4 York scientists' open letter

October 2004

Dear Secretary of State,

We write to protest in the strongest terms at the way in which the recently published fluoride bioavailability study at Newcastle has been conducted and presented. We have been advised in this by independent medical scientists of the highest standing who are experts in trial methodology and interpretation.

The Newcastle study was commissioned by the Department of Health in the wake of the York review of the evidence for water fluoridation (2000), and the subsequent MRC report (2002). There had been conflicting views, arising from weak and contradictory studies on bioavailability, on whether or not artificially and naturally fluoridated water had similar effects on humans. This has implications for policy and research. It is not legitimate to give sole emphasis to the view held by the dental community that the effects were identical. To resolve this question the present study was assigned to the School of Dental Sciences at Newcastle University.

We consider the outcome to be deeply disturbing, for the following reasons.

1. Conduct of the Newcastle Study. We accept that there was much that was well thought out in the design of this trial. The flaws, however, were so fundamental as to vitiate this entirely.

• Unbelievably, the study involved only 20 participants. The authors admitted that this gave no power to detect differences unless they were of the order of 33%, which is very large. The object of this study was to detect differences, if they existed.
• The study did detect a difference in the blood samples, which was very large and statistically significant, suggesting that artificially fluoridated water was much better absorbed. Faced with this, the authors discarded the results of one participant. This was not legitimate. With 19 subjects, the difference remained, but was no longer statistically significant.
• The conclusion was presented, first of "no statistically significant difference", and then of "no evidence for any differences". No reference was made to the trend that was found towards greater bioavailability of artificially fluoridated water. The conclusion was given "with some caution", because of the "limits imposed by the small number of subjects". These limits were wholly self-imposed by the chosen design study, which makes the conclusion virtually meaningless.
• There are serious questions about the validity of the urine calculations, which have no particular relevance for the absorption of fluoride from soft water which may be more bioavailable. These are the subject of ongoing correspondence.

The enclosed sheet of summary comments explains why the Newcastle study is not scientifically defensible. It needs to be re-run, with adequate numbers and statistical power, by experienced researchers with no association to the pro- or anti-fluoridation lobby.

2. Presentation of the study. The study, which had been delayed well beyond the expected date, was published in June 2004 as a report to the Department of Health. The terms of its announcement by the Department (see Written Answer in Lords Hansard for 28th June, WA 6) were inaccurate and misleading.

• It was announced only as showing "no evidence for any differences". There was no mention of the authors' specific caution over this, or of the consistent trend towards greater bioavailability of artificial fluoridation.
• It was also claimed as a "favourable result", providing "further reassurance of the safety of fluoridated water". This study did not address safety. The only reliable assessment of safety was carried out by the York systematic review, commissioned by the Department. It found that there was no good evidence from either naturally or artificially fluoridated areas, and that more research needed to be done. Until then, as the two public letters by independent scientists on that review make clear (enclosed), the safety issue remains unresolved.
• The study has been through an internal DoH process, but has not yet been published in a peer-reviewed journal. There has not therefore been the usual opportunity for external scientific criticism. (Nor will there be such opportunity, for a public health study for which so much is being claimed, if it is merely submitted to a dental journal.)

In spite of this, it is already being used to underpin the Chief Medical and Dental Officers' advice to Government (Written Answer, 28th June), and to support public calls by dental health spokespeople for further fluoridation schemes.

It is because of the seriousness of this last point that we are writing to you without waiting for further clarification from the authors of the study, which is proving difficult to obtain.

It is clear to us and our advisers that the York review itself, and everything that has appeared since then, has consistently been interpreted in the light of what Government and the dental profession previously believed about fluoridation; so that it was little surprise to read in another Written Answer (Lords Hansard, 7th September, WA 138) that the Newcastle results were "compatible with the conclusion" that was previously believed, namely that in bioavailibility "there is absolutely no difference between added and natural fluoride". This is known as 'confirmation bias'. It is bad science and it can lead to bad policy.

We are extremely disturbed, as are our scientific advisers, by the conduct and interpretation of this study, and propose to give objections wide circulation.

Yours sincerely,

Graham Brady MP, Brian Donohoe MP, Earl Baldwin of Bewdley
Vice-Chairs

Back to the Archive index.