Obsessive–compulsive disorder with and without bipolar disorder

Authors


Y.C. Janardhan Reddy, DPM, MD, OCD Clinic, NIMHANS, Bangalore 560029, India. Email: ycjreddy@gmail.com

Abstract

Aim:  Bipolar disorder (BD) is often comorbid with obsessive–compulsive disorder (OCD). In this study, we compared clinical profile and course of subjects with a primary diagnosis of OCD with and without BD.

Methods:  We compared 34 subjects with primary diagnosis of OCD with BD and 57 subjects with a diagnosis of OCD without BD. Structured interview schedules, clinical rating scales, and information from clinical charts were utilized to assess patients.

Results:  OCD with BD was characterized by: (i) an episodic course; (ii) a higher number of depressive episodes, greater suicidality and a higher rate of hospitalization; (iii) fewer pathological doubts and more miscellaneous compulsions; and (iv) poorer insight into obsessive–compulsive symptoms.

Conclusions:  Episodic course appears to be typical of OCD with BD. Bipolarity has a pathoplastic effect on OCD and it is possible that some forms of OCD and BD are pathophysiologically related. Bipolar OCD is associated with a higher rate of depressive episodes, higher suicidality and more frequent hospitalizations, suggesting greater morbidity. Long-term prospective follow-up studies and studies addressing pathophysiology and genetic basis are needed to understand the complexity of such comorbidity.

A STRONG ASSOCIATION between obsessive–compulsive disorder (OCD) and mood disorders has been reported in recent studies.1–4 Depression is the most common comorbid mood disorder in OCD with rates ranging from 13% to 75%.5–9 However, bipolar disorder (BD) is not uncommon in patients with OCD with prevalence rates ranging from 10% to 20%.1,3,10–12

Mood disorders have been shown to have a pathoplastic effect on the course and outcome of OCD.3,4,9,13–15 An episodic course rather than a chronic course has been described when OCD is comorbid with BD.3,4,16,17 OCD is known to worsen in depression and improve in hypomania/mania.3,16,18 OCD with BD is associated with a significantly higher rate of sexual, religious and symmetry obsessions3,12,19 and repeating, counting and ordering/arranging compulsions.3,12,19 In addition, higher rates of depression and anxiety disorders have been reported in OCD with BD.2,3,11,12,16

Studies on OCD and BD comorbidity broadly fall into two categories: those that have recruited patients with a primary diagnosis of OCD3,4,10,11,17,19–22 and those that have studied patients with a primary diagnosis of BD.2,23–25 In our previous study of subjects with a primary diagnosis of BD, OCD when comorbid with BD was characterized by an episodic course, a higher family loading for affective disorders (not for OCD), and less severe OC symptoms but higher rates of comorbid anxiety disorders and depressive episodes.16

Although the episodic nature of OCD is documented,3,4,17,21 it is not clear if clinical profile, insight and comorbidity patterns are unique to primary OCD when comorbid with bipolar disorder. It is possible that OCD may have a different clinical expression in primarily OCD and primarily BD populations. In this study, we examined the clinical characteristics of patients with a primary diagnosis of OCD with and without comorbid BD. Based on previous literature, we hypothesized that OCD when comorbid with BD runs an episodic course but we did not have any a priori hypotheses about other potential group differences. A primary diagnosis of OCD was made if OCD was the main reason for consultation.

METHODS

Sample

The study was approved by the Ethics Committee of the National Institute of Mental Health and Neuro Sciences (NIMHANS), Bangalore, India and conforms to the provisions of the Declaration of Helsinki in 1995. All participants gave written informed consent voluntarily. We recruited subjects from the specialty OCD clinic of the NIMHANS, Bangalore, India, from May 2006 to May 2007. All consecutive adult patients attending the OCD Clinic with a primary diagnosis of DSM-IV OCD and who had a lifetime diagnosis of either DSM-IV bipolar episode or iatrogenic bipolar episode (drug- and electroconvulsive therapy [ECT]-induced) were invited to participate in the study (n = 39). The diagnosis of BD had been made by the consultants of the clinic according to a detailed interview of patients and their immediate family members who live with them. This was further confirmed at the time of the study by detailed assessments as mentioned in the following section.

All the 39 OCD patients had received a diagnosis of BD before entry into the study. None of them were in a bipolar episode at the time of assessment. In all patients, OCD predated the bipolar episode but three of them reported an episode of major depression before onset of OCD. Of the 39 patients, five could not be recruited because they refused to give consent. BD included bipolar I (n = 11, 32%), bipolar II (n = 10, 29%), and cyclothymia (n = 1, 3%) according to DSM-IV, and drug- (n = 11, 32%), and ECT-induced switch (n = 1, 3%). We included drug/ECT-induced hypomania/mania under BD to have a broader and inclusive definition of bipolarity since propensity for antidepressant-induced switch may imply underlying diathesis to bipolarity. Previous researchers who studied the effect of bipolar disorders on OCD have included antidepressant-induced hypomania/mania under the bipolar disorder category.12

To compare OCD subjects with BD with those without BD, we approached 79 consecutive adult subjects with a primary diagnosis of DSM-IV OCD without BD attending the specialty OCD clinic over a 3-month period (July–September 2006). We could not assess 22 of the 79 subjects (28%) because of their inability to come for assessment or unwillingness to participate in the study. However, 22 subjects who could not be assessed did not differ from the 57 assessed subjects with respect to sociodemographic and baseline clinical characteristics.

Assessment

Initial diagnosis of OCD was made according to the DSM-IV criteria after detailed clinical assessment by at least two clinicians, one of whom was a senior consulting psychiatrist of the team (YCJR/BMS). The assessment included obtaining information as per the topical proforma specially developed for assessing OCD subjects, which focuses on symptom profile, comorbidity, family history and treatment history. The baseline clinical assessment included the Yale–Brown Obsessive Compulsive scale (YBOCS)26 symptom checklist and severity rating scale and the Clinical Global Impression-severity (CGI-S).27

All the subjects were evaluated further using the following instruments by the principal author: the Structured Clinical Interview for Axis I (SCID-I)28 and Axis II (SCID-II) DSM-IV disorders;29 the YBOCS symptom checklist, severity rating scale and YBOCS item 11 for insight;26 the Hamilton Depression Rating Scale-24 item (HDRS);30 the Scale for Suicidal Ideation (SSI);31 the Clinical Global Impression for severity and improvement (CGI-I & S);27 the Tics and Tourette's syndrome section of the Schedule for Tourette and Behavioral Syndromes;32 the Family Interview for Genetics Study (FIGS);33 and the Global Assessment of Functioning Scale (GAFS).34 The SSI is a 19-item clinician-administered instrument during the course of a semi-structured interview. It measures both the current (in the past 1 week of assessment) and the worst-ever suicidal ideation (a report of the suicidal ideation at any time in life that the patient reports to be the worst). The SSI-worst-ever includes either current or past, whichever is worst. For example, many may not have had suicidal ideas in the past but could well rate the current as their worst. Some could have had suicidal ideation both currently and previously but had to choose one of them as worst. Essentially, it means worst-ever lifetime suicidal ideation.

The time from the baseline assessment is shown in Table 1. The YBOCS and the CGI-S were the only instruments administered at the time of initial clinical evaluation. The remaining instruments were administered at the time of inclusion in the study.

Table 1.  Sociodemographic and clinical profile
 OCD with BD (n = 34) Mean (SD)OCD without BD (n = 57) Mean (SD)M–W UP-value
Age in years28.39 (7.10)29.36 (8.31)††934.5000.777
Years of education12.50 (2.92)13.01 (2.60)922.0000.692
Age of onset in years18.00 (6.24)19.57 (7.18)874.5000.437
Duration of OCD, months105.80 (80.98)109.75 (88.01)877.0000.935
Time since baseline assessment, months43.85 (33.04)41.98 (24.16)963.50.964
 n (%)n (%)χ2P
  • Mann–Whitney U-test.

  • χ2-test.

  • §

    Fisher's exact test.

  • Age varied from 19 to 45 years (range = 26).

  • ††

    Age varied from 19 to 55 years (range = 36).

  • BD, bipolar disorder; M–W U, Mann–Whitney U-test; OCD, obsessive–compulsive disorder.

Sex  0.6490.421
 Male23 (67)43 (75)
 Female11 (32)14 (25)
Marital status  5.9660.051
 Married19 (56)20 (35)
 Single14 (41)37 (65)
 Divorced1 (3)0 (0)
Background  0.0350.851
 Rural15 (44)24 (42)
 Urban19 (56)33 (58)
Monthly income, in rupees  2.4660.291
 <15005 (15)5 (9)
 1500–50008 (24)22 (39)
 >500021 (62)30 (53)
Occupation  6.8200.146
 Student7 (21)16 (28)
 Home-maker9 (27)12 (21)
 Employed4 (12)17 (30)
 Self-employed/business/agriculture11 (33)9 (16)
 Unemployed3 (9)3 (5)
Onset (early vs late)  0.0070.935
 Early (≤18 years)17 (50)29 (51)
 Late (>18 years)17 (50)28 (49)
Family history    
 Any psychiatric illness23 (68)30 (53)1.9740.160
 OCD6 (18)13 (23)0.3430.558
 Other anxiety disorder3 (9)2 (45)0.358§
 Depression9 (27)9 (16)1.5310.216
 BD1 (3)1 (2)1.000§
 Psychosis8 (24)7 (12)1.9570.162
 Substance use3 (9)4 (7)0.754§
 Suicide7 (21)10 (18)0.1300.719

OCD subjects with BD were assessed when they were not in a bipolar episode. However, five subjects (15%) in the OCD+BD group and four (7%) in the OCD group were in an episode of depression at the time of assessment.

We determined the course of OCD employing the following definitions that were used in previous studies.3,35–37 The definitions were modified to an extent by elaborating the terms ‘remission’ and ‘relapse’ as follows.

  • 1Chronic: Symptoms persisted for most of the course, causing significant distress and impairment in functioning.
  • 2Episodic: (i) History of remission and relapse; (ii) during remission, no OCD or at least subclinical course for ≥3 months; and (iii) during relapse, YBOCS score should be ≥15 for ≥3 months or significant distress and impairment in functioning should have been noted (severe enough to seek treatment or change of treatment).
  • 3Subclinical: Mild symptoms that did not cause any significant distress or impairment in functioning for most of the course. The subclinical state should persist for more than 3 months.
  • 4Recovered: No symptoms after recovery from the index episode of OCD, with no relapses.

We had access to the clinical charts of all the patients. In addition, information about the subjects' illness and the course was also obtained from their immediate family members. Life charting of the course of OCD and BD was made for each subject based on the information available from all the possible sources.38 The senior consultant (YCJR) expert in assessing OCD and BD reviewed the data obtained from all the sources, and the final opinions were arrived at by the consensus of the senior consultant and the first author.

Statistical analysis

The Kolmogorov–Smirnov test was used to determine the normality of the distribution. Continuous variables were compared using the Mann–Whitney U-Test because most of the variables were non-normatively distributed. The χ2-test/Fisher's exact test was used to compare categorical variables. To determine if worsening/improvement of OCD during mood episodes was statistically significant in the OCD with BD group, we used the single-sample χ2-test. In view of explorative nature of the study, no corrections were applied and a P < 0.05 was considered statistically significant. spss 13 (Chicago, IL, USA) was used for the analysis.

RESULTS

Our study did not reveal any difference between the groups in terms of age, sex, age at onset of OCD, mode of onset, duration of illness, and family history of psychiatric illnesses (Table 1).

Episodic course (53%) was overrepresented in the OCD with BD group whereas OCD without BD tended to have either a continuous (40%) or subclinical (32%) course (Table 2). In OCD with BD, worsening of OCD was seen in 21 (72%) of the 29 subjects during depressive episodes (χ2 = 5.828, P = 0.016); the remaining five patients did not have depression. Improvement in OCD was noted in 24 (71%) subjects during mania/hypomania (χ2 = 5.765, P = 0.016). Worsening in hypomanic/manic episodes was noted in only two (6%) of the bipolar subjects. Half of the subjects (50%) achieved complete remission of OC symptoms during hypomania or mania, but they relapsed following remission of mania/hypomania. The OCD with BD group had significantly less-frequent lifetime pathological doubts. Pathological slowness, reassurance-seeking and other miscellaneous compulsions were significantly higher in the OCD with BD group (Table 2).

Table 2.  Symptom profile and course of OCD with (n = 34) and without BD (n = 57)
 OCD with BD n (%)OCD without BD n (%)χ2P-value
Course of OCD  21.490<0.001
 Continuous12 (35)23 (40)
 Episodic18 (53)7 (12)
 Subclinical2 (6)18 (32)
 Recovered2 (6)9 (16)
Subtype of OCD  0.7550.686
 Predominantly obsessions5 (15)10 (18)
 Predominantly compulsions0 (0)1 (2)
 Mixed type29 (85)46 (81)
Obsessions, lifetime    
 Contamination20 (59)34 (60)0.0060.938
 Aggression19 (56)29 (51)0.2140.664
 Sexual13 (38)19 (33)0.2240.656
 Religious10 (29)22 (39)0.7880.375
 Hoarding4 (12)4 (7)0.466§
 Pathological doubt17 (50)43 (75)6.1360.013
 Need for symmetry9 (26)16 (28)0.0270.869
 Other obsessions14 (41)23 (40)0.0060.938
Compulsions, lifetime    
 Washing18 (53)31 (54)0.0180.894
 Checking14 (41)32 (56)1.9080.167
 Repeating and counting12 (35)31 (54)3.1140.078
 Collecting1 (3)1 (2)1.000§
 Ordering7 (21)14 (25)0.1890.663
Other compulsions, lifetime    
 Miscellaneous25 (74)34 (60)1.8000.180
 Mental rituals15 (44)24 (42)0.0350.851
 Excessive list-making2 (6)4 (7)1.00§
 Pathological slowness11 (32)5 (9)8.1720.004
 Need to confess/tell/ask8 (24)4 (7)0.050§
 Need to touch/tap/rub4 (12)3 (5)0.418§
 Superstitious behaviors5 (15)6 (11)0.741§
 Reassurance seeking11 (33)6 (11)6.6790.010
 Self-damaging behavior3 (9)1 (2)0.145§
 Blinking/staring rituals2 (6)0 (0)0.137§
 Others8 (24)3 (5)0.017§
 Mean (SD)Mean (SD) M–W UP
  • M–W U.

  • χ2-test.

  • §

    Fisher's exact test.

  • BD, bipolar disorder; M–W U, Mann–Whitney U-test; OCD, obsessive–compulsive disorder.

Total number of lifetime obsessions4.73 (2.93) 4.52 (2.86) 946.0000.849
Total number of lifetime compulsions4.35 (4.04) 3.54 (2.14) 960.5000.940

Severity of illness in those with and without BD is shown in Table 3. Baseline YBOCS scores were similar in both the groups, but the current compulsion score was somewhat higher in the OCD with BD group. Insight was found to be significantly poorer among OCD with BD subjects. The current total score on the SSI and frequency of attempters (10.5% vs 23.5%; P = 0.096) were similar in both the groups, but the worst-ever total score on SSI was significantly higher in the OCD with BD group.

Table 3.  Measures of severity of illness of OCD with (n = 34) and without BD (n = 57)
 OCD with BD Mean (SD)OCD without BD Mean (SD)M–W UP-value
  • M–W U. Bold type indicates statistical significance.

  • BD, bipolar disorder; CGI-S and -I, Clinical Global Impression for severity and improvement; GAFS, Global Assessment of Functioning Scale; HDRS, Hamilton Depression Rating Scale; M–W U, Mann–Whitney U-test; OCD, obsessive–compulsive disorder; SSI, Scale for Suicidal Ideation; YBOCS, Yale–Brown Obsessive Compulsive scale.

YBOCS score, current    
 Obsessions8.14 (6.46)6.12 (4.96)802.0000.167
 Compulsions6.91 (5.83)3.92 (4.72)675.5000.013
 Total15.05 (11.93)10.07 (8.44)744.0000.063
YBOCS score, baseline    
 Obsessions14.20 (3.33)12.91 (3.04)789.5000.138
 Compulsions11.11 (5.08)10.54 (4.92)898.5000.562
 Total25.44 (7.48)23.45 (7.11)802.0000.170
YBOCS-11 (Insight)1.11 (1.00)0.52 (0.71)642.5000.004
CGI-S    
 Baseline4.64 (0.84)4.36 (0.81)766.5000.073
 Current2.94 (1.84)2.21 (1.26)786.0000.120
CGI-I2.35 (1.45)1.94 (0.98)864.0000.364
GAFS65.94 (18.68)71.64 (16.18)801.5000.168
Total SSI score, Current3.67 (5.45)1.84 (3.12)781.0000.087
Total SSI score, Worst ever12.76 (8.38)6.91 (6.29)549.0000.001
Current HDRS5.52 (6.89)2.49 (4.11)693.5000.014

The total number of depressive episodes, severe depressive episodes and episodes of psychotic depression was significantly higher in the OCD with BD group (Table 4). However, the OCD with BD group had a significantly lower prevalence of any lifetime axis-I comorbidity and number of lifetime anxiety disorders. The number of trials with antidepressants and augmenting agents were similar in both the groups but the use of ECT and hospitalization was greater in the OCD with BD group (Table 5). In the OCD+BD group, 11 (32%) were hospitalized for mania and 13 (38%) for severe OCD, whereas in the OCD without BD group, eight subjects (14%) were hospitalized for severe OCD. Patients with bipolar disorder received mood stabilizers for bipolar disorder (Table 5). Serum lithium levels were in the therapeutic range in those subjects who received lithium (0.6 to 1 meq/L). Serum levels of valproic acid and carbamazepine could not be measured because of lack of laboratory facility. With treatment, all subjects remitted from their mood episodes, but whether mood stabilizers had significant impact on the course of bipolar disorder could not ascertained.

Table 4.  Comorbidity of OCD with (n = 34) and without BD (n = 57)
 OCD with BD n (%)OCD without BD n (%)χ2 valueP-value
Any axis I comorbidity, lifetime11 (32)42 (74)14.959<0.001
Any anxiety disorder, lifetime10 (29)26 (46)2.3380.126
Major depressive episode    
 Current episode5 (15)4 (7)1.4130.235
 Past28 (82)31 (54)7.3060.007
Dysthymia5 (15)11 (19)0.3100.578
Alcohol abuse2 (6)3 (5)1.000§
Panic disorder3 (9)4 (7)1.000§
Panic with agoraphobia1 (3)0 (0)0.374§
Agoraphobia without panic disorder0 (0)1 (2)1.000§
Social phobia7 (21)19 (33)1.6950.235
Generalized anxiety disorder2 (6)11 (19)0.121§
Chronic motor/vocal tic disorder1 (3)3 (5)1.000§
Any tic disorder1 (3)3 (5)1.000§
Any PD9 (27)20 (35)0.7280.488
Avoidant PD5 (15)9 (16)0.0190.890
Dependent PD1 (3)3 (5)1.000§
Obsessive–compulsive PD2 (6)9 (16)0.206§
Passive–aggressive PD0 (0)1 (2)1.000§
Paranoid PD0 (0)2 (4)0.527§
Histrionic PD1 (3)0 (0)0.370§
Narcissistic PD0 (0)1 (2)1.000§
Borderline PD2 (6)1 (2)0.553§
 Mean (SD) Mean (SD) M–W UP-value
  • M–W U. Bold type indicates statistical significance.

  • χ2-test. Bold type indicates statistical significance.

  • §

    Fisher's exact test.

  • BD, bipolar disorder; M–W U, Mann–Whitney U-test; OCD, obsessive–compulsive disorder; PD, personality disorder.

Number of axis I comorbid disorders0.35 (0.54) 1.29 (1.05) 445.400<0.001
Number of depressive episodes2.67 (2.40)1.71 (1.11)611.5000.002
Number of severe depressive episodes1.64 (1.33)1.44 (0.88)717.5000.007
Number of depressive episodes with psychotic symptoms0.32 (0.84)0.01 (0.13)813.5000.006
Number of anxiety disorders0.29 (0.52)0.61 (0.77)760.0000.047
Table 5.  Treatment profile of OCD with (n = 34) and without BD (n = 57)
 OCD with BD n (%)/Mean (SD), mgOCD without BD n (%)/Mean (SD), mgM–W UP-value
SSRI/TCA   
 Fluoxetine 18 (53)/57.91 (22.45)38 (67)/67 (17.27)3710.102
 Sertraline11 (32)/173.33 (45.77)20 (35)/197.61 (29.47)1100.045
 Fluvoxamine5 (15)/207 (104.56)6 (11)/191.66 (82.11)53.500.656
 Citalopram9 (27)/56 (20.65)11 (19)/53.63 (21.10)500.708
 Escitalopram19 (56)/25.23 (9.82)19 (33)/25.47 (7.40)206.500.705
 Paroxetine3 (9)/53.57 (26.88)4 (7)/57.91 (24.51)200.884
 Clomipramine5 (15)/145.83 (55.71)10 (18)/190.90 (51.56)180.113
Mood stabilizers    
 Lithium21 (62)/914.30 (141.54)
 Valproate13 (38)/1134.61 (332.53)
 Carbamazepine6 (18)/675 (194.30)
 Mean (SD)Mean (SD) M–W UP-value
Number of adequate trials with SSRI/TCA2.23 (1.41) 1.94 (1) 862.5000.359
Number of augmentations >2 months0.88 (1)0.61 (0.75)838.0000.242
Total number of hospitalizations1.55 (1.70)0.16 (0.41)381.5<0.001
 n (%) n (%) χ2P-value
  • M–W U. Bold type indicates statistical significance.

  • χ2-test. Bold type indicates statistical significance.

  • §

    Fisher's exact test.

  • Mean (SD) number of mood stabilizers used = 1.21 (0.59).

  • BD, bipolar disorder; ECT, electroconvulsive therapy; M–W U, Mann–Whitney U-test; OCD, obsessive–compulsive disorder; SSRI, selective serotonin reuptake inhibiter; TCA, tricyclic antidepressants.

Behavior therapy9 (27)13 (23)0.1560.693
ECT7 (21)2 (4)0.012§
Any hospitalization24 (71)8 (14)29.875<0.001

DISCUSSION

Our finding that episodic course is typical of OCD when comorbid with BD is corroborated by the results of previous studies, which reported episodic course as a characteristic feature of OCD comorbid with BD.3,4,11,16,17,39,40 The prevalence of episodic course in our study is similar to that reported by Perugi and colleagues (43–53%) in OCD cohorts,3,11 although it is lower than that reported in our previous study of a cohort of primarily bipolar subjects (75%).16 Higher rates of episodic course in our previous study could be due to the nature of the sample. The samples in the studies by Perugi and colleagues3 and the current study were primarily OCD cohorts, and therefore, similar rates are understandable.

The findings of our study and those of previous studies suggest that OCD tends to run an episodic course when associated with bipolar disorder irrespective of whether the cohorts are primarily bipolar or not. Worsening of OCD in depressed phases and improvement in hypomanic/manic phases is consistent with similar observations made previously3,16 and lends support to the argument that comorbid bipolar disorder influences the course of OCD in a distinct manner. Episodic course suggests a pathoplastic effect of bipolar disorder on OCD. It is also possible that both the disorders perhaps share a common pathophysiological basis. Affective dysregulation has been hypothesized as a basic abnormality in both OCD and bipolar disorder.41

The OCD with BD group had less-frequent pathological doubts and a higher rate of miscellaneous compulsions. Previous studies found higher rates of religious and sexual obsessions3 and obsessions with symmetry,19 as well as an inconsistent association with ordering/arranging compulsions and checking rituals3,12,19 in OCD with BD. Miscellaneous obsessions, such as philosophical, existential, bizarre, and superstitious beliefs, were more common in the bipolar OCD group in a study by Masi et al.12 and higher rates of reassurance-seeking compulsions were found in a study by Hantouche et al.,39 which are in line with our study findings. Lack of consistent association between the type of OC symptoms and bipolar comorbidity could be due to differences in the sample characteristics across studies. It is also possible that bipolarity does not have a specific effect on the phenomenology of OC symptoms. However, the presence of more miscellaneous compulsions could also generate and support another hypothesis, that a particular group of OCD cases could be an atypical expression of BD, as suggested by previous researchers.42,43

Insight was poorer in the OCD with BD group, which is in contrast to a previous study by Tukel et al.44 Most of the other studies have not evaluated insight in OCD with BD subjects. Systematic assessment of insight across phases of bipolar disorder in a prospective design may shed more light on the correlation between insight and bipolarity.

A greater number of depressive episodes in OCD with BD replicate the findings of previous studies.3,11 Understandably this may be due to the comorbid bipolar disorder. However, overall comorbidity was significantly lower in the OCD with BD group. Previous studies found higher rates of comorbid anxiety disorders2,3,11,12,16 and substance abuse1,3 in bipolar OCD subjects but such an association was not found in our study. Substance use is not a common comorbidity in the Indian OCD population as evident from previous studies.31,45–48 A previous study found higher rates of narcissistic and antisocial personality disorders in OCD with BD,22 whereas our study did not find any difference between both the groups. Variations in the pattern and rate of comorbid disorders across studies could be related to differing sample characteristics; for example some were primarily OCD cohorts (Perugi et al.3 and this study), others were primarily bipolar cohorts2,16 and one study included mainly children.12

Data on suicidality in OCD patients is sparse.46 Subjects with OCD and BD had higher worst-ever scores on the SSI, confirming the findings of previous studies that have noted higher suicidality in OCD with BD group.1,2,21,39 Therefore, it can be suggested that bipolar comorbidity increases the risk of suicide in OCD subjects. Higher suicidality could well be because of higher rates of lifetime depression in OCD with BD.

There was no difference in family history of any psychiatric illness between both the groups. This is similar to findings of the study by Perugi et al.3 but some studies have shown higher rates of OCD and mood disorder in relatives of bipolar OCD subjects.1,16,39 This difference in results can be attributed to the types of study populations, such as a community sample,1 cyclothymic subjects39 and primary bipolar subjects.16 We used the family history method, which may not be sensitive to diagnose minor bipolar disorders and OCD in relatives.

The strengths of this study are the systematic assessments and use of multiple sources of information, such as structured instruments, relatives and hospital files. The study has certain limitations. It is a cross-sectional study (with retrospective assessment of the course). The study had a relatively small sample of subjects with bipolar disorder and this limits certain interpretations. For example, we did not find significant differences between groups with respect to family loading for mood and psychotic disorders and personality disorders. This may be due to small sample sizes causing type 2 error. We employed a broader definition of bipolarity to include all forms of bipolar episodes, including iatrogenic bipolarity. Whether there are any differences in OCD phenotype based on subtype of comorbid bipolar disorder is not clear in view of the small sample size. Our sample was also somewhat different from those in previous studies3,4,20–22,39 in that we had almost equal representation of bipolar I and II subjects. This is possibly because of our decision to include all OCD+bipolar subjects a priori, unlike the previous studies, which evaluated all consecutive OCD patients prospectively for bipolarity and then compared those with and without bipolar disorder. In such samples, bipolar II has been overrepresented. We used mood stabilizers to treat comorbid bipolar disorder but their effect on the course of bipolar disorder could not be determined because of the cross-sectional nature of the study. Co-administration of antidepressants, particularly selective serotonin reuptake inhibiters and anticonvulsant mood stabilizers, may have resulted in complex drug interactions because of inhibition and induction of cytochrome P450 isozymes by selective serotonin reuptake inhibiters and certain anticonvulsants, respectively.49 The effect of combining these drugs on side-effects, tolerability and efficacy needs to be studied systematically. Finally, our sample was recruited from the specialty OCD clinic of a tertiary care hospital and this may restrict the generalizability of our results.

To conclude, our study suggests that OCD with BD differs from OCD without BD with respect to several clinical characteristics. The episodic nature of OCD is the most well-replicated finding across studies, irrespective of the type of cohort. This implies that bipolarity has a pathoplastic effect on OCD and it is possible that some forms of OCD and BD are pathophysiologically related. Whether all anxiety disorders share something in common with BD or they have a separate and unique correlation with BD is not clear now. This aspect is important because it is increasingly recognized that OCD is distinct from other anxiety disorders and that it belongs to a larger group of OCD spectrum disorders.50 That OCD with BD is associated with higher rates of depressive episodes, higher suicidality and a greater number of hospitalizations suggests greater morbidity. Studies examining the effect of mood stabilizers not only on the course of comorbid bipolar disorder but also on the course of OCD are needed. Prospective follow-up studies of the course and outcome and studies addressing pathophysiology and genetic basis are needed to understand the complexities of bipolar–OCD comorbidity.

ACKNOWLEDGMENT

The authors thank Prof Jagadisha Theerthalli, NIMHANS, Bangalore for his valuable help in the statistical analysis.

Ancillary