Fluoride is one of the most interesting stories of public health improvement. It all began over 100 years ago, with two simple observations: 1) that lots of people in some areas of the world had weird marks on their teeth, and 2) that somehow they never seemed to lose them to decay. After over a decade, scientists identified the reason—in these areas, there was fluoride in the water. In large amounts, fluoride causes unsightly blemishes on teeth, and it may even have some negative health consequences. But it wasn’t long before people realized that adding a tinybit of fluoride to public drinking water prevented dental decay and had no problematic impacts on health.
Despite the obvious and proven benefits of fluoridation for teeth, it’s very controversial. There’s been an ongoing debate for decades about whether government fluoridation programs could be causing children to lose IQ points. This theory is a favorite of the incoming Trump administration. It’s also recently hit headlines because of a new study published in the Journal of the American Medical Association Pediatrics that seems to show that fluoride is indeed making children less smart.
So is there something to this pet issue of RFK Jr.’s? What do the data really say?
The new paper is a systematic review and meta-analysis. This is a type of study in which researchers comb the literature on a topic to find all the research that’s been published in the area. They then put all that data together in a statistical model to get an idea of what the evidence shows altogether.
In this case, the researchers looked at studies that compared groups of kids who were exposed to more fluoride with those exposed to less. These studies looked at either those marks on the teeth that I mentioned earlier (which are called dental fluorosis), markers of fluoride exposure in urine, or sometimes estimates based on the amount of fluid people reported drinking. I’m an epidemiologist who fact-checks science news, and in general, the review looks pretty solid as a research paper. What’s more, to the great credit of the authors, they’ve put their methods and data online for all to read. This is very rare transparency for a scientific project.
Overall, the review paper found that fluoride exposure for kids was associated with lower IQs. The main model that the authors ran included a total of 74 studies (so, quite a few), and overall the effect size was quite big. But don’t get on board with the talk of banning fluoride just yet: These results come with some major caveats that are pretty obvious once you start reading the research that was included in the meta-analysis.
For one thing, most (77 percent) of the research in this review was conducted in China or India. China in particular is known for having very high levels of natural fluoride due to features of the groundwater in the country, with some regions having more than 60 times the Centers for Disease Control and Prevention’s recommended levels of fluoride occurring naturally in their water supply.
In addition, most of the research included in the meta-analysis was …terrible. The authors used an 11-question rating scale to determine whether they thought a study was reasonably solid. On this scale, only 13 of the 74 papers they found didn’t have significant issues with how they had reported the data or how the studies were conducted. And even the studies that were reasonably strong mostly had serious caveats.
It’s not that the people who did this research were bad at their jobs—it just comes back to the way this kind of research is done. All the research on whether fluoride in water is associated with IQ is observational, because you can’t really randomize people to drink fluoridated water, at least over any kind of span of time. People drink what comes out of their taps; imagine the work involved in supplying different kinds of water to many research subjects. This inherent limitation means that even high-quality research in this area is often riddled with potential pitfalls that make it hard to assess whether the fluoride is the problem or if something else is going on.
For example, take this study, which the review rated as high quality. The authors of the study looked at mother–child pairs in Mexico City. They took two databases of mothers who were enrolled in studies on lead exposure but had also had fluoride tests while pregnant, and followed their children up about a decade later to record their IQs. They found that there was a fairly large association between a mother’s having more fluoride in their urine and the child’s IQ at ages 6 to 12.
But the study had major limitations. The authors had to exclude 70 percent of the sample because they were missing data on either the mother or the child. This isn’t surprising, given that the researchers were following up on kids a decade after their mother had been in a completely different study. But it does mean that the results could have been seriously affected by this dropout. Although this may not have been considered a serious problem by the review paper, it renders the final numbers extremely unreliable.
What’s more, the review looked almost exclusively at studies of natural fluoridation. This is where researchers compare people who live in areas with different amounts of natural fluoride in the water. It’s a big issue because areas with high natural fluoride tend to be very different from areas with lower amounts: The former are often more rural, are lower income, and have other factors that can influence IQ.
If you examine the small number of studies that have compared fluoride levels in high-income countries where we’ve added fluoride to the water, the picture looks quite different. A detailed econometric paper from 2021 using a natural experiment in Sweden found that there was likely no relationship between cities’ putting small amounts of fluoride in their water and intelligence. In fact, the study found that putting fluoride in the water probably helped people from low-income areas earn more money, likely because they had fewer teeth problems that could prevent them from working.
Other studies from high-income places looking specifically at governments that put fluoride in the water show similar things. A Canadian study from 2023 followed what happened when Calgary stopped putting fluoride in the water in 2011, and found no major differences between kids who were exposed to fluoride and those who weren’t. A 2024 study looking at children born in Odense, Denmark, found that there was no association between fluoride in the water and children’s IQs. A recent Spanish paper showed that boys—but not girls—who were exposed to fluoride in the womb had higher IQs than did boys with no fluoride exposure.
From all of this data, we can say a couple of things with a fair deal of confidence. Children who are born in areas that have naturally high fluoride levels—particularly in China, where those levels can be extremely high—often have lower IQs than do kids in areas with little or no natural fluoride. This could be because of the fluoride, or it could be because kids in those areas have lots of disadvantages that affect their IQ scores. On the other hand, there appears to be no impact on IQ for kids when we add about 1 part per million of fluoride to drinking water, which is all it takes to get those dental benefits.
Finally, we also know that water fluoridation prevents dental disease. This has been shown time and again, and is proved in many of the studies I’ve cited in this piece. If we add a tiny amount of fluoride to drinking water, kids lose fewer teeth. The kids who benefit the most are also the poorest and most vulnerable. Despite the furor, there’s really no good evidence that adding fluoride to drinking water hurts children. It’s beneficial, and also really, really inexpensive; per capita, adding fluoride to the water can cost on the order of dimes each year. Overall, it’s a major public health win.
Original article online at: https://slate.com/technology/2025/01/fluoride-childrens-iq-study-pregnancy-science.html