The October 2009 issue of the JCPP brings together several diverse topics and themes. First, this issue presents several studies examining anxiety in pediatric and adolescent populations. Côté and colleagues examine the developmental course of anxiety in infants whereas Van Oort and colleagues study the developmental trajectory of anxiety in early adolescence. Hale and colleagues delve more deeply into the extent to which adolescent anxiety and depression develop as different manifestations of the same underlying etiology or as distinct disorders. Finally, Mueller and colleagues draw links between spatial navigation errors and anxiety, suggesting potential deficits in hippocampal functioning.

This issue also presents several papers related to Autism Spectrum Disorders (ASDs). In recent years, there has been greater recognition of the need for early identification and treatment of ASDs. A necessary condition for early identification and treatment is a stable construct of ASD in young children. This topic is addressed directly by Chawarska and colleagues. Another important condition for early identification and treatment is precision in the diagnosis of ASD. Lecavalier and colleagues examine this issue in the context of Pervasive Developmental Disorders. Hutman and colleagues examine the relationship between ASD and family characteristics, in their case the role of maternal synchrony. In addition to anxiety and autism spectrum disorders, this issue also presents papers covering several other important complex traits including oppositional defiant disorder (Munkvold and colleagues), characteristics of inpatient care (Jacobs and colleagues), fire setting (MacKay and colleagues), and bipolar disorder and post traumatic stress problems (Ayer and colleagues).

Finally, although all papers in this issue address the issue of etiology, three additional papers examine etiology more directly. Deater-Deckard and colleagues examine the relationship between household chaos and child cognitive outcomes in the context of other aspects of the home. Hicks and colleagues employ behavioral genetic designs to examine gene x environment interactions in internalizing behavior problems. Finally, Haworth and colleagues estimate the genetic and environmental contributions to the correlation among different kinds of learning problems.

Clearly, these are diverse studies, varying in terms of child age, sampling characteristics, substantive area, measurement strategy, as well as analytical method. So, what are the common themes across these diverse areas of study? One is the tension between measures of complex behavior and etiology. As researchers and clinicians, we are constantly juggling the need to develop constructs that are ecologically valid but are not overwhelmed by the numerous etiologies that yield individual differences within the construct. This is evident in several papers in this issue. For example, Munkvold and colleagues show that informant characteristics are important to measurement of oppositional defiant disorder, whereas Lecavalier and colleagues demonstrate how heterogeneity in sample, modality, and procedure influence the structure of pervasive developmental disorders. At the same time, other papers in this issue suggest that seemingly different constructs share a common etiology. Haworth and colleagues conclude that learning impairments share a common genetic etiology. Similarly, Ayer and colleagues show that bipolar disorder and post traumatic stress problems are influenced by a common underlying etiology.

So how do we reconcile this tension between constructs that seem to converge behaviorally but have numerous etiologies as well as the converse, where seemingly different manifestations of behavior are influenced by a common etiology? It is tempting to search for biological, environmental, or other reductionistic determinants of behavioral outcomes. However, whether genetically or environmentally influenced, these ‘lower level’ mechanisms are operationally defined by their relationship to the complex trait (such as anxiety) and are often still highly complex endophenotypes in and of themselves (e.g. spatial navigation). Also, reductionistic approaches usually result in models that are highly specified but explain little variance in the overall population whereas models based on complex traits usually explain more variance, but do not explain the mechanisms through which this variance arises (e.g. Haworth).

Thus, what we are left with is recognizing that all scientific approaches involve using empirical methods to make a leap between observation and theory. Thus, even the most seemingly obviously ‘simple’ biological process examined from the outside by behavioral researchers is, in reality, an incomplete, operationally defined construct. In the same way, even the most seemingly ‘simple’ behavioral outcome to a biologist is cut out of the world through observation. This is the central recognition that needs to take place; that one level of analysis does not precede another, nor can they be combined easily, but we as scientists should strive to cross levels of analysis where possible, or at least develop theories that are consistent with other domains. To date, most of our theories of child development and child psychopathology have been based on examining the relationship between complex traits or examining the relationship between complex traits and underlying process. However, there are other approaches; biological, genetic, sociological; and where possible using measures from these traditions, or at least becoming generally aware of the issues within these fields, may help to develop theories of child psychopathology that are more biologically plausible, or sociologically plausible. In the same vein, researchers focused on biology may find greater purchase on their domains of study with better understanding of the issues of measurement and sampling that are so important to the behavioral researcher. In sum, modern research in child psychopathology and behavior has the potential benefit of using not just the approaches historically common to our field, but also those emerging in other domains, and in an era where finding research and researchers in other domains is relatively easy, building true multidisciplinary collaborations focused around the central question of defining, assessing, and intervening to better psychiatric outcomes in children is not only possible, but necessary.