The issue is not just one of unproven effectiveness, but also one of unknown adverse effects. Even some treatments whose effectiveness may be backed up by placebo-controlled double-blind studies (B6/Magnesium, L-carnosine, Risperidone, etc.) may have short- or long-term adverse effects which are not well understood. The case of Ritalin should be taken into account as an example: There are now reasons to believe it may have cardiovascular adverse effects, psychiatric adverse effects, and is possibly a carcinogenic.
Certain treatments are just believed to be effective and touted as "evidence based" when in fact the evidence is somewhat questionable. This is the case of ABA, a very popular treatment whose effectiveness is taken for granted. But the initial study by Lovaas did not use a randomized control group, and evaluators were not blind to treatment. Follow-up studies on the same group of children of course suffer from the same methodological flaws introduced early on and new results may have been affected by parent perceptions following the initial treatment. The Lovaas findings have been criticized on methodological grounds on several occasions [ref][ref][ref]. If similar evidence were to be presented to support, say, the GFCF diet, it would be immediately qualified as nearly useless.
Michelle Dawson has, as everyone knows, criticized ABA on ethical grounds [ref]. Other adult autistics have expressed concerns (I'm being gentle) about ABA, such as the treatment of children as if they were trained seals [ref][ref]. Aversives are often criticized, but the usual response is that they are no longer used. (The early Lovaas study did rely on aversives, BTW). Some cases of PTSD have been reported following treatment [ref][ref]. One of my own concerns is that effects on late adolescence and adulthood are unknown.
I do believe there's probably an ideal educational approach that will work great for autistic children, a different approach to that used on NT children, but it's not at all clear ABA is that approach.
Parents often do indicate they see some improvement in their children while they undergo ABA. But you will probably find as many parents touting the GFCF diet or chelation therapy, both unproven treatment methods.
It is necessary to recognize that autistic children developing is not surprising. They just don't do it in the expected time frame. Autism is not a degenerative disease like Parkinson's or Alzheimer's, which ultimately end up killing the affected individual. When a neurotypical child learns to walk, do you believe the child "recovered" from a disability?
To put everything in perspective, I think it's helpful to look at what some placebo-controlled studies have found in relation to the placebo effect in autism.
- Sandler AD, et al. Lack of Benefit of a Single Dose of Synthetic Human Secretin in the Treatment of Autism and Pervasive Developmental Disorder (1999).
This study used standard behavioral measures, including the Autism Behavior Checklist, to evaluate children before and after administration of Secretin or Placebo. The initial score of the Placebo group was 63.2 (higher score means "more autistic"). After only 4 weeks, the score on the Placebo group was reduced by 17.8 points (compared to 8.9 in the Secretin group).
But what's most surprising, 69% of the parents of the children in this study said they remained interested in Secretin as a treatment for their children. I should emphasize that a large number of controlled studies have shown that Secretin is about as beneficial in the treatment of autism as drinking bottled water.
- Coplan J, et al. Children with autistic spectrum disorders. II: parents are unable to distinguish secretin from placebo under double-blind conditions. (2003)
This study demonstrated that parents were unable to distinguish the effects of Placebo from those of Secretin, using a swapping technique. But the surprising thing is that 77% of families reported the treatment was beneficial, only 6% saw no improvement, and 6% offered no guess.
- Adams J, et al. Vitamin, mineral supplements benefit people with autism.
This online paper is in preparation of a study that looks at the effectiveness of a multivitamin supplement compared to a Placebo. Parents are asked to rate their children's improvement after 3 months as follows: On a scale of 1-7, 1=much worse; 2=worse; 3=slightly worse; 4=same; 5=slightly better; 6=better; 7=much better.
On the "Sociability" category, parents reported scores of 5.1 for Placebo and 5.3 for the supplement. On the "Expressive language" scale it was 5.6 for Placebo and 5.9 for the supplement. That is, from "slightly better" to "better" on both measures, and not much difference with the supplement. (Side note: The study does show some statistically significant improvements on other measures, but it's not clear if the supplement only helps children who are "fussy eaters" - as always, caution is warranted given that vitamins are drugs).
I've come up with the following possible explanations for these results:
A) A regular placebo effect is at play, i.e. children realize they are undergoing treatment and improve as a result.
B) There's a placebo effect on the parents, i.e. they become optimistic and this positively reflects on the outcome of their children.
C) Parents and researchers are biased in their final evaluation, as they expect a positive outcome following treatment.
D) Autistic children, like any other children, are constantly developing and changing.
I think we can all agree that explanation A is impossible and may be discarded. Explanation B seems a bit far-fetched, but if true, it would bode well for the notion that optimism and enthusiasm are beneficial to children (something doctors and researchers are not very helpful with). Explanation C is quite plausible. Nevertheless, note that the Sandler (1999) study used a less subjective evaluation method carried out by researchers, not parents. And consider that researchers are notoriously pessimistic about autism and its outcome. Furthermore, in the Adams study, parents rate certain physiological outcomes (e.g. intestinal problems) as "no change" on Placebo. This is not surprising, but shows that parents might not be too biased on their expectations. This leaves explanation D as the most likely one: Autistic children do develop on their own, except parents usually only notice when they expect it.