Anti-fluoride hypothyroidism paper slammed yet again

lie.w.stats

Credit: Display of Statistical Data

Anti-fluoride campaigners often cite a limited number of papers published in scientific journals as evidence for their claims that community water fluoridation (CWF) is harmful. They do this to give some sort of scientific credibility to their claims – but the citations are far from scientific as they are usually cherry-picked, misrepresent the cited papers or use poor quality papers which are often not peer-reviewed, or poorly peer-reviewed, and written by fellow anti-fluoride campaigners.

This year, anti-fluoride campaigners have heavily promoted two such papers claiming harmful effects of CWF:

The hypothyroidism paper (Peckham et al., 2015) received negative reviews immediately after publication and a number of peer-reviewed responses have since been published. The latest response by Foley (2015) is useful because it gives a thorough critique of several faults in the Peckham paper. It is available in full text on-line:

Foley, M. (2015). Fluoridation and hypothyroidism – a commentary on Peckham et al. British Dental Journal, 219(9), 429–431.

Here is a summary of some of the points covered by Foley:

Confounding effects almost ignored

The sort of statistical analysis of data used in the Peckham et al., paper should always consider confounding effects – otherwise the analysis is simply an exercise in confirmation bias.  Foley says:

“Peckham et al. show little understanding of confounding factors, and have made only a token attempt at considering their impact. The authors mention the ecological fallacy, but then ignore its implications in strongly implying a causal link between water fluoridation and hypothyroidism.”

The confounding factor that sticks out like a sore thumb is iodine – or its deficiency. Iodine deficiency is associated with hypothyroidism and there are moderately to severely deficient levels of iodine across the UK. Foley points out:

“Vanderpump et al. have already shown that schoolgirls in fluoridated Birmingham appear to be more likely to show moderate-to-severe iodine deficiency than schoolgirls in many non-fluoridated UK cities . .”

statistics_joke_round_stickers

Credit: Mu-Peter

Iodine intake could potentially be associated with fluoridation status, yet Peckham et al. discounted its effects claiming “it is unlikely that there are significant differences [in dietary iodine intake] between people, living in fluoridated and non-fluoridated areas.” That is just not good enough and simply confirms the bias of the authors – especially as they had already referenced a study showing “greatly differing urinary iodine levels in groups of schoolgirls from major UK cities, with 17% of participants showing moderate to severe deficiency.”

As Foley points out:

“. . iodine intake is associated with the outcome variable, hypothyroidism. Iodine intake could certainly confound any statistical association between fluoridation status and hypothyroidism. So why wasn’t this considered by the authors? Authors also failed to consider the impacts of smoking, medications and other factors known to contribute to hypothyroidism.”

Ignoring or downplaying confounding effects is a common trick used by those who have a bias against CWF. I referred to this in my articles Connett misrepresents the fluoride and IQ data yet again, and Connett & Hirzy do a shonky risk assessment for fluoride. Connett uses data showing a correlation of IQ with urinary fluoride explaining only 3% of the variation in IQ. The statistical  analysis did not consider confounding effects which could easily have eliminated the correlation with fluoride.

Another example is the ADHD paper of Malin & Till where a significant correlation with CWF disappeared when confounding effects were included in the statistical analysis (see  ADHD linked to elevation not fluoridationADHD link to fluoridation claim undermined again).

Poor review and citation

Authors inevitably show some bias in the citations they select, but anti-fluoridation proponents carry this to the extreme. Citations are cherry-picked to support their case and the rest of the literature usually ignored (or ridiculed) to give the impression there is a large amount of scientific support for their case.

Foley criticises the Peckham paper for this sort of citation – and for “inappropriate self-citation.”

self-citations-is-it-worth-doing-them

Credit: Self-citations, is it worth to work on them?

Peckham et al., demonstrates how to use self-citation as a way of getting claims from a poor journal into a  mainstream journal. Foley mentions their self-citation of a earlier paper opposing CWF – Peckham & Awofeso (2014), Water Fluoridation: A Critical Review of the Physiological Effects of Ingested Fluoride as a Public Health Intervention, The Scientific World Journal Volume 2014 (2014):

Scientific World Journal has a well-publicised history of colluding with other journals to self-cite authors’ papers to increase journal impact factors. The referenced Peckham and Awofeso paper is unusual in that many unattributed paragraphs, including factual errors, are almost identical to paragraphs from a previous Awofeso paper. This paper in turn has unattributed paragraphs almost identical to paragraphs on webpages belonging to the National Institute of Dental and Craniofacial Research and a popular chemistry website.

I discussed this journal and Peckham & Awofeso (2014) in my article Peer review, shonky journals and misrepresenting fluoride science.

So, get a biased article published in a shonky journal (usually by paying publication fees to avoid proper peer-review) and then cite it in later papers to give dubious credibility to current claims.

But again and again Peckham et al., (2015) cites papers which don’t really support the claims they make. For example, they cite Feltman & Kosel to support their statement that “the effects of fluoride on the thyroid have long been observed.”

“But Feltman and Kosel’s only mention of the thyroid is a single sentence that other researchers ‘…report that fluoride is a thyroid inhibitor’. Their only references for this statement are a personal communication with US Public Health Service researcher Floyd De Eds, and a 1954 paper that studied delayed tooth eruption in rats following removal of the pituitary gland. The evidence supporting Peckham et al.’s statement is extremely weak.”

Similarly, Peckham et al referenced a 1958 Galletti and Joyet paper to claim fluoride reduced thyroid activity:

“However, Galletti and Joyet only investigated the effect of fluorides on patients with hyperthyroidism, and did not find that fluoride was linked to goitre. . . . . Regarding the reduction in thyroid activity in patients with hyperthyroidism treated with fluoride, Galletti and Joyet concluded ‘…such an action appears only occasionally among persons subjected to massive doses of this substance’, a situation clearly not comparable to community water fluoridation.”

Peckham et al., concluded there are “questions about the safety of community fluoridation and consideration should be given to reducing all sources of fluoride in the environment.” But Foley points out:

“Three references are given to support this statement; however, two of the references provided make no such alarmist recommendation. The only one of the three that does so is Peckham’s own 2014 paper.”

Poor conclusions

Foley’s opinion is that “the paper’s conclusions can and should be dismissed” as it has “serious biases and flaws.”

He says:

“Literature reviews have been highly selective and critical analysis of that literature has been poor. The authors show a disturbing tendency to focus on a small number of poor quality studies that reinforce their own views, while ignoring contradictory evidence from much stronger studies and reviews.”

And provides a bit of advice:

“Peckham et al. should have heeded the adage ‘correlation is not causation’ before coming to a conclusion at odds with a large body of reputable evidence from around the world.”

correlation

Credit: xkcd – correlation

Similar articles

159 responses to “Anti-fluoride hypothyroidism paper slammed yet again

  1. That’s rich coming from you lot – we all know the whole fluoridation scandal started with a big fraud!

    Like

  2. So cities thinking about fluoridation should first check whether that halogen can worsen the effects of a deficiency of another halogen, iodide, and if found yes for some genetic types, it would only be fair to attend to that first.

    Like

  3. Let’s see…

    We know lack of iodine (not iodide) can contribute to hypothyroidism.

    We know that 17% of schoolgirls in major UK cities have moderate to severe iodine deficiency.

    Historically New Zealand has had high levels of goitre and hypothyroidism. This has been corrected by adding iodine to the diet.

    Peckham ignores the above and claims that miniscule increases in the intake of fluoride are the cause of hypothyroidism. If he was correct then NZ, with its mostly low freshwater levels of fluoride, would have had increasing levels of hypothyroidism as fluoridation has occurred. In reality we see the opposite.

    Like

  4. Stuartg. We can eat a tiny amount of elemental iodine, as it reacts in Lugol’s Solution. Iodine is the least strongly electronegative of the halogens: fluorine, chlorine, bromine, all of which burn tissue in their elemental; form. (They have seven electrons in their outer shell and seek an eighth to complete 4 electron pairs.)

    Iodine tincture with elemental iodine may be used as a disinfectant on skin.

    Our diets are supplemented with iodised salt, but that does not have elemental iodine in it. Just potassium or sodium iodide or iodate. Even then it may stop proper bacterial action if used to try to make sauerkraut instead of non-iodised (common) salt.

    The chemical properties of the halogens are somewhat similar and if our intake is out of balance they can confuse the body.

    Excessive seaweed consumption with its high “iodine” dose can cause thyroid trouble as can a deficiency.

    Now that there has been a move to reduce salt intake that source of iodine is lessened as has the iodine from iodophors no longer used to wash milk machinery.

    Some relevant data in a thesis talking about prevalence and incidence of thyroid disease, which does not however deal with fluoride.

    https://researchspace.auckland.ac.nz/bitstream/handle/2292/8341/whole.pdf%3Fsequence%3D2

    ” living in close proximity to factories that
    expel substances, such as prolybrominated biphenyls and polychlorinated biphenyls, into the air, which have been reported to cause hypothyroidism in exposed workers)”

    Like

  5. “Does not however deal with fluoride.”

    …because fluoride has nothing to do with thyroid disease. See the Peckham references.

    Like

  6. The hormone produced by the thyroid gland is triiodothyronine. There are three iodine atoms in the molecule. No fluorine atoms.

    Since there are no fluorine atoms in triiodothyronine, fluoride cannot influence its production. (BTW, neither can the other halogens, bromine, chlorine and astatine)

    Like

  7. That is T3, Stuartg. T4 is another thryroid hormone. Deiodinases are important for thyroid metabolism, but affected by fluoride.

    Like

  8. soundhill,

    Perhaps you should look up things before you comment.

    T3 is the active hormone, triiodothyronine.

    T4 is the minimally active prohormone, thyroxine, that is enzymatically converted to the (much) more active T3.

    Neither of them involves any halogen other than iodine. Fluorine, chlorine, bromine and astatine are all halogens that have no effect whatsoever on thyroid function. Iodine is the only halogen that has an effect on thyroid function.

    The current clinical summary is that someone with hypothyroidism is given thyroxine. Within the body thyroxine is converted to triiodothyronine, the active hormone. The person then becomes euthyroid. Giving the person fluorine, chlorine, bromine or astatine will have no effect on their hypothyroidism. Peckham is wrong and Foley is right.

    Like

  9. PS. T4 has no fluorine atoms, either. CWF has nothing to do with thyroid disease.

    Like

  10. Stuartg: “Within the body thyroxine is converted to triiodothyronine, the active hormone.”

    Yes, by a deiodinase.

    “Triiodothyronine (T3) is identical to T4, but it has one less iodine atom per molecule.” wiki.

    Fluoride interferes with the deiodinase.

    Like

  11. Brian – could you provide a citation for you claim ” Fluoride interferes with the deiodinase.” – at concentrations used in CWF?

    Like

  12. The study
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3890436/pdf/40064_2013_Article_766.pdf
    claims that of children drinking water at about 1mg/l fluoride, 10% had derangement in thyroid hormone levels. However that was only in a group of 10.

    But in larger groups of 30, where drinking water fluoride ranged from 1.6 to 5.5 mg/l, 67 to 77% were deranged.

    Though no graph is shown a highly significant correlation of -0.7 between free T3 and water fluoride is stated.

    It would be interesting to extrapolate the curve backwards and think about doing a bigger study. Noone has cited this study yet, it seems.

    Like

  13. Seriously, Brian, is this your evidence for you claim ”Fluoride interferes with the deiodinase.” – at concentrations used in CWF?

    Like

  14. wiki suggests 1.2 mg/l is in the optimum range. And sports people drink more and get more.
    And a Dunedin council worker said the process is not accurate, and opposed lowering the level to what it is now, if I remember.

    When you say “concentrations used in CWF” do you mean current ones or what people have been exposed to from CWF earlier?

    The study I cited puts up a caution and maybe there will be another lowering.

    Like

  15. “In 2014, it was revealed that Springer had published 16 fake papers in its journals that had been computer-generated using SCIgen. Springer subsequently removed all the papers from these journals. IEEE had also done the same thing by removing more than 100 fake papers from its conference proceedings.

    In 2015, Springer retracted 64 of the papers it had published after it was found that they had gone through a fraudulent peer review process”

    Source https://en.wikipedia.org/wiki/Springer_Science%2BBusiness_Media

    Like

  16. Brian, you are straw clutching here. One person was involved and it does not support your claim of “Fluoride interferes with the deiodinase” at that concentration.

    This is pathetic.

    Like

  17. soundhill,

    You’re not serious about that as a reference, are you?

    In the discussion the authors say “Fluoride competes for the receptor sites
    on the thyroid gland which respond to TSH”. The authors are asserting, without reference, that an ion consisting of a single atom competes with a 28-30 kDa glycoprotein for the receptor designed for that glycoprotein.

    That’s like saying a lifebelt will compete with the Queen Mary for a dock. A citation is needed for such an assertion.

    There are several other statements in their discussion that lack references, but desperately need them, including “fluoride is known to interfere with the activity of the deiodinases.”

    Definitely clutching at straws.

    Like

  18. Stuartg, “competes” or somehow grabs them or bits of them so that the normal complex binding action cannot occur. Some info here: https://books.google.co.nz/books?id=R3jpCAAAQBAJ&pg=PA138&lpg=PA138&dq=reactivity+of+the+fluoride+ion+with+large+molecules&source=bl&ots=2UfwGqyAe5&sig=GdnodLO1Yo4KYT7shLXXFJJbkw0&hl=en&sa=X&ved=0ahUKEwietdHby6bJAhVC56YKHUw_BcQQ6AEISzAJ#v=onepage&q=reactivity%20of%20the%20fluoride%20ion%20with%20large%20molecules&f=false

    As for the deiodinases I see there may be too much confidence in the reference given. Lin was really only hypothesising if you check the fourth last paragraph before the refs in this: http://poisonfluoride.com/pfpc/lin-1991.pdf

    The study I cited which we are discussing, possibly in translation, has not been too clear in differentiating between hypotheses to explain the strong experimental results, and more accepted theories.

    Like

  19. soundhill,

    Perhaps a more recent reference is appropriate when debating current levels of understanding? Even a textbook from this century? The one you selected is 49 years old.

    Clutching at straws again.

    Like

  20. Stuartg, the author of the 2012 book: https://books.google.co.nz/books?hl=en&lr=&id=YfnlBwAAQBAJ&oi=fnd&pg=PA1&ots=JG_T1qWlsV&sig=BOcMfh0RiHZ_jJCFPJons8ZfBhA
    believes that Matthews in that “old” book (1970) “has reviewed the literature thoroughly on the use of fluoride as a metabolic inhibitor in cell culture studies and has emphasised that fluoride inhibits not only glycolysis but oxidative phosphorylation in general” p78 and adds more.

    While it is good to have a theory as to why things are happening that does not alter the fact that they do happen.

    Would you have denied the use of aspirin until its full mechanism of action were described?

    Like

  21. soundhill,

    About your Matthews reference: “This volume is the completion of work initially planned several years ago as a compilation of selected aspects of the biological effects of fluorides. The first portion appeared in Volume XXII (1966) of this Handbuch, under the title “Pharmacology of Fluorides, Part 1.” Part 2 is dated 1970 even though, as stated, it was planned prior to 1966. It’s now 2015 – so that’s a very old textbook for you to be using as a reference.

    I stand by my comment that your reference is approaching half a century old.

    You then produced another textbook that simply says he “reviewed the literature thoroughly” and no more. Ignoring everything published on the subject after 1970 is not exactly keeping up to date, is it? Research into fluoride (or even hypothyroidism) didn’t just stop in 1970.

    P.S. It pays to look at the printing data within the book: “© 1984 Plenum Press, New York. Softcover edition of the hardcover 1st edition 1984” Your book was published in 1984, 28 years before your stated 2012. It’s already well past a quarter century old.

    Like

  22. soundhill,

    Wow, Poisonfluoride! I don’t think that I could think of a name that says a website is anti-fluoride better than that.

    That appears to be an otherwise unpublished paper where the authors searched out children with congenital hypothyroidism (using that regretable old medical term “cretinism”) and described them. Two groups had “low iodine” in drinking water, 5.21 mcg/L and 0.96 mcg/L (a factor of over 5 between them). The control group used iodised salt or oil.

    They examined a total of 769 children (but their group figures add to 749, meaning poor editing and no peer review) in the three areas and of those 769 (or 749) children, 104 had “mental retardation”. 69% of the 104 had subclinical congenital hypothyroidism. That seems to me to be either a very unusual or a highly selected population.

    They mentioned “high” and “normal” fluoride water concentrations, 0.88 mg/L and 0.34 mg/L respectively, but since they neglected to measure/mention water fluoride concentrations in the control group any comments on fluoride intake are made irrelevant.

    Absolutely no relevance to a scientific discussion on hypothyroidism and fluoride.

    Like

  23. Lin FF, Aihaiti, Zhao HX, Lin J, Jiang JY,
    Ma L, Maimaiti, Ai K. The relationship of a low-iodine and high-fluoride environment to subclinical
    cretinism in Xinjiang. Endem Dis Bull 1991;6:62-8. [in Chinese]

    And Fluoride Alert and Poison Fluoride have got hold of a translation to publish. Probably 249 turned to 269 in the typesetting of the English version.

    Endem Dis Bull seems a respectable Chinese journal.

    Translaters are reported to be responsible for many articles being retracted.

    The experiment comparing two fluoride levels and low iodine intakes seems valid. It would have been nicer to have fluoride levels for iodine supplemented area. Maybe they did not give them since they show nothing if iodine is adequate.

    Like

  24. Brian, you are strrawclutching agian to resort to this one and a half page newsletter article. Here is what I wrote about that article in my critique of Theissen and Neurath’s article used by FFNZ.

    “Theissen stressed that in the Choi et al review “One study had “high” at 0.88 mg/L, quite relevant to CWF.” Neurath says “In fact, one of the Chinese IQ studies had an average water concentration of 0.88 mgL in the high exposure group.”

    At first sight this seems relevant to CWF and Paul Connett, like many anti-fluoride activists, stress this study in defending the relevance of Choi et al (2012). Strange then that none of them actually discuss the study details. Perhaps we should.

    The study is a one and a half pages newsletter report:

    Lin et al (1991). The relationship of of low-iodine and high-fluoride environment to subclinical cretinism in Xinjiang. Iodine Deficiency Disorder Newsletter, 24–25.

    It has few of the details we normally expect in scientific papers. For example, I would like to know what the range of fluoride concentrations was in the drinking water, what other dietary intake occurred, how was the “dental fluorosis” observed defined, etc.

    Children from low iodine areas were compared with a group from another area that had received iodine supplementation. About 15% of the children suffered mental retardation, 69% of these exhibited subclinical endemic cretinism. The effect of iodine supplementation was clear, the effect of fluoride not so clear. But anyway, hardly a report to hang any conclusion on about CWF in New Zealand.”

    Like

  25. Ken, I feel that your statement: “Children from low iodine areas were compared with a group from another area that had received iodine supplementation.” gives the impression that the study was two-pronged. However the low iodine prong was further divided, and made into a study involving water fluoride level, too.

    And where you say: “I would like to know what the range of fluoride concentrations was in the drinking water,” the concentrations were given for the divided low-iodine prongs.

    Children in the low iodine groups had worse thyroid trouble in the higher fluoride area.

    You say that is not relevant to NZ, however iodine deficiency in NZ was thought severe enough to start mandating the use of iodised salt in bread manufacture in 2009. Even that may not be a sufficent supplement in pregnancy.

    https://consultations.health.gov.au/chronic-disease-and-food-policy-branch/review-of-the-2006-nutrient-reference-values-for-a/supporting_documents/AustraliaandNewZealandNutrientReferenceValuesforIodine.pdf

    Like

  26. Brian – you may not have read the newsletter article you are talking about – here is a link to it: https://openparachute.files.wordpress.com/2015/11/fa-fu-et-al-1980-low-iodine-and-high-fluoride-environment-to-subclinical-cretinism-ln-xinjiang-term-effects-of-theiodine-slow-re.pdf

    I originally commented:

    “It has few of the details we normally expect in scientific papers. For example, I would like to know what the range of fluoride concentrations was in the drinking water, what other dietary intake occurred, how was the “dental fluorosis” observed defined, etc.”

    No range of fluoride concentration ind drinking water was given – only averages for two of the three regions:

    Region A “The goiter prevalence was 91 % and dental fluorosis 20.80%. The average level of iodine in drinking water was 5.21 ug/L,and that of fluoride 0.88 mg/L.”
    Region B: “The goiter prevalence was 82% and dental fluorosis of 16.00%. The average water iodine level was 0.96 ug/L and that of fluoride 0.34 mg/L.”

    They purely established statistically significant differences between these regions and a region where iodine was supplemented.

    And you want to draw conclusions from New Zeland from a very short newsletter article with limited data taken from a region where:

    “The detection rate of subclinical endemic cretinism in children with mental retardation was 69%, and the total attack rate of subclinical endemic cretinism 9%.”

    What is the level of cretinism in New Zealand?

    Extreme straw clutching.

    Like

  27. Ken you asked: “What is the level of cretinism in New Zealand?”

    Though the study was not about full blown cretinism it was about sub-clinical cretinism where the IQ, especially of progeny, may be reduced a few points.

    From the ref I gave:

    “Pettigrew-Porter et al, 2011
    New Zealand
    Cros-sectional
    N=170
    UIC (spot urine),
    TV (ultrasound),
    FT4 & TSH
    Semi-quantitative
    FFQ
    The median UIC of the women was 38 lg ⁄ L.
    7% of women had goitre.
    Iodine intake was 48 lg ⁄ day. The majority
    of women had TSH and FT4 concentrations
    within pregnant reference ranges,
    suggesting that despite the low UIC
    observed in these women, thyroid
    hormone production was not affected.”

    “Majority” is a very fuzzy term. In an election a majority is over 50%.

    That implies that in less than 50% of NZers they tested, thyroid hormone production was affected.

    In 7% ultrasound showed increased thyroid volume.

    You wrote: “They purely established statistically significant differences between these regions and a region where iodine was supplemented.”

    No it was more than that. The title even suggests it.
    “THE RELATIONSHIP
    OF A LOW-IODINE
    AND HIGH- FLUORIDE ENVIRONMENT
    TO SUBCLINICAL CRETINISM
    lN XINJIANG”

    They showed iodine deficiency brought about subclinical cretinism and in the iodine deficiency group fluoride worsened it.

    Like

  28. soundhill,

    I pointed out that the “high” and “normal” fluoride areas had a factor of ~2 between the levels in the low iodine areas.

    I also pointed out that the iodine levels of the same areas had a factor of ~5 between the two.

    Didn’t you read that?

    Before you attribute differences in an iodine dependent condition to fluoride levels (even ignoring no control), you have to account for the much greater differences in iodine intake. It’s much more likely any differences are related to differences in iodine levels.

    Like

  29. Stuartg, Iodine can cause thyroid trouble in large doses, as with seaweed eating. But seaweed is very high in iodine, it can be something like a 10th of a percent. So in 10 g of seaweed there could be of the order of 10,000 micrograms of iodine.

    At the higher of the study doses drinking say 5 liters of water would give 25 micrograms. That is still not enough iodine. Too little. NZers in the doc I gave are getting twice that. And if it is not in the water it won’t be in the soil and food.

    Like

  30. No, Brian, you are wrong. They did not “showed iodine deficiency brought about subclinical cretinism and in the iodine deficiency group fluoride worsened it.” They simply showed statistically singificant differences in a number of measurements between different areas. That does not “show” a mechanism.

    Like

  31. Thanks, Ken, I realised that after I had posted but we cannot edit posts.

    The study showed a significant negative correlation between T3, &c and fluoride in the non-iodine-supplemented group.
    So we wonder about a mechanism, but I don’t think the mechanism is the increased iodine, as I said to Stuartg, since the level has not risen enough.

    I also realised as regards my post to Stuartg that some land or freshwater algae and fish concentrate iodine.

    (If we have the correct organsims in our teeth pellicle they concentrate fluoride, I think.)

    http://www.atsdr.cdc.gov/toxprofiles/tp158-c6.pdf

    The iodine levels in that do not seem to agree with this data about NZ seaweeds

    http://www.tandfonline.com/doi/pdf/10.1080/01140671003619290

    Like

  32. soundhill,

    Did your reference have anything to do with seaweed eating? Short answer – no. Long answer – no.

    Your comment is irrelevant and trying to divert from the fact that your reference is completely separate from the discussion that fluoride levels in water are unrelated to hypothyroidism.

    Like

  33. soundhill,

    What is the measurement “Ig/L” you are using in your cut & paste?

    I cannot find any reference to it in the official SI units.

    (Is it related to the Ig Nobel prizes for improbable research?)

    Like

  34. No, Brian, the brief newsletter article does not report any correlations at all.

    Like

  35. Ken, though they have not quantified a correlation they have discussed one, and considered reasons for it.

    Stuartg, the ig, lg (small l), or Ig (capital I) was puzzling to me, too.

    This is another problem with typesetting or such.

    mg is milligrams, micrograms is mcg or µg. The font for the Greek letter mu (µ) is not always available so it seems the i is used since it follows the m in micro.

    Like

  36. Brian, you claimed:

    “The study showed a significant negative correlation between T3, &c and fluoride in the non-iodine-supplemented group.”

    Yet, what the newsletter note actually said is:

    “The relationships of IQs of 130 patients with complete data of 15 factors were studied by correlation
    analysis and stepwise regression analysis. Factors closely related to IQ were, in order, age, frequency of knock, weight/height X 100% and TSH.

    No mention of correlation with fluoride.

    Like

  37. Ken wrote:
    “Yet, what the newsletter note actually said is:

    “The relationships of IQs of 130 patients with complete data of 15 factors were studied by correlation
    analysis and stepwise regression analysis. Factors closely related to IQ were, in order, age, frequency of knock, weight/height X 100% and TSH.

    No mention of correlation with fluoride”

    Now go back to the title:
    ““THE RELATIONSHIP
    OF A LOW-IODINE
    AND HIGH- FLUORIDE ENVIRONMENT
    TO SUBCLINICAL CRETINISM
    lN XINJIANG”

    Really the study is in two parts:

    ALPHA: THE RELATIONSHIP
    OF A LOW-IODINE ENVIRONMENT
    TO SUBCLINICAL CRETINISM
    lN XINJIANG

    and BETA: THE RELATIONSHIP
    OF A LOW-IODINE
    ****PLUS**** HIGH- FLUORIDE ENVIRONMENT
    TO SUBCLINICAL CRETINISM
    lN XINJIANG.

    In part ALPHA they do not seem to have measured the fluoride water environment, for area C, (the iodine-supplemented control area) Though they have measured urinary fluoride in area C. The average was equal to that of Jiayi, the higher value area of the two low fluoride areas.

    With the overshadowing effect of adequate iodine in improving IQ maybe they did not get p<0.05 for negative "effect" of urinary fluoride.

    The results for part BETA are gleaned from these words:
    "(A) low iodine, high fluoride;
    (B) low iodine, normal fluoride;
    and
    (C) iodine supplemented, normal fluoride.
    Results for the following parameters
    for areas A, B, and C, respectively were:
    (a) average IQ: 71,77,96;
    (b) average auditory threshold
    (in dB): 24,20,16;
    (c) bone age retardation (%):
    28,13,4;
    (d) thyroid '3'1 uptake (%):
    60, 50,24;
    and (e) serum TSH (ILU/ml):
    21,11,6.
    Statistically significant differences
    existed between these areas,
    suggesting that a low iodine
    intake coupled with high fluoride
    intake exacerbates the ceptral
    nervous lesions and
    the somatic developmental
    disturbance of iodine deficiency."

    Just repeating those last few words which is what the ALPHA part of the experiment is about:
    the ceptral (should be central)
    nervous lesions and
    the somatic developmental
    disturbance of iodine deficiency."

    and the "exacerbation" is the BETA part.
    No signficance figures given but significance claimed and seems obvious from the drop of IQ from 77 to 71 between the two low iodine areas, but one of which had higher fluoride.

    Like

  38. Brian, despite your attempts you cannot support your claim:

    “The study showed a significant negative correlation between T3, &c and fluoride in the non-iodine-supplemented group.”

    A statistically significant difference between 2 populations is not the same as a correlation with any particular parameter.

    Like

  39. soundhill,

    You certainly didn’t understand my comment on “Ig/L”.

    To put it bluntly: if a reference contains a uses a word or measurement that you don’t understand, how can you be certain it says what you think it does?

    In this particular case, if a reference uses a non-existant unit of measurement, exactly how reliable can the reference be? You even admitted you didn’t understand the reference – so why did you use it?

    (In case you don’t understand, those are rhetorical questions. I believe I already know the answers – refer to Dunning-Kruger)

    Like

  40. soundhill,

    Official SI advice, shared by many others, is to write or type “mcg” for microgrammes if not using the mu.

    There’s no need for you to invent (repeated) typesetting errors or undocumented SI abbreviations. Just reject the reference because its meaning is not clear.

    Like

  41. Thanks Ken, going back to my original statement: “showed iodine deficiency brought about subclinical cretinism and in the iodine deficiency group fluoride worsened it.”
    And later I said: “though they have not quantified a correlation they have discussed one, and considered reasons for it.”

    OK I was using the word “correlation” loosely, which is fairly implicit since a correlation technically needs to be a figure to show a strength of a relationship, and I said they had not quantified.

    Doing a rank correlation on
    71 77 IQs
    2.56 1.46 fluorides in low iodine areas

    gives r = -1 but with no p value so it is a useless result or is it?

    Stuartg I brought in seaweed eating to admit extra iodine can cause thyroid trouble, but not in the extra quantity you were suggesting between the two low iodine areas.

    Like

  42. Stuartg, ug may also be used to indicate micrograms.

    Search Google ig micrograms

    better put in -immune to get rid of the immunoglobulin entries.

    With the doses the article I cited gave it was pretty obious they were micrograms.

    You seem to be looking for little nigglety things to exclude articles from discussion rather than deal with the substance of them.

    Like

  43. so, after all this diversion you acknowledge I was correct.

    But, come off it, you now attempt a regression with 2 points – not even data points but averages!!

    You are the one inventing little nigglety things to include an article in discussion rather than deal with the substance of them.”

    Extreme straw clutching.

    Like

  44. Ken, then may I try very slightly better.

    Taking the 3 low iodine areas, rather than just 2, whcih of course you were justified in complaining about, and doing correlations, then partial correlations, I get:
    IQ rel UF -0.90 -1 At a constant T3 IQ is very strongly negatively correlated with urinary fluoride.
    IQ rel T3 0.98 1. At a constant urinary fluoride IQ is very strongly postively correlated with T3.

    So far T3 seems a bit more causal than UF. (bigger jump from non-partial to partial)

    T3 rel UF -0.78 1 At a constant IQ, urinary fluoride is very strongly postively correlated with T3, but negatively if IQ not held constant. Still thinking on that. Any ideas?, though since p values are not near <0.05 I suppose that will be an escape from discussion.

    But including all 4 areas the 3 low iodine and the adequate iodine area
    IQ rel UF -0.49 0.17
    IQ rel T3 0.73 0.64
    T3 rel UF -0.77 -0.70

    which seem to show that when iodine is adequate, then partialling out T3 very much changes the IQ rel UF correlation.

    To me that seems to reinforce my opinion that fluoride is not the risk to IQ when iodine is adequate.

    And iodine has not been adequate in NZ or it would not have been mandated for bread in 2009. It's getting better.

    Like

  45. Brian, instead of this extreme straw clutching (3 “data points now!!) why not take the word of the authors who checked correlations with the original data?

    You are just coming across as pathetic.

    Like

  46. Ken they found a stronger thing with TSH. So I’ve used that and my effect is stronger, too, if I’ve got it right. I wonder if they tried correlating across the low iodine areas the way I have done, and compared the partials with correlating across all areas. Pathetic?

    With that language I just feel you are trying to detract from considering how fluoride may be a problem to people short of iodine.

    Like

  47. Brian, you arguments amount to extreme straw-clutching and are pathetic because you ignore their reported findings from the regression analysis and then attempt to submit 2 or 3 “data points” (in reality averages of data points) to an unwarranted statistical analysis.

    You don’t have access to the data so why not accept the authors point that fluoride does not show up in their regression analysis?

    Like

  48. Yes Ken, they said:
    “Factors closely related to
    IQ were, in order, age,
    frequency of knock,
    weight/heightX100%
    and TSH”

    That was for that part of the study.
    I have put their table into OpenOffice Calc and used the CORREL function to correlate each factor with IQ.
    With the caution that these are only taken on the four values (CORREL seems to allow for some missing values) Urinary F is down the list a bit, but not as far down as Knock frequency. Their list I have quoted at the top may be starting from lower and moving to higher correlation to fit with my results.

    Iodine 131 uptake (%) -0.99
    Hearing threshold -0.96
    T4 (lLg/dl) 0.93
    TSH (ILU/ml) -0.93
    T3 (ng/dl) 0.73
    Urine I mcg/g creatine 0.65
    urine f mg/L -0.49
    Grip strength ratio 0.46
    Knock frequency 0.44
    Action stability 0.31
    Mistake frequency -0.25
    rT3 (ng/dl) 0.18
    Reaction time (msec) -0.18

    Urinary fluoride was not on that particular list of theirs, but several times they refer to its possible or likely effect. Why isn’t weight/height in the table if it is
    notable, and age?

    Like

  49. soundhill,

    If neither “peer review” nor the article’s editor pick up that the authors are using non existent units of measurement, exactly how certain can you be that no other errors made it through the same process? Even though you are certain, everyone else has doubts.

    Next paper, my initial thoughts before going into details:
    Sure, they mentioned fluoride in the title, but as Sir Humphrey Appleby said “It does less damage there than in the text.” What is the water fluoride level in the control group? 0? 0.6? 6.0? Without the control level, the other levels are meaningless.
    What is the range of water fluoride levels in group A? Group B? Why are they completely silent about water fluoride levels in group C?
    What is the range of water iodine levels in the three groups?
    Why didn’t they discuss the 5x difference in water iodine levels between group A and B? Surely iodine intake is relevant to subclinical hypothyroidism?
    What is the normal range for TSH, T3, T4 in non-subject children in the same areas? A labs’ normal range is given for healthy adults and is usually different from the normal range for children.
    What is the normal range for urine fluoride in children? In these areas? How is it affected by diet? Does iodine supplementation in group C change it? (Fluoride levels in the salt? In the oils?)
    How did they select the children for inclusion?
    How do the selected children differ from the rest of their peer group?
    Why is there such a high level of subclinical hypothyroidism in the included group? Non-random selection?

    These are not questions about fluoride, or about hypothyroidism. They are questions that should have been raised at the peer review stage of publication. Actually, they should have been raised and answered in the text before it went to peer review!

    I’m a generalist. I’m not an expert in fluoride or hypothyroidism, although both are within my perview. I do know how to read scientific papers, though, and these are epic fails.

    Like

  50. Stuartg, That “Ig” representing micrograms in my quote from:
    “Australian and New Zealand
    Nutrient Reference Values for Iodine
    A report prepared for the Australian Government
    Department of Health
    By Expert Working Group for Iodine.”

    It is a draft copy from May and if it has not been fully published yet it may be worth writing to them and reporting that their cut and paste has turned the Greek letter µ into I.
    This is the article they got it from in which it shows as µ.
    http://www.ign.org/cm_data/2011_Prettigrew_Are_PW_in_NZ_ID_A_cross_sectional_survey_ANEJOG.pdf

    From that paper it seems that urinary iodine may not be a good indicator of deficiency in NZ pregnancies. I wish they had given the proportions from the various cities on their route map, they only say very few from Auckland. Maybe it is worth writing to them too, to get an idea of the proportion of their subjects from fluoridated vs non-fluoridated regions.

    As regards the units problem from cut and paste, it was not me that caused the µ going to I and it happened in several of the refs. But I note the same thing has happened to me a bit in my cutting and pasting into OpenOffice calc in my comment yesterday evening. I corrected to mcg sometimes.

    Like

  51. Brian, you are atempting to make far too much of this article. After all, it was only a note in a newsletter, it was very short and does not stand up to the standards required for a peer reviewed paper.

    it is idotic to use a statistical package in the way you have with only a few data ponts which are averages. Why not just accept what the authors say about their results from a regression analysis of the original data (they did not report F as having a signficant realtionship to IQ.

    All you can rely on legitimately is that statistically significant differences were found in a few paramters between 2 or 3 regions. That is not correlation and it is desperate to use this in your arguments.

    Like

  52. Thanks Ken,
    As I suggested to Stuartg I have communicated about both matters to Sheila Skeaff. I have drawn her attention to your link to the Lin FF article so she may comment.

    My use of the CORREL is just for rough scoping, a suggestion of where to be looking.

    The distribution of iodine intake vs TSH &c probably won’t be a straight line.
    Once iodine becomes sufficient it will level off. My request is to look at the earlier part of the graph, before it does so, the BETA part as I termed it.
    Though Lin FF et al discuss that part, before the change in slope, their main table loses that power.

    Like

  53. Brian, who is Sheila Skeaff and why do you involve her?

    There is no need to use a statistical package for “rough skoping” – besides being inappropriate the informaiton in Lin’s table is all you need , surely.

    Anything else require access to the data and/or further work.

    Like

  54. Ken, Sheila Skeaff, Otago is one of the authors of both of the publications I have been talking to Stuartg about.
    http://www.otago.ac.nz/humannutrition/staff/otago063239.html
    Lin FF et al did not have UF-IQ on their high list of correlations in the work which spread across all iodine levels, deficient and sufficient. But as my work roughly showed it seems to be there though a bit lower on the list.

    And as I have earlier showed, in the region before iodine becomes sufficient and changes the slope of the graph, it would be, UF level shows a stronger correlation, though with my poor figures.

    So let’s see what Sheila Skeaff says.

    Like

  55. Brian, “your work” doesn’t show anything. The table listed the parameters and indicated where there were statistically significant differences between regions. You cannot achieve anything more from that simply from the table. Without the original data regression analysis based solely on a few means is just silly.

    Like

  56. Ken I know people say I am trying to divert the subject when I bring in an analogy, but this is a bit like protecting a surface with coats of paint. By the time one gets to 20 coats the correlation is getting weaker between number of coats and protection of the surface. But for the first few it is quite high.

    There is a statistically significant difference between one and two coats but what about a correlation?

    So please don’t try to override the statistically significant fluoride effect just because a correlation is not supplied for that initial region.

    Like

  57. Brian, you are being silly. No one is attempting to “override” a significant difference between 2 regions in a measured parameter – that is just a fact of the data. The argument is about interpretation because the difference itself reveals nothing about the reasons for it.

    Your “regression analysis” is just silly, it adds nothing at all to the fact of the measured difference – after all it uses only those calculated means and does not use the data itself. And it is moronic to do a correlation with 2 or 3 data points and pretend it means something.

    Like

  58. So Ken do you disagree with Li et al’s interpretation?

    Like

  59. I guess you mean Lin et al (1991)?

    What specific interpretation are you asking about?

    If you mean the one usually used by Paul Connett (that it proves CWF would cause a lowering of IQ due to fluoide) no I do not agree.

    All that has been shown in the note is a statistically significant difference bweteen several regions – not the first to show that.

    But Xiang’s work showed urinary F explained only 3% of the variance in IQ – so it is far from a major factor – and once confounders are considered it probably wouldn’t explain any of the variance.

    It is simply straw clutching.

    Like

  60. soundhill,

    I take it that the silence about those questions I raised means – no answers.

    There’s plenty more questions that can be asked after a few minutes thought, all of which should have been asked, and answered before publication.

    Biggest question: since hypothyroidism is iodine related, why did they not discuss the factor of five between water levels of iodine?

    Which is more likely to have an effect – a difference of 5x between iodine intake (a proven relationship), or a 2x difference between fluoride intake (hypothesised relationship that has no control group)?

    Like

  61. Ken,

    At least on my tablet, the order of comments is getting somewhat confusing. Example: my comment at 6:59am is placed after yours at 10:16am. It happened once before, but I was overseas and hadn’t adjusted the time on the tablet. Today, I’m confused…

    Like

  62. Soundhill,

    I frequently use SI units of iu/ml. I know what SI units are, having used them for over 40 years. I know what the i stands for in the SI system.

    You are trying to explain errors that should have been caught by the authors, the editors, or the peer reviewers of the paper. You can’t, so how do we know the paper is accurate everywhere else?

    Like

  63. Addition to the comment above:

    iu/mL (the L didn’t capitalise – #@!* spellchecker).

    Like

  64. urinef mg/L2.56 1.61 1.6 1.34
    Urine I
    mcg/g creat
    97 96
    79 71
    64 79
    53 77

    with urinary iodine and fluoride as the independent variable axes, the IQ is on the array.

    Like

  65. soundhill,

    If group C had a water fluoride concentration of 6.0, what happens to your straw clutching?

    We can’t make any conclusions about fluoride in this paper because the authors (editors / peer reviewers) decided not to release the average water fluoride level for group C (nor its range).

    Perhaps you could ask them? Would you get a response? Would you tell us if it didn’t agree with your interpretation?

    Enquiring minds wish to know.

    Like

  66. Sorry the system removes spaces. Have to use dashes as place holders

    urine f mg/L 2.56 1.61 1.6 1.34
    Urine I
    mcg/g creat
    97—————————–96
    79—————71
    64———————-79
    53————————————77
    with urinary iodine and fluoride as the independent variable axes, the IQ is on the array.

    Like

  67. Stuartg wrote: “Which is more likely to have an effect – a difference of 5x between iodine intake (a proven relationship), or a 2x difference between fluoride intake (hypothesised relationship that has no control group)?”

    You have commented about that before. I’ll try another way to answer. I have made a 3D graph showing urinary iodide on the red horizontal axis, urinary fluoride on the green horizontal axis, and IQ on the blue vertical axis.
    I have made it a rotating video to visualise better.

    The 5x increase in iodine in the water does not have as great an effect on urinary iodide as does iodine supplementation, and that increases IQ.
    The blue ball at the top of the graph indicates normal fluoride in conjunction with supplemented iodine. It represents the value of highest IQ.

    The increase in urinary fluoride will show the result of increased fluoride intake. The higher value of it on the green horizontal axis is coupled with lowest IQ one of pink balls.

    Lower iodine plus lower fluoride gives intermediate IQ.

    My video graph portrays the values:
    Urine I———-urine f
    mcg/g creat mg/L——iQ
    79—————–2.56——71
    64—————–1.61——79
    53—————–1.34——77
    97—————–1.6——–96

    http://users.snap.net.nz/~bsandle/Iodine_Fluoride_IQ.MOV

    Like

  68. Stuartg wrote: “We can’t make any conclusions about fluoride in this paper because the authors (editors / peer reviewers) decided not to release the average water fluoride level for group C (nor its range).”

    We can presume something from the urinary levels.

    However for all groups except for the iodine-supplemented one we do not need to presume since those water levels are given and the associated parameters can be compared for those groups as a separate experiment.

    Ken wrote: “I guess you mean Lin et al (1991)?

    What specific interpretation are you asking about?”
    They said [with my addition in brackets]:

    “Statistically significant differences
    existed between these areas,
    suggesting that a low iodine
    intake coupled with high fluoride
    intake [I add that is in the non-iodine-supplemeted areas] exacerbates the ceptral [central]
    nervous lesions and
    the somatic developmental
    disturbance of iodine deficiency.”

    Like

  69. Ken wrote: “But Xiang’s work showed urinary F explained only 3% of the variance in IQ”

    Xiang’s iodine levels were Wamiao: 280.70 ± 87.16; Xinhuai: 300.96 ± 92.88.

    That is more than the sufficiency level for iodine suggested in this FAO report, so it would be expected for fluoride to have less effect.

    “Moulopoulos et al. (21) reported that a urinary iodine excretion between 151 and 200 µg/g creatinine (1.18-1.57 mmol/g creatinine), corresponding to a concentration of about 200 µg/l (1.57 mmol/l), gave the lowest values for serum TSH in a non-goitrous population” http://www.fao.org/docrep/004/y2809e/y2809e0i.htm

    Lin FF study had the higher fluoride problem at urinary iodine of 79 mcg/g creatine.

    Like

  70. Brain – your claim that “Lin FF study had the higher fluoride problem” is based on nothing more than a difference in parameters between tow areas and is not supported by regression analysis. That difference could have been caused by a whole host of factors (and the regression analysis suggests F is not involved).

    You have resorted to building a scenario based on minimal evidence in an attempt to defend Peckham’s obvious faults. He has made claims based on data analysis which specifically exclude iodine – something which the peer-reviewers must have pointed out (several subsequent articles have) and it surprises me that the paper could have been accepted in its current form.

    Like

  71. Ken wrote: “your claim that “Lin FF study had the higher fluoride problem” is based on nothing more than a difference in parameters between tow areas and is not supported by regression analysis.”

    If Lin FF et al made a mistake it could have been a result of failing to be able to use the old PC/XT computer to do at least linear-plateau regression.

    Like

  72. Ken, or segmented regression:

    https://en.wikipedia.org/wiki/Segmented_regression

    Quite a bit of my commenting has been aimed at suggesting two segments in the data, that is pre- and post-iodine sufficiency.

    Or there could be a gradual change of slope and even segmented linear regression would miss out and segmented non-linear regression might be in order.

    Like

  73. “Why is there such a high level of subclinical hypothyroidism in the included group? Non-random selection?”

    In this old publication from the days of widespread goitre in NZ: http://www.cawthron.org.nz/media_new/publications/pdf/2014_06/No_11_Goitre_in_the_Light_of_Recent_Research_.pdf (p14)

    ———————Water iodine——-goitre
    Otago—————-2.4 mcg/L——–19%
    Canterbury———-1.1 mcg/L———64%
    compare with China (Lin FF et al)
    low fluoride area—0.96 mcg/L—–82%
    higher fluoride——-5.21 mcg/L—–91%

    In 1929 when the NZ publication was published they would not have had the same sensitivity in detecting goitre.

    Infection may also swell the thyroid, as may the consumption of goitrogenic plants like brassica and soy. Such things may be looked into, though the Chinese tended to ferment their soybeans to lower antinutrients. (Fluffy tofu).

    I have emailed Sheila Sheaff to look into getting researchers to attend to plateau or segmented regression.

    Like

  74. It is axiomatic with data analysis that we should look first at the data itself and notice any peculiarities.

    Yet Brian says “too hell with the data, to hell with the real world” – he is going to apply “segmented regression.

    Bit difficult to do that without the data.

    But, again, he is straw clutching.

    Like

  75. Sheila has my sympathy.

    Like

  76. So, soundhill,

    You’re saying NZ goitre in the 1920s means subclinical hypothyroidism in China decades later?

    Words fail me.

    Like

  77. No Stuartg, both the Lin FF and the Hercus are talking of actual goitre.

    When the Lin FF uses the word “subclinical,” it is attached to cretinism.

    You said you do not like that word, “cretinism,” and you may be thinking it is being used as an alternative for the term, “hypothyroidism,” which you prefer. But no, it is being used in terms of when the hypothyroidism has affected mental development at some stage. Do you have another word for that?

    That is except when Lin FF et al are reporting others’ statements in the third sentence here::
    “ln this study, we found that 69% of the children with mental retardation
    had elevated TSH levels. IQ and TSH were negatively correlated.
    Many investigators regard an elevated TSH in the presence
    of normal T4 and T, levels as evidence for hypothyroidism
    that is subclinical but that can still affect the development
    of brain and cerebral function to some degree.”

    Like

  78. If I may rephrase something Ken said earlier:

    The understanding of when to use a statistical tool is a different, but equally important, skill when compared to knowing how to use that statistical tool.

    Just because a person is able to place figures into statistical analysis software, it doesn’t mean that they also know when it is appropriate to use the software.

    Like

  79. Stuartg, a statistical analysis gives probabilities. At p=0.05, normally accepted figure, if the experiment is done 20 times, one of the results is likely to show something which appears useful but is only a chance occurrence. So some people say you should not do correlating to look for relationships. They say only use correlating to test your hypothesis. But how does your mind get the hypothesis to test without some sort of correlating process going on in it? So why not help it along a bit, but being fully aware of the limitations?

    Then there are the things like which tests to use for parametric or non-parametric situations. Or like the point biserial when you have an “either-or” variable like sex, or maybe was the water fluoridated around a birth time. I’ve tried that for All Black captains. Seems to be a shortage from places like Auckland after fluoridation had been started when they were born.

    Then there is when something happens like the water rising in a tank until it is filled and only spills over after that. Then that can be called a plateau or segmented situation.

    A graph may help.

    With the Lin FF et al averages I did graphs for the quantities which rise and those which fall with TSH on the horizontal axis. Seems to be sort of plateau effects. Then there is analysis for that. See Wiki “segmented regression.”

    Ken seems to be delaying some of my posts, so you may not see them.

    Like

  80. Cretinism, hypothyroidism, subclinical hypothyroidism, goitre.

    I wonder why they have different names?

    Perhaps because they are different clinical entities?

    Research and clinical trials into one of the four cannot be directly applied to any of the others. Trying to do so will only result in errors and confusion.

    soundhill, when someone using the scientific method learns where and why their references are crap they immediately discard them. You are hanging on like grim death to your references no matter the evidence against them. That behaviour immediately tells others that you do not use the scientific method.

    Like

  81. soundhill,

    “So why not help it along a bit?”

    Because you’re not a statistical analyst.

    You have no idea when, or in what circumstances, the software should be used. Ken pointed out where your use was inappropriate. Instead of thanking him and discarding your errors you have clung to them and compounded them.

    To clarify: I’m not a statistician. I also have no idea when and in what circumstances to use statistical software. Because of that I have none on my computer. I know when something is out of my field and when to seek help from the professionals. Look up Dunning, Kruger.

    Like

  82. Stuartg how do scientists get the notion for a hypothesis to test?

    How did Ken get the notion to test for a correlation between ADHD and altitude? Did he help the process along by doing a correlation table first?

    Like

  83. Brian, I looked at the actual data – means would have been absolutely useless.

    This is your basic problem – you are trying to do statistical analysis without any data. That ios completely silly and a sign of extreme desperation.

    Like

  84. Ken
    Wiki: “The scientific method is an ongoing process, which usually begins with observations about the natural world. Human beings are naturally inquisitive, so they often come up with questions about things they see or hear and often develop ideas (hypotheses) about why things are the way they are.”
    So Ken you looked at the data and “came up with” with a question about whether it might be altitude. I am not having great success finding out about the process of “coming up with a question.” What aids are there to coming up with a question? One of them may be a sort of mental graph.

    I graphed some of Lin FF’s averages. Averages do mean something. Can’t looking at them be part of “coming up with a hypothesis”?

    A weather station might send average data every 30 mins. That might not be as good as every minute but still some things can be known.

    Like

  85. The fact is, Brian, I applied my statistical analysis to the data itself – not one or two means.

    You don’t have the original data to work with so all this playing around adds absolutely nothing. You can’t extract any more information than what Lin et al gave.

    Like

  86. Ken A: if the averages show segmentation then likely does the data.

    B: If the averages do not show segmentation that does not mean it is not present.

    I feel you are talking about B and I about A.

    I suggest the linear regression of all the data is not valid if segmentation is present. Some sort of average figure may be obtained, but it is better to look at each segment and get a figure for each. (the low and normal iodine levels)

    As regards my graph Openparachute may be showing an old cache version until you click on it.

    Like

  87. Brian, you are not able to “look at each segment or do any regressions” without the data.

    Like

  88. Ken, what I am suggesting is that Lin FF’s old PC/XT linear regression over their whole range was not capable of doing segmented analysis, but was helpful for their iodine deficiency part of their study. Separately they looked at the low iodine segment and gave fluoride findings and interpretation.

    Like

  89. You are suggesting this without any evidence, pretending you can read minds from the past, and purely to clutch at straws.

    You made a claim about the effect of F on I deficiency and this was the only citation you could find. No wonder you are desperate.

    Like

  90. Ken,
    http://www.nap.edu/catalog/11571/fluoride-in-drinking-water-a-scientific-review-of-epas-standards
    cites Lin FF and some others. pp262 263.

    I am not mind-reading Lin FF. Their work is in two parts.

    Like

  91. Brian – your claim “that Lin FF’s old PC/XT linear regression over their whole range was not capable of doing segmented analysis” is not factual – it is nothing more than pretending to read the minds of these authors from the past.

    Sure, as a review the NRC report mentions articles like Lin et al despite their unreliability. That does nothing to reinforce their claims. An intelligent reader always considers the evidence these claims are based on.

    Like

  92. Ken, Lin FF et al have not, in their correlations comments section acknowledged segmentation, though their averages show it, and the fact that they did comments on how the behaviour of fluoride at low iodide was different.

    Here is another translated study giving similar results, though its fluoride categories are more widely spaced.

    http://www.fluorideresearch.org/412/files/FJ2008_v41_n2_p156-160.pdf

    Like

  93. Combining the low water
    iodide parts of the studies of
    Fugui Hong et al
    and Lin Fa-Fu et al

    Water Fluoride-iodide–IQ
    —-mg/L—mcg/L
    Hong
    ———-2.94—–0.91—-68.3
    ———-0.48—–0.75—-75.3
    Lin FF
    ———-0.88—–5.21—- 71
    ———-0.34—–0.96—–78

    at these low water iodide
    values, which are relevant
    to New Zealand without
    supplementation, IQ seems
    to be further impaired by
    increased fluoride

    Like

  94. You may be able to gain from this,
    sometimes a 3D graph helps with visualization:
    IQ on vertical axis seems
    to be further impaired by
    increased fluoride on the
    green axis.
    Iodine, on the red, axis can
    only increase IQ if fluoride
    is very low.

    http://users.snap.net.nz/~bsandle/Lin_Hong_low_iodine_fluoride_IQ.mov

    Like

  95. soundhill,

    What was the fluoride water level in the control group? Why didn’t the authors publish it? Was the level higher than groups A and B? Or lower? Or the same? We don’t know because they didn’t publish it. Without that figure we can’t reach any conclusions about fluoride and you’re clutching at straws.

    What was the total population of areas A, B and C? What criteria did they use to select schoolchildren for further study? How representative were those schoolchildren of the rest of the population? Was this a highly selected group, or randomly selected? The authors don’t say. The authors found 104 children with mental retardation. That’s a very high proportion of the study group, which suggests the selection was not random.

    What was the TSH, T3, T4 levels in the rest of the population? Were the levels in the studied children the same as the rest of the population? Or different? The authors don’t say. Why not?

    The authors “studied a total of 769 schoolchildren”, 250 from area A, 256 from area B, and 243 from area C. All numbers are present in the body of the paper. Add them. There’s a discrepancy of 20. How many schoolchildren did they actually study? If the authors demonstrably can’t perform simple addition, how reliable are their statistics?

    “We examined 10-12 students…” Was it 10, 11, or 12 students? Weren’t they sure? Why couldn’t they have counted them rather than estimating? How good is their maths?

    These are simple questions about the science of the paper. Until they are answered the paper doesn’t actually tell us anything.

    Like

  96. Stuartg, you are talking about the Lin FF paper.

    As regards the fluoride in the control group, no real need to specify it since that part of the experiment was about iodide. They gave urinary fluoride so as to know nothing very out of the ordinary was happening with fluoride.

    As well as urinary fluoride, water fluoride was measured in the part of the experiment considering environmental fluoride and iodide, and the areas were compared against each other.

    In my first 3D video graph, using urinary fluoride, I was able to plot 4 points since the slightly different urinary figures were given for the two low iodide low fluoride areas. However Lin FF only gave one fluoride environmental level for those areas: the average of 0.34 mg/L.

    Fugui Hong’s experiment did not give urinary fluoride, so when comparing theirs and Lin’s results for iodide and fluoride I used just the two environmental/IQ figures from Lin, and joined them with Hong’s two to make my second 3D video graph.

    In the Lin fluoride/iodide study areas A and B are controlling against each other. They did have a “control area with iodine supplementation,” but I do not see that has anything to do with comparing the two low iodine groups against each other. It was to compare low iodine against adequate iodine.

    Like

  97. Stuartg : ““We examined 10-12 students…” Was it 10, 11, or 12 students? Weren’t they sure? Why couldn’t they have counted them rather than estimating? How good is their maths?”

    With 1632 students the groups cannot all be 10.

    Any Chinese translators reading?

    And some people may be quicker than I at getting through the co-citations &c.
    Appears to be the original:
    http://en.cnki.com.cn/Article_en/CJFDTotal-DFBT199102016.htm

    Like

  98. soundhill,

    Yes, I’m talking about the Lin paper.

    The one where they couldn’t add, casting doubts on the rest of their maths.

    The one where they publish incomplete information on one group (and you then tell us that information isn’t needed – how do you know without seeing it?)

    The one where they don’t tell us how they selected their study population.

    The one where they don’t tell us whether their study group was comparable to the rest of the population.

    The paper where many questions have been asked and the authors haven’t answered.

    The Royal Society motto is “Nullius in verba” – “On the word of no-one.” As long as (you think) a paper supports your ideas you accept it without criticism, the exact opposite of science. If you followed the scientific method, you would be questioning Lin et al. instead of the blind acceptance you demonstrate so well.

    Like

  99. soundhill,

    Until Lin et al tell us all their methods and results, and give us accurate maths, etc, you can talk until you are blue in the face. Your ideas remain unsupported by the paper and the paper does not reach reasonable standards for scientific acceptance.

    Like

  100. Stuartg For the low iodine results the Lin FF seems comparable to the Hong Fugui which you don’t seem to be complaining about.

    Like

  101. Ken wrote: “this one and a half page newsletter article.”

    That’s misleading use of language, Ken. It was *republished* in the “IDD Newsletter.” And it is closer to one and three quarter pages.

    I’m glad UNICEF did the project. So often the disadvantaged are just written off as a nuisance.

    Fluoridationists proclaim they are helping so why not go that little bit extra?

    Like

  102. Stuartg wrote: ““We examined 10-12 students…” Was it 10, 11, or 12 students? Weren’t they sure? Why couldn’t they have counted them rather than estimating? How good is their maths?”

    I’ll comment again, Stuartg.

    In each affected area they randomly selected a class and examined for endemic goiter, endemic fluorosis, body height, weight and IQ. They then selected 10 – 12 students from the class for further tests, hearing, bone age, thyroid iodine 131 uptake, thyroid hormone and the four psychomotilities. Those 10 – 12 students were selected from the class using stratified sampling, in other words making sure both sexes &c are represented in the 10 or 12.

    Like

  103. Lin et al was published in the Iodine Deficiency Disorder Newsletter. And really straw clutching again to extend from the general one and half page to one and three-quarters!

    Brian you are attempting to put lipstick on a pig.

    Like

  104. And really straw clutching again to extend from the general one and half page to one and three-quarters!

    almost one and three-quarters!

    Like

  105. Ken, that’s disappointing language.

    The Xinjiang, frequently Turkic Uyghur, people are suffering from suppression of their language, religion, and suffering from iodine deficiency which makes it hard to achieve in school because it spoils IQ.

    UNICEF has gone in to help them to try to lift themselves up.

    Parents had to pay to have children go to school, which may be changing now. But the class sizes were probably large, and frequently still religious at that stage.

    Probably a culture of economy has caused the authors to write a lot into a small space.

    Everything does not have to be spelled out because it can be seen.

    10 to 12 children were selected from each class to go through the complete set of tests. They totalled 130 so that makes about 12 classes in total or 4 from each region. The classes won’t all be the same size so the number of children selected from each for the complete tests will vary. (10 to 12). If it is a total of 749 children that may be 56 to 68 children per class since from 10 to 12 is 20% change. So in area A about 15 kids per class retarded. That’s to get a feeling for the scenario.

    Now how about asking something again, Stuartg.

    you wrote: “The one where they don’t tell us how they selected their study population.” But I think they did.

    “The one where they don’t tell us whether their study group was comparable to the rest of the population.”

    Randomly selected school classes. They haven’t afforded space to tell the reader it will only be from families who can afford to send the kids to school. Probably they will be above the total population of the regions’ average.

    “The one where they publish incomplete information on one group (and you then tell us that information isn’t needed – how do you know without seeing it?)”

    Are you speaking of the water fluoride in area C? Since the iodine was adequate there, knowing the water fluoride there is not necessary for their hypothesis that they are testing that fluoride is a problem in low iodide areas. What do you want to prove?

    Like

  106. Brain, my description of your tactics as attempting to put lipstick on a pig may be disappointing for you – but don’t be so hypocritical by pretending you have the interest of a particular group of people at heart.

    These people obviously suffer from a number of environmental, political and educational factors. So an ideologically-driven attempt to use them to support a poorly based belief in the evils of community water fluoridation is not only opportunistic – it shows you couldn’t give a stuff for the plight of these people. If you did you would sincerely support an approach based on good science which, after all, is the only way to alleviate some of these problems.

    You attempt to use a significant difference in a measurement between different regions as somehow proving your pet hangup when it does nothing of the sort. These differences could be caused by a whole range of factors and their causes just cannot be isolated by such superficial research. And it is inhumane to pretend they can as this diverts energy away from a more sensible approach which could genuinely help these people.

    Similarly, Peckham displays the same uncaring approach when he attempts to attribute hypothyroidism to CWF while purposely ignoring the known role of I deficiency.

    Neither you or Peckham are at all interested in solving these problems. You are only interested in pushing an ideological barrow. Such dogmatic ideology always leads to problems – not solutions.

    Like

  107. Ken here is another study which was going on at the same time as Lin FF.

    http://www.fluorideresearch.org/414/files/FJ2008_v41_n4_p319-320.pdf

    They do not give F or I readings just talk about high and low which would mean outside recommended levels I think. They are getting a similar result.

    I can remember the scar on my grandmother’s neck from an operation to reduce the size of her goiter. By the time my mother was growing up iodine was being added to salt so she did not suffer. The depression years reduced the amount of food people could eat, which may have been contributing to Grandmother’s goiter as she had to give her better food to her children.

    I wonder what would have happened had fluoridation been started before iodisation. Because now, without iodophors being used to wash milk machinery, and people being told salt is not good for them, and some people keeping a gluten free diet and not getting the iodised salt mandated in bread for the last few years, I believe it is only fair to correct iodine deficiency before fluoridating water. People need to know.

    Nor do we want excess iodine as the Hong study pointed out: fluoride also worsens the goitrogenous effect of excess iodine. So watch out if you are a seaweed eater.

    Peckham is losing a dimension by not including iodine.

    It is possible a number of studies presuming to be showing a risk of fluoridation could be explained by lack of iodine. The altitude relationship could be one.

    You seem to want to simplify out Lin’s iodine – fluoride effect by ignoring the toe of the curve – blurring the two parts of their experiment together. I don’t think that to be scientific. And I am always suspicious when writers start to use emotive language: I feel what you ciriticise me for is your own fault.

    Like

  108. soundhill,

    You “think” this, you “think” that…

    None of your speculation would be necessary if Lin et al had written the paper well, told us how and why they did things, gave us the full data, and ensured that their maths was accurate.

    You know, like scientists are supposed to do when they publish results of research.

    Like

  109. Brian, when will you learn that trolling through Connett’s “study tracker” is not research. Why not use the normal literature search tools? – Oh, I know, you prefer the preselection that Connett offers.

    This study suffers from the same deficiencies as Linet al and the others you produce. You cannot rely on a difference between 2 regions to determine a cause. Regression analysis is far more useful. Even then correlation is not causation and one must be careful to include confounding factors in the regressions. Peckham’s study is a prime example – only a fool would ignore I in such regressions, yet Peckham knowingly did. Its is stupid to describe that as “losing a dimension”.

    In these situations with endemic fluorosis and I deficiency rampant – not to mention educational, social and political factors, it is exceedingly naive to approach research in this way.

    Even more naive on your part to rely on the journals Fluoride which has done the trolling for you – found all the unreliable Chinese studies producing results they want – and then translated them for you.

    Yes, fools do make me angry when they behave the way you do. Suddenly you talk about the “toe of the curve” when you simply do not have a curve – how foolish can one get in attempting to distort and misrepresent the facts to fit a preconceived ideological position?

    I also get angry because discussion with you is such a waste of time with all the distortions and distractions you come up with. I do value an honest discussion of facts and ideas – but really I have far better things to do with my life that participate in such childishness as you promote.

    Like

  110. Stuartg it is always hard to write for people when you do not know how much they know about your environment.

    Lin et al gave the annual mean income of about 200 RMB per person.

    Then there is this which gives a Xinjiang example of 100 RMB per semester to attend primary school though it may have changed and will change.

    http://sites.miis.edu/comparativeeducation/files/2013/01/Education-Integration-and-the-Uyghurs.pdf
    I suggest again these kids would be healthier than the average population.

    If Lin et al were to have given the water fluoride level in region C how would that have affected the interpretation? It would be nice for people who want to look further but scientists would be cluttering up their papers so much if they put in everything they measured which was not related to their hypothesis. And they would be chided for trying to confuse.

    Like

  111. Ken I have not been using Connett’s stuff. I have been using the ordinary Google Scholar and looking at citations and now also cnki.net.

    Sometimes the “problem” jounrnals have made translations available. If you have an academic organisation address you might get a free trial on cnki and use google translate with caution.

    Now we have 3 studies pointing the same way. We must start to take note.

    Like

  112. soundhill,

    If Lin gave the water fluoride level of region C as 6.0 it would not make any difference to my interpretation: there’s too many other problems with the paper. It’s a crap paper because methods weren’t given, full data wasn’t given, comparison to population norms wasn’t given and they demonstrably couldn’t do their maths.

    I asked you the same question earlier. You declined to answer. I’ll give you a second chance. If Lin gave the water fluoride level of region C as 6.0, how would that affect your interpretation of the paper?

    I suspect that you won’t give an interpretable answer, because you have already demonstrated unquestioning acceptance of this and other cherry picked papers that you think support your ideas.

    Like

  113. soundhill,

    Actually, on thinking about things, there is an interpretation possible with this paper.

    Lin et al found differences between some of the children from three different areas of China.

    That’s not exactly an earth shattering relevation, is it? We can see differences between some of the children from different parts of Auckland.

    Any reasons for the differences? Well, for any hypothesis to be generated from this paper we would need to know their methods, data, population norms…

    Like

  114. soundhill,

    Scientists want to look further. They want the measurements. They want the methods. They want the controls. They want accurate maths.

    That’s not cluttering or confusion, that’s clarification.

    You don’t seem to understand that.

    Like

  115. Bazian (Independent critical appraisal of selected studies reporting an association between fluoride in drinking water and IQ) have done a very useful review of the papers usually flaunted by FAN to local councils regarding IQ and fluoride.

    Here is their summary of the Lin et al paper:

    “This study had a number of limitations:
    1: As with all observational studies, the results could be affected by confounding. Although the areas were reported to be similar in similar nationalities, habits, customs, and income, it is unclear whether a formal comparison of the areas was made. There may also have been other differences between the groups that could contribute to the differences seen.
    2: The study did not report whether IQ tests were performed blinded to which group the children were in. If assessors were not blinded the way in which they assessed the IQ tests
    may have been biased.
    3: As the study was cross sectional, it cannot establish the temporal sequence of events to support the claim that fluoride has directly affected IQ.
    4: Children’s fluoride exposures from their diet and the air were not assessed, so their overall exposures are unclear. However, the differences seen in urine levels of fluoride confirms that exposures did differ, at least at the time of measurement.
    5: The level of drinking water fluoride in the high fluoride area in this study (0.88mg/L) was lower than that proposed for use in Southampton (1mg/L). However, the children in the study may also have had considerable fluoride exposure from other sources, such as coal fires. In addition, the high fluoride levels were compounded by the presence of low iodine.
    As such, these results are not likely to be directly applicable to Southampton.”

    http://www.fairbanksalaska.us/wp-content/uploads/2011/07/20090211Bazian-Review-IQ-Studies.pdf

    Like

  116. Stuartg, Lin FF et al said of the iodine supplemented area: “Hetian where the nationalities, habits, customs, and income were basically similar.”

    It’s not like Auckland where pockets of people of differing skin colours will have differing vitamin D levels. Though there seems to be a few hundred meters difference in altitude, in these Chinese regions, which perhaps better be looked at for vitamin D purposes.

    The big differences are that Hetian was iodine-supplemented and that natural iodine and fluoride in water vary, and show up as varying fluorosis and goiter in the non-supplemented areas.

    As I have referenced, Lin’s study is paralleled by Hong’s which you haven’t criticised. Though in the Ren study though they haven’t tabulated the levels just said high or low which normally would mean outside recommended levels.

    Though each study may not be conclusive in itself in regards to the effect of fluoride on low iodine populations, when put together a trend is appearing.

    The Hong study reports of fluoride worsening both the low iodine and the high iodine goiter.

    Ken wants correlations to be attained in single studies. That may be easier in animal studies.

    A 29 Aug rat study by Jiang et al: “Effects of Excess Fluoride and Iodide on Thyroid Function and Morphology” is working with higher levels of fluoride than would be found in NZ fluoridated water, but over a period of 150 days without extra iodide, fluoride reduced total T4 and T3 though not TSH. And “iodide causes an increase in the areas of the colloid lumens and a decrease in the diameters of epithelial cells and nuclei; however, fluoride causes an increase in nuclear diameters.” http://link.springer.com/article/10.1007%2Fs12011-015-0479-0

    So a mechanism is starting to appear.

    As for Stuartg’s repeated question about 6ppm fluoride and Bazian’s suggestion of burning coal, it would seem the urinary fluoride levels give hints. About 50% of fluoride taken in appears in the urine of children, I think it is. Adults excrete more since their bones are already built.
    http://www.qucosa.de/fileadmin/data/qucosa/documents/13324/CRE2001035006451.pdf
    Bazian says: “However, the differences seen in urine levels of fluoride confirms that exposures did differ, at least at the time of measurement.” And the differences seem to have a fair relationship to the water level where it was given for three places.

    And Bazian seems to reinforce Lin et al’s hypothesis by saying: “the high fluoride levels were compounded by the presence of low iodine.”

    Like

  117. Ken “These differences could be caused by a whole range of factors and their causes just cannot be isolated by such superficial research. And it is inhumane to pretend they can as this diverts energy away from a more sensible approach which could genuinely help these people.”

    Though the authorities have thought iodine deficiency to be the problem and mandated to correct it. But now the worry is of excess iodine which also harms the thyroid, so the studies are tailored to that, and also find fluoride compounds that problem as well as compounding low iodine problem.

    Like

  118. Brian – you say “Ken wants correlations to be attained in single studies. That may be easier in animal studies.”

    Just not true. What I want is for people to read these studies critically and sensibly. It is not sensible to do a regression using 2 or 3 data points – you are trying to extract more confirmation bias from this poor quality study than is warranted.

    That said, in some of these studies (eg Xiang’s) regression analysis is possible – I am suspicious of the fact that Xiang did not include any other factor in his regression analysis or recognise the fact that he could only explain 3% of the variance using F.

    The Bazian analysis did not “reinforce Lin et al’s hypothesis by” – it pointed out that I is one of several confounding factors.

    Like

  119. Xiang was working without the confounding factor of low iodine which makes the distributions segmented.

    Lin et al, have a low iodine segment of their study and the limited graphs I have done of their work would seem to indicate Xiang’s 3% figure, based on linear regression, can not apply to that region, since the graph is not linear.

    Bazian said it was not necessary to worry about the confounding factor of low iodine in Britain. You have said something similar for NZ and I have disagreed for certain “idiosyncratic” diets where people cannot tolerate the current sources of iodine.

    Like

  120. Brian, you know nothing about distributions because you just don’t have the data. You are resorting to inventing it

    Xiang had all sort of confounding factors (and that is obvious from the very low R2 value for F) – he just didn’t include any in his regression – clearly he had a bias.

    You have absolutely no idea of wat explanatory power F may have for the Lin et al data because you don’t have the data.

    The Bazian company did not say “it was not necessary to worry about the confounding factor of low iodine in Britain.“

    Nor did I make the comment you attributed to me. Your comments are becoming 100% invention.

    Can you not see such dishonesty makes people angry – and justly so. J

    Like

  121. An interesting thing is that Bazian is in agreement with Peckham about not having to take low iodine into account in Britain. Though Peckham is not really saying that. He is just saying that if it is low for some people that group is likely to evenly distributed, isn’t he? That still does not rule out that the effect is due to them, does it?

    Like

  122. Ken: “The Bazian company did not say “it was not necessary to worry about the confounding factor of low iodine in Britain.“ ”

    Well its easy to make that interpretive mistake:

    Bazian “In addition, the high fluoride levels were compounded by the presence of low iodine.
    As such, these results are not likely to be directly applicable to Southampton.”

    I graphed Lin’s averages, if you are calling that invention. Usually averages mean a bit.

    Ken wrote: “The effect of iodine supplementation was clear, the effect of fluoride not so clear. But anyway, hardly a report to hang any conclusion on about CWF in New Zealand.”

    and I interpreted the second sentence there to be predicated by the first.

    Like

  123. Brian – you are lying again. The Bazian company’s review did nto even mention Peckham’s paper – how could they seeing it was published several years ago.

    Like

  124. It is easy “to make that interpretive mistake” when you are straw clutching to find anything to confirm you bias.

    AS for you quote of me –
    “Bazian said it was not necessary to worry about the confounding factor of low iodine in Britain. You have said something similar for NZ”

    This is completely dishonest. Both about the Bazian report and about me. I have clearly said that Peckham’s work is worthless without considering iodine confounding. And that comment has been made by many people – it is so obvious.

    Like

  125. So what else than adequate iodine, supposed lack of cretinism, distinguishes NZ that makes Lin’s work inappropriate for

    Like

  126. Ken : ” I have clearly said that Peckham’s work is worthless without considering iodine confounding. And that comment has been made by many people – it is so obvious.”

    Several people may have, but Bazian wrote off Lin’s work for Southampton on a basis Lin’s work involved the compounding low iodine problem.

    Like

  127. Ken: “Brian – you are lying again. The Bazian company’s review did nto even mention Peckham’s paper – how could they seeing it was published several years ago.”

    Scientists can get results “in agreement” independently without knowing each other, or even knowing that the other was working on a project.

    I had`written, “An interesting thing is that Bazian is in agreement with Peckham about not having to take low iodine into account in Britain.”

    Sorry I should perhaps have said, there is a sort of parallel between Bazian saying the Lin study, containing low-iodine compounding, and Peckham saying it is not needed to worry about iodine. But it was for two different reasons. Bazian implying Southapton does not have low iodine, but Peckham presumably meaning low iodine intake is reasonably evenly spread.

    Like

  128. Foley:
    “Peckham et al.’s hypothyroidism paper claims to have ‘…used a cross-sectional study design…’, but the study can be more correctly described as an ecological study. A cross-sectional study usually attempts to correlate exposure and outcome variables recorded from sampled individuals, whereas an ecological study derives data for either the exposure or outcome variable or both from population data, for example, censuses, government data or previous studies, rather than individual subjects. Attributing a causal association between exposure and outcome variables in ecological studies is difficult, as they are particularly susceptible to bias and confounding. The ecological fallacy, by which inferences about individuals can be incorrectly made based on findings from population data, is well known.”

    Wiki: ” concern about the ecological fallacy should not be used to disparage ecological studies.”

    “The term “ecological fallacy” means that the findings for the groups may not apply to individuals in the group. However, this term also applies to other observational studies and randomized controlled trials. All epidemiological studies include some people who have health outcomes related to the risk-modifying factors studied and some who do not. For example, genetic differences affect how people respond to pharmaceutical drugs. Thus, concern about the ecological fallacy should not be used to disparage ecological studies. The more important consideration is that ecological studies should include as many known risk-modifying factors for any outcome as possible, adding others if warranted. Then the results should be evaluated by other methods, using, for example, Hill’s criteria for causality in a biological system.”

    Like

  129. Stuartg: “The one where they couldn’t add, casting doubts on the rest of their maths.”

    Papers can be accepted in part.

    Here is a reviewer casting doubt on part of the Vanderpump study without disparaging the whole study:

    “(1) I don’t eat eggs
    (2) I eat eggs at least once per week
    (3) I eat eggs more than once per
    week
    Response (2), with the phrase “at
    least”, is non-specific and could range
    from “once” to “an infinite number
    of times” per week. Either response
    (2) or (3) could be given in several
    circumstances, showing that the
    question was not sensitive. Hence we
    urge caution over the conclusion that
    iodine intake is inversely related to egg
    consumption, particularly when other
    studies with more rigorous dietary
    methods have shown the converse.”

    http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736%2811%2961689-9.pdf

    Like

  130. “The more important consideration is that ecological studies should include as many known risk-modifying factors for any outcome as possible, adding others if warranted.”

    Like

  131. Vanderpump agrees with Sheila Sheaff’s comment about his article in terms of caution with urinary iodide readings. They can vary so much day to day, meal to meal in the same person that an average over a number of subjects has more meaning.

    And just brainstorming a bit about the amount of iodine excreted through bile and not reabsorbed, and any relation to fluoride. Maybe there is more than just dietary iodine level to the reported Birmingham (fluoridated) urinary iodine insufficency of girls, if that was what was being measured by Vanderpump.

    Like

  132. soundhill,

    You’re into the Gish gallop again.

    You were the person who brought up the Lin paper.

    It has been repeatedly pointed out to you where the paper fails and becomes irrelevant to your claims. Nothing you have said makes any difference, it’s still a crap paper.

    It has been repeatedly pointed out to you that your statistical “analysis” of Lin’s results cannot produce any more data (and we can’t believe Lin’s figures anyway, because the paper shows Lin et al can’t even add up).

    It’s been pointed out that you are making unwarranted assumptions about methods and results that cannot be justified – because Lin et al didn’t tell us what those methods and results were.

    It’s been pointed out that all Lin et al can really say is that some schoolchildren in one area are different from schoolchildren in other areas.

    It has been pointed out innumerable times, both here and elsewhere, that correlation does not mean causation. Perhaps you haven’t seen the “not” in that phrase, because you certainly appear to be correlating unrelated measurements and implying causation.

    Your initial defence of Lin et al references other papers. All have been analysed previously and shown to be of poor quality. None of them are directly related to the Lin paper; any relationships appear to be using the “six degrees” method.

    When directly asked about flaws in Lin et al, you have replied with “I think”, ” I assume”, and similar that demonstrate that the paper cannot be defended but you are not prepared to use the scientific method and look for better research. It gives the impression, true or not, that you are cherry picking and are hanging on to this paper because it agrees with your preconceived ideas. (“3 studies pointing the same way” – you are disregarding the hundreds, if not thousands, of papers that point other ways).

    Now, perhaps realising that Lin et al is indefencible, you have gone into the Gish gallop, referencing paper after paper. Unfortunately, this can only work in an oral discussion, where an opponent does not have the time to read those papers. Here we can have time to read those papers and realise exactly how bad they are and that they have nothing to do with the poor paper you are trying to defend.

    I am aware that no-one commenting here is going to change your ideas, simply because they are not science based. What I hope is that readers will recognise how and why the science contradicts your ideas, and the way your comments can be used to demonstrate the way science isn’t done.

    Like

  133. Stuartg: (“3 studies pointing the same way” – you are disregarding the hundreds, if not thousands, of papers that point other ways).

    Stuartg would you like to cite 3 which show that at low iodine and high iodine levels that fluoride improves thyroid function, then?

    Like

  134. Correction: Skeaff, not Sheaff:
    http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736%2811%2961691-7.pdf

    I note from another NZ commenter on that page:

    “Vanderpump and colleagues’
    paper illustrates a drawback of the
    prevailing paradigm that first insists
    on hard evidence of a problem
    rather than proactive prevention
    of its emergence in the first place. ”

    Science has to start with a notion, but mine about other mechanisms than low intake of iodine to bring about low urinary iodine may have put Stuartg into attack mode.

    But it is somewhat amusing he is not providing evidence.

    Like

  135. Brian, science is a very creative occupation. It thrives on ideas, no matter how silly. But for that reason the fact is most ideas in science are completely wrong. And we know that because when we do science we test ideas against reality.

    The refusal to test ideas against reality and yet impose them on society , as you are suggesting, is not scientific – it is harmful and inhumane.

    It is dishonest to call that proactive prevention.

    Like

  136. If fluoride affects deiodinase 1 in any way, then biliary excretion of iodine may change under fluoridation, invalidating to some degree urinary testing, and Vanderpumps results for Birmingham.

    “In normal rats, T3 glucuronide (T3G) is the major biliary T3 metabolite, but excretion of T3 sulfate (T3S) is greatly increased after inhibition of type I deiodinase, e.g. with 6-propyl-2-thiouracil (PTU). In this study, the fate of the T3 conjugates excreted with bile was studied to assess the significance of a putative enterohepatic circulation of T3 in rats.”

    http://press.endocrine.org/doi/abs/10.1210/endo-125-6-2822?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub=pubmed

    Like

  137. Ken: “The refusal to test ideas against reality and yet impose them on society , as you are suggesting, is not scientific – it is harmful and inhumane.”

    I am not doing that Ken, just wanting to be proactive about listening to small investigative voices even if they did a mistaken addition.

    To close ears to all but heavyweights seems cruel.

    Don’t miss any quirks when forming public policy, as this group purports to influence.

    Like

  138. I object to your use of the term “heavyweights“ as it is dishonestly emotive. You are dragging out extremely poor quality studies (which have been purposely mined by groups like the Connett crowd and translated for a specific political cause. They are “lightweight” because of their poor quality. The so-called “heavyweight” research is such because it is far better quality – not because it exerts unreasonable pressure.

    you are arguing for relying on extremely poor quality research fitting your bias and against good quality research which does not fit your bias.

    To close you ears to good quality research and take such extreme measures to find extremely poor quality research to promote is indeed inhumane and cruel.

    And you are performing gravity and reason-defying metal gymnastics rather than doing anything with your ears. Except close them to the reasoned critique we have made of the poor quality studies you are championing.

    Like

  139. Ken, anyone who does not accept all the little studies “events” for what they can give, may miss the greater picture, as in this study of the diffraction pattern of electrons with only few to start with:

    http://www.animations.physics.unsw.edu.au/jw/light/youngs-experiment-single-photons.html

    Like

  140. And it is not Paul, it is other Connetts studying biliary excretion of iodine.

    Like

  141. Brian, you seem oblivious to the fact that people here have “accepted” the studies you cite – in the sense of reading them and critically and intelligently analysing them. That is the scientific process of evaluation.

    You have “ace[ptedted” these studies uncritically with rose-tinted glasses and drawn conclusions from them which are not at all warranted. Extreme confirmation bias.

    Like

  142. I don’t think Paul Connett has ever studied anything, in the sense of honest research. Anyway, that is my conclusion from literature searching and the list of publication he supplied me with.

    Like

  143. The has Paul Connett ever actually referred to biliary excretion of iodine in your knowledge?

    Like

  144. Ken : “Brian, you seem oblivious to the fact that people here have “accepted” the studies you cite – in the sense of reading them and critically and intelligently analysing them. That is the scientific process of evaluation.

    You have “ace[ptedted” these studies uncritically with rose-tinted glasses and drawn conclusions from them which are not at all warranted. Extreme confirmation bias.”

    I am the one who has questioned whether linear analysis should have been used. I suggest those studying “uncritically” are just accepting unjustified linear analysis when the averages are not linearly distributed so the data won’t be.

    Like

  145. Brian, your so-called “questioning” is straw clutching. It is compoletley pointless becuase you do ntoi have the data to do any analysis on at all. And you are attempting to draw unwarranted conclusions from a few mean values to satisfy your bias.

    This is just putting lipstcik on a pig.

    Like

  146. In the Lin et al investigation the mean values for the primary school children who are otherwise said to be very similar societally, but have differing *fluoride* and *LOW IODIDE* environments, show differing thyroidal and IQ characteristics from one another.

    As part of an iodide supplementation investigation those differences between the low iodide groups were rather ignored and all those low iodide ones were compared to supplemented iodine children and of course the extra iodine was a much stronger factor in a correlation which was done for this wider range, and it overrides the effect of higher fluoride.

    Your fault, Ken, is to apply that correlation as if it could describe what was happening also within the low iodine segment, into which NZers have been moving again.

    The effect of fluoride on low iodine (and high iodine) groups has been replicated in two other studies I have cited, one of them not giving the actual I & F values, but the other one having similar low I and higher F values. You have not criticised those other studies, Ken and Stuartg.

    Stuartg, you have failed to cite any study in which an opposite effect shows, or no effect, under similar environmental low iodide conditions and variable fluoride. Therefore the way those studies together point stands for the mean time even if they have some weaknesses individually.

    Further, going on to other studies discussed, whereas some have claimed low iodide may be a cause of thyroid trouble, instead of fluoride, that is only a correlation. Fluoride may be causing the lower iodide by increasing biliary excretion of iodide via some mechanism. A mechanism like a part reduction in the effectiveness of normal deiodination, and unwanted resulting compounds being got rid of.

    Like

  147. I wrote: “Your fault, Ken, is to apply that correlation as if it could describe what was happening also within the low iodine segment,”

    As regards IQ over this wide iodine spectrum, a number of factors showed a correlation. But fluoride was not listed. Water fluoride level was not even stated by Lin et al for the supplemented iodine group. That result cannot be used to say zero or low correlation between water fluoride and IQ variation in the LOW IODINE range.

    Overall the situation would be more like, though probably a little stronger than, Xiang’s 3%, since Xiang et al’s groups were in a well supplemented iodine range, and they did not measure in a low iodide area. You seem to want to claim, Ken, that the Lin experiment’s main concentration on supplemented vs low iodide, and IQ, and the consequent lack of a fluoride correlation figure, that the lack of that figure overrides the significant differences found between the *LOW IODIDE* areas depending on fluoride. It is not valid to claim that.

    Like

  148. soundhill,

    Yet another painful sigh from me, especially as I can only connect every 48 hour or so.

    Correlation is not causation.

    You have 3 poor papers that suggest correlation, yet you ask me to provide 3 papers showing causation.

    You are saying this is causation, so it’s up to you to provide papers showing causation rather than correlation. You haven’t done it yet.

    Like

  149. Science works by critical analysis of published research.

    A paper gets even more criticism if it lacks requisite data (methods, controls, data, maths…) If a paper survives the criticism of other scientists, it is probably a good paper.

    Criticising a poor paper is not an attack, it’s just the way science works.

    Unfortunately, at least one commenter here doesn’t seem to be aware of how science works and interprets the scientific method as attacks.

    Like

  150. Stuartg, in the context I used the word “effect” is different from “cause and effect”. A common usage: “The Pearson product-moment correlation coefficient is measured on a standard scale — it can only range between -1.0 and +1.0. As such, we can interpret the correlation coefficient as representing an effect size.”

    But I don’t even ask for a correlation. Just something showing significant differences between LOW IODIDE areas sub areas of higher and lower fluoride. The 3 I cited give more thyroid trouble with higher fluoride in LOW IODIDE areas.

    You said many studies point other ways, so show me something that does.

    Like

  151. Stuartg. (On the newer thread I pointed to an article about what a mess science got into for Scott and scurvy.)

    Please give evidence for your comment: ” you are disregarding the hundreds, if not thousands, of papers that point other ways)” [than my three]

    in the context we were discussing, please.

    Like

  152. And Stuartg do you agree that thyroid metabolism might have anything to do with sports performance? Over the months I have posted on Openparachute about it. Now we have new All Blacks captain Kieran Read born in Drury which I think was unfluoridated. I have repeated my football stuff in three posts today on: https://www.facebook.com/Fluoride-Free-NZ-128729960477380/?fref=ts if you can access Facebook.
    Brian Sandle

    Like

  153. Whew. Fantastic to have reliable ‘net access again after so long without.

    soundhill, I’m not going to do your library searches for you. I’m just not interested enough to follow it further. I’ll trust the scientific consensus rather than rely on the three papers you think contradict the scientific consensus.

    You consider my comments to be attacks. Hardly. They are valid criticisms produced from basic questions. For example, let’s start from scratch and evaluate the basics in Lin et al.

    How many children were involved in the study? We don’t know because the figures given don’t add up. (Since their addition is demonstrably wrong, are there errors in the rest of their maths? It’s an unusual question to have to ask of a paper since most authors check their maths before publishing).

    How were the children selected for the study? We don’t know because they don’t tell us. (“We randomly selected a class of students in each affected area…” – each “class” being about 250 children? How did they do the randomisation? What about other classes, other schools? More detail is required.)

    Are the children representative of the populations? Or were they specially selected? We don’t know because they don’t tell us. (My own thoughts: if they are representative of the population then more than one in eight of the population is “mentally retarded”. That raises the possibility that the children were not representative of the population. I may be wrong, but we don’t know because they don’t tell us).

    What size were the populations that these children were selected from? We don’t know because they don’t tell us.

    How comparable are the populations in the three areas? How comparable are the three groups of children? We don’t know because they don’t tell us. The only information we are given is about the province as a whole.

    What is the range of water fluoride levels in the three areas? How representative was the intake of the children they selected? We don’t know because they don’t tell us.

    What is the range of water iodine concentration in the three areas? How representative was the intake of the selected children? We don’t know because they don’t tell us.

    How do they define high, normal and low water fluoride levels? We don’t know because they don’t tell us. Doubling the fluoride concentration from one area to another takes us from “normal” to “high”, but how far into the “high” category? We don’t know because they don’t tell us what they consider to be high, normal and low.

    How do they define high, normal and low water iodine levels? We don’t know because they don’t tell us. There’s a five times difference in iodine concentration between two areas, but they are both “low”. Why? We don’t know because they don’t tell us their definitions.

    How do they determine the incidence of goitre? We don’t know because they don’t tell us which definition of goitre they use.

    How do they define dental fluorosis to determine its incidence? There are multiple tools used to define dental fluorosis, so which did they use? We don’t know because they don’t tell us.

    All of the above are basic questions that I have been taught to ask. The answers to those questions, and others I haven’t raised, gives an idea of whether the findings of the paper can be generalised.

    The summary for Lin et al is that this is poor quality research that cannot be generalised. It does not contain the information necessary to duplicate the study anywhere else on the planet. It doesn’t even contain the information needed to replicate the study in the same areas to see if things have changed over time. With the lack of essential basic information, the comparative data table doesn’t actually mean much because we can’t directly relate it to anything else.

    Have you raised the same sort of questions with the other papers you have cited? Can they stand up to critical analysis?

    Valid criticism of a paper is not an attack, it’s just part of the process of science.

    Like

  154. Stuartg wrote:
    “soundhill, I’m not going to do your library searches for you. I’m just not interested enough to follow it further. I’ll trust the scientific consensus rather than rely on the three papers you think contradict the scientific consensus.”

    I did the literature search. As regards low iodide and fluoride I did not find an studies giving no result, or a help from fluoride. Only a hindrance by fluoride in low iodide areas. If you have seen hunders or thousands please try to remember one.

    “You consider my comments to be attacks.”

    No I just meant you suddenly started to write a lot of unrelated stuff trying to be persuasive when I got the brain storm of biliary excretion of iodide when fluoride is hampering the production of the desired iodine compounds.

    Hardly. They are valid criticisms produced from basic questions. For example, let’s start from scratch and evaluate the basics in Lin et al.
    How many children were involved in the study? We don’t know because the figures given don’t add up. (Since their addition is demonstrably wrong, are there errors in the rest of their maths?

    The translator or typesetter has written down 769 instead of 749. Also “central” was turned into “ceptral.” Fairly obvious errors, and I don’t think an academic would dismiss the whole paper because of them.

    “It’s an unusual question to have to ask of a paper since most authors check their maths before publishing).”

    Though it wasn’t the authors who published in English. It was a translator for IDD Newsletter which republised the article from a Chinese journal.

    “How were the children selected for the study? We don’t know because they don’t tell us. (“We randomly selected a class of students in each affected area…” – each “class” being about 250 children? How did they do the randomisation? What about other classes, other schools? More detail is required.)”

    They said the society is fairly uniform. Random means like pulling out of a hat.

    “Are the children representative of the populations? Or were they specially selected? We don’t know because they don’t tell us. (My own thoughts: if they are representative of the population then more than one in eight of the population is “mentally retarded”. That raises the possibility that the children were not representative of the population. I may be wrong, but we don’t know because they don’t tell us).”

    8%, or 1 in 12.5 were found to be retarded in the iodine supplemented area, compared to 1 in 6.25 in the low iodine normal fluoride area and 1 in 4 in the higher fluoride area (even though its iodide was possibly about half way up to the supplemented area).

    It’s the comparison that counts.

    Please note that the iodised salt or oil suburbs of Hotan (translated to Hetian by that translator) were only given that supplementation in 1982. So the sample of 7 to 14-year-olds in 1987-1989 when the study was done would still have had many born to mothers in a low iodine environment. The IQ could not have got to average of 100 in that time. Also note that in Timaru, South Canterbury, fluoridation stopped in January 1985. South Canterbury rugby had dropped from Division 1, Ranfurly Shield to Division 3 during fluoridation. 28 years after fluoridation stopped the team rose again to the Lochore Cup, though not as prestigious as the Ranfurly Shield. The team would once again be made of players born in non-fluoridated times.

    “What size were the populations that these children were selected from? We don’t know because they don’t tell us.”

    Doesn’t really matter, so long as things are randomised. Note that NZ study of IQ and Fluoridation, cited by the Royal Society, only had 99 control subjects and does not even specify where they lived, which wells they drank from, whether the watrer was treated to reduce acidty and so reduce possible lead in it, the way fluoridated water is.

    “How comparable are the populations in the three areas? How comparable are the three groups of children? We don’t know because they don’t tell us. The only information we are given is about the province as a whole.”

    It says: “In the suburbs of Hotan where the nationalities, habits, customs and income were basically similar [that must imply to the others]”

    You wrote: “What is the range of water fluoride levels in the three areas?”

    It was given for A and B areas where it was a variable of study. It was not a variable of study for area C.

    “How representative was the intake of the children they selected? We don’t know because they don’t tell us. “

    Similar habits.

    What is the range of water iodine concentration in the three areas?

    I admit they have not given the SD for each area, or in fact any Sds at all. But the range between the areas is very different.

    “How representative was the intake of the selected children? We don’t know because they don’t tell us.”

    Never can be very accurate with such studies, nor was the Dunedin study.

    How do they define high, normal and low water fluoride levels? We don’t know because they don’t tell us. Doubling the fluoride concentration from one area to another takes us from “normal” to “high”, but how far into the “high” category? We don’t know because they don’t tell us what they consider to be high, normal and low.
    “How do they define high, normal and low water iodine levels? We don’t know because they don’t tell us. There’s a five times difference in iodine concentration between two areas, but they are both “low”.

    The extremely low iodine area, though with low fluoride, did better than the moderately low iodine area with higher fluoride.

    Yes from extremely low to moderately low was 5 times. Still not as much urinary iodine as the supplemented ones. The urinary iodine is supposed to be an indicator of sufficiency.
    “Why? We don’t know because they don’t tell us their definitions.”

    Can get it from the UIC.

    “How do they determine the incidence of goitre? We don’t know because they don’t tell us which definition of goitre they use.”

    They had been doing other studies. We have a problem of only seeing limited translations. I’ll see if can find more.

    “How do they define dental fluorosis to determine its incidence? There are multiple tools used to define dental fluorosis, so which did they use? We don’t know because they don’t tell us.”

    Most studies use a similar test, don’t they? How many definitinos are there?

    “All of the above are basic questions that I have been taught to ask.”

    Though you do not bother to look much into the NZ study, persumably because it says what you want.

    “ The answers to those questions, and others I haven’t raised, gives an idea of whether the findings of the paper can be generalised.”

    “generalised” -by itself the paper may not be sufficient to make a policy for the general population, but when added to other studies it helps them, especially when no one is reporting contrary results.

    As I quoted from a NZ author we need to be proactive. That does not imply just waiting around shooting stuff down on shaky aspersions.

    “The summary for Lin et al is that this is poor quality research that cannot be generalised. It does not contain the information necessary to duplicate the study anywhere else on the planet. It doesn’t even contain the information needed to replicate the study in the same areas to see if things have changed over time.”

    I presume they are all being supplemented now, so couldn’t anyway.

    “ With the lack of essential basic information, the comparative data table doesn’t actually mean much because we can’t directly relate it to anything else.”

    The data table is from measurements of an approx 1 in 5 subset comparing very low,or low with supplemented iodine children.

    Lin Fa Fu added in the fluoride study in the very low and low iodine areas since they had already found such likelihood in their study:
    http://en.cnki.com.cn/Article_en/CJFDTOTAL-ZDFB198601019.htm

    “Have you raised the same sort of questions with the other papers you have cited? Can they stand up to critical analysis?
    Valid criticism of a paper is not an attack, it’s just part of the process of science.”

    Now how about giving a little thought to biliary excretion?

    Like

  155. Suartg: “How were the children selected for the study? We don’t know because they don’t tell us. (“We randomly selected a class of students in each affected area…” – each “class” being about 250 children? How did they do the randomisation? What about other classes, other schools? More detail is required.)”

    As I said before since there were groups of 10 to 12 making up 130 in the subset study there would have been about 12 classes or 4 in each of the 3 regions. You are confusing “classes” with “areas” or “regions.”

    Like

  156. soundhill,

    I can’t be bothered with this. This is the paper you selected. Now you excuse it as a limited translation. So why did you refer us to a limited translation in the first place?

    Ken has previously shown us what a difference the definition of fluorosis can make to a paper. That’s why we need to know which definition they use. Same applies for any medical condition.

    The paper uses average concentrations. That implies there was a range of concentrations within each area – what was that range? You’re confusing range and average.

    The authors found 104 “mentally retarded” children out of the 769 (or was it 749?) studied – that’s more than one in eight. Either that’s not a normal human population or the selection of children to study was non-random.

    You say “It’s the comparison that counts”. No, it’s the entirety of the paper that counts, not just the bits you choose.

    I’m not dismissing the paper. I never have. I just see many valid criticisms that mean it is not the devastating destructor of consensus you appear to think it is.

    You can carry on excusing the paper if you want, but facial cyanosis isn’t going to improve the poor quality of this paper that you chose as a reference.

    Like

  157. Stuartg: “The authors found 104 “mentally retarded” children out of the 769 (or was it 749?) studied – that’s more than one in eight. Either that’s not a normal human population or the selection of children to study was non-random.”

    “People living in areas affected by severe iodine deficiency may have an
    intelligence quotient (IQ) of up to 13.5 points below that of those from
    comparable communities in areas where there is no iodine deficiency ”
    from:
    World Health Organization, United Nations Children’s Fund, International
    Council for the Control of Iodine Deficiency Disorders (2007) Assessment of
    iodine deficiency disorders and monitoring their elimination: A guide for
    program managers. Geneva, Switzerland: World Health Organization.

    15.7% (1 in 6.4) of of a normal population have IQ below 85.That is beyond 1 SD. Take away 13.5 and get below 71.5. Lin FF et al are using IQ 50 – 69 as mild mental retardation. That must about match 1 in 8.

    Like

  158. soundhill,

    Research papers stand or fall on their own merits.

    Your excuses for Lin et al are now of greater length than the paper itself.

    Aren’t all these excuses your acknowledgement that the paper cannot stand by itself? 🤔

    Liked by 1 person

Leave a Reply: please be polite to other commenters & no ad hominems.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s