Practice Parameters for the Assessment and Treatment of Children and Adolescents With Bipolar Disorder


A multimodal treatment plan, combining medications with psychotherapeutic interventions, is needed to address the symptomatology and confounding psychosocial factors in children and adolescents with bipolar disorder. Cultural issues must be appropriately incorporated into the treatment plan. The goal of therapy is to ameliorate symptoms and prevent relapse, while also reducing long-term morbidity and promoting normal growth and development.

This guideline will focus primarily on the treatment of mania and mixed episodes. The treatment of bipolar depression will be more thoroughly addressed in a practice parameter that is being developed for the assessment and treatment of depression in children and adolescents.


Psychopharmacological intervention is necessary to effectively treat early-onset bipolar disorder. The literature regarding medication treatment for children and adolescents with bipolar disorder is limited. Many of the current recommendations are, therefore, based on studies of adults. Pharmacotherapy is usually instituted to address manic (or mixed) symptoms and depressive symptoms or to prevent relapse of the disorder.

Regardless of medication choice, the procedures that need to be followed when initiating pharmacotherapy include:

Informed Consent. Informed consent (addressing the rationale for treatment, as well as the potential risks and benefits of the therapy) should be obtained from both the youth and the youth's parents/guardians. If the mental status of the patient precludes this or if therapy is refused, invoking the relevant statutory mechanisms for involuntary treatment may become necessary. For patients under the legal age of consent, basic information regarding treatment should be provided in a developmentally appropriate manner.

Examinations. Prior to medication therapy, a thorough psychiatric evaluation is needed. The symptoms for which treatment is targeted should be documented. A thorough physical examination is also necessary, including any clinically indicated laboratory and/or neuroimaging studies to evaluate potential organic etiologies and to provide baseline assessments for monitoring specific medications.

Phase of the Illness. The clinician must assess which phase of the patient's illness (i.e., manic, depressed, mixed, or in remission) to effectively intervene. As patients are effectively treated for mania, they will frequently progress through the stages of the disorder. Thus, they may appear to be worsening when, in fact, they are undergoing a natural progression of symptoms. Some patients with mania will have a depressive phase before their mood returns to its baseline state. Similarly, mania often progresses from a euphoric/hyperexcited condition to a more disorganized state before showing clinical improvement. Recognition of these changes is important when considering the addition of other medications or a change in the primary antimanic agent.

Length of Treatment. Short-term, an antimanic agent needs to be given at an adequate dose for at least 4 to 6 weeks before its efficacy can be determined. Clinicians should avoid multiple medication changes/additions, since doing so frequently confuses the clinical picture and generally does not improve efficacy.

Long-term, the evidence to date shows that the relapse rate is quite high for early-onset bipolar disorder. Strober et al. (1990) prospectively followed 37 adolescents with bipolar disorder over 18 months and found that more than 90% of those who were noncompliant with their lithium treatment relapsed (12 of 13 cases). The relapse rate for those compliant with treatment was 37.5%. This study suggests that prophylactic therapy is needed for at least 18 months. However, further research is needed.

In the adult literature, over 80% of patients with a manic episode will have at least one episode of relapse (APA, 1994a). Withdrawal of maintenance lithium therapy has been associated with an increased risk of relapse, especially within the 6-month period following lithium discontinuation (APA, 1994a). In addition, there are reports of patients whose symptoms were well managed long-term on lithium, but after the lithium was discontinued, the patients relapsed and no longer responded adequately to lithium therapy (Post et al., 1992). The indications for, and the overall duration of, long-term maintenance therapy need further study. Clinically, however, some individuals may need lifelong prophylaxis.

Since no definitive studies address the long-term treatment of early-onset bipolar disorder, the clinician must balance the potential deleterious impact of relapse versus that of the side effects of the medications. Any attempts to taper prophylactic therapy must be closely monitored for symptoms of relapse. Further, patients and families must be thoroughly educated as to the signs and symptoms of relapse to provide for resumption of treatment as soon as possible, if necessary. Finally, the patient's diagnostic status must be reassessed over time, given the high rate of misdiagnosis in youth, to ensure that the course of medication therapy is appropriate.

Medication Choice. The choice of medication(s) should be made based on: (1) evidence of efficacy, (2) the phase of illness (see above), (3) the presence of confounding presentations (e.g., rapid-cycling mood swings, psychotic symptoms), (4) the agent's side effect spectrum, (5) the patient's history of medication response, and (6) the preferences of family/patients. In addition, different ethnic groups may vary in their pharmacokinetic response to lithium, antidepressants, and the other psychotropic agents, with a potential impact on side effects, blood levels, and efficacy (Lin et al., 1995). Multiple agents are often required, but care should be taken to avoid unnecessary polypharmacy. A discussion of the individual agents follows.


Lithium, traditionally the agent of first choice in the treatment of bipolar disorder, has the largest database supporting its efficacy (APA, 1994a). In adults, lithium has been shown to be effective for: (1) the treatment of acute manic and depressive episodes, (2) prevention of recurrent manic and depressive episodes, and (3) reduction of mood instability between episodes (Goodwin and Jamison, 1990). Approximately 80% of patients with bipolar disorder respond to lithium, both for acute mania and depression (APA, 1994a). However, the response rate for mania is quicker (2 weeks versus 6 to 8 weeks for depression). Lithium also helps recurrent episodes of both mania and major depression. Discontinuation of long-term lithium therapy increases the risk of relapse, at least in the short term (APA, 1994a). Some patients may develop a more treatment-resistant form of the illness after previously effective lithium prophylaxis has been discontinued (Post et al., 1992).

A paucity of data examines the efficacy of lithium for early-onset bipolar disorder (Alessi et al., 1994; Kafantaris, 1995; Strober, 1992a; Viesselman et al., 1993). The few double-blind placebo-controlled studies are limited by either small sample size and/or mixed diagnostic status, including more broadly defined psychotic disorders and/or the combination of manic symptoms and disruptive behavior disorders (Carlson et al., 1992; Delong and Aldershof, 1987; Gram and Rafaelsen, 1972; McKnew et al., 1981). These studies, plus case series and individual case reports, have found lithium to be effective (Alessi et al., 1994; Carlson and Strober, 1978; Hassanyeh and Davison, 1980; Horowitz, 1977; Hsu and Starzynski, 1986; Strober et al., 1988), with greater evidence published for those with adolescent onset (Alessi et al., 1994). However, the overall response may be less than that for adults, possibly because youth, especially adolescents, with mania often have either mixed manic-depressive syndromes and/or a predominance of psychotic symptoms, both of which are generally more refractory to treatment.

1. Pharmacology. Lithium, an alkali metal similar to sodium, is prescribed as one of two salt preparations, lithium carbonate or citrate (Viesselman et al., 1993). It has multiple complex neurochemical effects, with impact on ion channels, serotonin, dopamine, and norepinephrine neurotransmitter systems, as well as on second messenger systems (e.g., adenylate cyclase) (Alessi et al., 1994). How these neurochemical effects influence mania is not clearly understood (Viesselman et al., 1993).

Lithium is excreted almost entirely by the kidneys without undergoing hepatic metabolism (Viesselman et al., 1993). Serum levels increase linearly with dose, and steady-state levels are usually reached after approximately 1 week of administration. Since children generally have a higher glomerular filtration rate than adults, the required dose to weight ratio is usually higher for this age group (Viesselman et al., 1993).

2. Blood Levels. As in adults, the recommended therapeutic serum levels for the acute phase range from 0.6 to 1.2 mEq/ L (Viesselman et al., 1993). However, if manic symptoms persist, the serum levels may be carefully increased beyond this range, as long as the side effects are tolerated. In adults, maintenance serum levels generally have been in the range of 0.6 to 0.8 mEq/L, although this has not been adequately studied (APA, 1994a). Patients maintained on lower prophylactic levels (0.4 to 0.6 mEq/L) have fewer side effects but higher rates of relapse than those maintained on higher dosages (0.8 to 1.0 mEq/L) (Gelenberg et al., 1989). For some patients, the necessary acute phase and maintenance phase serum levels will be the same (APA, 1994a).

To achieve therapeutic levels, clinicians may choose to initiate a reasonable starting dose and monitor blood levels over time. Lithium levels are often needed at least weekly when starting treatment. Serum levels are assessed 12 hours after the last dose. Generally, starting dosages range from 300 mg to 900 mg per day, depending on the size of the child. Weller et al. (1986a) found that, for children, a starting dose of 900 mg/m2 (approximately 30 mg/kg daily) generally produced a therapeutic serum level within 5 days, although some children may develop levels greater than 1.4 mEq/L (Alessi et al., 1994). Also, published nomograms used to calculate dosages based on blood levels after single test doses (Alessi et al., 1994; Geller and Fetner, 1989) may be helpful in avoiding excessive dosages in children who are slow metabolizers and may permit more rapid dosage adjustments.

Several studies have examined the possibility of using saliva levels in children to avoid blood draws. However, methodological problems have limited the utility of this technique. Saliva levels are more variable than serum, and to establish an accurate serum-saliva ratio for a given patient, several blood draws are needed (Alessi et al., 1994). Therefore, serum levels remain the standard of care.

3. Side Effects. Youth generally tolerate lithium well and may have fewer side effects than adults (Alessi et al., 1994). Younger children may be more prone to side effects than older children (Campbell et al., 1991). Commonly reported adverse reactions include nausea, diarrhea, vomiting, tremor, weight gain, headache, polyuria, polydipsia, enuresis, fatigue, and ataxia (Alessi et al., 1994; Silva et al., 1992; Viesselman et al., 1993). Although the spectrum of lithium side effects in youth is similar to that in adults, there are few studies regarding its long-term effects in youth. Therefore, the potential side effects outlined below are primarily derived from the literature on adults:

Endocrine. Lithium has been associated with the development of hypothyroidism, goiters, and thyroid autoantibodies, including reported cases in children (Alessi et al., 1994). Thyroid stimulating hormone is elevated in lithium-induced hypothyroidism and, therefore, conducting the laboratory test for its level is useful in monitoring lithium therapy.

Renal. Lithium inhibits the action of antidiuretic hormone on the distal tubules and collecting ducts, leading to the common side effects of polyuria and polydipsia. This may evolve into diabetes insipidus, which is usually reversible if the lithium is discontinued (Gelenberg and Schoonover, 1991).

In adults, structural changes in the kidney, including focal glomerular atrophy, interstitial fibrosis, and impaired tubular functioning, have been reported with long-term lithium therapy (Gelenberg and Schoonover, 1991). However, the risk of significant glomerular damage with long-term lithium therapy appears to be minimal (Gelenberg and Schoonover, 1991; Hetmar, 1991; Meyers, 1989). Other reported renal problems include nephrotic syndrome, proteinuria, and renal failure with acute intoxication (Alessi et al., 1994; Gelenberg and Schoonover, 1991).

Cardiovascular. Lithium can have significant effects on cardiac conduction, including first-degree atrioventricular block, irregular sinus rhythms, and increased premature ventricular contractions. However, serious adverse reactions are rare (Gelenberg and Schoonover, 1991). In children, reversible conduction abnormalities have been reported (Campbell et al., 1972).

Hematological. Benign mild leukocytosis is commonly noted (Viesselman et al., 1993).

Dermatological. Lithium can induce or exacerbate dermatological problems, including acne, which may be a significant concern to adolescent patients (Viesselman et al., 1993).

Neurological. Lithium may produce a variety of neurological effects, including muscle weakness, tremor (which can be treated with propranolol), lethargy, cognitive blunting, and headaches (Gelenberg and Schoonover, 1991; Viesselman et al., 1993). These are often time-limited and remit quickly. Serious neurotoxicity may develop with higher blood levels, including ataxia, dysarthria, nystagmus, and confusion. With blood levels above 3.0 mEq/L, patients may develop more devastating neurological impairments, including seizures, coma, and death (Gelenberg and Schoonover, 1991).

In children, lithium has been reported to alter EEG patterns and to decrease performance on cognitive testing (Portteus Mazes test) (Alessi et al., 1994). However, at therapeutic blood levels, the impact on overall cognitive functioning does not appear to be significant (Viesselman et al., 1993).

Osseous. Given lithium's interaction with calcium metabolism, it theoretically might cause long-term problems with skeletal growth. However, it does not seem to increase the risk for osteoporosis in adults (Viesselman et al., 1993). How lithium potentially interacts with the developmental maturation of a child needs to be further researched.

Teratogenic. In utero exposure to lithium has been associated with various congenital abnormalities, especially cardiac anomalies (Ebstein's anomaly) (Jacobson et al., 1992). However, when examined prospectively in a multicenter study, women who used lithium during the first trimester of pregnancy had no greater risk of mania or congenital malformations in their offspring than controls (one out of 148 cases with in utero exposure to lithium developed Ebstein's anomaly) (Jacobson et al., 1992). It is generally recommended that lithium be avoided during pregnancy, especially during the first trimester (Green, 1995). If lithium is used, adequate monitoring for fetal congenital anomalies, including ultrasound and fetal echocardiography, is needed (Jacobson et al., 1992). Adolescent females should be evaluated for unsuspected pregnancies prior to making medication decisions.

Drug-Drug Interactions. Lithium is commonly used with antipsychotic agents without significant problems. However, there are reports of patients developing an encephalopathic syndrome or neuroleptic malignant syndrome with concomitant use of lithium and neuroleptics (Green, 1995). This may be dose-related.

Many agents influence lithium levels. For example, carbamazepine, nonsteroidal antiinflammatory agents, some antibiotics (tetracycline), and thiazide diuretics all may raise lithium levels (Green, 1995).

4. General Medical Evaluation. Prior to initiating lithium therapy, a thorough psychiatric and pediatric evaluation is warranted, including inquiring about any history of renal, endocrine, and/or cardiovascular problems. Whether or not the patient is on a low-salt diet or diuretic therapy should also be noted, since these will influence lithium levels.

Baseline laboratory assessment should include: blood counts, thyroid functions, urinalysis, blood urea nitrogen, creatinine, and a pregnancy test in adolescent females (Rosenberg et al., 1994). When clinically indicated, an electrocardiogram and a 24-hour creatinine clearance should also be obtained. Once a stable lithium dose is obtained, lithium levels, renal and thyroid functions, and urinalyses need to be regularly monitored (every 3 to 6 months) (Rosenberg et al., 1994).

Anticonvulsant Mood Stabilizers

In the adult literature, both open and controlled studies have supported the efficacy of the anticonvulsants carbamazepine and valproate for the acute treatment of bipolar disorder (APA, 1994a; McElroy et al., 1992). Greater evidence supports the use of valproate, which was recently approved as a treatment for mania by the U.S. Food and Drug Administration. Some patients who do not respond to, or are intolerant of, lithium, may respond to the anticonvulsants. These agents have been used by themselves, together, or in combination with lithium. Indications for their use include rapid cycling and dysphoric (or mixed) mania (both of which are associated with poor response to lithium) (APA, 1994a; McElroy et al., 1992). Some evidence supports their use as prophylactic agents during the maintenance phase of therapy, but controlled studies are lacking (APA, 1994a). The anticonvulsants are probably less effective for acute depressive symptoms than lithium, although valproate has been reported beneficial for the depressive phases of bipolar II disorder (Puzynski and Klosiewicz, 1984).

Only a few case reports examine the efficacy of anticonvulsant agents for early-onset bipolar disorder, but none is placebo-controlled (Pataki and Carlson, 1992; Viesselman et al., 1993). Therefore, the justification for their use is derived primarily from the adult literature. Valproate has been reported to be effective for treating mania in adolescents (West et al., 1994). However, the patients in these reports were also receiving adjunctive psychotropic agents (primarily neuroleptics). Carbamazepine has been reported to be effective in manic adolescents nonresponsive to lithium (Hsu, 1986). Since these agents have been widely used in children and adolescents with seizure disorders, there is considerable information regarding side effects in youth (Carpenter and Vining, 1993).

Valproate. Valproate, a commonly used anticonvulsant in youth, is effective for a wide variety of seizure disorders (Carpenter and Vining, 1993). Its antiepileptic effects are believed to be due to its increasing brain levels of (-aminobutyric acid (GABA) (Viesselman et al., 1993). It is hepatically metabolized and highly protein-bound (APA, 1994a). Valproate tends to inhibit the metabolism of other agents and, therefore, may cause an increase in the plasma levels of concurrent medications (APA, 1994a; Viesselman et al., 1993).

The relationship between valproate serum levels and its effectiveness for mania has not been adequately defined (APA, 1994a). Thus, the recommended therapeutic levels are the same as for treating seizure disorders -– 50 to 100 (g/mL (Viesselman et al., 1993). The usual dose for children is 10 to 60 mg/kg daily, and for adolescents, 1,000 to 3,000 mg daily (Pedley et al., 1995).

Most youth tolerate valproate well. Common side effects include sedation, nausea, and vomiting (Viesselman et al., 1993). Although rare, potentially serious side effects are reported. Valproate has been associated with fatal hepatic toxicity. Those at highest risk are very young children (less than 2 years old) during the first 6 months of therapy, especially those on multiple anticonvulsants: and/or with severe seizure disorders/organic impairment (PDR, 1994). Hematological side effects have been noted, including reported cases of coagulopathies, thrombocytopenia, neutropenia, pancytopenia, and/or aplastic anemia (PDR, 1994; Viesselman et al., 1993). Increased rates of polycystic ovary disease in adult women with seizure disorders have also been reported (Isojarvi,et al., 1993). Finally, valproate has been associated with neural tube defects when exposure occurred during the first trimester (PDR, 1994).

Prior to treatment, baseline liver functions, blood cell counts, and coagulation tests are recommended (Visselman et al., 1993). Once a stable therapeutic valproate level is obtained, serum valproate levels, plus hepatic and hematological measures, should then be periodically monitored every 3 to 6 months (Papatheodorou and Kutcher, 1993). However, periodic monitoring does not ensure that abnormalities will be identified; it is also important to advise patients and families about presenting symptoms of these side effects.

Carbamazapine. Carbamazapine is chemically related to tricyclic antidepressants. Its dosing kinetics are linear. However, it causes induction of hepatic metabolic enzymes, which accelerates its own metabolism, as well as that of other hepatically metabolized agents. This autoinduction often results in decreased blood levels over time without any change in dosage. Similar to valproate, the therapeutic range for serum levels has not been established for treating mania. The recommended range is the same as that used for seizure disorders–4 to 12 (g/mL (Rosenberg et al., 1994). The dosage for children generally is 10 to 20 mg/kg daily (200 to 600 mg per day), and for adolescents, dosages may go as high as 1,200 mg per day or more (Pedley et al., 1995; Viesselman et al., 1993). When initiating a trial of Carbamazapine, starting dosages begin at 100 to 300 mg per day and are increased as tolerated while monitoring blood levels. Daily dosages are usually divided in b.i.d. or t.i.d. schedules.

Upon initiation of therapy, transient side effects may include drowsiness, dizziness, nausea, and mild ataxia (Viesselman et al., 1993). Leukopenia occurs in children but is generally not clinically significant unless the total white blood count drops below 3,000/mm3. Carbamazapine has been reported to cause agranulocytosis and aplastic anemia (PDR, 1994). The greatest risk for agranulocytosis occurs within 3 months of initiating therapy, although it has been reported as late as 5 years after starting treatment (PDR, 1994). Finally, there are also reports of hepatic toxicity. In children with seizure disorders, reported cognitive and behavior effects include impaired performance on learning and memory tasks, irritability, agitation, insomnia, and emotional lability (Carpenter and Vining, 1993). However, Stores et al. (1992) found no significant cognitive or behavior effects after 1 year of therapy with either Carbamazapine or valproate.

Prior to initiating therapy, baseline blood counts and liver functions should be obtained. Carbamazapine blood levels are obtained frequently (once per week or more) at first to adjust the dose. Once the dosage is stabilized, periodic monitoring of serum levels, blood counts, and liver functions is needed (every 3 to 6 months) (Rosenberg et al., 1994). Since periodic monitoring does not ensure detection of agranulocytosis, the patient and family need to be advised of presenting symptoms (e.g., fever, easy bruising). If significant bone marrow suppression occurs, the Carbamazapine should be stopped and the blood counts closely monitored, with referral for specialized pediatric care. Finally, Carbamazapine may have teratogenic effects (especially with first trimester exposure) and has been associated with neural tube defects (Rosa, 1991).


Benzodiazepines may be useful for treating agitated manic states (APA, 1994a; Viesselman et al., 1993). These agents, used in conjunction with antimanic agents (i.e., lithium or anticonvulsants) and in place of neuroleptics, are helpful for psychomotor agitation, irritability, and insomnia in acutely manic patients (APA, 1994a; Viesselman et al., 1993). In adults, clonazepam and lorazepam have been most often studied, but no literature is available on their use in children and adolescents with mania (Werry and Arnan, 1993). The evidence is mixed as to whether benzodiazepines have any specific antimanic activity or whether their therapeutic effects are due solely to sedation (APA, 1994a). The few studies to date have had small sample sizes, and some were confounded by the coadministration of neuroleptics (APA, 1994a). No controlled studies examine the use of benzodiazepines for maintenance therapy or bipolar depression. Although alprazolam has been reported to be helpful for major depression, it also can induce mania (APA, 1994a).

Thus, benzodiazepines can be useful adjuncts to antimanic agents for patients with acute mania. However, their long term use in children and adolescents with bipolar disorder should be discouraged, given the lack of supporting research and potential dependency problems.


No studies examine the efficacy of neuroleptics for early-onset bipolar disorder (although they are commonly used in clinical practice). In the adult literature, neuroleptics have been shown to be effective for the treatment of acute mania (Goodwin and Jamison, 1990). However, it is not clear whether the effects of neuroleptics are actually antimanic or due to sedation (Goodwin and Jamison, 1990). Since their effects occur more rapidly than mood stabilizers, they may be useful during the initial phases when patients are highly agitated or psychotic (APA, 1994a). Little evidence indicates that neuroleptics, by themselves, should be used for maintenance treatment of bipolar disorder (APA, 1994a). Some patients whose symptoms do not respond adequately to antimanic agents alone may benefit from a combination of a mood stabilizer with an antipsychotic (APA, 1994a).

The side-effect spectrum of neuroleptics dictates close monitoring of their use and periodic reassessment of their ongoing use over the course of therapy (Campbell et al., 1993). Patients with mood disorders may be at greater risk to develop tardive dyskinesia (APA, 1991). In addition, the combination of lithium and neuroleptics has been associated with an increased risk of extrapyramidal side effects and neurotoxicity (Alessi et al., 1994).

Other Antimanic Agents

Clozapine and the other atypical antipsychotic agents (e.g., risperidone), may have mood stabilizing effects in patients with bipolar disorder, including those with psychotic features, mixed episodes, and/or rapid cycling (APA, 1994a). Clozapine was reported to be effective in an adolescent with bipolar disorder (Fuchs, 1994). Other agents with reported efficacy in the adult literature include calcium channel blockers and thyroid hormones (usually as an adjunct to other antimanic agents) (APA, 1994a). However, these agents have not been studied for the treatment of early-onset bipolar disorder, and their use in this population should only be considered after more standard regimens have been tried.

Antidepressant Medications

No studies with youth and only a few studies in the adult literature examine the efficacy of tricyclic antidepressants for the treatment of bipolar depression (Zornberg and Pope, 1993). They are more effective than placebos, but their usefulness either compared or used in conjunction with antimanic agents, such as lithium, has not been systematically studied (Zornberg and Pope, 1993). There are even fewer studies examining monoamine oxidase inhibitors and selective serotonin uptake inhibitors (Zornberg and Pope, 1993). Some evidence indicates that the antidepressant bupropion may be less likely than other antidepressants to induce mania in patients with bipolar disorder, although further study is needed (APA, 1994a; Haykal and Akiskal, 1990; Sachs et al., 1994). One case has been reported of the use of light therapy as an adjunct to antimanic agents for treating depressive symptoms in seven young adults, aged 16 to 22 years (Papatheodorou and Kutcher, 1995).

A major risk of antidepressant use in bipolar patients is the induction of mania and/or rapid cycling. This risk has been noted with all classes of antidepressant agents (APA, 1994a). Thus, they are generally used only as adjuncts to antimanic therapy for persistent depressive symptoms.


Psychostimulants may exacerbate or induce psychosis/mania (Waters and Mitch, 1993). However, there are case reports of stimulants having antimanic effects (Max et al., 1995). There are two reports of using lithium and methylphenidate together in treating preadolescent children (eight patients, total) with a combination of mood and disruptive behavior problems (Carlson et al., 1992; Licamele and Goldberg, 1989). Some improvements were noted, without significant side effects. Although the patients in these reports had some manic symptoms, none met the criteria for mania. Psychostimulants must be used with caution in patients with bipolar disorder and are best avoided during acute manic phases.

Electroconvulsive Therapy

In adults, electroconvulsive therapy (ECT) is as effective as lithium for the treatment of mania (APA, 1994a; Welch, 1989). Furthermore, it is the most effective therapy available for depression, particularly for patients with psychotic depression or for those nonresponsive or intolerant of medication therapy (APA, 1994a; Welch, 1989). Many patients with bipolar disorder respond quickly to ECT. It is extremely safe as long as modern methods are used (i.e., appropriate anesthesia, alterations in the delivery of the electrical stimulus, the selected use of unilateral treatment, and cardiopulmonary monitoring) (APA, 1994a). For a more thorough review of standard guidelines for the proper use and techniques, please see the APA Task Force on ECT report (APA, 1990).

ECT is generally considered the treatment of choice for bipolar disorder in the following clinical situations: (1) pregnancy; (2) catatonia; (3) neuroleptic malignant syndrome; and (4) any other medical condition where more standard medication regimens are contraindicated (APA, 1990).

Although the literature is sparse, case reports indicate that ECT is beneficial for children and adolescents with bipolar disorder, including mania, rapid cycling, and depressed phases (Bertagnoli and Borchardt, 1990). Potential side effects include short-term cognitive impairment, anxiety reactions, disinhibition, and altered seizure threshold (Bertagnoli and Borchardt, 1990). Despite its potential efficacy, many centers do not use ECT for patients with early-onset bipolar disorder due to a lack of experience and concern about its associated social stigma. Since ECT may be the most appropriate treatment for some patients, it should not be denied because of these barriers.

Treatment of Mixed Episodes and Rapid Cycling

Based on the adult literature, the treatment of mixed episodes is basically the same as that for mania, although lithium may be less often effective (APA, 1994a; Prien and Potter, 1990). Antidepressants are best avoided. The anticonvulsants may be particularly effective for mixed episodes, but further research is needed (APA, 1994a; Prien and Potter, 1990).

Similar to patients with mixed episodes, those with rapid cycling may be less responsive to lithium (APA, 1994a; Prien and Potter, 1990). Clinical experience suggests that antidepressants be avoided and that a combination of antimanic agents is often necessary to relieve symptoms (APA, 1994a). It is also important that patients with rapid cycling mania undergo a thorough medical evaluation to rule out potential medical conditions (e.g., thyroid disorder, substance abuse) that may be exacerbating the affective instability (APA, 1994a).


A comprehensive multimodal treatment approach combining psychopharmacology with adjunctive psychosocial therapies is almost always indicated for early-onset bipolar disorder. Medications help with the core symptoms of the illness, but they do not necessarily address the associated functional impairment. Preexisting behavioral disorders, substance abuse disorders, learning problems, and confounding psychosocial issues may all require additional treatment once the affective episode is stabilized. Furthermore, psychotherapeutic interventions are needed to promote medication compliance and avoid relapse.

Despite these clinical needs, the available research examining psycho social interventions for bipolar disorder is quite limited, and there is essentially no literature regarding children and adolescents (Kafantaris, 1995). No psychotherapy specific for the treatment of mania has been defined, although some pilot work is examining cognitive therapy interventions in adult patients (Kahn, 1990; Prien and Potter, 1990). A large literature addresses the effectiveness of cognitive behavioral and interpersonal psychotherapist for the treatment of depression in adults (APA, 1989). The use of these psychotherapies with children and adolescents will be reviewed in a separate practice parameter.

Extrapolating from the adult literature and clinical experience with this population suggests several areas where psychotherapeutic interventions should be directed:

Psychoeducational Therapy. Information should be provided to both the patients and their families on the disorder's symptoms and course, treatment options, potential impact on psychosocial and family functioning, and heritability (Goodwin and Jamison, 1990; Kahn, 1990; Prien and Potter, 1990).

Relapse Prevention. Education should be provided to the patient and family on the impact of noncompliance with medications, the recognition of emergent relapse symptoms, and other factors that may promote relapse (e.g., sleep deprivation, substance abuse). Medication noncompliance is a major contributor to relapse. Estimates of noncompliance in adults range from 25% to 50% (Prien and Potter, 1990). In a study of adolescents with bipolar disorder, 90% of those who were noncompliant with lithium relapsed over an 18-month period (Scrober et al., 1990). Thus, efforts must be made to educate both the patient and family of the importance of ongoing treatment. Establishing a strong therapeutic relationship and providing regular follow-up assessments are important in maintaining compliance (Kahn, 1990). Reports of supportive group therapies indicate that they are helpful for adult patients, but this modality has not been systematically studied (Prien and Potter, 1990).

Similar to schizophrenia, measures of family patterns of expressed emotion and interactional style are predictive of relapse in bipolar patients (Miklowitz et al., 1988). Preliminary evidence in studies of adults with bipolar disorder suggests that psychoeducational therapies with both the patient and family can effectively decrease relapse rates and improve overall functioning when used in conjunction with pharmacotherapy (APA, 1994a; Clarkin et al., 1990; Miklowitz and Goldstein, 1990; Prien and Potter, 1990). In these models, psychotherapeutic interventions are directed at stress reduction, problem-solving skills, and family functioning.

Miklowitz and Goldstein (1990) modified the Behavioral Family Management program, a home-based intervention designed for patients with schizophrenia (Falloon et al., 1984), to use with adult patients who have bipolar disorder. This program utilizes psychoeducation and training to enhance problem-solving and communication skills. Although further study is needed, use of this program, in conjunction with pharmacotherapy, does appear to decrease relapse rates (APA, 1994a).

Reduction of Morbidity. Bipolar disorder significantly affects social, family, academic, and developmental functioning. Therefore, in addition to efforts directed at reducing further episodes, psychosocial interventions are needed to address the myriad of disruptions that emerge in the wake of the disorder.

There is a dearth of systematic research regarding the efficacy of standard psychotherapeutic interventions with this population. Therefore, the use of these interventions is based solely on clinical consensus. Individual (usually supportive) and family psychotherapies may be indicated to deal with the various interpersonal, intrapsychic, and social conflicts that arise during the course of the disorder. Cultural issues must be taken into account when devising psychotherapeutic strategies. The use of more intensive forms of psychotherapy may be contraindicated in some patients with bipolar disorder because they cannot tolerate the affect generated (Bemporad, 1989).

The educational needs of youth with bipolar disorder must be adequately addressed to help promote long-term academic growth, especially given the high rates of comorbid disruptive behavior disorders. School consultation is often necessary to help develop an appropriate educational environment. Consultation with the school system and personnel can also be helpful to alert them to early warning signs of mania or depression. Some youth will need specialized educational programs, including day treatment/partial hospitalization programs.

Consultation may be needed with other involved community, juvenile justice, and/or social welfare programs. Some youth, either because of the severity of the symptoms or confounding environmental stressors, will need referral for intensive community-based services to continue living in their homes. Alternatively, some patients will need either foster care and/or residential services. Finally, patients and families often benefit from participation with community support and advocacy programs.

Bipolar disorder may first present during childhood and adolescence; therefore, clinicians working with youth should be familiar with its diagnosis and treatment. Although the same diagnostic criteria are used as for adults, issues regarding the developmental presentation of symptoms and the co-occurrence of other psychiatric disorders must be recognized.

Treatment involves the combination of pharmacotherapy and adjunctive psychosocial therapies. As in adults, early onset bipolar disorder may represent a lifelong condition that requires longitudinal assessment, medication monitoring, and ongoing psychosocial interventions. Treatment varies according to the phase of the disorder, with an overall focus on (1) amelioration of acute symptoms, (2) prevention of relapse, (3) reduction of long-term morbidity, and (4) promotion of long-term growth and development.

Next Section: Practice Parameters Beginning of Article