Chronic periodontitis and community-acquired pneumonia: a population-based cohort study

Periodontal disease is a chronic inflammatory disease of the periodontal tissue that affects approximately 20 to 50% of the world’s population [1]. It is caused by colonization of periapical periodontal pathogens that cause destruction of the ligaments and alveolar bone supporting the teeth [2, 3]. Chronic periodontitis (CP) results in tooth loss due to an advanced inflammatory form of periodontal disease caused by microorganisms [4]. Research has suggested that periodontal disease is a risk factor for systemic disease, and studies have shown that periodontitis is associated with an overall increased risk of mortality [5, 6].

Pneumonia is one of the most common serious infections and causes significant morbidity and mortality in both healthy and vulnerable individuals [7]. Pneumonia is divided into two categories: hospital-acquired pneumonia (HAP) and community-acquired pneumonia (CAP). HAP refers to pneumonia that develops 48 h after admission or within 10 days after discharge from a hospital without incubation [8], and CAP refers to an acute infection of the lungs in individuals who have not been recently hospitalized and are not regularly exposed to the healthcare system [9]. According to data from the Korea National Statistical Office, pneumonia was the 10th leading cause of death in 2004, but ranked 4th after cancer, cardiovascular disease, and cerebrovascular disease in 2015 [10]. The main causes of mortality, which increased significantly from 10 years ago, were pneumonia (22.9 persons, 246.8%), heart disease (17.1 persons, 41.5%), and lung cancer (6.5 persons, 22.5%) [10]. Numerous researchers have shown an association between periodontitis and respiratory disease such as pneumonia [11‐15].

Most previous studies on periodontal disease and pneumonia were case-control or cross-sectional studies and did not involve the general population, just elderly or vulnerable populations such as patients who have HAP or ventilator-associated pneumonia. Additionally, these studies were limited to HAP and not CAP. Moreover, the relationship between periodontal disease and CAP has yet to be established by longitudinal studies. This study aimed to analyze the relationship between CAP and CP using the Korean National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) data.

A total of 363,541 participants were followed for an average of 7.6 years (standard deviation, 1.3), resulting in 2,782,453 person-years. Table 2 shows the baseline characteristics of all study participants according to the severity of periodontal disease. Among 363,541 participants, 30.1% (n = 109,322) had CP. The number of CAP cases during the follow-up period was 14,838 (4.1%). The mean age of the participants was 57.8 years, and the mean age of the severe CP group was significantly higher than that of the Non-CP group. Table 3 shows the multivariable association results from Cox proportional hazard regression analyses between CP and CAP after adjustment for age, sex, household income, smoking status, alcohol consumption, physical exercise, CCI, BMI, fasting serum glucose level, and total cholesterol level. In those with CP, the severe CP group had the highest number of cases (n = 2798), while the moderate group (n = 568). The hazard ratio (HR) for CAP was not significant in any of the models regardless of the presence or absence of CP. Table 4 shows the multivariable association results from Cox proportional hazard regression analyses between severe CP and CAP after adjustment for age, sex, household income, smoking status, alcohol consumption, physical exercise, CCI, BMI, fasting serum glucose level, and total cholesterol level. There were no significant associations in any model (Model 4 HR, 1.00; 95% confidence interval, 0.96–1.04). Table 5 shows subgroup analysis of the association between CP and the risk of pneumonia. The results in all subgroups stratified by sex and BMI were similar to the main results. There was also no significant difference in smoking status, which is commonly accepted as a strong risk factor for CAP.

As a prospective study on a large Korean adult population over a long follow-up period, this study has advantages. To the best of our knowledge, no cohort study has previously examined the association between admission due to pneumonia and CP.

CP was not found to significantly affect CAP, even after adjustment for important health characteristics. Our results were consistent with those of Brown’s study, which reported that CAP was not related to dental biofilms [7]. The main causative agents of CAP are Streptococcus pneumoniae and Haemophilus influenzae. CAP is also caused by the spread of viral bacterial pathogens such as Mycoplasma pneumoniae, Chlamydia pneumoniae, and Legionella pneumophila. Porphyromonas gingivalis and Treponema denticola, which are the bacteria involved in CP [20], and are not typically CAP-related bacteria.

Many studies have investigated periodontal disease as a risk factor for various types of pneumonia, including aspiration pneumonia, HAP, and ventilator-associated pneumonia [13, 21‐26]. Aspiration of colonized pathogens has been proposed to be an important risk factor for pneumonia [11, 27‐29]. de Melo Neto et al. demonstrated that moderate and severe CP were associated with CAP [30]; however, the study had a small sample size (140 patients) and was conducted for only 17 months. Moreover, the control group consisted of hospitalized patients and did not involve the general population. In contrast, this cohort study was the result of an 8-year follow-up of patients hospitalized with CAP according to the presence of CP, with the general population at baseline at relatively low risk of disease compared to the subjects of de Melo Neto’s study. This could explain the difference in our results and those of previous studies.

Several hypotheses have been posited to explain the likelihood of developing periodontal disease and several types of pneumonia. Periodontal disease with periodontal pockets promotes accumulation of dental plaque, which can promote the growth and reproduction of pathogenic bacteria. Various pathogenic bacteria have been found in patients with deep periodontal pockets [31, 32]. The association between periodontal disease and pneumonia may be due to colonization by pathogenic bacteria in the periodontal pocket, as inhalation of a pathogen is considered a risk factor for pneumonia [33, 34]. Oral pathogens can be aspirated into the lower airways, which results in favorable conditions for the development of pneumonia [35, 36]. Excessive production of inflammatory cytokines induced by the major pathogens of periodontal disease plays a significant role in pneumonia [37]. The enzymes in saliva on the surface of the oral mucosa of patients with periodontal disease may facilitate the adhesion of respiratory disease pathogens [38]. However, it is difficult to clarify the cause of CAP, unlike those in other types of pneumonia. Most studies that identified the causes of CAP were conducted at tertiary referral hospitals, which may not represent the general population. Despite considerable efforts, rarely can the cause of CAP be clearly determined, and more rigorous investigations are needed [9, 39‐41].

Few studies have been conducted on oral health and pneumonia when distinguishing between CAP and aspiration pneumonia. CAP and aspiration pneumonia share common risk factors such as diabetes mellitus, malnutrition, alcohol consumption, smoking status, and aging; however, the major risk factors for CAP are chronic obstructive pulmonary disease, heart disease, chronic bronchitis, functional impairment, chronic renal failure, cancer, and human immunodeficiency virus [27, 28, 42]. The major risk factors for aspiration pneumonia are poor oral health, sputum suctioning, use of antipsychotic drugs, deterioration of swallowing function, dehydration, and dementia [43, 44]. Therefore, CAP and aspiration pneumonia involve different risk factors and underlying diseases. For this reason, elderly individuals are more susceptible to pneumonia, and the causes and prevention methods of aspiration pneumonia and CAP are different [45]. In the subgroup analysis (Table 5), important CAP risk factors, such as smoking status and CCI, were stratified. However, consistent with the main results of this study, no statistically significant relationships could be identified.

There were several limitations to this study, mostly stemming from the use of claims data based on ICD-10 and treatment codes. First, the definition of CP based on treatment codes could underestimate the actual number of CP cases, especially in the mild and moderate periodontal disease groups. Several people seldom visit dental clinics or hospital because of lack of awareness regarding oral health and economic reasons. However, codes associated with severe CP treatment such as tooth extraction and periodontal surgery can be reliably used as identifying variables because severe CP is typically more painful for patients than mild or moderate CP. Therefore, we dichotomized the 4 groups of CP in Table 3 into severe CP and non-severe CP, and the results were not altered (Table 4). A previous study in Taiwan also used ICD and treatment codes to classify CP severity [46]. Second, this study defined CP not by clinical attachment loss, but by treatment code. Clinical attachment loss is often used to diagnose periodontal disease, and this information was not available in the database of this study. However, periodontitis is diagnosed based on dental examinations, including periodontal examination, probing depth analysis, and radiographic checkups, which are covered by national insurance. Similarly, inadequate clinical information, such as chest radiographs, blood samples, and pulmonary gas exchange data, is a major limitation when defining CAP [47]. To evaluate the causality between CP and CAP, pneumonia diagnosis needs to be supported by specific pathogen information to assess the type and severity, so further studies involving specific laboratory data will be particularly informative on this topic. To overcome the limitations associated with these data, we attempted to avoid overestimation by setting strict definitions of the main exposure and outcome. Lastly, it is possible that the new operational definition of CP is inaccurate. As the severity of CP increases, it would be logical for the percentage of subjects to decrease, but the severe CP group had the largest number of subjects. However, in a recent study that classified the stages of periodontal disease, subjects who their had teeth removed were at high risk of periodontal disease [48]. Based on this study, subjects with extractions were also classified into the severe CP group in the present study. Although treatment procedure codes are not a formal basis for determining the severity of CP, we aimed to establish a new operational definition based on input from dental specialists.

This study suggested that CP is not a potential risk factor for CAP. The etiology and causality between periodontal disease and CAP remain unclear, and little is known about the underlying mechanism. In future studies, collection of clinical and epidemiological evidence is needed using a prospective study design to better understand the relationship between oral health and CAP. This study can be a starting point for discussion of the relationship between CP and CAP.