Update on: Eventually Dead and Buried?

The recent joint enterprise by the American Academy of Periodontology and European Federation of Periodontology to address, once again, the meanwhile well-known connection between periodontal and systemic disease held at La Granja de San Ildefonso, Segovia in Spain in November last year has resulted in a number of review papers and consensus reports which have right now been published by the Journal of Clinical Periodontology and the Journal of Periodontology. Some are in fact informative and present lots of new information. In a press release (strangely called “manifesto”) on the occasion of the publication of supplements of the respective journals, and as regards epidemiological data on periodontitis and cardiovascular disease, it is stated that “[t]here is consistent epidemiological evidence that periodontitis imparts increased risk for future [sic!] cardiovascular disease, independently of other confounding factors.”

Well, while this statement seems to imply indeed a causal relationship (periodontitis increases the risk for future cardiovasuclar disease) the respective review article by Dietrich et al. (2013) explicitly studied (once again) the association between periodontitis and incident atherosclerotic cardiovascular disease (ACVD). However, association does not automatically mean causation, something which is nowadays fortunately known by most undergraduate medical and dental students.

In an attempt to focus on “the most robust studies” in terms of endpoint definition (incidence of ACVD) and exposure (clinically or radiographically assessed periodontal disease), the authors omitted, for example, the long considered large longitudinal studies based on early National Health and Nutrition Examination Surveys (NHANES) which had adopted now outdated measures of periodontal disease; but, apart from a number of  well-conducted prospective studies which might be able to give a certain hint about possible causation, included several case-control studies. This is absolutely okay when assessing “association” and not addressing “causation” [1].

Unsurprisingly, Dietrich et al. (2013) [2] arrived at similar conclusions (when it comes to the desired “association”) as the recent extensive systematic review by Lockhart et al. (2012) [3], of which I have written before, see here. Lockhart et al. (2012) fortunately warn, in addition, that “statements that imply a causative association between PD and specific ASVD [atherosclerotic vascular diseases]events or claim that therapeutic interventions may be useful on the basis of that assumption are unwarranted.” Dietrich et al. (2013) claim that the findings suggest that intervention studies would not be suitable in older subjects, say 60+ years, due to general weaker associations, which “present yet another formidable challenge for the design and conduct of future clinical trials that aim to address the question of benefits of periodontal therapy on adverse cardiovascular events.”

So, what’s the meaning of this new effort. No new attempt to prove a causal relationship has been or can be made at the moment. After 25 years, intervention studies which might reveal the more interesting aspect (causality) are still by and large missing. At least for the time being, one may trumpet once more that “[a] large body of research has indicated a relationship between periodontal disease and other systemic diseases” as Nancy L. Newhouse (the AAP’s new president) informed me before yesterday in an email. Well, how long have we been told this news now?

One comment on the sloppy presentation of the findings in the review paper by Dietrich et al. (2013). While the absolutely necessary information about the 12 selected studies (to get information about the prospective or case-control design, the duration, the size of the study, its outcome) is buried in the supplementary material, the main tables in the review itself are completely messed up. Important measures of associations can hardly be found. Figures were even manipulated (rounded), which was revealed only after checking the referred original paper [4].

I am afraid that the common cursory reader of either JCP or JP can only miss these data, does not understand and is not informed that inconsistent results in several studies and multiple testing may only weaken any desired conclusion in favor of a substantial association. In the worst case, he or she has to rely on the, well misleading EFP “manifesto” which seems to suggests indeed causality [5].

Notes

[1] Dietrich T, Sharma P, Walter C, Weston P, Beck J. The epidemiological evidence behind the association between periodontitis and incident atherosclerotic cardiovascular disease. J Clin Periodontol 2013; 40(Suppl 14): S70-S84: “Observational studies to date support an association between PD and ASVD independent of known confounder.” While the authors mention that “meta-analyses were not attempted due to heterogeneity of studies in terms of virtually all study characteristics”, which had been done in the past when looking at these huge prospective, long-term studies, but with rather sobering, weak associations between periodontal and cardiovascular disease, it is amazing that risk ratios, relative risks, hazard ratios and odds ratios in the presently selected studies, where subgroup analyses were a common feature, were low or at most pretty moderate, inconclusive and frequently even insignificant. So, whether a decade old studies with crude methodologies had been omitted or not did not really matter for the current conclusions.

[2] Dietrich et al. (2013): “Observational studies to date support an association between PD and ASVD independent of known confounders.”

[3] Lockhart PB, Bolger AF, Papapanou PN Osinbowale O, Trevisan M, Levison ME, Taubert KA, Newburger JW, Gornik HL, Gewitz MH, Wilson WR, Smith SC Jr, Baddour LM. Periodontal disease and atherosclerotic vascular disease: Does the evidence support an independent association?: a scientific statement from the American Heart Association. Circulation 2012;125:2520-2544, conclude [pdf] [T]he current evidence supports the notion that the incidence of ACVD, as represented by incident CHD, cerebrovascular disease and periopheral arterial disease is higher in subjects with PD or with better periodontal status, independent of many established cardiovascular risk factors.”

[4] For instance, a not significant hazard ratio in younger than 60-yr old subjects in the original paper by Dietrich, Jimenez, Krall Kaye et al. (Circulation 2008; 117: 1668-1674) of 1.55 for a bone score of 1 to less than 1.5 had been rounded up to 1.6 in the systematic review by Dietrich et al. (2013).

[5] When scrolling down, the first recommendation for practitioners is in fact misleading. The U.S. Preventive Services Task Force had already in 2009 recommended not to screen asymptomatic subjects with no history of coronary heart disease for non-traditional risk factors such as periodontitis to prevent respective events. Remaining recommendations sound, on the other hand, rather moderate, just reflecting that there is nothing new in the Perio-Cardio relationship:

  • “Practitioners should be aware of the emerging and strengthening evidence that periodontitis is a risk factor for developing [sic!] atherosclerotic cardiovascular disease, advising patients of the risk of periodontal inflammation to general as well as oral health.

  • Based on the weight of evidence, periodontitis patients with other risk factors for atherosclerotic cardiovascular disease, such as hypertension, overweight/obesity, smoking, etc. who have not seen a physician within the last year, should be referred for medical examination.

  • Modifiable lifestyle associated risk factors for periodontitis (and atherosclerotic cardiovascular disease) should be addressed in the dental surgery/office and within the context of comprehensive periodontal therapy, i.e. smoking cessation programs and advice on lifestyle modifications (diet and exercise). This may be better achieved in collaboration with appropriate specialists and may bring health gains beyond the oral cavity.

  • Treatment of periodontitis in patients with a history of cardiovascular events needs to follow American Heart Association (AHA) guidelines for elective procedures.”

5 May 2013 @ 12:55 pm.

Last update May 5, 2013.

Advertisements

4 comments

  1. Lynne Slim

    I read the consensus report about oral/systemic links and they report there is strong epidemiological evidence demonstrating that periodontitis is a risk factor for cardiovascular disease. Would like to see the evidence. I cannot find it. This is in contradiction to what was written by the AHA committee in their scientific statement and what I read by A T Merchant in the J of Evidence Based Dental Practice. A consensus statement is a low level of evidence and you will notice that the effort was sponsored by industry. Shame on the European Federation and the AAP. They need to issue an important correction to this consensus paper if they want to remain credible.

    Like

    • Muller

      Correct. But I am afraid, they would never do this. This is a political issue. I really wonder how a consensus could be reached on a statement such as, “There is consistent epidemiological evidence that periodontitis imparts increased risk for future cardiovascular disease, independently of other confounding factors,” based on weak, inconsistent, often insignificant associations observed in a few longitudinal studies. This is by the way not new; in the omitted studies published more than a decade ago by Hujoel et al., Wu et al., and Howell et al. (huge, long-term, poor methodology) exactly the same had been observed.

      By the way, “a manifesto is a published verbal declaration of the intentions, motives, or views of the issuer, be it an individual, group, political party or government. A manifesto usually accepts a previously published opinion or public consensus and/or promotes a new idea with prescriptive notions for carrying out changes the author believes should be made. It often is political or artistic in nature, but may present an individual’s life stance. Manifestos relating to religious belief are generally referred to as creeds.” (From Wikipedia, emphasis added.)

      Like

  2. Lynne Slim

    Yes, it’s sad that special interest groups can no longer be objective. I would refer periodontists to AT Merchant’s review of the evidence in the J of EB Dent Practice. I’ll copy/paste his conclusion here:

    There is an association between PD and ASVD based on observational data. Even though there are plausible biological mechanisms to explain this relation, it is not possible to exclude the possibility of unmeasured confounding impacting the results. The absence of data from well-conducted randomized controlled trials is a significant gap in knowledge. Based on these results the statement concludes that assuming a causal link between PD and ASVD is unwarranted.

    Perhaps you should write to the European Federation . . . maybe they’ll listen to you. I emailed them and didn’t get a detailed response. Whatever happened to scientific scrutiny and professional ethics?

    Like

    • Muller

      The major problem is not authors’ conclusions based on the systematic reviews. They are in fact modest, conceding that most effects are small (and therefore irrelevant), see here: http://perioatuit.com/2013/05/08/strong-evidence-for-minor-effects/. The bigger problem is the press release, or “manifesto”, which is of course political as its wording indicates.

      I can easily imagine a “manifesto”, based on exactly the same systematic reviews, which would largely play down the Perio-Systemic issue, finishing it for the time being, maybe even for a long time.

      What we need, in my opinion, is re-allocating research resources to be able to seriously deal with the real questions, e.g., as to what extent is periodontitis just a sequel of the tobacco epidemic and a low socio-economic background, etc. And, new modes of how to treat periodontal disease beyond more or less stopping the disease process which, as seems to be evident right now, does not in essence decrease the risk for cardiovascular disease, or low-birth weight. And, as regards diabetic control, an overall 0.36% reduction of HbA1c after periodontal treatment is not very convincing, is it?

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s