**4.1 Basics**

Prevalence can be defined as the proportion of a population that has mental health problems at a specific point in time; it is often defined as *point prevalence*. Prevalence is calculated by dividing the number of cases by the total population. It is also possible to quantify the number of cases known to have the disorder at any time during a specified period. This so-


Epidemiology does not refer to a single scientific discipline or the use of one specific methodology. Instead, epidemiology derives and integrates concepts and methods from other areas such as biology, statistics, and sociology. In epidemiology, the combination of measurement principles and statistics is used for the development and testing of diagnostic assessment procedures. When applied to psychological concepts, the combination of measurement and statistics is called psychometrics. In epidemiology of child psychopathology, psychometric principles play an important role. A number of specialized areas that are derived from classical epidemiology are relevant to child psychopathology, including clinical epidemiology, genetic epidemiology, and pharmaco-epidemiology. Clinical epidemiological studies concern the development and application of diagnostic and screening tests, the prognosis of disorders, the effects of treatment and clinical decision-

Epidemiology is concerned with the study of the distribution and determinants of disease frequency in human populations. The quantification of the occurrence of psychopathology in populations, can be regarded the central task of epidemiology. Well-known measures of frequency are *prevalence* (see 4.1) or *incidence*. Incidence quantifies the number of new cases with a disorder that develop in a population during a specified period of time. *Cumulative incidence* is the proportion of individuals who become disordered during a specified period of time (Verhulst, 1995). The distribution of disorders, involves comparisons between different populations or subpopulations. The examination of factors that are associated with variations in the distribution of psychopathology is essential for testing etiological hypotheses. Measures of association and risk are quantifications of the influence of certain factors on the occurrence of disorder. In follow-up studies measures of risk for developing a disorder using categorical data are *relative risk*, and *attributable risk*. In case-control studies the measure often used is the *odds ratio* which reflects the likelihood for developing a disorder in the group with a possible aetiological factor versus the group without this factor

(Verhulst, 1995). For a more detailed discussion, see Verhulst and Koot (1992).

determination of the prevalence, the study is called prevalence study.

**4. Prevalence studies** 

**4.1 Basics** 

Epidemiological studies can be divided into prospective and retrospective studies, depending respectively on whether the measurement of exposure to a risk factor was done before or after the disorder occurred. A study, in which the presence or absence of a disorder and the presence or absence of associated factors are assessed at the same time, is called a cross-sectional study. If the aim of the cross-sectional study is limited to the

Prevalence can be defined as the proportion of a population that has mental health problems at a specific point in time; it is often defined as *point prevalence*. Prevalence is calculated by dividing the number of cases by the total population. It is also possible to quantify the number of cases known to have the disorder at any time during a specified period. This so-

the healthcare system due to mental disorders and their consequences.

**3. Epidemiological concepts and strategies** 

making.

called *period prevalence* (e.g. 6-month prevalence or lifetime prevalence) is frequently used in prevalence studies of child psychiatric conditions. There are two types of studies determining the prevalence of child psychopathology: (1) those that produce prevalence rates of psychiatric diagnoses, usually based on DSM (Diagnostic and Statistical Manual of Mental Disorders) criteria, and (2) those that generate scores on psychiatric symptom rating scales.

Many studies that determined prevalence rates of DSM diagnoses in general population samples of children have been conducted. Costello, Egger and Angold (2005), Roberts, Attkisson and Rosenblatt (1998), Verhulst (1995), and Waddell et al. (2002) provided reviews. Despite huge research efforts and many children involved, comparisons between these studies are seriously hampered by large differences in design and methodology including differences in sample size, age of children, assessment and sampling procedures, and case definition. Even for studies conducted in countries comparable in language, culture, and availability of services, differences in prevalence rates were extremely large and ranged from 10% to 20% (Waddell et al., 2002). It is more likely that these differences reflect variations in methodology than differences in true prevalence. Methodological variations and the lack of standardization among studies seriously limit the value of prevalence figures of categorical diagnoses.

The second approach, the use of rating scales for assessing parent- or self-reported emotional and behavioural problems of children in representative general population samples, is less vulnerable to methodological differences. This approach produces problem scores on continuous scales and does not generate prevalence rates for categorical diagnoses. Often, statistical criteria are used for distinguishing between cases and non-cases. Although dividing lines for caseness may be rather arbitrary, there are epidemiological methods for selecting effective cut-off points. However, prevalence figures will vary with the statistical criterion and cannot be used as absolute population prevalence measures without relating them to similar measures for other populations or subpopulations (Verhulst, 1995). In two recent multicultural prevalence studies, parents' reports and youths' self-reports of problems for children using the Child Behavior Checklist (CBCL) in 31 cultures, and the Youth Self-Report (YSR) in 24 cultures, were compared (Rescorla et al., 2007a; b). It was found that, when the same standardized assessment procedures are used for assessing children from different cultures, cultural differences per se do not lead to big differences in reported problems. Instead, individual differences within each cultural group are bigger than differences between the average scores obtained in different cultures. Assessment procedures with good cross-cultural track records and appropriate translations that capture individual differences in reliable and valid ways are apt to reflect the mental health needs of children that are robust across cultures (Achenbach & Rescorla, 2007).
