FIFTH INTERNATIONAL CONFERENCE 
ON BIPOLAR DISORDER

Poster Abstracts (201-220)

201. Olanzapine versus Lithium in Relapse Prevention in Bipolar Disorder: A Randomized Double-blind Controlled 12-month Clinical Trial
M. Tohen¹, A. Marneros², C. Bowden³, W. Greil⁴, A. Koukopoulis⁵, H. Belmaker⁶, T. Jacobs¹, R.W. Baker¹, D. Williamson¹, A.R. Evans¹, M. Dossenbach¹, G. Cassano^7
1Lilly Research Laboratories, Indianapolis, IN, USA, ²University Hospital for Psychiatry and Psychotherapy, Halle/Saale, Germany, ³Department of Psychiatry, University of Texas Health Science Center, San Antonio, TX, USA, ⁴Department of Psychiatry, Ludwig-Maximilians University of Munich, Germany, ⁵Centro Lucio Bini, Rome, Italy, ⁶Ministry of Health Mental Health Center, Ben-Gurion University of the Negev, Beer-Sheva, Israel, ⁷Department of Psychiatry, University of Pisa, Pisa, Italy

Purpose: This is the first randomized double-blind comparison of the efficacy and safety of olanzapine and lithium in the prevention of relapse into a manic, mixed, or depressed bipolar episode.
Methods: 543 patients with a diagnosis of bipolar I disorder, manic or mixed type, with a history of at least 2 manic or mixed episodes within 6 years, and a YMRS total score > 20 entered the study and received open label combination therapy of olanzapine and lithium for 6-12 weeks. Of these, 431 patients met symptomatic remission criteria (YMRS total score < 12 and HAMD-21 total score < 8) and were randomized to monotherapy with either olanzapine (N=217) (5-20 mg/d) or lithium (N=214) (serum level of 0.6 to 1.2 mEq/L in a dose range of 300-1800 mg/d) for 52 weeks of double-blind treatment.
Results: Significantly more olanzapine-treated patients (46.5%) completed the 52-week trial than those on lithium (32.7%; P=.004). Relapse to an affective episode, defined as YMRS total score > 15 and/or HAMD-21 total score > 15, occurred in 30.0% of olanzapine-treated and 38.8% of lithium-treated patients (P=.055). Olanzapine-treated patients had a statistically significantly lower incidence of relapse into a manic episode than lithium-treated patients (14.3% vs. 28.0%, respectively, P<.001), and both groups had similar incidences of relapse into a depressive episode (16.1% vs15.4%, respectively, P=.895). The rates of discontinuation due to adverse events were 18.9% for the olanzapine group and 25.7% for the lithium group (P=.105). Weight gain across open-label and double-blind therapy phases was statistically significantly greater in the olanzapine group compared to the lithium group (1.79 kg vs 1.38 kg, respectively, P<.001).
Conclusion: Olanzapine and lithium both appear to effectively and safely prolong remission in bipolar disorder, yet more patients remained on olanzapine throughout this year-long study and it was more effective than lithium in preventing relapse into mania.
Keywords: 1) Maintenance 2) Relapse Prevention 3) Mood Stabilizer

202. A Rating Scale for the Assessment of Eating Dysfunctions in Bipolar Disorders (BEDS)
C. Torrent, E. Vieta, J.A. Crespo, A. González-Pinto, J. Del Valle, J.M. Olivares, A. Rodriguez, C. De Arce, L. Sánchez-Planell, F. Colom
Bipolar Disorder Program, Hospital Clinic, Barcelona IDIBAPS, Stanley Research Center, Barcelona, Spain

Introduction: The presence of eating disorders in bipolar population is not rare, with rates over 10% according to the few available epidemiologic studies. However the literature on this issue is still scarce. A high percentage of bipolar individuals suffer from serious problems related to their eating behavior without fulfilling criteria for any DSM-IV eating disorder.
Methods: The Barcelona Eating Disorders Scale (BEDS) was designed on the basis of the existing eating scales, adjusted to the characteristics of bipolar disorders. The scale is aimed at assessing the intensity and frequency of several eating dysfunctions in the bipolar population.
Results: The BEDS is a 10-item, self-administered scale composed by five groups of items, namely: (1) items concerning habits regularity, (2) items assessing the influence of mood on hunger, (3) items concerning binge eating, (4) items regarding lack of control related to eating behavior and (5) items regarding carbohydrates craving. Each item has four possible punctuations –from 0 (never) to 3 (always)- and the final score goes from 0 to 30. Cut-off point has been established at 12. The BEDS has proved to be feasible. Patients receiving scores over 12 may require some specific intervention to improve their eating behavior.
Keywords: 1) Eating Behavior 2) Eating Dysfunctions 3) Self-administered Scale

203. Affective Temperamental Features in Turkey: Findings from the Validity and Reliability Study of TEMPS-A Turkish Version
S. Vahip¹, S. Kesebir¹, M. Kocadere¹, O. Yazici², H. Akiskal^3
1Ege University, School of Medicine, Department of Psychiatry, Affective Disorder Unit, İzmir, ²Istanbul University, School of Medicine, Department of Psychiatry, Mood Disorders Unit, Istanbul, Turkey, ³International Mood Clinic, San Diego, CA, USA

Introduction: The authors evaluated the reliability and validity of the Turkish Version of TEMPS-A (Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire), a scale that identifies the dominant affective temperament. In addition to reliability and validity assessments the prevalence rates of affective temperaments and the psychometric properties for Turkish population were also determined.
Method: The scale was administered to 658 subjects, selected from a wide range of professions (workers, academic faculty, artists and unemployed). The reliability and validity of the scale was assessed by item analysis, internal consistency, two-half reliability, test-retest reliability, factor analysis.
Results: The test-retest reliability and the internal consistency of the scale were very good. The findings of the factor analysis indicated that depressive and anxious temperament overlap considerably. The percentage of subjects with a z-score higher than the second positive standard deviation of depressive, cyclothymic, hyperthymic, irritable and anxious temperaments were 3.1 %, 1.7 %, 1.2 %, 3.7 %, 3.7 %, respectively. Males got higher scores for hyperthymic temperament (p=0.012) and females got higher scores for anxious temperament (p=0.025). There was no significant difference among age groups in frequencies of temperaments. An interesting observation was lack of high scores in all subtypes over age 61. The TEMPS-A Turkish version is a reliable and valid in the Turkish population. The authors suggest further studies to elaborate on the overlap of anxious and depressive temperaments, different distribution in two genders and maybe a semi-structured form to increase reliability and validity for small clinical samples.
Keywords: 1) Temperament 2) Reliability 3) Validity

204. The Impact of Positive Family History on Affective Temperamental Features in Bipolar Patients and Their First Degree Relatives: A Controlled Study
S. Vahip¹, S. Kesebir¹, H. Akiskal², Z. Yüncü^1
1Ege University, School of Medicine, Department of Psychiatry, Affective Disorder Unit, Bornova-Izmir, Turkey, ²International Mood Clinics, San Diego, CA, USA

Introduction: The aim of this study was to investigate the impact of positive family history of bipolar disorder on temperamental features in patients and their first degree relatives. Results would answer at least in familial and descriptive perspective to the questions about existence of subsyndromal states as parts of the bipolar spectrum.
Method: 100 bipolar patients (50 with, 50 without family history of bipolar disorder) and their 281 first degree relatives were taken into the study. ''No family history of bipolar disorder'' decision was relied on at least one year follow-up in our Unit. Temperament evaluation was done with TEMPS-A Turkish version. Two separate healthy matched control groups were formed; one for patients, one for relatives.
Results: In comparison of groups who have and have not family history of bipolar disorder with healthy controls, having at least one dominant temperament was significantly more common both in patients (p=0.02) and in first degree relatives (p=0.01) who have positive family history. Compared to control groups, cyclothymic, hyperthymic and irritable temperament were more frequent in patients with family history (respectively, p=0.006, p=0.013 and p=0.046), while it wasn't found for patients without family history. First degree relatives were differentiated from healthy controls by having higher frequencies of depressive, cyclothymic and hyperthymic temperaments (respectively, p=0.018, p=0.004 and p<0.001) in familial group, and only hyperthymic temperament (p<0.001) in nonfamilial group. Family history of bipolar disorder seems to cause more temperamental disregulation both in patients and their first degree relatives. Although there were some findings need to be clarified, it seems that cyclothymic and hyperthymic breed-true.
Keywords: 1) Temperament 2) Bipolar Disorder 3) Family History

205. Personality Changes during Light Therapy in Bipolar Patients
E.M. Van Gent, G.J.Bonte
Slingeland Hospital, GGNet, The Netherlands

Introduction: Personality changes in light therapy are proved using the Personality Disorders Questionnaire Revised (PDQ-R) for unipolar patients (1). In nonstable bipolar patients were found with the PDQ-R in a high percent (71%) personality disorders2. The question rose of: in bipolar patients after light therapy also a change was found and in which clusters.
Methods: Ten bipolar outpatients received light therapy 2500 lux, during two hours in five sequence days. Measurements: Beck Depression Inventory (BDI) on day one, four, eight and fifteen. PDQ-R on day one and day fifteen.
Results: Eight of the ten patients improved on light therapy. The mean BDI decreased from 22 to 8.5 (p<0.001). The number of personality disorders (DSMIV) reduced from mean 3.2 to mean 1.9. The total scores on the PDQ-R decreased from mean 38.8 to 29.9 (p<0.006). The cluster B changed from mean 11.5 to 9.0 (p<0.013). Five patients had no medication. The others used lithium, valproate, carbamazepine and citalopram.
Discussion: several authors discussed the symptoms of borderline disorder with bipolar disorder (2). Knoppert et al (1999) found 30% of cluster B disorder in a fresh outpatient; in stabled outpatients was this 22%. In our sample 60% had a cluster B disorder before and 30% after light therapy. In depression patients rated their wellbeing worse as expected. Personality disorder symptoms show an overlap with bipolar symptoms. If the latter decrease the personality will be normalized.
Keywords: 1) Personality 2) Bipolar 3) PDQ-R
References:
1.Reichborn-Kjennerud, T., Lingjaerde. O., Dahl, A.A. (1997) DSM III-R personality disorders in seasonal affective disorder: Change associated with depression. Comprehensive Psychiatry, 38, 43-48.

2.Knoppert-van der Klein, E.A.M.1999. From despondency to recklessness. Thesis, Nijmegen.

206. A Counseling Intervention for Adolescent Children of Bipolar Mothers
H. Verdeli, L. Mufson, M.M. Weissman
New York State Psychiatric Institute, New York, NY, USA

Introduction: Children of bipolar parents are exposed to a ’double risk’: 1) genetic factors that predispose them to a biological vulnerability for bipolar and other disorders, and 2) stressful life events, some of which are related to the bipolar parent’s variable clinical state. This is a feasibility study of a counseling intervention for adolescent children of mothers diagnosed with Bipolar I Disorder, which aims to: 1) educate the adolescents and the parents/caretakers about bipolar disorder; 2) help the family make realistic plans in the event of the mother’s relapse; 3)help the adolescents deal more adaptively with stressful interpersonal situations (including the mother’s disorder) and, when necessary, help them regulate their circadian rhythms, social stimulation and daily habits.
Methods: Ten adolescents whose mothers are currently being treated for Bipolar I Disorder at the New York State Psychiatric Institute Bipolar Disorder Research Clinic are being recruited. The adolescents receive 12 sessions of Interpersonal Counseling for Prevention (IPC-P), 3 of which involve family psychoeducation. In order to be included the adolescents (ages 12-17) need to have some functional impairment (C-GAS:75-51) and no history of depression, mania or psychosis (although they can have subsyndromal symptomatology of these disorders). The counseling incorporates a) IPT for adolescents (Mufson L, Moreau D, Weissman MM, et al, 1993), b) psychoeducation (modeled after D. Miklowitz’s psychoeducation in his Family Focused Treatment) and c) an optional modality of mood and circadian regulatory skills (Frank E, Kupfer DJ, Ehlers CL, et al, 1994). It is hypothesized that the intervention may improve adolescents’ understanding of their mother’s bipolar disorder, improve their interpersonal/social functioning, and help them regulate the amount and quality of daily activities (if that impacts on adolescent’s mood/behavior). The mothers and the adolescents are being rated on symptomatology, interpersonal and family functioning. Preliminary results will be presented.
Keywords: 1) Adolescent 2) Bipolar Mothers 3) Interpersonal Counseling

207. Effect of a Novel Long-acting Antipsychotic Formulation in Stable Patients with Schizoaffective Disorder
E. Vieta¹, G. Gharabawi², C. Bossie², Y. Zhu², R. Lasser^2
1Hospital Clinic, University of Barcelona, Barcelona, Spain, ²Janssen Pharmaceutica Products, Titusville, NJ, USA

Objective: Evaluate efficacy and safety of long-acting injectable risperidone (Risperdal Consta™) in stable patients with schizoaffective disorder.
Methods: After a 2-week run-in period during which patients received flexible doses of oral risperidone 1–6 mg, patients received intramuscular injections of 25, 50, or 75 mg of long-acting risperidone every 2 weeks for 50 weeks. Efficacy and safety measures included the Positive and Negative Syndrome Scale (PANSS) and Extrapyramidal Symptom Rating Scale (ESRS).
Results: Of the 725 patients receiving treatment in the study (RIS-INT-57), 110 had schizoaffective disorder: the latter are the subjects of this analysis. Mean age was 43.5 years; 52.7% were women; 67.3% completed the trial. The baseline mean (SE) PANSS total score was 62.3 (1.73). Mean PANSS total scores improved significantly (P < .001) from baseline at weeks 12 (–7.4), 24 (–10.0), 36 (–8.4), 50 (–10.2), and endpoint (–7.9). Significant reductions were observed at endpoint in mean PANSS scores for positive symptoms (–1.7, P < .01), negative symptoms (–2.9, P < .001), anxiety/depression (–1.2, P < .01), and disorganized thoughts (–1.5, P < .001). A ≥20% reduction in PANSS total scores was seen in 56.2% of patients at endpoint. According to Clinical Global Impressions scores, the proportion of patients rated not ill or with very mild or mild symptoms increased from 54.6% at baseline to 76.6% at endpoint. Significant improvements (P<0.001) from baseline ESRS scores were noted on the overall subjective rating and the parkinsonism exam. The most common adverse events were insomnia in 36.4%, anxiety in 30.0%, depression in 25.5%, and psychosis in 24.6%. The incidence of adverse events was reduced substantially from months 1–3 to months 10–12.
Conclusions: Stable patients with schizoaffective disorder can be safely and effectively switched from current treatment to long-acting risperidone.
Keywords: 1) Long Acting Risperidone 2) Schizoaffective Disorder 3) Extrapyramidal Symptoms

208. Safety and Tolerability of Lamotrigine in Bipolar I Disorder
E. Vieta¹, M. Berk², G. Asnis³, G. Sachs⁴, C. Bowden⁵, J. Calabrese^6
1University of Barcelona, Barcelona, Spain, ²University of Melbourne, Melbourne, Australia, ³Albert Einstein College of Medicine, New York, NY, ⁴Harvard University, Boston, MA, ⁵University of Texas Health Science Center at San Antonio, TX, ⁶Case Western Reserve University, Cleveland, OH, USA

Introduction: Tolerability and safety are important considerations in optimizing pharmacotherapy for bipolar disorder. We examined the tolerability of lamotrigine in a large bipolar disorder clinical trial database.
Methods: Pooled safety data were analyzed from 8 placebo-controlled, double-blind clinical studies, in which adults with DSM-IV bipolar disorder received lamotrigine at doses of 50-500mg (n=827) or placebo (n=685), for up to 18 months.
Results: Lamotrigine was well-tolerated with an adverse event profile comparable to placebo. The most common adverse event was headache (25% LTG, 21% PBO). Few patients experienced serious adverse events (8% LTG, 7% PBO), and the incidence of study withdrawals due to adverse events was low (12% LTG, 10% PBO). Lamotrigine did not destabilize mood and was not associated with sexual side effects, clinically significant weight changes, or withdrawal symptoms. The incidence of non-serious rash with lamotrigine (9%) did not differ from that with placebo (8%). Serious rash was rarely (0.1%) reported for lamotrigine.
Conclusions: Lamotrigine appears to be well tolerated as prophylaxis treatment in bipolar I disorder.
Keywords: 1) Lamotrigine 2) Bipolar Depression 3) Safety/tolerability

209. Neuropsychological Functioning of Children Diagnosed with Bipolar Disorder
G.T. Voelbel¹, M.E. Bates^1,2, G. Pandina³, E. Frenkle², L. Hyers², R. Hendren^4
1Rutgers University, ²UMDNJ-Robert Wood Johnson Medical School, ³Janssen Pharmaceutical Products, ⁴University of California-Davis Mind Institute, USA

There is limited knowledge about the cognitive deficits associated with bipolar disorder in children, and whether deficits are uniformly manifested in younger and older aged children. This study examined the cognitive and neuropsychological performance of children between the ages of 7-13 years who were diagnosed with Bipolar Disorder (BD) using the Kiddie Schedule of Affective Disorders and Schizophrenia (K-SADS: Ambrosini version). Based on the adult literature, it was hypothesized that the BD group would show reduced levels of neuropsychological performance primarily on memory and executive function tests. A comprehensive cognitive and neuropsychological battery was administered to 35 BD participants (25 male) and 23 matched comparison participants (15 male) without psychiatric diagnoses. The two groups did not differ by gender, age, or socioeconomic status. The mean age for the bipolar group was 9.94 years (SD = 1.88) and 10.48 years (SD = 1.76) for the comparison group. Preliminary analyses revealed that the BP group performed less well in terms of Full-Scale IQ, Verbal IQ, the Verbal Comprehension Index, Freedom of Distractibility Index, and Processing Speed Index. The BP group also performed significantly less well on executive function, attention, and verbal and nonverbal memory tests with age, gender, and socioeconomic status controlled. Statistically significant age by diagnostic group interactions suggested that between the ages of 7 to 13 years, relative deficits in verbal and nonverbal memory associated with bipolar disorder diminished, while executive function deficits conversely became more pronounced. Results were consistent with the hypothesis that bipolar disorder increases risk for frontal lobe and limbic area changes, which may contribute to clinical and behavioral presentation. In view of past evidence for adverse effects of mood stabilizers on memory and concentration, analyses are under way to also examine the influence of these medications on cognitive functions within this young population.
Keywords: 1) Children Diagnosed with Bipolar Disorder 2) Neuropsychological Functioning 3) Mood Stabilizers

210. The Psychosocial Impact of Bipolar Disorders
K.D. Wagner¹, J.R. Calabrese², R.M.A. Hirschfeld¹, M.A. Frye³, M. Reed^4
1University of Texas Medical Branch, Galveston TX, ²Case Western Reserve University, Cleveland, OH, ³University of California Los Angeles, Los Angeles, CA, ⁴Vedanta Research, Chapel Hill, NC, USA

Objective: To determine the impact of bipolar I and II disorders (BPD) in the United States (US).
Methods: 3059 subjects were selected from a large epidemiological study of bipolar prevalence that used the Mood Disorder Questionnaire (MDQ) to screen for BPD. Subjects were surveyed using the Social Adjustment Scale-SR, Sheehan Disability Scale, and Family History Screen. Comorbid disease data were also collected.
Results: The response rate was 80% (MDQ+ 1167, MDQ- 1283). MDQ+ subjects reported significantly (p<0.0001) more difficulties with work-role, social/leisure activities, and extended family interactions compared with MDQ- subjects. Positive screens for BPD were associated with more days of disruptive symptoms (p<0.0001) and risk of being fired or laid off (p<0.0001). Younger MDQ+ subjects (18 – 34 years) reported significantly (p=0.003) more symptom days than did older MDQ+ subjects. MDQ+ women reported more disruption in social and family life, while men reported being jailed, arrested, and convicted for crimes. Anxiety (30% vs. 6%), panic attacks (19% vs. 4%), migraine (12% vs. 5%), asthma (17% vs. 9%), obesity (16% vs. 4%), and allergies (43% vs. 29%) were significantly (p<0.0001) more common in MDQ+ versus MDQ- subjects.
Conclusions: Subjects with positive screens for bipolar I and II were more socially and functionally impacted by these disorders.
Keywords: 1) Bipolar Disorder 2) Psychosocial Impairment 3) Burden of Illness

211. Psychosocial Impairment Associated with an Arrest History Among People with Bipolar Disorder
K.D. Wagner¹, J.R. Calabrese², R.M.A. Hirschfeld¹, M.A. Frye³, M. Reed^4
1University of Texas Medical Branch, Galveston TX, ²Case Western Reserve University, Cleveland, OH, ³University of California Los Angeles, Los Angeles, CA, ⁴Vedanta Research, Chapel Hill, NC, USA

Objective: To assess psychosocial impairment associated with a prior arrest or incarceration among people with bipolar disorder (BPD) in the U.S.
Methods: 3059 subjects from a general population epidemiologic study of BPD were mailed a follow-up survey containing: Mood Disorders Questionnaire (MDQ), Social Adjustment Scale, and questions regarding work and prior arrest history (PAH).
Results: There was an 80% response rate to the survey and 1167 screened positive for BPD using the MDQ. MDQ+ subjects with a PAH were more likely to be male (p<.0001) and have lower income (p<.011). Demographics were controlled for in the analysis. 1141 completed the PAH question with 222 reporting “prior history of arrest or jail other than drunk driving”. Compared to no PAH, those with PAH were more likely (p<.01) to endorse MDQ items “excessive, foolish or risky behavior” and “spending money caused problems”, report greater psychosocial impairment (p<.001); report a prior job loss (p<.001) and report problems with alcohol/drug use (p<.001). BPD diagnosis is roughly the same for both groups but those with PAH were more likely to use mood stabilizers (p<.01).
Conclusions: Arrest history among those with BPD is associated with significantly greater psychosocial impairment, and drug abuse.
Keywords: 1) Bipolar Disorder 2) Arrest History 3) Psychosocial Impairment

212. Treatment of Bipolar Disorders – A Psychoeducational Approach
P. Wagner, G. Dietrich, P. Bräunig
Klinikum Chemnitz gGmbH - Klinik für Psychiatrie, Verhaltensmedizin und Psychosomatik, Chemnitz, Germany

Background: There is an increasing recognition that the efficacy of pharmacotherapy is far from satisfactory in the treatment of bipolar disorders. The influence of psychosocial factors on the frequency and course of the disorder is well documented (1, 2) and therefore the interest in psychotherapeutic interventions is growing. Especially psychoeducational elements seem to be effective (1) and are already used as a component in a variety of psychotherapeutic approaches (2). Nevertheless there is still a lack of systematic manualized psychoeducational programs. For this reason we developed a program for bipolar patients consisting of 12 units.
Method: Our psychoeducational program was administered to 18 bipolar outpatients. The treatment was delivered over a period of 2 months since November 2002. Different questionnaires and rating scales were used at the beginning and at the end of the program. Patient measures included those related to global status, symptomatology, quality of life, knowledge about the illness and the acceptance of the program.
Results: Data analysis has just started and final results will be presented in detail on the poster. The preview of the data showed a high acceptance by the patients and an increase of knowledge about their illness.
Conclusion: First results are encouraging and seem to confirm our assumption that the program enables patients to enhance their coping skills to deal with their illness and to improve treatment compliance. The study is still ongoing and for further evaluation a follow-up is planned.
Keywords: 1) Psychoeducation 2) Bipolar Disorder 3) Psychotherapy
References:
1. Colom, F., Vieta, E., Martinez, A. et al. (1998). What is the role of psychotherapy in the treatment of bipolar disorder? Psychotherapy & Psychosomatics 67(1): 3-9.

2. Huxley, N. A., Parikh, S. V. & Baldessarini, R. J. (2000). Effectiveness of psychosocial treatments in bipolar disorder: State of the evidence. Harvard Review of the Psychiatry, 8:126-140.

213. Psychopathologic Features of Rapid Cycling Bipolar Patients
M.M. Wankmuller, K.H. Sutherland, L.E. Oakley, J.F. Goldberg
Bipolar Disorder Research Center, Department of Psychiatry, NY Presbyterian Hospital-Weill Cornell Medical College, New York, NY, USA

Background: The extent to which depressive episodes dominate the illness course of rapid cycling bipolar disorder remains uncertain. Recently, the controversy over classifying rapid cycling as a distinct bipolar subtype, rather than simply a course specifier, has sparked considerable debate. We collected clinical and historical data from a cohort of rapid cycling and non-rapid cycling bipolar patients who presented for treatment in an academic research center.
Method: Thirty-nine consecutive bipolar outpatients meeting DSM-IV criteria for rapid cycling (RC) in the past 12 months and 24 bipolar outpatients who did not meet DSM-IV criteria for rapid cycling in the past 12 months (NRC) were assessed for manic, depressive, and mixed symptoms and for related illness characteristics.
Results: 1) 53% of RC patients sought treatment during mixed manic/hypomanic phases, while 31% presented during pure depressed phases, and 17% during pure manic/hypomanic episodes; 2) RC patients were more likely than NRC to manifest a significant history of comorbid substance abuse (p less than .01) and were also more likely to have a familial history of bipolar disorder (p less than .01); however, there was no significant difference in lifetime suicide rates between RC and NRC subjects; 3) Among the RC subgroup, 54% of episodes in the past year were either manic or hypomanic. Nevertheless, over half took antidepressant medications during that time.
Conclusions: Our results suggest that mixed states, such as depression with concurrent mania or hypomania, may typify clinical presentations in treatment-seeking RC bipolar patients, rather than pure depressive episodes. The suggestion that RC may be a unique subtype of bipolar disorder has important treatment implications. Namely, pharmacologic treatments that target mixed affective symptoms may be more applicable than unimodal treatments that have either antidepressant or antimanic properties.
Keywords: 1) Rapid Cycling 2) Mixed Mania/hypomania 3) Antidepressant Use

214. Relation Between Anxiety and Sleep Quality in Bipolar Disorder
C.A. Winett¹, S.L. Johnson², T.A. Mellman³, M.W. Otto^1
1Massachusetts General Hospital, Boston, MA, ²University of Miami, Coral Gables, FL, ³Dartmouth Hitchcock Medical Center, Lebanon, NH, USA

Introduction: A recent series of papers have documented alarmingly high rates of lifetime anxiety disorders in bipolar disorder (1-3). Prevalence estimates include 51% with any lifetime anxiety disorder, 17% with panic disorder, 9% with agoraphobia without panic disorder, 22% with SAD, 10% with OCD, 17% with PTSD, and 18% with GAD (3). Bipolar and anxiety disorder comorbidity has been associated with increased suicide rates and increased functional impairment (3). One additional concern about the high comorbidity rates of anxiety disorders is the subsequent impact on sleep quality. Sleep disturbances have long been recognized as a characteristic of mood disorders (4, 5), and multiple studies have emphasized the prospective impact of sleep deprivation on subsequent episodes of mania (6, 7). In previous work, we found that sleep disruptions influenced bipolar depression (8), congruent with polysomnography work demonstrating strong links between depression and sleep. Sleep disturbances in anxiety disorders have also been well-documented (9). However, to date, no study has examined the relation between anxiety and sleep quality in bipolar disorder.
Methods: For the current study, a subsample of individuals who participated in a longitudinal study of bipolar I disorder and were followed for at least two months were selected. Mania and depression severity interviews were conducted monthly using the Bech-Rafaelsen Mania Scale (BRMS) and Modified Hamilton Rating Scale for Depression (MHRSD), and anxiety and sleep quality assessments were obtained at months 2, 6, 12, 18, and 24 using the Pittsburgh Sleep Quality Index (PSQI) and Mood and Anxiety Symptom Questionnaire (MASQ).
Results: We will examine links between sleep quality and anxiety symptoms.
Keywords: 1) Sleep 2) Anxiety 3) Comorbidity
References:
1.  Frank, E, Cyranowski, JM, Rucci, P et al.  Clinical significance of lifetime panic spectrum symptoms in the treatment of patients with bipolar I disorder.  Archives of General Psychiatry 2002; 59:905-912.

2.  Kessler, RC, McGonagle, KA, Zhao, S et al.  Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States:  Results from the National Comorbidity Survey.  Archives of General Psychiatry 1994; 51:8-19.

3.  Simon, NM, Otto, MW, Wisniewski SR et al.  Anxiety disorder comorbidity in bipolar disorder:  data from the first 500 STEP BD participants.  Presented at the Annual Meeting of the American College of Neuropsychopharmacology.  San Juan, Puerto Rico, December 2002.

4.  Goodwin, FK & Jamison KR.  Manic-depressive illness.  New York, Oxford University Press, 1990.

5.  Wehr, T.  Sleep and biological rhythms in affective illness.  In:  Biological Rhythms and Mental Disorders.  Kupfer, DJ & Monk, TH, eds. New York, Guilford Press, 1988.

6.  Wehr, T. Sleep loss:  a preventable cause of mania and other excited states.  Journal of Clinical Psychiatry 1989; 50:8-16.

7.  Wehr, T, Goodwin F, Wirz-Justice, A, Breitmaier, J & Craig, C.  48-hour sleep-wake cycles in manic-depressive illness:  naturalistic observations and sleep deprivation experiments.  Archives of General Psychiatry 1982; 39:559-565.

8.  Winett, CA, Johnson, SL & Mellman, TA.  The prospective impact of sleep disturbance on depression and mania: defining a temporal relationship.  Manuscript in preparation.

9.  Ohayon, MM & Roth, T.  Place of chronic insomnia in the course of depressive anxiety disorders.  Journal of Psychiatric Research 2003; 37:9-15.

215. Canadian Network for Bipolar Disorder (CAN-BD): Preliminary Report of Data on the First 94 Patients
L.N. Yatham¹, P. Cervantes², S. Beaulieu³, M. Connolly⁴, V. Sharma^5
1University of British Columbia, Vancouver, ²Montreal General Hospital, Montreal, ³Clinique Verdun, Verdun, ⁴Royal Jubilee Hospital, Victoria, ⁵Regional Mental Health Care London, London, Canada

Background: CAN-BD began collecting prospective data to track clinical outcomes, treatment patterns, quality of life and resource utilization of bipolar patients treated at 15 centres across Canada about a year ago and the results on the first 94 patients will be reported here.
Method: Patients diagnosed with Bipolar I/II disorder, who within the past 3 months met criteria for a mood episode and required a change in treatment, are eligible. At baseline, data collected includes demographics, clinical/medical history, psychiatric medications, life events, symptom rating scales such as MADRS, YMRS, CGI, etc and side effect scales. Patients are managed under conditions of routine clinical practice and all behavioural scales are administered at least every 3 months, and YMRS and MADRS at each clinical visit. Data is collected directly into a database with some variables collected using paper CRFs and monthly patient diaries.
Results: Of the first 94 patients recruited, 54 are women, 40 are men. Of these, 59 have a diagnosis of BD I and 35 BD II. The mean age of the patients is 45.5. BD II patients showed a significantly earlier onset of depressive symptoms (p<0.005, mean age 17 vs. 24 years) but not hypomanic symptoms and were significantly more likely to be diagnosed at a later age (41 vs. 35 years, p<0.05). BD II patients had also experienced, on average, more hypomanic (p<0.001) and depressive episodes in the 2 years prior to study entry (p<0.05), and significantly higher scores at enrollment on the HAMD-21 and MADRS (p<0.05), but not on the YMRS.
Conclusion: BD II’s experience a longer course of illness, more depressive and hypomanic episodes, and greater severity of depressive symptoms. Considering that this population is also being diagnosed at a significantly later age than BD I patients, these findings suggest that patients with BD II are being under-diagnosed and under-treated.
Keywords: 1) Bipolar Disorder 2) Treatment Patterns 3) Clinical Outcomes

216. The Effects of Lithium and Lamotrigine in Prevention of Relapse/Recurrence of Bipolar I Depression
L. Yatham¹, G. Goodwin², C. Bowden³, J. Calabrese⁴, R. White⁵, R. Leadbetter⁵, A. Metz^5
1University of British Columbia, Vancouver, Canada, ²University of Oxford, Cambridge, UK, ³University of Texas Health Science Center at San Antonio, San Antonio, TX, ⁴Case Western Reserve University, Cleveland, OH, ⁵GlaxoSmithKline, Research Triangle Park, NC, USA

Introduction: Lithium, commonly used to treat bipolar mania, is believed to have antidepressant activity. Two large placebo-controlled prophylaxis trials recently demonstrated that lamotrigine, but not lithium, were superior to placebo in prolonging time to intervention for bipolar depression. We compared the incidence of relapse/recurrence of depression in lithium and lamotrigine treatment groups with placebo.
Methods: Placebo (PBO), lithium (Li), and lamotrigine (LTG) were studied as prophylaxis treatments for 18 months in bipolar I patients who were currently or recently symptomatic. Results from 2 clinical trials comprising 463 currently or recently depressed patients and 175 currently or recently manic patients were examined for incidence of depressive events, HAMD-17 scores, and DSM-IV depressive episodes by treatment group and index mood episode.
Results: In recently manic patients, fewer lamotrigine-treated patients required intervention for depression compared with placebo (PBO 30%, LTG 14%, Li 22%; PBO vs. LTG p=0.034), reported depressive adverse events (PBO 3%, LTG 0, Li 4%), had DSM-IV depression (PBO 28%, LTG 10%, Li 17%; PBO vs. LTG p=0.024), or had HAMD scores > 20 (PBO 19%, LTG 3%, Li 11%; PBO vs. LTG p=0.011). In recently depressed patients, the incidence of depressive symptoms varied numerically between treatment groups: intervention for depression (PBO 39%, LTG 34%, Li 38%), reported depressive adverse events (PBO 2%, LTG 4%, Li 3%), DSM-IV depression (PBO 36%, LTG 31%, Li 36%) or HAMD scores > 20 (PBO 26%, LTG 22%, Li 18%). Conclusions: Lamotrigine’s ability to protect against depressive symptoms in currently or recently manic patients suggests that administration of lamotrigine should be considered during or shortly after stabilization of mania, before depressive symptoms occur.
Keywords: 1) Lamotrigine 2) Bipolar Depression 3) Efficacy

217. Topiramate Treatment of Comorbid Tourette’s and Mood Disorders: Preliminary Report
C.M. Young, J.F Leckman, L.D. Scahill
Yale University School of Medicine, New Haven, CT, USA

Objective: To evaluate the efficacy and safety of topiramate in patients with comorbid tic and mood disorders.
Background: Recognized comorbidities of Tourette’s disorder (TD) include OCD, ADHD, and bipolar disorder. Broad-spectrum pharmacologic agents may benefit these multisymptomatic patients. Topiramate, indicated for epilepsy in children and adults, has shown efficacy for TD and as adjunctive therapy for bipolar disorder.
Methods: This was an open-label add-on trial of 8 subjects (6 male, 2 female; ages 9-43 yrs) with DSM-IV–diagnosed TD. Six patients met criteria for comorbid OCD; 4 of these met criteria for a mood disorder (3 bipolar and 1 chronic irritability/w outbursts). Two patients did not have comorbid OCD or mood disorder. At trial outset, 1 patient was on an SSRI, 3 on atypical antipsychotics, 3 on mood stabilizers, and 3 on no medication. Subjects were started on topiramate 25 mg/day, titrated by 25 mg/wk in twice-daily doses to a maximum of 200 mg/day. Follow-up ranged from 30 to 90 days. Rating measures were taken at baseline and final visit. Measures included Yale Global Tic Severity Scale– Total Tic Score (YGTSSTIC), Young Mania Rating Scale (YMRS), and Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS).
Results: Five of 8 subjects had improved YGTSSTIC scores (mean reduction 11.4[SD 10.9]; paired-t=2.97,7; p<0.02), including 2 of the 4 subjects without comorbid mood disorders. All 4 subjects with comorbid mood disorders had improved YMRS scores (mean reduction 11.8[SD 9.6]; paired-t=3.01,5; p<0.03). CY-BOCS changes were not statistically significant. The average daily topiramate dose was 100 mg/day. The most common side effect was mental confusion (5/8 subjects, including 2 responders). One subject discontinued due to appetite suppression and weight loss.
Conclusion: Topiramate may provide an alternative add-on therapy for patients with comorbid refractory TD and bipolar disorder, and may have independent benefits for chronic tics.
Keywords: 1) Topiramate 2) Tourette’s Disorder 3) Mood Disorders

218. Evaluating the Young Mania Rating Scale in a Comorbid Sample of Outpatients Age 5 to 17 Years
E.A. Youngstrom, C.K. Danielson, R.L. Findling, J.R. Calabrese
Case Western Reserve University/University Hospitals of Cleveland, Cleveland, OH, USA

Introduction: There is growing interest in pre-pubertal bipolar disorder, but there are currently no mania rating scales specifically designed for use with children and adolescents. The goal of the present study was to evaluate thoroughly the psychometrics of the YMRS in a large sample of outpatients presenting with a variety of different DSM-IV diagnoses, including frequent comorbid conditions.
Method: Participants were 612 youths (including 379 males and 88 African-Americans) presenting to an outpatient clinic. Diagnoses and YMRS ratings were determined by a semi-structured (KSADS) interview involving the parent and then youth sequentially.
Results: Horn’s Parallel Analysis and Minimum Average Partials indicated one factor solutions. Multigroup confirmatory factor analysis confirmed the one factor structure across male and female pre- and post-pubescent samples, Chi-squared (176) = 902.8, CFI = .91, RMSEA = .082. Alpha for the total score was .91. Three items showing poor item-total correlations and factor loadings: bizarre appearance, lack of insight into the illness, and hypersexuality. Regression analyses indicated that scores decreased by a half point per year of age (p < .0005), males scored similar to females (p > .05), and children with ADHD scored approximately five points higher than children without ADHD (p < .0005), whereas children with bipolar disorder scored approximately 20.3 points higher, even after controlling for comorbid ADHD.
Discussion: The YMRS showed acceptable general psychometric performance in terms of reliability, factor validity, and discriminative validity (i.e., total scores were higher for bipolar than for any other diagnostic group, including ADHD). However, several of the items perform poorly in a juvenile sample, probably reflecting developmental differences in symptom expression or the inappropriateness of applying adult anchors to the presentation of these symptoms in children. Although the YMRS appears adequate for use in juvenile samples, a more developmentally sensitive measure could be constructed.
Keywords: 1) Juvenile 2) Mania 3) Psychometrics

219. Rating Scales Discriminate BPSD versus ADHD in a Pre-Pubescent Clinical Sample
E.A. Youngstrom, R.L. Findling, S.Y. Kahana, C. Demeter, M. Price, G.M. Woldorf, J.R. Calabrese
Case Western Reserve University/University Hospitals of Cleveland, Cleveland, OH, USA

Introduction: There is considerable controversy about the relationship between ADHD and bipolar disorder in pediatric samples. The goal of the present study was to examine how well six measures (Achenbach CBCL Externalizing, CBCL Attention Problems, TRF Externalizing, TRF Attention Problems; Parent YMRS, Parent GBI) discriminated ADHD (including comorbid unipolar depression) from bipolar disorder (including bipolar I, II, NOS, and cyclothymia) in a pre-adolescent sample. Research questions included: (1) What is the diagnostic efficiency of each measure in detecting bipolar disorder? (2) Are there significant differences between the diagnostic performances of the measures?
Method: Participants were 248 youths aged 5 to 10 (M = 8.3 years; 71% male, 80% white; 97 bipolar I (39%), 63 other bipolar (25%), 88 with ADHD and no bipolar disorder (36%); 85% of total sample had ADHD due to comorbid ADHD + bipolar).
Results: Both groups had significant elevations of attention and externalizing problems according to parents and teachers (lowest average T-score = 62.5, TRF-Externalizing). All three parent reports significantly discriminated ADHD versus bipolar (including bipolar with comorbid ADHD). The bipolar group had significantly higher scores on the P-YMRS, CBCL Externalizing, and P-GBI (AUCs= .79, .77, and .76, all p’s < .0005). The performance of the three parent scales did not differ significantly (comparison ps>.05), but all three were significantly better than the TRF Externalizing or Attention problems (comparison ps<.0005). The TRF externalizing did not separate the two groups (AUC=.48, n.s.), nor did either parent or teacher report of attention problems (AUCs=.53 and .42).
Discussion: Findings suggest that parent report is a crucial component of diagnosing bipolar disorder in young children, contrary to clinical wisdom. Teachers reported similar levels of concern for both groups and did not separate ADHD from bipolar youths, consistent with prior findings that teachers interpret any disruptive behavior as reflecting attention problems.
Keywords: 1) Juvenile 2) Mania 3) Psychometrics

220. Predictive Value of Anxiety Symptoms for the Maintenance Treatment Response in Bipolar I Disorder
W. Zhang^1,2, R.A. Leadbetter², T.C. Spaulding², Z. Antonijevic², J.R.T. Davidson^1,2, A. Metz^2
1Duke University Medical Center, Durham, NC, ²GlaxoSmithKline, Research Triangle Park, NC, USA

Objective: To examine the predictive value of anxiety symptoms for response in the maintenance treatment of bipolar I disorder.
Methods: 638 currently or recently symptomatic bipolar I patients (DSM-IV) were stabilized and randomized to 18 months of double-blind monotherapy with lamotrigine (n=280; 50-400mg/day fixed and flexible dose), lithium (n=167; 0.8-1.1mEq) or placebo (n=191). The effects of anxiety symptoms as measured by the baseline psychic anxiety score (BPAS) and baseline somatic anxiety score (BSAS) on the Hamilton Depression Rating Scale (HAMD) were examined against time to intervention for a mood episode (TIME) using a Cox Proportional Hazard Model.
Results: BPAS=0 was predictive of an overall response when analyses were controlled by index mood (mania or depression) and treatment (p=0.002). BPAS=0 was predictive of a treatment response within lamotrigine (p=0.04) or lithium (p=0.04) but not the placebo (p>0.05) group; while BSAS=0 only predicted response within the placebo group (p=0.04). When compared with placebo, lamotrigine (p=0.003) and lithium (p=0.001) delayed TIME significantly among patients with BPAS=0 or BSAS=0.
Conclusions: Absence of baseline psychic or somatic anxiety symptoms predicted the most favorable response in the lamotrigine and lithium maintenance treatment of bipolar I disorder, consistent with previous reports for the acute treatment of bipolar I disorder. In addition, baseline somatic anxiety symptoms were associated with a high placebo response.
Keywords: 1) Lamotrigine 2) Bipolar Disorder 3) Anxiety