**6.1 Background**

148 Epidemiology Insights

education compared to higher maternal education (28.3% vs. 6.5 %; *χ*2 (2) = 15.82, *p* < .001). Rates of total problem behaviour were also significantly more common with lower education of fathers (< 9 years) in comparison to fathers with a higher education level (24.7% vs. 8.6 %; *χ*2 (2) = 12.72, *p* < .01). Caregivers reported no significant associations

The prevalence of *Internalizing* problems reported by both parents was higher in low-income families (mothers: 35.5 %, fathers: 32.7 %) than in the group with an income of over 3,000 € per month (10.8 % and 9.2 %; *χ*2 (2) = 20.71, *p* < .001; *χ*2 (2) = 12.14, *p* < .01). The prevalence of *Total Problems* reported by mothers is comparable in this context: Children from low-income households (24.3 %) were overrepresented compared to children from higher-income families (3.2 %; *χ*2 (2) = 18.67, *p* < .001). In contrast to these findings, caregivers reported more *Externalizing* problems for children from lower-income families. The results narrowly

The immigrant children's adjustment reported by informants was as follows: Non-German children (rated by their mothers) had higher prevalence rates on internalizing problems than those reported for native-German preschool children (48.5 % vs. 20.3 %; *χ*2 (2) = 24.23, *p* < .001). The odds ratio was 3.0 (95 %-CI 1.82-4.96). There were no significant associations between immigrant status, problem behaviour and other informants. There were also no significant discrepancies between parental marital status and prevalence rates of problem scales. The odds ratio for *Internalizing* problems in single-parent families versus dual-parent households was 1.96 (95 %-CI 1.07-2.68). The odds ratio for scoring T ≥ 60 on *Total Problems* was 1.80 (95 %-CI 1.08-3.00) for children with single parent versus dual-parent families.

This study was designed to determine the prevalence rates of behavioural and emotional problems for different informants among 3-6-year-old preschool children and to evaluate

In conclusion, 5 - 6 % predominantly clinically relevant internalizing problems were found. From the total of 447 children, 12 – 13 % met the criterion of the clinical range of internalizing mental health problems. Thus, an important finding across informants was higher prevalence for internalizing than externalizing problems in preschool children. Other studies in this age group have found higher rates of externalizing as opposed to internalizing problems (Bongers et al., 2003; Campbell, 2006). Therefore, this result was unexpected and might be important in understanding mental health in the preschool years. Such discrepancies may result from different assessment measures, procedures, and normative data. The use of the CBCL ASEBA preschool forms (Achenbach & Rescorla, 2000) is a particular strength of the study, as the CBCL is a well-established measure of mental health morbidity. While this advantage is important, the study was limited by the lack of representativeness. So far, no comparable studies using ASEBA preschool instruments have been published. When interpreting the results, it should be taken into account that the child mental health status was assessed by a symptom checklist questionnaire. Given a large number of subjects and multiple informants, the questionnaire approach is economical and offers useful information, but lacks the specificity and additional depth that structured

demographic factors which may be associated with prevalence.

between parents' education and child behaviour problems.

failed to reach statistical significance.

psychiatric interviews might provide.

**5.4 Discussion** 

Within the last century, considerable changes in the health and illness patterns of young children have been observed. One characteristic of this phenomenon, which is referred as the "new morbidity', is the growing importance of mental health concerns (Palfrey et al., 2005). Externalizing problem behaviours are the most common and persistent forms of childhood maladjustment (Campbell, 2006). Kraemer et al. (2000) reported that, at the same time, externalizing behaviours change so much in expression and frequency over the course of development that studies at any single point in development will provide only limited information or misrepresent the phenomenon. Therefore, there is a growing agreement that externalizing behaviour must be studied from a developmental perspective (Costello & Angold, 2006). The present article aims to describe the development of externalizing behaviours from preschool age to primary school.

Previous studies have investigated the development of externalizing behaviour in the general population (e.g. Bongers et al., 2004; Hofstra et al., 2000; Loeber et al., 2000). However, these studies used two different diagnostic approaches to describe externalizing problem behaviour. The *empirical approach* utilized multivariate statistical techniques to identify sets of problems that tend to occur together. This approach starts with empirical data derived from different informants who describe the behaviour of children and forms the basis of the empirical syndromes of rating scales such as the Child Behavior Checklist (CBCL; Achenbach, 2009) or the Conners' Rating Scales (Conners, 1997). The *diagnostic approach* is to take the diagnostic categories of one of the international nosological systems, the fourth edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) or the World Health Organization's International Classification of Diseases (ICD). Instead of disregarding one approach for the other,

The Epidemiology of Child Psychopathology: Basic Principles and Research Data 151

contacted all potentially eligible preschools (*N = 33*). Project staff members were present at preschool teacher meetings and explained the project. Twenty-three preschools (70 %) expressed interest in participating in the project. Seventeen of these preschools were then randomly selected to participate in the project, and then preschools were randomly assigned

The study population comprised 136 control preschool children. The baseline demographic characteristics of the 136 children and their families were follows: The age of the mothers ranged between 23 and 57 years (*M* = 35.0, *SD* = 5.4). The families had between one and five children (*M* = 2.0, *SD* = 0.9). The average age of the target children was 4.1 years (*SD* = 1.1), and 51% (*n* = 69) were girls. 33 % (*n* = 46) of the children lived with single parents. Fiftythree percent of mothers had a higher-track school qualification (= 13 years of schooling), and 32 % had completed a medium-track school (= 10 years of schooling). The net family income was equivalent to the German average; 33 % of the families were receiving social

The developmental course of child behaviour problems was established with self-report measures from mothers at pre, 1, 2, 3 and 4 years after the first assessment (follow-up 1 - 4). Across the four years after pre-test, 2 - 4 % of the families dropped out of the study, leaving 122 families (retention rate 90 %). The sample size at the 4-year follow-up assessment (primary school) consisted of 62 boys (50.8 %) and 60 girls. The mean age of the children

At pre-assessment, families provided information regarding their age, nationality, relationship to the child, education level, employment, receipt of social welfare assistance, and household income. They also provided data on the age and gender of the child of

Child mental health was measured during the preschool years by the German version of the Child Behavior Checklist (CBCL/1½-5; Achenbach & Rescorla, 2000; see 5.2.2). A recent study tested the generalizability of the seven-syndrome model in 23 societies (Ivanova et al., 2010). Findings from this study indicate that researchers (clinicians) can use the syndromes to assess preschool psychopathology. For the 2-year follow-up and after, the Child Behavior Checklist 4-18 (Arbeitsgruppe Deutsche Child Behavior Cheklist, 1998) was used. The scores from the parent-report were classified according to the manual into age- and sex-dependent

The German ADHD Rating scale (FBB-HKS) is part of the comprehensive Diagnostic System for Mental Disorders in Childhood and Adolescence (DISYPS-KJ; Döpfner & Lehmkuhl, 1998) and can be rated by parents and teachers. This ADHD rating scale includes 20 items addressing symptom criteria of both ICD-10 and DSM-IV as well as additional criteria assessing symptom onset, symptom duration, pervasiveness and functional impairment. Internal consistencies were satisfactory to very good in the different representative samples (Döpfner et al., 2008). The DSM-IV recognizes three subtypes of the disorder - the predominantly inattentive type, the predominantly hyperactive-impulsive type and the combined type. Children were diagnosed with any ADHD if parents reported that six or

categories which are based on the percentiles of the normative study.

more symptoms had persisted for at least 6 months.

to either the intervention or control condition.

security benefits, and 8 % of mothers were immigrants.

was 8;8 years (*SD* = 1;1; range 6;3 – 10;8 years).

**6.2.2 Measures** 

interest and any siblings.

Verhulst and Koot (1995) hold the view that both approaches are needed, and that combining both by adding information from one approach that is not captured by the other may increase our knowledge of children's psychopathology (Ferdinand et al., 2004).

A classification scheme of externalizing behaviours developed by Frick and colleagues (1993) distinguishes four types of externalizing behaviour problems based on a metaanalysis of 60 studies involving more than 28,000 youth. The four behavioural clusters that emerged may be ordered along two independent dimensions (overt vs. covert; destructive vs. non-destructive) and were labelled opposition, aggression, property violations, and status violations. These behaviour clusters were also confirmed in independent studies of adolescents. Most mental health practitioners and researchers distinguish between two types of childhood conduct problems based on the age at which children show first symptoms and the persistence of the symptoms across development (Moffitt, 2003). The differentiation between childhood-onset and adolescent-onset conduct problems is based on results from the Dunedin Multidisciplinary Health and Development Study, a 30-year longitudinal study of 1,000 New Zealand youths (Moffitt et al., 2001). Moffitt et al. (2001) identified two developmental pathways for childhood conduct problems: the life-course persistent path and the adolescence-limited path. Children with life-course persistent conduct problems first show symptoms in preschool or early primary school. Partly consistent with this theory, four developmental trajectories were identified for boys' externalizing problems from ages 2 to 8 and 6 to 15 years in two samples (Nagin & Tremblay, 1999): a *persistent problem* trajectory, a *high-level desister* trajectory, a *moderate-level desister* trajectory, and a *persistent low* trajectory. In addition, Campbell et al. (2006) identified for physical aggression from 24 months to age 9 the following trajectories: very-low, lowstable, moderate-decreasing, moderate-stable, and high stable aggression. Schaeffer et al. (2003) identified four somewhat different pathways of antisocial behaviour from first to seventh grade within an epidemiological sample of boys: *chronic high, moderate* (and stable), and *increasing aggression* trajectories as well as a *nonaggressive* trajectory. Early, persistent externalizing problems (e.g. aggression), however, predicts a range of negative outcomes including poor emotion regulation and impulsive behaviour, school failure and dropout, peer problems, and adolescent delinquency (Patterson et al., 1989; Tremblay, 2000).

Studies, in which researchers employed the empirical approach, have shown that between ages 2 and 9 children generally decline on externalizing behaviour measures (NICHD Early Child Care Research Network ECCRN; Shaw et al., 2003). Bongers et al. (2003) also found a decline in mother-reported externalizing behaviour problems for both boys and girls between ages 4 and 18 in a representative sample of over 2,000 Dutch children. The aims of the study were 1) to assess the prevalence and incidence of externalizing behavioural problems in children from kindergarten to primary school, and 2) to investigate the developmental course of externalizing behaviours.

### **6.2 Methods**

#### **6.2.1 Sample and design**

In the present study, families with children aged 3 to 6 years were recruited in preschools in the city of Braunschweig (Germany), a moderately sized city with 250,000 inhabitants. Families were recruited for a universal prevention efficacy study of child behaviour problems. Study details for recruitment were described by Hahlweg et al. (2010). We first contacted all potentially eligible preschools (*N = 33*). Project staff members were present at preschool teacher meetings and explained the project. Twenty-three preschools (70 %) expressed interest in participating in the project. Seventeen of these preschools were then randomly selected to participate in the project, and then preschools were randomly assigned to either the intervention or control condition.

The study population comprised 136 control preschool children. The baseline demographic characteristics of the 136 children and their families were follows: The age of the mothers ranged between 23 and 57 years (*M* = 35.0, *SD* = 5.4). The families had between one and five children (*M* = 2.0, *SD* = 0.9). The average age of the target children was 4.1 years (*SD* = 1.1), and 51% (*n* = 69) were girls. 33 % (*n* = 46) of the children lived with single parents. Fiftythree percent of mothers had a higher-track school qualification (= 13 years of schooling), and 32 % had completed a medium-track school (= 10 years of schooling). The net family income was equivalent to the German average; 33 % of the families were receiving social security benefits, and 8 % of mothers were immigrants.

The developmental course of child behaviour problems was established with self-report measures from mothers at pre, 1, 2, 3 and 4 years after the first assessment (follow-up 1 - 4). Across the four years after pre-test, 2 - 4 % of the families dropped out of the study, leaving 122 families (retention rate 90 %). The sample size at the 4-year follow-up assessment (primary school) consisted of 62 boys (50.8 %) and 60 girls. The mean age of the children was 8;8 years (*SD* = 1;1; range 6;3 – 10;8 years).
