Predicting whether children with ADHD will get better over time

ADHD is a controversial condition for several reasons:

  1. It is widespread and costly, so it attracts interest.
  2. It is treated with stimulant medication, and medication use has skyrocketed, raising questions about how diagnosis is conducted.
  3. Most crucially, over time, some children naturally get better while others have very poor outcomes, and we cannot yet predict which are which.

A major scientific focus is to identify biologically sound subtypes that will add to clinical prediction. In the OHSU ADHD Program, we are pursuing a neurobiologically informed model of subtypes that takes into account both the emotional and the attentional aspects of these children’s struggles.

The figure on the right shows that one cause of ADHD is in disruption of attention and another cause is in disruption of emotion. However, we don’t yet know if these are independent.

The approach also takes into account that their difficulties have a volitional component (partly under their control) and an involuntary component (partly beyond their deliberate self-control). For that reason the figure above depicts “bottom up” (involuntary) and “top down” (controlled) processes.

Evidence to support this approach has come from several studies conducted in the OHSU ADHD Program. The second figure shows a cluster analysis of a large sample of more than 500 children. On the graph, “normal” scores are 0 and “high” scores are “bad.”

When we organize the data based on performance on a broad battery of neuropsychological scores, we see that there is a group with problem in attention (controlled attention), and another group with problems in time processing; the other groups have problems in terms of low arousal or alertness.

This supports the idea that types of attention problems divide children into those with top down and those with bottom up attention breakdown. This is valuable because these attention types can be mapped onto
brain circuits and tested.

A second study (3rd diagram) shows a similar analysis from the perspective of emotion processing. It shows that one group of children with ADHD is characterized by highly irritable, angry behavior, another simply by exuberant behavior, and a third by normal emotional processing. Again, we can map these domains into known neurobiological pathways. In this last example, we also found that using this typology we could predict one year later which children were doing worse and which ones were not doing worse.

We could do so far more accurately than with existing clinical tools. Therefore, this approach is promising for improving clinical prediction in ADHD.

Joel T. Nigg, Ph.D.
Director, Division of Psychology
Director, ADHD Program
Professor, Departments of Psychology and Behavioral Neuroscience
OHSU School of Medicine

2 responses to “Predicting whether children with ADHD will get better over time

  1. What can you predict, in terms of adolescent/adult behaviors like school failure, drug abuse, and criminality, based on these subtypes?

  2. Thanks for the question, Pamela. This is what the OHSU ADHD Program is currently studying. They were just funded to follow the children to age 18 years, and will then be able to answer your question and others like it. Sorry we don’t have any additional information yet! –Lisa

Comments are closed.