When diagnoses enter the DSM, they become reified in the minds of most clinicians in psychiatry and allied fields. This raises special concerns for individuals considered to be at risk for psychosis based on current clinical research criteria. Although current estimates of risk for conversion to psychosis among individuals meeting UHR/COPS criteria are around 35 percent over a two year period (Cannon et al., 2008), these estimates are not yet stable across research samples and appear to be decreasing as the criteria get applied to broader, less clinically acute, community samples. No matter what efforts are made to encourage clinician to exercise caution and restraint in their treatment of individuals who meet UHR criteria, it seems inevitable that we will see more of these individuals treated with pharmacological treatments that are associated with medically significant and often stigmatizing side effects (e.g., motor side effects and obesity). Moreover, when pharmacological treatments are initiated, the research literature...  Read more

When diagnoses enter the DSM, they become reified in the minds of most clinicians in psychiatry and allied fields. This raises special concerns for individuals considered to be at risk for psychosis based on current clinical research criteria. Although current estimates of risk for conversion to psychosis among individuals meeting UHR/COPS criteria are around 35 percent over a two year period (Cannon et al., 2008), these estimates are not yet stable across research samples and appear to be decreasing as the criteria get applied to broader, less clinically acute, community samples. No matter what efforts are made to encourage clinician to exercise caution and restraint in their treatment of individuals who meet UHR criteria, it seems inevitable that we will see more of these individuals treated with pharmacological treatments that are associated with medically significant and often stigmatizing side effects (e.g., motor side effects and obesity). Moreover, when pharmacological treatments are initiated, the research literature provides no endpoint to guide clinicians in connection with how long UHR patients should remain treated. One simply cannot tell at this stage whether a preventative treatment is working, or whether the patient was one of the majority who, in fact, was not at imminent risk for psychosis (indeed, the best data we have to date provide no evidence that psychosis will ever arise in the majority of patients meeting UHR risk criteria). This state of affairs would not only be harmful to the mostly young patients who would almost certainly be targeted for drug interventions, including antipsychotics, it would be detrimental to the public's perception of our field.

A valid counterpoint is often raised that individuals meeting UHR criteria in many of the research samples collected to date are not only at risk, but are also currently in distress and seeking-help. While this certainly justifies treatment, the nature of the treatment is likely to remain more conservative if the criteria are not concretized in the pages of the DSM. We have a responsibility to "first do no harm" when approaching treatment of help-seeking patients in distress. Incorporation of the psychosis-risk criteria into the DSM at this time would seem more likely to increase treatment-related harmful side effects and stigma with much less clarity about increased benefit to patients. The current clinical criteria for UHR states are still evolving research criteria. The patients who meet these criteria are quite heterogenous in their presentation, and most importantly, in their courses and outcomes. We simply don't know enough about UHR states from the standpoint of their clinical boundaries and underlying pathophysiological processes to set the stage for widespread aggressive pharmacological treatments. We need to do much more research to better understand the UHR states before considering placing them in the DSM. Ironically, to the extent that diagnosis of psychosis risk will likely increase the rates of treatment with antipsychotics in the absence of clear data supporting this practice, the resulting medication confounds will impede the progress of the very research that is needed to position us to validly define a psychosis-prodrome syndrome in the DSM to guide clinicians. In my opinion, introduction of psychosis-risk criteria in the DSM-V is quite premature.

References:

Cannon TD, Cadenhead K, Cornblatt B, Woods SW, Addington J, Walker E, Seidman LJ, Perkins D, Tsuang M, McGlashan T, Heinssen R. Prediction of psychosis in youth at high clinical risk: a multisite longitudinal study in North America. Arch Gen Psychiatry. 2008 Jan 1 ; 65(1):28-37. Abstract

In my opinion, there are several arguments to not include a diagnostic “at-risk” category for schizophrenia in DSM-V. In the slidecast presented by Drs. McGlashan and Woods, five characteristic symptoms are listed to define this category that, however, can all be regarded to denote psychotic-like or attenuated psychotic symptoms. It seems to me very premature to define the “at risk” syndrome or schizophrenia prodrome merely on the basis of positive symptoms. Moreover, I think the concept of the schizophrenia prodrome itself can be criticized, also making the inclusion of the risk syndrome in DSM-V a risky business. In the following, I will elucidate on these points from the perspective of my own research on the SSQ (Schizotypic Syndrome Questionnaire) model (Van Kampen, 2005; 2006; Van Kampen et al., 2009a; 2009b). Although I basically agree with the comments made by Drs. Cornblatt and Yung on their slidecasts in rejecting the proposal of...  Read more

In my opinion, there are several arguments to not include a diagnostic “at-risk” category for schizophrenia in DSM-V. In the slidecast presented by Drs. McGlashan and Woods, five characteristic symptoms are listed to define this category that, however, can all be regarded to denote psychotic-like or attenuated psychotic symptoms. It seems to me very premature to define the “at risk” syndrome or schizophrenia prodrome merely on the basis of positive symptoms. Moreover, I think the concept of the schizophrenia prodrome itself can be criticized, also making the inclusion of the risk syndrome in DSM-V a risky business. In the following, I will elucidate on these points from the perspective of my own research on the SSQ (Schizotypic Syndrome Questionnaire) model (Van Kampen, 2005; 2006; Van Kampen et al., 2009a; 2009b). Although I basically agree with the comments made by Drs. Cornblatt and Yung on their slidecasts in rejecting the proposal of Drs. McGlashan and Woods, my views about the schizophrenia prodrome go even beyond some of their comments.

As noted in the above-stated publications, the 12 symptom scales in the SSQ model were specifically selected on the basis of previous studies by, for instance, Docherty et al. (1978) and Yung and McGorry (1996) to describe the main causal pathways in the process of schizophrenic prodromal unfolding. Postulating a detailed network of causal relationships between the selected symptoms against the background of the usual observation that negative symptoms precede positive ones (e.g., Häfner and Maurer, 1991), the SSQ model—which starts with social anxiety and active isolation and ends with the attenuated psychotic symptoms, perceptual disturbances, cognitive derailment, and delusional thinking—could be confirmed both in a general population sample (Van Kampen, 2005) and in a sample of first-episode schizophrenic patients with known IRAOS data (Van Kampen et al., 2009a). The SSQ breaks down in three correlated factors that were termed Negative Schizotypy, Asocial Schizotypy, and Positive Schizotypy (Van Kampen, 2006).

Although the SSQ model was presented as a model of pre-psychotic/prodromal unfolding, there are two issues concerning the status of that model that must be discussed in more detail. The first issue relates to the fact that the SSQ model embraces several symptoms that in present-day prospective investigations of the prodromal state are hardly emphasized. As the SSQ definition of schizotypy comprises both negative, asocial, and psychotic-like symptoms, this is in contrast with the current focus on only the last-mentioned symptom group (e.g., Woods et al., 2001; Miller et al., 2002). Although such a focus is understandable from a treatment perspective, because attenuated psychotic symptoms greatly enhance the risk for psychosis (Yung et al., 2003), the focus is nevertheless one-sided if it comes to the question of whether the resulting psychotic disorder can be diagnosed as schizophrenia. For instance, further analyzing the 10-year longitudinal schizotypy data provided by Chapman et al. (1994), Kwapil (1998) could demonstrate that the negative Revised Social Anhedonia Scaleand—an inventory measuring the same attributes as the SSQ Active Isolation scaleand—may identify individuals at specific risk for future development of schizophrenic spectrum disorders, whereas the positive Chapman scales (e.g., Perceptual Aberration) only had predictive validity for the development of psychosis in general. A similar view on the necessity to complement the positive symptom focus by paying attention to early negative and nonspecific prodromal features can also be found in the work of Barbara Cornblatt and collaborators (e.g., Cornblatt et al., 2003; Lencz et al., 2004). According to Cornblatt et al. (2003), for instance, the component of attenuated psychotic symptoms would not lead to schizophrenia in the absence of such features as cognitive deficits, social isolation, affective disturbances, and school failure that are hypothesized to directly reflect the genetically-based vulnerability core of schizophrenia. Indeed, speculating about the use of whatever schizotypy model in the context of psychosis prevention, investigators may not only emphasize attenuated psychotic symptoms, as is usually done by applying the CAARMS-UHR criteria developed by Yung and collaborators (2002, 2004) or the COPS criteria proposed by Miller et al. (2002), but also those early and non-psychotic manifestations that precede and causally influence the emergence of psychotic symptoms. From the perspective of the SSQ model, criteria may be developed that for instance relate to social anxiety and active isolation, providing of course that these symptoms occur in agreement with the temporal rankings in the SSQ model. It is clear, however, that at present no instrument exists that measures the degree of the “positivation” process indexed by the number of SSQ pathways walked on in the “right” order prior to psychosis (but of course corrected for the number of missing pathways due to the absence of SSQ symptoms as not all patients can be expected to have manifested all SSQ symptoms and/or their temporal orderings as depicted in the SSQ model).

The second issue that I will discuss relates to the first one. However, instead of merely criticizing the positive symptom focus in current prodromal studies, the concept of the prodrome itself will now be criticized, as the putatively prodromal stage might legitimately be considered as a manifestation of schizophrenia without psychosis. As indicated, for instance, by Cornblatt et al. (2003), emphasis on psychotic phenomena in recent definitions of schizophrenia is often misplaced, because of their demonstrated non-pathognomonic status (e.g., Peralta and Cuesta, 1999), their relatively weak or even absent relationship with schizophrenia”s pathophysiology and genetic background (e.g., Tsuang et al., 2000), and especially because many so-called prodromal symptoms coincide with what Kraepelin (1913) and Bleuler (1911) indicated as the most defining characteristics or Grundsymptome of schizophrenia, which may or may not precede the development of psychotic and other accessory symptoms. From this perspective, my additional characterization of the SSQ model as a model that describes the decompensation process in schizophrenia (see Van Kampen, 2009a) seems to be more sensible. Indeed, from that perspective, a description of the full-blown form of schizophrenia might be presented by the inclusion in the SSQ model of three extra pathways: 1) a pathway from cognitive derailment to formal thought disorder; 2) a pathway from perceptual disturbances to hallucinations; and 3) a pathway from delusional thinking to delusions, whereas the non-psychotic form (resembling the traditional diagnosis of simple schizophrenia) coincides with the presently formulated SSQ model. Of course, the extended SSQ model has not yet been tested, but (as indicated above) the fact that psychotic-like symptoms greatly enhance the risk of a transition to psychosis at least supports the possibility of such an extension. Overall, then, I am not so much interested in the impact of my findings with respect to current models of the schizophrenia prodrome as I am in the consequences of having ascertained that the SSQ model shows validity for a proper definition of the schizophrenia construct itself. For instance, from the SSQ model, we may conclude that schizophrenia is best typified both dimensionally and categorically. The necessity of a dimensional representation (see, e.g., Widiger and Samuel, 2005) stems from the fact that the SSQ scales can easily be considered to refer to continua and have even been found to relate to extreme positions on normal personality dimensions (Van Kampen, 2006). Moreover, the demonstration that the SSQ model holds both in a general population sample and in a sample of schizophrenic patients also agrees with the continuum hypothesis. Nevertheless, the same SSQ model, with or without the three additional pathways included, also suggests the necessity of a categorical representation. Although contrary to the beliefs of, for instance, Bentall (1990) and Boyle (1990), who regard schizophrenia as a “scientific delusion”, the SSQ model obviously refers to a distinct cluster of mutually related and causally dependent symptoms, thereby implying a set of regularities for inferring the real existence of a disorder. Thus, summarizing the above, we may conclude that the presently stated SSQ model, rather than a model of prodromal unfolding, offers a valid description of the core dimensional symptoms of schizophrenia as they unfold over time, and that an extension of that model in the direction of psychosis may also encompass those accessory symptoms that are nowadays often mistakenly emphasized. Of course, a conclusion like this does not justify the inclusion of a separate “at risk” category in DSM-V as suggested by Drs. McGlashan and Woods. Rather, I would suggest the inclusion of the “at risk” or “prodromal” state in a broader category of schizophrenia that refers to a set of pre-psychotic (negative, asocial, and psychotic-like) and psychotic symptoms that have been shown to sequentially unfold in time and to determine each other in terms of cause and effect.

References:

Van Kampen D. Pathways to schizophrenic psychosis: A LISREL-tested model of the unfolding of the schizophrenic prodrome. J Clin Psychol 2005; 61: 909-38. Abstract

Van Kampen D. The Schizotypic Syndrome Questionnaire (SSQ): Psychometrics, validation and norms. Schizophr Res 2006; 84: 305-22. Abstract

Van Kampen D, Maurer K, An der Heiden W, Häfner H: Prodromal unfolding: the validation of the Schizotypic Syndrome Questionnaire model in a sample of first-episode schizophrenic patients. Early Interv Psychiat 2009a, 3:143-156.

Van Kampen D, Deijen JB. SPEM dysfunction and general schizotypy as measured by the Schizotypic Syndrome Questionnaire. BMC Neurology 2009b; 9: 27. Abstract

(See for the other references these four publications.)

The risks of including a risk syndrome for psychosis have been well articulated by Drs. Yung and Cornblatt, and I personally agree with most of their points.

However, here I would like to bring more focus on risks of not continuing to consider a risk syndrome for inclusion in DSM-V.

When I give talks about a risk syndrome, some of the most poignant comments from the audience sound like this: “My son/daughter has been suffering with schizophrenia for a long time now. I only wish something like this had been available years ago, and maybe we could have caught it early and things could have gone much better.” I hear the same comments from parents who bring in their younger children for evaluation when an older child has already developed schizophrenia: “We took our older son/daughter to see several doctors when s/he was becoming ill but no one did anything.” If we stay with DSM-IV, a risk is that most patients who are truly on the course to develop schizophrenia will continue to develop a chronic disabling condition without having their early symptoms...  Read more

The risks of including a risk syndrome for psychosis have been well articulated by Drs. Yung and Cornblatt, and I personally agree with most of their points.

However, here I would like to bring more focus on risks of not continuing to consider a risk syndrome for inclusion in DSM-V.

When I give talks about a risk syndrome, some of the most poignant comments from the audience sound like this: “My son/daughter has been suffering with schizophrenia for a long time now. I only wish something like this had been available years ago, and maybe we could have caught it early and things could have gone much better.” I hear the same comments from parents who bring in their younger children for evaluation when an older child has already developed schizophrenia: “We took our older son/daughter to see several doctors when s/he was becoming ill but no one did anything.” If we stay with DSM-IV, a risk is that most patients who are truly on the course to develop schizophrenia will continue to develop a chronic disabling condition without having their early symptoms recognized.

Of course, we really don’t know what to do for these patients yet. In the last decade, with no DSM diagnosis, our field has managed to get funded and published only three randomized treatment/prevention studies for these patients. Even these studies have been tiny and inconclusive. If we conclude that DSM-IV is good enough for these patients for the next decade or so, the risk is that the progress of treatment research will continue to be only incremental.

In my view, inappropriate prescribing is a consequence of the paucity of treatment research. What is a doctor to do when sitting with an ill patient and his/her family, and we have not made available large studies that reveal who does and who does not benefit from which treatment? When we have done too little to develop treatments that could be specific for this phase of illness and are less or non-toxic? That a doctor must rely primarily on clinical judgment in this situation suggests to me that the problem is more with the treatment research database than with the doctor. Thus, in my view, a risk of staying with DSM-IV is a continuation of the current problem with inappropriate prescribing.

Another risk of staying with DSM-IV is that it is not possible from DSM-IV to determine who does not have a risk syndrome. Many of the patients I see who have been prescribed antipsychotic medication in the community do not meet risk syndrome criteria. The community prescriber does not know this because we have not provided him/her with criteria for diagnosing the risk syndrome.

Lastly, another risk of staying with DSM-IV is inadvertently encouraging inappropriate prescribing of other potentially risky mediations. For example, the differential diagnostic guidance in DSM-IV about complaints of attentional impairment does not focus on the possibility of a risk syndrome for psychosis. In my experience, it is fairly common for psychostimulants to be started in these patients who have attentional impairment often beginning de novo in adolescence. Risk syndrome patients can get provisional ADHD diagnoses, partly because ADHD is in DSM-IV and the risk syndrome is not. Similarly, risk syndrome patients can often get depressive disorder NOS diagnoses and SSRI prescriptions. While they may or may not have depressive disorder NOS comorbid with their risk syndrome, it is certainly true that we have little placebo-controlled data suggesting a benefit of SSRIs in adolescents with depressive disorder NOS, and of course there may be a rare risk of suicide promotion with SSRIs in this age group, as evidenced by the FDA black box warnings.

In summary, I think it would be a shame if we did not at least find out from DSM-V field trials whether average clinicians can or cannot use the proposed risk syndrome clinical criteria reliably.

Please see the abstract of the paper in press on validity of the risk syndrome.

Carpenter notes that "It is clear that experts can define risk and that a non-trivial proportion of at-risk individuals convert to a psychotic illness." And he rightly asks if "the evidence [is] sufficient to establish psychosis risk as a diagnostic class within the...DSM-V." Mathalon is concerned about the inevitability of the precipitous use of pharmacology and its stigmatizing side effects. I share that concern, but am also aware of the devastation wrought by schizophrenia in patients and their families where a genetic predisposition was evident from the number of individuals afflicted with schizophrenia and/or related factors. One potential risk factor for conversion that is compelling (at least to some) is marijuana. I say “potential” as it is still debated. Though there have been a number of convincing studies that appear to demonstrate the morbidity of this substance where predisposition is known, there are others who claim the cannabis question is still to be doubted. Nonetheless, I would ask if there might not be a way to make the general public more aware of what is...  Read more

Carpenter notes that "It is clear that experts can define risk and that a non-trivial proportion of at-risk individuals convert to a psychotic illness." And he rightly asks if "the evidence [is] sufficient to establish psychosis risk as a diagnostic class within the...DSM-V." Mathalon is concerned about the inevitability of the precipitous use of pharmacology and its stigmatizing side effects. I share that concern, but am also aware of the devastation wrought by schizophrenia in patients and their families where a genetic predisposition was evident from the number of individuals afflicted with schizophrenia and/or related factors. One potential risk factor for conversion that is compelling (at least to some) is marijuana. I say “potential” as it is still debated. Though there have been a number of convincing studies that appear to demonstrate the morbidity of this substance where predisposition is known, there are others who claim the cannabis question is still to be doubted. Nonetheless, I would ask if there might not be a way to make the general public more aware of what is thus far known about this and other factors currently under investigation. I strongly agree with Carpenter, therefore, that psychosis risk should be evaluated for its inclusion in the DSM-V as a diagnostic class. While the DSM has historically not served as a venue for raising awareness among clinicians, perhaps it should. Greater awareness on the part of clinicians would have some influence where it might really make a difference—which is to say in the area of prevention as well as in that of treatment.

Given the morbidity and difficulty of treating psychotic disorders, including schizophrenia, there has been a move toward identifying and treating adolescents and young adults who appear to be clinically at risk or "prodromal" to psychosis. The field now has greater specificity in identification, with rates of 40-50 percent conversion to frank psychosis within one to two years. There is further evidence that medications and other treatments may have some efficacy for "prodromal" patients, though with variable side effects. However, controversy remains about some of the inherent risks in prodromal research, such as medication exposure and stigma among false-positives. In this paper (Corcoran et al., 2005), we add to this discussion through an analysis of ethics in prodromal research from the more established field of predictive genetic testing. Issues are raised about the effects of information on patients, families, and institutions, as well as future insurability, the limits of confidentiality (as it relies on discretion of...  Read more

Given the morbidity and difficulty of treating psychotic disorders, including schizophrenia, there has been a move toward identifying and treating adolescents and young adults who appear to be clinically at risk or "prodromal" to psychosis. The field now has greater specificity in identification, with rates of 40-50 percent conversion to frank psychosis within one to two years. There is further evidence that medications and other treatments may have some efficacy for "prodromal" patients, though with variable side effects. However, controversy remains about some of the inherent risks in prodromal research, such as medication exposure and stigma among false-positives. In this paper (Corcoran et al., 2005), we add to this discussion through an analysis of ethics in prodromal research from the more established field of predictive genetic testing. Issues are raised about the effects of information on patients, families, and institutions, as well as future insurability, the limits of confidentiality (as it relies on discretion of patients and families), the autonomy of minors with psychiatric symptoms, and even the risks for the true-positive patient.

References:

Corcoran C, Malaspina D, Hercher L. Prodromal interventions for schizophrenia vulnerability: the risks of being "at risk". Schizophr Res. 2005 Mar 1;73(2-3):173-84. Abstract

The psychosis-risk syndrome is worth being recognized by the next DSM because the syndrome is in principle reversible and treatable in its early stages, as shown in animal models of psychotic-like behavior (Shuto et al., 2008; Seeman et al., 2006; 2009). There is, therefore, practical hope for the development of future treatment and desensitization of this syndrome.

References:

Shuto, T., Seeman, P., Kuroiwa, M., Nishi, A. Repeated administration of a dopamine D1 receptor agonist reverses the increased proportions of striatal dopamine D1High and D2High reeptors in methamphetamine-sensitized rats. Eur. J. Neurosci. 27: 2551-2557 (2008). Abstract

Seeman et al. Psychosis pathways converge via D2High dopamine receptors. Synapse 60: 319-346 (2006). Abstract

Seeman, P. Schizophrenia model of elevated D2High receptors: Haloperidol reverses the amphetamine-induced elevation in dopamine D2High. Schizophrenia Res. 91: 191-192 (2009). Abstract

View all comments by Philip Seeman

The idea of helping people when they are at high risk for developing an illness is laudable and has proved its worth many times.

The proposal to have an at-risk diagnosis is, of course, applicable to most of the classification of mental disorders. Individuals might be at risk for dementia, for personality disorders, for depression, for dependence on drugs, and so on.

At present, psychiatric treatment interventions are fairly non-specific for risk states. They are specific if the risks are known and can be removed, but this is not always the case. For most at-risk states the interventions will not be specific. Specifically, an at-risk-for-schizophrenia category might not correspond to a specific treatment intervention that is significantly different from an intervention that could be used to avoid the occurrence of other multifactorially caused disorders.

It would be wise to fight for improved access to mental health services, in particular, in cases where availability of services would in fact help individuals who are at risk for becoming mentally ill. This is a...  Read more

The idea of helping people when they are at high risk for developing an illness is laudable and has proved its worth many times.

The proposal to have an at-risk diagnosis is, of course, applicable to most of the classification of mental disorders. Individuals might be at risk for dementia, for personality disorders, for depression, for dependence on drugs, and so on.

At present, psychiatric treatment interventions are fairly non-specific for risk states. They are specific if the risks are known and can be removed, but this is not always the case. For most at-risk states the interventions will not be specific. Specifically, an at-risk-for-schizophrenia category might not correspond to a specific treatment intervention that is significantly different from an intervention that could be used to avoid the occurrence of other multifactorially caused disorders.

It would be wise to fight for improved access to mental health services, in particular, in cases where availability of services would in fact help individuals who are at risk for becoming mentally ill. This is a political and administrative battle, and thus it is important to assess whether this fight will be helped by having specific at-risk diagnoses.

While the arguments for and against inclusion of a risk syndrome have been well aired in this forum, less attention has been paid to the potential content of the proposed criteria.

Current research criteria for the risk syndrome were developed and operationalized more than a decade ago. The primary basis for diagnosis was presence of a single attenuated (subpsychotic) positive symptom. However, these original criteria have exhibited a pattern of decreasing predictive value as study populations have expanded over time (Yung et al., 2008; Cannon et al., 2008).

At the same time, numerous studies have demonstrated that predictive validity of these criteria can be enhanced by 1) increasing the threshold of attenuated positive symptoms required at baseline; and 2) addition of attenuated negative symptoms, basic symptoms, and/or functional deficits to the prediction model (Cornblatt et al., 2003; Lencz et al., 2003;   Read more

While the arguments for and against inclusion of a risk syndrome have been well aired in this forum, less attention has been paid to the potential content of the proposed criteria.

Current research criteria for the risk syndrome were developed and operationalized more than a decade ago. The primary basis for diagnosis was presence of a single attenuated (subpsychotic) positive symptom. However, these original criteria have exhibited a pattern of decreasing predictive value as study populations have expanded over time (Yung et al., 2008; Cannon et al., 2008).

At the same time, numerous studies have demonstrated that predictive validity of these criteria can be enhanced by 1) increasing the threshold of attenuated positive symptoms required at baseline; and 2) addition of attenuated negative symptoms, basic symptoms, and/or functional deficits to the prediction model (Cornblatt et al., 2003; Lencz et al., 2003; Yung et al., 2003, 2004, 2006; Cannon et al., 2008; Velthorst et al., 2009). Additional efforts at refining and operationalizing these criteria are ongoing.

If a risk syndrome is to be introduced in DSM-V (or any future edition), the existing body of empirical evidence should serve as a useful guide to criteria development prior to initiation of any large-scale field trials.

References:

Cannon TD, Cadenhead K, Cornblatt B, Woods SW, Addington J, Walker E, Seidman LJ, Perkins D, Tsuang M, McGlashan T, Heinssen R. Prediction of psychosis in youth at high clinical risk: a multisite longitudinal study in North America. Arch Gen Psychiatry. 2008 Jan;65(1):28-37. Abstract

Cornblatt BA, Lencz T, Smith CW, Correll CU, Auther AM, Nakayama E. The schizophrenia prodrome revisited: a neurodevelopmental perspective. Schizophr Bull. 2003;29(4):633-51. Abstract

Lencz T, Smith CW, Auther AM, Correll CU, Cornblatt BA. The assessment of "prodromal schizophrenia": unresolved issues and future directions. Schizophr Bull. 2003;29(4):717-28. Abstract

Velthorst E, Nieman DH, Becker HE, van de Fliert R, Dingemans PM, Klaassen R, de Haan L, van Amelsvoort T, Linszen DH. Baseline differences in clinical symptomatology between ultra high risk subjects with and without a transition to psychosis. Schizophr Res. 2009 Apr;109(1-3):60-5. Abstract

Yung AR, Phillips LJ, Yuen HP, Francey SM, McFarlane CA, Hallgren M, McGorry PD. Psychosis prediction: 12-month follow up of a high-risk ("prodromal") group. Schizophr Res. 2003 Mar 1;60(1):21-32. Abstract

Yung AR, Phillips LJ, Yuen HP, McGorry PD. Risk factors for psychosis in an ultra high-risk group: psychopathology and clinical features. Schizophr Res. 2004 Apr 1;67(2-3):131-42. Abstract

I side with Drs. Cornblatt and Yung on this controversy. Substantially more evidence is needed across various domains:

1. Capacity to accurately diagnose those who will convert.

2. Effective strategies for treatment.

3. Evidence that these early interventions produce meaningful and sustained effects over time.

I would not dispute the promising results addressing each of these points—others have alluded to this work. To suggest, though, that the present data substantiate incorporation into DSM-V is simply premature. The elephant in the room seems to be “real world” implications, which I believe would rapidly include further sanctioning (implicit or explicit) of antipsychotics in an age group where such use is already soaring for reasons that are, at best, open to question.

View all comments by Gary Remington

The “at-risk” state is an interesting clinical condition, and the question of including it as a diagnostic category in the DSM-V requires a critical appraisal of the work in the field of psychiatric research (i.e., ICOSR) accomplished over the past 20 years. It is imperative that we unequivocally make a case that “at-risk” psychosis is a clinical condition, which is aetiopathologically distinct, has its own natural history of course, outcome, response to treatment, and does not overlap with several other syndromes. The fundamental issue is whether it is a clinical syndrome or a phase specific condition in the longitudinal course of psychotic disorders.

As the title “at-risk” suggests, this condition is a precursor to a range of psychotic, non-affective, and affective disorders. There is no convincing evidence that at-risk psychosis is a distinct clinical entity. It is also not clear as to what happens to those who do not convert to major psychosis over time. Remarkable work has been done in this area that has brought some reasonable clarity to understanding the...  Read more

The “at-risk” state is an interesting clinical condition, and the question of including it as a diagnostic category in the DSM-V requires a critical appraisal of the work in the field of psychiatric research (i.e., ICOSR) accomplished over the past 20 years. It is imperative that we unequivocally make a case that “at-risk” psychosis is a clinical condition, which is aetiopathologically distinct, has its own natural history of course, outcome, response to treatment, and does not overlap with several other syndromes. The fundamental issue is whether it is a clinical syndrome or a phase specific condition in the longitudinal course of psychotic disorders.

As the title “at-risk” suggests, this condition is a precursor to a range of psychotic, non-affective, and affective disorders. There is no convincing evidence that at-risk psychosis is a distinct clinical entity. It is also not clear as to what happens to those who do not convert to major psychosis over time. Remarkable work has been done in this area that has brought some reasonable clarity to understanding the neurobiological and clinical complexity of psychosis. Having seen what has been presented in the excellent “backgrounder” and from other literature, my thoughts are as follows:

1. The “at risk” psychosis state does not qualify as a “diagnosis.” It’s not a unique illness, nor a syndrome, or a symptom complex.

2. At best, it can be defined as a “stage” of psychosis.

3. This area of research has shown that it is a set of behavioral changes seen amongst individuals who are vulnerable for schizophrenia

4. If DSM-V is going to have any category for “subsyndromal, sub-threshold, and subclinical” conditions, then including an “at-risk” state in such a category may be helpful. However, we need to be mindful of giving a pseudo diagnostic label to someone who does not qualify for any mental disorder, though there are number of other conditions such as psychosomatic illness and psychological factors affecting health which are not mental disorders, but have been included in other classification systems such as the ICD10.

5. It would have been a matter of great satisfaction if we were at stage where mental illness is equivalent to physical disorders in terms of stigma, and other related issues. In such a situation diagnosing a “preclinical” condition could have been a reasonable possibility.

6. Sub-clinical diabetes, sub-clinical hypothyroid states, sub-threshold depression, and a number of other conditions are cases that clearly warrant inclusion in a diagnostic category.

To sum up, I am of the opinion that an “at-risk” state is not an illness and it need not be presumed that it will lead to a major psychotic disorder and therefore has no place in any of the DSM-V diagnostic categories.

I agree with the concerns raised by Barbara Cornblatt. The criteria for identification of youth at ultra high risk of developing psychosis are still in the early stages of development, and as noted by Barbara, this identification or classification of ultra high risk can only be confirmed retrospectively. For the young people who are fortunate not to go on and develop a psychosis, “misidentification” of being at ultra being risk for psychosis can cause much trauma for young people and their families. This can be construed as having a similar traumatic impact as that of hospitalization in an adult psychiatric ward of a young person experiencing a first-episode psychosis. There has been much work to prevent this practice for youth, and I argue we should view placing ultra high risk for psychosis in DSM-V classification as being poor practice.

View all comments by Helen Stain

At this site there are very interesting observations which need to be synthesized. While it appears from a few observations that it may be premature to reject the idea of an at risk syndrome, there are some serious concerns:

1. Is it clarified what exactly will be the benefit of including this condition as a diagnosis in DSM-V?

2. The strongest argument for such inclusion is based upon the “staging model of illness” illustrated by Patrick McGorry. The concept of symptom progression in schizophrenia is neither universally accepted nor popular. It has been mostly “either-or” cases.

3. Is it then on a “severity continuum,” which can be coded independently?

4. There is value in the argument, but to define it as a clinical entity, which is distinct in diagnosis, is premature and non-validated.

5. Most of the objections are on the grounds of the following:
a. Non-specific nature of syndrome
b. Poor validation
c. Low conversion rate
d. High false positive cases
e. Concept of “diagnosis” and “disease”
f. Stigma
g. Inappropriate...  Read more

At this site there are very interesting observations which need to be synthesized. While it appears from a few observations that it may be premature to reject the idea of an at risk syndrome, there are some serious concerns:

1. Is it clarified what exactly will be the benefit of including this condition as a diagnosis in DSM-V?

2. The strongest argument for such inclusion is based upon the “staging model of illness” illustrated by Patrick McGorry. The concept of symptom progression in schizophrenia is neither universally accepted nor popular. It has been mostly “either-or” cases.

3. Is it then on a “severity continuum,” which can be coded independently?

4. There is value in the argument, but to define it as a clinical entity, which is distinct in diagnosis, is premature and non-validated.

5. Most of the objections are on the grounds of the following:
a. Non-specific nature of syndrome
b. Poor validation
c. Low conversion rate
d. High false positive cases
e. Concept of “diagnosis” and “disease”
f. Stigma
g. Inappropriate pharmacotherapy
h. “Label” of mental illness
i. Not a mental health priority
j. Dimensional approach to schizophrenia
k. Detrimental to early intervention initiatives and mental health promotion

I agree with Dr. Cornblatt and Professor Yung that it is premature given the current state of the evidence to include a risk syndrome for psychosis in DSM-V. There are a number of points I would like to raise on the issue:

1. The considerably lower rates of transition to a frank psychotic disorder experienced in international clinics other than the NAPLS group, such as the 10-15 percent quoted in recent years from the PACE clinic in Melbourne (Yung et al., 2007). These rates have significantly reduced from previous years and suggest that in such established clinics, perhaps only one in 10 individuals who meet the “at risk syndrome” would be correctly identified as developing a psychotic illness. Indeed, within the group that does make a transition to a psychotic illness, only around 55 percent develop a schizophreniform psychosis (Cannon et al., 2008). The question should be, Can we justify labeling and treating these individuals when using the present criteria if the positive...  Read more

I agree with Dr. Cornblatt and Professor Yung that it is premature given the current state of the evidence to include a risk syndrome for psychosis in DSM-V. There are a number of points I would like to raise on the issue:

1. The considerably lower rates of transition to a frank psychotic disorder experienced in international clinics other than the NAPLS group, such as the 10-15 percent quoted in recent years from the PACE clinic in Melbourne (Yung et al., 2007). These rates have significantly reduced from previous years and suggest that in such established clinics, perhaps only one in 10 individuals who meet the “at risk syndrome” would be correctly identified as developing a psychotic illness. Indeed, within the group that does make a transition to a psychotic illness, only around 55 percent develop a schizophreniform psychosis (Cannon et al., 2008). The question should be, Can we justify labeling and treating these individuals when using the present criteria if the positive predictive value is so low? This would be acceptable if the treatments we provide were of low risk and toxicity as Dr. Woods suggests. However, even within our own well-established clinic, we have difficulty in persuading our well-informed psychiatrists (not to mention our referring GPs and private psychiatrists) not to initiate antipsychotic medications in those that would meet the “at risk syndrome.” Within our increasingly pressured health services the reality of these individuals receiving CBT, for example, given the number of individuals receiving this for schizophrenia, is low (Lehman et al., 1998). The risks of treating with antipsychotic medications have been well described, and a recent audit of our own first-episode service found average weight gains of 7 kg over one year despite low-dose atypical antipsychotic regimes.

2. The assumption that the “at risk syndrome” represents a homogenous group. This does not appear to be the case with a wide range of diagnostic groupings and outcomes reported in these samples (Rosen et al., 2006). This is important if we are suggesting that a particular treatment may be effective in this population; our own experience and through the clinical trials performed at the PACE clinic is that one size does not fit all for this population. We therefore run the risk of reducing the complexity of presentation of these individuals to the “at risk syndrome” for psychosis rather than their varied clinical needs.

3. The lack of long-term data on the outcome of these individuals. At present the longest follow-up study has been in the order of five years (Phillips et al., 2007). This does not compare favorably with other putative risk syndromes such as mild cognitive impairment and the clinically isolated syndrome (CIS) in neurology where follow-up periods of 10 years have been reported. Longer-term follow-up studies will allow us to be more certain of the symptom course for these individuals.

References:

Cannon, T.D., Cadenhead, K., Cornblatt, B., et al., 2008. Prediction of psychosis in youth at high clinical risk: a multisite longitudinal study in North America. Arch Gen Psychiatry 65, 28-37. Abstract

Lehman, A.F., Steinwachs, D.M., & The Survey Co-Investigators of the, P.P., 1998. Patterns of Usual Care for Schizophrenia: Initial Results From the Schizophrenia Patient Outcomes Research Team (PORT) Client Survey. Schizophr Bull 24, 11-20. Abstract

Phillips, L.J., McGorry, P.D., Yuen, H.P., et al., 2007. Medium term follow-up of a randomized controlled trial of interventions for young people at ultra high risk of psychosis. Schizophr Res 96, 25-33. Abstract

Rosen, J.L., Miller, T.J., D'Andrea, J.T., et al., 2006. Comorbid diagnoses in patients meeting criteria for the schizophrenia prodrome. Schizophrenia Research Vol 85(1-3) Jul 2006, 124-131. Abstract

Yung, A.R., Yuen, H.P., Berger, G., et al., 2007. Declining transition rate in ultra high risk (prodromal) services: dilution or reduction of risk? Schizophr Bull 33, 673-681. Abstract

The intention to include the "at risk syndrome for psychosis" in DSM-V may be a noble one but appears to be misguided, not only because there is clearly lack of evidence for its specificity in terms of prediction as well as lack of any specific treatment, but also because of the potential for social harm. Clinicians once given the criteria for yet another ambiguous syndrome will most likely rush to using antipsychotic medications with all the attendant deleterious effects and unnecessary exposure to such harm. In addition, and perhaps more importantly, the real potential of introducing stigma associated more often with psychotic disorders than with depression, anxiety, etc., and possible rejection of future treatment by young people cannot be ignored.

I am bothered by the use of the term "prodromal" repeatedly in the DSM submission. This is at best intellectually reckless as prodrome (something that becomes before) can only be termed as such retrospectively, after the event it is supposed to precede has occurred. It is even more disconcerting when qualifications such as...  Read more

The intention to include the "at risk syndrome for psychosis" in DSM-V may be a noble one but appears to be misguided, not only because there is clearly lack of evidence for its specificity in terms of prediction as well as lack of any specific treatment, but also because of the potential for social harm. Clinicians once given the criteria for yet another ambiguous syndrome will most likely rush to using antipsychotic medications with all the attendant deleterious effects and unnecessary exposure to such harm. In addition, and perhaps more importantly, the real potential of introducing stigma associated more often with psychotic disorders than with depression, anxiety, etc., and possible rejection of future treatment by young people cannot be ignored.

I am bothered by the use of the term "prodromal" repeatedly in the DSM submission. This is at best intellectually reckless as prodrome (something that becomes before) can only be termed as such retrospectively, after the event it is supposed to precede has occurred. It is even more disconcerting when qualifications such as "severity of prodromal" symptoms are applied to such a nebulous concept.

There is indeed no doubt that within the confines of research programs for UHR individuals we all agree that the severity of their distress is no less than that of patients suffering from a full-fledged psychotic episode, but such distress is associated with non-psychotic symptoms, and we should continue to treat them for conditions such as depression, anxiety, PTSD, substance abuse, etc., and not make these conditions "comorbid" as suggested by the DSM submission.

Last, but not least, criteria of separating conversion to psychosis from UHR are not based on any solid and unbiased measurement where interviewer bias has been removed by blind assessments and the ultimate difficulty of separating one condition from the other by using a one-point difference in rating scales (between sub-threshold and threshold level).

We would be better off leaving the primary conditions that UHR patients suffer from and not promote a condition where the risk for conversion to a condition (psychosis), with reasonably adequate treatment available, is at best less than 30 percent (and most likely in newer programs as low as 10 percent).

We should allow clinicians to continue using their best judgment to treat the primary conditions (depression, anxiety, etc.) in so-called UHR individuals and not confuse them with yet another ambiguous condition which may better be conceptualized more on a continuum. Additionally, for research purposes, let us avoid yet another DSM artificial category.

In addition to echoing Carpenter’s points, I would like to suggest that the “risk syndrome” comprises signs that have not yet differentiated themselves into catatonic, disorganized, or paranoid type. Believers in the risk syndrome could, however, use the present ICD-10 diagnosis of undifferentiated schizophrenia (which has far less focus on deterioration than does DSM-IV’s). This would have the disadvantage of using “schizophrenia,” but would open up the concept of treating people before they differentiate. Restated, the social/occupational dysfunction requirements of DSM-IV are totally arbitrary, and loosening up on that criterion would give “schizophrenia” more of a sense of a disorder, much like the rest of medicine where function doesn’t get confused with diagnosis. Still, this approach would not avoid the issues highlighted by Carpenter.

Additionally, I hope that the DSM-V will not preserve the unfortunate “positive” and “negative” division of schizophrenia. A less problematic division is “florid,” “conative deficits,” and “cognitive deficits.”

View all comments by Roger Peele

I am rather skeptical about the benefits of inclusion of an at-risk state in DSM-V, but relatively convinced regarding many of the costs, so am in support of Professors Yung and Cornblatt. With regard to proposed benefits, UHR criteria already exist, so it is unclear why these cannot be utilized to allow more research to be done (focusing on predictive validity, reliability, and treatment, in the same way that they do currently); if these need additional criteria, such as help-seeking, distress, or impairment, then these could be amended. Similarly, if we require international consensus to allow convergence of the slightly different approaches to defining UHR status, an organization like the International Early Psychosis Association would seem more fit for the purpose than the DSM. The argument that such a category would allow access to treatment in an insurance company-led, managed care context is appealing, but since most of these patients meet criteria for another axis-I disorder (for example, see McGorry et al., 2002), these...  Read more

I am rather skeptical about the benefits of inclusion of an at-risk state in DSM-V, but relatively convinced regarding many of the costs, so am in support of Professors Yung and Cornblatt. With regard to proposed benefits, UHR criteria already exist, so it is unclear why these cannot be utilized to allow more research to be done (focusing on predictive validity, reliability, and treatment, in the same way that they do currently); if these need additional criteria, such as help-seeking, distress, or impairment, then these could be amended. Similarly, if we require international consensus to allow convergence of the slightly different approaches to defining UHR status, an organization like the International Early Psychosis Association would seem more fit for the purpose than the DSM. The argument that such a category would allow access to treatment in an insurance company-led, managed care context is appealing, but since most of these patients meet criteria for another axis-I disorder (for example, see McGorry et al., 2002), these diagnoses could presumably be used to justify payment for treatment.

The declining transition rates mentioned by several others also suggest we would be needlessly stigmatizing, pathologizing, and potentially treating the majority of people meeting such criteria inappropriately (the latter depending on the cost benefit ratio of particular treatments). Current transition rates seem more like 10-20 percent, meaning only a minority develop a psychotic disorder. This is likely to become even more problematic if services start canvassing for case identification in generic youth services and schools. McGorry and colleagues’ (1995) study of adolescents in schools demonstrated high rates of presumably “normal” teenagers meeting DSM-III-R criteria for prodromal states. It is highly likely that this would be the case for any revised definition in DSM-V, even if impaired functioning and help-seeking were included, since emotional difficulties, relationship problems, and trouble with education/employment are pretty common for young people. To introduce such a diagnostic category before extensive research examining the prevalence of these features in the general teenage population would be premature (and possibly irresponsible). This is especially so given the likelihood that this would result in an increase in prescription of antipsychotics, which have considerable risks for young people with developing brains (see Bentall and Morrison, 2002, for a discussion of these risks). It is worrying that those with the most to gain from inclusion of such a diagnosis in DSM-V appear to be researchers (including myself) and pharmaceutical companies, whereas those with the least to gain and most to lose would seem to be young people and their caregivers. Until this balance is reversed, I would strongly oppose the adoption of a risk syndrome for psychosis into the DSM.

References:

Bentall, R. P., and Morrison, A. P. (2002). More harm than good: The case against using antipsychotic drugs to prevent severe mental illness. Journal of Mental Health, 11, 351-365.

McGorry, P. D., McFarlane, C., Patton, G. C., Bell, R., Hibbert, M. E., H.J., J., and Bowes, G. (1995). The prevalence of prodromal features of schizophrenia in adolescence: A preliminary survey. Acta Psychiatrica Scandinavica, 92, 241-249. Abstract

McGorry, P. D., Yung, A. R., Phillips, L. J., Yuen, H. P., Francey, S., Cosgrave, E. M., Germano, D., Bravin, J., McDonald, T., Blair, A., Adlard, S., and Jackson, H. (2002). Randomized controlled trial of interventions designed to reduce the risk of progression to first-episode psychosis in a clinical sample with subthreshold symptoms. Archives of General Psychiatry, 59, 921-928. Abstract

Firstly, it is exciting to see that this debate is even taking place. It is an acknowledgement of the important work that has been undertaken by key researchers to develop strategies to identify and treat this group of individuals. However, despite some potential gains, it would currently be wrong to include an at risk category in the next iteration of the DSM.

As has been discussed in some of the slidecasts, the inclusion of an at risk category would certainly help focus attention on this population, and it would also improve the potential for larger, well-funded trials. But the problems associated with the creation of the at risk category are too great.

Having worked in both research and clinical settings, I have already seen a large numbers of clients who are being prescribed antipsychotic medication due to their being classified as “putatively prodromal.” As suggested, this terminology alone medicalizes these individuals. Additionally, it's problematic that the medication literature given to patients discusses schizophrenia, not risk. For young people in the midst...  Read more

Firstly, it is exciting to see that this debate is even taking place. It is an acknowledgement of the important work that has been undertaken by key researchers to develop strategies to identify and treat this group of individuals. However, despite some potential gains, it would currently be wrong to include an at risk category in the next iteration of the DSM.

As has been discussed in some of the slidecasts, the inclusion of an at risk category would certainly help focus attention on this population, and it would also improve the potential for larger, well-funded trials. But the problems associated with the creation of the at risk category are too great.

Having worked in both research and clinical settings, I have already seen a large numbers of clients who are being prescribed antipsychotic medication due to their being classified as “putatively prodromal.” As suggested, this terminology alone medicalizes these individuals. Additionally, it's problematic that the medication literature given to patients discusses schizophrenia, not risk. For young people in the midst of experiencing some frightening experiences, this is likely to compound their fears. We know these medications will have some fairly serious side effects. Also, the fact that the transition rate is likely to be less than 30 percent is problematic, and we know that studies have found declining conversion rates. The inclusion of an at risk category into the DSM will increase the number of young people classified as being at risk for psychosis and subsequently the number prescribed antipsychotic medication.

More time is required to refine the assessment and treatment strategies for individuals at risk for psychosis before this category is included in the DSM.

The concept of a psychosis risk syndrome based on 15 years of research is a valid one which is breaking new ground in psychiatry. It is congruent with clinical staging approaches as practiced in oncology and other areas of medicine to assist in treatment selection in early stages of illness when the disorder is more susceptible to successful treatment. There is a similar attraction to this approach in psychiatry. Based on our own work in Australia in the early 1990s and with refinement through important subsequent U.S. and other international work, a set of criteria has been developed which not only defines a group of people with an immediate need for non-specific symptom- and needs-based care, but who also carry a greatly enhanced risk of progression over the short term to more severe and often (but not always) more persistent psychotic symptoms, often qualifying for a later diagnosis of schizophrenia or a related illness. The risk is elevated up to 400-fold under certain sampling conditions; however, the false positive rate for psychosis is appreciable, ranging from 60 to 90...  Read more

The concept of a psychosis risk syndrome based on 15 years of research is a valid one which is breaking new ground in psychiatry. It is congruent with clinical staging approaches as practiced in oncology and other areas of medicine to assist in treatment selection in early stages of illness when the disorder is more susceptible to successful treatment. There is a similar attraction to this approach in psychiatry. Based on our own work in Australia in the early 1990s and with refinement through important subsequent U.S. and other international work, a set of criteria has been developed which not only defines a group of people with an immediate need for non-specific symptom- and needs-based care, but who also carry a greatly enhanced risk of progression over the short term to more severe and often (but not always) more persistent psychotic symptoms, often qualifying for a later diagnosis of schizophrenia or a related illness. The risk is elevated up to 400-fold under certain sampling conditions; however, the false positive rate for psychosis is appreciable, ranging from 60 to 90 percent depending on how the sample has been selected. If based on help seekers at early psychosis programs, then the false positive rate is acceptable at 60 percent. If samples are drawn from broader youth mental health settings, or worse still from the general community, then the rate is much higher, up to 90 percent plus. On the other hand, while these people may be false positives initially for psychosis, they are very frequently true positives for other persistent and often severe mood and related disorders. Hence, the spectrum of outcome risk even with risk criteria targeting non-affective psychosis is broad. This means that the concept of the risk syndrome, while it does possess validity, is trying to be too specific too early in the staging spectrum of illness. There should ideally be a psychotic/persistent mood disorder risk syndrome whose criteria should blend the current UHR criteria with additional items connoting risk for incipient bipolar and severe or persistent depressive disorder. This may need more research before it can be properly formulated.

If there is a decision to proceed with the narrower psychosis risk syndrome, then certain safeguards need to be put in place. Firstly, it should only be applied under conditions where help is being sought from an early intervention/early psychosis or youth mental health service and certainly not in school or general population settings, where the true base rate for psychosis is going to be so low as to render the criteria useless for identification of the group truly at risk. This is the “needle in the haystack” problem. We will only find needles when the haystack is small enough and the density of needles has been increased or enriched somehow. Help-seeking is one enrichment strategy, and there may be others, for example, sequential screening. Secondly, there is a risk, especially in countries where a highly reductionistic approach which equates treatment with drug therapy is dominant, that those meeting the psychosis risk syndrome will be treated with antipsychotics. Indeed, this is already occurring. The transition rates (falling in many centers to as low as 10 percent) and evidence base we have assembled so far means that the risk benefit ratio for use of antipsychotic medications in the UHR or psychosis risk syndrome is unfavorable. They should currently be withheld until transition, i.e., sustained psychosis, has developed (and then, of course, promptly commenced to minimize DUP). Placing the psychosis risk syndrome in the DSM-V without these caveats and warnings could lead to widespread overtreatment with medications which possess many harmful as well as beneficial effects.

Let's do a cost-benefit analysis.

In prodromal research, we consider true positives to be those at-risk individuals who develop the disorder, and false positives to be those who do not. In prodromal research, we know that the false positive rate was initially high and that it is increasing further.

However, in considering adding the psychosis risk syndrome to the DSM, we must consider yet another distinction between true and false positives. In this case, true positives are those individuals identified correctly as at risk (with academic centers as the gold standard) and false positives are those who are incorrectly believed to be at risk (incorrect use of the criteria, etc.). We don't have any sense yet what the false positive rates will be in terms of clinicians in the community, but per Michael First in a personal communication, in the best outcomes for previous new DSM diagnoses in field trials, typically this is 65-75 percent.

Putting these false positive rates together—say a 20 percent true positive conversion rate (for the academic prodromal...  Read more

Let's do a cost-benefit analysis.

In prodromal research, we consider true positives to be those at-risk individuals who develop the disorder, and false positives to be those who do not. In prodromal research, we know that the false positive rate was initially high and that it is increasing further.

However, in considering adding the psychosis risk syndrome to the DSM, we must consider yet another distinction between true and false positives. In this case, true positives are those individuals identified correctly as at risk (with academic centers as the gold standard) and false positives are those who are incorrectly believed to be at risk (incorrect use of the criteria, etc.). We don't have any sense yet what the false positive rates will be in terms of clinicians in the community, but per Michael First in a personal communication, in the best outcomes for previous new DSM diagnoses in field trials, typically this is 65-75 percent.

Putting these false positive rates together—say a 20 percent true positive conversion rate (for the academic prodromal designation) and a 35 percent true positive correct clinician identification rate (given the syndrome had its best shot in field trials)—only ~ 7 of 100 individuals identified by community clinicians will be at real risk for psychosis. And ~ 93 of 100 will not.

Next, let's consider what these 100 individuals are likely to experience. As yet, there is no evidence base for treatment for psychosis risk, so all 100 have no clear benefit from being identified as such and all incur the risk of being stigmatized and of exposure to unnecessary medications (including the 7!)

We don't know yet what is the stigma experienced by a young person in being identified as at risk for a serious and persistent mental illness, including schizophrenia. In academia, we strive to minimize the stigma of the designation but our success in this remains an empirical question. We don't have a clue what will happen in the community in terms of stigmatization and its effects, including both social exclusion and financial penalty.

Also, we can surmise that many (most?) individuals who receive this risk designation will be prescribed antipsychotic medications. If a person is at risk for a “something,” wouldn’t it make sense to give an anti-“something”? To take that further, what would the potential liability be for a psychiatrist who chose to watch and wait with an individual deemed at risk for psychosis who then went on to become psychotic? The unintended consequences of our good faith efforts could be a huge cohort of teens and young adults prescribed antipsychotic medications indefinitely and incurring their many side effects.

My conclusion is that it would be premature to reify this syndrome in the DSM-V. However, given the promise of prodromal research, psychosis risk could be included in the appendix of the DSM-V as a provisional syndrome that requires further investigation and clarification.

Schizophrenic disorder has been known as a psychotic spectrum disorder that encompasses heterogeneous clinical phenomena and diverse neurobiological pathogenesis. This has resulted in a long, ongoing debate as to whether these varied clinical manifestations constitute unique clinical phenomena with distinct neurobiological bases, or alternatively, are part of a broader schizophrenic spectrum process. Modifications to more permissible diagnostic revisions in earlier DSMs and increasing emphasis on the recognition and treatment of specific symptoms in recent years have advanced our understanding and clinical care of schizophrenic patients, as they heightened clinician awareness and research interest. Extensive epidemiological, clinical, and more recently genetic studies, have shown that the currently recognized schizophrenic symptoms often overlap with affective and anxiety disorders such as depressive, obsessive-compulsive, and panic disorders, as well as some impulse control disorders.

While such comorbid schizophrenic patients may constitute a distinct schizophrenic...  Read more

Schizophrenic disorder has been known as a psychotic spectrum disorder that encompasses heterogeneous clinical phenomena and diverse neurobiological pathogenesis. This has resulted in a long, ongoing debate as to whether these varied clinical manifestations constitute unique clinical phenomena with distinct neurobiological bases, or alternatively, are part of a broader schizophrenic spectrum process. Modifications to more permissible diagnostic revisions in earlier DSMs and increasing emphasis on the recognition and treatment of specific symptoms in recent years have advanced our understanding and clinical care of schizophrenic patients, as they heightened clinician awareness and research interest. Extensive epidemiological, clinical, and more recently genetic studies, have shown that the currently recognized schizophrenic symptoms often overlap with affective and anxiety disorders such as depressive, obsessive-compulsive, and panic disorders, as well as some impulse control disorders.

While such comorbid schizophrenic patients may constitute a distinct schizophrenic subtype, or a part of schizophrenic spectrum disorder, it warrants specific diagnostic considerations that will facilitate clinical management until it can be better defined and classified. A high risk category for psychosis should also include patients with well-recognized comorbid psychiatric conditions in schizophrenia that may eventually evolve into psychotic process during the course of the illness. This subgroup of psychotic affective (manic-depressive) and anxiety disorder patients often exhibit unique clinical profiles such as poor treatment response, greater impairment of cognitive function, and worse long-term outcome.

In 1900 Freud declared dreams to be the royal road to the unconscious and then went on to elaborate psychodynamic psychiatry from that initial observation. Now in 2000 we have the risk syndrome for first psychosis and we can identify it reliably and follow it with scientific precision in ways that it can become our prospective royal road into the machinations of psychosis.

I recognized the power of this perspective the first time I saw Alison Yung's work at the PACE Clinic in 1994, and I have pursued this line of inquiry ever since. I still maintain my initial enthusiasm and I advocate strongly that the ultimate benefit of promoting this perspective far outweighs the risk.

I cannot argue at all with the caution expressed in many, if not most, of the critiques of nosologizing the risk syndrome, but I believe there is an almost exclusive focus here on the risks for the false positive person/patient. Little concern or question has emerged about the true positive person in the risk syndrome discussions. All of the concern centers around the dangers of making a mistaken...  Read more

In 1900 Freud declared dreams to be the royal road to the unconscious and then went on to elaborate psychodynamic psychiatry from that initial observation. Now in 2000 we have the risk syndrome for first psychosis and we can identify it reliably and follow it with scientific precision in ways that it can become our prospective royal road into the machinations of psychosis.

I recognized the power of this perspective the first time I saw Alison Yung's work at the PACE Clinic in 1994, and I have pursued this line of inquiry ever since. I still maintain my initial enthusiasm and I advocate strongly that the ultimate benefit of promoting this perspective far outweighs the risk.

I cannot argue at all with the caution expressed in many, if not most, of the critiques of nosologizing the risk syndrome, but I believe there is an almost exclusive focus here on the risks for the false positive person/patient. Little concern or question has emerged about the true positive person in the risk syndrome discussions. All of the concern centers around the dangers of making a mistaken attribution and our need to hide the risk syndrome under a basket in order to avoid such an event. But how do the relative risks play out between true and false positive persons? For the false positive the biggest risks of applying the risk syndrome criteria is weeks, months, or a year or two of undue anxiety, perhaps stigma, and inappropriate biological treatment. For the true positive the risk of not applying the risk syndrome criteria could well be an uninformed, uncontrolled, unanticipated first psychotic break destructive to limb and reputation leading to the treadmill of revolving door treatments and ultimately to the irreversible brain deficits of chronic schizophrenia. Given this, how come everyone focuses on the false positives when their relative risk is miniscule compared to that of the true positives?

Perhaps this all distills down to one question I would put to all of us on this symposium. What would you do if your son or daughter met risk syndrome criteria?

Thanks for your attention.

There is always a concern about treating individuals at risk for schizophrenia with potent drugs, especially given that the majority of those individuals will not transition to psychosis. On the other hand, there is evidence from Johnstone et al. (2002) that among children at risk for developing psychosis, the ones that show abnormally high stress reactivity are the individuals most likely to convert. We have found that stress can affect the dopamine system (Valenti and Grace, 2008) in a manner analogous to what occurs in our animal models of schizophrenia (Lodge and Grace, 2008). This is consistent with a model we had advanced that, along with others, posits a means by which stress could lead to psychosis in susceptible individuals (Thompson et al., 2004). Based on these data, we examined whether treating the already-present anxiety component could circumvent the onset of psychosis in an animal model. We now have preliminary evidence that...  Read more

There is always a concern about treating individuals at risk for schizophrenia with potent drugs, especially given that the majority of those individuals will not transition to psychosis. On the other hand, there is evidence from Johnstone et al. (2002) that among children at risk for developing psychosis, the ones that show abnormally high stress reactivity are the individuals most likely to convert. We have found that stress can affect the dopamine system (Valenti and Grace, 2008) in a manner analogous to what occurs in our animal models of schizophrenia (Lodge and Grace, 2008). This is consistent with a model we had advanced that, along with others, posits a means by which stress could lead to psychosis in susceptible individuals (Thompson et al., 2004). Based on these data, we examined whether treating the already-present anxiety component could circumvent the onset of psychosis in an animal model. We now have preliminary evidence that peripubertal administration of a benzodiazepine (Fox and Grace, 2009) to rats that had been given mitotoxin methylazoxymethanol acetate prenatally (Moore et al., 2006) prevented the increase in dopamine neuron activity observed in the adult. If follow-up experiments in rats are consistent, this suggests that a relatively benign intervention to treat an already-present condition (i.e., stress hyperresponsivity) may be an effective preventative measure in those at risk for psychosis.

References:

Fox KM, Grace AA (2009) Peripubertal administration of diazepam prevents dopaminergic pathophysiology in the MAM developmental model of schizophrenia 2009 Neuroscience meeting Planner, Washington, DC: Society for Neuroscience, 2009. in press.

Johnstone EC, Lawrie SM, Cosway R (2002) What does the Edinburgh High-Risk Study tell us about schizophrenia? American Journal of Medical Genetics (Neuropsychiatric Genetics) 114:906-912. Abstract

Lodge, D.J. and Grace, A.A. (2008) Augmented hippocampal drive of mesolimbic dopamine neurons: A mechanism of psychostimulant sensitization. Journal of Neuroscience 28 7876-7882. Abstract

Moore, H., Ghajarnia, M.E., Jentsch, J.D., Geyer, M.A. and Grace, A.A. (2006) A Neurobehavioral Systems Analysis of Adult Rats Exposed to Methylazoxymethanol Acetate on E17: Implications for the Neuropathology of Schizophrenia. Biological Psychiatry 60: 253-264. Abstract

Thompson, J.L., Pogue-Geile, M.F. and Grace, A.A. (2004) Developmental pathology, dopamine, and stress: A model for the age of onset of schizophrenia symptoms. Schizophrenia Bulletin 30: 875-900. Abstract

Valenti, O. and Grace, A.A. (2008) Acute and repeated stress induce a pronounced and sustained activation of VTA DA neuron population activity. Program No. 479.11. 2008 Neuroscience meeting Planner, Washington, DC: Society for Neuroscience, 2008. Online.

During regular periodic revisions of the diagnostic systems, categories are assessed in terms of conformity with the current state of knowledge. In the event that new insights were gained since the last revision, categories are revised and adjusted.

Regarding the abnormalities that occur prior to psychotic episodes, the DSM-III listed eight symptoms and the DSM-III-R listed nine symptoms, of which at least two are needed to be evident in addition to a clear deterioration of performance, in order to fulfill the criteria for a prodromal phase. These symptoms, which at the same time were also utilized for the definition of the residual phase, were basically concordant with the symptoms used to diagnose a "schizotypal personality disorder" in DSM-III, DSM-III-R, and again in DSM-IV and its text revision. Additionally, evidence of these symptoms achieved a positive predictive power between 36 percent and 48 percent in the most thorough retrospective prediction analyses completed at that time (Jackson et al., 1995).

Similar to...  Read more

During regular periodic revisions of the diagnostic systems, categories are assessed in terms of conformity with the current state of knowledge. In the event that new insights were gained since the last revision, categories are revised and adjusted.

Regarding the abnormalities that occur prior to psychotic episodes, the DSM-III listed eight symptoms and the DSM-III-R listed nine symptoms, of which at least two are needed to be evident in addition to a clear deterioration of performance, in order to fulfill the criteria for a prodromal phase. These symptoms, which at the same time were also utilized for the definition of the residual phase, were basically concordant with the symptoms used to diagnose a "schizotypal personality disorder" in DSM-III, DSM-III-R, and again in DSM-IV and its text revision. Additionally, evidence of these symptoms achieved a positive predictive power between 36 percent and 48 percent in the most thorough retrospective prediction analyses completed at that time (Jackson et al., 1995).

Similar to the preceding DSM versions, the currently valid 1994 DSM-IV mentions the prodromal phase in the duration criterion (C) for schizophrenia, but waives the listing of the individual prodromal symptoms. Instead, it only exemplarily refers to attenuated forms of the positive symptoms used for the A-criterion. The A-criterion of the "criteria for the risk syndrome for first psychosis,” suggested by T. McGlashan and S. Woods, again contains the same attenuated positive symptoms in a more differentiated form that were already used for the definition of the prodromal phase in the DSM-III, DSM-III-R, and DSM-IV. However, if these symptoms constituted an additional diagnostic category, this would lead to an unprecedented upgrading in the schizophrenia chapter of the upcoming fifth version of the DSM.

Such a risk syndrome could best be compared with the category of "mild cognitive impairment (MCI),” which is nosologically non-specific, but often already in need of treatment. This condition indicates a risk for Alzheimer's dementia with a transition rate comparable to the risk syndrome for first psychosis. Therefore, MCI plays a role in the development of early recognition and prevention strategies in dementia comparable to the "at-risk-mental states (ARMS)" in the field of schizophrenia (Klosterkötter, 2008). In line with comparison and with regard to the suggested symptoms, for clinical purposes, it would perhaps be more appropriate to simply replace the suggested “risk syndrome for first psychosis” category in DSM-V with a "mild psychotic impairment” category.

McGlashan and Woods’s proposal addresses a very important issue within the field of psychiatry. Yet, does this sufficiently reflect the current state of knowledge in schizophrenia research? This question, which was also raised by a number of other commentators prior to and during the SRF online discussion, can be answered through the research efforts focused on early recognition and early treatment. In contemporary research, we generally no longer follow the conventional nosological rather than a molecular paradigm in medicine. Risk identification and prevention will increasingly replace diagnostics and therapy.

As is the case in many other physical and mental disorders, schizophrenia turned out to be a complex disorder with polygenic heritage and strong pathogenic influence of gene-gene as well as gene-environment interactions. At present, a new comprehensive cooperative project, supported by the European Commission (EU-GEI), is being initiated to analyze these interactions and their influence on the development of first-episode psychosis. This project also includes a large cohort of prodromal patients in accordance with the UHR-criteria (van Os, 2008). However, currently none of the already known molecular-genetic, neurobiological, and psychosocial risk factors alone (which would constitute the clinical stage 0 in the clinical staging model framework for psychotic and severe mood disorders as suggested by McGorry et al., 2006) seems to have sufficient predictive power for prevention. Therefore, prevention becomes possible when, at around age 16, the initial prodrome starts, lasting on average five to six years with many uncharacteristic, primarily affective symptoms accompanied by specific cognitive disturbances. A prospective study on the predictive validity of basic symptoms with an average 9.6 years of follow-up demonstrated a specificity of 85 percent and higher, a positive predictive power of at least 70 percent, and low false-positive rates (<7.5). After 12 months, 20 percent of the participants seeking medical help with such symptoms had developed a schizophrenic disorder, a further 17 percent after 24 months, another 13 percent after 36 months, and finally 70 percent after an average of 4.5 years (Klosterkötter et al., 2001). Based on these findings, a new "schizophrenia proneness instrument" with an adult version (SPI-A) was conceived as well as a child and youth version (SPI-CY), and cognitive and perceptive basic symptoms were proposed as "risk criteria" suitable for this early stage of the initial prodrome (Schultze-Lutter et al., 2007). Interestingly, this early stage of the initial prodrome corresponds with the clinical stage 1a of the Melbourne clinical staging model.

The attenuated psychotic symptoms proposed for inclusion in the DSM-V A-criterion for the new risk syndrome are partly overlapping but predominantly not occurring before stage 1b of the clinical staging model (Schultze-Lutter et al., 2008). In the European Prediction of Psychosis Study (EPOS, n = 245), a prospective study simultaneously investigating basic symptoms and UHR-criteria as inclusion criteria (Klosterkötter et al., 2005), combined the use of both criteria and resulted in a more sensitive and specific definition of the risk syndrome. Even more important for the further development of the indicated prevention was the development of a prognostic index, introduced for the first time in this study (Ruhrmann et al., in press). Similar to somatic medicine, e.g., oncology, such an index can be used to determine risk categories that allow for a more individualized estimate of the probability and time expected to pass until transition into first-episode psychosis.

Based on the most recent research, inclusion of the risk syndrome in the DSM-V would therefore only represent a comparatively short and late period in the development of psychosis over several risk stages. One could argue that this is a first step, and in relation to the delitescence of the prodromal symptoms in the DSM-IV C-criterion, an advancement. But, from my point of view, this proposal can only be approved if the complete diagnosis of schizophrenia were altered according to a clinical staging model. For example, according to the Melbourne concept, three stages aside from the risk stage ought to be implemented in such a model: 1) the first psychotic episode itself, 2) the critical period, 3a) stages of incomplete remission, 3b) recurrence or relapse, 3c) or multiple relapses; and 4) persistent and no longer remitting schizophrenia. For such a staging there are indeed many likewise arguable alternatives that could be invoked in creating a staging model.

A diagnostic structure of this kind would then make it possible to insert proposals for the definition of a risk stage or certain periods of such stages like the suggested risk syndrome within the context of the schizophrenia diagnosis. The advantage would be that this syndrome will remain distinguishable in its restricted and preliminary significance for the risk identification and it would not be viewed as an independent category with an already verified diagnostic and therapeutic validity. It is doubtful, though, that decisions will be made in the near future to fundamentally alter the DSM diagnoses based on a clinical staging model. Significant empirical prerequisites for this type of modeling are missing. The missing empirical prerequisites become more salient when one considers that this type of staging could potentially be developed for other diagnostic categories within the DSM. Furthermore, the clinical staging component ought to be compatible with the endophenotype approach that leads to other diagnostic consequences (Klosterkötter, 2008).

An independent diagnostic category, however, as suggested by McGlashan and Woods, would evoke all the dangers of the diagnostic and therapeutic misuse as specified in many of the comments. Its inclusion in DSM-V could only, if at all, be accepted, if as suggested by Corcoran in her commentary, the category were listed in the appendix under research criteria and provided with warning instructions as mentioned in McGorry’s commentary.

References:

Jackson HJ, McGorry PD, Dudgeon P. Prodromal symptoms of schizophrenia in first-episode psychosis. Prevalence and specificity. Compr Psychiatr 1995; 36: 241-250. Abstract

Klosterkötter J. Indicated Prevention of Schizophrenia. Dtsch Arztebl Int 2008; 105(30):532–9. Abstract

van Os J, Murray R. Gene-environment interactions in schizophrenia. Introduction. Schizophr Bull 2008 Nov; 34(6):1064-5. Abstract

McGorry PD, Hickie IB, Yung AR, Pantelis C, Jackson HJ. Clinical staging of psychiatric disorders: a heuristic framework for choosing earlier, safer and more effective interventions. Aust N Zeal J Psychiatry 2006; 40:616-22. Abstract

Klosterkötter J, Hellmich M, Steinmeyer EM, Schultze-Lutter F: Diagnosing schizophrenia in the initial prodromal phase. Arch Gen Psychiatry 2001; 58:158-164. Abstract

Schultze-Lutter F, Klosterkötter J, Picker H, Steinmeyer EM, Ruhrmann S. Predicting first-episode psychosis by basic symptom criteria. Clinical Neuropsychiatry 2007; 4,1:11-22.

Schultze-Lutter F, Ruhrmann S, Berning J, Maier W, Klosterkötter J. Basic Symptoms and Ultrahigh Risk Criteria: Symptom Development in the Initial Prodromal State. Schizophr Bull 2008 Jun 25. Abstract

Klosterkötter J, Ruhrmann S, Schultze-Lutter F, Salokangas RK, Linszen D, Birchwood M, Juckel G, Morrison A, Vazquez-Barquero JL, Hambrecht M, von Reventlow H: The European Prediction of Psychosis Study (EPOS): integrating early recognition and intervention in Europe. World Psychiatry 2005; 4:161-167. Abstract

Ruhrmann S, Schultze-Lutter F, Salokangas RK, Heinimaa M, Linszen D, Dingemans PM, Birchwood M, Patterson P, Juckel G, Heinz A, Morrison A, Lewis S, Graf von Reventlow H, Klosterkötter J. Prediction of psychosis in adolescents and young adults at high risk: results from the prospective European Prediction of Psychosis Study (EPOS). Arch Gen Psychiatry, in press.

Klosterkötter J. The clinical staging and the endophenotype approach as an integrative future perspective for psychiatry. World Psychiatry 2008 October; 7:3, pp 159-160.

At-risk states can be viewed in one of two ways. Namely, the "glass half full,” which translates into a low stigma, endangered health status; and the "glass half empty," which corresponds to a high stigma, emerging disease status.

The present structure of the DSM is designed to support only the "glass half empty" meaning of the at-risk syndrome.

An important question is whether adoption of the proposed at-risk syndrome into the next edition of the DSM will take place through a process of assimilation in which the syndrome will be pigeon-holed into the current disease oriented Procrustean bed of the DSM. Or, alternatively, will the syndrome be adopted through a process of accommodation in which the DSM will adapt its schema to accommodate both meanings of the at-risk state.

Like professors Cornblatt and Yung, and the many other commentators who share their concerns, I agree that the risks associated with adopting a psychosis risk syndrome into the DSM outweigh its potential benefits.

I believe that such adoption may be less risky if the DSM adapts itself to...  Read more

At-risk states can be viewed in one of two ways. Namely, the "glass half full,” which translates into a low stigma, endangered health status; and the "glass half empty," which corresponds to a high stigma, emerging disease status.

The present structure of the DSM is designed to support only the "glass half empty" meaning of the at-risk syndrome.

An important question is whether adoption of the proposed at-risk syndrome into the next edition of the DSM will take place through a process of assimilation in which the syndrome will be pigeon-holed into the current disease oriented Procrustean bed of the DSM. Or, alternatively, will the syndrome be adopted through a process of accommodation in which the DSM will adapt its schema to accommodate both meanings of the at-risk state.

Like professors Cornblatt and Yung, and the many other commentators who share their concerns, I agree that the risks associated with adopting a psychosis risk syndrome into the DSM outweigh its potential benefits.

I believe that such adoption may be less risky if the DSM adapts itself to accommodate the at-risk state by creating a new class of diagnoses that place more emphasis on an individual's endangered health status as opposed to hypothetical emerging disease.

Rather than including a “psychosis risk syndrome,” a category titled “endangered contact with reality syndrome” could be incorporated into the DSM. Adoption of a risk syndrome of this nature into the DSM seems worthy of consideration for two reasons. First, because the emphasis on the healthy side of the at-risk state is less stigmatizing and less likely to result in unnecessary drug treatments. Secondly, because the emphasis on endangerment will convey a stronger sense of medical necessity than V-code diagnoses, which are currently used for conditions that call for medical attention but are not considered disorders.

In conclusion, early identification of risk and consequent intervention in psychiatry can be viewed as a prospect of endangered health conservation and/or a prospect of emerging disease prevention. The first view is both more strongly supported by present data and also less stigmatizing than the second one. Hence, the appropriateness of adopting an at-risk category into the DSM will ultimately be dependent on the extent to which it is able to promote the health conservation side of early identification and intervention and not just disease prevention.

Weighing the benefits and harms of adding a "prodrome" for schizophrenia to the DSM as a separate diagnostic category is important at this time. The comments thus far reveal the values that many of us prioritize and the assumptions we bring to this issue. Early intervention may not always produce improved outcomes. We need to examine more closely under which conditions early interventions are more beneficial than harmful. If the screening, diagnosis, evidence base, and treatment outcomes are not robust, and if there are considerable social and medical harms probable or evident, early intervention may do more harm than good—it cannot be considered good public health policy. Recently, we have seen the controversy around possible and probable benefits and harms of breast and cervical cancer screening, with differing opinions based on competing views of the evidence base and differing values. The evidence base for these interventions is larger and clearer than for any psychiatric diagnoses. Also, long-term outcomes may not be better for early treatment versus delayed...  Read more

Weighing the benefits and harms of adding a "prodrome" for schizophrenia to the DSM as a separate diagnostic category is important at this time. The comments thus far reveal the values that many of us prioritize and the assumptions we bring to this issue. Early intervention may not always produce improved outcomes. We need to examine more closely under which conditions early interventions are more beneficial than harmful. If the screening, diagnosis, evidence base, and treatment outcomes are not robust, and if there are considerable social and medical harms probable or evident, early intervention may do more harm than good—it cannot be considered good public health policy. Recently, we have seen the controversy around possible and probable benefits and harms of breast and cervical cancer screening, with differing opinions based on competing views of the evidence base and differing values. The evidence base for these interventions is larger and clearer than for any psychiatric diagnoses. Also, long-term outcomes may not be better for early treatment versus delayed treatment where the correct diagnosis and prognosis are not clear or the evidence of further deterioration is not robust. For example, in epilepsy research, the question of the necessity of treating first seizures and the questions of treatment delay resulting in more resistant disease is far from clear (Feksi et al., 1991; Musicco et al., 1997; Sathasivam et al., 2008; Marson et al., 2005; French et al., 2004; Wilby et al., 2005; Chang et al., 2008). Also, we need to be more logical: we cannot equate the schizophrenia prodrome to aspects of metabolic syndrome. The screening process for many components of the metabolic syndrome is objective, and there is a significant body of research on drug and lifestyle interventions to reduce morbidity. In the case of the schizophrenia “prodrome,” I would argue that screening for the “prodrome” has poor predictive value and is non-specific, and the social and medical harm from the "diagnosis” that will immediately be considered as another psychiatric illness is substantial. Additionally, based on the fact that the "treatment" of the “prodrome” is almost exclusively drug treatment with clear evidence of medical harm, the addition of the prodrome to the DSM is not ready for scientific prime time. Extending the "spectrum" of the schizophrenia diagnosis in psychiatry is not a reasonable move at this time. If we frame the questions more broadly and include the theories of the social determinants of mental health (CSDH, 2008), we could learn that there are many interventions possible. Even if we limit the interventions to only biological ones, we need to consider sleep; light; essential nutrients which are not produced in the human body; chronic, multiple, and unrelenting stressors; circadian rhythms; neighborhood characteristics; and the built environment, and other biological aspects of human emotions and behaviors.

References:

Feksi AT, Kaamugisha J, Sander JW, Gatiti S, Shorvon SD. Comprehensive primary health care antiepileptic drug treatment programme in rural and semi-urban Kenya. ICBERG (International Community-based Epilepsy Research Group). Lancet. 1991 Feb 16;337(8738):406-9. Abstract

Musicco M, Beghi E, Solari A, Viani F. and First Seizure Trial Group (FIRST Group)* Treatment of first tonic-clonic seizure does not improve the prognosis of epilepsy. Neurology 1997;49:991-998. Abstract

Sathasivam S, Nicolson A. First seizure - to treat or not to treat? Int J Clin Pract. 2008 Dec;62(12):1920-5. Abstract

Marson A, Jacoby A, Johnson A, et al: Immediate versus deferred antiepileptic drug treatment for early epilepsy and single seizures: a randomized controlled trial. Lancet 365:2005. Abstract

French JA, Kanner AM, Bautista J, et al: Efficacy and tolerability of the new antiepileptic drugs I: treatment of new onset epilepsy. Report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 62:1252, 2004. Abstract

Wilby J, Kainth A, Hawkins N, et al: Clinical effectiveness, tolerability and cost-effectiveness of newer drugs for epilepsy in adults: a systematic review and economic evaluation. Health Technol Assess 9:15, 2005. Abstract

Chang YC, Huang CC, Huang SC. Long-term neuroplasticity effects of febrile seizures in the developing brain. Chang Gung Med J. 2008 Mar-Apr;31(2):125-35. Abstract

CSDH (2008). Closing the gap in a generation: health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. Geneva, World Health Organization. Abstract