Sir, I read the paper on toothwear/erosion, water fluoridation and deprivation (BDJ 2004, 197: 413) with interest. Although correctly described as the first to investigate fluoridation and erosion in the UK, it was preceded by a study in Ireland1 last year. A number of flaws require highlighting, to enable the results to be taken in context.

The authors correctly state that epidemiology studies are difficult to compare due to a wide range of indices used, but then use a new index and index teeth (canines), instead of following other national conventions2. They further muddy the waters by referring to tooth wear, erosion and smooth surface tooth wear, the latter two being the same thing. They then confuse and compromise their results by excluding molar data on the basis that attrition plays a part. Yet the differentiation between attrition and erosion on these surfaces is clear 3,4,5, and a greater amount of erosion has been identified on these surfaces than incisors6.

The literature on erosion and social class is not conflicting. For the permanent dentition in children, studies have found a significant correlation 2,7,8,9,10,11. The paper quoted involving primary dentitions can be dismissed as it was based upon social class of school area and not examinees residence, whereas others found a positive relationship1,12 or no difference2,13 between social groups.

All results quoted, are for fluoridation status at the time of examination (2,351 children) and not for those with lifelong exposure to fluoride (n – 1,331). This is a fundamental flaw. Although no results are provided this is dismissed in one sentence, confirming life long exposure did not result in a significant reduction in erosion.

Comparisons are made with the results of Teo et al14. This paper should be discounted as only 96 individuals were examined, 67% of the fluoridated group were aged 12 to 29 years, whilst 72.3% of the non-fluoridated residents were aged 30 to 60+, making any results a complete nonsense.

Stating 'the risk of smooth surface tooth wear (erosion) is decreased by a factor of 1.5 by water fluoridation' infers that children who have the same consumption rates of acidic drinks are at less risk by 1.5 times. But if they are drinking pop, they are not drinking the fluoridated water and thus not gaining from the effects.

It is known that gender, ethnicity, deprivation and consumption of carbonated drinks has an effect on the prevalence of erosion11,15. Thus when investigating a further factor such as the effect of fluoridation, these factors have to be controlled and balanced when choosing the case and control samples. Any results otherwise obtained, such as in this study, must be taken with a large pinch of salt.

The critical pH of hydroxyapatite is 5.5 and fluorapatite 4.5. Many acidic foods and drinks have a pH way below these, with a rapid and blanket effect, so the logical conclusion is that fluoridation should have little or no effect on the prevalence of erosion as in Ireland1.

One of the authors of the paper, Dr A Milosevic responds: There is no 'national convention' for measurement of wear although the national survey did examine the labial and palatal surfaces of the four upper incisors. The canines were included because twelve anterior teeth were reported to provide greater sensitivity and specificity in detecting wear cases 15 . Dr Dugmore is partly correct regarding the degree of molar involvement. Occlusal surfaces of the first molars exhibit dentine exposure more commonly than palatal surfaces of incisors but it cannot be stated with any certainty that this is solely from acid erosion 17 . I personally suspect that cupped cuspal tips with exposed dentine are primarily eroded but cusp tips will contact opposing fossae in this age group. Excluding attrition on occlusal surfaces is impossible. I still believe, however, that first molars should be included in any index, as a marker of dentinal exposure and susceptibility to wear. We strongly disagree that 'the results are confused and compromised'.

Although Dr. Dugmore states the literature regarding the relationship between erosion and social class is not conflicting, he cites several studies with different results.

Dr Dugmore mis-read the number of children who benefited from fluoride since birth. Only 383 children were in this category from a total number of 1,344 who had been resident at the same address since birth. Changing homes reduced the number with life long exposure to fluoridated water significantly and perhaps a larger sample would have produced a different result.

Between group analysis (fluoridated and non-fluoridated) for consumption rates was not assessed and we made no claim regarding this. Children would probably drink diluted squash and drink water occasionally not to mention brush their teeth with tap water. Topical effects are more important than systemic at this age.

It is agreed that “gender, ethnicity, deprivation and consumption” may influence the prevalence of erosion. However, carrying out statistical analyses on sub-samples in our study would have resulted in reduced power and increased the risk of type II error.

Studies on erosion and abrasion in vitro have found that fluoride reduces mineral loss18,19,20. It would be counter-intuitive if water fluoridation did not have a protective effect.