Tuesday, September 11, 2012

Fathers? Do they matter or are they just to blame for autism and schizophrenia?

In its Sunday Review, the New York Times has run an article entitled, "Why Fathers Really Matter," also posted online. The Twitter conversation among scientists and science writers that ensued about the article focused primarily on its confusing presentation of epigenetics and related science, but after I reading the entire thing, I found three problematic areas. So I'm partitioning this critique by theme: Science, autism and schizophrenia, and the blame game. Of interest, this piece that homes in on fathers as "culpable" in autism and schizophrenia comes hot on the heels of another recent NYT op-ed arguing for the mother's 'culpability' in autism. Good times.

Science
The writer is Judith Shulevitz, science editor for The New Republic. The science stumbles begin in the second paragraph, some as relatively minor oversights, such as when Shulevitz states that mothers "create the environments" in which babies develop. In fact, the womb environment results from complex interactions between the embryo/fetus and the mother, and it can be difficult to tell who's influencing whom. 

Then we get into genetics and genes. Shulevitz writes about a man's genes being bad or good, with a nod at environment's determination of this status. But this phrasing incorporates the oh-so-common error of describing the gene as having this good-or-bad quality when in reality, it's the gene variants that determine differences among us. If things are going the usual way, we all inherit the same basic suite of gene complements--on 45 to 47 or so total chromosomes--from our parents. That's what defines us as Homo sapiens. What makes us different from each other is the differences in the sequences of those genes or variations in their numbers and how we use them.

I've thought about introducing a drinking game for all articles about autism in which every time the writer mentions "toxins," you have to drink a fifth of your hard liquor of choice (I'll take vodka). That word comes up for the first time, unqualified, in the third paragraph in this piece. This paragraph presents one of the themes of the article: It describes epigenetics, the regulation of genes that doesn't involve changes in gene sequences themselves. In the current context (epigenetics can be broadly or narrowly defined), chemical tags on gene sequences serve as one of many varieties of on/off switches for genes. If you think of these sequences as text instructions, the cell hides (or reveals) certain sections of text during specific periods of development or in specific cell types. 

This cellular censorship is a handy way of ensuring that your liver cells don't act like they're part of your heart or that Embryo You didn't try to grow fingernails before you had fingers. The chemical tags the cell uses to do this are like reversible Wite-Out, making these instructions unavailable or available, as needed. Shulevitz writes, however, that we should "think of epigenetics as the way our bodies modify their genetic makeup." This phrasing is the first fuzzy conflation of sequence changes in genes and epigenetics. Our bodies also modify their "genetic makeup" through mutations, or changes in the gene sequence. This statement doesn't distinguish one process from the other. 

Shulevitz finds epigenetics both "game-changing and terrifying." Yet as a gene-regulation tactic, it appears to be present all over the mammalian family tree. It's been around as long as we have and shapes our development from conception. I could see some games that it changes, but why is it terrifying? She says it's because three environmental factors can switch our genes on or off, citing what we "ingest (food, drink, air, toxins --drink!); what we experience (stress, trauma); and how long we live." 

First, our uptake isn't limited to ingestion. We take up stuff we encounter by three natural routes--ingestion, inhalation (that'd be the air; please try to avoid ingesting air--gives you gas), and skin absorption (injection into various places is another possibility). Those routes and the exposure levels matter. For example, you can soak your fingers in water for a very long time with the only adverse effect being that they'll look like raisin-fingers. But if you inhale water, particularly at a certain volume, that'll kill you. And sometimes, what we take up, however we take it up, actually helps by its influence on what we do with our genes.

That takes me to ... second, our potentially influential experiences are not all negative. Positive experiences, like good friendships, good wine, and nice dogs, influence our physiological responses, too. Why so negative, people? Third, how long we live likely isn't causative in our gene expression as much as it allows time for accumulation of changes, whether epigenetic or mutational. Finally, beyond the genetic factors involved in regulating our genes and how we use them, many, many, many factors "switch our genes on or off," including being conceived, growing and developing, reproducing, reading bad novels, sleeping a lot, and living in New York City.

After citing a few very interesting epidemiological studies that hint at heritability of environmentally induced changes--presumably meaning inherited patterns of this chemical tagging--Shulevitz makes some more missteps. One relates to 'vinclozin' (sic; it's really vinclozolin), the antifungal she cites as an endocrine disruptor and environmental "toxin" (drink!),  saying that "it blocks production of testosterone." I think this interpretation comes from its having been described as an "anti-androgen." 'Anti-androgen' doesn't, however, mean that it blocks production of testosterone. It means that it gets in the way of the jobs that testosterone and other androgens do. That tends to translate as interference in signaling pathways for growth, development, and behavior. 

The point of the vinclozolin story, as told in the article, is that male rats born to a mother who received a "fat dose" of this antifungal had "defective testicles and reduced fertility," as one might expect from an anti-androgen. These problems then showed up in three subsequent generations. Shulevitz doesn't mention it, but researchers also found evidence of altered patterns of chemical tagging in the sperm-producing cells of these male rats in different generations, supporting an epigenetic explanation. Worth noting, the same group also looked in a later study at the female offspring and found changes in kidney pathology and tumor rates in subsequent generations. 

Then we get to the mice. Always, always the mice. The study described is about highly stressed male mice and the babies they father and seems intended to support the idea that paternal lifestyle--paternal mouse lifestyle, anyway--can result in genetic Wite-Outs that offspring then inherit. The only problem is, in this study, the one test that would have isolated a paternal epigenetic pattern as causative indicated that it wasn't. The researchers took sperm from the stressed-out male mice and impregnated the females with it. The "offspring were largely normal," according to a study author. Only when the females were impregnated Nature's way were the offspring more anxious than control mice. That points to a maternal influence in response to the father mouse's stress, not his chemically tagged sperm.

Age, writes Shulevitz, might play a role in epigenetic influences but then she cites a lot of studies associating age with mutation accumulations. She notes correctly that making sperm over a lifetime requires hundreds of cell divisions and that "mistakes may appear in the DNA chain" with each division. That description is of mutations, however, not epigenetic chemical patterning. She then writes that "some researchers" think that some of these mistakes are not "just random mutations but experience-based epigenetic markings that insinuate themselves from sperm to fetus and influence brain development." 

A "mistake in the DNA chain" can't suddenly become an epigenetic marking. Detection of gene sequence changes versus chemical tags and their patterns involves different processes, and one isn't mistaken biochemically for the other. Yet the article continues this conflation, going on to cite two studies. The first is a risk-association study showing a relationship between increasing paternal age and autism in offspring, leading Shulevitz to ask, "Can the aging of the parent population explain the apparent spike in autism cases?" I'd argue that there's very little to that spike, so I'd have to say "no." 

The other study is a high-profile Nature study that identified higher mutation numbers in the sperm of older fathers compared to the eggs of older mothers. But again ... those changes are in the sequence, not in the chemical tagging pattern associated with epigenetics. The Nature study linked some of these sequence changes to risk gene variants for autism or schizophrenia, thus linking older fathers and autism. [Note for all of you older fathers: Children of older fathers tend to live longer. So there.]

And now ... enter the autism. 

Autism and schizophrenia
The buzzy autism part of the article opens with a story about ... mice. Old mouse fathers are more likely to produce mouse offspring (not children, as Shulevitz generously describes mouse pups) that have "cognitive-behavioral handicaps" in adulthood and "less adventuresome personalities." I'm assuming that means they didn't impulsively try recreational drugs and drive at high speeds. "Super-high-performing" mice, on the other hand, were more likely to have come from younger mouse fathers. Of course, what's "super-high-performing" in a mouse might not necessarily be what is considered such in people.

And then this sentence:
One unanswered question about autism and schizophrenia is how they crop up in generation after generation; after all, wildly dysfunctional individuals don’t usually flourish romantically.
First, natural selection doesn’t care too much about what your genes or other factors do to you after you reproduce. Ironically, those desirably younger fathers (and mothers) who might later develop schizophrenia can reproduce before onset of the disorder, which in men is at an average age of 18 into the 20s and in women is at an average age of 25. In addition, schizophrenia has a pretty strong genetic component, as does autism, but the genetics are complex, not one-gene/one disorder. As with any complex behavioral manifestation, several pathways might need to merge for the condition to ultimately manifest. That doesn't mean those pathways don't exist in part in many members of the population. Most people in the world have brown eyes, but lurking behind many of those brown eyes is the capacity to produce a blue-eyed child, given the right genetic combination with a partner. I can personally attest to that.

Second, people with autism or schizophrenia are not by definition “wildly dysfunctional,” no matter how you translate that phrase. Autistic people are all around you, reproducing, sometimes working, yea verily even “flourishing romantically.” Ditto people with schizophrenia, a term that often gets vague use and can reference a diverse group of conditions with a range of manifestations. And given the popularity and romantic or reproductive success (if you can call it that) of truly “wildly” dysfunctional people, such as the Byronically dysfunctional Heathcliffs of the world, “bad boys,” murderers, or Keith Richards, I can't even buy the statement that "wildly dysfunctional" people don't flourish romantically (although, of course, romance isn't required; just ejaculation at the right time and place). Also, have I mentioned that a whole lot of people with autism or schizophrenia function well enough to read and understand articles like this one?

Of course, none of the scientific information presented in the article gives evidence supporting the involvement of inherited epigenetic patterning in these conditions or even in the outcomes observed in the mice. It describes heritable epigenetic associations for diabetes, weight, and reproductive endpoints but not for autism and schizophrenia. In fact, the article presents a series of findings that support a pretty straightforward and not-that-surprising role for accumulated mutations with age in influencing outcomes in the offspring. In this case, the "environment" is parental age--just living long enough to gather mutations in the germline (which was horribly misdefined in this article)--but the real influence, based on the two studies cited, is "changes in gene sequence." [NB: That's not to exclude an influence of epigenetic pattern changes accumulated in early life in schizophrenia, e.g., here, and some studies homing in on imprinting based on parental origin offer some interesting hints; probably the best-known example of parental-origin imprinting influence on a disorder is Prader-Willi/Angelman Syndrome, each of which respectively develops depending on the parent of origin for the mutation. The chemically tagged DNA from each of your parents is supposed to hit "reset" in preparation for forming a zygote so that the male-female chemical tagging patterns are mutually complementary and development can proceed as it should.] 

And then it gets rather ugly in this quote from self-described "inherently skeptical person" Jay Gingrich, a professor of psychobiology and the fellow who conducted the old-mouse-father study. Referencing people with autism and schizophrenia, he says: 
From a cruel Darwinian perspective, it’s not clear how much success these folks have at procreating, or how else these genes maintain themselves in the population.
This reference approaches the cruel social Darwinism of explaining away the relevance of "these folks" and why "these folks" still even manage to exist, confound their wildly dysfunctional hides. First, speaking of a cruel Darwinian perspective, as I noted, schizophrenia onset can easily occur after a person has already reproduced. In that sense, “cruel” Darwinism--i.e., natural selection--doesn’t give a mouse's less adventurous ass about whether or not a schizophrenia-related gene variant is present. Second of all, given the huge heterogeneity of autism, the number of rare autism-associated variants identified, and the apparently significant genetic component, it’s not a big stretch to think that “these folks” toting these things around are capable of, you know, doin’ it. Being autistic or carrying risk alleles for autism doesn’t mean you don’t or can't have sex. In fact, autism as a kind of family trait is not that rare.

Then we have the "autism is increasing" phrase, as necessary to any article about autism as the word "toxins" (which occurs a third time, so drink). Gingrich is again quoted in the piece:
I think we’re going to have to consider that advanced paternal age, with its epigenetic effects, may be a way of explaining the mysteries of schizophrenia and autism, insofar as the rates of these disorders have maintained themselves — and autism may be going up.  
That reference to epigenetics just comes out of nowhere, given that the preceding studies discussed in this context seem to rule out epigenetics and rule in sequence changes. As for "explaining the mysteries," just get in line behind the other 6471 people walking around with the key to the “mystery” (trope alert) of autism.

The blame game
The title of this article is, "Why fathers really matter," but it should have been "Now we can blame fathers, too!" It's not the first article to use the Nature old-dads+mutations=autism/schizophrenia study as a springboard to delighting in a newfound ability to blame men for something. The Daily Beast got the offensive scoop on that one. But this piece brings the imprimatur of a scientist--rather than an angry divorced woman--seizing the moment to throw some blame around. "Older mothers," writes Shulevitz, "no longer need to shoulder all the blame," and she quotes Dolores Malaspina, a professor of psychiatry, as saying,
It’s the aging man who damages the offspring.
I guess some people's damage is other people's treasure.  

It's true that mothers of autistic children have been blamed in the social sense and continue to be blamed, directly or indirectly, for their child's autism, whether it's hypothetically dangerous wombs or refrigerator-like personalities. Those associations without solid evidence were and are unfair. But obviously, there is a maternal contribution to neurobiology. Noting that fact shouldn't be a matter of "blame," it should be a matter of science. Stooping to revenge for unfair blame burdens by using the language of "blame" and "damage" against men does women no favors. It also does people with autism and schizophrenia no favors, treating them as the living, breathing negative consequences for the people who are blamed for having them.

Shulevitz casts our growing understanding of epigenetics as ushering in a "new age of sexual equality in which both sexes have to worry about threats to which women once felt uniquely exposed" and closes by saying, "Having twice endured the self-scrutiny and second-guessing that goes along with being pregnant, I wish them (the fathers) luck." As a blame facilitator or not, epigenetics isn't going to usher in a "new age" of sexual equality.  Based on epigenetics studies such as the cross-generational influence of DES exposure, women also can accumulate changes above the gene in their lifetimes that pass to their offspring and later generations. We are not excluded from making epigenetic contributions. 

So those fathers who now so delightfully have to worry about threats had mothers, in addition to fathers, from whom they presumably inherited some epigenetic baggage, as did their mothers and fathers, and so on. Yet men are never going to have the same level of scrutiny and second-guessing "that (go) along with being pregnant," no matter what epigenetics research tells us about them, because they simply don't get pregnant. There's no way to equalize the effects on offspring of a man's smoking in his teens and a woman's lighting up throughout her pregnancy.

The male contribution to offspring development continues to be the nuclear DNA and whatever potentially epigenetic baggage it carries. A woman? We contribute all of the original cellular components, all of the mitochondria and the mitochondrial genes, our nuclear genes, and whatever epigenetic baggage we carry ... and we carry the child. Findings pointing to lifestyle-related male epigenetic contributions are not going to level that field or equalize with our contributions to our children's genetic and epigenetic destiny. And I'm OK with that. It's the evolutionary duty I inherited as a human female and the first of many conscious responsibilities--not blame--I took on when I decided to become a mother. The mother of an autistic child, I add, who is not "wildly dysfunctional" and whom we have always viewed as a treasure.

-------------------------------------
7:22 EST: Minor changes made for paragraph transitions.

7 comments:

  1. Excellent work. It taught me much, and not just about the science. I hope everyone who reads the original gets to see this.

    ReplyDelete
  2. I really appreciated the way you described epigenetics etc. in non-technical terms without wrongifying it. And this is on top of appreciating the way you fight the derogatory references to autism and other conditions. Autistics not only read newspaper articles, some of us study biology and can be offended by the NYT author's ignorance of biology AND autism.

    ReplyDelete
  3. I think you are being excessively nitpicky to the detriment of your main points.

    Consider your criticism of the author's discussion of epigenetics. You take issue with one sentence of the definition she gives, saying that it confuses epigenetic changes with dna sequence changes, but ignore that the very next sentence completely clarifies that issue. In the next paragraph, you criticize the use of the phrase "game changing and terrifying" by suggesting that it can't possibly apply because epigenetic factors have been around for millions of years and are present in all mammals. The author quite obviously intends to suggest that a full accounting of paternal epigenetic effects on offspring could change the way we view men's behavior and its long term consequences. What do millions of years of evolution and 4000 other mammal species have to do with that? And then there is your criticism of the use of the term "genes" when "gene variants" is meant. This is absurd, as scientists themselves frequently use "genes" as shorthand for "gene variants" (a textbook example is the "good genes hypothesis" in sexual selection theory). It goes on.

    I get that your main objective is to criticize the underlying science brought to bear in this article, but your nitpicky and unfair characterizations of the author's writing make me mistrust the rest of your arguments, many of which are based on literature I'm not familiar with and facts I don't have time to verify.

    ReplyDelete
    Replies
    1. Your commentary on my commentary is in itself nitpicky, which is fine. I stand by what I wrote, the support I provided, and the criticisms I made. The sentence that was intended to encompass epigenetics simply lacked the necessary specificity. If you're going to write about molecular biology and genetics, specificity is important.

      No argument of "they do it all the time" is a defense. It is not absurd to argue for the use of "gene variant" (a.k.a. alleles, SNPs, etc.) as being more accurate and scientifically valid phrasing than "gene," and I'm not the only science writer to take issue with that. Indeed, I know many scientists who would completely disagree with that defense as valid.

      My main objective in the science writing context is not to criticize, it is to clarify and to argue for making as accurate and precise a presentation of scientific information as we can, particularly when writing for a publication with a read as broad as that of the NYT.

      I was quite careful to be as fair as possible in this analysis--there were many other areas that were very soft on the science, but I left those out because including them would have just been beating a dead horse. If that care and the accuracy I strive to use here makes you mistrust what I write ... that's a decision on your part that I can't help.

      ETA: Also, while I will cop to being nitpicky, the bigger issue with this article is that it tries to weave epigenetics into autism as an etiology, but what is presented that is specific to autism and schizophrenia does the exact opposite of that. This failure is global, not nitpicky or minute.

      Delete
    2. Now I have to say "thank you" for your commentary because I asked about it on Facebook and am having a useful discussion about how detailed is too detailed. I have learned that I should worry "about attacking targets that seem too small." Honestly, for this post, I debated keeping it shorter and sharper (like I might were I writing to a word count for an outlet), but because it's my personal blog, I stuck with being detailed. Something to reconsider, perhaps.

      Delete
    3. On a blog, there is no need to keep it short - readers are not paying by the word to read your work. I'd say you did what was necessary to make your points, which were worth making. And the difference between a gene and an allele is important for all but the most casual uses. We all have the same genes - the heritable differences among us come (largely) from the different versions of those genes that we inherit. That is not a difficult concept to impart to readers, and when a science journalist gets it wrong - as a Bio 101 student wouldn't - there's cause for concern in the rest of the work.

      I would add that we occasionally hear that epigenetics somehow overthrows the 'laws of genetics' somehow. The truth is that epigenetics is easily incorporated into our standard understanding of genetic mechanisms. Epigenetics is simply a modification, and a rare one at that. Some people - usually the undereducated - have a tendency to want to run with the next big breakthrough, and find it where it doesn't exist.

      Delete
  4. An interesting article and breakdown of points. I read the article, and winced. I A) do not consider myself to be "wildly dysfunctional", and B) suspect that if I had an interest in romantic connections, I could get them. As for my decision not to have children, it has little to nothing to do with my ASD (plus, I made it when I was 11, over a decade before I even learned I might be an Aspie). I do find this article at least somewhat insulting, especially in the negative view it takes of ASD, and the whole conflating of ASD with schizophrenia.

    And while I'm aware that one example does not a counter-argument make, I would throw out the point that I am the oldest child in my family, the second child has bipolar disorder, the third has had clinical depression, and the fourth has had no issues relating to any of those. Yet my father was older when the youngest of us was conceived than when I was, and I'm the one who is ASD. *shakes head*

    ;) tagAught

    ReplyDelete