Periodontal disease: associations with diabetes, glycemic control and complications

Authors


  • There are no financial relationships that may pose a conflict of interest.

George W Taylor, School of Dentistry and School of Public Health, University of Michigan, 1011 N. University, Ann Arbor, MI 48109, USA. Tel: 734 764 1737, Fax: 734 936 1597, E-mail: gwt@umich.edu

Abstract

Objective:  This report reviews the evidence for adverse effects of diabetes on periodontal health and periodontal disease on glycemic control and complications of diabetes.

Design:  MEDLINE search of the English language literature identified primary research reports published on (a) relationships between diabetes and periodontal diseases since 2000 and (b) effects of periodontal infection on glycemic control and diabetes complications since 1960.

Results:  Observational studies provided consistent evidence of greater prevalence, severity, extent, or progression of at least one manifestation of periodontal disease in 13/17 reports reviewed. Treatment and longitudinal observational studies provided evidence to support periodontal infection having an adverse effect on glycemic control, although not all investigations reported an improvement in glycemic control after periodontal treatment. Additionally, evidence from three observational studies supported periodontal disease increasing the risk for diabetes complications and no published reports refuted the findings.

Conclusion:  The evidence reviewed supports diabetes having an adverse effect on periodontal health and periodontal infection having an adverse effect on glycemic control and incidence of diabetes complications. Further rigorous study is necessary to establish unequivocally that treating periodontal infections can contribute to glycemic control management and to the reduction of the burden of diabetes complications.

Introduction

Diabetes mellitus and periodontal disease are two common chronic diseases that have long been considered to be biologically linked. Diabetes is an important chronic disease globally as reflected in the World Health Organization (WHO) declaring the rate of increase in diabetes prevalence is an epidemic. The WHO estimated there were 30 million people who had diabetes worldwide in 1985. This number increased to 135 million by 1995, and reached 217 million in 2005. By 2030 WHO predicts this number to increase to at least 366 million (Smyth and Heron, 2006). This growth in diabetes prevalence, driven principally by increasing prevalence of type 2 diabetes, is occurring in both developing and developed countries. The two countries with the largest predicted increases are India and China and the US ranked third (Smyth and Heron, 2006).

Susceptible individuals with diabetes and those with chronically poor metabolic control can experience microvascular and macrovascular complications leading to a significant burden for the individual and society. This burden includes direct costs of medical care and indirect costs, such as lost productivity, which result from diabetes-related morbidity and premature mortality (Harris, 1995; Hogan et al, 2003). Health care spending for people with diabetes is more than double what spending would be without diabetes, and direct and indirect expenditures attributable to diabetes in 2002 in the US were conservatively estimated at $132 billion, with slightly more spent on chronic complications attributable to diabetes than on diabetes care itself (Hogan et al, 2003). The International Diabetes Federation estimated that diabetes accounts for 5–10% of the total healthcare budget in many countries (Smyth and Heron, 2006).

Gingivitis and periodontitis are the most common periodontal diseases. For example, in the US approximately 50% of the population in all age groups exhibit reversible gingival inflammation (Albandar and Kingman, 1999). Moderate or severe periodontitis, with destruction of periodontal attachment tissues is much less common than gingivitis yet still a common chronic disease, affecting approximately 5–15% of any population (Albandar et al, 1999; Burt, 2005).

Current evidence regarding the biologic link between diabetes and periodontal disease supports diabetes and persisting hyperglycemia leading to an exaggerated immuno-inflammatory response to the periodontal pathogenic bacterial challenge (Southerland et al, 2006; Nishimura et al, 2007), resulting in more rapid and severe periodontal tissue destruction. In the metabolic dysregulation of diabetes, persisting hyperglycemia causes non-enzymatic glycation and oxidation of proteins and lipids, and the subsequent formation of advanced glycation endproducts (AGEs), which accumulate in the plasma and tissues (Brownlee, 1994; Schmidt et al, 1996b; Ramasamy et al, 2005). Hyperglycemia and resultant AGE formation are considered to be a major causal factor in the pathogenesis of diabetes complications (Brownlee, 1994; Vlassara, 1994). In subjects with diabetes who also have periodontitis, AGEs with accompanying markers for increased oxidant stress have been demonstrated in human gingiva (Schmidt et al, 1996a). Cell surface binding sites or receptors for AGE (RAGE) have been identified on the cell surfaces of several cell types exhibiting a heightened inflammatory response and involved with the pathogenesis of complications of diabetes. These cell types include mononuclear phagocytes, endothelial cells, fibroblasts, smooth muscle cells, lymphocytes, podocytes, and neurons (Brett et al, 1993; Ramasamy et al, 2005). The receptor for AGEs, RAGE, is the principal signal transducer for the AGE ligand (Schmidt et al, 2000).

The underlying postulate associated with these findings is that enhanced oxidant stress in the gingival tissues could contribute to more frequent and more severe periodontal tissue destruction in individuals with diabetes. For example, it has been hypothesized that the AGE-RAGE interaction induces an oxidant stress that may contribute to chronic monocytic upregulation, activation of NF-κB, and subsequent expression of mRNA and secretion of proinflammatory cytokines (such as TNF-α, IL-1β, and IL-6) by monocytic phagocytes involved in periodontal tissue inflammation and destruction (Baeuerle, 1991; Schreck et al, 1991; Moughal et al, 1992; Collins, 1993; Schmidt et al, 1994, 1996a; Takahashi et al, 1994; Yan et al, 1994). These mediators are recognized as effectors in periodontal tissue inflammation and destruction (Salvi et al, 1998). Blockade of RAGE has been shown to diminish Porphyromonas gingivalis-triggered alveolar bone loss in the periodontium and limit the enhanced inflammatory response in peripheral wounds, accelerating wound closure and facilitating angiogenesis (Lalla et al, 2000; Goova et al, 2001). Additionally, AGE interaction with endothelial cell RAGE has been shown to enhance endothelial cell vascular hyperpermeability and expression of vascular cell adhesion molecule-1, an adherence molecule capable of attracting mononuclear cells to the vascular wall (Schmidt et al, 1995; Wautier et al, 1996; Lalla et al, 1998b). Hence, AGE-RAGE interaction has been proposed to result in pertubation of cellular properties, exaggerated and sustained inflammatory response, impaired wound healing, and more severe diabetes-associated periodontal disease (Lalla et al, 1998a).

The specific ways in which diabetes-enhanced inflammation and apoptosis may specifically impact on periodontal tissues of was recently thoroughly reviewed (Graves et al, 2006). In their review, Graves and colleagues describe that diabetes has been reported to adversely affect bone repair by decreasing expression of genes that induce osteoblast differentiation, and diminishing growth factor and extracellular matrix production (Bouillon, 1991; Kawaguchi et al, 1994; Lu et al, 2003). One proposed mechanism for these adverse effects is through the contribution of AGEs to decreased extracellular matrix production and inhibition of osteoblast differentiation (McCarthy et al, 2001; Cortizo et al, 2003; Santana et al, 2003). AGEs may also delay wound-healing by inducing apoptosis of extracellular-matrix-producing cells. This enhanced apoptosis would reduce the number of osteoblastic and fibroblastic cells available for the repair of resorbed alveolar bone (Graves et al, 2006). In addition to promoting apoptosis, AGEs could affect oral tissue healing by reducing expression of collagen and promoting inflammation. The mechanisms suggested for AGE-enhanced apoptosis include the direct activation of caspase activity, and indirect pathways that increase oxidative stress or the expression of pro-apoptotic genes that regulate apoptosis (Graves et al, 2006).

Diabetes mellitus and its effects on periodontal disease

Evidence establishing the link between diabetes mellitus and adverse effects on periodontal health have been extensively reviewed (Taylor, 2001; Mealey et al, 2006). In a narrative review of the English language literature published between 1960 and 2000 Taylor (2001) reported that 44 of 48 observational studies provided supportive evidence of diabetes adversely affecting periodontal health provided (37 of the 41 cross-sectional and seven of the seven cohort studies).

The review conducted for this current report extends that 2001 review to include reports published into 2007. The search used MEDLINE as well as reviewed reference lists of relevant papers obtained from the search to identify primary research reports on investigations of relationships between diabetes/diabetes control and periodontal diseases/periodontal treatment. While the literature review is extensive in conducting the MEDLINE search, it is not exhaustive in that no other databases were searched. This review does not provide a formal assessment of the quality of the reports. The reports identified are displayed in table-form and the corresponding description is organized according to the following groupings of studies: (1) The effects of having diabetes on periodontal diseases in studies that include a non-diabetes comparison group (Table 1) and (2) Effect of the degree of glycemic control, usually measured by level of glycosylated hemoglobin, on periodontal status in studies that included assessment of degree of glycemic control while evaluating periodontal status in participants with diabetes (Table 2).

Table 1.   Effects of diabetes on periodontal diseases in studies including a non-diabetes control group; ordered by diabetes type and subject age
ReferenceCountryStudy design Diabetes typeaNo. subjects
a. Diabetes
b. Control
Agesb
a. Diabetes
b. Control
Perio. Measure: diabetes effectcOther diabetes-related variables considered
  1. aDM type = diabetes type: 1 = type 1 diabetes mellitus; 2 = type 2 diabetes mellitus; 1,2 = both subjects with type 1 and type 2 diabetes mellitus included; GDM = gestational diabetes mellitus; 9 = diabetes type not specified and not clearly ascertainable from other information in the report.

  2. bAges: subjects’ ages presented as minimum – maximum reported for those with a. diabetes (DM) and b. controls (Control) unless otherwise specified.

  3. cMeasure of periodontal disease status: Measures used include Ging = gingivitis or gingival bleeding, Ppd = probing pocket depth, Lpa = loss of periodontal attachment, XRBL = radiographic bone loss, JPS = juvenile periodontal score, MGI = modified gingival index, PI = Russell’s Periodontal Index, PDR = periodontal disease rate (proportion of teeth affected by periodontal disease). The number following the measure corresponds to greater disease in those with diabetes (1) or no difference between those with diabetes and controls (0). The letters following the number correspond to the parameter(s) assessed in the study: e = extent, i = incidence, p = prevalence, s = severity, r = progression.

Tervonen et al (2000)FinlandCross-sectional1a. 35
b. 10
a. 29.7 (mean)
b. 29.0 (mean)
XRBL: 1eGlycemic control
Duration of diabetes
Diabetes severity based on presence of complications
Endean et al (2004)AustraliaCross-sectional2a. 58
b. 231
All: 15–45+
a. Unknown
b. Unknown
Ppd: 1p, 1sNone
Mattout et al (2006)FranceCross-sectional2a. 71
b. 2073
All: 35–75
a. 54.5 (mean)
b. 49.0 (mean)
Ging: 1p, 1s
Ppd: 0p, 0s
Lpa: 1p, 1s
Fasting blood glucose
Campus et al (2005)ItalyCross-sectional2a. 71
b. 141
a. 36–75
b. 35–75
Ging: 1e
Ppd: 1e, 1s
Glycemic control
Orbak et al (2002)TurkeyCross-sectional2a. 40
b. 20
a1. 46 (mean)
a2. 43 (mean)
b. 41 (mean)
Ging: 1e, 1p, 1sGlycemic control
Diabetes complications
Tsai et al (2002)USACross-sectional2a. 502
b. 3841
a. 45+
b. 45+
Lpa & Ppd: 1pGlycemic control
Lu and Yang (2004)TaiwanCross-sectional2a. 72
b. 92
a. 54.3 (mean)
b. 54.9 (mean)
Ging: 1p, 1e, 1s
Lpa: 1p, 1e, 1s
Glycemic control
Duration of diabetes
Chuang et al (2005)TaiwanCross-sectional2a. 43
b. 85
All: 28–85
a. 60.2 (mean)
b. 56.1 (mean)
Ppd: 0sGlycemic control
Sandberg et al (2000)SwedenCross-sectional2a. 102
b. 102
a. 64.8 (mean)
b. 64.9 (mean)
Ging: 1e
Ppd: 1e
XRBL: 1p
Glycemic control
Duration of diabetes
Zielinski et al (2002)USACross-sectional2a. 32
b. 40
All: 60+
a. 71 (mean)
b. 74 (mean)
Ppd: 0e, 0p, 0sGlycemic control
Duration of diabetes
Borges-Yáñez et al (2006)MexicoCross-sectional2a. 247
b. 78
All: 60+
a. 73.4 (mean)
b. Unknown
Lpa: 0pFasting blood glucose
Lalla et al (2007)USACross-sectional1, 2a. 350
b. 350
a. 6–18
b. 6–18
Ging: 1e, 1p, 1s
Ppd: 1e, 1p, 1s
Lpa: 1e, 1p, 1s
Duration of diabetes
Glycemic control
Arrieta-Blanco et al (2003)SpainCross-sectional1, 2a. 70
b. 74
a. 11–81
b. 11–75
Ging: 1e
Ppd: 0s, 0e
Lpa: 1e, 1s
XRBL:0s, 0e
Glycemic control Duration of diabetes
Diabetes complications
Ogunbodede et al (2005)NigeriaCross-sectional1, 2a. 65
b. 54
a. 25–82
b. 25–82
Ppd: 0pDuration of diabetes
Xiong et al (2006)USACross-sectional1, 2, GDMa. 81
b. 4339
All: 15–44
a. Unknown
b. Unknown
Ppd or Lpa: 1pNone
Novak et al (2006)USACross-sectional2, GDMa. 113
b. 4131
All: 20–59
a. Unknown
b. Unknown
Ging & ppd & lpa: 1p, 1sGlycemic control
Duration of diabetes
Mittas et al (2006)GreeceCross-sectionalGDMa. 64
b. 88
a. 31.1 (mean)
b. 26.5 (mean)
Ging: 1sNone
Table 2.   Effect of degree of glycemic control on periodontal status, ordered by level of evidence, diabetes type, and subject age
ReferenceCountryStudy design Diabetes typebAge group EffectcNon-DM comparison groupdEvidence levela
  1. aHierarchy of evidence based on classification scheme used (U.S. Preventive Services Task Force, 1996) where: I = evidence obtained from at least one properly randomized controlled trial; II-1 = evidence obtained from well-designed controlled trial without randomization; II-2 = evidence obtained from well-designed cohort or case-control analytic studies, preferably from more than one center or research group; II-3 = evidence obtained from multiple time series with or without the intervention. Dramatic results in uncontrolled experiments (such as the results of the introduction of penicillin treatment in the 1940s) could also be regarded as this type of evidence; III = opinions of respected authorities, based on clinical experience; descriptive studies and case reports; or reports of expert committees.

  2. bDiabetes type: 1 = type 1 diabetes mellitus; 2 = type 2 diabetes mellitus; 1,2 = both subjects with type 1 and type 2 diabetes mellitus included; GDM = gestational diabetes mellitus; 9 = diabetes type not specified and not clearly ascertainable from other information in the report; *= diabetes type not specified but ascertained by reviewers from other information in the report or from other sources, such as direct communication with the authors.

  3. cEffect: 1 = subjects with poorer glycemic control had poorer health than the comparison group(s); 0 = no difference in the periodontal health status between subjects with poorer glycemic control and comparison group(s).

  4. dDiabetes types are 1 and 2 for all but one subject who had drug-induced diabetes mellitus.

Karikoski and Murtomaa (2003)FinlandProspective1, 2, otherAdults0NoII-2
Tervonen et al (2000)FinlandCross-sectional1Adults1YesIII
Sandberg et al (2000)SwedenCross-sectional2Adults0YesIII
Tsai et al (2002)USACross-sectional2Adults1YesIII
Lu and Yang (2004)TaiwanCross-sectional2Adults1YesIII
Campus et al (2005)ItalyCross-sectional2Adults1YesIII
Chuang et al (2005)TaiwanCross-sectional2Adults0NoIII
Peck et al (2006)South AfricaCross-sectional2Adults1NoIII
Jansson et al (2006)SwedenCross-sectional2Adults1NoIII
Arrieta-Blanco et al (2003)SpainCross-sectional1, 2Mixed ages0YesIII
Guzman et al (2003)USACross-sectional1, 2*Adults1NoIII
Negishi et al (2004)JapanCross-sectional1, 2*,dAdults1NoIII

The reports included in Table 1 were restricted to studies which compared periodontal health in subjects with and without diabetes. This subject has attracted increasing attention with greater numbers of publications in consecutive decades, ranging from six in the 1960s, eight in the 1970s, and 12 in the 1980s to 20 in the 1990s. This review identified 17 reports published in the current decade starting in the year 2000. Table 1 presents a summary of the evidence on the relationship between diabetes and periodontal disease. Studies were broadly classified and ordered by type of diabetes and age of subjects (Table 1). In contrast to seven reports of prospective studies published prior to 2000, all of the studies identified for this review are cross-sectional and thus limited in their ability to provide evidence for causal inferences. There was one study of type 1 diabetes and it reported more extensive radiographic bone loss in participants with type 1 diabetes (Tervonen et al, 2000).

Regarding the relationship between type 2 diabetes and periodontitis the review identified 10 reports. One report comprised 15–45+ year olds (Endean et al, 2004), and nine (Sandberg et al, 2000; Orbak et al, 2002; Tsai et al, 2002; Zielinski et al, 2002; Lu and Yang, 2004; Campus et al, 2005; Chuang et al, 2005; Borges-Yáñez et al, 2006; Mattout et al, 2006) included only adults. Seven of these 10 studies reported significantly poorer periodontal health in subjects with type 2 diabetes, whereas no significant difference was discerned in a study of mostly older Taiwanese dialysis patients with and without ‘insulin-dependent (type II) diabetes’ (Chuang et al, 2005) as well as in a study of U. S. university clinic patients 60+ years of age with good medical and dental care comparing well-controlled (mean HbA1c = 7.3% with 70% having HbA1c > 7.5%) subjects with diabetes to subjects without diabetes (Zielinski et al, 2002); whereas in a study of Mexicans 60+ years of age there was a marginally significantly greater prevalence (= 0.09) of periodontitis in the group with diabetes (61.5%) than in the group without diabetes (49.5%) (Borges-Yáñez et al, 2006).

Several reports consist of analyses in which subjects with type 1 and type 2 diabetes were not distinguished. All of the studies in this subset were cross-sectional. One study included children only (Lalla et al, 2007), and all other studies included adult subjects, although one also included children or adolescents (Arrieta-Blanco et al, 2003). Two of these three studies reported greater prevalence, extent, or severity of periodontal disease for at least one measure or index of periodontal disease (Arrieta-Blanco et al, 2003; Lalla et al, 2007). One report did not find significant differences in periodontal disease between subjects with and without diabetes (Ogunbodede et al, 2005).

Two studies report on analyses on National Health and Nutrition Examination Survey III data from over 4000 women with a history of gestational diabetes (GDM) in the US. One report included ages 15–44 (Xiong et al, 2006) the other ages 20–59 (Novak et al, 2006). Both reports concluded there is a strong relationship between GDM and periodontal disease. Xiong et al (2006) found periodontitis in 45% of pregnant women with GDM vs 13% in the group without diabetes, with an adjusted odds ratio of 9.11. In non-pregnant women, 40% of women with type 1 or 2 diabetes, 25% of those with a history of GDM, and 14% of women without diabetes had periodontal disease. The odds ratio for those with type 1 and 2 diabetes was 2.76 (Xiong et al, 2006).Novak et al (2006) found the prevalence of periodontal disease to be higher in women with a history of GDM and concluded that women with at history of GDM may be at greater risk for developing more severe periodontal disease. A smaller Greek study of 34–36 weeks pregnant women also concluded gingival inflammation was more prevalent in the women with GDM (Mittas et al, 2006), but also found more plaque in that group.

As with other complications of diabetes, current evidence also supports poorer glycemic control contributing to poorer periodontal health. Primary research reports in the literature published since 2000 investigating relationships between glycemic control level and periodontal disease have included studies with subjects with type 1 diabetes exclusively (one study), type 2 diabetes exclusively (seven studies), or a combination of individuals with either type 1 or type 2 diabetes (three studies) (Table 2). Only seven of the 12 reports published regard the association between degree of glycemic control and periodontal disease specifically in type 2 diabetes (Sandberg et al, 2000; Tsai et al, 2002; Lu and Yang, 2004; Campus et al, 2005; Chuang et al, 2005; Jansson et al, 2006; Peck et al, 2006). Five of the latter found poorer glycemic control to be a significant factor associated with poorer periodontal health, the association was borderline significant in one study of dialysis patients (Chuang et al, 2005) and no difference was found in the remaining study (Sandberg et al, 2000). Among the studies providing information on differences in periodontal health classified by glycemic control status, most have been cross-sectional, with eight of 12 publications reporting more prevalent or more severe periodontal disease in those with poorer glycemic control (Tervonen et al, 2000; Tsai et al, 2002; Guzman et al, 2003; Lu and Yang, 2004; Negishi et al, 2004; Campus et al, 2005; Jansson et al, 2006; Peck et al, 2006) and four reporting no differences (Sandberg et al, 2000; Arrieta-Blanco et al, 2003; Karikoski and Murtomaa, 2003; Chuang et al, 2005). There was one follow-up study identified (evidence level II-2) that was published since 2000 (Karikoski and Murtomaa, 2003).

The preponderance of studies included in this review of reports published since 2000 on the adverse effects of diabetes on periodontal health are cross-sectional and describe findings of convenience samples, principally from outpatients in hospitals and clinics. While limitations on causal inference must be considered, these reports continue to support previous consistent evidence of greater prevalence, severity or extent of at least one manifestation of periodontal disease in the large majority of studies. The reports reviewed also provide additional evidence to support a ‘dose-response’ relationship, i.e., as glycemic control worsens, the adverse effects of diabetes on periodontal health become greater. Further, focused study of the relationship between gestational diabetes and periodontal health is emerging in body of literature.

Finally, the findings and conclusions from this review are consistent with two published meta-analyses that have provided quantitative summaries of the adverse effects of diabetes on periodontal health (Papapanou, 1996; Khader et al, 2006).

Periodontal disease: its effects on glycemic control and complications of diabetes mellitus

In addition to the substantial evidence demonstrating diabetes as a risk factor for poor periodontal health, there is a growing body of evidence supporting periodontal infection adversely affecting glycemic control in diabetes and contributing to increased risk for the pathogenesis of diabetes complications. Because of the high vascularity of the inflamed periodontium, this inflamed tissue may serve as an endocrine-like source for TNF-α and other inflammatory mediators (Offenbacher et al, 1996; Grossi and Genco, 1998). Because of the predominance of Gram-negative anaerobic bacteria in periodontal infection, the ulcerated pocket epithelium is thought to constitute a chronic source of systemic challenge from bacteria, bacterial products and locally produced inflammatory mediators. TNF-α, IL6, and IL1, all mediators important in periodontal inflammation, have been shown to have important effects on glucose and lipid metabolism, particularly following an acute infectious challenge or trauma (Feingold et al, 1989; Ling et al, 1995; Grossi and Genco, 1998). TNF-α has been reported to interfere with lipid metabolism and to be an insulin antagonist (Grunfeld et al, 1990; Feingold and Grunfeld, 1992). IL6 and IL1 have also been reported to antagonize insulin action (Ling et al, 1995; Michie, 1996; Pickup et al, 1997).

More direct, empirical evidence regarding the effects of periodontal infection on glycemic control of diabetes comes from treatment studies using non-surgical periodontal therapy and observational studies (Table 3). The treatment studies are a heterogeneous set of reports that include randomized clinical trials (RCTs) and non-RCTs. The RCTs used control groups that were either non-treated controls (Aldridge et al, 1995; Kiran et al, 2005), positive controls (Grossi et al, 1997; Rodrigues et al, 2003; Skaleric et al, 2004), or controls advised to continue with their usual source of dental care (Jones et al, 2007). Of the seven RCTs, four reported a beneficial effect for periodontal therapy (Grossi et al, 1997; Rodrigues et al, 2003; Skaleric et al, 2004; Kiran et al, 2005).

Table 3.   Effects of periodontal disease and its treatment on glycemic control: clinical and epidemiological evidence, ordered by evidence level
ReferenceStudy design Diabetes typeb,*No. subjects
a. Treatment (Age)
b. Control (Age)
Follow-up timePeriodontal treatment a. Treatment group
b. Control group
Metabolic control outcome measureEffects on metabolic control Evidence level a
  1. aHierarchy of evidence based on classification scheme used (U.S. Preventive Services Task Force, 1996) where: I = evidence obtained from at least one properly randomized controlled trial; II-1 = evidence obtained from well-designed controlled trial without randomization; II-2 = evidence obtained from well-designed cohort or case-control analytic studies, preferably from more than one center or research group; II-3 = evidence obtained from multiple time series with or without the intervention. Dramatic results in uncontrolled experiments (such as the results of the introduction of penicillin treatment in the 1940s) could also be regarded as this type of evidence; III = opinions of respected authorities, based on clinical experience; descriptive studies and case reports; or reports of expert committees.

  2. bDM type = diabetes type: 1 = type 1 diabetes mellitus; 2 = type 2 diabetes mellitus; 1,2 = both subjects with type 1 and type 2 diabetes mellitus included; 9 = diabetes type not specified and not clearly ascertainable from other information in the report; *diabetes type not specified but ascertained by reviewers from other information in the report or from other sources, such as direct communication with the authors.

  3. cFive subjects at most might have diabetes type 1, but the majority have type 2.

  4. d38 subjects were followed for 1 year and 22 for 2 years. PIDD: poorly controlled insulin dependent diabetes; CIDD: controlled insulin dependent diabetes.

  5. eControl group consists of healthy subjects without diabetes mellitus, not of subjects with diabetes.

Aldridge et al (1995): Study 1RCT1a. 16 (16–40)
b. 15 (16–40)
2 monthsa. Oral hygiene instruction, scaling,  adjustment of restoration  margins, and reinforcement  after 1 month.
b. No treatment
Glycated hemoglobin FructosaminePeriodontal treatment had  no effect on change in  glycated hemoglobinI
Aldridge et al (1995): Study 2RCT1a. 12 (20–60)
b. 10 (20–60)
2 monthsa. Oral hygiene instruction, scaling  and root planing, extractions,  root canal therapy
b. No treatment
Glycated hemoglobinPeriodontal treatment had  no effect on change in  glycated hemoglobinI
Skaleric et al (2004)RCT1All: 26–58 (41.8 = mean)
a. 10 (42.0 = mean)
b. 10 (41.6 = mean)
24 weeksa. Scaling and root planing +    minocycline microshperes  (Arestin®) in pockets  ≥ 5 mm at baseline and  at 12 weeks
b. Scaling and root planning
Glycated hemoglobinDecreased glycated hemo  globin in test and  control groups;
  Adjunct local Arestin®  treatment is significantly  more effective than scaling  and root planning only
I
Grossi et al (1996, 1997)RCT2a. 89 (25–65)
b. 24 (25–65)
12 monthsa. Either systemic doxycycline  or placebo and ultrasonic  bactericidal curettage with  irrigation using either  H2O, chlorhexidine, or  povidone-iodine
b. Ultrasonic bacterial  curettage with H2O  irrigation and placebo
Glycated hemoglobinThe three groups receiving  doxycycline and ultra  sonic bacterial curettage  showed significant  reductions (≤ 0.05) in  mean glycated hemo  globin at 3 monthsI
Kiran et al (2005)RCT2a. 22 (31–79) (56 = mean)
b. 22 (31–79)
(53 = mean)
3 monthsa. Scaling and root planing
  Oral hygiene instruction
b. No treatment
Glycated hemoglobin
Fasting plasma glucose
2-h post-prandial glucose
Decreased glycated  hemoglobin and  2-h post-prandial  glucose levels in  treatment group onlyI
Rodrigues et al (2003)RCT2a. 15 (unknown)
b. 15 (unknown)
3 monthsa. Initial full-mouth scaling and  root planingSystemic amoxicillin/clavulanic  acid 875 mg Oral hygiene instruction  at baseline Control/re-instruction and  prophylaxis every two weeksb. Same as a, except no  medicationGlycated hemoglobin Fasting plasma glucosePeriodontal therapy was  associated with improved  glycemic control expressed  as glycated hemoglobin and  fasting plasma glucose,  (but only significant  improvement in glycated  hemoglobin in b)I
Jones et  al (2007)RCT2*,ca. 82 (59 = mean)
b. 83 (60 = mean)
4 monthsa. Early Tx: scaling/root planing;   100 mg doxycycline daily   for 14 days; two daily   30 cc chlorhexidine rinses   for 4 months.
b. Usual care: usual dental and medical care
Glycated hemoglobin
Insulin use
‘The results….suggest that the addition of periodontal therapy to current medical therapy may have promise in regard to improvement of glycemic control’.
No significant differences between early treatment and usual care groups
I
Talbert et al (2006)Treatment study,  non-RCT2a. 25 (16–64)
b. 0
3 monthsa. Scaling and root planing
b. No control group
Glycated hemoglobin
Fasting insulin
Fasting glucose
Periodontal treatment did not decrease HbA1c levelsII-2
Smith et al (1996)Treatment study,  non-RCT1a. 18 (26–57)
b. 0
2 monthsa. Scaling and root planning   with ultrasonic and curettes;   oral hygiene instruction
b. No control group
Glycated hemoglobinFound no statistically or clinically significant change in glycated hemoglobinII-1
Westfelt et al (1996)Treatment study,  non-RCT1, 2a. 20 (45–65)
b. 20 (45–65)e
5 yearsa. Baseline oral hygiene   instruction, scaling and   root planing followed by   periodic prophys, OHI,   localized subgingival   plaque removal, and   surgery at sites with   bleeding on probing   and PPD > 5 mm
b. Same as group a
Glycated hemoglobin‘The mean value of HbA1c between BL-24 months was not signif different from that between 24–60 months’II-1
Christgau et al (1998)Treatment  study, non-RCT1, 2a. 20 (30–66)
b. 20 (30–66)e
2 monthsa. Scaling/root planing;   subgingival irrigation   with chlorhexidine; OHI;   and extractions
b. Same as group a
Glycated hemoglobinNo effect on glycated hemoglobinII-1
Taylor et al (1996)Historical  prospective  cohort2a, b. No tx or  control subjects
  49 (sev. periodis)
  56 (less sev.
  periodis)
2–4 yearsNot applicableGlycated hemoglobinThose with severe periodontitis were ∼6 times more likely to have poor glycemic control at follow-upII-2
Collin et al (1998)Retrospective  cohort2a, b. No subjects received  treatment
  25 with diabetes (ages 58–76)
  40 without diabetes
  (ages 59–77)
2–3 yearsNot applicableGlycated hemoglobinAmong subjects with type 2 diabetes the HbA1c level significantly increased in those with advanced periodontitis, but not in those without advanced periodontitisII-2
Schara et al (2006)Treatment study, non-RCT1a. 10 (26–55)  (38.6 = mean)
b. 0
12 monthsa. At baseline: Full-mouth   disinfection; At 6 months:   ultrasonic debridement;   scaling & root planing;   crown polishing; chlorhexidine gel,   rinse, and irrigation followed   by 14 days of chlorhexidine rinsing
b. No control group
Glycated hemoglobinReduction in HbA1c 3 months after each treatment, but not at 6 months post-treatment. [Baseline mean HbA1c = 10.5% (range 8.4–16.4%)]III
Seppala et al (1993, 1994)Treatment study, non-RCT1a. 38-1y; 22-2y  26 PIDD-1y (48 ± 6)d  12 CIDD-1y (43 ± 5)  16 PIDD-2y  6 CIDD-2y
b. 0
1–2 yearsa. Scaling and root planing,   periodontal surgery, and   extractions
b. No control group
Glycated hemoglobin blood glucoseReported an improvement of the HBA1 levels of the PIDD and CIDD subjects (= 0.068, t-test)III
Miller et al (1992)Treatment study, non-RCT1a. 10 (Unknown)
b. 0
8 weeksa. Scaling and root planing,   systemic doxycycline
b. No control group
Glycated hemoglobin Glycated albuminFound decrease in glycated hemoglobin and glycated albumin in patients with improvement in gingival inflammation (< 0.01);
Patients with no improvement in gingival inflammation had either no change or increase in glycated hemoglobin post treatment
III
Wolf (1997)Treatment study, non-RCT1, 2a. 117 (16–60)
b. 0
8–12 monthsa. Scaling and home care instr.;   periodontal surgery; extractions;   endodontic treatment; restorations;   denture replacement or repair
b. No control group
Blood glucose, 24-h urinary glucose
Insulin dose
Compared 23 subjects with improved oral infect. with 23 who had no improvem aft tx for oral infec. and inflam. The subj. with improved oral inflam. and infect. tended to demonstrate diab. ctrl. improvement (< 0.1). However, Wolf states in discussion, ‘tx of periodontal inflam. and periapical lesions… does little to improve the control of diabetes’.III
Iwamoto et al (2001)Treatment study, non-RCT2a. 13 (19–65)
b. 0
1 montha. Local minocycline in every   perio-dontal pocket and   mechanical debridement   once a week for a month
b. No control group
Glycated hemoglobinAnti-infectious treatment is effective in improving metabolic controlIII
Faria-Almeida et al (2006)Treatment study, non-RCT2All: 35–70
a. 10 (Unknown)
b. 10 (Unknown)5
6 monthsa. Scaling and root planing
b. Same as group a
Glycated hemoglobinSignificant reductions in HbA1c values from baseline to 3- and 6-months follow-up, respectivelyIII
Stewart et al (2001)Treatment study, non-RCT2a. 36 (DM+) (62 = mean)
b. 36 (DM+) (67 = mean)
18 monthsa. Scaling, sub-gingival   curettage, and root planing   Oral hygiene instruction
b. No intervention
Glycated hemoglobin
Changes in medications/ dosages
Periodontal therapy was associated with improved glycemic controlIII
Promsudthi et al (2005)Treatment study, non-RCT2a. 27 (55–80)
b. 25 (55–73)
3 monthsa. Mechanical perio treatment   and systemic doxycycline  100 mg daily for 15 days
b. No intervention
Glycated hemoglobin
Fasting plasma glucose
Test group: the reductions in the levels of fasting plasma glucose and HbA1c did not reach significance; ‘no association between periodontal treatment with adjunctive antimicrobial treatment and changes in HbA1c levels’III
Williams and Mahan (1960)Descriptive clinical study9a. 9 (20–32)
b. 0
3–7 monthsa. Extractions, scaling and   curettage, gingivectomy,   systemic antibiotics
b. No control group
Insulin requirement
Diabetes control (not operationally defined)
7/9 subjects had ‘significant’ reduction in insulin requirementsIII

An important source of variation in the RCTs is the use of adjunctive antibiotics with the non-surgical periodontal therapy. Among the RCTs, four included adjunctive antibiotics used systemically (Grossi et al, 1997; Rodrigues et al, 2003; Jones et al, 2007) or delivered locally (Skaleric et al, 2004). Three of these four RCTs using antibiotics showed beneficial effects on glycemic control (Grossi et al, 1997; Rodrigues et al, 2003; Skaleric et al, 2004). However, it is important to note the significant improvement for one study was in the positive control group that did not receive the systemic antibiotic (Rodrigues et al, 2003) and one of the four RCTs reporting a beneficial effect did not use antibiotics (Kiran et al, 2005). Hence, to date there is no clear-cut evidence to support a requirement for the use of antibiotics in combination with non-surgical periodontal treatment in order to observe an improvement in glycemic control associated with periodontal therapy.

Among the set of thirteen periodontal treatment studies that were not RCTs, eight reported a beneficial effect on glycemic control (Williams and Mahan, 1960; Wolf, 1977; Miller et al, 1992; Seppala et al, 1993; Seppala and Ainamo, 1994; Iwamoto et al, 2001; Faria-Almeida et al, 2006; Schara et al, 2006) and five did not (Smith et al, 1996; Westfelt et al, 1996; Christgau et al, 1998; Promsudthi et al, 2005; Talbert et al, 2006). Only two of these studies had control or comparison groups (Stewart et al, 2001; Promsudthi et al, 2005). Like the RCTs there was marked variation in the use of adjunctive antibiotics, with three of the five studies that used systemic antibiotics reporting a beneficial effect on glycemic control (Williams and Mahan, 1960; Miller et al, 1992; Iwamoto et al, 2001).

As shown in Table 3, there is marked heterogeneity in the studies’ designs, conduct, length of follow-up, types of participants, and periodontal treatment protocols. The details of the variation in this body of literature have been extensively described in several detailed reviews (Grossi and Genco, 1998; Taylor, 1999; Janket et al, 2005).

Additional evidence to support the effect of severe periodontitis on increased risk for poorer glycemic control comes from two longitudinal observational studies. A longitudinal epidemiological study of the Pima Indians in Arizona, USA (Taylor et al, 1996) found subjects with type 2 diabetes in good to moderate control and with severe periodontitis at baseline were approximately six times more likely to have poor glycemic control at approximately 2-years follow-up than those without severe periodontitis at baseline. In another observational study of 25 adults with type 2 diabetes, aged 58–77 years, Collin et al (1998) also reported an association between advanced periodontal disease and impaired metabolic control.

It is well recognized that poor glycemic control is a major determinant for the development of the chronic complications of diabetes. Results from the landmark Diabetes Control and Complications Trial (type 1 diabetes) and the UK Prospective Diabetes Study (type 2 diabetes) demonstrated that attaining and maintaining good glycemic control could reduce the risk for and slow the progression of microvascular complications in patients with type 1 and type 2 diabetes (Anonymous, 1993, 1998a,b) (Diabetes Control and Complications Trial Research Group, 1993). Additionally, the UKPDS observed a 16% reduction (= 0.052) in the risk of combined fatal or nonfatal myocardial infarction and sudden death. Further epidemiological analysis from the UKPDS showed a continuous association between the risk of cardiovascular complications and glycemia; every percentage point decrease in HbAlc (e.g., 9–8%), was associated with 25% reduction in diabetes-related deaths, 7% reduction in all-cause mortality, and 18% reduction in combined fatal and nonfatal myocardial infarction (Genuth et al, 2003).

There is emerging evidence from observational studies regarding the association between periodontal disease and the risk for diabetes complications. Thorstensson et al (1996) studied 39 case-control pairs of individuals with type 1 and type 2 diabetes for 6 years median follow-up time in Jönköping, Sweden. In each pair the cases had severe alveolar bone loss and controls had gingivitis or minor alveolar bone loss. They found that cases were significantly more likely to have prevalent proteinuria, and cardiovascular complications including stroke, transient ischemic attacks, angina, myocardial infarction, and intermittent claudication than controls at their follow-up medical assessments.

Two recent reports from the on-going longitudinal study of diabetes and its complications in the Gila River Indian Community in Arizona, USA, conducted by the National Institute of Diabetes and Digestive and Kidney Diseases, address nephropathy and cardiovascular disease. Saremi et al (2005) studied a cohort of 628 individuals for a median follow-up time of 11 years. Individuals with severe periodontal disease had 3.2 times greater risk for cardio-renal mortality (i.e., ischemic heart disease and diabetic nephropathy combined) than those with no, mild, or moderate periodontal disease. This estimate of significantly greater risk persisted while controlling for several major risk factors of cardio-renal mortality including: age, sex, diabetes duration, HbA1c, body mass index (BMI), hypertension, blood glucose, cholesterol, electrocardiographic abnormalities, macroalbuminuria, and smoking.

In the second report Shultis et al (2007) investigated the effect of periodontitis on risk for development overt nephropathy (macroalbuminuria) and end-stage renal disease (ESRD) in a group of 529 Gila River Indian Community adults with type 2 diabetes. Their proportional hazards models analyses, adjusted for age, sex, diabetes duration, body mass index, and smoking, indicated periodontitis and edentulism were significantly associated with the risk of overt nephropathy and ESRD. The incidence of macroalbuminura was 2.0, 2.1, and 2.6 times greater in individuals with moderate or severe periodontitis or in those who were edentulous, respectively, than those with none/mild periodontitis. The incidence of ESRD was also 2.3, 3.5, and 4.9 times greater for individuals with moderate or severe periodontitis or for those who were edentulous, respectively, than those with none/mild periodontitis.

The clinical and epidemiological evidence reviewed provides support for the concept that periodontal infection contributes to poorer glycemic control and the risk for diabetes complications in people with diabetes mellitus. However, further rigorous, controlled trials in diverse populations are warranted to firmly establish that treating periodontal infections can be influential in contributing to glycemic control management and possibly to the reduction of the burden of complications of diabetes mellitus.

Summary and conclusion

The evidence reviewed in this report supports previous conclusions that diabetes is associated with increased occurrence and progression of periodontitis and periodontal infection is associated with poorer glycemic control in people with diabetes. There is also evidence emerging that gestational diabetes may adversely affect periodontal health. Additionally, evidence is emerging to suggest that periodontal disease is associated with increased risk for diabetes complications. While treating periodontal infection in people with diabetes is clearly an important component in maintaining oral health, it may also have an important role in establishing and maintaining glycemic control and possibly in delaying the onset or progression of diabetes complications. Further rigorous, systematic study in diverse populations is warranted to support existing evidence that treating periodontal infections can be influential in contributing to glycemic control management and possibly to the reduction of the burden of complications of diabetes mellitus.

Author contributions

Drs. Taylor and Borgnakke both searched the literature for reports for possible inclusion in this manuscript. Both authors reviewed reports, conferred on which articles to include, and completed article assessment forms, designed by Dr. Taylor, to summarize the content of relevance to this literature review for each included report. Dr. Taylor designed the format for the tables and both authors contributed contents in the tables. Both authors drafted sections of the manuscript and contributed in responding to reviewers' comments, participated in final review of the proofs, and approved the proofs for publication.

Ancillary