News The latest developments in autism research.

Genes dwarf environment in autism’s origins, study says

by  /  6 April 2015
Julia Yellow

The genetic makeup of an individual plays much a bigger role than environmental factors in whether he or she develops autism, according to one of the largest twin studies to date. The findings, published 4 March in JAMA Psychiatry, suggest that genes confer up to 95 percent of the risk for autism — nearly double that of previous estimates.

For example, a 2011 twin study pegged the autism risk from environmental factors at 58 percent. A study last year of more than 2 million people, including 37,570 pairs of twins, found that genetic and environmental factors each confer 50 percent of the risk for the disorder. However, the latter study loosely defined the term ‘environment’ to include spontaneous, or de novo, mutations.

Pinpointing the relative contributions of genetics and the environment is more than just an academic debate. “There is huge concern amongst families and professionals about possible environmental factors that may have a role in the etiology of autism,” says Francesca Happé, professor of cognitive neuroscience at King’s College London and lead researcher on the new study. “Genetically sensitive designs like twin studies are essential to estimate the magnitude of environmental effects in a population.”

To parse the source of autism risk, researchers often turn to twin pairs. By comparing autism rates in identical twins, who share all of their DNA, and fraternal twins, who are only as similar as non-twin siblings, they can tease out how strong of a role genetics and shared environmental factors play in autism risk.

Tapping twins:

Studies of autism in twins date back to 1977, when researchers provided the first evidence that autism is heritable rather than springing solely from environmental factors.

A 1995 follow-up of 40 twin pairs found a so-called concordance rate for autism of 92 percent for identical twins and 10 percent for fraternal twins2. Several other twin studies have yielded rates that suggest a strong genetic contribution.

The new findings are based on data from the Twins Early Development Study, which follows twins born in England and Wales between 1994 and 1996. The researchers selected participants from hundreds of families in which at least one twin has received a score suggestive of autism on the Childhood Autism Spectrum Test or has been diagnosed with autism.

From this sample, they found 146 twin pairs in which at least one sibling received an initial autism diagnosis. They then performed in-depth diagnostic tests as well as cognitive assessments during home visits to 129 of these families. Parents also provided reports on each twin’s development.From this information, the researchers confirmed 141 cases of autism and identified 40 children who have autism-like traits, or ‘broad autism phenotype.’

If one twin has autism, the odds of the other twin having the disorder range from 62 to 94 percent for identical twins and 5 to 61 percent for fraternal twins, the researchers found.

A statistical model then produced an estimate of the relative pull of genetic and environmental factors on autism risk: Genetic influences account for 56 to 95 percent of the risk and environmental factors, such as prenatal exposure to chemicals and parental age at conception, contribute 5 to 44 percent.

This statistical model also supports the idea that autism traits and skills are continuously distributed throughout the general population, with people who have autism at one extreme.

However, the model does not account for interactions between genes and the environment, says Janine LaSalle, professor of medical microbiology and immunology at the University of California, Davis.

“The reality is that there are many known complex gene-environment interactions that have major effects on shaping phenotypic traits in humans, and are beginning to be uncovered in autism spectrum disorders,” she says.

LaSalle also points out that 93 percent of the individuals in the new study are Caucasian and all are from the U.K. This homogeneity may explain why the estimate for genetic risk is higher than in the earlier twin studies, she says. “It’s pretty clear that environmental exposures vary by race, socioeconomic status, geography and occupation.”

One big challenge is to identify which exposures constitute environmental risk, says Joachim Hallmayer, associate professor of psychiatry at Stanford University and lead investigator of the 2011 study. “At the end of the day, we have to find the genes and environmental factors that impact the development of the disorder.”


1. Colvert E. et al. JAMA Psychiatry Epub ahead of print(2015) PubMed

2. Bailey A. et al. Psychol. Med. 25, 63-77 (1995) PubMed


23 responses to “Genes dwarf environment in autism’s origins, study says”

  1. Seth Bittker says:

    Take a look at the JAMA Psychiatry study that is cited here. The researchers take a number of sets of twins and observe that identical twins have a significantly higher probability of sharing an autism diagnosis. Based on a model, they then find that autism is primarily genetic. As Rachel notes other twin studies have obtained similar results using similar study designs.

    Any conclusion about the relative importance of genetics and the environment in development of autism is not warranted based on such studies. Think about it: in almost all cases twins share the same womb, same date and place of birth, same home after birth, same feeding practices and schedules, same chemical exposures, same foods, and same supplements and medical treatments. In fact in most cases for the first few years of their lives they are exposed to the same people and same activities. In essence for all practical purposes the environment of a twin is the same as his or her sibling twin. So is it any wonder if one does a twin study looking at genetic and environment factors in autism and one looks at variation in genetic make-up by categorizing by identical vs nonidentical twins but the environment is essentially the same for each twin pair that it appears that genetic factors dominate?

    • Elisabeth Whyte says:

      There are plenty of cases in the developmental psychology literature where only the environment (and not genes) predicted outcomes in behavioral genetics studies. That is why you compare both identical (monozygotic; MZ) and fraternal (dizygotic; DZ) twins. The fact that twins will share some aspects of the environment is the only way that the “Shared” environment value in the models can be accurately measured. For example, if you hypothetically tried to predict divorce rates in these models, you wouldn’t find any genetic influences at all because if one twin is from a divorced family, the other twin will also be from a divorced family regardless of whether they are MZ or DZ twins – so, environment will predict whether or not a child’s parents are divorced and not genes.

      The identical (MZ) twins are going to share close to 100% of their genes, whereas fraternal (DZ) twins came from different eggs (meaning they have the lower genetic similarity of non-twin siblings). Both MZ and DZ twins will share a prenatal environment, but both sets of twins will share different genetic risk properties. They can statistically estimate how the “shared” environment of being a twin differs from the levels of genetic risk based on degree of genetic similarity.

      So, if MZ twins (sharing higher genes, but the same amount of shared environment) with autism always have a 100% chance of having a twin with autism, but the DZ twins (sharing a lower amount of genes and the same amount of shared environment) ALSO have a 100% chance of autism if their sibling has autism, then genes will have a negligable value in the analysis. If genes didn’t play any role in the development of autism, then the twin studies would consistently find no genetic influence.

      However, we have convergent evidence across multiple twin studies as well as other types of genetic studies finding that genes play a very important role in brain development and we already understand a handful of specific gene mutations that can lead to autism symptoms. This is because identical twins have a much higher rate of both twins having a diagnosis compared to fraternal twins that have a much lower rate of both twins having a diagnosis – even though all the twins shared prenatal environments by being in the womb at the same time.

      • Seth Bittker says:

        I agree with you that genes play a role in development of autism, but the statements that are being made in much of the media and even in the article above are much stronger than this. Consider the statement, “Genes dwarf environment in autism’s origins”.

        The data underlying this twin study and other studies like it give no basis for making such an assertion because it involves the relative importance of genes and the environment and the environment is essentially constant in each twin pair. Consider some of the environmental factors which others have suggested may be relevant in the development of autism:
        1) Pollution.
        2) Pesticides.
        3) Maternal infection.
        4) Hygiene.
        5) Acetaminophen use.
        Could a twin study like this one say anything defensible about how important each of these factors are relative to genes? I would submit such a study could not. Consider pollution. Suppose you have a set of twins raised with their parents in a highly polluted environment (say Mexico City): Sofia and Lucia. Which one had greater exposure to pollution? In all likelihood the exposure is about the same and even if there was a very modest difference in exposure you would not know. In fact in all likelihood each of the environmental exposures mentioned above would be about the same for each twin pair. So based on such a study one might observe that identical twins almost always share an autism diagnosis and fraternal twins share a diagnosis much less frequently, but such a study in itself would not establish that pollution is a less important factor than genes in development of autism? To be clear, I don’t believe pollution is a more important factor than genes, I’m just using this for illustration purposes. My point is in order to do a truly effective twin study to measure the relative importance of environmental factors in autism, what you need are sets of identical and fraternal twins where the environments differ within each twin set (i.e. twins that are separated at birth). Of course most twins are not separated at birth, and perhaps for that reason I am unaware or such a study.

        • ASD Dad says:

          Adding further to the complexity is the 2015 Italian research that showed non shared environmental factors (56%) explained the individual differences in Autistic traits. (Genetic 44%).

          The 2011 Hallmayer research “A large proportion of the variance in liability can be explained by shared environmental factors (55%; 95% CI, 9%-81% for autism and 58%; 95% CI, 30%-80% for ASD) in addition to moderate genetic heritability (37%; 95% CI, 8%-84% for autism and 38%; 95% CI, 14%-67% for ASD).”

          The 2014 Swedish study “We found support for a disease etiology including only additive genetic and nonshared environmental effects. The ASD heritability was estimated to be 0.50 (95% CI, 0.45-0.56) and the autistic disorder heritability was estimated to 0.54 (95% CI, 0.44-0.64).”

          Differing pathways, differing effect … not one autism but many.

          Genetic and environmental influences underlying the relationship between autistic traits and temperament and character dimensions in adulthood.
          Picardi A

          Genetic heritability and shared environmental factors among twin pairs with autism.
          Hallmayer J1

          The familial risk of autism.
          Sandin S

        • Autism Parent says:

          A few years ago PBS ran an interesting news story called “The Ghost in Our Genes” that looked at disease in identical twins. One set of twins who were older women one developed cancer but the other did not implying that environment plays a strong role in triggering the onset of diseases. That disease is epigenetic. I think autism too is epigenetic and this gene plus environment interaction explains most of the cases of autism that we are seeing today. I don’t think one can have an honest discussion about autism without acknowledgement of environmental triggers. Even healthy, fit people have hidden mutations in their genes that predispose them to diseases so there’s more involved than genetics.

  2. Autism parent says:

    No genes do not “dwarf” environment and this is a misleading headline. Environment plays a very significant role in why the autism rates are skyrocketing. Just look a the numbers!! CDC reports that 1 in 68 American children are now affected. Genes alone could not possible account for the risk.

    Did this study even look at a range of possible post natal exposures such as pesticide residue on food; use of carpet cleaners; the presence of other household toxins or application of pesticides outside the home? I think a more accurate headline would “Autism risk is the result of a complex interplay between genes and environment”

    From the article:

    “A statistical model then produced an estimate of the relative pull of genetic and environmental factors on autism risk: Genetic influences account for 56 to 95 percent of the risk and environmental factors, such as prenatal exposure to chemicals and parental age at conception, contribute 5 to 44 percent.

    This statistical model also supports the idea that autism traits and skills are continuously distributed throughout the general population, with people who have autism at one extreme.

    However, the model does not account for interactions between genes and the environment, says Janine LaSalle, professor of medical microbiology and immunology at the University of California, Davis.

    “The reality is that there are many known complex gene-environment interactions that have major effects on shaping phenotypic traits in humans, and are beginning to be uncovered in autism spectrum disorders,” she says.”

    • Elisabeth Whyte says:

      The CDC doesn’t actually conduct diagnostic testing, so their crude estimate is incredibly misleading and isn’t really a reflection of whether or not more kids are having social communication difficulties. The CDC’s estimates only review medical records. So, all we know is that doctors are now more likely to write about these things in the medical records over time as the awareness of what autism is has increased. The CDC rates tell us nothing about all the times the doctors didn’t write down those autism symptoms on the pieces of paper. The CDC tells us absolutely NOTHING about whether the rate of the actual symptoms have changed over time because they don’t measure actual symptoms – they measure doctor’s perceptions and record keeping of those symptoms. This is how the last year’s CDC rate increase was only in minority (african american and latino) children and not in caucasian children – because the doctors got better at record keeping for those minority groups. This is also how the CDC rates vary so wildly from location to location (depending on the type and quality of records the CDC has access to in each state).

      The research doing actual diagnostic testing find that the rates are about 1% to 2% regardless of age group (adults or young children) or country of origin if you do consistent diagnostic screening and don’t rely on doctor’s written notes. In addition, the rates of “mental retardation” diagnoses have gone down as people have been given autism diagnoses instead. It’s not that kids stopped having intellectual disabilities, but we changed the words that we use to categorize behaviors.

      • Autism Parent says:

        The data collecting is much more rigorous than you describe Elizabeth. It is not a crude estimate. Children who are counted as having autism must meet DSM criteria according to the CDC ADDM Network; . Here is a description of Autism and Developmental Disabilities Monitoring Network. (ADDM) methodology in collecting data on autism prevalence:

        “The Autism and Developmental Disabilities Monitoring (ADDM) Network is an active surveillance system that estimates the prevalence of ASDs and describes other characteristics among children aged 8 years whose parents or guardians reside within 14 ADDM sites in the United States. ADDM does not rely on professional or family reporting of an existing ASD diagnosis or classification to ascertain case status. Instead, information is obtained from children’s evaluation records to determine the presence of ASD symptoms at any time from birth through the end of the year when the child reaches age 8 years. ADDM focuses on children aged 8 years because a baseline study conducted by CDC demonstrated that this is the age of identified peak prevalence. A child is included as meeting the surveillance case definition for an ASD if he or she displays behaviors (as described on a comprehensive evaluation completed by a qualified professional) consistent with the American Psychiatric Association’s Diagnostic and Statistical Manual-IV, Text Revision (DSM-IV-TR) diagnostic criteria for any of the following conditions: Autistic Disorder; Pervasive Developmental Disorder–Not Otherwise Specified (PDD-NOS, including Atypical Autism); or Asperger Disorder. The first phase of the ADDM methodology involves screening and abstraction of comprehensive evaluations completed by professional providers at multiple data sources in the community. Multiple data sources are included, ranging from general pediatric health clinics to specialized programs for children with developmental disabilities. In addition, many ADDM sites also review and abstract records of children receiving special education services in public schools. In the second phase of the study, all abstracted evaluations are reviewed by trained clinicians to determine ASD case status. Because the case definition and surveillance methods have remained consistent across all ADDM surveillance years to date, comparisons to results for earlier surveillance years can be made. This report provides updated ASD prevalence estimates from the 2008 surveillance year, representing 14 ADDM areas in the United States. In addition to prevalence estimates, characteristics of the population of children with ASDs are described, as well as detailed comparisons of the 2008 surveillance year findings with those for the 2002 and 2006 surveillance years.”. Source: Prevalence of Autism Spectrum Disorders — Autism and Developmental Disabilities Monitoring Network, 14 Sites, United States, 2008

        Surveillance Summaries
        March 30, 2012 / 61(SS03);1-19

  3. brian says:

    “Think about it: in almost all cases twins share the same womb, same date and place of birth, same home after birth, same feeding practices and schedules, same chemical exposures, same foods, and same supplements and medical treatments.”

    Indeed. Yet monozygotic (“identical”) twins are much more likely to share an ASD diagnosis (or, as significantly as a dichotomous ASD diagnosis, ASD-related traits) than are either fraternal twins or non-twin siblings, which have similar probabilities of developing ASD if their (dizyogtic twin or non-twin) sibling has ASD. That’s the crux of the matter, which is accurately reflected in the headline that highlights the key finding of this paper.

    • Seth Bittker says:

      I am not so sure. The headline says genetics dwarfs environmental factors. I think Colvert is making a similar claim.

      Suppose hypothetically that a major factor in development of autism is environmental exposure to acetaminophen prior to 1 year of age and relatedly the magnitude of that exposure. Suppose that genes play some role in how acetaminophen is metabolized (i.e. whether this factor is critical). Suppose also that parental tendency to provide acetaminophen is essentially the same for each of their children but varies significantly from family to family. I.e. some parents never provide it, others give only with circumcision, others give with high fevers and circumcision, others give with any fever and circumcision, others give with teething, fevers, and circumcision, others give it liberally for almost any ailment, etc. In such a world, the amount of acetaminophen consumed by 1 year of age for each twin in a twin pair would be quite similar to his or her sibling twin because consumption is largely a function of parental behavior and other aspects of the environment which are also quite similar within twin pairs.
      Suppose one does a twin study as Colvert has done without knowing any of this. Then as acetaminophen use is very similar at each twin pair level but varies between pairs, the amount of acetaminophen each twin receives will be very similar to his or her sibling twin. While this could be the key environmental factor in whether the child develops autism, the twin sets are categorized by fraternal versus identical twins. In other words, they are categorized by how much genetic material they share. As genes do have some affect on acetaminophen metabolism, in such a study the identical twins would much more frequently share autism diagnoses than non-identical twins. Based on such results one could conclude that autism is largely genetic when in fact the critical factor is environmental consumption of acetaminophen. In other words if there is any interaction between environmental and genetic factors (which seems highly likely), such a study may completely miss this and attribute the entire amount to genes as the study participants are categorized by genetic material shared.

  4. ASD Dad says:

    I was interested to note that the UK study only drew on twins born between January 1, 1994, through December 31, 1996 a particularly narrow temporal window.

    UK epidemiological studies have indicated clearly that “Following a fivefold increase in the annual incidence rates of autism during the 1990s in the UK, the incidence and prevalence rates in 8-year-old children reached a plateau in the early 2000s and remained steady through 2010.”

    Both the significant increase and subsequent plateau indicate an environmental effect. Of course an environmental insult for instance Paul Patterson’s elegant Maternal Immune Activation theory of Autism has both elements – genes and environment. Time to have a more complex reflection on genes and environment than them against us.

    BMJ Open. 2013 Oct 16;3(10):e003219. doi: 10.1136/bmjopen-2013-003219.
    Prevalence and incidence rates of autism in the UK: time trend from 2004-2010 in children aged 8 years.

    • brian says:

      It’s worth noting that changing diagnostic criteria and increased awareness of ASD among parents and healthcare providers are environmental effects.

  5. ASD Dad says:

    Also interestingly in this study there were 6423 twin pairs of which 207 gained a ASD ‘best estimate diagnosis’ giving a ‘rate’ of 1 in 31 or 3.22%

    To compare and contrast the CDC reported 1 in 125 to 1 in 150.(Birth Year 1994 – 1996)

  6. Omar says:

    Thanks for your helpful clarifications Elisabrth

  7. Planet Autism says:

    Whilst myself and both my children have autism, I still believe it is epigenetic.

    I suspect there are multiple possible toxic (man-made) environmental triggers for those with genetic susceptibility and I am very open-minded that this includes vaccinations in at least a sub-set of individuals with autism. (see P262 where autism is listed)

    “Parental Autoimmune Diseases Associated With Autism Spectrum Disorders in Offspring”

  8. RA Jensen says:

    The title of the Colvin article is the ‘heritability’ of ASD. The most ‘heritable’ of the developmentl disorders are Down syndrome and Rett syndrome. Nearly 100 percent of MZ twins in Down and Rett syndrome are concordant for Down or Rett syndrome and nearly 0 percent of DZ twins are discordant for Down or Rett syndrome. Both of these genetic syndromes are caused by sperm or egg mutations before conception but they are not ‘heritable’. Classicical twin study design cannot control for the increasing understanding of the role of de novo gene mutations in disease liability, ie Down or Rett syndrome.

  9. Jill Escher says:

    The archaic “genes or environment” viewpoint often expounded in the SFARI editorializing, as here, continues to baffle and mystify. “Genes” are not some imperturbable, sacrosanct marbles of DNA passed in kid gloves from generation to generation. The “environment” is not just something that happens during post-conception embryonic development.

    All “genes” have a history — an environmentally informed history, particularly once the entourage of the somewhat malleable epigenome is taken into account. Genomic/epigenomic history begins with the remodeling that occurs within the fetal germline, many decades prior to conception. There is no autism research that accounts for this highly sensitive earliest period of human development. We tend to focus on “genes” in a vacuum, and “environment” as it exists well after the formation of the germline.

    I would like to urge SFARI and its writers and editors to steep themselves in the reality of developmental biology, which encompasses all that molecular messiness of the early germline. There are many nasty things which messed with our genomes way back when — the idea of “spontaneous” de novo mutations and perturbations is a false, misleading assumption and historically and biologically inaccurate.

    Jill Escher (who, along with her nascent eggs, was exposed in utero to a veritable truckload of synthetic steroid hormone drugs back in the 1960s, as was a common medical practice of that Mad Men era — gosh, I wonder why the offspring of those eggs have such monstrously abnormal neurodevelopment, hm, let me think…)
    Escher Fund for Autism

  10. usethebrainsgodgiveyou says:

    Maybe someone said this in the comments, I don’t have time to read. What first stuck out is, the headline implies it is 95%, while in the article, it says chances are 62 to 95% for monozygotic, and 5-61% for dizygotic twins.

    Twin studies, in the age of the dinosaurs, when I learned about them, usually meant the children were raised in separate environments…thus “environment” was different for each one.This would be difficult, and therefore left untried. Time is money, have to get something out there…

  11. usethebrainsgodgiveyou says:

    Jill, nice to see you here.

Leave a Reply

Your email address will not be published. Required fields are marked *