Is there enough fluoride in a tube of toothpaste to kill a small child?

Is the following statement supported by epidemiological evidence or toxicological data?

sodium fluoride is so toxic that there is enough in a tube of toothpaste to kill a small child

A recent news article (actually, a submitted opinion article, which is often treated like “news”) says “sodium fluoride is so toxic that there is enough in a tube of toothpaste to kill a small child”. Is this a valid statement of risk? How can we check to see if this is in the ballpark of making sense, and how can we compare this to other risks? We need a measure of the likelihood of harm, including the most severe harm that the statement claims. We can use the acute oral lethal dose (LD) measure that is used in any material safety data sheet, in agriculture and the workplace, and in any comparative consideration of basic toxicology.

A more detailed toxicity of a product or compound should include assessment of the mode of entry, form, availability to the body, digestion, chronic ingestion, metabolism, excretion, etc. (and these sometimes explain why mortality is a bit less than expected, but often show what an acute one-big-dose study could not). The acute response occurs during a short time after treatment (days rather than months), and the experiments are conducted in standardized ways. Typically, rats are fed or injected with doses from small to massive, and the resulting mortalities to the acute challenges of the doses are used to estimate how much will kill 50% of the treated animals, oral or injected. The LD is estimated experimentally (meaning with a plus-or-minus confidence interval), in units of mg per kg live body weight (mg per kg is parts per million, ppm). An LD50 of 300 for something would mean that 300 mg per kg (summing to about 20 grams for an adult) would give a 50% probability of death (and a seriously ill feeling to many of the survivors). The LD50 for caffeine it is 192 mg/kg; for nicotine it is 50 mg/kg; for sodium arsenite is 41 mg/kg; and for sodium cyanide it is 6 mg/kg.

An LD25, LD90, or LD95 could be calculated, but the LD50 is often used because it is in the middle and therefore more precise (small confidence interval), and still high enough to be a cause for concern. One could argue that a lower probability should be the focus. The mortality response is non-linear, and almost always S-shaped. Picture the graph, with dose received along the x-axis, and probability of death along the y-axis, from 0 to 100%. Little mortality often occurs at low doses, but closer to the LD50, the probability of death rises more rapidly. Therefore, the LD25 is often close to LD50, and not halfway from 0 to 50. This is another reason why the LD50 is a good comparative indicator.

First: I am not an MD, nor do I have a graduate degree in toxicology (which would be more to the point than being an MD, since they often have only basic toxicology), although I have completed courses in toxicology (such as Advanced Principles of Toxicology, University of Guelph, and various others). I have applied it in environmental research, and teach data analysis in environmental sciences. I tend to think of my training in toxicology to be less than I would want (as in most topics), but it is enough for this. Therefore, this is not medical advice. It is meant to be a helpful examination of the probable relationship of the statement from the article, with reference to the numbers and to reality. This is the main question in any consideration of truth: does the statement “enough in a tube of toothpaste to kill a small child” fit the facts as represented by previous and current evidence.

First we should consider, is there a warning that says seek medical help if a large amount is swallowed. Yes, but this is more common sense than an indicator of high risk, and the same warning is often on vitamin bottles and supplements, and quite rightly. This is a fair warning but gives no way to estimate and compare risk. Next, let’s consider, are we aware of cases of children who have eaten a tube of fluoride toothpaste and died? Not that I can find, but this is not strong evidence against the statement in the article; it only means that it is not commonly occurring, for whatever reason. We want to see if the reason they are not dying from fluoride toothpaste is that the amount of fluoride is much too low for this result, given its toxicity. Let’s first check to see how likely a fatality would be, based on the contents of a tube.

I checked my kids’ current (open) tube of fluoride toothpaste. It has 0.115% fluoride w/w, which means weight for weight, g per g. We have three brands of fluoride-containing toothpaste (Colgate, Crest, and Pronamel), and I checked them all. I worked out the active ingredient rate from the actual contents of the two larger tubes. One brand is sodium fluoride and the other is sodium monofluorophosphate. Another one has small print in black on dark blue and I can’t make it out unless I get a magnifying glass – in that case, I would recommend that their license be revoked, no matter whether it is toxic or not… pet peeve.

The main brand we have says 0.115% w/w, and the tube weighs contains 18 mL. I am trying to give the author’s argument the benefit of the doubt, so let’s use the biggest tube I have (which aligns well with the big tubes on the drug store shelf). The label states fluoride 0.15% w/v (meaning g per mL), and the tube weighs 130 g. By the way, why are some labelled 0.115% w/w and others 0.15% w/v? It is because toothpaste sinks in water, and the specific gravity is 0.150/0.115 = 1.30. This is why the 100-mL tube of toothpaste is labelled 130 g. So these may look like different rates, but all of the brands I have looked at are 0.15% fluoride, g per ml. (I don’t think they should be allowed to mix these units and ways of showing the contents, but that is another issue.) So, 130 g at 0.115% g per g, = 0.1495 g fluoride in a full tube. This is 149.5 mg. (Remember that 1% would mean 1.3 g, or 1300 mg, and you will see that this is about right.) The LD50 (mammalian, acute) derived from controlled and standardized experiments ranges from 52 to 200. Let’s use 52, because it indicates the greatest estimate of toxicity in the range. My 7-year-olds weigh 23 and 28 kg. Let’s choose the lighter one, to be conservative, and lean toward trying to prove greater toxicity.

Before I work out the probability of death of one my kids, should they consume a tube, I want to first point out that an LD50 of 52 ppm (i.e., mg per kg body weight in order to reach a probability of mortality of 50%) does indeed indicate significant toxicity. The statement that fluoride is toxic is correct, they do have a point, but it is the amount that matters. That is what the anti-fluoride activists continually overlook. As the pioneering toxicologist Paracelsus showed in the 1500’s, everything is toxic if you take enough (by the way, his actual name was Auroleus Phillipus Theostratus Bombastus von Hohenheim, and even that seems a bit much). In perspective, many things around a home are in the range of an LD50 of a few hundred and up, and I would consider them somewhat toxic. Those that are 50 or lower should be used only in small amounts, for short periods of time. For example, 4-phenylcyclohexene from carpet backing (the new carpet smell) has LD50 of about 60 ppm, and the smell can pervade a home (I would avoid breathing it all night). Some candles release burning lead into the air (not a good idea). Para-dichlorobenzene (carpets and moth balls) has LD50 around 500, so less toxic, but still smelly. As I noted, the toxicity of nicotine is 50 ppm, and most people barely mention it.

So what does the 149.5 mg fluoride in a large (130 g) tube of toothpaste mean for a 23-kg child? If the LD50 is 52, then a 50% change of fatality, based on tests with mammals that have been shown to be similar in response, would be 23 X 52 = 1196 mg, exactly 8 tubes of toothpaste. In addition, 8 large tubes would be difficult to eat in the first place, and probably impossible to keep down. How “small” would a child have to be to have a good chance of being killed by eating one large (103 g) tube of fluoride tooth paste? About 3 kg. It is likely that the other components of the 130 g of toothpaste (beyond just the fluoride) would have adverse health effects on a baby. I have been conservative in choosing the values above (meaning, leaning toward evidence that would accept the claim of severe toxicity, and doing the opposite of stacking it in support of fluoride safety). To be totally sure, we could even double the apparent effect. That would mean 4 large tubes would have to be consumed during a short time and kept down, still very unlikely. However, the possibility of some strange sense of taste causing an unusual child to try to eat tube after tube after tube of toothpaste, perhaps trapped over the weekend in a drug store that has no candy on the shelf, could be a special but rare problem. A few children have died after consuming large quantities of table salt, for example. Many have died from sitting in running cars, normally not a problem because of low concentration of CO. It seems that it is enough to say that large supplies of many tubes of toothpaste should not be available to children (and perhaps the other components of toothpaste would be an even greater risk). Everything should be out of reach, unless it is needed in small amounts, including toothpaste, and especially cleaners and solvents.

Conclusion

The statement that “there is enough in a tube of toothpaste to kill a small child” is not supported by either the evidence or by the prediction based on toxicology. There could be enough to cause staining of teeth if a whole tube were eaten, but mortality risk seems low compared to other things. The claim in the article seems to be nonsense, but I would not say “total nonsense”, or as the British class it “complete bollocks”, which are terms I would reserve for claims in homeopathy or astrology, or for an example in toxicology, someone claiming that handling paper is likely to kill, or claiming that a battery pack and wire will remove a range of undiagnosed parasites. The article’s author is correct that fluoride has some significant toxicity; he is just confused regarding the amounts (plus he raves somewhat about chemical weapons turned on the public).

The last point to consider is, if the statement by the anti-fluoride activist(s) were true, then great care would be needed for many sources that they are apparently not considering. Fluoridated water has a concentration of 0.7 mg per L (0.7 ppm, because 1 L of water weighs 1000 g). What about food and drinks? I sorted the USDA data on fluoride in beverages and food, by ppm fluoride. Here is the list of the entries that exceed the 0.7 ppm that is in fluoridated water. Tea was a big surprise to me. Raisins, seafood, and certain juices are all multiple times higher than fluoridated water, although still not a source of risk relative to other things in everyday life.

Compared to fluoridated water, 0.70

Tea, instant, powder, with lemon and sugar

5.84

Tea, black, brewed, regular, Mid-West

3.93

Tea, black, brewed, regular, South

3.81

Tea, black, brewed, regular, all

3.73

Tea, black, brewed, regular, Northeast

3.57

Tea, black, brewed, regular, West

3.55

Tea, black, brewed, microwave, South

3.38

Tea, instant, powder, unsweetened, prepared with tap water

3.35

Tea, black, brewed, microwave, all

3.22

Tea, black, brewed, microwave, Mid-West

3.19

Tea, black, brewed, microwave, West

3.10

Tea, black, brewed, microwave, Northeast

3.09

Tea, black, decaffeinated, brewed, Mid-West

2.93

Tea, black, decaffeinated, brewed, Northeast

2.79

Tea, green, decaffeinated, brewed

2.72

Tea, black, decaffeinated, brewed, all

2.69

Tea, black, decaffeinated, brewed, South

2.64

Tea, black, decaffeinated, brewed, West

2.47

Raisins

2.34

Juice, grape, white

2.13

Crab, canned

2.10

Alcoholic beverage, wine, white

2.02

Shrimp, canned

2.01

Shrimp, fried

1.66

Fish sticks, baked

1.34

Soup, corn chowder

1.32

Tea, instant, powder, with lemon and sugar, prepared with tap water

1.27

Coffee, substitute, cereal grain beverage, prepared with water

1.25

Tea, iced, ARIZONA, ready-to-drink

1.23

French fries, McDONALD’S

1.15

Tea, green, brewed

1.15

Cream substitute, powdered

1.12

Fruit juice drink, apple, ready-to-drink

1.09

Juice blend (apple and grape), JUICY JUICE grape, ready-to-drink

1.07

Potato chip, baked

1.06

Carbonated, water, fruit-flavored

1.05

Alcoholic beverage, wine, red

1.05

Waters, tap, Mid-West, municipal $

0.99

Gravy, beef

0.99

Waters, tap, South, municipal $

0.93

Coffee, brewed

0.91

Carbonated, grape soda

0.91

Tea, herbal, peppermint, brewed

0.90

Tea, iced, COOL NESTEA Natural Lemon, ready-to-drink

0.90

Frozen novelties, ice type, sugar free, all flavors

0.89

Waters, tap, Mid-West, all (includes municipal and well)

0.88

Juice blend (apple, grape, and other juices), JUICY JUICE (berry, cherry, or punch), ready-to-drink

0.88

Cranberry-apple juice drink, ready-to-drink and prepared concentrate, ready-to-drink

0.87

Chocolate-flavor beverage, YOO-HOO, ready-to-drink

0.87

Soup, minestrone

0.86

Carbonated, orange soda

0.84

Carbonated, tonic water

0.84

Mashed potato and gravy

0.84

Waters, tap, all regions, municipal $

0.81

Carbonated, cola, fast food type, diet, without ice

0.78

Water, bottled, DANNON FLUORIDE TO GO

0.78

Frozen novelties, juice type

0.77

Juice, grape, WELCH’S, ready-to-drink

0.77

Waters, tap, South, all (includes municipal and well)

0.76

Soup, pea

0.76

Fruit drink, CAPRI-SUN, ready-to-drink

0.74

Waters, tap, Northeast, municipal $

0.74

Frozen novelties, ice type, regular, all flavors

0.74

Bread pudding

0.74

Carbonated, cola, fast food type, without ice

0.74

Carbonated, root beer

0.73

Jellies

0.73

Cranberry-raspberry juice drink, ready-to-drink and prepared concentrate, ready-to-drink

0.73

Carbonated, cola, DIET COKE, South

0.72

Tea, iced, LIPTON BRISK Lemon, ready-to-drink

0.72

Oatmeal, cooked

0.72

Fruit flavored drink, SUNNY DELIGHT, ready-to-drink

0.71

Waters, tap, all regions, all (includes municipal and well)

0.71

Cranberry juice cocktail, ready-to-drink and prepared concentrate, ready-to-drink

0.71

Cranberry juice cocktail and blends, light, ready-to-drink

0.71

Carbonated, ginger ale

0.70

Juice, apple and cherry

0.70

Sources for Further Reading

Print Friendly

Share: