Diagnostic change and the increased prevalence of autism

11 Sep

ResearchBlogging.orgHow real is the “epidemic” of autism? How much of the increase in the number of diagnoses have to do with factors other than a real increase in the number?

I am going to take some time with this paper. If you want the short version of this post–about 26% of the increase in autism counts in California can be attributed to changes in diagnositic practices leading to people being classified autistic (or autistic plus MR) who were classified with mental retardation by pre-1992 standards.

Perhaps the most used dataset for exploring the increase in autism, especially by amateur epidemiologists, is that of the California Department of Developmental Services, or CDDS. The CDDS serves people with developmental disabilities (not just autism) within the state of California. The CDDS made much of its data freely available. While the CDDS data show a large increase in the number of people getting services for autism as a developmental disability, it is difficult to ascertain how much (if any) of this is due to a real increase in the number of people who actually are autistic. This is because it is very hard to know how important external factors are in changing the administrative prevalence of autism.

Many factors influence the prevalence of autism. These include broadening of the criteria for what is called “autism”, such as the change in the 1990’s to include PDD-NOS and Asperger Syndrome in the Autism Spectrum Disorders.

Recently, Hertz-Picciotto and Delwiche found that “three artefacts—younger age at diagnosis, change in the accepted criteria and inclusion of milder cases—accounted for about one-third of a 12-year rise in incidence in California.”

Many people have misrepresented Hertz-Picciotto and Delwiche as showing that there has been a “true” increase in the autism prevalence when, in fact, they state quite clearly “Other artifacts have yet to be quantified, and as a result, the extent to which the continued rise represents a true increase in the occurrence of autism remains unclear.”

One of the artifacts that was not quantified by Hertz-Picciotto and Delwiche was the possibility of diagnostic change and accretion. When diagnostic practices change, a person who would have one diagnosis in one time period might get a different diagnosis in another.

For example, a common question that comes up is this: how many people diagnosed with autism today would have been given a diagnosis of mental retardation 20 years ago?

That is essentially the question posed by Marissa King and Peter Bearman of Columbia University in the paper
The paper, Diagnostic change and the increased prevalence of autism

This has the possibility to be a very important paper. I don’t think I am stretching when I state this as this paper was published together with commentary from no fewer than five four well known research groups.

Here’s the abstract:

Background Increased autism prevalence rates have generated considerable concern. However, the contribution of changes in diagnostic practices to increased prevalence rates has not been thoroughly examined. Debates over the role of diagnostic substitution also continue. California has been an important test case in these controversies. The objective of this study was to determine the extent to which the increased prevalence of autism in California has been driven by changes in diagnostic practices, diagnostic substitution and diagnostic accretion.

Methods Retrospective case record examination of 7003 patients born before 1987 with autism who were enrolled with the California Department of Developmental Services between 1992 and 2005 was carried out. Of principal interest were 631 patients with a sole diagnosis of mental retardation (MR) who subsequently acquired a diagnosis of autism. The outcome of interest was the probability of acquiring a diagnosis of autism as a result of changes in diagnostic practices was calculated. The probability of diagnostic change is then used to model the proportion of the autism caseload arising from changing diagnostic practices.

Results The odds of a patient acquiring an autism diagnosis were elevated in periods in which the practices for diagnosing autism changed. The odds of change in years in which diagnostic practices changed were 1.68 [95% confidence interval (CI) 1.11–2.54], 1.55 (95% CI 1.03–2.34), 1.58 (95% CI 1.05–2.39), 1.82 (95% CI 1.23–2.7) and 1.61 (95% CI 1.09–2.39). Using the probability of change between 1992 and 2005 to generalize to the population with autism, it is estimated that 26.4% (95% CI 16.25–36.48) of the increased autism caseload in California is uniquely associated with diagnostic change through a single pathway—individuals previously diagnosed with MR.

Conclusion Changes in practices for diagnosing autism have had a substantial effect on autism caseloads, accounting for one-quarter of the observed increase in prevalence in California between 1992 and 2005.

The authors started by looking at the individual records for the 7003 clients of the CDDS born before 1987 who had diagnoses of autism at any time in their CDDS records. These individuals would be at least 8 years old by the time the DSM-IV criteria for autism came out in 1994, so they should have already been diagnosed with autism by that time.

What they found was that 631 individuals started out with a diagnosis of mental retardation and later received a diagnosis of autism. 95% of these individuals retained the MR diagnosis. I.e. most moved from “autism” to a dual diagnoses of autism+MR.

They also found that 89 individuals “lost” their autism diagnosis. They are not discussed in detail, so we can’t tell if they held other diagnoses after “losing” their autism diagnosis.

The authors plotted the number of individuals who changed from MR to autism or autism+MR by year. They claim that the number is higher in years when significant changes in diagnostic practices were introduced.

Figure 1 from King and Bearman paper

Figure 1 from King and Bearman paper

A peak is fairly clear for 1994, when the DSM-IV was issued. Whether this peak and the others are real is a matter for discussion (as in Dr. Hertz-Picciotto’s commentaruy)

Some of the findings the authors reported are very interesting.

Examining the control variables, we see that the level of intellectual impairment of clients had a significant effect on the likelihood of observing diagnostic change. The relationship between severity and the odds of change appears to be non-linear with moderate and profound severity to be at greatest risk for diagnostic change.

The CDDS lists intellectual impairment by the categories mild, moderate, severe and profound. It strikes me as strange that mild ID is not the area with the highest odds of change. It is very strange that there is no clear trend that the odds of change/accretion go up (or down) with severity of intellectual disability.

Another interesting observation:

Changes in evaluation scores, which capture many of the requirements for an autism diagnosis, surprisingly had little discernable effect on the likelihood of diagnostic change [OR 1.02; 95% confidence interval (CI) 1.00–1.04].

This is quite strange to me as well. One would think, perhaps, that the lower the evaluation score, the more likely that someone would have been undiagnosed.

Or, to put it another way, why weren’t the more “obviously” autistic individuals identified before the diagnostic changes?

Finally, race and year of birth were also significantly associated with the odds of change. Persons born in later years, who were younger, were more likely to experience diagnostic accretion or substitution. Finally, African–Americans
were considerably less likely than Caucasians to have a change in diagnostic status.

To me, this speaks to the idea that not everyone who qualifies for an autism diagnosis under the changes is getting one. I.e. the CDDS still has a clients in this older cohort who are misclassified as MR instead of autism. For example, 0lder clients are less likely to have a family member to advocate for them, and are less likely to see a change in services under autism vs. MR classifications.

The authors then take this “micro level” data (looking at individuals) and apply their findings on a “macro level” (looking at groups of people). In other words, the usethese data to predict how much of the increase in CDDS caseload is due to this one pathway–shift from MR to autism. The results are shown in Figure 4, copied below.

Figure 4 from King an Bearman paper

Figure 4 from King an Bearman paper

They find that by 2005, 26% of the increase in the CDDS caseload can be attributed to the shift from MR to autism.

One important point the authors make is the difference between diagnostic substitution and diagnostic accretion. An example of substitution is an individual having his/her diagnosis change from MR to autism. An example of accretion is when an individual has autism added to the already existing MR diagnosis.

Accretion is harder to discern on a group (macro) level, since one would not see the MR count drop coincident with the autism increase. The authors have included both in their definition of diagnostic change.

The authors note that the MR to autism pathway is not the only possibility.

Diagnostic substitution and diagnostic accretion along other pathways, such as developmental language disorder or other learning disabilities, may be contributing to an increase in higher functioning cases. In a study applying contemporary diagnostic standards and practices to persons with a history of developmental language disorder 21% (8/38) of the individuals met the criteria for autism and 11% (4/38) met the criteria for milder forms of ASD. Thus, there are multiple pathways to an autism diagnosis from multiple disorders that contribute to increases along various parts of the spectrum. In this article, we have considered only one pathway and one part of the spectrum

In other words, more of the increase in the CDDS caseload may be due to diagnostic changes, but in ways not covered by this paper.

One factor the authors do not appear to be taking into account is the large regional disparities within California. The administrative prevalence in the CDDS system varies wildly depending on which part of the stat one looks at. In general, rural areas have much lower administrative prevalence values than urban areas, for example.

The authors’ concluding paragraph:

We have estimated that one in four children who are diagnosed with autism today would not have been diagnosed with autism in 1993. This finding does not rule out the possible contributions of other etiological factors, including environmental toxins, genetics or their interaction to the increased prevalence of autism. In fact, it helps us to recognize that such factors surely play an important role in increasing prevalence. There is no reason to believe that any of these frameworks are wrong and many reasons to believe that the increase in autism prevalence is in fact the outcome of multiple self-reinforcing processes. However, this study demonstrates that subsequent explanations for the increased prevalence of autism must take into account the effect of diagnostic change.

I think this is quite good–the paper does not rule out a true increase in autism prevalence. It does demonstrate that factors like diagnositic change and accretion are real and significant.

Many factors are involved with the increase in autism prevalence, including that in the CDDS data. Just because the number of people identified with autism went up doesn’t mean that all of that number is due to a real increase in the number of people who are autistic.

King, M., & Bearman, P. (2009). Diagnostic change and the increased prevalence of autism International Journal of Epidemiology DOI: 10.1093/ije/dyp261

23 Responses to “Diagnostic change and the increased prevalence of autism”

  1. brian September 11, 2009 at 19:53 #

    Thanks for that summary. Could you also please briefly summarize the comments from Drs. Fombonne and Rutter? (It will be a few days before I can get to the library.)

  2. Luke Tunyich September 12, 2009 at 00:51 #

    The fact is that today many famous people have autistic children.

    How many famous people have autistic children twenty years ago?

    Do you know any famous person that had autistic child twenty years ago.

    It seems to me that the autism epidemic is real.

  3. Ruth September 12, 2009 at 02:39 #

    James Watson and Bertrand Russel both had sons with developmental issues. In the old days such children were usually sent to institutions. Einstein’s son was schizophrenic. Would some of these kids have been given an autism label today? What would Rosemarie Kennedy’s diagnosis be today?

    Steven J. Gould had a son with autism-unlike Watson and Russell he was very supportive of him. How about Beverly Sills?

  4. C. S. Wyatt September 12, 2009 at 04:05 #

    During my doctoral coursework, I compared autism rates as reported by the U.S. Dept. of Education to rates of mental retardation and general disabilities. The charts from my research are online:


    You can clearly see a decline in retardation, from 12% to 8% of special needs students. At the same time, from 1991 to 2005, autism increased by roughly similar counts of students. Not a coincidence, as I suggested in my papers at the time.

    The tables and charts should be useful to others.

    I’m preparing to defend my dissertation this year, in the field of “Rhetoric and Science” — autism has plenty of rhetorical battles to study.

  5. RAJ September 12, 2009 at 04:19 #

    “James Watson and Bertrand Russel both had sons with developmental issues”.

    James Watson was awarded the Nobel Prize for his discovery of the structure of DNA along with his collaboraters, Francis Crick and Maurice Wilkins. The discovery of the structure of DNA was an outstanding discovery in science and well deserving of a Nobel Prize. Watson was the President and Director of the Cold Spring Harbor Laboratory for 35 years and among his many honors was having Cold Spring Laboratory name the Watson School of Biological Sciences after him.

    Watson’s interpretation of what genes do, if anything, to transmit disease can be vigerously questioned without demeaning his contributions which earned him a Nobel Prize, just as Leo Kanner’s discovery of the existance of what he called ‘Infantile Autism’ doesn’t demean his his discovery while questioning Kanner’s interpretation of what causes the condition, what Kanner called Freudian psychodynamics and the refrigerator mothers inability to relate herself in a normal way to her child.

    Cold Spring Harbor laboratory has a long and sordid history. Cold Spring Harbor was one of the leading authorities in the Eugenics Movement that took place in the early 20th century. They have a history of eugenics on their website where Cold Spring Harbor apologized for their role in the Eugenics Movement.


    James D. Watson the Pesident and Director of Cold Spring Laboratory for 35 years in 2007 was forced to resign from Cold Spring Harbor after his comments, published in the Times of London, that overall people of African descent were not as intelligent as people of European descent because genetics says this is the case.


    Watsons comments offer insight into the beliefs of Cold Spring Harbor and its faculty and its long history of defective genes ‘causing’ disease model.

    Even Professor Sir Michael Rutter, the retired editor of the Journal of Autism and Developmental Disorders, has correctly referred to some molecular geneticists and child psychiatrists as ‘genetic evangelists’, a label that would apply to virtually the entire research staff of Cold Spring Harbor.

    Among his statements that got Watson in trouble was the follwing outlined in the London Times article linked above:

    He talks of the “horror and destruction” of life that can arise from having a severely autistic child, and hopes that by diagnosing autism early, “we might prevent some [autism-prone] families having subsequent children”.

    And from the same article:

    “It was a hypothetical thing,” he explains. “If you could detect it pre-natally, could a woman abort a child who was homosexual? I said they should have the right to, because most women want to have grandchildren, period. We can’t do it, but it’s common sense. Anyways,” he says, shaking his head wearily, “it was a bad day when that headline hit. I was just arguing for the freedom of women to try and have the children they want, not what is right or wrong.”

    The comment about the intellectual inferiority of people of African descent to people of white European descent and the ensuing controversy was the final straw and he was forced to resigned from Cold Spring Harbor.

    Cold Spring Harbor Laboratory lives in a parallel universe where Darwin never existed and the therefore the environment plays no role in human disease and developmental disorders, what Rutter called ‘genetic evangelism’.

  6. brian September 12, 2009 at 04:52 #

    RAJ, that has to be the best example of trolling that I’ve ever seen on the internet.

    • Sullivan September 12, 2009 at 04:56 #


      I do have to say RAJ did an amazing job making a his comment into a basically independent blog post.

      I’m sorta used to it.

      What I found interesting was that he missed Watson’s comments where he called Roslyn Franklin (sp?) autistic. Somehow it struck me that it would have fit nicely into his statement.

  7. Michelle Dawson September 12, 2009 at 05:35 #

    Sullivan wrote: “I don’t think I am stretching when I state this as this paper was published together with commentary from no fewer than five well known research groups.”

    In fact there are 4 commentaries (by Fombonne, Rutter, Hertz-Picciotto, and Charman and colleagues) with respect to King and Bearman (in press), as well as a response (wrongly listed as an “original paper”) from the authors to these 4 commentaries.

    There is also another original paper (Nassar et al., in press), epublished the same date, which is wrongly listed as a “commentary.” King and Bearman (in press) is not mentioned in this paper.

  8. ebohlman September 12, 2009 at 08:01 #

    Luke: Twenty years ago, most people would have assumed that someone who said their child was autistic was mispronouncing “artistic.” The term was little known outside clinical circles. People were also more hesitant to talk about their kids’ special needs, and celebrity coverage didn’t go into quite as much detail about its subjects’ personal lives as it does today (if Sarah Palin had been the VP nominee in ’88, nobody would have known about Bristol’s pregnancy).

    Looking at news reports about famous people is about the worst possible way to determine trends involving anything but fashion.

  9. Mike Stanton September 12, 2009 at 08:29 #

    Regarding famous people, Prince John, the youngest son of King George V and Queen Mary would probably have been diagnosed with autism if alive today. He was kept from public view in an age of deference to royalty where the media made no mention of his disorder. He died folowing an epileptic siezure aged 13 in 1919.

    His story was finally told in an award winning drama by Stephen Poliakoff, The Lost Prince that was broadcast by the BBC in 2003.

  10. Sullivan September 12, 2009 at 15:44 #

    Thanks for the correction, Michelle.

    I had the Rutter and Herz-Picciotto and Charman responses open, but I should have checked the Nassar one before writing that.

  11. Tsu Dho Nimh September 12, 2009 at 18:29 #

    More famous people: Augustus Lamb, son of Lord Melbourne (Queen Victoria’s first prime minister) had behavior and symptoms that are clearly autistic.

  12. dr treg September 12, 2009 at 22:45 #

    “Changes in practices for diagnosing autism have had a substantial effect on autism caseloads, accounting for one-quarter of the observed increase in prevalence in California between 1992 and 2005.”
    Does that mean that 75% of the increase in prevalence is unexplained in terms of changes in practices for diagnosis and represents a real increase in prevalence?

  13. Sullivan September 12, 2009 at 23:19 #

    As to the question of famous people with autistic kids in the past–

    I would start with a more general question: can we name famous people in the past who had children with any disability? There are some examples, but I think it is safe to say that many famous people would have hidden their disabled children.

  14. Luke Tunyich September 13, 2009 at 00:52 #

    I didn’t clearly stated, but in my first post I was referring to classic autism.

    Fifty years ago as well as today criteria for classic autism diagnosis are the same.

    It is obvious that famous people are unable to hide if their child or children are affected with such condition, whether is fifty years ago or this day.

    It is simply to compare the prevalence of classic autism among the children of famous people in the period from 1989 to 2009 with the period from 1969 to 1989.

    Sullivan, to avoid misunderstanding we have to agree what is autism.

    Autism is not a disability.

    The great percentage of people affected with classic autism is unable to live independent life and that lead to misunderstanding that sometimes autism is wrongly labelled as disability.

    Explained on simply level, ASD including classic autism is characterised by developmental delay and/or developmental disorder.

    Classic autism is characterised by significant delay in mental development and in significant developmental disorder of mental and physical ability.
    I am sorry, that I couldn’t explain on more simply way but I hope it will make sense to most of you.

    • Sullivan September 13, 2009 at 15:10 #

      Luke Tanyich,

      you state both “autism is not a disaility” and that it is characterized by “…significant developmnetal disorder of phsyical and mental ability”.

      Your statements seem to be at odd with eachother.

      As to the autism criteria being the same 50 years ago, that is quite clearly incorrect. As Anne has pointed out, there has been much change in just the last 30 years. 50 years ago would be before Rutter’s criteria and would place you in the era when autism diagnoses were only given to people without intellectual disability. What is called “classic autism” today is very distinct from what autism was considered in the first two decades after Kanner’s first paper.

  15. Anne September 13, 2009 at 07:35 #

    “Fifty years ago as well as today criteria for classic autism diagnosis are the same.”

    Luke, the criteria were not the same. Fifty years ago, autism wasn’t in the DSM except as a manifestation of childhood schizophrenia.

    Since autism was included in the DSM-III in 1980, the criteria have changed over time. Roy Richard Grinker has a page on his website showing the criteria in the DSM from 1980 to the present: Diagnostic Criteria for Autism through the Years. It is interesting that King & Bearman found that changes in the diagnostic criteria were associated with a substantial amount of diagnostic change for people in the CDDS system.

  16. autieobssessive September 14, 2009 at 21:25 #

    Regarding famous people with autistic kids 20 or so years ago. Sylvester Stallone has an autistic son who I’m sure is at least 30 today. And I agree with the notion that any disorders celebrities’ children might have had in the past would have been hidden. Famous people didn’t champion causes years ago as they do now. Just an observation, not a judgement.

  17. Chris September 14, 2009 at 23:21 #

    The playwright, Arthur Miller, had a son named Daniel who had Down Syndrome. Daniel spent his childhood in an institution, and his father never spoke of him.

    • Sullivan September 14, 2009 at 23:31 #

      While probably not a disability issue, consider Albert Einstein’s daughter as an example of a child completely hidden.


      No one even knew she existed until 1986. No one now knows what really happened to her.


  1. Twitted by Club_Z - September 11, 2009

    […] This post was Twitted by Club_Z […]

  2. Tweets that mention Autism Blog - Diagnostic change and the increased prevalence of autism « Left Brain/Right Brain -- Topsy.com - September 12, 2009

    […] This post was mentioned on Twitter by Fran Bristow, Autism Hub, Trevor Sheffield, Joseph Karr O’Connor and others. Joseph Karr O’Connor said: Diagnostic change and the increased prevalence of autism: http://tinyurl.com/n7dycv #fb (via @gezlemon) […]

  3. Laura Carlson (lauracarlson) 's status on Sunday, 13-Sep-09 11:37:10 UTC - Identi.ca - September 13, 2009

    […] https://leftbrainrightbrain.co.uk/?p=3031 a few seconds ago from web […]

Leave a Reply

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 )

Google photo

You are commenting using your Google 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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: