Main

This systematic review of fluoridation is the fourth of the reviews commissioned by the NHMRC in Australia. The first two were carried out in 19854 and 19915 and focussed on the effectiveness of water fluoridation. The third one6 included a review of fluoride intake from discretionary fluoride supplements in addition to water fluoridation. The third review was published in 1999, and is presently available on the website of Australian Dental Association (www.ada.org.au/app_cmslib/media/lib/0703/m50958_v1_nhmrc%20fluoride.pdf). The fourth review7 published in 2007 has once again expanded its scope by including other methods of fluoride delivery, such as milk, salt, toothpaste, gel, varnish and mouthrinse. Fluoride supplements such as drops, chewable tablets and chewing gum tablets have not been explicitly included in the current review, however.

The aim of the most recent review was to synthesise the highest level of evidence to answer each clinical question. It should be noted that the levels of evidence accepted for fluoride intervention at the population level was based on those chosen for the systematic review of water fluoridation by McDonagh et al.2

The inclusion and exclusion criteria for the current review were explicit. The search strategy used to identify relevant studies could not be considered to be comprehensive as no controlled vocabulary was used in searching the electronic databases. Moreover, the range of electronic databases searched was rather limited and restricting studies to those published in the English language may also affect the findings. During the literature search, three reviewers assessed the eligibility of abstracts (approximately one third each). It is not clear whether study selection or data extraction was carried out independently or in duplicate.

Included studies were clearly laid out in table format in the appendix. This included information about the study design, population, intervention, comparator, outcomes and results. The quality of studies was assessed using the key questions from the NHMRC.7 For those study designs such as cross-sectional studies and ecological studies which had no guidance on assessment from the NHMRC, a summary of various factors relating to potential biases was provided. In addition, a global quality rating was given to each individual study. Post-hoc statistical analysis was carried out when necessary.

Two systematic reviews2, 8 and one additional, relevant, original study9 were identified in the literature search on water fluoridation and dental caries. The York review2 was chosen to form the evidence base for the effect of water fluoridation on dental caries in the current review, as it provided more detailed and comprehensive results than those shown in the review by Truman et al.6 It should be noted that 12 of the 21 studies included in the latter were among the 26 studies included in the York review.2 The lack of overlap between the two reviews is largely because the Truman review8 assessed both “fluoridation vs no fluoridation” and “fluoridation vs fluoridation at a lower level” whereas the York review5 assessed only “fluoridation vs no fluoridation”. Only one additional original study9 was identified in the current review and this did not change the conclusion from that of the York one.2 It should be noted that the benefits from fluoridated public water supply were weakened because beverages and food products processed in fluoridated communities were exported to surrounding non-fluoridated communities.10 This phenomenon is referred to as the halo effect: Griffin et al.11 attempted to quantify it by analysing data from the 1986–1987 National Survey of Oral Health in US School Children. Studies measuring the effectiveness of water fluoridation that consider only its direct benefit may have underestimated the total contribution of water fluoridation to caries reduction.

Regarding water fluoridation and dental fluorosis, the literature search identified two systematic reviews2, 12 and 10 additional original studies. It should be noted that in some cases there was a substantial difference in the prevalence of “any fluorosis,” both between different countries and within different countries. These differences result from a number of factors including methods (eg, different fluorosis indices), environmental influences (eg, phosphate mines) and lifestyles (eg, higher tea consumption). The authors concluded that although there was a fourfold risk of developing fluorosis of aesthetic concern with optimal water fluoridation compared with suboptimal water fluoridation, the absolute increase in prevalence was small (approx. 4–5%).

The studies cited in the report of the National Research Council13 have raised the possibility that infants could receive a greater than optimal amount of fluoride through liquid concentrate or powdered baby formula that has been mixed with water containing fluoride during a time when their developing teeth may be susceptible to dental fluorosis. Recently, a systematic review to investigate the association of fluorosis and infant formula has been completed.14 It concluded that the evidence suggests dental fluorosis might be caused by fluoride content in infant formula or the fluoride levels in the water used to reconstitute infant formula. Confounding factors could not be ruled out, however, and publication bias may also distort the evidence on infant formula and fluorosis.

Although the current review presents a summary of the relevant evidence, the potential effectiveness of any public health intervention must be considered in the context of practicalities associated with implementing the intervention, issues surrounding compliance, and issues related to equity of access.