Journal of Affective Disorders 147 (2013) 217–224

Contents lists available at SciVerse ScienceDirect

Journal of Affective Disorders

0165-03

http://d

n Corr

E-m

journal homepage: www.elsevier.com/locate/jad

Research report

Screening for bipolar disorder among patients undergoing a majordepressive episode: Report from the BRIDGE study in Egypt

Tarek Okasha a,n, Mohamed Fikry a, Aref Kowailed b, Tamer El-Guwiely b, Hisham Sadek a

a Okasha Institute of Psychiatry, Faculty of Medicine, Ain Shams University, Cairo, Egyptb Department of Psychiatry, Faculty of Medicine, Cairo University, Cairo, Egypt

a r t i c l e i n f o

Article history:

Received 20 April 2012

Received in revised form

30 October 2012

Accepted 1 November 2012Available online 26 November 2012

Keywords:

Bipolar disorder

Major depression

Egypt

Incidence

Screening

27/$ - see front matter & 2012 Elsevier B.V. A

x.doi.org/10.1016/j.jad.2012.11.007

esponding author. Tel.: þ202 29200900/1/2/

ail address: tokasha@internetegypt.com (T. O

a b s t r a c t

Background: To estimate the frequency of bipolar disorder (BPD) among patients with a major

depressive episode (MDE) and elucidate clinically-relevant factors predictive of bipolarity.

Methods: We evaluated 306 patients undergoing a MDE at facilities throughout Egypt. Patients were

given the HCL-32 R2 questionnaire to assess the presence of manic/hypomanic symptoms; those

scoring 414 were considered bipolar. We also investigated how various clinical criteria for bipolarity

changed the incidence of bipolar diagnosis. Finally, we examined if demographics, psychiatric history,

clinical characteristics, and the incidence of co-morbid conditions differed significantly between bipolar

and unipolar patients.

Results: The positive screen rate for BPD based on HCL-32 R2 scores was 62.2% (188/302). However,

only 26% (80/306) of patients had been diagnosed previously as bipolar. In contrast, when DSM-IV

criteria were used, only 13.7% (42/306) of patients qualified as bipolar. A number of factors were highly

predictive of bipolarity including: seasonality, number of past mood episodes, history of psychiatric

hospitalization, mixed state, and mood reactivity. Of the comorbidities examined, only borderline

personality disorder occurred at a higher rate in bipolar than in unipolar patients.

Limitations: Participating centers were not randomly selected and there could be a bias if only

psychiatrists having specific interest in BPD were included.

Conclusions: The positive HCL-32-R2-based bipolar screen rate of 62% suggests that a substantial

proportion of patients with a MDE may have BPD. Further, a number of factors in the patient’s

psychiatric history as well as clinical aspects of the episode itself may signal an increased likelihood of

bipolarity.

& 2012 Elsevier B.V. All rights reserved.

1. Introduction

Clinical depression is not only one of the most common mentaldisorders, affecting an estimated 121 million people worldwide(Cutcliffe and Lakeman, 2010), but it is also expected to become thechief cause of premature mortality and disability adjusted for lifeyears (DALYs) by the year 2030 (World_Health_Organization,2004). Persons with depression have shorter life expectanciesand are at a higher risk for suicide than their unaffected peers(Davies et al., 2001). Though depression can be successfullymanaged, only an estimated 25% of persons suffering from it haveaccess to effective treatment options (Gonzalez et al., 2010).

Major depressive disorder (also called unipolar depression) istypified by loss of pleasure or interest in normally enjoyableactivities and feelings of low self-worth. It often presents with

ll rights reserved.

3/4; fax: þ 202 29200908.

kasha).

changes in appetite, energy, sleep, as well as with difficulties inconcentrating symptoms which can interfere greatly with daily life.In a major depressive episode, one experiences a severely depressedmood for at least 2 weeks (American_Psychiatric_Association, 1994).However, if the person also experiences a markedly elevated mood(referred to as mania or hypomania if less severe), they may havebipolar disorder (BPD) rather than unipolar depression. Manicsymptoms include extreme irritability, racing thoughts, short atten-tion span, decreased need for sleep, increased energy, and impairedjudgment (American_Psychiatric_Association, 1994). When a patientexperiences manic symptoms alternating with their depressiveepisodes, they are usually diagnosed bipolar I (BP1), while thoseexperiencing less severe, hypomanic episodes usually are diagnosedbipolar II (BP2).

A diagnosis of BPD can be difficult to make conclusively evenfor the most experienced medical professionals. Bipolar patientspresenting with a major depressive episode may be diagnosed withunipolar depression if manic/hypomanic characteristics are notapparent. Commonly, patients may experience different degrees of

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224218

mania for different periods of time. Additionally, BPD is frequentlyassociated with other co-morbid psychiatric conditions which maypresent first, further complicating diagnosis (Keck et al., 2001;Kessler et al., 1999). For these reasons, as many as 40% of bipolarpatients get another first diagnosis, meaning it can take yearsbefore they are correctly diagnosed (Ghaemi et al., 2002). Thus, it isreasonable to theorize that a significant proportion of patientsdiagnosed with depression actually may have an undiagnosed BPD.

In the current study, we investigated the prevalence of BPDamong persons being treated for a major depressive episode. Ourmain goal was to estimate the frequency of BPDs in this patientpopulation and gain a better understanding of how frequentlyBPD presents as depression. Our secondary objectives were to:(1) examine the prevalence of various comorbidities in bipolarand unipolar patients, (2) identify the demographic and clinicalfeatures of patients with and without BPD, and (3) identifyhypomanic/manic components in patients with major depression.Our findings shed light on the magnitude of the public healthissue created by failing to diagnose a significant percentage ofbipolar cases.

2. Methods

2.1. Patients

Between August 2008 and October 2008, patients seekingtreatment for depressive symptoms were recruited for this studyduring a single visit to their local psychiatrists. Each psychiatriccenter (19 in total) was responsible for recruiting at least 10 patientsundergoing a major depressive episode. These mental health facil-ities represented different geographical areas of the country andwere reflective of the local psychiatric practice and patient manage-ment care. All subjects provided written informed consent, wereat least 18 years of age, and met the DSM-IV diagnostic criteriafor major depressive disorder (American_Psychiatric_Association,1994). As per this criteria the patient: (i) exhibits at least one ofthree abnormal mood criteria (abnormal depressed mood, loss ofinterest/pleasure, or irritable mood) which significantly impacts thepatient’s life nearly every day for at least two weeks, (ii) exhibits atleast five of nine characteristic depressive symptoms (depressedmood, diminished interest/pleasure, weight loss/decreased appetite,insomnia/hypersomnia, psychomotor agitation/retardation, fatigue/loss of energy, feelings of worthlessness/guilt, diminished concen-tration, recurrent thoughts of death), and (iii) has symptoms thatare not attributable to another cause, such as physical illness orsubstance abuse.

Patient and investigator data were treated in compliance withall local applicable laws and regulations. Appropriate measureswere taken to safeguard data confidentiality. Patients presentingwith an acute non-psychiatric condition, an emergency event, orthose unable to complete the HCL-32 R2 questionnaire were notincluded in this study. In total, 313 patients met these criteria andwere enrolled in the study. Of these, seven were not assessed forbipolarity according to DSM-IV criteria and were excluded fromfurther analysis, leaving a final sample number of 306 patients.

2.2. Study design

The current analysis is part of an international, multi-center,cross-sectional study in 18 countries known as the BRIDGEscreening program (Angst et al., 2011, 2012). The goal of thisscreening initiative was to measure the frequency of BPD amongpatients who have had at least one major depressive episodeand to elucidate the impact of various psychiatric characteristicsand comorbidities in this patient population. This program was

developed to test the hypotheses that: (i) patients with majordepression are more likely to be positively identified as bipolarthan the general population, (ii) patients with major depressionwho have co-morbid conditions are more likely to have BPD thanthose without these comorbidities, and (iii) bipolar patients havea higher number of comorbidities than those who are unipolar.

2.3. Reported measures

In a single visit, eligible patients responded to the HCL-32 R2Hypomania Self-Rating Scale questionnaire (Angst et al., 2005)and their physician completed the case report form (CRF). TheHCL-32 R2 is a screening tool for diagnosing BPD which assessesthe presence of various manic/hypomanic symptoms such as:changes in mood, level of energy, sleep, sex drive, sociability,willingness to take risks, talkativeness, and distractibility (Angstet al., 2005).

The CRF included socio-demographic factors, family andpatient psychiatric history, clinical characteristics of the depres-sive episode, presence of co-morbid conditions, global assessmentof functioning, current treatment(s), and illness progression.

In addition to the HCL-32 R2, the frequency of BPD in patientspresenting with a major depressive episode was determined bythree other diagnostic criteria namely, DSM-IV (American_Psychiatric_Association, 1994), modified DSM-IV (Angst et al.,2012), and other definitions of bipolarity (Angst et al., 2003a, b).Modified DSM-IV criteria factors into account specifically,(i) increased activity as a gate question, (ii) 1 or 2 to 3-dayhypomanic episodes, and (iii) the elimination of all exclusioncriteria in DSM-IV (Angst et al., 2012). Other definitions ofbipolarity takes into account the so-called softer expressions ofbipolarity which might otherwise be overlooked on the bipolarspectrum (Angst et al., 2003a, b).

2.4. Statistical analyses

All statistical analyses were performed at 5% significancecut-off using two-sided tests or two-sided confidence intervals.Population characteristics are summarized into count of non-missing data, mean, standard deviation, minimum, maximum,media, 95% confidence interval of the mean for quantitativevariables and count, and percentage with 95% confidence intervalof the population for categorical data. The frequency of BPD isprovided with 95% confidence interval and the presence/absenceof each co-morbid diagnosis. Determination of confidence inter-vals takes into account the cluster design effect. A sensitivityanalysis based on the heterogeneous ration of participatingpatients between centers was performed. Thus, different frequenciesof BPD are attributed to the non-participating patients, e.g. theobserved frequency � 2, or the observed frequency divided by 2.

Comparison of demographic, clinical, and co-morbid condi-tions in patients with and without BPD was performed usinganalysis of variance or chi-square test, according to the nature ofthe variable. Data regarding treatment received by patients withand without BPD were performed using the chi-square test.Bipolar disorder is presented as a function of population char-acteristics, thus the association between BPD and comorbidities,demographics, and clinical variables, was done using logisticregression models and odds ratio. This resulted in five differentunivariate analyses, which assessed bipolarity according to:(1) the HCL-32 R2 scale (Angst et al., 2005), (2) the DSM-IVcriteria (American_Psychiatric_Association, 1994), (3) a modifieddefinition of DSM-IV (Angst et al., 2012), (4) other definition ofbipolarity (Angst et al., 2003a, b), and (5) known bipolar I or II(BP1/2) status.

Table 2Patient psychiatric history.

Details of psychiatric history Value

Age at first psychiatric symptom (n¼304)

o15 14 (4.61)

15–19 42 (13.82)

20–29 142 (46.71)

30–39 61 (20.07)

40–49 28 (9.21)

50–59 14 (4.61)

Z60 3 (0.99)

Age at first setting of depressive diagnosis (n¼299)

o15 8 (2.68)

15–19 31 (10.37)

20–29 137 (45.82)

30–39 71 (23.75)

40–49 32 (10.70)

50–59 16 (5.35)

Z60 4 (1.34)

Postpartum depression (females; n¼150) 20 (13.33)

History of suicide attempts (n¼306) 59 (19.28)

Seasonality of mood episodes (n¼304) 46 (15.13)

Total number of mood episodes in the past (n¼305)

0 68 (22.30)

1 44 (14.43)

2–3 87 (28.52)

4–5 37 (12.13)

More 69 (22.62)

Total number of mood episodes in previous year (n¼306)

0 142 (46.41)

1 107 (34.97)

2–3 43 (14.05)

4–5 3 (0.98)

More 11 (3.59)

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224 219

3. Results

3.1. Demographics of the patient population

Of the 306 patients suffering a major depressive episode whowere recruited for this study, the majority (85.3%) were treatedon an out-patient basis. The subjects’ age ranged from 18 to 76years, and the average age was 37 years. Marital status wasdistributed fairly evenly; with 43.8% of patients married and41.2% single (Table 1). Of the nine symptoms of depressionassessed in this study, all patients exhibited at least five, with28.1%, 31.0%, and 21.6% exhibiting six, seven, and eight symp-toms, respectively (Table 1). The vast majority (98.7%) reporteddepressed mood, diminished interest of pleasure (94.4%), andeither insomnia or hypersomnia (90.5%) (Table 1).

3.2. Patient psychiatric history, co-morbid conditions, and current

treatments

Analysis of patient psychiatric histories revealed that 31/306(10.1%) had been diagnosed previously as BP1 and 49/306 (16%)as BP2, leaving 226/306 (73.9%) lacking a bipolar diagnosis.The most common age for psychiatric symptoms to be firstdiagnosed was between 20 and 29 years (Table 2). The majorityof patients (80.7%) had no history of attempted suicide and 68.3%had no previous psychiatric hospitalizations (Table 2). Most (77.7%)had at least one mood episode previously and 53.6% had at leastone episode in the last year (Table 2). When patients were scoredusing the global assessment of functioning scale (American_Psychiatric_Association, 1994), most fell into either the ‘‘severesymptoms’’ (31.8%) or ‘‘moderate symptoms’’ (25.9%) categories

Table 1Patient demographic information (N¼306).

Characteristics Value

Male, n (%) 150 (49.02)

Age in years, mean 7 SD 37.72 7 12.78

Age in years, n (%)

o20 4 (1.31)

20–29 93 (30.39)

30–39 87 (28.43)

40–49 61 (19.93)

50–59 40 (13.07)

60–69 16 (5.23)

70–79 5 (1.63)

Z80 0 (0.00)

Marital status, n (%)

Married 134 (43.79)

Divorced 36 (11.76)

Single 126 (41.18)

Other 10 (3.27)

Symptoms of depression, n (%)

Depressed mood 302 (98.69)

Markedly diminished interest of pleasure 289 (94.44)

Significant weight gain or weight loss 206 (67.32)

Insomnia or hypersomnia 277 (90.52)

Psychomotor agitation or retardation 219 (71.57)

Fatigue or loss of energy 262 (85.62)

Feelings of worthlessness 180 (58.82)

Diminished ability to think or concentrate 248 (81.05)

Recurrent thoughts of death 121 (39.54)

Number of symptoms of depression, n (%)

5 34 (11.11)

6 86 (28.10)

7 95 (31.05)

8 66 (21.57)

9 25 (8.17)

SD, standard deviation.

Number of days with depression in last year (n¼303),

mean 7 SD

99.2 7 87.2

Total number of suicide attempts (n¼306)

0 247 (80.72)

1 35 (11.44)

2–3 12 (3.92)

4–5 6 (1.96)

More 6 (1.96)

Previous psychiatric hospitalizations (n¼306)

None 209 (68.30)

One 50 (16.34)

Multiple 47 (15.36)

Note: All values are n (%), unless indicated otherwise. SD, standard deviation.

(Fig. 1). Comorbidity analysis revealed that 88/306 (29.3%) had atleast one comorbidity, the most common of which was anxietydisorder (20.9%), followed by borderline personality disorder (7.2%),substance abuse (4.0%), and eating disorders (0.3%) or attention-deficit hyperactivity disorder (ADHD; 0.3%). The most commonanxiety disorder was generalized anxiety disorder (GAD) at 10.3%,followed by panic disorder without agoraphobia (6.5%). Approxi-mately 26.9% of patients exhibited mixed state, while 20.6% werepositive for atypical depression.

We also analyzed that the treatment patients were receiving atthe time of the study (Table 3). Most patients were being treatedwith more than one prescribed medication, the most common ofwhich were antidepressants (88.2%) and anxiolytic drugs (75.2%,Table 3). Benzodiazepines were the most common anxiolytic drugprescribed, and selective serotonin reuptake inhibitors (SSRIs)were the most common antidepressant. In cases where antipsy-chotic medications were prescribed, more patients were receivingatypical/second generation antipsychotics (36.3%) than were tak-ing typical/first generation antipsychotics (13.1%, Table 3).Valproate was the most commonly prescribed mood stabilizer,at a frequency of 20.6% (Table 3).

Fig. 1. Global assessment of functioning in patient population.

Table 3Current patient treatments (N¼306).

Treatment Frequency, n (%)

Electroconvulsive therapy 56 (18.30)

Anxiolytic drugs 76 (24.84)

Benzodiazepines 71 (23.20)

Other anxiolytic 6 (1.96)

Antidepressant drugs 270 (88.24)

Selective serotonin reuptake inhibitors 182 (59.48)

Serotonin-norepinephrine reuptake inhibitors 67 (21.90)

Tricyclic antidepressants 86 (28.10)

Other antidepressant 7 (2.29)

Antipsychotic drugs 147 (48.04)

Typical/first generation 40 (13.07)

Atypical/second generation 111 (36.27)

Mood stabilizers 126 (41.18)

Lithium 42 (13.73)

Valproate 63 (20.59)

Carbamazepine 25 (8.17)

Other mood stabilizer 34 (11.11)

Table 4Incidence of BPD among patients suffering a major depressive episode: various

measures of bipolarity (N¼306).

Measure of bipolarity Incidence of BPD, n (%)

DSM-IV 42 (13.73)

HCL32a 188 (62.25)

Other definition of bipolarity 144 (47.06)

Modified DSM-IVb 86 (28.20)

a N¼302.b N¼305.

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224220

3.3. Prevalence of bipolar disorder among patients suffering a major

depressive episode

Of the 306 patients, 80 (26.1%) were known to be BP1/2 at thestart of the study. The number of bipolar patients more thandoubled when HCL-32 R2 scores was considered, with 188/302,(62.2%) positive for BPD (Table 4). Other definitions of bipolaritycharacterized 144/306 (47.1%) patients as bipolar (Table 4). WhenDSM-IV criteria was used, only 42/306 (13.7%) were consideredbipolar, a number lower than the known prevalence of BPD(Table 4). According to the modified DSM-IV criteria, 86/305(28.2%) of patients were bipolar, a number consistent with theknown BP1/2 prevalence (Table 4).

3.4. Analysis of demographics and psychiatric history in bipolar and

unipolar patients

Analysis of patients previously diagnosed as BP1/2 reveals thatnone of the demographic factors examined (gender, age, maritalstatus) differed significantly between the bipolar and unipolarpatient groups (Table 5). This was consistent with the majority ofanalyses, which failed to highlight any major differences ingender, age, or marital status between bipolar and unipolarpatient groups. Only when other definitions of bipolarity (Angstet al., 2003a, b) were considered, was age significant (p¼0.0064),with bipolar patients having a younger average age (35.45) thanunipolar patients (39.73, Table 5).

To determine if factors within a patient’s psychiatric historycan be used to predict bipolarity, we assessed a number of suchvariables. Some factors, such as history of suicide attempt(s),were weakly predictive of BPD, being significantly differentbetween the two groups in a minority of analyses (Table 5).In contrast, a number of psychiatric factors were found to be highlysignificant in elucidating the prevalence of bipolarity. Youngerage at first psychiatric symptom was a significant predictor ofBPD in many analyses (HCL-32 R2, p¼0.0439; other definition,p¼0.0014; known BP1/BP2, p¼0.0005). Likewise, bipolar patientswere more likely to report a family history of BPD, seasonalitywith their mood episodes, have a past history of mood episodes,and have a greater number of episodes in the past year thanunipolar patients in every analyses preformed except DSM-IV(Table 5). A number of analyses also revealed that a greaternumber of bipolar patients had been hospitalized than unipolarpatients (Table 5).

Patient response to previous antidepressant treatment alsoprovided clues about what experiences are more common inbipolar patients. The presence of manic/hypomanic switchingwhile on antidepressants was a highly significant predictor ofBPD by every definition of bipolarity except DSM-IV. For example,of those meeting the HCL-32 R2 criteria for bipolarity, 27%experienced manic/hypomanic switches compared to 6% of uni-polar patients. Additionally, bipolar patients were more likely toreport mood liability and/or irritability while on antidepressants,and have been previously treated with antidepressants thanunipolar group (Table 5).

3.5. Clinical characteristics predictive of BPD

We investigated whether various clinical characteristics of thepatients’ current depressive episodes differed between the bipolarand unipolar groups. Patients were evaluated for signs of atypicaldepression (hypersomnia, hyperphagia, and/or leaden paralysis)or mixed state (racing thoughts, mixed state, more talkative, and/or distractibility, Table 6). Bipolar patients tended to exhibitsignificantly higher prevalence of atypical depression and mixedstate than those who lacked a bipolar diagnosis (Table 6). Inter-estingly, the presence of mixed state symptoms proved highlysignificant in every analysis except when DSM-IV criteria wereconsidered. Additional factors regarding a patient’s current psy-chiatric state that were found to be significant predictors ofbipolarity included: psychomotor agitation, irritability, moodreactivity, and the total number of depressive symptoms exhib-ited (Table 6). In contrast, global assessment of functioning (GAF)scores differed between those with and without BPD only whenthe patients were separated according to DSM-IV-based criteria.Illness progression significantly differed between the two groupsunder most definitions of bipolarity, with bipolar patients muchmore likely to experience a major depressive episode with a freeinterval than unipolar patients (Table 6).

Table 5Psychiatric factors predictive of BPD.

Psychiatric predictive factors Analysis Number of bipolar positive, n (%) Number of unipolar positive, n (%) Significance

Postpartum depression Known 8/31 (25.8) 12/119 (10.1) p¼0.0348

History of suicide attempts HCL 43/188 (22.8) 15/114 (13.2) p¼0.0378

Mod 24/86 (27.9) 35/219 (15.9) p¼0.0177

Seasonality HCL 35/188 (18.6) 10/112 (8.9) p¼0.0230

Other 29/144 (20.1) 17/160 (10.6) p¼0.0208

Mod 24/86 (27.9) 22/217 (10.1) p¼0.0001

Known 23/80 (28.8) 23/224 (10.3) p¼0.0001

Previous psychiatric hospitalization DSM 21/42 (50) 76/264 (28.8) p¼0.0102

Other 57/144 (39.6) 40/162 (24.7) p¼0.0078

Mod 48/86 (55.8) 49/219 (22.4) po0.0001

Known 37/80 (46.3) 60/226 (26.5) p¼0.0032

Family history of BPD/mania/hypomania Other 38/144 (26.4) 19/162 (11.7) p¼0.0010

Mod 27/86 (31.4) 30/219 (13.7) p¼0.0004

Known 30/80 (37.5) 27/226 (11.9) po0.0001

Manic switching while on antidepressants HCL 50/184 (27.2) 7/110 (6.4) po0.0001

Other 55/141 (39) 3/157 (1.9) po0.000

Mod 40/83 (48.2) 17/214 (7.9) po0.0001

Known 43/78 (55.1) 15/220 (6.8) po0.0001

Mood liability Other 45/140 (32.1) 10/157 (6.4) po0.0001

Mod 29/82 (35.4) 25/214 (11.7) po0.0001

Known 34/78 (43.6) 21/219 (9.6) po0.0001

Irritability Other 45/138 (32.6) 26/157 (16.6) p¼0.0013

Mod 34/82 (41.5) 36/212 (17) po0.0001

Known 34/76 (44.7) 37/219 (16.9) po0.0001

First antidepressant treatment Other 50/138 (36.2) 77/153 (50.3) p¼0.0155

Mod 27/82 (32.9) 100/208 (48.1) p¼0.0192

Known 18/77 (23.4) 109/214 (51) po0.0001

DSM, DSM-IV analysis; HCL, HCL-32-R2 analysis; Other, other definitions of bipolarity analysis; Mod, modified DSM-IV analysis; Known, known BP1/2 analysis.

Table 6Clinical characteristics predictive of BPD.

Clinical predictive factors Analysis Number of bipolar positive, n (%) Number of unipolar positive, n (%) Significance

In-patient status DSM 13/42 (31) 32/264 (12.1) p¼0.0014

Mod 22/86 (25.6) 23/219 (10.5) p¼0.0008

Atypical depression Other 38/144 (26.4) 25/162 (15.4) p¼0.0180

Mod 26/86 (30.2) 37/219 (16.9) p¼0.0096

Known 30/80 (37.5) 33/226 (14.6) po0.0001

Mixed state HCL 65/188 (34.6) 16/113 (14.2) p¼0.0001

Other 62/143 (43.4) 20/162 (12.3) po0.0001

Mod 38/86 (44.2) 44/218 (20.2) po0.0001

Known 46/80 (57.5) 36/225 (16) po0.0001

Psychomotor agitation Other 69/144 (47.9) 48/162 (29.6) p¼0.0010

Mod 42/86 (48.8) 74/219 (33.8) p¼0.0149

Known 46/80 (57.5) 71/226 (31.4) po0.0001

Irritable mood Other 92/144 (63.9) 67/162 (41.4) p¼0.0001

Mod 53/86 (61.6) 105/219 (47.9) p¼0.0314

Known 57/80 (71.3) 102/226 (45.1) p¼0.0001

Psychotic features Known 35/80 (43.8) 45/225 (20) po0.0001

Mood reactivity DSM 4/42 (9.5) 61/263 (23.2) p¼0.0445

HCL 51/188 (27.1) 13/113 (11.5) p¼0.0013

Other 46/143 (32.2) 19/162 (11.7) po0.0001

Known 26/80 (32.5) 39/225 (17.3) p¼0.0044

Mood-free intervals DSM 33/42 (78.6) 153/264 (58) p¼0.0322

Other 92/144 (63.9) 94/162 (58) p¼0.0013

Mod 62/86 (72.1) 123/219 (56.2) p¼0.0003

Known 68/80 (85) 118/226 (52.2) po0.0001

Episode occurred on antidepressant HCL 51/173 (29.5) 16/92 (17.4) p¼0.0311

Other 53/144 (36.8) 14/125 (11.2) po0.0001

Mod 36/86 (41.9) 30/182 (16.5) po0.0001

Known 32/80 (40) 35/189 (18.5) p¼0.0002

DSM, DSM-IV analysis; HCL, HCL-32-R2 analysis; Other, other definitions of bipolarity analysis; Mod, modified DSM-IV analysis; Known, known BP1/2 analysis.

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224 221

3.6. Co-morbid conditions and current treatments of patients with

and without BPD

Interestingly, rates of substance abuse did not differ significantlybetween bipolar and unipolar patients, as we had hypothesized they

would. Similarly there were no differences in the prevalence ratesbetween the two groups for eating disorders, ADHD, anxietydisorder, obsessive compulsive disorder, or social phobia. Only ratesof borderline personality disorder were significantly differentbetween those with and without BPD. Borderline personality was

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224222

seen in 9.6–10.5% of bipolar patients as compared to 3.5–4.3% ofunipolar patients (p¼0.048 for HCL-32 R2 analysis and p¼0.0377for other definitions of bipolarity analysis, respectively).

The drug treatment which differed most in the rates ofprescription between those with and without BPD was moodstabilizers. Between 50% and 92% of patients diagnosed with BPD(depending on the defining criteria) were on prescription moodstabilizers, whereas only 23–37% of unipolar patients were beingtreated with this drug type. In most analyses, many more unipolarpatients were on antidepressants than bipolar. However, frequen-cies of other treatments between bipolar and unipolar patientswere highly dependent on the criteria used to determinebipolarity.

4. Discussion

The ability to differentiate between bipolar and depressivemood disorders is of critical importance in the field of psychiatry,as available treatments for these two conditions are quite differ-ent. Attempts to distinguish among different depressive disordersbegan in earnest with publication of the Diagnostic and Statistical

Manual of Mental Disorders, which provided standardized criteriafor classifying mental disorders (Gruenberg et al., 2008). Sincethat time, it has undergone numerous revisions as physiciansstrive for a more reliable classification system and fine-tuneterminology. Still, strictly delineating these disorders in theclinical setting can prove difficult, as patients exhibit a hetero-geneous range of symptoms; hence misdiagnosis, especially ofmore mild cases of BPD, is not infrequent (Hirschfeld et al., 2003;Manning, 2003; Manning et al., 1999). Often patients may fail tomention manic/hypomanic symptoms when being treated for adepression (Bowden, 2001) or their health care providers may failto screen them for BPD, even if they are at high risk (Brickmanet al., 2002). The typical age at onset of BPD is between 17 and 21years (Yatham et al., 2005), and the estimated prevalence inadults is 1% for bipolar I disorder, 1.1% for bipolar II disorder, and2.4% for sub-threshold BPD (Merikangas et al., 2007).

Depressive symptoms tend to dominate the illness presenta-tion and are approximately 3.5 times more frequent than manic/hypomanic symptoms in bipolar I disorder (Judd et al., 2002) and38.5 times more frequent in bipolar II disorder (Judd et al., 2003).Additionally, patients with BPD are at a high risk of suicidalideation; the estimated life time rate of suicide is between 10 and15% nearly 20 fold greater than that of the general population(McIntyre et al., 2008).

For these reasons, it takes an average of 8 years from the onsetof symptoms for a correct diagnosis of BPD to be made (Lish et al.,1994). A study conducted by Hirschfeld et al. (2003) found thatindividuals with BPD reported that the illness manifests itselfearly in life but the accurate diagnosis lags by many years. In fact,69% of over one-third of participants waited 10 years or morebefore receiving an accurate diagnosis of BPD. Those who werediagnosed consulted a mean of four physicians prior to receivingthe correct diagnosis, with the most frequent misdiagnosis beingunipolar depression.

To help resolve this issue, we designed our study to try todetermine the prevalence of bipolarity in patients presenting witha major depressive episode. At the start of our study, 26.1% ofpatients had been previously diagnosed BP1/2. We hypothesizedthat this number underestimated the true prevalence of bipolar-ity, so we used a number of different measures of BPD to elucidatethe issue. Remarkably, HCL-32 R2 responses revealed that 62.2%of our study population met the criteria for BPD. Similarly, whenother definitions of bipolarity (Angst et al., 2003a, b) wereconsidered (which aim to take into account so-called softer

expressions of bipolarity which might otherwise be overlookedon the bipolar spectrum), we found a much higher proportion(47.1%) of patients to be positive for BPD than originally diag-nosed. This number is consistent with previous studies that haveestimated that �40% of persons diagnosed with depression arelater found to have a bipolar spectrum disorder (Ghaemi et al.,1999, 2000). As our findings indicate that more than half ofpeople seeking treatment for a major depressive episode arelikely to have had manic/hypomanic symptoms, careful scrutinyof depressive patients for underlying BPD is warranted.

In contrast, we found that using the DSM-IV criteria under-estimated the prevalence of bipolarity, with only 13.7% of patientsscreening positive, the majority of these being manic rather thanhypomanic (73.8% vs. 26.1%, respectively). One explanation forthis apparent underestimation may be that DSM-IV is better atcapturing the more severe manic cases than milder forms of BPD.Modified DSM-IV criteria, which factors into account specifically,(i) increased activity as a gate question, (ii) 1 or 2–3-dayhypomanic episodes, and (iii) the elimination of all exclusioncriteria in DSM-IV (Angst et al., 2012), revealed the prevalenceconsistent with number of known BP1/2 patients (28.2%). Gau-ging by the number of known BP1/2 patients in this study whichis likely an underestimation, it appears that using the DSM-IVcriteria and modified DSM-IV criteria underestimates the trueprevalence of bipolar patients. Additionally, another recent studythat evaluated the long-term stability of the ICD-10 bipolardiagnosis found that there was a high prevalence of misdiagnosisand diagnostic shift from other psychiatric disorders to BPD(Baca-Garcia et al., 2007).

The current study examines a number of characteristics ofpatients presenting with a major depressive episode. Of thedemographic factors considered, none were predictive of bipolar-ity. More useful in predicting bipolarity were factors in a patient’spsychiatric background. We found that bipolar patients tend tohave an earlier onset of psychiatric symptoms, a higher preva-lence of suicide attempts (and a greater number of attempts), agreater likelihood of experiencing seasonality of mood episodes, agreater number of mood episodes, and were more likely to reportmanic switches in response to antidepressants compared totheir unipolar peers. The majority of these factors emerged assignificant in most analyses, suggesting the robustness of thesepredictors, and consistency with the findings of other studies.For example, patients with BPD reported experiencing greaterseasonality to their mood episodes than unipolar patients (Shinet al., 2005) and tend to have more depressive episodes whenseasonality is involved (Goikolea et al., 2007). Also, treatment ofbipolar patients with antidepressants has been linked to manicswitching in an estimated 20–30% of cases (Leverich et al., 2006),with the incidence of switching increasing with the increase inthe number of antidepressants are used (Truman et al., 2007).

In a 10-year follow-up prospective study on 300 bipolar I andII patients, polarity at onset was a good predictor of the polarity ofsubsequent episodes over time (Daban et al., 2006). Specifically,67% of patients experienced a depressive onset which is twice asfrequent as manic/hypomanic onset and carries more chronicityand cyclicity.

Assessing the proportion of time spent in mania, depressionand euthymia in a large cohort of bipolar subjects studied long-itudinally found that percentages of time spent ill for bipolar Iversus bipolar II patients were: euthymia 47.7% vs. 50.2%;depression 36% vs. 37%; hypomania 11.5% vs. 9.8%; mania 1% vs.0.2%; and cycling 3.7% vs. 2.8%. Depression represents the pre-dominant abnormal mood state for treated outpatients withbipolar I and II disorder (Kupka et al., 2007).

We also identified a number of clinical features of the patients’current depressive episodes that proved helpful in predicting

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224 223

bipolarity. Interestingly, we observed that using more expansivebipolar criteria (such as HCL-32 R2 responses) as opposed toDSM-IV criteria alone resulted in a greater number of qualitiespredictive of bipolarity. For instance, prevalence of mixed statewas some two-to-three times higher in the bipolar group than inunipolar patients in every analysis except DSM-IV. As bipolarpatients with mixed state are at a higher risk for early age ofonset, rapid cycling, and increased likelihood of suicide attempts(Goldberg et al., 2009), assessing mixed state may prove key tosuccessful treatment. We also found that mood reactivity was agood predictor of bipolarity in most analyses, consistent withreports that mood-reactivity is strongly associated with BP2, butnot unipolar depression (Benazzi, 2002).

In another study, researchers found that co-morbidity wasdetected in 31% of the sample and stressed that there was anassociation between depression, suicidality, and co-morbidity inbipolar I disorder (Vieta et al., 2001). Co-morbidity had a clearrelevance in the course and outcome of BPD.

One of the most surprising results of this study was theprevalence of co-morbidity in the two patient populations.Though we analyzed the rates of substance abuse and a rangeof co-morbid psychiatric conditions, only one co-morbidity wassignificantly associated with bipolarity. Borderline personalitydisorder was more than twice as frequent among bipolarpatients—an association consistent with previous studies(Magill, 2004; Perugi et al., 2011). The prevalence of substanceabuse (4%) was much lower than might be expected, with recentestimates contending that a much higher percentage of patientswith mood disorders also have a substance use disorder (Grantet al., 2004; Maremmani et al., 2006). However, this apparentdiscrepancy may stem from the fact that our study populationwas Egyptian, rather than western, and likely have different ratesof substance abuse overall.

Studies in bipolar patients (Faravelli et al., 2006; Henry et al.,2003) have suggested co-morbidity with anxiety disorders with atleast one lifetime anxiety disorder: 16% had panic disorder (withand without agoraphobia and panic attacks), 11% had phobia and3% had obsessive compulsive disorder. Co-morbidity of anxietydisorders was not correlated to the severity of illness. Also, othercomorbidities included personality disorders as borderline per-sonality disorder, paranoid and passive–aggressive personalitydisorders.

There are some limitations to this study. The non-randomselection of participating centers could have led to a bias withinclusion of only those psychiatrists who had a special interest inBPD. Additionally, the retrospective assessment of hypomanicsymptoms could have contributed to an imprecise calculation ofthe frequency.

Nevertheless, in conclusion, we believe that among patientsbeing treated for a major depressive episode, a significant percen-tage may have an undiagnosed, underlying BPD. Moreover, evaluat-ing bipolarity strictly on DSM-IV criteria alone may result in a failureto diagnose many patients who have symptoms of bipolarity,especially those with hypomanic symptoms. Physicians treatingpatients with major depression need to give special considerationto conducting an in-depth screening for BPD that utilizes multipleassessment tools. Additionally, a number of key factors within apatient’s psychiatric history and their clinical manifestations ofdepression may signal an increased likelihood of BPD.

Role of funding source

This research was made possible through an unrestricted educational grant

supported by Sanofi (Egypt).

Conflict of interestThe authors state no conflict of interest

AcknowledgmentsThe authors would like to acknowledge and thank all our colleagues who

participated and assisted in data gathering in this research: Drs. Adel Medany, Afaf

H. Khalil, Ahmed Rady Ahmed Refaat, Akl Atta, Hashem Bahary, Hoda Salama,

Mohamed Ghanem, Sherif Attalah, Ossama El Kholy, Tarek Assad, Tarek Molokhia,

Waguih Fares and Victor Sami.

Editorial assistance in preparing this manuscript was provided by Dr. Sujata

Shah and Anahita Gouri of Sanofi (India).

References

American_Psychiatric_Association, 1994. Diagnostic and Statistical Manual ofMental Disorders, Washington DC.

Angst, J., Adolfsson, R., Benazzi, F., Gamma, A., Hantouche, E., Meyer, T.D.,Skeppar, P., Vieta, E., Scott, J., 2005. The HCL-32: towards a self-assessmenttool for hypomanic symptoms in outpatients. Journal of Affective Disorders 88,217–233.

Angst, J., Azorin, J.M., Bowden, C.L., Perugi, G., Vieta, E., Gamma, A., Young, A.H.,2011. Prevalence and characteristics of undiagnosed bipolar disorders inpatients with a major depressive episode: the BRIDGE study. Archives ofGeneral Psychiatry 68, 791–798.

Angst, J., Gamma, A., Benazzi, F., Ajdacic, V., Eich, D., Rossler, W., 2003a. Diagnosticissues in bipolar disorder. European Neuropsychopharmacology : The Journalof the European College of Neuropsychopharmacology 13 (Suppl 2), S43–S50.

Angst, J., Gamma, A., Benazzi, F., Ajdacic, V., Eich, D., Rossler, W., 2003b. Toward are-definition of subthreshold bipolarity: epidemiology and proposed criteriafor bipolar-II, minor bipolar disorders and hypomania. Journal of AffectiveDisorders 73, 133–146.

Angst, J., Gamma, A., Bowden, C.L., Azorin, J.M., Perugi, G., Vieta, E., Young, A.H.,2012. Diagnostic criteria for bipolarity based on an international sample of5,635 patients with DSM-IV major depressive episodes. European Archives ofPsychiatry and Clinical Neuroscience 262, 3–11.

Baca-Garcia, E., Perez-Rodriguez, M.M., Basurte-Villamor, I., Lopez-Castroman, J.,Fernandez del Moral, A.L., Jimenez-Arriero, M.A., Gronzalez de Rivera, J.L., Saiz-Ruiz, J., Leiva-Murillo, J.M., de Prado-Cumplido, M., Santiago-Mozos, R., Artes-Rodriguez, A., Oquendo, M.A., de Leon, J., 2007. Diagnostic stability andevolution of bipolar disorder in clinical practice: a prospective cohort study.Acta Psychiatrica Scandinavica 115, 473–480.

Benazzi, F., 2002. Should mood reactivity be included in the DSM-IV atypicalfeatures specifier? European Archives of Psychiatry and Clinical Neuroscience252, 135–140.

Bowden, C.L., 2001. Strategies to reduce misdiagnosis of bipolar depression.Psychiatric Services 52, 51–55.

Brickman, A.L., LoPiccolo, C.J., Johnson, S.L., 2002. Screening for bipolar disorder.Psychiatric Services 53, 349.

Cutcliffe, J.R., Lakeman, R., 2010. Challenging normative orthodoxies in depression:Huxley’s Utopia or Dante’s Inferno? Archives of Psychiatric Nursing 24, 114–124.

Daban, C., Colom, F., Sanchez-Moreno, J., Garcia-Amador, M., Vieta, E., 2006.Clinical correlates of first-episode polarity in bipolar disorder. ComprehensivePsychiatry 47, 433–437.

Davies, S., Naik, P.C., Lee, A.S., 2001. Depression, suicide, and the national serviceframework. BMJ (Clinical Research Ed) 322, 1500–1501.

Faravelli, C., Rosi, S., Alessandra Scarpato, M., Lampronti, L., Amedei, S.G., Rana, N.,2006. Threshold and subthreshold bipolar disorders in the Sesto FiorentinoStudy. Journal of Affective disorders 94, 111–119.

Ghaemi, S.N., Boiman, E.E., Goodwin, F.K., 2000. Diagnosing bipolar disorder andthe effect of antidepressants: a naturalistic study. The Journal of ClinicalPsychiatry 61, 804–808.

Ghaemi, S.N., Ko, J.Y., Goodwin, F.K., 2002. ’’Cade’s disease’’ and beyond: mis-diagnosis, antidepressant use, and a proposed definition for bipolar spectrumdisorder. Canadian Journal of Psychiatry Revue Canadienne de Psychiatrie 47,125–134.

Ghaemi, S.N., Sachs, G.S., Chiou, A.M., Pandurangi, A.K., Goodwin, K., 1999.Is bipolar disorder still underdiagnosed? Are antidepressants overutilized?52, 135–144Journal of Affective Disorders 52, 135–144.

Goikolea, J.M., Colom, F., Martinez-Aran, A., Sanchez-Moreno, J., Giordano, A.,Bulbena, A., Vieta, E., 2007. Clinical and prognostic implications of seasonalpattern in bipolar disorder: a 10-year follow-up of 302 patients. PsychologicalMedicine 37, 1595–1599.

Goldberg, J.F., Perlis, R.H., Bowden, C.L., Thase, M.E., Miklowitz, D.J., Marangell, L.B.,Calabrese, J.R., Nierenberg, A.A., Sachs, G.S., 2009. Manic symptoms duringdepressive episodes in 1,380 patients with bipolar disorder: findings from theSTEP-BD. The American Journal of Psychiatry 166, 173–181.

Gonzalez, H.M., Vega, W.A., Williams, D.R., Tarraf, W., West, B.T., Neighbors, H.W.,2010. Depression care in the United States: too little for too few. Archives ofGeneral Psychiatry 67, 37–46.

Grant, B.F., Stinson, F.S., Dawson, D.A., Chou, S.P., Dufour, M.C., Compton, W.,Pickering, R.P., Kaplan, K., 2004. Prevalence and co-occurrence of substanceuse disorders and independent mood and anxiety disorders: results from the

T. Okasha et al. / Journal of Affective Disorders 147 (2013) 217–224224

National Epidemiologic Survey on Alcohol and Related Conditions. Archives ofGeneral Psychiatry 61, 807–816.

Gruenberg, A.M., Goldstein, R.D., Pincus, H.A., 2008. Classification of Depression:Research and Diagnostic Criteria: DSM-IV and ICD-10. Biology of Depression.Wiley-VCH Verlag GmbH, pp. 1–12.

Henry, C., Van den Bulke, D., Bellivier, F., Etain, B., Rouillon, F., Leboyer, M., 2003.Anxiety disorders in 318 bipolar patients: prevalence and impact on illness

severity and response to mood stabilizer. The Journal of Clinical Psychiatry 64,331–335.

Hirschfeld, R.M., Lewis, L., Vornik, L.A., 2003. Perceptions and impact of bipolardisorder: how far have we really come? Results of the national depressive andmanic-depressive association 2000 survey of individuals with bipolar disorder.

The Journal of Clinical Psychiatry 64, 161–174.Judd, L.L., Akiskal, H.S., Schettler, P.J., Coryell, W., Endicott, J., Maser, J.D., Solomon,

D.A., Leon, A.C., Keller, M.B., 2003. A prospective investigation of the naturalhistory of the long-term weekly symptomatic status of bipolar II disorder.Archives of General Psychiatry 60, 261–269.

Judd, L.L., Akiskal, H.S., Schettler, P.J., Endicott, J., Maser, J., Solomon, D.A., Leon,A.C., Rice, J.A., Keller, M.B., 2002. The long-term natural history of the weekly

symptomatic status of bipolar I disorder. Archives of General Psychiatry 59,530–537.

Keck Jr., P.E., McElroy, S.L., Arnold, L.M., 2001. Bipolar disorder. The Medical Clinicsof North America 85, 645–661.

Kessler, R.C., Stang, P., Wittchen, H.U., Stein, M., Walters, E.E., 1999. Lifetimeco-morbidities between social phobia and mood disorders in the US NationalComorbidity Survey. Psychological Medicine 29, 555–567.

Kupka, R.W., Altshuler, L.L., Nolen, W.A., Suppes, T., Luckenbaugh, D.A., Leverich,G.S., Frye, M.A., Keck Jr., P.E., McElroy, S.L., Grunze, H., Post, R.M., 2007. Three

times more days depressed than manic or hypomanic in both bipolar I andbipolar II disorder. Bipolar Disorders 9, 531–535.

Leverich, G.S., Altshuler, L.L., Frye, M.A., Suppes, T., McElroy, S.L., Keck Jr., P.E.,Kupka, R.W., Denicoff, K.D., Nolen, W.A., Grunze, H., Martinez, M.I., Post, R.M.,2006. Risk of switch in mood polarity to hypomania or mania in patients with

bipolar depression during acute and continuation trials of venlafaxine, sertra-line, and bupropion as adjuncts to mood stabilizers. The American Journal of

Psychiatry 163, 232–239.Lish, J.D., Dime-Meenan, S., Whybrow, P.C., Price, R.A., Hirschfeld, R.M., 1994.

The National Depressive and Manic-depressive Association (DMDA) survey ofbipolar members. Journal of Affective Disorders 31, 281–294.

Magill, C.A., 2004. The boundary between borderline personality disorder andbipolar disorder: current concepts and challenges. Canadian Journal ofPsychiatry Revue Canadienne de Psychiatrie 49, 551–556.

Manning, J.S., 2003. Bipolar disorder in primary care. Journal of Family Practice,Supplement, S6–S9.

Manning, J.S., Haykal, R.F., Akiskal, H.S., 1999. The role of bipolarity in depressionin the family practice setting. Psychiatric Clinics of North America 22,689–703.

Maremmani, I., Perugi, G., Pacini, M., Akiskal, H.S., 2006. Toward a unitaryperspective on the bipolar spectrum and substance abuse: opiate addictionas a paradigm. Journal of Affective Disorders 93, 1–12.

McIntyre, R.S., Muzina, D.J., Kemp, D.E., Blank, D., Woldeyohannes, H.O., Lofchy, J.,Soczynska, J.K., Banik, S., Konarski, J.Z., 2008. Bipolar disorder and suicide:research synthesis and clinical translation. Current Psychiatry Reports 10,66–72.

Merikangas, K.R., Akiskal, H.S., Angst, J., Greenberg, P.E., Hirschfeld, R.M.,Petukhova, M., Kessler, R.C., 2007. Lifetime and 12-month prevalence ofbipolar spectrum disorder in the National Comorbidity Survey replication.Archives of General Psychiatry 64, 543–552.

Perugi, G., Fornaro, M., Akiskal, H.S., 2011. Are atypical depression, borderlinepersonality disorder and bipolar II disorder overlapping manifestations of acommon cyclothymic diathesis? World psychiatry: official journal of theWorld Psychiatric Association (WPA) 10, 45–51.

Shin, K., Schaffer, A., Levitt, A.J., Boyle, M.H., 2005. Seasonality in a communitysample of bipolar, unipolar and control subjects. Journal of Affective Disorders86, 19–25.

Truman, C.J., Goldberg, J.F., Ghaemi, S.N., Baldassano, C.F., Wisniewski, S.R.,Dennehy, E.B., Thase, M.E., Sachs, G.S., 2007. Self-reported history of manic/hypomanic switch associated with antidepressant use: data from the Sys-tematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). TheJournal of Clinical Psychiatry 68, 1472–1479.

Vieta, E., Colom, F., Corbella, B., Martinez-Aran, A., Reinares, M., Benabarre, A.,Gasto, C., 2001. Clinical correlates of psychiatric comorbidity in bipolar Ipatients. Bipolar Disorders 3, 253–258.

World_Health_Organization, 2004 The Global Burden of Disease.Yatham, L.N., Kennedy, S.H., O’Donovan, C., Parikh, S., MacQueen, G., McIntyre, R.,

Sharma, V., Silverstone, P., Alda, M., Baruch, P., Beaulieu, S., Daigneault, A.,Milev, R., Young, L.T., Ravindran, A., Schaffer, A., Connolly, M., Gorman, C.P.,2005. Canadian Network for Mood and Anxiety Treatments (CANMAT) guide-lines for the management of patients with bipolar disorder: consensus andcontroversies. Bipolar Disorders 7 (Suppl 3), 5–69.