The paper, which is free, is here: Genetic Heritability and Shared Environmental Factors Among Twin Pairs With Autism
It's a twin study, and like all such research, it aims to estimate heritability, the proportion of the variability in autism risk caused by straightforward genetic effects. A heritability of 0% means no genetics and 100% means purely genetic. Note, however, that complex interactions between genes, epigenetics, and gene-environment interactions would throw the whole thing off.
Twin studies rely on the fact that there are two kinds of twins. Identical, or monozygotic (MZ), pairs have identical DNA, while dizygotic (DZ) twins are no more alike than any other brothers or sisters, genetically. So MZ twins ought to be more alike than DZ twins (have a higher "concordance"), and the size of the MZ-DZ difference is a measure of heritability.
There have been several previous twin studies of autism, and they've tended to find a heritability of around 90%, with high MZ concordance and very low DZ. However, these tended to be small and used outdated methods of diagnosis.
The new study used California records to find all twin pairs, born in the state between 1987 and 2004, where at least one of the twins had a diagnosis of autism on the DDS register of people receiving state services for developmental disorders.
They found 1156 twin pairs. Of these, they managed to recruit and get full data from 202 pairs. They gave all these 404 kids full autism diagnostic assessments. This is not a great response rate. Parents of responders tended to be slightly better educated and more likely to be white than the non-responders.
Here's the key data: concordance was higher in MZ twins, but not by nearly as much as previous studies would predict. Putting these data into a statistical model, assuming a baseline rate of autism of 1% in boys and 0.3% in girls, found that the most likely explanation was a heritability of about 35-40% and an effect of "shared environment", i.e. family factors, of 55-60%.So. Autism's not very genetic?
Maybe. This is certainly a major study and all autism researchers need to take note. But there's some caveats.
My major concern is that the DZ concordance might be too high, because a kid might be more likely to get diagnosed with autism if their twin already had a diagnosis. Suppose you're a parent and one of your twins is diagnosed - of course you're going to worry about the other one, and start thinking, are they really so different?
Although all the people in this study were (re)assessed for study purposes, the diagnostic instruments are hardly immune to the effects of prior diagnosis. The ADI interview is based on parental report of early childhood behaviour. Parents know whether the other twin has autism. The other interview, the ADOS, is based on direct observation of the patient, so it might avoid this - but you have to score on the ADI to get a diagnosis.
This, by itself, wouldn't explain the discrepency between these data and older twin studies. But we also know that diagnoses of autism in general has skyrocketed recently. People seem to be becoming more willing to accept that diagnosis, and more aware of the symptoms. So it's quite possible that some of the "unaffected" twins from older studies would get a diagnosis today if they were to have the kind of modern, formal assessment done in this study.
This doesn't mean that the new study is wrong. If this explanation is true, then the study is quite right - there is a strong shared environmental influence on autism diagnosis. But not necessarily on autism.
One reason to suspect that this is going on - and this is purely a hunch - is that the estimates of shared environmental influence, i.e family environment, was 55%. This is exceptionally high, because almost every other human disorder or trait for which twin studies have been done, have reported low shared environmental effects, and high individual environmental effects (smoking, alcoholism, anxiety, depression). In fact people have written books about this.
Maybe autism's different. Yet I'm more willing to accept that autism diagnosis is different.
A related, but seperate, point: it's very likely that some autism is more genetic than others. In particular we know that some cases are caused by single genetic variants, and these tend to be severe with associated low IQ and sometimes other abnormalities; this is sometimes called "syndromic" autism.
It's always easier to spot a severe case than a mild one. So it's quite possible that older studies had a higher proportion of these cases, because the diagnostic system was only able to pick up those ones. Maybe in more recent times, as diagnosis has expanded, "autism" is coming to cover a "less genetic" set of things.
The good thing about these data is that they span births from 1987 to 2004. So it would be possible to check this theory by looking to see whether the early data i.e. the older twins, have a higher heritability.
Finally, Michelle Dawson pointed out on Twitter that there's another large twin study from Wisconsin, as yet unpublished but presented at a conference. They found broadly comparable results.
Joachim Hallmayer, et al. (2011). Genetic Heritability and Shared Environmental Factors Among Twin Pairs With Autism Archives of General Psychiatry
21 comments:
"Suppose you're a parent and one of your twins is diagnosed - of course you're going to worry about the other one, and start thinking, are they really so different?"
That is definitely the case for parents of MZ twins but I'm not so sure about parents of DZ twins. In my experience, parents of DZ twins tend to view them more as individuals rather than as a set - especially if they aren't the same gender.
But then again, many parents of same sex twins don't know what type of twins they have and assume that whatever the ultrasound tech or doctor told them was true. We know quite a few families that are convinced that their twins are of a certain type because of what a tech told them and have never actually had their twins tested to find out for sure.
On a related note, it would be nice if one of these twin studies were able to to separate out MZ twins based on their configuration in the womb. Perhaps MZ twins that are dichorionic-diamniotic would have a lower rate of comorbidity than twins that have more sharing in the womb. Di-di twins are often misidentified as DZ twins by medical staff so I have to wonder if some of the past results where genetic testing wasn't done might have put some of the MZ twins in the DZ group.
It would also be nice to see a study that looked for shared genetic modifications and broke out the data by type of twin and prenatal configuration. For example, my MZ twin daughters were di-di, both have autism, but have different CNVs. I have to wonder if they have different CNVs because they were more like DZ twins and shared less resources in the womb or whether other types of MZ twins can have different CNVs.
The explanation of the definition of heritability in this interesting post is correct, but the concept and meaning of heritability is subtle, and there are puzzles.
Consider the number of fingers on normal human hands. This trait is definitely genetic and not environmental. A study of the heritability of the number of fingers, however, would produce a result of zero. There is no heritability because there is no variation.
Consider replicating the California twin study among normal twins and asking, “What language did the twins learn?” Here is a trait that is not genetic, but definitely environmental. Nevertheless, the heritability will be zero because there will be no variation. All of those children would learn English. In this case, there will be no variation between identical twins, between fraternal twins, or between non-twin siblings. If you were then to internationalize this study, you would find that the heritability was no longer zero because children from different countries learn different languages. Thus an environmental effect, the country of birth, would show up as non-zero heritability.
I am going to get repetitive, but which 'autism'?
DSM autism? I'll be happy to believe it. The criteria are such a hopeless jumble of assumption,observational pollution, circular logic and plain conformation bias that you can get any result you want.
The study is objectively meaningless, but vaccinists will have a field day.
LCB: That's quite true. I do think that the heritability of autism is a perfectly valid thing to measure, though. It's not one of those extreme cases.
Or to put it slightly less strongly, if it makes sense to measure the heritability of anything, then it makes sense for autism.
Several observations. The single gene disorders with high rates of co-ocuring autism such as Rhett Syndrome, Downs Syndrome, Tuberous Sclerosis, Williams Syndrome and others with the exception of Fragile X Syndrome are genetic but in the vast majority of cases not inherited.
One example:
http://ghr.nlm.nih.gov/condition/williams-syndrome
The NIH genetic resource is invaluable since the articles are generally written by leading experts in the specific single gene disorder.
Second. It raises an important question about classical twin studies and calculating 'heritability' estimates and whether classical twin study design has an inherent bias that will always overestimate hetitability..
Genetic papers in autism always reference the small European autism studies and usually state the autism is the most heritable developmental disorder with a 90% heritablity estimate. Not true. The concordance rates for Downs syndrome in MZ twins is 98% and the concordance rates for Downs Syndrome in DZ twins is 2% which should make Downs Syndrome the most heritable developmental disorder based on calculating heritabilty estimates based on the difference between concordance rates of MZ and DZ twin pairs..
Of course the problem is that Downs Syndrome is not heritable at all with 99% of the cases being the consequence of a germ line reproductive error (egg or sperm) and the Trisomy 21 mutation is not present in either parent.
Third. The lead author of the twin study is a behavioral geneticist at Stanford University with impeccable academic and institutional credentials. Joachim Hallmayer serves on the editorial board of the science journal Psychiatric Genetics and is a member of several large institutional consortiums that are engaged in genome research to identify autism 'susceptabilty' genes.
Hallmayer must have been surprised at these unexpected reults given his background as a behavioral geneticist.
Autism is not a universal single gene inherited disorder but is a complex multifactorial disorder.
This study, given the lead authors background might result is a paradigm shift in understanding the complexity of autism etiolgy and perhaps lead to interest in understanding the causes of these single gene and de novo copy number variations and environmental factors that may one day lead to at least some prevention measures.
The conclusions of this study seem overstated for several reasons and are very likely to be widely misinterpreted. First, the authors find that the concordance of MZ twins is about twice that of DZ twins - this seems to me to fit perfectly with high heritability and very little shared environmental effect. How their modeling leads to a different conclusion is not clear to me.
The major problem with the modeling is that it is based completely on a multifactorial threshold model for the disorder - the idea that autism arises when the collective burden of individually minor genetic or environmental insults passes some putative threshold. Sounds plausible, but there is in fact no evidence - at all - that this model applies to autism. In fact, it seems most likely that autism really is an umbrella term for a collection of distinct genetic disorders caused by mutations in separate genes, but which happen to cause common phenotypes (or symptoms). If that is the case, then calculating population heritability for "autism" as a whole actually becomes pretty meaningless. No one would do such a thing for mental retardation, for example, which we know can be caused by single mutations in a large number of distinct genes.
@Kevin Mitchell -
If the autism really is acting as a catchall for a number of different genetic mutations then why haven't more of them been found? There have been a sufficient number of studies with a large enough sample that we should have found more mutations that we have - even if we were talking 100 different mutations that can lead to autism.
Instead what we have found after evaluating thousands of subjects is a few shared mutations that are extremely rare and some mutations that might be related but have appeared in only one or two people.
So unless we are talking thousands of distinct mutations that all cause the same cluster of symptoms - a very dubious prospect - I think it is safe to say that pure genetic model of autism or autisms is dead. This twin study is just the final nail in the coffin of a theory whose time is past.
The classical twin study design which calculates a 'heritability' estimate based on the differences between concordance rates in MZ and DZ twins has a pround logical flaw. It assumes that genetic = inheritance. That profound flaw will always overstate heritablity estimates in conditions like autism which most investigators view as a complex multifactorial disorder.
To the behavioral geneticist (genetic determinist) environmental factors in autism is an irritant to be ignored.
In the single gene disorders and in the latest hot topic, de novo copy number variations, associated with higher rates of co-ococuring autism than is found in general population norms the majority of cases are not inherited but are the consequence of a germ line reproductive error (egg or sperm) and the mutation is not present in either parent.
There are named syndromes (tuberous sclerosis, Williams Synrome, 16P syndrome) where a minority of cases does involve transgenerational inheritance. The problem here is that the parent who has transmitted the genetic mutation is almost always unaffectd as far as autism is concerned.
Sir Michael Rutter, widely known as the father of modern child and adoloscent psychiatry for many years specialized in autism and served as the European editor of the Journal of Autism and Developmental Disorders. He gave a keynote speech at the 2009 gene-environment conference at Duke University.
Among the many statements he mad is that the behavioral geneticists are on their way to extinction.
Here is the app. 30 minute video of his keynote speech:
http://www.childandfamilypolicy.duke.edu/events/conferences/geneinteraction/presentations/videos/1SirMichaelRutter.mov
well, the link to Michael Rutters keynote on gene-environmnet interplay is broke. Anyone interested can Google:
'Duke university sir michael rutter 2009 gene environment conference'
MJ, we have already found over 100 mutations that can lead to autism. (See Betancur, C: Brain Res. 2011 Mar 22;1380:42-77. Etiological heterogeneity in autism spectrum disorders: more than 100 genetic and genomic disorders and still counting.)
These currently collectively account for about 25% of cases. And we have so far only looked for the most obvious mutations - mainly big deletions and duplications. Whole-genome sequencing is now revealing many more point mutations linked to the condition.
That doesn't discount any environmental influences of course, but the overall evidence has not strongly supported a major role for them. This current study comes up with pretty much the same figures for twin concordance as several other recent ones, but for some reason derives a large environmental parameter from those data. I can't see why they differ in their conclusions from the other studies, which all reinforced the large genetic component.
I wonder if the shared environmental effect may be a proxy for advanced paternal age effects?
Autism is one of a number of conditions strongly related to advanced paternal age, presumably because new mutations are more common in older men's sperm than in younger men.
If you use reasonable epidemiological models based on the assumption that age related cases are statistically attributable to first generation mutations, attributing 55% of autism cases to first generation mutations, and most of the rest to mutations which the parent inherited from a parent, isn't a bad fit. MZ and DZ twins, unlike non-twin siblings, both have parents of precisely the same age, while non-twins don't.
Interestingly, in this scenario:
(1) autism is still almost completely genetic even though it is not almost completely hereditary, since the definition of hereditary used in this case excludes mutations that arise for the first time in germline cells, and
(2) this reading would greatly reduce that possibility that neo-natal factors (e.g. SSRI use during pregnancy) are involved, which could be the next vacines cause autism drama if a paternal age mutation rate effect has not been properly considered.
Hi Kevin Mitchell -
The major problem with the modeling is that it is based completely on a multifactorial threshold model for the disorder - the idea that autism arises when the collective burden of individually minor genetic or environmental insults passes some putative threshold. Sounds plausible, but there is in fact no evidence - at all - that this model applies to autism.
I'd be interested in your thoughts regarding several recent papers that seem to show increased risk of autism based on some common, and increasingly common environmental conditions; i.e., season of birth, SSRI, obesity, and diabetes; most of which showed roughly a doubling of risk of autism.
Association of autistic spectrum disorder with season of birth and conception in a UK cohort
Antidepressant Use During Pregnancy and Childhood Autism Spectrum Disorders
The Role of Maternal Diabetes and Related Conditions In Autism and Other Developmental Delays
To my mind, these risk factors would speak towards small, but real, environmental factors that can push towards an autism diagnosis. Maybe I'm not understanding your position here, but wouldn't these environmental exposures amount to minor insults?
- pD
Neurosceptic;
The referenece to the Wisconsin twin study raises an even more persisteant and unsolveable problem. The ambiguity of autism definition.
That study used various diagnostic schemes and applied them to the twins producing widely fluctuating concordance rates. Which one is the more accurate?
The controversy over the impending publication of the new DSM (DMS-5) has resulted in Allan Frances to be one of the most persistent critic of the working groups charged with publishing DSM-5.
Alan Francis was the chair and editor that produced DSM-IV in 1994 a position that he still receives modest royalties for. He isn't only criticising DSM-5he even criticizes DSM-IV which he was the editor in chief for. Here's what he has to say about his own baby:
'The DSM IV field trials were meticulously conducted, but completely failed to predict the later false epidemics in attention deficit, bipolar, and autistic disorder'
Kevin Mitchells claim that molecular genetics have identified dozens, if not hundreds of rare chromosomal mutations and common genetic variations that he would claim represent autism risk. Every gene he might claim as an 'autism' gene has also been identified in schizophrenia, ADHD, childhood communication Disorders, intellectual disability, bi-polar disorder etc. As far as the common genetic variations are concerned (MET, COMT serotonin transporter gene etc) is concerned, these common genetic variations have also been implicated in all sorts of disorders but in the majority of cases they are also associated with normal variances of human behaviors. There is no boundary between disorders and these rare and common genetic mutations and variations are not diagnosis specific.
This ambiguity and non specificity affects every epidemiological, twin, molecular genetic and environmental paper related to 'autism'.
Every study produces a Rashomon effect. Different researchers studying exactly the same data arrive at completly different interpretations of the same data depending on their own bias. I plead guilty as well.
The behavioral genecists many of whom have been appointed to the DSM-V working group on autism view autism as dimensional, but no one has yet to define the boundary between disorder and normality and probably never will. I doubt if DSM-V will be any better than DSM-IV.
But it does make for interesting and heated debate and argument.
MJ you asked:
'On a related note, it would be nice if one of these twin studies were able to to separate out MZ twins based on their configuration in the womb'.
I asked one of the principals in the California twin study group several years ago if they intended to record chorion data and was told they would, where the data is available.
Apparently they were unable to secure a large enough sample of MZ twins with unambigous chorion data to offer any further insight into prenatal environmental effects in the womb segregated by chorion type. No autism twin study has ever recorded chorion type. The East Flanders Prospective Twin Survey is the only large twin registry that from its beginning has examined and recorded perinatal chorion data. In 2,000 pairs of MZ twins they reported a 2:1 ratio by chorion type (MC=66% - DC=34%).
http://www.ncbi.nlm.nih.gov/pubmed/12537855
Its interesting because all three major twin studies the 1995 British twin study, the IAN twin study and the CATS twin study have all found about the same 60% concordance in narrowly diagnosed MZ twins (genotype concordance and phenotype concordance) and they all found similar concordance rates in broadly diagnosed MZ twins (genotype concordance and phenotype discordance). Dichorionic MZ twins do not share exactly the same prenatal environment (seperate placenta) than do monochorionic twins(single placenta).
It does raise the question of whether MC twins who share almost the same prenatal environment (single placenta) are more similar than DC twins who share a more dissimilar prenatal environment (seperate placenta).
Several related studies that have examined the effect of chorionicity on behavior and brain structure have found that to be the case.
Here's a few and I hope the links work:
Personality and Cognitive Ability:
http://www.ncbi.nlm.nih.gov/pubmed/7487842?
X-inactivation patterns:
http://www.ncbi.nlm.nih.gov/pubmed/8790602?
Intelligence:
http://www.ncbi.nlm.nih.gov/pubmed/11545537?
Corpus Collosum structure:
http://www.ncbi.nlm.nih.gov/pubmed/12001090?
Schizophrenia:
http://www.ncbi.nlm.nih.gov/pubmed/7481567?
Twin studies in the future should attempt to secure unambigous chorion data.
PD, the epidemiological studies you mention do indeed show some statistical evidence of increased risk of autism associated with a variety of other factors. I expect some of these are real and some may not be. The problem with these kinds of studies is that they are correlational and so you cannot establish a causal link directly from them. In some cases, two different correlated parameters may actually be caused by an unmeasured third parameter. For example, SSRI use in mothers was associated with increased risk of autism in the progeny. (First note that that study was fairly small scale and the authors end the abstract by saying: "Further studies are needed to replicate and extend these findings". in any case, the association could be because SSRI's disrupt neural development in the fetus (perfectly plausible) or because there is a genetic link between risk of depression and risk of autism and it is known that rates of depression are higher in relatives of autistic people. The authors claim to have corrected for that by comparing rates of autism in the progeny of depressed mothers who were NOT prescribed SSRIs versus those who were but one might imagine that the severity of depression would be higher among those prescribed an antidpressant. So, I would say the jury should remain out on that one.
As for season of birth, that is better replicated and also found for schizophrenia. There is a theory that it relates to maternal vitamin D levels, which may also affect neural development. Could be true. The problem in really having any confidence in these findings is that they are population averages with fairly small effect sizes. Twice the risk may sound like a lot, but it's only a 1% increase, compared with some known mutations, which increase risk by 50-fold or more.
Overall, I don't doubt that the environment - especially the prenatal environment - can play a part in the etiology of autism. It is just very hard to pin that down from epidemiological studies alone, which yield a statistical association, but no mechanistic insights.
At the moment, genetics looks more likely to give us the entry points to the underlying biology.
Kevin Mitchell said:
'The problem with these kinds of studies is that they are correlational and so you cannot establish a causal link directly from them'
The gene associations with autism are also correlational (correlation does not imply causation). Take the finding of copy number variations found at higher rates in autism, schizophrenia, ADHD etc. compared to controls.
CNV's can occur de novo or can be inherited from an unaffected parent and are also present in lower numbers in controls. But why are the parents and controls almost always unaffected as far as autism is concerned.
Auitsm is a complex multifactorial isorder. Hallmayer has called for more research into gene-environment interaction.
Autism researchers could learn a lot from AIDS researchers who have conducted some of the best G x E research. AIDS researchers have identified a single gene CCl3L1 and have discovered that lower CNV's in CCl3L1 substantialy increases the risk for infection after exposure to HIV. No one would make the claim that lower CNV's in CC13L1 is an 'HIV' gene.
http://www.ncbi.nlm.nih.gov/pubmed/15882607
http://www.ncbi.nlm.nih.gov/pubmed/21209899
RAJ, you are right that the genetic associations with autism are correlational in a statistical sense (as all such genetic data are). But the causal chain is much more clear (or at least much more likely). Mutations cause disease (not the other way round, except for cancer). You can question whether any particular association is real based on the strength of the stats, but the logic is pretty clear once the association is established. (FMR1 mutations cause Fragile X syndrome, MeCP2 mutations cause Rett syndrome - both of which include autistic symptoms. You can simply extend the logic for hundreds of other genes).
Just stating that autism is a complex, multifactorial disease does not make it so. At least, it may be multifactorial across the population but down to single, definable causes in any individual.
For more on these particular papers, see:
http://wiringthebrain.blogspot.com/2011/07/environmental-influences-on-autism.html
Kevin you wrote:
'(FMR1 mutations cause Fragile X syndrome, MeCP2 mutations cause Rett syndrome - both of which include autistic symptoms'.
Fragile X, Rhett's, Downs etc are single gene mental retardation syndromes with high rates of co-occuring autism that is true.
The increasing rates of co-occuring autism in the single gene mental retardation syndromes has its roots in the introduction of DSM-IV which the editor in chief of DSM-IV has stated that DSM-IV failed to predict the false epidemics of attention, bi-polar and autistic disorder.
Moss & Howlin have spent their careers in working with researchers, clinicians, therapists and families with these devasting mental retardation syndromes.
They have published a comprehensive review of the single gene disorders and the association with co-occuring autism which they describe as superficial autistic type features directly related to intellectual disability and the severity of intellectual disability.
They concluded that :
'There is a need for caution in interpreting the significance of superficial similarities between ASD and the behavioural phenotypes of certain genetically determined syndromes'.
http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2788.2009.01197.x/full
Hi Kevin Mitchell -
Agreed regarding the difficult in assigning causation from those studies. The flip side of this argument is the frequency with which follow up studies on the genetic end fail to find modifications in areas previously implicated, even when we scan thousands of genomes.
Twice the risk may sound like a lot, but it's only a 1% increase, compared with some known mutations, which increase risk by 50-fold or more.
Sure, but I guess my thoughts are that we also have to take into consideration the number of people born to mothers with these risk factors (a great number), versus the number of children born with mutations that confer such a huge increase (a very small number). If being an obese mother does confer a double of risk, that's a whole lot more people at slightly increased risk compared to exceptionally rare, and static mutations.
By the way, I've only seen one mutation that increases risk anywhere close to 50 fold. Is there more than one that causes such a distinctly steep risk curve? Do you have links to these particular mutations?
Regarding what got me interested in your comments, the likelyhood or unlikelyhood of mutliple, small environmental influences reaching a phenotypic tipping point, here are my thoughts; I'd be interested in your comments.
As I stated previously, my understanding of most of the genetic studies implicate genes with far lower risks than 50 fold; the widely replicated MET-C allele being a good example, nearly half of everyone has it, but it is consistently over represented on the autism spectrum. Some of these studies implicate disturbances of gene networks, it seems that tinkering with the genetic recipe up or down the way has a similar effect of disturbing the very delicate and tightly correographed process of brain development. But if we accept that small changes to genetics along a stream of functionally related gene products can result in autism, we also can conclude that affecting those processes by mechanisms other than gene modifications can have the same outcome. When we look towards the biological plausibility of things like inflammation, manipulating neurotransmitters, vitamin d deficiency, or endocrine disruptors to participate in disturbing brain development, I don't see much of a problem in concluding that such external changes could have the same effect as a deletion on 16p11.2. Considering the amount of environment has changed with the relative stability of our gene pool, however, leads to some difficult questions regarding how much, if any, of our apparent observations of an increase in autism is real versus diagnostic artifact.
Thanks.
- pD
http://adf.ly/2Fycf autism
PKU is an autosomal recessive metabolic genetic disorder.
Change the metabolism via environment, and you change the trajectory of the disease.
Post a Comment