I finished my residency at New York University in 1986 with absolutely no intention to go into research, but I had an interest in schizophrenia from a clinical perspective. I took a job at a municipal hospital in New York and quickly found myself frustrated by how little we understood about schizophrenia as a field and how little we had to offer patients who had it. So after two years I took a post-doctoral fellowship with Nancy Andreasen at Iowa, which is the reason why I got into your database. I essentially rode her coattails. That’s the truth. She was a relative big shot when I came out to work as a post-doc in 1988, and since then she has risen steadily and I became sort of her second-in-command in her enlarging research operation.

Well, one of the biggest differences between doing clinical work and clinical research, at least for me, was that in clinical research every patient contact you have becomes a piece of data. What we do is try to quantify as many aspects of the patient presentation as we can. One of the real challenges is converting psychiatric presentations into a meaningful database. We’re talking about taking things like perception and cognition, etc., and turning those into a meaningful series of numbers. That’s the real challenge. I think we’ve been able to come up with methods that allow us to meaningfully quantify patient presentations at the symptom level, at the cognitive or neuropsychiatric level and then at the neurobiological level using whatever neurobiological indices we have. Here, that has primarily been neuro-imaging, both functional and structural.

My guess is a large MRI study published in the American Journal of Psychiatry in 1995. (M. Flaum, V.W. Swayze, D.S. O'Leary, W.T.C. Yuh, J.C. Ehrhardt, S.V. Arndt, and N.C. Andreasen. "Effects of diagnosis, laterality and gender on brain morphology in schizophrenia," American Journal of Psychiatry, 152[2]: 704-14, May 1995.) Still to this day there have been tons and tons of structural neuro-imaging studies on schizophrenia, probably over 200 of them. I think ours is the one with the largest sample of patients. And I have personally been very critical of schizophrenia imaging literature because the average sample sizes are extremely small. Given the almost certain heterogeneity of what we call schizophrenia, trying to draw conclusions from sample sizes of 20, 30, or 40 is just dubious science. I think a lot of the lack of reproducibility in the schizophrenia literature from one lab to another and even within labs has to do with that issue. And even our sample wasn’t anything to write home about it. But it was over 100 patients with schizophrenia compared to a similar number of very well and carefully matched controls. I think we had more power than just about anybody else to say things.

At the time, very few people were doing functional imaging studies of schizophrenia, and the findings were appealing. What’s interesting about that paper, and what’s interesting about the work that goes on in the lab and the reason I came to work with Nancy to begin with, is she’s really one of the few who focused as much on clinical phenomena as on various neurobiological techniques that could be used. Instead of just asking questions like, ‘how do people with schizophrenia differ from normals?’ she started asking questions within this broad construct of schizophrenia. What types of signs and symptoms go along with what types of neurobiological markers? In that 1992 paper we found that negative symptoms look different. Patients with prominent negative symptoms were those who had the most hypofrontality, which just means a relative decrease in cerebral blood flow to the frontal cortex. This is in the context of performing cognitive tasks that in normals usually is associated with a relative increase in blood flow to the frontal cortex. On the other hand, the sample size was still small, and even though the tools we were using at the time were state of the art—SPECT, which stands for single-photon emission computed tomography—they were still extremely primitive, considering what we can do now, especially with positron emission tomography and functional MRI.

It drew attention to the thalamus as a potential target of interest in schizophrenia. Even though there had been some very interesting postmortem studies that suggested major abnormalities in the thalamus, there had been no attention to it in the neuro-imaging literature, primarily because nobody could measure it. This was first time we had a method that allowed us to see it. The other interesting thing was that essentially all structural neuro-imaging up until then had involved this very laborious method, usually hand-tracing of structures. That paper was an example of an automated method that allowed us to get around this problem.

If you look at the whole picture, one of the biggest contributions of this lab—and Nancy deserves all the credit for this—is drawing attention to different aspects of symptomology in schizophrenia as core or fundamental. So 20 years ago, and even to this date in a lot of places, when people think of schizophrenia they think of psychotic symptoms like delusions and hallucinations as being core and fundamental and being the target of treatment. Nancy’s work drew attention to other symptoms of schizophrenia, initially negative symptoms. Probably the biggest single contribution is turning attention back to the fundamental place of negative symptoms in schizophrenia. And I say turning attention back, because when you read the initial historic descriptions of what we now call schizophrenia, they highlight the importance of negative symptoms as core and fundamental.

Commonly people divide the symptoms of schizophrenia into two broad categories, positive and negative. Positive includes things like hallucinations, delusions, and grossly agitated or disorganized behavior. These are the things a lot of us think about as craziness. But in fact, if you talk to families of people with schizophrenia, that’s not what accounts for most of the morbidity and illness. What accounts for that really is the kind of decrease in a lot of normal human functions. People with schizophrenia have very marked decrease in emotional experience that’s both subjective and objective. So they look as if they don’t have any emotions often. A lot of times, people with schizophrenia will sit in a chair or lie in a bed day after day, week after week, even year after year, and if you ask them if they’re bored, if they’re very honest they will tell you they’re not. They simply don’t have any drive, any volition. They don’t have much going on. That, in many ways, is the most tragic aspect of this illness. There’s clinical import to this. There are a lot of reasons the field started focusing on positive rather than negative symptoms. One is that the medications we had to offer, the treatments, were quite good at limiting or decreasing severity of delusions or hallucinations, but absolutely lousy about doing anything about negative symptoms. In fact, there is some evidence that they may make negative symptoms worse. One of the things this lab did was to bring attention to negative symptoms, not only from a theoretical and research perspective, but also from a clinical perspective. To say, ‘hey, these need to be the target of our therapies,’ and we can’t be satisfied with therapies that just target positive symptoms.

One is she developed instruments to assess negative symptoms. And probably her most-cited paper is an early paper on this Scale to Assess Negative Symptoms or SANS. The second important way I think this laboratory has influenced the field is by continuing to harness ongoing developments in technology, specifically neuro-imaging technology, toward the study of schizophrenia.

In truth, I’m not sure that I would say that. Within the last five years, probably the most important change is that people are starting to recognize this illness is unlikely to be explained by pathology in any single part of the brain. It’s much more likely to involve abnormalities in complex circuitry. The brain works as a unit. Everything is connected to everything else, and what we really need to be looking at more is abnormality in the circuitry level. The problem is that we use one word to describe the disorder—schizophrenia—but it’s likely that what we call schizophrenia will ultimately prove to be a very heterogeneous group of disorders, with very heterogeneous underlying pathophysiological mechanisms. And this brings us back to the sample size discussion. If that’s indeed the case then taking any 15 subjects is unlikely to shed any light on the underlying mechanisms because those 15 subjects may very well suffer from 15 different disease processes. A hundred subjects is not so great, but if in fact 10% of the people with schizophrenia share one type of pathology then you might be able to establish some patterns. The larger, the more extensive your sample, the more likely you are to find patterns that might tip you off to one of the more common disease processes. One of the problems with schizophrenia is that people think we know a lot more than we do. When you say the word schizophrenia and you have all these fancy studies and this fancy DSM [Diagnostic and Statistical Manual] people easily get confused and think it’s a single disease and not the syndrome that it is.

I’m not sure I want to say that. Although my guess is that if I thought about it for awhile, I’d end up going along with that; that that is the single biggest barrier. The other thing I should reiterate is that if in fact this is correct, then any one definition of it is not likely to be right and that, too, is a barrier.

Clearly advances in treatment are realistic. The treatments we have to offer people with schizophrenia today are markedly better than they were 20 years ago and they should get better still. We also talked about negative symptoms and how cognitive or neuropsychological dysfunction is an important aspect of research now. A reasonable short-term goal would be to take the next step. If we better understand neuropsychological deficits, maybe we could do more to come up with cognitive rehabilitative approaches to this illness. Just like we take stroke patients who have deficits and we can retrain them. Perhaps if we pay more attention to cognitive deficits of schizophrenia we can do better in helping people to negotiate those limitations.

  How about understanding the nature of the illness itself and beating it?

You’re asking ‘Am I going to understand schizophrenia in my lifetime?’ There I’m less optimistic. Last year, the big schizophrenia conference was in Sante Fe, and I heard people I really respected say I think we’re going to crack this thing. I thought, are they serious? I don’t think we have a chance in hell of cracking this thing. I had just come from an absolutely brilliant lecture on Huntington’s Disease. And here’s a disease in which we have the gene, we know a lot about it; it’s very simple in terms of inheritance patterns, and the bottom line of this one-hour lecture is they are a long, long way away from understanding how that one genetic aberration translates into this disease process. I think we are a long, long way from understanding the fundamental nature of what we call schizophrenia.

Dr. Michael Flaum
University of Iowa College of Medicine
Department of Psychiatry
Iowa City, IA, USA