Root caries and diabetes: risk assessing to improve oral and systemic health outcomes


Dr Pauline Ford
School of Dentistry
The University of Queensland
200 Turbot Street
Brisbane QLD 4000


This report reviews the current understanding of the relationship between diabetes and root caries, and investigates the evidence for dental chairside testing of gingival crevicular blood (GCB) glucose levels to assess risk for type 2 diabetes mellitus. Diabetes is linked with the progression of periodontal disease and a greater number of exposed root surfaces at risk for root caries. The rapidly increasing prevalence of type 2 diabetes coupled with a general trend towards retention of teeth means that the number of patients with increased risk for root caries is expanding significantly. Many patients with diabetes or abnormal blood glucose levels are unaware of their condition or that they are at increased risk for oral disease. Systemic blood glucose concentrations are comparable to those found in GCB and therefore may be a useful adjunctive clinical aid in determining appropriate care for patients and providing timely referrals to general medical practitioners. Use of GCB testing within the dental clinic is described. It is proposed that future studies be undertaken to provide clinicians with improved risk assessment strategies and to evaluate GCB glucose screening models.

Abbreviations and acronyms:

American Diabetes Association


advanced glycation end-products


Australian Type 2 Diabetes Risk Assessment Tool


cemento-enamel junction


gingival crevicular blood


general medical practitioner


US National Health and Nutrition Examinaton Survey


oral health therapist


random plasma glucose


As Australia’s population ages and retention of teeth increases, there will be increasing numbers of older patients at risk of root surface caries. Root caries has a higher prevalence among older adults than any other age group,1,2 and this is further increased among residents of long-term care facilities.3 Many of these individuals may be more likely to have chronic systemic disease. Root caries presents numerous restorative challenges and therefore prevention is a matter of increasing significance. Identification of patients at risk is essential for preventive strategies to be appropriate and effective. However, root caries risk assessment can be complicated by the presence of complex chronic medical conditions and, in particular, undiagnosed diabetes mellitus.

Type 2 diabetes mellitus is a chronic metabolic disorder which affects almost 1.6 million (7.4%) Australians.4 It is estimated that half of those living with type 2 diabetes remain undiagnosed.5 In its uncontrolled form, diabetes has the potential to severely impact oral and systemic health outcomes.4 Diabetes has been found to be bidirectionally linked with periodontal disease and subsequent loss of attachment.6 As a result, gingival recession can ensue, exposing the tooth’s root surface to the oral cavity and therefore contributing to the risk of root caries.7 Initial studies suggest that the dental setting provides a valuable opportunity to screen for this potentially overlooked risk factor. A standard glucometer (portable blood glucose testing device) can be employed to analyse gingival crevicular secretions for glucose concentration as part of a routine dental examination. This testing method is thought to be a reasonable estimate of glucose levels in the peripheral blood.8–10 Glucometric analyses which reveal an elevated blood glucose concentration indicate an increased risk of type 2 diabetes and require a referral to a general medical practitioner (GMP).8–10 With enhanced risk assessment for both root caries and diabetes, the dental team can better assist their patients in achieving positive oral and general health outcomes. Opportunistic screening for diabetes risk by the dental team is an illustration of the shift towards recognizing the importance of interdisciplinary health care and a holistic approach to health.

This review will provide dental practitioners with an overview of the current understanding of root caries, diabetes, and the links between diabetes and oral disease. The discussion of risk assessment strategies for root caries in this report will bring together the microbiological, clinical and metabolic indicators that may serve as markers of risk. Finally, diabetes screening will be reviewed for the dental context. The key message of this review is for dental practitioners to appreciate their important role as a member of the health care team responsible for the attainment of optimal general health for their patients. It is also important for the dental practitioner to understand the importance of diabetes and other general health conditions in determining oral disease risk and outcomes.

Root caries

The future of oral care will be significantly influenced by the challenge of root caries. Currently affecting an estimated 6.7%1 of the Australian population, the prevalence of root caries is predicted to increase alongside the ageing, dentate population.1 Studies investigating root caries report significant variation in terms of prevalence of the condition. These discrepancies11 can be attributed to studies reporting on populations with differing demographics, the lack of standardization of case definition and variations in measurement methods. As a result, the prevalence of root caries within the Australian population as a whole remains ill-defined at present.

Root caries is described as an oral disease that is expressed as a soft, progressive lesion found on the tooth root surface that has lost its connective tissue attachment and thereby has become exposed to the oral cavity.12 The lesions can be located at the cemento-enamel junction (CEJ) in degrees of a ‘ring-barking’ pattern, entirely on the root surface, spreading to undermine adjacent enamel or as recurrent lesions at a restoration margin.7,13,14 Lesions are most prevalent at proximal supragingival sites, often within 2 mm of the CEJ, yet lesions can occur subgingivally.12,15

Root caries poses a complex challenge for dental practitioners, which is different to those challenges presented by coronal caries.14 While the basic principles of disease that apply to coronal caries (i.e. host [tooth], organism [dental plaque biofilm] and substrate [oral fermentable carbohydrates]) also relate to root caries,14 the clinical management of this disease presents several difficulties for the clinician. Although root caries has the potential to initiate in the cementum of an exposed root surface, it is more likely that the majority of root caries is dentinal caries.16 The nature of dentine’s largely organic composition17 allows the structure to soften slowly when exposed to a prolonged low pH environment.14,16 If diagnosed and treated early in the disease process, this softened dentine has the potential to remineralize, without the need for surgical intervention.14,16 Non-surgical intervention in the treatment of root caries is preferable for several reasons. Dentine’s increased moisture content, as well as the root’s close proximity to the gingival margin, leads to the potential for moisture contamination during the restorative process.18 Root lesions develop within close proximity to the pulpal tissues, increasing the risk of bacterial ingress into pulpal tissues during cavity preparation, while also requiring a restorative material with comparable thermal conductivity levels.18 Degeneration of restoration margins leads to plaque retention and recurrent root caries, which remains increasingly difficult to manage.18 When considering these challenges, it is clear that the optimal approach for managing root caries is prevention. The ability to assess risk accurately and to implement appropriate preventive strategies will reduce the need for complex restorative treatments and associated dental morbidity.


Diabetes mellitus is a serious global health problem, resulting in substantial morbidity and mortality.19 The prevalence of diabetes mellitus has substantially increased over the last two decades and is expected to become the leading cause of disease burden by 2023.20 Defined by chronic hyperglycaemia, several classifications of diabetes exist. Type 1 diabetes, or insulin-dependent diabetes mellitus, is due to insulin deficiency and usually mediated by an autoimmune destruction of the insulin secreting beta cells in the pancreas.21,22 Type 1 diabetes constitutes approximately 10% of all diabetes cases in Australia.4 Type 2 diabetes is an acquired metabolic disease characterized by chronic hyperglycaemia, resulting from a defect in insulin secretion, insulin action, or both.5,22 Prediabetes is a condition in which blood glucose levels are elevated above the ‘normal’ threshold of <5.5 mmol/L, but do not satisfy the criteria for the diagnosis of diabetes mellitus.21,23 Glucose tolerance testing to detect impaired fasting glucose and impaired glucose tolerance is employed to diagnose these prediabetic conditions.23 Prediabetic patients display an increased risk of cardiovascular disease and progression to type 2 diabetes.23

Alongside conditions such as cancer and ischaemic heart disease, diabetes mellitus is listed as one of Australia’s top eight national priority conditions.24 Should type 2 diabetes remain undiagnosed or not be managed appropriately, long-term morbidity can include retinopathy, poor wound healing, nephropathy, neuropathy, cardiovascular disease,25 periodontal disease, limb amputation, psychological distress, and erectile dysfunction.5,26 During 2005–2006, 8.5% of all hospitalizations were associated with diabetes mellitus.20 This increased to 60% of hospitalizations for those aged 65 years and over, demonstrating that diabetes constitutes a significant burden for the older individual and for the health system.20

Risk for type 2 diabetes is influenced by both genetic and environmental factors. Non-modifiable risk factors for the disease include age, ethnicity and family history. The genetic basis of type 2 diabetes is stronger than for type 1 and there is a concordance rate of up to 90% for identical twins.27,28 However, the genes responsible are still largely unknown. Family history therefore is important, probably because it results in both genetic and behavioural/social inheritance of risk. The most important modifiable risk factor is obesity.27 Indigenous Australians are three times as likely to have diabetes in comparison to non-Indigenous Australians.26,29 Pacific Islanders, individuals of Chinese descent or from the Indian subcontinent,30 as well as women who have had gestational diabetes and their children, also maintain a higher risk of developing diabetes.5 Individuals of low socioeconomic status and those who reside in remote locations suffer a disproportionate burden of the disease in comparison to the general population.31

In 2004, type 2 diabetes was responsible for 2.7% of Australian deaths, while a further 8.9% of deaths recorded diabetes as an underlying cause.19,26 While the mortality rate for cardiovascular disease is falling, the increasing prevalence of diabetes in Australia offsets these life-expectancy gains.31 Diabetes also poses a significant burden on Australia’s economy and health system. The cost of this condition to the Australian health care system was estimated at $836 million during 2000–2001, representing 1.7% of the total national expenditure on diseases.26,32 Costs attributed to diabetes are predicted to increase between 2.5 to 3.5 times those for the year 2000–2001 by the year 2051.32 Furthermore, it is predicted that the prevalence of type 2 diabetes will continue to increase at a rate greater than that of general population growth. By 2051, the population aged 25 and over will have increased by 52%, while the population with type 2 diabetes will have increased by 128%.32

Diabetes and oral disease

The relationship between periodontitis and diabetes is now accepted to be a bidirectional association.6,33,34 It is widely understood that diabetes patients are at an increased risk for oral complications such as coronal caries, root caries, candidiasis, erosion, xerostomia and periodontal disease.35–39 Periodontal disease is an inflammatory condition in response to the presence of dental plaque.40 Chronic periodontal inflammation can result in loss of alveolar bone structure, loss of support for teeth and potential loss of teeth.40 Loss of attachment leading to gingival recession is common in periodontitis, and root exposure to the oral environment can follow.40 Consequently, these areas become a potential site for root caries.

The presence of diabetes mellitus modifies the risk for periodontal disease, with the potential to increase its prevalence, progression and severity.38 Poorly controlled diabetes increases the risk of developing periodontal disease three-fold.41 Increased systemic blood glucose concentrations translate to increased concentrations in the oral fluids, which may result in the proliferation of anaerobic microorganisms associated with periodontal pockets.42 The disease process also induces and accelerates inflammatory responses.6 Hyperglycaemia increases the concentration of advanced glycation end-products (AGEs). AGEs induce inflammatory responses that contribute to systemic degradation of connective tissues, including the periodontal tissues.6 Diabetic patients experience increased attachment loss, alveolar bone destruction and impaired wound healing due to the rapid degradation of localized collagen.41

Periodontal disease also has the potential to affect the diabetic condition. There is some evidence to suggest that a relationship exists between the clinical treatment of periodontitis and improving metabolic control in diabetic patients.43 This concept is of importance in reducing the risk of complications associated with uncontrolled diabetes and therefore reducing the associated burden. Patients with diabetes and severe periodontal disease display a three-fold increased risk of mortality from ischaemic heart disease compared with patients who have diabetes but a lesser severity of periodontal disease.44 Presence of periodontitis is also associated with a more than three-fold risk of end-stage renal disease in these patients.45

Obesity (body mass index ≥30 kg/m2 or a waist circumference over 102 cm for men and 88 cm for women) is well known as a risk factor for type 2 diabetes. A recent systematic review also provided evidence for an association of obesity with periodontitis.46 Obesity is an inflammatory condition associated with an increased presence of adipose tissue which is a source of inflammatory mediators. Elevated levels of systemic inflammation can induce insulin resistance and lead to diabetes. It is this contribution to the total inflammatory burden that underlies the hypothesized mechanism for the association between periodontitis and obesity.6,46

Currently, insufficient evidence exists to support or refute an association between diabetes and root caries. While biologically plausible explanations exist, there are too few well-designed human studies to support this hypothesis. One of the major reasons for inconsistencies between the studies47–50 is likely to be the failure to adequately account for confounding factors, particularly socioeconomic status and diet. In addition, few studies48,51–53 have examined root caries specifically and diabetes, and only one of these51 has shown a higher prevalence of root caries in diabetes patients. Animal models have demonstrated that periodontal disease and caries are exacerbated with diabetes.54 Salivary flow is known to be reduced in longstanding diabetes. This is thought to be due to neuropathy affecting the salivary glands as a result of chronic hyperglycaemia.55,56 Therefore, in diabetes, periodontal disease and associated attachment loss and gingival recession may mediate increased root caries, compounded by reductions in salivary flow and elevated gingival crevicular glucose levels in people with poorly controlled diabetes. However, even in the absence of periodontal disease, root surfaces may be exposed by other means, such as excessive brushing force and hence become at risk surfaces for root caries.

Risk assessment strategies for root caries

A definitive panel of risk factors for the assessment and management of root caries has yet to be defined. The presence of gingival recession and resultant exposed root surfaces is a common feature in ageing patients, thus increasing their susceptibility to root caries.16 Other predisposing factors for root caries include: living in residential care facilities; existing periodontal disease; xerostomia and/or xerostomogenic medications; systemic disease; reduced oral hygiene physical capabilities; past caries and restorative experience; removable dental prosthesis; tobacco use; lack of access to dental services (low socioeconomic status/low education level); eligibility for public dental services; and diabetes mellitus.1,14,18 It has been found that patients with fewer teeth record higher mean attachment loss scores and increased levels of coronal and root caries.57 Dietary risk factors contributing to the risk for root caries are: 2 to 9 intakes of sugar per day; and sucking sweets in the presence of a dry mouth.58,59 Low salivary buffering capacity, an increased number of missing teeth and existing coronal caries have also been shown to be associated with the development of root caries.51

Microbiological indicators

Research to date examining the microbiology of root caries has been considerably dominated by culture-based studies.60,61 Associated outcomes have shown partial repeatability, yet a definitive panel of organisms associated with root caries is yet to be established.11 Historically, root caries was thought to be associated with Streptococcus mutans, Lactobacilli (spp.) and Actinomyces (spp.).62,63 It has become apparent, however, that a more diverse range of oral microbiota may play a role in the root caries process. A number of oral organisms have been shown to be more frequently associated with root caries lesions.60,64 These organisms include Lactobacillus casei, Lactobacillus paracasei, Lactobacillus rhamnosus and Streptococcus mutans. Carious root surfaces had increased levels of yeasts, Lactobacilli (spp.) and Actinomyces israelii, while non-carious root surfaces were characterized by Streptococcus mutans/oralis/salivarius, Lactobacilli (spp.), Actinomyces naeslundii and Actinomyces gerencseriae.60Pseudoramibacter alactolyticus was found to be highly prevalent in dental plaque overlying non-carious root surfaces in patients with carious lesions in other sites.65 One study66 found higher levels of Treponema denticola, Prevotella nigrescens, Streptococcus sanguinis, Streptococcus oralis, and Streptococcus mutans in the supragingival plaque of patients with both type 2 diabetes and root caries. However, a limitation of this study was that plaque samples were pooled from all oral sites rather than specifically from root caries sites. Therefore, limited and inconclusive data exists on the specific relationship between diabetes and the root caries biofilm.

Clinical indicators

The clinical signs most often used to define a root caries lesion are based on colour, texture, surface smoothness, depth of lesion, and distinctiveness of the border, combined with whether the lesion is deemed to be active or inactive which is very subjective.7,12 Inclusion of secondary carious lesions adjacent to root surface restorations is a contentious issue and a potential barrier to straightforward study comparisons. Clinical investigators are in agreement regarding root caries being ‘soft’ when gently probed with an explorer, preferably a rounded-tip periodontal probe,7 yet employing measures of relative softness or hardness will result in disagreement among examiners.7 The colour of the lesion has been shown to have little correlation with caries activity status.67

Metabolic indicators

Assessing risk for root caries can be combined with assessing risk for type 2 diabetes. Strauss et al.68 used the National Health and Nutrition Examination Survey (NHANES) data to determine the proportion of dental patients who would meet the American Diabetes Association (ADA) guidelines for diabetes screening in the United States. Diabetes risk factors indicated in the ADA guidelines include: age ≥45 years; a body mass index ≥25 kg/m2; habitual physical inactivity; hypertension; hypercholesterolaemia; and a history of gestational diabetes.69 From the NHANES data it was shown that a total of 62.9% of those without periodontitis and 93.4% of those with periodontal disease should undergo diabetes testing. Of this indicated group, 60.4% had visited a dentist in the past 2 years.

Diabetes screening

Traditionally, diabetes screening is performed by GMPs. Recent trials of opportunistic diabetes screening in hospital accident and emergency departments have been found to be effective. These trials utilized a diabetes risk assessment questionnaire and a random capillary blood glucose test. Follow-up testing was performed at varying intervals after screening to confirm diabetes or prediabetes status. George et al.,70 Hewat et al.71 and Charfen et al.8 confirmed abnormal glucose tolerance in 4.2%, 8.9% and 26% of their respective study populations. However, the proportion of those identified patients attending for follow-up testing appointments differed among the studies. George et al.70 reported 100% return rates for those patients with diabetes risk factors and abnormal blood glucose levels. These authors performed the follow-up as part of the study and did not specify an interval between screening and follow-up. Charfen et al.8 lost 48.1% of those identified for follow-up testing after a six-week interval. Hewat et al.71 recommended that participants identified as at risk for diabetes consult with their GMP within one month of discharge and were followed up at two months. Of the group recommended for further assessment, 38% attended their GMP for further testing, while 23.6% reported being ‘too busy’ to attend.

Given the results of the accident and emergency department trials, it can be seen that screening can be an effective method for identifying impaired glucose metabolism. Not all patients will comply with the recommendation to consult with their doctor following a positive screening result; however, accident and emergency departments only cater for one-off encounters with their patients. The recall system employed by the dental profession on the other hand, would allow for a more systematic approach to follow-up and provision of health education interventions.

Screening for diabetes can easily be performed within the dental setting. Random plasma glucose (RPG) testing in the dental setting has been proposed previously, with the earliest report published over 40 years ago.9,10,72,73 This test requires no fasting period prior to the analysis and is performed irrelevant of the duration since the patient’s last consumed food or caloric drink. The NHMRC recommends an RPG cut-off of ≥5.5 mmol/L to describe elevated glucose levels, at which point referral to a GMP for assessment and diagnostic testing is indicated.21 While some providers have employed the traditional finger-stick method,74,75 it has been shown that routine periodontal examination produces ample blood volumes (0.3–5 μL required) for glucometric analysis during a dental examination.9,10,72 Studies have shown that gingival crevicular blood (GCB)9,10,72,76 glucose levels are a reliable correlate of peripheral blood levels, as long as an adequate volume of sample is obtained. Therefore, testing of GCB glucose levels may be a useful and less invasive means of screening for diabetes in the dental setting. One study77 was unable to determine a significant correlation between the sampling sites (peripheral blood and GCB); however, in this study an adequate GCB volume was not able to be obtained for approximately one-third of the patients sampled.77 Shetty et al.78 recently reported that they were able to detect and confirm diabetes in 40.7% of their periodontal patient population using reagent test strips to analyse GCB glucose levels. Alternatively, the GCB analysis can be performed using an inexpensive, readily available glucometer, such as those used for daily monitoring by patients with diabetes.9,10,72 GCB analysis is of particular interest given the appeal of an integrated diabetes screening procedure without the additional discomfort of the traditional finger-stick method. Furthermore, Greenberg et al.79 showed that 77% of US dentists felt that it was important for dentists to conduct screening for type 2 diabetes, while only 56% felt comfortable testing blood via the finger-stick method. Gingival crevicular fluid (GCF)80 has also been used to assess glucose levels; however, its correlation with peripheral blood was not as high as that of GCB samples.

In summary, while the finger-tip blood sample is the currently accepted standard sample for assessing blood glucose levels, GCB samples have been shown to be a comparable alternative. Since both patients and dental practitioners are more familiar and comfortable with intraoral interventions, GCB testing would appear appropriate and acceptable for dental clinic screening. Screening could occur as part of the regular examination process for the patient because if a periodontal assessment is being performed, the screening takes only one or two minutes more to incorporate. The extremely small volume of sample required means that almost all patients will have at least some suitable sites for collection of gingival blood.

The authors have recently trialled the use of GCB testing within a clinical dental setting with the aim of describing the process to assist practitioners in implementing the screening process in their own clinics. The trial was performed at a public dental clinic in Brisbane, Australia. Institutional ethical approval was obtained. The Accu-Chek Performa Nano (Roche Diagnostics Australia Pty Limited, Castle Hill, Australia) glucometer was selected, having a good sensitivity to detect blood glucose concentrations within a wide range (0.6–33.3 mmol/L), while requiring only a small volume of blood (0.6 μL) to perform testing. Other advantageous features were a quick sample processing time (5 seconds), hand-held design that was simple to use and was available for public purchase. A blood glucose concentration value of 5.5 mmol/L or above was considered to require a GMP referral for further assessment.21 An adapted version of the Australian Type 2 Diabetes Risk Assessment Tool (AUSDRISK)81 was used to assess risk criteria.

Use of the blood glucose monitor as a screening technique in the clinic proved to be a fairly simple and quick procedure in this trial. The monitor was easily set up. A small cardboard test-strip was loaded into the barrier-wrapped monitor just prior to use during the periodontal assessment (Fig. 1). Anterior test sites with a periodontal probing depth of at least 3 mm were preferable for site access and suitable GCB volume. Once a test site was selected, the area was gently cleared of any plaque or debris, rinsed and dried and reprobed to ensure an uncontaminated GCB sample (Fig. 2). The loaded test-strip was then brought into contact with the GCB, being careful to avoid contact with either tooth or gingival tissue (Fig. 3). The GCB sample was quickly collected and processed within 5 seconds (Fig. 4). The monitor could then be set aside and the periodontal assessment completed. The monitor was small, simple to barrier-wrap, and easy to operate and handle while wearing clinical personal protective equipment.

Figure 1.

 Accu-Chek Performa Nano. Blood glucose monitor with unused test strip.

Figure 2.

 Gingival crevicular bleeding following periodontal probing.

Figure 3.

 Collection of gingival crevicular blood using a test strip.

Figure 4.

 Blood glucose monitor with used test strip loaded and digital display showing reading in mmol/L.

The researcher who performed this trial was an oral health therapist (OHT). Risk assessment and health education are routine components of the OHT role in clinical care and should include relevant general health considerations, including diabetes. Expansion of this role to include offering and explaining a GCB glucose screen posed no problems for the OHT involved. When verbally questioned, participants found the screening procedure to be acceptable and were much more interested in learning more about diabetes and its importance to oral health as a result. This trial should now be extended to a larger study to incorporate an increased number of participants in order to determine the effectiveness and acceptability of this screening procedure on a larger scale.


Future research should determine the prevalence and incidence of root caries to quantify the burden of this disease within the Australian context for the general population and also for people with diabetes. Furthermore, research is required to understand the practices currently employed by dental professionals to diagnose root caries and to assess risk for the development of future lesions. A reliable panel of risk factors for root caries, perhaps including microbiological markers and GCB glucose levels would be very useful. In addition, the ability to screen for diabetes at a dental appointment would appear to be a very cost-effective measure to address the burden of diabetes. It is recommended that appropriate follow-up with screened patients is performed to ascertain whether or not the patient has been seen by a GMP and to determine the diabetes diagnostic testing outcomes. Furthermore, this screening should be combined with health education so that changes to health literacy and behaviours can be targeted where necessary. OHTs have expertise in risk assessment and health promotion and are therefore ideally placed to perform the screening and the accompanying health education. In order to assess the feasibility of diabetes screening within the dental setting, a cost-effectiveness analysis should be undertaken. Clinical trials should investigate not only the efficacy of the GCB glucose screening model, but also the patient and clinician acceptability of the procedures.


The authors would like to acknowledge the financial support provided by Clinical Education and Training Queensland (Oral Health), Queensland Health.