|
Sign In to gain access to subscriptions and/or personal tools.
|
Optimal Drinking Water Composition for Caries Control in Populations
M. Bruvo1,
K. Ekstrand1,
E. Arvin2,
H. Spliid3,
D. Moe1,
S. Kirkeby1 and
A. Bardow1,*
1 Dental School of Copenhagen, University of Copenhagen, Nørre Alle 20, 2200 Copenhagen N, Denmark; and
2 Institute of Environment & Resources and
3 Informatics and Mathematical Modelling, Technical University of Denmark, 2800 Kgs. Lyngby, Denmark
Correspondence: * corresponding author, ab{at}odont.ku.dk
 |
ABSTRACT
|
|---|
Apart from the well-documented effect of fluoride in drinking water on dental caries, little is known about other chemical effects. Since other ions in drinking water may also theoretically influence caries, as well as binding of fluoride in the oral environment, we hypothesized that the effect of drinking water on caries may not be limited to fluoride only. Among 22 standard chemical variables, including 15 ions and trace elements as well as gases, organic compounds, and physical measures, iterative search and testing identified that calcium and fluoride together explained 45% of the variations in the numbers of decayed, filled, and missing tooth surfaces (DMF-S) among 52,057 15-year-old schoolchildren in 249 Danish municipalities. Both ions had reducing effects on DMF-S independently of each other, and could be used in combination for the design of optimal drinking water for caries control in populations.
Key Words: drinking water • fluoride • calcium • dental caries
|
INTRODUCTION
|
|---|
Shortage of drinking water in many parts of the world will increase the use of seawater and brackish water as a drinking water resource, especially in densely populated coastal areas (Dawoud, 2005). After treatment, such water is depleted of ions. To obtain a satisfactory taste and to reduce the corrosive potential of the water when distributed to consumers, engineers normally return some ions after desalination. We hypothesized that the return of a mixture of selected ions could make desalinated drinking water optimal for caries control. However, apart from the well-documented effects of fluoride (McDonagh et al., 2000), the literature is sparse on reports dealing with effects of drinking water composition on dental caries. Past investigations have focused mostly on rare trace elements, such as lithium (Schamschula et al., 1981), copper and lead (Ludwig et al., 1970), strontium, boron, and molybdenum (Losee and Adkins, 1968, 1969; Curzon et al., 1970), as well as water hardness and calcium (Mills, 1937; Glass et al., 1973). No analysis exists on the relationship between all standard chemical characteristics of drinking water and dental caries on population data.
Denmark offers an ideal model system for testing such effects. Geographic variations in caries experience among children are recorded in detail (Helm, 1973), and the drinking water quality is monitored for every waterworks. Close to 100% of the Danish drinking water supply is currently based on groundwater. Analysis of these data has shown that fluoride in drinking water by itself explains 35% of inter-municipality variations in numbers of decayed, missing, and filled tooth surfaces (DMF-S) among Danish schoolchildren (Ekstrand et al., 2003). Here, we aimed to take past investigations further by testing the effects of all standard chemical characteristics of drinking water on DMF-S in schoolchildren.
|
MATERIALS & METHODS
|
|---|
Data Collection
From the Geological Survey of Denmark and Greenland (GEUS), chemical data were retrieved as average concentrations from 1995 to 2004 for each waterworks in the 275 Danish municipalities. In total, data from 3364 waterworks were obtained, and a weighted average was calculated based on the drinking water flow from each waterworks in a given municipality and the total drinking water consumption for the same municipality. Depending on the amount of water produced, the quality of the water was monitored every second year in half the waterworks, and annually or more in the other half of the waterworks. Hydrogen carbonate was by far the most abundant ion, constituting more than half the compounds in the water by weight, followed by sulphate, calcium, sodium, and chloride (Table 1 ). DMF-S data for Danish schoolchildren aged 15 in 2004 were retrieved from the Danish National Board of Health. We chose data on 15-year-olds because this recording is the last compulsory recording, it has the most children registered, and it is made while the majority are still in secondary school. Therefore, the general conditions for these children are relatively alike, while still allowing for sufficient exposure time for variations in caries to develop. The data cover 249 Danish municipalities, where DMF-S values on more than 60% (mean 85%) of the children were reported, amounting to 52,057 children and, geographically, the land of Denmark. DMF-S data were obtained anonymously and by permission from the Danish National Board of Health, according to the requirements of the National IRB.
Statistics
Statistical analyses were done with Excel and the R statistical program (R Development Core Team, 2006). The identification and estimation of the most significant drinking water characteristics for DMF-S were performed by an iterative search and testing algorithm based on generalized linear modeling. Consider a municipality where n children have been observed with a total number of DMF-S Y and with mean DMF-S = Y/n. We assume that Y follows a Poisson distribution with a mean equal to n · E · . The term E, a random factor with a mean of 1, accounts for unobserved sources of variation, such as socio-economic factors in the municipality in question. The effect of drinking water on DMF-S is described by , which depends on chemical characteristics of the water in the municipality. For example, with 2 chemical characteristics, the natural model is:
 | (Eq. 1) |
where A and B are 2 chemical characteristics in the drinking water in the municipality in question, while µ,  , and β are the general parameters of interest, describing the effects of A and B on DMF-S. The formulation and statistical analysis of this model have been described previously (Venables and Ripley, 2002).
|
RESULTS
|
|---|
All chemical water characteristics were subjected to iterative search and testing, with DMF-S as the outcome resulting in the isolation of fluoride and calcium (Fig. A). With fluoride and calcium being identified to be of major importance, the model was:
View larger version (32K):
[in this window]
[in a new window]
|
Figure. Drinking water and DMF-S. (A) The relationship among DMF-S, calcium, and fluoride in Danish drinking water, where each column represents one of the 249 Danish municipalities, covering a total of 52,057 children, which were included in the analyses. (B) The nonlinear relationship among calcium, fluoride, and DMF-S according to the model for DMF-S (Eq. 2), where both the estimate for fluoride (–0.18 ± 0.03; p < 0.001) and the estimate for calcium (–0.11 ± 0.02; p < 0.001) were negative and highly significant.
|
|
 | (Eq. 2) |
where 1.05 is the value for µ (the intercept in the regression analysis), 0.18 the estimate for , and 0.11 the estimate for β, with fluoride and calcium in standardized forms. Standard deviations for the estimates of µ, , and β were 0.02, 0.03, and 0.02 (p < 0.001), respectively. When the mean national calcium and fluoride concentrations (Table 1 ) are inserted into this model, only the value 1.05 (i.e., µ) will remain, giving rise to a mean DMF-S for 15-year-old schoolchildren of 2.86, which is equal to the mean DMF-S for the 52,057 children in the analysis. The combined explanatory power for this model, showing both fluoride and calcium to reduce DMF-S, was 45%, and no significant interactions or higher-order terms were found. How various concentrations of calcium and fluoride in the drinking water will affect DMF-S is illustrated in the Fig. (B) and Table 2. Due to the non-linear relations with calcium and fluoride, the DMF-S estimate will never become zero. Therefore, the effects of both ions become less at higher concentrations, where the effects tend to flatten out (Fig., B).
|
DISCUSSION
|
|---|
Fluoride makes the tooth surface more resistant to acid, probably by surface fluoridation of apatite crystals (Shellis and Duckworth, 1994), and this resistance increases with the concentration of fluoride (ten Cate and Duijsters, 1983). Since the fluoride concentration is normally much higher in Danish drinking water than the concentration of dissolved fluoride in the aqueous environment of the dental tissues, fluoride from the water protects against caries. We speculated that calcium from drinking water favors remineralization and also reduces demineralization in early caries lesions. Another effect of calcium on DMF-S could relate to the positive relationship between this ion and fluoride levels in plaque (Whitford et al., 2002). In this way, calcium may diffuse into plaque and provide extra binding sites for fluoride (Vogel et al., 2006). In both cases, the concentration of free calcium has to be higher in the drinking water than in saliva. Given the average concentration of total calcium in the drinking water, the concentration of free calcium was calculated to be 2.0 mmol/L (93% of total calcium) at 25°C (Parkhurst and Appelo, 1999). For comparison, human whole saliva contains only around 1.0 mmol/L of total calcium (Matsuo and Lagerlöf, 1991). In addition, human saliva also contains high concentrations of bicarbonate (Bardow et al., 2000) as well as proteins and phosphorus, which all result in considerable binding of calcium (Gron, 1973; Hay et al., 1982). Therefore, the free calcium concentration in human saliva is often only around half the total calcium concentration (Gron, 1973; Matsuo and Lagerlöf, 1991). Accordingly, the average concentration of free calcium in Danish drinking water is 4 times higher, and in some areas more than 7 times higher, than it is in human saliva. Remineralization and effects related to fluoride uptake in plaque may therefore occur upon frequent exposure to such drinking water. In this context, Mills (1937) showed that, among 75 cities in the US, those having the highest drinking water hardness had the lowest DMF-S. We propose that Mills’ results were indirect measures of calcium, a main determinant of water hardness, and reflect the effect obtained in the present study. In support of this, Glass et al.(1973), in 2 isolated villages in Colombia, South America, also identified calcium as caries-protective in drinking water.
Apart from the calcium/fluoride model given above, an expanded model containing calcium and fluoride as well as pH, bicarbonate, and chloride could also be developed, thereby increasing the explanatory power of the effect of drinking water composition on DMF-S to 51% (p < 0.001). In this expanded model, the estimates for the effects of calcium, fluoride, pH, and bicarbonate on DMF-S were negative, whereas the estimate for chloride was positive. Although these effects may be theoretically explained, and clearly illustrate the complexity of the effects of drinking water on DMF-S, we believe that the simplicity of the smaller calcium/fluoride model makes it the core model for describing effects of drinking water on DMF-S.
This is the first time a mathematical model with 2 ions has been developed for the effect of drinking water on DMF-S. This calcium/fluoride model could be used for the design of optimal drinking water for caries control in populations, especially from desalinated sea- and brackish water. But the model could also be used for optimizing drinking water from groundwater resources in countries where water fluoridation is common. We do not suggest any replacement of fluoride with calcium, which, according to the model, may be done with 170 milligrams of calcium for each milligram of fluoride, but merely suggest the use of both ions combined whenever possible. Thus, high concentrations of calcium are unwanted, because it leads to relatively high consumption of laundry detergents in households and calcium carbonate precipitation on surfaces in water distribution pipes and in bathrooms. Instead, a supply of drinking water with 0.75 mg/L fluoride and 90 mg/L of calcium, the latter giving a free calcium concentration more than 4 times higher than that in saliva, will result in a substantial reduction in DMF-S, while still allowing for a supply of drinking water with moderate concentrations of both ions.
|
ACKNOWLEDGMENTS
|
|---|
We are grateful for funding given to the project by The Danish Dental Association. A preliminary report was presented at the ORCA 2007 conference in Ellsinore, Denmark.
Received for publication January 15, 2007.
Revision received January 10, 2008.
Accepted for publication January 23, 2008.
|
REFERENCES
|
|---|
- Bardow A, Madsen J, Nauntofte B (2000). The bicarbonate concentration in human saliva does not exceed the plasma level under normal physiological conditions. Clin Oral Investig 4:245–253.[CrossRef][Medline]
[Order article via Infotrieve]
- Curzon ME, Adkins BL, Bibby BG, Losee FL (1970). Combined effect of trace elements and fluorine on caries. J Dent Res 49:526–528.
- Dawoud MA (2005). The role of desalination in augmentation of water supply in GCC countries. Desalination 186:187–198.
- Ekstrand KR, Christiansen ME, Qvist V (2003). Influence of different variables on the inter-municipality variation in caries experience in Danish adolescents. Caries Res 37:130–141.[Medline]
[Order article via Infotrieve]
- Glass RL, Rothman KJ, Espinal F, Velez H, Smith NJ (1973). The prevalence of human dental caries and water-borne trace metals. Arch Oral Biol 18:1099–1104.[CrossRef][Medline]
[Order article via Infotrieve]
- Gron P (1973). The state of calcium and inorganic orthophosphate in human saliva. Arch Oral Biol 18:1365–1378.[Medline]
[Order article via Infotrieve]
- Hay DI, Schluckebier SK, Moreno EC (1982). Equilibrium dialysis and ultrafiltration studies of calcium and phosphate binding by human salivary proteins. Implications for salivary supersaturation with respect to calcium phosphate salts. Calcif Tissue Int 34:531–538.[CrossRef][Medline]
[Order article via Infotrieve]
- Helm S (1973). National statistics on caries and oral hygiene derived from the Danish Child Dental Health recording system. Community Dent Oral Epidemiol 1:121–126.[Medline]
[Order article via Infotrieve]
- Losee FL, Adkins BL (1968). Anti-cariogenic effect of minerals in food and water. Nature 219:630–631.[Medline]
[Order article via Infotrieve]
- Losee FL, Adkins BL (1969). A study of the mineral environment of caries-resistant Navy recruits. Caries Res 3:23–31.[Medline]
[Order article via Infotrieve]
- Ludwig TG, Adkins BL, Losee FL (1970). Relationship of concentrations of eleven elements in public water supplies to caries prevalence in American schoolchildren. Aust Dent J 15:126–132.[Medline]
[Order article via Infotrieve]
- Matsuo S, Lagerlöf F (1991). Relationship between total and ionized calcium concentrations in human whole saliva and dental plaque fluid. Arch Oral Biol 36:525–521.[Medline]
[Order article via Infotrieve]
- McDonagh MS, Whiting PF, Wilson PM, Sutton AJ, Chestnutt I, Cooper J, et al. (2000). Systematic review of water fluoridation. BMJ 321:855–859.[Abstract/Free Full Text]
- Mills CA (1937). Factors affecting the incidence of dental caries in population groups. J Dent Res 16:417–430.
- Parkhurst DL, Appelo CAJ (1999). User’s guide to PHREEQC (version 2)—a computer program for speciation, batch reaction, one-dimensional transport, and inverse geochemical calculations. Water-Resources Investigations Report 99-4259. US Geological Survey. http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc.
- R Development Core Team (2006). R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. http://www.r-project.org.
- Schamschula RG, Cooper MH, Agus HM, Un PS (1981). Oral health of Australian children using surface and artesian water supplies. Community Dent Oral Epidemiol 9:27–31.[Medline]
[Order article via Infotrieve]
- Shellis RP, Duckworth RM (1994). Studies on the cariostatic mechanisms of fluoride. Int Dent J 44:263–273.[Medline]
[Order article via Infotrieve]
- ten Cate JM, Duijsters PP (1983). Influence of fluoride in solution on tooth demineralization. I. Chemical data. Caries Res 17:193–199.[Medline]
[Order article via Infotrieve]
- Venables WN, Ripley BD (2002). Modern applied statistics with S. 4th ed. New York: Springer Science & Business Media LLC, pp. 183–206.
- Vogel GL, Shim D, Schumacher GE, Carey CM, Chow LC, Takagi S (2006). Salivary fluoride from fluoride dentifrices or rinses after use of a calcium pre-rinse or calcium dentifrice. Caries Res 40:449–454.[Medline]
[Order article via Infotrieve]
- Whitford GM, Wasdin JL, Schafer TE, Adair SM (2002). Plaque fluoride concentrations are dependent on plaque calcium concentrations. Caries Res 36:256–265.[Medline]
[Order article via Infotrieve]
Journal of Dental Research, Vol. 87, No. 4,
340-343 (2008)
DOI: 10.1177/154405910808700407
 CiteULike  Complore  Connotea  Del.icio.us  Digg  Reddit  Technorati  Twitter What's this?
|
|
TOP
ABSTRACT
INTRODUCTION