A recent flurry of "fluoride is bad!" posts are appearing on my social network news feeds. Usually I can simply ignore them after a brief look, but this one, stemming from a recent article in The Lancet, isn't easy enough for me to understand and apply or ignore.

The posts link to a March, 2014 article, Neurobehavioural effects of developmental toxicity, from The Lancet Neurology, Volume 13, Issue 3, which has the summary (emphasis added):

Neurodevelopmental disabilities, including autism, attention-deficit hyperactivity disorder, dyslexia, and other cognitive impairments, affect millions of children worldwide, and some diagnoses seem to be increasing in frequency. Industrial chemicals that injure the developing brain are among the known causes for this rise in prevalence. In 2006, we did a systematic review and identified five industrial chemicals as developmental neurotoxicants: lead, methylmercury, polychlorinated biphenyls, arsenic, and toluene. Since 2006, epidemiological studies have documented six additional developmental neurotoxicants—manganese, fluoride, chlorpyrifos, dichlorodiphenyltrichloroethane, tetrachloroethylene, and the polybrominated diphenyl ethers. We postulate that even more neurotoxicants remain undiscovered. To control the pandemic of developmental neurotoxicity, we propose a global prevention strategy. Untested chemicals should not be presumed to be safe to brain development, and chemicals in existing use and all new chemicals must therefore be tested for developmental neurotoxicity. To coordinate these efforts and to accelerate translation of science into prevention, we propose the urgent formation of a new international clearinghouse.

I don't have full access to the article, so I can't refer to the epidemiological study it says shows fluoride to be a developmental neurotoxicant.

  • What is a developmental neurotoxicant?
  • How strong is the research that claims fluoride is a developmental neurotoxicant?
  • What relationship, if any, does this report or research have with typical fluoride compounds used in US water treatment at the levels typically present in US water (sodium fluoride, fluorosilicic acid, or sodium fluorosilicate at 0.7 mg/L)?
  • $\begingroup$ The PDF of the paper can be found here. I'll have a look at it later and write an answer about it. $\endgroup$
    – Chris
    May 21, 2014 at 17:35
  • $\begingroup$ The paper references the following paper regarding the fluoride meta-analysis: Choi AL, Sun G, Zhang Y, Grandjean P. Developmental fl uoride neurotoxicity: a systematic review and meta-analysis. Environ Health Perspect 2012; 120: 1362–68. This appears to be available here: ehp.niehs.nih.gov/1104912 Thanks for reviewing this! $\endgroup$
    – Adam Davis
    May 21, 2014 at 18:00
  • $\begingroup$ I don't really understand your third question. $\endgroup$
    – Chris
    May 21, 2014 at 18:47
  • $\begingroup$ @Chris I should probably rewrite that. Just because fluoride may be considered toxic doesn't mean that any or all of the compounds used deliver effective levels of fluoride to the brain, so I'm asking if this information about fluoride toxicity applies to fluoride as it is currently used in our water systems. $\endgroup$
    – Adam Davis
    May 21, 2014 at 18:49
  • $\begingroup$ Well, that is the question. The Choi paper does its meta-analysis on research which has been conducted on water fluorine levels. So this is not unimportant. $\endgroup$
    – Chris
    May 21, 2014 at 18:51

1 Answer 1


First of all: Yes, fluoride is toxic, but the toxicity depends largely on the form (soluble vs. unsoluble, which fluoride salt etc.) occurs. It also depends on the environment since insoluble salts which are subjected to strong acids can release fluorine ions. The certain toxic dose for adults is 32-64mg/kg body weight, a 75kg adult needs to take up between 2,4 and 4,8 grams of fluoride ions (for sodium fluoride this would be between 5,3 and 10,6 gr of the salt). For children this is about 5mg/kg body weight. For the water the added dose of fluorine is much lower, somewhere between 0,7 and 1.2 mg/l. Other sources of fluroride are food (especially sea food) and also black tea.

The toxicity of fluoride is mediated by different ways: First it forms an insoluble salt with calcium in the cells, effectively causing hypocalcemia. Calcium is an important second messenger in different signal transduction pathways as well as highly important for the functioning of nerve cells (which also explains why it is neurotoxic in the development of nerve and brain cells). It also causes the inhibition of a number of different enzymes as enolase (disturbs the glycolysis) but also phosphatases and katalase. For all of this have a look into the Wikipedia if you want to know more details.

About the paper: It is a review of published research articles which sums up the situation in this field. The paper mentions fluoride only twice: Once on the abstract and once in a table about newly identified developmental neurotoxins (table 2 to be precise). It references only one meta-analysis on this topic.

The cited meta-analysis ("Developmental fluoride neurotoxicity: a systematic review and meta-analysis") which did a statistical meta-analysis of published papers and their data. They find a significant relation between the fluoride content of the drinking water and the IQ of the children. However this is not undisputed (to say it mildly), since while a finding can be statistically significant, it doesn't necessarily mean that there is clinical importance. The other important point which @Doctor Whom mentioned is that a correlation doesn't imply a causation. So even if there seems to be a correlation between two things, they are not necessarily connected (see the famous example of storchs and birth rates).

This paper ("Developmental fluoride neurotoxicity: clinical importance versus statistical significance.") is published as a comment on the original paper and claims exactly this. They say the the mean difference in IQ is about 7 points, which is not of clinical importance. Another paper critizes the relatively low quality of the original papers used for the study ("Fluoride and children's IQ.").

They also mention a low consensus about what is a high concentration of fluoride in the drinking water and also show the very wide variation of the fluoride concentration in the original studies. The highest concentrations mentioned are around 10mg/l which is about 10x the concentration used for adding fluoride to the water. Even if we take in account that children are more sensitive to fluoride and also need lower fluoride concentrations to cuse some developmental damages (below the fatal concentration) this would still add up to quite absurd amounts of water they have to drink.

Regarding your questions:

  1. A developmental neurotoxin is a substance which interferes with developing neurons or nerve cells. It either inhibits their growth completely or at least impairs the cells.

  2. Given on the articles I found (and cited above) I would say that fluoride is a developmental neurotoxin. Nevertheless it seems to me that it is not very strong in this context as the current studies are contradictory. Thinking about the amounts of fluoride present in water (even if the concentration is much higher than recommended) makes me doubt if this is really the causative agent.

  • 1
    $\begingroup$ Well written, Chris. I would recommend the addition of two points. One, to the "significance" discussion, the point that correlation does not infer causality. That study had a strong likelihood for bias, as there are MANY regional factors beyond yes/no fluoridated water that could influence this marginal difference in IQ results. Secondly, to the final point (#2), I would comment on the issue of dose. $\endgroup$
    – DoctorWhom
    May 21, 2014 at 22:23
  • $\begingroup$ Good points, thank you for mentioning. I added them. $\endgroup$
    – Chris
    May 22, 2014 at 7:19
  • $\begingroup$ 7 IQ points is rather significant - 10 points is supposed to be one standard deviation. Do you mean that the IQ difference was not statistically significant? $\endgroup$
    – shigeta
    May 22, 2014 at 13:36
  • $\begingroup$ The original authors say that 7 IQ points difference is statistical significant. We are less than 1 sigma away from the mean here. $\endgroup$
    – Chris
    May 22, 2014 at 14:48

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .