(*Update on 11/19/13: Friend and science reporter Brandon Keim takes my critics to task for sloppy journalism. [And me, too, for not sufficiently caveating.] I repeatedly urged him not to get involved, but he wouldn’t be deterred. It’s an interesting read and an interesting take. But I have to ask: Do I really not have a big enough science writer gang? Does that really matter? )
Some of my critics are raising questions that I feel compelled to address.
1) Why is the NYT promoting my book?
It’s not. I’m a science writer. I pitch stories like other freelancers. These ideas are accepted on perceived merit. In the Sunday Review where my latest piece appeared, as far as I can tell, everyone who’s not a staffer is identified either by affiliation and/or what books they’ve written. Same in the NYT magazine. Same in other sections of the paper where freelancers write. So the argument that the NYT is actively promoting my book is — well, I only wish. Really, though, I pitch stories that may or may not be accepted like everyone else’s. And when they are, I’m identified by what I’ve done / who I am — like everyone else.
2) ..which they panned?
Abigail Zuger didn’t like my book — wasn’t convinced, didn’t think I had the bona fides to separate the “gold from the dross” as she put it. Fine. But surely you know enough about papers to understand that she’s a reviewer for one desk at the Times. She’s not THE editor overseeing all desks issuing edicts on who can write about what. And anyway, if all freelancers worked according to the worst review of their book, there wouldn’t be many working.
3) What other MDs say about the book
Zuger’s middling review is in the minority. Here’s what the science writer Matt Ridley said — also a skeptic of the so-called “hygiene hypothesis,” by the way, but familiar enough with principles of evolutionary biology (he wrote the Red Queen) that he absorbed the argument once clearly presented.
But he’s not an MD. Here’s what other MDs, with expertise in the relevant fields, say:
A favorable one by Jonathan Bernstein in the Annals of Allergy, Asthma & Immunology.
Here’s another by Dr. David Katz, head of Yale’s Prevention Research Center in Nature.
And here’s a third in Nature Medicine by Paul H. Plotz at NIH’s center for Arthritis and Musculoskeletal and Skin Diseases.
Plotz takes me to task on number of points. My hypothesizing is irritating. I give too much time to stories of people. I don’t understand the full complexity of inflammation. All were calculated decisions—not appreciated by a scientist maybe, but meant to make the book, which is too full of science as it is, more readable for lay readers. To simplify and include color when possible.
Anyway, after expounding on how research in the microbiome is some of the most exciting in his lifetime, Plotz ultimately calls the hypothesis I explored entirely plausible and deems the book worth reading. He should know, given his specialty.
So: Abigail Zuger aside, scientists who actually work in the relevant fields have a very different take than hers.
4) You’re casting the “hygiene hypothesis” net too widely
I’m assuming my critics haven’t read the book based on — well, on their not being familiar with any of its content.
It’s not titled the “hygiene hypothesis”. And I use the phrase in the book maybe three times. That’s because the phrase, many argue — and I agree — is misleading. It’s antiquated. It doesn’t get at what’s important. Which is simple: the organisms that live on and in your body — or sometimes just pass through — affect how your body works, especially your immune system. That includes parasites, microbes, viruses, fungi. These communities, which push and pull your immune system in various directions, may have shifted in the past 150 years. We’ve lost some residents, such as H. pylori. Maybe we’ve gained others. They may also have shifted with our changing diet.
So, if what you mean when you say “hygiene hypothesis” is David Strachan’s original formulation ~25 years ago—that early life infections lower the risk of hay fever later—that’s not what I’m talking about. It’s not what the vast majority of researchers in the field are talking about. (A few other proposed phrasings: “the microbial deprivation hypothesis”; “the old friends hypothesis”; “the deranged microflora hypothesis.” None are perfect, but they illustrate the inadequacy of the first.)
My book explores how these shifts in our flora and fauna may have contributed to allergy, asthma and autoimmunity, as well as, briefly, the greater “diseases of civilization.” These are not my hypotheses, of course. They come from scientists across a number of fields. And by definition, they’re complicated. But they deserve to be explored.
Consider: ulcerative colitis increases the risk of colon cancer. Why? Probably the chronic inflammation. What causes inflammatory bowel disease? There’s a genetic component — a NOD2 variant, for example. But turns out that the NOD2 variant may have its effect by shaping the microbiota in an unfavorable way. This seems to be a feedback loop between us and our microbes. Question is, how to interrupt it?
Well, it’s very strange that parasites can seemingly interrupt this feedback loop in monkeys with spontaneous colitis. How? Perhaps in part by restoring the microbiota.
Another example.
Why, in metabolic syndrome — the condition linked to heart disease, diabetes, some cancers and dementia — do scientists observe elevated microbial detritus in circulation (metabolic endotoxemia)?
Endotoxin comes from your gut microbes. Surely it’s not causal? Well, when injected into mice, endotoxin speeds up the development of obesity and diabetes.
Inject endotoxin in human volunteers, and you inflame adipose tissue and increase insulin resistance. Not good. A step closer to diabetes.
Maybe this shouldn’t be too surprising. Decades ago, scientists noticed that insulin resistance increased during acute infections—during inflammation. And some scientists been arguing for a very long time that the inflammation seen in Met S is not just a result, but perhaps a contributor to the disease.
These are complex interactions. But here’s what’s seemingly important: the closer scientists look, the more the immune system appears to be acting in diseases that have not been traditionally considered immunological at all. It’s REacting in fact, perhaps to bacterial detritus leaking into circulation from the gut (Met S), perhaps to diet, perhaps to who knows what. But it’s activated in the absence of an obvious infectious challenge. That’s the takeaway.
I discuss these complicated dynamics in my book (and elsewhere). So when I’m accused of casting the hygiene hypothesis net too wide, frankly, I’m not sure what the critics are referring to. Because 1) the hygiene hypothesis is a dated somewhat meaningless phrase at this point; I don’t really use it; and 2) you overlook the tremendous variation of phenomena under consideration, all of them acting through different pathways, but all of them relating to this question of how we interact with the organisms on and in our body, and also around us.
Anyone who read the book knows this. They’d also know that after I guide the reader through 350 pages of possibilities and exploration, I ultimately don’t make any recommendations aside from eating well (not too much junkfood) and exercising regularly. Which we already know. At this point the science isn’t certain enough to recommend more.
And I am far from the only or highest profile person writing about this stuff. See Michael Specter’s piece.
See Michael Pollan’s.
See this one in Smithsonian.
Carl Zimmer constantly writes about it. (Gina Kolata occasionally writes about it at the NYT. Others do as well.)
They all cover territory I also covered in my book. And that’s because this science is happening now. It’s burgeoning. It’s interesting and illuminating. And it’s worth writing about.
To the crux of the matter.
5) The autism op-ed
There are two issues with that autism oped. One is the science. The other is how I wrote it. I’ll address them in reverse order.
I wrote the piece in an oped voice — simple declarative language. Few caveats. Not in the conditional tense, as per most science writing.
Two reasons for this: I hoped to get this largely ignored but robust hypothesis explained in very limited room. (I’m talking the prenatal immune activation model, not the link to the so-called “hygiene hypothesis,” on which more below.) So I made a decision to mostly omit caveats and the qualified language. It was a matter of fitting a complicated argument in a smallish space.
I did this, in part, because that’s how op-eds are generally written. They live in the Opinion section. Readers assume, I think, that they exist bracketed in the conditional tense, not least because they literally have “opinion” stamped on them. You’re meant to consider the argument, the evidence presented, not swallow it as fact.
I’m sure we could argue about how true this is, whether it’s appropriate to assume, etc.
But here’s what I want to say:
In retrospect, writing this way may have been unwise. Especially for autism. And especially because it prompted some to write off the science being discussed. That’s exactly the opposite of the piece’s intention.
But that says nothing about the validity of the underlying science and what it suggests.
Now, it’s no secret that we’re plagued these days by what are often called, as a group, “inflammatory diseases.” It’s discussed nonstop in the literature. Here’s a nice recent discussion.
Autoimmune disease and metabolic disorders are often included in this category. So if these diseases, when in the mother, are really a risk factor for some subset of ASD — and if the relationship is not just coincidental, but causal, as suggested by the many animal experiments (see more below) — then the logical question to ask is, What’s behind the rise of inflammatory diseases?
Many ideas exist — pollution, endocrine disruptors, stress. But a very good hypothesis with loads of evidence suggesting it has merit is the one I explored in the book. (Another good one is diet. But with diet, you can’t escape the microbiome. Germ-free mice can eat crap food without becoming obese. That’s Jeffrey Gordon’s study that, arguably, launched the microbiome field. Experimental helminth infections, meanwhile, also protect against metabolic disorders by, in part, shifting immune function.)
Again, without knowing what people mean when they say “hygiene hypothesis,” I should emphasize that am not talking about acute childhood colds preventing anything (Strachan’s original idea). Or not taking a shower. I’m talking about how what lives on and in us — or just passes through — affects our immune system and metabolism, and how those communities may have shifted away from what was likely the ancestral state. This is not my idea, of course. It comes from literally hundreds, possibly thousands, of scientists working across many fields, from anthropology to microbiology.
Moreover, this link between ASD and prenatal immune activation is not going away.
Here was my source list then.
6) What’s happened since last year’s op-ed
Alan Brown has found levels of CRP in pregnant mothers to correlate with ASD in offpsring in a dose-dependent fashion. The higher the CRP, the greater the risk. Brown of course pioneered this hypothesis in schizophrenia.
This corroborates an earlier, smaller study on mothers in California.
Also, elevated pro-inflammatory cytokines in amniotic fluid of children who develop ASD.
Van de Water et al show that those fetal brain-binding antibodies associated with ASD in offspring show up in about 23 percent of her cohort.
In a larger (2431 mothers) cohort, Betty Diamond finds them in only 10 percent of mothers of children with ASD (versus 2.6% of mothers without). But that’s replication of the association.
Diamond also finds an increased prevalence of rheumatoid arthritis and lupus in these mothers. That observation of autoimmune disease in some ASD mothers has been made several times before in various size cohorts, the largest being in Denmark.
But here’s another in Canada, showing a more than twofold increased risk of ASD in offspring born to mothers with lupus.
What about the causation question?
The UC Davis group showed that these antibodies not only prompt behavioral changes in offspring when injected into pregnant macaques, they also seem to change brain size, causing enlargement.
That mimics what Eric Courchesne (UCSD) has been observing in a subset of ASD human children for years—”brain overgrowth.”
Some years ago, Chris Coe also prompted “brain overgrowth” in macaques, albeit with a generalized inflammatory hit consisting of low-dose LPS (endotoxin). He also saw behavioral disturbances in these offspring. It’s not exactly replication — a different immunological stimulus was used — but it’s at least corroboration.
Meanwhile, search Pubmed for maternal obesity and inflammatory markers in amniotic fluid, as well as maternal obesity and behavioral disorders in offspring. Plenty of stuff. And several animal studies suggesting Met S / obesity alters fetal development in various ways, including in the brain.
So: these observations are being replicated, often in triplicate. Good primate models suggest causation, also by different groups. Meanwhile, the genetic argument is probably not as strong as geneticists, who don’t argue it’s that strong anyway, might have you believe. (See here.)
Older twin studies suggested 90 percent heritability. But more recent, much larger twin studies strongly suggest environment in the etiology of ASD, not genetics.
“Where did all the heritability go?” asked one scientist recently.
More importantly, these arguments are not mutually exclusive. The abiding question is: why may prenatal immune activation NOT be a problem for the great majority, but only for a minority? Genetics will certainly play a role in defining that minority. This is likely a gene-environment interaction — not solely mediated by either factor.
And partly due to inconsistent genetic evidence (not counting de novo mutations, by the way, which can be caused by environmental factors), as well as the emergence of viable “environmental” hypotheses like this one, NIH’s Thomas Insel is calling for a renewed emphasis on possible environmental factors.
Long story short, this idea is not disappearing. The opposite, in fact. It’s rapidly accreting. I’d like nothing more than to see an informed discussion about the evidence– the science. But that does not mean simply ignoring it. I’d be happy to discuss my book, provided you read it. An intelligent debate. A little research perhaps. A little reporting. Falsify it. Verify it. Dig up the really good caveats (Such as: when you alter brain development in mice, you also produce disturbances in immune function — a blow to the unidirectional causation argument.)
Something like that would be productive.
But these drummed-up insinuations about the NYT and digs at a book you haven’t read — informed by a misunderstanding of an hypothesis whose complexity you either ignore or don’t comprehend — they’re unfair to the science, to the considerable research and reporting carried out while looking into the science, and to readers.
