When Martin was diagnosed, during the time when Adrian and I were learning what autism means, Adrian asked whether I’d ever seen Martin “go backwards,” i.e., lose skills or developments that he once had.

No, I told him. No, I hadn’t seen that. There wasn’t anything that Martin had been able to do and no longer could.

He’d once made eye contact, once been interactive. I didn’t think about that. The changes were so gradual.

That’s good, Adrian replied. He’d read that going backwards was somehow worse than not having progressed in the first instance.

Oh? Well, phew. I haven’t seen Martin go backwards. I guess, on that point, we’re lucky.

Four-and-a-half years change a lot. Today I have a more informed opinion on regressive versus classic (non-regressive) autism. The idea that classic autism is somehow less threatening, wherever Adrian got that idea, must be mainstream. It must come from the school of thought that says autism is untreatable, recovery is not possible, and the best you can do is to teach a child with autism to live with his disorder, and to hope the condition doesn’t become more severe.

A few words about terminology. Regressive autism seems to appear following some insult to the immune system, like a serious illness, a vaccination, or an allergic reaction. In such cases, I believe—and I can’t say often enough that I am a non-scientifically minded lay person treading water in an ocean of evolving knowledge—the child likely has some preexisting immune shortcomings, or a genetic predisposition to these shortcomings, and whatever injury occurs throws the child’s whole system into disarray.

Classic, non-regressive autism is more of a mystery. The classically autistic child, as I generally consider Martin to be, does not acquire skills and them lose them; he simply never meets expected developmental milestones, or at some point stops meeting them. In Martin’s earliest months, we thought he was physiologically advanced. “You should video that,” his first pediatrician said, when I showed her that Martin already could turn over, front-to-back, at 17 days old. “Otherwise no one will believe you.” We marveled at Martin’s extensive vocabulary and his uncanny memory. “That is incredible,” the same pediatrician said at Martin’s 24-month check-up, when she realized that he’d memorized all the characters from the 1995 BBC version of Pride & Prejudice and was repeating their lines (hello, echolalia). But the thousands of words never came together into original sentences. The walking, a bit late at 14 months, never became running or skipping. The interest in videos gave way to obsession with wheels and mechanical parts. Martin stopped meeting expectations.

As it turns out, from everything I understand, and from all the families I’ve met who are on journeys like ours, regressive autism is more readily treatable than classic autism. Take that with a grain of salt, of course; I don’t know any form of autism that is really “readily” treatable, and certainly none that is easily treatable. Still, there appears, at least anecdotally, to be a distinction: Regressed kids, when treated biomedically, recover better.

True confession, and one I’ve made before, on this blog: Sometimes I feel disheartened when I see another family making more rapid progress healing their child than I’m making. What am I missing? I ask. What are they doing that I’m not? Then, where possible, I comfort myself by thinking, well, that boy’s autism was regressive. He was typically developing until age two-and-a-half. That’s why he’s recovering faster than Martin. These thoughts help. A little.

One school of thought holds that all autism is regressive, and the differences arise only based on when the regression occurred. For example, a pregnant mother’s (concurrent or earlier) exposure to certain pollutants or contaminants may crash her fetus’s developing system. In Martin’s case, I believe that the circumstances surrounding his birth—I was pressured to induce labor with Pitocin, leading to an un-planned Caesarian section, and then (healthy) Martin was taken to the NICU against my wishes and pumped with intravenous antibiotics—contributed heavily to the health and neurological problems that later became evidence. The injury, whatever it is, need not result from a lone tipping point, either. Many families report a series of mini-regressions following vaccinations and acetaminophen use.

If all autism is, to whatever extent, regressive, what does it mean that obviously regressive autism, cases in which a family watches language or eye contact or socialization fade rapidly, seems more readily treatable? Does it mean just that it is easier to recover skills if the child had them once? The ability to ride a bicycle or operate a manual automotive transmission, according to conventional wisdom, never disappears once acquired. Research shows that autism does not cause permanent brain injury, and that the brain can return to healthy functioning once the neural misfires and inflammation are eliminated. Perhaps a child who once spoke well can return to full language function faster than a child who never acquired spoken words. Playing catch-up for time lost is easier than starting from scratch.

Maybe all autism is regressive. Maybe there really is a difference between regressive and classic autism. Maybe regressive autism is more readily treatable. Maybe the important factor is the date of immune injury. Maybe the important factor is the extent of genetic predisposition. Maybe this, maybe that. For scientists, these questions matter. For a parent, fuhgeddaboudit. What does it matter, anyway? Keep chipping away at this monster, this autism, however long it takes.