Onset patterns in autism: correspondence between home video and parent report

27 Jul

Regression is a major topic in autism. Children who lose abilities at a very young age. In Onset patterns in autism: correspondence between home video and parent report, Sally Ozonoff, Ph.D. and a team of researchers at the U.C. Davis MIND Institute looked at the developmental trajectories of children, autistic and non autistic. They reviewed home videos of the children to map those trajectories. They monitored social communication as a function of time.

What they found was even more complex than expected. Instead of finding that some children show low levels of social communication from very early in life. A second group has early high levels of social communication, followed by significant decreases over time (regression). But, there was a third group: a group which was more typical in development followed by not regression, but a plateau in progress in social communication.

The numbers of children in the study are small (53 autistic children), making it unlikely to get a precise idea of what fraction of the children follow each trajectory. According to the IMFAR abstract for this study:

Bayesian Information Criteria were used to select the number of trajectories that best fit the data. There was strong support from coded home video for 3 onset trajectories. The first “early onset” trajectory (n = 20) displayed low rates of social-communication behavior at all ages. The second “regression” trajectory (n = 20) displayed high levels of social-communication behavior early in life and significantly declined over time. The third “plateau” trajectory (n = 12) was similar to the typical children early in life but did not progress as expected. There was no support for a mixed (early signs + regression) trajectory.

Here is the abstract for the published paper:

The onset of autism is usually conceptualized as occurring in one of two patterns, early onset or regressive. This study examined the number and shape of trajectories of symptom onset evident in coded home movies of children with autism and examined their correspondence with parent report of onset.

Four social-communicative behaviors were coded from the home video of children with autism (n = 52) or typical development (n = 23). All home videos from 6 through 24 months of age were coded (3199 segments). Latent class modeling was used to characterize trajectories and determine the optimal number needed to describe the coded home video. These trajectories were then compared with parent reports of onset patterns, as defined by the Autism Diagnostic Interview-Revised.

A three-trajectory model best fit the data from the participants with autism. One trajectory displayed low levels of social-communication across time. A second trajectory displayed high levels of social-communication early in life, followed by a significant decrease over time. A third trajectory displayed initial levels of behavior that were similar to the typically developing group but little progress in social-communication with age. There was poor correspondence between home video-based trajectories and parent report of onset.

More than two onset categories may be needed to describe the ways in which symptoms emerge in children with autism. There is low agreement between parent report and home video, suggesting that methods for improving parent report of early development must be developed.

The last statement in the results is obviously intriguing. “There was poor correspondence between home video-based trajectories and parent report of onset.”

Here is the segment of the IMFAR abstract:

There was poor correspondence between parent report and home video classifications (kappa = .11, p = .30). Only 9 of 20 participants whose home video displayed clear evidence of a major decline in social-communication behavior were reported to have had a regression by parents. Only 8 of 20 participants with evidence of early delays in social-communication on video were reported to demonstrate an early onset pattern by parents. Of the 10 whose parents described a plateau, only 3 had home video consistent with this pattern.

So, none of the three groups were able to correctly recall the trajectory. Not the parents of kids who regressed. Not the parents of kids who plateaued. Not the parents of kids who had early onset autism.

As parents we’d like to see ourselves as the experts of our children. And, frankly, we are. No one else knows them like we do. But that doesn’t make us infallible.

The study was presented at IMFAR and Shannon Rosa wrote about it for The Thinking Person’s Guide to Autism.

20 Responses to “Onset patterns in autism: correspondence between home video and parent report”

  1. Barbara July 27, 2011 at 10:12 #

    That IS the most important finding in the study – the poor correlation of video evidence with parental memory. But it does beg the question of whether the video clips were typical, and whether the raters’ observations can, in fact, be correlated with the far more ‘instinctive’ and holistic view of parental experience? Observations have a behavioural bias, which can sometimes be superficial. But on the other hand, diagnosis is usually done on the basis of behaviours.

    I’m going to have to think some more about this one…..

  2. Stuart Duncan July 27, 2011 at 12:11 #

    Sadly, I have talked to a few parents that swear their child regressed around the age of 2 only to find that a lot of signs were really there much earlier.
    So yeah, I could definitely understand how stories might not always match up.

  3. Bad Mommy July 27, 2011 at 13:59 #

    That is not surprising, actually. You don’t know what you are looking at with a very young child as a parent – and before you reach an awareness that something is not as expected, you might miss all kinds of signs. As humans we really only see what we are looking for, and have a remarkable capacity to reshape facts to fit a pattern that we expect to find — in the present or in our memories of the past.

    Considering that babies don’t even social smile before a certain point, how is a parent to realize that things are not progressing as they should? There are some big steps around 1 year to 16 months that so many kids fail to take, but the cumulative effect of those missed steps is finally apparent around 20 months. The difficulty is in making parents understand that their perception of the situation might not match an objective description for many reasons that have nothing to do with your qualities as a parent.

  4. passionlessDrone July 27, 2011 at 14:50 #

    Hello friends –

    The Kennedy Kreiger group has been talking about plateau for a while now.


    Plateau (n=17%): Display of only mild developmental delays until the child experiences a gradual to abrupt developmental halt that restricts further advancement of skills

    Does this mean that we can finally, finally put an end to the narrative friendly, “autism is not developmental stasis, but developmental delay” canard?

    – pD

  5. livsparents July 27, 2011 at 16:04 #

    The critical piece, in my mind, is getting these ‘groups’ right so that the right questions can be asked early. Our pediatrician completely missed Livie’s regression because he asked whether she had “10 or more words” and not “has she lost a significant number of words over the past XX months”. The other obvious benefit is that one or two of these subgroups may be a more ‘natural’ indication of autism, while rapid regression is more of an indication of a possible medical issue…

  6. Leila July 27, 2011 at 16:33 #

    The thing is, looking back, I’m not 100% sure if my child belongs to either of the groups. I want to say my son is on the “plateau” group, but you may have found some mild early signs, or you may have found some mild regression. Could he belong to a 4th group (which doesn’t exist, according to this study), or am I also being a poor reporter of his early development? I’d love to have Dr. Ozonoff review my home videos!

  7. farmwifetwo July 27, 2011 at 19:35 #

    The problem with slow regression is that you don’t notice it until it becomes an issue. It’s like the day nearly a year ago I realized all of a sudden we’d just had “joint attention” and I’d missed the opportunity to continue the dialogue. Or all of a sudden your child has mastered a skill and they probably did days ago but living with them day in and day out you’ve missed it.

    My severely autistic one was the one that developed “normally” except for the speech that never happened. He is speaking more and more now at the age of 9. It wasn’t until 2.5yrs that we saw the “autism”, it was 3yrs at our Dev Ped appt that I told her that he’s finally “autistic” (no longer social, the touching, sensory etc). Yet, the mild one hit every milestone with a twist, had the behaviour, speech disappearing and that was what we were looking for. We did start speech at 18mths with the youngest but even the SLP who’d worked with all kinds of children for many years didn’t know if he was autistic or not.

    I simply think each child is different and even though you can “guess” early. Age 2.5 to 3, IMO is when you can tell there is something going on.

  8. Tim July 27, 2011 at 22:23 #

    Count us in the group that would have bombed the video vs. recollection test, and I consider my wife and I to be pretty observant people about a lot of things. Looking back at videos now, it’s obvious, but we didn’t know enough back then to see it. So this study doesn’t surprise me a bit.

    Seems like this also casts doubt about the reliability of the ADI-R. I know no doctor worth much relies solely on it but on a broad array of diagnostic methods when doing an eval, but having gone through the ADI-R interview three times now for research studies we’ve participated in, its limitations start showing. You have to get parental input somehow, but controlling for our memory limitations may be too complex for any one method.

  9. Chris July 27, 2011 at 22:57 #

    I guess we would not be among the parents with bad recall. I had signed up my son in an infant/toddler language survey where I had to fill out a circle next to a word that he uttered. I usually sent it back without any of the little circles filled in.

    He got dropped from the study about the time I failed to send one in during the time of his multiple trips to the ER for croup (about eighteen months). He was obviously an outlier, which was not unexpected due to his history of seizures.

  10. Nightstorm July 27, 2011 at 23:01 #

    Does this mean that we can finally, finally put an end to the narrative friendly, “autism is not developmental stasis, but developmental delay” canard?

    No. Because then you’re making a blanket statement to other individuals that have delays and not stasii.

    Seems like this also casts doubt about the reliability of the ADI-R. I know no doctor worth much relies solely on it but on a broad array of diagnostic methods when doing an eval, but having gone through the ADI-R interview three times now for research studies we’ve participated in, its limitations start showing.

    I think this is why the ADI-R was done last when I was reDX’d it was after the barage of testing over the months that last tests were the ADI-R. It was more of a confirmation on the other testing that has been done than a basis of DX. It basically told yes this data syncs up with this data here.

    My mother also saw early onset signs when I was an infant. It was only after the testing that she realized what she saw, but she also other kinds of delays that weren’t neuro-developmental (I was late in the physical end like rolling over and walking). She didn’t put two and two together until she did the ADI-R. The lack of eye contact the unwillingness to socialize as an infant or play social games as a child. Those all added up

  11. Kasey July 27, 2011 at 23:14 #

    I am with Leila. I think mine would be in a 4th group. Our son (now 7) seemed to develop ‘nomrally’ in most ways/hit milestones on time. But, looking back I think there were some signs even if it wasn’t true regression. We seemed to hit a plateau with development. There was no obvious regression.

  12. passionlessDrone July 28, 2011 at 17:03 #

    Hi Nightstorm –

    No. Because then you’re making a blanket statement to other individuals that have delays and not stasii.

    The only statement I am making is that “autism is not developmental stasis, but developmental delay” does not match up with reality, nothing else; I am not making any blanket statements.

    For example, I’m completely OK with:

    “Some children with autism are delayed, some aren’t delayed, some regress, and some plateau.”

    This is the reality.

    The upshot of this is that the oft quoted, never substantiated, statement that autism is categorized by developmental delay, as opposed to stasis, can no longer be used to beat up on parents who believe they have helped their children by treating them medically.

    Although, it does occur to me that no one ever invoked the stasis canard if someone were to claim that ABA or speech therapy had helped their child. Go figure.

    – pD

  13. Sunshine July 29, 2011 at 07:06 #

    LOL why is it surprising that parents have less than perfect memories, especially on a subject that is likely to be pretty emotional? This is pretty predictable given the human brain’s need to make connections and fit pieces together. The fact that autism is not usually apparent until a seemingly normal child is nearing 2 or 3 years old makes it predictable that parents would look back retrospectively and search for clues, and often remember things wrong.

    As far as I can tell, my son had some early onset symptoms and definitely some regression. The study you mentioned (small as it is) didn’t even find any evidence this happens. (shrug) What I know for sure is that when I figured out my son was autistic, I looked back on all his behaviors up until that point and tried to fit them with my new knowledge. Now how accurate is that going to be?

  14. Sunshine July 29, 2011 at 07:16 #

    Does this mean that we can finally, finally put an end to the narrative friendly, “autism is not developmental stasis, but developmental delay” canard?

    Passionless, you’re wrong. The plateau the study is referring to is a plateau as an ONSET of the symptoms. As in, your kid hit all his milestones up until he was 12 months, and then he remained at a 12 month skill level for some time. 12 months would be considered the onset, then. This study does not address where these kids are a year or 2 down the road. They are not referring to a permanent stasis- and, in fact, the term “plateau” rarely refers to permanent stasis. ASD is not retardation, period. This isn’t an emotional or political issue- it’s a medical issue. That even those with the severest forms of ASD, with intellectual disabilities, are highly likely to make intellectual progress throughout their lifetime. What you are suggesting is just not supported by science, and certainly not supported by this study.

  15. sharon July 29, 2011 at 10:00 #

    Thanks Sunshine, that’s the point re. pD’s comment I was going to make. A plateau at the time of onset does mean the child, and eventual adult does not continue to develop. I’m sure you know that pD so find it an odd remark.

    As for this study, I wonder if my son might have plateaued? Even though retrospectively his Autism was apparent from a very early age, as seen by avoidance of eye contact, there was no regression. My question is how these differing onset patterns manifest over time? Are there any differences in developmental trajectories for each group?

  16. Terrell Spanswick March 11, 2013 at 06:49 #

    New findings published in Pediatrics (Epub ahead of print) by the Kennedy Krieger Institute’s Center for Autism and Related Disorders reveal that 70 percent of children with autism spectrum disorders (ASD) who have a history of severe language delay, achieved phrase or fluent speech by age eight. This suggests that more children presenting with ASD and severe language delay at age four can be expected to make notable language gains than was previously thought. Abnormalities in communication and language are a defining feature of ASD, yet prior research into the factors predicting the age and quality of speech attainment has been limited. ;

    My own, personal web site

  17. Lara Lohne April 8, 2013 at 17:53 #

    Either the above comment is major spam or you are selling products from your blog that I’m not aware of. That hack job really affected a lot of different things, didn’t it?

    • Sullivan (Matt Carey) April 10, 2013 at 18:19 #

      We got a lot of spam getting through the filters lately.

      Are you referring to the hack at Science Based Medicine? We were not affected by that.

      • Lara Lohne April 11, 2013 at 01:34 #

        I thought the hack was ‘wordpress.com’ not specifically SBM. I’ll need to re-read the email they sent me lol.


  1. Autism Blog – Onset patterns in autism: correspondence between … | My Autism Site | All About Autism - July 27, 2011

    […] Visit link: Autism Blog – Onset patterns in autism: correspondence between … […]

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 )

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: