Root caries: a periodontal perspective

Authors


Abstract

Background and objective

A prevailing dental problem in the periodontal patient is root caries. Specifically, periodontal involvement often results in root surfaces becoming exposed and at risk for this condition. Periodontal therapy often leads to increased gingival recession as well, and the associated increased root caries risk may compromise the long-term success and survival of periodontally treated teeth.This narrative review will address the topic of root caries in the periodontal patient, focusing on unmet research needs.

Material and Methods

The Medline database was searched to identify items dealing with root caries, in terms of clinical features, diagnosis, pathogenic mechanisms and histopathology, as well as epidemiology, focusing then on the relationship between root caries and periodontal disorders.

Results

Although there is extensive literature on root caries, consensus is lacking regarding certain aspects, such as diagnostic criteria, prevalence within populations and indisputable risk factors. Advancing age could be an aggravating factor in susceptibility to root caries for the periodontal patient; however, definitive evidence in this regard is still missing. Similarly, full awareness of the increased risk of root caries in patients with periodontal disease or long-term periodontal treatment appears to be still lacking.

Conclusion

Research regarding root caries in age-specific (elderly) periodontal patients is needed. Improved oral hygiene practices, locally applied preventive measures, good dietary habits and regular dental check-ups are crucial approaches to prevent both periodontal disease progression and root caries. Periodontal patients with root exposure should follow a strict root caries prevention protocol, as an integral component of their periodontal maintenance therapy.

Gingival recession (GR) is defined as “the apical migration of the gingival margin beyond the cemento-enamel junction” [1]. This condition is inevitably associated with exposed root surfaces and the consequently increased risk of root caries (RC). GR is a multifactorial condition, with periodontal disease and inappropriate dental therapy being common contributing factors [2].

The purpose of this narrative review is to provide a periodontal perspective on RC in the periodontal patient; specifically, this article will provide an overview of the main pathogenic mechanisms leading to RC and of the preventive and therapeutic approaches to address these tooth root lesions in periodontally involved patients. The review focuses on: the role of periodontal disease and therapy as risk factors for tooth root exposure and RC; the benefits and limitations of periodontal surgical procedures, with or without adjunctive restorative treatments, as a viable prevention and treatment strategy for RC; and the significance of RC as a factor complicating or reducing the long-term prognosis of teeth treated with resective periodontal surgery. In addition, relevant aspects of RC in age-specific (elderly) populations are addressed. The review also highlights areas of unmet research needs in relation to RC and periodontal involvement.

Root caries: general features

RC commonly presents as a progressive lesion found on a tooth root that, due to some degree of periodontal attachment loss, has become exposed to the oral environment [3, 4]. Sumney et al. [3] defined RC as a cavitation below the cemento-enamel junction (CEJ), not including the adjacent enamel, usually discolored, softened, ill-defined and involving both cementum and underlying dentin. RC lesions have been variously described in the literature, where some authors proposed distinctive criteria such as location at the CEJ, position entirely on the root surface [5], spread to undermine the adjacent enamel [6] or extension to more than half on the root cementum [7].

RC lesions are most often located on exposed root surfaces, although it has been reported that up to 10–20% of lesions may occur subgingivally [8]. Banting et al. [9] reported that the most frequent RC location is close to the gingival margin, but Ravald et al. [10] identified the margin of previous restorations (51%), the CEJ (25%), and points of confluence with other lesions (17%) as areas at risk for RC in periodontally treated patients. Katz et al. [6] reported that RC are found – in descending order of frequency – at buccal and interproximal surfaces of mandibular posterior teeth, interproximal surfaces of maxillary anterior teeth, lingual and interproximal surfaces of maxillary posterior teeth and buccal and interproximal surfaces of mandibular anterior teeth. In contrast, Heegaard et al. [11] reported that RCs are frequently found on labial surfaces and evenly distributed within the dentition. Imazato et al. [12] reported canines and first premolars as the most frequently involved teeth, while Kularatne and Ekanayake [13] indicated instead the molars of both arches as most frequently involved. Regardless of possible relative differences in prevalence among sites and tooth type, RC can affect the entire dentition. Although RC lesions tend to spread along the CEJ and, in general, along the root surface, more aggressive RC lesions may progress toward the pulp similarly to dentinal caries of the crown [14].

Investigators have based the diagnosis and staging of RC lesions on location [3, 8, 15, 16], color [17-19], texture [6, 16, 20, 21], cavitation [21, 22] and contour [15, 23, 24] of the involved surface. Investigators have also tried to distinguish between active and inactive RC lesions based on characteristics cited above [20, 25, 26]. Gaengler et al. [27] classified RC into three main categories, based on clinical and histopathological features. Specifically, depth and progression of the lesion as well as degree of demineralization were considered to distinguish initial, stagnating and progressing lesions.

Diagnosis of RC is based primarily on traditional visual–tactile methods, although some reservations have been raised regarding the reliability and repeatability of this RC screening method [26]. In this context, radiographs and existing microbiological tests may supplement the clinical evaluation to increase sensitivity and specificity of the RC diagnostic process [28], which remains an issue. The difficulty in rapid screening and early detection of carious lesions is a general concern in cariology. The clinical detection of dental caries at the early stages, which precede cavitation, has always been a challenge. The majority of caries detection systems proposed over the years, both for coronal and RC, used a dichotomous system (cavitation/not) that did not measure the disease process at different stages, but only identified carious lesions at an advanced stage of their natural history [29]. Recently, the International Caries Detection and Assessment System (ICDAS) has been developed from the best aspects of previously published systems to address and overcome inconsistencies, set diagnostic threshold and criteria, as well as reliably and repeatedly identify and stage noncavitated lesions [29]. The ICDAS visual detection codes have been proven effective for assessing enamel and dentinal coronal caries in a reliable, valid and reproducible manner in both permanent and deciduous teeth [30, 31]. Unfortunately, despite the fact that ICDAS includes diagnostic criteria and decision trees specifically designed for RCs, data on the validity and reliability of this ICDAS section are not yet available, because no clinical, laboratory or epidemiological studies have been reported to date in this respect [32]. Moreover, RC diagnosis may present unique clinical issues, such as the unreliability of color [33] and texture [28] as reference diagnostic criteria or the reduced accessibility of the affected area to direct view in case of interproximal [6] or subgingival [8] lesions, especially if complicated by the presence of prosthetic crowns [34].

As clinical examination may be not sensitive enough to identify RC at the very early clinical stages, several technologies have been proposed to give the clinician the possibility to identify subclinical, noncavitated lesions. The use of emerging technologies aims to anticipate the diagnostic time, allowing clinicians to manage the caries process at an earlier stage of its natural history and take preventive rather than therapeutic measures. These technologies range from fluorescence-based or electrical caries monitors [35, 36] to optical coherence tomography [37]. These technologies remain to be conclusively validated before their use on a large scale to serve epidemiologic or dental practice needs.

In the context of RC diagnosis, conclusive validation of a RC clinical diagnostic system that is comprehensive of all clinical signs, but also rapid, feasible and repeatable, as ICDAS has been shown to be for coronal caries, appears to be necessary. Future research should address the in vitro and in vivo validation of ICDAS for RCs, taking into consideration intra- and inter-examiner agreement, to pursue standardization in diagnosis and staging; studies that address the specific challenges of RC diagnosis should be encouraged.

Root caries: etiologic factors

The main etiologic factors for onset and progression of RC in the elderly are the presence of bacteria and fermentable carbohydrates on the root surface [38]. Lactobacillus, Streptococcus mutans and Actinomyces salivary counts have been significantly correlated with RC occurrence [39]. Scanning electron microscopy observations revealed various patterns of bacterial coaggregation in RC lesions [40], while synergistic growth was observed in cocultures of Lactobacillus acidophilus with Streptococcus mutans, Actinomyces israelii with Lactobacillus acidophilus and Actinomyces israelii with Lactobacillus acidophilus and Streptococcus mutans, seemingly resulting in greater acidogenic and cariogenic effects on the root surface [39]. Although Candida albicans has been identified in root surface soft lesions, it is not considered an etiologic factor in RC onset [41-43].

Subjects with unaffected root surfaces show a more nonspecific microflora, with greater interindividual differences in biofilm composition, while in patients with RC the bacterial plaque of root surfaces tends to have lower interpatient diversity, decreasing further at affected roots [44]. Although these findings corroborate the hypothesis of a plaque-specific RC etiology, presently the microbiological agents causing this condition have not been conclusively identified [44]. The oral microflora is highly diversified and may comprise up to 600 bacterial species, many of which have not yet been cultured [45, 46]. Because of their inherent limitations, culture-dependent identification techniques pose the risk of limited or even misleading information, particularly when single site analysis is performed, given the site-specificity of microflora [44].

The use of culture-independent methods has allowed the discovery of microbial species not previously identified in the oral cavity [47]. Culture-independent, open-ended approaches comparing the biofilm of patients/sites with or without RC are advocated. The microbiological profile of patients with RC with periodontal disease or subjected to long-term periodontal treatment should be characterized as well, to clarify the possible impact of periodontal status on root surface bacterial biofilm and its role in RC pathogenesis.

Root caries: pathogenic mechanism and relevant bacterial metabolic pathways

Bacteria metabolize sugar into organic acids, which initiate root surface demineralization by removing calcium and phosphate ions from surface apatite crystals. For enamel, this process takes place as the pH reaches the critical value of 5.5; however, pH 6.4 is sufficient for cementum and dentin demineralization, due to their lower degree of mineralization [48]. Under normal circumstances, this loss of calcium and phosphate ions is balanced by the uptake of minerals from the surrounding microenvironment. An unfavorable oral environment, with high levels of pathogenic bacteria and rate of carbohydrate fermentation, may alter the balance of this demineralizingremineralizing cycle in favor of caries development.

In brief, some characteristics of dental plaque have been considered to be relevant with regard to cariogenic potential in several studies over the years; these include the dynamics of pH and free calcium concentration in the bacterial biofilm fluids after exposure to fermentable carbohydrate [49] and collagenase activity [50] with consequent organic matrix degradation [51], long-lasting glycogen synthetic and degradative activities at acidic pH levels [52, 53], and the ability to induce significant concentrations of microbial-derived organic acids with high dissociation constants – such as formic and pyruvic acids – in dentin matrix [54].

Brailsford et al. [55] suggested that not only the ability to produce acid from carbohydrates (acidogenicity), but also the ability to survive and grow in a low pH environment (aciduricity) might be a virulence determinant crucial for inducing RC. The authors proposed a dynamic model of bacterial etiology, with a progressive selection for a bacterial plaque with significant acidogenic and aciduric properties, that seems to include (but is not limited to) bacteria commonly believed to be involved, such as S. mutans and lactobacilli.

Recently, Hashimoto et al. [56] performed a pilot study in six subjects aged 48–73 years and used 16S rRNA sequencing analysis and anaerobic culture on blood agar plates to identify the proportions of protein-degrading and protein-coagulating bacteria in plaque samples collected from RC lesions, healthy supragingival sites and periodontal pockets ≥ 5 mm deep. In RC sites, a predominance of Propionibacterium, Actinomyces, Streptococcus, Lactobacillus and Bifidobacterium was found, with evidence of prevailing protein-coagulating metabolic pathways, and less represented protein-degrading activities. Subgingival periodontitis sites had the highest proportion of protein-denaturing bacterial strains, with specific predominance of protein-degrading ones, in agreement with previous findings [57]. The authors noted that periodontally involved root surfaces, exposed to oral environment and challenged by organic acids derived from fermentable carbohydrates are likely to have their organic components denatured and then degraded through the synergistic action of protein-degrading and protein-coagulating bacterial strains, resulting in the onset and progression of RC [56]. In this setting, robust host-derived proteolytic/collagenolytic activity (from periodontally involved gingival tissues) might play a contributory role in the RC process in patients with periodontitis.

Although the various pathogenic activities summarized above have been reported in the issue-focused literature, the complete RC pathogenetic puzzle has not yet been solved. Given the apparent significance of proteolysis as a pathogenic mechanism in RC it appears that targeted studies, on larger patient samples, with specific attention to protein-denaturing bacterial activities, as well as their potential association or synergism with host-derived proteases, could be helpful in elucidating the significance of the various mechanisms presently implicated in RC pathogenesis. In this context, intervention studies targeting host-derived proteolytic enzymes could be of help in illuminating the contribution of periodontally involved tissues to RC pathogenesis.

Root caries: histopathology

Root cementum and dentin are structurally different from enamel and coronal dentin and react differently to cariogenic challenges (Table 1) [14, 28, 58, 59]. As mentioned above, Gaengler et al. [27] classified RC into initial, stagnating and progressing lesions, based on clinical and histopathological features. Microstructural observations of initial RC lesions reveal an external layer with dentinal tubules partly occluded by peritubular and intratubular dentin deposition and an underlying layer of translucent dentin [60]. Dentinal tubules appear to be sclerosed by precipitation of calcium and phosphate ions, and tubules containing ghosts of bacteria and fine-granular crystals may be observed as well [61]. Stagnating RC lesions are characterized by a surface layer with high mineral content [62], absence of viable bacteria in dentin tubules [59, 63], impermeability to dyes and isotopes [64] and high resistance to acid and proteolytic enzymes [65] suggesting that arrest and remineralization of active lesions may depend upon the balance between severity of cariogenic bacterial infection of dentin and degree of active sclerosis of dentinal tubules in areas underlying the lesion.

Table 1. Comparison between coronal and root caries, adapted from Banting [28]
 Coronal cariesRoot caries
  1. GR, gingival recession; CAL, clinical attachment loss.

Etiology

Cariogenic bacteria (S. mutans, Lactobacilli)

Fermentable carbohydrates

Cariogenic bacteria (S. mutans, Lactobacilli, Actinomyces) [39]

Fermentable carbohydrates [38]

Predisposing factors

Plaque index

Frequency of carbohydrate intake

Reduced salivary flow

Fluoride exposure (−)

Plaque index [80, 88, 89]

Frequency of carbohydrate intake [79]

Reduced salivary flow [90, 228]

Fluoride exposure (−) [8]

GR/CAL [69, 79, 94, 95, 97-99]

Advanced age [81, 206, 207]

Low socio-economic status [95]

Reduced manual dexterity [93]

Cognitive decline [229]

Surface tissueEnamel/dentinCementum/dentin
Tissue composition (by weight)

Enamel:

 95–97% mineral

 3–5% organic and water

Dentin:

 65–70% mineral

 30–35% organic and water

Dentin:

 65–70% mineral

 30–35% organic and water

Cementum

 45%–55% mineral

 45–55% organic and water

Demineralization onsetpH ≤ 5.5pH ≤ 6.4 [48]
Carious process

In enamel:

Bacterial invasion followed by demineralization

In cementum:

Bacterial invasion followed by simultaneous demineralization and proteolysis [14, 58, 59]

In dentin:

Bacterial penetration of tubules; demineralization of intertubular dentin and proteolysis of organic component;

Sclerosis of dentin tubules, destruction of lumens and peritubular dentin deposition

In dentin:

Bacterial penetration of tubules; demineralization of intertubular dentin and proteolysis of organic component;

Sclerosis of dentin tubules, destruction of lumens and peritubular dentin deposition [14, 28, 58-60]

Conversely, in progressing RC lesions a superficial layer with diffuse bacterial penetration and an intermediate area of demineralizing dentin may be observed above the translucent dentin [14]. Progressing lesions may be clinically distinguished from stagnating ones by their surface texture and color: progressing RC lesions are commonly characterized by a soft and yellowish discolored surface, while in stagnating RC lesions a hard, brown to black surface may be observed [17, 66]. In more advanced lesions, the surface zone and demineralizing dentin are commonly not well demarcated, and a massive lateral spread of bacteria into intertubular dentin may be observed, with unaffected dentinal areas becoming continuously undermined. Notably, the mineral content in different zones within the lesion may vary considerably, and this aspect could result in underestimation of lesion depth [67].

Root caries: epidemiology (prevalence, incidence and risk indicators)

The apparent lack of unanimous consensus on the criteria to diagnose RC lesions and the variability of RC descriptors used in the literature [28] result in subjectivity impacted estimates of prevalence, incidence and severity of RC in the studied pop-ulations [8, 68]. The inclusion or exclusion of restored root surfaces and/or recurrent lesions in any assessment introduce further variability on reported prevalence rates [21]. Several population characteristics have been investigated as putative RC risk indicators in cross-sectional and longitudinal studies [41, 69-77]; unfortunately, the considerable heterogeneity among studies in terms of analyzed variables (e.g. sample size, follow-up period and analytical techniques) has prevented conclusive validation of a risk model and identification of population features that may significantly increase the risk of RC occurrence [78].

In general, all the identified predisposing conditions have to be considered risk indicators rather than actual risk factors for RC occurrence, because of the inability to establish cause–effect relationships through the available descriptive, cross-sectional studies and the limited agreement among reported longitudinal investigations. Table 2 summarizes the results of longitudinal studies, reporting the calculated incidence rates and the variables identified as putative risk factors for RC occurrence.

Table 2. Summary of cohort longitudinal studies reporting incidence rates and risk modelling (continued)
StudySample size (N)Observation period (months)Population characteristicsRC incidence (% subjects affected)Risk modelSignificant predictors
  1. RCI, Root Caries Index; Regr Coeff, Regression Coefficient; SE, Standard Error; CI, Confidence Interval; RC, Root Caries; rDFS, root Decayed Filled Surface; RPD, Removable Partial Denture; GR, Gingival recession; PD, Probing Depth; DFS, Decayed Filled Surface; DS, Decayed Surface; NR, Not Reported; Rel Risk, Relative Risk; DMFT, Decayed Missing Filled Teeth; FS, Filled Surface; CAL, Clinical Attachment Loss; OH habits, Oral Hygiene habits; BDLA, Basic Daily Life Activity.

Powell et al., 1991 (230)2312≥ 65 years old instituzionalized61.9%Logistic regression Variables Regr Coeff (SE)   
Gender1.06 [0.41]  
RCIlog0.42 [0.15]  
Ravald and Birkhed, 1992 (88)992433–76 years old community-dwelling, periodontally treated50%Stepwise multiple regression Variables R 2 Partial correlation p-value
Baseline RC0.180.43= 0.0002
Plaque score0.230.24= 0.0069
No. teeth0.280.24= 0.0246
Joshi et al., 1993 (89)1309–24 (median = 16)Middle-aged/older community-dwelling50.7%Stepwise logistic regression Variables Estimate (SE) OR (95% CI) p-value
Baseline RC0.13 [0.04]1.14 (1.05–1.23)= 0.004
No. teeth0.97 [0.39]2.63 (1.22–5.66)= 0.01
Plaque score0.99 [0.45]2.69 (1.11–6.50)= 0.02
Ravald et al., 1993 (80)27144≥ 46 years old community-dwelling periodontally treated88.9%Stepwise multiple regression Variables   p-value
Plaque score  = 0.00
Smoking  p < 0.05
Scheinin et al., 1994 (43)963647–79 years old community-dwelling51%Logistic regression Variables OR (95% CI)   
rDFS12.8 [2.8–58.5]  
Candida2.8 [0.9–8.1]  
Lactobacilli8.6 [2.4–31.2]  
Lawrence et al., 1995 (95)5536

≥ 65 years old community-dwelling

234 blacks

218 whites

29% (blacks)

39% (whites)

Logistic regression Variables OR (95% CI)   
Blacks:
RPD3.43 [1.52–7.76]  
Root fragments3.31 [1.28–8.58]  
Average GR1.75 [0.75–4.12]  
P. Intermedia 2.74 [1.27–5.92]  
Impact on appearance2.24 [1.09–4.60]  
Daily activities1.67 [1.12–2.5]  
Whites:
Worst GR4.5 [1.95–10.39]  
Average PD3.81 [1.24–11.69]  
Worst GR + average PD3.35 [1.28–8.76]  
Antihistamines4 [1.45–11.07]  
Calcium therapy2.45 [1.08–5.57]  
Age-related problems (> 40)4.99 [2.17–11.47]  
Employement status3.17 [1.32–7.61]  
Locker, 1996 (81)49336≥ 50 years old community-dwelling27.4%Logistic regression Variables OR  p-value
DFS model:
Baseline RC2.3 p < 0.001
DS model
Age2.5 p < 0.05
Regular check-up1.9 p < 0.05
Baseline RC2.2 p < 0.05
Lawrence et al., 1996 (22)70260

≥ 65 years old community-dwelling

379 blacks

323 whites

39% (blacks)

52% (whites)

Logistic regressionNRNR  
Powell et al., 1998 (42)5536≥ 60 years old community-dwelling77%Poisson regression Variables Regr Coeff (SE) Rel Risk (95% C) p-value
Bacterial countslog0.258 [0.085]1.29 [1.10–1.53]= 0.002
Asian ethnicity0.617 [0.237]1.85 [1.16–2.95]= 0.009
Gilbert et al., 2001 (75)72324≥ 45 years old community-dwelling36%Logistic regression Variables OR (95% CI) Estimate (SE) p-value
Baseline RC2.3 [1.4–3.7]0.83 [0.24]= 0.01
Baseline FS3.6 [2.4–5.7]1.29 [0.22]= 0.01
No. teeth1.9 [1.1–3.4]0.64 [0.30]= 0.03
No. lost teeth1.6 [1.0–2.5]0.46 [0.22]= 0.04
Chalmers et al., 2002 (229)21612

Older, institutionalized

103 dementia affected 113 control

62.1% (dementia)

44.2 (control)

Linear regression Variables β (SE) F p-value
Gender (male)0.071[0.291]0.059p < 0.8
Dementia0.945[0.287]10.824= 0.001
Takano et al., 2003 (69)3732470 years old community-dwelling35.9%Logistic regression Variables OR (95% CI)   p-value
Baseline RC3.71[2.07–6.67] = 0.00
Prosthetic crow2.33[1.24–4.39] = 0.009
CAL2.32[1.31–4.12] = 0.004
OH habits2.05[1.26–3.35] = 0.004
Fure et al., 2004 (41)10212065–85 years old community-dwelling12.4%Stepwise regression Variables Partial correlation β-coefficient  
Lactobacilli log0.372.78 
No. teeth0.450.85 
Drugs/day0.271.76 
Sánchez-García et al., 2011 (94)53112≥ 60 years old21.7%Bivariate and multivariate logistic regression Variables Bivariate OR (95% CI) Multivariate OR (95% CI)  
Limitations in BDLA2.2 [0.9–5.0]3.1 [1.0–9.5] 
Smoking2.0 [1.1–3.7]2.0 [1.0–4.1] 
Self-perceived oral health1.6 [1.0–2.7]1.4 [0.8–2.5] 
Dental mouthwash (no)1.6 [1.0–2.5]1.7 [1.0–2.8] 
Mutans streptococci (‡105 CFU⁄ ml)1.6 [1.0–2.7]2.1 [1.1–4.0] 
DMFT index (≥ 17)1.6 [1.0–2.5]1.3 [0.8–2.3] 
Exposed root surfaces (≥ 6)5.1 [3.2–8.2]5.4 [3.2–9.1] 
RCI ≥ 8%1.6 [1.1–2.5]1.8 [1.1–3.1] 
CAL ≥ 4 mm1.6 [1.0–2.5]0.8 [0.5–1.4] 

Among the numerous variables that have been investigated over the years as putative RC risk indicators, 30 have been shown to be significantly associated with increased RC incidence in at least one study. Specifically, RC prevalence at baseline, number of teeth at baseline, patient age, plaque index and salivary counts of Lactobacillus and S. mutans are independent variables that have been repeatedly shown to have a positive association with new RC occurrence during a 1–10 year (median = 3) incidence period [78].

Steele et al. [79] conducted a multivariate analysis on 462 patients (405 community-living; 57 institutionalized) aged ≥ 65 years; they used the presence of primary RC as dependent variable and behavioral or background characteristics as independent variables. Based on this study, there seemed to be a greater likelihood to develop RC for patients with a higher mean frequency of sugar intake, with an odds ratio (OR) of 2.4, infrequent tooth cleaning (OR = 4.7) and ineffective tooth cleaning in presence of a partial denture (OR = 1.6). The most relevant background risk variables were severe (≥ 9 mm) loss of clinical attachment (OR = 2.4) and exposed roots (OR = 1.05, per “vulnerable” tooth), while having sound coronal restorations was somehow protective (OR = 0.94 per restored tooth). A further stepwise multiple regression analysis was carried out considering as dependent variable the root caries index for decay, calculated as the proportion of all vulnerable teeth, i.e. with some degree of exposed root surface, being affected by active decay. Under such analysis, the habit of sucking sweets to relieve dry mouth (β = 0.101, = 0.037), living in an institution (β = 0.172, = 0.000) or not attending regular dental check-ups (β = 0.098, = 0.002) were the behavioral conditions associated with RC occurrence, in addition to the previously identified variables of wearing a partial denture and having infrequent personal tooth cleaning.

The potential contribution of some sociodemographic parameters to RC prevalence may be hypothesized but it is not easily evaluable due to lack of homogeneity among the relevant studies. Several cohort studies have revealed a wide range of incidence rates, varying by population characteristics, sample size and study length. Despite such differences in incidence rate, older, medically compromised or institutionalized subjects and subjects with advanced periodontal disease commonly exhibit higher RC incidence compared to the general adult population [9, 22, 41, 42, 70, 75, 80-83].

Social, psychological and cogni-tive status also seems to affect oral health conditions in noninstitutionalized elderly subjects. Avlund et al. [84] examined 157 independently living subjects, aged over 80 years, and reported substantial differences in oral health – measured as active coronal caries, active RC, edentulism and regular use of dental services – in two subgroups having different social class and education. In two other studies involving similar populations of community-dwelling individuals aged 80 years or older, cognitive decline, reduced functional ability as well as unsatisfactory social relations were significantly related to higher risk of active coronal caries and RC and, in general, poorer oral health and infrequent use of dental services [85, 86].

Therefore, on the basis of the available literature, the following variables should be considered among the main local factors involved in RC pathogenesis: high salivary counts of pathogenic microorganisms [41, 87]; high percentage of dental surfaces harboring plaque [43, 88, 89]; unrestored coronal caries and RC [69, 75, 88]; and factors promoting plaque accumulation, such as presence of restorations [10] and reduced salivary flow [90]. The population's systemic and behavioral conditions that might play a role in RC occurrence or progression include advanced age (41,72–74), low socio-economic status and educational level [84], discontinuity in oral health care and dental check-up [79, 86], frequency of carbohydrate intake [91, 92], impaired manual dexterity and cognitive decline [85, 93]. It must be recognized that, despite the fact that disease prevalence is frequently used as a rough estimate of risk, incidence data from longitudinal studies (i.e. cohort studies) may provide a higher level of evidence and more reliable risk estimates.

Surprisingly, the contribution of periodontal parameters to RC has received limited attention. Only a few studies [69, 94, 95] considered clinical attachment loss (CAL) and/or GR among the targeted putative risk indicators. Nevertheless, considering that GR is almost endemic in the adult population, as well as a supposedly necessary condition for RC occurrence in the majority of cases, it may not be proper, from a methodological point of view, to include GR as an independent variable in a risk model [78]. However, it is undeniable that, due to the unique spatial relationship between root surface and marginal periodontium, RC and periodontal disorders often coexist, not only as concurrent events, but also as conditions linked by two-way cause–effect relationships.

Root caries: relationship with periodontal disorders

In the context of RC predisposing clinical factors, particular attention should be paid to patients with periodontal diseases. CAL and GR have been recognized as relevant predisposing factors in the onset of RC, given the increased vulnerability of exposed root surfaces. RC is considered a serious problem affecting the long-term prognosis of both treated and untreated periodontally involved teeth [96]. Epidemiologic as well as large clinical trials agree that presence of GR and CAL are associated with a higher incidence of RC [79, 97-99].

Fadel et al. in a recent cross-sectional study [100] reported a high prevalence of RC lesions and high caries risk rates in 20% of patients referred for periodontal treatment, whether affected by gingivitis (mean age 25.8), mild-to-moderate periodontitis (mean age 41.9), or severe periodontitis (mean age 49.7); the prevalence of root lesions (caries and/or fillings) was 9%, 15% and 29% for each group, respectively. However, estimated RC risk profiles appeared correlated with a combination of several factors and not specifically with periodontal disease severity. The undeniable correlation between periodontal disorders and RC incidence rate emerged also from a RC prediction model involving a population of 698 subjects aged ≥ 60 years, where it has been shown that having ≥ 75 years of age (OR = 1.3, CI 95% 0.8–2.0), ≥ 6 exposed roots (OR = 5.1, CI 95%: 3.2–8.2) or ≥ 4 mm CAL (OR = 1.6, CI 95%: 1.0–2.5) are significant risk predictors for RC development in a 12-month period [94].

Root caries: relationship with gingival recession

As discussed above, despite the fact that a small proportion of RC lesions can occur subgingivally [8], in most cases GR might be considered a necessary condition for RC development [101, 102]. GR is highly prevalent in patients with either high [2, 103] or poor [104, 105] oral hygiene standard, but presents with different characteristics. Miller [106] categorized GR defects into four classes, based on the relation of the gingival margin to the mucogingival junction and the degree of interproximal CAL, highlighting how these clinical variables concerning GR are significant for root coverage outcomes [106].

Notably, in patients with good oral hygiene and without signs of inflammatory periodontal disease, GR is commonly located at buccal tooth surfaces and putative causative factors include traumatic tooth-brushing [107-109], eccentric orthodontic movement [110], tooth malposition [111], presence of alveolar bone dehiscence [112], or high frenal attachment [113]. In this subgroup of patients, GR defects are not associated with interproximal CAL and are typically classified as either Miller I or Miller II [2, 103, 114] (Fig. 1A); these types of GR defects are much more likely to fully resolve (i.e. to exhibit complete root coverage) after appropriate periodontal plastic surgery procedures [106].

Figure 1.

(A) Patient with good oral hygiene and multiple deep Miller I and Miller II gingival recession defects. (B) Severe Miller III and Miller IV gingival recession defects in a patient with periodontitis. (C) Severe periodontitis treated by osseous surgery (left panel); the resulting gingival recession is evident at the 1 wk postoperative visit (right panel).

Conversely, GR that manifests as a component of inflammatory periodontitis is associated with interproximal CAL [105], and may involve buccal, lingual and interproximal surfaces [2, 105] (Fig. 1B). This type of GR lesion typically belongs to Miller III or IV class [103], and is, therefore, unlikely to be completely resolved following regenerative periodontal surgical intervention [106, 115]. Regardless of the underlying etiology or associated attachment loss, GR lesions predispose to several different complications, such as hypersensitivity [116], loss of esthetic appearance [117], tooth abrasion [118], plaque retention [119] and RC [120]. The RC prevalence rate in subjects with GR was significantly higher (90%) than that (20–40%) reported in matched recession-free subjects [120].

Root caries: relationship with nonsurgical periodontal therapy

The interrelationship between RC and periodontal disease extends beyond the pathological entities per se. Periodontal therapy alone may lead to GR and thus to an increased vulnerability of the root surface to caries. While the crucial role of subgingival mechanical debridement (scaling and root planing) in the treatment of periodontitis is undisputed [121], the occurrence of GR development or increase following mechanical instrumentation has been widely reported [122-124]. GR development or pro-gression following scaling and root planing, especially for initially shallow or moderately deep probing depths [125], is particularly evident in patients with thin gingiva [126]. Cause-related periodontal therapy, through ultrasonic or hand instrumentation, acts on the root surface by removing or exposing cementum and dentin; these tissues are characterized by a lower degree of mineralization, limited amount of fluoride and poorer caries resistance than enamel [48]. Furthermore, repeated professional oral hygiene procedures may lead to an imbalance of normal competitive microflora, promoting the growth of cariogenic bacterial colonies on the root surface [127].

Ravald et al. (10,80,128,129) studied longitudinally the incidence of RC and the main reasons for tooth loss in a population of periodontally treated patients at four (128), eight (10), 12 (80) and 14 (129) years of maintenance periodontal therapy. The mean age of patients was 52 ± 10.6 years (range 30–78) at baseline and 64 ± 8.3 years (range 49–91) at the final examination. During the first 4 years of follow-up, approximately two-thirds of patients developed RC lesions. This incidence of new RC lesions was confirmed during all observation periods, exhibiting a certain degree of clustering in some patients and sporadic or recurrent correlation with advanced age [10, 80, 128, 129], salivary counts of S. mutans and Lactobacilli [10, 80, 128, 129], plaque score [10, 80], dietary habits [10], previous RC experience [10, 128], low saliva secretion rates [128] and smoking habit [80, 129]. In these studies, patients with ongoing periodontal problems were more prevalent (49%) than patients with RC alone (3%), and recurrence of periodontal breakdown was the main cause of tooth loss during the period of observation. However, a combination of periodontal problems and RC was evident in 20% of the considered population, and RC lesions alone accounted for 13% of failing teeth [129]. It should be emphasized that subgingival mechanical debridement is indicated, and must be performed, only in pathologically involved periodontal sites; subgingival instrumentation performed on physiologic crevices, with shallow probing depth values and absence of bleeding on probing, may result in iatrogenic CAL and GR, which could promote long-term RC occurrence.

The nonsurgical therapy of periodontitis may include systemic and/or local chemotherapeutic agents that target either the pathogenic biofilm or host-derived catabolic processes [130-133]. Although most of the chemotherapeutic drug studies are of short duration, which precludes consideration of RC development, even studies of sufficient duration (≥ 12 mo) have not included RC assessment [134-137]. Therefore, the potential impact of periodontal chemotherapeutics to RC incidence or prevention remains to be investigated.

Root caries: relationship with surgical periodontal therapy

Although surgical pocket elimination remains an important component of periodontal treatment, it can be anticipated that periodontal surgery, especially when incorporating resective procedures, will have an even greater impact than nonsurgical therapy in promoting increased RC incidence. Kaldahl et al. [138, 139] followed 82 patients with chronic periodontitis (initial mean age 43.5 years) treated by either coronal scaling, root planing, modified Widman surgery, or osseous resection surgery in a four-quadrant split-mouth design, evaluating probing depth, CAL and GR after 2 [138] and 7 [139] years of maintenance therapy. In general, osseous resective surgery produced both the greatest probing depth decrease and the greatest amount of GR compared with other therapies, showing the worst probing depth/GR ratio in sites of medium (5–6 mm) and shallow (1–4 mm) probing depth. Even though it is important to emphasize that sites treated by resective surgery seem to have the least incidence of breakdown during long-term follow-up [139], it is fair to point out that the potential significant amount of GR resulting from this surgical approach (Fig. 1C) may constitute a limit of the obvious and generally recognized clinical benefit of osseous surgery in certain cases.

In this context, when surgical intervention is deemed necessary, an alternative to osseous surgery, such as open flap debridement, may be considered to gain access to the affected root surfaces without drastically changing the periodontal tissue architectural relationships, thus minimizing post-surgical GR occurrence. Furthermore, as two determining factors for the prognosis of surgically treated sites are the quality of the patient's postoperative plaque control and the frequency of maintenance care appointments [140, 141], patients who have difficulty in maintaining a high standard of self-performed daily oral hygiene and/or regular attendance for dental check-ups are not the best candidates for resective periodontal surgery (or any periodontal surgical therapy, for that matter).

Other site-specific putative predisposing factors; noncarious cervical lesions

In sites affected by GR, it is not uncommon to observe an associated noncarious cervical lesion (NCCL) on the exposed root surface. NCCL is defined as the wear of tooth substance along the gingival margin of the tooth, due to mechanical abrasion, erosion or abfraction. The different underlying etiologies result in NCCL with diverse clinical characteristics.

Dental abrasion is frequently associated with repeated mechanical trauma from compulsive, excessive toothbrushing and use of abrasive toothpaste or hard toothbrush. The frequency and duration of toothbrushing, force applied and structure of dental tissues influence the degree of tooth wear in each individual [142]. Root surface abrasion lesions appear as well-demarcated cervical concavities, usually more broad than deep, showing a hard and smooth surface (Fig. 2A). Dental erosion is tooth structure dissolution without bacterial involvement, due to intrinsic (e.g. gastro-esophageal reflux) or extrinsic (e.g. acidic foods such as fruit juices) acid sources [143]. Root surface erosion lesions appear as not well-demarcated, hard and quite smooth areas (Fig. 2B). Abfraction lesions appear to be the consequence of tooth bending and strain, possibly due to eccentric occlusal forces [144]. They commonly present as deep, narrow, v-shaped notches on the facial aspect of the cervical area (Fig. 2C).

Figure 2.

(A) Patient with multiple noncarious cervical lesions, likely due to mechanical trauma from horizontal toothbrushing (abrasion lesions). (B) Gingival recession defects associated with noncarious cervical lesions apparently due to erosion from acidic foods and drinks. (C) Noncarious cervical lesions probably related to occlusal imbalance (abfractions). Frontal view (left panel) and lateral view (right panel); note the wear facets on canines and premolars. (D) Advanced noncarious cervical lesion resulting in pulpal exposure (endodontically treated left central incisor) in elderly patient; note the good level of oral hygiene and superimposed inactive carious lesion. (E) Carious cervical lesion superimposed on noncarious cervical lesion. (F) Multiple restored noncarious cervical lesions and coexisting buccal and interproximal root caries, under conditions of suboptimal oral hygiene, in an elderly patient.

Recent clinical and experimental findings suggest that it is the synergistic action of the various processes cited above, with mechanical causes such as abrasion and abfraction being potentiated by chemical dental erosion, which results in tooth structure wear, as opposed to the action of a single damaging mechanism [145]. NCCLs have the potential to evolve and in later stages result in dentin hypersensitivity [146], severe loss of tooth structure integrity and pulpal involvement (Fig. 2D). More importantly, for the context of the present topic, NCCLs may be complicated by superimposed RCs [147] (Fig. 2E).

It is widely recognized that NCCL and RC are distinct nosologic entities, and risk models generally do not consider NCCL as a predisposing factor for RC [78]. Nevertheless, these two pathological conditions share some clinical features (such as progressive impaired structural integrity of cervical dental hard tissues), pathogenic factors (such as cavitation from intermittent demineralization processes [148] and long-term consequences, as mentioned above; these shared traits establish a conceptual link between NCCL and RC. In this context, it is reasonable to postulate that the presence of NCCL, i.e. concave area near the gingival margin, may promote bacterial plaque accumulation on tissues with decreased microstructural resistance and thus result in RC development [149].

Specifically focusing on dental erosion, long-term high intake of acidic drinks or foods has been shown to reduce surface hardness of enamel and dentin [150]. Frequent exposure to erosive substances, in the presence of good oral hygiene, results primarily in dental erosion without caries. Hard tissue loss after acid exposure and immediate toothbrushing is significantly greater than acid erosion alone [151, 152] and, paradoxically, it would be advisable to avoid toothbrushing for at least 30 min after an erosive attack to protect dentin [153]. Nevertheless, in a patient who presents with multiple NCCLs a decline in self-performed oral hygiene standard can quickly lead to the onset of deep RC lesions superimposed on the previous inactive ones [154] (Fig. 2F).

Root caries: preventive and therapeutic considerations

From a clinical standpoint, there is general agreement in the literature that deep cervical lesions, regardless of their carious or noncarious nature, require a full understanding of their etiology; this will allow the clinician to provide the patient with appropriate information on lifestyle changes and adoption of preventive measures [145], as well as proper restorative treatment to preclude further damage [154].

Preventive measures

Proper preventive measures should be taken based on the estimated level of risk for the individual patient. In general, all modifiable behaviors that have been implicated, as outlined above, in the multifactorial model of pathogenesis should be addressed, to the extent possible. Particularly, prevention of RC, similar to coronal caries, is based on modification of harmful dietary habits, inhibition of demineralization, promotion of remineralization and reduction of the cariogenic dental biofilm [155].

The high consumption of low molecular weight carbohydrates, especially in industrialized countries, appears to be strongly implicated in the observed high caries prevalence, as organic acids resulting from the anaerobic metabolism of fermentable dietary sugars is the main cause of demineralization and cavity development [91]. Accordingly, patient education regarding diet, i.e. avoiding high-sugar containing foods and snacks and limiting the frequency of sugar intake, should be an essential component of a caries prevention program [91]. Nevertheless, sugar restriction may be a difficult goal to achieve for most individuals [156], considering that sweet taste may be responsible for addictive-like behaviors [157]. A possible support strategy to address this concern is to replace fermentable carbohydrates with noncariogenic sugar alcohols such as sorbitol or xylitol, to decrease sugar intake and thus reduce caries risk [158]. The replacement of sugar with xylitol was able to produce a substantial (85%) decrease in caries incidence in an adult Finnish population during a 2-year observation period [159].

Furthermore, when polyol sweeteners are delivered in the form of chewing gum, the chewing action in itself and the resulting salivary flow stimulation may provide further protective effects [160]. Accordingly, xylitol and sorbitol in chewing gum or candies were able to significantly reduce RC incidence in a geriatric patient population sample over a 6–30 month period [161]. Although the widespread adoption of a xylitol regimen for caries prevention appears to be nonsustainable, because of low cost-effectiveness [160], this prevention strategy remains viable when applied to high-risk populations such as disabled, elderly and most frail individuals [155].

Fluorides are widely recognized to be a key factor in the prevention and control of dental caries, through inhibition of demineralization and enhancement of remineralization. Topical fluoride treatments may be indicated, especially in patients with reduced salivary flow or impaired self-performed oral hygiene ability [162], but higher fluoride concentrations seem to be required for remineralization of RC in comparison with coronal caries [163]. There are several options available, such as daily 5000 ppm sodium fluoride toothpaste or 0.025–0.1% fluoride mouth rinses, neutral 1.2% sodium fluoride gel (in a 5-min tray), as well as sodium fluoride or silver diamine fluoride varnishes applied on to the exposed root surfaces three to four times a year; finally, fluoride tablets, chewing gum, toothpick and flossing have been proposed, although their use may be questioned due to unfavorable cost–benefit ratio [164].

Griffin et al. [165], analyzing through a systematic review with meta-analysis the effectiveness of various forms of fluorides in preventing caries in adults, reported that any form of fluoride administration (self- and professional application or water fluoridation) results in a RC prevented fraction of 22% per year. There is some evidence [166] of a synergistic effect of amorphous calcium phosphate and casein amorphous calcium phosphate complexes with fluoride, through the induction of a favorable chemical gradient of calcium and phosphorous at the tooth surface able to induce remineralization and reduced incidence of RC even in patients at particular risk [167].

Recently, encouraging in vitro results have been obtained testing a proanthocyanidin-rich grape extract, a widely available naturally occurring plant metabolite, in reducing root dentin demineralization during a pH cycling artificial caries protocol [168, 169]. Based on the principle that proper mineralization of a collagen-based tissue (i.e. root dentin) strictly depends on structure and stability of its extracellular organic matrix, the mechanism of action hypothesized for a proanthocyanidin-rich grape extract in inhibiting RC relies on increased collagen cross-links and reduced enzymatic degradation of the root collagen matrix [168]. A synergistic effect with fluoride has been speculated, but further studies are needed to deepen the knowledge about this promising natural agent.

Moreover, as already underlined, it is essential to motivate the patient to perform satisfactory oral hygiene routines and attend regular dental check-ups, as a good standard of oral hygiene is a necessary condition in any caries prevention program, but especially when addressing RC prevention [170] in the periodontal patient. Focusing on patients with a history of periodontal disease and therapy, typically exhibiting generalized CAL and numerous tooth surfaces with GR, the resulting anatomical complications must be considered a nonmodifiable risk factor; however, the patient's motivation to achieve a high standard of oral hygiene, along with close monitoring of at-risk sites, integrated with the patient's periodontal maintenance program, may be useful in preventing RC occurrence [171]. Antimicrobial agents, such as chlorhexidine, have been investigated as additional tools in reducing cariogenic oral biofilm [172], but – specifically considering RC prevention – the current lack of evidence on long-term clinical outcomes and ascertained side effects makes chlorhexidine devoid of any added clinical benefit compared to conventional and well-established preventive measures, i.e. diet modification, fluoride delivery and improved oral hygiene [173]. In patients with impaired salivary gland function, due to aging, radiotherapy, diseases or medications, the administration of salivary substitutes to buffer the acidic environment and inhibit adhesion and growth of bacteria might also be of help [174, 175].

The general preventive measures considered above are useful and effective in preventing RC by reducing the impact of risk factors throughout the oral cavity. In addition to general prevention, it is possible to implement site-specific measures that, from a periodontal standpoint, could represent an important clinical strategy to prevent and arrest RC [176-178]. Whenever possible, the ideal site-specific approach to RC prevention in GR cases, with or without initial RC or NCCL, is the restitutio ad integrum of the root–periodontium complex [102]. Complete root coverage will significantly diminish, if not eliminate, the possibility of RC development and progression for the specific site. State of the art of root coverage periodontal techniques have been the subject of several clinical trials [179-183] and systematic reviews [184-186]. Although great progress has been made in the surgical treatment of GR, and the available techniques show excellent results [185, 186], it must be emphasized that most, if not all, of the published studies have not considered specifically the patient with a history of inflammatory periodontal involvement. Therefore, clinical studies focusing on the periodontal management of GR in this class of patients are warranted. Given the progressing aging of the population, such studies will be particularly applicable to periodontal practice.

Restorative therapeutic approaches

The potential consequences of both RC and NCCL, e.g. dental pulp and/or marginal periodontal tissue involvement, hypersensitivity and compromised aesthetics, demand a timely and comprehensive therapeutic approach for each lesion. Such a site-specific approach may include the use of restorations (Fig. 3). Glass ionomer, resin-modified glass ionomer, resin composite, fluoride-containing resin composite or compomer restorations may be recommended based on individual patient risk [187]. In this context, it should be emphasized that unique aspects of the tooth cervical area, i.e. less than ideal characteristics as adhesion substrate [188], concentration of biomechanical stress [189], spatial relationship with the marginal periodontium and difficulties in access and field isolation, may render challenging the achievement of perfect marginal seal and long-term success of composite restorations used to treat RC. However, as restorative issues are not the focus of this review, the reader is referred to specific sources on this topic [154, 190].

Figure 3.

Root caries occurrence in patient with periodontitis treated with resective surgery and root amputation; clinical view at baseline (A) and at 3 mo (B), 2 years (C), 3 years (D) and 6 years (E, F) postoperatively. Recommended restorative procedure was performed (G, H) following the long-awaited patient consent.

Periodontal therapeutic approaches

Deep RC may occur, involving deep layers of dentin and/or affecting the root at a more apical level than the cervical area or the marginal gingiva. In such cases, the maximum apical extension of RC should be carefully considered to achieve complete removal of the decayed tissue and proper cavity restoration. RC that extends below the gingival margin may impinge on the biologic width and interfere with the gingival attachment apparatus [191].

A deeply placed restorative margin, violating the biologic width, will result in loss of periodontal support [192]. Tal et al. [193] considered in a beagle dog model class V cavities prepared so that the apical margin was in close spatial relationship with the alveolar crest (test) and other cavities not exceeding the CEJ (control). All cavities were restored and, at 1 year following treatment, GR and bone loss were significantly greater at test sites in comparison with controls. More recently, Günay et al. [194] reported increased inflammatory parameters and periodontal attachment loss at prosthetic restorative margins placed in proximity of the bone crest, i.e. within the biologic width. There is general agreement in recommending that when RC extends below the gingival margin the restoration must be preceded by, or be part of, a crown lengthening procedure (Fig. 4) to re-establish a proper biologic width apical to the proposed restoration margin [195].

Figure 4.

Root caries under fixed dental prosthesis; note the extension of the carious lesion apical to the gingival margin. Clinical view at initial presentation (A), during crown lengthening surgery (B), intraoperative restoration application (C), and 3 mo postoperatively (D).

Root caries: a risk factor for tooth loss after periodontal therapy

From a periodontal perspective RC is significant not only because periodontal disease (attachment loss) and therapy lead to conditions (i.e. GR) that favor RC development, as outlined in the previous paragraphs, but also because RC can be a leading factor compromising the long-term outcomes of definitive periodontal therapy. Longitudinal studies have shown that in cases of advanced periodontal disease, where resective therapeutic approaches (such as osseous surgery combined with root amputation, hemisection or tunneling procedures) have been employed, it is not uncommon for the initially successful therapeutic outcome to be compromised long-term, primarily because of complications other than periodontal ones, and especially because of RC [196]. In this regard, several authors reported RC occurrence to account for 7.9% (202) to 25% (203), and up to 50% (204) of the nonsalvageable complications leading to tooth extraction.

The frequency of periodontally involved teeth treated by resective surgery and subsequently condemned because of RC suggests that patients with periodontitis who are candidates for osseous surgery must undergo a stringent assessment for all endodontics, prosthetic and periodontal prognostic factors before implementing such a treatment approach [197]. Ultimately, the level of patient compliance and feasibility of an adequate maintenance program, in conjunction with an individualized caries prevention protocol during follow-up, will allow the patient who received resective periodontal therapy to remain free of RC and retain the function of his/her periodontally treated dentition.

Root caries and periodontal issues in the elderly

Longer life expectancy and decreased edentulism result in an ever-increasing number of elderly patients retaining their natural dentition in advanced age and being in need of special dental services [198]. Aging is associated with changes, such as reduced manual dexterity and altered oral anatomical structures, which can hinder the performance of proper oral hygiene. Senescence is also associated with reduced salivary flow, because of age-related structural salivary gland changes [199, 200] or functional impairment induced by various systemic disorders and medications. Hyposalivation is a debilitating condition that can potentially induce difficulty in chewing and speaking, as well as taste loss, dysphagia, and increased incidence of dental caries and soft tissue pathologies [201, 202]. Furthermore, other factors can also contribute to deficiencies in the structural integrity of teeth in elderly people through an incremental effect over the years: dietary acid exposure, oral parafunctional habits, occlusal trauma, attrition, abrasion from toothpastes or powders, erosion by alteration of oral pH due to medications or gastro-esophageal reflux [145].

Consistently, the prevalence of RC in adults has been shown to progressively increase with age [203-205]. The 1988–1991 NHANES III national survey [203] provided evidence for RC prevalence among nearly 6726 people; 55.9% of patients aged ≥ 75 years had one or more RC lesions, with severity being age-dependent. The NHANES III study established that, past the age of 34 years, the expected probability for a person of having one or more decayed or filled root surface will be 20–22% less than his age. This means that for a person aged 65 years, a 43–45% probability of having one or more RC may be assumed.

More recently, different prevalence rates in the elderly have been reported in the literature, ranging from 28.4% to 96.5% [12, 13, 206-210]. This wide range of RC prevalence probably reflects the highly diverse ethnic, cultural and social background of the populations examined in the aforementioned studies; such differences are likely to represent different dietary habits, standards of oral hygiene, and other relevant parameters, such as availability of water fluoridation, community-dwelling or institutionalized subjects, and urban or rural inhabitants. Nevertheless, it is clear that RC in the elderly is a problem with a global impact. In older adults it is not unusual to find RC, especially at the margin of previous prosthetic restorations, which often extend below the gingival margin.

Increased RC susceptibility of the elderly may be due to age-specific behavioral and socio-demographic characteristics, as mentioned above. However, from a clinical point of view, the main local factor that may be associated with increased RC prevalence is root surfaces becoming exposed simply due to aging or progressive periodontal disease [41, 69]. Coherently, in periodontal patients the estimated RC risk profile appears correlated with a number of predisposing factors but as reported above, the prevalence increases with the amount of CAL and with advancing age [108]. A correlation with advanced age recurs also in studies, including periodontally involved patients receiving long-term supportive periodontal therapy [10, 80, 128, 129]. In this regard, the cumulative effect over time of repeated professional root instrumentation procedures should be considered.

With respect to resective periodontal surgery, age per se is not a contraindication for this approach, which – when used in older adults – seems to provide clinical outcomes similar to those obtained in younger patients [140]. However, careful consideration concerning the advisability of performing resective surgery in the elderly patient is necessary [211]. Considering the documented increased RC risk with advancing age [41, 69, 80] and greater GR [79, 97, 99], the overall therapeutic advantage of radical surgical pocket elimination should be carefully weighed in the context of the overall dental health management of the older adult.

As previously mentioned, furcation-involved molars treated with flap surgery and root resection are prone to the development of a series of complications, including RC [197, 212, 213]. It is reasonable to assume that in elderly populations the coexistence of several predisposing factors, such as hyposalivation or impaired manual dexterity may result in increased postoperative RC incidence; nevertheless, targeted studies on age-specific populations, which may corroborate this assumption, are still lacking [196].

Similarly, knowledge on some other issues regarding RC in the elderly is still scarce. Current models of RC microbiological etiology are based mainly on studies of young and middle-aged adults. Bacterial species other than those commonly found have been isolated in RC lesions of elderly patients [214], suggesting that the putative RC etiologic agents in older adults belong to microbial communities more complex than previously presumed [44, 87]. Recently Preza et al. [215], using an rRNA-based, culture-independent, microarray technique in a sample of 41 patients aged 70 or older, did not find specific strains to be consistently associated with RC, suggesting the possibility of a certain degree of intersubject variability regarding RC bacterial etiology in this age-specific setting and the need for additional research. Moreover, although the caries process is endemic in the elderly and more than 50% of all individuals retaining their dentition after age 65 have at least one RC lesion [205, 216, 217], clustering has been observed so that only a third of the older adults account for the majority of RC cases [218-220]; the latter observation suggests the possible contribution of yet unknown genetic factors [90]. Specifically focusing on patients with periodontitis, the intense protein-degrading activity found at RC-affected and periodontally involved sites may suggest the possibility of a special role for protein-degrading and protein-coagulating bacterial strains in the onset and progression of RC [56]. This putative pathogenic mechanism deserves further investigation, with special emphasis on the elderly.

Other special issues of RC in the aging patient should be considered as well. RC diagnosis in elderly people may be complicated by the necessity to perform examinations under nonideal clinical settings [221] and the reduced ability to maintain a good standard of oral hygiene, resulting in greater plaque deposits on the root surfaces [222]. Such diagnostic difficulties lessen the reliability of the reported epidemiological data in the age-specific populations and, more importantly, may result in inadequate patient care.

As regards prevention, a good standard of oral hygiene may be a challenge for the elderly, because of physical and mental disabilities. Changes in dietary habits and gravitation towards softer foods rich in simple sugars – typical of the elderly population – may complicate further prevention efforts. Therefore, elderly patients need to receive adequate motivation and support for the necessary changes in diet, with particular emphasis on the frequency of fermentable carbohydrate ingestion and the possible replacement with noncariogenic sweeteners [161]. Moreover, instructions to maintain the best possible plaque control using electric or manual toothbrush, floss, interdental brush and adjuvants such as chlorhexidine and fluoride mouthwashes, and fluoride or anti-plaque dentifrices [223] should be provided, especially in elderly patients with a history of periodontal involvement. Unfortunately, other age-specific socio-demographic characteristics, such as institutionalization, lower socio-economic status and age-related cognitive decline, may severely impair effective patient communication or essentially curtail the patient's ability to comply with periodontal maintenance.

In conclusion, advanced age, periodontitis and RC are a triad of conditions frequently interrelated. The recent epidemiological data indicating that moderate or severe periodontitis affects 64% of adults aged ≥ 65 years in the United States [224] underscores the relevance of periodontal disease for the older adult. Therefore, future studies on RC in periodontally involved elderly populations are needed to better understand epidemiologic, diagnostic, preventive and therapeutic aspects predominantly or uniquely relevant to this patient subgroup.

Root caries and periodontal disease: need for greater awareness

Although the evidence linking periodontal disease to RC development is well established, it appears that implications of the interrelationship between these two clinical entities have not been routinely considered in the course of RC investigations. Despite the fact that GR and, in general, periodontal attachment loss are biological factors essential for RC onset and progression, often these variables are not investigated in the proposed risk models [78]. The inclusion of these parameters in future investigations should help determine further the significance of the periodontal component in RC pathophysiology.

Despite the recognition of the significance of RC prevention protocols as integral parts of a periodontal maintenance program [80, 225], it appears that RC prevention is commonly not given due weight and importance during periodontal supportive therapy [226]. Given that high patient compliance with periodontal maintenance has been shown to result in protection from further attachment loss and the onset of new RC [227], a fact that underscores again the interrelationship between the two conditions, it is evident that periodontists, hygienists or other oral healthcare providers responsible for periodontal maintenance, can play a key role in RC prevention.

Specifically emerging from this review, future research should assess provider awareness about the impact of RC as a risk factor for further tooth loss in periodontally compromised patients of advanced age, with the inevitable consequences of edentulism and reduced quality of life. Clinical studies regarding the effect of specific RC prevention protocols – possibly based on targeted check-ups, dietary advice, improved oral hygiene techniques supplemented by topical fluoride therapy – long-term results of periodontal maintenance care and rate of tooth loss would help improve the oral care of older patients and should help increase awareness of the significance of RC in this growing segment of the population.

Conclusions

This review highlighted the need for careful consideration of the potential role of RC in compromising the long-term survival of the dentition of periodontal patients. RC appears to be an underestimated clinical problem, despite its high prevalence and contribution to tooth loss in this patient population. The exposure of root surfaces to the oral environment due to aging, i.e. progressive GR, makes them prone to carious involvement. The risk of RC in patients with a history of periodontitis and resective surgery, especially in multi-rooted teeth, is of particular concern. Therefore, general preventive measures such as noncariogenic diet and fluoride application together with plaque control, must be regarded as mandatory components of any periodontal maintenance program. Careful assessment of each individual case is needed to outline a patient-centered comprehensive preventive and therapeutic approach for RC lesions. The high prevalence of periodontitis in older adults, increasing life expectancy and decreasing edentulism of the population, and apparent coexistence of RC with advanced age and periodontitis, collectively suggest that RC could become an increasingly frequent challenge for periodontal care providers. From this standpoint, future studies are needed to address special diagnostic, preventive and therapeutic measures to best meet the needs of the elderly periodontal patient.

Acknowledgements

We gratefully acknowledge Dr Luca Landi and Dr Takuichi Sato for the scientific material provided.

Ancillary