Growth Charts for Brain Development?


by Kayt Sukel

June 27, 2016

During each check-up, my son’s pediatrician has compared his development to a series of growth charts. He noted his height and weight, as well as his head circumference—monthly during the first two years of life, then annually—to gauge the rate and proportion of his development and make sure it was on track. Now, a new study from researchers at the University of Michigan suggests that pediatricians could also use a growth charting technique to follow the brain’s intrinsic connectivity networks as they mature. The researchers argue that this sort of tracking could help doctors predict which children are at most risk for attention impairments such as attention deficit hyperactivity disorder (ADHD), a neurodevelopmental disorder characterized by impulsive and hyperactive behavior as well as problems with sustained attention.

Diagnosing ADHD today

Diagnosing the disorder in children today requires a “clinical investigation,” says Anthony Rostain, a professor of psychiatry at the University of Pennsylvania and the Children’s Hospital of Philadelphia who specializes in ADHD. “There is no lab test for ADHD—or any simple way to go about diagnosing it. It’s a disease that exists on a continuum. It’s not always easy to distinguish between a child that really has ADHD and just a more rambunctious kid, because every child with ADHD looks a little bit different,” he says. “So we ask a lot of questions of the parents, observe the child, ask teachers to verify the reported behavior using more-or-less standardized neuropsychological tools. We take a look at the matter of degree, intensity, and frequency of symptoms and how much they interfere with functioning. But even after that, there can still be quite a bit of debate about who has ADHD and who doesn’t. There’s a lot of heterogeneity in this developmental disorder.”

It’s that heterogeneity that led Daniel Kessler and Chandra Sripada, researchers at the University of Michigan’s department of psychiatry, to wonder if there might be a better way to diagnose attention problems. Converging data on intrinsic connectivity networks (ICNs)—networks of functionally connected brain regions whose activity can be measured when a person is at rest—suggests that ADHD may be the result of ICNs that are improperly “synced” and therefore unable to effectively communicate.

“Now we do a very complicated battery of task-based tests that take half a day, then a doctor interviews the patient for an hour and we talk to parents and teachers. And we have to do all that before we discuss what the diagnosis could be,” says Sripada. “The fact is that many studies have shown that this kind of diagnosis method gives us very little discriminant information about the disorder. But, in the future, work from functional magnetic resonance imaging (fMRI) studies may give us more to work with so we can have a clearer idea of what might be going on.”

Growth charting the brain

Kessler says that growth charts are already very useful to pediatricians, helping to flag potential problems in maturity and development. Such a method could also be used with ICNs. To test the idea, Kessler, Sripada, and colleagues used publicly available neuroimaging and neuropsychological data from more than 500 children with an average age of 16 years. In that group, approximately 5 percent met the criteria for an ADHD diagnosis. The researchers simplified ICN data from the study participants by using a machine learning technique to create growth chart deviation “scores” based on how the behavior of the ICNs compared with normal maturation and functional activity. They found that the children with strong deviation scores also showed worse performance on sustained attention tasks. The study was published online in April in JAMA Psychiatry.

“Using this technique enabled us to take this really sophisticated and nuanced data, things we’ve measured in people in hundreds of thousands different ways, and boil it down to a smaller number of scores of deviation patterns in the brain’s network growth, where those scores indexed the pathological organization of the different ICNs measured and how it different from normal maturation,” says Kessler. “It’s possible that we can use those scores like a growth chart to help clinicians identify when there might be problems with attention in the future.”

A feasible plan?

Sripada says such an fMRI-based growth charting technique would not replace the current method but could be another tool in a psychiatrist’s arsenal to make a good clinical diagnosis. It could even direct doctors to good interventions based on how and where a child may be lagging. As for the cost of such brain-growth tracking, Sripada admits that brain imaging remains quite expensive, but he says it isn’t that much more expensive than current tools used by psychiatrists for diagnosis—those common half-day evaluations can cost thousands of dollars. Kessler hopes to distill the growth chart method to a less-expensive tool, such as electroencephalogram (EEG). He and his team are currently working towards that end.

“fMRI is really useful for discovery and can help us do the discovery science to figure out the meaningful pathological features that we need to track,” he says. “But our hope is that we can transfer the deeper, more detailed picture from the fMRI to characterize the population but then use EEG to locate at-risk individuals against a really well-defined population curve.”  

Both Sripada and Kessler say there is a long road ahead to replicate their initial findings and then transfer them to a useful and useable diagnostic tool. They also caution that this study focused on sustained attention; future work needs to look at other symptoms of ADHD like impulsivity and hyperactivity. But they think this could be a useful technique one day.

Rostain says he finds these initial findings “appealing on a lot of levels,” but argues it will take a lot more time and work to see how applicable this model will really be to diagnosis outside the laboratory.

“This is a hugely fascinating way of thinking about ADHD—it’s not just that your wires are there but whether the wires are firing the way they ought to be, in sync with one another. The notion of how intrinsic connectivity is altered in disease states is one of the biggest paradigms going in neuroimaging these days,” he says. “But there’s still a lot left to understand. ADHD is not a simple problem, to identify or to fix. But hopefully these kinds of brain studies are going to show us the complexity involved and will give us some good, cost-effective approaches to better diagnose and treat these kinds of conditions in the future.”