Association Between Periodontitis and All-cause and Cancer Mortality

Retrospective Elderly Community Cohort Study

Ping-Chen Chung; Ta-Chien Chan

Disclosures

BMC Oral Health. 2020;20(168) 

In This Article

Results

Characteristics of the Study Population

In the baseline, 24,806 participants had periodontitis (30.05%). The mean age of the sample at baseline was 73.59 years, and the slight majority were males (52.15%). More participants had 7–14 years of schooling (43.12%), were not married and living together (72.95%), had not smoked in the past 6 months (90.86%), ate fruits and vegetables daily (77.41%) and were not diagnosed with diabetes (65.8%). The association of periodontal status with demographic and health behaviors is presented in Table 1. The following are significant based on p < 0.05. When comparing participants with healthy periodontium to participants with periodontitis, the latter were more likely to be male (32.98%), be illiterate (32.70%), have higher frequency of smoking (smoke daily 39.82%) and have no fruit and vegetable intake daily (33.10%). However, there was no obvious difference (p > 0.05) in the percentages between the healthy periodontium group and periodontitis group by marital status and diabetes diagnosis based on univariate logistic regression. At the end of study, the number of deaths was 11,160 participants, among which about 33.15% had periodontitis.

As of December 31, 2012, among 82,548 participants the average number of visits was 3.17 times (Figure 2). The maximum of visits was 8 times, involving 1666 participants (2.02%).

Figure 2.

Distribution of elderly health exam frequency at baseline

Association Between Periodontal Status and Risk of All-cause Mortality and All-cancer Mortality

At the midpoint of the study (1500 days), the survival probability of the periodontitis group was lower than that of the healthy periodontium group with regard to both all-cause mortality and all-cancer mortality (Figures 3, 4). Of the 82,548 participants, 7460 of 57,742 (12.9%) in the healthy periodontium group and 3700 of 24,806 (14.9%) in the periodontitis group died by the end of the study. The estimated rate of overall survival at 3000 days in the Kaplan–Meier analysis was 80.9% (95% CI, 80.1 to 81.8) in the periodontitis group and 82.3% (95% CI, 81.3 to 83.3) in the healthy periodontium group. There were significant differences in the rates of survival between the two groups (P < 0.001).

Figure 3.

Kaplan-Meier graph of time to all-cause mortality by periodontal status

Figure 4.

Kaplan-Meier graph of time to all-cancer mortality by periodontal status

Of the 82,548 participants, 2362 of 57,742 (4.1%) in the healthy periodontium group and 1153 of 24,806 (4.6%) in the periodontitis group died from cancer. The estimated rate of overall survival at 3000 days in the Kaplan–Meier analysis was 93.5% (95% CI, 92.9 to 94.1) in the periodontitis group and 94.2% (95% CI, 93.7 to 94.7) in the healthy periodontium group. There were significant differences in the rates of survival among the two groups (P = 0.004).

Table 2 shows the adjusted association of periodontitis with risk of all-cause mortality and all-cancer mortality in the baseline. After controlling for other covariates, participants with periodontitis had significantly higher hazard ratios (HRs) for all-cause mortality (HR = 1.077, 95% CI:1.027 to 1.130). A multivariate Cox proportional hazards model showed that being male (HR = 1.696, 95% CI:1.606 to 1.791), being elderly, and smoking (daily, HR = 1.253, 95% CI:1.126 to 1.394) were risk factors for all-cause mortality. Participants with a high education level (above 14 years of schooling, HR = 0.527, 95% CI: 0.480 to 0.579) had lower mortality. In regard to all-cancer mortality, after controlling for other covariates, hazard ratios (HRs) of all covariates had the same trend as that this result was not statistically significant for all-cancer mortality (HR = 1.036, 95% CI: 0.952 to 1.128).

Table 3 took annual health examinations results per participant into account. With regard to all-cause mortality and all-cancer mortality, there were significant associations with periodontitis (HR = 1.092, 95% CI: 1.038 to 1.149; HR = 1.114, 95% CI: 1.032 to 1.203) in multivariate Cox frailty models, after controlling for other covariates. Being male, having a low education level and being a smoker were risk factors for both all-cause mortality and all-cancer mortality when considering each visit.

Association Between Periodontal Status and Risk of Specific Cancer Mortality

Comparing mortality of lung cancer in the periodontitis group to the healthy periodontium group, the hazard ratio was 1.185 (95% CI: 1.027 to 1.368) after adjusting for age and sex in a multivariate Cox frailty model (Table S1). After adjusting for age and sex, the hazard ratio was 1.305 (95% CI: 0.856 to 1.989) for esophageal cancer, 1.019 (95% CI: 0.790 to 1.313) for pancreatic cancer, 0.960 (95% CI: 0.789 to 1.168) for liver and gallbladder cancer, 1.164 (95% CI: 0.952 to 1.423) for colorectal cancer, and 1.340 (95% CI: 1.019 to 1.762) for prostate cancer.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....