Reposting my comment:
When I see a headline like this my instinct is to jump to "overfitting".
In a study like this they use P=0.05 as their cut off for significance. One in 20 such studies would be expected to give meaningless results. But they had the choice of cohort (all kids or boys and girls separately), so 1 in 10. They measured urine concentrations of fluorine at three time points during pregnancy, which makes for 7 different potential combinations of input variables, for a total of 14 different potential analyses, where 1 in 20 is expected to be significant. Of course, they didn't preregister, so who knows whether their analysis changed on the fly..
But not having published in this field or journal, I have no idea what their norms are.
Acute, chronic, cancer studies, Alzheimer's links?
Comparison vs bad dental hygiene in the above?
What is the reference group you propose?
How much would you pay for this relatively superb level of detail? (Which is on par with 50+ years of longitudinal and cross sectional studies. About 1 billion USD.)
Surely there are some population effects or something else going on here, or the effect, especially later on in life, evens out between the groups.
Likewise, I understand several council areas in Queensland stopped fluoridating in around 2016.
"If the work holds up—a big “if,” as the paper’s findings are already coming under heavy scrutiny—it could hold serious implications for public policy."
Why should scientific data be doubted simply because its implications, if taken seriously, would be to cause more inconvenience or less money-making in our lives?
>Sci-Hub Link: https://sci-hub.se/10.1001/jamapediatrics.2019.1728
>Study itself: https://sci-hub.se/10.1001/jamapediatrics.2019.1729
>A decent amount of this is focused on Black Tea consumption during pregnancy which is apparently a problem: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4808922/
>Collected skeptical commentary at the Science Media Centre: https://www.sciencemediacentre.org/expert-reaction-to-study-...
>First, the claim that maternal fluoride exposure is associated with a decrease in IQ of children is false. This finding was non-significant (but not reported in the abstract). They did observe a decrease for male children and a slight increase in IQ (but non-significant) for girls. This is an example of subgroup analysis – which is frowned upon in these kinds of studies because it is nearly always possible to identify some subgroup which shows an effect if the data are noisy. Here the data are very noisy. A further issue is that the estimate of the decrease in IQ for male offspring is unfeasibility large – at 4-5 IQ points. This level of average deficit would be readily detectable in previous studies and is likely a reflection of bias or very noisy data (the interval estimate here is very wide). As high fluoride areas are not randomly assigned there are also countless uncontrolled confounders. While they did correct for a limited set of covariates, the overall effect was non-significant with and without covariates. In summary it is not correct to imply that the data here show evidence of a link between maternal fluoride exposure and IQ. The average change in IQ is not statistically significant.
The scales chosen in the study make it sound a lot more scary, though, given the unfluoridated mean intake was 0.3 mg and the fluoridated mean intake was 0.93mg. Still, this is an estimated difference of two whole IQ points between the populations.
I am personally a bit spectical, sounds too big to be true. Here is an illustration of what I am saying https://xkcd.com/882/