Schizophreniaforum - Print Discussion

Stephan Heckers		Rajiv Tandon

SRF Live Discussion Series: Anticipating the DSM-V
Schizophrenia Research Forum is presenting a series of live discussions focusing on areas of contention within the evolution of the Diagnostic and Statistical Manual (DSM) psychotic disorders area.

The first discussion on 22 July 2009, Is the Risk Syndrome for Psychosis Risky Business?, focused on the proposal to create a diagnostic category for people possibly in the prodrome for schizophrenia.

The discussion on 15 December 2009 was led by Stephan Heckers of Vanderbilt University and Rajiv Tandon of the University of Florida, and addressed the value of the schizoaffective diagnosis. Please read the backgrounder below and the article by Heckers mentioned therein. Then add your comments.

Suggested Reading: Heckers S. Is schizoaffective disorder a useful diagnosis? Curr Psychiatry Rep. 2009 Aug ;11(4):332-7. Abstract

Our apologies; due to copyright issues, we will not be able to provide access to Stephan Heckers's article, "Is schizoaffective disorder a useful diagnosis?"

See Draft of proposed DSM-V modifications.

Click on the images below to launch the slidecasts.

Emil Kraepelin divided the psychoses into non-affective (dementia praecox, later schizophrenia) and affective (manic depression, later bipolar disorder) types. This dichotomy continues to this day in diagnostic manuals (see SRF related Live Discussion). As discussed by Stephan Heckers in the background text below, Jacob Kasanin introduced the diagnosis schizoaffective disorder in 1933 (Kasanin, 1933) to describe patients with both prominent psychotic and affective symptoms. The diagnosis, with minor variations, has been part of the DSM since its first edition in 1952.

“However, the current DSM-IV-TR diagnosis of schizoaffective disorder is not reliable and is of limited clinical utility,” Heckers writes. He traces the evolution of the current schizoaffective disorder diagnosis in the DSM and reviews options for revision (see list of options below, courtesy of S. Heckers). Some modifications are minor, whereas others are more radical. Do we have enough evidence to remove the diagnosis from the DSM? What does a revision of the diagnosis schizoaffective disorder mean for the more fundamental dichotomy of affective and non-affective psychoses? We invite your preliminary commentary on the options presented.

References:
Kasanin J. The acute schizoaffective psychoses. Am J Psychiatry 1933, 90:97–126.

Heckers S. Is schizoaffective disorder a useful diagnosis? Curr Psychiatry Rep. 2009 Aug ;11(4):332-7. Abstract

Carla Canuso, Ortho-McNeil Janssen Scientific Affairs
William Carpenter, University of Maryland
Nick Craddock, Cardiff University, Wales
Angela Epshtein, Schizophrenia Research Forum
Pamela DeRosse, NSLIJHS Zucker Hillside Hospital
Stephan Heckers, Vanderbilt University
Hakon Heimer, Schizophrenia Research Forum
J. Kando
Dara Manoach, Massachusetts General Hospital
Andrey Potapov, Moscow Research Institute of Psychiatry
Eugenia Radulescu, Bucharest
Abraham Rudnick, University of Western Ontario
Maristela Schaufelberger, University of São Paulo
Amresh Srivastava, University of Western Ontario
Rajiv Tandon, Florida Department of Children and Families
Norris Turner, Ortho-McNeil Janssen Scientific Affairs
Neil Woodward, Vanderbilt University

Angela Epshtein
I would like to introduce and thank our chat leaders: Rajiv Tandon, who is currently chief of psychiatry at the Florida Department of Children and Families and professor at the University of Florida, and Stephan Heckers, who is chair of psychiatry and professor of radiology at Vanderbilt University. Both Rajiv and Stephan are involved in the DSM-V work group on psychotic disorders.

Stephan Heckers
Hi; thank you for joining us today. This is a roundtable discussion on the topic of schizoaffective disorder (SAD).

We hope that this session is interesting and informative for all participants. Our discussion, in conjunction with the preliminary comments that have been posted, will be important feedback for the DSM-V Psychotic Disorders Work Group as well as the Mood Disorders Work Group. We will shape the discussion into roughly three 20-minute segments:

1. Given the history of the SAD diagnostic category, has it been a useful construct for clinicians and researchers? In what ways?

2. Will a longitudinal versus an episodic approach to SAD diagnosis help us address the reliability and validity issues of the diagnosis?

3. Finally, we’ll review the options for the diagnosis of schizoaffective disorder.

Pamela DeRosse
Stephan, I believe it serves as a bridge between the affective and psychotic dimensions, so in terms of identifying the molecular substrates of symptoms, I would argue it is very useful.

Nick Craddock
One benefit is that it identifies cases with a specific set of symptoms, which can get lost if subsumed under either schizophrenia or bipolar disorder.

Eugenia Radulescu
It’s still a compromise. It’s difficult to give up SAD for the moment.

Amresh Srivastava
It has been neither useful to clinicians, researchers, nor patients. It would be a bold decision to give up SAD.

Amresh Srivastava
Pamela, it is for the time being that diagnosis does not change treatment. That may be history soon.

Stephan Heckers
If the diagnosis of SAD is not useful, why is it being used so often?

Nick Craddock
Stephan, it describes a lot of patients! Perhaps it is actually the "typical" psychosis.

Stephan Heckers
Nick, SAD describes a lot of patients only if it is vaguely defined. A more stringent diagnosis makes it much less prevalent than schizophrenia.

Nick Craddock
Stephan, yes, quite correct. Current DSM usage of SAD is exceptionally narrow—hence, the poor reliability. However, the "essence" captured (i.e., mood and schizophrenia-like psychosis) is very common.

Pamela DeRosse
Amresh, I think the most relevant issue is the severity and pervasiveness of the mood syndrome.

Amresh Srivastava
Pamela, quantitative measure need not decide diagnosis; it needs to be qualitative difference.

Rajiv Tandon
We considered difficulties with the category: poor validity, low reliability, limited diagnostic stability. But it is widely used. Why?

Amresh Srivastava
It’s used mainly by clinicians in a rush for a shortcut diagnosis.

William Carpenter
For your information, SAD and psychosis not otherwise specified are by far the most common diagnoses in the psychosis chapter of DSM-IV.

Rajiv Tandon
We considered giving up the diagnosis. But there are many patients with an admixture of mood and psychotic symptoms (concurrently and over time) that cannot easily be categorized into schizophrenia or major mood disorders. We hence decided that we needed to retain this diagnosis either dimensionally or categorically, or both.

Stephan Heckers
Few people are left who are in favor of SAD as a diagnostic class, but even fewer want to change the status quo. It seems that we have not tested alternative diagnoses well enough.

Amresh Srivastava
Nick, can the three—major mood disorders, schizophrenia, and SAD—be distinct entities?

Rajiv Tandon
That was the challenge, Amresh. Mood symptoms occur in schizophrenia. Schizophrenic symptoms occur in major mood disorders. Where are the boundaries, and does schizoaffective disorder serve a useful purpose? The boundaries of the condition have varied over time, but it is widely used, suggesting the need to redefine it but retain it in some way.

Nick Craddock
Amresh, the problem with the current schizophrenia concept is that the mood symptoms get "lost" or "forgotten" in the diagnosis. Also, the balance of mood to psychosis at any time determines whether someone gets put in one of three categories (schizophrenia, SAD, bipolar disorder). That seems pretty unhelpful.

Stephan Heckers
If our primary interest is to capture mood symptoms in a person with a primary psychotic disorder, then we can do this with specifiers added to a psychotic disorder. This does not require the diagnosis of SAD.

Rajiv Tandon
In DSM-III, SAD was virtually abolished and has slowly been brought back. The reality is that there are patients not easily categorized in the two other categories.

Hakon Heimer
Ragiv et al., on why SAD is still used, Ray DePaulo told me this past week that at his institution, they train their residents not to use it, the implication being that some training programs are not as rigorous (he alluded to this in his comment). If it is a grab bag for clinicians who are not up to deciding among disorders, is it worth keeping?

Stephan Heckers
Hakon, the effect of training site is significant. It’s not too different from what we saw in the 1970s, when bipolar disorder became more "prevalent" after the introduction of lithium as a major mood stabilizer.

Rajiv Tandon
Let's step back. Question #1: Should we retain some concept of schizoaffective disorder at the ill-defined boundary of schizophrenia and major mood disorder? If the answer is “yes, but not as at present,” then we could get to the next issues of how best to do this.

William Carpenter
To all, should SAD be a diagnosis for overall course of illness, or just an episode designator? If an episode, why would it not be better to have a depression dimension, a manic dimension, and a psychosis dimension?

Amresh Srivastava
Will, that’s a better perspective, throughout the course of illness.

Stephan Heckers
Will, that is a key question. DSM-III-R looked at lifetime illness; DSM-IV switched to episode (with the goal to increase reliability). We are now poised to go back to DSM-III-R.

Amresh Srivastava
Stephan, what you are saying has been historically the position with SAD, and we are revolving back into it.

Nick Craddock
All, dimensions are a good way to go, particularly for research (Craddock and Owen, 2007). However, we do not yet know exactly what to do. It is important not to make major changes for which there is no clear evidence. Retaining SAD makes sense because it is used—and there are genetic studies that support the possibility of some specificity (Craddock et al., 2010).

Pamela DeRosse
Nick, isn't it possible at the biological level that SAD is an admixture of both psychotic and affective illness?

Nick Craddock
Pamela, yes, it is possible. But we don't yet know, so why make major changes on the basis of personal hunches?

Stephan Heckers
Rajiv Tandon, it seems to me that geneticists and those who have done family studies are the primary advocates for keeping SAD in the DSM. There is less enthusiasm from other camps.

Rajiv Tandon
If we agree that we need to retain some kind of boundary condition between schizophrenia and major mood disorder, then we can get to the questions of how: 1) dimensionally, 2) categorically with better definitions, or 3) both.

Stephan Heckers
Rajiv, we do not need categories, but there is no proven alternative.

Rajiv Tandon
Also, if we agree about the need for such a condition, should it be an episode diagnosis (better reliability but poorer validity) or lifelong diagnosis (better validity and utility)?

Stephan Heckers
Rajiv, lifetime diagnosis makes more sense, but we need better patient histories. Our clinicians are not encouraged to go the extra mile for this.

Amresh Srivastava
Stephan, you are right, and that’s the problem. Is there a common neurobiology between schizophrenia and mood disorder at all?

Rajiv Tandon
I agree, Stephan. The category is not working well, but the alternatives are not very good. As we explore what to do, can we agree that schizoaffective disorder is useful in some fashion?

Neil Woodward
Stephan and Nick, to what extent can premorbid personality, social functioning, and childhood diagnosis be used to inform diagnosis made in adulthood?

Nick Craddock
If diagnosis is about directing treatment and prognosis, we must be making lifetime diagnoses.

William Carpenter
Nick, if longitudinal (and I agree), we need to recognize that most often diagnosis will be made with little information about co-occurrence over the lifetime. Imagine the primary care doctor or emergency room when a psychotic patient with mood disturbance is seen, and you try to piece together the lifetime course of the two dimensions.

Nick Craddock
Will, it is an aspect of psychiatry that it may take a period of observation (perhaps long) to be sure of diagnosis. There is nothing new here.

Pamela DeRosse
Stephan, do you think the differential between schizophrenia and SAD alters the way a clinician treats the patient?

Stephan Heckers
Pamela, yes, definitely. It has a huge effect on the patient. Most prefer SAD over schizophrenia.

Rajiv Tandon
I think the diagnosis of schizophrenia (with or without mood symptoms) versus mood disorder (with or without psychotic symptoms) and schizoaffective disorders do come with different treatment and prognostic implications (not easily or logically separated or distinguished, but still different).

Carla Canuso
I do think differential diagnosis does result in different treatment. While patients with schizophrenia and SAD are equally likely to receive antipsychotic medication and antidepressants, patients with SAD are much more likely to receive mood stabilizers or mood stabilizers and antidepressants in addition to antipsychotics.

Abraham Rudnick
What do people think of viewing schizophrenia and major mood disorders as possibly comorbid?

Stephan Heckers
Abraham, I see little value in comorbidity of two diagnostic classes.

Nick Craddock
Abraham, I dislike the idea of "comorbidity" because it reifies the components.

William Carpenter
Nick, we have to write the DSM-V for all the usual clinical situations, not for the research application.

Pamela DeRosse
Will, but shouldn't the DSM-V also function to specify the diagnostic groups that we are studying?

Nick Craddock
Will, I agree DSM is not primarily a research instrument, but we have to remember that it has a pretty big impact on the way people think.

Amresh Srivastava
Nick, I agree; a well-defined diagnosis needs to have a definite origin point.

Rajiv Tandon
Can we move to a discussion of whether this should be an episode or lifelong construct?

Stephan Heckers
To all, a major question for the DSM-V work group on psychosis is, Should SAD be diagnosed based on the episode (as it is currently in DSM-IV-TR)? It seems there is little support to keep it linked to the episode, correct?

Rajiv Tandon
I agree, Stephan. It should be a lifelong (not episode) construct; that would be more meaningful from both clinical and research perspectives.

Pamela DeRosse
Stephan, I think there are too few data on the question, but longitudinally seems to make more sense.

Nick Craddock
Stephan, I think linking to episode is not helpful (of course, an episode might be described as "mood/psychosis" or "schizoaffective," but the concept should be longitudinally based).

Stephan Heckers
To all, why do we hold on to a diagnosis that has low (<0.4) reliability ratings? What would we do if MDD or schizophrenia had similar reliability scores? The poor reliability of SAD is primarily due to our inability to capture mood symptoms accurately over time.

Pamela DeRosse
Stephan, I think it also has to do with the poor reliability of patient reporting.

William Carpenter
Pamela, our work group is assigned all of the schizophrenia and related disorders group of psychoses as the starting point and will be very close to that at the end. Problems are within each class. Here it is: Should we be focusing on the current episode, as in DSM-IV, or over the course of illness, as in DSM-III? I think that only the overall course is conceptually coherent, but that approach has the big problem of needing detailed history. If only episode, I would favor capturing with dimensions.

Rajiv Tandon
I completely agree with Will's view: If only episode diagnosis, we do not need a category but could address with dimensional assessments. But longitudinally, there is value to having a schizoaffective category still instead of artificially assigning some patients to schizophrenia and others to mood disorders.

Amresh Srivastava
Is there any evidence that SAD retains its character in longitudinal course?

Stephan Heckers
Amresh, there is strong evidence that it does not. Please look at my slide set; it lists some recent studies on this topic.

Amresh Srivastava
Stephan, that’s what I am saying; where is the compulsion to retain this as diagnosis?

Nick Craddock
Stephan, the low reliability is because of the narrow definition. If the definition were more inclusive, it would have greater reliability (Craddock et al., 2009). Narrowing the concept of either schizophrenia or bipolar disorder would make them have low reliability!!

Stephan Heckers
Nick, very important. You are suggesting to make SAD broader, correct?

Amresh Srivastava
Is this diagnosis different in different parts of the world?

Carla Canuso
ICD-10 (International Classification of Diseases, 10th Revision) is different from DSM, and more inclusive.

Stephan Heckers
Amresh, yes, ICD is broader. It does not require psychotic symptoms without mood symptoms for two weeks, as in DSM.

Amresh Srivastava
Will, we need to look at the entire dataset to decide this question. In reactive psychosis, it is more common, perhaps.

Stephan Heckers
Nick, do you use ICD or DSM criteria for your genetic studies?

Dara Manoach
In research practice, in essence, schizoaffective is divided up, with schizoaffective depression being included in the schizophrenia group, and schizoaffective bipolar not included.

Nick Craddock
Stephan, currently SAD is pretty much written as a diagnosis of last resort after not being able to opt for bipolar disorder or schizophrenia because of the very fine "balance." That is narrow. If the wording allowed greater overlap—but clearly specified—that would be more reliable, because people would not easily change category according to small changes in available information, longitudinal history, or the opinion of the doctors.

Rajiv Tandon
Nick and Stephan, you make an interesting suggestion—broaden the concept of schizoaffective disorder—but that would be at the expense of schizophrenia with varying amounts of mood symptoms and major mood disorders with psychotic symptoms. Would that really make things better?

Pamela DeRosse
Rajiv, I would argue not. There are different types of affective episodes that show up superimposed on schizophrenia.

Rajiv Tandon
The division of schizoaffective disorder into bipolar type (with major mood disorders) and depressive type (with schizophrenia) is based on a 1987 Andreasen study (Andreasen et al., 1987). It has not been replicated.

Stephan Heckers
Rajiv, a broader concept of SAD might increase reliability. That is Nick's point.

Nick Craddock
Rajiv, if a diagnosis of SAD flags up both mood and schizophrenia-like problems, that seems to me to serve a clinically useful function. It will also make it easier to research this very common overlap.

William Carpenter
A big problem with DSM-IV is that the mood disturbance is required to meet full criteria for a mood disorder. It is absurd to think that clinicians can adhere to this requirement over the life course of illness. I suppose it is based on the impression that mood is disturbed much of the time rather than considering whether someone meets full criteria in the past three episodes.

Hakon Heimer
Will, Rajiv, Stephan, if SAD is in the purview of the psychosis group, and mood disorder specialists are not part of this group, does that weaken your efforts to come to grips with the diagnosis? (Was that a provocative question or just naive?)

Amresh Srivastava
Perhaps it’s not a question for mood disorder researchers.

Angela Epshtein
Hakon, we have a comment from Jan Fawcett, who is the mood disorder work group leader.

William Carpenter
Hakon, Ray DePaulo is liaison with us, Bill Coryell has been in conference with us, and it is discussed more generally with the mood disorders group.

Stephan Heckers
Hakon, I would not worry too much about buy-in from different groups. Nobody has the perfect alternative, but the mood disorder group is more content with the status quo.

Rajiv Tandon
Hakon, a very good question. We did try working with the mood disorders group, and we had different sets of concerns and boundary issues. Initially, we were hoping to be able to eliminate SAD as a category and address it only dimensionally. But the mood disorders group was opposed to the idea, in part because it would have necessitated the same set of dimensional assessments for both major mood disorders and the psychotic disorders. Hence, we adopted our present set of recommendations.

Pamela DeRosse
Will, how do you think the loss of the depressive disorder and bipolar disorder not otherwise specified diagnoses being suggested by the affective disorders work group will affect the diagnosis of mood syndromes superimposed on schizophrenia?

William Carpenter
Pamela, I don't know. In general, DSM-V will try to decrease the use of “not otherwise specified,” and the assumption is that, in the mood disorders, it will force more specific mood disorder diagnosis. I doubt that it will push towards psychoses.

Pamela DeRosse
Will, perhaps it will force more people into the SAD category?

Stephan Heckers
To all, let's review the options for the diagnosis of schizoaffective disorder in DSM-V. We posted four options for DSM-V. It seems likely that we will end up with #2 (see SRF website).

Rajiv Tandon
Stephan, there are four criteria. The first defines severity and need for both psychotic (schizophrenia criterion A) and major mood symptoms; we are not proposing a change there. Criterion 2 tries to distinguish SAD from psychotic mood disorders; we propose a change to clarify here. Criterion 3 to help separate SAD from schizophrenia with mood symptoms; we propose a change to clarify here. Criterion 4 is the substance abuse and organicity exclusion—no changes proposed here.

Stephan Heckers
To all, can we remove Criterion B and get DSM and ICD closer together?

Carla Canuso
I don't think Criterion B is all that useful clinically, or in distinguishing SAD from a major mood disorder.

William Carpenter
Stephan, I like dropping the mood-comes-later-and-leaves-earlier criterion, since I think that is the reverse of natural history. That would be a good place to harmonize with ICD.

Rajiv Tandon
Will and Stephan, the problem with eliminating Criterion B is that the mood disorder group objects, as it wants a separation between SAD and psychotic mood disorder. It is somewhat artificial (based on the Coryell studies in the 1980s), but it is the best we have.

Pamela DeRosse
Stephan, I think that would make it harder for us to distinguish between a psychotic affective disorder (i.e., bipolar disorder with psychotic features) and SAD.

Stephan Heckers
Pamela, you are correct. But why are U.S. psychiatrists so concerned about it when the Europeans are not? Criterion B makes little sense (other than keeping psychotic mood disorder further away from SAD).

Rajiv Tandon
Absolutely Stephan, but as you know, the mood disorders group is adamant about this (with some good reason); we over-diagnosed schizophrenia in the U.S. prior to DSM-III.

Pamela DeRosse
Stephan, I don't agree. I think it functions to bridge the gap between the psychotic and affective disorders and is more consistent with the dimensional model that will likely emerge in later editions of the DSM.

Stephan Heckers
Pamela, that’s another way of looking at the same boundary.

William Carpenter
All, our big problem with dimensions is that we only think DSM-V can manage them as concurrent measures. Clinicians don’t have enough time or informants for reliable life course dimensions. Therefore, dimensions can only solve the SAD problem if SAD is a current-episode diagnosis.

Rajiv Tandon
Maybe we need to change Criterion B, but we need some better separation between SAD and psychotic mood disorders. Perhaps we could look at our proposal to see if this might help.

Carla Canuso
Is the proposed language, “delusions and hallucinations for two or more weeks in the absence of psychopathology” meeting criteria for a major mood episode?

Rajiv Tandon
Carla, yes, that is the proposal. I think it helps a clearer separation from psychotic disorders.

Nick Craddock
Stephan, it varies. DSM is very influential because of publishing (many U.S. journals...). In the Wellcome Trust Case Control Consortium genomewide association study (Wellcome Trust Case Control Consortium, 2007), I used RDC (Research Diagnostic Criteria)!! (Because it captures better the "mixed" cases.)

Stephan Heckers
Nick, does it make sense to compare the genetics of SAD based on either DSM or ICD or RDC?

Nick Craddock
Stephan, we have a paper in Molecular Psychiatry (Craddock et al., 2010) showing a relatively specific association at the family of GABAA receptor genes and RDC SAD, bipolar disorder (and this is not present in bipolar disorder or schizophrenia cases). RDC schizoaffective disorder, bipolar type (SABP) is a broad definition!

Hakon Heimer
Nick, would there be any data or discussion from the psychiatric genomewide association study group (especially the cross-disorders group) that are germane to this discussion?

Rajiv Tandon
Could we consider our proposed change to Criterion C (that tries to better differentiate SAD from schizophrenia with some mood symptoms)?

Stephan Heckers
Rajiv, yes, please. Does an explicit threshold (say, 30 percent) make the task easier and more reliable?

Rajiv Tandon
Basically, our proposed changes to Criterion C have to do with making this a lifetime (not episode) diagnosis and suggesting 30 percent prevalence of mood symptoms as the separator of SAD from schizophrenia with some mood symptoms.

Pamela DeRosse
Stephan, 30 percent is certainly better than the current exclusion criteria of "relatively brief."

Carla Canuso
I think that objectifying "substantial proportion" helps, but doesn't get around the subjectivity that goes into making that assessment.

Rajiv Tandon
It seems that Stephan, Carla, and Pam are all in agreement with Criterion C as proposed.

Stephan Heckers
Nick, we need to hear from you. Does a cut-off of 30 percent help?

William Carpenter
All, does anyone think meeting full criteria for mood disorder is either feasible or desirable? Is the concept of serious mood disturbance what clinicians must use, especially regarding the past?

Rajiv Tandon
Will, I think the requirement for a major mood symptom versus prominent mood symptom is helpful (more objective). Also, it helps better delineate SAD from schizophrenia with some mood symptoms.

Nick Craddock
All, I suspect that will still be somewhat difficult to use. Why not be radical and, for SABP, go with a clear-cut episode of mania?!

Nick Craddock
Stephan, I think one episode could be appropriate for SABP. Mania is a pretty recognizable syndrome, and a person has an ongoing susceptibility to such disturbances. Of course, it might be more difficult thinking of doing this for depression.

Pamela DeRosse
Nick, I agree. Will, I think the full criteria for major mood episodes is feasible.

William Carpenter
Pamela, imagine a patient with a 10-year history and the doctor has 15 minutes to do everything. How does he/she figure out meeting full criteria over time?

Pamela DeRosse
Will, clinically, that would be very difficult, so I see your point. However, the same holds true for the 30 percent criterion.

William Carpenter
Pamela, 30 percent is being offered as a magic bullet, and we don't dare look too closely.

Angela Epshtein
Everyone, we have just a few minutes left. Many of us will need to leave, though the chatroom will remain open. Does anyone have any last thoughts or comments to make before some of us leave?

Hakon Heimer
Stephan, all, what is needed in the next five to 10 years before DSM-VI? Dimensions? Genes?

Nick Craddock
Dimensions, but they need to be the right dimensions!! Genes, imaging, biology, and psychology will help.

Stephan Heckers
I agree with Nick. Dear all, thank you for joining us today. No revelations, but a good set of questions.

Rajiv Tandon
I think this has been a very valuable conversation. I hope we continue the process. We will incorporate suggestions proposed here into further revisions.

William Carpenter
Hakon and Angela, thanks for doing this, and Rajiv and Stephan.

Nick Craddock
Stephan, very interesting and worthwhile. Many thanks! Best wishes, Nick.

Angela Epshtein
Thank you, Rajiv and Stephan, for leading. Thanks to all for their comments.



	Comments on Online Discussion



	Comment by: Rajiv Tandon
	Submitted 4 December 2009	Posted 4 December 2009
	Schizoaffective Disorder Considerations for DSM-V Although dementia praecox or schizophrenia has been considered a unique disease entity for the past century, its definitions and boundaries have varied over this period. The relationship between schizophrenia and the major mood disorders (bipolar disorder and major depressive disorder) has been a topic of much ambiguity with ill-defined and fluctuating boundaries. Schizoaffective disorder has been at the interface of the boundary between schizophrenia and the major mood disorders, and opinions about its nature vary considerably. The following are four characterizations of schizoaffective disorder: 1. A unique condition clearly demarcated from both schizophrenia and major mood disorders 2. A subtype of schizophrenia 3. A subtype of major mood disorders with psychosis 4. Not a unique entity but an admixture of schizophrenia and major mood disorders that are difficult to differentiate Despite its uncertain nature, schizoaffective disorder is widely diagnosed and is considered by clinicians to be a useful construct. On the other hand, its validity is questionable, with most data suggesting that patients diagnosed with this condition represent a mixture of those with schizophrenia and those with major mood disorder. Additionally, it has marginal reliability and low diagnostic stability. Given its perceived clinical utility in the context of poor reliability and validity, is it possible that eliminating it as a categorical entity but providing a dimensional approach toward measuring both psychotic and mood symptoms might provide for a more valid, reliable, and clinically useful construct? Alternatively, one could ask the question of whether or not schizoaffective disorder should be retained as a distinct entity but be better defined so that it can be more clearly discriminated from both schizophrenia and major mood disorders. These issues were considered by the DSM-V workgroup on schizophrenia and related psychotic disorders, and the second approach was selected; the proposed revisions are detailed in the Draft of proposed DSM-V modifications. References: Heckers S. Is schizoaffective disorder a useful diagnosis? Current Psychiatry Reports, 11: 332-337, 2009. Abstract Jager M, Becker T, Weinmann S, et al. Treatment of schizoaffective disorder-a challenge to evidence-based psychiatry. Acta Psychiatrica Scandinavica, 2009. Abstract Kaymaz N, van Os J. Murray et al. (2004) revisited: is bipolar disorder identical to schizophrenia without cognitive impairment? Acta Psychiatrica Scandinavica, 120:249-252, 2009. Abstract Tandon R, Maj M. Nosological status and definition of schizophrenia: some considerations for DSM-V and ICD-11. Asian Journal of Psychiatry, 1:22-27, 2008. Tandon R, Nasrallah HA, Keshavan MS. Schizophrenia ”Just the Facts” IV. Clinical features and conceptualization. Schizophrenia Research, 110: 1-23, 2009. Abstract View all comments by Rajiv Tandon

	Comment by: Amresh Shrivastava
	Submitted 5 December 2009	Posted 5 December 2009
	The question, which has come up for discussion, is a significant one. One of my professors taught me that the most important thing in creating effective psychiatric treatment is the diagnosis. While this point may be obvious, it becomes problematic when a diagnostic dilemma is brought to the foreground. The fact that there is a DSM-V position paper that questions the existence of one of the subgroups of schizophrenia is significant (Srivastava, 2009). Before I comment on whether or not the diagnosis of schizoaffective disorder ought to be contained under the rubric of schizophrenias, mood disorders, or as an independent psychiatric disorder, it is useful to recap what constitutes a diagnosis. Clinical psychiatry continues to function through an exploratory model of descriptive symptoms, signs, and syndromes. A definitive diagnosis is exclusively clinical in nature. Considerable differences of opinion and conflicting viewpoints exist regarding whether a particular symptom in a given disorder is part of the disease or exists outside the realm of that disease, thus qualifying as a comorbid symptom. For example, many clinicians would refrain from a diagnosis of schizophrenia with OCD or GAD, and would be happy to consider these phenomena as part of the spectrum of the illness. Thus, our first problem is to be able to decide criteria for diagnosis. I don’t think the DSM-V is addressing this fundamental difficulty, and it may be the case that we require more evidence and phenomenological knowledge before we can create a definition and criteria for psychiatric diagnosis. The question regarding whether or not schizoaffective disorder is a useful diagnosis entails an examination of whether or not SAD is indeed a sub-diagnosis of schizophrenia. No one is surprised with this proposal as history repeats itself. Positioning of the schizoaffective disorder has been shifting from one end of the diagnostic spectrum to the other (Tsuang, 1979; Singh and Sachdeva, 1982). Jacob Kasanin (1933) originally proposed this group for atypical syndromes in terms of schizophrenia and mood disorder. In 1961 Leonard called it the “third psychosis,” retaining a similar description as Kasanin’s definition. The International Classification System of World Health Organization, ICD 9, subsumed it under schizophrenia. However, in the Research Diagnostic Criteria (RDC), SAD was kept as a separate entity. In the 1986 draft of the ICD 10, SAD was assigned to the affective disorder category, but in the final version of ICD 10, it was placed alongside the category of schizophrenic disorder. The idea of SAD as a separate entity continued within the DSM-III R; however, in DSM-IV, it again became a subgroup within schizophrenic disorders. The story continues. The silently implied issue is that we are not talking about mood symptoms across schizophrenia. We are discussing whether there is a symptom cluster that is based on the current understanding of psychiatric illnesses that we can designate as schizoaffective disorder, and whether the strength of the clinical and research evidence suggests retaining SAD as a variety of schizophrenia. This topic has been reviewed from time to time in the literature, only to conclude that the dilemma exists (Cheniaux et al., 2008; Lake and Hurwitz, 2007; Lake and Hurwitz, 2006). We conducted a study in which we followed individuals diagnosed with schizoaffective disorder for two years. Our goal was to test the stability of the diagnosis, and we found that only 11 percent of patients retained the SAD diagnosis. The majority (72 percent) changed over to schizophrenia, and 17 percent changed to unipolar and bipolar mood disorder (Shrivastava and Rao, 1997). Several authors have written that SAD is an inconsistent syndrome as a diagnosis. If we examine the available evidence of parameters of clinical course and outcome, neurochemistry, neuroimaging, genetics, and treatment response, which are normally the main parameters to consider a symptom-cluster or syndrome as a diagnosis, we find that schizoaffective disorder does not maintain an exclusive diagnostic entity on any of these parameters. There is some evidence of a neurochemistry parameter that is equivocal; however, neurochemical parameters are not considered to be very strong parameters. There is no valid argument justifying the removal of SAD from the DSM-V. Its presence as a diagnostic category provides a scope of ambiguity and non-specificity in clinical practice. Research in the field of course and phenomenology of brief reactive psychosis, early psychosis, and schizophrenia spectrum disorder have provided enough evidence that mood symptoms exist and arise from common neurobiological basis, and the pathogenesis of schizoaffective disorder can be articulated along these lines. However, this again does not solve the problem. Leaders within the research community have the responsibility to provide a logical placement of schizoaffective disorder. The philosophy driving the DSM-V is based on a movement toward a dimensional approach and away from a categorical system, which is highly comforting for those of us engaged in clinical work. Hopefully, the classification as domains of schizophrenia may be retained. In my opinion, schizoaffective disorder would best be suited as an affective domain of schizophrenia. References: Srivastava A. "Should Schizoaffective Disorder Be Dropped from DSM V" (2009). Psychiatry Presentations. Paper 25. Tsuang MT. 'Schizoaffective disorder': dead or alive? Arch Gen Psychiatry. 1979 Jun 1;36(6):633-4. Abstract Singh, G and Sachdeva JS. (1982) Schizoaffective Psychosis, Are They Schizophrenic? Indian Journal of Psychiatry, 24, 28-42. Kasanin, J. (1933) The acute schzioaffecytiv psychosis. American Journal of Psyhciatry 90,97-126. Cheniaux E, Landeira-Fernandez J, Lessa Telles L, Lessa JL, Dias A, Duncan T, Versiani M. Does schizoaffective disorder really exist? A systematic review of the studies that compared schizoaffective disorder with schizophrenia or mood disorders. J Affect Disord. 2008 Mar 1;106(3):209-17. Abstract Lake CR, Hurwitz N. Schizoaffective disorder merges schizophrenia and bipolar disorders as one disease--there is no schizoaffective disorder. Curr Opin Psychiatry. 2007 Jul 1;20(4):365-79. Abstract Lake CR, Hurwitz N. Schizoaffective disorders are psychotic mood disorders; there are no schizoaffective disorders. Psychiatry Res. 2006 Aug 30;143(2-3):255-87. Abstract Shrivastava A, and Rao, S. Diagnostic Validity of Schizoaffective Disorders. Indian Journal of Psychiatry. 1997. View all comments by Amresh Shrivastava

	Comment by: Eugenia Radulescu
	Submitted 9 December 2009	Posted 9 December 2009
	The proposed discussion is fundamental as the clinical diagnosis imposes the therapeutic management and the orientation of the scientific research in the psychoses field. When confronted with a possible case of schizoaffective disorder (SAD), the clinician has to make important decisions, often with profound consequences. The background paper by Heckers (Heckers, 2009) highlights the difficulty of the longitudinal diagnosis due to the lack of reliable information and the doubtful accuracy of self-report, especially from a patient with psychosis. Consequently, the diagnosis is merely cross-sectional, based on clinical observation and interview, but necessary for the immediate therapeutic endeavor. Moreover, as it is well known, the context of the therapeutic decision does not simplify treatment choices, since the available therapies mostly comprise symptomatic medication and are not etiologically based. Practically, the clinician who needs to provide care for a patient with SAD has to answer questions like, Are the depressive symptoms substantial enough to add an antidepressant to the antipsychotics?” or Can the manic symptoms be controlled by antipsychotics only, or, Is a mood stabilizer (i.e., lithium, valproate) also necessary? Obviously, the decision is not trivial since the spectrum of adverse effects is exponentially complicated by the various pharmacological combinations. In clinical practice, there are other aspects of the SAD diagnosis that require specific attention (e.g., substance abuse, increased suicidal risk). It is clear that clinicians need a way to describe the cohabitation between psychotic and affective/mood symptoms. Tandon concludes his slidecast with the (Tandon, 2009) proposal that the DSM-V keep the SAD diagnosis with the amendment of the operational criteria and consequently the improving of diagnosis reliability. While this is a judicious approach, the issue of how one conceptualizes psychosis remains, along with the “compromise” diagnoses of SAD. Obviously, the leaders in the field who work on DSM-V still need to decide if the definition of psychosis should be based on the Kraepelinian dichotomy, or should be viewed on a “continuum,” or even better, as a mixture of symptomatic dimensions. From this angle, a redefinition of psychosis on a dimensional basis with diagnostic criteria adjusted for practical purposes might be helpful for clinicians and researchers. Some consistent symptomatic dimensions such as positive, negative, and disorganization (Bilder et al., 1985; Liddle et al., 1989) or negative, positive, cognitive, excitation, depression/anxiety (Lindenmayer et al., 1994; Von Knorring and Lindstrom, 1995; Radulescu, 2003) have emerged from factor analytical studies, with some variation due to the assessment instruments. In this respect, an affective dimension is distinguishable within the psychotic spectrum. However, it might be useful to remember that the affective dimension is complex and encompasses various phenomenological aspects that can align the dimension closer to the schizophrenic spectrum or to the mood disturbance spectrum. The affective domain may suffer qualitative changes manifested through affective incongruence/inadequacy, or affective ambivalence more characteristic of schizophrenia. Similarly, there could be quantitative changes in the direction of hypothymia/affective impoverishment (the affective flattening characteristic of the deficit syndrome described by Carpenter, 1994) or hyperthymia, with the exacerbation of affective tone toward the depressive or manic end. Similar fluctuations are described within perceptual or thought domains, which also contributes to the complexity of the clinical picture across these dimensions. Further complicating the diagnostic process, the weight of symptomatic dimensions may vary longitudinally for the same individual, so it might be useful to view psychosis in a broader manner—as a dynamic ensemble of intermixed symptomatic dimensions: perceptual, thought, affective, etc. While this would add an element of flexibility that could be useful to clinicians, it would constrain clinicians to a careful evaluation of each symptomatic dimension with clear implications for the therapeutic decision. From this point of view, the cross-sectional SAD diagnosis might be equivalent to psychotic disorder with co-dominance of perceptual (hallucinations), thought (delusions), affective (depression) disturbances. The dimensional approach might have important implications for research, too. It’s noted in Heckers’s paper (Heckers, 2009) that within categorical diagnosis, scientists usually group SAD with schizophrenia. This implies that in terms of biology, SAD is actually viewed as a type of schizophrenia, which raises concern. It is well known that within the biological etiology of psychosis, scientists have a justified preference to focus on endophenotypes, not on clinical symptoms (Meyer-Lindenberg and Weinberger 2006). However, here intervenes a paradoxical situation: In the procrustean manner imposed by the DSM criteria, clinicians use diagnostic labels that subsequently are used in research for grouping psychiatric patients and comparing these groups with normal individuals or unaffected relatives with respect to endophenotypes. According to criteria proposed by Gottesman and Gould (2003), the endophenotype (aka intermediate phenotype or vulnerability marker) is associated with the illness, is heritable, state-independent, and within families segregates with the illness. It is also suggested that an endophenotype, less complex than clinical phenotype, is closer to the biological determinants and correlated with the illness, suggesting a linear relationship. However, is this the case, especially for SAD, which has a poor validity? Is it, therefore, justifiable to group it with schizophrenia? As it was specified in Heckers’s (2009) paper and in Shrivastava’s comment, longitudinal studies show that a substantial proportion of SAD-diagnosed individuals will evolve similarly to individuals diagnosed with schizophrenia. However, some will follow the path of individuals diagnosed with a form of mood disorder. While this is the case, could one assume that the relationship between any schizophrenia endophenotype and SAD is the same as the relationship between that endophenotype and the classical schizophrenia? This question is really important in research nowadays, when studies of great amplitude (especially in psychiatric genetics) are ongoing and the homogeneity of diagnosis is a prerequisite of reliable results. Maybe a dimensional approach could be one way to better group the affected subjects. Working with dimensions might help us discern which brain circuitry is more severely affected, and consequently, which endophenotypes are suitable for group comparisons. One might even attempt to assess which symptomatic dimension could have an endophenotype value. In this respect the cognitive dimension has already been evaluated as an endophenotype specific to schizophrenia (Goldberg et al., 1995). In conclusion, as a clinician, I hope that the DSM-V will offer a more flexible diagnostic guide, with room for a dimensional approach to evaluating psychosis. In the particular case of the SAD diagnosis, even if it will remain in use, one could specify that it be used in a provisional manner until longitudinal validation is well documented. As a researcher, I would operate more comfortably with symptomatic dimensions that might be closer to endophenotypes and the subjacent biology. References: Heckers S. Is schizoaffective disorder a useful diagnosis? Curr Psychiatry Rep. 2009 Aug ;11(4):332-7. Abstract Bilder, RM, Mukherjee, S, Rieder, RO, Pandurangi, AK: Symptomatic and neuropsychological components of defect states. Schizophrenia Bull, 1985, 11(3): 409-419. Abstract Liddle, PF, Barnes, TRE, Morris, D: Three Syndromes in Chronic Schizophrenia. British Journal of Psychiatry, 1989, 155(suppl. 7): 119-122. Abstract Lindenmayer, JP, Bernstein-Hyman, R, Grochowski, S: Five Factor Model of Schizophrenia: Initial Validation. J Nerv Ment Dis 1994, 182: 631-638. Abstract Von Knorring, I, Lindstrom, E: Principal Components and further possibilities with PANSS, Acta Psychiatr Scand, 1995, 91(suppl. 388): 5-10. Abstract Radulescu, E: PhD Thesis 2003. Carpenter Jr, WT: The Deficit Syndrome. Am J Psychiatry, 1994, March, 151(3): 327-9. Abstract Meyer-Lindenberg, A and Weinberger, DR: Intermediate phenotypes and genetic mechanisms of psychiatric disorders. Nat Rev Neurosci, 2006, Oct 7(10): 818-27. Abstract Goldberg, TE, Torrey, EF, Gold, JM, Bigelow, LB, Ragland, RD, Taylor, E, Weinberger, DR: Genetic risk of neuropsychological impairment of schizophrenia: a study of monozygotic twins discordant and concordant for schizophrenia. Schizophrenia Res 1995, Sep, 17(1): 77-84. Abstract Gottesman, II, Gould, TD: The Endophenotype Concept in Psychiatry: Etymology and Strategic Intentions. Am J Psychiatry 2003, 160: 636-645. Abstract View all comments by Eugenia Radulescu

	Comment by: Pamela DeRosse
	Submitted 14 December 2009	Posted 14 December 2009
	Contrary to most other areas of investigative medicine, the term “diagnosis” in psychiatry does not define the etiology of an illness but rather specifies a set of observable behavioral phenomena that arise as a result of some presumably common, yet unidentified etiological factor. Thus, the utility of a psychiatric diagnosis depends upon how well it lends itself to the discovery of these etiological factors. From a research perspective, then, it seems that the primary question that needs to be addressed is whether or not a change in the diagnostic criteria for schizoaffective disorder, by either including it as a subtype of schizophrenia or the affective disorders or by providing clearer diagnostic criteria, will enhance our ability to identify the pathophysiological mechanisms responsible for the symptoms associated with the diagnosis. Presently, there is a paucity of data on the molecular mechanisms that might differentiate schizoaffective disorder from other psychotic and affective disorders. In molecular genetic studies, schizoaffective cases may be viewed as problematic and excluded from genetic analyses, or, more commonly, are combined with either a predominantly psychotic group or a predominantly affective group. The few studies that have treated schizoaffective disorder as a separate diagnostic group, however, have produced some intriguing results. To date, a single linkage study of schizoaffective disorder has been reported with findings indicating a significant linkage peak near the gene encoding disrupted in schizophrenia 1 (DISC1) at chromosome 1q42 (Hashmere et al., 2005). Data from genetic association studies has provided further support for the relation between DISC1 and schizoaffective disorder. For example, Hodgkinson and colleagues (2004) identified several single nucleotide polymorphisms (SNPs) in DISC1 that were associated with schizoaffective disorder but not with schizophrenia or bipolar disorder. Moreover, other variants were associated with risk for schizophrenia and bipolar disorder but not schizoaffective disorder, suggesting that genetic heterogeneity at this locus may function to alter the phenotypic expression of psychiatric illness. To our knowledge, only a single study has compared the frequencies of a single risk locus across diagnostic categories ranging from purely affective to purely psychotic disorders. Lencz et al. (2009) compared the frequencies of a BDNF risk haplotype across diagnostic groups including schizophrenia, schizoaffective disorder, affective disorders, and healthy individuals, and found that the frequency of the risk haplotype in the schizoaffective disorder group was similar to other affective disorders, and dissimilar from schizophrenia and healthy controls. Taken together, data derived from these molecular genetic studies, although exceedingly limited, suggest that at the molecular genetic level there is both overlap and independence amongst these diagnostic groups. At the clinical level, the topic “Do we need schizoaffective disorder?” is an empirical question which we are currently not well equipped to answer. Specifically, while the diagnosis of schizoaffective disorder requires a longitudinal clinical assessment, most of the data seeking to validate the diagnosis rely on cross-sectional rather than longitudinal data. To date, the limited data that are available to address the longitudinal course of schizoaffective disorder indicate that the clinical outcomes for these patients are substantially worse than the outcomes observed in patients with psychotic affective disorder and substantially better than the outcomes observed in schizophrenia (Harrow et al., 2000; Jäger et al., 2004). Despite our growing understanding of the factors mediating clinical outcomes, however, there are few data seeking to elucidate differences in these factors amongst these diagnostic groups. Our group has recently completed an analysis of a large cohort of patients (N = 993) in which we tested longitudinal hypotheses about the cognitive and clinical differences between the aforementioned patient groups. Among the most striking findings in these analyses was that schizoaffective disorder patients appear to undergo a more severe cognitive decline than either patients with schizophrenia or psychotic bipolar disorder. These findings were somewhat surprising, given previous reports of the relation between impaired cognitive function and worse clinical outcomes (Harvey et al., 2009). However, the most significant symptom variable that separated these diagnostic groups, which is also strongly correlated with clinical outcome (Leung et al., 2008), was the negative symptom cluster. Specifically, the schizoaffective disorder group showed levels of negative symptoms intermediate to those observed in the schizophrenia and psychotic bipolar disorder group. Moreover, we have found that many of the demographic characteristics of patients with schizoaffective disorder are more consistent with those observed in the psychotic bipolar disorder, while many of the clinical characteristics observed are more consistent with the schizophrenia diagnostic group. Thus, rather than providing direct support for the independence of schizoaffective disorder from either schizophrenia or psychotic affective disorders, these data directly challenge the assignment of the schizoaffective diagnosis to a mere subtype of either schizophrenia or the affective psychoses. The convergence of the molecular genetic and clinical data currently available seems to provide little evidence for either abolishing the schizoaffective disorder diagnosis or subsuming it under the schizophrenia or psychotic affective disorder clusters. Therefore, we would argue in favor of revising the diagnostic criteria for schizoaffective disorder. This revision would theoretically provide a better phenotype for the type of longitudinal studies that will be required to validate or remove the diagnosis from future editions of the DSM. References: Hamshere ML, Bennett P, Williams N, Segurado R, Cardno A, Norton N, Lambert D, Williams H, Kirov G, Corvin A, Holmans P, Jones L, Jones I, Gill M, O'Donovan MC, Owen MJ, Craddock N (2005). Genomewide linkage scan in schizoaffective disorder: significant evidence for linkage at 1q42 close to DISC1, and suggestive evidence at 22q11 and 19p13. Arch Gen Psychiatry; 62(10):1081-8. Abstract Harrow M, Grossman LS, Herbener ES, Davies EW (2000). Ten-year outcome: patients with schizoaffective disorders, schizophrenia, affective disorders and mood-incongruent psychotic symptoms. Br J Psychiatry; 177:421-6. Abstract Harvey PD, Keefe RS, Patterson TL, Heaton RK, Bowie CR (2009). Abbreviated neuropsychological assessment in schizophrenia: prediction of different aspects of outcome. J Clin Exp Neuropsychol; 31(4):462-71. Abstract Hodgkinson CA, Goldman D, Jaeger J, Persaud S, Kane JM, Lipsky RH, Malhotra AK (2005). Disrupted in schizophrenia 1 (DISC1): association with schizophrenia, schizoaffective disorder, and bipolar disorder. Am J Hum Genet; 75(5):862-72. Abstract Jäger M, Bottlender R, Strauss A, Möller HJ (2004). Fifteen-year follow-up of ICD-10 schizoaffective disorders compared with schizophrenia and affective disorders. Acta Psychiatr Scand; 109(1):30-7. Abstract Lencz T, Lipsky RH, DeRosse P, Burdick KE, Kane JM, Malhotra AK (2009). Molecular differentiation of schizoaffective disorder from schizophrenia using BDNF haplotypes. Br J Psychiatry; 194(4):313-8. Abstract Leung WW, Bowie CR, Harvey PD (2008). Functional implications of neuropsychological normality and symptom remission in older outpatients diagnosed with schizophrenia: A cross-sectional study. J Int Neuropsychol Soc; 14(3):479-88. Abstract View all comments by Pamela DeRosse

	Comment by: Abraham Rudnick
	Submitted 11 December 2009	Posted 14 December 2009
	The diagnosis of schizoaffective disorder may be either too broad or redundant as it stands. It could be too broad as major depression may be a comorbidity of schizophrenia, similar to comorbid anxiety disorders, thus requiring the narrowing down of schizoaffective disorder to schizophrenia with bipolar disorder. It may be redundant, as if the century-old assumption, inherited from Kraepelin, that dementia praecox (schizophrenia) and manic-depressive illness (bipolar disorder) are mutually exclusive, is not endorsed anymore; bipolar disorder can also be viewed as a comorbidity of schizophrenia. Research on such comorbidity possibilities is needed in order to determine the nosological status and value of schizoaffective disorder. View all comments by Abraham Rudnick

	Comment by: Nick Craddock
	Submitted 14 December 2009	Posted 14 December 2009
	I am afraid that due to access difficulties I have read the abstract only. I am aware that it has been suggested that the schizoaffective category be abolished. I do not think this would be helpful at the present time. I have some comments: 1. The difficulty with poor reliability comes in large part from the very "narrow" definition currently used and common interpretation of "schizoaffective" as being used only when a case cannot be fitted to either schizophrenia or mood disorder categories (i.e., starting with the assumption of a dichotomy). A broader definition would be more reliable, stable—and useful. 2. Incorporating dimensional measures to sit alongside categories is an excellent idea and will obviously help with recognition of overlapping psychosis and mood features. 3. However, abolishing the schizoaffective category whilst retaining mood and schizophrenia categories would send a completely inappropriate message that would reinforce the idea of a dichotomy when we really need to be thinking about clinical spectra. 4. Accumulating genetic data suggest that there may be particular value in recognizing, at least for genetic investigations, cases with a rich mix of schizophrenia-like psychotic and bipolar mood features. 5. Making major changes to classification categories in the absence of compelling reasons would change risks of untoward consequences (financial, training, logistics, time, and confusion). In the face of evidence that a category may be useful, it makes no sense at all. 6. "Schizoaffective” cases vary enormously in the nature of symptomatology and their temporal relationship. Definitions of schizoaffective should be applicable longitudinally. Mike Owen, Mick O'Donovan, and I have written both data and review/opinion papers on this topic. Some relevant papers are listed below. References: Hamshere ML, Green EK, Jones IR, Jones L, Moskvina V, Kirov G, Grozeva D, Nikolov I, Vukcevic D, Caesar S, Gordon-Smith K, Fraser C, Russell E, Breen G, St Clair D, Collier DA, Young AH, Ferrier IN, Farmer A, McGuffin P; Wellcome Trust Case Control Consortium, Holmans PA, Owen MJ, O'Donovan MC, Craddock N. Genetic utility of broadly defined bipolar schizoaffective disorder as a diagnostic concept. Br J Psychiatry. 2009 Jul;195(1):23-9. Abstract Craddock N, O'Donovan MC, Owen MJ. Psychosis genetics: modeling the relationship between schizophrenia, bipolar disorder, and mixed (or "schizoaffective") psychoses. Schizophr Bull. 2009 May;35(3):482-90. Epub 2009 Mar 27. Abstract Craddock N, Jones L, Jones IR, Kirov G, Green EK, Grozeva D, Moskvina V, Nikolov I, Hamshere ML, Vukcevic D, Caesar S, Gordon-Smith K, Fraser C, Russell E, Norton N, Breen G, St Clair D, Collier DA, Young AH, Ferrier IN, Farmer A, McGuffin P, Holmans PA; Wellcome Trust Case Control Consortium (WTCCC), Donnelly P, Owen MJ, O'Donovan MC. Strong genetic evidence for a selective influence of GABA(A) receptors on a component of the bipolar disorder phenotype. Mol Psychiatry. 2008 Jul 1. Abstract Caddock N, Owen MJ. Rethinking psychosis: the disadvantages of a dichotomous classification now outweigh the advantages. World Psychiatry. 2007 Jun;6(2):84-91. Abstract Craddock N, Owen MJ. The beginning of the end for the Kraepelinian dichotomy. Br J Psychiatry. 2005 May;186:364-6. Abstract Hamshere ML, Bennett P, Williams N, Segurado R, Cardno A, Norton N, Lambert D, Williams H, Kirov G, Corvin A, Holmans P, Jones L, Jones I, Gill M, O'Donovan MC, Owen MJ, Craddock N. Genomewide linkage scan in schizoaffective disorder: significant evidence for linkage at 1q42 close to DISC1, and suggestive evidence at 22q11 and 19p13. Arch Gen Psychiatry. 2005 Oct;62(10):1081-8. Abstract View all comments by Nick Craddock

	Comment by: Ray DePaulo, Fernando Goes
	Submitted 14 December 2009	Posted 14 December 2009
	As noted by Marneros, schizoaffective disorder remains “a nosological nuisance but a clinical reality” (Marneros, 2003). It is a category that lacks reliability and is used with highly varying frequency in the field. Though we use the diagnosis only infrequently here at Johns Hopkins, it is applied to 20 percent or more of admissions at other academic and non-academic psychiatric inpatient units. Since the higher rate is the norm, removing the diagnoses from DSM-V would leave many clinicians in the lurch, particularly since some of the alternative proposals (e.g., the use of dimensions) might increase reliability, but are, as yet, untested constructs of uncertain clinical utility. While the use of dimensional measures (particularly in complement to categorical constructs) may turn out to have greater validity and/or clinical utility, we believe that, at present, this is a question that must be addressed by empirical research before it can be implemented in DSM-V. To implement such a major change without a substantial empirical database could increase confusion without adding clinical value. At the same time, it is imperative that researchers consider alternatives to DSM-IV SA with regard to biological origins (i.e., validity) and clinical utility. This will be an important but daunting task that is unlikely to be accomplished by the DSM-V field trials. Hence, for the time being, our suggestion is to leave the category largely as it is, though perhaps with the minor changes being proposed by Drs. Heckers and Tandon that are likely to improve the clarity of the diagnostic construct. In strong agreement with the original text in DSM-III (American Psychiatric Association, 1980; p. 202), however, we urge that clinicians use the schizoaffective disorder category only sparingly, and with the knowledge that it is an unreliable and unstable diagnosis that requires ongoing re-evaluation. References: Marneros A. The schizoaffective phenomenon: the state of the art.Acta Psychiatr Scand 2003: 108 (Suppl. 418): 29–33. Abstract American Psychiatric Association, 1980; DSM-III., p202. View all comments by Ray DePaulo View all comments by Fernando Goes

	Comment by: Jan Fawcett
	Submitted 15 December 2009	Posted 15 December 2009
	The Mood Disorders Committee in general currently supports the retention of the schizoaffective diagnosis in DSM-V. One basis for this is the very frequent use of this diagnosis in the Medicaid and Privately Insured Data Base published June 23, 2009, by Mark Olfson, MD, MPH. A study of first-episode psychoses by Mauricie Tohen found that these patients, initially diagnosed Psychosis NOS, when followed, showed the highest switch in diagnosis to schizoaffective disorder than any other diagnosis. My own personal take is based on the study by Andreasen et al. (1987) from the Collaborative Depression Study which showed that the relatives of patients with schizoaffective disorder, depressed, had a preponderance of relatives with schizophrenia, while the relatives of patients with schizoaffective disorder, bipolar type, had a preponderance of bipolar I disorder. This suggests to me a Solomonesque solution of dividing the baby, schizoaffective disorder, in half, the schizoaffective bipolars being diagnosed as bipolar I with psychosis, and the schizoaffective depressed being diagnosed as schizophrenia with major depression. References: Andreasen NC, Rice J, Endicott J, Coryell W, Grove WM, Reich T. Familial rates of affective disorder. A report from the National Institute of Mental Health Collaborative Study. Arch Gen Psychiatry . 1987 May 1 ; 44(5):461-9. Abstract View all comments by Jan Fawcett

Live Discussion: Do We Need Schizoaffective Disorder?