University of Groningen. Children of bipolar parents Wals, Marjolein

Transcription

1 University of Groningen Children of bipolar parents Wals, Marjolein IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2004 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Wals, M. (2004). Children of bipolar parents: prevalence of psychopathology and antecedents of mood disorders Groningen: s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2

3 Children of Bipolar Parents Prevalence of psychopathology and antecedents of mood disorders Marjolein Wals

4 The study reported in this thesis was supported by a grant from NWO Nederlandse Organisatie voor Wetenschappelijk Onderzoek, Dutch Organisation for Scientific Research and by the Stanley Medical Research Institute, United States. The Stanley Medical Research Institute 2004 Marjolein Wals Printed by Optima Grafische Communicatie Rotterdam, the Netherlands Cover design by Francien Verhulst Technical support by Thomas Veltman ISBN

5 RIJKSUNIVERSITEIT GRONINGEN Children of Bipolar Parents Prevalence of psychopathology and antecedents of mood disorders Proefschrift ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op woensdag 8 december 2004 om uur door Marjolein Wals geboren op 1 maart 1969 te Leiden

6 Promotores: Prof. Dr. J. Ormel Prof. Dr. W.A. Nolen Beoordelingscommissie: Prof. Dr. R. B. Minderaa Prof. Dr. J. van Os Prof. Dr. F. Verheij

7 Contents Contents Chapter 1: General Introduction... 1 Overall aim of KBO-project...1 Global aims of this thesis...2 Definition and epidemiology of bipolar disorder...2 Definition and epidemiology of unipolar disorder...3 Background and theoretical framework...4 Main research questions of this thesis...8 Sample...8 Measures...10 Chapter 2: Prevalence of psychopathology in children of bipolar parents.. 13 (Journal of the American Academy of Child and Adolescent Psychiatry, 2001, 40, ) Abstract...14 Introduction...14 Methods...17 Results...19 Discussion...23 Chapter 3: Multiple dimensions of familial psychopathology affect risk of mood disorder in children of bipolar parents (American Journal of Medical Genetics, part B, Neuropsychiatric Genetics, 2004, 127B (1), 35-41) Abstract...28 Introduction...28 Methods...29 Results...32 Discussion...36 Appendix...40 Chapter 4: The impact of birth weight and genetic liability on psychopathology in children of bipolar parents (Journal of the American Academy of Child and Adolescent Psychiatry, 2003, 42, ) Abstract...42 Introduction...42 Methods...44 Results...47 Discussion...49 Appendix...51

8 Contents Chapter 5: Prediction of change in level of problem behavior among children of bipolar parents (Submitted for publication) Abstract...54 Introduction...54 Methods...56 Results...59 Discussion...62 Acknowledgements...63 Chapter 6: Stressful life events and onset of mood disorders in children of bipolar parents during 14-month follow-up (Submitted for publication) Abstract...66 Introduction...66 Methods...68 Results...73 Discussion...76 Chapter 7: General Discussion and Conclusions Recapitulation of Research Questions...79 Strengths and limitations of the study...87 Research implications...89 Final conclusion...91 References Summary Nederlandse samenvatting (Summary in Dutch) Dankwoord Curriculum Vitae English Curriculum Vitae Nederlands

9 General Introduction Chapter 1: General Introduction He felt good. Really good. Better than good, actually... nearly invincible. He felt like he had limitless energy and could go without sleep for days. He was full of ideas and plans and would be frustrated by the inability of others to keep up with him. He could barely express one thought before barreling onto another at a dizzying pace that would leave his listeners bewildered. But then there were the times when he felt sad. But it was more than that. The burden of life felt too heavy, too daunting to face. Nothing cheered him up. All the things he loved no longer gave him pleasure. At times his feelings of despair were so great he would even consider suicide. He often wondered how he could feel so unbelievably good at some times and so horribly bad at others. Is it possible for one person to experience such emotional extremes? Cited from the website of the American Psychiatric Association (1999) The person described above suffers from a psychiatric illness called bipolar disorder. Bipolar disorder is characterized by severe mood swings. A person with bipolar disorder suffers from episodes of depression and episodes of (hypo)mania (mood elevation, including exaggerated euphoria, irritability, or both), which alternate with periods of normal mood. Overall aim of KBO-project By the end of 1997 the Kinderen van Bipolaire Ouders (KBO)-project (Dutch for Children of Bipolar Parents ) was launched. This project was initiated because most patients with bipolar disorder are diagnosed as such many years after the onset of their illness (Goodwin and Jamison, 1990). As a consequence, adequate treatments may be postponed or inappropriate treatments may be given. It is therefore important to accurately identify early signs of bipolar disorder as well as the risk factors that increase the probability that bipolar disorder emerges. The overall aim of this project was therefore to study the early development of bipolar disorder and factors influencing this development. The most straightforward way to study the development of bipolar disorder would be to follow prospectively a large cohort of children from the general population and to identify new cases with bipolar disorder. However, because such a study would require a very large number of individuals, it was decided to study children from a parent with bipolar disorder. Children of bipolar parents have an increased risk to develop psychopathology, including bipolar disorder, compared to children from the general population. The lifetime risk for bipolar disorder in relatives of a bipolar proband was found to be 5-10% for first-degree relatives compared with % for unrelated individuals (Craddock and Jones, 1999). Therefore, the study of children of bipolar parents is an economic way to elucidate factors influencing the emergence of bipolar disorder. 1

10 Chapter 1 Global aims of this thesis The focus of this thesis is on the current functioning and development of children of bipolar parents for two main reasons. First, children of bipolar parents are at genetic risk for developing psychopathology, including bipolar disorder themselves (Alda, 1997; Gershon et al., 1987; McGuffin and Katz, 1989). Second, children of bipolar parents are subject to environmental stressors associated with parental psychopathology and its consequences. These multiple risks may be responsible for the emergence of multiple types of psychopathology that subsequently are subject to developmental change. For example, it is possible that the first signs of bipolar disorder are aspecific problems, such as oppositional behavior, later followed by the development of a unipolar disorder that subsequently evolves into a bipolar disorder. Little is known about the effects of environmental influences on the functioning of children of bipolar parents. Therefore, the global aims of this thesis are: 1. To determine the lifetime and current prevalence of psychopathology including mood disorders among offspring of bipolar parents. 2. To test the effects of factors influencing the presence and onset of psychopathology, including mood disorders, among offspring of bipolar parents. To achieve these aims we studied a sample of 140 children of 86 bipolar parents aged 12 to 21 years. This thesis is based on two measurements with an interval of 14 months. Definition and epidemiology of bipolar disorder Bipolar disorder is characterized by episodes of extreme moods, mania or hypomania and depression, which alternate with periods of normal mood. In this study, we used DSM-IV-criteria for defining bipolar disorder (American Psychiatric Association, 1994). Mania is the term to describe episodes of severe mood elevations. Hypomania is the term used to describe episodes with mild elevations of mood. Bipolar disorder with hypomanic and severe depressed episodes is called Bipolar II disorder, whereas manic episodes alternating with more or less severe depressed episodes indicate Bipolar I disorder. The spectrum of severity ranges from mild cyclothymia (only hypomanic episodes and minor depression, not reaching the criteria for more severe major depressed episodes) to severe bipolar disorder with episodes that are accompanied by psychotic features. The lifetime prevalence of bipolar disorder in The Netherlands is 1.8 to 1.9% (Bijl et al., 1998; Ten Have et al., 2002) which is comparable to the prevalence of bipolar disorder in other western countries (Regier et al., 1993; Kessler et al., 1994). Both sexes are equally affected. The age of onset of bipolar disorder is typically between ages 15 and 25 years. Many patients experience their first episode of bipolar disorder before the age of 20 years (Goodwin and Jamison, 1990). 2

11 General Introduction In a survey among National Depressive and Manic Depressive Association (NDMDA) members it was found that 88% of patients with bipolar disorder have ever been treated in a psychiatric hospital at least once and that 66% have been hospitalized twice or more (Lish et al., 1994). Days lost from work, difficulty regaining employment and loss of employment all contribute to the disability associated with bipolar disorder. In The Netherlands, individuals diagnosed with bipolar disorder reported more bed rest days and absenteeism days than healthy people or individuals suffering from unipolar mood disorders or anxiety disorders (Ten Have et al., 2002). In the NDMDA survey it was found that 37% of patients with bipolar disorder were unemployed at the time of assessment (Lish et al., 1994). In a study in New Zealand patients who were hospitalized for bipolar disorder were employed in only 34% of cases, compared to 75% in the general population (McPherson et al., 1992). In a U.S. study it was found that only 43% of patients with bipolar disorder were employed six months after discharge from psychiatric hospitalization (Dion et al., 1988), and only 21% were functioning at their expected level of employment. Bipolar disorder can place enormous stress on intimate relationships. The New Zealand study reported that only 20% of patients were married prior to their hospitalization compared to 55% in the general population (McPherson et al., 1992). The lifetime risk of death by suicide in bipolar disorder was estimated to be 18.9% in a meta-analysis of 29 studies (Goodwin and Jamison, 1990). Mortality rates for untreated bipolar disorder were estimated to be comparable to most types of heart disease and many types of cancer. In The Netherlands, individuals diagnosed with bipolar disorder reported more suicidal ideation and prior suicide attempts than healthy people or individuals with unipolar mood disorders or anxiety disorders (Ten Have et al., 2002). Episodes of mania or depression can last from a few weeks to several months and are frequently severe enough to affect day-to-day functioning at work and at home. For most patients, the illness is recurring. Men tend to have more episodes of mania; women show more episodes of depression. Bipolar disorder can be a lifelong condition for which no curative treatment is available. However, adequate treatment reduces the number and severity of episodes and thereby reduces the level of handicap in the majority of patients. Definition and epidemiology of unipolar disorder A person who suffers from a major depressive disorder has either a depressed mood or a loss of interest or pleasure in daily activities consistently for at least a 2-week period. The mood represents a change from the person's normal mood. In addition, other symptoms may be present including, significant weight loss or gain or decrease or increase in appetite, insomnia or hypersomnia, psychomotor agitation or retardation, fatigue or loss of energy, feelings of worthlessness or excessive guilt, diminished ability to think or concentrate, or recurrent thoughts of death. Finally, social, occupational, educational or other important areas of 3

12 Chapter 1 functioning also are negatively impaired by the change in mood. A depressed mood caused by substances (such as drugs, alcohol, and medications) is not considered a major depressive disorder, nor is one that is caused by a general medical condition. Major depressive disorder cannot be diagnosed if a person has a history of manic, hypomanic, or mixed episodes (i.e. the person has a bipolar disorder) or if the depressed mood is better accounted for by schizoaffective disorder or is superimposed on schizophrenia, a delusional disorder or another primary psychotic disorder. The symptoms are not better accounted for by bereavement, i.e., after the loss of a loved one, the symptoms persist for longer than 2 months or are characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation (American Psychiatric Association, 1994). The lifetime prevalence of major depression in The Netherlands is 14 to 15.4% (Alonso et al., 2004; Bijl et al., 1998) which is comparable to the prevalence in the US of 11.5% (Jonas et al., 2003). There are clear gender differences in the overall prevalence of mood disorders. Mood disorders are almost twice as common amongst women as amongst men (prevalence of major depression for females: 20.1% and for males: 10.9%; Bijl et al., 1998). Background and theoretical framework Mood and non-mood disorders in children of bipolar parents Several studies have reported on the risk of children of parents with bipolar disorder to develop both mood and non-mood disorders (e.g. DelBello and Geller, 2001; Lapalme et al., 1997; Chang et al., 2001). The reported risk of lifetime mood disorders is approximately 4 times greater for children of bipolar parents than for children of parents without psychiatric disorder or without major psychiatric disorder. Lapalme and colleagues (1997) summarized the prevalences of psychiatric disorders in children of bipolar parents in a meta-analysis of 17 studies. The authors reported that 52% of the children of parents suffering from bipolar disorder met criteria for some psychiatric disorder, compared with 29% of the children in comparison groups of parents without psychiatric disorder or without major psychiatric disorder. They found that 26.5% of the offspring of bipolar parents met criteria for a mood disorder versus 8.3% of the children of parents from the comparison groups. In a more recent review of 17 studies, DelBello and Geller (2001) reported that the lifetime prevalences of mood disorders in child and adolescent offspring of bipolar parents ranged from 5 to 67% versus 0 to 38% compared to offspring of healthy volunteers. In the present thesis, we reviewed all published studies that included diagnostic assessments resulting in psychiatric diagnoses, including the 17 studies of the meta-analysis of Lapalme et al. (1997, see table 1 in chapter 2) and concluded that the prevalences of current psychiatric disorders among children of bipolar parents varied greatly, ranging from 6% (LaRoche et al., 1981) to 86% (Zahn-Waxler et al., 1988), whereas the prevalences of lifetime diagnoses ranged from 43% (Klein et al., 1985) to 92% (Hammen et al., 1987). We also concluded 4

13 General Introduction that there are a number of serious methodological shortcomings in existing studies, hampering the comparability across studies and the interpretation of findings. Firstly, for many studies the authors did not specify whether prevalence figures reflected current or lifetime diagnoses. Evidently, this will have led to large differences in the reported prevalences of psychopathology. Secondly, it is often not clear if results reflect prevalences of disorders or prevalences of disordered subjects. Because many subjects fulfill criteria of multiple disorders, prevalences based on numbers of disorders are higher than those based on numbers of individuals with one or more disorders. Thirdly, most studies used small and highly selected samples of convenience as comparison groups (e.g. Zahn-Waxler et al. 1988; Carlson and Weintraub, 1993). Consequently, they do not form an adequate basis for contrasting the prevalences in bipolar offspring. Finally, there is large variation in sample composition. Most studies used selected clinical samples of adults with bipolar disorder that consisted of inpatients (e.g. Carlson and Weintraub, 1993) or outpatients (e.g. La Roche et al., 1987) or a mix (e.g. Anderson and Hammen, 1993) and of bipolar I (e.g. Duffy et al., 1998), or a mix of bipolar I and bipolar II patients (e.g. Decina et al., 1983). The majority of studies used clinical interviews, mostly the Schedule for Affective Disorders and Schizophrenia for School-Age Children (K-SADS; Kaufman et al., 1997) to assess psychopathology in the bipolar offspring. Because there are no data with the K-SADS for representative normative samples, it is difficult to interpret the findings. In conclusion, the methodological variations across studies and the lack of reliable reference data make it very difficult to draw firm conclusions about the prevalence of psychopathology in offspring of bipolar parents. In the present study, we therefore aimed to compare the prevalence of psychopathology in bipolar offspring with that in normative samples partly using the same standardized assessment procedures. An additional rationale for studying the prevalence of psychopathology in bipolar offspring is that the prevalence of disorders among children of bipolar parents in The Netherlands need not necessarily be the same to the prevalence among children of bipolar parents across countries. For example, it may be that differences in accessibility to mental health care across countries may differentially affect the risk of developmental problems in children. 5

14 Chapter 1 Determinants of mood disorders in children of bipolar parents An important question is whether we can identify risk factors for the development of mood disorders (Wals and Verhulst, in press). If we are able to gain insight into the etiology of mood disorders this may facilitate the development of preventive interventions, aimed at reducing the risk of developing mood disorders in children of bipolar parents. Genetic factors Given the frequently reported finding that familial clustering of mood disorders is largely the result of shared genes rather than shared family environment (Alda, 1997; Gershon et al., 1987; McGuffin and Katz, 1989), we expected that one important determinant of mood disorder development among bipolar offspring would be familial loading of mood disorders among first- and second-degree relatives. A number of studies reported an elevated prevalence of mood disorders in the families of patients with mood disorders (e.g. Gershon et al., 1982; Rice et al., 1987). To our knowledge, there are no studies that determined to what extent the risk of mood disorders in bipolar offspring varies with the degree of familial loading of mood and related disorders in the extended family. Since a number of studies reported an association between bipolar disorder and substance use disorder (Regier et al., 1990; Strakowski et al., 2000; Strakowski and Debello, 2000; Winokur et al., 1995), and because substance use disorder may be genetically related to mood disorder (Kendler et al., 1993; Winokur et al., 1998), we expected that substance use disorders among first- and second-degree relatives may also be a risk factor for bipolar offspring to develop mood disorders. Environmental factors Birth weight. Among environmental factors increasing the risk of health problems in adult life, low birth weight can be viewed as a marker of prenatal environmental influences. The relevance of prenatal factors that may influence birth weight has been stressed by Barker (1998) who studied the relationship between prenatal influences, especially fetal malnutrition resulting in low birth weight, and increased risk of adult physical diseases including cardiovascular diseases and diabetes. Several studies have assessed the associations between obstetric complications (including birth weight) and unipolar as well as bipolar disorder (e.g. Botting et al., 1997; Browne et al., 2000; Buka and Fan, 1999; Guth et al., 1993; Kinney et al., 1998; Sigurdsson et al., 1999), but the results of these studies are not unequivocal. A possible reason for this is that previous studies did not take account of possible effects of genetic factors. The association between birth weight and psychopathology might be altered by genetic liability. This is important because the association between low birth weight and psychopathology may represent a true causal influence, but it could also represent the effect of a third genetic or environmental variable influencing both characteristics (van Os et al., 2001; Wichers et al., 2002). We therefore decided to assess the impact of birth weight and its interaction with familial loading on the development of mood disorders in bipolar offspring. 6

15 General Introduction Family factors. Another environmental contributor to the risk of psychopathology in offspring of unipolar or bipolar parents is poor family functioning. For instance, Billings and Moos (1983) found variations in family stressors and resources to be strongly related to the probability of disturbance among 133 children of depressed parents as compared to 135 children of nondepressed parents. Chang et al. (2001) found families with a bipolar parent to report differences in their family environment compared to population means. Families with a bipolar parent differed from the average family in having less cohesion and organization, and more conflict. However, Family Environment Scale (FES, Moos and Moos, 1986) scores did not differ significantly for 6- to 18- year-old children with versus children without a DSM Axis I disorder. A drawback of these studies is their cross-sectional nature. It is therefore not possible to determine the temporal relationship between family problems and child psychopathology. We decided to study the impact of family problems on the course of psychopathology during a 14-month follow-up in our sample of bipolar offspring. Stressful life events (SLEs). The degree and nature of stressful life events (SLEs) may be an important environmental factor in the development of mood disorders. In a review, Paykel (2003) reported that episodes of unipolar depression in adults are preceded by SLEs at higher rates than in general population or non-depressed patient samples. Kessling et al. (2004) found that first admissions of adults with mania were often preceded by the occurrence of death by suicide in the family or other major SLEs. Petti et al. (2004) found that the depressed offspring of bipolar parents showed higher levels of dependent negative SLEs preceding onset of their disorder than offspring without affective disorders. Studies examining the impact of SLEs on the development of mood disorders among children and adolescents also reported associations between prior SLEs and onset of depressive disorder (e.g. Goodyer et al., 1985, 1987; Williamson et al., 1998). We decided to study the impact of SLEs on the development of first or recurrent onset of mood disorder episodes (MDEs) during the 14-month follow-up of our sample of bipolar offspring. Transactional developmental model Sameroff (2000) proposed a model of transactional development in which there is a biological organization, the genotype, that regulates the physical outcome of each individual, as well as a social organization that regulates the way human beings fit into their society. The latter organization operates through family and cultural socialization patterns and has been postulated to compose an environtype analogous to the biological genotype. Behavior is a product of transactions between the phenotype (i.e., the child), the environtype (i.e., the source of external experience), and the genotype (i.e., the source of biological organization). Thus, both genetic and environmental factors, alone or in interaction, are expected to influence mental health. As described above, we examined the impact of three environmental variables and one predominantly genetic variable on the development of psychopathology in children of bipolar parents. In addition to unique effects of these variables we thought to determine 7

16 Chapter 1 whether there were interactions between each of the environmental variables with familial loading. Furthermore, we were interested in mediation effects. For instance, birth weight or stressful life events might mediate the association between familial loading of mood disorders and occurrence of mood disorder episodes between first and second measurement. Main research questions of this thesis The main questions of this thesis were: 1. What is the prevalence of psychopathology including mood disorders among offspring of bipolar parents? (chapter 2) 2. Are familial loading of mood disorder and familial loading of substance use disorder associated with lifetime mood disorders in bipolar offspring? (chapter 3) 3. Does birth weight predict mood and non-mood disorder in bipolar offspring independent from, in addition to, or in interaction with familial loading of mood or substance use disorder? (chapter 4) 4. Are familial loading of mood and substance use disorder, birth weight, and family problems predictive of changes in the level of parent reported emotional and behavioral problems across a 14-month interval? (chapter 5) 5. Is there an association between stressful life events and the onset of mood disorders across a 14-month interval? (chapter 6) Sample Recruitment The recruitment of the KBO-cohort is shown in figure 1. By the end of 1997 we sent a survey to all 1961 members of the Dutch Association for Manic- Depressives and Relatives (Nederlandse Vereniging voor Manisch-Depressieven en Betrokkenen-VMDB). This survey explained the aims of the study and included questions concerning the illness of the members, family composition and age of the offspring. Of the 712 (36%) who returned the survey, 110 reported that they had one or more children in the age range of 12 to 21 years. Eventually, 62 out of the eligible 110 parents, which was 56% of the bipolar parents with children in the age range of 12 to 21 years, agreed to participate with a total of 102 children. In addition, we contacted 9 psychiatric hospitals with an assigned outpatient clinic for patients with bipolar disorder in different regions widely spread over the country. These 9 psychiatric hospitals identified 91 bipolar patients with children aged 12 to 21 years. Eventually, 24 (26%) of these patients, with a total of 38 children, agreed to participate. All patients were outpatients at the moment of recruitment. 8

17 General Introduction Dutch Patient Association Psychiatric hospitals 110 eligible families 92 eligible families 62 families participated (56%) 24 families participated (26%) 102 offspring aged 12 to offspring aged 12 to families, 140 offspring Figure 1 Recruitment of the sample Subjects Subjects were enrolled in the study between November 97 and April 99. A sample of 140 children, aged 12 to 21 years, was included in the study (T1). At the second measurement (T2), 14 months after T1, 132 (94%) offspring and at third measurement (T3) 129 (92%) offspring agreed to participate again (see figure 2). This thesis is based on the first two measurements. The characteristics of the sample are shown in table 1. There were 52 bipolar mothers and 34 bipolar fathers; 64 had a bipolar I disorder and 22 a bipolar II disorder. Twenty-one (24%) of the parents were divorced. The number of participating boys and girls was almost equal. Boys and girls did not differ in mean age or IQ. T n = months T months n =132 T n = 129 Figure 2 Assessment scheme 9

18 Chapter 1 Table 1: Demographic characteristics of KBO-cohort n (%) p Bipolar fathers 34 (40) Bipolar mothers 52 (60).060 Bipolar I parents 64 (74) Bipolar II parents 22 (26) Parents Married 65 (76) Parents Divorced 21 (24) Mean SD p Nr. of male offspring 72 (51) Nr. of female offspring 68 (49).800 Age male offspring Age female offspring IQ male offspring IQ female offspring IQ children of total sample offspring SES offspring Measures The instruments and sources of information are shown in table 2. We used three sources of information to assess psychopathology in bipolar offspring: the offspring themselves, parents and teachers of the offspring. Bipolar offspring and their parents were interviewed by one of the researchers or an intensively trained psychologist using the Schedule for Affective Disorders and Schizophrenia for School Age Children Kiddie-SADS-Present and Lifetime Version (K-SADS). The K-SADS is designed to assess current and past DSM-IV diagnoses in children and adolescents, by interviewing the parent(s) and child separately. The Child Behavior Checklist (CBCL; Achenbach, 1991a,b) is a questionnaire to be completed by parents for assessing behavioral and emotional problems for 4 to 18-year-old children. The Youth Self-Report (YSR; Achenbach, 1991a,d) has the same format as the CBCL but has to be completed by 11 to 18- year-olds themselves. The Teacher's Report Form (TRF; Achenbach, 1991a,c) can be completed by teachers of 5 to 18-year-olds. Good reliability and validity of the CBCL, YSR, and TRF have been replicated for the Dutch translations (Verhulst et al., 1996; 1997b,c). 10

19 General Introduction The Young Adult Self-Report (YASR) and Young Adult Behavior Checklist (YABCL) (Achenbach, 1997a,b) are upward extensions of the YSR and CBCL for ages 18 years and older. The YASR is filled out by young adults and the YABCL by parents. The YASR and YABCL can be scored on similar scales as those of the CBCL. In addition, the offspring were interviewed with the Life Event and Difficulties Schedule (LEDS; Brown and Harris, 1978) which is a semi-structured interview for assessing stressful life events (SLEs) and long-term difficulties in adults. With the LEDS it is possible to rate a variety of dimensions of events and difficulties including loss, danger, humiliation, entrapment and required adjustment. It guarantees a relatively objective measurement of the meaning of events and difficulties, and its reliability and validity are satisfactory (Brown and Harris, 1989). Monck and Dobbs (1985) adapted the Bedford College LEDS methodology for use with adolescents (the Kiddie LEDS; K-LEDS). The parents were interviewed with the International Diagnostic Checklist (IDCL, Hiller, 1993) in order to check their diagnosis of bipolar disorder. In addition, both parents were interviewed with the Family History-Research Diagnostic Criteria (FH-RDC; Andreasen et al., 1977). In this interview, the RDC criteria (which are comparable to DSM-IV) are screened concerning all first- and second-degree relatives of the bipolar offspring. We only applied sections of the FH-RDC yielding the following diagnoses: unipolar disorder, bipolar disorder, and substance use disorder. Finally, parents gave us the birth weight of their offspring, which was recorded in birth records that they obtained just after birth. Table 2: Sources of information and instruments used in the KBO-study Offspring Parents Teachers Youth Self-Report (YSR)/ Young Adult Self-Report (YASR) K-SADS-PL Life Event and Difficulties Schedule for adolescents (K-LEDS) Child Behavior Checklist (CBCL)/ Young Adult Behavior Checklist (YABCL) International Diagnostic Checklist (IDCL) Schedule for Affective Disorders and Schizophrenia for School Age Children Kiddie-SADS-Present and Lifetime Version (K-SADS-PL) Family History-Research Diagnostic Criteria (FH-RDC) Birth weight record Teacher's Report Form (TRF) 11

20

21 Prevalence of Psychopathology Chapter 2: Prevalence of psychopathology in children of bipolar parents Marjolein Wals Manon H. J. Hillegers Catrien G. Reichart Johan Ormel Willem A. Nolen Frank C. Verhulst Journal of the American Academy of Child and Adolescent Psychiatry, 2001, 40,

22 Chapter 2 Abstract Objective: To determine psychopathology in adolescent children of a bipolar parent living in The Netherlands, using multiple sources of information (selfreport, parent, and teacher reports). Methods: Problem behavior in 140 offspring (aged years) of 86 bipolar parents was assessed with the Child Behavior Checklist (CBCL), the Teacher's Report Form (TRF), and the Youth Self-Report (YSR) between 1997 and All adolescents, bipolar parents and their available spouses were interviewed with the Schedule for Affective Disorders and Schizophrenia for School Age Children Kiddie-SADS-Present and Lifetime Version (K-SADS-PL). Results: Higher problem scores were found for 8 of the 11 CBCL scales for girls, 4 of the 11 CBCL scales for boys, compared to a Dutch normative sample and 1 YASR scale for girls compared to an American normative sample. Lower problem scores were found on 4 YSR and 4 YASR scales for boys, 1 TRF scale for girls and 1 TRF scale for boys. The prevalence of current DSM-IV diagnoses in the offspring was 29%, and of lifetime DSM-IV diagnoses 44%. Conclusions: The prevalence of problem behavior and DSM-IV diagnoses found in our sample did not support the notion that the level of psychopathology in children aged years of bipolar parents is highly elevated. Limitations: This study, similar to prior studies, suffers from lack on information on the representativeness of the sample, and a rather low response rate. Keywords: bipolar, psychopathology, adolescents, K-SADS, CBCL Introduction Several studies have reported on an increased prevalence of psychiatric disorders in children of parents with bipolar disorder. Table 1 gives the prevalences of children of bipolar parents with a current or lifetime psychiatric disorder for all published studies that included diagnostic assessments resulting in psychiatric diagnoses. All studies used clinical interviews to assess psychopathology, mostly the Schedule for Affective Disorders and Schizophrenia for School-Age Children (K-SADS). Whereas the majority of studies included in table 1 reported, in addition to DSM-IV diagnoses, parent or teacher rating scale scores of psychopathology in children of bipolar parents, only four compared these data with those obtained in normative or comparison samples. One study, not listed in table 1, only reported on parent and teacher ratings of problem behavior and not on psychiatric diagnoses (Anderson and Hammen, 1993). To our knowledge, there are no studies that have used adolescents' self-reported problems to assess the level and type of psychopathology. Two studies using rating scales did not report elevated levels of psychopathology in children of bipolar parents (Anderson and Hammen, 1993; Kashani et al., 1985), whereas three other studies using rating scales did find elevated levels of psychopathology in children of bipolar parents (Carlson and Weintraub, 1993; Radke-Yarrow, 1992; Zahn-Waxler et al., 1988). 14

23 Prevalence of Psychopathology Prevalence rates reported in existing studies varied greatly. However, there are a number of methodological issues that hamper the comparability across studies and the interpretation of findings. First, there is large variation in sample composition. Most studies used selected clinical samples of adults with bipolar disorder that consisted of either inpatients, outpatients or a mix, with varying degrees of severity of the condition, and of bipolar I, or a mix of bipolar I and II patients (see table 1). Also, the comparison samples varied greatly and comprised children of parents from nonrepresentative nonclinical samples of convenience, or of children of parents with no major mental disorder. Second, the diagnostic procedures varied greatly. The level of standardization of the interviews used in the various studies differed as well as the diagnostic criteria used and the number of diagnoses included. Most studies used clinical interviews of parents and children separately and combined the often conflicting information in idiosyncratic ways. The way parent and child information from standardized interviews is combined affects the prevalence considerably (Verhulst et al., 1997a). It is important to use information from multiple informants in order to obtain a comprehensive picture of a child's functioning. Only few studies used teachers in addition to parents and children (e.g. Anderson and Hammen,1993). The use of informants other than parents is relevant because mothers who themselves are depressed may over-report problems in their children (e.g. Fergusson et al., 1993). Third, the comparison samples used in the studies are usually small and highly unrepresentative. Therefore, they do not form an adequate basis for comparing results in the at risk samples. The apparently high prevalences reported in many studies using clinical interviews with children of bipolar parents can only be meaningfully interpreted if we know the population base rates obtained with similar assessment procedures. Purpose of this study The main aims were: 1) to compare parent's, teacher's and self ratings of problems in children of bipolar parents with ratings for children from the general population that were published previously (Achenbach, 1997a,b; Verhulst et al., 1996; 1997b,c); and 2) to determine the current and lifetime prevalence of DSM- IV diagnoses based on interviews with both parents and adolescents with the K- SADS. 15

24 Chapter 2 Table 1: Prevalences of current and lifetime diagnoses based on psychiatric interviews in studies among bipolar offspring Study Number of children Age in years Interview with child (C) or parent (P) Kuyler et al., Structured interview based on Steward and Gath, 1978: P Waters & Marchenko- Bouer, 1980 Diagnostic criteria % of cases with current diagnoses % of cases with life-time diagnoses Type of Bipolar disorder in parent Treatment setting of parent no DSM a,b 45% - I + II Outpatients? SADS-L c : C DSM-III? 50%?? Outpatients LaRoche et al., CPRS d, CARS e : C -? Inpatients DSM-III 6% -? Outpatients Decina et al., MHAF f : C DSM-III/RDC 52% - I + II Outpatients Gershon et al., K-SADS-E: C + P DSM-III 72% - I In- outpatients Kashani et al., DICA-P, DICA g : C + P DSM-III C: 52%-P: 48% h Klein et al., SADS-L: C DSM-III/RDC? 43%? I Inpatients LaRoche et al., CPRS, CARS: C DSM-III 23% -? Outpatients Hammen et al., K-SADS: C + P DSM-III - 92% I? In- outpatients LaRoche et al., CPRS, CARS: C DSM-III 24% -? Outpatients Weintraub, >18 SADS, SCI i : C DSM-III 20% -? Inpatients Nurnberger et al., SADS-L: C DSM-III 74% -?? Zahn-Waxler et al., CAS j : C DSM-III 86% -?? Grigoroiu-Serbanescu K-SADS-E: C + P DSM-III 61% - I Inpatients et al., 1989 Hammen et al., K-SADS: C + P DSM-III 33% 72% I + II In/outpatients Radke-Yarrow et al., CAS : C DSM-III 56% k - I + II Outpatients Carlson & Weintraub, >18 SADS-L, SCI: C DSM-III 53% -? Inpatients Todd et al., DICA-R-P, DICA-R DSM-III-R - 44% I + II Outpatients Duffy et al., K-SADS: C + P DSM-IV - 53% I? Chang et al., K-SADS: C + P DSM-IV - 55% I + II Outpatients Note: a Diagnostic Statistic Manual; b > 3 symptoms = "diagnosis"; c Schedule for Affective Disorders and Schizophrenia; d Children's Psychiatric Rating Scale; e Children's Affective Rating Scale; f Mental Health Assessment Form; g Diagnostic Interview Schedule for (parents of) Children and Adolescents; h these rates are for bipolar and unipolar offspring together; i Structured Clinic Interview for DSM-III; j Child Assessment Schedule; k This is the prevalence at follow-up 16

25 Prevalence of Psychopathology Methods Population and procedure Subjects were enrolled in the study between November 1997-March A survey was sent to all (n = 1961) members of the Dutch Association for manicdepressives (NSMD) explaining the aims of the study and including questions concerning the illness of the subjects, family composition and age of the offspring. Of the 712 (36%) who returned the survey, 110 reported that they had bipolar disorder as well as one or more children in the age range of years. Eventually, 62 out of the eligible 110 parents (56%) agreed to participate with a total of 102 children. In addition, we contacted 9 psychiatric hospitals with an assigned outpatient clinic for patients with bipolar disorder in different regions of the country, representing rural as well as urbanized areas. Psychiatrists of these hospitals were requested to identify all potential participants and to ask them for permission to be contacted by one of us; of the 91 bipolar patients with children aged 12 to 21 years, 58 (64%) agreed to be contacted, and 24 (26%) with a total of 38 children, agreed to participate. A family was included into the study only if all adolescents of the family aged 12 to 21 years agreed to participate. All bipolar parents were outpatients at the moment of recruitment. After a complete description of the study was given, written informed consent from 86 bipolar parents and their spouses, and 140 offspring, was obtained. Finally, only adolescents without a severe physical disease or handicap and with an IQ of at least 70 were included. After initial contact, all consenting adolescents fulfilled these criteria. Instruments Parental characteristics DSM-IV criteria for bipolar I or II disorder were checked by applying the mood disorders section of the International Diagnostic Check List (IDCL, Hiller et al., 1993) in the interview with the bipolar parent and, if available, his/her partner. We compared the IDCL-based diagnoses with the DSM-IV diagnoses made by the treating psychiatrist. No discrepancies were found. Socioeconomic status (SES) was scored on a 9-point scale of parental occupational level with 1 = lowest and 9 = highest. If both parents worked, the highest score was used. The mean SES of the parents in our sample of 4.9 (SD = 2.1) did not differ significantly from the mean of 4.5 (SD = 2.1) from a Dutch general population sample (Netherlands Central Bureau of Statistics, 1993) (t = 1.941, p =.06). Child characteristics The K-SADS-present and lifetime version (K-SADS-PL; Kaufman et al., 1997) is an interviewer oriented diagnostic interview designed to assess current and past DSM-IV symptomatology resulting in diagnoses in children and adolescents, by interviewing the parent(s) and child separately. If parents and child disagreed on the presence of a symptom, greater weight was typically given to parents' reports 17

26 Chapter 2 of observable behavior and children's reports of subjective experiences (Kaufman et al., 1997). The K-SADS-PL was conducted by three of the authors (MW, MH, and CR), and by five intensively trained interviewers with graduate degrees in psychology. In addition to the K-SADS derived diagnoses (for mood, anxiety, attention deficit, conduct, substance abuse, eating, post traumatic stress, adjustment and tic disorders; enuresis/encopresis), we also screened for DSM-IV pervasive developmental disorders. The Child Behavior Checklist (CBCL; Achenbach, 1991a,b) is a questionnaire to be completed by parents of 4- to 18-year-olds and can be scored on 8 syndrome scales and two broad-band groupings of syndromes: Internalizing (consisting of the Withdrawn, Somatic Complaints, and Anxious/Depressed scales), and Externalizing (consisting of the Delinquent and Aggressive Behavior scales). A total problem score is derived by summing the individual item scores. The Youth Self-Report (YSR; Achenbach, 1991a,d) has the same response format as the CBCL but has to be completed by year-olds themselves. It consists of the same scales as the CBCL plus a self destructive/identity problems scale (for boys only). The Teacher's Report Form (TRF; Achenbach, 1991a,c) can be completed by teachers of 5-18-year-olds who have known a pupil in a school setting for at least two months. It consists of the same scales as the CBCL. Good reliability and validity of the CBCL, YSR, and TRF have been replicated for the Dutch translation (Verhulst et al., 1996; 1997b,c). The Young Adult Self-Report (YASR) and Young Adult Behavior Checklist (YABCL) (Achenbach, 1997a,b) are upward extensions of the YSR and CBCL for ages 18 years and older. The YASR is filled out by young adults and the YABCL by parents. The YASR and YABCL can be scored on similar syndrome scales as those of the CBCL, except for the Social Problems scale which was replaced by the Intrusive Behavior scale. The syndrome scales can be scored on two broadband groupings of syndromes: Internalizing (consisting of the Anxious/Depressed and Withdrawn scales) and Externalizing (consisting of the Intrusive Behavior, Delinquent Behavior, and Aggressive Behavior scales). CBCL, YSR and TRF scores for bipolar offspring (boys and girls separately) were compared with those for children of a Dutch normative sample. For the YABCL and YASR, comparisons between bipolar offspring and children of an American normative sample were made because no Dutch normative data were available. The percentage of subjects who could be regarded as deviant using cutoffs recommended by Achenbach (1991a,b,c, 1997) were computed for each sex, age group, 12-18, and > 18 years, and for each instrument separately. In this way sex and age differences were taken into account. For all questionnaires, the cutoff was set at the 82nd percentile for Externalizing, Internalizing, and total problems scores, and for the syndrome scales at the 95th percentile. To assess the children's IQs one verbal (Vocabulary) and one Performance (Block Design) subtest of the Wechsler Intelligence Scale for Children-Revised and Wechsler Adult Intelligence Scale (WISC-R and WAIS), was used. These subtests were chosen because of their high correlations with the full scale score of the verbal and performance Intelligence Quotient (Silverstein, 1972, 1982). 18

27 Prevalence of Psychopathology Results Demographics The demographic characteristics of the sample are shown in table 2. Our sample did not differ significantly from the Dutch general population with regard to sex distribution and SES. Our sample did differ from the general population in mean IQ, which was higher in the offspring of bipolar parents. Table 2: Demographics n % p Males Females Bipolar fathers Bipolar mothers Bipolar I parents Bipolar II parents Married Divorced Mean SD p Age males Age females IQ males IQ females IQ children of bipolar parents IQ general population SES parents of bipolar offspring SES general population

28 Chapter 2 Problem behavior: rating scale scores for children of bipolar parents compared with normative samples Table 3 shows the mean problem scores as reported by mothers for adolescents aged < 19 years on the CBCL, and for adolescents aged > 19 years on the YABCL; by the adolescents aged < 19 years on the YSR or aged > 19 years on the YASR; and by the teachers for adolescents aged on the TRF. We will report the differences that were significant at the level of p<.05 and p<.01. CBCL/YABCL. Girls of bipolar parents obtained significantly higher scores than girls of the Dutch normative sample on the following CBCL scales: total problems, Internalizing, Externalizing, Somatic Complaints, Anxious/Depressed, Social Problems, Delinquent Behavior and Aggressive Behavior. Boys of bipolar parents obtained significantly higher scores on the total problems, Externalizing, Thought Problems and Aggressive Behavior scales than boys of the normative sample. No YABCL scale showed a significant difference in mean scores for the bipolar offspring versus the normative samples. YSR/YASR. The mean problem scores of the YSR scales as reported by girls < 19 years of bipolar parents yielded no significant differences versus the normative sample. Boys of bipolar parents aged < 19 years scored significantly lower than normative boys on the total problems, Social Problems, Attention Problems and Self Destructive/Identity Problems scales. On the self-report measure for the older children (YASR), girls of bipolar parents scored significantly higher than a normative sample of US girls on Attention Problems. Boys of bipolar parents scored significantly lower on the Externalizing, Withdrawn, Thought Problems and Aggressive Behavior scales than boys in the normative sample. TRF. The mean problems scores reported by teachers yielded two significant differences: girls of bipolar parents scored significantly lower than girls in the normative sample on the Thought Problems scale and boys scored significantly lower than boys in the normative sample on Social Problems. 20