Chapter 6 David J. Miklowitz and Sheri L. Johnson Over the past two decades, there has been a considerable resurgence of interest in bipolar disorder (BD; formerly known as manic-depressive illness). This resurgence is attributable in part to the availability of new pharmacological agents, disorder-specific psychosocial treatments, and data on genetic mechanisms and neurophysiological and neuroanatomical correlates. It has also been driven by the increasing recognition that the onset of the disorder is often in childhood or adolescence. In this chapter we describe the disorder from the vantage points of diagnostic criteria, diagnostic controversies, epidemiology, and course and prognosis, with particular attention to developmental considerations pertinent to early-onset bipolar illness. A case study illustrates these issues. We discuss the etiology and prognosis of BD from both a genetic and a psychosocial viewpoint. Current etiological models view bipolar disorder as a primarily genetic illness whose onset can be elicited by environmental stressors, although the nature of the inherited biological vulnerability is unclear. Although few studies have examined psychosocial stressors relevant to the onset of bipolar disorder, there is now a considerable literature on psychosocial stressors that affect the course and outcome of the disease. In the final sections, we summarize the major recent findings concerning the treatment of the disorder and offer directions for further research. The core symptoms believed to constitute bipolar mania—elation, grandiosity, and hyperactivation—have not fundamentally changed from one edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) to the next. In this section, we describe the core features of mania/hypomania and depression and provide an update on the rules of classification according to DSM-5 (American Psychiatric Association [APA], 2013). Bipolar disorder (BD) is defined by manic symptoms. According to the DSM-5(APA, 2013), people with bipolar, manic episodes experience elated, expansive, or irritable mood (or any combination of these) and increased activity, plus at least three (four if the mood is only irritable) of the following symptoms: decreased need for sleep; racing thoughts or flight of ideas; rapid speech; inflated self-esteem (also called grandiosity); impulsive, reckless behavior (e.g., spending sprees, hypersexuality); and distractibility. These symptoms must be present for at least 1 week or interrupted by hospitalization or emergency treatment. They must also cause functional impairment. Hypomania is characterized by parallel symptoms, but the criteria specify only that the symptoms last at least 4 days and result in a distinct, observable change in functioning rather than severe impairment. Although persons with bipolar disorder often experience both depression and mania, a person who has only had one lifetime manic episode and no depression is still diagnosed with bipolar I disorder. DSM-5 criteria differ from DSM-IV-TR (APA, 2000) in listing increased activity as one of the cardinal symptoms of mania and hypomania. It has been argued that changes in activity may be more easily recognized than changes in subjective mood state. As such, it is hoped that the DSM-5 criteria may provide for better accuracy and reliability in the detection of manic and hypomanic episodes. Depressive episodes, when present, last for at least 2 weeks and are characterized by at least five symptoms, of which sad mood or loss of interest or pleasure in daily activities must be one (criterion A symptoms), plus insomnia or hypersomnia, psychomotor agitation or retardation, increases or decreases in weight or appetite, loss of energy, difficulty concentrating or making decisions, feelings of worthlessness, and suicidal ideation or behavior (criterion B symptoms). Depression must also be associated with functional impairment. Manic or depressive episodes that are clearly related to an ingested substance or to biological treatments—including antidepressant medications—are classified as substance-induced mood disorders. A major change in the DSM-5 is the classification of episodes in which manic and depressive symptoms occur simultaneously—what used to be called a mixed episode. The DSM-5 refers to mixed features as a course specifier that can occur in mania or depression, whether of the unipolar or bipolar variety. A patient with mania has mixed features if he or she has three co-occurring symptoms of depression. If he or she has major depressive disorder, the mixed specifier is met by having three simultaneous manic or hypomanic symptoms. Thus, a manic patient meets the mixed specifier criteria if he or she has irritable mood, increased activity, decreased need for sleep, impulsive behavior, grandiosity, and pressure of speech (all manic symptoms) along with loss of interests, suicidal thinking, and feelings of worthlessness (depressive symptoms). The new criteria, although more complicated than those in the DSM-IV, are intended to match what clinicians see in practice. Bipolar I disorder is defined by the presence of a single manic episode that is not substance-induced (see the following case study). In other words, patients need not have experienced a major depressive episode to be called bipolar I. Rates of unipolar mania are between 25% and 33% in community samples but only about 10% in clinical samples (Depue & Monroe, 1978; Karkowski & Kendler, 1997; Kessler, Chiu, Demler, & Walters, 2005; Weissman & Myers, 1978). There is some evidence, however, that most patients with unipolar mania eventually develop depressive episodes. In a 20-year study of unipolar mania, 20 of 27 patients had episodes of depression during the follow-up period (Solomon et al., 2003). Bipolar II disorder is characterized by major depressive episodes alternating with hypomanic episodes. Both must be present; one cannot be diagnosed with bipolar II disorder on the basis of hypomanic episodes alone. About 1 in 10 bipolar II patients eventually develops a full manic or mixed episode over a 10-year follow-up, and, thus, converts to bipolar I disorder (Coryell et al., 1995). Cyclothymia is a variant of bipolar disorder characterized by 2 or more years (or 1 year among adolescents) of alternations between hypomanic and depressive symptoms, but none of these alternations meet the full DSM-5 criteria for a hypomanic or major depressive episode. Bipolar disorder, not elsewhere classified (previously “not otherwise specified”) is reserved for patients whose disorder meets the minimum number of required symptoms but not the duration requirements for a full manic, hypomanic, or depressive, episode. Many childhood-onset patients receive this subthreshold diagnosis, and up to 50% “convert” to bipolar I or II disorder over 5 years (Axelson, Birmaher, Strober, et al., 2011). Differences in epidemiological estimates vary across studies, depending in part on culture and on how broadly the bipolar spectrum is defined. A multinational study conducted by the World Health Organization reported lifetime prevalence rates in 61,392 adults in 11 countries: 0.6% for bipolar I disorder, 0.4% for bipolar II disorder, and 1.4% for subthreshold bipolar disorder. The highest rates were observed in the United States; lifetime prevalence rates were 1.0% for bipolar I disorders, 1.1% for bipolar II, and 2.4% for “subthreshold” BD (Merikangas et al., 2007). Although there are fewer epidemiological data on cyclothymic disorder, it is believed to affect as much as 4.2% of the general population (Regeer et al., 2004). The onset of mood disorders appears to be getting younger in successive birth cohorts (Kessler et al., 2005; Wickramaratne, Weissman, Leaf, & Holford, 1989). Kessler et al. (2005) reported that the lifetime risk of bipolar I or II disorders in 18- to 29-year-olds was 22 times higher than in persons over 60. It is possible, however, that younger persons feel less stigmatized by psychiatric symptoms and are more likely than older persons to report mood disorder (notably manic) symptoms. In a large (N = 10,123) community sample of adolescents in the United States, 2.5% met lifetime DSM-IV criteria for bipolar I or II disorder (Merikangas et al., 2012). Although many investigators suspect that bipolar disorder is more likely to be diagnosed by practitioners in U.S. children (e.g., Reichart & Nolen, 2004), the base rates of the disorder appear comparable across countries when standardized diagnostic interviews are used. An aggregate analysis of 12 clinical studies of 16,222 children (ages 7 to 21) across the world reported a rate of bipolar spectrum disorder (1.8%) that did not significantly differ across countries (Van Meter, Moreira, & Youngstrom, 2011). The prevalence of bipolar spectrum disorders (i.e., bipolar disorder not elsewhere classified, bipolar I, and bipolar II) may be as high as 6% in U. S. treatment-seeking samples of youth (Youngstrom, Findling, Youngstrom, & Calabrese, 2005). The mean age of onset of bipolar I disorder averages 18.2 years, and 20.3 years for bipolar II (Merikangas et al., 2007), but there is substantial variability. Between 50% and 67% of patients develop the disease by age 18, and between 15% and 28% before age 13 (Perlis et al., 2004). Earlier age at onset is associated with rapid cycling and other negative outcomes in adulthood (Coryell et al., 2003; Schneck et al., 2004). Women and men are equally likely to develop bipolar I disorder. Women, however, report more depressive episodes than men and, correspondingly, are more likely to be diagnosed with bipolar II disorder (e.g., Leibenluft, 1997; Schneck et al., 2004). Women are also more likely to meet experience rapid-cycling BD (at least four episodes per year), again potentially related to the greater vulnerability to depression (Schneck et al., 2004). Less is known about racial and ethnic differences in the diagnosis and treatment of BD. In the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD), African American and Latino patients received fewer prescriptions for psychiatric medications than European American patients (Gonzalez et al., 2007). Moreover, African American bipolar patients were less likely than Caucasian patients to have an outpatient follow-up visit within 3 months of the initial diagnosis (Kilbourne et al., 2005). Evidence suggests major racial disparities in the treatment of bipolar disorder. In the NCS-R major epidemiological study, none of the African American persons who were diagnosed with bipolar disorder were receiving adequate treatment (Johnson & Johnson, 2014). One study found that adult African American patients were less likely than Caucasians to be prescribed mood stabilizers or benzodiazepines and more likely to have been given antipsychotic medications (Kupfer, Frank, Grochocinski, Houck, & Brown, 2005). African American adolescents with bipolar disorder are treated for longer periods than Caucasian adolescents with atypical antipsychotics, even after adjusting for the severity of psychotic symptoms (Patel, DelBello, Keck, & Strakowski, 2005). Not surprisingly given the treatment deficits, the course of BD illness may be worse among African American patients, who are more likely to have attempted suicide and been hospitalized than Caucasian patients. The reasons for these racial disparities in treatment are unclear, but are not explained by lack of insurance, lack of treatment-seeking, or lack of acknowledgment of symptoms among African American patients (Johnson & Johnson, 2014). Recent studies have examined the factor structure of the manic syndrome. A principal component analysis of data from 576 diagnosed manic patients identified seven stable underlying factors: depressive mood, irritable aggression, insomnia, depressive inhibition, pure manic symptoms, emotional lability/agitation, and psychosis (Sato, Bottlender, Kleindienst, & Moller, 2002). Through cluster analysis, Sato et al. identified four phenomenological subtypes of acute mania: pure, aggressive, psychotic, and depressive-mixed mania. The pattern of manic symptoms has been investigated in youths with bipolar disorder. A meta-analysis of seven studies examining the phenomenological characteristics of mania among children and adolescents (aged 5 to 18; Kowatch, Youngstrom, Danielyan, & Findling, 2005) found that the most common symptoms during manic episodes were increased energy, distractibility, and pressure of speech. Approximately 80% showed irritability and grandiosity, whereas 70% had the “cardinal” manic symptoms of elated mood, decreased need for sleep, or racing thoughts. Less common symptoms included hypersexuality and psychotic symptoms. Thus, most manic children showed symptoms that also characterize adult mania. Bipolar disorder is associated with multiple threats to health and livelihood, but the most fundamental concern is the risk of suicide. The true prevalence of suicide in BD is unclear. Early estimates suggested that rates of completed suicide in BD were 12 to 15 times higher than in the general population (Angst, Angst, Gerber-Werder, & Gamma, 2005; Harris & Barraclough, 1997; Jamison & Baldessarini, 1999) and 4 times higher than rates among patients with recurrent major depression (Brown, Beck, Steer, & Grisham, 2000). In a 44-year follow-up of patients with major depressive disorder and BD, 11% died by suicide (Angst et al., 2005). Rates are especially high among younger, recent-onset male patients and those who have comorbid alcohol or substance abuse, social isolation, depression, significant anxiety, aggression, impulsiveness, a family history of suicide, or combinations of these (Angst et al., 2005; Fawcett, Golden, & Rosenfeld, 2000; Jamison, 2000). Some of these figures may overestimate the true suicide prevalence in BD (Bostwick & Pankratz, 2000). First, estimates based on “proportionate mortality prevalence” (percentage of the dead who died by suicide) tend to be higher than the “case fatality prevalence” (proportion of people who died by suicide). The former method is particularly error-prone for younger samples, who less commonly die from medical causes. Second, some studies base their estimates on hospitalized samples, who are at much higher risk than outpatient samples. In one study 4% of inpatients with affective disorders eventually committed suicide, compared to 2.2% of mixed inpatient/outpatient samples (Bostwick & Pankratz, 2000). Many patients with bipolar disorder experience ongoing impairments in social, occupational, and familial functioning even between episodes, especially if they have unresolved depressive symptoms (Altshuler et al., 2006; Fagiolini et al., 2005; Gitlin, Mintz, Sokolski, Hammen, & Altshuler, 2011). In a Stanley Foundation Network study of 253 adult patients with bipolar I or II disorder, only about one in three worked full-time outside of the home (Suppes et al., 2001). More than half (57%) were unable to work or worked only in sheltered settings. Only 42% of bipolar I, manic patients show “steady” work performance an average of 1.7 years after hospital discharge (Harrow, Goldberg, Grossman, & Meltzer, 1990). In community samples, people with bipolar disorder are at 4 times greater likelihood of work disability compared to the general population (Mitchell, Slade, & Andrews, 2004). There is considerable variability in functional impairment, however. For example, there appears to be a link between BD and creativity or productivity: Many famous artists, musicians, writers, and politicians probably had the disorder (Jamison, 1993). Patients with bipolar disorder and highly creative persons without psychiatric disorder appear to have temperamental commonalities, such as high levels of drive, ambition, openness to new experiences and novelty seeking (Johnson, Edge, Holmes, & Carver, 2012; Nowakowska, Strong, Santosa, Wang, & Ketter, 2005). Children diagnosed with bipolar disorder and children who are the offspring of bipolar parents scored higher than healthy control children on a creativity index (Simeonova, Chang, Strong, & Ketter, 2005). Interestingly, the unaffected family members of patients with bipolar disorder demonstrate higher accomplishment and creativity than do their affected relatives. Virtually all patients with bipolar disorder have illness recurrences. The rates of recurrence, even when patients are treated with mood stabilizers, average 37% in 1 year, 60% over 2 years, and 73% over 5 years (Gitlin, Swendsen, Heller, & Hammen, 1995). Approximately one in five patients meets criteria for rapid cycling, defined by four or more distinct episodes of mania, hypomania, mixed, or depressive disorder within 1 year (Schneck et al., 2004). Biological and psychosocial factors that predict illness recurrences are discussed further in the next section. Even more significant are the persistent, mild-to-moderate residual symptoms that most patients experience between episodes, even when undergoing pharmacotherapy (Judd et al., 2002; Keller, Lavori, Coryell, Endicott, & Mueller, 1993; Post et al., 2003). Over a 13-year follow-up, subsyndromal symptoms were present during approximately half the weeks of follow-up (Judd et al., 2002). These symptoms were predominantly depressive rather than manic. A study of children with bipolar I, II, and NOS disorders observed similar patterns of residual symptoms over a 15-month follow-up (Birmaher et al., 2009). Indeed, one of the most formidable issues in the treatment of the disorder is the stabilization of depressive symptoms (e.g., Perlis et al., 2006). By the end of the 1980s, researchers began to acknowledge that biological and genetic models of BD did not explain the enormous heterogeneity in the course of the illness over time (Prien & Potter, 1990). This recognition contributed to a renewed emphasis on psychosocial predictors in the course of the disorder. For example, Ellicott, Hammen, Gitlin, Brown, and Jamison (1990) found that BD patients with high life-events-stress scores were at 4.5 times greater risk for relapse in a 2-year follow-up than patients with medium or low life-events-stress scores. Miklowitz, Goldstein, Nuechterlein, Snyder, and Mintz (1988) found that BD I manic patients who returned after a hospitalization to families rated high on expressed emotion (EE) attitudes (criticism, hostility, or emotional overinvolvement) or who showed high levels of caregiver-to-patient affective negativity (criticism, hostility, or guilt induction) during face-to-face interactions were at high risk for relapse. Those whose families had both high EE and high affective negativity were highly likely to relapse within 9 months (94%), whereas those whose families rated low on both attributes were unlikely to relapse within this time frame (17%). Although these first-generation studies established the prognostic role of psychosocial factors, they did not address how psychosocial variables influence depression versus mania. A second generation of research has examined which psychosocial variables influence the course of BD depression and which influence the course of mania. The symptomatology and neurobiology of unipolar and BD depression have many strong parallels (Cuellar, Johnson, & Winters, 2005). Given these parallels, one might expect that psychosocial predictors of unipolar depression would influence BD depression. Here, we focus on well-established predictors of unipolar depression, including negative life events (Monroe, Harkness, Simons, & Thase, 2001), low social support (Brown & Andrews, 1986), EE (Butzlaff & Hooley, 1998), neuroticism (Gunderson, Triebwasser, Phillips, & Sullivan, 1999), and negative cognitive styles (Alloy, Reilly-Harrington, Fresco, Whitehouse, & Zechmeister, 1999). Negative life events are perhaps the most comprehensively examined predictors of bipolar depression. Fortunately, a body of BD studies have now used interview-based measures of life events. As reviewed by Johnson (2005a), three of these cross-sectional studies found that negative life events are equally common before episodes of BD depression and unipolar depression (Malkoff-Schwartz et al., 2000; Pardoen et al., 1996; Perris, 1984). Findings of prospective studies with interview-based measures also indicate that stressful life events are correlated with slow recovery from depression (Johnson & Miller, 1997) and predict increases in BD depression over several months (Johnson et al., 2008). Thus, the most methodologically rigorous studies suggest that negative life events are precipitants of bipolar depression. Other variables involved in unipolar depression also appear to have validity as predictors of bipolar depression. For example, neuroticism (Heerlein, Richter, Gonzalez, & Santander, 1998; Lozano & Johnson, 2001), low social support (Johnson, Winett, Meyer, Greenhouse, & Miller, 1999), family EE (Kim & Miklowitz, 2004; Yan, Hammen, Cohen, Daley, & Henry, 2004), and rejection sensitivity (Ng & Johnson, 2013) have been found to predict increases in depressive symptoms but not manic symptoms over time. Although negative cognitive styles are often documented in BD (see Cuellar et al., 2005, for review), they (a) are most likely to be found during depression compared with well periods (Johnson & Kizer, 2002), (b) predict depression better than mania (Johnson & Fingerhut, 2004; Johnson, Meyer, Winett, & Small, 2000), and (c) can be explained by the presence of depressive history rather than manic history (Alloy et al., 1999). In sum, variables that influence the course of unipolar depression also influence BD depression, including negative life events, poor social support, family EE, negative cognitive styles, and low self-esteem. Few studies have examined the additive or interactive effects of these risk variables in the course of BD. Compared with BD depression, less is known about the psychosocial predictors of mania. Available models highlight two sets of predictors: goal engagement and sleep/schedule disruption. Drawing on biological models of overly sensitive reward pathways, the goal dysregulation model suggests that people with BD may show more extreme responses to rewarding stimuli (Johnson et al., 2012). A large number of studies indicate that people with a history of mania and students who are vulnerable to mania describe themselves as more sensitive to rewards (Johnson et al., 2012). People with BD also place a stably high emphasis on goal pursuit, even when they are not in an episode (Johnson, Eisner, & Carver, 2009). This reward sensitivity would be expected to influence reactions to life events that involve major successes. Consistent with this idea, life events involving goal attainments (such as new relationships, births of children, or career successes) predict increases in manic symptoms but not depressive symptoms (Johnson et al., 2000; Johnson et al., 2008). Such effects were apparent even after controlling for baseline levels of manic symptoms and excluding life events that could have been caused by the patients’ symptoms. Why might attaining an important goal trigger manic symptoms? A set of studies suggest that for people with BD, cognition becomes much more positive during good moods than it does for other people. Available evidence suggests that mood states are associated with distinct positive shifts in confidence (Johnson et al., 2012; Stern & Berrenberg, 1979), autobiographical recall (Eich, Macaulay, & Lam, 1997), and attention to valenced stimuli (Murphy et al., 1999). Impulsivity, or the tendency to pursue rewards without awareness of potential negative consequences, also becomes elevated as people become manic (Swann, Dougherty, Pazzaglia, Pham, & Moeller, 2004) or even more mildly happy (Muhtadie, Johnson, Carver, Gotlib, & Ketter, 2014). Mood-state-dependent shifts in confidence may contribute to increased goal setting (Johnson, 2005b). In turn, investment in goal pursuit predicts increases in manic symptoms over several months (Lozano & Johnson, 2001). Consistent with these findings, experience sampling data suggest that goal attainments trigger increased goal engagement for those with bipolar disorder (Fulford, Johnson, Llabre, & Carver, 2010), which, in turn, may trigger manic symptoms (Johnson, Carver, & Gotlib, 2012). Reward engages confidence and goal pursuit but may also bring to mind memories of failure or feelings of low self-worth (Eisner, Johnson, & Carver, 2008). In one study (Miklowitz, Alatiq, Geddes, Goodwin, & Williams, 2010), remitted or partially remitted patients with BD, patients with MDD, and healthy controls unscrambled six-word strings into five-word sentences, leaving out one word. The extra word allowed the sentences to be completed in a negative, neutral, or “hyperpositive” (manic/goal-oriented) way. Under conditions of reward (a pleasing bell tone for every four sentences completed), all subjects completed more sentences than in the nonreward conditions. However, patients with BD unscrambled more negative sentences under conditions of reward than did patients with MDD. Thus, a simple reward may increase the accessibility of negative thoughts among bipolar patients, consistent with earlier notions of mania as a defense against threat or low self-esteem (Adler, 1964; Lyon, Startup, & Bentall, 1999). Experimental studies (Barbini et al., 1998) as well as longitudinal studies (Leibenluft, Albert, Rosenthal, & Wehr, 1996) suggest that sleep deprivation is an important trigger of manic symptoms. Wehr, Sack, and Rosenthal (1987) hypothesized that sleep disruption might be one way in which life events trigger episodes of BD, noting that illness episodes are often preceded by life events interfering with the ability to sleep (e.g., transmeridian flights, childbearing). This theory was broadened by Ehlers and colleagues (Ehlers, Frank, & Kupfer, 1988; Ehlers, Kupfer, Frank, & Monk, 1993), who suggested that social disruptions to other aspects of circadian rhythms could trigger symptoms (the “social zeitgebers model”). Consistent with this idea, Jones, Hare, and Evershed (2005) have documented that people with bipolar disorder demonstrate more variability in their daily schedules than do healthy controls. In two studies, Malkoff-Schwartz et al. (1998; 2000) found that bipolar patients reported more life events involving social-rhythm disruption (events that affect sleep or wake times, patterns of social stimulation, or daily routines) in the weeks preceding manic episodes compared to the weeks preceding depressive episodes. Laboratory mice with mutations in “CLOCK” genes behave in ways that resemble people with mania (e.g., increases in activity, decreased sleep, reward-seeking behavior). These changes are reversed when mice are given lithium (Roybal et al., 2007). These animal models may inform the role of sleep disruptions and possibly broader schedule disruptions in the onset and course of mania. In sum, sleep or schedule disruption is proposed to trigger manic symptoms. As with research on the predictors of bipolar depression, very few longitudinal studies of sleep disruption are available. As discussed next, some of these risk factors are amenable to modification through psychosocial intervention. Optimally, treatments for BD should involve combinations of pharmacological and psychosocial interventions. Unfortunately, in the era of managed care cost containment, drug treatments are often the only treatment provided. Perhaps as a result, the average duration of lithium treatment for patients in community settings is only 76 days (Johnson & McFarland, 1996). Distinctions are usually made between acute pharmacological treatment and maintenance treatment. The goal of acute treatment is to stabilize an existing manic or depressive episode (for recent reviews of acute pharmacotherapy studies, see Malhi, Adams, & Berk [2009] and Geddes & Miklowitz [2013]). Adjunctive psychotherapy is usually introduced during the postepisode stabilization phase and continued throughout maintenance treatment. The goals of adjunctive psychotherapy are to minimize residual symptoms and prevent recurrences. Untreated residual symptoms of mania or depression are prospectively associated with illness recurrences (Perlis et al., 2006). Current pharmacotherapy algorithms for mania for adult and childhood-onset patients (e.g., Kowatch, Fristad, et al., 2005; McAllister-Williams, 2006) combine mood stabilizers (e.g., lithium carbonate, divalproex sodium, carbamazepine) with atypical antipsychotic medications (e.g., olanzapine, quetiapine, risperidone, aripiprazole, ziprasidone, and less frequently, clozapine). Adjunctive antidepressant agents are often recommended for bipolar depression. When given alone, antidepressant medications can cause manic switching and acceleration of cycles in a significant number of patients (Ghaemi, Lenox, & Baldessarini, 2001). When given with mood stabilizers, however, antidepressants do not appear to cause an increase in cycling (Altshuler et al., 2003; Sachs et al., 2007). The anticonvulsant lamotrigine is an option for many patients given its antidepressant properties and lower propensity to cause cycle acceleration (Malhi et al., 2009). Despite the clear effectiveness of many forms of pharmacotherapy, patients with BD are prone to discontinuing their medications, with as many as 60% being fully or partially noncompliant after a major episode of illness (Keck et al., 1998; Strakowski et al., 1998). Patients who discontinue their pharmacotherapy abruptly are at a high risk for recurrence and suicide attempts (Keck, McElroy, Strakowski, Bourne, & West, 1997; Suppes, Baldessarini, Faedda, Tondo, & Tohen, 1993; Tondo & Baldessarini, 2000). For example, in a naturalistic study of adolescent BD patients followed 18 months after a hospitalization, relapse was 3 times more likely among patients who discontinued lithium than among patients who remained on it. Patients describe barriers to compliance that include missing high periods, objecting to having one’s moods controlled by medications, lack of information about the disorder, or lack of social or familial supports (for review, see Colom et al., 2000). Randomized controlled trials of psychotherapy indicate positive benefits for psychoeducational, skill-oriented, and interpersonal treatments. The modalities investigated in the trials have included individual, family, and group formats. Table 6.1 lists some of the key components of empirically supported treatments for BD. Treatments supported by at least one randomized controlled trial include individual psychoeducation (Perry, Tarrier, Morriss, McCarthy, & Limb, 1999), group psychoeducation (e.g., Colom et al., 2003; Simon et al., 2005), family-focused therapy (FFT; e.g., Miklowitz, George, Richards, Simoneau, & Suddath, 2003; Rea et al., 2003), cognitive-behavioral therapy (CBT; e.g., Lam et al., 2003), and interpersonal and social rhythm therapy (IPSRT; e.g., Frank et al., 2005) (for review, see Miklowitz & Scott, 2009). Table 6.1 Key Therapeutic Elements of Psychoeducational Treatment These treatments have several elements in common. First, all include an active psychoeducational component, which often involves teaching patients (and in some cases, family members) to recognize and obtain early treatment for manic episodes before they develop fully. All emphasize medication adherence, avoiding alcohol and street drugs, and the use of skills to cope more effectively with stress triggers. They differ in the emphasis on specific strategies, including involvement of family members in psychoeducation (FFT), stabilizing sleep/wake rhythms (interpersonal and social rhythm therapy), and cognitive restructuring (CBT; Miklowitz, Goodwin, Bauer, & Geddes, 2008). The treatments listed in Table 6.1 have all been found in at least one study to delay relapses of BD when combined with pharmacotherapy. The control conditions vary and have included treatment as usual, individual supportive therapy, unstructured group support, and active clinical management. There are failed replication studies as well: In a large-scale multicenter study, Scott et al. (2006) found that CBT was no more effective than treatment as usual in delaying recurrences. Post-hoc analyses, however, revealed that relative to treatment as usual, CBT was associated with longer time to recurrence among patients who had had fewer than 12 lifetime episodes, and less time to recurrence among patients with 12 or more prior episodes. The large-scale Systematic Treatment Enhancement Program for BD (STEP-BD) attempted to examine the effects of psychosocial interventions in a practical clinical trial across 15 U.S. treatment centers (Miklowitz et al., 2007). In this trial, 293 bipolar I and II patients were randomly assigned to one of three intensive psychosocial interventions (30 sessions over 9 months of family-focused treatment, IPSRT, or CBT, or a control treatment called collaborative care [CC]). The CC involved three psychotherapy sessions over 6 weeks and focused on developing a relapse prevention plan. All patients were in an acute episode of bipolar depression at the time of randomization. All received “best practice” pharmacotherapy (i.e., mood stabilizers, atypical antipsychotic agents, or antidepressants, in various combinations) in combination with psychotherapy. Over 1 year, being in any of the intensive psychotherapies was associated with a faster recovery from depression than being in CC. On average, patients in intensive treatment recovered within 169 days, as compared to 279 days in the CC condition. Patients in intensive treatment were also 1.6 times more likely than patients in CC to be clinically well in any given study month. Rates of recovery over 1 year were as follows: FFT, 77%; IPSRT, 65%; CBT, 60%; and CC, 52%. The differences among the three intensive modalities were not significant. There are still many gaps in research on psychosocial treatment. For example, effect sizes for depression have been larger than those for mania prevention, and it remains unclear whether particular subgroups of patients respond more completely to individual, family, or group treatment. Few studies have identified the mechanisms of action of psychosocial interventions (for example, whether they enhance medication adherence, improve the patient’s ability to recognize prodromal symptoms of recurrence, or increase insight or self-awareness). The applicability of psychosocial interventions to early-onset BD, or children at risk for developing the illness, has begun to be investigated systematically. Several randomized trials find that family-based treatments are effective in reducing symptom severity in children with bipolar spectrum disorders (Fristad et al., 2009) and adolescents with bipolar I or II disorder (Miklowitz et al., 2008). FFT has also been found to protect against the worsening of depressive and manic symptoms in children at high genetic risk for developing bipolar disorder (Miklowitz et al., 2013). Although the diagnosis of BD appears straightforward by DSM-5, there is considerable controversy as to its diagnostic boundaries with other conditions. Possibly, conceiving of bipolar disorder as a series of dimensions may enhance the reliability of diagnoses. Here, we address controversies regarding the definition of the bipolar spectrum, diagnosis in pediatric populations, and dual diagnosis (comorbidity) considerations. Although it has traditionally been conceptualized as a disease involving shifts between the dramatic poles of mania and depression, there is increasing recognition that many patients have bipolar spectrum disorders, which, depending on the definition, may include subsyndromal manic episodes, manic or hypomanic episodes triggered by antidepressants, subsyndromal mixed episodes, or agitated depression (Akiskal, Benazzi, Perugi, & Rihmer, 2005; Akiskal et al., 2000). For example, Smith, Harrison, Muir, and Blackwood (2005) observed that many young adults with recurrent depression have symptoms that would be best considered within the bipolar spectrum. In a sample of young adults treated for recurrent depression at a university health service, only 16% met the DSM-IV criteria for bipolar disorder and 83% for recurrent major depressive illness. When broader definitions of the bipolar spectrum were used, between 47% and 77% received bipolar diagnoses. Caution is warranted in interpreting these results, as the interrater reliability of interviews designed to assess milder spectrum forms of bipolar disorder is relatively low (Kessler et al., 2006). What is the evidence that spectrum patients are really bipolar in the classic sense? Although they do not necessarily follow the same course of illness patterns as patients with BD I or II, patients with subsyndromal forms of BD are more likely to have family histories of BD and higher rates of hypomania induced by antidepressants than people without subsyndromal BD symptoms. They also have high rates of suicide, marital disruption, and mental health service utilization compared to those with no history of mental illness (Judd & Akiskal, 2003; Nusslock & Frank, 2011). Some researchers have examined the bipolar spectrum as a risk factor for the development of fully syndromal BD. Assessment instruments designed to identify persons at risk include the Temperament Evaluation of Memphis, Pisa, Paris, and San Diego—autoquestionnaire version (TEMPS-A; Akiskal et al., 2005). The TEMPS-A is a self- or parent-rated assessment of temperaments believed to precede the onset of full BD and to persist during the euthymic states. Although the TEMPS-A measures a range of mood-related symptoms and tendencies, a specific tendency toward high moods and mood variability, as captured on the cyclothymia subscale, predicted onset of BD in a 2-year follow-up of 80 children and adolescents with depression (Kochman et al., 2005). The 79-item General Behavior Inventory (GBI; Depue, Kleinman, Davis, Hutchinson, & Krauss, 1985) evaluates lifetime experiences of depressive and manic symptoms, as well as mood variability, on 1–4 scales of frequency. High scores on the GBI during adolescence are associated with psychosocial impairment in adulthood (Klein & Depue, 1984), and a childhood version rated by parents discriminated pediatric BD from attention deficit hyperactivity disorder (ADHD; Danielson, Youngstrom, Findling, & Calabrese, 2003). Whereas the Klein et al. studies focused on the prognostic value of cyclothymic temperament, Reichart et al. (2005) found that higher scores on the GBI depression subscale were the best predictors of BD onset at 5-year follow-up among adolescents with depression. Finally, the Hypomanic Personality Scale (Eckblad & Chapman, 1986) has been used to assess subsyndromal manic symptoms and related traits. Kwapil et al. (2000) found that high scores on this scale predicted the onset of bipolar spectrum disorders over a 13-year follow-up of a large sample of college students. Thus, self-report measures of spectrum symptoms and temperamental characteristics predict the onset of BDs in longitudinal studies. Perhaps the most controversial diagnostic dilemma is where to draw the boundaries of the bipolar diagnosis in children. Before about 1980, belief was widespread that neither mania nor depression could occur before puberty, although individual case reports of the phenomenon existed (e.g., Anthony & Scott, 1960; Strecker, 1921). This belief has changed significantly in the past two decades (for review, see Luby & Navsaria, 2010). Unfortunately, no separate diagnostic criteria exist for juvenile BD patients, and, as a result, there is little agreement on the operational definition of a manic or mixed episode, whether well-demarcated episodes of mania and depression must occur or whether the symptoms can be chronic and unremitting, the minimum duration of episodes, and what constitutes a symptom (Biederman et al., 2003; Leibenluft, Charney, Towbin, Bhangoo, & Pine, 2003; McClellan, 2005). The reliability of the diagnosis appears to decrease with age (Meyer & Carlson, 2010). Unfortunately, DSM-5 has done little to improve the situation. One diagnosis, disruptive mood dysregulation disorder (DMDD), has been added to the mood disorders section to give a “diagnostic home” to children and adolescents who have recurrent irritable outbursts as well as chronic negative affectivity. These youth have often been called bipolar in community settings. Unfortunately, the DMDD diagnosis has significant problems, such as the lack of distinction with oppositional defiant disorder, the lack of a distinctive set of associated symptoms, and few data to guide treatment (Axelson, Birmaher, Findling, et al., 2011). One research group, led by Geller and colleagues (2002), has recommended that BD not be diagnosed in children unless elevated mood and grandiosity are present. Another group (Leibenluft et al., 2003) has recommended that BD in children be divided into three phenotypes: (1) a narrow phenotype (meets DSM-IV criteria for mania or hypomania with elated mood and grandiosity); (2) an intermediate phenotype (meets DSM-IV criteria but with irritable mood only); and (3) a broad phenotype (e.g., mania that does not meet the duration criteria for bipolar I or II or is one symptom short). Does childhood-onset BD develop into adult BD, or are they separate conditions? The NIMH Course and Outcome of Bipolar Youth (COBY) study provides strong empirical support for one approach to phenotyping high-risk youth. This long-term prospective study showed that youth with bipolar disorder, NOS characterized by (a) one DSM-IV
Bipolar and Related Disorders
Description of the Disorder
Manic, Hypomanic, Depressive, and Mixed Episodes
Bipolar Subtypes
Epidemiology
Age at Onset
Gender and Racial-Ethnic Issues
Clinical Picture
Suicide
Functional Impairment
Course and Prognosis
Psychosocial Predictors of the Course of Bipolar Disorder
Psychosocial Predictors of Depression Within Bipolar Disorder
Psychosocial Predictors of Mania
Goal Dysregulation
Sleep and Schedule Disruption
Treatment of Bipolar Disorder: Current Trends
Pharmacological Treatments
Psychotherapy as an Adjunct to Medication Maintenance
Diagnostic Considerations
The Bipolar Spectrum
Diagnosis in Children and Adolescents
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
