A CLASSIC EPIDEMIOLOGIC STUDY

Dr. Frederick McKay, as a new dental graduate in the early 1900s, headed west and opened a practice in Colorado Springs, Colorado. He soon noticed that many of his patients had a curious blotching of the enamel that he had not encountered before. People in the area called it “Colorado brown stain,” and to them it was just a local oddity. It seemed harmless enough, though it was disfiguring in some cases. McKay was clearly a born scientist; he had an inquiring mind and fine powers of observation, and Colorado brown stain piqued his curiosity. In 1908 he began to investigate the extent of Colorado brown stain in the surrounding area.

In his travels over the next few years, McKay found that the condition was highly prevalent in the Colorado Springs area. It was found only in long-term residents, individuals who had been born there or who had come to the area as babies. Because the stain was difficult to polish off, McKay reasoned that it must be caused by an environmental agent that was active during the period of enamel formation. To ensure that his findings attracted some attention, McKay was shrewd enough to enlist the collaboration of G. V. Black, a major figure in dental history, in writing the first description of what then came to be called mottled enamel.¹⁵ This detailed report, in the elegant prose of the time, is a tribute to McKay’s investigative thoroughness.

McKay found that mottled enamel was endemic in many other communities along the Continental Divide and the plains to the east. It was most prevalent where deep artesian wells were the source of drinking water, and within any community the persons affected had almost invariably been users of the same water supply. By the 1920s McKay had reached the conclusion that the etiologic agent had to be a constituent of some community water supplies, despite the fact that chemical analyses all failed to identify likely constituents. In communities such as Andover and Britton, South Dakota, where he found severe mottling, he advised mothers to obtain their children’s drinking water from sources other than the community supply. In Oakley, Idaho, McKay found that children living on the outskirts of the city, using water from a private spring, were free of mottling. He advised the citizens of Oakley to abandon their old supply and tap this spring for a new source, which the community did in 1925. McKay was right, for children born in Oakley subsequent to the change were free of mottled enamel.¹²⁸

By 1930 new methods of spectrographic analysis of water had been developed. In 1931 McKay sent several samples of suspected water to an Alcoa Company chemist named Churchill, who was using these new methods. Churchill identified F in each of the samples, in amounts ranging up to 14 parts per million (ppm).²⁶ At around the same time, similar findings were reported by investigators at the University of Arizona¹⁶¹ and by a veterinary group in Morocco, then still a French colony, that was studying le darmous, the local name given to an extreme degree of mottled enamel found in Moroccan sheep.¹⁷⁴

The immediate reaction of the scientific community to the identification of F in drinking water was one of concern, because F in high concentrations was known to be a protoplasmic poison. The discovery led to the appointment in 1931 of the first dentist to the newly established National Institute of Health. This was H. Trendley Dean, who was transferred from elsewhere in the U.S. Public Health Service to become the one-person Dental Hygiene Unit, an odd name for a unit formed to investigate mottled enamel (it subsequently became the National Institute of Dental Research in 1948).

Amidst this flurry of concern, however, McKay had also noted some benefits that seemed to accompany mottled enamel. In 1928, 3 years before F was identified in drinking water, he was confident enough to publish his view that caries experience was reduced by the same waters that produced mottled enamel.¹²⁷ A similar observation was made shortly afterward in England by Ainsworth,² who like McKay was an observant dentist with an inquiring mind. Although McKay was not the first to make this suggestion, none of the earlier observers took the idea any further. McKay, and Dean as well, are good examples of history’s putting the right people in the right place at the right time.

The task of defining the relationship of F to mottled enamel now passed to Dean. His first job was to map out the prevalence of mottled enamel in the United States. Dean began like an investigative reporter, writing extensively to dental societies all around the country to ask for their experiences. He received a good response and published his first map on the distribution of mottled enamel in 1933.³² His next step was to develop a seven-point ordinal-scale index to classify the full range of mottled enamel he had seen, from the finest of lacy markings to the stained and highly friable enamel seen with extreme hypomineralization.³³

Dean began using the term fluorosis to replace mottled enamel in the mid-1930s.³⁹ He patiently surveyed children in many parts of the country, using his original fluorosis index, and built up a substantial body of information (what today we would call a database) for analysis. He devised his Community Fluorosis index based on his original seven grades of severity.³⁸ Studies through the mid-1930s analyzed many drinking water samples for minerals and other chemical constituents, but none apart from F could be related to fluorosis.^39,41 Dean chose his words carefully to define a desirable F concentration as follows:

By the mid-1930s, Dean had concluded that this “minimal threshold” level was 1 ppm F,³⁴ and that fluorosis seen in communities with water below 1 ppm F was “of no public health significance.”⁴⁰ Soon afterward he defined 1 ppm F as “the permissible maximum.”⁴⁵ Later in that decade of the Great Depression, Dean condensed his original 1934 fluorosis index to one using a six-point ordinal scale (see Chapter 17) by combining the categories of “moderately severe” and “severe.”⁴² He then added numerical values to the categories to permit quantitative comparisons among populations. By 1942 Dean had documented the prevalence of fluorosis for most of the United States.³⁶

Although he still documented fluorosis in his studies, the main theme of Dean’s research from then on was the F-caries relationship. In the mid-1930s Dean matched his fluorosis data for children in parts of South Dakota, Colorado, and Wisconsin with the caries data from an earlier 26-state survey in what today would be called an ecologic study (see Chapter 13). Although he could hardly have failed to notice the low caries experience in communities with F-bearing water during his early surveys, this was his first report in which he commented on the inverse relationship between fluorosis and caries.³⁵

Encouraged by these preliminary data, Dean chose four cities in central Illinois as study sites in which to test the hypothesis that consumption of F-containing water was associated with a reduced prevalence of caries. Galesburg and Monmouth, where Dean had already studied fluorosis,⁴⁰ used water from deep wells that averaged 1.8 and 1.7 ppm F. Macomb and Quincy used surface water averaging 0.2 ppm F. Clinical examinations of children ages 12-14 years, all with lifetime residence in their respective cities, showed that more than twice as many children in Galesburg and Monmouth were free of detectable caries than in the two cities with low-F water, and the mean number of permanent teeth affected by caries in Galesburg and Monmouth was half of that in the two cities with low-F water.⁴⁴ The evidence to support the F-caries hypothesis was now stronger.

Although caries prevalence and severity were low in Galesburg and Monmouth, fluorosis was an obvious problem in both communities. Dean and the other investigators put it this way, in the flowing prose of the time:

The next logical step was therefore to define the lowest F level at which caries was clearly inhibited. This was done through a series of investigations that have become known collectively as the 21 cities study and that are rightly considered a landmark in dental research. The first part consisted of the results of clinical examinations of children ages 12-14 with lifetime residence in eight suburban Chicago communities with various but stable mean F levels in their domestic water.⁴³ The project was later expanded by adding data from 13 additional cities in Illinois, Colorado, Ohio, and Indiana.³⁷ The collective findings of the 21 cities study are depicted graphically in Figs. 24-1 and 24-2. Fig. 24-1 shows that dental caries experience in different communities dropped sharply as F concentration rose toward 1 ppm, then leveled off. Fig. 24-2 shows the dental fluorosis experience that Dean found among the 12- to 14-year-olds in the 21 cities. Dean’s practice was to show questionable fluorosis separately in his reports, as we have done in Fig. 24-2. The data in Fig. 24-2 suggest that, had “questionable” fluorosis been included in the prevalence figures, then the level for “acceptable” fluorosis might have been set at concentrations lower than 1 ppm F.

Fig. 24-1 Decayed, missing, and filled permanent teeth (DMFT) in children ages 12-14 in 21 cities in the late 1930s, related to the fluoride concentration.^37,43

Fig. 24-2 Fluorosis experience of children ages 12-14 in 21 cities in the late 1930s, related to the fluoride concentration of drinking water.^37,43

Because the 21 cities study had a cross-sectional design, the results confirmed the association but could not by themselves establish the cause-and-effect relationship between fluoridated water and reduced caries prevalence. But the data in Figs. 24-1 and 24-2 did lead to the adoption of 1.0-1.2 ppm as the appropriate concentration of F in drinking water in temperate climates, a standard that remains in place today in the United States. These results also set the stage for a prospective test of the F-caries hypothesis, first suggested in 1943.⁸ The years of study of people using water with F levels much higher than the proposed 1 ppm (detailed later in this chapter) were sufficient to convince public authorities that the prospective tests could be carried out in safety. These first prospective studies are described in Chapter 25.

Trendley Dean died in 1962. Those who knew him said that he could be autocratic, and the reverence he inspired in his colleagues may too often have been expressed as uncritical acceptance of all he did. But Dean had the researcher’s virtues of dogged perseverance and remorseless logic as he progressed from one stage to the next. His grasp of the concept of “databasing” and his analytic mind make us wonder what he could have done with modern computer technology. Although knowledge of F’s effects has advanced considerably since those days, we enjoy the benefits of F today because of the pioneering research of Dean and his colleagues. In time, it led to the substantial decline in dental caries in high-income nations, one of history’s major public health success stories.

ENVIRONMENTAL FLUORIDE

Fluorine is one of the most reactive elements and therefore is never found naturally in its elemental form. The F ion, however, is abundant in nature and occurs almost universally in soils and waters in varying, but generally low, concentrations. Seawater contains 1.2-1.4 ppm F.¹⁸¹ Fresh surface water generally has very low concentrations, 0.2 ppm F or less, whereas concentrations of up to 29.5 ppm F have been recorded in deep well water in Arizona¹²⁰ and concentrations of over 40 ppm in boreholes in Kenya.¹¹² F’s ubiquity in soil and water means that all plants and animals contain F to some extent. Given this environmental omnipresence, it seems likely that all forms of life have evolved to thrive with continuous exposure to small amounts of F.

SOURCES AND AMOUNTS OF FLUORIDE INTAKE

Humans absorb F from air, food, and water. F intake from air is usually negligible, around 0.04 mg F/day.¹⁸⁰ Exceptions can occur around some industrial plants that work with F-rich material, such as aluminum smelters with inadequate safeguards to prevent the escape of F-containing compounds. Such environmental hazards should be controlled to the extent possible, an issue that has nothing whatever to do with the use of F to control caries.

F’s abundance in soils and plants means that everyone consumes it to some extent. Studies to estimate the average daily intake of F from all sources have provided fairly consistent results, despite both the variability of the human diet and methodologic difficulties inherent in analyzing such minute amounts. Estimates for an adult North American male in an area with fluoridated water fall within the range of 1-3 mg F/day from food and beverages,^{30,60,102,141,159,160,165} decreasing to 1 mg F/day or less in an area without fluoridation.^{30,102,141,159,160} “Market basket” analyses indicated that 6-month-old infants ingested 0.21-0.54 mg F/day in four American cities with different F concentrations in the drinking water. For 2-year-olds in the same cities, the range was 0.41-0.61 mg F/day.^139,140

F ingestion in infancy is a matter of some concern because of the risk of dental fluorosis. Methodical estimates of F intake by infants have come from the Iowa F studies, initiated in the 1990s, which showed that F exposure among infants is extensive and variable. The Iowa studies documented the F exposures of newborn infants at periodic intervals through extensive interviews about F exposure from drinking water, toothpaste, and dietary supplements. During the first 3 years of life, F ingestion from these sources averaged 0.37-0.45 mg daily from birth to 3 months, 0.5 mg at 6-9 months, 0.36 mg at 12 and 16 months, and 0.5-0.63 mg from 16 to 36 months.¹⁰⁷ Although mean intakes in Iowa were similar to those estimated in the earlier market basket surveys, there was a considerable range of intakes, with 90th percentile values well above the means and medians. When values were expressed as F intake per kilogram of body weight, nearly half of the children up to 6 months of age were found to be exceeding the desirable limits, although this proportion dropped considerably at later ages. The upper limit of intake for 12-month-old children, beyond which fluorosis risk is greatly increased, has been estimated at 0.43 mg F/day.²⁰ The 12-month-old children in the Iowa studies averaged 0.36 mg F/day,¹⁰⁷ but more than 25% of them were ingesting F above that upper limit.

The principal overall finding of the Iowa studies is the great variability of F intake during infancy, which is to be expected given the range of dietary choices for infants and the range of F concentrations in drinking waters. Despite the difficulty in conducting studies like these, the results are similar to those found in the market basket surveys described earlier and in studies from other countries.^70,115

For most people, water and other beverages provide some 75% of F intake, whether or not the drinking water is fluoridated.¹⁶⁰ This may occur because many soft drinks and fruit juices are processed in cities with fluoridated water, or it may reflect variable F content of the ingredients. One brand of grape juice in North Carolina, for example, was found to contain more than 1.6 ppm F, even when reconstituted with deionized water.¹⁴³ Even in soft drinks of the same brand, F levels can vary considerably due to production at different sites.⁷⁶ Another source of high F intake for infants is powdered infant formula reconstituted with fluoridated drinking water.²² The F content of some tested beverages is shown in Table 24-1.

FLUORIDE PHYSIOLOGY

Although the use of F is a contribution to the public’s health of which dentistry can be proud, F compounds must be handled responsibly and with respect. Everyone in dentistry should understand how the human body handles ingested F so that the material can be used safely and efficiently.

FLUORIDE AND HUMAN HEALTH