Fluoride in Drinking Water

Public water fluoridation has been a core component of preventing dental caries for more than 70 years. Studies before 1975 consistently demonstrated substantial reductions in decayed, missing, or filled teeth among children because of water fluoridation,1 which may have averted billions of dollars per year globally in health care costs. Largely considered safe at drinking water concentrations of 0.7 mg/L or lower, public water fluoridation has long been viewed as one of the great public health achievements of our time. In the United States, 62.8% of the population received fluoridated water in 2022.2Increasingly, the average American is exposed to other sources of fluoride. Fluoride is commonly found in toothpastes, mouthwashes, tooth varnishes, and gels. As such, the benefits of public water fluoridation may be reduced. A 2024 Cochrane review found that the substantial reductions in adverse dental outcomes seen before 1975 (before widespread use of fluoride-containing toothpaste) had been attenuated.1 Five studies3–7 among 5709 children before 1975 found that on average, children younger than 8 years living in areas with fluoridated water had 2 fewer decayed, missing, or filled primary teeth per year compared with children living in areas without fluoridated water. Conversely, 2 studies8,9 among 2908 children after 1975 found that fluoridated water averted an average of 0.24 decayed, missing, or filled primary teeth per child per year, with the pooled effect estimates no longer statistically significant.1 That is not to say that fluoride should be removed from drinking water; rather, any discussion of the benefits and harms must be informed by recent data.The National Institute for Occupational Safety and Health guidelines consider urinary fluoride concentrations of 0.2 to 3.2 mg/L as expected for individuals without occupational exposures.10 Approximately 50% of the ingested fluoride is retained in body tissues, while the other half is excreted in urine.11 Exposure to high concentrations of fluoride in drinking water (>4 mg/L) has been associated with numerous adverse health impacts. Several studies over the last decade have documented worse neurocognitive outcomes among infants and children exposed to fluoridated water at concentrations between 0.5 and 1.38 mg/L (Table).Among 512 mother-child pairs in Canada, a 1-mg/L increase from a median of 0.41 mg/L (range, 0.06–2.44 mg/L) in maternal urine fluoride concentration during pregnancy was associated with a 4.5-fold reduction in IQ scores (95% CI, −8.38 to −0.06) among boys aged 3 to 4 years.12 A prospective study in Canada comparing women in areas with access to fluoridated water throughout pregnancy against those in areas without fluoridated water (n = 616 mother-child dyads) did not find significant reductions in IQ scores but did observe a significant reduction in performance on standardized assessments of inhibition control among children aged 3 to 4 years overall (aOR, 0.53; 95% CI, 0.31–0.93) and among girls (aOR, 0.30; 95% CI, 0.13–0.74).13 Among 299 mother-child pairs in Mexico, each 0.5-mg/L increase in maternal urine fluoride concentration from a mean of 0.9 mg/L (SD, 0.35) was associated with a 3-fold reduction in IQ scores among 4-year-olds (95% CI, −5.42 to −0.87) and a 2.5-fold reduction in IQ scores among children between 6 and 12 years old (95% CI, −4.12 to −0.59).14 A similar study found significant increases in measures of attention-deficit/hyperactivity disorder.15 Finally, among a cohort of 229 women living in California, each 0.7-mg/L increase in maternal third-trimester spot urine fluoride concentration from a median of 0.76 mg/L (IQR, 0.51–1.19 mg/L) was associated with a 1.8-fold increase in the odds of infants aged 36 months demonstrating neurobehavioral problems (95% CI, 1.17–2.86) after controlling for blood lead levels.16 A 2025 meta-analysis of 59 studies (n = 20 932 children) found a dose response between prenatal fluoride exposure and lower mean standardized IQ scores among children (n = 3 studies; pooled standardized mean difference, −1.70; 95% CI, −4.23 to 0.84) and postnatal fluoride exposure (n = 10 studies; pooled standardized mean difference, −1.65; 95% CI, −2.39 to 0.90).17While the epidemiologic evidence is compelling, interpretation of the research findings can be challenging. Spot urine measurements can be impacted by time of collection with respect to fluoride intake, hydration status, and duration of fluoride accumulation and thus do not necessarily accurately reflect total fluoride exposure.18 Additionally, even though statistically significant differences were observed, the absolute changes in the measured outcomes were small and of unclear clinical significance.Furthermore, it is not possible to assign causality to outcomes influenced by multiple factors and that occur many years after an exposure. The studies highlighted were observational in nature; thus, unmeasured influences on cognitive outcomes may have impacted results. Additionally, the measured outcomes differed between studies (eg, IQ at 3–4 years, IQ at 6–12 years, inhibition control at 3–4 years, symptoms of attention-deficit/hyperactivity disorder at 6–12 years, and behavioral domains at 36 months). Even when studies measured the same outcome, results differed between studies; for example, one study reported an associated reduction in IQ scores among boys,12 another found an associated reduction in IQ scores across sexes,14 and another found no association with IQ scores.13 Likewise, the 2025 meta-analysis grouped various cognitive measures under the umbrella of IQ, some of which more accurately measure developmental rather than intellectual quotients. Furthermore, no studies from the United States were included in the meta-analysis; among included studies, most were found to have a high risk of bias and were from settings in which environmental fluoride levels can be substantially higher than in the United States.Such heterogeneity makes drawing a unified association of harm difficult and limits generalizing the findings to other communities. However, given multiple studies have now raised the possibility that neurocognitive outcomes of children exposed to fluoride in utero may be worse than those not exposed, attention must be paid to the potential harms of fluoridated water, particularly considering the lower certainty of benefits in the current era with fluoride-containing dental hygiene products.Rigorous studies are urgently needed to determine whether public water fluoridation worsens childhood neurocognitive outcomes. Prospective studies using national registries and standardized neurocognitive measurements may partially address the question. However, work among animal models using analogous fluoride concentrations will be important. Contemporaneously, additional studies into the benefits of water fluoridation in the era of fluoride-containing dental products will provide further context against which to weigh the potential risks. Should water fluoridation prove harmful, there are numerous additional questions that need to be answered, such as defining the gestational period during which the fetus is susceptible to fluoride toxicity, if and to what extent fluoride impacts development during early childhood, and a better understanding of the harms of fluoride avoidance during pregnancy on dental health.In the meantime, it is imperative that public health officials and physicians maintain transparency. Communities and families must be provided with an honest assessment of the risks and benefits of water fluoridation. In addition to being ethically justified, transparency will be an important step toward regaining trust while assuring preventive dental care for all populations.The benefits afforded by adding fluoride to drinking water in the early part of the 20th century are unequivocal. However, such benefits are likely far less today than they once were. Furthermore, we are now seeing mounting evidence of harm. Considering such uncertainty, and in accordance with the precautionary principal, those factors together warrant further research into the issue of fluoride safety and a critical rethinking of our public health strategy to prevent adverse dental health outcomes.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.