Article

Relationship of Parent and Child Informants to Prevalence of Mania Symptoms in Children With a Prepubertal and Early Adolescent Bipolar Disorder Phenotype Rebecca Tillman, M.S. Barbara Geller, M.D. James L. Craney, M.S., M.P.H., J.D. Kristine Bolhofner, B.S. Marlene Williams, R.N. Betsy Zimerman, M.A.

Objective: A controversy regarding pediatric bipolar disorder is whether to use child in addition to parent informants. To investigate this issue, the authors conducted a study comparing separate child and parent interview data for child bipolar disorder. Method: Responses on the Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia from 93 child and 93 parent informants were compared by using kappa statistics. Research nurses, blind to subject information, separately interviewed parents about their children and children about themselves. Different nurses were used for the parent and child in each family to avoid bias from the same research nurse interviewing a child after interviewing that child’s parent. Mania was defined by DSM-IV criteria, with at least one of the two cardinal symptoms of mania (elated mood and/or grandiosity), to avoid diag-

nosing mania by symptoms that overlapped with those for attention deficit hyperactivity disorder (ADHD). Results: Parent-child concordance was poor to fair for all cardinal and noncardinal mania symptoms. Kappas were not significantly different by age within the 7– 14-year-old age range. Conclusions: Symptoms endorsed by just the child included substantial proportions of bipolar symptoms that have been shown to best differentiate mania from ADHD (i.e., elation, grandiosity, flight of ideas, racing thoughts, decreased need for sleep). These findings support the need for child informants in research on prepubertal and early adolescent bipolar disorder in children ages 7–14. Differences in mania symptom profiles between investigative groups may be, in part, due to whether child informants were assessed. (Am J Psychiatry 2004; 161:1278–1284)

T

o our knowledge, no other study has compared parent to child informants in systematically assessed subjects with a prepubertal and early adolescent bipolar disorder phenotype. Child informants, however, may be especially important for the assessment and diagnosis of prepubertal and early adolescent mania because of differences in symptom profiles reported by major investigative groups (1–3). Specifically, studies that examined child informants found a symptom distribution that resembled that described for severely ill adults with bipolar I disorder (4). In contrast to this picture of adult-type bipolar I disorder, other investigations did not include child interviews and found a profile characterized primarily by irritable mood and conduct-disordered behavior (3). Because some groups describing nonadult-type symptom profiles did not interview children under the age of 12 (3), there was a question of whether differences between studies might be due, in part, to whether child informants were used. It was hypothesized that child informants would be important in diagnosing prepubertal and early adolescent

1278

http://ajp.psychiatryonline.org

mania because of reports in the literature of low agreement beyond chance between parent and child ratings (obtained by direct interviews of the informants) for other internalizing disorders (e.g., depression and anxiety) (5–12). In child psychiatry, internalizing disorders refer to those primarily characterized by nonobservable symptoms (e.g., racing thoughts) as opposed to externalizing disorders that include observable pathology (e.g., hyperactivity). In 1995, when the Phenomenology and Course of Pediatric Bipolar Disorder study began, it was the first investigation funded by the National Institute of Mental Health on the phenomenology and longitudinal course of children with prepubertal and early adolescent mania. To establish credibility for the existence of mania in children amid contention in the field (1), our group elected to study a phenotype with conservative diagnostic criteria. To address ambiguities in the field of pediatric bipolar disorders (1), including how to differentiate prepubertal mania from attention deficit hyperactivity disorder (ADHD), several research strategies were developed (4). First, subjects needed to fit DSM-IV criteria for mania with at least Am J Psychiatry 161:7, July 2004

TILLMAN, GELLER, CRANEY, ET AL.

one of the cardinal mania criteria (i.e., elated mood and/ or grandiosity). This schema followed the DSM-IV pattern of needing a cardinal symptom of depression (i.e., sad mood or anhedonia) to fit the diagnosis of major depressive disorder. This cardinal symptom approach obviated the problem of diagnosing pediatric bipolar disorder by criteria that overlapped with those for ADHD (e.g., hyperactivity, distractibility) (13, 14). Elated mood and grandiosity were selected because they are highly specific to mania at all ages (2, 15). As etiopathogenetic studies become available (16), both the cardinal symptom definition of DSM-IV major depressive disorder and the cardinal symptom approach to prepubertal and early adolescent mania may need modification. Another aspect of using the cardinal symptom approach addressed the issue of irritability. Several authors have stressed that child mania is characterized by irritable rather than elated or concurrent elated and irritable moods (3). Indeed, irritability across the age span is a common symptom of mania, as has been reported in numerous studies of bipolar I disorder in adults and in work with bipolar children (2, 15). For example, 87.1% of the subjects with a prepubertal and early adolescent bipolar disorder phenotype had both elated mood and irritability (2). Irritability, however, although very sensitive (i.e., it is present in most cases of mania), is highly nonspecific because it also occurs in multiple other child diagnoses. For example, Aman et al. (17) reported a controlled study of risperidone for irritability and aggression in subjects with a low IQ. Another randomized clinical trial examined risperidone for irritability and aggression in autism (18). More recently, Kim-Cohen et al. (19) investigated whether young adults with multiple psychiatric diagnoses (e.g., eating, substance use, schizophreniform disorders) had a childhood diagnosis. These investigators found that 20%– 60% of adults with a psychiatric diagnosis had a childhood disorder characterized by irritability and aggression (e.g., oppositional defiant disorder, conduct disorder). Based on the nonspecificity of irritability, subjects with only irritable mood (i.e., without euphoria) needed to have grandiosity in order to fit the study phenotype. Of note, although subjects only needed to have one of the two cardinal symptoms of mania, 77.4% had both elated mood and grandiosity. Moreover, 87.1% had concurrent irritability and elation (2). A second strategy was to construct an assessment instrument, the Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS), which included mania items specific to the developmental stage of prepubertal children (20, 21). This instrument was necessary because children are developmentally incapable of many manifestations of mania observed in adults with bipolar disorder (e.g., children will not have had four marriages or have “maxed out” their credit cards) (22). The WASH-U-KSADS has demonstrated reliability and 6-month stability and is used in the Am J Psychiatry 161:7, July 2004

majority of federally funded grants on pediatric bipolar disorder (1, 21, 23). By using these research strategies, a prepubertal and early adolescent bipolar disorder phenotype has been validated by both longitudinal (4, 24, 25) and family (Geller, unpublished data) studies. Subjects with this phenotype were suitable for study of the relationship of child informants to symptom profiles.

Method Inclusion and Exclusion Criteria Details of the study inclusion and exclusion criteria have been previously reported (2). Briefly, inclusion criteria were boys and girls ages 7–16 years who were in good physical health with a current DSM-IV diagnosis of mania (manic or mixed phase) for at least 2 weeks. A Children’s Global Assessment Scale (26, 27) score ≤60 was needed to establish significant clinical impairment. Exclusion criteria were having an IQ