Pedigree Puzzle: Why is there a Gender Bias in Autism?
★ Autism Spectrum Disorders are more frequently diagnosed in males than females: commonly four times as often, although that bias climbs to 11:1 in the case of Asperger’s syndrome. The underlying reasons are complex and many plausible theories have been proposed. Let’s begin by looking at one example of pronounced gender bias in autism, seen in this pedigree chart. A pedigree chart is used by geneticists to track genotypes (such as a particular mutation) and phenotypes (such as appearance of a disease) over many generations of an extended family. If you’ve never deciphered one before, this is your chance to figure out what those squares and circles mean!
★ How to Read a Pedigree Chart: To begin, girls are circles and boys are squares – helpfully colorized to pink or blue to fit the stereotype 😉 There are four generations in this chart (I-IV), each in a separate row. Offspring from a pair are shown by the T bars: for example, the first pair (now deceased) had four children, two males and two females. One of the females produced the four children shown in generation III. Progeny from three pairs are shown in generation IV. Makes sense so far?
★ Linking Genes to Autism: Back to the Science. Researchers monitored the SHANK1 gene in ~2,600 people with autism and ~15,000 “controls”. They found a large deletion that wiped out most of one copy of the gene in four people with autism. Three were in the family shown in the pedigree chart. None in the control group had this deletion, so this was a statistically significant difference. In gene speak, we say there is a Copy Number Variation (CNV) in this gene. The Shank proteins act as scaffolds around which the synapse, or junction between nerve cells, is built. Other SHANK genes have already been linked to autism, so they used pedigree analysis on SHANK1. Six members of the family carried the CNV but surprisingly, only males with CNV were diagnosed with autism (labeled A in the chart). In case you are wondering, SHANK1 is not on the X chromosome, so the gene is not sex linked. So why are only males in this family autists even though they carry the same mutation as some of the females? This is an extreme case of gender bias in autism. Although the precise answer is not known for the SHANK1 mutation, we will follow some testable hypotheses in future posts!
★ Reference (free read): Shank1 deletions in males with autism spectrum disorder. Sato et al., 2012
Links to my older posts on autism are here:
The Genetics of Autism ▶ http://goo.gl/AzTuAX
Autism Spectrum Disorders from Mechanism to Therapy ▶ http://goo.gl/y751QH
A Part of the Puzzle: NHE9 and Autism ▶http://goo.gl/YXbOkN
Very interesting, Rajini
What is the p value for the statistical difference?
Cynthia Bush , the p value for that data was 0.009. Fisher’s Exact test two-tailed.
That seems a bit shaky for the numbers, but I trust your data.
Cynthia Bush , these are not my data! 🙂 I’m using the reference cited in my post to (i) explain a pedigree chart, and (ii) begin a discussion on gender differences in autism diagnoses. There is more information on sequencing in the linked paper, and I used only the pedigree analysis in this post. Autism associated mutations are always very rare, with no single mutation contributing to more than 1% of all non-syndromic cases. This makes it hard to understand the mechanism. One common feature is that many genes identified work at the synapse.
Will share with friend having autistic son. Thank You for your research!
Your reserch very good
The diagnostic bias is known to be too far on the male side, as many female autistics either end up undiagnosed, cannot be diagnosed correctly because diagnosticians fallaciously assume that being female excludes an autism diagnosis, or nonsense like Baron-Cohen’s ‘extreme male brain’ tosh.
Rajini Rao Is there a basis for estimating to what degree an actual sexual bias for autism prevalence is found? I’d be interested in seeing that. Certainly, the 11:1 AS ratio is probably fallacious, in my view.
Interestingly the male:female ration for ADHD shows a similar pattern, though the ratio is much lower, 2.28:1 Ref: http://www.ncbi.nlm.nih.gov/pubmed/20410711
I’m just wondering if there isn’t a gene difference but a gender stereotype difference? In the above article they dismiss the ratio difference to under-reporting of adhd in females. Again I’m wondering if a) the syndromes (can I call it that?) symptoms are gender biased, and hence we don’t see it in women as often? Or b) maybe it’s still poorly understood?
Gender stereotyping and under diagnosis are both good (but partial) explanations for the bias in autism (and some other neurological conditions, Satyr Icon ). I hope to cover the literature in support of this in my next post. However, there’s is also some biological evidence for a “female protective effect” (i.e., differences in threshold and tolerance to environmental/genetic burden). I’ve not formed an opinion yet on the extreme male brain theory, but I’ve seen some papers on this based on imaging.
Wooshed over my head!! Only thing I could grasp is autism hates/loves (if that’s of any consolation) blue squares.
R Prakash Prakash , nooo…this chart should easily slide into your head and churn away! It’s just a family tree. Look at the top: great grandpa marries great grandma and they have four kids, none with autism. And so on. Give it a try, and ask if you have doubts.
Gert Sønderby and Satyr Icon , I forgot to add that in the family featured in this pedigree, the females with the mutation had no features of autism but showed symptoms of anxiety, which is considered distinct from autism. Also the father in generation 1 (striped blue box) was also tentatively diagnosed with autism based on anecdotal evidence (since he is deceased). So this gender bias is quite mysterious. They experimentally ruled out a role for the X chromosome in this study, and say that paternal silencing (that is, the gene on the chromosome from the father is turned off epigenetically) is a possibility.
Rajini Rao well then I can’t wait for your next article. I have moderate ADHD, meaning I will probably get distracted by work and completely forget about this. Plus me in please! 🙂
BTW, I thought you were just making up terms with “extreme male brain theory”. I was thinking, what does she mean by that? Extra macho di@&heads?;) Anyway, I’m glad I did a quick search before I opened my proverbial about that because it turns out that is actually a technical term. The EMBT, not the macho di%$heads. Though that ought to be too.
Very cool, thanks for the puzzle link Praveen Kulkarni !!
LOL, scientists do come up with some bizarre terms and theories Satyr Icon ! Fortunately, they quickly turn them into acronyms that sound more official 😉
Re. ADHD, one of my favorite transporters, that we study in our lab, is strongly linked to ADHD. The gene name is SLC9A9 (protein is NHE9) and alter pH inside endosomes and trafficking of cargo within the cell.
Thanks Rajini Rao for your help.
> Gene responsible for autism is called SHANK1 Gene?
> What is meant by ‘controls’? who are they?
> All the members on the family in 3rd generation carried CNV, right? >The article says six members, where as there are only five in the chart. I might have got it wrong and that makes my confusion hard.
R Prakash Prakash , (1) in this family the SHANK1 gene appears to be responsible for autism. At least, that is what the researchers are trying to prove.
(2) Controls are people who do not have a diagnosis of autism. Like you and I. The researchers looked at the SHANK1 gene in 15,000 people without autism and did not find this CNV (or mutation). This suggests that people with a defect in SHANK1 may have autism.
(3) Only the members who have CNV labeled under the square/circle have carry this mutation. If you count the red colored CNV label, there are six in all. Notice that the females (pink circles) who have CNV do not have autism. But the four males (blue squares) who have the CNV all have autism. That’s the evidence for gender bias that we need to understand.
Maybe Autism develops more in persons with higher electrical activity in the brain 🙂
Our brains do indeed use spikes of charges in a form of electrical signaling, ajith kunnathully . But it’s a lot more complex than “more” or “less” activity 🙂
Thanks Rajini Rao . Many thanks for your kindness.
We had a family at our church that had two sons; one was deeply autistic and the other was so gifted that he had taught himself to read at four years of age. His parents we somewhat autistic themselves.
Geoffrey Swenson , I’m guessing the precocious reader falls into the autistic savant category. There are many wonderful traits of autism that should be nurtured and celebrated. Then there are the miserable gastrointestinal and immune problems, absence of speech and seizures in others that we would like to alleviate. Good point about the range of symptoms in autism spectrum disorders, thank you!
SHANK you for posting this Rajini Rao. I’ve mentioned before that two of my nephews have Autism Spectrum Disorders and one has Asperger’s syndrome. I think this could be an example where 23andme could be helpful and hurtful. We are still trying to understand the genetic aspects of Autism Spectrum Disorders. I could see where having the genetic information available to parents and their primary care physicians could be useful. I could also see where people could make poor decisions with this information because we simply don’t know enough yet.
Chad Haney , I agree! Hopefully, 23andMe will be back with more resources and oversight to offer a more complete service. I noticed that you mentioned nephews…anecdotally, many of us know young boys with autism. One friend mentioned that her mother was relieved when she heard that she (my friend) was carrying a girl baby because of the risk of autism in boys. I was quite taken aback by the thought. The 1:50 diagnosis for boys is really quite extraordinary (whatever we may think of the diagnosis).
In current news, the singer Susan Boyle has revealed her Asperger’s syndrome diagnosis.
What are the specific symptoms and severity of the SHANK related autism cases? Autism is applied to a very wide range of behaviors and disabilities. It’s only a hunch based on my own observations, but autism is a diagnosis applied to a variety of conditions with only a few things in common. The anxiety thing is interesting, the one woman I work with who has an official autism diagnosis certainly has disabling levels of anxiety.
Rajini Rao have you thought about discussing 23andme separately? I don’t have enough of a genetic background to wade into that topic. There are some people treating it as the genetic version of OpenAccess which I think is oversimplifying things.
Part of my interest in this topic is that I’m somewhat autistic myself. I lost two jobs in a row a few months ago because I need a little extra help adapting to new environments. I seem to be doing ok at my current job where they are happy to let me be a sometimes awkward but very smart geek. I am interested also in anything you can tell me about gastrointestinal issues. I had no idea these were related.
Barry Blatt , you are right, hence the term autism spectrum disorders. The uniqueness of each autist makes sense given the nearly unique genetic make-up and environmental experience of each. As I mentioned, this makes the condition very hetergenous and challenging. Anxiety disorders are more common in women, and they may or may not be associated with autism.
In the interest of space, I’ll just say that the cases in this family were of the high functioning kind (e.g., all had language ability for example). The link (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3376495/) describes each individual in detail.
William Rutiser thank you for the link! Imagine going through life being told that you had brain damage. Life is hard enough without these labels, so thanks for the information on Susan Boyle.
Geoffrey Swenson , I like smart geeks 🙂 I’ll find you some info on the GI issues that autists face. They can be extremely debilitating. Did you see the recent news report on the link between autism and gut bacteria? Here is a post by joe breskin on “good bacteria” relieving symptoms http://goo.gl/RBCE9l
I have a laymen hypothesis; that because until recently it was believed that autists did not have empathy more males were diagnosed because more work is traditionally done with females to care for other’s emotional states. Recent studies are showing that autists do in fact have empathy, perhaps so much of it that they must withdraw from people or be drawn under from it. Also studies show that while the children have difficulty in learning facial expressions, they can learn to do so. This makes me suspect that had they checked when I was young I could have ended up diagnosed. However due to childhood issues I had strong incentive to develope vigilance to learn facial expressions, both parents being abusive.
Brenna Bliss , I agree..the empathy issue is an autism label that recent studies show is incorrect. Interesting idea that there may be a case of empathy overload to explain withdrawal. You’ve given us something to think about, thank you for sharing your perspective on this complex condition.
Chad Haney , the 23andMe story would be great for a hangout. Perhaps we can persuade Ian Bosdet to join in..he is our resident human genetics expert on Google+. There are others here who have used the service (Matt Kuenzel ?) , I wonder what they think?
Good idea Rajini Rao
Some may find this article interesting: http://blogs.scientificamerican.com/mind-guest-blog/2013/12/04/can-synesthesia-in-autism-lead-to-savantism/
Interesting! Thank you Rajini Rao for sharing.
Excellent account of this topic and great follow up comments. I had a couple of points to raise but they have all been covered I think.
On the face of it this seems to be a reasonably well constructed bit of research. It will be interesting to see how replicable it is over time.
Jim Donegan , I was just commenting to a friend on another post that the frustrating point is connecting the dots between changes in genes/cell function to the complexities of behavior in an individual. We can learn a lot about the former, and improve the quality and quantity of the data, but to integrate it all together seems too complex.
Rajini Rao Early days yet, really, to generate an overarching theory of exactly how genes ‘work’ at the highest scales. I’m not a geneticist but the principles involved in describing the mathematical complexity seen at the highest scales is very familiar to me.
Is it a computational problem? No, not yet – I wouldn’t know what to compute. But understanding how divergence from given system states over time with changed boundary constraints and the extent to which those changes feed back to and interplay with genes resulting in ‘ons’ and ‘offs’ is something we might begin to understand once at least the basic mechanisms are fully grasped. So we see patterns – dots – and try to understand what it all means.
For me one of the great things about all of this is that it spurs medicine on to become truly scientific and evidence-based rather than the art form it has been in the past.
I’m looking forward to the day when we can computationally model complex systems. Until then, I’ll do my best to churn out individual bits of data, although it seems dauntingly futile at times.
this is really interesting. I just learned about the gender bias as I prepared a post on autism also for today
at the risk of adding too large a jump via this tangent, I offer another very interesting paper, not related to autism, but very much related to Rajini Rao’s observation at 11:26 AM about the difficulties in connecting the dots between genome and behavior, and note that it links directly to the incredibly important (and almost inconceivably complex and complicating) role(s) of gut microbes in this conversation, and how potentially fragile the structure of this community may turn out to be.
Read somewhere that there are a lot of gender-biased illnesses where the genes are on the X chomosome, because the women only get the faulty gene half the time.
It seem that you would need a very large set diagnosticians composed of an equal number of males and females diagnosing the very same large set of an equal number of male and female patients to arrive at a better set of test results if this was possible so there would be less incidence of error in the diagnosis of autism. I was labelled a paranoid schizophrenic and can’t remember if I was diagnosed as bipolar or not the first time I was hospitalized at age 49, 2 years after I lost my mother and my job as a computer program supporting a husband and family of 4 and soon after quitting nursing school and attempting to return to programming I got fired. I quit taking the medication soon after release from the hospital. 3 years later I was diagnosed as Bipolar I and hospitalized again. I had a long period of depression after the first hospitalization. In another 3 years they just asked me what my last diagnosis was and I don’t know if I got the same diagnosis or a different one. Again I was asked about my prior diagnoses 3 years later and guess since it was the same doctor the 4th time as the 3rd time I got the same diagnosis as the time before which was probably Bipolar I. Thing is when I was depressed and drugged the first 3 times I suffered periods of depression. The only time I get hospitalized is when I am in a mania state and hallucinating which could be from anxiety, lack of sleep and nutrition. I have always have had anxiety about failing and not measuring up. I have had mood swings all my life. I worked 24 years at the same job for the same company which kept changing names. Really got depressed after the 3rd and 4th child were born. Had a miscarriage between them. Initially was relieved about it then felt tremendous guilt. Learned that all of my brother had anti-anxiety meds at some point in their lives, but they were all in management positions which I never wanted. I was never on any meds til 49 not even asprin. Did have PMS and easy arousal to anger especially those times. My mother had anxiety. Her sister was even put in a mental hospital until her other sister got her out. My moms brother committed suicide at age 45. I am the youngest of 4 boys. My mom lost her first baby and still told me about it more than 50 years later. My dad may have been bipolar. He was a heavy drinker. The sister that was hospitalized for mental illness, both her son and husband were alcoholics. My brothers are drinkers and one has serious problems with it. After 26, I never touched alcohol again and married someone who did not drink. I do have trouble focusing. Always had reading comprehension problems, but was better than avg in math, but put in lots of effort, too. When I was working in computers, I realized that sometimes I said the opposite of what I meant to say, but was unaware of it until I had a disagreement with another person about what I said. My husband and family say I am argumentative, both in childhood and in my own family with husband and kids. They say I jump from topic to topic and don’t stay on the topic. I do interrupt. Often times when I try to wait patiently, I cannot recall what I was going to say especially if I have to listen and digest what others are saying first. I might do better in writing than speaking. My high school teacher checked my IQ because she said I talked like a very slowly and when she taped me and I listened, I thought I sounded retarded, too. My oldest girl 26 has OCDC by my diagnosis. She is a good student, but things don’t come easy for her. My next boy 25.5 is very bright and ambitious . My next boy is bright, but slacks off and is addicted to computer games age 20.5, my last boy is 14.5 and very sensitive to noise and disagreements. He is big on computer games, too, but he has better study habits than the 20.5, but not as good as 25.5. I worry if my children will be diagnosed as bipolar down the road. Is bipolar genetic? Sometimes I think I have ADHD, too. My nurse sister-in-law once called me hyperactive when I was younger. When I read, I find my self rereading the same material over and over instead of running right through it. Lots of times I used to write what I was reading about to stay attuned to what I had been reading. In 7th grade we used to take turns reading the material in history class. I would just move my finger along the page so that if I got called upon I would know where we were. I didn’t read many of the books in English that I was supposed to read. Would get the footnotes. I did have to push myself to read.
Satyr Icon To make sure you don’t miss the series of posts on this topic, why not simply ‘follow’ Rajini? I don’t think I have even seen a post from her that was not worth reading.
Note – you may see the same post in her ‘shared’ stream and also in one of the science streams. Go with the one that has the big comment hit rate, as this version does:-)
She is also a great cook and posts (elsewhere) about that, too:-)
Geoffrey Swenson Rajini Rao If you do find some GI info, can you post it in the comments here, or make a post in it’s own right? This impacts a lot of people and it would be a good addition.
Nick James , will do re. the GI info.
Satyr Icon does “follow” me, and I just realized that he is not in my circles so I circled back. Unfortunately, G+ cherry picks the posts that appear on my stream (in FB it’s called the edge rank algorithm, I think). So I tend to see the same posts/posters all the time in my general stream. I’ve been advised to look at individual circles separately, to get more of a sampling of what’s out there.
My understanding and which is supported by long-standing personal experience is that autistics have affective and motivational empathy, but not cognitive. So their wonderful empathy may not get triggered if one does not understand this aspect. Their inability to pick up on non verbal communication (which comprises the majority of communication for non-autistics) means that you need to clearly and directly communicate via words (not easy to do when you are upset and in need of instant empathy) in order for their empathy to be triggered.
Rajini Rao I’d like to thank for your delightful post and yes, I laughed at the joke, too;-) Even more, Thank You for all the effort you have put into answering so sensitively and comprehensively all the people that have commented here. It is obviously a very deeply felt topic for them and I’m sure they all appreciate your effort today.
E.E. Giorgi has a really good, detailed post on autism today. Check it out on her blog here: http://chimerasthebooks.blogspot.com/2013/12/autism-not-one-disease-but-spectrum-of.html
She brings up many of the difficulties in the study of autism, including the statistics that we only touched on here. The lack of replication in the gene hunts, and very large number of candidate genes, has made it difficult to separate the chaff from the grain, as it were.
Donna Pezeshki your story illustrates how complicated and intertwined mood disorders really are. While there are genetic components to most of the disorders you describe, it is the mix of our genes and environment that takes us down individual paths of development. I hope you are getting good therapy. Those career bumps sound tough. As for the boys, aren’t they all gaming addicts now? I felt better about my rascal’s video game obsession when I saw the medical student lounge at Hopkins full of gaming consoles and zombie players 🙂
Nick James this link – re GI stuff – gets close to the brink of “23andme problem” – which from my POV has little to do with the FDA’s stated concerns – or even the IRB concerns – and almost everything to do with getting a large enough database, (same problems here as 23andme) populating it with information on which action-related decisions can be made, and developing an appropriate research interface to that data – but I am seeing a lot of interesting if not immediately useful information about GIT microbes coming out of their social media posts.
here is a pretty good summary of complaints
joe breskin Geoffrey Swenson Thanks for the pointers, Joe.
Interesting how regulation and enterprise sometimes have a different focus;-)
Rajini Rao I agonized over the decision whether to buy the 23andMe genetic profile. Eventually I decided against it and this is why: No doubt my profile would show numerous variants associated with heightened risk of many adverse conditions. (I assume that would be true for anyone.) If, for instance, I found out today that I am at moderately increased risk for diabetes, I could decide whether to change my lifestyle or not. If I didn’t have children then any increased anxiety would affect only me. But I have two children in their twenties. So for each of those worrisome possibilities in my profile, I would have to decide whether to inform my children and worry them or to silently suffer anxiety on their behalf. If I were to tell them that they also are at heightened risk for diabetes would it benefit them or just increase their own anxiety? And if I didn’t tell them how would I feel when they later developed the condition? It seems paradoxical but, in my case, I prefer not having the information.
Matt Kuenzel I get your reasoning, although diabetes is more amenable to control in the long term compared to cancer. I think I would want to know for something like the BRCA mutations linked to breast cancer (I’m not in the relevant risk group for these though). I can imagine how hard the decision must have been for Angelina Jolie to go through the double mastectomy after she found she was carrying the risk allele.
To see Rajini Rao posts or even find your own, just type the name of the individual in the Google+ search box like visiting a person’s wall. Then if you go regularly you won’t miss anything.
Believe you can just click on a person’s name and get to see all of their posts, too, if they comment on your post without even being in the same circle. Just learning Google+
more news on autism: http://www.medicalnewstoday.com/categories/autism
More GIT stuff
Interesting point on the anxiety gene. Autism and anxiety sometimes have co-morbidity. I wish I had a science degree. It would be interesting to look at correlations.
ADHD or ADD are often co-morbids also. Both of those have been linked to social issues as well and having either can make one miss social cues or facial expressions.
I am interested in the research (http://goo.gl/oGEU7o) that shows increased autism risk caused by de novo mutations in the gametes of the parents (i.e. sperm from elderly Dad). This would not explain the gender bias but it would explain why older parents have higher risk of having an autistic child.
As per +Systems Biology manuscript, In case of mutaion ratio, i find that the females ratio is bit higher than males. Can you update me on any environmental origins??
But any ways great topic. Thanks for sharing thoughtful and intellectual topics… 🙂
I know of 3 autistic people and all 3 are males. Statistics say I should at least one or more should be female if Uttam sanDy is correct, but I know… this is a very small sample and he /she could be correct. Not gonna say anymore.
Systems Biology the paper you linked to is a great example of how autism shows polygenic characteristics: spontaneous mutations arise in any of a large number of genes (more with age) and in certain combinations, increase risk. Individually, the risk is small but together they place a burden on a critical brain pathway to change function.
Drew Sowersby but paternity tests have been around for every. I assume they too are regulated by the FDA?
Uttam sanDy some environmental effects that increase risk are well documented, such as taking anti-depressants during pregnancy. Others are suspected but still being validated. There must surely be environmental contribution to autism (as well as other neurological traits).
Rajini Rao Did you get the eBook link I posted for you? Dec 8 on my stream.
No! Checking now, Madjid Boukri .
My husband wonders if the father’s medications during conception could affect things. I was taking nothing, but he was on Gabapentin when we conceived our daughter.
Great post, Rajini Rao. And not just because it contains a pedigree. 😛
The gender-specific SHANK1 CNV effect is interesting. I gave a talk just yesterday on the spatial organization of chromosomes in the nucleus (i.e. each chromosome actually has a specific place and shape in the nucleus, both of which are important for gene function). You could speculate that the SHANK1 deletions change the conformation and interactions of chromosome 19q, with the effect being different depending on whether there are two Xs or an X and a Y in there too. It’s highly unlikely that’s really the case, but it’s a good reminder that connecting the dots between gene changes and behavior is going to take a while: there are whole new types of dots we’re just discovering!
Eric Hollander’s group has been doing some interesting work with autism and the possible role of inflammation. Some (not all) kids have improved symptoms when they have a fever, so he found that putting them in a hot bath once a day helped them. He also tried dosing adults with benign worm eggs (ya, ick) and that helped them too. Perhaps it also relates back to the microbiome work that joe breskin mentioned.
I’d be happy to talk about 23andMe sometime (although the “resident” is a bit generous, since it took me 4 days to see your post). It’s a fascinating time for consumer genetics. I spent some time on 23andMe’s user forums the other day and it was disheartening. The FDA is worried about whether the data interpretation is correct (as they should be). The other big problem is that many people really don’t seem to understand what they’re being told, meaning we have to find a better way to tell them.
I just pasted [Ian Bosdet 23andme] into the Google+ search box and now I can see LOTS of stuff that you have written/posted Ian Bosdet …
Now, I would like to add one more seriously off-topic tangent here, in the interest of nudging our adventures deeper into serendipity. This journal
http://www.hindawi.com/ has provided me with some seriously interesting papers. The paper I am linking, about TCM and epigenetic changes is not very far very from some of the ideas the present discussion has already touched on or engaged
Epigenetics in Traditional Chinese Pharmacy: A Bioinformatic Study at Pharmacopoeia Scale
Thank You Rajini Rao
Este es un pedigree chart mostrando porque autismo es mas diagnosticado en los niños que en las niñas.
Alfinal del artículo, hay mas links de posts sobre el asunto.
Muchas gracias!! Susana M. He visto además que hay varios enlaces en el post. Lo miraré con mucho interés 🙂