STRENGTHS AND LIMITATIONS OF THIS STUDY

The study compares antibodies to four anaerobic oral bacteria of self-reported non-malignant and incidence of malignant respiratory diseases.

The study population is a randomised sample of participants of a large cohort of men.

The dissemination of oral bacteria is a probable cause or association with chronic bronchitis/emphysema.

This study analyses four oral bacteria, but the oral microbiome may include other relevant bacteria.

The study presents self-reported non-malignant diseases at the time of screening simultaneously with cases of cancer in prospective analyses of 17.5 years follow-up.

Introduction

Several diseases, such as pneumonia, chronic obstructive pulmonary disease (COPD) and asthma, have been related to the oral microbiome.1 The natural barrier of cilia of the epithelium, together with secreted mucus, acts as a protective function by moving and expelling mucus, bacteria, viruses and minor irritants in an outward motility movement.2

The oral cavity functions as a reservoir for potential lung pathogens and common oral bacteria have been identified in infectious pneumonia.3 4 To this can be added, that alveolar bone loss in periodontitis has been associated with and seen as a predictive factor for COPD in a prospective study.5

Treatment of oral infections, including periodontitis, and oral health preventive measures have shown to be beneficial for lung function both by reducing exacerbations of COPD and slowing lung function decline.6–13 The associations between periodontal disease and asthma, COPD, and pneumonia were further supported in a recent systematic review.14 There is also evidence that persons with asthma have more gingivitis and periodontal disease.15 16

In chronic periodontitis, particular interest has been devoted to the so-called red complex of periodontal bacteria. This consists of the strict anaerobic Gram-negative rods Tannerella forsythia (TF), Treponema denticola (TD) and Porphyromonas gingivalis (PG). The characteristics of the four periodontal bacteria and their main microbiological mechanisms for a breakdown of the tooth-supporting structures of the mandible and maxilla are well described.17 We have previously shown that low levels of antibodies to TF are associated with cardiovascular disease mortality in men with a history of myocardial infarction (MI).18 Further, low levels of antibodies to TF are associated with bladder cancer, and TD to be associated with bladder cancer and colon cancer.19 Low antibody level in the presence of a microbiological infection is indicative of a hampered immunological response. The primary aim of this study is to test whether low antibody levels of four major periodontal bacteria were associated with self-reported asthma and chronic bronchitis in a cross-sectional setting and prospectively with cancer of the airways. The secondary aim is to assess the association between smoking and the antibody levels of these four periodontal bacteria.

MethodsStudy population

This study includes a randomised age-stratified sample of 621 men with no former history of cancer that were available from a population-based study carried out in 2000 in Oslo, Norway, also called the Oslo-II study.20 The participants in 2000 had previously been invited to take part in the Oslo study 1972/1973. In all, 12 764 men were invited, 6530 attended and 5323 men participated in both health surveys. The participants answered questionnaires on general health issues, underwent anthropometric and blood pressure measurements and had blood samples taken. One tube of EDTA/full blood and the remaining serum after planned risk factor analyses were stored at −80°C for the future analyses. During the period 2004–2005, analyses for antibodies to four oral bacteria (1172) and high sensitive C reactive protein (hs-CRP) (n=5323) were performed. The study on antibodies included 548 men with a self-reported history of MI and a random sample of controls from the remaining cohort (n=624). This study sample consisted of these randomised controls and a randomised sample (n=73) among the men with an MI history totalling 697. Excluding the men with a history of cancer yielded 621 men available for analysis (figure 1). The cancer incidence follow-up was from 1 July 2000 to 31 December 2017. Thus, the analyses use screening information from 2000 in cross-sectional analyses for asthma and bronchitis/emphysema and prospective follow-up analyses over 17.5 years follow-up on cancer incidence for case identification. The SPSS (version 25) random generator was used.

Figure 1

Flow chart of the antibody substudy of the Oslo II study.

Serum analyses

For this analysis, hs-CRP and antibodies to the four periodontal bacteria were used. The bacteria were the red complex bacteria; the strict anaerobic Gram-negative rods TF, TD and PG. Also, the very common facultative anaerobic Aggregatibacter actinomycetemcomitans (AA) was included for comparison with the red complex. The laboratory procedures have previously been described.20

Statistics

The outcomes were either the self-reported diseases asthma or bronchitis/emphysema, or cancer incidence in a 17.5 years of follow-up. Cancer information was provided by the Norwegian Cancer Registry. The antibody levels are described by the mean, median and SD of each bacterium of the chosen outcomes. The number and percentages for the cases per quartile of the bacteria and their distribution were tested by the Pearson χ2 test for the significant differences between cases and non-cases. Analyses excluding versus including cancer incident cases prior to the health screening in 2000 were performed. The incident cases prior to 2000 are excluded from the results presented. Differences between daily cigarette smokers and non-smokers of all included in the study and the four different outcomes were also performed. However, results from the four different outcomes yielded few or no cases in some quartiles. These analyses are not presented. The results were tested by the Pearson χ2 test (χ2 p value). The significant level was p≤0.05.

Results

Variations in antibody levels are presented in table 1 for the four bacteria and the outcomes of asthma, bronchitis/emphysema, cancer incidence of bronchi and lungs combined and all cancers of the respiratory system. By comparing asthma versus bronchitis/emphysema, mean antibody levels were higher for TF, TD and PG, but lower for AA. For all cancers of the respiratory system, mean antibody levels were higher for TF and TD but lower for PG and AA compared with cancers of bronchi and lungs. For all diagnoses, PG levels were the highest and TD the lowest. SDs were the highest for PG.

Table 1

Antibody values from serum ELISA measurements by self-reported conditions in 2000 and cancer incidence by relevant ICD-10* codes for the respiratory system, the Oslo II study

The distribution of cases across the quartiles is presented in tables 2 and 3 in the order of TF, TD, PG and AA. Comparing cases to non-cases showed statistically significant low levels of TD antibodies for bronchitis/emphysema (p=0.035). Comparative analyses with or without incident cancer cases prior to 2000 did not greatly alter these results. A number of incident cancer cases prior to the screening in 2000 were for asthma 6, bronchitis/emphysema 5, lung and bronchitis cancer 4, all cancers of the respiratory system 4 and the total study 79 cases.

Table 2

Distribution of individuals per quartiles of antibody values for self-reported conditions, asthma, chronic bronchitis/emphysema versus the rest of the study population

Table 3

Distribution of individuals per quartiles of antibody values for cancer and respiratory system cancers versus the rest of the study population

Daily smoking versus non-smoking was examined as shown in table 4. TF (p=0.019) and TD (p≤0.001) were significant for increased low antibody levels. Stratified analyses of the four disease outcomes separately were of low power as the number of cases in some quartiles was 0 (see online supplemental file l). In asthma, there was a different distribution (p=0.049) between smokers and non-smokers for AA. In bronchitis/emphysema, there were 0 cases in Q1 and Q4 for non-smokers for TF (trend over quartiles p=0.061). For TD, there were 0 cases in Q2, with trend p=0.041. For bronchi/lung cancers, non-smokers, 0 cases for Q4 in TF, for TD in Q3, non-smoker, for PG in Q1, smoker and for AA in Q4, non-smoker. In respiratory cancers for PG 0 cases in Q1, smokers (borderline p=0.053).

Table 4

Daily smokers versus non-smokers; antibody levels by quartile values of the study cohort, n=621

The participants changed their smoking habits between the first and second health screenings. In the 1972/1973 health examination, 44.2% reported that they were daily smokers in contrast to the next health screening in 2000 when 17.4% reported to be daily smokers.21

Discussion

The risk by low quartile level of antibodies to the periodontal bacterium TD was found to be increased in bronchitis/emphysema compared with the remaining participants of the study cohort. Low levels of TF and TD were observed in daily smokers compared with non-smokers. A trend of increased risk by low antibody levels was apparent.

These results on respiratory diseases such as cancer, asthma and bronchitis/emphysema support previous findings on the importance of low levels of antibodies both for increased risk of mortality among men with heart disease and men with specific cancers such as colon and bladder cancers. Several bacteria are involved in the development of chronic and aggressive periodontitis. They are, with increasing levels of anaerobe conditions, grouped into yellow, green, purple, orange and red complexes.17 22–24 The formation of an oral biofilm is an ongoing dynamic process on the teeth, the gingiva, the gingival crevices, the buccal surfaces and the tongue. Oral bacteria also spread down the pharynx and larynx and into the bronchi and lungs by aspiration and can pass past pharyngeal cilia and overcome the secreted mucus from the epithelial lining of the airway tube. Their described enzymatic activity makes the bacteria able to break down tissue distant to the mouth in addition to oral infections.25 This is a process closely linked to the immune system.

Brodin and Davis have explored the immune system’s development by heritable factors and the influence of symbiotic and pathogenic microbes.26 The mechanisms behind the variation observed are becoming clearer. This knowledge will help the understanding of the immune-mediated and infectious diseases that so often occur and the underlying mechanisms and hence the ability to prevent, modify and treat infections. A greater understanding is necessary as several studies have linked oral diseases to different systemic diseases.27 28 An association between lung cancer and periodontal and other oral bacteria has been studied in the Atherosclerosis Risk in Communities Study, which investigated IgG antibodies to 18 oral bacteria in 1287 participants.29 In a 17.5-year follow-up study, the authors found bacteria of the orange complex to be positively associated with lung cancer. The orange complex includes bacteria that are moderately pathogenic for chronic periodontitis.17 Lung cancer risk and periodontal disease were also the issues in the study on women, the Women’s Health Initiative Observational Cohort. They observed that the risk of total cancer increased in older women whether they were smokers or non-smokers.30

A limitation of the study is that no clinical measurements were taken at the screening and the Oslo II study included men only. Although this is a random sample of the full study, the number of cases is not high. The ELISA analyses provided exact measurements of systemic antibody levels and the results are reliable. As lung disorders are frequently linked to the adverse effects of smoking, comparative analyses between smokers and non-smokers were performed for each antibody for all. However, previous prospective follow-up analyses were adjusted for smoking status and the predictivity of low antibody levels remained.18 19

Conclusion

These major periodontal bacteria are anaerobes (strict/facultative) that survive and produce enzymes that break down teeth-supportive tissues in the oral cavity. Our studies indicate that they migrate as reflected by the systemic antibody levels and may cause negative tissue effects at distant sites from the oral cavity with serious consequences for disease development.

Data availability statement

Data may be available on request. The data are based on the study data and registries, and permission is dependent on their acceptance.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by the Norwegian Data Inspectorate and The Regional Committee for Medical and Health Research Ethics approved (REK), reference no. DT ref. 03/111 and granted permission. The Declaration of Helsinki for medical research involving persons was followed. Participants gave informed consent to participate in the study before taking part.

Acknowledgments

The study was planned jointly but carried out ahead of the Oslo Health Study 2000 in collaboration with the National Health Screening Service of Norway (now part of the Norwegian Institute for Public Health), the City of Oslo, the University of Oslo, and the Ullevål University Hospital (now Oslo University Hospital, Oslo). Professor Per Nafstad, University of Oslo, is thanked for his contribution to the hs-CRP analyses and former Department Director Per Everhard Schwarze, Norwegian Institute for Public Health, for his contribution to the ELISA analyses.