- The bastards are telling us that our kids will never get better!
- Loss of label in autism is so miraculous that some innovative cure must have made little Timmi recover.
- My child would not have skills X, Y, and Z if it weren't for [insert woo].
- My 2 year-old child will never be able to do X.
- My child will necessarily end up in an institution.
When we deal with flawed premises of this nature, rational discussion obviously becomes very difficult. That's why I want to make this post a sort of all-purpose summary of what the science tells us in regards to developmental progress, adult outcomes, loss of label in childhood, speech, institutionalization, and shifts in parental beliefs. I will discuss factors that seem to be predictive of outcome, but don't expect any magical formulas or guarantees from me. I can only speak of possibilities and I will only report what the data tells me. I will not sugar-coat what the research says.
As always, if you find that I failed to include some relevant data, or if something I said is inaccurate, feel free to let me know in the comments section.
Developmental ProgressRecently there was some publicity sorrounding a study by Paul Shattuck which found that autistics "get better" in average (source). This type of finding is not new, however.
Even early papers by Kanner show that autistic children tend to make substantial developmental progress as time goes by. There are several other published findings as well.
Findings from Harris et al. (1991), for example, "support the notion that young children with autism can make very significant developmental gains."
Tager-Flusberg et al. (1990) found that language acquisition in most autistic children followed the same developmental path as that found in children with Down Syndrome, and that of normal children as reported in the literature.
Harris et al. (1990) found that all autistic children in the study made normative progress in their rate of language development.
Piven et al. (1996) reported "significant change over time in autistic behaviors, generally in the direction of improvement" in high-IQ autistic adolescents.
Charman et al. (2004) found that autistic children had "encouraging developmental progress" over a 1-year period, although no change in "symptom severity as measured by the Social Communication Questionnaire."
Of course, there is such a thing as "regression" and there are developmental plateaus, but these are the exception rather than the norm, and they will not necessarily persist indefinitely.
Adult OutcomesThe first adult outcomes reported in the literature can be found in Kanner (1971) and Kanner et al. (1972). The 1972 paper details the case histories of 11 (actually 13) of 96 autistic adults Kanner considered "success stories." The paper contains some discussion of predictive factors for a successful outcome, such as the presence of some expressive language before the age of 5, and effort spent to "compensate for the lack of inherent sociability" when the individuals were in their middle to late teens.
The only environmental factor Kanner found significant was a total lack of institutionalization. I see this as a key factor, and it would be difficult to do it justice with a risk ratio. Lack of institutionalization was practically a requirement for a successful outcome. On the other hand, from Kanner's 1971 paper it would seem that institutionalization from an early age practically ensured a "dismal" outcome.
There have been other adult outcome studies since then, with varying results. For example, Howlin et al. (2004) found that 12% of adults had "very good outcome" and another 10% "good outcome", with the majority (58%) having "poor" or "very poor" outcome. Billstedt et al. (2005) found that autistics diagnosed in 60s, 70s and 80s had worse outcome than previously reported, with only 3% achieving independence, and 78% having a "poor outcome."
A recent adult outcome study reports that half of the autistic adults had a good or fair outcome, and 46% had poor outcomes (Eaves & Ho, 2007). The authors indicate that "current young people had more opportunities" and thus better outcomes were expected. I would venture a guess that one of those opportunities was the opportunity not to be institutionalized.
Some factors in outcome, such as baseline IQ or symptom severity, are usually discussed in these types of outcome studies.
Szatmari et al. (1989) studied the outcome of "non-retarded" autistic adults, and found that 4 of 16 (25%), a "surprising number", had a very good outcome and could be considered "recovered." An interesting finding from this study was that symptom severity was not predictive of outcome.
Outcome studies generally use various standard measures to determine if an outcome is good or poor. Ruble & Dalrymple (1996) questions the traditional view of outcome and suggests that some good outcomes are sometimes "invisible" to traditional measures. It also found that the parents of autistic adults who were doing well had pushed for integration and mainstreaming.
You might have heard neurodiversity advocates promote the notion that "presuming competence" is beneficial. There is actually some evidentiary support for this if you look at Donohue et al. (2000), which is not autism specific, but is referred by Ivey (2007), a survey of teacher expectations of outcome in autism. The following is quoted from its summary.
Research shows that when the significant people in a child’s life do not believe that he or she has capability to achieve an outcome, it is unlikely that the outcome will be realized. For example, if a teacher feels that a student has
social obstacles, then their perceptions may increase the undesirable behavior and the student may see him or herself in that light (Donohue, Weinstein, Cowan, & Cowan, 2000).
The environmental factor that seems to be of interest to medical researchers is, of course, treatment. Unfortunately, there are no adult outcome studies that could tell us if modern treatments, particulary early intensive behavioral intervention (EIBI), are predictive of a good outcome. The evidence is rather ambiguous. To take an example, you have Howlin (1997) which tells us that "appropriately structured programmes for education and management in the early years can play a significant role in enhancing functioning in later life." But later the same author, in Howlin (2003), states that "despite claims to the contrary, there is little evidence that very early, intensive interventions can significantly alter the long-term course of the disorder." It would be of interest to know, for example, how the experimental group of Lovaas (1987) is doing, now in their 20s and 30s (even if the group was not representative) but unfortunately no one has taken the time to document these adult outcomes in the peer-reviewed literature thus far.
Loss of Label in ChildhoodIt seems obvious that the earlier the diagnosis, the harder it is to predict if the diagnosis will hold some years later. There are some studies that support this view, such as Turner et al. (2006) which found that only 88% of children with autism or PDD-NOS diagnoses at age 2 were found to be in the spectrum at age 9. See also Sutera et al. (2006) and Charman et al. (2005).
While loss of label in childhood is rare, it's obviously not as miraculous as some claim. A gap in the science is that nothing is known about the adult outcomes of autistic children who lose their label at a young age.
Speech DevelopmentTurner et al. (2006) tells us that 88% of children diagnosed with autism or PDD-NOS at age 2 could demonstrate some functional language at age 9. However, only 32% were able to engage in a conversation.
Kobayashi & Murata (1998) found that 48% of a sample of 187 autistic adults had good or very good speech.
Speech delay has been found to be non-significant in autistic children with normal intelligence, i.e. their outcome appears to be no different to those with a diagnosis of Asperger's (Calhoun, 2001).
InstitutionalizationIt is important to realize that institutionalization of developmentally disabled persons has been in decline in recent decades, probably in large part due to the deinstitutionalization movement.
Krauss et al. (2005) reports that 63% of autistic adults did not live at home with their parents. Of those, the majority (73%) lived in a community residential programme. About 17% had a semi-independent living arrangement. Only 3% lived in an institutional setting.
This is consistent with the latest data from California DDS, where we find that at least 66% of adults in the system live at home or independently, and no more than 4% live in institutional care facilities (source).
Currently about 10 in 10,000 persons in California are registered with CDDS and do not live at home or independently. There is a slight downward trend in this proportion. I would estimate that autistic children diagnosed today in California have a relatively small chance (5-10%) of not living at home or independently as adults, due to the trend in institutionalization and the broadening of the spectrum. To put it in perspective, I think they have better odds of "recovering" than of being institutionalized.