Latest News

Swedish migration: Only specific groups of immigrants — those from low-income countries and those who migrated near or during pregnancy — have an increased risk of autism, suggests a new study.

Immigrating to another country during pregnancy appears to boost the risk of having a child who has low-functioning autism, according to a comprehensive, population-based study in Sweden1.

The study, which was published online 23 February in the British Journal of Psychiatry, analyzed health data from all children — nearly 60,000 — living in Stockholm county between 2001 and 2007.

It found that the risk of low-functioning autism is also greater for mothers coming from countries with a low human development index, a measure of a country’s life expectancy, income and education.

The new findings follow several small studies in Sweden and elsewhere suggesting that rates of autism are higher among immigrants. But this effort is unique in size, encompassing nearly 5,000 children with autism. It is also the first to look specifically at the timing of pregnancy in relation to immigration as a risk factor.

“This is extremely important because it reinforces the idea that it is not necessarily the region where one comes from that influences [autism] risk, but that factors associated with the migration itself do,” says Brian Lee, assistant professor of epidemiology and biostatistics at Drexel University School of Public Health in Philadelphia, who was not involved in the study.

The study’s size also allowed the researchers to independently assess the risk associated with high- and low-functioning autism, distinguished as an IQ above or below 70.

The fact that immigration factors boosted rates only of low-functioning autism might suggest that there are different risk factors for these two groups.

But the finding might also result from cultural factors, says Eric Fombonne, Canada Research Chair in Child Psychiatry at McGill University in Montreal, who was not involved in the study. For example, the link between immigration and autism with intellectual disability may reflect a cultural bias in the tests used to assess IQ rather than a true difference, he says. He points out that children with high- and low-functioning autism were considered together, parents’ immigration had no effect on risk.

Previous research has linked parents’ immigration to an increased risk of autism, but the causes behind this association remain unclear.

Some studies have found links to specific countries of origin. Among the highest profile are those linking higher autism rates specifically to children of Somali immigrants in Sweden and Minnesota.

These and other reports have prompted theories that genetic factors or vitamin D deficiency are to blame. Others have pointed toward cultural factors, such as increased use of health or social services, and hence greater diagnosis, among some immigrant groups. There is little evidence to back any specific explanations, however.

In the new study, researchers used medical and other records, including those from education and occupational therapy services, to determine autism diagnosis, parents’ country of origin and timing of immigration, as well as whether the children diagnosed with autism also had intellectual disability. Because both education and healthcare in Sweden are publicly funded, information-rich records from different agencies can be easily synthesized using national identification numbers.

About 20 percent of the children in the study had parents born outside of Sweden, most commonly in countries in East Africa, Northern Europe and Western and Southern Asia. Nearly 800 of the 5,000 children with autism were born to immigrants.

“The large sample size meant we were able to do more robust statistical analysis than previous studies were able to do,” says Dheeraj Rai, clinical lecturer at the University of Bristol in the U.K. and lead scientist on the study.

Children whose mothers immigrated during pregnancy had double the risk of having low-functioning autism compared with those whose mothers had immigrated 15 years or more before birth. 

Neither parental age nor obstetric complications explained the difference.

The increase remained regardless of country of origin and human development index, which argues against ethnicity as the sole explanation for increased risk among immigrants, says Rai. “If it was just related to ethnicity, there shouldn’t be any change in risk over time.”

While the researchers did not have enough information to study migration-associated factors that might predispose a child to autism, they speculate that maternal stress may play a role.

Animal studies show that prenatal stress can alter brain development, including in ways that mimic some aspects of autism. But research on the effects of stress during pregnancy in humans is limited.

“We really need to understand more about the context of migration for those families,” says Catherine Rice, who leads the Autism and Developmental Disabilities Monitoring Network at the U.S. Centers for Disease Control and Prevention and was not involved in the study. “Were there stressful events at the time? Are there other risk factors in those families? Illnesses or exposures that can be tracked?”

It’s also not yet clear whether the same risk profiles are seen in children with intellectual disability but not autism.

Evaluating the possible theories behind the link between immigration and autism will require extensive examination of rates of the disorder around the world.

“In order to sort out issues about prevalence and risk, we need a better understanding of how autism is distributed by geography, ethnicity or socioeconomic factors,” says Rice. Information on autism rates in lower- and middle-income countries is currently limited, especially compared with that available for wealthier countries.

One of the major limitations of this and other autism studies on immigration is the challenge of assessing the role of cultural factors. Previous research has shown that autism rates differ among ethnic groups in part because of how they access health services2. And members of some ethnic groups have more severe symptoms when diagnosed, most likely because they are diagnosed later than their peers3.

Indeed, the researchers found immigrant parents had a lower risk of having a child with high-functioning autism compared with the general population, which Rai and collaborators speculate is due to cultural factors. They say that Swedish-born parents may be more likely to seek assistance for children with mild symptoms compared with immigrants.

“When looking at the issue of immigration in relation to autism, you want to make sure that detection is equally efficient in all the different populations,” says Fombonne.

The effectiveness of diagnostic tests for autism is also an issue. Little is known about how well they work in different populations.

“The diagnostic tools used to identify autism are very Western-based,” says Rai.

Fombonne also points out that a spike in risk for women who immigrated to Sweden while pregnant may be due to the fact that women who already have a child with autism did not leave their home country or are prevented from doing so.

Rai and his collaborators now plan to look at other risk factors for autism within the same group. They also hope to further study the contribution of stress during pregnancy by examining the effect of different types of stress, such as losing a loved one or suffering from depression or anxiety.

1: Magnusson C. et al. Br. J. Psychiatry Epub ahead of print (2012) Abstract

2: Liptak G.S. et al. J. Dev. Behav. Pediatr. 29, 152-160 (2008) PubMed

3: Tek S. and R.J. Landa J. Autism Dev. Disord. Epub ahead of print (2012) PubMed

Happy helmet: Researchers have created child-friendly MRI devices that look like Homer Simpson. 

Researchers have developed functional magnetic imaging (fMRI) devices that are optimized to fit children’s heads, according to a study published in the December issue of Magnetic Resonance in Medicine1. These tools increase the quality of the data and deter head motion, which research suggests can lead to spurious results.

Several studies in the past few years have detected differences between the brains of children with autism and those of controls. But these can be confounded by artifacts, most notably head-motion in the scanner.

It is more difficult to image the brains of young children than those of adults, particularly children with autism, because they are more likely to be frightened by the process and to fidget during scans. To overcome this, some researchers sedate children while they are in the imaging apparatus, but this practice is not compatible with fMRI studies, which seek to measure brain activity while individuals watch videos or perform tasks.

In the new study, researchers constructed five MRI coils, the part of the device that detects the magnetic properties of the brain, that were optimized to fit newborns, 6-month-olds, and children aged 1, 4 and 7. These coils were designed to fit snugly around a child’s head in order to eliminate the gap that exists when researchers use standard, adult-sized MRI machines.

Well-fitting coils improve the quality of the signal and reduce noise compared with larger ones, the study found.

The new tool also offered another benefit: The tightly fitting coils — which gently grip the head and even the neck — also reduce the amount of head motion in the scanner. The average movement of a child in these devices is half a millimeter, compared with up to seven using the typical set up, reports a researcher involved in the study.

Researchers hope that the new device could prevent spurious data resulting from head motion, although this has yet to be directly tested.

The team is using a number of approaches to comfort and engage the children during the scans, such as showing them colorful videos instead of still images and decorating the top of the imaging helmet to look like Homer Simpson. They are also developing a quieter scanner, because the loud noises produced by the machines frighten some children.

The goal of all this work is to make it possible to scan children for up to 45 minutes at a time, which is 1.5 to 2 times longer than is normal without sedation, researchers say.

Future plans include using the new devices to look for early signs of the disorder by comparing brain scans of 4-month-old siblings of children who have autism with those of typical infants.

1: Keil B. et al. Magn. Reson. Med. 66, 1777-1787 (2011) PubMed

Linked risk: Ankylosing spondylitis, a form of arthritis that in severe cases can lead to fusion of vertebrae, may share a genetic basis with autism.

The genetic risk factors for autism may also increase a person’s risk of developing ankylosing spondylitis, a form of arthritis of the spine, and decrease the risk of multiple sclerosis, according to a study published 13 December in Translational Psychiatry1.

The results link the genes associated with autism and those associated with autoimmune disease and suggest that similar abnormalities could underlie both disorders.

Epidemiological studies have shown that women with rheumatoid arthritis are more likely than controls to give birth to a child with autism. Studies have also shown that women who have children with autism are five times more likely than controls to have antibodies that attack mouse brain tissue. When these anti-brain antibodies are injected into pregnant mice, they lead to social deficits in their offspring.

In the new study, researchers investigated whether autoimmune disorders and autism have a common genetic basis. They looked at the gene variants that increase the risk of autism in 941 families that have more than one child with the disorder and in genomic data from 3,000 individuals with one of a variety of autoimmune disorders, including ankylosing spondylitis, multiple sclerosis, rheumatoid arthritis and Crohn’s disease, compared with 4,500 controls.

The researchers identified all the single nucleotide polymorphisms, or SNPs — alterations to single DNA base pairs — associated with each disorder. They then compared the group of SNPs associated with autism and those linked to autoimmune disease.

There is a strong correlation between gene variants in individuals with autism and in those with ankylosing spondylitis or multiple sclerosis, the study found. Specifically, autism-associated SNPs increase the likelihood of developing ankylosing spondylitis, whereas they appear to protect against multiple sclerosis.

Autoimmune thyroid disease is also slightly associated with an increased risk of autism, the study found. However, the researchers saw no association between autism and Crohn’s disease or rheumatoid arthritis.

As in autism, ankylosing spondylitis is more common in males than it is in females, whereas multiple sclerosis is more prevalent in women, the researchers note.

1: Jung J-Y. et al. Transl. Psych. 1, e63 (2011) Article

We’ve learned a lot in the past few decades about how brain connections in children with autism go awry during early development, and the genetic and environmental factors that contribute to these changes. Veteran autism researchers often quip, “We’ve come a long way from refrigerator mothers,” referring to the notorious theory from the 1950s that cold, unaffectionate mothering causes the disorder.

In France, however, much of the psychiatric establishment has not moved on.

At least, that’s the message of Sophie Robert’s documentary film, Le Mur (The Wall), which purports that four out of every five psychologists in France follow Sigmund Freud’s psychoanalytic method and shun biological explanations for autism. Their ignorance is shocking and almost comical, except that it’s preventing thousands of children with the disorder from accessing behavioral therapies.

Over the course of four years, Robert interviewed 30 psychiatrists and psychoanalysts in France regarding their beliefs about autism, and followed one family that has two boys with the disorder and was searching for help. One of the boys, Guillaume, showed marked improvements in speech after his parents began behavioral intervention — the very approach eschewed by analyst after analyst. Autism, the analysts insist, is the result of ”la folie maternelle,“ or maternal madness, and the best way to treat it is to remove the child from her care.

The film was released online last September, prompting outrage from the psychologists who were featured in the film, as well as their colleagues. Three sued Robert for allegedly misrepresenting their views through editing. Then last month, a court ruled in their favor, censoring the film until Robert removes their interviews and pays more than 30,000 euros in damages and fees. Robert, who denies taking the interviewees’ comments out of context and says they all signed detailed releases, is filing an appeal.

The uncut film is still available online with rough English subtitles. (See this site for a handy transcript.) But even with the poor translation, it’s easy to see why these psychologists — most of whom are in their later years — are embarrassed by the way they’re portrayed.

During the interviews, several of them accused mothers of having incestuous relationships with their sons, or of wishing them dead. Some claimed that the father’s natural role in a family is to prevent a mother from destroying her child. “The mother is on the side of nature,” said psychoanalyst Aldo Naouri (who renounces the way he’s depicted in the film on his website). “She is animal, if you want, whereas the father was the one who founded culture.”

The analysts champion the writings of several 20th-century psychiatrists. The first, Bruno Bettelheim, observed that several symptoms of autism — averted gaze, shuffled gait, repeating lists of names or dates, and a “blotting out of pain” — are similar to what he saw in fellow prisoners during his 11-month experience in a German concentration camp. So autism, he argued in his 1967 book The Empty Fortress, must come from a similar kind of deprivation.

Around the same time, French psychoanalyst Jacques Lacan proposed the Freudian idea that a mother’s desire for her child is like the jaws of a crocodile, meaning she’s a suffocating influence. In the documentary, several analysts strain to apply this idea to autism (even going so far as to use a stuffed crocodile to illustrate their point). When mothers refuse to let go of the “fusion” of pregnancy, the analysts say, the child will never learn to speak or make connections with other people.

Mothers just can’t win, apparently. Don’t give a child enough attention, or give him too much — either way you’ll ruin his mind, the psychologists appear to be saying. When Robert asks them about this seeming contradiction, they cite yet another European psychoanalyst of the last century, Donald Winnicott, who wrote that mothers should be just “good enough.”

The most infuriating — and, perhaps, most dangerous — aspect of these beliefs is the way the psychologists manipulate real scientific findings to fit their theories.

For example, studies of Romanian orphans have shown that when babies are completely neglected, it stunts their physical and mental development. It’s also true that some mothers of children with autism have depression or bipolar disorder. That’s partly due to genetics, as the genes for bipolar disorder, depression and autism tend to crop up in the same families. A mother’s depression may also come from the stress of raising a child with special needs. But there’s absolutely no evidence that a mother’s neglect or her disposition causes autism.

Last year, a large research team published a 900-page systematic review of autism treatments and found no supporting evidence for the psychoanalytic approach.

The analysts admit as much. When Robert asks them how their methods could help children with autism, they say that’s simply not the point. Psychoanalysis, they insist, demands no memory, no expectation and no progress.

Some analysts in France have said that Robert chose to spotlight some of the most dogmatic proponents of Freudian theory. Even if that’s true, there’s no denying that the philosophy is widespread in the country. The Classification Française des Troubles Mentaux de l’Enfant et de l’Adolescent, France’s official reference of psychiatric disorders, listed autism as a psychosis until 2004. A barbaric treatment called ‘le packing,’ in which children are bundled in wet towels when they have a tantrum, is common. And the Council of Europe has officially chastised France for its narrow definition of autism, which is likely behind the country’s notoriously low prevalence rates.

By the end of the documentary, I had written off these analysts as clumsy dinosaurs, irrelevant in the global autism research landscape. But I was reminded of their very real impact when I saw the fate of the family’s older son, Julien, who has more severe autism and does not speak. France’s psychoanalytic culture left him, and too many others, out in the cold.

***

Virginia Hughes is a freelance journalist based in Brooklyn, New York. She earned an M.A. in science writing from Johns Hopkins University in 2006, and has written for Nature, New Scientist, Popular Science and Scientific American.

Cautionary tail: An environmental toxin that accumulates in breast milk leads to social deficits in the offspring of exposed mice.

A commonly used flame retardant may lead to deficits in sociability, learning and memory in healthy female mice and those that model Rett syndrome, according to a study published 15 February in Human Molecular Genetics1. The effects are different in Rett syndrome models compared with healthy mice, suggesting underlying gene-environment interactions.

DNA methylation, a type of epigenetic modification — chemical changes that turn genes on and off without altering the DNA sequence — likely mediates these changes, the study also shows.

Studies have identified numerous genetic risk factors for autism, but none of them account for the majority of cases of the disorder. Many researchers propose instead that a combination of multiple mutations and environmental factors could lead to the disorder. However, the molecular mechanisms underlying such gene-environment interactions are poorly understood.

In the new study, researchers investigated the effect of BDE-47 — a type of polybrominated diphenyl ether, or PBDE, which are commonly used as commercial flame retardants — on mice that model Rett syndrome, a single-gene disorder associated with autism. Studies show that BDE-47 can be found in human tissue and is present at particularly high concentrations in breast milk2.

Rett syndrome mice have a mutation in MeCP2, which detects DNA methylation and regulates the expression of thousands of genes. Because MeCP2 is on the X chromosome, a single MeCP2 mutation leads to more severe symptoms in males than in females.

Female mice with a MeCP2 mutation mated with male controls and gave birth to mixed litters of male and female mice, some of which have MeCP2 mutations. The researchers fed these mothers low levels of BDE-47 from four weeks pre-conception to three weeks of lactation, then observed the effect of the chemical on their offspring.

The resulting levels of BDE-47 in the brains of the mother mice are similar to those found in postmortem human brains, the study found. This suggests that the BDE-47 treatment corresponds with our environmental exposure to the chemical, the researchers say.

Genetically typical female pups that were exposed to BDE-47 in utero emit fewer ultrasonic vocalizations when separated from their mothers compared with untreated mice, the study found. They also show less interest in a novel mouse in their environment than in an object and take longer to learn the location of a hidden platform in a water maze, suggesting social deficits and problems with learning and memory.

Male pups exposed to the chemical have fewer symptoms than females do, only showing a reduction in vocalizations when separated from their mothers. 

However, having the MeCP2 mutation appears to worsen some and improve other symptoms of BDE-47 exposure in healthy female mice, the study found. Specifically, these mice are more likely to investigate a novel mouse than are control females that were also exposed to BDE-47 in utero, but take even longer to learn the location of the water platform. These results suggest that the MeCP2 mutation modulates the effects of BDE-47 exposure.

The exaggerated effect of BDE-47 on female compared with male mice may be due to differences in DNA methylation, the study found. Both control and MeCP2 mutant female mice exposed to the PBDE have decreased DNA methylation in their brains compared with unexposed mice, whereas males do not.  

The combined effects of BDE-47 exposure and MeCP2 mutation could be due to elevated levels of DNMT3A, a protein that modulates methylation and plays a role in learning and memory in mice. Female MeCP2 mutant mice treated with BDE-47, but not unexposed MeCP2 mutants or healthy females treated with BDE-47, have elevated levels of the genetic message that codes for the protein.

1: Woods R. et al. Hum. Mol. Genet. Epub ahead of print (2012) PubMed

2: Costa L.G. and G. Giordano Neurotoxicology 28, 1047-1067 (2007) PubMed

Lost DNA: Healthy people have about 100 mutations in their genomes that alter gene function.

More than 250 genes in the human genome — about one percent of our genes — can be eliminated without serious health effects, according to research published last week in Science1.

The findings are the result of an exhaustive effort to catalogue ‘loss-of-function’ mutations — those that disrupt the function of a gene  — in 185 people from the 1000 Genomes Project.

Although previous genome sequencing studies have shown that apparently healthy people harbor many genetic mistakes, even sometimes possessing two flawed copies of a gene, the new study is unique in its scope.

“This is the first study to look at so many possible types of loss-of-function variants, and also the first to invest in really correcting all of the different kinds of errors that can pop up in high-throughput human genome sequencing,” says lead researcher Daniel MacArthur, who completed the work as a postdoctoral fellow at the Wellcome Trust Sanger Institute in Hinxton, U.K., and is starting a new research group at Massachusetts General Hospital.

The researchers created a catalogue of these mutations, which should help scientists interpret the results of whole-genome sequencing studies of disease. When mutations that disrupt the function of genes are identified, they’re usually fingered as candidates for the genetic culprit behind the disease under study. But knowing that certain genes can be eliminated without serious consequence will allow researchers to send those to the bottom of the list of those potentially causing the disease. 

Ever since scientists began sequencing individuals’ genomes, it became apparent that genomes from even healthy adults are full of flaws.

“The challenge has always been how to interpret that variation,” says Nicholas Katsanis, director of the Center for Human Disease Modeling at Duke University in Durham, North Carolina, who was not involved in the study. “It will help us understand the buffering capacity of the genome, the ability to tolerate deleterious mutations.”

MacArthur and his collaborators started by compiling a list of 3,000 possible loss-of-function mutations, which can range from single-letter mutations that make a non-functional protein to larger deletions that take out an entire gene. 

The data came from the 1000 Genomes Project, an international collaboration to catalogue human genetic variation. Participants are outwardly healthy adults from all over the world.

They then filtered out mutations that were due to sequencing or other errors, such as incorrect labeling of a gene. This process eliminated about half the list.

“Without applying so many filtering steps we would never have been able to generate a high-quality catalogue of loss-of-function variants for further study,” says MacArthur. “Now that we have this catalogue, we can really start to figure out what effects these variants have on human variation and disease risk.”

They found that everyone has approximately 100 of these mutations in their genome, about 20 of which affect both copies of a gene. At least one individual in the group had two dysfunctional copies of 253 genes in the genome.

“Even after going through all these filters, there are still a lot of genes that appear to be knocked out in healthy people,” says Mark Gerstein, professor of biomedical informatics at Yale University and one of the researchers on the study.

The most common of these were genes involved in the olfactory system but also included those that determine blood type and the ability to metabolize drugs.

To confirm that the most common loss-of-function mutations aren’t strongly linked to common complex diseases, researchers looked for correlations in data from the Wellcome Trust Case Control Consortium, a database of genetic information on 16,000 people who have one of seven different diseases. They found only one exception, a mutation that has previously been linked to Crohn’s disease.

The new gene catalogue will initially help researchers searching for the source of Mendelian diseases, such as cystic fibrosis and muscular dystrophy, in which disrupting two copies of a gene definitively leads to the disease, to eliminate mutations from the potentially long list of candidate mutations in an individual’s genome.

But it should also aid in the search for the genetic mutations that increase risk for common diseases, such as autism, because scientists can use the catalogue to prioritize the list of candidate mutations.

The researchers compared their catalogue of genes to known disease-linked genes, and found broad differences between the two classes. 

On average, the former “were less evolutionarily conserved and had fewer protein-protein interactions,” says Chris Tyler-Smith, head of the Human Evolution team at the Wellcome Trust Sanger Institute and a researcher involved in the study. That finding is not surprising; one would expect that genes that can be knocked out in people without serious effects would play a less central role in the cell, and that such genes would not be under strong evolutionary pressure.

While these criteria can’t be used to sort out expendable genes from those that may cause disease, the findings should help researchers assess the importance of newly identified loss-of-function mutations.

The team now plans to look at the full dataset of the 1000 Genomes Project, due to be completed this year, and aims to find additional loss-of-function mutations. They will also look at mutations outside the protein-coding regions of the genome, which comprises 99 percent of our DNA. Such mutations can influence gene expression and other cellular processes.

One of the limitations of the data from the 1000 Genomes Project is that participants were not phenotyped. So while they were broadly defined as healthy adults at the time they donated DNA, some individuals may have had minor health problems or gone on to develop medical issues.

“In the longer term, we would like to be able to go back to people carrying loss-of-function variants and look for subtle effects on their phenotype, health-related or otherwise,” says Tyler-Smith.   

Most of the loss-of-function mutations identified in the study were rare, found in less than two percent of individuals in the group. This suggests that they were prevented from becoming more common by natural selection, as a result of having mildly or even seriously harmful effects, says MacArthur. Indeed, rare loss-of-function variants may be “where most of the action is in terms of effects on disease risk,” he says.

MacArthur is planning to create a custom genotype array — a tool designed to quickly detect specific mutations — to cost-effectively assess loss-of-function variants in people with specific diseases and controls.

“If there are loss of function variants playing a major role in these diseases, this approach should be able to track them down,” he says.

A second limitation of the new catalogue is that it doesn’t identify mutations that may only cause problems when combined with a second genetic mistake.  

“Mutations do not exert their effects in a vacuum but rather in the context of the entire genome,” says Katsanis. Some cases of autism have been linked to mutations in two or even three genes. “If studied in isolation, we might conclude [that these loss-of-function mutations] are tolerated,” he says. But it’s possible that “it’s only when they come together that they give rise to a clinical phenotype.”

1: MacArthur D.G. et al. Science 335, 823-828 (2012) PubMed

Behavioral therapy can have a profound effect on children with autism, and the earlier it is started the better. But how effective is it in adults?

The answer is a mystery. A new meta-analysis shows a severe paucity of research into interventions for people age 20 or older.

Analyzing 148 studies of interventions for autism from four journals — Autism, Focus on Autism and Other Developmental Disabilities, Journal of Autism and Developmental Disorders and Research in Autism Spectrum Disorders — researchers found that the average age of participants ranged from 4 to 8 years old. Only 1.7 percent of participants were aged 20 or older.

The researchers point out that the study has limitations; it analyzes data from only four journals, for example. But the findings do fall in line with previous estimates, which suggest that adults with autism receive almost no behavioral therapy.

In general, very little is known about the course of autism in adults, though that is starting to change. More studies are following children into adulthood, or trying to recruit adults with autism, in order to better understand the evolution of the disorder.

Those studies highlight just how important it is to find effective therapies that can be used beyond childhood. As much as 75 percent of adults with autism are jobless, and even those with milder forms of the disorder are unlikely to marry or hold down jobs.

Fortunately, the research community is starting to recognize the problem. Last fall, the Autism-in-Older-Adults Working Group outlined six research priorities. One of these is to study whether drug treatments and behavioral interventions have the same benefits in children and adults.

Early research suggests there could be a big difference. Drugs can have different effects depending on age, and a study published in December found that Prozac, which has no benefit in children with autism, improves some symptoms in adults.

Assessing the effectiveness of behavioral interventions, such as efforts to teach social skills, in adults is a good first step, but common sense suggests that adults will need therapies specially tailored to their needs. And that will require research. Let’s hope that analyses such as this one will draw attention to the current dearth and attract more researchers into this area.

European Journal of Human Genetics

Gene splicing: A deletion within an autism-associated chromosomal region may predispose the DNA to recombination, resulting in a neighboring deletion.

Twin brothers who have autism carry a non-inherited duplication that overlaps with 16p11.2, a chromosomal region that has been associated with autism. Their brother, who also has autism, has a nearby deletion within 16p11.2, according to a study published 11 January in the European Journal of Human Genetics1.

The results highlight the genetic diversity of autism and that too many or too few gene copies can lead to neurological disorders.

Duplications and deletions of chromosomal regions spanning several genes, also called copy number variations (CNVs), have been associated with autism and other psychiatric disorders. One of these, the deletion of a 600-kilobase region on chromosome 16, called 16p11.2, is present in about one percent of people with autism.

About 30 percent of people with the 16p11.2 deletion have autism, according to one study. The deletion is also associated with intellectual disability and obesity. Duplications of the same region have been linked to schizophrenia.

In the new study, researchers identified chromosomal alterations that overlap with the 16p11.2 region in a pair of identical twins and their older brother, all of whom have autism.

The eldest boy carries a small deletion of a part of 16p11.2 that has been linked to obesity. He was diagnosed with autism at age 4, is obese and has unusual facial features, or dysmorphology.

He inherited the deletion from his now-deceased father, who does not have a psychiatric evaluation but is described by his wife as having been introverted and having had few friends.

The twins have a duplication, called 16p11.2p12.2, that spans part of 16p11.2. Previous studies have identified six boys who have a deletion of this region and a developmental disability, but not autism2,3. A duplication of this region has also been identified in three boys who have autism, but the boundaries of their duplications are unclear, the researchers say4.

The twins were diagnosed with autism at 3 years of age and have similar symptoms, including intellectual disability and dysmorphology.  

The sites at which the twins’ DNA is rearranged suggest that the father’s 16p11.2 deletion may predispose the 16p.11.2p12.2 region to duplication during meiosis, a type of cell division in which chromosomes separate. Further research is needed to confirm this, but it suggests a genetic connection between the father’s and sons’ variants, the researchers say.

1: Tabet A.C. et al. Eur. J. Hum. Genet. Epub ahead of print (2012) PubMed

2: Ballif B.C. et al. Nat. Genet. 39, 1070-1073 (2007) PubMed

3: Hempel M. et al. Am. J. Med. Genet. A. 149A, 2106-2112 (2009) PubMed

4: Finelli P. et al. J. Med. Genet. 41, e90 (2004) PubMed

Pointing practices: Translations of autism questionnaires must take into account language and cultural differences. People in some cultures do not point with the index finger to signal joint attention, for example.

As autism awareness increases around the world, translations of English-language autism screening and diagnostic instruments are proliferating. For example, nine translations of the developmental, dimensional and diagnostic interview, or 3Di, are underway or complete. The Modified Checklist for Autism in Toddlers (M-CHAT) exists in at least 40 different languages.

Yet despite the increasing availability of such translations, there is little research evaluating how well they work. Validation is necessary not only to ensure that children with autism get the services they need, but also to accurately measure the disorder’s prevalence in different countries, researchers say.

To be able to compare autism rates, “you’ve got to know that what you’re looking at is approximately the same thing in every culture,” says David Skuse, professor of behavioral sciences at University College London and one of the developers of the 3Di.

The most widely used autism questionnaires and diagnostic instruments were developed in English — the global lingua franca of science and medicine — and are based on the English-language definition of the disorder enshrined in the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases.

But English idioms can be difficult to render in other languages. For example, the Social Responsiveness Scale (SRS), a 65-item questionnaire designed to identify autism-related traits in children age 4 to 18, contains a question about whether a child ‘separates easily from caregivers.’

“That item was totally rubbish in German,” says Sven Bölte, professor of child and adolescent psychiatry at the Karolinska Institute in Stockholm, Sweden, who translated the test1. His team tried several different ways of wording the question, but none encapsulated the idea as neatly as the English phrase. When Bölte made a Swedish translation of the SRS, he had similar results. “The question did not help to differentiate at all between autism and non-autism, but it actually did in the U.S.”

Some questions are lost in translation, but a different sort of linguistic problem crops up as well: In Korea, children show deference to their elders by adding different suffixes to words depending on whom they’re speaking to. “One of the things that parents consistently report as an impairment associated with their child’s autism is an inability to change the language to fit the status of the person that they’re speaking to or answering,” says Roy Richard Grinker, professor of anthropology at George Washington University in Washington, D.C. Yet such suffixes don’t exist in English, so autism diagnostic instruments simply don’t ask about them.

Often, translation challenges are more cultural than linguistic, however. Two questions on the M-CHAT, a 23-item screening questionnaire, ask whether a child points with his or her index finger — an important marker in Western cultures of joint attention, which is often impaired in children with autism.

“There are places in the world where people don’t tend to point with their fingers,” notes Grinker, who has collaborated on studies of autism in various cultures. In some cultures, people might ‘point’ with the palm of their hand, or by pushing out their lips. A literal translation of the M-CHAT wouldn’t work in those places.

The 3Di, a computerized assessment composed of some 300 questions, includes several items concerning whether a child shows social initiative, such as inviting other children over to play, which have been difficult to capture in Cantonese and Mandarin translations. “It’s no good saying, ‘Do they invite little kids round to play?’ because that’s just not done” in those cultures, says Richard Warrington, 3Di software developer and training manager at University College London.

He suggests that in the future, the team could include possible substitutes for problematic questions in translated versions of the 3Di, then gather data on how well these substitutes track with other autism symptoms.

These complexities mean that translating an autism diagnostic instrument also requires adapting it to a new culture. “A person making a translation should have a lot of experience with autism diagnosis,” says Young-Shin Kim, associate professor in the Child Study Center at the Yale School of Medicine. “They should be fluent in both languages. And they should be fluent in both cultures.”

Kim has made Korean translations of several autism-related questionnaires and used them as part of a study of autism prevalence in South Korea. Working on the translation of the Autism Diagnostic Interview-Revised (ADI-R), Kim realized that questions about whether a child mixes up the pronouns ‘I’ and ‘you’ would have little meaning in the Korean context.

“We don’t use these pronouns very often in Korean,” she says. But questions about saying goodbye might get at the same issues, she thought.

People with autism tend to have difficulty putting themselves in another person’s shoes, which is known as perspective-taking. In spoken English, this manifests itself through the reversal of pronouns. In Korean, saying goodbye requires perspective-taking, because the proper phrase to use depends on whether the person being spoken to is staying or leaving.

Kim worked with Cathy Lord, a developer of the original, English version of the ADI-R, to tweak the scoring system in the Korean version so that questions about saying goodbye would count as measuring perspective-taking.

Despite wide agreement that translation is a complex and tricky endeavor, there’s very little published research on the validity of translated autism questionnaires.

Recently, Skuse and Warrington collaborated with a team of Thai researchers on such a study. They administered the Thai version of the 3Di to the parents of three groups of children: those with autism, those with other neurodevelopmental disorders and those with typical development. Then they compared the patterns of Thai scores with those in a large database of results from people in the U.K. who took the original English version.

“The results are very similar to the original interview, despite the difference in culture and the considerable difference in language,” Skuse says. Full results of the study are slated to be published in Autism later this year. Skuse adds that his team has similar unpublished data on the Finnish version of the 3Di.

Working in Germany, Bölte and his collaborators found that, aside from the question about how children separate from caregivers, the German translation of the SRS yields similar results to the English version administered to children in the U.S.

More recently, Bölte has shown that a German translation of the adult version of the SRS also discriminates between those with and without autism2. However, the difference in scores between the two groups is much narrower among adults than it is among children, which could indicate the adult version of the test isn’t as accurate as the child version, or could simply be a result of the study’s small size. It had 245 participants, including just 20 adults who have autism, and Bölte says a much larger study is necessary.

Other results aren’t as reassuring. One study in Sri Lanka found that a Sinhalese translation of the M-CHAT has a specificity of only 70 percent and a sensitivity of just 25 percent — meaning that it identified some children who didn’t turn out to have autism, and missed even more children who did3. In English, the M-CHAT’s sensitivity and specificity are well above 90 percent.

As autism prevalence studies get underway around the world, the need for accurate questionnaires increases. One difficulty is securing funding for validation of translations, especially for ‘small-market’ languages, says Bölte. “There’s extremely little funding for development of psychological instruments.”

Nevertheless, more researchers are paying attention to the complexities involved in translation says Grinker, who reports that his perspective as an anthropologist is more and more in demand. “There’s increasing appreciation for just how challenging it is to employ these instruments in other cultures.”

1: Bölte S. et al. Autism Research 1 354-363 (2008) PubMed

2: Bölte S. J. Autism Dev. Disord. Epub ahead of print (2011) PubMed

3: Perera H. et al. J.Trop. Pediatr. 55 402-405 (2009) PubMed

Pathway analysis: A receptor that regulates signaling at synapses, the junctions between neurons, has abnormal protein interactions in mice that model fragile X syndrome.

Researchers have identified a new mechanism that may underlie a runaway cell-signaling pathway in fragile X syndrome. The results were published 22 January in Nature Neuroscience1.

Fragile X syndrome is caused by a mutation that inactivates the fragile X mental retardation protein, or FMRP, which regulates gene expression by stalling the machinery that translates genetic messages into proteins. Because FMRP silences many genes that are activated by metabotropic glutamate receptors, or mGluRs, researchers are developing fragile X syndrome treatments that block mGluRs.

However, studies also suggest that mGluR activity may be enhanced in individuals with fragile X syndrome through a separate mechanism involving FMRP that is unrelated to its role in regulating gene expression.

In the new study, researchers show that in male mice lacking FMR1, the fragile X gene, mGluR binds preferentially to an altered version of a scaffolding protein, called homer1a (H1A). H1A is a short version of the homer1 protein that permanently activates mGluR signaling. 

Homer proteins connect mGluRs with other cell-signaling molecules at the synapse, or junction between neurons, most notably the SHANK3 protein. A SHANK3 mutant disrupts the protein’s ability to bind to homer proteins, leading to autism symptoms in mice.

Deleting the H1A gene in fragile X mice restores mGluR binding to the longer homer proteins and normalizes many of the consequences of overactive mGluR5 signaling, the study found. These include excess protein synthesis in the brain.

Mice that lack both H1A and FMR1 also have fewer seizures and exhibit more typical anxiety levels than FMR1 mutants alone, which tend to be less anxious than controls, the study found.

Blocking the interaction between the H1A protein and mGluR also restores mGluR signaling in fragile X mice to more typical levels.

However, other features of fragile X syndrome are not recovered when H1A activity is inhibited, the study found. These include enhanced long-term depression, which is a dampening of signals at the junctions between neurons after they fire.

This suggests that H1A binding to mGluRs only partially accounts for the role the receptor plays in fragile X syndrome.

However, compounds that modulate the interaction between mGluR and H1A may still be potential therapeutics for fragile X syndrome, the researchers say.

1: Ronesi J.A. et al. Nat. Neurosci. Epub ahead of print (2012) PubMed

Time span: Researchers used 108 postmortem brains to map how a form of gene regulation changes from before birth to old age.

Researchers have charted patterns of DNA methylation — a chemical alteration to DNA that modifies gene expression — in the planning center of the brain from before birth to old age. The results were published 10 February in The American Journal of Human Genetics1 and are also available in a searchable online database.

Results show that several genes switch their methylation status as the brain matures. These include the autism-associated gene neurexin-1, which functions at synapses, the junctions between neurons, and a number of genes that have been linked to schizophrenia. Altered methylation during crucial developmental stages could contribute to neurological disorders, the researchers say.

DNA methylation is one form of epigenetic regulation, which influences gene expression without changing the DNA sequence. Epigenetic modifications can be inherited, but also vary over an individual’s lifespan, often in response to experience.

In a study published last year, researchers show that several autism-associated genes are regulated through DNA methylation.

In the new study, researchers examined DNA methylation patterns in 108 postmortem brains ranging in age from fetal week 14 to more than 80 years. They analyzed more than 27,000 sites that are prone to methylation in the regulatory regions of about 14,500 genes in the prefrontal cortex, a brain region important for thinking, planning and social behavior.

The brains of fetuses, young children and individuals older than 10 years of age each have a distinct pattern of methylation, the study found. Fetal brains also show the greatest change over time, with the methylation marks varying at a rate of 80 percent per year in fetal tissue. However, these changes take place at fewer sites than those seen in the brains of young children and individuals older than 10: 865 sites compared with 5,506 and 10,578, respectively.

A number of these sites switch methylation states as brains mature from the fetal stage into childhood, with the majority gaining a new methyl mark, the study found.

Changes in methylation are not always followed by changes in gene expression, the study also found. And, in contrast to previous reports suggesting that methylation serves as a repressor, methylation can result in either increases or decreases in gene expression.

Using the online database, researchers can track the epigenetic modifications that underlie brain development, which may be awry in developmental disorders such as autism.

1: Numata S. et al. Am. J. Hum. Genet. 90, 260-272 (2012) PubMed

A friend of mine who has a 3-year-old son with autism says that he could sing dozens of songs before he ever spoke a word of  command. That sharp contrast between language and music ability is not unusual in children with the disorder.

Though these children often have language difficulties, a significant number appear to have exceptional musical talent, such as perfect pitch.

Given that the brain regions that process speech and song overlap, what explains this disparity in autism? A new brain imaging study seeks to explore its cause and, in so doing, shed light on some of the roots of the disorder.

Researchers used functional magnetic resonance imaging (fMRI) to measure brain activity in 12 low-functioning children with autism with a mean age of 12, and 21 similarly aged controls as they listened to a parent speaking or their favorite song.

Activity in part of the brain called the left inferior frontal gyrus, which is involved in processing both music and speech, is weaker in participants with autism than in controls when they listen to speech, but stronger than in controls when they listen to music, the study found. 

Researchers also analyzed brain structure in these participants and in an additional 24 low-functioning people with autism, by using fMRI to explore the functional connections among brain regions, and a type of structural MRI known as diffusion tensor imaging.

One theory for the language deficits in autism is a problem with the long-range connections in the brain. In the new study, the researchers found some structural differences between the two groups in various parts of the brain, but conclude that problems with the brain’s long-range connections are unlikely to account for language deficits in the people with autism.

They suggest instead that changes in local connectivity or lower-level auditory processing may account for the differences in the way the brains of people with autism respond to speech and song. For example, speech requires discrimination of quickly changing sounds, whereas the fluctuations in music are slower and therefore may be easier to process.

The specific conclusions of this kind of structural study are up for debate, especially given recent research pointing to a potential flaw in brain imaging studies of children with developmental disabilities. But the findings provide support for incorporating music into behavioral therapy.

Previous research has shown that making music a part of therapy may help children with autism learn social skills and language, but more rigorous research is needed on the topic. It has also shown that children with the disorder process emotional cues from music, despite having difficulty understanding facial emotions.

Perhaps song will offer a way for lower-functioning children with autism to better learn to communicate.

Self destruction: Antibodies carried in the blood of mothers of children with autism (top), but not those of mothers of typically developing children (bottom) attack proteins in mouse brains.

The possibility that autoimmune mechanisms are a contributing factor in autism spectrum disorders has been entertained for decades, ever since early studies suggested that individuals with autism have a family history of autoimmune disease1,2.

Much of the early data were acquired from a fairly small number of individuals with autism. However, a 2009 Danish study examined autoimmune disorders in more than 600,000 children born between 1993 and 2004, and found an association between autism and both rheumatoid arthritis and celiac disease1. In fact, the study concluded that the risk of autism more than doubles for children who have a mother with one of these disorders.

The presence of autoantibodies, which are immune proteins that mistakenly attack the body’s own cells, in both diseases raises the possibility of a relationship between maternal autoantibodies and autism. In this model, maternal autoantibodies cross the placenta and enter the fetal brain, leading to alterations in its development. A variety of data published in the past few years provide evidence that such a model is biologically plausible.

The passage of maternal antibodies across the placenta is a well-known mechanism for fetal immune protection. Maternal antibodies reach all fetal tissues, even the brain. In adults, the entry of circulating soluble molecules and cells into brain tissue is limited by the blood-brain barrier, but in recent years it has become increasingly clear that the brain is less of an immune-privileged organ than it was previously considered to be. 

In the fetus, the blood-brain barrier is not fully formed, making the developing brain vulnerable to blood-borne substances. In fact, acquired changes or genetic impairments in cognition and behavior have been shown to be a consequence of circulating brain-specific antibodies that can alter function if they gain access to brain tissue3.

The children of patients with systemic lupus erythematosus (SLE) provide compelling data supporting this hypothesis. Congenital heart block, a type of arrhythmia, and skin rash are clearly transmissible to offspring by autoantibodies that are commonly present in mothers with SLE.

The children of mothers with SLE also have a high frequency of learning disorders4. This effect has been linked to autoantibodies in mice, but not in humans, however. Intriguingly, many anti-DNA antibodies that are characteristic of SLE cross-react with the N-methyl-D-aspartate receptor (NMDAR), which is involved in learning and memory5.

We have shown that pregnant female mice harboring DNA- and NMDAR-specific antibodies have pups with abnormal fetal brain development. When the pups are born, their reflexes don’t develop as quickly as those of controls and, as adults, they have selected impairments in cognitive tasks6.

Several investigators have identified the presence of antibodies that bind to human fetal brain tissue in a subset of women who have children with autism7,8. When researchers gave these antibodies to pregnant mice and monkeys, they caused abnormal behavior in their offspring9.

In a 2003 study, researchers gave serum with anti-brain antibodies from mothers of children with autism to pregnant mice. The offspring had deficits in social behavior and motor skills, as well as cerebellar abnormalities10.

In a subsequent study, pregnant mice were given immunoglobulin antibodies isolated from the blood of mothers of children with autism. In this case, the offspring were more anxious during adolescence, had alterations in sociability and were more sensitive to noise than controls were11.

Researchers have also administered similar antibodies to pregnant rhesus monkeys. The offspring had more social deficits, increased motor activity and increased repetitive behaviors compared with offspring born to mothers given immunoglobulin from mothers of typically developing children12.

The results from these studies suggest that maternal antibodies targeting the brain can affect brain development in their offspring, resulting in altered cognition, behavior and motor skills.

By studying families in the Simons Simplex Collection (SSC), we have confirmed that mothers of children with autism are five times more likely to have anti-brain antibodies than are members of a control group consisting of healthy women of childbearing age. The SSC is a database of genetic and clinical information from families that have one child with autism, and unaffected parents and siblings, funded by SFARI.org’s parent organization.

Anti-nuclear antibodies, which are directed against the cell’s nucleus, are characteristic of many autoimmune diseases. We found that they are elevated in the blood of mothers of children with autism who also carry anti-brain antibodies compared with those who do not have anti-brain antibodies. This is consistent with the theory that autoimmunity predisposes mothers to having brain-reactive antibodies and giving birth to children with autism.

Interestingly, mothers with rheumatoid arthritis are as likely to have anti-brain antibodies as are mothers of children with autism. We are investigating whether genetic variants that have been linked to rheumatic arthritis and celiac disease are present in mothers who have both anti-brain antibodies and a child with autism.

Confirming an immune mechanism for some proportion of autism cases may help identify at-risk pregnancies, by allowing pregnant women or women planning to become pregnant to be screened for harmful anti-brain antibodies. It could eventually lead to the development of drugs that block these antibodies, thereby preventing autism from developing in vulnerable offspring.

These observations suggest that genetic studies in autism should be integrated with investigations into environmental exposures, including the maternal immune repertoire, in order to fully understand the genetic susceptibility of autism. Studying the targets of harmful anti-brain antibodies may also provide insights into disease mechanisms and pathways — which is a top priority for our future studies.

Betty Diamond is head of the Center for Autoimmune and Musculoskeletal Disorders at The Feinstein Institute for Medical Research in Long Island, New York. Lior Brimberg is a postdoctoral fellow in her laboratory. Peter Gregersen is head of the Robert S. Boas Center for Genomics and Human Genetics at the Feinstein Institute.

1: Atladóttir H.O. et al. Pediatrics 124, 687-694 (2009) PubMed

2: Keil A. et al. Epidemiology 21, 805-808 (2010) PubMed

3: Diamond B. et al. Nat. Rev. Immunol. 9, 449-456 (2009) PubMed

4: Lahita R.G. Psychoneuroendocrinology 13, 385-396 (1988) PubMed

5: DeGiorgio L.A. et al. Nat. Med. 7, 1189-1193 (2001) PubMed

6: Lee J.Y. et al. Nat. Med. 15, 91-96 (2009) PubMed

7: Croen L.A. et al.. Biol. Psychiatry 64, 583-588 (2008) PubMed

8: Singer H.S. et al. J. Neuroimmunol. 194, 165-172 (2008) PubMed

9: Enstrom A.M. et al. Curr. Opin. Investig. Drugs 10, 463-473 (2009) PubMed

10: Dalton P. et al. Ann. Neurol. 53, 533-537 (2003) PubMed

11: Singer H.S. et al. J. Neuroimmunol. 211, 39-48 (2009) PubMed

12: Martin L.A. et al. Brain Behav. Immun. 22, 806-816 (2008) PubMed

Climbing cases: An epidemiology study shows that the rates of autism rise linearly for children born between 1992 and 2003 based on their year of birth.

The likelihood of being diagnosed with autism has increased for children born each year since 1992, especially for individuals at the higher-functioning end of the autism spectrum, reports a study published 7 December in The International Journal of Epidemiology1.

The results provide a lens through which to investigate theories about the rising prevalence of the disorder. Specifically, they suggest that greater social awareness of the disorder and changes to diagnostic standards are responsible for the increase, the researchers say.

Autism diagnoses have risen significantly in the past few years: For example, studies report a 634 percent increase in autism cases in California between 1987 and 2003. The question is: Why?

Epidemiological studies suggest that increased awareness of autism and changes to diagnostic criteria account for much of this trend. Others have proposed that environmental factors may be fueling the steep rise.

In the new study, researchers looked at more than six million individuals born in California between 1992 and 2003 to investigate the relative contribution of three factors to the likelihood of an autism diagnosis: the year they were born, the year they were diagnosed and the age at which they received a diagnosis.

The study found that individuals are most likely to be diagnosed with autism around age 3. Specifically, 3-year-olds in California are 37 times more likely to receive a diagnosis than 2-year-olds. This is consistent with the age at which symptoms of the disorder become obvious.

The study also found that the later an individual was born, the higher the likelihood that they would be diagnosed with autism. The yearly increase in autism rates is linear, with each subsequent birth year carrying with it a higher risk of autism than the previous one. For example, individuals born in 1999 are about four times more likely to have an autism diagnosis than those born in 1992, and people born in 2003 have almost 17 times the odds, the study found.

This trend is more pronounced for high-functioning individuals with autism than for those with greater social and language impairments. For example, high-functioning individuals with autism born in 2002 are almost 15 times more likely to have an autism diagnosis than those born in 1992. Lower-functioning individuals born the same year have only four times the odds.

The overall number of individuals with autism diagnosed in a given year does not rise as significantly, the study found. For example, there were only four times the number of autism diagnoses in 2003 compared with 1994. This finding suggests that the year in which someone was born has a bigger effect on the likelihood of receiving a diagnosis than the year in which they were diagnosed with the disorder.  

For a theory about the rise in autism prevalence to fit the new data, it has to explain the gradual increase in rates of diagnosis over time and the skew toward diagnosing high-functioning people with autism. Greater social awareness fits these data, the researchers say. However, because the rise in diagnosis is tied to an individual’s year of birth, and not the year of diagnosis, this theory only fits if parents become aware of the disorder when their child is a certain age.

Broadening of diagnostic criteria also matches the pronounced trend toward diagnosing more children at the high-functioning end of the spectrum, the researchers say.

On the other hand, environmental factors fit poorly with the data. For environmental factors to explain the change in prevalence, they would have to increase incrementally over that time period, which would increase the number of diagnoses across all ages.

1: Keyes K.M. et al. Int. J. Epidemiol. (2011) Epub ahead of print PubMed

Building bridges: In 6-month-old babies who later develop autism, white matter bundles have more structural integrity than do those in infants who do not develop the disorder. 

The development of white matter tracts, the nerve bundles that join one brain region to another, is different in babies who go on to develop autism compared with those who do not, according to a new study.

Researchers scanned the brains of infant siblings of children with autism — who have an increased risk of developing the disorder themselves — several times during their first two years of life. The so-called ‘baby sibs’ who go on to receive a diagnosis of autism at 24 months of age have distinct brain patterns at 6 months and abnormal neural development from 6 to 24 months, according to the study. The results were published 17 February in the American Journal of Psychiatry1.

“The story is that autism is an unfolding process, not something that happens in the third trimester and then is done,” says lead investigator Joseph Piven, professor of psychiatry at the University of North Carolina-Chapel Hill. “We see the brain changing over time in a dynamic way.”

It’s too early to speculate how these specific brain changes might ultimately lead to autism, the researchers say.

But the findings add to other published and unpublished studies identifying neural signatures of autism in 6-month-olds.

The new study “was music to my ears, because it superimposes on other infant-sib studies finding all kinds of brain signatures by 6 months,” says Charles Nelson, professor of pediatrics and neuroscience at Harvard Medical School, who was not involved in the new study. Nelson is using a different imaging technology to measure baby sibs’ brain waves. “It’s making us all wish that we’d started [scanning] even younger.”

These kinds of biomarkers, especially if paired with information from genetic and behavioral screens, could potentially identify children with autism long before symptoms appear, Piven says. Early detection would ideally lead to interventions that improve behaviors.

“It gives you this sense that you can jump in during this window when the brain is unfolding and maybe make a difference,” he says.

Since 2007, Piven’s group and three other centers in the Infant Brain Imaging Study have been collecting longitudinal data on hundreds of baby sibs, including several types of brain scans, blood samples for future genetic studies and, beginning at 2 years old, standardized tests for autism.

The new study analyzed diffusion tensor imaging (DTI) scans from 92 of these infants, 28 of whom received an autism diagnosis at 24 months.

DTI measures the integrity of brain connections by tracking the movement of water molecules. In white matter, water tends to move along nerve fibers in one direction, whereas in gray matter, the tissue in which neuronal cell bodies reside, “water moves all over the place,” says Jason Wolff, a postdoctoral fellow in Piven’s lab.

This study examined 15 white matter tracts, each chosen because it had been previously linked to autism. In scans done at 6 months, 12 tracks were significantly different in the babies that would go on to develop autism than in those who would not.

More specifically, babies who would later develop autism had higher ‘fractional anisotropy,’ or FA, a measure of how strongly water moves in one direction. “A higher fractional anisotropy means the bundles are more developed and water moves more efficiently along them than across them,” Wolff says.

FA increases with age, and is typically thought to mark brain maturity2. “You’d expect it to be lower in autism,” notes Kevin Pelphrey, director of the Yale Child Neuroscience Lab, who was not involved in the study.

Still, he points out that it’s hard to interpret what a high FA means at 6 months because white matter undergoes so many changes during early development. “Just showing that there’s a difference is important,” Pelphrey says. “What more FA or less FA means at this point, I’m not sure.”

For instance, a high FA in the autism group could mean that the millions of nerve fibers that make up the white matter bundles are not getting pruned appropriately at 6 months.

One major caveat in the new study is the lack of a control group of babies whose older siblings do not have autism. Pelphrey’s studies suggest that baby sibs have unusual brain signatures even if they don’t go on to develop autism.

Another important factor is time: The brain changes rapidly in the first few years of life. For example, the study found that at 24 months, toddlers with autism have lower FA values than those who do not. Piven plans to continue to scan the same group of children until age 3, because other baby-sib studies have shown that some children lose their autism diagnoses between age 2 and 3.

These abnormal trajectories of brain development only come to light when studies follow children over many years, Pelphrey notes. “That seems to be the story in autism, over and over again,” he says. “At this point, I think all imaging studies should be longitudinal or we shouldn’t be publishing them.”

1: Wolff J.J. et al. Am. J. Psychiatry Epub ahead of print (2012) Abstract

2: Gao W. et al. AJNR Am. J. Neuroradiol. 30, 290-296 (2009) PubMed