Jul 27 2012

Anti-Fluoride Propaganda as News

A news article in the Sacramento Bee declares: “Harvard Study Finds Fluoride Lowers IQ – Published in Federal Gov’t Journal.” Except – this is not a news item, and it’s not really a study. The article is about a recently published systematic review and meta-analysis – not new data. The term “study” is vague, and I find it often causes confusion.

Far worse than this common imprecision is the fact that this article, under the “News” tab on the Bee website, is not actually a news report. It is a propaganda article written by the NYS Coalition Opposed to Fluoridation, Inc. and distributed as a press release. The Bee does post a disclaimer at the top of the page, reading:

This section contains unedited press releases distributed by PR Newswire. These releases reflect the views of the issuing entity and are not reviewed or edited by the Sacramento Bee staff. More information on PR Newswire can be found on their web site.

That’s better than nothing, but I wonder how many people reading the press release will notice and read the disclaimer. In my opinion, a news outlet should not reprint press releases sent out from advocacy organizations clearly intended to promote an agenda. They especially should not print them under the banner of “News.” The disclaimer is not adequate.

The review itself is being spread around the internet and multiple readers sent me links to the article or reporting about it. Not surprisingly, the press release from an anti-fluoride organization did not put the review into proper scientific context. The review concludes:

“The results support the possibility of an adverse effect of high fluoride exposure on children’s neurodevelopment. Future research should include detailed individual-level information on prenatal exposure, neurobehavioral performance, and covariates for adjustment.”

“The possibility” of an effect justifying future research is not the same as concluding that there is an effect. But the problems with the way this review is being presented go far deeper. The implication being implied by anti-fluoride groups is that the fluoridation program is the US and elsewhere is putting children’s IQs at risk. This data, however, cannot be used to support that conclusion.

First it should be noted that almost all of the studies reviewed were conducted in China (one was conducted in Iran) – not in the US. China had a limited fluoridation program for a time, and has had no fluoridation of drinking water since 2002. So why, then, are most of the studies from China?

There are many rural areas of China that have naturally high levels of fluoride in the well water. The studies were largely looking at this exposure. Two studies looked at fluoride exposure from inhaling smoke from coal burning. So the question is – how do these levels of exposure relate to the amount of fluoride being added to water in the US (because toxicity is always all about dose)? There was a lot of variability across the studies, but generally the high fluoride groups were in the 2-10 mg/L range, while the reference low fluoride groups were in the 0.5-1.0 mg/L range (not including the coal burning studies, which had much higher fluoride levels).

The recommended fluoride level for fluoridated drinking water is 0.5-1.0 mg/L (similar in most countries – Australia, for example, uses 0.6-1.1 mg/L). The EPA set the upper safe limit at 4.0, with a secondary (voluntary) recommendation of 2.0. Areas with high natural fluoride actually have some of the fluoride removed from the drinking water.

In other words – fluoridated water in the US has the same level of fluoride as the control or low fluoride groups in the China studies reviewed in the recent article, and the negative association with IQ was only found where fluoride levels were much higher – generally above EPA limits.


There are many weaknesses to the epidemiological studies reviewed in the recent article – high heterogeneity, poor controlling for other variables, no indication of blinding of IQ assessments, and many others. But even taken at face value they do not indicate any association of between lower IQ and the fluoride levels added to drinking water in the US. In fact, those levels of fluoride were used as the controls in these studies showing higher IQ. (There was a lot of variance of the effect size, but the net effect size on IQ in the meta-analysis was -0.45 IQ points).

This fact has not stopped anti-fluoridation groups from exploiting the review for their own propaganda purposes. Otherwise respectable news outlets are unwittingly collaborating in this anti-scientific propaganda campaign by lazily reprinting these press releases in their news sections – without any editorial filter.

10 responses so far

10 Responses to “Anti-Fluoride Propaganda as News”

  1. ConspicuousCarlon 27 Jul 2012 at 2:28 pm

    Wow. The disclaimer is really just an announcement that they didn’t do their job.

  2. nybgruson 27 Jul 2012 at 10:37 pm

    I actually read this myself as part of a post on Reddit where someone was claiming it was “new data.” I had essentially the same objections to it. In my comment, however, I primarily focused on the things people should be looking at to realize how little they should trust or care about the “article.” Many of the claims for veracity of the article hinged on Reuters being a reputable news source (obviously not having read the disclaimer), the fact that it is a Harvard study (I mentioned Kaptchuk and Moerman), and that there were references (I pointed out that all but one reference was a link to the anti-fluoride group’s website and other similarly dubious sources).

  3. LindaRosaRNon 28 Jul 2012 at 2:43 pm

    The Chinese IQ/fluoride studies have found to be flawed. Not taken into account was the practice of burning in the home coal that contains high levels of fluoride, as well as other likely contaminants (e.g. arsenic and tainted grain).

  4. Nitpickingon 29 Jul 2012 at 10:48 am

    @ConspicuousCarl: it all depends on what you think “their job” is. If the job is “fill this site with stuff people will look at to maximize our ad revenue” it was a huge success. Papers in general seem to be moving to the PuffHo model.

  5. Mlemaon 10 Aug 2012 at 2:46 am

    it’s a shame this anti-fluoride group has resorted to such means as this.
    I guess the gobs of money from the big chemical corporations who don’t want to lose the increased profit they gain from being able to dump their waste into municipal drinking water and charge for it must have got the better of them. Why is most of the fluoride in our water provided by the addition of Fluorosilicic acid? It’s the most expensive to distribute BUT cheap to manufacture: an inexpensive liquid by-product of phosphate fertilizer manufacture. Hey, the municipalities can pay for the shipping right? Is that why?

    there’s a lot of propaganda on both sides of this issue and we don’t need any more.

    just a couple of things from the CDC: the suggested concentration was lowered to .7 in January of 2011. And if you’re feeding your infant exclusively on formula that must be mixed with water, don’t use fluoridated water.

    The rates of fluoridosis have increased over the last 25 years.
    from HHS:
    “…In 1986-87 and 1999-2004 the prevalence of dental fluorosis was 23% and 41%, respectively, among adolescents aged 12 to 15. (Beltrán-Aguilar ED, et al, 2010a). Similarly, the prevalence of very mild fluorosis (17.2% and 28.5%), mild fluorosis (4.1% and 8.6%) and moderate and severe fluorosis combined (1.3% and 3.6%) have increased.”

    i’ts the presence of fluoride in saliva that may be beneficial for preventing tooth decay. But the ingestion of fluoride to provide that fluoride has little proven benefit, and has proven detriment to small children. There is no benefit to fluoride in the saliva before the teeth have erupted, while fluorosis is likely to occur because small children uptake a greater amount in relationship to their size.

    Dr. Novella linked to this site:
    “In infants 80–90% of absorbed fluoride is retained, with the rest excreted, mostly via urine; in adults about 60% is retained. About 99% of retained fluoride is stored in bone, teeth, and other calcium-rich areas, where excess quantities can cause fluorosis.”
    also from that page:
    “Existing evidence strongly suggests that water fluoridation reduces tooth decay. There is also consistent evidence that it causes dental fluorosis, most of which is mild and not usually of aesthetic concern.[10] There is no clear evidence of other adverse effects. Moderate-quality research exists as to water fluoridation’s effectiveness and its potential association with cancer; research into other potential adverse effects has been almost all of low quality. Little high-quality research has been performed.[11]”

    it’s a good page to read if you’d like to familiarize yourself with the issue.

    Personally, the fact that the presence of fluoride in our food and environment is increasing, and the fact that there are people who shouldn’t be drinking fluoridated water (babies and people with reduced kidney function) and the fact that children are developing fluorosis (which means they’re getting too much fluoride) should indicate that fluoride supplementation would be better done on an individualized basis. There are too many people drinking fluoridated water that have no idea that they shouldn’t. There’s no public education about this. The CDC makes its recommendations, but they’re not disseminated in any way. just seems wrong to me. Why should I have to pay for and drink fluoridated water if I don’t want to? Because somebody somewhere decided it’s good for me? What if I’m one of the people it’s not good for? Then I have to pay to get it out. This practice is discriminatory against the poor, who have less access to information, medical advice, and non-fluoridated water.

  6. Billy Buddon 23 Aug 2012 at 2:12 am

    The use of this meta-analysis by those opposing community water fluoridation (CWF) should be instructive. It was listed as a news story on Reuters by the New York State Coalition Opposing Fluoridation.

    A explanation of that may be conveniently read at: http://www.cracked.com/quick-fixes/fluoride-lowers-your-iq-b.s.-headline-week/

    As does Mlema, they make a lot of the fact that two of the water additives are co-products of the phosphate fertilizer industry.

    The truth is that both hexafluorosilicic acid (HFSA) and its sodium salt completely dissociate by hydrolysis into fluoride ion, sodium, and silicate (ie sand). NMR studies have shown this to be complete, there are no intermediaries at drinking water pH.

    Mlema isn’t quite as clear as some that they tell citizens that the fertilizer companies distribute HFSA to get rid of toxic industrial wastes for which disposal would be otherwise inconvenient and expensive.

    The truth is, US water additive sales are unimportant to phosphate mining waste disposal. The US mines about 37 million tons of phosphate rock. PR contains 3.5 to 4.0% fluorine. About 40,000 tons of fluorosilicic acid are recovered. This represents only about 2% of the total fluorine mined. The fluoride which will become water additives is an inconsequential portion of the fluorine waste. Excess fluoride goes with the mining tailings to “gyp stacks.” This is where the 98% of the fluoride in the ore is going, not to America’s water supply. Florida requires land reclamation. The fluoride removed from ore has nothing to do with the cost nor the methods used to basically rebury the unused mineral materials.

    Mlema bewails dental fluorosis, laying the blame on CWF, and doubly complains that swallowing doesn’t do any good anyway.

    The truth is that all the fluorosis from CWF is of the type only noticed on careful professional examinations and that new research shows it to protect teeth from cavities. This latter effect quite obviously requires that children swallow fluoride to be incorporated into the tooth. Another recent study showed that kids who swallow fluoride become adults with more teeth.

    Oh and cancer. .

    The truth is that in Nov 2011 the California Cancer Identification Committee unanimously ruled fluoride (at any concentration) does not cause cancer.

    Mlema says that CWF discriminates against poor people.

    The truth is that 2/3rd of the operations for terrible cavities in kids on Medicaid are avoided with community water fluoridation (CWF). Done under anesthesia procedures include extractions root canals & stainless steel crowns & cost up to $15,000. In that Louisiana study 50% of the dental bills for the kids studied were avoided. A subsequent Colorado study found that the beneficial effect on these young kids mouths alone returns 150% in lower dental bills. This saved money is available for state Medicaid programs to spend for more disease prevention and treatment of these socially disadvantaged kids.

    What Mlema does not tell you of is the enormous passion which persons like him/herself bring to state and local fluoridation politics. Often as not good public health, fewer operations and better teeth loose to ideas more mistaken and misleading than these you have read here. Only yesterday in Portland, Or activists marched on City Hall protesting that among other harms, fluoridation would ruin the beer. In 2004 in Hood River, OR both beer and good whiskey were seen as harmed.

    CWF is a really big deal. About 72% of Americans on public water systems enjoy its benefits. A few months ago Pinellas County Florida Commission discontinued it for about 700,000. Quotes from one of the off the edge right wing leaders accomplishing this could well have been in the Dr. Strangelove fluoridation scene.

    There is an overwhelming professional consensus the CWF is safe, effective and inexpensive. Readers can see the long list of respected scientific and health advocacy groups endorsing, in their own words, CWF at the Pew Charitable Trust’s web site: http://www.ilikemyteeth.org/fluoridation/respected-organizations/

    I write this with the hope of energizing scientifically minded people to participate in advocating for CWF. Fluoride Wars are underway in Portland, OR and Wichita, KS.

    References available if anyone wishes . . I’ll try to check back.

  7. iAccidentallyon 25 Aug 2012 at 8:19 pm

    “So the question is – how do these levels of exposure relate to the amount of fluoride being added to water in the US (because toxicity is always all about dose)?”
    For me the question is, rather: Is fluoridation worth the risk of ANY neurotoxicity? That exposure to higher levels of fluoride was deemed responsible for the lower IQ scores doesn’t imply that exposure to the lower levels of fluoride had no effect. I agree with the paper that more studies are warranted, but in the meantime, is the potential upside worth the risk?

  8. jwalms2on 03 Feb 2013 at 10:53 am

    Big shock: Mercola wrote about this is his 28 Jan HuffPo column, “Harvard Study Confirms Fluoride Reduces Children’s IQ”.

  9. Mlemaon 03 Feb 2013 at 6:05 pm

    Billy Budd,

    nothing you’ve written changes anything I wrote:

    1) fluorosis in children means they’re ingesting too much fluoride

    2) the public is paying for fluoridation of water. Chemical companies and others profit from fluoridation while there are no quality studies of its purported benefits or risks.

    3) there are subsets of the population who are harmed by fluoridation and killed when accidental over-fluoridation happens.

    there’s no way you can substantiate the financial benefit you claim – fluoridated toothpaste and hygiene education allows more than enough dental benefit for those who choose to utilize them.

    And choice is what we’re supposed to have as Americans and consumers.

    your facts are vague. I’m guessing that the studies and opinions you mention can be traced back to financially-interested parties. Big money has big voice, and one tactic used is to paint those who dissent as ignorant or subscribing to a naturalistic fallacy or fear of big business or public, government run programs. My hope is that people who have enough physiology background will be able to understand that the science which originally supported development of the practice of fluoridation of public water supplies was incomplete. And if not inaccurate at the time, certailny outdated now. Fluoride in our diets continues to increase. We don’t need to pay and have it artificially added to our water. It’s a toxin above all but trace amounts.

  10. gwernon 06 Jun 2014 at 12:39 pm

    > There was a lot of variance of the effect size, but the net effect size on IQ in the meta-analysis was -0.45 IQ points

    (The fulltext of Choi et al 2012 is available at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3491930/ )

    This isn’t correct; you’re misinterpreting the units used in the meta-analysis. Meta-analyses always use standardized effect sizes like odds ratios or mean differences in standard deviations, and the abstract is perfectly clear that they aren’t using a converted unit like IQ points when they reported the -0.45 point-estimate:

    > The standardized weighted mean difference in IQ score between exposed and reference populations was –0.45 (95% confidence interval: –0.56, –0.35) using a random-effects model.

    As in, the two groups differed by 0.45 standard deviations. If the -0.45 was being reported as IQ points, then the actual SMD would be (assuming the usual IQ scale of 15 points per standard deviation) -0.03, but you can check the forest plot and see that the result definitely isn’t -0.03: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3491930/figure/f2/

    So the actual correlation is 15x larger than you think it is: -6.75 IQ points. This is a concerning result.

    To put -6.75 in perspective, if you have a population at 100 IQ and drop the mean by 10 points, then you triple the number of people 35,000!* A look at the forest plot, and one knows that the total sample size was nowhere near that big…

    * estimate from a t-test power analysis for d=-0.03 and power of 80%; a fixed-effects meta-analysis of SMDs is similar to a t-test; random-effects meta-analyses usually require more samples than a fixed-effects; hence, 35k is a lower bound

    > In other words – fluoridated water in the US has the same level of fluoride as the control or low fluoride groups in the China studies reviewed in the recent article, and the negative association with IQ was only found where fluoride levels were much higher – generally above EPA limits.

    Why would the harmful effect stop at the EPA limit? Did they ever run randomized experiments measuring cognitive functioning showing that there is a sharp threshold where at 4mg/L harmful effects abruptly go to zero? The linear-no-threshold model is used for a lot of things, why not fluoride?


    > Not taken into account was the practice of burning in the home coal that contains high levels of fluoride, as well as other likely contaminants (e.g. arsenic and tainted grain).

    Measurement error biases coefficients to zero. Similarly, unless there’s systematic covariation between the coal usage and fluoride levels, that’s just part of the random heterogeneity which we already know about from the high I^2.


    > I agree with the paper that more studies are warranted, but in the meantime, is the potential upside worth the risk?

    Depends on what the risk is, and also how bad ‘neurotoxicity’ is. I hope we all agree here that cost-benefit is the right approach for this sort of question…?

    It’s fashionable to argue that IQ doesn’t matter, but it affects a lot of things, as we can see in natural experiments like the iodization of America and Switzerland, and we can extract direct financial estimates of its value. To reuse the $12k figure from above, across the American population, that gives an upper bound of $3 trillion a year lost. America spends ~$100 billion annually on dentistry or ~$333 per capita. Let’s be conservative and imagine that fluoridation fully halves dentistry bills and that fluoridated toothpaste isn’t an option, so not fluoridating would mean spending $666 per capita. Then the two choices, compared to the status quo, are $12000 – (666 – 333) = $11667 (we reap the gain of higher IQ but we have to pay on the margin for more dentistry), and $333 (no gain from higher IQ, but also cheaper dentistry); if there’s a >3% chance fluoridation does such damage, then the expected-value is higher with the no-fluoridate option. So that’s an example of how you might calculate whether the upside of cheaper dentistry is worth the risk of neurotoxicity.

    Definitely the sort of problem which urgently justifies more studies, and specifically, *randomized* experiments. We can argue about correlational results from China until the cows come home without reaching much more confident conclusions (‘garbage in, garbage out’), but it’s only randomized experiments which will give us causal inferences. The good news is that if the loss is half a standard deviation for high levels vs normal levels, then the loss for normal vs low/zero levels should be fairly easy to measure (halve the effect to d=0.225, demand extreme power of 99%, and you only need ~1500 people). And it’s not like experiments have to be hard to run: you could settle for cluster randomization of local drinking suppliers – and how much does it cost to *not* fluoridate water and then monitor the next decade of local school grades & ACT/SAT scores (an already-being-collected proxy for intelligence)?

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.