The Great Divide In Autism Research: Genetics Vs. Environment

By Karen Weintraub
Guest Blogger


For years, I’ve been confused about who to believe in the autism research world: the people who say autism is all in our genes or the ones who believe something bad in our environment is triggering a terrible epidemic.

A paper published Monday in the Archives of General Psychiatry seemed particularly appealing, because it argued for a middle ground. Genes are a risk factor, the Stanford University study found, but environment plays a major role, too, accounting for more than half the risk of developing autism.

Argument settled, I naively thought. All this week though, scientists on both sides have continued to snipe at each other.

Millions At Stake
This would be merely an academic argument except that the U.S. government spends well over $150 million annually on autism research. Over the last 10 years, taxpayers have devoted roughly $1 billion to studying genes involved in autism and only about $40 million digging into possible environmental causes. Whoever wins this argument could sway future spending.

Plus, of course, there are millions of families struggling with autism: the communication difficulties, repetitive behaviors and social challenges. Settling this dispute will effect which treatments they try, and what all of us can do, if anything, to prevent autism.

I finally realized today while talking to a geneticist that this difference may never be resolved. The two camps are so distinct – their world views, and the language they use, are altogether different.

Then there are the politics. Mention “environment and autism” and you’re bound to trigger the intense debate over whether vaccines cause autism (a debate that continues despite multiple solid studies that have found no link — though some say research hasn’t yet established whether a small subset of already-vulnerable kids could be harmed by vaccines.)

But if you want to understand autism research and where it’s heading, I think you need to understand where both groups are coming from, and why they can look at the same data and come up with opposite conclusions.

The Great Divide
First, the world according to most geneticists and molecular biologists:

By training and temperament, these folks only trust things they can see and count. If it doesn’t show up in a gene array, it doesn’t exist.

Michael Ronemus fits in this camp. He’s a genetics researcher at Cold Spring Harbor Laboratory, a genetics and molecular biology research institute on New York’s Long Island. He desperately wants to understand autism to help his 9-year-old son who was once diagnosed with the condition, but he’s sticking to what he sees as hard, provable data.

Ronemus thinks we need to first understand all the genes involved in autism before we look to the environmental influences. Knowing which genes are involved will give us clues to the environment, he says. And starting from the environment and trying to work backward is just too squishy.

He wanted to believe, for instance, that eating gluten – a protein in most grains – might contribute to autism symptoms. After all, many parents say their kids improve dramatically when breads and pasta are taken away. But Ronemus says it was hard to get his son to give up those foods, it didn’t seem to help, and none of the autism-related genes that have been found play a role in gluten digestion. So, he scratched that theory off his list. “The things that are in DNA, we can find,” he told me.

To Ronemus, Monday’s study drew the wrong conclusions out of good data.

What Doctors See
The other world view is held mainly by doctors who believe the evidence they see in their clinics, rather than what an abstract number supposedly reveals.

They know that when they came into the profession a few decades ago, they almost never saw a child with autism. One doctor told me he remembered being called into the room during his training to watch an exam with a child who had autism. It was important to see, his supervisor told him, because it might be years before he saw another such child. Now these doctors, see kids with autism every week, if not every day.

They can’t believe that a 40-fold increase in diagnosis is due mainly to an expanded definition of the condition.

To this group, Monday’s Stanford study is validation. “I think it’s going to make the funders wake up a little bit,” said Irva Hertz-Picciotto, an epidemiologist at UC Davis who is firmly in the second camp. “If we’d had this 10 years ago, we’d been a lot further ahead. Hopefully it’s going to change the pattern for the next 10 years.”

The study, led by Stanford psychiatrist Dr. Joachim Hallmayer, pinpoints pregnancy or early childhood as the most likely times for the development of autism.

Now What?

Researchers looking for environmental causes of autism are focusing on five basic areas:

1. Immune problems
Mothers with rheumatoid arthritis and fathers with type 1 diabetes are more likely to have children with autism, suggesting an immune risk. It’s also theoretically possible, but unproven, that a virus the mother catches during pregnancy might contribute to autism (as seems to be the case in schizophrenia); or a virus the child catches early in life.

2. Chemicals in the air, water, food or on the ground
There’s some research suggesting that children of farmworkers exposed to pesticides are more likely to develop autism, and plenty of people are worried about chemicals, like BPA and flame retardants, which are omnipresent and which seem to impact hormones, some perhaps involved in autism.

3. Problems during birth and delivery
There is a possible link between C-sections and autism, and potentially also with the use of pitocin, a drug given to speed the birth process. Complications during birth may also be connected to autism, though it’s not clear whether the complications cause the autism or the other way around.

4. Medications
Another study out this week suggests that antidepressant drugs may be linked to autism, though mothers who are truly depressed may put their child at higher risk for problems by forgoing antidepressants. Also it’s theoretically possible that the “hygiene hypothesis” we hear about in connection to asthma is at play with autism, too. This hypothesis suggests that by cleaning up our environment too much, we’re getting rid of good bacteria along with bad, and could be making our children more vulnerable to new problems.

5. Nutrition
One recent study by Hertz-Picciotto shows that women who took prenatal vitamins just before and in the first two months of pregnancy were less likely to have children with autism than women who didn’t supplement. Researchers are also exploring rates of essential fatty acids, selenium, and molybdenum to see if they might be linked to autism.

The bottom line, according to Dr. Isaac Kohane of Children’s Hospital Boston and Harvard Medical School, is that for now, it’s easier to study genetics – though environment clearly plays a key role. “With genetics, we’re given a very nice hammer and of course something that looks like a nail,” he told me via Skype while on a family vacation. “In environment, it’s not even clear that we’ve found the right hammer. With neither the right hammer nor the right nails, our [research] money might be wasted.”

Hopefully, this week’s study will start to convince people that we need to come up with better ways to study the environment, too.

If you want to read more, see this CDC site that follows autism research, and check out the advocacy group Autism Speaks which funds and follows autism research.

Karen Weintraub, a freelance Health and Science journalist, has just finished writing a book with someone who subscribes to this second world view.

Please follow our community rules when engaging in comment discussion on this site.
  • Guest

    I have idea on this, it relates to genes, environment and MTHFR.

  • Anca Diaconu D
  • joe joe

    Since the study said it should look into chemicals, I’d like to know what makes the vaccines sacred that we have to look the other way? They are not tested to true scientific standards, such as double blind studies, I have not seen one yet to this date, so then why are they being excluded?

  • axnreaxn

    my son was diagnosed with autism on july 5 2011. i was aware that this could be a possiblity and a year + later it is a reality. i am just now learning more about this. i had been casually aware but now i am getting involved. i am particularly interested in the pitocin and as one of the commenters mentioned the long term effects of vaccines. my son was very healthy when he was born 4 days late feb 29 2008. i was at risk for pregnant hypertension, under an immense amount of stress, and i am bipolar but i was not taking meds before or while i was pregnant. he was 8lbs 13.9 oz and i was in labor for almost 3 days before he made his first appearance. i was not dialating and he was upside down but facing the wrong way. they were prepping the c-section room when he started crowning, the nurse did not even have her gloves on. i was given pitocin (for almost 2 days i would estimate) and i had an epidural prior to his birth. other than that exciting experience my son has been virtually perfect health wise (both in the womb and out). some minor hiccups i.e. a runny nose or sore throat, thrush one time, a little constipation, and sensitive skin. i noticed early on that he was different, but i chalked it up to being an only child who spends a lot of time with mum. he is currently being evaluated for adhd and hyperlexia. all i can say is this is very scary to me. the prospect of treating his “disorders” with medication when the cause cannot be pinpointed. my son is my whole world, and i wouldn’t change him a bit not even when he is at his most difficult. it is very helpful for me to be able to read your comments. i don’t think there is any data to support vaccines being part of the problem, but i don’t see why they should not still be considered part of the equation. i was 23 going on 24 when my son was born. his father younger still. i was raised in the states, my son’s father is from a very rural area (tibetan refugee from india). as far as i know there is no known autism in either family, but i am not sure as i do not think this is something that would have been recognized in previous generations. i would like to know what data is out there that is concrete about autism.  i understand that new breakthroughs reshape our understanding of things, but what information is out there that is pretty unanimously agreed upon?

    • Karen Weintraub

      Some key things that are universally agreed upon: 
      Boys are more likely to be diagnosed with autism than girls.There is some role for genetics and environment – it’s the balance that’s the question.The only scientifically supported treatment is behavioral therapy started as young as possible. (There are some medications to treat the symptoms of autism, but no drug designed to treat its core symptoms, though several are currently being studied).The diagnosis is subjective. There is no biological standard – no way to “see” autism on a brain scan, or count it in the bloodstream, or whatever. So, deciding who is on the spectrum is not as cut and dried as, say, diagnosing type 1 diabetes. The federal government’s guideline for diagnosing is summarized here:’s autism is different. It’s very hard to predict a child’s outcome and what to do that will help. Some children seem to outgrow autism, some parents work like crazy to help their kids outgrow it, some parents work like crazy and their kids don’t outgrow it. There’s no way to tell in early childhood which category any particular child will fall into. 

  • Ed Guise

    I think this will come down to a house hold chemical. I think labs that test the products we use everyday, Don’t look at the human for the future. but to see how it effects humans now. I never trusted Miracle Grow, Dias anon, or sugar supplements. I bet its right under our nose. just like moldy bread that created penicillin. The EPA and the FDA look at the now….   This works yes, and its good but there’s underlining things that happen in time. Ed Guise Ronkonkoma NY

  • WorthE

    All the money in the world isn’t going to “cure” Autism. Indeed, if you were to ask my son who has Asperger’s if he needs cured, he would be insulted. What we need are evidence-based therapies to help them cope with the disorder. As it is, doctors prescribe a cocktail of anti-depressants, anti-psychotics and ADD medications with significant side-effects, not to mention the charlatans out there selling snake oil like “Truehope.”

    There simply isn’t much empirical evidence for any particular therapy. Insurers don’t pay for ASD therapies–except for the drugs. Since it is classified as a developmental disorder, the expectation is for the schools to intervene. There is no consensus as to how to intervene and the intervention needs to occur long before school age.

    It is clear that the genes provide the predisposition and the environment provides the trigger. Given the significant social deficits and the fact that only the more severe Autism symptoms were diagnoses prior to the DSM-IV, there are a lot of people in their 40′s and 50′s who weren’t classified. The “information age” has given them an opportunity to meet others, get married and have children that didn’t exist a few decades ago. Back then, they simply didn’t have the opportunity to breed, as it were. Silicon Valley has ASD incidence of as high as 20%. Why? They have a natural affinity for computer science and when similar people with similar genetics get together, they have children with more pronounced symptoms. They take a little longer than neurotypicals do to get there, hence the correlation to advanced age for both mothers and fathers. It has far more to do with their social deficits than their degrading genes. [I was 22 when I had my son and his father was 24.]

    Billions have been spent on dead ends. Meanwhile, there are kids who need help now and parents who don’t know where to turn. The money needs to be spent on developing therapies rather than finding culprits at the cost of those already “afflicted.”

  • Mjkropf

    The problem with defining causes/partial causes in the case of Autism, Cerebral Palsy, Fibromyalgia, Breast Cancer or other conditions is the lack of clarity as to the process and site of involvement.  We cannot currently decide where a defect in the brain might be with some evidence for a Cerebellar site from UCSD and more prevalent sites suspected in the Cerebrum.  With no evident focus or neuropathology, trying to ‘dope out’ the condition may be difficult to assess factors.    Epidemiologic studies can find all sorts of associations that may be spurious.    In my first year of medical school, I criticized a study by Duncan Clark (an otherwise superb epidemiologist!) in the NEJM that showed Coffee intake was associated with Pancreatic Cancer.    I doubt anybody believes this anymore.    Statistics don’t lie, but they can surely lead us down a primrose path!    We need a priori evidence, preferably not association data to generate provable hypotheses.    Until we have this, all we can claim to do with most studies is ‘spin our wheels’.    One wonders whether it was appropriate to study all breast cancers together!   We now are starting to study different patterns separately.    We know of at least 3 types of Neurofibromatosis, 180 or more forms of Cystic Fibrosis and Autosomal Dominant and Autosomal Recessive forms of Marfan’s Syndrome.    Even where we have genetic conditions, it may be hard to have clear cut patterns.    Will Autism be proven to perhaps be overlapping disorders?    Is it the same as what has been called Asperger’s and is now to be called Autism Spectrum Disorder?    If we cannot define the nosology, how can we claim to be able to find causation?

    • Karen Weintraub

      Yes, you’re exactly right – the difficulty of studying autism is that it may be the end result of so many different things. Clearly, there’s a genetic component, because there are people who have a single gene mutation, who have autism. But not everyone with those mutations has autism, and simple gene mutations cannot explain the vast majority of cases of autism – at least not that scientists have found so far. It’s theoretically possible that some autism is caused by genes, some by immune problems, some by medication, some by the age of the parents, some by nutritional deficiencies, some by infections….This level of complexity makes autism EXTREMELY difficult to study, and probably explains why there are so many competing theories about its causes and potential treatments.

  • Midwest Mom

    You did a nice job of summarizing some of the major – and sensible – environmental possibilities.  But you said very little about genetic theories, or the basis behind them, *besides* the one dad of an Autistic kid, who focuses on genetics simply because environmental factors seem too hard to pinpoint.

    It only seems logical, to me, that IF Autism rates are truly increasing (and it’s not JUST that we’ve expanded the definition of Autism – which we *have* – and that we’re less likely to hide Autistic kids in institutions – which we *are*), then something about how we eat and/or live must be responsible.

    However, I have Autistic twin sons.  Their father seems clearly on the spectrum (albeit on the mild end – his personality was not considered “Autistic”, back when we were kids).  Their father’s young son by his current wife may also be on the spectrum.  He’s significantly delayed in speech and captivated by objects with repetitive motion.  Their father’s paternal grandfather had an “odd” personality, which *today* might well be diagnosed as Asperger’s.  And one of their father’s few, male first cousins is Autistic.  Whereas, I have a very large extended family and not *one* relative – siblings, cousins, 2nd cousins – is Autistic, besides my twins.  So, I simply can’t ignore what appears like an obvious genetic link, with my own children.

    I’d love to hear more about genetic theories and how common families like ours are; and whether they’re more common in the US, than elsewhere.  But despite where you report that research dollars are going, I hear a lot more about interesting, but usually unproven, environmental theories.   

    • Midwest Mom

      Forgive me for replying to my own post, but I read DethWench’s post about maternal age.  I don’t dispute that it may be a factor in some cases, but not it’s definitely not, in our family.  I was just shy of 22 when I had the twins.  Their father was born when his mom was in her mid-twenties.  His current wife was also in her twenties, when she had their son.

    • Karen Weintraub

      Most of the environmental theories are unproven, because not enough research has been done. The strongest evidence is on aging of parents – mother, father and/or the gap between the two – and on immune problems in the parents. 

      On your point about the family genetics, there are many who think autism rates are unchanged, we’re just labeling more people. The family you describe supports that idea. 

      But there are other theories, too…A writer I admire tremendously, Steve Silberman wrote a piece a decade ago in Wired magazine ( suggesting that the increase in autism may be due to what he called “Geek Syndrome” – that people with mild versions of these autistic traits are now more easily able to find each other and marry, producing children with more extreme versions of the traits. 

      Another theoretical possibility: perhaps members of his family share a genetically-based sensitivity to something in the environment, let’s say the protein gluten, found in most bread products. If they never ate bread, they might not have the symptoms you describe, but when they eat bread, they develop those symptoms. I’m NOT saying this is the case, just that there’s not enough data yet to rule it out.

      I haven’t seen any statistics about how common families like yours are. Anyone else?

      • joe joe

        The age group is not a factor, if it was we would have seen this over the course of history. Does anyone really think that there wern’t older parents before in any combination? Does anyone really think this is new? It’s not, many girls marry much older men in Europe, Asia, etc.

  • Graiae66

    I’m not a scientist. I’m a history and education professor at a small college, and I spend a lot of time studying cultural childbirth practices in the US. For 3 years now, I have been telling people (and students who take my educ courses) that I think there might be a correlation between the greatly increased use of Pitocin and the rise in autism. But it was just a theory–and clearly, I had no academic credentials or research skills to prove it. I kept seeing data showing a rise in autism, and then data showing a rise in Pitocin-augmented delivery (inductions), and thought, this is odd. Wow. Now there might be some evidence to show this is true. Makes me glad I fought tooth and nail not to have an induction (including refusing to show up for the appointment) for a non-medical reason (doctor thought 39 weeks was long enough). Sometimes, nature can do just fine on its own.

  • DethWench

    I am an epidemiologist and I feel you are missing an important risk factor in your report.  Older maternal age (especially after age 30) has been definitively linked to autism.  One of the reasons proposed for autism rates increasing has been the increasing tendency for women (in the U.S., at least) to wait longer and be older before having children as compared to the past.  It’s important to counsel potential mothers and fathers on this environmental risk factor, as it may change their plans as to when to have children.

    • Karen Weintraub

      Yes, advancing age of parents is one of the likely reasons for the rise of autism. It’s not entirely clear, though, whether it’s the mother’s age, the father’s age or the gap between the two.

  • Cheryl Gaudino

    I only see 3 basic areas..what happened to 3 & 4?

    • rzimmerman

      Hi Cheryl –

      Here are the five areas of research listed in the post:1. Immune problems
      2. Chemicals in the air, water, food or on the ground3. Problems during birth and delivery4. Medications5. Nutrition 

      • rzimmerman

        …and sorry for the weird formatting.

        • Cheryl Gaudino

          Thanks…I kept refreshing the page but couldn’t see nos. 3 or 4.

          And thanks so much for the article.  I have a 15 year old with autism, PANDAS, encephalopathy, immune dysfunction, severe GI issues, and on and on and on. 

          He had a horrific reaction to the first MMR and 4th Hib (see pic) at 13 months (erythema multiforme and vasculitis, 105 degree fevers, couldn’t eat, talk, walk, lost all skills).  He had been on antibiotics several times in his first year for recurrent ear infections and had just recovered from the wild chicken pox a month before those shots.  His immune system just couldn’t handle the triple live viruses in the MMR–not to mention the load of mercury he got hit with starting at birth.

          My son was normal at birth.  He hit his developmental milestones until the vaccine injury.  He is a severely affected young man with a myriad of complex health issues.  It frustrates me that so much is spent on genetic research.  A genetic epidemic in the span of 10 or 20 years?  Hmmm…I don’t think so. 

          • Donna

            Hi Cheryl,
            So glad to see your contribution here.  How the author never thought to include it in the article is beyond me.
            My sister has 2 autistic sons (she was 22, 29 at their births).  My son has an aspergers son.  Definitely more common in the boys.. All began showing evidence of problems after the vaccines.
            I believe the propensity of our DNA to produce an autism affected child is due to the vaccines that we endured as children and that our parents endured.   These kids are 4th generation vaccinated children.   What do vaccines do??  They change they change the immune system.. how it reacts.   Maybe the vaccines that we had impacted the children to the point of “propensity” then piling on another 38 shots finished the deal.   I begged my son not to take my grandson for shots.. he did and thus began the slide of eye contact, verbal development and socialization.   My sister,  (another sister) has a grandchild (girl) the same age as mine. .. with no vaccines , she is fine.