So there I was last Sunday, filling my son’s sippy cup with tap water, when my friend told me about a Huffington Post article making the parenting rounds—the one titled “Harvard Study Confirms Fluoride Reduces Children's IQ.” I set down my Brita pitcher, knowing that fluoride was one of the few chemicals it doesn’t remove, and sighed. Well, shit.
Then I read the actual article, realized it was written by Dr. Joseph Mercola, the alternative physician who distinguishes himself as not being driven by “whatever has the most profit potential,” yet who sells fluoride-free toothpaste and hundreds of other products on his website (some of which have been slammed by the FDA for illegal marketing), and felt much better. Mercola frequently overstates the science and misleads his many readers—among other things, he preaches that vaccines cause autism and that homeopathy cures it, and oh, that animals are psychic—and this story (thankfully) is no different: The study on which he based his HuffPo article did find an association between high fluoride consumption and child IQ, but the findings aren’t applicable to American kids for a number of reasons.
Before I get into that, though, let me say this: I’m not staunchly pro-fluoride. The idea to add the chemical to drinking water grew from work that showed, in the 1930s, that children had fewer cavities in states in which fluoride naturally contaminated the drinking water (mainly from rocks). Three U.S. cities began adding the chemical to their water in 1945, and today, 66 percent of Americans drink fluoridated water. Scientists have pointed out that research on fluoride’s health effects is generally of poor quality, and that along with the upside, there is potential for harm: A 2006 report by the National Research Council, a nonprofit institution that provides expert advice on science, technology, engineering, and health issues, concluded that U.S. water that contains more than the recommended amount of fluoride may boost the risk of bone fractures and, ironically, discolor our teeth. The EPA might, then, be wise to lower its legal water fluoridation limit, which is now more than three times higher than what the Department of Health and Human Services recommends as an optimal concentration. (The EPA’s limit is legally enforced; the HHS recommendation is voluntary. The EPA is currently considering lowering its limit.)
But Mercola’s piece is not about bone fractures—it’s about IQ, and research on the link between fluoride and IQ is muddy at best. Moreover, the study Mercola highlights certainly doesn’t suggest that America’s water is stunting anyone’s cognitive potential. The research is a systematic review and meta-analysis of 27 studies, and it found that kids exposed to high levels of fluoride are nearly twice as likely to have low IQ scores as are unexposed kids. But 25 of the studies were conducted in China (the other two in Iran), a highly pertinent fact considering that water is not fluoridated the same way there as it is here. Much of China’s water is heavily contaminated with fluoride (when this happens in the U.S., the chemical is reduced to a safe level); some children in the studies were consuming nearly three times the EPA’s limit. And in two of the studies, the kids weren’t getting fluoride from their water at all: They were inhaling it from coal burning, which could have much more acute health effects. So the “high fluoride exposure” causing IQ problems in these studies is vastly different from what U.S. children are exposed to—in fact, some Chinese children in the reference groups (i.e., those who were supposedly “unexposed”) were drinking water that was fluoridated at levels found in the United States.
There’s another problem: The research reported an association—kids exposed to lots of fluoride have lower IQs—yet it didn’t account for other potential differences between the children that could explain it. Were the kids in fluoride-contaminated areas also exposed to lead? The authors of the meta-analysis don’t know. (They know that some of the kids were also exposed to arsenic and iodine; when they tried to statistically control for these exposures, fluoride seemed less dangerous, but still bad.) The children could also have differed in socio-economic status or education, both of which can affect intelligence.
I reached out to the U.S.-based co-author of the study, Anna Choi, an environmental health scientist at the Harvard School of Public Health. She declined an interview but emailed me that the results “do not allow us to make any judgment regarding possible levels of risk at levels of exposure typical for water fluoridation in the U.S.” She added, however, that it’s also impossible to conclude that American kids are not at any risk.
Right: If fluoride is potentially dangerous in large amounts, isn’t it best to avoid it altogether? Not necessarily. Yale clinical neurologist Steven Novella, one of the authors of the well-respected blog Science-Based Medicine, put it to me this way: “Everything is toxic at a high enough dose; everything is safe at a low enough dose.” Yes, even water and vitamin C can be deadly when you consume too much. And the idea that something bad at high doses is also necessarily bad at low doses is based in part on the assumption that dose-response effects follow a linear pattern, but many scientists now think that biological responses are more complex than that. Some substances may only be dangerous beyond a certain threshold, while others may follow U- or inverted-U-shaped dose-response curves, such that substances have unexpected effects at high or low doses. (The anti-cancer drug tamoxifen, for instance, can stimulate tumor growth in small amounts.)
There is, however, some evidence suggesting that we should take a closer look at fluoride’s effects on the brain, even at the moderate exposures to which some Americans are exposed. In a 2003 study, researchers gave IQ tests to 222 children living in a Chinese village with water fluoridated to an average level 2.5 times higher than recommended U.S. levels (but still below the EPA’s legal limit) and 290 children of similar ages and backgrounds consuming water fluoridated to lower-than-recommended levels. They found that the children drinking the more heavily fluoridated water had an average IQ of 92, while the less fluoridated kids had an average IQ of 100. In addition, significantly more kids in the high-fluoride village were mentally disabled. Another Chinese study found that 21.6 percent of children consuming an average water fluoride level close to the EPA’s limit had IQs below 70, whereas only 3.4 percent of kids had such low IQ scores in a village with water containing an eighth of that fluoride concentration. (The authors of the 2006 National Research Council report considered these studies when they were evaluating the possible health effects of fluoride but were “unable to assess” the studies’ strength and come to firm conclusions.)
What should American parents take away from all this? Consider that a lot of U.S. water is nowhere near the EPA’s limit of 4 mg/L—New York City, for instance, aims for a fluoridation level of 1 mg/L. If you want to check your water’s status, visit this CDC webpage or contact your local water utility. And if you think your kids are getting too much (particularly if you’re making infant formula with highly fluoridated tap water), consider using bottled water (but check first that it doesn’t also contain high levels—yes, some do). Finally … try not to overreact. If you’re anything like me, you are having urges to order this $200 water cooler and disconnect your household water pipes. But while we do need more and better research on fluoride’s neurological effects, evidence of harm in a few studies on a very different continent where fluoridation is less tightly controlled doesn’t mean that your tap water is turning your kids’ brains to mush. Just like most everything else, fluoride probably isn’t all bad or all good. There may well be a cavity-fighting, intelligence-conserving sweet spot. I just hope my local water utility has found it.