Yesterday, new NHANES 2009-2012 data on prevalence, extent and severity of periodontitis have gone online in the Journal of Periodontology. On first sight, authors confirm findings of the previous, 2009-2010, survey in that prevalence is much larger as previously reported. So, 46% of U.S. adults have periodontitis with almost 9% having severe disease (Eke et al. 2015). The previous report (Eke et al. 2012a) contained an unclear description of how attachment loss was measured concealing that a “signed” measure of recession was used to calculate clinical attachment loss as difference of probing pocket depth and recession (from Latin, recessus, retreat). In a letter to the Editor-in-Chief of the Journal of Dental Research, Professor Giannobile, I had raised more problems in the article when I wrote,
“Teaching in particular undergraduates about how probing parameters periodontal probing depth, attachment level, and recession are measured is quite an effort but usually straightforward. In order to avoid undue exaggeration of prevalence, extent and severity of periodontitis both in the population and in patients attending a common office and to be able to assess treatment outcomes, metric periodontal probing parameters have to be properly defined. I would therefore appreciate if authors could comment on the apparent redefinition of attachment loss in their paper. When analyzing the Figure in the paper by Eke et al. (2012a), what immediately hits the eye is that there seems to be higher prevalence of attachment loss at different thresholds (a) than of pocket depth at respective thresholds (b) in all age groups. Such a pattern may actually be a result of how attachment loss had erroneously been redefined, most probably due to convenience. Just as a trivial example, a 4 mm probing depth without recession may be associated with either 0, 1, 2, 3, or 4 mm attachment loss, but the NHANES oral health data management program would have “instantly calculated” 4 mm. Based on the new case definition using attachment loss in addition to probing depth, prevalence of all periodontitis in the adult population of 30 years and older in the U.S. has now been estimated to exceed 47%, after 35% found in NHANES III during 1988-1994. This much higher prevalence may be due to the redefinition of attachment loss, too. Moreover, as to Eke et al. (2012a), mild periodontitis has a rather low prevalence in all age groups while moderate periodontitis is widespread (Figure c). The picture was different in NHANES III when severe periodontitis occurred with lowest, moderate periodontitis with intermediate and mild periodontitis with highest prevalence, a pattern which, I suppose, applies to many other widespread chronic diseases. The strange new pattern might indeed be explained partly by the redefinition of attachment loss as well, ultimately leading to a different distribution of cases.” (Emphasis added.)
Subtracting recession from periodontal probing depth makes sense only when true recession (the free gingival margin is located apical to the cemento-enamel junction) gets a minus sign. This was circumstantially explained to me in an email by the authors forwarded to me by Professor Giannobile, who never published my original letter. Eke et al. (2012a) had actually concealed that a signed recession definition was used. In the new update of NHANES 2009-2012, calculation of clinical attachment is now correctly described, including the signed recession definition. Eke et al. (2015) may also have realized that there is no complex chronic disease where moderate severity is more prevalent than its mild form. It is rather perplexing to see that authors have now abandoned the differentiation between moderate and mild periodontitis which they call “other” periodontitis (other than severe). They give the following reason,
“These subgroups [mild and moderate periodontitis] are not truly ordinal [sic] as the label suggests because many ‘moderate’ cases had insufficient pocket depth to qualify as ‘mild’ and we have therefore combined them and used the label ‘other‘ periodontitis.”
As I have mentioned in a previous post I had sent a letter to the editor of Journal of Dental Research, Professor W. Giannobile, in which I had asked the authors of a recent article on the periodontal findings of the 2009-2010 NHANES for some clarification regarding the quite confusing definition of attachment loss. There, Eke et al. (2012) wrote in the Material and Methods section,
“All periodontal examinations were conducted in a mobile examination center (MEC) by dental hygienists registered in at least one U.S. state. Gingival recession [= the distance between the free gingival margin (FGM) and the cemento-enamel junction (CEJ)] and pocket depth (PD) (= the distance from FGM to the bottom of the sulcus or periodontal pocket) were measured at 6 sites per tooth (mesio-, mid-, and disto-buccal; mesio-, mid-, and disto-lingual) for all teeth, excluding third molars. For measurements at each site, a periodontal probe (Hu-Friedy PVP 2TM, Chicago, IL, USA) with 2-, 4-, 6-, 8-, 10-, and 12-mm graduations was positioned parallel to the long axis of the tooth at each site. Data were recorded directly into an NHANES oral health data management program that instantly calculated attachment loss (AL) as the difference between probing depth and recession. Bleeding from probing and the presence of dental furcations were not assessed.” (Emphasis added.)
In my letter [pdf] I had pointed to (1) a possible glitch when calculating attachment loss as the difference, rather than the sum, between probing depth and recession; and (2) the fact that in the absence of any recession, periodontal probing depth must not automatically be regarded as attachment loss as the above description of the method suggested. I had provided a most trivial example regarding a 4 mm probing depth without recession which may be associated with either 0, 1, 2, 3, or 4 mm attachment loss, but it seems that the NHANES oral health data management program would have “instantly calculated” 4 mm. Then, (3) I had expressed my concern that either observation that attachment loss at certain thresholds was consistently higher than periodontal probing depth at respective thresholds in all age groups of the adult population of the United States, and the rather conspicuous finding that moderate periodontitis occurred at much higher prevalence than mild periodontal disease may in part be due to the apparently amended definition of attachment loss. And finally, as a constructive suggestion, (4) I asked the authors to compare the new 2009-2010 data with those of 1988-1994 by using the previous case definition (solely based on probing depth) by Albandar et al. (1999) in order to be able to “forget what I was told”, namely that prevalence, extent and severity of periodontal disease had in fact not decreased since NHANES III.
About two weeks ago, I had written a letter to Professor William Giannobile, Editor-in-Chief of prestigious Journal of Dental Research, as regards the paper by Eke et al. (2012) in which new data on prevalence, extent and severity of periodontal disease in the adult population of the U.S. are presented. I had written about an apparent redefinition of clinical attachment loss which might have led to an undue exaggeration of the results here on this blog.
Yesterday, Dr. Giannobile confirmed that the letter has been received. Since he explains that he might, after thorough examination “in concert with the authors”, consider an Erratum by the authors rather than publishing the letter and respective authors’ response, I have decided to link to my letter here [pdf].
I will continue to report on the issue as appropriate.
16 December 2012 @ 7:27 am.
Last modified December 16, 2012.