Rheumatoid arthritis and periodontal disease: biological, clinical, and therapeutic relations


Niki M. Moutsopoulos

National Institute of Dental and Craniofacial Research, Bethesda, MD, USA


Clio P. Mavragani

Department of Experimental Physiology, School of Medicine, University of Athens, Athens, Greece



A number of clinical and epidemiological studies have been conducted in the past decade to investigate a possible association between periodontal disease and rheumatoid arthritis (RA). The majority of such studies indicate that patients with RA have an increased prevalence of periodontitis and tooth loss. However, because of the great variability in study designs, the strength of the association remains unclear. Furthermore, most available studies are cross-sectional, and therefore the temporality of the association between periodontitis and RA cannot be evaluated. Emerging evidence indicates that infection with Porphyromonas gingivalis, a bacterium highly implicated in chronic periodontal disease, might have a direct role in the etiology and pathogenesis of the chronic inflammatory response in RA. Additional noncausal pathways include genetic, environmental, and behavioral factors that are common to both conditions. Linking both disorders are the pathways involved in inflammatory bone destruction, suggesting possible shared therapeutic targets and research aims.

Analecta Periodontologica 2009; 20:135-148


Introduction to rheumatoid arthritis

Rheumatoid arthritis (RA) is a chronic, immune-mediated inflammatory disease that is characterized by synovial inflammation. Left untreated, it can lead to progressive destruction of cartilage and bone that may result in structural damage and functional disability (Gabriel 2001). RA affects women more than men at a ratio of 3-4:1, at virtually any age, but most commonly in middle adult life (Lipsky 2008). Although clinical presentation can vary, the typical picture of established RA is bilateral symmetric inflammatory polyarthritis involving small and large joints in both the upper and lower extremities (Figure 1), with sparing of the axial skeleton except the cervical spine. The predominant symptoms of RA are pain, morning stiffness, and swelling of peripheral joints. Although quite uncommon, systemic (extra-articular) features such as subcutaneous nodules, Sjogren syndrome, lung involvement, or vasculitis can occur. Laboratory findings include anemia of chronic type and increased inflammatory markers in active disease; autoantibodies against the Fc portion of IgG –the so-called rheumatoid factor– and against citrullinated peptides that incorporate the amino acid citrulline (McInnes and Schett 2007) are detected with varying sensitivity and specificity (Gabriel 2001, Lipsky 2008). Diverse genetic, hormonal, immunologic, and environmental factors have been identified as potential players in RA development, leading to heterogeneity of disease presentation and severity. A distinct environmental and genetic background could account for the milder nature of the disease in Greek compared with North European populations. RA in Greek populations is characterized by less inflammatory articular disease, fewer extra-articular manifestations, less severe joint damage, and a high frequency of Ro(SSA) antibodies linked to a higher prevalence of secondary Sjögren syndrome (Drosos et al. 1992, Drosos and Moutsopoulos 1995). Furthermore, Greek RA patients are characterized by a lack of association with human leukocyte antigens (HLA), the absence of amyloidosis alleles, and a weak association with HLA-DR1 and -DR4 antigens (Boki et al. 1992, 1993, Mavragani et al. 2007). In patients with aggressive joint disease, the long-term prognosis of RA is poor. In such patients, studies have shown that the risk of mortality for men with RA is 38% greater (55% for women) than the general population and that life expectancy is reduced by an average of 3-18 years (Yelin and Wanke 1999).


Is there a link between periodontal disease and RA?

To date, a large number of clinical studies have been conducted in an effort to investigate a possible association between periodontitis and RA (Albandar 1990, Tolo and Jorkjend 1990, Kässer et al. 1997, Mercado et al. 2000, 2001, Abou-Raya et al. 2005, Al-Shammari et al. 2005, Havemose-Poulsen et al. 2006, Marotte et al. 2006, Abou-Raya et al. 2008, de Pablo et al. 2008, Pischon et al. 2008, de Pablo et al. 2009).

The majority of studies conducted are case-control studies with relatively small cohorts of fewer than 100 participants. Great variability occurs in study design and particularly in the definitions of periodontitis and RA. Periodontitis is typically assessed clinically by measuring probing pocket depth (PD) and clinical attachment level (CAL) at various sites of dentition. Definitions of periodontitis in the clinical studies of RA vary widely from self-reported disease and tooth loss as measures of periodontitis to varying levels of attachment loss in one or more sites of the mouth being considered as presence of disease (de Pablo et al. 2009). Different studies have also investigated the association of RA with varying types of periodontitis (chronic and aggressive, localized and generalized). The definition of RA has been more consistent throughout the studies because the majority of studies employed the American College of Rheumatology (ACR) criteria for defining RA; however, some did use subjective and inaccurate criteria such as self-reported RA (de Pablo et al. 2009).

Some of the larger studies conducted to date are briefly described below (Table 1) as a sample of the type of studies investigating the degree of association between periodontal disease and RA. Pischon et al. (2008) examined 57 patients with RA and 52 healthy controls matched by age, gender, and sociodemographic characteristics. The researchers reported a significant 5.7-fold increase in odds of periodontal disease (mean CAL >4 mm) in RA patients compared with healthy controls with a 95% confidence interval (CI) 2.35-13.84. CAL and PD were significantly higher in subjects with RA compared with control subjects (4.37 ± 1.30 mm vs 3.40 ± 0.89 mm, p <0.001 and 3.71 ± 0.73 mm vs 3.16 ± 0.58 mm, p <0.001, respectively). Significant differences were also found in plaque accumulation assessed by the plaque index (PI) (0.71 ± 0.46 vs 0.44 ± 0.28, p <0.001) and inflammatory gingival status evaluated by the gingival index (GI) (0.83 ± 0.48 vs 0.57 ± 0.39, p <0.003). Pischon et al. (2008) also examined the association of RA with periodontal disease after adjusting for potential confounders. PI and GI were markedly associated with CAL (Spearman correlation coefficient r = 0.59, p <0.001 and r = 0.70, p <0.001, respectively). The strength of the association of RA with periodontal disease decreased but remained statistically significant after further adjustment for PI, GI, or both.

Abou-Raya et al. (2005) included in their study 50 patients with RA, diagnosed according to the 1987 revised American Rheumatism Association (ARA) criteria, and 50 controls matched for age, sex, social status, and dental hygiene. The authors evaluated parameters of inflammation and oral care, including gingival bleeding and calculus. In addition, they evaluated indicators of periodontal disease such as PD >4 mm, loss of attachment >4 mm, and missing teeth (a parameter that is not always associated with periodontal disease). Significant correlations of RA with gingival inflammation, calculus, periodontal measures, and tooth loss were reported. Nevertheless, the severity of disease was not clearly described in the patient and control populations. The percentage of sites with PD >4 mm was not reported; only the number of patients with such sites were documented. Finally, the presence of increased calculus and inflammation was not controlled and could account for the increased presence or severity of periodontal disease. The authors also found positive associations between a number of RA parameters and periodontal disease severity, suggesting a possible contribution of the one condition on the severity of the other. The severity of PD was found to be positively correlated to RA disease duration (r = 0.62, p <0.001), number of tender and swollen joints (r = 0.62, p <0.001), erythrocyte sedimentation rate (ESR) (r = 0.44, p <0.005), visual analogue scale (r = 0.34, p <0.001), and C-reactive protein (r = 0.41, p <0.003). The authors also correlated RA parameters with the pantomography index, which is not a measure of periodontal disease but of general oral health.

De Pablo et al. (2008) analyzed data from the US National Health and Nutrition Examination Survey (NHANES III, 1988-1994) and included participants older than 60 years who had undergone dental and musculoskeletal examinations. In this study, RA diagnosis was based on ACR criteria, and periodontitis was defined as at least one site exhibiting both attachment loss and PD >4 mm. Dental status was classified as (1) no periodontitis, (2) periodontitis, and (3) edentulism. Findings indicate that of 4,461 participants, 103 were diagnosed with RA. Diagnosis of RA and edentulism were strongly associated with an odds ratio (OR) of 2.3 (95% CI 1.5-3.3) after adjusting for age, sex, ethnicity, and smoking. The association with edentulism was particularly strong for those with seropositive RA (OR 4.5, 95% CI 1.2-17). Diagnosis of RA was also associated with a greater number of missing teeth (20 vs 16, p <0.001), although it was not documented whether these teeth were lost as a result of periodontal disease. The association of RA with the presence of periodontitis was less profound, though present (OR 1.82, 95% CI 1.0-3.2). However, the patient population in this study was substantially older than the populations of most other studies. Therefore, patients with periodontal disease may have already lost some teeth with advanced disease, potentially contributing to the increased tooth loss in the RA population.

Kässer et al. (1997) evaluated 50 RA patients and 101 controls matched for age, sex, smoking status, and oral hygiene. The authors reported that patients with long-standing active RA (mean ± SD: 13 ± 8 years) who were receiving treatment with disease-modifying antirheumatic drugs (Ν = 46), corticosteroids (Ν = 38), or nonsteroidal anti-inflammatory drugs (Ν = 43) had a higher rate of gingival bleeding (increased by 50%), greater PD (increased by 26%, 2.9 ± 0.8 vs 2.3 ± 0.4 mm, p <0.001), greater attachment loss (increased by 173%, 2.6 ± 1.7 vs 0.95 ± 0.7mm, p <0.001), and more missing teeth (increased by 29%, 7.2 ± 4.4 vs 5.6 ± 4.4 p<0.01) compared with controls.

Finally, Mercado et al. (2001) examined 65 consecutive patients attending a rheumatology clinic for their levels of periodontitis and RA. No differences were noted in plaque accumulation or bleeding on probing between the control and RA groups. Periodontal evaluation included PD and CAL measurements, missing teeth, and radiographic bone loss. The most severe bone loss was used to classify each patient as follows: P0 (no bone loss), P1 (mild bone loss), P2 (moderate bone loss corresponding up to 2/3 of the root length), and P3 (severe bone loss). The percentage of RA patients in the P0 to P1 category was 30.82% vs 66.18% (p <0.05) in the control population. More patients were in the P2/P3 bone loss category in the RA population, 69.2 vs 33.8% (p <0.05). When analyzed with logistic regression, RA patients were twice as likely to have moderate to severe bone loss (OR 2.27, 95% CI 1.1-4.6).

A number of studies have also indicated that periodontitis may contribute to the severity of RA. Periodontal treatment with scaling and root planing has been shown to reduce signs and symptoms of RA, including disease activity scores (DAS) and ESR in multiple studies (Ribeiro et al. 2005, Al-Katma et al. 2007).

As illustrated by these studies, there has been great variability in the selection criteria, definition of periodontitis and RA, and selection of controls throughout the study populations; thus, it is difficult to draw conclusions about the level of association between the two conditions. However, the majority of studies conducted to date report a level of association between periodontitis and RA with an OR range of 1.8-6,0 (de Pablo et al. 2009). Future larger scale studies are needed to further characterize this association. Such studies should also document additional parameters, such as years since diagnosis of RA, inflammatory indices of disease activity, severity of disease for both conditions, and serological markers, all of which may influence the level of association. Future studies should also investigate possible common genetic and environmental parameters (such as smoking and HLA-DR loci), which may contribute to the development of both conditions.

This association or coexistence between the conditions could arise either because one is contributing to the etiology and pathogenesis of the other or because similar environmental or genetic parameters govern the manifestation and progression of both. Although many common factors may influence both conditions, interestingly, unlike RA, periodontal disease is not seen primarily in the female population. On the contrary, epidemiological studies indicate a greater prevalence and severity of periodontal disease in men (Albandar 1990, Hyman and Reid 2003). The academic community is currently investigating possible contributions of periodontal disease to the pathogenesis of RA while keeping in mind that similar patterns of genetic susceptibility and common environmental factors such as tobacco use could contribute to both conditions. In addition, symptoms of advanced RA, such as lack of manual dexterity, malaise, and hospitalizations, may influence oral hygiene and level of dental care, rendering patients susceptible to oral diseases.


Could periodontal disease contribute to the pathogenesis of RA?

The etiology and pathogenesis of RA continue to be investigated but remain largely unknown. The general consensus is that various environmental triggers may act on a susceptible genetic background to initi66,18 ate an autoimmune process. Several genetic loci have been linked with the susceptibility and severity of RA. The most established association has been with HLA-DR4 alleles. Other loci, including PTPN22 (protein tyrosine phosphatase, nonreceptor type 22), PADΙ4 (peptidyl arginine deiminase, type IV), CTLA4 (cytotoxic T lymphocyte antigen 4), STAT4 (signal transducer and activator of transcription 4), FcγRs (Fc receptors for IgG), and various cytokine and cytokine-receptor loci, such as those encoding tumor necrosis factor (TNF) and interleukin (IL)-1, IL-10 and IL-18, have been implicated, with various degrees of association to distinct populations (McInnes and Schett 2007). Many of the genes are important in immune regulation; HLA-DRB1, PTPN22, STAT4, CTLA4 and IL2 are all involved in T-cell activation and signaling pathways. Environmental factors also have an effect on the induction, magnitude and rate of progression of the disease. Numerous infectious organisms have been implicated, but recent data most strongly implicate smoking as an important environmental risk factor for the development of disease in HLA-DR4-positive individuals (McInnes and Schett 2007).

Initial manifestation of disease in RA involves production of autoantibodies specific for the human immunoglobulin IgG (known as rheumatoid factors) or specific for cyclic citrullinated peptides (CCPs). This family of autoantibodies includes antiperinuclear factor, antikeratin antibody, antivimentin, and antifilaggrin antibody. These autoantibodies all recognize epitopes containing citrulline, a nonessential amino acid, which is incorporated into proteins during post-translational modification. Citrullinated residues are generated by deimination of the guanidino group of carboxy-terminal arginine residues by the enzyme peptidyl arginine deiminase (PAD) (Liao et al 2009). Anti-CCP antibodies are the most specific autoantibodies of RA and can be detected in almost 80% of RA sera, with a specificity of 99% (van Venrooij et al. 2004, Vossenaar et al. 2004). Besides being specific for RA, the antibodies can be detected very early in the disease and can predict clinical disease outcome. To date, five isoforms of PAD (PAD1, 2, 3, 4, and 6) have been identified on human chromosome 1p36, a region related to susceptibility to RA (Suzuki et al. 2003).

A possible etiologic/pathogenic link between periodontal disease and RA comes from the realization that Porphyromonas gingivalis, a recognized major pathogenic microorganism in periodontal disease, is the only bacterium known to express a PAD enzyme (McGraw et al. 1999, Liao et al. 2009). Although distinct from the mammalian form, bacterial PAD is responsible for the post-translational conversion of arginine to citrulline. The ability of P. gingivalis to express PAD suggests that infection with this microorganism could impact RA onset and progression by facilitating the presentation of autoantigens and the production of disease-specific autoantibody targeting citrullinated peptides. An important feature of microbial PAD is its capacity to deiminate arginine in fibrin found in the periodontal lesion (Travis et al. 1997). This feature is especially noteworthy because citrullinated variants of the α- and β-fibrin chains have been proposed as target “autoantigens” in the rheumatoid joint (Liao et al. 2009).

The possible role of P. gingivalis in RA is also supported by studies demonstrating that antibody titers to P. gingivalis are significantly increased in patients with RA and are significantly correlated with anti-CCP antibody isotypes (most notably anti-CCP-IgM and -IgG2 subclass) that are specific to RA (Moen et al. 2006). The association of immune responses to P. gingivalis with anti-CCP-IgG2 subclass concentrations may be particularly noteworthy because the IgG2 subclass represents the predominant serum IgG response to infection with P. gingivalis (Whitney et al. 1992). P. gingivalis DNA and antibodies to P. gingivalis have also been recovered in the synovial fluid of patients with RA (Moen et al. 2003, 2006, Martinez-Martinez et al. 2009).

With this currently available data, it appears likely that P. gingivalis infection influences the early prearticular phase of RA and may help set the stage of an autoimmune background in which T-cell and B-cell tolerance are breached, allowing a subsequent event (which remains poorly understood) to trigger articular localization, inflammatory synovitis, and clinical presentation.


Inflammatory bone loss in RA and perio-dontal disease: Are there common treatment modalities?

Although there could be a causal or noncausal association between periodontal disease and RA, the two conditions are also linked by similarities in their pathogenesis and presentation (Mercado et al. 2003, Herman et al. 2008, Smolik et al. 2009). Key features shared between the two conditions are the mechanisms involved in inflammatory-induced tissue and, in particular, bone degradation. During the course of different inflammatory diseases such as RA, psoriatic arthritis, and periodontal disease (Moutsopoulos and Madianos 2006), a confined set of common inflammatory pathways triggers the process of local bone destruction (Herman et al. 2008). Since the early 1970s, accumulating evidence has indicated that the immune and skeletal system share several factors, such as cytokines, transcription factors, and receptors. Furthermore, immune cells and osteoclasts are derived from the same hematopoietic precursor cells. Therefore, there is compelling evidence that these two systems influence each another under both physiological and pathological circumstances.

A clear example of the interrelated nature between the two systems comes from autoimmune and chronic inflammatory diseases such as RA and periodontal disease, in which prolonged and enhanced activation of the immune system leads to massive bone destruction. During inflammation, the balance between bone formation and bone resorption is disturbed in favor of osteoclast-mediated bone resorption. Bone resorption depends on the influx of osteoclast precursor cells into the inflamed tissue and their differentiation into mature osteoclasts. Both processes, which are accelerated during an inflammatory reaction, are controlled via a cytokine signaling network formed among cells of the osteoclast lineage. The major factor responsible for osteoclast differentiation is the receptor activator of nuclear factor-kappa (RANK)/RANK ligand (RANKL) system (Boyle et al. 2003). RANKL expression is regulated by inflammatory cytokines, such as TNF, IL-1β, IL-6, and IL-17, but is also influenced by noncytokine inflammatory mediators such as prostaglandin E2 (McInnes and Schett 2007). RANKL induces the final differentiation of osteoclasts and their bone-resorbing activity. The interaction of RANKL with its receptor RANK is modulated by osteoprotegerin (OPG; also known as TNFRSF11B), a soluble decoy receptor, which is expressed by mesenchymal cells in the RA synovium. In RA and in periodontal disease, an imbalance between OPG and RANKL expression promotes RANKL-induced bone loss (McInnes and Schett 2007, Cochran 2008).

Anti-inflammatory therapies have been extensively used in RA in an effort to control the destructive chronic inflammation. The use of nonsteroidal anti-inflammatory drugs (NSAIDs) widely applied in RA have shown some limited promise for use in periodontal disease. The discovery that NSAIDs block the enzyme cyclooxygenase and reduce prostaglandin synthesis led to in vitro studies evaluating NSAIDs as inhibitors of bone resorption. Inhibition of periodontal disease progression using an NSAID was first demonstrated with indomethacin in a ligature-induced canine periodontitis model (Nyman et al. 1979). Multiple NSAIDs, including indomethacin, flurbiprofen, ibuprofen, naproxen, meclofenamic acid, and piroxicam, demonstrated the ability to inhibit gingivitis and progression of periodontitis in animal models (Williams et al. 1987, 1988a, 1988b, 1989, Howell et al. 1991).

In humans, clinical trials have assessed the efficacy of topically and systemically administered NSAIDs in the treatment of experimental gingivitis and chronic and aggressive periodontitis, with contradictory results (Jeffcoat et al. 1991). The most extensive clinical trial that investigated a systemically administered NSAID (flurbiprofen) demonstrated significantly lower bone loss rates over an 18-month period, though the effect of the agent was not sustainable in the long term (Heasman et al. 1993).

During the last decade, remarkable progress has been made in the development of targeted biological therapies such as monoclonal antibodies and fusion proteins. Monoclonal antibody technology was first invented more than three decades ago (Kohler and Milstein 1975) and quickly revolutionized the management of malignancies in the context of stem cell and solid organ transplantation, rheumatologic disorders, autoimmune disease, inflammatory bowel disease, in fections, and atopic disorders (Shirota et al. 2008).

The most commonly used “biologics” interfere with the action of cytokines, which are soluble mediators of inflammation. Cytokines exert their action by binding to cell surface receptors. The action of cytokines can be blocked by binding the soluble cytokine or preventing its binding to its cognate receptor. The former can be achieved by using monoclonal antibodies against the cytokine or by using a decoy soluble receptor, which will bind to the cytokines in a manner similar to that of cell surface receptors, thereby reducing the levels of free, biologically active cytokines. All of these options have already been successfully used in the treatment of RA. Of three anti-TNF agents approved in the United States, two (infliximab and adalimumab) are monoclonal antibodies, whereas the third, etanercept, is a soluble receptor fusion protein (Keystone 2006, Scott and Kingsley 2006). TNF blockade has been shown not only to reduce clinical symptoms of inflammation, but also to suppress progression of inflammatory bone destruction. Research has identified numerous pathogenetic pathways in RA that represent potential targets for biological therapy beyond the already successful anti-TNF inhibitors (Keystone 2006, Scott and Kingsley 2006). Addition al members of the proinflammatory cytokine network, such as IL-6, IL-15, IL-17, and RANKL, are emerging as suitable therapeutic targets for effective amelioration of inflammation and bone destruction.

Among the cytokine targets of interest are TNF-α, IL-6, IL-1, and IL-17, which we, and others, have shown to contribute to the immunopathological lesions of periodontal disease (Stashenko et al. 1991, Kornman et al. 1997, Page and Kornman 1997, Nares et al. 2009). The role of IL-17 and T-helper 17 (TH17) cells, a unique CD4+ T-cell subset characterized by production of IL-17, has recently emerged as key in the pathogenesis of periodontal disease. Levels of IL-17 have been shown to be high in periodontal tissues, particularly adjacent to bone-destructive sites (Ohyama et al. 2009), though the role of this cytokine remains controversial. TH17 cells are protective against microorganisms and fungi, but they have also been implicated in the destructive phase of periodontal disease (Gaffen and Hajishengallis 2008). TH17 cells are now a recognized osteoclastogenic lymphocyte population that links T-cell activation to bone resorption (Sato et al. 2006). Ultimately, studies of periodontitis in humans treated with anticytokine therapeutics, and particularly anti-IL-17 or anti-IL-23, may reveal the most relevant players in the pathogenesis of disease, but may also lead to the development of therapeutic interventions in this locally destructive disease (Giannobile 2008, Shirota et al. 2008).



Although the extent and etiology of the association between periodontal disease and RA continue to be investigated, it has become clear that similar biological mechanisms govern the immune-mediated tissue destruction in both conditions. This realization brings periodontal diseases to the forefront of biomedical research as a model disease for the understanding of the communication between the immune and skeletal systems and for the study of pathways that lead to chronic inflammation, impaired immune regulation, and defective tissue remodeling. The accessibility of the oral tissues and the high prevalence of this chronic inflammatory condition, which affects more than 10% of the general population, renders it an ideal study model. Conversely, the advances already made in the field of research and therapeutics in RA may likely be applicable to periodontal disease and should be seriously evaluated.



The authors declare no conflicts of interest related to this article. This work was partially funded by the National Institute of Dental and Craniofacial Research (NIDCR), National Institutes of Health, Bethesda, MD, USA.