This article is presented here in four sections to facilitate access by slower internet connections:
These practice parameters describe the assessment and treatment of early-onset bipolar disorder based on scientific evidence regarding diagnosis and effective treatment and on the current state of clinical practice. Given the paucity of research on bipolar disorder in children and adolescents, many of the treatment recommendations are drawn from the adult literature. Although the same diagnostic criteria are used as for adults, youth may differ with regard to the developmental presentation of symptoms and comorbid psychiatric disorders. Treatment involves the combination of pharmacotherapy and adjunctive psychosocial interventions. Antimanic agents (primarily lithium or valproic acid) are the mainstays of pharmacotherapy. The treatment focuses on (1) amelioration of acute symptoms; (2) the prevention of relapse; (3) the reduction of long-term morbidity; and (4) the promotion of long-term growth and development. These parameters were approved by Council of the American Academy of Child and Adolescent Psychiatry on June 5, 1996. J. Am. Acad. Child Adolesc. Psychiatry, 1997, 36(l):138-157.
Principal Authors: Jon McClellan. M.D., and John Werry, M.D.
These parameters were developed by the Work Group on Quality Issues: William Ayres, M.D., and John Dunne, MD., Chairs; Members: Elissa Benedek, M. D., Gail Bernstein, M.D., Richard L. Gross, M.D., Robert King, M. D., Henrietta Leonard, M. D., and William Licamele, M. D. AACAP Staff: Mary Graham, Leslie Seigle, Carolyn A. Heier, Michelle E. Wright, and Diane Wiegand, R. N.
A draft of these parameters was distributed to the entire AACAP membership for comments. The parameters were approved by the AACAP Council on June 15, 1996.
© 1997 by the American Academy of Child and Adolescent Psychiatry. Republished on the BPSO Public Pages with the permission of the Academy.
Bipolar disorder was once thought to occur only rarely in youth. However, approximately 20% of all bipolar patients have their first episode during adolescence, with a peak age of onset between 15 and 19 years of age. Developmental variations in presentation, symptomatic overlap with other disorders, and lack of clinician awareness have all led to underdiagnosis or misdiagnosis in children and adolescents. Therefore, clinicians need to be aware of some unique clinical characteristics associated with the early-onset form.
Childhood-onset mania generally went unrecognized over the first part of this century, despite Kraepelin's (1921) observations that mania occurred rarely in children and that the onset of first episodes increased significantly after puberty. Anthony and Scott (1960) reached similar conclusions by using criteria derived from the adult literature to review reported cases of manic-depressive psychosis in children. They established a basis for further research by separating childhood onset from that during adolescence and by applying a criterion-based diagnostic schema to youth (Carlson, 1990).
The clinical biases that mania did not occur in adolescents persisted until large-scale studies of bipolar adults indicated that approximately one fifth of cases presented prior to age 19 (Carlson et al., 1977; Winokur et al., 1969). Subsequent studies have confirmed these findings (Joyce, 1984; Loranger and Levine, 1978).
Another previously held clinical bias, that schizophrenia was more common in youth, was complicated by the fact that manic adolescents frequently present with psychotic symptoms (McGlashan et al., 1988). Carlson and Strober (1978) reported on six bipolar adolescents originally misdiagnosed as schizophrenic, a diagnostic tendency further noted in subsequent studies (Bashir et al., 1987; Joyce, 1984; McClellan et al., 1993; Werry et al., 1991). Although clinicians have become increasingly aware of the confusion between early-onset bipolar disorder and schizophrenia, bipolar disorder in youth continues to be underrecognized and misdiagnosed (Carlson et al., 1994).
Historically considered rare, childhood-onset bipolar disorder is now being reported more often, although its frequency remains an area of some controversy (Carlson, 1990). Further research is needed to establish the specificity of symptoms distinguishing childhood mania from behavior disorders. Furthermore, if severity and duration are not included in the diagnostic criteria, estimated lifetime prevalence rates are greatly increased (Carlson and Kashani, 1988). This is an important issue when reviewing the literature since DSM-III-R (APA, 1987) removed the 7-day duration requirement specified in DSM-III (APA, 1980). Thus, studies reporting on early-onset mania using DSM-III-R (APA, 1987) criteria may have overdiagnosed mania. With DSM-IV (APA, 1994b) the 7-day duration criterion has been reinstituted, and a severity requirement has been added.
The existing research examining early-onset bipolar disorder is limited. Methodological problems include small sample sizes, lack of comparison groups, retrospective designs, and tack of standardized measures. More research is clearly needed in all aspects of this disorder but especially in examining the efficacy of various modes of treatment, longitudinal course, and diagnostic issues. Given these limitations, some of the information presented in this review had to be drawn from the adult literature. When discussing aspects of the disorder in relation to age of onset, we refer to early onset as prior to 18 years of age and very early onset as prior to 13 years of age. The literature has referred to the latter group as prepubescent but often on the basis of age rather than actual physiological development.
Children and adolescents are diagnosed with bipolar disorder based on the same criteria used for adults as outlined in DSM-IV (American Psychiatric Association [APA], 1994b). The existing data are sufficient to suggest that bipolar disorder with onset before the age of 18 years is essentially the same disorder as that in adults, but further studies are needed to clarify the long-term course and outcome of the early-onset forms, especially for those with very early onset. The following definitions should be used:
DSM-IV (APA, 1994b) has outlined several subtypes of bipolar disorder, including:
Whether there is either a seasonal pattern or rapid cycling should be specified for both bipolar I and II. A seasonal pattern represents a course of illness where the major depressive episodes occur consistently at a particular time of year. Rapid cycling is diagnosed when the patient has at least four episodes of a mood disturbance (major depression, mania, mixed, or hypomania) over a 12-month period.
None of the above three disorders is diagnosed if the symptoms are either better accounted for, or superimposed on, schizoaffective disorder, schizophrenia, schizophreniform disorder, delusional disorder, or a psychotic disorder not otherwise specified.
The diagnostic assessment needs to incorporate both current and past history regarding symptomatic presentation, treatment response, and psychosocial stressors. Cross-cultural issues may influence the expression or interpretation of symptoms and/or treatment response; therefore, they must be assessed (APA, 1994a). It is helpful to organize the clinical information using a life chart to characterize the course of illness, patterns of episodes, severity, and treatment response (APA, 1994a). Using such a longitudinal perspective to conceptualize the disorder helps with diagnostic accuracy, since the presenting symptoms during the acute phases often can be confused with other disorders.
Similarly, it is important to recognize the various phases and patterns of episodes associated with bipolar disorder. Youth may first present with either manic or depressive episodes. Twenty to thirty percent of youth with major depressions go on to have manic episodes (Geller et al., 1994; Rao et al., 1995; Strober and Carlson, 1982). Risk factors predicting eventual mania include: (1) a depressive episode characterized by rapid onset, psychomotor retardation, and psychotic features; (2) a family history of affective disorders, especially bipolar disorder; and (3) a history of mania or hypomania after treatment with antidepressants (Strober and Carlson, 1982). The same risk factors are also noted in the adult literature (Goodwin and Jamison, 1990). Similarly, psychotic depression, a family history of bipolar disorder, and a history of attention-deficit hyperactive disorder (ADHD) with associated affective instability are factors associated with youth switching to mania on antidepressant therapy (Venkataraman et al., 1992). In children with major depressive disorder, Geller et al. (1994) found that 31.7% (n = 25) went on to develop either mania or hypomania; 80 percent of them were 12 years old or younger (mean age 11.1 years) at the time of onset. Neither exposure to antidepressants, nor atypical features, were predictive in this sample.
In studies of adults, characteristic stages of mania have been described: stage I includes euphoria, increased psychomotor activity, and mood lability. This evolves into stage II, with symptoms of irritability, racing thoughts, dysphoria, and disorganization. Finally, in stage III, the patient's cognitive status deteriorates, with significant confusion and florid psychosis (Carlson and Goodwin, 1973). Significant depressive symptoms may either precede, occur conjointly (mixed episodes), and/or follow those of mania within the same episode. Depressive episodes in bipolar disorder are typically characterized by psychomotor retardation and hypersomnia, significant suicide attempts, and often psychotic symptoms (Goodwin and Jamison, 1990). Severe cases may progress to catatonia.
Children with mania frequently present with symptoms that are considered atypical (Bowring and Kovacs, 1992). The changes noted in mood, level of psychomotor agitation, and mental excitement are often markedly labile and erratic, rather than persistent. Irritability, belligerence, and mixed manic-depressive features are more common than euphoria. The types of reckless behaviors seen are constrained by the child's developmental and social boundaries, and thereby limited to typical childhood behavior problems, such as school failure, fighting, dangerous play, and inappropriate sexualized activity. Thus, hallmark manic symptoms, such as grandiosity, psychomotor agitation, and reckless behavior, must be differentiated from those of other more common childhood disorders, as well as from the normal childhood phenomena of boasting, imaginary play, overactivity, and youthful indiscretions.
The incidence and validity of the diagnosis in children remains controversial (Carlson, 1990). Dating back to Kraepelin (1921), large surveys have found that onset prior to age 10 occurs in only 0.3% to 0.5% of bipolar patients (Goodwin and Jamison, 1990; Kraepelin, 1921; Loranger and Levine, 1978). Since the estimated lifetime prevalence in the general population is 0.8% (APA, 1994a), it has generally been accepted that childhood-onset bipolar disorder is rare. However, more recent case reports question these estimates, noting that many seriously emotionally disturbed children meet criteria for mania by virtue of their problems with irritability, emotional lability, increased energy and reckless/dangerous behaviors (Weller et al., 1986b; Wozniak et al., 1995). These symptoms, which often have a chronic and, at times, rapidly fluctuating course, may represent the child's baseline state, rather than a marked change in functioning (Wozniak et al., 1995). Such presentations differ from the episodic course classically ascribed to the illness and are more typical of childhood behavioral disorders (Bowring and Kovacs, 1992). Further research is needed to clarify whether these reflect "true cases" of mania, or are, rather, a nonspecific collection of behavioral and emotional difficulties. Of particular importance will be longitudinal studies following these youth into adolescence/adulthood to determine if their symptoms evolve into the more typical presentation of the disorder (Goodwin and Jamison, 1990).
Adolescents with mania frequently have complicated presentations, including: (1) psychotic symptoms, including mood-incongruent hallucinations, paranoia, and marked thought disorder; (2) markedly labile moods, with mixed manic and depressive features; and (3) severe deterioration in their behavior (Akiskal et al., 1985; Goodwin and Jamison, 1990). These varying presentations have led to underdiagnosis of bipolar disorder in teenagers (Carlson et al., 1994), including very high rates of misdiagnosis as schizophrenia (Carlson, 1990; McClellan et al., 1993; Werry et al., 1991). Although the early course of bipolar disorder in adolescents is often more chronic and refractory to treatment, the long-term prognosis is probably similar to that of adults (Carlson, 1990; McClellan et al., 1993; McGlashan, 1988; Werry et al., 1991).
A diagnosis of bipolar disorder should be considered for any youth with a marked deterioration in functioning associated with either mood or psychotic symptoms. Since diagnostic accuracy does improve if DSM criteria are reliably applied (Carlson et al., 1994), using the available structured and semistructured diagnostic interviews may be warranted. However, even with these instruments, there is still a risk of overdiagnosis in youth with conduct disorder and ADHD (Weller et al., 1995). The Mania Rating Scale (Fristad et al., 1992) and subscales on the Child Behavior Checklist (Achenbach and Edelbrock, 1983; Biederman et al., 1995) have been used to distinguish manic children from those with ADHD.
Youth presenting with symptoms suggestive of bipolar disorder need to have a thorough psychiatric evaluation, including any necessary pediatric and neurological assessments. In this section, we highlight the most pertinent diagnostic conditions that need to be considered.
Schizophrenia. Early-onset bipolar disorder is frequently misdiagnosed (rates of 50% or more) as schizophrenia, especially in patients with onset during adolescence (Carlson, 1990; McClellan et al., 1993; Werry et al., 1991). Adolescents with mania more often have "schizophrenic-like" symptoms (i.e., hallucinations and delusions), and are more likely to be diagnosed as having schizophrenia or schizoaffective disorder than patients with adult-onset bipolar disorder (Bashir et al., 1987; McGlashan, 1988). Using standardized diagnostic techniques improves accuracy, although differentiation can still be problematic (Carlson et al., 1994).
Schizoaffective Disorder. The diagnosis of schizoaffective disorder requires a period of illness in which the patient has both a significant mood disorder (either major depression, mania, or a mixed episode) and psychotic symptoms fulfilling the requirements for schizophrenia (APA, 1994b). During the same period of illness, there also must be at least a 2-week period where hallucinations and delusions persist in the absence of predominant mood symptoms. Finally, the mood symptoms must be present for a substantial portion of the overall illness.
Schizoaffective disorder has not been well defined in youth. Eggers (1989) found that 28% of patients with early-onset schizophrenia had schizoaffective psychoses at follow-up. This is an ICD-9 diagnosis that overlaps with DSM-III-R (APA, 1987) diagnoses of bipolar disorder and schizoaffective disorder. Other follow-up studies of psychotic youth have found that the diagnosis of schizoaffective disorder was made infrequently, was associated with the most severe impairment, and was somewhat unreliable (McClellan et al., 1993; Werry et al., 1991).
Agitated Depression. An agitated depression may be confused with a mixed episode due to symptoms of increased psychomotor activity and anxiety. Ratings of manic symptoms have been used to differentiate the two conditions in adults (Swann et al., 1993). Conversely, patients with agitated depressions may eventually develop bipolar disorder.
Posttraumatic Stress Disorder. Youth with significant histories of trauma, including childhood maltreatment, often present with mood instability, hypervigilance, irritability, dissociative symptoms, and sleep disturbances. These symptoms may be confused with mania/mixed episodes. Some youth will have both disorders (Borchardt and Bernstein, 1995).
Borderline Personality Disorder. The affective instability, poor impulse control, and erratic behaviors associated with borderline personality disorder may be misdiagnosed as bipolar disorder. To differentiate, the traits of a personality disorder should be pervasive and persistent, whereas the symptoms of bipolar disorder represent a marked change in the patient's baseline mental status and global functioning. However, patients with bipolar disorder may have chronic symptoms of irritability, cyclothymia, and/or dysthymia, and are, therefore, at risk of being misdiagnosed as borderline (Akiskal, 1981). Some youth may have both (Kutcher et al., 1990). An additional problem is that a diagnosis of borderline personality disorder in adolescents may lack specificity (Garnet et al., 1994).
Childhood Disruptive Behavioral Disorders. The impulsivity, hyperactivity and irritability of ADHD or the antisocial and provocative behaviors of conduct disorder may all be confused with mania. Many of the symptoms overlap, including aggression, school failure, psychomotor agitation, restless sleep, distractibility, and sexually inappropriate behaviors. Although historically, mania was undoubtedly underdiagnosed in youth felt to be behaviorally disordered, the increased recognition of the phenomenon may now be leading to overdiagnosis, especially in preadolescents. Adding to this confusion is the fact that many youth with bipolar disorder have comorbid ADHD and/or conduct disorder (Carlson, 1990). However, the reverse does [not (bpso ed.)] seem to be true, since longitudinal follow-up studies of ADHD have not shown an increased incidence of bipolar disorder (Gittelman et al., 1985). Finally, Wozniak et al. (1994) found that first-degree relatives of children who meet DSM-III-R (APA, 1987) criteria for mania (94% also had ADHD) had high rates of both ADHD and bipolar disorder, and relatives of children with only ADHD had high rates of ADHD but not bipolar disorder.
Differentiating between these conditions can usually be done via history and mental status examination (Bowring and Kovacs, 1992). Both ADHD and conduct disorder are chronic persistent disorders of impulse control and behavioral regulation, and they represent stable patterns of functioning. ADHD begins before age 7 and may evolve into conduct disorder during late childhood or early adolescence. Bipolar disorder is usually episodic, with onset usually after age 12. It is a disorder of affect regulation characterized by abnormal mood and mental excitement usually presenting as a marked change in a youth's baseline functioning.
Cross-Cultural Issues and Culture-Bound Syndromes. Ethnic minorities and individuals from lower socioeconomic settings have had a greater risk of misdiagnosis of schizophrenia (Goodwin and Jamison, 1990). Ki1gus et al. (1995) found that psychiatrically hospitalized African-American adolescents were more often diagnosed with organic/psychotic disorders and less often diagnosed with affective/anxiety disorders than Caucasian teenagers.
Culture-bound syndromes represent recurrent patterns of maladaptive behaviors and/or troubling experiences specifically associated with different cultures or localities (APA, 1994b). Clinicians should refer to DSM-IV for a glossary of the best-studied culture-bound syndromes (APA, 1994b). These syndromes may be confused with bipolar disorder, and in many cases, the clinician will need to obtain consultation from a professional familiar with the particular cultural issues involved.
Mood Disorder Due to a Medical Condition. Symptoms of mania can be produced by a variety of different medical conditions (Cummings, 1985), including: (1) neurological disorders, such as brain tumors and CNS infections, including human immunodeficiency virus (HIV), multiple sclerosis, temporal lobe seizures, and Klein-Levin syndrome; (2) systemic conditions, such as hyperthyroidism, uremia, Wilson's disease, and porphyra; (3) prescribed medications, including antidepressant agents, sympathomimetics, bromocriptine, stimulants, and steroids; and (4) substances of abuse, including amphetamines, cocaine, phencyclidine, inhalants, and methylenedioxymethamphetamine (ecstasy). Youth presenting with manic symptoms need to have a thorough physical evaluation. Decisions regarding more extensive laboratory and neuroimaging studies should be made based on the clinical findings of the psychiatric, pediatric, and neurological examinations.
Associated Clinical Features
Although the number of studies examining early-onset bipolar disorder is limited, the data are sufficient to outline some of the associated features of the disorder:
Prevalence and Age of Onset. Epidemiological surveys of childhood psychiatric disorders have generally not addressed bipolar disorder (Costello, 1989a). A community school survey of older adolescents (14 to 18 years of age) found the lifetime prevalence rate to be approximately 1% (Lewinsohn et al., 1995). Most of the patients identified had bipolar II disorder. An additional 5.7% had subthreshold symptomatology. Carlson and Kashani (1988), in an epidemiological survey of 14- to 16-year-old youth, found that the estimated lifetime prevalence of mania varied from 0.6% to 13.3%, depending on whether duration and severity criteria were applied; for comparison, the lifetime prevalence in adults is estimated to be 0.8% (APA, 1994a). The incidence appears to increase after onset of puberty. Despite anecdotal reports of onset prior to 6 years of age, further research is needed to establish whether such cases represent true bipolar disorder.
Subtypes of Bipolar Disorder. Manic symptoms in youth frequently do not persist long enough to meet the 1-week duration criteria required by DSM-IV (APA, 1994b) for a manic episode (Akiskal, 1995; Akiskal et al., 1985; Carlson and Kashani, 1988; Lewinsohn et al., 1995). This is especially true for children. Therefore, youth are more likely to have a diagnosis of either bipolar II or cyclothymic disorder, rather than bipolar I disorder. Children and adolescents may also be more likely than adults to present with rapid-cycling episodes. Geller et al. (1995) found that in 26 patients with early-onset bipolar disorder (ages 7 to 18 years), 81% had a rapid-cycling course.
Gender Ratio. 0verall, bipolar disorder affects both sexes equally. However, in studies of early-onset cases, males seem to be more often affected, especially in those with onset before the age of 13 years. Females are more likely overall to have depressive disorders, although for children younger than 12 years of age, boys again appear to be at greater risk (Costello, 1989b).
Rapidity of Onset. Studies of the rapidity of onset have found that the majority of youth have either an acute (less than 2 weeks) or subacute (less than 3 months) prodromal course (McClellan et al., 1993; Werry et al., 1991).
Premorbid Functioning. Many youth with bipolar disorder have normal premorbid histories. However, preexisting behavioral problems, including ADHD and/or conduct disorder, are also found in a significant number (Carlson, 1990; McClellan et al., 1993; Werry et al., 1991). Premorbid anxiety and emotional problems are also common, including among those whose first affective episode is a depressive disorder. Dysthymic, cyclothymic, or hyperthymic (irritable, driven) temperaments may presage eventual bipolar disorder (Akiskal, 1995).
Intellectual Functioning. Studies have generally reported that over 90% of youth with bipolar disorder have normal IQs (Geller et al., 1994; McClellan et al., 1993; Werry et al., 1991). However, bipolar disorder, including rapid cycling, has been reported in patients with moderate to severe mental retardation, autism, and trisomy 21 (Carlson, 1990).
Family History. Bipolar disorder definitely has been shown to be a familial disorder (Rice et al., 1987; Strober, 1992b). The degree of familiarity is increased in early-onset cases, with reported lifetime rates of bipolar disorder of approximately 15% in first-degree relatives (Strober, 1992b; Todd et al., 1993, 1994).
Comorbid Conditions. A significant number of youth displays comorbid ADHD and/or conduct disorder (Borchardt and Bernstein, 1995; Carlson, 1990; Kovacs and Pollock, 1995; West et at., 1995). High rates of substance abuse are also noted in some samples (Borchardt and Bernstein, 1995; Carlson, 1990; McClellan et al., 1993). The presence of comorbid behavioral disorders and/or substance abuse negatively influences prognosis and treatment response (Carlson, 1990; Kovacs and Pollock, 1995; Strober, 1992a).
Course and Prognosis. In adults, bipolar disorder is generally an episodic disorder with a variable course (APA, 1994a). The majority of patients will have multiple episodes, usually 10 or more in untreated patients (Goodwin and Jamison, 1990). Episodes tend to come more frequently over time, until the cycle length stabilizes after the fourth or fifth episode (Goodwin and Jamison, 1990).
In a 5-year naturalistic prospective follow-up study of 54 adolescents with bipolar disorder, two patients never achieved complete remission (Strober et al., 1995). Of the remaining patients, 44% had a relapsing course (either major depression or mania), and 21% had two or more further episodes (Strober et al., 1995). Recovery from the index episode took longer for patients with depression (median time to recovery, 26 weeks) than for either mania or mixed episodes (median time, 9 and 11 weeks, respectively) (Strober et al., 1995).
Compared with adults, adolescents with bipolar disorder may have a more prolonged early course and less responsiveness to treatment (McGlashan, 1988; Strober et al., 1995). This may be due to the fact that adolescents with bipolar disorder frequently present with either mixed features, psychotic symptoms, and/or comorbid behavior/substance abuse problems, all of which predict a more refractory response to lithium therapy. However, the few available studies suggest that the long-term prognosis of early-onset bipolar disorder is similar to that of adult onset; with approximately one half of patients showing significant functional impairment compared with their premorbid state (McClellan et al., 1993; McGlashan, 1988; Werry et al., 1991). Premorbid characteristics, including intellectual functioning, also strongly influence outcome (Werry et al., 1992). Further research is needed to examine how bipolar disorder affects evolving developmental processes, given the disruptive impact of the episodes on academic, social, and family functioning.
Adolescents with bipolar disorder are at increased risk for completed suicides (Brent et al., 1988, 1993, 1994; Welner et al., 1979). Strober et al. (1995) found that 20% of their adolescent patients made at least one medically significant suicide attempt. In the adult literature, a large review of studies examining depressive and manic depressive disorders found the mean rate of completed suicides to be 19% (Goodwin and Jamison, 1990). Patients who are male or who are in the depressed phase of their illness are at the highest risk.
|Next Section: Treatment||Beginning of Article|
For suggestions and corrections, please contact WebMaster