Abstract
One of the many problems plaguing developmental psychopathology is behavioral phenotyping of disorders. Although most diagnostic criteria are designed to cover the lifespan; they have few, if any, modifications to accommodate for age and developmental level of the affected individual. This could not be clearer than with the disruptive behavior disorders (Wakschlag et al. 2008), but it applies to other conditions as well. The examination of early onset psychopathology requires a trained eye and the expectation that, over time, one of several events may occur, either as part of the natural history of the disorder or in response to treatment: 1. Diagnostic stability – the phenotype is largely unchanged over time 2. Resolution – the disorder phenotype ends and typical development proceeds 3. Heterotypic continuity – the underlying pathological developmental process(es) stay the same, but the phenotype changes 4. Homotypic continuity – the underlying pathological developmental process(es) change but the phenotype stays the same
What is the cause of these changes? Is it so-called “normal development” that includes remarkable resilience and neuroplasticity? Is it the time-limited course of some disorders? Is this the end of one episode in a disorder that will have an exacerbation at some later date? Or, is this a response to one or several treatments or other environmental factors? Presumably, longitudinal observational studies and well-controlled treatment trials can help answer these questions; however, these are rarely of the size and scope required to have enough power to analyze the relative effects of each of these possibilities.
Recent reports may give us a glimmer of what lies ahead. For example, Bufferd and colleagues (Bufferd 2012) examined 462 community-ascertained (not representative) children at age 3 and then re-assessed them again at age 6. The findings suggest that the rates of emotional, anxiety, or behavioral disorder did not change in the 3 year interval. Analyses further suggested that there was homotypic continuity between age 3 and age 6 anxiety, attention-deficit/hyperactivity disorder (ADHD) and oppositional defiant disorder (ODD). Interestingly, this did not apply to depression, as no individuals who were depressed at age 3 met criteria for depression at age 6. Additionally, there was significant heterotypic continuity between age 3 depression and age 6 anxiety, age 3 anxiety and age 6 depression, age 3 anxiety and age 6 ODD, age 3 ADHD and age 6, ODD, and age 3 ODD and age 5 ADHD. Unfortunately, there are no data on treatment and how this may have affected these cases, but it seems clear that children who are affected early, remain affected by some condition, and that those who have a diagnosis of psychopathology at age 6 are likely to have been affected by some disorder at age 3.
Does this, however, apply to all disorders, everywhere? Apparently, that is the case for ADHD. In a number of recent studies from around the world, the stability of ADHD symptoms has been noted over extended periods of time. As examples, Mordre and colleagues (2012) followed a group of Norwegian child psychiatric patients over 28 years and found persistent problems, including low levels of psychosocial functioning and high rates of enrollment for disability pensions. Similarly, Kan et al. (2013) used the Netherlands Twin Register to find that attention problems have a high level of persistence into adulthood, with genetic factors contributing more early on and environmental factors playing a greater role in symptom stability later. And, finally, another 33 year longitudinal follow-up study from Klein's group (Klein et al. 2012) found adverse outcomes in the probands, with 22% still meeting criteria for ADHD along with higher rates of substance use and antisocial behaviors (heterotypic continuity?). However, in each of these studies, a significant number of individuals no longer met diagnostic criteria and seemed to do well. What makes the difference?
What about chronic disorders such as autism spectrum disorder (ASD)? From a well-executed and very striking study by Fein and colleagues at the University of Connecticut (Fein et al. 2013), comes a report on a group of 34 children who were diagnosed with ASD by skilled clinicians before the age of 5, but who no longer met diagnostic ASD criteria as early adolescents. Indeed, the so-called “optimal outcome” individuals were distinctly different from controls with high functioning autism (HFA) and more like typically developing, matched controls on measures of socialization, communication, face recognition, and most language scales. They also reportedly attend regular education classes without assistance. The use of both “typically developing” and HFA controls helps clarify the distinctions, however, as with many studies, there are some important limitations in terms of selection (more affluent, higher cognitive level) and many questions left unanswered. It is clear that a significant subset of individuals with “optimal outcomes” (at least 7 to 20%) have identifiable forms of psychopathology: “some form of impairment (although not marked) in nonverbal social interaction, either in facial expressiveness, gesture, or eye contact; these difficulties were judged to be secondary to inhibition, anxiety, depression, inattention and impulsivity, embarrassment, or hostility, and not to have an autistic quality” (Fein et al. 2013, p. 198). These are interesting observations, especially as no specific screening for other forms of developmental psychopathology is reported in this article. And, if there is screening, might there be other such cases? Is this heterotypic continuity? Could these children have had other disorders all along? Additionally, what role, if any, did treatment play in this outcome? Wheras it is clear that the clinicians see these children differently at two different time points, Fein and colleagues challenge us to think deeply about what accounts for these changes.
These and similar reports expose the serious problems that result from the lack of longitudinal normative developmental data from epidemiologic samples, especially those that assessed symptoms of developmental psychopathology. Although those data may not be available, we can learn that the study of both diagnostic stability and treatment response in developmental psychopathology must consider the role of that which is treasured most by the field: Development. How do typical and atypical developmental processes, including biological and environmental events, affect behavioral phenotypes, over time? This requires that diagnostic studies must consider heterotypic continuity for changing diagnoses or changes resulting from treatment. Similarly, homotypic continuity must be considered when there is a lack of change.
All of this serves to highlight the complexity of understanding and treating developmental psychopathology. In short, we often think we understand why children change over time; however, we should remember that, just as often, we do not. And, whereas many think that “change is good,” just because something changes does not mean that it is not the same thing.