Several studies have been published in the fall of 2026 that offer a detailed look into how fluoride, at levels of exposure commonly seen in fluoridated communities, can affect the brain.

ECHO Study 1 (Simon 2026): Prenatal exposure to fluoride in public water and child cognition in the US ECHO cohort, 2006-2019

The study, authored by a 25-member team of scientists and physicians from 13 universities and medical schools across the nation, followed 2,514 children in 23 states. It found that fluoride in public drinking water during pregnancy is linked to lower cognition scores in children – even at levels considered “optimal” for dental health. 

The study found that for every 0.5 mg/L increase of fluoride level above 0.675 mg/L, cognition scores decreased by an average of 3.36 points. However, when tests under age 7 were included, the losses were much higher, at 13.5 points.

The ECHO study found diminished cognition levels s at 0.675 milligrams per liter (mg/L) or higher, close to the current fluoridation levels of 0.7 mg/L. However, this is an average and should not be interpreted to mean that water fluoride concentrations at or below 0.7 mg/L can be considered safe. This study shows there is no margin of safety.

Its findings complement the comprehensive 2024 National Toxicology Program systematic review of studies that consistently linked higher fluoride levels in pregnant women with lower IQs in their children.

ECHO Study 2 (Hernandez 2026): Prenatal Water Fluoride Exposure and Children’s Behavioral Problems

This cohort study used observational pregnancy cohort data and found that exposure to prenatal fluoride levels in regulated public water was NOT associated with internalizing and externalizing problems.

Reading past the abstract’s plain-language summary, the study is not actually a clean null result. Its own pre-specified sensitivity analysis found a statistically significant, dose-dependent association between prenatal fluoride exposure and externalizing behavior problems above a threshold that sits below the fluoridation target.

Study Analysis:

• The exposure measure is a residence-based proxy, not an individual biomarker — the authors say so themselves. Their own Limitations section states plainly that water fluoride concentration “represents a residence-based proxy and may not capture individual-level intake or total fluoride exposure,” and that this “measurement error … may reduce power to detect an association.” This is not this comment’s inference; it is the study authors’ own stated conclusion, and it supports the same principle Appendix B applies to Simon et al.: a water-only exposure measure is expected to understate, not overstate, a true effect.
• A more serious version of that problem: the design cannot distinguish a household that drinks the local tap water from one that avoids it — and avoidance is not randomly distributed. Parents who filter their water, buy bottled water, or otherwise avoid tap water in a higher-fluoride area are often, as a matter of general behavior clustering documented across public health research, the same parents most likely to adopt other protective practices — extended breastfeeding, a less-processed or organic diet, closer attention to other environmental exposures — each of which has its own independent, separately documented association with better child behavioral outcomes. This is a recognized phenomenon sometimes called “healthy-user bias”: one protective choice tends to travel with several others, and a design that measures only the fluoride available at an address, rather than what the household actually consumed or did, cannot separate fluoride’s own effect from the effect of everything else that comes bundled with the choice to avoid it.
• The study could not account for postnatal fluoride exposure or any other fluoride source — diet, tea, dental products — again by the authors’ own statement. It measures one source of exposure among several, not a child’s or mother’s total fluoride burden.
• Most important: the study’s own headline framing understates its own results. The abstract’s plain-language “Findings” statement reports only that fluoride “was not associated with internalizing and externalizing problems,” based on the continuous linear models (adjusted mean difference per 500 µg/L: ?0.15 for internalizing, 0.06 for externalizing — genuinely null at that level). But the study’s own pre-specified change-point sensitivity analysis found a statistically significant, positive, dose-dependent association for externalizing behavior problems above an inflection point of 625 µg/L: an adjusted mean difference of 1.59 points (95% CI, 0.72–2.47) per 500 µg/L increase beyond that threshold. That 625 µg/L is strikingly close to the 675 µg/L change point Simon et al. (2026) identified for cognitive outcomes in a substantially overlapping ECHO sample. A reader relying on the abstract’s Key Points alone would not learn this; it appears only in the full Results and Discussion.

Ashley Malin further explains contradictory evidence in the literature.

Long-term exposure to prenatal fluoride and social-emotional and behavioural outcomes in children in the New Hampshire Birth Cohort Study (Till/Lanphear 2026)

Found that higher maternal toenail fluoride concentrations, but not water fluoride concentrations, were associated with greater odds of at-risk externalizing problems in preschoolers.

Association of Maternal Bone Resorption with Urinary Fluoride Levels During Pregnancy (Bashash/Hu 2026)

What are the main findings?

Maternal bone resorption was independently and positively associated with maternal urinary fluoride in each trimester of pregnancy. This is consistent with fluoride stored in the maternal skeleton being released into the body as bone breaks down during pregnancy.

What is the implication of the main finding?

Maternal urinary fluoride during pregnancy appears to reflect both external fluoride exposure and endogenous fluoride released from the maternal skeleton during bone resorption. This has implications for how prenatal fluoride exposure is measured and interpreted, since urinary fluoride may capture a bone-derived component in addition to recent exposure.