Let's begin by talking about the evolution of child psychiatry, from Freud to neuroscience. ... How have you seen that evolution?

Really what we're talking about here is the evolution of psychiatry itself, which has gone through a transformation in the last five decades. I trained in psychiatry in the 1970s, and much of our training was about what was then psychoanalytic theory, with a little bit of theory from Jungian psychology and a few other places.

The essence of it was to try to understand the unconscious by looking at infantile conflicts, exploring your past. And there was a sense in which diagnosis wasn't important, and treatment options were really only going to be played out within which kind of psychotherapy conflict you would follow. So the questions that I was trained to study would be, is this going to be someone that you should see in a short-term psychotherapy to really focus on the most recent issues, or is this someone who should be in three-times-a-week, long-term therapy, to focus on how they understand and how they can overcome childhood conflicts and issues that they've been avoiding throughout their life?

That seems like a long time ago, because the field largely has been swept by understanding that many of these disorders can be thought of in more biological terms; that psychiatric disorders are, at the end of the day, brain disorders.

With the revolution that's taken place in understanding the brain and how the brain works, there's now an opportunity to begin to apply that to think about a, how do we understand all of these very complicated behavioral disorders; and b, how do we treat them? Are we able to now begin to define them in ways that were not available 15, 20, 30 years ago? And are we able, then, to come up with novel kinds of treatments that could make a big difference for people who before would have spent five, six years really focusing on their distant past to be able to, either through medication or through relatively brief interventions, get back to work, take care of their families and move on?

What do you think triggered that change? What was the moment?

Well, there were a lot of factors. The thing to remember is it's still very much in transition. ... The mental health field in terms of the workforce ... is still very much a mixed bag. You've got people who on the one hand may think of themselves as sort of applied pharmacologists who are real experts in medications and how to mix them and how to individualize them and how one would follow through with different kinds of treatments. Then you've got people who say, well, the treatment really is swimming with dolphins or play therapy or finding something we can do with the family. So you've got this remarkable range of treatments that are out there, only some of which have any real scientific basis to them.

One of the things I think that people struggle with the most is that the treatment they are likely to be given may depend much more on who they call and not on what problem they're dealing with. And that's a bit of a change. That's not as true in cancer. It's not quite as true in heart disease, although there's some of that. But in the case of mental disorders, there's still this huge variation in the treatments that people are given, and a lot of it depends not so much on a thorough understanding of these disorders, [but] much more on what it is the therapist is most comfortable in doing.

So it's too many bags of tricks in a way ... and not enough uniformity in the field?

Well, in some sense we're dealing with simplicity on this side of complexity when we need to be at the point of simplicity on the other side of complexity. These are complicated problems. We don't fully understand them. Now, it's true we don't fully understand most forms of cancer, but we're further into that than we are with understanding bipolar disorder, schizophrenia, understanding even autism. These are really complicated. We now think of them as brain disorders. But where in the brain, and how? And even when is not really clear.

I think the next few years are going to be this very exciting, transformative time. Yes, there have been huge changes over the last 30 to 40 years. Certainly the advent of medications, the institution of new diagnostic criteria, the ability to begin to categorize these disorders in different ways, those are important.

But I think the most important changes are just about to happen, and that's where we're going to be able to use genomics, imaging, a much more thorough understanding of how to study cognition and behavior. And we'll be able to, I think, really transform the way we understand these disorders and also the way we treat them.

... I listened to your talk in Pittsburgh at the conference on bipolar disorder, and you spoke about the three decades [in which you broke down the evolution of psychiatry]. The first was the decade of the brain, and it was 1990 to 2000. Can you walk me through what that period was about?

... There's certainly a period before 1990 when much of what we thought about was based on I would say the psychological theories of the first half of the 20th century.

The evolution in neuroscience and in biology could probably be encapsulated as this decade of the brain, which Congress mandated in 1990, and it went from 1990 to 2000, and it was an extraordinary period, just an amazing period of discovery and of beginning to, for the first time, to think about mental events as neural events; that you could think about the life of the mind in terms of the life of the brain.

One implication of that, which happened rather late in that decade, was we began, for the first time, to get serious about thinking about mental disorders as brain disorders. Much of what we were able to do through the availability of neuroimaging, which allowed us to see the brain and the brain activity for the first time, get beyond the black box, was to begin to locate particular areas or particular circuits within the brain that seemed to be associated with features of these very complicated mental -- now brain -- disorders.

We're in a somewhat different decade now. I would call this the decade of discovery. By that I mean that we're in a period, mostly through the wonders of genomics and genomic technology, where we are, for the first time, beginning to understand many of the players at the level of molecules and cells. We're beginning to finally get a sense that this isn't just about serotonin and dopamine and a few other obvious factors in the brain.

The idea that these disorders are somehow a chemical imbalance, that you're a quart low in serotonin with depression or you have too much dopamine in the case of schizophrenia, we're well beyond that. We now have a list of scores of factors that seem to be important, but we don't know quite how, and we don't know exactly where they work within the brain. So this period, I'd say from 2000 to 2010, is going to be this really astonishing era of, for the first time, really laying out this frontier of finding out ... what are the major factors that will be important for these mental/brain disorders.

We still have another big decade in front of us -- and maybe we're just edging into it now -- and that decade is going to be translation. It's translating what we've learned about the brain, translating what we're learning about the molecules and the cells involved in these disorders to, for the first time, come up with truly novel, effective treatments.

Now, it's true, this field has had some pretty good medications for most of the disorders dating back to the 1970s, and in some way, that actually was both a boon and a curse. It was great that we were able to help a lot of people with these illnesses, but the curse was that the medications didn't help everyone.

What it did, I think, was it gave us this false sense of security that we thought that actually we would be able to understand depression and schizophrenia if we could just understand how the medications worked. Turns out, that doesn't really deliver for us.

The medications aren't the answer. They're part of the answer, they are helpful, but they're not cures. So depression is not just the absence of an SSRI [selective serotonin reuptake inhibitor] drug, and schizophrenia is not just the absence of an antipsychotic drug. These are disorders that we really have to understand in a much more careful way.

And I think the pharmacology in some way has taken us off-track a bit. It has given us a good tool, but what we still need are drugs, possibly entirely new types of drugs, and other kinds of treatments. That could be a whole range of everything from psychosocial treatments to other kinds of interventions which are far more effective, and which are based not just on what we currently have, but they're based on trying to really get an understanding of what's wrong here, what we call the pathophysiology -- where is the lesion? What's the problem that causes someone to have hallucinations and delusions, or to have a dysregulated mood? -- and then targeting that with the treatments that are designed accordingly. So that's the next decade, and that's really where much of our effort is going.

... So the wrong question has been asked all along? How did this affect you not getting further ahead in understanding where these lesions were?

It's a great question: Have we been chasing the wrong leads? I wouldn't put it that way. I think the leads we've been chasing have been, in some ways, the easy ones. In science you go after the possible; science really is, in some ways, the art of the possible. You follow the leads that you've got. So if you know that a particular medication is really effective in reducing hallucinations and delusions, it makes some sense to try to understand how does it do that. And if it turns out that that medication blocks a particular brain receptor like the dopamine receptor, gee, that's a really important lead to help you to understand, why was that receptor overactive? Why did it need to be blocked? And is there some connection between that receptor and the actual instigation of hallucinations or delusions? So it's not a necessarily impossible lead to follow, but it's one that we've followed now for 30 years, and it's taken us only so far.

I think now is the point to say, OK, we've got medications that work on that particular target. We've probably gone about as far as we're going to go on that target; let's see what else is involved with this disorder, because there's much more here than just hallucinations and delusions. The medicines we have, as effective as they may be against those symptoms, they really aren't taking care of a lot of the other problems with a disorder like schizophrenia or bipolar illness. Let's see what we can do to actually treat the whole person and really make sure that people with this illness can recover.

And the meds aren't doing that. And that's where we need to really rethink, step back, look at the disorders, understand much more about how they happen, where's the pathophysiology, and then design treatments accordingly.

It's not just the art of the possible; it was also the lack of money and the fact that the pharmaceutical industry was giving the resources for the research that benefited their sales. Is that correct?

No, I wouldn't say that. ... NIMH [National Institute of Mental Health] is the single largest source of funding for research in understanding mental illnesses in the United States, and actually worldwide. The pharmaceutical industry, ... their mission is, at least to some extent, to develop the new generation of medications. The frustration that many of us have had is that that's been much more difficult than any of us thought.

In fact, most of the medications that have been developed over the last 10 years are just really knockoffs of meds that we've had before. What all of us would like to see now is, both through public funding and some private funding, through the industry as well, that we look at the next generation, that we think much more in an innovative track here about what can we do that we haven't done: How can we get to parts of these disorders that haven't been affected [by the medications]?

For us, it's also really important to think about the non-medication treatments, because those are going to be critical. We know that [for] bipolar disorder and schizophrenia and depression, that medication by itself isn't going to be enough. It's probably going to be necessary, but not sufficient in understanding how to get at some of the more complicated aspects that aren't simply going to go away with a pill. ...

Where are you now [in that process]? Do you feel you're coming close to getting rid of the focus on how do these drugs work and moving on to a new research phase?

This is an extraordinary moment in time. I guess people have said that at many points in time when they think about these very complicated problems, but I do think we're at a somewhat different moment right now.

Part of that has to do with where we are in being able to understand the genetic susceptibility to these disorders. These are not going to be a single-gene disorder like sickle-cell anemia or cystic fibrosis. It's going to be far more complicated.

For most of these illnesses, there is a significant genetic susceptibility, a component that sets you up to be vulnerable to developing this disorder, whether it's schizophrenia or bipolar illness or autism. We have not really had the tools to nail that until very recently. By recent I mean the last year. But we now have the ability to look across the entire genome at 20-some thousand genes at probably 2 to 3 million points of variation.

To do that quickly in large numbers of people -- to be able to get a better handle in a comprehensive way about what are the likely candidates, where are the factors within the genome, the points of variation that set you up to have the vulnerability -- again, it's not going to explain everything, but it will give us a start. ... [T]hat will tell us some of the factors that we haven't known about that contribute to getting these illnesses, and it will also help to lay out some pathways that could be important targets for new treatment development.

In Pittsburgh, I sensed you were scolding the audience. You were saying you've got to change the way you've been doing research, and you've got to ask different questions. Was that a right perception?

Well, I'm not sure that I would consider myself a very good scolder, but I like to think of myself as someone who's here to change the landscape. My job is to push and push and push and to remind the research field that the needs are urgent.

If there's one word we use here on a daily basis, it's "urgency;" [to describe] the needs of families for new treatments and for better understanding and for better access to care, ... there's no better word than "urgent." We have to be able to respond to what we're hearing over and over again from families across the country.

There's a real mismatch between what they see as the needs and what we're providing in the services, and they need to know that we're doing everything we can to fix that. So if that's scolding, yes, I plead guilty. But most of all, I do think that someone in my position, which is working for taxpayers -- it's essentially entirely supported by the taxpayer base -- needs to do everything that can be done to deliver what people who are paying our salaries are asking for.

Let's talk about the conundrum of doing research in kids. It's kind of a taboo subject: You do not test children's brains. How do you deal with that, and how do we need to change that?

Tough one. So the question [is] how do we begin to answer some of these complicated questions in kids, particularly when you're thinking about treatment development -- and that's what everybody would see as the end game: We want to cure autism; we want to cure childhood bipolar disorder. What you see over and over again is much of what we're doing is testing treatments that have already been around for a long time in adults.

And why do we do that? We do that because we have, understandably, a very high threshold for risk in children. I suppose my part of pushing in this domain is to say we've actually been willing to lower that risk a bit for children with cancer because we feel that the risk is worth taking for the potential benefit.

... [W]e need to understand that for some of these childhood disorders, for autism and for some of the worst mood disorders in kids, we're taking on a risk by not doing things here, and we need to figure out a way that we can allow ourselves some flexibility that we don't currently have. There is an important discussion to have in how do we begin to develop the ability to do treatment trials and innovation in children, even for therapies that aren't currently accepted and haven't been thoroughly proven in adults.

It may be, as we know in the rest of medicine, that the treatments for kids will look different than the treatments for adults.

So how do we begin conducting clinical trials involving children?

Well, one of the things we'll need to do is really understand the landscape here. We do need to step back and look at what's currently available, what are we currently doing, how do we currently structure, let's say, clinical trials for children, and where are the needs. One of the things we've done is to form this very large-scale practice network called ... CAPTN [Child and Adolescent Psychiatry Trials Network]. It's about 200 sites across the country which work together to begin to address the opportunities to do trials in children, possibly even for medications that haven't been fully developed for adults.

This would be at a practice level very much like what we've done for childhood oncology. In pediatric cancer, most children who end up with a form of cancer become participants in a clinical trial. We haven't been there for psychiatric disorders, but why not? The needs are just as great, and the understanding is, as I said before, just as urgent. We need to really move things along quickly, and this is one way to do it.

People argue about definitions. There are some things that are very well defined, like ADHD, and it doesn't seem like people are arguing about what that is as much. But with bipolar, I sense complete confusion.

For bipolar in adults, I think there's pretty good agreement about what this looks like. For bipolar in children, there is some considerable debate about where are the boundaries. At the mild end, are these just kids who are active? Is this the class clown at the very severe -- is this something other than a mood disorder?

So absolutely there's debate. I should clarify, though, that I think for most psychiatric disorders, we're still in a bit of an area of controversy. Psychiatric diagnosis is still a little bit different than diagnosis in the rest of medicine. We've come quite a ways. Thirty years ago, we really didn't worry so much about diagnosis. Everyone got whatever treatment there was available. The Diagnostic and Statistical Manual [DSM], which was developed most of all in the '80s and '90s, served a very important function. It allowed everybody to speak the same language, and it provided a set of standardized criteria, used now widely for reimbursement, but also used so that clinicians can talk to each other and know that when they say the words "bipolar," it has a reference that everybody can agree to.

It's important to recognize, though, that there's a difference between reliability, which is what this manual gives us, and validity. Validity comes from really understanding the disorders, and we're not there yet. We need to get some of the biology to really play out in a much more rigorous and robust way so that we understand what it is that leads to these behavioral criteria that we call bipolar. The ability to use biomarkers, the ability to perhaps have neuroimaging as a tool -- maybe the genetics will help. I think we're at that point where we're finally able to bring some of the tools to bear, but it hasn't happened yet. And I think we can expect over the next few years that a lot of this confusion and controversy can be settled by some of the tools we've got.

... [W]hat does the DSM say about children?

As with the rest of the manual, the DSM, for children, has given [us] this kind of reliability index. It tells us a set of criteria that we can use. But what you see -- and you've probably already gotten a sense of this if you've spent time with families and kids and clinicians who are struggling with these disorders -- is that very few kids come in in a pure form. As many people have said, it's likely that nature never read the DSM manual; that what you see is most often, particularly in this realm of mood disorders, impulse control problems. Kids will fit into many categories at the same time, and it's probably unlikely that the diagnoses themselves will explain very much of the variants.

This may be a place where we need to step back and ask, is there a better way to do this? Should we be thinking about understanding behavior in children along some kind of continuum or along an axis; that maybe having these categories really isn't going to be very satisfying and isn't going to explain very much for us? There's a pretty lively discussion going on now in the field about ... how do we capture some of the distress that kids are feeling without losing the complexity and without losing the dynamic nature of symptoms in children. ...

When you're looking at a bell curve, you know that there will be some very extreme cases, and I've seen it. But you do go to these schools -- and I'm no clinician and no psychiatrist -- and you come out a little bit perplexed and wondering, is it really necessary to medicate some of this behavior? It's a little bit disruptive for the teacher, it's a little bit disruptive at home. And not knowing what these medications do in the long term, ... why [is it] that medication is a first resort?

I get asked this a lot: Are we overusing meds in kids? I think the answer is probably the same for childhood allergies and asthma and lots of other disorders in kids, that we are both overusing and underusing meds; that there are kids who receive meds who probably either could get by without them or might be able to now stop them, and there are a lot of kids who would benefit who aren't getting treated.

So we haven't found quite the right balance. There is this higher sensitivity to giving meds for behavioral disorders than there is for giving meds for what are considered more medical disorders, and from my perspective that's a false separation. I really believe that these so-called behavioral disorders are medical ultimately; that these kids have a genetic susceptibility; that there's some kind of a change in the way that the brain is processing the information that makes these kids, in a particular environment, unable to cope. We'd love to say, "Well, just fix the environment," or "Just pull yourself up and get tough; walk it off." We don't say that for asthma, hopefully, and I don't think we should say it for mood disorders or for anxiety disorders.

If medications are part of what helps kids cope, why not give them every chance? The tough part comes with recognizing that meds by themselves aren't the answer, that they may be in some cases necessary but not sufficient. And we may not be doing enough on the other side of this, providing them additional kinds of resources and also figuring out, on the very mild side, are there kids who should begin an exercise program, begin other kinds of interventions before one reaches for the nearest pharmaceutical pad? I think most docs, most family members and most people who look at this would agree, absolutely, that you don't reach for medications as the very first thing to think about. ...

Being that the U.S. is sort of so unusual in its prescribing habits compared to other countries, what does that mean about the U.S.? We're further ahead? We're more able to prescribe because we aren't willing to invest in other treatments? What's going on here?

It really depends on the disorder. One of the things that has been interesting in looking at recent prescribing practices -- and by recent I mean the last three or four years -- is there has been a bit of a shift. So what you're seeing is increasing use of atypical antipsychotics in children. You're seeing a somewhat leveling-off or even decreased use of antidepressants. And the psychostimulants which have been around for some time are still being used at about the same level -- and in some cases, in some areas being used more in the States than in some other countries.

... If you look at the child psychiatry workforce, what most strikes you is the variation in what's available. There are about 7,000 child and adolescent psychiatrists in the United States, about 300 new ones being produced every year, coming out of training.

If you do the math, that translates to about seven and a half per 100,000 children and adolescents in the U.S., which is about ... one-fifth to 1/10th of what we actually need, given the burden of illness and the needs for thorough evaluations and careful management of either medication or other kinds of treatments. ... But what's really striking, as you look at it carefully, is it's not just the overall numbers; it's the variation.

So you have states like Massachusetts that have about 17.5 child psychiatrists per 100,000, and then you have rural states like West Virginia where you're about 1.3 or less. If you look at that kind of path, you're actually seeing a drop-off. There are actually fewer [child psychiatrists] as the number of kids is going up, and the demand for services is going up in many of these places. The number of board-certified trained physicians who can deal with these kids who have these very severe disorders is actually going down. So it's a continuing problem.

So you're relying on pediatricians to help children with psychiatric problems, is that [correct]? What problems does that cause?

Yeah. In this country, pediatricians [or family-practice doctors] write most of the prescriptions for medications for children, and for the adults for depression for that matter. They often are the both first and last resource that people see, because in much of this country there just isn't the availability to find a specialist.

It's asking an awful lot of a pediatrician or a family-practice doc to do this. These are complicated problems. Whether they start that way or they evolve that way, usually a kid with a serious emotional disorder is going to be embedded in a family that's really struggling in all kinds of ways, and that takes time to evaluate. It takes time to know what is the best way of helping, and it's a very unusual pediatrician or family-practice doc who's got that kind of time to either do the evaluation or to do the follow-through of what would be needed.

So often what you see is people who are trying to do the very best thing in a difficult situation will use medication to try to at least get a handle on some of the behavioral problems. But what they don't have time to do is to look at any of the underlying dynamics of what the kid and the parents and everybody else may be dealing with. That takes a different kind of approach that most pediatricians don't have an opportunity to really get into.

In the '90s it was sort of the decade of ADHD. Suddenly that was the diagnosis of the moment. There was a 700 percent increase in a matter of years in that diagnosis. Is something similar happening to bipolar, and did ADHD open the door for this somehow?

Are we in the era of bipolar disorder? You would think so from looking at the covers of some national magazines and the number of press articles. It's actually hard to say. We don't have good figures right now that say that there's a real increase. What's not really clear is whether many of the kids who are called bipolar have anything that's related to this very well-studied disorder in adults.

It may be an unfortunate name. It may be better to talk about children with problems with mood and impulse control. Calling it bipolar provides this kind of unfortunate wink to what we used to call manic-depressive disorder in adulthood. It's not clear that people with that adult illness started with what we're now calling bipolar in children. Nor is it clear that the kids who have this disorder are going to grow up to have what we used to call manic-depressive illness in adulthood.

But there is an awful lot of attention given to this right now, and a lot of interest. And there is a sense -- I've heard this over and over again as I travel around and talk to clinicians and families -- that there is an increase; that they're seeing more children, usually between the ages of 8 and 12, who have problems with mood and impulse control. ADHD is certainly part of that. They also have problems with activity and problems with attention.

But they're seeing more of the kind of mood side to this than what they claimed they were seeing a decade ago. Is this a real change? We're doing those kinds of studies to try to get population-based epidemiology to really put numbers on this. But the numbers we have so far don't show it. We don't actually see increases in mood disorders in those carefully constructed population-based studies; what we're hearing are the anecdotes. And what we still need is the science to demonstrate that this is a true increase.

So you think it hasn't really changed over time all that much? It might be media hype?

Well, I'm saying that the absence of evidence is not really the same as the evidence of an absence here. We would need to look at that. We need a study that carefully monitors in the same population, using the same measures, whether there has been a true increase in any of these disorders. The evidence we have thus far doesn't support it. But I don't think we have enough evidence, especially recently, to say that this couldn't be happening. Certainly the anecdotes make you wonder.

You were alluding to the fact that the use of antidepressants is down because of the black-box warnings, [warnings placed directly on a medication's package describing severe side effects and risks associated with it]. How has that impacted the use of perhaps more dangerous medications like the antipsychotic meds?

We are seeing an increase in the use of atypical antipsychotics, especially in children. One thought is that those are being used in place of the antidepressants, which, as you mentioned, have a black-box warning. I worry about that for several reasons.

All the evidence we have is that, at least for the treatment of depression in children, the SSRI drugs, the antidepressants are effective. They're effective, and they are largely safe. There have been very rare and sometimes tragic side effects.

Overall, there's no question that the benefits greatly, greatly, greatly outweigh the risks here. We would say the same thing about a lot of medications. ... There's no risk-free medicine that I know of.

We've now seen a decrease in the use of those, partly because of this concern about very rare but tragic side effects. And docs are looking for what else they can do for families and kids who they feel need something. The atypical antipsychotics have not been adequately studied in children. We don't know about their efficacy. We don't know about their safety.

But we do know from studies that have been done in adults that these medications are associated with high levels of certain kinds of side effects, particularly weight gain and what we call a metabolic syndrome. They increase the risk for diabetes and other metabolic diseases. So we're concerned about the risk of putting children on these drugs when we don't yet have good studies that demonstrate their efficacy. In a sense, what we've done is we've taken a drug, or class of drugs [antidepressants], that has well-known efficacy and fairly well-known but very limited risk and replaced these drugs often with a class of drugs that have unknown efficacy but quite well-known risks. I'm not sure that that's progress.

It seems like a big step forward, and rather ironic. I mean, shouldn't the black box be on the atypicals?

... [T]here's the concern about using atypicals in children because they're being used off-label, with a couple of exceptions. ... Particularly, one atypical antipsychotic is approved for use for children with autism. There are a couple of other places -- Tourette Syndrome and a few others -- where medications have been approved.

I wouldn't say that it's a mistake for people to use medications off-label. You do what you feel will help. What we're arguing for is that people look at the full range of opportunities. I think it's a mistake, actually, to reduce the use of antidepressants, which we know really can be helpful. What we would say for any of these medications is, remember that medication for treating these complicated disorders is part of the answer; it is not the full answer. This is not like treating strep throat, where you write a prescription and say, "Come back in a month if you're not better." In any of these disorders, if medication is going to be part of the treatment, the treatment has to be carefully monitored, and it has to be within an entire treatment plan that looks at everything else that can be done for a child who's really struggling with either thought disorder or with mood dysregulation.

Did you sit in any of the black-box hearings?

I wasn't part of any of the black-box hearings.

And is it a bad idea then?

To have a black box? You know, there are lots of ideas that have unintended consequences. If the consequence of having a black-box label is that people recognize there's a risk that previously they may have overlooked, and if as a result of that risk they use the medication more wisely, it's a very good idea. That's really terrific if we get better monitoring of antidepressant use in 16-year-olds who are depressed and may have suicidal thoughts.

That would be huge progress. What isn't progress is if people say: "Well, because of this label and the possibility that I could be sued, I'm not going to use any of the medications in this class. I'm going to use medications in a different class. Maybe they'll be helpful." But in fact, there's very little evidence that for a 16-year-old who's got suicidal thoughts and who's depressed and withdrawn and can't sleep that the antipsychotic drugs will really be the treatment of choice.

[Are there any misperceptions about mood disorders that you hear a lot?]

... [T]here's a tendency for people to think that children with mental disorders or emotional disorders will just grow out of it; if we just let them be, that somehow these disorders aren't as serious as real medical disorders, and they don't require the same kinds of interventions or the same kind of study.

I think that's really selling all of us short. One of the things that we've learned, it's not only that there's a relatively high prevalence of some of these disorders -- perhaps 5 percent in the general population of kids under 18 -- but when you talk to people in their 20s and their 30s who are really disabled by serious illness like schizophrenia, like bipolar illness, like major depressive disorder, as many as 50 percent of them tell us that, from their perspective, subjectively, the illness, the symptoms began by age 14. Maybe they didn't receive treatment for 10 or 15 years afterward, but they lived for an awfully long time struggling with some really difficult experiences. I think we ought to think about, how do we do better by them? How can we make sure that we're able to help people get treatments that will be useful much earlier in the process?

We're doing much, much better in the case of heart disease. We're getting to people before their heart attack, and often pre-empting or preventing the worst phases of cardiovascular illness. We need to do that. That ought to really be the vision for how we think about these other disorders. It really ought to be that we are able to find out how to predict who's going to get them and we're able to pre-empt them at a much earlier stage. We're there for much of the rest of medicine. That's really got to be where we are as we think about these mental disorders as well.