The lost girls of autism

It's been a longstanding notion that Autism Spectrum Disorders are significantly more common among males than females, but I'm not clear in whether this is a real sex difference. There were prevalence studies from the 1980s that reported something like a 4/1 m/f ratio in autism/ASD diagnoses. It could be biological, sure. It could also be a sampling bias. More explicitly, I can think of at least these options:

1. Biological, mechanistic explanation: similar to how color blindness is predominantly male, ASD could be partially caused by yet unidentified genetic factors that happened to be recessive and X- chromosome linked.
2. Biological, functional (hyperfunctionalist/hyperadaptionist) explanation: its more common in males because the selective optimum for males but not for females is one that makes it so that a normal distribution around that optimum includes many people with diagnosable conditions on the spectrum.
3. Girls/women are simply better at masking their autistic traits so that they often go unassessed by clinicians without anyone to blame.
4. Autism Spectrum Disorders tend to present differently in male and female patients, and the standard diagnostic criteria are geared towards a typical male presentation.
5. It's become a self fulfilling prophecy of sorts, since it has become general knowledge that autism is a mostly male phenomenon, clinicians are prone to consider different diagnoses first and may never conduct a formal assessment for autism when a girl/woman presents with the exact same symptoms that would immediately trigger the autism bell if it were a male patient.

I have my suspicions, but I genuinely don't know. I think #2 is rather far fetched, but that's more because I generally find hyperadaptationist hypotheses rather farfetched if they don't come with strong evidence, rather than because of having any reason to think this is particularly implausible in this specific case.

Here's some related content (an interview by New Scientist with an experts who appears to lean towards 3 and 5)



Any comments?

There's an interesting skew in college degrees--the more social (are you working with people vs are you working with equipment) a field is the more it skews female. But this is not nearly as pronounced in women from a poor financial background. Clearly they are being influenced by the money.

This makes me think there's something to the old comparison that women are more social, it's not just socialization doing it. Now, let's consider autism very differently: Instead of being a specific malady it simply is a sufficient lack in a range--there's a normal distribution around a center point, we call a point a certain distance from that center to be "autism".

And as with most cases, small changes in the middle result in large differences in the tails. If women start out with a slight advantage in social ability a lot fewer of them end up across the line into "autism". (For a clear example of this at work, look at the racial composition of the NBA. Blacks are on average slightly taller, Asians are on average slightly shorter.)

Loren Pechtel said:

There's an interesting skew in college degrees--the more social (are you working with people vs are you working with equipment) a field is the more it skews female. But this is not nearly as pronounced in women from a poor financial background. Clearly they are being influenced by the money.

This makes me think there's something to the old comparison that women are more social, it's not just socialization doing it. Now, let's consider autism very differently: Instead of being a specific malady it simply is a sufficient lack in a range--there's a normal distribution around a center point, we call a point a certain distance from that center to be "autism".

Click to expand...

I think you might be equivocating between motivation and skills here. When you say that women tend to be more social, I believe we should interpret this as having a higher motivation to engage in social behaviours, not necessarily as being innately more skilled. On the other hand, when you say that ASD might boil down to being a certain distance from the normal, I believe we should interpret this in terms of innate skills. To complicate matters, autistic individuals are often able to compensate their lack of innate skills with consciously adopted behaviours, a phenomenon described as "masking", and of course masking behaviours will be more prominent in individuals with high motivation. So a person with high motivation but low skills might for example consciously count to three in their head while keeping eye contact or practice facial expressions in front of the mirror so that their innate lack of skills becomes less obvious to their surroundings. If it's true women tend to score higher on motivation, that alone might explain why women who score low on innate skills might be better at masking.

Loren Pechtel said:

And as with most cases, small changes in the middle result in large differences in the tails. If women start out with a slight advantage in social ability a lot fewer of them end up across the line into "autism". (For a clear example of this at work, look at the racial composition of the NBA. Blacks are on average slightly taller, Asians are on average slightly shorter.)

Click to expand...

I'm not saying this is implausible, but do we have any strong indication that it is true in this case? Apparently it's an accepted fact that girls in particular often have their autism Spectrum Disorders misdiagnosed initially. That could be partially because women show more masking behaviors, but possibly in part also due to caregivers and clinicians having grown up believing that ASD is a condition that mostly affects males. There appear to be studies where educators and clinicians were given a written description of a child's behavioral patterns, and the name at the top alone significantly affected the likelihood that they'd recommend a systematic screening for ASD given the exact same patterns. This is specifically brought up in the interview I linked in the OP.

Do we have a positive reason to believe this type of effects is insufficient to explain the observed gender bias in confirmed diagnoses?

I would pose that autism presents differently depending on gender.

The hard part here is the high comorbidity between gender dysphoria and autism, though, because this suggests that the same parts of the brain that have gender differences are the parts that differ in autism as well, and my experience of observing autistic transitioners strongly reinforces the intuition of differing presentations.

Prior to transition, autistic people I have known have been quiet, withdrawn, socially awkward.

Put some estrogen in them, though, and you probably wouldn't even recognize them for who they had been, the result is just so different in expression.

I imagine at one point in time, the stark differences between members of certain species were so pronounced on the gender difference that people might not even realize these were the same species.

Likewise, I think the difference between autistic trans people pre/post transition and how it influences the expression of autism are valuable.

I would expect that autistic girls will present "in their own way", and that the principle identifying factor will be rejection of the arbitrary social expectations assigned to them: autistic girls will reject girly just as much as autistic boys reject boyish, instead focusing doggedly on their actual personal interests.

As a result, I think the OP is largely right, and there isn't a difference in the regularity of expression across sex so much as there is a difference in how it manifests mostly due to our social expectations and how we see them violated.

It took years of work for my ex-wife to finally be given the autism diagnosis she'd known was appropriate for herself from the start. Her symptoms were obvious to her, but getting the medical system to agree so she would actually qualify for care under state insurance was a wild and expensive journey of medical bureaucratic hoop-jumping. So I am inclined to believe that the disorder is under-reported.