If you’ve been following autism research in recent years, you have probably read—many times—that familial, or inherited, risk is seldom the whole picture. A few inherited genes are sufficient by themselves to cause autism. But most so-called “autism genes” only increase the risk that an infant will go on to develop this developmental disorder. As is the case in many complex diseases, it appears that autism often results from a combination of genetic susceptibility and environmental triggers.
This is where epigenetics comes in. Epigenetics is the study of the factors that control gene expression, and this control is mediated by chemicals that surround a gene’s DNA. Environmental epigenetics looks at how outside influences modify these epigenetic chemicals, or “markers,” and so affect genetic activity.
It is important to remember that scientists use the term “environment” to refer to much more than pollutants and other chemical exposures. Researchers use this term to refer to pretty much any influence beyond genetic mutation. Parental age at time of conception, for example, is an environmental influence associated with increased risk of autism, as are birth complications that involve oxygen deprivation to an infant’s brain.
Because epigenetics gives us a way to look at the interaction between genes and environment, it holds great potential for identifying ways to prevent or reduce the risk of autism. It may also help us develop medicines and other interventions that can target disabling symptoms. We have written about epigenetics previously on this blog (here and here). So in this answer, I’d like to focus on the progress reported at a recent meeting hosted by Autism Speaks.
The Environmental Epigenetics of Autism Spectrum Disorders symposium, held in Washington, D.C. on Dec. 8, was the first of its kind. The meeting brought together more than 30 leaders in autism neurobiology, genetics and epidemiology with investigators in the epigenetics of other complex disorders to promote cross-disciplinary collaborations and identify opportunities for future studies.
Rob Waterland, of Baylor College of Medicine in Texas, described epidemiological studies and animal research that suggested how maternal nutrition during pregnancy can affect epigenetic markers in the brain cells of offspring.
Julie Herbstman, of Columbia University, described research that associated epigenetic changes in umbilical cord blood with a mother’s exposure to air pollutants known as polycyclic aromatic hydrocarbons (PAHs). PAHs are already infamous for their association with cancer and heart disease.
Rosanna Weksberg, of the Hospital for Sick Kids in Toronto, discussed findings that suggest how assisted reproductive technology may lead to changes in epigenetically regulated gene expression. This was of particular interest because assisted reproduction has been associated with ASD. Taking this one step further, Michael Skinner, of Washington State University, discussed “transgenerational epigenetic disease” and described research suggesting that exposures during pregnancy produce epigenetic changes that are then inherited through subsequent generations.
Arthur Beaudet, of Baylor College of Medicine, discussed a gene mutation that controls availability of the amino acid carnitine. This genetic mutation has been found to be more prevalent among children with ASD than among non-affected children, suggesting that it might be related to some subtypes of autism. Further study is needed to follow up on the suggestion that dietary supplementation of carnitine might help individuals with ASD who have this mutation. Caution is needed, however. As Laura Schaevitz, of Tufts University in Massachusetts, pointed out, studies with animal models of autism suggest that dietary supplementation may produce only temporary improvements in symptoms of neurodevelopmental disorders.
So what does this all mean for research that aims to help those currently struggling with autism? The meeting participants agreed that the role of epigenetics in ASD holds great promise but remains understudied and insufficiently understood. For clearer answers, they called for more research examining epigenetic changes in brain tissues. This type of research depends on bequeathed postmortem brain tissue, and Autism Speaks Autism Tissue Program is one of the field’s most important repositories. (Find more information on becoming an ATP family here).
The field also needs large epidemiological studies looking at epigenetic markers in blood samples taken over the course of a lifetime. One such study is the Early Autism Risk Longitudinal Investigation (EARLI). More information on participating in EARLI can be found here.
Autism Speaks remains committed to supporting and guiding environmental epigenetics as a highly important area of research. We look forward to reporting further results in the coming year and years.
Got more questions? Send them to email@example.com.
Read more autism research news and perspective on the science page.
Posted by Autism Speaks scientific advisory board member Irva Hertz-Picciotto, PhD, MPH. As an epidemiologist at the University of California-Davis MIND Institute, Hertz-Picciotto studies exposures to environmental chemicals, their interactions with nutrition and pregnancy and their effects on prenatal and early child development.
Alan Zarembo’s series on autism in the LA Times covers a great deal of territory and has brought to light a wide range of personal stories that exemplify the complexity of issues surrounding autism diagnosis, treatment choices and effectiveness, impact on families and population incidence. Zarembo should be commended for the substantial work he has done researching inequities in the delivery of services. Of notable concern, he has put a spotlight on what appear to be serious racial and ethnic disparities in the California Department of Developmental Services (DDS) system and the provision of therapies. If his figures are correct, this result should stimulate an analysis of how to right this situation and ensure that appropriate opportunities are made available to all families with affected children.
Zarembo has also highlighted adults living with an autism spectrum disorder (ASD) but diagnosed late in life. We have too long overlooked the struggles faced by adults with autism as they try to find ways to be productive, live independently and connect with others.
My remaining comments pertain to Zarembo’s conclusions about the rise in autism diagnoses and the role of environmental factors. First, he is right that there is substantial variability in rates of diagnoses in different regions, and that we should not confuse diagnoses with the actual incidence of disease. Not all persons who meet criteria have been correctly diagnosed, and the degree to which this is true has likely changed over time.
Nevertheless, impressions are not the same as a scientific analysis. Zarembo has not demonstrated that the rise is purely social and cultural. My colleague Lora Delwiche and I published the first quantitative analysis of how much of the increase in diagnosed cases in California could be explained by artifacts (changes in diagnostic criteria, earlier age at diagnosis and inclusion of milder cases).1 We used California state data that provided statistics over many years and found that the numbers simply do not add up. In other words, the actual increase has been far larger than these artifacts could have produced. Combining our results with those of another research team, it appears that about half of the increase in diagnoses in California is due to changes in diagnostic criteria or practices.2 These results left about a three-fold increase unexplained as of 2007. And autism diagnoses in California have continued to rise both in areas with low rates and in areas with high rates. Zarembo is interested in explaining the geographic variation, but the explanations for variation spatially are not necessarily the explanations for variation over time.
These statements were particularly misleading:
“No study points to an environmental reason for the worldwide explosion in cases over the last two decades.
Given the slow pace of genetic change in large populations, genes can’t account for the surge either.
That suggests the explanation for the boom lies mainly in social and cultural forces, notably a broader concept of autism and greater vigilance in looking for it.”
The logic that leads from the first two sentences to the third involves huge assumptions.
How many studies have been done of environmental causes? Very few! And of these, most were extremely poor studies involving very small samples or lacking individual-level data. Is it surprising we’ve uncovered few leads? The funding for environmental factors has been paltry – somewhere around $40 to $60 million over the last 10 years, while more than $1 billion has been spent on studying autism genetics. To imply that environmental factors can be dismissed and that only social/cultural factors should be pursued is nonsensical.
It should be noted, however, that if anyone is looking for “one” environmental factor to explain the increase, they will certainly be disappointed. It doesn’t exist. Autism is far too complex. Moreover, to the extent the increase is due to diagnostic differences over time, we need to find explanations both for the increasing numbers of diagnoses and for the autism that has been around “all along.” In fact, data are emerging about quite a number of environmental factors. In 2011, major papers were published supporting contributions from maternal nutrition around the time of conception (here and here), traffic-related air pollution, and season of conception.3-6 Earlier papers indicated associations with pesticides (here and here) and air pollution.7-9
One concern raised about the increase in diagnoses is a type of ‘inflation’ from inclusion of a growing number of high-functioning persons whose diagnosis is more likely to be Asperger syndrome than classic autism. This may apply to some studies of changes over time, but in our analysis of CHARGE study data, most of the cases were low functioning.10 This would likely be true for the majority of persons with ASD served by the California Department of Developmental Services (DDS), because in order to qualify for state services, they must have “significant functional limitations” in three areas of major life activities. This requirement would exclude most of those who are higher functioning.
With regard to genetics, Zarembo’s article leaves out the most current information: the largest and most statistically robust study of twin pairs found that 38 percent of concordance is due to shared genetics with 58 percent due to shared environmental factors (most likely prenatal and perinatal).11 This result completely overturns the widespread assumption, based on a number of previous small studies, that the causes of autism are overwhelmingly heritable, or genetic. Unfortunately, most analyses of twins make the incorrect assumption that genes and environment do not interact to influence risk for disordered brain development. This interaction is real, and one study has already shown a whole class of genes that primarily affect development in children whose mothers had not taken prenatal vitamin supplements during the months preceding and immediately following conception. 3
In summary, Zarembo’s investigative journalism provides unusual depth into difficult aspects of autism occurrence and the social policies that bear on the lives of affected families. He has raised several critical challenges facing the autism community. What was lacking from his series is a more balanced view of the role environment likely plays and the need to advance the agenda of discovering modifiable causative factors.
Autism Speaks is one of a few organizations that have begun to turn in this direction. I look forward to a continued strong commitment by Autism Speaks and others willing to support and significantly expand the scientific research aimed at identifying and understanding environmental contributions to autism, factors driving increased prevalence and ways to intervene so as to eliminate or lower human exposure levels. This will be the fastest road to reducing the occurrence of ASD in the next generation.
1 Hertz-Picciotto I, Delwiche L. The rise in autism and the role of age at diagnosis. Epidemiology 2009;20: 84-90.
2 King M, Bearman P. Diagnostic change and the increased prevalence of autism. Int J Epidemiol. 2009; 38:1224-34.
3 Schmidt, R J, et al. Prenatal vitamins, one-carbon metabolism gene variants, and risk for autism. Epidemiology 2011;22:476-85.
4 Cheslack-Postava K, Liu K, Bearman PS. Closely spaced pregnancies are associated with increased odds of autism in California sibling births. Pediatrics 2011;127:246-53.
5 Volk HE, Hertz-Picciotto I, Delwiche L, Lurmann F, McConnell R. Residential proximity to freeways and autism in the CHARGE study. Environ Health Perspect 2011;119: 873-7.
6 Zerbo O, Iosif AM, Delwiche L, Walker C, Hertz-Picciotto I. Month of conception and risk of autism. Epidemiology 2011;22:469-75.
7 Roberts EM, et al. Maternal residence near agricultural pesticide applications and autism spectrum disorders among children in the California Central Valley. Environ Health Perspect. 2007;115:1482-9.
8 Eskenazi B, et al. Organophosphate pesticide exposure and neurodevelopment in young Mexican-American children. Environ Health Perspect 2007;115:792-8.
9 Windham G, Zhang L, Gunier R, Croen L, Grether J. Autism spectrum disorders in relation to distribution of hazardous air pollutants in the San Francisco Bay Area. Environ Health Perspect. 2006; 114(9):1438-44.
10 Hertz-Picciotto, I. et al. The CHARGE study: an epidemiologic investigation of genetic and environmental factors contributing to autism. Environ Health Perspect. 2006;114: 1119-25.
11 Hallmayer, J. et al. Genetic heritability and shared environmental factors among twin pairs with autism. Arch Gen Psychiatry. 2011(68):1095-102.
On a day to day basis, I can get so immersed in the mechanics, data and details of what I do that I sometimes forget to step back and see the bigger picture: The tremendous value of the research information systems that Autism Speaks has created within its Autism Genetic Resource Exchange (AGRE).
I am feeling particularly sensitive to this “big picture” with this week’s announcement of the historic addition of AGRE information to the National Database for Autism Research (NDAR), which is supported and maintained by the National Institutes of Health.
I’m reminded that the goal of all my work is to increase the availability and usefulness of the vital information that we collect—information that can advance the scientific understanding of autism and speed the development of better treatments.
I promise to spare you the technical details. Suffice it to say, our role is to take the anonymous information (de-identified data) we gather from our participating families and put it in an easy-to-use format that autism researchers can use to increase the power and accuracy of their scientific findings and insights.
In joining our data with NDAR’s, we are making more comprehensive data available to the broader scientific community and also linking data collected on participants within AGRE with additional data on the same participants across a variety of other research studies (all anonymous). This is adding significantly to the autism field’s body of scientific knowledge. And this is our obligation to our families: To maximize their contributions and make sure their de-identified data will always be available to qualified scientists who are working to improve the lives of those who struggle with autism. We take pride in providing researchers with the most comprehensive and highest quality of data possible, so that they can do what they do best: science.
Special thanks go to Reinis Berzins, our AGRE data projects coordinator, whose position was made possible by the $1 million National Institutes of Health grant we received to integrate the AGRE and NDAR databases. To learn more about AGRE, please visit its website. To learn more about the AGRE-NDAR federation, please see this week’s related news item.
Read more science news and perspective on the Science Page.
Research has taught us that there’s no simple explanation for what causes autism. We know that genes play a role, but they aren’t the whole picture. Environment also matters.
However “environment” can be a tricky term, as pediatrician Perri Klass recently noted in her New York Times column. In autism research, we use the word to refer to pretty much any influence beyond inherited genes—not just exposure to pollutants or other toxic chemicals.
In fact, the environmental factors that research most strongly links to autism are influences such as maternal infection during pregnancy (especially rubella), birth complications (especially those involving oxygen deprivation), and parental age at time of conception (dad as well as mom). Parents who wait less than one year between pregnancies may be at a slightly higher risk for having a child with autism. (Conversely, there is strong evidence that mothers who take prenatal vitamins before conceiving reduce the odds that their children will develop autism.)
Clearly, countless fetuses and babies are exposed to “environmental risk factors” such as these without ever developing autism. But if a child is genetically predisposed to autism, it appears that these influences further increase the risk. For this reason, we say that environmental factors increase the risk of autism rather than cause it.
Research has suggested that many other environmental, or nongenetic, factors may increase the risk for autism. But scientists can’t yet say whether these involve direct (versus coincidental) links. Such factors include a pregnant woman’s exposure to certain chemicals such as pesticides and phthalates (commonly found in plastics) or certain drugs such as terbutaline (used to stop premature labor), valproic acid (to control seizures), and some antipsychotics and mood stabilizers. Of course, in the case of medications, any possible increased risk of autism must be balanced against a woman’s medical needs—which can likewise affect the health of her pregnancy and children.
In addition, most environmental factors associated with autism appear to increase risk only slightly and only in combination with other factors such as genetic predisposition. So it is difficult, in most cases, to pinpoint any one environmental influence. For these reasons, Autism Speaks continues to fund research on a wide range of environmental risk factors. Importantly, the more we learn about how these influences affect brain development, the better we can help the children, adults and families who are affected by autism.
Want to learn more about the research Autism Speaks is funding? On our Science Grant Search page, you can browse studies by topic and location. Finally, if you or your child is affected by autism, please consider participating in one of our clinical studies. Thanks, and please keep sending us your questions.
This is a guest blog post from Autism Speaks Science Board member John Elder Robison, author of Look Me in the Eye: My Life with Asperger’s and Be Different: Adventures of a Free-Range Aspergian.
As reported last week, a large twin study supported by Autism Speaks compared the frequency with which identical and fraternal twins both share a diagnosis of autism. This approach enabled the investigators to use statistical techniques to calculate the degree to which environmental factors shared by twins contribute to their risk of developing autism. Such factors include conditions in the womb and during birth.
The results of the California Autism Twin study were game-changing because they revealed a much larger environmental influence than had previously been estimated—accounting for about 58% of the risk of developing autism. By contrast, much smaller twin studies had previously suggested that genes largely accounted for a child’s risk of autism.
The results underscore the need to investigate the role that non-inherited risk factors play in the development of autism spectrum disorders (ASD). So what’s being done to help speed this research? And what role are Autism Speaks and its donor dollars playing in this effort?
Autism Speaks has funded over $21 million in the study of environmental risk factors, an initiative we call (obviously enough) the Environmental Factors of Autism Initiative. Already, we have a large body of evidence suggesting that it is not any one environmental factor, but many different factors working together, that elevate the risk and severity of autism in individuals with a genetic predisposition for this condition. In other words, autism is seldom caused by any one thing and neither is it an “all or nothing” condition. Furthermore, different combinations of genetic and environmental risk factors contribute to individual cases of autism.
Within the Environmental Factors of Autism Initiative are ongoing studies focusing on environmental exposures that occur before and during pregnancy and throughout the first year of life—crucial periods for human brain development. These studies look at such possible risk factors as maternal and paternal age, socioeconomic status, season of birth, exposure to chemicals or toxic agents, nutrition and exposure to various pharmaceutical drugs during pregnancy, the difficulty of labor and delivery, and various other forms of prenatal stress. The researchers we support are also investigating the mechanisms by which genes and the environment may interact (so-called epigenetics) and the role of the immune system. We are also supporting large scale epidemiological studies that focus on pregnancy and the first year of life. These include the EARLI study and the IBIS study.
Are you interested in learning more about the studies Autism Speaks is funding with donor dollars? We are proud to debut the new Grant Search function on our website. Please use it to explore past and present research studies by topic or location. And if you or your family is affected by autism, please consider participating in one of our clinical studies.
by Alycia Halladay, Ph.D, Director of Environmental Science
Research using identical and fraternal twins is typically used to identify genetic influences on the development of ASD. This year, researchers studied a large group of twins and examined the concordance of different types of symptoms (1). Using this approach, the researchers found that the concordance of severe autism between identical twins and fraternal twins was about the same, indicating a strong environmental component to ASD severity. But what are those environmental factors? Epidemiological studies are providing clues.
At this year’s IMFAR, new data was presented that focused on studying groups of people and their exposures to a number of environmental factors. Each used different designs with their own unique advantages. For example, at UC Davis, the CHARGE study (www.beincharge.ucdavis.edu) examined the risk of developing autism following exposure to a number of factors that were identified through self report or medical records. Those that showed an association were antidepressant SSRI use (2) and metabolic disorders including hypertension and diabetes (3). On the other hand, a previously identified factor, maternal infection, was not associated (4). Why not? The researchers suggested that fever, not infection per se, may be a factor. Using self-report and medical records obtained prior to study entry may not accurately capture all relevant information, and an infection or fever may be missed in some reports. However, other types of information, such as method of birth, is easier to gather accurately. An analysis revealed that non-emergency or elective c-section deliveries did not show a significant association with autism, addressing a concern that many public and community stakeholders have expressed (5).
As an alternative to retrospective reports, the Early Markers of Autism Study in California is obtaining samples of blood from pregnant women by obtaining extra blood taken during the alpha-fetal protein screen that is banked. Not all states bank these samples for research, so this is a unique resource. By examining the levels of mercury in blood taken during pregnancy together with newborn blood spots, the researchers can get a more comprehensive picture of the prenatal environment. They reported no difference in mercury levels compared to those of non-affected children during gestation, and also reported no difference in thyroid hormone levels (6,7). Examination of subgroups of autism with regression did not change the results. While these data are incredibly novel and valuable, these studies were not designed to capture information throughout the entire pregnancy nor capture factors after birth
Another way to study exposures during pregnancy is through birth certificate data. In some states, the birth certificate contains information such as the place of birth and the occupation of the mother and the father. Using this information, scientists found that occupational exposures in mothers to certain chemicals resulted in an increased risk of ASD in offspring (8).
While each approach brings unique strengths, all researchers agree that the most comprehensive way to capture all information accurately, is a prospective design. This means identifying children as soon as possible and following them from that point on to gather every piece of relevant information from medical reports to blood samples. Autism Speaks is proud to co-sponsor such a study: the Early Autism Risk Longitudinal Investigation (EARLI). This groundbreaking project will provide even more answers to what causes autism, and needs the help of the community to do so.
So how can researchers blend or expand their research if they are using only one type of design? Autism Speaks and the National Institutes for Environmental Health Sciences are sponsoring a network of projects called the Environmental Epidemiology of Autism Research Network (EEARN). The goal of this network is to improve communication among researchers in this field, identify opportunities for collaborative projects and improve research tools for both existing, and new projects. Over 20 studies from 8 countries are represented in the network. We will keep you updated on the activity of the network, and we hope you will keep checking in for updates.
1. Understanding Clinical Variability In Autism: Results From a California Twin Study. W. Froehlich*1, S.
Cleveland1, A. Torres1, J. M. Phillips1, B. Cohen2, A. Fedele3, T. Torigoe2, J. Collins4, K. S. Smith5, L. Lotspeich1, L. A. Croen4, S. Ozonoff6, C. Lajonchere7, J. K. Grether5, N. Risch8 and J. Hallmayer1, (1)Stanford University, Stanford, CA, (2)Autism Genetic ResourceExchange, Los Angeles, CA, (3)Autism Speaks, Westmont, NJ,
United States, (4)Kaiser Permanente, Division of Research, Oakland, CA, (5)California Department of Public Health, Richmond , CA, (6)UC Davis MIND Institute, Sacramento, CA, (7)Autism Speaks, Los Angeles, CA, United States, (8)University of California San Francisco, San Francisco, CA
2. SSRI Use During Pregnancy and Risk of ASD or Developmental Delay In Children. R. A. Harrington*1,L. C. Lee1, C. K. Walker2, R. L. Hansen3, S. Ozonoff3 and I. Hertz-Picciotto4, (1)Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (2)Department of Public Health Sciences, University of California at Davis, Davis, CA, (3)MIND Institute, University of California at Davis, Sacramento, CA, (4)Department of Public Health Sciences, University of California Davis, Davis, CA
3. The Role of Maternal Diabetes and Related Conditions In Autism and Other Developmental Delays. P. Krakowiak*1,2, A. A. Bremer3, A. S. Baker1, C. K. Walker1,4, R. L. Hansen2,3 and I. Hertz-Picciotto1,2, (1)Public Health Sciences, University of California, Davis, Davis, CA, (2)M.I.N.D. Institute, Sacramento, CA, (3)Pediatrics, University of California, Davis, Sacramento, CA, (4)Obstetrics & Gynecology, University of California, Davis, Sacramento, CA
4. Prenatal Influenza or Fever and Risk of Autism/Autism Spectrum Disorders. O. Zerbo*1, I. Hertz- Picciotto2,3, A. M. Iosif4, R. L. Hansen5,6,7 and C. K. Walker8, (1)Sacramento, CA, (2)University of California, Davis, Davis, CA, (3)Department of Public Health Sciences, University of California Davis, Davis, CA, (4)UC Davis, Davis, CA, (5)University of California, Davis, MIND Institute, Sacramento, CA, (6)MIND Institute, University of California at Davis, Sacramento, CA, (7)MIND Institute and Dept. of Pediatrics, University of California Davis, Davis, CA, (8)Department of Public Health Sciences, University of California at Davis, Davis, CA
5. Cesarean Birth and Autism Spectrum Disorder. C. K. Walker*1, P. Krakiowiak2, A. S. Baker3, R. L. Hansen4, S. Ozonoff5 and I. Hertz-Picciotto6, (1)Obstetrics & Gynecology, UC Davis, Sacramento, CA, (2)Public Health Sciences, UC Davis, Sacramento, CA, (3)Public Health Sciences, UC Davis, Davis, CA, (4)Pediatrics, M.I.N.D. Institute, UC Davis, Sacramento, CA, (5)Psychiatry and Behavioral Sciences, M.I.N.D. Institute, UC Davis, Sacramento, CA, (6)Public Health Sciences, M.I.N.D. Institute, UC Davis, Davis, CA
6. Prenatal and Neonatal Peripheral Blood Mercury Levels and Autism Spectrum Disorders. L. A. Croen*1, M. A. Lutsky1, C. Yoshida1, C. P. Alaimo2, M. Kharrazi3, J. K. Grether4 and P. Green2, (1)Kaiser Permanente Division of Research, Oakland, CA, (2)Civil and Environmental Engineering, Univ. of California Davis, Davis, CA, (3)Genetic Disease Screening Program, California Department of Public Health, Richmond, CA, (4)California Department of Public Health, Richmond, CA
7. Prenatal and Neonatal Thyroid Stimulating Hormone Levels and Autism Spectrum Disorder. M. A. Lutsky*1, C. Yoshida1, B. Lasley2, M. Kharrazi3, J. K. Grether4, G. Windham4 and L. A. Croen1, (1)Kaiser Permanente Division of Research, Oakland, CA, (2)Department of Population Health and Reproduction, UC Davis, Davis, CA, (3)Genetic Disease Screening Program, California Department of Public Health, Richmond, CA, (4)California Department of Public Health, Richmond, CA
8. Autism Spectrum Disorders In Relation to Parental Occupational Exposures During Pregnancy. G. Windham*1, J. K. Grether2, A. Sumner3, S. Li4, E. Katz5 and L. A. Croen6, (1)California Department of Public Health, Richmond, CA, (2)California Department of Public Health, Richmond, CA, (3)Vermont Department of Health, Burlington, VT, (4)Kaiser Permanente Divison of Research, Oakland, CA, (5)Occupational Health Branch, CA Department of Public Health, Richmond, CA, (6)Kaiser Permanente Division of Research, Oakland, CA