How Essential is Fluoride?

By Guy Armstrong

What Do The Experts Say?

And just as importantly What do they NOT say?

This book argues and demonstrates that experts have consistently exaggerated in their claim that fluorine should be considered an essential element for humans and suggests reasons why.

Expert focus on fluorine has detracted from the importance of elements and compounds that are definitely essential in the growth of teeth. In 2017 Guy Armstrong spent six months in the New Zealand Health Department archives looking through the fluoridation files of the 1950s. This has given his work a unique aspect which makes it very relevant to countries with F added to their public water supply.

A number of countries have been under the influence of American expertise and under the influence of the pervasive bias he shows. Any reader with a basic interest in science and health will be fascinated by this comprehensive account.

Media have consistently reported poorly on the fluoridation issue, partly because they did not want to be abused. This book looks at some very detailed aspects of expert and media interaction and behaviour, notably the presupposition of "objectivity" which allows the quoting of experts while disclaiming any responsibility to point out contradictions, and not ask uncomfortable or even obvious questions, when an evidence-based attitude would suggest this acceptable.

This book will help journalists, city councillors and the public understand not only the exaggeration of essentiality but the reasons behind it. The role of the sugar, aluminium and toothpaste industries are explored. The public relations industry has a presence here.

Contains six chapters, many sub-chapters, a detailed reference section, nine appendices, and an index.

• Language: English
• Publisher: The Nameless Publisher
• Release date: Apr 5, 2021
• ISBN: 9780473510190

Book preview

1. Is Fluorine/Fluoride an Essential Nutrient?

How do we test if something is an essential nutrient? Just what is an essential nutrient, anyway? This of course is a technical question, in that it means different things to different people, as you will see. I apologise if this is feels like an obtuse answer. Whether it should be something necessary or something beneficial is a cornerstone of this work. A far more important cornerstone of this work is how conclusions of necessity and benefit are treated in public, compared to how they are treated in scientific and more obscure and influential documents. Some definitions of common terms are given prior to Chapter 1 .

How do we test if something is an essential nutrient? This question has been addressed by three people I will mention here. Full documentation and citations are given in the text.

On the 14th of October, 2013, an expert panel from the University of Waikato appeared on a Google Hangout. This is an online group discussion filmed with interaction from the public, who could ask questions through social media. I don’t think it received much media attention; it may have been drowned out because this was near the time New Zealand’s city of Hamilton, Waikato, was embroiled in fluoridation arguments.

A viewer asked if fluorine was essential or beneficial. The experts discussed the question. Regarding specific criteria, Dr. Mucalo, Senior Lecturer in physical chemistry told viewers:

Yes, you need to design very careful experiments to specifically exclude fluoride from the diet before you could categorically prove whether it was or was not an essential element, and I think they’ve done that for things like silicon, where they’ve tried to prove whether it was essential for the diet but you’d have to do very excrutiatingly careful experiments to do that, so… at this stage I don’t know whether anyone’s done that yet. [1]

Many people have attempted such experiments.

Most of the experiments I have collected here have been cited in work like the American National Academy of Sciences and the World Health Organization, but also writings of people supportive of, and opposed to fluoridation.

In 2007, the National Academy of Sciences published a document which cited the 8th edition of a textbook called Essentials of Medical Geology, published in 2005. The sixth chapter is called Biological Functions of the Elements. It begins by looking at what constitutes an essential element. Author Ulf Lindh, Senior Researcher at the Biology Education Centre, Uppsala University in Sweden, says definitions have provoked much discussion, and that the earliest was borrowed from protein chemistry.

To paraphrase: the element should be present in living tissues in a reasonably constant concentration, it should cause problems – anomalies in several species when removed, and these anomalies should be corrected upon reinstatement of the element.

Lindh proposes the current (1998) definition:

An element is considered essential to an organism when reduction of its exposure below a certain limit results consistently in a reduction in a physiologically important function, or when the element is an integral part of an organic structure performing a vital function in the organism. [2]

Lindh notes problems associated with proving necessity and ascertaining exact requirements. Further into Chapter 1 we will see it is nearly impossible to completely remove fluorine from a diet. Reduction of only one element is difficult in food preparation - fluorine is apparently not the only element difficult to remove. Removal and reduction of one element may effect uptake of others leading to ambiguity of results.

Detection is also sometimes difficult. In terms of knowledge regarding necessary trace elements, we can be more certain about animal needs than human needs. Lindh doesn’t say it, but this is probably true for toxic effects and upper limitations of tolerance as well. In a section called The Functional Value of Trace Elements he writes:

The paramount function is to be necessary for the structure and function of significant biomolecules, mainly enzymes.

All of the experiments in the first chapter have tested fluorine’s essentiality on rodents. They have largely adhered to this criteria in terms of experimental design and intent, though some have used more or less imaginative, precise, and technological methods than others. This leads to the question of how well can these be applied to humans?

In Chapter 6 , I discuss the ‘for and against’ arguments that have been given with regard to using conclusions of rodent experiments on humans, and look at some human experiments. Relevant here may be a sentence from the World Health Organization’s 1970 monograph, Fluoride and Human Health:

Where fluoride data for man are unavailable, corollary studies on experimental animals are presented. [3]

Regarding experiments, I obtained almost every one I, or someone who had written on the subject, thought relevant. I also quoted from reviews that simply observed and critiqued experiments already in the literature. I feel we are quite lucky in this regard, because in cases of opposition or promotion, scientists seemed quite open about it, and none seemed set in stone regarding their conclusions, though a few were a little persistent. The reader will be pleased to note that there is almost no difference of opinion in terms of conclusion regarding individual experimentation in ‘pro’ or ‘anti’ literature with a couple of small exceptions, yet the overall conclusion – the ‘yes, no or maybe’ answer to the title of this chapter, is varied and nuanced, depending on the person to whom we go.

Different people have different criteria for what constitutes dietary essentiality. What I have quoted from Dr. Mucalo and Dr. Lindh is best termed traditional, maybe even standard – I will introduce other criteria throughout, notably prevention and benefit have both been suggested.

Read through these experiments slowly and patiently. Some have concluded fluorine to be essential, while others have not.

These are mainly primary sources, meaning original experiments, though there are a couple of exceptions. I have avoided reviews for the most part until the next chapter, although if they have comments neglected by others that help enlighten understanding of experiments, these have been included.

I have kept to a reasonably chronological order. When a sequence of conclusions has been argued over for many years, for the sake of convenience I have given them breathing room before going back to the main stream of research. Please consider some disagreements take a long time to work through, some are argued about then forgotten, and some criticisms may simply be ignored.

Usually the researchers create diets that are as low as possible in fluoride, and add differing levels of fluoride to the rats’ drinking water, or use foods with defined amounts.

Lindh’s and Mucalo’s discussions are very similar to what Drs. Messer, Armstrong and Singer from the University of Minneapolis wrote in a 1973 paper regarding two of their experiments that attempted the reduction of fluorine in the diets of mice.

A specific deficiency state should be produced by a diet lacking the element in question, but which is otherwise adequate and satisfactory.

The deficiency should be prevented or cured by addition to the diet of that element alone. [4]

This is what I will call Messer’s first criterion. I don’t know if the words originated in his mind, but I will attribute it to him because another scientist did so in a 1974 presentation. In the text I will refer to the set of four papers in the early 1970s by this group as being by either Messer et al. or Armstrong et al.

The statement quoted is from 1973. I will introduce many experiments before that year, but it should be understood that this criterion is what all of these experiments have aimed for, though they had different ways of doing so.

Because of nuances in methods, one cannot simply count the number of times a particular conclusion appears and claim one conclusion wins. This is shortsighted and presumptuous, as will be shown.

Obviously conclusive strength depends on a myriad of factors. I will discuss individual aims, methods, results and conclusions of each experiment. Chapter 1.2 will elaborate on some of these, and introduce more.

As far back as the 1800s researchers were considering the question of fluoride’s essentiality. Dr. Gerald Cox, writing in a 1952 National Academy of Sciences publication, pointed out that before the year 1933, accuracy regarding analysis of fluoride was poor. In his words little reliance can be placed on any quantitative analyses for fluorine reported [5] before Willard and Winter’s 1933 method of isolation by distillation. I point this out because I show a couple of experiments and papers before this date, and I do not want you to feel that the conclusions of an experiment done almost a century ago should be set in stone. Each experiment should be viewed in and of itself. Some yield plenty of information on their own, others more when compared with the rest.

1.1 What Do Scientists Who Performed Animal Experiments Say?

In 1933, the Journal of Nutrition published an experiment [6] by George Sharpless and E. V. McCollum from the School of Hygiene and Public Health, at the Johns Hopkins University of Baltimore. They stated at the beginning of their experiment that the question of fluoride’s essentiality was unanswered. The purpose of their experiment was to test if nutritional requirements can be satisfied with diets containing as little fluoride as possible. Ten rats, five of each gender, were fed a low fluoride ration, while another three of each gender were fed the same but with 0.001 percent (10 parts per million, abbreviated to ppm) fluoride added. Once per week iodine was added to (distilled) drinking water in order to avoid iodine deficiency. Rats were kept in a room where no roach powder was used, and when females became pregnant they were taken to individual cages, and given filter paper clippings (fluoride free) in which to nest. Offspring did not live long; Sharpless and McCollum attribute this to neither the absence or presence of fluoride, but noted that the results on reproduction seemed to favour low fluoride.

Sharpless and McCollum state their diet was

very low in fluorine but not quite free.

The rats on the low fluorine diet looked very well, were quite fat and on the whole appeared normal.

There were no notable differences between the control rats (10 ppm) and the rats on the low fluorine diet, except the rats on the low fluorine diet had hair that was a little coarser but this did not appear striking. Both groups of animals had incisors that were a deep orange.

The teeth showed no caries and from gross appearances seemed to be perfect.

Regarding a necessity for fluoride in teeth, Sharpless and McCollum claimed there was no detectable fluorine in the teeth of rats on the low fluorine ration.

"Thus, if fluorine is present as a factor in the consolidation of the tooth, abnormalities should appear in the histological ² picture. There is, however, no marked abnormality in the structure. The teeth seem to be excellent, no indication of caries, and seemingly perfect calcification."

The scientists claimed that fluorine could be removed from teeth until only between six and twenty-five ppm remained, with no defect, though of course perfect determination in the 1930s was difficult. There were no structural issues, nor were there any noticed changes in the calcium to phosphorous ratio in bone caused by the low fluorine diet. Quantities of fluorine were measured in the bones and teeth of the rats. The scientists wrote a lot on the preparation and accuracy of their method, it follows a colorimetric determination. Reported accuracy by duplicating determinations of values is shown (Table 1 ).

While they had a difficult time preventing the death of rats while young, the low fluoride groups reproduced more. They suggested some unexplained factor may have caused early deaths in young rats. Two tables from this experiment are reproduced (Table 2 & 3 ).

FLUORINE DETERMINATIONS TO SHOW ABILITY TO DUPLICATE VALUES

Table 1. Fluorine determinations to show ability to duplicate values. Source: Sharpless and McCollum, 1933 [6].

REPRODUCTION RECORD (TOTAL LITTERS)

Table 2. Reproduction Record (Total Litters). Source: Sharpless and McCollum, 1933 [6].

REPRODUCTION RECORD (FIRST LITTERS)

Table 3. Reproduction Record (First Litters). Source: Sharpless and McCollum, 1933 [6].

While some scientists have claimed that fluoride is a requirement for reproduction in the rat, Sharpless and McCollum concluded otherwise:

Rats grow normally on a diet low in fluorine. A diet low in fluorine does not affect reproduction in any way. [6]

Floyd DeEds was Senior Toxicologist at the Bureau of Chemistry and Soils, U.S Dept. of Agriculture, and lectured at the Department of Pharmacology at Stanford University School of Medicine in San Francisco. In 1933 the journal Medicine published his 60-page review [7] of the literature on fluoride from the previous fifty years. His review serves as perhaps the first of its kind. He looked at over one hundred experiments, that he noted had been largely uncorrelated. I believe he wanted to take stock of what did and did not need to be studied in future.

I include the review by DeEds here because it contains information that was relevant in the past. It may no longer be relevant to an appraisal of a requirement for fluoride in human nutrition, but it has certainly shaped the research and serves as an interesting benchmark of our understanding. While the title of the review suggests a toxicological appraisal, I will remind the reader that (obviously) anything in excess is almost always harmful, and keep strictly to the study of essentiality/non-essentiality, or what DeEds called a biological role of fluorine.

DeEds discussed work by a scientist named Gautier in 1914. Gautier proposed that fluorine’s role is in helping to assure the fixation of phosphorous in the organism and of the soft nitrogenous organic matter. Gautier looked at the ratio of fluorine to phosphorous in animal tissues used for ornamentation, protection and defense (hair, feathers, nails etc), found them to be close to the ratio existing in mineral fluoro-phosphates (about one part fluorine to five parts phosphorous). For less active tissues like bone, tendon and cartilage, Gautier suggested the ratio was one part fluorine to between one hundred and thirty and one hundred and eighty times its weight of phosphorous. I think Gautier believed that in the shedding of these tissues, unwanted or unnecessary fluorine could be removed from the body.

DeEds disagreed with Gautier’s conclusion, and mentioned the 1933 Sharpless and McCollum study, wherein the fluorine content of bones could

be reduced to between six and twenty-five ppm, and could be eliminated from the teeth, without showing any gross deleterious effect. [7]

He mentioned four scientists who had found fluoride in all tissues examined. This was around the end of the nineteenth century. He wrote:

… the constant occurrence of fluorine in tissues does not constitute proof that fluorine has a biological role. In view of the widespread distribution of fluorine in soils, waters, and plants, it is entirely possible that the presence of fluorine in animal tissues is an expression of the inevitable tendency to establish a chemical equilibrium between the organism and its environment.

Like others, DeEds believed one would have to minimize fluorine and create a deficiency in a population with definite symptoms of that deficiency, before claiming a necessity for the element.

Three scientists from the University of Wisconsin, two of whom were from the Department of Agricultural Chemistry and the third from the Department of Animal Husbandry performed an experiment five years in length culminating in late 1933 [8]. The experiment was published in 1934 in the Journal of Biological Chemistry. Doctors Phillips, Hart and Bohstedt wanted to determine if fluorine in mineral supplements for dairy cattle would increase fluorine content of milk to harmful levels.

The authors claimed the food used in the experiment contained ample amounts of other nutrients. They divided the cows into six groups, which were fed grain, corn silage, and hay, with differing ratios of bone meal and rock phosphate. Seven groups of rats (one control, then one for each lot of cows) were fed milk from cows within the six groups. Milks were mineralized with iron, copper and manganese. Rats were sacrificed after 140 days on the diet then fluorine body ash determined.

Phillips’ team noted that

our results are interesting in view of the concept that fluorine has an essential role in the animal body because small graded amounts of fluorine in the diet did not visibly increase growth in these rats.

They suggested that renal (meaning kidney) mechanisms influenced the ability of the rats to dispel an excess of fluorine, and while they encountered toxic levels of fluorine in the cows, they did not in the rats. This is evidence that mammalian milk is low in fluorine. The last sentence of their experiment is relevant to our purposes:

It is apparent from these data that fluorine in concentrations greater than 1 part in 10,000,000 in the diet has no essential function in the metabolism of the rat.

One part in ten million is 0.1 ppm. With regard to teeth, Phillips’ team wrote no evidence of typical fluorine toxicosis was apparent. By typical fluorine toxicosis I am assuming they meant dental fluorosis, because the next sentence reads the characteristic bleaching effect upon the incisor teeth was entirely absent…

Paul Phillips, a biochemist mentioned in the previous, 1934 experiment, and Robert Evans, a Moorman Manufacturing Company fellow, used milk mineralized with iron, copper and manganese in an experiment to test fluoride’s necessity in the skeleton, growth and reproduction of rats. Their experiment was published in 1939, also in the Journal of Nutrition [9].

One male and three females were used in each lot. There were eight different diets. The additions of aluminium and percomorph oil (a fish oil containing vitamins A and D) were to test if either of these substances in excess would inhibit the effects of the higher levels of fluoride. (Table 4 )

AVERAGE SKELETAL FLUORINE CONTENT (IN PPM) OF ADULT RATS

Table 4. Average skeletal fluorine content (in ppm) of adult rats. Source: R. J. Evans and P. H. Phillips, 1939 [9]. ‘mg/kilo/day’ refers to milligrams per kilogram of body weight, daily.

Evans and Phillips wrote that

In general the levels of fluorine administered had no beneficial or adverse effect upon reproduction.

However lots five, six and seven all failed to reproduce beyond the first generation. There was no real explanation given for this, although excess fluorine and aluminium seems probable at first glance. To quote:

Fluorine is not necessary for the rat in amounts larger than the 0.1 to 0.2 ppm found in milk… Food consumption records showed that only 0.05 to 0.06 mg. of fluorine per kilogram of body weight were ingested per day [basal diet]. It is plainly evident that in this species, which can on occasion withstand relatively large quantities of fluorine, very little fluorine is necessary.

They claimed that fifty micrograms of fluorine per kilogram of body weight per day is enough for growth and general wellbeing in rats. The rats weighed 40 gram each. Adding 0.1 to 20.0 ppm did not give any measurable improvement.

Phillips and Evans also claimed that bleaching of teeth occurred above an intake of 3.0 mg of fluorine per kilogram of body weight per day. Difficulties can arise in the appraisal of these measurements – mg of fluorine per kilogram of body weight per day is different to total daily intake, or mg of fluorine per day.

The work of the 1930s is mentioned in early American scientific literature, but not often discussed in detail beyond a simple acceptance of conclusion. It is also mentioned in some of the World Health Organizaton (WHO) documents discussed in Chapter 2.3 .

In 1945, H. H. Mitchell and Marjorie Edman from the Division of Animal Nutrition in the University of Illinois, Urbana, wrote a paper for the journal Soil Science that contained a section called Dispensability of Fluorine in Animal Nutrition [10]. They wrote,

Though various essential functions of fluorine in the animal body have been proposed, convincing evidence of their reality has not been forthcoming.

Attempts to raise laboratory animals (rats) on rations containing extremely low concentrations of fluorine have been uniformly successful.

They discussed an unpublished experiment by Lawrenz in which 14 pairs of albino rats were fed diets with 0.47 and 2.5 ppm fluorine for 207 days. These rats were produced by mothers that had lived on a low fluorine diet (exact amount unspecified, presumably the 0.47 ppm diet). These rats contained 0.8 ppm fluorine, compared with rats containing 5 to 7 ppm raised on a diet of natural foods. After 207 days there were no differences in growth rate, or dry weights of skeleton or teeth. Retained were averages of 0.210 mg and 1.648 mg for the low and high fluorine diets respectively (see Table 5 ).

RETENTION OF FLUORINE – UNPUBLISHED DATA FROM LAWRENZ

Table 5. Retention of fluorine – unpublished data from Lawrenz. Source: Mitchell and Edman, 1945 [10].

Mitchell and Edman wrote that the teeth of four rats on the basal diet (0.47 ppm fluorine) were examined by Dr. Isaac Schour, a dental histologist of Illinois University’s College of Dentistry. Though the teeth were not fixed in formalin as soon after death as they should have been for the most effective examination, Dr. Schour detected no abnormalities in them. No amount of time was given from death to fixation and there is no year mentioned for this experiment.

Mitchell and Edman said the rats on the low fluorine diet received a total of 0.88 milligrams of fluorine over 207 days, equivalent to 27 microgram per kilogram of body weight daily with no signs of malnutrition.

I haven’t seen this work discussed except for a couple of mentions.

None of the above experiments showed problems with a low fluoride diet.

However, a study that concluded fluorine to be essential appeared in the American Journal of Medical Science, in May, 1945 [11]. It was performed by Jesse Francis McClendon and Wm. C. Foster, from the Hahnemann Medical College Department of Physiology, and aided by a grant from ALCOA, the Aluminum Company of America. An abstract was published in The American Institute of Nutrition’s Federation Proceedings in 1944.

There are three experiments I have found in which McClendon concluded fluorine a dietary essential in rats. In all three his method was similar. He performed the experiments on a farm to minimize exposure of rats to fluoride particles from city air.

The experiment did not discuss previous experiments relating to the topic of fluorine’s essentiality at all. If he knew of these experiments, they are not mentioned in the work attributed to him here. In the 1953 and 1954 experiments (discussed below) he mostly cited his own work.

Regarding the 1944/1945 experiment, I have seen the abstract cited more frequently than the full experiment, which is only slightly longer.

The diet McClendon used was half corn, thirty percent glucose, and the rest beans, bean and corn leaves, salt, yeast and oils. In the 1945 experiment McClendon mentioned that the plants were grown hydroponically as a way to produce fluorine-free plants. No amount or concentration was given regarding the fluorine levels in these plants, or that of any other compound or element.

Two rats were put on a diet at the age of twenty-one days. Their mother’s diet contained 0.3 ppm fluoride (the 1939 experiment by Phillips and Evans [9] claimed fluoride is transferred through the placenta, therefore these rats would have retained some). One of the rats starved to death after forty-eight days - its teeth had decayed such that it could not chew. The other rat was saved from starvation with milk containing 1 microgram of fluorine, but died of starvation in 70 days.

McClendon wrote that

The crowns of the 12 molar teeth of each rat were practically removed by caries except that the caries of the 3rd molars (which are of little use in chewing) was less extensive. This is the most extensive caries seen in any rat on any diet for any number of days.

Also,

Fluorine is passed from mother to offspring and the less fluorine in the mother’s diet, the fewer young are born and the less is the milk supply of the mother. All these rats had extensive decay of the molar teeth; in some cases all of the crowns were lost.

The first statement is from his study in the Federation Proceedings and the second is from the American Journal of Medical Science. In the abstract McClendon compared these results of 12 carious teeth per rat in 48 and 70 days with 0.3 carious teeth per rat in 75 days on a diet

… which was apparently not superior except that it contained 0.3 ppm fluorine, and no caries in 75 days when 10 ppm fluorine was added to the drinking water.

His article ends with the statement:

fluorine is necessary in a diet that has to be chewed. [11]

His experiment used very small sample sizes (three sets of two rats), but results were consistent.

McClendon created a fluorine-free diet using water-culture crops designed specifically for studying deficiencies in animals. This is discussed in his 1953 experiment [12] with Jacob Gershon-Cohen, from the Albert Einstein Medical Center in Philadelphia. This experiment was published in the Journal of Agricultural and Food Chemistry. The two men went into a lot of detail with the preparations, their previous and following experiments did not focus so much on method. They used rain water (with 0.002 to 0.004 ppm fluorine) that ran off an aluminium roof at the experimental farm twenty-five miles from Philadelphia. The rain water was passed through ion exchange resins to remove halogens and sodium, before being used for the growth of plants in 200-litre tanks. No figure was given for fluorine in rain water after halogen removal, but one could assume extremely low. Nutrient solutions were added to this, daily additions of iron and manganese were given, and salts added once a month. They took frequent measurements of the solution in which the plants grew. They used salts of reagent purity.

The tanks were covered with aluminium foil, except for holes cut where plants could grow through. Of the health of the plants, the researchers wrote:

The best crop yields were obtained from sunflowers. Corn was not always well pollinated. Legumes produced lower yields and rice did not head in the short season. [12]

The rats used were litter-mates, 21 days old at the beginning of the experiment. Diets contained corn, sunflower seed and leaves, and ten percent each yeast and sucrose. One group of 19 rats were fed the food grown in the fluorine-free water-culture, with added distilled water.

The controls were either on the same diet with the missing elements added or on a diet of the same proportions made from the same varieties of plants but grown in soil.

Normally grown crops comprised the diet of the control rats, whose drinking water contained 20 ppm of fluorine. [12]

The words missing elements refer to either fluorine or iodine. The water-grown plants were deficient in iodine. This experiment [12] was also looking at need for iodine. In the fluorine-free group, obviously iodine was added. I am not reproducing the results of the iodine-related work here. A research grant was given from the Chilean Iodine Educational Bureau. There is no mention of research grants or funding from any other persons or organizations in this experiment.

McClendon and Gershon-Cohen wanted to use yeast with which to supplement rat food, but couldn’t find a halogen-free variety. Instead, a long-winded method of juicing halogen-free corn and sunflower stalks followed. The stalks and juice were boiled, with fructose and sucrose, ammonium nitrate, potassium dihydrogen phosphate, l-aspartic acid, and inositol added. When the mixture cooled, bacteria (Saccharomyces cerevisiae or Torulopsis utilis) were added. Calcium, boron, vitamins and transition metals were supplied by this mixture.

One may inquire what ratio of soil-grown to water-culture plants were used in the controls. This is not mentioned. The control rats were fed 20 ppm fluorine in water. The following pictures (Figure 1 ) are x-rays of one rat jaw from each group.

ROENTGENOGRAMS OF LOWER JAW OF RAT

Figure 1. Roentgenograms of lower jaw of rat. Upper: Healthy molars in control rat on water-grown and soil-grown (normal) diet containing 20 ppm of fluorine in water. Lower: Extensive caries of molars in litter mate rat in fluorine-free water-grown diet. McClendon and Gershon-Cohen, 1953 [12].

After two months on the experiment, soft tissues, teeth and bones were studied.

This experiment found a lot of difference between the growth of rats on control diets and fluorine-free diets, shown below (Table 6 ).

EFFECT OF FLUORINE-FREE WATER-GROWN DIET ON WEIGHT AND DENTAL CARIES IN RATS

Table 6. Effect of fluorine-free water-grown diet on weight and dental caries in rats. Source: McClendon, 1953 [12].

Discussing their results, McClendon and Gershon-Cohen in 1953 [12] said:

"A number of female rats fed a fluorine-free diet and mated to normal males did not produce viable offspring.

Therefore we suspect that fluorine is necessary in the diet of the rat."

In 1954, McClendon and Gershon-Cohen repeated their water-culture experiment, using 16 rats. Again, 21-day-old rats at weaning were used. The water-culture method was justified:

In attempts to extract halogens from natural foodstuffs, two difficulties arise: (1) complete extraction is impossible, and (2) when extraction is done, other essential nutrients are also removed. To obviate these difficulties, foodstuffs might be grown lacking these nutrients.

Of the diets, they wrote:

The water-cultured diets were made up of yellow corn, sunflower seeds, leaves and yeast, all grown in fluorine-free water, with additions of fluorine-free chemicals: glucose, corn oil, cod liver oil (containing 9 ppm iodine), and sodium chloride. [13]

The rats on the diets lower in fluorine did not grow well. They were given methionine, cystine and lysine when their growth was especially slow. The purest quality reagent calcium citrate was added to the others fed the low-fluorine food, though this did result in an extra 0.004 ppm fluorine to the diet.

They did not discuss the health or vitality of the plants in this experiment, therefore it is probably safe to assume not much had changed from the previous experiment published the year before. Results are shown in Table 7 .

DENTAL CARIES IN RATS FED VARYING CONCENTRATIONS OF FLUORINE IN FOOD AND WATER

Table 7. Dental caries in rats fed varying concentrations of fluorine in food and water. Source: McClendon and Gershon-Cohen, 1954 [13].

The scientists wrote:

Animals fed on water-cultured fluorine-free foods tended to die within two to three months and this premature death could be partially delayed by a feeding of 10 cc [cubic centimeter] milk containing 1 mg of fluorine. [13]

This experiment also made the statement that female rats fed fluorine-free diets did not produce viable offspring. From this, and for dental reasons, the scientists suspected fluorine is necessary in rat nutrition. McClendon said that no fluorine was found in the faeces or urine of fluorine-free rats, which was indicative of retention from parents.

The scientists noted that their experiments indicated that foods low in fluorine uniformly produced periodontoclasia.

Therefore we suspect that fluorine is necessary in the diet of the rat.

At no point did the scientists cite any of the previous experiments beyond their own, regarding fluorine as a necessity. This is true for all three of the McClendon experiments here. Regarding the length of this 1954 experiment, I believe it lasted for only one generation. They wrote:

If the element is transmitted from the mother through the placenta and milk, it would be desirable to continue the diet through two generations.

The teeth of our rats on a fluorine-free diet contained some fluorine inherited from the mother. Since the chemical tests showed a lower concentration of fluorine in the dentine of these animals, and since no excretion of fluorine was noted in the urine or feces, this small concentration of maternally derived fluorine seems to be fixed or retained.

Research grants were given for this 1954 experiment from Lever Brothers, a toothpaste manufacturer, and from the Aluminum Company of America (ALCOA). Some scientists have expressed ideas on these experiments, which are discussed in the next chapter.

Assistant Professor of Chemistry Joseph Muhler was a man who spent many years studying fluorides. He worked in the Department of Chemistry at Indiana University. Countless boxes of his research exist today in storage. His PhD thesis published in 1951 [14] has been cited in much of the literature. According to the New York Times (15):

He received his D.D.S. in 1948 from Indiana University, followed by a doctorate in chemistry in 1951, the year he joined the faculty as an assistant professor of chemistry. In 1972 he was named a research professor of dental science and director of the School of Dentistry's Dental Research Institute. [15]

In 1954, Muhler performed an experiment in which he attempted to create a diet completely free of fluoride [16]. It was published in America’s Journal of Nutrition. Muhler looked at weight, growth and reproduction, as well as fluoride levels in femurs of male and female rats.

He stated that the criterion for dietary use of fluorine is the uptake of fluorine by the skeleton. He used four diets with fluoride concentrations in ppm of 3.1, 2.2, 1.8, and less than 0.1. Muhler concentrated largely on storage of fluoride but pointed out very little real difference in rats fed different diets. One of his objectives, necessary for testing the essentiality of fluoride, was to make a diet completely free of the element, however this could not be done (obviously this was the reason McClendon grew his crops in water-culture). Muhler stated that

even with exceptionally careful purification the diet still contained traces. [16]

By ‘traces’ he meant less than 0.1 microgram per gram (0.1 ppm). He knew there were traces due to its presence in the skeleton. For this reason, Muhler wrote that this experiment could not conclude anything definite about fluoride’s essentiality.

The most common finding was related to the difficulty of obtaining young from rats receiving the highly purified diet. After many attempts to obtain second generation rats, a different diet had to be used. Attempts to determine if this was due to the highly purified nature of the diet or to the lack of fluorine were inconclusive…

In 1957, Dr. Richard Maurer and Professor Harry Day (a mentor of Joseph Mulher’s), from the Department of Chemistry, Indiana University in Bloomington, performed an experiment which they introduced with the recognition that while fluorine was of practical significance in the protection of teeth from decay, studies were inconclusive regarding its essentiality.

The experiment was supported by the Medical Research and Development Board, Office of the Surgeon General, Department of the Army and published in the Journal of Nutrition. It was a smaller part of a thesis submitted as part of Richard Maurer’s Doctor of Philosophy degree. Mentioning the conclusions of experiments (see above) done by Sharpless and McCollum (1933), Phillips, Hart and Bohstedt (1934), and Evans and Phillips (1939) [6, 8, 9], Maurer and Day concluded that under the conditions used in these experiments fluorine had not been proved necessary in nutrition.

They also noted that McClendon (1944) and McClendon and Gershon-Cohen (1953) [11, 12] had concluded that fluorine was essential, based on rats fed a diet of mainly corn and sunflowers grown in rain water between 0.002 and 0.004 ppm fluorine (although no data were given on the fluorine content of either the diet or the animals). They also mentioned McClendon and Gershon-Cohen’s 1954 conclusion that periodontoclasia was a symptom of fluorine deficiency [13].

Maurer and Day spend three pages discussing their own methods, which to some may seem overly precise. For example, each individual animal was housed in a stainless steel cage, fabricated without using any fluorine in the welding or soldering fluxes.

The diet used had casein, corn oil, salts and corn starch as its primary components. Supplements like vitamins were added.

After purification, the diet contained no more than 0.007 ppm fluorine. Maurer and Day wrote:

"Under the extremely rigorous conditions of this study fluorine was not found to have any influence on the growth and well-being of rats.

Thus it is justifiable to conclude that under some conditions fluorine may not have any value in nutrition or even in the maintenance of dental health." [17]

They commented on the animals with the low fluorine diet having sleek coats and appearing in good condition, indistinguishable from the animals that were given fluorine. Animal teeth were examined with a dental probe. They found no dental defects.

The diet was not conducive to impaction and it contained no fermentable sugar; thus the results indicate that fluorine is dispensable in the maintenance of sound teeth if cariogenic factors such as high concentrations of sugar are not operative. [17]

With regard to reproduction, this experiment found the element unnecessary. Four generations, 110 animals were raised. All females had a successful pregnancy, giving birth to live young.

The experiment ended with the conclusion that fluorine was not a dietary essential. Maurer and Day described their experimental conditions as rigorous. The final sentence of their experiment says fluorine’s value is:

… apparently limited to the promotion of resistance to dental caries.

This experiment did not find any dental difference between the two groups of rats.

In 1959, the research of R. E. Wuthier and Paul Phillips from the Department of Biochemistry at the University of Wisconsin was published in the April issue of the Journal of Nutrition [18]. They performed an experiment which was designed to examine the long-term effects of differing levels of food and water-borne fluoride on rats over a two-year period.

They noted that

no demonstration of absolute essentiality

had been made for fluoride. They cited experiments by Maurer and Day [17] from Indiana University, and Evans and Phillips [9] from the University of Wisconsin, but suggested more existed.

Funds for this experiment were supplied partially by the Research Committee of the Graduate School from funds supplied by the Wisconsin Alumni Research Foundation. Merck and Co. supplied vitamins (of which no details were given).

Wuthier and Phillips used seven groups of twenty-five litter-mates, fed a semi-natural cariogenic diet for a year. They planned to remove three rats from each group at 3, 6, 12 and 18 months, and the rest at two years for testing. They would look at fluoride femur levels, dental caries incidence and severity, wear of molars, general health, reproduction, growth rate, number and weight of litters and weanlings. Unfortunately, due to respiratory and glandular infections causing death among rats, the experiment could only last one year.

The main constituents of the diet fed to the rats were ground oat groats, soybean oil, sucrose and dry skim milk powder. Corn and halibut liver oils were added, as were calcium, salt, and iodine. The diet contained approximately 0.5 ppm fluoride.

Of the seven groups of twenty-five rats each, one was given the basal diet with distilled water, and had only the fluoride that occurred in the food (0.5 ppm). All rats had the same diet, although three groups (given distilled water) had differing levels of fluoride added to food – 1.2, 3.2 and 7.2 ppm. Two groups had 1.2 and 3.2 ppm added to distilled water instead of food, and one group had city water (1.0 ppm) instead of distilled.

The scientists wrote:

The severity of dental caries, molar wear, and periodontal effects increased progressively with time. Fluoridation of either water or food at these low levels did not alter the pattern of these changes.

No significant protective effect against dental caries was observed from any of these fluoride levels studied.

Growth rates, mature weights, reproduction, and lactation were all normal and reflected no effect due to fluoride.

After the fourth month the general health of this strain of rats was poor, but again, no effect attributable to fluoride was indicated. [18]

The only places I have ever seen this experiment mentioned are a 1974 National Academy of Sciences publication, Effects of Fluorides in Animals, and Eric Underwood’s textbook Trace Elements in Human and Animal Nutrition. These are discussed a little in Chapter 2 .

In 1963, five scientists from the Departments of Agricultural Biochemistry, Botany and Poultry Science at the University of Arizona, Tuscon, carried out an experiment with a goal similar to those mentioned here. They discussed previous attempts to show an essential role for fluoride. Their study [19] was abstracted in a 1963 volume of Federation Proceedings and published in full in the Journal of Experimental Biology and Medicine in 1964. It was supported partially by the USPHS (United States Public Health Service).

Doberenz and co-workers pointed to the minimal diets produced by Sharpless and McCollum (1933), Phillips, Hart and Bohstedt (1934), Evans and Phillips (1939), Lawrenz (1945) and Maurer and Day (1957) [6, 8, 9, 10, 17]. They cited five of the experiments mentioned here previously to conclude that

in no case has an indication of an essential function for fluoride been noted.

They did point out the 1953 experiment by McClendon and Gershon-Cohen [12] as an exception to this trend, mentioning the reduced growth rate and dental caries on the rats fed the fluorine-free diet. However, they also pointed to a lack of information regarding levels of fluorine in the tissues of the animals.

Doberenz and co-workers cultured soybean and grain sorghum in a greenhouse built for the purpose of minimizing fluorine in hydroponically grown crops. The greenhouse was fitted with

evaporative coolers with microfilters to minimize dust. [19]

Silica sand was used as potting medium, after treatment to remove fluorides. Water was double distilled, deionized, and purity was checked throughout.

Chemicals used contained less than 0.01 ppm fluoride. The nutrient solution contained less than 0.001 ppm fluoride. The details involved in minimizing the fluoride levels in the food ingredients were quite particular. The main constituent of the diet was soybeans (55%). The rest included grain sorghum (23%), sucrose (10%), and minor constituents vitamin mix, salts and corn oil. Sodium bromide was also added. Doberenz and coworkers said the diet

was supplemented with vitamins and minerals adequately to satisfy the requirements for these micronutrients.

They claimed the diet contained less than 0.005 ppm fluoride.

This was fed to the first group (no added fluoride), and also to the second group (2.0 ppm fluoride added). A third group was fed field-grown sorghum and soybean, which contained 2.67 ppm fluoride.

Nine rats were used in each group, in a ten-week experimental time. An average carcass fluoride analysis was 0.72 ppm (fresh weight). After the ten weeks, average tibiae content of the first group (less than 0.005 ppm) of fluoride was 2.92 ppm. The second and third groups had 34.63 ppm and 12.54 ppm fluoride in tibiae respectively.

While starting weights were near identical, the rats with the added sodium fluoride had final weights less than the others. Doberenz and co-workers wrote in their summary that the only significant differences in enzyme activity were an increase in serum isocitric dehydrogenase and a decrease of this enzyme in the liver. This work has been cited frequently in the literature of the United States National Academy of Sciences before 1990.

Reproduced in Table 8 are the starting and final weights, and the fluoride storage in tibiae.

MINIMAL FLUORIDE DIET EFFECT ON RAT GROWTH AND FLUORIDE BONE DEPOSITION

Table 8. Minimal fluoride diet effect on rat growth and fluoride bone deposition. Source: Doberenz et al., 1964 [19].

In 1972, the journal Bioinorganic Chemistry published an experiment by Dr. Klaus Schwarz and David Milne from the Veterans Administration Hospital in Long Beach, California.

Schwarz and Milne claimed that:

fluorine, supplied as potassium fluoride, is essential for optimal growth in rats which are fed highly purified amino acid diets and kept in trace element controlled isolators. [20]

A trace element controlled isolator is a plastic housing for animals which inhibits movement of breathable particles with a couple of filters.

Schwarz and Milne used

highly purified, chemically defined amino acid diets

that varied in fluoride levels, as low as below 0.04 ppm, however some control diets contained 0.21-0.46 ppm. Like other scientists, they found it very difficult to keep fluoride levels minimal. To these diets they added 1.0, 2.5, and 7.5 ppm fluorine.

The authors claimed that an added 1.0 ppm caused a 17% increase in average growth rate (17 rats used), 2.5 ppm caused a 31% increase in average growth rate (17 rats used), and 7.5 ppm caused a 28% increase in average growth rate (19 rats used).

Rats on a second basal diet with 2.5 ppm added grew 20% better than controls. They said:

Rats receiving fluoride were generally better developed, as evident from their dimensions and from x-ray pictures of their skeletons.

According to Schwarz and Milne, growth is the most sensitive indicator of an existing deficiency in a trace element isolator³ . However, other problems occurred in almost all animals, such as shaggy fur, loss of hair and seborrhoea (an over-activity of oil-producing glands on skin), which were apparently not affected by fluorine additions. It is not mentioned whether these things were caused by dietary or other factors.

They suggested a level of 1.5-2.5 ppm fluorine as essential for rats, and said the same is appropriate for humans. They also mentioned the Messer, Armstrong, and Singer experiment [28] (via personal communication, it had yet to be published) as another proof that mice fed diets with low levels of fluoride had impaired fertility, and that 50 ppm fluoride added to drinking water prevented small litter sizes.

Schwarz would state publicly that his work applied to humans (discussed in Chapter 5.2 ). In this experiment he claimed:

The metabolism of fluorine in mammals presents several features which support the concept that it is essential. [20]

This remark is made with citations to three textbooks – Fluorine Chemistry Volumes 3 and 4 edited by J. J. Simons (1965), Fluorides and Human Health by the World Health Organization (1970), and Pharmacology of Fluorides by Smith (1966).

Fluorine Chemistry Volume 3 by Harold C. Hodge, Frank Smith and Phillip S. Chen, focussed a lot on excess fluorine, with chapters on fluoroacetate, phosphofluoridates, and organic fluorine compounds [21]. I found nothing on the topic of a nutritional role.

Volume 4 by Harold C. Hodge and Frank Smith, published in 1965, contained some interesting observations, but concluded:

At present there is no generally accepted demonstration of the essentiality of fluorine in the diet of man or animals using growth or reproduction as indices. In the diet of civilized man, traces of fluoride (1 ppm fluoride in the drinking water) confer such notable improvement in tooth health that one is tempted to describe this role as essential. [22]

This book is discussed a little more in Chapter 1.2 .

The 1970 World Health Organization monograph is discussed in detail in Chapter 2.3 . The sections on essentiality in the second and sixth chapters of the monograph looked only at experiments on rodents and claimed more research was needed before a conclusion of essentiality for fluorine could be reached, though the introduction of the monograph, written by a different author and without recourse to experiments, claimed fluorine essential. Schwarz and Milne also cite page 183 of this 1970 document at the start of their experiment which is where the sixth chapter’s discussion on essentiality begins.

The 1966 textbook by Smith et al. is highly technical and goes into great detail regarding the chemistry of elemental fluorine [23]. Regarding essentiality, we are referred to work by Muhler [24] and Underwood [25] (both are discussed in the following chapter as neither are original experiments but summaries of others’ work), as well as the 1965 book edited by Simons [22].

The following is Frank A. Smith’s conclusion:

Suffice it to say that unequivocal evidence that fluoride is required for growth, reproduction, or skeletal development, has not yet been forthcoming. Unquestionably, fluoride does improve dental health and in its absence this condition is impaired; however, the animal does not die in the absence of this optimal dentition. On the basis of present knowledge, fluoride is classed as essential by Muhler and as probably essential by Underwood. [23]

Schwarz and Milne presented their findings on the 28th of December, 1971 at the 138th meeting of the American Association for the Advancement of Science in Philadelphia, and again on the 10th of April, 1972 at the International Atomic Energy Agency Symposium on Nuclear Activation Techniques in the Life Sciences in Bled, Yugoslavia.

On the 20th of April 1973, Dr. Schwarz presented his research at the 57th Annual Meeting of the Federation of American Societies for Experimental Biology in Atlantic City, New Jersey. He was the sole author of a paper on fluorine and other elements. He cited his previous work [20] as one demonstration of fluoride’s essentiality in the written version of this paper on fluorine and silicon, chromium, vanadium and tin, published in Federation Proceedings in 1974.

Again, Schwarz suggested a level of 1.5-2.5 ppm fluorine in water to yield optimal growth for rats, with varying levels for humans. Note that this is a concentration in water, not a dose or defined amount. He wrote:

Compared to the amount of fluorine found in the earth's crust, the dietary requirement for fluoride appears very small.

Rats in the trace element sterile isolator fail to develop normal incisor pigmentation. [26]

He believed that tooth pigmentation was improved by fluoride, if compromised by a lack of vitamins E, D, and A, or deficiencies in calcium, phosphorous, magnesium and iron. In 1973 two people wrote of Dr. Schwarz’s work. Nielsen and Sandstead, a Research Chemist and a Director from the United States Department of Agriculture’s Agricultural Research Service, presented a review on nickel, vanadium, silicon, fluorine and tin at the Annual Meeting of the Institute of Food Technologists in Miami. The paper was published in 1974, in The American Journal of Clinical Nutrition [27].

With regard to work on the dental and skeletal benefits of fluorine since the 1930s, they quoted Dr. Mark Hegsted:

If an essential element were defined as one which has a beneficial effect in health and well-being, under the usual conditions in which individuals live, then in the light of the above evidence, fluorine would be considered an essential element in human nutrition.

(Dr. Hegsted’s work is discussed in the next chapter.)

Nielsen and Sandstead also looked at the claim by Messer and Armstrong, that fluoride was required for growth and reproduction (see below). They didn’t go into details, but their conclusion indicates their disagreement:

At present, a requirement for fluorine cannot be estimated.

Their review also contains the criticism/caution of Schwarz’s experiments using the trace element isolators:

This observation must, until confirmed, be viewed with reservation for the following reasons: 1) The control rats experienced suboptimal growth, even though they were supplemented with fluoride. 2) Although significant, the differences in weight gain between the deficient and the control animals were small, approximately 6 g over 26 days, even though the diet contained all known essential elements including vanadium, silicon, and tin. 3) Others have not been able to confirm this finding even though they have fed diets containing less fluorine. Clearly more research will be necessary before it can be stated that fluorine is essential for growth. [27]

The article concluded in a summary by saying that fluorine and tin had not been shown essential for humans, although it seemed probable that they were needed somewhere in the human metabolism, and therefore in nutrition.

Here, a rat suffering suboptimal growth is compared with a normal rat. These photos (Figure 2 a and b) are taken from Schwarz’s 1974 paper. The sick rat is explained by Schwarz’s statement that

… chemicals needed in diets to produce trace element deficiencies… must be of a degree of purity which greatly exceeds that normally encountered in chemistry.

(a)

(b)

Figure 2. Both animals were on the same diet. The animal on the left (a) was typical of animals kept for 20 days in the trace element sterile environment system. The animal on the right (b) was an outside control, kept under conventional conditions on a purified diet. Schwarz, 1974 [26].

Dr. Schwarz pointed out in his second paper that previous researchers, notably Maurer and Day, and Doberenz (both of whose means he described as rather sophisticated), had found a much lower amount of fluorine necessary than what he had. Schwarz said the reason for this was not clear to him.

It is well known, however, that the fluoride requirement is influenced by the levels of calcium, phosphorus, magnesium, and various other constituents of the diet.

In our experiments, these were contained in adequate but balanced amounts in an amino acid diet which was completely defined. The fluoride supplement was added to the diet while most other investigators supplied it in the drinking water.

Schwarz believed that the absorption (he used the word ‘utilization’) of fluoride in water was different to that in foods; this is something believed today. Ions of fluorine floating freely in water are obviously absorbed in greater amounts than fluorine bound to food. He also claimed that silicates could interfere with fluoride use in the body, 50 mg/100 ml sodium metasilicate inhibited the effect on growth.

The level of fluoride required to produce growth is approximately 2.5 ppm. This amount is physiological; it occurs normally in foods and feeds, including average human diets. [26]

Dr. Schwarz did not go into detail in either of these papers about why he thought earlier work was useless – due to its inadequate methods. Experiments discussed here in 1939, 1945, 1954 and 1959 would perhaps demonstrate his discrepancy – they didn’t all use fluoride in water, they used it in food, and as Nielsen and Sandstead pointed out, at levels below what he used. Perhaps he considered McClendon and Gershon-Cohen’s approach extreme, but said nothing in the two papers mentioned here.

Schwarz did not elaborate on his claim that a need for fluorine was influencd by other minerals. His work made headlines (discussed in Chapter 5.2 ).

In September 1972, the journal Science published the work of Drs. Messer, Armstrong and Singer, working in the Biochemistry Department (Health Sciences) of the University of Minnesota, Minneapolis.

This is the first of four of their papers I will discuss here. As noted previously I will refer to these papers throughout this book as being authored by either Messer et al. or Armstrong et al. They introduced this experiment by drawing attention to the fact that

satisfactory evidence of a deficiency state with respect to fluorine has not been demonstrated despite several investigations with this purpose. [28]

In support of this statement, they cited experiments by Muhler (1954), Maurer and Day (1957), and Doberenz et al. (1963) [16, 17, 19].

Female mice were put in two groups and given deionized water (58 mice) or deionized water with 50 ppm fluoride as sodium fluoride added (55 mice). Both of these groups were fed a diet containing from 0.1 to 0.3 ppm fluoride. They were grouped four females to one male for 25 weeks. The dietary effects were exclusive to the females, the males were fed normal laboratory rodent food and swapped between the low and high fluoride animals. Armstrong and co-workers wrote:

Litter production was observed over the 25-week period to a maximum of four litters. Each litter was reduced to six pups and all pups were removed at 5 days of age to promote a rapid breeding rate.

Mice in the low fluoride group showed a progressive impairment in reproductive capacity.

All of the first generation animals gave birth. Below half the mice produced four litters. The second-generation mice on the low-fluoride diet suffered a

… progressive decline in litter production.

Twenty percent of this group did not produce even one litter, over half did not produce four litters. Regarding earlier experiments that concluded there was no need for fluoride in reproduction, Armstrong and co-workers wrote that

The failure of previous studies to demonstrate a role of fluoride in reproduction may be attributed to the small numbers of animals involved and the short duration of the studies, since in this work the infertility developed slowly in each generation. [28]

This is an inaccurate comment, as the Maurer and Day study, cited by Armstrong, Messer and Singer, went on for four generations. Armstrong et al. did not mention this. Maurer and Day observed most animals for 150 days, but kept two pairs, one with and one without fluorine, on their diet for 325 days [17].

The Doberenz study was ten weeks long [19].

Armstrong and his co-workers mated their mice at eight weeks.

The Wuthier experiment (no effect from low fluoride) lasted for a year, but Armstrong, Messer and Singer didn’t mention it. It was not a popular experiment. They also didn’t mention the five-year 1934 study, which claimed to have kept rats on the diet for 140 days [8]. This was mentioned in the 1957 experiment, but not acknowledged by Armstrong et al. This was possibly not considered long enough, which I think from their point of view, is justified.

They did point out that they had no definite basis for infertility, no physiological or molecular reason why low fluoride should cause restrictions in reproduction. They suggested that infertility was common in nutritional deficiencies, so infertility caused by low fluoride intake,

…may represent, at least in part, a nonspecific response to the stress of a nutritional deficiency. [28]

They also investigated if additions of fluoride could restore fertility to fertility-impaired mice. Female mice were kept on the low fluoride diet and mated after 8 weeks as before, and put into two groups. Half the animals were given added fluoride (50 ppm), and both groups were re-mated after one week away from males, then their litter production was monitored over the following 20 weeks.

Results were as expected: only 40 to 50 percent of mice on the low fluoride intake produced four litters, while 85 to 90 percent of mice on the high fluoride intake produced four litters. Armstrong and co-workers wrote:

This study demonstrates that fluorine satisfies the major criteria for an essential trace element: (i) A deficiency state, characterized by a delayed production of the first litter and a progressive infertility, has been produced in mice on a diet low in fluoride. (ii) The deficiency is prevented and cured by addition of fluoride alone to the diet. (iii) The deficiency correlates well with low tissue (bone) levels of fluoride. Thus, there is good evidence that fluorine is an essential element, at least in the diet of the mouse. [28]

In this experiment Armstrong and his co-workers wrote that after they had sent their experiment to be printed, they received