.G., Lancrenon, S., 2003. BP-II with
and without cyclothymic temperament: ‘‘dark’’ and ‘‘sunny’’
avoid unnecessary complications by an early recog- expressions. J. Affect. Disord. 73, 49–57.
nition of complex cyclothymic OCD, and to protect Allilaire, J.F., Hantouche, E.G., Sechter, D., Bourgeois, M.L.,
the patient mood-stabilizers, possibly even before ˆ
Azorin, J.M., Lancrenon, S., Chatenet-Duchene, ˆ L., Akiskal,
exposure to serotonergic agents. In the experience of ´
H.S., 2001. Frequence et aspects cliniques du trouble BP-II
the authors, in some cases, mainly very young OCD ´
dans une etude multicentrique française: EPIDEP. Encephale
27, 149–158.
patients, OCD was cleared by mood-stabilizer mono-
Altman, E., 1998. Rating scales for mania: is self-rating reliable?
therapy. In brief, we believe, on clinical grounds, J. Affect. Disord. 50, 283–286.
that ‘‘cyclothymic OCD’’ should be diagnosed at an Angst, J., 1998. The emerging epidemiology of hypomania and
early stage, to bring the benefit of mood stabilizers bipolar II disorder. J. Affect. Disord. 50, 143–151.
and / or atypical antipsychotics. Angst, J., Hantouche, E.G., 2002. The epidemiology of minor
bipolar disorder and hypomania: new territory. In: Vieta, E.
(Ed.), Hypomania, Aula Medica ´ Bibliotecas, Madrid, pp. 13–
31.
5. Conclusion ´
Ballet, G., 1902. La melancolie intermittente. Presse Med. 39,
459–462.
Publications from the Pisa Psychiatric Institute, Berk, M., Koopowitz, L.E., Szabo, C.P., 1996. Antidepressant-
the Zurich epidemiologic project and the French induced mania in OCD. Eur. Neuropsychopharmacol. 6, 9–11.
Chen, Y.W., Dilsaver, S.C., 1995. Comorbidity for OCD in bipolar
AFTOC survey converge in delineating a complex
and unipolar disorders. Psychiatry Res. 59, 57–64.
form of OCD with concurrent soft bipolarity. This De, H.L., Beuk, N., Hoogenboom, B., Dingemans, P., Linszen, D.,
condition seems to affect a substantial number of 2002. Obsessive-compulsive symptoms during treatment with
OCD patients (almost 50% of those seeking treat- olanzapine and risperidone: a prospective study of 113 patients
10 E.G. Hantouche et al. / Journal of Affective Disorders 75 (2003) 1–10

with recent-onset schizophrenia or related disorders. J. Clin. McDougle, C.J., Epperson, C.N., Pelton, G.H., Wasylink, S., Price,
Psychiatry 63, 104–107. L.H., 2000. A double-blind, placebo-controlled study of ris-
Demonfaucon, C., Hantouche, E.G., 2001. Approche dimension- peridone addition in serotonin reuptake inhibitor-refractory
nelle des rapports entre TOC et bipolarite: ´ interactions entre obsessive-compulsive disorder. Arch. Gen. Psychiatry 57, 794–
´ ´ impulsivite,
emotivite, ´ lenteur et place centrale de la symetrie.
´ 801.
Synapse 180, 31–40. Mohr, N., Vythilingum, B., Emsley, R.A., Stein, D.J., 2002.
Fitzgerald, K.D., Stewart, C.M., Tawile, V., Rosenberg, D.R., Quetiapine augmentation of serotonin reuptake inhibitors in
1999. Risperidone augmentation of serotonin reuptake in- obsessive-compulsive disorder. Int. Clin. Psychopharmacol. 17,
hibitors in obsessive-compulsive and related disorders. J. Child. 37–40.
Adolesc. Psychopharmacol. 9, 115–123. Perugi, G., Akiskal, H.S., Pfanner, C., Presta, S., Gemignani, A.,
Francobandiera, G., 2001. Olanzapine augmentation of serotonin Milanfranchi, A., Lensi, P., Ravagli, S., Cassano, G.B., 1997.
uptake inhibitors in obsessive-compulsive disorder: an open The clinical impact of bipolar and unipolar affective comor-
study. Can. J. Psychiatry 46, 356–358. bidity on obsessive-compulsive disorder. J. Affect. Disord. 46,
Hantouche, E.G., Akiskal, H.S., Lancrenon S Allilaire, J.F., 15–23.
Sechter, D., Azorin, J.M., Bourgeois, M., Fraud, J.P., Chatenet- Perugi, G., Akiskal, H.S., Gemignani, A., Pfanner, C., Presta, S.,
Duchene, L., 1998. Systematic clinical methodology for val- Milanfranchi, A., Lensi, P., Ravagli, S., Maremmani, I.,
idating BP-II disorder: data in mid-stream from a French Cassano, G.B., 1998. Episodic course in obsessive-compulsive
national study (EPIDEP). J. Affect. Disord. 50, 163–175. disorder. Eur. Arch. Psychiatr. Clin. Neurosci. 248, 240–244.
Hantouche, E.G., Akiskal, H.S., Bourgeois, M.L., Azorin, J.M., Perugi, G., Akiskal, H.S., Ramacciotti, S., Nassini, S., Toni, C.,
Sechter, D., Allilaire, J.F., Chatenet-Duchene, L., Lancrenon, Milanfranchi, A., Musetti, L., 1999. Depressive comorbidity of
S., 2001. The feasibility of self-assessment of dysphoric mania panic, social phobia, and OCD re-examined: is there a bipolar
in the French national EPIMAN study. J. Affect. Disord. 67, II connection? J. Psychiatry Res. 33, 53–61.
97–103. Perugi, G., Toni, C., Frare, F., Travierso, C., Hantouche, E.,
Hantouche, E.G., Akiskal, H.S., Demonfaucon, C., Barrot, I., Akiskal, H.S., 2003. Obsessive-compulsive-bipolar comorbidi-
Kochman, F., Millet, B., Lancrenon, S., Allilaire, J.F., 2002a. ty: a systematic exploration of clinical features and treatment
Bipolarite´ cachee
´ dans le trouble obsessionnel compulsif: outcomes. J. Clin. Psychiatry, in press.
ˆ collaboratıve
enquete ˆ avec l’AFTOC. Ann. Med. Psychol. 160, Pfanner, C., Marazziti, D., Dell’Osso, L., Presta, S., Gemignani,
34–41. A., Milanfranchi, A., Cassano, G.B., 2000. Risperidone aug-
Hantouche, E.G., Kochman, F., Demonfaucon, C., Barrot, I., mentation in refractory obsessive-compulsive disorder: an
Millet, B., Lancrenon, S., Akiskal, H.S., 2002b. TOC bipolaire: open-label study. Int. Clin. Psychopharmacol. 15, 297–301.
´
confirmation des resultats ˆ «ABC-TOC» dans deux
de l’enquete Ravizza, L., Barzega, G., Bellino, S., Bogetto, F., Maina, G.,
´
populations de patients adherents ´
versus non adherents a` une 1996. Therapeutic effect and safety of adjunctive risperidone in
association. Encephale 28, 21–28. refractory obsessive-compulsive disorder (OCD). Psycho-
Hantouche, E.G., Angst, J., Akiskal, H.S., 2003. Factor structure pharmacol. Bull. 32, 677–682.
of hypomania: Interrelationships with cyclothymia and soft Saxena, S., Wang, D., Bystritsky, A., Baxter, Jr. L.R., 1996.
bipolar spectrum. J. Affect. Disord. 73, 39–47. Risperidone augmentation of serotonin reuptake inhibitor treat-
Jacobsen, F.M., 1995. Risperidone in the treatment of affective ment of pediatric obsessive compulsive disorder. J. Clin.
illness and obsessive-compulsive disorder. J. Clin. Psychiatry Psychiatry 57, 303–306.
56, 423–429.
Stein, D.J., Bouwer, C., Hawkridge, S., Emsley, R.A., 1997.
Khullar, A., Chue, P., Tibbo, P., 2001. Quetiapine and obsessive-
Risperidone augmentation of SRI treatment for refractory
compulsive symptoms (OCS): case report and review of
obsessive-compulsive disorder. J. Clin. Psychiatry 58, 119–
atypical antipsychotic-induced OCS. J. Psychiatry Neurosci.
122.
26, 55–59.
Weiss, E.L., Potenza, M.N., McDougle, C.J., Epperson, C.N.,
Koran, L.M., Ringold, A.L., Elliott, M.A., 2000. Olanzapine
1999. Olanzapine addition in obsessive-compulsive disorder
augmentation for treatment-resistant obsessive-compulsive dis-
refractory to selective serotonin reuptake inhibitors: an open-
order. J. Clin. Psychiatry 61, 514–517.
label case series. J. Clin. Psychiatry 60, 524–527.
Lester, D., Abdel-Khalek, A.M., 1999. Manic-depression,
suicidality, and obsessive-compulsive tendencies. Psychol.
Rep. 85, 1100.