The National Toxicology Program (NTP), part of the U.S. Department of Health and Human Services, published a systematic review in 2024 examining 72 studies on fluoride and neurodevelopment. The review took over six years and underwent multiple rounds of peer review.

The NTP concluded with "moderate confidence" that fluoride exposures above 1.5 mg/L in drinking water are consistently associated with lower IQ in children. Of 22 studies rated as higher quality, the majority showed inverse associations between fluoride exposure and IQ scores.

The pooled effect across studies showed approximately 0.45 standard deviations lower IQ in high-fluoride groups, which translates to roughly 6-7 IQ points on a standard scale. This was one of the most comprehensive government-sponsored reviews of fluoride neurotoxicity ever conducted.

Researchers at the Harvard School of Public Health and China Medical University conducted a systematic review and meta-analysis of 27 epidemiological studies (mostly from China) examining the relationship between fluoride exposure and children''s intelligence.

The pooled analysis found a standardized weighted mean difference of 0.45, which translates to approximately 7 IQ points on a standard scale with a standard deviation of 15. Of the 27 studies included, 26 showed that children in high-fluoride areas had lower IQ scores than children in low-fluoride areas.

The studies compared villages with naturally high fluoride levels (typically 2-10 mg/L) to villages with lower levels (typically under 1 mg/L). The consistency of the finding across 26 of 27 studies from different regions, time periods, and research groups strengthens the case that the association is not due to chance or a single methodological flaw.

Published in JAMA Pediatrics in 2019, researchers from York University and other Canadian institutions studied 601 mother-child pairs from the MIREC (Maternal-Infant Research on Environmental Chemicals) cohort, a multi-city Canadian birth cohort.

Maternal urinary fluoride was measured during pregnancy and children''s IQ was assessed at ages 3-4 years. A 1 mg/L increase in maternal urinary fluoride was associated with a 4.49-point decrease in IQ in boys (95% confidence interval: -8.38 to -0.60). The effect in girls was smaller and not statistically significant.

This was significant because it was the first high-quality prospective study from a country with community water fluoridation (Canada fluoridates at 0.7 mg/L) to find an association. Previous evidence came mostly from cross-sectional studies in high-fluoride regions. The prospective design and individual-level biomarker measurement addressed many limitations of earlier research.

Published in Environment International in 2020, researchers examined 398 mother-child dyads from the MIREC birth cohort, focusing specifically on fluoride exposure during infancy through reconstituted formula.

Infants who were formula-fed in fluoridated communities showed a 9.3-point decrease in Performance IQ compared to formula-fed infants in non-fluoridated communities. Breastfed children showed a smaller 6.2-point decrease associated with each 0.5 mg/L increase in water fluoride concentration.

The formula feeding pathway is particularly concerning because formula reconstituted with fluoridated water provides a much higher fluoride dose per kilogram of body weight to infants compared to adults drinking the same water. A formula-fed infant in a fluoridated area receives approximately 100-200 times more fluoride per kilogram than a breastfed infant, creating a natural dose contrast for studying effects.

Published in Environmental Health Perspectives in 2017, researchers from the University of Toronto, University of Michigan, and Mexican institutions studied 299 mother-child pairs from the ELEMENT (Early Life Exposures in Mexico to Environmental Toxicants) project in Mexico City.

Maternal urinary fluoride was measured during pregnancy and children were assessed for cognitive function at age 4 (using the Bayley Scales) and ages 6-12 (using the WASI IQ test). Each 0.5 mg/L increase in maternal urinary fluoride predicted 3.15 fewer IQ points at age 4 (95% CI: -5.42 to -0.87) and 2.50 fewer points at ages 6-12 (95% CI: -4.12 to -0.59).

The study was notable for its prospective design, individual-level exposure measurement (urinary biomarkers rather than ecological water levels), and long follow-up period. Mexico City does not fluoridate water but has fluoride exposure through naturally occurring fluoride and fluoridated salt programs.

Published in JAMA Pediatrics in 2025 by NTP researchers, this meta-analysis synthesized 74 studies (64 cross-sectional, 10 cohort) from 10 countries involving 20,932 children total.

The pooled standardized mean difference was -0.45 (95% CI: -0.57 to -0.33), indicating lower IQ in higher-fluoride-exposed children. In dose-response analysis, every 1 mg/L increase in urinary fluoride was associated with a 1.63-point decrease in IQ (95% CI: -2.33 to -0.93, p<0.001). The inverse association persisted when the analysis was restricted to studies rated as having low risk of bias.

This was the most comprehensive quantitative synthesis to date and was notable because the dose-response relationship showed effects beginning at relatively low fluoride concentrations, not just at very high levels. The consistency across 10 different countries also argues against confounding by any single cultural or regional factor.

Published in Environmental Health in 2024, researchers combined data from three prospective cohorts: the Odense Child Cohort (Denmark), MIREC (Canada), and ELEMENT (Mexico), totaling over 1,500 mother-child pairs with individual-level prenatal fluoride measurements and child IQ assessments.

Using benchmark dose analysis, they estimated that the concentration of maternal urinary fluoride associated with a 1-point IQ loss was 0.45 mg/L, with a lower confidence limit (BMCL) of 0.28 mg/L. A 1 mg/L increase in maternal urinary fluoride was associated with a 2.06-point decrease in IQ across the combined cohorts.

The benchmark dose approach is considered more rigorous than simple threshold analysis because it estimates the dose at which a defined adverse effect begins, using the full dose-response curve rather than relying on arbitrary cutoff comparisons. The finding that the benchmark concentration falls below levels commonly seen in fluoridated communities raised concerns about the adequacy of current fluoridation standards.

Published in Fluoride (the journal of the International Society for Fluoride Research) in 2003, Chinese researchers compared cognitive development in children from two villages in Jiangsu Province with different naturally occurring fluoride levels.

Wamiao village had drinking water fluoride at 2.47 mg/L while Xinhuai village had 0.36 mg/L. Among 222 children aged 8-13 years tested in the high-fluoride village, the mean IQ was 92.02, significantly lower than in the control village. Using benchmark dose analysis, the researchers estimated that the concentration at which IQ drops below 80 (intellectual disability threshold) was 2.32 mg/L, with a lower confidence limit of 1.85 mg/L.

While this study examined levels above those used in community water fluoridation, it provided early dose-response data suggesting that fluoride''s neurotoxic effects follow a continuous relationship without a clear "safe" threshold. The study was later cited in the Harvard meta-analysis and the NTP review as part of the cumulative evidence base.

Published in Environmental Health in 2015, researchers from York University used CDC water fluoridation statistics (1992-2008) and parent-reported ADHD prevalence data for children aged 4-17 years (2003-2011) to examine associations at the state level.

After controlling for median household income (a proxy for socioeconomic status and healthcare access), each 1% increase in artificial fluoridation prevalence was associated with approximately 67,000 to 131,000 additional ADHD diagnoses nationally. States with higher fluoridation prevalence had significantly higher rates of ADHD diagnosis.

While ecological studies cannot establish causation (they compare populations, not individuals), the finding is relevant because ADHD and reduced IQ both involve disrupted neurodevelopment. The biological plausibility is supported by animal research showing fluoride accumulates in the hippocampus and prefrontal cortex, brain regions involved in attention and executive function.

Published in Risk Analysis in 2021 by Philippe Grandjean and colleagues, this study applied benchmark dose modeling to individual-level data from the MIREC (Canada) and ELEMENT (Mexico) prospective cohorts.

The estimated benchmark concentration (BMC) of maternal urinary fluoride associated with a 1-point IQ loss was 0.33 mg/L, with a benchmark concentration lower confidence limit (BMCL) of 0.2 mg/L. Based on pharmacokinetic models relating water intake to urinary fluoride, these values correspond to drinking water fluoride concentrations below the current U.S. recommended level of 0.7 mg/L.

This was the first benchmark dose analysis applied specifically to fluoride neurotoxicity at community-relevant exposure levels. The approach is the same methodology used by regulatory agencies to set safe exposure limits for other toxic substances, and the finding that the benchmark dose falls below the fluoridation level directly challenges the assumption that 0.7 mg/L provides an adequate margin of safety for neurodevelopmental protection.