Keywords

Background and Significance

Introduction

Bipolar disorder (BD) is a familial illness characterized by episodes of abnormally elevated mood that are above and beyond the child’s developmental stage [1, 2]. It is now widely accepted that Bipolar Disorder (BD) occurs in children and adolescents and the controversy has shifted from a debate about whether it can be diagnosed in youth, now to how it can be diagnosed, differentiated from more common psychiatric disorders in childhood such as attention deficit hyperactivity disorder (ADHD), and how it can be treated and prevented [3].

BD in youth is increasingly recognized as a significant public health problem that is often associated with impaired family and peer relationships, poor academic performance, high rates of chronic mood symptoms and mixed presentations, psychosis, disruptive behavior disorders, anxiety disorders, substance use disorders, medical problems (e.g., obesity, thyroid problems, diabetes), hospitalizations, and suicide attempts and completions [3]. Early identification of BD in youth is essential for not only stabilizing their mood but also for enabling them to follow a normative developmental path and prevent an unrecoverable loss in their psychosocial development and education [4]. Moreover, youth with BD can have greater utilization of medical services and higher behavioral health costs relative to youth with unipolar depression or non-mood disorders. Youth with undiagnosed BD may have even more behavioral health costs than those with diagnosed BD [3]. This chapter reviews the epidemiology, clinical aspects, differential diagnosis, natural course, and treatment of pediatric BD with ADHD.

Epidemiology

Recent studies have shown dramatic increases in recognition and rates of BD in youth over the past 20 years and some authors have questioned the possibility of overdiagnosing children with BD in the US, whereas many others have brought up the possibility of long neglecting the presence of this condition in childhood [5]. The prevalence BD spectrum in adults is around 5% (BD I is around 1%), and the majority had the onset of their mood symptoms before age 20 years [6]. Those with ADHD usually have an early onset BD.

In clinical populations of youth in the US, the prevalence of BD has been reported between 0.6 and 15% depending on the setting, the referral source, and the methodology to ascertain BD [7]. Based on findings from two separate samples of juveniles with ADHD, Wozniak, Biederman, and colleagues have reported 20–21% had comorbid bipolar disorders [8, 9]. Other investigators have been skeptical about so many patients with ADHD meeting true diagnostic criteria for bipolar disorders. The skeptics have argued that: (1) the chronic irritability of these children was inconsistent with the episodic nature of mania described in DSM criteria; (2) the investigators did not consider the overlap of ADHD and manic symptoms (e.g., talkativeness, distractibility and hyperactivity); and (3) there was also an unusually high rate of bipolar disorders among subjects in the control groups of these samples. Instead, Klein and colleagues have argued that the children with ADHD described by the MGH group as having bipolar disorders most likely had severe conduct disorders or intermittent explosive disorders instead [10].

A recent meta-analysis about epidemiology of BD in youth around the world reported that the overall rate of BD was 1.8% (95% CI, 1.1–3.0%), enrolling 16,222 youth between the ages of 7 and 21 years during a period from 1985 to 2007 [11]. This meta-analysis suggested that there was no significant difference in the mean rates of BD-I in youth between the US and the non-US studies, but the US studies had a wider range of rates, especially when a broader definition of BD was used. In addition, the prevalence of BD-I in youth is similar to the current prevalence estimates of BD in adults, and while BD it is being diagnosed more commonly in clinical settings, the prevalence of BD in youth in the community is not increasing [11]. The same meta-analysis concluded that BD can begin in childhood, but the prevalence is much higher during the adolescence [11]. A large epidemiological study in the US reported slightly higher rates of BD-I and BD-II in female than in male adolescents (3.3% vs. 2.6%, respectively) with increasing rates of BD with older ages [12]. Studies in clinical populations have reported that the rates of bipolar spectrum disorders in youth are equally common in males and females [1, 13]. The rates of BD-II and adolescent-onset BD are more common in females [2].

Etiological Factors

Twin and adoption studies have demonstrated that the heritability of BD is quite high at 80%, indicating that 80% of the condition is determined by genetic rather than environmental factors [14,15,16]. However, several genes and epigenetic factors seem relevant to the manifestation of this illness, and identification of the genes associated with BD has not been conclusive. Other factors such as variation in ascertainment, phenotype definition, control selection, and limited power have led to inconsistent results.

Studies evaluating the risk for BD in offspring of parents with BD and in first-degree relatives of youth with BD have provided further evidence that BD runs in families [2, 16, 17]. It is suggested that offspring of parents with BD have up to 25-fold greater rates of BD when compared to offspring of control parents [2, 18, 19]. Importantly, a large prospective, high-risk BD study has suggested that offspring with mood lability, depression/anxiety, and in particular having subsyndromal manic symptoms and parents who had early-onset BD were at 50% risk to develop BD [20]. However, children of parents with BD are also at risk to develop depression, anxiety, ADHD, and behavioral problems [2, 21]. It is suggested that ADHD was associated with a significantly higher risk for switches from unipolar to bipolar disorder (28% vs. 6%, over 7 years) [22]; however, a review of high-risk studies suggested that the clinical diagnosis of childhood ADHD is not a reliable predictor of the development of BD in offspring studies [23]. Similarly, in Pittsburgh Bipolar Offspring Study , the rates of ADHD were not statistically different (38.9% vs. 19.8%) between bipolar offspring who did or did not develop BD during follow-up [2]. However, symptoms of inattention (in addition to depression and anxiety) may be part of a mixed clinical presentation during the early stages of evolving BD in high-risk offspring [23]. The same analysis reported preliminary evidence that childhood ADHD may form part of a neurodevelopmental phenotype in offspring at risk for developing a subtype of bipolar disorder unresponsive to lithium stabilization [23].

Research and clinical experience also suggest that trauma or stressful life events can trigger an episode of BD, though many BD episodes occur without an obvious or identifiable cause. In brief, the etiology is multifactorial, with complex interactions of biological vulnerabilities and environmental influences. The few studies that have evaluated the effects of psychosocial factors on the onset and maintenance of pediatric BD have found that low socioeconomic status (SES), exposure to negative events, and high “expressed-emotion” (EE) in the family are associated with poor prognosis [2, 24,25,26,27].

Assessment and Differential Diagnosis

Bipolar Symptoms and Subtypes

It is very important to have a common language and to use similar terms appropriately between professionals (and with patients and families) when describing, reporting, and monitoring mood changes in youth. According to the DSM-IV, there were four subtypes of BD: Bipolar I, Bipolar II, Cyclothymia and Other Specified/Unspecified Bipolar and Related Disorders (BD-Not Otherwise Specified; NOS) and their diagnostic criteria were the same between adults and children, with the exception of cyclothymia [28]. Of note, subtypes of BD in youth may not be stable over time. In a 4-year follow-up study, 25% of youth with BD-II converted to bipolar I and 45% of those with BD-NOS converted to BD-I or II [29]. Similar to the offspring studies [2, 23], it is important to note that the rates of ADHD in converters vs. non-converters were not significantly different (61.9% vs. 63.6%, respectively), nor were rates of stimulant use (33.33% vs. 28.6%) or family history of ADHD (42.9% vs. 40.3%), respectively [29].

To date, almost all studies in pediatric BD have employed the Diagnostic and Statistical Manual (DSM)-IV criteria [30]. However, there are important differences to keep in mind regarding the diagnostic criteria of BD in DSM-5 [31]. Key changes regarding criteria for bipolar and related disorders in DSM-5 relative to DSM-IV include: (1) requiring “increased energy or goal-directed activity” as a main symptom criteria for a manic episode (in addition to “elated mood or irritability”); (2) replacing mixed episodes with the new specifier of “mixed features (presence of three or more depressive symptoms during the course of a manic episode); and (3) introducing the new specifier of “anxiety features (presence of anxiety symptoms specifically during the course of an manic episode)” [31].

There are also now several important differences in the diagnostic criteria of DSM-5 relative to the latest International Classification System (ICD-10) [32]. In order to make the bipolar affective disorder diagnosis, in DSM-5 only a single manic episode is required [31], while in ICD-10, at least two mood episodes are required, one of which is manic or mixed (e.g., a mixture or rapid alteration of manic and depressive symptoms), while the other could be a depressive, hypomanic, manic, or mixed [32]. In both DSM-5 and ICD-10, duration criteria for hypomanic and a manic episodes are 4 and 7 days, respectively. Having a history of a hypomanic episode plus at least one major depressive episode is classified as a “bipolar II disorder” in DSM-5, and as an “other bipolar affective disorder” in ICD 10. Finally, the DSM-5 criteria require the presence of functional impairment with these altered moods, while ICD-10 criteria require that mood and activity levels must be “significantly disturbed.”

Figure 8.1 shows various presentations of either BD-I or BD-2 based on their combination of manic, hypomanic, and major depressive episodes. BD-I requires presence or history of a manic episode with or without major depressive episode. For BD-I diagnosis, both symptom criteria (three or four symptoms in addition to elation or irritability, respectively) and duration criteria should be met, in addition to the “significant functional impairment or psychosis” during mania. For the duration criteria, a manic episode should last at least seven consecutive days or lead to an inpatient admission anytime during the episode. Mixed presentations of mania in DSM-5 require three or more depressive symptoms during a manic episode, while mixed episodes in DSM-IV required meeting symptom criteria of both mania and major depression during the same mood episode (simultaneously or in rapid sequence).

Fig. 8.1
figure 1

This figure illustrates the minimum combinations of mood episodes required over time for patients with the two main types of bipolar disorders. Figure (a) depicts one potential presentation of a patient with bipolar I (BD-I), who has had an early major depressive episode followed by a manic episode. Figure (b) depicts another potential presentation of BD-I, in a patient who has had a single episode of mania but no other mood episodes. Figure (c) depicts a potential presentation of a bipolar II disorder (BD-II), with at least one major depressive episode, followed by an episode of hypomania with insufficient duration and symptom severity to meet criteria for full manic episode, which would make it convert to a BD-I diagnosis

Full size image

BD-II is characterized by at least one major depressive episode (for at least 2 weeks with functional impairment) and at least one hypomanic episode (lasting at least four consecutive days). Hypomania is described as the milder form of a manic episode during which the patient has a “distinct change” from the baseline functioning (sometimes patients may appreciate these changes in functioning, for example being able to work on more projects), but should “not have marked functional impairment” during the course of the hypomanic episode.

Cyclothymia is characterized by numerous periods of hypomania alternating with numerous periods of depressive mood or loss of interest or pleasure, which do not meet full criteria for a BD or a major depressive episode. The abnormal mood swings must be present for at least 1 year in youth or for 2 years in adults for the diagnosis.

Other specified BD (a.k.a. BD-NOS) is used when there are features of hypomanic or mixed episodes that do not meet the diagnostic criteria for any of the more specific BD subtypes. Because BD-NOS criteria is vague in DSM, researchers have developed clearer definitions to identify a BD-NOS diagnosis, such as having had at least one episode of hypomanic symptoms for 2 days, or at least four episodes of hypomania lasting 4 h each but being one symptom shy of meeting full symptom criteria [13, 15].

Clinical Presentations and Assessment of Bipolar Disorder

There is consensus in the field that children and adolescents may fulfill the strict DSM criteria for BD-I and II. The American Academy of Child and Adolescent Psychiatry (AACAP) has released the practice parameters for BD and recommended that clinicians should adhere to the DSM, including the duration criteria (requirement of an episodic change in mood lasting at least 4 days for hypomania and 7 days for mania) [33]. Typically, youth are given a BD NOS diagnosis, when they do not meet the duration criteria for a diagnosis of BD-I or BD-II [29].

The assessment of symptoms of mania, hypomania, and depression requires careful probing and in most cases, longitudinal assessment. In addition to the specific manic/hypomanic and depressive symptoms, it is important to ascertain the frequency, intensity, number, and duration (FIND) of the depressive and manic/hypomanic episodes [1, 3]. Most experts agree that these mood symptoms should exist as a collection of concurrent symptoms and behaviors (i.e., “cluster together”), occur episodically, and reflect a change in the child’s baseline characteristics. For example, preexisting hyperactivity and inattentiveness should not be counted towards mood symptoms unless there is significant worsening of these symptoms during the mood episode. It is also imperative to obtain information from caregivers, teachers, and other providers in order to accurately assess symptoms and potential change in functioning. Children’s chronological age, intellectual capabilities, and environmental factors are important when assessing their levels of functional impairment or improvement.

The most widely used interviews in BD studies are the Kiddie Schedule for Affective Disorders and Schizophrenia for school-age children-Present and Lifetime version (K-SADS-PL) (available for free at http://www.wpic.pitt.edu/research under tools and assessments) and the Washington University KSADS (WASH-U-KSADS) [1, 2]. However, these are mainly used for research purposes, require training of the interviewer, and are lengthy and time-consuming. Thus, symptom checklists for BD are also useful, such as Parent General Behavior Inventory (GBI-10; [34]) and the parent Child Mania Rating Scale (CMRS-10; [35]). Dimensional scales such as Child Behavior Checklist (CBCL) [36] and some of its subscales such as dysregulation profile (attention, aggression, and anxiety/depression) and externalizing problems can help identify significant psychopathology in those with ADHD and/or at risk for BD [37], but they are not specific for BD [38].

As expected, because of varying methodologies and conceptualizations of pediatric BD, there is also significant variability in the prevalence of individual manic symptoms among the studies. Similar to the meta-analysis published in 2005 [39], a recent meta-analysis (average age: 11.5 years old, 64% male, mainly Caucasian, 2226 youth in 20 published studies) reported that the most common symptoms across BD subtypes include increased energy, irritability and mood lability, distractibility, and goal directed activity (all approximately 75%) [40]. Grandiosity and hypersexuality were the most specific symptoms, but they were less common (57% and 32%, respectively). There are key issues about making an accurate BD diagnosis in youth, such as the requirement of clearly identified mood episodes (e.g., episodicity), the importance of cardinal symptoms and irritability, subthreshold presentations, bipolar depression, and preschool presentations.

Episodicity

Despite suggestions by some investigators that episodicity is not needed to diagnose pediatric BD, most other investigators and clinicians, as well as the AACAP guidelines [33], recommend that episodicity be required to diagnose BD diagnosis in youth. In fact, it is suggested to first focus on determining the presence of mood episodes based on the DSM manic/hypomanic symptoms, and then to determine how much these DSM manic/hypomanic symptoms occur during an identifiable time frame.

Cardinal Symptoms

Few investigators, aiming to be more specific and avoid misdiagnosing BD in youth, have suggested that elated mood and grandiosity must be present together to diagnose pediatric BD [41]. However, this is not required in DSM and some youth may have BD without elation and grandiosity. There is considerable heterogeneity among studies in the rates of these symptoms reflecting differences in samples (e.g., origin and age) and methodologies. It is important to consider difficulties in identifying elation and grandiosity, especially in younger children.

Irritability

Irritability is a very common symptom present in BD youth [39] and the absence of any episodes with irritability decreases the likelihood of a BD diagnosis. On the other hand, irritability rarely occurs in manic youth without elation. Irritability is also part of the other disorders such as Oppositional Defiant, Major Depressive, Generalized Anxiety, and Post-Traumatic Stress Disorders, and is also frequently present in youth with other psychiatric diagnoses such as ADHD and Autism Spectrum Disorders. Thus, irritability has low specificity for BD, and is analogous to how having fever/pain suggests only suggests that “something is wrong” [39]. DSM criteria for a manic episode explicitly allows for the presence of irritable mood alone to satisfy the main criterion. However, they require an additional manic symptom to meet the manic episode criteria (e.g., four or more manic symptoms should accompany irritability during the same time frame (clustering) for the mood episode). It is suggested that the severity and duration of irritability (the “super-angry/grouchy/cranky-type” of irritability, but not the “mad/cranky” or oppositional defiant disorder-type irritability) [42] are important clinical factors when assessing BD subjects. Furthermore, in order to be counted as a symptom of manic episode, irritability needs to be episodic even if the child has preexisting irritability (e.g., worsening of irritability during the manic episode when other comorbid disorders exist such as anxiety disorders, ADHD, or ODD) [43].

In contrast to episodic irritability, chronic irritability has recently been conceptualized as the core feature of a new diagnosis category that is included in the DSM-5 (a.k.a. disruptive mood dysregulation disorder (DMDD) ). DMDD has previously been referred to as temper dysregulation disorder with dysphoria (TDDD) and severe mood dysregulation (SMD) . Chronic irritability has also been associated with attention deficit hyperactivity disorder, oppositional defiant disorder, and major depressive disorder rather than with BD, whereas episodic irritability has been associated with BD and anxiety [44].

Depression

Most youth seen in psychiatric clinics experience their first episode of BD as depression (Birmaher et al. 2007) [2]. Similar to adults, depressive episodes are reported to be the most common manifestation of BD in children and adolescents based on both frequency and duration [15]. The presence of psychosis, family history of BD, and pharmacologically induced mania/hypomania may indicate susceptibility to develop BD [45,46,47]. Early identification and treatment of BD depression is of vital importance, because it is associated with increased risk for psychosocial impairment and suicide as compared to unipolar depression [48].

Preschool Presentations

Validity of manic symptoms such as grandiosity and elation have been questioned in preschool children (aged from 3 to 7 year old) given the emotional and cognitive developmental stage of the younger children. Few available studies have suggested that preschool children may have BD diagnosis [49]. Irritability is more common, and grandiosity and elation are also suggested as helpful in preschool children when differentiating BD from other disorders such as major depressive disorder and disruptive behavior disorder (DBD) [49].

Bipolarity and Comorbidities

The presence of the comorbid disorders affects the child’s and adolescent’s response to treatment and prognosis, indicating the need to identify and treat them effectively. Comorbid disorders, particularly disruptive behavior disorders (DBD; 30–70%), ADHD, (50–80%), and anxiety disorders (30–70%) are very common [1, 3]. Beginning in adolescence, the rate of comorbid substance abuse steadily increases [50, 51]. The prevalence of these disorders will depend on the methods used to diagnose them, the source of the sample studied (e.g., more common in clinical versus community), and the age range of its members, with more ADHD and Oppositional Defiant Disorder (ODD) in children, and more conduct and substance use disorders in adolescents.

Bipolar Disorders and ADHD

Clinicians must be cautious about attributing symptoms to mania or hypomania unless they show a clear temporal association with the abnormally elevated, expansive and/or irritable mood (plus increased activity/energy levels). Clinicians should also carefully observe in a child with possible BD whether the symptoms of the comorbid disorder disappear or persist while the suspected child with BD is euthymic, and whether the symptoms associated with BD worsen during the mood episode. Biederman and colleagues have argued that comorbidity between ADHD and BD cannot be dismissed due to the shared features of the two disorders (e.g., distractibility, motor hyperactivity, and talkativeness) [52]. These investigators have shown that even after removing overlapping diagnostic criteria, children with each condition can still be distinguished. Geller and colleagues have reported that several manic symptoms such as elated mood, grandiosity, hypersexuality, decreased need for sleep, racing thoughts (but not hyperenergetic and distractibility) are substantially and significantly more frequent among youth with BD relative to youth with ADHD [53]. The following features, when present, were suggested to help distinguish the diagnosis of BD in a child with ADHD: (a) the ADHD symptoms appear later in life (e.g., after age 12 years old), (b) the symptoms of ADHD appear abruptly in an otherwise healthy child, (c) the ADHD symptoms previously responded to stimulants but now do not, (d) the ADHD symptoms come and go and tend to occur with mood changes, (e) periods of exaggerated elation, grandiosity, depression, no need for sleep, or inappropriate sexual behaviors, (f) recurrent, severe mood swings, temper outbursts, or rages, (g) hallucinations and/or delusions, (h) a strong family history of BD, particularly if the child is not responding to appropriate ADHD treatments [54].

Longitudinal Course and Differential Diagnosis

The differential diagnosis may require longitudinal rather than a cross-sectional assessment. Mood time-lines or diaries and using school years, birthdays, and holidays as anchors are very helpful in the assessment and monitoring mood symptoms and episodes. The mood time lines (mood monitoring) instruments should be user/child-friendly and can be modified (regarding the child’s age, culture, and interests) to increase compliance. In addition to mood, energy levels can be monitored simultaneously for diagnostic clarification. From this perspective, Dr. Diler has developed a novel self- report mood rating (the “Mood and Energy Thermometer, MET © ”) for daily assessment and monitoring of mood state in bipolar track adolescents (as shown in Fig. 8.2 ) [55]. This scale aims to provide a practical way of monitoring complex mood cycles and daily schedule. Given the confusion about several 1 to 10 scales (e.g., a 10 could mean extreme depression or extreme mania or no depression), a common language between the youth, care givers, and providers is necessary. Moreover, many children report their energy levels more accurately than their mood and therefore, energy levels are incorporated in the mood rating. The Mood and Energy Thermometer (MET©) rates mania and increased energy on a 1 to 10 scale and rates depression and tiredness on −1 to −10 scale, and aims to form a common language between patients, families, and clinicians (please see MET© at the end of the chapter). The inclusion of energy level measurements is consistent with the DSM-5, because energy level is now considered as a main symptom criterion for mania. Bipolar or depressed patients are encouraged to rate their mood and energy levels every day on this scale to make them better able to identify and record their mood symptoms, which has significant clinical value for not only treatment but also to detect and prevent a future episode early.

Fig. 8.2
figure 2

This figure shows the Mood and Energy Thermometer , which allows patients with potential bipolar disorders to quantify how their levels of mood and energy vary over time

Full size image

It is important to be aware of various potential trajectories of mood and ADHD symptoms over time. For example, though youth with ADHD may show different trajectories, their hyperactivity and impulsivity may lessen as they age into adolescence and adulthood [56]. Recent studies have shown that BD is mainly characterized by recovery and recurrences, and that ongoing fluctuations in mood symptoms, especially subsyndromal depressive and mixed symptoms, are common [15, 26]. In a sample of youth with BD followed over a 9-year period, a recent study used latent growth class analyses to identify the proportions of the overall sample falling into one of four different mood trajectories: (1) 24.0% had a “predominantly euthymic” course; (2) 34.6% had a “moderately euthymic” course, (3) 19.1% had an “ill with improving” course, and (4) 22.3% had a “predominantly ill” course [57]. While a majority of all youth with BD in this study (59%) had at least a somewhat favorable course (groups 1 and 2), each less favorable trajectory group had a higher rate of ADHD (42.1% in “predominantly euthymic” group, 59.1% in “moderately euthymic” group, 64.3% in “ill with improving” group, and 72% in “predominantly ill” group) [57]. However, whether subjects managed to achieve a euthymic mood of ≥18 months during the study, was not associated with having ADHD. While many children and adolescents with BD may also have ADHD, most longitudinal studies have not reported an association between ADHD or stimulant treatment with conversions to mania [23, 29]. In another study, which involved the longitudinal assessment 707 children for manic symptoms, 421 had ADHD alone, 45 had manic-symptoms alone, 117 had both ADHD and manic symptoms, and 124 had neither [58]. Comorbidity (16.5%) was slightly less than expected by chance (17.5%), suggesting ADHD was not a risk factor for manic symptoms [58]. The investigators speculated that the increased rates of bipolarity noted in other studies of ADHD samples are likely due to excessive rates of these disorders in child outpatient settings, not because ADHD is a true risk factor for BD [58]. Once again, the co-occurrence of ADHD and BD was associated with significantly worse global functioning, symptom severity, and comorbidities relative than in either condition alone [58]. However, the rates of ADHD were not significantly different between the four different trajectories of manic symptoms during the 24-month follow-up [59].

Treatment

Psychoeducation and support start with the assessment phase and are always indicated at any phases of treatment. Family members and the patient should be educated about the causes, symptoms, course, and different treatments of BD and the risks associated with treatment options [1]. Sleep hygiene and routine are also very important, especially because sleep deprivation leading to worsening of mood symptoms. Ensuring a stable circadian rhythm has a positive effect on physiology and daily functioning. In addition to supportive psychotherapy, specific psychosocial treatment packages for youth with BD target acute affective symptoms and prevention, or delay of recurrences, improvement of adherence to treatment, and management of comorbid conditions. A central feature of all psychosocial treatment models [such as Child and Family Focused Cognitive Behavior Therapy (CFF-CBT) [61], Multi-family Psychoeducation Groups (MFPG) and Individual Family Psychoeducation (IFP) [61], Family Focused Therapy (FFT) specifically for adolescents with BD (FFT-A) [62], Interpersonal and Social Rhythm Therapy (IPSRT) [63], and Dialectical Behavior Therapy (DBT) [64] for pediatric BD] is that they include psychoeducation, problem-solving, and coping skills. Parents are closely engaged in their child’s therapy and are referred for treatment themselves if they too have clinically significant symptoms.

Current studies suggest that the most efficacious and fastest way to yield response for acute manic/mixed episodes is with the atypical antipsychotics such as aripiprazole [65], asenapine [66], olanzapine [67], risperidone [68], quetiapine [69], and ziprasidone [70]. Response rates to the atypical antipsychotics in studies of acute manic and mixed episodes among children and adolescents are comparable to those among adults, but youth are more sensitive to these medications’ metabolic side effects [71, 72]. The antiepileptic mood stabilizers and lithium can also be helpful, but seem less efficacious in younger patients than in adults. However, most of the studies in children or adolescents have lasted only 8 weeks, possibly an inadequate time to observe a full response to these medications [1].

BD comorbid with ADHD presents unique treatment considerations. The AACAP treatment guidelines advise that symptoms of BD should be stabilized first, and if impairing symptoms of ADHD persist, they may be judiciously treated, with stimulants as first-line [73]. Several studies have suggested that comorbid ADHD may reduce the responsivity of acute mania to pharmacotherapy [74,75,76,77], especially in patients who are adolescents or have BD-I [76]. Although longitudinal, naturalistic studies did not find an association between stimulant treatment and emergence of BD [2, 29], there have been concerns about the risk of treatment-emergent mania or mood destabilization with stimulant treatment [78]. For example, 2.5–10% of BD youth treated with stimulants or atomoxetine (adjunctive to mood-stabilizing medication) experience psychiatric adverse events (i.e., hypo/mania and/or suicidality), and discontinuation of stimulants is often associated with an improvement in such events [79,80,81,82].

Tricyclic antidepressants have been reported to help improve comorbid ADHD symptoms, but also significantly increase the risk for manic symptoms [83]. In an open-label trial of valproate in 40 youth with BD and ADHD, only 7.5% had a positive response for their ADHD, while 80% had a positive response for their bipolar disorder. In the same study, using a 2-week crossover design with mixed amphetamine salts (MAS; 5 mg twice daily) or placebo, ADHD symptom reduction was significantly greater for MAS compared to placebo [81]. Similarly, a subsequent 4-week, randomized control trial of adjunctive methylphenidate vs. placebo reported significant improvement with methylphenidate [80]. In a randomized trial of adjunctive methylphenidate or placebo added to aripiprazole, ADHD symptoms showed no significant between-group differences, but self-reported depressive symptoms improved more with methylphenidate [79]. An open-treatment study of atomoxetine suggested that it may be effective for comorbid BD and ADHD [82].

Conclusions and Future Directions

Much evidence now shows that youth may manifest classical symptoms of BD; however, many youth do not fulfill the current DSM BD-I or II criteria for such diagnoses, primarily because they lack the required duration of symptoms. Moreover, many youth referred for evaluation for BD have severe mood lability, irritability, inattentiveness, hyperactivity, verbal and/or physical aggression that need careful baseline assessment, and longitudinal follow-up to ascertain whether these symptoms are indeed manifestations of BD (e.g., clustering of manic symptoms, change from baseline and functional impairment). Early recognition and acute and maintenance treatment of BD in children and adolescents is of vital importance to ameliorate ongoing syndromal and subsyndromal symptomatology and to reduce or prevent the serious psychosocial morbidity and risk for suicide. There are growing numbers of imaging [84,85,86] and neurocognitive [87,88,89,90] studies attempting to identify core features of BD that can distinguish it from other disorders including ADHD; however, larger longitudinal studies are needed in BD and ADHD youth for this expanding knowledge about neurobiology/cognition to be implemented into clinical practice for each patient.

The extant pharmacological studies suggest that the atypical antipsychotics are helpful for the acute treatment of manic/mixed symptoms, and stimulants can be added for ADHD after mood stabilization. However, studies of longer duration, including maintenance trials to reduce the risk of relapses and recurrences, as well as to examine the effects of development, family environment and psychopathology, and side effects are urgently needed in BD and ADHD youth. Future studies evaluating possible preventative strategies for ADHD and depressed youth at high risk for BD are indicated. In addition, studies to evaluate protective factors (e.g., cognitive development, social and coping skills, environmental factors) are warranted. Genetic and other biological studies, including pharmacogenetic studies that correlate the effects of treatment and biochemical changes on the brain, are also needed to improve precision in matching potentially helpful treatments with the most suitable patients.