| Sign In to gain access to subscriptions and/or personal tools. |
A Critical Review of Non-carious Cervical (Wear) Lesions and the Role of Abfraction, Erosion, and AbrasionDepartment of Prosthodontics, Floor 25, Guy’s Tower, St. Thomas’ Street, London Bridge, London SE1 9RT, UK Correspondence: * corresponding author, david.bartlett{at}kcl.ac.uk
The terms ‘abfraction’ and ‘abrasion’ describe the cause of lesions found along the cervical margins of teeth. Erosion, abrasion, and attrition have all been associated with their formation. Early research suggested that the cause of the V-shaped lesion was excessive horizontal toothbrushing. Abfraction is another possible etiology and involves occlusal stress, producing cervical cracks that predispose the surface to erosion and abrasion. This article critically reviews the literature on abrasion, erosion, and abrasion, and abfraction. The references were obtained by a MEDLINE search in March, 2005, and from this, hand searches were undertaken. From the literature, there is little evidence, apart from laboratory studies, to indicate that abfraction exists other than as a hypothetical component of cervical wear.
Key Words: abrasion • erosion • attrition • abfraction • tooth wear
‘Tooth surface loss’ or ‘tooth wear’ refers to the pathological loss of tooth tissue by a disease process other than dental caries (Eccles, 1982). Non-carious cervical lesions (cervical wear) are defined as the loss of tooth substance at the cemento-enamel junction (Mair, 1992). Terms also used to describe these are ‘cervical erosion/abrasion’ lesions and ‘abfractions’. The term ‘abfraction’ evolved from the work by McCoy (1982), Lee and Eakle (1984), and Grippo (1991a) and describes a theoretical process whereby occlusal forces create stresses in enamel and dentin along the cervical area and predispose it to erosion and abrasion. The aim of this paper is to critically review the literature and assess whether abfraction exists as a phenomenon.
Tooth wear can be separated into attrition, erosion, and abrasion. Attrition is defined as the loss of enamel, dentin, or restoration by tooth-to-tooth contact (Pindborg, 1970). Erosion is the loss of dental hard tissues by chemical action not involving bacteria (Eccles, 1982). It is further classified, according to the source of the acid, as either intrinsic or extrinsic. Intrinsic sources of acids originate in the stomach and are associated with eating disorders, such as anorexia and bulimia nervosa (Scheutzel, 1996), or with acid reflux and regurgitation (Bartlett et al., 1996). Extrinsic sources are acids contained in dietary components, such as carbonated soft drinks and fruit, and fruit juices (Zero, 1996; Lussi et al., 2004). Abrasion is the loss of tooth substance from factors other than tooth contact (Pindborg, 1970). Perceptions relating to the relative importance of erosion, attrition, and abrasion are geographically polarized, with apparently lower recognition in North America of the potential consequences of acids in tooth wear. This apparent conflict arises from a differing interpretation of the definitions relating to the etiology of tooth wear (Bartlett et al., 1999). Non-carious cervical lesions present in a variety of forms. Historically, these have been classified according to appearance: wedge-shaped, disc-shaped, flattened, irregular, and figured areas. Generally, they vary from shallow grooves to broad dished-out lesions to large wedge-shaped defects with sharp internal and external line angles. Clinical studies and observations have shown that cervical wear lesions are often situated on the vestibular surfaces of teeth, seldom on lingual surfaces and rarely on proximal surfaces (Kitchin, 1941). They are also more pronounced on incisors, canines, and premolars and more prevalent in the maxilla than in the mandible (Kitchin, 1941). There is some suggestion that the shape of the lesion is related to its etiology (Sognnaes et al., 1972; Brady and Woody, 1977; Lee and Eakle, 1984). One group of authors suggested, in a literature review, that those lesions with sharply defined margins could be caused by abrasive factors, whereas erosion produces broader, dish-shaped but shallower lesions (Levitch et al., 1994). The same authors classified palatal erosion as a non-carious cervical lesion, despite the clinical observation that this pattern of erosion is not limited to the cervical area.
The prevalence of cervical wear has been reported to vary between 5 and 85% (Bergström and Lavstedt, 1979; Levitch et al., 1994; Piotrowski et al., 2001; Aw et al., 2002; Oginni et al., 2003; Borcic et al., 2004). This large variation reflects the relatively few studies reporting the prevalence of cervical wear alone. Most studies report general levels of tooth wear in children and adolescents (Smith and Robb, 1996; Bartlett et al., 1998; Bardsley et al., 2004). From these studies, it is accepted that tooth wear is an almost universal condition, but that severe dentin exposure on non-cervical sites is relatively uncommon, at 2–4% (Lussi et al., 1991; Millward et al., 1994; Smith and Robb, 1996; Hugoson et al., 1996; Bartlett et al., 1998; Deery et al., 2000; Al Dlaigan et al., 2002; Dugmore and Rock, 2004; Bardsley et al., 2004). A few studies reporting the prevalence of tooth wear include data on cervical wear (Smith and Robb, 1996; Bartlett et al., 1998). These studies support the findings of others that cervical wear involving loss of enamel is common, but that exposure of dentin is less frequent, with quotes ranging between 2 and 6%. From the few studies reporting the prevalence of cervical wear, maxillary teeth seem to be more frequently affected, perhaps because of their lingual tilt (Levitch et al., 1994). Almost exclusively, the prevalence is reported on the buccal surfaces and rarely on the lingual surfaces of teeth (Khan et al., 1999). The teeth most commonly affected, from a study of 1002 patients from 4 dental practices in Croatia, are premolars and molars, with incisors being the least affected (Borcic et al., 2004). Enamel wear in the Croatian study was observed to be common, at around 60–70%, while dentin exposure varied between 0.6 and 5.6% of sites. In this sample, the prevalence and severity of cervical wear lesions appeared to increase with age, and this finding is supported by two other studies of 428 and 1007 adults recruited in practices in Turkey and the southeast of England, respectively (Smith and Robb, 1996; Akgul et al., 2003). Another study reported 10–35% of tooth sites with enamel cervical wear in a cluster sample of 210 adolescents in South London (Bartlett et al., 1998). Sangnes and Gjermo (1976) reported similarly high levels of enamel wear in the cervical region in 533 subjects. From these studies, it seems that cervical tooth wear of enamel is common, but this inference is based on relatively few epidemiological studies compared with the investigations on the prevalence of tooth wear as a whole (Table 1  ).
The role of abrasion has been investigated in the laboratory and clinically. Most of the early laboratory studies were based on the hypothesis that brushing with toothpastes containing abrasives will wear teeth. However, the values of less than 1 µm obtained in more recent laboratory and in situ studies suggest that the amount of enamel wear by toothpastes may not be clinically significant (Svinnseth et al., 1987; Joiner et al., 2004). There has been comparatively little work on dentin and abrasion, because the tissue is more complex and more difficult to measure (Ashmore et al., 1972). Early clinical work suggested that the orientation of the toothbrush influenced the wear of the teeth. From these studies, horizontal brushing was suggested as causing two to three times as much wear compared with vertical brushing (Mannerberg, 1960). This evidence was derived from scratches seen on replicas of teeth taken from 32 dental nurses. Prolonged contact time between bristles and tooth surface assessed in a laboratory study was reported to increase this further (Bjorn and Lindhe, 1966). Manly et al.(1965), in a laboratory study, reported that wear is, to some extent, dependent upon the force and frequency applied to the brush. The hypotheses of these early studies reflected the understanding at the time that cervical wear was caused by abrasion. A large cross-sectional study of 818 subjects showed an increase in the frequency of cervical lesions in patients (12% of the total) who brushed twice daily, compared with those who brushed less frequently (Bergström and Lavstedt, 1979). The results were supported by a smaller study on 100 military personnel of whom fewer than 50 had some cervical wear (Radentz et al., 1976). Although males might be expected to apply greater force during brushing, this may not influence the prevalence of cervical lesions when the genders are compared (Radentz et al., 1976; Sangnes and Gjermo, 1976; Bergström and Lavstedt, 1979). If the force during toothbrushing was important, it might be expected that lesions would be observed more frequently on the side of the mouth opposite the hand holding the brush (Kitchin, 1941; Sangnes and Gjermo, 1976). One study of 106 subjects reported that more lesions were found on the left side of right-handed subjects, but the difference was not statistically significant (Oginni et al., 2003). Another study on the effect of handedness on gingival cleaning and abrasion found no association with abrasion of the gingival tissues (Niemi et al., 1987). Overall, there appears to be little evidence for the effect of which hand holds the brush and the outcome on the tissues. However, other older laboratory studies indicate that the force of brushing varies with the brushing technique, the groups of teeth brushed, the stiffness of the bristles, the age of the brusher, and also individual brushing habits (Bjorn and Lindhe, 1966; Kakudo et al., 1969).
Some studies suggest that toothpaste has more relevance to abrasion than does the toothbrush (Litonjua et al., 2004a). Very small and insignificant changes on the tooth surface are caused by the toothbrush itself (Manly et al., 1965; Bjorn et al., 1966; Volpe et al., 1975; Bergström and Lavstedt, 1979). The initial lesions have been described as small scratches on the tooth surface (Mannerberg, 1960). There is some suggestion that softer brushes result in greater abrasion than do harder brushes, since they carry more paste (Dyer et al., 2000). But the toothbrush acts only as a device to carry the abrasive paste to the tooth surface. Even if the toothpaste does cause wear, the amount has been estimated to be a few microns rather than the more extreme wear seen clinically (Saxton and Cowell, 1981). Further doubts about toothbrush abrasion causing cervical wear were reported by Volpe et al.(1975) in a study of 120 subjects using 2 dentifrices. The results showed no correlation between development of cervical lesions and abrasivity of the toothpastes. In conclusion, from results of laboratory and clinical studies, there is little evidence to suggest that cervical wear lesions are solely caused by abrasion (Table 2
One of the first papers to introduce the concept that abrasion is accelerated with acid-softening or dissolution was a classic laboratory investigation by Davis and Winter (1980). This work has been supported by numerous laboratory studies showing that the combined effect of erosion and abrasion is greater than the effect of either operating on its own (Azzopardi et al., 2001; Eisenburger et al., 2003). One laboratory study comparing abrasion with erosion demonstrated a 50% increase in wear (Eisenburger et al., 2003). Despite this evidence, one study reported no statistically significant difference in wear after 5 erosive/abrasive cycles on extracted teeth (De Menezes et al., 2004). This study immersed bovine root dentin in common dietary drinks for 5 min and subjected the specimens to 5000 brushing stokes with a remineralizing phase. The remineralizing phase consisted of inorganic mineral ions, including potassium chloride and calcium phosphate, and was the likely reason why the acid’s effect was reduced. The choice of 5000 cycles was first proposed by Davis and Winter, but has been used by several different authors (Kelly and Smith, 1988; Bartlett et al., 1994; Azzopardi et al., 2001). However, the majority of laboratory and in situ studies support the hypothesis that a combination of erosion and abrasion increases wear (Jaeggi and Lussi, 1999; Hooper et al., 2003; Amaechi et al., 2003; Azzopardi et al., 2004; Attin et al., 2004). The combined etiology of cervical wear is further supported by an investigation analyzing study casts from 264 individuals and involving data from saliva and dietary questionnaires (Bader et al., 1996). The case-controlled study reported that cervical wear is a multifactorial process in both initiation and progression. There is only one clinical study that has investigated the progression of cervical wear lesions (Lussi and Schaffner, 2000). This six-year longitudinal clinical study reported the results for 55 people with cervical wear, and the authors observed that both consumption of dietary acids and frequency of toothbrushing correlated to increased wear.
If the combination of erosion and abrasion is cumulative, it might be expected that, theoretically, reversal is possible. Two different groups of researchers (Jaeggi and Lussi, 1999; Attin et al., 2004), using in situ studies, reported that if the tooth surface is not exposed to mechanical abrasion while in the softened state, remineralization can reverse the softening after a prolonged period of exposure to saliva. From this laboratory study, they surmised that a delay of at least one hour before brushing, but after an acid challenge, can increase the resistance of the tooth surfaces to abrasion (Jaeggi and Lussi, 1999; Attin et al., 2000). In conclusion, there is overwhelming evidence from laboratory and clinical studies that erosion and abrasion are linked in cervical wear (Table 3
Abfraction means ‘to break away’ (Braem et al., 1992). There remain many lesions that cannot be explained by toothbrush abrasion or erosion alone—for example, single-tooth lesions with unaffected teeth on either side. In addition, some studies have found no association between the incidence of lesions and horizontal toothbrushing (Bevenius et al., 1993). It is suggested that abfraction is the cause of such lesions. Abfraction lesions present primarily at the cervical region of the dentition and are typically wedge-shaped, with sharp internal and external line angles. Engineering studies have demonstrated that when teeth are loaded in a horizontal direction, the effect of the stress becomes concentrated in the cervical region, causing flexure (Hammadeh and Rees, 2001; Lee et al., 2002; Rees et al., 2003). McCoy (1982) proposed that bruxism may be the primary cause of angled notches at the cemento-enamel junction. The author postulated that tooth flexure from tensile stresses led to cervical tooth breakdown. Later, Lee and Eakle (1984) hypothesized that the primary etiological factor in wedge-shaped cervical erosions was the impact of tensile stress from mastication and malocclusion. The wear is suggested to be created by a combination of bending and barrelling deformations that cause alternating tensile and compressive stresses, which lead to a weakening of the enamel and dentin. The cyclic tension and compression may reach a fatigue limit and result in cracking or breakage of the tooth structure. At the same time, the opposite region comes under compressive stress. When the direction of the force changes (e.g., in bruxism), the tooth bends in the opposite direction, and the stresses correspondingly reverse at this cervical area. This bending of the tooth from side to side results in fatigue and fracture of the most-flexed zone. The interocclusal forces create physical microfractures, or abfractions, at the cervical region, which, in turn, result in wear. It has been reported that enamel is weak in tension (Powers et al., 1973), and the tensile phase of deformation therefore may cause disruption, allowing water and other small molecules to penetrate between the prisms and prevent reestablishment of interprismatic bonds on release of the stress (Levitch et al., 1994). Lee and Eakle (1996) suggested, in a review, that the cervical fulcrum area of a tooth might be subject to unique stress, torque, and moments resulting from occlusal function, bruxing, and parafunctional activity. These flexural forces would then disrupt the normally ordered crystalline structure of the thin enamel and underlying dentin by cyclic fatigue, leading to cracks, chips, and rupture (Spranger, 1995; Lee and Eakle, 1996; De Las Casas et al., 2003). Ultimately, the enamel breaks away at the cervical margin and progressively exposes the dentin, in which the process continues. Grippo (1991b, 1992) subsequently coined the term ‘abfraction’ and defined it as the pathological loss of tooth substance caused by biomechanical loading forces that result in flexure and failure of enamel and dentin at a location away from the loading. Grippo (1991b) suggested that abfraction is the basic cause of all non-carious cervical lesions, whereas Lee and Eakle (1984) proposed a multifactorial etiology, with a combination of occlusal stress, abrasion, and erosion. Spranger (1995), in a review of the literature, supported the multifactorial nature of the cervical area. The authors suggested that the wear was related to the anatomy, the distribution of forces calculated from elastic deformation studies, development of caries, and occlusion and parafunction. But much of the evidence for abfraction has been derived from finite element studies (Rees, 1998; Piotrowski et al., 2001; Rees et al., 2003; Geramy and Sharafoddin, 2003; Rees and Hammadeh, 2004). These computerized numerical procedures use discrete triangular elements, joined together to form nodes. The elements are ascribed certain physical properties, and a series of theoretical applied loads produces complex algebraic equations that are used to assess the impact of forces applied along the length of the model. Rees and coauthors (Rees, 1998; Rees and Jacobsen, 1998; Rees and Hammadeh, 2004) used this concept to estimate the effect of repeated loading on restoration of cervical cavities. The hypothesis suggests that continual occlusal loading produced displacements and stresses under the buccal cervical enamel and dentin, increasing crack initiation and encouraging loss of the restoration. The same authors subsequently investigated the effect of stress formation on unrestored teeth (Rees and Jacobsen, 1998). Lateral forces applied to the occlusal surfaces of premolars increased stress concentrations along the buccal cervical areas. Contradictory evidence from the finite element studies suggests that the lingual walls of teeth should be equally as susceptible to cervical wear as are the buccal walls (Rees, 2002), but this is not supported by the clinical findings, where lingual lesions are comparatively rare (Radentz et al., 1976). Further finite element studies showed that exposed dentin could be eroded by acid undermining enamel, causing more breakdown and increasing wear (Rees and Hammadeh, 2004). Although mathematical prediction models attempt to assess the impact of the periodontal ligament and the associated tissues, the actual role of the ligament in absorbing occlusal forces remains difficult to prove. Certainly, the theory behind abfraction seems to be plausible. Clinical and laboratory studies do not necessarily support such a simple etiology. A controlled study on extracted premolars with axial and non-axial loading indicated that no relationship between the direction of the load and the creation of cervical lesions could be established (Litonjua et al., 2004b). The same group investigated the effects of continuous and non-continuous loading on cervical wear on extracted premolars (Litonjua et al., 2004b). The authors reported that there was no evidence of microfracture, and that the application of occlusal load may not have a role in the progression of cervical lesions. In contrast, Staninec et al.(2005), in a laboratory study on thin sections of enamel and dentin taken from the cervical area of extracted unworn teeth, showed increased loss of mineralized tissue after stress and immersion in acid. The combination of erosion and stress upon the sections increased wear evenly over the length to a greater extent than when stress was applied singularly. But a clinical study investigating matching occlusal contacts to cervical wear lesions in 32 subjects reported no correlation between lesion dimensions and facet areas on the teeth (Piotrowski et al., 2001). These authors reported that control teeth had less gingival recession than did affected teeth, and therefore proposed that the lesions were more likely a result of toothbrushing than occlusal stress. Another recent study, from the articulated casts of 299 dental students, observed no correlation between occlusal wear and cervical lesions (Estafan et al., 2005). A study on 52 modern skulls, collected between 1920 and 1958, showed that attrition was a common finding, but of the 415 teeth examined, only 15 were found to have cervical wear lesions (Horning et al., 2000). The authors reported that, from the skulls, there was little evidence of occlusal forces causing either bone formation in the periodontal ligament or abfraction.
One paper reported the results of a study on 122 subjects with cervical wear (Khan et al., 1999). Cervical wear was compared with the appearance of the occlusal surface on the same tooth, and within the same subject, different occlusal lesions could be attributed to erosion or attrition. The study showed increased frequency of attrition on the occlusal surfaces in those with cervical wear. But they also showed an equal frequency between erosion on the occlusal surface with cervical wear lesions. The numbers of sites exhibiting an association between attrition and cervical wear were relatively low, and the authors did not report the results of any statistical tests. The results of the study add support to the multifactorial nature of cervical wear. Weak support was obtained from a clinical case study where strain gauges were used in three subjects (Nohl and Setchell, 2000). In the healthy subjects, increased strain developed along the cervical areas, but there was no information available to compare the results with those from controls. There is some evidence supporting the association between occlusal stress and cervical wear, but most of this is derived from finite element analysis and laboratory studies, with little direct data to confirm this clinically. In conclusion, there appears to be little evidence to correlate occlusal stress positively to cervical wear (Table 4
Stress corrosion and piezoelectric effects have also been theorized to have an effect on cervical wear (Grippo and Simring, 1995). More recently, new terms have been introduced to describe cervical tooth wear. ‘Biodental engineering factors’ (Grippo and Masi, 1991) have been defined as the effect of piezoelectricity at the cervical area, and ‘stress corrosion’ (Grippo and Simring, 1995) has been used to describe a multifactorial physiochemical degradation of the CEJ area. Also, ‘dental compression syndrome’ (McCoy, 1999) is tooth deformation related to malocclusion, parafunctional habits, and temporomandibular joint disorders. Several factors indicate a multifactorial etiology of cervical wear (Osborne-Smith et al., 1999). Cervical wear has been reported on the lingual surfaces of teeth in sites where toothbrush access is limited (Lee and Eakle, 1984). Cervical wear has been reported adjacent to restorations that remain unaffected (Braem et al., 1992), and, generally, mobile teeth are not as frequently affected as non-mobile teeth (Grippo, 1992). The hypothesis that occlusion has some influence has been recognized with increased incidence of cervical wear in bruxists as compared with non-bruxists (Xhonga, 1977). In this study on 30 subjects, the authors reported a significant association (p > 0.001) between the presence of occlusal and cervical wear. However, the authors grouped the subjects into bruxists and non-bruxists, based upon the appearance of occlusal wear. They defined erosion as wear occurring along the cervical margin. Therefore, those in the bruxist group, by selection, had more wear and consequently were more likely to develop cervical wear. There was no attempt to compare the location of the cervical with that of the occlusal wear lesions. This comparison was undertaken by Estafan et al.(2005), who compared cervical wear lesion with the occlusal contacts and observed no correlation between these factors in their study.
Tooth wear is a universal consequence of aging (Smith and Robb, 1996). The cause of pathological levels of tooth wear is difficult to diagnose and is generally a result of abrasion, attrition, and erosion (Smith and Knight, 1984). It is therefore difficult to select specific etiologies and make assumptions based on theories that assume that they do not co-exist with each other. Therefore, cervical wear lesions, like other forms, are probably created by a combination of erosion, abrasion, and attrition. There is strong support that erosion and abrasion are important in the development of wedge-shaped lesions along the cervical margins of teeth, but, as yet, there is insufficient evidence to confirm that abfraction truly exists. More research, particularly clinical, is needed to establish the validity of abfraction existing as an entity. Much stronger evidence suggests that the result of cervical wear is a combination of erosion, abrasion, and attrition.
Received for publication May 11, 2005. Accepted for publication November 8, 2005.
Journal of Dental Research, Vol. 85, No. 4,
306-312 (2006) This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
