Enamel defects and dental caries among Southland children
Mackay T D, Thomson W M,
NZ Dental Journal 101, No. 2, June 2005 (p35-43)

Key findings from this study were as follows.

Dental fluorosis incidence

  • There were no socioeconomic differences in relation to dental fluorosis
  • The prevalence of diffuse opacities (and therefore the overall prevalence of any defect) was higher amongst those who continuously resided in fluoridated areas up to the age of 4. (Note: This should be read in conjunction with the 2008 Auckland study, which did not consider the age of exposure)

Tooth decay rates

  • Socioeconomic status (SES) did not affect tooth decay rates (note: international studies show that SES is the main determinant of dental health/ tooth decay. See in particular Armfield and Spencer 2004)
  • There was no benefit from fluoridation to the deciduous (“baby”) teeth
  • There was no benefit to permanent teeth in the initial results. Following furtehr data manipulation, it was claimed that those 9 year olds who lived all their lives in fluoridated areas had half the decay (equating to ½ a filling) than those who had never lived in a fluoridated area. This finding conflicts with the 2008 Auckland study, which found no such benefit. However, at age 9 there are few permanent teeth, and they have had only brief exposure to decay, as noted in the Auckland study. 12 years old is the WHO-prescribed age for examining permanent teeth effects. International studies show there is often an apparent benefit at younger ages, likely due to delayed tooth eruption caused by fluoridation, but which disappears by age 12 to 15 – see Armfield and Spencer (2004); Newbrun (1989).

Implications of increased dental fluorosis

The study notes that “the clinical… and public health significance of [diffuse opacities (i.e. dental fluorosis) remains unclear”. In other words, we do not know what associated health risks there may be! The US National Research Council Review, published in 2006, found a number of studies linking dental fluorosis to other adverse health effects.


436 children between the ages of 9 and 10 were examined. 137 children had not lived in fluoridated communities (but may have been given fluoride tablets – the analysis does not control for this; 116 had intermittent residence in fluoridated communities; and 183 had always lived in fluoridated communities.

Socioeconomic status was determined by the school decile rating, rather than the SES status of each child. This approach is not considered appropriate. It is not clear from the results whether this has had a significant impact.

Although a range of intersecting population characteristics was recorded, individual analysis tables look at only one factor at a time. These results are therefore meaningless, as a correlation between high decile and fluoridation status, for example, precludes any conclusion as to which factor was relevant. The fact that high decile children had the highest rates of dental fluorosis suggests there was such a correlation. The use of fluoride tablet would be exclusively among non-fluoridated children (including, possibly, intermittently fluoridated). Without controlling for this in each set of results, an “unfluoridated” child becomes a “fluoridated” child for analytical purposes.

The authors conducted a multivariate analysis for permanent tooth decay only. This approach is intended to find a common factor giving rise to results. There are many ways of conducting such an analaysis. It is not possible to determine whether the particular approach used was appropriate or not without the raw data. On the face of it, it appears that the data may have been manipulated to show a desired result when that result was not shown by standard analysis. As the saying goes “you can ‘prove’ anything with statistics.”

Overall, we feel compelled to recommend that their conclusions be viewed with extreme caution without an independent statistical analysis being available.