Before reviewing the differential diagnosis of bipolar disorder, it is important to understand the epidemiology, incident, and prevalence of the disorder as well as to divide it by type: type I, type II, or the phenomenon of not otherwise specified, which is defined as subthreshold mania or hypomania. Subthreshold mania or hypomania entails fewer than the requisite number of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition symptoms of mania or hypomania, or an insufficient duration of time with which to make a diagnosis of a syndrome of mania or hypomania. Currently, the most reliable epidemiology of prevalence rates were reported in a replication study of 9,282 respondents to the National Comorbidity Survey conducted by Merikangas and colleagues.

Those data show there is a lifetime prevalence rate of bipolar I disorder of 1 % and bipolar II disorder of 1.1%. In this study, not otherwise specified was labeled subthreshold bipolar disorder as the illness duration was not long enough or the number of symptoms was not extensive enough to constitute syndromal mania or hypomania. This type had a prevalence rate of 2.4%, for a total lifetime prevalence of ~4.4%. This rate is somewhat higher than has been described in the past, but included the subdivision that researchers believe comprises the bipolar spectrum.

There has been much underdiagnosis or lack of diagnostic accuracy of bipolar disorder, with relative overdiagnosis of major depressive disorder, which historically has been well documented. However, in the last 10 years, there has been a substantial increase in the making of diagnoses of bipolar disorder in various clinical settings. One dataset examined ambulatory medical care survey data and found that among adults over an ~10-year period, there was approximately a doubling in the incidence of diagnosing bipolar disorder of any type (Slide 1).

Bipolar disorder, particularly bipolar type I disorder, is at least as highly comorbid with other psychiatric and behavioral disorders as any Axis I medical disorder (Slide 1). Two iterations of the National Comorbidity Survey study, which is epidemio-logic-based and not anecdotal evidence from clinical offices and hospital emergency rooms, have shown that these data are reflective of the overall population in the United States and potentially for all other countries across the world.

In particular, if the prevalence rates of all anxiety disorders are grouped together, they are almost as prevalent as bipolar disorder itself, and clinicians rarely see a patient with bipolar disorder who does not have an anxiety disorder. This finding raises the question that anxiety may not be a separate entity, but an additional fundamental component of bipolar symptomatology, at least for a substantial number of patients. Regardless, the role of anxiety in bipolar disorder is unique. Perugi and colleagues studied the time sequence of different anxiety disorders in relation to first presence and clinical recognition of bipolar disorder (Slide 2). In contrast to panic disorder/agoraphobia and obsessive-compulsive disorder (OCD), social anxiety or social phobia was shown to be prevalent in 94.7% of patients prior to onset of hypomania and the clinically recognizable problematic school and home situations before the diagnosis of the bipolar disorder. This finding suggests that there may be some fundamental aspect of social anxiety and other anxiety disorders that is a ties. These patients tend to be more unstable symptomatically and have multiple comorbidities.

Regarding the treatment of patients with bipolar disorder, positive findings have shown that patients engaged in treatment may live longer. Angst and Sellaro studied standardized mortality ratios for patients with bipolar disorder and found a significant reduction in mortality ratio in our bipolar patients who were actively in treatment (Slide 1). Data from the Collaborative Depression Study by Judd and colleagues showed that bipolar patients are quite symptomatic. Judd and colleagues followed bipolar patients over a 13-year period and found that these patients were symptomatic ~50% of the study period. What were these patients most likely to be symptomatic with: bipolar depression (Slide 2). Thus, it is important to focus on bipolar depression and examine placebo-controlled trials that were adequately powered in its treatment.

Over the last decade, there have been positive trials studying lamotrigine, quetiapine, olanzapine, and olanzapine-fluoxetine combination. Two of these medications are United States Food and Drug Administration—approved in bipolar disorder treatment: quetiapine and olanzapine-fluoxetine. There are a number of negative trials in bipolar depression, including a number of antidepressant monotherapy trials and antidepressants in combination with mood stabilizers (Slide 3). However, what is the most frequent pharmacotherapy treatment used for bipolar depression? It is the use of antidepressants alone; antidepressant monotherapy is twice as commonly prescribed as mood stabilizers (Slide 4). Thus, it is necessary for clinicians to possess knowledge of the evidence and results from both positive and negative antidepressant trials for bipolar disorder.

Medication is the mainstay of treatment for bipolar disorder. However, no medication will be effective if patients do not take it, and the rates of medication compliance in bipolar disorder are very low. Johnson and McFarland found that the modal length of compliance with a mood stabilizer was only 2 months. Keck and colleagues found that 50% to 66% of patients with bipolar disorder exhibit poor compliance within the first 12 months of treatment. In addition, even with adequate medication compliance, high rates of relapse persist.

Adjunctive psychosocial treatments can help reduce relapse and provide patients as well as their families with tools to manage bipolar disorder more effectively. Several forms of intensive psychotherapy have shown promise for the treatment of bipolar disorder. In the Systematic Treatment Enhancement Program for Bipolar Disorder, Miklowitz and colleagues compared three forms of intensive interventions: cognitive-behavioral therapy (CBT), interpersonal and social rhythm therapy, and family-focused treatment. These were compared to a brief, 3-session psychoeducational intervention known as collaborative care. A total of 293 depressed patients with bipolar type I or type II disorder were treated with protocol pharmacotherapy and were randomly assigned to either one of the three intensive interventions or the brief psychoeducational intervention.

The three intensive interventions provided up to 30 sessions of treatment over a 9-month period. The collaborative care intervention consisted of three sessions administered over a 6-week period. The authors found that patients who received one of the intensive interventions had a median time to recovery 110 days earlier than patients who had received the collaborative care conditions. Patients who received one of the three intensive psychotherapies also had significantly higher year-end recovery rates, and are more than 1 to 1.5 times more likely to be clinically well during any study month. No statistically significant differences were found between the 3 intensive treatments.