I am pretty sure the gender bias of autism is caused by a mutation on the 'X' Chromosome, which is what alters our social abilities, but since females are 'XX' they are able to compensate with the extra 'X' and become more social while still harboring the other mutations associated with Autism.
There is a mutation they found there, but it only accounts for a very small percentage of autistics.
Interesting study linked below that suggests that male brain characteristics among females with Aspergers, are similar to male brain characteristics in males with Aspergers.
The study challenges the assumption of an extreme male brain, in males with Aspergers, because the brain characteristics associated with the male gender are similar to a control group of males without Aspergers.
The difference in the volumes of total white matter and local gray matter between men and women is smaller in the group with Asperger syndrome than in controls. This suggests that the brains of women with the syndrome have more ‘male’ brain characteristics than those of controls, the researchers say.
The men with Asperger syndrome do not have more total white matter or more gray matter in the right parietal operculum compared with controls, however. The finding challenges the idea that men with Asperger syndrome have brains that are structurally more ‘male’ than those of controls.
Among controls, the men’s brains also have greater connectivity — as measured by the flow of water through the brain — than the women’s brains in a number of regions, including the corpus callosum, which connects the two hemispheres. This difference in connectivity between males and females is also smaller among the brains of individuals with Asperger syndrome.
2D/4D digit ratio to test for the potential of exposure to prenatal testosterone wasn't mentioned in the study, but it might stand to reason that there is a pretty good likely-hood that females diagnosed with Aspergers, will have a low digit ratio.
If one is not familiar with 2D/4D digit ratio, one can figure it out by measuring their index finger on their right hand from the crease of the palm by millimeters with a ruler, and doing a similar measurement with the ring finger, and then dividing the ring finger measurement into the index finger measurement to determine how low the ratio is. Normal ratios for a female exceed .97. There is the potential that exposure to high levels of prenatal testosterone may influence brain development per characteristics that have been associated with male brains.
2D/4D digit ratio measured by Samuel Baron Cohen, in his research with autistic individuals, provided evidence of very low ratios among males with autism disorder. The ratio was not as low among males with Aspergers. His study was very limited per the number of females with Aspergers, but they had a lower 2D/4D ratio than would be expected as well. He used it as evidence for the "Extreme Male Brain" theory per systemizing/empathizing and the AQ test he developed.
While the association of systemizing hold fairly strong across the spectrum, in the AQ test, that is only one area of brain gender, associated with many other characteristics in the test provided by the BBC, linked below.
At least in the case of Aspergers, it might be more appropriate overall to call the theory, the "neutral gender brain" theory, per real world results of the BBC provided "Sex I.D. test taken by autistic individuals in online communities. Per self-reports many people diagnosed with Aspergers, male and female, score close to the middle of the brain gender spectrum.
Other research by Michelle Dawson indicates that verbal intelligence among individuals with Aspergers is higher than performance intelligence measured in standard measures of IQ testing, whereas those with Autism Disorder score higher in performance measures in comparison to their verbal measures of intelligence in standard IQ testing. That could impact some of the brain gender associations in the test provided by the BBC, linked below.
There is a strong correlation of symptoms of non-verbal learning disorders measured among those diagnosed with Aspergers syndrome, which could potentially explain some of the differences measured in IQ testing, done by Dawson. She didn't explore that area, though.
Of course none of it is a definitive cause of autism, just interesting associations, per the 2d/4d digit ratio, and this latest research on actual differences noted per male brain characteristics.
Here is the link to the interesting test that indicates associations per gender ID, not necessarily reflective of sexual orientation, provided by the BBC, if one hasn't come across it before. One of the questions on the test requires an individual to measure their 2d/4d digit ratio.
Physical and behavioral characteristics associated with masculinity and femininity, interestingly, do not always match the results of what might be expected from the test linked above that attempts to measure brain gender, nor what Cohen describes as strong systemizers with an extreme male brain, per traits measured in the AQ test.
And finally, per these interesting gender and hormonal associations, an individual diagnosed with autism, Andrew Lehmann, has provided a very interesting theory, of an Estrogen associated theory of evolution, autism, & social change that pursues a much greater area, than what Cohen pursued in his theory and research.
It is linked here, in free book form, on this site:
It has nothing to do with autism supremacy, but it provides an interesting theory as to why the broader autism phenotype remains a part of the population, regardless of reproductive success among those in the phenotype.
While there are similar behavioral characteristics measured in the definition of Autism Spectrum Disorders, across the spectrum, the recent research that has shown abnormal brain growth, almost limited to males whom develop regressive autism; the research per Aspergers and brain characteristics quoted above; the differences seen in intelligence testing by Dawson, per autism disorder opposed to aspergers; along with the work of Cohen, provides some evidence that these conditions, while similar per psychological behavioral assessment are likely diverse in potential causal factors, per biology and environment, including levels of prenatal hormones, associated with physical, cognitive, and behavioral characteristics seen in other mammals.
These many factors, considered together, seem to suggest that the broader autism phenotype is likely a part of human nature that existed far into the past, and will likely continue to be a part of that nature, far into the future, regardless of changes in culture or medical technology.