New Research on Autism

There are lots of good reasons to try and follow a healthier diet–you lose weight, you feel good, but the main reason–to live a longer, happy, productive life.  Sounds good, right?   And though it may sound deceptively easy, the devil is in the details.   Welcome to the Nutrition Facts podcast.  I’m your host, Dr. Michael Greger. 

On today’s show, we take a look at new research on autism, the neurodevelopmental disorder characterized by impaired social interaction, impaired verbal and non-verbal communication, and restricted and repetitive behavior.

Before speculating about the reason for the “explosive” increase in autism, one has to make sure the explosion is real.

In 1943, a famous paper was published in which a child psychiatrist at Hopkins described a series of children with so-called “fascinating peculiarities.” He thought these characteristics formed “a unique ‘syndrome’” he called autism, which seemed rare at the time. But, who knew how many were out there undiagnosed?

What causes it? Autism is currently “considered…a multi-factorial disorder resulting from [both] genetic and non-genetic risk factors…” Yes, it can run in families, but “genetic factors [may] account for only 10 to 20% of…[autism] cases.” This is based in part on the fact that you can have identical twins, with identical DNA—the same genes, and one identical twin may have autism, and the other not. So, “While genetic susceptibility may be a key contributor to [these autism spectrum disorders], it may…just ‘load the gun’ so to speak, with prenatal, perinatal, and…postnatal environmental exposures [some kinds of exposures during, around, or after pregnancy] being the events that ‘pull the trigger’ and may [actually] give rise to [the disease].”

This is good news. I mean, the larger role these non-genetic factors play in causing autism, the more “modifiable [the] risk factors” may be, potentially “open[ing] up avenues for the primary prevention of…autism” in the first place.

Since autism as a medical condition was first described, the prevalence of autism has apparently exploded from like “1 in 5000 individuals to 1 in 68…” now—more than 1% of the population. That would be like a 7,000% increase. And, indeed, you’ll see exponential increase in the prevalence of autism, from like no diagnosed cases in the early 1900s, to the prevalence shooting through the roof in the ‘80s and ’90s.  And so, that immediately gets you thinking: what happened around that time that could account for the explosion? But, wait a second. Of course, there were no diagnosed cases in the early 1900s; it didn’t even have a name until 1943. As Kanner said in the original paper, there’s probably more cases out there, but they just hadn’t been looking. The prevalence of autism diagnoses depends on what diagnostic criteria you’re using, and whether you’re out there looking for it or not.

“Put another way, historical prevalence estimates for autism…[and its rarity] might well have been underestimates of the true prevalence” back then; may have just been missing lots of cases. The “[i]ncreased recognition among doctors” and society at large, the “broadening of the diagnostic concept over time” and different studies using different criteria may account for [much] of the apparent increase in prevalence, although this [can’t] be quantified.” So, before we start speculating about the reason for the explosive increase, maybe we should first make sure the explosion is real. The bottom line is that, while we may never really know what the prevalence was a half century ago, we do have decent data over the last few decades that really does point to a considerable increase in the true prevalence.

So yeah, maybe there wasn’t actually a “22…-fold increase” in autism in the ‘80s and ‘90s; maybe there was actually only an “8…-fold increase.” So yeah, we may quibble whether the increase was 800%, or closer to 2,000%, but the bottom line seems to be autism rates are really increasing. And so, the question legitimately then turns to: Why?

Our next story takes a close look at the effects of smoking, pesticides, vaccine mercury, and air pollution on autistic spectrum disorder risk.

“The prevalence of [autism] has increased dramatically in the US…” We’re not exactly sure how prevalent it used to be, but these days, about “1 in 68” kids born in the United States will have it. Despite a massive influx of research funding, we still don’t know: what’s the cause? This “sharp increase in prevalence remains unresolved.” Yes, “[c]hanges in awareness [and diagnosis] can account for some of the increase,” but it also might be something we’re exposed to in the environment.

For example, all the new chemicals we’re now exposed to. Out of 80,000 agents the chemical industries now put out, we have evidence that at least a thousand may have neurotoxic properties, yet “only a small fraction have been studied in humans during critical windows of development.” “The current chemical risk assessment approach is typically based on the toxicity caused by a single chemical…without acknowledging” the effects of different chemical mixtures. There are hundreds of chemicals currently “allowed in food that may have potential harmful effects on the developing brain.” “Each individual chemical may or may not have a harmful effect [on its own], but we know next to nothing about their cumulative biological effects on the brain.”

“If we really want to protect children’s brains from chemicals…, we first need to recognize and deal with the massive dearth of basic hazard and exposure information [about any] risks posed by environmental chemicals.” A lot of the data we do have is from “preclinical studies”—meaning like test tubes and lab animals. Well, yeah, it’s not like you can expose people on purpose. That’s where epidemiological research can step in: studying exposures within populations. That’s how we found out about other well-known hazards, like benzene and asbestos. You can’t ethically force people to smoke for a study, but you can note how many people exposed to cigarette smoke, day in day out, get lung cancer, compared to nonsmokers.

Industries spewing poison like to patronize the public that we shouldn’t worry our pretty little heads, but what does the science show? What does the epidemiological evidence have to say about exposures to environmental chemicals and autism?

For example, there’s a bunch of studies on the effects of smoking during pregnancy on autism rates. I’ve talked about these so-called forest plots before, where values greater than one suggest at least a tendency towards an increased risk, whereas the values less than one suggest more of a protective association. And should the whiskers cross the line, then that means that particular finding did not reach statistical significance. This allows you, with a quick glance, to interpret a huge amount of data from multiple studies.

Does it look like smoking during pregnancy has any effect on autism rates? No; about half skewed to one side, half to the other, and hardly any of the results reached statistical significance—meaning they may have just been largely chance findings.

Same thing with thimerosal exposure, the mercury-containing preservative used in vaccines. Yeah, one study found significantly increased risk of autism, but three other studies found that those with more vaccine mercury exposure had lower the risk of autism. But mostly, they just showed no relationship either way—not that it matters much at this point, since thimerosal has been removed from most vaccines.

But check out the air pollution data. This is exposure to car exhaust and diesel fumes during pregnancy and infancy. But, correlation doesn’t necessarily mean causation. This could be a true cause-and-effect relationship, or maybe it’s just that poor people live in the inner cities next to polluted highways, or the traffic noise is causing stress, or some other confounding factor.

The other suspicious link they found was “pesticide exposure.” Lots of environmental toxicants have been implicated in autism: plastics, chemicals, PCBs, solvents, toxic waste sites, and heavy metals. But the “strongest evidence [has been] found for air pollutants and pesticides.”

Yeah, “[c]linicians [can] advise pregnant women and parents [to try to avoid exposures to] harmful substances in their environment.” But, instead of just advising patients one-by-one, a “more powerful” strategy would be for us clinicians to band together and take a “leadership role in advocating for a healthy environment” and society-wide actions to help everyone.

What happens when children with autism on gluten- and casein-free diets are covertly challenged with wheat and dairy? Here’s the story.

The first randomized, controlled study of a gluten-free, casein-free diet for autism found that parents in the diet group reported their kids did “significantly better.” But, that could just be the placebo effect—where parents attribute changes to the diet, particularly since they’re putting so much work into it that they may be “biased towards seeing evidence of success” that may not actually be there.

Okay; so, what if you don’t just rely on parental report? What if you do a blinded study of a gluten- and casein-free diet? The parents know what the kids are eating, but you don’t just ask the parents how the kids are doing; you have investigators objectively assess all the kids, without knowing who was in which group—the diet group or the control group. And, they found “a significant beneficial…effect at 8, 12, and 24 months…on core autistic…behaviors” with a gluten- and casein-free diet. And, this was one of the largest such studies ever done—starting out with 73 children. But, they didn’t end up with 73, since about a fifth dropped out—mostly from the diet group.

“If a family didn’t feel their child was making strides on the diet, they may have been more likely to drop out of the study, thereby skewing the analysis toward those” in which the diet appeared to work better. And so, the remarkable results they got in terms of improved social interaction and fewer ADHD-type symptoms may have ended up exaggerating the effects of the diet, since the kids for which it didn’t help may have been, gotten kind of, disproportionately weeded out. Also, because the parents were very much aware whether they were in the diet group versus the control group—because they were the ones cooking the meals—they may have changed their own behavior towards their children.

Like in that famous sugar study where mothers were lied to, and told their children had just received a whopping dose of sugar (even though they hadn’t). Not only did the mothers rate their children “as significantly more hyperactive”; they inadvertently changed their own behavior. They were videotaped, and the mothers who falsely believed their kids just got a load of sugar “exercised more control” and were more critical. So, their expectation of an effect may have actually ended up having an actual effect in changing their children’s behavior. So, in these autism studies, here the parents are upending their family’s diet, hoping, expecting that their child will get better, and maybe unconsciously treating them differently—such that they end up behaving differently when assessed later by the blinded investigators.

That’s why we need double-blind studies, where no one knows who’s in the diet group, who’s in the control group—not the parents, not the kids. So, why didn’t the researchers do it? Why didn’t they secretly sneak some gluten or casein into their diets to see if they got worse again? Same excuse as before; it wouldn’t be “ethical.” But, that’s pre-deciding the outcome; that’s like circular logic. We can’t test to see if it really works, because it may really work—but we can’t test that.

Finally, however, researchers at the University of Florida broke through the impasse by performing a double-blind study. Now, this is not an easy thing to do. I mean, you have to provide “all meals and snacks,” so that the families remain clueless as to whether they were randomized into the gluten- and casein-free diet group, or were actually in the control group getting the same foods, but with gluten and casein slipped in. Then, after six weeks, they switched the foods, so the gluten-free, casein-free group started getting wheat and dairy, and the wheat and dairy control group were covertly switched to gluten- and casein-free.

Before they broke the codes to see who was in which group, the “parents were asked [as to] whether they thought their child was on the [special] diet during the first or second 6 weeks.” And: “Five [got it right], two had ‘‘no idea’’, and six [got it wrong].” In other words, no better than a flip-of-a-coin chance. About half pretty much thought their kids got better on the casein and gluten; so, “no statistically significant findings”—though several parents did report an “improvement…in their children,” claiming “marked improvements in…language, decreased hyperactivity and decreased tantrums”—so much so that a bunch of the parents “decided to keep the children on the…diet,” even though the researchers just told them that it didn’t work.

Look; you don’t want to miss anything, though. Some of the parents claimed significant improvement. So, what the researchers did was to go back and examine the videotapes they made of the kids before and after and showed them to blinded examiners. Did their language really get better? Apparently not. The videotapes showed no improvement, so again, did “not support the efficacy of [a gluten-free, casein-free diet]” for improving some of the core symptoms of autism—or, at least, within six weeks’ time. The non-double-blind studies that showed an effect had kids on the diet for a year or two. And so, the failure to find an effect in the double-blind study “should not necessarily be interpreted to mean that [diet] does not work,” given “the relatively shorter duration of [dietary] intervention…”

The same issue cropped up years later in a 2014 study in Texas. Simple study design: put everyone on a gluten-free, casein-free diet, and then randomize parents to get baggies each week filled with either gluten-free, casein-free brown rice flour—thereby sticking to the diet—or, an identical-looking powder with gluten and milk mixed in. So, no one knew until the end who really remained gluten- and casein-free, and…no, meaningful changes were found in either diet group. Okay, but this study only lasted four weeks, and diet proponents suggest it may take months off gluten and casein to properly assess a response.

To see any graphs, charts, graphics, images or studies mentioned here, please go to the Nutrition Facts podcast landing page. There, you’ll find all the detailed information you need, plus links to all of the sources we cite for each of these topics.

Be sure to check out my new How Not to Die Cookbook. It’s beautifully designed, with more than 120 recipes for delicious, plant-based meals, snacks, and beverages. All the proceeds from the sales of all my books all go to charity. I just want you to be healthier. 

NutritionFacts.org is a nonprofit, science-based public service, where you can sign up for free daily updates on the latest in nutrition research via bite-sized videos and articles.

Everything on the website is free. There’s no ads, no corporate sponsorship.  It’s strictly non-commercial.  I’m not selling anything. I just put it up as a public service, as a labor of love—as a tribute to my grandmother – whose own life was saved with evidence-based nutrition.

Thanks for listening to Nutrition Facts.  I’m Dr. Michael Greger.

This is just an approximation of the audio content, contributed by Allyson Burnett.