3.4.1. Cardiovascular Disease

Kaplan–Meier survival analyses have shown that cardiovascular disease-free survival rates in HD patients with moderate/severe periodontitis (defined as 2 or more teeth with at least 6 mm of inter-proximal attachment loss) were significantly worse (*p* = 0.01) compared to those with no/mild periodontitis [3]. In addition, a multivariate analysis model including age, center, sex, dialysis vintage, smoking status, cause of ESRD, DM, and hypertension demonstrated that moderate/severe periodontitis was an independent predictor of cardiovascular diseases in HD patients (hazard ratio = 5.0, 95% CI; 1.2–19.1, *p* = 0.02) compared to no/mild periodontitis [3]. However, the study also showed that periodontitis does not play a significant role in all-cause mortality among patients with HD (hazard ratio = 1.8 and 95% CI; 0.7–4.5) [3]. In contrast, other investigators showed that periodontal disease was significantly associated with risks of both cardiovascular-related and all-cause mortality [11]. Briefly, Kaplan–Meier survival curves showed that the cumulative survival rates in HD patients with severe periodontitis were significantly worse (*p* < 0.001) than in HD patients with no/mild or moderate periodontitis [11]. In addition, they found that the overall mortality rates in the no/mild (*n* = 104), moderate (*n* = 98), and severe periodontitis (*n* = 51) groups were 24.0%, 41.8%, and 70.6%, respectively, a statistically significant di fference (*p* < 0.001) [11]. Interestingly, this study also showed that in a multivariate analysis model including age, serum levels of albumin, hs-CRP, the Charlson Comorbidity index score, education level, and history of smoking, severe periodontitis was an independent risk factor for all-cause mortality (hazard ratio = 1.83, 95% CI; 1.04–3.24, *p* < 0.05), but not cardiovascular-related diseases (hazard ratio = 1.95, 95% CI; 0.90–4.23, *p* = 0.09) [11]. To summarize, while a prior study had demonstrated that periodontitis is closely associated with mortality from cardiovascular disease, but not with all-cause mortality, the latter study showed opposing results.

A contrasting opinion regarding how periodontitis a ffects the risk for all-cause and cardiovascular mortality in patients on HD has been raised [44]. In short, the risk of all-cause and cardiovascular disease morality in HD patients with moderate-severe periodontitis was lower (hazard ratio = 0.74, CI = 0.61–0.90 and 0.67, 0.51–0.88, respectively) compared to those with none/mild periodontitis [44]. Importantly, these analyses were performed in propensity-matched cohorts.

### 3.4.2. Metabolic Syndromes and Pneumonia

Poor oral heath was reported to be associated with metabolic syndrome in 312 HD patients [76]. Briefly, HD patients with metabolic syndrome (*n* = 145) had a higher score compared to those without metabolic syndrome (*n* = 108) in terms of PI (mean/SD; 2.23/0.05 vs. 2.03/0.06; *p* = 0.01), GI (1.63/0.07 vs. 1.33/0.08; *p* = 0.003), and PDI (4.35/0.10 vs. 3.84/0.14; *p* = 0.002) [76]. In addition, they found that metabolic syndrome was positively associated with the severity of periodontitis (*p* = 0.002), and multivariate analysis also showed that moderate or severe periodontitis was an independent risk factor for metabolic syndrome in HD patients (odds ratio; 2.74, 95% CI; 1.29–5.79, *p* = 0.008) [76]. Furthermore, other investigators have compared periodontal indices in HD patients with and without metabolic syndrome [77]. The authors found that bone resorption in the metabolic syndrome group (mean/SD: 1.99/0.39 mm) was significantly higher than in the non-metabolic syndrome group (1.45/0.12 mm) [77]. In addition, PPD showed significant differences between the metabolic syndrome and non-metabolic syndrome groups (mean/SD: 2.73/0.47 and 2.17/0.18, respectively; *p* < 0.05).

Conversely, the cumulative incidence of pneumonia mortality in HD patients with periodontal disease was found to be significantly higher than in HD patients without periodontal disease (*p* < 0.01) [41]. Interestingly, another report showed contrasting findings where intensive treatment of periodontal disease led to a reduced risk of acute and sub-acute pneumonia (hazard ratio; 0.77, 95% CI; 0.65–0.78, *p* < 0.001) in patients with HD [78]. Incidentally, this study also demonstrated that periodontal disease treatment in HD patients was associated with a lower risk of endocarditis (hazard ratio; 0.54, 95% CI; 0.35–0.84, *p* < 0.01) and osteomyelitis (hazard ratio; 0.77, 95% CI; 0.62–0.96, *p* < 0.05) [78]. Thus, periodontal disease is speculated to have played an important role in the pathogenesis and mortality due to metabolic syndrome and pneumonia in HD patients.

### *3.5. Diabetic and Non-Diabetic Nephropathy*

Currently, there is a general agreemen<sup>t</sup> that DM is a major cause of dialysis. In other words, many patients with dialysis have been receiving treatment for DM for a long time. Conversely, DM is considered to be one of the major causes of periodontal diseases [79,80]. It has been speculated that HD patients with DM have higher risk of periodontal diseases compared to those without DM. Unfortunately, few studies on the prevalence and severity of periodontal disease in HD patients with DM have been performed, and there is no systematic review on this issue to date.

### 3.5.1. Comparison of the Prevalence

In Japan, a cross-sectional study composed of HD patients with DM (*n* = 29), HD patients with chronic kidney nephropathy (CGN; *n* = 69), and a control group (*n* = 106) was performed. The study showed that mean/SD number of teeth present in HD patients with DM (17.9/9.8) was significantly lower (*p* < 0.05) compared to those with CGN (24.1/6.8) and the control group (25.3/5.8) [81]. In addition, the same study showed that the percentage of sites with bleeding on probing in HD patients with DM (22.2%) was significantly higher (*p* < 0.05) than in those with CGN (15.9%) and in controls (9.3%) [81]. The authors indicated that the reason for this difference in periodontal conditions between HD patients with DM and those with CGN were unclear; however, they speculated that various factors, such as healthy behavior, social, economic, and environmental status, and mental and systemic conditions, might influence the oral status in these patients [81]. Furthermore, the authors found that the SFR and total score of xerostomia in HD patients with DM or those with CGN were significantly different than those of the control group; however, these periodontitis-related parameters were similar between DM and CGN patients [81]. Thus, the oral health status in patients with renal dysfunction was worse than that of individuals with normal renal function, albeit with similarities observed between HD patients with DM or GCN. Interestingly, they showed that smoking status was significantly associated with the number of teeth in HD patients with DM. Smoking is a well-known risk factor for periodontitis and teeth loss, and it is also associated with development of DM. [82,83]. Thus, smoking played an important role, both directly and indirectly, in the oral health of HD patients with DM [81].

In contrast, the prevalence of moderate or severe periodontitis in HD patients with DM (74/76 patients = 97.4%) was reported to be similar to that of those without DM (51/53 ones = 96.2%), and the severity of periodontitis was not significantly associated with DM status (*p* = 0.71) [84]. In addition, this study showed that various periodontal findings, such as the PBI, PPD, CAL, and bleeding on probing, showed no significant differences in HD patients with and without DM (*p* = 0.72, 0.40, 0.58, and 0.79, respectively) [84]. Thus, the impact of diabetes on the prevalence and/or severity of periodontal disease in HD patients must be clarified by further investigations.

Interestingly, the prevalence of a variety of bacteria di ffered between HD patients with and without DM (*Capnocytophaga* species *p* = 0.02; *Eubacterium nodatum*, *p* = 0.02; *Parvimonas micra*, *p* = 0.03, *Porphyromonas gingivalis, p* = 0.02) [84]. However, the authors could not definitively conclude on the relationships between periodontitis and DM in patients with HD due to several limitations; for example, the plaque index was not assessed, the mean age of patients was di fferent, and the study was performed across multiple centers. However, they concluded that DM has no decisive influence on periodontal conditions in HD patients [84]. Nevertheless, we should consider that their study population included HD patients with well-controlled DM (mean/SD hemoglobin A1c level = 6.3/2.7).

### 3.5.2. Objective and Subjective Manifestations

The first comparative study of oral health, dental conditions, and periodontal status in HD patients with and without DM was reported in 2005 [85]. Regarding objective manifestations, the study showed that the percentage of HD patients with DM exhibiting uremic odor (12/43 patients; 27.9%) was significantly lower (*p* = 0.018) compared to those without DM (42/85 patients; 49.4%). However, a similar significant di fference was not found in tongue coating, mucosa petechia, or ecchymosis [85]. Conversely, other investigators found that all of these objective manifestations, including uremic odor, did not di ffer significantly between HD patients with (*n* = 46) and without DM (*n* = 54) [86]. Nevertheless, another report showed that uremic odor, tongue coating, and mucosa petechia were significantly di fferent between the two patient groups (*p* = 0.044, 0.026, and 0.008, respectively) [87].

Another report using a visual analogue scale to assess subjective symptoms, including dry mouth (xerostomia), taste change (dysgeusia), and tongue/mucosa pain, showed that HD patients with DM had significantly higher scores than those without DM (*p* = 0.040, 0.004, and 0.018, respectively) [85]. This was in contrast to findings indicating that DM status in HD patients was significantly associated with dysgeusia (*p* = 0.008), but not with dry mouth or mucosal pain [87]. Thus, these 2 studies showed that taste change (dysgeusia) in HD patients di ffered significantly among patients with and without DM; however, another report showed that the frequency of dysgeusia was similar [86]. In addition, no significant relationships with regard to diabetic status or xerostomia have been reported by other studies [81,87]. Thus, there is no general agreemen<sup>t</sup> concerning di fferences in objective and subjective manifestations of HD patients with and without DM.

### 3.5.3. The Community Periodontal Index and DMFT Index

As mentioned above, the DMFT index has been commonly used to determine the incidence of dental caries, while the community periodontal index (CPI) has been used to assess periodontal condition using a mouth mirror and a probe according to the World Health Organization criteria [88]. The DMFT index consists of 4 parameters (decayed, missing, fillings, and total teeth) and the CPI consists of 6 (healthy periodontium, bleeding on probing, calculus deposition, probing depth of 4 to 5 mm, probing depth of 6 mm or deeper, and 3 or more teeth missing).

We compared the DMFT index of HD patients with and without DM in Table 5. The overall DMFT score in HD patients, which was calculated as the sum of the number of decayed, missing, and filled teeth, was remarkably higher (*p* = 0.001) in DM patients (19.93/81.9) than in non-DM patients (14.26/9.19) [85]. This finding was supported by other investigators [87]. However, the overall DMFT score was not associated with DM status in HD patients (Table 5) [83,85]. Nevertheless, as shown in Table 4, one report showed that the DM status was significantly associated with each index item, while others showed opposing results [84,85,87].


**Table 5.** Decayed, missing, and filled teeth (DMFT) index in hemodialysis patients with and without diabetes.

↑ meansthatvariablesinthediabeticgroup were highercomparedtothenon-diabeticgroup.

Evaluating the CPI in HP patients, Chuang et al. [85] reported that there was a borderline significant di fference in patients with diabetic and non-diabetic nephropathy (*p* = 0.055). Murali et al. [86] also found no significant association between DM status and CPI in HD patients. These results were obtained by using the chi-square test to compare diabetic nephropathy and CPI codes. In contrast, a di fferent study reported that one CPI variable, probing depth of 6 mm or deeper, in the DM patients receiving HD group was significantly higher (*p* = 0.015) than in non-DM patients, whereas other codes, such as calculus deposition and probing depth of 4 to 5 mm, were not [87]. Thus, with the exception of one code (probing depth 6 mm or deeper), a significant di fference in CPI scores was not found in previous studies. However, in contrast to this report, another study reported that the percentage of bleeding on probing and sites in HD patients with DM was significantly higher than in non-DM patients (mean/SD%: 13.3/22.2 versus 8.2/15.9; *p* < 0.05) [81]. However, we must note that this result was obtained using the Tukey HSD test, which is a multiple comparison test, to compare among patients on HD for diabetic nephropathy and chronic glomerulonephritis and healthy controls.

### 3.5.4. Properties of Saliva

Regarding SFRs in patients with HD, unstimulated and stimulated salivary flow, and bu ffering capacity of stimulated saliva have been reported to have no significant relationship with diabetic status (*p* = 0.15, 0.20, and 1.0, respectively) [83], which was also supported by another study [81]. However, Sung et al. [89] reported that the unstimulated whole SFR in 68 diabetic HD patients was significantly lower (*p* = 0.032) than that in 116 non-diabetic HD patients.

The salivary pH level in DM patients (mean/SD; 7.97/0.67) showed a trend (*p* = 0.063) towards being lower than that of patients without DM (8.22/0.44) [85]. Similarly, another study showed that the frequency of salivary pH >7.0 in the DM patients was lower compared to the non-DM patients (17.0% versus 34.0%, respectively; *p* = 0.056); however, the frequency of salivary pH <7.0 in both patient groups was similar (DM; 36.2% and non-DM; 34.0%; *p* = 0.823) [87]. Chi-square analysis showed that the relationship between diabetic status and salivary pH level was not significant (*p* = 0.623) [87]. In contrast, Schmalz et al. [84] reported an opposing finding, where the mean/SD salivary pH level of HD patients with DM (6.7/0.7) was significantly lower than that of non-DM patients (7.0/0.9; *p* < 0.01) [84].

As shown in Table 6, the changes in the SFR and pH level according to diabetic status are unclear. While the precise reasons underlying this discrepancy are unknown, di fferences in patients' backgrounds, history of chronic renal failure and HD, duration of DM, and hemoglobin A1c level may be factors. Conversely, an investigation into the relationships between periodontal condition-related parameters and glycemic control in HD patients with DM showed that tongue coating, dry mouth, and tongue/mucosal pain were significantly correlated with hemoglobin A1c levels (*p* = 0.001, 0.001, and 0.004, respectively); however, salivary pH levels were decreased with higher hemoglobin A1c levels (mean/SD pH level in hemoglobin A1c level ≤6%; 8.20/0.36, 6.1%–9%; 8.00/0.58, ≥9.1%; 7.46/1.07), but the di fferences were not statistically significant (*p* = 0.086) [85]. Therefore, we emphasized the importance of more detailed investigations with a larger study population to arrive at a conclusion.


**Table 6.** Comparison of properties of saliva in diabetic and non-diabetic patients.

DM; diabetic mellitus, Ref; Reference.

### *3.6. Periodontal Indices and Salivary Status with Peritoneal Dialysis*

In the above sections, we discussed the changes in periodontal indices and salivary findings in response to HD. However, several reports have shown di fferences in these parameters between PD and HD patients, and controls (Table 7). For example, Bayraktar et al. [90]. reported that plaque and calculus accumulation in the HD and PD groups were higher than in controls. However, gingival inflammation in PD patients was significantly lower than in HD patients [22,68,90]. Moreover, the SFR in the PD group was significantly higher than that in the HD group, although both groups had significantly lower values than the control group [68]. The reduction in the SFR has been known to increase the risk of caries [91]. In fact, the authors found that the number of filled teeth was higher in the PD group than in the HD group. Di fferences in the SFR might be associated with better volume status in the PD group and in the relatively more liberal fluid intake because of residual renal function. In fact, the fluid dynamics of the gingiva might influence gingival health in children undergoing PD therapy [92]. We have summarized these results in Table 6. These results seem to show that the pattern of di fferences in periodontal indices and salivary status between healthy controls and patients with PD was not replicated between HD and PD patients. However, the data was insu fficient to draw definitive conclusions.


**Table 7.** Periodontal parameters in the peritoneal dialysis, hemodialysis, and healthy groups.

GI; gingival index, PBI; papillary bleeding index, PI; plaque index, CSI; calculus surface index, S-; salivary-, SFR; salivary flow rate, PPD; probing pocket depth, NS: not significant, Ref; reference.

### **4. Management of Periodontal Disease**

Many investigators have suggested that correct diagnosis and appropriate treatment of periodontal disease are important not only to the improvement of oral infection and inflammation, but also to maintain systemic health in patients receiving dialysis, and that managing oral health can e ffectively prolong survival [15–17,23,41,61,70]. In fact, brushing the teeth twice daily led to a reduced chance of developing periodontal diseases than brushing once daily, and it was identified as an independent factor [36]. Furthermore, a lower frequency of visits to the dentist has also been reported to be positively associated with higher mortality in HD patients [37]. Conversely, this report also showed that although the mortality of HD patients with chronic periodontitis is worse than of those without, the survival rate of patients treated for chronic periodontitis was not significantly di fferent from that of untreated

patients (*p* = 0.774) [37]. Furthermore, Hou et al. emphasized that special e fforts for the prevention and managemen<sup>t</sup> of periodontal disease are important in elderly HD patients because aging is another risk factor for periodontal disease in HD patients [36]. In addition to the prevention of di fferent complex diseases and the prolongation of survival, maintenance of oral health and treatment of periodontal disease are essential for dialysis patients waiting for renal transplantation because patients with active inflammation and/or severe periodontal disease are usually judged as unfit for transplantation.

Thus, many studies support the importance of maintenance of dental health and periodontal treatment in patients with dialysis. However, in the real world, approximately 70% of Japanese HD facilities have no registered dental clinic [93]. Consequently, collaboration with dental facilities was promoted as beneficial for maintenance and managemen<sup>t</sup> of oral health in HD patients [93], and found support from other investigators [53]. In addition, education on preventive dental care is important to the collaboration with dentists [94]. The prevalence and severity of periodontal disease are reported to vary substantially according to country, rather than being rotted in individual patient characteristics or healthcare [75].

Thus, conscientious e fforts are necessary to establish an e ffective managemen<sup>t</sup> and/or treatment approach to oral health in patients receiving dialysis treatment. In addition, this system must be modified for di fferent countries in accordance with their specific conditions, including causes of ESRD, complications, and lifestyle habits. However, we believe that such a managemen<sup>t</sup> approach in collaboration with dentists is well worth implementing because managemen<sup>t</sup> of oral health and periodontal disease leads to maintenance or improvement in the QoL, reduction of complications, and prolongation of survival in dialysis patients.

### **5. Conclusions and Future Perspectives**

This review showed that periodontal diseases a ffect the inflammation, immune response, and nutritional status in patients on dialysis. In fact, the severity of periodontal disease was significantly associated with serum levels of CRP, albumin, and a variety of minerals. In addition, several inflammation-related cytokines and molecules, such as IL-6, Il-8, TNFα, and PTX-3 were influenced by periodontal conditions. As a result, a significant association between periodontal disease and various pathological conditions, including cardiovascular disease, metabolic syndrome, and pneumonia, is observed. Furthermore, we showed how dialysis, especially HD, exacerbates oral conditions via disruption of salivary characteristics, such as pH and flow rate. The deterioration in oral and periodontal tissues subsequently leads to a high frequency and severity of periodontal disease. Importantly, DM plays important roles in these pathological mechanisms. These findings are summarized in Figure 1. Furthermore, based on these facts, we demonstrated that maintenance of oral health and treatment of periodontal disease are important to maintaining QoL, prevention of various pathological conditions, and prolongation of survival in patients with dialysis. Unfortunately, we found that it was di fficult to conduct a focused and systematic study into the relationships between periodontal disease and dialysis-related factors. Although numerous studies have been performed, they exhibit significant heterogeneity in patient characteristics, such as age, diabetic nephropathy, duration of chronic kidney disease; and methods of dialysis, such as PD or HD, the specific machine and agents used, timing of sampling, and duration. Data on the pathological roles of periodontal disease in patients with PD in particular is insu fficient for a systematic review. However, continuous technological development could enable us to identify new pathogens, determine their interactions, and assess periodontal disease-related parameters in patients on dialysis. In fact, we have developed a research strategy aimed at elucidating the molecular and immune-related mechanisms of periodontal disease in patients on dialysis using novel approaches [95,96]. We emphasized that further studies with larger populations and uniform design are required to determine the pathological significance of periodontal disease and the clinical utility of oral care in patients on dialysis.

**Figure 1.** Schema of pathological roles played by periodontal disease in patients on hemodialysis.

**Funding:** This research was funded in part by a Grant-in-Aid from the Japan Society for the Promotion of Science (to Yasuyoshi Miyata; No. 16K15690).

**Acknowledgments:** We would like to thank Editage (www.editage.jp) for English language editing.

**Conflicts of Interest:** The authors declare no conflicts of interest.
