Previous retrospective analyses prove that impacted mandibular third molars (M3s) increase the risk of angle fractures and decrease the risk of concomitant fractures to the condyle. The authors have attempted to verify these relationships and identify the underlying mechanism of injury. A retrospective cohort was designed for patients attending the Division of Oral and Maxillofacial Surgery from January 2001 till October 2008. The primary predictor variable was M3. The secondary predictor variables were: M3 position, classified using the Pell and Gregory system; angulation, classified using Shiller’s method; and the number of visible dental roots. The outcome variables were angle and condyle fractures. Hospital charts and radiographs were used to determine and classify these variables. The study sample comprised 1102 mandibular fractures in 600 patients. For patients injured by moderate traumatic force resulting in two fractures of the mandible, the presence/absence of impacted M3s played an important role in angle/condylar fractures. Patients with impacted M3s were three times more likely to develop angle fractures and less likely to develop condylar fractures than those without impacted M3s. This study provides clinical evidence to suggest that the removal of unerupted mandibular third molars predisposes the mandible to condyle fractures.
Recent literature has led credence to the fact that the presence and state of eruption of mandibular third molars (M3s) contribute appreciably to the weakness of the angle region, thereby predisposing it to fracture . This is regardless of the fact that fractures of the mandible are usually influenced by factors such as the direction, severity and impact of force, the presence of soft tissue bulk, occlusal loading pattern and biomechanical characteristics such as bone density, mass and anatomic structures creating weak areas .
Reitzik hypothesized that, as sharp angulation concentrates stress, the angle of the mandible becomes a ‘weak’ area, and certain injuries deform the mandible beyond its yield point . The ‘weakest’ area of the dentulous mandible is the condyle and, if forces inflicted remain well absorbed at the angle, the mandible fractures at the angle, sparing the condyle.
Recently, many studies have reported a two- to threefold increased risk for mandibular angle fractures and a concomitant reduction in the incidence of condyle fractures when impacted mandibular M3s are present . Based on the current biomechanical model, it has been hypothesized that the M3 weakens the angle by decreasing the bone mass in the region, making the mandibular angle more susceptible to fracture , probably preventing another fracture at the condyle. If this model is correct, the risk of angle and condyle fractures would vary as a function of the position of the M3.
Despite the consistency of the reported relationships, between M3s, angle and condyle fractures, there are theoretical concerns that the relationship may be spurious and confounding. There has been no formal critique, but fracture risk and the influence of impacted M3 remains unclear. The purpose of this study was to clarify, the relationships between M3 presence and position and the risk for angle and condyle fracture. The authors hypothesize that: the previously observed relationship between M3 presence and subsequent angle and condyle fractures were not spurious; and different M3 positions are associated with risks of angle and condyle fracture.
Patients and methods
A retrospective cohort was designed, consisting of 600 patients who presented for treatment of mandibular fractures between January 2001 and October 2008. Hospital case records and panoramic radiographs were used to assess the presence, position and angulation of impacted M3s (predictor variables) and the incidence of mandibular angle and condyle fractures (outcome variables) were determined.
Excluded from the study were 72 patients aged 16 years and under and 68 patients for whom there was inadequate follow up. To analyze the position of M3s the Pell and Gregory Classification was used to identify horizontal position (Class I, Class II and Class III) and vertical position (Class A, Class B and Class C) ( Table 1 ). The absence of M3s was indicated by Class 0.
|Horizontal and vertical position of M3s|
|Horizontal||Amount of space available between ramus and second molar|
|Class I||Adequate space for eruption|
|Class II||Inadequate space for eruption|
|Class III||M3 located partially or completely in ramus|
|Vertical||Relationship of M3 crown to second molar crown|
|Class A||Level at occlusal plane|
|Class B||Between the cemento-enamel junction of the second molar and occlusal plane|
|Class C||Below the cemento-enamel junction of the second molar|
Angulation of M3s was measured using Shiller’s method. The inclination of the occlusal surface of the M3 was measured in relation to the occlusal surface of the second molar. The M3s were classified as: vertical if they were ±0–10°; mesioangular and distoangular if they were ±11–70°; and horizontal if they were more than ±71° The number of visible dental roots was counted from panoramic radiographs.
Mandibular angle fracture was determined using the definition given by Kelly and Harrigan: a fracture located posterior to the second molar, extending from any point on the curve formed by the junction of the body and ramus in the retromolar area to any point on the curve formed by the inferior border of the body and posterior border of the ramus of the mandible .
Condylar fracture was defined as a fracture with the fracture line superior to the sigmoid notch.
The database was constructed and analysis performed using SPSS version 10.0 (SPSS, Inc., Chicago, IL, USA). Data were analyzed by calculating the means and standard deviation, and the cohort comparisons were made using the χ 2 -test, Student’s t -test, and analysis of variance. Data were considered significant if P < 0.05.
The cohort consisted of 460 patients with 870 mandibular fractures caused by: road traffic accident (49%), assault (35%) and fall (16%). The only method for assessing severity of the trauma force was on the basis of the number of fracture sites. Low trauma force was considered to produce one fracture, a moderate trauma force produced two fractures and a high trauma force produced three or more fracture sites. The most common fracture pattern was a bi-fracture pattern (216; 47%), followed by mono-fracture (147; 32%) and multi-fracture (97; 21%). Fractures of the mandibular symphysis were observed most frequently (33%), followed by condyle (30%), angle (23%), body (11%) and ramus (2%). Distribution by site and pattern are shown in Table 2 .
|Most common fracture pattern||Sites|
|Bi-fracture (216 patients, 47%)||Symphysis/parasymphysis (287, 33%)|
|Mono fracture (147 patients, 32%)||Condyle (261, 30%)|
|Multiple fractures (97 patients, 21%)||Angle (200, 23%)|
|Body (100, 11%)|
|Ramus (22, 2%)|
Mandibular angle fractures were seen in 175 patients, with 25 patients presenting with bilateral angle fractures. Mandibular condyle fractures were seen in 187 patients, with 74 patients presenting with bilateral condyle fractures.
The cohort of 460 patients had 920 mandibular halves, of which 200 patients had no M3s (43%). Among the remaining 260 patients (57%), 520 mandibular halves, impacted mandibular M3s were present in the positions listed in Table 3 .
|Variable||Impacted M3s||Angle fracture||Condyle fracture|
|Present, N = 260||Absent, N = 200||Present, N = 175||Absent, N = 285||Present, N = 187||Absent, N = 273|
|Age||25.8 ± 9.2||32.6 ± 8.6||29.3 ± 10.8||29.7 ± 10.6||35.8 ± 10.1||30.0 ± 11.6|
|Male (345)||205||140||115 (33%)||230||103 (30%)||242|
|Female (115)||55||60||60 (52%)||55||84 (73%)||31|
The unerupted M3 present group ( n = 260 patients) sustained 442 fractures. The unerupted M3 absent group ( n = 200 patients) sustained 428 fractures. There were no significant differences between the groups for incidence of fractures. In accordance with the authors’ hypothesis, there were significant differences between both groups in the incidence of angle and condyle fractures. The unerupted M3 present group had a higher proportion of angle fracture (150 patients, 75%) than those in the unerupted M3 absent group (50 patients, 25%). Condyle fractures were more common in the unerupted M3 absent group (172 patients, 86%) than in the unerupted M3 present group (89 patients, 35%) ( Table 4 ). The frequency of fractures at the other sites did not show a significant difference between the two groups.
|No. of patients with impacted M3s||Condyle fractures *||Angle fractures †|
|Absent ( n = 200)||28 (14%)||172 (86%)||150 (75%)||50 (25%)|
|Present ( n = 260)||171 (65%)||89 (35%)||110 (42%)||150 (75%)|
Details of ramus position, occlusal position, angulation of M3s, as well as number of dental roots were determined ( Table 5 ). Concerning the ramus position of impacted M3s, the highest incidence of angle fractures was observed in Class II (53%). The highest incidence of condylar fractures was typically seen in the absence of the impacted M3, designated as Class 0 followed by Class III (27%) ( Table 6 ). Interesting correlations were observed between the two sites of fractures and M3 vertical position. The highest incidence of angle fractures was observed in Class B (41%). In keeping with the authors’ hypothesis, the highest incidence of condylar fractures was noted in the absence of M3, followed by Class C (25%) ( Table 7 ).
|Ramus position||Occlusal position||Angulation||Number of dental roots|
|Class I – 234 (45%)||Class A – 218(42%)||Mesioangular – 250 (48%)||Single – 302 (58%)|
|Two or more – 202 (39%)|
|Class II – 188 (36%)||Class B – 197 (38%)||Distoangular – 75(14%)||Tooth germ only – 15 (3%)|
|Vertical – 151 (29%)|
|Class III – 98 (19%)||Class C – 105 (20%)||Horizontal – 35(6%)|
|Germ – 9 (2%)|