The truth behind that viral study on fluoride and IQ

A version of this essay first appeared in Jess Steier’s newsletter, “Unbiased Science.”

Fluoride is having a rough day in the court of public opinion. The National Toxicology Program (NTP), part of the Department of Health and Human Services, has released a comprehensive study titled “Monograph on the State of Science Concerning Fluoride Exposure and Neurodevelopment and Cognition,” and longtime fluoride foes argue that it includes a supposed bombshell finding: It linked high levels of fluoride with lower IQs in children.

This association between elevated fluoride exposure and decreased cognitive function in children has caused a lot of fear and anxiety. Adding weight to these findings, the NTP expressed “moderate confidence” in this association, which seems significant coming from a federal agency. Let’s be real: Before this report was published, many people were already concerned about fluoride in water and toothpaste — much to the chagrin of dentists — and this isn’t helping the situation.

The NTP defines “high levels” as 1.5 milligrams per liter of water or higher.  This is more than twice the fluoride concentration typically found in U.S. tap water: The Centers for Disease Control and Prevention recommends that U.S. community water systems contain about 0.7 milligrams per liter.

According to the NTP report, about 0.6% of the U.S. population — approximately 1.9 million people — are on water systems with naturally occurring fluoride levels of 1.5 milligrams per liter or higher. It’s important enough that I’ll repeat it: These high levels are naturally occurring and not the result of artificial water fluoridation programs. So let’s make this crystal clear: This study does not provide evidence against current community water fluoridation practices. 

Interestingly, the publication of this report was delayed for several years. As Linda Birnbaum, a toxicologist and former head of the NTP, told NPR, “There was a great deal of concern raised, especially by the dental community, and they were concerned that this report would be misconstrued to say that any fluoridation of drinking water is a problem.” This delay underscores the sensitivity of the issue and the potential for misinterpretation.

The American Dental Association (ADA) is not happy with the report. In their statement, they highlight “significant limitations and biases within the report” that must be considered before accepting its conclusions. The ADA also pointed out that there were two previous drafts of the report in 2019 and 2020, both of which raised major red flags. The National Academies of Sciences, Engineering and Medicine, which was the original peer reviewer, reported that these earlier drafts “would not survive scientific scrutiny without major revision.”

Of particular concern in these drafts was a hazard assessment in which the authors state that fluoride is “presumed to be a cognitive neurodevelopmental hazard to humans,” regardless of exposure level. This assessment was later removed after the peer review committee found that the report fell short of providing a clear and convincing argument to support this claim.

The ADA criticizes the NTP for abandoning the original peer review process and instead hand-picking its own panel for the final draft. They also point out that the NTP failed to adequately address concerns raised by the National Academies, including inconsistent application of risk of bias criteria, inadequate statistical rigor, and selective reporting of nonsignificant study results. These issues, according to the ADA, persist in the latest version of the report, undermining its credibility and conclusions.

The study’s finding of lower IQ scores, which has garnered significant attention and concern, was based on 19 high-quality studies, with 18 reporting an inverse association between fluoride exposure and IQ. That sounds damning. But only three of these were prospective cohort studies, which limits the ability to draw causal inferences. Moreover, these studies were conducted in five different countries with varied methodologies.

Now, putting on my data scientist hat: While meta-analyses can be powerful, it is important to remember that any threats to validity in each individual study get pooled (and potentially compounded) when the data are combined. This means that if there were any biases or confounding factors in the original studies, they could be amplified in the meta-analysis, potentially overstating the strength of the association. Some of the most important limitations are:

1. The study focused on fluoride levels significantly higher than those in U.S. water fluoridation programs.
2. None of the 19 IQ studies were conducted in the U.S. or in other places with fluoridation. Instead, they took place in areas with naturally high fluoride levels, limiting their generalizability.
3. The review didn’t consider fluoride’s dental health benefits or provide a risk-benefit analysis.
4.  It did not address risks at lower fluoride levels relevant to most public water systems. 
5. Adult fluoride exposure effects weren’t addressed.

The ADA’s criticisms include inconsistent application of risk of bias criteria, inadequate statistical rigor, and selective reporting of non-significant study results, all of which persist in the latest report.

Nevertheless, you may still be wondering: How much of an impact are we talking? The NTP report suggested IQ reductions in the range of 2 to 5 points in children with higher fluoride exposures (>1.5 mg/L). However, this should be interpreted cautiously:

• Again, this range was observed in studies with fluoride levels higher than U.S. water fluoridation programs.
• Effect sizes varied across studies.
• These are population averages, not individual guarantees.
• More research is needed, especially at lower exposure levels.

We can’t ignore fluoride’s well-established dental health benefits. One study not included in the meta-analysis found a significant association between lower cognitive ability and poorer dental health, persisting even after controlling for confounders like income and ethnicity. This highlights the complexity of the issue and the need for a holistic approach to public health.

Many people understand that “the dose makes the poison.” As one commenter in my DMs aptly put it, “Anything at twice the normal limit will have negative health consequences. That’s why there are limits.” This principle is crucial in understanding the NTP’s findings.

Some individuals express concern about cumulative fluoride exposure from multiple sources, including water, toothpaste, and dietary sources. While this is a valid point for further research, current evidence doesn’t suggest that normal use of fluoridated products in combination with optimally fluoridated water poses a risk.

The bottom line: While this study raises important questions about high-level fluoride exposure, it doesn’t challenge current water fluoridation practices.

The CDC recognizes water fluoridation as one of the great public health achievements of the 20th century. As we continue researching fluoride’s effects, it’s crucial to maintain a balanced perspective, considering both potential risks at high exposures and well-established benefits at recommended levels.

While we’ve focused on the potential cognitive effects of high fluoride exposure, it’s worth briefly mentioning dental fluorosis. Fluorosis is a separate condition that can occur when developing teeth are exposed to excessive fluoride, resulting in visible changes to tooth enamel. While mild fluorosis is generally considered cosmetic, severe cases can cause more significant enamel defects.

Once again: The fluoride levels associated with potential IQ effects in this study are significantly higher than those typically linked to fluorosis risk. As always, consult with your dentist or healthcare provider for personalized advice on fluoride use.

But for most people, this report means very little.

Jess Steier is a public health scientist and the host of the podcast “Unbiased Science.”