The US Media’s Coordinated Attack on Science and MAHA
On November 19, 2025, a strikingly synchronized wave of articles appeared across major outlets – Scientific American, Gizmodo, CNN, and even Utah’s Deseret News – all singing the same tune
Each ran headlines touting a new “massive” study by John R. Warren purportedly debunking claims that fluoridated water harms children’s brain development.
The messaging was uncannily aligned.
In Scientific American, the news story’s subtitle flatly declared: “Fluoride in Tap Water Not Linked to Lower Child IQ, Massive Study Finds”, noting that researchers tracked thousands of Americans and found “no links” between recommended fluoride levels and cognitive skills.
The piece explicitly framed the study as challenging Robert F. Kennedy, Jr.’s claim that adding fluoride to water may harm cognition.
On the same day, Gizmodo ran “New Research Shatters Common Claim About Fluoride and Intelligence” by Ed Cara, opening with the assertion that “the case for removing fluoride from the drinking water supply just got weaker”.
CNN’s headline was similarly emphatic: “Fluoride in drinking water does not negatively affect cognitive ability — and may actually provide benefit, study finds.” Its story by Deidre McPhillips reported that young people exposed to recommended fluoride levels performed better on cognitive tests than those without fluoride.
Even the Deseret News jumped in with a political angle, running a piece on Sen. Elizabeth Warren warning that halting fluoridation could harm U.S. military readiness, implicitly rebuking RFK Jr.’s anti-fluoride stance.
Such message discipline across disparate outlets on the very same day was no coincidence – it bore the hallmarks of a coordinated media operation.
The articles not only reported the new study but did so using nearly identical framing: that this “much-needed” U.S.-based research definitively shows community water fluoridation is safe and even beneficial, thereby debunking the fluoride-IQ link raised by RFK Jr. and the health freedom movement.
It’s as if a media playbook went out: Emphasize a “massive” dataset, declare no evidence of IQ harm, highlight a slight possible benefit, and portray fluoride critics as fear-mongers.
Indeed, Scientific American quoted a pediatrician praising the Warren study as “a much needed addition to a broad literature that shows the safety and benefits of community water fluoridation… the first study… in the United States, which makes it a stronger basis for… policy decisions.”
Gizmodo, after summarizing the results, went further by pointedly editorializing that these findings “obviously undercut the rationale to remove fluoride” and that it would take far stronger evidence than currently exists to justify ending fluoridation over “hypothetical” IQ concerns.
In Gizmodo’s words: “What we’re showing is that this IQ story — it doesn’t hold up… at levels of fluoride that are actually relevant for policy,” lead author John “Rob” Warren said. The piece concluded with a caustic twist:
“Fluoride might not be lowering anyone’s IQ. But the fear of it sure seems to be doing a number on some people’s brains.”
A snide swipe at fluoride skeptics’ intelligence.
Other outlets struck the same notes. CNN noted the longstanding practice of water fluoridation was under “heavy scrutiny” but that “new research challenges recent claims about the risks… and instead suggests it may have additional positive effects.”
The CNN report explicitly contextualized this “debunking” in opposition to RFK Jr., mentioning that Kennedy (now U.S. Health Secretary) had called fluoride “an industrial waste” and cited “IQ loss” in pledging to end federal fluoridation recommendations – and that Utah and Florida had become the first states to ban fluoridation.
The clear subtext: the new study proves those concerns misguided. CNN quoted Dr. Warren’s colorful analogy that earlier studies showing harm were like testing a heart drug at 1,000x the dose – “that doesn’t tell you anything about…100 milligrams versus nothing,” he said, arguing prior fluoride-IQ research isn’t relevant to real-world exposure.
All three science/health outlets (SciAm, Gizmodo, CNN) showcased quotes from Warren stressing how reasonable, mainstream science is on the side of fluoridation, while portraying opposing views as alarmist.
And then Deseret News amplified Sen. Warren’s attack line that RFK Jr.’s “fluoride crusade” is “anti-science,” threatening troops’ dental health and military deployability.
In short, the media blitz painted a united front: that fluoride is safe, beneficial, and only conspiracy theorists think otherwise.
The Warren et al. Study – Big Data, Big Headlines, Big Flaws
At the center of this PR tempest is the new study titled “Evidence- based water fluoridation policy” by Warren et al. published November 19 in Science Advances.
On the surface, it is indeed a “massive” analysis, leveraging data from the ‘High School and Beyond’ cohort that tracked 26,000+ Americans from adolescence in 1980 through mid-life (2021).
Warren’s team estimated each person’s fluoride exposure based on where they lived (fluoridated community or not) from birth through high school, then compared those exposures to two sets of cognitive outcomes: standardized test scores in high school, and memory/mental tests around age 60.
The headline finding gleefully repeated by the media: children who grew up with fluoridated water had slightly higher test scores in high school on average than those who never had fluoride, and by age 60 there was no difference in cognitive performance between the fluoridated and non-fluoridated groups.
In other words, no sign that fluoride lowered anyone’s IQ – if anything, kids with fluoride did marginally better in school. The implication (much played up by commentators) is that fluoridation might even help cognition indirectly by preventing dental problems.
Warren hypothesized that better dental health from fluoride meant fewer sick days and thus better academic performance. (It’s worth noting this is speculative; the study did not measure anyone’s actual dental health or school attendance, only test scores.)
The sheer size and American scope of the dataset was used as a selling point to claim this study as definitive. “It is the first study that looks at this information in the United States, which makes it a stronger basis for policy decisions,” said one expert in Scientific American.
And indeed, the data span over four decades and thousands of individuals. But size does not equal quality, and beneath the hype lie serious methodological flaws – many readily acknowledged in the fine print, though glossed over in the media coverage.
1. Ecological Exposure Measures: Warren’s team did not measure anyone’s actual fluoride ingestion or bodily fluoride levels. Instead, exposure was inferred simply from whether the community’s water was fluoridated during a person’s youth.
This is an extremely crude proxy – essentially an ecological metric. It assumes that if you lived in a fluoridated town, you drank enough tap water to matter. In reality, individual fluoride exposure varies with water consumption, use of filters, bottled water, etc.
The study could not account for these differences. Even the accompanying Science Advances commentary admitted “Assigning exposure based simply on whether an individual resides in an area served by fluoridated water may seem crude relative to studies that evaluate urinary fluoride concentrations” (which directly measure fluoride in the body).
The authors argue this is acceptable since policy is set by community fluoridation status, not personal urine levels. But that logic is circular – it essentially bakes the policy assumption (that everyone in a fluoridated area is significantly exposed) into the study design.
This ecological fallacy can easily mask susceptible subgroups or individual differences. For example, a child in a fluoridated city who mainly drank well water or milk would be misclassified as “exposed” when they weren’t, potentially diluting any real effect among those truly ingesting fluoride daily.
2. Non-Specific Dosage and Duration: Relatedly, the study lacked any granular data on how much fluoride each participant actually consumed or when. All fluoridation is not equal – “recommended levels” have changed over time (from ~1.0 ppm in the 1980s to 0.7 ppm today), and natural fluoride can vary.
The study essentially treated exposure as a yes/no binary (fluoridated community or not). It also presumed people stayed in one place through childhood. The Gizmodo article even noted “many people grow up in the towns where they went to high school” as justification, implying that if you went to high school in a fluoridated town, your whole childhood was fluoridated – which may not hold true for those who moved.
All this introduces misclassification error that would tend to obscure an exposure-harm relationship (biasing results toward “no effect”). It’s telling that Warren’s Focus commentary explicitly defends the coarse exposure metric, while conceding that precise biomarkers “may be helpful” for understanding individual intake.
In short, the study’s exposure assessment is a blunt instrument, arguably incapable of detecting anything but a very large effect.
3. Outcome Measures – Test Scores vs. IQ: Perhaps the most glaring issue is that this study did not actually measure IQ at all – despite every headline trumpeting “IQ”. Warren et al. used standardized test scores (in reading, math, vocabulary) as a proxy for cognitive ability.
While test scores do correlate with IQ to an extent, they also reflect quality of education, socio-economic factors, and other confounders. Even Warren openly admits “the study is not perfect… standardized test scores are not the same as an IQ test.”.
At age 60, they administered some cognitive assessments, but again not formal IQ tests. By using these indirect measures, the study opens itself to numerous confounding factors.
For example, fluoridated areas might differ systematically – perhaps fluoridated cities tend to have better-funded schools (since many big cities fluoridate) or other advantages that boost test scores.
The study did not publish a detailed adjustment for socio-economic status or school quality. If such factors weren’t fully controlled, the slight academic “benefit” it found for fluoridated groups could simply reflect those unmeasured advantages, not fluoride per se.
In fact, a Swedish study (Aggeborn & Öhman 2021) similarly found no IQ deficit with fluoridation and even better dental outcomes, but noted that isolating fluoride’s effect in observational data is tricky.
Warren’s study did attempt some statistical adjustments, but given the limitations of the proxy outcomes, any conclusion that fluoride does or doesn’t impact “IQ” is premature.
Warren himself is reportedly working on a follow-up that will “directly assess the link between fluoride and IQ” using actual IQ tests in a separate cohort – an implicit acknowledgment that this initial study was an approximation.
4. Selection and Nonresponse Bias: The High School and Beyond (HS&B) dataset used had major attrition over 40 years. Of the initial ~26,000 teens, only about half were followed up in 2021 with cognitive tests.
The rest dropped out of the study. If, for instance, those who struggled cognitively or were in poorer health (possibly including more heavily exposed individuals) were less likely to remain in the study, the results could be skewed.
The Science Advances commentary explicitly notes the attempt at a representative sample was “impaired by extensive nonresponse” in the follow-up. That is a polite way of saying there may be bias in who ended up counted.
Without careful analysis, we don’t know if the lost participants disproportionately came from certain communities or had certain traits. It’s quite conceivable that any subtle neurotoxic effects of fluoride might manifest in subpopulations that also have higher dropout rates (e.g. lower socioeconomic status, etc.), thus washing out of the final analysis.
The study’s authors downplay this concern, but it cannot be brushed aside.
5. Statistical Power and Reporting: Counterintuitively, an extremely large sample can sometimes mask small effects by focusing on aggregate differences. The Warren study looked at broad comparisons (ever vs never fluoridated, etc.).
If fluoride’s impact is modest or affects only a fraction of susceptible individuals (for example, children with particular nutritional deficiencies or genetic predispositions), a coarse analysis of thousands might dilute that signal.
Meanwhile, the media’s portrayal of the findings as “no effect, case closed” is an overstatement. What the study actually found was no statistically significant harm on these particular measures – which is not proof of absolute safety, just an absence of detected effect in this setup.
The authors themselves hedge that it “moves the needle” toward concluding fluoridation policy is not broken, yet also acknowledge “the study alone can’t tell us whether water fluoridation is a net positive” for overall health.
Nuances like that did not make the headlines.
In sum, the Warren et al. study – while valuable as a large-scale data point – has critical limitations that the coordinated media narrative conveniently glossed over.
Instead, we got triumphalist sound bites: “no evidence” of harm, “IQ story doesn’t hold up”, “massive study debunks… claims”.
The reality is far murkier. By design, the study could easily miss subtle developmental effects of fluoride, especially given it didn’t measure actual IQ or exposure levels.
As one commentator (Dr. David Savitz) noted in the Science Advances editorial, the focus on lifetime cognitive ability (test performance) is “broadly relevant” but “less precise than rigorous neurobehavioral testing in children.”
Even Warren conceded the study “is not perfect”. Yet perfection was exactly how the media treated it – as the decisive trump card against decades of dissenting science.
What the Media Omitted: Decades of Evidence on Fluoride & Neurotoxicity
Perhaps even more telling than the flaws downplayed is the evidence outright ignored in this media blitz. By focusing solely on the new Warren study (and hailing it as the definitive word), the coverage omitted a large and growing body of research that underpins RFK Jr.’s and others’ concerns about fluoride’s neurodevelopmental effects.
This omission skews public understanding, because casual readers would think “no link found, end of story” – apparently without realizing numerous high-quality studies have found links between fluoride and reduced IQ or other neurological harm:
- Bashash et al. (2017) – A NIH-funded study in Mexico that measured pregnant women’s urinary fluoride (a direct exposure biomarker) and later tested their children’s IQ. It found a clear association: higher prenatal fluoride levels were linked to lower IQ scores in offspring. Notably, Mexico doesn’t fluoridate water, but many residents ingest fluoride through fluoridated salt and other sources. Bashash et al. was a high-quality study (published in Environmental Health Perspectives), controlling for confounders, and it reported that an increase of 1 mg/L in maternal urine fluoride corresponded to about a 5-6 point IQ decrement in children – a potentially significant effect.
- Green et al. (2019) – A study in Canada (published in JAMA Pediatrics) that made waves for similar reasons. Green’s team examined pregnant women in fluoridated vs. non-fluoridated cities (Toronto vs Montreal, primarily) and their children’s IQ scores. They found that boys born to mothers with higher fluoride exposure had significantly lower IQ (the effect was less clear in girls). This study took place in communities with ~0.7 ppm fluoride in water – exactly “recommended levels” – making it highly relevant to U.S. policy. The findings provoked intense debate and even pushback from fluoridation proponents, but the methodology was rigorous. One of the co-authors, Dr. Christine Till, has since published further research suggesting neurodevelopmental risks (including a 2020 study linking early-life fluoride exposure to ADHD symptoms).
- Till et al. / Riddell et al. (2020) – Following up in Canada, researchers looked at infants fed formula reconstituted with fluoridated water versus non-fluoridated water. They found that babies in fluoridated areas had markedly lower IQ by age 3-4, presumably because using fluoridated water in formula during a critical brain development window increased fluoride exposure compared to breastfed or non-fluoridated counterparts. This study (Riddell/Till 2020) underscored concerns about bottle-fed infants, who get a high dose per body weight if formula is mixed with fluoridated tap water. Again, none of the Nov 19 media stories reference this vulnerable subgroup or any infant studies.
- Spanish & Danish Cohorts (2021-2022) – In Spain, a study (Ibarluzea et al. 2022) similarly examined prenatal fluoride exposure and toddler development, while in Denmark, another (2019/2021) used nationwide data to see if community fluoride levels associated with intelligence. Results were mixed – the Spanish study oddly found a positive association (which some suspect might be due to methodological quirks), and the Danish study (at very low fluoride levels) found no effect. Rather than conclusively exonerating fluoride, experts have viewed these as part of an “array of inconsistent findings” that still leave questions open. The media articles did not delve into any of this nuance; they simply presented the new Warren study as if it single-handedly settles what dozens of previous studies have been exploring.
- Grandjean et al. (2024) – Notably, one of the newest papers before Warren’s was a cross-analysis by Dr. Philippe Grandjean (a renowned environmental health scientist) and colleagues, published in European Journal of Public Health in 2024. They pooled data from multiple prospective cohorts (including the Mexico City and Canadian studies) to assess fluoride’s dose-response relationship with IQ. Grandjean’s analysis reported dose-dependent cognitive effects – higher prenatal fluoride associated with lower cognitive scores, even at levels down to around 0.2–0.4 mg/L in maternal urine (which corresponds to typical U.S. exposures). In plain language, it suggested there may be no clear safe threshold for fluoride’s neurotoxicity, or that the “safe” level is lower than current policy. This directly challenges the notion that sticking to 0.7 ppm in water is beyond reproach. Again, zero mention of this in mainstream coverage.
- The National Toxicology Program Draft (2022) – To reiterate, the NTP’s meta-analysis of 27 studies found strong evidence that elevated fluoride exposure is associated with reduced IQ in children, and it rated this evidence as of “high confidence” for fluoride as a developmental neurotoxin in humans. The draft NTP report (eventually leaked in 2023) also found that such effects were observed even in subsets of studies at ~1.5 ppm and below. This undercuts the media narrative that only crazy high exposures show harm. Indeed, in 2021 a member of the NTP panel, Dr. Bruce Lanphear (co-author on Green 2019), commented that the loss of IQ from maternal fluoride at 0.7–1.0 ppm in water might be on the order of 3-5 IQ points population-wide – a subtle effect, but enough to be concerning if true. None of this context made it into the glowing press coverage of the Warren study.
This is taken from a long document. Read the rest here substack.com