Fluoride Action Network

Three figures to help NTP see the whole truth

Fluoride Action Network | Bulletin | November 20, 2020

The NTP Review: “A Chipped Vase”

Before FAN deals with the “Chip” let us again make very clear that we are happy with the “vase.” The NTP review is one of the most comprehensive reviews of fluoride’s neurotoxicity ever conducted. Its overall findings that fluoride is a “presumed neurotoxicant” and the evidence that it lowers children’s IQ at 1.5 ppm or above is sufficient in itself to end the reckless practice of water fluoridation. We could then, like any proud homeowner, simply turn the vase so the “chip” faces the wall, but we can’t. The NTP’s conclusion in the section dealing with the relevance to America (i.e. water fluoridation) is that the evidence is “less robust and less consistent” in studies at levels of exposure less than 1.5 ppm than above, is simply not true. The science does not support such a conclusion. In fact, the strongest studies, published since 2017 – and funded by the NIH – were conducted at levels less than 1.5 ppm and at the same exposure levels experienced in artificially fluoridated communities at the recommended 0.7 ppm.

Our suspicion is that this false information was slipped in to the NTP Review by a pro-fluoridation element to provide a thread on which government bodies can hang continued support for this outdated policy. Those who are unfamiliar with normal risk assessment practices might argue that 1.5 ppm is about twice as high as 0.7 ppm and it offers an adequate margin of safety for water fluoridation programs, but they would be wrong. It might be very tempting for the EPA to set a maximum contaminant level (MCL) of 1.5 ppm for fluoride, which would bring it in line with Mexico, Canada and the WHO. However, this would also be wrong – in fact all of these countries and WHO will eventually have to tighten their own standards to fit reality and the science. Simply put, fluoride research on neurotoxicity is following the same trajectory as the research on lead’s neurotoxicity from the 1970s to 1990s; namely, as far as early life is concerned there is virtually no safe level of exposure.

At a time where the public’s trust in government agencies is reaching an all-time low, it is absolutely essential that we maintain trust in both the NTP and its parent agency NIEHS by making sure that its reports are not tainted with any political interference. That is why we are again giving chapter and verse on the science here. We are doing it in the form of three figures which all point to the simple truth that the science shows that the studies conducted at exposure levels less than 1.5 ppm are as robust – if not more so – as those studies conducted above 1.5 ppm. There is no “chip”, the “vase” is whole.

The first figure is a table summarizing the 29 studies the NTP rated as higher quality (lower Risk of Bias is NTP’s terminology). Two of the studies were essentially duplications of a third so have been excluded. Of the 27 different high quality studies, 25 found statistically significant adverse neurotoxic effects and only 2 found no significant effect. None found a significant beneficial effect.

Of the 25 finding adverse effects, 15 were at exposures below 1.5 mg/L and 11 of those found effects below 0.7 mg/L, with the 10 remaining finding effects at or above 1.5 mg/L. This represents overwhelming consistency both below and above 1.5 mg/L.

FAN has submitted the detailed explanation of how each of the 25 studies were classified by exposure to the NAS and NTP.

Table 1: Summary of the key data from the highest quality studies reviewed by NTP.

The second figure is a graph summarizing Table 1.

Figure 2: Bar graph summarizing the findings presented in Table 1 above.

The third figure summarizes results of a meta-analysis comparing the main findings for studies above 1.5 ppm versus below 1.5 ppm for all the higher quality studies for which exposure data is available at the individual level (as opposed to group-level).  This is a forest plot (named because to someone it resembles a tree) which shows the effect size found in each study by its position on the x-axis. The degree of weight given to each study is indicated by the size of its gray box.  The pooled overall estimates of effect size and significance are shown by the blue diamonds.  The units are IQ points lost for every 1 mg/L increase in the concentration of water fluoride or urine fluoride. Black dot point estimates that are to the left of the vertical black zero line indicate loss of IQ.  There are no black dots to the right of the zero line which would indicate a beneficial effect of fluoride on IQ.  The black horizontal lines extending on either side of the black dots indicate the 95% confidence interval and when they are entirely below zero the finding is considered to be statistically significant.

Figure 3.  Forest plot summarizing meta-analysis of high quality studies.

As can be seen, multiple high quality studies, conducted in varied populations, consistently found strong associations between loss of IQ or other adverse developmental neurobehavioral outcomes at fluoride exposures commonly occurring from artificial fluoridation.  At exposures below 1.5 mg/L the effect size was actually greater than in the studies above 1.5 mg/L (-4.04 versus -2.40 IQ points per 1 mg/L fluoride).  While the NTP review’s authors included a statement in the narrative portion of their review stating, “the evidence that fluoride lowers IQ is less robust and less consistent at levels below 1.5ppm,” this biased claim is not supported by the data NTP themselves extracted from the original studies.

In fact, the highest quality studies, including several funded by NIH, strongly suggest that artificial fluoridation poses an unprecedented neurotoxic risk to a large proportion of the children in the US and other fluoridated countries.


Paul Connett
Executive Director

See all FAN bulletins online