Tuesday, 7 May 2019

For F's sake?

F is for Fluorine, just above chlorine on the right-hand side of the periodic table. My mad-scientist PhD student waged a war with his supervisors to let him order a bowser of HF hydrofluoric acid for him to play with. Most of us [poor bugger had four (4!) supervisors in the all-colonels-few-footsoldiers traditions of The Institute] r'ared back in terror and nixed his persistent requests; so it never happened. HF is far more active that hydrochloric acid HCl  - it is used to etch glass for example and so cannot be stored in glass bottles. A spill of that stuff was likely to be significantly more damaging than when another student on my watch rested her arm in a puddle of sulphuric acid. But there's a place for everything and context is important. Chlorine Cl is a pale green gas that was used to kill and injure soldiers in WWI, sodium Na and water mix with a dramatic exothermic >!boom!< reaction - but we sprinkle NaCl [too] liberally on our chips (and on the roads) every day. Fluorine is used in two important 21stC processes.
  1. Uranium hexafluoride is used to separate the extremely fizzy U235 from boring old U238which makes up 99% of the natural uranium on the planet. without that enrichment we wouldn't have nuclear powerplants and more of medieval central Hiroshima would be still standing.
  2. Hexafluorosilicic acid [HFSA structure top R] is added to Irish water by Irish Water in the long-standing but still controversial practice of water fluoridation.
Q. Why are we adulterating the water with fluoride?
A. Because the evidence is that it reduces tooth decay in the young.
The evidence was a lot of case-controlled and epidemiological studies in which the decayed, missing, filled, tooth DMFT index was tallied up for children in areas with fluoride in the water and adjacent regions without fluoride Geordieland for example. It looked like there was clear evidence that a little jolt of HFSA in the communal water supply would reduce the number of fillings in children and so The Patriarchy authorised it. In the same way that The Patriarchy mandated MMR vaccination and polio prophylaxis and, more recently, HPV vaccination against a future clouded with the chance of cervical cancer. It's the herd immunity, silly. But then The People were wroth and they rose up against The Patriarchy and shouted that it was all a con. The trouble is that a go[o]dly proportion of the people would rather heed a shouty and articulate advocate-against [aka Con] rather than The Man even if The Man [often a Con but here Pro] can muster a lot of scientific evidence to support that position. It's up to each of us to find out what the truth is by doing a bit of reading and digging and critical thinking - rather than nodding at tendentious assertions from The Blob.

I was all behind The Man  on fluoridation until I polled my students on their prior beliefs and then  tasked them to do some research, marshal their best evidence and present it in a formal debate with the opposition among their class mates. That was interesting because I was forced to change my position. The most compelling evidence was the more or less uniform decay in the rate of decay since WWII across European countries whether they fluoridated the water or not. It's like a rising tide that floats all boats - better tooth brushing, better diet (yer yer: even with all the sweets and sodas kids are better fed than they were when I grew up). Indeed vaccination might play a part in this. We didn't MMR The Boy and aged two he got The Measles. Later on he had a symmetrical pair of really crappy spongy teeth. The dentist said the problem might have stemmed from those teeth being at a crucial stage in development when they were assaulted by measles virus.

Well now, up in Republican Alaska where everyone is free to make their own mistakes and nobody can expect a hand-out from the government - because that makes you weak - they decided to stop water fluoridation after a long and vociferous campaign to get The Man out of our taps.  It's kinda nutty: Juneau is the Capital of Alaska but has a population of only 32,000 [Kilkenny 26K, Navan 30K Dundalk 39K] and goes for a tedious amount of direct democracy like it was a 17thC New England town with a town meeting for all adults. In 2007 several  propositions were put to the people about how their tax dollars would be spent:
Proposition Money Y N
Raise 1% sales tax ??? 6104 3523
Fluoridate water ??? 3802 5975
Returf baseball field $4m 3999 5792
Build new swim pool facility $20m5402 4416
Renovate 2 elementary schools $22m 6683 3133
So they switched off the fluoride tap in accordance with the people's wishes. Ten years later Jennifer Meyer, a Professor of Public Health up the road in Anchorage, led a study to measure the effects on the dental health of the children [exec summary from sciencealert] in Juneau. Their parents had voted for a couple of refurbished schools and a new swimming pool in exchange for no fluoride in the water. Instead of comparing two neighbouring communities they measured two different cohorts of children before [2003] and after [2012]  the end of fluoride in Juneau. Their conclusion is that there has been an up-blip in the amount dental work children required under conditions of sub-optimal community water fluoridation CWF compared to under the previous optimal CWF. Sub-optimal and optimal are the phrases used by the researchers so you know that they are in flavour of CWF.
Age Mean (SD) N 2003 Mean (SD) N 2012 U-test
0 to < 7 2.01 (4.22) n = 303 2.68 (4.57) n = 461 0.004
7 to < 13 1.61 (3.38) n = 352 1.64 (2.60) n = 400 NS
13 to 18 2.75 (4.73) n = 198 3.04 (4.66) n = 191 NS
0 to 18 2.02 (4.05) n = 853 2.35 (3.99) n = 1052 0.001
U-test is the Mann-Whitney U test to compare samples that are not normally distributed; but mean and SD [standard deviation] are statistics used where the data is normally distributed. The significance of the difference reported above lies entirely in the youngest cohort where, on average, they had 2.68/2.01 = 33% more fillings without fluoride than with it. But check out the enormous standard deviations: SD gives an estimate of  distribution of the data about the mean.  If I assume that the data is normally distributed, then I can do a t-test and this gives t = 2.07 for the under 7s which is on the borderline (0.05 > p > 0.01) of significance and t = 1.77 for all the children in the study which is definitely nothing to see here not significant

Now it happens that Dorothy Bishop has written a piece to camera for Nature 2 weeks ago: Rein in the four horsemen of irreproducibility. The discussion of Meyer's paper has a whole section on 'Limitations' which suggests that she suspects her findings won't be reproducible. The study has a number of limitations. First, due to funding limits, only two years of data were purchased for comparison rather than five to ten years of data, which would have enabled a more sophisticated trend analysis. Whoa! What? Public Health analysts have to buy data from Medicaid? Isn't Medicaid a federal scheme funded by tax-payers? Shouldn't that data be available [anonymized] to other authorised / qualified public employees?  IF they know after the event that their study is underpowered and isn't going to give a clear answer about the benefits of fluoridation THEN they should have known before all their hard work & analysis that it was likely to be a busted flush for actually answering the question. Ante hoc Power Analysis please - G*Power will do.

And what about the NS no significant difference in dental treatment for kids aged 7 to 18? That is explained by the authors as the 2012 kids getting some fluoride at the crucial time of development of tooth enamel before fluoride was taken off the market. But a more skeptical assessor might just read it straight that F has no effect on tooth costs for that cohort. Let the plain people of Juneau move on; fluoride is history, there are far more worrying things come down the tracks for Alaska - like melting glaciers. Five years ago, when we last had local elections [much less exciting than the direct government by proposition that they have in Alaska], I wrote about jangling the fluoride chains of aspirants to public office. Local elections are coming round here again on the 24th May - I'll dust off my fluoride arguments to test the mettle of candidates.

No comments:

Post a Comment