Meta-analysis Finding of a Fluoride-IQ Link Controversial

Results published from a systematic review and meta-analysis find an inverse association and a dose-response link between fluoride exposure and children’s IQ scores.

Kyla W. Taylor, PhD, with the Division of Translational Toxicology at the National Institutes of Health, Morrisville, North Carolina, led the multicountry study published online in JAMA Pediatrics.

Two accompanying editorials offer two very different perspectives on how to interpret the researchers’ conclusions.

Study Results

The authors note that of the 74 studies included in the review (64 cross-sectional and 10 cohort studies), most (45) were conducted in China. Other countries included were Canada (3), Denmark (1), India (12), Iran (4), Mexico (4), New Zealand (1), Pakistan (2), Spain (1), and Taiwan (1). “Fifty-two studies were rated high risk of bias, and 22 were rated low risk of bias,” the authors stated.

Researchers found that 64 of the 74 studies reported inverse associations between fluoride exposure measures and children’s IQ scores. Their analysis of 59 studies with group-level measures of fluoride in drinking water, dental fluorosis, or other measures of fluoride exposure showed an inverse relationship between fluoride exposure and IQ (pooled Standardized Mean Difference [SMD], −0.45; 95% CI, −0.57 to −0.33; P 

In 31 studies that reported fluoride measurements in drinking water, a dose-response relationship was found between exposed and reference groups (SMD, −0.15; 95% CI, −0.20 to −0.11; P

In 20 studies reporting fluoride measured in urine, there was an inverse dose-response association (SMD, −0.15; 95% CI, −0.23 to −0.07; P

For perspective, in the United States, the US Public Health Service in 2015 lowered the recommended concentration of fluoride in drinking water from a range of 0.7-1.2 mg/L to 0.7 mg/L to reduce the risk for dental fluorosis while keeping its protective effect against dental caries. 

When Taylor’s team analyzed 13 studies with individual-level measures, they found an IQ score decrease of 1.63 points (95% CI, −2.33 to −0.93; P P

The authors conclude both that inverse relationships and a dose-response association between fluoride measured in urine and drinking water and children’s IQs were found across the literature examined but also that “there were limited data and uncertainty in the dose-response association between fluoride exposure and children’s IQ when fluoride exposure was estimated by drinking water alone at concentrations less than 1.5 mg/L.”

The authors point out that, “To our knowledge, no studies of fluoride exposure and children’s IQ have been performed in the United States and no nationally representative urinary fluoride levels are available, hindering application of these findings to the US population.”

Editorial: Time to Reassess Systemic Fluoride

Bruce P. Lanphear, MD, MPH, with Simon Fraser University, Vancouver, British Columbia, Canada, is the lead author on an editorial that suggests these data point to the need to reassess systemic fluoride exposure.

“Their study is the largest and includes the most rigorous series of meta-analyses of fluoride ever conducted,” Lanphear and colleagues wrote. “It is time for health organizations and regulatory bodies to reassess the risks and benefits of fluoride, particularly for pregnant women and infants.”

Lanphear’s team says distinguishing between water fluoride and urinary fluoride levels is important in these results “because regulatory and public health agencies must consider total fluoride intake when assessing risks.”

Taylor and colleagues’ finding that there was no statistically significant association between water fluoride

Editorial: Be Cautious About the Conclusions

Steven M. Levy, DDS, MPH, cites “major areas of concern” in the meta-analysis in his editorial.

He points to the large majority of studies in the meta-analysis that were at “high risk of bias” (47 high risk vs 12 that were low risk). He also cited information from a further look at the low-risk-of-bias studies included in the supplement.

“The studies with lower risk of bias showed a negligible effect (standardized mean difference [SMD], −0.19; 95% CI, −0.35 to −0.04) with very high heterogeneity (I² = 87%), and a majority of publications (8 of 12) did not show a negative association between fluoride and childhood IQ,” Levy writes. 

“Taylor et al do not adequately justify selection or omission of studies or explain or justify the calculated individual effect sizes presented in the main analysis. Also, readers are not told which studies with lower risk of bias are included in the subanalyses for water fluoride levels less than 1.5 mg/L, less than 2.0 mg/L, and less than 4.0 mg/L; therefore, readers cannot independently consider important differences across these studies.”

Levy also states that the magnitudes of the possible IQ differences are unfairly inflated. For the United Staes and most of the world, he points out, the recommended community water fluoridation level is 0.7 mg/L. Therefore, the difference between a community with low fluoride levels (about 0.2 mg/L) and one with optimal levels is about 0.5 mg/L. 

“However, Taylor and colleagues use a difference of 1.0 mg/L in their calculations, artificially doubling the estimated impact on IQ,” Levy writes.

The meta-analysis should not affect public policy on adding fluoride to community water systems “and the widespread use of fluoride for caries prevention should continue,” Levy concludes.

Concerns About Quality of Studies Included

Charlotte Lewis, MD, MPH, associate professor of pediatrics at the University of Washington School of Medicine and part of Seattle Children’s Multidisciplinary Infant Nutrition and Feeding Team, Seattle, Washington, who was not involved in the meta-analysis or editorials, also told Medscape Medical News that systemic fluoridation should not change based on these results, citing what she says are problems with methodology.

“There are many concerns about the quality of studies included in this meta-analysis,” Lewis said. “Although the authors claim to have separated out low-bias studies, it is important to note that many of these same studies have substantial methodological flaws.”

She said studies deemed low-bias and included in the meta-analysis “relied on multiple examiners for cognitive testing without consideration for inter-rater variation or reliability measures.” She added that “a number of the studies failed to account for maternal IQ scores, breastfeeding, lead exposure, or other factors that could affect cognitive development, further contributing to biased conclusions.”

Importantly, she said, many of the studies, including one by Green et al published in JAMA Pediatrics in 2019, relied on maternal spot urinary fluoride to assess fetal exposure to fluoride. “This is not a valid way to assess fetal exposure to fluoride and including such studies in this meta-analysis has led to inappropriate conclusions because they are based on studies using a flawed exposure measure.”

She pointed to recent longitudinal, population-based studies, including one by Kumar et al that have found no adverse impact on IQ, or other cognitive tests, of drinking water with low levels of fluoride present, comparable to US community water fluoridation standards. 

“Relative to the small convenience-sample, cross-sectional studies included in this meta-analysis, longitudinal, population-based studies are considered significantly more reliable for establishing cause and effect,” she said.

Fluoride Levels Different Globally

Lewis said in some parts of the world fluoride is present in the environment in much higher levels than in fluoridated water in the United States.

“There are known adverse health effects of high fluoride ingestion in these endemic regions found primarily in China, India, and Iran. This points to the importance of dose response. What is beneficial at low levels can be toxic at high levels and that appears to be the case, not surprisingly, for fluoride as well. However, at 0.7 ppm, the level of fluoride in community water fluoridation, we experience fluoride’s beneficial effects when we regularly drink optimally fluoridated water.”

“Water fluoridation is an important public health approach available and beneficial to all, even those unable to afford or access dental care,” she says. “Water fluoridation diminishes oral health disparities, and its removal threatens to worsen disparities and increased suffering from dental disease. I remain confident in the benefits and safety of community water fluoridation.”

Taylor and colleagues reported no relevant financial relationships. 

Editorialist Lanphear reported grants from the National Institute of Environmental Health Sciences (NIEHS) and the Canadian Institute for Health Research (CIHR) and having served as a nonretained and unpaid expert witness in a federal fluoride suit against the US EPA. Coeditorialist Till reports grants from the NIEHS and CIHR, an honorarium from the Irish Health Research Board for consulting work, and an honorarium and travel support for a lecture from Icahn School of Medicine at Mount Sinai, New York. 

Levy reported past grants from the National Institute of Dental and Craniofacial Research related to fluoride, dental caries, dental fluorosis, and bone development. He reported small grant funding from the Centers for Disease Control and Prevention related to fluoride, dental caries, and fluorosis. He consults for the Centers for Disease Control and Prevention and the National Institute of Dental and Craniofacial Research and serves on the National Fluoride Advisory Committee for the American Dental Association.

Marcia Frellick is an independent healthcare journalist based in Chicago.