Recent headlines spoke to yet another study demonstrating fluoride’s detrimental impact on IQ. Both sides of the controversy took to the media to push their narrative. But it was all smoke and mirrors because, in September, the US District Court for California’s Northern District told the EPA that our current standard for fluoride in drinking water poses an unreasonable risk and must be reviewed. It is time to re-examine how science fares in Court.
Under the Amended Toxic Substances Control Act (TSCA), the Court determined whether fluoridation of drinking water at the current level of 0.7 mg/L poses an unreasonable risk under the Amended TSCA.
Under the Amended TSCA, plaintiffs must prove by a preponderance of the evidence that a chemical poses an unreasonable risk to health or the environment, including risks to susceptible populations, without consideration of costs. The Court reviews the issue “de novo,” meaning without deferring to the EPA's prior decisions. The Court determines, based upon expert testimony, whether the scientific evidence demonstrates a risk exists and that it is unreasonable. The Court found that while the evidence did not definitively prove harm, water fluoridation poses an unreasonable risk concerning reduced IQ in children, compelling additional EPA regulatory action.
Getting into the Details - TSCA risk evaluation
The EPA risk evaluation involves four steps: hazard assessment, exposure assessment, risk characterization, and risk determination. The Court followed this framework in making its judgment.
Hazard Assessment
The EPA begins by identifying a hazard or chemical exposure associated with an adverse effect; there is no requirement for proof of causation. There is broad agreement that fluoride is hazardous at certain levels. The National Toxicology Program (NTP) reviewed 72 studies linking maternal fluoride exposure during pregnancy to reduced IQ in children. The NTP concluded that fluoride exposure at or above 1.5 mg/L is associated with cognitive harm. The Court, based on the NTP report, found,
“exposure to the chemical fluoride is associated with the adverse effect of reduced IQ in children, and particularly in boys.”
Dose-Response Assessment
The EPA next evaluates the weight of scientific evidence to determine which adverse effects and high or medium-quality studies are suitable for dose-response assessment. After some legal back and forth over EPA procedures [1], which the Court sidestepped by addressing both approaches, the Court found,
“The weight of the scientific evidence regarding fluoride’s association with reduced IQ is sufficient to proceed to the dose-response assessment… As discussed previously, not every epidemiological study on fluoride has found associations with reduced IQ in children. However, the evidence at issue is overall consistent as to the finding that fluoride is associated with reduced IQ in children, and there is a vast amount of experimental support for the association.” [emphasis added]
Dose-response assessment begins with a "point of departure" (POD), the threshold at which a chemical transitions from safe to hazardous. There are three main approaches to establishing the POD:
- NOAEL - no observed adverse effect level, the highest exposure level with no significant adverse effects to the exposed population.
- LOAEL – lowest observable adverse effect level, the lowest dose with no significant adverse effects to the exposed population.
- Benchmark Dose (BMD) Approach - the chemical dose or concentration that produces a predetermined change in the response rate of an adverse effect.
BMD is the preferred approach. [2] The Court identified multiple valid points of departure (PODs) for evaluating fluoride's risks to IQ, supported by strong evidence. The Court found, based on the NTP findings, that the BMD and LOAEL were
“well-supported point of departure to utilize in the risk evaluation. …supported by a preponderance of high-quality evidence.”
Exposure Assessment
The EPA identifies sources and levels of chemical exposure under specific conditions of use, considering factors like duration, intensity, frequency, and population affected. These exposure estimates are crucial in risk characterization, where they are compared to the point of departure (POD) to assess risk. They consider a range of exposures, from typical to high-exposure scenarios, calculating a benchmark, Margin of Exposure (MOE) – the ratio of safe exposure to the estimated level of human exposure – a higher MOE denotes a larger buffer. [3]
The Court found fluoride exposure exceeded safe levels under all identified PODs, including the more conservative LOAEL 4mg/ threshold, even when human exposure estimates are halved.
Risk Determination
The EPA then determines whether the identified risk is unreasonable based on the hazard's severity, exposure extent, and vulnerable populations. Confidence in the data supporting the hazard and exposure assessments and the strength of the overall evidence play critical roles in scrutinizing the uncertainties or assumptions made during the evaluation process.
Exposure does not need to exceed hazardous levels to pose a risk; an inadequate safety margin is sufficient. A pooled analysis showed a 1-point IQ drop for every 0.28 mg/L of fluoride in maternal urine. Evidence shows typical US levels of maternal urinary fluoride (0.8-1.89 mg/L) lack sufficient safety margins and often exceed the hazard threshold, even under conservative estimates.
The EPA argued that maternal urinary fluoride is not ideal for assessing risk in drinking water because it reflects total fluoride exposure, including sources beyond drinking water, such as food and toothpaste. However, the Court found this metric appropriate and meaningful in evaluating aggregate fluoride exposure, particularly since fluoridated drinking water is the primary contributor.
The Court noted that even if maternal urinary fluoride was not used, evidence shows that fluoridated water poses a health risk. Using the most conservative hazard threshold of 4 mg/L in water fluoride, studies, including the NTP assessment, consistently show a significant association between fluoride exposure and IQ reductions in children. Applying the default, 10-fold safety margin, the safe fluoride exposure level would be 0.4 mg/L. The current "optimal" US fluoridation level of 0.7 mg/L exceeds this threshold by nearly double, posing potential risks to pregnant women and their offspring. Thus, fluoridated water presents a clear risk.
In ordering the EPA to reassess optimal recommendations for fluoridation of public drinking water, the Court concluded,
“In all, there is substantial and scientifically credible evidence establishing that fluoride poses a risk to human health; it is associated with a reduction in the IQ of children and is hazardous at dosages that are far too close to fluoride levels in the drinking water of the United States. And this risk is unreasonable under Amended TSCA.”
The Court’s ruling sends a clear message: the days of unquestioned water fluoridation might be numbered. It also demonstrates that those claiming that the end of Chevron deference only involves the Court in procedural or legal questions, not scientific ones, are incorrect. Whether this decision sparks meaningful reform or just another round of bureaucratic foot-dragging, one thing’s certain—this debate isn’t going down the drain anytime soon.
[1] Plaintiffs argue that the weight-of-evidence step is a distinct qualitative evaluation of fluoride’s neurotoxicity, separate from a quantitative dose-response assessment. The EPA disagreed, claiming the two are interconnected.
[2] The Benchmark Dose (BMD) method is preferred over NOAEL and LOAEL due to its greater precision and reliability. It utilizes statistical modeling to analyze the entire dose-response curve, providing a more comprehensive and consistent assessment across studies. Unlike the fixed thresholds of NOAEL and LOAEL, BMD offers flexibility to target specific endpoints and reduces bias from experimental design limitations.
[3] The MOE is calculated by applying an uncertainty factor to the POD. Under the Amended TSCA, the EPA applies a default safety margin of 10x between the hazard and exposure levels to account for variability in human sensitivities, meaning exposure must be below 1/10th of the hazard level to be deemed safe.
Chuck Dinerstein, MD, MBA
Director of Medicine
Dr. Charles Dinerstein, M.D., MBA, FACS is Director of Medicine at the American Council on Science and Health. He has over 25 years of experience as a vascular surgeon.
