dissecting the predictable (in this case, ffnz on broadbent et al)

A paper just out by Broadbent et al (2014) describes research which used data from a 38-year-long longitudinal study of Dunedin children to examine claims that exposure to fluoride in childhood has a negative effect on children's IQ. The paper found these claims wanting, and thus – quite predictably – it's now subject to attack by antifluoride activists, both on twitter and via press release. 

And so on twitter (among many other scurrilous tweets) we get this: "@JayMan471 @OpenParachute A longitudinal study? This is hilarious. [The] twit has little concept of science."

Which is hilarious, all right, but only in that it clearly shows that the tweeter has no idea how a longitudinal study works.

And on scoop, a Fluoride Free New Zealand release entitled "Dunedin IQ-fluoride study Sloppy Science", in which the writer schools peer-review referees in how to do their job.

"Not only were there too few unfluoridated children in this study to give reliable results…"

However, careful statistical analysis overcomes this hurdle – the paper identifies the statistical tools used in their analysis. The FFNZ spokesperson continues:

[It] is likely that they were a large proportion of the 139 children who took fluoride tablets. If so, there is no ‘unfluoridated’ group. Why was this not disclosed?

Probably because a careful reading of the paper shows that this question has been considered by the authors; fluoride tablets had no effect on children's IQ (nor did other sources of F- in their environment). From the paper: 

Residence in a CWF area, use of fluoride dentifrice and intake of 0.5-milligram fluoride tablets were assessed in early life (prior to age 5 years); we assessed IQ repeatedly between ages 7 to 13 years and at age 38 years (Broadbent et al, 2014).

Back to the FFNZ release:

But worse, the study fails to allow for a whole range of confounding factors. The most important period for IQ damage is in the womb, yet the mothers’ fluoride intake and other factors like iodine deficiency were not controlled for.

Well, no: 

Preschool fluoride exposure was used in these analyses because this is when brain development is rapid and vulnerable, and thereafter the IQ is known to be relatively stable. Studies of twins indicate that environmental effects on IQ are greatest in the early years, and genetic effects are least during that period (Broadbent et al, 2014).

The charge of not measuring mothers' fluoride intake, possible iodine deficiency, & so on can also be levelled at the 'Harvard' study. A major EU review found many of the various fluoride-IQ studies also evaluated in the Harvard paper to have methodologies with little or no control for various confounding factors such as iodine, lead, & socioeconomic status.

Other confounders in those studies are highlighted by Broadbent et al (2014), who note that these include other environmental sources of fluoride (many of the villages sampled were exposed to F- from industrial exhausts), differences in the size of village populations, proximity to schools and "the lack of relevance of the studies included in the meta-analysis to the use of CWF or fluoride toothpastes" (ibid.).

However, FFNZ's spokesperson continues:

Similarly, there was poor information on total fluoride intake by these infants. Had the study actually been prospective as claimed, rather than retrospective, this essential information could have been available.

The researchers state – quite explicitly – that fluoride intake was not measured directly. 

FFNZ again:

The Dunedin research report begins with the conclusion it set out to “prove” – that fluoridation is harmless. The first two named ‘researchers’ are two of NZ’s leading political promoters of fluoridation. They are dentists, not developmental neurotoxicologists.

The paper by Broadbent and his colleagues (a published, peer-reviewed paper, not just a 'report') begins with a statement that is accepted by health authorities the world over: that CWF is cost-effective in reducing dental caries, safe, and environmentally friendly. They did not 'set out to prove' this, but to examine the purported link between fluoride exposure and IQ. 

In contrast, a Harvard University meta-analysis of studies was conducted by some of the world’s leading expert researchers into developmental neurotoxicology, who have no known bias on fluoridation policy. There were 27 studies reviewed. The total number now available is 43. The Dunedin authors wrongly dismiss this as a single study.

The authors of Broadbent et al correctly point out that the metanalysis by Grandjean & Choi was just that, a single review of a large number of papers. It's interesting to note at this point that FFNZ is intent on a not-so-subtle attempt to downgrade the paper by Broadbent et al, with the laboured contrast between 'Harvard University' and 'Dunedin' (rather than Otago University), and the insinuation that because two of the authors of this 2014 paper are "dentists, not developmental neurotoxicologists" they are somehow incapable of performing statistical analyses of data. In fact, every author of the paper holds a PhD ie a research degree – the same qualification as that held by Grandjean and Choi. In addition, FFNZ is really being quite inconsistent here, given that recent TV ads paid for by this advocacy group feature a dentist giving his opinion on the safety of fluoride. I doubt that the individual concerned is a developmental neurotoxicologist either.

FFNZ is also glossing over the fact that the paper has multiple authors from a range of disciplines, not all of whom are faculty members of Otago University (itself an internationally-respected institution), and conveniently ignoring the additional fact that the paper has gone through a rigorous peer review process for an international journal. But let's go on.

The Harvard review rightly caused concern to decision-makers as it showed a consistent lowering of IQ associated with fluoride intake. It is clearly the reason this Dunedin study has been published – as a political posturing, just as the tobacco companies funded and published ‘research’ showing cigarette smoking did not cause lung cancer.

And what are we missing here? That the "Harvard review" found that IQ was lowered by exposure to fluoride levels much higher than that found in fluoridated water. The authors themselves conceded that their findings did not provide evidence of risk at the levels of fluoride used in community water fluoridation. Regardless of this, FFNZ continues:

The Dunedin IQ-fluoride study is missing just about all the confounding factors that the authors have criticised in the studies reviewed by the Harvard team. This is outright sloppy science. Broadbent criticizes the studies reviewed by Harvard for not controlling for these factors (when in fact some of them did) and then fails to control for them when the data is readily available to him.

Really? How does that fit with statements such as this, for example?

Cohort families represent the full range of socioeconomic status (SES) in the general population of New Zealand’s South Island (Broadbent et al., 2014).

This (says the FFNZ spokesperson) is a single ‘study’ by politically driven dentists against a robust analysis of 27 studies by world-leading experts in this field, from one of the world’s foremost universities. You’d have to have a pretty low IQ to not know whose opinion carries the more credible weight.

No, FFNZ, I'm afraid that your statement above is an example of the logical fallacy known as 'appeal to authority'. It's also worth pointing out that 12 months ago the Deans of the Harvard Medical School, School of Dental Medicine and School of Public Health distanced themselves from this study, saying that they

continue to support community water fluoridation as an effective and safe public health measure for people of all ages. Numerous reputable studies over the years have consistently demonstrated that community water fluoridation is safe, effective, and practical. Fluoridation has made an enormous impact on improving the oral health of the American people.

I'll look forward to seeing a peer-reviewed letter from Fluoride Free NZ to the American Journal of Public Health, critiquing the paper so that their specific criticisms can be replied to by the original authors in the usual way.


J.M.Broadbent, W.M.Thomson, S.Ramrakha, T.E.Moffitt, J.Zeng, L.A.Foster Page & R.Poulton (2014) Community Water Fluoridation and Intelligence: Prospective Study in New Zealand. American Journal of Public Health Published online ahead of print May 15, 2014: e1–e5. doi:10.2105/AJPH.2013.301857

4 thoughts on “dissecting the predictable (in this case, ffnz on broadbent et al)”

Leave a Reply

Your email address will not be published. Required fields are marked *