CRITIQUE of the All Party Parliamentary Group on Primary Care and Public Health Inquiry into Water Fluoridation
1.1 This report was published in March 2003, following three 90-minute sessions of oral evidence and the receipt of seventeen written submissions. Because of the way in which the inquiry was set up a number of significant figures in the fluoridation debate declined to take part. In particular the writer of the present critique was concerned  that
(a) An All-Party Group was not the proper vehicle for reaching well-founded conclusions about an issue with as many contentious aspects as fluoridation;
(b) The organisation and structure of the inquiry were not well enough ordered to provide adequate representation of differing points of view;
(c) In particular the failure to ensure that oral evidence was heard from the leading independent scientists from the York systematic review, which remains the best source of scientific evidence, was a serious omission;
(d) This inquiry was not, unlike York, designed so as to minimise bias.
1.2 In the event these concerns have been borne out. The report has emerged as a low-quality contribution to the debate, whose weaknesses this paper examines under the headings of Science, Bias, and Omissions. These occasionally overlap. If the present critique goes into detail it is because the devil is in the detail, and there is much of it. It is not possible to give an accurate and credible response without being at least reasonably specific.
2.1 The principal weaknesses were the failure to take account of the unique status of the York review, and the failure fully to acknowledge and draw conclusions from the generally poor quality of the evidence in the fluoridation literature.
2.2 York is the only systematic, fully transparent and thorough review of the evidence that has been conducted. This means that, unlike other reviews, it looked as far as possible at every potentially eligible study, and unlike other reviews, it assessed them by internationally recognised criteria. It involved the general public in so doing. It relied on no secondary sources, i.e. other people's reviews or opinions, but made its judgements from a clean slate. Also, unlike the MRC and some others, it took care that participants who favoured fluoridation, notably dentists, did not predominate on the inquiry. The transparency was further enhanced by peer review.
2.2.1 This means that it is not legitimate to call in aid other inquiries conducted to a lower standard in areas examined by York, as the All-Party Group does under 6. with Acheson and the Department of Health, under 9.2 with Ireland, and under 14.2 with a number of other well-known bodies - and as the dental profession has done since York reported in 2000. It matters not how 'expert' these bodies may appear, if the careful science of the York review shows their findings to have been unsupported by the evidence. This, after all, is how science advances.
2.3 The quality of the available evidence is all-important. The All-Party Group refers briefly (7.8) to the "little high quality research" that has been undertaken, but it fails to follow through with the self-evident conclusion that any reliance placed on it must therefore be limited. In an age of evidence-based medicine this is not good enough, especially as York were explicit about the unreliable nature of the fluoridation evidence. (The section Omissions below looks in 4.5 at the key question without which it is not possible to decide the scientific case for or against fluoridation on a rational footing, namely "How good does the evidence have to be to support a public health measure treating whole populations?") The Group accepts with little question that the evidence surrounding fluoridation, most of which was categorised by York as "Level C (lowest quality of evidence, high risk of bias)", is good enough.
2.3.1 A further question-mark over evidence arises from the phrases "no conclusive evidence" of a link with cancer or hip fractures (8.1), "no consistent evidence" of an association with cancer (9.1), "no conclusive link" with various ill effects (12.2). Qualifying words indicate that there is some evidence of danger, and to readers of scientific journals such phraseology sends warning signals. The conclusion might be that, if there is some evidence of cancer, better research is now needed to determine whether the inconclusive evidence stands up, because not to clarify this would be irresponsible. At the very least this is arguable, but again the Group does not follow the argument through. At all events the phrases cited are incompatible with the claim in 12.2 that "There is no scientific basis whatsoever" to safety concerns about fluoridation. York has shown that we do not know whether it is safe or unsafe : there are studies that go both ways, across a range of potential harms, but the evidence is insufficient to decide. There is a firm scientific basis in questioning fluoridation's safety, and contrary to what the report states (17.1) Dr. Mansfield's "individualistic" view on this subject is fully in line with York. If the Group had interviewed Professors Kleijnen, Sheldon, Davey Smith or Sir Iain Chalmers from the York review, they would hardly have made this mistake.
2.3.2 League tables and uncontrolled comparisons are not good science, in the absence of proper methodology. It is not legitimate to build arguments on Eire v. Northern Ireland (9.2) or Birmingham v. Manchester (14.1). It has been pointed out a number of times to dental associations that these studies, where they have been done at all, were not even worthy of inclusion under York's Level C; yet they are still being cited to uncritical audiences. The All-Party Group was unfortunately one of these.
2.3.3 The Group has accepted another disreputable argument: that because no patterns of ill health are visible among large populations who drink fluoridated water over time, such patterns can be discounted (12.2; 17.2). One would have expected a health Group to have some awareness of how carefully designed studies are needed to pick up associations between a plethora of environmental factors and the range of ailments that all populations are subject to (as for example with smoking and lung cancer). This is standard epidemiology. Although patterns rarely spring to view, especially in a culture when they are not looked for, increases in cancer death rates were noted in Birmingham after fluoridation was introduced, and 1991 data showed the city to be high in the infant mortality tables (cf. the report's "no reported ill effects" in 17.2).
2.3.4 York did NOT "confirm" the beneficial effects of fluoridation on children. It was the MRC that said that they did. York found, at this stage of the evidence, a likelihood of an effect . The evidential distinction, for policy, is crucial, as it is with 2.3.5 below.
2.3.5 The All-Party Group's statements that the evidence is "strongly" supportive of fluoridation's effectiveness with vulnerable groups (17.4), and (particularly) that there is a "strong probability" of helping with dental inequalities, are quite at variance with what York found, and go way beyond the bounds of good science (York scientists have described the evidence on inequalities as "weak, contradictory and unreliable" ).
3.1 Bias is an enemy of good science, and one of the purposes of modern research methodology is to eliminate it where possible. It can be conscious or unconscious; it is a common belief that the other fellow has biases whereas one is free of them oneself. Bias is not always easy to deal with at other levels in the process, but the problem must nevertheless be faced.
3.2 The history of the fluoridation controversy shows bias to be endemic, on both sides of the argument, and this has been reflected in selective presentation of the evidence. Views are held with great tenacity, and the report is wrong to say that anti-fluoridationists hold their views more strongly than others (17.1). It would not be inaccurate to describe fluoridation as a dentist-led crusade, considering how the issue has been presented from its beginnings on slender evidence after the second world war. One student of the long and politicised debate has argued persuasively that the science and sociology of fluoridation are inseparable: one has always informed the other, for example over which studies are conducted, and how results and arguments are presented.
3.3 Because of the resources, inherent credibility and firepower available to organised dentistry, but not to its opponents, an 'establishment' culture of acceptance of fluoridation was quickly embedded in much of the English-speaking world (but not significantly elsewhere), and it has persisted. The dangers of premature scientific consensus can rarely have been so clearly illustrated, and this has produced a powerful bias against any change of view which, given the absolute terms in which fluoridation's effectiveness and safety have been promoted over the years, would leave the authorities (including Government) considerably exposed. There is nothing uncommon in this, human nature and professions being as they are; but with fluoridation it is unusually strong, and needs to be carefully guarded against.
3.4 The All-Party Group has singularly failed to guard against it - even perhaps to be aware of it. Since the principal arguer for fluoridation is the British Dental Association, with its supporting organisations, it is simply unacceptable to have assigned a key role in assisting with the inquiry and drafting the report to a lecturer in a school of dentistry (1.). The present writer's invitation to participate came from the BDA. The point should be obvious - as obvious as are the examples of bias which abound in the eventual report to an informed observer.
3.5 These examples are both of language and substance. The linguistic devices are familiar to readers of BDA and British Fluoridation Society literature, serving to diminish the opposition or prejudge the case. 5.1 speaks of fluoridation's "benefits", before the science has been considered. 7.2 and 12.2 have anti-fluoridationists as "campaigners", a mildly pejorative term which has not been applied equally to the BDA whose pressure on Government and public has been unceasing; it would be just as accurate to describe them as reactors to a 40-year crusade. They are "high profile" (12.), despite their profile being low in comparison with the well-resourced fluoridationists. Several "failed" to participate, whereas in fact they declined, in company with others who were neutral in orientation, for well-considered reasons which an even-handed report might have detailed. Evidence from the National Pure Water Association (12.1) and Dr. Mansfield (17.1) is presented in slanted language, which suggests (if it does not say outright) that the NPWA were happy about the safety issue, which they were not, and that the objections of both parties were necessarily confined to single issues. On the latter the report contradicts itself between 12.1 which lists the NPWA's three points, and 17.1 which says the Association argued "solely" on one point. Dr. Mansfield's designation as "highly individualistic" reflected his desire to present the medical evidence which he knew best; his presentation of this, and his views on safety as exemplified by York, might not have appeared so individualistic if some of his more senior colleagues on the York review had been called to give evidence, as they undoubtedly should have been.
3.5.1 The report's description of the civil liberties argument as "legalistic" is remarkable (see 4.4 below), and 17.2 shows that the Group has not grasped what this issue is about. It is tendentious to write (17.6) that "it is hard not to have sympathy with the view expressed by the British Dental Association", that no further research is needed, when it is in fact extremely easy not to have sympathy with a view which is difficult to justify after York and the MRC, and which many would regard as dangerously complacent. Language is important, as it predisposes the reader to accept the conclusions put forward, as well as betraying the writer's own bias. In this report it is hostile to opponents of fluoridation.
3.6 Bias over matters of substance may be more serious. The oral evidence did NOT reflect "the full range of opinions" (4.), because those responsible for the best scientific evidence were not given a platform, and may not have thought that the status of this inquiry merited a speculative appearance under a format which they considered unhelpful. 5.1 paints a picture of many millions who "benefit" from fluoridation and of areas in the USA which have recently come on board; because this inquiry was not even-handed it neglected, as do BDA briefings, to say (for example) that more than a hundred US cities have ceased or rejected fluoridation since 1990, or that, fluoridation being a minority activity world-wide, many countries have rejected this measure for a variety of reasons, and the formerly fluoridated Czechoslovakia, DDR, Finland, Japan, Netherlands, Sweden, Switzerland, USSR and W.Germany have ceased to fluoridate (after a minimum period of 19 years). Thus the picture given is a biased one.
3.6.1 5.2 is seriously misleading in citing a 1985 Act as "the reason" that fluoridation remains an outstanding issue: there are a number of reasons, notably the ethical, environmental, and not least the scientific reasons thrown up by York's recent highlighting of the weakness of the evidence, and local communities (not just Government, as implied) will need to grapple with them all. It is a familiar approach among fluoridationists to minimise the objections that have to be met.
3.6.2 In 7.2 it is not just "campaigners" who state that the "benefits found" were less than previously claimed: it is a verifiable fact, unless York's 14.6% is thought to equate with the "massive" caries reduction that dentists were still announcing in 2000, which has traditionally ranged between 30% - 65%. In 9.2 the report is selective in citing Ireland on effectiveness. A balanced presentation would have included European countries which have taken a contrary view; and if fluoridated Eire is compared favourably with Northern Ireland, it should also have been compared with those other unfluoridated European countries which have have been shown to have better dental health than Eire.
3.6.3 9.3 introduces clear bias in the citation of one study (Phipps et al.) which supports the case for safety over hip fractures. The report does not mention that York had already assessed this at Level C ("lowest quality, high risk of bias"); nor that Phipps found an increased risk of wrist fracture; nor that of the studies that reached statistical significance 5 indicated an increased risk of bone fracture and 4 a decreased risk (York 8.5). To say that "the results supported the safety of fluoridation" is therefore highly misleading.
3.6.4 12.2 cites "experts" who are satisfied of fluoridation's safety and effectiveness; but there are experts, who include a Nobel Prize winner in medicine, who are not so satisfied, and this should have been stated. The report's claims about the degree of public support (16.) are open to serious objection, besides raising questions about how the case is presented to the public. It is tendentious to publish the "personal opinion" of one scientist (17.3) without seeking the opinions of other senior scientists, some of whom would not agree with his assessment. And to prefer the opinion of a long-time advocate of fluoridation (the BDA) and its supporters about the need for research, over the clear recommendations of two recent scientific inquiries which drew attention to the significant gaps in the knowledge base, without supporting argument (17.3 & 6, or elsewhere), verges on the perverse. Credibility is not helped by the suggestion in 17.3 that the Chair of the MRC working group did not support his own recommendations for research, or else that this was how the All Party Group chose to interpret his evidence.
3.7 No instances of anti-fluoridation bias could be found in the Group's report, but the writer would be glad to have any such drawn to his attention.
4.1 The report cited the "Key issues" (8.1) from the MRC report. A significant omission was the new emphasis on the importance of total fluoride exposure from all sources (MRC 3.), which may change the nature of the debate on water fluoridation. Citation without examination of the MRC's three "main health outcomes" avoided consideration of whether some of the potential problem areas flagged by York (e.g. infant mortality, congenital defects) might be as important as hip fractures. Once again, where York and the MRC do not coincide, some debate is called for.
4.2 Any reasonably well-informed student of fluoridation is aware that there are a number of strands to the debate which have been addressed over the years, both here and abroad, when schemes have been considered. The ecological/environmental argument, for one, has weighed heavily with Nordic countries, and Denmark has rejected fluoridation on these grounds. At a time when the EU is reflecting widespread concern over the use of toxic chemicals which are not strictly necessary, and in view of the fact that over 95% of fluoridated water goes rapidly down waste-pipes and drains, this aspect deserves a high priority. It received no mention in the report.
4.3 The York reviewers pointed out (York 12.7) that the scientific argument was only one factor in a wider debate, and that in addition to the ecological and environmental effects of fluoridation the cost-effectiveness, legal and ethical dimensions would need to be explored before decisions were taken. If York, which was a purely scientific inquiry, could register this, it is surprising that neither of the subsequent bodies that have pronounced on fluoridation, the MRC or the Audit Commission, have paid them any attention, and that the All-Party Group failed to pick them up also.
4.4 The Group did, however, make brief reference to the ethical or "libertarian" issue (17.1 & 2), but dismissed it in such a way that it was clear that they had not grasped what it was about. It has little to do with whether the measure is "safe and efficacious". It is about the fundamental right of the patient to accept or refuse treatment as s/he alone may decide. This right is accepted in civilised societies, is codified under European conventions, and features in the Patient's Charter. Indeed it is a corner-stone of a modern patient-centred approach to medicine, in contrast to the view that the state or the doctor (or with fluoridation, the health authority or the community) knows best. Any debate on this issue must focus on why fluoridation should be the exception to this rule. This is a major issue of medical ethics and public policy, and in not addressing it adequately the Group has misdirected itself.
4.5 The MRC report, in framing questions for public involvement and discussion, omitted a key question: "How strong does the scientific evidence have to be for a compulsory health measure given to populations?", compared with (say) a drug prescribed one-to-one after consultation. Without this there is nothing against which to judge the available evidence, and people are giving opinions in a vacuum. The All-Party Group has reported what the MRC said about informing the public (16.4), but without any indication of having thought through the adequacy or otherwise of section 2. in the MRC's report, which had other gaps beside.
4.6 One of the principal omissions in the Group's report, therefore, is of any genuinely critical approach to the pro-fluoridation case. It is not enough to call in aid well-known organisations that support previously held beliefs, and report what they say in the same kind of language that has characterised dental association briefings down the years. Readers need to know why best evidence was not sought on the scientific front, and why what the distinguished independent scientists who were closest to the York review have recently written to correct misinterpretations has been disregarded; why it should be right to breach the fundamental right of the patient to refuse treatment in this case (and this case only); why environmental concerns should go by default when the trend within Europe is to use fewer toxic chemicals, not more; why no transparent cost-effectiveness arguments have been presented, even by the Audit Commission; what are the views of those industries that have to remove fluoride from their water supplies before manufacturing; why fluorides which are given to millions indiscriminately should not have to go through the normal licensing procedure for medicinal substances as laid down in European law. It may be because these questions have seldom been put before the public in the selective presentations which have characterised this debate, that fluoridation has been supported by many communities. (It has been opposed by many others.) There is much thinking still to be done, and this report has not helped the thinking process.
5.1 It was not to be expected that an All-Party Group, lacking official status, secretariat and access to funding, could produce a meaningful contribution to a debate as wide-ranging, contentious and open to bias as fluoridation. This is the job for a major inquiry, or perhaps a select committee of one or other House. But the results must be addressed.
5.2 The failings of this report are manifold. Chief among them are these.
(a) The Group were visibly linked to one party to the debate, one of their principal witnesses (the dental profession), both in the setting up and the writing of the report. This was totally unacceptable;
(b) The results can be seen in a bias, both in substance and language, which vitiates any credibility the report might otherwise have had;
(c) The inquiry, in not calling the authors of the York report to the oral sessions, did not obtain best scientific evidence. This evidence was in dispute;
(d) There was too much bad science in the report, which went beyond occasional error;
(e) Whether because of the compressed timescale of the inquiry or an unwillingness to look beyond well-trodden paths, some of the key considerations necessary for any decision on fluoridation were not dealt with at all, or were barely examined.
5.3 Those who decided to stand apart from the inquiry were on the face of it justified, for it was unlikely to have been their failure to attend that resulted in so one-sided a report. The concern now is that uninformed readers, or those who are impressed by a Parliamentary cachet and are unaware that All-Party Groups have no official standing, will take this as a worthwhile contribution to the debate; and that it will join the circle of citations of verdicts from bodies, distinguished and less so, that have passed for good evidence in the years preceding York, and even in the two and a half years since. It is the kind of evidence that piles up, but does not gain weight. That York should still be not only misinterpreted but misquoted is an indictment of the lack of scientific and intellectual rigour that still prevails over fluoridation. The issue, which concerns the health of millions, deserves much better.
28th April 2003
1 Letter, Earl Baldwin of Bewdley to Stephen Hesford MP, 19th January 2003
2 Two letters from scientists on the York review, 3rd January 2001 and 11th December 2002
3 Fluoridation. Chemical & Engineering News, 1 August 1988: 26-41. A British Fluoridation Society briefing Water Fluoridation: Questions and Answers, extant at the time of the York review, states: "Water Fluoridation . . reduces tooth decay by 50%."
4 For a recent example see Andy Burnham MP, giving his "passionately held view" of the benefits of fluoridation on 13th March 2003, Hansard col. 520
5 Brian Martin, Scientific Knowledge in Controversy: The Social Dynamics of the Fluorid ation Debate, State University of New York Press 1991
6 For fluoride as a medicinal product, see Lord Jauncey in the Scottish case M'Coll v. Strathclyde Regional Council, 1983 Sessions Cases, 244
7 Council of Europe Convention for the Protection of Human Rights and Dignity of Human Beings with Regard to the Application of Biology and Medicine, Articles 2 & 5
8 Report Dentistry, September 2002
9 EU Codified Pharmaceutical Directive 2001/83/EEC, Article 1.