Castellanos is a pediatrician and child psychiatrist conducting neuroimaging and genetic studies of ADHD. He is the head of ADHD research at the National Institute of Mental Health (NIMH). FRONTLINE interviewed Castellanos on October 10, 2000.

ADHD is defined as symptoms of hyperactivity, impulsivity, and inattention beyond what's usual for a developmental age. A 6-year-old is expected to be more hyperactive than a 10-year-old. But a 6-year-old with ADHD can be much more hyperactive than a 6-year-old without ADHD in certain settings; likewise inattention; likewise impulsivity. Those are observed in the home, in school, in play situations. But we don't have a test. We don't have an objective way of definitively saying, "This person has ADHD, or does not," in part, because we don't really understand what it is. ...

How does ADHD work on the brain? What do we know about it?

We don't yet know what's going on in ADHD. We've approached it in a number of different ways, and one way has been to look at what brain regions are smaller, or different. What we've found is that there are a few regions that are smaller in kids who have ADHD. There are other groups that are looking at functional MRI, or at SPECTA (Single Photon Emission Computomography), usually in adults who have ADHD. And most of the evidence converges and suggests that regions that are rich in dopamine are involved. And that's interesting, because dopamine is one of the chemicals that Ritalin boosts.

... What does dopamine have to do with any of this?

... Dopamine is an important signal that certain parts of the brain use to regulate movement. If you don't have dopamine in those brain regions, you develop Parkinson's disease. But it's also apparently used to send a signal that something important is happening. ...

It used to be said that dopamine was the reward chemical--that if something was rewarding, then you would release dopamine. It turns out to be more complicated than that. It's not just whether something's going to feel good, or be rewarded; it's more if there's a possibility that something would feel good.

If an animal knows that they're going to be rewarded when they correctly do a task, then dopamine is no longer involved. But when the animal thinks that maybe this is the way to solve the task, dopamine is leading the way, saying, "This, try this, try this." So it's a very important signaling molecule. And it's classified officially, I guess, or technically, as a neurotransmitter. But it's probably more appropriately described as a neuro-modulator, which means that it sets or modulates the tone for these complex systems.

So we think that it's important to have optimal levels of dopamine in various brain regions; if they're not optimal, then things don't work as well. So that's a very sort of cut-and-dried, simplistic possible explanation of what's going on in dopamine and ADHD.

... If I have ADHD, what do I have in my brain? A deficiency of dopamine? A small cerebellum?

The leading theory is probably that, effectively, your brain doesn't save its dopamine and use it as well as it might. Now, that's not been conclusively shown. There are two studies in adults with ADHD that suggest that that's true in the basal ganglia. And because they use radiation, it's going to be very difficult to repeat those studies in children. But it is sort of exciting to have that new information.

In fact, in one of the studies, they took individuals who had never been treated for their ADHD, tested them, and then gave them Ritalin and re-tested them--and found that the Ritalin, in fact, did affect the molecule that they were trying to have an effect on. And the result of that should be an increase in dopamine in the basal ganglia--we know in the basal ganglia, because we can measure it. We don't know what's going on in the frontal lobes; we don't know what's going on in the cerebellum, because there's much less dopamine there, and so it's much more difficult to measure.

So it's only a partial answer. If you're an adult with ADHD, I can make some tentative conclusions. But if you're a child with ADHD, I don't yet know what's going on in your brain.

Why is it that you may know more about adults than children? Is the brain not yet developed? And when is it fully developed?

No. There are some studies that can't be done in children, because the regulations that we work under require that we limit the level of risk to negligible levels, especially if we're going to be studying healthy children as control subjects. So we have to think of other ways to get indirect information.

But the brain is constantly evolving... . And it changes dramatically from childhood, to puberty, to adulthood. Can't you outgrow ADHD?

The manifestations of ADHD change with age. And so I think some individuals can be said to outgrow their ADHD. It used to be believed that all kids with ADHD would outgrow it. And now we also know that that's not true. For some people, they continued to have difficulties, whether it's in high school, or college, or at work, or with relationships. But it's not everyone. A significant proportion get better . We don't know what the proportion is, because we'd need a very large study, and we'd need to follow people over a long period of time. ...

I hear you say that there are a lot of limitations in what you know. Why are there so many limitations?

Well, there are limitations because we've only just started. Until about 10 years ago, we couldn't really quantify or measure anything in the brain, especially in children. And so we had to rely on impressions. If you go back and look at the journals that were written in the 1960s, 1970s, 1980s, some of them were almost like novels. They're very rich with detail and texture, but have very few numbers. And until you could start counting things, or measuring them, it's hard to really make progress. Once we had magnetic resonance imaging and other sorts of techniques like that, we're able to start measuring and studying the brain.

... When I get discouraged, I think about a map that I have in my house that came from an exhibit in the Smithsonian from 1500. It's a map of the New World, and you can tell that they knew about Florida--there's a peninsula. They have an idea about the East Coast, and even Texas, but you wouldn't want to use that to drive from Tallahassee to Atlanta, because it's just too crude. So we're about at that level of getting an idea of where the major pieces are, and what their relationships may be. And it's better than it used to be. But we've only been at it for about 10 or 12 years. ...

... Give me one true fact about ADHD.

The posterior inferior vermis of the cerebellum is smaller in ADHD. I think that that is a true fact. It's taken about five years to convince myself that that's the case. That's about as much as I know--that I'm confident about.

What does that mean?

I don't know what it means, but it's true, and it's a fact. And that's the next step.

So show me a bit about how it works in the brain.

... Even though it's only about 10 percent of the size of the total brain, there are more neurons in the cerebellum than in the entire rest of the brain combined, which is fascinating. But the cerebellum has never been thought to be that important, because you can remove it, and nothing terrible happens. But recently, people have noticed that the cerebellum is very involved.

One person says that the cerebellum is a little bit like a co-processor--it's useful, but it's not necessary. And he believes that its function is really to help the other parts of the brain work better. That makes some sense to me, that maybe that's what's not working so well in ADHD. People with ADHD can do anything; they just don't do it quite so well. It's a disorder of efficiency, or inefficiency, as much as anything, I believe. I'm excited about the possibility that this is an important clue to understanding this. But how to build on that is something that I'm still struggling with. ...

What is a question that you're trying to answer in your research?

Well, I'm trying to answer probably too many different questions. One question is whether this difference in the cerebella vermis is the most important difference, and if that somehow has effects in other brain regions, and therefore produces a kind of downstream effect of ADHD.

I'm also involved in collecting samples for genetic analysis, to see if we can find what genes increase the risk that someone's going to have ADHD. We know that it's not likely to be a simple association between a gene and the disorder. But likely, almost certainly, genes do influence how hyperactive, how impulsive, how inattentive someone is. And that's a long-term kind of project. I'd like to think of some newer ways of imaging the brain in children, to see if we can see the functioning of the brain. But those are perhaps a year or two away.

So how close are you to finding a biological marker for ADHD?

I don't know. I don't think we'll know until we've found it. ... We'd like to find a biological marker. We'd like to find some sort of objective way, something that gives us a real sense that we understand what's going on in ADHD. The problem is, we're searching in the dark, and don't know where that clue is going to be. My personal opinion is that we'll trip over it in the next three to five years. ...

A lot of people say Ritalin is "kiddie cocaine." What's the truth?

... A very respected researcher wrote an article that was entitled, "How is Ritalin, or Methylphenidate, Like Cocaine?" The researcher wrote a paper in which she examined the similarities between Methylphenidate, or Ritalin, and cocaine. They go to some of the same places in the brain; but there are also some major differences. Cocaine leaves the basal ganglia much more rapidly. She also studied the differences between injecting Ritalin and swallowing Ritalin, and it makes a huge difference in the kind of response you get. So, Ritalin, when it's taken as a pill, has a very safe profile. When it's injected or when it's snorted, it becomes a very dangerous drug.

So I think that that's the truth in this. And it's not something to play with. There have been deaths from kids who thought it was fun to take Ritalin recreationally. But when it's taken as prescribed, it's remarkably safe. There's always a risk. But there are fewer adverse events than from vitamins with iron, which are quite toxic when children take too many of them, or aspirin, or Tylenol, or antibiotics.

We're prescribing this powerful substance to kids, and yet we don't quite know what's happening.

There are two parts to that. When you work with these children and their families, there's a real level of distress and unhappiness that makes us want to do something to help. Over 60 years ago, a man named Charles Bradley found out by accident that medicines like Ritalin or Dexedrine have profound effects on the behavior of children or adults that have these kinds of problems. It was completely by accident. He was hoping that this would help their headaches. It didn't help the headaches, but the teachers were just amazed at how these children--who were locked up because they were not safe to be outside an institution--were all of a sudden able to sit in class and learn math.

And so, part of the explanation is that there are some very unhappy children and parents. And for many of them, these medications reduce that unhappiness--not for all--but for many.

The other part of the question--our level of ignorance--is true. But we're pretty ignorant about almost everything. And if we only used treatments that we understand comprehensively, we wouldn't do very much, and people would be the worse off. The attempt is to learn from our ignorance, by systematic trials, by advancing what we know. And we have advanced--appreciably--which isn't to say that we know enough. But we know quite a bit. The opinion of the vast majority of physicians in this country, and in several other countries . . . is that the balance of benefit and risk is positive enough for most cases to make this worth trying.

Even if we don't know the long-term consequences of using these medications?

We don't know long-term consequences of many things. And sometimes, the long-term consequences of not doing something have to be weighed as well. The best way to learn the long-term consequences of a treatment ... is to do an impossible study--to take 1,000 children and randomly decide who's going to get Ritalin and who's going to get placebo for the next 20 years, and not let their parents change their minds about what they're going to do. That's not going to be done. ...

We also know, from the fact that millions of children have taken these medicines, that the risks are not dramatic or obvious, because those are things that people notice. We can't be glib or certain that there are no long-term risks. But there's no large mass of doubts amassing and suggesting that we have a generation of children developing cancer or things of that type.

But it is an open question. And when the decision is made whether or not to use medications in the child, the uncertainty about that needs to be acknowledged. For some people, that weighs more than the potential benefit; and for other people, the distress is more important. But that's a personal decision that, I think, in most cases, parents have to make with recommendations and advice from a physician. I don't think that that's an easy call, but uncertainty is a constant against all medicine. ...

So what is your research about?

We use the MRI scanner to measure the brain, and to measure regions of the brain. We're very interested in the basal ganglia, the cerebellum, because those are regions that we think are important, and also, they're measurable. We've been doing this for almost 10 years. We've had hundreds of kids who've come through two, three, four times. And one of our goals is to develop growth curves for the brain, like pediatricians have had for height and weight, so that we can see what's the expected pattern of growth.

And I see you're measuring girls.

Several years ago, we started a study of ADHD in girls, because it's always been thought to be almost exclusively in boys. It turns out that a lot of girls have it and no one ever notices. ...

And do girls have as much hyperactivity as boys do?

It depends how you compare it. Some girls can be very hyperactive. But depending on the age, some girls will be much more hyperactive than girls who don't have ADHD, but they're usually less hyperactive than hyperactive boys of the same age. The problems with concentrating on schoolwork, which we normally describe as inattention, can certainly be present to the same degree. ...

How big is your sample size? And is it big enough?

It's never as big as you want it to be. ... We've scanned 187 individual children who have ADHD over the last 10 years. Many of those children have had more than one scan, so we have over 300 to 400 scans. We're talking about both boys and girls of different ages.

... The brain of a 6-year-old is not exactly like the brain of a 12-year-old, or an 18-year-old. We'd like to have a few more. ...

What are you hoping to find? What's your thesis here?

Well, our first question was, if ADHD is related to the brain, what parts of the brain are different? We have some initial answers to that. The next question is, what happens to those brain regions as kids get older? And our preliminary answers seem to be that they develop fairly normally, at the same rate as they do in our control kids. But if they were smaller to begin with, they stay somewhat smaller. We're still doing measurements in some brain regions. So we may find other things, but we don't have any other results yet.

So, in normal versus ADHD, what is it that you see?

We compare 50 or 60 kids who are healthy to 50 or 60 kids who have ADHD. We see that the posterior inferior vermis, which is the small part of the cerebellum, is smaller by about 12 percent in the kids who have ADHD. We also find that the caudate is about 6 percent smaller. Other findings are more tenuous, but those are our two main findings.

The critics would say, "So Ritalin shrinks the brain."

... A number of years ago, we knew that we needed to check to see whether or not the changes that we observed, or the differences we have observed, were because of ADHD, or because the kids were taking medicines like Ritalin, or Dexedrine, or Adderall.

We can't do a study that puts kids on medicine, and puts other kids on placebo, and then follows them. Instead, we've been recruiting children who were never medicated, and then we follow them over time. We compare them to kids who have been on medicine. Within the next year, we think we'll have that completed. Our preliminary findings are that it's not related to the medication. But we're not finished getting the scans or measuring them.

How do the medications work on the brain?

We know the first step of how medicines like the stimulants work, but we don't know many other things. We know that they increase dopamine and norepinephrine, which are important neurochemicals, in regions where those chemicals are being released normally. We know that they enhance the amount that's available to those neurons. But we don't know if that is important in all of the brain regions that have those chemicals, or if they're interacting in more complicated ways. We just flat-out don't know.

... Why use a stimulant, and why not a tranquilizer? People have a hard time understanding that.

... The first person who found out that stimulants can be helpful for hyperactive children was Charles Bradley. And he guessed, or hypothesized, that they must be stimulating some of the centers that allow inhibition and self-control more than they stimulate other parts of the brain. That was in 1937, and that's still a pretty good explanation of what we know. So we haven't progressed as much as we'd like. We know that they do work in many kids.

... Are you uncomfortable with how little you know about this?

Yes. But I'm also glad that I know more than I used to. It's both half full and half empty. It's not even half full. It's about a tenth full. ...

Six thousand studies, and 160 double-blind control placebos. Why is there a controversy about ADHD?

It reaches into some core issues. The idea that not all children are born perfect is a very hard one to deal with. The question about the long-term risks, I think, is an important one. And then there are historical reasons for people to be worried about what psychiatrists would propose in some cases. . . .

If you had a child with ADHD, would you give him medication?

If I had a child that had ADHD, with what I know, I'd be willing to have that child take medication, but I'd want to make sure that it was really necessary. And I would put it off as long as I could; I would not be comfortable medicating a child who was two, three, four years of age, unless it was the only option. ...

Is there something you'd like to say that I haven't quite addressed?

I think it's important to look at the real issues. One of the real issues is that this can be a complicated process to sort out. And if physicians or clinicians don't have the time ... if someone's not willing to pay for that and to say, "This is an important thing," then there are going to be cases where kids fall through the cracks, or instant diagnoses are made that may be incomplete, or incorrect. ...