[nilsb]

Or maybe the "rise" in autism rates is due to the fact that our diagnostic tools have become more advanced in the past 20 years which has resulted in more people getting their diagnoses.

In particular older people (i.e. 30+) are getting diagnosed just now because doctors previously didn't know what they were looking for.

[typeformer]

Do you not think that the highly competent researchers didn't account for this very well know the rise in the data. Their findings here are novel, critically important and are related to a host of other works revealing the intricate links between the gut biome and neurobiology previously unknown. Parkinsons is another disease that recently has been shown to potentially be related to changes in the gut biome.

[olliej]

The highly competent researchers have to answer the question:

“why have the number of children diagnosed with ASD increased?”

They can’t “control” for changing diagnostic rules: that would literally mean “controlling” the exact thing they’re trying to analyze.

What they are doing (as medicinal chemists?) is identify a compound(s) that correlate with the increased diagnoses.

Another study might investigate the method by which ASD is diagnosed - eg applying modern diagnoses to old diagnoses, and vice versa.

There are also definitely sociological factors that could contribute - in the US specialized educational support for kids diagnosed with ASD is much better funded per-capita than most other learning difficulties, so there is a clear advantage to pushing for (or shopping around for) a dr that will give an ASD diagnosis - eg If you can get FAS diagnoses as ASD you’ll get much more financial and educational support.

Personally I expect there to be a parental age component (which would explain the increased prevalence, relative to overall average, of ASD in middle and upper class Caucasian families, at least in the US).

I suspect however it’s going to be a combination of all those factors, and probably a few more for good measure.

[TrinaryWorksToo]

I've seen brilliant people make mistakes and not recognize it for decades. Consider the Intel processor Spectre vulnerabilities. Being smart doesn't make the logic correct

[loceng]

Or if we want to keep guessing instead of doing thorough long-term research, a more likely answer is that it's a mixture of many things.