Fluoride has substantial benefits in the prevention of tooth decay. Numerous studies, taken together, clearly establish a causal relationship between water fluoridation2 and the prevention of dental caries. While dental decay is reduced by fluoridated toothpaste and mouth rinses, professional fluoride treatments and fluoride dietary supplements, fluoridation of water is the most cost-effective method. It provides the greatest benefit to those who can least afford preventive and restorative dentistry and reduces dental disease, loss of teeth, time away from work or school, and anesthesia-related risks associated with dental treatment.

In the 1940's, children in communities with fluoridated drinking water had reductions in caries scores of about 60 percent as compared to those living in non-fluoridated communities. Recent studies still reveal that caries scores are lower in naturally or adjusted fluoridated areas; however, the differences in caries scores between fluoridated and non-fluoridated areas are not as great as those observed in the 1940's. This apparent change is likely explained by the presence, in non-fluoridated areas, of fluoride in beverages, food, dental products, and dietary supplements.

Fluoride has been used for nearly 30 years as an experimental therapy to treat osteoporosis, but has only recently been evaluated in controlled clinical trials. Two new U.S. clinical trials showed no significant reduction in the rates of bone fractures related to the administration of fluoride. An FDA advisory panel has concluded that fluoride therapy has not been shown to be effective in reducing the frequency of vertebral fractures.


Among the more significant health conditions evaluated in relation to fluoride intake are cancer, dental fluorosis, and bone fractures. Other conditions are evaluated in the full report.

Cancer. Two major scientific approaches have been used to determine whether an association exists between the use of fluoride and cancer: carcinogenicity studies in rodents, and human epidemiological analyses which compare cancer incidence and mortality between communities with fluoridated water and those with negligible amounts of fluoride in drinking water.

Five carcinogenicity studies in animals have been reported in the biomedical literature. Three studies, conducted before 1970 and interpreted as negative, had significant methodological limitations, as judged by current standards of experimental design. Two subsequent studies were conducted using current standards to evaluate the carcinogenicity of sodium fluoride in experimental animals.

One of the two carcinogenicity studies was conducted by the National Toxicology Program (NTP). This peer-reviewed study provided sodium fluoride in drinking water to rats and mice and determined the occurrence of tumor formation many different organ systems. The peer review panel concluded that, "Under the conditions of these 2-year dosed water studies, there was equivocal evidence of carcinogenic activity of sodium fluoride in male F344/N rats, based on the occurrence of a small number of osteosarcomas in dosed animals. There was no evidence of carcinogenic activity in female F344/N rats receiving sodium fluoride at concentrations of 25, 100, or 175 ppm (0, 11, 45, 79 ppm fluoride) in drinking water for 2 years. There was no evidence of carcinogenic activity of sodium fluoride in male or female mice receiving sodium fluoride at concentrations of 25, 100, or 175 ppm in drinking water for 2 years." The Ad Hoc Subcommittee on Fluoride concurs with this conclusion.

The other carcinogenicity study was sponsored by the Procter and Gamble Company using Cr:CD (Sprague-Dawley) rats and Crl:CD-1 (ICR) mice both treated with 0, 4, 10, or 25 milligrams/kilogram/day sodium fluoride added to a low fluoride-basal diet. A second control group received powdered rodent chow. There was no evidence of malignant tumors associated with sodium fluoride in mice and rats of either sex in the Procter and Gamble study. While there were two osteosarcomas in the low dose female rats, one osteosarcoma in a high dose male rat, and one fibroblastic sarcoma in a mid-dose male rat, these findings in treated animals were not statistically different from controls. Male and female mice in the study did have a statistically significant increase in benign bone tumors (osteomas). The significance of a Type C retrovirus, detected in the osteomas, remains to be determined. Osteomas and osteosarcomas are different in anatomical site and clinical course. The FDA also noted difficulty in assessing the dose-related aspects of the osteomas in mice (see page 75-6). Furthermore, osteomas and osteosarcomas are so rare normally in rodents that the relationship between these tumors cannot be accurately stated.

When the NTP and the Procter and Gamble studies are combined, a total of eight individual sex/species groups are available for analysis. Seven of these groups showed no significant evidence of malignant tumor formation. One of these groups, male rats from the NTP study, showed "equivocal" evidence of carcinogenicity, which is defined by NTP as a marginal increase in neoplasms—i.e., osteosarcomas—that may be chemically related. Taken together, the two animal studies available at this time fail to establish an association between fluoride and cancer.

There have been over 50 human epidemiology studies of the relationship between water fluoridation and cancer. Epidemiological studies of fluoride usually attempt to identify statistical associations between cancer rates and county- or city-wide patterns of water fluoridation. Expert panels which reviewed this international body of literature agree that there is no credible evidence of an association between either natural fluoride or adjusted fluoride in drinking water and human cancer (LARC, 1982; Knox, 1985). Interpretation of these studies is limited by the inability to measure individual fluoride exposures or to measure other individual predictors of cancer risk, such as smoking or occupational exposures.

In March of 1990, the National Cancer Institute (NCI) updated and expanded an earlier analysis of cancer deaths, by county in the United States, to determine whether there is or is not an association between cancer and fluoride in drinking water. The new studies evaluated an additional 16 years of cancer mortality data, and also examined patterns of cancer incidence between 1973 and 1987 in the Surveillance, Epidemiology and End Results (SEER) Program cancer registries. SEER, an NCI sponsored network of population-based cancer incidence registries, started in 1973 and represents about 10 percent of the U.S. population. The SEER registries were used to obtain incidence data on all cancers, with special emphasis placed on trends in osteosarcoma. Because mortality data do not contain information on tumor-specific pathology, analysis of osteosarcomas is limited to the incidence data.

The NCI study identified no trends in cancer risk which could be attributed to the introduction of fluoride into drinking water. The study examined nationwide moronity data and incidence data from counties in Iowa and the Seattle, Washington, metropolitan area. There were no consistent differences in the trends in cancer mortality rates among males and females living in counties having initiated relative mortality rates from cancer, including cancer of the bones and joints, were similar after 20-35 years of fluoridation as they were in the years preceding fluoridation. In addition, there was no relationship between the introduction and duration of fluoridation and the patterns of cancer incidence rates, including those of the bone and joint, and the subset of osteosarcomas (Appendix E). For example, there were 91 observed cases of osteosarcoma in the fluoridated areas, when 93 cases were expected based on rates in non-fluoridated areas.

The NCI also conducted a more detailed evaluation of osteosarcomas using nationwide age-adjusted incidence data from the entire SEER database for the years 1973-1987 (Appendix F). Osteosarcoma is a rare form of bone cancer, the cause of which is under study. Approximately 750 newly diagnosed cases occur each year in the United States, representing about 0.1 percent of all reported cancers. Between two time periods, 1973-1980 and 1981-1987, there was an unexplained increase in the annual incidence rates of osteosarcoma in young males under age 20 from 3.6 cases per 1,000,000 people (88 registry cases) to 5.5 cases per 1,000,000 people (100 registry cases). This compares to a decrease in young females of the same age group from 3.8 cases per 1,000,000 people (87 registry cases) to 3.7 cases per 1,000,000 people (63 registry cases). The amount of increase observed in young males was greater in fluoridated than in non-fluoridated areas. Although the reason for the increase in young males remains to be clarified, an extensive analysis reveals that it is unrelated to the introduction and duration of fluoridation.

In studying rare cancers, such as osteosarcoma, small increases in risk, on the order of 5 to 10 percent, would not likely be detected. While descriptive epidemiological studies are useful in determining whether or not there is a credible association, the qualitative nature of any association, if one exists, can best be determined through more refined methods, such as case-control studies.

Dental Fluorosis. Dental fluorosis has been recognized since the turn of the century in people with high exposure to naturally occurring fluoride in drinking water. It has always been more prevalent in fluoridated than non-fluoridated areas. Dental fluorosis only occurs during tooth formation and becomes apparent upon eruption of the teeth. It ranges from very mild symmetrical whitish areas on teeth (very mild dental fluorosis) to pitting of the enamel, frequently associated with brownish discoloration (severe dental fluorosis). The very mild form barely is detectable even by experienced dental personnel. Moderate and severe forms of dental fluorosis, considered by some investigators as presenting a cosmetic problem, do not appear to produce adverse dental health effects, such as the loss of tooth function, and represents less than 6 percent of the cases of fluorosis nationally.

In the 1940's, about 10 percent of the population displayed very mild and mild dental fluorosis when the concentration of fluoride found naturally in the drinking water was about 1 part per million (ppm). Over the last 40 years, in areas where fluoride is added to the drinking water to bring the total level of fluoride to about 1 ppm (optimally fluoridated areas), there may have been an increase in the total prevalence of dental fluorosis. In non-fluoridated areas, there is clear evidence that the total prevalence of dental fluorosis has increased over the last 40 years.

The greater the fluoride exposure during tooth development, the greater the likelihood of dental fluorosis. In the 1940's and 1950's, the major sources of fluoride were from drinking water and food. Since then, numerous sources of fluoride have become available, including dental products containing fluoride (e.g., toothpastes and mouth rinses) and fluoride dietary supplements. The inappropriate use of these products can contribute significantly to total fluoride intake.

Increases in the prevalence of dental fluorosis in a population should be taken as evidence that fluoride exposure is increasing. Because dental fluorosis does not compromise oral health or tooth function, an increase in dental fluorosis, by itself, is not as much of a dental public health concern as it is an indication that total fluoride exposure may be more than necessary to prevent tooth decay. Prudent public health practice generally dictates using no more of a substance than the amount necessary to achieve a desired effect.

Bone Fractures. There is some suggestion from epidemiological studies that the incidence of certain bone fractures may be greater in some communities with either naturally high or adjusted fluoride levels. However, there are a number of confounding factors that need resolution to determine whether or not an association exists. Additionally, other studies do not show an increase in the incidence of bone fractures; one study provided evidence of a lower incidence of bone fractures in an optimally fluoridated community as compared to a similar community with trace levels of fluoride in the water. Therefore, further research is required.


1For other findings, see Other Effects section page 87 of the report.

2Fluoridated area means that drinking water supplied to that area contains fluoride at least at the optimal level (0.7-1.2 ppm depending on temperature) either naturally or by the water's having been adjusted to that level. Non-fluoridated area means that the drinking water does not contain fluoride at the optimal level. The level of fluoride naturally present in the non-fluoridated areas varies from study to study, but generally ranges from negligible levels to less than 0.3 ppm.

Return to Table of Contents

Return to Committee Reports Page