Diagnostic validity of comorbid bipolar disorder and obsessive–compulsive disorder: a systematic review

Authors

  • A. Amerio,

    Corresponding author
    1. Section of Psychiatry, Department of Neuroscience, University of Parma, Parma, Italy
    2. Mood Disorders Program, Tufts Medical Center, Boston, MA, USA
    • Andrea Amerio, Section of Psychiatry, Department of Neuroscience, University of Parma, c/o Ospedale Maggiore, Pad. 21 - Braga, Viale A. Gramsci 14, 43126 Parma, Italy.

      E-mail: andrea.amerio@studenti.unipr.it

    Search for more papers by this author
  • A. Odone,

    1. School of Medicine-Public Health Unit, University of Parma, Parma, Italy
    2. Department of Global Health and Social Medicine, Harvard Medical School, Boston, MA, USA
    Search for more papers by this author
  • C. C. Liapis,

    1. Mood Disorders Program, Tufts Medical Center, Boston, MA, USA
    2. First Department of Psychiatry, Eginition Hospital, Athens University Medical School, Athens, Greece
    Search for more papers by this author
  • S. N. Ghaemi

    1. Departments of Psychiatry and Pharmacology, Mood Disorders Program, Tufts Medical Center, Tufts University Medical School, Boston, MA, USA
    Search for more papers by this author

Abstract

Objective

At least 50% of bipolar disorder (BD) patients have an additional diagnosis, one of the most difficult to manage being obsessive–compulsive disorder (OCD). Defining the nosology of BD-OCD comorbidity has important clinical implications, given that treatments for OCD can worsen BD outcomes.

Method

A systematic review was conducted on: i) BD-OCD comorbidity lifetime prevalence and ii) on standard diagnostic validators: phenomenology, course of illness, heredity, biological markers, and treatment response. Relevant papers published through March 30th 2013 were identified searching the electronic databases MEDLINE, Embase, PsycINFO, and the Cochrane Library.

Results

Sixty-four articles met inclusion criteria. Lifetime comorbidity prevalence was 11–21% in BD patients and 6–10% in OCD patients. Compared to non-comorbid subjects, BD-OCD has a more episodic course of OC symptoms (up to 75% vs. 3%), typically with worsening during depression (78%) and improvement during mania/hypomania (64%), as well as a higher total mean number of depressive episodes (8.9 ± 4.2 vs. 4.1 ± 2.7) and perhaps more antidepressant-induced mania/hypomania (39% vs. 9%).

Conclusion

In this first systematic review of BD-OCD comorbidity, it appears that OC symptoms are usually secondary to BD, rather than representing a separate disease.

Summations

  • Bipolar disorder–obsessive–compulsive disorder (BD-OCD) comorbidity's burden and clinical management is a critical issue in psychiatry.
  • The evidence so far on BD-OCD nosology—mainly based on the course of illness—supports the view that the majority of cases of comorbid BD-OCD are in fact cases of BD. OC symptoms usually are secondary manifestations of depressive or manic mood episodes.
  • BD-OCD patients may respond for both groups of symptoms with adequate doses of mood stabilizers and atypical antipsychotics. Because of the risk of worsening BD via SRI-induced mania/hypomania, antidepressants should only be used in a minority of refractory OCD.

Considerations

  • Course of illness is the only diagnostic validator that clearly supports the OCD-BD construct as a subtype of BD. Findings on the phenomenology, heredity, biological markers, and treatment were fragmented and heterogeneous and mainly suggested increased severity of illness or functioning without helping us to answer the nosological question.
  • The use of retrospective assessment scales with unclear sensitivity in discriminating true ego-dystonic obsessions from depressive ruminations may have biased results toward an overestimation of obsession's prevalence. In addition, small sample sizes, mainly from out-patients units, may have limited the generalizability of the results to a whole community.
  • Further original studies are needed to clarify BD-OCD diagnostic validity and treatment approach.

Introduction

Psychiatric comorbidity is extremely common in bipolar disorder (BD). Nearly 50% of patients with BD have an additional diagnosis [1], one of the most difficult to manage being obsessive–compulsive disorder (OCD). In a classic 1860 text, French psychiatrist Bénédict-Augustin Morel described patients with comorbidity of symptoms typical of what are now considered BD and OCD [2]. Current large epidemiological data indicate that this comorbidity occurs in 21% of BD patients, substantially more than comorbidity rates for OCD in unipolar depression (12.2%) and in the general population (2.5%) [3].

What remains unclear with whether this high comorbidity represents the frequent occurrence of two independent diseases or whether it represents the occurrence of symptoms of one kind (e.g., OCD symptoms) in a different disease (i.e., BD). This question is highlighted in a standard 1969 psychiatry textbook by Mayer-Gross and colleagues: ‘Compulsive symptoms are quite common in manic-depressive psychoses, especially in depressions. … From time to time one sees the rare cases of recurrent endogenous obsessional neuroses. These people who, in time of health, show no noteworthy obsessional traits, but who have phases in which compulsive symptoms appear out of the blue and rapidly mount up to complete incapacitation. … Nevertheless these illnesses remit and relapse in very much the same way as cyclothymic illnesses, may show just as much regularity of timing, and are probably to be included, from the etiological point of view, in the manic-depressive disorders' [4].

Recent reviews have not systematically assessed the relative independence, or not, of OCD symptoms in subjects with BD, which has important nosological and clinical implications. The nosological question is whether this common ‘comorbidity’ represents two diseases or multiple symptoms of one disease. Thus, the comorbidity may represent a severe subtype of BD, or a severe subtype of OCD, or two separate disease entities that co-occur by chance. The clinical question is whether, and how, to treat the comorbidity because the main treatment for one disease (serotonin reuptake inhibitors, SRIs, for OCD) [5, 6] can worsen the other (antidepressants, like SRIs, can cause mania and/or more mood episodes in BD) [7-11].

This is the first systematic review of the diagnostic and clinical meaning of the BD-OCD comorbidity.

Aims of the study

We systematically review the evidence from the literature to define the nosological validity of bipolar disorder–obsessive–compulsive disorder (BD-OCD) comorbidity and to provide recommendations for clinical management and future research.

Material and methods

We conducted this review according to the methods recommended by the Cochrane Collaboration and documented the process and results in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines [12, 13].

Information sources and search strategy

Studies were identified searching the electronic databases MEDLINE, Embase, PsycINFO, and the Cochrane Library. We combined the search strategy of free text terms and exploded MESH headings for the topics of bipolar disorder and obsessive–compulsive disorder combined as following: (((((((Bipolar disorder) OR BD) OR Bipolar) OR Manic-depressive disorder) OR Manic-depressive) OR Manic)) AND (((Obsessive–compulsive disorder) OR Obsessive–compulsive) OR OCD). The strategy was first developed in MEDLINE and then adapted for use in the other databases (Appendix). Studies published in English through March 30, 2013 were included. In addition, further studies were retrieved from reference listing of relevant articles and consultation with experts in the field.

Inclusion criteria

Study population and study design

We considered studies that included BD and OCD subjects if diagnostic criteria used were specified. Among BD study populations, studies that only focused on BD-I, BD-II, or BD not otherwise specified [14] were included. Studies that considered subjects with bipolar and obsessive–compulsive spectrums were also included if diagnostic criteria used were specified [15-17]. Participants of both sexes older than 6 years of age were considered. Studies conducted on subjects with physical comorbidities such as lupus erythematous or Asperger's disorder were excluded as non-representative of the study population [18, 19].

Both population-based and hospital-based studies were included. Among hospital-based studies, in-patients, day-hospital, and out-patient subjects were included, while emergency care records were excluded as considered non-representative. All experimental and observational study designs were included apart from case reports. Narrative and systematic reviews, letters to the editor, and book chapters were excluded.

Outcome measures

Primary outcomes were comorbidity prevalence rates and specific validating criteria for BD-OCD clinical construct. With regard to comorbidity prevalence rates, we retrieved data on: i) lifetime prevalence of comorbid OCD in BD patients and ii) lifetime prevalence of comorbid BD in OCD patients. Studies that reported only data about current prevalence were excluded [20]. With regard to validating criteria for BD-OCD clinical construct, we used the approach described by Robins and Guze [21] and retrieved data on: i) phenomenology, ii) course of illness, iii) heredity, iv) biological markers, and v) treatment response.

Study selection and data extraction

Identified studies were independently reviewed for eligibility by two authors (AA, CCL) in a two-step based process; a first screening was performed based on title and abstract, while full texts were retrieved for the second screening. At both stages, disagreements by reviewers were resolved by consensus. Data were extracted by one author (AA) and supervised by a second author (CCL) using an ad-hoc developed data extraction spreadsheet. The data extraction spreadsheet was piloted on 10 randomly selected papers and modified accordingly.

Quality assessment

The same authors who performed data extraction (AA, CCL) independently assessed the quality of selected studies using the checklist developed by Downs and Black both for randomized and non-randomized studies [22]. Disagreements by reviewers were resolved by consensus. Table 1 shows the quality assessment total score assigned to each study.

Table 1. Studies that met inclusion/exclusion criteria for systematic review
ReferencesStudy designCountryStudy populationSample sizeDiagnosis assessmentOutcomesQualitya
  1. BD, bipolar disorder; BD-I, bipolar disorder type I; BD-II, bipolar disorder type II; OCD, obsessive–compulsive disorder; MDE, major depressive episode; CYC, cyclothymic; PR, prevalence; PH, phenomenology; CI, course of illness; HE, heredity; BM, biological markers; TR, treatment; NS, not specified; Pt., patients; DSM, Diagnostic and Statistical Manual of Mental Disorders; SCID, Structured Clinical Interview; SCID-P, Structured Clinical Interview Patient Version; SCID-CV, Structured Clinical Interview Clinical Version; CIDI, WHO Composite International Diagnostic Interview; DIS, Diagnostic Interview Schedule; MINI, Mini International Neuropsychiatric Interview; DIGS, Diagnostic Instrument for Genetic Studies; DICA-R, Diagnostic Interview for Children and Adolescents-Revised; SADS-L, Schedule for Affective Disorders and Schizophrenia, Lifetime Version; K-SADS-E, Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children-Epidemiologic Version; K-SADS-PL, Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version; WASH-U-K-SADS, Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia; DIS, Diagnostic Interview Schedule; DIA-X/M-CIDI, Munich Composite International Diagnostic Interview.

  2. a

    Checklist for measuring study quality developed by Downs and Black.

  3. b

    They refer to the same study but were included separately as one focused on prevalence of BD-OCD comorbidity and the other focused more in details on BD-OCD family history, association with substance abuse and consequences of comorbidity.

  4. c

    They refer to the same Pt. sample.

Population-based studies
Adam et al. 2012 [47]Cross-sectional studyGermany4181 subjects (age range = 18–65): OCD (n = 30)30Broad concept of OCD; DIA-X/M-CIDI; DSM-IVPR25/31
Angst et al. 2004b [15]Prospective cohort studySwitzerland591 subjects recruited at age 19/20 and assessed over 20 years: OCD (n = 30)30Broad definition for BD and OCD; DSM-IVPR26/31
Angst et al. 2005b [16]Prospective cohort studySwitzerland591 subjects recruited at age 19/20 and assessed over 20 years: OCD (n = 30), BD (n = 93)123Broad definition for BD and OCD; DSM-IVHE, CI26/31
Chen et al. 1995 [3]Cross-sectional studyUSAPt. with BD, unipolar disorder or any Axis I disorder other than bipolar or unipolar disorder (n = 6622, age > 18): BD (n = 167)167DIS; DSM-IIIPR, CI25/31
Faravelli et al. 2004 [23]Cross-sectional studyItaly2500 subjects randomly selected from the lists of 15 GPs: BD (n = 19, age > 14), OCD (n = 57, age > 14)76SCID; DSM-IVPR24/31
Fireman et al. 2001 [46]Cross-sectional studyUSA1.728.480 subjects (age > 6): adults with OCD (n = 1078), children and adolescents with OCD (198)1078/198NS; DSM-IVPR24/31
Merikangas et al. 2007 [24]Cross-sectional studyUSA9282 subjects recruited (age ≥ 18): adults with any BD (n = 408)408CIDI; SCID; DSM-IVPR25/31
Hospital-based studies: adults
Bogetto et al. 1999 [88]Case–control studyItalyOCD (n = 160, mean age: males 32.1 ± 13.0, females 36.9 ± 11.4)160NS; DSM-IVPR21/31
Boylan et al. 2004 [28]Prospective cohort studyCanadaBD (n = 138, age range = 16–65)138SCID; DSM-IVPR22/31
Cannon et al. 2006 [77]Case–control studyUSABD (n = 18, mean age = 30 ± 9)18NS; DSM-IVBM23/31
Cassano et al. 1999 [89]Cross-sectional studyItalyPt. with psychotic symptoms consecutively hospitalized (n = 77, mean age = 33.5 ± 10.3): BD-I (n = 48), schizoaffective disorder, bipolar type (n = 11), unipolar depression (n = 18)48SCID-P; DSM-III-RPR23/31
Centorrino et al. 2006 [76]Case–control studyUSAAdults (n = 62) with BD, OCD, or BD-OCD62NS; DSM-IVCI20/31
Cosoff et al. 1998 [34]Case–control studyAustraliaSubjects with a psychotic disorder (n = 100, mean age: men = 34.8 ± 10.0, women = 34.9 ± 9.6): BD (n = 20)20SCID-P; DSM-III-RPR20/31
Craig et al. 2002 [25]Cross-sectional studyUSA450 subjects: BD with psychosis (n = 138, mean age range = 15–60)138SCID; DSM-III-RPR22/31
Dell'Osso et al. 2000 [67]Case–control studyItalyBD with psychotic features (n = 125, mean age: 33.3 ± 11.1)125SCID-P; DSM-III-RPH21/31
Dilsaver et al. 2008 [90]Case–control studyUSA187 Latino pt. enrolled consecutively from 2001 to 2003: BD-I (n = 69, mean age = 34.9 ± 11.8)69SCID-CV; DSM-IVPR20/31
Diniz et al. 2004 [56]Cross-sectional studyBrazilOCD (n = 161, mean age = 30 ± 10))161SCID; DSM-IVPR21/31
Edmonds et al. 1998 [33]Case–control studyNew ZealandBD (n = 55,mean age = 41.6), first-degree relatives (n = 67, mean age = 50.3)122DIGS; DSM-IVPR20/31
Goes et al. 2012 [75]Cross-sectional studyUSABD (n = 1416, mean age = 42.0), first-degree relatives with BD (n = 850)1416/850DIGS; DSM-IVCI23/31
Hantouche et al. 2003 [17]Case–control studyFranceOCD (n = 628, mean age CYC-OCD = 35 ± 12,mean age NC-OCD = 36 ± 14)628NS; DSM-IVPR24/31
Hasler et al. 2005 [69]Cross-sectional studyUSAOCD (n = 317, age > 18)317SCID; DSM-IVPH23/31
Henry et al. 2003 [26]Cross-sectional studyFranceBD (n = 318, mean age = 53.3 ± 15.1)318DIGS; DSM-IVPR23/31
Issler et al. 2005 [70]Case–control studyBrazilOCD-BD (n = 15, mean age = 38.9 ± 10.7)15SCID-P; DSM-IVPH, CI14/31
Issler et al. 2010 [74]Case–control studyBrazilBD (n = 30, mean age: BD = 41.8 ± 10.5, OCD-BD = 38.9 ± 10.7)30SCID-P; DSM-IVCI17/31
Koyuncu et al. 2010 [30]Case–control studyTurkeyBD (n = 214, mean age: BD = 34.8 ± 10.3, OCD-BD = 36.2 ± 15.9)214SCID; DSM-IVPR, HE, CI20/31
Kruger et al. 1995 [36]Cross-sectional studyGermanyMajor affective disorder (n = 149, mean age = 49 ± 12): BD (n = 44)37DIS; DSM-IIIPR21/31
Kruger et al. 2000 [37]Case–control studyGermanyBD-I or BD-II (n = 143, mean age = 44)143SCID; DSM-III-RPR, CI22/31
LaSalle-Ricci et al. 2006 [72]Cross-sectional studyUSAOCD (n = 204, age > 18)204SCID-P for DSM-IVPH20/31
Lensi et al. 1996 [58]Case–control studyItalyOCD (n = 263, mean age = 33.1)263NS; DSM-III-RPR23/31
Magalhaes et al. 2010 [31]Case–control studyBrazilBD (n = 259, mean age = 41)259SCID; DSM-IVPR, CI23/31
Mahasuar et al. 2011 [65]Case–control studyIndiaOCD (n = 91, mean age: OCD = 29.36 ± 8.31, BD-OCD = 28.39 ± 7.10)91SCID for DSM-IVPH, HE, CI19/31
Maina et al. 2007 [49]Case–control studyItalyOCD (n = 204, mean age = 34.7 ± 12.1)204SCID; DSM-IVPR, PH, CI21/31
Marazziti et al. 2002 [50]Cross-sectional studyItalyOCD (n = 117, mean age = 30 ± 9.3)117SCID-P; DSM-IVPR, PH21/31
McElroy et al. 1995 [66]Case–control studyUSABD (n = 71, mean age = 35 ± 16)71SCID-P; DSM-III-RPH23/31
McElroy et al. 2001 [38]Case–control studyUSABD-I or BD-II (n = 288, mean age = 42.8 ± 11.3)288SCID-P; DSM-IVPR23/31
Ortiz et al. 2011 [91]Cross-sectional studyCanadaBD (n = 379, mean age = 25.1 ± 10.6)379SADS-L; DSM-IVCI23/31
Perugi et al. 1997 [53]Case–control studyItalyOCD (n = 315, mean age: BD-OCD = 32.8 ± 12.2, OCD = 32.5 ± 12.6)345NS; DSM-III-RPR, PH, CI22/31
Perugi et al. 1998 [52]Case–control studyItalyOCD (n = 135, mean age = 38.4 ± 13.3)135NS; DSM-III-RPR, HE, CI21/31
Perugi et al. 1999 [54]Case–control studyItaly269 pt. enrolled consecutively from 1993 to 1995: OCD (n = 79, mean age = 30.4 ± 11.8)79SCID-Up-R; DSM-III-RPR, TR20/31
Perugi et al. 2002 [55]Case–control studyItalyOCD-MDE (n = 68, mean age = 34.2 ± 12.5); BD-OCD (n = 38, mean age = 35.9 ± 12.2)68SCID; DSM-IVPR, PH, CI, HE, TR20/31
Pini et al. 1997 [40]Case–control studyItalyCurrent episode of depression (n = 87): bipolar depression (n = 24, mean age = 37.9 ± 12.0), unipolar depression (n = 38, mean age = 47.0 ± 15.0), dysthymia (n = 25, mean age = 43.0 ± 12.0)24SCID-P; DSM-III-RPR20/31
Pini et al. 1999 [92]Cross-sectional studyItalyBD (n = 125, age > 16)125SCID-P; DSM-III-RPR21/31
Saunders et al. 2012 [27]Cross-sectional studyUSABD, type I or Schizoaffective Disorder, bipolar type (n = 736, mean age = 42 ± 12)736DIGS; DSM-IVPR19/31
Shabani et al. 2008 [68]Case–control studyIranOCD 78: BD-OCD (n = 39, mean age = 26.6 ± 7.23), OCD (n = 39, mean age = 30.1 ± 6.52)78SCID-CV; DSM-IVPH20/31
Simon et al. 2003 [32]Case–control studyUSA236 subjects: BD (n = 122, mean age = 40.8 ± 12.2)122SCID; DSM-IVPR23/31
Simon et al. 2004 [39]Cross-sectional studyUSABD (n = 475, mean age = 41.7 ± 12.8)475MINI; DSM-IVPR, CI23/31
Strakowski et al. 1992 [35]Cross-sectional studyUSABD (n = 41, mean age = 40.4 ± 11.7)41SCID; DSM-III-RPR22/31
Strakowski et al. 1998 [29]Cross-sectional studyUSABD, manic or mixed with psychosis (n = 77, mean age = 25 ± 6)77SCID-P; DSM-III-RPR, CI22/31
Tamam et al. 2002 [42]Cross-sectional studyTurkeyBD-I in remission (n = 70, mean age = 33.4 ± 10.3)70SCID-CV; DSM-IVPR20/31
Timpano et al. 2012 [51]Case–control studyUSAOCD (n = 605, mean age = 39.2)605SCID-P; DSM-IVPR, PH, CI20/31
Tukel et al. 2006 [71]Case–control studyTurkeyOCD (n = 115, age > 18)117SCID-CV for DSM-IVPH, CI21/31
Tukel et al. 2007 [57]Case–control studyTurkeyOCD (n = 128, mean age = 29.3 ± 10.8)128SCID-CV; DSM-IVPR, CI21/31
Zutshi et al. 2006 [41]Case–control studyIndiaBD in remission (n = 80, mean age = 30.06 ± 7.77)80SCID-CV; DSM-IVPR22/31
Zutshi et al. 2007 [64]Case–control studyIndiaOCD (n = 106, mean age: BD-OCD = 27.93 ± 6.71, OCD = 26.47 ± 7.38)106SCID-CV; DSM-IVPH, HE, CI20/31
Hospital-based studies: children, adolescents
Dilsaver et al. 2006 [45]Case–control studyUSALatino adolescents (n = 313): BD (n = 115, mean age = 14.6 ± 1.5)115SCID-CV; DSM-IVPR, CI18/31
Joshi, Mick et al. 2010 [78]Open-label trialUSABD enrolled in the olanzapine trials (n = 52, mean age = 8.4 ± 3.1)52K-SADS-E; DSM-IVTR20/31
Joshi, Wozniak et al. 2010 [43]Case–control studyUSAOCD (n = 125, age range = 6–17), BD (n = 82, age range = 6–17)207K-SADS-E; DSM-III-RPR, PH, CI19/31
Masi et al. 2004 [73]Case–control studyItalyBD, OCD, BD-OCD (n = 102, mean age = 14.2 ± 3.2)102DICA-R; DSM-IVPH, CI21/31
Masi et al. 2005 [59]Prospective cohort studyItalyOCD (n = 94, mean age = 13.6 ± 2.8)94DICA-R; DSM-IVPR20/31
Masi et al. 2007 [62]Prospective cohort studyItalyOCD (n = 120, mean age = 13.7 ± 2.8)120K-SADS-PL or DICA-R; DSM-IVPR, PH, TR, CI21/31
Masi et al. 2009c [60]Case–control studyItalyOCD (n = 257, mean age = 13.6 ± 2.7)257K-SADS-PL; DSM-IVPR, TR22/31
Masi et al. 2010c [61]Cross-sectional studyItalyOCD (n = 257, mean age = 13.6 ± 2.7)257K-SADS-PL; DSM-IVPR, PH22/31
Reddy et al. 2000 [63]Cross-sectional studyIndiaOCD (n = 54, age = 16 or less)54DICA-R; DSM-III-RPR15/31
Tillman et al. 2003 [44]Case–control studyUSABD (n = 93, mean age = 10.9 ± 2.6)93WASH-U-K-SADS; DSM-IVPR20/31

Definition of valid diagnostic construct

A diagnosis can be considered valid when one can differentiate it from other diagnoses. To define the validity of the diagnostic construct of interest, we applied the criteria described by Robins and Guze in a classic 1970 paper based on the availability of data on: phenomenology, course of illness, heredity, biological markers, and treatment response [21].

Results

One thousand one hundred and 59 potential studies were identified from searching the selected databases and listing references of relevant articles. After removing duplicates, 749 articles were retrieved. Studies were screened and selected on the basis of prespecified inclusion and exclusion criteria (Fig. 1). The search identified 64 articles that were included in the systematic review.

Figure 1.

Flow diagram of selected articles. *Search strategy limited to March 2013, English language and human subjects older than 6 years old.

Included studies

The characteristics of included studies are reported in Table 1. Twenty-three of the 64 studies were cross-sectional studies, 35 case–control studies, five prospective cohort studies, and one clinical trial. Seven studies (11%) were population-based, while the majority (89%) were hospital-based. Among them, five were conducted on remitted patients. Ten studies considered specifically child and adolescent populations. The smallest study included 15 subjects, while the largest considered a sample of 1416 subjects. The study population was less than 50 subjects in 16% (n = 10) of the included studies, between 50 and 100 in 24% (n = 15), between 100 and 250 in 34% (n = 22), between 250 and 500 in 17% (n = 11), and more than 500 in 9% (n = 6). The majority of the studies were conducted in Europe (n = 29, 45%) and North America (n = 24, 38%). In all the considered studies, diagnosis of BD and OCD was based on the Diagnostic and Statistical Manual (DSM) criteria and were established using validated assessment scales (Table 1).

Comorbidity rates

Forty-seven studies assessed comorbidity rates between BD and OCD. Among them, 26 assessed the lifetime prevalence of comorbid OCD in BD patients (Table 2), 20 assessed the lifetime prevalence of comorbid BD in OCD patients (Table 3), and one study assessed both.

Table 2. Lifetime prevalence of obsessive–compulsive disorder (OCD) comorbidity in bipolar disorder (BD) subjects
BD type n BDBD-IBD-II
References%%%
  1. BD, bipolar disorder; BD-I, bipolar disorder type I; BD-II, bipolar disorder type II.

  2. Polarity of mood episode is not specified in most studies.

Population-based studies
Chen et al. 1995 [3]16721.0  
Faravelli et al. 2004 [23]1911.1  
Merikangas et al. 2007 [24]40813.625.220.8
Hospital-based studies: adults
Boylan et al. 2004 [28]1388.7  
Cassano et al. 1999 [89]48   
Cosoff et al. 1998 [34]2030  
Craig et al. 2002 [25]1383.8  
Dilsaver et al. 2008 [90]69 62.3 
Edmonds et al. 1998 [33]551.8  
Henry et al. 2003 [26]3183  
Koyuncu et al. 2010 [30]21416.311.923.1
Kruger et al. 1995 [36]3735.1  
Kruger et al. 2000 [37]14370.0100
Magalhaes et al. 2010 [31]25912.4  
McElroy et al. 2001 [38]2889910
Pini et al. 1997 [40]24   
Pini et al. 1999 [92]125   
Saunders et al. 2012 [27]7366  
Simon et al. 2003 [32]12213.4  
Simon et al. 2004 [39]4759.910.97.0
Strakowski et al. 1992 [35]417.3  
Strakowski et al. 1998 [29]7716  
Tamam et al. 2002 [42]70   
Zutshi et al. 2006 [41]80   
Hospital-based studies: children, adolescents
Dilsaver et al. 2006 [45]11546.9  
Joshi, Wozniak et al. 2010 [43]8220.7  
Tillman et al. 2003 [44]9324.7  
Table 3. Lifetime prevalence of bipolar disorder (BD) comorbidity in obsessive–compulsive (OCD) subjects by BD type
BD type n BDBD-IBD-II
References%%%
  1. BD, bipolar disorder; BD-I, bipolar disorder type I; BD-II, bipolar disorder type II.

Population-based studies
Adam et al. 2012 [47]3010.0  
Angst et al. 2004 [15]3053.3 30.0
Fireman et al. 2001 [46]1078 (adults)6.0  
198 (children and adolescents)5.0
Hospital-based studies: adults
Bogetto et al. 1999 [88]160  6.8
Diniz et al. 2004 [56]1619.0  
Hantouche et al. 2003 [17]62811.03.08.0
Lensi et al. 1996 [58]263 1.512.1
Maina et al. 2007 [49]20410.32.08.3
Marazziti et al. 2002 [50]11720.55.015.3
Perugi et al. 1997 [53]34515.72.013.6
Perugi et al. 1998 [52]13519.21.417.7
Perugi et al. 1999 [54]7921.53.817.7
Perugi et al. 2002 [55]6855.817.638.2
Timpano et al. 2012 [51]60513.18.44.6
Tukel et al. 2007 [57]1287.0  
Hospital-based studies: children, adolescents
Joshi, Wozniak et al. 2010 [43]12515.2  
Masi et al. 2005 [59]9424.4  
Masi et al. 2007 [62]12035.811.716.7
Masi et al. 2009 [60]25734.2  
Masi et al. 2010 [61]25734.2  
Reddy et al. 2000 [63]541.8  
Comorbid OCD in BD patients

Three population-based studies conducted in Italy and USA, reported lifetime prevalence rates of comorbid OCD in BD patients ranging between 11.1% and 21% [3, 23, 24]. In hospital-based studies, the lifetime prevalence of comorbid OCD in BD patients ranged between 1.8% and 35.1% [25-40] depending on the features of the BD study population included (schizoaffective disorder, bipolar type). When restricting the analysis to studies with sample size greater than 250, the OCD-BD prevalence range was narrower (3–13.6%) [24, 26, 27, 31, 38, 39], and the mean prevalence rate of BD-I comorbidity was 15%, while BD-II was 12.6% [24, 38, 39]. When considering remitted patients, comorbidity prevalence rates were 35% (OCD-BD patients) [41] and 38.6% (OCD-BD-I patients) [42]. Studies focusing on prepubertal and adolescent BD subjects reported OCD comorbidity rates ranging between 20.7% and 46.9% [43-45].

Comorbid BD in OCD patients

Two population-based studies conducted, respectively, in USA and Germany, reported 6% and 10% of lifetime prevalence of comorbid BD in OCD adults [46, 47] and 5% in OCD children and adolescents [46]. Using the bipolar spectrum concept in the general population [48], a study reported 53.3% and 30% of BD-OCD and BD-II-OCD patients respectively [15]. In hospital-based studies, the lifetime prevalence of comorbid BD in OCD adult patients ranged between 7% and 55.8% [17, 49-57]. When restricting the analysis to studies with sample size greater than 250, the BD-OCD prevalence range was narrower (11–15.7%) [17, 51, 53], and the mean prevalence rate of BD-I comorbidity was 3.7%, while BD-II was 9.5% [17, 51, 53, 58]. Hospital-based studies conducted in OCD children and adolescents described prevalence of comorbid BD ranging between 1.8% and 35.8% [43, 59-63]. Only one study differentiated BD comorbidity by type in child and adolescent population [62]. When considering a dimensional approach to BD diagnose, the comorbid prevalence rates were 30% for lifetime hypomanic episodes and almost 50% for cyclothymic traits [17].

Phenomenology

Eighteen studies focused on the phenomenology of patients with comorbidity between BD and OCD (Table 4). Obsessions and compulsions were assessed through the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) except for one study that used a semistructured in-depth interview to collect demographic data, family history, psychopathological features, course of illness, and symptoms [52, 53, 55]. Y-BOCS total score did not provide unequivocal evidence that BD-OCD patients have more severe OCD compared to non-BD-OCD patients [51, 64] while Y-BOCS-11 was associated with poorer insight into obsessive–compulsive symptoms in patients with comorbidity [50, 65]. In BD patients, OCD comorbidity manifested more likely during mixed mania and depression episodes [66, 67].

Table 4. Diagnostic validators: phenomenology, heredity, biological markers, and treatment
ReferencesResults
  1. BD, bipolar disorder; BD-I, bipolar disorder type I; BD-II, bipolar disorder type II; OCD, obsessive–compulsive disorder; OCS, obsessive–compulsive syndrome; Pt., patients; vs., versus; YBOCS, Yale-Brown Obsessive–Compulsive Scale; YMRS, Young Mania Rating Scale; CGI-S, Clinical Global Impression Severity; SRIs, serotonin reuptake inhibitors; 5-HTT, serotonin transport protein. Differences statistically significant (P < 0.05).

Phenomenology
Dell'Osso et al. 2000 [67]BD pt. with depression had a higher frequency of OCD than those with mania
Hasler et al. 2005 [69]Obsessions of symmetry and repeating, counting, ordering/arranging compulsions were associated with BD (OR = 1.5, 95% CI = 1.1–2.2)
Issler et al. 2005 [70]Higher frequency of aggression, symmetry, contamination, hoarding and miscellaneous obsessions and cleaning, checking, ordering, and several other compulsions in BD-OCD pt. vs. non-BD-OCD pt.
Joshi, Wozniak et al. 2010 [43]Higher rates of hoarding/saving obsessions (58% vs. 23%) and compulsions (63% vs. 20%) in BD-OCD pt. vs. non-BD-OCD pt.
LaSalle-Ricci et al. 2006 [72]Diagnosis of BD-I was significantly and positively correlated with hoarding
Mahasuar et al. 2011 [65]Fewer pathological doubts (50% vs. 75%), higher rates of pathological slowness (32% vs. 9%), and reassurance seeking (33% vs. 11%), poorer insight into obsessive–compulsive symptoms (YBOCS-11: 1.11 ± 1.00 vs. 0.52 ± 0.71) in BD-OCD pt. vs. non-BD-OCD pt.
Maina et al. 2007 [49]Higher rates of hoarding (33.3% vs. 10.9%) and sexual obsessions (42.9% vs. 19.7%) and repeating compulsions (71.4% vs. 42.6%) in BD-OCD pt. vs. non-BD-OCD pt.
Marazziti et al. 2002 [50]Lower level of insight in BD pt. with a positive history of repeated manic or hypomanic episode
Masi et al. 2004 [73]Higher rates of ordering compulsions (30% vs. 5.7%) in BD-OCD pt. vs. non-BD-OCD pt.
Masi et al. 2007 [62]Higher rates of hoarding obsessions and compulsions in BD-OCD pt. vs. non-BD-OCD pt. (14% vs. 2.6%)
Masi et al. 2010 [61]Higher frequency of BD in OCD pt. with hoarding compulsions
McElroy et al. 1995 [66]Pt. with episodes of mixed mania were more likely to have comorbid OCD than those with pure euphoric mania (21% vs. 4%)
Perugi et al. 1997 [53]Higher rates of sexual (41.1% vs. 20.3%) and religious obsessions (23.5% vs. 9.8%) and lower rate of checking rituals (56.8% vs. 73.4%) in BD-OCD pt. vs. non-BD-OCD pt.
Perugi et al. 2002 [55]Higher rates of sexual (55.3% vs. 26.7%) and lower rate of ordering rituals (18.4% vs. 50.0%) in BD-OCD pt. vs. non-BD-OCD pt.
Shabani et al. 2008 [68]Higher rates of sexual (87.2% vs. 0.0%) and religious (38.5% vs. 15.4%) and aggressive obsessions (35.9% vs. 7.7%) and lower rate of contamination obsessions (64.1% vs. 84.6%) in BD-OCD pt. vs. non-BD-OCD pt. Higher prevalence of miscellaneous compulsions (53.8% vs. 25.6%, P = 0.01) and lower prevalence of washing compulsion (48.7% vs. 84.6%) in BD-OCD pt. vs. non-BD-OCD pt.
Timpano et al. 2012 [51]More severe OC symptoms (YBOCS total score: 22.7 ± 9.8 vs. 18.2 ± 8.5) in BD-OCD pt. vs. non-BD-OCD pt.
Tukel et al. 2006 [71]Higher rates of symmetry/exactness obsessions (73.1% vs. 40.8%) and ordering/arranging compulsions (61.5% vs. 36.7%) and lower rates of somatic obsessions (7.7% vs. 14.3%) in BD-OCD pt. vs. non-BD-OCD pt.
Zutshi et al. 2007 [64]Less severe OCD (YBOCS total score: 14.43 ± 7.27 vs. 25.53 ± 7.54) and lower rates of washing (32% vs. 62%), repeating (7% vs. 54%), and ordering (3% vs. 27%) compulsions in BD-OCD pt. vs. non-BD-OCD pt.
Heredity
Angst et al. 2005 [16]No statistically significant differences about the family history for OCD, depression, and mania in OCS pt. with or without BD comorbidity
Koyuncu et al. 2010 [30]Higher frequency rate of OCD in first-degree relatives of OCD-BD pt. vs. non-OCD-BD pt. (45.7% vs. 5.7%); no statistical significant differences in family history for BD
Mahasuar et al. 2011 [65]Statistically non-significant trends of higher prevalence of family history for mood disorders in BD-OCD pt. and lower prevalence of family history for OCD vs. non-BD-OCD pt.
Perugi et al. 2002 [55]Statistically non-significant trends of higher prevalence of family history for mood disorders in BD-OCD pt. and lower prevalence of family history for OCD vs. non-BD-OCD pt.
Perugi et al. 1998 [52]Positive correlation between episodic OCD and family history for mood disorders compared with pt. with continuous OCD (54.1% vs. 34.7%)
Zutshi et al. 2007 [64]Higher prevalence of family history for mood disorders in BD-OCD pt. (36% vs. 6%) and lower prevalence of family history for OCD (0.0% vs. 21%) vs. non-BD-OCD pt.
Biological markers
Cannon et al. 2006 [77]Higher 5-HTT binding potential in OCD-BD pt. vs. non-OCD-BD pt. in the insula, in the posterior cingular cortex, in the subgenual anterior cingulate cortex, and in the dorsal cingulate cortex
Treatment
Joshi, Mick et al. 2010 [78]Lower treatment response in OCD-BD pt. vs. non-OCD-BD pt. (YMRS mean reduction: −5.9 ± 13.1 vs. −13.7 ± 18.8; ≥30% reduction: 25% vs. 63%; CGI-S Improvement score ≤ 2: 25% vs. 68%). No statistically significant differences reported in the rate of drop-outs or adverse effects
Masi et al. 2007 [62]All OCD patients treated with SRIs (sertraline, fluvoxamine, fluoxetine, paroxetine, citalopram, clomipramine) in mono- or poly-therapy (89.2%) and/or mood stabilizers (valproic acid, lithium) (37.5%) and/or antipsychotics (risperidone, olanzapine, quetiapine) (30.8%). BD-OCD pt. received more mood stabilizers vs. non-BD-OCD pt. (86.0% vs. 10.4%); 69.8% of BD-OCD pt. were treated with SRIs; higher rate of non-responders to pharmacological treatment in BD-OCD pt. vs. non-BD-OCD pt. (54.7% vs. 25.6%)
Masi et al. 2009 [60]BD-OCD pt. more frequently received poly-pharmacy than taking SRIs alone (51.1% vs. 5.6%)
Perugi et al. 2002 [55]In BD-OCD pt., compared to non-BD-OCD pt., clomipramine and, to a lesser extent, SRIs were related to a higher rate of manic/hypomanic switches (clomipramine: 39.1% vs. 4.1%; SSRIs: 13.9% vs. 0.0%). More frequent pharmacologic mania/hypomania in BD-OCD pt. that did not concomitantly receive mood stabilizers (38.7% vs. 8.8%). Polypharmacologic treatments were required in 31.6% (a combination of mood stabilizers, lithium plus antiepilectis) and 10.5% (a combination of mood stabilizers with atypical antipsychotics, clozapine, olanzapine, risperidone) of the BD-OCD pt

Sexual [49, 53, 55, 68], symmetry [69-71], aggressive [68, 70], religious [53, 68], contamination [70], and hoarding [49] obsessions were positively associated with BD-OCD comorbidity. In addition, three studies reported lower rates of somatic and contamination obsessions and pathological doubt in patients with comorbidity as compared to non-comorbid patients [65, 68, 71].

With regard to compulsions, patients with comorbidity were reported to have higher rates of order [55, 69-71], control/checking [70] and repeating [49, 69] rituals, pathological slowness [65], reassurance seeking [65], cleaning [70], and counting [69] compulsions. Conversely, three studies showed lower rates of control [53], washing [64, 68], ordering, and repeating [64] compulsions in subjects with comorbidity. In the presence of BD, children and adolescents with OCD more frequently were associated with hoarding obsessions and compulsions [43, 61, 62, 72] and ordering [73] compulsions.

Course of illness

Table 5 summarizes findings on course of illness in comorbid and non-comorbid patients reported in the included studies.

Table 5. Diagnostic validator: Course of illness
ReferencesResults
  1. BD, bipolar disorder; BD-I, bipolar disorder type I; BD-II, bipolar disorder type II; OCD, obsessive–compulsive disorder; OCS, obsessive–compulsive syndrome; Pt., patients; vs., versus; GAD, generalized anxiety disorder; PD, panic disorder; PTSD, post-traumatic stress disorder; ED, eating disorder; ADHD, attention-deficit/hyperactivity disorder; CD, conduct disorder; GAF, Global Assessment of Functioning; CGI-S, Clinical Global Impression Severity; CGI-I, Clinical Global Impression Improvement; C-GAS, Children's Global Assessment Scale; WHOQOL, World Health Organization Quality of Life Brief; Q-LES-Q, Quality of Life Enjoyment and Satisfaction Questionnaire; LIFE-RIFT, Range of Impaired Functioning Tool. Differences statistically significant (P < 0.05).

  2. a

    Episodic OCD: at least one circumscribed symptom-free interval (6 months) was present.

  3. b

    Chronic OCD: if symptoms persisted for most of the course, causing significant distress and impairment in functioning.

  4. c

    Chronic BD: if all criteria for a major mood episode were met continuously for at least 2 years.

Age and type of onset
Chen et al. 1995 [3]Similar mean age of onset of OCD and BD
Maina et al. 2007 [49]In most cases (52.4%), the onset of OCD in BD-OCD pt. was concomitant to the first mood episode rather than beginning before (38.1%) or following it (9.5%)
Ortiz et al. 2011 [91]In BD pt., an early onset of BD and an early onset of depressive episodes were associated with comorbid OCD
Perugi et al. 1997 [53]BD-OCD pt. more inclined to a gradual onset vs. non-BD-OCD pt. (68.5% vs. 49.0%)
Perugi et al. 2002 [55]In most cases (52.6%), the onset of OCD in BD-OCD pt. was concomitant to the first mood episode rather than beginning before (31.6%) or following it (15.8%)
Zutshi et al. 2007 [64]In most cases (54%), BD-OCD pt. had onset of OCD before the onset of BD
Course of illness
Goes et al. 2012 [75]Higher number of depressive episodes in OCD-BD pt. vs. non-OCD-BD pt. (19.7 vs. 10.3)
Issler et al. 2005 [70]Higher rate of chronic course of OCD in OCD-BD pt. vs. non-OCD-BD pt. (86.7% vs. 13.3%)b
Issler et al. 2010 [74]Higher number of depressive episodes in OCD-BD pt. vs. non-OCD-BD pt. (8.9 ± 4.2 vs. 4.1 ± 2.7). Higher rate of chronic course of BD in OCD-BD pt. vs. non-OCD-BD pt. (66.7% vs. 20%)c
Koyuncu et al. 2010 [30]Higher rate of chronic course of BD in OCD-BD pt. vs. non-OCD-BD pt. (17.1% vs. 0.0%)c
Mahasuar et al. 2011 [65]Episodic course overrepresented in BD-OCD pt. vs. pt. without comorbidity (53% vs. 12%); higher rates of continuous (40% vs. 35%) or subclinical (32% vs. 6%) course in non-BD-OCD pt. vs. BD-OCD pta,b
Perugi et al. 1997 [53]BD-OCD pt. were more inclined toward an episodic course of OCD symptoms vs. non-BD-OCD pt. (42.6% vs. 26.9%)a
Perugi et al. 1998 [52]Statistically non-significant trends of higher rate of episodic OCD course in OCD pt. with comorbid BD-II vs. pt. without comorbiditya
Perugi et al. 2002 [55]Higher rate of episodic course of OCD in BD-OCD pt. vs. non-BD-OCD pt. (52.6% vs. 16.7%)a
Strakowski et al. 1998 [29]BD and OCD cycled together in 44% of BD-OCD pt. followed for 12 months after a first hospitalization
Tukel et al. 2006 [71]Higher number of BD-OCD pt. with an episodic course of OCD vs. non-BD-OCD pt. (42.3% vs. 10.9%)a
Tukel et al. 2007 [57]Higher frequency of BD in pt. with episodic OCD than in pt. with chronic OCD (20.8% vs. 3.8%)a
Zutshi et al. 2007 [64]Higher rate of episodic course of OCD in BD-OCD pt. vs. non-BD-OCD pt. (75% vs. 3%); higher rate of chronic course of OCD in non-BD-OCD pt. vs. pt. with comorbidity (97% vs. 14%). Most BD-OCD pt. (78%) either had OCD confined exclusively to depressive episodes or reported worsening of OCD during depression. Improvement in OC symptoms was noted in 64% pt. during manic/hypomanic episodesa,b
Global functioning and quality of life
Angst et al. 2005 [16]Statistically significant difference between BD-OCD pt. and ‘pure’ OCD pt. in terms of global (100.0% vs. 87.5%) and work (100.0% vs. 78.1%) impairment
Centorrino et al. 2006 [76]No statistically significant differences in terms of GAF and CGI-S scores between ‘pure’ OCD and ‘pure’ BD pt. and pt. with comorbidity
Joshi, Wozniak et al. 2010 [43]Lower lifetime and current level of functioning on the GAF scale in BD-OCD and OCD-BD pt. vs. pt. without comorbidity
Magalhaes et al. 2010 [31]Lower score on all domains of the WHOQOL (physical, psychological, social, environmental) in BD-OCD pt. vs. non-OCD-BD pt.
Mahasuar et al. 2011 [65]No statistically significant differences in terms of GAF, CGI-S, and CGI-I scores between OCD pt. with or without comorbid BD
Masi et al. 2004 [73]Higher mean global severity score, measured during the first visit by means of CGI-S, in BD-OCD pt. vs. pure OCD (4.80 ± 0.7 vs. 4.4 ± 0.5); similar scores in BD and BD-OCD pt. (5.0 ± 0.8 and 4.80 ± 0.7 respectively). Higher rates of severity at the end of the follow-up in BD and BD-OCD pt. vs. OCD pt. (3.1 ± 1.0 and 2.7 ± 1.0, respectively, vs. 2.0 ± 0.7)
Masi et al. 2007 [62]Greater functional impairment (C-GAS) (43.1 ± 8.7 vs. 46.4 ± 9.0) and greater severity (CGI-S) at the baseline and at the end of the follow-up in BD-OCD pt. vs. non-BD-OCD pt.
Simon et al. 2004 [39]The presence of current anxiety disorder (with the exception of OCD) in BD pt. was associated with poorer functioning (LIFE-RIFT) and poorer quality of life (Q-LES-Q)
Timpano et al. 2012 [51]No statistically significant differences in terms of GAF total score between OCD pt. with or without comorbid BD
Suicide ideations and attempts
Chen et al. 1995 [3]Higher lifetime rates of ‘thoughts of suicide’, ‘suicide attempts’, ‘thoughts of death’, and ‘wanting to die’ in BD-OCD pt. vs. non-OCD-BD pt.
Dilsaver et al. 2006 [45]OCD-BD pt. associated with a 2.4-fold increase in the odds of suicidal ideation vs. non-OCD-BD pt. (95% CI = 1.0–5.8)
Goes et al. 2012 [75]Higher rates of history of suicide attempts in OCD-BD pt. vs. non-OCD-BD pt. (48.3 vs. 29.6%)
Koyuncu et al. 2010 [30]No statistically significant differences between OCD-BD pt. and non-OCD-BD pt. in terms of lifetime suicide attempts
Kruger et al. 2000 [37]Higher incidence of prior suicide attempts in OCD-BD pt. vs. non-OCD-BD pt. (90% vs. 38%)
Magalhaes et al. 2010 [31]Higher rates of history of suicide attempts in OCD-BD pt. vs. non-OCD-BD pt. (70% vs. 35%)
Simon et al. 2004 [39]Higher rates of history of suicide attempts in OCD-BD pt. vs. non-OCD-BD pt.
Hospitalization
Centorrino et al. 2006 [76]Re-hospitalizations were similar among BD-OCD pt. and BP pt., but 2.9-times more frequent among BD-OCD pt. vs. non-BD-OCD pt.
Joshi, Wozniak et al. 2010 [43]Greater rates of hospitalization in BD-OCD and OCD-BD pt. vs. pt. without comorbidity (BD-OCD: 41.2% vs. 10.4%; OCD-BD: 31.6% vs. 10.4%)
Mahasuar et al. 2011 [65]Higher total number of hospitalization in BD-OCD pt. vs. non-BD-OCD pt. (1.55 ± 1.70 vs. 0.16 ± 0.41)
Masi et al. 2007 [62]More frequent need of hospitalization (62.8% vs. 28.6%) in BD-OCD pt. vs. non-BD-OCD pt.
Timpano et al. 2012 [51]More frequent need of hospitalization (58.2% vs. 13.8%) in BD-OCD pt. vs. non-BD-OCD pt.
Substance and alcohol abuse
Angst et al. 2005 [16]Higher rates of substance (50% vs. 31.3%) and alcohol (37.5% vs. 18.8%) abuse/dependence in OCS-BD pt. vs. ‘pure’ OCS pt.
Boylan et al. 2004 [28]Higher rates of past substance abuse or dependence in OCD-BD pt. vs. non-OCD-BD pt. (50% vs. 20%)
Chen et al. 1995 [3]OCD-BD pt. have rates of alcohol and drug abuse of 31.4 and 37.1% respectively. Non-comorbid pt. have rates of alcohol and drug abuse of 36.4% and 28.2% respectively
Magalhaes et al. 2010 [31]Higher rates of lifetime alcohol dependence in OCD-BD pt. vs. non-OCD-BD pt. (31% vs. 10%)
Maina et al. 2007 [49]Strong association between BD-OCD pt. and substance use disorders (28.6% vs. 4.9%) vs. non-BD-OCD pt.
Perugi et al. 2002 [55]Higher association between BD-OCD comorbidity and sedatives, nicotine, alcohol, and coffee (P = 0.03) vs. pt. without comorbidity
Timpano et al. 2012 [51]Strong association between BD-OCD pt. and substance use disorders (OR 3.18, 95% CI = 1.81–5.58) and alcohol use disorders (OR 2.21, 95% CI = 1.34–3.65) vs. non-BD-OCD pt.
Axis I and Axis II comorbidities
Issler et al. 2010 [74]Higher prevalence rates of any anxiety disorder other than OCD (93.3 vs. 53.3%), impulse control disorders (60 vs. 13.3%), ED (33.3 vs. 0%), and tic disorder (33.3 vs. 0%) in OCD-BD pt. vs. non-OCD-BD pt.
Masi et al. 2004 [73]Higher rates of ADHD-CD (51.3% vs. 20.0%) and CD (29.7% vs. 13.3%) in BD pt. vs. BD-OCD pt. Lower rates of ADHD (2.8% vs. 16.7%) and higher rates of GAD (77.1% vs. 16.7%) in OCD pt. vs. BD-OCD pt.
Maina et al. 2007 [49]Higher rates of at least one Cluster A personality disorder (P = 0.027), at least one Cluster B personality disorder, and narcissistic and antisocial personality disorder in BD-OCD pt. vs. non-BD-OCD pt.
Masi et al. 2007 [62]Higher comorbidity with ADHD and ODD, and a lower comorbidity with GAD in BD-OCD pt.
Perugi et al. 2002 [55]Higher rates of current comorbidity with panic disorder/agoraphobia in BD-OCD pt. vs. non-BD-OCD pt. (52.6% vs. 16.7%)
Timpano et al. 2012 [51]BD-OCD pt. were over two times as likely to be also diagnosed with PD (OR 2.26, 95% CI = 1.38–3.71), agoraphobia (OR 2.30, 95% CI = 1.35–3.91), PTSD (OR 2.85, 95% CI = 1.49–5.45), and ED (OR 2.03, P < 0.001, 95% CI = 1.14–3.60) vs. pt. without comorbidity
Age and type of onset

Some investigators reported that the onset of OCD usually was concomitant with the first mood episode, rather than preceding or following it [49, 55]. In line with these findings, the ECA population-based study reported similar mean age of onset of OCD and BD among BD subjects [3]. In contrast, a smaller study from India reported that the majority of BD-OCD patients had onset of OCD before the onset of BD [64]. In one study, OCD-BD patients were more likely to have a gradual onset of OCD symptoms than OCD patients without BD [53].

Course of illness

Twelve studies assessed the course of illness and classified patients in chronic and episodic subgroups (Table 5). Higher rates of episodic OCD were reported in BD-OCD patients as compared to patients without comorbidity [52, 53, 55, 57, 64, 65, 71] with episodic frequency rates reaching 75% in one study [64]. In particular, one Italian study reported that episodic OCD more frequently was associated with BD-II comorbidity as compared to chronic OCD [52]. The only study that found a more profound chronic course of OCD in OCD-BD patients was conducted in a small group of 15 women mainly with BD-I [70]. Two studies conducted in BD patients reported higher rates of chronic course of BD in OCD-BD patients vs. non-comorbid patients [30, 74].

In other studies, most BD-OCD subjects either had OCD confined exclusively to depressive episodes or reported worsening of OCD during depression [64]; one report showed that in 44% of BD-OCD patients, BD and OCD cycled together [29]. When assessed, the total number of depressive episodes was higher in patients with OCD-BD comorbidity than in BD alone [74, 75].

Global functioning and quality of life

As showed in Table 5, studies conducted in adult and child patients reported that the presence of comorbidity between BD and OCD was associated with poorer functioning and poorer quality of life in all the domains (physical, psychological, social, environmental) vs. patients with ‘pure’ BD or ‘pure’ OCD [16, 31, 39, 43, 51, 62, 73]. Two hospital-based studies presented similar trends without, although not showing statistically significant differences [65, 76].

Suicide ideation and attempts

In the ECA database, OCD-BD patients had statistically significant higher lifetime rates of ‘thoughts of suicide’, ‘suicide attempts’, ‘thoughts of death’, and ‘wanting to die’ vs. non-OCD-BD patients [3]. These features were confirmed in most recent studies both in OCD-BD patients [31, 37, 39, 75] and in BD-OC syndrome patients [16]. OCD also was associated with a 2.4-fold increase in the odds of suicidal ideation among BD adolescents as compared to non-comorbid adolescents [45]. One small sample size study (n = 35) did not report statistically significant difference in terms of suicide attempts between ‘pure’ BD patients and OCD-BD patients [30].

Hospitalization

Five studies report higher hospitalization rates in patients BD-OCD vs. those without comorbidity [43, 51, 62, 65, 76].

Substance and alcohol abuse

Five studies investigated the rates of substance abuse in patients with comorbidity between BD and OCD (Table 5). A large hospital-based study (n > 500) found that BD-OCD patients had an over two times higher likelihood to be also diagnosed with substance use disorders (OR 3.18, 95% CI = 1.81–5.58) and alcohol use disorders (OR 2.21, 95% CI = 1.34–3.65) vs. non-comorbid patients [51]. Similar findings were reported in previous studies [16, 28, 31, 49]. In addition, one report showed a positive association between BD-OCD comorbidity and sedatives, nicotine, alcohol, and coffee use [55]. The ECA study confirmed the higher rate of drug abuse in OCD-BD subjects vs. non-comorbid subjects (37.1% vs. 28.2%) [3].

Axis I and Axis II comorbidities

Six studies assessed the presence of Axis I and Axis II disorders in patients with BD and OCD comorbidity (Table 5). Confirming previous studies [55, 74], one report showed that BD-OCD patients were over two times as likely to be also diagnosed with panic disorder, agoraphobia, post-traumatic stress disorder, and eating disorder vs. patients without comorbidity [51]. When assessed in child and adolescent populations, higher rates of attention-deficit/hyperactivity disorder, oppositional defiant disorder, and lower rates of generalized anxiety disorder were reported in BD-OCD patients as compared to non-comorbid patients [62, 73].

Only one study conducted in Italy assessed Axis II comorbidities comparing BD-OCD and non-BD-OCD patients; it reported higher rates of at least one Cluster A personality disorder, at least one Cluster B personality disorder, and narcissistic and antisocial personality disorder in subjects with comorbidity [49].

Heredity

No studies were found that examined familial transmission of comorbid BD-OCD (Table 4).

Five studies (four hospital-based and one population-based) assessed family history for OCD or BD in comorbid BD-OCD probands using semi- or unstructured clinical interview and clinical records [16, 30, 55, 64, 65]. Three studies reported BD-OCD patients to have higher prevalence of family history for mood disorders and lower prevalence of family history for OCD vs. non-BD-OCD patients [55, 64, 65], while one study found the reverse [30]. The Zurich study, which is the sole population-based study, found no statistically significant differences about the family history for OCD, depression, and mania in OCD patients with or without BD comorbidity [16]. In addition, with regard to the course of illness, a family history for mood disorders was reported to be more frequent in episodic OCD patients than in continuous/chronic OCD patients [52].

Biological markers

One study was found on biological markers of comorbidity between BD and OCD (Table 4). Investigators from USA analyzed the serotonin transporter (5-HTT) binding potential (BP) using positron-emission tomography (PET) and [11C] DASB, a radioligand with high sensitivity and specificity for 5-HTT. They reported higher 5-HTT binding potential in OCD-BD subjects as compared to non-OCD-BD in the insula, in the posterior cingular cortex, in the subgenual anterior cingulate cortex, and in the dorsal cingulate cortex [77].

Treatment

No double-blind randomized clinical trials (RCTs) have been performed in patients with comorbidity of BD and OCD. Four studies were retrieved on non-randomized treatment response in those subjects (Table 4).

One study reported that in BD-OCD patients, clomipramine and, to a lesser extent, SRIs were associated with a higher rate of manic switches vs. non-comorbid patients. Besides, BD-OCD patients that did not concomitantly receive mood stabilizers presented more frequent pharmacologic mania as compared to non-BD-OCD patients. A combination of mood stabilizers, alone and with atypical antipsychotics, was required, respectively, in 31.6% and 10.5% of BD-OCD patients [55].

From secondary analysis of data collected from three identically designed open-label trials on olanzapine treatment in youth BD patients, investigators from USA found that the treatment response in OCD-BD patients was significantly lower than in non-comorbid patients, and no statistically significant differences were reported in the rates of drop-outs or adverse effects [78].

One report showed that child and adolescent patients with BD and OCD more frequently received poly-pharmacy than SRIs alone [60]; they also received more mood stabilizers and presented higher rates of non-response to pharmacological treatment vs. non-BD-OCD patients [62].

Discussion

In agreement with Kraepelin's thought, a psychiatric diagnosis is best established by its longitudinal course of illness [79], we can conclude that the evidence so far supports the view that the majority of cases of ‘comorbid’ BD-OCD are in fact cases of BD, with OC symptoms as secondary manifestations of depressive or manic mood episodes. Findings on the other diagnostic validators were heterogeneous and mainly suggested increased severity of illness or functioning without clarifying nosological validity. These conclusions are based on careful weighing of the quality of the included observational studies, as shown in Table 1, based on research design, sample size, and statistical analysis.

In the course of illness studies, the most definitive finding was that about 50–75% of OCD cases were limited to mood episodes in BD. In other words, the majority of comorbid OCD cases appear to be secondary to mood episodes. A substantial minority of comorbid OCD cases are not episodic and may represent ‘true’ OCD independent of BD. Overall, OC symptoms in comorbid patients emerge more often—and sometimes exclusively—during depressive episodes, and comorbid BD and OCD even cycle together, with OC symptoms often remitting during manic/hypomanic episodes. In addition, evidence reported that SRI antidepressants produce more mania/hypomania in BD-OCD patients than in non-comorbid patients. These findings are of fundamental importance to guide clinical management of patients with comorbidity between BD and OCD and suggest that physicians might avoid treating comorbid OC symptoms in BD with antidepressants. In fact, such an approach might worsen BD via SRI-induced mania/hypomania [7-11].

These results can also be interpreted by means of the concept of a diagnostic hierarchy in psychiatry [80], as opposed to the egalitarian DSM system. In the hierarchical approach, advocated in classic European psychopathology, anxiety presentations, like OCD, are not diagnosed as separate conditions when co-occurring with mood presentations, like BD. In other words, OCD was not diagnosed unless BD was ruled out. If correct, as seems to be the case based on this review, OCD patients should be evaluated about family history for mood disorders and other evidence of bipolarity.

If most OCD symptoms are secondary to BD, then it may be that both groups of symptoms may respond to adequate mood stabilizer treatment. In a minority with persistent OCD, despite improvement in mood episodes, addition of low doses of antidepressants could be considered while strictly monitoring emerging symptoms of mania or mixed states. Several case reports report improvement of OC symptoms with mood stabilizers [81, 82] and atypical antipsychotics [83] in comorbid patients. This review suggests that more definitive treatment research with such agents alone, without SRIs, is needed in BD-OCD.

It could be argued that more depressive episodes with OCD could be secondary to lack of response to the treatment of OCD. Data from the STEP-BD study [84] showed that co-occurrence anxiety disorder including OCD predicted increased risk of relapse into depression. This claim has two major flaws: First, it would not explain OCD during manic episodes. Second, it would further require an unlikely assumption: since patients with mood illnesses (bipolar and unipolar) tend to have dozens of episodes or more in a lifetime, this perspective would have to claim that every time someone has a depressive episode, an OCD exacerbation causes it, and this co-occurrence happens coincidentally dozens of times in a lifetime. Although possible, this opinion has not been studied or proven. The other perspective, that mood illnesses exist with the occurrence of dozens of mood episodes in a lifetime, is based on over a century of solid scientific research [79]. The observation of OCD symptoms during some of those dozens of mood episodes would seem to be more likely related to the ongoing mood illness rather than the intermittent OCD symptoms. We agree, however, that this question of causation cannot be definitively adjudicated based on clinical data, whether retrospective or prospective. Correlation and causation are not exactly the same, as is well known, and other aspects of causation need to be taken into account, as in the classic formulation of the famous epidemiologist A. Bradford Hill [85].

With regard to the meaning of comorbidity, we define it in accordance with Feinstein's definition, that considered the presence of a ‘distinct additional clinical entity’ [86]. In Feinstein's formulation, the implication was that a completely different and independent disease occurred at the same time as another disease. These two diseases co-occurred, more often than not, randomly. On the contrary, DSM explicitly produces overlapping clinical criteria for many diagnoses, especially mood and anxiety disorders, guaranteeing comorbidity in quite a different sense than in the medical meaning of the term as co-occurrence of independent diseases. Using DSM definition, it is unclear whether concomitant diagnoses actually reflect the presence of distinct clinical entities or refer to multiple manifestations of a single clinical entity [87]. According to Feinstein's definition, ‘true’ OCD comorbid with BD would represent the random co-occurrence of two independent diseases, estimated by multiplication of prevalence rates, which are about 1% for each condition, producing a very low expected true comorbidity of 0.1% in the general population.

The main limitation of this systematic review is linked to the study design and analysis strategy of the included studies, as documented in the quality assessment scale used. Most studies are observational and based on retrospective assessments. Some of them do not include a control group. Small sample size and enrollment of subjects mainly from BD-OCD out-patient units may limit generalizability of these results. Only a few hospital-based studies assessed patients during clinical remission, and BD subjects were more frequently studied during episodes of depression rather than mania. Potential confounding factors in these studies include demographic and historical illness variables, which often were not appropriately analyzed through multivariate modeling. The strength of the selected studies is that the diagnosis of BD, OCD, and comorbid psychiatric disorders was consistently based on the DSM criteria and were established by trained investigators using validated assessment scales mainly with interrater reliability. The main strength of this review is its being systematic and its including the entire scientific evidence published so far on the main medical databases.

Further studies are needed to confirm or refute our findings and consequent clinical recommendations. In particular, studies addressing hereditary and biological markers are essential to illuminate pathogenetic mechanisms that underlie BD-OCD comorbidity and clarify its diagnostic validity and treatment approach.

Declaration of interest

Dr. Amerio, Dr. Odone, Dr. Liapis: no competing interests. Dr. Ghaemi: in the past 12 months has received a research grant from Takeda Pharmaceuticals. Neither he nor his family holds equity positions in pharmaceutical corporations.

Appendix

MEDLINE search strategy.

SETMEDLINE

Note

  1. Words written in italic were used as MeSH headings, the others were used as free text.

1Bipolar disorder
2BD
3Bipolar
4Manic-depressive disorder
5Manic-depressive
6Manic
7 Bipolar disorder
8Sets 1–7 were combined with ‘OR’
9Obsessive–compulsive disorder
10Obsessive–compulsive
11OCD
12 Obsessive–compulsive disorder
13Sets 9–12 were combined with ‘OR’
14Sets 8 and 13 were combined with ‘AND’
15Set 14 was limited to March 2013, Humans, English language, Child: 6–12 years, Adolescent: 13–18 years, Adult: 19+ years

Ancillary