Comment by almog

There are natural experiments where the correlation between IQ and lower amounts of fluoride can be studied. For example in Israel, until 2014, the fluoride level was 0.8-1.0 mg/L depending on the region but starting in 2014 water is no longer fluoridated (naturally occurring fluoride level is in the 0.1-0.3 mg/L range for the vast majority of the population) so the data should be there, it's just a matter of collecting it. Some countries can provide data for the opposite natural experiment (adding fluoride in recent years).

To me, using the 1.5 mg/L doesn't tell anything about lower levels but it does make me interested in seeing such study whereas before I'd have just dismissed it.