I was an employee of the Fred Hutchinson Cancer Research Center when we first began to suspect that our son had ASD. My job at the time was to write programs to analyze data collected in various drug protocols. I didn’t need to know anything about clinical trials to do my job, but after a few years I picked up quite a bit via osmosis. I could speak knowledgeably about phases and informed consent, blinding and double-blinding, and the critical distinction between causation and correlation.
My wife and I also come from scientific backgrounds: she has a degree in chemistry (and works as a botanist), I have a degree in environmental science (and work in the field of software engineering). Not to brag, but we know our way around a null hypothesis.
So when we were approached by the University of Washington Autism Clinic, and asked if we wanted to participate in the Toddler Assessment Project, we enthusiastically agreed. Not only would we get a diagnosis out of the deal, but our participation would help the doctors identify the early signs of autism in other children.
Our son was given a battery of standardized tests and exams before an official assessment was made. The rigor of the process and the staff’s commitment to objective, impartial conclusions were a great comfort to us. They never hesitated to explain to us exactly what they were doing and why; better yet, they were very open about what they didn’t know. I sat by the MRI operator as he gave me a guided tour of my son’s brain, freely admitting that they didn’t know specifically what they were looking for, but were gathering data to advance the science and aid in future diagnoses.
All of this was in stark contrast to our experiences after the study was concluded. When we began seeking treatment options for our son, we were inundated with suggestions from friends and family and casual acquaintances, ranging from the benign (vitamins) to the severe (chelation therapy). Each recommendation came with an assurance that the treatment had worked wonders for someone, be it the child of the speaker, or a son of a friend, or a kid they had read about on the internet.
These suggestions were well-meaning, and I sincerely appreciated the motives of those who made them, but they also gave me fits. Anecdotal evidence is anathema to disciples of the scientific method such as I, but in the fledgling field of autism treatment it was all anyone had to offer. And everything was proposed as possible treatments: dietary modification, craniosacral therapy, acupuncture, magnets, auditory integration training, horseback riding, supplements, pharmaceuticals, homeopathic tinctures. Pretty much every form of medicine, Western or Eastern or complementary or alternative, has at one time or another been championed as aiding, or even “curing”, those with autism.
But who really knows? In a clinical trial, you have an intervention group and a control group; you administer the drug or therapy to the first, and leave the latter unaffected. Then, after some period of time, you compare the results. Did the intervention group fare better than the control group, using whatever rigorously defined definition of “better” you’ve adopted? If so, then you had evidence—not proof, never proof—that the intervention was beneficial.
But when you introduce a treatment method to a three-year-old with autism, it is extraordinarily difficult to evaluate its efficacy. What if his communication skills suddenly increase? Isn’t that evidence? You can interpret it as such (and, if you are hoping that the treatment works, confirmation bias will ensure that you do), but maybe the jump in verbal skills was just a normal advance, something that would have happened even had you not administered anything.
Worse, from a clinical point of view, parents will often try several things at once, introducing three new supplements and altering the diet and enrolling the child in art therapy all in the same week, thereby ensuring that they will have no way to determine which of the treatments is responsible for which of the outcomes, assuming that the outcomes are a result of any of the intervention and not just a manifestation of the child’s natural development.
It is maddening. We tried a lot of things, and I’m not sure we can point to any of them and say with conviction that they “helped”, whatever that even means. Our son is on a gluten-free, casein-free diet, for instance, which we adopted shortly after his diagnosis because it is one of the few “alternative” treatments that actually has some scientific support (albeit a single study, with largely inconclusive findings). We keep him on this diet because GFCF foods have become so ubiquitous in Seattle that it’s no more difficult to stick to this diet than any other (protip: Udi’s), and also because … *shrug*, I dunno. Because maybe it’s helping? Because maybe he would have been “worse” had we not kept him on the diet?
That’s what you wind up doing: using an imaginary version of your own child, in a universe where you didn’t administer some treatment, as your control group. You can’t compare him to his peers, because every child with autism develops uniquely, thwarting your attempts to draw conclusions from differing parenting techniques.
Think of a baseball player with an elaborate ritual before every at-bat. He once touched the brim of his ballcap before hitting a home run, so now he always touches his ballcap while warming up; he once hit a triple after tapping his shoe with the bat three times, so shoe tapping goes into the routine. Now he can’t swing until he’s gone through a full minute of seemingly random actions and tics, some that may actually help, others that are pure superstition. He’s unconvinced that any one of them is effective, but he’d be a fool to tempt fate by stopping.
That’s what you become, as a parent of an autistic child. You try various things, latch onto those that preceded a positive development, and stick with them in the hope that you are getting some return for your investment. Empiricism is great on paper, but in the absence of evidence you go with your gut.