Since the sub-categories were not really useful in terms of prognosis, treatment, and education, and since they were not used consistently (even within countries there were differences in the application of e.g., Asperger versus high-functioning autism), I think this indeed makes sense. Etiological factors such as Rett can still be added as specifier, so mostly AD, PDD-NOs and Asperger are collapsed.

The previous classification gave us the illusion that we knew how to describe the ASD heterogeneity. The current classification at least admits that we do not know what all the various "Developmental disorders" under the ASD spectrum are and in that sense I think it is a positive development. On the other hand, the classification has made no progress in helping us understand the heterogeneity with the exception of the addition of certain specifiers.

Agreed, The research didn't really support the reliability and validity of these subgroups across sites. ASD, like other developmental disorders, is quite heterogeneous and it is that heterogeneity that needs to be explained in terms of both genetic and environmental (in the broad sense) factors.

Thanks to all of you who have responded thus far. I do understand that a great deal of heterogeneity exists in ASD - the consensus in the responses. But there is an absolutely huge difference in the presentation of an virtually catatonic, unresponsive child or head-banging autistic child, and one who might have mostly or exclusively social adjustment problems (formerly Aspberger's). How does DSM-V handle those differences in terms of assigning a descriptive, clinically helpful diagnosis?

Yes you are correct. They attempt to do so through the severity ratings system.

Ira,

Severity ratings, while a move in the right direction - all diagnoses should have these scaled ratings, in my opinion - I do not believe the ratings address the qualitative differences between ASD kids.

Neil

Nor do I. Notice I said "attempt." The question is why do you want those qualitative differences to be captured in a yes/no diagnosis? I suggest having quantitative, dimensional measures are much more informative.

Dimensional rankings are important, but I think one thing that has driven controversy about the changes is the sense that there are also categorical differences which are no longer recognized. I think there are categorical differences, but I also think the DSM-5 added a diagnostic category that seems to have slipped under the radar. The new category, Social (Pragmatic) Communication Disorder (315.39) seems to be like a step forward. In my opinion, many patients who previously would have been classified as Asperger's, "Atypical Autism", or PDD NOS fit well into this category.

DSM-V has changed the required 3 key areas of impairment to 2 , thereby broadening the definition of ASD. It has included CDD but excludes Rett's syndrome . The age criterion remains at 3 years.

Neil, Ira,

I agree that there is a great deal of heterogeneity in persons with ASD. I think the question is what you want/need DSM for. Do you want a yes/no diagnosis, indicating that the person needs help and giving them access to professional help? then perhaps yes/no is fine. if you want detailed assessment helping you to develop a treatment plan, you need much more information: indeed the kind of dimensional severity scales, perhaps also social subtype according to Wing, but I think you also need to have information on the context and netwerk of the person, because much of your treatment option will depend on that. Same goes for presence of behaviour problems, anxiety, ... . this can not all go into a diagnostic system, and I do not think it is easy to draw the line between what should and what should not be included, once you move beyond a yes/no decision. Any idea's on that?

Petra,

My opinion is that DSM's main function is to define the disorder as precisely as possible, so that a treatment plan can be formulated. You are absolutely correct that there is no way to include all the assessment information one needs into the DSM - that comprehensive assessment actually comes first and is carefully analyzed, both quantitatively and qualitatively, to assign a diagnosis and suggest treatment options.

Establishing a definitive diagnosis is important with regard to treatment. For example, a DSM differential diagnosis would help a clinician distinguish between Bipolar Disorder and Major Depression with psychotic features. The treatment for these two disorders would be very different based upon the DSM diagnosis and all the research that is associated with those two diagnoses.

So the main question is, I think, how is DSM able to distinguish between the different levels of ASD, both for treatment and research purposes?

Neil,

I agree that DSM would need to include information that is relevant in distinguishing disorder that have different treatment needs. I think this was exactly the problem with the PDD subtypes, that they did not distinguish differtial treatment needs, and that the variables that do matter, such as social subtype and the other mentioned above, are as variable within as they are between categories. I'm afraid that the scientific advantages have not been made so far to distinguish ASD categories for this. It is also, in spite of its importance, not a field that is easily funded, unfortunately. In intervention studies people like to include groups as homogeneous as possible, in order to have a higher chance to detect treatment effects, which often are already small.

Neill

That is the issue. The categories listed in the DSM-IV were found to not be used in a reliable manner and hence were of limited value for research or for treatment recommendations. The definitions were vague for PDD-NOS and, for Aspergers', easily confused with autism and PDD-NOS. Parents were also confused about their child's label. Thus, not comparable to bipolar vs MD. We're not there yet and questions are also being raised about the meaningfulness of the whole diagnostic system from an etiological standpoint. Of course, when you change how a disorder is defined, it will likely impact prevalence estimates so the extent of this potential effect remains to be seen.

That is why, at this point, it is better to have an overarching diagnosis of ASD (where there is good reliability across clinicians) and then provide more dimensional (and categorical) information that can be informative for treatment, e.g., co-occurring behavioral syndromes such as an anxiety disorder and/or ADHD, communication skills, non-verbal skills, severity of the various dimensions that comprise ASD, neurological complications, other medical complications, family history, attitudes, etc. These can be obtained through a careful history, observation, and rating scales, and, perhaps, the severity and level of support ratings listed in the DSM-V.

Ira,

Excellent points, all. I do not necessarily agree that there is agreement and reliability across clinicians with regard to the spectrum of kids who may or may not be given a diagnosis of ASD. The reason for this is the very co-occurring factors you mentioned that cloud the clinical picture. I know of a young girl who was given PDD-NOS (because of her unusual and varied presentations to various clinicians), but yet as she moved out of childhood and into adolescence her unsual behaviors subsided (which were likely, in retrospect, due to high introversion and a rich imagination) and she began to socialize normally. Thus my opinion is this case was that there was both uncertainty in the early diagnosis - thus NOS, and the diagnosis was ultimately not valid. So we have issues in ASD of both reliability and validity. I am not at all sure DSM-V will address either of these key diagnostic issues adequately.

Is the DSM research based? I thought it was more dependent on the philosophy of the group working on each revision and very political in nature.

Susan,

DSM is definitely research-based. However, as with any data, it can be interpreted in different ways and that process lends itself to the exercise of political factors. I am, of course, using the term "political" in the sense of research data being either emphasized or discounted based on the biases, perceived power, or prestige of individuals within the work groups.

Really? Have you read "The Book of Woe" by Gary Greenberg? There is a lot of literature indicating how certain diagnoses were developed or excluded and they are not all based on research. The National Institute of Mental Health has always been critical, calling it unscientific. It just happens to be the primary money maker for the APA. I don't know many clinicians that take it too seriously except to submit the diagnostic code to insurance companies.

Susan,

I think we'll have to agree to disagree on the DSM. Though it has faults in its either/or orientation, I actually know of no clinicians or researchers who trash it, though they may be critical of certain parts of it. And it is certainly not the easiest system to use, especially around differential diagnosis.

I would be interested to hear of diagnoses that have no research basis. It may be old or controversial research, but research nonetheless. Take a look at Appendix J in DSM-IV, which includes 20 pages of contributors from many different fields and countries and lists the field trials of the diagnoses.

Having said that, the DSM itself knows that our knowledge is evolving and hence the release of DSM-5 and its changes.

Neil, Start off by reading "The Book of Woe". Off the top of my head the Antisocial Personality Disorder has been a controversial inclusion in the DSM IV. It is based on social constructs and was developed by a woman, who if I remember correctly, had her roots in Sociology. It has been trashed, as you say, by Robert Hare, who is considered the foremost authority on psychopathy and by Theodore Millon, who pretty much disagrees with all of the personality disorders in the DSM. I always found his book on personality disorders so much more helpful.

As a former clinician for 17 years, I can say that people do not fit into neat little boxes. I did the best I could to put someone in one to appease the insurance companies. The real use I have found for the DSM (and when I started working we were using DSM II) is for research purposes. Say, you are doing research on a certain intervention on children with ASD. It is helpful if everyone has been diagnosed with that disorder to make your research clean. I

I have no problem with agreeing to disagree. None at all.

Susan,

You are certainly entitled to your opinion, and I'm sure we will find lots of different ones on this topic of diagnostic categories. But this particular question was not meant to be a referendum on DSM, but getting input on the most controversial of the changes that was made to DSM-5, viz, the collapsing of several childhood ASD disorders into one ASD diagnosis. I'd like to keep the discussion on this topic.

"DSM-V uses Autististic Spectrum Disorders to encompass many developmental disorders that previously were distinct disorders"

I don't think many people thought these were distinct categorical disorders, more variants of closely related disorders, or on a dimension, differing quantitatively rather than qualtiatively.

"Is this change warranted?"

It is too late to worry about this. DSM 5 is here to stay so we will have to get used to it. In particular, sensory abnormalities are now recognised, which means researchers can now get grants to study them.

For years I have been convinced that there is a preliminary very common form of Meniere's Disease, but I struggled to find a suitable name. I have been incorrectly quoted as believing that Darwin had Meniere's Disease, albeit my own fault as I did not have a suitable alternative name. Indeed, he would not have been given the MD diagnosis by any modern otologist. So thank you DSM 5 for legitimising ASD, as it suggested what I think is a precise term for the disabling condition that Darwin and many others have had, ie Meniere Spectrum Disorder.

As someone who as been training therapists about autism and diagnosing pervasive developmental disorders since DSM III-R was in effect, I feel that the appropriate evolution of categories has taken place.

Asperger's Syndrome was added in the DSM-IV to cover people who had the social communication difficulties and the repetitive and rigid behaviors of autism but not the deficits in language or intelligence. You can have autistic features but have a full vocabulary!

But the DSM-IV acknowledged that people can vary along a variety of dimensions, that people with Asperger's did have language issues, and that intelligence is not strictly related to functioning in autism.

The woman who made the US aware of Asperger's work never meant for these kids to be in a separate category; Lorna Wing said that she wanted to call attention to a full spectrum of kids.

And now we are there, recognizing that there are a variety of kids with persistent, pervasive problems in social understanding, social communication, and have sensory, coordination issues and rigid, restricted patterns of behavior (especially under stress. As I read the research, these are people with brains that with disorganized local neural networks (except in the visual cortex) and poor communication and connection across brain regions, thus not handling complexity well.

And these kids are not all along one single spectrum: as the DSM-5 notes and asks us to define in diagnosing, they can vary in language ability, intelligence, and degree of impairment of social communication/perception. This last point is why I titled my autism blog "AutismSpectrA",* not SpectrUM; there are a few dimensions to consider separately instead of still assuming that all traits vary together.

*autismspectra.blogspot.com, where I address the question you're raising.

About PDD-NOS: the type of case you mentioned would probably not be diagnosed with ASD, but the DSM-5 does have a place for people who either don't meet the full criteria for ASD or for whom more information or time is needed to be sure: "315.9 Unspecified Neurodevelopmental Disorder". By the way, another specific disorder like Williams Syndrome or Angleman Syndrome (not so sure about Meniere's) affecting many areas of functioning but not specifically listed in the DSM-5 are also not excluded as some think: they would be diagnosed as 315. "Other Specified Neurodevelopmental Disorder." NOS is gone, but two new categories replace it, each with a more clear and limited use.

Hope this helps!

"not so sure about Meniere's"

I am not expecting Meniere Spectrum Disorder to be in DSM any time soon. However, many autistic adults writing autobiographies clearly describe relevant MSD, and Temple Grandin actually puts her hand up to Meniere's disease.

As you say, the DSM-IV listed four PDDs. But one of these (Childhood Disintegrative Disorder) was vanishingly rare - I don't think 100 cases were diagnosed outside of the Yale Child Study Center before its inclusion. It was also very ill-defined: basically, just "really severe autism that arose a bit later than usual." Another (Rett's Disorder), while rare, seems very well-defined. But it seems like a very specific neurological disorder whose basis is not yet fully understood. As such, it should group with Fragile X, Cornelia deLange, and other such disorders - it isn't defined mainly in terms of behavior.

Annd then there were two: Autistic Disorder (basically Kanner's infantile autism) and Asperger's Disorder. The big dividing line here was, of course, language (at least in its semantic and syntactic aspects). In addition, those with Asperger's Disorder were almost always in the average (or better) range of intelligence, while those with Autistic Disorder were usually in the intellectually disabled or borderline range. I think the distinction between "autism with serious language impairment" and "autism despite grossly intact language functioning" is real and important. In this respect, the DSM-5 could be seen as a mistake. But on balance, I have to acknowledge the force of the authors' reasoning. In practice, lots of kids with more isolated social communication deficits were getting diagnosed with Asperger's Disorder, a very expensive error. But since Nonverbal Learning Disability wasn't in the DSM-IV, that diagnosis was little used (like Asperger's Syndrome before DSM-IV). The hope is that the current criteria will make it harder to shoehorn a learning-disabled person into the autistic category. And now that Social Communication Disorder is available, its use should start to expand.

"those with Autistic Disorder were usually in the intellectually disabled or borderline range."

This may have been true in the past, at least for diagnosed cases, but there have always been high IQ persons with autism, as Kanner noted. In fact, I don't think there is any upper limit to the IQ in autism, unlike Down's syndrome for example.

Categories are properly judged by their usefulness, not their truth (what would that mean anyhow?). The degree to which categories are accepted and used is influenced by this and other factors such as the power of their advocates. Busy clinicians need diagnoses to have a way of categorising and addressing their patients' problems, and communicating about them, just as any profession need its working language (its jargon). The ASD category has proven useful, but has, like ADHD, expanded to include the behaviour of more and more people.

The definition of ASD and most other categories in DSM5 is crucially in terms of behaviour. It is important to remember that, strictly speaking, there are no autistic people, only autistic behaviours. There is nothing new about this way of thinking in medicine, a person is are not “a pancreatic cancer”, or “an influenza”, they show signs and symptoms of these “illnesses” and test positive if the condition, or sub class of it, can be identified this way, but the sufferers differ in many other ways. Similarly people can show autistic behaviour but differ in many other ways. Unsurprising heterogeneity.

The two features of autistic behaviour, a tendency to avoid communicative interactions and to avoid unpredictability can occur separately and would probably get labelled as social communication disorder and OCD respectively.

The intensity with which these two behavioural tendencies is shown varies both within and between individuals, from strong social avoidance and obsessionally through gauche over-the-top social behaviour to some social timidity and a tendency to be orderly. More unsurprising heterogeneity.

Moreover the frequency with which this behaviour is shown varies greatly amongst individuals according to the prevailing causal conditions, some show it most of the time in most environments and their behaviour would be described a classical autistic; others show it only under considerable stress and at other times behaviour is ordinary ways. Yet others fall between these extremes. Yet more unsurprising heterogeneity.

The term "disorder" refers crucially a claim by clinicians that some problems come within their competence (which is not to deny the suffering of many patients). It is an implicit part of the job description.

Syndromically defined “disorders” are distinguished to make it easier to initiate and communicate different treatments, in a state of inadequate knowledge, and there is the hope, occasionally fulfilled, that they differ in aetiology. Whether these “disorders” are useful scientific categories is quite another matter, although many protagonists would like to argue that the categories are scientifically kosher, to add respectability to them.

So is the change warranted? Well, it comes closer to the dimensional reality of autistic behaviour revealed by ethological direct observation and experiment. It also accepts that this behaviour can associated with a wide variety of other features. Both steps in the right direction.

Apologies for some typos: 4th paragraph from end should read:

Moreover the frequency with which this behaviour is shown varies greatly amongst individuals according to the prevailing causal conditions, some show it most of the time in most environments and their behaviour would be described as classically autistic; others show it only under considerable stress and at other times behave in ordinary ways. Yet others fall between these extremes. Yet more unsurprising heterogeneity.

From a clinician's perspective, it is good to see ASD as a term formalised. However, it is interesting that in this arena other syndromes are rubbished because they are not in the ICD or DSM classifications, such as PDA or DAMP. We have been using ASD as a label for at least 10 years before it was formalised in DSM. The point made above regarding the tendency of broadly defined diagnoses to attract more and more within them can be partly explained by the need for a label or diagnosis to access ever more limited public funds for interventions educational or otherwise, partly because the awareness has increased significantly although in my experience within adult mental health has a long way to go, and in part because, and this is the controversial point, true prevalence may be increasing due to epigenetic sociodemographic risk factors- e.g. later age of parents at conception. Whilst a healthy discussion on all these points is warranted, I am more concerned that the removal of PDD NOS to social communication disorder has relegated a highly complex group of children to receipt of fewer services as ASD exclusion is a requirement of the social communication disorder diagnosis. This group includes children who are often the most difficult to place educationally and who have high comorbidity in terms of risk factors for social exclusion. Indeed, this paper actually says the only clear difference between the autistic disorder and PDD NOS cohort was that the latter were at greater risk for mental disorder: http://www.ncbi.nlm.nih.gov/pubmed/20556500

And this paper is a good review of the potential benefits of a dimensional approach in managing the inherent heterogeneity of ASD http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3671160/

It is important to remember that diagnosed individuals do not see themselves as a cluster of behaviours but adopt their diagnosis as part of their identity. Diagnosis therefore takes on more complex meanings then those intially intended. By collapsing the diagnostic label we have produced tension between clinicans who want to alleviate distress and high functioing individuals who do not want to be seen as disordered but to have their difference recognised and validated. Both positions are valid in my opinion, but is there a place for both in the DSM?

"we have produced tension between clinicans who want to alleviate distress and high functioing individuals who do not want to be seen as disordered but to have their difference recognised and validated. Both positions are valid in my opinion, but is there a place for both in the DSM?"

Many high functioning paranoid schizophrenics want above all to have their views validated, but I still think they should be in the DSM.

If someone with "high functioning autism" really is happy with his life in every respect, and would not consider he needed to see a psychiatrist, then he would presumably not qualify for DSM.

I think the previous answer highlights another important question. Whose distress are we trying to allevaiate? An individual with "high functioning autism" may be very happy but their spouse and family may be less so. Who is disordered in this instance and why?

A second but related point is that the DSM presents a dichotmous view of individuals. they are either disordered or they are not. Ironically many with ASD prefer to see things as more nuanced then this.

Rachel Lewis makes an important point about how diagnosis is viewed by the diagnosed individual themselves. Part of the clinician's art is discussing the "diagnosis" (I leave aside the question of quality of the diagnostic system) so that it makes sense in the first person and is helpful to that particular patient. But the inside view will differ from person to person even when their observed behaviour falls into the same category. The inside view cannot dictate the diagnostic categories otherwise the system would be chaotic.

This is definitely not to say that the inside view, the individual's own subjective experience, is not important, of course it is, and clinicians come to understand the range of such accounts associated with certain behaviour and relationships. Also it is not to say that it is not important heuristically, pointing investigations in certain directions.

This good practice is not helped by the poor quality of many diagnostic categories, with their pseudo objectivity. (see chapter on "Ethology and Mental Illness" in Evolutionary Science of Human Behavior: An Interdisciplinary Approach. Eds. P LaFreniere and G Weisfeld To be published Autumn 2014)

Thank you to John Richer for what looks like a very helpful reference. Although I do not diagnose, I can appreciate that it is a complex art form, requiring careful interplay between objective observations and the subjective accounts of patients. I cannot speak for the quality of the diagnostic category of ASD. I can only draw attention to the difficulties of applying a set of behavioural criteria to a heterogenous and mysterious condition.

There is a great deal of variability in the significance that might be attributed to any given behaviour. There is consequently a great deal of variability in how individuals are diagnosed and who gets diagnosed. Moreover the purpose of diagnosis is unclear when treatments for autism are varied, patchy and in some cases controversial.

The inside view of the patient might therefore be one of the best indicators a cliician might have for evaluating the value of a diagnosis. Arguably, it is the patient and their family who can most accurately judge whether a diagnosis is going to be helpful or not.

Rachel Lewis is quite right. A diagnosis in this area is judged more by its utility not by truth. And utility to the diagnosed person is paramount but in two senses, (i) how that person understands and uses that diagnosis (ii) how the diagnosis helps the clinician to help the person.

One point I would not agree with: heterogeneity does not mean that behavioural criteria cannot be applied and be useful. And mystery applies to the diagnostician's problems in understanding, and that is simply a challenge not a barrier, as is the case is all science.

To repeat, the challenge, and privilege and satisfaction, of clinical work is to juggle with the dialogue between these two views of the world: the scientific / observer's view on the one hand and the insider's / subjective view on the other. But then this is only an extreme intensification of what we all do everyday.

"the DSM presents a dichotmous view of individuals. they are either disordered or they are not. Ironically many with ASD prefer to see things as more nuanced then this."

I think most of those involved with DSM realize quite well that behavior and mental illness is dimensional, and that it was mainly for practical reasons that a predominantly categorical classification was retained in DSM 5.

As an assistant psychologist working in a diagnostic service for both Autism and ADHD, I personally think it was a rather bizarre move to ignore deficits in social communication as this is one of the major criteria we as a service look for when using tools such as the ADOS or ADI-R!

Social communication is known as one part of the 'triad of impairment' used to define the characteristics of Autism... ignoring it just does not make a lot of sense to me.

diagnostic criteria A. in DSM V deals with deficits in social communication. I think the range within this criteria is quite broad and encompasses communication at the high functioning end.  These individuals are able to operate at quite a sosphisticated level but may need more complex support with intimate adult relationships or with parenting. One of the questions is whether or not it is appropriate to group high and low functioning individuals together within the same diagnostic category. Jemma Coates concern may be a reflection of that concern. However, my experience with the autistic community is that definitions of high and low functioning are more fluid then our diagnostic categories allow. There seems to be a strong sense of social cohesion across the spectrum, suggesting that many individuals wth ASD are happy to be lumped together no matter at what level of communication they operate.