The systematic review by Wang et al. is well conducted and satisfies the main methodological standards of the Cochrane Collaboration - although it is not a Cochrane systematic review per se. Most importantly, this is the first systematic review that takes into account the results from the ‘Diabetes and Periodontal Therapy Trial’ (DPTT), a large randomised controlled trial published in December 2013 in the Journal of the American Medical Association.1 Results from this sufficiently powered trial (n=514) failed to demonstrate a significant effect of periodontal treatment on glycaemic control among diabetic patients.

This topic has been much discussed in the literature during recent years, including in previous issues of Evidence-Based Dentistry.2,3,4 It is worth highlighting that effect sizes reported in successive meta-analyses during the last ten years have been globally decreasing over time. Until 2010, reported effects for the mean percentage difference in HbA1c for periodontal treatment ranged between −0.46% and −0.40%.5,6,7 In 2013, values reported in two meta-analyses decreased slightly, to between −0.38% and −0.36%.8,9 Finally, since 2014 and the publication of the DPTT,1 the decrease has continued, with values from two other meta-analyses, including the one by Wang et al., showing significant values of −0.36% to −0.27%.10,11 In these two most recent meta-analyses, the main result from the DPTT trial had a marked influence on the calculated mean differences, respectively 18.1%10 and 27%.11

Considering original trials, the decreasing trend in effect sizes with time is a well-known phenomenon. Time lag in the publication of negative findings has been reported as a possible explanation.12 Another more subtle explanation is the ‘Proteus phenomenon’,13 which describes the phenomenon of rapidly alternating extreme research claims and extremely opposite refutations.14

It is possible that the widespread attention given to the potential effect of periodontal treatment on HbA1c values led to major excitement in the scientific community. Then, ‘negative’ results may have become ‘attractive’.14 Inconsistent conclusions between results from meta-analyses and large trials create confusion in clinical guidelines.15 Today, in 2015, the topic is at the crossroads, and some key points should be kept in mind:

1) When all the available evidence is combined, the effect of periodontal treatment on HbA1c still points to benefits of periodontal treatment to improve glycaemic control in diabetic patients

2) Those concerned with health care should remain attentive to the evolution of these data because the trends are decreasing, and several trials are still ongoing

3) Researchers and clinicians should consider that patient-centred outcomes (such as assessment of quality of life) could also be of interest in the topic.

Practice point

  • The available evidence in 2015 points to benefits of periodontal treatment to improve glycaemic control in diabetic patients.