zlacker

Only 34 girls with autism were included in the study that concludes it is linked to autism in males and not females, while the study included 203 males with autism.

The control side was 54 males without autism and 67 females without autism.

Maybe 34 is enough to get that statistical conclusion, but it seems pretty small. It's based on recall of if they had -frequent diet drinks- (edit: cutoff variable was one single diet drink per day) during pregnancy, so it might be being determined by like 15 people for the female side that were diet drink users and more for the controls, further divided by 6 different demographic subsets? Will try and find that number.

>>cma+(OP)
For various reasons autism is often under-diagnosed in women, so I wouldn't be surprised that if this effect holds up it's not actually gender-specific. That's a bunch of big "ifs" though.

>>cma+(OP)
The claim is that there is a link for male offspring, not that there is no link for female offspring---that would be a very different claim.

From the abstract of the study [0]:

> No statistically significant associations were found in females.

[0] https://www.mdpi.com/2072-6643/15/17/3772

>>cma+(OP)
Women tend to not be diagnosed as much as men either, mainly as they're forced to mask more throughout life -- I don't think we'll ever get an accurate representation there, unfortunately.

I feel like there are things not being said in the study that should have been. I could be wrong, but I don't see a mention of whether the parents of the autistic children were themselves autistic or not and feel that could be important criteria to have noted.

>>cma+(OP)
If memory serves, only a few decades ago, it was believed that autism incidence was 10 boys for every 1 girl.

Since then I think they discovered they were under-diagnosing it in females due to gender norms, but my impression is that it still skews more heavily towards males (3:1 or 4:1).

>>cvoss+o1
If there had been the same percentage link on the female side could it have even then been claimed as statistically significant? Was it even possible for the study to show the same link with significance?