Do accidents play a role here? Everyone's always talking about suicide and the possibility of that being more common among those on the autistic spectrum, but...
As someone on said spectrum, I'd say I might be more accident prone than someone who's normal, and potentially more likely to get into trouble in general. For example, while I wasn't seriously injuried by any of it, I've been nearly hit by traffic, fallen off various things, actually been hit by traffic and done quite a few things which would probably be seen as rather 'stupid' (in an 'endangers your health' sense).
Perhaps it's possible that due to various behavioural factors, those on the spectrum might be prone to getting into dangerous situations and paying the price for them more than 'normal' people. Being generally clumsy seems to be a common trait here.
There's also a potential darker side here, in that being on the spectrum might make some more likely to end up in a violent confrontation than someone who isn't. There were quite a few news stories in the past where people with these traits have been assaulted or even killed because they did something that somehow annoyed some random scumbag. Maybe people outside of the autistic spectrum might be more likely to see a situation is trouble and stay out of it.
Those things plus the risk of suicide could all explain this difference in mortality.
The original paper (paywall-free link: http://sci-hub.io/10.1192/bjp.bp.114.160192 ) itemises the frequencies of causes of death by control population and autism spectrum disorder population.
It suggests that autistic people are an order of magnitude more likely to commit suicide. But the figures also suggest that autistic people were more at risk from all causes of death, including common causes of death amongst the general populations (approx twice as likely to die of cancer; are autistic people more at risk of not getting the best available treatments in time or more prone to [uncontrolled] unhealthy behaviours?)
Some caution is needed in interpreting the figures, which appears to escape the notice of a Washington Post subeditor who didn't even get the figure in the headline right (the difference between a sample life expectancy of 70.20 years and 53.87 years is not 18... )
It's worth noting that sometimes autism is likely to have been a symptom of the causes of death recorded (for example, people diagnosed as being on the autism spectrum were more than eight times likely to have "congenital malformation" as a course of death, which is a cause of death which by definition doesn't have a behavioural component).
Another factor is the survey sample is unavoidably biased because the individuals were separated into autistic and non-autistic groups based on whether they had been in contact with clinical psychiatrists over their autism, something plausibly much more likely if they're suffering from unrelated health problems; the short average period between diagnosis of autism and death stands out here.
It sounds like the ascertainment bias there will be improved as time passes:
> Weaknesses of the present study include exclusive reliance on the National Patient Register for case ascertainment, leading to a selected and perhaps severely affected sample by only including individuals with ASD who had been in contact with clinical psychiatry services. This selection bias may be particularly relevant for individuals diagnosed before the year 2001 (i.e. before contact with out-patient psychiatric care services was included in the National Patient Register). However, after the year 2001 (88.2% in current data) all individuals having received an ASD diagnosis are registered in the National Patient Register because of the diagnostic assessment per se, i.e. also in cases with no further contact with psychiatric services.
Another way around this would be genotyping: use an autism polygenic score to quantify genetic liability. If the increased mortality is observed even with elevated polygenic scores which are insufficient to lead to diagnosable symptoms, that would suggest the cause is greater general genetic risk to disease rather than sequelae to symptoms.
I wonder how the mortality for people diagnosed with ASD compares with the mortality for people diagnosed with anything by a psychiatrist. It seems likely that by the time people see psychiatrists they're already on average less healthy than the general population.
I also think we are less scared of death. For example, I didn't feel anything more than anxiety and a little fear when we were robed, but my friends felt like it was the end of the world. When this situation ended, I automatically recovered all my senses, but the others needed some time to recover.
This also happens with other things. I think my ego is too big, and if someone dares me to do something, I might do it, even if it risks my life.
I talked with several other autists about this, and we all can't feel a fear of death as neurotypical seem to feel.
As someone on said spectrum, I'd say I might be more accident prone than someone who's normal, and potentially more likely to get into trouble in general. For example, while I wasn't seriously injuried by any of it, I've been nearly hit by traffic, fallen off various things, actually been hit by traffic and done quite a few things which would probably be seen as rather 'stupid' (in an 'endangers your health' sense).
Perhaps it's possible that due to various behavioural factors, those on the spectrum might be prone to getting into dangerous situations and paying the price for them more than 'normal' people. Being generally clumsy seems to be a common trait here.
There's also a potential darker side here, in that being on the spectrum might make some more likely to end up in a violent confrontation than someone who isn't. There were quite a few news stories in the past where people with these traits have been assaulted or even killed because they did something that somehow annoyed some random scumbag. Maybe people outside of the autistic spectrum might be more likely to see a situation is trouble and stay out of it.
Those things plus the risk of suicide could all explain this difference in mortality.