Bipolar disorder subtypes in children and adolescents: Demographic and clinical characteristics from an Australian sample

Author's Accepted Manuscript

Bipolar disorder subtypes in children and adolescents: Demographic and clinical characteristics from an Australian sample Stephen J. Hirneth, Philip L. Hazell, Tanya L. Hanstock, Terry J. Lewin

www.elsevier.com/locate/jad

PII: DOI: Reference:

S0165-0327(14)00807-6 http://dx.doi.org/10.1016/j.jad.2014.12.021 JAD7173

To appear in:

Journal of Affective Disorders

Received date: 11 December 2013 Revised date: 2 December 2014 Accepted date: 4 December 2014 Cite this article as: Stephen J. Hirneth, Philip L. Hazell, Tanya L. Hanstock, Terry J. Lewin, Bipolar disorder subtypes in children and adolescents: Demographic and clinical characteristics from an Australian sample, Journal of Affective Disorders, http://dx.doi.org/10.1016/j.jad.2014.12.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Bipolar disorder subtypes in children and adolescents: Demographic and clinical characteristics from an Australian sample

Stephen J. Hirnetha,b*, Philip L. Hazellb,c, Tanya L. Hanstockd, Terry J. Lewine a

Hunter New England Local Health District, Child and Adolescent Mental Health Service, Newcastle, NSW

Australia b

University of Sydney, Sydney Medical School, Discipline of Psychiatry, Sydney, NSW, Australia

c

Sydney and South Western Sydney Local Health Districts, Infant Child and Adolescent Mental Health

Services, Sydney, NSW, Australia d

School of Psychology, University of Newcastle, Callaghan, NSW, Australia

e

University of Newcastle and Hunter New England Mental Health Services, Centre for Translational

Neuroscience and Mental Health, Newcastle, NSW, Australia *

Corresponding author at: Newcastle Child and Adolescent Mental Health Service, 621 Hunter Street,

Newcastle West, NSW 2302 Australia. Tel: +61 2 49257800; fax: +61 2 49257863.

E-Mail address: [email protected]

Abstract Background: Bipolar disorder (BD) phenomenology in children and adolescents remains contentious. The study investigated Australian children and adolescents with bipolar I disorder (BD-I), bipolar II disorder (BD-II), or BD not otherwise specified (BD-NOS).

Methods: Index episode demographics, symptomatology, functioning and diagnostic data were compared for 88 participants (63 female) aged 8-18 years (M = 14.8, SD = 2.5) meeting DSM-IV-TR criteria for BD-I (n = 24), BD-II (n = 13) or BD-NOS (n = 51).

Results: BD-I had higher rates of previous episodes, psychotropic medication (compared to BD-II but not BD-NOS), rates of inpatient admissions (compared to BD-NOS), and number of inpatient admissions (compared to BD-II). BD-II had lower rates of lifetime depression and anxiety disorders, higher frequency of hypomania, shorter duration of illness, and fewer previous episodes. BD-NOS had younger age of onset,

2

chronic course, irritability and mixed presentation. All BD subtypes had high rates of self-harm (69.3%), suicidal ideation (73.9%), suicide attempts (36.4%), psychiatric admission (55.7%), and psychosis (36.4%).

Limitations: There were relatively small numbers of BD-I and BD-II. Diagnoses were based on retrospective recall.

Conclusions: All BD subtypes had high levels of acuity and clinical risk. In accord with previous results, BD-I and BD-II participants’ phenomenology was consistent with classical descriptions of these subtypes. BD-NOS participants were younger, with less euphoric mania but otherwise phenomenologically on a continuum with BD-I, suggesting that child and adolescent BD-NOS may be an early and less differentiated phase of illness of BD-I or BD-II and hence a target for early intervention.

Keywords Bipolar disorder; Subtypes; Children; Adolescents; Phenomenology

1.

Introduction The phenomenology of bipolar disorder (BD) in children and adolescents remains a contentious topic in

psychiatry practice and research. There is general acceptance that BD symptomatology often has an onset in the mid to late teenage years (American Psychiatric Association, 2002; Mitchell et al., 2003) and that BD can present in children and younger adolescents, but with a much less certain prognosis (Carlson and Meyer, 2006; Kowatch et al., 2005; Kyte et al., 2006). Researchers remain divided as to whether instances of BD in children and adolescents represent a more potent presentation of the classic adult illness, a separate subtype of BD, a variant of other conditions such as attention deficit-hyperactivity disorder (ADHD), or a newly recognized illness unrelated to adult-onset BD (Kyte et al., 2006). Researchers attempting to define the core features of child and adolescent BD have generally focused on bipolar I disorder (BD-I) (Biederman et al., 1999; DelBello et al., 2007; Findling et al., 2001; Geller et al., 2002b; Jairam et al., 2004; Srinath et al., 1998). Such studies have been reviewed comprehensively elsewhere (Birmaher and Axelson, 2006; Jairam et al., 2006; Kowatch et al., 2005). In summary, current research suggests that children and adolescents with BD-I tend to be predominantly male, present with an episodic course, history of functional impairment, high rates of prior hospitalization, medication treatment, mixed mood episodes

3

longitudinally, suicidal ideation, psychotic symptoms and suicide attempts (Axelson et al., 2006; Birmaher et al., 2009; Geller et al., 2002b; Hazell et al., 1999; Masi et al., 2007). In contrast to the many BD-I samples described in the literature, far fewer studies have described other variants of child and adolescent BD such as bipolar II disorder (BD-II) and bipolar disorder not otherwise specified (BD-NOS). Only two large-scale studies have described in detail the phenomenological features of BD-I, BD-II and BD-NOS subtypes in children and adolescents: First, the Course and Outcomes of Bipolar Youth (COBY) study (Axelson et al., 2006) is a US multi-site prospective naturalistic study that examined 438 children and adolescents (M = 12.7 years) with BD-I (n = 255), BD-II (n = 30), and BD-NOS (n = 153) recruited from 3 US academic medical centres, with primarily an outpatient focus. Second, Masi et al. (2007) recruited 217 patients (M = 13.6 years; n = 78 BD-I, 97 BD-II and 42 BD-NOS) from a unit for inpatient and outpatient children and adolescents with mood and anxiety disorders based in Pisa, Italy. Between them, the studies found BD-I participants had higher scores on overall symptom severity, and comparatively high rates of psychosis, psychiatric hospitalization, suicide attempts, and use of psychotropic medication; BD-II participants had more anxiety disorders and lifetime depressive episodes; and participants with BD-NOS exhibited an earlier age of onset, and more frequent irritability and comorbid ADHD and oppositional-defiant disorder (ODD) (Axelson et al., 2006; Masi et al., 2007). Other studies have included children and adolescents with bipolar spectrum diagnoses, but without examining subtype differences (Faedda et al., 2004; Findling et al., 2010; Lewinsohn et al., 1995; Scott et al., 2013; Soutullo et al., 2009; Starling et al., 2013). One major challenge in researching BD subtypes in children and adolescents is the lack of clear and consistent diagnostic instruments and criteria for defining BD-NOS. For example, Axelson et al. (2006) and Masi et al. (2007) both used the Schedule for Affective Disorders and Schizophrenia for School-Age ChildrenPresent and Lifetime Version (KSADS-PL) (Kaufman et al., 1997) to determine diagnoses, however, the COBY study modified the present episode version to include severity ratings for mania and depression (e.g. KMRS, Axelson et al., 2003). A modified version of the KSADS – the Washington University in St Louis KSADS (WASH-U-KSADS) - has been purpose-built for examining bipolar spectrum disorders in children and adolescents to account for such methodological limitations (Geller et al., 1998; Geller et al., 1996). No studies to date however, have reported on the characteristics of BD subtypes using purpose-built instruments such as the WASH-U-KSADS. We have provided a comparison of the methodological differences between WASH-UKSADS and KSADS-PL, and how the instruments have been applied across studies of child and adolescent BD

4

subtypes (see Table 1). Although an extensive discussion of the differences is outside the scope of this study, the reader is referred to Galanter et al. (2012) for a thorough discussion. Given the limited existing research, there is a need for further studies of child and adolescent BD subtypes in other cultural contexts and using instruments that are designed specifically for this population. The present study reports on the index episode characteristics of an Australian child and adolescent BD sample, analysing BD subtypes and comparing findings with existing US and non-US studies to add to the phenomenological description of child and adolescent BD subtypes internationally. To our knowledge, the study is the first examining the phenomenology of bipolar spectrum disorders in an Australian child and adolescent sample.

[Insert Table 1 about here]

1.1 Hypotheses 1.1.1 Clinical presentation BD-I participants will score higher on measures of overall clinical acuity (higher rates of psychosis, psychiatric hospitalization, suicide attempts, and psychotropic medication). BD-NOS participants will have higher levels of irritability (Axelson et al., 2006; Masi et al., 2007). BD-I participants will have lowest scores, and BD-II participants highest scores, on measures of psychosocial functioning (Axelson et al., 2006; Masi et al., 2007).

1.1.2 Course of illness BD-I participants will have higher rates of multiple mood episodes, BD-II participants will have higher rates of lifetime depressive episodes, and BD-NOS participants will have an earlier age of onset (Axelson et al., 2006; Birmaher et al., 2009; Masi et al., 2007).

1.1.3 Patterns of comorbidity BD-II participants will have more internalizing (anxiety) disorders, and BD-NOS participants will have more externalizing disorders (ADHD, ODD and conduct disorder [CD]) (Axelson et al., 2006; Birmaher et al., 2009; Masi et al., 2007).

5

2.

Methods

2.1 Participant selection and assessment Participants were recruited between 2005 and 2010 from The Bipolar Program (TBP), a specialist public community mental health clinic for the assessment and management of possible bipolar disorder in young people aged 5-18 years, embedded within a local child and adolescent mental health service (CAMHS) located in Newcastle, Australia. Newcastle is a metropolitan area of New South Wales with a population of about 308,000; approximately the population of the US city of Pittsburgh, Pa. Sample characteristics are described in detail in the following section. The study was approved by the relevant ethics committees (Hunter New England Local Health District and University of Sydney). Index episode data were collected at service admission, based on current symptoms and function using the standard assessment procedures of the clinic, which included initial assessment (clinical interview and history taking), subsequent structured diagnostic interview and final review by psychiatrist. Psychosocial functioning was assessed using the Children’s Global Assessment Scale (CGAS) (Shaffer et al., 1983). Socioeconomic status (SES) was assessed based on parental occupation using the Australian Socioeconomic Index 2006 (AUSEI06) (McMillan et al., 2009), a numerical occupational status score (0-100) based on the Australian and New Zealand Standard Classification of Occupations (ANZSCO). Diagnoses were established using the Washington University in St Louis Kiddie Schedule for Affective disorders and Schizophrenia (WASH-U-KSADS) (Geller et al., 1998; Geller et al., 1996; Geller et al., 2001) conducted with both parents and clients. The WASH-U-KSADS was chosen as it is designed specifically for assessing child and adolescent mood disorders, with the methodology being quite similar to the modified protocol of Axelson et al. (2006). Assessments were conducted by clinical psychologists (SH and TH, one of whom [TH] has received formal training in WASH-U-KSADS administration with Dr Geller’s research team in St Louis, achieved acceptable inter-rater reliability with Dr Geller’s research team and subsequently provided comprehensive training for the other clinician, including direct observation and feedback). All diagnoses were verified by a child and adolescent psychiatrist. Clients with schizophrenia, autism, intellectual disability, or a primary medical or substance-related mood disorder were excluded. Inclusion criteria were current or past diagnosis of BD-I, BD-II or BD-NOS. Diagnoses of BD-I and BD-II were made based on standard DSM-IV-TR criteria (American Psychiatric Association, 2000) using the WASHU-KSADS scoring procedures as described by Geller and colleagues (Geller et al., 1996; Geller et al., 2002b;

6

Geller et al., 2002c) and with reference to considerations described by Leibenluft et al. (2003): Mania and hypomania were defined as episodes of elevated/expansive mood, irritability and / or grandiosity meeting DSMIV-TR criteria and with duration criteria of 1 week for mania and 4 days for hypomania. The inclusion of irritability as a criterion in this ‘narrowband’ definition is controversial (Hunt et al., 2009; Leibenluft et al., 2003) and it should be noted that no participants in our study were diagnosed with BD-I or BD-II solely on the basis of irritability. Clients with DSM-IV defined rapid cycling of moods meeting DSM-IV duration criteria for mania or hypomania were defined as having BD-I or BD-II. Mixed episodes were defined as overlapping time periods of mania/hypomania and major depressive disorder on WASH-U-KSADS items. Age at onset of bipolar disorder was defined as per the COBY study (Axelson et al., 2006) as the age at onset of a DSM-IV mood episode or an episode fulfilling our criteria for BD-NOS (see below). Episodes were defined as a period of active symptoms (≥ 4 or ‘moderate’ on WASH-U-KSADS items; ≥ 3 or ‘mild’ for hypomania) demarcated at each end by a period of nil or minimal mood and behavioral symptoms consistent with the person’s baseline level of functioning; that is, ratings of 1 (nil) or 2 (slight) on WASH-U-KSADS items. Durations of BD diagnoses were calculated, as per Geller et al. (2004) as the number of weeks between onset and offset of full DSM-IV syndromal criteria or BD-NOS criteria. Our definition of BD-NOS was informed by considerations regarding episode duration and severity criteria, as addressed by Leibenluft et al. (2003), the WASH-U-KSADS scoring manual (Geller et al., 1996), and research studies using the WASH-U-KSADS (Geller et al., 2002a; Geller et al., 2002b). The key consideration in defining BD-NOS is the level of resolution used for defining the duration of ‘episodes’: Regarding the WASH-U-KSADS, there has been particular debate regarding the instrument’s measurement of cycles that last for hours to days: Geller and colleagues for example (Geller et al., 2002a; Geller et al., 2002b; Leibenluft et al., 2003), using the WASH-U-KSADS, defined those with ‘ultra-rapid’ (cycling most days) or ‘ultradian’ cycling as having BD-I with ‘continuous rapid cycling’, whereas most other studies have conceptualised chronic daily or intra-day cycles as BD-NOS (Axelson et al., 2006; Leibenluft et al., 2003; Masi et al., 2007). Even those authors that argue against Geller and colleagues’ approach have differing opinions: Leibenluft et al. (2003) recommend that BD-NOS should be defined as episodes lasting 1-3 days; the COBY study (Axelson et al., 2006) defines BD-NOS as episodes lasting > 4 hours per day, with at least 4 lifetime days; and Masi et al. (2007) do not set any minimum duration of symptoms. In addressing this issue of defining BD-NOS, we have incorporated recommendations from Leibenluft et al. (2003), and previous similar studies (Axelson et al., 2006), in order to set a definition that is comparable to

7

previous research, whilst also being compatible with the WASH-U-KSADS protocols: BD-NOS episodes were defined as periods of (hypo)mania not otherwise specified (symptoms meeting full ‘narrow phenotype’ DSM-IV severity criteria for mania or hypomania, but with sub-threshold duration criteria), for ≥ 4 hours per day and less than 4 continuous days. Thus, rapidly cycling moods over days are included in our definition of BD-NOS rather than BD-I or BD-II. For BD-NOS, mixed episodes were defined as per previous authors (Geller et al., 2002b; Leibenluft et al., 2003) as symptoms meeting these (hypo)mania NOS criteria and major depressive episode diagnostic criteria, either occurring simultaneously or rapidly alternating over hours to days. Regarding the issue of conceivably diagnosing bipolar disorder on the basis of a single episode lasting only 4 hours, the WASH-UKSADS protocol requires mood symptoms to be present ‘daily or most days’ (rather than within a single day) for symptoms to be rated as clinically significant (Geller et al., 2008). Conversely, BD-NOS episode offset, using this WASH-U-KSADS protocol was defined as a period where no clinically significant mood disturbance was present ‘daily or most days’. Using this definition, BD-NOS episodes are very heterogeneous including a range of mostly chronic presentations including those with frequent periods of short-lived (hypo)mania, those with fluctuating manic and depressive symptoms alternating over periods of 1-4 days, and those with ongoing depressive symptoms and regular periods of (hypo)manic symptoms.

2.2 Sample Characteristics The present study reports on a subset (n = 88) of the overall Bipolar Program clinic sample (N = 127), after excluding participants due to no diagnosis of BD (28%), intellectual disability (3%), and primary substance – induced or medical cause (0.8%). Data were not available to compare the Bipolar Program sample to clients from the wider CAMHS clinic. Regarding characteristics of the overall clinic sample (N = 127, 38 male, 89 female), clients were referred from a range of sources including local community mental health services (51.2%), inpatient units (15.8%), general medical practitioners (22.8%), school counsellors (6.3%) and paediatricians (1.6%). One third of the referred patients had a prior diagnosis of bipolar disorder. Most (78%) were adolescents (aged 13-18). Clients not included were diagnosed with depression (30%), anxiety (28%), other mood disorder (7%), ADHD (5%) and nil diagnosis (18%). There were no differences between clients included versus excluded from the study in terms of age, gender, education or employment status. Participants included in the study were significantly more likely than those excluded to have received a previous mental health service (89.8% vs 62.1%; Fisher’s exact test, p = .001), to be referred on discharge from an inpatient facility (20.5% vs 5.1%; Fisher’s exact test, p = .034), to be referred by a

8

psychiatrist (51.1% vs 15.4%; Fisher’s exact test, p < .001), and less likely to be referred by a general medical practitioner (13.6% vs 46.2%; Fisher’s exact test, p < .001). Included participants (n = 88, 25 male, 63 female) were aged from 8 to 18 years (M = 14.8, SD = 2.5), with most (58%) diagnosed with BD-NOS at intake, and a smaller proportion with BD-I (27.3%) and BD-II (14.8%) (see Table 2). BD-NOS participants were significantly younger (M = 14.1, SD = 2.7) than BD-I participants (M = 16.1, SD = 1.2, t(72.96) = 3.45, p < .001). Participants were primarily female (71.6%). SES tended to be low (around 40/100 on AUSEI06). A minority (34.1%) resided with both parents, most (76.1%) stated they were attending some form of educational facility (school, technical college, university or private training academy), and a minority were employed (17.0%). There was no difference in sex ratios, SES, household composition, or education / employment status across BD subtypes.

2.3 Statistical analyses Analyses used IBM SPSS Statistics version 21.0 (IBM Corp, 2012). A priori hypotheses regarding BD subtype differences were tested with planned pairwise comparisons (ANOVA LSD method for quantitative data, correcting for unequal variances where necessary; and Fisher’s exact test for categorical data). All significance tests were 2-tailed (α = .05). Further exploratory analyses were conducted on measures of symptom course and severity, functioning, comorbidity, and family psychiatric history, using parametric and nonparametric analyses. Uncorrected p values are reported, however a Bonferroni correction for multiple pairwise comparisons was applied to all post-hoc analyses, specifying α = .05/3 = .0167.

3. Results 3.1 Overall clinical characteristics Table 2 presents participant demographics and clinical characteristics. Onset of BD was typically in early adolescence (M = 12.32 years, SD = 3.04). There was a high rate of mixed mood episodes (83.0%) and comorbidity (87.5%), in particular anxiety disorders (50.0%), psychosis (36.4%), ADHD (30.7%), ODD (17.0%) and CD (12.5%). There were low rates of substance abuse or dependence. Most (75.0%) were in their first episode of BD, with a long duration of untreated BD (DUB) (M = 97.01 weeks, SD = 85.53) and a high rate of psychopharmacological treatment (85.2% lifetime, 75% current at study intake). Medications prescribed included sodium valproate (51.1% lifetime, 37.5% current), SSRI (53.4% lifetime, 30.7% current),

9

antipsychotics (52.9% lifetime, 39.7% current), lithium (26.1% lifetime, 18.2% current), lamotrigine (15.9% lifetime, 10.2% current), methylphenidate (17.0% lifetime, 8.0% current) and dexamphetamine (9.1% lifetime, 3.4% current). Regarding participants who were taking SSRI at study intake, 74% were being treated for a comorbid anxiety disorder, and 81.5% were also prescribed a mood stabiliser. Rates of medication use were highest in BP-I participants, but bipolar subgroup differences were not statistically significant for individual medications. No participants had an index presentation of major depressive episode with history of manic episode. All participants with a previous episode (25%) had a depressive episode as their previous episode. There was a high rate of clinical distress and acuity, including self-harm or suicidal gestures (69.3%), suicidal ideation (73.9%), and suicide attempts (36.4%). There was poor functioning overall (mean CGAS score = 53.39, SD = 9.81). Just over half of participants (55.7%) had experienced a psychiatric admission, with an average length of 1.98 weeks (SD = 2.35). The reasons for previous admissions were generally related to issues of suicidality (75.5%), aggression (16.3%) and mania (8.2%). All admissions due to mania occurred in the BP-I sample (22.2%), but there were no statistically significant differences overall in admission reason across bipolar subtypes.

[Insert Table 2 about here]

3.2 Clinical presentation 3.2.1 A priori hypotheses BD-I participants had significantly higher rates of previous psychiatric admissions (75.0%) compared to BDNOS (47.1%, Fisher’s exact test, p = .027) but not BD-II (53.8%, Fisher’s exact test, p = .76). BD-I participants had significantly more inpatient admissions (per patient admitted, M = 2.39, SD = 1.65) compared to the BD-II participants (M = 1.43, SD = 0.54, t(22.73) = 2.19, p = .04) but not compared to BD-NOS (M = 3.50, SD = 4.15, t(40) = 1.07, p = .29). BD-I participants had higher rates of psychotropic medication (95.8%) compared to BD-II (69.2%; Fisher’s exact test, p = .04) but not compared to BD-NOS (84.3%, Fisher’s exact test, p = .26). BD-I participants did not show significantly higher rates of psychosis, number of suicide attempts or rate of suicide attempts (see Table 2). There was a significantly higher rate of index episode irritability in BD-NOS participants (100%) compared to BD-I participants (87.5%; Fisher’s exact test, p = .03) and BD-II participants (76.9%;

10

Fisher’s exact test, p = .007), although it should be noted only 11.8% of BD-NOS participants presented with pure irritability in the absence of elation. BD-I participants were not different to BD-II or BD-NOS on psychosocial functioning (CGAS; see Table 2).

3.2.2 Post-hoc analyses BD-NOS participants had a significantly higher rate of mixed index mood episodes (96.1%) compared to BD-I (62.5%; Fisher’s exact test, p < .001) and BD-II (69.2%; Fisher’s exact test, p = .01; see Table 2). BDNOS participants had a significantly longer duration of index episode (M = 118.98, SD = 110.33) compared to BD-I (M = 72.00, SD = 53.12, t(85) = 2.10, p = .039) and BD-II (M = 60.38, SD = 42.26, t(85) = 2.09, p = .040). However, these findings did not remain significant after Bonferroni correction.

3.3 Course of illness 3.3.1 A priori hypotheses BD-I participants had significantly higher rates of previous mood episodes (45.8%) compared to BD-II (7.7%; Fisher’s exact test, p = .027) and BD-NOS (19.6%; Fisher’s exact test, p = .027; see Table 2). All previous mood episodes across all subtypes were depressive episodes; consequently, the BD-II participants had significantly lower rates of lifetime depressive episodes (7.7%) compared to BD-I (45.8%; Fisher’s exact test, p = .027) but were not significantly different to BD-NOS (19.6%, Fisher’s exact test, p = .44). BD-NOS participants (M = 11.40, SD = 3.21) had a significantly younger age at onset compared to BD-I (M = 13.50, SD = 2.45; t(57.87) = 3.13, p = .003) and BD-II (M = 13.75, SD = 2.01; t(29.60) = 3.28, p = .003).

3.3.2 Post-hoc analyses BD-II participants had a significantly shorter duration of illness (see Table 2) from onset of first episode to baseline (M = 65.15 weeks, SD = 38.99) compared to BD-I (M = 127.33, SD = 105.15, t(32.18) = 2.59, p = .014) and BD-NOS (M = 133.84, SD = 108.53, t(62) = 2.23, p = .029), although the latter comparison did not remain significant after Bonferroni correction. There were no other group differences in duration of illness or DUB.

11

3.4 Patterns of comorbidity 3.4.1 A priori hypotheses BD-II participants had a significantly lower number of anxiety disorders (M = 0.54, SD = 0.88) compared to BD-I (M = 1.33, SD = 1.49; t(34.67) = 2.04, p = .049) but not reaching statistical significance compared to BDNOS (M = 0.76, SD = 0.97, see Table 2). Overall rates of anxiety disorders were (although not reaching statistical significance) lowest in the BD-II participants (38.5%) compared to BD-NOS (49.0%) and BD-I (58.3%). Rates of externalizing disorders were not higher in participants with BD-NOS (47.1%) compared to BD-I (45.8%) and BD-II (30.8%).

3.4.2 Post-hoc Analyses There were no other significant group differences in comorbidity after Bonferroni correction.

4. Discussion 4.1 Bipolar subtype differences There was a pattern of more established and ‘classic’ BD presentation for BD-I participants (longer duration of illness, presenting with manic or mixed index episodes characterized by elation, and high rates of previous episodes), consistent with the DSM-IV-TR diagnostic criteria for BD-I. Consistent with previous research, we found higher rates of hospitalization and psychotropic medication prescription for BD-I participants compared to BD-II and BD-NOS (Axelson et al., 2006; Masi et al., 2007). Our rates of psychosis for BD-I (45.8%) are within the range reported by BD-I samples using varying assessment protocols, with some reporting lower rates - such as Axelson et al. (2006) (34.5%); DelBello et al. (2007) (41%); Hazell et al. (1999) (40%) - and some reporting higher rates - Geller et al. (2002b) (59.3%), Jairam et al. (2004) (60%), Masi et al. (2007) (52.6%). Comparing our results to other BD-I studies using the WASH-U-KSADS, our rates of psychosis are considerably lower than Geller et al.’s (2002b) study, but are consistent with DelBello et al. (2007) who report a sample similar to ours, with predominantly older adolescent females (58%), and high rates of mixed episodes (73%). Geller et al.’s (2002b) study by contrast contains younger, mainly preadolescent males with lower rates of internalising / anxiety disorders and very high rates of ADHD and disruptive behaviour disorders. It is possible that the extreme presentation seen in Geller’s sample of difficult children with BD-I translates somehow in their methodology into higher levels of symptoms rating positive for psychosis (as discussed below

12

for example, the WASH-U-KSADS counts symptoms towards multiple diagnostic categories). In contrast to hypotheses and previous findings (Axelson et al., 2006; Masi et al., 2007), BD-I participants exhibited similar levels of clinical severity (psychosocial functioning, symptom severity ratings, suicidality, and psychosis) to BD-II and BD-NOS, lending support to the notion that BD-II and BD-NOS in children and adolescents are on a continuum with BD-I (Axelson et al., 2006; Birmaher et al., 2009). Consistent with previous research (Axelson et al., 2006; Masi et al., 2007), BD-II participants had milder presentations than BD-I and BD-NOS participants, characterized by significantly higher frequency of index episode hypomania, a significantly shorter duration of illness and fewer previous episodes. In contrast to previous studies (Axelson et al., 2006; Masi et al., 2007), the BD-II participants were less likely than BD-I participants to have experienced lifetime depression or anxiety disorders. BD-NOS was characterized by early onset, a chronic, less episodic pattern and involving irritable / mixed presentations. Nevertheless, BD-NOS participants displayed a duration of illness, level of functional impairment and symptomatology similar to BD-I, including high numbers of hospital admissions (in fact the highest, although not reaching statistical significance), and high rates of psychotropic medication, suicide attempts, comorbid psychosis, anxiety and disruptive behaviour disorders. Although all BD-NOS participants had symptoms of irritability, only (11.8%) had (hypo)mania symptoms characterised solely by irritability; there was no significant difference in rates of elation across BD subtypes. Our rates of psychosis for BD-II (23.1%) and particularly for BD-NOS (35.3%) are higher than the COBY study’s rates (BD-II 20%, BD-NOS 17.6%) (Axelson et al., 2006; Birmaher et al., 2009) and considerably higher than those of Masi et al. (2007) (BD-II 1%, BD-NOS 2.3%). As mentioned below, our BD-NOS sample was defined by full symptom severity criteria, rather than the ‘one less than DSM-IV’ approach used by the COBY study and Masi et al. (2007), and this may explain why higher rates of psychosis were found in our BD-NOS sample. Our BD-NOS sample was also considerably older than previous samples, and this is likely to have had an effect on our rates of psychosis being higher and more comparable to our other subtypes. The findings suggest that BD-NOS in children and adolescents may be an early and less differentiated phase of illness of BD-I or BD-II. Although longitudinal data on our sample would be required to support this assertion, data from other longitudinal studies of bipolar spectrum disorders generally support this conclusion: Birmaher and colleagues (Birmaher and Axelson, 2006; Birmaher et al., 2009) for example found that 25% of BD-NOS participants converted to BD-I or BD-II within 2 years, and 38% within 4 years.

13

4.2 Comparison of overall findings with previous studies Overall, in contrast to many previous studies of child and adolescent BD (Axelson et al., 2006; DelBello et al., 2007; Geller et al., 2002b; Hazell et al., 1999; Jairam et al., 2004; Masi et al., 2007; Soutullo et al., 2009), the current sample was older, predominantly female, with a higher rate of internalizing compared to externalizing disorders, predominantly BD-NOS, rather than BD-I, and with more consistently high levels of clinical severity across subtypes, in particular our BD-NOS group more closely resembling the other BD subtypes. There are a number of considerations that may explain the broader demographic and clinical differences to previous studies: Firstly, the demographic differences in our study compared to previous studies may reflect wider community trends. There is evidence from broader Australian surveys that the pattern of child and adolescent mood disorder presentations is consistent with the trends seen in our clinic: Surveys indicate approximately twice as many Australian female youths compared to males report high levels of depressive symptoms and affective disorders (possibly due to greater reporting by females); and that there are very low rates of BD amongst Australian 10-14 year olds, with rates increasing throughout teenage years (ABS, 2008; McDermott et al., 2010). The low rates of externalizing disorders seen in our study are likely to be a consequence of the predominance of adolescent females – the most recent national survey of Australian children and adolescents found that externalizing disorders are significantly more prevalent in males (Sawyer et al., 2000). Other studies in New South Wales have also noted similar demographic patterns to our study: A recent study of older adolescents and young adults with bipolar spectrum disorders in the nearby Sydney region (Scott et al., 2013), found 54.4% females, a high proportion of BD-NOS (46%), and low rates of externalising disorders (21%). A recent study of early-onset psychotic disorders in children and adolescents in the Sydney region (Starling et al., 2013) described a sample with affective psychosis that was also largely adolescent (mean age = 13.2) females (71%) with high rates of anxiety disorders (67%). A previous study in the Sydney region (McShane et al., 2006) included a small child and adolescent sample with BD – 100% female and average age 15.1 years. A previous study from the Newcastle region reported on an exclusively male sample with younger age and high rates of comorbid externalising disorders (Hazell et al., 1999); however, that study screened for mania in clients referred to a developmental clinic for pediatric ADHD and behavior problems. The present sample of predominantly mid-adolescent females, with less externalising disorders than previous samples, may have led to less easily diagnosing a manic episode and thus BD-I. Research also suggests that in Australia there

14

is a relatively conservative approach to diagnosing prepubertal BD (Parry et al., 2009), and this may have biased results towards adolescent samples. A second factor that may explain differences to previous studies is the recruitment approach. Previous studies (Axelson et al., 2006; Masi et al., 2007) have reported predominantly BD-I samples, but have recruited from a mixture of inpatient and outpatient facilities, including large research centres and these authors have acknowledged that their results may not generalize to the true population of child and adolescent BD (Axelson et al., 2006). Our study is based in a public psychiatric outpatient setting, so it may be that our predominantly BD-NOS sample is more representative of the wider child and adolescent BD population. In support of this assertion, our data is more similar to US epidemiological studies (Lewinsohn et al., 1995) that have similarly reported a majority of females, a higher rate of anxiety disorders compared to externalizing disorders, and a majority of individuals with diagnoses of BD-NOS, BD-II and cyclothymia (all with substantial impairment, as in our study). Such community studies suggest that the milder forms of the disorder may be more prevalent than, and equally as disabling as BD-I (Lewinsohn et al.). The majority of our sample (55.7%) had experienced previous inpatient psychiatric admissions, but frequently the reason for admission involved issues of suicidality, self-harm and depression, rather than exclusively mania. Accordingly, only 20.5% of referrals came as a result of hospital discharge. Lastly, our study used more restricted criteria for BD-NOS than previous studies of child and adolescent BD. Axelson et al. (2006) defined BD-NOS as a minimum of elated mood plus two associated DSM symptoms, or irritable mood plus three associated DSM symptoms and a change in functioning, with a minimum duration of four hours within a 24-hour period, occurring for at least four cumulative lifetime days. Masi et al. (2007) used similar severity criteria to Axelson et al. (2006) but did not set a minimum duration of mania/hypomania for BD-NOS. Our study used full DSM-IV-TR severity criteria for BD-NOS and with a minimum duration of 4 hours per day (as per WASH-U-KSADS instructions), but in practice the pattern identified was of chronic severe mood fluctuations and mixed symptomatology rather than individuals receiving a diagnosis based on periods of mania / hypomania lasting less than 1 day. Our use of full DSM-IV-TR severity criteria may have identified a more robust BD-NOS phenotype, and may also explain why our BD-NOS sample more closely resembled the BD-I sample compared to previous studies (Axelson et al., 2006; Masi et al., 2007).

15

4.3 Methodological limitations There were small numbers of BD-I and BD-II compared to our BD-NOS sample, and this may have reduced the power to detect effects in these subtypes. Rates of personality disorders were not assessed. The diagnostic data are based on semi structured interviews, which unavoidably summarise complex client presentations into coarse symptom descriptors and diagnostic groups, limiting the fidelity of client phenomenology. The structured diagnostic interview data also rely on the retrospective recall of participants and carers, which may affect the validity of the results, as with all such studies. Inter-rater reliability data were not obtained (although it should be noted that all diagnoses were confirmed by child and adolescent psychiatrists). The data are based on index episode presentations and do not allow conclusions to be drawn about the overall disorder. Methodological differences as described above (see Table 1) also limit the comparability to other studies to some degree (Galanter et al., 2012; Kowatch et al., 2005). As noted earlier (see Table 1) different semi structured interviews define bipolar subgroups slightly differently (particularly in the case of BD-NOS, where there is no consensus definition in the literature). The WASH-U-KSADS for example counts symptoms towards multiple diagnostic categories, whereas the KSADS-PL counts symptoms (eg concentration difficulties) towards a new disorder only if they represent a significant worsening from the subject’s previous state. Therefore the WASH-U-KSADS methodology used in our study may have led to a higher proportion of comorbid diagnoses, or conversely the possibility that some mood symptoms represented pre-existing behavioural issues or other disorders. Our study however reports lower rates of comorbid ADHD and externalising disorders compared to other studies using WASH-U-KSADS across both prepubertal (Geller et al., 2004; Geller et al., 2002b) and adolescent (DelBello et al., 2007) samples, and also compared to KSADS-PL studies (Axelson et al., 2006; Masi et al., 2007) arguing against our methodology inflating rates of comorbidity or mistaking ADHD or ODD for BD. Anxiety disorders could also be conceivably mistaken for BD; and although our study had higher rates of anxiety disorders compared to previous WASH-U-KSADS studies, the rates are within the range reported by other studies of BD subtypes, being higher than Axelson et al. (2006) but lower than Masi et al. (2007).

16

4.4 Conclusion This study is the first to our knowledge to examine the phenomenology of BD subtypes in an Australian sample of children and adolescents. The study findings are in general agreement with previous studies of child and adolescent bipolar spectrum disorders (Axelson et al., 2006; Masi et al., 2007). There were identifiable differences between child and adolescent BD subtypes that were broadly in line with expectations, supporting the notion that, in this age group, the DSM-IV-TR subtypes can be usefully applied. The BD-I sample appeared to represent a group of older teenagers whose illness had progressed often from depressive episodes through to severe manic episodes and attendant dysfunction. In general, the BD-II sample was milder in symptomatology and overall dysfunction. The findings for the BD-NOS sample were consistent with the notion that children and adolescents with BD-NOS are in the early stages of an evolving severe illness (Birmaher and Axelson, 2006), although the cross-sectional nature of this study precludes specific conclusions about the illness course. The implications of our findings are that all bipolar spectrum disorders in children and adolescents are associated with high levels of clinical morbidity and distress, and that BD-NOS in particular may be an early and less differentiated phase of illness of BD-I or BD-II in children and adolescents and hence a target for early intervention, although longitudinal data are required to support this theory. Research is in progress to map the longitudinal course of BD in our sample.

Conflict of interest All authors declare that they have no conflicts of interest.

Contributors Stephen Hirneth was the primary researcher responsible for study design; overseeing data collection, collation and entry; supervising research assistants; data analysis and interpretation; and writing the manuscript. Stephen Hirneth and Tanya Hanstock were the clinicians responsible for data collection. Philip Hazell was responsible for overseeing study design and research funding protocols. Terry Lewin provided consultation regarding statistical analysis. All authors contributed to final interpretation and editing of the manuscript.

17

Role of funding source The research was supported by funding from the Australian Rotary Health Research Fund (Mental Health Evaluation Grants Scheme). The grant provider did not have any role or input regarding study design, data collection, analysis, interpretation, conclusions, report-writing or decisions regarding publication.

Acknowledgements The authors would like to thank research assistants Tonelle Handley, Leanne Gulliver, and Julie Searl for their assistance in data collation and entry. The research was supported by funding from the Australian Rotary Health Research Fund (Mental Health Evaluation Grants Scheme).

References ABS, 2008. National Survey of Mental Health and Wellbeing: Summary of Results, 2007. Australian Bureau of Statistics, Canberra. American Psychiatric Association, 2000. Diagnostic and statistical manual of mental disorders, fourth edition, text revision. American Psychiatric Association, Washington, DC. American Psychiatric Association, 2002. Practice guideline for the treatment of patients with bipolar disorder. Am. J. Psychiatry 159, supplement. Axelson, D., Birmaher, B., Brent, D., Wassick, S., Hoover, C., Bridge, J., Ryan, N., 2003. A preliminary study of the Kiddie Schedule for Affective Disorders and Schizophrenia for School-age Children mania rating scale for children and adolescents. J. Child Adolesc. Psychopharmacol. 13, 463-470. Axelson, D., Birmaher, B., Strober, M., Gill, M.K., Valeri, S., Chiapetta, L., Ryan, N., Leonard, H., Hunt, J., Iyengar, S., Bridge, J., Keller, M., 2006. Phenomenology of children and adolescents with bipolar spectrum disorders. Arch. Gen. Psychiatry 63, 1139-1148. Biederman, J., Faraone, S.V., Chu, M.P., Wozniak, J., 1999. Further evidence of a bidirectional overlap between juvenile mania and conduct disorder in children. J. Am. Acad. Child Adolesc. Psychiatry 38, 468-476.

18

Birmaher, B., Axelson, D., 2006. Course and outcome of bipolar spectrum disorder in children and adolescents: A review of the existing literature. Dev. Psychopathol. 18, 1023-1035. Birmaher, B., Axelson, D., Goldstein, B., Strober, M., Gill, M.K., Hunt, J., Houck, P., Ha, W., Iyengar, S., Kim, E., Yen, S., Hower, H., Esposito-Smythers, C., Goldstein, T., Ryan, N., Keller, M., 2009. Four-year longitudinal course of children and adolescents with bipolar spectrum disorders: the Course and Outcome of Bipolar Youth (COBY) study. Am. J. Psychiatry 166, 795-804. Carlson, G., Meyer, S., 2006. Phenomenology and diagnosis of bipolar disorder in children, adolescents, and adults: Complexities and developmental issues. Dev. Psychopathol. 18, 939-969. DelBello, M.P., Hanseman, D., Adler, C.M., Fleck, D.E., Strakowski, S.M., 2007. Twelve month outcome of adolescents with bipolar disorder following first hospitalization for a manic or mixed episode. Am. J. Psychiatry 164, 582-590. Faedda, G.L., Baldessarini, R.J., Glovinsky, I.P., Austin, N.B., 2004. Pediatric bipolar disorder: phenomenology and course of illness. Bipolar Disord. 6, 305-313. Findling, R.L., Gracious, B.L., McNamara, N.K., Youngstrom, E.A., Demeter, C.A., Branicky, L.A., Calabrese, J.R., 2001. Rapid, continuous cycling and psychiatric co-morbidity in pediatric bipolar I disorder. Bipolar Disord. 3, 202-210. Findling, R.L., Youngstrom, E.A., Fristad, M., Birmaher, B., Kowatch, R.A., Arnold, L.E., Frazier, J., Axelson, D., Ryan, N., Demeter, C.A., Gill, M.K., Fields, B., Depew, J., Kennedy, S.M., Marsh, L., Rowles, B.M., Horwitz, S.M., 2010. Characteristics of children with elevated symptoms of mania: The Longitudinal Assessment of Manic Symptoms (LAMS) study. J. Clin. Psychiatry 71, 1664-1672. Galanter, C.A., Hundt, S.R., Goyal, P., Le, J., Fisher, P.W., 2012. Variability among research diagnostic interview instruments in the application of DSM-IV-TR criteria for pediatric bipolar disorder. J. Am. Acad. Child Adolesc. Psychiatry 51, 605-612. Geller, B., Craney, J.L., Bolhofner, K., Nickelsburg, M.J., Williams, M., Zimerman, B., 2002a. Two-Year Prospective Follow-Up of Children With a Prepubertal and Early Adolescent Bipolar Disorder Phenotype. Am. J. Psychiatry 159, 927-933. Geller, B., Tillman, R., Bolhofner, K., Zimerman, B., 2008. Child bipolar 1 disorder: Prospective continuity with adult bipolar 1 disorder; characteristic of second and third episodes; predictors of 8-year outcome. Arch. Gen. Psychiatry 65, 1125-1133.

19

Geller, B., Tillman, R., Craney, J.L., Bolhofner, K., 2004. Four-year prospective outcome and natural history of mania in children with a prepubertal and early adolescent bipolar disorder phenotype. Arch. Gen. Psychiatry 61, 459-467. Geller, B., Warner, K., Williams, M., Zimerman, B., 1998. Prepubertal and young adolescent bipolarity versus ADHD: assessment and validity using the WASH-U-KSADS, CBCL and TRF. J. Affect. Disord. 51, 93-100. Geller, B., Williams, M., Zimerman, B., Frazier, J., 1996. Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS). Washington University, St Louis, MO. Geller, B., Zimerman, B., Williams, M., Bolhofner, K., Craney, J.L., DelBello, M.P., Soutullo, C.A., 2001. Reliability of the Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS) Mania and Rapid Cycling Sections. J. Am. Acad. Child Adolesc. Psychiatry 40, 450-455. Geller, B., Zimerman, B., Williams, M., DelBello, M.P., Bolhofner, K., Craney, J.L., Frazier, J., Beringer, L., Nickelsburg, M.J., 2002b. DSM-IV mania symptoms in a prepubertal and early adolescent bipolar disorder phenotype compared to attention-deficit hyperactive and normal controls. J. Child Adolesc. Psychopharmacol. 12, 11-25. Geller, B., Zimerman, B., Williams, M., DelBello, M.P., Frazier, J., Beringer, L., 2002c. Phenomenology of prepubertal and early adolescence bipolar disorder: examples of elated mood, grandiose behaviours, decreased need for sleep, racing thoughts and hypersexuality. J. Child Adolesc. Psychopharmacol. 12, 3-9. Hazell, P.L., Lewin, T.J., Carr, V.J., 1999. Confirmation that Child Behavior Checklist clinical scales discriminate juvenile mania from attention deficit hyperactivity disorder. J. Paediatr. Child Health 35, 199-203. Hunt, J., Birmaher, B., Leonard, H., Strober, M., Axelson, D., Ryan, N., Yang, M., Gill, M., Dyl, J., EspositoSmythers, C., Swenson, L., Goldstein, B., Goldstein, T., Stout, R., Keller, M., 2009. Irritability without elation in a large bipolar youth sample: frequency and clinical description. J. Am. Acad. Child Adolesc. Psychiatry 48, 730-739. IBM Corp, 2012. IBM SPSS Statistics for Windows, Version 22.0. IBM Corp, Armonk, NY. Jairam, R., Anderson, R., Redwin, R., 2006. Pediatric bipolar disorder: Perspectives on course and outcome, In: Kotler, M.B. (Ed.), New Developments in Mania Research. Nova Science, New York, pp. 1-25.

20

Jairam, R., Srinath, S., Girimaji, S.C., Seshadri, S.P., 2004. A prospective 4-5 year follow-up of juvenile onset bipolar disorder. Bipolar Disord. 6, 386-394. Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C., Moreci, P., Williamson, D., Ryan, N., 1997. Schedule for affective disorders and schizophrenia for school-age children-Present and lifetime version (KSADS-PL): Initial reliability and validity data. J. Am. Acad. Child Adolesc. Psychiatry 36, 980-988. Kowatch, R.A., Youngstrom, E.A., Danielyan, A., Findling, R.L., 2005. Review and meta-analysis of the phenomenology and clinical characteristics of mania in children and adolescents. Bipolar Disord. 7, 483-496. Kyte, Z.A., Carlson, G.A., Goodyer, I.M., 2006. Clinical and neuropsychological characteristics of child and adolescent bipolar disorder. Psychol. Med. 36, 1197. Leibenluft, E., Charney, D.S., Towbin, K.E., Bhangoo, R.K., Pine, D.S., 2003. Defining clinical phenotypes of juvenile mania. Am. J. Psychiatry 160, 430-437. Lewinsohn, P.M., Klein, D.N., Seeley, J.R., 1995. Bipolar disorders in a community sample of older adolescents: prevalence, phenomenology, comorbidity, and course. J. Am. Acad. Child Adolesc. Psychiatry 34, 454-463. Masi, G., Perugi, G., Millepiedi, S., Mucci, M., Pari, C., Pfanner, C., Berloffa, S., Toni, C., 2007. Clinical implications of DSM-IV subtyping of bipolar disorders in referred children and adolescents. J. Am. Acad. Child Adolesc. Psychiatry 46, 1299-1306. McDermott, B., Baigent, M., Chanen, A., Fraser, L., Graetz, B., Hayman, N., Newman, L., Parikh, N., Peirce, B., Proimos, J., Smalley, T., Spence, S., 2010. beyondblue expert working committee clinical practice guidelines: depression in adolescents and young adults. beyondblue: the national depression initiative, Melbourne. McMillan, J., Beavis, A., Jones, F.L., 2009. The AUSEI06: A new socioeconomic index for Australia. Journal of Sociology 45, 123-149. McShane, G., Mihalich, M., Walter, G., Rey, J., 2006. Outcome of patients with unipolar, bipolar and psychotic disorders admitted to a specialist child and adolescent mental health service. Australas. Psychiatry 14, 198-201. Mitchell, P., Malhi, G., Redwood, B.L., Ball, J., 2003. Summary of guideline for the treatment of bipolar disorder. Australas. Psychiatry 11, 39-53.

21

Parry, P., Furber, G., Allison, S., 2009. The paediatric bipolar hypothesis: the view from Australia and New Zealand. Child. Adolesc. Ment. Health 14, 140-147. Sawyer, M.G., Arney, F.M., Baghurst, P.A., Clark, J.J., Graetz, B., Kosky, R.J., Nurcombe, B., Patton, G.C., Prior, M.R., Raphael, B., Rey, J., Whaites, L.C., Zubrick, S.R., 2000. The Mental Health of Young People in Australia. Commonwealth Department of Health and Aged Care, Mental Health and Special Programs Branch, Canberra. Scott, E.M., Hermens, D.F., Naismith, S.L., Guastella, A.J., De Regt, T., White, D., Lagopoulos, J., Hickie, I., 2013. Distinguishing young people with emerging bipolar disorders from those with unipolar depression. J. Affect. Disord. 144, 208-215. Shaffer, D., Gould, M.S., Brasic, J., Ambrosini, P., Fisher, P., Bird, H., 1983. A children’s global assessment scale (CGAS). Arch. Gen. Psychiatry 40, 1228-1231. Soutullo, C.A., Escamilla-Canales, I., Wozniak, J., Gamazo-Garrán, P., Figueroa-Quintana, A., Biederman, J., 2009. Pediatric bipolar disorder in a Spanish sample: Features before and at the time of diagnosis. J. Affect. Disord. 118, 39-47. Srinath, S., Reddy, Y.C., Girimaji, S.C., Seshadri, S.P., Subbakrishna, D.K., 1998. A prospective study of bipolar disorder in children and adolescents from India. Acta Psychiatr. Scand. 98, 437-442. Starling, J., Williams, L.M., Hainsworth, C., Harris, A.W., 2013. The presentation of early-onset psychotic disorders. Aust. N. Z. J. Psychiatry 47, 43-50.

22

Table 1 Comparison of WASH-U-KSADS and KSADS-PL methodology for studies of BD subtypes in children and adolescents.a KSADS-PL WASH-U-KSADS General characteristics Age range

Instructions Axelson et al. (Kaufman et al., 1997) (2006) 6-18b 7-17 inclusive

Scaling for criteria 3-point scalec

Score required for 3 (threshold) threshold symptom

Masi et al. (2007) 8-18

Instructions Current Study (Geller et al., 1996) 6-18b 8-18

Mood disorders: as per WASH-UKSADS. Non-mood disorders: as per KSADS-PL Not stated; Scales for mood sections are similar to WASH-UKSADS items. Non-mood items assumed to be as per KSADS-PL.

Not explicitly stated; assumed to be as per KSADSPL

6-point Likert scale, As per WASH-UKSADS 7-point for depressed mood

Not explicitly stated; assumed to be as per KSADSPL

≥4 for most items, ≥3 for hypomania and dysthymia

As per WASH-UKSADS

Only if symptoms intensify with the onset of abnormal mood. Comorbid diagnoses not assigned if they occurred exclusively during a mood episode Present: most severe week in last month Lifetime: Most severe week everd

Not explicitly stated; assumed to be as per KSADSPL

Yes. All items that occur within overlapping time frames are counted towards diagnoses

Yes, as per WASH-UKSADS

Not explicitly stated; assumed to be as per KSADSPL

Onset(s) & offset(s) As per WASH-UKSADS requested for all symptoms for each episode across the lifespan

Symptoms count towards multiple diagnoses

No

Time period assessed for present and lifetime episodes

Present: Past year without 2 month remission. Lifetime: most severe episode

Age at onset BD

Not defined

Age when the Not defined subject first met DSM-IV criteria for a major mood episode or the BD-NOS criteria

Not defined

Age when the subject first met DSM-IV criteria for a major mood episode or the BD-NOS criteria

Episode duration

NA

Not specified

Period ≥ 7 days without a free interval (return to baseline) of ≥ 1 month

NA

Training level

Clinician

Trained research clinician

Child and adolescent psychiatrists

Clinician

Number of weeks between onset and offset of full threshold DSMIV syndromal criteria or BDNOS criteria Clinical Psychologist

23

Skip-out criteria

Informants

Parent vs child interviewer; scoring

Yes

Mood sections: No (as per WASH-UKSADS). Non-mood sections: Yes Child or adolescent and Child or parente (other sources of adolescent and a parent or primary information are incorporated into parent caregiver interview)

Same or different interviewer; same interviewer implied in instructions

Integrating parent Both parent and child's vs child response response scored; either informant's response can satisfy criteria

Handling of divergent information

Introductory clinical interview conducted

Irritability Questions Location

Interviews of the child or adolescent and a parent or primary caregiver (about the subject) Not explicitly stated; assumed to be as per KSADS-PL

Not explicitly stated; assumed to be as per KSADSPL

No

No

Child or adolescent and parent; other sources of information incorporated during historytaking and psychosocial assessment Same interviewer

Child or adolescent and parent (other sources of information are incorporated into parent interview)

Child or adolescent and parent; other sources of information incorporated during historytaking and psychosocial assessment Same interviewer used

Different interviewers requirede

Both parent and child's response scored; either informant's response can satisfy criteria Child and parent Interviewers use clinical Not explicitly stated; assumed responses judgement to decide reviewed for to be as per which score is most consensus by the KSADS-PL accurate; child and research parent responses clinicians; further integrated into consultation with summary score for each parent and child question when no clear consensus; responses integrated into summary score for each question Comprehensive Introductory Instructions specify evaluation process unstructured introductory clinical including history interview; interview lifetime treatment taking, history, course of psychosocial observation of illness, family participant, past psychiatric records history

Both parent and child's response scored; either informant's response can satisfy criteria

Includes sections for assessment of medical history, psychiatric history, and demographics

Comprehensive evaluation process including psychosocial and developmental history, past records, standardized questionnaires

Depression section

Depression section

As per WASH-UKSADS

Depression and mania sections

Not explicitly stated; assumed to be as per KSADSPL

Interviewers use clinical judgement to decide which score is most accurate; child and parent responses integrated into summary score for each question

Both parent and child's response scored; either informant's response can satisfy criteria Interviewers use clinical judgement to decide which score is most accurate; child and parent responses integrated into summary score for each question

24

Elation vs Irritability

Assessed separately, either can count towards mania criterion A

Assessed separately, either can count towards mania criterion A Symptoms and Substance-related Scoring requires that mood changes pharmacological symptoms; exclusion criteria aetiology has been ruled that occurred during or as a out result of substance use or antidepressant treatment did not count toward BD Study Axelson et al. (2006) characteristics

Assessed separately, either can count towards mania criterion A

Assessed separately, As per WASH-Ueither can count KSADS towards mania criterion A

Not stated

Scoring manual contains criteria that the symptoms are not due substance or medical condition

Participants with primary substance-related mood disorder were excluded

Masi et al. (2007)

Current Study

Sample ascertainment

3 academic medical centers; recruited through internal clinical referrals, community referrals, and advertisements. Brief screening interview with parent, assessing for likely manic symptoms and exclusion criteria (current or lifetime DSM-IV diagnosis of schizophrenia, mental retardation, autism, or severe autistic spectrum disorders or mood disorders due to substance abuse, due to a medical condition, or secondary to use of medications)

Research unit for inpatient and outpatient children and adolescents with mood and anxiety disorders; selected from consecutive presentations with a first putative diagnosis of a mood and/or an anxiety disorder

Specialist public community mental health clinic for young people 5-18 years old with diagnosed or suspected BD; consecutive referrals

BD-NOS criteria

“Distinct period of abnormally elevated, expansive, or irritable mood plus the following: (1) 2 DSM-IV manic symptoms (3 if the mood is irritability only) that were clearly associated with the onset of abnormal mood, (2) a clear change in functioning, (3) mood and symptom duration of a minimum of 4 hours within a 24-hour period for a day to be considered meeting the diagnostic threshold, and (4) a minimum of 4 days (not necessarily consecutive) meeting the mood, symptom, duration, and functional change criteria over the subject’s lifetime, which could be two 2-day episodes, four 1-day episodes, or another variation"

"Defined BD-NOS according to DSM-IV, Leibenluft et al. (2003), and Carlson et al. (2004), including in this subtype patients with rapid and prolonged alternation between manic and depressive symptoms that do not meet duration criteria for manic or depressive symptoms (at least 4 days), or those who present with abnormal mood (sadness, anger, elated mood) and hyperarousal associated with a minimum of three manic symptoms (one less than the DSM-IV B criteria). Multiple episodes with rapid cycling can occur, but no specific minimum duration of the manic episodes is required"

Symptoms meeting WASH-UKSADS mania / hypomania diagnostic criteria with duration of greater than 4 hours and less than 4 continuous days. Those with ‘ultra-rapid’ or ‘ultradian’ cycling were defined as per Leibenluft et al. (2003) as having “(hypo)mania NOS” (ie, BD-NOS)

Mixed episodes definition

Not stated; assumed to be as per KSADS-PL.f Subsyndromal mixed episode definition not stated

Not stated; assumed to be as per KSADS-PL.f Subsyndromal mixed episode definition not stated

Psychosis symptom thresholds

Delusions or hallucinations meeting threshold on KSADS-PLh

Not explicitly stated; assumed to be as per KSADS-PLh

As per WASH-U-KSADS:g Overlapping time periods of mania/hypomania/(hypo)mania NOS and major depressive disorder Delusions or hallucinations meeting threshold on WASHU-KSADSi

25 a

Sections adapted from Galanter et al. (2012). In practice, the interviews have been used to assess both older and younger ages (Galanter et al., 2012). c 1= not present, 2= subthreshold, 3= threshold. d Assessed by the KSADS–Present and Lifetime Version, in the first 84 subjects enrolled in the study, and with the KSADS MRS and KSADS-P mood disorder sections for the remaining subjects. e "Although in practice, the same interviewer may be used" (Galanter et al., 2012). f KSADS-PL defines mixed mood episodes as per DSM-IV. g WASH-U-KSADS defines mixed episodes as overlapping time periods meeting criteria for mania/ hypomania and depression. h Psychosis in KSADS-PL is defined as follows: Delusions: “fixed false beliefs that are above and beyond what would be expected from a child of same age”. Hallucinations: no anchoring statements Excludes illusions, name being called, and bereavement. Required to last “throughout the day for several days or several times a week for several weeks”. Also items pertaining to thought disorder and unusual / disorganised behaviour and affect (Kaufman et al., 1997). i Psychosis in WASH-U-KSADS is defined as “malignant, pathological delusions” or “true psychopathologically significant hallucinations”…that do not only occur hypnogogically or hypnopompically…only if they occur when the subject was fully awake, and neither febrile nor under the influence of alcohol or some drug. The definition explicitly excludes elaborated fantasises, eidetic imagery, and illusions. The threshold for severity is that the subject “generally believes in the reality” of the experience. Frequency / duration criteria for hallucinations are > 5 times during the current episode or more than twice during the last week. The criteria for delusions also include ratings of 6 on the grandiosity, hopelessness, hypochondriasis, and guilt items. Also includes items assessing formal thought disorder, & unusual / disorganised behaviour and affect (Geller et al., 1996). b

26

Table 2 Demographics and clinical characteristics by bipolar subtype. Pairwise comparisons, pf All

BD-I

BD-II

BD-NOS BD-I BD-I BD-II

participants (N=88)

vs (n=24)

(n=13)

(n=51)

vs

vs

BD-II BD-

BD-

NOS

NOS

Demographic data Age, M (SD)

14.8 (2.47) 16.1 (1.23) 15.0 (2.27 14.1 (2.71) 1

Male sex, n (%)

25 (28.4)

3

9)

5 (20.8)

3 (23.1

3

.08 g

<.00

.25g

1h

17 (33.3)

1.0

.42

.74

8.80 (2.35) 10.0 (1.32) 9.46 (2.30 8.02 (2.45)

.30

<.00

.06g

.92

.51

.69

.75

.99

.71

17 (33.3)

.73

.80

1.0

39 (76.5)

.45

.77

.72

8 (15.7)

1.0

.75

1.0

0 (0.0)

.01

<.00

) Education; years, M (SD)

8 AUSEI06 father, M (SD)

40.9 (22.23 43.2 (24.69 42.4 (24.4 39.5 (20.78 9)

AUSEI06 mother, M (SD)

2)

2 0)

30 (34.1)

8)

0 8)

9 (37.5)

4 (30.8

1h

8)

40.4 (20.83 40.1 (20.44 42.5 (22.3 40.1 (21.01 8)

Resides with both parents, n (%)

g

)

0)

) Attending education, n (%)

67 (76.1)

17 (70.8)

11 (84.6 )

Currently employed, n (%)

15 (17.0)

5 (20.8)

2 (15.4 )

Clinical presentation Index episode polarity, n (%) Manic

9 (10.2)

9 (37.5)

0 (0.0)

5

1

27

Hypomanic

6 (6.8)

0 (0.0)

4 (30.8

2 (3.9)

) Mixed

73 (83.0)

15 (62.5)

9 (69.2

.01

1.00

.013

.73

<.00

.013

1 49 (96.1)

) Depressed Elation

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

81 (92.0)

24 (100.0

12 (92.3

45 (88.2)

.35

.17

1.00

51 (100)

.64

.030

.007

24 (47.1)

.27

.027

.76

2.80 (3.14) 2.39 (1.65) 1.43 (0.54 3.50 (4.15)

.04

.29g

.20g

.62

.52

.96

1.0

.44

.52

38 (74.5)

.28

.78

.49

18 (35.3)

.72

.62

1.00

2.91 (2.29) 2.60 (1.71) 2.25 (0.96 3.22 (2.76)

.80

.51

.46

18 (35.3)

.29

.45

.52

43 (84.3)

.04

.26

.24

53.3 (9.81) 53.2 (10.83 56.9 (12.0 52.5 (8.66)

.29

.75

.15

) Irritability

1

82 (93.2)

21 (87.5)

) 10 (76.9 )

Psychiatric admission, n (%)

49 (55.7)

18 (75.0)

7 (53.8 )

Number of admissions, M (SD)a

h

) ALOS, weeks, M (SD)a

1.98 (2.35) 2.29 (2.84) 1.76 (3.23 1.81 (1.65) )

Self-harm / suicidal gestures, n (%)

61 (69.3)

18 (75.0)

10 (76.9

33 (64.7)

) Suicidal ideation, n (%)

65 (73.9)

19 (79.2)

8 (61.5

0

) Suicide attempt, n (%)

32 (36.4)

10 (41.7)

4 (30.8 )

Number of suicide attempts, M (SD)b

) Psychosis, n (%)

32 (36.4)

11 (45.8)

3 (23.1 )

Psychopharmacological treatment,

75 (85.2)

23 (95.8)

lifetime, n (%) CGAS, M (SD)

)

9 Course of illness

9 (69.2

9)

2 0)

3

28

Age of 1st episode, M (SD)

12.3 (3.04) 13.5 (2.45) 13.7 (2.01 11.4 (3.21) 2

Multiple episodes, n (%)

22 (25.0)

0 11 (45.8)

5) 1 (7.7)

0

.76 g

10 (19.6)

.003 h

.003 h

.02

.027

.44

.02

.027

.44

.10

.74

.12

.01

.81g

.029

7 Previous episode depressed

22 (25.0)

11(45.8)

1 (7.7)

10 (19.6) 7

Previous episode (hypo)manic

0 (0.0)

0(0.0)

0 (0.0)

0 (0.0)

Previous episode mixed

0 (0.0)

0(0.0)

0 (0.0)

0 (0.0)

DUB, weeks, M (SD)c

97.0 (85.53 108.(107.0 59.9 (41.4 101. (81.33 1)

Illness duration, weeks, M (SD)d

(SD)e

2 2)

18 )

121.9(102.3 127.(105.1 65.1 (38.9 133.(108.5 25)

Duration of index episode, weeks, M

251)

335)

5 9)

843)

4h

97.5 (92.97 72.0 (53.12 60.3 (42.2 118.(110.3 1)

0)

g,i

.71

.039i

.040i

8 6)

983)

5 (38.5

25 (49.0)

.31

.47

.55

7 (13.7)

.38

.026

1.00

6 (11.8)

1.0

.505

.411

Comorbid disorders, n (%) Anxiety disorder

44 (50.0)

14 (58.3)

) Eating disorder

15 (17.0)

5 (20.8)

3 (23.1 )

Panic disorder

13 (14.8)

4 (16.7)

3 (23.1 )

Specific phobia Social phobia

0

13 (14.8)

6 (25.0)

0 (0.0)

7 (13.7)

.07

.32

.33

8 (9.1)

3 (12.5)

1 (7.7)

4 (7.8)

1.0

.67

1.00

1.0

.38

.27

0 OCD

8 (9.1)

3 (12.5)

2 (15.4

3 (5.9)

)

0

GAD

11 (12.5)

4 (16.7)

0 (0.0)

7 (13.7)

.28

.74

.33

PTSD

13 (14.8)

5 (20.8)

0 (0.0)

8 (15.7)

.14

.75

.19

0.89 (1.15) 1.33 (1.49) 0.54 (0.88 0.76 (0.97)

.04

.10h

.45g

Number of anxiety disorders, M (SD)

)

9h

29

Externalising disorder

39 (44.3)

11 (45.8)

4 (30.8

24 (47.1)

.49

1.00

.36

19 (37.3)

.41

.11

.76

) ADHD

27 (30.7)

4 (16.7)

4 (30.8 )

CD

11 (12.5)

4 (16.7)

1 (7.7)

6 (11.8)

.64

.72

1.00

ODD

15 (17.0)

6 (25.0)

0 (0.0)

9 (17.6)

.07

.54

.19

0.60 (0.78) 0.58 (0.78) 0.38 (0.65 0.67 (0.82)

.46

.67

.25

1.0

.14

.59

46 (90.2)

.64

0.71

.34

2.18 (1.60) 2.67 (1.97) 1.46 (1.33 2.14 (1.41)

.02

.18

.17

Number of externalising disorders, M (SD) Substance use disorder

) 11 (12.5)

5 (20.8)

2 (15.4

4 (7.8)

) Any psychiatric comorbidity

77 (87.5)

21 (87.5)

10 (76.9

0

) Number of comorbid conditions, M (SD)

)

9

Abbreviations: BD-I = bipolar I disorder; BD-II = bipolar II disorder; BD-NOS = bipolar disorder not otherwise specified; AUSEI06 = Australian Socioeconomic Index 2006; CGAS = Children’s Global Assessment Scale; DUB = duration of untreated bipolar; OCD = obsessive-compulsive disorder; GAD = generalised anxiety disorder; PTSD = posttraumatic stress disorder; ADHD = attention deficit-hyperactivity disorder; ODD = oppositional-defiant disorder; CD = conduct disorder; ALOS = average length of hospital stay. a Based on subset of participants with psychiatric admission(s). b Based on subset of participants with suicide attempt(s). c Weeks from onset of first episode to first mental health contact. d Weeks from onset of first episode to initial assessment. e Time period of index episode prior to study enrolment. f p values uncorrected for multiple comparisons; pairwise comparisons calculated using Fisher exact test for categorical variables or ANOVA LSD method for quantitative variables, except where specified. g t test using unpooled variances, due to unequal variances across bipolar subgroups. h t test, correction applied for unequal variances. i Post-hoc comparison, nonsignificant after Bonferroni correction.