Figure 4-3 shows 143 prevalence estimates from studies of children outside the United States. Not shown are results for three Ethiopian communities with extremely high water fluoride concentrations of 26, 34 and 36 mg/L and prevalences of 18%, 48% and 25%, respectively (Haimanot et al. 1987). Although a positive association may be discernible, it is much less obvious than in the U.S. studies. There is little evidence of an approximate population threshold as in the results in U.S. communities (Figure 4-1). In many regions around the world, water intake among children whose permanent teeth are forming can be much more variable than in the United States, susceptibility may differ more widely, sources of fluoride intake other than the community water supply may be more prevalent, or the ascertainment of severe enamel fluorosis may be more often compromised by other determinants of dental discoloration and pitting.
One question is whether the most severe forms of enamel fluorosis, specifically those involving confluent pitting, occur at water concentrations in the range of the current MCLG of 4 mg/L. This question cannot be an-