Could Fluoride From Tap Water During Pregnancy Affect Children’s Thinking? New Study Finds A Link

Fluoride in tap water has long been celebrated as one of public health’s greatest achievements, protecting teeth across generations. A large new study is now raising uncomfortable questions about whether prenatal exposure to it is linked to lower scores on tests of how children think and reason, particularly when exposure begins before birth.

Researchers analyzed data from more than 2,500 children born across 23 U.S. states between 2006 and 2019, estimating the fluoride levels in the public water serving each mother’s home during pregnancy. The findings were published in the American Journal of Epidemiology. What they found adds weight to growing concerns: higher prenatal fluoride exposure was linked to lower scores on tests measuring a child’s ability to reason, learn new things, and process information, with the effect most pronounced in younger children.

These associations appeared at fluoride concentrations below the thresholds currently set by major health agencies, raising fresh questions about whether existing safety limits are adequate to protect fetal brain development. With millions of Americans drinking water that contains fluoride above the levels where researchers began seeing a measurable effect, the findings are hard to dismiss.

What the Fluoride Study Measured and How

Not all thinking works the same way. Scientists distinguish between two broad categories: one reflects raw mental processing, including the ability to solve new problems, hold information in mind, and adapt thinking on the fly. The other reflects knowledge and skills built up over time through education and experience, like vocabulary. Prior research had suggested the first type might be more vulnerable to environmental exposures. This study tested that idea using a nationally recognized cognitive assessment tool developed by the National Institutes of Health.

Drawing on the Environmental influences on Child Health Outcomes (ECHO) Cohort, a large research effort spanning 17 study sites across the United States, the team examined data on 2,514 children. Researchers estimated each mother’s fluoride exposure through public drinking water during pregnancy by linking her residential address history to fluoride measurements that water systems routinely report to the U.S. Environmental Protection Agency. This allowed month-by-month exposure tracking across an entire pregnancy, even when families moved.

Prenatal fluoride concentrations in the sample ranged widely, from barely detectable levels up to 1,940 micrograms per liter, with an average of about 397 micrograms per liter. For context, the U.S. Public Health Service recommends an optimal fluoride concentration of 700 micrograms per liter, the World Health Organization sets a guidance limit of 1,500 micrograms per liter, and the EPA’s maximum legal limit sits at 4,000 micrograms per liter.

Where Fluoride’s Link to Lower Scores Gets Steeper

Fluoride’s relationship with brain test scores was not a simple straight line. Instead, the association appeared to level off at lower concentrations before declining more noticeably above a certain point.

Using an analysis designed to find where the relationship changes most sharply, the researchers identified 675 micrograms per liter as the concentration at which scores on the raw-processing cognition test began to fall more steeply. Above that point, every additional 100 micrograms per liter of fluoride in the water was associated with a 0.67-point drop in cognitive scores. Scaled up to a 500-microgram-per-liter increase, that translates to a potential decline of roughly 3.36 points. A separate curve-fitting analysis found statistically significant drops in those same scores at concentrations above about 1,107 micrograms per liter.

Scores on the vocabulary-based, knowledge-built-over-time type of cognition showed no statistically significant association with fluoride levels, consistent with earlier research.

Young Children Showed the Strongest Link to Prenatal Fluoride

One notable finding emerged when researchers split the sample by age. Children under 7 showed considerably larger declines in raw-processing cognitive scores associated with higher prenatal fluoride than children 7 and older. For the younger group, the drop above the 675-microgram threshold was 2.70 points per 100 micrograms per liter of additional fluoride, compared to 0.45 points for the older group. Scaled to a 500-microgram increase, that represents a potential decline of up to 13.5 points among younger children.

Researchers offered a plausible explanation: the youngest children are still in a phase of very rapid brain development, which may make the downstream effects of prenatal exposure more visible and measurable at that age. As children grow older, accumulated experiences like schooling may partially mask or compensate for earlier developmental effects, though associations remained detectable even in the older group.

Because the exposure measurements came directly from regulatory compliance monitoring data rather than individual urine samples or self-reported water consumption, the findings translate more cleanly into a policy context than some prior studies. Over 2.9 million U.S. residents currently receive public water with fluoride concentrations exceeding 1,500 micrograms per liter, according to estimates cited by the authors. While many public water systems that add fluoride intentionally keep levels near the recommended 700-microgram mark, higher concentrations often occur naturally in source water.

Authors of the study state that these findings “highlight the need to re-evaluate the current USEPA MCL for fluoride in public water systems (4,000 µg/L), which may not adequately protect neurodevelopment.” They also note that the dental health benefits of water fluoridation need to be weighed against these potential risks, pointing to a recent review suggesting those benefits may be smaller today than in past decades, when fluoride toothpaste was less widely used.

Several limitations temper the conclusions. Fluoride concentrations were measured at the neighborhood level rather than at individual taps, meaning some families may have been misclassified. Researchers could not account for bottled water use, filtration, or fluoride from food and other sources, and they did not measure postnatal fluoride exposure. When researchers applied a statistical correction for the possibility that some families in the study were not representative of the broader population, the effect estimates became smaller, though they still pointed in the same direction. The study sample also skewed White, non-Hispanic, and college-educated. Despite those caveats, results held up across a wide range of sensitivity checks, including analyses that removed individual study sites one at a time and adjusted for additional potential confounding factors.

Fluoridated water remains one of the most debated topics in public health and environmental science. This study does not call for an immediate ban on water fluoridation, and its authors explicitly acknowledge fluoride’s established value for dental health. But with a dataset spanning nearly two decades and more than two dozen states, the link between prenatal fluoride exposure and measurable differences in children’s cognitive test scores is a result regulators and policymakers will have a hard time setting aside.

Disclaimer: This article summarizes findings from a single observational study. Observational research can identify associations but cannot prove that fluoride causes changes in cognition, and the authors note limitations including neighborhood-level exposure estimates, an unmeasured amount of individual water intake, and a sample that skewed White, non-Hispanic, and college-educated. The findings do not establish that water fluoridation is unsafe, and the study’s authors acknowledge fluoride’s established dental-health benefits. Anyone with questions about fluoride and pregnancy should consult a healthcare provider.

---