The President's Council on Bioethics click here to skip navigation



Session 1: Pediatric Psychopharmacology

Steven E. Hyman, MD, Provost of Harvard University;
Professor of Neurobiology, Harvard Medical School;
and former Director, US National Institutes of Mental Health

DR. HYMAN:  Thank you very much for inviting me.  As you and I discussed, we want to have an overview of some very difficult issues, such as how we set diagnostic boundaries in psychiatry in general.  There are even more difficulties, as you know, in children where the science is less advanced. 

I prepared a Power Point presentation.  This has become a necessary crutch for all of us for about 30 or 35 minutes.  I can't possibly address all of the wide-ranging issues that we need to think about in order to grapple with these topics, but would be happy afterwards as we engage in discussion to go very, very far from what's in the Power Point.  Also, I think to be maximally useful, I'm happy to be interrupted during this Power Point discussion if something is opaque, or alternatively if you already have thought about an issue, and feel that time would be better spent elsewhere. 

I note that I wear my intellectual biases on my sleeve.  If you can even judge by the back drop of my title slide, you can see that there's a brain there, so we have a number of different questions when it comes to the use of psychotropic drugs in children.  As the absolute medical groundwork, we have to ask whether particular drugs, Methylphenidate, Prozac-like drugs, anti-psychotic drugs are safe and effective.  And we have to ask it for different age groups, because brain development means a fortiori that children are not simply small adults.

Secondly, we have to worry, effective for what?  When we're testing the utility of a drug, are we looking, as we most often do, for diagnosable disorders, or are we thinking as we do increasingly in general medicine about early intervention or even prevention?  We have to ask whether we treat risk states for psychiatric disorders in the same way as we do for heart disease.  So, for example, I wonder if you had planned to have an ethical panel on the early and widespread use of the statin drugs to lower cholesterol.  And if not, I think it's not a question beneath contempt, why we consider drugs to treat the diseases of the mind differently than we treat drugs to prevent general medical disorders.

We have to think about unknown, of course it's a great conundrum on how we would learn about long term unwanted drug effects.  We'll come back to this.  And then we also have to ask a question that is so often forgotten, which is how does an untreated mental disorder, or even - and here I think we get to some of the issues you're grappling with - milder impairment affect the developing brain, and how does it affect real-life outcomes?

Unfortunately, we do not have adequate empirical data on almost any of these outstanding questions.  There's some things we -- I just would like to assert.  I have a few picture slides afterwards, and I don't want to get into scientific detail, but I think it's very important that we not fall into the ordinary trap of thinking that drugs are about the brain, and lived experience is about the mind, and they do not mechanistically interact.

In fact, both psychotropic drugs and lived experience affect behavior, both brain, physical substrate and behavior, both short-term and long-term.  And both psychotropic drugs and lived experience alter the brain probably by remodeling of synapses, which is something that I know that those of you who are not scientists don't think about every day.  This is really just a picture to help me make a few points.  This is a picture.  It's a drug experiment.  It's from Terry Robinson at the University of Michigan, and I just thought it would be worth your seeing physically what I mean by this.

On the screen in front of you are two dendrites, the receptive processes of nerve cells, and these are a kind of neuron aptly named medium spiny neurons because they have dendritic spines.  And we think that the communicative connections of different neurons make these connections or synapses on these spines.  And what you can see is after a certain drug treatment, just comparing the left and the right, that the architecture of these spines is different.

Now there's a lot we don't know.  Are synaptic connections really made?  Is information processing really changed?  Is this rat thinking and behaving differently in a way that correlates with these synaptic changes.  But I think what's really important to recognize, again just to grapple with this mind/body distinction, is that even beginning 20 years ago, Bill Greenough, a psychologist at the University of Illinois, began comparing rats raised in the normal laboratory shoebox cage with rats raised in what might be called a rat park, you know, with interesting toys and interesting food, not just Purina Rat Chow, but also chocolate chips and so forth.  And when he looked at dendritic complexity and cortical thickness, they were really quite different in the rats raised in this enriched versus the shoebox cage environment.

The point is that there is no fundamental mechanistic physical difference using drugs to affect the brain and lived experience.  In fact, drugs act on mechanisms that are present for normal lived experience in the brain.  Now, of course, the potency of the effects and precise effects will vary enormously, but there is no categorical difference.  I think that's really quite important to recognize.

Now one of the things I understand you're grappling with are the boundaries of psychiatric disorders, so if I can switch from that assertion to the issue of psychiatric diagnosis.  This will be fast.  I misspelled the title a bit, but -- and we can come back to this.  Psychiatric diagnosis today, whether it's for ADHD or depression or anything, is made based on operationalized criteria derived from symptoms and signs.  There are no objective laboratory tests for any mental disorders that we normally think about them.  There are some appearing, so narcolepsy, which could be considered a mental disorder, has been discovered to be related to defective expression of a particular gene, the orexin gene or its receptor, and there will be tests for this.  But in general, we are talking about the scoring of symptoms and signs by human observers.

In the DSM-IV, the current American Diagnostic Manual, reliability is king, not validity.  Now let me tell you what this is.  Reliability is the notion that two independent observers will arrive at the same diagnosis.  This is not a trivial feat.  Prior to the DSM-III, published in 1978, psychiatry had the Tower of Babel problem-- different people could not understand each other.  People had idiosyncratic diagnostic criteria.  The putative prevalence of schizophrenia in the United States was said to be twice what it was in Great Britain.  None of this was credible, and we needed shared diagnostic criteria that would be simple enough, and good enough to permit two observers to arrive at the same diagnosis.

This is not the same as validity.  This is -- I have another slide later which mentions this, but this is -- validity is about picking out natural kinds, carving nature at the joints, and the two are obviously related to each other but are not the same thing.

Now the fact that the DSM has only tenuous claims to validity, although not zero claim, really reflects the early stages of the sciences of brain and behavior.  And also, the difficulties of the genetics of behavior.  Now at the same time that we have these difficulties, we still have a need to make treatment decisions and to communicate with our patients, so we've come up with what really has to be considered a provisional diagnostic schema.  The problem is that that diagnostic schema has become reified in a way that isn't warranted by the science, but is all too predictable.

Now I'm just going to show you two pictures again, because to say that the science is difficult and that it's early is just -- it's such an important point to make.  And to say that it's difficult and early doesn't mean we won't ultimately understand the biological basis of mental disorders.  And again remember, the biological basis means the integration of genetic experience and lived experience, and other environmental factors integrated by the functioning of the brain.  So this is a recent paper from Judy Rappaport and colleagues showing over time the loss of  gray matter, of cells in the cerebral cortex of children who have an early onset of schizophrenia.  And what you can see in these pseudocolor scales -- where  always in pseudocolor red is bad -- comparing a normal adolescent with an individual with schizophrenia, that there has been loss of gray matter, and here you can actually, by scanning the same person at intervals, you can look five years later and actually map the loss of gray matter.

I'm showing you this because the technology to map, find differences, small differences with reliability in the brain, even to parcelate different brain regions is really a technology of the last years of the 20th Century, and it's still under development.  People like Thomas Szaz, who were true enemies of psychiatric diagnosis, were mistaking the difficulty and the earliness of the science for its impossibility.

Now we don't have answers.  We can't turn these pictures of regression of gray matter in the cortex of children with schizophrenia into clinical diagnostic criteria.  Maybe this will never be a criterion, and if it is, maybe it will take a decade.  The point is that we are really at the birth of the kinds of neuroscience technologies, and as you'll see, genetic technologies that can help us get some handle on the underlying biology of mental disorders.

Okay.  Let's turn to genetics, because this is also very, very important in understanding the boundaries of what are called psychiatric disorders.  And here, the glass is both half-empty and half-full.  Family studies with appropriate designs, meaning properly designed studies comparing monozygotic and dizygotic twins, that is twins that share 100 percent of their DNA versus twins that share about 50 percent of their DNA.  Studies that have been performed of children adopted out of their biological families early in life and so forth do not fully corroborate the current DSM-IV diagnosis, which is to say these diagnoses, based on symptoms and signs, have chosen certain symptoms and signs that experts felt could be reliably scored, and have decided that this cluster equals a disease. 

We can then go to experiments in nature, to genetics and ask whether the symptom clusters co-segregate; that is, whether they are transmitted as a group, as a single entity from generation to generation.  And what we find when we look very carefully is that there are, you know, again not complete invalidation, but not complete validation of the way we do things, that symptoms and signs mix and match in a rather uncomfortably loose way often.

Now there's also reassurance.  Despite these problems, despite the fact that for, for example, many families with manic depressive illness, you'll find that they have psychotic symptoms, delusions and hallucinations, and the delusions and hallucinations are not passed down exactly together with the mood disorder, for example.  Even though we have situations like that, there is reassurance.  If our categories were arbitrary, if they were completely fantastical chimeras, we wouldn't see any significant passage of these traits from generation to generation.  They would not cohere.  And in fact, what family and genetic studies show us, is that the categories we have, as problematic as they are, are picking out something real, so let me make you ersatz geneticists just for a few moments.

A useful if imperfect measure of heredity, because it doesn't perfectly separate genes and environment, but it goes something like this.  The likelihood of expressing a trait if something is genetic should increase as a person shares a higher percentage of DNA with someone else who expresses that trait.  That means that if you share 100 percent of your DNA with somebody and they have schizophrenia, you should have a higher likelihood of having schizophrenia or ADHD, or depression, or diabetes than if you share about 50 percent of DNA with somebody, if they are your dizygotic twin, your sibling, your parent.  And there should be correspondingly less association with second degree relatives.

And when we, using this crude but useful measure, when we look at what are called recurrence risk ratios, what you can see is, and let me explain this.  And I think after this, we're done with the most technical slides, but I do think some detail, rather than bald assertion, is important here.  What you're seeing here, this recurrence risk ratio is as follows.  Your increased risk of having a, in this case disorder, given that in the first column you have a sibling with a disorder, and then in the second column you have an identical twin with the disorder, where the ratio is your increased risk over the population base rate.  And the higher the risk ratio given a, in this case a monozygotic twin or a first degree relative with the disorder, the higher the ratio, the more genetic the disorder.   And this kind of information is used by people using the tools of genetics, even to pick what diseases they want to work on.

So let's look at schizophrenia.  We've been talking about it.  The population base rate of schizophrenia is 1 percent.  If you have an identical twin with schizophrenia, you have about a 50-fold increased risk over the population, so in this case about a 50 percent, 48 percent risk overall of having schizophrenia.  If you have an ordinary sibling, on average you'll share about half of your DNA, you have a 9-fold increased risk.  So I want to compare this with Type II or Adult Onset Diabetes, which is widely understood to be a genetic disease, and you can see that genes have a lot more to say about schizophrenia than they do about Type II Diabetes.  And even with our problematic classification of ADHD, it has recurrence risk ratios that look very much like Type II Diabetes.  So again, these are not ridiculous chimeras, as imperfect as they are.  They are picking out something that is transmitted as disease risk across generations.

Now some of you who remember genetics would say ah-ha, but those ratios - I should actually go back - like 9 and 48 are not what Gregor Mendel would have taught us; that is, if -- I don't want to get into the details, but rather say this, and we can come back to it in discussion.  What the studies tell us overall is that genes play a substantial role in, I've shown you only pathology on these slides, but also in normal behavioral variation.  And this entails another whole set of debates, but I would like to forestall them with the second bullet here, which is that behavioral variation, including psychopathology, is genetically complex.  That is, when we say in common parlance that something is genetic, what most people have is a deterministic notion that there is a single gene for Attention Deficit Hyperactivity Disorder, or schizophrenia, or Diabetes Mellitus, or criminal tendencies, when in fact, what genetic complexity means is that, as we're going to see, this idea, this simple idea is really of little which is relevant to some rare single gene disorders like Huntington's Disease or Cystic Fibrosis, is of little relevance to behavior, and actually the most common human illnesses.

Where it is currently thought that many different genes, many different places in the genome or loci interact to produce risk, and that they interact equally with, or in some ratio with environmental factors, and chance.  Chance is often forgotten in this.  You can't wire up a hundred trillion synapses in the brain in a mechanistic and reliable way.  There are a lot of stochastic effects in this, and that these interactions are non-linear.  Genotypes help select the salience of environmental experiences, and those affect gene expression, so these are extremely complex interactions.

The other thing which is very important in thinking about disease, something like depression, depression which affects 15 percent of the population, these are not likely to be deleterious mutations in the genome, the truncation of an important protein, something important not being made.  Instead, these are more likely to be just variants, that in some combinations are neutral, in others are advantageous, and only in certain infelicitous combinations and in  interaction with environmental factors and chance do they produce disease.

So let's just think about this again.  There's some evolutionary speculation, and it is speculation about, you know, Diabetes Mellitus, that the thrifty genes, the metabolically thrifty genes that would have helped you store fat before the neolithic revolution are not a good thing to have in the era of McDonald's.  And these are not harmful mutations, these are just particular variants that are interacting, and give you a particular metabolic situation.  And similarly, for behavioral variants, including mental disorders. 

Now, of course, certain combinations of these variants can give you the risk of very severe illness.  I mean, there's no doubt that even if Autism is a group of variants, that depending on which ones you get set you up for being somewhere on a spectrum of being a little bit aloof, which might help you to be a great mathematician, versus somebody who can't communicate at all with other humans, that still doesn't make these things mutations.  It means that these -- again, this is an interaction of genetic variants.  And let me just say again, we think this is right.  I mean, there's a lot of data that this is right, but this is -- we're right here on the edge of genomics and genetics. 

Okay.  Now I just also want to point out for our later discussions, that the boundaries of mental disorders are not the only disorders that are going to be redefined once we understand better the genetics and the genomics.  So let me show you a picture of a breast cancer that looked the same to pathologists. So how does one make a diagnosis of a cancer?  Well, a tumor is removed by a surgeon, and still today in most cases, a pathologist will put an acidic dye, and a basic dye on the tissue specimen, a 19th Century technology.  And the pathologist is very learned in comparing patterns of -- seen under a microscope with certain clinical outcomes.  And the pathologist will say this is a certain kind of lymphoma or breast cancer.

Well, here is a gene chip, and the investigators here basically arrayed 5,000 different human genes.  And they asked whether a gene was on being expressed in a cell, or not on, not being expressed in a cell.  And then they asked the computer just to array patterns of gene expression in 98 different breast tumors which are arrayed on the left to see whether there was any pattern of correlation between the likelihood of metastasis.  And indeed, with this gene chip you can see that what looked for many years, eons, like one disease, was actually at least two diseases.  There are two different patterns of gene expression here shown by red and green.  Green is off, red is on, and these are genes that might be very much involved in invasiveness and metastasis and so on, but the point is that genomics and genetics are redefining the boundaries of disease everywhere.

The difference between this and behavior, frankly, is again, this is an objective test outside the body; whereas, in behavior, despite the early Judy Rappaport picture of this brain of the young person with schizophrenia that I showed you, we are still lacking any objective tests that we can look at outside of the behaving human being.

Well, just to bring this full circle, and to make sure we don't get too focused on the genes, and I've already said this, if genetic complexity were not enough, ultimately it's not our genes but our brain that regulates our behavior.  And that while genes are very important, neural circuits are shaped by gene environment interactions over a life time.  This is the idea of synaptic remodeling by lived experience.  Thus, behavioral traits reflect the interaction of multiple genetic, environmental and stochastic factors.

Okay.  So it's very complicated, but we shouldn't despair.  Our genetic studies to date, for example, show us that we have some coherence.  I'm getting pretty close to the end of this.  The other thing, and this makes your job harder as you ponder the questions that you are, is that depression, ADHD, autism, and many other conditions appear to be quantitative or dimensional traits. 

Now the current DSM actually says, you know, to have depression you need to have five out of nine DSM-IV symptoms of depression, and you have to have those symptoms every day for two weeks.  Well, that's really quite arbitrary.  In fact, of course, the boundaries between -- really getting the boundaries right between normal and diseased phenotypes have enormous impact on treatment decisions.  And the truth is that even though these are handled relatively arbitrarily in the DSM-IV, medicine is very comfortable dealing with quantitative traits like hypertension.

Basically, how did we decide that 120/80 is terrific, and 140/90 is a disease and warrants treatment?  Well, we did long-term follow-up studies, and it turned out that if you had 140/90, there seems to be a bit of a discontinuity, actually, and you have a clear increased risk of strokes and myocardial infarctions.  And whereas, the 5 out of 9 DSM-IV criteria for major depression just felt right to a committee, they haven't been exposed to that kind of empirical test, to understand what is a risk state and so forth.  The other thing is that we understand that there is early, and transient, and mild, and moderate, and severe hypertension.  And again, in the DSM we have no capacity to do that.

Okay.  Now this is a problem, this issue of not having rational diagnostic thresholds defining either risk states or actual pathology, has -- does a disservice, especially to children.  It creates a lot of lack of clarity with respect to when one intervenes in children, both in the over-treatment domain, and in the under-treatment domain.  In the DSM approach, you know, you only achieve it if you have 5 of 9 DSM-IV criteria.  If that were applied to cardiology, we might limit our diagnoses to angina and heart attacks, and not to early coronary disease or even elevated cholesterol.

Okay.  The complexity of mental disorders is humbling at the levels of genetics, neurobiology and behavior.  But then again, many illnesses in general medicine are not so simple.  I showed you a cancer example.  It's just true for all common disorders that these are not unitary disorders, but have very complex boundaries.  The difference is again, that in general medical disorders there are objective tests that get you within the pathophysiologic family, and we have to live with this.  We can't tell people in distress to come back when our science is more advanced.  We have to do as well as we can.

Okay.  So now in the last five minutes, just to give us a -- put this to work, so Attention Deficit Hyperactivity Disorder, you already know this.  I assumed you've discussed these things.  It's defined as age inappropriate inattention, impulsivity and hyperactivity.  It's critical that the diagnosis requires symptoms and disability in multiple settings.  Ideally, the workup, therefore, involves looking at the child, interviewing the child, and also the parent and the teacher.  If symptoms only occur in Mrs. Smith's classroom but not at home or on the playground, that's not ADHD.  It is a clinical diagnosis.  There is still no objective test.

Even with these diagnostic criteria, when they are well applied by experts with a proper workup, they are associated, it is a high risk state.  It is associated with bad outcomes.  It's associated with academic and occupational under-achievement compared with abilities otherwise measured.  There is an increased risk for substance abuse which, by the way, is decreased with Ritalin treatment, even though Ritalin is abusable.  And there is an increased risk for arrest.

Anecdotes, you know, one takes with a big grain of salt, but when I was NIMH Director, some of our ADHD research was co-funded by the Justice Department, because there was such an over-representation of kids with untreated ADHD who are incarcerated.

It is also associated with an increased risk of other disorders, depression, anxiety and conduct disorder.  It's the most common behavioral disturbance that results in a clinical referral.  Thirty to 50 percent of kids who are referred clinically for a psychiatric illness have this diagnosis.  It has to start before the age of 7.  And as you already know, it runs in families and it's likely that some of the risk is genetic.

Treatment has been very well studied.  Stimulant drugs in particular have been around for a time.  We know that behavior therapy is effective.  We know that stimulant medication, Methylphenidate is effective.  We know that from the very extensive MTA trial which was conducted in 1999 through the NIMH, I have -- it was initiated before I arrived, so I don't have a sense of ownership, that careful use of medication is more effective than behavior therapy.

Behavior therapy is interesting.  It doesn't generalize across different contexts, so behavior therapy in school doesn't necessarily generalize to home or playground.  Behavior therapy in the home doesn't generalize to school. Combining medications and behavior therapy is cost effective only for children, not uncommon, but only for children with co-occurring disorders; that is, ADHD plus anxiety or ADHD plus depression.

We also know that community treatment of ADHD is not as good as it should be; that is, there are lots of kids who are on Ritalin but don't have good outcomes, their doses are wrong, they don't know about side effects, they're not being monitored.  A certain number of children who are said to have ADHD in our communities and are being treated with stimulants also do not meet criteria for ADHD.  Adrian Angold in 2000 had a paper that suggested - he's at Duke - that 50 percent of children carrying this diagnosis wouldn't really meet standardized criteria.

There's enormous variability with some communities treating it far less than the epidemiology suggests.  The epidemiology again -- so now you know how, you know, problematic these criteria are, but if you apply them, and apply them rigorously, the suggestion is that 3 to 5 percent of boys have ADHD, but we see rates of treatment often close to zero among non-caucasian minorities and some inner cities to the best documented.  There was a paper looking at several suburban Virginia counties - it's apt that we're in Virginia - where 20 percent of the boys in some counties were getting Ritalin, and anecdotally in some schools, especially among middle class or upper middle class caucasian males, 30 or 40 percent of the boys may be on Ritalin.  That's not rigorous, but even rigorous epidemiology shows this disparity between a 3 to 5 percent incidence, and 20 percent.

There is also an unexplained increase in stimulants and other psychotropic drug use.  I know that the Julie Zito studies have alarmed many people.  I would caution you though, the studies are very difficult to interpret since drug prescription data is divorced from good diagnostic workup, so we just really have no idea what's going on with these kids.

The bottom line, there's a mismatch between children in need of treatment and children getting a diagnosis, both under-treatment and over-treatment.  We need better diagnostic methods.  We need to understand the best use of existing treatment in all age groups, and we need better treatments.  I'll skip over where we need to do research.  So to conclude, how do we take this -- how do we think about these things?  Why did I give you the first 20 or 25 minutes on genes, and brain, and behavior and the complexity of diagnosis?

Well, partly we have to remember that while we're worried about the risks of drugs on the brain, you also have to remember that for childhood mental disorders there may also be risks of no treatment.  You have to ask how easy is it to recover from persistent problems, in well diagnosed kids with school, and parents, and peers, especially if the symptoms persist.  There may be a downward spiral for depression, anxiety, ADHD in which a kid has symptoms, aggregates with deviant peers, gets in trouble with the law and so forth, and these may leave their permanent traces on brain and behavior, and life trajectories.

We live in a world of diagnostic gray zones, as Dr. Kass said right at the outset.  And increasingly in medicine, we have an ethic of prevention and early intervention.  We're told that the time to treat -- that Osteoporosis is actually a pediatric disease that expresses itself in old age, you know, you're building your bone density as a child.  Is there a moral difference between lowering cholesterol levels and altering neurotransmitter levels?  Why might it be okay for everyone to be on a statin but not an SSRI stimulant or modafinil so that we can stay awake and work all night?

In treatment there's always a balance between risks and benefits, and the balance is influenced by severity of symptoms and stage of illness.  Difficulties related to children with psychological symptoms.  For most conditions -- see, if we had this answer it would be easy, but we do not know enough to project trajectories with or without treatment.

There is abundant evidence that appropriate treatment decisions, however, are not currently being made with ADHD in particular.  We could also talk about depression, because of outside pressures to treat coming from family, school, drug advertising and so forth.  Inadequately trained physicians.  There are very few child psychiatrists, and pediatric neurologists, and behavioral pediatricians in our communities.  Almost all prescribing is done by family physicians who are working in earnest, but don't have the training, and they don't have adequate time or reimbursement for full work-ups and follow-up.  A full work-up to demonstrate ADHD would take several hours, or would require talking to several people.  And also, medications are seen in many health plans as cheaper and easier than behavioral treatments.    

And when we go beyond well-defined disorders and high risk states toward early intervention and prevention, the balance of medical risk side effects shifts.  Right?  It shifts against treatment.  There are unknown long term effects of drugs.  Now I've talked about unknown effects of going untreated, but now there are unknown long term effects of drugs on a child's symptomatic trajectory.  Does giving drugs now mean that a child might need drugs later that they might not otherwise have needed?  Well, we don't know.

There are symbolic messages to children about self-efficacy.  Behavioral control comes from a bottle.  We have the problem of anabolic steroids for the soul.  Can we really separate health and prevention issues from performance enhancement issues?  I think that's a very, very difficult line to draw.  And then there's the issue of social coercion or unilateral disarmament. I find it working at the University an anathema that all of our kids, well many of our kids feel the need to get coaching for their SATs, and they get their resumes spruced up, and one couldn't possibly be a serious professional football player, I would imagine, without the use of some performance enhancing drugs.  And one might take an ethical position, but taking that position puts one at a competitive disadvantage.  And in the case of ADHD in schools, and maybe one day Prozac in the work place, there might be a sense that you'll get the message that to do well, you know, you shouldn't be the only one not being treated.

So let me end there.  You can see that you have -- well, actually as you know, you've gotten yourself into an area of extraordinary difficulty because of the intrinsic complex nature of psychiatric disorders, because the science of understanding and diagnosis is young, because despite that, we've empirically discovered treatments that have efficacy and a low side-effect burden, because if I'm right and these disorders really are the left-tail of a bell curve and not something discontinuous, these effective treatments will work not only for those severely effected, but those less severely effected, and those who wouldn't even receive a diagnosis.  And you're addressing this at a time when we really don't understand fully how sets of behavioral symptoms portend a certain trajectory for children.  Thank you.

CHAIRMAN KASS:  Thank you very much for a very clear, comprehensive presentation.  The floor is open for discussion.  Elizabeth Blackburn.

PROF. BLACKBURN:  With relation to your table with the risk ratios, you gave some very precise numbers, and then you talked about the degrees and the quantitative aspects of diagnoses.  And I was curious about those numbers, did they come from the extreme very, you know, say 9 out of 9 symptoms for depression, for example, category?

DR. HYMAN:  Yes.

PROF. BLACKBURN:  Because I could see --

DR. HYMAN:  What we did basically, and it's interesting.  The genesis of this slide was when I was  NIMH Director, and we were about to make an enormous investment in genetics, we were working out -- that is, which diseases warranted an early attack on the genetics given the complexity as genomic approaches were maturing.  And it was quite clear that autism, schizophrenia, and manic depressive illness, because of their very high risk recurrence ratios, were apt targets for genetic studies, and actually other things were a lower priority.  And the way we went about it is basically to do a search and find every credible well-designed, adequate "N" study that had been done comparing identical twins, dizygotic twins and siblings.  Per force they were done in different countries using different diagnostic criteria, and if we did this in a more technical meeting and you saw error bars, you would be somewhat alarmed, but in the end, I think there's a general consensus that these numbers are not far off, representing the heritabilities.

PROF. BLACKBURN:  Yes.  No, I take the point about the relative contributions of heritability, but I was curious, for example, bipolar, you know, 7-fold for sibs, 60-fold risk ratio for monozygotics.  But then if depression, which is only part --

DR. HYMAN:  Yeah.

PROF. BLACKBURN:  I mean bipolar is only part of that, 15 percent so clearly there must have been a subset --

DR. HYMAN:  Oh, bipolar --

PROF. BLACKBURN:  -- how broadly you define these things.

DR. HYMAN:  I'm sorry.  Okay.  So depression, when we -- so if you've ever had a manic episode you're not in the unipolar depression category, you're in the bipolar category.

PROF. BLACKBURN:  Yes, but from what you're saying, since you took studies from multiple countries --

DR. HYMAN:  Yes.

PROF. BLACKBURN:  -- it must have been a fairly broad set of criteria that were included then in these things.

DR. HYMAN:  Yes.

PROF. BLACKBURN:  I was curious about whether this represented, you know, as you say, the right hand side of the bell curve for these diseases.

DR. HYMAN:  Right.

PROF. BLACKBURN:  It sounds from what you're saying as though it's a fairly broad set of definitions.

DR. HYMAN:  Yeah.  I would have to -- I mean, for autism it's not so broad.  For major depression, it is broad.  We were dependent on the world's existing literature, but I take your -- your point is -- if your point is that the input data has certain infirmities, it certainly does.

PROF. BLACKBURN:  No, I wasn't --

DR. HYMAN:  Yeah.

PROF. BLACKBURN:  I think the point is well-taken, and I certainly think there's every reason to think this.  But I was just curious about how broadly  some of these things were defined.  Were they looking at the extreme ones where you could be very, very clear.

DR. HYMAN:  I think, yes.

PROF. BLACKBURN:  Or did they segue into them --

DR. HYMAN:  So for autism, bipolar disorder and schizophrenia, they're very, very clear.  I think for major depression there is no clarity. 

PROF. BLACKBURN:  Okay.  Thank you very much.

CHAIRMAN KASS:  Rebecca Dresser.

PROF. DRESSER:  You were very balanced and fair in your presentation, and I wondered if I could get you to express a point of view on something that's relevant, not just to this project on "Beyond Therapy", but also another project we have, which is regulation.

You seem to say that in some situations that diagnostic criteria are fairly defensible, but there's still this problem of getting professionals to apply them stringently, or even with reasonable rigor.  And also, dealing with parental demand, I suppose sometimes patient demand.  I wondered if you had any ideas on what might be done to address that problem?

DR. HYMAN:  I have no problem expressing a direct opinion on this topic.  I think that in medicine in general, in psychiatry in particular, and in pediatric mental disorders most particularly, there is -- as early as our criteria are, if they were well-used, and if children were followed-up appropriately to ensure that the first diagnostic hypotheses and treatment suggestions were optimal, it would be a much better world.

I think the data suggests, as my slide listed, that most prescriptions, as far as we can tell, for psychotropic drugs are made by family physicians, well-meaning to be sure.  I do not want to bash family physicians.  They are working at an extreme disadvantage.  They have been untrained.  They have no time.  I mean, it's fine for me to say that you should spend several hours on a diagnosis, and one of the things you want to do is make sure not only that the child is in trouble in the classroom with the teacher, but also is being rejected by peers on the playground, so it's fine for me to say that.  The family physician is not trained to know what questions to ask, very often has 11 minutes to get from the beginning of the intervention to writing a scrip, and would not be reimbursed to take the time after the visit to make those phone calls.

As a result, and in combination with outside pressures that exist in certain school systems, and through advertising, I think that what we -- there is a yawning gap between what we know, which is imperfect but good enough in a rough-and-ready way and what we do. 

I would also say that while I focused on diagnosis, I think the follow-up issues are as or more problematic; that is, if the medicine isn't working and has side effects, it may or may not be stopped.  Kids who really need it may not be complying, and nobody notices, so I think there is an enormous problem of medical practice.

Let me say one other thing about coercion, and let me give you a thought experiment, because I think it's -- coercion already carries with it negative connotations, you know, the evil empire forcing a child to take the medication.  We can certainly imagine a situation.  Why you don't risk a unilateral disarmament situation, where you have a few symptoms, and you're not performing so well, and all of your friends are taking this drug, and their parents and the teachers are happy consumers.  That's not an ideal situation.  I don't think anyone could defend that.

On the other hand, there's a case where there's 25 kids in a classroom or 28, and one teacher, and there are two kids who have symptoms of severe behavioral disorder, who absorb almost all of the teacher's time.  And any of you who have children or have been in a classroom know that this is a reality, and so the 23 kids don't really get much of an education.

Now the issue here is that schools should never be making diagnoses, and schools should never be saying, as has been rumored, you know, get Johnny on Ritalin.  That's absolutely wrong.   But for schools to demand that somebody be worked-up or treated in a certain way so that everybody could learn I think is a different issue.  And I think when we think about the roles of schools, you have to think not only about the dark 1984ish scenario, but also about the second scenario, and weigh those.

CHAIRMAN KASS:  Do you want to follow-up, please?

PROF. DRESSER:  This is rather provocative.  I was involved in a group of pediatricians addressing the growth hormone question, and one idea which didn't really go very far was that only pediatric endocrinologists should be able to prescribe because of this desire to do a good work-up, and keep the boundaries.  Has anybody ever proposed something like that in this area?

DR. HYMAN:  It is absolutely infeasible.  The number of -- I mean, in most -- if you think of inner cities or most of rural America, the number of trained pediatric psychopharmacologists, behavioral neurologists, or behavioral pediatricians tend towards zero, and so the burden of doing this correctly must be on the family physician. 

The pediatric growth hormone issue is, as you know, exactly analysis.  Short stature, but my child does have short stature.  He was going to be a center for an NBA team, you know, and I'm drawing that line.

CHAIRMAN KASS:  If I -- Rebecca when she first started to ask about the -- whether you had some suggestions, in fact, about the regulation of this practice, you've gone and indicated why it is in need of such attention, but in effect said this is a problem for professional practice.  And almost all of the pressures make it almost impossible for it be done.  Is that the best that we can do?

DR. HYMAN:  I hope not.  On the other hand, regulation that flies in the face of reality is -- decreases respect for regulation and fails.

CHAIRMAN KASS:  Of course.

DR. HYMAN:  You know, I would -- I have long believed -- so I've told you I believe that family doctors are going to carry the burden of most of this.  I think certainly that at least in certification requirements, there has to be increased attention among family physicians to pediatric psychopharmacology.  I mean, if you look at the numbers of young people receiving Ritalin or an SSRI, this must make up a very substantial part of general  pediatric and family physicians practice, and I think some kind of training is necessary.

I must say in my years of looking at the gap, as I call it, between what we know and what we do, and physician behavior, I've become not quite despondent about the power of education by itself, and I think that some kind of accountability to make sure that these -- that what is being taught is being practiced in some way is very, very important.  But that accountability, if it just accrues to the physician who doesn't have time or resources, and is not reimbursed, becomes very, very problematic, so I think it has to be -- it really has to look at health systems in some sense, and really look at the quality of the overall work-ups that are being performed and reimbursed within a system.


DR. FOSTER:  You used the term that on a "rough-and-ready" basis if the things that we know were followed, we would be in a much better world, the sense being that although there might not be a bright line between disease and --

DR. HYMAN:  There isn't.

DR. FOSTER:  There isn't and so it's broad.

DR. HYMAN:  There is not, yes.

DR. FOSTER:  At least there would be some broad category of defense about prevention and so forth.  And you talked earlier a little bit about a curve that went on into normality, where there might still be an enhancement from excitatory drugs and so forth.  I just wondered if you -- I think I know what you would say about this, but merging out of the issue of children with - as an example here - what is your opinion about the enhancement virtues that kids in college and so forth, you know, are using these drugs extensively and so forth?  Is that defensible or not defensible in your view?

DR. HYMAN:  I really struggle enormously with this.  It really comes down -- I'm not going to give you a crisp answer because I'm not done struggling.  It comes down to the anabolic steroids issue in some sense.  Athletics has decided that you are supposed to -- I mean, even though we know that illicit and problematic use of drugs continues in sports at all levels, tragically sometimes leading to death, the sports community will assert that it wants human beings to compete with each other in athletics, only based on their bodies and the physical training that they undergo.  And that the use of drugs is just not to be part of an athletic contest.  And in some sense, if you -- that idea, which is so -- always breached, or often breached, that idea which is so attractive, really doesn't translate very easily to life as it is truly lived.  And here my impulses, my libertarian impulses, and impulses as a physician begin to take over.

If somebody feels distressed and doesn't meet any criteria, and finally a physician says well, I don't really know what's going on with you, but let's try this thing.  It's marketed as an anti-depressant, but since everything is really probably on a bell curve, you know, it doesn't stop working if you don't have 5 out of 9 DSM-IV symptoms.  And the person has no side effects, or model tolerable side effects, and really it relieves their distress, and they're a better husband or wife, and they're functioning better at work.  It seems very difficult in America to say that they can't feel better.  I think it's actually impossible to say that.  But then that scenario shades into the dystopic notion that but if everybody in the work place is -- if these drugs get better and better, and everybody in the work place is on these drugs, might we not have a coercive situation where somebody  who wanted to engage life without a psychopharmacologic agent is now being in some sense forced to join the crowd, which is the situation if you want to an interior lineman in the NFL.  You might have moral questions about anabolic steroids, but you would be ill-advised if that's your chosen career to avoid them, so I can't give you a crisp answer.

I think it's really -- if you think about the case, the first case I gave you -- I mean, I've given you extreme cases, but in the first case it would be very hard to say that somebody couldn't have the drug, and in the second case one begins to really imagine some fairly nightmarish scenarios.  And I think finding -- if there is some way of finding an appropriate middle ground, I imagine that's what you are all engaged in, and I can't make it easy.

You might think my interim solution while I struggle is not strong enough, but my interim solution is really that physicians, because these are prescription drugs, physicians really have to be a lot better.  They can't say -- somebody can't say I've read an ad for -- you know, I've been shy all my life, and I read an ad on a side of a bus, and I want you to prescribe X.  I mean, the physician really has to understand the person's symptoms, and treat a prescription as an empirical trial, and not a birthright, and stop it if it's not working, and weigh risks and benefits.  And maybe it doesn't make you happy as an answer, but we're not even there, we're not nearly there.

CHAIRMAN KASS:  Could I pursue this with you unless, Dan, you want to follow-up?  Are you okay?  Yeah.  In the presentation, on the one hand at the very beginning, you make clear your view that both lived experience and drugs affect the brain.

DR. HYMAN:  That's correct.

CHAIRMAN KASS:  And that there is no fundamental categorical difference.

DR. HYMAN:  At the level of the brain.

CHAIRMAN KASS:  At the level of the brain.

DR. HYMAN:  Right.

CHAIRMAN KASS:  Yet toward the end when you talked about the kinds of problems that might -- and among the list of problems about the use of these drugs was the concern that you would be sending symbolic messages about self-efficacy to children.

DR. HYMAN:  Uh-huh. 

CHAIRMAN KASS:  That somehow drugs were a better way to deal with their difficulties.

DR. HYMAN:  Uh-huh. 

CHAIRMAN KASS:  And then this question about the anabolic steroids for the soul which we talked about.

DR. HYMAN:  Right.

CHAIRMAN KASS:  If you bracket the question of social coercion that might come if you were dealing with competitive situations, and simply talked about  the fact that look, we don't know whether there's a diagnosis here.  In fact, it's not clear this is a medical condition, but there's some kind of self-discontent.  And you've already pointed out that in the absence of certain kinds of treatment, the possible plasticity of the brain might, in fact, be under-developed and not used.

CHAIRMAN KASS:  Why wouldn't you say that -- why don't these concerns sort of vanish, concerns at the end if you begin to think about the thought at the beginning; namely, that we could use pharmacological agents to help lots of people feel somehow better about themselves, in the wake of which feeling better, all kinds of other experiences might go better.  The brain might, in fact, develop and become richer.  Isn't this just sort of priggish concerns that you have?

DR. HYMAN:  Yes.  No, no.  No, no.  Well, I think that's why I began with athletics, where society has made a decision that you're supposed to be competing naked of pharmacology.  And we haven't made that decision when it comes to the rest of life.  WE're uneasy.  I mean, we are generally uneasy.  And the whole reason I struggle is because I can't refute that final position that you took.

Let me also say at the beginning, there are also certain kinds of experience that can have bad effects on life trajectory presumably mediated by the brain, by long-term changes in the brain.  I mean, we're worried about pathologic gambling, you know.  That's a lived experience.  That's not a pharmacologic experience, and it can have very profound effect on brain function.  I was really just making the point that we can't -- you shouldn't rest your deliberations on the idea that there's something fundamentally categorically different between drugs and experience and their effect on the organism.  There are other issues that we are struggling with, which has to do with symbolism, with relative moral weights, and of course, with medical side effects, risks and benefits.

CHAIRMAN KASS:  But bracket the side effects.

DR. HYMAN:  Yes.  Right.

CHAIRMAN KASS:  Why isn't, in a way, the philosophical teaching about the brain sort of at odds with the desire to privilege the symbolic meaning of do it for yourself without drugs? 

DR. HYMAN:  Yes.

CHAIRMAN KASS:  Bracketing your -- no one is going to ban anything here.

DR. HYMAN:  Right.

CHAIRMAN KASS:  Bracket libertarian concerns.

DR. HYMAN:  Right.

CHAIRMAN KASS:  And just talk about the kind of anthropological question, moral self-understanding.

DR. HYMAN:  Well, I think that there is a symbolic difference that we don't -- that we often don't recognize that plays itself out, often in irrational ways in other aspects of medical practice.  For example, we are quite well aware that most people will not control their cholesterol with diet and exercise over the long run.  And yet, every medical textbook tells you to start with diet and exercise, and force somebody to fail through this exercise in self-control before you would prescribe a statin, even though the statins are, you know, they're -- we can bracket side effects, but they have side effects.  And I think what we'r talking about here is what kind of society -- I think what we're engaged in is what kind of society we want to have, and whether we, as a society, want to recognize as valid, unstigmatized, and in no way diminishing of someone's humanity, that they gain appropriate treatment even early, or preventive treatment for identifiable high risk states, but that in general as a society, we would like people to work with their own native abilities, with their own struggles.

The alternative view is that if we can find medications which will enhance our performance, lengthen our life, decrease the stress, do we really think that Aldous Huxley was right and we'll end up, you know, in some "Brave New World", or is it more likely that we will end up in a happier, healthier world where people can all function well?  And I think we don't know the answer to that, and what you and -- what we are all here together struggling with is we're worrying about the Huxley and dystopic, but it would be hard to say that we should not be aiming for the more utopian version of this.  And in fact, if such drugs were to be developed, no regulation in the world would keep them from general use.

CHAIRMAN KASS:  Michael Sandel.

PROF. SANDEL:  I'd like to continue along Leon's lines.  What he's been doing has been -- you gave us a very balanced and elegant, and terrifically, for me, informative overview of these issues, and you have an instinct that's emerged in the discussion for which you call the appropriate middle ground, but Leon, and now I am trying to drag you into the fray that we've been occupying.

DR. HYMAN:  Uh-huh. 

PROF. SANDEL:  Which is partly an ethical and ideological fray, as it intersects with the medical and scientific account that you've given.

DR. HYMAN:  Absolutely right.  I mean, medical practices -- there are a lot of problems in medical practice that have to be addressed, but we're talking about something very separate from medical practice now.


DR. HYMAN:  So you're not letting me hide.  Yes.

PROF. SANDEL:  Okay.  So if I could continue along these lines.  First, when you gave your general, your opening account of the mind, calling into question the sharp metaphysical mind/body distinction, both drugs and experience have effects on the brain, to some ears, to some people listening to this, see that finding, if it's true, as threatening to something they believe about ethics and about freedom, and about moral responsibility.  I don't think that they're correct in seeing this as a threat -- 

DR. HYMAN:  Correct. 

PROF. SANDEL:  -- to proper understandings of freedom and moral responsibility, but there's a powerful philosophical tradition that supports their worry that this would be a threat.

DR. HYMAN:  Right.  I agree.

PROF. SANDEL:  But in any case, to come closer to the surface of these ethical and ideological issues, so you signalled -- those were fighting words your account, your brief account about drugs and experience having effects on the brain.  And then you reinforced them, and you posed as a question if statin, why not stimulants?

DR. HYMAN:  Right.

PROF. SANDEL:  What's in principle the difference between the use of the two for prevention and early intervention?Maybe we stigmatize the second more than the first, but is that defensible in the light of this earlier thing, so that gets closer.  And then -- so to take one step further along the lines of Leon's questions, go back to the sports case.  Now you dodged that by saying well, we as a society have decided that we want sports athletes to play just with their bodies and physical training.

DR. HYMAN:  Right, but you could imagine a different --

PROF. SANDEL:  Well, first of all, it's not clear that that's true sociologically, and even if it were, it might be mistaken.

DR. HYMAN:  Right.  Right.

PROF. SANDEL:  And one could raise questions about improved running shoes, or graphite tennis racquets and so on, which have nothing to do with bodies or with physical training, but which are external in just the way that drugs would be external.

DR. HYMAN:  Uh-huh. 

PROF. SANDEL:  You don't object to that, so at the level of sociology, it's not so clear that we've accepted that.  In the Sports Illustrated expose of the use of steroids that they pointed out that players now speak to each other in the language of "playing naked", you're not going to play naked, are you?

DR. HYMAN:  Yes.  That was my unilateral disarmament.


DR. HYMAN: Right.

PROF. SANDEL:  But here's the issue.  Suppose we put aside questions of safety, we find something that doesn't have bad side effects as steroids do.

DR. HYMAN:  Uh-huh. 

PROF. SANDEL:  And we put aside the question of fairness, that by making it available to everyone if they want to, to use it let's say in sports, some drug that will enhance performance, that doesn't pose medical risk, and that's equally available on a voluntary basis to anyone who wants to use it, so you remove the fairness.

Now you may say well, there's some people who want to be play naked, they'll be effectively coerced, but if it's safe, then it's no more coercion in that direction than it would be in the other direction if you said people who want to use it, can't use it.

DR. HYMAN:  Right.

PROF. SANDEL:  So it would be a draw.  So the coercion is gone, the fairness is gone, the safety objection is gone -- something objectionable about enhancing performance through a drug rather than through, let's say, a more rigorous training or better genetic luck.  That's question one.

The second question is the parallel to that in the case of Ritalin, in the case of stimulants.  And here, there's been a lot of discussion about over-treatment, and over-treatment reflecting an ideology of the, you know, the diagnostic manual, this ideological --

DR. HYMAN:  Right.

PROF. SANDEL:  It's led, or permitted and over-treatment.

DR. HYMAN:  And there's also under-treatment.

PROF. SANDEL:  Well, but you --

DR. HYMAN:  As are more problems --

PROF. SANDEL:  Well, that's a different emphasis you brought from the ones we've had in previous discussions.  You said well, we have to look at the risks of not treating, we have to look at the incidence of under-treatment.  And if it's safe, and let's say even hypothetically to isolate the issue, if it doesn't have adverse health side effects, then the under-treatment would be worse than the over-treatment.

DR. HYMAN:  Uh-huh. 

PROF. SANDEL:  So the question then, the first question is this -- the hypothetical in the sports case, and the analogy in the Ritalin case.  If we do away hypothetically with the medical risk, is there anything objectionable to letting any kid whose parents let's say agree to have Ritalin to improve behavior or concentration on the SAT, provided it's available to everyone, and assuming for the sake of argument it weren't risky.

DR. HYMAN:  So you'll force me to take off my medical hat that makes me worry, and to address in very naked terms where I would come down.  I would say that if there were drugs that could enhance performance, that were perfectly safe, but I want to define safe on my terms. 

The key for me is, and this is very unlike the soma of "Brave New World", there must be no clouding of consciousness, there must be no alteration of -- no artificial control of overall moods, but rather -- in short, we don't want people getting high all the time.  I would include that in a risk.  It's very important for me to segregate, you know, artificial elevation of moods and change in human judgments into the category of side effects.  But if we had drugs that were free of that, I would think that it would be -- they would be acceptable.  I think that --  

DR. GAZZANIGA:  We do.  That's Ritalin.

DR. HYMAN:  Well, not quite Ritalin.  I mean, Ritalin is not -- well, let me come back to the second point.  The reason that you've found me so worried about this in talking about the moral symbolism is I also believe that it is better for humans -- we never want somebody to have to struggle against difficulties with some rectifiable disorder, defect or disability.  We want to bring everybody up to a certain normal level of function, with the understanding that that's a very gray zone.  But it's also important -- I would argue that a society which -- if a society errs in the direction of increased personal responsibility and sense of moral agency, that's a better society than a society which errs in the opposite direction.  And that while I said yes ultimately to your first question, the way you set it up, the way that such drugs would be used, the way Ritalin can be used, worries me because of the risk of undercutting a broad societal sense of moral agency and responsibility.

I don't think that -- and the two positions are not really fully congruent.  Right?  Because I know that one of the risks of having risk-free performance enhancers is potentially to undercut a set of moral agency and responsibility.  And yet, it is very hard for me to find a bright line where if these things were truly risk-free, didn't make people -- give people artificial emotions and impair their human judgments, that I could say that we can't have.  We will have.  What I'm saying is we have to learn to manage them, and we have to learn to manage them in a way that minimizes the loss of a sense of moral agency.

DR. FOSTER:  I just want to interrupt for a second to say, I'm not sure at all about the issue of safety of Ritalin in young children.  I mean, I'm not sure that it's a safe drug anywhere, but particularly I don't think that the neurological circuits and everything are completely, you know, at 7 years of age you're not worried about that.  You remember that --  if you look at something like alcohol and pregnancy, I mean four hours of alcohol at 200 milligrams percent you've got a defective brain.

DR. HYMAN:  You know, Dr. Foster, I don't want to cut you off, but time -- I'm going to have to go exactly at 10:30.

DR. FOSTER:  All right.

DR. HYMAN:  Let me just say, if you actually look at the data, imperfect as it is in any clinical trial, the data would tell you that properly used by any criteria that we have, Ritalin is a safe drug.  Can it be misused?  Yes.  Can it be abused?  Yes.  Do all drugs we have today have side effects?  Yes.  But compared with almost actually another drug we have that we use in psychopharmacology, oddly the data that we have suggests that Ritalin is the safest.  Now you and I might argue about this, but I don't think that's where these questions are going.

CHAIRMAN KASS:  I have Gil, Paul, Bill.  You have how much time?

DR. HYMAN:  I have to leave at 10:30, about five minutes.

CHAIRMAN KASS:  Okay.  Well, let's try to --  briefly then.

PROF. MEILAENDER:  Yeah.  I'm confused.  If I follow what you said to Leon, and then to Michael, it would seem that you're not prepared to draw a line against enhancing performance for perfectly normal people.

DR. HYMAN:  Right.

PROF. MEILAENDER:  Enhancing performance so long as it doesn't have bad side effects, one of which would be altering of moods. 

DR. HYMAN:  Artificial elevation.

PROF. MEILAENDER:  Yes. So somehow a person who just, you know, is not clinically depressed or anything, but just goes through life a little discontented and would like to be happier, has less claim to be helped than someone who without help can get a 1400 SAT and would really like to hit 1550.  If I'm understanding your right, I just don't see why anybody would draw that line.

DR. HYMAN:  Would draw the line.

PROF. MEILAENDER:  Yes.  Why do you have more sympathy for the person who wants to notch up the SAT than the person who would like to be a little happier.

DR. HYMAN:  I'm sorry.  I don't have more sympathy.  I just --

PROF. MEILAENDER:  Well, you're more prepared to permit them to look with some --

DR. HYMAN:  I must not have expressed myself -- well, I just said I have much more sympathy with the sickest person, and --

PROF. MEILAENDER:  I'm not talking about a sick person, just a person who'd like to be a little happier.

DR. HYMAN:  What I'm saying is, I don't find a rational way to draw the line to say that at some point we can't -- I've been trying to take refuge in good medical practice, and I've been not allowed to do that, and ask whether in principle I would draw a line in which somebody with no diagnosable illness - right - with no diagnosable illness would not be allowed to take a medication.  And what I'm saying is that if you ask me naked of all my protections about medical practice, follow-up, and this and that, I would have to say I can't find that line right now, but that I'm worried about the impact of that on the sense of moral agency and personal responsibility.


DR. McHUGH:  I'm sorry you have to leave, Steve, because we could go on for a long time talking about your presentation, and how thorough it was.  But I want to come, if you would, just briefly to talk about the foundations of your opinions here today as expressed.  I know you have other opinions, and they turn on this idea that we're dealing with reality out there, and we have to make accounts of reality.  But, Steve, you know perfectly well that those realities of our making, particularly from the psychiatric side.  And that DSM-III-R, IV, and IV-TR are based on a particular approach to things that had a reason 25 years ago, to try to get us to talk, but now has its own deep problems, very deep problems.

DR. HYMAN:  Right.

DR. McHUGH:  Including the expansion of psychiatric disorder so that it's now a huge, huge thing.

DR. HYMAN:  Right.

DR. McHUGH:  The belief in appropriate treatments now being offered for various forms of conditions that are placed in there, certain forms of PTSD, social phobia, multiple personality disorder, and all of that rests upon this idea that a top-down approach of checking off a checklist is the right way to diagnosis.  With your hope that ultimately we'll find a validation of that from --

DR. HYMAN: A different set of --

DR. McHUGH:  Yeah, entirely different.  Now since that reality is of our making, that's what makes for the 11 minutes is all you, and that you can't teach people appropriate approaches to psychiatry, I have two questions.  What are we going to do about getting a psychiatric approach to diagnosis that approximates medicine?  DSM-IV does not approximate anything like medicine in ICD-9 and 10, as you know that, so that's the first question. 

And the second question is, do we understand development, maturation and acculturation for it's psychosocial tasks adequately, and its responsibilities for the development of the child right now to be able -- for most children to be able to intervene with medications?  You said that drugs and life experience do the same thing to the brain.  And you know I agree with that.  They certainly change the synapses.  But we also know that both drugs and live experiences can do terrible things in the process of building --

DR. HYMAN:  Right.

DR. McHUGH:  So question one, where are you going with DSM-IV to DSM-V?  Secondly, do we understand anything to be able to satisfy the gentlemen on this side, and ladies on this side, to say that we know how to build a child sufficiently well to make the gains on your SAT scores adequate to the things you lose in depriving them of the opportunity to play more, to have different friends.

DR. HYMAN:  Right.  Okay.  So this is -- right, these were not exactly yes or no questions.

DR. McHUGH:  No.

DR. HYMAN:  You know because you've read my criticisms of current psychiatric diagnostic nomenclature, that I find enormous problems, and again, in one minute, we have a difference between trying to coalesce around and name important conditions that we really see that people have, like manic depressive illness or schizophrenia, versus the exigencies of a profession that wants to be reimbursed for its work and is forced by, you know, the reimbursement system to give a lot of things names, and we know that there is a lot of -- this is not theoretically neutral or apolitical, but at the same time, I have a certain amount of sympathy in my better moments given the -- the real difficulties of doing better given the state of our science, so we have to change, but I don't have any easy prescription for how to change.  And I'm very conservative about blowing up an existing system.  I'd like to fight its excesses.  You and I would both like to fights its excesses, but I don't want to blow it up until I have something really that -- until our genetics is farther along, our neuro imaging is farther along, because we'll go back to Babel.

The other issue is exactly my problem, and Professor Sandel here smoked me out, that I am living -- on the one hand I can't find, you know, a bright line that if we really had perfect drugs which didn't create, you know, the 1980 -- I'm sorry, the "Brave New World" scenario.  I could find a bright line to outlaw them, but at the same time, my -- as I said my precise concern is the messages that we deliver in terms of human self-efficacy and moral responsibility.  And if you can find a way to manage the reality that will dawn on us, there is just no doubt that minimizes the undercutting of a sense of human agency and moral responsibility, without unfairly stigmatizing those who are truly in need of treatment.  You have my blessing and best wishes.  Unfortunately, I have to be off --

DR. McHUGH:  If I could just reply to that just to say that maybe the thing that we need is not more neuroscience at some level and better psychiatry at some level, both to develop a classification and to answer some of these questions about what maturation itself is doing to individuals, and that's where my problem.

CHAIRMAN KASS:  Dr. Hyman has to return to teach a class. It's heroic of him to have come on a teaching day to spend time with us.  Thank you enormously for a wonderful session.  We're adjourned for 15 minutes.


(The session then went off the record for a break.)


CHAIRMAN KASS:  This is a session on the topic of "Beyond Therapy: Ageless Bodies?"  Before we start into that, at least a couple of the people who had their hands up when time ran out have spoken to me about a possibility of at least putting into the record certain kinds of questions or concern, not so much to interrogate Dr. Hyman, who unfortunately had to leave, but if people would briefly like to simply put into the discussion the questions or concerns that they had, since that was a conversation that was just about to take off, I think it would be appropriate if we allowed room for that.  I had on my list Bill May, and Bill Hurlbut, and I don't know if there was anyone else in the queue.  Bill, do you want to add something?

DR. MAY:  Well, I felt that Dr. Hyman talked about the gap between what we know and what we do, which tends to lead to too much over-treatment, and too much under-treatment.  But earlier he really talked about the gap in science really, it's a young science we're talking about.  And also, there's a second gap between what we know now, and what we would like to know to feel comfortable about what we do.  And both of those gaps seems to lead in the direction of saying no regulations, no bans, but it makes you very dependent upon a guardian class in the interval before you narrow the gap between what we know now, and what we should know to feel fully confident in what we do.

But what was quite depressing, it seemed to me, is you depend upon the guardian class, but it turns out that our guardian class doesn't have the timing, or the time or the training to do what it ought to do in order to guard.  So no regulations given the gaps with which we live, real dependency upon the clinicians, but we discover the clinicians are woefully under-trained in operating in a system where they have no time, even if they had the training, to do what needs to be done.  Which again leads bioethics out in the direction of systems, institutions and structures, away from some of the issues that we've dealt with.

CHAIRMAN KASS:  Bill Hurlbut.

DR. HURLBUT:  I was going to inquire of Dr. Hyman whether he thinks that he would see ADHD in a hunter/gatherer community, and in the same amount.  And whether it really is in some way an artifact of narrowing our definition of normal, because it seems to me that -- well, if it were a Mendelian trait, we'd define polymorphism as 1 percent of the population.  Here we have 3 to 5 percent.  It seems like well, 3 to 5 percent of the population could have some kind of a deficit, but then the question becomes well, is this just part of the spectrum of human variation in a positive aspect of our society which should not necessarily be treated or narrowed, but given a different educational process and opportunity to develop in its own trajectory.

I think it's interesting that ADHD is kind of a derivative diagnosis based on the educational system. Isn't that where it's first picked up, as inability to sit at a little desk, read little black symbols off of white pages, and not playing out with the activity that normally accompanies childhood for most of human history?  What worries me about this is it also goes deeper than that.  I think the idea that this is a genetic disease, which is -- certainly, there's a strong corollary between monozygotic twins, but that doesn't necessarily mean it's even genetic, of course, because it's -- they share nine months in the womb.  And there are some people that believe that maternal stress during gestation provokes this problem, and that even if it were correlated through generations doesn't prove its genetic because there are now evidences that stress itself is echoed generation after generation.  And what worries me in that is that when you label something genetic, it's much easier to justify the concept that it has a single unitive source as maybe a missing enzyme or something like that, and then justify a medical diagnosis and a pharmacologic intervention, so I think those are worth saying.

And then just two final points.  One is that I disagree with the statement that was made, if I understood it right, that there was no difference between treating somebody who seems to fall below the norm and somebody above.  I think there's an intrinsic difference there, and one case you might argue that it is more like therapy in the sense that it's normalizing, that it's bringing somebody into community.

On the other hand, that's when you treat the left side of the bell curve.  But when you treat the right side of the bell curve, it's intrinsically competitive because it's moving you away from the norm into a realm of superior performance.  I mean, there's a lot more to be said about that, but I think if you think of human community as the ground not just of human strength in sociology, but also human meaning, then there's an intrinsic difference between the treatment of those two sides of the bell curve.

And finally, it seems to me that one of the fundamental dangers in this simplistic notion of a genetic deficit is that the reification of a very complex human phenomenon where we tend to think of complicated human realities as treatable by some kind of magic bullet, when in fact the closer you get to meaningful human existence, the less easy it will be to intervene, because human beings have evolved, or have been created, however you want to see it, to be distanced from determinism, and even simple molecular interventions.  The most meaningful human existence is somehow the comprehensive willed self-governance of our humanizing activities.  And to the degree that we give over easily to notions that there are simple deficit disorders, we have to be very cautious about that because that's a very dangerous assumption in a complicated species like our's.

That's not to say I don't believe there are disorders like that, and I'm not saying this one isn't, but it just strikes me as a very, very important point.  And actually, let me add one final point to that.  Even if you say this is directly related to say a genetic cause, there are historical conditions that suggest that we need to be very, very careful about what we think that cause is.  An example of that would be cleft palate and the relationship with mental retardation, which was long assumed to be correlated.  And then when it got to the point where our medical treatments could go in and do surgical interventions early enough, we realized that that so-called genetic mental retardation was actually a byproduct of the fact that the cleft palate was blocking the eustachian tubes, causing earaches and otitis media, and muffling speech, and therefore, causing the children to not be able to keep up with their peers because they couldn't hear and understand what was going on.  So what looked like a genetic cause of mental retardation, turned out to be just a secondary.

CHAIRMAN KASS:  Okay.  I think there are people who are also in the midst of developing their own thoughts in relation to Dr. Hyman's presentation.  And Paul and others, if you'd be so inclined, a couple of paragraphs leading to some kind of question, we would welcome them at the office, and we can send them on to him and see if we can elicit from him some further elaboration on some of the things of concern to us.

Let me turn to --

DR. FOSTER:  Leon, let -- could I just --

CHAIRMAN KASS:  Please, Dan.

DR. FOSTER:  I want to sort of give a reference for the library that people might have here, because one of the things we talked about in the last hour was the issue, the moral issue of one struggling without the help of drugs and so forth in mental illness, and maybe some of you have seen it, but Leon Rosenberg, who happens to be a close friend of mine, Leon Rosenberg has published his experience with manic depressive illness.  Leon was a Dean at Yale Medical School for 10 years, and now works at Princeton, and Robby probably knows him in molecular biology and so forth.  He published an article called "Brainsick", and it's in Cerebrum, Vol. IV.  I can't remember the pages, last year.  It's one of the most remarkable documents that I have ever read, a confession about a struggle from youth, ending up finally in a suicide attempt in which ultimately then somebody who struggled against this for life, was a great investigator, became the Dean of Yale, and all these things, trying to fight this alone, and eventually succumbed and surrendered to a suicide, requiring ultimately electroconvulsive therapy and Lithium and so forth.  It also is a familial, as he outlines, it's a familial illness, and it simply emphasizes to me the difference between a highly moral struggle to try to do it by one's own in an illness which ultimately requires medical therapy.  And so it's heroic in one sense, and in another sense it's a tragedy.  I mean, not that he didn't achieve everything, but he might have had a happier life if he could have done that.  So I simply wanted, if you haven't seen it.  It's a rather obscure journal, and so -- but I thought it would be an interesting article to put into the archives of the library that the Council is putting together.

CHAIRMAN KASS:  Yeah.  Thank you, Dan.  We'll find it and actually circulate it to members.

DR. FOSTER:  Yeah.  Leon Rosenberg is the author.  It's called "Brainsick" in Cerebrum.  Okay.

  - The President's Council on Bioethics -  
Home Site Map Disclaimers Privacy Notice Accessibility NBAC HHS