TSCA Ruling in Fluoridation trial: Findings of Fact and Conclusion of Law. September 24, 2024.

Abstract

UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA

FOOD & WATER WATCH, INC., et al.,
Plaintiffs,
v.
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY, et al.,
Defendants.

Case No. 17-cv-02162-EMC
FINDINGS OF FACT AND CONCLUSIONS OF LAW

I. INTRODUCTION
In 2016, Congress amended the Toxic Substances Control Act (“TSCA”), empowering United States citizens to petition the Environmental Protection Agency (“EPA”) to consider whether a chemical presents an unreasonable risk of injury to health. See Pub. L. No. 114-182, 114th Congress (Frank R. Lautenberg Chemical Safety for the 21st Century Act) (the “Act”). The Act addresses the modern day reality that “human beings and the environment are being exposed each year to a large number of chemical substances and mixtures,” 15 U.S.C. § 2601(a)(1), and that, “among the many chemical substances and mixtures which are constantly being developed and produced, there are some whose manufacture, processing, distribution in commerce, use, or disposal may present an unreasonable risk of injury to health or the environment,” id. § 2601(a)(2).

To this end, under TSCA, as amended by the Act (“Amended TSCA”), a citizen is entitled to judicial review of the EPA’s denial of the citizen’s petition, wherein a court considers whether the chemical poses an unreasonable risk de novo, i.e., without deference to the EPA’s decision. See id. § 2620(b)(4)(B). Amended TSCA sets up a system of judicial review that is remarkably different from the usual scope of judicial review of administrative actions under the Administrative Procedure Act, which confers substantial deference to administrative agencies. See id. Under Amended TSCA, the Court owes no deference to the EPA in assessing the risk posed by chemical substances. See id. If the Court finds anew that the chemical at issue presents an unreasonable risk, it then orders the EPA to engage in rulemaking regarding the chemical. Seeid. The EPA is afforded in the first instance the authority to respond; regulatory actions can range from requiring a mere warning label to banning the chemical. See id. § 2605(a)(1)-(7). The EPA, in short, has options. See id.

The issue before this Court is whether the Plaintiffs have established by a preponderance of the evidence that the fluoridation of drinking water at levels typical in the United States poses an unreasonable risk of injury to health of the public within the meaning of Amended TSCA. For the reasons set forth below, the Court so finds. Specifically, the Court finds that fluoridation of water at 0.7 milligrams per liter (“mg/L”) – the level presently considered “optimal” in the United States – poses an unreasonable risk of reduced IQ in children. It should be noted that this finding does not conclude with certainty that fluoridated water is injurious to public health; rather, as required by the Amended TSCA, the Court finds there is an unreasonable risk of such injury, a risk sufficient to require the EPA to engage with a regulatory response. This order does not dictate precisely what that response must be. Amended TSCA leaves that decision in the first instance to the EPA. One thing the EPA cannot do, however, in the face of this Court’s finding, is to ignore that risk.

A. Context
Water fluoridation has a long history in the United States and has been a source of politicaldiscord, at times. See, e.g., Dkt. No. 429-3, Trial Ex. 13 at 15.1 In 1975 the EPA recommended adding fluoride to water, with an optimal level up to 1.2 mg/L for its dental health benefits. Id. at 16.Between 1981 and 1984, fluoride’s association with adverse effects including osteosclerosis,enamel fluorosis, and psychological and behavioral problems was contested. Id. at 17-18. Still, asof 1986, up to 1.2 mg/L water fluoridation was considered optimal, and the maximum level was 4mg/L. Id. at 14-18. After evidence increasingly established fluoride’s connection to adverse effects, including severe enamel fluorosis, risk of bone fracture, and potential skeletal fluorosis, recommended levels were lowered in 2006. Id. at 10. Community water fluoridation has since continued at levels believed to be safe for its dental health benefits. At present, fluoride is added to tap water in the United States, with an optimal level of 0.7 mg/L.

However, scientific evidence has increasingly identified a link between fluoride exposure and adverse cognitive effects in children (reduced IQ). Accordingly, Plaintiffs exercised their power under Amended TSCA and petitioned the EPA to consider whether fluoride in drinking water presents an unreasonable risk of injury to human health. Notwithstanding the growing and robust body of evidence indicating an association between fluoride intake and cognitive impairment in children, the EPA denied Plaintiffs’ petition. Plaintiffs filed suit in this Court, arguing that the EPA was wrong and that community water fluoridation at 0.7 mg/L (the “condition of use”) poses an unreasonable risk of injury to human health.

B. Summary
To succeed in a suit brought under the Amended TSCA, Plaintiffs must prove, by a preponderance of the evidence, that a risk of injury to human health is present and that such risk is unreasonable. For a risk to be present, Plaintiffs must show that some segment of the United States population is exposed to the chemical at issue at levels that either exceed, or are too close to the dosage at which the chemical presents a hazard.2 The reasonableness of the risk is informed by several factors, including inter alia, the size and susceptibility of impacted populations, severity of the harm at issue, and the frequency and duration of exposure.

There is little dispute in this suit as to whether fluoride poses a hazard to human health. Indeed, EPA’s own expert agrees that fluoride is hazardous at some level of exposure. And ample evidence establishes that a mother’s exposure to fluoride during pregnancy is associated with IQ decrements in her offspring. The United States National Toxicology Program (“NTP”) – the federal agency regarded as experts in toxicity – undertook a systematic review of all available literature near the time of publication considering whether fluoride poses cognitive harm, reviewing 72 human epidemiological studies considering this question. The NTP concluded that fluoride is indeed associated with reduced IQ in children, at least at exposure levels at or above 1.5 mg/L (i.e., “higher” exposure levels). And notwithstanding inherent difficulties in observing effects at lower exposure levels, explained in further detail below, scientists have observed a statistically significant association between fluoride and adverse effects in children even at such “lower” exposure levels (less than 1.5 mg/L).

Notwithstanding recognition by EPA’s expert that fluoride is hazardous, the EPA points to technicalities at various steps of the risk evaluation to conclude that fluoride does not present an unreasonable risk. Primarily, the EPA argues the hazard level and the precise relationship between dosage and response at lower exposure levels are not entirely clear. These arguments are not persuasive.

Importantly, the chemical at issue need not be found hazardous at the exposure level to establish that a risk is present under Amended TSCA. Instead, the EPA requires a margin exist between the hazard level and exposure level to ensure safety; if there is an insufficient margin then the chemical poses a risk. The trial evidence in this case establishes that even if there is some uncertainty as to the precise level at which fluoride becomes hazardous (hazard level), under even the most conservative estimates of this level, there is not enough of a margin between the accepted hazard level and the actual human exposure levels to find that fluoride is safe. Simply put, the risk to health at exposure levels in United States drinking water is sufficiently high to trigger regulatory response by the EPA under Amended TSCA.

To this end, as mentioned previously, the NTP compiled and analyzed all relevant studies it could find and concluded that, at least at dosages of 1.5 mg/L or higher, fluoride is associated with reduced IQ in children. Subsequently, toxicology experts endeavored to put a finer point on the impact of fluoride on children’s IQ at “lower” exposure levels, i.e., those below 1.5 mg/L, and conducted a pooled benchmark dose analysis to define the precise hazard level of fluoride. For reasons described below, this pooled benchmark dose analysis benefited from increased statistical power relative to the NTP’s assessment due to its methodology (i.e., the benchmark dose analysis used individualized, continuous data, while the NTP assessment did not, due to quantity and variety of studies the NTP reviewed in that assessment). The pooled benchmark dose analysis concluded that a 1-point drop in IQ of a child is to be expected for each 0.28 mg/L of fluoride in a pregnant mother’s urine. This is highly concerning, because maternal urinary fluoride levels for pregnant mothers in the United States range from 0.8 mg/L at the median and 1.89 mg/L depending upon the degree of exposure. Not only is there an insufficient margin between the hazard level and these exposure levels, for many, the exposure levels exceed the hazard level of 0.28 mg/L.

The EPA challenges, for a variety of reasons, whether this 0.28 mg/L hazard level (measured in maternal urinary fluoride) is appropriate for this risk evaluation. The EPA argues, among other things, that the hazard and exposure levels should not be expressed in maternal urinary fluoride because that metric reflects total fluoride exposure – not just exposure resulting from drinking fluoridated water from one’s community. Fluoride may also be ingested through, e.g., tea, fish, toothpaste, and commercial food and beverage made with fluoridated water. Nonetheless, the risk analysis should consider the additive effect of the chemical under the subjected condition of use (here, fluoridated community drinking water), especially where, as here, the fluoridated drinking water is a significant (and likely primary) contributor to aggregate exposure to fluoride. Indeed, the Amended TSCA, expressly contemplates that the aggregate exposure to a chemical will be considered when conducting a risk assessment. See 15 U.S.C. § 2605(b)(4)(F). In this sense, maternal urinary fluoride is not just an acceptable metric, it is highly useful in assessing the real-world end result of exposure from drinking fluoridated water along with other sources.Even if urinary fluoride were not the appropriate metric in assessing health risk, or even if the toxicologically determined hazard level of 0.28 mg/L were deemed insufficiently substantiated, evidence in the record still establishes with little doubt that fluoridated drinking water presents a risk of injury to health. Using a highly conservative estimate of the hazard level of 4 mg/L measured in drinking water fluoride (well above the 1.5 mg/L identified as hazardous to children by the NTP) based on the consistent and repeated observation of adverse effects summarized in the NTP’s assessment, a risk is present. There is little dispute that there is a statistically significant association between IQ decrements in children and fluoride concentration levels at 4 mg/L.

The EPA’s default margin of error requires a factor of 10 between the hazard level and exposure level due to variability in human sensitivities. Put differently, only an exposure that is below 1/10th of the hazard level would be deemed safe under Amended TSCA, given the margin of error required. Here, an even greater margin (100x) is owed because the methodology (which yields the 4 mg/L hazard level) uses the lowest observed adverse effect level (“LOAEL”); this methodology adds an additional level of uncertainty (and hence the application of a 100x rather than 10x margin). But even if only the default 10x margin is required, the safe level of fluoride exposure would be 0.4 mg/L (4 mg/L (hazard level) divided by 10). The “optimal” water fluoridation level in the United States of 0.7 mg/L is nearly double that safe level of 0.4 mg/L for pregnant women and their offspring.

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. Reduced IQ poses serious harm. Studies have linked IQ decrements of even one or two points to e.g., reduced educational attainment, employment status, productivity, and earned wages. Indeed, the EPA recognizes that reduction of IQ poses a serious community health issue. Moreover, highly susceptible populations are impacted, including over two million pregnant women and babies, a number far exceeding population size the EPA has looked to in determining whether regulatory action was warranted in other risk evaluations (i.e., 500 people or less).

Thus, the Court finds Plaintiffs have met their burden in establishing, by a preponderance of the evidence, that community water fluoridation at 0.7 mg/L presents an unreasonable risk of injury to health under Amended TSCA and that the EPA is thus obliged to take regulatory action in response. The Court does not in this order prescribe what that response should be.