Discussion
ORAL-D is a large-scale prospective cohort study that was designed to examine oral diseases, including periodontitis, and associations with clinical outcomes in adults with kidney failure treated with long-term hemodialysis. Moderate to severe periodontitis was highly prevalent and affected about 40% of the cohort. Overall, there was no evidence of an association with all-cause or cardiovascular mortality in people treated with dialysis and who had moderate to severe periodontitis, when controlling for potentially confounding variables. In analyses in which participants were otherwise matched for key clinical and sociodemographic characteristics, those with moderate to severe periodontitis had a lower risk of all-cause and cardiovascular mortality, than those with no or mild periodontitis. Findings were similar when limited to adults with a minimum of 12 natural teeth and when competing causes for cardiovascular death were considered.
These findings may be in contrast to some studies in the general population and in adults with other chronic diseases, which show that periodontitis is associated with increased mortality and cardiovascular events [
10,
31]. The study findings are consistent with other studies which show no association of periodontitis with mortality among patients with coronary artery disease and older adults [
13,
32]. The lack of association between periodontitis and increased mortality among people with end-stage kidney disease has several potential explanations. First, the most likely explanation is that periodontitis may be associated with factors that are also linked to mortality, such as age, smoking, employment, and time treated with dialysis, but that the effects of these potential confounding variables linked to both periodontitis and mortality were reduced in these analyses through restriction to propensity-matched patients or statistical adjustment. Indeed, dental disease is linked to socioeconomic disadvantage, lower health literacy, rural dwelling, [
33] cardiovascular risk, and reduced receptiveness to health advice and engagement with health services, [
34] all of which are related to increased mortality [
35‐
37]. It is possible these factors were minimized in this study by controlling for measured confounders that might be indicative of these broader determinants of both dental health and survival. Second, loss of tooth attachment in older adults caused by periodontal destruction due to long-term gingival inflammation might have led to misclassification of adults with milder periodontal pocket depth due to tooth attachment loss as having no or mild periodontitis when, in fact, shallow periodontal pocket depth may have been indicative of more advanced periodontal disease. While some prognostic studies have similarly classified periodontitis by considering gingivitis and periodontal pocket depth, and concluded periodontitis is linked to worse survival, [
19] other studies have evaluated alveolar bone loss and tooth mortality as indicating of the natural evolution of severe periodontal loss [
10]. Third, ORAL-D provided short-term outcome information for participants at 12–24 months after baseline, at which time the negative effect of periodontitis per se may not have been sufficiently relevant to risks of mortality.
Although observational studies in the non-dialysis setting have generally demonstrated that periodontitis is associated with poor health outcomes, a recent trial of nonsurgical periodontal treatment did not improve glycemic control [
24]. A recent uncontrolled study in adults treated with peritoneal dialysis showed improved C-reactive protein and blood urea nitrogen levels following periodontal treatment, although the clinical relevance of the measured changes in these biochemical endpoints is uncertain [
38]. In contrast, a randomized trial of periodontal treatment showed no evidence of sustained improvement in periodontal health or inflammatory markers at 6 months among 53 dialysis patients [
25]. Ultimately, understanding any effects of periodontal treatment on survival in the setting of chronic kidney disease requires a large adequately-powered trial, although based on the findings of ORAL-D, a more convincing basis within observational studies is required before trials are planned.
The ORAL-D study has a number of strengths. First, the sample size of ORAL-D was sufficiently large to detect a meaningful difference in mortality adjusted for relevant socio-demographic and clinical variables and was geographically diverse, increasing the generalizability of the findings. Second, the oral examination was calibrated across study sites and valid methods for oral health assessment and periodontitis were used. WHO criteria in fact recommend CPI and a specially designed probe for the evaluation of the periodontal status [
27]. Despite this may be not as sensitive as the Williams probe, mostly used in the routinely periodontal practice, the WHO probe was conceived for epidemiological studies. Because of the large ORAL-D study sample, CPI and not PISA score was used as validated epidemiological screening procedure in populations (at the time of recruitment, PISA score had been only used in one small size cohort study [
32]). CPI remains till today the most used and validated method to evaluate severity and degree of periodontal disease. Third, participants showed important differences in key periodontal characteristics at baseline suggesting patients had differing exposures to periodontal pathology during the course of follow up.
However, potential limitations of ORAL-D should also be considered. Periodontal disease was measured after starting hemodialysis, long after any potential biological pathway between exposure and outcome was established. As with any cohort study, residual confounding is possible; in particular, ORAL-D did not measure variables linked to dental care access, oral health literacy, early life socioeconomic status, or other health behaviors that may have explained any linkages between periodontal disease and longer life expectancy. Because ORAL-D did not include patients with earlier stages of chronic kidney disease, the possibility that periodontitis is associated with mortality in adults with milder kidney disease cannot be ruled out. Although the oral examination was done using the World Health Organization methods and standardized and discussed across countries using a shared operational protocol, variation in the oral examination in each country was possible. We did not conduct formal testing of inter-examiner agreement between centers, which may have confounded the diagnosis of periodontal disease between countries.
Acknowledgements
We thank all the participants for their involvement in the study. The following people contributed to the study and gave consent to be mentioned in this list:
Participating centers, facilitators and organizing committee.
Argentina: S Raña, M Serrano, S Claros, M Arias, L Petracci, M Arana, P De Rosa, A Gutierrez, M Simon, V Vergara, M Tosi, M Cernadas, I Vilamajó, D Gravac, M Paulón, L Penayo, G Carrizo, M Ghiani, G Perez, O Da Cruz, D Galarce, M Gravielle, E Vescovo, R Paparone, C Mato Mira, E Mojico, O Hermida, D Florio, M Yucoswky, W Labonia, D Rubio, G Di Napoli, A Fernandez, H Altman, J Rodriguez, S Serrano, G Valle, M Lobos, V Acosta, G Corpacci, M Jofre, L Gianoni, G Chiesura, M Capdevila, J Montenegro, J Bequi, J Dayer, A Gómez, C Calderón, E Abrego, C Cechín, J García, J Corral, M Natiello, A Coronel, M Muñiz, V Muñiz, A Bonelli, F Sanchez, S Maestre, S Olivera, M Camargo, V Avalos, E Geandet, M Canteli, A Escobar, E Sena, S Tirado, A Peñalba, G Neme, M Cisneros, R Oliszewski, V Nascar, M Daud, S Mansilla, A Paredes Álvarez, L Gamín, M Arijón, M Coombes, M Zapata. France: C Boriceanu, S Frantzen-Trendel, UFSBD 13.
Hungary: K Albert, I Csaszar, E Kiss, D Kosa, A Orosz, J Redl, L Kovacs, E Varga, M Szabo, K Magyar, G Kriza, E Zajko, A Bereczki, J Csikos, A Kuti, A Mike, K Steiner, E Nemeth, K Tolnai, A Toth, J Vinczene, Sz Szummer, E Tanyi, R Toth, M Szilvia. Italy: N Dambrosio, G Paparella, M Sambati, C Donatelli, F Pedone, VA Cagnazzo, R Antinoro, F Torsello, C Saturno, G Giannoccaro, S Maldera, E Boccia, M Mantuano, R Di Toro Mammarella, M Meconizzi, PF Steri, C Riccardi, A Flammini, L Moscardelli, M Murgo, N San Filippo, S Pagano, G Marino, G Montalto, S Cantarella, B Salamone, G Randazzo, D Rallo, A Maniscalco, M Fici, A Lupo, P Pellegrino, R Fichera, A D’Angelo, N Falsitta. Poland: E. Bocheńska-Nowacka, A. Jaroszyński, J. Drabik, M. Birecka, D. Daniewska, M. Drobisz, K. Doskocz, G. Wyrwicz. Portugal: L Inchaustegui, C Outerelo, D Sousa Mendes, A Mendes, J Lopes, J Barbas, C Madeira, A Fortes, R Vizinho, A Cortesão, E Almeida. Spain: A Bernat, B De la Torre, A Lopez, J Martín, G Cuesta, RM Rodriguez, F Ros, M Garcia, E Orero, E Ros.
Dentists.
Argentina: G Bava (Buenos Aires), A Bora (Caleta Olivia), H Gorena (Córdoba), T Calderón (Mendoza), R Dupuy, N Alonso (Tucumán), V Siciliano (Comodoro Rivadavia). France: S Frantzen-Trendel, UFSBD 13. Hungary: K Nagy, Ö Bajusz, I Pinke, G Decsi, L Gyergyoi, Zs Jobba, Zs Zalai, Á Zsedenyi, G Kiss, M Pinter, M Kereszturi (Faculty of Dentistry University of Szeged). Italy: M Petruzzi (Bari), M De Benedittis (Bari). Poland: J Szkutnik (Lublin), J Sieczkarek (Lublin).
Portugal: A Capelo (Lisboa), A Caetano, K MacGregor, M Santos, S Silva Pinheiro, L Martins, D Leitão, C Izidoro (independent dentists). Spain: M Garcia Gallart (independent dentist), C Mendieta (Barcelona).
Authors’ contributions
MR, SCP, GW, JCC, MP, MDB, PF, DWJ, MT, GFMS designed the sudy; MR, MP, MDB, PN, VS, EC, RG, ML, MT, PS, AB, JD, LF, DDC, AGB, JH, CW, LG acquired data; MR, SCP, GW, JCC, MP, MDB, PF, DWJ, MT, PN, VS, FP, EC, RG, ML, MT, PS, AB, JD, LF, DDC, SS, AGB, JH, CW, LG, CCB, GFMS analyzed, interpreted data and revised manuscript; SCP drafted the manuscript; statistical analysis were conducted by SCP, GW, FP, GFMS; GFMS, JH obtained funding; GFMS supervised the study. All authors gave final approval and agree to be accountable for all aspects of work.