Deconstructing Psychosis (February 15-17, 2006)

Prepared by Michael B. First, M.D., DSM Consultant to the American Psychiatric Institute for
Research and Education (APIRE), a subsidiary of the American Psychiatric Association

Deconstructing Psychosis, the fifth diagnosis-related research planning session convened under
the conference series on the Future of Psychiatric Diagnosis: Refining the Research Agenda, was
held at APA headquarters in Arlington, Virginia on February 16
th
and 17
th
, 2006. . APAs American
Psychiatric Institute for Research and Education (APIRE) is sponsoring the project in collaboration
with the World Health Organization and the funding agency, the National Institutes of Health. The
five-year effort represents an unprecedented scientific preparatory phase in advance of the next
revision of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) and other psychiatric
classification systems.
Carol A. Tamminga, M.D., University of Texas Southwestern Medical Center in Dallas, TX, and Jim
Van Os, M.D., Maastricht University, Maastricht, the Netherlands co-chaired the psychosis
meeting. Twenty-one invited scientists from around the world participated. The goal of the
conference was to examine psychosis dimensionally, the definition of, and new data concerning,
schizophrenia, bipolar disorder, major depressive psychosis, and substance-induced
psychosis. "Functional" psychotic states (e.g. paranoia) and those related to defined pathological
changes in the brain (e.g., dementia) were not a focus of the reviews or recommendations.

Three initial presentations reviewed the diagnostic features of schizophrenia, bipolar psychosis, and
major depressive psychosis. Jim Van Os, M.D. (Maastricht, the Netherlands) presented evidence
for and against the diagnostic validity and usefulness of a schizophrenia diagnosis and discussed
possible alternative approaches. Noting that the current phenotype of psychosis originated in
psychiatric hospitals at the time of Kraepelin, Van Os observed that little is known today about how
psychosis presents in general practice settings or in the general population. Given that psychosis
results from the interaction of many genes, it is likely to be a continuous phenotype with mild forms
present in the general population. At present, 114 combinations of symptoms can lead to meeting
the DSM-IV definition of schizophrenia; that, and the fact that different populations of patients are
defined by different diagnostic systems (i.e., ICD-10, RDC, DSM-III-R, DSM-IV), raise questions
about the validity of the DSM-IV definition. One means of improving diagnosis will be to find groups
that are more homogeneous, such as patients who are characterized primarily by the deficit
syndrome. Another approach that Dr. Van Os proposed entails grouping symptoms into dimensions
rather than people into categories, with the aim of developing quantitative phenotypes that
correspond more closely to clinical realities than do the current diagnostic categories. While some
evidence suggests that a dimensional approach may be superior to a categorical approach in terms
of clinical usefulness and prognostic ability, additional research is needed to determine which type
of dimensional representation would have the most diagnostic usefulness. Commenting on Dr. Van
Os paper, Wolfgang Gaebel, M.D. (Dusseldorf, Germany), noted that the diagnostic entity of
schizophrenia comprises a loose boundary around a heterogeneous collection of interrelated and
relatively distinct phenotypes. These variants relate to relatively distinct brain-behavioral modules
each with either overlapping or separate etiology, pathophysiology, course characteristics, and
treatment response. Dr. Gaebel suggested that the diagnostic concept of schizophrenia be
abandoned or supplemented in favor of a modular illness concept that focuses on pathological
functions and etiopathogensis (e.g., emotional, cognitive, social) as opposed to diagnostic
categories.

In the next presentation, Eduard Vieta, MD, PhD (Barcelona, Spain) examined the diagnostic
usefulness of bipolar disorder and the separation of bipolar psychoses and schizophrenia into
distinct diagnostic categories. Imaging data examining volume loss in brain structures reveal that
the genetic risk for schizophrenia is associated with volume loss in gray matter in left frontal-
striatum-thalamic and temporal areas, whereas the genetic risk for bipolar disorder is associated
with volume loss in gray matter in the right anterior cingulate cortex and in ventral striatum. An
association between genetic risk for both schizophrenia and bipolar disorder, however, is evident in
volume loss in white matter in frontal and temporo-parietal areas. Differences may be seen in the
cognitive profiles of patients with schizophrenia and bipolar disorder, although mediocre sensitivity
and specificity of neuropsychological testing instruments limit their usefulness as a diagnostic tool in
differentiating the two disorders. Nonetheless, an overlap in symptoms between the two i.e., the
presence of affective symptoms in schizophrenia, and psychotic symptoms in bipolar disorder
suggests that psychotic disorders may lie along a continuum that extends from schizophrenia to
schizoaffective depression to schizoaffective bipolar, and bipolar I disorder. Dr Vieta outlined
limitations in the current bipolar criteria including 1) the absence of psychotic symptoms in the
diagnostic criteria for bipolar disorder reinforces the idea that psychosis is a core feature of
schizophrenia but not bipolar disorder; 2) mood-congruent vs mood-incongruent psychotic
symptoms are not well-defined; 3) mixed episodes are too narrowly defined; 4) cognitive symptoms
are not included; 5) medication-induced mania and hypomania are excluded from a diagnosis of
bipolar disorder; 6) the difficulty of judging what direct physiological consequence of a drug,
medication, or somatic treatment means; 7) the failure of the definitions to take account of family
history and biological markers; 8) the four-day duration required for a diagnosis of hypomania and 1
week for mania may be too long; and 9) Bipolar Disorder NOS may include the majority of cases,
particularly in children and adolescents. He proposed a new specifier for indicating predominant
polarity given that polarity may be the best predictor of long-term treatment response; current
evidence, he said, indicates that predominantly depressive polarity may respond better to
lamotrigine, and predominantly manic polarity to atypical antipsychotics).

In her discussion of Dr. Vietas paper, Mary Phillips, MD (Pittsburgh, PA) proposed using what she
described as the psychiatric toolbox (i.e., neuropsychological tests, neuroimaging, genotyping) to
develop disorder biomarkers that are persistent, rather than state-dependent, in order to improve
early diagnosis based on neurobiologic and neurocognitive markers. She cited need for research
on behavioral and neural endophenotypes in order to detect risk of future disorder and on
neurobiological predictors of treatment response in order to improve treatment strategy. Dr. Phillips
advocated a multidisciplinary approach over the longer term in order to understand the
pathophysiological mechanisms of the full spectrum of bipolar disorder.

Alan Schatzberg, MD (Stanford, CA) reviewed the diagnostic usefulness of psychotic major
depression. Key characteristics that differentiate it from other manisfestation of psychosis are its
clinical features (i.e., psychotic thinking and neuropsychological impairment in areas such as
attention, executive function, verbal declarative memory, visual memory, and visual-spatial
perception), biological features (HPA axis abnormality, low DBH activity, elevated platelet MAO
activity, and alterations in REM sleep), treatment response (poor placebo response, response to
ECT, amoxipine, and combined antidepressant/antipsychotic therapy, and possibly GR
antagonists), and course (psychotic recurrences, higher mortality rates, higher rates of suicide
ideation). Problems with the current DSM classification, Dr. Schatzberg said, include the implication
that psychosis is associated only with severe depression, and inadequacies in the definition, which
includes only delusions and hallucinations. Among issues requiring further study are whether the
definition can be broadened to include other positive symptoms such as conceptual disorganization
and unusual thought content, whether neuropsychological or HPA axis measures can be employed
in the diagnosis, and determining the prognostic implications of a past diagnosis of psychotic major
depression. Although no single laboratory or neuropsychological test can differentiate accurately
between psychotic and non-psychotic depression, a combination of clinical ratings,
neuropsychological testing, and cortisol levels has been shown to differentiate between these
groups with a high degree of accuracy. Responding to these comments, Mario Maj, MD, PhD
(Naples, Italy) pointed out that the boundary between psychotic and nonpsychotic depression is not
clear-cutthoughts and feelings of guilt and worthlessness may be present in various degrees with
fluctuations within same episode, and hallucinations may occur only occasionally. Dr. Maj
contended that while extant research evidence supports the usefulness of a psychosis specifier, it
does not support the introduction of a new diagnostic subtype. He agreed that the specifier should
probably be kept separate from the issue of severity and that it should be possible to record the
presence of both mood-congruent and mood-incongruent psychotic features in the same patient
since both may be present simultaneously.

Robin Murray, MD (London, England) addressed three questions in his presentation: 1) what are
the critical similarities and differences between schizophrenia and bipolar disorder; 2) what is the
relation between psychotic illness and minor psychotic symptoms in the general population; and 3)
does the biological meaning of having schizophrenic symptoms differ among different
ethnic/cultural groups? From a genetic perspective, considerable overlap exists between
schizophrenia and bipolar disorder; for example, if one member of a monozygotic twin pair has
schizophrenia, there is a 40% chance that the co-twin will have met criteria for schizophrenia, an
8% chance the co-twin will have met criteria for schizoaffective disorder, and an 8% chance the co-
twin will have met criteria for mania. Yet given the fact that pre-schizophreniform subjects have
demonstrable impairment in motor and language development whereas pre-bipolar subjects have
no general neurocognitive impairment, Dr. Murray proposed a model wherein, given a background
of shared genetic predisposition to psychosis, schizophrenia but not bipolar disorder is subject to
additional genetic or early environmental insults that impair neurodevelopment. With respect to the
relation between non-clinical psychotic symptoms and schizophrenia, minor psychotic symptoms in
the general population are related to the same variables as is schizophrenia (i.e., low verbal IQ, low
education, living in cities, cannabis dependence, life events/victimization, neurotic symptoms, being
a member of an ethnic minority). This suggests that psychosis, much like hypertension, represents
one extreme of a distribution of psychotic symptoms. Regarding the question of ethnic differences
in psychotic symptoms, evidence supports significant cultural differences in schizophrenia
symptoms for blacks in the UK as compared to other ethnic groups. For example, black patients
diagnosed as having affective psychosis have more first rank symptoms (37%) than white patients
(21%) and black patients diagnosed with schizophrenia have more positive and manic symptoms
and less chronicity than white patients, suggesting that the application of the traditional Kraepelinian
diagnostic system to black patients in the UK is problematic. From all of these lines of evidence, Dr.
Murray concluded that the current dichotomous system is no longer useful. He suggested that
psychosis be re-termed dopamine dysregulation disorder and that individuals so diagnosed should
be rated on developmental impairment and on symptom factor scores.

Jean Addington, PhD (Toronto, Canada) one of three respondents to Dr. Murray, focused on the
challenge of identifying individuals who are at ultra-high risk of developing schizophrenia. Current
criteria for schizophrenia include the construct of prodrome but currently this can only be
determined retrospectively. Various research criteria for identifying this ultra or clinical high-risk
group have been proposed and include items such as attenuated positive symptoms, negative
symptoms, and genetic risk plus recent deterioration in functioning. While individuals at genetic
high risk (i.e., first-degree relatives with schizophrenia) have a 10-20% risk of developing
schizophrenia, these ultra-high risk groups convert to psychosis at rates of 25%-54% in 12
months. Dr. Addington proposed using current prodromal/ultra high risk studies to develop models
of prediction that might make the diagnostic criteria for schizophrenia more useful and suggested
that future studies test biopsychosocial and genetic markers of the ultra high risk group. Kwame
McKenzie BM (London, England), emphasized in his comments the impact of culture on the
diagnosis of psychotic disorders. Noting that culture has a ubiquitous multi-level impact that is
impossible to ignore, he observed that the differences in rates of illness in different minority ethnic
populations, different course, and different reactions to getting a diagnosis of a psychotic disorder
can make specific categorical diagnoses untenable in particular cultural groups. With world-wide
migration increasing, it is crucial to develop a diagnostic system that will be useful across cultural
minority groups. Furthermore, given that the positive predictive value of a psychotic symptom for a
diagnosis of a psychotic disorder varies among countries and different ethnic groups within a
country, and that there are different balances of etiologies and outcomes among ethnic groups
within a country, assumptions about cross-cultural similarities of categorical diagnoses are difficult
to support. Michael Phillips, MD (Beijing, China) provided a perspective from China in his
discussion of culture and psychiatric diagnosis. He noted that the experience and understanding of
psychotic symptoms are imbedded in a network of local meanings that vary from nation to nation,
within different sub-cultural groups in a single nation, and over time (as communities undergo socio-
cultural changes). The nosological paradigms developed to categorize different types of psychotic
symptoms are also imbedded in specific professional cultures; these paradigms change as
professional cultures evolve. Dr. Phillips proposed developing standardized qualitative and
quantitative methods that can be employed in a wide range of different communities to conduct
culture-sensitive assessments of psychotic symptoms. Only then will it be possible to determine if
universal gold standard criteria for psychotic diagnoses are feasible.

Richard Keefe, PhD (Durham, NC) presented on the cognitive psychology of psychosis. Cognitive
impairment is not included among the criteria for schizophrenia, although it is considered to be a
core component of the disorder and is more predictive of dysfunction in increasingly complex
societies than positive and negative symptoms. Various studies demonstrate that schizophrenia
and schizoaffective disorder share a similar pattern of cognitive impairments, which differs from
patterns in major depression, bipolar disorder, and Alzheimers dementia. There is, however, no
point of rarity in cognitive functioning between schizophrenia and other psychotic disorders. Dr.
Keefe pointed out several challenges involved in defining cognitive impairment in
schizophrenia. It would be unrealistic to expect formal cognitive testing on every individual
suspected of having schizophrenia. Because clinician ascertainment of cognitive impairment is
notoriously poor, input from other informants (e.g., family, caregivers) is very important. Still,
determining whether there has been a decline from pre-morbid functioning or from expected
cognitive level is complicated by the fact that pre-morbid levels are decreased in people who later
develop schizophrenia and by the fact that expected cognitive level is based upon early school
performance and test scores, parental/family education, and reading level. It is important to
demonstrate that the cognitive impairment is relatively stable over time, as this pattern differentiates
it from the cognitive impairment characteristic of affective disorders. Even if cognitive impairment
does not increase the point of rarity between schizophrenia and other psychotic disorders, including
cognitive impairment in the definition of schizophrenia may draw clinicians attention towards
neglected treatment options, especially given current and future interest in the development of
treatments for cognitive impairment in schizophrenia. To that point, Wayne Fenton, MD (Bethesda,
MD), outlined in his discussion ongoing NIMH initiatives targeting cognition in
schizophrenia. These include the MATRICS project (Measurement and Treatment Research to
Improve Cognition in Schizophrenia), assessment standardization, and the TURNS project
(Treatment Units for Research on Neurocognition and Schizophrenia). Dr. Fenton noted that
cognitive impairment (specifically immediate verbal memory) is more highly correlated with
functional outcome than delusions, hallucinations, or thought disorder. A variety of compounds
(e.g., glutamatergic modulators, dopamine D1 receptor agonists, serotonin 5-HT2A receptor
antagonists, cholinesterase inhibitors) potentially have promise in the treatment of cognitive
impairment. In order to facilitate needed research, the MATRICS project has developed a
consensus battery of neuropsychological tests that cover the most important cognitive domains
affected by schizophrenia.

Michael Owen, MD PhD (Cardiff, Wales), in his presentation, provided a genetic deconstruction of
psychosis. The categorical distinction between schizophrenia and bipolar disorder has allowed
some genes to be identified, and has increased the power to find some genes but reduced the
power to find others. Genetic epidemiology suggests overlap between schizophrenia,
schizoaffective disorder, and bipolar disorder. Although it is widely assumed that schizophrenia
and bipolar disorder breed-true, schizoaffective disorder shows familial overlap, with evidence
accumulating that siblings of schizophrenic and bipolar probands have increased risk of other
disorders. Furthermore a number of recently identified candidate genes (e.g., NRG1, Dysbindin,
G72, and DISC 1 ) appear to confer risk across both disorders . Dr. Owen concluded that there is
no compelling genetic argument for continuing to distinguish between schizophrenia and bipolar
disorder. He called for research that combines inductive and hypothesis-driven approaches aimed
at relating genes and biological processes to symptoms and syndromes defined at both clinical and
endophenotypic levels. In response to the presentation, Assen Jablensky MD (Perth, Australia)
noted that, as yet, no candidate gene or combination of genes has been shown to be either
necessary or sufficient to cause schizophrenia or bipolar disorder and that the effect sizes of genes
that have so far been implicated in psychiatric disorders are small to modest. With few exceptions,
genetic studies of schizophrenia to date have been based solely on the broad clinical diagnosis as
the phenotype. Currently known cognitive, neuroanatomical, and biobehavioral markers might
account for more of the genetic variation than does the clinical diagnosis of
schizophrenia. Consequently, there is a need for robust endophenotyping strategies that focus on
traits associated with brain structure or function, that are stable, objectively measurable, heritable,
and correlate with the transmission of schizophrenia within families. Dr. Jablensky remarked that
although the current state of knowledge about the genetic bases of psychotic disorders cannot
support any radical re-designing of the DSM (or ICD) nosology of psychoses, the diagnostic
categories can be supplemented with research-oriented, evidence-based, annotations on
dimensional models, recommended endophenotype measures and testable subtyping of disorders.

Francine Benes MD, PhD (Belmont, MA) described her use of a top-down approach in her
presentation on the molecular biology of schizophrenia and bipolar disorder. Working with
postmortem tissues, investigators first look for regions-of-interest, then circuits within those regions,
and, finally, molecular regulatory mechanisms within specific subpopulations of neurons that show
abnormalities in schizophrenia and/or bipolar disorder. Dr. Benes proposed a two-factor model for
defining cellular endophenotypes of schizophrenia and bipolar disorder; the model assumes
environmental genes that are regulated by pre-natal or post-natal stress and susceptibility genes
that are turned on or off in adolescence. GABA-ergic interneurons are common sites for cellular
and molecular abnormalities in both schizophrenia and bipolar disorder, and GABA cells in these
two disorders show similarities in gene expression that may reflect environmental risk factors in
both schizophrenia and bipolar disorder. Support for this proposal derives from a rodent model
that simulates changes in neural circuitry found in postmortem studies of the hippocampus in
psychotic disorders. In human postmortem studies, GABA cells also show differences in gene
expression that may reflect heritable susceptibility factors in schizophrenia versus bipolar
disorder. The endophenotypes for schizophrenia and bipolar disorder probably involve changes in
specific gene clusters within specific subpopulations of GABA neurons. Dr. Benes suggested that
targeting specific functional clusters of genes expressed in specific GABA cells will lead to novel
therapies that are unique and specific for schizophrenia and bipolar disorder, respectively. Dr.
Benes discussant, Mayada Akil MD (Bethesda, MD), asked whether postmortem studies can truly
differentiate the two disorders. To illustrate the complexity of the problem, she cited a comparison of
one hundred RNAs, proteins and other markers in multiple brain regions in post-mortem brains by
56 research groups. The various teams reported a 65% overlap in molecular alterations between
bipolar disorder and schizophrenia. Dr. Akil enumerated challenges for future research efforts using
post-mortem brains: 1) crossing levels from molecular/cellular to circuits and systems to behavior;
2) molecular and cellular abnormalities are more easily linked to a functional disturbance (e.g.,
impaired working memory) or an endophenotype than to a syndrome or even a symptom; 3) the
effects of chronicity, treatment, and limited function on the brain cannot easily be disentangled from
pathophysiology; and 4) post-mortem studies do not provide any information about timing or
context. Future research strategies for confronting these challenges include: 1) combining
postmortem studies with imaging approaches and animal models; 2) thinking in terms of symptom
domains; 3) trying to understand normal development as a context for the development of
psychiatric disorders; and 4) pursuing discovery as well as hypothesis driven research.

Raquel Gur, MD PhD. (Philadelphia, PA) reviewed the state of the art and future challenges in the
use of neuroimaging to study psychotic disorders. Neuroimaging potentially may contribute
significantly to elucidating brain circuitry across disorders, for providing a developmental
perspective; in its application to genetic paradigms, and in studying treatment effects. To date,
most neuroimaging studies have examined one population meeting DSM-IV criteria. There is a
growing MRI literature comparing diagnostic groups including studies that compare patients with
schizophrenia to those with bipolar disorder and normal controls; studies that compare scans of
patients with first-episode psychosis to longitudinally followed patients with an established
diagnosis; and family studies that examine individuals with schizophrenia and compare them to
others with a family history of schizophrenia or bipolar disorder. Imaging research has been
hampered, however, by the lack of a conceptual approach to examining psychosis in fMRI
paradigms; moreover, studies generally have not been geared to address fundamental questions
because of small sample size, their cross-sectional nature, and the absence of clinical
correlates. Dr. Gur proposed promising hypotheses that could be tested by using neuroimaging
techniques, including: 1) there is both common and distinct brain circuitry that contribute to
schizophrenia and bipolar disorder; 2) this pattern can be identified early and relate to clinical status
including differential diagnosis; and 3) neuroimaging may elucidate endophenotypes. In summary,
Dr. Gur noted that while increasingly sophisticated imaging technology exists, it will not
substantively inform DSM-V.

Stephen Lawrie MD (Edinburgh, Scotland) agreed with Dr. Gur that neuroimaging data is more
likely to affect DSM-VIrather than DSM-V. He noted that current neuroimaging evidence indicates
that there is a subtle gross structural neuroanatomy of schizophrenia, which is evident to some
extent in relatives. One of the most consistent findings is a reduction in medial temporal lobe
volume, which likely occurs around the onset of the illness but is at least partly non-specific and of
largely unknown cause. To further investigate this, he recommended more longitudinal structural
MRI studies within 5 years of onset and urgent harmonization of techniques across research
centers. He suggested that if there is a gross functional neuroanatomy of schizophrenia, it is a
partly dopamine-modulated frontal (symptom-related) disconnectivity. Matcheri Keshvan MD
(Detroit, MI) noted that the advantages of neuroimaging include its reliability and that its capability
to reflect altered biology. Limitations include the fact that patient-control differences are quantitative
rather than categorical, that there is overlap in MRI data between disorders like schizophrenia and
bipolar disorder, and that neuroimaging data lacks etiologic specificity. Future research using brain
imaging studies should help to elucidate structural and functional endophenotypes that map on to
genetic polymorphisms and will hopefully offer ways of characterizing illness onset and progression
and help predict treatment response and outcome.

Carol Tamminga, MD (Dallas, TX), presented on the similarities and the differences in the
neuropharmacology of schizophrenia and bipolar disorder. For schizophrenia, medications that
may be efficacious include first and second generation antipsychotics, clozapine, mood stabilizers,
antidepressants, and cognitive enhancement medications. For bipolar disorder, efficacious
medications include lithium, valproate, first and second generation antipsychotic medications, and
new antiepileptic agents (e.g., lamotrigine). Antipsychotic medications treat psychosis in both
schizophrenia and in bipolar disorder. Although mood stabilizers alone can treat psychotic
symptoms in acute mania, they are not effective in treating psychosis in schizophrenia. Thus,
treating mood destabilization in bipolar disorder can resolve psychosis. Dr. Tamminga offered
three possible hypotheses: 1) psychosis has a distinct pathophysiology, common to both
schizophrenia and bipolar disorder, and antipsychotics target that molecular mechanism;
2) psychosis is mediated by neural systems which are different in schizophrenia and bipolar
disorder, which can be stimulated and treated from multiple points and by many different
pharmacologic strategies; and 3) psychosis is an unimportant neural response analogous to fever
and should not be a primary target for treatment. Dr. Tamminga suggested that systematic
pharmacology with pharmacological probes (e.g., ketamine, amphetamine, nicotine,
benzodiazepines, cognitive drugs), which has been done far less often in bipolar disorder than in
schizophrenia, would be helpful to elucidate the underlying mechanisms. Respondent John Davis
MD (Chicago, IL) urged more systematic presentation of clinical trial results to help elucidate
treatment-specific subtypes for example, determining if there were a subgroup of bipolar patients
that needed both antipsychotic medications and mood stabilizers. Dr. Davis also recommended
early intervention and treatment regardless of whether such treatment interrupts a pathological
process or makes it less severe, preventing the illness from doing irreparable damage to the
persons life course. Given the potentially severe side effects of these medications, it is important
to determine whether individuals with disorders such as schizophrenia and bipolar disorder need
lifelong continuous treatment or if treatment can be intermittent.

Given the common high comorbidity of substance use disorders (SUD) with other psychiatric
disorders, the conference grant supports a presenter at the SUD conference to participate in the
other diagnosis-specific conferences, a role assumed for this meeting by Bruce Rounsaville, MD
(New Haven, CN). He noted that there are three types of relationships between psychotic
symptoms and substance use: they can be independent (i.e., a primary psychotic disorder with
comorbid substance use), they can substance-induced, or they can be the expected effects of a
substance (e.g., hallucinations resulting from taking an hallucinogen). In a first episode of comorbid
psychosis and substance use, unless there is a clear history of psychotic symptoms prior to using
substances or during a drug-free period, the episode should be considered to be substance-
induced until proven otherwise. The most prevalent types of drugs self-administered by patients
with schizophrenia are nicotine (probably used to improve working memory), alcohol, cannabis, and
stimulants; conversely, little opioid use is seen. Patients with schizophrenia tend to use lower
amounts of drugs than do patients with primary substance use disorders. Furthermore, since some
minimal level of social functioning is generally needed to obtain hard drugs, patients with
schizophrenia who use drugs tend to show higher positive symptoms and higher social functioning
than those patient who do not use substaunces. . Diagnostic challenges in determining whether a
psychosis is primary or secondary to drug use include: 1) the absence of persistent drug-free
episodes in many individuals; 2) abuse of multiple substances, 3) poor recall of the temporal
sequencing of drug use and psychotic symptoms, 4) the fact that drug effects can endure (e.g., 1-
15% of patients who experience psychosis subsequent to stimulant use continue to have symptoms
more than one month after abstinence), and 5) primary and substance-induced psychotic disorders
have similar symptoms. Dr. Rounsaville predicted that for DSM-V, changes based on new data
would most likely affect the text rather than the diagnostic criteria. New data could help elucidate
symptoms that mark or predict independent vs. substance induced psychotic disorders, clarify the
casual role of different drugs to causing enduring psychosis, and lead to the development of
substance-specific data on how long the drug-free interval should be in order to determine that a
psychosis is primary vs. substance-induced.

At the conclusion of the presentations, the participants formed two breakout groups to discuss the
presentations in greater detail and to formulate recommendations for future research and
suggestions for future versions of the classifications of mental disorders (e.g. in the DSM-V, ICD-11
and others).

Recommendations from the first group included:
replacing the current categories with a general psychosis syndrome that would cover a
broad range of disorders ranging from schizophrenia, schizoaffective, delusional, and brief
psychotic disorders, to bipolar disorder and psychotic depression;
reducing the duration criterion for schizophrenia from 6 months to 1 month;
including dimensionally-specified criteria in the research appendix for positive symptoms,
negative symptoms, depression, mania, cognitive decline, and functional impairment;
including research specifiers for duration, time course, and mode of onset.
The second group recommended:
Harmonizing the discrepancy in the schizophrenia duration criteria between DSM and ICD
by conducting reanalysis of prospective studies that have data points at periodic (e.g.,
monthly) intervals up to 12 months, with the aim of ascertaining which duration cut point
maximizes chosen validators (e.g., diagnostic stability, outcome, social adjustment).
While agreeing that the schizophrenia/bipolar dichotomy has some validity problems, group
members expressed reservations about the impact on clinical utility of abandoning this
distinction. Since a large proportion of individuals fall into the overlap area between
schizophrenia and bipolar disorder and are currently diagnosed as schizoaffective, or
Psychotic/Mood NOS, studies are needed to focus on these problematic cases with respect
to treatment response, laboratory studies, prognosis and outcome. Another approach is to
parse out homogeneous subcategories of schizophrenia and bipolar disorder with respect
to etiology, treatment response, outcome; e.g., deficit syndrome. If the categorical
approach is retained in DSM-V, there is a pressing need to improve definition of
schizoaffective disorder.
Finally, the group agreed that ultimately there will need to be a dimensional approach to
psychosis/mood disorders and suggested that research be carried out to determine which
dimensions to use and how they should