Non-communicable diseases (NCDs) are rising in prevalence because of aging, unhealthy diet and sedentary lifestyle, and common NCDs are caries and periodontitis (here oral diseases) and cardiovascular disease (CVD). Association between oral diseases and CVD has been observed in epidemiological studies, and suggested mechanisms include transfer of oral pathogen bacteria and pro-inflammatory mediators to other organs triggering immune response and systemic inflammation. Circulating mediators may initiate a response in the liver with production of C-reactive protein, serum amyloid A and increased tryptophan degradation, that contribute to CV inflammation and atherosclerosis. We have demonstrated that higher serum levels of these markers are associated with presence of hypertension and obesity and with higher risk of CVD. However, CVD and oral diseases also share many of the same risk factors, including hypertension, obesity, diabetes and smoking. Since the association of oral disease with CV risk factors mostly has been demonstrated in cross-sectional studies, the direction has not been ascertained. These CVD risk factors lead to changes in the heart and arteries (preclinical CVD); more harmful in women than men. If these conditions can impact development of oral diseases is not been investigated in large studies.

Socio-economic inequalities have been reported for oral diseases and are linked to low socio-economic status (SES).

It is unknown whether CV health and SES in midlife may impact prevalence of oral diseases later in life. It is also unknown if the oral microbiome differ by periodontitis severity and can be associated with inflammatory biomarkers, CV risk factors and preclinical CVD. The project will be performed in adults approaching the retirement age in Vestland county. We will combine their data from the ongoing Hordaland Oral Health Survey with their data from three surveys in the longitudinal Hordaland Health Study performed in the period 1992-2020.

Title:

Oral health in adults: predictors, social gradients and correlation with cardiovascular health in women and men. Date: 24.02.2023

1. Excellence 1.1 State of the art, knowledge needs and project objectives Non-communicable diseases (NCD) are rising in prevalence globally because of aging, refined diets and sedentary lifestyle. Two of the most common NCDs in humans, caries and periodontitis (here named oral diseases), contribute substantially to years lived with disability in age-standardized prevalence rates worldwide. Cardiovascular (CV) disease (CVD) is the leading cause of death and years lived with disability globally. Co-presence of CVD and oral diseases is therefore common, and an association between oral diseases and CVD has been observed in epidemiological studies.

Mechanisms linking oral diseases to CVD include transfer of oral pathogen bacteria and pro-inflammatory mediators to other organs triggering immune response and systemic and vascular inflammation. Circulating inflammatory cytokines may initiate an acute-phase response in the liver with production of C-reactive protein and serum amyloid A and increased tryptophan degradation, that contribute to CV inflammation and atherosclerosis. We have demonstrated that higher serum levels of these inflammatory markers are associated with presence of hypertension and obesity in midlife in the Hordaland Health Study, and with higher risk of subsequent CVD independent of clinical risk factors. However, CVD and oral diseases also share many of the same risk factors, including hypertension, obesity, diabetes, and smoking. Since the association of oral disease with CV risk factors so far mostly has been demonstrated in cross-sectional studies, the direction of these associations has not been ascertained. These CVD risk factors lead to chronic CV inflammation which promotes atherosclerosis as well as structural and functional changes in the heart and arteries (preclinical CVD) and are all more harmful in women than men. Whether these conditions can impact development of oral diseases is not fully explored.

Socio-economic inequalities have been reported for many oral diseases. In a study in subjects older than 65 yrs from Northern Norway, severe periodontitis was more prevalent in men, smokers and subjects with low socio-economic status (SES). Thus, sex and SES may modify the association between oral diseases and CVD. It is presently unknown whether CV health in midlife, measured by CV risk factor burden, circulating inflammatory markers and SES, may impact presence and severity of oral disease in women and men later in life. It is also unknown if the structural and functional oral microbiome differ by periodontitis severity and can be associated with inflammatory biomarkers, CV risk factors and presence or absence of preclinical CVD. Finally, the association between oral diseases and preclinical CVD has to date not been explored in large studies.

The current project will explore the interface between odontology, microbiology, medicine and social science in a cohort of community dwelling adults approaching the retirement age (born in 1950-1951), in Hordaland County (now Vestland). We will combine data from the ongoing Hordaland Oral Health Survey (HUSK-T) with their data from three surveys in the longitudinal Hordaland Health Study (HUSK) performed in the period 1992-2020. A total of 2254 participants were invited to HUSK-T which was completed in 2022 with a participation rate of 63%. The project objectives will focus on the interaction of oral diseases with CV health:

1. Describe oral diseases in the cohort and their relation to quality of life in women and men.
2. Investigate the structural and functional oral microbiome in relation to periodontitis severity, CV health and systemic inflammatory biomarkers in women and men.
3. Investigate whether CV risk factors, inflammatory biomarkers and SES in midlife predict oral disease later in life in women and men.
4. Explore the associations between oral disease prevalence and severity and preclinical CVD in women and men.

1.2 Research questions and hypotheses, theoretical approach and methodology WP1A Oral health and quality of life We will describe oral health status as periodontal health or disease, tooth loss and caries experience (DMFT= decayed+missed+filled teeth) among participants in HUSK-T and test the hypothesis that oral health related quality of life is associated with oral health status in women and men.

Task 1.1 Description of oral diseases Periodontal disease Periodontitis is a common chronic inflammatory disease characterized by bone resorption, deepening of the pocket and bleeding upon probing. It is a multifactorial disease affected by interaction between subgingival microbiota, host response and environmental modifying factors. A dysbiosis in the dental biofilm results in activation of an inflammatory host response followed by tissue destruction. The tissue destruction is largely irreversible and therefore cumulative over lifetime. Social gradients are reported among periodontitis patients and they are age and area of residence dependent. Sex differences in this context have not been explored in Norway.

Methods: Periodontal examination in HUSK-T includes assessment of bleeding on probing (BOP), periodontal probing depth (PPD), and gingival recession (GR) at six sites per tooth for all teeth. PPD is registered as the distance from the marginal gingiva to the probable base of the pocket and GR as the distance from enamel cement junction (CEJ) to marginal gingiva.

Clinical attachment level (CAL) is calculated adding GR and PPD. Periodontitis is considered present if interdental CAL 1-2 mm is present at ≥ 2 non-adjacent teeth, or buccal/oral CAL ≥3 mm with PPD > 3 mm at ≥ 2 teeth. Severity is grouped as mild, moderate and severe. Periodontal health in a reduced periodontium is clinically defined as < 10% bleeding sites with no bleeding pockets > 3 mm.

Dental caries and endodontic treatment In HUSK-T, caries is diagnosed based on clinical and radiologic findings and root-filled teeth detected in the panoramic radiograph. Methods: Caries is registered on tooth surfaces and classified as decayed (d) from 1-5. For root caries; only active lesions are registered.

Tooth loss: as part of the oral health status registration, missed teeth are calculated as the total number minus number of lost teeth. Periodontitis and caries are the two most important causes of loss of teeth. Number of teeth as an outcome is strongly associated to income in older adults.

Task 1.2. Self-reported oral health Little is known about how oral conditions impact the quality of life of older women and men in Norway. Some studies have reported associations between tooth loss and oral health related quality of life (OHRQoL), but the association to periodontal disease is less documented. Methods: Self-reported data related to the conceptual model of Gilbert et al. (17) (i.e. OHRQoL, self-perceived oral health, symptoms, access to dental care services and socio-demographic factors) are collected by questionnaires. For measurement of OHRQoL, the eight item oral impacts on daily performances inventory (OIDP) has proven appropriate as a discriminative and descriptive measure in cross-sectional studies. We previously (19) used the OIDP scale in a national sample of adults and found it to be reliable and valid in the Norwegian context. All eight items are included in our questionnaire and OIDP frequency SC score will be performed.

WP1B Structural and functional oral microbiome in periodontitis and relation to CV health We will test the hypotheses that 1) The microbiome composition in subgingival plaque vary between participants with severe, moderate, and mild periodontitis and can be linked to the prevalence of hexa-acylated lipo-polysaccharide (LPS) producing bacteria; 2) The microbiome composition varies with presence of CV risk factors or/and preclinical CVD in women and men.

Task 1.3 We will investigate if the structural and functional oral microbiome differ by periodontitis severity and Task.1.4 can be associated to inflammatory biomarkers in the kynurenine-tryptophan pathway, CV risk factors (see WP2) and presence or absence of preclinical CVD (see WP3) in women and men.

The oral bacteria community composition in health has been found to be rather stable over time. Research has suggested that the transition from health to disease may be attributed to a shift in the global balance of the microbial flora rather than to the appearance of individual bacteria. Porphyromonas gingivalis, a bacterium known to be important for human periodontal disease, has been detected in atherosclerotic lesions, where they may act as pro-atherogenic stimuli. Further, atherosclerotic susceptible mice exposed to P. gingivalis developed advanced atherosclerotic lesions. It has been reported that low gut (21) microbiome diversity correlates with higher arterial stiffness in woman, but overall there are very few studies that have explored the association between microbiome composition, CV risk factors and preclinical CVD. The pro-inflammatory function of the oral microbiome in periodontitis in such setting has never been investigated. The lipid A component of LPS in the cell wall of all gram-negative bacteria trigger the immune response in the host, and five lipid chains has 100-fold less inflammatory activity than the lipid A with six lipid chains. Using the well-conserved lipid A structure, the bacteria can be classified based on which lipid A coding gene they carry. No previous studies have linked Lipid A functions to periodontal disease and severity, nor to systemic inflammation, CV risk factors or preclinical CVD outcomes.

Methods: Bacterial composition will be characterized in subgingival plaque samples from participants with mild, moderate, and severe periodontitis and from healthy participants (e.g. 200 participants in each periodontitis severity group and 400 healthy participants). For bacterial quantification we will use qPCR for bacterial DNA yield, 16S rRNA Illumina HiSeq sequencing and reduced metagenome sequencing (RMS) for bacterial determination. To assign taxonomy we will use the Human Oral Microbiome Database (www.homd.org).The reduced metagenome sequences can be applied in the functional annotation model to identify genes that code for the different lipid As. The output data can then be imported into any statistical framework.

WP2 A CV risk factors in midlife: impact on oral health status in later life. We will test the hypothesis that CV risk factors and markers of immune system activation measured in midlife predict oral health later in life, and whether these associations differ between sexes.

Task 2.1 Prospectively explore the association of CV risk factors (HUSK-2 and -3) with presence of oral diseases (HUSK/T) and test if these associations differ between sexes.

Associations between oral diseases like periodontitis, caries and tooth loss and CV risk factors have been suggested from smaller studies and meta-analyses (24, 25), but many of them are based on self-reported information. Periodontitis, caries, hypertension and diabetes also share some common risk factors, including obesity. Thus, CV risk factor burden may influence oral health. Prospective studies, using objective measures for CV risk factors and oral health later in life are missing and warranted. Sex differences in prevalence of CV risk factors and CVD are well described, but it remains unknown whether sex differences exist in the association between oral and CV health.

Task 2.2 Explore underlying inflammatory pathways of the association of CV risk factors with prevalence and severity of oral diseases test if the strength of these associations differs between sexes.

Little is known about underlying pathways explaining the associations of oral health with CV risk factors and CVD. Periodontitis has been postulated to promote systemic inflammation through activation of the sympathetic nervous system and of immune cells. A small study in 30 women demonstrated that reduction in serum neopterin, a marker of macrophage activation, paralleled improvement of periodontitis disease (26). Periodontitis is associated with increased circulatory levels of C reactive protein and pro-inflammatory cytokines like Interleukin 1β, interleukin-6 and Tumor Necrosis Factor-α, which all stimulates the first and rate-limiting step of the kynurenic pathway of tryptophan degradation. Tryptophan is an essential amino acid, mainly catabolized (90%) through the kynurenine pathway. Increased conversion of tryptophan to kynurenine results in higher levels of the kynurenine-tryptophan ratio, which is a marker of immune system activation. Previous reports from HUSK have documented that higher serum levels of neopterin and kynurenine-tryptophan ratio both predicted increased risk of CVD. Recently, calprotectin and serum amyloid A have been suggested as biomarkers of periodontitis severity. We will explore whether these inflammatory biomarkers also prospectively predict oral diseases independent of CV risk factor burden.

WP2B Socioeconomic status (SES) and oral health We hypothesize that SES indicators in midlife determine oral health status later in life and can be modified by sex Task 2.3. Investigate the association between SES indicators earlier in life (HUSK-1 and 2) and oral diseases (HUSK-T) later in life in women and men.

Data from the Norwegian Institute of Public Health has documented that people with low education use dental services less often than their affluent counterparts, and financial constraints were identified as the main factor explaining these differences (report 2016/28). As expenses related to dental treatment in adults are only in part reimbursed in Norway, and oral health deteriorates with age, the financial burden related to oral disease increases in older adults. Despite an universal health coverage in Norway, clear SES gradients have been reported with regard to CVD and CV risk factors (https://www.fhi.no/en/op/hin/groups/social-inequalities/). A cross-sectional study conducted in Tromsø, found that social indicators (education, income and urbanisation) were associated with oral disease status, but the modifying role of social network and marital status in the association between SES and oral health was not investigated. Furthermore, longitudinal studies which explore whether SES in midlife is associated with oral disease 20 years later and modified by sex, has never been performed.

Materials and Methods We will use data from the HUSK-1-3 and from HUSK-T. An overview of information collected in HUSK-1-3 surveys and used in this WP is presented in Table 1 below.

CV risk factors Hypertension is defined as SBP ≥140 mmHg and/or DBP ≥90 mmHg measured in accordance with guidelines or use of antihypertensive drug therapy (6). Height and weight were measured, and body mass index (BMI) was calculated. Waist/hip circumferences were measured using a standardized method. Diabetes is identified as HbA1c≥48mmol/mol or use of antidiabetic drug therapy. Information about medical history, smoking habits and drug therapy is collected from questionnaires. Non-fasting blood samples were analyzed for serum lipids, glucose, HbA1c and creatinine. Serum cotinine will be measured to objectively quantitate smoking status.

Measurement of circulating inflammation biomarkers Serum neopterin, tryptophan, kynurenines, high-sensitive C-reactive protein, calprotectin and serum amyloid A will be measured (at the BEVITAL academic laboratory (www.bevital.no)) in biobank samples from the HUSK-3 survey. These have all been measured in the HUSK-2, allowing assessment of prospective as well as longitudinal performance.

WP3 Oral health: Association with preclinical CVD We will test the hypothesis that oral disease is associated with presence of preclinical CVD independent of sex, CVD risk factors and socio-economic factors.

Task 3.1 Explore the associations of oral disease prevalence and severity with preclinical CVD in women and men.

Prevalence of preclinical CVD increases when several CV risk factors cluster and may be regarded as a measure of CV risk factor burden (29). Common types of preclinical CVD include coronary atherosclerosis, arterial stiffness and left ventricular hypertrophy. The two former are more prevalent in men at all ages, but the latter is more common in women. Sex-specific immune activation has been implicated in development of preclinical CVD in obesity and hypertension. Thus, systemic inflammation is important in development of preclinical CVD. Oral diseases may promote preclinical CVD through systemic vascular inflammation. However, larger clinical studies in humans are lacking in this field.

We will do the following: compare oral disease prevalence and severity in women and men with and without preclinical CVD in HUSK-T and HUSK-3, respectively. Subgroups with higher prevalence and more severe oral disease will be identified in sex-specific analysis.

Task 3.2 Explore whether inflammatory biomarkers of preclinical CVD may classify oral disease in women and men The analysis will focus on circulating inflammatory markers with documented association with CVD and periodontitis in previous research (see WP2). We will explore whether high-sensitive C-reactive protein, serum neopterin, kynurenine/tryptophan ratio, calprotectin and serum amyloid A levels differ between groups with combined oral diseases and preclinical CVD compared to those with either or neither of these disorders in sex-specific analyses combining data from the HUSK-3 and HUSK-T. Analyses will be adjusted for CV risk factors and SES factors documented as predictors of oral diseases in WP1-2.

Methods Preclinical CVD will be assessed by validated prognostic markers following the European guidelines. Ultrasound of the heart (echocardiography) is used for detection of left ventricular hypertrophy by sex-specific cut-off values. Digital ultrasound images from HUSK-3 are post-processed at the Bergen core laboratory for echocardiography. Appropriate intra- and inter-observer variation assessments will be performed. Arterial stiffness is assessed by the state-of-the-art method (carotid-femoral pulse wave velocity) using aplanation tonometry. Coronary atherosclerosis is assessed by coronary computed tomography (CT) angiography calcium score, and presence of obstructive and non-obstructive coronary artery disease following the American Heart Association recommendations.

Inclusion Criteria:

Participated in 3 Health surveys + one oral health survey

-

A cohort of community dwelling adults approaching the retirement age (born in 1950-1951), in Hordaland County (now Vestland).