Strong evidence of the impact of fluoridated drinking water on cognitive development was a major reason for HHS’s altered stance on fluoride. A widely cited new fluoridation study—relying on old data—claims there are no risks to brain development; but its flawed design and failure to address available evidence make its value limited.
Listen to the audio version of this article:
THE TOPLINE
- The study concludes fluoridated water doesn’t affect IQ, but relies on indirect, county-level estimates of exposure rather than measuring actual fluoride levels in individuals.
- Key methodological issues—including grouping different types of fluoride exposure, poorly timed variables, and use of a decades-old cohort—undermine its relevance and reliability.
- The findings conflict with more recent and comprehensive research using stronger data, raising concerns that the study may reflect policy advocacy more than rigorous science.
A newly published paper in Proceedings of the National Academy of Sciences is already being touted as definitive proof that water fluoridation poses no risk to brain development. The paper represents an effort to counter a ‘gold standard’ systematic review and meta-analysis published last year in one of the most influential journals, JAMA Pediatrics, which evaluated 74 cross-sectional and prospective cohort studies and found significant inverse associations between fluoride exposure and children’s IQ scores. This study, along with others showing how fluoride in drinking water disturbs the gut microbiota impacting children’s immunity, were key to the change in stance on fluoride of HHS under Secretary Kennedy’s leadership.
In his Substack article, James Lyons-Weiler, PhD, reveals a catalogue of problems with the supposed debunking of the JAMA Pediatrics study that is being used for all it’s worth by the pro-fluoride, anti-Kennedy brigade. Beneath the headline-grabbing conclusion lies a study design that raises more questions than it answers and may ultimately obscure, rather than clarify, the real science on fluoride and cognition.
What the Study Claims
The study, drawing on data from the Wisconsin Longitudinal Study, concludes that exposure to community water fluoridation (CWF) is not associated with lower IQ in adolescence or reduced cognitive function later in life.
At first glance, that sounds reassuring. But the reliability of any scientific conclusion depends entirely on how the study was conducted. And that’s where the problems begin.
The authors present their work as a replication of earlier findings. But it’s not independent replication—the gold standard for building scientific consensus. Instead, it’s the same lead researcher, working with essentially the same dataset, confirming a prior result.
The Core Problem: Exposure Was Never Measured
Perhaps the most significant issue is this: the study never directly measures fluoride exposure in any individual.
There are no urine or blood fluoride levels, no measurements from participants’ drinking water, no clinical indicators like dental fluorosis. Instead, exposure is inferred from county-level records—essentially guessing who was exposed based on where they lived and when fluoridation began.
This kind of shortcut introduces what researchers call the ecological fallacy: assuming that group-level data accurately reflects individual experience.
A Critical Category Error
The study also treats all fluoride exposure as equivalent—but it isn’t.
Participants categorized as “exposed from birth” often lived in areas with naturally occurring fluoride in well water, not water treated with fluoridation chemicals used in modern systems.
These are chemically distinct substances with potentially different biological effects. Combining them into a single “exposure” category undermines the validity of the comparison from the start.
Timing Matters—and It’s Off
Another major flaw: the study measures “confounding variables” after years of supposed fluoride exposure have already occurred.
But if you’re trying to isolate cause and effect, your baseline measurements need to come before the exposure—not in the middle of it. Otherwise, you risk adjusting away the very effects you’re trying to detect.
A Study from a Different Era
There’s also a fundamental mismatch between the study population (high school graduates from 1957) and today’s reality.
Participants were born in the early 1940s—before widespread fluoridation standards, fluoride toothpaste, and modern dietary fluoride exposure.
Today’s children are exposed to fluoride from multiple sources simultaneously, creating a total exposure profile that simply didn’t exist in mid-20th century Wisconsin.
Using this cohort to answer today’s policy questions is like using 1950s air pollution data to evaluate modern emissions standards. It doesn’t work.
What About the Broader Evidence?
This study stands in contrast to more recent analyses, including:
- A 2024 review by the National Toxicology Program, which found moderate evidence linking fluoride exposure to lower IQ in children
- A 2025 meta-analysis in JAMA Pediatrics reporting similar concerns
These analyses rely more heavily on individual-level exposure data—the kind missing from the PNAS paper.
Policy Framing Raises Concerns
The study explicitly positions itself as a response to recent policy decisions to roll back fluoridation in some states.
That doesn’t automatically invalidate the research—but it does raise an important question: Was the study designed to investigate a question—or to defend a position?
The rapid publication timeline and methodological problems—each tending toward a null result—suggest the latter.
Why This Matters
Public health decisions affect millions of people—especially children. They should be grounded in the best possible science, not studies with built-in blind spots.
Let’s not forget: in a 2024 ruling, a federal court determined that fluoride levels in U.S. drinking water present an unreasonable risk to public health. The court found compelling evidence from human studies, including a review of over 70 epidemiological studies, which confirm that even lower levels of fluoride can have detrimental effects on children’s cognitive development.
One flawed study to the contrary doesn’t change this.
Science isn’t a fixed set of conclusions—it’s a process. The strength of the scientific method lies in its willingness to be challenged: to revisit assumptions, refine methods, and update conclusions when new data emerge. That means questioning isn’t a threat to science—it’s the engine that drives it forward.
On complex public health questions like fluoride exposure, where evidence is developing and stakes are high, that principle matters even more. Conflicting findings shouldn’t be dismissed or spun into certainty; they should prompt better research, more precise measurement, and a deeper look at what we may be missing.
If science is doing its job, the conversation doesn’t end with a single study—or even a consensus moment. It continues, evolves, and corrects itself over time. And public policy should remain open to that same process: grounded in evidence, but always ready to adapt as our understanding grows.
Please share this article widely in your networks.
