Drastic drop in ASD diagnosis variations from DSM-IV (2027+) to the proposed DSM-5 (just 11!)

Light bedtime reading

Image by Richard Masoner / Cyclelicious via Flickr

Back in February 2011, Jon Brock of the most excellent blog, “Cracking the Enigma,” wrote a post about the current number of variations in autism diagnoses. He explained that the variations are important to researchers in the field, and to understanding autism (and autistic individuals) more generally. He performed a calculation of the different ways (essentially different combinations of symptoms) that can lead to an autism diagnosis under the current DSM-IV, and reached the staggering number of 2027. That doesn’t even include the full range of what has become known as the “autism spectrum,” such as Aspergers and PDD-NOS.

That “autism spectrum” gets the official name (instead of just the five “pervasive developmental disorders”) under the proposed DSM-5, due to come in to effect in May 2013. The proposed DSM-5 makes significant changes to the autism spectrum, such as removing the terms “Aspergers” and PDD-NOS, pushing Retts out of the DSM altogether, and shuffling around the criteria for an autism diagnosis.

Not surprisingly, these changes accompany a change in the number of ways you can get an autism diagnosis. There would no longer be 12 tick-the-box options under category A of the DSM-IV criteria for autism, as discussed in Brock’s post. Instead, under the proposed DSM-5, there are four criteria that must be met for a diagnosis (A, B, C and D). Only criteria B has variations of things an autistic person might or might not have – the variant tick-the-box style that dominated the DSM-IV criteria. Within criteria B there are four options, at least two of which must be present for a diagnosis.

Essentially, the criteria from DSM-IV has been shuffled into a new formation for the proposed DSM-5. By my own breakdown, the proposed criteria A (all of which would be compulsory for a diagnosis) groups together all the current criteria A (1) (social interaction) and some of A (2) (communication). The proposed criteria B (the one with 4 options), is a mix of some of the current A (2) and A (3) (repetitive behaviours and interests), plus a new consideration for sensory issues. The only significant removal from the current criteria, is (2) (a) about a delay or lack of spoken language.

When it gets to the numbers, what all this means is that instead of over 2027 variations for an autism diagnosis under DSM-IV, the proposed DSM-5 brings that number down to only 11.

Considering how much the huge variation under the old diagnostic criteria (even if it had been far less than the number calculated by Brock), can and does affect research and general statements and understanding of autism, that significant drop in variation numbers would be a real game-changer. With just 11 variations, it would be easier to research and draw wider-ranging conclusions about autism and those with an autism diagnosis. It’s an interesting and important change to ponder, putting it into numbers just makes that change more stark and astounding.

A huge thank you to Jon Brock of course for posing the original question and calculating the 2027 number. Another thank you to his commenter on that post, mamafog, for wondering how this would change in the DSM-5 thus prompting me to find out for myself. And an advanced apology if my 11 figure is somehow wrong; I ran the process past my husband who lectures in Economics and he reached the same figure as I did independently, so I’m hopeful that means it’s correct. Do let me know if I’ve made a mistake and where the mistake has been made if you spot one! Thanks.

This entry was posted in Diagnosis and tagged , , , , , , . Bookmark the permalink.

6 Responses to Drastic drop in ASD diagnosis variations from DSM-IV (2027+) to the proposed DSM-5 (just 11!)

  1. mamafog says:

    Wow that is a big difference. I agree that I can see some advantages to only 11 “flavors” of autism. Perhaps instead of just describing individuals with autism in terms of severity, we will be able to talk about it in terms of which check boxes our child has.

    Thank you for figuring this out. It is interesting to think about. It actually gives me a little hope that “they” are not just rearranging things for the benefit of the insurance companies.

    • Interesting thought about talking about individuals in terms of checked boxes; that would indeed be easier with only 11 possibilities. I don’t think it would remove the severity references (and the DSM-5 does introduce specific reference to severity in diagnoses), but references to groupings of issues might become a more important or equally important reference point (?).

  2. Sharon says:

    I find this interesting. My son is currently pdd-nos but it’s thought he may have Aspergers yet is too young for that dx currently. That aside, he will still meet the new criteria for the DSM based on what I read above. It will indeed be of interest to see how and if stats change after 2013.

  3. I guess the point of my original post was that there are many different ways of getting an autism diagnosis under DSM IV and that this demonstrates the inherent heterogeneity within autism – even if we go for a narrow definition, excluding Asperger’s and PDD-NOS. Given this heterogeneity, I find it frustrating that research focuses more on the autism vs not-autism distinction as opposed to making a serious attempt to make sense of the variation within autism. I’m certainly not the first person to make this argument (see for example Sue Gerrard’s paper referenced below) but I think it’s one that needs to be reiterated.

    It’s certainly the way my colleagues and I are trying to go in our own research but because we’re setting studies up to look at individual differences rather than group differences, they’re not always perfectly designed for looking at group differences. As a result, we’re having a relatively tough time publishing this work because journal reviewers are conditioned to evaluate autism research where group differences are the main focus. So even when we say “we’re looking at variation in X within autism” and we find that not everyone with autism has a problem with X, and we show that being able to do X is associated with something else that might explain this variation, we still get reviewers saying “This paper is fundamentally flawed because the autism group were not matched to a control group on nonverbal mental age”. It’s frustrating but I’d like to think before too long people will come around to this way of thinking.

    As for DSM 5, you’re right that there are going to be fewer ways of getting an autism diagnosis because of the way that it the “algorithm” is set up. But the intention, as I understand it, is that most people who currently meet DSM IV-R criteria for autistic disorder should also meet DSM 5 criteria for ASD. So the underlying heterogeneity isn’t going away, it’s just being repackaged.

    For example, under DSM 5 everyone with an ASD diagnosis will need to show “Deficits in social-emotional reciprocity” but this can range “from abnormal social approach and failure of normal back and forth conversation through reduced sharing of interests, emotions, and affect and response to total lack of initiation of social interaction”. That’s one box to tick but a huge number of different ways of meeting that single criterion.

    On top of that, there’s the issue (that I didn’t touch) that similar behaviours can be observed for very different underlying reasons. So different people on the spectrum could be described as showing “reduced sharing of interests” (and thereby tick the “Social-emotional reciprocity” box) because they (a) don’t care what other other people are interested in; (b) assume that other people will be interested in the same thing so sharing interests is redundant; (c) realise that nobody else *is* interested in the same thing so there is no point sharing their interest; (d) don’t know how to initiate an interest-sharing-interaction; (e) have difficulty switching attention between the activity of interest and the other person and so sharing the interest lessens enjoyment of the activity. I’m just brainstorming here and we could play the same game for every phrase in DSM 5. I don’t know how many of these possibilities have any truth in them (I’d be interested to know from people with first-hand experience). The point being that defining autism in terms of a checklist of behaviours allows for huge heterogeneity, however the checklist is structured.

    I’m certainly not saying that we should abandon this approach to diagnosing autism because out there in the real world we need *something* and there aren’t any practical alternatives at this moment in time. But I think it’s important to bear all of these issues in mind when both conducting and consuming *research* on autism.

    Sue Gerrard’s article: http://www.springerlink.com/content/j75v5146620k120q/

    My more recent paper on similar issues: https://sites.google.com/site/drjonbrock/publications/complementary-approaches-to-the-developmental-cognitive-neuroscience-of-autism

    • Thanks Jon.

      Perhaps I should point out that I wasn’t trying to form any different or additional conclusions than your own (very interesting and important) original post did, rather than just musing that this is one more area where the difference between DSM-IV and DSM-5 becomes quite stark. (And indeed, directly following up mamafog’s own question on your post.) I’d hoped that anyone interested in the “numbers game,” clicks through to your original post first.

      Overall, I found it an interesting exercise for myself, since going through and breaking down how the criteria will potentially change (and not change) enhanced my understanding of both the current and suggested criteria for an autism diagnosis.

  4. Sue Gerrard’s comment (originally wiped by the spam filter):

    “Autism is a syndrome. That is, a group of symptoms that often occur together. Symptoms that occur together often do so because they share the same cause. But that isn’t always the case. You can get groups of similar symptoms that have different origins or quite different symptoms in different people from the same cause. In other words, not everybody with the same syndrome necessarily has the same medical condition.

    It’s important to bear in mind that ‘autism’ is a label given to a bunch of symptoms, not a label given to the underlying medical condition(s) that produce the symptoms. It’s easy to forget that, because the word ‘autism’ is widely used to describe both the symptoms and an unknown underlying medical condition or conditions that results in autism, without making a distinction between the two.

    It’s possible that there are children whose only symptoms are the autism syndrome, but most appear to have other symptoms as well. The new UK National Institute of Clinical Excellence consultation on management and support of children and young people on the autism spectrum lists sleep disturbance, feeding problems, bladder, bowel and other gastrointestinal problems , epilepsy and regression as problems associated with autism. In other words, for many children, autism is only one aspect of their symptoms. But whatever other symptoms their developmental disorder might have, those are marginalised by a diagnosis in terms of autism only.

    Even if DSM-5 reduces the number of ways of getting an autism diagnosis to 11 rather than 2027, the problem still remains: if autism is one aspect of your symptoms, you will get the same diagnosis as everybody else who has autism as one aspect of their symptoms, even if your other symptoms are very different – which strongly suggests a different underlying disorder.

    For me, the point Jon Brock was making was summed up in the last paragraph of his blog post;

    “The point is that a lot of autism research begins with the premise that even though there are 2027 different ways of getting an autism diagnosis, being in the autism ‘gang’ makes you (a) fundamentally similar to all the other gang members and (b) fundamentally different to everyone that doesn’t make the grade. That’s a big, big assumption. And I for one am not willing to make it.”

    Even if you substitute 11 for 2027, because the DSM implies that autistic characteristics form a distinct disorder, research will begin with the same misleading premise that everyone showing autistic characteristics is fundamentally similar. As he says, that’s a big, big assumption.”

    My reply:

    “Sue, you appear to have misunderstood the point of my post. You’ve also taken me to be saying and implying things I have not said or implied. I do believe what we have here is a failure to communicate. And not just because your original comment was swallowed by the filter.”

Share your thoughts:

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s