Introduction: Our aims in this investigation were to assess asymmetric mandibular growth relative to skeletal maturation and to determine whether asymmetric growth occurs during a period of high growth velocity. Methods: We evaluated lateral oblique and hand-wrist radiographs of 30 male and 30 female Class I participants in the Burlington Growth Study who were assessed at 3 time periods with the skeletal maturation index (SMI). The body, the ramus, the effective length, and the gonial angle on each side of the mandible were measured. Asymmetry between the right and left sides was analyzed with the SMI and for sex dependency. Results: The left ramus was consistently longer than the right in all evaluation periods ( P <.05). The right body was consistently longer than the left in all evaluation periods ( P <.05). The effective length showed no asymmetry until the last maturation group, when the right side was longer ( P <.05). The gonial angle had no significant differences. Tests to determine differences between the sexes showed no significance in asymmetry, but the body, the ramus, and the effective length were longer in males than in females ( P <.05). Conclusions: Asymmetry does not occur or increase during any specific growth period.
The mandible develops from endochondral ossification, and growth is due to surface deposition and remodeling. Growth of the mandible can be viewed in 2 ways. If the cranium is the reference point, the chin will move down and forward. Conversely, if the mandible is isolated, the principal growth sites are the posterior surface of the ramus, and the condylar and coronoid processes. There is little change in the anterior part of the mandible or the chin. The mandible appears to grow down and forward due to growth at the condyle and the posterior surface of the ramus. The body of the mandible lengthens by periosteal apposition on its posterior surface, whereas ramal growth occurs by endochondral replacement at the condyle accompanied by surface remodeling.
In an infant, the anterior surface of the ramus is near the second deciduous molar. As the child develops, the anterior surface of the ramus remodels, creating space for the permanent molars while deposition occurs at the posterior border. Growth of the mandible not only is strongly influenced by genetic factors but also can be significantly altered by environmental factors such as nutrition, physical activity, and systemic or localized illness.
Craniofacial asymmetry is a commonly studied subject. The causes range from syndromic to normal asymmetric development. Mandibular asymmetry is also thoroughly researched, having been evaluated in syndromic patients, hyperplastic and hypoplastic mandibles, and normal development. In a cephalometric study of 200 children from the Burlington Growth Center in Toronto, Ontario, Canada, assumed to represent a normal population, Melnik demonstrated that mandibular asymmetry occurs during normal growth and development. He found that on average the effective length of the mandible was longer on the left side than the right in boys aged 6 to 9 years, but, by 16 years of age, the right side had become longer. Melnik also found a similar trend in the girls, except that the left-to-right side shift occurred earlier, by age 12. Melnik thus concluded that “there were varying degrees of mandibular asymmetry in the effective length at all ages and in both sexes” but no significant differences in the gonial angle. Generally, he found males had longer effective lengths than the females. In addition to Melnik, Tadej et al and Bishara et al also found asymmetry to be related to age and sex. Melnik’s study provokes the question of what role the lengths of the body and the ramus have on the effective length. This was the framework for our investigation.
We reviewed the literature, exploring both random and specifically targeted samples. These studies suggested that mandibular asymmetry is more common than not. Studies evaluating direction of asymmetry are conflicting. The dominant side of the mandible is determined by comparing measurements from both sides. Peck et al, Shah and Joshi, and Melnik found right-sided dominance, whereas Vig and Hewitt, Williamson and Simmons, Letzer and Kronman, and Chebib and Chamma found left-sided dominance. Conversely, Rose et al found no significant asymmetry in the evaluation of ramal and effective mandibular lengths in Angle Class I and Class II subjects. Peck et al found varying amounts of asymmetry depending on the measurement used. In general, their “data showed less asymmetry and more dimensional stability as the cranium is approached,” when moving from an inferior to a more superior position on the skull. The concept of side-dominant asymmetry was also studied in the long bones of the body, with equally conflicting results.
The purpose of this study was to assess asymmetric mandibular growth relative to skeletal maturation rather than chronological age. Because every person has a unique maturational profile, increased variability associated with chronological age will be significantly reduced. We also wanted to investigate whether asymmetric growth occurs during a period of high growth velocity. The clinical implication of this would be in coordinating timing of treatment. “Normal” asymmetry is difficult to define but perhaps can be defined by identifying factors that can cause “abnormal” asymmetry and excluding them. More than 300 syndromes are associated with craniofacial asymmetry, through the loss of normal genetic control or by chromosomal defects. Studies have shown that increased vulnerability to environmental stress due to developmental defects—eg, congenital muscular torticollis—can cause mandibular asymmetry.
Mandibular asymmetry has been suggested to be caused by trauma, degenerative joint disease, juvenile rheumatoid arthritis, and condylar hypoplasia or hyperplasia. Mouth breathing and sucking habits also have been associated with changes in the shape of the mandible. The suggestion has also been made of dental malocclusion as an etiologic factor in the development of mandibular asymmetry.
Our hypotheses were (1) male mandible measurements are greater than female measurements, (2) in both sexes, there is asymmetry between the left and right sides, (3) the degree of asymmetry is the same in both sexes, (4) asymmetry increases with maturation stage, (5) the velocity of growth for the left side of the mandible is equal to that of the right over the maturation stages, and (6) asymmetry increases during a period of high growth velocity as shown by an increase in asymmetry between groups I and II assessed with the skeletal maturation index (SMI) compared with SMI groups II and III.
Material and methods
The sample for this study was taken from the Burlington Growth Centre in Toronto, Ontario, Canada. Longitudinal records from 30 males and 30 females were randomly selected without bias. Hand-wrist and lateral oblique radiographs were evaluated. The subjects’ ages ranged from 9 to 20 years to include the 3 maturation-stage subgroups. Inclusion criteria were ANB angle less than 4° and greater than 0°, permanent dentition, and Angle Class I. No subject had been determined to be a “vertical grower” as suggested by Thompson and Popovich. There were no mouthbreathers. These subjects had no orthodontic treatment.
The skeletal maturation assessment was used in analyzing the hand-wrist radiographs. The SMI was used to determine the stage of maturation. Although numerous methods of maturation assessment are available, we used the technique developed by Fishman. The adolescent SMI stages (1-11) were categorized into 3 subgroups: group I, a period of accelerating growth velocity, included maturation levels between SMI stages 1 to 3; group II, a period of very high growth velocity, included SMI stages 4 to 7; group III, a period of decelerating growth velocity, included SMI stages 8 to 11. This grouping was found to be valid in previous studies when significant differences in growth velocity were demonstrated as a child matures. The records were used only if they included SMI stages in all 3 developmental groups. When a record occurred more than once in a subgroup, an average was taken for that record.
The radiographs were taken with the central ray directed at an angle of 45° to the midsagittal plane of the skull with exposure factors of 120 kV and 25 mA. This projection allowed for more accurate measurement of the length of the basal core of the mandible than is possible with a lateral cephalogram, because the lower border of the mandible is more parallel to the film. The measurements were not corrected for magnification because the cephalograms were taken with a standardized technique, with fixed distances from anode to subject and subject to film. Acetate tracings, measurements of the oblique radiographs, and evaluations of the hand-wrist radiographs were done by an author (J.D.).
The oblique radiographs were used to measure the body, the ramal, the effective length, and the gonial angle by using the following landmarks ( Fig 1 ).
Symphysis point (S): the midpoint of the inferior border of the mandibular symphysis.
Condylion (C): the most posterosuperior point on the head of the mandibular condyle.
Effective Length (E): the linear distance from the symphysis point to condylion.
Body length (B): the linear distance from symphysis point to the intersection of a line from the symphysis point to the antegonial notch (the point on the lower border of the mandible immediately anterior to the insertion of the masseter and medial pterygoid muscles) and a line tangent to the posterior margin of the ramus originating at condylion.
Ramal height (R): the linear distance from condylion to the intersection of a line from the symphysis point to the antegonial notch and a line tangent to the posterior margin of the ramus originating at condylion.
Gonial angle (GA): the angle formed by a line from the symphysis point and the antegonial notch to a line tangent to the posterior margin of the ramus originating at condylion.
Barber et al found that measurements from oblique radiographs varied less than 0.3 mm compared with direct skull measurements. They also found that, when compared with the standard lateral cephalogram, the severity of distortion for an oblique radiograph was less. Stellingsma et al found that, when Frankfort horizontal was altered from –20° to +20°, the image changed less than 1%.
In addition, other inherent potential errors in this study involved the tracing and location of the points. Five radiographs were retraced; the intraoperator coefficient correlation (r) was 0.9. To reduce error, all measurements are expressed in millimeters and as percentages to avoid deception by the larger male measurements. Percentages were calculated with the following formula: [(left – right)/left] × 100. A percentage was also used when analyzing the growth from 1 SMI group to another.
The Student t test for paired samples was used to test for differences between the right and left sides, and between time 1 vs time 2. The Student t test for unpaired samples was used to test the significance between the sexes. The confidence interval was 95%. Asymmetry and growth were considered significant at P <.05. The data are presented as left side minus right; a positive value indicated that the left side was longer than the right.
Asymmetry was identified in both sexes for body length ( Table I ) and ramal height ( Table II ) ( P <.05) in all SMI groups. There was asymmetry in effective length ( Table III ) only in SMI group III ( P <.05). There was no asymmetry in the gonial angle ( Table IV ). There were no sex differences, but the measurements of the males were larger than those of the females at all levels, with the exception of gonial angle ( P <.05). Although mean values were used in all calculations, there was much variation of measurements between subjects. Differences between sides ranged from 0 to 7.4 mm for body length, 0 to 6.3 mm for ramus height, 0 to 6.7 mm for effective length, and 0° to 13.8° for gonial angle.
|Body length (mm)||Group I||Group II||Group III|
|Male (± SD) (n = 30)||81.5 (3.2)||82.5 (3.4)||86.2 (3.4)||87.2 (2.5)||90.2 (3.0)||91.4 (2.7)|
|Female (± SD) (n = 30)||75.8 (3.0)||76.7 (3.2)||79.7 (3.6)||80.5 (2.0)||83.3 (4.2)||84.1 (1.7)|
|Ramal height (mm)||Group I||Group II||Group III|
|Male (± SD) (n = 30)||55.6 (3.6)||54.6 (3.8)||60.4 (3.9)||59.5 (2.2)||66.6 (4.0)||65.5 (2.3)|
|Female (± SD) (n = 30)||52.2 (3.0)||51.1 (2.9)||55.7 (3.1)||54.5 (2.4)||59.6 (3.1)||58.8 (1.8)|
|Effective length (mm)||Group I||Group II||Group III|
|Male (± SD) (n = 30)||123.0 (4.5)||122.5 (4.2)||130.2 (4.7)||130.6 (2.8)||138.1 (4.2)||139.0 (2.2)|
|Female (± SD) (n = 30)||114.4 (4.1)||114.4 (3.7)||120.5 (4.0)||120.8 (2.6)||126.3 (3.8)||127.2 (2.0)|
|Gonial angle (°)||Group I||Group II||Group III|
|Male (± SD) (n = 30)||125.7 (3.4)||125.7 (4.3)||124.5 (4.0)||124.8 (3.9)||123.0 (5.3)||124.1 (4.6)|
|Female (± SD) (n = 30)||126.1 (4.3)||126.5 (4.3)||125.3 (4.1)||126.1 (4.3)||124.0 (5.6)||125.4 (5.4)|
Figure 2 shows that the ramus was longer on the left side (+ value) in all SMI groups ( P <.05), and the body length was longer on the right side (– value) in all SMI groups ( P <.05), but effective length had a right-sided dominance and was significant only in SMI group III ( P <.05). Asymmetry was not significant between the sexes. This finding supports the hypothesis that the sex variable was eliminated with the SMI.
Asymmetry in a growing patient is often expected to increase as he or she grows. The asymmetry in this study, when evaluated in absolute millimeters, increased in each period when growth in the left components of the mandible was compared with the right. This result was expected because, as the size of the mandible increases, the difference also increases. However, when asymmetry was tested as a percentage, it did not increase or get worse but remained constant for ramal height and body length ( Figs 3 and 4 ). This could indicate that asymmetry is not the result of a cause-and-effect relationship but a normal pattern of development. With respect to effective length, however, the right side grew a greater percentage than the left side during the entire adolescent period (SMI groups I to III). Thus, effective length became more asymmetric as the subjects matured ( Fig 5 ).
The data were also analyzed to determine whether there were differences in how the components of the mandible grew in the first half of adolescence (SMI groups I to II) compared with the second half (SMI groups II to III) to assess how and when the asymmetry occurred. There were no differences in growth of any component of the mandible on either side. Our results demonstrated that asymmetric growth was consistent throughout the groups with the exception of effective length; the right side slowly, but consistently, grew more than the left. There were no changes in the results when the subjects were separated by sex. For treatment timing decisions, asymmetry does not appear to increase at any particular time during the growth period. Therefore, asymmetry will not likely respond better to treatment during any particular stage of growth. This is important from a clinicians’ standpoint in considerations for initiation or timing of treatment.
Because of inconsistencies in the results of past studies about the dominant side of growth, we suggest potential explanations for these discrepancies. The possibilities include variation in head size and head position in the cephalometer. Verhoeven et al showed that even slight head tilting can result in a large measurement error in the oblique cephalometric radiograph, so this is important to keep in mind during analysis of such radiographs. The same concept applies to lateral cephalograms as suggested by Barber et al.