The association between chronic periodontal and atherosclerotic cardiovascular diseases


Maria Retzepi

Clinical Lecturer


Francesco D’Aiuto

Senior Clinical Lecturer, Periodontology Unit, Clinical Research Division, UCL Eastman Dental Institute, London, UK



Over the last two decades, the association between periodontal diseases and atherosclerotic cardiovascular diseases (CVDs) has been intensively investigated. The scope of the present review is to (a) discuss the epidemiological evidence for the correlation between CVD or surrogate CVD markers and periodontal diseases, (b) synthesize our current understanding of the plausible molecular mechanisms underlying the pathophysiology of the relation between CVD and periodontitis, and (c) present and discuss the available intervention studies on the association between CVDs and periodontal diseases.

The available preclinical and clinical evidence indicates that untreated moderate to severe periodontitis is associated with endothelial dysfunction and with an independent increase in the risk for subclinical atherosclerosis and atherosclerotic CVD, possibly via an increased systemic infectious and inflammatory burden in susceptible individuals. Further, evidence emerging from randomized controlled intervention studies indicates that periodontal treatment may attenuate systemic infl ammation and endothelial dysfunction in a dose-dependent manner in systemically healthy individuals with moderate to severe generalized periodontitis. However, at present, there is no evidence to support a direct effect of periodontal therapy on the incidence of CVD. Future research should focus on the pathophysiological mechanisms underlying the effect of CVDs on periodontal diseases and on the performance of properly designed clinical trials. This direction will allow further investigation into the role of confounding factors and the effect of periodontal therapy on CVD incidence.

Analecta Periodontologica 2009; 20:55-69

Cardiovascular diseases (CVDs) comprise a variety of conditions including ischemia, atherosclerosis, peripheral artery disease, infective endocarditis, and acute myocardial infarction. Myocardial infarctions, stroke, and thromboembolic events result from atherosclerosis, often combined with a superimposed coronary thrombosis (Karchmer 1997). Atherosclerosis is considered to occur as a response to injury of the vascular endothelium in the context of an inflammatory process (Libby et al. 2002). Periodontitis is a chronic inflammatory disease resulting in the destruction of the supporting tissues of the teeth. Infection and inflammation have both been implicated in the pathology of various chronic diseases, including atherosclerotic CVD and periodontitis (Roivainen et al. 2000, Mattila et al. 2005). During the last two decades, research into the association between periodontal diseases and CVDs has evolved dramatically. The scope of the present review of the literature is to provide a better understanding of the link between atherosclerotic CVD and periodontitis with a focus on more recent publications.


Epidemiological studies on periodontitis and CVD association

Epidemiological studies on the association between periodontitis and CVD have reported varying results. The initial data linking dental infections and CVD were provided by case-control studies (Mattila et al. 1989), whereas subsequent cohort studies demonstrated an independent modest association between periodontitis and CVD (Scannapieco et al. 2003). A significant association between acute myocardial infarction and periodontal status based on clinical measures of probing pocket depths and clinical attachment loss has been reported (Cueto et al. 2005). On the other hand, several studies have reported a weak or no association between periodontitis and CVD (Nakib et al. 2004, Shimazaki et al. 2004, Beck et al. 2005, Spahr et al. 2006). Interestingly, studies on alveolar bone loss have reported higher odds ratios (ORs) for the association between periodontitis and CVD compared with studies that measured probing pocket depth and clinical attachment level (Persson et al. 2003, Engebretson et al. 2005, Geismar et al. 2006). This observation may be attributed to the fact that alveolar bone loss constitutes a more representative assessment of the cumulative effect of periodontal disease on systemic health compared with clinical periodontal measures, i.e. bleeding on probing, probing pocket depth, and clinical attachment levels (Beck et al. 2005). The alternative explanation is that alveolar bone loss is closely linked to age, which in turn is also associated with periodontal measures. Furthermore, the heterogeneity in prevalence rates of periodontitis among different study populations poses additional challenges when researchers evaluate the likelihood of an association between periodontitis and CVD. Therefore, differential periodontitis measures and definitions in epidemiological research render the assessment of the associations between periodontitis and CVD difficult (Kinane and Bouchard 2008).

Two recent meta-analyses supported the evidence of a positive although weak association between periodontitis and CVD (Bahekar et al. 2007, Mustapha et al. 2007). Bahekar et al. (2007) included five prospective cohort studies and reported that periodontal patients presented with a 1.14 times higher relative risk (RR 1.14, 95% CI 1.01-1.2) for developing coronary heart disease compared with healthy individuals. In this meta-analysis, the odds ratio for the association in cross-sectional studies was higher (OR 1.6, 95% CI 1.3-1.9). Similarly, other meta-analyses of prospective and retrospective studies confirmed that periodontal disease may slightly increase the risk for CVD (Meurman et al. 2003, Mustapha et al. 2007).


Common risk factors between periodontitis and CVD

Periodontitis and atherosclerotic CVD share a series of common risk factors, which may serve as confounders when researchers evaluate the association between the two diseases. Common risk factors between periodontitis and CVD include age, gender, socioeconomic status, smoking, metabolic factors, and stress.



Age is an important risk factor associated with both periodontal disease and CVD (Holm-Pedersen et al. 2006, Oyama et al. 2008). An association between age and dental conditions in relation to stroke has been reported (Lee et al. 2006). In addition, Persson et al. (2002) reported that approximately 50% of subjects older than 60 years of age had periodontal disease, whereas approximately 55% had either a diagnosis of atherosclerosis or a history of stroke, hypertension, or acute coronary syndrome. The role of aging as a potentially confounding factor in the link between periodontal disease and CVD will always be present, as it will be very difficult to control for; therefore, researchers should interpret the pertinent data with greater caution.



The role of gender as an effect modifier in the association between periodontitis and risk for CVD is unclear at present, as some investigators have presented evidence of gender-based differences in the association between periodontal disease and CVD (Desvarieux et al. 2004, Buhlin et al. 2005), whereas other clinicians have failed to do so (Andriankaja et al. 2007).


Socioeconomic status

Although lower socioecononomic status has been associated with both higher prevalence of CVD and periodontal disease (Borrell et al. 2006, Chaix et al. 2007), most studies evaluating the relationship between socioeconomic status and periodontitis in the context of CVD development have reported an attenuation of the degree of association between these two factors.



Large epidemiological case-control studies have demonstrated that the association between periodontitis and CVD is present independently from any confounding effects related to smoking (Holmlund et al. 2006, Andriankaja et al. 2007). However, the statistical adjustment of cigarette smoking has often not been conducted in a proper manner. Indeed, a continuous variable of exposure to cigarette smoking (number of cigarettes or pack/years) is a more appropriate measure when reseachers analyze the association between periodontitis and CVD.


Metabolic factors

An association has been demonstrated between high body mass index (BMI) and CVD (Balkau et al. 2007), as well as between high BMI and periodontitis (DallaVecchia et al. 2005, Linden et al. 2007). However, the occurence of a periodontal condition in women has been associated with coronary heart disease after adjusting for BMI (Buhlin et al. 2005). Cross-sectional epidemiological studies in various populations worldwide have associated periodontitis with increased levels of total cholesterol, low-density lipoprotein (LDL), triglycerides, and blood glucose, as well as with reduced levels of high-density lipoprotein (HDL) (Cutler et al. 1999, Katz et al. 2002, Buhlin et al. 2003, Craig et al. 2003, Morita et al. 2004, Losche et al. 2005, Nibali et al. 2007).

Recently, it has also become apparent that a moderate association exists between periodontitis and metabolic syndrome (D’Aiuto et al. 2008). This syndrome represents a cluster of cardiometabolic factors, including dyslipidemia, obesity, insulin resistance, and hypertension, and has been associated with an almost doubled future risk of developing CVD or Type 2 diabetes. Future research on the association between periodontitis and metabolic syndrome should include intervention trials to test whether improvements in oral health lead to reductions in the risk of metabolic syndrome and vice versa.



Stress has been associated with elevated risk for CVDs, possibly owing to impaired cortisol response (Taylor et al. 2006a). In addition, stress has been correlated with periodontal disease severity (Genco et al. 1999, Wimmer et al. 2002, Ng and Keung Leung 2006) and with poorer treatment outcomes (Elter et al. 2002). Stress may therefore constitute a risk factor for both CVD and periodontal disease. Further studies are required to evaluate the impact of stress on the relationship between periodontitis and CVD.

In conclusion, the relationship between periodontal disease and CVD is complicated by a series of shared risk factors, including smoking, age, socioeconomic status, metabolic factors, and psychosocial stress. Although the available epidemiological evidence currently shows a significant but modest relationship between periodontitis and CVD, there is an urgent need for further controlled prospective studies with large sample sizes to explore the extent and true nature of the association between periodontitisdiseases and CVDs.


Epidemiological studies on periodontitis and surrogate CVD marker associations

Several non-invasive examination methods and surrogate markers of future CVD exist, including computed tomography of the coronary arteries, ultrasound of the carotid arteries, magnetic resonance imaging, ankle-brachial index, flow-mediated dilation (FMD) of the brachial artery, pulse wave form analysis, microalbuminuria, and several serum biomarkers such as HDL, LDL, cholesterol, triglyceride, fibrinogen, and high-sensitivity C-reactive protein (CRP). However, these surrogate endpoints of future CVD present with variable predictive value for the atherosclerotic phase of plaque rupture and are not recognized as definitive diagnostic markers of CVD. As such, they cannot fully substitute for studies of clinical CVD endpoints (Jacobs and Crow 2007).


Endothelial dysfunction

Endothelial dysfunction precedes clinical manifestations of atherosclerosis, as it is considered to be the first inflammatory change of the vascular endothelium, ultimately leading to atherosclerosis (Pellegrino et al. 2005). In an early case-control study, Amar et al. (2003) reported that periodontitis is associated with endothelial dysfunction as accessed via FMD of the brachial artery.


Subclinical carotid atherosclerosis

Increased carotid artery intimal medial thickness (IMT) is highly correlated with an elevated risk for acute myocardial infarction in healthy subjects and constitutes a well-validated index of subclinical atheroma (Simon et al. 2002). Clinical measures of periodontal disease and the presence of periodontal pathogens have been associated with subclinical atherosclerosis assessed by IMT (Beck et al. 2001, Desvarieux et al. 2005). More specifically, data from the Atherosclerosis Risk in Community (ARIC) study indicated that subjects with severe generalized periodontitis had a significantly higher risk (OR 1.3, 95% CI1.03-1.66) of having a carotid IMT ≥1mm. A subsequent report from the same research group indicated an association between carotid atherosclerosis and elevated IgG titers to Campylobacter rectus and Peptostreptococcus micros (Beck et al. 2005).


Serum biomarkers

A series of association studies have indicated a negative impact of periodontitis on CVD surrogates, including interleukin-6 (IL-6), CRP, white blood cell (WBC) counts, and plasminogen factors. At present, these biomarkers are considered nonspecifi c indicators of a chronic infl ammatory response (Kinane and Bouchard 2008, Persson and Persson 2008). IL-6 is a proinflammatory cytokine that stimulates an immune response to tissue damage and is considered a potential risk marker of future CVD (Giannessi et al. 2007, Woodward et al. 2007). Several studies have reported elevated serum IL-6 levels in untreated periodontal patients (Loos et al. 2000, Ide et al. 2004).

CRP serves as a biomarker because high CRP levels present a predictive value for future acute myocardial infarction and unstable angina pectoris. The kinetics of high-sensitivity CRP is particularly useful in monitoring disease progression and the treatment effectiveness for diseases eliciting a systemic inflammatory response (Tsimikas et al. 2006). Interestingly, the CVD risk seems to be highest among patients who present evidence of chronic low-grade infection in combination with elevated CRP levels (D’Aiuto et al. 2004c). The presence of periopathogenic bacteria, i.e. Porphyromonas gingivalis, Prevotella intermedia, C. rectus and Tannerella forsythia, in subgingival samples has been also associated with elevated CRP levels (Noack et al. 2001). Paraskevas et al. (2008) performed a meta-analysis of case-control studies, demonstrating that periodontal patients present with 1.7 mg/l (95% CI 1.1 2.2) higher serum CRP concentrations compared with periodontally healthy controls. This calculated mean weighted difference is considered clinically significant, as it may shift patients between different classes of CRP-associated CVD risk.

Elevated WBC counts are present in patients with acute coronary syndrome (Avramakis et al. 2007). Furthermore, it has been reported that systemically healthy periodontal patients may present with higher serum WBC counts than those found in periodontally healthy controls (Persson et al. 2005, Buhlin et al. 2005, Renvert et al. 2006).

Based on the role of plasminogen activator inhibitor-1 (PAI-1) in fibrinolysis, increased PAI-1 concentrations are considered independent markers for the development of atherosclerosis and major adverse cardiac events (Marcucci et al. 2006). Elevated levels of plasminogen activator-1 (PA-1) have been reported in periodontal patients (Montebugnoli et al. 2005, Bizzarro et al. 2007), and therefore PA-1 may be implicated in higher risk for CVD through impaired fibrinolysis and thrombosis.


Infection and immune response

A series of studies have provided evidence that both acute and chronic infections of either viral or bacterial origin are linked to CVD. However, consensus is lacking for the direct role of infection in acute coronary syndrome and atherosclerosis (Mattila et al. 2005). Periodontal pathogens can invade the epithelium (Papapanou et al. 2004), and, furthermore, thepresence of periodontal pathogens has been linked to the development of infectious complications, i.e. brain abscesses, pulmonary infections, and endovascular infections (De Soyza et al. 2000, Ewald et al. 2006). Bacteremia may constitute a mechanism linking periodontitis and CVD if one takes into consideration that 80% of patients present with positive bacterial cultures immediately following subgingival debridement (Forner et al. 2006, Lafaurie et al. 2007). However, other reports have indicated a low incidence of bacteremia following periodontal treatment (Hartzell et al. 2005, Kinane et al. 2005).

Several studies have reported the presence of DNA from periodontal pathogens and also of viable periopathogens in samples from atheromatic lesions in the aorta and heart valves (Haraszthy et al. 2000, Beck et al. 2005, Kozarov et al. 2005). However, the mere existence of various microbes in human atherosclerotic lesions does not necessarily prove that they have induced the lesions. Recent systematic reviews and meta-analyses of epidemiological studies have indicated that elevated serum antibody titers to periodontal pathogens are linked to higher CVD risk (Meurman et al. 2003, Mustapha et al. 2007). Interestingly, elevated antibody titers to Aggregatibacter actinomycetemcomitans have been specifically correlated with higher prevalence and future incidence of coronary heart disease (Pussinen et al. 2004, Beck et al. 2005, Pussinen et al. 2005, 2007). In addition, high IgA-class antibody levels to P. gingivalis increased the risk for myocardial infarction by 300% (OR 4.0, 95% CI 1.2-13.1) in a Finnish population with exceptionally high prevalence of CVD (Pussinen et al. 2004).

Evidence from in vivo studies supports the hypothesis of a direct role of infection in the atherogenesis process. More specifically, Li et al. (2002) indicated that repeated intravenous administration of P. gingivalis accelerated the progression of atherosclerosis in the proximal aortas of apolipoprotein E (apoE)-null mice. Similarly, oral and anal application of P. gingivalis led to the formation of early atherosclerotic lesions in apoE-null mice (Lalla et al. 2003, Gibson et al. 2004). In addition, the severity of P. gingivalis-induced periodontitis was positively associated with enhanced lipid accumulation in the aortas of New Zealand white rabbits (Jain et al. 2003).


Increased levels of systemic inflammation

Inflammation is considered to integrally participate in all the stages characterizing the development of atherosclerotic CVD, although its precise role as a direct, causative factor in chronic atherogenesis remains largely unknown. Acute infections infl uence lipoprotein levels and therefore promote the atherogenesis process, owing to the acute phase response (Khovidhunkit et al. 2000). However, the impact of chronic low-grade infections, such as periodontitis, is relatively unclear. Chemokines and cytokines, including IL-6, IL-8, IL-10, IL-18 , tumor necrosis factor-α (TNF-α), and monocyte chemoattractant protein-1 (MCP-1) are frequently present in abnormal levels in patients with acute coronary syndromes (Armstrong et al. 2006).

As discussed earlier, epidemiological studies have demonstrated an association between moderate to severe periodontitis and systemic inflammatory and immune markers, including elevated blood levels of IL-6, CRP, PA-1, and WBC counts. These associations suggest that periodontitis elicits low-grade systemic infl ammation. In addition, a recent study has established an association between peripheral arterial disease and periodontitis with concomitantly significant increased serum levels of IL-6 and TNF-α (Chen et al. 2008).


Intervention studies

The epidemiological data support the presence of an association between CVD and periodontitis, whereas animal studies have confirmed the biological plausibility of the link between the two diseases. However, randomized controlled intervention studies are necessary to document the causative nature of this association and to exclude the contribution of undetected confounders (Tonetti 2009).

At present, there are no available trials evaluating the effect of periodontal therapy on the incidence of CVD. The Periodontitis and Vascular Events (PAVE) clinical study was a pilot study designed to assess the impact of periodontal treatment in reducing patients’ risk for a second myocardial infarction (Beck et al. 2008, Couper et al. 2008). This study failed to demonstrate a significant effect of routine nonsurgical periodontal treatment in reducing the future risk of serious CVD events. It should be noted, however, that, according to the design of the study, intervention was provided when both CVD and periodontal disease were established and as such, the observed lack of an attenuating effect on the future risk for a cardiac event should be regarded with caution (Offenbacher et al. 2009, Tonetti 2009).

Several intervention trials have provided modest evidence indicating that periodontal therapy reduces the systemic levels of inflammatory mediators (Glurich et al. 2002, D’Aiuto et al. 2004a, b, Pussinen et al. 2004, Tonetti et al. 2007) and improves endothelial function (Mercanoglu et al. 2004, Piconi et al. 2009), although there is considerable heterogeneity among patients (Behle et al. 2009).



Intervention studies have reported reduced (D’Aiuto et al. 2004a, b) or unaltered (Ide et al. 2004, Yamazaki et al. 2005, Elter et al. 2006) IL-6 levels following periodontal treatment. Interestingly, D’Aiuto et al. (2004a, b) revealed a dose-dependent relationship of the magnitude of IL-6 decrease and the clinical outcome of periodontal treatment, which included oral hygiene instructions, scaling and root planing and extraction of teeth with poor prognosis. In addition, Tonetti et al. (2007) reported that development of an acute short-term systemic inflammatory response was observed 24 hours following intensive periodontal treatment, as indicated by elevated IL-6 levels.


C-reactive protein

Treatment of periodontitis has been shown to reduce CRP levels (Glurich et al. 2002, Mattila et al. 2002). More recent studies have provided evidence that a sharp increase of CRP levels occurs within 24 hours of periodontal therapy (D’Aiuto et al. 2005, 2006) and that this increase may be sustained over 6 months following periodontal intervention (Tonetti et al. 2007). A recent meta-analysis reported a weighted mean reduction in serum CRP of 0.5 mg (95% CI 0.08-0.93) following periodontal treatment, which further supports the notion that periodontitis contributes to the systemic inflammatory burden (Paraskevas et al. 2008). The clinical significance of these findings is underlined by the significance of CRP as a predictor of the development of atherosclerosis and myocardial infarction.


WBC counts

Intervention studies have reported a decrease in WBC counts following periodontal therapy in patients with aggressive periodontitis (Dietrich et al. 2002).


Plasminogen activators

Intervention studies have reported a short-term increase in PA-1 values following periodontal treatment of systemically healthy periodontal patients, which was no longer evident 6 months following completion of treatment (Tonetti et al. 2007). In contrast, Taylor et al. (2006b) reported a decrease in PA-1 levels following tooth eradication in medically compromised individuals.



Pussinen et al. (2004) reported increases in serum HDL cholesterol concentration, HDL2/HDL3 ratio, and HDL phospholipid content and sphingomyelin/phosphatidylcholine ratio following periodontal treatment. The same research group also reported that, in periodontal patients who presented a reduction in CRP values following periodontal treatment and who were positive for A. actinomycetemcomitans, periodontal treatment promoted the antiatherogenic pathway of cholesterol efflux from macrophages (Pussinen et al. 2004).


Surrogate endpoints of CVD

Initial pilot studies reported that periodontal treatment can improve brachial artery fl ow rate in systemically healthy periodontal patients (Mercanoglu et al. 2004, Seinost et al. 2005, Elter et al. 2006). A subsequent study reported that, following intensive periodontal therapy in systemically healthy subjects with severe generalized periodontitis, a biphasic change of the FMD values of the brachial artery occured. More specifically, the authors reported a decrease in FMD values during the first day post-treatment, indicating an initial worsening followed by a dose-dependent significant improvement of the endothelial dysfunction compared with the baseline and with the control group (Tonetti et al. 2007). In a recent pilot study, periodontal treatment in systemically healthy patients with mild or moderate periodontitis led to a significant decrease in carotid IMT values following observation periods of 6 and 12 months (Piconi et al. 2009).


Recommendations for clinical practice

On the basis of the available evidence from preclinical, epidemiological, and intervention trials, the Consensus Report of the 6th European Workshop in Periodontology (Kinane and Bouchard 2008) and the Editors’ Consensus Report of the American Journal of Cardiology and the Journal of Periodontology (Friedewald et al. 2009) have introduced a series of recommendations for clinical practice, which may be summarized as follows:

  • The clinician should be aware of the evidence linking CVD and periodontal disease and of the clinical ramifi cations.
  • Patients with moderate to severe periodontal disease should be informed of a potential increased risk for atherosclerotic CVD.
  • Periodontal treatment can reduce the levels of systemic markers of infl ammation and improve endothelial function. However, no prospective periodontitis intervention studies have evaluated CVD outcomes.
  • Periodontal patients presenting with at least one known major risk factor for atherosclerotic CVD, i.e. smoking, immediate family history, or dyslipidemia, should be advised to consider a medical evaluation if one has not been performed during the last 12 months.
  • In the case of periodontal patients with atheroscle-rotic CVD, close collaboration between the perio-dontist and the physician is advocated to optimize CVD risk reduction and periodontal care provision.



Over the last two decades, the study of the impact of periodontal disease on CVDs has been a field of intensive research interest. Epidemiological studies have indicated that untreated moderate to severe periodontitis is associated with endothelial dysfunction and with an independent increase in the risk for subclinical atherosclerosis and atherosclerotic CVD via increased systemic infectious and inflammatory burden in susceptible individuals. More specifically, the systemic levels of IL-6, PA-1, CRP, and WBC counts are elevated in the presence of periodontal disease. In addition, evidence emerging from randomized controlled intervention studies indicates that periodontal treatment may attenuate systemic infl ammation and endothelial dysfunction in a dose-dependent manner in systemically healthy individuals with moderate to severe generalized periodontitis.

Future clinical research may focus on the design of cross-sectional and prospective studies with better defined criteria and/or markers of both exposure and disease to further investigate the role of confounding factors and the causal relationship between periodontitis and CVD. In addition, properly designed intervention studies prospectively evaluating the effect of periodontal therapy on CVD incidence and on validated surrogate markers of CVD events are warranted. Although the hypothesis of increased systemic inflammatory burden linking periodontitis and atherosclerosis is plausible, further preclinical and clinical investigations of the pathogenetic mechanisms underlying this association are needed.




The authors declare that there are no financial or other conflicts of interest related to this publication.