The following letter is in response to the recently released York Review. The writer, Paul Connett, was a peer reviewer of the study. A synopsis of the York Review can be accessed on the web at http://www.bmj.com/cgi/content/full/321/7265/855
October 11, 2000
British Medical Journal
As an invited peer reviewer of the York Review of fluoridation, please permit me to add a few concerns about this report and the spin it has been given in the press. While a number of people have correctly pointed out that this York Review deals a very serious blow to water fluoridation based upon the findings of a much lower benefit for caries reduction than claimed by promoters, and a much higher level of dental fluorosis than is acceptable, I am very disturbed that the authors are not making it clear that dental fluorosis is in fact an indication of a toxic effect of fluoride, in a similar fashion that the blue line on gums was an early indicator of lead poisoning. The fact that neither tooth mottling nor the blue line causes any obvious pain is beside the point. Tooth mottling indicates that fluoride has poisoned the enzyme involved in removing protein from between the mineral prisms prior to enamel formation (DenBesten, 1999). It raises – or should raise -the obvious question of what other enzymes might have been poisoned by fluoride which have no visible effects. British researcher Jennifer Luke (1998) may have answered this question in her Ph.D thesis. It would appear that fluoride accumulates in the human pineal gland and experiments in animals indicate that fluoride lowers melatonin levels. Unfortunately, the York reviewers didn’t consider Luke’s work, nor others’ work on the thyroid gland or many animal experiments conducted over the last 10 years which wave very large red flags of concern about fluoride’s potential health effects. I have reviewed many of these concerns elsewhere (see www.fluoridealert.org).
Turning to an issue that the York team did examine, the authors claim that there is no evidence of a relationship between increased hip fracture in the elderly and exposure to fluoride in drinking water. I feel that properly interpreted the current literature provides considerable evidence of such harm.Paul Wilson of the York team is quoted in an Associated Press report (Oct 5, 2000) as saying that “the studies he examined tracked the effects of up to 4 mg per liter in drinking water”. This is not accurate. In their review of hip fracture the York team included an unpublished study by Li et al (1999) which they indicated showed no association between fluoride and hip fracture. The York team was only able to draw this conclusion because they limited their comparison to hip fractures in the village with low fluoride levels with the hip fractures at 1 ppm (i.e. 1 mg per liter). When I reviewed this study, as part of my invited review, I found Li et al examined the hip fracture rates in six Chinese villages with different water fluoride concentrations: 0.25 – 0.34 ppm; 0.58 – 0.73 ppm; 1.00 – 1.06 ppm; 1.45 – 2.19 ppm; 2.62 – 3.56 ppm, and 4.32 – 7.97 ppm. The authors chose the village with 1 – 1.06 ppm fluoride for their baseline for comparison. Their reported odds ratios for hip fracture were 0.99; 1.12; 1.00 (control); 2.13; 1.75 and 3.26. In other words they found an approximate doubling of the hip fracture rate for fluoridation levels over 1.5 ppm and tripling for water levels over 4 ppm.
In my view, the York team did a disservice by confining themselves to hip fractures in villages with fluoride concentrations at 1 ppm or below. Paul Wilson compounds this disservice by claiming that they actually considered studies which looked at effects up to 4 ppm (4 mg per liter), when in this critical case they clearly did not.
Paul Wilson further distorts the scientific record when he claims in the same AP report that, the fear of osteoporosis has never been based on strong evidence. In actual fact, since 1990 there have been 18 studies (4 unpublished, including one abstract) which have examined a possible association with hip fracture in the elderly and fluoride in the drinking water. 10 of these found a positive association (increased hip fracture rates) and 8 did not. Moreover, these studies were prompted by the fact that when fluoride (50 – 80 mg per day) was used to treat patients with osteoporosis in an effort to harden their bones and reduce fracture, it was found that the treatment increased not decreased hip fracture rates (Hedlund and Gallagher, 1989 and Riggs et al, 1990). The fact that 50% of all the fluoride that we swallow accumulates in our bones should give us pause. Increased hip fracture in the elderly is not a minor problem. One in four of those suffering a hip fracture in the U.S.are dead within a year of their treatment and one in two never regain an independent existence.
It is unfortunate that the York team has managed to convey the message that there is “no evidence of harm” when in actual fact there is “mixed evidence of harm”. A critical difference. They also beg the question of how one should resolve mixed epidemiological findings when one is giving advice to a government which is considering putting a toxic substance into the drinking water of every man, woman and child in the population. The precautionary principle says “If in doubt, leave it out”. As far as hip fractures are concerned, the York Team says, based upon meta-analysis of a mixed bag of poor studies, we find no evidence of harm. Not very reassuring and not very helpful when they themselves avoided the key piece of dose-response evidence presented by Li and co-workers.
Dr. Paul Connett,
Professor of Chemistry,
St. Lawrence University, Canton, NY 13617.