Thread by @phylogenomics

There are, alas, some fundamental flaws with some of the inferences made in this paper (and in the comments on this paper) on autism spectrum disorder and infections 1/n

https://twitter.com/ScienceMagazine/status/1439680732679528450

I am particularly concerned with the section on clinical histories of children. In this analysis they looked at whether "prevalence of hospitalizations due to infections in males (but not in females) is associated with future development of autism spectrum disorders (ASD)." 2/n

That wording is from the abstract and seems fine to me as all they were doing was looking for correlations / associations. The problem is that in the main text of the paper they switch to inferring that the hospitalizations actually contributed to the ASD development 3/n

For example consider the first sentence on this work in the main text - where now they are claiming they are looking at whether the hospitalization "contributes" (i.e., causes in some way) the ASD severity and phenotypes. But they have no means of inferring causality here 4/n

If a child has some trait that leads to a higher risk of hospitalization due to infection and also a higher risk of ASD then they will see a correlation in hospitalization and ASD even if hospitalization has nothing at all to do with ASD 5/n

I mean, yes, the study is certainly interesting. But it is simply unsound to conclude from this that they have any evidence that hospitalizations due to infection contribute in any way to ASD or ASD severity 6/n

Alas some the news coverage of this has basically failed to mention that all they have found in humans is a correlation and have not shown any causal connection eg the-scientist.com/news-opinion/s… 7/n

Some of the next coverage has mentioned this though, so that is good - e.g., upi.com/Health_News/20… "While the study does not prove the illnesses cause autism" 8/n

So, no, the study did not show that severe infections in children contribute to autism development or severity. 9/n

And I think it is troubling that some of the people involved are using this flawed inference to argue that this is why people should get their kids vaccinated. Yes, vaccines are important. But not based on the flawed logic presented in this work. the-scientist.com/news-opinion/s… 10/n

@threadreaderapp unroll please