The US National Academies of Science (NAS) found the National Toxicity Programme’s (NTP) conclusion in their recent draft report that fluoride is neurotoxic was not supported by the reviewed evidence (see Another embarrassment for anti-fluoride campaigners as neurotoxic claim found not to be justified). This was a huge defeat for anti-fluoride activists who had been roundly promoting the conclusion of the draft report – despite the warning on every page that it “should not be construed to represent any NTP determination or policy” and that the draft had been “distributed solely for the purpose of pre-dissemination peer review.”
So the anti-fluoridationists declared belief that the NTP draft was going to mean the end of community water fluoridation (CWF) was dashed. But that has not stopped Paul Connett, head of the main US anti-fluoride group Fluoride Action Network (FAN) from putting his own spin on the NAS peer review report. In his attempt to present this huge defeat as somehow positive he issued a press release where he claims:
“If the recommendations are adopted, they will make the final report ‘iron-clad’ against criticisms. The NAS suggestions should strengthen the draft report’s conclusion that fluoride is a presumed neurotoxin in children, which is based on 149 human studies. This finding brings into question the long-standing assurances from public health officials that water fluoridation is safe.”
The NAS peer review is very critical – and not just about the fact that the draft conclusion was not supported by the reviewed evidence. There are criticisms of biased study selection, insufficient consideration of the effect of confounders and problems with statistical analyses used. But besides playing with words to misrepresent the peer review, Paul Connett has declared a belief that the NTP will now consider new studies, published since the draft report was released, and these will provide the evidence missing in the original draft. He is hoping the revised review will include new “strong” (in his opinion) studies which will finally swing things his way:
“Multiple strong scientific studies, at exposures relevant to fluoridation, have been published after the NTP’s review. They link fluoridation in Canada to greatly lowered IQ in formula-fed infants (Till 2020) and 300% higher rates of ADHD (Ridell 2019); fluoridation in USA with sleep disturbances in adolescents (Malin 2020); and fluoride with lower IQ by thyroid disruption (Wang 2020).”
Notably, Connett actively ignores, yet again, a new study which does not fit his bias because it reports a positive relationship between maternal prenatal urinary F and child IQ – the complete opposite of what he wishes (see The anti-fluoride brigade won’t be erecting billboards about this study). But Connett has been specific so we can objectively consider just how strong or weak these new studies are and evaluate for ourselves the likelihood that they can be used to provide support for the unwarranted conclusion expressed in the draft NTP report.
I have discussed several of these studies in previous articles but briefly consider them together here.
Till et al (2020) formula-fed child study
I discussed this study in my article Anti-fluoride propagandists appear not to read the articles they promote. It’s citation is:
Till, C., Green, R., Flora, D., Hornung, R., Martinez-mier, E. A., Blazer, M., … Lanphear, B. (2020). Fluoride exposure from infant formula and child IQ in a Canadian birth cohort. Environment International, 134(September 2019), 105315.
This has been blindly promoted by anti-fluoride propagandists as finding a decrease in IQ for children formula-fed as babies and a negative relationship of IQ for these children with drinking water F. However, the relationships are not statistically significant. It’s possible those anti-fluoride activists claiming a significant relationship had been influenced by Till’s previous conference paper where she indulged in special pleading for a non-significant relationship.
The findings reported by Till et al (2020) are summarised below:
While there are no significant relationships for IQ (FSIQ – Full-Scale IQ), the authors did report a significant negative relationship of performance IQ (PIQ) with water fluoride for children both breastfed and formula-fed as babies (and also for estimated F-intake for formula). Strangely, the anti-fluoride promoters of this paper generally don’t refer to this – perhaps because very few of them actually read the papers they promote.
Why there is a difference between IQ and PIQ is a bit of a mystery but there is a recommendation that PIQ, a subset of IQ tests, should not be used for clinical decision making because it is not consistent with current standards (see Beware of scientific paper abstracts – read the full text to avoid being fooled).
Riddell et al (2018) ADHD study
My article ADHD and fluoride – wishful thinking supported by statistical manipulation? discussed this study. Its citation is:
Riddell, J. K., Malin, A., Flora, D., McCague, H., & Till, C. (2019). Association of water fluoride and urinary fluoride concentrations with Attention Deficit Hyperactivity Disorder in Canadian Youth. Submitted to Environment International, 133(May), 105190.
The authors expressed disappointment that they could not find any relationship of ADHD prevalence to urinary F which they considered the best measure of F exposure. However, they were able to squeeze out some significant relationships with water F when the data was divided according to youth age. Significant for older youth but not younger. The findings of Riuddell et al (2019) are summarised in this figure.
The relationships are tenuous with large confidence intervals indicating their weakness. Manipulation of the data by age to find significant relationships reminds me of the saying that if data is tortured enough it will produce the answer you want
So, again, hardly a “strong” study and I cannot see Connett’s wish that the NTP will be able to justify its unwarranted draft conclusion with this.
Malin et al (2020) sleep “disturbances” study
I reviewed this study in my article Sleep disorders and fluoride: dredging data to confirm a bias. The paper citation is:
Malin, A. J., Bose, S., Busgang, S. A., Gennings, C., Thorpy, M., Wright, R. O., … Arora, M. (2019). Fluoride exposure and sleep patterns among older adolescents in the United States: a cross-sectional study of NHANES 2015 – 2016. Environmental Health, 1–9.
First of all, this paper has been poorly peer-reviewed. Effectively just an “in-house” review by coworkers all involved in publishing the same work (see Some fluoride-IQ researchers seem to be taking in each other’s laundry). This figure illustrates the relationships between the authors and their coworkers working with the same and similar data:
Relationships between Malin and her peer reviewers as indicated by joint publications. Links to the papers listed from the top are: Malin & Till (2015), Thomas et al (2014), Bashash et al (2017), Marlin et al (2018), Malin et al (2018), Bashash et al (2018), Thomas et al (2018), and Riddell et al (2018)
I think this sort of self-selected peer-review during journal publication is almost scientifically corrupt. In my article, I concluded that the journal was chosen for publication and the absence of genuine peer review, together with the weak findings, indicated “this is another poor quality paper on fluoride and health effects which make unwarranted claims – and which will be used by anti-fluoride activists in their campaign against community water fluoridation.”
The study used a range of parameters which facilitated dredging of the data to find apparently statistically significant relationships (see Statistical manipulation to get publishable results). None of the relationships with blood plasma F were statistically significant (despite a bit of special pleading by the authors). They did report significant relationships with water F for sleep apnea, snoring, bedtime and waketime. An apparent reduction of snoring with an increase in water F baffled the authors. The confidence interval for sleep apnea was very large and small differences in bedtime and wake time are easily understood as connected with social differences between rural and city residents. There were no relationships with sleep duration.
So much for Connett’s description of this as a “strong scientific study” capable of reinstalling NTP confidence in their unsupported draft conclusion of a neurotoxic effect of fluoride.
Wang et al (2020) IQ study
This study is reported in:
Wang, M., Liu, L., Li, H., Li, Y., Liu, H., Hou, C., … Wang, A. (2020). Thyroid function, intelligence, and low-moderate fluoride exposure among Chinese school-age children. Environment International, 134(September 2019), 105229. https://doi.org/10.1016/j.envint.2019.105229
Although Connett claims the study was “at exposures relevant to fluoridation” this is not so. The study was aimed at measuring “low-moderate fluoride exposure affects” in the Chinese context and involved comparison of children from areas of “endemic and non-endemic fluorosis areas in Tianjin, China.” Water fluoride concentrations ranged from 0.20 to 3.90 mg/L and urinary F from 0.01 to 5.54 mg/L. Not at all comparable to the concentration where community water fluoridation (CWF) is present.
However, the study did separate results into different quartiles according to water F and urinary F concentrations and the data for the lowest quartiles are worth considering because they are for concentrations similar to that present where CWF is used.
I have extracted that data for quartile 2 of water F (0.7 to 1.00 mg/L) – a bit high compared with CWF but the data for the more relevant quartile 1 was not presented. Also the data for the second and third quartile of urinary F (UF) with the ranges 0.15 to 0.41 mg/L and 0.451 top 2.28 mg/L. Again these values are high.
Linear regression in these ranges showed no significant relationship of child IQ with either water F or Urinary F (UF) (see figure below). So much for Connett’s claim that this is a “strong scientific studies, at exposures relevant to fluoridation” that will lead to the NTP strengthening the NTP’s unsupported draft conclusion of a neurotoxic effect.
The elephant in the room – Santa-marina et al (2019)
Of course, Connett studiously ignores this study because it reported a positive relationship between maternal prenatal urinary F and child IQ (see The anti-fluoride brigade won’t be erecting billboards about this study).
The citation for this study is:
Santa-Marina, L., Jimenez-Zabala, A., Molinuevo, A., Lopez-Espinosa, M., Villanueva, C., Riano, I., … Ibarluzea, J. (2019). Fluorinated water consumption in pregnancy and neuropsychological development of children at 14 months and 4 years of age. Environmental Epidemiology, 3.
I have summarised the findings reported by Santa Marina et al (2019) below. These are certainly not going to provide any hope to Connett that the unsupported NTP conclusion will somehow be retrieved.
Connett is simply attempting, in his press release, to put a brave face on the embarrassment of the NAS peer review report. He and his supporters had been actively promoting the unsupported conclusion of the draft report – even suggesting it would shortly lead to the end of CWF. So a public embarrassment and he responds by holding out hope that his position will be retrieved by new studies.
His reliance on new studies is effectively an acknowledgement that the studies considered by the NTP were not adequate for the draft conclusion presented. However, he is wrong to describe these studies as “strong.” They are actually very weak and are hardly likely to change the final NTP assessment. He has also actively ignored, and hidden, the fact that the new studies are contradictory (Santa-marina et al 2019 produced the opposite result to what Connett wants) and conflicting (the different studies rely on different measures of cognitive ability and fluoride intake as well as manipulation of the data to find significant relationships).
The revised NTP report will not be produced soon as the peer reviewers identified a range of problems which will require major changes. There may well be even newer studies reported in the interim and these may not help Connett’s case any more the ones he cited. On top of that, the NTP may well discover older studies that their biased selection process missed in their original literature search (I still hold out hope that they will this time in include Perrott (2018) which, I believe, raises important issues about the effect of confounders ignored by other studies).