The bidirectional relationship between periodontal disease and diabetes mellitus


Panagiotis Koromantzos



Phoebus Madianos

Associate Professor, Department of Periodontology, School of Dentistry, University of Athens, Athens, Greece



Both periodontal disease and diabetes mellitus (DM) are chronic conditions that affect a large percentage of the global population. The relationship between these two disease entities, and more specifically the way one affects the other, has been researched extensively. In the mid 1960s, the first reports were published on the effect of DM on the periodontal health of patients, and since then a large body of evidence has proved that the severity, extent, and progression of periodontal inflammation is increased in patients with DM. Inflammation is accompanied by more severe and faster destruction of alveolar bone than occurs in nondiabetic subjects. The biological mechanisms that explain the relationship between periodontal disease and DM have been clarified to a large extent. In addition, glycemic control has been shown to play an important role in the periodontal condition of diabetic patients.

During the last decade, the research community has focused on the effect of periodontal disease on glycemic control. Studies have demonstrated that diabetic patients with severe periodontal disease have worse glycemic control than do those with mild or no periodontitis and that they even experience more diabetic complications. The effect that nonsurgical periodontal treatment has on the glycemic control of diabetic patients has also been studied, and the results show a favorable effect of periodontal therapy, thus establishing the association between periodontal disease and DM. The purpose of this review is to present contemporary evidence for the bidirectional nature of this relationship with emphasis on type 2 DM, as it affects 80-85% of all diabetics.



Periodontal disease and diabetes mellitus (DM) are chronic diseases that affect a large percentage of the world’s population. Periodontal disease is a chronic inflammatory condition that occurs in response to dental biofilm bacteria. Periodontal diseasse may affect part or all of the supporting periodontal tissues, resulting in inflammation and destruction of alveolar bone (Mantzavinos and Vrotsos 2002). DM is a metabolic syndrome that is characterized by a carbohydrate, fat, and protein metabolism disorder; hyperglycemia is the main result. DM constitutes a major health problem, as the number of diabetic patients is constantly increasing. In developed countries, it affects 6-8% of the population on average. In 1990, the frequency of DM in Greek people 50-79 years of age was reported to be 7.7-19.5%, with an increasing trend (Katsilambros et al. 1993). Furthermore, the relatively high number of undiagnosed DM cases should be added to these percentages. In 30% of patients with type 2 DM, the diagnosis is made by chance as part of another check (e.g., presurgical) or because of the presence of a diabetic complication (e.g., sight disorder) (Katsilambros 2005). The number of diabetic patients continuously increases; the World Health Organization has estimated that by 2025 the number will have doubled worldwide. According to the latest classification of the American Diabetes Association (ADA), there are four types of DM: type 1 DM, type 2 DM, gestational diabetes, and other specific types (ADA 2009b).

The relationship between periodontal disease and DM, and more specifically, how the one condition affects the other, has been researched extensively. From the 1960s onwards, papers have been published in the international literature demonstrating the effect of DM on the progress and severity of periodontal disease and discussing the possible mechanisms through which this interaction takes place (Glavind et al. 1968, Cohen et al. 1970, Sznajder et al. 1978). In 1993, periodontal disease was reported for the first time as a complication of DM (Löe 1993). In addition, the way in which glycemic control affects the periodontal condition of DM patients was studied on a smaller scale. In the last decade, research interest has turned to the study of the effect of periodontal disease on DM, whether the presence of inflammation in the periodontal tissues affects the severity of DM and impedes glycemic control of diabetic patients, and whether periodontal treatment can result in better glycemic control.

The aim of this paper is to present data on the bidirectional relationship of the two diseases, more specifically, the most recent data on the effect that nonsurgical periodontal treatment has on the glycemic control of patients with DM. A short report will also be included on the effect of DM and glycemic control in diabetic patients, on the periodontal condition so as to make clear the unprecedented need for periodontal treatment in diabetic patients. Particular emphasis will be on studies of patients with type 2 DM, as this DM type affects 80-85% of diabetic patients.


Effect of DM on periodontal disease

The first report on the influence of DM on periodontal disease was published in 1960 (Sandler and Stahl 1960). The majority of studies until the 1980s generally dealt with type 1 DM (insulin-dependent or juvenile diabetes according to past classifications), mainly on children or teenagers, although most often the type of diabetes was not clarified. Fewer trials in the literature refer to patients with type 2 DM only. Table 1 shows the studies that address the influence of type 2 DM on periodontal disease.

Μost of the studies addressing type 2 DM only are not directly comparable because of differences in the periodontal parameters assessed. Depending on the study, these measures include pocket depth (PD), clinical attachment level (CAL), bleeding on probing (BoP), plaque index (PI), bleeding index (BI), community periodontal index of treatment needs (CPITN) and radiographic bone loss. However, enough studies have been conducted to draw safe conclusions because most of them analyzed an adequate sample population and used correct methodology (Mealey and Ocampo 2007). Tests for glycosylated hemoglobin (HbA1c), considered the most reliable measure of metabolic control (ADA 2009a), were used in most trials; in only one study, fasting plasma glucose was also measured. A negative effect of type 2 DM and a deterioration of periodontal status were reported in all studies (Table 1). More specifically, an increase in the severity of periodontal disease expressed by elevated PD and attachment loss, was observed in four studies (Emrich et al. 1991, Morton et al. 1995, Novaes et al. 1996, Almas et al. 2001). An increase in the incidence of periodontal disease in patients with DM was observed in seven studies (Nelson et al. 1990, Shlossman et al. 1990, Emrich et al. 1991, Morton et al. 1995, Sandberg et al. 2000, Tsai et al. 2002, Campus et al. 2005), and three studies reported an enhanced rate of periodontal destruction in patients with DM (Novaes et al. 1996, Taylor et al. 1998a, 1998b).

Emrich et al. (1991) reported that patients with DM have an increased risk of periodontal attachment loss with an odds ratio (OR) of 2.81 (95% CI 1.9-4.1) and an increased risk for alveolar bone loss (OR 3.43, 95% CI 2.3-5.1) in comparison with nondiabetic patients, independent of age, sex, and oral hygiene. In a 6-year study on Pima Indians (a North American Indian population with the greatest incidence of type 2 DM in the world), periodontal disease was initially observed in 60% of people with DM, whereas it was observed in only 36% of those without DM (Nelson et al. 1990). After a 2.5-year observation period, without differentiating between sexes, a greater incidence of periodontal disease (2.6×) was reported in patients with DM than in those without DM. Although periodontal disease was also common in nondiabetic Pima Indians, type 2 DM was considered an accountable risk factor for periodontal disease. In another 2-year study, researchers concluded that patients with type 2 DM run a 4-fold greater risk for further radiographicbone loss compared with healthy people (Taylor et al. 1998a). An interesting conclusion emerged from studies that were carried out on children and teenagers with type 1 DM (Lalla et al. 2006, 2007a, 2007b). Researchers stated that the periodontal destruction in diabetic patients starts at a much earlier stage in life than was initially believed. They reported that in children and teenagers with type 1 DM, the periodontal destruction was higher in relation to people of the same age without DM (OR ranged from 1.84 to 3.72). The difference remained statistically significant, even with an age differentiation (6-11 and 12-18 years) of children with DM.

The cellular and molecular mechanisms through which DM negatively affects the periodontal condition of diabetic patients have been studied and clarified to a great extent. The same mechanisms that are in effect for the micro- and macrovascular complications of DM also apply to the negative effect of DM on periodontic tissues (Mealey and Rose 2008). Taking into account that periodontal diseases are chronic inflammatory conditions of microbial etiology that affect the supportive periodontal tissues, and because the subgingival microbiota are similar in subjects with and without DM (Zambon et al. 1988), it becomes apparent that the host reaction to the periopathogenic bacteria is the reason for the detrimental effect of DM on the periodontal tissues. This hypothesis is supported by the fact that patients with DM exhibit reduced chemotaxis and impaired activity of the neutrophils (Karima et al. 2005), paired with an increase in the production of proinflammatory cytokines and biological mediators of inflammation (Salvi et al. 1994a, 1997b, Muller et al. 2002, Engebretson et al. 2004, Gyurko et al. 2006). The increased cytokine expression is a result of the interaction between advanced glycosylation endproducts (AGEs) and the receptors of AGEs (RAGEs) on the periodontal tissues. This process causes the activation of RAGEs, which results in a severe reaction to periopathogenic bacteria, the presence of more severe inflammation, and a faster rate of alveolar bone destruction. Based on the aforementioned mechanisms, many researchers proposed the hypothesis of the impaired repair concept of periodontal tissues in patients with DM (Lalla et al. 2001, Graves et al. 2006, Liu et al. 2006, Nassar et al. 2007).

The formation of AGEs, which is nonreversible and responsible for many complications of DM, appears to play an important role in the progression of periodontal disease. The accumulation of AGEs in the periodontal tissues of diabetic patients and their binding to RAGEs on the surface of endothelial cells, macrophages, and fibroblasts, induces increased permeability of the local vasculature and activation of phagocytes (Lalla et al. 2000). These alterations result in increased levels of metalloproteinases and proinflammatory cytokines, including interleukin (IL)-6 and tumor necrosis factor (TNF), boosting the inflammatory response of the host to the periopathogenic bacteria.


Effects of glycemic control on periodontal disease

Glycemic control refers to the maintenance of plasma glucose at normal concentrations. There are different ways to determine the level of glucose control, including the home blood glucose monitoring and the glycosylated hemoglobin (HbA1c) test. The glucometer provides at home an instant blood glucose level with a puncture in the fingertip. HbA1c measurement constitutes the most reliable method of monitoring and assessing glycemic control, as it gives clinicians the average blood glucose levels for the past 3 months. The normal HbA1c is 6% or less, whereas the goal for patients with DM is to maintain HbA1c values <7% (ADA 2009a). Diabetic patients with HbA1c >8% require immediate treatment to improve glycemic control (ADA 2009a).

The correlation between glycemic control of patients with type 2 DM and the extent or severity of periodontal disease has received less attention than other aspects of the relationship between the two conditions. Although most related trials refer to patients with type 1 DM, Table 2 includes the studies on the relationship between glycemic control and periodontal disease in patients with type 2 DM.

As it is evident from Table 2, all but one study (Sandberg et al. 2000) correlate glycemic control with the extent of periodontal disease: the worse the glycemic control of the patients, the greater the extent of periodontal disease. However, the studies are not directly comparable mostly because of differences in the periodontal parameters assessed, as all articles with the exception of two (Ainamo et al. 1990, Unal et al. 1993) use the HbA1c test for glycemic control. Ainamo et al. (1990) noted that alveolar bone loss is correlated with elevated plasma glucose level, although the paper is a report of only two cases.

Tsai et al. (2002) analyzed data from the US National Health and Nutrition Examination Survey (NHANES III, 1988-1994) and included 4,343 participants over 45 years old. Patients with poorly controlled DB (HbA1c >9%) presented a higher prevalence of severe periodontal disease compared with people without DM (OR 2.9, 95% CI 1.4-6.0), whereas patients with satisfactory glycemic control had a lower risk (OR 1.5, 95% CI 0.9-2.7).


Effect of periodontal treatment on patients with type 2 DM 

Clinical periodontal parameters

The negative effect of type 2 DM on periodontal health has been established and proven to be correlated to glycemic control. The question that therefore arises is whether the treatment of periodontal disease is effective in diabetic patients, considering the medical background and complications of DM. Most studies published on the effectiveness of periodontal treatment in diabetic patients are parts of broader trials on the relationship between the two pathologic conditions, and, therefore, there is a great diversity in the methodology employed. Although the number of available studies is considered relatively low, they all conclude that there is no statistically significant difference in the response to periodontal treatment between controlled diabetic patients and nondiabetic patients (Table 3). Only one study reported statistically better PD outcomes for the control group of nondiabetic patients after a 6-month reexamination period (Faria-Almeida et al. 2006). However, this study was not randomized and included a small sample size (10 healthy and 10 diabetic patients). In a prospective study on well-controlled diabetic patients, the results of periodontal treatment that included a combination of surgical and nonsurgical therapy could be maintained for 5 years if the patients strictly followed the supportive periodontal therapy (Westfelt et al. 1996). Long-term maintenance of the therapeutic result is more difficult in patients who do not maintain good glycemic control (Tervonen and Karjalainen 1997).


Glycemic control

Given the above-mentioned negative effect of DM, and more specifically type 2 DM, on the extent, severity, and progression of periodontal disease, researchers developed working hypotheses and undertook investigations to study the effect of periodontal treatment on the glycemic control of diabetic patients. Investigators considered that the elimination or reduction of inflammation from the periodontal tissues can assist in the glycemic control of the diabetic patients. It was also suggested that the inflamed periodontal tissues appeared to be functioning as an “endocrine gland” that produces biological mediators of inflammation such as TNF-α and IL-1, which, in turn, have been shown to intervene in the metabolism of lipids and decrease insulin action (Grossi and Genco 1998). This process results in a negative effect of periodontal disease on the severity of DM and difficulty in maintaining satisfactory glycemic control.

In a 6-year study with two groups of 39 DM patients, Thorstensson et al. (1996) reported that the group with severe periodontal disease exhibited more diabetic complications (cardiovascular and kidney conditions) compared with the control group of diabetic patients with gingivitis or mild periodontitis. In another study carried out on 80 patients, Taylor et al. (1996) noted that patients with type 2 DM and severe periodontal disease had a sixfold risk of experiencing difficulty in glycemic control compared with patients who have DM without severe periodontal disease.

In a prospective 11-year study with 628 type 2 diabetic Pima Indians that aimed to determine whether periodontal disease is a prognostic factor for death in patients with DM, Saremi et al. (2005) concluded that patients with severe periodontal disease had 3.2 times the risk (95% CI 1.1-9.3) of cardiorenal mortality compared with the reference group that presented with no, mild, and moderate periodontal disease. The effect of periodontal disease should be considered as an additional effect on ischemic heart disease or nephropathy, both of which are diabetic complications.

Finally, in a retrospective 20-year study with 9,296 nondiabetic participants (part of the NHANES I data set), Demmer et al. (2008) concluded that the baseline periodontal status of patients can be considered as a prognostic factor for the presence of type 2 DM later in life (OR 2.26, 95% CI 1.56-3.27). The mechanisms underlying the bidirectional relationship between periodontal disease and DM are illustrated in Figure 1.

Few clinical studies address the effect of periodontal treatment on the glycemic control of diabetic patients and most refer to patients with type 1 DM or do not differentiate between the types of DM. Trials that have been exclusively conducted in patients with type 2 DM are presented in Table 4. All these studies report an improvement of HbA1c levels after nonsurgical periodontal therapy with or without the use of topical or systemic antibiotics. As evident from Table 4, all clinical studies, with the exception of two (Promsudthi et al. 2005, Jones et al. 2007) showed that a statistically significant improvement is observed in the level of glycemic control, as assessed by the HbA1c test. Two studies carried out on Pima Indians reported a significant 1% reduction in HbA1c levels 3 months after periodontal treatment, but a return to the previous condition within a year (Grossi et al. 1996, 1997). Patients in these two studies were randomized into groups that, apart from scaling and root planing (SRP) and systemic use of doxycycline, were also administered local povidone-iodine, chlorhexidine, or placebo. In a nonrandomized study, Stewart et al. (2001) investigated the effect of nonsurgical periodontal therapy on glycemic control in patients with type 2 DM, whereas diabetic patients in the control group did not receive any periodontal treatment. After a 9-month observation period, the levels of HbA1c decreased significantly in the treatment group (9.2 ± 2.2% to 7.6 ± 1.4%) compared with the control group (8.5 ± 2.1% to 7.7 ± 1.4%). Two more studies, cοnducted by the same research team, also indicated statistically significant improvements in the levels of HbA1c 6 months after periodontal therapy (Faria-Almeida et al. 2006, Navarro-Sanchez et al. 2007). However, the sample population was limited to 20 patients and the control groups were comprised of nondiabetic individuals. Finally, in the study with the greatest sample population to date (165 diabetic patients), Jones et al. (2007) observed improvement of HbA1c in the treatment group compared with the control group, but without statistical significance (mean absolute HbA1c change: -0.65% and -0.51%, respectively). Particularly interesting is the observation that the control group patients had double the possibility of increasing their insulin dosage in the time span from the initial examination until reexamination 4 months later, in comparison with the patients that received periodontal treatment after the initial examination (Jones et al. 2007).

We recently conducted a related study in the Department of Periodontology, University of Athens Dental School, in collaboration with the Diabetologic Center of the 1st Pathology Propedeutic Clinic, University of Athens Medical School (Koromantzos et al., unpublished data). The sample population consisted of 60 poorly controlled type 2 DM patients with HbA1c >7, randomly assigned into two groups of 30 subjects. The treatment group received nonsurgical treatment without the use of antiseptics or antibiotics, whereas the control group was treated in a similar manner after the completion of the study. A statistically significant improvement (p<0.05) of the HbA1c levels was reported in the treatment group (from 7.89 to 7.09) compared with the control group (from 7.51 to 7.36) at the 6-month evaluation.

Among the nine clinical studies published in the literature to date, five used systemic antibiotics including doxycycline and a combination of amoxicillin with clavulanic acid (Grossi et al. 1996, 1997, Rodrigues et al. 2003, Promsudthi et al. 2005, Jones et al. 2007). The use of adjunctive antibiotics has been proven to contribute to the resolution of inflammation in the periodontal tissues, but in the specific hypothesis that periodontal treatment can improve the glycemic control of diabetics, it may be a confounding factor. In a meta-analysis that includes clinical trials up to 2005, Janket et al. (2005) noted that the weighted average improvement in HbA1c for the type 2 diabetic patients who were treated only with SRP was 0.44%, whereas for the patients who received both SRP and antibiotics the respective decrease was 0.71%.



It is evident from the data presented that the severity, extent, and progression of periodontal disease are greater in patients with DM compared with healthy individuals. The relationship between periodontal didease and DM has been established with clinical studies, and the underlying cellular and molecular mechanisms have been clarified to a great extent. As a result, periodontal disease is considered today one of the diabetic complications. It has also been documented that the level of glycemic control is the main determinant in this relationship. In addition, published clinical studies prove that, compared with well-controlled diabetic individuals, DM patients with unsatisfactory glycemic control present greater severity, extent, and progression of periodontal disease. Regarding the effectiveness of periodontal therapy, a statistically significant difference between well-controlled diabetic patients and healthy people has not been established, although the few studies in the literature have not been adequately substantiated.

Finally, the studies that refer to the bidirectional relationship of the two diseases clearly demonstrate a positive effect of periodontal treatment on glycemic control. All clinical trials show a reduction in HbA1c after treatment, although it is not statistically significant in some of them. More randomized controlled clinical trials that include an adequate sample population are necessary to draw safe conclusions with greater statistical power.