(Proc. Natl.Acad. Sci., USA 94:1612-1614, 1997)
H. Barondes
a, Bruce
M. Alberts
b, Nancy
C. Andreasen
d, Francine
e, Patricia
Irving Gottesman
Stephen F. Heinemann
Edward G. Jones
Marc Kirschner
David Lewis
k, Martin
l,Allen Rosesm,
John Rubenstein
Solomon Snyder
Stanley J. Watson
Daniel R. Weinberger
and Robert H. Yolken

of Psychiatry, Center for Neurobiology and Psychiatry, University
of California, San Francisco, CA 94143-0984;bNational
Academy of Sciences, Washington, DC 20418;cThe
University of Iowa, MHCRC, 2911 JPP, 200 Hawkins Drive, Iowa
City, IA 522-1533;dDepartment of Anatomy, University
of California, San Francisco, CA 94143;eMailman
Research Center, McLean Hospital, Belmont, MA 02178;fSection
on Neurobiology, Yale University School of Medicine, New Haven,
CT 06510;gDepartment of Psychology, University of
Virginia, Charlottesville, VA 22903;hThe Salk
Institute, P.O. Box 85800, San Diego, CA 92186;iDepartment
of Anatomy and Neurobiology, College of Medicine, University of
California, Irvine, CA 92717;jDepartment of Cell
Biology, Harvard Medical School, Boston, MA 02115;kCenter
for the Neuroscience of Mental Disorders, University of
Pittsburgh Medical Center, Pittsburgh, PA 15213-2593;lLaboratory
of Molecular Cell Biology, Department of Biology, Medawar
Building, University College, London Gower Street, London WC1E
6BT, United Kingdom;mDivision of Neurology, Department
of Medicine, Duke University Medical Center, Durham, NC 27710;nDepartment
of Psychiatry, University of California, San Francisco, CA 94143;oDepartment
of Neuroscience, Johns Hopkins University, Baltimore, MD 21205;pMental
Health Research Institute, University of Michigan, Ann Arbor, MI
48109-0720;qNational Institute of Mental Health, St.
Elizabeth’s Neuroscience Center, Washington DC 20032; and rDepartment
of Pediatrics and Infectious Diseases, Johns Hopkins Medical
School, Baltimore, MD 21205

November 29-30, 1995, the National Academy of Sciences and the
Institute of Medicine brought together experts in schizophrenia
and specialists in other areas of the biological sciences in a
workshop aimed at promoting the application of the latest
biological information to this clinical problem. The workshop
paid particular attention to evidence of pathology in the brains
of people with schizophrenia, and to the possibility that this
reflects an abnormality in brain development that eventually
leads to the appearance of symptoms. The participants were
impressed with the complexity of the problem, and felt that
multiple approaches would be required to understand this disease.
They recommended that a major focus should be on the search for
predisposing genes, but that there should be parallel research in
many other areas.

Schizophrenia is
one of the most devastating human diseases. It afflicts about 1
in 100 people in the United States, generally becoming manifest
in late adolescence or early adulthood and persisting thereafter
(1). Because it begins so early and tends to interfere with
education, employment, and marriage, schizophrenia is a great
burden not only to those who suffer from it, but also to their
families and to society; thus, the number of people who are
seriously affected by this diseases and the economic and personal
cost to society are enormous.

treatments—which combine long-term medications with
supportive care—usually alleviate the more florid symptoms
of schizophrenia, such as paranoid delusions and hallucinations.
However, most affected individuals have substantial lifelong
impairment; indeed, more than one-half require continuous support
whether living in the community or in long-term institutions. The
chronic suffering of people with schizophrenia is highlighted by
the fact that 10-15% of them ultimately commit suicide (2) and
that many suffer homelessness. One-third of the nation’s
mental hospital beds are occupied by individuals with
schizophrenia. The annual direct cost of their care was estimated
to be $19 billion in 1991, with an additional indirect cost of
$46 billion due to lost productivity (3).

extensive study for many years and many initially promising, but
ultimately disappointing hypotheses, we do not have a clear
understanding of either the causes of schizophrenia or how the
causative factors lead to the clinical features. But, the
striking recent advancement in our understanding of cellular,
molecular, and integrative neuroscience, combined with the clear
recognition that schizophrenia is a well-defined disease with
documented changes in brain structure and function, present real
opportunities for research into its causes and treatment. To that
end, the National Academy of Sciences and the Institute of
Medicine brought together experts in schizophrenia and
specialists in other areas of the biological sciences in a
workshop aimed at promoting the application of the latest
biological information, technology, and experience to this
clinical problem.

The workshop,
held on November 29-30, 1995, paid particular attention to signs
of pathology in the brains of people with schizophrenia and to
the possibility that this reflects an abnormality in brain
development (4,5) that eventually leads to the appearance of
symptoms. The nonexperts in schizophrenia research at the meeting
were senior and young scientists in a wide variety of fields,
including cellular biology, biochemistry, receptor molecular
biology, the neurobiology of model systems, neuropharmacology,
control of cell growth and death, human genetics, and cognitive
neuroscience. Most of these individuals had little prior
knowledge of either the nature of schizophrenia or of the
hypotheses regarding its causes. In addition to educating both
experts and nonexperts about the disorder, the discussions were
designed to identify some promising areas for interdisciplinary
research in schizophrenia.

There was a
general consensus among the experts on the following points.

  1. Twin,
    family, and adoption studies suggest genetic factors are
    of substantial etiologic importance in schizophrenia
    (6,7). Their role is complicated because single gene
    models do not fit current family transmission data. A
    reasonable interpretation of these data is that several
    genes (or, more precisely, variants of genes, i.e.,
    alleles) act together. To date, none of the multiple
    genes that might contribute to schizophrenia has been
    identified, although scientists are presently focusing on
    regions of chromosomes 3, 6, 8, 9, 20, and 22.
  2. Genes do
    not act alone to cause schizophrenia (6, 7).
    Environmental factors are also important (8), although
    their nature is not established. Environmental factors
    that have been proposed include obstetric complications,
    intrauterine abnormalities and viruses. Although family
    and social factors causing stress may affect the course
    of the diseases, there is no convincing evidence that
    they play a major causative role.
  3. The brains
    of people with schizophrenia differ anatomically from the
    brains of normal people, and the differences can often be
    demonstrated by brain-imaging techniques (4) and by
    examination of postmortem brain specimens (4, 9-12).
    Although no anatomical change is shared by all affected
    individuals diagnosed with schizophrenia, there is a
    growing consensus from postmortem studies and from
    functional imaging of affected individuals that there is
    both a volume loss and cellular pathology in areas of the
    cerebral cortex. But there is overlap between normal and
    affected individuals, and there is no consensus about
    which, if any, of these differences are either
    characteristic of or essential for the development of
  4. The
    neuroanatomical findings, in many cases, are compatible
    with the presence of an abnormality in brain development
    (4, 5), possibly in fetal life. They are different from
    the anatomical changes seen in degenerative brain
    diseases, whether of genetic, infectious, or traumatic

In brief,
schizophrenia can be conceptualized as a complex biological
disorder in which
genes play a role (but not an exclusive one) and in which brain
development is likely to be abnormal.

The nonexpert
participants were stimulated and intrigued by their exposure to
the details of this disease. For them, the following observations
were especially useful for thinking about future research

    1. Although
      psychological stress was once thought by some
      clinicians to cause schizophrenia, the consensus
      is that it is instead a disorder of brain
      function that is due to complex factors—with
      developmental processes, including the onset of
      puberty, thought of as possible triggers for the
      full display of the symptoms of the disease. In
      prospective studies, many affected people have
      been found to be somewhat developmentally delayed
      in childhood (13, 14), which is consistent with
      the subtle but detectable abnormalities in brain
      anatomy of people with schizophrenia, including
      those examined prior to treatment with
      antipsychotic medications. Twin studies provide
      evidence of a genetic contribution: if one
      identical twin gets the disease, the other has
      approximately a 30-40% chance of getting it (with
      a mean delay of 4 years), even if the two have
      been brought up in different families (15, 16).
      Nevertheless, the shared genes of identical twins
      are not sufficient to give rise to the disease in
      all instances, pointing to the participation of
      other factors.
    2. People
      with schizophrenia may display a variety of
      patterns of aberrant behavior, raising the
      possibility that each pattern reflects a
      different form of the disorder due to a different
      cause. But even identical twins who both have
      schizophrenia, and who are presumed to have been
      afflicted because of shared causative factors,
      often have different patterns of symptoms.
      Therefore, the variability in symptoms does not
      appear to be a basis for distinguishing between
      schizophrenia due to different causes.
    3. The
      unusual time course of appearance and development
      of the symptoms of schizophrenia is an important
      characteristic that should guide research
      strategies. The fact that most cases arise in
      early adulthood is an important clue with respect
      to pathophysiology, because developmental changes
      may continue to occur in the brain during this
      period. The onset of the illness, with subsequent
      progression of symptoms during the first few
      years and later stabilization for the remaining
      years, makes schizophrenia very unusual among
      brain disorders. Evidence that schizophrenia is
      not a degenerative brain disease is the absence
      of gliosis, the proliferation of nonneuronal
      brain cells that accompanies central nervous
      system injury.
    4. Brain
      imaging studies suggest that people with
      schizophrenia have abnormalities in the
      prefrontal, temporal, and anterior cingulate
      regions known to be involved in cognition and
      affective integration, as well as in midline
      regions of the brain. Moreover, memory problems
      resembling those produced by lesions of the
      frontal lobe in monkeys and humans are evident in
      schizophrenia (17) and may account for the
      cardinal symptoms of disorganization in thinking,
      planning, and expressing thoughts.

There was a consensus among experts and nonexperts that
understanding schizophrenia would ultimately require a much
deeper understanding of many aspects of brain structure and
function. Because schizophrenia is widely believed to result from
abnormalities in brain development, interest was expressed in the
study of genes that control brain development. These include
homeobox genes that control the formation of specific brain
regions and genes that encode proteins that control the processes
of cell adhesion and cell migration responsible for the precise
cellular connections in the brain.

The participants
expressed great interest in extending basic physiological and
anatomical studies of neuronal circuits in the temporal and
frontal lobes and midline regions that control aspects of
cognition and emotion, because these regions and functions appear
to be affected in schizophrenia. If more were learned about the
details of such neuronal circuits, mechanisms of specific
functional defects in schizophrenia should be easier to unravel.
Study of these circuits at the anatomical level would be greatly
facilitated by detailed mapping of key proteins involved in
neurotransmission in the human brain. Several participants
expressed interest in a national program to develop a library of
sections of regions of human brain suitable for histochemical
examination that could be made available to all qualified
investigators. The ability to image brain function in living
individuals should also be very informative, especially as the
resolution of this technology increases. It would be particularly
useful to understand the changes that occur as the disease
develops, e.g. in initially unaffected twins who later develop
schizophrenia or individuals presenting symptoms early in the
course of their disorder.
A major focus of the workshop was on evidence of microscopic
pathology in postmortem specimens of the brains of people with
schizophrenia. Most participants were convinced by the evidence
that is suggestive of a defect in interneuronal connectivity in
the frontal and temporal cortical and related subcortical regions
of the brains of individuals with schizophrenia. Although it is
not certain whether the defect is primary or secondary to the
disease, the evidence of cortical volume loss in first-episode
patients indicates that it may play a causative role. There was a
consensus that the disease does not reflect a classical
neuodegenerative condition, in that it is not accompanied by the
changes in glial cells that are typically found with
neurodegeneration. The evidence of a defect in connectivity may
be a reflection of either an abnormality in brain development or
of the later onset of atrophic changes in the frontal and
temporal cortex. The histological data that can be interpreted as
indicating abnormal migration of cortical cells to their normal
destinations is consistent with a failure of normally programmed
death of these cells during brain development, but is seen in
only a fraction of the cases. Immunohistochemical and in situ hybridization
techniques are providing more specific evidence of
neuroanatomical abnormalities, but this work is still in its
early stages.

assessment of the microscopic pathology of the brains of people
with schizophrenia depends on the availability of properly
preserved specimens from well-characterized affected individuals.
There is a national program to collect such specimens, but
several participants expressed dissatisfaction with either the
quality of some of the material or its availability. Some
participants expend a great deal of effort in collecting
specimens on their own. The national program would be greatly
facilitated by cooperative collection, better characterization,
preservation, and distribution of brain tissue. All agreed that a
reference collection of material would be particularly valuable,
making possible direct correlations of observations on tissue
from the same cases by different investigators.

Despite the
great importance of medications in the treatment of
schizophrenia, the workshop steered away from issues of drug
development because they are already under intensive
investigation by the pharmaceutical industry and at academic
centers. Nevertheless, strong support was expressed for
refinement of available drugs to minimize side effects while
retaining therapeutic effects. To achieve this goal it may be
necessary to find drugs with much greater specificity for
particular neurotransmitter receptors. Because there is a
suspicion that the therapeutic efficacy of currently used drugs
is based on their interactions with more than one receptor,
appropriate combinations of several specific drugs might best
target the relevant receptors without the side effects caused by
additional interactions.

There was also
strong support for continuing basic research on the molecular and
physiological basis of neurotransmission on the molecular and
physiological basis of neurotransmission. Even though drugs
targeted to dopamine receptors and to the mesolimbic and
mesocortical dopaminergic pathways have proven to be highly
effective in alleviating many symptoms of schizophrenia, drugs
directed to other targets may also produce some symptomatic
relief. Therefore, additional work on other neurotransmitters and
receptors should help in the development of new classes of drugs.

promising subject for detailed study is the unaffected identical
twin of a co-twin newly diagnosed with schizophrenia, who often
will subsequently develop the disease, with a mean delay of 4
years. A research project focused on a close observation of large
numbers of as-yet-unaffected twins during this period of
vulnerability could detail the progression of the disease, and
test possible preventive therapies. As one example, repeated
imaging of the brains of both twins could reveal whether the
brains of those twins that develop the disease show reproducible
changes as their status changes from normal to schizophrenic.

Because of the
availability of genetic maps with a high density of markers and
the development of new statistical approaches that facilitate the
identification of multiple genetic loci responsible for human
diseases, the rate-limiting factor in finding the alleles that
predispose to schizophrenia is the availability of families with
sufficient numbers of rigorously diagnosed cases. Some
participants expressed concern that even the existing cooperative
programs established by the National Institute for Mental Health
might not provide adequate material, sine very large samples may
be needed to find relevant genes for a disorder that probably
involves interaction of a number of genes, and that also may be
genetically heterogeneous. For this reason it may be necessary to
collect thousands of DNA samples from groups of relatives with
schizophrenia, including pairs of siblings who are both affected,
to provide sufficient material for genetic studies.

Despite the
consensus that genes are important in schizophrenia, there was
strong interest in also encouraging other avenues of
investigation. Even when relevant genes are found, interpretation
of their actions could be difficult because of our limited
understanding of fundamental aspects of brain functional
organization and development.


This workshop
was designed to facilitate the application of biological sciences
to a very challenging and complicated medical problem of immense
personal and social significance. The conclusion of all the
participants was that many aspects of the problem could, in fact,
be approached by applying the explosive growth of knowledge and
techniques derived from fundamental research in neuroscience and
molecular biology. In fact, one main goal of the workshop was to
help stimulate the interest of a new generation of basic and
clinical scientists in schizophrenia. For even though there has
been an impressive growth of knowledge about this devastating
disorder, there was general agreement that we are still far from
understanding its cause and pathogenesis.

There was also a
consensus that multiple approaches would be required to
understand this disease. Clearly, a major focus should be on the
search for predisposing genes. But even though all agreed that
the availability of (and continuing improvements in) the
necessary technology make this a very valuable approach, there
was lively disagreement on the degree of emphasis that it
warranted relative to other approaches. The advocates of the
genetic approach felt that it should be central because
identification of the relevant genes would open up so many other
avenues of research by providing aids to diagnosis, clues to new
treatments, and reagents (such as nucleic acid probes and
monoclonal antibodies) for more detailed studies of the
microscopic pathology of schizophrenia. Others felt that this
approach should not draw resources away from work in other areas.
And even the strongest advocates of genetic studies of
schizophrenia agreed that parallel research in basic
neurobiology, neurophysiology, microscopic neuroanatomy of the
brain, and integrative neuroscience, as well as a continued
search for environmental factors, will be essential for
understanding the actions of the relevant genes and for devising
new treatments.

Funding for
the workshop was provided by the Theodore and Vada Stanley
Foundation and the National Alliance for the Mentally Ill.

  1. Sartorius
    N, Jablensky A, Korten A, et al. (1986) Psychol Med 16,
  2. Caldwell,
    CB & Gottesman II. (1992) Suicide Life-Threatening
    Behav. 2,479-493.
  3. Wyatt RJ,
    Henter I, Leary MC & Taylor E. (1995) Soc. Psychiatry
    Psychiatric Epidemiol 30, 196-205.
  4. Weinberger
    DR (1995) Lancet 346, 553-557.
  5. Royston MC
    & Roberts GW (1995) Curr Biol 5, 342-344.
  6. Gottesman
    II (1991) Schizophrenia Genesis (Freeman, New York).
  7. McGuffin P,
    Owen MJ & Farmer AE (1995) Lancet 346, 678-682.
  8. Torrey EF
    (1992) Schizophr Bull 18 (2), 159-170.
  9. Shapiro RM
    (1993) Schizophr Res 10, 187-239.
  10. Akbarian S,
    Kim J, Potkin S, Hetrick W, Bunney W & Jones EG
    (1996) Arch Gen Psychiatry 53, 425-436.
  11. Selemon LD,
    Rajkowska G & Goldman-Rakic PS (1995) Arch Gen
    Psychiatry 52, 805-818.
  12. Benes FM,
    Davidson J & Bird ED (1986) Arch Gen Psychiatry 43,
  13. Done DJ,
    Crow TJ, Johnstone EC & Sacker A (1994) Br Med J 309,
  14. Jones P,
    Rodgers B, Murray R & Marmot M (1994) Lancet 334,
  15. Gottesman
    II (1994) Acta Psychiatr Scand 90, Suppl. 384, 26-33.
  16. Torrey EF,
    Bowler AE, Taylor EH & Gottesman II (1994)
    Schizophrenia and Manic Depressive Disorder: The
    Biological Roots of Mental Illness as Revealed by a
    Landmark Study of Identical Twins (Basic Books, New
  17. Goldman-Rakic
    PS (1994) J Neuropsychiatry Clin. Neurosci 6, 348-357.