Regression Analysis

Regression. Possibly THE most frustrating aspect of Joy's development, for me anyway.

For those not living in autism-land themselves, it might surprise you to learn that big regressions are not necessarily a part of the autism trajectory. The most highly-publicized autism stories (especially in the "what causes autism" controversy accounts) often involve a great big regression at some point in the first two years of life. The child retreats into him- or herself, becomes withdrawn, loses language, almost seems like a different person -- and when the stunned parents go looking for a diagnosis, autism is what they discover.

While this is definitely one piece of the autism story, it is by no means what everyone experiences. Some kids with autism don't regress at all, others plateau, others show symptoms from the beginning and then regress, others have multiple regressions over time. There's a nice round-up of several studies on autism-related regression in a June 2008 IAN Research Report (the Interactive Autism Network, sponsored by Autism Speaks). One of the studies they cite found 46% of their sample of kids on the autism spectrum had experienced regression, but only 30% of those reported normal development up to that point. Another study with a stricter definition of regression found regression in their sample in only 11.8% of children with autism and 5.5% of children elsewhere on the autism spectrum.

Joy's in the "early symptoms" and "multiple regressions over time" categories.

I blogged in July about the "early symptoms" (When Did You First Notice...?) We definitely had a whole cascade of symptoms before we hit a regression, from language problems to interaction deficits to sensory issues, with whipped cream and seizures on top.

The first regression was subtle enough that we didn't catch it in progress. Sometime in late fall of 2006, at the age of 2 1/2, Joy started losing words and started ramping up with behavior issues. The timing was probably pretty good, if you can say such a thing, because we'd been waiting for an appointment with a prominent developmentalist who specialized in autism, and the appointment was Dec. 21. That's when we got the autism diagnosis, and the doctor pointed out autism issues left and right. She chastised me for having filled out forms describing Joy in far too glowing terms, and told me that I'd better mend my ways when applying for intensive autism therapy, because you have to paint the bleakest version of truth to have the best chance of getting the therapies you need.

And let me digress here to say what a soul-killing reality this is, on an annual basis or more, for parents of kids with special needs. At least every year for the school district, and sometimes more often than that, we have to focus on what our children can't do, in order to advocate for them and get them the best services we can. How screwed up is that?!

Anyway. The doctor also chastized our Birth-to-Three representative, who was so kind as to accompany us to the appointment, for not recognizing the autism. We all felt rather wrung-out by the end of the appointment. Merry Christmas 2006 to all!

What we hadn't yet put together was that Joy's first big regression had been happening between the time I filled out those forms and the time she got the diagnosis. The most obvious piece of the regression was that she went from about 80 words to about 20. Even that, though, can be hard to track. Think about it -- if you're even keeping track of your kiddo's words for the baby book, you write them down when you start hearing them, and then you consider them mastered. What you don't anticipate is that any utterance of that word might be the last time you hear it for goodness-knows-how-long. When do you realize that you haven't heard that word for a while? In a week? In a month?

Our tracking system got much better when we started a private Yahoo! group for Joy and started trading daily updates with Lynda the daycare lady, wherein I also began describing the Birth-to-Three therapy sessions. In this way we were able to track how Joy came out of the regression over the course of the summer. Then in the fall she started with the school district and those therapists started contributing reports of every session to our lovely online archive, documenting some very nice progress. We were back up over 80 words, starting to do some color matching, stringing chunky beads, beginning to stack Legos....

And then came Christmas 2007. This time we saw it right away, though it probably was only a little more extreme than the first time. It happened over Christmas break. Joy started losing words again, wouldn't tolerate being in the Sunday school classroom, had a behavioral down-spiral. In February 2008 we think she had a virus that made things even worse.

Good things have happened since then, for sure. Joy has begun to lead people by the hand to what she wants. We've established some nice new helper routines. Some words have come back (like ma-ma-ma-ma-ma!) She re-entered the Sunday school classroom in June. She has begun to initiate peek-a-boo games. Much to be happy about!

On the other hand, we haven't seen the level of language recovery that we did the prior summer. And Christmas is coming again in three months...

That's the worst of it, I think. Christmas is coming again, or we may see a regression sooner, or maybe later, or maybe never again. And we never know, from day to day, whether this day will be the last day that she says ma-ma-ma for the next year. We can't take any gain for granted. And that's very, very hard.

The IAN report I linked to earlier was also reporting on its own data collection about regression. Very unscientific, not peer-reviewed, self-selected respondents, self-reported information. However, despite all those caveats, there was this fascinating chart, with the big question on the chart in red:


In other words, maybe someone will start taking notice of late regressions, or multiple regressions? Maybe there's some fodder here for future research, instead of focusing autistic-regression research solely on those big initial regressions, which, devastating as they are, by no means tell the whole regression story.