Fluoridation of Water Supplies and Cancer—A Possible Association?

1. In the recent book Fluoride, Teeth and Health published by the Royal College of Physicians, it was asserted inter alia that there was no evidence that fluoride increases cancer mortality. 2. This statement was challenged by the presentation of data showing crude cancer mortality rates in two groups of large American cities. The rise in the crude mortality rate between 1950 and 1970 in 10 cities which had fluoride added to their water supplies early in that period was undoubtedly larger than the rise in ten unfluoridated cities. 3. The US National Cancer Institute asserted that this rise could be explained by the different age‐sex‐race structures of the populations in the two groups of cities. 4. The Royal College of Physicians invited the Council of the Royal Statistical Society to comment. 5. Our analysis shows that the two groups of 10 cities differed in their age‐sex‐race structure in 1950. The cities which were to be fluoridated started with many fewer elderly white females, somewhat fewer elderly white males and more non‐whites at all ages below 50. 6. When these differences are taken into consideration, the two groups both had higher‐than‐expected cancer mortality in standardized comparison with national rates; this conforms with world‐wide experience that cancer is generally more common in towns than in rural areas. But the two groups were not similar in 1950, despite similar crude death rates: the cities which were to be fluoridated had at that time an excess of cancer deaths which was 10·3 per 100,000 population greater than that of the control cities. 7. By 1970, the two sets of cities differed much more in their demographic structure than they had in 1950. The fluoridated cities now had many more non‐whites of all ages in their populations, and many fewer whites under the age of 55. These demographic changes made the fluoridated cities much more likely to have deaths from cancer. 8. When these demographic changes are taken into account we find, in proportional terms, that the excess cancer rate increased by 1 per cent over the 20 years in the fluoridated cities, but it also increased by 4 per cent in the unfluoridated control cities, giving a difference of 3 per cent to the advantage of the fluoridated cities. 9. Alternatively, but not in keeping with normal epidemiological practice, these results could be expressed in absolute rather than relative terms. The excess cancer rate increased by 8·8 per 100,000 population over the 20 years in the fluoridated cities, but it also increased by 7·7 per 100,000 in the control cities, giving a difference of 1·1 per 100,000 to the disadvantage of the fluoridated cities. 10. In conclusion, we must disagree with the suggestion that the data here cited are evidence that fluoridation of water supplies throughout the USA would lead to 25,000 excess cancer deaths per year. We concur with the RCP statement on the basis of the data and reports that we have examined.