here have been claims over the years that autism is caused by maternal virus infection
(such as CMV or influenza), maternal immune response, parental schizoaffective disorders,
infertility treatments, inter-pregnancy interval, alcohol and vitamin intake, cigarette
smoking (of the parent), SSRI usage during pregnancy, proximity to highways, or chemicals
such as PCBs, BPA, pesticides, diesel fumes, styrene, mercury, nickel, beryllium,
and air pollution. Some of these are environmental and some are heritable.
Last month a big study (Taylor et al., 2020) in Sweden tried to resolve the question. The authors studied records from 37,957 pairs of twins whose parents volunteered to register them in one of two Swedish twin registries. The subjects were binned into four-year time intervals between 1982 and 2008. The authors argued that, contrary to prevailing belief, autism rates are rising and it was important to identify any environmental factors.
Unfortunately, the database showed that autism rates recorded by the doctors of the time, which the authors call “Clinical ASD,” were decreasing. So the authors assumed that many individuals must not have received a correct diagnosis. They called 37,000 people on the phone—some of whom were in the database and some not—and gave them a yes-or-no ten-question survey, which they say is highly accurate. And lo and behold, they found that ASD actually increased by 60% between 1982 and 2008.
Next they ran statistical tests using the standard trick of comparing identical twins with non-identical twins and found that the heritability of autism in their cohorts ranged from 0.88 to 0.93, which means that any environmental effect would have been nearly unobservable.
After some more statistical manipulations, they concluded that the importance of environmental factors didn't increase over time, so the environment doesn't explain the increase in prevalence ASD. Which is a bit odd because it would have to mean that hereditary factors are increasing, which is impossible to explain unless people in Sweden are spontaneously mutating.
We've all had experiments that give you the opposite of whatever assumptions you start with, but I've never seen such a determined attempt to hang on to an impossible hypothesis after nature pounds you over the head again and again trying to tell you that it's false.
There is no epidemic of autism
The evidence that ASD is increasing is far too weak to support claims that there is anything like an increase, let alone an “epidemic” as some claim.
For ASD, any change over time is likely to be the result of differences in diagnosis rather than changes in incidence. The diagnostic criteria for ASD in the DSM, which reflects international standards, were changed repeatedly between DSM-III, DSM-IV, and DSM-5. Asperger's or “high-functioning” autism, for example, is no longer considered a real disorder. French et al.[1] point out that in earlier years some patients who would be diagnosed as autistic were classified as having other disorders such as “pervasive developmental disorders not otherwise specified” or PDD-NOS. No points for guessing where the term PDD-NOS comes from.
Maybe the DSM dropped Asperger's because too many people were diagnosing themselves with it to make themselves feel special. There's also a growing feeling that ASD itself may be an autoimmune disorder. That doesn't rule out a heritable cause, and that's the point: we don't have a clue. Trying to determine whether the incidence of ASD is changing when the diagnostic criteria change from year to year, the cause of ASD is unknown, and even whether it is a single entity is a fool's errand.
Almost all articles on this topic pointedly omit vaccines. The single paper that claimed MMR vaccine as a cause has become a huge embarrassment for the field, partly because it spawned a huge industry in conspiracy theories. And that's what I really want to talk about.
Acetaminophen
It always raises suspicion in science when a single cause explains everything. But when something is poorly understood it becomes a crank magnet, which is to say people who are obsessed with a single cause join in and loudly proclaim that their idea is the right one. Autism is one. Glyphosate is another. And bees. Oh, the poor bees. And then there's acetaminophen.
Acetaminophen is the most widely used drug in the world. Its mechanism of action is reasonably well understood, and there's no question that an overdose will cause acute liver failure. So you might be surprised that it has become a crank magnet, but it has. It's been claimed to cause allergic rhinitis, asthma, renal impairment, ADHD, impaired neurodevelopment, autism, birth defects, and Parkinson's disease. Don't get me wrong: these are all legitimate hypotheses, well worth considering, but that point somehow keeps getting lost in how the findings are reported.
Glyphosate
Glyphosate is probably the biggest crank magnet around at the moment. There are claims that glyphosate causes ADHD, leaky gut syndrome, celiac disease, neurological problems, breast cancer, gluten intolerance, liver damage, antibiotic resistance, birth defects, kidney damage, Alzheimer's disease, and Parkinson's disease. Glyphosate is accused of harming worms, bees, wasps, snails, fish, amphibians, birds, rats, bacteria, pigs, and of course humans.
So it was only a matter of time before somebody claimed that glyphosate causes autism. One paper went for a three-fer: autism initiated by acetaminophen, they claimed, is aggravated by the oral antibiotic amoxicillin and glyphosate.
One paper in Environmental Pollution claims the supposed toxic effects aren't due to glyphosate but polyethoxylated tallow amine (POEA) and MON 0818, detergents used in the glyphosate formulation. Others are claiming that it's a metabolite of glyphosate called aminomethylphosphonic acid produced by microbial degradation, which destroys up to 70% of any glyphosate in the soil within four days.
Again, these are all hypotheses that are worthy of consideration, but they are miles away from being a reason to ban glyphosate. Doing so would abrogate any potential insights the research might provide. You never know: the researchers might find a cure for the common cold.
Conspiracy theories are a luxury good
When an association is shown to be false, protagonists split into two groups. One side tries to save the theory by proposing more and more complicated biological mechanisms. The other attacks the integrity of the scientists who showed that the association was false. Often the topic becomes politicized and no resolution is possible.
There are also two markets for conspiracy theories. One market is found in developing countries, where people are unable to improve their economic status (largely because their governments are corrupt). In this market, the conspiracy allows them to point the blame onto an external target, which relieves pressure on the corrupt government and may even result in increased foreign aid.
The other market is wealthy urbanites in advanced countries for whom removal of any possible risk is worth any price. They provide a lucrative market for such ideas, and scientists will supply the ideas provided they can be re-framed as a problem to be solved. And indeed, there are many scientific papers where the data point to undetectable levels of risk, but the authors conclude that the risk is high and is ominously increasing. The Taylor et al. autism study is a perfect example.
The conspiracy theory matryoshka
The market for conspiracy theories is so big that some people are inventing conspiracy theories within conspiracy theories. They point, for example, to Russian hackers who are inventing them because, like all foreigners, they love President Trump so much.
Well, who doesn't? But many conspiracy theories actually come from people who read something in the scientific literature and who trust the results. In this way science creates a perpetual motion machine: we publish conclusions about possible dangers that go well beyond the actual data—often even contradicting them. This creates a sort of Gemino curse: for every spurious finding that gets published, ten other studies are needed to refute it. I suspect this may not be entirely accidental.
One reason people overstate their conclusions is to justify to the editors that their manuscript is worthy of publication. If they wrote “We recognize that this is not a problem and hardly worth studying, and our whole idea was actually kind of stupid” they wouldn't get published. But the main purpose, the one that's implicit in all scientific papers, is that the results must justify continuation of government funding.
When a false result is published, such as the vaccine result in Wakefield's infamous paper, we're surprised when people invent conspiracy theories when the hypothesized problem is never solved. And that gives us a new problem to try to solve. Indeed, there's a growing literature about conspiracy theories, their harmful effects, and many suggestions on how to eliminate them, most of which involve highly sophisticated forms of censorship.
That's a classic case of a perpetual motion machine. Censorship turns a conspiracy theory into a genuine conspiracy. Censoring a conspiracy theory would be throwing gasoline onto a fire.
If anybody ever figures out that all the nutty stuff about 5G and autism and glyphosate wasn't pulled out of thin air but from more or less reputable scientific journals, we'll have a brand new problem to find a solution for. And if we can prove that the problem is bad enough, we'll be able to say that we need massive funding to go along with it. As they say in industry, it's a win-win.
We can't blame scientists. They are no longer seekers of truth. Their job is to beg the government for money. That's what our system has devolved into, and we're starting to see the effects.
1. French LR, Bertone A, Hyde KL, Fombonne E (2013). Epidemiology of autism spectrum disorders. In The Neuroscience of Autism Spectrum Disorders, Elsevier, http://dx.doi.org/10.1016/B978-0-12-391924-3.00014-4.
2. Lysall K, Schmidt RJ, Hertz-Picciotto I (2013). The environment in autism spectrum disorders. In The Neuroscience of Autism Spectrum Disorders, Elsevier, http://dx.doi.org/10.1016/B978-0-12-391924-3.00014-4.
jun 01 2020, 7:49 am. revised and edited 3:55 pm
