A bit of a storm arose at Scienceblogs the other day when a blogger normally dealing with climate change allowed his father, James Beck, to do a guest post about a new anti-fluoridation book he had co-authored. Needless to say there were strong reactions on either side of the issue, including a response at Respectful Insolence. Since I’d been planning for a while to post something about fluoridation and just hadn’t got around to it, now seems like a good time.
Some of the arguments against fluoridation involve questions about how effective it is – I’m not going to get into that here (though in general the anti-fluoridation sites I’ve looked at use some pretty questionable analyses when downplaying its effectiveness). I’m also not going to get into the ethical side - whether it is right for the government to administer fluoride in drinking water when some of the water users may not want it (though to me it’s not that much different than fortifying milk with vitamin D, enriching flour, etc.). I’m going to focus instead on my own area of expertise, specifically whether fluoride is toxic at the concentrations in drinking water.
First off, it’s important to remember that fluoride occurs naturally in water at a wide range of concentrations. In some places the fluoride is actually naturally higher than desirable and is treated to reduce the concentration.
The original post by James Beck made the following claims about fluoride:
- Fluoride intake is a cause of impaired thyroid function. Indeed fluoride was once used medically to suppress thyroid function.
- Deleterious effects on reproductive systems in humans have been found to be associated with fluoridation: in girls, early onset of menstruation; in men, low sperm counts.
- We now have strong evidence of the association of osteosarcoma in boys and young adult males with fluoridation. Osteosarcoma is a bone cancer which is often fatal.
- The possible incidence of bone fracture with fluoridation has been studied with mixed results. One of the strongest studies is presented in a paper by Li et al. published in 2001 which shows a rising prevalence of hip fracture correlated with a rising intake of fluoride starting with concentrations comparable with those used in fluoridation in North America. And this is just one example that suggests that hip fracture is caused by fluoridated water.
- Fluoride adversely affects kidneys.
- Two recent studies have shown adverse effects on the heart and the aorta.
There are some strong claims here – unfortunately not much linking it to actual concentrations or doses and no references. Dr. Beck later provided a link to the bibliography to his book in the comments, though it still isn’t easy to track down the references for the specific effects, and only a few of the references are to original studies in reputable scientific journals. I imagine the references are more traceable in the book, but I don’t have a copy of it (and probably don’t have enough readers to warrant getting a free copy from the publisher, even if I did have time to read it). However, the effects claimed are based on available scientific literature which is readily available, along with reviews by major health agencies.
I’m going to start with bone toxicity, since fluoride in the body resides primarily in the bone, and it has been shown to affect bone density. Some studies have suggested a biphasic effect on bone strength and fractures, with low doses increasing bone strength and higher doses resulting in a decrease. The paper by Li et al. (2001) referenced by Dr. Beck, which looked at fracture rates in different villages in China, actually shows some indication of this effect, with the lowest rate of overall fractures appearing in communities with fluoride levels around 1 ppm. For hip fractures, the rate was pretty constant at concentrations around 1 ppm and lower, and showed an increase at higher concentrations. The relevant graphs from the paper are below:
Based on these data, concentrations higher than 1 ppm in drinking water appear to be associated with some increased risk of fractures (while the increase isn’t statistically significant until the highest dose, the overall dose-response trend can’t be ignored either). Several other similar studies have been conducted, and included in reviews by agencies such as NRC and Health Canada; results have been fairly mixed, and most of these studies have some fairly significant limitations and confounding factors. However, overall it appears that high fluoride concentrations in water are associated with increased rates of hip fractures in particular, but these effects generally occur above 1 ppm.
Several studies have looked at the relationship between fluoride and osteosarcoma (a cancer of the bone), again with mixed results. Perhaps most notably an exploratory epidemiological study by Bassin et al. (2006) found some indication of a relationship between fluoride exposure and osteosarcoma in males, but not females. However, exploratory studies are notorious for false positives, and require follow-up to confirm their conclusions; even the authors of this study cautioned against reading too much into the results. Analysis of an expanded data set, using the same hospitals, the same pathology departments and similar methods of fluoride exposure estimation did not find any association between fluoride and osteosarcoma. A recent expert panel convened by Health Canada to study the issue concluded the current weight of scientific evidence does not support an association between fluoride and cancer.
As Dr. Beck noted, high fluoride exposures have been associated with other effects such as thyroid effects and kidney stones, but these effects were generally observed at concentrations on the order of 4 ppm. Some studies have shown evidence of reproductive effects, but have had issues such as poor control of confounding factors, and again involved high fluoride concentrations. Available studies have been summarized in the recent reviews by NRC and Health Canada.
Overall the data do not appear to show evidence of significant adverse effects (other than fluorosis) at the “target” concentration in drinking water of approximately 1 ppm. Higher concentrations, particularly in the range of 4 ppm and above, do appear to be associated with a variety of adverse effects and should be avoided – areas where fluoride naturally exceeds this concentration should use appropriate treatment to reduce the concentration. There are certainly limitations to the existing data; in particular, the epidemiological studies on which the conclusions are based would likely only detect population-level effects, and may not detect rare effects in very sensitive individuals. While based on the mechanisms of toxicity and pharmacokinetics of fluoride I wouldn’t necessarily expect huge differences in response between different individuals (at least compared to a lot of organic chemicals), further study is worthwhile. There is also some concern about exposure from sources other than drinking water, which can be significant at concentrations of 1 ppm; generally the better-designed studies have incorporated this, but many of the foreign studies may involve populations with different exposures to fluoride from sources other than water than typical North American populations.